{ "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": 4.266282558441162, "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": 4.47427225112915, "learning_rate": 4.23908435777872e-09, "loss": 1.0326, "mean_token_accuracy": 0.7414199113845825, "num_tokens": 79961.0, "step": 2 }, { "epoch": 0.0003816308357715303, "ewc_loss": 5.719869022868806e-13, "ewc_loss_diag": 1.8648277366750676e-17, "ewc_loss_parallel": 5.689893001203927e-16, "grad_norm": 4.357933044433594, "learning_rate": 8.47816871555744e-09, "loss": 0.9019, "mean_token_accuracy": 0.7778021097183228, "num_tokens": 118063.0, "step": 3 }, { "epoch": 0.0005088411143620405, "ewc_loss": 7.332801033044234e-12, "ewc_loss_diag": 2.1250362580715887e-16, "ewc_loss_parallel": 7.327471962526033e-15, "grad_norm": 4.466577053070068, "learning_rate": 1.271725307333616e-08, "loss": 0.9762, "mean_token_accuracy": 0.7545428276062012, "num_tokens": 149306.0, "step": 4 }, { "epoch": 0.0006360513929525506, "ewc_loss": 5.4569682106375694e-11, "ewc_loss_diag": 1.6653345369377348e-15, "ewc_loss_parallel": 5.440092820663267e-14, "grad_norm": 4.378961563110352, "learning_rate": 1.695633743111488e-08, "loss": 0.9982, "mean_token_accuracy": 0.7503748536109924, "num_tokens": 192477.0, "step": 5 }, { "epoch": 0.0007632616715430606, "ewc_loss": 3.128661774098873e-10, "ewc_loss_diag": 9.71445146547012e-15, "ewc_loss_parallel": 3.1086244689504383e-13, "grad_norm": 4.845970630645752, "learning_rate": 2.11954217888936e-08, "loss": 0.9607, "mean_token_accuracy": 0.7625337839126587, "num_tokens": 226246.0, "step": 6 }, { "epoch": 0.0008904719501335708, "ewc_loss": 1.0986695997416973e-09, "ewc_loss_diag": 4.440892098500626e-14, "ewc_loss_parallel": 1.0942358130705543e-12, "grad_norm": 5.3440656661987305, "learning_rate": 2.543450614667232e-08, "loss": 1.0125, "mean_token_accuracy": 0.7553364038467407, "num_tokens": 263593.0, "step": 7 }, { "epoch": 0.001017682228724081, "ewc_loss": 2.473825588822365e-09, "ewc_loss_diag": 7.682743330406083e-14, "ewc_loss_parallel": 2.4584778657299466e-12, "grad_norm": 4.186656951904297, "learning_rate": 2.967359050445104e-08, "loss": 0.9102, "mean_token_accuracy": 0.7727815508842468, "num_tokens": 298202.0, "step": 8 }, { "epoch": 0.001144892507314591, "ewc_loss": 3.841705620288849e-09, "ewc_loss_diag": 1.1368683772161603e-13, "ewc_loss_parallel": 3.836930773104541e-12, "grad_norm": 4.462976932525635, "learning_rate": 3.391267486222976e-08, "loss": 1.0362, "mean_token_accuracy": 0.7469140887260437, "num_tokens": 340163.0, "step": 9 }, { "epoch": 0.0012721027859051012, "ewc_loss": 1.501757651567459e-08, "ewc_loss_diag": 4.902744876744691e-13, "ewc_loss_parallel": 1.5006662579253316e-11, "grad_norm": 4.654847145080566, "learning_rate": 3.815175922000848e-08, "loss": 1.0421, "mean_token_accuracy": 0.7430824041366577, "num_tokens": 372703.0, "step": 10 }, { "epoch": 0.0013993130644956112, "ewc_loss": 2.1187588572502136e-08, "ewc_loss_diag": 6.927791673660977e-13, "ewc_loss_parallel": 2.1032064978498966e-11, "grad_norm": 4.193572044372559, "learning_rate": 4.23908435777872e-08, "loss": 0.9768, "mean_token_accuracy": 0.7515010833740234, "num_tokens": 414187.0, "step": 11 }, { "epoch": 0.0015265233430861213, "ewc_loss": 2.8405338525772095e-08, "ewc_loss_diag": 8.704148513061227e-13, "ewc_loss_parallel": 2.830802259268239e-11, "grad_norm": 4.621038913726807, "learning_rate": 4.662992793556592e-08, "loss": 1.0088, "mean_token_accuracy": 0.7500497698783875, "num_tokens": 446133.0, "step": 12 }, { "epoch": 0.0016537336216766315, "ewc_loss": 9.033828973770142e-08, "ewc_loss_diag": 3.382183422218077e-12, "ewc_loss_parallel": 9.00399754755199e-11, "grad_norm": 4.722043037414551, "learning_rate": 5.086901229334464e-08, "loss": 1.0518, "mean_token_accuracy": 0.7319566011428833, "num_tokens": 477130.0, "step": 13 }, { "epoch": 0.0017809439002671415, "ewc_loss": 1.3504177331924438e-07, "ewc_loss_diag": 4.234834705130197e-12, "ewc_loss_parallel": 1.3460521586239338e-10, "grad_norm": 4.267655372619629, "learning_rate": 5.510809665112336e-08, "loss": 1.009, "mean_token_accuracy": 0.7584123611450195, "num_tokens": 521754.0, "step": 14 }, { "epoch": 0.0019081541788576518, "ewc_loss": 1.695007085800171e-07, "ewc_loss_diag": 5.5706550483591855e-12, "ewc_loss_parallel": 1.6825651982799172e-10, "grad_norm": 4.44929313659668, "learning_rate": 5.934718100890208e-08, "loss": 0.9557, "mean_token_accuracy": 0.7663767337799072, "num_tokens": 560311.0, "step": 15 }, { "epoch": 0.002035364457448162, "ewc_loss": 2.0954757928848267e-07, "ewc_loss_diag": 6.508571459562518e-12, "ewc_loss_parallel": 2.0827428670600057e-10, "grad_norm": 4.512139797210693, "learning_rate": 6.358626536668081e-08, "loss": 1.0335, "mean_token_accuracy": 0.7409235835075378, "num_tokens": 595214.0, "step": 16 }, { "epoch": 0.002162574736038672, "ewc_loss": 4.1909515857696533e-07, "ewc_loss_diag": 1.8417267710901797e-11, "ewc_loss_parallel": 4.1654857341200113e-10, "grad_norm": 4.865448474884033, "learning_rate": 6.782534972445952e-08, "loss": 1.0434, "mean_token_accuracy": 0.7420413494110107, "num_tokens": 631427.0, "step": 17 }, { "epoch": 0.002289785014629182, "ewc_loss": 7.674098014831543e-07, "ewc_loss_diag": 2.808064891723916e-11, "ewc_loss_parallel": 7.639755494892597e-10, "grad_norm": 4.861464023590088, "learning_rate": 7.206443408223825e-08, "loss": 1.094, "mean_token_accuracy": 0.7347823977470398, "num_tokens": 671155.0, "step": 18 }, { "epoch": 0.0024169952932196924, "ewc_loss": 1.0132789611816406e-06, "ewc_loss_diag": 3.2514435588382185e-11, "ewc_loss_parallel": 1.0113581083714962e-09, "grad_norm": 4.520163059234619, "learning_rate": 7.630351844001696e-08, "loss": 0.9979, "mean_token_accuracy": 0.7530702352523804, "num_tokens": 704515.0, "step": 19 }, { "epoch": 0.0025442055718102024, "ewc_loss": 1.2069940567016602e-06, "ewc_loss_diag": 3.8198777474462986e-11, "ewc_loss_parallel": 1.2005330063402653e-09, "grad_norm": 4.482278347015381, "learning_rate": 8.054260279779568e-08, "loss": 0.9737, "mean_token_accuracy": 0.7598851919174194, "num_tokens": 745414.0, "step": 20 }, { "epoch": 0.0026714158504007124, "ewc_loss": 1.4230608940124512e-06, "ewc_loss_diag": 4.615685611497611e-11, "ewc_loss_parallel": 1.418811734765768e-09, "grad_norm": 4.605969429016113, "learning_rate": 8.47816871555744e-08, "loss": 1.0251, "mean_token_accuracy": 0.7420562505722046, "num_tokens": 779384.0, "step": 21 }, { "epoch": 0.0027986261289912225, "ewc_loss": 1.6316771507263184e-06, "ewc_loss_diag": 5.206857167650014e-11, "ewc_loss_parallel": 1.622538547962904e-09, "grad_norm": 4.450732231140137, "learning_rate": 8.902077151335312e-08, "loss": 1.0137, "mean_token_accuracy": 0.7495447993278503, "num_tokens": 812994.0, "step": 22 }, { "epoch": 0.0029258364075817325, "ewc_loss": 2.3096799850463867e-06, "ewc_loss_diag": 1.0913936421275139e-10, "ewc_loss_parallel": 2.2992026060819626e-09, "grad_norm": 4.4110493659973145, "learning_rate": 9.325985587113183e-08, "loss": 0.9537, "mean_token_accuracy": 0.7671468257904053, "num_tokens": 849351.0, "step": 23 }, { "epoch": 0.0030530466861722425, "ewc_loss": 4.559755325317383e-06, "ewc_loss_diag": 1.864464138634503e-10, "ewc_loss_parallel": 4.540197551250458e-09, "grad_norm": 4.610292434692383, "learning_rate": 9.749894022891056e-08, "loss": 0.9976, "mean_token_accuracy": 0.7482016086578369, "num_tokens": 880319.0, "step": 24 }, { "epoch": 0.003180256964762753, "ewc_loss": 6.198883056640625e-06, "ewc_loss_diag": 2.28283170145005e-10, "ewc_loss_parallel": 6.170012056827545e-09, "grad_norm": 4.56417179107666, "learning_rate": 1.0173802458668929e-07, "loss": 1.0919, "mean_token_accuracy": 0.7258732318878174, "num_tokens": 921902.0, "step": 25 }, { "epoch": 0.003307467243353263, "ewc_loss": 7.420778274536133e-06, "ewc_loss_diag": 2.546585164964199e-10, "ewc_loss_parallel": 7.392372936010361e-09, "grad_norm": 4.6700758934021, "learning_rate": 1.05977108944468e-07, "loss": 1.027, "mean_token_accuracy": 0.742982029914856, "num_tokens": 959531.0, "step": 26 }, { "epoch": 0.003434677521943773, "ewc_loss": 8.702278137207031e-06, "ewc_loss_diag": 2.801243681460619e-10, "ewc_loss_parallel": 8.672941476106644e-09, "grad_norm": 4.129568576812744, "learning_rate": 1.1021619330224672e-07, "loss": 0.9438, "mean_token_accuracy": 0.7642400860786438, "num_tokens": 997280.0, "step": 27 }, { "epoch": 0.003561887800534283, "ewc_loss": 9.655952453613281e-06, "ewc_loss_diag": 3.310560714453459e-10, "ewc_loss_parallel": 9.604264050722122e-09, "grad_norm": 4.663846969604492, "learning_rate": 1.1445527766002544e-07, "loss": 1.0816, "mean_token_accuracy": 0.7312512993812561, "num_tokens": 1034608.0, "step": 28 }, { "epoch": 0.003689098079124793, "ewc_loss": 1.0609626770019531e-05, "ewc_loss_diag": 3.6197889130562544e-10, "ewc_loss_parallel": 1.05355866253376e-08, "grad_norm": 4.240644454956055, "learning_rate": 1.1869436201780416e-07, "loss": 0.9637, "mean_token_accuracy": 0.7616682052612305, "num_tokens": 1075907.0, "step": 29 }, { "epoch": 0.0038163083577153036, "ewc_loss": 1.1980533599853516e-05, "ewc_loss_diag": 3.92901711165905e-10, "ewc_loss_parallel": 1.1932570487260818e-08, "grad_norm": 4.578054428100586, "learning_rate": 1.229334463755829e-07, "loss": 1.0136, "mean_token_accuracy": 0.7547609806060791, "num_tokens": 1114492.0, "step": 30 }, { "epoch": 0.003943518636305814, "ewc_loss": 1.2814998626708984e-05, "ewc_loss_diag": 4.256435204297304e-10, "ewc_loss_parallel": 1.2747477740049362e-08, "grad_norm": 4.27765417098999, "learning_rate": 1.2717253073336161e-07, "loss": 0.936, "mean_token_accuracy": 0.7733161449432373, "num_tokens": 1151596.0, "step": 31 }, { "epoch": 0.004070728914896324, "ewc_loss": 1.7762184143066406e-05, "ewc_loss_diag": 9.204086381942034e-10, "ewc_loss_parallel": 1.7695128917694092e-08, "grad_norm": 4.824753761291504, "learning_rate": 1.3141161509114031e-07, "loss": 1.0234, "mean_token_accuracy": 0.7515901327133179, "num_tokens": 1191475.0, "step": 32 }, { "epoch": 0.004197939193486834, "ewc_loss": 2.9802322387695312e-05, "ewc_loss_diag": 1.3533281162381172e-09, "ewc_loss_parallel": 2.9685907065868378e-08, "grad_norm": 4.296551704406738, "learning_rate": 1.3565069944891904e-07, "loss": 1.0034, "mean_token_accuracy": 0.7537984848022461, "num_tokens": 1230726.0, "step": 33 }, { "epoch": 0.004325149472077344, "ewc_loss": 3.933906555175781e-05, "ewc_loss_diag": 1.673470251262188e-09, "ewc_loss_parallel": 3.91155481338501e-08, "grad_norm": 4.19180154800415, "learning_rate": 1.3988978380669777e-07, "loss": 0.9667, "mean_token_accuracy": 0.7633091807365417, "num_tokens": 1273523.0, "step": 34 }, { "epoch": 0.004452359750667854, "ewc_loss": 4.792213439941406e-05, "ewc_loss_diag": 1.8408172763884068e-09, "ewc_loss_parallel": 4.7730281949043274e-08, "grad_norm": 4.253842353820801, "learning_rate": 1.441288681644765e-07, "loss": 0.9692, "mean_token_accuracy": 0.7627644538879395, "num_tokens": 1312013.0, "step": 35 }, { "epoch": 0.004579570029258364, "ewc_loss": 5.412101745605469e-05, "ewc_loss_diag": 1.9936123862862587e-09, "ewc_loss_parallel": 5.3783878684043884e-08, "grad_norm": 4.7169084548950195, "learning_rate": 1.4836795252225522e-07, "loss": 0.9965, "mean_token_accuracy": 0.7515853047370911, "num_tokens": 1350139.0, "step": 36 }, { "epoch": 0.004706780307848874, "ewc_loss": 5.841255187988281e-05, "ewc_loss_diag": 2.1100277081131935e-09, "ewc_loss_parallel": 5.820766091346741e-08, "grad_norm": 4.296267986297607, "learning_rate": 1.5260703688003392e-07, "loss": 0.9676, "mean_token_accuracy": 0.7583296895027161, "num_tokens": 1387670.0, "step": 37 }, { "epoch": 0.004833990586439385, "ewc_loss": 6.341934204101562e-05, "ewc_loss_diag": 2.240994945168495e-09, "ewc_loss_parallel": 6.332993507385254e-08, "grad_norm": 4.0502800941467285, "learning_rate": 1.5684612123781264e-07, "loss": 0.9724, "mean_token_accuracy": 0.758994460105896, "num_tokens": 1431658.0, "step": 38 }, { "epoch": 0.004961200865029895, "ewc_loss": 6.866455078125e-05, "ewc_loss_diag": 2.561137080192566e-09, "ewc_loss_parallel": 6.798654794692993e-08, "grad_norm": 4.296767711639404, "learning_rate": 1.6108520559559137e-07, "loss": 0.9044, "mean_token_accuracy": 0.7776188850402832, "num_tokens": 1467751.0, "step": 39 }, { "epoch": 0.005088411143620405, "ewc_loss": 7.152557373046875e-05, "ewc_loss_diag": 2.8230715543031693e-09, "ewc_loss_parallel": 7.12461769580841e-08, "grad_norm": 4.190687656402588, "learning_rate": 1.6532428995337007e-07, "loss": 1.0038, "mean_token_accuracy": 0.7465680837631226, "num_tokens": 1507078.0, "step": 40 }, { "epoch": 0.005215621422210915, "ewc_loss": 7.534027099609375e-05, "ewc_loss_diag": 2.954038791358471e-09, "ewc_loss_parallel": 7.497146725654602e-08, "grad_norm": 4.268731117248535, "learning_rate": 1.695633743111488e-07, "loss": 0.9901, "mean_token_accuracy": 0.756094753742218, "num_tokens": 1544856.0, "step": 41 }, { "epoch": 0.005342831700801425, "ewc_loss": 7.867813110351562e-05, "ewc_loss_diag": 3.157765604555607e-09, "ewc_loss_parallel": 7.82310962677002e-08, "grad_norm": 4.459619522094727, "learning_rate": 1.7380245866892752e-07, "loss": 0.9781, "mean_token_accuracy": 0.7604909539222717, "num_tokens": 1576776.0, "step": 42 }, { "epoch": 0.005470041979391935, "ewc_loss": 8.153915405273438e-05, "ewc_loss_diag": 3.346940502524376e-09, "ewc_loss_parallel": 8.102506399154663e-08, "grad_norm": 4.166018486022949, "learning_rate": 1.7804154302670624e-07, "loss": 0.9685, "mean_token_accuracy": 0.7596904039382935, "num_tokens": 1613819.0, "step": 43 }, { "epoch": 0.005597252257982445, "ewc_loss": 8.96453857421875e-05, "ewc_loss_diag": 4.132743924856186e-09, "ewc_loss_parallel": 8.89413058757782e-08, "grad_norm": 4.4805378913879395, "learning_rate": 1.8228062738448497e-07, "loss": 1.0268, "mean_token_accuracy": 0.7475086450576782, "num_tokens": 1654834.0, "step": 44 }, { "epoch": 0.005724462536572955, "ewc_loss": 0.0001430511474609375, "ewc_loss_diag": 7.799826562404633e-09, "ewc_loss_parallel": 1.424923539161682e-07, "grad_norm": 4.108485698699951, "learning_rate": 1.8651971174226367e-07, "loss": 0.9962, "mean_token_accuracy": 0.7507072687149048, "num_tokens": 1691238.0, "step": 45 }, { "epoch": 0.005851672815163465, "ewc_loss": 0.00020122528076171875, "ewc_loss_diag": 1.0011717677116394e-08, "ewc_loss_parallel": 2.0023435354232788e-07, "grad_norm": 3.997955799102783, "learning_rate": 1.907587961000424e-07, "loss": 0.884, "mean_token_accuracy": 0.772165060043335, "num_tokens": 1734446.0, "step": 46 }, { "epoch": 0.005978883093753975, "ewc_loss": 0.00025177001953125, "ewc_loss_diag": 1.1757947504520416e-08, "ewc_loss_parallel": 2.4959444999694824e-07, "grad_norm": 3.911594867706299, "learning_rate": 1.9499788045782112e-07, "loss": 0.9495, "mean_token_accuracy": 0.7589887380599976, "num_tokens": 1770266.0, "step": 47 }, { "epoch": 0.006106093372344485, "ewc_loss": 0.0002899169921875, "ewc_loss_diag": 1.30385160446167e-08, "ewc_loss_parallel": 2.8870999813079834e-07, "grad_norm": 4.026203632354736, "learning_rate": 1.9923696481559985e-07, "loss": 1.0077, "mean_token_accuracy": 0.7412751913070679, "num_tokens": 1807979.0, "step": 48 }, { "epoch": 0.006233303650934996, "ewc_loss": 0.000308990478515625, "ewc_loss_diag": 1.3853423297405243e-08, "ewc_loss_parallel": 3.073364496231079e-07, "grad_norm": 3.795013189315796, "learning_rate": 2.0347604917337857e-07, "loss": 0.898, "mean_token_accuracy": 0.7695391774177551, "num_tokens": 1851061.0, "step": 49 }, { "epoch": 0.006360513929525506, "ewc_loss": 0.000335693359375, "ewc_loss_diag": 1.4842953532934189e-08, "ewc_loss_parallel": 3.334134817123413e-07, "grad_norm": 3.545194625854492, "learning_rate": 2.0771513353115727e-07, "loss": 0.854, "mean_token_accuracy": 0.784693717956543, "num_tokens": 1887647.0, "step": 50 }, { "epoch": 0.006487724208116016, "ewc_loss": 0.000347137451171875, "ewc_loss_diag": 1.5599653124809265e-08, "ewc_loss_parallel": 3.4458935260772705e-07, "grad_norm": 4.145073413848877, "learning_rate": 2.11954217888936e-07, "loss": 1.0203, "mean_token_accuracy": 0.743596613407135, "num_tokens": 1924843.0, "step": 51 }, { "epoch": 0.006614934486706526, "ewc_loss": 0.000354766845703125, "ewc_loss_diag": 1.594889909029007e-08, "ewc_loss_parallel": 3.520399332046509e-07, "grad_norm": 4.108866214752197, "learning_rate": 2.1619330224671472e-07, "loss": 0.9467, "mean_token_accuracy": 0.7624076008796692, "num_tokens": 1967116.0, "step": 52 }, { "epoch": 0.006742144765297036, "ewc_loss": 0.0003643035888671875, "ewc_loss_diag": 1.6763806343078613e-08, "ewc_loss_parallel": 3.632158041000366e-07, "grad_norm": 3.740668535232544, "learning_rate": 2.2043238660449345e-07, "loss": 0.9283, "mean_token_accuracy": 0.7584682703018188, "num_tokens": 1999575.0, "step": 53 }, { "epoch": 0.006869355043887546, "ewc_loss": 0.0003719329833984375, "ewc_loss_diag": 1.83936208486557e-08, "ewc_loss_parallel": 3.7066638469696045e-07, "grad_norm": 4.11323881149292, "learning_rate": 2.2467147096227215e-07, "loss": 1.0159, "mean_token_accuracy": 0.7415099740028381, "num_tokens": 2039396.0, "step": 54 }, { "epoch": 0.006996565322478056, "ewc_loss": 0.0003814697265625, "ewc_loss_diag": 2.0023435354232788e-08, "ewc_loss_parallel": 3.7997961044311523e-07, "grad_norm": 4.038576602935791, "learning_rate": 2.2891055532005087e-07, "loss": 1.0476, "mean_token_accuracy": 0.7322359681129456, "num_tokens": 2080258.0, "step": 55 }, { "epoch": 0.007123775601068566, "ewc_loss": 0.000392913818359375, "ewc_loss_diag": 2.1420419216156006e-08, "ewc_loss_parallel": 3.91155481338501e-07, "grad_norm": 3.9922494888305664, "learning_rate": 2.331496396778296e-07, "loss": 1.0247, "mean_token_accuracy": 0.7412066459655762, "num_tokens": 2116552.0, "step": 56 }, { "epoch": 0.007250985879659076, "ewc_loss": 0.0003986358642578125, "ewc_loss_diag": 2.200249582529068e-08, "ewc_loss_parallel": 3.9674341678619385e-07, "grad_norm": 3.6873841285705566, "learning_rate": 2.3738872403560833e-07, "loss": 0.9226, "mean_token_accuracy": 0.7655600309371948, "num_tokens": 2161097.0, "step": 57 }, { "epoch": 0.007378196158249586, "ewc_loss": 0.000408172607421875, "ewc_loss_diag": 2.293381839990616e-08, "ewc_loss_parallel": 4.0605664253234863e-07, "grad_norm": 3.547177791595459, "learning_rate": 2.4162780839338705e-07, "loss": 0.8788, "mean_token_accuracy": 0.775458037853241, "num_tokens": 2201190.0, "step": 58 }, { "epoch": 0.007505406436840096, "ewc_loss": 0.000415802001953125, "ewc_loss_diag": 2.3981556296348572e-08, "ewc_loss_parallel": 4.1350722312927246e-07, "grad_norm": 4.0050859451293945, "learning_rate": 2.458668927511658e-07, "loss": 1.0168, "mean_token_accuracy": 0.7402915358543396, "num_tokens": 2233725.0, "step": 59 }, { "epoch": 0.007632616715430607, "ewc_loss": 0.0004253387451171875, "ewc_loss_diag": 2.5262124836444855e-08, "ewc_loss_parallel": 4.2282044887542725e-07, "grad_norm": 3.7992382049560547, "learning_rate": 2.501059771089445e-07, "loss": 0.9634, "mean_token_accuracy": 0.7512845993041992, "num_tokens": 2274459.0, "step": 60 }, { "epoch": 0.007759826994021117, "ewc_loss": 0.0004863739013671875, "ewc_loss_diag": 2.7241185307502747e-08, "ewc_loss_parallel": 4.842877388000488e-07, "grad_norm": 6.612071990966797, "learning_rate": 2.5434506146672323e-07, "loss": 0.9116, "mean_token_accuracy": 0.7659997940063477, "num_tokens": 2317530.0, "step": 61 }, { "epoch": 0.007887037272611627, "ewc_loss": 0.00078582763671875, "ewc_loss_diag": 3.632158041000366e-08, "ewc_loss_parallel": 7.82310962677002e-07, "grad_norm": 3.693325996398926, "learning_rate": 2.5858414582450195e-07, "loss": 1.0015, "mean_token_accuracy": 0.7477725744247437, "num_tokens": 2355260.0, "step": 62 }, { "epoch": 0.008014247551202136, "ewc_loss": 0.00110626220703125, "ewc_loss_diag": 5.052424967288971e-08, "ewc_loss_parallel": 1.0952353477478027e-06, "grad_norm": 3.209998369216919, "learning_rate": 2.6282323018228063e-07, "loss": 0.886, "mean_token_accuracy": 0.7711548805236816, "num_tokens": 2397135.0, "step": 63 }, { "epoch": 0.008141457829792647, "ewc_loss": 0.001251220703125, "ewc_loss_diag": 6.146728992462158e-08, "ewc_loss_parallel": 1.2442469596862793e-06, "grad_norm": 3.2107274532318115, "learning_rate": 2.6706231454005935e-07, "loss": 0.9256, "mean_token_accuracy": 0.7585434913635254, "num_tokens": 2440206.0, "step": 64 }, { "epoch": 0.008268668108383158, "ewc_loss": 0.00131988525390625, "ewc_loss_diag": 6.798654794692993e-08, "ewc_loss_parallel": 1.3113021850585938e-06, "grad_norm": 3.479689836502075, "learning_rate": 2.713013988978381e-07, "loss": 0.9634, "mean_token_accuracy": 0.7408866286277771, "num_tokens": 2469565.0, "step": 65 }, { "epoch": 0.008395878386973667, "ewc_loss": 0.00141143798828125, "ewc_loss_diag": 7.543712854385376e-08, "ewc_loss_parallel": 1.4007091522216797e-06, "grad_norm": 2.973215103149414, "learning_rate": 2.755404832556168e-07, "loss": 0.8927, "mean_token_accuracy": 0.7619765996932983, "num_tokens": 2514314.0, "step": 66 }, { "epoch": 0.008523088665564178, "ewc_loss": 0.00146484375, "ewc_loss_diag": 8.149072527885437e-08, "ewc_loss_parallel": 1.4603137969970703e-06, "grad_norm": 3.128146171569824, "learning_rate": 2.7977956761339553e-07, "loss": 0.936, "mean_token_accuracy": 0.7515219449996948, "num_tokens": 2558466.0, "step": 67 }, { "epoch": 0.008650298944154687, "ewc_loss": 0.0015411376953125, "ewc_loss_diag": 8.707866072654724e-08, "ewc_loss_parallel": 1.5348196029663086e-06, "grad_norm": 3.040696859359741, "learning_rate": 2.8401865197117426e-07, "loss": 0.8736, "mean_token_accuracy": 0.7630621194839478, "num_tokens": 2596843.0, "step": 68 }, { "epoch": 0.008777509222745198, "ewc_loss": 0.0015869140625, "ewc_loss_diag": 9.220093488693237e-08, "ewc_loss_parallel": 1.5795230865478516e-06, "grad_norm": 2.968855381011963, "learning_rate": 2.88257736328953e-07, "loss": 0.8619, "mean_token_accuracy": 0.7663230895996094, "num_tokens": 2632933.0, "step": 69 }, { "epoch": 0.008904719501335707, "ewc_loss": 0.0016021728515625, "ewc_loss_diag": 9.73232090473175e-08, "ewc_loss_parallel": 1.5944242477416992e-06, "grad_norm": 2.865215301513672, "learning_rate": 2.924968206867317e-07, "loss": 0.8435, "mean_token_accuracy": 0.7799892425537109, "num_tokens": 2673526.0, "step": 70 }, { "epoch": 0.009031929779926218, "ewc_loss": 0.00157928466796875, "ewc_loss_diag": 1.019798219203949e-07, "ewc_loss_parallel": 1.5720725059509277e-06, "grad_norm": 3.06756591796875, "learning_rate": 2.9673590504451043e-07, "loss": 0.8843, "mean_token_accuracy": 0.7645642757415771, "num_tokens": 2707770.0, "step": 71 }, { "epoch": 0.009159140058516728, "ewc_loss": 0.00156402587890625, "ewc_loss_diag": 1.043081283569336e-07, "ewc_loss_parallel": 1.55717134475708e-06, "grad_norm": 2.9014694690704346, "learning_rate": 3.009749894022891e-07, "loss": 0.8459, "mean_token_accuracy": 0.7765671014785767, "num_tokens": 2745236.0, "step": 72 }, { "epoch": 0.009286350337107238, "ewc_loss": 0.0015411376953125, "ewc_loss_diag": 1.0803341865539551e-07, "ewc_loss_parallel": 1.5348196029663086e-06, "grad_norm": 3.0180678367614746, "learning_rate": 3.0521407376006783e-07, "loss": 0.9212, "mean_token_accuracy": 0.7540191411972046, "num_tokens": 2780291.0, "step": 73 }, { "epoch": 0.009413560615697748, "ewc_loss": 0.00151824951171875, "ewc_loss_diag": 1.126900315284729e-07, "ewc_loss_parallel": 1.5124678611755371e-06, "grad_norm": 3.0584182739257812, "learning_rate": 3.0945315811784656e-07, "loss": 0.9726, "mean_token_accuracy": 0.7465370893478394, "num_tokens": 2814771.0, "step": 74 }, { "epoch": 0.009540770894288259, "ewc_loss": 0.00150299072265625, "ewc_loss_diag": 1.1641532182693481e-07, "ewc_loss_parallel": 1.4901161193847656e-06, "grad_norm": 2.803338050842285, "learning_rate": 3.136922424756253e-07, "loss": 0.8468, "mean_token_accuracy": 0.7739905714988708, "num_tokens": 2861375.0, "step": 75 }, { "epoch": 0.00966798117287877, "ewc_loss": 0.00146484375, "ewc_loss_diag": 1.2293457984924316e-07, "ewc_loss_parallel": 1.4528632164001465e-06, "grad_norm": 2.6331582069396973, "learning_rate": 3.17931326833404e-07, "loss": 0.8411, "mean_token_accuracy": 0.7701063752174377, "num_tokens": 2896139.0, "step": 76 }, { "epoch": 0.009795191451469279, "ewc_loss": 0.00144195556640625, "ewc_loss_diag": 1.2759119272232056e-07, "ewc_loss_parallel": 1.4230608940124512e-06, "grad_norm": 2.6034958362579346, "learning_rate": 3.2217041119118274e-07, "loss": 0.8003, "mean_token_accuracy": 0.7854072451591492, "num_tokens": 2940269.0, "step": 77 }, { "epoch": 0.00992240173005979, "ewc_loss": 0.0013885498046875, "ewc_loss_diag": 1.30385160446167e-07, "ewc_loss_parallel": 1.3709068298339844e-06, "grad_norm": 2.741795063018799, "learning_rate": 3.2640949554896146e-07, "loss": 0.8492, "mean_token_accuracy": 0.770720362663269, "num_tokens": 2971115.0, "step": 78 }, { "epoch": 0.010049612008650299, "ewc_loss": 0.001373291015625, "ewc_loss_diag": 1.3504177331924438e-07, "ewc_loss_parallel": 1.3560056686401367e-06, "grad_norm": 2.448208808898926, "learning_rate": 3.3064857990674013e-07, "loss": 0.7425, "mean_token_accuracy": 0.7949132919311523, "num_tokens": 3003727.0, "step": 79 }, { "epoch": 0.01017682228724081, "ewc_loss": 0.0013427734375, "ewc_loss_diag": 1.4062970876693726e-07, "ewc_loss_parallel": 1.3262033462524414e-06, "grad_norm": 2.4906623363494873, "learning_rate": 3.3488766426451886e-07, "loss": 0.7809, "mean_token_accuracy": 0.7881709337234497, "num_tokens": 3045557.0, "step": 80 }, { "epoch": 0.010304032565831319, "ewc_loss": 0.0012969970703125, "ewc_loss_diag": 1.4621764421463013e-07, "ewc_loss_parallel": 1.2814998626708984e-06, "grad_norm": 2.619741439819336, "learning_rate": 3.391267486222976e-07, "loss": 0.8505, "mean_token_accuracy": 0.7654210329055786, "num_tokens": 3080478.0, "step": 81 }, { "epoch": 0.01043124284442183, "ewc_loss": 0.00128936767578125, "ewc_loss_diag": 1.5087425708770752e-07, "ewc_loss_parallel": 1.2740492820739746e-06, "grad_norm": 2.53307843208313, "learning_rate": 3.433658329800763e-07, "loss": 0.8751, "mean_token_accuracy": 0.7624169588088989, "num_tokens": 3119205.0, "step": 82 }, { "epoch": 0.010558453123012339, "ewc_loss": 0.001312255859375, "ewc_loss_diag": 1.6111880540847778e-07, "ewc_loss_parallel": 1.296401023864746e-06, "grad_norm": 2.7737083435058594, "learning_rate": 3.4760491733785504e-07, "loss": 0.8972, "mean_token_accuracy": 0.7549859285354614, "num_tokens": 3151816.0, "step": 83 }, { "epoch": 0.01068566340160285, "ewc_loss": 0.00144195556640625, "ewc_loss_diag": 1.7229467630386353e-07, "ewc_loss_parallel": 1.4230608940124512e-06, "grad_norm": 2.7010340690612793, "learning_rate": 3.5184400169563376e-07, "loss": 0.8664, "mean_token_accuracy": 0.7615218162536621, "num_tokens": 3189765.0, "step": 84 }, { "epoch": 0.010812873680193359, "ewc_loss": 0.00176239013671875, "ewc_loss_diag": 1.8533319234848022e-07, "ewc_loss_parallel": 1.7434358596801758e-06, "grad_norm": 2.4896650314331055, "learning_rate": 3.560830860534125e-07, "loss": 0.8812, "mean_token_accuracy": 0.7594732642173767, "num_tokens": 3228988.0, "step": 85 }, { "epoch": 0.01094008395878387, "ewc_loss": 0.0022125244140625, "ewc_loss_diag": 2.0209699869155884e-07, "ewc_loss_parallel": 2.1904706954956055e-06, "grad_norm": 2.5863499641418457, "learning_rate": 3.603221704111912e-07, "loss": 0.8899, "mean_token_accuracy": 0.7572320699691772, "num_tokens": 3270648.0, "step": 86 }, { "epoch": 0.01106729423737438, "ewc_loss": 0.002655029296875, "ewc_loss_diag": 2.2724270820617676e-07, "ewc_loss_parallel": 2.637505531311035e-06, "grad_norm": 3.6081020832061768, "learning_rate": 3.6456125476896994e-07, "loss": 0.8362, "mean_token_accuracy": 0.771554708480835, "num_tokens": 3312134.0, "step": 87 }, { "epoch": 0.01119450451596489, "ewc_loss": 0.0033111572265625, "ewc_loss_diag": 2.477318048477173e-07, "ewc_loss_parallel": 3.2782554626464844e-06, "grad_norm": 2.385901927947998, "learning_rate": 3.688003391267486e-07, "loss": 0.844, "mean_token_accuracy": 0.7695115208625793, "num_tokens": 3351904.0, "step": 88 }, { "epoch": 0.0113217147945554, "ewc_loss": 0.0037384033203125, "ewc_loss_diag": 2.6263296604156494e-07, "ewc_loss_parallel": 3.7103891372680664e-06, "grad_norm": 2.1715192794799805, "learning_rate": 3.7303942348452734e-07, "loss": 0.7948, "mean_token_accuracy": 0.7802520990371704, "num_tokens": 3389610.0, "step": 89 }, { "epoch": 0.01144892507314591, "ewc_loss": 0.0040283203125, "ewc_loss_diag": 2.849847078323364e-07, "ewc_loss_parallel": 3.993511199951172e-06, "grad_norm": 2.16152286529541, "learning_rate": 3.7727850784230606e-07, "loss": 0.8683, "mean_token_accuracy": 0.7558006644248962, "num_tokens": 3425779.0, "step": 90 }, { "epoch": 0.01157613535173642, "ewc_loss": 0.00421142578125, "ewc_loss_diag": 3.03611159324646e-07, "ewc_loss_parallel": 4.172325134277344e-06, "grad_norm": 2.1535110473632812, "learning_rate": 3.815175922000848e-07, "loss": 0.8433, "mean_token_accuracy": 0.7698609828948975, "num_tokens": 3463938.0, "step": 91 }, { "epoch": 0.01170334563032693, "ewc_loss": 0.004364013671875, "ewc_loss_diag": 3.334134817123413e-07, "ewc_loss_parallel": 4.32133674621582e-06, "grad_norm": 2.0036158561706543, "learning_rate": 3.857566765578635e-07, "loss": 0.8035, "mean_token_accuracy": 0.7794984579086304, "num_tokens": 3500494.0, "step": 92 }, { "epoch": 0.01183055590891744, "ewc_loss": 0.004425048828125, "ewc_loss_diag": 3.5390257835388184e-07, "ewc_loss_parallel": 4.380941390991211e-06, "grad_norm": 1.9866023063659668, "learning_rate": 3.8999576091564224e-07, "loss": 0.7644, "mean_token_accuracy": 0.7855725288391113, "num_tokens": 3540901.0, "step": 93 }, { "epoch": 0.01195776618750795, "ewc_loss": 0.004425048828125, "ewc_loss_diag": 3.762543201446533e-07, "ewc_loss_parallel": 4.380941390991211e-06, "grad_norm": 1.8539185523986816, "learning_rate": 3.9423484527342097e-07, "loss": 0.7918, "mean_token_accuracy": 0.7764837741851807, "num_tokens": 3579275.0, "step": 94 }, { "epoch": 0.012084976466098461, "ewc_loss": 0.00439453125, "ewc_loss_diag": 3.855675458908081e-07, "ewc_loss_parallel": 4.351139068603516e-06, "grad_norm": 1.82904052734375, "learning_rate": 3.984739296311997e-07, "loss": 0.7613, "mean_token_accuracy": 0.7828728556632996, "num_tokens": 3614755.0, "step": 95 }, { "epoch": 0.01221218674468897, "ewc_loss": 0.004302978515625, "ewc_loss_diag": 4.0605664253234863e-07, "ewc_loss_parallel": 4.26173210144043e-06, "grad_norm": 1.8891286849975586, "learning_rate": 4.027130139889784e-07, "loss": 0.8155, "mean_token_accuracy": 0.7678079009056091, "num_tokens": 3655744.0, "step": 96 }, { "epoch": 0.012339397023279481, "ewc_loss": 0.004302978515625, "ewc_loss_diag": 4.302710294723511e-07, "ewc_loss_parallel": 4.26173210144043e-06, "grad_norm": 1.82692289352417, "learning_rate": 4.0695209834675715e-07, "loss": 0.7978, "mean_token_accuracy": 0.779268205165863, "num_tokens": 3697257.0, "step": 97 }, { "epoch": 0.012466607301869992, "ewc_loss": 0.004241943359375, "ewc_loss_diag": 4.5262277126312256e-07, "ewc_loss_parallel": 4.202127456665039e-06, "grad_norm": 1.7666374444961548, "learning_rate": 4.111911827045358e-07, "loss": 0.7499, "mean_token_accuracy": 0.7910207509994507, "num_tokens": 3735737.0, "step": 98 }, { "epoch": 0.012593817580460501, "ewc_loss": 0.004180908203125, "ewc_loss_diag": 4.76837158203125e-07, "ewc_loss_parallel": 4.112720489501953e-06, "grad_norm": 1.7504273653030396, "learning_rate": 4.1543026706231454e-07, "loss": 0.7632, "mean_token_accuracy": 0.7862638235092163, "num_tokens": 3774212.0, "step": 99 }, { "epoch": 0.012721027859051012, "ewc_loss": 0.004058837890625, "ewc_loss_diag": 4.880130290985107e-07, "ewc_loss_parallel": 3.993511199951172e-06, "grad_norm": 1.7739871740341187, "learning_rate": 4.1966935142009327e-07, "loss": 0.752, "mean_token_accuracy": 0.7868748903274536, "num_tokens": 3807640.0, "step": 100 }, { "epoch": 0.012848238137641521, "ewc_loss": 0.00396728515625, "ewc_loss_diag": 4.954636096954346e-07, "ewc_loss_parallel": 3.904104232788086e-06, "grad_norm": 1.7278790473937988, "learning_rate": 4.23908435777872e-07, "loss": 0.7818, "mean_token_accuracy": 0.7827883362770081, "num_tokens": 3848908.0, "step": 101 }, { "epoch": 0.012975448416232032, "ewc_loss": 0.0037384033203125, "ewc_loss_diag": 5.066394805908203e-07, "ewc_loss_parallel": 3.6954879760742188e-06, "grad_norm": 1.8054680824279785, "learning_rate": 4.281475201356507e-07, "loss": 0.7906, "mean_token_accuracy": 0.7749413251876831, "num_tokens": 3886341.0, "step": 102 }, { "epoch": 0.013102658694822541, "ewc_loss": 0.0036468505859375, "ewc_loss_diag": 5.178153514862061e-07, "ewc_loss_parallel": 3.606081008911133e-06, "grad_norm": 1.6226816177368164, "learning_rate": 4.3238660449342945e-07, "loss": 0.6675, "mean_token_accuracy": 0.8064512014389038, "num_tokens": 3917111.0, "step": 103 }, { "epoch": 0.013229868973413052, "ewc_loss": 0.0035247802734375, "ewc_loss_diag": 5.21540641784668e-07, "ewc_loss_parallel": 3.471970558166504e-06, "grad_norm": 1.671181321144104, "learning_rate": 4.3662568885120817e-07, "loss": 0.7505, "mean_token_accuracy": 0.7869089841842651, "num_tokens": 3959642.0, "step": 104 }, { "epoch": 0.013357079252003561, "ewc_loss": 0.0033721923828125, "ewc_loss_diag": 5.289912223815918e-07, "ewc_loss_parallel": 3.3229589462280273e-06, "grad_norm": 1.6309775114059448, "learning_rate": 4.408647732089869e-07, "loss": 0.7747, "mean_token_accuracy": 0.7816381454467773, "num_tokens": 4001028.0, "step": 105 }, { "epoch": 0.013484289530594072, "ewc_loss": 0.003326416015625, "ewc_loss_diag": 5.401670932769775e-07, "ewc_loss_parallel": 3.2633543014526367e-06, "grad_norm": 1.63315749168396, "learning_rate": 4.451038575667656e-07, "loss": 0.7849, "mean_token_accuracy": 0.7742131948471069, "num_tokens": 4045249.0, "step": 106 }, { "epoch": 0.013611499809184581, "ewc_loss": 0.0031585693359375, "ewc_loss_diag": 5.476176738739014e-07, "ewc_loss_parallel": 3.0994415283203125e-06, "grad_norm": 1.5888855457305908, "learning_rate": 4.493429419245443e-07, "loss": 0.7871, "mean_token_accuracy": 0.7775830030441284, "num_tokens": 4082369.0, "step": 107 }, { "epoch": 0.013738710087775092, "ewc_loss": 0.0030670166015625, "ewc_loss_diag": 5.62518835067749e-07, "ewc_loss_parallel": 3.0100345611572266e-06, "grad_norm": 1.587127447128296, "learning_rate": 4.53582026282323e-07, "loss": 0.7693, "mean_token_accuracy": 0.7780968546867371, "num_tokens": 4117448.0, "step": 108 }, { "epoch": 0.013865920366365603, "ewc_loss": 0.00299072265625, "ewc_loss_diag": 5.699694156646729e-07, "ewc_loss_parallel": 2.9355287551879883e-06, "grad_norm": 1.5014616250991821, "learning_rate": 4.5782111064010175e-07, "loss": 0.7076, "mean_token_accuracy": 0.8002164363861084, "num_tokens": 4159132.0, "step": 109 }, { "epoch": 0.013993130644956112, "ewc_loss": 0.0029296875, "ewc_loss_diag": 5.885958671569824e-07, "ewc_loss_parallel": 2.86102294921875e-06, "grad_norm": 1.5016268491744995, "learning_rate": 4.620601949978805e-07, "loss": 0.7411, "mean_token_accuracy": 0.7844639420509338, "num_tokens": 4199800.0, "step": 110 }, { "epoch": 0.014120340923546623, "ewc_loss": 0.0028839111328125, "ewc_loss_diag": 6.146728992462158e-07, "ewc_loss_parallel": 2.816319465637207e-06, "grad_norm": 1.5342025756835938, "learning_rate": 4.662992793556592e-07, "loss": 0.7464, "mean_token_accuracy": 0.791232705116272, "num_tokens": 4244513.0, "step": 111 }, { "epoch": 0.014247551202137132, "ewc_loss": 0.0028533935546875, "ewc_loss_diag": 6.221234798431396e-07, "ewc_loss_parallel": 2.7865171432495117e-06, "grad_norm": 1.4946098327636719, "learning_rate": 4.705383637134379e-07, "loss": 0.6907, "mean_token_accuracy": 0.8011561632156372, "num_tokens": 4282858.0, "step": 112 }, { "epoch": 0.014374761480727643, "ewc_loss": 0.002777099609375, "ewc_loss_diag": 6.370246410369873e-07, "ewc_loss_parallel": 2.7120113372802734e-06, "grad_norm": 1.4392473697662354, "learning_rate": 4.7477744807121665e-07, "loss": 0.741, "mean_token_accuracy": 0.7876119613647461, "num_tokens": 4325092.0, "step": 113 }, { "epoch": 0.014501971759318152, "ewc_loss": 0.002777099609375, "ewc_loss_diag": 6.556510925292969e-07, "ewc_loss_parallel": 2.7120113372802734e-06, "grad_norm": 1.4107617139816284, "learning_rate": 4.790165324289954e-07, "loss": 0.674, "mean_token_accuracy": 0.8054571151733398, "num_tokens": 4363459.0, "step": 114 }, { "epoch": 0.014629182037908663, "ewc_loss": 0.0027923583984375, "ewc_loss_diag": 6.780028343200684e-07, "ewc_loss_parallel": 2.726912498474121e-06, "grad_norm": 1.405043363571167, "learning_rate": 4.832556167867741e-07, "loss": 0.7481, "mean_token_accuracy": 0.7871387600898743, "num_tokens": 4400065.0, "step": 115 }, { "epoch": 0.014756392316499172, "ewc_loss": 0.002899169921875, "ewc_loss_diag": 7.003545761108398e-07, "ewc_loss_parallel": 2.816319465637207e-06, "grad_norm": 1.412116527557373, "learning_rate": 4.874947011445528e-07, "loss": 0.7325, "mean_token_accuracy": 0.7885193228721619, "num_tokens": 4437828.0, "step": 116 }, { "epoch": 0.014883602595089683, "ewc_loss": 0.0030059814453125, "ewc_loss_diag": 7.189810276031494e-07, "ewc_loss_parallel": 2.9355287551879883e-06, "grad_norm": 1.456479549407959, "learning_rate": 4.917337855023316e-07, "loss": 0.7015, "mean_token_accuracy": 0.7956205010414124, "num_tokens": 4475556.0, "step": 117 }, { "epoch": 0.015010812873680193, "ewc_loss": 0.0031585693359375, "ewc_loss_diag": 7.413327693939209e-07, "ewc_loss_parallel": 3.084540367126465e-06, "grad_norm": 1.4149997234344482, "learning_rate": 4.959728698601102e-07, "loss": 0.7036, "mean_token_accuracy": 0.7998701930046082, "num_tokens": 4512087.0, "step": 118 }, { "epoch": 0.015138023152270703, "ewc_loss": 0.003204345703125, "ewc_loss_diag": 7.487833499908447e-07, "ewc_loss_parallel": 3.129243850708008e-06, "grad_norm": 1.3435126543045044, "learning_rate": 5.00211954217889e-07, "loss": 0.6736, "mean_token_accuracy": 0.8025075197219849, "num_tokens": 4552116.0, "step": 119 }, { "epoch": 0.015265233430861214, "ewc_loss": 0.0032958984375, "ewc_loss_diag": 7.82310962677002e-07, "ewc_loss_parallel": 3.2186508178710938e-06, "grad_norm": 1.378007173538208, "learning_rate": 5.044510385756677e-07, "loss": 0.7195, "mean_token_accuracy": 0.7930575013160706, "num_tokens": 4591723.0, "step": 120 }, { "epoch": 0.015392443709451724, "ewc_loss": 0.003448486328125, "ewc_loss_diag": 8.009374141693115e-07, "ewc_loss_parallel": 3.3676624298095703e-06, "grad_norm": 1.2889164686203003, "learning_rate": 5.086901229334465e-07, "loss": 0.6456, "mean_token_accuracy": 0.8144766688346863, "num_tokens": 4635800.0, "step": 121 }, { "epoch": 0.015519653988042234, "ewc_loss": 0.0034942626953125, "ewc_loss_diag": 8.307397365570068e-07, "ewc_loss_parallel": 3.4123659133911133e-06, "grad_norm": 1.323401689529419, "learning_rate": 5.129292072912251e-07, "loss": 0.713, "mean_token_accuracy": 0.7945448160171509, "num_tokens": 4679908.0, "step": 122 }, { "epoch": 0.015646864266632744, "ewc_loss": 0.0035858154296875, "ewc_loss_diag": 8.493661880493164e-07, "ewc_loss_parallel": 3.5017728805541992e-06, "grad_norm": 1.3400375843048096, "learning_rate": 5.171682916490039e-07, "loss": 0.6945, "mean_token_accuracy": 0.7944124937057495, "num_tokens": 4720120.0, "step": 123 }, { "epoch": 0.015774074545223254, "ewc_loss": 0.0037689208984375, "ewc_loss_diag": 8.828938007354736e-07, "ewc_loss_parallel": 3.680586814880371e-06, "grad_norm": 1.3525785207748413, "learning_rate": 5.214073760067826e-07, "loss": 0.7171, "mean_token_accuracy": 0.7841470837593079, "num_tokens": 4758777.0, "step": 124 }, { "epoch": 0.015901284823813765, "ewc_loss": 0.0038604736328125, "ewc_loss_diag": 9.201467037200928e-07, "ewc_loss_parallel": 3.769993782043457e-06, "grad_norm": 1.3563944101333618, "learning_rate": 5.256464603645613e-07, "loss": 0.6784, "mean_token_accuracy": 0.8032070398330688, "num_tokens": 4794781.0, "step": 125 }, { "epoch": 0.016028495102404273, "ewc_loss": 0.00396728515625, "ewc_loss_diag": 9.5367431640625e-07, "ewc_loss_parallel": 3.874301910400391e-06, "grad_norm": 1.3149466514587402, "learning_rate": 5.2988554472234e-07, "loss": 0.7259, "mean_token_accuracy": 0.7875306606292725, "num_tokens": 4836391.0, "step": 126 }, { "epoch": 0.016155705380994784, "ewc_loss": 0.003997802734375, "ewc_loss_diag": 9.760260581970215e-07, "ewc_loss_parallel": 3.904104232788086e-06, "grad_norm": 1.3859994411468506, "learning_rate": 5.341246290801187e-07, "loss": 0.6981, "mean_token_accuracy": 0.7969847321510315, "num_tokens": 4874193.0, "step": 127 }, { "epoch": 0.016282915659585295, "ewc_loss": 0.00408935546875, "ewc_loss_diag": 9.98377799987793e-07, "ewc_loss_parallel": 3.993511199951172e-06, "grad_norm": 1.3586089611053467, "learning_rate": 5.383637134378975e-07, "loss": 0.7053, "mean_token_accuracy": 0.7912318706512451, "num_tokens": 4909311.0, "step": 128 }, { "epoch": 0.016410125938175806, "ewc_loss": 0.004180908203125, "ewc_loss_diag": 1.0207295417785645e-06, "ewc_loss_parallel": 4.082918167114258e-06, "grad_norm": 1.3269020318984985, "learning_rate": 5.426027977956762e-07, "loss": 0.6952, "mean_token_accuracy": 0.7989416122436523, "num_tokens": 4947409.0, "step": 129 }, { "epoch": 0.016537336216766316, "ewc_loss": 0.004241943359375, "ewc_loss_diag": 1.043081283569336e-06, "ewc_loss_parallel": 4.1425228118896484e-06, "grad_norm": 1.334097981452942, "learning_rate": 5.468418821534549e-07, "loss": 0.6984, "mean_token_accuracy": 0.7932260632514954, "num_tokens": 4982792.0, "step": 130 }, { "epoch": 0.016664546495356824, "ewc_loss": 0.004302978515625, "ewc_loss_diag": 1.0579824447631836e-06, "ewc_loss_parallel": 4.202127456665039e-06, "grad_norm": 1.299081563949585, "learning_rate": 5.510809665112336e-07, "loss": 0.6694, "mean_token_accuracy": 0.8012450933456421, "num_tokens": 5020795.0, "step": 131 }, { "epoch": 0.016791756773947335, "ewc_loss": 0.00433349609375, "ewc_loss_diag": 1.0654330253601074e-06, "ewc_loss_parallel": 4.231929779052734e-06, "grad_norm": 1.319149374961853, "learning_rate": 5.553200508690124e-07, "loss": 0.7076, "mean_token_accuracy": 0.7918691635131836, "num_tokens": 5061519.0, "step": 132 }, { "epoch": 0.016918967052537846, "ewc_loss": 0.00439453125, "ewc_loss_diag": 1.0654330253601074e-06, "ewc_loss_parallel": 4.291534423828125e-06, "grad_norm": 1.3547570705413818, "learning_rate": 5.595591352267911e-07, "loss": 0.7039, "mean_token_accuracy": 0.7904974222183228, "num_tokens": 5094608.0, "step": 133 }, { "epoch": 0.017046177331128357, "ewc_loss": 0.00445556640625, "ewc_loss_diag": 1.0952353477478027e-06, "ewc_loss_parallel": 4.32133674621582e-06, "grad_norm": 1.3454777002334595, "learning_rate": 5.637982195845697e-07, "loss": 0.7038, "mean_token_accuracy": 0.7952449321746826, "num_tokens": 5131130.0, "step": 134 }, { "epoch": 0.017173387609718864, "ewc_loss": 0.0045166015625, "ewc_loss_diag": 1.1026859283447266e-06, "ewc_loss_parallel": 4.380941390991211e-06, "grad_norm": 1.3075181245803833, "learning_rate": 5.680373039423485e-07, "loss": 0.6876, "mean_token_accuracy": 0.8002269268035889, "num_tokens": 5168921.0, "step": 135 }, { "epoch": 0.017300597888309375, "ewc_loss": 0.0045166015625, "ewc_loss_diag": 1.1175870895385742e-06, "ewc_loss_parallel": 4.380941390991211e-06, "grad_norm": 1.3284300565719604, "learning_rate": 5.722763883001272e-07, "loss": 0.6843, "mean_token_accuracy": 0.7967262864112854, "num_tokens": 5206417.0, "step": 136 }, { "epoch": 0.017427808166899886, "ewc_loss": 0.0045166015625, "ewc_loss_diag": 1.1399388313293457e-06, "ewc_loss_parallel": 4.380941390991211e-06, "grad_norm": 1.357301115989685, "learning_rate": 5.76515472657906e-07, "loss": 0.7288, "mean_token_accuracy": 0.7851212024688721, "num_tokens": 5239984.0, "step": 137 }, { "epoch": 0.017555018445490397, "ewc_loss": 0.00457763671875, "ewc_loss_diag": 1.1622905731201172e-06, "ewc_loss_parallel": 4.470348358154297e-06, "grad_norm": 1.332899808883667, "learning_rate": 5.807545570156846e-07, "loss": 0.7011, "mean_token_accuracy": 0.7960382699966431, "num_tokens": 5276444.0, "step": 138 }, { "epoch": 0.017682228724080904, "ewc_loss": 0.0045166015625, "ewc_loss_diag": 1.1846423149108887e-06, "ewc_loss_parallel": 4.380941390991211e-06, "grad_norm": 2.764127254486084, "learning_rate": 5.849936413734634e-07, "loss": 0.6307, "mean_token_accuracy": 0.8145818114280701, "num_tokens": 5312064.0, "step": 139 }, { "epoch": 0.017809439002671415, "ewc_loss": 0.005584716796875, "ewc_loss_diag": 1.2218952178955078e-06, "ewc_loss_parallel": 5.453824996948242e-06, "grad_norm": 1.4789645671844482, "learning_rate": 5.892327257312421e-07, "loss": 0.7196, "mean_token_accuracy": 0.7943037152290344, "num_tokens": 5350405.0, "step": 140 }, { "epoch": 0.017936649281261926, "ewc_loss": 0.005157470703125, "ewc_loss_diag": 1.259148120880127e-06, "ewc_loss_parallel": 5.036592483520508e-06, "grad_norm": 1.3293181657791138, "learning_rate": 5.934718100890209e-07, "loss": 0.6693, "mean_token_accuracy": 0.8050506711006165, "num_tokens": 5388338.0, "step": 141 }, { "epoch": 0.018063859559852437, "ewc_loss": 0.004547119140625, "ewc_loss_diag": 1.3113021850585938e-06, "ewc_loss_parallel": 4.410743713378906e-06, "grad_norm": 1.2591047286987305, "learning_rate": 5.977108944467995e-07, "loss": 0.6708, "mean_token_accuracy": 0.803256630897522, "num_tokens": 5432228.0, "step": 142 }, { "epoch": 0.018191069838442948, "ewc_loss": 0.004486083984375, "ewc_loss_diag": 1.341104507446289e-06, "ewc_loss_parallel": 4.351139068603516e-06, "grad_norm": 1.309996485710144, "learning_rate": 6.019499788045782e-07, "loss": 0.6493, "mean_token_accuracy": 0.8088300228118896, "num_tokens": 5469969.0, "step": 143 }, { "epoch": 0.018318280117033455, "ewc_loss": 0.00457763671875, "ewc_loss_diag": 1.3634562492370605e-06, "ewc_loss_parallel": 4.4405460357666016e-06, "grad_norm": 1.293582797050476, "learning_rate": 6.06189063162357e-07, "loss": 0.6432, "mean_token_accuracy": 0.8066266775131226, "num_tokens": 5506262.0, "step": 144 }, { "epoch": 0.018445490395623966, "ewc_loss": 0.00445556640625, "ewc_loss_diag": 1.3783574104309082e-06, "ewc_loss_parallel": 4.291534423828125e-06, "grad_norm": 1.306872010231018, "learning_rate": 6.104281475201357e-07, "loss": 0.6628, "mean_token_accuracy": 0.800946056842804, "num_tokens": 5542870.0, "step": 145 }, { "epoch": 0.018572700674214477, "ewc_loss": 0.00445556640625, "ewc_loss_diag": 1.4007091522216797e-06, "ewc_loss_parallel": 4.291534423828125e-06, "grad_norm": 1.2700806856155396, "learning_rate": 6.146672318779144e-07, "loss": 0.675, "mean_token_accuracy": 0.7980151176452637, "num_tokens": 5579958.0, "step": 146 }, { "epoch": 0.018699910952804988, "ewc_loss": 0.004425048828125, "ewc_loss_diag": 1.4081597328186035e-06, "ewc_loss_parallel": 4.26173210144043e-06, "grad_norm": 1.3449907302856445, "learning_rate": 6.189063162356931e-07, "loss": 0.6885, "mean_token_accuracy": 0.7945975661277771, "num_tokens": 5623306.0, "step": 147 }, { "epoch": 0.018827121231395495, "ewc_loss": 0.004425048828125, "ewc_loss_diag": 1.4230608940124512e-06, "ewc_loss_parallel": 4.26173210144043e-06, "grad_norm": 1.2675161361694336, "learning_rate": 6.231454005934719e-07, "loss": 0.6162, "mean_token_accuracy": 0.8182538747787476, "num_tokens": 5660801.0, "step": 148 }, { "epoch": 0.018954331509986006, "ewc_loss": 0.004364013671875, "ewc_loss_diag": 1.4454126358032227e-06, "ewc_loss_parallel": 4.202127456665039e-06, "grad_norm": 1.2783114910125732, "learning_rate": 6.273844849512506e-07, "loss": 0.6497, "mean_token_accuracy": 0.8040317893028259, "num_tokens": 5696625.0, "step": 149 }, { "epoch": 0.019081541788576517, "ewc_loss": 0.00439453125, "ewc_loss_diag": 1.4603137969970703e-06, "ewc_loss_parallel": 4.231929779052734e-06, "grad_norm": 1.2525646686553955, "learning_rate": 6.316235693090293e-07, "loss": 0.6675, "mean_token_accuracy": 0.8008696436882019, "num_tokens": 5733491.0, "step": 150 }, { "epoch": 0.019208752067167028, "ewc_loss": 0.004364013671875, "ewc_loss_diag": 1.475214958190918e-06, "ewc_loss_parallel": 4.202127456665039e-06, "grad_norm": 1.2745782136917114, "learning_rate": 6.35862653666808e-07, "loss": 0.686, "mean_token_accuracy": 0.7942519187927246, "num_tokens": 5773550.0, "step": 151 }, { "epoch": 0.01933596234575754, "ewc_loss": 0.004364013671875, "ewc_loss_diag": 1.4975666999816895e-06, "ewc_loss_parallel": 4.202127456665039e-06, "grad_norm": 1.3368291854858398, "learning_rate": 6.401017380245868e-07, "loss": 0.6797, "mean_token_accuracy": 0.7966376543045044, "num_tokens": 5813114.0, "step": 152 }, { "epoch": 0.019463172624348046, "ewc_loss": 0.004364013671875, "ewc_loss_diag": 1.5050172805786133e-06, "ewc_loss_parallel": 4.202127456665039e-06, "grad_norm": 1.3052769899368286, "learning_rate": 6.443408223823655e-07, "loss": 0.7089, "mean_token_accuracy": 0.7900581359863281, "num_tokens": 5852702.0, "step": 153 }, { "epoch": 0.019590382902938557, "ewc_loss": 0.004364013671875, "ewc_loss_diag": 1.5348196029663086e-06, "ewc_loss_parallel": 4.202127456665039e-06, "grad_norm": 1.7680290937423706, "learning_rate": 6.485799067401442e-07, "loss": 0.7775, "mean_token_accuracy": 0.770654559135437, "num_tokens": 5894481.0, "step": 154 }, { "epoch": 0.019717593181529068, "ewc_loss": 0.004608154296875, "ewc_loss_diag": 1.55717134475708e-06, "ewc_loss_parallel": 4.4405460357666016e-06, "grad_norm": 1.310477375984192, "learning_rate": 6.528189910979229e-07, "loss": 0.633, "mean_token_accuracy": 0.8055554628372192, "num_tokens": 5929501.0, "step": 155 }, { "epoch": 0.01984480346011958, "ewc_loss": 0.004364013671875, "ewc_loss_diag": 1.5795230865478516e-06, "ewc_loss_parallel": 4.202127456665039e-06, "grad_norm": 1.2447205781936646, "learning_rate": 6.570580754557017e-07, "loss": 0.6318, "mean_token_accuracy": 0.8063169717788696, "num_tokens": 5966407.0, "step": 156 }, { "epoch": 0.019972013738710086, "ewc_loss": 0.004302978515625, "ewc_loss_diag": 1.601874828338623e-06, "ewc_loss_parallel": 4.1425228118896484e-06, "grad_norm": 1.205289602279663, "learning_rate": 6.612971598134803e-07, "loss": 0.5888, "mean_token_accuracy": 0.8211476802825928, "num_tokens": 6004059.0, "step": 157 }, { "epoch": 0.020099224017300597, "ewc_loss": 0.004302978515625, "ewc_loss_diag": 1.646578311920166e-06, "ewc_loss_parallel": 4.1425228118896484e-06, "grad_norm": 1.2533055543899536, "learning_rate": 6.65536244171259e-07, "loss": 0.6449, "mean_token_accuracy": 0.8049287796020508, "num_tokens": 6040516.0, "step": 158 }, { "epoch": 0.020226434295891108, "ewc_loss": 0.00433349609375, "ewc_loss_diag": 1.6689300537109375e-06, "ewc_loss_parallel": 4.172325134277344e-06, "grad_norm": 1.2761895656585693, "learning_rate": 6.697753285290377e-07, "loss": 0.716, "mean_token_accuracy": 0.7850193977355957, "num_tokens": 6077801.0, "step": 159 }, { "epoch": 0.02035364457448162, "ewc_loss": 0.00433349609375, "ewc_loss_diag": 1.7061829566955566e-06, "ewc_loss_parallel": 4.1425228118896484e-06, "grad_norm": 1.2123867273330688, "learning_rate": 6.740144128868165e-07, "loss": 0.6326, "mean_token_accuracy": 0.8089746832847595, "num_tokens": 6114102.0, "step": 160 }, { "epoch": 0.020480854853072127, "ewc_loss": 0.0042724609375, "ewc_loss_diag": 1.7061829566955566e-06, "ewc_loss_parallel": 4.082918167114258e-06, "grad_norm": 1.2123699188232422, "learning_rate": 6.782534972445952e-07, "loss": 0.6535, "mean_token_accuracy": 0.8035264015197754, "num_tokens": 6157342.0, "step": 161 }, { "epoch": 0.020608065131662637, "ewc_loss": 0.004302978515625, "ewc_loss_diag": 1.7210841178894043e-06, "ewc_loss_parallel": 4.112720489501953e-06, "grad_norm": 1.2718714475631714, "learning_rate": 6.82492581602374e-07, "loss": 0.6554, "mean_token_accuracy": 0.8039358854293823, "num_tokens": 6193058.0, "step": 162 }, { "epoch": 0.02073527541025315, "ewc_loss": 0.00433349609375, "ewc_loss_diag": 1.7508864402770996e-06, "ewc_loss_parallel": 4.1425228118896484e-06, "grad_norm": 1.2886263132095337, "learning_rate": 6.867316659601526e-07, "loss": 0.6485, "mean_token_accuracy": 0.8079063892364502, "num_tokens": 6223616.0, "step": 163 }, { "epoch": 0.02086248568884366, "ewc_loss": 0.00433349609375, "ewc_loss_diag": 1.7657876014709473e-06, "ewc_loss_parallel": 4.1425228118896484e-06, "grad_norm": 1.2199759483337402, "learning_rate": 6.909707503179314e-07, "loss": 0.6479, "mean_token_accuracy": 0.8075760006904602, "num_tokens": 6264920.0, "step": 164 }, { "epoch": 0.02098969596743417, "ewc_loss": 0.004302978515625, "ewc_loss_diag": 1.780688762664795e-06, "ewc_loss_parallel": 4.112720489501953e-06, "grad_norm": 1.2244770526885986, "learning_rate": 6.952098346757101e-07, "loss": 0.6897, "mean_token_accuracy": 0.7920055389404297, "num_tokens": 6305270.0, "step": 165 }, { "epoch": 0.021116906246024678, "ewc_loss": 0.004302978515625, "ewc_loss_diag": 1.8030405044555664e-06, "ewc_loss_parallel": 4.112720489501953e-06, "grad_norm": 1.2283517122268677, "learning_rate": 6.994489190334889e-07, "loss": 0.5943, "mean_token_accuracy": 0.818498969078064, "num_tokens": 6341945.0, "step": 166 }, { "epoch": 0.02124411652461519, "ewc_loss": 0.00433349609375, "ewc_loss_diag": 1.8402934074401855e-06, "ewc_loss_parallel": 4.1425228118896484e-06, "grad_norm": 1.199223518371582, "learning_rate": 7.036880033912675e-07, "loss": 0.5936, "mean_token_accuracy": 0.821101725101471, "num_tokens": 6381565.0, "step": 167 }, { "epoch": 0.0213713268032057, "ewc_loss": 0.00433349609375, "ewc_loss_diag": 1.8551945686340332e-06, "ewc_loss_parallel": 4.1425228118896484e-06, "grad_norm": 1.218207597732544, "learning_rate": 7.079270877490463e-07, "loss": 0.644, "mean_token_accuracy": 0.806036114692688, "num_tokens": 6419771.0, "step": 168 }, { "epoch": 0.02149853708179621, "ewc_loss": 0.00433349609375, "ewc_loss_diag": 1.862645149230957e-06, "ewc_loss_parallel": 4.1425228118896484e-06, "grad_norm": 1.2459795475006104, "learning_rate": 7.12166172106825e-07, "loss": 0.6014, "mean_token_accuracy": 0.8170993328094482, "num_tokens": 6451413.0, "step": 169 }, { "epoch": 0.021625747360386718, "ewc_loss": 0.004364013671875, "ewc_loss_diag": 1.8775463104248047e-06, "ewc_loss_parallel": 4.172325134277344e-06, "grad_norm": 1.2226662635803223, "learning_rate": 7.164052564646038e-07, "loss": 0.6057, "mean_token_accuracy": 0.8168115615844727, "num_tokens": 6488883.0, "step": 170 }, { "epoch": 0.02175295763897723, "ewc_loss": 0.00433349609375, "ewc_loss_diag": 1.8924474716186523e-06, "ewc_loss_parallel": 4.1425228118896484e-06, "grad_norm": 1.3278456926345825, "learning_rate": 7.206443408223824e-07, "loss": 0.6575, "mean_token_accuracy": 0.801633358001709, "num_tokens": 6523669.0, "step": 171 }, { "epoch": 0.02188016791756774, "ewc_loss": 0.00439453125, "ewc_loss_diag": 1.9222497940063477e-06, "ewc_loss_parallel": 4.202127456665039e-06, "grad_norm": 1.2273736000061035, "learning_rate": 7.248834251801612e-07, "loss": 0.62, "mean_token_accuracy": 0.8095381855964661, "num_tokens": 6561795.0, "step": 172 }, { "epoch": 0.02200737819615825, "ewc_loss": 0.00433349609375, "ewc_loss_diag": 1.9371509552001953e-06, "ewc_loss_parallel": 4.1425228118896484e-06, "grad_norm": 1.2088180780410767, "learning_rate": 7.291225095379399e-07, "loss": 0.6047, "mean_token_accuracy": 0.8170306086540222, "num_tokens": 6604102.0, "step": 173 }, { "epoch": 0.02213458847474876, "ewc_loss": 0.004364013671875, "ewc_loss_diag": 1.996755599975586e-06, "ewc_loss_parallel": 4.1425228118896484e-06, "grad_norm": 1.2193529605865479, "learning_rate": 7.333615938957187e-07, "loss": 0.6894, "mean_token_accuracy": 0.7943819761276245, "num_tokens": 6645412.0, "step": 174 }, { "epoch": 0.02226179875333927, "ewc_loss": 0.004364013671875, "ewc_loss_diag": 1.996755599975586e-06, "ewc_loss_parallel": 4.1425228118896484e-06, "grad_norm": 1.2374430894851685, "learning_rate": 7.376006782534972e-07, "loss": 0.5913, "mean_token_accuracy": 0.821211576461792, "num_tokens": 6678407.0, "step": 175 }, { "epoch": 0.02238900903192978, "ewc_loss": 0.004364013671875, "ewc_loss_diag": 1.996755599975586e-06, "ewc_loss_parallel": 4.1425228118896484e-06, "grad_norm": 1.1996541023254395, "learning_rate": 7.41839762611276e-07, "loss": 0.6244, "mean_token_accuracy": 0.8094684481620789, "num_tokens": 6720968.0, "step": 176 }, { "epoch": 0.02251621931052029, "ewc_loss": 0.004364013671875, "ewc_loss_diag": 2.0116567611694336e-06, "ewc_loss_parallel": 4.1425228118896484e-06, "grad_norm": 1.2237753868103027, "learning_rate": 7.460788469690547e-07, "loss": 0.6282, "mean_token_accuracy": 0.8129236698150635, "num_tokens": 6757286.0, "step": 177 }, { "epoch": 0.0226434295891108, "ewc_loss": 0.00439453125, "ewc_loss_diag": 2.041459083557129e-06, "ewc_loss_parallel": 4.172325134277344e-06, "grad_norm": 1.271677017211914, "learning_rate": 7.503179313268335e-07, "loss": 0.6605, "mean_token_accuracy": 0.7991984486579895, "num_tokens": 6790260.0, "step": 178 }, { "epoch": 0.02277063986770131, "ewc_loss": 0.00439453125, "ewc_loss_diag": 2.0563602447509766e-06, "ewc_loss_parallel": 4.172325134277344e-06, "grad_norm": 1.2179207801818848, "learning_rate": 7.545570156846121e-07, "loss": 0.6039, "mean_token_accuracy": 0.8155422210693359, "num_tokens": 6828882.0, "step": 179 }, { "epoch": 0.02289785014629182, "ewc_loss": 0.004364013671875, "ewc_loss_diag": 2.086162567138672e-06, "ewc_loss_parallel": 4.1425228118896484e-06, "grad_norm": 1.2212440967559814, "learning_rate": 7.587961000423909e-07, "loss": 0.6196, "mean_token_accuracy": 0.8149885535240173, "num_tokens": 6867332.0, "step": 180 }, { "epoch": 0.02302506042488233, "ewc_loss": 0.00439453125, "ewc_loss_diag": 2.086162567138672e-06, "ewc_loss_parallel": 4.172325134277344e-06, "grad_norm": 1.2586162090301514, "learning_rate": 7.630351844001696e-07, "loss": 0.7327, "mean_token_accuracy": 0.7786372900009155, "num_tokens": 6914101.0, "step": 181 }, { "epoch": 0.02315227070347284, "ewc_loss": 0.00445556640625, "ewc_loss_diag": 2.1010637283325195e-06, "ewc_loss_parallel": 4.231929779052734e-06, "grad_norm": 1.2133228778839111, "learning_rate": 7.672742687579484e-07, "loss": 0.627, "mean_token_accuracy": 0.8093645572662354, "num_tokens": 6952981.0, "step": 182 }, { "epoch": 0.02327948098206335, "ewc_loss": 0.004425048828125, "ewc_loss_diag": 2.1010637283325195e-06, "ewc_loss_parallel": 4.202127456665039e-06, "grad_norm": 1.2248319387435913, "learning_rate": 7.71513353115727e-07, "loss": 0.6038, "mean_token_accuracy": 0.8139918446540833, "num_tokens": 6987534.0, "step": 183 }, { "epoch": 0.02340669126065386, "ewc_loss": 0.004425048828125, "ewc_loss_diag": 2.115964889526367e-06, "ewc_loss_parallel": 4.202127456665039e-06, "grad_norm": 1.21579909324646, "learning_rate": 7.757524374735058e-07, "loss": 0.6788, "mean_token_accuracy": 0.7949894666671753, "num_tokens": 7029343.0, "step": 184 }, { "epoch": 0.02353390153924437, "ewc_loss": 0.004425048828125, "ewc_loss_diag": 2.1457672119140625e-06, "ewc_loss_parallel": 4.202127456665039e-06, "grad_norm": 1.1975959539413452, "learning_rate": 7.799915218312845e-07, "loss": 0.6158, "mean_token_accuracy": 0.8125838041305542, "num_tokens": 7067558.0, "step": 185 }, { "epoch": 0.02366111181783488, "ewc_loss": 0.004425048828125, "ewc_loss_diag": 2.1457672119140625e-06, "ewc_loss_parallel": 4.202127456665039e-06, "grad_norm": 1.2031893730163574, "learning_rate": 7.842306061890633e-07, "loss": 0.5947, "mean_token_accuracy": 0.8136730790138245, "num_tokens": 7106164.0, "step": 186 }, { "epoch": 0.023788322096425393, "ewc_loss": 0.004425048828125, "ewc_loss_diag": 2.175569534301758e-06, "ewc_loss_parallel": 4.202127456665039e-06, "grad_norm": 1.220270037651062, "learning_rate": 7.884696905468419e-07, "loss": 0.6493, "mean_token_accuracy": 0.8043076395988464, "num_tokens": 7145549.0, "step": 187 }, { "epoch": 0.0239155323750159, "ewc_loss": 0.00445556640625, "ewc_loss_diag": 2.175569534301758e-06, "ewc_loss_parallel": 4.231929779052734e-06, "grad_norm": 1.2523797750473022, "learning_rate": 7.927087749046207e-07, "loss": 0.6149, "mean_token_accuracy": 0.8140692114830017, "num_tokens": 7179697.0, "step": 188 }, { "epoch": 0.02404274265360641, "ewc_loss": 0.0045166015625, "ewc_loss_diag": 2.175569534301758e-06, "ewc_loss_parallel": 4.291534423828125e-06, "grad_norm": 1.2272027730941772, "learning_rate": 7.969478592623994e-07, "loss": 0.6912, "mean_token_accuracy": 0.7896154522895813, "num_tokens": 7220103.0, "step": 189 }, { "epoch": 0.024169952932196922, "ewc_loss": 0.004547119140625, "ewc_loss_diag": 2.1904706954956055e-06, "ewc_loss_parallel": 4.32133674621582e-06, "grad_norm": 1.1964352130889893, "learning_rate": 8.011869436201782e-07, "loss": 0.5469, "mean_token_accuracy": 0.8346002101898193, "num_tokens": 7258059.0, "step": 190 }, { "epoch": 0.024297163210787433, "ewc_loss": 0.0045166015625, "ewc_loss_diag": 2.2202730178833008e-06, "ewc_loss_parallel": 4.291534423828125e-06, "grad_norm": 1.3013269901275635, "learning_rate": 8.054260279779568e-07, "loss": 0.6602, "mean_token_accuracy": 0.8038949966430664, "num_tokens": 7295111.0, "step": 191 }, { "epoch": 0.02442437348937794, "ewc_loss": 0.004608154296875, "ewc_loss_diag": 2.2202730178833008e-06, "ewc_loss_parallel": 4.380941390991211e-06, "grad_norm": 1.3187278509140015, "learning_rate": 8.096651123357356e-07, "loss": 0.6615, "mean_token_accuracy": 0.7991409301757812, "num_tokens": 7332703.0, "step": 192 }, { "epoch": 0.02455158376796845, "ewc_loss": 0.004608154296875, "ewc_loss_diag": 2.250075340270996e-06, "ewc_loss_parallel": 4.380941390991211e-06, "grad_norm": 1.233016848564148, "learning_rate": 8.139041966935143e-07, "loss": 0.6647, "mean_token_accuracy": 0.7978389263153076, "num_tokens": 7376446.0, "step": 193 }, { "epoch": 0.024678794046558962, "ewc_loss": 0.00457763671875, "ewc_loss_diag": 2.2649765014648438e-06, "ewc_loss_parallel": 4.351139068603516e-06, "grad_norm": 1.204512596130371, "learning_rate": 8.18143281051293e-07, "loss": 0.6066, "mean_token_accuracy": 0.8145190477371216, "num_tokens": 7416378.0, "step": 194 }, { "epoch": 0.024806004325149473, "ewc_loss": 0.0045166015625, "ewc_loss_diag": 2.2798776626586914e-06, "ewc_loss_parallel": 4.291534423828125e-06, "grad_norm": 1.2008323669433594, "learning_rate": 8.223823654090716e-07, "loss": 0.6164, "mean_token_accuracy": 0.8138020038604736, "num_tokens": 7458178.0, "step": 195 }, { "epoch": 0.024933214603739984, "ewc_loss": 0.004608154296875, "ewc_loss_diag": 2.3096799850463867e-06, "ewc_loss_parallel": 4.351139068603516e-06, "grad_norm": 1.16750168800354, "learning_rate": 8.266214497668504e-07, "loss": 0.5613, "mean_token_accuracy": 0.8298507332801819, "num_tokens": 7502436.0, "step": 196 }, { "epoch": 0.02506042488233049, "ewc_loss": 0.004547119140625, "ewc_loss_diag": 2.339482307434082e-06, "ewc_loss_parallel": 4.291534423828125e-06, "grad_norm": 1.2094029188156128, "learning_rate": 8.308605341246291e-07, "loss": 0.6196, "mean_token_accuracy": 0.8101630806922913, "num_tokens": 7539625.0, "step": 197 }, { "epoch": 0.025187635160921002, "ewc_loss": 0.004638671875, "ewc_loss_diag": 2.3692846298217773e-06, "ewc_loss_parallel": 4.380941390991211e-06, "grad_norm": 1.2404919862747192, "learning_rate": 8.350996184824079e-07, "loss": 0.5957, "mean_token_accuracy": 0.8159143924713135, "num_tokens": 7570904.0, "step": 198 }, { "epoch": 0.025314845439511513, "ewc_loss": 0.004669189453125, "ewc_loss_diag": 2.3990869522094727e-06, "ewc_loss_parallel": 4.410743713378906e-06, "grad_norm": 1.211055040359497, "learning_rate": 8.393387028401865e-07, "loss": 0.5903, "mean_token_accuracy": 0.819817841053009, "num_tokens": 7608652.0, "step": 199 }, { "epoch": 0.025442055718102024, "ewc_loss": 0.004608154296875, "ewc_loss_diag": 2.428889274597168e-06, "ewc_loss_parallel": 4.351139068603516e-06, "grad_norm": 1.2614295482635498, "learning_rate": 8.435777871979653e-07, "loss": 0.6631, "mean_token_accuracy": 0.8006998300552368, "num_tokens": 7647759.0, "step": 200 }, { "epoch": 0.02556926599669253, "ewc_loss": 0.00469970703125, "ewc_loss_diag": 2.4437904357910156e-06, "ewc_loss_parallel": 4.470348358154297e-06, "grad_norm": 1.2857210636138916, "learning_rate": 8.47816871555744e-07, "loss": 0.5849, "mean_token_accuracy": 0.8178712129592896, "num_tokens": 7678918.0, "step": 201 }, { "epoch": 0.025696476275283042, "ewc_loss": 0.0047607421875, "ewc_loss_diag": 2.4437904357910156e-06, "ewc_loss_parallel": 4.5299530029296875e-06, "grad_norm": 1.2134630680084229, "learning_rate": 8.520559559135228e-07, "loss": 0.5611, "mean_token_accuracy": 0.8268632888793945, "num_tokens": 7715797.0, "step": 202 }, { "epoch": 0.025823686553873553, "ewc_loss": 0.004608154296875, "ewc_loss_diag": 2.4586915969848633e-06, "ewc_loss_parallel": 4.351139068603516e-06, "grad_norm": 1.2034729719161987, "learning_rate": 8.562950402713014e-07, "loss": 0.6656, "mean_token_accuracy": 0.800405740737915, "num_tokens": 7763053.0, "step": 203 }, { "epoch": 0.025950896832464064, "ewc_loss": 0.004669189453125, "ewc_loss_diag": 2.4586915969848633e-06, "ewc_loss_parallel": 4.410743713378906e-06, "grad_norm": 1.2913627624511719, "learning_rate": 8.605341246290802e-07, "loss": 0.632, "mean_token_accuracy": 0.8119243383407593, "num_tokens": 7793956.0, "step": 204 }, { "epoch": 0.026078107111054575, "ewc_loss": 0.0047607421875, "ewc_loss_diag": 2.4884939193725586e-06, "ewc_loss_parallel": 4.5299530029296875e-06, "grad_norm": 1.2612214088439941, "learning_rate": 8.647732089868589e-07, "loss": 0.6099, "mean_token_accuracy": 0.8139574527740479, "num_tokens": 7833997.0, "step": 205 }, { "epoch": 0.026205317389645082, "ewc_loss": 0.004730224609375, "ewc_loss_diag": 2.5480985641479492e-06, "ewc_loss_parallel": 4.500150680541992e-06, "grad_norm": 1.2273582220077515, "learning_rate": 8.690122933446377e-07, "loss": 0.655, "mean_token_accuracy": 0.8036231398582458, "num_tokens": 7877711.0, "step": 206 }, { "epoch": 0.026332527668235593, "ewc_loss": 0.00469970703125, "ewc_loss_diag": 2.562999725341797e-06, "ewc_loss_parallel": 4.470348358154297e-06, "grad_norm": 1.2498236894607544, "learning_rate": 8.732513777024163e-07, "loss": 0.5781, "mean_token_accuracy": 0.8260310292243958, "num_tokens": 7918136.0, "step": 207 }, { "epoch": 0.026459737946826104, "ewc_loss": 0.0047607421875, "ewc_loss_diag": 2.592802047729492e-06, "ewc_loss_parallel": 4.5299530029296875e-06, "grad_norm": 1.2201430797576904, "learning_rate": 8.774904620601951e-07, "loss": 0.62, "mean_token_accuracy": 0.8136269450187683, "num_tokens": 7960847.0, "step": 208 }, { "epoch": 0.026586948225416615, "ewc_loss": 0.004791259765625, "ewc_loss_diag": 2.60770320892334e-06, "ewc_loss_parallel": 4.5299530029296875e-06, "grad_norm": 1.2355912923812866, "learning_rate": 8.817295464179738e-07, "loss": 0.5811, "mean_token_accuracy": 0.8190047740936279, "num_tokens": 7996567.0, "step": 209 }, { "epoch": 0.026714158504007122, "ewc_loss": 0.004791259765625, "ewc_loss_diag": 2.6226043701171875e-06, "ewc_loss_parallel": 4.5299530029296875e-06, "grad_norm": 1.2513692378997803, "learning_rate": 8.859686307757526e-07, "loss": 0.6048, "mean_token_accuracy": 0.8154153227806091, "num_tokens": 8032441.0, "step": 210 }, { "epoch": 0.026841368782597633, "ewc_loss": 0.00482177734375, "ewc_loss_diag": 2.637505531311035e-06, "ewc_loss_parallel": 4.559755325317383e-06, "grad_norm": 1.2535388469696045, "learning_rate": 8.902077151335312e-07, "loss": 0.6391, "mean_token_accuracy": 0.8051663637161255, "num_tokens": 8072020.0, "step": 211 }, { "epoch": 0.026968579061188144, "ewc_loss": 0.00482177734375, "ewc_loss_diag": 2.6673078536987305e-06, "ewc_loss_parallel": 4.559755325317383e-06, "grad_norm": 1.2969355583190918, "learning_rate": 8.944467994913099e-07, "loss": 0.6842, "mean_token_accuracy": 0.7940149307250977, "num_tokens": 8105736.0, "step": 212 }, { "epoch": 0.027095789339778655, "ewc_loss": 0.0048828125, "ewc_loss_diag": 2.682209014892578e-06, "ewc_loss_parallel": 4.6193599700927734e-06, "grad_norm": 1.2407461404800415, "learning_rate": 8.986858838490886e-07, "loss": 0.5343, "mean_token_accuracy": 0.8311353325843811, "num_tokens": 8139021.0, "step": 213 }, { "epoch": 0.027222999618369163, "ewc_loss": 0.00482177734375, "ewc_loss_diag": 2.6971101760864258e-06, "ewc_loss_parallel": 4.559755325317383e-06, "grad_norm": 1.2169238328933716, "learning_rate": 9.029249682068674e-07, "loss": 0.6066, "mean_token_accuracy": 0.8126785755157471, "num_tokens": 8182361.0, "step": 214 }, { "epoch": 0.027350209896959674, "ewc_loss": 0.00482177734375, "ewc_loss_diag": 2.726912498474121e-06, "ewc_loss_parallel": 4.559755325317383e-06, "grad_norm": 1.2586517333984375, "learning_rate": 9.07164052564646e-07, "loss": 0.6214, "mean_token_accuracy": 0.8095966577529907, "num_tokens": 8218131.0, "step": 215 }, { "epoch": 0.027477420175550184, "ewc_loss": 0.0048828125, "ewc_loss_diag": 2.7418136596679688e-06, "ewc_loss_parallel": 4.6193599700927734e-06, "grad_norm": 1.2555952072143555, "learning_rate": 9.114031369224248e-07, "loss": 0.5974, "mean_token_accuracy": 0.8213463425636292, "num_tokens": 8255886.0, "step": 216 }, { "epoch": 0.027604630454140695, "ewc_loss": 0.0048828125, "ewc_loss_diag": 2.7418136596679688e-06, "ewc_loss_parallel": 4.6193599700927734e-06, "grad_norm": 1.2127572298049927, "learning_rate": 9.156422212802035e-07, "loss": 0.5536, "mean_token_accuracy": 0.8281166553497314, "num_tokens": 8297196.0, "step": 217 }, { "epoch": 0.027731840732731206, "ewc_loss": 0.0048828125, "ewc_loss_diag": 2.7567148208618164e-06, "ewc_loss_parallel": 4.6193599700927734e-06, "grad_norm": 1.2591145038604736, "learning_rate": 9.198813056379823e-07, "loss": 0.5956, "mean_token_accuracy": 0.8173816800117493, "num_tokens": 8331857.0, "step": 218 }, { "epoch": 0.027859051011321714, "ewc_loss": 0.00494384765625, "ewc_loss_diag": 2.771615982055664e-06, "ewc_loss_parallel": 4.678964614868164e-06, "grad_norm": 1.2362474203109741, "learning_rate": 9.24120389995761e-07, "loss": 0.5904, "mean_token_accuracy": 0.8177626132965088, "num_tokens": 8371120.0, "step": 219 }, { "epoch": 0.027986261289912225, "ewc_loss": 0.004913330078125, "ewc_loss_diag": 2.8014183044433594e-06, "ewc_loss_parallel": 4.649162292480469e-06, "grad_norm": 1.2943487167358398, "learning_rate": 9.283594743535397e-07, "loss": 0.7128, "mean_token_accuracy": 0.7848273515701294, "num_tokens": 8407502.0, "step": 220 }, { "epoch": 0.028113471568502735, "ewc_loss": 0.004974365234375, "ewc_loss_diag": 2.8014183044433594e-06, "ewc_loss_parallel": 4.708766937255859e-06, "grad_norm": 1.2491698265075684, "learning_rate": 9.325985587113184e-07, "loss": 0.5638, "mean_token_accuracy": 0.8273640275001526, "num_tokens": 8444360.0, "step": 221 }, { "epoch": 0.028240681847093246, "ewc_loss": 0.004913330078125, "ewc_loss_diag": 2.8312206268310547e-06, "ewc_loss_parallel": 4.649162292480469e-06, "grad_norm": 1.2603949308395386, "learning_rate": 9.368376430690972e-07, "loss": 0.5758, "mean_token_accuracy": 0.8234947919845581, "num_tokens": 8476749.0, "step": 222 }, { "epoch": 0.028367892125683754, "ewc_loss": 0.0050048828125, "ewc_loss_diag": 2.8461217880249023e-06, "ewc_loss_parallel": 4.738569259643555e-06, "grad_norm": 1.2555853128433228, "learning_rate": 9.410767274268759e-07, "loss": 0.6059, "mean_token_accuracy": 0.8167203664779663, "num_tokens": 8517157.0, "step": 223 }, { "epoch": 0.028495102404274265, "ewc_loss": 0.0050048828125, "ewc_loss_diag": 2.8908252716064453e-06, "ewc_loss_parallel": 4.708766937255859e-06, "grad_norm": 1.2507786750793457, "learning_rate": 9.453158117846546e-07, "loss": 0.5776, "mean_token_accuracy": 0.8227342367172241, "num_tokens": 8556677.0, "step": 224 }, { "epoch": 0.028622312682864776, "ewc_loss": 0.0050048828125, "ewc_loss_diag": 2.8908252716064453e-06, "ewc_loss_parallel": 4.708766937255859e-06, "grad_norm": 1.2538186311721802, "learning_rate": 9.495548961424333e-07, "loss": 0.6123, "mean_token_accuracy": 0.8170680999755859, "num_tokens": 8594335.0, "step": 225 }, { "epoch": 0.028749522961455286, "ewc_loss": 0.005035400390625, "ewc_loss_diag": 2.9355287551879883e-06, "ewc_loss_parallel": 4.738569259643555e-06, "grad_norm": 1.2496106624603271, "learning_rate": 9.53793980500212e-07, "loss": 0.5328, "mean_token_accuracy": 0.8350958824157715, "num_tokens": 8626259.0, "step": 226 }, { "epoch": 0.028876733240045797, "ewc_loss": 0.00506591796875, "ewc_loss_diag": 2.9653310775756836e-06, "ewc_loss_parallel": 4.76837158203125e-06, "grad_norm": 1.2725321054458618, "learning_rate": 9.580330648579909e-07, "loss": 0.6953, "mean_token_accuracy": 0.7930349111557007, "num_tokens": 8665481.0, "step": 227 }, { "epoch": 0.029003943518636305, "ewc_loss": 0.00506591796875, "ewc_loss_diag": 2.995133399963379e-06, "ewc_loss_parallel": 4.76837158203125e-06, "grad_norm": 1.272605538368225, "learning_rate": 9.622721492157695e-07, "loss": 0.5603, "mean_token_accuracy": 0.828978955745697, "num_tokens": 8703136.0, "step": 228 }, { "epoch": 0.029131153797226816, "ewc_loss": 0.005096435546875, "ewc_loss_diag": 2.995133399963379e-06, "ewc_loss_parallel": 4.798173904418945e-06, "grad_norm": 1.2903887033462524, "learning_rate": 9.665112335735482e-07, "loss": 0.5753, "mean_token_accuracy": 0.8212007880210876, "num_tokens": 8734989.0, "step": 229 }, { "epoch": 0.029258364075817327, "ewc_loss": 0.005096435546875, "ewc_loss_diag": 3.0249357223510742e-06, "ewc_loss_parallel": 4.798173904418945e-06, "grad_norm": 1.2593733072280884, "learning_rate": 9.707503179313269e-07, "loss": 0.5398, "mean_token_accuracy": 0.8318878412246704, "num_tokens": 8768145.0, "step": 230 }, { "epoch": 0.029385574354407838, "ewc_loss": 0.005096435546875, "ewc_loss_diag": 3.0547380447387695e-06, "ewc_loss_parallel": 4.798173904418945e-06, "grad_norm": 1.2687602043151855, "learning_rate": 9.749894022891056e-07, "loss": 0.5934, "mean_token_accuracy": 0.8185303211212158, "num_tokens": 8804668.0, "step": 231 }, { "epoch": 0.029512784632998345, "ewc_loss": 0.005126953125, "ewc_loss_diag": 3.0547380447387695e-06, "ewc_loss_parallel": 4.827976226806641e-06, "grad_norm": 1.2492215633392334, "learning_rate": 9.792284866468842e-07, "loss": 0.6155, "mean_token_accuracy": 0.8071575164794922, "num_tokens": 8848343.0, "step": 232 }, { "epoch": 0.029639994911588856, "ewc_loss": 0.00506591796875, "ewc_loss_diag": 3.0547380447387695e-06, "ewc_loss_parallel": 4.76837158203125e-06, "grad_norm": 1.259220838546753, "learning_rate": 9.834675710046631e-07, "loss": 0.5807, "mean_token_accuracy": 0.8178436756134033, "num_tokens": 8883794.0, "step": 233 }, { "epoch": 0.029767205190179367, "ewc_loss": 0.005126953125, "ewc_loss_diag": 3.069639205932617e-06, "ewc_loss_parallel": 4.827976226806641e-06, "grad_norm": 1.2661789655685425, "learning_rate": 9.877066553624418e-07, "loss": 0.6011, "mean_token_accuracy": 0.8149930238723755, "num_tokens": 8921597.0, "step": 234 }, { "epoch": 0.029894415468769878, "ewc_loss": 0.005157470703125, "ewc_loss_diag": 3.084540367126465e-06, "ewc_loss_parallel": 4.857778549194336e-06, "grad_norm": 1.3260279893875122, "learning_rate": 9.919457397202205e-07, "loss": 0.6228, "mean_token_accuracy": 0.8102706670761108, "num_tokens": 8956260.0, "step": 235 }, { "epoch": 0.030021625747360385, "ewc_loss": 0.005218505859375, "ewc_loss_diag": 3.11434268951416e-06, "ewc_loss_parallel": 4.9173831939697266e-06, "grad_norm": 1.3382568359375, "learning_rate": 9.961848240779991e-07, "loss": 0.6941, "mean_token_accuracy": 0.7930988669395447, "num_tokens": 8992492.0, "step": 236 }, { "epoch": 0.030148836025950896, "ewc_loss": 0.0052490234375, "ewc_loss_diag": 3.129243850708008e-06, "ewc_loss_parallel": 4.947185516357422e-06, "grad_norm": 1.2813031673431396, "learning_rate": 1.000423908435778e-06, "loss": 0.5984, "mean_token_accuracy": 0.8172202706336975, "num_tokens": 9028300.0, "step": 237 }, { "epoch": 0.030276046304541407, "ewc_loss": 0.00518798828125, "ewc_loss_diag": 3.1441450119018555e-06, "ewc_loss_parallel": 4.887580871582031e-06, "grad_norm": 1.2560011148452759, "learning_rate": 1.0046629927935567e-06, "loss": 0.5916, "mean_token_accuracy": 0.8192002177238464, "num_tokens": 9066620.0, "step": 238 }, { "epoch": 0.030403256583131918, "ewc_loss": 0.00518798828125, "ewc_loss_diag": 3.159046173095703e-06, "ewc_loss_parallel": 4.887580871582031e-06, "grad_norm": 1.2718881368637085, "learning_rate": 1.0089020771513354e-06, "loss": 0.5364, "mean_token_accuracy": 0.8361213803291321, "num_tokens": 9105213.0, "step": 239 }, { "epoch": 0.03053046686172243, "ewc_loss": 0.005218505859375, "ewc_loss_diag": 3.1888484954833984e-06, "ewc_loss_parallel": 4.9173831939697266e-06, "grad_norm": 1.299623966217041, "learning_rate": 1.013141161509114e-06, "loss": 0.5968, "mean_token_accuracy": 0.8170572519302368, "num_tokens": 9139673.0, "step": 240 }, { "epoch": 0.030657677140312936, "ewc_loss": 0.005279541015625, "ewc_loss_diag": 3.2186508178710938e-06, "ewc_loss_parallel": 4.947185516357422e-06, "grad_norm": 1.2894986867904663, "learning_rate": 1.017380245866893e-06, "loss": 0.5991, "mean_token_accuracy": 0.8132562637329102, "num_tokens": 9172435.0, "step": 241 }, { "epoch": 0.030784887418903447, "ewc_loss": 0.005279541015625, "ewc_loss_diag": 3.2335519790649414e-06, "ewc_loss_parallel": 4.947185516357422e-06, "grad_norm": 1.305808186531067, "learning_rate": 1.0216193302246716e-06, "loss": 0.6236, "mean_token_accuracy": 0.8035202026367188, "num_tokens": 9206015.0, "step": 242 }, { "epoch": 0.030912097697493958, "ewc_loss": 0.005340576171875, "ewc_loss_diag": 3.2633543014526367e-06, "ewc_loss_parallel": 5.0067901611328125e-06, "grad_norm": 1.2491307258605957, "learning_rate": 1.0258584145824503e-06, "loss": 0.5584, "mean_token_accuracy": 0.826631486415863, "num_tokens": 9247224.0, "step": 243 }, { "epoch": 0.03103930797608447, "ewc_loss": 0.005279541015625, "ewc_loss_diag": 3.293156623840332e-06, "ewc_loss_parallel": 4.947185516357422e-06, "grad_norm": 1.2665472030639648, "learning_rate": 1.030097498940229e-06, "loss": 0.5442, "mean_token_accuracy": 0.8319539427757263, "num_tokens": 9282171.0, "step": 244 }, { "epoch": 0.031166518254674976, "ewc_loss": 0.005340576171875, "ewc_loss_diag": 3.337860107421875e-06, "ewc_loss_parallel": 5.0067901611328125e-06, "grad_norm": 1.2753702402114868, "learning_rate": 1.0343365832980078e-06, "loss": 0.5957, "mean_token_accuracy": 0.8163831830024719, "num_tokens": 9324078.0, "step": 245 }, { "epoch": 0.03129372853326549, "ewc_loss": 0.00537109375, "ewc_loss_diag": 3.3676624298095703e-06, "ewc_loss_parallel": 5.036592483520508e-06, "grad_norm": 1.274093508720398, "learning_rate": 1.0385756676557865e-06, "loss": 0.5913, "mean_token_accuracy": 0.818448543548584, "num_tokens": 9364516.0, "step": 246 }, { "epoch": 0.031420938811855995, "ewc_loss": 0.00537109375, "ewc_loss_diag": 3.382563591003418e-06, "ewc_loss_parallel": 5.036592483520508e-06, "grad_norm": 1.287397027015686, "learning_rate": 1.0428147520135652e-06, "loss": 0.5691, "mean_token_accuracy": 0.8244961500167847, "num_tokens": 9403804.0, "step": 247 }, { "epoch": 0.03154814909044651, "ewc_loss": 0.005401611328125, "ewc_loss_diag": 3.4123659133911133e-06, "ewc_loss_parallel": 5.066394805908203e-06, "grad_norm": 1.2896530628204346, "learning_rate": 1.0470538363713438e-06, "loss": 0.6112, "mean_token_accuracy": 0.8114200830459595, "num_tokens": 9441589.0, "step": 248 }, { "epoch": 0.031675359369037016, "ewc_loss": 0.005401611328125, "ewc_loss_diag": 3.427267074584961e-06, "ewc_loss_parallel": 5.066394805908203e-06, "grad_norm": 1.2937930822372437, "learning_rate": 1.0512929207291225e-06, "loss": 0.612, "mean_token_accuracy": 0.8107078671455383, "num_tokens": 9480240.0, "step": 249 }, { "epoch": 0.03180256964762753, "ewc_loss": 0.005401611328125, "ewc_loss_diag": 3.4570693969726562e-06, "ewc_loss_parallel": 5.066394805908203e-06, "grad_norm": 1.3607654571533203, "learning_rate": 1.0555320050869012e-06, "loss": 0.5744, "mean_token_accuracy": 0.8193553686141968, "num_tokens": 9512704.0, "step": 250 }, { "epoch": 0.03192977992621804, "ewc_loss": 0.005462646484375, "ewc_loss_diag": 3.4570693969726562e-06, "ewc_loss_parallel": 5.125999450683594e-06, "grad_norm": 1.3442820310592651, "learning_rate": 1.05977108944468e-06, "loss": 0.6296, "mean_token_accuracy": 0.8031957149505615, "num_tokens": 9545057.0, "step": 251 }, { "epoch": 0.032056990204808546, "ewc_loss": 0.00543212890625, "ewc_loss_diag": 3.471970558166504e-06, "ewc_loss_parallel": 5.0961971282958984e-06, "grad_norm": 1.2881147861480713, "learning_rate": 1.0640101738024587e-06, "loss": 0.5711, "mean_token_accuracy": 0.8246780633926392, "num_tokens": 9585797.0, "step": 252 }, { "epoch": 0.03218420048339906, "ewc_loss": 0.005401611328125, "ewc_loss_diag": 3.4868717193603516e-06, "ewc_loss_parallel": 5.066394805908203e-06, "grad_norm": 1.3011146783828735, "learning_rate": 1.0682492581602374e-06, "loss": 0.5322, "mean_token_accuracy": 0.8357502222061157, "num_tokens": 9617890.0, "step": 253 }, { "epoch": 0.03231141076198957, "ewc_loss": 0.005462646484375, "ewc_loss_diag": 3.4868717193603516e-06, "ewc_loss_parallel": 5.125999450683594e-06, "grad_norm": 1.2947101593017578, "learning_rate": 1.072488342518016e-06, "loss": 0.5313, "mean_token_accuracy": 0.8322662115097046, "num_tokens": 9652895.0, "step": 254 }, { "epoch": 0.03243862104058008, "ewc_loss": 0.005401611328125, "ewc_loss_diag": 3.4868717193603516e-06, "ewc_loss_parallel": 5.066394805908203e-06, "grad_norm": 1.2965774536132812, "learning_rate": 1.076727426875795e-06, "loss": 0.5925, "mean_token_accuracy": 0.8161677122116089, "num_tokens": 9691804.0, "step": 255 }, { "epoch": 0.03256583131917059, "ewc_loss": 0.00543212890625, "ewc_loss_diag": 3.5017728805541992e-06, "ewc_loss_parallel": 5.0961971282958984e-06, "grad_norm": 1.2912466526031494, "learning_rate": 1.0809665112335736e-06, "loss": 0.6177, "mean_token_accuracy": 0.8119515180587769, "num_tokens": 9736178.0, "step": 256 }, { "epoch": 0.0326930415977611, "ewc_loss": 0.0054931640625, "ewc_loss_diag": 3.546476364135742e-06, "ewc_loss_parallel": 5.125999450683594e-06, "grad_norm": 1.274632453918457, "learning_rate": 1.0852055955913523e-06, "loss": 0.5079, "mean_token_accuracy": 0.8413954973220825, "num_tokens": 9773867.0, "step": 257 }, { "epoch": 0.03282025187635161, "ewc_loss": 0.00543212890625, "ewc_loss_diag": 3.546476364135742e-06, "ewc_loss_parallel": 5.066394805908203e-06, "grad_norm": 1.291248083114624, "learning_rate": 1.089444679949131e-06, "loss": 0.6158, "mean_token_accuracy": 0.8106715679168701, "num_tokens": 9813764.0, "step": 258 }, { "epoch": 0.03294746215494212, "ewc_loss": 0.005523681640625, "ewc_loss_diag": 3.606081008911133e-06, "ewc_loss_parallel": 5.155801773071289e-06, "grad_norm": 1.285904884338379, "learning_rate": 1.0936837643069099e-06, "loss": 0.5693, "mean_token_accuracy": 0.8228795528411865, "num_tokens": 9855447.0, "step": 259 }, { "epoch": 0.03307467243353263, "ewc_loss": 0.0054931640625, "ewc_loss_diag": 3.606081008911133e-06, "ewc_loss_parallel": 5.125999450683594e-06, "grad_norm": 1.3048350811004639, "learning_rate": 1.0979228486646885e-06, "loss": 0.5299, "mean_token_accuracy": 0.8374948501586914, "num_tokens": 9889806.0, "step": 260 }, { "epoch": 0.03320188271212314, "ewc_loss": 0.00555419921875, "ewc_loss_diag": 3.6209821701049805e-06, "ewc_loss_parallel": 5.185604095458984e-06, "grad_norm": 1.3229531049728394, "learning_rate": 1.1021619330224672e-06, "loss": 0.6715, "mean_token_accuracy": 0.7940137386322021, "num_tokens": 9929257.0, "step": 261 }, { "epoch": 0.03332909299071365, "ewc_loss": 0.005584716796875, "ewc_loss_diag": 3.6507844924926758e-06, "ewc_loss_parallel": 5.21540641784668e-06, "grad_norm": 1.3081004619598389, "learning_rate": 1.1064010173802459e-06, "loss": 0.6036, "mean_token_accuracy": 0.8131611347198486, "num_tokens": 9967822.0, "step": 262 }, { "epoch": 0.03345630326930416, "ewc_loss": 0.00555419921875, "ewc_loss_diag": 3.6507844924926758e-06, "ewc_loss_parallel": 5.185604095458984e-06, "grad_norm": 1.296783208847046, "learning_rate": 1.1106401017380248e-06, "loss": 0.5333, "mean_token_accuracy": 0.8338041305541992, "num_tokens": 10004873.0, "step": 263 }, { "epoch": 0.03358351354789467, "ewc_loss": 0.00555419921875, "ewc_loss_diag": 3.680586814880371e-06, "ewc_loss_parallel": 5.185604095458984e-06, "grad_norm": 1.3656046390533447, "learning_rate": 1.1148791860958034e-06, "loss": 0.6151, "mean_token_accuracy": 0.8067987561225891, "num_tokens": 10033761.0, "step": 264 }, { "epoch": 0.03371072382648518, "ewc_loss": 0.005645751953125, "ewc_loss_diag": 3.680586814880371e-06, "ewc_loss_parallel": 5.27501106262207e-06, "grad_norm": 1.3173247575759888, "learning_rate": 1.1191182704535821e-06, "loss": 0.6021, "mean_token_accuracy": 0.8141771554946899, "num_tokens": 10072586.0, "step": 265 }, { "epoch": 0.03383793410507569, "ewc_loss": 0.005584716796875, "ewc_loss_diag": 3.7103891372680664e-06, "ewc_loss_parallel": 5.21540641784668e-06, "grad_norm": 1.2957404851913452, "learning_rate": 1.123357354811361e-06, "loss": 0.5952, "mean_token_accuracy": 0.8170045018196106, "num_tokens": 10112210.0, "step": 266 }, { "epoch": 0.0339651443836662, "ewc_loss": 0.005615234375, "ewc_loss_diag": 3.7103891372680664e-06, "ewc_loss_parallel": 5.245208740234375e-06, "grad_norm": 1.2925786972045898, "learning_rate": 1.1275964391691395e-06, "loss": 0.578, "mean_token_accuracy": 0.8233952522277832, "num_tokens": 10155908.0, "step": 267 }, { "epoch": 0.03409235466225671, "ewc_loss": 0.005645751953125, "ewc_loss_diag": 3.7997961044311523e-06, "ewc_loss_parallel": 5.27501106262207e-06, "grad_norm": 1.3259981870651245, "learning_rate": 1.1318355235269181e-06, "loss": 0.6293, "mean_token_accuracy": 0.8050726056098938, "num_tokens": 10195944.0, "step": 268 }, { "epoch": 0.03421956494084722, "ewc_loss": 0.00567626953125, "ewc_loss_diag": 3.7997961044311523e-06, "ewc_loss_parallel": 5.304813385009766e-06, "grad_norm": 1.3208428621292114, "learning_rate": 1.136074607884697e-06, "loss": 0.616, "mean_token_accuracy": 0.8100935220718384, "num_tokens": 10233602.0, "step": 269 }, { "epoch": 0.03434677521943773, "ewc_loss": 0.00567626953125, "ewc_loss_diag": 3.814697265625e-06, "ewc_loss_parallel": 5.304813385009766e-06, "grad_norm": 1.3377519845962524, "learning_rate": 1.1403136922424757e-06, "loss": 0.5701, "mean_token_accuracy": 0.8222012519836426, "num_tokens": 10267001.0, "step": 270 }, { "epoch": 0.03447398549802824, "ewc_loss": 0.0057373046875, "ewc_loss_diag": 3.844499588012695e-06, "ewc_loss_parallel": 5.334615707397461e-06, "grad_norm": 1.305284857749939, "learning_rate": 1.1445527766002544e-06, "loss": 0.5789, "mean_token_accuracy": 0.8187037110328674, "num_tokens": 10308264.0, "step": 271 }, { "epoch": 0.03460119577661875, "ewc_loss": 0.0057373046875, "ewc_loss_diag": 3.874301910400391e-06, "ewc_loss_parallel": 5.334615707397461e-06, "grad_norm": 1.374164342880249, "learning_rate": 1.148791860958033e-06, "loss": 0.6801, "mean_token_accuracy": 0.7922770977020264, "num_tokens": 10343440.0, "step": 272 }, { "epoch": 0.034728406055209264, "ewc_loss": 0.005828857421875, "ewc_loss_diag": 3.904104232788086e-06, "ewc_loss_parallel": 5.424022674560547e-06, "grad_norm": 1.3390041589736938, "learning_rate": 1.153030945315812e-06, "loss": 0.6073, "mean_token_accuracy": 0.8146287798881531, "num_tokens": 10383402.0, "step": 273 }, { "epoch": 0.03485561633379977, "ewc_loss": 0.0057373046875, "ewc_loss_diag": 3.904104232788086e-06, "ewc_loss_parallel": 5.334615707397461e-06, "grad_norm": 1.3137433528900146, "learning_rate": 1.1572700296735906e-06, "loss": 0.5828, "mean_token_accuracy": 0.817167341709137, "num_tokens": 10421700.0, "step": 274 }, { "epoch": 0.03498282661239028, "ewc_loss": 0.0057373046875, "ewc_loss_diag": 3.904104232788086e-06, "ewc_loss_parallel": 5.334615707397461e-06, "grad_norm": 1.3525837659835815, "learning_rate": 1.1615091140313693e-06, "loss": 0.6229, "mean_token_accuracy": 0.80995774269104, "num_tokens": 10460295.0, "step": 275 }, { "epoch": 0.03511003689098079, "ewc_loss": 0.005828857421875, "ewc_loss_diag": 3.904104232788086e-06, "ewc_loss_parallel": 5.424022674560547e-06, "grad_norm": 1.3569003343582153, "learning_rate": 1.165748198389148e-06, "loss": 0.5705, "mean_token_accuracy": 0.8234367370605469, "num_tokens": 10501619.0, "step": 276 }, { "epoch": 0.0352372471695713, "ewc_loss": 0.005859375, "ewc_loss_diag": 3.933906555175781e-06, "ewc_loss_parallel": 5.453824996948242e-06, "grad_norm": 1.3291430473327637, "learning_rate": 1.1699872827469268e-06, "loss": 0.5739, "mean_token_accuracy": 0.8230764865875244, "num_tokens": 10540465.0, "step": 277 }, { "epoch": 0.03536445744816181, "ewc_loss": 0.005828857421875, "ewc_loss_diag": 3.933906555175781e-06, "ewc_loss_parallel": 5.424022674560547e-06, "grad_norm": 1.306778073310852, "learning_rate": 1.1742263671047055e-06, "loss": 0.5807, "mean_token_accuracy": 0.8198155164718628, "num_tokens": 10580980.0, "step": 278 }, { "epoch": 0.03549166772675232, "ewc_loss": 0.005828857421875, "ewc_loss_diag": 3.9637088775634766e-06, "ewc_loss_parallel": 5.424022674560547e-06, "grad_norm": 1.3557838201522827, "learning_rate": 1.1784654514624842e-06, "loss": 0.6532, "mean_token_accuracy": 0.7979149222373962, "num_tokens": 10621332.0, "step": 279 }, { "epoch": 0.03561887800534283, "ewc_loss": 0.005859375, "ewc_loss_diag": 3.9637088775634766e-06, "ewc_loss_parallel": 5.453824996948242e-06, "grad_norm": 1.3739221096038818, "learning_rate": 1.1827045358202629e-06, "loss": 0.6068, "mean_token_accuracy": 0.8170582056045532, "num_tokens": 10654468.0, "step": 280 }, { "epoch": 0.035746088283933344, "ewc_loss": 0.005859375, "ewc_loss_diag": 3.9637088775634766e-06, "ewc_loss_parallel": 5.453824996948242e-06, "grad_norm": 1.3205957412719727, "learning_rate": 1.1869436201780417e-06, "loss": 0.5268, "mean_token_accuracy": 0.8365564346313477, "num_tokens": 10692191.0, "step": 281 }, { "epoch": 0.03587329856252385, "ewc_loss": 0.00592041015625, "ewc_loss_diag": 3.9637088775634766e-06, "ewc_loss_parallel": 5.513429641723633e-06, "grad_norm": 1.3239719867706299, "learning_rate": 1.1911827045358204e-06, "loss": 0.5669, "mean_token_accuracy": 0.8288506269454956, "num_tokens": 10731568.0, "step": 282 }, { "epoch": 0.03600050884111436, "ewc_loss": 0.005950927734375, "ewc_loss_diag": 3.9637088775634766e-06, "ewc_loss_parallel": 5.543231964111328e-06, "grad_norm": 1.3520405292510986, "learning_rate": 1.195421788893599e-06, "loss": 0.6393, "mean_token_accuracy": 0.807439386844635, "num_tokens": 10772166.0, "step": 283 }, { "epoch": 0.036127719119704874, "ewc_loss": 0.00592041015625, "ewc_loss_diag": 3.993511199951172e-06, "ewc_loss_parallel": 5.513429641723633e-06, "grad_norm": 1.343524694442749, "learning_rate": 1.199660873251378e-06, "loss": 0.5808, "mean_token_accuracy": 0.8209748268127441, "num_tokens": 10807527.0, "step": 284 }, { "epoch": 0.03625492939829538, "ewc_loss": 0.00592041015625, "ewc_loss_diag": 3.993511199951172e-06, "ewc_loss_parallel": 5.513429641723633e-06, "grad_norm": 1.3336336612701416, "learning_rate": 1.2038999576091564e-06, "loss": 0.5547, "mean_token_accuracy": 0.8283910155296326, "num_tokens": 10846064.0, "step": 285 }, { "epoch": 0.036382139676885895, "ewc_loss": 0.0059814453125, "ewc_loss_diag": 4.023313522338867e-06, "ewc_loss_parallel": 5.5730342864990234e-06, "grad_norm": 1.3310965299606323, "learning_rate": 1.208139041966935e-06, "loss": 0.5312, "mean_token_accuracy": 0.8378642797470093, "num_tokens": 10884496.0, "step": 286 }, { "epoch": 0.0365093499554764, "ewc_loss": 0.005950927734375, "ewc_loss_diag": 4.023313522338867e-06, "ewc_loss_parallel": 5.543231964111328e-06, "grad_norm": 1.3733835220336914, "learning_rate": 1.212378126324714e-06, "loss": 0.5754, "mean_token_accuracy": 0.8191268444061279, "num_tokens": 10917542.0, "step": 287 }, { "epoch": 0.03663656023406691, "ewc_loss": 0.006072998046875, "ewc_loss_diag": 4.082918167114258e-06, "ewc_loss_parallel": 5.662441253662109e-06, "grad_norm": 1.4004294872283936, "learning_rate": 1.2166172106824927e-06, "loss": 0.6406, "mean_token_accuracy": 0.8087962865829468, "num_tokens": 10949596.0, "step": 288 }, { "epoch": 0.036763770512657425, "ewc_loss": 0.006134033203125, "ewc_loss_diag": 4.082918167114258e-06, "ewc_loss_parallel": 5.7220458984375e-06, "grad_norm": 1.342920184135437, "learning_rate": 1.2208562950402713e-06, "loss": 0.5449, "mean_token_accuracy": 0.828179657459259, "num_tokens": 10986143.0, "step": 289 }, { "epoch": 0.03689098079124793, "ewc_loss": 0.0059814453125, "ewc_loss_diag": 4.082918167114258e-06, "ewc_loss_parallel": 5.5730342864990234e-06, "grad_norm": 1.3754628896713257, "learning_rate": 1.22509537939805e-06, "loss": 0.584, "mean_token_accuracy": 0.8167932033538818, "num_tokens": 11024678.0, "step": 290 }, { "epoch": 0.03701819106983844, "ewc_loss": 0.006103515625, "ewc_loss_diag": 4.082918167114258e-06, "ewc_loss_parallel": 5.692243576049805e-06, "grad_norm": 1.3368101119995117, "learning_rate": 1.2293344637558289e-06, "loss": 0.6004, "mean_token_accuracy": 0.8143924474716187, "num_tokens": 11068950.0, "step": 291 }, { "epoch": 0.037145401348428954, "ewc_loss": 0.00604248046875, "ewc_loss_diag": 4.082918167114258e-06, "ewc_loss_parallel": 5.632638931274414e-06, "grad_norm": 1.3543281555175781, "learning_rate": 1.2335735481136076e-06, "loss": 0.5493, "mean_token_accuracy": 0.8256491422653198, "num_tokens": 11108697.0, "step": 292 }, { "epoch": 0.03727261162701946, "ewc_loss": 0.006134033203125, "ewc_loss_diag": 4.082918167114258e-06, "ewc_loss_parallel": 5.751848220825195e-06, "grad_norm": 1.356231451034546, "learning_rate": 1.2378126324713862e-06, "loss": 0.5676, "mean_token_accuracy": 0.8277815580368042, "num_tokens": 11148482.0, "step": 293 }, { "epoch": 0.037399821905609976, "ewc_loss": 0.006134033203125, "ewc_loss_diag": 4.082918167114258e-06, "ewc_loss_parallel": 5.751848220825195e-06, "grad_norm": 1.3431068658828735, "learning_rate": 1.242051716829165e-06, "loss": 0.5833, "mean_token_accuracy": 0.8132500648498535, "num_tokens": 11189139.0, "step": 294 }, { "epoch": 0.03752703218420048, "ewc_loss": 0.006134033203125, "ewc_loss_diag": 4.082918167114258e-06, "ewc_loss_parallel": 5.751848220825195e-06, "grad_norm": 1.3517677783966064, "learning_rate": 1.2462908011869438e-06, "loss": 0.5683, "mean_token_accuracy": 0.8215941786766052, "num_tokens": 11226484.0, "step": 295 }, { "epoch": 0.03765424246279099, "ewc_loss": 0.006134033203125, "ewc_loss_diag": 4.082918167114258e-06, "ewc_loss_parallel": 5.751848220825195e-06, "grad_norm": 1.3677830696105957, "learning_rate": 1.2505298855447223e-06, "loss": 0.5216, "mean_token_accuracy": 0.8364061117172241, "num_tokens": 11263806.0, "step": 296 }, { "epoch": 0.037781452741381505, "ewc_loss": 0.006195068359375, "ewc_loss_diag": 4.082918167114258e-06, "ewc_loss_parallel": 5.811452865600586e-06, "grad_norm": 1.390837550163269, "learning_rate": 1.2547689699025011e-06, "loss": 0.6464, "mean_token_accuracy": 0.802544891834259, "num_tokens": 11299822.0, "step": 297 }, { "epoch": 0.03790866301997201, "ewc_loss": 0.0062255859375, "ewc_loss_diag": 4.112720489501953e-06, "ewc_loss_parallel": 5.811452865600586e-06, "grad_norm": 1.4161646366119385, "learning_rate": 1.2590080542602798e-06, "loss": 0.606, "mean_token_accuracy": 0.8117498755455017, "num_tokens": 11330796.0, "step": 298 }, { "epoch": 0.03803587329856253, "ewc_loss": 0.00628662109375, "ewc_loss_diag": 4.112720489501953e-06, "ewc_loss_parallel": 5.8710575103759766e-06, "grad_norm": 1.3673253059387207, "learning_rate": 1.2632471386180587e-06, "loss": 0.5975, "mean_token_accuracy": 0.8145877122879028, "num_tokens": 11369529.0, "step": 299 }, { "epoch": 0.038163083577153034, "ewc_loss": 0.006256103515625, "ewc_loss_diag": 4.1425228118896484e-06, "ewc_loss_parallel": 5.841255187988281e-06, "grad_norm": 1.3942638635635376, "learning_rate": 1.2674862229758372e-06, "loss": 0.5699, "mean_token_accuracy": 0.8297548890113831, "num_tokens": 11405044.0, "step": 300 }, { "epoch": 0.03829029385574354, "ewc_loss": 0.00628662109375, "ewc_loss_diag": 4.1425228118896484e-06, "ewc_loss_parallel": 5.8710575103759766e-06, "grad_norm": 1.368496060371399, "learning_rate": 1.271725307333616e-06, "loss": 0.5649, "mean_token_accuracy": 0.8259751796722412, "num_tokens": 11445409.0, "step": 301 }, { "epoch": 0.038417504134334056, "ewc_loss": 0.00628662109375, "ewc_loss_diag": 4.172325134277344e-06, "ewc_loss_parallel": 5.8710575103759766e-06, "grad_norm": 1.3334380388259888, "learning_rate": 1.2759643916913947e-06, "loss": 0.5594, "mean_token_accuracy": 0.8246922492980957, "num_tokens": 11488285.0, "step": 302 }, { "epoch": 0.03854471441292456, "ewc_loss": 0.00628662109375, "ewc_loss_diag": 4.202127456665039e-06, "ewc_loss_parallel": 5.8710575103759766e-06, "grad_norm": 1.3979418277740479, "learning_rate": 1.2802034760491736e-06, "loss": 0.5672, "mean_token_accuracy": 0.823566198348999, "num_tokens": 11520970.0, "step": 303 }, { "epoch": 0.03867192469151508, "ewc_loss": 0.006317138671875, "ewc_loss_diag": 4.202127456665039e-06, "ewc_loss_parallel": 5.900859832763672e-06, "grad_norm": 1.3697179555892944, "learning_rate": 1.284442560406952e-06, "loss": 0.6259, "mean_token_accuracy": 0.8051064014434814, "num_tokens": 11563590.0, "step": 304 }, { "epoch": 0.038799134970105585, "ewc_loss": 0.00628662109375, "ewc_loss_diag": 4.202127456665039e-06, "ewc_loss_parallel": 5.8710575103759766e-06, "grad_norm": 1.4216276407241821, "learning_rate": 1.288681644764731e-06, "loss": 0.6652, "mean_token_accuracy": 0.7929937839508057, "num_tokens": 11598629.0, "step": 305 }, { "epoch": 0.03892634524869609, "ewc_loss": 0.00640869140625, "ewc_loss_diag": 4.231929779052734e-06, "ewc_loss_parallel": 5.990266799926758e-06, "grad_norm": 1.3900359869003296, "learning_rate": 1.2929207291225096e-06, "loss": 0.5634, "mean_token_accuracy": 0.8274639248847961, "num_tokens": 11635526.0, "step": 306 }, { "epoch": 0.03905355552728661, "ewc_loss": 0.00634765625, "ewc_loss_diag": 4.291534423828125e-06, "ewc_loss_parallel": 5.930662155151367e-06, "grad_norm": 1.3943136930465698, "learning_rate": 1.2971598134802885e-06, "loss": 0.5827, "mean_token_accuracy": 0.8202666640281677, "num_tokens": 11672974.0, "step": 307 }, { "epoch": 0.039180765805877114, "ewc_loss": 0.00640869140625, "ewc_loss_diag": 4.351139068603516e-06, "ewc_loss_parallel": 5.990266799926758e-06, "grad_norm": 1.4118006229400635, "learning_rate": 1.301398897838067e-06, "loss": 0.6028, "mean_token_accuracy": 0.814939022064209, "num_tokens": 11708886.0, "step": 308 }, { "epoch": 0.03930797608446762, "ewc_loss": 0.00640869140625, "ewc_loss_diag": 4.351139068603516e-06, "ewc_loss_parallel": 5.990266799926758e-06, "grad_norm": 1.3757282495498657, "learning_rate": 1.3056379821958458e-06, "loss": 0.6128, "mean_token_accuracy": 0.8156818151473999, "num_tokens": 11749217.0, "step": 309 }, { "epoch": 0.039435186363058136, "ewc_loss": 0.006378173828125, "ewc_loss_diag": 4.351139068603516e-06, "ewc_loss_parallel": 5.9604644775390625e-06, "grad_norm": 1.3847562074661255, "learning_rate": 1.3098770665536245e-06, "loss": 0.574, "mean_token_accuracy": 0.8236973285675049, "num_tokens": 11788380.0, "step": 310 }, { "epoch": 0.039562396641648644, "ewc_loss": 0.00640869140625, "ewc_loss_diag": 4.380941390991211e-06, "ewc_loss_parallel": 5.990266799926758e-06, "grad_norm": 1.4104201793670654, "learning_rate": 1.3141161509114034e-06, "loss": 0.5455, "mean_token_accuracy": 0.8291829824447632, "num_tokens": 11820587.0, "step": 311 }, { "epoch": 0.03968960692023916, "ewc_loss": 0.0064697265625, "ewc_loss_diag": 4.380941390991211e-06, "ewc_loss_parallel": 6.0498714447021484e-06, "grad_norm": 1.3945358991622925, "learning_rate": 1.3183552352691819e-06, "loss": 0.5518, "mean_token_accuracy": 0.8293718099594116, "num_tokens": 11854473.0, "step": 312 }, { "epoch": 0.039816817198829665, "ewc_loss": 0.0064697265625, "ewc_loss_diag": 4.410743713378906e-06, "ewc_loss_parallel": 6.0498714447021484e-06, "grad_norm": 1.3730820417404175, "learning_rate": 1.3225943196269605e-06, "loss": 0.59, "mean_token_accuracy": 0.8214778900146484, "num_tokens": 11894744.0, "step": 313 }, { "epoch": 0.03994402747742017, "ewc_loss": 0.0064697265625, "ewc_loss_diag": 4.410743713378906e-06, "ewc_loss_parallel": 6.0498714447021484e-06, "grad_norm": 1.3789632320404053, "learning_rate": 1.3268334039847394e-06, "loss": 0.5518, "mean_token_accuracy": 0.8290811777114868, "num_tokens": 11932346.0, "step": 314 }, { "epoch": 0.04007123775601069, "ewc_loss": 0.006500244140625, "ewc_loss_diag": 4.4405460357666016e-06, "ewc_loss_parallel": 6.0498714447021484e-06, "grad_norm": 1.4299005270004272, "learning_rate": 1.331072488342518e-06, "loss": 0.6098, "mean_token_accuracy": 0.808502197265625, "num_tokens": 11967683.0, "step": 315 }, { "epoch": 0.040198448034601195, "ewc_loss": 0.006591796875, "ewc_loss_diag": 4.4405460357666016e-06, "ewc_loss_parallel": 6.139278411865234e-06, "grad_norm": 1.3840382099151611, "learning_rate": 1.335311572700297e-06, "loss": 0.5963, "mean_token_accuracy": 0.8140596747398376, "num_tokens": 12010521.0, "step": 316 }, { "epoch": 0.04032565831319171, "ewc_loss": 0.006500244140625, "ewc_loss_diag": 4.470348358154297e-06, "ewc_loss_parallel": 6.0498714447021484e-06, "grad_norm": 1.3759355545043945, "learning_rate": 1.3395506570580754e-06, "loss": 0.5669, "mean_token_accuracy": 0.8239150047302246, "num_tokens": 12053317.0, "step": 317 }, { "epoch": 0.040452868591782216, "ewc_loss": 0.006622314453125, "ewc_loss_diag": 4.500150680541992e-06, "ewc_loss_parallel": 6.16908073425293e-06, "grad_norm": 1.4880201816558838, "learning_rate": 1.3437897414158543e-06, "loss": 0.6248, "mean_token_accuracy": 0.8068469762802124, "num_tokens": 12081589.0, "step": 318 }, { "epoch": 0.040580078870372724, "ewc_loss": 0.0068359375, "ewc_loss_diag": 4.500150680541992e-06, "ewc_loss_parallel": 6.377696990966797e-06, "grad_norm": 1.4305775165557861, "learning_rate": 1.348028825773633e-06, "loss": 0.5655, "mean_token_accuracy": 0.8220285177230835, "num_tokens": 12114826.0, "step": 319 }, { "epoch": 0.04070728914896324, "ewc_loss": 0.006622314453125, "ewc_loss_diag": 4.500150680541992e-06, "ewc_loss_parallel": 6.16908073425293e-06, "grad_norm": 1.3952935934066772, "learning_rate": 1.3522679101314119e-06, "loss": 0.5457, "mean_token_accuracy": 0.8296536207199097, "num_tokens": 12150769.0, "step": 320 }, { "epoch": 0.040834499427553746, "ewc_loss": 0.006744384765625, "ewc_loss_diag": 4.559755325317383e-06, "ewc_loss_parallel": 6.288290023803711e-06, "grad_norm": 1.402727484703064, "learning_rate": 1.3565069944891903e-06, "loss": 0.5677, "mean_token_accuracy": 0.825538158416748, "num_tokens": 12189035.0, "step": 321 }, { "epoch": 0.04096170970614425, "ewc_loss": 0.006683349609375, "ewc_loss_diag": 4.559755325317383e-06, "ewc_loss_parallel": 6.22868537902832e-06, "grad_norm": 1.4109116792678833, "learning_rate": 1.3607460788469692e-06, "loss": 0.558, "mean_token_accuracy": 0.8267139196395874, "num_tokens": 12225974.0, "step": 322 }, { "epoch": 0.04108891998473477, "ewc_loss": 0.006805419921875, "ewc_loss_diag": 4.559755325317383e-06, "ewc_loss_parallel": 6.3478946685791016e-06, "grad_norm": 1.4104312658309937, "learning_rate": 1.364985163204748e-06, "loss": 0.6264, "mean_token_accuracy": 0.8077008724212646, "num_tokens": 12266971.0, "step": 323 }, { "epoch": 0.041216130263325275, "ewc_loss": 0.006805419921875, "ewc_loss_diag": 4.649162292480469e-06, "ewc_loss_parallel": 6.3478946685791016e-06, "grad_norm": 1.4113397598266602, "learning_rate": 1.3692242475625268e-06, "loss": 0.5366, "mean_token_accuracy": 0.8387369513511658, "num_tokens": 12306365.0, "step": 324 }, { "epoch": 0.04134334054191579, "ewc_loss": 0.0068359375, "ewc_loss_diag": 4.649162292480469e-06, "ewc_loss_parallel": 6.377696990966797e-06, "grad_norm": 1.4089475870132446, "learning_rate": 1.3734633319203052e-06, "loss": 0.592, "mean_token_accuracy": 0.8259141445159912, "num_tokens": 12345343.0, "step": 325 }, { "epoch": 0.0414705508205063, "ewc_loss": 0.0068359375, "ewc_loss_diag": 4.678964614868164e-06, "ewc_loss_parallel": 6.377696990966797e-06, "grad_norm": 1.4674572944641113, "learning_rate": 1.3777024162780841e-06, "loss": 0.6571, "mean_token_accuracy": 0.7979215383529663, "num_tokens": 12384532.0, "step": 326 }, { "epoch": 0.041597761099096804, "ewc_loss": 0.006988525390625, "ewc_loss_diag": 4.678964614868164e-06, "ewc_loss_parallel": 6.5267086029052734e-06, "grad_norm": 1.4083845615386963, "learning_rate": 1.3819415006358628e-06, "loss": 0.5207, "mean_token_accuracy": 0.8376708030700684, "num_tokens": 12424955.0, "step": 327 }, { "epoch": 0.04172497137768732, "ewc_loss": 0.006866455078125, "ewc_loss_diag": 4.708766937255859e-06, "ewc_loss_parallel": 6.407499313354492e-06, "grad_norm": 1.3724631071090698, "learning_rate": 1.3861805849936417e-06, "loss": 0.5416, "mean_token_accuracy": 0.8347150683403015, "num_tokens": 12469583.0, "step": 328 }, { "epoch": 0.041852181656277826, "ewc_loss": 0.0069580078125, "ewc_loss_diag": 4.738569259643555e-06, "ewc_loss_parallel": 6.496906280517578e-06, "grad_norm": 1.4394944906234741, "learning_rate": 1.3904196693514201e-06, "loss": 0.587, "mean_token_accuracy": 0.8152598142623901, "num_tokens": 12504997.0, "step": 329 }, { "epoch": 0.04197939193486834, "ewc_loss": 0.00701904296875, "ewc_loss_diag": 4.738569259643555e-06, "ewc_loss_parallel": 6.556510925292969e-06, "grad_norm": 1.416439414024353, "learning_rate": 1.394658753709199e-06, "loss": 0.5886, "mean_token_accuracy": 0.8202507495880127, "num_tokens": 12546248.0, "step": 330 }, { "epoch": 0.04210660221345885, "ewc_loss": 0.00701904296875, "ewc_loss_diag": 4.798173904418945e-06, "ewc_loss_parallel": 6.5267086029052734e-06, "grad_norm": 1.4488530158996582, "learning_rate": 1.3988978380669777e-06, "loss": 0.6078, "mean_token_accuracy": 0.8120110034942627, "num_tokens": 12580741.0, "step": 331 }, { "epoch": 0.042233812492049355, "ewc_loss": 0.00714111328125, "ewc_loss_diag": 4.857778549194336e-06, "ewc_loss_parallel": 6.645917892456055e-06, "grad_norm": 1.4178828001022339, "learning_rate": 1.4031369224247562e-06, "loss": 0.6132, "mean_token_accuracy": 0.8120264410972595, "num_tokens": 12622539.0, "step": 332 }, { "epoch": 0.04236102277063987, "ewc_loss": 0.00701904296875, "ewc_loss_diag": 4.857778549194336e-06, "ewc_loss_parallel": 6.5267086029052734e-06, "grad_norm": 1.4352256059646606, "learning_rate": 1.407376006782535e-06, "loss": 0.5684, "mean_token_accuracy": 0.8234984874725342, "num_tokens": 12660137.0, "step": 333 }, { "epoch": 0.04248823304923038, "ewc_loss": 0.007232666015625, "ewc_loss_diag": 4.887580871582031e-06, "ewc_loss_parallel": 6.735324859619141e-06, "grad_norm": 1.4211387634277344, "learning_rate": 1.4116150911403137e-06, "loss": 0.5943, "mean_token_accuracy": 0.8172375559806824, "num_tokens": 12702341.0, "step": 334 }, { "epoch": 0.04261544332782089, "ewc_loss": 0.00714111328125, "ewc_loss_diag": 4.9173831939697266e-06, "ewc_loss_parallel": 6.645917892456055e-06, "grad_norm": 1.418991208076477, "learning_rate": 1.4158541754980926e-06, "loss": 0.5362, "mean_token_accuracy": 0.8348268270492554, "num_tokens": 12739560.0, "step": 335 }, { "epoch": 0.0427426536064114, "ewc_loss": 0.007232666015625, "ewc_loss_diag": 4.9173831939697266e-06, "ewc_loss_parallel": 6.735324859619141e-06, "grad_norm": 1.4141978025436401, "learning_rate": 1.420093259855871e-06, "loss": 0.5731, "mean_token_accuracy": 0.8289796710014343, "num_tokens": 12781534.0, "step": 336 }, { "epoch": 0.042869863885001906, "ewc_loss": 0.00726318359375, "ewc_loss_diag": 4.9173831939697266e-06, "ewc_loss_parallel": 6.765127182006836e-06, "grad_norm": 1.4546715021133423, "learning_rate": 1.42433234421365e-06, "loss": 0.5785, "mean_token_accuracy": 0.8216186761856079, "num_tokens": 12815840.0, "step": 337 }, { "epoch": 0.04299707416359242, "ewc_loss": 0.00726318359375, "ewc_loss_diag": 4.947185516357422e-06, "ewc_loss_parallel": 6.765127182006836e-06, "grad_norm": 1.457509160041809, "learning_rate": 1.4285714285714286e-06, "loss": 0.586, "mean_token_accuracy": 0.8194944858551025, "num_tokens": 12849201.0, "step": 338 }, { "epoch": 0.04312428444218293, "ewc_loss": 0.007293701171875, "ewc_loss_diag": 4.947185516357422e-06, "ewc_loss_parallel": 6.794929504394531e-06, "grad_norm": 1.4631690979003906, "learning_rate": 1.4328105129292075e-06, "loss": 0.5894, "mean_token_accuracy": 0.8205030560493469, "num_tokens": 12888490.0, "step": 339 }, { "epoch": 0.043251494720773435, "ewc_loss": 0.00732421875, "ewc_loss_diag": 5.036592483520508e-06, "ewc_loss_parallel": 6.794929504394531e-06, "grad_norm": 1.4269527196884155, "learning_rate": 1.437049597286986e-06, "loss": 0.5641, "mean_token_accuracy": 0.8257450461387634, "num_tokens": 12927554.0, "step": 340 }, { "epoch": 0.04337870499936395, "ewc_loss": 0.00732421875, "ewc_loss_diag": 5.036592483520508e-06, "ewc_loss_parallel": 6.8247318267822266e-06, "grad_norm": 1.5369467735290527, "learning_rate": 1.4412886816447649e-06, "loss": 0.6117, "mean_token_accuracy": 0.8106664419174194, "num_tokens": 12960801.0, "step": 341 }, { "epoch": 0.04350591527795446, "ewc_loss": 0.007568359375, "ewc_loss_diag": 5.036592483520508e-06, "ewc_loss_parallel": 7.063150405883789e-06, "grad_norm": 1.474873423576355, "learning_rate": 1.4455277660025435e-06, "loss": 0.5295, "mean_token_accuracy": 0.8337363004684448, "num_tokens": 12993119.0, "step": 342 }, { "epoch": 0.04363312555654497, "ewc_loss": 0.00732421875, "ewc_loss_diag": 5.0961971282958984e-06, "ewc_loss_parallel": 6.794929504394531e-06, "grad_norm": 1.4116030931472778, "learning_rate": 1.4497668503603224e-06, "loss": 0.5897, "mean_token_accuracy": 0.8187210559844971, "num_tokens": 13033966.0, "step": 343 }, { "epoch": 0.04376033583513548, "ewc_loss": 0.00738525390625, "ewc_loss_diag": 5.0961971282958984e-06, "ewc_loss_parallel": 6.854534149169922e-06, "grad_norm": 1.4955486059188843, "learning_rate": 1.4540059347181009e-06, "loss": 0.567, "mean_token_accuracy": 0.8258072137832642, "num_tokens": 13063649.0, "step": 344 }, { "epoch": 0.043887546113725986, "ewc_loss": 0.007598876953125, "ewc_loss_diag": 5.0961971282958984e-06, "ewc_loss_parallel": 7.092952728271484e-06, "grad_norm": 1.445556640625, "learning_rate": 1.4582450190758798e-06, "loss": 0.6202, "mean_token_accuracy": 0.8088889122009277, "num_tokens": 13106926.0, "step": 345 }, { "epoch": 0.0440147563923165, "ewc_loss": 0.007415771484375, "ewc_loss_diag": 5.0961971282958984e-06, "ewc_loss_parallel": 6.884336471557617e-06, "grad_norm": 1.440261960029602, "learning_rate": 1.4624841034336584e-06, "loss": 0.5652, "mean_token_accuracy": 0.8260889649391174, "num_tokens": 13142961.0, "step": 346 }, { "epoch": 0.04414196667090701, "ewc_loss": 0.007537841796875, "ewc_loss_diag": 5.125999450683594e-06, "ewc_loss_parallel": 7.033348083496094e-06, "grad_norm": 1.4429724216461182, "learning_rate": 1.4667231877914373e-06, "loss": 0.5461, "mean_token_accuracy": 0.8305658102035522, "num_tokens": 13183544.0, "step": 347 }, { "epoch": 0.04426917694949752, "ewc_loss": 0.007568359375, "ewc_loss_diag": 5.0961971282958984e-06, "ewc_loss_parallel": 7.063150405883789e-06, "grad_norm": 1.4666554927825928, "learning_rate": 1.470962272149216e-06, "loss": 0.5632, "mean_token_accuracy": 0.8244975209236145, "num_tokens": 13222609.0, "step": 348 }, { "epoch": 0.04439638722808803, "ewc_loss": 0.007598876953125, "ewc_loss_diag": 5.155801773071289e-06, "ewc_loss_parallel": 7.092952728271484e-06, "grad_norm": 1.4683743715286255, "learning_rate": 1.4752013565069945e-06, "loss": 0.5617, "mean_token_accuracy": 0.8260489702224731, "num_tokens": 13258267.0, "step": 349 }, { "epoch": 0.04452359750667854, "ewc_loss": 0.007598876953125, "ewc_loss_diag": 5.185604095458984e-06, "ewc_loss_parallel": 7.092952728271484e-06, "grad_norm": 1.4696681499481201, "learning_rate": 1.4794404408647733e-06, "loss": 0.5679, "mean_token_accuracy": 0.8254774808883667, "num_tokens": 13294681.0, "step": 350 }, { "epoch": 0.04465080778526905, "ewc_loss": 0.00762939453125, "ewc_loss_diag": 5.185604095458984e-06, "ewc_loss_parallel": 7.12275505065918e-06, "grad_norm": 1.4337763786315918, "learning_rate": 1.483679525222552e-06, "loss": 0.5164, "mean_token_accuracy": 0.8365415334701538, "num_tokens": 13335469.0, "step": 351 }, { "epoch": 0.04477801806385956, "ewc_loss": 0.007598876953125, "ewc_loss_diag": 5.21540641784668e-06, "ewc_loss_parallel": 7.092952728271484e-06, "grad_norm": 1.4452528953552246, "learning_rate": 1.487918609580331e-06, "loss": 0.4959, "mean_token_accuracy": 0.8413029313087463, "num_tokens": 13372465.0, "step": 352 }, { "epoch": 0.04490522834245007, "ewc_loss": 0.007598876953125, "ewc_loss_diag": 5.27501106262207e-06, "ewc_loss_parallel": 7.092952728271484e-06, "grad_norm": 1.4601603746414185, "learning_rate": 1.4921576939381094e-06, "loss": 0.5451, "mean_token_accuracy": 0.8323934078216553, "num_tokens": 13409218.0, "step": 353 }, { "epoch": 0.04503243862104058, "ewc_loss": 0.00762939453125, "ewc_loss_diag": 5.27501106262207e-06, "ewc_loss_parallel": 7.12275505065918e-06, "grad_norm": 1.4446935653686523, "learning_rate": 1.4963967782958882e-06, "loss": 0.513, "mean_token_accuracy": 0.8405451774597168, "num_tokens": 13452548.0, "step": 354 }, { "epoch": 0.04515964889963109, "ewc_loss": 0.00775146484375, "ewc_loss_diag": 5.304813385009766e-06, "ewc_loss_parallel": 7.241964340209961e-06, "grad_norm": 1.4972424507141113, "learning_rate": 1.500635862653667e-06, "loss": 0.6012, "mean_token_accuracy": 0.8168280124664307, "num_tokens": 13488325.0, "step": 355 }, { "epoch": 0.0452868591782216, "ewc_loss": 0.0078125, "ewc_loss_diag": 5.304813385009766e-06, "ewc_loss_parallel": 7.3015689849853516e-06, "grad_norm": 1.4732856750488281, "learning_rate": 1.5048749470114458e-06, "loss": 0.6011, "mean_token_accuracy": 0.8154482245445251, "num_tokens": 13528203.0, "step": 356 }, { "epoch": 0.04541406945681211, "ewc_loss": 0.0076904296875, "ewc_loss_diag": 5.304813385009766e-06, "ewc_loss_parallel": 7.18235969543457e-06, "grad_norm": 1.4475144147872925, "learning_rate": 1.5091140313692243e-06, "loss": 0.5281, "mean_token_accuracy": 0.8345516920089722, "num_tokens": 13567242.0, "step": 357 }, { "epoch": 0.04554127973540262, "ewc_loss": 0.0078125, "ewc_loss_diag": 5.364418029785156e-06, "ewc_loss_parallel": 7.271766662597656e-06, "grad_norm": 1.4775888919830322, "learning_rate": 1.5133531157270031e-06, "loss": 0.5797, "mean_token_accuracy": 0.8170125484466553, "num_tokens": 13604521.0, "step": 358 }, { "epoch": 0.04566849001399313, "ewc_loss": 0.0078125, "ewc_loss_diag": 5.364418029785156e-06, "ewc_loss_parallel": 7.3015689849853516e-06, "grad_norm": 1.4691795110702515, "learning_rate": 1.5175922000847818e-06, "loss": 0.541, "mean_token_accuracy": 0.8334438800811768, "num_tokens": 13642787.0, "step": 359 }, { "epoch": 0.04579570029258364, "ewc_loss": 0.0078125, "ewc_loss_diag": 5.364418029785156e-06, "ewc_loss_parallel": 7.271766662597656e-06, "grad_norm": 1.4696178436279297, "learning_rate": 1.5218312844425607e-06, "loss": 0.565, "mean_token_accuracy": 0.8195292949676514, "num_tokens": 13684113.0, "step": 360 }, { "epoch": 0.045922910571174154, "ewc_loss": 0.00787353515625, "ewc_loss_diag": 5.364418029785156e-06, "ewc_loss_parallel": 7.331371307373047e-06, "grad_norm": 1.4761515855789185, "learning_rate": 1.5260703688003392e-06, "loss": 0.4955, "mean_token_accuracy": 0.8413887023925781, "num_tokens": 13719208.0, "step": 361 }, { "epoch": 0.04605012084976466, "ewc_loss": 0.00787353515625, "ewc_loss_diag": 5.364418029785156e-06, "ewc_loss_parallel": 7.361173629760742e-06, "grad_norm": 1.4564754962921143, "learning_rate": 1.530309453158118e-06, "loss": 0.5171, "mean_token_accuracy": 0.8388335704803467, "num_tokens": 13760159.0, "step": 362 }, { "epoch": 0.04617733112835517, "ewc_loss": 0.0078125, "ewc_loss_diag": 5.364418029785156e-06, "ewc_loss_parallel": 7.271766662597656e-06, "grad_norm": 1.49056077003479, "learning_rate": 1.5345485375158967e-06, "loss": 0.6042, "mean_token_accuracy": 0.813421368598938, "num_tokens": 13798375.0, "step": 363 }, { "epoch": 0.04630454140694568, "ewc_loss": 0.0079345703125, "ewc_loss_diag": 5.364418029785156e-06, "ewc_loss_parallel": 7.3909759521484375e-06, "grad_norm": 1.454585313796997, "learning_rate": 1.5387876218736756e-06, "loss": 0.4815, "mean_token_accuracy": 0.8482503890991211, "num_tokens": 13841924.0, "step": 364 }, { "epoch": 0.04643175168553619, "ewc_loss": 0.00787353515625, "ewc_loss_diag": 5.3942203521728516e-06, "ewc_loss_parallel": 7.331371307373047e-06, "grad_norm": 1.5091090202331543, "learning_rate": 1.543026706231454e-06, "loss": 0.5928, "mean_token_accuracy": 0.8197035789489746, "num_tokens": 13881436.0, "step": 365 }, { "epoch": 0.0465589619641267, "ewc_loss": 0.00811767578125, "ewc_loss_diag": 5.3942203521728516e-06, "ewc_loss_parallel": 7.569789886474609e-06, "grad_norm": 1.5197087526321411, "learning_rate": 1.547265790589233e-06, "loss": 0.605, "mean_token_accuracy": 0.8134722113609314, "num_tokens": 13916295.0, "step": 366 }, { "epoch": 0.04668617224271721, "ewc_loss": 0.00799560546875, "ewc_loss_diag": 5.3942203521728516e-06, "ewc_loss_parallel": 7.4803829193115234e-06, "grad_norm": 1.4925137758255005, "learning_rate": 1.5515048749470116e-06, "loss": 0.5695, "mean_token_accuracy": 0.8219290375709534, "num_tokens": 13950189.0, "step": 367 }, { "epoch": 0.04681338252130772, "ewc_loss": 0.0079345703125, "ewc_loss_diag": 5.424022674560547e-06, "ewc_loss_parallel": 7.3909759521484375e-06, "grad_norm": 1.4998221397399902, "learning_rate": 1.55574395930479e-06, "loss": 0.5765, "mean_token_accuracy": 0.8216679096221924, "num_tokens": 13988422.0, "step": 368 }, { "epoch": 0.046940592799898234, "ewc_loss": 0.008056640625, "ewc_loss_diag": 5.424022674560547e-06, "ewc_loss_parallel": 7.539987564086914e-06, "grad_norm": 1.4692013263702393, "learning_rate": 1.559983043662569e-06, "loss": 0.5734, "mean_token_accuracy": 0.8230719566345215, "num_tokens": 14029199.0, "step": 369 }, { "epoch": 0.04706780307848874, "ewc_loss": 0.00799560546875, "ewc_loss_diag": 5.424022674560547e-06, "ewc_loss_parallel": 7.450580596923828e-06, "grad_norm": 1.496444821357727, "learning_rate": 1.5642221280203476e-06, "loss": 0.5753, "mean_token_accuracy": 0.8238698244094849, "num_tokens": 14067976.0, "step": 370 }, { "epoch": 0.04719501335707925, "ewc_loss": 0.008056640625, "ewc_loss_diag": 5.453824996948242e-06, "ewc_loss_parallel": 7.539987564086914e-06, "grad_norm": 1.4899969100952148, "learning_rate": 1.5684612123781265e-06, "loss": 0.4971, "mean_token_accuracy": 0.844811201095581, "num_tokens": 14102858.0, "step": 371 }, { "epoch": 0.04732222363566976, "ewc_loss": 0.00811767578125, "ewc_loss_diag": 5.513429641723633e-06, "ewc_loss_parallel": 7.569789886474609e-06, "grad_norm": 1.5087474584579468, "learning_rate": 1.572700296735905e-06, "loss": 0.578, "mean_token_accuracy": 0.8267259001731873, "num_tokens": 14138560.0, "step": 372 }, { "epoch": 0.04744943391426027, "ewc_loss": 0.0081787109375, "ewc_loss_diag": 5.513429641723633e-06, "ewc_loss_parallel": 7.62939453125e-06, "grad_norm": 1.536096453666687, "learning_rate": 1.5769393810936839e-06, "loss": 0.5661, "mean_token_accuracy": 0.8253620266914368, "num_tokens": 14168007.0, "step": 373 }, { "epoch": 0.047576644192850785, "ewc_loss": 0.00823974609375, "ewc_loss_diag": 5.513429641723633e-06, "ewc_loss_parallel": 7.68899917602539e-06, "grad_norm": 1.4978569746017456, "learning_rate": 1.5811784654514625e-06, "loss": 0.4949, "mean_token_accuracy": 0.8470562696456909, "num_tokens": 14201779.0, "step": 374 }, { "epoch": 0.04770385447144129, "ewc_loss": 0.0081787109375, "ewc_loss_diag": 5.543231964111328e-06, "ewc_loss_parallel": 7.62939453125e-06, "grad_norm": 1.5769926309585571, "learning_rate": 1.5854175498092414e-06, "loss": 0.6239, "mean_token_accuracy": 0.8076012134552002, "num_tokens": 14245917.0, "step": 375 }, { "epoch": 0.0478310647500318, "ewc_loss": 0.00830078125, "ewc_loss_diag": 5.543231964111328e-06, "ewc_loss_parallel": 7.748603820800781e-06, "grad_norm": 1.505763053894043, "learning_rate": 1.5896566341670199e-06, "loss": 0.5819, "mean_token_accuracy": 0.8176189661026001, "num_tokens": 14284032.0, "step": 376 }, { "epoch": 0.047958275028622314, "ewc_loss": 0.0081787109375, "ewc_loss_diag": 5.5730342864990234e-06, "ewc_loss_parallel": 7.62939453125e-06, "grad_norm": 1.5188062191009521, "learning_rate": 1.5938957185247988e-06, "loss": 0.6312, "mean_token_accuracy": 0.8027209043502808, "num_tokens": 14322077.0, "step": 377 }, { "epoch": 0.04808548530721282, "ewc_loss": 0.00823974609375, "ewc_loss_diag": 5.5730342864990234e-06, "ewc_loss_parallel": 7.68899917602539e-06, "grad_norm": 1.5110416412353516, "learning_rate": 1.5981348028825774e-06, "loss": 0.5802, "mean_token_accuracy": 0.8201466798782349, "num_tokens": 14358031.0, "step": 378 }, { "epoch": 0.048212695585803336, "ewc_loss": 0.00823974609375, "ewc_loss_diag": 5.5730342864990234e-06, "ewc_loss_parallel": 7.68899917602539e-06, "grad_norm": 1.5593833923339844, "learning_rate": 1.6023738872403563e-06, "loss": 0.6698, "mean_token_accuracy": 0.7925280928611755, "num_tokens": 14390230.0, "step": 379 }, { "epoch": 0.048339905864393844, "ewc_loss": 0.00830078125, "ewc_loss_diag": 5.5730342864990234e-06, "ewc_loss_parallel": 7.748603820800781e-06, "grad_norm": 1.5105206966400146, "learning_rate": 1.6066129715981348e-06, "loss": 0.5221, "mean_token_accuracy": 0.8379214406013489, "num_tokens": 14427394.0, "step": 380 }, { "epoch": 0.04846711614298435, "ewc_loss": 0.00830078125, "ewc_loss_diag": 5.632638931274414e-06, "ewc_loss_parallel": 7.748603820800781e-06, "grad_norm": 1.4796252250671387, "learning_rate": 1.6108520559559137e-06, "loss": 0.5173, "mean_token_accuracy": 0.8385394215583801, "num_tokens": 14466606.0, "step": 381 }, { "epoch": 0.048594326421574865, "ewc_loss": 0.00830078125, "ewc_loss_diag": 5.632638931274414e-06, "ewc_loss_parallel": 7.748603820800781e-06, "grad_norm": 1.5654116868972778, "learning_rate": 1.6150911403136924e-06, "loss": 0.5851, "mean_token_accuracy": 0.8215964436531067, "num_tokens": 14494795.0, "step": 382 }, { "epoch": 0.04872153670016537, "ewc_loss": 0.00830078125, "ewc_loss_diag": 5.632638931274414e-06, "ewc_loss_parallel": 7.748603820800781e-06, "grad_norm": 1.4950376749038696, "learning_rate": 1.6193302246714712e-06, "loss": 0.5855, "mean_token_accuracy": 0.8208431601524353, "num_tokens": 14537400.0, "step": 383 }, { "epoch": 0.04884874697875588, "ewc_loss": 0.00830078125, "ewc_loss_diag": 5.662441253662109e-06, "ewc_loss_parallel": 7.748603820800781e-06, "grad_norm": 1.5320419073104858, "learning_rate": 1.62356930902925e-06, "loss": 0.6044, "mean_token_accuracy": 0.8210662603378296, "num_tokens": 14575215.0, "step": 384 }, { "epoch": 0.048975957257346395, "ewc_loss": 0.00836181640625, "ewc_loss_diag": 5.692243576049805e-06, "ewc_loss_parallel": 7.808208465576172e-06, "grad_norm": 1.556445837020874, "learning_rate": 1.6278083933870286e-06, "loss": 0.6279, "mean_token_accuracy": 0.805426299571991, "num_tokens": 14608486.0, "step": 385 }, { "epoch": 0.0491031675359369, "ewc_loss": 0.00830078125, "ewc_loss_diag": 5.692243576049805e-06, "ewc_loss_parallel": 7.748603820800781e-06, "grad_norm": 1.512369155883789, "learning_rate": 1.6320474777448073e-06, "loss": 0.5557, "mean_token_accuracy": 0.826830267906189, "num_tokens": 14644583.0, "step": 386 }, { "epoch": 0.049230377814527417, "ewc_loss": 0.00836181640625, "ewc_loss_diag": 5.7220458984375e-06, "ewc_loss_parallel": 7.808208465576172e-06, "grad_norm": 1.5167872905731201, "learning_rate": 1.636286562102586e-06, "loss": 0.5806, "mean_token_accuracy": 0.8186614513397217, "num_tokens": 14682976.0, "step": 387 }, { "epoch": 0.049357588093117924, "ewc_loss": 0.00836181640625, "ewc_loss_diag": 5.7220458984375e-06, "ewc_loss_parallel": 7.808208465576172e-06, "grad_norm": 1.535002589225769, "learning_rate": 1.6405256464603648e-06, "loss": 0.5513, "mean_token_accuracy": 0.8242713212966919, "num_tokens": 14717926.0, "step": 388 }, { "epoch": 0.04948479837170843, "ewc_loss": 0.00836181640625, "ewc_loss_diag": 5.7220458984375e-06, "ewc_loss_parallel": 7.808208465576172e-06, "grad_norm": 1.5590816736221313, "learning_rate": 1.6447647308181433e-06, "loss": 0.5925, "mean_token_accuracy": 0.8163919448852539, "num_tokens": 14750484.0, "step": 389 }, { "epoch": 0.049612008650298946, "ewc_loss": 0.00836181640625, "ewc_loss_diag": 5.751848220825195e-06, "ewc_loss_parallel": 7.808208465576172e-06, "grad_norm": 1.5255132913589478, "learning_rate": 1.6490038151759222e-06, "loss": 0.6095, "mean_token_accuracy": 0.8124276399612427, "num_tokens": 14790044.0, "step": 390 }, { "epoch": 0.04973921892888945, "ewc_loss": 0.0084228515625, "ewc_loss_diag": 5.781650543212891e-06, "ewc_loss_parallel": 7.808208465576172e-06, "grad_norm": 1.5126193761825562, "learning_rate": 1.6532428995337008e-06, "loss": 0.5301, "mean_token_accuracy": 0.8326547145843506, "num_tokens": 14828768.0, "step": 391 }, { "epoch": 0.04986642920747997, "ewc_loss": 0.0084228515625, "ewc_loss_diag": 5.781650543212891e-06, "ewc_loss_parallel": 7.808208465576172e-06, "grad_norm": 1.5241849422454834, "learning_rate": 1.6574819838914797e-06, "loss": 0.5184, "mean_token_accuracy": 0.8431554436683655, "num_tokens": 14865848.0, "step": 392 }, { "epoch": 0.049993639486070475, "ewc_loss": 0.0084228515625, "ewc_loss_diag": 5.781650543212891e-06, "ewc_loss_parallel": 7.867813110351562e-06, "grad_norm": 1.534327745437622, "learning_rate": 1.6617210682492582e-06, "loss": 0.5563, "mean_token_accuracy": 0.8263024091720581, "num_tokens": 14903973.0, "step": 393 }, { "epoch": 0.05012084976466098, "ewc_loss": 0.0084228515625, "ewc_loss_diag": 5.781650543212891e-06, "ewc_loss_parallel": 7.808208465576172e-06, "grad_norm": 1.5351253747940063, "learning_rate": 1.665960152607037e-06, "loss": 0.5909, "mean_token_accuracy": 0.8190243244171143, "num_tokens": 14944958.0, "step": 394 }, { "epoch": 0.0502480600432515, "ewc_loss": 0.0084228515625, "ewc_loss_diag": 5.781650543212891e-06, "ewc_loss_parallel": 7.867813110351562e-06, "grad_norm": 1.5401493310928345, "learning_rate": 1.6701992369648157e-06, "loss": 0.5815, "mean_token_accuracy": 0.8184165358543396, "num_tokens": 14985924.0, "step": 395 }, { "epoch": 0.050375270321842004, "ewc_loss": 0.0084228515625, "ewc_loss_diag": 5.781650543212891e-06, "ewc_loss_parallel": 7.867813110351562e-06, "grad_norm": 1.5451867580413818, "learning_rate": 1.6744383213225946e-06, "loss": 0.5476, "mean_token_accuracy": 0.8276803493499756, "num_tokens": 15022346.0, "step": 396 }, { "epoch": 0.05050248060043251, "ewc_loss": 0.00848388671875, "ewc_loss_diag": 5.8710575103759766e-06, "ewc_loss_parallel": 7.867813110351562e-06, "grad_norm": 1.524189829826355, "learning_rate": 1.678677405680373e-06, "loss": 0.4995, "mean_token_accuracy": 0.8420689105987549, "num_tokens": 15058656.0, "step": 397 }, { "epoch": 0.050629690879023026, "ewc_loss": 0.00848388671875, "ewc_loss_diag": 5.8710575103759766e-06, "ewc_loss_parallel": 7.867813110351562e-06, "grad_norm": 1.539178729057312, "learning_rate": 1.682916490038152e-06, "loss": 0.5342, "mean_token_accuracy": 0.8300353288650513, "num_tokens": 15094165.0, "step": 398 }, { "epoch": 0.05075690115761353, "ewc_loss": 0.00848388671875, "ewc_loss_diag": 5.8710575103759766e-06, "ewc_loss_parallel": 7.867813110351562e-06, "grad_norm": 1.5161981582641602, "learning_rate": 1.6871555743959306e-06, "loss": 0.5149, "mean_token_accuracy": 0.8368527889251709, "num_tokens": 15136979.0, "step": 399 }, { "epoch": 0.05088411143620405, "ewc_loss": 0.00848388671875, "ewc_loss_diag": 5.8710575103759766e-06, "ewc_loss_parallel": 7.867813110351562e-06, "grad_norm": 1.532487154006958, "learning_rate": 1.6913946587537095e-06, "loss": 0.5403, "mean_token_accuracy": 0.8311722874641418, "num_tokens": 15178765.0, "step": 400 }, { "epoch": 0.051011321714794555, "ewc_loss": 0.008544921875, "ewc_loss_diag": 5.8710575103759766e-06, "ewc_loss_parallel": 7.927417755126953e-06, "grad_norm": 1.551129698753357, "learning_rate": 1.695633743111488e-06, "loss": 0.5425, "mean_token_accuracy": 0.8322607278823853, "num_tokens": 15215892.0, "step": 401 }, { "epoch": 0.05113853199338506, "ewc_loss": 0.008544921875, "ewc_loss_diag": 5.900859832763672e-06, "ewc_loss_parallel": 7.927417755126953e-06, "grad_norm": 1.5605641603469849, "learning_rate": 1.6998728274692669e-06, "loss": 0.5136, "mean_token_accuracy": 0.8419578075408936, "num_tokens": 15248693.0, "step": 402 }, { "epoch": 0.05126574227197558, "ewc_loss": 0.00848388671875, "ewc_loss_diag": 5.900859832763672e-06, "ewc_loss_parallel": 7.867813110351562e-06, "grad_norm": 1.5478421449661255, "learning_rate": 1.7041119118270455e-06, "loss": 0.5381, "mean_token_accuracy": 0.8331764936447144, "num_tokens": 15289476.0, "step": 403 }, { "epoch": 0.051392952550566084, "ewc_loss": 0.008544921875, "ewc_loss_diag": 5.930662155151367e-06, "ewc_loss_parallel": 7.927417755126953e-06, "grad_norm": 1.574709415435791, "learning_rate": 1.708350996184824e-06, "loss": 0.5771, "mean_token_accuracy": 0.8187857270240784, "num_tokens": 15323032.0, "step": 404 }, { "epoch": 0.0515201628291566, "ewc_loss": 0.008544921875, "ewc_loss_diag": 5.930662155151367e-06, "ewc_loss_parallel": 7.927417755126953e-06, "grad_norm": 1.5333201885223389, "learning_rate": 1.7125900805426029e-06, "loss": 0.6074, "mean_token_accuracy": 0.8117043972015381, "num_tokens": 15369155.0, "step": 405 }, { "epoch": 0.051647373107747106, "ewc_loss": 0.0086669921875, "ewc_loss_diag": 5.930662155151367e-06, "ewc_loss_parallel": 8.046627044677734e-06, "grad_norm": 1.5656828880310059, "learning_rate": 1.7168291649003816e-06, "loss": 0.5246, "mean_token_accuracy": 0.833002507686615, "num_tokens": 15402623.0, "step": 406 }, { "epoch": 0.051774583386337614, "ewc_loss": 0.0086669921875, "ewc_loss_diag": 5.990266799926758e-06, "ewc_loss_parallel": 8.046627044677734e-06, "grad_norm": 1.6065969467163086, "learning_rate": 1.7210682492581604e-06, "loss": 0.562, "mean_token_accuracy": 0.8262203335762024, "num_tokens": 15440485.0, "step": 407 }, { "epoch": 0.05190179366492813, "ewc_loss": 0.00872802734375, "ewc_loss_diag": 5.990266799926758e-06, "ewc_loss_parallel": 8.106231689453125e-06, "grad_norm": 1.5549894571304321, "learning_rate": 1.725307333615939e-06, "loss": 0.5327, "mean_token_accuracy": 0.832926869392395, "num_tokens": 15479641.0, "step": 408 }, { "epoch": 0.052029003943518635, "ewc_loss": 0.00860595703125, "ewc_loss_diag": 6.020069122314453e-06, "ewc_loss_parallel": 7.987022399902344e-06, "grad_norm": 1.5664503574371338, "learning_rate": 1.7295464179737178e-06, "loss": 0.5741, "mean_token_accuracy": 0.8229367733001709, "num_tokens": 15519085.0, "step": 409 }, { "epoch": 0.05215621422210915, "ewc_loss": 0.0086669921875, "ewc_loss_diag": 6.0498714447021484e-06, "ewc_loss_parallel": 8.046627044677734e-06, "grad_norm": 1.5577338933944702, "learning_rate": 1.7337855023314965e-06, "loss": 0.5086, "mean_token_accuracy": 0.841084361076355, "num_tokens": 15557642.0, "step": 410 }, { "epoch": 0.05228342450069966, "ewc_loss": 0.00872802734375, "ewc_loss_diag": 6.079673767089844e-06, "ewc_loss_parallel": 8.106231689453125e-06, "grad_norm": 1.55793035030365, "learning_rate": 1.7380245866892753e-06, "loss": 0.5094, "mean_token_accuracy": 0.8425205945968628, "num_tokens": 15593178.0, "step": 411 }, { "epoch": 0.052410634779290165, "ewc_loss": 0.0087890625, "ewc_loss_diag": 6.109476089477539e-06, "ewc_loss_parallel": 8.165836334228516e-06, "grad_norm": 1.5673798322677612, "learning_rate": 1.7422636710470538e-06, "loss": 0.5641, "mean_token_accuracy": 0.8240542411804199, "num_tokens": 15630167.0, "step": 412 }, { "epoch": 0.05253784505788068, "ewc_loss": 0.0087890625, "ewc_loss_diag": 6.109476089477539e-06, "ewc_loss_parallel": 8.165836334228516e-06, "grad_norm": 1.5662287473678589, "learning_rate": 1.7465027554048327e-06, "loss": 0.5953, "mean_token_accuracy": 0.8171447515487671, "num_tokens": 15669380.0, "step": 413 }, { "epoch": 0.05266505533647119, "ewc_loss": 0.00885009765625, "ewc_loss_diag": 6.109476089477539e-06, "ewc_loss_parallel": 8.225440979003906e-06, "grad_norm": 1.5957012176513672, "learning_rate": 1.7507418397626114e-06, "loss": 0.5365, "mean_token_accuracy": 0.8340898752212524, "num_tokens": 15701791.0, "step": 414 }, { "epoch": 0.052792265615061694, "ewc_loss": 0.0087890625, "ewc_loss_diag": 6.109476089477539e-06, "ewc_loss_parallel": 8.165836334228516e-06, "grad_norm": 1.561529517173767, "learning_rate": 1.7549809241203902e-06, "loss": 0.5119, "mean_token_accuracy": 0.8382686376571655, "num_tokens": 15738223.0, "step": 415 }, { "epoch": 0.05291947589365221, "ewc_loss": 0.00885009765625, "ewc_loss_diag": 6.139278411865234e-06, "ewc_loss_parallel": 8.225440979003906e-06, "grad_norm": 1.5847606658935547, "learning_rate": 1.759220008478169e-06, "loss": 0.5931, "mean_token_accuracy": 0.8175950646400452, "num_tokens": 15773159.0, "step": 416 }, { "epoch": 0.053046686172242716, "ewc_loss": 0.00885009765625, "ewc_loss_diag": 6.139278411865234e-06, "ewc_loss_parallel": 8.225440979003906e-06, "grad_norm": 1.58534574508667, "learning_rate": 1.7634590928359476e-06, "loss": 0.5598, "mean_token_accuracy": 0.8272497653961182, "num_tokens": 15813132.0, "step": 417 }, { "epoch": 0.05317389645083323, "ewc_loss": 0.00885009765625, "ewc_loss_diag": 6.139278411865234e-06, "ewc_loss_parallel": 8.225440979003906e-06, "grad_norm": 1.563980221748352, "learning_rate": 1.7676981771937263e-06, "loss": 0.5407, "mean_token_accuracy": 0.8291997909545898, "num_tokens": 15849752.0, "step": 418 }, { "epoch": 0.05330110672942374, "ewc_loss": 0.00885009765625, "ewc_loss_diag": 6.139278411865234e-06, "ewc_loss_parallel": 8.225440979003906e-06, "grad_norm": 1.582177996635437, "learning_rate": 1.7719372615515052e-06, "loss": 0.5189, "mean_token_accuracy": 0.8406234383583069, "num_tokens": 15885331.0, "step": 419 }, { "epoch": 0.053428317008014245, "ewc_loss": 0.0089111328125, "ewc_loss_diag": 6.16908073425293e-06, "ewc_loss_parallel": 8.285045623779297e-06, "grad_norm": 1.650259256362915, "learning_rate": 1.7761763459092838e-06, "loss": 0.5602, "mean_token_accuracy": 0.8293783068656921, "num_tokens": 15924649.0, "step": 420 }, { "epoch": 0.05355552728660476, "ewc_loss": 0.009033203125, "ewc_loss_diag": 6.16908073425293e-06, "ewc_loss_parallel": 8.404254913330078e-06, "grad_norm": 1.5864850282669067, "learning_rate": 1.7804154302670625e-06, "loss": 0.5298, "mean_token_accuracy": 0.8334692716598511, "num_tokens": 15959342.0, "step": 421 }, { "epoch": 0.05368273756519527, "ewc_loss": 0.00897216796875, "ewc_loss_diag": 6.198883056640625e-06, "ewc_loss_parallel": 8.344650268554688e-06, "grad_norm": 1.5649547576904297, "learning_rate": 1.7846545146248412e-06, "loss": 0.5171, "mean_token_accuracy": 0.8373861312866211, "num_tokens": 15998479.0, "step": 422 }, { "epoch": 0.05380994784378578, "ewc_loss": 0.00897216796875, "ewc_loss_diag": 6.258487701416016e-06, "ewc_loss_parallel": 8.344650268554688e-06, "grad_norm": 1.603604793548584, "learning_rate": 1.7888935989826198e-06, "loss": 0.5603, "mean_token_accuracy": 0.8238976001739502, "num_tokens": 16034197.0, "step": 423 }, { "epoch": 0.05393715812237629, "ewc_loss": 0.00897216796875, "ewc_loss_diag": 6.288290023803711e-06, "ewc_loss_parallel": 8.344650268554688e-06, "grad_norm": 1.5933722257614136, "learning_rate": 1.7931326833403987e-06, "loss": 0.5731, "mean_token_accuracy": 0.8200205564498901, "num_tokens": 16071448.0, "step": 424 }, { "epoch": 0.054064368400966796, "ewc_loss": 0.00897216796875, "ewc_loss_diag": 6.288290023803711e-06, "ewc_loss_parallel": 8.344650268554688e-06, "grad_norm": 1.5549907684326172, "learning_rate": 1.7973717676981772e-06, "loss": 0.497, "mean_token_accuracy": 0.8460460305213928, "num_tokens": 16116007.0, "step": 425 }, { "epoch": 0.05419157867955731, "ewc_loss": 0.009033203125, "ewc_loss_diag": 6.288290023803711e-06, "ewc_loss_parallel": 8.404254913330078e-06, "grad_norm": 1.587754726409912, "learning_rate": 1.801610852055956e-06, "loss": 0.5341, "mean_token_accuracy": 0.8352015614509583, "num_tokens": 16158070.0, "step": 426 }, { "epoch": 0.05431878895814782, "ewc_loss": 0.0091552734375, "ewc_loss_diag": 6.288290023803711e-06, "ewc_loss_parallel": 8.52346420288086e-06, "grad_norm": 1.5957549810409546, "learning_rate": 1.8058499364137347e-06, "loss": 0.5748, "mean_token_accuracy": 0.8248500823974609, "num_tokens": 16196310.0, "step": 427 }, { "epoch": 0.054445999236738325, "ewc_loss": 0.00921630859375, "ewc_loss_diag": 6.3478946685791016e-06, "ewc_loss_parallel": 8.58306884765625e-06, "grad_norm": 1.5854188203811646, "learning_rate": 1.8100890207715136e-06, "loss": 0.622, "mean_token_accuracy": 0.8056217432022095, "num_tokens": 16237927.0, "step": 428 }, { "epoch": 0.05457320951532884, "ewc_loss": 0.0091552734375, "ewc_loss_diag": 6.3478946685791016e-06, "ewc_loss_parallel": 8.52346420288086e-06, "grad_norm": 1.6047132015228271, "learning_rate": 1.814328105129292e-06, "loss": 0.5757, "mean_token_accuracy": 0.818745493888855, "num_tokens": 16274865.0, "step": 429 }, { "epoch": 0.05470041979391935, "ewc_loss": 0.00921630859375, "ewc_loss_diag": 6.377696990966797e-06, "ewc_loss_parallel": 8.58306884765625e-06, "grad_norm": 1.6470175981521606, "learning_rate": 1.818567189487071e-06, "loss": 0.5525, "mean_token_accuracy": 0.8237946033477783, "num_tokens": 16302503.0, "step": 430 }, { "epoch": 0.05482763007250986, "ewc_loss": 0.0091552734375, "ewc_loss_diag": 6.407499313354492e-06, "ewc_loss_parallel": 8.52346420288086e-06, "grad_norm": 1.597713589668274, "learning_rate": 1.8228062738448496e-06, "loss": 0.607, "mean_token_accuracy": 0.8132807016372681, "num_tokens": 16342472.0, "step": 431 }, { "epoch": 0.05495484035110037, "ewc_loss": 0.0091552734375, "ewc_loss_diag": 6.407499313354492e-06, "ewc_loss_parallel": 8.52346420288086e-06, "grad_norm": 1.599118947982788, "learning_rate": 1.8270453582026285e-06, "loss": 0.5665, "mean_token_accuracy": 0.8232022523880005, "num_tokens": 16378423.0, "step": 432 }, { "epoch": 0.055082050629690876, "ewc_loss": 0.0091552734375, "ewc_loss_diag": 6.407499313354492e-06, "ewc_loss_parallel": 8.52346420288086e-06, "grad_norm": 1.5931105613708496, "learning_rate": 1.831284442560407e-06, "loss": 0.5357, "mean_token_accuracy": 0.8346028327941895, "num_tokens": 16418430.0, "step": 433 }, { "epoch": 0.05520926090828139, "ewc_loss": 0.00921630859375, "ewc_loss_diag": 6.467103958129883e-06, "ewc_loss_parallel": 8.52346420288086e-06, "grad_norm": 1.6040332317352295, "learning_rate": 1.8355235269181859e-06, "loss": 0.5487, "mean_token_accuracy": 0.8297663331031799, "num_tokens": 16460984.0, "step": 434 }, { "epoch": 0.0553364711868719, "ewc_loss": 0.00921630859375, "ewc_loss_diag": 6.467103958129883e-06, "ewc_loss_parallel": 8.52346420288086e-06, "grad_norm": 1.6189874410629272, "learning_rate": 1.8397626112759646e-06, "loss": 0.5266, "mean_token_accuracy": 0.8355793952941895, "num_tokens": 16493460.0, "step": 435 }, { "epoch": 0.05546368146546241, "ewc_loss": 0.00921630859375, "ewc_loss_diag": 6.467103958129883e-06, "ewc_loss_parallel": 8.52346420288086e-06, "grad_norm": 1.6120299100875854, "learning_rate": 1.8440016956337434e-06, "loss": 0.5045, "mean_token_accuracy": 0.8406527638435364, "num_tokens": 16528998.0, "step": 436 }, { "epoch": 0.05559089174405292, "ewc_loss": 0.00927734375, "ewc_loss_diag": 6.496906280517578e-06, "ewc_loss_parallel": 8.58306884765625e-06, "grad_norm": 1.6183862686157227, "learning_rate": 1.848240779991522e-06, "loss": 0.5301, "mean_token_accuracy": 0.8344857096672058, "num_tokens": 16564041.0, "step": 437 }, { "epoch": 0.05571810202264343, "ewc_loss": 0.0091552734375, "ewc_loss_diag": 6.5267086029052734e-06, "ewc_loss_parallel": 8.463859558105469e-06, "grad_norm": 1.6591647863388062, "learning_rate": 1.8524798643493008e-06, "loss": 0.6041, "mean_token_accuracy": 0.8107244968414307, "num_tokens": 16597967.0, "step": 438 }, { "epoch": 0.05584531230123394, "ewc_loss": 0.00921630859375, "ewc_loss_diag": 6.5267086029052734e-06, "ewc_loss_parallel": 8.52346420288086e-06, "grad_norm": 1.6146409511566162, "learning_rate": 1.8567189487070795e-06, "loss": 0.587, "mean_token_accuracy": 0.8192811012268066, "num_tokens": 16636756.0, "step": 439 }, { "epoch": 0.05597252257982445, "ewc_loss": 0.00927734375, "ewc_loss_diag": 6.556510925292969e-06, "ewc_loss_parallel": 8.58306884765625e-06, "grad_norm": 1.5921940803527832, "learning_rate": 1.860958033064858e-06, "loss": 0.4994, "mean_token_accuracy": 0.8427536487579346, "num_tokens": 16675811.0, "step": 440 }, { "epoch": 0.05609973285841496, "ewc_loss": 0.00933837890625, "ewc_loss_diag": 6.616115570068359e-06, "ewc_loss_parallel": 8.64267349243164e-06, "grad_norm": 1.6272718906402588, "learning_rate": 1.8651971174226368e-06, "loss": 0.5843, "mean_token_accuracy": 0.821837306022644, "num_tokens": 16713714.0, "step": 441 }, { "epoch": 0.05622694313700547, "ewc_loss": 0.0093994140625, "ewc_loss_diag": 6.616115570068359e-06, "ewc_loss_parallel": 8.702278137207031e-06, "grad_norm": 1.6097391843795776, "learning_rate": 1.8694362017804155e-06, "loss": 0.5394, "mean_token_accuracy": 0.8315080404281616, "num_tokens": 16754317.0, "step": 442 }, { "epoch": 0.05635415341559598, "ewc_loss": 0.0093994140625, "ewc_loss_diag": 6.616115570068359e-06, "ewc_loss_parallel": 8.702278137207031e-06, "grad_norm": 1.6121916770935059, "learning_rate": 1.8736752861381944e-06, "loss": 0.501, "mean_token_accuracy": 0.8438778519630432, "num_tokens": 16793189.0, "step": 443 }, { "epoch": 0.05648136369418649, "ewc_loss": 0.0093994140625, "ewc_loss_diag": 6.616115570068359e-06, "ewc_loss_parallel": 8.702278137207031e-06, "grad_norm": 1.6325998306274414, "learning_rate": 1.8779143704959728e-06, "loss": 0.5108, "mean_token_accuracy": 0.8400088548660278, "num_tokens": 16825046.0, "step": 444 }, { "epoch": 0.056608573972777, "ewc_loss": 0.0093994140625, "ewc_loss_diag": 6.645917892456055e-06, "ewc_loss_parallel": 8.702278137207031e-06, "grad_norm": 1.6566723585128784, "learning_rate": 1.8821534548537517e-06, "loss": 0.5735, "mean_token_accuracy": 0.8228214383125305, "num_tokens": 16857769.0, "step": 445 }, { "epoch": 0.05673578425136751, "ewc_loss": 0.009521484375, "ewc_loss_diag": 6.645917892456055e-06, "ewc_loss_parallel": 8.821487426757812e-06, "grad_norm": 1.6523528099060059, "learning_rate": 1.8863925392115304e-06, "loss": 0.532, "mean_token_accuracy": 0.8349751234054565, "num_tokens": 16900878.0, "step": 446 }, { "epoch": 0.05686299452995802, "ewc_loss": 0.0093994140625, "ewc_loss_diag": 6.67572021484375e-06, "ewc_loss_parallel": 8.702278137207031e-06, "grad_norm": 1.6077392101287842, "learning_rate": 1.8906316235693093e-06, "loss": 0.5459, "mean_token_accuracy": 0.8284565210342407, "num_tokens": 16942832.0, "step": 447 }, { "epoch": 0.05699020480854853, "ewc_loss": 0.0093994140625, "ewc_loss_diag": 6.67572021484375e-06, "ewc_loss_parallel": 8.702278137207031e-06, "grad_norm": 1.6435184478759766, "learning_rate": 1.8948707079270877e-06, "loss": 0.5492, "mean_token_accuracy": 0.8220034837722778, "num_tokens": 16975609.0, "step": 448 }, { "epoch": 0.057117415087139044, "ewc_loss": 0.009521484375, "ewc_loss_diag": 6.67572021484375e-06, "ewc_loss_parallel": 8.821487426757812e-06, "grad_norm": 1.6277656555175781, "learning_rate": 1.8991097922848666e-06, "loss": 0.5201, "mean_token_accuracy": 0.8367922306060791, "num_tokens": 17012620.0, "step": 449 }, { "epoch": 0.05724462536572955, "ewc_loss": 0.0093994140625, "ewc_loss_diag": 6.67572021484375e-06, "ewc_loss_parallel": 8.702278137207031e-06, "grad_norm": 1.6315503120422363, "learning_rate": 1.9033488766426453e-06, "loss": 0.5652, "mean_token_accuracy": 0.8212533593177795, "num_tokens": 17053820.0, "step": 450 }, { "epoch": 0.05737183564432006, "ewc_loss": 0.00946044921875, "ewc_loss_diag": 6.735324859619141e-06, "ewc_loss_parallel": 8.761882781982422e-06, "grad_norm": 1.8043383359909058, "learning_rate": 1.907587961000424e-06, "loss": 0.5186, "mean_token_accuracy": 0.8404548168182373, "num_tokens": 17091885.0, "step": 451 }, { "epoch": 0.05749904592291057, "ewc_loss": 0.00970458984375, "ewc_loss_diag": 6.735324859619141e-06, "ewc_loss_parallel": 9.059906005859375e-06, "grad_norm": 1.623203992843628, "learning_rate": 1.911827045358203e-06, "loss": 0.5655, "mean_token_accuracy": 0.825050413608551, "num_tokens": 17133432.0, "step": 452 }, { "epoch": 0.05762625620150108, "ewc_loss": 0.00946044921875, "ewc_loss_diag": 6.765127182006836e-06, "ewc_loss_parallel": 8.761882781982422e-06, "grad_norm": 1.6234941482543945, "learning_rate": 1.9160661297159817e-06, "loss": 0.5875, "mean_token_accuracy": 0.8192061185836792, "num_tokens": 17171845.0, "step": 453 }, { "epoch": 0.057753466480091595, "ewc_loss": 0.00958251953125, "ewc_loss_diag": 6.765127182006836e-06, "ewc_loss_parallel": 8.940696716308594e-06, "grad_norm": 1.6599918603897095, "learning_rate": 1.92030521407376e-06, "loss": 0.5763, "mean_token_accuracy": 0.8186559677124023, "num_tokens": 17206137.0, "step": 454 }, { "epoch": 0.0578806767586821, "ewc_loss": 0.00946044921875, "ewc_loss_diag": 6.765127182006836e-06, "ewc_loss_parallel": 8.761882781982422e-06, "grad_norm": 1.619860053062439, "learning_rate": 1.924544298431539e-06, "loss": 0.5561, "mean_token_accuracy": 0.8294466733932495, "num_tokens": 17247619.0, "step": 455 }, { "epoch": 0.05800788703727261, "ewc_loss": 0.009521484375, "ewc_loss_diag": 6.794929504394531e-06, "ewc_loss_parallel": 8.821487426757812e-06, "grad_norm": 1.682185173034668, "learning_rate": 1.9287833827893175e-06, "loss": 0.5786, "mean_token_accuracy": 0.8194795846939087, "num_tokens": 17285053.0, "step": 456 }, { "epoch": 0.058135097315863124, "ewc_loss": 0.00970458984375, "ewc_loss_diag": 6.8247318267822266e-06, "ewc_loss_parallel": 9.059906005859375e-06, "grad_norm": 1.6598503589630127, "learning_rate": 1.9330224671470964e-06, "loss": 0.5527, "mean_token_accuracy": 0.825764536857605, "num_tokens": 17320303.0, "step": 457 }, { "epoch": 0.05826230759445363, "ewc_loss": 0.00958251953125, "ewc_loss_diag": 6.8247318267822266e-06, "ewc_loss_parallel": 8.940696716308594e-06, "grad_norm": 1.6713038682937622, "learning_rate": 1.9372615515048753e-06, "loss": 0.6045, "mean_token_accuracy": 0.817713737487793, "num_tokens": 17354102.0, "step": 458 }, { "epoch": 0.05838951787304414, "ewc_loss": 0.0096435546875, "ewc_loss_diag": 6.854534149169922e-06, "ewc_loss_parallel": 9.000301361083984e-06, "grad_norm": 1.6543762683868408, "learning_rate": 1.9415006358626538e-06, "loss": 0.5437, "mean_token_accuracy": 0.8307119011878967, "num_tokens": 17389958.0, "step": 459 }, { "epoch": 0.05851672815163465, "ewc_loss": 0.0096435546875, "ewc_loss_diag": 6.884336471557617e-06, "ewc_loss_parallel": 9.000301361083984e-06, "grad_norm": 1.676563024520874, "learning_rate": 1.9457397202204326e-06, "loss": 0.5302, "mean_token_accuracy": 0.8340225219726562, "num_tokens": 17421488.0, "step": 460 }, { "epoch": 0.05864393843022516, "ewc_loss": 0.00970458984375, "ewc_loss_diag": 6.884336471557617e-06, "ewc_loss_parallel": 9.059906005859375e-06, "grad_norm": 1.653337001800537, "learning_rate": 1.949978804578211e-06, "loss": 0.4935, "mean_token_accuracy": 0.8467695116996765, "num_tokens": 17455172.0, "step": 461 }, { "epoch": 0.058771148708815675, "ewc_loss": 0.0096435546875, "ewc_loss_diag": 6.9141387939453125e-06, "ewc_loss_parallel": 9.000301361083984e-06, "grad_norm": 1.6553717851638794, "learning_rate": 1.95421788893599e-06, "loss": 0.5913, "mean_token_accuracy": 0.8149784207344055, "num_tokens": 17494798.0, "step": 462 }, { "epoch": 0.05889835898740618, "ewc_loss": 0.009765625, "ewc_loss_diag": 7.0035457611083984e-06, "ewc_loss_parallel": 9.059906005859375e-06, "grad_norm": 1.7024108171463013, "learning_rate": 1.9584569732937684e-06, "loss": 0.5562, "mean_token_accuracy": 0.8325195908546448, "num_tokens": 17527619.0, "step": 463 }, { "epoch": 0.05902556926599669, "ewc_loss": 0.00970458984375, "ewc_loss_diag": 6.973743438720703e-06, "ewc_loss_parallel": 9.059906005859375e-06, "grad_norm": 1.636559009552002, "learning_rate": 1.9626960576515473e-06, "loss": 0.4743, "mean_token_accuracy": 0.8509654998779297, "num_tokens": 17564299.0, "step": 464 }, { "epoch": 0.059152779544587204, "ewc_loss": 0.0096435546875, "ewc_loss_diag": 6.973743438720703e-06, "ewc_loss_parallel": 9.000301361083984e-06, "grad_norm": 1.6801538467407227, "learning_rate": 1.9669351420093262e-06, "loss": 0.5899, "mean_token_accuracy": 0.8193221688270569, "num_tokens": 17598333.0, "step": 465 }, { "epoch": 0.05927998982317771, "ewc_loss": 0.00982666015625, "ewc_loss_diag": 6.973743438720703e-06, "ewc_loss_parallel": 9.179115295410156e-06, "grad_norm": 1.6514681577682495, "learning_rate": 1.971174226367105e-06, "loss": 0.5719, "mean_token_accuracy": 0.8256192207336426, "num_tokens": 17642106.0, "step": 466 }, { "epoch": 0.059407200101768226, "ewc_loss": 0.009765625, "ewc_loss_diag": 7.033348083496094e-06, "ewc_loss_parallel": 9.059906005859375e-06, "grad_norm": 1.66447913646698, "learning_rate": 1.9754133107248836e-06, "loss": 0.5148, "mean_token_accuracy": 0.8346198797225952, "num_tokens": 17676412.0, "step": 467 }, { "epoch": 0.059534410380358734, "ewc_loss": 0.0098876953125, "ewc_loss_diag": 7.063150405883789e-06, "ewc_loss_parallel": 9.179115295410156e-06, "grad_norm": 1.688578724861145, "learning_rate": 1.9796523950826624e-06, "loss": 0.5967, "mean_token_accuracy": 0.8169132471084595, "num_tokens": 17717079.0, "step": 468 }, { "epoch": 0.05966162065894924, "ewc_loss": 0.009765625, "ewc_loss_diag": 7.063150405883789e-06, "ewc_loss_parallel": 9.059906005859375e-06, "grad_norm": 1.6693922281265259, "learning_rate": 1.983891479440441e-06, "loss": 0.486, "mean_token_accuracy": 0.8466586470603943, "num_tokens": 17753603.0, "step": 469 }, { "epoch": 0.059788830937539755, "ewc_loss": 0.00994873046875, "ewc_loss_diag": 7.12275505065918e-06, "ewc_loss_parallel": 9.238719940185547e-06, "grad_norm": 1.6612110137939453, "learning_rate": 1.98813056379822e-06, "loss": 0.5529, "mean_token_accuracy": 0.8292558789253235, "num_tokens": 17795960.0, "step": 470 }, { "epoch": 0.05991604121613026, "ewc_loss": 0.0098876953125, "ewc_loss_diag": 7.12275505065918e-06, "ewc_loss_parallel": 9.179115295410156e-06, "grad_norm": 1.701310157775879, "learning_rate": 1.9923696481559983e-06, "loss": 0.5704, "mean_token_accuracy": 0.8242899775505066, "num_tokens": 17829298.0, "step": 471 }, { "epoch": 0.06004325149472077, "ewc_loss": 0.00994873046875, "ewc_loss_diag": 7.12275505065918e-06, "ewc_loss_parallel": 9.238719940185547e-06, "grad_norm": 1.6733309030532837, "learning_rate": 1.996608732513777e-06, "loss": 0.6057, "mean_token_accuracy": 0.8142633438110352, "num_tokens": 17872027.0, "step": 472 }, { "epoch": 0.060170461773311285, "ewc_loss": 0.00982666015625, "ewc_loss_diag": 7.12275505065918e-06, "ewc_loss_parallel": 9.119510650634766e-06, "grad_norm": 1.6575168371200562, "learning_rate": 2.000847816871556e-06, "loss": 0.4922, "mean_token_accuracy": 0.8427748084068298, "num_tokens": 17911846.0, "step": 473 }, { "epoch": 0.06029767205190179, "ewc_loss": 0.00982666015625, "ewc_loss_diag": 7.12275505065918e-06, "ewc_loss_parallel": 9.119510650634766e-06, "grad_norm": 1.6852844953536987, "learning_rate": 2.005086901229335e-06, "loss": 0.4891, "mean_token_accuracy": 0.84425950050354, "num_tokens": 17946095.0, "step": 474 }, { "epoch": 0.060424882330492306, "ewc_loss": 0.01007080078125, "ewc_loss_diag": 7.18235969543457e-06, "ewc_loss_parallel": 9.357929229736328e-06, "grad_norm": 1.6904387474060059, "learning_rate": 2.0093259855871134e-06, "loss": 0.5201, "mean_token_accuracy": 0.8365399837493896, "num_tokens": 17983232.0, "step": 475 }, { "epoch": 0.060552092609082814, "ewc_loss": 0.00994873046875, "ewc_loss_diag": 7.18235969543457e-06, "ewc_loss_parallel": 9.238719940185547e-06, "grad_norm": 1.6882561445236206, "learning_rate": 2.0135650699448923e-06, "loss": 0.6095, "mean_token_accuracy": 0.809468150138855, "num_tokens": 18022455.0, "step": 476 }, { "epoch": 0.06067930288767332, "ewc_loss": 0.01007080078125, "ewc_loss_diag": 7.18235969543457e-06, "ewc_loss_parallel": 9.357929229736328e-06, "grad_norm": 1.6953129768371582, "learning_rate": 2.0178041543026707e-06, "loss": 0.575, "mean_token_accuracy": 0.8217644095420837, "num_tokens": 18059844.0, "step": 477 }, { "epoch": 0.060806513166263836, "ewc_loss": 0.00994873046875, "ewc_loss_diag": 7.212162017822266e-06, "ewc_loss_parallel": 9.238719940185547e-06, "grad_norm": 1.6980595588684082, "learning_rate": 2.022043238660449e-06, "loss": 0.6244, "mean_token_accuracy": 0.8067919015884399, "num_tokens": 18099569.0, "step": 478 }, { "epoch": 0.06093372344485434, "ewc_loss": 0.010009765625, "ewc_loss_diag": 7.212162017822266e-06, "ewc_loss_parallel": 9.298324584960938e-06, "grad_norm": 1.7071489095687866, "learning_rate": 2.026282323018228e-06, "loss": 0.5867, "mean_token_accuracy": 0.8173770904541016, "num_tokens": 18134747.0, "step": 479 }, { "epoch": 0.06106093372344486, "ewc_loss": 0.0101318359375, "ewc_loss_diag": 7.271766662597656e-06, "ewc_loss_parallel": 9.417533874511719e-06, "grad_norm": 1.6887152194976807, "learning_rate": 2.030521407376007e-06, "loss": 0.5941, "mean_token_accuracy": 0.8197326064109802, "num_tokens": 18173989.0, "step": 480 }, { "epoch": 0.061188144002035365, "ewc_loss": 0.010009765625, "ewc_loss_diag": 7.271766662597656e-06, "ewc_loss_parallel": 9.298324584960938e-06, "grad_norm": 1.7236148118972778, "learning_rate": 2.034760491733786e-06, "loss": 0.5652, "mean_token_accuracy": 0.8226850032806396, "num_tokens": 18205065.0, "step": 481 }, { "epoch": 0.06131535428062587, "ewc_loss": 0.01025390625, "ewc_loss_diag": 7.3015689849853516e-06, "ewc_loss_parallel": 9.5367431640625e-06, "grad_norm": 1.6802188158035278, "learning_rate": 2.0389995760915643e-06, "loss": 0.5145, "mean_token_accuracy": 0.840285062789917, "num_tokens": 18242999.0, "step": 482 }, { "epoch": 0.06144256455921639, "ewc_loss": 0.01007080078125, "ewc_loss_diag": 7.331371307373047e-06, "ewc_loss_parallel": 9.357929229736328e-06, "grad_norm": 1.6950920820236206, "learning_rate": 2.043238660449343e-06, "loss": 0.5968, "mean_token_accuracy": 0.8160540461540222, "num_tokens": 18280642.0, "step": 483 }, { "epoch": 0.061569774837806894, "ewc_loss": 0.01031494140625, "ewc_loss_diag": 7.361173629760742e-06, "ewc_loss_parallel": 9.59634780883789e-06, "grad_norm": 1.6886707544326782, "learning_rate": 2.0474777448071216e-06, "loss": 0.5205, "mean_token_accuracy": 0.8393411636352539, "num_tokens": 18317894.0, "step": 484 }, { "epoch": 0.0616969851163974, "ewc_loss": 0.01025390625, "ewc_loss_diag": 7.420778274536133e-06, "ewc_loss_parallel": 9.5367431640625e-06, "grad_norm": 1.7090729475021362, "learning_rate": 2.0517168291649005e-06, "loss": 0.5487, "mean_token_accuracy": 0.8279008865356445, "num_tokens": 18350481.0, "step": 485 }, { "epoch": 0.061824195394987916, "ewc_loss": 0.0103759765625, "ewc_loss_diag": 7.450580596923828e-06, "ewc_loss_parallel": 9.655952453613281e-06, "grad_norm": 1.723507285118103, "learning_rate": 2.0559559135226794e-06, "loss": 0.5579, "mean_token_accuracy": 0.8222766518592834, "num_tokens": 18384390.0, "step": 486 }, { "epoch": 0.06195140567357842, "ewc_loss": 0.0103759765625, "ewc_loss_diag": 7.450580596923828e-06, "ewc_loss_parallel": 9.655952453613281e-06, "grad_norm": 1.6799896955490112, "learning_rate": 2.060194997880458e-06, "loss": 0.5166, "mean_token_accuracy": 0.8359056711196899, "num_tokens": 18423958.0, "step": 487 }, { "epoch": 0.06207861595216894, "ewc_loss": 0.01043701171875, "ewc_loss_diag": 7.4803829193115234e-06, "ewc_loss_parallel": 9.715557098388672e-06, "grad_norm": 1.7104277610778809, "learning_rate": 2.0644340822382368e-06, "loss": 0.4816, "mean_token_accuracy": 0.8477802872657776, "num_tokens": 18459622.0, "step": 488 }, { "epoch": 0.062205826230759445, "ewc_loss": 0.01031494140625, "ewc_loss_diag": 7.4803829193115234e-06, "ewc_loss_parallel": 9.59634780883789e-06, "grad_norm": 1.7095940113067627, "learning_rate": 2.0686731665960156e-06, "loss": 0.5386, "mean_token_accuracy": 0.8350310325622559, "num_tokens": 18495568.0, "step": 489 }, { "epoch": 0.06233303650934995, "ewc_loss": 0.010498046875, "ewc_loss_diag": 7.510185241699219e-06, "ewc_loss_parallel": 9.775161743164062e-06, "grad_norm": 1.6995525360107422, "learning_rate": 2.072912250953794e-06, "loss": 0.5056, "mean_token_accuracy": 0.8460729122161865, "num_tokens": 18534440.0, "step": 490 }, { "epoch": 0.06246024678794047, "ewc_loss": 0.01043701171875, "ewc_loss_diag": 7.4803829193115234e-06, "ewc_loss_parallel": 9.715557098388672e-06, "grad_norm": 1.715100646018982, "learning_rate": 2.077151335311573e-06, "loss": 0.5413, "mean_token_accuracy": 0.8298213481903076, "num_tokens": 18570583.0, "step": 491 }, { "epoch": 0.06258745706653097, "ewc_loss": 0.010498046875, "ewc_loss_diag": 7.510185241699219e-06, "ewc_loss_parallel": 9.775161743164062e-06, "grad_norm": 1.7020570039749146, "learning_rate": 2.0813904196693514e-06, "loss": 0.5198, "mean_token_accuracy": 0.8379532098770142, "num_tokens": 18611801.0, "step": 492 }, { "epoch": 0.06271466734512149, "ewc_loss": 0.010498046875, "ewc_loss_diag": 7.569789886474609e-06, "ewc_loss_parallel": 9.775161743164062e-06, "grad_norm": 1.7297049760818481, "learning_rate": 2.0856295040271303e-06, "loss": 0.5774, "mean_token_accuracy": 0.8174028396606445, "num_tokens": 18646985.0, "step": 493 }, { "epoch": 0.06284187762371199, "ewc_loss": 0.0106201171875, "ewc_loss_diag": 7.748603820800781e-06, "ewc_loss_parallel": 9.834766387939453e-06, "grad_norm": 1.7104952335357666, "learning_rate": 2.0898685883849092e-06, "loss": 0.5559, "mean_token_accuracy": 0.8307191729545593, "num_tokens": 18688818.0, "step": 494 }, { "epoch": 0.0629690879023025, "ewc_loss": 0.01055908203125, "ewc_loss_diag": 7.748603820800781e-06, "ewc_loss_parallel": 9.775161743164062e-06, "grad_norm": 1.720839023590088, "learning_rate": 2.0941076727426877e-06, "loss": 0.5365, "mean_token_accuracy": 0.8344649076461792, "num_tokens": 18727208.0, "step": 495 }, { "epoch": 0.06309629818089302, "ewc_loss": 0.0106201171875, "ewc_loss_diag": 7.748603820800781e-06, "ewc_loss_parallel": 9.834766387939453e-06, "grad_norm": 1.6934553384780884, "learning_rate": 2.0983467571004666e-06, "loss": 0.5218, "mean_token_accuracy": 0.8362892270088196, "num_tokens": 18772174.0, "step": 496 }, { "epoch": 0.06322350845948353, "ewc_loss": 0.0106201171875, "ewc_loss_diag": 7.808208465576172e-06, "ewc_loss_parallel": 9.834766387939453e-06, "grad_norm": 1.7233617305755615, "learning_rate": 2.102585841458245e-06, "loss": 0.5564, "mean_token_accuracy": 0.8274662494659424, "num_tokens": 18808111.0, "step": 497 }, { "epoch": 0.06335071873807403, "ewc_loss": 0.0106201171875, "ewc_loss_diag": 7.808208465576172e-06, "ewc_loss_parallel": 9.834766387939453e-06, "grad_norm": 1.7312452793121338, "learning_rate": 2.106824925816024e-06, "loss": 0.5138, "mean_token_accuracy": 0.8330979347229004, "num_tokens": 18841912.0, "step": 498 }, { "epoch": 0.06347792901666455, "ewc_loss": 0.0106201171875, "ewc_loss_diag": 7.808208465576172e-06, "ewc_loss_parallel": 9.834766387939453e-06, "grad_norm": 1.7302945852279663, "learning_rate": 2.1110640101738024e-06, "loss": 0.5234, "mean_token_accuracy": 0.8381739258766174, "num_tokens": 18877694.0, "step": 499 }, { "epoch": 0.06360513929525506, "ewc_loss": 0.01068115234375, "ewc_loss_diag": 7.808208465576172e-06, "ewc_loss_parallel": 9.894371032714844e-06, "grad_norm": 1.7260195016860962, "learning_rate": 2.1153030945315813e-06, "loss": 0.5283, "mean_token_accuracy": 0.8346590995788574, "num_tokens": 18912253.0, "step": 500 }, { "epoch": 0.06373234957384556, "ewc_loss": 0.0106201171875, "ewc_loss_diag": 7.808208465576172e-06, "ewc_loss_parallel": 9.834766387939453e-06, "grad_norm": 1.7347629070281982, "learning_rate": 2.11954217888936e-06, "loss": 0.5296, "mean_token_accuracy": 0.8366900086402893, "num_tokens": 18945700.0, "step": 501 }, { "epoch": 0.06385955985243608, "ewc_loss": 0.0107421875, "ewc_loss_diag": 7.808208465576172e-06, "ewc_loss_parallel": 9.953975677490234e-06, "grad_norm": 1.7231634855270386, "learning_rate": 2.123781263247139e-06, "loss": 0.555, "mean_token_accuracy": 0.8262535929679871, "num_tokens": 18985260.0, "step": 502 }, { "epoch": 0.06398677013102659, "ewc_loss": 0.01068115234375, "ewc_loss_diag": 7.808208465576172e-06, "ewc_loss_parallel": 9.894371032714844e-06, "grad_norm": 1.7317954301834106, "learning_rate": 2.1280203476049175e-06, "loss": 0.6262, "mean_token_accuracy": 0.8024717569351196, "num_tokens": 19025364.0, "step": 503 }, { "epoch": 0.06411398040961709, "ewc_loss": 0.0107421875, "ewc_loss_diag": 7.867813110351562e-06, "ewc_loss_parallel": 9.953975677490234e-06, "grad_norm": 1.7257658243179321, "learning_rate": 2.1322594319626964e-06, "loss": 0.5613, "mean_token_accuracy": 0.8245534300804138, "num_tokens": 19064262.0, "step": 504 }, { "epoch": 0.0642411906882076, "ewc_loss": 0.01068115234375, "ewc_loss_diag": 7.867813110351562e-06, "ewc_loss_parallel": 9.894371032714844e-06, "grad_norm": 1.7802380323410034, "learning_rate": 2.136498516320475e-06, "loss": 0.5633, "mean_token_accuracy": 0.8267245292663574, "num_tokens": 19094780.0, "step": 505 }, { "epoch": 0.06436840096679812, "ewc_loss": 0.01080322265625, "ewc_loss_diag": 7.867813110351562e-06, "ewc_loss_parallel": 1.0013580322265625e-05, "grad_norm": 1.7353508472442627, "learning_rate": 2.1407376006782537e-06, "loss": 0.5012, "mean_token_accuracy": 0.8426322937011719, "num_tokens": 19131166.0, "step": 506 }, { "epoch": 0.06449561124538863, "ewc_loss": 0.01068115234375, "ewc_loss_diag": 7.867813110351562e-06, "ewc_loss_parallel": 9.894371032714844e-06, "grad_norm": 1.7337380647659302, "learning_rate": 2.144976685036032e-06, "loss": 0.5159, "mean_token_accuracy": 0.8403583765029907, "num_tokens": 19167371.0, "step": 507 }, { "epoch": 0.06462282152397913, "ewc_loss": 0.01080322265625, "ewc_loss_diag": 7.867813110351562e-06, "ewc_loss_parallel": 1.0013580322265625e-05, "grad_norm": 1.7203478813171387, "learning_rate": 2.149215769393811e-06, "loss": 0.5661, "mean_token_accuracy": 0.824158251285553, "num_tokens": 19206751.0, "step": 508 }, { "epoch": 0.06475003180256965, "ewc_loss": 0.0107421875, "ewc_loss_diag": 7.867813110351562e-06, "ewc_loss_parallel": 9.953975677490234e-06, "grad_norm": 1.7410016059875488, "learning_rate": 2.15345485375159e-06, "loss": 0.5213, "mean_token_accuracy": 0.8372386693954468, "num_tokens": 19243227.0, "step": 509 }, { "epoch": 0.06487724208116016, "ewc_loss": 0.0108642578125, "ewc_loss_diag": 7.927417755126953e-06, "ewc_loss_parallel": 1.0073184967041016e-05, "grad_norm": 1.739974021911621, "learning_rate": 2.157693938109369e-06, "loss": 0.5729, "mean_token_accuracy": 0.8248544335365295, "num_tokens": 19285050.0, "step": 510 }, { "epoch": 0.06500445235975066, "ewc_loss": 0.0107421875, "ewc_loss_diag": 7.927417755126953e-06, "ewc_loss_parallel": 9.953975677490234e-06, "grad_norm": 1.730662226676941, "learning_rate": 2.1619330224671473e-06, "loss": 0.558, "mean_token_accuracy": 0.8263748288154602, "num_tokens": 19328780.0, "step": 511 }, { "epoch": 0.06513166263834118, "ewc_loss": 0.01080322265625, "ewc_loss_diag": 7.927417755126953e-06, "ewc_loss_parallel": 1.0013580322265625e-05, "grad_norm": 1.7437187433242798, "learning_rate": 2.166172106824926e-06, "loss": 0.5541, "mean_token_accuracy": 0.8310911059379578, "num_tokens": 19373011.0, "step": 512 }, { "epoch": 0.06525887291693169, "ewc_loss": 0.01080322265625, "ewc_loss_diag": 7.927417755126953e-06, "ewc_loss_parallel": 1.0013580322265625e-05, "grad_norm": 1.7528526782989502, "learning_rate": 2.1704111911827046e-06, "loss": 0.5439, "mean_token_accuracy": 0.8305482864379883, "num_tokens": 19409641.0, "step": 513 }, { "epoch": 0.0653860831955222, "ewc_loss": 0.0108642578125, "ewc_loss_diag": 7.927417755126953e-06, "ewc_loss_parallel": 1.0073184967041016e-05, "grad_norm": 1.742845892906189, "learning_rate": 2.174650275540483e-06, "loss": 0.5856, "mean_token_accuracy": 0.8215302228927612, "num_tokens": 19451285.0, "step": 514 }, { "epoch": 0.06551329347411271, "ewc_loss": 0.01092529296875, "ewc_loss_diag": 7.927417755126953e-06, "ewc_loss_parallel": 1.0132789611816406e-05, "grad_norm": 1.7676957845687866, "learning_rate": 2.178889359898262e-06, "loss": 0.5821, "mean_token_accuracy": 0.8226733803749084, "num_tokens": 19487584.0, "step": 515 }, { "epoch": 0.06564050375270322, "ewc_loss": 0.01092529296875, "ewc_loss_diag": 7.987022399902344e-06, "ewc_loss_parallel": 1.0132789611816406e-05, "grad_norm": 1.7439464330673218, "learning_rate": 2.183128444256041e-06, "loss": 0.5917, "mean_token_accuracy": 0.8203883171081543, "num_tokens": 19531200.0, "step": 516 }, { "epoch": 0.06576771403129372, "ewc_loss": 0.0108642578125, "ewc_loss_diag": 7.987022399902344e-06, "ewc_loss_parallel": 1.0073184967041016e-05, "grad_norm": 1.7494102716445923, "learning_rate": 2.1873675286138197e-06, "loss": 0.5894, "mean_token_accuracy": 0.8145842552185059, "num_tokens": 19573126.0, "step": 517 }, { "epoch": 0.06589492430988424, "ewc_loss": 0.01092529296875, "ewc_loss_diag": 7.987022399902344e-06, "ewc_loss_parallel": 1.0132789611816406e-05, "grad_norm": 1.7473680973052979, "learning_rate": 2.191606612971598e-06, "loss": 0.6045, "mean_token_accuracy": 0.8125758171081543, "num_tokens": 19613499.0, "step": 518 }, { "epoch": 0.06602213458847475, "ewc_loss": 0.01092529296875, "ewc_loss_diag": 7.987022399902344e-06, "ewc_loss_parallel": 1.0132789611816406e-05, "grad_norm": 1.761278510093689, "learning_rate": 2.195845697329377e-06, "loss": 0.5942, "mean_token_accuracy": 0.8113102912902832, "num_tokens": 19652459.0, "step": 519 }, { "epoch": 0.06614934486706527, "ewc_loss": 0.010986328125, "ewc_loss_diag": 7.987022399902344e-06, "ewc_loss_parallel": 1.0192394256591797e-05, "grad_norm": 1.7549710273742676, "learning_rate": 2.2000847816871556e-06, "loss": 0.4751, "mean_token_accuracy": 0.8492400646209717, "num_tokens": 19691874.0, "step": 520 }, { "epoch": 0.06627655514565577, "ewc_loss": 0.010986328125, "ewc_loss_diag": 8.046627044677734e-06, "ewc_loss_parallel": 1.0192394256591797e-05, "grad_norm": 1.745896816253662, "learning_rate": 2.2043238660449344e-06, "loss": 0.5239, "mean_token_accuracy": 0.838848352432251, "num_tokens": 19731424.0, "step": 521 }, { "epoch": 0.06640376542424628, "ewc_loss": 0.010986328125, "ewc_loss_diag": 8.046627044677734e-06, "ewc_loss_parallel": 1.0192394256591797e-05, "grad_norm": 1.7572649717330933, "learning_rate": 2.2085629504027133e-06, "loss": 0.5179, "mean_token_accuracy": 0.8361971378326416, "num_tokens": 19768530.0, "step": 522 }, { "epoch": 0.0665309757028368, "ewc_loss": 0.010986328125, "ewc_loss_diag": 8.046627044677734e-06, "ewc_loss_parallel": 1.0192394256591797e-05, "grad_norm": 1.7893270254135132, "learning_rate": 2.2128020347604918e-06, "loss": 0.5515, "mean_token_accuracy": 0.8280661106109619, "num_tokens": 19807836.0, "step": 523 }, { "epoch": 0.0666581859814273, "ewc_loss": 0.01092529296875, "ewc_loss_diag": 8.046627044677734e-06, "ewc_loss_parallel": 1.0132789611816406e-05, "grad_norm": 1.7392500638961792, "learning_rate": 2.2170411191182707e-06, "loss": 0.5313, "mean_token_accuracy": 0.8334373831748962, "num_tokens": 19854338.0, "step": 524 }, { "epoch": 0.06678539626001781, "ewc_loss": 0.010986328125, "ewc_loss_diag": 8.046627044677734e-06, "ewc_loss_parallel": 1.0192394256591797e-05, "grad_norm": 1.7932777404785156, "learning_rate": 2.2212802034760496e-06, "loss": 0.55, "mean_token_accuracy": 0.8263866901397705, "num_tokens": 19888469.0, "step": 525 }, { "epoch": 0.06691260653860832, "ewc_loss": 0.01104736328125, "ewc_loss_diag": 8.046627044677734e-06, "ewc_loss_parallel": 1.0251998901367188e-05, "grad_norm": 1.809654712677002, "learning_rate": 2.225519287833828e-06, "loss": 0.6007, "mean_token_accuracy": 0.8170961141586304, "num_tokens": 19925137.0, "step": 526 }, { "epoch": 0.06703981681719882, "ewc_loss": 0.0111083984375, "ewc_loss_diag": 8.046627044677734e-06, "ewc_loss_parallel": 1.0311603546142578e-05, "grad_norm": 1.7812316417694092, "learning_rate": 2.229758372191607e-06, "loss": 0.5584, "mean_token_accuracy": 0.8264284133911133, "num_tokens": 19959658.0, "step": 527 }, { "epoch": 0.06716702709578934, "ewc_loss": 0.01104736328125, "ewc_loss_diag": 8.046627044677734e-06, "ewc_loss_parallel": 1.0251998901367188e-05, "grad_norm": 1.7614364624023438, "learning_rate": 2.2339974565493854e-06, "loss": 0.5163, "mean_token_accuracy": 0.8383554220199585, "num_tokens": 19998702.0, "step": 528 }, { "epoch": 0.06729423737437985, "ewc_loss": 0.01104736328125, "ewc_loss_diag": 8.046627044677734e-06, "ewc_loss_parallel": 1.0251998901367188e-05, "grad_norm": 1.7858178615570068, "learning_rate": 2.2382365409071642e-06, "loss": 0.5878, "mean_token_accuracy": 0.8214136362075806, "num_tokens": 20038406.0, "step": 529 }, { "epoch": 0.06742144765297035, "ewc_loss": 0.0111083984375, "ewc_loss_diag": 8.046627044677734e-06, "ewc_loss_parallel": 1.0311603546142578e-05, "grad_norm": 1.781916856765747, "learning_rate": 2.242475625264943e-06, "loss": 0.5434, "mean_token_accuracy": 0.8290829062461853, "num_tokens": 20073185.0, "step": 530 }, { "epoch": 0.06754865793156087, "ewc_loss": 0.01116943359375, "ewc_loss_diag": 8.046627044677734e-06, "ewc_loss_parallel": 1.0371208190917969e-05, "grad_norm": 1.7867908477783203, "learning_rate": 2.246714709622722e-06, "loss": 0.5251, "mean_token_accuracy": 0.8376836776733398, "num_tokens": 20106828.0, "step": 531 }, { "epoch": 0.06767586821015138, "ewc_loss": 0.01116943359375, "ewc_loss_diag": 8.046627044677734e-06, "ewc_loss_parallel": 1.0371208190917969e-05, "grad_norm": 1.8070300817489624, "learning_rate": 2.2509537939805005e-06, "loss": 0.5277, "mean_token_accuracy": 0.8335241079330444, "num_tokens": 20137541.0, "step": 532 }, { "epoch": 0.0678030784887419, "ewc_loss": 0.01123046875, "ewc_loss_diag": 8.046627044677734e-06, "ewc_loss_parallel": 1.043081283569336e-05, "grad_norm": 1.7747173309326172, "learning_rate": 2.255192878338279e-06, "loss": 0.5398, "mean_token_accuracy": 0.8298099040985107, "num_tokens": 20174452.0, "step": 533 }, { "epoch": 0.0679302887673324, "ewc_loss": 0.01116943359375, "ewc_loss_diag": 8.046627044677734e-06, "ewc_loss_parallel": 1.0371208190917969e-05, "grad_norm": 1.7844892740249634, "learning_rate": 2.259431962696058e-06, "loss": 0.5565, "mean_token_accuracy": 0.8299411535263062, "num_tokens": 20212195.0, "step": 534 }, { "epoch": 0.06805749904592291, "ewc_loss": 0.01129150390625, "ewc_loss_diag": 8.046627044677734e-06, "ewc_loss_parallel": 1.049041748046875e-05, "grad_norm": 1.7753596305847168, "learning_rate": 2.2636710470538363e-06, "loss": 0.4698, "mean_token_accuracy": 0.8455036282539368, "num_tokens": 20248344.0, "step": 535 }, { "epoch": 0.06818470932451343, "ewc_loss": 0.01129150390625, "ewc_loss_diag": 8.046627044677734e-06, "ewc_loss_parallel": 1.049041748046875e-05, "grad_norm": 1.7697433233261108, "learning_rate": 2.267910131411615e-06, "loss": 0.5155, "mean_token_accuracy": 0.8389238715171814, "num_tokens": 20289874.0, "step": 536 }, { "epoch": 0.06831191960310393, "ewc_loss": 0.01129150390625, "ewc_loss_diag": 8.106231689453125e-06, "ewc_loss_parallel": 1.049041748046875e-05, "grad_norm": 1.8248367309570312, "learning_rate": 2.272149215769394e-06, "loss": 0.5264, "mean_token_accuracy": 0.8327659368515015, "num_tokens": 20319582.0, "step": 537 }, { "epoch": 0.06843912988169444, "ewc_loss": 0.01141357421875, "ewc_loss_diag": 8.106231689453125e-06, "ewc_loss_parallel": 1.0609626770019531e-05, "grad_norm": 1.778839111328125, "learning_rate": 2.276388300127173e-06, "loss": 0.5047, "mean_token_accuracy": 0.8383667469024658, "num_tokens": 20356944.0, "step": 538 }, { "epoch": 0.06856634016028496, "ewc_loss": 0.01141357421875, "ewc_loss_diag": 8.106231689453125e-06, "ewc_loss_parallel": 1.0609626770019531e-05, "grad_norm": 1.7781624794006348, "learning_rate": 2.2806273844849514e-06, "loss": 0.5582, "mean_token_accuracy": 0.8229185938835144, "num_tokens": 20398099.0, "step": 539 }, { "epoch": 0.06869355043887546, "ewc_loss": 0.0113525390625, "ewc_loss_diag": 8.106231689453125e-06, "ewc_loss_parallel": 1.055002212524414e-05, "grad_norm": 1.8048856258392334, "learning_rate": 2.2848664688427303e-06, "loss": 0.5293, "mean_token_accuracy": 0.8321362733840942, "num_tokens": 20435155.0, "step": 540 }, { "epoch": 0.06882076071746597, "ewc_loss": 0.0115966796875, "ewc_loss_diag": 8.106231689453125e-06, "ewc_loss_parallel": 1.0788440704345703e-05, "grad_norm": 1.8339921236038208, "learning_rate": 2.2891055532005087e-06, "loss": 0.5878, "mean_token_accuracy": 0.8187752962112427, "num_tokens": 20468369.0, "step": 541 }, { "epoch": 0.06894797099605648, "ewc_loss": 0.01165771484375, "ewc_loss_diag": 8.106231689453125e-06, "ewc_loss_parallel": 1.0848045349121094e-05, "grad_norm": 1.8290594816207886, "learning_rate": 2.2933446375582876e-06, "loss": 0.5642, "mean_token_accuracy": 0.8272970914840698, "num_tokens": 20500303.0, "step": 542 }, { "epoch": 0.06907518127464699, "ewc_loss": 0.0115966796875, "ewc_loss_diag": 8.165836334228516e-06, "ewc_loss_parallel": 1.0788440704345703e-05, "grad_norm": 1.7880593538284302, "learning_rate": 2.297583721916066e-06, "loss": 0.5157, "mean_token_accuracy": 0.8406131267547607, "num_tokens": 20541290.0, "step": 543 }, { "epoch": 0.0692023915532375, "ewc_loss": 0.01165771484375, "ewc_loss_diag": 8.165836334228516e-06, "ewc_loss_parallel": 1.0848045349121094e-05, "grad_norm": 1.7888959646224976, "learning_rate": 2.301822806273845e-06, "loss": 0.5356, "mean_token_accuracy": 0.8342598676681519, "num_tokens": 20583486.0, "step": 544 }, { "epoch": 0.06932960183182801, "ewc_loss": 0.01177978515625, "ewc_loss_diag": 8.165836334228516e-06, "ewc_loss_parallel": 1.0967254638671875e-05, "grad_norm": 1.8279106616973877, "learning_rate": 2.306061890631624e-06, "loss": 0.5882, "mean_token_accuracy": 0.8216540813446045, "num_tokens": 20618655.0, "step": 545 }, { "epoch": 0.06945681211041853, "ewc_loss": 0.01165771484375, "ewc_loss_diag": 8.165836334228516e-06, "ewc_loss_parallel": 1.0848045349121094e-05, "grad_norm": 1.8016777038574219, "learning_rate": 2.3103009749894027e-06, "loss": 0.5402, "mean_token_accuracy": 0.8322706818580627, "num_tokens": 20655733.0, "step": 546 }, { "epoch": 0.06958402238900903, "ewc_loss": 0.01171875, "ewc_loss_diag": 8.165836334228516e-06, "ewc_loss_parallel": 1.0907649993896484e-05, "grad_norm": 1.7869335412979126, "learning_rate": 2.314540059347181e-06, "loss": 0.5689, "mean_token_accuracy": 0.8243540525436401, "num_tokens": 20699338.0, "step": 547 }, { "epoch": 0.06971123266759954, "ewc_loss": 0.01177978515625, "ewc_loss_diag": 8.165836334228516e-06, "ewc_loss_parallel": 1.0967254638671875e-05, "grad_norm": 1.8916773796081543, "learning_rate": 2.31877914370496e-06, "loss": 0.5948, "mean_token_accuracy": 0.8149627447128296, "num_tokens": 20731841.0, "step": 548 }, { "epoch": 0.06983844294619006, "ewc_loss": 0.01202392578125, "ewc_loss_diag": 8.285045623779297e-06, "ewc_loss_parallel": 1.1146068572998047e-05, "grad_norm": 1.8428336381912231, "learning_rate": 2.3230182280627385e-06, "loss": 0.5159, "mean_token_accuracy": 0.8404078483581543, "num_tokens": 20761097.0, "step": 549 }, { "epoch": 0.06996565322478056, "ewc_loss": 0.01190185546875, "ewc_loss_diag": 8.285045623779297e-06, "ewc_loss_parallel": 1.1026859283447266e-05, "grad_norm": 1.7955352067947388, "learning_rate": 2.3272573124205174e-06, "loss": 0.4903, "mean_token_accuracy": 0.8454477787017822, "num_tokens": 20797013.0, "step": 550 }, { "epoch": 0.07009286350337107, "ewc_loss": 0.01202392578125, "ewc_loss_diag": 8.285045623779297e-06, "ewc_loss_parallel": 1.1146068572998047e-05, "grad_norm": 1.833472728729248, "learning_rate": 2.331496396778296e-06, "loss": 0.5265, "mean_token_accuracy": 0.8320705890655518, "num_tokens": 20833727.0, "step": 551 }, { "epoch": 0.07022007378196159, "ewc_loss": 0.01202392578125, "ewc_loss_diag": 8.285045623779297e-06, "ewc_loss_parallel": 1.1146068572998047e-05, "grad_norm": 1.7889959812164307, "learning_rate": 2.3357354811360748e-06, "loss": 0.5559, "mean_token_accuracy": 0.8295693397521973, "num_tokens": 20877587.0, "step": 552 }, { "epoch": 0.07034728406055209, "ewc_loss": 0.01202392578125, "ewc_loss_diag": 8.285045623779297e-06, "ewc_loss_parallel": 1.1146068572998047e-05, "grad_norm": 1.8500176668167114, "learning_rate": 2.3399745654938537e-06, "loss": 0.5643, "mean_token_accuracy": 0.8242551684379578, "num_tokens": 20909623.0, "step": 553 }, { "epoch": 0.0704744943391426, "ewc_loss": 0.0120849609375, "ewc_loss_diag": 8.285045623779297e-06, "ewc_loss_parallel": 1.1205673217773438e-05, "grad_norm": 1.823438048362732, "learning_rate": 2.344213649851632e-06, "loss": 0.5457, "mean_token_accuracy": 0.831454873085022, "num_tokens": 20946348.0, "step": 554 }, { "epoch": 0.07060170461773312, "ewc_loss": 0.01214599609375, "ewc_loss_diag": 8.285045623779297e-06, "ewc_loss_parallel": 1.1265277862548828e-05, "grad_norm": 1.819177508354187, "learning_rate": 2.348452734209411e-06, "loss": 0.6271, "mean_token_accuracy": 0.8093889355659485, "num_tokens": 20989831.0, "step": 555 }, { "epoch": 0.07072891489632362, "ewc_loss": 0.0120849609375, "ewc_loss_diag": 8.404254913330078e-06, "ewc_loss_parallel": 1.1205673217773438e-05, "grad_norm": 1.8346985578536987, "learning_rate": 2.3526918185671895e-06, "loss": 0.5751, "mean_token_accuracy": 0.8200093507766724, "num_tokens": 21027007.0, "step": 556 }, { "epoch": 0.07085612517491413, "ewc_loss": 0.01220703125, "ewc_loss_diag": 8.404254913330078e-06, "ewc_loss_parallel": 1.1324882507324219e-05, "grad_norm": 1.8129620552062988, "learning_rate": 2.3569309029249684e-06, "loss": 0.5242, "mean_token_accuracy": 0.8360450863838196, "num_tokens": 21066488.0, "step": 557 }, { "epoch": 0.07098333545350465, "ewc_loss": 0.01214599609375, "ewc_loss_diag": 8.404254913330078e-06, "ewc_loss_parallel": 1.1265277862548828e-05, "grad_norm": 1.8272308111190796, "learning_rate": 2.3611699872827472e-06, "loss": 0.6137, "mean_token_accuracy": 0.8112305402755737, "num_tokens": 21109049.0, "step": 558 }, { "epoch": 0.07111054573209516, "ewc_loss": 0.01226806640625, "ewc_loss_diag": 8.404254913330078e-06, "ewc_loss_parallel": 1.138448715209961e-05, "grad_norm": 1.837886095046997, "learning_rate": 2.3654090716405257e-06, "loss": 0.5561, "mean_token_accuracy": 0.8266865611076355, "num_tokens": 21146661.0, "step": 559 }, { "epoch": 0.07123775601068566, "ewc_loss": 0.01220703125, "ewc_loss_diag": 8.404254913330078e-06, "ewc_loss_parallel": 1.1324882507324219e-05, "grad_norm": 1.849815011024475, "learning_rate": 2.3696481559983046e-06, "loss": 0.5735, "mean_token_accuracy": 0.8195052146911621, "num_tokens": 21179568.0, "step": 560 }, { "epoch": 0.07136496628927617, "ewc_loss": 0.01226806640625, "ewc_loss_diag": 8.404254913330078e-06, "ewc_loss_parallel": 1.138448715209961e-05, "grad_norm": 1.8462623357772827, "learning_rate": 2.3738872403560835e-06, "loss": 0.5755, "mean_token_accuracy": 0.8209528923034668, "num_tokens": 21216535.0, "step": 561 }, { "epoch": 0.07149217656786669, "ewc_loss": 0.01226806640625, "ewc_loss_diag": 8.463859558105469e-06, "ewc_loss_parallel": 1.138448715209961e-05, "grad_norm": 1.8378881216049194, "learning_rate": 2.378126324713862e-06, "loss": 0.5372, "mean_token_accuracy": 0.8304297924041748, "num_tokens": 21253887.0, "step": 562 }, { "epoch": 0.07161938684645719, "ewc_loss": 0.01226806640625, "ewc_loss_diag": 8.463859558105469e-06, "ewc_loss_parallel": 1.138448715209961e-05, "grad_norm": 1.8394081592559814, "learning_rate": 2.382365409071641e-06, "loss": 0.5409, "mean_token_accuracy": 0.8317009806632996, "num_tokens": 21289813.0, "step": 563 }, { "epoch": 0.0717465971250477, "ewc_loss": 0.01226806640625, "ewc_loss_diag": 8.463859558105469e-06, "ewc_loss_parallel": 1.138448715209961e-05, "grad_norm": 1.8393532037734985, "learning_rate": 2.3866044934294193e-06, "loss": 0.4935, "mean_token_accuracy": 0.8441318273544312, "num_tokens": 21326121.0, "step": 564 }, { "epoch": 0.07187380740363822, "ewc_loss": 0.0123291015625, "ewc_loss_diag": 8.463859558105469e-06, "ewc_loss_parallel": 1.1444091796875e-05, "grad_norm": 1.8482847213745117, "learning_rate": 2.390843577787198e-06, "loss": 0.5878, "mean_token_accuracy": 0.8162662982940674, "num_tokens": 21363644.0, "step": 565 }, { "epoch": 0.07200101768222872, "ewc_loss": 0.0123291015625, "ewc_loss_diag": 8.463859558105469e-06, "ewc_loss_parallel": 1.1444091796875e-05, "grad_norm": 1.8090859651565552, "learning_rate": 2.395082662144977e-06, "loss": 0.473, "mean_token_accuracy": 0.8525142669677734, "num_tokens": 21407119.0, "step": 566 }, { "epoch": 0.07212822796081923, "ewc_loss": 0.01226806640625, "ewc_loss_diag": 8.52346420288086e-06, "ewc_loss_parallel": 1.138448715209961e-05, "grad_norm": 1.8396580219268799, "learning_rate": 2.399321746502756e-06, "loss": 0.5479, "mean_token_accuracy": 0.8305277228355408, "num_tokens": 21447648.0, "step": 567 }, { "epoch": 0.07225543823940975, "ewc_loss": 0.0123291015625, "ewc_loss_diag": 8.52346420288086e-06, "ewc_loss_parallel": 1.1444091796875e-05, "grad_norm": 1.8513542413711548, "learning_rate": 2.4035608308605344e-06, "loss": 0.5906, "mean_token_accuracy": 0.8246682286262512, "num_tokens": 21488335.0, "step": 568 }, { "epoch": 0.07238264851800025, "ewc_loss": 0.0123291015625, "ewc_loss_diag": 8.58306884765625e-06, "ewc_loss_parallel": 1.150369644165039e-05, "grad_norm": 1.8664665222167969, "learning_rate": 2.407799915218313e-06, "loss": 0.5924, "mean_token_accuracy": 0.8135747909545898, "num_tokens": 21525987.0, "step": 569 }, { "epoch": 0.07250985879659076, "ewc_loss": 0.0123291015625, "ewc_loss_diag": 8.58306884765625e-06, "ewc_loss_parallel": 1.1444091796875e-05, "grad_norm": 1.8281753063201904, "learning_rate": 2.4120389995760917e-06, "loss": 0.4829, "mean_token_accuracy": 0.8521325588226318, "num_tokens": 21564671.0, "step": 570 }, { "epoch": 0.07263706907518128, "ewc_loss": 0.0123291015625, "ewc_loss_diag": 8.58306884765625e-06, "ewc_loss_parallel": 1.1444091796875e-05, "grad_norm": 1.8386610746383667, "learning_rate": 2.41627808393387e-06, "loss": 0.5302, "mean_token_accuracy": 0.8323889374732971, "num_tokens": 21608590.0, "step": 571 }, { "epoch": 0.07276427935377179, "ewc_loss": 0.0123291015625, "ewc_loss_diag": 8.58306884765625e-06, "ewc_loss_parallel": 1.150369644165039e-05, "grad_norm": 1.8744250535964966, "learning_rate": 2.420517168291649e-06, "loss": 0.5512, "mean_token_accuracy": 0.831762969493866, "num_tokens": 21641684.0, "step": 572 }, { "epoch": 0.07289148963236229, "ewc_loss": 0.01239013671875, "ewc_loss_diag": 8.58306884765625e-06, "ewc_loss_parallel": 1.1563301086425781e-05, "grad_norm": 1.8518282175064087, "learning_rate": 2.424756252649428e-06, "loss": 0.4927, "mean_token_accuracy": 0.8442918658256531, "num_tokens": 21674923.0, "step": 573 }, { "epoch": 0.0730186999109528, "ewc_loss": 0.01239013671875, "ewc_loss_diag": 8.58306884765625e-06, "ewc_loss_parallel": 1.1563301086425781e-05, "grad_norm": 1.8590335845947266, "learning_rate": 2.428995337007207e-06, "loss": 0.5775, "mean_token_accuracy": 0.8236913084983826, "num_tokens": 21714006.0, "step": 574 }, { "epoch": 0.07314591018954332, "ewc_loss": 0.012451171875, "ewc_loss_diag": 8.58306884765625e-06, "ewc_loss_parallel": 1.1622905731201172e-05, "grad_norm": 1.8664014339447021, "learning_rate": 2.4332344213649853e-06, "loss": 0.5298, "mean_token_accuracy": 0.8396324515342712, "num_tokens": 21750580.0, "step": 575 }, { "epoch": 0.07327312046813382, "ewc_loss": 0.01239013671875, "ewc_loss_diag": 8.58306884765625e-06, "ewc_loss_parallel": 1.1563301086425781e-05, "grad_norm": 1.8542795181274414, "learning_rate": 2.437473505722764e-06, "loss": 0.5037, "mean_token_accuracy": 0.8411420583724976, "num_tokens": 21787400.0, "step": 576 }, { "epoch": 0.07340033074672433, "ewc_loss": 0.0123291015625, "ewc_loss_diag": 8.58306884765625e-06, "ewc_loss_parallel": 1.150369644165039e-05, "grad_norm": 1.8648419380187988, "learning_rate": 2.4417125900805427e-06, "loss": 0.5197, "mean_token_accuracy": 0.8320856690406799, "num_tokens": 21823159.0, "step": 577 }, { "epoch": 0.07352754102531485, "ewc_loss": 0.01239013671875, "ewc_loss_diag": 8.64267349243164e-06, "ewc_loss_parallel": 1.1563301086425781e-05, "grad_norm": 1.8704184293746948, "learning_rate": 2.4459516744383215e-06, "loss": 0.559, "mean_token_accuracy": 0.8270680904388428, "num_tokens": 21858067.0, "step": 578 }, { "epoch": 0.07365475130390535, "ewc_loss": 0.01251220703125, "ewc_loss_diag": 8.64267349243164e-06, "ewc_loss_parallel": 1.1682510375976562e-05, "grad_norm": 1.8629844188690186, "learning_rate": 2.4501907587961e-06, "loss": 0.5187, "mean_token_accuracy": 0.8362158536911011, "num_tokens": 21893332.0, "step": 579 }, { "epoch": 0.07378196158249586, "ewc_loss": 0.012451171875, "ewc_loss_diag": 8.64267349243164e-06, "ewc_loss_parallel": 1.1622905731201172e-05, "grad_norm": 1.8717658519744873, "learning_rate": 2.454429843153879e-06, "loss": 0.6308, "mean_token_accuracy": 0.8052562475204468, "num_tokens": 21932696.0, "step": 580 }, { "epoch": 0.07390917186108638, "ewc_loss": 0.01251220703125, "ewc_loss_diag": 8.64267349243164e-06, "ewc_loss_parallel": 1.1682510375976562e-05, "grad_norm": 1.887353539466858, "learning_rate": 2.4586689275116578e-06, "loss": 0.5426, "mean_token_accuracy": 0.8252407312393188, "num_tokens": 21964825.0, "step": 581 }, { "epoch": 0.07403638213967688, "ewc_loss": 0.0125732421875, "ewc_loss_diag": 8.702278137207031e-06, "ewc_loss_parallel": 1.1742115020751953e-05, "grad_norm": 1.861608862876892, "learning_rate": 2.4629080118694367e-06, "loss": 0.5052, "mean_token_accuracy": 0.8413678407669067, "num_tokens": 22003396.0, "step": 582 }, { "epoch": 0.0741635924182674, "ewc_loss": 0.012451171875, "ewc_loss_diag": 8.702278137207031e-06, "ewc_loss_parallel": 1.1622905731201172e-05, "grad_norm": 1.9385939836502075, "learning_rate": 2.467147096227215e-06, "loss": 0.512, "mean_token_accuracy": 0.8410210609436035, "num_tokens": 22045107.0, "step": 583 }, { "epoch": 0.07429080269685791, "ewc_loss": 0.0126953125, "ewc_loss_diag": 8.64267349243164e-06, "ewc_loss_parallel": 1.1861324310302734e-05, "grad_norm": 1.849655032157898, "learning_rate": 2.471386180584994e-06, "loss": 0.487, "mean_token_accuracy": 0.847287118434906, "num_tokens": 22082754.0, "step": 584 }, { "epoch": 0.07441801297544842, "ewc_loss": 0.01251220703125, "ewc_loss_diag": 8.761882781982422e-06, "ewc_loss_parallel": 1.1682510375976562e-05, "grad_norm": 1.867645263671875, "learning_rate": 2.4756252649427725e-06, "loss": 0.5068, "mean_token_accuracy": 0.8401470184326172, "num_tokens": 22122501.0, "step": 585 }, { "epoch": 0.07454522325403892, "ewc_loss": 0.0126953125, "ewc_loss_diag": 8.761882781982422e-06, "ewc_loss_parallel": 1.1861324310302734e-05, "grad_norm": 1.871220350265503, "learning_rate": 2.4798643493005513e-06, "loss": 0.5443, "mean_token_accuracy": 0.8321726322174072, "num_tokens": 22158110.0, "step": 586 }, { "epoch": 0.07467243353262944, "ewc_loss": 0.0125732421875, "ewc_loss_diag": 8.702278137207031e-06, "ewc_loss_parallel": 1.1742115020751953e-05, "grad_norm": 1.8924626111984253, "learning_rate": 2.48410343365833e-06, "loss": 0.494, "mean_token_accuracy": 0.8386285901069641, "num_tokens": 22190097.0, "step": 587 }, { "epoch": 0.07479964381121995, "ewc_loss": 0.0126953125, "ewc_loss_diag": 8.702278137207031e-06, "ewc_loss_parallel": 1.1861324310302734e-05, "grad_norm": 1.8878356218338013, "learning_rate": 2.4883425180161087e-06, "loss": 0.5706, "mean_token_accuracy": 0.8229561448097229, "num_tokens": 22227568.0, "step": 588 }, { "epoch": 0.07492685408981045, "ewc_loss": 0.0126953125, "ewc_loss_diag": 8.761882781982422e-06, "ewc_loss_parallel": 1.1861324310302734e-05, "grad_norm": 1.8961269855499268, "learning_rate": 2.4925816023738876e-06, "loss": 0.5323, "mean_token_accuracy": 0.8371944427490234, "num_tokens": 22265376.0, "step": 589 }, { "epoch": 0.07505406436840097, "ewc_loss": 0.0126953125, "ewc_loss_diag": 8.761882781982422e-06, "ewc_loss_parallel": 1.1861324310302734e-05, "grad_norm": 1.866267442703247, "learning_rate": 2.496820686731666e-06, "loss": 0.5555, "mean_token_accuracy": 0.8294990062713623, "num_tokens": 22310398.0, "step": 590 }, { "epoch": 0.07518127464699148, "ewc_loss": 0.0128173828125, "ewc_loss_diag": 8.761882781982422e-06, "ewc_loss_parallel": 1.1980533599853516e-05, "grad_norm": 1.8988194465637207, "learning_rate": 2.5010597710894445e-06, "loss": 0.4602, "mean_token_accuracy": 0.8554180264472961, "num_tokens": 22341413.0, "step": 591 }, { "epoch": 0.07530848492558198, "ewc_loss": 0.01300048828125, "ewc_loss_diag": 8.881092071533203e-06, "ewc_loss_parallel": 1.2099742889404297e-05, "grad_norm": 1.8768290281295776, "learning_rate": 2.5052988554472234e-06, "loss": 0.5362, "mean_token_accuracy": 0.833762526512146, "num_tokens": 22383228.0, "step": 592 }, { "epoch": 0.0754356952041725, "ewc_loss": 0.01287841796875, "ewc_loss_diag": 8.881092071533203e-06, "ewc_loss_parallel": 1.1980533599853516e-05, "grad_norm": 1.8948726654052734, "learning_rate": 2.5095379398050023e-06, "loss": 0.5557, "mean_token_accuracy": 0.8284752368927002, "num_tokens": 22423750.0, "step": 593 }, { "epoch": 0.07556290548276301, "ewc_loss": 0.01300048828125, "ewc_loss_diag": 8.881092071533203e-06, "ewc_loss_parallel": 1.2099742889404297e-05, "grad_norm": 1.9131509065628052, "learning_rate": 2.513777024162781e-06, "loss": 0.5121, "mean_token_accuracy": 0.8395524621009827, "num_tokens": 22454717.0, "step": 594 }, { "epoch": 0.07569011576135352, "ewc_loss": 0.0130615234375, "ewc_loss_diag": 8.940696716308594e-06, "ewc_loss_parallel": 1.2159347534179688e-05, "grad_norm": 1.8792885541915894, "learning_rate": 2.5180161085205596e-06, "loss": 0.5112, "mean_token_accuracy": 0.8411707878112793, "num_tokens": 22492798.0, "step": 595 }, { "epoch": 0.07581732603994402, "ewc_loss": 0.012939453125, "ewc_loss_diag": 8.940696716308594e-06, "ewc_loss_parallel": 1.2040138244628906e-05, "grad_norm": 1.896647334098816, "learning_rate": 2.5222551928783385e-06, "loss": 0.5552, "mean_token_accuracy": 0.8286828398704529, "num_tokens": 22530774.0, "step": 596 }, { "epoch": 0.07594453631853454, "ewc_loss": 0.0130615234375, "ewc_loss_diag": 8.940696716308594e-06, "ewc_loss_parallel": 1.2159347534179688e-05, "grad_norm": 1.893966555595398, "learning_rate": 2.5264942772361174e-06, "loss": 0.5352, "mean_token_accuracy": 0.8330258131027222, "num_tokens": 22566355.0, "step": 597 }, { "epoch": 0.07607174659712505, "ewc_loss": 0.0130615234375, "ewc_loss_diag": 9.000301361083984e-06, "ewc_loss_parallel": 1.2159347534179688e-05, "grad_norm": 1.8944441080093384, "learning_rate": 2.5307333615938963e-06, "loss": 0.5203, "mean_token_accuracy": 0.8402464985847473, "num_tokens": 22603633.0, "step": 598 }, { "epoch": 0.07619895687571555, "ewc_loss": 0.01312255859375, "ewc_loss_diag": 9.000301361083984e-06, "ewc_loss_parallel": 1.2218952178955078e-05, "grad_norm": 1.9033974409103394, "learning_rate": 2.5349724459516743e-06, "loss": 0.5489, "mean_token_accuracy": 0.8316518068313599, "num_tokens": 22641481.0, "step": 599 }, { "epoch": 0.07632616715430607, "ewc_loss": 0.0130615234375, "ewc_loss_diag": 9.000301361083984e-06, "ewc_loss_parallel": 1.2159347534179688e-05, "grad_norm": 1.9118037223815918, "learning_rate": 2.539211530309453e-06, "loss": 0.4924, "mean_token_accuracy": 0.8448613882064819, "num_tokens": 22672853.0, "step": 600 }, { "epoch": 0.07645337743289658, "ewc_loss": 0.01318359375, "ewc_loss_diag": 9.000301361083984e-06, "ewc_loss_parallel": 1.2278556823730469e-05, "grad_norm": 1.9203553199768066, "learning_rate": 2.543450614667232e-06, "loss": 0.5666, "mean_token_accuracy": 0.8217101097106934, "num_tokens": 22704851.0, "step": 601 }, { "epoch": 0.07658058771148708, "ewc_loss": 0.01318359375, "ewc_loss_diag": 9.000301361083984e-06, "ewc_loss_parallel": 1.2278556823730469e-05, "grad_norm": 1.916445016860962, "learning_rate": 2.547689699025011e-06, "loss": 0.5846, "mean_token_accuracy": 0.8183513283729553, "num_tokens": 22742578.0, "step": 602 }, { "epoch": 0.0767077979900776, "ewc_loss": 0.01318359375, "ewc_loss_diag": 9.000301361083984e-06, "ewc_loss_parallel": 1.2278556823730469e-05, "grad_norm": 1.8827977180480957, "learning_rate": 2.5519287833827894e-06, "loss": 0.5267, "mean_token_accuracy": 0.8316076397895813, "num_tokens": 22783960.0, "step": 603 }, { "epoch": 0.07683500826866811, "ewc_loss": 0.01324462890625, "ewc_loss_diag": 9.119510650634766e-06, "ewc_loss_parallel": 1.233816146850586e-05, "grad_norm": 1.9310612678527832, "learning_rate": 2.5561678677405683e-06, "loss": 0.4894, "mean_token_accuracy": 0.8451162576675415, "num_tokens": 22813643.0, "step": 604 }, { "epoch": 0.07696221854725861, "ewc_loss": 0.01324462890625, "ewc_loss_diag": 9.119510650634766e-06, "ewc_loss_parallel": 1.233816146850586e-05, "grad_norm": 1.9094178676605225, "learning_rate": 2.560406952098347e-06, "loss": 0.5791, "mean_token_accuracy": 0.8178250789642334, "num_tokens": 22850290.0, "step": 605 }, { "epoch": 0.07708942882584913, "ewc_loss": 0.0133056640625, "ewc_loss_diag": 9.119510650634766e-06, "ewc_loss_parallel": 1.239776611328125e-05, "grad_norm": 1.8995434045791626, "learning_rate": 2.5646460364561252e-06, "loss": 0.5325, "mean_token_accuracy": 0.8312044143676758, "num_tokens": 22894009.0, "step": 606 }, { "epoch": 0.07721663910443964, "ewc_loss": 0.01336669921875, "ewc_loss_diag": 9.119510650634766e-06, "ewc_loss_parallel": 1.245737075805664e-05, "grad_norm": 1.9268516302108765, "learning_rate": 2.568885120813904e-06, "loss": 0.5342, "mean_token_accuracy": 0.8325688242912292, "num_tokens": 22931418.0, "step": 607 }, { "epoch": 0.07734384938303016, "ewc_loss": 0.01324462890625, "ewc_loss_diag": 9.119510650634766e-06, "ewc_loss_parallel": 1.233816146850586e-05, "grad_norm": 1.9053118228912354, "learning_rate": 2.573124205171683e-06, "loss": 0.5226, "mean_token_accuracy": 0.839162290096283, "num_tokens": 22970074.0, "step": 608 }, { "epoch": 0.07747105966162066, "ewc_loss": 0.0133056640625, "ewc_loss_diag": 9.179115295410156e-06, "ewc_loss_parallel": 1.239776611328125e-05, "grad_norm": 1.9391281604766846, "learning_rate": 2.577363289529462e-06, "loss": 0.5563, "mean_token_accuracy": 0.8297240734100342, "num_tokens": 23008185.0, "step": 609 }, { "epoch": 0.07759826994021117, "ewc_loss": 0.01336669921875, "ewc_loss_diag": 9.238719940185547e-06, "ewc_loss_parallel": 1.245737075805664e-05, "grad_norm": 1.91072416305542, "learning_rate": 2.5816023738872403e-06, "loss": 0.4747, "mean_token_accuracy": 0.8494632244110107, "num_tokens": 23039667.0, "step": 610 }, { "epoch": 0.07772548021880168, "ewc_loss": 0.0133056640625, "ewc_loss_diag": 9.238719940185547e-06, "ewc_loss_parallel": 1.239776611328125e-05, "grad_norm": 1.9329978227615356, "learning_rate": 2.5858414582450192e-06, "loss": 0.5431, "mean_token_accuracy": 0.8319372534751892, "num_tokens": 23075012.0, "step": 611 }, { "epoch": 0.07785269049739219, "ewc_loss": 0.01336669921875, "ewc_loss_diag": 9.238719940185547e-06, "ewc_loss_parallel": 1.245737075805664e-05, "grad_norm": 1.9184026718139648, "learning_rate": 2.590080542602798e-06, "loss": 0.4747, "mean_token_accuracy": 0.8500005006790161, "num_tokens": 23108397.0, "step": 612 }, { "epoch": 0.0779799007759827, "ewc_loss": 0.0133056640625, "ewc_loss_diag": 9.238719940185547e-06, "ewc_loss_parallel": 1.239776611328125e-05, "grad_norm": 1.9253478050231934, "learning_rate": 2.594319626960577e-06, "loss": 0.5518, "mean_token_accuracy": 0.8331552743911743, "num_tokens": 23146419.0, "step": 613 }, { "epoch": 0.07810711105457321, "ewc_loss": 0.01336669921875, "ewc_loss_diag": 9.238719940185547e-06, "ewc_loss_parallel": 1.245737075805664e-05, "grad_norm": 1.8956995010375977, "learning_rate": 2.598558711318355e-06, "loss": 0.5454, "mean_token_accuracy": 0.831521213054657, "num_tokens": 23191750.0, "step": 614 }, { "epoch": 0.07823432133316371, "ewc_loss": 0.01336669921875, "ewc_loss_diag": 9.238719940185547e-06, "ewc_loss_parallel": 1.245737075805664e-05, "grad_norm": 1.9380989074707031, "learning_rate": 2.602797795676134e-06, "loss": 0.5562, "mean_token_accuracy": 0.8277143239974976, "num_tokens": 23230199.0, "step": 615 }, { "epoch": 0.07836153161175423, "ewc_loss": 0.013427734375, "ewc_loss_diag": 9.298324584960938e-06, "ewc_loss_parallel": 1.2516975402832031e-05, "grad_norm": 1.9321812391281128, "learning_rate": 2.607036880033913e-06, "loss": 0.545, "mean_token_accuracy": 0.8240002989768982, "num_tokens": 23264184.0, "step": 616 }, { "epoch": 0.07848874189034474, "ewc_loss": 0.013427734375, "ewc_loss_diag": 9.298324584960938e-06, "ewc_loss_parallel": 1.2516975402832031e-05, "grad_norm": 1.9465829133987427, "learning_rate": 2.6112759643916917e-06, "loss": 0.5011, "mean_token_accuracy": 0.8436336517333984, "num_tokens": 23295543.0, "step": 617 }, { "epoch": 0.07861595216893524, "ewc_loss": 0.01348876953125, "ewc_loss_diag": 9.357929229736328e-06, "ewc_loss_parallel": 1.2576580047607422e-05, "grad_norm": 1.93916654586792, "learning_rate": 2.61551504874947e-06, "loss": 0.5272, "mean_token_accuracy": 0.8383979797363281, "num_tokens": 23328148.0, "step": 618 }, { "epoch": 0.07874316244752576, "ewc_loss": 0.01348876953125, "ewc_loss_diag": 9.357929229736328e-06, "ewc_loss_parallel": 1.2576580047607422e-05, "grad_norm": 1.9184718132019043, "learning_rate": 2.619754133107249e-06, "loss": 0.5147, "mean_token_accuracy": 0.8390636444091797, "num_tokens": 23369441.0, "step": 619 }, { "epoch": 0.07887037272611627, "ewc_loss": 0.01348876953125, "ewc_loss_diag": 9.357929229736328e-06, "ewc_loss_parallel": 1.2576580047607422e-05, "grad_norm": 1.964369535446167, "learning_rate": 2.623993217465028e-06, "loss": 0.5741, "mean_token_accuracy": 0.8199794888496399, "num_tokens": 23402977.0, "step": 620 }, { "epoch": 0.07899758300470679, "ewc_loss": 0.013427734375, "ewc_loss_diag": 9.357929229736328e-06, "ewc_loss_parallel": 1.2516975402832031e-05, "grad_norm": 1.9348750114440918, "learning_rate": 2.628232301822807e-06, "loss": 0.5508, "mean_token_accuracy": 0.8317515850067139, "num_tokens": 23441988.0, "step": 621 }, { "epoch": 0.07912479328329729, "ewc_loss": 0.01348876953125, "ewc_loss_diag": 9.357929229736328e-06, "ewc_loss_parallel": 1.2576580047607422e-05, "grad_norm": 1.947283387184143, "learning_rate": 2.6324713861805853e-06, "loss": 0.5796, "mean_token_accuracy": 0.819772481918335, "num_tokens": 23480481.0, "step": 622 }, { "epoch": 0.0792520035618878, "ewc_loss": 0.01348876953125, "ewc_loss_diag": 9.357929229736328e-06, "ewc_loss_parallel": 1.2576580047607422e-05, "grad_norm": 1.9599580764770508, "learning_rate": 2.6367104705383637e-06, "loss": 0.5758, "mean_token_accuracy": 0.8188304901123047, "num_tokens": 23515188.0, "step": 623 }, { "epoch": 0.07937921384047832, "ewc_loss": 0.013671875, "ewc_loss_diag": 9.47713851928711e-06, "ewc_loss_parallel": 1.2695789337158203e-05, "grad_norm": 1.9325354099273682, "learning_rate": 2.6409495548961426e-06, "loss": 0.563, "mean_token_accuracy": 0.8272110223770142, "num_tokens": 23557777.0, "step": 624 }, { "epoch": 0.07950642411906882, "ewc_loss": 0.013671875, "ewc_loss_diag": 9.47713851928711e-06, "ewc_loss_parallel": 1.2695789337158203e-05, "grad_norm": 2.0003466606140137, "learning_rate": 2.645188639253921e-06, "loss": 0.5513, "mean_token_accuracy": 0.8337125778198242, "num_tokens": 23588357.0, "step": 625 }, { "epoch": 0.07963363439765933, "ewc_loss": 0.013671875, "ewc_loss_diag": 9.47713851928711e-06, "ewc_loss_parallel": 1.2755393981933594e-05, "grad_norm": 1.9250435829162598, "learning_rate": 2.6494277236117e-06, "loss": 0.512, "mean_token_accuracy": 0.8366419672966003, "num_tokens": 23628656.0, "step": 626 }, { "epoch": 0.07976084467624985, "ewc_loss": 0.013671875, "ewc_loss_diag": 9.47713851928711e-06, "ewc_loss_parallel": 1.2755393981933594e-05, "grad_norm": 1.969826579093933, "learning_rate": 2.653666807969479e-06, "loss": 0.6119, "mean_token_accuracy": 0.8118149638175964, "num_tokens": 23666390.0, "step": 627 }, { "epoch": 0.07988805495484035, "ewc_loss": 0.0137939453125, "ewc_loss_diag": 9.47713851928711e-06, "ewc_loss_parallel": 1.2814998626708984e-05, "grad_norm": 1.9470560550689697, "learning_rate": 2.6579058923272577e-06, "loss": 0.4697, "mean_token_accuracy": 0.855156660079956, "num_tokens": 23699620.0, "step": 628 }, { "epoch": 0.08001526523343086, "ewc_loss": 0.0137939453125, "ewc_loss_diag": 9.47713851928711e-06, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 1.9624160528182983, "learning_rate": 2.662144976685036e-06, "loss": 0.5509, "mean_token_accuracy": 0.8298313021659851, "num_tokens": 23735883.0, "step": 629 }, { "epoch": 0.08014247551202137, "ewc_loss": 0.0137939453125, "ewc_loss_diag": 9.5367431640625e-06, "ewc_loss_parallel": 1.2814998626708984e-05, "grad_norm": 1.95307195186615, "learning_rate": 2.666384061042815e-06, "loss": 0.5112, "mean_token_accuracy": 0.8396041393280029, "num_tokens": 23769918.0, "step": 630 }, { "epoch": 0.08026968579061187, "ewc_loss": 0.013916015625, "ewc_loss_diag": 9.655952453613281e-06, "ewc_loss_parallel": 1.2934207916259766e-05, "grad_norm": 1.9409916400909424, "learning_rate": 2.670623145400594e-06, "loss": 0.5202, "mean_token_accuracy": 0.8376294374465942, "num_tokens": 23815402.0, "step": 631 }, { "epoch": 0.08039689606920239, "ewc_loss": 0.01385498046875, "ewc_loss_diag": 9.655952453613281e-06, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 1.9409465789794922, "learning_rate": 2.6748622297583724e-06, "loss": 0.5165, "mean_token_accuracy": 0.8388785123825073, "num_tokens": 23858205.0, "step": 632 }, { "epoch": 0.0805241063477929, "ewc_loss": 0.013916015625, "ewc_loss_diag": 9.655952453613281e-06, "ewc_loss_parallel": 1.2934207916259766e-05, "grad_norm": 1.9731289148330688, "learning_rate": 2.679101314116151e-06, "loss": 0.5312, "mean_token_accuracy": 0.838014543056488, "num_tokens": 23893506.0, "step": 633 }, { "epoch": 0.08065131662638342, "ewc_loss": 0.0137939453125, "ewc_loss_diag": 9.655952453613281e-06, "ewc_loss_parallel": 1.2814998626708984e-05, "grad_norm": 1.9683119058609009, "learning_rate": 2.6833403984739298e-06, "loss": 0.5214, "mean_token_accuracy": 0.8343169689178467, "num_tokens": 23929457.0, "step": 634 }, { "epoch": 0.08077852690497392, "ewc_loss": 0.01397705078125, "ewc_loss_diag": 9.655952453613281e-06, "ewc_loss_parallel": 1.2993812561035156e-05, "grad_norm": 1.9429761171340942, "learning_rate": 2.6875794828317086e-06, "loss": 0.5171, "mean_token_accuracy": 0.8422045111656189, "num_tokens": 23974577.0, "step": 635 }, { "epoch": 0.08090573718356443, "ewc_loss": 0.013916015625, "ewc_loss_diag": 9.775161743164062e-06, "ewc_loss_parallel": 1.2934207916259766e-05, "grad_norm": 1.9997340440750122, "learning_rate": 2.6918185671894875e-06, "loss": 0.5937, "mean_token_accuracy": 0.8180227279663086, "num_tokens": 24007490.0, "step": 636 }, { "epoch": 0.08103294746215495, "ewc_loss": 0.01397705078125, "ewc_loss_diag": 9.775161743164062e-06, "ewc_loss_parallel": 1.2993812561035156e-05, "grad_norm": 1.975460171699524, "learning_rate": 2.696057651547266e-06, "loss": 0.6319, "mean_token_accuracy": 0.8038982152938843, "num_tokens": 24048058.0, "step": 637 }, { "epoch": 0.08116015774074545, "ewc_loss": 0.0140380859375, "ewc_loss_diag": 9.775161743164062e-06, "ewc_loss_parallel": 1.3053417205810547e-05, "grad_norm": 1.9488770961761475, "learning_rate": 2.700296735905045e-06, "loss": 0.4983, "mean_token_accuracy": 0.8434640169143677, "num_tokens": 24089776.0, "step": 638 }, { "epoch": 0.08128736801933596, "ewc_loss": 0.0140380859375, "ewc_loss_diag": 9.775161743164062e-06, "ewc_loss_parallel": 1.3053417205810547e-05, "grad_norm": 1.9771335124969482, "learning_rate": 2.7045358202628238e-06, "loss": 0.549, "mean_token_accuracy": 0.8256363868713379, "num_tokens": 24135730.0, "step": 639 }, { "epoch": 0.08141457829792648, "ewc_loss": 0.0140380859375, "ewc_loss_diag": 9.775161743164062e-06, "ewc_loss_parallel": 1.3053417205810547e-05, "grad_norm": 1.9773069620132446, "learning_rate": 2.7087749046206026e-06, "loss": 0.5038, "mean_token_accuracy": 0.8422160148620605, "num_tokens": 24168513.0, "step": 640 }, { "epoch": 0.08154178857651698, "ewc_loss": 0.0140380859375, "ewc_loss_diag": 9.775161743164062e-06, "ewc_loss_parallel": 1.3053417205810547e-05, "grad_norm": 1.9803714752197266, "learning_rate": 2.7130139889783807e-06, "loss": 0.5666, "mean_token_accuracy": 0.8231127262115479, "num_tokens": 24206555.0, "step": 641 }, { "epoch": 0.08166899885510749, "ewc_loss": 0.01397705078125, "ewc_loss_diag": 9.775161743164062e-06, "ewc_loss_parallel": 1.2993812561035156e-05, "grad_norm": 1.946038007736206, "learning_rate": 2.7172530733361596e-06, "loss": 0.4741, "mean_token_accuracy": 0.8514044284820557, "num_tokens": 24248919.0, "step": 642 }, { "epoch": 0.081796209133698, "ewc_loss": 0.0140380859375, "ewc_loss_diag": 9.834766387939453e-06, "ewc_loss_parallel": 1.3053417205810547e-05, "grad_norm": 1.9721364974975586, "learning_rate": 2.7214921576939385e-06, "loss": 0.4714, "mean_token_accuracy": 0.85347580909729, "num_tokens": 24290493.0, "step": 643 }, { "epoch": 0.0819234194122885, "ewc_loss": 0.0140380859375, "ewc_loss_diag": 9.834766387939453e-06, "ewc_loss_parallel": 1.3053417205810547e-05, "grad_norm": 1.9740570783615112, "learning_rate": 2.725731242051717e-06, "loss": 0.5276, "mean_token_accuracy": 0.8378545045852661, "num_tokens": 24327968.0, "step": 644 }, { "epoch": 0.08205062969087902, "ewc_loss": 0.01416015625, "ewc_loss_diag": 9.834766387939453e-06, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 1.9799330234527588, "learning_rate": 2.729970326409496e-06, "loss": 0.5592, "mean_token_accuracy": 0.8288306593894958, "num_tokens": 24370019.0, "step": 645 }, { "epoch": 0.08217783996946953, "ewc_loss": 0.01416015625, "ewc_loss_diag": 9.834766387939453e-06, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 1.9969758987426758, "learning_rate": 2.7342094107672747e-06, "loss": 0.5294, "mean_token_accuracy": 0.837286114692688, "num_tokens": 24403911.0, "step": 646 }, { "epoch": 0.08230505024806005, "ewc_loss": 0.01416015625, "ewc_loss_diag": 9.834766387939453e-06, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 1.9850038290023804, "learning_rate": 2.7384484951250536e-06, "loss": 0.5477, "mean_token_accuracy": 0.832021951675415, "num_tokens": 24445039.0, "step": 647 }, { "epoch": 0.08243226052665055, "ewc_loss": 0.0142822265625, "ewc_loss_diag": 9.894371032714844e-06, "ewc_loss_parallel": 1.329183578491211e-05, "grad_norm": 2.013120412826538, "learning_rate": 2.7426875794828316e-06, "loss": 0.5482, "mean_token_accuracy": 0.8263980150222778, "num_tokens": 24479543.0, "step": 648 }, { "epoch": 0.08255947080524106, "ewc_loss": 0.01416015625, "ewc_loss_diag": 9.894371032714844e-06, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 1.9728394746780396, "learning_rate": 2.7469266638406105e-06, "loss": 0.5069, "mean_token_accuracy": 0.8395885229110718, "num_tokens": 24517483.0, "step": 649 }, { "epoch": 0.08268668108383158, "ewc_loss": 0.014404296875, "ewc_loss_diag": 9.894371032714844e-06, "ewc_loss_parallel": 1.341104507446289e-05, "grad_norm": 1.999523401260376, "learning_rate": 2.7511657481983894e-06, "loss": 0.5387, "mean_token_accuracy": 0.8291463255882263, "num_tokens": 24557702.0, "step": 650 }, { "epoch": 0.08281389136242208, "ewc_loss": 0.01434326171875, "ewc_loss_diag": 9.894371032714844e-06, "ewc_loss_parallel": 1.33514404296875e-05, "grad_norm": 1.9831024408340454, "learning_rate": 2.7554048325561683e-06, "loss": 0.5425, "mean_token_accuracy": 0.8332559466362, "num_tokens": 24595145.0, "step": 651 }, { "epoch": 0.0829411016410126, "ewc_loss": 0.014404296875, "ewc_loss_diag": 9.953975677490234e-06, "ewc_loss_parallel": 1.341104507446289e-05, "grad_norm": 1.9962064027786255, "learning_rate": 2.7596439169139467e-06, "loss": 0.5319, "mean_token_accuracy": 0.8345940709114075, "num_tokens": 24631564.0, "step": 652 }, { "epoch": 0.08306831191960311, "ewc_loss": 0.01446533203125, "ewc_loss_diag": 9.953975677490234e-06, "ewc_loss_parallel": 1.3470649719238281e-05, "grad_norm": 2.0003881454467773, "learning_rate": 2.7638830012717256e-06, "loss": 0.5857, "mean_token_accuracy": 0.8188061714172363, "num_tokens": 24668563.0, "step": 653 }, { "epoch": 0.08319552219819361, "ewc_loss": 0.01446533203125, "ewc_loss_diag": 1.0013580322265625e-05, "ewc_loss_parallel": 1.3470649719238281e-05, "grad_norm": 2.0099539756774902, "learning_rate": 2.7681220856295045e-06, "loss": 0.5715, "mean_token_accuracy": 0.8203578591346741, "num_tokens": 24703662.0, "step": 654 }, { "epoch": 0.08332273247678412, "ewc_loss": 0.0146484375, "ewc_loss_diag": 1.0013580322265625e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 1.9966830015182495, "learning_rate": 2.7723611699872834e-06, "loss": 0.5246, "mean_token_accuracy": 0.839523196220398, "num_tokens": 24739504.0, "step": 655 }, { "epoch": 0.08344994275537464, "ewc_loss": 0.0146484375, "ewc_loss_diag": 1.0013580322265625e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 2.0353755950927734, "learning_rate": 2.7766002543450614e-06, "loss": 0.5341, "mean_token_accuracy": 0.8300089836120605, "num_tokens": 24781884.0, "step": 656 }, { "epoch": 0.08357715303396514, "ewc_loss": 0.0146484375, "ewc_loss_diag": 1.0073184967041016e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 1.9840248823165894, "learning_rate": 2.7808393387028403e-06, "loss": 0.5884, "mean_token_accuracy": 0.8166500926017761, "num_tokens": 24823964.0, "step": 657 }, { "epoch": 0.08370436331255565, "ewc_loss": 0.0147705078125, "ewc_loss_diag": 1.0073184967041016e-05, "ewc_loss_parallel": 1.3709068298339844e-05, "grad_norm": 2.019223690032959, "learning_rate": 2.785078423060619e-06, "loss": 0.5312, "mean_token_accuracy": 0.8358717560768127, "num_tokens": 24859552.0, "step": 658 }, { "epoch": 0.08383157359114617, "ewc_loss": 0.0147705078125, "ewc_loss_diag": 1.0073184967041016e-05, "ewc_loss_parallel": 1.3709068298339844e-05, "grad_norm": 2.0099077224731445, "learning_rate": 2.789317507418398e-06, "loss": 0.5512, "mean_token_accuracy": 0.8275672197341919, "num_tokens": 24897580.0, "step": 659 }, { "epoch": 0.08395878386973668, "ewc_loss": 0.0147705078125, "ewc_loss_diag": 1.0073184967041016e-05, "ewc_loss_parallel": 1.3709068298339844e-05, "grad_norm": 2.0057177543640137, "learning_rate": 2.7935565917761765e-06, "loss": 0.499, "mean_token_accuracy": 0.8412923812866211, "num_tokens": 24934616.0, "step": 660 }, { "epoch": 0.08408599414832718, "ewc_loss": 0.0147705078125, "ewc_loss_diag": 1.0132789611816406e-05, "ewc_loss_parallel": 1.3709068298339844e-05, "grad_norm": 2.0254228115081787, "learning_rate": 2.7977956761339554e-06, "loss": 0.5713, "mean_token_accuracy": 0.8221317529678345, "num_tokens": 24972149.0, "step": 661 }, { "epoch": 0.0842132044269177, "ewc_loss": 0.01483154296875, "ewc_loss_diag": 1.0132789611816406e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 2.0042953491210938, "learning_rate": 2.8020347604917343e-06, "loss": 0.451, "mean_token_accuracy": 0.8596667051315308, "num_tokens": 25008682.0, "step": 662 }, { "epoch": 0.08434041470550821, "ewc_loss": 0.01483154296875, "ewc_loss_diag": 1.0132789611816406e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 1.9950892925262451, "learning_rate": 2.8062738448495123e-06, "loss": 0.4783, "mean_token_accuracy": 0.8480808138847351, "num_tokens": 25048454.0, "step": 663 }, { "epoch": 0.08446762498409871, "ewc_loss": 0.01495361328125, "ewc_loss_diag": 1.0132789611816406e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 2.029042959213257, "learning_rate": 2.8105129292072912e-06, "loss": 0.5547, "mean_token_accuracy": 0.8214453458786011, "num_tokens": 25086772.0, "step": 664 }, { "epoch": 0.08459483526268922, "ewc_loss": 0.0150146484375, "ewc_loss_diag": 1.0132789611816406e-05, "ewc_loss_parallel": 1.3947486877441406e-05, "grad_norm": 2.0106213092803955, "learning_rate": 2.81475201356507e-06, "loss": 0.5032, "mean_token_accuracy": 0.8397310376167297, "num_tokens": 25122215.0, "step": 665 }, { "epoch": 0.08472204554127974, "ewc_loss": 0.0150146484375, "ewc_loss_diag": 1.0251998901367188e-05, "ewc_loss_parallel": 1.3947486877441406e-05, "grad_norm": 2.022366762161255, "learning_rate": 2.818991097922849e-06, "loss": 0.5454, "mean_token_accuracy": 0.8340785503387451, "num_tokens": 25159966.0, "step": 666 }, { "epoch": 0.08484925581987024, "ewc_loss": 0.0150146484375, "ewc_loss_diag": 1.0251998901367188e-05, "ewc_loss_parallel": 1.3947486877441406e-05, "grad_norm": 2.03344464302063, "learning_rate": 2.8232301822806274e-06, "loss": 0.5004, "mean_token_accuracy": 0.8449482321739197, "num_tokens": 25191929.0, "step": 667 }, { "epoch": 0.08497646609846075, "ewc_loss": 0.0150146484375, "ewc_loss_diag": 1.0251998901367188e-05, "ewc_loss_parallel": 1.4007091522216797e-05, "grad_norm": 2.015413761138916, "learning_rate": 2.8274692666384063e-06, "loss": 0.522, "mean_token_accuracy": 0.8347073793411255, "num_tokens": 25233385.0, "step": 668 }, { "epoch": 0.08510367637705127, "ewc_loss": 0.0150146484375, "ewc_loss_diag": 1.0251998901367188e-05, "ewc_loss_parallel": 1.3947486877441406e-05, "grad_norm": 2.0090715885162354, "learning_rate": 2.8317083509961852e-06, "loss": 0.5418, "mean_token_accuracy": 0.831371545791626, "num_tokens": 25276943.0, "step": 669 }, { "epoch": 0.08523088665564178, "ewc_loss": 0.0150146484375, "ewc_loss_diag": 1.0371208190917969e-05, "ewc_loss_parallel": 1.4007091522216797e-05, "grad_norm": 2.023669719696045, "learning_rate": 2.835947435353964e-06, "loss": 0.5192, "mean_token_accuracy": 0.8380703926086426, "num_tokens": 25316552.0, "step": 670 }, { "epoch": 0.08535809693423228, "ewc_loss": 0.0150146484375, "ewc_loss_diag": 1.0251998901367188e-05, "ewc_loss_parallel": 1.3947486877441406e-05, "grad_norm": 2.0411674976348877, "learning_rate": 2.840186519711742e-06, "loss": 0.574, "mean_token_accuracy": 0.8280364871025085, "num_tokens": 25357517.0, "step": 671 }, { "epoch": 0.0854853072128228, "ewc_loss": 0.01507568359375, "ewc_loss_diag": 1.0371208190917969e-05, "ewc_loss_parallel": 1.4066696166992188e-05, "grad_norm": 2.050835609436035, "learning_rate": 2.844425604069521e-06, "loss": 0.5649, "mean_token_accuracy": 0.8252803087234497, "num_tokens": 25392693.0, "step": 672 }, { "epoch": 0.08561251749141331, "ewc_loss": 0.0150146484375, "ewc_loss_diag": 1.0371208190917969e-05, "ewc_loss_parallel": 1.4007091522216797e-05, "grad_norm": 2.0250465869903564, "learning_rate": 2.8486646884273e-06, "loss": 0.527, "mean_token_accuracy": 0.8358426094055176, "num_tokens": 25430623.0, "step": 673 }, { "epoch": 0.08573972777000381, "ewc_loss": 0.01507568359375, "ewc_loss_diag": 1.0371208190917969e-05, "ewc_loss_parallel": 1.4066696166992188e-05, "grad_norm": 2.0238208770751953, "learning_rate": 2.852903772785079e-06, "loss": 0.4665, "mean_token_accuracy": 0.8531680107116699, "num_tokens": 25465043.0, "step": 674 }, { "epoch": 0.08586693804859433, "ewc_loss": 0.0150146484375, "ewc_loss_diag": 1.043081283569336e-05, "ewc_loss_parallel": 1.4007091522216797e-05, "grad_norm": 2.0255658626556396, "learning_rate": 2.8571428571428573e-06, "loss": 0.4903, "mean_token_accuracy": 0.8474138975143433, "num_tokens": 25503483.0, "step": 675 }, { "epoch": 0.08599414832718484, "ewc_loss": 0.01507568359375, "ewc_loss_diag": 1.049041748046875e-05, "ewc_loss_parallel": 1.4066696166992188e-05, "grad_norm": 2.0249764919281006, "learning_rate": 2.861381941500636e-06, "loss": 0.5431, "mean_token_accuracy": 0.8306725025177002, "num_tokens": 25547079.0, "step": 676 }, { "epoch": 0.08612135860577534, "ewc_loss": 0.01507568359375, "ewc_loss_diag": 1.049041748046875e-05, "ewc_loss_parallel": 1.4066696166992188e-05, "grad_norm": 2.0366480350494385, "learning_rate": 2.865621025858415e-06, "loss": 0.5784, "mean_token_accuracy": 0.8190491795539856, "num_tokens": 25586494.0, "step": 677 }, { "epoch": 0.08624856888436586, "ewc_loss": 0.01519775390625, "ewc_loss_diag": 1.049041748046875e-05, "ewc_loss_parallel": 1.4185905456542969e-05, "grad_norm": 2.060234308242798, "learning_rate": 2.869860110216193e-06, "loss": 0.5775, "mean_token_accuracy": 0.8187738656997681, "num_tokens": 25618998.0, "step": 678 }, { "epoch": 0.08637577916295637, "ewc_loss": 0.01513671875, "ewc_loss_diag": 1.055002212524414e-05, "ewc_loss_parallel": 1.4126300811767578e-05, "grad_norm": 2.038248300552368, "learning_rate": 2.874099194573972e-06, "loss": 0.519, "mean_token_accuracy": 0.8375617265701294, "num_tokens": 25663147.0, "step": 679 }, { "epoch": 0.08650298944154687, "ewc_loss": 0.0152587890625, "ewc_loss_diag": 1.055002212524414e-05, "ewc_loss_parallel": 1.424551010131836e-05, "grad_norm": 2.058440923690796, "learning_rate": 2.878338278931751e-06, "loss": 0.5681, "mean_token_accuracy": 0.8280753493309021, "num_tokens": 25696830.0, "step": 680 }, { "epoch": 0.08663019972013739, "ewc_loss": 0.0152587890625, "ewc_loss_diag": 1.055002212524414e-05, "ewc_loss_parallel": 1.424551010131836e-05, "grad_norm": 2.0369672775268555, "learning_rate": 2.8825773632895297e-06, "loss": 0.4954, "mean_token_accuracy": 0.8489928245544434, "num_tokens": 25732828.0, "step": 681 }, { "epoch": 0.0867574099987279, "ewc_loss": 0.0152587890625, "ewc_loss_diag": 1.0609626770019531e-05, "ewc_loss_parallel": 1.424551010131836e-05, "grad_norm": 2.0294017791748047, "learning_rate": 2.886816447647308e-06, "loss": 0.4862, "mean_token_accuracy": 0.8464552164077759, "num_tokens": 25773545.0, "step": 682 }, { "epoch": 0.08688462027731841, "ewc_loss": 0.0152587890625, "ewc_loss_diag": 1.0609626770019531e-05, "ewc_loss_parallel": 1.424551010131836e-05, "grad_norm": 2.035274028778076, "learning_rate": 2.891055532005087e-06, "loss": 0.5238, "mean_token_accuracy": 0.8371478319168091, "num_tokens": 25814792.0, "step": 683 }, { "epoch": 0.08701183055590891, "ewc_loss": 0.01544189453125, "ewc_loss_diag": 1.0609626770019531e-05, "ewc_loss_parallel": 1.4424324035644531e-05, "grad_norm": 2.055835247039795, "learning_rate": 2.895294616362866e-06, "loss": 0.5401, "mean_token_accuracy": 0.8296211957931519, "num_tokens": 25850713.0, "step": 684 }, { "epoch": 0.08713904083449943, "ewc_loss": 0.01544189453125, "ewc_loss_diag": 1.0609626770019531e-05, "ewc_loss_parallel": 1.4424324035644531e-05, "grad_norm": 2.0496180057525635, "learning_rate": 2.899533700720645e-06, "loss": 0.5092, "mean_token_accuracy": 0.8357721567153931, "num_tokens": 25888320.0, "step": 685 }, { "epoch": 0.08726625111308994, "ewc_loss": 0.01544189453125, "ewc_loss_diag": 1.0609626770019531e-05, "ewc_loss_parallel": 1.4424324035644531e-05, "grad_norm": 2.0580801963806152, "learning_rate": 2.9037727850784233e-06, "loss": 0.5525, "mean_token_accuracy": 0.8271877765655518, "num_tokens": 25924941.0, "step": 686 }, { "epoch": 0.08739346139168044, "ewc_loss": 0.01544189453125, "ewc_loss_diag": 1.0609626770019531e-05, "ewc_loss_parallel": 1.4424324035644531e-05, "grad_norm": 2.0453755855560303, "learning_rate": 2.9080118694362018e-06, "loss": 0.5186, "mean_token_accuracy": 0.8382200598716736, "num_tokens": 25964051.0, "step": 687 }, { "epoch": 0.08752067167027096, "ewc_loss": 0.0155029296875, "ewc_loss_diag": 1.0669231414794922e-05, "ewc_loss_parallel": 1.4483928680419922e-05, "grad_norm": 2.049139976501465, "learning_rate": 2.9122509537939806e-06, "loss": 0.5217, "mean_token_accuracy": 0.8384333848953247, "num_tokens": 26007300.0, "step": 688 }, { "epoch": 0.08764788194886147, "ewc_loss": 0.015625, "ewc_loss_diag": 1.0669231414794922e-05, "ewc_loss_parallel": 1.4543533325195312e-05, "grad_norm": 2.063701629638672, "learning_rate": 2.9164900381517595e-06, "loss": 0.5787, "mean_token_accuracy": 0.8175234794616699, "num_tokens": 26045833.0, "step": 689 }, { "epoch": 0.08777509222745197, "ewc_loss": 0.015625, "ewc_loss_diag": 1.0669231414794922e-05, "ewc_loss_parallel": 1.4543533325195312e-05, "grad_norm": 2.050016403198242, "learning_rate": 2.920729122509538e-06, "loss": 0.5705, "mean_token_accuracy": 0.8226109743118286, "num_tokens": 26092277.0, "step": 690 }, { "epoch": 0.08790230250604249, "ewc_loss": 0.015625, "ewc_loss_diag": 1.0669231414794922e-05, "ewc_loss_parallel": 1.4603137969970703e-05, "grad_norm": 2.0588622093200684, "learning_rate": 2.924968206867317e-06, "loss": 0.5236, "mean_token_accuracy": 0.8336787223815918, "num_tokens": 26130082.0, "step": 691 }, { "epoch": 0.088029512784633, "ewc_loss": 0.01556396484375, "ewc_loss_diag": 1.0788440704345703e-05, "ewc_loss_parallel": 1.4483928680419922e-05, "grad_norm": 2.0565407276153564, "learning_rate": 2.9292072912250958e-06, "loss": 0.4995, "mean_token_accuracy": 0.8454325199127197, "num_tokens": 26170087.0, "step": 692 }, { "epoch": 0.0881567230632235, "ewc_loss": 0.0157470703125, "ewc_loss_diag": 1.0788440704345703e-05, "ewc_loss_parallel": 1.4662742614746094e-05, "grad_norm": 2.0848348140716553, "learning_rate": 2.9334463755828746e-06, "loss": 0.5799, "mean_token_accuracy": 0.8200283050537109, "num_tokens": 26207645.0, "step": 693 }, { "epoch": 0.08828393334181402, "ewc_loss": 0.0157470703125, "ewc_loss_diag": 1.0788440704345703e-05, "ewc_loss_parallel": 1.4662742614746094e-05, "grad_norm": 2.0619308948516846, "learning_rate": 2.937685459940653e-06, "loss": 0.5018, "mean_token_accuracy": 0.8403548002243042, "num_tokens": 26242417.0, "step": 694 }, { "epoch": 0.08841114362040453, "ewc_loss": 0.0157470703125, "ewc_loss_diag": 1.0907649993896484e-05, "ewc_loss_parallel": 1.4662742614746094e-05, "grad_norm": 2.1032159328460693, "learning_rate": 2.941924544298432e-06, "loss": 0.5701, "mean_token_accuracy": 0.8236583471298218, "num_tokens": 26275427.0, "step": 695 }, { "epoch": 0.08853835389899505, "ewc_loss": 0.015625, "ewc_loss_diag": 1.0907649993896484e-05, "ewc_loss_parallel": 1.4543533325195312e-05, "grad_norm": 2.066699981689453, "learning_rate": 2.9461636286562104e-06, "loss": 0.5223, "mean_token_accuracy": 0.8369756937026978, "num_tokens": 26308692.0, "step": 696 }, { "epoch": 0.08866556417758555, "ewc_loss": 0.015625, "ewc_loss_diag": 1.0907649993896484e-05, "ewc_loss_parallel": 1.4543533325195312e-05, "grad_norm": 2.058870792388916, "learning_rate": 2.950402713013989e-06, "loss": 0.516, "mean_token_accuracy": 0.8394951820373535, "num_tokens": 26352155.0, "step": 697 }, { "epoch": 0.08879277445617606, "ewc_loss": 0.0157470703125, "ewc_loss_diag": 1.0907649993896484e-05, "ewc_loss_parallel": 1.4662742614746094e-05, "grad_norm": 2.0681633949279785, "learning_rate": 2.954641797371768e-06, "loss": 0.558, "mean_token_accuracy": 0.827937662601471, "num_tokens": 26392410.0, "step": 698 }, { "epoch": 0.08891998473476657, "ewc_loss": 0.015869140625, "ewc_loss_diag": 1.1026859283447266e-05, "ewc_loss_parallel": 1.4781951904296875e-05, "grad_norm": 2.072624444961548, "learning_rate": 2.9588808817295467e-06, "loss": 0.4849, "mean_token_accuracy": 0.847489595413208, "num_tokens": 26430120.0, "step": 699 }, { "epoch": 0.08904719501335707, "ewc_loss": 0.015869140625, "ewc_loss_diag": 1.1026859283447266e-05, "ewc_loss_parallel": 1.4722347259521484e-05, "grad_norm": 2.0825202465057373, "learning_rate": 2.9631199660873256e-06, "loss": 0.5478, "mean_token_accuracy": 0.8272815346717834, "num_tokens": 26466159.0, "step": 700 }, { "epoch": 0.08917440529194759, "ewc_loss": 0.0159912109375, "ewc_loss_diag": 1.1026859283447266e-05, "ewc_loss_parallel": 1.4841556549072266e-05, "grad_norm": 2.06594181060791, "learning_rate": 2.967359050445104e-06, "loss": 0.4685, "mean_token_accuracy": 0.8545840978622437, "num_tokens": 26502382.0, "step": 701 }, { "epoch": 0.0893016155705381, "ewc_loss": 0.015869140625, "ewc_loss_diag": 1.1026859283447266e-05, "ewc_loss_parallel": 1.4722347259521484e-05, "grad_norm": 2.080972671508789, "learning_rate": 2.971598134802883e-06, "loss": 0.5065, "mean_token_accuracy": 0.8428076505661011, "num_tokens": 26540911.0, "step": 702 }, { "epoch": 0.0894288258491286, "ewc_loss": 0.0159912109375, "ewc_loss_diag": 1.1026859283447266e-05, "ewc_loss_parallel": 1.4841556549072266e-05, "grad_norm": 2.0889859199523926, "learning_rate": 2.975837219160662e-06, "loss": 0.5682, "mean_token_accuracy": 0.8223851919174194, "num_tokens": 26580081.0, "step": 703 }, { "epoch": 0.08955603612771912, "ewc_loss": 0.0159912109375, "ewc_loss_diag": 1.1026859283447266e-05, "ewc_loss_parallel": 1.4841556549072266e-05, "grad_norm": 2.085066318511963, "learning_rate": 2.9800763035184403e-06, "loss": 0.5337, "mean_token_accuracy": 0.8340452313423157, "num_tokens": 26622033.0, "step": 704 }, { "epoch": 0.08968324640630963, "ewc_loss": 0.0159912109375, "ewc_loss_diag": 1.1026859283447266e-05, "ewc_loss_parallel": 1.4901161193847656e-05, "grad_norm": 2.0945072174072266, "learning_rate": 2.9843153878762187e-06, "loss": 0.5348, "mean_token_accuracy": 0.8331421613693237, "num_tokens": 26657042.0, "step": 705 }, { "epoch": 0.08981045668490013, "ewc_loss": 0.0159912109375, "ewc_loss_diag": 1.1026859283447266e-05, "ewc_loss_parallel": 1.4841556549072266e-05, "grad_norm": 2.0939085483551025, "learning_rate": 2.9885544722339976e-06, "loss": 0.5455, "mean_token_accuracy": 0.8303499221801758, "num_tokens": 26696668.0, "step": 706 }, { "epoch": 0.08993766696349065, "ewc_loss": 0.0159912109375, "ewc_loss_diag": 1.1026859283447266e-05, "ewc_loss_parallel": 1.4901161193847656e-05, "grad_norm": 2.087268829345703, "learning_rate": 2.9927935565917765e-06, "loss": 0.5494, "mean_token_accuracy": 0.8320331573486328, "num_tokens": 26736550.0, "step": 707 }, { "epoch": 0.09006487724208116, "ewc_loss": 0.0159912109375, "ewc_loss_diag": 1.1086463928222656e-05, "ewc_loss_parallel": 1.4901161193847656e-05, "grad_norm": 2.0901122093200684, "learning_rate": 2.9970326409495554e-06, "loss": 0.5556, "mean_token_accuracy": 0.8286980986595154, "num_tokens": 26776311.0, "step": 708 }, { "epoch": 0.09019208752067168, "ewc_loss": 0.01611328125, "ewc_loss_diag": 1.1086463928222656e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 2.0945796966552734, "learning_rate": 3.001271725307334e-06, "loss": 0.4946, "mean_token_accuracy": 0.8457457423210144, "num_tokens": 26812094.0, "step": 709 }, { "epoch": 0.09031929779926218, "ewc_loss": 0.01611328125, "ewc_loss_diag": 1.1086463928222656e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 2.106630802154541, "learning_rate": 3.0055108096651127e-06, "loss": 0.5295, "mean_token_accuracy": 0.8364442586898804, "num_tokens": 26846317.0, "step": 710 }, { "epoch": 0.09044650807785269, "ewc_loss": 0.0162353515625, "ewc_loss_diag": 1.1086463928222656e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 2.0738043785095215, "learning_rate": 3.0097498940228916e-06, "loss": 0.47, "mean_token_accuracy": 0.8518580198287964, "num_tokens": 26885247.0, "step": 711 }, { "epoch": 0.0905737183564432, "ewc_loss": 0.0162353515625, "ewc_loss_diag": 1.1086463928222656e-05, "ewc_loss_parallel": 1.5079975128173828e-05, "grad_norm": 2.101804494857788, "learning_rate": 3.0139889783806705e-06, "loss": 0.5345, "mean_token_accuracy": 0.8350517749786377, "num_tokens": 26925578.0, "step": 712 }, { "epoch": 0.0907009286350337, "ewc_loss": 0.016357421875, "ewc_loss_diag": 1.1086463928222656e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 2.1005117893218994, "learning_rate": 3.0182280627384485e-06, "loss": 0.5585, "mean_token_accuracy": 0.8296661972999573, "num_tokens": 26961924.0, "step": 713 }, { "epoch": 0.09082813891362422, "ewc_loss": 0.0162353515625, "ewc_loss_diag": 1.1086463928222656e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 2.093867540359497, "learning_rate": 3.0224671470962274e-06, "loss": 0.4882, "mean_token_accuracy": 0.8446526527404785, "num_tokens": 26999433.0, "step": 714 }, { "epoch": 0.09095534919221474, "ewc_loss": 0.0162353515625, "ewc_loss_diag": 1.1146068572998047e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 2.0856542587280273, "learning_rate": 3.0267062314540063e-06, "loss": 0.5113, "mean_token_accuracy": 0.8398616313934326, "num_tokens": 27044427.0, "step": 715 }, { "epoch": 0.09108255947080524, "ewc_loss": 0.0162353515625, "ewc_loss_diag": 1.1146068572998047e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 2.108647584915161, "learning_rate": 3.0309453158117847e-06, "loss": 0.5184, "mean_token_accuracy": 0.8375698328018188, "num_tokens": 27080645.0, "step": 716 }, { "epoch": 0.09120976974939575, "ewc_loss": 0.016357421875, "ewc_loss_diag": 1.1146068572998047e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 2.1224358081817627, "learning_rate": 3.0351844001695636e-06, "loss": 0.5867, "mean_token_accuracy": 0.8192740678787231, "num_tokens": 27118213.0, "step": 717 }, { "epoch": 0.09133698002798626, "ewc_loss": 0.0162353515625, "ewc_loss_diag": 1.1265277862548828e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 2.1071760654449463, "learning_rate": 3.0394234845273425e-06, "loss": 0.5035, "mean_token_accuracy": 0.8428959846496582, "num_tokens": 27156827.0, "step": 718 }, { "epoch": 0.09146419030657676, "ewc_loss": 0.016357421875, "ewc_loss_diag": 1.1265277862548828e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 2.1246986389160156, "learning_rate": 3.0436625688851214e-06, "loss": 0.5371, "mean_token_accuracy": 0.8283575773239136, "num_tokens": 27192016.0, "step": 719 }, { "epoch": 0.09159140058516728, "ewc_loss": 0.016357421875, "ewc_loss_diag": 1.1265277862548828e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 2.129058361053467, "learning_rate": 3.0479016532428994e-06, "loss": 0.5828, "mean_token_accuracy": 0.8159842491149902, "num_tokens": 27226037.0, "step": 720 }, { "epoch": 0.0917186108637578, "ewc_loss": 0.0162353515625, "ewc_loss_diag": 1.1324882507324219e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 2.1121199131011963, "learning_rate": 3.0521407376006783e-06, "loss": 0.4953, "mean_token_accuracy": 0.8438389897346497, "num_tokens": 27260399.0, "step": 721 }, { "epoch": 0.09184582114234831, "ewc_loss": 0.016357421875, "ewc_loss_diag": 1.1324882507324219e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 2.1155354976654053, "learning_rate": 3.056379821958457e-06, "loss": 0.5063, "mean_token_accuracy": 0.8432564735412598, "num_tokens": 27297293.0, "step": 722 }, { "epoch": 0.09197303142093881, "ewc_loss": 0.016357421875, "ewc_loss_diag": 1.1324882507324219e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 2.115159034729004, "learning_rate": 3.060618906316236e-06, "loss": 0.503, "mean_token_accuracy": 0.8428469896316528, "num_tokens": 27333045.0, "step": 723 }, { "epoch": 0.09210024169952932, "ewc_loss": 0.016357421875, "ewc_loss_diag": 1.1265277862548828e-05, "ewc_loss_parallel": 1.52587890625e-05, "grad_norm": 2.124809980392456, "learning_rate": 3.0648579906740146e-06, "loss": 0.5002, "mean_token_accuracy": 0.8421620726585388, "num_tokens": 27369044.0, "step": 724 }, { "epoch": 0.09222745197811984, "ewc_loss": 0.016357421875, "ewc_loss_diag": 1.1265277862548828e-05, "ewc_loss_parallel": 1.52587890625e-05, "grad_norm": 2.1199915409088135, "learning_rate": 3.0690970750317934e-06, "loss": 0.5547, "mean_token_accuracy": 0.8266842365264893, "num_tokens": 27407486.0, "step": 725 }, { "epoch": 0.09235466225671034, "ewc_loss": 0.016357421875, "ewc_loss_diag": 1.1265277862548828e-05, "ewc_loss_parallel": 1.52587890625e-05, "grad_norm": 2.1373846530914307, "learning_rate": 3.0733361593895723e-06, "loss": 0.5194, "mean_token_accuracy": 0.8389225006103516, "num_tokens": 27441835.0, "step": 726 }, { "epoch": 0.09248187253530085, "ewc_loss": 0.016357421875, "ewc_loss_diag": 1.1324882507324219e-05, "ewc_loss_parallel": 1.52587890625e-05, "grad_norm": 2.1037168502807617, "learning_rate": 3.077575243747351e-06, "loss": 0.5313, "mean_token_accuracy": 0.8327952027320862, "num_tokens": 27483105.0, "step": 727 }, { "epoch": 0.09260908281389137, "ewc_loss": 0.0164794921875, "ewc_loss_diag": 1.1324882507324219e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 2.1244146823883057, "learning_rate": 3.0818143281051292e-06, "loss": 0.5515, "mean_token_accuracy": 0.831377387046814, "num_tokens": 27528888.0, "step": 728 }, { "epoch": 0.09273629309248187, "ewc_loss": 0.0164794921875, "ewc_loss_diag": 1.1324882507324219e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 2.126614570617676, "learning_rate": 3.086053412462908e-06, "loss": 0.5887, "mean_token_accuracy": 0.8175760507583618, "num_tokens": 27571763.0, "step": 729 }, { "epoch": 0.09286350337107238, "ewc_loss": 0.0164794921875, "ewc_loss_diag": 1.1324882507324219e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 2.137024402618408, "learning_rate": 3.090292496820687e-06, "loss": 0.5204, "mean_token_accuracy": 0.8408163189888, "num_tokens": 27606980.0, "step": 730 }, { "epoch": 0.0929907136496629, "ewc_loss": 0.0164794921875, "ewc_loss_diag": 1.138448715209961e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 2.1617281436920166, "learning_rate": 3.094531581178466e-06, "loss": 0.5115, "mean_token_accuracy": 0.8400638699531555, "num_tokens": 27640969.0, "step": 731 }, { "epoch": 0.0931179239282534, "ewc_loss": 0.0166015625, "ewc_loss_diag": 1.1444091796875e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 2.1315348148345947, "learning_rate": 3.0987706655362444e-06, "loss": 0.6502, "mean_token_accuracy": 0.7974038124084473, "num_tokens": 27684360.0, "step": 732 }, { "epoch": 0.09324513420684391, "ewc_loss": 0.0166015625, "ewc_loss_diag": 1.1444091796875e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 2.1329452991485596, "learning_rate": 3.1030097498940232e-06, "loss": 0.49, "mean_token_accuracy": 0.8448498845100403, "num_tokens": 27723596.0, "step": 733 }, { "epoch": 0.09337234448543442, "ewc_loss": 0.0164794921875, "ewc_loss_diag": 1.1444091796875e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 2.1329097747802734, "learning_rate": 3.107248834251802e-06, "loss": 0.5257, "mean_token_accuracy": 0.8291244506835938, "num_tokens": 27759469.0, "step": 734 }, { "epoch": 0.09349955476402494, "ewc_loss": 0.0166015625, "ewc_loss_diag": 1.1444091796875e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 2.145982503890991, "learning_rate": 3.11148791860958e-06, "loss": 0.5183, "mean_token_accuracy": 0.8383479118347168, "num_tokens": 27795783.0, "step": 735 }, { "epoch": 0.09362676504261544, "ewc_loss": 0.0166015625, "ewc_loss_diag": 1.138448715209961e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 2.1539320945739746, "learning_rate": 3.115727002967359e-06, "loss": 0.5275, "mean_token_accuracy": 0.8327587842941284, "num_tokens": 27826146.0, "step": 736 }, { "epoch": 0.09375397532120595, "ewc_loss": 0.0166015625, "ewc_loss_diag": 1.150369644165039e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 2.151287078857422, "learning_rate": 3.119966087325138e-06, "loss": 0.5316, "mean_token_accuracy": 0.8369026184082031, "num_tokens": 27865581.0, "step": 737 }, { "epoch": 0.09388118559979647, "ewc_loss": 0.0166015625, "ewc_loss_diag": 1.150369644165039e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 2.12712025642395, "learning_rate": 3.124205171682917e-06, "loss": 0.574, "mean_token_accuracy": 0.8230664730072021, "num_tokens": 27911209.0, "step": 738 }, { "epoch": 0.09400839587838697, "ewc_loss": 0.0166015625, "ewc_loss_diag": 1.150369644165039e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 2.1464052200317383, "learning_rate": 3.1284442560406953e-06, "loss": 0.5376, "mean_token_accuracy": 0.8316293954849243, "num_tokens": 27951639.0, "step": 739 }, { "epoch": 0.09413560615697748, "ewc_loss": 0.0166015625, "ewc_loss_diag": 1.150369644165039e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 2.129542827606201, "learning_rate": 3.132683340398474e-06, "loss": 0.4706, "mean_token_accuracy": 0.8499611020088196, "num_tokens": 27990808.0, "step": 740 }, { "epoch": 0.094262816435568, "ewc_loss": 0.0166015625, "ewc_loss_diag": 1.150369644165039e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 2.1649105548858643, "learning_rate": 3.136922424756253e-06, "loss": 0.494, "mean_token_accuracy": 0.847411572933197, "num_tokens": 28025421.0, "step": 741 }, { "epoch": 0.0943900267141585, "ewc_loss": 0.0167236328125, "ewc_loss_diag": 1.150369644165039e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 2.1232054233551025, "learning_rate": 3.141161509114032e-06, "loss": 0.4648, "mean_token_accuracy": 0.8565864562988281, "num_tokens": 28068652.0, "step": 742 }, { "epoch": 0.09451723699274901, "ewc_loss": 0.0166015625, "ewc_loss_diag": 1.150369644165039e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 2.1635992527008057, "learning_rate": 3.14540059347181e-06, "loss": 0.5093, "mean_token_accuracy": 0.8381364345550537, "num_tokens": 28104852.0, "step": 743 }, { "epoch": 0.09464444727133953, "ewc_loss": 0.0167236328125, "ewc_loss_diag": 1.150369644165039e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 2.15716290473938, "learning_rate": 3.149639677829589e-06, "loss": 0.4801, "mean_token_accuracy": 0.8481771349906921, "num_tokens": 28141646.0, "step": 744 }, { "epoch": 0.09477165754993004, "ewc_loss": 0.016845703125, "ewc_loss_diag": 1.1563301086425781e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 2.1756956577301025, "learning_rate": 3.1538787621873677e-06, "loss": 0.579, "mean_token_accuracy": 0.8253967761993408, "num_tokens": 28178951.0, "step": 745 }, { "epoch": 0.09489886782852054, "ewc_loss": 0.0166015625, "ewc_loss_diag": 1.150369644165039e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 2.1496424674987793, "learning_rate": 3.1581178465451466e-06, "loss": 0.5942, "mean_token_accuracy": 0.8194855451583862, "num_tokens": 28221911.0, "step": 746 }, { "epoch": 0.09502607810711106, "ewc_loss": 0.0166015625, "ewc_loss_diag": 1.150369644165039e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 2.195125102996826, "learning_rate": 3.162356930902925e-06, "loss": 0.5126, "mean_token_accuracy": 0.8397613167762756, "num_tokens": 28252828.0, "step": 747 }, { "epoch": 0.09515328838570157, "ewc_loss": 0.0167236328125, "ewc_loss_diag": 1.150369644165039e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 2.138913631439209, "learning_rate": 3.166596015260704e-06, "loss": 0.4886, "mean_token_accuracy": 0.8461124897003174, "num_tokens": 28293459.0, "step": 748 }, { "epoch": 0.09528049866429207, "ewc_loss": 0.0167236328125, "ewc_loss_diag": 1.150369644165039e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 2.174797773361206, "learning_rate": 3.170835099618483e-06, "loss": 0.5234, "mean_token_accuracy": 0.836097240447998, "num_tokens": 28332775.0, "step": 749 }, { "epoch": 0.09540770894288259, "ewc_loss": 0.0167236328125, "ewc_loss_diag": 1.150369644165039e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 2.1436264514923096, "learning_rate": 3.1750741839762617e-06, "loss": 0.5008, "mean_token_accuracy": 0.8438147306442261, "num_tokens": 28370367.0, "step": 750 }, { "epoch": 0.0955349192214731, "ewc_loss": 0.016845703125, "ewc_loss_diag": 1.1622905731201172e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 2.1945431232452393, "learning_rate": 3.1793132683340398e-06, "loss": 0.534, "mean_token_accuracy": 0.8345810770988464, "num_tokens": 28404913.0, "step": 751 }, { "epoch": 0.0956621295000636, "ewc_loss": 0.016845703125, "ewc_loss_diag": 1.1622905731201172e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 2.15871262550354, "learning_rate": 3.1835523526918187e-06, "loss": 0.4756, "mean_token_accuracy": 0.8485893607139587, "num_tokens": 28441057.0, "step": 752 }, { "epoch": 0.09578933977865411, "ewc_loss": 0.016845703125, "ewc_loss_diag": 1.1742115020751953e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 2.1904635429382324, "learning_rate": 3.1877914370495975e-06, "loss": 0.5656, "mean_token_accuracy": 0.8222619295120239, "num_tokens": 28479691.0, "step": 753 }, { "epoch": 0.09591655005724463, "ewc_loss": 0.016845703125, "ewc_loss_diag": 1.1742115020751953e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 2.161639928817749, "learning_rate": 3.192030521407376e-06, "loss": 0.5599, "mean_token_accuracy": 0.8279933929443359, "num_tokens": 28520469.0, "step": 754 }, { "epoch": 0.09604376033583513, "ewc_loss": 0.016845703125, "ewc_loss_diag": 1.1742115020751953e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 2.178173065185547, "learning_rate": 3.196269605765155e-06, "loss": 0.5001, "mean_token_accuracy": 0.8420453071594238, "num_tokens": 28560622.0, "step": 755 }, { "epoch": 0.09617097061442564, "ewc_loss": 0.016845703125, "ewc_loss_diag": 1.1742115020751953e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 2.1802306175231934, "learning_rate": 3.2005086901229338e-06, "loss": 0.5392, "mean_token_accuracy": 0.8297008275985718, "num_tokens": 28594890.0, "step": 756 }, { "epoch": 0.09629818089301616, "ewc_loss": 0.016845703125, "ewc_loss_diag": 1.1742115020751953e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 2.181145191192627, "learning_rate": 3.2047477744807127e-06, "loss": 0.5326, "mean_token_accuracy": 0.8324023485183716, "num_tokens": 28633112.0, "step": 757 }, { "epoch": 0.09642539117160667, "ewc_loss": 0.016845703125, "ewc_loss_diag": 1.1742115020751953e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 2.1750824451446533, "learning_rate": 3.208986858838491e-06, "loss": 0.5602, "mean_token_accuracy": 0.8226885795593262, "num_tokens": 28672195.0, "step": 758 }, { "epoch": 0.09655260145019717, "ewc_loss": 0.016845703125, "ewc_loss_diag": 1.1742115020751953e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 2.177500009536743, "learning_rate": 3.2132259431962696e-06, "loss": 0.4979, "mean_token_accuracy": 0.8465055823326111, "num_tokens": 28710315.0, "step": 759 }, { "epoch": 0.09667981172878769, "ewc_loss": 0.016845703125, "ewc_loss_diag": 1.1742115020751953e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 2.1819825172424316, "learning_rate": 3.2174650275540485e-06, "loss": 0.5377, "mean_token_accuracy": 0.8372197151184082, "num_tokens": 28748713.0, "step": 760 }, { "epoch": 0.0968070220073782, "ewc_loss": 0.016845703125, "ewc_loss_diag": 1.1742115020751953e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 2.167588710784912, "learning_rate": 3.2217041119118274e-06, "loss": 0.5049, "mean_token_accuracy": 0.8407429456710815, "num_tokens": 28789832.0, "step": 761 }, { "epoch": 0.0969342322859687, "ewc_loss": 0.0169677734375, "ewc_loss_diag": 1.1742115020751953e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 2.2074460983276367, "learning_rate": 3.225943196269606e-06, "loss": 0.5283, "mean_token_accuracy": 0.8363009095191956, "num_tokens": 28822781.0, "step": 762 }, { "epoch": 0.09706144256455922, "ewc_loss": 0.016845703125, "ewc_loss_diag": 1.1742115020751953e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 2.1984972953796387, "learning_rate": 3.2301822806273847e-06, "loss": 0.514, "mean_token_accuracy": 0.8353993892669678, "num_tokens": 28857734.0, "step": 763 }, { "epoch": 0.09718865284314973, "ewc_loss": 0.0169677734375, "ewc_loss_diag": 1.1742115020751953e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 2.181025981903076, "learning_rate": 3.2344213649851636e-06, "loss": 0.546, "mean_token_accuracy": 0.8298115134239197, "num_tokens": 28901296.0, "step": 764 }, { "epoch": 0.09731586312174023, "ewc_loss": 0.0169677734375, "ewc_loss_diag": 1.1801719665527344e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 2.178783893585205, "learning_rate": 3.2386604493429425e-06, "loss": 0.5121, "mean_token_accuracy": 0.8372530341148376, "num_tokens": 28942053.0, "step": 765 }, { "epoch": 0.09744307340033075, "ewc_loss": 0.0169677734375, "ewc_loss_diag": 1.1742115020751953e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 2.2083499431610107, "learning_rate": 3.242899533700721e-06, "loss": 0.4839, "mean_token_accuracy": 0.8430832624435425, "num_tokens": 28975320.0, "step": 766 }, { "epoch": 0.09757028367892126, "ewc_loss": 0.01708984375, "ewc_loss_diag": 1.1861324310302734e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 2.1721646785736084, "learning_rate": 3.2471386180585e-06, "loss": 0.5296, "mean_token_accuracy": 0.8288239240646362, "num_tokens": 29019941.0, "step": 767 }, { "epoch": 0.09769749395751176, "ewc_loss": 0.0172119140625, "ewc_loss_diag": 1.1861324310302734e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 2.194458246231079, "learning_rate": 3.2513777024162783e-06, "loss": 0.5603, "mean_token_accuracy": 0.8279435634613037, "num_tokens": 29062608.0, "step": 768 }, { "epoch": 0.09782470423610228, "ewc_loss": 0.0172119140625, "ewc_loss_diag": 1.1861324310302734e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 2.1833529472351074, "learning_rate": 3.255616786774057e-06, "loss": 0.5058, "mean_token_accuracy": 0.8430927991867065, "num_tokens": 29106423.0, "step": 769 }, { "epoch": 0.09795191451469279, "ewc_loss": 0.017333984375, "ewc_loss_diag": 1.1861324310302734e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 2.1878011226654053, "learning_rate": 3.2598558711318356e-06, "loss": 0.4525, "mean_token_accuracy": 0.8561456799507141, "num_tokens": 29148717.0, "step": 770 }, { "epoch": 0.0980791247932833, "ewc_loss": 0.01708984375, "ewc_loss_diag": 1.1980533599853516e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 2.1922109127044678, "learning_rate": 3.2640949554896145e-06, "loss": 0.4994, "mean_token_accuracy": 0.8432014584541321, "num_tokens": 29189464.0, "step": 771 }, { "epoch": 0.0982063350718738, "ewc_loss": 0.0172119140625, "ewc_loss_diag": 1.1980533599853516e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 2.201737642288208, "learning_rate": 3.2683340398473934e-06, "loss": 0.5291, "mean_token_accuracy": 0.8346329927444458, "num_tokens": 29228384.0, "step": 772 }, { "epoch": 0.09833354535046432, "ewc_loss": 0.017333984375, "ewc_loss_diag": 1.1980533599853516e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 2.198638439178467, "learning_rate": 3.272573124205172e-06, "loss": 0.5196, "mean_token_accuracy": 0.8415156602859497, "num_tokens": 29270942.0, "step": 773 }, { "epoch": 0.09846075562905483, "ewc_loss": 0.0174560546875, "ewc_loss_diag": 1.1980533599853516e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 2.200310707092285, "learning_rate": 3.2768122085629507e-06, "loss": 0.4946, "mean_token_accuracy": 0.8453572392463684, "num_tokens": 29313770.0, "step": 774 }, { "epoch": 0.09858796590764533, "ewc_loss": 0.0174560546875, "ewc_loss_diag": 1.1980533599853516e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 2.2167608737945557, "learning_rate": 3.2810512929207296e-06, "loss": 0.4908, "mean_token_accuracy": 0.8442230224609375, "num_tokens": 29347924.0, "step": 775 }, { "epoch": 0.09871517618623585, "ewc_loss": 0.017578125, "ewc_loss_diag": 1.1980533599853516e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 2.20487117767334, "learning_rate": 3.2852903772785085e-06, "loss": 0.5424, "mean_token_accuracy": 0.8306623697280884, "num_tokens": 29386899.0, "step": 776 }, { "epoch": 0.09884238646482636, "ewc_loss": 0.0174560546875, "ewc_loss_diag": 1.1980533599853516e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 2.2104077339172363, "learning_rate": 3.2895294616362865e-06, "loss": 0.4865, "mean_token_accuracy": 0.8509972095489502, "num_tokens": 29423349.0, "step": 777 }, { "epoch": 0.09896959674341686, "ewc_loss": 0.017578125, "ewc_loss_diag": 1.1980533599853516e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 2.2059361934661865, "learning_rate": 3.2937685459940654e-06, "loss": 0.4957, "mean_token_accuracy": 0.8423500061035156, "num_tokens": 29461127.0, "step": 778 }, { "epoch": 0.09909680702200738, "ewc_loss": 0.017333984375, "ewc_loss_diag": 1.1980533599853516e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 2.224424123764038, "learning_rate": 3.2980076303518443e-06, "loss": 0.5788, "mean_token_accuracy": 0.8240475654602051, "num_tokens": 29498335.0, "step": 779 }, { "epoch": 0.09922401730059789, "ewc_loss": 0.0174560546875, "ewc_loss_diag": 1.1980533599853516e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 2.2052512168884277, "learning_rate": 3.302246714709623e-06, "loss": 0.5057, "mean_token_accuracy": 0.8412393927574158, "num_tokens": 29541174.0, "step": 780 }, { "epoch": 0.09935122757918839, "ewc_loss": 0.017578125, "ewc_loss_diag": 1.1980533599853516e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 2.223630666732788, "learning_rate": 3.3064857990674017e-06, "loss": 0.5318, "mean_token_accuracy": 0.8363324403762817, "num_tokens": 29578684.0, "step": 781 }, { "epoch": 0.0994784378577789, "ewc_loss": 0.0174560546875, "ewc_loss_diag": 1.2099742889404297e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 2.2237770557403564, "learning_rate": 3.3107248834251805e-06, "loss": 0.4793, "mean_token_accuracy": 0.8497283458709717, "num_tokens": 29611236.0, "step": 782 }, { "epoch": 0.09960564813636942, "ewc_loss": 0.017578125, "ewc_loss_diag": 1.2099742889404297e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 2.21309494972229, "learning_rate": 3.3149639677829594e-06, "loss": 0.4968, "mean_token_accuracy": 0.8442846536636353, "num_tokens": 29653769.0, "step": 783 }, { "epoch": 0.09973285841495994, "ewc_loss": 0.017578125, "ewc_loss_diag": 1.2099742889404297e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 2.215822458267212, "learning_rate": 3.3192030521407383e-06, "loss": 0.5097, "mean_token_accuracy": 0.8424308896064758, "num_tokens": 29695231.0, "step": 784 }, { "epoch": 0.09986006869355044, "ewc_loss": 0.0174560546875, "ewc_loss_diag": 1.2099742889404297e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 2.2473857402801514, "learning_rate": 3.3234421364985163e-06, "loss": 0.4908, "mean_token_accuracy": 0.8469927310943604, "num_tokens": 29725755.0, "step": 785 }, { "epoch": 0.09998727897214095, "ewc_loss": 0.0174560546875, "ewc_loss_diag": 1.2159347534179688e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 2.2387290000915527, "learning_rate": 3.3276812208562952e-06, "loss": 0.5103, "mean_token_accuracy": 0.8384126424789429, "num_tokens": 29760270.0, "step": 786 }, { "epoch": 0.10011448925073146, "ewc_loss": 0.017578125, "ewc_loss_diag": 1.2159347534179688e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 2.23689341545105, "learning_rate": 3.331920305214074e-06, "loss": 0.5619, "mean_token_accuracy": 0.8314427137374878, "num_tokens": 29797292.0, "step": 787 }, { "epoch": 0.10024169952932196, "ewc_loss": 0.0177001953125, "ewc_loss_diag": 1.2218952178955078e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 2.2250661849975586, "learning_rate": 3.3361593895718526e-06, "loss": 0.4851, "mean_token_accuracy": 0.8517838716506958, "num_tokens": 29833265.0, "step": 788 }, { "epoch": 0.10036890980791248, "ewc_loss": 0.0177001953125, "ewc_loss_diag": 1.2218952178955078e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 2.2192001342773438, "learning_rate": 3.3403984739296315e-06, "loss": 0.5633, "mean_token_accuracy": 0.8243690133094788, "num_tokens": 29876399.0, "step": 789 }, { "epoch": 0.100496120086503, "ewc_loss": 0.0177001953125, "ewc_loss_diag": 1.2278556823730469e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 2.250250816345215, "learning_rate": 3.3446375582874103e-06, "loss": 0.5238, "mean_token_accuracy": 0.8371452689170837, "num_tokens": 29913883.0, "step": 790 }, { "epoch": 0.1006233303650935, "ewc_loss": 0.0177001953125, "ewc_loss_diag": 1.2278556823730469e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 2.2488276958465576, "learning_rate": 3.3488766426451892e-06, "loss": 0.5795, "mean_token_accuracy": 0.8259620666503906, "num_tokens": 29951691.0, "step": 791 }, { "epoch": 0.10075054064368401, "ewc_loss": 0.0177001953125, "ewc_loss_diag": 1.2278556823730469e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 2.2278332710266113, "learning_rate": 3.3531157270029673e-06, "loss": 0.5431, "mean_token_accuracy": 0.8315691351890564, "num_tokens": 29992978.0, "step": 792 }, { "epoch": 0.10087775092227452, "ewc_loss": 0.017822265625, "ewc_loss_diag": 1.239776611328125e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 2.226243019104004, "learning_rate": 3.357354811360746e-06, "loss": 0.5322, "mean_token_accuracy": 0.830856204032898, "num_tokens": 30035602.0, "step": 793 }, { "epoch": 0.10100496120086502, "ewc_loss": 0.017822265625, "ewc_loss_diag": 1.239776611328125e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 2.254704713821411, "learning_rate": 3.361593895718525e-06, "loss": 0.5302, "mean_token_accuracy": 0.8308696150779724, "num_tokens": 30070063.0, "step": 794 }, { "epoch": 0.10113217147945554, "ewc_loss": 0.017822265625, "ewc_loss_diag": 1.239776611328125e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 2.2580978870391846, "learning_rate": 3.365832980076304e-06, "loss": 0.5698, "mean_token_accuracy": 0.8205128908157349, "num_tokens": 30101636.0, "step": 795 }, { "epoch": 0.10125938175804605, "ewc_loss": 0.0179443359375, "ewc_loss_diag": 1.239776611328125e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 2.244338274002075, "learning_rate": 3.3700720644340824e-06, "loss": 0.5272, "mean_token_accuracy": 0.828088104724884, "num_tokens": 30137429.0, "step": 796 }, { "epoch": 0.10138659203663657, "ewc_loss": 0.0179443359375, "ewc_loss_diag": 1.239776611328125e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 2.2401692867279053, "learning_rate": 3.3743111487918613e-06, "loss": 0.5046, "mean_token_accuracy": 0.8433189988136292, "num_tokens": 30176173.0, "step": 797 }, { "epoch": 0.10151380231522707, "ewc_loss": 0.0179443359375, "ewc_loss_diag": 1.239776611328125e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 2.2690281867980957, "learning_rate": 3.37855023314964e-06, "loss": 0.4996, "mean_token_accuracy": 0.8433764576911926, "num_tokens": 30206588.0, "step": 798 }, { "epoch": 0.10164101259381758, "ewc_loss": 0.0179443359375, "ewc_loss_diag": 1.239776611328125e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 2.264235734939575, "learning_rate": 3.382789317507419e-06, "loss": 0.5299, "mean_token_accuracy": 0.8344254493713379, "num_tokens": 30237345.0, "step": 799 }, { "epoch": 0.1017682228724081, "ewc_loss": 0.0179443359375, "ewc_loss_diag": 1.245737075805664e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 2.2588775157928467, "learning_rate": 3.387028401865197e-06, "loss": 0.5414, "mean_token_accuracy": 0.8346958160400391, "num_tokens": 30274543.0, "step": 800 }, { "epoch": 0.1018954331509986, "ewc_loss": 0.0179443359375, "ewc_loss_diag": 1.245737075805664e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 2.244100570678711, "learning_rate": 3.391267486222976e-06, "loss": 0.5166, "mean_token_accuracy": 0.8351266384124756, "num_tokens": 30312687.0, "step": 801 }, { "epoch": 0.10202264342958911, "ewc_loss": 0.0179443359375, "ewc_loss_diag": 1.245737075805664e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 2.2733120918273926, "learning_rate": 3.395506570580755e-06, "loss": 0.5545, "mean_token_accuracy": 0.8238610625267029, "num_tokens": 30343993.0, "step": 802 }, { "epoch": 0.10214985370817962, "ewc_loss": 0.0179443359375, "ewc_loss_diag": 1.245737075805664e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 2.253549814224243, "learning_rate": 3.3997456549385337e-06, "loss": 0.4999, "mean_token_accuracy": 0.8441075086593628, "num_tokens": 30378998.0, "step": 803 }, { "epoch": 0.10227706398677013, "ewc_loss": 0.0181884765625, "ewc_loss_diag": 1.2576580047607422e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 2.328806161880493, "learning_rate": 3.403984739296312e-06, "loss": 0.5699, "mean_token_accuracy": 0.8221633434295654, "num_tokens": 30412205.0, "step": 804 }, { "epoch": 0.10240427426536064, "ewc_loss": 0.0181884765625, "ewc_loss_diag": 1.2576580047607422e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 2.2302474975585938, "learning_rate": 3.408223823654091e-06, "loss": 0.4893, "mean_token_accuracy": 0.8451627492904663, "num_tokens": 30448489.0, "step": 805 }, { "epoch": 0.10253148454395115, "ewc_loss": 0.0184326171875, "ewc_loss_diag": 1.2576580047607422e-05, "ewc_loss_parallel": 1.71661376953125e-05, "grad_norm": 2.2777211666107178, "learning_rate": 3.41246290801187e-06, "loss": 0.543, "mean_token_accuracy": 0.8376320004463196, "num_tokens": 30488741.0, "step": 806 }, { "epoch": 0.10265869482254165, "ewc_loss": 0.01806640625, "ewc_loss_diag": 1.2576580047607422e-05, "ewc_loss_parallel": 1.6808509826660156e-05, "grad_norm": 2.238694190979004, "learning_rate": 3.416701992369648e-06, "loss": 0.54, "mean_token_accuracy": 0.8387666940689087, "num_tokens": 30535715.0, "step": 807 }, { "epoch": 0.10278590510113217, "ewc_loss": 0.018310546875, "ewc_loss_diag": 1.2695789337158203e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 2.2850613594055176, "learning_rate": 3.420941076727427e-06, "loss": 0.4976, "mean_token_accuracy": 0.8465515375137329, "num_tokens": 30574992.0, "step": 808 }, { "epoch": 0.10291311537972268, "ewc_loss": 0.0181884765625, "ewc_loss_diag": 1.2695789337158203e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 2.2583911418914795, "learning_rate": 3.4251801610852058e-06, "loss": 0.5677, "mean_token_accuracy": 0.8194756507873535, "num_tokens": 30615384.0, "step": 809 }, { "epoch": 0.1030403256583132, "ewc_loss": 0.0185546875, "ewc_loss_diag": 1.2695789337158203e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 2.2959470748901367, "learning_rate": 3.4294192454429847e-06, "loss": 0.5182, "mean_token_accuracy": 0.8377759456634521, "num_tokens": 30651011.0, "step": 810 }, { "epoch": 0.1031675359369037, "ewc_loss": 0.0181884765625, "ewc_loss_diag": 1.2755393981933594e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 2.2611351013183594, "learning_rate": 3.433658329800763e-06, "loss": 0.5012, "mean_token_accuracy": 0.8440214395523071, "num_tokens": 30687213.0, "step": 811 }, { "epoch": 0.10329474621549421, "ewc_loss": 0.018310546875, "ewc_loss_diag": 1.2755393981933594e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 2.2957093715667725, "learning_rate": 3.437897414158542e-06, "loss": 0.5392, "mean_token_accuracy": 0.8335437774658203, "num_tokens": 30726950.0, "step": 812 }, { "epoch": 0.10342195649408473, "ewc_loss": 0.0181884765625, "ewc_loss_diag": 1.2755393981933594e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 2.2503726482391357, "learning_rate": 3.442136498516321e-06, "loss": 0.4749, "mean_token_accuracy": 0.8498872518539429, "num_tokens": 30769359.0, "step": 813 }, { "epoch": 0.10354916677267523, "ewc_loss": 0.0184326171875, "ewc_loss_diag": 1.2755393981933594e-05, "ewc_loss_parallel": 1.71661376953125e-05, "grad_norm": 2.29272723197937, "learning_rate": 3.4463755828740998e-06, "loss": 0.5418, "mean_token_accuracy": 0.8329557776451111, "num_tokens": 30812480.0, "step": 814 }, { "epoch": 0.10367637705126574, "ewc_loss": 0.0181884765625, "ewc_loss_diag": 1.2755393981933594e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 2.2749905586242676, "learning_rate": 3.450614667231878e-06, "loss": 0.4819, "mean_token_accuracy": 0.845886766910553, "num_tokens": 30847287.0, "step": 815 }, { "epoch": 0.10380358732985626, "ewc_loss": 0.0186767578125, "ewc_loss_diag": 1.2874603271484375e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 2.285423517227173, "learning_rate": 3.4548537515896567e-06, "loss": 0.531, "mean_token_accuracy": 0.8353329300880432, "num_tokens": 30887050.0, "step": 816 }, { "epoch": 0.10393079760844676, "ewc_loss": 0.018310546875, "ewc_loss_diag": 1.2993812561035156e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 2.2737011909484863, "learning_rate": 3.4590928359474356e-06, "loss": 0.5297, "mean_token_accuracy": 0.8358404636383057, "num_tokens": 30926543.0, "step": 817 }, { "epoch": 0.10405800788703727, "ewc_loss": 0.0184326171875, "ewc_loss_diag": 1.2993812561035156e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 2.288296937942505, "learning_rate": 3.4633319203052145e-06, "loss": 0.5499, "mean_token_accuracy": 0.8282128572463989, "num_tokens": 30966439.0, "step": 818 }, { "epoch": 0.10418521816562779, "ewc_loss": 0.018310546875, "ewc_loss_diag": 1.2993812561035156e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 2.294171094894409, "learning_rate": 3.467571004662993e-06, "loss": 0.4895, "mean_token_accuracy": 0.8463910222053528, "num_tokens": 30999860.0, "step": 819 }, { "epoch": 0.1043124284442183, "ewc_loss": 0.0184326171875, "ewc_loss_diag": 1.3053417205810547e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 2.2842304706573486, "learning_rate": 3.471810089020772e-06, "loss": 0.5356, "mean_token_accuracy": 0.8353480100631714, "num_tokens": 31037040.0, "step": 820 }, { "epoch": 0.1044396387228088, "ewc_loss": 0.0184326171875, "ewc_loss_diag": 1.3172626495361328e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 2.313525676727295, "learning_rate": 3.4760491733785507e-06, "loss": 0.5347, "mean_token_accuracy": 0.827742338180542, "num_tokens": 31071508.0, "step": 821 }, { "epoch": 0.10456684900139931, "ewc_loss": 0.0184326171875, "ewc_loss_diag": 1.3172626495361328e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 2.2801899909973145, "learning_rate": 3.4802882577363296e-06, "loss": 0.5156, "mean_token_accuracy": 0.8330056667327881, "num_tokens": 31107358.0, "step": 822 }, { "epoch": 0.10469405927998983, "ewc_loss": 0.0184326171875, "ewc_loss_diag": 1.3172626495361328e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 2.305121421813965, "learning_rate": 3.4845273420941076e-06, "loss": 0.5273, "mean_token_accuracy": 0.8348672389984131, "num_tokens": 31144986.0, "step": 823 }, { "epoch": 0.10482126955858033, "ewc_loss": 0.018310546875, "ewc_loss_diag": 1.3172626495361328e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 2.3251421451568604, "learning_rate": 3.4887664264518865e-06, "loss": 0.5135, "mean_token_accuracy": 0.8346129655838013, "num_tokens": 31179709.0, "step": 824 }, { "epoch": 0.10494847983717084, "ewc_loss": 0.0184326171875, "ewc_loss_diag": 1.3232231140136719e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 2.296051502227783, "learning_rate": 3.4930055108096654e-06, "loss": 0.5275, "mean_token_accuracy": 0.8374314308166504, "num_tokens": 31222242.0, "step": 825 }, { "epoch": 0.10507569011576136, "ewc_loss": 0.0184326171875, "ewc_loss_diag": 1.3232231140136719e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 2.2755489349365234, "learning_rate": 3.497244595167444e-06, "loss": 0.5041, "mean_token_accuracy": 0.8411840200424194, "num_tokens": 31264793.0, "step": 826 }, { "epoch": 0.10520290039435186, "ewc_loss": 0.0186767578125, "ewc_loss_diag": 1.329183578491211e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 2.3179900646209717, "learning_rate": 3.5014836795252227e-06, "loss": 0.5654, "mean_token_accuracy": 0.8274031281471252, "num_tokens": 31300254.0, "step": 827 }, { "epoch": 0.10533011067294237, "ewc_loss": 0.0186767578125, "ewc_loss_diag": 1.329183578491211e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 2.2874491214752197, "learning_rate": 3.5057227638830016e-06, "loss": 0.5035, "mean_token_accuracy": 0.8418622016906738, "num_tokens": 31339786.0, "step": 828 }, { "epoch": 0.10545732095153289, "ewc_loss": 0.0186767578125, "ewc_loss_diag": 1.329183578491211e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 2.3227336406707764, "learning_rate": 3.5099618482407805e-06, "loss": 0.5422, "mean_token_accuracy": 0.8294374346733093, "num_tokens": 31375100.0, "step": 829 }, { "epoch": 0.10558453123012339, "ewc_loss": 0.0184326171875, "ewc_loss_diag": 1.33514404296875e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 2.2862582206726074, "learning_rate": 3.514200932598559e-06, "loss": 0.5937, "mean_token_accuracy": 0.8175966143608093, "num_tokens": 31419000.0, "step": 830 }, { "epoch": 0.1057117415087139, "ewc_loss": 0.018798828125, "ewc_loss_diag": 1.341104507446289e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 2.317617893218994, "learning_rate": 3.518440016956338e-06, "loss": 0.5354, "mean_token_accuracy": 0.834723949432373, "num_tokens": 31459287.0, "step": 831 }, { "epoch": 0.10583895178730442, "ewc_loss": 0.018798828125, "ewc_loss_diag": 1.341104507446289e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 2.28721284866333, "learning_rate": 3.5226791013141163e-06, "loss": 0.5434, "mean_token_accuracy": 0.8367733955383301, "num_tokens": 31502617.0, "step": 832 }, { "epoch": 0.10596616206589493, "ewc_loss": 0.018798828125, "ewc_loss_diag": 1.341104507446289e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 2.318580389022827, "learning_rate": 3.526918185671895e-06, "loss": 0.558, "mean_token_accuracy": 0.8267871141433716, "num_tokens": 31545159.0, "step": 833 }, { "epoch": 0.10609337234448543, "ewc_loss": 0.0186767578125, "ewc_loss_diag": 1.341104507446289e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 2.3138949871063232, "learning_rate": 3.5311572700296736e-06, "loss": 0.4827, "mean_token_accuracy": 0.8512251973152161, "num_tokens": 31579193.0, "step": 834 }, { "epoch": 0.10622058262307595, "ewc_loss": 0.018798828125, "ewc_loss_diag": 1.3470649719238281e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 2.3164520263671875, "learning_rate": 3.5353963543874525e-06, "loss": 0.5012, "mean_token_accuracy": 0.8448256850242615, "num_tokens": 31618995.0, "step": 835 }, { "epoch": 0.10634779290166646, "ewc_loss": 0.0186767578125, "ewc_loss_diag": 1.3470649719238281e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 2.3470497131347656, "learning_rate": 3.5396354387452314e-06, "loss": 0.5287, "mean_token_accuracy": 0.8355255126953125, "num_tokens": 31657024.0, "step": 836 }, { "epoch": 0.10647500318025696, "ewc_loss": 0.0186767578125, "ewc_loss_diag": 1.3470649719238281e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 2.293377637863159, "learning_rate": 3.5438745231030103e-06, "loss": 0.5186, "mean_token_accuracy": 0.8388954401016235, "num_tokens": 31700521.0, "step": 837 }, { "epoch": 0.10660221345884748, "ewc_loss": 0.0186767578125, "ewc_loss_diag": 1.3589859008789062e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 2.3240489959716797, "learning_rate": 3.5481136074607888e-06, "loss": 0.4869, "mean_token_accuracy": 0.8506271839141846, "num_tokens": 31735942.0, "step": 838 }, { "epoch": 0.10672942373743799, "ewc_loss": 0.018798828125, "ewc_loss_diag": 1.3589859008789062e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 2.309666633605957, "learning_rate": 3.5523526918185676e-06, "loss": 0.5157, "mean_token_accuracy": 0.839939534664154, "num_tokens": 31774092.0, "step": 839 }, { "epoch": 0.10685663401602849, "ewc_loss": 0.018798828125, "ewc_loss_diag": 1.3589859008789062e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 2.3240816593170166, "learning_rate": 3.5565917761763465e-06, "loss": 0.4948, "mean_token_accuracy": 0.8454397916793823, "num_tokens": 31816292.0, "step": 840 }, { "epoch": 0.106983844294619, "ewc_loss": 0.018798828125, "ewc_loss_diag": 1.3589859008789062e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 2.3347010612487793, "learning_rate": 3.560830860534125e-06, "loss": 0.5173, "mean_token_accuracy": 0.83948814868927, "num_tokens": 31848641.0, "step": 841 }, { "epoch": 0.10711105457320952, "ewc_loss": 0.018798828125, "ewc_loss_diag": 1.3649463653564453e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 2.349616527557373, "learning_rate": 3.5650699448919035e-06, "loss": 0.566, "mean_token_accuracy": 0.8238109350204468, "num_tokens": 31881292.0, "step": 842 }, { "epoch": 0.10723826485180002, "ewc_loss": 0.018798828125, "ewc_loss_diag": 1.3649463653564453e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 2.3274970054626465, "learning_rate": 3.5693090292496823e-06, "loss": 0.5244, "mean_token_accuracy": 0.8400362133979797, "num_tokens": 31915810.0, "step": 843 }, { "epoch": 0.10736547513039053, "ewc_loss": 0.0189208984375, "ewc_loss_diag": 1.3649463653564453e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 2.325927257537842, "learning_rate": 3.5735481136074612e-06, "loss": 0.5077, "mean_token_accuracy": 0.842147946357727, "num_tokens": 31953621.0, "step": 844 }, { "epoch": 0.10749268540898105, "ewc_loss": 0.0189208984375, "ewc_loss_diag": 1.3649463653564453e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 2.3342819213867188, "learning_rate": 3.5777871979652397e-06, "loss": 0.5499, "mean_token_accuracy": 0.8280868530273438, "num_tokens": 31988738.0, "step": 845 }, { "epoch": 0.10761989568757156, "ewc_loss": 0.018798828125, "ewc_loss_diag": 1.3649463653564453e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 2.320840358734131, "learning_rate": 3.5820262823230186e-06, "loss": 0.4922, "mean_token_accuracy": 0.8508837819099426, "num_tokens": 32029184.0, "step": 846 }, { "epoch": 0.10774710596616206, "ewc_loss": 0.0189208984375, "ewc_loss_diag": 1.3768672943115234e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 2.347731590270996, "learning_rate": 3.5862653666807975e-06, "loss": 0.5746, "mean_token_accuracy": 0.8206343650817871, "num_tokens": 32064413.0, "step": 847 }, { "epoch": 0.10787431624475258, "ewc_loss": 0.0189208984375, "ewc_loss_diag": 1.3828277587890625e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 2.331928014755249, "learning_rate": 3.5905044510385763e-06, "loss": 0.5386, "mean_token_accuracy": 0.8362449407577515, "num_tokens": 32104043.0, "step": 848 }, { "epoch": 0.10800152652334309, "ewc_loss": 0.0189208984375, "ewc_loss_diag": 1.3768672943115234e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 2.3506643772125244, "learning_rate": 3.5947435353963544e-06, "loss": 0.5243, "mean_token_accuracy": 0.839186429977417, "num_tokens": 32140956.0, "step": 849 }, { "epoch": 0.10812873680193359, "ewc_loss": 0.0189208984375, "ewc_loss_diag": 1.3828277587890625e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 2.3552253246307373, "learning_rate": 3.5989826197541333e-06, "loss": 0.5729, "mean_token_accuracy": 0.8242841958999634, "num_tokens": 32181368.0, "step": 850 }, { "epoch": 0.1082559470805241, "ewc_loss": 0.018798828125, "ewc_loss_diag": 1.3828277587890625e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 2.3328287601470947, "learning_rate": 3.603221704111912e-06, "loss": 0.5222, "mean_token_accuracy": 0.8342679142951965, "num_tokens": 32222714.0, "step": 851 }, { "epoch": 0.10838315735911462, "ewc_loss": 0.01904296875, "ewc_loss_diag": 1.3887882232666016e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 2.323948383331299, "learning_rate": 3.607460788469691e-06, "loss": 0.4882, "mean_token_accuracy": 0.851649820804596, "num_tokens": 32267870.0, "step": 852 }, { "epoch": 0.10851036763770512, "ewc_loss": 0.0189208984375, "ewc_loss_diag": 1.3887882232666016e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 2.331512451171875, "learning_rate": 3.6116998728274695e-06, "loss": 0.464, "mean_token_accuracy": 0.8531028032302856, "num_tokens": 32307408.0, "step": 853 }, { "epoch": 0.10863757791629564, "ewc_loss": 0.0189208984375, "ewc_loss_diag": 1.3887882232666016e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 2.372556686401367, "learning_rate": 3.6159389571852484e-06, "loss": 0.5982, "mean_token_accuracy": 0.8207957148551941, "num_tokens": 32343503.0, "step": 854 }, { "epoch": 0.10876478819488615, "ewc_loss": 0.01904296875, "ewc_loss_diag": 1.3887882232666016e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 2.3311660289764404, "learning_rate": 3.6201780415430273e-06, "loss": 0.5399, "mean_token_accuracy": 0.8329225778579712, "num_tokens": 32386375.0, "step": 855 }, { "epoch": 0.10889199847347665, "ewc_loss": 0.0189208984375, "ewc_loss_diag": 1.3887882232666016e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 2.368717908859253, "learning_rate": 3.624417125900806e-06, "loss": 0.554, "mean_token_accuracy": 0.8299205899238586, "num_tokens": 32424370.0, "step": 856 }, { "epoch": 0.10901920875206716, "ewc_loss": 0.01904296875, "ewc_loss_diag": 1.3887882232666016e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 2.3464465141296387, "learning_rate": 3.628656210258584e-06, "loss": 0.5609, "mean_token_accuracy": 0.8250701427459717, "num_tokens": 32461998.0, "step": 857 }, { "epoch": 0.10914641903065768, "ewc_loss": 0.0191650390625, "ewc_loss_diag": 1.3887882232666016e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 2.360107421875, "learning_rate": 3.632895294616363e-06, "loss": 0.4979, "mean_token_accuracy": 0.8458948135375977, "num_tokens": 32498124.0, "step": 858 }, { "epoch": 0.1092736293092482, "ewc_loss": 0.01904296875, "ewc_loss_diag": 1.3947486877441406e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 2.3407652378082275, "learning_rate": 3.637134378974142e-06, "loss": 0.4914, "mean_token_accuracy": 0.8455355167388916, "num_tokens": 32537735.0, "step": 859 }, { "epoch": 0.1094008395878387, "ewc_loss": 0.0191650390625, "ewc_loss_diag": 1.3947486877441406e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 2.366342306137085, "learning_rate": 3.641373463331921e-06, "loss": 0.537, "mean_token_accuracy": 0.8348111510276794, "num_tokens": 32576219.0, "step": 860 }, { "epoch": 0.10952804986642921, "ewc_loss": 0.0189208984375, "ewc_loss_diag": 1.3947486877441406e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 2.353647232055664, "learning_rate": 3.6456125476896993e-06, "loss": 0.498, "mean_token_accuracy": 0.8383074402809143, "num_tokens": 32612868.0, "step": 861 }, { "epoch": 0.10965526014501972, "ewc_loss": 0.01904296875, "ewc_loss_diag": 1.3947486877441406e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 2.3544960021972656, "learning_rate": 3.649851632047478e-06, "loss": 0.4931, "mean_token_accuracy": 0.8492926359176636, "num_tokens": 32655412.0, "step": 862 }, { "epoch": 0.10978247042361022, "ewc_loss": 0.0191650390625, "ewc_loss_diag": 1.3947486877441406e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 2.352569103240967, "learning_rate": 3.654090716405257e-06, "loss": 0.4608, "mean_token_accuracy": 0.8576119542121887, "num_tokens": 32692579.0, "step": 863 }, { "epoch": 0.10990968070220074, "ewc_loss": 0.01904296875, "ewc_loss_diag": 1.3947486877441406e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 2.373347282409668, "learning_rate": 3.658329800763035e-06, "loss": 0.5564, "mean_token_accuracy": 0.8225787281990051, "num_tokens": 32731642.0, "step": 864 }, { "epoch": 0.11003689098079125, "ewc_loss": 0.01904296875, "ewc_loss_diag": 1.3947486877441406e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 2.3842670917510986, "learning_rate": 3.662568885120814e-06, "loss": 0.5232, "mean_token_accuracy": 0.8350729942321777, "num_tokens": 32765707.0, "step": 865 }, { "epoch": 0.11016410125938175, "ewc_loss": 0.01904296875, "ewc_loss_diag": 1.3947486877441406e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 2.3729236125946045, "learning_rate": 3.666807969478593e-06, "loss": 0.5545, "mean_token_accuracy": 0.8245671391487122, "num_tokens": 32804613.0, "step": 866 }, { "epoch": 0.11029131153797227, "ewc_loss": 0.01904296875, "ewc_loss_diag": 1.4066696166992188e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 2.388324022293091, "learning_rate": 3.6710470538363718e-06, "loss": 0.5399, "mean_token_accuracy": 0.8304988145828247, "num_tokens": 32841198.0, "step": 867 }, { "epoch": 0.11041852181656278, "ewc_loss": 0.019287109375, "ewc_loss_diag": 1.4066696166992188e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 2.362314462661743, "learning_rate": 3.6752861381941502e-06, "loss": 0.5037, "mean_token_accuracy": 0.8435466885566711, "num_tokens": 32878241.0, "step": 868 }, { "epoch": 0.11054573209515328, "ewc_loss": 0.0191650390625, "ewc_loss_diag": 1.4126300811767578e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 2.389017105102539, "learning_rate": 3.679525222551929e-06, "loss": 0.5659, "mean_token_accuracy": 0.8263287544250488, "num_tokens": 32915510.0, "step": 869 }, { "epoch": 0.1106729423737438, "ewc_loss": 0.019287109375, "ewc_loss_diag": 1.4126300811767578e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 2.351109504699707, "learning_rate": 3.683764306909708e-06, "loss": 0.4645, "mean_token_accuracy": 0.8536734580993652, "num_tokens": 32953371.0, "step": 870 }, { "epoch": 0.11080015265233431, "ewc_loss": 0.019287109375, "ewc_loss_diag": 1.424551010131836e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 2.377549171447754, "learning_rate": 3.688003391267487e-06, "loss": 0.5254, "mean_token_accuracy": 0.8392587900161743, "num_tokens": 32998167.0, "step": 871 }, { "epoch": 0.11092736293092482, "ewc_loss": 0.019287109375, "ewc_loss_diag": 1.424551010131836e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 2.3457789421081543, "learning_rate": 3.692242475625265e-06, "loss": 0.4953, "mean_token_accuracy": 0.8436353206634521, "num_tokens": 33046660.0, "step": 872 }, { "epoch": 0.11105457320951533, "ewc_loss": 0.019287109375, "ewc_loss_diag": 1.424551010131836e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 2.4141252040863037, "learning_rate": 3.696481559983044e-06, "loss": 0.5387, "mean_token_accuracy": 0.828999400138855, "num_tokens": 33080672.0, "step": 873 }, { "epoch": 0.11118178348810584, "ewc_loss": 0.019287109375, "ewc_loss_diag": 1.424551010131836e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 2.3661632537841797, "learning_rate": 3.7007206443408227e-06, "loss": 0.4759, "mean_token_accuracy": 0.8500005006790161, "num_tokens": 33117023.0, "step": 874 }, { "epoch": 0.11130899376669635, "ewc_loss": 0.01953125, "ewc_loss_diag": 1.424551010131836e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 2.408392906188965, "learning_rate": 3.7049597286986016e-06, "loss": 0.5945, "mean_token_accuracy": 0.8168392777442932, "num_tokens": 33154843.0, "step": 875 }, { "epoch": 0.11143620404528685, "ewc_loss": 0.019287109375, "ewc_loss_diag": 1.424551010131836e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 2.3856732845306396, "learning_rate": 3.70919881305638e-06, "loss": 0.5469, "mean_token_accuracy": 0.8318960666656494, "num_tokens": 33193500.0, "step": 876 }, { "epoch": 0.11156341432387737, "ewc_loss": 0.0196533203125, "ewc_loss_diag": 1.424551010131836e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 2.3800597190856934, "learning_rate": 3.713437897414159e-06, "loss": 0.5384, "mean_token_accuracy": 0.8335343599319458, "num_tokens": 33240528.0, "step": 877 }, { "epoch": 0.11169062460246788, "ewc_loss": 0.0191650390625, "ewc_loss_diag": 1.424551010131836e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 2.3837778568267822, "learning_rate": 3.717676981771938e-06, "loss": 0.5455, "mean_token_accuracy": 0.8299427032470703, "num_tokens": 33276959.0, "step": 878 }, { "epoch": 0.11181783488105838, "ewc_loss": 0.0194091796875, "ewc_loss_diag": 1.430511474609375e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 2.393127918243408, "learning_rate": 3.721916066129716e-06, "loss": 0.4581, "mean_token_accuracy": 0.8558911085128784, "num_tokens": 33313573.0, "step": 879 }, { "epoch": 0.1119450451596489, "ewc_loss": 0.0194091796875, "ewc_loss_diag": 1.430511474609375e-05, "ewc_loss_parallel": 1.800060272216797e-05, "grad_norm": 2.403791666030884, "learning_rate": 3.7261551504874947e-06, "loss": 0.5311, "mean_token_accuracy": 0.8386242985725403, "num_tokens": 33347408.0, "step": 880 }, { "epoch": 0.11207225543823941, "ewc_loss": 0.0196533203125, "ewc_loss_diag": 1.430511474609375e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 2.3880929946899414, "learning_rate": 3.7303942348452736e-06, "loss": 0.4646, "mean_token_accuracy": 0.8524314761161804, "num_tokens": 33381894.0, "step": 881 }, { "epoch": 0.11219946571682991, "ewc_loss": 0.0196533203125, "ewc_loss_diag": 1.430511474609375e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 2.392280340194702, "learning_rate": 3.7346333192030525e-06, "loss": 0.4777, "mean_token_accuracy": 0.8528180122375488, "num_tokens": 33421746.0, "step": 882 }, { "epoch": 0.11232667599542043, "ewc_loss": 0.019775390625, "ewc_loss_diag": 1.4424324035644531e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 2.379204750061035, "learning_rate": 3.738872403560831e-06, "loss": 0.4995, "mean_token_accuracy": 0.8483281135559082, "num_tokens": 33465203.0, "step": 883 }, { "epoch": 0.11245388627401094, "ewc_loss": 0.0196533203125, "ewc_loss_diag": 1.4424324035644531e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 2.4249818325042725, "learning_rate": 3.74311148791861e-06, "loss": 0.5399, "mean_token_accuracy": 0.8321393728256226, "num_tokens": 33500333.0, "step": 884 }, { "epoch": 0.11258109655260146, "ewc_loss": 0.019775390625, "ewc_loss_diag": 1.4424324035644531e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 2.388692617416382, "learning_rate": 3.7473505722763887e-06, "loss": 0.5584, "mean_token_accuracy": 0.8249961137771606, "num_tokens": 33535180.0, "step": 885 }, { "epoch": 0.11270830683119196, "ewc_loss": 0.019775390625, "ewc_loss_diag": 1.4424324035644531e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 2.414515972137451, "learning_rate": 3.7515896566341676e-06, "loss": 0.5098, "mean_token_accuracy": 0.8395565748214722, "num_tokens": 33571298.0, "step": 886 }, { "epoch": 0.11283551710978247, "ewc_loss": 0.0198974609375, "ewc_loss_diag": 1.4483928680419922e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 2.381276845932007, "learning_rate": 3.7558287409919456e-06, "loss": 0.5355, "mean_token_accuracy": 0.8315510153770447, "num_tokens": 33613693.0, "step": 887 }, { "epoch": 0.11296272738837299, "ewc_loss": 0.0198974609375, "ewc_loss_diag": 1.4483928680419922e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 2.4634835720062256, "learning_rate": 3.7600678253497245e-06, "loss": 0.5001, "mean_token_accuracy": 0.8461685180664062, "num_tokens": 33649760.0, "step": 888 }, { "epoch": 0.11308993766696349, "ewc_loss": 0.019775390625, "ewc_loss_diag": 1.4543533325195312e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 2.3617746829986572, "learning_rate": 3.7643069097075034e-06, "loss": 0.4618, "mean_token_accuracy": 0.8560126423835754, "num_tokens": 33692441.0, "step": 889 }, { "epoch": 0.113217147945554, "ewc_loss": 0.02001953125, "ewc_loss_diag": 1.4543533325195312e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 2.4593145847320557, "learning_rate": 3.7685459940652823e-06, "loss": 0.5152, "mean_token_accuracy": 0.8390053510665894, "num_tokens": 33723034.0, "step": 890 }, { "epoch": 0.11334435822414451, "ewc_loss": 0.0196533203125, "ewc_loss_diag": 1.4543533325195312e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 2.368875026702881, "learning_rate": 3.7727850784230608e-06, "loss": 0.5287, "mean_token_accuracy": 0.837925910949707, "num_tokens": 33764102.0, "step": 891 }, { "epoch": 0.11347156850273502, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 1.4543533325195312e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 2.4494094848632812, "learning_rate": 3.7770241627808396e-06, "loss": 0.4882, "mean_token_accuracy": 0.8482171297073364, "num_tokens": 33803293.0, "step": 892 }, { "epoch": 0.11359877878132553, "ewc_loss": 0.01953125, "ewc_loss_diag": 1.4543533325195312e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 2.38114857673645, "learning_rate": 3.7812632471386185e-06, "loss": 0.4891, "mean_token_accuracy": 0.8465490341186523, "num_tokens": 33841681.0, "step": 893 }, { "epoch": 0.11372598905991604, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 1.4603137969970703e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 2.4645068645477295, "learning_rate": 3.7855023314963974e-06, "loss": 0.5137, "mean_token_accuracy": 0.8345305919647217, "num_tokens": 33877207.0, "step": 894 }, { "epoch": 0.11385319933850654, "ewc_loss": 0.0196533203125, "ewc_loss_diag": 1.4603137969970703e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 2.376455068588257, "learning_rate": 3.7897414158541754e-06, "loss": 0.5286, "mean_token_accuracy": 0.8334587216377258, "num_tokens": 33917703.0, "step": 895 }, { "epoch": 0.11398040961709706, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 1.4662742614746094e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 2.464677095413208, "learning_rate": 3.7939805002119543e-06, "loss": 0.5151, "mean_token_accuracy": 0.8420453071594238, "num_tokens": 33957007.0, "step": 896 }, { "epoch": 0.11410761989568757, "ewc_loss": 0.0196533203125, "ewc_loss_diag": 1.4662742614746094e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 2.3844566345214844, "learning_rate": 3.7982195845697332e-06, "loss": 0.4493, "mean_token_accuracy": 0.8581929802894592, "num_tokens": 33992615.0, "step": 897 }, { "epoch": 0.11423483017427809, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 1.4662742614746094e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 2.4887678623199463, "learning_rate": 3.8024586689275117e-06, "loss": 0.5695, "mean_token_accuracy": 0.8288055658340454, "num_tokens": 34032049.0, "step": 898 }, { "epoch": 0.11436204045286859, "ewc_loss": 0.0196533203125, "ewc_loss_diag": 1.4662742614746094e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 2.3885338306427, "learning_rate": 3.8066977532852906e-06, "loss": 0.4982, "mean_token_accuracy": 0.8426721096038818, "num_tokens": 34065978.0, "step": 899 }, { "epoch": 0.1144892507314591, "ewc_loss": 0.0206298828125, "ewc_loss_diag": 1.4662742614746094e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 2.4690308570861816, "learning_rate": 3.8109368376430694e-06, "loss": 0.4773, "mean_token_accuracy": 0.8517091870307922, "num_tokens": 34103454.0, "step": 900 }, { "epoch": 0.11461646101004962, "ewc_loss": 0.0198974609375, "ewc_loss_diag": 1.4662742614746094e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 2.3975398540496826, "learning_rate": 3.815175922000848e-06, "loss": 0.4551, "mean_token_accuracy": 0.8523856401443481, "num_tokens": 34137970.0, "step": 901 }, { "epoch": 0.11474367128864012, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 1.4662742614746094e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 2.472865581512451, "learning_rate": 3.819415006358627e-06, "loss": 0.55, "mean_token_accuracy": 0.8322175741195679, "num_tokens": 34176997.0, "step": 902 }, { "epoch": 0.11487088156723063, "ewc_loss": 0.0201416015625, "ewc_loss_diag": 1.4662742614746094e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 2.3968474864959717, "learning_rate": 3.823654090716406e-06, "loss": 0.52, "mean_token_accuracy": 0.8352051973342896, "num_tokens": 34215790.0, "step": 903 }, { "epoch": 0.11499809184582115, "ewc_loss": 0.0205078125, "ewc_loss_diag": 1.4662742614746094e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 2.4762163162231445, "learning_rate": 3.8278931750741846e-06, "loss": 0.5309, "mean_token_accuracy": 0.8412539958953857, "num_tokens": 34255212.0, "step": 904 }, { "epoch": 0.11512530212441165, "ewc_loss": 0.02001953125, "ewc_loss_diag": 1.4781951904296875e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 2.412322759628296, "learning_rate": 3.8321322594319634e-06, "loss": 0.4651, "mean_token_accuracy": 0.855314314365387, "num_tokens": 34288432.0, "step": 905 }, { "epoch": 0.11525251240300216, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 1.4781951904296875e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 2.4750945568084717, "learning_rate": 3.8363713437897415e-06, "loss": 0.5748, "mean_token_accuracy": 0.8208174705505371, "num_tokens": 34328991.0, "step": 906 }, { "epoch": 0.11537972268159268, "ewc_loss": 0.020263671875, "ewc_loss_diag": 1.4781951904296875e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 2.4168055057525635, "learning_rate": 3.84061042814752e-06, "loss": 0.5632, "mean_token_accuracy": 0.8292615413665771, "num_tokens": 34366712.0, "step": 907 }, { "epoch": 0.11550693296018319, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 1.4781951904296875e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 2.4724619388580322, "learning_rate": 3.844849512505299e-06, "loss": 0.5503, "mean_token_accuracy": 0.8267927169799805, "num_tokens": 34405757.0, "step": 908 }, { "epoch": 0.11563414323877369, "ewc_loss": 0.020263671875, "ewc_loss_diag": 1.4781951904296875e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 2.4110324382781982, "learning_rate": 3.849088596863078e-06, "loss": 0.5048, "mean_token_accuracy": 0.8448496460914612, "num_tokens": 34448019.0, "step": 909 }, { "epoch": 0.1157613535173642, "ewc_loss": 0.0206298828125, "ewc_loss_diag": 1.4841556549072266e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 2.4633986949920654, "learning_rate": 3.853327681220856e-06, "loss": 0.5218, "mean_token_accuracy": 0.8389304876327515, "num_tokens": 34491237.0, "step": 910 }, { "epoch": 0.11588856379595472, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 1.4841556549072266e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 2.440415382385254, "learning_rate": 3.857566765578635e-06, "loss": 0.4902, "mean_token_accuracy": 0.8465291857719421, "num_tokens": 34522842.0, "step": 911 }, { "epoch": 0.11601577407454522, "ewc_loss": 0.0206298828125, "ewc_loss_diag": 1.4841556549072266e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 2.463670492172241, "learning_rate": 3.861805849936414e-06, "loss": 0.5145, "mean_token_accuracy": 0.843813419342041, "num_tokens": 34560551.0, "step": 912 }, { "epoch": 0.11614298435313573, "ewc_loss": 0.0205078125, "ewc_loss_diag": 1.4960765838623047e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 2.4283149242401123, "learning_rate": 3.866044934294193e-06, "loss": 0.5627, "mean_token_accuracy": 0.8312861919403076, "num_tokens": 34603064.0, "step": 913 }, { "epoch": 0.11627019463172625, "ewc_loss": 0.0206298828125, "ewc_loss_diag": 1.4960765838623047e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 2.4769413471221924, "learning_rate": 3.870284018651972e-06, "loss": 0.5393, "mean_token_accuracy": 0.8331938982009888, "num_tokens": 34640283.0, "step": 914 }, { "epoch": 0.11639740491031675, "ewc_loss": 0.020751953125, "ewc_loss_diag": 1.4960765838623047e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 2.423396348953247, "learning_rate": 3.874523103009751e-06, "loss": 0.4891, "mean_token_accuracy": 0.8495423197746277, "num_tokens": 34679840.0, "step": 915 }, { "epoch": 0.11652461518890726, "ewc_loss": 0.020751953125, "ewc_loss_diag": 1.5020370483398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 2.472118616104126, "learning_rate": 3.8787621873675295e-06, "loss": 0.4766, "mean_token_accuracy": 0.8507025241851807, "num_tokens": 34717045.0, "step": 916 }, { "epoch": 0.11665182546749778, "ewc_loss": 0.0206298828125, "ewc_loss_diag": 1.5020370483398438e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 2.421839952468872, "learning_rate": 3.8830012717253075e-06, "loss": 0.4781, "mean_token_accuracy": 0.8516691327095032, "num_tokens": 34755849.0, "step": 917 }, { "epoch": 0.11677903574608828, "ewc_loss": 0.020751953125, "ewc_loss_diag": 1.5020370483398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 2.47723650932312, "learning_rate": 3.887240356083086e-06, "loss": 0.4535, "mean_token_accuracy": 0.8564120531082153, "num_tokens": 34795202.0, "step": 918 }, { "epoch": 0.11690624602467879, "ewc_loss": 0.0206298828125, "ewc_loss_diag": 1.5020370483398438e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 2.463639259338379, "learning_rate": 3.891479440440865e-06, "loss": 0.4918, "mean_token_accuracy": 0.8492962121963501, "num_tokens": 34824279.0, "step": 919 }, { "epoch": 0.1170334563032693, "ewc_loss": 0.020751953125, "ewc_loss_diag": 1.5020370483398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 2.5077850818634033, "learning_rate": 3.895718524798644e-06, "loss": 0.5328, "mean_token_accuracy": 0.8344546556472778, "num_tokens": 34857631.0, "step": 920 }, { "epoch": 0.11716066658185982, "ewc_loss": 0.020751953125, "ewc_loss_diag": 1.5020370483398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 2.4379682540893555, "learning_rate": 3.899957609156422e-06, "loss": 0.5039, "mean_token_accuracy": 0.8442167639732361, "num_tokens": 34896969.0, "step": 921 }, { "epoch": 0.11728787686045032, "ewc_loss": 0.020751953125, "ewc_loss_diag": 1.5020370483398438e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 2.4717953205108643, "learning_rate": 3.904196693514201e-06, "loss": 0.5212, "mean_token_accuracy": 0.836693525314331, "num_tokens": 34942177.0, "step": 922 }, { "epoch": 0.11741508713904084, "ewc_loss": 0.020751953125, "ewc_loss_diag": 1.519918441772461e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 2.453230857849121, "learning_rate": 3.90843577787198e-06, "loss": 0.5903, "mean_token_accuracy": 0.8186976313591003, "num_tokens": 34980440.0, "step": 923 }, { "epoch": 0.11754229741763135, "ewc_loss": 0.02099609375, "ewc_loss_diag": 1.52587890625e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 2.4920785427093506, "learning_rate": 3.912674862229759e-06, "loss": 0.5418, "mean_token_accuracy": 0.8315906524658203, "num_tokens": 35015343.0, "step": 924 }, { "epoch": 0.11766950769622185, "ewc_loss": 0.020751953125, "ewc_loss_diag": 1.52587890625e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 2.448859691619873, "learning_rate": 3.916913946587537e-06, "loss": 0.533, "mean_token_accuracy": 0.8347456455230713, "num_tokens": 35050265.0, "step": 925 }, { "epoch": 0.11779671797481236, "ewc_loss": 0.02099609375, "ewc_loss_diag": 1.52587890625e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 2.4965126514434814, "learning_rate": 3.921153030945316e-06, "loss": 0.5565, "mean_token_accuracy": 0.8267961144447327, "num_tokens": 35089235.0, "step": 926 }, { "epoch": 0.11792392825340288, "ewc_loss": 0.02099609375, "ewc_loss_diag": 1.52587890625e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 2.4603865146636963, "learning_rate": 3.925392115303095e-06, "loss": 0.5041, "mean_token_accuracy": 0.8428720235824585, "num_tokens": 35122887.0, "step": 927 }, { "epoch": 0.11805113853199338, "ewc_loss": 0.02099609375, "ewc_loss_diag": 1.52587890625e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 2.495828628540039, "learning_rate": 3.9296311996608736e-06, "loss": 0.5084, "mean_token_accuracy": 0.8399757146835327, "num_tokens": 35158686.0, "step": 928 }, { "epoch": 0.1181783488105839, "ewc_loss": 0.0211181640625, "ewc_loss_diag": 1.537799835205078e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 2.4468581676483154, "learning_rate": 3.9338702840186524e-06, "loss": 0.5049, "mean_token_accuracy": 0.8411310911178589, "num_tokens": 35195215.0, "step": 929 }, { "epoch": 0.11830555908917441, "ewc_loss": 0.02099609375, "ewc_loss_diag": 1.537799835205078e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 2.49367618560791, "learning_rate": 3.938109368376431e-06, "loss": 0.5099, "mean_token_accuracy": 0.8429023027420044, "num_tokens": 35236499.0, "step": 930 }, { "epoch": 0.11843276936776491, "ewc_loss": 0.021240234375, "ewc_loss_diag": 1.5497207641601562e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 2.454125165939331, "learning_rate": 3.94234845273421e-06, "loss": 0.5632, "mean_token_accuracy": 0.8261252045631409, "num_tokens": 35278809.0, "step": 931 }, { "epoch": 0.11855997964635542, "ewc_loss": 0.0213623046875, "ewc_loss_diag": 1.537799835205078e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 2.507990837097168, "learning_rate": 3.946587537091989e-06, "loss": 0.5097, "mean_token_accuracy": 0.8388633728027344, "num_tokens": 35318342.0, "step": 932 }, { "epoch": 0.11868718992494594, "ewc_loss": 0.021240234375, "ewc_loss_diag": 1.537799835205078e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 2.4535179138183594, "learning_rate": 3.950826621449767e-06, "loss": 0.508, "mean_token_accuracy": 0.8414291739463806, "num_tokens": 35361802.0, "step": 933 }, { "epoch": 0.11881440020353645, "ewc_loss": 0.0213623046875, "ewc_loss_diag": 1.5497207641601562e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 2.511948585510254, "learning_rate": 3.955065705807546e-06, "loss": 0.4801, "mean_token_accuracy": 0.8500888347625732, "num_tokens": 35401244.0, "step": 934 }, { "epoch": 0.11894161048212695, "ewc_loss": 0.021240234375, "ewc_loss_diag": 1.5497207641601562e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 2.4753830432891846, "learning_rate": 3.959304790165325e-06, "loss": 0.5192, "mean_token_accuracy": 0.8412785530090332, "num_tokens": 35438557.0, "step": 935 }, { "epoch": 0.11906882076071747, "ewc_loss": 0.021240234375, "ewc_loss_diag": 1.5497207641601562e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 2.5235750675201416, "learning_rate": 3.963543874523103e-06, "loss": 0.5076, "mean_token_accuracy": 0.8434761762619019, "num_tokens": 35471686.0, "step": 936 }, { "epoch": 0.11919603103930798, "ewc_loss": 0.021240234375, "ewc_loss_diag": 1.5497207641601562e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 2.4498393535614014, "learning_rate": 3.967782958880882e-06, "loss": 0.5338, "mean_token_accuracy": 0.8341314792633057, "num_tokens": 35514288.0, "step": 937 }, { "epoch": 0.11932324131789848, "ewc_loss": 0.021240234375, "ewc_loss_diag": 1.5497207641601562e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 2.540921449661255, "learning_rate": 3.972022043238661e-06, "loss": 0.5417, "mean_token_accuracy": 0.8322386741638184, "num_tokens": 35547588.0, "step": 938 }, { "epoch": 0.119450451596489, "ewc_loss": 0.021240234375, "ewc_loss_diag": 1.5497207641601562e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 2.4597418308258057, "learning_rate": 3.97626112759644e-06, "loss": 0.4796, "mean_token_accuracy": 0.8527570962905884, "num_tokens": 35582099.0, "step": 939 }, { "epoch": 0.11957766187507951, "ewc_loss": 0.0213623046875, "ewc_loss_diag": 1.5497207641601562e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 2.538114070892334, "learning_rate": 3.980500211954218e-06, "loss": 0.4584, "mean_token_accuracy": 0.8549596667289734, "num_tokens": 35616173.0, "step": 940 }, { "epoch": 0.11970487215367001, "ewc_loss": 0.021240234375, "ewc_loss_diag": 1.5497207641601562e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 2.4640495777130127, "learning_rate": 3.9847392963119965e-06, "loss": 0.5571, "mean_token_accuracy": 0.8296416997909546, "num_tokens": 35654801.0, "step": 941 }, { "epoch": 0.11983208243226053, "ewc_loss": 0.0213623046875, "ewc_loss_diag": 1.5616416931152344e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 2.546433448791504, "learning_rate": 3.988978380669775e-06, "loss": 0.5027, "mean_token_accuracy": 0.8465409278869629, "num_tokens": 35692219.0, "step": 942 }, { "epoch": 0.11995929271085104, "ewc_loss": 0.021240234375, "ewc_loss_diag": 1.5616416931152344e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 2.462332248687744, "learning_rate": 3.993217465027554e-06, "loss": 0.564, "mean_token_accuracy": 0.8281835913658142, "num_tokens": 35732886.0, "step": 943 }, { "epoch": 0.12008650298944154, "ewc_loss": 0.021484375, "ewc_loss_diag": 1.5616416931152344e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 2.5575811862945557, "learning_rate": 3.997456549385333e-06, "loss": 0.4517, "mean_token_accuracy": 0.8590850234031677, "num_tokens": 35766796.0, "step": 944 }, { "epoch": 0.12021371326803205, "ewc_loss": 0.021240234375, "ewc_loss_diag": 1.5616416931152344e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 2.4699277877807617, "learning_rate": 4.001695633743112e-06, "loss": 0.5195, "mean_token_accuracy": 0.8375169634819031, "num_tokens": 35803065.0, "step": 945 }, { "epoch": 0.12034092354662257, "ewc_loss": 0.021484375, "ewc_loss_diag": 1.5616416931152344e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 2.5732500553131104, "learning_rate": 4.005934718100891e-06, "loss": 0.5086, "mean_token_accuracy": 0.841685950756073, "num_tokens": 35836490.0, "step": 946 }, { "epoch": 0.12046813382521308, "ewc_loss": 0.0213623046875, "ewc_loss_diag": 1.5616416931152344e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 2.45658278465271, "learning_rate": 4.01017380245867e-06, "loss": 0.4675, "mean_token_accuracy": 0.8557920455932617, "num_tokens": 35877239.0, "step": 947 }, { "epoch": 0.12059534410380358, "ewc_loss": 0.0216064453125, "ewc_loss_diag": 1.5616416931152344e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 2.5814058780670166, "learning_rate": 4.014412886816448e-06, "loss": 0.5759, "mean_token_accuracy": 0.8230976462364197, "num_tokens": 35917779.0, "step": 948 }, { "epoch": 0.1207225543823941, "ewc_loss": 0.0213623046875, "ewc_loss_diag": 1.5735626220703125e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 2.4564576148986816, "learning_rate": 4.018651971174227e-06, "loss": 0.5008, "mean_token_accuracy": 0.8446016311645508, "num_tokens": 35955083.0, "step": 949 }, { "epoch": 0.12084976466098461, "ewc_loss": 0.0216064453125, "ewc_loss_diag": 1.5735626220703125e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 2.583096504211426, "learning_rate": 4.022891055532006e-06, "loss": 0.5313, "mean_token_accuracy": 0.8345873951911926, "num_tokens": 35999581.0, "step": 950 }, { "epoch": 0.12097697493957511, "ewc_loss": 0.0213623046875, "ewc_loss_diag": 1.5735626220703125e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 2.470147132873535, "learning_rate": 4.0271301398897845e-06, "loss": 0.5734, "mean_token_accuracy": 0.8269869089126587, "num_tokens": 36040891.0, "step": 951 }, { "epoch": 0.12110418521816563, "ewc_loss": 0.021728515625, "ewc_loss_diag": 1.5735626220703125e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 2.605893135070801, "learning_rate": 4.0313692242475625e-06, "loss": 0.4986, "mean_token_accuracy": 0.8427219986915588, "num_tokens": 36074958.0, "step": 952 }, { "epoch": 0.12123139549675614, "ewc_loss": 0.0213623046875, "ewc_loss_diag": 1.5735626220703125e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 2.4620823860168457, "learning_rate": 4.0356083086053414e-06, "loss": 0.5284, "mean_token_accuracy": 0.834476113319397, "num_tokens": 36115211.0, "step": 953 }, { "epoch": 0.12135860577534664, "ewc_loss": 0.0220947265625, "ewc_loss_diag": 1.5735626220703125e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 2.6041626930236816, "learning_rate": 4.03984739296312e-06, "loss": 0.5302, "mean_token_accuracy": 0.8384534120559692, "num_tokens": 36156876.0, "step": 954 }, { "epoch": 0.12148581605393716, "ewc_loss": 0.0216064453125, "ewc_loss_diag": 1.5735626220703125e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 2.4589223861694336, "learning_rate": 4.044086477320898e-06, "loss": 0.5067, "mean_token_accuracy": 0.8400475382804871, "num_tokens": 36197277.0, "step": 955 }, { "epoch": 0.12161302633252767, "ewc_loss": 0.0220947265625, "ewc_loss_diag": 1.5854835510253906e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 2.6321964263916016, "learning_rate": 4.048325561678677e-06, "loss": 0.5404, "mean_token_accuracy": 0.8319859504699707, "num_tokens": 36232839.0, "step": 956 }, { "epoch": 0.12174023661111817, "ewc_loss": 0.021728515625, "ewc_loss_diag": 1.5854835510253906e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 2.468745708465576, "learning_rate": 4.052564646036456e-06, "loss": 0.5303, "mean_token_accuracy": 0.835211455821991, "num_tokens": 36273531.0, "step": 957 }, { "epoch": 0.12186744688970869, "ewc_loss": 0.0220947265625, "ewc_loss_diag": 1.5854835510253906e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 2.6391799449920654, "learning_rate": 4.056803730394235e-06, "loss": 0.5439, "mean_token_accuracy": 0.8297321796417236, "num_tokens": 36309325.0, "step": 958 }, { "epoch": 0.1219946571682992, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 1.5854835510253906e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 2.474428653717041, "learning_rate": 4.061042814752014e-06, "loss": 0.5205, "mean_token_accuracy": 0.8399487733840942, "num_tokens": 36344457.0, "step": 959 }, { "epoch": 0.12212186744688971, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 1.5854835510253906e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 2.6631298065185547, "learning_rate": 4.065281899109793e-06, "loss": 0.5262, "mean_token_accuracy": 0.8340058326721191, "num_tokens": 36376731.0, "step": 960 }, { "epoch": 0.12224907772548022, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 1.5854835510253906e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 2.460618019104004, "learning_rate": 4.069520983467572e-06, "loss": 0.4764, "mean_token_accuracy": 0.85248863697052, "num_tokens": 36415529.0, "step": 961 }, { "epoch": 0.12237628800407073, "ewc_loss": 0.0224609375, "ewc_loss_diag": 1.5854835510253906e-05, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 2.6724798679351807, "learning_rate": 4.0737600678253505e-06, "loss": 0.5056, "mean_token_accuracy": 0.8410937786102295, "num_tokens": 36449522.0, "step": 962 }, { "epoch": 0.12250349828266124, "ewc_loss": 0.0220947265625, "ewc_loss_diag": 1.5974044799804688e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 2.4790029525756836, "learning_rate": 4.077999152183129e-06, "loss": 0.4562, "mean_token_accuracy": 0.8567755818367004, "num_tokens": 36483010.0, "step": 963 }, { "epoch": 0.12263070856125174, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 1.5974044799804688e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 2.6525368690490723, "learning_rate": 4.0822382365409075e-06, "loss": 0.5263, "mean_token_accuracy": 0.8391716480255127, "num_tokens": 36527993.0, "step": 964 }, { "epoch": 0.12275791883984226, "ewc_loss": 0.022216796875, "ewc_loss_diag": 1.5974044799804688e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 2.48056697845459, "learning_rate": 4.086477320898686e-06, "loss": 0.4975, "mean_token_accuracy": 0.8434609770774841, "num_tokens": 36568479.0, "step": 965 }, { "epoch": 0.12288512911843277, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 1.5974044799804688e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 2.667332887649536, "learning_rate": 4.090716405256465e-06, "loss": 0.5373, "mean_token_accuracy": 0.8342182040214539, "num_tokens": 36607196.0, "step": 966 }, { "epoch": 0.12301233939702327, "ewc_loss": 0.022216796875, "ewc_loss_diag": 1.5974044799804688e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 2.512261390686035, "learning_rate": 4.094955489614243e-06, "loss": 0.5031, "mean_token_accuracy": 0.842430591583252, "num_tokens": 36640845.0, "step": 967 }, { "epoch": 0.12313954967561379, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 1.5974044799804688e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 2.651029348373413, "learning_rate": 4.099194573972022e-06, "loss": 0.5023, "mean_token_accuracy": 0.8437588810920715, "num_tokens": 36682182.0, "step": 968 }, { "epoch": 0.1232667599542043, "ewc_loss": 0.022216796875, "ewc_loss_diag": 1.5974044799804688e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 2.5071816444396973, "learning_rate": 4.103433658329801e-06, "loss": 0.5455, "mean_token_accuracy": 0.8326475024223328, "num_tokens": 36723629.0, "step": 969 }, { "epoch": 0.1233939702327948, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 1.5974044799804688e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 2.645382881164551, "learning_rate": 4.10767274268758e-06, "loss": 0.5291, "mean_token_accuracy": 0.8345851302146912, "num_tokens": 36763560.0, "step": 970 }, { "epoch": 0.12352118051138532, "ewc_loss": 0.022216796875, "ewc_loss_diag": 1.5974044799804688e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 2.5241663455963135, "learning_rate": 4.111911827045359e-06, "loss": 0.5942, "mean_token_accuracy": 0.8134046196937561, "num_tokens": 36802267.0, "step": 971 }, { "epoch": 0.12364839078997583, "ewc_loss": 0.0224609375, "ewc_loss_diag": 1.5974044799804688e-05, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 2.6385838985443115, "learning_rate": 4.116150911403137e-06, "loss": 0.4836, "mean_token_accuracy": 0.8523839116096497, "num_tokens": 36840047.0, "step": 972 }, { "epoch": 0.12377560106856635, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 1.5974044799804688e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 2.5326590538024902, "learning_rate": 4.120389995760916e-06, "loss": 0.5466, "mean_token_accuracy": 0.8331276774406433, "num_tokens": 36876788.0, "step": 973 }, { "epoch": 0.12390281134715685, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 1.5974044799804688e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 2.640834093093872, "learning_rate": 4.124629080118695e-06, "loss": 0.4946, "mean_token_accuracy": 0.8442786931991577, "num_tokens": 36909954.0, "step": 974 }, { "epoch": 0.12403002162574736, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 1.5974044799804688e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 2.527695417404175, "learning_rate": 4.1288681644764735e-06, "loss": 0.5348, "mean_token_accuracy": 0.8321448564529419, "num_tokens": 36948371.0, "step": 975 }, { "epoch": 0.12415723190433788, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 1.5974044799804688e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 2.6424400806427, "learning_rate": 4.133107248834252e-06, "loss": 0.545, "mean_token_accuracy": 0.8310684561729431, "num_tokens": 36983304.0, "step": 976 }, { "epoch": 0.12428444218292838, "ewc_loss": 0.022216796875, "ewc_loss_diag": 1.5974044799804688e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 2.517375946044922, "learning_rate": 4.137346333192031e-06, "loss": 0.464, "mean_token_accuracy": 0.8538858890533447, "num_tokens": 37019644.0, "step": 977 }, { "epoch": 0.12441165246151889, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 1.5974044799804688e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 2.6386778354644775, "learning_rate": 4.141585417549809e-06, "loss": 0.535, "mean_token_accuracy": 0.8327075839042664, "num_tokens": 37056906.0, "step": 978 }, { "epoch": 0.1245388627401094, "ewc_loss": 0.0224609375, "ewc_loss_diag": 1.5974044799804688e-05, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 2.530473470687866, "learning_rate": 4.145824501907588e-06, "loss": 0.4912, "mean_token_accuracy": 0.8443488478660583, "num_tokens": 37090396.0, "step": 979 }, { "epoch": 0.1246660730186999, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 1.5974044799804688e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 2.6370904445648193, "learning_rate": 4.150063586265367e-06, "loss": 0.5284, "mean_token_accuracy": 0.8363528251647949, "num_tokens": 37129246.0, "step": 980 }, { "epoch": 0.12479328329729042, "ewc_loss": 0.0224609375, "ewc_loss_diag": 1.5974044799804688e-05, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 2.5211782455444336, "learning_rate": 4.154302670623146e-06, "loss": 0.4757, "mean_token_accuracy": 0.8511008620262146, "num_tokens": 37166344.0, "step": 981 }, { "epoch": 0.12492049357588093, "ewc_loss": 0.022705078125, "ewc_loss_diag": 1.5974044799804688e-05, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 2.63631534576416, "learning_rate": 4.158541754980924e-06, "loss": 0.4605, "mean_token_accuracy": 0.8579144477844238, "num_tokens": 37202827.0, "step": 982 }, { "epoch": 0.12504770385447145, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 1.5974044799804688e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 2.5284578800201416, "learning_rate": 4.162780839338703e-06, "loss": 0.5254, "mean_token_accuracy": 0.8329818844795227, "num_tokens": 37239606.0, "step": 983 }, { "epoch": 0.12517491413306195, "ewc_loss": 0.022705078125, "ewc_loss_diag": 1.5974044799804688e-05, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 2.649942636489868, "learning_rate": 4.167019923696482e-06, "loss": 0.5212, "mean_token_accuracy": 0.8353980779647827, "num_tokens": 37275154.0, "step": 984 }, { "epoch": 0.12530212441165245, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 1.621246337890625e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 2.525322198867798, "learning_rate": 4.171259008054261e-06, "loss": 0.4761, "mean_token_accuracy": 0.8532928228378296, "num_tokens": 37311849.0, "step": 985 }, { "epoch": 0.12542933469024298, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 1.621246337890625e-05, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 2.6509592533111572, "learning_rate": 4.1754980924120395e-06, "loss": 0.5298, "mean_token_accuracy": 0.8377944231033325, "num_tokens": 37353174.0, "step": 986 }, { "epoch": 0.12555654496883348, "ewc_loss": 0.0224609375, "ewc_loss_diag": 1.621246337890625e-05, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 2.5286929607391357, "learning_rate": 4.1797371767698184e-06, "loss": 0.4852, "mean_token_accuracy": 0.850676953792572, "num_tokens": 37392081.0, "step": 987 }, { "epoch": 0.12568375524742398, "ewc_loss": 0.02294921875, "ewc_loss_diag": 1.633167266845703e-05, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 2.656924247741699, "learning_rate": 4.183976261127597e-06, "loss": 0.4838, "mean_token_accuracy": 0.8497351408004761, "num_tokens": 37430639.0, "step": 988 }, { "epoch": 0.1258109655260145, "ewc_loss": 0.0224609375, "ewc_loss_diag": 1.633167266845703e-05, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 2.5445520877838135, "learning_rate": 4.188215345485375e-06, "loss": 0.5691, "mean_token_accuracy": 0.8216540813446045, "num_tokens": 37467126.0, "step": 989 }, { "epoch": 0.125938175804605, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 1.633167266845703e-05, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 2.6610939502716064, "learning_rate": 4.192454429843154e-06, "loss": 0.4997, "mean_token_accuracy": 0.8462071418762207, "num_tokens": 37505657.0, "step": 990 }, { "epoch": 0.12606538608319554, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 1.633167266845703e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 2.5277059078216553, "learning_rate": 4.196693514200933e-06, "loss": 0.5045, "mean_token_accuracy": 0.8442995548248291, "num_tokens": 37549262.0, "step": 991 }, { "epoch": 0.12619259636178604, "ewc_loss": 0.02294921875, "ewc_loss_diag": 1.633167266845703e-05, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 2.6666581630706787, "learning_rate": 4.200932598558712e-06, "loss": 0.5201, "mean_token_accuracy": 0.8363250494003296, "num_tokens": 37589967.0, "step": 992 }, { "epoch": 0.12631980664037654, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 1.633167266845703e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 2.5270276069641113, "learning_rate": 4.20517168291649e-06, "loss": 0.5214, "mean_token_accuracy": 0.8402177095413208, "num_tokens": 37635686.0, "step": 993 }, { "epoch": 0.12644701691896706, "ewc_loss": 0.023193359375, "ewc_loss_diag": 1.6450881958007812e-05, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 2.6873059272766113, "learning_rate": 4.209410767274269e-06, "loss": 0.5625, "mean_token_accuracy": 0.8311663269996643, "num_tokens": 37673753.0, "step": 994 }, { "epoch": 0.12657422719755757, "ewc_loss": 0.022705078125, "ewc_loss_diag": 1.6450881958007812e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 2.549558162689209, "learning_rate": 4.213649851632048e-06, "loss": 0.4727, "mean_token_accuracy": 0.8535137176513672, "num_tokens": 37706345.0, "step": 995 }, { "epoch": 0.12670143747614807, "ewc_loss": 0.023193359375, "ewc_loss_diag": 1.6450881958007812e-05, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 2.675041913986206, "learning_rate": 4.217888935989827e-06, "loss": 0.5427, "mean_token_accuracy": 0.8298376798629761, "num_tokens": 37750356.0, "step": 996 }, { "epoch": 0.1268286477547386, "ewc_loss": 0.02294921875, "ewc_loss_diag": 1.6570091247558594e-05, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 2.543269634246826, "learning_rate": 4.222128020347605e-06, "loss": 0.536, "mean_token_accuracy": 0.8345845341682434, "num_tokens": 37791466.0, "step": 997 }, { "epoch": 0.1269558580333291, "ewc_loss": 0.0235595703125, "ewc_loss_diag": 1.6570091247558594e-05, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 2.6787524223327637, "learning_rate": 4.226367104705384e-06, "loss": 0.5425, "mean_token_accuracy": 0.8308106064796448, "num_tokens": 37837257.0, "step": 998 }, { "epoch": 0.1270830683119196, "ewc_loss": 0.02294921875, "ewc_loss_diag": 1.6570091247558594e-05, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 2.541499614715576, "learning_rate": 4.2306061890631625e-06, "loss": 0.5297, "mean_token_accuracy": 0.8368589282035828, "num_tokens": 37880223.0, "step": 999 }, { "epoch": 0.12721027859051012, "ewc_loss": 0.0235595703125, "ewc_loss_diag": 1.6570091247558594e-05, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 2.7095062732696533, "learning_rate": 4.234845273420941e-06, "loss": 0.6106, "mean_token_accuracy": 0.8191498517990112, "num_tokens": 37917662.0, "step": 1000 }, { "epoch": 0.12733748886910062, "ewc_loss": 0.02294921875, "ewc_loss_diag": 1.6570091247558594e-05, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 2.537557363510132, "learning_rate": 4.23908435777872e-06, "loss": 0.5343, "mean_token_accuracy": 0.8317806124687195, "num_tokens": 37961985.0, "step": 1001 }, { "epoch": 0.12746469914769112, "ewc_loss": 0.023681640625, "ewc_loss_diag": 1.6570091247558594e-05, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 2.7156620025634766, "learning_rate": 4.243323442136499e-06, "loss": 0.5242, "mean_token_accuracy": 0.8414188623428345, "num_tokens": 37998960.0, "step": 1002 }, { "epoch": 0.12759190942628165, "ewc_loss": 0.02294921875, "ewc_loss_diag": 1.6570091247558594e-05, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 2.547990560531616, "learning_rate": 4.247562526494278e-06, "loss": 0.4959, "mean_token_accuracy": 0.8475585579872131, "num_tokens": 38040920.0, "step": 1003 }, { "epoch": 0.12771911970487215, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 1.6570091247558594e-05, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 2.748143196105957, "learning_rate": 4.251801610852057e-06, "loss": 0.5438, "mean_token_accuracy": 0.8327374458312988, "num_tokens": 38071814.0, "step": 1004 }, { "epoch": 0.12784632998346265, "ewc_loss": 0.02294921875, "ewc_loss_diag": 1.6570091247558594e-05, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 2.5438807010650635, "learning_rate": 4.256040695209835e-06, "loss": 0.5606, "mean_token_accuracy": 0.8298880457878113, "num_tokens": 38112784.0, "step": 1005 }, { "epoch": 0.12797354026205318, "ewc_loss": 0.0240478515625, "ewc_loss_diag": 1.6570091247558594e-05, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 2.74106502532959, "learning_rate": 4.260279779567614e-06, "loss": 0.5782, "mean_token_accuracy": 0.8269027471542358, "num_tokens": 38152140.0, "step": 1006 }, { "epoch": 0.12810075054064368, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 1.6570091247558594e-05, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 2.540846347808838, "learning_rate": 4.264518863925393e-06, "loss": 0.4794, "mean_token_accuracy": 0.8511518239974976, "num_tokens": 38188293.0, "step": 1007 }, { "epoch": 0.12822796081923418, "ewc_loss": 0.024169921875, "ewc_loss_diag": 1.6570091247558594e-05, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 2.7566614151000977, "learning_rate": 4.268757948283171e-06, "loss": 0.5358, "mean_token_accuracy": 0.8349645137786865, "num_tokens": 38229136.0, "step": 1008 }, { "epoch": 0.1283551710978247, "ewc_loss": 0.0234375, "ewc_loss_diag": 1.6570091247558594e-05, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 2.544645309448242, "learning_rate": 4.27299703264095e-06, "loss": 0.5518, "mean_token_accuracy": 0.8306527137756348, "num_tokens": 38273330.0, "step": 1009 }, { "epoch": 0.1284823813764152, "ewc_loss": 0.0244140625, "ewc_loss_diag": 1.6570091247558594e-05, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 2.760775327682495, "learning_rate": 4.2772361169987285e-06, "loss": 0.5173, "mean_token_accuracy": 0.8408670425415039, "num_tokens": 38315066.0, "step": 1010 }, { "epoch": 0.1286095916550057, "ewc_loss": 0.023681640625, "ewc_loss_diag": 1.6570091247558594e-05, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 2.565140962600708, "learning_rate": 4.281475201356507e-06, "loss": 0.5211, "mean_token_accuracy": 0.8389321565628052, "num_tokens": 38353430.0, "step": 1011 }, { "epoch": 0.12873680193359624, "ewc_loss": 0.0244140625, "ewc_loss_diag": 1.6689300537109375e-05, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 2.7670743465423584, "learning_rate": 4.2857142857142855e-06, "loss": 0.5544, "mean_token_accuracy": 0.8297538161277771, "num_tokens": 38391919.0, "step": 1012 }, { "epoch": 0.12886401221218674, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 1.6689300537109375e-05, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 2.5752220153808594, "learning_rate": 4.289953370072064e-06, "loss": 0.4946, "mean_token_accuracy": 0.8478772640228271, "num_tokens": 38430104.0, "step": 1013 }, { "epoch": 0.12899122249077727, "ewc_loss": 0.0245361328125, "ewc_loss_diag": 1.6689300537109375e-05, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 2.765016794204712, "learning_rate": 4.294192454429843e-06, "loss": 0.5428, "mean_token_accuracy": 0.8332140445709229, "num_tokens": 38468775.0, "step": 1014 }, { "epoch": 0.12911843276936777, "ewc_loss": 0.023681640625, "ewc_loss_diag": 1.6689300537109375e-05, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 2.560795545578003, "learning_rate": 4.298431538787622e-06, "loss": 0.4295, "mean_token_accuracy": 0.8688720464706421, "num_tokens": 38510330.0, "step": 1015 }, { "epoch": 0.12924564304795827, "ewc_loss": 0.0244140625, "ewc_loss_diag": 1.6689300537109375e-05, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 2.7858364582061768, "learning_rate": 4.302670623145401e-06, "loss": 0.5722, "mean_token_accuracy": 0.8250361680984497, "num_tokens": 38543840.0, "step": 1016 }, { "epoch": 0.1293728533265488, "ewc_loss": 0.023681640625, "ewc_loss_diag": 1.6689300537109375e-05, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 2.5892844200134277, "learning_rate": 4.30690970750318e-06, "loss": 0.5239, "mean_token_accuracy": 0.8409121036529541, "num_tokens": 38581670.0, "step": 1017 }, { "epoch": 0.1295000636051393, "ewc_loss": 0.024658203125, "ewc_loss_diag": 1.6808509826660156e-05, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 2.7862212657928467, "learning_rate": 4.311148791860959e-06, "loss": 0.5116, "mean_token_accuracy": 0.8395086526870728, "num_tokens": 38612642.0, "step": 1018 }, { "epoch": 0.1296272738837298, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 1.6808509826660156e-05, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 2.5907578468322754, "learning_rate": 4.315387876218738e-06, "loss": 0.5486, "mean_token_accuracy": 0.8352618217468262, "num_tokens": 38647579.0, "step": 1019 }, { "epoch": 0.12975448416232033, "ewc_loss": 0.0244140625, "ewc_loss_diag": 1.6808509826660156e-05, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 2.7806243896484375, "learning_rate": 4.319626960576516e-06, "loss": 0.5495, "mean_token_accuracy": 0.8276681303977966, "num_tokens": 38683114.0, "step": 1020 }, { "epoch": 0.12988169444091083, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 1.6808509826660156e-05, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 2.5861423015594482, "learning_rate": 4.3238660449342946e-06, "loss": 0.4919, "mean_token_accuracy": 0.8452955484390259, "num_tokens": 38722990.0, "step": 1021 }, { "epoch": 0.13000890471950133, "ewc_loss": 0.0245361328125, "ewc_loss_diag": 1.6808509826660156e-05, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 2.7602620124816895, "learning_rate": 4.3281051292920735e-06, "loss": 0.5187, "mean_token_accuracy": 0.8408243656158447, "num_tokens": 38765041.0, "step": 1022 }, { "epoch": 0.13013611499809186, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 1.6808509826660156e-05, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 2.5880696773529053, "learning_rate": 4.332344213649852e-06, "loss": 0.4836, "mean_token_accuracy": 0.84869384765625, "num_tokens": 38802241.0, "step": 1023 }, { "epoch": 0.13026332527668236, "ewc_loss": 0.0245361328125, "ewc_loss_diag": 1.6808509826660156e-05, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 2.776735544204712, "learning_rate": 4.33658329800763e-06, "loss": 0.5123, "mean_token_accuracy": 0.8433856964111328, "num_tokens": 38839444.0, "step": 1024 }, { "epoch": 0.13039053555527286, "ewc_loss": 0.02392578125, "ewc_loss_diag": 1.6808509826660156e-05, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 2.5866620540618896, "learning_rate": 4.340822382365409e-06, "loss": 0.4935, "mean_token_accuracy": 0.8471167087554932, "num_tokens": 38882102.0, "step": 1025 }, { "epoch": 0.13051774583386339, "ewc_loss": 0.0247802734375, "ewc_loss_diag": 1.6808509826660156e-05, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 2.7682061195373535, "learning_rate": 4.345061466723188e-06, "loss": 0.4835, "mean_token_accuracy": 0.8524465560913086, "num_tokens": 38925855.0, "step": 1026 }, { "epoch": 0.13064495611245389, "ewc_loss": 0.02392578125, "ewc_loss_diag": 1.6808509826660156e-05, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 2.6088178157806396, "learning_rate": 4.349300551080966e-06, "loss": 0.5397, "mean_token_accuracy": 0.8343798518180847, "num_tokens": 38962293.0, "step": 1027 }, { "epoch": 0.1307721663910444, "ewc_loss": 0.024658203125, "ewc_loss_diag": 1.6927719116210938e-05, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 2.7746031284332275, "learning_rate": 4.353539635438745e-06, "loss": 0.5079, "mean_token_accuracy": 0.8410482406616211, "num_tokens": 38999000.0, "step": 1028 }, { "epoch": 0.13089937666963491, "ewc_loss": 0.02392578125, "ewc_loss_diag": 1.6927719116210938e-05, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 2.6187076568603516, "learning_rate": 4.357778719796524e-06, "loss": 0.5686, "mean_token_accuracy": 0.8222159743309021, "num_tokens": 39032819.0, "step": 1029 }, { "epoch": 0.13102658694822542, "ewc_loss": 0.0245361328125, "ewc_loss_diag": 1.6927719116210938e-05, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 2.759439706802368, "learning_rate": 4.362017804154303e-06, "loss": 0.4908, "mean_token_accuracy": 0.8500844836235046, "num_tokens": 39070247.0, "step": 1030 }, { "epoch": 0.13115379722681592, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 1.6927719116210938e-05, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 2.609788417816162, "learning_rate": 4.366256888512082e-06, "loss": 0.4879, "mean_token_accuracy": 0.8505134582519531, "num_tokens": 39105310.0, "step": 1031 }, { "epoch": 0.13128100750540644, "ewc_loss": 0.0245361328125, "ewc_loss_diag": 1.6927719116210938e-05, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 2.801053285598755, "learning_rate": 4.370495972869861e-06, "loss": 0.519, "mean_token_accuracy": 0.8387925624847412, "num_tokens": 39144218.0, "step": 1032 }, { "epoch": 0.13140821778399694, "ewc_loss": 0.024169921875, "ewc_loss_diag": 1.6927719116210938e-05, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 2.6334211826324463, "learning_rate": 4.3747350572276395e-06, "loss": 0.5091, "mean_token_accuracy": 0.8414293527603149, "num_tokens": 39177248.0, "step": 1033 }, { "epoch": 0.13153542806258745, "ewc_loss": 0.024658203125, "ewc_loss_diag": 1.6927719116210938e-05, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 2.760084629058838, "learning_rate": 4.378974141585418e-06, "loss": 0.5203, "mean_token_accuracy": 0.8399556875228882, "num_tokens": 39214230.0, "step": 1034 }, { "epoch": 0.13166263834117797, "ewc_loss": 0.02392578125, "ewc_loss_diag": 1.6927719116210938e-05, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 2.6242098808288574, "learning_rate": 4.383213225943196e-06, "loss": 0.5103, "mean_token_accuracy": 0.8404106497764587, "num_tokens": 39252189.0, "step": 1035 }, { "epoch": 0.13178984861976847, "ewc_loss": 0.024658203125, "ewc_loss_diag": 1.6927719116210938e-05, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 2.7533862590789795, "learning_rate": 4.387452310300975e-06, "loss": 0.4611, "mean_token_accuracy": 0.8579563498497009, "num_tokens": 39288058.0, "step": 1036 }, { "epoch": 0.13191705889835897, "ewc_loss": 0.024169921875, "ewc_loss_diag": 1.6927719116210938e-05, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 2.6159048080444336, "learning_rate": 4.391691394658754e-06, "loss": 0.469, "mean_token_accuracy": 0.8525491952896118, "num_tokens": 39322157.0, "step": 1037 }, { "epoch": 0.1320442691769495, "ewc_loss": 0.024658203125, "ewc_loss_diag": 1.704692840576172e-05, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 2.76177978515625, "learning_rate": 4.395930479016533e-06, "loss": 0.4858, "mean_token_accuracy": 0.8474866151809692, "num_tokens": 39357678.0, "step": 1038 }, { "epoch": 0.13217147945554, "ewc_loss": 0.0244140625, "ewc_loss_diag": 1.704692840576172e-05, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 2.643573522567749, "learning_rate": 4.400169563374311e-06, "loss": 0.5061, "mean_token_accuracy": 0.843393862247467, "num_tokens": 39398909.0, "step": 1039 }, { "epoch": 0.13229868973413053, "ewc_loss": 0.0247802734375, "ewc_loss_diag": 1.704692840576172e-05, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 2.761366367340088, "learning_rate": 4.40440864773209e-06, "loss": 0.5841, "mean_token_accuracy": 0.821926474571228, "num_tokens": 39437934.0, "step": 1040 }, { "epoch": 0.13242590001272103, "ewc_loss": 0.0244140625, "ewc_loss_diag": 1.704692840576172e-05, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 2.6205615997314453, "learning_rate": 4.408647732089869e-06, "loss": 0.5035, "mean_token_accuracy": 0.8400404453277588, "num_tokens": 39471772.0, "step": 1041 }, { "epoch": 0.13255311029131153, "ewc_loss": 0.02490234375, "ewc_loss_diag": 1.704692840576172e-05, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 2.7655704021453857, "learning_rate": 4.412886816447648e-06, "loss": 0.5182, "mean_token_accuracy": 0.8434047102928162, "num_tokens": 39512398.0, "step": 1042 }, { "epoch": 0.13268032056990206, "ewc_loss": 0.0245361328125, "ewc_loss_diag": 1.704692840576172e-05, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 2.616058826446533, "learning_rate": 4.417125900805427e-06, "loss": 0.5121, "mean_token_accuracy": 0.8424605131149292, "num_tokens": 39554735.0, "step": 1043 }, { "epoch": 0.13280753084849256, "ewc_loss": 0.02490234375, "ewc_loss_diag": 1.704692840576172e-05, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 2.7780251502990723, "learning_rate": 4.4213649851632055e-06, "loss": 0.562, "mean_token_accuracy": 0.8306064009666443, "num_tokens": 39595397.0, "step": 1044 }, { "epoch": 0.13293474112708306, "ewc_loss": 0.0244140625, "ewc_loss_diag": 1.704692840576172e-05, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 2.6159746646881104, "learning_rate": 4.4256040695209836e-06, "loss": 0.5445, "mean_token_accuracy": 0.8339195251464844, "num_tokens": 39632474.0, "step": 1045 }, { "epoch": 0.1330619514056736, "ewc_loss": 0.025146484375, "ewc_loss_diag": 1.704692840576172e-05, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 2.795761823654175, "learning_rate": 4.4298431538787625e-06, "loss": 0.5326, "mean_token_accuracy": 0.8326354026794434, "num_tokens": 39666408.0, "step": 1046 }, { "epoch": 0.1331891616842641, "ewc_loss": 0.0245361328125, "ewc_loss_diag": 1.704692840576172e-05, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 2.610013484954834, "learning_rate": 4.434082238236541e-06, "loss": 0.5679, "mean_token_accuracy": 0.8236794471740723, "num_tokens": 39709033.0, "step": 1047 }, { "epoch": 0.1333163719628546, "ewc_loss": 0.025390625, "ewc_loss_diag": 1.704692840576172e-05, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 2.8022854328155518, "learning_rate": 4.43832132259432e-06, "loss": 0.5401, "mean_token_accuracy": 0.8358721733093262, "num_tokens": 39749732.0, "step": 1048 }, { "epoch": 0.13344358224144512, "ewc_loss": 0.0245361328125, "ewc_loss_diag": 1.704692840576172e-05, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 2.6203410625457764, "learning_rate": 4.442560406952099e-06, "loss": 0.5474, "mean_token_accuracy": 0.829390823841095, "num_tokens": 39788125.0, "step": 1049 }, { "epoch": 0.13357079252003562, "ewc_loss": 0.0255126953125, "ewc_loss_diag": 1.704692840576172e-05, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 2.8007657527923584, "learning_rate": 4.446799491309877e-06, "loss": 0.4908, "mean_token_accuracy": 0.8501118421554565, "num_tokens": 39830125.0, "step": 1050 }, { "epoch": 0.13369800279862612, "ewc_loss": 0.0245361328125, "ewc_loss_diag": 1.704692840576172e-05, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 2.6210129261016846, "learning_rate": 4.451038575667656e-06, "loss": 0.5453, "mean_token_accuracy": 0.8297954797744751, "num_tokens": 39868236.0, "step": 1051 }, { "epoch": 0.13382521307721665, "ewc_loss": 0.025634765625, "ewc_loss_diag": 1.704692840576172e-05, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 2.816333055496216, "learning_rate": 4.455277660025435e-06, "loss": 0.5141, "mean_token_accuracy": 0.8399142622947693, "num_tokens": 39913254.0, "step": 1052 }, { "epoch": 0.13395242335580715, "ewc_loss": 0.0245361328125, "ewc_loss_diag": 1.704692840576172e-05, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 2.6386308670043945, "learning_rate": 4.459516744383214e-06, "loss": 0.5237, "mean_token_accuracy": 0.8362716436386108, "num_tokens": 39948242.0, "step": 1053 }, { "epoch": 0.13407963363439765, "ewc_loss": 0.0255126953125, "ewc_loss_diag": 1.704692840576172e-05, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 2.8275821208953857, "learning_rate": 4.463755828740992e-06, "loss": 0.539, "mean_token_accuracy": 0.8345897793769836, "num_tokens": 39989196.0, "step": 1054 }, { "epoch": 0.13420684391298818, "ewc_loss": 0.0245361328125, "ewc_loss_diag": 1.704692840576172e-05, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 2.633981466293335, "learning_rate": 4.467994913098771e-06, "loss": 0.5367, "mean_token_accuracy": 0.8415429592132568, "num_tokens": 40028662.0, "step": 1055 }, { "epoch": 0.13433405419157868, "ewc_loss": 0.025634765625, "ewc_loss_diag": 1.704692840576172e-05, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 2.8327934741973877, "learning_rate": 4.47223399745655e-06, "loss": 0.5396, "mean_token_accuracy": 0.8334540128707886, "num_tokens": 40066178.0, "step": 1056 }, { "epoch": 0.13446126447016918, "ewc_loss": 0.0245361328125, "ewc_loss_diag": 1.704692840576172e-05, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 2.6236023902893066, "learning_rate": 4.4764730818143285e-06, "loss": 0.5285, "mean_token_accuracy": 0.8351101875305176, "num_tokens": 40106908.0, "step": 1057 }, { "epoch": 0.1345884747487597, "ewc_loss": 0.025634765625, "ewc_loss_diag": 1.704692840576172e-05, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 2.843810796737671, "learning_rate": 4.480712166172107e-06, "loss": 0.4929, "mean_token_accuracy": 0.847602903842926, "num_tokens": 40146872.0, "step": 1058 }, { "epoch": 0.1347156850273502, "ewc_loss": 0.024658203125, "ewc_loss_diag": 1.704692840576172e-05, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 2.6268410682678223, "learning_rate": 4.484951250529886e-06, "loss": 0.5221, "mean_token_accuracy": 0.8382750153541565, "num_tokens": 40186717.0, "step": 1059 }, { "epoch": 0.1348428953059407, "ewc_loss": 0.0260009765625, "ewc_loss_diag": 1.704692840576172e-05, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 2.867549419403076, "learning_rate": 4.489190334887665e-06, "loss": 0.5626, "mean_token_accuracy": 0.8322063684463501, "num_tokens": 40227119.0, "step": 1060 }, { "epoch": 0.13497010558453124, "ewc_loss": 0.0247802734375, "ewc_loss_diag": 1.704692840576172e-05, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 2.6510584354400635, "learning_rate": 4.493429419245444e-06, "loss": 0.5386, "mean_token_accuracy": 0.8349220752716064, "num_tokens": 40262302.0, "step": 1061 }, { "epoch": 0.13509731586312174, "ewc_loss": 0.026123046875, "ewc_loss_diag": 1.71661376953125e-05, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 2.8714301586151123, "learning_rate": 4.497668503603222e-06, "loss": 0.4865, "mean_token_accuracy": 0.8498643636703491, "num_tokens": 40296783.0, "step": 1062 }, { "epoch": 0.13522452614171224, "ewc_loss": 0.024658203125, "ewc_loss_diag": 1.71661376953125e-05, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 2.650665760040283, "learning_rate": 4.501907587961001e-06, "loss": 0.5475, "mean_token_accuracy": 0.8289364576339722, "num_tokens": 40331323.0, "step": 1063 }, { "epoch": 0.13535173642030277, "ewc_loss": 0.0260009765625, "ewc_loss_diag": 1.71661376953125e-05, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 2.863461971282959, "learning_rate": 4.50614667231878e-06, "loss": 0.5209, "mean_token_accuracy": 0.8397592306137085, "num_tokens": 40369965.0, "step": 1064 }, { "epoch": 0.13547894669889327, "ewc_loss": 0.02490234375, "ewc_loss_diag": 1.71661376953125e-05, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 2.640979766845703, "learning_rate": 4.510385756676558e-06, "loss": 0.5697, "mean_token_accuracy": 0.8282709121704102, "num_tokens": 40410820.0, "step": 1065 }, { "epoch": 0.1356061569774838, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 1.71661376953125e-05, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 2.8820292949676514, "learning_rate": 4.514624841034337e-06, "loss": 0.4758, "mean_token_accuracy": 0.8533430695533752, "num_tokens": 40446942.0, "step": 1066 }, { "epoch": 0.1357333672560743, "ewc_loss": 0.02490234375, "ewc_loss_diag": 1.71661376953125e-05, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 2.6474766731262207, "learning_rate": 4.518863925392116e-06, "loss": 0.49, "mean_token_accuracy": 0.8481440544128418, "num_tokens": 40484684.0, "step": 1067 }, { "epoch": 0.1358605775346648, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 1.71661376953125e-05, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 2.8837802410125732, "learning_rate": 4.5231030097498945e-06, "loss": 0.4712, "mean_token_accuracy": 0.8516985177993774, "num_tokens": 40521164.0, "step": 1068 }, { "epoch": 0.13598778781325532, "ewc_loss": 0.0250244140625, "ewc_loss_diag": 1.71661376953125e-05, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 2.6469614505767822, "learning_rate": 4.5273420941076726e-06, "loss": 0.507, "mean_token_accuracy": 0.8420136570930481, "num_tokens": 40566192.0, "step": 1069 }, { "epoch": 0.13611499809184582, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 1.71661376953125e-05, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 2.9022610187530518, "learning_rate": 4.5315811784654514e-06, "loss": 0.5428, "mean_token_accuracy": 0.8331742286682129, "num_tokens": 40605534.0, "step": 1070 }, { "epoch": 0.13624220837043632, "ewc_loss": 0.02490234375, "ewc_loss_diag": 1.71661376953125e-05, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 2.659850597381592, "learning_rate": 4.53582026282323e-06, "loss": 0.4946, "mean_token_accuracy": 0.845375657081604, "num_tokens": 40645478.0, "step": 1071 }, { "epoch": 0.13636941864902685, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 1.728534698486328e-05, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 2.9202053546905518, "learning_rate": 4.540059347181009e-06, "loss": 0.5805, "mean_token_accuracy": 0.8208853006362915, "num_tokens": 40684591.0, "step": 1072 }, { "epoch": 0.13649662892761735, "ewc_loss": 0.0250244140625, "ewc_loss_diag": 1.728534698486328e-05, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 2.6701364517211914, "learning_rate": 4.544298431538788e-06, "loss": 0.5133, "mean_token_accuracy": 0.8409220576286316, "num_tokens": 40724437.0, "step": 1073 }, { "epoch": 0.13662383920620785, "ewc_loss": 0.026611328125, "ewc_loss_diag": 1.728534698486328e-05, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 2.9248955249786377, "learning_rate": 4.548537515896567e-06, "loss": 0.5604, "mean_token_accuracy": 0.8270227909088135, "num_tokens": 40758324.0, "step": 1074 }, { "epoch": 0.13675104948479838, "ewc_loss": 0.0250244140625, "ewc_loss_diag": 1.728534698486328e-05, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 2.6901910305023193, "learning_rate": 4.552776600254346e-06, "loss": 0.542, "mean_token_accuracy": 0.8336470127105713, "num_tokens": 40794182.0, "step": 1075 }, { "epoch": 0.13687825976338888, "ewc_loss": 0.026611328125, "ewc_loss_diag": 1.728534698486328e-05, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 2.906147003173828, "learning_rate": 4.557015684612125e-06, "loss": 0.4933, "mean_token_accuracy": 0.8488700985908508, "num_tokens": 40832293.0, "step": 1076 }, { "epoch": 0.13700547004197938, "ewc_loss": 0.025146484375, "ewc_loss_diag": 1.728534698486328e-05, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 2.677809953689575, "learning_rate": 4.561254768969903e-06, "loss": 0.4874, "mean_token_accuracy": 0.8475760221481323, "num_tokens": 40871078.0, "step": 1077 }, { "epoch": 0.1371326803205699, "ewc_loss": 0.026611328125, "ewc_loss_diag": 1.728534698486328e-05, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 2.9028942584991455, "learning_rate": 4.565493853327682e-06, "loss": 0.4993, "mean_token_accuracy": 0.8477779030799866, "num_tokens": 40909061.0, "step": 1078 }, { "epoch": 0.1372598905991604, "ewc_loss": 0.0250244140625, "ewc_loss_diag": 1.728534698486328e-05, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 2.69956111907959, "learning_rate": 4.5697329376854606e-06, "loss": 0.536, "mean_token_accuracy": 0.8394045829772949, "num_tokens": 40945623.0, "step": 1079 }, { "epoch": 0.1373871008777509, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 1.728534698486328e-05, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 2.9051709175109863, "learning_rate": 4.573972022043239e-06, "loss": 0.6023, "mean_token_accuracy": 0.8136733770370483, "num_tokens": 40985009.0, "step": 1080 }, { "epoch": 0.13751431115634144, "ewc_loss": 0.0250244140625, "ewc_loss_diag": 1.728534698486328e-05, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 2.6885247230529785, "learning_rate": 4.5782111064010175e-06, "loss": 0.5499, "mean_token_accuracy": 0.832263171672821, "num_tokens": 41027377.0, "step": 1081 }, { "epoch": 0.13764152143493194, "ewc_loss": 0.026611328125, "ewc_loss_diag": 1.728534698486328e-05, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 2.8979721069335938, "learning_rate": 4.582450190758796e-06, "loss": 0.5552, "mean_token_accuracy": 0.8297070264816284, "num_tokens": 41068816.0, "step": 1082 }, { "epoch": 0.13776873171352244, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 1.728534698486328e-05, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 2.6939384937286377, "learning_rate": 4.586689275116575e-06, "loss": 0.4745, "mean_token_accuracy": 0.852766752243042, "num_tokens": 41106603.0, "step": 1083 }, { "epoch": 0.13789594199211297, "ewc_loss": 0.026611328125, "ewc_loss_diag": 1.728534698486328e-05, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 2.9064297676086426, "learning_rate": 4.590928359474353e-06, "loss": 0.5652, "mean_token_accuracy": 0.8268067240715027, "num_tokens": 41142862.0, "step": 1084 }, { "epoch": 0.13802315227070347, "ewc_loss": 0.0250244140625, "ewc_loss_diag": 1.728534698486328e-05, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 2.688319206237793, "learning_rate": 4.595167443832132e-06, "loss": 0.5131, "mean_token_accuracy": 0.8426512479782104, "num_tokens": 41182761.0, "step": 1085 }, { "epoch": 0.13815036254929397, "ewc_loss": 0.02685546875, "ewc_loss_diag": 1.728534698486328e-05, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 2.903167486190796, "learning_rate": 4.599406528189911e-06, "loss": 0.4656, "mean_token_accuracy": 0.8568284511566162, "num_tokens": 41221318.0, "step": 1086 }, { "epoch": 0.1382775728278845, "ewc_loss": 0.0250244140625, "ewc_loss_diag": 1.728534698486328e-05, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 2.689485788345337, "learning_rate": 4.60364561254769e-06, "loss": 0.504, "mean_token_accuracy": 0.8401232957839966, "num_tokens": 41257748.0, "step": 1087 }, { "epoch": 0.138404783106475, "ewc_loss": 0.027099609375, "ewc_loss_diag": 1.728534698486328e-05, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 2.9137887954711914, "learning_rate": 4.607884696905469e-06, "loss": 0.5084, "mean_token_accuracy": 0.8427374362945557, "num_tokens": 41299287.0, "step": 1088 }, { "epoch": 0.1385319933850655, "ewc_loss": 0.025146484375, "ewc_loss_diag": 1.7404556274414062e-05, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 2.6923422813415527, "learning_rate": 4.612123781263248e-06, "loss": 0.5472, "mean_token_accuracy": 0.8299573659896851, "num_tokens": 41344600.0, "step": 1089 }, { "epoch": 0.13865920366365603, "ewc_loss": 0.027099609375, "ewc_loss_diag": 1.7404556274414062e-05, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 2.9233102798461914, "learning_rate": 4.616362865621027e-06, "loss": 0.442, "mean_token_accuracy": 0.8605016469955444, "num_tokens": 41380681.0, "step": 1090 }, { "epoch": 0.13878641394224653, "ewc_loss": 0.025146484375, "ewc_loss_diag": 1.7404556274414062e-05, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 2.7012603282928467, "learning_rate": 4.6206019499788055e-06, "loss": 0.5364, "mean_token_accuracy": 0.8369665741920471, "num_tokens": 41418751.0, "step": 1091 }, { "epoch": 0.13891362422083706, "ewc_loss": 0.0272216796875, "ewc_loss_diag": 1.7404556274414062e-05, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 2.9361917972564697, "learning_rate": 4.6248410343365835e-06, "loss": 0.5853, "mean_token_accuracy": 0.8178845643997192, "num_tokens": 41455617.0, "step": 1092 }, { "epoch": 0.13904083449942756, "ewc_loss": 0.0255126953125, "ewc_loss_diag": 1.7404556274414062e-05, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 2.694626808166504, "learning_rate": 4.629080118694362e-06, "loss": 0.4944, "mean_token_accuracy": 0.847452700138092, "num_tokens": 41494480.0, "step": 1093 }, { "epoch": 0.13916804477801806, "ewc_loss": 0.027099609375, "ewc_loss_diag": 1.7404556274414062e-05, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 2.9403889179229736, "learning_rate": 4.633319203052141e-06, "loss": 0.511, "mean_token_accuracy": 0.8443673849105835, "num_tokens": 41529294.0, "step": 1094 }, { "epoch": 0.13929525505660859, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 1.7642974853515625e-05, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 2.6935017108917236, "learning_rate": 4.63755828740992e-06, "loss": 0.4851, "mean_token_accuracy": 0.8491894006729126, "num_tokens": 41570787.0, "step": 1095 }, { "epoch": 0.1394224653351991, "ewc_loss": 0.0274658203125, "ewc_loss_diag": 1.7642974853515625e-05, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 2.9693493843078613, "learning_rate": 4.641797371767698e-06, "loss": 0.5014, "mean_token_accuracy": 0.848051130771637, "num_tokens": 41606596.0, "step": 1096 }, { "epoch": 0.1395496756137896, "ewc_loss": 0.0257568359375, "ewc_loss_diag": 1.7642974853515625e-05, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 2.698967218399048, "learning_rate": 4.646036456125477e-06, "loss": 0.5218, "mean_token_accuracy": 0.8353928327560425, "num_tokens": 41645317.0, "step": 1097 }, { "epoch": 0.13967688589238011, "ewc_loss": 0.0277099609375, "ewc_loss_diag": 1.7642974853515625e-05, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 2.9694314002990723, "learning_rate": 4.650275540483256e-06, "loss": 0.4891, "mean_token_accuracy": 0.8455973863601685, "num_tokens": 41680424.0, "step": 1098 }, { "epoch": 0.13980409617097062, "ewc_loss": 0.0260009765625, "ewc_loss_diag": 1.7642974853515625e-05, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 2.725059747695923, "learning_rate": 4.654514624841035e-06, "loss": 0.5141, "mean_token_accuracy": 0.8423972129821777, "num_tokens": 41712492.0, "step": 1099 }, { "epoch": 0.13993130644956112, "ewc_loss": 0.0277099609375, "ewc_loss_diag": 1.7642974853515625e-05, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 2.9672443866729736, "learning_rate": 4.658753709198813e-06, "loss": 0.5186, "mean_token_accuracy": 0.8416897058486938, "num_tokens": 41747666.0, "step": 1100 }, { "epoch": 0.14005851672815164, "ewc_loss": 0.026123046875, "ewc_loss_diag": 1.7642974853515625e-05, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 2.733738660812378, "learning_rate": 4.662992793556592e-06, "loss": 0.4965, "mean_token_accuracy": 0.8473932147026062, "num_tokens": 41784638.0, "step": 1101 }, { "epoch": 0.14018572700674214, "ewc_loss": 0.02783203125, "ewc_loss_diag": 1.7642974853515625e-05, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 3.0480926036834717, "learning_rate": 4.667231877914371e-06, "loss": 0.5859, "mean_token_accuracy": 0.820793628692627, "num_tokens": 41826547.0, "step": 1102 }, { "epoch": 0.14031293728533265, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 1.7762184143066406e-05, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 2.7350027561187744, "learning_rate": 4.6714709622721496e-06, "loss": 0.5315, "mean_token_accuracy": 0.8363637328147888, "num_tokens": 41859526.0, "step": 1103 }, { "epoch": 0.14044014756392317, "ewc_loss": 0.028076171875, "ewc_loss_diag": 1.7642974853515625e-05, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 2.971156597137451, "learning_rate": 4.6757100466299284e-06, "loss": 0.534, "mean_token_accuracy": 0.837502121925354, "num_tokens": 41901177.0, "step": 1104 }, { "epoch": 0.14056735784251367, "ewc_loss": 0.0263671875, "ewc_loss_diag": 1.7642974853515625e-05, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 2.7240591049194336, "learning_rate": 4.679949130987707e-06, "loss": 0.5433, "mean_token_accuracy": 0.8310168981552124, "num_tokens": 41941579.0, "step": 1105 }, { "epoch": 0.14069456812110417, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 1.7762184143066406e-05, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 2.980565071105957, "learning_rate": 4.684188215345486e-06, "loss": 0.4758, "mean_token_accuracy": 0.8527576327323914, "num_tokens": 41975315.0, "step": 1106 }, { "epoch": 0.1408217783996947, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 1.7762184143066406e-05, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 2.735786199569702, "learning_rate": 4.688427299703264e-06, "loss": 0.4638, "mean_token_accuracy": 0.8518671989440918, "num_tokens": 42011950.0, "step": 1107 }, { "epoch": 0.1409489886782852, "ewc_loss": 0.0286865234375, "ewc_loss_diag": 1.7881393432617188e-05, "ewc_loss_parallel": 2.682209014892578e-05, "grad_norm": 2.9720065593719482, "learning_rate": 4.692666384061043e-06, "loss": 0.5284, "mean_token_accuracy": 0.8378759622573853, "num_tokens": 42051764.0, "step": 1108 }, { "epoch": 0.1410761989568757, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 1.7881393432617188e-05, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 2.7394752502441406, "learning_rate": 4.696905468418822e-06, "loss": 0.5416, "mean_token_accuracy": 0.8305314779281616, "num_tokens": 42094218.0, "step": 1109 }, { "epoch": 0.14120340923546623, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 1.7881393432617188e-05, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 2.983684778213501, "learning_rate": 4.701144552776601e-06, "loss": 0.5179, "mean_token_accuracy": 0.8406556844711304, "num_tokens": 42129120.0, "step": 1110 }, { "epoch": 0.14133061951405673, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 1.7881393432617188e-05, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 2.750140905380249, "learning_rate": 4.705383637134379e-06, "loss": 0.5323, "mean_token_accuracy": 0.8361120820045471, "num_tokens": 42167437.0, "step": 1111 }, { "epoch": 0.14145782979264723, "ewc_loss": 0.0286865234375, "ewc_loss_diag": 1.811981201171875e-05, "ewc_loss_parallel": 2.682209014892578e-05, "grad_norm": 2.9775218963623047, "learning_rate": 4.709622721492158e-06, "loss": 0.5071, "mean_token_accuracy": 0.8453824520111084, "num_tokens": 42203143.0, "step": 1112 }, { "epoch": 0.14158504007123776, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 1.811981201171875e-05, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 2.7606308460235596, "learning_rate": 4.713861805849937e-06, "loss": 0.5012, "mean_token_accuracy": 0.8453178405761719, "num_tokens": 42237024.0, "step": 1113 }, { "epoch": 0.14171225034982826, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 1.811981201171875e-05, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 2.979048252105713, "learning_rate": 4.718100890207716e-06, "loss": 0.5799, "mean_token_accuracy": 0.8234186172485352, "num_tokens": 42272847.0, "step": 1114 }, { "epoch": 0.1418394606284188, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 1.811981201171875e-05, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 2.769930124282837, "learning_rate": 4.7223399745654945e-06, "loss": 0.5321, "mean_token_accuracy": 0.8346021175384521, "num_tokens": 42309335.0, "step": 1115 }, { "epoch": 0.1419666709070093, "ewc_loss": 0.0279541015625, "ewc_loss_diag": 1.811981201171875e-05, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 2.9516091346740723, "learning_rate": 4.726579058923273e-06, "loss": 0.5287, "mean_token_accuracy": 0.8408826589584351, "num_tokens": 42346349.0, "step": 1116 }, { "epoch": 0.1420938811855998, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 1.811981201171875e-05, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 2.758315086364746, "learning_rate": 4.730818143281051e-06, "loss": 0.5546, "mean_token_accuracy": 0.8316679000854492, "num_tokens": 42389310.0, "step": 1117 }, { "epoch": 0.14222109146419032, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 1.811981201171875e-05, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 2.955780506134033, "learning_rate": 4.73505722763883e-06, "loss": 0.4959, "mean_token_accuracy": 0.8467288613319397, "num_tokens": 42427808.0, "step": 1118 }, { "epoch": 0.14234830174278082, "ewc_loss": 0.026611328125, "ewc_loss_diag": 1.823902130126953e-05, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 2.7558367252349854, "learning_rate": 4.739296311996609e-06, "loss": 0.5183, "mean_token_accuracy": 0.8382450342178345, "num_tokens": 42465531.0, "step": 1119 }, { "epoch": 0.14247551202137132, "ewc_loss": 0.028076171875, "ewc_loss_diag": 1.823902130126953e-05, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 2.9387905597686768, "learning_rate": 4.743535396354388e-06, "loss": 0.514, "mean_token_accuracy": 0.8445003032684326, "num_tokens": 42507613.0, "step": 1120 }, { "epoch": 0.14260272229996185, "ewc_loss": 0.026611328125, "ewc_loss_diag": 1.823902130126953e-05, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 2.7629637718200684, "learning_rate": 4.747774480712167e-06, "loss": 0.5374, "mean_token_accuracy": 0.8369185924530029, "num_tokens": 42542876.0, "step": 1121 }, { "epoch": 0.14272993257855235, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 1.823902130126953e-05, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 2.9561498165130615, "learning_rate": 4.752013565069945e-06, "loss": 0.5552, "mean_token_accuracy": 0.8303620219230652, "num_tokens": 42577587.0, "step": 1122 }, { "epoch": 0.14285714285714285, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 1.823902130126953e-05, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 2.759897232055664, "learning_rate": 4.756252649427724e-06, "loss": 0.5217, "mean_token_accuracy": 0.839178204536438, "num_tokens": 42612733.0, "step": 1123 }, { "epoch": 0.14298435313573338, "ewc_loss": 0.028564453125, "ewc_loss_diag": 1.823902130126953e-05, "ewc_loss_parallel": 2.6702880859375e-05, "grad_norm": 2.9557836055755615, "learning_rate": 4.760491733785503e-06, "loss": 0.4685, "mean_token_accuracy": 0.8550781011581421, "num_tokens": 42646831.0, "step": 1124 }, { "epoch": 0.14311156341432388, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 1.823902130126953e-05, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 2.7676217555999756, "learning_rate": 4.764730818143282e-06, "loss": 0.517, "mean_token_accuracy": 0.8351452946662903, "num_tokens": 42681351.0, "step": 1125 }, { "epoch": 0.14323877369291438, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 1.823902130126953e-05, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 2.9481866359710693, "learning_rate": 4.76896990250106e-06, "loss": 0.5689, "mean_token_accuracy": 0.8252362608909607, "num_tokens": 42723491.0, "step": 1126 }, { "epoch": 0.1433659839715049, "ewc_loss": 0.02685546875, "ewc_loss_diag": 1.823902130126953e-05, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 2.749027729034424, "learning_rate": 4.7732089868588386e-06, "loss": 0.5261, "mean_token_accuracy": 0.8366762399673462, "num_tokens": 42763349.0, "step": 1127 }, { "epoch": 0.1434931942500954, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 1.823902130126953e-05, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 2.9600963592529297, "learning_rate": 4.7774480712166174e-06, "loss": 0.5404, "mean_token_accuracy": 0.8331924676895142, "num_tokens": 42803977.0, "step": 1128 }, { "epoch": 0.1436204045286859, "ewc_loss": 0.02685546875, "ewc_loss_diag": 1.823902130126953e-05, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 2.748912811279297, "learning_rate": 4.781687155574396e-06, "loss": 0.5425, "mean_token_accuracy": 0.8324710726737976, "num_tokens": 42847313.0, "step": 1129 }, { "epoch": 0.14374761480727644, "ewc_loss": 0.029052734375, "ewc_loss_diag": 1.823902130126953e-05, "ewc_loss_parallel": 2.7179718017578125e-05, "grad_norm": 2.973396062850952, "learning_rate": 4.785926239932175e-06, "loss": 0.491, "mean_token_accuracy": 0.8467947244644165, "num_tokens": 42887468.0, "step": 1130 }, { "epoch": 0.14387482508586694, "ewc_loss": 0.0269775390625, "ewc_loss_diag": 1.823902130126953e-05, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 2.7530181407928467, "learning_rate": 4.790165324289954e-06, "loss": 0.518, "mean_token_accuracy": 0.8416773080825806, "num_tokens": 42925502.0, "step": 1131 }, { "epoch": 0.14400203536445744, "ewc_loss": 0.0289306640625, "ewc_loss_diag": 1.8358230590820312e-05, "ewc_loss_parallel": 2.7060508728027344e-05, "grad_norm": 2.996243476867676, "learning_rate": 4.794404408647733e-06, "loss": 0.5223, "mean_token_accuracy": 0.8419268727302551, "num_tokens": 42961930.0, "step": 1132 }, { "epoch": 0.14412924564304797, "ewc_loss": 0.0269775390625, "ewc_loss_diag": 1.823902130126953e-05, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 2.7350077629089355, "learning_rate": 4.798643493005512e-06, "loss": 0.4651, "mean_token_accuracy": 0.8554985523223877, "num_tokens": 43002121.0, "step": 1133 }, { "epoch": 0.14425645592163847, "ewc_loss": 0.0294189453125, "ewc_loss_diag": 1.823902130126953e-05, "ewc_loss_parallel": 2.753734588623047e-05, "grad_norm": 3.032088279724121, "learning_rate": 4.80288257736329e-06, "loss": 0.5661, "mean_token_accuracy": 0.8287191987037659, "num_tokens": 43039343.0, "step": 1134 }, { "epoch": 0.14438366620022897, "ewc_loss": 0.0269775390625, "ewc_loss_diag": 1.823902130126953e-05, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 2.74225115776062, "learning_rate": 4.807121661721069e-06, "loss": 0.4664, "mean_token_accuracy": 0.8572086691856384, "num_tokens": 43075035.0, "step": 1135 }, { "epoch": 0.1445108764788195, "ewc_loss": 0.029541015625, "ewc_loss_diag": 1.823902130126953e-05, "ewc_loss_parallel": 2.765655517578125e-05, "grad_norm": 3.0486812591552734, "learning_rate": 4.811360746078848e-06, "loss": 0.5229, "mean_token_accuracy": 0.841152012348175, "num_tokens": 43114350.0, "step": 1136 }, { "epoch": 0.14463808675741, "ewc_loss": 0.02734375, "ewc_loss_diag": 1.8477439880371094e-05, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 2.770948648452759, "learning_rate": 4.815599830436626e-06, "loss": 0.5187, "mean_token_accuracy": 0.8396573066711426, "num_tokens": 43146445.0, "step": 1137 }, { "epoch": 0.1447652970360005, "ewc_loss": 0.029541015625, "ewc_loss_diag": 1.8477439880371094e-05, "ewc_loss_parallel": 2.765655517578125e-05, "grad_norm": 3.0547258853912354, "learning_rate": 4.819838914794405e-06, "loss": 0.5053, "mean_token_accuracy": 0.845924437046051, "num_tokens": 43182216.0, "step": 1138 }, { "epoch": 0.14489250731459102, "ewc_loss": 0.0274658203125, "ewc_loss_diag": 1.8596649169921875e-05, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 2.772488832473755, "learning_rate": 4.8240779991521835e-06, "loss": 0.5117, "mean_token_accuracy": 0.8398797512054443, "num_tokens": 43216014.0, "step": 1139 }, { "epoch": 0.14501971759318152, "ewc_loss": 0.02978515625, "ewc_loss_diag": 1.8596649169921875e-05, "ewc_loss_parallel": 2.7894973754882812e-05, "grad_norm": 3.0586953163146973, "learning_rate": 4.828317083509962e-06, "loss": 0.5223, "mean_token_accuracy": 0.8393483757972717, "num_tokens": 43252909.0, "step": 1140 }, { "epoch": 0.14514692787177205, "ewc_loss": 0.027587890625, "ewc_loss_diag": 1.8596649169921875e-05, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 2.7801284790039062, "learning_rate": 4.83255616786774e-06, "loss": 0.4965, "mean_token_accuracy": 0.8499748110771179, "num_tokens": 43285653.0, "step": 1141 }, { "epoch": 0.14527413815036255, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 1.8596649169921875e-05, "ewc_loss_parallel": 2.777576446533203e-05, "grad_norm": 3.0662190914154053, "learning_rate": 4.836795252225519e-06, "loss": 0.4941, "mean_token_accuracy": 0.8508868217468262, "num_tokens": 43322087.0, "step": 1142 }, { "epoch": 0.14540134842895305, "ewc_loss": 0.02783203125, "ewc_loss_diag": 1.8715858459472656e-05, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 2.789006233215332, "learning_rate": 4.841034336583298e-06, "loss": 0.5033, "mean_token_accuracy": 0.8467826247215271, "num_tokens": 43360526.0, "step": 1143 }, { "epoch": 0.14552855870754358, "ewc_loss": 0.0299072265625, "ewc_loss_diag": 1.8715858459472656e-05, "ewc_loss_parallel": 2.8133392333984375e-05, "grad_norm": 3.102005958557129, "learning_rate": 4.845273420941077e-06, "loss": 0.484, "mean_token_accuracy": 0.8528786301612854, "num_tokens": 43400267.0, "step": 1144 }, { "epoch": 0.14565576898613408, "ewc_loss": 0.0279541015625, "ewc_loss_diag": 1.8715858459472656e-05, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 2.817786693572998, "learning_rate": 4.849512505298856e-06, "loss": 0.5231, "mean_token_accuracy": 0.8352031707763672, "num_tokens": 43430411.0, "step": 1145 }, { "epoch": 0.14578297926472458, "ewc_loss": 0.02978515625, "ewc_loss_diag": 1.8715858459472656e-05, "ewc_loss_parallel": 2.8014183044433594e-05, "grad_norm": 3.0615451335906982, "learning_rate": 4.853751589656635e-06, "loss": 0.5095, "mean_token_accuracy": 0.8446265459060669, "num_tokens": 43472199.0, "step": 1146 }, { "epoch": 0.1459101895433151, "ewc_loss": 0.0279541015625, "ewc_loss_diag": 1.8715858459472656e-05, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 2.8005335330963135, "learning_rate": 4.857990674014414e-06, "loss": 0.5312, "mean_token_accuracy": 0.8365612626075745, "num_tokens": 43515896.0, "step": 1147 }, { "epoch": 0.1460373998219056, "ewc_loss": 0.030029296875, "ewc_loss_diag": 1.8835067749023438e-05, "ewc_loss_parallel": 2.8252601623535156e-05, "grad_norm": 3.0662801265716553, "learning_rate": 4.862229758372193e-06, "loss": 0.5556, "mean_token_accuracy": 0.8271652460098267, "num_tokens": 43556088.0, "step": 1148 }, { "epoch": 0.1461646101004961, "ewc_loss": 0.0279541015625, "ewc_loss_diag": 1.8835067749023438e-05, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 2.811542510986328, "learning_rate": 4.866468842729971e-06, "loss": 0.5575, "mean_token_accuracy": 0.8261451721191406, "num_tokens": 43596623.0, "step": 1149 }, { "epoch": 0.14629182037908664, "ewc_loss": 0.0299072265625, "ewc_loss_diag": 1.8835067749023438e-05, "ewc_loss_parallel": 2.8133392333984375e-05, "grad_norm": 3.066983938217163, "learning_rate": 4.8707079270877495e-06, "loss": 0.5327, "mean_token_accuracy": 0.8377389907836914, "num_tokens": 43634757.0, "step": 1150 }, { "epoch": 0.14641903065767714, "ewc_loss": 0.02783203125, "ewc_loss_diag": 1.895427703857422e-05, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 2.8154757022857666, "learning_rate": 4.874947011445528e-06, "loss": 0.5363, "mean_token_accuracy": 0.8373873829841614, "num_tokens": 43671979.0, "step": 1151 }, { "epoch": 0.14654624093626764, "ewc_loss": 0.030029296875, "ewc_loss_diag": 1.895427703857422e-05, "ewc_loss_parallel": 2.8252601623535156e-05, "grad_norm": 3.0529112815856934, "learning_rate": 4.879186095803307e-06, "loss": 0.4619, "mean_token_accuracy": 0.8578832149505615, "num_tokens": 43711898.0, "step": 1152 }, { "epoch": 0.14667345121485817, "ewc_loss": 0.02783203125, "ewc_loss_diag": 1.895427703857422e-05, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 2.822319269180298, "learning_rate": 4.883425180161085e-06, "loss": 0.4914, "mean_token_accuracy": 0.848439633846283, "num_tokens": 43748394.0, "step": 1153 }, { "epoch": 0.14680066149344867, "ewc_loss": 0.02978515625, "ewc_loss_diag": 1.895427703857422e-05, "ewc_loss_parallel": 2.8014183044433594e-05, "grad_norm": 3.0625669956207275, "learning_rate": 4.887664264518864e-06, "loss": 0.5446, "mean_token_accuracy": 0.8398594856262207, "num_tokens": 43784730.0, "step": 1154 }, { "epoch": 0.14692787177203917, "ewc_loss": 0.02783203125, "ewc_loss_diag": 1.895427703857422e-05, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 2.8297641277313232, "learning_rate": 4.891903348876643e-06, "loss": 0.5302, "mean_token_accuracy": 0.840650200843811, "num_tokens": 43823993.0, "step": 1155 }, { "epoch": 0.1470550820506297, "ewc_loss": 0.02978515625, "ewc_loss_diag": 1.9073486328125e-05, "ewc_loss_parallel": 2.777576446533203e-05, "grad_norm": 3.059279680252075, "learning_rate": 4.896142433234421e-06, "loss": 0.568, "mean_token_accuracy": 0.8326644897460938, "num_tokens": 43861673.0, "step": 1156 }, { "epoch": 0.1471822923292202, "ewc_loss": 0.0279541015625, "ewc_loss_diag": 1.9073486328125e-05, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 2.829730272293091, "learning_rate": 4.9003815175922e-06, "loss": 0.5613, "mean_token_accuracy": 0.8327500224113464, "num_tokens": 43900123.0, "step": 1157 }, { "epoch": 0.1473095026078107, "ewc_loss": 0.0299072265625, "ewc_loss_diag": 1.9073486328125e-05, "ewc_loss_parallel": 2.8014183044433594e-05, "grad_norm": 3.0462801456451416, "learning_rate": 4.904620601949979e-06, "loss": 0.518, "mean_token_accuracy": 0.8409314751625061, "num_tokens": 43939371.0, "step": 1158 }, { "epoch": 0.14743671288640123, "ewc_loss": 0.028076171875, "ewc_loss_diag": 1.919269561767578e-05, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 2.835095167160034, "learning_rate": 4.908859686307758e-06, "loss": 0.5527, "mean_token_accuracy": 0.8324838876724243, "num_tokens": 43972960.0, "step": 1159 }, { "epoch": 0.14756392316499173, "ewc_loss": 0.02978515625, "ewc_loss_diag": 1.919269561767578e-05, "ewc_loss_parallel": 2.777576446533203e-05, "grad_norm": 3.045919895172119, "learning_rate": 4.913098770665537e-06, "loss": 0.5561, "mean_token_accuracy": 0.8338151574134827, "num_tokens": 44009919.0, "step": 1160 }, { "epoch": 0.14769113344358223, "ewc_loss": 0.028076171875, "ewc_loss_diag": 1.9311904907226562e-05, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 2.818027973175049, "learning_rate": 4.9173378550233155e-06, "loss": 0.5005, "mean_token_accuracy": 0.8428213000297546, "num_tokens": 44050491.0, "step": 1161 }, { "epoch": 0.14781834372217276, "ewc_loss": 0.0299072265625, "ewc_loss_diag": 1.9431114196777344e-05, "ewc_loss_parallel": 2.8014183044433594e-05, "grad_norm": 3.0454001426696777, "learning_rate": 4.9215769393810944e-06, "loss": 0.5212, "mean_token_accuracy": 0.8447397947311401, "num_tokens": 44087999.0, "step": 1162 }, { "epoch": 0.14794555400076326, "ewc_loss": 0.028076171875, "ewc_loss_diag": 1.9431114196777344e-05, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 2.828436851501465, "learning_rate": 4.925816023738873e-06, "loss": 0.5701, "mean_token_accuracy": 0.8267119526863098, "num_tokens": 44126014.0, "step": 1163 }, { "epoch": 0.14807276427935376, "ewc_loss": 0.0301513671875, "ewc_loss_diag": 1.9431114196777344e-05, "ewc_loss_parallel": 2.8252601623535156e-05, "grad_norm": 3.055788516998291, "learning_rate": 4.930055108096651e-06, "loss": 0.4607, "mean_token_accuracy": 0.8583210706710815, "num_tokens": 44154713.0, "step": 1164 }, { "epoch": 0.1481999745579443, "ewc_loss": 0.028076171875, "ewc_loss_diag": 1.9431114196777344e-05, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 2.839757204055786, "learning_rate": 4.93429419245443e-06, "loss": 0.5844, "mean_token_accuracy": 0.8222717046737671, "num_tokens": 44189624.0, "step": 1165 }, { "epoch": 0.1483271848365348, "ewc_loss": 0.0301513671875, "ewc_loss_diag": 1.9550323486328125e-05, "ewc_loss_parallel": 2.8252601623535156e-05, "grad_norm": 3.0509326457977295, "learning_rate": 4.938533276812209e-06, "loss": 0.5269, "mean_token_accuracy": 0.8402224779129028, "num_tokens": 44225010.0, "step": 1166 }, { "epoch": 0.14845439511512531, "ewc_loss": 0.028076171875, "ewc_loss_diag": 1.9550323486328125e-05, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 2.8102946281433105, "learning_rate": 4.942772361169988e-06, "loss": 0.4391, "mean_token_accuracy": 0.8591216802597046, "num_tokens": 44264987.0, "step": 1167 }, { "epoch": 0.14858160539371582, "ewc_loss": 0.030029296875, "ewc_loss_diag": 1.9550323486328125e-05, "ewc_loss_parallel": 2.8133392333984375e-05, "grad_norm": 3.0596375465393066, "learning_rate": 4.947011445527766e-06, "loss": 0.5183, "mean_token_accuracy": 0.8432389497756958, "num_tokens": 44306705.0, "step": 1168 }, { "epoch": 0.14870881567230632, "ewc_loss": 0.028076171875, "ewc_loss_diag": 1.9550323486328125e-05, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 2.807110071182251, "learning_rate": 4.951250529885545e-06, "loss": 0.4705, "mean_token_accuracy": 0.8558766841888428, "num_tokens": 44347227.0, "step": 1169 }, { "epoch": 0.14883602595089684, "ewc_loss": 0.0303955078125, "ewc_loss_diag": 1.9550323486328125e-05, "ewc_loss_parallel": 2.849102020263672e-05, "grad_norm": 3.1040356159210205, "learning_rate": 4.955489614243324e-06, "loss": 0.6075, "mean_token_accuracy": 0.8114762306213379, "num_tokens": 44385273.0, "step": 1170 }, { "epoch": 0.14896323622948734, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 1.9669532775878906e-05, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 2.8145740032196045, "learning_rate": 4.959728698601103e-06, "loss": 0.5248, "mean_token_accuracy": 0.8352215886116028, "num_tokens": 44426177.0, "step": 1171 }, { "epoch": 0.14909044650807785, "ewc_loss": 0.03076171875, "ewc_loss_diag": 1.9669532775878906e-05, "ewc_loss_parallel": 2.8848648071289062e-05, "grad_norm": 3.112344741821289, "learning_rate": 4.963967782958881e-06, "loss": 0.5025, "mean_token_accuracy": 0.8469475507736206, "num_tokens": 44468609.0, "step": 1172 }, { "epoch": 0.14921765678666837, "ewc_loss": 0.0283203125, "ewc_loss_diag": 1.9669532775878906e-05, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 2.8196022510528564, "learning_rate": 4.96820686731666e-06, "loss": 0.5017, "mean_token_accuracy": 0.8431364297866821, "num_tokens": 44507212.0, "step": 1173 }, { "epoch": 0.14934486706525887, "ewc_loss": 0.03076171875, "ewc_loss_diag": 1.9669532775878906e-05, "ewc_loss_parallel": 2.8848648071289062e-05, "grad_norm": 3.1491503715515137, "learning_rate": 4.9724459516744385e-06, "loss": 0.4868, "mean_token_accuracy": 0.8488922119140625, "num_tokens": 44538551.0, "step": 1174 }, { "epoch": 0.14947207734384937, "ewc_loss": 0.028564453125, "ewc_loss_diag": 1.9669532775878906e-05, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 2.8330671787261963, "learning_rate": 4.976685036032217e-06, "loss": 0.5773, "mean_token_accuracy": 0.822969377040863, "num_tokens": 44577105.0, "step": 1175 }, { "epoch": 0.1495992876224399, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 1.9669532775878906e-05, "ewc_loss_parallel": 2.8967857360839844e-05, "grad_norm": 3.1437501907348633, "learning_rate": 4.980924120389996e-06, "loss": 0.4568, "mean_token_accuracy": 0.859827995300293, "num_tokens": 44615548.0, "step": 1176 }, { "epoch": 0.1497264979010304, "ewc_loss": 0.02880859375, "ewc_loss_diag": 1.9669532775878906e-05, "ewc_loss_parallel": 2.682209014892578e-05, "grad_norm": 2.8311915397644043, "learning_rate": 4.985163204747775e-06, "loss": 0.462, "mean_token_accuracy": 0.8578261137008667, "num_tokens": 44654447.0, "step": 1177 }, { "epoch": 0.1498537081796209, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 1.9669532775878906e-05, "ewc_loss_parallel": 2.8967857360839844e-05, "grad_norm": 3.1529603004455566, "learning_rate": 4.989402289105554e-06, "loss": 0.5178, "mean_token_accuracy": 0.844331681728363, "num_tokens": 44697995.0, "step": 1178 }, { "epoch": 0.14998091845821143, "ewc_loss": 0.0289306640625, "ewc_loss_diag": 1.9669532775878906e-05, "ewc_loss_parallel": 2.6941299438476562e-05, "grad_norm": 2.865586042404175, "learning_rate": 4.993641373463332e-06, "loss": 0.5773, "mean_token_accuracy": 0.8226193785667419, "num_tokens": 44734376.0, "step": 1179 }, { "epoch": 0.15010812873680193, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 1.9669532775878906e-05, "ewc_loss_parallel": 2.8967857360839844e-05, "grad_norm": 3.210181474685669, "learning_rate": 4.997880457821111e-06, "loss": 0.5278, "mean_token_accuracy": 0.8367175459861755, "num_tokens": 44772745.0, "step": 1180 }, { "epoch": 0.15023533901539243, "ewc_loss": 0.0291748046875, "ewc_loss_diag": 1.9669532775878906e-05, "ewc_loss_parallel": 2.7179718017578125e-05, "grad_norm": 2.8509392738342285, "learning_rate": 5.002119542178889e-06, "loss": 0.5145, "mean_token_accuracy": 0.8432968258857727, "num_tokens": 44816530.0, "step": 1181 }, { "epoch": 0.15036254929398296, "ewc_loss": 0.03125, "ewc_loss_diag": 1.9788742065429688e-05, "ewc_loss_parallel": 2.9325485229492188e-05, "grad_norm": 3.1909403800964355, "learning_rate": 5.006358626536668e-06, "loss": 0.4761, "mean_token_accuracy": 0.8528306484222412, "num_tokens": 44853428.0, "step": 1182 }, { "epoch": 0.15048975957257346, "ewc_loss": 0.029296875, "ewc_loss_diag": 1.9788742065429688e-05, "ewc_loss_parallel": 2.7298927307128906e-05, "grad_norm": 2.8687222003936768, "learning_rate": 5.010597710894447e-06, "loss": 0.4957, "mean_token_accuracy": 0.8468990325927734, "num_tokens": 44891118.0, "step": 1183 }, { "epoch": 0.15061696985116396, "ewc_loss": 0.03125, "ewc_loss_diag": 1.9788742065429688e-05, "ewc_loss_parallel": 2.9325485229492188e-05, "grad_norm": 3.253838062286377, "learning_rate": 5.014836795252226e-06, "loss": 0.4871, "mean_token_accuracy": 0.8500872254371643, "num_tokens": 44929143.0, "step": 1184 }, { "epoch": 0.1507441801297545, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 1.9788742065429688e-05, "ewc_loss_parallel": 2.765655517578125e-05, "grad_norm": 2.890876293182373, "learning_rate": 5.0190758796100045e-06, "loss": 0.4928, "mean_token_accuracy": 0.8508577942848206, "num_tokens": 44966361.0, "step": 1185 }, { "epoch": 0.150871390408345, "ewc_loss": 0.031494140625, "ewc_loss_diag": 2.002716064453125e-05, "ewc_loss_parallel": 2.944469451904297e-05, "grad_norm": 3.1846461296081543, "learning_rate": 5.0233149639677834e-06, "loss": 0.5294, "mean_token_accuracy": 0.8365470767021179, "num_tokens": 45007993.0, "step": 1186 }, { "epoch": 0.1509986006869355, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 2.002716064453125e-05, "ewc_loss_parallel": 2.765655517578125e-05, "grad_norm": 2.9194247722625732, "learning_rate": 5.027554048325562e-06, "loss": 0.5662, "mean_token_accuracy": 0.8263643383979797, "num_tokens": 45040337.0, "step": 1187 }, { "epoch": 0.15112581096552602, "ewc_loss": 0.03125, "ewc_loss_diag": 2.002716064453125e-05, "ewc_loss_parallel": 2.9325485229492188e-05, "grad_norm": 3.1566736698150635, "learning_rate": 5.031793132683341e-06, "loss": 0.5765, "mean_token_accuracy": 0.8236880302429199, "num_tokens": 45080730.0, "step": 1188 }, { "epoch": 0.15125302124411652, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 2.002716064453125e-05, "ewc_loss_parallel": 2.765655517578125e-05, "grad_norm": 2.9152793884277344, "learning_rate": 5.036032217041119e-06, "loss": 0.5337, "mean_token_accuracy": 0.8340906500816345, "num_tokens": 45116748.0, "step": 1189 }, { "epoch": 0.15138023152270705, "ewc_loss": 0.031005859375, "ewc_loss_diag": 2.002716064453125e-05, "ewc_loss_parallel": 2.9087066650390625e-05, "grad_norm": 3.119302272796631, "learning_rate": 5.040271301398898e-06, "loss": 0.5254, "mean_token_accuracy": 0.83528071641922, "num_tokens": 45158596.0, "step": 1190 }, { "epoch": 0.15150744180129755, "ewc_loss": 0.029541015625, "ewc_loss_diag": 2.002716064453125e-05, "ewc_loss_parallel": 2.753734588623047e-05, "grad_norm": 2.907804012298584, "learning_rate": 5.044510385756677e-06, "loss": 0.4985, "mean_token_accuracy": 0.8443827629089355, "num_tokens": 45202433.0, "step": 1191 }, { "epoch": 0.15163465207988805, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 2.002716064453125e-05, "ewc_loss_parallel": 2.8967857360839844e-05, "grad_norm": 3.1044843196868896, "learning_rate": 5.048749470114456e-06, "loss": 0.5087, "mean_token_accuracy": 0.8452306985855103, "num_tokens": 45239202.0, "step": 1192 }, { "epoch": 0.15176186235847858, "ewc_loss": 0.0291748046875, "ewc_loss_diag": 2.002716064453125e-05, "ewc_loss_parallel": 2.7179718017578125e-05, "grad_norm": 2.9361910820007324, "learning_rate": 5.052988554472235e-06, "loss": 0.5241, "mean_token_accuracy": 0.8379624485969543, "num_tokens": 45270150.0, "step": 1193 }, { "epoch": 0.15188907263706908, "ewc_loss": 0.0306396484375, "ewc_loss_diag": 2.002716064453125e-05, "ewc_loss_parallel": 2.872943878173828e-05, "grad_norm": 3.073286771774292, "learning_rate": 5.057227638830014e-06, "loss": 0.5231, "mean_token_accuracy": 0.841657280921936, "num_tokens": 45306394.0, "step": 1194 }, { "epoch": 0.15201628291565958, "ewc_loss": 0.029296875, "ewc_loss_diag": 2.014636993408203e-05, "ewc_loss_parallel": 2.7179718017578125e-05, "grad_norm": 2.9326281547546387, "learning_rate": 5.0614667231877925e-06, "loss": 0.5392, "mean_token_accuracy": 0.8354254961013794, "num_tokens": 45344945.0, "step": 1195 }, { "epoch": 0.1521434931942501, "ewc_loss": 0.030517578125, "ewc_loss_diag": 2.014636993408203e-05, "ewc_loss_parallel": 2.86102294921875e-05, "grad_norm": 3.0568456649780273, "learning_rate": 5.06570580754557e-06, "loss": 0.4933, "mean_token_accuracy": 0.8507051467895508, "num_tokens": 45378179.0, "step": 1196 }, { "epoch": 0.1522707034728406, "ewc_loss": 0.029296875, "ewc_loss_diag": 2.014636993408203e-05, "ewc_loss_parallel": 2.7179718017578125e-05, "grad_norm": 2.8973052501678467, "learning_rate": 5.069944891903349e-06, "loss": 0.5085, "mean_token_accuracy": 0.8416630029678345, "num_tokens": 45421194.0, "step": 1197 }, { "epoch": 0.1523979137514311, "ewc_loss": 0.03076171875, "ewc_loss_diag": 2.014636993408203e-05, "ewc_loss_parallel": 2.872943878173828e-05, "grad_norm": 3.058896064758301, "learning_rate": 5.0741839762611275e-06, "loss": 0.5194, "mean_token_accuracy": 0.8394073247909546, "num_tokens": 45458226.0, "step": 1198 }, { "epoch": 0.15252512403002164, "ewc_loss": 0.029296875, "ewc_loss_diag": 2.0265579223632812e-05, "ewc_loss_parallel": 2.7179718017578125e-05, "grad_norm": 2.894425868988037, "learning_rate": 5.078423060618906e-06, "loss": 0.5191, "mean_token_accuracy": 0.8415660858154297, "num_tokens": 45500007.0, "step": 1199 }, { "epoch": 0.15265233430861214, "ewc_loss": 0.030517578125, "ewc_loss_diag": 2.0265579223632812e-05, "ewc_loss_parallel": 2.849102020263672e-05, "grad_norm": 3.0712084770202637, "learning_rate": 5.082662144976685e-06, "loss": 0.5135, "mean_token_accuracy": 0.8383692502975464, "num_tokens": 45536605.0, "step": 1200 }, { "epoch": 0.15277954458720264, "ewc_loss": 0.029296875, "ewc_loss_diag": 2.0265579223632812e-05, "ewc_loss_parallel": 2.7179718017578125e-05, "grad_norm": 2.8901641368865967, "learning_rate": 5.086901229334464e-06, "loss": 0.5076, "mean_token_accuracy": 0.847109854221344, "num_tokens": 45572137.0, "step": 1201 }, { "epoch": 0.15290675486579317, "ewc_loss": 0.0306396484375, "ewc_loss_diag": 2.0265579223632812e-05, "ewc_loss_parallel": 2.86102294921875e-05, "grad_norm": 3.0846059322357178, "learning_rate": 5.091140313692243e-06, "loss": 0.5151, "mean_token_accuracy": 0.8413741588592529, "num_tokens": 45608659.0, "step": 1202 }, { "epoch": 0.15303396514438367, "ewc_loss": 0.0291748046875, "ewc_loss_diag": 2.0265579223632812e-05, "ewc_loss_parallel": 2.7060508728027344e-05, "grad_norm": 2.873575210571289, "learning_rate": 5.095379398050022e-06, "loss": 0.4991, "mean_token_accuracy": 0.844603419303894, "num_tokens": 45647002.0, "step": 1203 }, { "epoch": 0.15316117542297417, "ewc_loss": 0.0306396484375, "ewc_loss_diag": 2.0265579223632812e-05, "ewc_loss_parallel": 2.86102294921875e-05, "grad_norm": 3.121006488800049, "learning_rate": 5.0996184824078e-06, "loss": 0.528, "mean_token_accuracy": 0.835820734500885, "num_tokens": 45680194.0, "step": 1204 }, { "epoch": 0.1532883857015647, "ewc_loss": 0.0291748046875, "ewc_loss_diag": 2.0384788513183594e-05, "ewc_loss_parallel": 2.7060508728027344e-05, "grad_norm": 2.8647501468658447, "learning_rate": 5.103857566765579e-06, "loss": 0.499, "mean_token_accuracy": 0.8458758592605591, "num_tokens": 45719629.0, "step": 1205 }, { "epoch": 0.1534155959801552, "ewc_loss": 0.03125, "ewc_loss_diag": 2.0384788513183594e-05, "ewc_loss_parallel": 2.9206275939941406e-05, "grad_norm": 3.140732526779175, "learning_rate": 5.108096651123358e-06, "loss": 0.4694, "mean_token_accuracy": 0.8559682369232178, "num_tokens": 45754630.0, "step": 1206 }, { "epoch": 0.1535428062587457, "ewc_loss": 0.029052734375, "ewc_loss_diag": 2.0384788513183594e-05, "ewc_loss_parallel": 2.6941299438476562e-05, "grad_norm": 2.8481688499450684, "learning_rate": 5.112335735481137e-06, "loss": 0.5046, "mean_token_accuracy": 0.8424593210220337, "num_tokens": 45791904.0, "step": 1207 }, { "epoch": 0.15367001653733622, "ewc_loss": 0.031494140625, "ewc_loss_diag": 2.0384788513183594e-05, "ewc_loss_parallel": 2.944469451904297e-05, "grad_norm": 3.1760175228118896, "learning_rate": 5.1165748198389155e-06, "loss": 0.4247, "mean_token_accuracy": 0.87168288230896, "num_tokens": 45830553.0, "step": 1208 }, { "epoch": 0.15379722681592672, "ewc_loss": 0.029052734375, "ewc_loss_diag": 2.0503997802734375e-05, "ewc_loss_parallel": 2.6941299438476562e-05, "grad_norm": 2.853548765182495, "learning_rate": 5.120813904196694e-06, "loss": 0.4973, "mean_token_accuracy": 0.8481740951538086, "num_tokens": 45865957.0, "step": 1209 }, { "epoch": 0.15392443709451722, "ewc_loss": 0.03173828125, "ewc_loss_diag": 2.0503997802734375e-05, "ewc_loss_parallel": 2.9802322387695312e-05, "grad_norm": 3.2389819622039795, "learning_rate": 5.125052988554473e-06, "loss": 0.5609, "mean_token_accuracy": 0.8303415775299072, "num_tokens": 45898993.0, "step": 1210 }, { "epoch": 0.15405164737310775, "ewc_loss": 0.029541015625, "ewc_loss_diag": 2.0503997802734375e-05, "ewc_loss_parallel": 2.7418136596679688e-05, "grad_norm": 2.8607983589172363, "learning_rate": 5.1292920729122505e-06, "loss": 0.4891, "mean_token_accuracy": 0.8499665260314941, "num_tokens": 45932843.0, "step": 1211 }, { "epoch": 0.15417885765169825, "ewc_loss": 0.0322265625, "ewc_loss_diag": 2.0623207092285156e-05, "ewc_loss_parallel": 3.0279159545898438e-05, "grad_norm": 3.2590951919555664, "learning_rate": 5.133531157270029e-06, "loss": 0.4903, "mean_token_accuracy": 0.8521484136581421, "num_tokens": 45973247.0, "step": 1212 }, { "epoch": 0.15430606793028875, "ewc_loss": 0.02978515625, "ewc_loss_diag": 2.0623207092285156e-05, "ewc_loss_parallel": 2.765655517578125e-05, "grad_norm": 2.8832967281341553, "learning_rate": 5.137770241627808e-06, "loss": 0.448, "mean_token_accuracy": 0.860892653465271, "num_tokens": 46008641.0, "step": 1213 }, { "epoch": 0.15443327820887928, "ewc_loss": 0.032470703125, "ewc_loss_diag": 2.0623207092285156e-05, "ewc_loss_parallel": 3.0517578125e-05, "grad_norm": 3.2849206924438477, "learning_rate": 5.142009325985587e-06, "loss": 0.4834, "mean_token_accuracy": 0.8560659885406494, "num_tokens": 46046789.0, "step": 1214 }, { "epoch": 0.15456048848746978, "ewc_loss": 0.030029296875, "ewc_loss_diag": 2.0623207092285156e-05, "ewc_loss_parallel": 2.7894973754882812e-05, "grad_norm": 2.8931119441986084, "learning_rate": 5.146248410343366e-06, "loss": 0.5191, "mean_token_accuracy": 0.8429161310195923, "num_tokens": 46086904.0, "step": 1215 }, { "epoch": 0.1546876987660603, "ewc_loss": 0.033203125, "ewc_loss_diag": 2.0623207092285156e-05, "ewc_loss_parallel": 3.123283386230469e-05, "grad_norm": 3.2919280529022217, "learning_rate": 5.150487494701145e-06, "loss": 0.4724, "mean_token_accuracy": 0.8577964305877686, "num_tokens": 46129743.0, "step": 1216 }, { "epoch": 0.1548149090446508, "ewc_loss": 0.0303955078125, "ewc_loss_diag": 2.0623207092285156e-05, "ewc_loss_parallel": 2.8371810913085938e-05, "grad_norm": 2.9325759410858154, "learning_rate": 5.154726579058924e-06, "loss": 0.5139, "mean_token_accuracy": 0.8421825170516968, "num_tokens": 46164902.0, "step": 1217 }, { "epoch": 0.1549421193232413, "ewc_loss": 0.03271484375, "ewc_loss_diag": 2.0623207092285156e-05, "ewc_loss_parallel": 3.075599670410156e-05, "grad_norm": 3.283663749694824, "learning_rate": 5.158965663416703e-06, "loss": 0.5131, "mean_token_accuracy": 0.8450660705566406, "num_tokens": 46209689.0, "step": 1218 }, { "epoch": 0.15506932960183184, "ewc_loss": 0.030517578125, "ewc_loss_diag": 2.0623207092285156e-05, "ewc_loss_parallel": 2.849102020263672e-05, "grad_norm": 2.9382147789001465, "learning_rate": 5.163204747774481e-06, "loss": 0.4603, "mean_token_accuracy": 0.857203483581543, "num_tokens": 46251365.0, "step": 1219 }, { "epoch": 0.15519653988042234, "ewc_loss": 0.03271484375, "ewc_loss_diag": 2.0623207092285156e-05, "ewc_loss_parallel": 3.075599670410156e-05, "grad_norm": 3.282221555709839, "learning_rate": 5.1674438321322596e-06, "loss": 0.5159, "mean_token_accuracy": 0.8407183289527893, "num_tokens": 46290961.0, "step": 1220 }, { "epoch": 0.15532375015901284, "ewc_loss": 0.030517578125, "ewc_loss_diag": 2.0623207092285156e-05, "ewc_loss_parallel": 2.849102020263672e-05, "grad_norm": 2.9763734340667725, "learning_rate": 5.1716829164900385e-06, "loss": 0.4905, "mean_token_accuracy": 0.8504356145858765, "num_tokens": 46326317.0, "step": 1221 }, { "epoch": 0.15545096043760337, "ewc_loss": 0.032470703125, "ewc_loss_diag": 2.0623207092285156e-05, "ewc_loss_parallel": 3.0517578125e-05, "grad_norm": 3.249711275100708, "learning_rate": 5.175922000847817e-06, "loss": 0.4627, "mean_token_accuracy": 0.857675313949585, "num_tokens": 46361018.0, "step": 1222 }, { "epoch": 0.15557817071619387, "ewc_loss": 0.0303955078125, "ewc_loss_diag": 2.086162567138672e-05, "ewc_loss_parallel": 2.8371810913085938e-05, "grad_norm": 2.9822590351104736, "learning_rate": 5.180161085205596e-06, "loss": 0.49, "mean_token_accuracy": 0.8472021818161011, "num_tokens": 46397344.0, "step": 1223 }, { "epoch": 0.15570538099478437, "ewc_loss": 0.031982421875, "ewc_loss_diag": 2.086162567138672e-05, "ewc_loss_parallel": 2.9921531677246094e-05, "grad_norm": 3.219306707382202, "learning_rate": 5.184400169563375e-06, "loss": 0.5534, "mean_token_accuracy": 0.8311759233474731, "num_tokens": 46437257.0, "step": 1224 }, { "epoch": 0.1558325912733749, "ewc_loss": 0.0303955078125, "ewc_loss_diag": 2.09808349609375e-05, "ewc_loss_parallel": 2.8371810913085938e-05, "grad_norm": 2.978060483932495, "learning_rate": 5.188639253921154e-06, "loss": 0.5263, "mean_token_accuracy": 0.8394798040390015, "num_tokens": 46477844.0, "step": 1225 }, { "epoch": 0.1559598015519654, "ewc_loss": 0.03173828125, "ewc_loss_diag": 2.09808349609375e-05, "ewc_loss_parallel": 2.956390380859375e-05, "grad_norm": 3.1954829692840576, "learning_rate": 5.192878338278933e-06, "loss": 0.4923, "mean_token_accuracy": 0.8487975597381592, "num_tokens": 46514556.0, "step": 1226 }, { "epoch": 0.1560870118305559, "ewc_loss": 0.0302734375, "ewc_loss_diag": 2.09808349609375e-05, "ewc_loss_parallel": 2.8252601623535156e-05, "grad_norm": 2.9709184169769287, "learning_rate": 5.19711742263671e-06, "loss": 0.5644, "mean_token_accuracy": 0.8283135890960693, "num_tokens": 46559274.0, "step": 1227 }, { "epoch": 0.15621422210914643, "ewc_loss": 0.031494140625, "ewc_loss_diag": 2.09808349609375e-05, "ewc_loss_parallel": 2.944469451904297e-05, "grad_norm": 3.1902928352355957, "learning_rate": 5.201356506994489e-06, "loss": 0.4946, "mean_token_accuracy": 0.8443516492843628, "num_tokens": 46591104.0, "step": 1228 }, { "epoch": 0.15634143238773693, "ewc_loss": 0.030029296875, "ewc_loss_diag": 2.110004425048828e-05, "ewc_loss_parallel": 2.8014183044433594e-05, "grad_norm": 2.9739577770233154, "learning_rate": 5.205595591352268e-06, "loss": 0.5842, "mean_token_accuracy": 0.8204394578933716, "num_tokens": 46631296.0, "step": 1229 }, { "epoch": 0.15646864266632743, "ewc_loss": 0.031494140625, "ewc_loss_diag": 2.110004425048828e-05, "ewc_loss_parallel": 2.944469451904297e-05, "grad_norm": 3.172238826751709, "learning_rate": 5.209834675710047e-06, "loss": 0.5218, "mean_token_accuracy": 0.8424595594406128, "num_tokens": 46673079.0, "step": 1230 }, { "epoch": 0.15659585294491796, "ewc_loss": 0.030029296875, "ewc_loss_diag": 2.110004425048828e-05, "ewc_loss_parallel": 2.8014183044433594e-05, "grad_norm": 2.974701404571533, "learning_rate": 5.214073760067826e-06, "loss": 0.584, "mean_token_accuracy": 0.8224382996559143, "num_tokens": 46706820.0, "step": 1231 }, { "epoch": 0.15672306322350846, "ewc_loss": 0.031494140625, "ewc_loss_diag": 2.1219253540039062e-05, "ewc_loss_parallel": 2.9325485229492188e-05, "grad_norm": 3.1595213413238525, "learning_rate": 5.2183128444256045e-06, "loss": 0.5125, "mean_token_accuracy": 0.8455790281295776, "num_tokens": 46742466.0, "step": 1232 }, { "epoch": 0.15685027350209896, "ewc_loss": 0.030029296875, "ewc_loss_diag": 2.1219253540039062e-05, "ewc_loss_parallel": 2.7894973754882812e-05, "grad_norm": 2.9654061794281006, "learning_rate": 5.222551928783383e-06, "loss": 0.4889, "mean_token_accuracy": 0.8476170301437378, "num_tokens": 46778882.0, "step": 1233 }, { "epoch": 0.1569774837806895, "ewc_loss": 0.03125, "ewc_loss_diag": 2.1219253540039062e-05, "ewc_loss_parallel": 2.9206275939941406e-05, "grad_norm": 3.1469223499298096, "learning_rate": 5.226791013141161e-06, "loss": 0.5498, "mean_token_accuracy": 0.8351870179176331, "num_tokens": 46821509.0, "step": 1234 }, { "epoch": 0.15710469405928, "ewc_loss": 0.0299072265625, "ewc_loss_diag": 2.1219253540039062e-05, "ewc_loss_parallel": 2.777576446533203e-05, "grad_norm": 2.94608211517334, "learning_rate": 5.23103009749894e-06, "loss": 0.496, "mean_token_accuracy": 0.8468675017356873, "num_tokens": 46862247.0, "step": 1235 }, { "epoch": 0.1572319043378705, "ewc_loss": 0.031494140625, "ewc_loss_diag": 2.1219253540039062e-05, "ewc_loss_parallel": 2.944469451904297e-05, "grad_norm": 3.1535568237304688, "learning_rate": 5.235269181856719e-06, "loss": 0.52, "mean_token_accuracy": 0.8419179320335388, "num_tokens": 46906645.0, "step": 1236 }, { "epoch": 0.15735911461646102, "ewc_loss": 0.02978515625, "ewc_loss_diag": 2.1219253540039062e-05, "ewc_loss_parallel": 2.765655517578125e-05, "grad_norm": 2.9581758975982666, "learning_rate": 5.239508266214498e-06, "loss": 0.5651, "mean_token_accuracy": 0.8236449956893921, "num_tokens": 46940492.0, "step": 1237 }, { "epoch": 0.15748632489505152, "ewc_loss": 0.031494140625, "ewc_loss_diag": 2.1219253540039062e-05, "ewc_loss_parallel": 2.944469451904297e-05, "grad_norm": 3.1733834743499756, "learning_rate": 5.243747350572277e-06, "loss": 0.5262, "mean_token_accuracy": 0.8417569398880005, "num_tokens": 46976871.0, "step": 1238 }, { "epoch": 0.15761353517364202, "ewc_loss": 0.030029296875, "ewc_loss_diag": 2.1219253540039062e-05, "ewc_loss_parallel": 2.8014183044433594e-05, "grad_norm": 2.942922353744507, "learning_rate": 5.247986434930056e-06, "loss": 0.5167, "mean_token_accuracy": 0.8442285060882568, "num_tokens": 47013689.0, "step": 1239 }, { "epoch": 0.15774074545223254, "ewc_loss": 0.031982421875, "ewc_loss_diag": 2.1219253540039062e-05, "ewc_loss_parallel": 2.9802322387695312e-05, "grad_norm": 3.2017388343811035, "learning_rate": 5.252225519287835e-06, "loss": 0.5794, "mean_token_accuracy": 0.823179304599762, "num_tokens": 47050995.0, "step": 1240 }, { "epoch": 0.15786795573082305, "ewc_loss": 0.0299072265625, "ewc_loss_diag": 2.1219253540039062e-05, "ewc_loss_parallel": 2.777576446533203e-05, "grad_norm": 2.9202284812927246, "learning_rate": 5.256464603645614e-06, "loss": 0.5038, "mean_token_accuracy": 0.8487184643745422, "num_tokens": 47088571.0, "step": 1241 }, { "epoch": 0.15799516600941357, "ewc_loss": 0.0322265625, "ewc_loss_diag": 2.1219253540039062e-05, "ewc_loss_parallel": 3.0159950256347656e-05, "grad_norm": 3.2177693843841553, "learning_rate": 5.260703688003392e-06, "loss": 0.5187, "mean_token_accuracy": 0.8419869542121887, "num_tokens": 47130547.0, "step": 1242 }, { "epoch": 0.15812237628800407, "ewc_loss": 0.030029296875, "ewc_loss_diag": 2.1338462829589844e-05, "ewc_loss_parallel": 2.7894973754882812e-05, "grad_norm": 2.923003911972046, "learning_rate": 5.2649427723611705e-06, "loss": 0.5173, "mean_token_accuracy": 0.8397489786148071, "num_tokens": 47162724.0, "step": 1243 }, { "epoch": 0.15824958656659457, "ewc_loss": 0.032958984375, "ewc_loss_diag": 2.1338462829589844e-05, "ewc_loss_parallel": 3.075599670410156e-05, "grad_norm": 3.2717416286468506, "learning_rate": 5.269181856718949e-06, "loss": 0.5601, "mean_token_accuracy": 0.8286788463592529, "num_tokens": 47200081.0, "step": 1244 }, { "epoch": 0.1583767968451851, "ewc_loss": 0.030029296875, "ewc_loss_diag": 2.1338462829589844e-05, "ewc_loss_parallel": 2.7894973754882812e-05, "grad_norm": 2.9222421646118164, "learning_rate": 5.2734209410767275e-06, "loss": 0.4966, "mean_token_accuracy": 0.850785493850708, "num_tokens": 47235074.0, "step": 1245 }, { "epoch": 0.1585040071237756, "ewc_loss": 0.033447265625, "ewc_loss_diag": 2.1338462829589844e-05, "ewc_loss_parallel": 3.123283386230469e-05, "grad_norm": 3.2886126041412354, "learning_rate": 5.277660025434506e-06, "loss": 0.4582, "mean_token_accuracy": 0.8579898476600647, "num_tokens": 47274258.0, "step": 1246 }, { "epoch": 0.1586312174023661, "ewc_loss": 0.030517578125, "ewc_loss_diag": 2.1338462829589844e-05, "ewc_loss_parallel": 2.8371810913085938e-05, "grad_norm": 2.910419225692749, "learning_rate": 5.281899109792285e-06, "loss": 0.4824, "mean_token_accuracy": 0.8512755632400513, "num_tokens": 47314232.0, "step": 1247 }, { "epoch": 0.15875842768095663, "ewc_loss": 0.033935546875, "ewc_loss_diag": 2.1338462829589844e-05, "ewc_loss_parallel": 3.170967102050781e-05, "grad_norm": 3.3401737213134766, "learning_rate": 5.286138194150064e-06, "loss": 0.5317, "mean_token_accuracy": 0.8401685357093811, "num_tokens": 47355245.0, "step": 1248 }, { "epoch": 0.15888563795954713, "ewc_loss": 0.03076171875, "ewc_loss_diag": 2.1338462829589844e-05, "ewc_loss_parallel": 2.86102294921875e-05, "grad_norm": 2.9390265941619873, "learning_rate": 5.290377278507842e-06, "loss": 0.5604, "mean_token_accuracy": 0.833714485168457, "num_tokens": 47394356.0, "step": 1249 }, { "epoch": 0.15901284823813763, "ewc_loss": 0.034423828125, "ewc_loss_diag": 2.1338462829589844e-05, "ewc_loss_parallel": 3.218650817871094e-05, "grad_norm": 3.367037057876587, "learning_rate": 5.294616362865621e-06, "loss": 0.5035, "mean_token_accuracy": 0.8474224805831909, "num_tokens": 47435220.0, "step": 1250 }, { "epoch": 0.15914005851672816, "ewc_loss": 0.03125, "ewc_loss_diag": 2.1338462829589844e-05, "ewc_loss_parallel": 2.9087066650390625e-05, "grad_norm": 2.970268964767456, "learning_rate": 5.2988554472234e-06, "loss": 0.5768, "mean_token_accuracy": 0.8321524858474731, "num_tokens": 47477294.0, "step": 1251 }, { "epoch": 0.15926726879531866, "ewc_loss": 0.0341796875, "ewc_loss_diag": 2.1457672119140625e-05, "ewc_loss_parallel": 3.1948089599609375e-05, "grad_norm": 3.391345262527466, "learning_rate": 5.303094531581179e-06, "loss": 0.5365, "mean_token_accuracy": 0.8374476432800293, "num_tokens": 47512216.0, "step": 1252 }, { "epoch": 0.15939447907390916, "ewc_loss": 0.03173828125, "ewc_loss_diag": 2.1457672119140625e-05, "ewc_loss_parallel": 2.956390380859375e-05, "grad_norm": 2.9847147464752197, "learning_rate": 5.307333615938958e-06, "loss": 0.4536, "mean_token_accuracy": 0.8626784682273865, "num_tokens": 47550075.0, "step": 1253 }, { "epoch": 0.1595216893524997, "ewc_loss": 0.034423828125, "ewc_loss_diag": 2.1457672119140625e-05, "ewc_loss_parallel": 3.218650817871094e-05, "grad_norm": 3.403059720993042, "learning_rate": 5.3115727002967366e-06, "loss": 0.5418, "mean_token_accuracy": 0.8351602554321289, "num_tokens": 47581595.0, "step": 1254 }, { "epoch": 0.1596488996310902, "ewc_loss": 0.031982421875, "ewc_loss_diag": 2.1457672119140625e-05, "ewc_loss_parallel": 2.9802322387695312e-05, "grad_norm": 3.025163412094116, "learning_rate": 5.3158117846545155e-06, "loss": 0.4412, "mean_token_accuracy": 0.8609740734100342, "num_tokens": 47617336.0, "step": 1255 }, { "epoch": 0.1597761099096807, "ewc_loss": 0.0341796875, "ewc_loss_diag": 2.1457672119140625e-05, "ewc_loss_parallel": 3.1948089599609375e-05, "grad_norm": 3.3571879863739014, "learning_rate": 5.320050869012294e-06, "loss": 0.5177, "mean_token_accuracy": 0.8441038727760315, "num_tokens": 47662701.0, "step": 1256 }, { "epoch": 0.15990332018827122, "ewc_loss": 0.031982421875, "ewc_loss_diag": 2.1457672119140625e-05, "ewc_loss_parallel": 2.9921531677246094e-05, "grad_norm": 3.042433261871338, "learning_rate": 5.324289953370072e-06, "loss": 0.5654, "mean_token_accuracy": 0.8301965594291687, "num_tokens": 47703335.0, "step": 1257 }, { "epoch": 0.16003053046686172, "ewc_loss": 0.033935546875, "ewc_loss_diag": 2.1457672119140625e-05, "ewc_loss_parallel": 3.170967102050781e-05, "grad_norm": 3.3182713985443115, "learning_rate": 5.328529037727851e-06, "loss": 0.5075, "mean_token_accuracy": 0.8456588983535767, "num_tokens": 47746980.0, "step": 1258 }, { "epoch": 0.16015774074545222, "ewc_loss": 0.03173828125, "ewc_loss_diag": 2.1457672119140625e-05, "ewc_loss_parallel": 2.968311309814453e-05, "grad_norm": 3.055687189102173, "learning_rate": 5.33276812208563e-06, "loss": 0.5025, "mean_token_accuracy": 0.844289243221283, "num_tokens": 47784606.0, "step": 1259 }, { "epoch": 0.16028495102404275, "ewc_loss": 0.033447265625, "ewc_loss_diag": 2.1576881408691406e-05, "ewc_loss_parallel": 3.123283386230469e-05, "grad_norm": 3.2900688648223877, "learning_rate": 5.337007206443409e-06, "loss": 0.5408, "mean_token_accuracy": 0.8362795114517212, "num_tokens": 47821298.0, "step": 1260 }, { "epoch": 0.16041216130263325, "ewc_loss": 0.031494140625, "ewc_loss_diag": 2.1576881408691406e-05, "ewc_loss_parallel": 2.944469451904297e-05, "grad_norm": 3.0434648990631104, "learning_rate": 5.341246290801188e-06, "loss": 0.5027, "mean_token_accuracy": 0.8482800722122192, "num_tokens": 47863102.0, "step": 1261 }, { "epoch": 0.16053937158122375, "ewc_loss": 0.032958984375, "ewc_loss_diag": 2.1576881408691406e-05, "ewc_loss_parallel": 3.075599670410156e-05, "grad_norm": 3.25022029876709, "learning_rate": 5.345485375158966e-06, "loss": 0.4867, "mean_token_accuracy": 0.8503360748291016, "num_tokens": 47903124.0, "step": 1262 }, { "epoch": 0.16066658185981428, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 2.1696090698242188e-05, "ewc_loss_parallel": 2.8967857360839844e-05, "grad_norm": 3.0467822551727295, "learning_rate": 5.349724459516745e-06, "loss": 0.4817, "mean_token_accuracy": 0.8492473363876343, "num_tokens": 47939441.0, "step": 1263 }, { "epoch": 0.16079379213840478, "ewc_loss": 0.032958984375, "ewc_loss_diag": 2.1576881408691406e-05, "ewc_loss_parallel": 3.075599670410156e-05, "grad_norm": 3.228468179702759, "learning_rate": 5.353963543874523e-06, "loss": 0.5504, "mean_token_accuracy": 0.8343560695648193, "num_tokens": 47981236.0, "step": 1264 }, { "epoch": 0.1609210024169953, "ewc_loss": 0.031005859375, "ewc_loss_diag": 2.1696090698242188e-05, "ewc_loss_parallel": 2.8848648071289062e-05, "grad_norm": 3.0511393547058105, "learning_rate": 5.358202628232302e-06, "loss": 0.5165, "mean_token_accuracy": 0.844783365726471, "num_tokens": 48011431.0, "step": 1265 }, { "epoch": 0.1610482126955858, "ewc_loss": 0.0322265625, "ewc_loss_diag": 2.1696090698242188e-05, "ewc_loss_parallel": 3.0159950256347656e-05, "grad_norm": 3.1999571323394775, "learning_rate": 5.362441712590081e-06, "loss": 0.5044, "mean_token_accuracy": 0.8460187911987305, "num_tokens": 48051003.0, "step": 1266 }, { "epoch": 0.1611754229741763, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 2.1696090698242188e-05, "ewc_loss_parallel": 2.872943878173828e-05, "grad_norm": 3.028047561645508, "learning_rate": 5.3666807969478595e-06, "loss": 0.548, "mean_token_accuracy": 0.8335592746734619, "num_tokens": 48091239.0, "step": 1267 }, { "epoch": 0.16130263325276684, "ewc_loss": 0.0322265625, "ewc_loss_diag": 2.1696090698242188e-05, "ewc_loss_parallel": 3.0159950256347656e-05, "grad_norm": 3.2147555351257324, "learning_rate": 5.370919881305638e-06, "loss": 0.5366, "mean_token_accuracy": 0.8410215973854065, "num_tokens": 48126021.0, "step": 1268 }, { "epoch": 0.16142984353135734, "ewc_loss": 0.03076171875, "ewc_loss_diag": 2.1696090698242188e-05, "ewc_loss_parallel": 2.86102294921875e-05, "grad_norm": 3.024913787841797, "learning_rate": 5.375158965663417e-06, "loss": 0.5183, "mean_token_accuracy": 0.8444390296936035, "num_tokens": 48160459.0, "step": 1269 }, { "epoch": 0.16155705380994784, "ewc_loss": 0.032470703125, "ewc_loss_diag": 2.1696090698242188e-05, "ewc_loss_parallel": 3.0279159545898438e-05, "grad_norm": 3.198213815689087, "learning_rate": 5.379398050021196e-06, "loss": 0.508, "mean_token_accuracy": 0.8460283279418945, "num_tokens": 48200199.0, "step": 1270 }, { "epoch": 0.16168426408853837, "ewc_loss": 0.0306396484375, "ewc_loss_diag": 2.1696090698242188e-05, "ewc_loss_parallel": 2.849102020263672e-05, "grad_norm": 3.0139904022216797, "learning_rate": 5.383637134378975e-06, "loss": 0.4889, "mean_token_accuracy": 0.846927285194397, "num_tokens": 48233926.0, "step": 1271 }, { "epoch": 0.16181147436712887, "ewc_loss": 0.032470703125, "ewc_loss_diag": 2.1696090698242188e-05, "ewc_loss_parallel": 3.0279159545898438e-05, "grad_norm": 3.2085068225860596, "learning_rate": 5.387876218736753e-06, "loss": 0.5847, "mean_token_accuracy": 0.8195251822471619, "num_tokens": 48279667.0, "step": 1272 }, { "epoch": 0.16193868464571937, "ewc_loss": 0.0306396484375, "ewc_loss_diag": 2.1696090698242188e-05, "ewc_loss_parallel": 2.849102020263672e-05, "grad_norm": 2.9862754344940186, "learning_rate": 5.392115303094532e-06, "loss": 0.4709, "mean_token_accuracy": 0.8558787107467651, "num_tokens": 48319185.0, "step": 1273 }, { "epoch": 0.1620658949243099, "ewc_loss": 0.03271484375, "ewc_loss_diag": 2.1696090698242188e-05, "ewc_loss_parallel": 3.0517578125e-05, "grad_norm": 3.2361271381378174, "learning_rate": 5.396354387452311e-06, "loss": 0.4611, "mean_token_accuracy": 0.8606874942779541, "num_tokens": 48357250.0, "step": 1274 }, { "epoch": 0.1621931052029004, "ewc_loss": 0.0306396484375, "ewc_loss_diag": 2.1696090698242188e-05, "ewc_loss_parallel": 2.849102020263672e-05, "grad_norm": 2.976182222366333, "learning_rate": 5.40059347181009e-06, "loss": 0.5418, "mean_token_accuracy": 0.83573317527771, "num_tokens": 48396577.0, "step": 1275 }, { "epoch": 0.1623203154814909, "ewc_loss": 0.033447265625, "ewc_loss_diag": 2.181529998779297e-05, "ewc_loss_parallel": 3.123283386230469e-05, "grad_norm": 3.2991931438446045, "learning_rate": 5.404832556167869e-06, "loss": 0.5435, "mean_token_accuracy": 0.8331880569458008, "num_tokens": 48431360.0, "step": 1276 }, { "epoch": 0.16244752576008142, "ewc_loss": 0.0306396484375, "ewc_loss_diag": 2.181529998779297e-05, "ewc_loss_parallel": 2.849102020263672e-05, "grad_norm": 2.9509975910186768, "learning_rate": 5.4090716405256475e-06, "loss": 0.5048, "mean_token_accuracy": 0.8481326699256897, "num_tokens": 48471640.0, "step": 1277 }, { "epoch": 0.16257473603867192, "ewc_loss": 0.0341796875, "ewc_loss_diag": 2.181529998779297e-05, "ewc_loss_parallel": 3.1948089599609375e-05, "grad_norm": 3.3554506301879883, "learning_rate": 5.413310724883426e-06, "loss": 0.5649, "mean_token_accuracy": 0.8225943446159363, "num_tokens": 48509870.0, "step": 1278 }, { "epoch": 0.16270194631726242, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 2.181529998779297e-05, "ewc_loss_parallel": 2.872943878173828e-05, "grad_norm": 2.9434051513671875, "learning_rate": 5.417549809241205e-06, "loss": 0.5399, "mean_token_accuracy": 0.838209867477417, "num_tokens": 48554174.0, "step": 1279 }, { "epoch": 0.16282915659585295, "ewc_loss": 0.034912109375, "ewc_loss_diag": 2.181529998779297e-05, "ewc_loss_parallel": 3.266334533691406e-05, "grad_norm": 3.429910659790039, "learning_rate": 5.4217888935989825e-06, "loss": 0.5488, "mean_token_accuracy": 0.8354661464691162, "num_tokens": 48588590.0, "step": 1280 }, { "epoch": 0.16295636687444345, "ewc_loss": 0.03125, "ewc_loss_diag": 2.181529998779297e-05, "ewc_loss_parallel": 2.9206275939941406e-05, "grad_norm": 2.961182117462158, "learning_rate": 5.426027977956761e-06, "loss": 0.563, "mean_token_accuracy": 0.8278368711471558, "num_tokens": 48633567.0, "step": 1281 }, { "epoch": 0.16308357715303395, "ewc_loss": 0.035400390625, "ewc_loss_diag": 2.193450927734375e-05, "ewc_loss_parallel": 3.314018249511719e-05, "grad_norm": 3.4868879318237305, "learning_rate": 5.43026706231454e-06, "loss": 0.5882, "mean_token_accuracy": 0.8222897052764893, "num_tokens": 48673764.0, "step": 1282 }, { "epoch": 0.16321078743162448, "ewc_loss": 0.0322265625, "ewc_loss_diag": 2.193450927734375e-05, "ewc_loss_parallel": 2.9921531677246094e-05, "grad_norm": 2.9922661781311035, "learning_rate": 5.434506146672319e-06, "loss": 0.5605, "mean_token_accuracy": 0.8302515745162964, "num_tokens": 48722531.0, "step": 1283 }, { "epoch": 0.16333799771021498, "ewc_loss": 0.035888671875, "ewc_loss_diag": 2.193450927734375e-05, "ewc_loss_parallel": 3.361701965332031e-05, "grad_norm": 3.5350406169891357, "learning_rate": 5.438745231030098e-06, "loss": 0.5551, "mean_token_accuracy": 0.8293350338935852, "num_tokens": 48758273.0, "step": 1284 }, { "epoch": 0.16346520798880548, "ewc_loss": 0.033203125, "ewc_loss_diag": 2.193450927734375e-05, "ewc_loss_parallel": 3.0994415283203125e-05, "grad_norm": 3.0687267780303955, "learning_rate": 5.442984315387877e-06, "loss": 0.447, "mean_token_accuracy": 0.8622649312019348, "num_tokens": 48793016.0, "step": 1285 }, { "epoch": 0.163592418267396, "ewc_loss": 0.035888671875, "ewc_loss_diag": 2.205371856689453e-05, "ewc_loss_parallel": 3.361701965332031e-05, "grad_norm": 3.5122125148773193, "learning_rate": 5.447223399745656e-06, "loss": 0.5233, "mean_token_accuracy": 0.8441903591156006, "num_tokens": 48832204.0, "step": 1286 }, { "epoch": 0.1637196285459865, "ewc_loss": 0.033447265625, "ewc_loss_diag": 2.205371856689453e-05, "ewc_loss_parallel": 3.123283386230469e-05, "grad_norm": 3.1188344955444336, "learning_rate": 5.451462484103434e-06, "loss": 0.4922, "mean_token_accuracy": 0.8509756326675415, "num_tokens": 48867199.0, "step": 1287 }, { "epoch": 0.163846838824577, "ewc_loss": 0.03564453125, "ewc_loss_diag": 2.205371856689453e-05, "ewc_loss_parallel": 3.337860107421875e-05, "grad_norm": 3.4647481441497803, "learning_rate": 5.455701568461213e-06, "loss": 0.5131, "mean_token_accuracy": 0.8455236554145813, "num_tokens": 48901423.0, "step": 1288 }, { "epoch": 0.16397404910316754, "ewc_loss": 0.033447265625, "ewc_loss_diag": 2.205371856689453e-05, "ewc_loss_parallel": 3.123283386230469e-05, "grad_norm": 3.1520893573760986, "learning_rate": 5.459940652818992e-06, "loss": 0.5927, "mean_token_accuracy": 0.8208407759666443, "num_tokens": 48941883.0, "step": 1289 }, { "epoch": 0.16410125938175804, "ewc_loss": 0.034912109375, "ewc_loss_diag": 2.205371856689453e-05, "ewc_loss_parallel": 3.266334533691406e-05, "grad_norm": 3.407400131225586, "learning_rate": 5.4641797371767705e-06, "loss": 0.5235, "mean_token_accuracy": 0.8399218320846558, "num_tokens": 48971764.0, "step": 1290 }, { "epoch": 0.16422846966034857, "ewc_loss": 0.033203125, "ewc_loss_diag": 2.205371856689453e-05, "ewc_loss_parallel": 3.0994415283203125e-05, "grad_norm": 3.1485962867736816, "learning_rate": 5.468418821534549e-06, "loss": 0.4632, "mean_token_accuracy": 0.8574371337890625, "num_tokens": 49003687.0, "step": 1291 }, { "epoch": 0.16435567993893907, "ewc_loss": 0.0341796875, "ewc_loss_diag": 2.205371856689453e-05, "ewc_loss_parallel": 3.1948089599609375e-05, "grad_norm": 3.3390491008758545, "learning_rate": 5.472657905892328e-06, "loss": 0.4723, "mean_token_accuracy": 0.8565598726272583, "num_tokens": 49036896.0, "step": 1292 }, { "epoch": 0.16448289021752957, "ewc_loss": 0.03271484375, "ewc_loss_diag": 2.205371856689453e-05, "ewc_loss_parallel": 3.0517578125e-05, "grad_norm": 3.1348869800567627, "learning_rate": 5.476896990250107e-06, "loss": 0.5876, "mean_token_accuracy": 0.8211359977722168, "num_tokens": 49076549.0, "step": 1293 }, { "epoch": 0.1646101004961201, "ewc_loss": 0.033447265625, "ewc_loss_diag": 2.2172927856445312e-05, "ewc_loss_parallel": 3.123283386230469e-05, "grad_norm": 3.2815680503845215, "learning_rate": 5.481136074607886e-06, "loss": 0.4732, "mean_token_accuracy": 0.8507136106491089, "num_tokens": 49114507.0, "step": 1294 }, { "epoch": 0.1647373107747106, "ewc_loss": 0.032470703125, "ewc_loss_diag": 2.2292137145996094e-05, "ewc_loss_parallel": 3.0159950256347656e-05, "grad_norm": 3.1074559688568115, "learning_rate": 5.485375158965663e-06, "loss": 0.4821, "mean_token_accuracy": 0.8547055721282959, "num_tokens": 49159794.0, "step": 1295 }, { "epoch": 0.1648645210533011, "ewc_loss": 0.033447265625, "ewc_loss_diag": 2.2292137145996094e-05, "ewc_loss_parallel": 3.123283386230469e-05, "grad_norm": 3.2661399841308594, "learning_rate": 5.489614243323442e-06, "loss": 0.5223, "mean_token_accuracy": 0.8406546115875244, "num_tokens": 49198766.0, "step": 1296 }, { "epoch": 0.16499173133189163, "ewc_loss": 0.0322265625, "ewc_loss_diag": 2.2292137145996094e-05, "ewc_loss_parallel": 2.9921531677246094e-05, "grad_norm": 3.096085786819458, "learning_rate": 5.493853327681221e-06, "loss": 0.4652, "mean_token_accuracy": 0.8570379018783569, "num_tokens": 49236481.0, "step": 1297 }, { "epoch": 0.16511894161048213, "ewc_loss": 0.033447265625, "ewc_loss_diag": 2.2411346435546875e-05, "ewc_loss_parallel": 3.123283386230469e-05, "grad_norm": 3.27077579498291, "learning_rate": 5.498092412039e-06, "loss": 0.5909, "mean_token_accuracy": 0.8249216079711914, "num_tokens": 49271730.0, "step": 1298 }, { "epoch": 0.16524615188907263, "ewc_loss": 0.031982421875, "ewc_loss_diag": 2.2411346435546875e-05, "ewc_loss_parallel": 2.9802322387695312e-05, "grad_norm": 3.0616202354431152, "learning_rate": 5.502331496396779e-06, "loss": 0.4799, "mean_token_accuracy": 0.851171612739563, "num_tokens": 49316208.0, "step": 1299 }, { "epoch": 0.16537336216766316, "ewc_loss": 0.033447265625, "ewc_loss_diag": 2.2411346435546875e-05, "ewc_loss_parallel": 3.123283386230469e-05, "grad_norm": 3.2737369537353516, "learning_rate": 5.506570580754558e-06, "loss": 0.5186, "mean_token_accuracy": 0.8421190977096558, "num_tokens": 49355343.0, "step": 1300 }, { "epoch": 0.16550057244625366, "ewc_loss": 0.031982421875, "ewc_loss_diag": 2.2411346435546875e-05, "ewc_loss_parallel": 2.9802322387695312e-05, "grad_norm": 3.0645899772644043, "learning_rate": 5.5108096651123365e-06, "loss": 0.4937, "mean_token_accuracy": 0.8501702547073364, "num_tokens": 49394055.0, "step": 1301 }, { "epoch": 0.16562778272484416, "ewc_loss": 0.03369140625, "ewc_loss_diag": 2.2411346435546875e-05, "ewc_loss_parallel": 3.147125244140625e-05, "grad_norm": 3.2910468578338623, "learning_rate": 5.5150487494701146e-06, "loss": 0.5219, "mean_token_accuracy": 0.8430128693580627, "num_tokens": 49432466.0, "step": 1302 }, { "epoch": 0.1657549930034347, "ewc_loss": 0.0322265625, "ewc_loss_diag": 2.2530555725097656e-05, "ewc_loss_parallel": 2.9921531677246094e-05, "grad_norm": 3.0406813621520996, "learning_rate": 5.5192878338278934e-06, "loss": 0.5034, "mean_token_accuracy": 0.8452360033988953, "num_tokens": 49470746.0, "step": 1303 }, { "epoch": 0.1658822032820252, "ewc_loss": 0.033935546875, "ewc_loss_diag": 2.2530555725097656e-05, "ewc_loss_parallel": 3.170967102050781e-05, "grad_norm": 3.3459572792053223, "learning_rate": 5.523526918185672e-06, "loss": 0.5518, "mean_token_accuracy": 0.8311339616775513, "num_tokens": 49502879.0, "step": 1304 }, { "epoch": 0.1660094135606157, "ewc_loss": 0.031982421875, "ewc_loss_diag": 2.2530555725097656e-05, "ewc_loss_parallel": 2.968311309814453e-05, "grad_norm": 3.0255117416381836, "learning_rate": 5.527766002543451e-06, "loss": 0.4663, "mean_token_accuracy": 0.8534674644470215, "num_tokens": 49534875.0, "step": 1305 }, { "epoch": 0.16613662383920622, "ewc_loss": 0.034912109375, "ewc_loss_diag": 2.2530555725097656e-05, "ewc_loss_parallel": 3.266334533691406e-05, "grad_norm": 3.3715837001800537, "learning_rate": 5.53200508690123e-06, "loss": 0.5371, "mean_token_accuracy": 0.8413802981376648, "num_tokens": 49573537.0, "step": 1306 }, { "epoch": 0.16626383411779672, "ewc_loss": 0.0322265625, "ewc_loss_diag": 2.2530555725097656e-05, "ewc_loss_parallel": 3.0040740966796875e-05, "grad_norm": 2.999905824661255, "learning_rate": 5.536244171259009e-06, "loss": 0.4995, "mean_token_accuracy": 0.8486203551292419, "num_tokens": 49615190.0, "step": 1307 }, { "epoch": 0.16639104439638722, "ewc_loss": 0.03515625, "ewc_loss_diag": 2.2530555725097656e-05, "ewc_loss_parallel": 3.2901763916015625e-05, "grad_norm": 3.441119909286499, "learning_rate": 5.540483255616788e-06, "loss": 0.6004, "mean_token_accuracy": 0.8182637691497803, "num_tokens": 49649788.0, "step": 1308 }, { "epoch": 0.16651825467497774, "ewc_loss": 0.0322265625, "ewc_loss_diag": 2.276897430419922e-05, "ewc_loss_parallel": 2.9921531677246094e-05, "grad_norm": 3.015699863433838, "learning_rate": 5.544722339974567e-06, "loss": 0.5088, "mean_token_accuracy": 0.8445776700973511, "num_tokens": 49683728.0, "step": 1309 }, { "epoch": 0.16664546495356825, "ewc_loss": 0.035888671875, "ewc_loss_diag": 2.288818359375e-05, "ewc_loss_parallel": 3.361701965332031e-05, "grad_norm": 3.4842875003814697, "learning_rate": 5.548961424332344e-06, "loss": 0.5051, "mean_token_accuracy": 0.8411580920219421, "num_tokens": 49719331.0, "step": 1310 }, { "epoch": 0.16677267523215875, "ewc_loss": 0.032470703125, "ewc_loss_diag": 2.288818359375e-05, "ewc_loss_parallel": 3.0159950256347656e-05, "grad_norm": 3.0208141803741455, "learning_rate": 5.553200508690123e-06, "loss": 0.4604, "mean_token_accuracy": 0.8592890501022339, "num_tokens": 49760158.0, "step": 1311 }, { "epoch": 0.16689988551074927, "ewc_loss": 0.036376953125, "ewc_loss_diag": 2.288818359375e-05, "ewc_loss_parallel": 3.409385681152344e-05, "grad_norm": 3.546365261077881, "learning_rate": 5.557439593047902e-06, "loss": 0.5659, "mean_token_accuracy": 0.827698826789856, "num_tokens": 49795285.0, "step": 1312 }, { "epoch": 0.16702709578933977, "ewc_loss": 0.032958984375, "ewc_loss_diag": 2.288818359375e-05, "ewc_loss_parallel": 3.075599670410156e-05, "grad_norm": 3.0641586780548096, "learning_rate": 5.561678677405681e-06, "loss": 0.516, "mean_token_accuracy": 0.8432436585426331, "num_tokens": 49839071.0, "step": 1313 }, { "epoch": 0.16715430606793028, "ewc_loss": 0.03662109375, "ewc_loss_diag": 2.300739288330078e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 3.5726730823516846, "learning_rate": 5.5659177617634595e-06, "loss": 0.4975, "mean_token_accuracy": 0.8500539064407349, "num_tokens": 49870063.0, "step": 1314 }, { "epoch": 0.1672815163465208, "ewc_loss": 0.033447265625, "ewc_loss_diag": 2.300739288330078e-05, "ewc_loss_parallel": 3.123283386230469e-05, "grad_norm": 3.119701385498047, "learning_rate": 5.570156846121238e-06, "loss": 0.5319, "mean_token_accuracy": 0.8367213010787964, "num_tokens": 49908232.0, "step": 1315 }, { "epoch": 0.1674087266251113, "ewc_loss": 0.036376953125, "ewc_loss_diag": 2.288818359375e-05, "ewc_loss_parallel": 3.409385681152344e-05, "grad_norm": 3.551805019378662, "learning_rate": 5.574395930479017e-06, "loss": 0.5572, "mean_token_accuracy": 0.8329127430915833, "num_tokens": 49944676.0, "step": 1316 }, { "epoch": 0.16753593690370183, "ewc_loss": 0.033935546875, "ewc_loss_diag": 2.300739288330078e-05, "ewc_loss_parallel": 3.170967102050781e-05, "grad_norm": 3.1705691814422607, "learning_rate": 5.578635014836796e-06, "loss": 0.4975, "mean_token_accuracy": 0.8513516187667847, "num_tokens": 49978472.0, "step": 1317 }, { "epoch": 0.16766314718229233, "ewc_loss": 0.035888671875, "ewc_loss_diag": 2.300739288330078e-05, "ewc_loss_parallel": 3.361701965332031e-05, "grad_norm": 3.497643232345581, "learning_rate": 5.582874099194574e-06, "loss": 0.5518, "mean_token_accuracy": 0.8340710401535034, "num_tokens": 50010564.0, "step": 1318 }, { "epoch": 0.16779035746088283, "ewc_loss": 0.03369140625, "ewc_loss_diag": 2.288818359375e-05, "ewc_loss_parallel": 3.147125244140625e-05, "grad_norm": 3.1821422576904297, "learning_rate": 5.587113183552353e-06, "loss": 0.5353, "mean_token_accuracy": 0.8374737501144409, "num_tokens": 50049538.0, "step": 1319 }, { "epoch": 0.16791756773947336, "ewc_loss": 0.03564453125, "ewc_loss_diag": 2.288818359375e-05, "ewc_loss_parallel": 3.337860107421875e-05, "grad_norm": 3.425689220428467, "learning_rate": 5.591352267910132e-06, "loss": 0.4942, "mean_token_accuracy": 0.8510140180587769, "num_tokens": 50090779.0, "step": 1320 }, { "epoch": 0.16804477801806386, "ewc_loss": 0.033203125, "ewc_loss_diag": 2.300739288330078e-05, "ewc_loss_parallel": 3.0994415283203125e-05, "grad_norm": 3.186565637588501, "learning_rate": 5.595591352267911e-06, "loss": 0.538, "mean_token_accuracy": 0.8385326862335205, "num_tokens": 50129527.0, "step": 1321 }, { "epoch": 0.16817198829665436, "ewc_loss": 0.03515625, "ewc_loss_diag": 2.300739288330078e-05, "ewc_loss_parallel": 3.2901763916015625e-05, "grad_norm": 3.3804171085357666, "learning_rate": 5.59983043662569e-06, "loss": 0.5052, "mean_token_accuracy": 0.846413254737854, "num_tokens": 50167950.0, "step": 1322 }, { "epoch": 0.1682991985752449, "ewc_loss": 0.033203125, "ewc_loss_diag": 2.300739288330078e-05, "ewc_loss_parallel": 3.0994415283203125e-05, "grad_norm": 3.175361156463623, "learning_rate": 5.604069520983469e-06, "loss": 0.5054, "mean_token_accuracy": 0.8476668000221252, "num_tokens": 50205079.0, "step": 1323 }, { "epoch": 0.1684264088538354, "ewc_loss": 0.034423828125, "ewc_loss_diag": 2.300739288330078e-05, "ewc_loss_parallel": 3.218650817871094e-05, "grad_norm": 3.3493518829345703, "learning_rate": 5.6083086053412475e-06, "loss": 0.4961, "mean_token_accuracy": 0.8485097885131836, "num_tokens": 50237909.0, "step": 1324 }, { "epoch": 0.1685536191324259, "ewc_loss": 0.033203125, "ewc_loss_diag": 2.300739288330078e-05, "ewc_loss_parallel": 3.0994415283203125e-05, "grad_norm": 3.165041446685791, "learning_rate": 5.612547689699025e-06, "loss": 0.5572, "mean_token_accuracy": 0.8343067765235901, "num_tokens": 50272825.0, "step": 1325 }, { "epoch": 0.16868082941101642, "ewc_loss": 0.03466796875, "ewc_loss_diag": 2.3245811462402344e-05, "ewc_loss_parallel": 3.218650817871094e-05, "grad_norm": 3.310706853866577, "learning_rate": 5.6167867740568036e-06, "loss": 0.5439, "mean_token_accuracy": 0.8389561176300049, "num_tokens": 50312964.0, "step": 1326 }, { "epoch": 0.16880803968960692, "ewc_loss": 0.03271484375, "ewc_loss_diag": 2.3245811462402344e-05, "ewc_loss_parallel": 3.0517578125e-05, "grad_norm": 3.134861946105957, "learning_rate": 5.6210258584145824e-06, "loss": 0.543, "mean_token_accuracy": 0.8341400027275085, "num_tokens": 50355912.0, "step": 1327 }, { "epoch": 0.16893524996819742, "ewc_loss": 0.03466796875, "ewc_loss_diag": 2.3245811462402344e-05, "ewc_loss_parallel": 3.218650817871094e-05, "grad_norm": 3.3025248050689697, "learning_rate": 5.625264942772361e-06, "loss": 0.5265, "mean_token_accuracy": 0.8440589904785156, "num_tokens": 50395875.0, "step": 1328 }, { "epoch": 0.16906246024678795, "ewc_loss": 0.03271484375, "ewc_loss_diag": 2.3245811462402344e-05, "ewc_loss_parallel": 3.0517578125e-05, "grad_norm": 3.1306817531585693, "learning_rate": 5.62950402713014e-06, "loss": 0.5779, "mean_token_accuracy": 0.8317317366600037, "num_tokens": 50435628.0, "step": 1329 }, { "epoch": 0.16918967052537845, "ewc_loss": 0.03466796875, "ewc_loss_diag": 2.3245811462402344e-05, "ewc_loss_parallel": 3.218650817871094e-05, "grad_norm": 3.3103909492492676, "learning_rate": 5.633743111487919e-06, "loss": 0.5833, "mean_token_accuracy": 0.820349395275116, "num_tokens": 50474812.0, "step": 1330 }, { "epoch": 0.16931688080396895, "ewc_loss": 0.03271484375, "ewc_loss_diag": 2.3245811462402344e-05, "ewc_loss_parallel": 3.039836883544922e-05, "grad_norm": 3.1075398921966553, "learning_rate": 5.637982195845698e-06, "loss": 0.4614, "mean_token_accuracy": 0.8622487783432007, "num_tokens": 50510058.0, "step": 1331 }, { "epoch": 0.16944409108255948, "ewc_loss": 0.03466796875, "ewc_loss_diag": 2.3245811462402344e-05, "ewc_loss_parallel": 3.218650817871094e-05, "grad_norm": 3.331157922744751, "learning_rate": 5.642221280203477e-06, "loss": 0.4659, "mean_token_accuracy": 0.8592830896377563, "num_tokens": 50546282.0, "step": 1332 }, { "epoch": 0.16957130136114998, "ewc_loss": 0.03271484375, "ewc_loss_diag": 2.3245811462402344e-05, "ewc_loss_parallel": 3.0279159545898438e-05, "grad_norm": 3.110239028930664, "learning_rate": 5.646460364561255e-06, "loss": 0.4786, "mean_token_accuracy": 0.8514779806137085, "num_tokens": 50577457.0, "step": 1333 }, { "epoch": 0.16969851163974048, "ewc_loss": 0.03466796875, "ewc_loss_diag": 2.3245811462402344e-05, "ewc_loss_parallel": 3.24249267578125e-05, "grad_norm": 3.339818239212036, "learning_rate": 5.650699448919034e-06, "loss": 0.5783, "mean_token_accuracy": 0.82547926902771, "num_tokens": 50619754.0, "step": 1334 }, { "epoch": 0.169825721918331, "ewc_loss": 0.03271484375, "ewc_loss_diag": 2.3245811462402344e-05, "ewc_loss_parallel": 3.0279159545898438e-05, "grad_norm": 3.0853989124298096, "learning_rate": 5.654938533276813e-06, "loss": 0.5129, "mean_token_accuracy": 0.8436065316200256, "num_tokens": 50660674.0, "step": 1335 }, { "epoch": 0.1699529321969215, "ewc_loss": 0.03515625, "ewc_loss_diag": 2.3245811462402344e-05, "ewc_loss_parallel": 3.266334533691406e-05, "grad_norm": 3.3687891960144043, "learning_rate": 5.6591776176345916e-06, "loss": 0.5075, "mean_token_accuracy": 0.8450380563735962, "num_tokens": 50698404.0, "step": 1336 }, { "epoch": 0.170080142475512, "ewc_loss": 0.032470703125, "ewc_loss_diag": 2.3245811462402344e-05, "ewc_loss_parallel": 3.0159950256347656e-05, "grad_norm": 3.0649514198303223, "learning_rate": 5.6634167019923704e-06, "loss": 0.5145, "mean_token_accuracy": 0.8427913188934326, "num_tokens": 50736329.0, "step": 1337 }, { "epoch": 0.17020735275410254, "ewc_loss": 0.03564453125, "ewc_loss_diag": 2.3245811462402344e-05, "ewc_loss_parallel": 3.314018249511719e-05, "grad_norm": 3.4291739463806152, "learning_rate": 5.667655786350149e-06, "loss": 0.5755, "mean_token_accuracy": 0.8183311223983765, "num_tokens": 50770065.0, "step": 1338 }, { "epoch": 0.17033456303269304, "ewc_loss": 0.03271484375, "ewc_loss_diag": 2.3245811462402344e-05, "ewc_loss_parallel": 3.0279159545898438e-05, "grad_norm": 3.043600559234619, "learning_rate": 5.671894870707928e-06, "loss": 0.5073, "mean_token_accuracy": 0.848497211933136, "num_tokens": 50811603.0, "step": 1339 }, { "epoch": 0.17046177331128357, "ewc_loss": 0.0361328125, "ewc_loss_diag": 2.3245811462402344e-05, "ewc_loss_parallel": 3.3855438232421875e-05, "grad_norm": 3.487795829772949, "learning_rate": 5.676133955065705e-06, "loss": 0.4516, "mean_token_accuracy": 0.8602604866027832, "num_tokens": 50846141.0, "step": 1340 }, { "epoch": 0.17058898358987407, "ewc_loss": 0.03271484375, "ewc_loss_diag": 2.3365020751953125e-05, "ewc_loss_parallel": 3.039836883544922e-05, "grad_norm": 3.0405468940734863, "learning_rate": 5.680373039423484e-06, "loss": 0.4962, "mean_token_accuracy": 0.8490259647369385, "num_tokens": 50885511.0, "step": 1341 }, { "epoch": 0.17071619386846457, "ewc_loss": 0.037353515625, "ewc_loss_diag": 2.3365020751953125e-05, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 3.574481964111328, "learning_rate": 5.684612123781263e-06, "loss": 0.5453, "mean_token_accuracy": 0.8322412967681885, "num_tokens": 50926374.0, "step": 1342 }, { "epoch": 0.1708434041470551, "ewc_loss": 0.033447265625, "ewc_loss_diag": 2.3365020751953125e-05, "ewc_loss_parallel": 3.0994415283203125e-05, "grad_norm": 3.070106029510498, "learning_rate": 5.688851208139042e-06, "loss": 0.555, "mean_token_accuracy": 0.8275721669197083, "num_tokens": 50965453.0, "step": 1343 }, { "epoch": 0.1709706144256456, "ewc_loss": 0.03759765625, "ewc_loss_diag": 2.3365020751953125e-05, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 3.6342806816101074, "learning_rate": 5.693090292496821e-06, "loss": 0.5492, "mean_token_accuracy": 0.8395569324493408, "num_tokens": 51007921.0, "step": 1344 }, { "epoch": 0.1710978247042361, "ewc_loss": 0.033935546875, "ewc_loss_diag": 2.3365020751953125e-05, "ewc_loss_parallel": 3.147125244140625e-05, "grad_norm": 3.1125426292419434, "learning_rate": 5.6973293768546e-06, "loss": 0.5211, "mean_token_accuracy": 0.8417098522186279, "num_tokens": 51046982.0, "step": 1345 }, { "epoch": 0.17122503498282662, "ewc_loss": 0.0380859375, "ewc_loss_diag": 2.3365020751953125e-05, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 3.6664555072784424, "learning_rate": 5.701568461212379e-06, "loss": 0.4636, "mean_token_accuracy": 0.8609984517097473, "num_tokens": 51083192.0, "step": 1346 }, { "epoch": 0.17135224526141712, "ewc_loss": 0.034912109375, "ewc_loss_diag": 2.3603439331054688e-05, "ewc_loss_parallel": 3.24249267578125e-05, "grad_norm": 3.174152135848999, "learning_rate": 5.705807545570158e-06, "loss": 0.4696, "mean_token_accuracy": 0.8565855622291565, "num_tokens": 51125258.0, "step": 1347 }, { "epoch": 0.17147945554000762, "ewc_loss": 0.0380859375, "ewc_loss_diag": 2.3603439331054688e-05, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 3.6537039279937744, "learning_rate": 5.710046629927936e-06, "loss": 0.5104, "mean_token_accuracy": 0.8455476760864258, "num_tokens": 51164036.0, "step": 1348 }, { "epoch": 0.17160666581859815, "ewc_loss": 0.03515625, "ewc_loss_diag": 2.3603439331054688e-05, "ewc_loss_parallel": 3.266334533691406e-05, "grad_norm": 3.2308576107025146, "learning_rate": 5.7142857142857145e-06, "loss": 0.4722, "mean_token_accuracy": 0.8544250130653381, "num_tokens": 51201288.0, "step": 1349 }, { "epoch": 0.17173387609718865, "ewc_loss": 0.037353515625, "ewc_loss_diag": 2.3603439331054688e-05, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 3.6070988178253174, "learning_rate": 5.718524798643493e-06, "loss": 0.544, "mean_token_accuracy": 0.8364945650100708, "num_tokens": 51235931.0, "step": 1350 }, { "epoch": 0.17186108637577915, "ewc_loss": 0.034912109375, "ewc_loss_diag": 2.3603439331054688e-05, "ewc_loss_parallel": 3.24249267578125e-05, "grad_norm": 3.2695281505584717, "learning_rate": 5.722763883001272e-06, "loss": 0.5467, "mean_token_accuracy": 0.8356276750564575, "num_tokens": 51277188.0, "step": 1351 }, { "epoch": 0.17198829665436968, "ewc_loss": 0.03662109375, "ewc_loss_diag": 2.3603439331054688e-05, "ewc_loss_parallel": 3.409385681152344e-05, "grad_norm": 3.534595012664795, "learning_rate": 5.727002967359051e-06, "loss": 0.5173, "mean_token_accuracy": 0.8432120680809021, "num_tokens": 51313829.0, "step": 1352 }, { "epoch": 0.17211550693296018, "ewc_loss": 0.034912109375, "ewc_loss_diag": 2.3603439331054688e-05, "ewc_loss_parallel": 3.24249267578125e-05, "grad_norm": 3.2613418102264404, "learning_rate": 5.73124205171683e-06, "loss": 0.5769, "mean_token_accuracy": 0.8218676447868347, "num_tokens": 51355678.0, "step": 1353 }, { "epoch": 0.17224271721155068, "ewc_loss": 0.0361328125, "ewc_loss_diag": 2.3603439331054688e-05, "ewc_loss_parallel": 3.361701965332031e-05, "grad_norm": 3.457613229751587, "learning_rate": 5.735481136074609e-06, "loss": 0.4884, "mean_token_accuracy": 0.851915717124939, "num_tokens": 51390125.0, "step": 1354 }, { "epoch": 0.1723699274901412, "ewc_loss": 0.034423828125, "ewc_loss_diag": 2.372264862060547e-05, "ewc_loss_parallel": 3.1948089599609375e-05, "grad_norm": 3.2519214153289795, "learning_rate": 5.739720220432386e-06, "loss": 0.4874, "mean_token_accuracy": 0.8496619462966919, "num_tokens": 51426384.0, "step": 1355 }, { "epoch": 0.1724971377687317, "ewc_loss": 0.035888671875, "ewc_loss_diag": 2.372264862060547e-05, "ewc_loss_parallel": 3.337860107421875e-05, "grad_norm": 3.406334400177002, "learning_rate": 5.743959304790165e-06, "loss": 0.5843, "mean_token_accuracy": 0.8274555206298828, "num_tokens": 51467087.0, "step": 1356 }, { "epoch": 0.1726243480473222, "ewc_loss": 0.033935546875, "ewc_loss_diag": 2.372264862060547e-05, "ewc_loss_parallel": 3.147125244140625e-05, "grad_norm": 3.2352583408355713, "learning_rate": 5.748198389147944e-06, "loss": 0.5366, "mean_token_accuracy": 0.8385473489761353, "num_tokens": 51505126.0, "step": 1357 }, { "epoch": 0.17275155832591274, "ewc_loss": 0.03564453125, "ewc_loss_diag": 2.372264862060547e-05, "ewc_loss_parallel": 3.314018249511719e-05, "grad_norm": 3.3702151775360107, "learning_rate": 5.752437473505723e-06, "loss": 0.5625, "mean_token_accuracy": 0.8307989239692688, "num_tokens": 51546623.0, "step": 1358 }, { "epoch": 0.17287876860450324, "ewc_loss": 0.03369140625, "ewc_loss_diag": 2.372264862060547e-05, "ewc_loss_parallel": 3.123283386230469e-05, "grad_norm": 3.2167468070983887, "learning_rate": 5.756676557863502e-06, "loss": 0.5307, "mean_token_accuracy": 0.8373839855194092, "num_tokens": 51583175.0, "step": 1359 }, { "epoch": 0.17300597888309374, "ewc_loss": 0.035400390625, "ewc_loss_diag": 2.372264862060547e-05, "ewc_loss_parallel": 3.2901763916015625e-05, "grad_norm": 3.348076581954956, "learning_rate": 5.7609156422212805e-06, "loss": 0.4889, "mean_token_accuracy": 0.8502970933914185, "num_tokens": 51621454.0, "step": 1360 }, { "epoch": 0.17313318916168427, "ewc_loss": 0.03369140625, "ewc_loss_diag": 2.372264862060547e-05, "ewc_loss_parallel": 3.123283386230469e-05, "grad_norm": 3.196213960647583, "learning_rate": 5.7651547265790594e-06, "loss": 0.5114, "mean_token_accuracy": 0.845513641834259, "num_tokens": 51659482.0, "step": 1361 }, { "epoch": 0.17326039944027477, "ewc_loss": 0.03515625, "ewc_loss_diag": 2.384185791015625e-05, "ewc_loss_parallel": 3.266334533691406e-05, "grad_norm": 3.3544821739196777, "learning_rate": 5.769393810936838e-06, "loss": 0.5134, "mean_token_accuracy": 0.8438827991485596, "num_tokens": 51695478.0, "step": 1362 }, { "epoch": 0.17338760971886527, "ewc_loss": 0.033447265625, "ewc_loss_diag": 2.384185791015625e-05, "ewc_loss_parallel": 3.0994415283203125e-05, "grad_norm": 3.183685779571533, "learning_rate": 5.773632895294616e-06, "loss": 0.5597, "mean_token_accuracy": 0.8279603719711304, "num_tokens": 51736436.0, "step": 1363 }, { "epoch": 0.1735148199974558, "ewc_loss": 0.03515625, "ewc_loss_diag": 2.384185791015625e-05, "ewc_loss_parallel": 3.266334533691406e-05, "grad_norm": 3.361651659011841, "learning_rate": 5.777871979652395e-06, "loss": 0.492, "mean_token_accuracy": 0.8506574034690857, "num_tokens": 51772502.0, "step": 1364 }, { "epoch": 0.1736420302760463, "ewc_loss": 0.033203125, "ewc_loss_diag": 2.384185791015625e-05, "ewc_loss_parallel": 3.075599670410156e-05, "grad_norm": 3.143768787384033, "learning_rate": 5.782111064010174e-06, "loss": 0.5238, "mean_token_accuracy": 0.8396239280700684, "num_tokens": 51816343.0, "step": 1365 }, { "epoch": 0.17376924055463683, "ewc_loss": 0.03564453125, "ewc_loss_diag": 2.384185791015625e-05, "ewc_loss_parallel": 3.314018249511719e-05, "grad_norm": 3.3889541625976562, "learning_rate": 5.786350148367953e-06, "loss": 0.5189, "mean_token_accuracy": 0.8430920839309692, "num_tokens": 51861317.0, "step": 1366 }, { "epoch": 0.17389645083322733, "ewc_loss": 0.033203125, "ewc_loss_diag": 2.384185791015625e-05, "ewc_loss_parallel": 3.075599670410156e-05, "grad_norm": 3.1225273609161377, "learning_rate": 5.790589232725732e-06, "loss": 0.5199, "mean_token_accuracy": 0.8381892442703247, "num_tokens": 51902994.0, "step": 1367 }, { "epoch": 0.17402366111181783, "ewc_loss": 0.035888671875, "ewc_loss_diag": 2.384185791015625e-05, "ewc_loss_parallel": 3.337860107421875e-05, "grad_norm": 3.454472541809082, "learning_rate": 5.794828317083511e-06, "loss": 0.507, "mean_token_accuracy": 0.8469290733337402, "num_tokens": 51936374.0, "step": 1368 }, { "epoch": 0.17415087139040836, "ewc_loss": 0.033203125, "ewc_loss_diag": 2.396106719970703e-05, "ewc_loss_parallel": 3.075599670410156e-05, "grad_norm": 3.109772205352783, "learning_rate": 5.79906740144129e-06, "loss": 0.5046, "mean_token_accuracy": 0.8461138010025024, "num_tokens": 51970235.0, "step": 1369 }, { "epoch": 0.17427808166899886, "ewc_loss": 0.036865234375, "ewc_loss_diag": 2.396106719970703e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 3.511408805847168, "learning_rate": 5.8033064857990685e-06, "loss": 0.4965, "mean_token_accuracy": 0.8513616323471069, "num_tokens": 52006220.0, "step": 1370 }, { "epoch": 0.17440529194758936, "ewc_loss": 0.033203125, "ewc_loss_diag": 2.396106719970703e-05, "ewc_loss_parallel": 3.075599670410156e-05, "grad_norm": 3.0740537643432617, "learning_rate": 5.807545570156847e-06, "loss": 0.4745, "mean_token_accuracy": 0.8551980257034302, "num_tokens": 52046634.0, "step": 1371 }, { "epoch": 0.1745325022261799, "ewc_loss": 0.037353515625, "ewc_loss_diag": 2.396106719970703e-05, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 3.6161234378814697, "learning_rate": 5.811784654514625e-06, "loss": 0.5686, "mean_token_accuracy": 0.8287883400917053, "num_tokens": 52079768.0, "step": 1372 }, { "epoch": 0.1746597125047704, "ewc_loss": 0.033447265625, "ewc_loss_diag": 2.396106719970703e-05, "ewc_loss_parallel": 3.0994415283203125e-05, "grad_norm": 3.0836031436920166, "learning_rate": 5.8160237388724035e-06, "loss": 0.4902, "mean_token_accuracy": 0.8486090898513794, "num_tokens": 52118160.0, "step": 1373 }, { "epoch": 0.1747869227833609, "ewc_loss": 0.038330078125, "ewc_loss_diag": 2.396106719970703e-05, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 3.698578357696533, "learning_rate": 5.820262823230182e-06, "loss": 0.5356, "mean_token_accuracy": 0.8384355902671814, "num_tokens": 52160219.0, "step": 1374 }, { "epoch": 0.17491413306195142, "ewc_loss": 0.034423828125, "ewc_loss_diag": 2.396106719970703e-05, "ewc_loss_parallel": 3.1948089599609375e-05, "grad_norm": 3.129265785217285, "learning_rate": 5.824501907587961e-06, "loss": 0.4867, "mean_token_accuracy": 0.8504840135574341, "num_tokens": 52194746.0, "step": 1375 }, { "epoch": 0.17504134334054192, "ewc_loss": 0.0390625, "ewc_loss_diag": 2.396106719970703e-05, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 3.776181221008301, "learning_rate": 5.82874099194574e-06, "loss": 0.5834, "mean_token_accuracy": 0.821668803691864, "num_tokens": 52232625.0, "step": 1376 }, { "epoch": 0.17516855361913242, "ewc_loss": 0.035400390625, "ewc_loss_diag": 2.396106719970703e-05, "ewc_loss_parallel": 3.2901763916015625e-05, "grad_norm": 3.2085001468658447, "learning_rate": 5.832980076303519e-06, "loss": 0.5794, "mean_token_accuracy": 0.8246113061904907, "num_tokens": 52273030.0, "step": 1377 }, { "epoch": 0.17529576389772294, "ewc_loss": 0.039306640625, "ewc_loss_diag": 2.396106719970703e-05, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 3.7865636348724365, "learning_rate": 5.837219160661297e-06, "loss": 0.4733, "mean_token_accuracy": 0.8555181622505188, "num_tokens": 52306490.0, "step": 1378 }, { "epoch": 0.17542297417631345, "ewc_loss": 0.035888671875, "ewc_loss_diag": 2.4080276489257812e-05, "ewc_loss_parallel": 3.337860107421875e-05, "grad_norm": 3.282256841659546, "learning_rate": 5.841458245019076e-06, "loss": 0.4651, "mean_token_accuracy": 0.8580793142318726, "num_tokens": 52346717.0, "step": 1379 }, { "epoch": 0.17555018445490395, "ewc_loss": 0.038818359375, "ewc_loss_diag": 2.4080276489257812e-05, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 3.7403438091278076, "learning_rate": 5.845697329376855e-06, "loss": 0.5171, "mean_token_accuracy": 0.8468074798583984, "num_tokens": 52378165.0, "step": 1380 }, { "epoch": 0.17567739473349447, "ewc_loss": 0.0361328125, "ewc_loss_diag": 2.4080276489257812e-05, "ewc_loss_parallel": 3.361701965332031e-05, "grad_norm": 3.345395803451538, "learning_rate": 5.849936413734634e-06, "loss": 0.5781, "mean_token_accuracy": 0.8261246681213379, "num_tokens": 52418719.0, "step": 1381 }, { "epoch": 0.17580460501208497, "ewc_loss": 0.0380859375, "ewc_loss_diag": 2.4080276489257812e-05, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 3.635453701019287, "learning_rate": 5.854175498092413e-06, "loss": 0.4898, "mean_token_accuracy": 0.852691650390625, "num_tokens": 52459094.0, "step": 1382 }, { "epoch": 0.17593181529067548, "ewc_loss": 0.03564453125, "ewc_loss_diag": 2.4080276489257812e-05, "ewc_loss_parallel": 3.314018249511719e-05, "grad_norm": 3.346064567565918, "learning_rate": 5.8584145824501915e-06, "loss": 0.5236, "mean_token_accuracy": 0.8412986993789673, "num_tokens": 52503117.0, "step": 1383 }, { "epoch": 0.176059025569266, "ewc_loss": 0.037109375, "ewc_loss_diag": 2.4080276489257812e-05, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 3.552417516708374, "learning_rate": 5.86265366680797e-06, "loss": 0.5469, "mean_token_accuracy": 0.8336547017097473, "num_tokens": 52542643.0, "step": 1384 }, { "epoch": 0.1761862358478565, "ewc_loss": 0.035400390625, "ewc_loss_diag": 2.4080276489257812e-05, "ewc_loss_parallel": 3.2901763916015625e-05, "grad_norm": 3.3320846557617188, "learning_rate": 5.866892751165749e-06, "loss": 0.5226, "mean_token_accuracy": 0.8419227600097656, "num_tokens": 52585885.0, "step": 1385 }, { "epoch": 0.176313446126447, "ewc_loss": 0.03662109375, "ewc_loss_diag": 2.4080276489257812e-05, "ewc_loss_parallel": 3.409385681152344e-05, "grad_norm": 3.497868061065674, "learning_rate": 5.871131835523527e-06, "loss": 0.571, "mean_token_accuracy": 0.8277509212493896, "num_tokens": 52621629.0, "step": 1386 }, { "epoch": 0.17644065640503753, "ewc_loss": 0.03515625, "ewc_loss_diag": 2.4199485778808594e-05, "ewc_loss_parallel": 3.266334533691406e-05, "grad_norm": 3.3047008514404297, "learning_rate": 5.875370919881306e-06, "loss": 0.5245, "mean_token_accuracy": 0.8375734686851501, "num_tokens": 52667056.0, "step": 1387 }, { "epoch": 0.17656786668362803, "ewc_loss": 0.036376953125, "ewc_loss_diag": 2.4199485778808594e-05, "ewc_loss_parallel": 3.3855438232421875e-05, "grad_norm": 3.4495787620544434, "learning_rate": 5.879610004239085e-06, "loss": 0.5812, "mean_token_accuracy": 0.8282576203346252, "num_tokens": 52705857.0, "step": 1388 }, { "epoch": 0.17669507696221853, "ewc_loss": 0.034912109375, "ewc_loss_diag": 2.4199485778808594e-05, "ewc_loss_parallel": 3.24249267578125e-05, "grad_norm": 3.28533935546875, "learning_rate": 5.883849088596864e-06, "loss": 0.5645, "mean_token_accuracy": 0.8290217518806458, "num_tokens": 52748908.0, "step": 1389 }, { "epoch": 0.17682228724080906, "ewc_loss": 0.03564453125, "ewc_loss_diag": 2.4199485778808594e-05, "ewc_loss_parallel": 3.314018249511719e-05, "grad_norm": 3.431365966796875, "learning_rate": 5.888088172954642e-06, "loss": 0.5396, "mean_token_accuracy": 0.8389602303504944, "num_tokens": 52783600.0, "step": 1390 }, { "epoch": 0.17694949751939956, "ewc_loss": 0.0341796875, "ewc_loss_diag": 2.4199485778808594e-05, "ewc_loss_parallel": 3.170967102050781e-05, "grad_norm": 3.2534339427948, "learning_rate": 5.892327257312421e-06, "loss": 0.4614, "mean_token_accuracy": 0.8606438636779785, "num_tokens": 52819873.0, "step": 1391 }, { "epoch": 0.1770767077979901, "ewc_loss": 0.03564453125, "ewc_loss_diag": 2.4199485778808594e-05, "ewc_loss_parallel": 3.314018249511719e-05, "grad_norm": 3.4191737174987793, "learning_rate": 5.8965663416702e-06, "loss": 0.4654, "mean_token_accuracy": 0.8538782000541687, "num_tokens": 52858883.0, "step": 1392 }, { "epoch": 0.1772039180765806, "ewc_loss": 0.0341796875, "ewc_loss_diag": 2.4199485778808594e-05, "ewc_loss_parallel": 3.170967102050781e-05, "grad_norm": 3.242649793624878, "learning_rate": 5.900805426027978e-06, "loss": 0.5162, "mean_token_accuracy": 0.8444867730140686, "num_tokens": 52895501.0, "step": 1393 }, { "epoch": 0.1773311283551711, "ewc_loss": 0.03564453125, "ewc_loss_diag": 2.4199485778808594e-05, "ewc_loss_parallel": 3.314018249511719e-05, "grad_norm": 3.4200544357299805, "learning_rate": 5.905044510385757e-06, "loss": 0.4648, "mean_token_accuracy": 0.8584463000297546, "num_tokens": 52934678.0, "step": 1394 }, { "epoch": 0.17745833863376162, "ewc_loss": 0.033935546875, "ewc_loss_diag": 2.4199485778808594e-05, "ewc_loss_parallel": 3.147125244140625e-05, "grad_norm": 3.2214035987854004, "learning_rate": 5.909283594743536e-06, "loss": 0.5693, "mean_token_accuracy": 0.824226438999176, "num_tokens": 52972540.0, "step": 1395 }, { "epoch": 0.17758554891235212, "ewc_loss": 0.035888671875, "ewc_loss_diag": 2.4199485778808594e-05, "ewc_loss_parallel": 3.337860107421875e-05, "grad_norm": 3.4450626373291016, "learning_rate": 5.9135226791013145e-06, "loss": 0.5076, "mean_token_accuracy": 0.8420060276985168, "num_tokens": 53010819.0, "step": 1396 }, { "epoch": 0.17771275919094262, "ewc_loss": 0.033935546875, "ewc_loss_diag": 2.4199485778808594e-05, "ewc_loss_parallel": 3.147125244140625e-05, "grad_norm": 3.196295738220215, "learning_rate": 5.917761763459093e-06, "loss": 0.5685, "mean_token_accuracy": 0.8267360329627991, "num_tokens": 53050142.0, "step": 1397 }, { "epoch": 0.17783996946953315, "ewc_loss": 0.036376953125, "ewc_loss_diag": 2.4437904357910156e-05, "ewc_loss_parallel": 3.3855438232421875e-05, "grad_norm": 3.4868738651275635, "learning_rate": 5.922000847816872e-06, "loss": 0.5, "mean_token_accuracy": 0.8505642414093018, "num_tokens": 53087306.0, "step": 1398 }, { "epoch": 0.17796717974812365, "ewc_loss": 0.03369140625, "ewc_loss_diag": 2.4437904357910156e-05, "ewc_loss_parallel": 3.123283386230469e-05, "grad_norm": 3.1735615730285645, "learning_rate": 5.926239932174651e-06, "loss": 0.5549, "mean_token_accuracy": 0.8318325281143188, "num_tokens": 53131850.0, "step": 1399 }, { "epoch": 0.17809439002671415, "ewc_loss": 0.037109375, "ewc_loss_diag": 2.4437904357910156e-05, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 3.544647216796875, "learning_rate": 5.93047901653243e-06, "loss": 0.513, "mean_token_accuracy": 0.8422799110412598, "num_tokens": 53168902.0, "step": 1400 }, { "epoch": 0.17822160030530468, "ewc_loss": 0.03369140625, "ewc_loss_diag": 2.4437904357910156e-05, "ewc_loss_parallel": 3.123283386230469e-05, "grad_norm": 3.1435258388519287, "learning_rate": 5.934718100890208e-06, "loss": 0.4864, "mean_token_accuracy": 0.8502122163772583, "num_tokens": 53208497.0, "step": 1401 }, { "epoch": 0.17834881058389518, "ewc_loss": 0.037841796875, "ewc_loss_diag": 2.4437904357910156e-05, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 3.6338562965393066, "learning_rate": 5.938957185247987e-06, "loss": 0.5447, "mean_token_accuracy": 0.833141028881073, "num_tokens": 53245618.0, "step": 1402 }, { "epoch": 0.17847602086248568, "ewc_loss": 0.03369140625, "ewc_loss_diag": 2.4437904357910156e-05, "ewc_loss_parallel": 3.123283386230469e-05, "grad_norm": 3.1482949256896973, "learning_rate": 5.943196269605766e-06, "loss": 0.5015, "mean_token_accuracy": 0.8443686962127686, "num_tokens": 53280315.0, "step": 1403 }, { "epoch": 0.1786032311410762, "ewc_loss": 0.038818359375, "ewc_loss_diag": 2.4437904357910156e-05, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 3.7266955375671387, "learning_rate": 5.947435353963545e-06, "loss": 0.5457, "mean_token_accuracy": 0.8381685018539429, "num_tokens": 53313291.0, "step": 1404 }, { "epoch": 0.1787304414196667, "ewc_loss": 0.03466796875, "ewc_loss_diag": 2.4437904357910156e-05, "ewc_loss_parallel": 3.218650817871094e-05, "grad_norm": 3.1652541160583496, "learning_rate": 5.951674438321324e-06, "loss": 0.5285, "mean_token_accuracy": 0.8405798077583313, "num_tokens": 53347764.0, "step": 1405 }, { "epoch": 0.1788576516982572, "ewc_loss": 0.039306640625, "ewc_loss_diag": 2.4437904357910156e-05, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 3.795865058898926, "learning_rate": 5.9559135226791025e-06, "loss": 0.5325, "mean_token_accuracy": 0.8358036875724792, "num_tokens": 53385184.0, "step": 1406 }, { "epoch": 0.17898486197684774, "ewc_loss": 0.035400390625, "ewc_loss_diag": 2.4437904357910156e-05, "ewc_loss_parallel": 3.2901763916015625e-05, "grad_norm": 3.2287416458129883, "learning_rate": 5.9601526070368805e-06, "loss": 0.603, "mean_token_accuracy": 0.8153973817825317, "num_tokens": 53425850.0, "step": 1407 }, { "epoch": 0.17911207225543824, "ewc_loss": 0.0400390625, "ewc_loss_diag": 2.4437904357910156e-05, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 3.812217950820923, "learning_rate": 5.964391691394659e-06, "loss": 0.541, "mean_token_accuracy": 0.8391962051391602, "num_tokens": 53472444.0, "step": 1408 }, { "epoch": 0.17923928253402874, "ewc_loss": 0.035888671875, "ewc_loss_diag": 2.4437904357910156e-05, "ewc_loss_parallel": 3.337860107421875e-05, "grad_norm": 3.3048298358917236, "learning_rate": 5.968630775752437e-06, "loss": 0.5755, "mean_token_accuracy": 0.8273688554763794, "num_tokens": 53506996.0, "step": 1409 }, { "epoch": 0.17936649281261927, "ewc_loss": 0.039794921875, "ewc_loss_diag": 2.4437904357910156e-05, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 3.796684503555298, "learning_rate": 5.972869860110216e-06, "loss": 0.5533, "mean_token_accuracy": 0.8309798240661621, "num_tokens": 53542005.0, "step": 1410 }, { "epoch": 0.17949370309120977, "ewc_loss": 0.036376953125, "ewc_loss_diag": 2.4437904357910156e-05, "ewc_loss_parallel": 3.3855438232421875e-05, "grad_norm": 3.3563358783721924, "learning_rate": 5.977108944467995e-06, "loss": 0.5865, "mean_token_accuracy": 0.8286799192428589, "num_tokens": 53582130.0, "step": 1411 }, { "epoch": 0.17962091336980027, "ewc_loss": 0.0390625, "ewc_loss_diag": 2.4557113647460938e-05, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 3.716501474380493, "learning_rate": 5.981348028825774e-06, "loss": 0.507, "mean_token_accuracy": 0.8510512709617615, "num_tokens": 53624528.0, "step": 1412 }, { "epoch": 0.1797481236483908, "ewc_loss": 0.036376953125, "ewc_loss_diag": 2.4557113647460938e-05, "ewc_loss_parallel": 3.3855438232421875e-05, "grad_norm": 3.386289119720459, "learning_rate": 5.985587113183553e-06, "loss": 0.553, "mean_token_accuracy": 0.8336474895477295, "num_tokens": 53663701.0, "step": 1413 }, { "epoch": 0.1798753339269813, "ewc_loss": 0.0380859375, "ewc_loss_diag": 2.467632293701172e-05, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 3.6355183124542236, "learning_rate": 5.989826197541332e-06, "loss": 0.5429, "mean_token_accuracy": 0.8392610549926758, "num_tokens": 53700655.0, "step": 1414 }, { "epoch": 0.18000254420557182, "ewc_loss": 0.0361328125, "ewc_loss_diag": 2.467632293701172e-05, "ewc_loss_parallel": 3.361701965332031e-05, "grad_norm": 3.3618111610412598, "learning_rate": 5.994065281899111e-06, "loss": 0.5628, "mean_token_accuracy": 0.8334740400314331, "num_tokens": 53747387.0, "step": 1415 }, { "epoch": 0.18012975448416232, "ewc_loss": 0.037353515625, "ewc_loss_diag": 2.467632293701172e-05, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 3.5670037269592285, "learning_rate": 5.998304366256889e-06, "loss": 0.5313, "mean_token_accuracy": 0.8389829397201538, "num_tokens": 53789416.0, "step": 1416 }, { "epoch": 0.18025696476275282, "ewc_loss": 0.0361328125, "ewc_loss_diag": 2.467632293701172e-05, "ewc_loss_parallel": 3.361701965332031e-05, "grad_norm": 3.3527493476867676, "learning_rate": 6.002543450614668e-06, "loss": 0.4543, "mean_token_accuracy": 0.8656954765319824, "num_tokens": 53824300.0, "step": 1417 }, { "epoch": 0.18038417504134335, "ewc_loss": 0.037109375, "ewc_loss_diag": 2.467632293701172e-05, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 3.5317680835723877, "learning_rate": 6.0067825349724465e-06, "loss": 0.5304, "mean_token_accuracy": 0.8405120968818665, "num_tokens": 53861366.0, "step": 1418 }, { "epoch": 0.18051138531993385, "ewc_loss": 0.035888671875, "ewc_loss_diag": 2.467632293701172e-05, "ewc_loss_parallel": 3.337860107421875e-05, "grad_norm": 3.3491201400756836, "learning_rate": 6.011021619330225e-06, "loss": 0.5533, "mean_token_accuracy": 0.834406852722168, "num_tokens": 53899655.0, "step": 1419 }, { "epoch": 0.18063859559852435, "ewc_loss": 0.03662109375, "ewc_loss_diag": 2.467632293701172e-05, "ewc_loss_parallel": 3.409385681152344e-05, "grad_norm": 3.494868516921997, "learning_rate": 6.015260703688004e-06, "loss": 0.532, "mean_token_accuracy": 0.8352360129356384, "num_tokens": 53935563.0, "step": 1420 }, { "epoch": 0.18076580587711488, "ewc_loss": 0.035400390625, "ewc_loss_diag": 2.467632293701172e-05, "ewc_loss_parallel": 3.2901763916015625e-05, "grad_norm": 3.32513689994812, "learning_rate": 6.019499788045783e-06, "loss": 0.5294, "mean_token_accuracy": 0.8403920531272888, "num_tokens": 53970490.0, "step": 1421 }, { "epoch": 0.18089301615570538, "ewc_loss": 0.03662109375, "ewc_loss_diag": 2.491474151611328e-05, "ewc_loss_parallel": 3.409385681152344e-05, "grad_norm": 3.4752187728881836, "learning_rate": 6.023738872403562e-06, "loss": 0.4953, "mean_token_accuracy": 0.8482519388198853, "num_tokens": 54000460.0, "step": 1422 }, { "epoch": 0.18102022643429588, "ewc_loss": 0.035400390625, "ewc_loss_diag": 2.491474151611328e-05, "ewc_loss_parallel": 3.2901763916015625e-05, "grad_norm": 3.296185255050659, "learning_rate": 6.027977956761341e-06, "loss": 0.5216, "mean_token_accuracy": 0.8463295698165894, "num_tokens": 54038007.0, "step": 1423 }, { "epoch": 0.1811474367128864, "ewc_loss": 0.03662109375, "ewc_loss_diag": 2.491474151611328e-05, "ewc_loss_parallel": 3.409385681152344e-05, "grad_norm": 3.4968574047088623, "learning_rate": 6.032217041119118e-06, "loss": 0.5326, "mean_token_accuracy": 0.8348832130432129, "num_tokens": 54074515.0, "step": 1424 }, { "epoch": 0.1812746469914769, "ewc_loss": 0.03515625, "ewc_loss_diag": 2.491474151611328e-05, "ewc_loss_parallel": 3.266334533691406e-05, "grad_norm": 3.269315481185913, "learning_rate": 6.036456125476897e-06, "loss": 0.567, "mean_token_accuracy": 0.8282574415206909, "num_tokens": 54109355.0, "step": 1425 }, { "epoch": 0.1814018572700674, "ewc_loss": 0.036865234375, "ewc_loss_diag": 2.491474151611328e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 3.4933090209960938, "learning_rate": 6.040695209834676e-06, "loss": 0.469, "mean_token_accuracy": 0.8548312187194824, "num_tokens": 54150966.0, "step": 1426 }, { "epoch": 0.18152906754865794, "ewc_loss": 0.03515625, "ewc_loss_diag": 2.5153160095214844e-05, "ewc_loss_parallel": 3.266334533691406e-05, "grad_norm": 3.249040365219116, "learning_rate": 6.044934294192455e-06, "loss": 0.5612, "mean_token_accuracy": 0.8295477628707886, "num_tokens": 54188398.0, "step": 1427 }, { "epoch": 0.18165627782724844, "ewc_loss": 0.037353515625, "ewc_loss_diag": 2.5153160095214844e-05, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 3.5255603790283203, "learning_rate": 6.049173378550234e-06, "loss": 0.5186, "mean_token_accuracy": 0.8426548838615417, "num_tokens": 54233669.0, "step": 1428 }, { "epoch": 0.18178348810583894, "ewc_loss": 0.035400390625, "ewc_loss_diag": 2.5153160095214844e-05, "ewc_loss_parallel": 3.2901763916015625e-05, "grad_norm": 3.223639726638794, "learning_rate": 6.0534124629080126e-06, "loss": 0.4991, "mean_token_accuracy": 0.8485691547393799, "num_tokens": 54272377.0, "step": 1429 }, { "epoch": 0.18191069838442947, "ewc_loss": 0.03759765625, "ewc_loss_diag": 2.5153160095214844e-05, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 3.5878663063049316, "learning_rate": 6.0576515472657915e-06, "loss": 0.5715, "mean_token_accuracy": 0.8293823003768921, "num_tokens": 54313831.0, "step": 1430 }, { "epoch": 0.18203790866301997, "ewc_loss": 0.034912109375, "ewc_loss_diag": 2.5153160095214844e-05, "ewc_loss_parallel": 3.24249267578125e-05, "grad_norm": 3.182668447494507, "learning_rate": 6.0618906316235695e-06, "loss": 0.4737, "mean_token_accuracy": 0.8533530235290527, "num_tokens": 54356724.0, "step": 1431 }, { "epoch": 0.18216511894161047, "ewc_loss": 0.03857421875, "ewc_loss_diag": 2.5153160095214844e-05, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 3.6699953079223633, "learning_rate": 6.066129715981348e-06, "loss": 0.5113, "mean_token_accuracy": 0.8442884683609009, "num_tokens": 54398407.0, "step": 1432 }, { "epoch": 0.182292329220201, "ewc_loss": 0.034912109375, "ewc_loss_diag": 2.5153160095214844e-05, "ewc_loss_parallel": 3.24249267578125e-05, "grad_norm": 3.185297966003418, "learning_rate": 6.070368800339127e-06, "loss": 0.4669, "mean_token_accuracy": 0.8591289520263672, "num_tokens": 54430023.0, "step": 1433 }, { "epoch": 0.1824195394987915, "ewc_loss": 0.039794921875, "ewc_loss_diag": 2.5272369384765625e-05, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 3.768873453140259, "learning_rate": 6.074607884696906e-06, "loss": 0.5024, "mean_token_accuracy": 0.8463561534881592, "num_tokens": 54467154.0, "step": 1434 }, { "epoch": 0.182546749777382, "ewc_loss": 0.035400390625, "ewc_loss_diag": 2.5272369384765625e-05, "ewc_loss_parallel": 3.2901763916015625e-05, "grad_norm": 3.2090492248535156, "learning_rate": 6.078846969054685e-06, "loss": 0.528, "mean_token_accuracy": 0.8403708934783936, "num_tokens": 54501139.0, "step": 1435 }, { "epoch": 0.18267396005597253, "ewc_loss": 0.040283203125, "ewc_loss_diag": 2.5272369384765625e-05, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 3.8512914180755615, "learning_rate": 6.083086053412464e-06, "loss": 0.5125, "mean_token_accuracy": 0.8458825945854187, "num_tokens": 54541084.0, "step": 1436 }, { "epoch": 0.18280117033456303, "ewc_loss": 0.035888671875, "ewc_loss_diag": 2.5272369384765625e-05, "ewc_loss_parallel": 3.337860107421875e-05, "grad_norm": 3.2515716552734375, "learning_rate": 6.087325137770243e-06, "loss": 0.4895, "mean_token_accuracy": 0.8504608869552612, "num_tokens": 54576745.0, "step": 1437 }, { "epoch": 0.18292838061315353, "ewc_loss": 0.041015625, "ewc_loss_diag": 2.5272369384765625e-05, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 3.901095390319824, "learning_rate": 6.091564222128022e-06, "loss": 0.5737, "mean_token_accuracy": 0.8317604064941406, "num_tokens": 54619039.0, "step": 1438 }, { "epoch": 0.18305559089174406, "ewc_loss": 0.036376953125, "ewc_loss_diag": 2.5272369384765625e-05, "ewc_loss_parallel": 3.3855438232421875e-05, "grad_norm": 3.331979274749756, "learning_rate": 6.095803306485799e-06, "loss": 0.5096, "mean_token_accuracy": 0.8437105417251587, "num_tokens": 54660009.0, "step": 1439 }, { "epoch": 0.18318280117033456, "ewc_loss": 0.041015625, "ewc_loss_diag": 2.5272369384765625e-05, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 3.8825299739837646, "learning_rate": 6.100042390843578e-06, "loss": 0.5495, "mean_token_accuracy": 0.8363721966743469, "num_tokens": 54701253.0, "step": 1440 }, { "epoch": 0.1833100114489251, "ewc_loss": 0.036865234375, "ewc_loss_diag": 2.5272369384765625e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 3.4146108627319336, "learning_rate": 6.104281475201357e-06, "loss": 0.4987, "mean_token_accuracy": 0.8486693501472473, "num_tokens": 54738406.0, "step": 1441 }, { "epoch": 0.1834372217275156, "ewc_loss": 0.04052734375, "ewc_loss_diag": 2.5272369384765625e-05, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 3.828378915786743, "learning_rate": 6.1085205595591355e-06, "loss": 0.5556, "mean_token_accuracy": 0.8331031203269958, "num_tokens": 54771808.0, "step": 1442 }, { "epoch": 0.1835644320061061, "ewc_loss": 0.036865234375, "ewc_loss_diag": 2.5272369384765625e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 3.4447243213653564, "learning_rate": 6.112759643916914e-06, "loss": 0.4509, "mean_token_accuracy": 0.86306232213974, "num_tokens": 54807337.0, "step": 1443 }, { "epoch": 0.18369164228469662, "ewc_loss": 0.039794921875, "ewc_loss_diag": 2.5272369384765625e-05, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 3.729051113128662, "learning_rate": 6.116998728274693e-06, "loss": 0.5404, "mean_token_accuracy": 0.8395148515701294, "num_tokens": 54844932.0, "step": 1444 }, { "epoch": 0.18381885256328712, "ewc_loss": 0.037109375, "ewc_loss_diag": 2.5272369384765625e-05, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 3.4421284198760986, "learning_rate": 6.121237812632472e-06, "loss": 0.5402, "mean_token_accuracy": 0.8339791297912598, "num_tokens": 54884492.0, "step": 1445 }, { "epoch": 0.18394606284187762, "ewc_loss": 0.0380859375, "ewc_loss_diag": 2.5391578674316406e-05, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 3.6427879333496094, "learning_rate": 6.12547689699025e-06, "loss": 0.4906, "mean_token_accuracy": 0.854810893535614, "num_tokens": 54924164.0, "step": 1446 }, { "epoch": 0.18407327312046814, "ewc_loss": 0.03662109375, "ewc_loss_diag": 2.5391578674316406e-05, "ewc_loss_parallel": 3.409385681152344e-05, "grad_norm": 3.4213967323303223, "learning_rate": 6.129715981348029e-06, "loss": 0.5224, "mean_token_accuracy": 0.845291018486023, "num_tokens": 54961676.0, "step": 1447 }, { "epoch": 0.18420048339905865, "ewc_loss": 0.037841796875, "ewc_loss_diag": 2.5391578674316406e-05, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 3.592594623565674, "learning_rate": 6.133955065705808e-06, "loss": 0.5487, "mean_token_accuracy": 0.832016110420227, "num_tokens": 55000889.0, "step": 1448 }, { "epoch": 0.18432769367764915, "ewc_loss": 0.03662109375, "ewc_loss_diag": 2.5391578674316406e-05, "ewc_loss_parallel": 3.409385681152344e-05, "grad_norm": 3.3955414295196533, "learning_rate": 6.138194150063587e-06, "loss": 0.5115, "mean_token_accuracy": 0.8453618288040161, "num_tokens": 55039471.0, "step": 1449 }, { "epoch": 0.18445490395623967, "ewc_loss": 0.03759765625, "ewc_loss_diag": 2.5391578674316406e-05, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 3.556793689727783, "learning_rate": 6.142433234421366e-06, "loss": 0.5469, "mean_token_accuracy": 0.8329925537109375, "num_tokens": 55078408.0, "step": 1450 }, { "epoch": 0.18458211423483017, "ewc_loss": 0.035888671875, "ewc_loss_diag": 2.5391578674316406e-05, "ewc_loss_parallel": 3.337860107421875e-05, "grad_norm": 3.3810644149780273, "learning_rate": 6.146672318779145e-06, "loss": 0.5468, "mean_token_accuracy": 0.8346444368362427, "num_tokens": 55116068.0, "step": 1451 }, { "epoch": 0.18470932451342068, "ewc_loss": 0.03759765625, "ewc_loss_diag": 2.5391578674316406e-05, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 3.548649549484253, "learning_rate": 6.1509114031369235e-06, "loss": 0.5359, "mean_token_accuracy": 0.8410172462463379, "num_tokens": 55149933.0, "step": 1452 }, { "epoch": 0.1848365347920112, "ewc_loss": 0.035888671875, "ewc_loss_diag": 2.5510787963867188e-05, "ewc_loss_parallel": 3.337860107421875e-05, "grad_norm": 3.3498754501342773, "learning_rate": 6.155150487494702e-06, "loss": 0.4626, "mean_token_accuracy": 0.8591694235801697, "num_tokens": 55184674.0, "step": 1453 }, { "epoch": 0.1849637450706017, "ewc_loss": 0.037353515625, "ewc_loss_diag": 2.5510787963867188e-05, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 3.527393102645874, "learning_rate": 6.15938957185248e-06, "loss": 0.4974, "mean_token_accuracy": 0.8495351076126099, "num_tokens": 55227918.0, "step": 1454 }, { "epoch": 0.1850909553491922, "ewc_loss": 0.03564453125, "ewc_loss_diag": 2.562999725341797e-05, "ewc_loss_parallel": 3.314018249511719e-05, "grad_norm": 3.3164803981781006, "learning_rate": 6.1636286562102585e-06, "loss": 0.5065, "mean_token_accuracy": 0.8464570045471191, "num_tokens": 55274312.0, "step": 1455 }, { "epoch": 0.18521816562778273, "ewc_loss": 0.03759765625, "ewc_loss_diag": 2.562999725341797e-05, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 3.5696167945861816, "learning_rate": 6.167867740568037e-06, "loss": 0.576, "mean_token_accuracy": 0.8259463310241699, "num_tokens": 55310583.0, "step": 1456 }, { "epoch": 0.18534537590637323, "ewc_loss": 0.03564453125, "ewc_loss_diag": 2.562999725341797e-05, "ewc_loss_parallel": 3.314018249511719e-05, "grad_norm": 3.281022787094116, "learning_rate": 6.172106824925816e-06, "loss": 0.5021, "mean_token_accuracy": 0.8452200293540955, "num_tokens": 55352365.0, "step": 1457 }, { "epoch": 0.18547258618496373, "ewc_loss": 0.0380859375, "ewc_loss_diag": 2.562999725341797e-05, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 3.622440814971924, "learning_rate": 6.176345909283595e-06, "loss": 0.5197, "mean_token_accuracy": 0.8419450521469116, "num_tokens": 55391268.0, "step": 1458 }, { "epoch": 0.18559979646355426, "ewc_loss": 0.03564453125, "ewc_loss_diag": 2.562999725341797e-05, "ewc_loss_parallel": 3.314018249511719e-05, "grad_norm": 3.2614309787750244, "learning_rate": 6.180584993641374e-06, "loss": 0.4966, "mean_token_accuracy": 0.8483394384384155, "num_tokens": 55428842.0, "step": 1459 }, { "epoch": 0.18572700674214476, "ewc_loss": 0.0390625, "ewc_loss_diag": 2.562999725341797e-05, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 3.691534996032715, "learning_rate": 6.184824077999153e-06, "loss": 0.4796, "mean_token_accuracy": 0.8542401790618896, "num_tokens": 55464341.0, "step": 1460 }, { "epoch": 0.18585421702073526, "ewc_loss": 0.03564453125, "ewc_loss_diag": 2.562999725341797e-05, "ewc_loss_parallel": 3.2901763916015625e-05, "grad_norm": 3.2444422245025635, "learning_rate": 6.189063162356932e-06, "loss": 0.5727, "mean_token_accuracy": 0.8255040645599365, "num_tokens": 55498985.0, "step": 1461 }, { "epoch": 0.1859814272993258, "ewc_loss": 0.03955078125, "ewc_loss_diag": 2.562999725341797e-05, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 3.76554012298584, "learning_rate": 6.19330224671471e-06, "loss": 0.5044, "mean_token_accuracy": 0.8465113639831543, "num_tokens": 55537762.0, "step": 1462 }, { "epoch": 0.1861086375779163, "ewc_loss": 0.03564453125, "ewc_loss_diag": 2.562999725341797e-05, "ewc_loss_parallel": 3.314018249511719e-05, "grad_norm": 3.2420291900634766, "learning_rate": 6.197541331072489e-06, "loss": 0.5454, "mean_token_accuracy": 0.8336121439933777, "num_tokens": 55572571.0, "step": 1463 }, { "epoch": 0.1862358478565068, "ewc_loss": 0.040771484375, "ewc_loss_diag": 2.574920654296875e-05, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 3.850045680999756, "learning_rate": 6.201780415430268e-06, "loss": 0.5141, "mean_token_accuracy": 0.8476456999778748, "num_tokens": 55611858.0, "step": 1464 }, { "epoch": 0.18636305813509732, "ewc_loss": 0.036376953125, "ewc_loss_diag": 2.574920654296875e-05, "ewc_loss_parallel": 3.361701965332031e-05, "grad_norm": 3.26363468170166, "learning_rate": 6.2060194997880465e-06, "loss": 0.5502, "mean_token_accuracy": 0.8334070444107056, "num_tokens": 55649168.0, "step": 1465 }, { "epoch": 0.18649026841368782, "ewc_loss": 0.041259765625, "ewc_loss_diag": 2.574920654296875e-05, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 3.927828550338745, "learning_rate": 6.210258584145825e-06, "loss": 0.4935, "mean_token_accuracy": 0.8488760590553284, "num_tokens": 55686127.0, "step": 1466 }, { "epoch": 0.18661747869227835, "ewc_loss": 0.03662109375, "ewc_loss_diag": 2.586841583251953e-05, "ewc_loss_parallel": 3.3855438232421875e-05, "grad_norm": 3.332475423812866, "learning_rate": 6.214497668503604e-06, "loss": 0.5474, "mean_token_accuracy": 0.8317978978157043, "num_tokens": 55724062.0, "step": 1467 }, { "epoch": 0.18674468897086885, "ewc_loss": 0.041748046875, "ewc_loss_diag": 2.586841583251953e-05, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 3.973633050918579, "learning_rate": 6.218736752861383e-06, "loss": 0.5099, "mean_token_accuracy": 0.847758412361145, "num_tokens": 55752137.0, "step": 1468 }, { "epoch": 0.18687189924945935, "ewc_loss": 0.037353515625, "ewc_loss_diag": 2.6106834411621094e-05, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 3.417894124984741, "learning_rate": 6.22297583721916e-06, "loss": 0.5421, "mean_token_accuracy": 0.8434809446334839, "num_tokens": 55793171.0, "step": 1469 }, { "epoch": 0.18699910952804988, "ewc_loss": 0.041748046875, "ewc_loss_diag": 2.6106834411621094e-05, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 3.931034564971924, "learning_rate": 6.227214921576939e-06, "loss": 0.5745, "mean_token_accuracy": 0.8334368467330933, "num_tokens": 55831118.0, "step": 1470 }, { "epoch": 0.18712631980664038, "ewc_loss": 0.03759765625, "ewc_loss_diag": 2.6106834411621094e-05, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 3.473804473876953, "learning_rate": 6.231454005934718e-06, "loss": 0.4663, "mean_token_accuracy": 0.8595949411392212, "num_tokens": 55865568.0, "step": 1471 }, { "epoch": 0.18725353008523088, "ewc_loss": 0.041015625, "ewc_loss_diag": 2.6106834411621094e-05, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 3.849844455718994, "learning_rate": 6.235693090292497e-06, "loss": 0.5874, "mean_token_accuracy": 0.8290456533432007, "num_tokens": 55904624.0, "step": 1472 }, { "epoch": 0.1873807403638214, "ewc_loss": 0.03759765625, "ewc_loss_diag": 2.6106834411621094e-05, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 3.479975700378418, "learning_rate": 6.239932174650276e-06, "loss": 0.4637, "mean_token_accuracy": 0.8605731725692749, "num_tokens": 55944714.0, "step": 1473 }, { "epoch": 0.1875079506424119, "ewc_loss": 0.04052734375, "ewc_loss_diag": 2.6106834411621094e-05, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 3.7740249633789062, "learning_rate": 6.244171259008055e-06, "loss": 0.5521, "mean_token_accuracy": 0.8313587307929993, "num_tokens": 55979444.0, "step": 1474 }, { "epoch": 0.1876351609210024, "ewc_loss": 0.037353515625, "ewc_loss_diag": 2.6345252990722656e-05, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 3.4804530143737793, "learning_rate": 6.248410343365834e-06, "loss": 0.4851, "mean_token_accuracy": 0.8528691530227661, "num_tokens": 56017171.0, "step": 1475 }, { "epoch": 0.18776237119959294, "ewc_loss": 0.0390625, "ewc_loss_diag": 2.6345252990722656e-05, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 3.687389612197876, "learning_rate": 6.2526494277236125e-06, "loss": 0.5039, "mean_token_accuracy": 0.8499449491500854, "num_tokens": 56060562.0, "step": 1476 }, { "epoch": 0.18788958147818344, "ewc_loss": 0.037109375, "ewc_loss_diag": 2.6345252990722656e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 3.471482515335083, "learning_rate": 6.2568885120813906e-06, "loss": 0.4995, "mean_token_accuracy": 0.8475891351699829, "num_tokens": 56098366.0, "step": 1477 }, { "epoch": 0.18801679175677394, "ewc_loss": 0.03857421875, "ewc_loss_diag": 2.6345252990722656e-05, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 3.6261022090911865, "learning_rate": 6.2611275964391694e-06, "loss": 0.5616, "mean_token_accuracy": 0.8288314342498779, "num_tokens": 56143323.0, "step": 1478 }, { "epoch": 0.18814400203536447, "ewc_loss": 0.037109375, "ewc_loss_diag": 2.6345252990722656e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 3.441797971725464, "learning_rate": 6.265366680796948e-06, "loss": 0.4569, "mean_token_accuracy": 0.8598693609237671, "num_tokens": 56183846.0, "step": 1479 }, { "epoch": 0.18827121231395497, "ewc_loss": 0.038330078125, "ewc_loss_diag": 2.6345252990722656e-05, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 3.590806007385254, "learning_rate": 6.269605765154727e-06, "loss": 0.5186, "mean_token_accuracy": 0.8428410291671753, "num_tokens": 56227412.0, "step": 1480 }, { "epoch": 0.18839842259254547, "ewc_loss": 0.037109375, "ewc_loss_diag": 2.6345252990722656e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 3.4293811321258545, "learning_rate": 6.273844849512506e-06, "loss": 0.5139, "mean_token_accuracy": 0.8438147306442261, "num_tokens": 56267649.0, "step": 1481 }, { "epoch": 0.188525632871136, "ewc_loss": 0.038330078125, "ewc_loss_diag": 2.6345252990722656e-05, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 3.5742688179016113, "learning_rate": 6.278083933870285e-06, "loss": 0.444, "mean_token_accuracy": 0.8656752705574036, "num_tokens": 56310802.0, "step": 1482 }, { "epoch": 0.1886528431497265, "ewc_loss": 0.037109375, "ewc_loss_diag": 2.6345252990722656e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 3.4088220596313477, "learning_rate": 6.282323018228064e-06, "loss": 0.4835, "mean_token_accuracy": 0.8560498952865601, "num_tokens": 56345972.0, "step": 1483 }, { "epoch": 0.188780053428317, "ewc_loss": 0.0380859375, "ewc_loss_diag": 2.6464462280273438e-05, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 3.5776736736297607, "learning_rate": 6.286562102585841e-06, "loss": 0.5685, "mean_token_accuracy": 0.8322548866271973, "num_tokens": 56391778.0, "step": 1484 }, { "epoch": 0.18890726370690752, "ewc_loss": 0.036376953125, "ewc_loss_diag": 2.6464462280273438e-05, "ewc_loss_parallel": 3.361701965332031e-05, "grad_norm": 3.373350143432617, "learning_rate": 6.29080118694362e-06, "loss": 0.4491, "mean_token_accuracy": 0.8606536984443665, "num_tokens": 56426306.0, "step": 1485 }, { "epoch": 0.18903447398549802, "ewc_loss": 0.038330078125, "ewc_loss_diag": 2.6464462280273438e-05, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 3.606961965560913, "learning_rate": 6.295040271301399e-06, "loss": 0.4559, "mean_token_accuracy": 0.8603437542915344, "num_tokens": 56461974.0, "step": 1486 }, { "epoch": 0.18916168426408853, "ewc_loss": 0.036376953125, "ewc_loss_diag": 2.6464462280273438e-05, "ewc_loss_parallel": 3.361701965332031e-05, "grad_norm": 3.3350014686584473, "learning_rate": 6.299279355659178e-06, "loss": 0.5018, "mean_token_accuracy": 0.8500288128852844, "num_tokens": 56507205.0, "step": 1487 }, { "epoch": 0.18928889454267905, "ewc_loss": 0.039306640625, "ewc_loss_diag": 2.6464462280273438e-05, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 3.66086483001709, "learning_rate": 6.303518440016957e-06, "loss": 0.5157, "mean_token_accuracy": 0.8418276309967041, "num_tokens": 56542076.0, "step": 1488 }, { "epoch": 0.18941610482126955, "ewc_loss": 0.036376953125, "ewc_loss_diag": 2.6702880859375e-05, "ewc_loss_parallel": 3.361701965332031e-05, "grad_norm": 3.318821668624878, "learning_rate": 6.3077575243747355e-06, "loss": 0.5082, "mean_token_accuracy": 0.84950852394104, "num_tokens": 56575980.0, "step": 1489 }, { "epoch": 0.18954331509986008, "ewc_loss": 0.03955078125, "ewc_loss_diag": 2.6702880859375e-05, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 3.727351665496826, "learning_rate": 6.311996608732514e-06, "loss": 0.5431, "mean_token_accuracy": 0.8397444486618042, "num_tokens": 56613410.0, "step": 1490 }, { "epoch": 0.18967052537845058, "ewc_loss": 0.035888671875, "ewc_loss_diag": 2.6702880859375e-05, "ewc_loss_parallel": 3.314018249511719e-05, "grad_norm": 3.2776243686676025, "learning_rate": 6.316235693090293e-06, "loss": 0.5318, "mean_token_accuracy": 0.8375015258789062, "num_tokens": 56649105.0, "step": 1491 }, { "epoch": 0.18979773565704108, "ewc_loss": 0.04052734375, "ewc_loss_diag": 2.6702880859375e-05, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 3.823017120361328, "learning_rate": 6.320474777448071e-06, "loss": 0.5389, "mean_token_accuracy": 0.8361833095550537, "num_tokens": 56686049.0, "step": 1492 }, { "epoch": 0.1899249459356316, "ewc_loss": 0.035888671875, "ewc_loss_diag": 2.6702880859375e-05, "ewc_loss_parallel": 3.314018249511719e-05, "grad_norm": 3.261630058288574, "learning_rate": 6.32471386180585e-06, "loss": 0.4973, "mean_token_accuracy": 0.846761167049408, "num_tokens": 56725659.0, "step": 1493 }, { "epoch": 0.1900521562142221, "ewc_loss": 0.041748046875, "ewc_loss_diag": 2.682209014892578e-05, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 3.95297908782959, "learning_rate": 6.328952946163629e-06, "loss": 0.5946, "mean_token_accuracy": 0.8249473571777344, "num_tokens": 56760612.0, "step": 1494 }, { "epoch": 0.1901793664928126, "ewc_loss": 0.03662109375, "ewc_loss_diag": 2.682209014892578e-05, "ewc_loss_parallel": 3.3855438232421875e-05, "grad_norm": 3.288536787033081, "learning_rate": 6.333192030521408e-06, "loss": 0.464, "mean_token_accuracy": 0.857244610786438, "num_tokens": 56796081.0, "step": 1495 }, { "epoch": 0.19030657677140314, "ewc_loss": 0.04296875, "ewc_loss_diag": 2.6941299438476562e-05, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 4.049662113189697, "learning_rate": 6.337431114879187e-06, "loss": 0.5265, "mean_token_accuracy": 0.8430543541908264, "num_tokens": 56833477.0, "step": 1496 }, { "epoch": 0.19043378704999364, "ewc_loss": 0.037109375, "ewc_loss_diag": 2.6941299438476562e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 3.3920700550079346, "learning_rate": 6.341670199236966e-06, "loss": 0.4984, "mean_token_accuracy": 0.8518258929252625, "num_tokens": 56870898.0, "step": 1497 }, { "epoch": 0.19056099732858414, "ewc_loss": 0.0439453125, "ewc_loss_diag": 2.6941299438476562e-05, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 4.07037353515625, "learning_rate": 6.345909283594745e-06, "loss": 0.4746, "mean_token_accuracy": 0.857447624206543, "num_tokens": 56916980.0, "step": 1498 }, { "epoch": 0.19068820760717467, "ewc_loss": 0.03857421875, "ewc_loss_diag": 2.6941299438476562e-05, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 3.505356550216675, "learning_rate": 6.3501483679525235e-06, "loss": 0.5195, "mean_token_accuracy": 0.8381345272064209, "num_tokens": 56952564.0, "step": 1499 }, { "epoch": 0.19081541788576517, "ewc_loss": 0.042724609375, "ewc_loss_diag": 2.6941299438476562e-05, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 4.021286964416504, "learning_rate": 6.354387452310301e-06, "loss": 0.4859, "mean_token_accuracy": 0.8508585095405579, "num_tokens": 56987381.0, "step": 1500 }, { "epoch": 0.19094262816435567, "ewc_loss": 0.03857421875, "ewc_loss_diag": 2.6941299438476562e-05, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 3.576996326446533, "learning_rate": 6.3586265366680796e-06, "loss": 0.4866, "mean_token_accuracy": 0.8523855209350586, "num_tokens": 57023402.0, "step": 1501 }, { "epoch": 0.1910698384429462, "ewc_loss": 0.0419921875, "ewc_loss_diag": 2.6941299438476562e-05, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 3.9082565307617188, "learning_rate": 6.3628656210258584e-06, "loss": 0.4811, "mean_token_accuracy": 0.8568161725997925, "num_tokens": 57066464.0, "step": 1502 }, { "epoch": 0.1911970487215367, "ewc_loss": 0.038818359375, "ewc_loss_diag": 2.7060508728027344e-05, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 3.5745856761932373, "learning_rate": 6.367104705383637e-06, "loss": 0.495, "mean_token_accuracy": 0.8493155241012573, "num_tokens": 57104495.0, "step": 1503 }, { "epoch": 0.1913242590001272, "ewc_loss": 0.040283203125, "ewc_loss_diag": 2.7060508728027344e-05, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 3.800558090209961, "learning_rate": 6.371343789741416e-06, "loss": 0.5157, "mean_token_accuracy": 0.8439805507659912, "num_tokens": 57146246.0, "step": 1504 }, { "epoch": 0.19145146927871773, "ewc_loss": 0.038330078125, "ewc_loss_diag": 2.7298927307128906e-05, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 3.560431480407715, "learning_rate": 6.375582874099195e-06, "loss": 0.5053, "mean_token_accuracy": 0.8449342846870422, "num_tokens": 57184861.0, "step": 1505 }, { "epoch": 0.19157867955730823, "ewc_loss": 0.0390625, "ewc_loss_diag": 2.7298927307128906e-05, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 3.736598014831543, "learning_rate": 6.379821958456974e-06, "loss": 0.5619, "mean_token_accuracy": 0.8291951417922974, "num_tokens": 57219395.0, "step": 1506 }, { "epoch": 0.19170588983589873, "ewc_loss": 0.037841796875, "ewc_loss_diag": 2.7298927307128906e-05, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 3.53110671043396, "learning_rate": 6.384061042814752e-06, "loss": 0.5296, "mean_token_accuracy": 0.839638352394104, "num_tokens": 57255816.0, "step": 1507 }, { "epoch": 0.19183310011448926, "ewc_loss": 0.038818359375, "ewc_loss_diag": 2.7298927307128906e-05, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 3.6705219745635986, "learning_rate": 6.388300127172531e-06, "loss": 0.522, "mean_token_accuracy": 0.843629002571106, "num_tokens": 57296347.0, "step": 1508 }, { "epoch": 0.19196031039307976, "ewc_loss": 0.03759765625, "ewc_loss_diag": 2.7298927307128906e-05, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 3.4843883514404297, "learning_rate": 6.39253921153031e-06, "loss": 0.5216, "mean_token_accuracy": 0.8388411402702332, "num_tokens": 57341789.0, "step": 1509 }, { "epoch": 0.19208752067167026, "ewc_loss": 0.038330078125, "ewc_loss_diag": 2.7298927307128906e-05, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 3.650817632675171, "learning_rate": 6.396778295888089e-06, "loss": 0.4917, "mean_token_accuracy": 0.8543890714645386, "num_tokens": 57379399.0, "step": 1510 }, { "epoch": 0.1922147309502608, "ewc_loss": 0.03759765625, "ewc_loss_diag": 2.7298927307128906e-05, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 3.457655429840088, "learning_rate": 6.4010173802458676e-06, "loss": 0.5272, "mean_token_accuracy": 0.839954137802124, "num_tokens": 57418936.0, "step": 1511 }, { "epoch": 0.1923419412288513, "ewc_loss": 0.038818359375, "ewc_loss_diag": 2.7298927307128906e-05, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 3.6423842906951904, "learning_rate": 6.4052564646036464e-06, "loss": 0.5135, "mean_token_accuracy": 0.8432926535606384, "num_tokens": 57461772.0, "step": 1512 }, { "epoch": 0.1924691515074418, "ewc_loss": 0.037109375, "ewc_loss_diag": 2.7298927307128906e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 3.426668882369995, "learning_rate": 6.409495548961425e-06, "loss": 0.4891, "mean_token_accuracy": 0.8504315614700317, "num_tokens": 57498388.0, "step": 1513 }, { "epoch": 0.19259636178603232, "ewc_loss": 0.0390625, "ewc_loss_diag": 2.7298927307128906e-05, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 3.675955295562744, "learning_rate": 6.413734633319204e-06, "loss": 0.4996, "mean_token_accuracy": 0.8475111722946167, "num_tokens": 57534123.0, "step": 1514 }, { "epoch": 0.19272357206462282, "ewc_loss": 0.036865234375, "ewc_loss_diag": 2.7298927307128906e-05, "ewc_loss_parallel": 3.409385681152344e-05, "grad_norm": 3.3978874683380127, "learning_rate": 6.417973717676982e-06, "loss": 0.4734, "mean_token_accuracy": 0.8560289144515991, "num_tokens": 57568351.0, "step": 1515 }, { "epoch": 0.19285078234321335, "ewc_loss": 0.03955078125, "ewc_loss_diag": 2.7298927307128906e-05, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 3.709545135498047, "learning_rate": 6.422212802034761e-06, "loss": 0.5482, "mean_token_accuracy": 0.8337078094482422, "num_tokens": 57617842.0, "step": 1516 }, { "epoch": 0.19297799262180385, "ewc_loss": 0.03662109375, "ewc_loss_diag": 2.7298927307128906e-05, "ewc_loss_parallel": 3.3855438232421875e-05, "grad_norm": 3.3644425868988037, "learning_rate": 6.426451886392539e-06, "loss": 0.5636, "mean_token_accuracy": 0.8334645628929138, "num_tokens": 57657658.0, "step": 1517 }, { "epoch": 0.19310520290039435, "ewc_loss": 0.04052734375, "ewc_loss_diag": 2.7298927307128906e-05, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 3.7790987491607666, "learning_rate": 6.430690970750318e-06, "loss": 0.4916, "mean_token_accuracy": 0.850355327129364, "num_tokens": 57695964.0, "step": 1518 }, { "epoch": 0.19323241317898487, "ewc_loss": 0.036376953125, "ewc_loss_diag": 2.7298927307128906e-05, "ewc_loss_parallel": 3.361701965332031e-05, "grad_norm": 3.3563272953033447, "learning_rate": 6.434930055108097e-06, "loss": 0.538, "mean_token_accuracy": 0.8373473882675171, "num_tokens": 57725722.0, "step": 1519 }, { "epoch": 0.19335962345757537, "ewc_loss": 0.041015625, "ewc_loss_diag": 2.7298927307128906e-05, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 3.860957622528076, "learning_rate": 6.439169139465876e-06, "loss": 0.5537, "mean_token_accuracy": 0.8332244753837585, "num_tokens": 57761944.0, "step": 1520 }, { "epoch": 0.19348683373616588, "ewc_loss": 0.03662109375, "ewc_loss_diag": 2.765655517578125e-05, "ewc_loss_parallel": 3.3855438232421875e-05, "grad_norm": 3.3213343620300293, "learning_rate": 6.443408223823655e-06, "loss": 0.5085, "mean_token_accuracy": 0.8441101908683777, "num_tokens": 57796853.0, "step": 1521 }, { "epoch": 0.1936140440147564, "ewc_loss": 0.0419921875, "ewc_loss_diag": 2.765655517578125e-05, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 3.9631850719451904, "learning_rate": 6.447647308181433e-06, "loss": 0.5155, "mean_token_accuracy": 0.8433330059051514, "num_tokens": 57835295.0, "step": 1522 }, { "epoch": 0.1937412542933469, "ewc_loss": 0.037109375, "ewc_loss_diag": 2.765655517578125e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 3.325740337371826, "learning_rate": 6.451886392539212e-06, "loss": 0.487, "mean_token_accuracy": 0.8518099188804626, "num_tokens": 57876294.0, "step": 1523 }, { "epoch": 0.1938684645719374, "ewc_loss": 0.043701171875, "ewc_loss_diag": 2.765655517578125e-05, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 4.0790114402771, "learning_rate": 6.4561254768969905e-06, "loss": 0.524, "mean_token_accuracy": 0.8459396958351135, "num_tokens": 57915332.0, "step": 1524 }, { "epoch": 0.19399567485052793, "ewc_loss": 0.037841796875, "ewc_loss_diag": 2.765655517578125e-05, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 3.403804063796997, "learning_rate": 6.460364561254769e-06, "loss": 0.4747, "mean_token_accuracy": 0.8567681908607483, "num_tokens": 57955624.0, "step": 1525 }, { "epoch": 0.19412288512911843, "ewc_loss": 0.04541015625, "ewc_loss_diag": 2.765655517578125e-05, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 4.14313268661499, "learning_rate": 6.464603645612548e-06, "loss": 0.4749, "mean_token_accuracy": 0.8585200309753418, "num_tokens": 57993692.0, "step": 1526 }, { "epoch": 0.19425009540770893, "ewc_loss": 0.0390625, "ewc_loss_diag": 2.7894973754882812e-05, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 3.5176899433135986, "learning_rate": 6.468842729970327e-06, "loss": 0.511, "mean_token_accuracy": 0.8488609194755554, "num_tokens": 58035813.0, "step": 1527 }, { "epoch": 0.19437730568629946, "ewc_loss": 0.044677734375, "ewc_loss_diag": 2.7894973754882812e-05, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 4.120560169219971, "learning_rate": 6.473081814328106e-06, "loss": 0.5546, "mean_token_accuracy": 0.8412325978279114, "num_tokens": 58077118.0, "step": 1528 }, { "epoch": 0.19450451596488996, "ewc_loss": 0.04052734375, "ewc_loss_diag": 2.7894973754882812e-05, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 3.61376690864563, "learning_rate": 6.477320898685885e-06, "loss": 0.525, "mean_token_accuracy": 0.8433124423027039, "num_tokens": 58120853.0, "step": 1529 }, { "epoch": 0.19463172624348046, "ewc_loss": 0.04248046875, "ewc_loss_diag": 2.7894973754882812e-05, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 4.01201057434082, "learning_rate": 6.481559983043663e-06, "loss": 0.5418, "mean_token_accuracy": 0.8354759216308594, "num_tokens": 58160279.0, "step": 1530 }, { "epoch": 0.194758936522071, "ewc_loss": 0.0400390625, "ewc_loss_diag": 2.8014183044433594e-05, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 3.6434473991394043, "learning_rate": 6.485799067401442e-06, "loss": 0.4859, "mean_token_accuracy": 0.8553230166435242, "num_tokens": 58200902.0, "step": 1531 }, { "epoch": 0.1948861468006615, "ewc_loss": 0.0419921875, "ewc_loss_diag": 2.8014183044433594e-05, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 3.8827176094055176, "learning_rate": 6.490038151759221e-06, "loss": 0.4516, "mean_token_accuracy": 0.8647421598434448, "num_tokens": 58237179.0, "step": 1532 }, { "epoch": 0.195013357079252, "ewc_loss": 0.03955078125, "ewc_loss_diag": 2.8014183044433594e-05, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 3.6368439197540283, "learning_rate": 6.494277236117e-06, "loss": 0.4521, "mean_token_accuracy": 0.8619914054870605, "num_tokens": 58271072.0, "step": 1533 }, { "epoch": 0.19514056735784252, "ewc_loss": 0.04052734375, "ewc_loss_diag": 2.8014183044433594e-05, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 3.7788095474243164, "learning_rate": 6.4985163204747785e-06, "loss": 0.6003, "mean_token_accuracy": 0.8220735788345337, "num_tokens": 58317186.0, "step": 1534 }, { "epoch": 0.19526777763643302, "ewc_loss": 0.0390625, "ewc_loss_diag": 2.8014183044433594e-05, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 3.6055140495300293, "learning_rate": 6.5027554048325566e-06, "loss": 0.5039, "mean_token_accuracy": 0.8445543050765991, "num_tokens": 58353396.0, "step": 1535 }, { "epoch": 0.19539498791502352, "ewc_loss": 0.03955078125, "ewc_loss_diag": 2.8014183044433594e-05, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 3.7395403385162354, "learning_rate": 6.5069944891903354e-06, "loss": 0.5783, "mean_token_accuracy": 0.8315832614898682, "num_tokens": 58392466.0, "step": 1536 }, { "epoch": 0.19552219819361405, "ewc_loss": 0.03857421875, "ewc_loss_diag": 2.8014183044433594e-05, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 3.5535261631011963, "learning_rate": 6.511233573548114e-06, "loss": 0.462, "mean_token_accuracy": 0.8601772785186768, "num_tokens": 58432745.0, "step": 1537 }, { "epoch": 0.19564940847220455, "ewc_loss": 0.03955078125, "ewc_loss_diag": 2.8014183044433594e-05, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 3.7019779682159424, "learning_rate": 6.515472657905892e-06, "loss": 0.5153, "mean_token_accuracy": 0.8437312841415405, "num_tokens": 58464288.0, "step": 1538 }, { "epoch": 0.19577661875079505, "ewc_loss": 0.0380859375, "ewc_loss_diag": 2.8014183044433594e-05, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 3.528904438018799, "learning_rate": 6.519711742263671e-06, "loss": 0.567, "mean_token_accuracy": 0.8309311866760254, "num_tokens": 58501232.0, "step": 1539 }, { "epoch": 0.19590382902938558, "ewc_loss": 0.03955078125, "ewc_loss_diag": 2.8014183044433594e-05, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 3.6843841075897217, "learning_rate": 6.52395082662145e-06, "loss": 0.5667, "mean_token_accuracy": 0.8266894817352295, "num_tokens": 58539749.0, "step": 1540 }, { "epoch": 0.19603103930797608, "ewc_loss": 0.0380859375, "ewc_loss_diag": 2.8014183044433594e-05, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 3.485342502593994, "learning_rate": 6.528189910979229e-06, "loss": 0.4922, "mean_token_accuracy": 0.8522939085960388, "num_tokens": 58576436.0, "step": 1541 }, { "epoch": 0.1961582495865666, "ewc_loss": 0.03955078125, "ewc_loss_diag": 2.8014183044433594e-05, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 3.680626392364502, "learning_rate": 6.532428995337008e-06, "loss": 0.5017, "mean_token_accuracy": 0.8497621417045593, "num_tokens": 58618677.0, "step": 1542 }, { "epoch": 0.1962854598651571, "ewc_loss": 0.0380859375, "ewc_loss_diag": 2.8014183044433594e-05, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 3.4581921100616455, "learning_rate": 6.536668079694787e-06, "loss": 0.525, "mean_token_accuracy": 0.8377737998962402, "num_tokens": 58654404.0, "step": 1543 }, { "epoch": 0.1964126701437476, "ewc_loss": 0.04052734375, "ewc_loss_diag": 2.8014183044433594e-05, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 3.7384092807769775, "learning_rate": 6.540907164052566e-06, "loss": 0.5203, "mean_token_accuracy": 0.8422131538391113, "num_tokens": 58697312.0, "step": 1544 }, { "epoch": 0.19653988042233814, "ewc_loss": 0.037841796875, "ewc_loss_diag": 2.8252601623535156e-05, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 3.416414499282837, "learning_rate": 6.545146248410344e-06, "loss": 0.5266, "mean_token_accuracy": 0.8400385975837708, "num_tokens": 58738940.0, "step": 1545 }, { "epoch": 0.19666709070092864, "ewc_loss": 0.041259765625, "ewc_loss_diag": 2.849102020263672e-05, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 3.796250104904175, "learning_rate": 6.549385332768123e-06, "loss": 0.5373, "mean_token_accuracy": 0.8414850234985352, "num_tokens": 58779626.0, "step": 1546 }, { "epoch": 0.19679430097951914, "ewc_loss": 0.037841796875, "ewc_loss_diag": 2.849102020263672e-05, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 3.3858630657196045, "learning_rate": 6.5536244171259015e-06, "loss": 0.5317, "mean_token_accuracy": 0.8370141983032227, "num_tokens": 58818938.0, "step": 1547 }, { "epoch": 0.19692151125810967, "ewc_loss": 0.041748046875, "ewc_loss_diag": 2.849102020263672e-05, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 3.8889002799987793, "learning_rate": 6.55786350148368e-06, "loss": 0.5519, "mean_token_accuracy": 0.8326093554496765, "num_tokens": 58855973.0, "step": 1548 }, { "epoch": 0.19704872153670017, "ewc_loss": 0.03759765625, "ewc_loss_diag": 2.849102020263672e-05, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 3.3495547771453857, "learning_rate": 6.562102585841459e-06, "loss": 0.4888, "mean_token_accuracy": 0.8481698632240295, "num_tokens": 58894624.0, "step": 1549 }, { "epoch": 0.19717593181529067, "ewc_loss": 0.04345703125, "ewc_loss_diag": 2.849102020263672e-05, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 4.012784481048584, "learning_rate": 6.566341670199238e-06, "loss": 0.5054, "mean_token_accuracy": 0.8485627174377441, "num_tokens": 58934440.0, "step": 1550 }, { "epoch": 0.1973031420938812, "ewc_loss": 0.037841796875, "ewc_loss_diag": 2.849102020263672e-05, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 3.355508327484131, "learning_rate": 6.570580754557017e-06, "loss": 0.5013, "mean_token_accuracy": 0.8494696617126465, "num_tokens": 58972150.0, "step": 1551 }, { "epoch": 0.1974303523724717, "ewc_loss": 0.04638671875, "ewc_loss_diag": 2.849102020263672e-05, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 4.150111198425293, "learning_rate": 6.574819838914795e-06, "loss": 0.4863, "mean_token_accuracy": 0.8546890616416931, "num_tokens": 59006179.0, "step": 1552 }, { "epoch": 0.1975575626510622, "ewc_loss": 0.038818359375, "ewc_loss_diag": 2.849102020263672e-05, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 3.430997848510742, "learning_rate": 6.579058923272573e-06, "loss": 0.533, "mean_token_accuracy": 0.8388853073120117, "num_tokens": 59046583.0, "step": 1553 }, { "epoch": 0.19768477292965272, "ewc_loss": 0.0478515625, "ewc_loss_diag": 2.849102020263672e-05, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 4.244755744934082, "learning_rate": 6.583298007630352e-06, "loss": 0.5314, "mean_token_accuracy": 0.8427181839942932, "num_tokens": 59078023.0, "step": 1554 }, { "epoch": 0.19781198320824323, "ewc_loss": 0.04052734375, "ewc_loss_diag": 2.849102020263672e-05, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 3.56968355178833, "learning_rate": 6.587537091988131e-06, "loss": 0.5289, "mean_token_accuracy": 0.839946985244751, "num_tokens": 59121160.0, "step": 1555 }, { "epoch": 0.19793919348683373, "ewc_loss": 0.04736328125, "ewc_loss_diag": 2.849102020263672e-05, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 4.205909252166748, "learning_rate": 6.59177617634591e-06, "loss": 0.5062, "mean_token_accuracy": 0.8464111089706421, "num_tokens": 59164564.0, "step": 1556 }, { "epoch": 0.19806640376542425, "ewc_loss": 0.042724609375, "ewc_loss_diag": 2.849102020263672e-05, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 3.672463893890381, "learning_rate": 6.596015260703689e-06, "loss": 0.5054, "mean_token_accuracy": 0.850937008857727, "num_tokens": 59211558.0, "step": 1557 }, { "epoch": 0.19819361404401475, "ewc_loss": 0.045166015625, "ewc_loss_diag": 2.849102020263672e-05, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 4.098156452178955, "learning_rate": 6.6002543450614675e-06, "loss": 0.5012, "mean_token_accuracy": 0.8496180772781372, "num_tokens": 59251260.0, "step": 1558 }, { "epoch": 0.19832082432260525, "ewc_loss": 0.042236328125, "ewc_loss_diag": 2.849102020263672e-05, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 3.7156002521514893, "learning_rate": 6.604493429419246e-06, "loss": 0.5597, "mean_token_accuracy": 0.8343075513839722, "num_tokens": 59293949.0, "step": 1559 }, { "epoch": 0.19844803460119578, "ewc_loss": 0.04248046875, "ewc_loss_diag": 2.849102020263672e-05, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 3.9621660709381104, "learning_rate": 6.6087325137770244e-06, "loss": 0.5716, "mean_token_accuracy": 0.8381098508834839, "num_tokens": 59341549.0, "step": 1560 }, { "epoch": 0.19857524487978628, "ewc_loss": 0.041259765625, "ewc_loss_diag": 2.872943878173828e-05, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 3.700263738632202, "learning_rate": 6.612971598134803e-06, "loss": 0.5097, "mean_token_accuracy": 0.8456456065177917, "num_tokens": 59379355.0, "step": 1561 }, { "epoch": 0.19870245515837678, "ewc_loss": 0.04150390625, "ewc_loss_diag": 2.872943878173828e-05, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 3.854795455932617, "learning_rate": 6.617210682492582e-06, "loss": 0.5574, "mean_token_accuracy": 0.829771876335144, "num_tokens": 59416340.0, "step": 1562 }, { "epoch": 0.1988296654369673, "ewc_loss": 0.0400390625, "ewc_loss_diag": 2.872943878173828e-05, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 3.660983085632324, "learning_rate": 6.621449766850361e-06, "loss": 0.4816, "mean_token_accuracy": 0.8543246388435364, "num_tokens": 59453100.0, "step": 1563 }, { "epoch": 0.1989568757155578, "ewc_loss": 0.040771484375, "ewc_loss_diag": 2.872943878173828e-05, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 3.767677068710327, "learning_rate": 6.62568885120814e-06, "loss": 0.5256, "mean_token_accuracy": 0.8420954346656799, "num_tokens": 59491164.0, "step": 1564 }, { "epoch": 0.19908408599414834, "ewc_loss": 0.0390625, "ewc_loss_diag": 2.872943878173828e-05, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 3.6331207752227783, "learning_rate": 6.629927935565919e-06, "loss": 0.5295, "mean_token_accuracy": 0.8398431539535522, "num_tokens": 59525752.0, "step": 1565 }, { "epoch": 0.19921129627273884, "ewc_loss": 0.039794921875, "ewc_loss_diag": 2.872943878173828e-05, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 3.727384328842163, "learning_rate": 6.634167019923698e-06, "loss": 0.5681, "mean_token_accuracy": 0.8341255187988281, "num_tokens": 59565206.0, "step": 1566 }, { "epoch": 0.19933850655132934, "ewc_loss": 0.03857421875, "ewc_loss_diag": 2.8848648071289062e-05, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 3.599184274673462, "learning_rate": 6.638406104281477e-06, "loss": 0.506, "mean_token_accuracy": 0.8465461730957031, "num_tokens": 59597578.0, "step": 1567 }, { "epoch": 0.19946571682991987, "ewc_loss": 0.040283203125, "ewc_loss_diag": 2.8848648071289062e-05, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 3.691412925720215, "learning_rate": 6.642645188639254e-06, "loss": 0.482, "mean_token_accuracy": 0.8529364466667175, "num_tokens": 59640073.0, "step": 1568 }, { "epoch": 0.19959292710851037, "ewc_loss": 0.03857421875, "ewc_loss_diag": 2.8848648071289062e-05, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 3.5632362365722656, "learning_rate": 6.646884272997033e-06, "loss": 0.5349, "mean_token_accuracy": 0.8365894556045532, "num_tokens": 59680259.0, "step": 1569 }, { "epoch": 0.19972013738710087, "ewc_loss": 0.0400390625, "ewc_loss_diag": 2.8848648071289062e-05, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 3.7018980979919434, "learning_rate": 6.651123357354812e-06, "loss": 0.5198, "mean_token_accuracy": 0.8435602784156799, "num_tokens": 59716372.0, "step": 1570 }, { "epoch": 0.1998473476656914, "ewc_loss": 0.0380859375, "ewc_loss_diag": 2.8848648071289062e-05, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 3.5401549339294434, "learning_rate": 6.6553624417125905e-06, "loss": 0.5379, "mean_token_accuracy": 0.8354345560073853, "num_tokens": 59751294.0, "step": 1571 }, { "epoch": 0.1999745579442819, "ewc_loss": 0.04052734375, "ewc_loss_diag": 2.8848648071289062e-05, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 3.714200258255005, "learning_rate": 6.659601526070369e-06, "loss": 0.5727, "mean_token_accuracy": 0.8339559435844421, "num_tokens": 59794395.0, "step": 1572 }, { "epoch": 0.2001017682228724, "ewc_loss": 0.03857421875, "ewc_loss_diag": 2.8848648071289062e-05, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 3.5232880115509033, "learning_rate": 6.663840610428148e-06, "loss": 0.5977, "mean_token_accuracy": 0.8251432776451111, "num_tokens": 59825503.0, "step": 1573 }, { "epoch": 0.20022897850146293, "ewc_loss": 0.04052734375, "ewc_loss_diag": 2.8848648071289062e-05, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 3.7435648441314697, "learning_rate": 6.668079694785927e-06, "loss": 0.4837, "mean_token_accuracy": 0.855634331703186, "num_tokens": 59864634.0, "step": 1574 }, { "epoch": 0.20035618878005343, "ewc_loss": 0.03857421875, "ewc_loss_diag": 2.8848648071289062e-05, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 3.471860647201538, "learning_rate": 6.672318779143705e-06, "loss": 0.5002, "mean_token_accuracy": 0.8466818332672119, "num_tokens": 59898693.0, "step": 1575 }, { "epoch": 0.20048339905864393, "ewc_loss": 0.04150390625, "ewc_loss_diag": 2.8848648071289062e-05, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 3.8124821186065674, "learning_rate": 6.676557863501484e-06, "loss": 0.5268, "mean_token_accuracy": 0.8427742123603821, "num_tokens": 59930234.0, "step": 1576 }, { "epoch": 0.20061060933723446, "ewc_loss": 0.03857421875, "ewc_loss_diag": 2.8848648071289062e-05, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 3.4118170738220215, "learning_rate": 6.680796947859263e-06, "loss": 0.5183, "mean_token_accuracy": 0.843706488609314, "num_tokens": 59970932.0, "step": 1577 }, { "epoch": 0.20073781961582496, "ewc_loss": 0.0419921875, "ewc_loss_diag": 2.8848648071289062e-05, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 3.918975591659546, "learning_rate": 6.685036032217042e-06, "loss": 0.5138, "mean_token_accuracy": 0.8450310826301575, "num_tokens": 60007282.0, "step": 1578 }, { "epoch": 0.20086502989441546, "ewc_loss": 0.0380859375, "ewc_loss_diag": 2.8848648071289062e-05, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 3.3676137924194336, "learning_rate": 6.689275116574821e-06, "loss": 0.4944, "mean_token_accuracy": 0.8517834544181824, "num_tokens": 60043336.0, "step": 1579 }, { "epoch": 0.200992240173006, "ewc_loss": 0.044921875, "ewc_loss_diag": 2.8848648071289062e-05, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 4.080028533935547, "learning_rate": 6.6935142009326e-06, "loss": 0.5281, "mean_token_accuracy": 0.845832109451294, "num_tokens": 60083284.0, "step": 1580 }, { "epoch": 0.2011194504515965, "ewc_loss": 0.0380859375, "ewc_loss_diag": 2.8848648071289062e-05, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 3.35457444190979, "learning_rate": 6.6977532852903785e-06, "loss": 0.5247, "mean_token_accuracy": 0.8395510911941528, "num_tokens": 60125115.0, "step": 1581 }, { "epoch": 0.201246660730187, "ewc_loss": 0.048828125, "ewc_loss_diag": 2.8967857360839844e-05, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 4.28136682510376, "learning_rate": 6.701992369648157e-06, "loss": 0.4502, "mean_token_accuracy": 0.8660813570022583, "num_tokens": 60161497.0, "step": 1582 }, { "epoch": 0.20137387100877752, "ewc_loss": 0.03955078125, "ewc_loss_diag": 2.8967857360839844e-05, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 3.4473986625671387, "learning_rate": 6.7062314540059345e-06, "loss": 0.4682, "mean_token_accuracy": 0.8578182458877563, "num_tokens": 60202282.0, "step": 1583 }, { "epoch": 0.20150108128736802, "ewc_loss": 0.05224609375, "ewc_loss_diag": 2.8967857360839844e-05, "ewc_loss_parallel": 4.935264587402344e-05, "grad_norm": 4.4328813552856445, "learning_rate": 6.7104705383637134e-06, "loss": 0.5161, "mean_token_accuracy": 0.8471100330352783, "num_tokens": 60241211.0, "step": 1584 }, { "epoch": 0.20162829156595852, "ewc_loss": 0.044189453125, "ewc_loss_diag": 2.8967857360839844e-05, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 3.6836531162261963, "learning_rate": 6.714709622721492e-06, "loss": 0.5469, "mean_token_accuracy": 0.8374319076538086, "num_tokens": 60273677.0, "step": 1585 }, { "epoch": 0.20175550184454905, "ewc_loss": 0.050537109375, "ewc_loss_diag": 2.8967857360839844e-05, "ewc_loss_parallel": 4.76837158203125e-05, "grad_norm": 4.3702616691589355, "learning_rate": 6.718948707079271e-06, "loss": 0.5119, "mean_token_accuracy": 0.8496239185333252, "num_tokens": 60313904.0, "step": 1586 }, { "epoch": 0.20188271212313955, "ewc_loss": 0.046630859375, "ewc_loss_diag": 2.8967857360839844e-05, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 3.8176422119140625, "learning_rate": 6.72318779143705e-06, "loss": 0.5083, "mean_token_accuracy": 0.8491672277450562, "num_tokens": 60351534.0, "step": 1587 }, { "epoch": 0.20200992240173005, "ewc_loss": 0.04833984375, "ewc_loss_diag": 2.8967857360839844e-05, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 4.209682941436768, "learning_rate": 6.727426875794829e-06, "loss": 0.5533, "mean_token_accuracy": 0.8373458981513977, "num_tokens": 60393362.0, "step": 1588 }, { "epoch": 0.20213713268032057, "ewc_loss": 0.045654296875, "ewc_loss_diag": 2.8967857360839844e-05, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 3.8279404640197754, "learning_rate": 6.731665960152608e-06, "loss": 0.5223, "mean_token_accuracy": 0.8425601720809937, "num_tokens": 60437735.0, "step": 1589 }, { "epoch": 0.20226434295891108, "ewc_loss": 0.04443359375, "ewc_loss_diag": 2.8967857360839844e-05, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 4.063358783721924, "learning_rate": 6.735905044510387e-06, "loss": 0.5801, "mean_token_accuracy": 0.8225322365760803, "num_tokens": 60476032.0, "step": 1590 }, { "epoch": 0.2023915532375016, "ewc_loss": 0.043212890625, "ewc_loss_diag": 2.8967857360839844e-05, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 3.7968931198120117, "learning_rate": 6.740144128868165e-06, "loss": 0.5298, "mean_token_accuracy": 0.8412183523178101, "num_tokens": 60513330.0, "step": 1591 }, { "epoch": 0.2025187635160921, "ewc_loss": 0.04248046875, "ewc_loss_diag": 2.8967857360839844e-05, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 3.9322702884674072, "learning_rate": 6.744383213225944e-06, "loss": 0.5575, "mean_token_accuracy": 0.832817792892456, "num_tokens": 60552829.0, "step": 1592 }, { "epoch": 0.2026459737946826, "ewc_loss": 0.041259765625, "ewc_loss_diag": 2.9087066650390625e-05, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 3.7496399879455566, "learning_rate": 6.7486222975837225e-06, "loss": 0.4885, "mean_token_accuracy": 0.848116934299469, "num_tokens": 60585333.0, "step": 1593 }, { "epoch": 0.20277318407327313, "ewc_loss": 0.041748046875, "ewc_loss_diag": 2.9087066650390625e-05, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 3.8479490280151367, "learning_rate": 6.7528613819415014e-06, "loss": 0.5217, "mean_token_accuracy": 0.8408567905426025, "num_tokens": 60621091.0, "step": 1594 }, { "epoch": 0.20290039435186363, "ewc_loss": 0.040283203125, "ewc_loss_diag": 2.9087066650390625e-05, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 3.6869049072265625, "learning_rate": 6.75710046629928e-06, "loss": 0.5214, "mean_token_accuracy": 0.8449669480323792, "num_tokens": 60661602.0, "step": 1595 }, { "epoch": 0.20302760463045413, "ewc_loss": 0.04150390625, "ewc_loss_diag": 2.9087066650390625e-05, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 3.8278372287750244, "learning_rate": 6.761339550657059e-06, "loss": 0.529, "mean_token_accuracy": 0.8366248607635498, "num_tokens": 60689085.0, "step": 1596 }, { "epoch": 0.20315481490904466, "ewc_loss": 0.03955078125, "ewc_loss_diag": 2.9087066650390625e-05, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 3.6331324577331543, "learning_rate": 6.765578635014838e-06, "loss": 0.5208, "mean_token_accuracy": 0.8461687564849854, "num_tokens": 60727259.0, "step": 1597 }, { "epoch": 0.20328202518763516, "ewc_loss": 0.04150390625, "ewc_loss_diag": 2.9087066650390625e-05, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 3.8036417961120605, "learning_rate": 6.769817719372615e-06, "loss": 0.4981, "mean_token_accuracy": 0.8493102788925171, "num_tokens": 60761880.0, "step": 1598 }, { "epoch": 0.20340923546622566, "ewc_loss": 0.038818359375, "ewc_loss_diag": 2.9087066650390625e-05, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 3.5968360900878906, "learning_rate": 6.774056803730394e-06, "loss": 0.5374, "mean_token_accuracy": 0.8380665183067322, "num_tokens": 60795416.0, "step": 1599 }, { "epoch": 0.2035364457448162, "ewc_loss": 0.04150390625, "ewc_loss_diag": 2.9206275939941406e-05, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 3.8115224838256836, "learning_rate": 6.778295888088173e-06, "loss": 0.5173, "mean_token_accuracy": 0.8450490236282349, "num_tokens": 60833373.0, "step": 1600 }, { "epoch": 0.2036636560234067, "ewc_loss": 0.03857421875, "ewc_loss_diag": 2.9206275939941406e-05, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 3.5535738468170166, "learning_rate": 6.782534972445952e-06, "loss": 0.5303, "mean_token_accuracy": 0.8366581201553345, "num_tokens": 60870024.0, "step": 1601 }, { "epoch": 0.2037908663019972, "ewc_loss": 0.041748046875, "ewc_loss_diag": 2.9206275939941406e-05, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 3.86952805519104, "learning_rate": 6.786774056803731e-06, "loss": 0.5262, "mean_token_accuracy": 0.8439061641693115, "num_tokens": 60909790.0, "step": 1602 }, { "epoch": 0.20391807658058772, "ewc_loss": 0.03857421875, "ewc_loss_diag": 2.9206275939941406e-05, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 3.499678134918213, "learning_rate": 6.79101314116151e-06, "loss": 0.5303, "mean_token_accuracy": 0.8382136225700378, "num_tokens": 60948793.0, "step": 1603 }, { "epoch": 0.20404528685917822, "ewc_loss": 0.042724609375, "ewc_loss_diag": 2.9206275939941406e-05, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 3.954169988632202, "learning_rate": 6.795252225519289e-06, "loss": 0.5616, "mean_token_accuracy": 0.8305157423019409, "num_tokens": 60991370.0, "step": 1604 }, { "epoch": 0.20417249713776872, "ewc_loss": 0.03857421875, "ewc_loss_diag": 2.9206275939941406e-05, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 3.4544994831085205, "learning_rate": 6.7994913098770675e-06, "loss": 0.5214, "mean_token_accuracy": 0.8410576581954956, "num_tokens": 61028137.0, "step": 1605 }, { "epoch": 0.20429970741635925, "ewc_loss": 0.045166015625, "ewc_loss_diag": 2.9206275939941406e-05, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 4.087722301483154, "learning_rate": 6.8037303942348455e-06, "loss": 0.4843, "mean_token_accuracy": 0.8545211553573608, "num_tokens": 61060843.0, "step": 1606 }, { "epoch": 0.20442691769494975, "ewc_loss": 0.03857421875, "ewc_loss_diag": 2.944469451904297e-05, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 3.4401838779449463, "learning_rate": 6.807969478592624e-06, "loss": 0.5238, "mean_token_accuracy": 0.8431506156921387, "num_tokens": 61096272.0, "step": 1607 }, { "epoch": 0.20455412797354025, "ewc_loss": 0.04736328125, "ewc_loss_diag": 2.944469451904297e-05, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 4.2314558029174805, "learning_rate": 6.812208562950403e-06, "loss": 0.5532, "mean_token_accuracy": 0.8421571254730225, "num_tokens": 61139119.0, "step": 1608 }, { "epoch": 0.20468133825213078, "ewc_loss": 0.039794921875, "ewc_loss_diag": 2.956390380859375e-05, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 3.481682062149048, "learning_rate": 6.816447647308182e-06, "loss": 0.5781, "mean_token_accuracy": 0.8267782330513, "num_tokens": 61183327.0, "step": 1609 }, { "epoch": 0.20480854853072128, "ewc_loss": 0.05126953125, "ewc_loss_diag": 2.956390380859375e-05, "ewc_loss_parallel": 4.839897155761719e-05, "grad_norm": 4.3684773445129395, "learning_rate": 6.820686731665961e-06, "loss": 0.5218, "mean_token_accuracy": 0.8455467224121094, "num_tokens": 61218787.0, "step": 1610 }, { "epoch": 0.20493575880931178, "ewc_loss": 0.042724609375, "ewc_loss_diag": 2.956390380859375e-05, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 3.6276867389678955, "learning_rate": 6.82492581602374e-06, "loss": 0.5209, "mean_token_accuracy": 0.844256579875946, "num_tokens": 61256517.0, "step": 1611 }, { "epoch": 0.2050629690879023, "ewc_loss": 0.051513671875, "ewc_loss_diag": 2.956390380859375e-05, "ewc_loss_parallel": 4.863739013671875e-05, "grad_norm": 4.365507125854492, "learning_rate": 6.829164900381519e-06, "loss": 0.5177, "mean_token_accuracy": 0.8475913405418396, "num_tokens": 61301065.0, "step": 1612 }, { "epoch": 0.2051901793664928, "ewc_loss": 0.046142578125, "ewc_loss_diag": 2.968311309814453e-05, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 3.765190362930298, "learning_rate": 6.833403984739296e-06, "loss": 0.5175, "mean_token_accuracy": 0.8445863723754883, "num_tokens": 61336368.0, "step": 1613 }, { "epoch": 0.2053173896450833, "ewc_loss": 0.048828125, "ewc_loss_diag": 2.968311309814453e-05, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 4.265403747558594, "learning_rate": 6.837643069097075e-06, "loss": 0.5977, "mean_token_accuracy": 0.8228344917297363, "num_tokens": 61373742.0, "step": 1614 }, { "epoch": 0.20544459992367384, "ewc_loss": 0.0458984375, "ewc_loss_diag": 2.968311309814453e-05, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 3.8170719146728516, "learning_rate": 6.841882153454854e-06, "loss": 0.5052, "mean_token_accuracy": 0.8477990627288818, "num_tokens": 61408454.0, "step": 1615 }, { "epoch": 0.20557181020226434, "ewc_loss": 0.0458984375, "ewc_loss_diag": 2.968311309814453e-05, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 4.1169891357421875, "learning_rate": 6.846121237812633e-06, "loss": 0.4753, "mean_token_accuracy": 0.8591094613075256, "num_tokens": 61445420.0, "step": 1616 }, { "epoch": 0.20569902048085487, "ewc_loss": 0.044921875, "ewc_loss_diag": 2.968311309814453e-05, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 3.8046646118164062, "learning_rate": 6.8503603221704115e-06, "loss": 0.5033, "mean_token_accuracy": 0.849481463432312, "num_tokens": 61487607.0, "step": 1617 }, { "epoch": 0.20582623075944537, "ewc_loss": 0.044189453125, "ewc_loss_diag": 2.968311309814453e-05, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 3.9957938194274902, "learning_rate": 6.85459940652819e-06, "loss": 0.5102, "mean_token_accuracy": 0.846511960029602, "num_tokens": 61521053.0, "step": 1618 }, { "epoch": 0.20595344103803587, "ewc_loss": 0.04248046875, "ewc_loss_diag": 2.968311309814453e-05, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 3.7750141620635986, "learning_rate": 6.858838490885969e-06, "loss": 0.5677, "mean_token_accuracy": 0.8279134631156921, "num_tokens": 61562211.0, "step": 1619 }, { "epoch": 0.2060806513166264, "ewc_loss": 0.04296875, "ewc_loss_diag": 2.9802322387695312e-05, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 3.899794101715088, "learning_rate": 6.863077575243748e-06, "loss": 0.5805, "mean_token_accuracy": 0.8306818604469299, "num_tokens": 61600089.0, "step": 1620 }, { "epoch": 0.2062078615952169, "ewc_loss": 0.041015625, "ewc_loss_diag": 2.9802322387695312e-05, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 3.741220712661743, "learning_rate": 6.867316659601526e-06, "loss": 0.4999, "mean_token_accuracy": 0.8475604057312012, "num_tokens": 61636226.0, "step": 1621 }, { "epoch": 0.2063350718738074, "ewc_loss": 0.04248046875, "ewc_loss_diag": 2.9802322387695312e-05, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 3.8336164951324463, "learning_rate": 6.871555743959305e-06, "loss": 0.5278, "mean_token_accuracy": 0.8435764908790588, "num_tokens": 61677224.0, "step": 1622 }, { "epoch": 0.20646228215239792, "ewc_loss": 0.040771484375, "ewc_loss_diag": 2.9802322387695312e-05, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 3.708305835723877, "learning_rate": 6.875794828317084e-06, "loss": 0.5025, "mean_token_accuracy": 0.8488593101501465, "num_tokens": 61715174.0, "step": 1623 }, { "epoch": 0.20658949243098843, "ewc_loss": 0.0419921875, "ewc_loss_diag": 2.9802322387695312e-05, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 3.801804542541504, "learning_rate": 6.880033912674863e-06, "loss": 0.492, "mean_token_accuracy": 0.8528991937637329, "num_tokens": 61760819.0, "step": 1624 }, { "epoch": 0.20671670270957893, "ewc_loss": 0.040771484375, "ewc_loss_diag": 2.9802322387695312e-05, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 3.679819107055664, "learning_rate": 6.884272997032642e-06, "loss": 0.6, "mean_token_accuracy": 0.8216283321380615, "num_tokens": 61800502.0, "step": 1625 }, { "epoch": 0.20684391298816945, "ewc_loss": 0.0419921875, "ewc_loss_diag": 2.9802322387695312e-05, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 3.798550844192505, "learning_rate": 6.888512081390421e-06, "loss": 0.4864, "mean_token_accuracy": 0.854773223400116, "num_tokens": 61834922.0, "step": 1626 }, { "epoch": 0.20697112326675995, "ewc_loss": 0.039794921875, "ewc_loss_diag": 2.9802322387695312e-05, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 3.6480774879455566, "learning_rate": 6.8927511657481995e-06, "loss": 0.5442, "mean_token_accuracy": 0.8355792760848999, "num_tokens": 61874675.0, "step": 1627 }, { "epoch": 0.20709833354535045, "ewc_loss": 0.041748046875, "ewc_loss_diag": 2.9802322387695312e-05, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 3.8147833347320557, "learning_rate": 6.896990250105978e-06, "loss": 0.4885, "mean_token_accuracy": 0.8556307554244995, "num_tokens": 61914310.0, "step": 1628 }, { "epoch": 0.20722554382394098, "ewc_loss": 0.039794921875, "ewc_loss_diag": 2.9802322387695312e-05, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 3.6075122356414795, "learning_rate": 6.901229334463756e-06, "loss": 0.4925, "mean_token_accuracy": 0.848261833190918, "num_tokens": 61949142.0, "step": 1629 }, { "epoch": 0.20735275410253148, "ewc_loss": 0.041748046875, "ewc_loss_diag": 2.9802322387695312e-05, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 3.8460099697113037, "learning_rate": 6.9054684188215345e-06, "loss": 0.5122, "mean_token_accuracy": 0.8514455556869507, "num_tokens": 61989939.0, "step": 1630 }, { "epoch": 0.20747996438112198, "ewc_loss": 0.039306640625, "ewc_loss_diag": 2.9802322387695312e-05, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 3.5670528411865234, "learning_rate": 6.909707503179313e-06, "loss": 0.5139, "mean_token_accuracy": 0.8431203365325928, "num_tokens": 62029263.0, "step": 1631 }, { "epoch": 0.2076071746597125, "ewc_loss": 0.042724609375, "ewc_loss_diag": 2.9802322387695312e-05, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 3.9051079750061035, "learning_rate": 6.913946587537092e-06, "loss": 0.5051, "mean_token_accuracy": 0.8497254252433777, "num_tokens": 62067963.0, "step": 1632 }, { "epoch": 0.207734384938303, "ewc_loss": 0.03857421875, "ewc_loss_diag": 2.9802322387695312e-05, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 3.5274744033813477, "learning_rate": 6.918185671894871e-06, "loss": 0.5833, "mean_token_accuracy": 0.8245174884796143, "num_tokens": 62105313.0, "step": 1633 }, { "epoch": 0.2078615952168935, "ewc_loss": 0.04443359375, "ewc_loss_diag": 2.9802322387695312e-05, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 4.01933479309082, "learning_rate": 6.92242475625265e-06, "loss": 0.5569, "mean_token_accuracy": 0.8449794054031372, "num_tokens": 62140836.0, "step": 1634 }, { "epoch": 0.20798880549548404, "ewc_loss": 0.038818359375, "ewc_loss_diag": 2.9802322387695312e-05, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 3.46116304397583, "learning_rate": 6.926663840610429e-06, "loss": 0.4676, "mean_token_accuracy": 0.8589850664138794, "num_tokens": 62178951.0, "step": 1635 }, { "epoch": 0.20811601577407454, "ewc_loss": 0.04736328125, "ewc_loss_diag": 2.9802322387695312e-05, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 4.1762261390686035, "learning_rate": 6.930902924968207e-06, "loss": 0.5124, "mean_token_accuracy": 0.8462916612625122, "num_tokens": 62216941.0, "step": 1636 }, { "epoch": 0.20824322605266504, "ewc_loss": 0.03955078125, "ewc_loss_diag": 3.0040740966796875e-05, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 3.4396917819976807, "learning_rate": 6.935142009325986e-06, "loss": 0.5428, "mean_token_accuracy": 0.8372201919555664, "num_tokens": 62257676.0, "step": 1637 }, { "epoch": 0.20837043633125557, "ewc_loss": 0.05078125, "ewc_loss_diag": 3.0040740966796875e-05, "ewc_loss_parallel": 4.792213439941406e-05, "grad_norm": 4.384043216705322, "learning_rate": 6.939381093683765e-06, "loss": 0.5664, "mean_token_accuracy": 0.834147036075592, "num_tokens": 62296878.0, "step": 1638 }, { "epoch": 0.20849764660984607, "ewc_loss": 0.04052734375, "ewc_loss_diag": 3.0159950256347656e-05, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 3.5184710025787354, "learning_rate": 6.943620178041544e-06, "loss": 0.4765, "mean_token_accuracy": 0.8568719029426575, "num_tokens": 62335292.0, "step": 1639 }, { "epoch": 0.2086248568884366, "ewc_loss": 0.0546875, "ewc_loss_diag": 3.0279159545898438e-05, "ewc_loss_parallel": 5.173683166503906e-05, "grad_norm": 4.558631420135498, "learning_rate": 6.9478592623993225e-06, "loss": 0.5311, "mean_token_accuracy": 0.8504328727722168, "num_tokens": 62371182.0, "step": 1640 }, { "epoch": 0.2087520671670271, "ewc_loss": 0.046630859375, "ewc_loss_diag": 3.0279159545898438e-05, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 3.742297649383545, "learning_rate": 6.952098346757101e-06, "loss": 0.5126, "mean_token_accuracy": 0.8455976247787476, "num_tokens": 62407803.0, "step": 1641 }, { "epoch": 0.2088792774456176, "ewc_loss": 0.0546875, "ewc_loss_diag": 3.0279159545898438e-05, "ewc_loss_parallel": 5.173683166503906e-05, "grad_norm": 4.53984260559082, "learning_rate": 6.95633743111488e-06, "loss": 0.5095, "mean_token_accuracy": 0.852342963218689, "num_tokens": 62446320.0, "step": 1642 }, { "epoch": 0.20900648772420813, "ewc_loss": 0.048828125, "ewc_loss_diag": 3.0279159545898438e-05, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 3.932159185409546, "learning_rate": 6.960576515472659e-06, "loss": 0.5479, "mean_token_accuracy": 0.8394052386283875, "num_tokens": 62486919.0, "step": 1643 }, { "epoch": 0.20913369800279863, "ewc_loss": 0.0517578125, "ewc_loss_diag": 3.0279159545898438e-05, "ewc_loss_parallel": 4.887580871582031e-05, "grad_norm": 4.369575500488281, "learning_rate": 6.964815599830436e-06, "loss": 0.5296, "mean_token_accuracy": 0.84693843126297, "num_tokens": 62521511.0, "step": 1644 }, { "epoch": 0.20926090828138913, "ewc_loss": 0.048583984375, "ewc_loss_diag": 3.0279159545898438e-05, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 3.9624710083007812, "learning_rate": 6.969054684188215e-06, "loss": 0.5829, "mean_token_accuracy": 0.8237552642822266, "num_tokens": 62564702.0, "step": 1645 }, { "epoch": 0.20938811855997966, "ewc_loss": 0.0478515625, "ewc_loss_diag": 3.0279159545898438e-05, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 4.1739630699157715, "learning_rate": 6.973293768545994e-06, "loss": 0.4927, "mean_token_accuracy": 0.854796290397644, "num_tokens": 62605850.0, "step": 1646 }, { "epoch": 0.20951532883857016, "ewc_loss": 0.046142578125, "ewc_loss_diag": 3.0279159545898438e-05, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 3.9124045372009277, "learning_rate": 6.977532852903773e-06, "loss": 0.52, "mean_token_accuracy": 0.8430494666099548, "num_tokens": 62650497.0, "step": 1647 }, { "epoch": 0.20964253911716066, "ewc_loss": 0.0458984375, "ewc_loss_diag": 3.0279159545898438e-05, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 4.0532402992248535, "learning_rate": 6.981771937261552e-06, "loss": 0.5741, "mean_token_accuracy": 0.8312141299247742, "num_tokens": 62685538.0, "step": 1648 }, { "epoch": 0.2097697493957512, "ewc_loss": 0.04443359375, "ewc_loss_diag": 3.0279159545898438e-05, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 3.8539657592773438, "learning_rate": 6.986011021619331e-06, "loss": 0.528, "mean_token_accuracy": 0.842431902885437, "num_tokens": 62725076.0, "step": 1649 }, { "epoch": 0.2098969596743417, "ewc_loss": 0.043701171875, "ewc_loss_diag": 3.0279159545898438e-05, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 3.9709784984588623, "learning_rate": 6.99025010597711e-06, "loss": 0.5449, "mean_token_accuracy": 0.8377429842948914, "num_tokens": 62762335.0, "step": 1650 }, { "epoch": 0.2100241699529322, "ewc_loss": 0.042236328125, "ewc_loss_diag": 3.0279159545898438e-05, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 3.7907276153564453, "learning_rate": 6.994489190334888e-06, "loss": 0.5018, "mean_token_accuracy": 0.8483627438545227, "num_tokens": 62804507.0, "step": 1651 }, { "epoch": 0.21015138023152272, "ewc_loss": 0.043212890625, "ewc_loss_diag": 3.0279159545898438e-05, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 3.908535957336426, "learning_rate": 6.9987282746926666e-06, "loss": 0.4715, "mean_token_accuracy": 0.859026312828064, "num_tokens": 62844499.0, "step": 1652 }, { "epoch": 0.21027859051011322, "ewc_loss": 0.04150390625, "ewc_loss_diag": 3.039836883544922e-05, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 3.7545740604400635, "learning_rate": 7.0029673590504455e-06, "loss": 0.5129, "mean_token_accuracy": 0.8510191440582275, "num_tokens": 62880061.0, "step": 1653 }, { "epoch": 0.21040580078870372, "ewc_loss": 0.04296875, "ewc_loss_diag": 3.039836883544922e-05, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 4.261139392852783, "learning_rate": 7.007206443408224e-06, "loss": 0.462, "mean_token_accuracy": 0.8602467775344849, "num_tokens": 62916726.0, "step": 1654 }, { "epoch": 0.21053301106729425, "ewc_loss": 0.04052734375, "ewc_loss_diag": 3.039836883544922e-05, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 3.621805191040039, "learning_rate": 7.011445527766003e-06, "loss": 0.5064, "mean_token_accuracy": 0.8449020385742188, "num_tokens": 62953341.0, "step": 1655 }, { "epoch": 0.21066022134588475, "ewc_loss": 0.046630859375, "ewc_loss_diag": 3.039836883544922e-05, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 4.137123107910156, "learning_rate": 7.015684612123782e-06, "loss": 0.5055, "mean_token_accuracy": 0.8479477167129517, "num_tokens": 62995047.0, "step": 1656 }, { "epoch": 0.21078743162447525, "ewc_loss": 0.039794921875, "ewc_loss_diag": 3.039836883544922e-05, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 3.542449474334717, "learning_rate": 7.019923696481561e-06, "loss": 0.5283, "mean_token_accuracy": 0.8412948250770569, "num_tokens": 63039520.0, "step": 1657 }, { "epoch": 0.21091464190306577, "ewc_loss": 0.049560546875, "ewc_loss_diag": 3.039836883544922e-05, "ewc_loss_parallel": 4.673004150390625e-05, "grad_norm": 4.320939064025879, "learning_rate": 7.02416278083934e-06, "loss": 0.5549, "mean_token_accuracy": 0.8308367729187012, "num_tokens": 63071243.0, "step": 1658 }, { "epoch": 0.21104185218165628, "ewc_loss": 0.041015625, "ewc_loss_diag": 3.039836883544922e-05, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 3.5926365852355957, "learning_rate": 7.028401865197118e-06, "loss": 0.5426, "mean_token_accuracy": 0.8363569378852844, "num_tokens": 63114496.0, "step": 1659 }, { "epoch": 0.21116906246024678, "ewc_loss": 0.052490234375, "ewc_loss_diag": 3.039836883544922e-05, "ewc_loss_parallel": 4.9591064453125e-05, "grad_norm": 4.408350944519043, "learning_rate": 7.032640949554897e-06, "loss": 0.4626, "mean_token_accuracy": 0.8599194884300232, "num_tokens": 63158111.0, "step": 1660 }, { "epoch": 0.2112962727388373, "ewc_loss": 0.04443359375, "ewc_loss_diag": 3.0517578125e-05, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 3.699826717376709, "learning_rate": 7.036880033912676e-06, "loss": 0.5546, "mean_token_accuracy": 0.8339506387710571, "num_tokens": 63198305.0, "step": 1661 }, { "epoch": 0.2114234830174278, "ewc_loss": 0.0537109375, "ewc_loss_diag": 3.0517578125e-05, "ewc_loss_parallel": 5.078315734863281e-05, "grad_norm": 4.430737495422363, "learning_rate": 7.041119118270454e-06, "loss": 0.5131, "mean_token_accuracy": 0.848196268081665, "num_tokens": 63235053.0, "step": 1662 }, { "epoch": 0.2115506932960183, "ewc_loss": 0.04736328125, "ewc_loss_diag": 3.0517578125e-05, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 3.8301620483398438, "learning_rate": 7.045358202628233e-06, "loss": 0.5552, "mean_token_accuracy": 0.8380652666091919, "num_tokens": 63275495.0, "step": 1663 }, { "epoch": 0.21167790357460883, "ewc_loss": 0.0517578125, "ewc_loss_diag": 3.0517578125e-05, "ewc_loss_parallel": 4.887580871582031e-05, "grad_norm": 4.329288482666016, "learning_rate": 7.0495972869860115e-06, "loss": 0.5543, "mean_token_accuracy": 0.836121678352356, "num_tokens": 63314338.0, "step": 1664 }, { "epoch": 0.21180511385319933, "ewc_loss": 0.0478515625, "ewc_loss_diag": 3.0517578125e-05, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 3.90006160736084, "learning_rate": 7.05383637134379e-06, "loss": 0.5031, "mean_token_accuracy": 0.8504971861839294, "num_tokens": 63348582.0, "step": 1665 }, { "epoch": 0.21193232413178986, "ewc_loss": 0.04833984375, "ewc_loss_diag": 3.0517578125e-05, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 4.173372268676758, "learning_rate": 7.058075455701568e-06, "loss": 0.5484, "mean_token_accuracy": 0.8315300941467285, "num_tokens": 63388496.0, "step": 1666 }, { "epoch": 0.21205953441038036, "ewc_loss": 0.04638671875, "ewc_loss_diag": 3.0517578125e-05, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 3.90447735786438, "learning_rate": 7.062314540059347e-06, "loss": 0.4964, "mean_token_accuracy": 0.855347752571106, "num_tokens": 63426216.0, "step": 1667 }, { "epoch": 0.21218674468897086, "ewc_loss": 0.04638671875, "ewc_loss_diag": 3.0517578125e-05, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 4.0126118659973145, "learning_rate": 7.066553624417126e-06, "loss": 0.4984, "mean_token_accuracy": 0.8500325679779053, "num_tokens": 63460627.0, "step": 1668 }, { "epoch": 0.2123139549675614, "ewc_loss": 0.044921875, "ewc_loss_diag": 3.0517578125e-05, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 3.8882627487182617, "learning_rate": 7.070792708774905e-06, "loss": 0.5293, "mean_token_accuracy": 0.841941237449646, "num_tokens": 63501911.0, "step": 1669 }, { "epoch": 0.2124411652461519, "ewc_loss": 0.04443359375, "ewc_loss_diag": 3.0517578125e-05, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 3.8803670406341553, "learning_rate": 7.075031793132684e-06, "loss": 0.5515, "mean_token_accuracy": 0.8385268449783325, "num_tokens": 63548764.0, "step": 1670 }, { "epoch": 0.2125683755247424, "ewc_loss": 0.0439453125, "ewc_loss_diag": 3.0517578125e-05, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 3.888275384902954, "learning_rate": 7.079270877490463e-06, "loss": 0.4994, "mean_token_accuracy": 0.8503167629241943, "num_tokens": 63583347.0, "step": 1671 }, { "epoch": 0.21269558580333292, "ewc_loss": 0.04296875, "ewc_loss_diag": 3.0517578125e-05, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 3.8030049800872803, "learning_rate": 7.083509961848242e-06, "loss": 0.5799, "mean_token_accuracy": 0.8287500143051147, "num_tokens": 63624115.0, "step": 1672 }, { "epoch": 0.21282279608192342, "ewc_loss": 0.043701171875, "ewc_loss_diag": 3.075599670410156e-05, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 3.8824996948242188, "learning_rate": 7.087749046206021e-06, "loss": 0.5001, "mean_token_accuracy": 0.8548098802566528, "num_tokens": 63660850.0, "step": 1673 }, { "epoch": 0.21295000636051392, "ewc_loss": 0.042236328125, "ewc_loss_diag": 3.075599670410156e-05, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 3.7387332916259766, "learning_rate": 7.091988130563799e-06, "loss": 0.5445, "mean_token_accuracy": 0.8357740640640259, "num_tokens": 63701530.0, "step": 1674 }, { "epoch": 0.21307721663910445, "ewc_loss": 0.043701171875, "ewc_loss_diag": 3.075599670410156e-05, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 3.8999552726745605, "learning_rate": 7.0962272149215775e-06, "loss": 0.5425, "mean_token_accuracy": 0.8365225195884705, "num_tokens": 63740400.0, "step": 1675 }, { "epoch": 0.21320442691769495, "ewc_loss": 0.04150390625, "ewc_loss_diag": 3.075599670410156e-05, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 3.6958673000335693, "learning_rate": 7.100466299279356e-06, "loss": 0.5232, "mean_token_accuracy": 0.8452388644218445, "num_tokens": 63775019.0, "step": 1676 }, { "epoch": 0.21333163719628545, "ewc_loss": 0.044189453125, "ewc_loss_diag": 3.075599670410156e-05, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 3.938483476638794, "learning_rate": 7.104705383637135e-06, "loss": 0.5472, "mean_token_accuracy": 0.8416372537612915, "num_tokens": 63813154.0, "step": 1677 }, { "epoch": 0.21345884747487598, "ewc_loss": 0.040771484375, "ewc_loss_diag": 3.075599670410156e-05, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 3.6276984214782715, "learning_rate": 7.108944467994914e-06, "loss": 0.5065, "mean_token_accuracy": 0.8494755029678345, "num_tokens": 63851982.0, "step": 1678 }, { "epoch": 0.21358605775346648, "ewc_loss": 0.04541015625, "ewc_loss_diag": 3.075599670410156e-05, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 4.02687931060791, "learning_rate": 7.113183552352693e-06, "loss": 0.5805, "mean_token_accuracy": 0.8246533870697021, "num_tokens": 63888983.0, "step": 1679 }, { "epoch": 0.21371326803205698, "ewc_loss": 0.040283203125, "ewc_loss_diag": 3.075599670410156e-05, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 3.5592825412750244, "learning_rate": 7.117422636710471e-06, "loss": 0.4636, "mean_token_accuracy": 0.8602023124694824, "num_tokens": 63927131.0, "step": 1680 }, { "epoch": 0.2138404783106475, "ewc_loss": 0.048095703125, "ewc_loss_diag": 3.075599670410156e-05, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 4.170498847961426, "learning_rate": 7.12166172106825e-06, "loss": 0.6063, "mean_token_accuracy": 0.8172829151153564, "num_tokens": 63962362.0, "step": 1681 }, { "epoch": 0.213967688589238, "ewc_loss": 0.040283203125, "ewc_loss_diag": 3.075599670410156e-05, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 3.4990220069885254, "learning_rate": 7.125900805426028e-06, "loss": 0.4726, "mean_token_accuracy": 0.8592357635498047, "num_tokens": 63997189.0, "step": 1682 }, { "epoch": 0.2140948988678285, "ewc_loss": 0.052001953125, "ewc_loss_diag": 3.075599670410156e-05, "ewc_loss_parallel": 4.887580871582031e-05, "grad_norm": 4.379997253417969, "learning_rate": 7.130139889783807e-06, "loss": 0.5237, "mean_token_accuracy": 0.8454742431640625, "num_tokens": 64032763.0, "step": 1683 }, { "epoch": 0.21422210914641904, "ewc_loss": 0.040771484375, "ewc_loss_diag": 3.075599670410156e-05, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 3.5008232593536377, "learning_rate": 7.134378974141586e-06, "loss": 0.5061, "mean_token_accuracy": 0.8473125696182251, "num_tokens": 64070176.0, "step": 1684 }, { "epoch": 0.21434931942500954, "ewc_loss": 0.055908203125, "ewc_loss_diag": 3.075599670410156e-05, "ewc_loss_parallel": 5.269050598144531e-05, "grad_norm": 4.635523796081543, "learning_rate": 7.138618058499365e-06, "loss": 0.5221, "mean_token_accuracy": 0.8494631052017212, "num_tokens": 64105632.0, "step": 1685 }, { "epoch": 0.21447652970360004, "ewc_loss": 0.04638671875, "ewc_loss_diag": 3.075599670410156e-05, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 3.672868251800537, "learning_rate": 7.1428571428571436e-06, "loss": 0.4852, "mean_token_accuracy": 0.857781171798706, "num_tokens": 64148729.0, "step": 1686 }, { "epoch": 0.21460373998219057, "ewc_loss": 0.057861328125, "ewc_loss_diag": 3.075599670410156e-05, "ewc_loss_parallel": 5.459785461425781e-05, "grad_norm": 4.762661457061768, "learning_rate": 7.1470962272149224e-06, "loss": 0.5216, "mean_token_accuracy": 0.8485890626907349, "num_tokens": 64189466.0, "step": 1687 }, { "epoch": 0.21473095026078107, "ewc_loss": 0.051025390625, "ewc_loss_diag": 3.075599670410156e-05, "ewc_loss_parallel": 4.792213439941406e-05, "grad_norm": 3.981567144393921, "learning_rate": 7.151335311572701e-06, "loss": 0.492, "mean_token_accuracy": 0.8557431697845459, "num_tokens": 64226325.0, "step": 1688 }, { "epoch": 0.21485816053937157, "ewc_loss": 0.056396484375, "ewc_loss_diag": 3.075599670410156e-05, "ewc_loss_parallel": 5.316734313964844e-05, "grad_norm": 4.638919353485107, "learning_rate": 7.155574395930479e-06, "loss": 0.5407, "mean_token_accuracy": 0.8392109274864197, "num_tokens": 64262178.0, "step": 1689 }, { "epoch": 0.2149853708179621, "ewc_loss": 0.0517578125, "ewc_loss_diag": 3.075599670410156e-05, "ewc_loss_parallel": 4.863739013671875e-05, "grad_norm": 4.120793342590332, "learning_rate": 7.159813480288258e-06, "loss": 0.5148, "mean_token_accuracy": 0.8443584442138672, "num_tokens": 64299933.0, "step": 1690 }, { "epoch": 0.2151125810965526, "ewc_loss": 0.052978515625, "ewc_loss_diag": 3.075599670410156e-05, "ewc_loss_parallel": 4.982948303222656e-05, "grad_norm": 4.607570171356201, "learning_rate": 7.164052564646037e-06, "loss": 0.4792, "mean_token_accuracy": 0.8561256527900696, "num_tokens": 64335836.0, "step": 1691 }, { "epoch": 0.21523979137514312, "ewc_loss": 0.051025390625, "ewc_loss_diag": 3.075599670410156e-05, "ewc_loss_parallel": 4.792213439941406e-05, "grad_norm": 4.128443717956543, "learning_rate": 7.168291649003816e-06, "loss": 0.5061, "mean_token_accuracy": 0.8549184203147888, "num_tokens": 64377190.0, "step": 1692 }, { "epoch": 0.21536700165373363, "ewc_loss": 0.04931640625, "ewc_loss_diag": 3.075599670410156e-05, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 4.244914531707764, "learning_rate": 7.172530733361595e-06, "loss": 0.5513, "mean_token_accuracy": 0.8364062309265137, "num_tokens": 64414693.0, "step": 1693 }, { "epoch": 0.21549421193232413, "ewc_loss": 0.049072265625, "ewc_loss_diag": 3.075599670410156e-05, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 4.029741287231445, "learning_rate": 7.176769817719374e-06, "loss": 0.4659, "mean_token_accuracy": 0.868690013885498, "num_tokens": 64454256.0, "step": 1694 }, { "epoch": 0.21562142221091465, "ewc_loss": 0.048095703125, "ewc_loss_diag": 3.075599670410156e-05, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 4.112407684326172, "learning_rate": 7.181008902077153e-06, "loss": 0.5185, "mean_token_accuracy": 0.8492873907089233, "num_tokens": 64491513.0, "step": 1695 }, { "epoch": 0.21574863248950515, "ewc_loss": 0.04736328125, "ewc_loss_diag": 3.075599670410156e-05, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 3.96557879447937, "learning_rate": 7.1852479864349316e-06, "loss": 0.5378, "mean_token_accuracy": 0.8437914848327637, "num_tokens": 64531349.0, "step": 1696 }, { "epoch": 0.21587584276809565, "ewc_loss": 0.046630859375, "ewc_loss_diag": 3.075599670410156e-05, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 4.027034759521484, "learning_rate": 7.189487070792709e-06, "loss": 0.5162, "mean_token_accuracy": 0.8485552072525024, "num_tokens": 64568007.0, "step": 1697 }, { "epoch": 0.21600305304668618, "ewc_loss": 0.04541015625, "ewc_loss_diag": 3.075599670410156e-05, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 3.89695143699646, "learning_rate": 7.193726155150488e-06, "loss": 0.5179, "mean_token_accuracy": 0.8470509052276611, "num_tokens": 64610287.0, "step": 1698 }, { "epoch": 0.21613026332527668, "ewc_loss": 0.045654296875, "ewc_loss_diag": 3.075599670410156e-05, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 3.9809749126434326, "learning_rate": 7.1979652395082665e-06, "loss": 0.5197, "mean_token_accuracy": 0.8462276458740234, "num_tokens": 64648757.0, "step": 1699 }, { "epoch": 0.21625747360386718, "ewc_loss": 0.043701171875, "ewc_loss_diag": 3.075599670410156e-05, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 3.852896213531494, "learning_rate": 7.202204323866045e-06, "loss": 0.5165, "mean_token_accuracy": 0.8475064635276794, "num_tokens": 64690137.0, "step": 1700 }, { "epoch": 0.2163846838824577, "ewc_loss": 0.04541015625, "ewc_loss_diag": 3.075599670410156e-05, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 3.957942008972168, "learning_rate": 7.206443408223824e-06, "loss": 0.5154, "mean_token_accuracy": 0.8449803590774536, "num_tokens": 64730218.0, "step": 1701 }, { "epoch": 0.2165118941610482, "ewc_loss": 0.043701171875, "ewc_loss_diag": 3.075599670410156e-05, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 3.7938358783721924, "learning_rate": 7.210682492581603e-06, "loss": 0.4928, "mean_token_accuracy": 0.8514840602874756, "num_tokens": 64762213.0, "step": 1702 }, { "epoch": 0.2166391044396387, "ewc_loss": 0.045166015625, "ewc_loss_diag": 3.075599670410156e-05, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 3.9675521850585938, "learning_rate": 7.214921576939382e-06, "loss": 0.5362, "mean_token_accuracy": 0.8365679979324341, "num_tokens": 64802471.0, "step": 1703 }, { "epoch": 0.21676631471822924, "ewc_loss": 0.04296875, "ewc_loss_diag": 3.075599670410156e-05, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 3.7392282485961914, "learning_rate": 7.21916066129716e-06, "loss": 0.4809, "mean_token_accuracy": 0.8544472455978394, "num_tokens": 64838139.0, "step": 1704 }, { "epoch": 0.21689352499681974, "ewc_loss": 0.0458984375, "ewc_loss_diag": 3.075599670410156e-05, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 4.014814376831055, "learning_rate": 7.223399745654939e-06, "loss": 0.5243, "mean_token_accuracy": 0.843778669834137, "num_tokens": 64877721.0, "step": 1705 }, { "epoch": 0.21702073527541024, "ewc_loss": 0.04150390625, "ewc_loss_diag": 3.075599670410156e-05, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 3.673523426055908, "learning_rate": 7.227638830012718e-06, "loss": 0.5364, "mean_token_accuracy": 0.8387609124183655, "num_tokens": 64916315.0, "step": 1706 }, { "epoch": 0.21714794555400077, "ewc_loss": 0.047607421875, "ewc_loss_diag": 3.075599670410156e-05, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 4.118847846984863, "learning_rate": 7.231877914370497e-06, "loss": 0.4974, "mean_token_accuracy": 0.8537585139274597, "num_tokens": 64951303.0, "step": 1707 }, { "epoch": 0.21727515583259127, "ewc_loss": 0.040771484375, "ewc_loss_diag": 3.075599670410156e-05, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 3.6108782291412354, "learning_rate": 7.236116998728276e-06, "loss": 0.5466, "mean_token_accuracy": 0.8360265493392944, "num_tokens": 64984677.0, "step": 1708 }, { "epoch": 0.21740236611118177, "ewc_loss": 0.0498046875, "ewc_loss_diag": 3.075599670410156e-05, "ewc_loss_parallel": 4.673004150390625e-05, "grad_norm": 4.258233070373535, "learning_rate": 7.2403560830860545e-06, "loss": 0.4464, "mean_token_accuracy": 0.8678998947143555, "num_tokens": 65021960.0, "step": 1709 }, { "epoch": 0.2175295763897723, "ewc_loss": 0.041259765625, "ewc_loss_diag": 3.075599670410156e-05, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 3.5640578269958496, "learning_rate": 7.244595167443833e-06, "loss": 0.5408, "mean_token_accuracy": 0.8385798931121826, "num_tokens": 65065576.0, "step": 1710 }, { "epoch": 0.2176567866683628, "ewc_loss": 0.05419921875, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 5.1021575927734375e-05, "grad_norm": 4.4648356437683105, "learning_rate": 7.248834251801612e-06, "loss": 0.5198, "mean_token_accuracy": 0.8513737916946411, "num_tokens": 65107188.0, "step": 1711 }, { "epoch": 0.2177839969469533, "ewc_loss": 0.042236328125, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 3.5838987827301025, "learning_rate": 7.2530733361593895e-06, "loss": 0.4787, "mean_token_accuracy": 0.8560617566108704, "num_tokens": 65145740.0, "step": 1712 }, { "epoch": 0.21791120722554383, "ewc_loss": 0.05712890625, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 5.3882598876953125e-05, "grad_norm": 4.699367523193359, "learning_rate": 7.257312420517168e-06, "loss": 0.4823, "mean_token_accuracy": 0.8579006791114807, "num_tokens": 65176572.0, "step": 1713 }, { "epoch": 0.21803841750413433, "ewc_loss": 0.04833984375, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 3.7845587730407715, "learning_rate": 7.261551504874947e-06, "loss": 0.5011, "mean_token_accuracy": 0.8484624624252319, "num_tokens": 65214839.0, "step": 1714 }, { "epoch": 0.21816562778272486, "ewc_loss": 0.058349609375, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 5.507469177246094e-05, "grad_norm": 4.763702869415283, "learning_rate": 7.265790589232726e-06, "loss": 0.5063, "mean_token_accuracy": 0.8533709645271301, "num_tokens": 65251438.0, "step": 1715 }, { "epoch": 0.21829283806131536, "ewc_loss": 0.05126953125, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 4.8160552978515625e-05, "grad_norm": 4.03361177444458, "learning_rate": 7.270029673590505e-06, "loss": 0.5363, "mean_token_accuracy": 0.8423641920089722, "num_tokens": 65294256.0, "step": 1716 }, { "epoch": 0.21842004833990586, "ewc_loss": 0.05615234375, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 5.2928924560546875e-05, "grad_norm": 4.613742828369141, "learning_rate": 7.274268757948284e-06, "loss": 0.5877, "mean_token_accuracy": 0.8289854526519775, "num_tokens": 65332917.0, "step": 1717 }, { "epoch": 0.2185472586184964, "ewc_loss": 0.0517578125, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 4.863739013671875e-05, "grad_norm": 4.140547275543213, "learning_rate": 7.278507842306063e-06, "loss": 0.4964, "mean_token_accuracy": 0.8526795506477356, "num_tokens": 65366678.0, "step": 1718 }, { "epoch": 0.2186744688970869, "ewc_loss": 0.0537109375, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 5.054473876953125e-05, "grad_norm": 4.397442817687988, "learning_rate": 7.282746926663842e-06, "loss": 0.5861, "mean_token_accuracy": 0.8279387950897217, "num_tokens": 65404797.0, "step": 1719 }, { "epoch": 0.2188016791756774, "ewc_loss": 0.05029296875, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 4.7206878662109375e-05, "grad_norm": 4.097477436065674, "learning_rate": 7.28698601102162e-06, "loss": 0.4953, "mean_token_accuracy": 0.8526671528816223, "num_tokens": 65441855.0, "step": 1720 }, { "epoch": 0.21892888945426792, "ewc_loss": 0.049560546875, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 4.207396984100342, "learning_rate": 7.291225095379399e-06, "loss": 0.5401, "mean_token_accuracy": 0.8436326384544373, "num_tokens": 65486895.0, "step": 1721 }, { "epoch": 0.21905609973285842, "ewc_loss": 0.048828125, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 4.03356409072876, "learning_rate": 7.2954641797371775e-06, "loss": 0.5206, "mean_token_accuracy": 0.8481876254081726, "num_tokens": 65524862.0, "step": 1722 }, { "epoch": 0.21918331001144892, "ewc_loss": 0.047607421875, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 4.082278251647949, "learning_rate": 7.299703264094956e-06, "loss": 0.5142, "mean_token_accuracy": 0.8463504910469055, "num_tokens": 65561538.0, "step": 1723 }, { "epoch": 0.21931052029003945, "ewc_loss": 0.04736328125, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 3.9887747764587402, "learning_rate": 7.303942348452735e-06, "loss": 0.5352, "mean_token_accuracy": 0.8406970500946045, "num_tokens": 65602376.0, "step": 1724 }, { "epoch": 0.21943773056862995, "ewc_loss": 0.04638671875, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 4.00937032699585, "learning_rate": 7.308181432810514e-06, "loss": 0.4802, "mean_token_accuracy": 0.854052722454071, "num_tokens": 65640817.0, "step": 1725 }, { "epoch": 0.21956494084722045, "ewc_loss": 0.045654296875, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 3.9219484329223633, "learning_rate": 7.312420517168293e-06, "loss": 0.5082, "mean_token_accuracy": 0.8439177870750427, "num_tokens": 65679238.0, "step": 1726 }, { "epoch": 0.21969215112581097, "ewc_loss": 0.04541015625, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 3.977414846420288, "learning_rate": 7.31665960152607e-06, "loss": 0.5311, "mean_token_accuracy": 0.8376713395118713, "num_tokens": 65720141.0, "step": 1727 }, { "epoch": 0.21981936140440148, "ewc_loss": 0.043701171875, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 3.8801729679107666, "learning_rate": 7.320898685883849e-06, "loss": 0.5027, "mean_token_accuracy": 0.847457766532898, "num_tokens": 65758511.0, "step": 1728 }, { "epoch": 0.21994657168299198, "ewc_loss": 0.044677734375, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 3.979081869125366, "learning_rate": 7.325137770241628e-06, "loss": 0.6253, "mean_token_accuracy": 0.815971851348877, "num_tokens": 65798485.0, "step": 1729 }, { "epoch": 0.2200737819615825, "ewc_loss": 0.043212890625, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 3.815338611602783, "learning_rate": 7.329376854599407e-06, "loss": 0.5296, "mean_token_accuracy": 0.8413313627243042, "num_tokens": 65840414.0, "step": 1730 }, { "epoch": 0.220200992240173, "ewc_loss": 0.04541015625, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 4.0050530433654785, "learning_rate": 7.333615938957186e-06, "loss": 0.5112, "mean_token_accuracy": 0.8483613729476929, "num_tokens": 65878425.0, "step": 1731 }, { "epoch": 0.2203282025187635, "ewc_loss": 0.04296875, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 3.759084939956665, "learning_rate": 7.337855023314965e-06, "loss": 0.5122, "mean_token_accuracy": 0.844322919845581, "num_tokens": 65916958.0, "step": 1732 }, { "epoch": 0.22045541279735403, "ewc_loss": 0.046142578125, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 4.06544828414917, "learning_rate": 7.3420941076727435e-06, "loss": 0.5662, "mean_token_accuracy": 0.8292626142501831, "num_tokens": 65958511.0, "step": 1733 }, { "epoch": 0.22058262307594453, "ewc_loss": 0.042236328125, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 3.6896679401397705, "learning_rate": 7.346333192030522e-06, "loss": 0.5125, "mean_token_accuracy": 0.8487367630004883, "num_tokens": 65994735.0, "step": 1734 }, { "epoch": 0.22070983335453503, "ewc_loss": 0.048828125, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 4.183917999267578, "learning_rate": 7.3505722763883004e-06, "loss": 0.5736, "mean_token_accuracy": 0.828575849533081, "num_tokens": 66036017.0, "step": 1735 }, { "epoch": 0.22083704363312556, "ewc_loss": 0.0419921875, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 3.6156511306762695, "learning_rate": 7.354811360746079e-06, "loss": 0.515, "mean_token_accuracy": 0.8433523178100586, "num_tokens": 66073203.0, "step": 1736 }, { "epoch": 0.22096425391171606, "ewc_loss": 0.051513671875, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 4.839897155761719e-05, "grad_norm": 4.3683857917785645, "learning_rate": 7.359050445103858e-06, "loss": 0.5417, "mean_token_accuracy": 0.8411810398101807, "num_tokens": 66109623.0, "step": 1737 }, { "epoch": 0.22109146419030656, "ewc_loss": 0.041259765625, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 3.5652124881744385, "learning_rate": 7.363289529461637e-06, "loss": 0.5082, "mean_token_accuracy": 0.8468592166900635, "num_tokens": 66148426.0, "step": 1738 }, { "epoch": 0.2212186744688971, "ewc_loss": 0.056640625, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 5.340576171875e-05, "grad_norm": 4.639998912811279, "learning_rate": 7.367528613819416e-06, "loss": 0.4814, "mean_token_accuracy": 0.859647274017334, "num_tokens": 66189164.0, "step": 1739 }, { "epoch": 0.2213458847474876, "ewc_loss": 0.044677734375, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 3.654304265975952, "learning_rate": 7.371767698177195e-06, "loss": 0.4863, "mean_token_accuracy": 0.852850615978241, "num_tokens": 66223523.0, "step": 1740 }, { "epoch": 0.22147309502607812, "ewc_loss": 0.06005859375, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 5.698204040527344e-05, "grad_norm": 4.874337196350098, "learning_rate": 7.376006782534974e-06, "loss": 0.4917, "mean_token_accuracy": 0.8579589128494263, "num_tokens": 66256819.0, "step": 1741 }, { "epoch": 0.22160030530466862, "ewc_loss": 0.05126953125, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 4.8160552978515625e-05, "grad_norm": 3.946474075317383, "learning_rate": 7.380245866892751e-06, "loss": 0.5122, "mean_token_accuracy": 0.8477595448493958, "num_tokens": 66291619.0, "step": 1742 }, { "epoch": 0.22172751558325912, "ewc_loss": 0.060546875, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 5.745887756347656e-05, "grad_norm": 4.8456244468688965, "learning_rate": 7.38448495125053e-06, "loss": 0.5247, "mean_token_accuracy": 0.8478174805641174, "num_tokens": 66334034.0, "step": 1743 }, { "epoch": 0.22185472586184965, "ewc_loss": 0.05419921875, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 5.1021575927734375e-05, "grad_norm": 4.191059589385986, "learning_rate": 7.388724035608309e-06, "loss": 0.5058, "mean_token_accuracy": 0.8543967008590698, "num_tokens": 66368786.0, "step": 1744 }, { "epoch": 0.22198193614044015, "ewc_loss": 0.05615234375, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 5.2928924560546875e-05, "grad_norm": 4.637209892272949, "learning_rate": 7.392963119966088e-06, "loss": 0.5002, "mean_token_accuracy": 0.8528158664703369, "num_tokens": 66400034.0, "step": 1745 }, { "epoch": 0.22210914641903065, "ewc_loss": 0.052978515625, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 4.982948303222656e-05, "grad_norm": 4.222482681274414, "learning_rate": 7.3972022043238665e-06, "loss": 0.5326, "mean_token_accuracy": 0.8450194597244263, "num_tokens": 66436954.0, "step": 1746 }, { "epoch": 0.22223635669762118, "ewc_loss": 0.053955078125, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 5.078315734863281e-05, "grad_norm": 4.427250862121582, "learning_rate": 7.401441288681645e-06, "loss": 0.4996, "mean_token_accuracy": 0.854367733001709, "num_tokens": 66476360.0, "step": 1747 }, { "epoch": 0.22236356697621168, "ewc_loss": 0.051025390625, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 4.792213439941406e-05, "grad_norm": 4.144738674163818, "learning_rate": 7.405680373039424e-06, "loss": 0.4755, "mean_token_accuracy": 0.8629705905914307, "num_tokens": 66516205.0, "step": 1748 }, { "epoch": 0.22249077725480218, "ewc_loss": 0.050537109375, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 4.744529724121094e-05, "grad_norm": 4.262657165527344, "learning_rate": 7.409919457397203e-06, "loss": 0.5393, "mean_token_accuracy": 0.8386399745941162, "num_tokens": 66548732.0, "step": 1749 }, { "epoch": 0.2226179875333927, "ewc_loss": 0.04931640625, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 4.0695343017578125, "learning_rate": 7.414158541754981e-06, "loss": 0.5768, "mean_token_accuracy": 0.8267797827720642, "num_tokens": 66587888.0, "step": 1750 }, { "epoch": 0.2227451978119832, "ewc_loss": 0.048583984375, "ewc_loss_diag": 3.123283386230469e-05, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 4.167117118835449, "learning_rate": 7.41839762611276e-06, "loss": 0.5109, "mean_token_accuracy": 0.8571512699127197, "num_tokens": 66617846.0, "step": 1751 }, { "epoch": 0.2228724080905737, "ewc_loss": 0.047607421875, "ewc_loss_diag": 3.123283386230469e-05, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 3.9817442893981934, "learning_rate": 7.422636710470539e-06, "loss": 0.5241, "mean_token_accuracy": 0.8445862531661987, "num_tokens": 66659703.0, "step": 1752 }, { "epoch": 0.22299961836916424, "ewc_loss": 0.048583984375, "ewc_loss_diag": 3.123283386230469e-05, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 4.120870590209961, "learning_rate": 7.426875794828318e-06, "loss": 0.5382, "mean_token_accuracy": 0.8439548015594482, "num_tokens": 66697604.0, "step": 1753 }, { "epoch": 0.22312682864775474, "ewc_loss": 0.045654296875, "ewc_loss_diag": 3.123283386230469e-05, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 3.9127564430236816, "learning_rate": 7.431114879186097e-06, "loss": 0.5627, "mean_token_accuracy": 0.832889974117279, "num_tokens": 66736839.0, "step": 1754 }, { "epoch": 0.22325403892634524, "ewc_loss": 0.04833984375, "ewc_loss_diag": 3.123283386230469e-05, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 4.115421295166016, "learning_rate": 7.435353963543876e-06, "loss": 0.5257, "mean_token_accuracy": 0.8460589647293091, "num_tokens": 66777477.0, "step": 1755 }, { "epoch": 0.22338124920493577, "ewc_loss": 0.044189453125, "ewc_loss_diag": 3.123283386230469e-05, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 3.8403432369232178, "learning_rate": 7.4395930479016545e-06, "loss": 0.4701, "mean_token_accuracy": 0.8561186790466309, "num_tokens": 66816294.0, "step": 1756 }, { "epoch": 0.22350845948352627, "ewc_loss": 0.049072265625, "ewc_loss_diag": 3.123283386230469e-05, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 4.165693759918213, "learning_rate": 7.443832132259432e-06, "loss": 0.5403, "mean_token_accuracy": 0.8383238315582275, "num_tokens": 66852740.0, "step": 1757 }, { "epoch": 0.22363566976211677, "ewc_loss": 0.043701171875, "ewc_loss_diag": 3.123283386230469e-05, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 3.7802438735961914, "learning_rate": 7.4480712166172105e-06, "loss": 0.5137, "mean_token_accuracy": 0.8444162607192993, "num_tokens": 66890511.0, "step": 1758 }, { "epoch": 0.2237628800407073, "ewc_loss": 0.049560546875, "ewc_loss_diag": 3.123283386230469e-05, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 4.251399040222168, "learning_rate": 7.4523103009749894e-06, "loss": 0.4795, "mean_token_accuracy": 0.8554304838180542, "num_tokens": 66924724.0, "step": 1759 }, { "epoch": 0.2238900903192978, "ewc_loss": 0.043212890625, "ewc_loss_diag": 3.123283386230469e-05, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 3.708876132965088, "learning_rate": 7.456549385332768e-06, "loss": 0.5531, "mean_token_accuracy": 0.8356380462646484, "num_tokens": 66964102.0, "step": 1760 }, { "epoch": 0.2240173005978883, "ewc_loss": 0.052978515625, "ewc_loss_diag": 3.123283386230469e-05, "ewc_loss_parallel": 4.982948303222656e-05, "grad_norm": 4.390591621398926, "learning_rate": 7.460788469690547e-06, "loss": 0.4457, "mean_token_accuracy": 0.8670365214347839, "num_tokens": 67003234.0, "step": 1761 }, { "epoch": 0.22414451087647883, "ewc_loss": 0.04296875, "ewc_loss_diag": 3.123283386230469e-05, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 3.6796483993530273, "learning_rate": 7.465027554048326e-06, "loss": 0.5925, "mean_token_accuracy": 0.8214872479438782, "num_tokens": 67045952.0, "step": 1762 }, { "epoch": 0.22427172115506933, "ewc_loss": 0.0556640625, "ewc_loss_diag": 3.123283386230469e-05, "ewc_loss_parallel": 5.245208740234375e-05, "grad_norm": 4.57459831237793, "learning_rate": 7.469266638406105e-06, "loss": 0.5898, "mean_token_accuracy": 0.8285808563232422, "num_tokens": 67090169.0, "step": 1763 }, { "epoch": 0.22439893143365983, "ewc_loss": 0.046142578125, "ewc_loss_diag": 3.123283386230469e-05, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 3.7322216033935547, "learning_rate": 7.473505722763884e-06, "loss": 0.4992, "mean_token_accuracy": 0.8500716686248779, "num_tokens": 67125075.0, "step": 1764 }, { "epoch": 0.22452614171225035, "ewc_loss": 0.05859375, "ewc_loss_diag": 3.123283386230469e-05, "ewc_loss_parallel": 5.53131103515625e-05, "grad_norm": 4.73777437210083, "learning_rate": 7.477744807121662e-06, "loss": 0.5562, "mean_token_accuracy": 0.8381705284118652, "num_tokens": 67162892.0, "step": 1765 }, { "epoch": 0.22465335199084085, "ewc_loss": 0.050048828125, "ewc_loss_diag": 3.123283386230469e-05, "ewc_loss_parallel": 4.696846008300781e-05, "grad_norm": 3.905625581741333, "learning_rate": 7.481983891479441e-06, "loss": 0.5302, "mean_token_accuracy": 0.8445742130279541, "num_tokens": 67198324.0, "step": 1766 }, { "epoch": 0.22478056226943138, "ewc_loss": 0.05908203125, "ewc_loss_diag": 3.147125244140625e-05, "ewc_loss_parallel": 5.5789947509765625e-05, "grad_norm": 4.736145973205566, "learning_rate": 7.48622297583722e-06, "loss": 0.5217, "mean_token_accuracy": 0.8443480134010315, "num_tokens": 67239548.0, "step": 1767 }, { "epoch": 0.22490777254802188, "ewc_loss": 0.05126953125, "ewc_loss_diag": 3.147125244140625e-05, "ewc_loss_parallel": 4.8160552978515625e-05, "grad_norm": 4.0772199630737305, "learning_rate": 7.4904620601949985e-06, "loss": 0.478, "mean_token_accuracy": 0.855569064617157, "num_tokens": 67273373.0, "step": 1768 }, { "epoch": 0.22503498282661238, "ewc_loss": 0.056640625, "ewc_loss_diag": 3.147125244140625e-05, "ewc_loss_parallel": 5.340576171875e-05, "grad_norm": 4.595829963684082, "learning_rate": 7.4947011445527774e-06, "loss": 0.5635, "mean_token_accuracy": 0.835393488407135, "num_tokens": 67307074.0, "step": 1769 }, { "epoch": 0.2251621931052029, "ewc_loss": 0.05126953125, "ewc_loss_diag": 3.147125244140625e-05, "ewc_loss_parallel": 4.8160552978515625e-05, "grad_norm": 4.1466498374938965, "learning_rate": 7.498940228910556e-06, "loss": 0.5349, "mean_token_accuracy": 0.8406638503074646, "num_tokens": 67347905.0, "step": 1770 }, { "epoch": 0.2252894033837934, "ewc_loss": 0.0537109375, "ewc_loss_diag": 3.147125244140625e-05, "ewc_loss_parallel": 5.054473876953125e-05, "grad_norm": 4.383006572723389, "learning_rate": 7.503179313268335e-06, "loss": 0.5086, "mean_token_accuracy": 0.8470654487609863, "num_tokens": 67382594.0, "step": 1771 }, { "epoch": 0.2254166136623839, "ewc_loss": 0.05078125, "ewc_loss_diag": 3.147125244140625e-05, "ewc_loss_parallel": 4.76837158203125e-05, "grad_norm": 4.12793493270874, "learning_rate": 7.507418397626114e-06, "loss": 0.5518, "mean_token_accuracy": 0.8349921107292175, "num_tokens": 67422867.0, "step": 1772 }, { "epoch": 0.22554382394097444, "ewc_loss": 0.050048828125, "ewc_loss_diag": 3.147125244140625e-05, "ewc_loss_parallel": 4.696846008300781e-05, "grad_norm": 4.2098388671875, "learning_rate": 7.511657481983891e-06, "loss": 0.5381, "mean_token_accuracy": 0.8361701369285583, "num_tokens": 67463524.0, "step": 1773 }, { "epoch": 0.22567103421956494, "ewc_loss": 0.049560546875, "ewc_loss_diag": 3.147125244140625e-05, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 4.096466064453125, "learning_rate": 7.51589656634167e-06, "loss": 0.4927, "mean_token_accuracy": 0.8470867276191711, "num_tokens": 67499224.0, "step": 1774 }, { "epoch": 0.22579824449815544, "ewc_loss": 0.0478515625, "ewc_loss_diag": 3.147125244140625e-05, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 4.080566883087158, "learning_rate": 7.520135650699449e-06, "loss": 0.5205, "mean_token_accuracy": 0.8426970839500427, "num_tokens": 67536745.0, "step": 1775 }, { "epoch": 0.22592545477674597, "ewc_loss": 0.048828125, "ewc_loss_diag": 3.147125244140625e-05, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 4.087054252624512, "learning_rate": 7.524374735057228e-06, "loss": 0.5471, "mean_token_accuracy": 0.8408570885658264, "num_tokens": 67572725.0, "step": 1776 }, { "epoch": 0.22605266505533647, "ewc_loss": 0.04638671875, "ewc_loss_diag": 3.147125244140625e-05, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 3.9948525428771973, "learning_rate": 7.528613819415007e-06, "loss": 0.4987, "mean_token_accuracy": 0.8553314208984375, "num_tokens": 67607588.0, "step": 1777 }, { "epoch": 0.22617987533392697, "ewc_loss": 0.048095703125, "ewc_loss_diag": 3.1948089599609375e-05, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 4.073627948760986, "learning_rate": 7.532852903772786e-06, "loss": 0.5356, "mean_token_accuracy": 0.8409148454666138, "num_tokens": 67643064.0, "step": 1778 }, { "epoch": 0.2263070856125175, "ewc_loss": 0.04541015625, "ewc_loss_diag": 3.1948089599609375e-05, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 3.9362354278564453, "learning_rate": 7.537091988130565e-06, "loss": 0.5348, "mean_token_accuracy": 0.8376486897468567, "num_tokens": 67677985.0, "step": 1779 }, { "epoch": 0.226434295891108, "ewc_loss": 0.047607421875, "ewc_loss_diag": 3.1948089599609375e-05, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 4.078185558319092, "learning_rate": 7.541331072488343e-06, "loss": 0.5244, "mean_token_accuracy": 0.8444614410400391, "num_tokens": 67713814.0, "step": 1780 }, { "epoch": 0.2265615061696985, "ewc_loss": 0.044921875, "ewc_loss_diag": 3.1948089599609375e-05, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 3.8833086490631104, "learning_rate": 7.5455701568461215e-06, "loss": 0.5218, "mean_token_accuracy": 0.8434211611747742, "num_tokens": 67752806.0, "step": 1781 }, { "epoch": 0.22668871644828903, "ewc_loss": 0.048095703125, "ewc_loss_diag": 3.1948089599609375e-05, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 4.106468200683594, "learning_rate": 7.5498092412039e-06, "loss": 0.5087, "mean_token_accuracy": 0.8501055240631104, "num_tokens": 67792804.0, "step": 1782 }, { "epoch": 0.22681592672687953, "ewc_loss": 0.0439453125, "ewc_loss_diag": 3.1948089599609375e-05, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 3.814998149871826, "learning_rate": 7.554048325561679e-06, "loss": 0.4691, "mean_token_accuracy": 0.8587828278541565, "num_tokens": 67831361.0, "step": 1783 }, { "epoch": 0.22694313700547003, "ewc_loss": 0.049072265625, "ewc_loss_diag": 3.1948089599609375e-05, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 4.198994159698486, "learning_rate": 7.558287409919458e-06, "loss": 0.5639, "mean_token_accuracy": 0.8379519581794739, "num_tokens": 67868981.0, "step": 1784 }, { "epoch": 0.22707034728406056, "ewc_loss": 0.04345703125, "ewc_loss_diag": 3.1948089599609375e-05, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 3.750227689743042, "learning_rate": 7.562526494277237e-06, "loss": 0.5353, "mean_token_accuracy": 0.8372664451599121, "num_tokens": 67905501.0, "step": 1785 }, { "epoch": 0.22719755756265106, "ewc_loss": 0.05224609375, "ewc_loss_diag": 3.218650817871094e-05, "ewc_loss_parallel": 4.9114227294921875e-05, "grad_norm": 4.315673351287842, "learning_rate": 7.566765578635016e-06, "loss": 0.5134, "mean_token_accuracy": 0.8508355021476746, "num_tokens": 67946001.0, "step": 1786 }, { "epoch": 0.22732476784124156, "ewc_loss": 0.04345703125, "ewc_loss_diag": 3.218650817871094e-05, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 3.6668100357055664, "learning_rate": 7.571004662992795e-06, "loss": 0.5265, "mean_token_accuracy": 0.8441457748413086, "num_tokens": 67983868.0, "step": 1787 }, { "epoch": 0.2274519781198321, "ewc_loss": 0.055908203125, "ewc_loss_diag": 3.218650817871094e-05, "ewc_loss_parallel": 5.269050598144531e-05, "grad_norm": 4.538135051727295, "learning_rate": 7.575243747350573e-06, "loss": 0.5957, "mean_token_accuracy": 0.8263826370239258, "num_tokens": 68023664.0, "step": 1788 }, { "epoch": 0.2275791883984226, "ewc_loss": 0.044677734375, "ewc_loss_diag": 3.218650817871094e-05, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 3.6266226768493652, "learning_rate": 7.579482831708351e-06, "loss": 0.522, "mean_token_accuracy": 0.8466123342514038, "num_tokens": 68063950.0, "step": 1789 }, { "epoch": 0.2277063986770131, "ewc_loss": 0.06005859375, "ewc_loss_diag": 3.218650817871094e-05, "ewc_loss_parallel": 5.698204040527344e-05, "grad_norm": 4.842607021331787, "learning_rate": 7.58372191606613e-06, "loss": 0.5344, "mean_token_accuracy": 0.8455892205238342, "num_tokens": 68097812.0, "step": 1790 }, { "epoch": 0.22783360895560362, "ewc_loss": 0.04931640625, "ewc_loss_diag": 3.218650817871094e-05, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 3.7725789546966553, "learning_rate": 7.587961000423909e-06, "loss": 0.5253, "mean_token_accuracy": 0.8445766568183899, "num_tokens": 68135219.0, "step": 1791 }, { "epoch": 0.22796081923419412, "ewc_loss": 0.0654296875, "ewc_loss_diag": 3.218650817871094e-05, "ewc_loss_parallel": 6.198883056640625e-05, "grad_norm": 5.074655532836914, "learning_rate": 7.5922000847816875e-06, "loss": 0.5945, "mean_token_accuracy": 0.8326349258422852, "num_tokens": 68169757.0, "step": 1792 }, { "epoch": 0.22808802951278465, "ewc_loss": 0.05517578125, "ewc_loss_diag": 3.218650817871094e-05, "ewc_loss_parallel": 5.1975250244140625e-05, "grad_norm": 4.141324043273926, "learning_rate": 7.5964391691394664e-06, "loss": 0.5285, "mean_token_accuracy": 0.8453625440597534, "num_tokens": 68204807.0, "step": 1793 }, { "epoch": 0.22821523979137515, "ewc_loss": 0.06396484375, "ewc_loss_diag": 3.218650817871094e-05, "ewc_loss_parallel": 6.079673767089844e-05, "grad_norm": 4.967891216278076, "learning_rate": 7.600678253497245e-06, "loss": 0.5331, "mean_token_accuracy": 0.8488475680351257, "num_tokens": 68245037.0, "step": 1794 }, { "epoch": 0.22834245006996565, "ewc_loss": 0.058349609375, "ewc_loss_diag": 3.218650817871094e-05, "ewc_loss_parallel": 5.507469177246094e-05, "grad_norm": 4.352451324462891, "learning_rate": 7.604917337855023e-06, "loss": 0.4991, "mean_token_accuracy": 0.8537841439247131, "num_tokens": 68281938.0, "step": 1795 }, { "epoch": 0.22846966034855618, "ewc_loss": 0.058349609375, "ewc_loss_diag": 3.218650817871094e-05, "ewc_loss_parallel": 5.507469177246094e-05, "grad_norm": 4.703710079193115, "learning_rate": 7.609156422212802e-06, "loss": 0.4813, "mean_token_accuracy": 0.8591258525848389, "num_tokens": 68319363.0, "step": 1796 }, { "epoch": 0.22859687062714668, "ewc_loss": 0.0556640625, "ewc_loss_diag": 3.218650817871094e-05, "ewc_loss_parallel": 5.245208740234375e-05, "grad_norm": 4.33394193649292, "learning_rate": 7.613395506570581e-06, "loss": 0.5035, "mean_token_accuracy": 0.8533545732498169, "num_tokens": 68357428.0, "step": 1797 }, { "epoch": 0.22872408090573718, "ewc_loss": 0.0556640625, "ewc_loss_diag": 3.218650817871094e-05, "ewc_loss_parallel": 5.245208740234375e-05, "grad_norm": 4.48668098449707, "learning_rate": 7.61763459092836e-06, "loss": 0.5084, "mean_token_accuracy": 0.8513601422309875, "num_tokens": 68390558.0, "step": 1798 }, { "epoch": 0.2288512911843277, "ewc_loss": 0.052734375, "ewc_loss_diag": 3.24249267578125e-05, "ewc_loss_parallel": 4.9591064453125e-05, "grad_norm": 4.244648456573486, "learning_rate": 7.621873675286139e-06, "loss": 0.541, "mean_token_accuracy": 0.8415495157241821, "num_tokens": 68432250.0, "step": 1799 }, { "epoch": 0.2289785014629182, "ewc_loss": 0.052734375, "ewc_loss_diag": 3.24249267578125e-05, "ewc_loss_parallel": 4.9591064453125e-05, "grad_norm": 4.32891845703125, "learning_rate": 7.626112759643918e-06, "loss": 0.5409, "mean_token_accuracy": 0.8397626876831055, "num_tokens": 68469427.0, "step": 1800 }, { "epoch": 0.2291057117415087, "ewc_loss": 0.05078125, "ewc_loss_diag": 3.24249267578125e-05, "ewc_loss_parallel": 4.76837158203125e-05, "grad_norm": 4.155282497406006, "learning_rate": 7.630351844001696e-06, "loss": 0.516, "mean_token_accuracy": 0.8475961685180664, "num_tokens": 68503442.0, "step": 1801 }, { "epoch": 0.22923292202009923, "ewc_loss": 0.051025390625, "ewc_loss_diag": 3.24249267578125e-05, "ewc_loss_parallel": 4.792213439941406e-05, "grad_norm": 4.240562915802002, "learning_rate": 7.634590928359475e-06, "loss": 0.5032, "mean_token_accuracy": 0.8534728288650513, "num_tokens": 68540836.0, "step": 1802 }, { "epoch": 0.22936013229868973, "ewc_loss": 0.049072265625, "ewc_loss_diag": 3.24249267578125e-05, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 4.065019130706787, "learning_rate": 7.638830012717254e-06, "loss": 0.535, "mean_token_accuracy": 0.8394037485122681, "num_tokens": 68581118.0, "step": 1803 }, { "epoch": 0.22948734257728023, "ewc_loss": 0.0498046875, "ewc_loss_diag": 3.24249267578125e-05, "ewc_loss_parallel": 4.673004150390625e-05, "grad_norm": 4.1904473304748535, "learning_rate": 7.643069097075032e-06, "loss": 0.504, "mean_token_accuracy": 0.8490321040153503, "num_tokens": 68620988.0, "step": 1804 }, { "epoch": 0.22961455285587076, "ewc_loss": 0.0478515625, "ewc_loss_diag": 3.24249267578125e-05, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 3.9866902828216553, "learning_rate": 7.647308181432811e-06, "loss": 0.4767, "mean_token_accuracy": 0.8566172122955322, "num_tokens": 68662141.0, "step": 1805 }, { "epoch": 0.22974176313446126, "ewc_loss": 0.0498046875, "ewc_loss_diag": 3.24249267578125e-05, "ewc_loss_parallel": 4.673004150390625e-05, "grad_norm": 4.215173721313477, "learning_rate": 7.65154726579059e-06, "loss": 0.5641, "mean_token_accuracy": 0.8329132795333862, "num_tokens": 68700922.0, "step": 1806 }, { "epoch": 0.22986897341305176, "ewc_loss": 0.0458984375, "ewc_loss_diag": 3.266334533691406e-05, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 3.916444778442383, "learning_rate": 7.655786350148369e-06, "loss": 0.5369, "mean_token_accuracy": 0.839917778968811, "num_tokens": 68741137.0, "step": 1807 }, { "epoch": 0.2299961836916423, "ewc_loss": 0.05078125, "ewc_loss_diag": 3.266334533691406e-05, "ewc_loss_parallel": 4.76837158203125e-05, "grad_norm": 4.259430408477783, "learning_rate": 7.660025434506148e-06, "loss": 0.5434, "mean_token_accuracy": 0.8405338525772095, "num_tokens": 68785679.0, "step": 1808 }, { "epoch": 0.2301233939702328, "ewc_loss": 0.044921875, "ewc_loss_diag": 3.266334533691406e-05, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 3.8366949558258057, "learning_rate": 7.664264518863927e-06, "loss": 0.518, "mean_token_accuracy": 0.8458973169326782, "num_tokens": 68823482.0, "step": 1809 }, { "epoch": 0.2302506042488233, "ewc_loss": 0.052734375, "ewc_loss_diag": 3.266334533691406e-05, "ewc_loss_parallel": 4.9591064453125e-05, "grad_norm": 4.386960506439209, "learning_rate": 7.668503603221706e-06, "loss": 0.5821, "mean_token_accuracy": 0.8266299366950989, "num_tokens": 68863854.0, "step": 1810 }, { "epoch": 0.23037781452741382, "ewc_loss": 0.044677734375, "ewc_loss_diag": 3.266334533691406e-05, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 3.7554619312286377, "learning_rate": 7.672742687579483e-06, "loss": 0.4684, "mean_token_accuracy": 0.8582122921943665, "num_tokens": 68901341.0, "step": 1811 }, { "epoch": 0.23050502480600432, "ewc_loss": 0.055908203125, "ewc_loss_diag": 3.266334533691406e-05, "ewc_loss_parallel": 5.269050598144531e-05, "grad_norm": 4.585894584655762, "learning_rate": 7.676981771937262e-06, "loss": 0.4597, "mean_token_accuracy": 0.8611700534820557, "num_tokens": 68931774.0, "step": 1812 }, { "epoch": 0.23063223508459482, "ewc_loss": 0.04541015625, "ewc_loss_diag": 3.266334533691406e-05, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 3.7289822101593018, "learning_rate": 7.68122085629504e-06, "loss": 0.4888, "mean_token_accuracy": 0.8530311584472656, "num_tokens": 68972114.0, "step": 1813 }, { "epoch": 0.23075944536318535, "ewc_loss": 0.06005859375, "ewc_loss_diag": 3.266334533691406e-05, "ewc_loss_parallel": 5.698204040527344e-05, "grad_norm": 4.842380523681641, "learning_rate": 7.68545994065282e-06, "loss": 0.526, "mean_token_accuracy": 0.8447051048278809, "num_tokens": 69007272.0, "step": 1814 }, { "epoch": 0.23088665564177585, "ewc_loss": 0.0498046875, "ewc_loss_diag": 3.266334533691406e-05, "ewc_loss_parallel": 4.673004150390625e-05, "grad_norm": 3.863132953643799, "learning_rate": 7.689699025010598e-06, "loss": 0.512, "mean_token_accuracy": 0.8489639163017273, "num_tokens": 69050011.0, "step": 1815 }, { "epoch": 0.23101386592036638, "ewc_loss": 0.0634765625, "ewc_loss_diag": 3.266334533691406e-05, "ewc_loss_parallel": 6.031990051269531e-05, "grad_norm": 4.997368335723877, "learning_rate": 7.693938109368377e-06, "loss": 0.5467, "mean_token_accuracy": 0.8397694826126099, "num_tokens": 69087914.0, "step": 1816 }, { "epoch": 0.23114107619895688, "ewc_loss": 0.05419921875, "ewc_loss_diag": 3.266334533691406e-05, "ewc_loss_parallel": 5.1021575927734375e-05, "grad_norm": 4.176501274108887, "learning_rate": 7.698177193726156e-06, "loss": 0.5259, "mean_token_accuracy": 0.8440712690353394, "num_tokens": 69125301.0, "step": 1817 }, { "epoch": 0.23126828647754738, "ewc_loss": 0.061279296875, "ewc_loss_diag": 3.266334533691406e-05, "ewc_loss_parallel": 5.817413330078125e-05, "grad_norm": 4.864847183227539, "learning_rate": 7.702416278083933e-06, "loss": 0.5469, "mean_token_accuracy": 0.8413260579109192, "num_tokens": 69159268.0, "step": 1818 }, { "epoch": 0.2313954967561379, "ewc_loss": 0.05615234375, "ewc_loss_diag": 3.2901763916015625e-05, "ewc_loss_parallel": 5.269050598144531e-05, "grad_norm": 4.304388046264648, "learning_rate": 7.706655362441712e-06, "loss": 0.5224, "mean_token_accuracy": 0.8471199870109558, "num_tokens": 69202207.0, "step": 1819 }, { "epoch": 0.2315227070347284, "ewc_loss": 0.056640625, "ewc_loss_diag": 3.2901763916015625e-05, "ewc_loss_parallel": 5.340576171875e-05, "grad_norm": 4.625212669372559, "learning_rate": 7.710894446799491e-06, "loss": 0.4985, "mean_token_accuracy": 0.8522616624832153, "num_tokens": 69234757.0, "step": 1820 }, { "epoch": 0.2316499173133189, "ewc_loss": 0.05419921875, "ewc_loss_diag": 3.2901763916015625e-05, "ewc_loss_parallel": 5.078315734863281e-05, "grad_norm": 4.296210289001465, "learning_rate": 7.71513353115727e-06, "loss": 0.4993, "mean_token_accuracy": 0.8576931953430176, "num_tokens": 69274995.0, "step": 1821 }, { "epoch": 0.23177712759190944, "ewc_loss": 0.0546875, "ewc_loss_diag": 3.2901763916015625e-05, "ewc_loss_parallel": 5.125999450683594e-05, "grad_norm": 4.3987507820129395, "learning_rate": 7.719372615515049e-06, "loss": 0.5655, "mean_token_accuracy": 0.8337256908416748, "num_tokens": 69318706.0, "step": 1822 }, { "epoch": 0.23190433787049994, "ewc_loss": 0.05224609375, "ewc_loss_diag": 3.314018249511719e-05, "ewc_loss_parallel": 4.887580871582031e-05, "grad_norm": 4.314244270324707, "learning_rate": 7.723611699872828e-06, "loss": 0.5117, "mean_token_accuracy": 0.8481503129005432, "num_tokens": 69361895.0, "step": 1823 }, { "epoch": 0.23203154814909044, "ewc_loss": 0.051025390625, "ewc_loss_diag": 3.314018249511719e-05, "ewc_loss_parallel": 4.76837158203125e-05, "grad_norm": 4.198923110961914, "learning_rate": 7.727850784230607e-06, "loss": 0.5278, "mean_token_accuracy": 0.8404285907745361, "num_tokens": 69400877.0, "step": 1824 }, { "epoch": 0.23215875842768097, "ewc_loss": 0.051513671875, "ewc_loss_diag": 3.314018249511719e-05, "ewc_loss_parallel": 4.8160552978515625e-05, "grad_norm": 4.242483615875244, "learning_rate": 7.732089868588386e-06, "loss": 0.5513, "mean_token_accuracy": 0.8403525352478027, "num_tokens": 69437875.0, "step": 1825 }, { "epoch": 0.23228596870627147, "ewc_loss": 0.04931640625, "ewc_loss_diag": 3.314018249511719e-05, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 4.0653204917907715, "learning_rate": 7.736328952946165e-06, "loss": 0.4608, "mean_token_accuracy": 0.8645234704017639, "num_tokens": 69481770.0, "step": 1826 }, { "epoch": 0.23241317898486197, "ewc_loss": 0.05078125, "ewc_loss_diag": 3.314018249511719e-05, "ewc_loss_parallel": 4.744529724121094e-05, "grad_norm": 4.244130611419678, "learning_rate": 7.740568037303943e-06, "loss": 0.4864, "mean_token_accuracy": 0.8526848554611206, "num_tokens": 69514643.0, "step": 1827 }, { "epoch": 0.2325403892634525, "ewc_loss": 0.048095703125, "ewc_loss_diag": 3.314018249511719e-05, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 3.9837758541107178, "learning_rate": 7.744807121661722e-06, "loss": 0.4514, "mean_token_accuracy": 0.8686920404434204, "num_tokens": 69551906.0, "step": 1828 }, { "epoch": 0.232667599542043, "ewc_loss": 0.051513671875, "ewc_loss_diag": 3.314018249511719e-05, "ewc_loss_parallel": 4.8160552978515625e-05, "grad_norm": 4.261355400085449, "learning_rate": 7.749046206019501e-06, "loss": 0.5279, "mean_token_accuracy": 0.8478684425354004, "num_tokens": 69593621.0, "step": 1829 }, { "epoch": 0.2327948098206335, "ewc_loss": 0.046875, "ewc_loss_diag": 3.314018249511719e-05, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 3.917613983154297, "learning_rate": 7.75328529037728e-06, "loss": 0.4583, "mean_token_accuracy": 0.8626683950424194, "num_tokens": 69629586.0, "step": 1830 }, { "epoch": 0.23292202009922403, "ewc_loss": 0.052001953125, "ewc_loss_diag": 3.314018249511719e-05, "ewc_loss_parallel": 4.863739013671875e-05, "grad_norm": 4.330195903778076, "learning_rate": 7.757524374735059e-06, "loss": 0.567, "mean_token_accuracy": 0.8315093517303467, "num_tokens": 69667409.0, "step": 1831 }, { "epoch": 0.23304923037781453, "ewc_loss": 0.046142578125, "ewc_loss_diag": 3.314018249511719e-05, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 3.837533473968506, "learning_rate": 7.761763459092838e-06, "loss": 0.4683, "mean_token_accuracy": 0.8596185445785522, "num_tokens": 69704645.0, "step": 1832 }, { "epoch": 0.23317644065640503, "ewc_loss": 0.054931640625, "ewc_loss_diag": 3.314018249511719e-05, "ewc_loss_parallel": 5.14984130859375e-05, "grad_norm": 4.457907199859619, "learning_rate": 7.766002543450615e-06, "loss": 0.5077, "mean_token_accuracy": 0.8502159714698792, "num_tokens": 69742216.0, "step": 1833 }, { "epoch": 0.23330365093499555, "ewc_loss": 0.04541015625, "ewc_loss_diag": 3.314018249511719e-05, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 3.767580986022949, "learning_rate": 7.770241627808394e-06, "loss": 0.4607, "mean_token_accuracy": 0.860308051109314, "num_tokens": 69779119.0, "step": 1834 }, { "epoch": 0.23343086121358606, "ewc_loss": 0.057861328125, "ewc_loss_diag": 3.314018249511719e-05, "ewc_loss_parallel": 5.435943603515625e-05, "grad_norm": 4.6522040367126465, "learning_rate": 7.774480712166173e-06, "loss": 0.5506, "mean_token_accuracy": 0.8388558030128479, "num_tokens": 69819980.0, "step": 1835 }, { "epoch": 0.23355807149217656, "ewc_loss": 0.046875, "ewc_loss_diag": 3.314018249511719e-05, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 3.751595973968506, "learning_rate": 7.778719796523952e-06, "loss": 0.5194, "mean_token_accuracy": 0.842516303062439, "num_tokens": 69859840.0, "step": 1836 }, { "epoch": 0.23368528177076708, "ewc_loss": 0.06201171875, "ewc_loss_diag": 3.314018249511719e-05, "ewc_loss_parallel": 5.8650970458984375e-05, "grad_norm": 4.913211822509766, "learning_rate": 7.78295888088173e-06, "loss": 0.5236, "mean_token_accuracy": 0.8447442650794983, "num_tokens": 69893446.0, "step": 1837 }, { "epoch": 0.23381249204935758, "ewc_loss": 0.05078125, "ewc_loss_diag": 3.314018249511719e-05, "ewc_loss_parallel": 4.744529724121094e-05, "grad_norm": 3.893357276916504, "learning_rate": 7.78719796523951e-06, "loss": 0.475, "mean_token_accuracy": 0.8566756844520569, "num_tokens": 69930258.0, "step": 1838 }, { "epoch": 0.23393970232794808, "ewc_loss": 0.06640625, "ewc_loss_diag": 3.314018249511719e-05, "ewc_loss_parallel": 6.29425048828125e-05, "grad_norm": 5.095839977264404, "learning_rate": 7.791437049597288e-06, "loss": 0.5972, "mean_token_accuracy": 0.8271631002426147, "num_tokens": 69960691.0, "step": 1839 }, { "epoch": 0.2340669126065386, "ewc_loss": 0.0556640625, "ewc_loss_diag": 3.314018249511719e-05, "ewc_loss_parallel": 5.221366882324219e-05, "grad_norm": 4.195858001708984, "learning_rate": 7.795676133955067e-06, "loss": 0.5363, "mean_token_accuracy": 0.8424961566925049, "num_tokens": 69998503.0, "step": 1840 }, { "epoch": 0.2341941228851291, "ewc_loss": 0.06494140625, "ewc_loss_diag": 3.314018249511719e-05, "ewc_loss_parallel": 6.151199340820312e-05, "grad_norm": 4.983316421508789, "learning_rate": 7.799915218312844e-06, "loss": 0.5422, "mean_token_accuracy": 0.8423662781715393, "num_tokens": 70032381.0, "step": 1841 }, { "epoch": 0.23432133316371964, "ewc_loss": 0.058349609375, "ewc_loss_diag": 3.314018249511719e-05, "ewc_loss_parallel": 5.4836273193359375e-05, "grad_norm": 4.371229648590088, "learning_rate": 7.804154302670623e-06, "loss": 0.4966, "mean_token_accuracy": 0.8546392917633057, "num_tokens": 70070616.0, "step": 1842 }, { "epoch": 0.23444854344231014, "ewc_loss": 0.06005859375, "ewc_loss_diag": 3.314018249511719e-05, "ewc_loss_parallel": 5.6743621826171875e-05, "grad_norm": 4.738396644592285, "learning_rate": 7.808393387028402e-06, "loss": 0.522, "mean_token_accuracy": 0.8462005257606506, "num_tokens": 70111709.0, "step": 1843 }, { "epoch": 0.23457575372090064, "ewc_loss": 0.056640625, "ewc_loss_diag": 3.314018249511719e-05, "ewc_loss_parallel": 5.316734313964844e-05, "grad_norm": 4.372495174407959, "learning_rate": 7.812632471386181e-06, "loss": 0.564, "mean_token_accuracy": 0.8309234380722046, "num_tokens": 70148516.0, "step": 1844 }, { "epoch": 0.23470296399949117, "ewc_loss": 0.056396484375, "ewc_loss_diag": 3.337860107421875e-05, "ewc_loss_parallel": 5.2928924560546875e-05, "grad_norm": 4.510881423950195, "learning_rate": 7.81687155574396e-06, "loss": 0.5273, "mean_token_accuracy": 0.8463041186332703, "num_tokens": 70185688.0, "step": 1845 }, { "epoch": 0.23483017427808167, "ewc_loss": 0.05322265625, "ewc_loss_diag": 3.361701965332031e-05, "ewc_loss_parallel": 4.982948303222656e-05, "grad_norm": 4.298891067504883, "learning_rate": 7.821110640101739e-06, "loss": 0.5422, "mean_token_accuracy": 0.8375282883644104, "num_tokens": 70228901.0, "step": 1846 }, { "epoch": 0.23495738455667217, "ewc_loss": 0.053466796875, "ewc_loss_diag": 3.361701965332031e-05, "ewc_loss_parallel": 5.0067901611328125e-05, "grad_norm": 4.376039981842041, "learning_rate": 7.825349724459518e-06, "loss": 0.5221, "mean_token_accuracy": 0.8481123447418213, "num_tokens": 70268341.0, "step": 1847 }, { "epoch": 0.2350845948352627, "ewc_loss": 0.05224609375, "ewc_loss_diag": 3.361701965332031e-05, "ewc_loss_parallel": 4.887580871582031e-05, "grad_norm": 4.224982738494873, "learning_rate": 7.829588808817295e-06, "loss": 0.5182, "mean_token_accuracy": 0.845996618270874, "num_tokens": 70309285.0, "step": 1848 }, { "epoch": 0.2352118051138532, "ewc_loss": 0.052490234375, "ewc_loss_diag": 3.361701965332031e-05, "ewc_loss_parallel": 4.9114227294921875e-05, "grad_norm": 4.308488368988037, "learning_rate": 7.833827893175074e-06, "loss": 0.5501, "mean_token_accuracy": 0.8332539200782776, "num_tokens": 70339694.0, "step": 1849 }, { "epoch": 0.2353390153924437, "ewc_loss": 0.05029296875, "ewc_loss_diag": 3.361701965332031e-05, "ewc_loss_parallel": 4.696846008300781e-05, "grad_norm": 4.134157657623291, "learning_rate": 7.838066977532853e-06, "loss": 0.5101, "mean_token_accuracy": 0.8483429551124573, "num_tokens": 70377684.0, "step": 1850 }, { "epoch": 0.23546622567103423, "ewc_loss": 0.0517578125, "ewc_loss_diag": 3.361701965332031e-05, "ewc_loss_parallel": 4.839897155761719e-05, "grad_norm": 4.246707916259766, "learning_rate": 7.842306061890632e-06, "loss": 0.5489, "mean_token_accuracy": 0.8397420644760132, "num_tokens": 70417690.0, "step": 1851 }, { "epoch": 0.23559343594962473, "ewc_loss": 0.049560546875, "ewc_loss_diag": 3.361701965332031e-05, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 4.071319580078125, "learning_rate": 7.84654514624841e-06, "loss": 0.5347, "mean_token_accuracy": 0.8390589356422424, "num_tokens": 70454485.0, "step": 1852 }, { "epoch": 0.23572064622821523, "ewc_loss": 0.0517578125, "ewc_loss_diag": 3.361701965332031e-05, "ewc_loss_parallel": 4.839897155761719e-05, "grad_norm": 4.264503479003906, "learning_rate": 7.85078423060619e-06, "loss": 0.5627, "mean_token_accuracy": 0.8377843499183655, "num_tokens": 70490337.0, "step": 1853 }, { "epoch": 0.23584785650680576, "ewc_loss": 0.048828125, "ewc_loss_diag": 3.361701965332031e-05, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 3.984469413757324, "learning_rate": 7.855023314963968e-06, "loss": 0.5362, "mean_token_accuracy": 0.8428869247436523, "num_tokens": 70535194.0, "step": 1854 }, { "epoch": 0.23597506678539626, "ewc_loss": 0.053466796875, "ewc_loss_diag": 3.361701965332031e-05, "ewc_loss_parallel": 5.0067901611328125e-05, "grad_norm": 4.3099775314331055, "learning_rate": 7.859262399321747e-06, "loss": 0.5427, "mean_token_accuracy": 0.837756872177124, "num_tokens": 70577046.0, "step": 1855 }, { "epoch": 0.23610227706398676, "ewc_loss": 0.047607421875, "ewc_loss_diag": 3.361701965332031e-05, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 3.9078378677368164, "learning_rate": 7.863501483679526e-06, "loss": 0.4981, "mean_token_accuracy": 0.8509411215782166, "num_tokens": 70616483.0, "step": 1856 }, { "epoch": 0.2362294873425773, "ewc_loss": 0.05517578125, "ewc_loss_diag": 3.361701965332031e-05, "ewc_loss_parallel": 5.173683166503906e-05, "grad_norm": 4.4322190284729, "learning_rate": 7.867740568037305e-06, "loss": 0.5037, "mean_token_accuracy": 0.848537802696228, "num_tokens": 70654073.0, "step": 1857 }, { "epoch": 0.2363566976211678, "ewc_loss": 0.04736328125, "ewc_loss_diag": 3.361701965332031e-05, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 3.836087465286255, "learning_rate": 7.871979652395084e-06, "loss": 0.5223, "mean_token_accuracy": 0.8404955267906189, "num_tokens": 70687530.0, "step": 1858 }, { "epoch": 0.2364839078997583, "ewc_loss": 0.057373046875, "ewc_loss_diag": 3.361701965332031e-05, "ewc_loss_parallel": 5.3882598876953125e-05, "grad_norm": 4.6081156730651855, "learning_rate": 7.876218736752863e-06, "loss": 0.4889, "mean_token_accuracy": 0.8605970144271851, "num_tokens": 70729156.0, "step": 1859 }, { "epoch": 0.23661111817834882, "ewc_loss": 0.0478515625, "ewc_loss_diag": 3.361701965332031e-05, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 3.7795283794403076, "learning_rate": 7.880457821110642e-06, "loss": 0.5494, "mean_token_accuracy": 0.8372585773468018, "num_tokens": 70769781.0, "step": 1860 }, { "epoch": 0.23673832845693932, "ewc_loss": 0.061767578125, "ewc_loss_diag": 3.409385681152344e-05, "ewc_loss_parallel": 5.841255187988281e-05, "grad_norm": 4.866812705993652, "learning_rate": 7.88469690546842e-06, "loss": 0.4967, "mean_token_accuracy": 0.8582573533058167, "num_tokens": 70811822.0, "step": 1861 }, { "epoch": 0.23686553873552982, "ewc_loss": 0.050537109375, "ewc_loss_diag": 3.409385681152344e-05, "ewc_loss_parallel": 4.7206878662109375e-05, "grad_norm": 3.874490261077881, "learning_rate": 7.8889359898262e-06, "loss": 0.5323, "mean_token_accuracy": 0.8391635417938232, "num_tokens": 70850470.0, "step": 1862 }, { "epoch": 0.23699274901412035, "ewc_loss": 0.06591796875, "ewc_loss_diag": 3.409385681152344e-05, "ewc_loss_parallel": 6.246566772460938e-05, "grad_norm": 5.070105075836182, "learning_rate": 7.893175074183978e-06, "loss": 0.5389, "mean_token_accuracy": 0.8441381454467773, "num_tokens": 70889270.0, "step": 1863 }, { "epoch": 0.23711995929271085, "ewc_loss": 0.0556640625, "ewc_loss_diag": 3.409385681152344e-05, "ewc_loss_parallel": 5.221366882324219e-05, "grad_norm": 4.129327297210693, "learning_rate": 7.897414158541755e-06, "loss": 0.5215, "mean_token_accuracy": 0.8468818664550781, "num_tokens": 70928787.0, "step": 1864 }, { "epoch": 0.23724716957130135, "ewc_loss": 0.06591796875, "ewc_loss_diag": 3.409385681152344e-05, "ewc_loss_parallel": 6.246566772460938e-05, "grad_norm": 5.047125339508057, "learning_rate": 7.901653242899534e-06, "loss": 0.5685, "mean_token_accuracy": 0.8395384550094604, "num_tokens": 70965726.0, "step": 1865 }, { "epoch": 0.23737437984989188, "ewc_loss": 0.05810546875, "ewc_loss_diag": 3.409385681152344e-05, "ewc_loss_parallel": 5.459785461425781e-05, "grad_norm": 4.365268707275391, "learning_rate": 7.905892327257313e-06, "loss": 0.5228, "mean_token_accuracy": 0.8438305258750916, "num_tokens": 71002603.0, "step": 1866 }, { "epoch": 0.23750159012848238, "ewc_loss": 0.06103515625, "ewc_loss_diag": 3.409385681152344e-05, "ewc_loss_parallel": 5.7697296142578125e-05, "grad_norm": 4.817054748535156, "learning_rate": 7.910131411615092e-06, "loss": 0.5177, "mean_token_accuracy": 0.8503873348236084, "num_tokens": 71039930.0, "step": 1867 }, { "epoch": 0.2376288004070729, "ewc_loss": 0.056884765625, "ewc_loss_diag": 3.409385681152344e-05, "ewc_loss_parallel": 5.340576171875e-05, "grad_norm": 4.4001054763793945, "learning_rate": 7.914370495972871e-06, "loss": 0.5215, "mean_token_accuracy": 0.8471971750259399, "num_tokens": 71077789.0, "step": 1868 }, { "epoch": 0.2377560106856634, "ewc_loss": 0.05810546875, "ewc_loss_diag": 3.409385681152344e-05, "ewc_loss_parallel": 5.459785461425781e-05, "grad_norm": 4.580239295959473, "learning_rate": 7.91860958033065e-06, "loss": 0.5645, "mean_token_accuracy": 0.8359982967376709, "num_tokens": 71121325.0, "step": 1869 }, { "epoch": 0.2378832209642539, "ewc_loss": 0.055419921875, "ewc_loss_diag": 3.409385681152344e-05, "ewc_loss_parallel": 5.1975250244140625e-05, "grad_norm": 4.350803852081299, "learning_rate": 7.922848664688429e-06, "loss": 0.5609, "mean_token_accuracy": 0.8396549224853516, "num_tokens": 71159764.0, "step": 1870 }, { "epoch": 0.23801043124284443, "ewc_loss": 0.054931640625, "ewc_loss_diag": 3.409385681152344e-05, "ewc_loss_parallel": 5.14984130859375e-05, "grad_norm": 4.385004043579102, "learning_rate": 7.927087749046206e-06, "loss": 0.4667, "mean_token_accuracy": 0.866042971611023, "num_tokens": 71196559.0, "step": 1871 }, { "epoch": 0.23813764152143493, "ewc_loss": 0.05322265625, "ewc_loss_diag": 3.409385681152344e-05, "ewc_loss_parallel": 4.982948303222656e-05, "grad_norm": 4.299678802490234, "learning_rate": 7.931326833403985e-06, "loss": 0.5013, "mean_token_accuracy": 0.849618673324585, "num_tokens": 71229925.0, "step": 1872 }, { "epoch": 0.23826485180002543, "ewc_loss": 0.052978515625, "ewc_loss_diag": 3.409385681152344e-05, "ewc_loss_parallel": 4.9591064453125e-05, "grad_norm": 4.249495506286621, "learning_rate": 7.935565917761764e-06, "loss": 0.5126, "mean_token_accuracy": 0.8487486839294434, "num_tokens": 71273764.0, "step": 1873 }, { "epoch": 0.23839206207861596, "ewc_loss": 0.052734375, "ewc_loss_diag": 3.409385681152344e-05, "ewc_loss_parallel": 4.935264587402344e-05, "grad_norm": 4.265060901641846, "learning_rate": 7.939805002119543e-06, "loss": 0.5332, "mean_token_accuracy": 0.841776967048645, "num_tokens": 71309090.0, "step": 1874 }, { "epoch": 0.23851927235720646, "ewc_loss": 0.051025390625, "ewc_loss_diag": 3.409385681152344e-05, "ewc_loss_parallel": 4.76837158203125e-05, "grad_norm": 4.16654109954834, "learning_rate": 7.944044086477321e-06, "loss": 0.5284, "mean_token_accuracy": 0.8430213928222656, "num_tokens": 71348867.0, "step": 1875 }, { "epoch": 0.23864648263579696, "ewc_loss": 0.05224609375, "ewc_loss_diag": 3.409385681152344e-05, "ewc_loss_parallel": 4.887580871582031e-05, "grad_norm": 4.235188007354736, "learning_rate": 7.9482831708351e-06, "loss": 0.4778, "mean_token_accuracy": 0.8580491542816162, "num_tokens": 71385707.0, "step": 1876 }, { "epoch": 0.2387736929143875, "ewc_loss": 0.049560546875, "ewc_loss_diag": 3.409385681152344e-05, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 4.097888946533203, "learning_rate": 7.95252225519288e-06, "loss": 0.4431, "mean_token_accuracy": 0.8674835562705994, "num_tokens": 71428516.0, "step": 1877 }, { "epoch": 0.238900903192978, "ewc_loss": 0.0517578125, "ewc_loss_diag": 3.409385681152344e-05, "ewc_loss_parallel": 4.839897155761719e-05, "grad_norm": 4.233247756958008, "learning_rate": 7.956761339550658e-06, "loss": 0.4729, "mean_token_accuracy": 0.8593647480010986, "num_tokens": 71466913.0, "step": 1878 }, { "epoch": 0.2390281134715685, "ewc_loss": 0.04931640625, "ewc_loss_diag": 3.409385681152344e-05, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 4.0667829513549805, "learning_rate": 7.961000423908435e-06, "loss": 0.5228, "mean_token_accuracy": 0.8487899303436279, "num_tokens": 71506362.0, "step": 1879 }, { "epoch": 0.23915532375015902, "ewc_loss": 0.052001953125, "ewc_loss_diag": 3.409385681152344e-05, "ewc_loss_parallel": 4.863739013671875e-05, "grad_norm": 4.258444786071777, "learning_rate": 7.965239508266214e-06, "loss": 0.4996, "mean_token_accuracy": 0.8533077239990234, "num_tokens": 71544303.0, "step": 1880 }, { "epoch": 0.23928253402874952, "ewc_loss": 0.04833984375, "ewc_loss_diag": 3.409385681152344e-05, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 4.003222465515137, "learning_rate": 7.969478592623993e-06, "loss": 0.4766, "mean_token_accuracy": 0.8577847480773926, "num_tokens": 71582295.0, "step": 1881 }, { "epoch": 0.23940974430734002, "ewc_loss": 0.05322265625, "ewc_loss_diag": 3.409385681152344e-05, "ewc_loss_parallel": 4.982948303222656e-05, "grad_norm": 4.315585613250732, "learning_rate": 7.973717676981772e-06, "loss": 0.5051, "mean_token_accuracy": 0.8515007495880127, "num_tokens": 71621880.0, "step": 1882 }, { "epoch": 0.23953695458593055, "ewc_loss": 0.047607421875, "ewc_loss_diag": 3.409385681152344e-05, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 3.9303274154663086, "learning_rate": 7.97795676133955e-06, "loss": 0.4602, "mean_token_accuracy": 0.8609681129455566, "num_tokens": 71657997.0, "step": 1883 }, { "epoch": 0.23966416486452105, "ewc_loss": 0.0546875, "ewc_loss_diag": 3.409385681152344e-05, "ewc_loss_parallel": 5.125999450683594e-05, "grad_norm": 4.450343608856201, "learning_rate": 7.98219584569733e-06, "loss": 0.6129, "mean_token_accuracy": 0.8176369667053223, "num_tokens": 71697622.0, "step": 1884 }, { "epoch": 0.23979137514311155, "ewc_loss": 0.046875, "ewc_loss_diag": 3.409385681152344e-05, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 3.832103967666626, "learning_rate": 7.986434930055109e-06, "loss": 0.4656, "mean_token_accuracy": 0.8579144477844238, "num_tokens": 71733354.0, "step": 1885 }, { "epoch": 0.23991858542170208, "ewc_loss": 0.058349609375, "ewc_loss_diag": 3.409385681152344e-05, "ewc_loss_parallel": 5.4836273193359375e-05, "grad_norm": 4.66098690032959, "learning_rate": 7.990674014412887e-06, "loss": 0.5339, "mean_token_accuracy": 0.8440932035446167, "num_tokens": 71777172.0, "step": 1886 }, { "epoch": 0.24004579570029258, "ewc_loss": 0.04638671875, "ewc_loss_diag": 3.409385681152344e-05, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 3.750648260116577, "learning_rate": 7.994913098770666e-06, "loss": 0.5101, "mean_token_accuracy": 0.844675600528717, "num_tokens": 71816165.0, "step": 1887 }, { "epoch": 0.24017300597888308, "ewc_loss": 0.064453125, "ewc_loss_diag": 3.409385681152344e-05, "ewc_loss_parallel": 6.103515625e-05, "grad_norm": 4.99922513961792, "learning_rate": 7.999152183128445e-06, "loss": 0.5585, "mean_token_accuracy": 0.8366783857345581, "num_tokens": 71852313.0, "step": 1888 }, { "epoch": 0.2403002162574736, "ewc_loss": 0.05078125, "ewc_loss_diag": 3.409385681152344e-05, "ewc_loss_parallel": 4.744529724121094e-05, "grad_norm": 3.8342530727386475, "learning_rate": 8.003391267486224e-06, "loss": 0.4955, "mean_token_accuracy": 0.8517451286315918, "num_tokens": 71892462.0, "step": 1889 }, { "epoch": 0.2404274265360641, "ewc_loss": 0.0703125, "ewc_loss_diag": 3.409385681152344e-05, "ewc_loss_parallel": 6.67572021484375e-05, "grad_norm": 5.343531608581543, "learning_rate": 8.007630351844003e-06, "loss": 0.5245, "mean_token_accuracy": 0.8522053956985474, "num_tokens": 71924079.0, "step": 1890 }, { "epoch": 0.24055463681465464, "ewc_loss": 0.058349609375, "ewc_loss_diag": 3.409385681152344e-05, "ewc_loss_parallel": 5.4836273193359375e-05, "grad_norm": 4.240107536315918, "learning_rate": 8.011869436201782e-06, "loss": 0.4988, "mean_token_accuracy": 0.8543292880058289, "num_tokens": 71959546.0, "step": 1891 }, { "epoch": 0.24068184709324514, "ewc_loss": 0.0703125, "ewc_loss_diag": 3.409385681152344e-05, "ewc_loss_parallel": 6.67572021484375e-05, "grad_norm": 5.335010051727295, "learning_rate": 8.01610852055956e-06, "loss": 0.5465, "mean_token_accuracy": 0.8426061868667603, "num_tokens": 72002565.0, "step": 1892 }, { "epoch": 0.24080905737183564, "ewc_loss": 0.0654296875, "ewc_loss_diag": 3.409385681152344e-05, "ewc_loss_parallel": 6.198883056640625e-05, "grad_norm": 4.601503372192383, "learning_rate": 8.02034760491734e-06, "loss": 0.4994, "mean_token_accuracy": 0.8563128709793091, "num_tokens": 72037178.0, "step": 1893 }, { "epoch": 0.24093626765042617, "ewc_loss": 0.0654296875, "ewc_loss_diag": 3.409385681152344e-05, "ewc_loss_parallel": 6.198883056640625e-05, "grad_norm": 5.018632888793945, "learning_rate": 8.024586689275117e-06, "loss": 0.5258, "mean_token_accuracy": 0.8480300903320312, "num_tokens": 72076562.0, "step": 1894 }, { "epoch": 0.24106347792901667, "ewc_loss": 0.062255859375, "ewc_loss_diag": 3.409385681152344e-05, "ewc_loss_parallel": 5.888938903808594e-05, "grad_norm": 4.607112407684326, "learning_rate": 8.028825773632896e-06, "loss": 0.53, "mean_token_accuracy": 0.8480185866355896, "num_tokens": 72113379.0, "step": 1895 }, { "epoch": 0.24119068820760717, "ewc_loss": 0.06005859375, "ewc_loss_diag": 3.409385681152344e-05, "ewc_loss_parallel": 5.6743621826171875e-05, "grad_norm": 4.751057147979736, "learning_rate": 8.033064857990675e-06, "loss": 0.4562, "mean_token_accuracy": 0.8670451641082764, "num_tokens": 72149225.0, "step": 1896 }, { "epoch": 0.2413178984861977, "ewc_loss": 0.058349609375, "ewc_loss_diag": 3.409385681152344e-05, "ewc_loss_parallel": 5.4836273193359375e-05, "grad_norm": 4.4776153564453125, "learning_rate": 8.037303942348453e-06, "loss": 0.4997, "mean_token_accuracy": 0.852526068687439, "num_tokens": 72185774.0, "step": 1897 }, { "epoch": 0.2414451087647882, "ewc_loss": 0.057373046875, "ewc_loss_diag": 3.409385681152344e-05, "ewc_loss_parallel": 5.3882598876953125e-05, "grad_norm": 4.554229259490967, "learning_rate": 8.041543026706232e-06, "loss": 0.4943, "mean_token_accuracy": 0.8560681939125061, "num_tokens": 72224270.0, "step": 1898 }, { "epoch": 0.2415723190433787, "ewc_loss": 0.054931640625, "ewc_loss_diag": 3.409385681152344e-05, "ewc_loss_parallel": 5.14984130859375e-05, "grad_norm": 4.3743672370910645, "learning_rate": 8.045782111064011e-06, "loss": 0.501, "mean_token_accuracy": 0.8510841727256775, "num_tokens": 72260629.0, "step": 1899 }, { "epoch": 0.24169952932196923, "ewc_loss": 0.055419921875, "ewc_loss_diag": 3.409385681152344e-05, "ewc_loss_parallel": 5.1975250244140625e-05, "grad_norm": 4.446946620941162, "learning_rate": 8.05002119542179e-06, "loss": 0.5473, "mean_token_accuracy": 0.8383808135986328, "num_tokens": 72300040.0, "step": 1900 }, { "epoch": 0.24182673960055973, "ewc_loss": 0.052734375, "ewc_loss_diag": 3.409385681152344e-05, "ewc_loss_parallel": 4.935264587402344e-05, "grad_norm": 4.272029399871826, "learning_rate": 8.054260279779569e-06, "loss": 0.5132, "mean_token_accuracy": 0.8466038703918457, "num_tokens": 72335396.0, "step": 1901 }, { "epoch": 0.24195394987915023, "ewc_loss": 0.054443359375, "ewc_loss_diag": 3.409385681152344e-05, "ewc_loss_parallel": 5.1021575927734375e-05, "grad_norm": 4.397994518280029, "learning_rate": 8.058499364137346e-06, "loss": 0.5539, "mean_token_accuracy": 0.8393349647521973, "num_tokens": 72375215.0, "step": 1902 }, { "epoch": 0.24208116015774075, "ewc_loss": 0.052001953125, "ewc_loss_diag": 3.4332275390625e-05, "ewc_loss_parallel": 4.863739013671875e-05, "grad_norm": 4.172528266906738, "learning_rate": 8.062738448495125e-06, "loss": 0.4634, "mean_token_accuracy": 0.863013744354248, "num_tokens": 72416407.0, "step": 1903 }, { "epoch": 0.24220837043633126, "ewc_loss": 0.053955078125, "ewc_loss_diag": 3.4332275390625e-05, "ewc_loss_parallel": 5.054473876953125e-05, "grad_norm": 4.404537200927734, "learning_rate": 8.066977532852904e-06, "loss": 0.4667, "mean_token_accuracy": 0.8666738271713257, "num_tokens": 72455031.0, "step": 1904 }, { "epoch": 0.24233558071492176, "ewc_loss": 0.049560546875, "ewc_loss_diag": 3.4332275390625e-05, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 4.093484878540039, "learning_rate": 8.071216617210683e-06, "loss": 0.5281, "mean_token_accuracy": 0.8438859581947327, "num_tokens": 72487186.0, "step": 1905 }, { "epoch": 0.24246279099351228, "ewc_loss": 0.0556640625, "ewc_loss_diag": 3.4332275390625e-05, "ewc_loss_parallel": 5.221366882324219e-05, "grad_norm": 4.469991683959961, "learning_rate": 8.075455701568462e-06, "loss": 0.502, "mean_token_accuracy": 0.851645827293396, "num_tokens": 72525659.0, "step": 1906 }, { "epoch": 0.24259000127210278, "ewc_loss": 0.049072265625, "ewc_loss_diag": 3.4332275390625e-05, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 3.9779763221740723, "learning_rate": 8.07969478592624e-06, "loss": 0.4889, "mean_token_accuracy": 0.8559601306915283, "num_tokens": 72567912.0, "step": 1907 }, { "epoch": 0.24271721155069328, "ewc_loss": 0.05712890625, "ewc_loss_diag": 3.4332275390625e-05, "ewc_loss_parallel": 5.364418029785156e-05, "grad_norm": 4.629563331604004, "learning_rate": 8.08393387028402e-06, "loss": 0.4551, "mean_token_accuracy": 0.8676887154579163, "num_tokens": 72599525.0, "step": 1908 }, { "epoch": 0.2428444218292838, "ewc_loss": 0.048583984375, "ewc_loss_diag": 3.4332275390625e-05, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 3.894117832183838, "learning_rate": 8.088172954641797e-06, "loss": 0.5005, "mean_token_accuracy": 0.8487756252288818, "num_tokens": 72635391.0, "step": 1909 }, { "epoch": 0.2429716321078743, "ewc_loss": 0.060791015625, "ewc_loss_diag": 3.4332275390625e-05, "ewc_loss_parallel": 5.745887756347656e-05, "grad_norm": 4.891911506652832, "learning_rate": 8.092412038999576e-06, "loss": 0.5546, "mean_token_accuracy": 0.8363817930221558, "num_tokens": 72666058.0, "step": 1910 }, { "epoch": 0.24309884238646481, "ewc_loss": 0.05029296875, "ewc_loss_diag": 3.4332275390625e-05, "ewc_loss_parallel": 4.696846008300781e-05, "grad_norm": 3.9088659286499023, "learning_rate": 8.096651123357354e-06, "loss": 0.5509, "mean_token_accuracy": 0.8318318724632263, "num_tokens": 72699869.0, "step": 1911 }, { "epoch": 0.24322605266505534, "ewc_loss": 0.06787109375, "ewc_loss_diag": 3.457069396972656e-05, "ewc_loss_parallel": 6.437301635742188e-05, "grad_norm": 5.14210319519043, "learning_rate": 8.100890207715133e-06, "loss": 0.5637, "mean_token_accuracy": 0.8399293422698975, "num_tokens": 72735660.0, "step": 1912 }, { "epoch": 0.24335326294364584, "ewc_loss": 0.053955078125, "ewc_loss_diag": 3.457069396972656e-05, "ewc_loss_parallel": 5.054473876953125e-05, "grad_norm": 4.112579345703125, "learning_rate": 8.105129292072912e-06, "loss": 0.5196, "mean_token_accuracy": 0.8471811413764954, "num_tokens": 72774334.0, "step": 1913 }, { "epoch": 0.24348047322223634, "ewc_loss": 0.06884765625, "ewc_loss_diag": 3.457069396972656e-05, "ewc_loss_parallel": 6.532669067382812e-05, "grad_norm": 5.24181604385376, "learning_rate": 8.109368376430691e-06, "loss": 0.5609, "mean_token_accuracy": 0.8392567038536072, "num_tokens": 72808387.0, "step": 1914 }, { "epoch": 0.24360768350082687, "ewc_loss": 0.059814453125, "ewc_loss_diag": 3.457069396972656e-05, "ewc_loss_parallel": 5.650520324707031e-05, "grad_norm": 4.427833557128906, "learning_rate": 8.11360746078847e-06, "loss": 0.4903, "mean_token_accuracy": 0.8574923276901245, "num_tokens": 72846144.0, "step": 1915 }, { "epoch": 0.24373489377941737, "ewc_loss": 0.06591796875, "ewc_loss_diag": 3.457069396972656e-05, "ewc_loss_parallel": 6.246566772460938e-05, "grad_norm": 5.025547504425049, "learning_rate": 8.117846545146249e-06, "loss": 0.4734, "mean_token_accuracy": 0.8629617691040039, "num_tokens": 72887381.0, "step": 1916 }, { "epoch": 0.2438621040580079, "ewc_loss": 0.060546875, "ewc_loss_diag": 3.457069396972656e-05, "ewc_loss_parallel": 5.7220458984375e-05, "grad_norm": 4.551898956298828, "learning_rate": 8.122085629504028e-06, "loss": 0.5532, "mean_token_accuracy": 0.8350967168807983, "num_tokens": 72927126.0, "step": 1917 }, { "epoch": 0.2439893143365984, "ewc_loss": 0.060302734375, "ewc_loss_diag": 3.457069396972656e-05, "ewc_loss_parallel": 5.698204040527344e-05, "grad_norm": 4.733218193054199, "learning_rate": 8.126324713861807e-06, "loss": 0.5288, "mean_token_accuracy": 0.8461946845054626, "num_tokens": 72968770.0, "step": 1918 }, { "epoch": 0.2441165246151889, "ewc_loss": 0.058837890625, "ewc_loss_diag": 3.4809112548828125e-05, "ewc_loss_parallel": 5.53131103515625e-05, "grad_norm": 4.514573097229004, "learning_rate": 8.130563798219586e-06, "loss": 0.5034, "mean_token_accuracy": 0.8556252121925354, "num_tokens": 73008831.0, "step": 1919 }, { "epoch": 0.24424373489377943, "ewc_loss": 0.056884765625, "ewc_loss_diag": 3.4809112548828125e-05, "ewc_loss_parallel": 5.340576171875e-05, "grad_norm": 4.488602638244629, "learning_rate": 8.134802882577364e-06, "loss": 0.5373, "mean_token_accuracy": 0.8458319902420044, "num_tokens": 73048469.0, "step": 1920 }, { "epoch": 0.24437094517236993, "ewc_loss": 0.0556640625, "ewc_loss_diag": 3.4809112548828125e-05, "ewc_loss_parallel": 5.221366882324219e-05, "grad_norm": 4.453908920288086, "learning_rate": 8.139041966935143e-06, "loss": 0.5161, "mean_token_accuracy": 0.8505043983459473, "num_tokens": 73096067.0, "step": 1921 }, { "epoch": 0.24449815545096043, "ewc_loss": 0.05419921875, "ewc_loss_diag": 3.4809112548828125e-05, "ewc_loss_parallel": 5.078315734863281e-05, "grad_norm": 4.321532249450684, "learning_rate": 8.143281051292922e-06, "loss": 0.5755, "mean_token_accuracy": 0.8354958295822144, "num_tokens": 73132680.0, "step": 1922 }, { "epoch": 0.24462536572955096, "ewc_loss": 0.05517578125, "ewc_loss_diag": 3.4809112548828125e-05, "ewc_loss_parallel": 5.173683166503906e-05, "grad_norm": 4.4336838722229, "learning_rate": 8.147520135650701e-06, "loss": 0.5323, "mean_token_accuracy": 0.8429635763168335, "num_tokens": 73168436.0, "step": 1923 }, { "epoch": 0.24475257600814146, "ewc_loss": 0.05224609375, "ewc_loss_diag": 3.4809112548828125e-05, "ewc_loss_parallel": 4.887580871582031e-05, "grad_norm": 4.195718288421631, "learning_rate": 8.151759220008478e-06, "loss": 0.5055, "mean_token_accuracy": 0.8487459421157837, "num_tokens": 73210961.0, "step": 1924 }, { "epoch": 0.24487978628673196, "ewc_loss": 0.05517578125, "ewc_loss_diag": 3.4809112548828125e-05, "ewc_loss_parallel": 5.173683166503906e-05, "grad_norm": 4.446990489959717, "learning_rate": 8.155998304366257e-06, "loss": 0.5156, "mean_token_accuracy": 0.8534013032913208, "num_tokens": 73244580.0, "step": 1925 }, { "epoch": 0.2450069965653225, "ewc_loss": 0.05078125, "ewc_loss_diag": 3.4809112548828125e-05, "ewc_loss_parallel": 4.744529724121094e-05, "grad_norm": 4.1139702796936035, "learning_rate": 8.160237388724036e-06, "loss": 0.5485, "mean_token_accuracy": 0.8375347852706909, "num_tokens": 73278176.0, "step": 1926 }, { "epoch": 0.245134206843913, "ewc_loss": 0.05615234375, "ewc_loss_diag": 3.4809112548828125e-05, "ewc_loss_parallel": 5.269050598144531e-05, "grad_norm": 4.4906229972839355, "learning_rate": 8.164476473081815e-06, "loss": 0.5284, "mean_token_accuracy": 0.8456565737724304, "num_tokens": 73315678.0, "step": 1927 }, { "epoch": 0.2452614171225035, "ewc_loss": 0.04931640625, "ewc_loss_diag": 3.4809112548828125e-05, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 4.015956878662109, "learning_rate": 8.168715557439594e-06, "loss": 0.4662, "mean_token_accuracy": 0.860381007194519, "num_tokens": 73356336.0, "step": 1928 }, { "epoch": 0.24538862740109402, "ewc_loss": 0.057373046875, "ewc_loss_diag": 3.4809112548828125e-05, "ewc_loss_parallel": 5.3882598876953125e-05, "grad_norm": 4.599572658538818, "learning_rate": 8.172954641797373e-06, "loss": 0.4971, "mean_token_accuracy": 0.8532067537307739, "num_tokens": 73396605.0, "step": 1929 }, { "epoch": 0.24551583767968452, "ewc_loss": 0.049072265625, "ewc_loss_diag": 3.504753112792969e-05, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 3.939730167388916, "learning_rate": 8.177193726155152e-06, "loss": 0.5987, "mean_token_accuracy": 0.8227049112319946, "num_tokens": 73444278.0, "step": 1930 }, { "epoch": 0.24564304795827502, "ewc_loss": 0.06005859375, "ewc_loss_diag": 3.504753112792969e-05, "ewc_loss_parallel": 5.6743621826171875e-05, "grad_norm": 4.796696662902832, "learning_rate": 8.18143281051293e-06, "loss": 0.5684, "mean_token_accuracy": 0.835497260093689, "num_tokens": 73483755.0, "step": 1931 }, { "epoch": 0.24577025823686555, "ewc_loss": 0.050048828125, "ewc_loss_diag": 3.504753112792969e-05, "ewc_loss_parallel": 4.673004150390625e-05, "grad_norm": 3.8969614505767822, "learning_rate": 8.185671894870708e-06, "loss": 0.5728, "mean_token_accuracy": 0.830329418182373, "num_tokens": 73522819.0, "step": 1932 }, { "epoch": 0.24589746851545605, "ewc_loss": 0.0654296875, "ewc_loss_diag": 3.504753112792969e-05, "ewc_loss_parallel": 6.198883056640625e-05, "grad_norm": 5.065377235412598, "learning_rate": 8.189910979228487e-06, "loss": 0.4887, "mean_token_accuracy": 0.855683445930481, "num_tokens": 73554973.0, "step": 1933 }, { "epoch": 0.24602467879404655, "ewc_loss": 0.052490234375, "ewc_loss_diag": 3.504753112792969e-05, "ewc_loss_parallel": 4.9114227294921875e-05, "grad_norm": 3.999098777770996, "learning_rate": 8.194150063586265e-06, "loss": 0.5607, "mean_token_accuracy": 0.8341217041015625, "num_tokens": 73592541.0, "step": 1934 }, { "epoch": 0.24615188907263708, "ewc_loss": 0.06982421875, "ewc_loss_diag": 3.504753112792969e-05, "ewc_loss_parallel": 6.628036499023438e-05, "grad_norm": 5.274673938751221, "learning_rate": 8.198389147944044e-06, "loss": 0.5654, "mean_token_accuracy": 0.8363168835639954, "num_tokens": 73629328.0, "step": 1935 }, { "epoch": 0.24627909935122758, "ewc_loss": 0.05810546875, "ewc_loss_diag": 3.504753112792969e-05, "ewc_loss_parallel": 5.459785461425781e-05, "grad_norm": 4.2805280685424805, "learning_rate": 8.202628232301823e-06, "loss": 0.5276, "mean_token_accuracy": 0.8483761548995972, "num_tokens": 73664944.0, "step": 1936 }, { "epoch": 0.24640630962981808, "ewc_loss": 0.06982421875, "ewc_loss_diag": 3.504753112792969e-05, "ewc_loss_parallel": 6.628036499023438e-05, "grad_norm": 5.239986896514893, "learning_rate": 8.206867316659602e-06, "loss": 0.5745, "mean_token_accuracy": 0.8340849280357361, "num_tokens": 73712403.0, "step": 1937 }, { "epoch": 0.2465335199084086, "ewc_loss": 0.062255859375, "ewc_loss_diag": 3.504753112792969e-05, "ewc_loss_parallel": 5.888938903808594e-05, "grad_norm": 4.51298189163208, "learning_rate": 8.211106401017381e-06, "loss": 0.4437, "mean_token_accuracy": 0.8697596788406372, "num_tokens": 73752716.0, "step": 1938 }, { "epoch": 0.2466607301869991, "ewc_loss": 0.0654296875, "ewc_loss_diag": 3.504753112792969e-05, "ewc_loss_parallel": 6.198883056640625e-05, "grad_norm": 4.998098373413086, "learning_rate": 8.21534548537516e-06, "loss": 0.5763, "mean_token_accuracy": 0.8317920565605164, "num_tokens": 73796900.0, "step": 1939 }, { "epoch": 0.2467879404655896, "ewc_loss": 0.060791015625, "ewc_loss_diag": 3.504753112792969e-05, "ewc_loss_parallel": 5.745887756347656e-05, "grad_norm": 4.573781490325928, "learning_rate": 8.219584569732939e-06, "loss": 0.5392, "mean_token_accuracy": 0.8396918773651123, "num_tokens": 73836426.0, "step": 1940 }, { "epoch": 0.24691515074418013, "ewc_loss": 0.060791015625, "ewc_loss_diag": 3.528594970703125e-05, "ewc_loss_parallel": 5.745887756347656e-05, "grad_norm": 4.727661609649658, "learning_rate": 8.223823654090718e-06, "loss": 0.4772, "mean_token_accuracy": 0.8594177961349487, "num_tokens": 73872147.0, "step": 1941 }, { "epoch": 0.24704236102277063, "ewc_loss": 0.058837890625, "ewc_loss_diag": 3.528594970703125e-05, "ewc_loss_parallel": 5.53131103515625e-05, "grad_norm": 4.50996208190918, "learning_rate": 8.228062738448497e-06, "loss": 0.5145, "mean_token_accuracy": 0.8470203280448914, "num_tokens": 73912878.0, "step": 1942 }, { "epoch": 0.24716957130136116, "ewc_loss": 0.057861328125, "ewc_loss_diag": 3.528594970703125e-05, "ewc_loss_parallel": 5.435943603515625e-05, "grad_norm": 4.522037029266357, "learning_rate": 8.232301822806274e-06, "loss": 0.5396, "mean_token_accuracy": 0.8421893119812012, "num_tokens": 73952152.0, "step": 1943 }, { "epoch": 0.24729678157995166, "ewc_loss": 0.055419921875, "ewc_loss_diag": 3.528594970703125e-05, "ewc_loss_parallel": 5.1975250244140625e-05, "grad_norm": 4.44343900680542, "learning_rate": 8.236540907164053e-06, "loss": 0.5249, "mean_token_accuracy": 0.8478076457977295, "num_tokens": 73988760.0, "step": 1944 }, { "epoch": 0.24742399185854216, "ewc_loss": 0.05517578125, "ewc_loss_diag": 3.528594970703125e-05, "ewc_loss_parallel": 5.173683166503906e-05, "grad_norm": 4.389520645141602, "learning_rate": 8.240779991521831e-06, "loss": 0.5014, "mean_token_accuracy": 0.8508380651473999, "num_tokens": 74028107.0, "step": 1945 }, { "epoch": 0.2475512021371327, "ewc_loss": 0.0546875, "ewc_loss_diag": 3.528594970703125e-05, "ewc_loss_parallel": 5.125999450683594e-05, "grad_norm": 4.389003276824951, "learning_rate": 8.24501907587961e-06, "loss": 0.5415, "mean_token_accuracy": 0.8392730951309204, "num_tokens": 74071236.0, "step": 1946 }, { "epoch": 0.2476784124157232, "ewc_loss": 0.053955078125, "ewc_loss_diag": 3.528594970703125e-05, "ewc_loss_parallel": 5.054473876953125e-05, "grad_norm": 4.312613010406494, "learning_rate": 8.24925816023739e-06, "loss": 0.5417, "mean_token_accuracy": 0.845440149307251, "num_tokens": 74105734.0, "step": 1947 }, { "epoch": 0.2478056226943137, "ewc_loss": 0.053955078125, "ewc_loss_diag": 3.528594970703125e-05, "ewc_loss_parallel": 5.054473876953125e-05, "grad_norm": 4.355526447296143, "learning_rate": 8.253497244595168e-06, "loss": 0.5409, "mean_token_accuracy": 0.8400334119796753, "num_tokens": 74142569.0, "step": 1948 }, { "epoch": 0.24793283297290422, "ewc_loss": 0.052978515625, "ewc_loss_diag": 3.528594970703125e-05, "ewc_loss_parallel": 4.9591064453125e-05, "grad_norm": 4.257322788238525, "learning_rate": 8.257736328952947e-06, "loss": 0.5588, "mean_token_accuracy": 0.8343132734298706, "num_tokens": 74185395.0, "step": 1949 }, { "epoch": 0.24806004325149472, "ewc_loss": 0.0537109375, "ewc_loss_diag": 3.528594970703125e-05, "ewc_loss_parallel": 5.030632019042969e-05, "grad_norm": 4.326493263244629, "learning_rate": 8.261975413310726e-06, "loss": 0.5176, "mean_token_accuracy": 0.8467094302177429, "num_tokens": 74223416.0, "step": 1950 }, { "epoch": 0.24818725353008522, "ewc_loss": 0.052490234375, "ewc_loss_diag": 3.528594970703125e-05, "ewc_loss_parallel": 4.9114227294921875e-05, "grad_norm": 4.205644130706787, "learning_rate": 8.266214497668505e-06, "loss": 0.4766, "mean_token_accuracy": 0.8613159656524658, "num_tokens": 74265356.0, "step": 1951 }, { "epoch": 0.24831446380867575, "ewc_loss": 0.05419921875, "ewc_loss_diag": 3.528594970703125e-05, "ewc_loss_parallel": 5.078315734863281e-05, "grad_norm": 4.353874683380127, "learning_rate": 8.270453582026284e-06, "loss": 0.5544, "mean_token_accuracy": 0.8359287977218628, "num_tokens": 74303807.0, "step": 1952 }, { "epoch": 0.24844167408726625, "ewc_loss": 0.051513671875, "ewc_loss_diag": 3.528594970703125e-05, "ewc_loss_parallel": 4.8160552978515625e-05, "grad_norm": 4.1598124504089355, "learning_rate": 8.274692666384063e-06, "loss": 0.501, "mean_token_accuracy": 0.8518695831298828, "num_tokens": 74337299.0, "step": 1953 }, { "epoch": 0.24856888436585675, "ewc_loss": 0.0546875, "ewc_loss_diag": 3.528594970703125e-05, "ewc_loss_parallel": 5.125999450683594e-05, "grad_norm": 4.403230667114258, "learning_rate": 8.278931750741841e-06, "loss": 0.523, "mean_token_accuracy": 0.8439623117446899, "num_tokens": 74370817.0, "step": 1954 }, { "epoch": 0.24869609464444728, "ewc_loss": 0.050048828125, "ewc_loss_diag": 3.528594970703125e-05, "ewc_loss_parallel": 4.673004150390625e-05, "grad_norm": 4.095251560211182, "learning_rate": 8.283170835099619e-06, "loss": 0.5064, "mean_token_accuracy": 0.8560501337051392, "num_tokens": 74408009.0, "step": 1955 }, { "epoch": 0.24882330492303778, "ewc_loss": 0.05615234375, "ewc_loss_diag": 3.528594970703125e-05, "ewc_loss_parallel": 5.269050598144531e-05, "grad_norm": 4.484434604644775, "learning_rate": 8.287409919457398e-06, "loss": 0.4911, "mean_token_accuracy": 0.855907678604126, "num_tokens": 74446669.0, "step": 1956 }, { "epoch": 0.24895051520162828, "ewc_loss": 0.04931640625, "ewc_loss_diag": 3.528594970703125e-05, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 3.9962961673736572, "learning_rate": 8.291649003815176e-06, "loss": 0.5267, "mean_token_accuracy": 0.8435525298118591, "num_tokens": 74484264.0, "step": 1957 }, { "epoch": 0.2490777254802188, "ewc_loss": 0.058837890625, "ewc_loss_diag": 3.528594970703125e-05, "ewc_loss_parallel": 5.53131103515625e-05, "grad_norm": 4.663036346435547, "learning_rate": 8.295888088172955e-06, "loss": 0.5545, "mean_token_accuracy": 0.8418353199958801, "num_tokens": 74523956.0, "step": 1958 }, { "epoch": 0.2492049357588093, "ewc_loss": 0.04931640625, "ewc_loss_diag": 3.528594970703125e-05, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 3.8752641677856445, "learning_rate": 8.300127172530734e-06, "loss": 0.4458, "mean_token_accuracy": 0.8668405413627625, "num_tokens": 74560748.0, "step": 1959 }, { "epoch": 0.2493321460373998, "ewc_loss": 0.06494140625, "ewc_loss_diag": 3.528594970703125e-05, "ewc_loss_parallel": 6.151199340820312e-05, "grad_norm": 4.993662357330322, "learning_rate": 8.304366256888513e-06, "loss": 0.5105, "mean_token_accuracy": 0.8517020344734192, "num_tokens": 74596465.0, "step": 1960 }, { "epoch": 0.24945935631599034, "ewc_loss": 0.051025390625, "ewc_loss_diag": 3.528594970703125e-05, "ewc_loss_parallel": 4.76837158203125e-05, "grad_norm": 3.8449974060058594, "learning_rate": 8.308605341246292e-06, "loss": 0.5042, "mean_token_accuracy": 0.8498774170875549, "num_tokens": 74632788.0, "step": 1961 }, { "epoch": 0.24958656659458084, "ewc_loss": 0.0732421875, "ewc_loss_diag": 3.528594970703125e-05, "ewc_loss_parallel": 6.961822509765625e-05, "grad_norm": 5.431804656982422, "learning_rate": 8.312844425604069e-06, "loss": 0.5255, "mean_token_accuracy": 0.8557433485984802, "num_tokens": 74667723.0, "step": 1962 }, { "epoch": 0.24971377687317134, "ewc_loss": 0.057373046875, "ewc_loss_diag": 3.528594970703125e-05, "ewc_loss_parallel": 5.3882598876953125e-05, "grad_norm": 4.128637790679932, "learning_rate": 8.317083509961848e-06, "loss": 0.5425, "mean_token_accuracy": 0.8445532321929932, "num_tokens": 74709205.0, "step": 1963 }, { "epoch": 0.24984098715176187, "ewc_loss": 0.07568359375, "ewc_loss_diag": 3.5762786865234375e-05, "ewc_loss_parallel": 7.2479248046875e-05, "grad_norm": 5.6787285804748535, "learning_rate": 8.321322594319627e-06, "loss": 0.5033, "mean_token_accuracy": 0.8559963703155518, "num_tokens": 74749437.0, "step": 1964 }, { "epoch": 0.24996819743035237, "ewc_loss": 0.068359375, "ewc_loss_diag": 3.5762786865234375e-05, "ewc_loss_parallel": 6.4849853515625e-05, "grad_norm": 4.700998306274414, "learning_rate": 8.325561678677406e-06, "loss": 0.5625, "mean_token_accuracy": 0.8361560106277466, "num_tokens": 74790602.0, "step": 1965 }, { "epoch": 0.2500954077089429, "ewc_loss": 0.072265625, "ewc_loss_diag": 3.5762786865234375e-05, "ewc_loss_parallel": 6.866455078125e-05, "grad_norm": 5.4142069816589355, "learning_rate": 8.329800763035185e-06, "loss": 0.5737, "mean_token_accuracy": 0.8379106521606445, "num_tokens": 74830845.0, "step": 1966 }, { "epoch": 0.25022261798753337, "ewc_loss": 0.068359375, "ewc_loss_diag": 3.5762786865234375e-05, "ewc_loss_parallel": 6.4849853515625e-05, "grad_norm": 4.8746538162231445, "learning_rate": 8.334039847392964e-06, "loss": 0.5186, "mean_token_accuracy": 0.8498603105545044, "num_tokens": 74864208.0, "step": 1967 }, { "epoch": 0.2503498282661239, "ewc_loss": 0.06689453125, "ewc_loss_diag": 3.5762786865234375e-05, "ewc_loss_parallel": 6.341934204101562e-05, "grad_norm": 5.062401294708252, "learning_rate": 8.338278931750742e-06, "loss": 0.489, "mean_token_accuracy": 0.8600720167160034, "num_tokens": 74897217.0, "step": 1968 }, { "epoch": 0.2504770385447144, "ewc_loss": 0.064453125, "ewc_loss_diag": 3.5762786865234375e-05, "ewc_loss_parallel": 6.103515625e-05, "grad_norm": 4.740579605102539, "learning_rate": 8.342518016108521e-06, "loss": 0.5084, "mean_token_accuracy": 0.8506933450698853, "num_tokens": 74932105.0, "step": 1969 }, { "epoch": 0.2506042488233049, "ewc_loss": 0.06298828125, "ewc_loss_diag": 3.5762786865234375e-05, "ewc_loss_parallel": 5.936622619628906e-05, "grad_norm": 4.818975448608398, "learning_rate": 8.3467571004663e-06, "loss": 0.5131, "mean_token_accuracy": 0.8519816398620605, "num_tokens": 74965229.0, "step": 1970 }, { "epoch": 0.2507314591018954, "ewc_loss": 0.059814453125, "ewc_loss_diag": 3.5762786865234375e-05, "ewc_loss_parallel": 5.626678466796875e-05, "grad_norm": 4.587625026702881, "learning_rate": 8.350996184824079e-06, "loss": 0.5544, "mean_token_accuracy": 0.8369225263595581, "num_tokens": 75002290.0, "step": 1971 }, { "epoch": 0.25085866938048595, "ewc_loss": 0.059326171875, "ewc_loss_diag": 3.5762786865234375e-05, "ewc_loss_parallel": 5.555152893066406e-05, "grad_norm": 4.664925575256348, "learning_rate": 8.355235269181858e-06, "loss": 0.533, "mean_token_accuracy": 0.8415833115577698, "num_tokens": 75038113.0, "step": 1972 }, { "epoch": 0.2509858796590764, "ewc_loss": 0.056396484375, "ewc_loss_diag": 3.5762786865234375e-05, "ewc_loss_parallel": 5.269050598144531e-05, "grad_norm": 4.450837135314941, "learning_rate": 8.359474353539637e-06, "loss": 0.492, "mean_token_accuracy": 0.853679895401001, "num_tokens": 75078145.0, "step": 1973 }, { "epoch": 0.25111308993766696, "ewc_loss": 0.05810546875, "ewc_loss_diag": 3.5762786865234375e-05, "ewc_loss_parallel": 5.435943603515625e-05, "grad_norm": 4.586063861846924, "learning_rate": 8.363713437897416e-06, "loss": 0.584, "mean_token_accuracy": 0.8308508396148682, "num_tokens": 75126473.0, "step": 1974 }, { "epoch": 0.2512403002162575, "ewc_loss": 0.054443359375, "ewc_loss_diag": 3.5762786865234375e-05, "ewc_loss_parallel": 5.078315734863281e-05, "grad_norm": 4.330821990966797, "learning_rate": 8.367952522255195e-06, "loss": 0.5133, "mean_token_accuracy": 0.8474145531654358, "num_tokens": 75166785.0, "step": 1975 }, { "epoch": 0.25136751049484796, "ewc_loss": 0.05810546875, "ewc_loss_diag": 3.5762786865234375e-05, "ewc_loss_parallel": 5.435943603515625e-05, "grad_norm": 4.58306360244751, "learning_rate": 8.372191606612974e-06, "loss": 0.5209, "mean_token_accuracy": 0.8463834524154663, "num_tokens": 75204599.0, "step": 1976 }, { "epoch": 0.2514947207734385, "ewc_loss": 0.053466796875, "ewc_loss_diag": 3.600120544433594e-05, "ewc_loss_parallel": 4.982948303222656e-05, "grad_norm": 4.227005481719971, "learning_rate": 8.37643069097075e-06, "loss": 0.4725, "mean_token_accuracy": 0.8582367300987244, "num_tokens": 75235295.0, "step": 1977 }, { "epoch": 0.251621931052029, "ewc_loss": 0.05859375, "ewc_loss_diag": 3.600120544433594e-05, "ewc_loss_parallel": 5.4836273193359375e-05, "grad_norm": 4.640495300292969, "learning_rate": 8.38066977532853e-06, "loss": 0.482, "mean_token_accuracy": 0.8604706525802612, "num_tokens": 75275549.0, "step": 1978 }, { "epoch": 0.25174914133061954, "ewc_loss": 0.05224609375, "ewc_loss_diag": 3.62396240234375e-05, "ewc_loss_parallel": 4.863739013671875e-05, "grad_norm": 4.111713886260986, "learning_rate": 8.384908859686308e-06, "loss": 0.4748, "mean_token_accuracy": 0.8605577945709229, "num_tokens": 75316920.0, "step": 1979 }, { "epoch": 0.25187635160921, "ewc_loss": 0.06103515625, "ewc_loss_diag": 3.62396240234375e-05, "ewc_loss_parallel": 5.745887756347656e-05, "grad_norm": 4.797308921813965, "learning_rate": 8.389147944044087e-06, "loss": 0.521, "mean_token_accuracy": 0.8495213985443115, "num_tokens": 75360091.0, "step": 1980 }, { "epoch": 0.25200356188780054, "ewc_loss": 0.051513671875, "ewc_loss_diag": 3.62396240234375e-05, "ewc_loss_parallel": 4.792213439941406e-05, "grad_norm": 4.0250654220581055, "learning_rate": 8.393387028401866e-06, "loss": 0.5385, "mean_token_accuracy": 0.839295506477356, "num_tokens": 75397123.0, "step": 1981 }, { "epoch": 0.25213077216639107, "ewc_loss": 0.06591796875, "ewc_loss_diag": 3.62396240234375e-05, "ewc_loss_parallel": 6.246566772460938e-05, "grad_norm": 5.06833028793335, "learning_rate": 8.397626112759645e-06, "loss": 0.5779, "mean_token_accuracy": 0.8332173228263855, "num_tokens": 75429983.0, "step": 1982 }, { "epoch": 0.25225798244498154, "ewc_loss": 0.05322265625, "ewc_loss_diag": 3.62396240234375e-05, "ewc_loss_parallel": 4.9591064453125e-05, "grad_norm": 4.044698238372803, "learning_rate": 8.401865197117424e-06, "loss": 0.4951, "mean_token_accuracy": 0.8552147150039673, "num_tokens": 75460322.0, "step": 1983 }, { "epoch": 0.25238519272357207, "ewc_loss": 0.07080078125, "ewc_loss_diag": 3.62396240234375e-05, "ewc_loss_parallel": 6.723403930664062e-05, "grad_norm": 5.325287342071533, "learning_rate": 8.406104281475203e-06, "loss": 0.5005, "mean_token_accuracy": 0.8557317852973938, "num_tokens": 75499733.0, "step": 1984 }, { "epoch": 0.2525124030021626, "ewc_loss": 0.058349609375, "ewc_loss_diag": 3.62396240234375e-05, "ewc_loss_parallel": 5.459785461425781e-05, "grad_norm": 4.272007465362549, "learning_rate": 8.41034336583298e-06, "loss": 0.4942, "mean_token_accuracy": 0.8545406460762024, "num_tokens": 75541182.0, "step": 1985 }, { "epoch": 0.2526396132807531, "ewc_loss": 0.072265625, "ewc_loss_diag": 3.62396240234375e-05, "ewc_loss_parallel": 6.866455078125e-05, "grad_norm": 5.3938446044921875, "learning_rate": 8.414582450190759e-06, "loss": 0.5239, "mean_token_accuracy": 0.8491573333740234, "num_tokens": 75585052.0, "step": 1986 }, { "epoch": 0.2527668235593436, "ewc_loss": 0.06298828125, "ewc_loss_diag": 3.62396240234375e-05, "ewc_loss_parallel": 5.9604644775390625e-05, "grad_norm": 4.613109111785889, "learning_rate": 8.418821534548538e-06, "loss": 0.6039, "mean_token_accuracy": 0.8274120688438416, "num_tokens": 75625208.0, "step": 1987 }, { "epoch": 0.25289403383793413, "ewc_loss": 0.0673828125, "ewc_loss_diag": 3.647804260253906e-05, "ewc_loss_parallel": 6.389617919921875e-05, "grad_norm": 5.130674839019775, "learning_rate": 8.423060618906317e-06, "loss": 0.5276, "mean_token_accuracy": 0.8487722873687744, "num_tokens": 75665205.0, "step": 1988 }, { "epoch": 0.2530212441165246, "ewc_loss": 0.06298828125, "ewc_loss_diag": 3.647804260253906e-05, "ewc_loss_parallel": 5.9604644775390625e-05, "grad_norm": 4.690825462341309, "learning_rate": 8.427299703264096e-06, "loss": 0.5201, "mean_token_accuracy": 0.849102795124054, "num_tokens": 75708278.0, "step": 1989 }, { "epoch": 0.25314845439511513, "ewc_loss": 0.0625, "ewc_loss_diag": 3.647804260253906e-05, "ewc_loss_parallel": 5.91278076171875e-05, "grad_norm": 4.807666778564453, "learning_rate": 8.431538787621875e-06, "loss": 0.5402, "mean_token_accuracy": 0.8436595797538757, "num_tokens": 75750565.0, "step": 1990 }, { "epoch": 0.25327566467370566, "ewc_loss": 0.061279296875, "ewc_loss_diag": 3.647804260253906e-05, "ewc_loss_parallel": 5.7697296142578125e-05, "grad_norm": 4.6664886474609375, "learning_rate": 8.435777871979653e-06, "loss": 0.5028, "mean_token_accuracy": 0.8534963726997375, "num_tokens": 75790519.0, "step": 1991 }, { "epoch": 0.25340287495229613, "ewc_loss": 0.057861328125, "ewc_loss_diag": 3.62396240234375e-05, "ewc_loss_parallel": 5.412101745605469e-05, "grad_norm": 4.54557991027832, "learning_rate": 8.440016956337432e-06, "loss": 0.5275, "mean_token_accuracy": 0.8478127717971802, "num_tokens": 75828256.0, "step": 1992 }, { "epoch": 0.25353008523088666, "ewc_loss": 0.05908203125, "ewc_loss_diag": 3.647804260253906e-05, "ewc_loss_parallel": 5.53131103515625e-05, "grad_norm": 4.638792514801025, "learning_rate": 8.44425604069521e-06, "loss": 0.5409, "mean_token_accuracy": 0.847368061542511, "num_tokens": 75869238.0, "step": 1993 }, { "epoch": 0.2536572955094772, "ewc_loss": 0.0556640625, "ewc_loss_diag": 3.647804260253906e-05, "ewc_loss_parallel": 5.1975250244140625e-05, "grad_norm": 4.352779865264893, "learning_rate": 8.448495125052988e-06, "loss": 0.5274, "mean_token_accuracy": 0.8458612561225891, "num_tokens": 75906776.0, "step": 1994 }, { "epoch": 0.25378450578806766, "ewc_loss": 0.059326171875, "ewc_loss_diag": 3.647804260253906e-05, "ewc_loss_parallel": 5.555152893066406e-05, "grad_norm": 4.647729873657227, "learning_rate": 8.452734209410767e-06, "loss": 0.463, "mean_token_accuracy": 0.865339994430542, "num_tokens": 75940271.0, "step": 1995 }, { "epoch": 0.2539117160666582, "ewc_loss": 0.0537109375, "ewc_loss_diag": 3.62396240234375e-05, "ewc_loss_parallel": 5.0067901611328125e-05, "grad_norm": 4.221621990203857, "learning_rate": 8.456973293768546e-06, "loss": 0.491, "mean_token_accuracy": 0.8543472290039062, "num_tokens": 75974536.0, "step": 1996 }, { "epoch": 0.2540389263452487, "ewc_loss": 0.059326171875, "ewc_loss_diag": 3.6716461181640625e-05, "ewc_loss_parallel": 5.555152893066406e-05, "grad_norm": 4.809370994567871, "learning_rate": 8.461212378126325e-06, "loss": 0.5495, "mean_token_accuracy": 0.8405104875564575, "num_tokens": 76013809.0, "step": 1997 }, { "epoch": 0.2541661366238392, "ewc_loss": 0.052978515625, "ewc_loss_diag": 3.6716461181640625e-05, "ewc_loss_parallel": 4.935264587402344e-05, "grad_norm": 4.103968620300293, "learning_rate": 8.465451462484104e-06, "loss": 0.5699, "mean_token_accuracy": 0.8335355520248413, "num_tokens": 76049559.0, "step": 1998 }, { "epoch": 0.2542933469024297, "ewc_loss": 0.064453125, "ewc_loss_diag": 3.6716461181640625e-05, "ewc_loss_parallel": 6.079673767089844e-05, "grad_norm": 4.931816101074219, "learning_rate": 8.469690546841883e-06, "loss": 0.5233, "mean_token_accuracy": 0.8475945591926575, "num_tokens": 76087091.0, "step": 1999 }, { "epoch": 0.25442055718102025, "ewc_loss": 0.052978515625, "ewc_loss_diag": 3.6716461181640625e-05, "ewc_loss_parallel": 4.935264587402344e-05, "grad_norm": 4.052725791931152, "learning_rate": 8.473929631199662e-06, "loss": 0.4883, "mean_token_accuracy": 0.854809582233429, "num_tokens": 76121466.0, "step": 2000 }, { "epoch": 0.2545477674596107, "ewc_loss": 0.06884765625, "ewc_loss_diag": 3.695487976074219e-05, "ewc_loss_parallel": 6.4849853515625e-05, "grad_norm": 5.149065971374512, "learning_rate": 8.47816871555744e-06, "loss": 0.525, "mean_token_accuracy": 0.8538854122161865, "num_tokens": 76158291.0, "step": 2001 }, { "epoch": 0.25467497773820125, "ewc_loss": 0.055908203125, "ewc_loss_diag": 3.695487976074219e-05, "ewc_loss_parallel": 5.221366882324219e-05, "grad_norm": 4.140809535980225, "learning_rate": 8.48240779991522e-06, "loss": 0.5249, "mean_token_accuracy": 0.8444479703903198, "num_tokens": 76197350.0, "step": 2002 }, { "epoch": 0.2548021880167918, "ewc_loss": 0.072265625, "ewc_loss_diag": 3.695487976074219e-05, "ewc_loss_parallel": 6.818771362304688e-05, "grad_norm": 5.319915294647217, "learning_rate": 8.486646884272998e-06, "loss": 0.5341, "mean_token_accuracy": 0.846007227897644, "num_tokens": 76235367.0, "step": 2003 }, { "epoch": 0.25492939829538225, "ewc_loss": 0.060302734375, "ewc_loss_diag": 3.719329833984375e-05, "ewc_loss_parallel": 5.6743621826171875e-05, "grad_norm": 4.352539539337158, "learning_rate": 8.490885968630777e-06, "loss": 0.5562, "mean_token_accuracy": 0.8366241455078125, "num_tokens": 76270462.0, "step": 2004 }, { "epoch": 0.2550566085739728, "ewc_loss": 0.072265625, "ewc_loss_diag": 3.719329833984375e-05, "ewc_loss_parallel": 6.818771362304688e-05, "grad_norm": 5.310552597045898, "learning_rate": 8.495125052988556e-06, "loss": 0.5567, "mean_token_accuracy": 0.8446513414382935, "num_tokens": 76315032.0, "step": 2005 }, { "epoch": 0.2551838188525633, "ewc_loss": 0.0634765625, "ewc_loss_diag": 3.719329833984375e-05, "ewc_loss_parallel": 5.9604644775390625e-05, "grad_norm": 4.542507171630859, "learning_rate": 8.499364137346335e-06, "loss": 0.5298, "mean_token_accuracy": 0.8490753173828125, "num_tokens": 76356337.0, "step": 2006 }, { "epoch": 0.2553110291311538, "ewc_loss": 0.06884765625, "ewc_loss_diag": 3.719329833984375e-05, "ewc_loss_parallel": 6.4849853515625e-05, "grad_norm": 5.133205413818359, "learning_rate": 8.503603221704114e-06, "loss": 0.5575, "mean_token_accuracy": 0.8392097353935242, "num_tokens": 76391329.0, "step": 2007 }, { "epoch": 0.2554382394097443, "ewc_loss": 0.06298828125, "ewc_loss_diag": 3.719329833984375e-05, "ewc_loss_parallel": 5.936622619628906e-05, "grad_norm": 4.6114325523376465, "learning_rate": 8.507842306061891e-06, "loss": 0.5139, "mean_token_accuracy": 0.849992036819458, "num_tokens": 76429015.0, "step": 2008 }, { "epoch": 0.25556544968833483, "ewc_loss": 0.0654296875, "ewc_loss_diag": 3.719329833984375e-05, "ewc_loss_parallel": 6.151199340820312e-05, "grad_norm": 4.895371913909912, "learning_rate": 8.51208139041967e-06, "loss": 0.5518, "mean_token_accuracy": 0.8479918241500854, "num_tokens": 76470343.0, "step": 2009 }, { "epoch": 0.2556926599669253, "ewc_loss": 0.061767578125, "ewc_loss_diag": 3.719329833984375e-05, "ewc_loss_parallel": 5.817413330078125e-05, "grad_norm": 4.587342739105225, "learning_rate": 8.516320474777449e-06, "loss": 0.5324, "mean_token_accuracy": 0.8473114967346191, "num_tokens": 76505140.0, "step": 2010 }, { "epoch": 0.25581987024551583, "ewc_loss": 0.06103515625, "ewc_loss_diag": 3.719329833984375e-05, "ewc_loss_parallel": 5.745887756347656e-05, "grad_norm": 4.680964469909668, "learning_rate": 8.520559559135228e-06, "loss": 0.488, "mean_token_accuracy": 0.8621218800544739, "num_tokens": 76541701.0, "step": 2011 }, { "epoch": 0.25594708052410636, "ewc_loss": 0.05908203125, "ewc_loss_diag": 3.719329833984375e-05, "ewc_loss_parallel": 5.53131103515625e-05, "grad_norm": 4.541444301605225, "learning_rate": 8.524798643493007e-06, "loss": 0.5482, "mean_token_accuracy": 0.8459244966506958, "num_tokens": 76580208.0, "step": 2012 }, { "epoch": 0.25607429080269684, "ewc_loss": 0.05859375, "ewc_loss_diag": 3.719329833984375e-05, "ewc_loss_parallel": 5.4836273193359375e-05, "grad_norm": 4.533412456512451, "learning_rate": 8.529037727850785e-06, "loss": 0.4936, "mean_token_accuracy": 0.8555073738098145, "num_tokens": 76617750.0, "step": 2013 }, { "epoch": 0.25620150108128736, "ewc_loss": 0.0576171875, "ewc_loss_diag": 3.743171691894531e-05, "ewc_loss_parallel": 5.3882598876953125e-05, "grad_norm": 4.489477634429932, "learning_rate": 8.533276812208564e-06, "loss": 0.4636, "mean_token_accuracy": 0.8632235527038574, "num_tokens": 76652678.0, "step": 2014 }, { "epoch": 0.2563287113598779, "ewc_loss": 0.0576171875, "ewc_loss_diag": 3.743171691894531e-05, "ewc_loss_parallel": 5.3882598876953125e-05, "grad_norm": 4.449672222137451, "learning_rate": 8.537515896566342e-06, "loss": 0.4985, "mean_token_accuracy": 0.852332353591919, "num_tokens": 76695375.0, "step": 2015 }, { "epoch": 0.25645592163846836, "ewc_loss": 0.056640625, "ewc_loss_diag": 3.743171691894531e-05, "ewc_loss_parallel": 5.2928924560546875e-05, "grad_norm": 4.462967872619629, "learning_rate": 8.54175498092412e-06, "loss": 0.5098, "mean_token_accuracy": 0.8513387441635132, "num_tokens": 76728695.0, "step": 2016 }, { "epoch": 0.2565831319170589, "ewc_loss": 0.056396484375, "ewc_loss_diag": 3.743171691894531e-05, "ewc_loss_parallel": 5.269050598144531e-05, "grad_norm": 4.405852794647217, "learning_rate": 8.5459940652819e-06, "loss": 0.484, "mean_token_accuracy": 0.8598238825798035, "num_tokens": 76761081.0, "step": 2017 }, { "epoch": 0.2567103421956494, "ewc_loss": 0.056640625, "ewc_loss_diag": 3.743171691894531e-05, "ewc_loss_parallel": 5.2928924560546875e-05, "grad_norm": 4.426926136016846, "learning_rate": 8.550233149639678e-06, "loss": 0.5803, "mean_token_accuracy": 0.8289281129837036, "num_tokens": 76800335.0, "step": 2018 }, { "epoch": 0.2568375524742399, "ewc_loss": 0.055419921875, "ewc_loss_diag": 3.743171691894531e-05, "ewc_loss_parallel": 5.173683166503906e-05, "grad_norm": 4.365833282470703, "learning_rate": 8.554472233997457e-06, "loss": 0.5429, "mean_token_accuracy": 0.839128315448761, "num_tokens": 76839653.0, "step": 2019 }, { "epoch": 0.2569647627528304, "ewc_loss": 0.056884765625, "ewc_loss_diag": 3.743171691894531e-05, "ewc_loss_parallel": 5.316734313964844e-05, "grad_norm": 4.426000595092773, "learning_rate": 8.558711318355236e-06, "loss": 0.5752, "mean_token_accuracy": 0.8281567692756653, "num_tokens": 76876557.0, "step": 2020 }, { "epoch": 0.25709197303142095, "ewc_loss": 0.05517578125, "ewc_loss_diag": 3.743171691894531e-05, "ewc_loss_parallel": 5.14984130859375e-05, "grad_norm": 4.342251777648926, "learning_rate": 8.562950402713015e-06, "loss": 0.465, "mean_token_accuracy": 0.8683956265449524, "num_tokens": 76910000.0, "step": 2021 }, { "epoch": 0.2572191833100114, "ewc_loss": 0.056884765625, "ewc_loss_diag": 3.743171691894531e-05, "ewc_loss_parallel": 5.316734313964844e-05, "grad_norm": 4.419568061828613, "learning_rate": 8.567189487070794e-06, "loss": 0.5584, "mean_token_accuracy": 0.8382337093353271, "num_tokens": 76946861.0, "step": 2022 }, { "epoch": 0.25734639358860195, "ewc_loss": 0.054443359375, "ewc_loss_diag": 3.743171691894531e-05, "ewc_loss_parallel": 5.078315734863281e-05, "grad_norm": 4.313060283660889, "learning_rate": 8.571428571428571e-06, "loss": 0.5723, "mean_token_accuracy": 0.8357239961624146, "num_tokens": 76990831.0, "step": 2023 }, { "epoch": 0.2574736038671925, "ewc_loss": 0.05712890625, "ewc_loss_diag": 3.743171691894531e-05, "ewc_loss_parallel": 5.340576171875e-05, "grad_norm": 4.431861877441406, "learning_rate": 8.57566765578635e-06, "loss": 0.4777, "mean_token_accuracy": 0.8628016710281372, "num_tokens": 77033145.0, "step": 2024 }, { "epoch": 0.25760081414578295, "ewc_loss": 0.054443359375, "ewc_loss_diag": 3.743171691894531e-05, "ewc_loss_parallel": 5.078315734863281e-05, "grad_norm": 4.277463436126709, "learning_rate": 8.579906740144129e-06, "loss": 0.5088, "mean_token_accuracy": 0.8523378372192383, "num_tokens": 77076621.0, "step": 2025 }, { "epoch": 0.2577280244243735, "ewc_loss": 0.057373046875, "ewc_loss_diag": 3.743171691894531e-05, "ewc_loss_parallel": 5.364418029785156e-05, "grad_norm": 4.489561557769775, "learning_rate": 8.584145824501908e-06, "loss": 0.547, "mean_token_accuracy": 0.8390101790428162, "num_tokens": 77115951.0, "step": 2026 }, { "epoch": 0.257855234702964, "ewc_loss": 0.054443359375, "ewc_loss_diag": 3.743171691894531e-05, "ewc_loss_parallel": 5.078315734863281e-05, "grad_norm": 4.2171783447265625, "learning_rate": 8.588384908859686e-06, "loss": 0.5417, "mean_token_accuracy": 0.8443832397460938, "num_tokens": 77154674.0, "step": 2027 }, { "epoch": 0.25798244498155454, "ewc_loss": 0.05859375, "ewc_loss_diag": 3.743171691894531e-05, "ewc_loss_parallel": 5.4836273193359375e-05, "grad_norm": 4.577738285064697, "learning_rate": 8.592623993217465e-06, "loss": 0.5261, "mean_token_accuracy": 0.8481175899505615, "num_tokens": 77192970.0, "step": 2028 }, { "epoch": 0.258109655260145, "ewc_loss": 0.05224609375, "ewc_loss_diag": 3.743171691894531e-05, "ewc_loss_parallel": 4.863739013671875e-05, "grad_norm": 4.125927925109863, "learning_rate": 8.596863077575244e-06, "loss": 0.5523, "mean_token_accuracy": 0.8401109576225281, "num_tokens": 77231258.0, "step": 2029 }, { "epoch": 0.25823686553873554, "ewc_loss": 0.060302734375, "ewc_loss_diag": 3.743171691894531e-05, "ewc_loss_parallel": 5.6743621826171875e-05, "grad_norm": 4.759513854980469, "learning_rate": 8.601102161933023e-06, "loss": 0.535, "mean_token_accuracy": 0.8455597162246704, "num_tokens": 77269829.0, "step": 2030 }, { "epoch": 0.25836407581732607, "ewc_loss": 0.051513671875, "ewc_loss_diag": 3.743171691894531e-05, "ewc_loss_parallel": 4.792213439941406e-05, "grad_norm": 3.976764440536499, "learning_rate": 8.605341246290802e-06, "loss": 0.505, "mean_token_accuracy": 0.850063681602478, "num_tokens": 77309611.0, "step": 2031 }, { "epoch": 0.25849128609591654, "ewc_loss": 0.06787109375, "ewc_loss_diag": 3.743171691894531e-05, "ewc_loss_parallel": 6.389617919921875e-05, "grad_norm": 5.114315986633301, "learning_rate": 8.609580330648581e-06, "loss": 0.5154, "mean_token_accuracy": 0.8541195392608643, "num_tokens": 77341937.0, "step": 2032 }, { "epoch": 0.25861849637450707, "ewc_loss": 0.0517578125, "ewc_loss_diag": 3.743171691894531e-05, "ewc_loss_parallel": 4.8160552978515625e-05, "grad_norm": 3.891552448272705, "learning_rate": 8.61381941500636e-06, "loss": 0.567, "mean_token_accuracy": 0.837472677230835, "num_tokens": 77380484.0, "step": 2033 }, { "epoch": 0.2587457066530976, "ewc_loss": 0.07666015625, "ewc_loss_diag": 3.7670135498046875e-05, "ewc_loss_parallel": 7.295608520507812e-05, "grad_norm": 5.653844356536865, "learning_rate": 8.618058499364139e-06, "loss": 0.5863, "mean_token_accuracy": 0.8342894315719604, "num_tokens": 77421220.0, "step": 2034 }, { "epoch": 0.25887291693168807, "ewc_loss": 0.0595703125, "ewc_loss_diag": 3.7670135498046875e-05, "ewc_loss_parallel": 5.5789947509765625e-05, "grad_norm": 4.173290729522705, "learning_rate": 8.622297583721918e-06, "loss": 0.5154, "mean_token_accuracy": 0.854220986366272, "num_tokens": 77465136.0, "step": 2035 }, { "epoch": 0.2590001272102786, "ewc_loss": 0.083984375, "ewc_loss_diag": 3.7670135498046875e-05, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 6.067989826202393, "learning_rate": 8.626536668079696e-06, "loss": 0.5516, "mean_token_accuracy": 0.8458698391914368, "num_tokens": 77500120.0, "step": 2036 }, { "epoch": 0.2591273374888691, "ewc_loss": 0.07421875, "ewc_loss_diag": 3.7670135498046875e-05, "ewc_loss_parallel": 7.009506225585938e-05, "grad_norm": 4.979894638061523, "learning_rate": 8.630775752437475e-06, "loss": 0.5281, "mean_token_accuracy": 0.8490958213806152, "num_tokens": 77532860.0, "step": 2037 }, { "epoch": 0.2592545477674596, "ewc_loss": 0.078125, "ewc_loss_diag": 3.7670135498046875e-05, "ewc_loss_parallel": 7.43865966796875e-05, "grad_norm": 5.703977584838867, "learning_rate": 8.635014836795252e-06, "loss": 0.5693, "mean_token_accuracy": 0.8404043316841125, "num_tokens": 77569154.0, "step": 2038 }, { "epoch": 0.2593817580460501, "ewc_loss": 0.07421875, "ewc_loss_diag": 3.7670135498046875e-05, "ewc_loss_parallel": 7.009506225585938e-05, "grad_norm": 5.188782215118408, "learning_rate": 8.639253921153031e-06, "loss": 0.4966, "mean_token_accuracy": 0.8578816056251526, "num_tokens": 77602918.0, "step": 2039 }, { "epoch": 0.25950896832464065, "ewc_loss": 0.0703125, "ewc_loss_diag": 3.7670135498046875e-05, "ewc_loss_parallel": 6.628036499023438e-05, "grad_norm": 5.280862808227539, "learning_rate": 8.64349300551081e-06, "loss": 0.5114, "mean_token_accuracy": 0.8537619113922119, "num_tokens": 77637233.0, "step": 2040 }, { "epoch": 0.2596361786032311, "ewc_loss": 0.068359375, "ewc_loss_diag": 3.7670135498046875e-05, "ewc_loss_parallel": 6.437301635742188e-05, "grad_norm": 4.950799942016602, "learning_rate": 8.647732089868589e-06, "loss": 0.5905, "mean_token_accuracy": 0.8304316997528076, "num_tokens": 77678344.0, "step": 2041 }, { "epoch": 0.25976338888182166, "ewc_loss": 0.0673828125, "ewc_loss_diag": 3.7670135498046875e-05, "ewc_loss_parallel": 6.341934204101562e-05, "grad_norm": 5.043264865875244, "learning_rate": 8.651971174226368e-06, "loss": 0.4997, "mean_token_accuracy": 0.854923665523529, "num_tokens": 77712706.0, "step": 2042 }, { "epoch": 0.2598905991604122, "ewc_loss": 0.06396484375, "ewc_loss_diag": 3.7670135498046875e-05, "ewc_loss_parallel": 6.031990051269531e-05, "grad_norm": 4.747499942779541, "learning_rate": 8.656210258584147e-06, "loss": 0.5606, "mean_token_accuracy": 0.8360222578048706, "num_tokens": 77749841.0, "step": 2043 }, { "epoch": 0.26001780943900266, "ewc_loss": 0.0654296875, "ewc_loss_diag": 3.7670135498046875e-05, "ewc_loss_parallel": 6.151199340820312e-05, "grad_norm": 4.9116315841674805, "learning_rate": 8.660449342941926e-06, "loss": 0.5621, "mean_token_accuracy": 0.836815357208252, "num_tokens": 77788464.0, "step": 2044 }, { "epoch": 0.2601450197175932, "ewc_loss": 0.0595703125, "ewc_loss_diag": 3.7670135498046875e-05, "ewc_loss_parallel": 5.602836608886719e-05, "grad_norm": 4.5710625648498535, "learning_rate": 8.664688427299705e-06, "loss": 0.53, "mean_token_accuracy": 0.8435373306274414, "num_tokens": 77826493.0, "step": 2045 }, { "epoch": 0.2602722299961837, "ewc_loss": 0.0634765625, "ewc_loss_diag": 3.7670135498046875e-05, "ewc_loss_parallel": 5.9604644775390625e-05, "grad_norm": 4.859702110290527, "learning_rate": 8.668927511657482e-06, "loss": 0.4918, "mean_token_accuracy": 0.8560213446617126, "num_tokens": 77857814.0, "step": 2046 }, { "epoch": 0.2603994402747742, "ewc_loss": 0.05712890625, "ewc_loss_diag": 3.7670135498046875e-05, "ewc_loss_parallel": 5.340576171875e-05, "grad_norm": 4.426565647125244, "learning_rate": 8.67316659601526e-06, "loss": 0.5327, "mean_token_accuracy": 0.844889760017395, "num_tokens": 77891263.0, "step": 2047 }, { "epoch": 0.2605266505533647, "ewc_loss": 0.06396484375, "ewc_loss_diag": 3.7670135498046875e-05, "ewc_loss_parallel": 6.008148193359375e-05, "grad_norm": 4.89874792098999, "learning_rate": 8.67740568037304e-06, "loss": 0.5083, "mean_token_accuracy": 0.8565021753311157, "num_tokens": 77931683.0, "step": 2048 }, { "epoch": 0.26065386083195524, "ewc_loss": 0.0556640625, "ewc_loss_diag": 3.7670135498046875e-05, "ewc_loss_parallel": 5.1975250244140625e-05, "grad_norm": 4.282475471496582, "learning_rate": 8.681644764730819e-06, "loss": 0.5252, "mean_token_accuracy": 0.8454254269599915, "num_tokens": 77965204.0, "step": 2049 }, { "epoch": 0.2607810711105457, "ewc_loss": 0.06591796875, "ewc_loss_diag": 3.814697265625e-05, "ewc_loss_parallel": 6.198883056640625e-05, "grad_norm": 5.05312967300415, "learning_rate": 8.685883849088597e-06, "loss": 0.4766, "mean_token_accuracy": 0.8644591569900513, "num_tokens": 78003427.0, "step": 2050 }, { "epoch": 0.26090828138913624, "ewc_loss": 0.0546875, "ewc_loss_diag": 3.814697265625e-05, "ewc_loss_parallel": 5.078315734863281e-05, "grad_norm": 4.195130348205566, "learning_rate": 8.690122933446376e-06, "loss": 0.5221, "mean_token_accuracy": 0.8493934869766235, "num_tokens": 78045524.0, "step": 2051 }, { "epoch": 0.26103549166772677, "ewc_loss": 0.07177734375, "ewc_loss_diag": 3.814697265625e-05, "ewc_loss_parallel": 6.771087646484375e-05, "grad_norm": 5.311459064483643, "learning_rate": 8.694362017804155e-06, "loss": 0.5146, "mean_token_accuracy": 0.8525463342666626, "num_tokens": 78085485.0, "step": 2052 }, { "epoch": 0.26116270194631724, "ewc_loss": 0.05810546875, "ewc_loss_diag": 3.814697265625e-05, "ewc_loss_parallel": 5.412101745605469e-05, "grad_norm": 4.28193473815918, "learning_rate": 8.698601102161932e-06, "loss": 0.5341, "mean_token_accuracy": 0.8425642848014832, "num_tokens": 78128212.0, "step": 2053 }, { "epoch": 0.26128991222490777, "ewc_loss": 0.07421875, "ewc_loss_diag": 3.814697265625e-05, "ewc_loss_parallel": 7.009506225585938e-05, "grad_norm": 5.507245063781738, "learning_rate": 8.702840186519711e-06, "loss": 0.5034, "mean_token_accuracy": 0.8585773706436157, "num_tokens": 78170959.0, "step": 2054 }, { "epoch": 0.2614171225034983, "ewc_loss": 0.0634765625, "ewc_loss_diag": 3.814697265625e-05, "ewc_loss_parallel": 5.984306335449219e-05, "grad_norm": 4.565584182739258, "learning_rate": 8.70707927087749e-06, "loss": 0.4916, "mean_token_accuracy": 0.8574700355529785, "num_tokens": 78211181.0, "step": 2055 }, { "epoch": 0.2615443327820888, "ewc_loss": 0.07275390625, "ewc_loss_diag": 3.814697265625e-05, "ewc_loss_parallel": 6.866455078125e-05, "grad_norm": 5.425472736358643, "learning_rate": 8.711318355235269e-06, "loss": 0.5871, "mean_token_accuracy": 0.828460693359375, "num_tokens": 78246281.0, "step": 2056 }, { "epoch": 0.2616715430606793, "ewc_loss": 0.06591796875, "ewc_loss_diag": 3.838539123535156e-05, "ewc_loss_parallel": 6.198883056640625e-05, "grad_norm": 4.814415454864502, "learning_rate": 8.715557439593048e-06, "loss": 0.4974, "mean_token_accuracy": 0.8548254370689392, "num_tokens": 78279604.0, "step": 2057 }, { "epoch": 0.26179875333926983, "ewc_loss": 0.06787109375, "ewc_loss_diag": 3.838539123535156e-05, "ewc_loss_parallel": 6.389617919921875e-05, "grad_norm": 5.083785057067871, "learning_rate": 8.719796523950827e-06, "loss": 0.5845, "mean_token_accuracy": 0.8294647932052612, "num_tokens": 78317534.0, "step": 2058 }, { "epoch": 0.2619259636178603, "ewc_loss": 0.06591796875, "ewc_loss_diag": 3.838539123535156e-05, "ewc_loss_parallel": 6.198883056640625e-05, "grad_norm": 4.845610618591309, "learning_rate": 8.724035608308606e-06, "loss": 0.4619, "mean_token_accuracy": 0.8685922622680664, "num_tokens": 78349744.0, "step": 2059 }, { "epoch": 0.26205317389645083, "ewc_loss": 0.0615234375, "ewc_loss_diag": 3.838539123535156e-05, "ewc_loss_parallel": 5.7697296142578125e-05, "grad_norm": 4.703319072723389, "learning_rate": 8.728274692666385e-06, "loss": 0.5302, "mean_token_accuracy": 0.8447553515434265, "num_tokens": 78390841.0, "step": 2060 }, { "epoch": 0.26218038417504136, "ewc_loss": 0.06396484375, "ewc_loss_diag": 3.838539123535156e-05, "ewc_loss_parallel": 6.008148193359375e-05, "grad_norm": 4.852642059326172, "learning_rate": 8.732513777024163e-06, "loss": 0.5275, "mean_token_accuracy": 0.8482420444488525, "num_tokens": 78425604.0, "step": 2061 }, { "epoch": 0.26230759445363183, "ewc_loss": 0.057373046875, "ewc_loss_diag": 3.838539123535156e-05, "ewc_loss_parallel": 5.340576171875e-05, "grad_norm": 4.419988632202148, "learning_rate": 8.736752861381942e-06, "loss": 0.568, "mean_token_accuracy": 0.8326594829559326, "num_tokens": 78469444.0, "step": 2062 }, { "epoch": 0.26243480473222236, "ewc_loss": 0.06396484375, "ewc_loss_diag": 3.838539123535156e-05, "ewc_loss_parallel": 6.031990051269531e-05, "grad_norm": 4.9318718910217285, "learning_rate": 8.740991945739721e-06, "loss": 0.5235, "mean_token_accuracy": 0.8474670648574829, "num_tokens": 78507931.0, "step": 2063 }, { "epoch": 0.2625620150108129, "ewc_loss": 0.055419921875, "ewc_loss_diag": 3.838539123535156e-05, "ewc_loss_parallel": 5.14984130859375e-05, "grad_norm": 4.2348856925964355, "learning_rate": 8.7452310300975e-06, "loss": 0.5104, "mean_token_accuracy": 0.8509837985038757, "num_tokens": 78550572.0, "step": 2064 }, { "epoch": 0.26268922528940336, "ewc_loss": 0.06787109375, "ewc_loss_diag": 3.838539123535156e-05, "ewc_loss_parallel": 6.389617919921875e-05, "grad_norm": 5.124270439147949, "learning_rate": 8.749470114455279e-06, "loss": 0.5007, "mean_token_accuracy": 0.8566863536834717, "num_tokens": 78586975.0, "step": 2065 }, { "epoch": 0.2628164355679939, "ewc_loss": 0.05517578125, "ewc_loss_diag": 3.838539123535156e-05, "ewc_loss_parallel": 5.125999450683594e-05, "grad_norm": 4.159778118133545, "learning_rate": 8.753709198813058e-06, "loss": 0.5857, "mean_token_accuracy": 0.8283036947250366, "num_tokens": 78630535.0, "step": 2066 }, { "epoch": 0.2629436458465844, "ewc_loss": 0.072265625, "ewc_loss_diag": 3.838539123535156e-05, "ewc_loss_parallel": 6.818771362304688e-05, "grad_norm": 5.399853229522705, "learning_rate": 8.757948283170837e-06, "loss": 0.5071, "mean_token_accuracy": 0.8597261309623718, "num_tokens": 78672486.0, "step": 2067 }, { "epoch": 0.2630708561251749, "ewc_loss": 0.058837890625, "ewc_loss_diag": 3.838539123535156e-05, "ewc_loss_parallel": 5.4836273193359375e-05, "grad_norm": 4.297210216522217, "learning_rate": 8.762187367528614e-06, "loss": 0.486, "mean_token_accuracy": 0.8572609424591064, "num_tokens": 78710208.0, "step": 2068 }, { "epoch": 0.2631980664037654, "ewc_loss": 0.0751953125, "ewc_loss_diag": 3.838539123535156e-05, "ewc_loss_parallel": 7.104873657226562e-05, "grad_norm": 5.567534923553467, "learning_rate": 8.766426451886393e-06, "loss": 0.5525, "mean_token_accuracy": 0.8425343036651611, "num_tokens": 78754233.0, "step": 2069 }, { "epoch": 0.26332527668235595, "ewc_loss": 0.064453125, "ewc_loss_diag": 3.838539123535156e-05, "ewc_loss_parallel": 6.0558319091796875e-05, "grad_norm": 4.5985941886901855, "learning_rate": 8.770665536244172e-06, "loss": 0.5756, "mean_token_accuracy": 0.8372137546539307, "num_tokens": 78794530.0, "step": 2070 }, { "epoch": 0.2634524869609464, "ewc_loss": 0.07373046875, "ewc_loss_diag": 3.838539123535156e-05, "ewc_loss_parallel": 6.961822509765625e-05, "grad_norm": 5.465675354003906, "learning_rate": 8.77490462060195e-06, "loss": 0.5013, "mean_token_accuracy": 0.856977641582489, "num_tokens": 78828620.0, "step": 2071 }, { "epoch": 0.26357969723953695, "ewc_loss": 0.06640625, "ewc_loss_diag": 3.838539123535156e-05, "ewc_loss_parallel": 6.246566772460938e-05, "grad_norm": 4.788087368011475, "learning_rate": 8.77914370495973e-06, "loss": 0.4973, "mean_token_accuracy": 0.8562362194061279, "num_tokens": 78867658.0, "step": 2072 }, { "epoch": 0.2637069075181275, "ewc_loss": 0.0703125, "ewc_loss_diag": 3.838539123535156e-05, "ewc_loss_parallel": 6.628036499023438e-05, "grad_norm": 5.204908847808838, "learning_rate": 8.783382789317508e-06, "loss": 0.5426, "mean_token_accuracy": 0.8462011814117432, "num_tokens": 78903744.0, "step": 2073 }, { "epoch": 0.26383411779671795, "ewc_loss": 0.0654296875, "ewc_loss_diag": 3.838539123535156e-05, "ewc_loss_parallel": 6.151199340820312e-05, "grad_norm": 4.789918422698975, "learning_rate": 8.787621873675287e-06, "loss": 0.5603, "mean_token_accuracy": 0.8345974683761597, "num_tokens": 78934098.0, "step": 2074 }, { "epoch": 0.2639613280753085, "ewc_loss": 0.06689453125, "ewc_loss_diag": 3.838539123535156e-05, "ewc_loss_parallel": 6.29425048828125e-05, "grad_norm": 4.959383487701416, "learning_rate": 8.791860958033066e-06, "loss": 0.5004, "mean_token_accuracy": 0.8558889627456665, "num_tokens": 78969473.0, "step": 2075 }, { "epoch": 0.264088538353899, "ewc_loss": 0.06298828125, "ewc_loss_diag": 3.8623809814453125e-05, "ewc_loss_parallel": 5.91278076171875e-05, "grad_norm": 4.693235874176025, "learning_rate": 8.796100042390843e-06, "loss": 0.4937, "mean_token_accuracy": 0.8591791391372681, "num_tokens": 79008799.0, "step": 2076 }, { "epoch": 0.2642157486324895, "ewc_loss": 0.06298828125, "ewc_loss_diag": 3.8623809814453125e-05, "ewc_loss_parallel": 5.936622619628906e-05, "grad_norm": 4.802280426025391, "learning_rate": 8.800339126748622e-06, "loss": 0.5017, "mean_token_accuracy": 0.8530822396278381, "num_tokens": 79042502.0, "step": 2077 }, { "epoch": 0.26434295891108, "ewc_loss": 0.06005859375, "ewc_loss_diag": 3.8623809814453125e-05, "ewc_loss_parallel": 5.626678466796875e-05, "grad_norm": 4.604338645935059, "learning_rate": 8.804578211106401e-06, "loss": 0.5691, "mean_token_accuracy": 0.8313616514205933, "num_tokens": 79079641.0, "step": 2078 }, { "epoch": 0.26447016918967053, "ewc_loss": 0.062255859375, "ewc_loss_diag": 3.8623809814453125e-05, "ewc_loss_parallel": 5.841255187988281e-05, "grad_norm": 4.719899654388428, "learning_rate": 8.80881729546418e-06, "loss": 0.5192, "mean_token_accuracy": 0.8504689335823059, "num_tokens": 79116120.0, "step": 2079 }, { "epoch": 0.26459737946826106, "ewc_loss": 0.058349609375, "ewc_loss_diag": 3.8623809814453125e-05, "ewc_loss_parallel": 5.435943603515625e-05, "grad_norm": 4.510500431060791, "learning_rate": 8.813056379821959e-06, "loss": 0.4968, "mean_token_accuracy": 0.8584309816360474, "num_tokens": 79150322.0, "step": 2080 }, { "epoch": 0.26472458974685154, "ewc_loss": 0.060791015625, "ewc_loss_diag": 3.8623809814453125e-05, "ewc_loss_parallel": 5.698204040527344e-05, "grad_norm": 4.714634895324707, "learning_rate": 8.817295464179738e-06, "loss": 0.5262, "mean_token_accuracy": 0.8456630706787109, "num_tokens": 79184944.0, "step": 2081 }, { "epoch": 0.26485180002544206, "ewc_loss": 0.056884765625, "ewc_loss_diag": 3.8623809814453125e-05, "ewc_loss_parallel": 5.2928924560546875e-05, "grad_norm": 4.408426284790039, "learning_rate": 8.821534548537517e-06, "loss": 0.5898, "mean_token_accuracy": 0.8236420750617981, "num_tokens": 79226137.0, "step": 2082 }, { "epoch": 0.2649790103040326, "ewc_loss": 0.0615234375, "ewc_loss_diag": 3.8623809814453125e-05, "ewc_loss_parallel": 5.7697296142578125e-05, "grad_norm": 4.763366222381592, "learning_rate": 8.825773632895296e-06, "loss": 0.4991, "mean_token_accuracy": 0.855860710144043, "num_tokens": 79267721.0, "step": 2083 }, { "epoch": 0.26510622058262306, "ewc_loss": 0.055419921875, "ewc_loss_diag": 3.8623809814453125e-05, "ewc_loss_parallel": 5.14984130859375e-05, "grad_norm": 4.317591667175293, "learning_rate": 8.830012717253074e-06, "loss": 0.4978, "mean_token_accuracy": 0.8548109531402588, "num_tokens": 79305355.0, "step": 2084 }, { "epoch": 0.2652334308612136, "ewc_loss": 0.06396484375, "ewc_loss_diag": 3.8623809814453125e-05, "ewc_loss_parallel": 6.008148193359375e-05, "grad_norm": 4.879908084869385, "learning_rate": 8.834251801610853e-06, "loss": 0.5403, "mean_token_accuracy": 0.84255450963974, "num_tokens": 79349714.0, "step": 2085 }, { "epoch": 0.2653606411398041, "ewc_loss": 0.05517578125, "ewc_loss_diag": 3.8623809814453125e-05, "ewc_loss_parallel": 5.125999450683594e-05, "grad_norm": 4.209295749664307, "learning_rate": 8.838490885968632e-06, "loss": 0.4834, "mean_token_accuracy": 0.858190655708313, "num_tokens": 79387208.0, "step": 2086 }, { "epoch": 0.2654878514183946, "ewc_loss": 0.068359375, "ewc_loss_diag": 3.8623809814453125e-05, "ewc_loss_parallel": 6.437301635742188e-05, "grad_norm": 5.099706649780273, "learning_rate": 8.842729970326411e-06, "loss": 0.4855, "mean_token_accuracy": 0.8648535013198853, "num_tokens": 79424396.0, "step": 2087 }, { "epoch": 0.2656150616969851, "ewc_loss": 0.055908203125, "ewc_loss_diag": 3.8623809814453125e-05, "ewc_loss_parallel": 5.1975250244140625e-05, "grad_norm": 4.165435791015625, "learning_rate": 8.846969054684188e-06, "loss": 0.5948, "mean_token_accuracy": 0.8281245827674866, "num_tokens": 79465105.0, "step": 2088 }, { "epoch": 0.26574227197557565, "ewc_loss": 0.072265625, "ewc_loss_diag": 3.8623809814453125e-05, "ewc_loss_parallel": 6.818771362304688e-05, "grad_norm": 5.379206657409668, "learning_rate": 8.851208139041967e-06, "loss": 0.5006, "mean_token_accuracy": 0.8574072122573853, "num_tokens": 79504000.0, "step": 2089 }, { "epoch": 0.2658694822541661, "ewc_loss": 0.05859375, "ewc_loss_diag": 3.886222839355469e-05, "ewc_loss_parallel": 5.459785461425781e-05, "grad_norm": 4.28580904006958, "learning_rate": 8.855447223399746e-06, "loss": 0.5472, "mean_token_accuracy": 0.8390651941299438, "num_tokens": 79538954.0, "step": 2090 }, { "epoch": 0.26599669253275665, "ewc_loss": 0.0751953125, "ewc_loss_diag": 3.886222839355469e-05, "ewc_loss_parallel": 7.152557373046875e-05, "grad_norm": 5.563364028930664, "learning_rate": 8.859686307757525e-06, "loss": 0.5835, "mean_token_accuracy": 0.8343049883842468, "num_tokens": 79577264.0, "step": 2091 }, { "epoch": 0.2661239028113472, "ewc_loss": 0.06396484375, "ewc_loss_diag": 3.886222839355469e-05, "ewc_loss_parallel": 6.031990051269531e-05, "grad_norm": 4.558898448944092, "learning_rate": 8.863925392115304e-06, "loss": 0.5178, "mean_token_accuracy": 0.849061131477356, "num_tokens": 79612803.0, "step": 2092 }, { "epoch": 0.26625111308993765, "ewc_loss": 0.07421875, "ewc_loss_diag": 3.910064697265625e-05, "ewc_loss_parallel": 7.009506225585938e-05, "grad_norm": 5.466028213500977, "learning_rate": 8.868164476473083e-06, "loss": 0.4813, "mean_token_accuracy": 0.8647092580795288, "num_tokens": 79651595.0, "step": 2093 }, { "epoch": 0.2663783233685282, "ewc_loss": 0.0673828125, "ewc_loss_diag": 3.910064697265625e-05, "ewc_loss_parallel": 6.341934204101562e-05, "grad_norm": 4.778445243835449, "learning_rate": 8.872403560830862e-06, "loss": 0.5135, "mean_token_accuracy": 0.8549144268035889, "num_tokens": 79691587.0, "step": 2094 }, { "epoch": 0.2665055336471187, "ewc_loss": 0.0703125, "ewc_loss_diag": 3.933906555175781e-05, "ewc_loss_parallel": 6.628036499023438e-05, "grad_norm": 5.185121536254883, "learning_rate": 8.87664264518864e-06, "loss": 0.5687, "mean_token_accuracy": 0.8367677927017212, "num_tokens": 79729803.0, "step": 2095 }, { "epoch": 0.2666327439257092, "ewc_loss": 0.06640625, "ewc_loss_diag": 3.933906555175781e-05, "ewc_loss_parallel": 6.246566772460938e-05, "grad_norm": 4.818030834197998, "learning_rate": 8.88088172954642e-06, "loss": 0.4823, "mean_token_accuracy": 0.8610736131668091, "num_tokens": 79765866.0, "step": 2096 }, { "epoch": 0.2667599542042997, "ewc_loss": 0.06640625, "ewc_loss_diag": 3.933906555175781e-05, "ewc_loss_parallel": 6.246566772460938e-05, "grad_norm": 4.873310089111328, "learning_rate": 8.885120813904198e-06, "loss": 0.4927, "mean_token_accuracy": 0.8572163581848145, "num_tokens": 79808973.0, "step": 2097 }, { "epoch": 0.26688716448289024, "ewc_loss": 0.06494140625, "ewc_loss_diag": 3.933906555175781e-05, "ewc_loss_parallel": 6.103515625e-05, "grad_norm": 4.816391944885254, "learning_rate": 8.889359898261977e-06, "loss": 0.5916, "mean_token_accuracy": 0.8303440809249878, "num_tokens": 79843677.0, "step": 2098 }, { "epoch": 0.2670143747614807, "ewc_loss": 0.06103515625, "ewc_loss_diag": 3.933906555175781e-05, "ewc_loss_parallel": 5.7220458984375e-05, "grad_norm": 4.608211517333984, "learning_rate": 8.893598982619754e-06, "loss": 0.5019, "mean_token_accuracy": 0.8567804098129272, "num_tokens": 79882680.0, "step": 2099 }, { "epoch": 0.26714158504007124, "ewc_loss": 0.06396484375, "ewc_loss_diag": 3.933906555175781e-05, "ewc_loss_parallel": 6.008148193359375e-05, "grad_norm": 4.838837623596191, "learning_rate": 8.897838066977533e-06, "loss": 0.5187, "mean_token_accuracy": 0.8499875068664551, "num_tokens": 79929691.0, "step": 2100 }, { "epoch": 0.26726879531866177, "ewc_loss": 0.0576171875, "ewc_loss_diag": 3.933906555175781e-05, "ewc_loss_parallel": 5.364418029785156e-05, "grad_norm": 4.42047119140625, "learning_rate": 8.902077151335312e-06, "loss": 0.5117, "mean_token_accuracy": 0.849820077419281, "num_tokens": 79970515.0, "step": 2101 }, { "epoch": 0.26739600559725224, "ewc_loss": 0.0654296875, "ewc_loss_diag": 3.933906555175781e-05, "ewc_loss_parallel": 6.151199340820312e-05, "grad_norm": 4.932118892669678, "learning_rate": 8.906316235693091e-06, "loss": 0.6123, "mean_token_accuracy": 0.8171349763870239, "num_tokens": 80007550.0, "step": 2102 }, { "epoch": 0.26752321587584277, "ewc_loss": 0.056640625, "ewc_loss_diag": 3.933906555175781e-05, "ewc_loss_parallel": 5.269050598144531e-05, "grad_norm": 4.301597595214844, "learning_rate": 8.91055532005087e-06, "loss": 0.5172, "mean_token_accuracy": 0.8467413187026978, "num_tokens": 80044106.0, "step": 2103 }, { "epoch": 0.2676504261544333, "ewc_loss": 0.068359375, "ewc_loss_diag": 3.933906555175781e-05, "ewc_loss_parallel": 6.437301635742188e-05, "grad_norm": 5.0682573318481445, "learning_rate": 8.914794404408649e-06, "loss": 0.5293, "mean_token_accuracy": 0.8487573862075806, "num_tokens": 80084429.0, "step": 2104 }, { "epoch": 0.26777763643302377, "ewc_loss": 0.05615234375, "ewc_loss_diag": 3.9577484130859375e-05, "ewc_loss_parallel": 5.221366882324219e-05, "grad_norm": 4.222436904907227, "learning_rate": 8.919033488766428e-06, "loss": 0.4557, "mean_token_accuracy": 0.8660076856613159, "num_tokens": 80123317.0, "step": 2105 }, { "epoch": 0.2679048467116143, "ewc_loss": 0.07177734375, "ewc_loss_diag": 3.981590270996094e-05, "ewc_loss_parallel": 6.771087646484375e-05, "grad_norm": 5.293704986572266, "learning_rate": 8.923272573124205e-06, "loss": 0.5598, "mean_token_accuracy": 0.843011200428009, "num_tokens": 80157658.0, "step": 2106 }, { "epoch": 0.2680320569902048, "ewc_loss": 0.058349609375, "ewc_loss_diag": 3.9577484130859375e-05, "ewc_loss_parallel": 5.435943603515625e-05, "grad_norm": 4.254581451416016, "learning_rate": 8.927511657481984e-06, "loss": 0.5555, "mean_token_accuracy": 0.8366273641586304, "num_tokens": 80200482.0, "step": 2107 }, { "epoch": 0.2681592672687953, "ewc_loss": 0.0751953125, "ewc_loss_diag": 3.9577484130859375e-05, "ewc_loss_parallel": 7.152557373046875e-05, "grad_norm": 5.518293857574463, "learning_rate": 8.931750741839763e-06, "loss": 0.4993, "mean_token_accuracy": 0.8580049872398376, "num_tokens": 80231844.0, "step": 2108 }, { "epoch": 0.2682864775473858, "ewc_loss": 0.061767578125, "ewc_loss_diag": 3.981590270996094e-05, "ewc_loss_parallel": 5.793571472167969e-05, "grad_norm": 4.438811302185059, "learning_rate": 8.935989826197541e-06, "loss": 0.5058, "mean_token_accuracy": 0.8533761501312256, "num_tokens": 80269734.0, "step": 2109 }, { "epoch": 0.26841368782597635, "ewc_loss": 0.0751953125, "ewc_loss_diag": 3.9577484130859375e-05, "ewc_loss_parallel": 7.152557373046875e-05, "grad_norm": 5.592130184173584, "learning_rate": 8.94022891055532e-06, "loss": 0.4897, "mean_token_accuracy": 0.8656346797943115, "num_tokens": 80306309.0, "step": 2110 }, { "epoch": 0.2685408981045668, "ewc_loss": 0.06640625, "ewc_loss_diag": 3.9577484130859375e-05, "ewc_loss_parallel": 6.246566772460938e-05, "grad_norm": 4.682204246520996, "learning_rate": 8.9444679949131e-06, "loss": 0.5361, "mean_token_accuracy": 0.842586874961853, "num_tokens": 80345390.0, "step": 2111 }, { "epoch": 0.26866810838315736, "ewc_loss": 0.0732421875, "ewc_loss_diag": 3.9577484130859375e-05, "ewc_loss_parallel": 6.914138793945312e-05, "grad_norm": 5.425198078155518, "learning_rate": 8.948707079270878e-06, "loss": 0.461, "mean_token_accuracy": 0.870907723903656, "num_tokens": 80384118.0, "step": 2112 }, { "epoch": 0.2687953186617479, "ewc_loss": 0.0673828125, "ewc_loss_diag": 3.9577484130859375e-05, "ewc_loss_parallel": 6.341934204101562e-05, "grad_norm": 4.803987979888916, "learning_rate": 8.952946163628657e-06, "loss": 0.5129, "mean_token_accuracy": 0.8527141213417053, "num_tokens": 80416703.0, "step": 2113 }, { "epoch": 0.26892252894033836, "ewc_loss": 0.07080078125, "ewc_loss_diag": 3.981590270996094e-05, "ewc_loss_parallel": 6.67572021484375e-05, "grad_norm": 5.174912929534912, "learning_rate": 8.957185247986436e-06, "loss": 0.4668, "mean_token_accuracy": 0.8688621520996094, "num_tokens": 80453732.0, "step": 2114 }, { "epoch": 0.2690497392189289, "ewc_loss": 0.06591796875, "ewc_loss_diag": 3.981590270996094e-05, "ewc_loss_parallel": 6.198883056640625e-05, "grad_norm": 4.777040004730225, "learning_rate": 8.961424332344215e-06, "loss": 0.5165, "mean_token_accuracy": 0.8522570729255676, "num_tokens": 80495265.0, "step": 2115 }, { "epoch": 0.2691769494975194, "ewc_loss": 0.0673828125, "ewc_loss_diag": 3.981590270996094e-05, "ewc_loss_parallel": 6.341934204101562e-05, "grad_norm": 4.931638717651367, "learning_rate": 8.965663416701994e-06, "loss": 0.51, "mean_token_accuracy": 0.8515642285346985, "num_tokens": 80533789.0, "step": 2116 }, { "epoch": 0.2693041597761099, "ewc_loss": 0.06494140625, "ewc_loss_diag": 3.981590270996094e-05, "ewc_loss_parallel": 6.079673767089844e-05, "grad_norm": 4.737061500549316, "learning_rate": 8.969902501059773e-06, "loss": 0.4837, "mean_token_accuracy": 0.8627234697341919, "num_tokens": 80570802.0, "step": 2117 }, { "epoch": 0.2694313700547004, "ewc_loss": 0.064453125, "ewc_loss_diag": 3.981590270996094e-05, "ewc_loss_parallel": 6.031990051269531e-05, "grad_norm": 4.764901161193848, "learning_rate": 8.974141585417551e-06, "loss": 0.4536, "mean_token_accuracy": 0.8689539432525635, "num_tokens": 80608031.0, "step": 2118 }, { "epoch": 0.26955858033329094, "ewc_loss": 0.06298828125, "ewc_loss_diag": 3.981590270996094e-05, "ewc_loss_parallel": 5.91278076171875e-05, "grad_norm": 4.684147357940674, "learning_rate": 8.97838066977533e-06, "loss": 0.5407, "mean_token_accuracy": 0.8437687158584595, "num_tokens": 80646271.0, "step": 2119 }, { "epoch": 0.2696857906118814, "ewc_loss": 0.0615234375, "ewc_loss_diag": 3.981590270996094e-05, "ewc_loss_parallel": 5.7697296142578125e-05, "grad_norm": 4.674500942230225, "learning_rate": 8.98261975413311e-06, "loss": 0.5359, "mean_token_accuracy": 0.8438573479652405, "num_tokens": 80683325.0, "step": 2120 }, { "epoch": 0.26981300089047194, "ewc_loss": 0.061279296875, "ewc_loss_diag": 3.981590270996094e-05, "ewc_loss_parallel": 5.745887756347656e-05, "grad_norm": 4.627607345581055, "learning_rate": 8.986858838490888e-06, "loss": 0.5433, "mean_token_accuracy": 0.8437572717666626, "num_tokens": 80726403.0, "step": 2121 }, { "epoch": 0.26994021116906247, "ewc_loss": 0.060302734375, "ewc_loss_diag": 3.981590270996094e-05, "ewc_loss_parallel": 5.650520324707031e-05, "grad_norm": 4.61240816116333, "learning_rate": 8.991097922848665e-06, "loss": 0.5034, "mean_token_accuracy": 0.8545212745666504, "num_tokens": 80767293.0, "step": 2122 }, { "epoch": 0.27006742144765294, "ewc_loss": 0.059814453125, "ewc_loss_diag": 3.981590270996094e-05, "ewc_loss_parallel": 5.602836608886719e-05, "grad_norm": 4.59257698059082, "learning_rate": 8.995337007206444e-06, "loss": 0.5312, "mean_token_accuracy": 0.843997597694397, "num_tokens": 80806218.0, "step": 2123 }, { "epoch": 0.2701946317262435, "ewc_loss": 0.059814453125, "ewc_loss_diag": 3.981590270996094e-05, "ewc_loss_parallel": 5.602836608886719e-05, "grad_norm": 4.5836567878723145, "learning_rate": 8.999576091564223e-06, "loss": 0.4992, "mean_token_accuracy": 0.8573459386825562, "num_tokens": 80841694.0, "step": 2124 }, { "epoch": 0.270321842004834, "ewc_loss": 0.0595703125, "ewc_loss_diag": 3.981590270996094e-05, "ewc_loss_parallel": 5.555152893066406e-05, "grad_norm": 4.5713605880737305, "learning_rate": 9.003815175922002e-06, "loss": 0.5041, "mean_token_accuracy": 0.8527361154556274, "num_tokens": 80874879.0, "step": 2125 }, { "epoch": 0.2704490522834245, "ewc_loss": 0.0595703125, "ewc_loss_diag": 4.029273986816406e-05, "ewc_loss_parallel": 5.53131103515625e-05, "grad_norm": 4.562943935394287, "learning_rate": 9.00805426027978e-06, "loss": 0.5194, "mean_token_accuracy": 0.8497005701065063, "num_tokens": 80916317.0, "step": 2126 }, { "epoch": 0.270576262562015, "ewc_loss": 0.0595703125, "ewc_loss_diag": 4.029273986816406e-05, "ewc_loss_parallel": 5.53131103515625e-05, "grad_norm": 4.5543694496154785, "learning_rate": 9.01229334463756e-06, "loss": 0.5438, "mean_token_accuracy": 0.8428685665130615, "num_tokens": 80952304.0, "step": 2127 }, { "epoch": 0.27070347284060553, "ewc_loss": 0.05908203125, "ewc_loss_diag": 4.029273986816406e-05, "ewc_loss_parallel": 5.507469177246094e-05, "grad_norm": 4.550358772277832, "learning_rate": 9.016532428995339e-06, "loss": 0.4836, "mean_token_accuracy": 0.8582898378372192, "num_tokens": 80989886.0, "step": 2128 }, { "epoch": 0.27083068311919606, "ewc_loss": 0.05908203125, "ewc_loss_diag": 4.029273986816406e-05, "ewc_loss_parallel": 5.4836273193359375e-05, "grad_norm": 4.537550926208496, "learning_rate": 9.020771513353116e-06, "loss": 0.5374, "mean_token_accuracy": 0.8423686027526855, "num_tokens": 81031041.0, "step": 2129 }, { "epoch": 0.27095789339778653, "ewc_loss": 0.0595703125, "ewc_loss_diag": 4.029273986816406e-05, "ewc_loss_parallel": 5.53131103515625e-05, "grad_norm": 4.554991245269775, "learning_rate": 9.025010597710895e-06, "loss": 0.5302, "mean_token_accuracy": 0.8455645442008972, "num_tokens": 81068915.0, "step": 2130 }, { "epoch": 0.27108510367637706, "ewc_loss": 0.05908203125, "ewc_loss_diag": 4.029273986816406e-05, "ewc_loss_parallel": 5.507469177246094e-05, "grad_norm": 4.527692794799805, "learning_rate": 9.029249682068674e-06, "loss": 0.4951, "mean_token_accuracy": 0.8577992916107178, "num_tokens": 81106044.0, "step": 2131 }, { "epoch": 0.2712123139549676, "ewc_loss": 0.0595703125, "ewc_loss_diag": 4.029273986816406e-05, "ewc_loss_parallel": 5.555152893066406e-05, "grad_norm": 4.5535569190979, "learning_rate": 9.033488766426452e-06, "loss": 0.5055, "mean_token_accuracy": 0.8516170382499695, "num_tokens": 81142589.0, "step": 2132 }, { "epoch": 0.27133952423355806, "ewc_loss": 0.0595703125, "ewc_loss_diag": 4.029273986816406e-05, "ewc_loss_parallel": 5.53131103515625e-05, "grad_norm": 4.522963523864746, "learning_rate": 9.037727850784231e-06, "loss": 0.483, "mean_token_accuracy": 0.8594261407852173, "num_tokens": 81178415.0, "step": 2133 }, { "epoch": 0.2714667345121486, "ewc_loss": 0.06005859375, "ewc_loss_diag": 4.029273986816406e-05, "ewc_loss_parallel": 5.602836608886719e-05, "grad_norm": 4.571036338806152, "learning_rate": 9.04196693514201e-06, "loss": 0.525, "mean_token_accuracy": 0.8483989238739014, "num_tokens": 81216952.0, "step": 2134 }, { "epoch": 0.2715939447907391, "ewc_loss": 0.0595703125, "ewc_loss_diag": 4.029273986816406e-05, "ewc_loss_parallel": 5.555152893066406e-05, "grad_norm": 4.49421501159668, "learning_rate": 9.046206019499789e-06, "loss": 0.4732, "mean_token_accuracy": 0.863673746585846, "num_tokens": 81257879.0, "step": 2135 }, { "epoch": 0.2717211550693296, "ewc_loss": 0.06005859375, "ewc_loss_diag": 4.029273986816406e-05, "ewc_loss_parallel": 5.602836608886719e-05, "grad_norm": 4.6136980056762695, "learning_rate": 9.050445103857568e-06, "loss": 0.5502, "mean_token_accuracy": 0.843512773513794, "num_tokens": 81298045.0, "step": 2136 }, { "epoch": 0.2718483653479201, "ewc_loss": 0.05859375, "ewc_loss_diag": 4.029273986816406e-05, "ewc_loss_parallel": 5.459785461425781e-05, "grad_norm": 4.45380163192749, "learning_rate": 9.054684188215345e-06, "loss": 0.4619, "mean_token_accuracy": 0.8682719469070435, "num_tokens": 81332342.0, "step": 2137 }, { "epoch": 0.27197557562651065, "ewc_loss": 0.060546875, "ewc_loss_diag": 4.029273986816406e-05, "ewc_loss_parallel": 5.6743621826171875e-05, "grad_norm": 4.676103591918945, "learning_rate": 9.058923272573124e-06, "loss": 0.5391, "mean_token_accuracy": 0.8417405486106873, "num_tokens": 81367270.0, "step": 2138 }, { "epoch": 0.2721027859051011, "ewc_loss": 0.0576171875, "ewc_loss_diag": 4.029273986816406e-05, "ewc_loss_parallel": 5.340576171875e-05, "grad_norm": 4.375472545623779, "learning_rate": 9.063162356930903e-06, "loss": 0.4868, "mean_token_accuracy": 0.8592863082885742, "num_tokens": 81406644.0, "step": 2139 }, { "epoch": 0.27222999618369165, "ewc_loss": 0.0634765625, "ewc_loss_diag": 4.029273986816406e-05, "ewc_loss_parallel": 5.9604644775390625e-05, "grad_norm": 4.8013458251953125, "learning_rate": 9.067401441288682e-06, "loss": 0.5266, "mean_token_accuracy": 0.8506765961647034, "num_tokens": 81443225.0, "step": 2140 }, { "epoch": 0.2723572064622822, "ewc_loss": 0.056640625, "ewc_loss_diag": 4.076957702636719e-05, "ewc_loss_parallel": 5.245208740234375e-05, "grad_norm": 4.247346878051758, "learning_rate": 9.07164052564646e-06, "loss": 0.4873, "mean_token_accuracy": 0.8538475036621094, "num_tokens": 81482168.0, "step": 2141 }, { "epoch": 0.27248441674087265, "ewc_loss": 0.06884765625, "ewc_loss_diag": 4.076957702636719e-05, "ewc_loss_parallel": 6.4849853515625e-05, "grad_norm": 5.066225528717041, "learning_rate": 9.07587961000424e-06, "loss": 0.5415, "mean_token_accuracy": 0.8442907929420471, "num_tokens": 81518127.0, "step": 2142 }, { "epoch": 0.2726116270194632, "ewc_loss": 0.055419921875, "ewc_loss_diag": 4.124641418457031e-05, "ewc_loss_parallel": 5.125999450683594e-05, "grad_norm": 4.070852756500244, "learning_rate": 9.080118694362018e-06, "loss": 0.5294, "mean_token_accuracy": 0.8411386013031006, "num_tokens": 81559857.0, "step": 2143 }, { "epoch": 0.2727388372980537, "ewc_loss": 0.07666015625, "ewc_loss_diag": 4.124641418457031e-05, "ewc_loss_parallel": 7.295608520507812e-05, "grad_norm": 5.550530910491943, "learning_rate": 9.084357778719797e-06, "loss": 0.5433, "mean_token_accuracy": 0.8442994356155396, "num_tokens": 81596509.0, "step": 2144 }, { "epoch": 0.2728660475766442, "ewc_loss": 0.05810546875, "ewc_loss_diag": 4.124641418457031e-05, "ewc_loss_parallel": 5.3882598876953125e-05, "grad_norm": 4.024962425231934, "learning_rate": 9.088596863077576e-06, "loss": 0.5033, "mean_token_accuracy": 0.8564902544021606, "num_tokens": 81632693.0, "step": 2145 }, { "epoch": 0.2729932578552347, "ewc_loss": 0.0888671875, "ewc_loss_diag": 4.124641418457031e-05, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 6.266396522521973, "learning_rate": 9.092835947435355e-06, "loss": 0.548, "mean_token_accuracy": 0.8480896949768066, "num_tokens": 81670067.0, "step": 2146 }, { "epoch": 0.27312046813382523, "ewc_loss": 0.0712890625, "ewc_loss_diag": 4.124641418457031e-05, "ewc_loss_parallel": 6.723403930664062e-05, "grad_norm": 4.701478958129883, "learning_rate": 9.097075031793134e-06, "loss": 0.5699, "mean_token_accuracy": 0.8361805081367493, "num_tokens": 81706902.0, "step": 2147 }, { "epoch": 0.2732476784124157, "ewc_loss": 0.0908203125, "ewc_loss_diag": 4.124641418457031e-05, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 6.442305564880371, "learning_rate": 9.101314116150913e-06, "loss": 0.5022, "mean_token_accuracy": 0.8652700185775757, "num_tokens": 81743340.0, "step": 2148 }, { "epoch": 0.27337488869100623, "ewc_loss": 0.083984375, "ewc_loss_diag": 4.124641418457031e-05, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.540408611297607, "learning_rate": 9.105553200508692e-06, "loss": 0.5106, "mean_token_accuracy": 0.8599435687065125, "num_tokens": 81779573.0, "step": 2149 }, { "epoch": 0.27350209896959676, "ewc_loss": 0.08154296875, "ewc_loss_diag": 4.124641418457031e-05, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.813288688659668, "learning_rate": 9.10979228486647e-06, "loss": 0.5351, "mean_token_accuracy": 0.8512363433837891, "num_tokens": 81822004.0, "step": 2150 }, { "epoch": 0.27362930924818724, "ewc_loss": 0.0771484375, "ewc_loss_diag": 4.124641418457031e-05, "ewc_loss_parallel": 7.343292236328125e-05, "grad_norm": 5.442481994628906, "learning_rate": 9.11403136922425e-06, "loss": 0.5344, "mean_token_accuracy": 0.8490642309188843, "num_tokens": 81860472.0, "step": 2151 }, { "epoch": 0.27375651952677776, "ewc_loss": 0.07470703125, "ewc_loss_diag": 4.124641418457031e-05, "ewc_loss_parallel": 7.05718994140625e-05, "grad_norm": 5.464084625244141, "learning_rate": 9.118270453582027e-06, "loss": 0.5916, "mean_token_accuracy": 0.8364903926849365, "num_tokens": 81899151.0, "step": 2152 }, { "epoch": 0.2738837298053683, "ewc_loss": 0.0712890625, "ewc_loss_diag": 4.1484832763671875e-05, "ewc_loss_parallel": 6.723403930664062e-05, "grad_norm": 5.133580207824707, "learning_rate": 9.122509537939806e-06, "loss": 0.5131, "mean_token_accuracy": 0.8507332801818848, "num_tokens": 81941213.0, "step": 2153 }, { "epoch": 0.27401094008395877, "ewc_loss": 0.072265625, "ewc_loss_diag": 4.1484832763671875e-05, "ewc_loss_parallel": 6.771087646484375e-05, "grad_norm": 5.259894371032715, "learning_rate": 9.126748622297584e-06, "loss": 0.5593, "mean_token_accuracy": 0.8352276086807251, "num_tokens": 81978055.0, "step": 2154 }, { "epoch": 0.2741381503625493, "ewc_loss": 0.0673828125, "ewc_loss_diag": 4.1484832763671875e-05, "ewc_loss_parallel": 6.29425048828125e-05, "grad_norm": 4.9065961837768555, "learning_rate": 9.130987706655363e-06, "loss": 0.5451, "mean_token_accuracy": 0.8438065052032471, "num_tokens": 82015599.0, "step": 2155 }, { "epoch": 0.2742653606411398, "ewc_loss": 0.0703125, "ewc_loss_diag": 4.1484832763671875e-05, "ewc_loss_parallel": 6.580352783203125e-05, "grad_norm": 5.158125400543213, "learning_rate": 9.135226791013142e-06, "loss": 0.5739, "mean_token_accuracy": 0.8422818183898926, "num_tokens": 82055113.0, "step": 2156 }, { "epoch": 0.2743925709197303, "ewc_loss": 0.0634765625, "ewc_loss_diag": 4.1484832763671875e-05, "ewc_loss_parallel": 5.9604644775390625e-05, "grad_norm": 4.718398571014404, "learning_rate": 9.139465875370921e-06, "loss": 0.511, "mean_token_accuracy": 0.8494886159896851, "num_tokens": 82083052.0, "step": 2157 }, { "epoch": 0.2745197811983208, "ewc_loss": 0.0703125, "ewc_loss_diag": 4.1484832763671875e-05, "ewc_loss_parallel": 6.580352783203125e-05, "grad_norm": 5.1729817390441895, "learning_rate": 9.1437049597287e-06, "loss": 0.5358, "mean_token_accuracy": 0.8440085649490356, "num_tokens": 82118850.0, "step": 2158 }, { "epoch": 0.27464699147691135, "ewc_loss": 0.060302734375, "ewc_loss_diag": 4.1484832763671875e-05, "ewc_loss_parallel": 5.626678466796875e-05, "grad_norm": 4.547746181488037, "learning_rate": 9.147944044086477e-06, "loss": 0.5478, "mean_token_accuracy": 0.8395088911056519, "num_tokens": 82151261.0, "step": 2159 }, { "epoch": 0.2747742017555018, "ewc_loss": 0.0712890625, "ewc_loss_diag": 4.1484832763671875e-05, "ewc_loss_parallel": 6.723403930664062e-05, "grad_norm": 5.297628402709961, "learning_rate": 9.152183128444256e-06, "loss": 0.5617, "mean_token_accuracy": 0.8462576866149902, "num_tokens": 82192018.0, "step": 2160 }, { "epoch": 0.27490141203409235, "ewc_loss": 0.0595703125, "ewc_loss_diag": 4.1484832763671875e-05, "ewc_loss_parallel": 5.53131103515625e-05, "grad_norm": 4.424200534820557, "learning_rate": 9.156422212802035e-06, "loss": 0.5576, "mean_token_accuracy": 0.8328631520271301, "num_tokens": 82231044.0, "step": 2161 }, { "epoch": 0.2750286223126829, "ewc_loss": 0.07421875, "ewc_loss_diag": 4.1484832763671875e-05, "ewc_loss_parallel": 7.009506225585938e-05, "grad_norm": 5.567756175994873, "learning_rate": 9.160661297159814e-06, "loss": 0.5021, "mean_token_accuracy": 0.8579566478729248, "num_tokens": 82265334.0, "step": 2162 }, { "epoch": 0.27515583259127335, "ewc_loss": 0.06103515625, "ewc_loss_diag": 4.1484832763671875e-05, "ewc_loss_parallel": 5.698204040527344e-05, "grad_norm": 4.471010684967041, "learning_rate": 9.164900381517593e-06, "loss": 0.5058, "mean_token_accuracy": 0.8506090641021729, "num_tokens": 82303995.0, "step": 2163 }, { "epoch": 0.2752830428698639, "ewc_loss": 0.0791015625, "ewc_loss_diag": 4.1484832763671875e-05, "ewc_loss_parallel": 7.534027099609375e-05, "grad_norm": 5.824389457702637, "learning_rate": 9.169139465875372e-06, "loss": 0.5301, "mean_token_accuracy": 0.8505971431732178, "num_tokens": 82345157.0, "step": 2164 }, { "epoch": 0.2754102531484544, "ewc_loss": 0.0673828125, "ewc_loss_diag": 4.1484832763671875e-05, "ewc_loss_parallel": 6.341934204101562e-05, "grad_norm": 4.797837734222412, "learning_rate": 9.17337855023315e-06, "loss": 0.526, "mean_token_accuracy": 0.8469350934028625, "num_tokens": 82380723.0, "step": 2165 }, { "epoch": 0.2755374634270449, "ewc_loss": 0.0771484375, "ewc_loss_diag": 4.1484832763671875e-05, "ewc_loss_parallel": 7.295608520507812e-05, "grad_norm": 5.743776321411133, "learning_rate": 9.17761763459093e-06, "loss": 0.4939, "mean_token_accuracy": 0.8605544567108154, "num_tokens": 82418985.0, "step": 2166 }, { "epoch": 0.2756646737056354, "ewc_loss": 0.0703125, "ewc_loss_diag": 4.1484832763671875e-05, "ewc_loss_parallel": 6.628036499023438e-05, "grad_norm": 5.075071811676025, "learning_rate": 9.181856718948707e-06, "loss": 0.4674, "mean_token_accuracy": 0.8662823438644409, "num_tokens": 82455498.0, "step": 2167 }, { "epoch": 0.27579188398422594, "ewc_loss": 0.072265625, "ewc_loss_diag": 4.1484832763671875e-05, "ewc_loss_parallel": 6.818771362304688e-05, "grad_norm": 5.369653701782227, "learning_rate": 9.186095803306485e-06, "loss": 0.5122, "mean_token_accuracy": 0.8535482883453369, "num_tokens": 82491117.0, "step": 2168 }, { "epoch": 0.2759190942628164, "ewc_loss": 0.0693359375, "ewc_loss_diag": 4.1484832763671875e-05, "ewc_loss_parallel": 6.4849853515625e-05, "grad_norm": 5.1124982833862305, "learning_rate": 9.190334887664264e-06, "loss": 0.5576, "mean_token_accuracy": 0.8411474227905273, "num_tokens": 82533640.0, "step": 2169 }, { "epoch": 0.27604630454140694, "ewc_loss": 0.06689453125, "ewc_loss_diag": 4.172325134277344e-05, "ewc_loss_parallel": 6.246566772460938e-05, "grad_norm": 4.975661277770996, "learning_rate": 9.194573972022043e-06, "loss": 0.6098, "mean_token_accuracy": 0.830132007598877, "num_tokens": 82568812.0, "step": 2170 }, { "epoch": 0.27617351481999747, "ewc_loss": 0.068359375, "ewc_loss_diag": 4.172325134277344e-05, "ewc_loss_parallel": 6.389617919921875e-05, "grad_norm": 5.0977678298950195, "learning_rate": 9.198813056379822e-06, "loss": 0.5993, "mean_token_accuracy": 0.8283476829528809, "num_tokens": 82611777.0, "step": 2171 }, { "epoch": 0.27630072509858794, "ewc_loss": 0.061767578125, "ewc_loss_diag": 4.172325134277344e-05, "ewc_loss_parallel": 5.7697296142578125e-05, "grad_norm": 4.637478828430176, "learning_rate": 9.203052140737601e-06, "loss": 0.5382, "mean_token_accuracy": 0.8441066145896912, "num_tokens": 82646938.0, "step": 2172 }, { "epoch": 0.27642793537717847, "ewc_loss": 0.0703125, "ewc_loss_diag": 4.172325134277344e-05, "ewc_loss_parallel": 6.580352783203125e-05, "grad_norm": 5.163607120513916, "learning_rate": 9.20729122509538e-06, "loss": 0.4867, "mean_token_accuracy": 0.8619369268417358, "num_tokens": 82684316.0, "step": 2173 }, { "epoch": 0.276555145655769, "ewc_loss": 0.05859375, "ewc_loss_diag": 4.172325134277344e-05, "ewc_loss_parallel": 5.435943603515625e-05, "grad_norm": 4.408999443054199, "learning_rate": 9.211530309453159e-06, "loss": 0.5183, "mean_token_accuracy": 0.8463859558105469, "num_tokens": 82723204.0, "step": 2174 }, { "epoch": 0.27668235593435947, "ewc_loss": 0.07275390625, "ewc_loss_diag": 4.172325134277344e-05, "ewc_loss_parallel": 6.818771362304688e-05, "grad_norm": 5.364508628845215, "learning_rate": 9.215769393810938e-06, "loss": 0.5281, "mean_token_accuracy": 0.8504075407981873, "num_tokens": 82759634.0, "step": 2175 }, { "epoch": 0.27680956621295, "ewc_loss": 0.057861328125, "ewc_loss_diag": 4.172325134277344e-05, "ewc_loss_parallel": 5.364418029785156e-05, "grad_norm": 4.321627140045166, "learning_rate": 9.220008478168717e-06, "loss": 0.6023, "mean_token_accuracy": 0.8234748244285583, "num_tokens": 82800446.0, "step": 2176 }, { "epoch": 0.2769367764915405, "ewc_loss": 0.07666015625, "ewc_loss_diag": 4.172325134277344e-05, "ewc_loss_parallel": 7.2479248046875e-05, "grad_norm": 5.6654953956604, "learning_rate": 9.224247562526495e-06, "loss": 0.5367, "mean_token_accuracy": 0.8471722602844238, "num_tokens": 82840754.0, "step": 2177 }, { "epoch": 0.277063986770131, "ewc_loss": 0.062255859375, "ewc_loss_diag": 4.172325134277344e-05, "ewc_loss_parallel": 5.817413330078125e-05, "grad_norm": 4.472116947174072, "learning_rate": 9.228486646884274e-06, "loss": 0.5212, "mean_token_accuracy": 0.8492565155029297, "num_tokens": 82881859.0, "step": 2178 }, { "epoch": 0.2771911970487215, "ewc_loss": 0.08056640625, "ewc_loss_diag": 4.172325134277344e-05, "ewc_loss_parallel": 7.62939453125e-05, "grad_norm": 5.8652567863464355, "learning_rate": 9.232725731242053e-06, "loss": 0.5828, "mean_token_accuracy": 0.8340145945549011, "num_tokens": 82922146.0, "step": 2179 }, { "epoch": 0.27731840732731206, "ewc_loss": 0.0693359375, "ewc_loss_diag": 4.1961669921875e-05, "ewc_loss_parallel": 6.4849853515625e-05, "grad_norm": 4.841535568237305, "learning_rate": 9.236964815599832e-06, "loss": 0.5618, "mean_token_accuracy": 0.8382039666175842, "num_tokens": 82960017.0, "step": 2180 }, { "epoch": 0.2774456176059026, "ewc_loss": 0.0771484375, "ewc_loss_diag": 4.1961669921875e-05, "ewc_loss_parallel": 7.295608520507812e-05, "grad_norm": 5.719231605529785, "learning_rate": 9.241203899957611e-06, "loss": 0.5081, "mean_token_accuracy": 0.8568748235702515, "num_tokens": 83001778.0, "step": 2181 }, { "epoch": 0.27757282788449306, "ewc_loss": 0.07177734375, "ewc_loss_diag": 4.220008850097656e-05, "ewc_loss_parallel": 6.723403930664062e-05, "grad_norm": 5.035375595092773, "learning_rate": 9.245442984315388e-06, "loss": 0.5146, "mean_token_accuracy": 0.8533990979194641, "num_tokens": 83037023.0, "step": 2182 }, { "epoch": 0.2777000381630836, "ewc_loss": 0.0732421875, "ewc_loss_diag": 4.1961669921875e-05, "ewc_loss_parallel": 6.866455078125e-05, "grad_norm": 5.419190406799316, "learning_rate": 9.249682068673167e-06, "loss": 0.5557, "mean_token_accuracy": 0.8412555456161499, "num_tokens": 83075355.0, "step": 2183 }, { "epoch": 0.2778272484416741, "ewc_loss": 0.06982421875, "ewc_loss_diag": 4.220008850097656e-05, "ewc_loss_parallel": 6.532669067382812e-05, "grad_norm": 5.010940074920654, "learning_rate": 9.253921153030946e-06, "loss": 0.5884, "mean_token_accuracy": 0.8321232199668884, "num_tokens": 83116233.0, "step": 2184 }, { "epoch": 0.2779544587202646, "ewc_loss": 0.0712890625, "ewc_loss_diag": 4.220008850097656e-05, "ewc_loss_parallel": 6.67572021484375e-05, "grad_norm": 5.146730422973633, "learning_rate": 9.258160237388725e-06, "loss": 0.5044, "mean_token_accuracy": 0.8603761196136475, "num_tokens": 83152238.0, "step": 2185 }, { "epoch": 0.2780816689988551, "ewc_loss": 0.06689453125, "ewc_loss_diag": 4.220008850097656e-05, "ewc_loss_parallel": 6.246566772460938e-05, "grad_norm": 4.907832622528076, "learning_rate": 9.262399321746504e-06, "loss": 0.5083, "mean_token_accuracy": 0.8524720668792725, "num_tokens": 83186779.0, "step": 2186 }, { "epoch": 0.27820887927744564, "ewc_loss": 0.06884765625, "ewc_loss_diag": 4.220008850097656e-05, "ewc_loss_parallel": 6.437301635742188e-05, "grad_norm": 4.978464126586914, "learning_rate": 9.266638406104283e-06, "loss": 0.5707, "mean_token_accuracy": 0.8349741697311401, "num_tokens": 83226001.0, "step": 2187 }, { "epoch": 0.2783360895560361, "ewc_loss": 0.06591796875, "ewc_loss_diag": 4.220008850097656e-05, "ewc_loss_parallel": 6.151199340820312e-05, "grad_norm": 4.813765525817871, "learning_rate": 9.270877490462061e-06, "loss": 0.5276, "mean_token_accuracy": 0.8478554487228394, "num_tokens": 83257196.0, "step": 2188 }, { "epoch": 0.27846329983462664, "ewc_loss": 0.06689453125, "ewc_loss_diag": 4.220008850097656e-05, "ewc_loss_parallel": 6.246566772460938e-05, "grad_norm": 4.907232284545898, "learning_rate": 9.27511657481984e-06, "loss": 0.5329, "mean_token_accuracy": 0.8477750420570374, "num_tokens": 83293261.0, "step": 2189 }, { "epoch": 0.27859051011321717, "ewc_loss": 0.06298828125, "ewc_loss_diag": 4.220008850097656e-05, "ewc_loss_parallel": 5.8650970458984375e-05, "grad_norm": 4.700463771820068, "learning_rate": 9.279355659177618e-06, "loss": 0.5428, "mean_token_accuracy": 0.8470052480697632, "num_tokens": 83327096.0, "step": 2190 }, { "epoch": 0.27871772039180764, "ewc_loss": 0.06689453125, "ewc_loss_diag": 4.220008850097656e-05, "ewc_loss_parallel": 6.246566772460938e-05, "grad_norm": 4.909554481506348, "learning_rate": 9.283594743535396e-06, "loss": 0.5646, "mean_token_accuracy": 0.8392437696456909, "num_tokens": 83360558.0, "step": 2191 }, { "epoch": 0.2788449306703982, "ewc_loss": 0.060302734375, "ewc_loss_diag": 4.220008850097656e-05, "ewc_loss_parallel": 5.626678466796875e-05, "grad_norm": 4.572556972503662, "learning_rate": 9.287833827893175e-06, "loss": 0.5208, "mean_token_accuracy": 0.8455590009689331, "num_tokens": 83400563.0, "step": 2192 }, { "epoch": 0.2789721409489887, "ewc_loss": 0.06787109375, "ewc_loss_diag": 4.220008850097656e-05, "ewc_loss_parallel": 6.341934204101562e-05, "grad_norm": 4.985593318939209, "learning_rate": 9.292072912250954e-06, "loss": 0.5399, "mean_token_accuracy": 0.8431469202041626, "num_tokens": 83432961.0, "step": 2193 }, { "epoch": 0.2790993512275792, "ewc_loss": 0.059326171875, "ewc_loss_diag": 4.220008850097656e-05, "ewc_loss_parallel": 5.507469177246094e-05, "grad_norm": 4.449453830718994, "learning_rate": 9.296311996608733e-06, "loss": 0.4652, "mean_token_accuracy": 0.8631162643432617, "num_tokens": 83472183.0, "step": 2194 }, { "epoch": 0.2792265615061697, "ewc_loss": 0.07080078125, "ewc_loss_diag": 4.220008850097656e-05, "ewc_loss_parallel": 6.628036499023438e-05, "grad_norm": 5.148288249969482, "learning_rate": 9.300551080966512e-06, "loss": 0.6, "mean_token_accuracy": 0.8268848657608032, "num_tokens": 83512376.0, "step": 2195 }, { "epoch": 0.27935377178476023, "ewc_loss": 0.05810546875, "ewc_loss_diag": 4.220008850097656e-05, "ewc_loss_parallel": 5.3882598876953125e-05, "grad_norm": 4.358313083648682, "learning_rate": 9.304790165324291e-06, "loss": 0.5089, "mean_token_accuracy": 0.8492274880409241, "num_tokens": 83546141.0, "step": 2196 }, { "epoch": 0.2794809820633507, "ewc_loss": 0.07470703125, "ewc_loss_diag": 4.220008850097656e-05, "ewc_loss_parallel": 7.009506225585938e-05, "grad_norm": 5.4028639793396, "learning_rate": 9.30902924968207e-06, "loss": 0.5268, "mean_token_accuracy": 0.8523997664451599, "num_tokens": 83584223.0, "step": 2197 }, { "epoch": 0.27960819234194123, "ewc_loss": 0.05908203125, "ewc_loss_diag": 4.220008850097656e-05, "ewc_loss_parallel": 5.4836273193359375e-05, "grad_norm": 4.322581768035889, "learning_rate": 9.313268334039847e-06, "loss": 0.5325, "mean_token_accuracy": 0.8434620499610901, "num_tokens": 83621773.0, "step": 2198 }, { "epoch": 0.27973540262053176, "ewc_loss": 0.07861328125, "ewc_loss_diag": 4.220008850097656e-05, "ewc_loss_parallel": 7.43865966796875e-05, "grad_norm": 5.730058670043945, "learning_rate": 9.317507418397626e-06, "loss": 0.4735, "mean_token_accuracy": 0.86892169713974, "num_tokens": 83659129.0, "step": 2199 }, { "epoch": 0.27986261289912223, "ewc_loss": 0.06591796875, "ewc_loss_diag": 4.220008850097656e-05, "ewc_loss_parallel": 6.151199340820312e-05, "grad_norm": 4.534885406494141, "learning_rate": 9.321746502755405e-06, "loss": 0.5697, "mean_token_accuracy": 0.8380594253540039, "num_tokens": 83699146.0, "step": 2200 }, { "epoch": 0.27998982317771276, "ewc_loss": 0.08154296875, "ewc_loss_diag": 4.220008850097656e-05, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.881999492645264, "learning_rate": 9.325985587113184e-06, "loss": 0.5475, "mean_token_accuracy": 0.8473733067512512, "num_tokens": 83733266.0, "step": 2201 }, { "epoch": 0.2801170334563033, "ewc_loss": 0.07177734375, "ewc_loss_diag": 4.220008850097656e-05, "ewc_loss_parallel": 6.723403930664062e-05, "grad_norm": 4.910865306854248, "learning_rate": 9.330224671470962e-06, "loss": 0.5422, "mean_token_accuracy": 0.8520016670227051, "num_tokens": 83766610.0, "step": 2202 }, { "epoch": 0.28024424373489376, "ewc_loss": 0.07666015625, "ewc_loss_diag": 4.220008850097656e-05, "ewc_loss_parallel": 7.2479248046875e-05, "grad_norm": 5.672167778015137, "learning_rate": 9.334463755828741e-06, "loss": 0.5652, "mean_token_accuracy": 0.8364648818969727, "num_tokens": 83797891.0, "step": 2203 }, { "epoch": 0.2803714540134843, "ewc_loss": 0.07275390625, "ewc_loss_diag": 4.220008850097656e-05, "ewc_loss_parallel": 6.818771362304688e-05, "grad_norm": 5.076590061187744, "learning_rate": 9.33870284018652e-06, "loss": 0.5607, "mean_token_accuracy": 0.8366146087646484, "num_tokens": 83837046.0, "step": 2204 }, { "epoch": 0.2804986642920748, "ewc_loss": 0.07275390625, "ewc_loss_diag": 4.220008850097656e-05, "ewc_loss_parallel": 6.818771362304688e-05, "grad_norm": 5.289333820343018, "learning_rate": 9.342941924544299e-06, "loss": 0.5535, "mean_token_accuracy": 0.8423661589622498, "num_tokens": 83876457.0, "step": 2205 }, { "epoch": 0.2806258745706653, "ewc_loss": 0.07177734375, "ewc_loss_diag": 4.220008850097656e-05, "ewc_loss_parallel": 6.723403930664062e-05, "grad_norm": 5.060664653778076, "learning_rate": 9.347181008902078e-06, "loss": 0.4902, "mean_token_accuracy": 0.8635244369506836, "num_tokens": 83912729.0, "step": 2206 }, { "epoch": 0.2807530848492558, "ewc_loss": 0.0693359375, "ewc_loss_diag": 4.220008850097656e-05, "ewc_loss_parallel": 6.4849853515625e-05, "grad_norm": 4.956699848175049, "learning_rate": 9.351420093259857e-06, "loss": 0.4891, "mean_token_accuracy": 0.8596885800361633, "num_tokens": 83945958.0, "step": 2207 }, { "epoch": 0.28088029512784635, "ewc_loss": 0.0703125, "ewc_loss_diag": 4.220008850097656e-05, "ewc_loss_parallel": 6.580352783203125e-05, "grad_norm": 5.0656352043151855, "learning_rate": 9.355659177617636e-06, "loss": 0.4781, "mean_token_accuracy": 0.8606007099151611, "num_tokens": 83976767.0, "step": 2208 }, { "epoch": 0.2810075054064368, "ewc_loss": 0.06396484375, "ewc_loss_diag": 4.220008850097656e-05, "ewc_loss_parallel": 5.9604644775390625e-05, "grad_norm": 4.698287010192871, "learning_rate": 9.359898261975415e-06, "loss": 0.4663, "mean_token_accuracy": 0.8653462529182434, "num_tokens": 84017996.0, "step": 2209 }, { "epoch": 0.28113471568502735, "ewc_loss": 0.07080078125, "ewc_loss_diag": 4.220008850097656e-05, "ewc_loss_parallel": 6.628036499023438e-05, "grad_norm": 5.11452054977417, "learning_rate": 9.364137346333194e-06, "loss": 0.6031, "mean_token_accuracy": 0.8306001424789429, "num_tokens": 84057528.0, "step": 2210 }, { "epoch": 0.2812619259636179, "ewc_loss": 0.0615234375, "ewc_loss_diag": 4.220008850097656e-05, "ewc_loss_parallel": 5.745887756347656e-05, "grad_norm": 4.539424896240234, "learning_rate": 9.368376430690972e-06, "loss": 0.4774, "mean_token_accuracy": 0.8613548278808594, "num_tokens": 84091009.0, "step": 2211 }, { "epoch": 0.28138913624220835, "ewc_loss": 0.07177734375, "ewc_loss_diag": 4.220008850097656e-05, "ewc_loss_parallel": 6.723403930664062e-05, "grad_norm": 5.243674278259277, "learning_rate": 9.372615515048751e-06, "loss": 0.5648, "mean_token_accuracy": 0.8374024629592896, "num_tokens": 84125042.0, "step": 2212 }, { "epoch": 0.2815163465207989, "ewc_loss": 0.06005859375, "ewc_loss_diag": 4.220008850097656e-05, "ewc_loss_parallel": 5.602836608886719e-05, "grad_norm": 4.435940742492676, "learning_rate": 9.376854599406528e-06, "loss": 0.5567, "mean_token_accuracy": 0.8388596773147583, "num_tokens": 84169340.0, "step": 2213 }, { "epoch": 0.2816435567993894, "ewc_loss": 0.0751953125, "ewc_loss_diag": 4.2438507080078125e-05, "ewc_loss_parallel": 7.05718994140625e-05, "grad_norm": 5.443849563598633, "learning_rate": 9.381093683764307e-06, "loss": 0.5323, "mean_token_accuracy": 0.8475471138954163, "num_tokens": 84205347.0, "step": 2214 }, { "epoch": 0.2817707670779799, "ewc_loss": 0.061279296875, "ewc_loss_diag": 4.2438507080078125e-05, "ewc_loss_parallel": 5.7220458984375e-05, "grad_norm": 4.410269260406494, "learning_rate": 9.385332768122086e-06, "loss": 0.505, "mean_token_accuracy": 0.8506287932395935, "num_tokens": 84245788.0, "step": 2215 }, { "epoch": 0.2818979773565704, "ewc_loss": 0.078125, "ewc_loss_diag": 4.2438507080078125e-05, "ewc_loss_parallel": 7.390975952148438e-05, "grad_norm": 5.678960800170898, "learning_rate": 9.389571852479865e-06, "loss": 0.5262, "mean_token_accuracy": 0.8541958332061768, "num_tokens": 84286744.0, "step": 2216 }, { "epoch": 0.28202518763516093, "ewc_loss": 0.06591796875, "ewc_loss_diag": 4.2438507080078125e-05, "ewc_loss_parallel": 6.151199340820312e-05, "grad_norm": 4.555597305297852, "learning_rate": 9.393810936837644e-06, "loss": 0.5204, "mean_token_accuracy": 0.847037136554718, "num_tokens": 84326031.0, "step": 2217 }, { "epoch": 0.2821523979137514, "ewc_loss": 0.0810546875, "ewc_loss_diag": 4.267692565917969e-05, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.805851459503174, "learning_rate": 9.398050021195423e-06, "loss": 0.5638, "mean_token_accuracy": 0.8439907431602478, "num_tokens": 84367697.0, "step": 2218 }, { "epoch": 0.28227960819234194, "ewc_loss": 0.07080078125, "ewc_loss_diag": 4.267692565917969e-05, "ewc_loss_parallel": 6.628036499023438e-05, "grad_norm": 4.79368782043457, "learning_rate": 9.402289105553202e-06, "loss": 0.4997, "mean_token_accuracy": 0.8568621873855591, "num_tokens": 84410647.0, "step": 2219 }, { "epoch": 0.28240681847093246, "ewc_loss": 0.0771484375, "ewc_loss_diag": 4.267692565917969e-05, "ewc_loss_parallel": 7.295608520507812e-05, "grad_norm": 5.717457294464111, "learning_rate": 9.406528189910979e-06, "loss": 0.5309, "mean_token_accuracy": 0.8518859148025513, "num_tokens": 84446839.0, "step": 2220 }, { "epoch": 0.28253402874952294, "ewc_loss": 0.07177734375, "ewc_loss_diag": 4.267692565917969e-05, "ewc_loss_parallel": 6.723403930664062e-05, "grad_norm": 4.970598220825195, "learning_rate": 9.410767274268758e-06, "loss": 0.5009, "mean_token_accuracy": 0.8548233509063721, "num_tokens": 84485017.0, "step": 2221 }, { "epoch": 0.28266123902811346, "ewc_loss": 0.0751953125, "ewc_loss_diag": 4.267692565917969e-05, "ewc_loss_parallel": 7.05718994140625e-05, "grad_norm": 5.469114303588867, "learning_rate": 9.415006358626537e-06, "loss": 0.6055, "mean_token_accuracy": 0.8250122666358948, "num_tokens": 84524518.0, "step": 2222 }, { "epoch": 0.282788449306704, "ewc_loss": 0.07177734375, "ewc_loss_diag": 4.267692565917969e-05, "ewc_loss_parallel": 6.723403930664062e-05, "grad_norm": 4.993584156036377, "learning_rate": 9.419245442984316e-06, "loss": 0.5854, "mean_token_accuracy": 0.8325433731079102, "num_tokens": 84559850.0, "step": 2223 }, { "epoch": 0.28291565958529447, "ewc_loss": 0.072265625, "ewc_loss_diag": 4.291534423828125e-05, "ewc_loss_parallel": 6.771087646484375e-05, "grad_norm": 5.200024127960205, "learning_rate": 9.423484527342095e-06, "loss": 0.5246, "mean_token_accuracy": 0.8490670919418335, "num_tokens": 84594332.0, "step": 2224 }, { "epoch": 0.283042869863885, "ewc_loss": 0.068359375, "ewc_loss_diag": 4.291534423828125e-05, "ewc_loss_parallel": 6.389617919921875e-05, "grad_norm": 4.949124813079834, "learning_rate": 9.427723611699873e-06, "loss": 0.527, "mean_token_accuracy": 0.8487831354141235, "num_tokens": 84628759.0, "step": 2225 }, { "epoch": 0.2831700801424755, "ewc_loss": 0.06884765625, "ewc_loss_diag": 4.291534423828125e-05, "ewc_loss_parallel": 6.437301635742188e-05, "grad_norm": 4.980748176574707, "learning_rate": 9.431962696057652e-06, "loss": 0.5448, "mean_token_accuracy": 0.8434902429580688, "num_tokens": 84672022.0, "step": 2226 }, { "epoch": 0.283297290421066, "ewc_loss": 0.0673828125, "ewc_loss_diag": 4.291534423828125e-05, "ewc_loss_parallel": 6.29425048828125e-05, "grad_norm": 4.894896030426025, "learning_rate": 9.436201780415431e-06, "loss": 0.4895, "mean_token_accuracy": 0.8622559309005737, "num_tokens": 84709266.0, "step": 2227 }, { "epoch": 0.2834245006996565, "ewc_loss": 0.06689453125, "ewc_loss_diag": 4.267692565917969e-05, "ewc_loss_parallel": 6.246566772460938e-05, "grad_norm": 4.862697601318359, "learning_rate": 9.44044086477321e-06, "loss": 0.5404, "mean_token_accuracy": 0.8466453552246094, "num_tokens": 84745502.0, "step": 2228 }, { "epoch": 0.28355171097824705, "ewc_loss": 0.06640625, "ewc_loss_diag": 4.267692565917969e-05, "ewc_loss_parallel": 6.198883056640625e-05, "grad_norm": 4.839620113372803, "learning_rate": 9.444679949130989e-06, "loss": 0.5432, "mean_token_accuracy": 0.8422929048538208, "num_tokens": 84786061.0, "step": 2229 }, { "epoch": 0.2836789212568376, "ewc_loss": 0.0654296875, "ewc_loss_diag": 4.267692565917969e-05, "ewc_loss_parallel": 6.103515625e-05, "grad_norm": 4.808649063110352, "learning_rate": 9.448919033488768e-06, "loss": 0.5556, "mean_token_accuracy": 0.8395488858222961, "num_tokens": 84822550.0, "step": 2230 }, { "epoch": 0.28380613153542805, "ewc_loss": 0.06494140625, "ewc_loss_diag": 4.291534423828125e-05, "ewc_loss_parallel": 6.079673767089844e-05, "grad_norm": 4.784969329833984, "learning_rate": 9.453158117846547e-06, "loss": 0.542, "mean_token_accuracy": 0.844114363193512, "num_tokens": 84860608.0, "step": 2231 }, { "epoch": 0.2839333418140186, "ewc_loss": 0.064453125, "ewc_loss_diag": 4.291534423828125e-05, "ewc_loss_parallel": 6.008148193359375e-05, "grad_norm": 4.790513038635254, "learning_rate": 9.457397202204326e-06, "loss": 0.5704, "mean_token_accuracy": 0.8358461856842041, "num_tokens": 84903065.0, "step": 2232 }, { "epoch": 0.2840605520926091, "ewc_loss": 0.06396484375, "ewc_loss_diag": 4.291534423828125e-05, "ewc_loss_parallel": 5.9604644775390625e-05, "grad_norm": 4.7374067306518555, "learning_rate": 9.461636286562103e-06, "loss": 0.4927, "mean_token_accuracy": 0.8603100776672363, "num_tokens": 84938583.0, "step": 2233 }, { "epoch": 0.2841877623711996, "ewc_loss": 0.064453125, "ewc_loss_diag": 4.291534423828125e-05, "ewc_loss_parallel": 6.031990051269531e-05, "grad_norm": 4.788768291473389, "learning_rate": 9.465875370919882e-06, "loss": 0.5756, "mean_token_accuracy": 0.8352276086807251, "num_tokens": 84975939.0, "step": 2234 }, { "epoch": 0.2843149726497901, "ewc_loss": 0.0625, "ewc_loss_diag": 4.291534423828125e-05, "ewc_loss_parallel": 5.841255187988281e-05, "grad_norm": 4.699852466583252, "learning_rate": 9.47011445527766e-06, "loss": 0.5424, "mean_token_accuracy": 0.8421233892440796, "num_tokens": 85011304.0, "step": 2235 }, { "epoch": 0.28444218292838064, "ewc_loss": 0.06494140625, "ewc_loss_diag": 4.291534423828125e-05, "ewc_loss_parallel": 6.079673767089844e-05, "grad_norm": 4.799041271209717, "learning_rate": 9.47435353963544e-06, "loss": 0.5115, "mean_token_accuracy": 0.8523572087287903, "num_tokens": 85053163.0, "step": 2236 }, { "epoch": 0.2845693932069711, "ewc_loss": 0.062255859375, "ewc_loss_diag": 4.315376281738281e-05, "ewc_loss_parallel": 5.793571472167969e-05, "grad_norm": 4.657679080963135, "learning_rate": 9.478592623993218e-06, "loss": 0.5088, "mean_token_accuracy": 0.8497424125671387, "num_tokens": 85085919.0, "step": 2237 }, { "epoch": 0.28469660348556164, "ewc_loss": 0.06591796875, "ewc_loss_diag": 4.315376281738281e-05, "ewc_loss_parallel": 6.151199340820312e-05, "grad_norm": 4.839883804321289, "learning_rate": 9.482831708350997e-06, "loss": 0.5145, "mean_token_accuracy": 0.8516817092895508, "num_tokens": 85124323.0, "step": 2238 }, { "epoch": 0.28482381376415217, "ewc_loss": 0.061279296875, "ewc_loss_diag": 4.315376281738281e-05, "ewc_loss_parallel": 5.698204040527344e-05, "grad_norm": 4.584743976593018, "learning_rate": 9.487070792708776e-06, "loss": 0.5515, "mean_token_accuracy": 0.8427297472953796, "num_tokens": 85163735.0, "step": 2239 }, { "epoch": 0.28495102404274264, "ewc_loss": 0.0673828125, "ewc_loss_diag": 4.315376281738281e-05, "ewc_loss_parallel": 6.29425048828125e-05, "grad_norm": 4.930502891540527, "learning_rate": 9.491309877066555e-06, "loss": 0.5459, "mean_token_accuracy": 0.8429229259490967, "num_tokens": 85206533.0, "step": 2240 }, { "epoch": 0.28507823432133317, "ewc_loss": 0.060546875, "ewc_loss_diag": 4.315376281738281e-05, "ewc_loss_parallel": 5.626678466796875e-05, "grad_norm": 4.487408638000488, "learning_rate": 9.495548961424334e-06, "loss": 0.5609, "mean_token_accuracy": 0.8358372449874878, "num_tokens": 85248435.0, "step": 2241 }, { "epoch": 0.2852054445999237, "ewc_loss": 0.06982421875, "ewc_loss_diag": 4.315376281738281e-05, "ewc_loss_parallel": 6.532669067382812e-05, "grad_norm": 5.113375663757324, "learning_rate": 9.499788045782113e-06, "loss": 0.5327, "mean_token_accuracy": 0.8493579030036926, "num_tokens": 85285686.0, "step": 2242 }, { "epoch": 0.28533265487851417, "ewc_loss": 0.059814453125, "ewc_loss_diag": 4.315376281738281e-05, "ewc_loss_parallel": 5.53131103515625e-05, "grad_norm": 4.338639736175537, "learning_rate": 9.50402713013989e-06, "loss": 0.5356, "mean_token_accuracy": 0.8441138863563538, "num_tokens": 85327518.0, "step": 2243 }, { "epoch": 0.2854598651571047, "ewc_loss": 0.0751953125, "ewc_loss_diag": 4.315376281738281e-05, "ewc_loss_parallel": 7.05718994140625e-05, "grad_norm": 5.461028099060059, "learning_rate": 9.508266214497669e-06, "loss": 0.5027, "mean_token_accuracy": 0.859611988067627, "num_tokens": 85364309.0, "step": 2244 }, { "epoch": 0.2855870754356952, "ewc_loss": 0.058837890625, "ewc_loss_diag": 4.315376281738281e-05, "ewc_loss_parallel": 5.435943603515625e-05, "grad_norm": 4.187091827392578, "learning_rate": 9.512505298855448e-06, "loss": 0.5223, "mean_token_accuracy": 0.8502721190452576, "num_tokens": 85407231.0, "step": 2245 }, { "epoch": 0.2857142857142857, "ewc_loss": 0.0859375, "ewc_loss_diag": 4.315376281738281e-05, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 6.0491862297058105, "learning_rate": 9.516744383213227e-06, "loss": 0.4902, "mean_token_accuracy": 0.8640661835670471, "num_tokens": 85443753.0, "step": 2246 }, { "epoch": 0.2858414959928762, "ewc_loss": 0.06640625, "ewc_loss_diag": 4.315376281738281e-05, "ewc_loss_parallel": 6.198883056640625e-05, "grad_norm": 4.402705192565918, "learning_rate": 9.520983467571005e-06, "loss": 0.5516, "mean_token_accuracy": 0.8411166667938232, "num_tokens": 85482874.0, "step": 2247 }, { "epoch": 0.28596870627146675, "ewc_loss": 0.09814453125, "ewc_loss_diag": 4.3392181396484375e-05, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.5969038009643555, "learning_rate": 9.525222551928784e-06, "loss": 0.62, "mean_token_accuracy": 0.8376587629318237, "num_tokens": 85521648.0, "step": 2248 }, { "epoch": 0.2860959165500572, "ewc_loss": 0.08203125, "ewc_loss_diag": 4.3392181396484375e-05, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.263987064361572, "learning_rate": 9.529461636286563e-06, "loss": 0.5163, "mean_token_accuracy": 0.8563936948776245, "num_tokens": 85553078.0, "step": 2249 }, { "epoch": 0.28622312682864776, "ewc_loss": 0.09033203125, "ewc_loss_diag": 4.3392181396484375e-05, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 6.332550048828125, "learning_rate": 9.533700720644342e-06, "loss": 0.5042, "mean_token_accuracy": 0.8611365556716919, "num_tokens": 85589034.0, "step": 2250 }, { "epoch": 0.2863503371072383, "ewc_loss": 0.08544921875, "ewc_loss_diag": 4.3392181396484375e-05, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.629337787628174, "learning_rate": 9.53793980500212e-06, "loss": 0.5606, "mean_token_accuracy": 0.8477525115013123, "num_tokens": 85631541.0, "step": 2251 }, { "epoch": 0.28647754738582876, "ewc_loss": 0.08251953125, "ewc_loss_diag": 4.3392181396484375e-05, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.852032661437988, "learning_rate": 9.542178889359898e-06, "loss": 0.5277, "mean_token_accuracy": 0.8539775609970093, "num_tokens": 85668498.0, "step": 2252 }, { "epoch": 0.2866047576644193, "ewc_loss": 0.07763671875, "ewc_loss_diag": 4.3392181396484375e-05, "ewc_loss_parallel": 7.343292236328125e-05, "grad_norm": 5.4296369552612305, "learning_rate": 9.546417973717677e-06, "loss": 0.5559, "mean_token_accuracy": 0.8451015949249268, "num_tokens": 85708419.0, "step": 2253 }, { "epoch": 0.2867319679430098, "ewc_loss": 0.0771484375, "ewc_loss_diag": 4.3392181396484375e-05, "ewc_loss_parallel": 7.295608520507812e-05, "grad_norm": 5.565485954284668, "learning_rate": 9.550657058075456e-06, "loss": 0.5637, "mean_token_accuracy": 0.8427184820175171, "num_tokens": 85751564.0, "step": 2254 }, { "epoch": 0.2868591782216003, "ewc_loss": 0.07373046875, "ewc_loss_diag": 4.3392181396484375e-05, "ewc_loss_parallel": 6.914138793945312e-05, "grad_norm": 5.178339004516602, "learning_rate": 9.554896142433235e-06, "loss": 0.5632, "mean_token_accuracy": 0.8374916315078735, "num_tokens": 85789776.0, "step": 2255 }, { "epoch": 0.2869863885001908, "ewc_loss": 0.07421875, "ewc_loss_diag": 4.3392181396484375e-05, "ewc_loss_parallel": 6.961822509765625e-05, "grad_norm": 5.4080963134765625, "learning_rate": 9.559135226791014e-06, "loss": 0.576, "mean_token_accuracy": 0.8332471251487732, "num_tokens": 85828427.0, "step": 2256 }, { "epoch": 0.28711359877878134, "ewc_loss": 0.06982421875, "ewc_loss_diag": 4.3392181396484375e-05, "ewc_loss_parallel": 6.532669067382812e-05, "grad_norm": 4.964992046356201, "learning_rate": 9.563374311148793e-06, "loss": 0.565, "mean_token_accuracy": 0.8387628197669983, "num_tokens": 85877598.0, "step": 2257 }, { "epoch": 0.2872408090573718, "ewc_loss": 0.07373046875, "ewc_loss_diag": 4.3392181396484375e-05, "ewc_loss_parallel": 6.914138793945312e-05, "grad_norm": 5.359003067016602, "learning_rate": 9.567613395506572e-06, "loss": 0.517, "mean_token_accuracy": 0.854968786239624, "num_tokens": 85908395.0, "step": 2258 }, { "epoch": 0.28736801933596234, "ewc_loss": 0.06689453125, "ewc_loss_diag": 4.363059997558594e-05, "ewc_loss_parallel": 6.246566772460938e-05, "grad_norm": 4.786309242248535, "learning_rate": 9.57185247986435e-06, "loss": 0.4962, "mean_token_accuracy": 0.8571261167526245, "num_tokens": 85945344.0, "step": 2259 }, { "epoch": 0.28749522961455287, "ewc_loss": 0.07373046875, "ewc_loss_diag": 4.363059997558594e-05, "ewc_loss_parallel": 6.914138793945312e-05, "grad_norm": 5.4140214920043945, "learning_rate": 9.57609156422213e-06, "loss": 0.4709, "mean_token_accuracy": 0.8660497069358826, "num_tokens": 85982484.0, "step": 2260 }, { "epoch": 0.28762243989314334, "ewc_loss": 0.06396484375, "ewc_loss_diag": 4.38690185546875e-05, "ewc_loss_parallel": 5.9604644775390625e-05, "grad_norm": 4.646635055541992, "learning_rate": 9.580330648579908e-06, "loss": 0.5465, "mean_token_accuracy": 0.8391774892807007, "num_tokens": 86019496.0, "step": 2261 }, { "epoch": 0.2877496501717339, "ewc_loss": 0.0771484375, "ewc_loss_diag": 4.38690185546875e-05, "ewc_loss_parallel": 7.295608520507812e-05, "grad_norm": 5.616008281707764, "learning_rate": 9.584569732937687e-06, "loss": 0.5518, "mean_token_accuracy": 0.8460426926612854, "num_tokens": 86059344.0, "step": 2262 }, { "epoch": 0.2878768604503244, "ewc_loss": 0.0634765625, "ewc_loss_diag": 4.38690185546875e-05, "ewc_loss_parallel": 5.888938903808594e-05, "grad_norm": 4.599275588989258, "learning_rate": 9.588808817295466e-06, "loss": 0.5249, "mean_token_accuracy": 0.8449926376342773, "num_tokens": 86093090.0, "step": 2263 }, { "epoch": 0.2880040707289149, "ewc_loss": 0.08154296875, "ewc_loss_diag": 4.410743713378906e-05, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.883153438568115, "learning_rate": 9.593047901653245e-06, "loss": 0.5387, "mean_token_accuracy": 0.8481043577194214, "num_tokens": 86127149.0, "step": 2264 }, { "epoch": 0.2881312810075054, "ewc_loss": 0.06787109375, "ewc_loss_diag": 4.410743713378906e-05, "ewc_loss_parallel": 6.341934204101562e-05, "grad_norm": 4.781558036804199, "learning_rate": 9.597286986011024e-06, "loss": 0.4868, "mean_token_accuracy": 0.8592344522476196, "num_tokens": 86163541.0, "step": 2265 }, { "epoch": 0.28825849128609593, "ewc_loss": 0.08251953125, "ewc_loss_diag": 4.410743713378906e-05, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.972220420837402, "learning_rate": 9.601526070368801e-06, "loss": 0.5841, "mean_token_accuracy": 0.8373672962188721, "num_tokens": 86200124.0, "step": 2266 }, { "epoch": 0.2883857015646864, "ewc_loss": 0.0732421875, "ewc_loss_diag": 4.410743713378906e-05, "ewc_loss_parallel": 6.866455078125e-05, "grad_norm": 5.0975141525268555, "learning_rate": 9.60576515472658e-06, "loss": 0.5256, "mean_token_accuracy": 0.8514050245285034, "num_tokens": 86236464.0, "step": 2267 }, { "epoch": 0.28851291184327693, "ewc_loss": 0.0771484375, "ewc_loss_diag": 4.410743713378906e-05, "ewc_loss_parallel": 7.295608520507812e-05, "grad_norm": 5.703194618225098, "learning_rate": 9.610004239084359e-06, "loss": 0.543, "mean_token_accuracy": 0.8465614318847656, "num_tokens": 86277504.0, "step": 2268 }, { "epoch": 0.28864012212186746, "ewc_loss": 0.0732421875, "ewc_loss_diag": 4.410743713378906e-05, "ewc_loss_parallel": 6.866455078125e-05, "grad_norm": 5.2464704513549805, "learning_rate": 9.614243323442138e-06, "loss": 0.6012, "mean_token_accuracy": 0.8269994854927063, "num_tokens": 86314404.0, "step": 2269 }, { "epoch": 0.28876733240045793, "ewc_loss": 0.07275390625, "ewc_loss_diag": 4.410743713378906e-05, "ewc_loss_parallel": 6.818771362304688e-05, "grad_norm": 5.282010555267334, "learning_rate": 9.618482407799916e-06, "loss": 0.5127, "mean_token_accuracy": 0.8518407344818115, "num_tokens": 86350690.0, "step": 2270 }, { "epoch": 0.28889454267904846, "ewc_loss": 0.0732421875, "ewc_loss_diag": 4.410743713378906e-05, "ewc_loss_parallel": 6.866455078125e-05, "grad_norm": 5.249955654144287, "learning_rate": 9.622721492157695e-06, "loss": 0.4777, "mean_token_accuracy": 0.8660768270492554, "num_tokens": 86389294.0, "step": 2271 }, { "epoch": 0.289021752957639, "ewc_loss": 0.0673828125, "ewc_loss_diag": 4.410743713378906e-05, "ewc_loss_parallel": 6.29425048828125e-05, "grad_norm": 4.899041652679443, "learning_rate": 9.626960576515474e-06, "loss": 0.5268, "mean_token_accuracy": 0.8469998836517334, "num_tokens": 86420234.0, "step": 2272 }, { "epoch": 0.28914896323622946, "ewc_loss": 0.07373046875, "ewc_loss_diag": 4.410743713378906e-05, "ewc_loss_parallel": 6.914138793945312e-05, "grad_norm": 5.314916133880615, "learning_rate": 9.631199660873251e-06, "loss": 0.5124, "mean_token_accuracy": 0.8525683283805847, "num_tokens": 86458435.0, "step": 2273 }, { "epoch": 0.28927617351482, "ewc_loss": 0.06298828125, "ewc_loss_diag": 4.4345855712890625e-05, "ewc_loss_parallel": 5.8650970458984375e-05, "grad_norm": 4.59340238571167, "learning_rate": 9.63543874523103e-06, "loss": 0.5058, "mean_token_accuracy": 0.8548570871353149, "num_tokens": 86499082.0, "step": 2274 }, { "epoch": 0.2894033837934105, "ewc_loss": 0.07666015625, "ewc_loss_diag": 4.4345855712890625e-05, "ewc_loss_parallel": 7.2479248046875e-05, "grad_norm": 5.526393413543701, "learning_rate": 9.63967782958881e-06, "loss": 0.5535, "mean_token_accuracy": 0.840002179145813, "num_tokens": 86536989.0, "step": 2275 }, { "epoch": 0.289530594072001, "ewc_loss": 0.0615234375, "ewc_loss_diag": 4.4345855712890625e-05, "ewc_loss_parallel": 5.7220458984375e-05, "grad_norm": 4.443125247955322, "learning_rate": 9.643916913946588e-06, "loss": 0.4916, "mean_token_accuracy": 0.8571571111679077, "num_tokens": 86572154.0, "step": 2276 }, { "epoch": 0.2896578043505915, "ewc_loss": 0.08203125, "ewc_loss_diag": 4.4345855712890625e-05, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.8645758628845215, "learning_rate": 9.648155998304367e-06, "loss": 0.4903, "mean_token_accuracy": 0.86394202709198, "num_tokens": 86610908.0, "step": 2277 }, { "epoch": 0.28978501462918205, "ewc_loss": 0.06640625, "ewc_loss_diag": 4.4345855712890625e-05, "ewc_loss_parallel": 6.198883056640625e-05, "grad_norm": 4.560227870941162, "learning_rate": 9.652395082662146e-06, "loss": 0.5922, "mean_token_accuracy": 0.8272462487220764, "num_tokens": 86653585.0, "step": 2278 }, { "epoch": 0.2899122249077726, "ewc_loss": 0.08984375, "ewc_loss_diag": 4.4345855712890625e-05, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 6.158513069152832, "learning_rate": 9.656634167019925e-06, "loss": 0.5529, "mean_token_accuracy": 0.8450337648391724, "num_tokens": 86690794.0, "step": 2279 }, { "epoch": 0.29003943518636305, "ewc_loss": 0.07373046875, "ewc_loss_diag": 4.4345855712890625e-05, "ewc_loss_parallel": 6.914138793945312e-05, "grad_norm": 4.983960151672363, "learning_rate": 9.660873251377704e-06, "loss": 0.5549, "mean_token_accuracy": 0.842261552810669, "num_tokens": 86729742.0, "step": 2280 }, { "epoch": 0.2901666454649536, "ewc_loss": 0.08544921875, "ewc_loss_diag": 4.4345855712890625e-05, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 6.048730373382568, "learning_rate": 9.66511233573548e-06, "loss": 0.5782, "mean_token_accuracy": 0.843164324760437, "num_tokens": 86765837.0, "step": 2281 }, { "epoch": 0.2902938557435441, "ewc_loss": 0.076171875, "ewc_loss_diag": 4.4345855712890625e-05, "ewc_loss_parallel": 7.200241088867188e-05, "grad_norm": 5.279359817504883, "learning_rate": 9.66935142009326e-06, "loss": 0.5043, "mean_token_accuracy": 0.8566609621047974, "num_tokens": 86799697.0, "step": 2282 }, { "epoch": 0.2904210660221346, "ewc_loss": 0.0771484375, "ewc_loss_diag": 4.4345855712890625e-05, "ewc_loss_parallel": 7.295608520507812e-05, "grad_norm": 5.689570903778076, "learning_rate": 9.673590504451039e-06, "loss": 0.5298, "mean_token_accuracy": 0.8488272428512573, "num_tokens": 86832264.0, "step": 2283 }, { "epoch": 0.2905482763007251, "ewc_loss": 0.07470703125, "ewc_loss_diag": 4.458427429199219e-05, "ewc_loss_parallel": 7.009506225585938e-05, "grad_norm": 5.267982482910156, "learning_rate": 9.677829588808817e-06, "loss": 0.512, "mean_token_accuracy": 0.8482289910316467, "num_tokens": 86864492.0, "step": 2284 }, { "epoch": 0.29067548657931563, "ewc_loss": 0.0732421875, "ewc_loss_diag": 4.458427429199219e-05, "ewc_loss_parallel": 6.866455078125e-05, "grad_norm": 5.357184410095215, "learning_rate": 9.682068673166596e-06, "loss": 0.5843, "mean_token_accuracy": 0.8370632529258728, "num_tokens": 86898186.0, "step": 2285 }, { "epoch": 0.2908026968579061, "ewc_loss": 0.072265625, "ewc_loss_diag": 4.458427429199219e-05, "ewc_loss_parallel": 6.771087646484375e-05, "grad_norm": 5.157025337219238, "learning_rate": 9.686307757524375e-06, "loss": 0.5338, "mean_token_accuracy": 0.8464245796203613, "num_tokens": 86935850.0, "step": 2286 }, { "epoch": 0.29092990713649663, "ewc_loss": 0.07177734375, "ewc_loss_diag": 4.458427429199219e-05, "ewc_loss_parallel": 6.723403930664062e-05, "grad_norm": 5.135386943817139, "learning_rate": 9.690546841882154e-06, "loss": 0.4604, "mean_token_accuracy": 0.8707947731018066, "num_tokens": 86972732.0, "step": 2287 }, { "epoch": 0.29105711741508716, "ewc_loss": 0.07080078125, "ewc_loss_diag": 4.458427429199219e-05, "ewc_loss_parallel": 6.628036499023438e-05, "grad_norm": 5.039344310760498, "learning_rate": 9.694785926239933e-06, "loss": 0.5383, "mean_token_accuracy": 0.8476401567459106, "num_tokens": 87012538.0, "step": 2288 }, { "epoch": 0.29118432769367764, "ewc_loss": 0.0703125, "ewc_loss_diag": 4.482269287109375e-05, "ewc_loss_parallel": 6.580352783203125e-05, "grad_norm": 5.036881923675537, "learning_rate": 9.699025010597712e-06, "loss": 0.5818, "mean_token_accuracy": 0.8336538076400757, "num_tokens": 87046917.0, "step": 2289 }, { "epoch": 0.29131153797226816, "ewc_loss": 0.06787109375, "ewc_loss_diag": 4.458427429199219e-05, "ewc_loss_parallel": 6.341934204101562e-05, "grad_norm": 4.930147647857666, "learning_rate": 9.70326409495549e-06, "loss": 0.5412, "mean_token_accuracy": 0.843738317489624, "num_tokens": 87086002.0, "step": 2290 }, { "epoch": 0.2914387482508587, "ewc_loss": 0.0693359375, "ewc_loss_diag": 4.482269287109375e-05, "ewc_loss_parallel": 6.4849853515625e-05, "grad_norm": 4.998393535614014, "learning_rate": 9.70750317931327e-06, "loss": 0.5232, "mean_token_accuracy": 0.8525367379188538, "num_tokens": 87126138.0, "step": 2291 }, { "epoch": 0.29156595852944917, "ewc_loss": 0.06689453125, "ewc_loss_diag": 4.482269287109375e-05, "ewc_loss_parallel": 6.246566772460938e-05, "grad_norm": 4.8276777267456055, "learning_rate": 9.711742263671049e-06, "loss": 0.514, "mean_token_accuracy": 0.8513718843460083, "num_tokens": 87163062.0, "step": 2292 }, { "epoch": 0.2916931688080397, "ewc_loss": 0.06982421875, "ewc_loss_diag": 4.482269287109375e-05, "ewc_loss_parallel": 6.532669067382812e-05, "grad_norm": 5.029181957244873, "learning_rate": 9.715981348028827e-06, "loss": 0.5074, "mean_token_accuracy": 0.8547389507293701, "num_tokens": 87203836.0, "step": 2293 }, { "epoch": 0.2918203790866302, "ewc_loss": 0.06494140625, "ewc_loss_diag": 4.506111145019531e-05, "ewc_loss_parallel": 6.031990051269531e-05, "grad_norm": 4.717039585113525, "learning_rate": 9.720220432386606e-06, "loss": 0.5075, "mean_token_accuracy": 0.8552749752998352, "num_tokens": 87236854.0, "step": 2294 }, { "epoch": 0.2919475893652207, "ewc_loss": 0.0712890625, "ewc_loss_diag": 4.506111145019531e-05, "ewc_loss_parallel": 6.67572021484375e-05, "grad_norm": 5.126378059387207, "learning_rate": 9.724459516744385e-06, "loss": 0.4994, "mean_token_accuracy": 0.8600050210952759, "num_tokens": 87276316.0, "step": 2295 }, { "epoch": 0.2920747996438112, "ewc_loss": 0.06298828125, "ewc_loss_diag": 4.506111145019531e-05, "ewc_loss_parallel": 5.841255187988281e-05, "grad_norm": 4.586432456970215, "learning_rate": 9.728698601102162e-06, "loss": 0.4864, "mean_token_accuracy": 0.8601685166358948, "num_tokens": 87314504.0, "step": 2296 }, { "epoch": 0.29220200992240175, "ewc_loss": 0.07421875, "ewc_loss_diag": 4.506111145019531e-05, "ewc_loss_parallel": 6.961822509765625e-05, "grad_norm": 5.332152843475342, "learning_rate": 9.732937685459941e-06, "loss": 0.528, "mean_token_accuracy": 0.8474320769309998, "num_tokens": 87348403.0, "step": 2297 }, { "epoch": 0.2923292202009922, "ewc_loss": 0.06103515625, "ewc_loss_diag": 4.506111145019531e-05, "ewc_loss_parallel": 5.650520324707031e-05, "grad_norm": 4.440547943115234, "learning_rate": 9.73717676981772e-06, "loss": 0.5544, "mean_token_accuracy": 0.8423210978507996, "num_tokens": 87393044.0, "step": 2298 }, { "epoch": 0.29245643047958275, "ewc_loss": 0.0791015625, "ewc_loss_diag": 4.506111145019531e-05, "ewc_loss_parallel": 7.486343383789062e-05, "grad_norm": 5.6880669593811035, "learning_rate": 9.741415854175499e-06, "loss": 0.596, "mean_token_accuracy": 0.8348285555839539, "num_tokens": 87430930.0, "step": 2299 }, { "epoch": 0.2925836407581733, "ewc_loss": 0.0625, "ewc_loss_diag": 4.506111145019531e-05, "ewc_loss_parallel": 5.793571472167969e-05, "grad_norm": 4.401177883148193, "learning_rate": 9.745654938533278e-06, "loss": 0.5419, "mean_token_accuracy": 0.8411523103713989, "num_tokens": 87467941.0, "step": 2300 }, { "epoch": 0.29271085103676375, "ewc_loss": 0.08935546875, "ewc_loss_diag": 4.506111145019531e-05, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 6.163758754730225, "learning_rate": 9.749894022891057e-06, "loss": 0.5905, "mean_token_accuracy": 0.8380457162857056, "num_tokens": 87507453.0, "step": 2301 }, { "epoch": 0.2928380613153543, "ewc_loss": 0.0693359375, "ewc_loss_diag": 4.506111145019531e-05, "ewc_loss_parallel": 6.4849853515625e-05, "grad_norm": 4.703843593597412, "learning_rate": 9.754133107248836e-06, "loss": 0.486, "mean_token_accuracy": 0.8602330684661865, "num_tokens": 87545983.0, "step": 2302 }, { "epoch": 0.2929652715939448, "ewc_loss": 0.09619140625, "ewc_loss_diag": 4.5299530029296875e-05, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.416731357574463, "learning_rate": 9.758372191606615e-06, "loss": 0.5992, "mean_token_accuracy": 0.8346688151359558, "num_tokens": 87582352.0, "step": 2303 }, { "epoch": 0.2930924818725353, "ewc_loss": 0.07958984375, "ewc_loss_diag": 4.5299530029296875e-05, "ewc_loss_parallel": 7.534027099609375e-05, "grad_norm": 5.304421424865723, "learning_rate": 9.762611275964392e-06, "loss": 0.4919, "mean_token_accuracy": 0.863431453704834, "num_tokens": 87616668.0, "step": 2304 }, { "epoch": 0.2932196921511258, "ewc_loss": 0.0859375, "ewc_loss_diag": 4.5299530029296875e-05, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 6.096672534942627, "learning_rate": 9.76685036032217e-06, "loss": 0.5692, "mean_token_accuracy": 0.8412399888038635, "num_tokens": 87653008.0, "step": 2305 }, { "epoch": 0.29334690242971634, "ewc_loss": 0.08154296875, "ewc_loss_diag": 4.553794860839844e-05, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.487593173980713, "learning_rate": 9.77108944467995e-06, "loss": 0.513, "mean_token_accuracy": 0.8561681509017944, "num_tokens": 87694985.0, "step": 2306 }, { "epoch": 0.2934741127083068, "ewc_loss": 0.0771484375, "ewc_loss_diag": 4.57763671875e-05, "ewc_loss_parallel": 7.295608520507812e-05, "grad_norm": 5.663783073425293, "learning_rate": 9.775328529037728e-06, "loss": 0.553, "mean_token_accuracy": 0.8428395986557007, "num_tokens": 87730841.0, "step": 2307 }, { "epoch": 0.29360132298689734, "ewc_loss": 0.07666015625, "ewc_loss_diag": 4.57763671875e-05, "ewc_loss_parallel": 7.2479248046875e-05, "grad_norm": 5.385138034820557, "learning_rate": 9.779567613395507e-06, "loss": 0.5839, "mean_token_accuracy": 0.8391147255897522, "num_tokens": 87762215.0, "step": 2308 }, { "epoch": 0.29372853326548787, "ewc_loss": 0.0732421875, "ewc_loss_diag": 4.57763671875e-05, "ewc_loss_parallel": 6.866455078125e-05, "grad_norm": 5.323825359344482, "learning_rate": 9.783806697753286e-06, "loss": 0.5234, "mean_token_accuracy": 0.8522486686706543, "num_tokens": 87799150.0, "step": 2309 }, { "epoch": 0.29385574354407834, "ewc_loss": 0.07421875, "ewc_loss_diag": 4.57763671875e-05, "ewc_loss_parallel": 6.961822509765625e-05, "grad_norm": 5.230409145355225, "learning_rate": 9.788045782111065e-06, "loss": 0.4907, "mean_token_accuracy": 0.8595998287200928, "num_tokens": 87842463.0, "step": 2310 }, { "epoch": 0.29398295382266887, "ewc_loss": 0.0712890625, "ewc_loss_diag": 4.57763671875e-05, "ewc_loss_parallel": 6.67572021484375e-05, "grad_norm": 5.11325216293335, "learning_rate": 9.792284866468842e-06, "loss": 0.4796, "mean_token_accuracy": 0.860914945602417, "num_tokens": 87880093.0, "step": 2311 }, { "epoch": 0.2941101641012594, "ewc_loss": 0.07177734375, "ewc_loss_diag": 4.57763671875e-05, "ewc_loss_parallel": 6.723403930664062e-05, "grad_norm": 5.138679504394531, "learning_rate": 9.796523950826621e-06, "loss": 0.4864, "mean_token_accuracy": 0.8596149682998657, "num_tokens": 87912713.0, "step": 2312 }, { "epoch": 0.29423737437984987, "ewc_loss": 0.0693359375, "ewc_loss_diag": 4.57763671875e-05, "ewc_loss_parallel": 6.4849853515625e-05, "grad_norm": 4.9939141273498535, "learning_rate": 9.8007630351844e-06, "loss": 0.5577, "mean_token_accuracy": 0.8423575162887573, "num_tokens": 87946357.0, "step": 2313 }, { "epoch": 0.2943645846584404, "ewc_loss": 0.07080078125, "ewc_loss_diag": 4.57763671875e-05, "ewc_loss_parallel": 6.628036499023438e-05, "grad_norm": 5.0782575607299805, "learning_rate": 9.805002119542179e-06, "loss": 0.5889, "mean_token_accuracy": 0.830715537071228, "num_tokens": 87981102.0, "step": 2314 }, { "epoch": 0.2944917949370309, "ewc_loss": 0.0673828125, "ewc_loss_diag": 4.57763671875e-05, "ewc_loss_parallel": 6.29425048828125e-05, "grad_norm": 4.880861759185791, "learning_rate": 9.809241203899958e-06, "loss": 0.5207, "mean_token_accuracy": 0.8527433276176453, "num_tokens": 88022325.0, "step": 2315 }, { "epoch": 0.2946190052156214, "ewc_loss": 0.0703125, "ewc_loss_diag": 4.57763671875e-05, "ewc_loss_parallel": 6.580352783203125e-05, "grad_norm": 5.068162441253662, "learning_rate": 9.813480288257737e-06, "loss": 0.546, "mean_token_accuracy": 0.8410890102386475, "num_tokens": 88057254.0, "step": 2316 }, { "epoch": 0.2947462154942119, "ewc_loss": 0.0673828125, "ewc_loss_diag": 4.601478576660156e-05, "ewc_loss_parallel": 6.29425048828125e-05, "grad_norm": 4.793894290924072, "learning_rate": 9.817719372615516e-06, "loss": 0.4974, "mean_token_accuracy": 0.8582190275192261, "num_tokens": 88095667.0, "step": 2317 }, { "epoch": 0.29487342577280246, "ewc_loss": 0.07080078125, "ewc_loss_diag": 4.601478576660156e-05, "ewc_loss_parallel": 6.628036499023438e-05, "grad_norm": 5.106782913208008, "learning_rate": 9.821958456973294e-06, "loss": 0.4715, "mean_token_accuracy": 0.865382730960846, "num_tokens": 88129634.0, "step": 2318 }, { "epoch": 0.29500063605139293, "ewc_loss": 0.064453125, "ewc_loss_diag": 4.601478576660156e-05, "ewc_loss_parallel": 5.984306335449219e-05, "grad_norm": 4.699107646942139, "learning_rate": 9.826197541331073e-06, "loss": 0.4793, "mean_token_accuracy": 0.8609001040458679, "num_tokens": 88164592.0, "step": 2319 }, { "epoch": 0.29512784632998346, "ewc_loss": 0.07177734375, "ewc_loss_diag": 4.601478576660156e-05, "ewc_loss_parallel": 6.723403930664062e-05, "grad_norm": 5.222464084625244, "learning_rate": 9.830436625688852e-06, "loss": 0.5437, "mean_token_accuracy": 0.8455254435539246, "num_tokens": 88202390.0, "step": 2320 }, { "epoch": 0.295255056608574, "ewc_loss": 0.06298828125, "ewc_loss_diag": 4.601478576660156e-05, "ewc_loss_parallel": 5.8650970458984375e-05, "grad_norm": 4.54865837097168, "learning_rate": 9.834675710046631e-06, "loss": 0.4909, "mean_token_accuracy": 0.8568302989006042, "num_tokens": 88243104.0, "step": 2321 }, { "epoch": 0.29538226688716446, "ewc_loss": 0.07568359375, "ewc_loss_diag": 4.601478576660156e-05, "ewc_loss_parallel": 7.152557373046875e-05, "grad_norm": 5.482663631439209, "learning_rate": 9.83891479440441e-06, "loss": 0.4917, "mean_token_accuracy": 0.8639122247695923, "num_tokens": 88284038.0, "step": 2322 }, { "epoch": 0.295509477165755, "ewc_loss": 0.0625, "ewc_loss_diag": 4.601478576660156e-05, "ewc_loss_parallel": 5.793571472167969e-05, "grad_norm": 4.409456729888916, "learning_rate": 9.843153878762189e-06, "loss": 0.5535, "mean_token_accuracy": 0.8424594402313232, "num_tokens": 88320705.0, "step": 2323 }, { "epoch": 0.2956366874443455, "ewc_loss": 0.083984375, "ewc_loss_diag": 4.601478576660156e-05, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.912508487701416, "learning_rate": 9.847392963119968e-06, "loss": 0.5232, "mean_token_accuracy": 0.854129433631897, "num_tokens": 88359636.0, "step": 2324 }, { "epoch": 0.295763897722936, "ewc_loss": 0.0654296875, "ewc_loss_diag": 4.601478576660156e-05, "ewc_loss_parallel": 6.103515625e-05, "grad_norm": 4.4150309562683105, "learning_rate": 9.851632047477747e-06, "loss": 0.5273, "mean_token_accuracy": 0.8456449508666992, "num_tokens": 88402029.0, "step": 2325 }, { "epoch": 0.2958911080015265, "ewc_loss": 0.09521484375, "ewc_loss_diag": 4.601478576660156e-05, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.460097312927246, "learning_rate": 9.855871131835524e-06, "loss": 0.4778, "mean_token_accuracy": 0.8700559735298157, "num_tokens": 88438245.0, "step": 2326 }, { "epoch": 0.29601831828011704, "ewc_loss": 0.07568359375, "ewc_loss_diag": 4.601478576660156e-05, "ewc_loss_parallel": 7.152557373046875e-05, "grad_norm": 4.958386421203613, "learning_rate": 9.860110216193303e-06, "loss": 0.4972, "mean_token_accuracy": 0.8605628609657288, "num_tokens": 88477987.0, "step": 2327 }, { "epoch": 0.2961455285587075, "ewc_loss": 0.09814453125, "ewc_loss_diag": 4.601478576660156e-05, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.552977561950684, "learning_rate": 9.864349300551082e-06, "loss": 0.5104, "mean_token_accuracy": 0.8624299764633179, "num_tokens": 88510759.0, "step": 2328 }, { "epoch": 0.29627273883729804, "ewc_loss": 0.08642578125, "ewc_loss_diag": 4.601478576660156e-05, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.576662540435791, "learning_rate": 9.86858838490886e-06, "loss": 0.5132, "mean_token_accuracy": 0.8556616306304932, "num_tokens": 88544819.0, "step": 2329 }, { "epoch": 0.2963999491158886, "ewc_loss": 0.08642578125, "ewc_loss_diag": 4.601478576660156e-05, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 6.083737850189209, "learning_rate": 9.87282746926664e-06, "loss": 0.5171, "mean_token_accuracy": 0.8560726046562195, "num_tokens": 88584560.0, "step": 2330 }, { "epoch": 0.2965271593944791, "ewc_loss": 0.0830078125, "ewc_loss_diag": 4.601478576660156e-05, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.5653557777404785, "learning_rate": 9.877066553624418e-06, "loss": 0.5095, "mean_token_accuracy": 0.8567103147506714, "num_tokens": 88620733.0, "step": 2331 }, { "epoch": 0.2966543696730696, "ewc_loss": 0.07763671875, "ewc_loss_diag": 4.601478576660156e-05, "ewc_loss_parallel": 7.343292236328125e-05, "grad_norm": 5.705792427062988, "learning_rate": 9.881305637982197e-06, "loss": 0.5214, "mean_token_accuracy": 0.8532763123512268, "num_tokens": 88661306.0, "step": 2332 }, { "epoch": 0.2967815799516601, "ewc_loss": 0.078125, "ewc_loss_diag": 4.601478576660156e-05, "ewc_loss_parallel": 7.390975952148438e-05, "grad_norm": 5.375067710876465, "learning_rate": 9.885544722339976e-06, "loss": 0.5023, "mean_token_accuracy": 0.8578047752380371, "num_tokens": 88696886.0, "step": 2333 }, { "epoch": 0.29690879023025063, "ewc_loss": 0.0751953125, "ewc_loss_diag": 4.601478576660156e-05, "ewc_loss_parallel": 7.05718994140625e-05, "grad_norm": 5.427128791809082, "learning_rate": 9.889783806697753e-06, "loss": 0.4758, "mean_token_accuracy": 0.864083468914032, "num_tokens": 88733242.0, "step": 2334 }, { "epoch": 0.2970360005088411, "ewc_loss": 0.07470703125, "ewc_loss_diag": 4.601478576660156e-05, "ewc_loss_parallel": 7.009506225585938e-05, "grad_norm": 5.22067928314209, "learning_rate": 9.894022891055532e-06, "loss": 0.4786, "mean_token_accuracy": 0.8658663034439087, "num_tokens": 88769599.0, "step": 2335 }, { "epoch": 0.29716321078743163, "ewc_loss": 0.0732421875, "ewc_loss_diag": 4.601478576660156e-05, "ewc_loss_parallel": 6.866455078125e-05, "grad_norm": 5.25918436050415, "learning_rate": 9.898261975413311e-06, "loss": 0.4598, "mean_token_accuracy": 0.8709481954574585, "num_tokens": 88802885.0, "step": 2336 }, { "epoch": 0.29729042106602216, "ewc_loss": 0.07177734375, "ewc_loss_diag": 4.601478576660156e-05, "ewc_loss_parallel": 6.723403930664062e-05, "grad_norm": 5.079748630523682, "learning_rate": 9.90250105977109e-06, "loss": 0.5159, "mean_token_accuracy": 0.8535852432250977, "num_tokens": 88837061.0, "step": 2337 }, { "epoch": 0.29741763134461263, "ewc_loss": 0.072265625, "ewc_loss_diag": 4.601478576660156e-05, "ewc_loss_parallel": 6.771087646484375e-05, "grad_norm": 5.172412872314453, "learning_rate": 9.906740144128869e-06, "loss": 0.527, "mean_token_accuracy": 0.8455209732055664, "num_tokens": 88874672.0, "step": 2338 }, { "epoch": 0.29754484162320316, "ewc_loss": 0.0693359375, "ewc_loss_diag": 4.6253204345703125e-05, "ewc_loss_parallel": 6.4849853515625e-05, "grad_norm": 4.953679084777832, "learning_rate": 9.910979228486648e-06, "loss": 0.4745, "mean_token_accuracy": 0.8622499108314514, "num_tokens": 88910520.0, "step": 2339 }, { "epoch": 0.2976720519017937, "ewc_loss": 0.072265625, "ewc_loss_diag": 4.6253204345703125e-05, "ewc_loss_parallel": 6.723403930664062e-05, "grad_norm": 5.1660661697387695, "learning_rate": 9.915218312844427e-06, "loss": 0.5479, "mean_token_accuracy": 0.8459820747375488, "num_tokens": 88949495.0, "step": 2340 }, { "epoch": 0.29779926218038416, "ewc_loss": 0.0673828125, "ewc_loss_diag": 4.6253204345703125e-05, "ewc_loss_parallel": 6.246566772460938e-05, "grad_norm": 4.826284408569336, "learning_rate": 9.919457397202205e-06, "loss": 0.5409, "mean_token_accuracy": 0.8441202044487, "num_tokens": 88988776.0, "step": 2341 }, { "epoch": 0.2979264724589747, "ewc_loss": 0.072265625, "ewc_loss_diag": 4.6253204345703125e-05, "ewc_loss_parallel": 6.771087646484375e-05, "grad_norm": 5.248124122619629, "learning_rate": 9.923696481559984e-06, "loss": 0.4949, "mean_token_accuracy": 0.8621225357055664, "num_tokens": 89018983.0, "step": 2342 }, { "epoch": 0.2980536827375652, "ewc_loss": 0.0654296875, "ewc_loss_diag": 4.6253204345703125e-05, "ewc_loss_parallel": 6.0558319091796875e-05, "grad_norm": 4.688836574554443, "learning_rate": 9.927935565917761e-06, "loss": 0.5417, "mean_token_accuracy": 0.8425043821334839, "num_tokens": 89055834.0, "step": 2343 }, { "epoch": 0.2981808930161557, "ewc_loss": 0.076171875, "ewc_loss_diag": 4.6253204345703125e-05, "ewc_loss_parallel": 7.152557373046875e-05, "grad_norm": 5.428810119628906, "learning_rate": 9.93217465027554e-06, "loss": 0.5413, "mean_token_accuracy": 0.8519613742828369, "num_tokens": 89098612.0, "step": 2344 }, { "epoch": 0.2983081032947462, "ewc_loss": 0.0634765625, "ewc_loss_diag": 4.6253204345703125e-05, "ewc_loss_parallel": 5.888938903808594e-05, "grad_norm": 4.554471492767334, "learning_rate": 9.93641373463332e-06, "loss": 0.5896, "mean_token_accuracy": 0.8343020081520081, "num_tokens": 89130169.0, "step": 2345 }, { "epoch": 0.29843531357333675, "ewc_loss": 0.0810546875, "ewc_loss_diag": 4.6253204345703125e-05, "ewc_loss_parallel": 7.62939453125e-05, "grad_norm": 5.756786346435547, "learning_rate": 9.940652818991098e-06, "loss": 0.4952, "mean_token_accuracy": 0.8642511367797852, "num_tokens": 89170346.0, "step": 2346 }, { "epoch": 0.2985625238519272, "ewc_loss": 0.064453125, "ewc_loss_diag": 4.6253204345703125e-05, "ewc_loss_parallel": 5.984306335449219e-05, "grad_norm": 4.472876071929932, "learning_rate": 9.944891903348877e-06, "loss": 0.5451, "mean_token_accuracy": 0.8407915830612183, "num_tokens": 89211713.0, "step": 2347 }, { "epoch": 0.29868973413051775, "ewc_loss": 0.0927734375, "ewc_loss_diag": 4.6253204345703125e-05, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 6.240184307098389, "learning_rate": 9.949130987706656e-06, "loss": 0.5219, "mean_token_accuracy": 0.8594256639480591, "num_tokens": 89252842.0, "step": 2348 }, { "epoch": 0.2988169444091083, "ewc_loss": 0.0712890625, "ewc_loss_diag": 4.6253204345703125e-05, "ewc_loss_parallel": 6.67572021484375e-05, "grad_norm": 4.76965856552124, "learning_rate": 9.953370072064435e-06, "loss": 0.4835, "mean_token_accuracy": 0.8606844544410706, "num_tokens": 89291914.0, "step": 2349 }, { "epoch": 0.29894415468769875, "ewc_loss": 0.09765625, "ewc_loss_diag": 4.6253204345703125e-05, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.515519618988037, "learning_rate": 9.957609156422214e-06, "loss": 0.5952, "mean_token_accuracy": 0.8381655812263489, "num_tokens": 89333912.0, "step": 2350 }, { "epoch": 0.2990713649662893, "ewc_loss": 0.0830078125, "ewc_loss_diag": 4.649162292480469e-05, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.396835803985596, "learning_rate": 9.961848240779993e-06, "loss": 0.5242, "mean_token_accuracy": 0.8542461395263672, "num_tokens": 89363603.0, "step": 2351 }, { "epoch": 0.2991985752448798, "ewc_loss": 0.0888671875, "ewc_loss_diag": 4.649162292480469e-05, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 6.160119533538818, "learning_rate": 9.966087325137771e-06, "loss": 0.5082, "mean_token_accuracy": 0.8619019985198975, "num_tokens": 89404392.0, "step": 2352 }, { "epoch": 0.2993257855234703, "ewc_loss": 0.08447265625, "ewc_loss_diag": 4.673004150390625e-05, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.569301605224609, "learning_rate": 9.97032640949555e-06, "loss": 0.5711, "mean_token_accuracy": 0.8373454809188843, "num_tokens": 89444180.0, "step": 2353 }, { "epoch": 0.2994529958020608, "ewc_loss": 0.078125, "ewc_loss_diag": 4.673004150390625e-05, "ewc_loss_parallel": 7.343292236328125e-05, "grad_norm": 5.712460994720459, "learning_rate": 9.97456549385333e-06, "loss": 0.5704, "mean_token_accuracy": 0.8356689810752869, "num_tokens": 89483875.0, "step": 2354 }, { "epoch": 0.29958020608065133, "ewc_loss": 0.08056640625, "ewc_loss_diag": 4.673004150390625e-05, "ewc_loss_parallel": 7.581710815429688e-05, "grad_norm": 5.4693603515625, "learning_rate": 9.978804578211108e-06, "loss": 0.5635, "mean_token_accuracy": 0.8423289060592651, "num_tokens": 89519544.0, "step": 2355 }, { "epoch": 0.2997074163592418, "ewc_loss": 0.07373046875, "ewc_loss_diag": 4.673004150390625e-05, "ewc_loss_parallel": 6.866455078125e-05, "grad_norm": 5.350372314453125, "learning_rate": 9.983043662568887e-06, "loss": 0.5455, "mean_token_accuracy": 0.8471692204475403, "num_tokens": 89552635.0, "step": 2356 }, { "epoch": 0.29983462663783234, "ewc_loss": 0.07666015625, "ewc_loss_diag": 4.673004150390625e-05, "ewc_loss_parallel": 7.200241088867188e-05, "grad_norm": 5.361106872558594, "learning_rate": 9.987282746926664e-06, "loss": 0.5108, "mean_token_accuracy": 0.8575481176376343, "num_tokens": 89588514.0, "step": 2357 }, { "epoch": 0.29996183691642286, "ewc_loss": 0.07177734375, "ewc_loss_diag": 4.673004150390625e-05, "ewc_loss_parallel": 6.67572021484375e-05, "grad_norm": 5.093102931976318, "learning_rate": 9.991521831284443e-06, "loss": 0.5141, "mean_token_accuracy": 0.8538689613342285, "num_tokens": 89629029.0, "step": 2358 }, { "epoch": 0.30008904719501334, "ewc_loss": 0.0751953125, "ewc_loss_diag": 4.673004150390625e-05, "ewc_loss_parallel": 7.009506225585938e-05, "grad_norm": 5.322713375091553, "learning_rate": 9.995760915642222e-06, "loss": 0.4609, "mean_token_accuracy": 0.8694629669189453, "num_tokens": 89659214.0, "step": 2359 }, { "epoch": 0.30021625747360386, "ewc_loss": 0.06982421875, "ewc_loss_diag": 4.673004150390625e-05, "ewc_loss_parallel": 6.4849853515625e-05, "grad_norm": 4.918934345245361, "learning_rate": 1e-05, "loss": 0.48, "mean_token_accuracy": 0.8603367209434509, "num_tokens": 89695342.0, "step": 2360 }, { "epoch": 0.3003434677521944, "ewc_loss": 0.07568359375, "ewc_loss_diag": 4.673004150390625e-05, "ewc_loss_parallel": 7.05718994140625e-05, "grad_norm": 5.358338832855225, "learning_rate": 1e-05, "loss": 0.5352, "mean_token_accuracy": 0.8495854139328003, "num_tokens": 89731209.0, "step": 2361 }, { "epoch": 0.30047067803078487, "ewc_loss": 0.0673828125, "ewc_loss_diag": 4.673004150390625e-05, "ewc_loss_parallel": 6.246566772460938e-05, "grad_norm": 4.778491497039795, "learning_rate": 1e-05, "loss": 0.5666, "mean_token_accuracy": 0.8338465094566345, "num_tokens": 89771516.0, "step": 2362 }, { "epoch": 0.3005978883093754, "ewc_loss": 0.07666015625, "ewc_loss_diag": 4.7206878662109375e-05, "ewc_loss_parallel": 7.200241088867188e-05, "grad_norm": 5.473872184753418, "learning_rate": 1e-05, "loss": 0.57, "mean_token_accuracy": 0.8388355374336243, "num_tokens": 89817929.0, "step": 2363 }, { "epoch": 0.3007250985879659, "ewc_loss": 0.06689453125, "ewc_loss_diag": 4.7206878662109375e-05, "ewc_loss_parallel": 6.198883056640625e-05, "grad_norm": 4.674295425415039, "learning_rate": 1e-05, "loss": 0.6353, "mean_token_accuracy": 0.8227249383926392, "num_tokens": 89850714.0, "step": 2364 }, { "epoch": 0.3008523088665564, "ewc_loss": 0.080078125, "ewc_loss_diag": 4.7206878662109375e-05, "ewc_loss_parallel": 7.534027099609375e-05, "grad_norm": 5.675057888031006, "learning_rate": 1e-05, "loss": 0.5485, "mean_token_accuracy": 0.8484828472137451, "num_tokens": 89885686.0, "step": 2365 }, { "epoch": 0.3009795191451469, "ewc_loss": 0.0673828125, "ewc_loss_diag": 4.7206878662109375e-05, "ewc_loss_parallel": 6.246566772460938e-05, "grad_norm": 4.59160852432251, "learning_rate": 1e-05, "loss": 0.4944, "mean_token_accuracy": 0.857168972492218, "num_tokens": 89926549.0, "step": 2366 }, { "epoch": 0.30110672942373745, "ewc_loss": 0.0859375, "ewc_loss_diag": 4.7206878662109375e-05, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.977332592010498, "learning_rate": 1e-05, "loss": 0.5775, "mean_token_accuracy": 0.8378018140792847, "num_tokens": 89969031.0, "step": 2367 }, { "epoch": 0.3012339397023279, "ewc_loss": 0.06982421875, "ewc_loss_diag": 4.7206878662109375e-05, "ewc_loss_parallel": 6.4849853515625e-05, "grad_norm": 4.714666843414307, "learning_rate": 1e-05, "loss": 0.539, "mean_token_accuracy": 0.8444942235946655, "num_tokens": 90000348.0, "step": 2368 }, { "epoch": 0.30136114998091845, "ewc_loss": 0.091796875, "ewc_loss_diag": 4.7206878662109375e-05, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 6.217443943023682, "learning_rate": 1e-05, "loss": 0.5506, "mean_token_accuracy": 0.8495907783508301, "num_tokens": 90033945.0, "step": 2369 }, { "epoch": 0.301488360259509, "ewc_loss": 0.0751953125, "ewc_loss_diag": 4.7206878662109375e-05, "ewc_loss_parallel": 7.009506225585938e-05, "grad_norm": 5.013567924499512, "learning_rate": 1e-05, "loss": 0.5081, "mean_token_accuracy": 0.8605670928955078, "num_tokens": 90069537.0, "step": 2370 }, { "epoch": 0.30161557053809945, "ewc_loss": 0.0908203125, "ewc_loss_diag": 4.7206878662109375e-05, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 6.16084623336792, "learning_rate": 1e-05, "loss": 0.5137, "mean_token_accuracy": 0.8603540658950806, "num_tokens": 90107923.0, "step": 2371 }, { "epoch": 0.30174278081669, "ewc_loss": 0.07958984375, "ewc_loss_diag": 4.7206878662109375e-05, "ewc_loss_parallel": 7.486343383789062e-05, "grad_norm": 5.286211967468262, "learning_rate": 1e-05, "loss": 0.5693, "mean_token_accuracy": 0.8392865061759949, "num_tokens": 90152211.0, "step": 2372 }, { "epoch": 0.3018699910952805, "ewc_loss": 0.08154296875, "ewc_loss_diag": 4.76837158203125e-05, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.84613037109375, "learning_rate": 1e-05, "loss": 0.5514, "mean_token_accuracy": 0.8460561037063599, "num_tokens": 90193687.0, "step": 2373 }, { "epoch": 0.301997201373871, "ewc_loss": 0.080078125, "ewc_loss_diag": 4.76837158203125e-05, "ewc_loss_parallel": 7.534027099609375e-05, "grad_norm": 5.3350419998168945, "learning_rate": 1e-05, "loss": 0.4874, "mean_token_accuracy": 0.8660441637039185, "num_tokens": 90232639.0, "step": 2374 }, { "epoch": 0.3021244116524615, "ewc_loss": 0.076171875, "ewc_loss_diag": 4.76837158203125e-05, "ewc_loss_parallel": 7.152557373046875e-05, "grad_norm": 5.485772609710693, "learning_rate": 1e-05, "loss": 0.5467, "mean_token_accuracy": 0.8430290222167969, "num_tokens": 90273344.0, "step": 2375 }, { "epoch": 0.30225162193105204, "ewc_loss": 0.0771484375, "ewc_loss_diag": 4.76837158203125e-05, "ewc_loss_parallel": 7.2479248046875e-05, "grad_norm": 5.3052287101745605, "learning_rate": 1e-05, "loss": 0.5429, "mean_token_accuracy": 0.8488860130310059, "num_tokens": 90309999.0, "step": 2376 }, { "epoch": 0.3023788322096425, "ewc_loss": 0.07275390625, "ewc_loss_diag": 4.76837158203125e-05, "ewc_loss_parallel": 6.771087646484375e-05, "grad_norm": 5.184545993804932, "learning_rate": 1e-05, "loss": 0.5085, "mean_token_accuracy": 0.8529165983200073, "num_tokens": 90349049.0, "step": 2377 }, { "epoch": 0.30250604248823304, "ewc_loss": 0.0751953125, "ewc_loss_diag": 4.76837158203125e-05, "ewc_loss_parallel": 7.009506225585938e-05, "grad_norm": 5.267798900604248, "learning_rate": 1e-05, "loss": 0.5558, "mean_token_accuracy": 0.8436425924301147, "num_tokens": 90388795.0, "step": 2378 }, { "epoch": 0.30263325276682357, "ewc_loss": 0.0712890625, "ewc_loss_diag": 4.76837158203125e-05, "ewc_loss_parallel": 6.628036499023438e-05, "grad_norm": 4.992819786071777, "learning_rate": 1e-05, "loss": 0.5686, "mean_token_accuracy": 0.8327637910842896, "num_tokens": 90423785.0, "step": 2379 }, { "epoch": 0.3027604630454141, "ewc_loss": 0.07421875, "ewc_loss_diag": 4.76837158203125e-05, "ewc_loss_parallel": 6.914138793945312e-05, "grad_norm": 5.25752592086792, "learning_rate": 1e-05, "loss": 0.5403, "mean_token_accuracy": 0.8487758636474609, "num_tokens": 90461625.0, "step": 2380 }, { "epoch": 0.30288767332400457, "ewc_loss": 0.0693359375, "ewc_loss_diag": 4.76837158203125e-05, "ewc_loss_parallel": 6.437301635742188e-05, "grad_norm": 4.854196071624756, "learning_rate": 1e-05, "loss": 0.5038, "mean_token_accuracy": 0.8566598892211914, "num_tokens": 90500286.0, "step": 2381 }, { "epoch": 0.3030148836025951, "ewc_loss": 0.07470703125, "ewc_loss_diag": 4.76837158203125e-05, "ewc_loss_parallel": 6.961822509765625e-05, "grad_norm": 5.3015313148498535, "learning_rate": 1e-05, "loss": 0.5033, "mean_token_accuracy": 0.8532163500785828, "num_tokens": 90533463.0, "step": 2382 }, { "epoch": 0.3031420938811856, "ewc_loss": 0.0673828125, "ewc_loss_diag": 4.76837158203125e-05, "ewc_loss_parallel": 6.246566772460938e-05, "grad_norm": 4.743682861328125, "learning_rate": 1e-05, "loss": 0.4996, "mean_token_accuracy": 0.8568591475486755, "num_tokens": 90570536.0, "step": 2383 }, { "epoch": 0.3032693041597761, "ewc_loss": 0.076171875, "ewc_loss_diag": 4.76837158203125e-05, "ewc_loss_parallel": 7.152557373046875e-05, "grad_norm": 5.4209699630737305, "learning_rate": 1e-05, "loss": 0.578, "mean_token_accuracy": 0.8359601497650146, "num_tokens": 90603030.0, "step": 2384 }, { "epoch": 0.3033965144383666, "ewc_loss": 0.0654296875, "ewc_loss_diag": 4.76837158203125e-05, "ewc_loss_parallel": 6.079673767089844e-05, "grad_norm": 4.637969493865967, "learning_rate": 1e-05, "loss": 0.5566, "mean_token_accuracy": 0.8408053517341614, "num_tokens": 90638459.0, "step": 2385 }, { "epoch": 0.30352372471695716, "ewc_loss": 0.0791015625, "ewc_loss_diag": 4.76837158203125e-05, "ewc_loss_parallel": 7.43865966796875e-05, "grad_norm": 5.6391496658325195, "learning_rate": 1e-05, "loss": 0.5261, "mean_token_accuracy": 0.8491976261138916, "num_tokens": 90671894.0, "step": 2386 }, { "epoch": 0.30365093499554763, "ewc_loss": 0.0654296875, "ewc_loss_diag": 4.76837158203125e-05, "ewc_loss_parallel": 6.079673767089844e-05, "grad_norm": 4.542469501495361, "learning_rate": 1e-05, "loss": 0.5018, "mean_token_accuracy": 0.8577825427055359, "num_tokens": 90708583.0, "step": 2387 }, { "epoch": 0.30377814527413816, "ewc_loss": 0.0869140625, "ewc_loss_diag": 4.76837158203125e-05, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.971584320068359, "learning_rate": 1e-05, "loss": 0.5699, "mean_token_accuracy": 0.8410382270812988, "num_tokens": 90749914.0, "step": 2388 }, { "epoch": 0.3039053555527287, "ewc_loss": 0.06884765625, "ewc_loss_diag": 4.76837158203125e-05, "ewc_loss_parallel": 6.389617919921875e-05, "grad_norm": 4.625794410705566, "learning_rate": 1e-05, "loss": 0.5254, "mean_token_accuracy": 0.8483539819717407, "num_tokens": 90791955.0, "step": 2389 }, { "epoch": 0.30403256583131916, "ewc_loss": 0.0947265625, "ewc_loss_diag": 4.76837158203125e-05, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.290909767150879, "learning_rate": 1e-05, "loss": 0.4906, "mean_token_accuracy": 0.8677901029586792, "num_tokens": 90827666.0, "step": 2390 }, { "epoch": 0.3041597761099097, "ewc_loss": 0.076171875, "ewc_loss_diag": 4.76837158203125e-05, "ewc_loss_parallel": 7.152557373046875e-05, "grad_norm": 5.005838394165039, "learning_rate": 1e-05, "loss": 0.5362, "mean_token_accuracy": 0.8438049554824829, "num_tokens": 90860647.0, "step": 2391 }, { "epoch": 0.3042869863885002, "ewc_loss": 0.09521484375, "ewc_loss_diag": 4.76837158203125e-05, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.268777847290039, "learning_rate": 1e-05, "loss": 0.5897, "mean_token_accuracy": 0.8364664912223816, "num_tokens": 90899737.0, "step": 2392 }, { "epoch": 0.3044141966670907, "ewc_loss": 0.08154296875, "ewc_loss_diag": 4.792213439941406e-05, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.334640979766846, "learning_rate": 1e-05, "loss": 0.507, "mean_token_accuracy": 0.8558313846588135, "num_tokens": 90939934.0, "step": 2393 }, { "epoch": 0.3045414069456812, "ewc_loss": 0.0830078125, "ewc_loss_diag": 4.792213439941406e-05, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.940793514251709, "learning_rate": 1e-05, "loss": 0.5551, "mean_token_accuracy": 0.8423312902450562, "num_tokens": 90971538.0, "step": 2394 }, { "epoch": 0.30466861722427174, "ewc_loss": 0.0810546875, "ewc_loss_diag": 4.8160552978515625e-05, "ewc_loss_parallel": 7.62939453125e-05, "grad_norm": 5.4024763107299805, "learning_rate": 1e-05, "loss": 0.5228, "mean_token_accuracy": 0.8531834483146667, "num_tokens": 91012119.0, "step": 2395 }, { "epoch": 0.3047958275028622, "ewc_loss": 0.0771484375, "ewc_loss_diag": 4.8160552978515625e-05, "ewc_loss_parallel": 7.2479248046875e-05, "grad_norm": 5.5537638664245605, "learning_rate": 1e-05, "loss": 0.5418, "mean_token_accuracy": 0.8479310274124146, "num_tokens": 91051342.0, "step": 2396 }, { "epoch": 0.30492303778145274, "ewc_loss": 0.07861328125, "ewc_loss_diag": 4.8160552978515625e-05, "ewc_loss_parallel": 7.390975952148438e-05, "grad_norm": 5.341487884521484, "learning_rate": 1e-05, "loss": 0.5145, "mean_token_accuracy": 0.8554064631462097, "num_tokens": 91087650.0, "step": 2397 }, { "epoch": 0.30505024806004327, "ewc_loss": 0.07421875, "ewc_loss_diag": 4.8160552978515625e-05, "ewc_loss_parallel": 6.914138793945312e-05, "grad_norm": 5.23328971862793, "learning_rate": 1e-05, "loss": 0.5196, "mean_token_accuracy": 0.8520702123641968, "num_tokens": 91129786.0, "step": 2398 }, { "epoch": 0.30517745833863374, "ewc_loss": 0.076171875, "ewc_loss_diag": 4.8160552978515625e-05, "ewc_loss_parallel": 7.104873657226562e-05, "grad_norm": 5.295351982116699, "learning_rate": 1e-05, "loss": 0.5213, "mean_token_accuracy": 0.8486902117729187, "num_tokens": 91165273.0, "step": 2399 }, { "epoch": 0.3053046686172243, "ewc_loss": 0.07177734375, "ewc_loss_diag": 4.8160552978515625e-05, "ewc_loss_parallel": 6.67572021484375e-05, "grad_norm": 5.035994052886963, "learning_rate": 1e-05, "loss": 0.4968, "mean_token_accuracy": 0.8630053400993347, "num_tokens": 91198034.0, "step": 2400 }, { "epoch": 0.3054318788958148, "ewc_loss": 0.07470703125, "ewc_loss_diag": 4.8160552978515625e-05, "ewc_loss_parallel": 6.961822509765625e-05, "grad_norm": 5.2724809646606445, "learning_rate": 1e-05, "loss": 0.5177, "mean_token_accuracy": 0.8536883592605591, "num_tokens": 91241415.0, "step": 2401 }, { "epoch": 0.3055590891744053, "ewc_loss": 0.0703125, "ewc_loss_diag": 4.8160552978515625e-05, "ewc_loss_parallel": 6.532669067382812e-05, "grad_norm": 4.888904094696045, "learning_rate": 1e-05, "loss": 0.5354, "mean_token_accuracy": 0.8446561098098755, "num_tokens": 91284713.0, "step": 2402 }, { "epoch": 0.3056862994529958, "ewc_loss": 0.07470703125, "ewc_loss_diag": 4.8160552978515625e-05, "ewc_loss_parallel": 6.961822509765625e-05, "grad_norm": 5.300996780395508, "learning_rate": 1e-05, "loss": 0.5334, "mean_token_accuracy": 0.8460147380828857, "num_tokens": 91322143.0, "step": 2403 }, { "epoch": 0.30581350973158633, "ewc_loss": 0.0673828125, "ewc_loss_diag": 4.8160552978515625e-05, "ewc_loss_parallel": 6.246566772460938e-05, "grad_norm": 4.781978130340576, "learning_rate": 1e-05, "loss": 0.5448, "mean_token_accuracy": 0.8448387980461121, "num_tokens": 91366825.0, "step": 2404 }, { "epoch": 0.3059407200101768, "ewc_loss": 0.07666015625, "ewc_loss_diag": 4.8160552978515625e-05, "ewc_loss_parallel": 7.200241088867188e-05, "grad_norm": 5.405058860778809, "learning_rate": 1e-05, "loss": 0.5786, "mean_token_accuracy": 0.8429107069969177, "num_tokens": 91404469.0, "step": 2405 }, { "epoch": 0.30606793028876733, "ewc_loss": 0.06689453125, "ewc_loss_diag": 4.8160552978515625e-05, "ewc_loss_parallel": 6.198883056640625e-05, "grad_norm": 4.651154041290283, "learning_rate": 1e-05, "loss": 0.5527, "mean_token_accuracy": 0.8398934602737427, "num_tokens": 91444799.0, "step": 2406 }, { "epoch": 0.30619514056735786, "ewc_loss": 0.07958984375, "ewc_loss_diag": 4.839897155761719e-05, "ewc_loss_parallel": 7.486343383789062e-05, "grad_norm": 5.620596885681152, "learning_rate": 1e-05, "loss": 0.5428, "mean_token_accuracy": 0.8468008637428284, "num_tokens": 91483843.0, "step": 2407 }, { "epoch": 0.30632235084594833, "ewc_loss": 0.0654296875, "ewc_loss_diag": 4.839897155761719e-05, "ewc_loss_parallel": 6.079673767089844e-05, "grad_norm": 4.549701690673828, "learning_rate": 1e-05, "loss": 0.5272, "mean_token_accuracy": 0.8489655256271362, "num_tokens": 91523922.0, "step": 2408 }, { "epoch": 0.30644956112453886, "ewc_loss": 0.0859375, "ewc_loss_diag": 4.839897155761719e-05, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.967166423797607, "learning_rate": 1e-05, "loss": 0.5269, "mean_token_accuracy": 0.8534094095230103, "num_tokens": 91561155.0, "step": 2409 }, { "epoch": 0.3065767714031294, "ewc_loss": 0.06884765625, "ewc_loss_diag": 4.839897155761719e-05, "ewc_loss_parallel": 6.389617919921875e-05, "grad_norm": 4.613344192504883, "learning_rate": 1e-05, "loss": 0.5151, "mean_token_accuracy": 0.8506217002868652, "num_tokens": 91596072.0, "step": 2410 }, { "epoch": 0.30670398168171986, "ewc_loss": 0.095703125, "ewc_loss_diag": 4.863739013671875e-05, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.300248622894287, "learning_rate": 1e-05, "loss": 0.5521, "mean_token_accuracy": 0.8477863073348999, "num_tokens": 91640369.0, "step": 2411 }, { "epoch": 0.3068311919603104, "ewc_loss": 0.076171875, "ewc_loss_diag": 4.863739013671875e-05, "ewc_loss_parallel": 7.152557373046875e-05, "grad_norm": 4.9737067222595215, "learning_rate": 1e-05, "loss": 0.5219, "mean_token_accuracy": 0.8506098985671997, "num_tokens": 91674941.0, "step": 2412 }, { "epoch": 0.3069584022389009, "ewc_loss": 0.09619140625, "ewc_loss_diag": 4.887580871582031e-05, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.319573402404785, "learning_rate": 1e-05, "loss": 0.5505, "mean_token_accuracy": 0.8507125377655029, "num_tokens": 91709023.0, "step": 2413 }, { "epoch": 0.3070856125174914, "ewc_loss": 0.08203125, "ewc_loss_diag": 4.887580871582031e-05, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.344903469085693, "learning_rate": 1e-05, "loss": 0.5311, "mean_token_accuracy": 0.851373016834259, "num_tokens": 91750684.0, "step": 2414 }, { "epoch": 0.3072128227960819, "ewc_loss": 0.08642578125, "ewc_loss_diag": 4.887580871582031e-05, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.997428894042969, "learning_rate": 1e-05, "loss": 0.5554, "mean_token_accuracy": 0.8475906848907471, "num_tokens": 91789369.0, "step": 2415 }, { "epoch": 0.30734003307467245, "ewc_loss": 0.08154296875, "ewc_loss_diag": 4.887580871582031e-05, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.4311909675598145, "learning_rate": 1e-05, "loss": 0.5813, "mean_token_accuracy": 0.838624894618988, "num_tokens": 91828228.0, "step": 2416 }, { "epoch": 0.3074672433532629, "ewc_loss": 0.07861328125, "ewc_loss_diag": 4.887580871582031e-05, "ewc_loss_parallel": 7.390975952148438e-05, "grad_norm": 5.609492778778076, "learning_rate": 1e-05, "loss": 0.5046, "mean_token_accuracy": 0.8578131198883057, "num_tokens": 91866191.0, "step": 2417 }, { "epoch": 0.30759445363185345, "ewc_loss": 0.07861328125, "ewc_loss_diag": 4.887580871582031e-05, "ewc_loss_parallel": 7.390975952148438e-05, "grad_norm": 5.357443332672119, "learning_rate": 1e-05, "loss": 0.5295, "mean_token_accuracy": 0.852114200592041, "num_tokens": 91907945.0, "step": 2418 }, { "epoch": 0.307721663910444, "ewc_loss": 0.07421875, "ewc_loss_diag": 4.887580871582031e-05, "ewc_loss_parallel": 6.914138793945312e-05, "grad_norm": 5.292481899261475, "learning_rate": 1e-05, "loss": 0.4975, "mean_token_accuracy": 0.8575236797332764, "num_tokens": 91938651.0, "step": 2419 }, { "epoch": 0.30784887418903445, "ewc_loss": 0.076171875, "ewc_loss_diag": 4.887580871582031e-05, "ewc_loss_parallel": 7.104873657226562e-05, "grad_norm": 5.311532020568848, "learning_rate": 1e-05, "loss": 0.5563, "mean_token_accuracy": 0.8488048315048218, "num_tokens": 91977557.0, "step": 2420 }, { "epoch": 0.307976084467625, "ewc_loss": 0.07177734375, "ewc_loss_diag": 4.887580871582031e-05, "ewc_loss_parallel": 6.67572021484375e-05, "grad_norm": 5.0658416748046875, "learning_rate": 1e-05, "loss": 0.5496, "mean_token_accuracy": 0.8414860963821411, "num_tokens": 92012316.0, "step": 2421 }, { "epoch": 0.3081032947462155, "ewc_loss": 0.07470703125, "ewc_loss_diag": 4.9114227294921875e-05, "ewc_loss_parallel": 6.961822509765625e-05, "grad_norm": 5.291236877441406, "learning_rate": 1e-05, "loss": 0.5301, "mean_token_accuracy": 0.8478817939758301, "num_tokens": 92049262.0, "step": 2422 }, { "epoch": 0.308230505024806, "ewc_loss": 0.0703125, "ewc_loss_diag": 4.9114227294921875e-05, "ewc_loss_parallel": 6.532669067382812e-05, "grad_norm": 4.910317420959473, "learning_rate": 1e-05, "loss": 0.5568, "mean_token_accuracy": 0.8453829288482666, "num_tokens": 92086048.0, "step": 2423 }, { "epoch": 0.3083577153033965, "ewc_loss": 0.07470703125, "ewc_loss_diag": 4.9114227294921875e-05, "ewc_loss_parallel": 6.961822509765625e-05, "grad_norm": 5.305714130401611, "learning_rate": 1e-05, "loss": 0.4996, "mean_token_accuracy": 0.8581960797309875, "num_tokens": 92123540.0, "step": 2424 }, { "epoch": 0.30848492558198704, "ewc_loss": 0.06787109375, "ewc_loss_diag": 4.9114227294921875e-05, "ewc_loss_parallel": 6.29425048828125e-05, "grad_norm": 4.805456638336182, "learning_rate": 1e-05, "loss": 0.5847, "mean_token_accuracy": 0.8294589519500732, "num_tokens": 92158275.0, "step": 2425 }, { "epoch": 0.3086121358605775, "ewc_loss": 0.076171875, "ewc_loss_diag": 4.9114227294921875e-05, "ewc_loss_parallel": 7.104873657226562e-05, "grad_norm": 5.382936954498291, "learning_rate": 1e-05, "loss": 0.4783, "mean_token_accuracy": 0.8671890497207642, "num_tokens": 92196444.0, "step": 2426 }, { "epoch": 0.30873934613916804, "ewc_loss": 0.06689453125, "ewc_loss_diag": 4.9114227294921875e-05, "ewc_loss_parallel": 6.198883056640625e-05, "grad_norm": 4.67795467376709, "learning_rate": 1e-05, "loss": 0.4865, "mean_token_accuracy": 0.8580271005630493, "num_tokens": 92233988.0, "step": 2427 }, { "epoch": 0.30886655641775856, "ewc_loss": 0.0791015625, "ewc_loss_diag": 4.935264587402344e-05, "ewc_loss_parallel": 7.43865966796875e-05, "grad_norm": 5.579714775085449, "learning_rate": 1e-05, "loss": 0.5031, "mean_token_accuracy": 0.8587421178817749, "num_tokens": 92271102.0, "step": 2428 }, { "epoch": 0.3089937666963491, "ewc_loss": 0.06591796875, "ewc_loss_diag": 4.935264587402344e-05, "ewc_loss_parallel": 6.103515625e-05, "grad_norm": 4.569049835205078, "learning_rate": 1e-05, "loss": 0.5187, "mean_token_accuracy": 0.8487870693206787, "num_tokens": 92306112.0, "step": 2429 }, { "epoch": 0.30912097697493957, "ewc_loss": 0.0859375, "ewc_loss_diag": 4.935264587402344e-05, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.912718772888184, "learning_rate": 1e-05, "loss": 0.5183, "mean_token_accuracy": 0.8565124273300171, "num_tokens": 92347796.0, "step": 2430 }, { "epoch": 0.3092481872535301, "ewc_loss": 0.06787109375, "ewc_loss_diag": 4.935264587402344e-05, "ewc_loss_parallel": 6.29425048828125e-05, "grad_norm": 4.587843418121338, "learning_rate": 1e-05, "loss": 0.546, "mean_token_accuracy": 0.8440096974372864, "num_tokens": 92382018.0, "step": 2431 }, { "epoch": 0.3093753975321206, "ewc_loss": 0.0947265625, "ewc_loss_diag": 4.935264587402344e-05, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.257744789123535, "learning_rate": 1e-05, "loss": 0.5604, "mean_token_accuracy": 0.8473974466323853, "num_tokens": 92428352.0, "step": 2432 }, { "epoch": 0.3095026078107111, "ewc_loss": 0.07470703125, "ewc_loss_diag": 4.935264587402344e-05, "ewc_loss_parallel": 6.961822509765625e-05, "grad_norm": 4.9027886390686035, "learning_rate": 1e-05, "loss": 0.5364, "mean_token_accuracy": 0.8460366725921631, "num_tokens": 92462826.0, "step": 2433 }, { "epoch": 0.3096298180893016, "ewc_loss": 0.09619140625, "ewc_loss_diag": 4.935264587402344e-05, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.350412845611572, "learning_rate": 1e-05, "loss": 0.5268, "mean_token_accuracy": 0.8560006618499756, "num_tokens": 92498371.0, "step": 2434 }, { "epoch": 0.30975702836789215, "ewc_loss": 0.08203125, "ewc_loss_diag": 4.935264587402344e-05, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.313444137573242, "learning_rate": 1e-05, "loss": 0.533, "mean_token_accuracy": 0.8495597839355469, "num_tokens": 92536921.0, "step": 2435 }, { "epoch": 0.3098842386464826, "ewc_loss": 0.0869140625, "ewc_loss_diag": 4.935264587402344e-05, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 6.064405918121338, "learning_rate": 1e-05, "loss": 0.577, "mean_token_accuracy": 0.8404468297958374, "num_tokens": 92570627.0, "step": 2436 }, { "epoch": 0.31001144892507315, "ewc_loss": 0.08154296875, "ewc_loss_diag": 4.935264587402344e-05, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.433304309844971, "learning_rate": 1e-05, "loss": 0.5174, "mean_token_accuracy": 0.8544896841049194, "num_tokens": 92608683.0, "step": 2437 }, { "epoch": 0.3101386592036637, "ewc_loss": 0.0791015625, "ewc_loss_diag": 4.935264587402344e-05, "ewc_loss_parallel": 7.43865966796875e-05, "grad_norm": 5.671452045440674, "learning_rate": 1e-05, "loss": 0.5066, "mean_token_accuracy": 0.8569104671478271, "num_tokens": 92644689.0, "step": 2438 }, { "epoch": 0.31026586948225415, "ewc_loss": 0.080078125, "ewc_loss_diag": 4.935264587402344e-05, "ewc_loss_parallel": 7.534027099609375e-05, "grad_norm": 5.380402565002441, "learning_rate": 1e-05, "loss": 0.542, "mean_token_accuracy": 0.8489176034927368, "num_tokens": 92684471.0, "step": 2439 }, { "epoch": 0.3103930797608447, "ewc_loss": 0.07568359375, "ewc_loss_diag": 4.935264587402344e-05, "ewc_loss_parallel": 7.05718994140625e-05, "grad_norm": 5.339962482452393, "learning_rate": 1e-05, "loss": 0.5698, "mean_token_accuracy": 0.8407889604568481, "num_tokens": 92724991.0, "step": 2440 }, { "epoch": 0.3105202900394352, "ewc_loss": 0.076171875, "ewc_loss_diag": 4.935264587402344e-05, "ewc_loss_parallel": 7.152557373046875e-05, "grad_norm": 5.322084426879883, "learning_rate": 1e-05, "loss": 0.5365, "mean_token_accuracy": 0.8416441082954407, "num_tokens": 92759302.0, "step": 2441 }, { "epoch": 0.3106475003180257, "ewc_loss": 0.072265625, "ewc_loss_diag": 4.935264587402344e-05, "ewc_loss_parallel": 6.723403930664062e-05, "grad_norm": 5.102234840393066, "learning_rate": 1e-05, "loss": 0.5476, "mean_token_accuracy": 0.8452852368354797, "num_tokens": 92794585.0, "step": 2442 }, { "epoch": 0.3107747105966162, "ewc_loss": 0.07470703125, "ewc_loss_diag": 4.935264587402344e-05, "ewc_loss_parallel": 6.961822509765625e-05, "grad_norm": 5.313282489776611, "learning_rate": 1e-05, "loss": 0.5561, "mean_token_accuracy": 0.84235018491745, "num_tokens": 92832405.0, "step": 2443 }, { "epoch": 0.31090192087520674, "ewc_loss": 0.07080078125, "ewc_loss_diag": 4.9591064453125e-05, "ewc_loss_parallel": 6.580352783203125e-05, "grad_norm": 4.922987937927246, "learning_rate": 1e-05, "loss": 0.5503, "mean_token_accuracy": 0.8404500484466553, "num_tokens": 92877960.0, "step": 2444 }, { "epoch": 0.3110291311537972, "ewc_loss": 0.07470703125, "ewc_loss_diag": 4.9591064453125e-05, "ewc_loss_parallel": 6.961822509765625e-05, "grad_norm": 5.3256707191467285, "learning_rate": 1e-05, "loss": 0.4918, "mean_token_accuracy": 0.8625361323356628, "num_tokens": 92914421.0, "step": 2445 }, { "epoch": 0.31115634143238774, "ewc_loss": 0.068359375, "ewc_loss_diag": 4.9591064453125e-05, "ewc_loss_parallel": 6.341934204101562e-05, "grad_norm": 4.796546936035156, "learning_rate": 1e-05, "loss": 0.53, "mean_token_accuracy": 0.8470021486282349, "num_tokens": 92954844.0, "step": 2446 }, { "epoch": 0.31128355171097827, "ewc_loss": 0.076171875, "ewc_loss_diag": 4.9591064453125e-05, "ewc_loss_parallel": 7.152557373046875e-05, "grad_norm": 5.4401445388793945, "learning_rate": 1e-05, "loss": 0.5144, "mean_token_accuracy": 0.8628459572792053, "num_tokens": 92987719.0, "step": 2447 }, { "epoch": 0.31141076198956874, "ewc_loss": 0.0673828125, "ewc_loss_diag": 4.9591064453125e-05, "ewc_loss_parallel": 6.246566772460938e-05, "grad_norm": 4.674871444702148, "learning_rate": 1e-05, "loss": 0.5331, "mean_token_accuracy": 0.8486104011535645, "num_tokens": 93027335.0, "step": 2448 }, { "epoch": 0.31153797226815927, "ewc_loss": 0.0810546875, "ewc_loss_diag": 4.9591064453125e-05, "ewc_loss_parallel": 7.62939453125e-05, "grad_norm": 5.655369281768799, "learning_rate": 1e-05, "loss": 0.5424, "mean_token_accuracy": 0.8495949506759644, "num_tokens": 93063325.0, "step": 2449 }, { "epoch": 0.3116651825467498, "ewc_loss": 0.06689453125, "ewc_loss_diag": 4.982948303222656e-05, "ewc_loss_parallel": 6.198883056640625e-05, "grad_norm": 4.576992034912109, "learning_rate": 1e-05, "loss": 0.5405, "mean_token_accuracy": 0.8404052257537842, "num_tokens": 93104057.0, "step": 2450 }, { "epoch": 0.31179239282534027, "ewc_loss": 0.08740234375, "ewc_loss_diag": 5.030632019042969e-05, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 6.005488872528076, "learning_rate": 1e-05, "loss": 0.5656, "mean_token_accuracy": 0.8443111181259155, "num_tokens": 93138912.0, "step": 2451 }, { "epoch": 0.3119196031039308, "ewc_loss": 0.06982421875, "ewc_loss_diag": 5.030632019042969e-05, "ewc_loss_parallel": 6.4849853515625e-05, "grad_norm": 4.650618553161621, "learning_rate": 1e-05, "loss": 0.5816, "mean_token_accuracy": 0.8385390043258667, "num_tokens": 93178658.0, "step": 2452 }, { "epoch": 0.3120468133825213, "ewc_loss": 0.09619140625, "ewc_loss_diag": 5.030632019042969e-05, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.325953960418701, "learning_rate": 1e-05, "loss": 0.5546, "mean_token_accuracy": 0.8508225679397583, "num_tokens": 93217723.0, "step": 2453 }, { "epoch": 0.3121740236611118, "ewc_loss": 0.078125, "ewc_loss_diag": 5.030632019042969e-05, "ewc_loss_parallel": 7.343292236328125e-05, "grad_norm": 5.01922082901001, "learning_rate": 1e-05, "loss": 0.5133, "mean_token_accuracy": 0.8566406965255737, "num_tokens": 93253917.0, "step": 2454 }, { "epoch": 0.3123012339397023, "ewc_loss": 0.0966796875, "ewc_loss_diag": 5.030632019042969e-05, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.319284439086914, "learning_rate": 1e-05, "loss": 0.5416, "mean_token_accuracy": 0.8580692410469055, "num_tokens": 93289735.0, "step": 2455 }, { "epoch": 0.31242844421829286, "ewc_loss": 0.08447265625, "ewc_loss_diag": 5.030632019042969e-05, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.3843207359313965, "learning_rate": 1e-05, "loss": 0.5406, "mean_token_accuracy": 0.8491498231887817, "num_tokens": 93331549.0, "step": 2456 }, { "epoch": 0.31255565449688333, "ewc_loss": 0.0849609375, "ewc_loss_diag": 5.030632019042969e-05, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.974950790405273, "learning_rate": 1e-05, "loss": 0.5854, "mean_token_accuracy": 0.8361356854438782, "num_tokens": 93372977.0, "step": 2457 }, { "epoch": 0.31268286477547386, "ewc_loss": 0.0830078125, "ewc_loss_diag": 5.030632019042969e-05, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.439734935760498, "learning_rate": 1e-05, "loss": 0.5023, "mean_token_accuracy": 0.8614873290061951, "num_tokens": 93406842.0, "step": 2458 }, { "epoch": 0.3128100750540644, "ewc_loss": 0.07763671875, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 7.295608520507812e-05, "grad_norm": 5.590152740478516, "learning_rate": 1e-05, "loss": 0.5289, "mean_token_accuracy": 0.8518223762512207, "num_tokens": 93444113.0, "step": 2459 }, { "epoch": 0.31293728533265486, "ewc_loss": 0.07861328125, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 7.390975952148438e-05, "grad_norm": 5.359463691711426, "learning_rate": 1e-05, "loss": 0.5703, "mean_token_accuracy": 0.8391880393028259, "num_tokens": 93490618.0, "step": 2460 }, { "epoch": 0.3130644956112454, "ewc_loss": 0.07470703125, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 6.961822509765625e-05, "grad_norm": 5.283951759338379, "learning_rate": 1e-05, "loss": 0.5483, "mean_token_accuracy": 0.842963457107544, "num_tokens": 93524711.0, "step": 2461 }, { "epoch": 0.3131917058898359, "ewc_loss": 0.076171875, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 7.104873657226562e-05, "grad_norm": 5.307863712310791, "learning_rate": 1e-05, "loss": 0.5284, "mean_token_accuracy": 0.8489867448806763, "num_tokens": 93559169.0, "step": 2462 }, { "epoch": 0.3133189161684264, "ewc_loss": 0.07275390625, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 6.771087646484375e-05, "grad_norm": 5.067580699920654, "learning_rate": 1e-05, "loss": 0.5297, "mean_token_accuracy": 0.8517110347747803, "num_tokens": 93598941.0, "step": 2463 }, { "epoch": 0.3134461264470169, "ewc_loss": 0.07470703125, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 6.961822509765625e-05, "grad_norm": 5.274325370788574, "learning_rate": 1e-05, "loss": 0.5114, "mean_token_accuracy": 0.8535035848617554, "num_tokens": 93634979.0, "step": 2464 }, { "epoch": 0.31357333672560744, "ewc_loss": 0.07080078125, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 6.580352783203125e-05, "grad_norm": 4.939634799957275, "learning_rate": 1e-05, "loss": 0.5814, "mean_token_accuracy": 0.8281517028808594, "num_tokens": 93676736.0, "step": 2465 }, { "epoch": 0.3137005470041979, "ewc_loss": 0.07470703125, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 6.961822509765625e-05, "grad_norm": 5.278015613555908, "learning_rate": 1e-05, "loss": 0.5007, "mean_token_accuracy": 0.8615081906318665, "num_tokens": 93720118.0, "step": 2466 }, { "epoch": 0.31382775728278844, "ewc_loss": 0.06884765625, "ewc_loss_diag": 5.078315734863281e-05, "ewc_loss_parallel": 6.389617919921875e-05, "grad_norm": 4.8217549324035645, "learning_rate": 1e-05, "loss": 0.5054, "mean_token_accuracy": 0.8558952808380127, "num_tokens": 93760814.0, "step": 2467 }, { "epoch": 0.313954967561379, "ewc_loss": 0.07568359375, "ewc_loss_diag": 5.078315734863281e-05, "ewc_loss_parallel": 7.05718994140625e-05, "grad_norm": 5.363522529602051, "learning_rate": 1e-05, "loss": 0.4658, "mean_token_accuracy": 0.8664995431900024, "num_tokens": 93795326.0, "step": 2468 }, { "epoch": 0.31408217783996945, "ewc_loss": 0.06689453125, "ewc_loss_diag": 5.078315734863281e-05, "ewc_loss_parallel": 6.198883056640625e-05, "grad_norm": 4.701047897338867, "learning_rate": 1e-05, "loss": 0.5126, "mean_token_accuracy": 0.8513389825820923, "num_tokens": 93832993.0, "step": 2469 }, { "epoch": 0.31420938811856, "ewc_loss": 0.07861328125, "ewc_loss_diag": 5.078315734863281e-05, "ewc_loss_parallel": 7.390975952148438e-05, "grad_norm": 5.554156303405762, "learning_rate": 1e-05, "loss": 0.4991, "mean_token_accuracy": 0.8628860116004944, "num_tokens": 93871282.0, "step": 2470 }, { "epoch": 0.3143365983971505, "ewc_loss": 0.06591796875, "ewc_loss_diag": 5.078315734863281e-05, "ewc_loss_parallel": 6.103515625e-05, "grad_norm": 4.5809125900268555, "learning_rate": 1e-05, "loss": 0.539, "mean_token_accuracy": 0.8461674451828003, "num_tokens": 93912698.0, "step": 2471 }, { "epoch": 0.314463808675741, "ewc_loss": 0.083984375, "ewc_loss_diag": 5.078315734863281e-05, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.8697614669799805, "learning_rate": 1e-05, "loss": 0.5243, "mean_token_accuracy": 0.85328608751297, "num_tokens": 93952499.0, "step": 2472 }, { "epoch": 0.3145910189543315, "ewc_loss": 0.0673828125, "ewc_loss_diag": 5.1021575927734375e-05, "ewc_loss_parallel": 6.246566772460938e-05, "grad_norm": 4.560880184173584, "learning_rate": 1e-05, "loss": 0.5633, "mean_token_accuracy": 0.8364236354827881, "num_tokens": 93992277.0, "step": 2473 }, { "epoch": 0.31471822923292203, "ewc_loss": 0.0947265625, "ewc_loss_diag": 5.14984130859375e-05, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 6.273946285247803, "learning_rate": 1e-05, "loss": 0.5308, "mean_token_accuracy": 0.8528040647506714, "num_tokens": 94032012.0, "step": 2474 }, { "epoch": 0.3148454395115125, "ewc_loss": 0.07421875, "ewc_loss_diag": 5.125999450683594e-05, "ewc_loss_parallel": 6.914138793945312e-05, "grad_norm": 4.840084552764893, "learning_rate": 1e-05, "loss": 0.4978, "mean_token_accuracy": 0.8581925630569458, "num_tokens": 94072210.0, "step": 2475 }, { "epoch": 0.31497264979010303, "ewc_loss": 0.09912109375, "ewc_loss_diag": 5.14984130859375e-05, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.464406490325928, "learning_rate": 1e-05, "loss": 0.5715, "mean_token_accuracy": 0.8471136093139648, "num_tokens": 94112269.0, "step": 2476 }, { "epoch": 0.31509986006869356, "ewc_loss": 0.08349609375, "ewc_loss_diag": 5.14984130859375e-05, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.341251850128174, "learning_rate": 1e-05, "loss": 0.6237, "mean_token_accuracy": 0.8253786563873291, "num_tokens": 94153596.0, "step": 2477 }, { "epoch": 0.31522707034728403, "ewc_loss": 0.09033203125, "ewc_loss_diag": 5.14984130859375e-05, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 6.188189506530762, "learning_rate": 1e-05, "loss": 0.5742, "mean_token_accuracy": 0.843644917011261, "num_tokens": 94195146.0, "step": 2478 }, { "epoch": 0.31535428062587456, "ewc_loss": 0.08544921875, "ewc_loss_diag": 5.14984130859375e-05, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.5031585693359375, "learning_rate": 1e-05, "loss": 0.5053, "mean_token_accuracy": 0.8594023585319519, "num_tokens": 94233196.0, "step": 2479 }, { "epoch": 0.3154814909044651, "ewc_loss": 0.08056640625, "ewc_loss_diag": 5.14984130859375e-05, "ewc_loss_parallel": 7.534027099609375e-05, "grad_norm": 5.781061172485352, "learning_rate": 1e-05, "loss": 0.56, "mean_token_accuracy": 0.8462789058685303, "num_tokens": 94275314.0, "step": 2480 }, { "epoch": 0.3156087011830556, "ewc_loss": 0.08154296875, "ewc_loss_diag": 5.14984130859375e-05, "ewc_loss_parallel": 7.62939453125e-05, "grad_norm": 5.425349235534668, "learning_rate": 1e-05, "loss": 0.5363, "mean_token_accuracy": 0.852332353591919, "num_tokens": 94316673.0, "step": 2481 }, { "epoch": 0.3157359114616461, "ewc_loss": 0.07568359375, "ewc_loss_diag": 5.14984130859375e-05, "ewc_loss_parallel": 7.009506225585938e-05, "grad_norm": 5.444436550140381, "learning_rate": 1e-05, "loss": 0.5127, "mean_token_accuracy": 0.8541953563690186, "num_tokens": 94354118.0, "step": 2482 }, { "epoch": 0.3158631217402366, "ewc_loss": 0.078125, "ewc_loss_diag": 5.14984130859375e-05, "ewc_loss_parallel": 7.295608520507812e-05, "grad_norm": 5.338117599487305, "learning_rate": 1e-05, "loss": 0.5363, "mean_token_accuracy": 0.8445309400558472, "num_tokens": 94386476.0, "step": 2483 }, { "epoch": 0.31599033201882715, "ewc_loss": 0.0732421875, "ewc_loss_diag": 5.14984130859375e-05, "ewc_loss_parallel": 6.771087646484375e-05, "grad_norm": 5.206414222717285, "learning_rate": 1e-05, "loss": 0.5284, "mean_token_accuracy": 0.8487737774848938, "num_tokens": 94420146.0, "step": 2484 }, { "epoch": 0.3161175422974176, "ewc_loss": 0.0751953125, "ewc_loss_diag": 5.14984130859375e-05, "ewc_loss_parallel": 6.961822509765625e-05, "grad_norm": 5.272863864898682, "learning_rate": 1e-05, "loss": 0.5431, "mean_token_accuracy": 0.8466132879257202, "num_tokens": 94455057.0, "step": 2485 }, { "epoch": 0.31624475257600815, "ewc_loss": 0.072265625, "ewc_loss_diag": 5.14984130859375e-05, "ewc_loss_parallel": 6.67572021484375e-05, "grad_norm": 5.0317182540893555, "learning_rate": 1e-05, "loss": 0.5002, "mean_token_accuracy": 0.8579171895980835, "num_tokens": 94500666.0, "step": 2486 }, { "epoch": 0.3163719628545987, "ewc_loss": 0.0751953125, "ewc_loss_diag": 5.14984130859375e-05, "ewc_loss_parallel": 6.961822509765625e-05, "grad_norm": 5.24271821975708, "learning_rate": 1e-05, "loss": 0.4786, "mean_token_accuracy": 0.8629201054573059, "num_tokens": 94537623.0, "step": 2487 }, { "epoch": 0.31649917313318915, "ewc_loss": 0.0712890625, "ewc_loss_diag": 5.173683166503906e-05, "ewc_loss_parallel": 6.580352783203125e-05, "grad_norm": 4.9217352867126465, "learning_rate": 1e-05, "loss": 0.4781, "mean_token_accuracy": 0.8621480464935303, "num_tokens": 94568563.0, "step": 2488 }, { "epoch": 0.3166263834117797, "ewc_loss": 0.0751953125, "ewc_loss_diag": 5.173683166503906e-05, "ewc_loss_parallel": 6.961822509765625e-05, "grad_norm": 5.267302513122559, "learning_rate": 1e-05, "loss": 0.4868, "mean_token_accuracy": 0.8634495735168457, "num_tokens": 94600931.0, "step": 2489 }, { "epoch": 0.3167535936903702, "ewc_loss": 0.068359375, "ewc_loss_diag": 5.173683166503906e-05, "ewc_loss_parallel": 6.29425048828125e-05, "grad_norm": 4.816693305969238, "learning_rate": 1e-05, "loss": 0.5854, "mean_token_accuracy": 0.8314141035079956, "num_tokens": 94634345.0, "step": 2490 }, { "epoch": 0.3168808039689607, "ewc_loss": 0.07666015625, "ewc_loss_diag": 5.173683166503906e-05, "ewc_loss_parallel": 7.104873657226562e-05, "grad_norm": 5.3458967208862305, "learning_rate": 1e-05, "loss": 0.5002, "mean_token_accuracy": 0.8632333874702454, "num_tokens": 94673157.0, "step": 2491 }, { "epoch": 0.3170080142475512, "ewc_loss": 0.0673828125, "ewc_loss_diag": 5.173683166503906e-05, "ewc_loss_parallel": 6.198883056640625e-05, "grad_norm": 4.687458515167236, "learning_rate": 1e-05, "loss": 0.5605, "mean_token_accuracy": 0.8351129293441772, "num_tokens": 94712727.0, "step": 2492 }, { "epoch": 0.31713522452614173, "ewc_loss": 0.07861328125, "ewc_loss_diag": 5.173683166503906e-05, "ewc_loss_parallel": 7.343292236328125e-05, "grad_norm": 5.561357021331787, "learning_rate": 1e-05, "loss": 0.6022, "mean_token_accuracy": 0.8309420943260193, "num_tokens": 94752341.0, "step": 2493 }, { "epoch": 0.3172624348047322, "ewc_loss": 0.06640625, "ewc_loss_diag": 5.173683166503906e-05, "ewc_loss_parallel": 6.103515625e-05, "grad_norm": 4.549551010131836, "learning_rate": 1e-05, "loss": 0.548, "mean_token_accuracy": 0.8421785831451416, "num_tokens": 94791287.0, "step": 2494 }, { "epoch": 0.31738964508332274, "ewc_loss": 0.0869140625, "ewc_loss_diag": 5.173683166503906e-05, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.926205158233643, "learning_rate": 1e-05, "loss": 0.5516, "mean_token_accuracy": 0.8448620438575745, "num_tokens": 94833029.0, "step": 2495 }, { "epoch": 0.31751685536191326, "ewc_loss": 0.06787109375, "ewc_loss_diag": 5.173683166503906e-05, "ewc_loss_parallel": 6.246566772460938e-05, "grad_norm": 4.532995700836182, "learning_rate": 1e-05, "loss": 0.5252, "mean_token_accuracy": 0.8485445976257324, "num_tokens": 94873400.0, "step": 2496 }, { "epoch": 0.31764406564050374, "ewc_loss": 0.09716796875, "ewc_loss_diag": 5.173683166503906e-05, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.3954315185546875, "learning_rate": 1e-05, "loss": 0.6001, "mean_token_accuracy": 0.8321813344955444, "num_tokens": 94914603.0, "step": 2497 }, { "epoch": 0.31777127591909426, "ewc_loss": 0.076171875, "ewc_loss_diag": 5.1975250244140625e-05, "ewc_loss_parallel": 7.05718994140625e-05, "grad_norm": 4.904883861541748, "learning_rate": 1e-05, "loss": 0.47, "mean_token_accuracy": 0.8706760406494141, "num_tokens": 94955925.0, "step": 2498 }, { "epoch": 0.3178984861976848, "ewc_loss": 0.10107421875, "ewc_loss_diag": 5.1975250244140625e-05, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.575725078582764, "learning_rate": 1e-05, "loss": 0.4985, "mean_token_accuracy": 0.8687804341316223, "num_tokens": 94989420.0, "step": 2499 }, { "epoch": 0.31802569647627527, "ewc_loss": 0.08740234375, "ewc_loss_diag": 5.1975250244140625e-05, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.481822967529297, "learning_rate": 1e-05, "loss": 0.567, "mean_token_accuracy": 0.8434560298919678, "num_tokens": 95028835.0, "step": 2500 }, { "epoch": 0.3181529067548658, "ewc_loss": 0.09130859375, "ewc_loss_diag": 5.1975250244140625e-05, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 6.213146686553955, "learning_rate": 1e-05, "loss": 0.6158, "mean_token_accuracy": 0.8361898064613342, "num_tokens": 95069246.0, "step": 2501 }, { "epoch": 0.3182801170334563, "ewc_loss": 0.0869140625, "ewc_loss_diag": 5.1975250244140625e-05, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.610269546508789, "learning_rate": 1e-05, "loss": 0.6536, "mean_token_accuracy": 0.8161661624908447, "num_tokens": 95110184.0, "step": 2502 }, { "epoch": 0.3184073273120468, "ewc_loss": 0.08203125, "ewc_loss_diag": 5.1975250244140625e-05, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.786946773529053, "learning_rate": 1e-05, "loss": 0.5338, "mean_token_accuracy": 0.854925274848938, "num_tokens": 95149118.0, "step": 2503 }, { "epoch": 0.3185345375906373, "ewc_loss": 0.08203125, "ewc_loss_diag": 5.1975250244140625e-05, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.463169097900391, "learning_rate": 1e-05, "loss": 0.5297, "mean_token_accuracy": 0.8518670201301575, "num_tokens": 95195605.0, "step": 2504 }, { "epoch": 0.31866174786922785, "ewc_loss": 0.076171875, "ewc_loss_diag": 5.1975250244140625e-05, "ewc_loss_parallel": 7.05718994140625e-05, "grad_norm": 5.4628005027771, "learning_rate": 1e-05, "loss": 0.5482, "mean_token_accuracy": 0.8423556089401245, "num_tokens": 95234602.0, "step": 2505 }, { "epoch": 0.3187889581478183, "ewc_loss": 0.07763671875, "ewc_loss_diag": 5.221366882324219e-05, "ewc_loss_parallel": 7.2479248046875e-05, "grad_norm": 5.33193302154541, "learning_rate": 1e-05, "loss": 0.5085, "mean_token_accuracy": 0.8581241369247437, "num_tokens": 95275646.0, "step": 2506 }, { "epoch": 0.31891616842640885, "ewc_loss": 0.07421875, "ewc_loss_diag": 5.221366882324219e-05, "ewc_loss_parallel": 6.866455078125e-05, "grad_norm": 5.253137111663818, "learning_rate": 1e-05, "loss": 0.5314, "mean_token_accuracy": 0.8469942212104797, "num_tokens": 95309948.0, "step": 2507 }, { "epoch": 0.3190433787049994, "ewc_loss": 0.07470703125, "ewc_loss_diag": 5.221366882324219e-05, "ewc_loss_parallel": 6.914138793945312e-05, "grad_norm": 5.239895343780518, "learning_rate": 1e-05, "loss": 0.5398, "mean_token_accuracy": 0.8502564430236816, "num_tokens": 95347506.0, "step": 2508 }, { "epoch": 0.31917058898358985, "ewc_loss": 0.0732421875, "ewc_loss_diag": 5.221366882324219e-05, "ewc_loss_parallel": 6.771087646484375e-05, "grad_norm": 5.115141868591309, "learning_rate": 1e-05, "loss": 0.532, "mean_token_accuracy": 0.848648190498352, "num_tokens": 95386809.0, "step": 2509 }, { "epoch": 0.3192977992621804, "ewc_loss": 0.07373046875, "ewc_loss_diag": 5.221366882324219e-05, "ewc_loss_parallel": 6.818771362304688e-05, "grad_norm": 5.176584720611572, "learning_rate": 1e-05, "loss": 0.5884, "mean_token_accuracy": 0.8429955244064331, "num_tokens": 95429287.0, "step": 2510 }, { "epoch": 0.3194250095407709, "ewc_loss": 0.072265625, "ewc_loss_diag": 5.221366882324219e-05, "ewc_loss_parallel": 6.67572021484375e-05, "grad_norm": 5.008431911468506, "learning_rate": 1e-05, "loss": 0.4516, "mean_token_accuracy": 0.8718197345733643, "num_tokens": 95466443.0, "step": 2511 }, { "epoch": 0.3195522198193614, "ewc_loss": 0.07275390625, "ewc_loss_diag": 5.221366882324219e-05, "ewc_loss_parallel": 6.723403930664062e-05, "grad_norm": 5.148586750030518, "learning_rate": 1e-05, "loss": 0.506, "mean_token_accuracy": 0.8560173511505127, "num_tokens": 95511593.0, "step": 2512 }, { "epoch": 0.3196794300979519, "ewc_loss": 0.07080078125, "ewc_loss_diag": 5.221366882324219e-05, "ewc_loss_parallel": 6.532669067382812e-05, "grad_norm": 4.936550617218018, "learning_rate": 1e-05, "loss": 0.5109, "mean_token_accuracy": 0.8505355715751648, "num_tokens": 95546756.0, "step": 2513 }, { "epoch": 0.31980664037654244, "ewc_loss": 0.07275390625, "ewc_loss_diag": 5.221366882324219e-05, "ewc_loss_parallel": 6.723403930664062e-05, "grad_norm": 5.15773344039917, "learning_rate": 1e-05, "loss": 0.5281, "mean_token_accuracy": 0.8494811654090881, "num_tokens": 95586434.0, "step": 2514 }, { "epoch": 0.3199338506551329, "ewc_loss": 0.0693359375, "ewc_loss_diag": 5.221366882324219e-05, "ewc_loss_parallel": 6.389617919921875e-05, "grad_norm": 4.8582658767700195, "learning_rate": 1e-05, "loss": 0.4683, "mean_token_accuracy": 0.8681081533432007, "num_tokens": 95616980.0, "step": 2515 }, { "epoch": 0.32006106093372344, "ewc_loss": 0.0751953125, "ewc_loss_diag": 5.221366882324219e-05, "ewc_loss_parallel": 6.961822509765625e-05, "grad_norm": 5.222420692443848, "learning_rate": 1e-05, "loss": 0.5311, "mean_token_accuracy": 0.8514854907989502, "num_tokens": 95656919.0, "step": 2516 }, { "epoch": 0.32018827121231397, "ewc_loss": 0.068359375, "ewc_loss_diag": 5.221366882324219e-05, "ewc_loss_parallel": 6.29425048828125e-05, "grad_norm": 4.754428386688232, "learning_rate": 1e-05, "loss": 0.5243, "mean_token_accuracy": 0.8523475527763367, "num_tokens": 95691785.0, "step": 2517 }, { "epoch": 0.32031548149090444, "ewc_loss": 0.07666015625, "ewc_loss_diag": 5.245208740234375e-05, "ewc_loss_parallel": 7.104873657226562e-05, "grad_norm": 5.367463111877441, "learning_rate": 1e-05, "loss": 0.535, "mean_token_accuracy": 0.8473567366600037, "num_tokens": 95731055.0, "step": 2518 }, { "epoch": 0.32044269176949497, "ewc_loss": 0.06591796875, "ewc_loss_diag": 5.245208740234375e-05, "ewc_loss_parallel": 6.079673767089844e-05, "grad_norm": 4.596426010131836, "learning_rate": 1e-05, "loss": 0.461, "mean_token_accuracy": 0.8702911138534546, "num_tokens": 95767609.0, "step": 2519 }, { "epoch": 0.3205699020480855, "ewc_loss": 0.0810546875, "ewc_loss_diag": 5.245208740234375e-05, "ewc_loss_parallel": 7.581710815429688e-05, "grad_norm": 5.665459632873535, "learning_rate": 1e-05, "loss": 0.4931, "mean_token_accuracy": 0.8603286147117615, "num_tokens": 95805156.0, "step": 2520 }, { "epoch": 0.32069711232667597, "ewc_loss": 0.06494140625, "ewc_loss_diag": 5.245208740234375e-05, "ewc_loss_parallel": 5.9604644775390625e-05, "grad_norm": 4.44210147857666, "learning_rate": 1e-05, "loss": 0.4835, "mean_token_accuracy": 0.8604296445846558, "num_tokens": 95848809.0, "step": 2521 }, { "epoch": 0.3208243226052665, "ewc_loss": 0.091796875, "ewc_loss_diag": 5.245208740234375e-05, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 6.206310272216797, "learning_rate": 1e-05, "loss": 0.5559, "mean_token_accuracy": 0.8445389270782471, "num_tokens": 95882961.0, "step": 2522 }, { "epoch": 0.320951532883857, "ewc_loss": 0.0703125, "ewc_loss_diag": 5.245208740234375e-05, "ewc_loss_parallel": 6.4849853515625e-05, "grad_norm": 4.561683177947998, "learning_rate": 1e-05, "loss": 0.4766, "mean_token_accuracy": 0.862392008304596, "num_tokens": 95916360.0, "step": 2523 }, { "epoch": 0.3210787431624475, "ewc_loss": 0.1064453125, "ewc_loss_diag": 5.245208740234375e-05, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.769919395446777, "learning_rate": 1e-05, "loss": 0.5487, "mean_token_accuracy": 0.8522887229919434, "num_tokens": 95952064.0, "step": 2524 }, { "epoch": 0.32120595344103803, "ewc_loss": 0.08447265625, "ewc_loss_diag": 5.245208740234375e-05, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.2840423583984375, "learning_rate": 1e-05, "loss": 0.536, "mean_token_accuracy": 0.8497503399848938, "num_tokens": 95993073.0, "step": 2525 }, { "epoch": 0.32133316371962856, "ewc_loss": 0.1044921875, "ewc_loss_diag": 5.245208740234375e-05, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.673623085021973, "learning_rate": 1e-05, "loss": 0.5856, "mean_token_accuracy": 0.8392099738121033, "num_tokens": 96028659.0, "step": 2526 }, { "epoch": 0.32146037399821903, "ewc_loss": 0.095703125, "ewc_loss_diag": 5.245208740234375e-05, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 5.810467720031738, "learning_rate": 1e-05, "loss": 0.5949, "mean_token_accuracy": 0.8447885513305664, "num_tokens": 96064802.0, "step": 2527 }, { "epoch": 0.32158758427680956, "ewc_loss": 0.08837890625, "ewc_loss_diag": 5.245208740234375e-05, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 6.152336120605469, "learning_rate": 1e-05, "loss": 0.5696, "mean_token_accuracy": 0.8425657749176025, "num_tokens": 96112980.0, "step": 2528 }, { "epoch": 0.3217147945554001, "ewc_loss": 0.0869140625, "ewc_loss_diag": 5.245208740234375e-05, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.684244632720947, "learning_rate": 1e-05, "loss": 0.5709, "mean_token_accuracy": 0.8397992849349976, "num_tokens": 96144114.0, "step": 2529 }, { "epoch": 0.3218420048339906, "ewc_loss": 0.0830078125, "ewc_loss_diag": 5.245208740234375e-05, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.800882339477539, "learning_rate": 1e-05, "loss": 0.5328, "mean_token_accuracy": 0.8542878031730652, "num_tokens": 96177383.0, "step": 2530 }, { "epoch": 0.3219692151125811, "ewc_loss": 0.0810546875, "ewc_loss_diag": 5.245208740234375e-05, "ewc_loss_parallel": 7.581710815429688e-05, "grad_norm": 5.454827308654785, "learning_rate": 1e-05, "loss": 0.5314, "mean_token_accuracy": 0.8542600870132446, "num_tokens": 96216513.0, "step": 2531 }, { "epoch": 0.3220964253911716, "ewc_loss": 0.07763671875, "ewc_loss_diag": 5.245208740234375e-05, "ewc_loss_parallel": 7.2479248046875e-05, "grad_norm": 5.558335781097412, "learning_rate": 1e-05, "loss": 0.5273, "mean_token_accuracy": 0.857120156288147, "num_tokens": 96256588.0, "step": 2532 }, { "epoch": 0.32222363566976214, "ewc_loss": 0.076171875, "ewc_loss_diag": 5.245208740234375e-05, "ewc_loss_parallel": 7.05718994140625e-05, "grad_norm": 5.2559733390808105, "learning_rate": 1e-05, "loss": 0.528, "mean_token_accuracy": 0.8487738370895386, "num_tokens": 96295365.0, "step": 2533 }, { "epoch": 0.3223508459483526, "ewc_loss": 0.0771484375, "ewc_loss_diag": 5.269050598144531e-05, "ewc_loss_parallel": 7.200241088867188e-05, "grad_norm": 5.436649799346924, "learning_rate": 1e-05, "loss": 0.561, "mean_token_accuracy": 0.8460871577262878, "num_tokens": 96328048.0, "step": 2534 }, { "epoch": 0.32247805622694314, "ewc_loss": 0.0732421875, "ewc_loss_diag": 5.269050598144531e-05, "ewc_loss_parallel": 6.771087646484375e-05, "grad_norm": 5.0800652503967285, "learning_rate": 1e-05, "loss": 0.5206, "mean_token_accuracy": 0.8525969386100769, "num_tokens": 96362088.0, "step": 2535 }, { "epoch": 0.32260526650553367, "ewc_loss": 0.07666015625, "ewc_loss_diag": 5.269050598144531e-05, "ewc_loss_parallel": 7.104873657226562e-05, "grad_norm": 5.419621467590332, "learning_rate": 1e-05, "loss": 0.5565, "mean_token_accuracy": 0.8463535308837891, "num_tokens": 96396161.0, "step": 2536 }, { "epoch": 0.32273247678412414, "ewc_loss": 0.07080078125, "ewc_loss_diag": 5.269050598144531e-05, "ewc_loss_parallel": 6.532669067382812e-05, "grad_norm": 4.9109673500061035, "learning_rate": 1e-05, "loss": 0.5194, "mean_token_accuracy": 0.8514172434806824, "num_tokens": 96433402.0, "step": 2537 }, { "epoch": 0.3228596870627147, "ewc_loss": 0.07763671875, "ewc_loss_diag": 5.269050598144531e-05, "ewc_loss_parallel": 7.2479248046875e-05, "grad_norm": 5.498176574707031, "learning_rate": 1e-05, "loss": 0.5408, "mean_token_accuracy": 0.8435862064361572, "num_tokens": 96473011.0, "step": 2538 }, { "epoch": 0.3229868973413052, "ewc_loss": 0.068359375, "ewc_loss_diag": 5.269050598144531e-05, "ewc_loss_parallel": 6.29425048828125e-05, "grad_norm": 4.755317211151123, "learning_rate": 1e-05, "loss": 0.4959, "mean_token_accuracy": 0.8577325344085693, "num_tokens": 96511595.0, "step": 2539 }, { "epoch": 0.3231141076198957, "ewc_loss": 0.08154296875, "ewc_loss_diag": 5.269050598144531e-05, "ewc_loss_parallel": 7.62939453125e-05, "grad_norm": 5.7527570724487305, "learning_rate": 1e-05, "loss": 0.5436, "mean_token_accuracy": 0.8459911346435547, "num_tokens": 96546192.0, "step": 2540 }, { "epoch": 0.3232413178984862, "ewc_loss": 0.06787109375, "ewc_loss_diag": 5.269050598144531e-05, "ewc_loss_parallel": 6.246566772460938e-05, "grad_norm": 4.639413833618164, "learning_rate": 1e-05, "loss": 0.4337, "mean_token_accuracy": 0.876862645149231, "num_tokens": 96582613.0, "step": 2541 }, { "epoch": 0.32336852817707673, "ewc_loss": 0.0908203125, "ewc_loss_diag": 5.269050598144531e-05, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 6.134662628173828, "learning_rate": 1e-05, "loss": 0.5607, "mean_token_accuracy": 0.8451488018035889, "num_tokens": 96620219.0, "step": 2542 }, { "epoch": 0.3234957384556672, "ewc_loss": 0.07177734375, "ewc_loss_diag": 5.269050598144531e-05, "ewc_loss_parallel": 6.628036499023438e-05, "grad_norm": 4.804102897644043, "learning_rate": 1e-05, "loss": 0.5457, "mean_token_accuracy": 0.8458028435707092, "num_tokens": 96661126.0, "step": 2543 }, { "epoch": 0.32362294873425773, "ewc_loss": 0.0966796875, "ewc_loss_diag": 5.269050598144531e-05, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.394965171813965, "learning_rate": 1e-05, "loss": 0.5613, "mean_token_accuracy": 0.8436775207519531, "num_tokens": 96697995.0, "step": 2544 }, { "epoch": 0.32375015901284826, "ewc_loss": 0.07958984375, "ewc_loss_diag": 5.269050598144531e-05, "ewc_loss_parallel": 7.43865966796875e-05, "grad_norm": 5.264489650726318, "learning_rate": 1e-05, "loss": 0.5226, "mean_token_accuracy": 0.8537845611572266, "num_tokens": 96732542.0, "step": 2545 }, { "epoch": 0.32387736929143873, "ewc_loss": 0.08984375, "ewc_loss_diag": 5.2928924560546875e-05, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 6.188952922821045, "learning_rate": 1e-05, "loss": 0.5781, "mean_token_accuracy": 0.83909010887146, "num_tokens": 96761787.0, "step": 2546 }, { "epoch": 0.32400457957002926, "ewc_loss": 0.0849609375, "ewc_loss_diag": 5.2928924560546875e-05, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.529621601104736, "learning_rate": 1e-05, "loss": 0.551, "mean_token_accuracy": 0.8467164635658264, "num_tokens": 96804918.0, "step": 2547 }, { "epoch": 0.3241317898486198, "ewc_loss": 0.080078125, "ewc_loss_diag": 5.2928924560546875e-05, "ewc_loss_parallel": 7.486343383789062e-05, "grad_norm": 5.700824737548828, "learning_rate": 1e-05, "loss": 0.5193, "mean_token_accuracy": 0.8519855737686157, "num_tokens": 96843646.0, "step": 2548 }, { "epoch": 0.32425900012721026, "ewc_loss": 0.08203125, "ewc_loss_diag": 5.2928924560546875e-05, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.548648834228516, "learning_rate": 1e-05, "loss": 0.5961, "mean_token_accuracy": 0.8305251002311707, "num_tokens": 96874796.0, "step": 2549 }, { "epoch": 0.3243862104058008, "ewc_loss": 0.07470703125, "ewc_loss_diag": 5.2928924560546875e-05, "ewc_loss_parallel": 6.914138793945312e-05, "grad_norm": 5.226119041442871, "learning_rate": 1e-05, "loss": 0.5067, "mean_token_accuracy": 0.8573915958404541, "num_tokens": 96912777.0, "step": 2550 }, { "epoch": 0.3245134206843913, "ewc_loss": 0.0810546875, "ewc_loss_diag": 5.2928924560546875e-05, "ewc_loss_parallel": 7.581710815429688e-05, "grad_norm": 5.569797515869141, "learning_rate": 1e-05, "loss": 0.5414, "mean_token_accuracy": 0.8454090356826782, "num_tokens": 96952045.0, "step": 2551 }, { "epoch": 0.3246406309629818, "ewc_loss": 0.07080078125, "ewc_loss_diag": 5.2928924560546875e-05, "ewc_loss_parallel": 6.532669067382812e-05, "grad_norm": 4.896910667419434, "learning_rate": 1e-05, "loss": 0.5018, "mean_token_accuracy": 0.8552893400192261, "num_tokens": 96984625.0, "step": 2552 }, { "epoch": 0.3247678412415723, "ewc_loss": 0.08251953125, "ewc_loss_diag": 5.2928924560546875e-05, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.730680465698242, "learning_rate": 1e-05, "loss": 0.5125, "mean_token_accuracy": 0.8583929538726807, "num_tokens": 97022409.0, "step": 2553 }, { "epoch": 0.32489505152016285, "ewc_loss": 0.06884765625, "ewc_loss_diag": 5.2928924560546875e-05, "ewc_loss_parallel": 6.341934204101562e-05, "grad_norm": 4.7394256591796875, "learning_rate": 1e-05, "loss": 0.5442, "mean_token_accuracy": 0.845029354095459, "num_tokens": 97062461.0, "step": 2554 }, { "epoch": 0.3250222617987533, "ewc_loss": 0.0869140625, "ewc_loss_diag": 5.2928924560546875e-05, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.981016635894775, "learning_rate": 1e-05, "loss": 0.6175, "mean_token_accuracy": 0.8262650966644287, "num_tokens": 97102010.0, "step": 2555 }, { "epoch": 0.32514947207734385, "ewc_loss": 0.07080078125, "ewc_loss_diag": 5.2928924560546875e-05, "ewc_loss_parallel": 6.532669067382812e-05, "grad_norm": 4.780837535858154, "learning_rate": 1e-05, "loss": 0.5071, "mean_token_accuracy": 0.8542264103889465, "num_tokens": 97140304.0, "step": 2556 }, { "epoch": 0.3252766823559344, "ewc_loss": 0.09228515625, "ewc_loss_diag": 5.2928924560546875e-05, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 6.208467960357666, "learning_rate": 1e-05, "loss": 0.4833, "mean_token_accuracy": 0.8714784383773804, "num_tokens": 97180460.0, "step": 2557 }, { "epoch": 0.32540389263452485, "ewc_loss": 0.0751953125, "ewc_loss_diag": 5.316734313964844e-05, "ewc_loss_parallel": 6.961822509765625e-05, "grad_norm": 5.037827491760254, "learning_rate": 1e-05, "loss": 0.5614, "mean_token_accuracy": 0.8425043225288391, "num_tokens": 97223439.0, "step": 2558 }, { "epoch": 0.3255311029131154, "ewc_loss": 0.09130859375, "ewc_loss_diag": 5.316734313964844e-05, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 6.211256980895996, "learning_rate": 1e-05, "loss": 0.5996, "mean_token_accuracy": 0.8371807336807251, "num_tokens": 97257050.0, "step": 2559 }, { "epoch": 0.3256583131917059, "ewc_loss": 0.07958984375, "ewc_loss_diag": 5.316734313964844e-05, "ewc_loss_parallel": 7.43865966796875e-05, "grad_norm": 5.28878927230835, "learning_rate": 1e-05, "loss": 0.5509, "mean_token_accuracy": 0.8443727493286133, "num_tokens": 97296790.0, "step": 2560 }, { "epoch": 0.3257855234702964, "ewc_loss": 0.0849609375, "ewc_loss_diag": 5.316734313964844e-05, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.961029529571533, "learning_rate": 1e-05, "loss": 0.5358, "mean_token_accuracy": 0.8471401929855347, "num_tokens": 97336511.0, "step": 2561 }, { "epoch": 0.3259127337488869, "ewc_loss": 0.07861328125, "ewc_loss_diag": 5.316734313964844e-05, "ewc_loss_parallel": 7.343292236328125e-05, "grad_norm": 5.328920841217041, "learning_rate": 1e-05, "loss": 0.5176, "mean_token_accuracy": 0.8497672080993652, "num_tokens": 97373589.0, "step": 2562 }, { "epoch": 0.32603994402747744, "ewc_loss": 0.07958984375, "ewc_loss_diag": 5.316734313964844e-05, "ewc_loss_parallel": 7.43865966796875e-05, "grad_norm": 5.692415237426758, "learning_rate": 1e-05, "loss": 0.4795, "mean_token_accuracy": 0.8666768074035645, "num_tokens": 97413986.0, "step": 2563 }, { "epoch": 0.3261671543060679, "ewc_loss": 0.076171875, "ewc_loss_diag": 5.316734313964844e-05, "ewc_loss_parallel": 7.05718994140625e-05, "grad_norm": 5.255247116088867, "learning_rate": 1e-05, "loss": 0.5371, "mean_token_accuracy": 0.8461565375328064, "num_tokens": 97457878.0, "step": 2564 }, { "epoch": 0.32629436458465844, "ewc_loss": 0.07666015625, "ewc_loss_diag": 5.316734313964844e-05, "ewc_loss_parallel": 7.152557373046875e-05, "grad_norm": 5.499479293823242, "learning_rate": 1e-05, "loss": 0.5109, "mean_token_accuracy": 0.8551088571548462, "num_tokens": 97495463.0, "step": 2565 }, { "epoch": 0.32642157486324896, "ewc_loss": 0.07421875, "ewc_loss_diag": 5.316734313964844e-05, "ewc_loss_parallel": 6.866455078125e-05, "grad_norm": 5.151513576507568, "learning_rate": 1e-05, "loss": 0.5443, "mean_token_accuracy": 0.8436036109924316, "num_tokens": 97533748.0, "step": 2566 }, { "epoch": 0.32654878514183944, "ewc_loss": 0.07666015625, "ewc_loss_diag": 5.316734313964844e-05, "ewc_loss_parallel": 7.104873657226562e-05, "grad_norm": 5.403998374938965, "learning_rate": 1e-05, "loss": 0.5173, "mean_token_accuracy": 0.8531781435012817, "num_tokens": 97572476.0, "step": 2567 }, { "epoch": 0.32667599542042997, "ewc_loss": 0.07177734375, "ewc_loss_diag": 5.316734313964844e-05, "ewc_loss_parallel": 6.628036499023438e-05, "grad_norm": 5.026008605957031, "learning_rate": 1e-05, "loss": 0.4888, "mean_token_accuracy": 0.8563768863677979, "num_tokens": 97606658.0, "step": 2568 }, { "epoch": 0.3268032056990205, "ewc_loss": 0.07666015625, "ewc_loss_diag": 5.316734313964844e-05, "ewc_loss_parallel": 7.104873657226562e-05, "grad_norm": 5.395273685455322, "learning_rate": 1e-05, "loss": 0.5023, "mean_token_accuracy": 0.8561289310455322, "num_tokens": 97645083.0, "step": 2569 }, { "epoch": 0.32693041597761097, "ewc_loss": 0.0693359375, "ewc_loss_diag": 5.316734313964844e-05, "ewc_loss_parallel": 6.389617919921875e-05, "grad_norm": 4.910671234130859, "learning_rate": 1e-05, "loss": 0.5283, "mean_token_accuracy": 0.8475962281227112, "num_tokens": 97679440.0, "step": 2570 }, { "epoch": 0.3270576262562015, "ewc_loss": 0.0771484375, "ewc_loss_diag": 5.316734313964844e-05, "ewc_loss_parallel": 7.200241088867188e-05, "grad_norm": 5.472333908081055, "learning_rate": 1e-05, "loss": 0.604, "mean_token_accuracy": 0.8307102918624878, "num_tokens": 97719295.0, "step": 2571 }, { "epoch": 0.327184836534792, "ewc_loss": 0.068359375, "ewc_loss_diag": 5.316734313964844e-05, "ewc_loss_parallel": 6.29425048828125e-05, "grad_norm": 4.786942005157471, "learning_rate": 1e-05, "loss": 0.4861, "mean_token_accuracy": 0.8596017360687256, "num_tokens": 97759417.0, "step": 2572 }, { "epoch": 0.3273120468133825, "ewc_loss": 0.08056640625, "ewc_loss_diag": 5.316734313964844e-05, "ewc_loss_parallel": 7.534027099609375e-05, "grad_norm": 5.635074138641357, "learning_rate": 1e-05, "loss": 0.5673, "mean_token_accuracy": 0.8382596373558044, "num_tokens": 97804004.0, "step": 2573 }, { "epoch": 0.327439257091973, "ewc_loss": 0.06884765625, "ewc_loss_diag": 5.316734313964844e-05, "ewc_loss_parallel": 6.341934204101562e-05, "grad_norm": 4.7243170738220215, "learning_rate": 1e-05, "loss": 0.5851, "mean_token_accuracy": 0.8347427845001221, "num_tokens": 97845496.0, "step": 2574 }, { "epoch": 0.32756646737056355, "ewc_loss": 0.08447265625, "ewc_loss_diag": 5.316734313964844e-05, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.859482288360596, "learning_rate": 1e-05, "loss": 0.5572, "mean_token_accuracy": 0.8490316867828369, "num_tokens": 97887184.0, "step": 2575 }, { "epoch": 0.327693677649154, "ewc_loss": 0.06982421875, "ewc_loss_diag": 5.316734313964844e-05, "ewc_loss_parallel": 6.437301635742188e-05, "grad_norm": 4.7902960777282715, "learning_rate": 1e-05, "loss": 0.5568, "mean_token_accuracy": 0.8397486209869385, "num_tokens": 97924498.0, "step": 2576 }, { "epoch": 0.32782088792774455, "ewc_loss": 0.0869140625, "ewc_loss_diag": 5.316734313964844e-05, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 6.004527568817139, "learning_rate": 1e-05, "loss": 0.517, "mean_token_accuracy": 0.8569384813308716, "num_tokens": 97960225.0, "step": 2577 }, { "epoch": 0.3279480982063351, "ewc_loss": 0.07373046875, "ewc_loss_diag": 5.316734313964844e-05, "ewc_loss_parallel": 6.818771362304688e-05, "grad_norm": 4.985360622406006, "learning_rate": 1e-05, "loss": 0.5649, "mean_token_accuracy": 0.8344639539718628, "num_tokens": 98001746.0, "step": 2578 }, { "epoch": 0.3280753084849256, "ewc_loss": 0.0859375, "ewc_loss_diag": 5.316734313964844e-05, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.9356369972229, "learning_rate": 1e-05, "loss": 0.5242, "mean_token_accuracy": 0.8560855984687805, "num_tokens": 98038812.0, "step": 2579 }, { "epoch": 0.3282025187635161, "ewc_loss": 0.076171875, "ewc_loss_diag": 5.316734313964844e-05, "ewc_loss_parallel": 7.05718994140625e-05, "grad_norm": 5.201021194458008, "learning_rate": 1e-05, "loss": 0.5493, "mean_token_accuracy": 0.8472161889076233, "num_tokens": 98073782.0, "step": 2580 }, { "epoch": 0.3283297290421066, "ewc_loss": 0.078125, "ewc_loss_diag": 5.316734313964844e-05, "ewc_loss_parallel": 7.295608520507812e-05, "grad_norm": 5.614582061767578, "learning_rate": 1e-05, "loss": 0.4898, "mean_token_accuracy": 0.8622515201568604, "num_tokens": 98109379.0, "step": 2581 }, { "epoch": 0.32845693932069714, "ewc_loss": 0.078125, "ewc_loss_diag": 5.316734313964844e-05, "ewc_loss_parallel": 7.295608520507812e-05, "grad_norm": 5.30975341796875, "learning_rate": 1e-05, "loss": 0.4884, "mean_token_accuracy": 0.8633188009262085, "num_tokens": 98141661.0, "step": 2582 }, { "epoch": 0.3285841495992876, "ewc_loss": 0.0732421875, "ewc_loss_diag": 5.340576171875e-05, "ewc_loss_parallel": 6.771087646484375e-05, "grad_norm": 5.1965651512146, "learning_rate": 1e-05, "loss": 0.5591, "mean_token_accuracy": 0.8378545045852661, "num_tokens": 98182634.0, "step": 2583 }, { "epoch": 0.32871135987787814, "ewc_loss": 0.07763671875, "ewc_loss_diag": 5.340576171875e-05, "ewc_loss_parallel": 7.2479248046875e-05, "grad_norm": 5.429118633270264, "learning_rate": 1e-05, "loss": 0.5248, "mean_token_accuracy": 0.8533807992935181, "num_tokens": 98225356.0, "step": 2584 }, { "epoch": 0.32883857015646867, "ewc_loss": 0.0693359375, "ewc_loss_diag": 5.340576171875e-05, "ewc_loss_parallel": 6.389617919921875e-05, "grad_norm": 4.832954406738281, "learning_rate": 1e-05, "loss": 0.5214, "mean_token_accuracy": 0.8520143032073975, "num_tokens": 98263575.0, "step": 2585 }, { "epoch": 0.32896578043505914, "ewc_loss": 0.08154296875, "ewc_loss_diag": 5.340576171875e-05, "ewc_loss_parallel": 7.62939453125e-05, "grad_norm": 5.662797451019287, "learning_rate": 1e-05, "loss": 0.5407, "mean_token_accuracy": 0.8469499945640564, "num_tokens": 98304993.0, "step": 2586 }, { "epoch": 0.32909299071364967, "ewc_loss": 0.0673828125, "ewc_loss_diag": 5.340576171875e-05, "ewc_loss_parallel": 6.198883056640625e-05, "grad_norm": 4.609836578369141, "learning_rate": 1e-05, "loss": 0.499, "mean_token_accuracy": 0.8540052771568298, "num_tokens": 98339963.0, "step": 2587 }, { "epoch": 0.3292202009922402, "ewc_loss": 0.08740234375, "ewc_loss_diag": 5.340576171875e-05, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 6.031007289886475, "learning_rate": 1e-05, "loss": 0.514, "mean_token_accuracy": 0.8561736345291138, "num_tokens": 98375507.0, "step": 2588 }, { "epoch": 0.32934741127083067, "ewc_loss": 0.06982421875, "ewc_loss_diag": 5.340576171875e-05, "ewc_loss_parallel": 6.437301635742188e-05, "grad_norm": 4.66731595993042, "learning_rate": 1e-05, "loss": 0.5169, "mean_token_accuracy": 0.8516315221786499, "num_tokens": 98413577.0, "step": 2589 }, { "epoch": 0.3294746215494212, "ewc_loss": 0.09716796875, "ewc_loss_diag": 5.340576171875e-05, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.369370937347412, "learning_rate": 1e-05, "loss": 0.5813, "mean_token_accuracy": 0.8419064283370972, "num_tokens": 98451615.0, "step": 2590 }, { "epoch": 0.3296018318280117, "ewc_loss": 0.0771484375, "ewc_loss_diag": 5.340576171875e-05, "ewc_loss_parallel": 7.200241088867188e-05, "grad_norm": 5.0740132331848145, "learning_rate": 1e-05, "loss": 0.6188, "mean_token_accuracy": 0.8267964124679565, "num_tokens": 98490728.0, "step": 2591 }, { "epoch": 0.3297290421066022, "ewc_loss": 0.09716796875, "ewc_loss_diag": 5.340576171875e-05, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.326788902282715, "learning_rate": 1e-05, "loss": 0.5054, "mean_token_accuracy": 0.8644003868103027, "num_tokens": 98526645.0, "step": 2592 }, { "epoch": 0.3298562523851927, "ewc_loss": 0.08251953125, "ewc_loss_diag": 5.340576171875e-05, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.408952236175537, "learning_rate": 1e-05, "loss": 0.551, "mean_token_accuracy": 0.8435268998146057, "num_tokens": 98568447.0, "step": 2593 }, { "epoch": 0.32998346266378326, "ewc_loss": 0.08544921875, "ewc_loss_diag": 5.340576171875e-05, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.9852728843688965, "learning_rate": 1e-05, "loss": 0.5358, "mean_token_accuracy": 0.8499170541763306, "num_tokens": 98608952.0, "step": 2594 }, { "epoch": 0.33011067294237373, "ewc_loss": 0.08154296875, "ewc_loss_diag": 5.340576171875e-05, "ewc_loss_parallel": 7.62939453125e-05, "grad_norm": 5.42966890335083, "learning_rate": 1e-05, "loss": 0.4695, "mean_token_accuracy": 0.8688507080078125, "num_tokens": 98649180.0, "step": 2595 }, { "epoch": 0.33023788322096426, "ewc_loss": 0.07861328125, "ewc_loss_diag": 5.340576171875e-05, "ewc_loss_parallel": 7.343292236328125e-05, "grad_norm": 5.650240898132324, "learning_rate": 1e-05, "loss": 0.5407, "mean_token_accuracy": 0.8488653898239136, "num_tokens": 98692506.0, "step": 2596 }, { "epoch": 0.3303650934995548, "ewc_loss": 0.078125, "ewc_loss_diag": 5.340576171875e-05, "ewc_loss_parallel": 7.295608520507812e-05, "grad_norm": 5.329733371734619, "learning_rate": 1e-05, "loss": 0.5315, "mean_token_accuracy": 0.8467273116111755, "num_tokens": 98727817.0, "step": 2597 }, { "epoch": 0.33049230377814526, "ewc_loss": 0.07568359375, "ewc_loss_diag": 5.340576171875e-05, "ewc_loss_parallel": 7.009506225585938e-05, "grad_norm": 5.409660339355469, "learning_rate": 1e-05, "loss": 0.5111, "mean_token_accuracy": 0.8541162014007568, "num_tokens": 98769714.0, "step": 2598 }, { "epoch": 0.3306195140567358, "ewc_loss": 0.0751953125, "ewc_loss_diag": 5.340576171875e-05, "ewc_loss_parallel": 6.961822509765625e-05, "grad_norm": 5.211667537689209, "learning_rate": 1e-05, "loss": 0.4956, "mean_token_accuracy": 0.8596862554550171, "num_tokens": 98811894.0, "step": 2599 }, { "epoch": 0.3307467243353263, "ewc_loss": 0.0751953125, "ewc_loss_diag": 5.340576171875e-05, "ewc_loss_parallel": 6.961822509765625e-05, "grad_norm": 5.307365417480469, "learning_rate": 1e-05, "loss": 0.5508, "mean_token_accuracy": 0.8442080020904541, "num_tokens": 98847117.0, "step": 2600 }, { "epoch": 0.3308739346139168, "ewc_loss": 0.07373046875, "ewc_loss_diag": 5.3882598876953125e-05, "ewc_loss_parallel": 6.818771362304688e-05, "grad_norm": 5.083259105682373, "learning_rate": 1e-05, "loss": 0.514, "mean_token_accuracy": 0.8536691665649414, "num_tokens": 98882745.0, "step": 2601 }, { "epoch": 0.3310011448925073, "ewc_loss": 0.0751953125, "ewc_loss_diag": 5.3882598876953125e-05, "ewc_loss_parallel": 6.961822509765625e-05, "grad_norm": 5.273924350738525, "learning_rate": 1e-05, "loss": 0.4991, "mean_token_accuracy": 0.861609935760498, "num_tokens": 98918039.0, "step": 2602 }, { "epoch": 0.33112835517109784, "ewc_loss": 0.0712890625, "ewc_loss_diag": 5.3882598876953125e-05, "ewc_loss_parallel": 6.580352783203125e-05, "grad_norm": 4.952362060546875, "learning_rate": 1e-05, "loss": 0.5148, "mean_token_accuracy": 0.8508886098861694, "num_tokens": 98958571.0, "step": 2603 }, { "epoch": 0.3312555654496883, "ewc_loss": 0.07666015625, "ewc_loss_diag": 5.3882598876953125e-05, "ewc_loss_parallel": 7.104873657226562e-05, "grad_norm": 5.333338737487793, "learning_rate": 1e-05, "loss": 0.5131, "mean_token_accuracy": 0.8534237146377563, "num_tokens": 99004970.0, "step": 2604 }, { "epoch": 0.33138277572827884, "ewc_loss": 0.06982421875, "ewc_loss_diag": 5.3882598876953125e-05, "ewc_loss_parallel": 6.437301635742188e-05, "grad_norm": 4.8311848640441895, "learning_rate": 1e-05, "loss": 0.5699, "mean_token_accuracy": 0.8397262096405029, "num_tokens": 99041112.0, "step": 2605 }, { "epoch": 0.3315099860068694, "ewc_loss": 0.07763671875, "ewc_loss_diag": 5.3882598876953125e-05, "ewc_loss_parallel": 7.2479248046875e-05, "grad_norm": 5.475688934326172, "learning_rate": 1e-05, "loss": 0.5166, "mean_token_accuracy": 0.853073000907898, "num_tokens": 99076771.0, "step": 2606 }, { "epoch": 0.33163719628545985, "ewc_loss": 0.06787109375, "ewc_loss_diag": 5.3882598876953125e-05, "ewc_loss_parallel": 6.246566772460938e-05, "grad_norm": 4.711187839508057, "learning_rate": 1e-05, "loss": 0.5202, "mean_token_accuracy": 0.848482608795166, "num_tokens": 99112183.0, "step": 2607 }, { "epoch": 0.3317644065640504, "ewc_loss": 0.08203125, "ewc_loss_diag": 5.3882598876953125e-05, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.70461893081665, "learning_rate": 1e-05, "loss": 0.5162, "mean_token_accuracy": 0.851157546043396, "num_tokens": 99148789.0, "step": 2608 }, { "epoch": 0.3318916168426409, "ewc_loss": 0.068359375, "ewc_loss_diag": 5.3882598876953125e-05, "ewc_loss_parallel": 6.29425048828125e-05, "grad_norm": 4.629613876342773, "learning_rate": 1e-05, "loss": 0.5935, "mean_token_accuracy": 0.8288277387619019, "num_tokens": 99189697.0, "step": 2609 }, { "epoch": 0.3320188271212314, "ewc_loss": 0.08837890625, "ewc_loss_diag": 5.3882598876953125e-05, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 6.040578365325928, "learning_rate": 1e-05, "loss": 0.5622, "mean_token_accuracy": 0.8463097214698792, "num_tokens": 99228269.0, "step": 2610 }, { "epoch": 0.3321460373998219, "ewc_loss": 0.07080078125, "ewc_loss_diag": 5.3882598876953125e-05, "ewc_loss_parallel": 6.532669067382812e-05, "grad_norm": 4.727248191833496, "learning_rate": 1e-05, "loss": 0.5231, "mean_token_accuracy": 0.8499819040298462, "num_tokens": 99264275.0, "step": 2611 }, { "epoch": 0.33227324767841243, "ewc_loss": 0.0966796875, "ewc_loss_diag": 5.3882598876953125e-05, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.320553779602051, "learning_rate": 1e-05, "loss": 0.5362, "mean_token_accuracy": 0.8526937961578369, "num_tokens": 99294598.0, "step": 2612 }, { "epoch": 0.3324004579570029, "ewc_loss": 0.07763671875, "ewc_loss_diag": 5.3882598876953125e-05, "ewc_loss_parallel": 7.2479248046875e-05, "grad_norm": 5.091677188873291, "learning_rate": 1e-05, "loss": 0.5104, "mean_token_accuracy": 0.8562116026878357, "num_tokens": 99336307.0, "step": 2613 }, { "epoch": 0.33252766823559343, "ewc_loss": 0.095703125, "ewc_loss_diag": 5.3882598876953125e-05, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.231895446777344, "learning_rate": 1e-05, "loss": 0.5109, "mean_token_accuracy": 0.8624860048294067, "num_tokens": 99373983.0, "step": 2614 }, { "epoch": 0.33265487851418396, "ewc_loss": 0.0830078125, "ewc_loss_diag": 5.3882598876953125e-05, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.40665340423584, "learning_rate": 1e-05, "loss": 0.5581, "mean_token_accuracy": 0.8446931838989258, "num_tokens": 99415794.0, "step": 2615 }, { "epoch": 0.33278208879277443, "ewc_loss": 0.0830078125, "ewc_loss_diag": 5.3882598876953125e-05, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.84495210647583, "learning_rate": 1e-05, "loss": 0.549, "mean_token_accuracy": 0.8437941074371338, "num_tokens": 99457572.0, "step": 2616 }, { "epoch": 0.33290929907136496, "ewc_loss": 0.08203125, "ewc_loss_diag": 5.435943603515625e-05, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.462204456329346, "learning_rate": 1e-05, "loss": 0.5364, "mean_token_accuracy": 0.8513367772102356, "num_tokens": 99496074.0, "step": 2617 }, { "epoch": 0.3330365093499555, "ewc_loss": 0.07666015625, "ewc_loss_diag": 5.435943603515625e-05, "ewc_loss_parallel": 7.104873657226562e-05, "grad_norm": 5.412470817565918, "learning_rate": 1e-05, "loss": 0.547, "mean_token_accuracy": 0.8424893617630005, "num_tokens": 99539655.0, "step": 2618 }, { "epoch": 0.33316371962854596, "ewc_loss": 0.08154296875, "ewc_loss_diag": 5.435943603515625e-05, "ewc_loss_parallel": 7.62939453125e-05, "grad_norm": 5.4888811111450195, "learning_rate": 1e-05, "loss": 0.5638, "mean_token_accuracy": 0.843804657459259, "num_tokens": 99574696.0, "step": 2619 }, { "epoch": 0.3332909299071365, "ewc_loss": 0.0732421875, "ewc_loss_diag": 5.435943603515625e-05, "ewc_loss_parallel": 6.771087646484375e-05, "grad_norm": 5.053735256195068, "learning_rate": 1e-05, "loss": 0.5116, "mean_token_accuracy": 0.8531240224838257, "num_tokens": 99612960.0, "step": 2620 }, { "epoch": 0.333418140185727, "ewc_loss": 0.07958984375, "ewc_loss_diag": 5.435943603515625e-05, "ewc_loss_parallel": 7.43865966796875e-05, "grad_norm": 5.575196266174316, "learning_rate": 1e-05, "loss": 0.6089, "mean_token_accuracy": 0.8264197111129761, "num_tokens": 99644107.0, "step": 2621 }, { "epoch": 0.3335453504643175, "ewc_loss": 0.0703125, "ewc_loss_diag": 5.435943603515625e-05, "ewc_loss_parallel": 6.4849853515625e-05, "grad_norm": 4.825349807739258, "learning_rate": 1e-05, "loss": 0.5219, "mean_token_accuracy": 0.8493549227714539, "num_tokens": 99682014.0, "step": 2622 }, { "epoch": 0.333672560742908, "ewc_loss": 0.08251953125, "ewc_loss_diag": 5.435943603515625e-05, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.741227626800537, "learning_rate": 1e-05, "loss": 0.5557, "mean_token_accuracy": 0.8449201583862305, "num_tokens": 99718969.0, "step": 2623 }, { "epoch": 0.33379977102149855, "ewc_loss": 0.06884765625, "ewc_loss_diag": 5.435943603515625e-05, "ewc_loss_parallel": 6.341934204101562e-05, "grad_norm": 4.733940124511719, "learning_rate": 1e-05, "loss": 0.5707, "mean_token_accuracy": 0.8358973860740662, "num_tokens": 99756955.0, "step": 2624 }, { "epoch": 0.333926981300089, "ewc_loss": 0.08740234375, "ewc_loss_diag": 5.435943603515625e-05, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.979804039001465, "learning_rate": 1e-05, "loss": 0.5062, "mean_token_accuracy": 0.8569002151489258, "num_tokens": 99792216.0, "step": 2625 }, { "epoch": 0.33405419157867955, "ewc_loss": 0.07177734375, "ewc_loss_diag": 5.435943603515625e-05, "ewc_loss_parallel": 6.628036499023438e-05, "grad_norm": 4.820825099945068, "learning_rate": 1e-05, "loss": 0.5347, "mean_token_accuracy": 0.8412898182868958, "num_tokens": 99823973.0, "step": 2626 }, { "epoch": 0.3341814018572701, "ewc_loss": 0.091796875, "ewc_loss_diag": 5.435943603515625e-05, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 6.136226654052734, "learning_rate": 1e-05, "loss": 0.5611, "mean_token_accuracy": 0.845222532749176, "num_tokens": 99857206.0, "step": 2627 }, { "epoch": 0.33430861213586055, "ewc_loss": 0.0751953125, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 6.961822509765625e-05, "grad_norm": 5.031576633453369, "learning_rate": 1e-05, "loss": 0.5951, "mean_token_accuracy": 0.8324726819992065, "num_tokens": 99891146.0, "step": 2628 }, { "epoch": 0.3344358224144511, "ewc_loss": 0.08935546875, "ewc_loss_diag": 5.435943603515625e-05, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 6.098262786865234, "learning_rate": 1e-05, "loss": 0.6057, "mean_token_accuracy": 0.8325793743133545, "num_tokens": 99931456.0, "step": 2629 }, { "epoch": 0.3345630326930416, "ewc_loss": 0.078125, "ewc_loss_diag": 5.435943603515625e-05, "ewc_loss_parallel": 7.295608520507812e-05, "grad_norm": 5.2022786140441895, "learning_rate": 1e-05, "loss": 0.5408, "mean_token_accuracy": 0.8438736200332642, "num_tokens": 99965968.0, "step": 2630 }, { "epoch": 0.33469024297163213, "ewc_loss": 0.0849609375, "ewc_loss_diag": 5.459785461425781e-05, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.880512237548828, "learning_rate": 1e-05, "loss": 0.5759, "mean_token_accuracy": 0.8378764390945435, "num_tokens": 100010874.0, "step": 2631 }, { "epoch": 0.3348174532502226, "ewc_loss": 0.0791015625, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 7.390975952148438e-05, "grad_norm": 5.241141319274902, "learning_rate": 1e-05, "loss": 0.5196, "mean_token_accuracy": 0.8544861078262329, "num_tokens": 100052166.0, "step": 2632 }, { "epoch": 0.33494466352881314, "ewc_loss": 0.0791015625, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 7.390975952148438e-05, "grad_norm": 5.642301559448242, "learning_rate": 1e-05, "loss": 0.5219, "mean_token_accuracy": 0.8552826642990112, "num_tokens": 100093927.0, "step": 2633 }, { "epoch": 0.33507187380740366, "ewc_loss": 0.07666015625, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 7.104873657226562e-05, "grad_norm": 5.2151994705200195, "learning_rate": 1e-05, "loss": 0.5488, "mean_token_accuracy": 0.8445972204208374, "num_tokens": 100128414.0, "step": 2634 }, { "epoch": 0.33519908408599414, "ewc_loss": 0.07666015625, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 7.152557373046875e-05, "grad_norm": 5.4404473304748535, "learning_rate": 1e-05, "loss": 0.4949, "mean_token_accuracy": 0.8571702241897583, "num_tokens": 100160334.0, "step": 2635 }, { "epoch": 0.33532629436458466, "ewc_loss": 0.0751953125, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 6.961822509765625e-05, "grad_norm": 5.147219657897949, "learning_rate": 1e-05, "loss": 0.5157, "mean_token_accuracy": 0.8492381572723389, "num_tokens": 100195188.0, "step": 2636 }, { "epoch": 0.3354535046431752, "ewc_loss": 0.076171875, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 7.05718994140625e-05, "grad_norm": 5.314585208892822, "learning_rate": 1e-05, "loss": 0.5226, "mean_token_accuracy": 0.8550688028335571, "num_tokens": 100235338.0, "step": 2637 }, { "epoch": 0.33558071492176567, "ewc_loss": 0.07373046875, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 6.818771362304688e-05, "grad_norm": 5.064279079437256, "learning_rate": 1e-05, "loss": 0.5292, "mean_token_accuracy": 0.8524672389030457, "num_tokens": 100276267.0, "step": 2638 }, { "epoch": 0.3357079252003562, "ewc_loss": 0.07568359375, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 7.009506225585938e-05, "grad_norm": 5.267767906188965, "learning_rate": 1e-05, "loss": 0.5858, "mean_token_accuracy": 0.8311212062835693, "num_tokens": 100316201.0, "step": 2639 }, { "epoch": 0.3358351354789467, "ewc_loss": 0.07177734375, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 6.628036499023438e-05, "grad_norm": 4.9757232666015625, "learning_rate": 1e-05, "loss": 0.5798, "mean_token_accuracy": 0.8371186256408691, "num_tokens": 100353278.0, "step": 2640 }, { "epoch": 0.3359623457575372, "ewc_loss": 0.07666015625, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 7.104873657226562e-05, "grad_norm": 5.3054986000061035, "learning_rate": 1e-05, "loss": 0.552, "mean_token_accuracy": 0.8444874286651611, "num_tokens": 100384185.0, "step": 2641 }, { "epoch": 0.3360895560361277, "ewc_loss": 0.07080078125, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 6.532669067382812e-05, "grad_norm": 4.873110771179199, "learning_rate": 1e-05, "loss": 0.5692, "mean_token_accuracy": 0.8332079648971558, "num_tokens": 100423232.0, "step": 2642 }, { "epoch": 0.33621676631471825, "ewc_loss": 0.07666015625, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 7.152557373046875e-05, "grad_norm": 5.3987345695495605, "learning_rate": 1e-05, "loss": 0.5976, "mean_token_accuracy": 0.8269809484481812, "num_tokens": 100457912.0, "step": 2643 }, { "epoch": 0.3363439765933087, "ewc_loss": 0.0693359375, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 6.389617919921875e-05, "grad_norm": 4.749094009399414, "learning_rate": 1e-05, "loss": 0.5147, "mean_token_accuracy": 0.8538762331008911, "num_tokens": 100499894.0, "step": 2644 }, { "epoch": 0.33647118687189925, "ewc_loss": 0.080078125, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 7.486343383789062e-05, "grad_norm": 5.598693370819092, "learning_rate": 1e-05, "loss": 0.4975, "mean_token_accuracy": 0.8641291856765747, "num_tokens": 100540961.0, "step": 2645 }, { "epoch": 0.3365983971504898, "ewc_loss": 0.068359375, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 6.29425048828125e-05, "grad_norm": 4.655335903167725, "learning_rate": 1e-05, "loss": 0.5322, "mean_token_accuracy": 0.8442004323005676, "num_tokens": 100581302.0, "step": 2646 }, { "epoch": 0.33672560742908025, "ewc_loss": 0.08544921875, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.885240077972412, "learning_rate": 1e-05, "loss": 0.5092, "mean_token_accuracy": 0.8548227548599243, "num_tokens": 100615289.0, "step": 2647 }, { "epoch": 0.3368528177076708, "ewc_loss": 0.06982421875, "ewc_loss_diag": 5.4836273193359375e-05, "ewc_loss_parallel": 6.437301635742188e-05, "grad_norm": 4.668736457824707, "learning_rate": 1e-05, "loss": 0.4954, "mean_token_accuracy": 0.8600051403045654, "num_tokens": 100659911.0, "step": 2648 }, { "epoch": 0.3369800279862613, "ewc_loss": 0.09326171875, "ewc_loss_diag": 5.507469177246094e-05, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 6.184946060180664, "learning_rate": 1e-05, "loss": 0.5376, "mean_token_accuracy": 0.8516532182693481, "num_tokens": 100700861.0, "step": 2649 }, { "epoch": 0.3371072382648518, "ewc_loss": 0.07470703125, "ewc_loss_diag": 5.507469177246094e-05, "ewc_loss_parallel": 6.914138793945312e-05, "grad_norm": 4.925995826721191, "learning_rate": 1e-05, "loss": 0.5346, "mean_token_accuracy": 0.8501655459403992, "num_tokens": 100737329.0, "step": 2650 }, { "epoch": 0.3372344485434423, "ewc_loss": 0.09521484375, "ewc_loss_diag": 5.507469177246094e-05, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.242916107177734, "learning_rate": 1e-05, "loss": 0.5012, "mean_token_accuracy": 0.8654448986053467, "num_tokens": 100774278.0, "step": 2651 }, { "epoch": 0.33736165882203284, "ewc_loss": 0.08056640625, "ewc_loss_diag": 5.53131103515625e-05, "ewc_loss_parallel": 7.534027099609375e-05, "grad_norm": 5.255927085876465, "learning_rate": 1e-05, "loss": 0.5382, "mean_token_accuracy": 0.8473860621452332, "num_tokens": 100819083.0, "step": 2652 }, { "epoch": 0.3374888691006233, "ewc_loss": 0.0869140625, "ewc_loss_diag": 5.53131103515625e-05, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.957594871520996, "learning_rate": 1e-05, "loss": 0.602, "mean_token_accuracy": 0.836180567741394, "num_tokens": 100859711.0, "step": 2653 }, { "epoch": 0.33761607937921384, "ewc_loss": 0.08154296875, "ewc_loss_diag": 5.53131103515625e-05, "ewc_loss_parallel": 7.62939453125e-05, "grad_norm": 5.41700553894043, "learning_rate": 1e-05, "loss": 0.6023, "mean_token_accuracy": 0.8268651962280273, "num_tokens": 100893915.0, "step": 2654 }, { "epoch": 0.33774328965780437, "ewc_loss": 0.0771484375, "ewc_loss_diag": 5.53131103515625e-05, "ewc_loss_parallel": 7.200241088867188e-05, "grad_norm": 5.519402027130127, "learning_rate": 1e-05, "loss": 0.5065, "mean_token_accuracy": 0.8581291437149048, "num_tokens": 100932609.0, "step": 2655 }, { "epoch": 0.33787049993639484, "ewc_loss": 0.08154296875, "ewc_loss_diag": 5.53131103515625e-05, "ewc_loss_parallel": 7.62939453125e-05, "grad_norm": 5.44379186630249, "learning_rate": 1e-05, "loss": 0.5275, "mean_token_accuracy": 0.8524262309074402, "num_tokens": 100966539.0, "step": 2656 }, { "epoch": 0.33799771021498537, "ewc_loss": 0.07373046875, "ewc_loss_diag": 5.53131103515625e-05, "ewc_loss_parallel": 6.818771362304688e-05, "grad_norm": 5.1371169090271, "learning_rate": 1e-05, "loss": 0.5796, "mean_token_accuracy": 0.8385008573532104, "num_tokens": 100999979.0, "step": 2657 }, { "epoch": 0.3381249204935759, "ewc_loss": 0.08056640625, "ewc_loss_diag": 5.53131103515625e-05, "ewc_loss_parallel": 7.534027099609375e-05, "grad_norm": 5.521259307861328, "learning_rate": 1e-05, "loss": 0.5626, "mean_token_accuracy": 0.8410801887512207, "num_tokens": 101036546.0, "step": 2658 }, { "epoch": 0.33825213077216637, "ewc_loss": 0.0703125, "ewc_loss_diag": 5.53131103515625e-05, "ewc_loss_parallel": 6.4849853515625e-05, "grad_norm": 4.86692476272583, "learning_rate": 1e-05, "loss": 0.506, "mean_token_accuracy": 0.8483514785766602, "num_tokens": 101067861.0, "step": 2659 }, { "epoch": 0.3383793410507569, "ewc_loss": 0.08251953125, "ewc_loss_diag": 5.53131103515625e-05, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.6782546043396, "learning_rate": 1e-05, "loss": 0.4671, "mean_token_accuracy": 0.8694645762443542, "num_tokens": 101104268.0, "step": 2660 }, { "epoch": 0.3385065513293474, "ewc_loss": 0.0703125, "ewc_loss_diag": 5.53131103515625e-05, "ewc_loss_parallel": 6.4849853515625e-05, "grad_norm": 4.731169700622559, "learning_rate": 1e-05, "loss": 0.4597, "mean_token_accuracy": 0.8678081035614014, "num_tokens": 101143364.0, "step": 2661 }, { "epoch": 0.3386337616079379, "ewc_loss": 0.0859375, "ewc_loss_diag": 5.53131103515625e-05, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.90704870223999, "learning_rate": 1e-05, "loss": 0.577, "mean_token_accuracy": 0.8385206460952759, "num_tokens": 101178359.0, "step": 2662 }, { "epoch": 0.33876097188652843, "ewc_loss": 0.07080078125, "ewc_loss_diag": 5.53131103515625e-05, "ewc_loss_parallel": 6.532669067382812e-05, "grad_norm": 4.766583442687988, "learning_rate": 1e-05, "loss": 0.4823, "mean_token_accuracy": 0.8624386787414551, "num_tokens": 101215993.0, "step": 2663 }, { "epoch": 0.33888818216511896, "ewc_loss": 0.0908203125, "ewc_loss_diag": 5.53131103515625e-05, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 6.098449230194092, "learning_rate": 1e-05, "loss": 0.4994, "mean_token_accuracy": 0.8621011972427368, "num_tokens": 101254924.0, "step": 2664 }, { "epoch": 0.33901539244370943, "ewc_loss": 0.07470703125, "ewc_loss_diag": 5.53131103515625e-05, "ewc_loss_parallel": 6.914138793945312e-05, "grad_norm": 4.947254657745361, "learning_rate": 1e-05, "loss": 0.4854, "mean_token_accuracy": 0.8605995178222656, "num_tokens": 101295632.0, "step": 2665 }, { "epoch": 0.33914260272229996, "ewc_loss": 0.091796875, "ewc_loss_diag": 5.5789947509765625e-05, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 6.126298904418945, "learning_rate": 1e-05, "loss": 0.5125, "mean_token_accuracy": 0.8569892644882202, "num_tokens": 101333295.0, "step": 2666 }, { "epoch": 0.3392698130008905, "ewc_loss": 0.078125, "ewc_loss_diag": 5.5789947509765625e-05, "ewc_loss_parallel": 7.295608520507812e-05, "grad_norm": 5.156805038452148, "learning_rate": 1e-05, "loss": 0.5145, "mean_token_accuracy": 0.8542537093162537, "num_tokens": 101372319.0, "step": 2667 }, { "epoch": 0.33939702327948096, "ewc_loss": 0.08642578125, "ewc_loss_diag": 5.5789947509765625e-05, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.9587812423706055, "learning_rate": 1e-05, "loss": 0.5285, "mean_token_accuracy": 0.8511784076690674, "num_tokens": 101405300.0, "step": 2668 }, { "epoch": 0.3395242335580715, "ewc_loss": 0.07958984375, "ewc_loss_diag": 5.5789947509765625e-05, "ewc_loss_parallel": 7.43865966796875e-05, "grad_norm": 5.245472431182861, "learning_rate": 1e-05, "loss": 0.5875, "mean_token_accuracy": 0.8363693356513977, "num_tokens": 101445576.0, "step": 2669 }, { "epoch": 0.339651443836662, "ewc_loss": 0.0810546875, "ewc_loss_diag": 5.5789947509765625e-05, "ewc_loss_parallel": 7.581710815429688e-05, "grad_norm": 5.722064971923828, "learning_rate": 1e-05, "loss": 0.6056, "mean_token_accuracy": 0.832127571105957, "num_tokens": 101481840.0, "step": 2670 }, { "epoch": 0.3397786541152525, "ewc_loss": 0.0771484375, "ewc_loss_diag": 5.5789947509765625e-05, "ewc_loss_parallel": 7.200241088867188e-05, "grad_norm": 5.249242305755615, "learning_rate": 1e-05, "loss": 0.5662, "mean_token_accuracy": 0.8402368426322937, "num_tokens": 101511968.0, "step": 2671 }, { "epoch": 0.339905864393843, "ewc_loss": 0.0771484375, "ewc_loss_diag": 5.5789947509765625e-05, "ewc_loss_parallel": 7.200241088867188e-05, "grad_norm": 5.475844383239746, "learning_rate": 1e-05, "loss": 0.5342, "mean_token_accuracy": 0.8543026447296143, "num_tokens": 101549909.0, "step": 2672 }, { "epoch": 0.34003307467243354, "ewc_loss": 0.076171875, "ewc_loss_diag": 5.5789947509765625e-05, "ewc_loss_parallel": 7.05718994140625e-05, "grad_norm": 5.191101551055908, "learning_rate": 1e-05, "loss": 0.558, "mean_token_accuracy": 0.8425053358078003, "num_tokens": 101580892.0, "step": 2673 }, { "epoch": 0.340160284951024, "ewc_loss": 0.076171875, "ewc_loss_diag": 5.5789947509765625e-05, "ewc_loss_parallel": 7.05718994140625e-05, "grad_norm": 5.321089267730713, "learning_rate": 1e-05, "loss": 0.5099, "mean_token_accuracy": 0.8554971218109131, "num_tokens": 101614589.0, "step": 2674 }, { "epoch": 0.34028749522961454, "ewc_loss": 0.07470703125, "ewc_loss_diag": 5.5789947509765625e-05, "ewc_loss_parallel": 6.914138793945312e-05, "grad_norm": 5.100606441497803, "learning_rate": 1e-05, "loss": 0.5134, "mean_token_accuracy": 0.855053186416626, "num_tokens": 101658988.0, "step": 2675 }, { "epoch": 0.3404147055082051, "ewc_loss": 0.07568359375, "ewc_loss_diag": 5.5789947509765625e-05, "ewc_loss_parallel": 7.009506225585938e-05, "grad_norm": 5.262749195098877, "learning_rate": 1e-05, "loss": 0.5988, "mean_token_accuracy": 0.8347588777542114, "num_tokens": 101697838.0, "step": 2676 }, { "epoch": 0.34054191578679555, "ewc_loss": 0.07373046875, "ewc_loss_diag": 5.5789947509765625e-05, "ewc_loss_parallel": 6.818771362304688e-05, "grad_norm": 5.01204252243042, "learning_rate": 1e-05, "loss": 0.4839, "mean_token_accuracy": 0.8597540855407715, "num_tokens": 101733831.0, "step": 2677 }, { "epoch": 0.3406691260653861, "ewc_loss": 0.076171875, "ewc_loss_diag": 5.5789947509765625e-05, "ewc_loss_parallel": 7.05718994140625e-05, "grad_norm": 5.266571998596191, "learning_rate": 1e-05, "loss": 0.5244, "mean_token_accuracy": 0.8524723052978516, "num_tokens": 101772697.0, "step": 2678 }, { "epoch": 0.3407963363439766, "ewc_loss": 0.07177734375, "ewc_loss_diag": 5.5789947509765625e-05, "ewc_loss_parallel": 6.628036499023438e-05, "grad_norm": 4.925832271575928, "learning_rate": 1e-05, "loss": 0.594, "mean_token_accuracy": 0.832689642906189, "num_tokens": 101810107.0, "step": 2679 }, { "epoch": 0.34092354662256713, "ewc_loss": 0.07666015625, "ewc_loss_diag": 5.5789947509765625e-05, "ewc_loss_parallel": 7.104873657226562e-05, "grad_norm": 5.333343505859375, "learning_rate": 1e-05, "loss": 0.5134, "mean_token_accuracy": 0.854812741279602, "num_tokens": 101847239.0, "step": 2680 }, { "epoch": 0.3410507569011576, "ewc_loss": 0.0703125, "ewc_loss_diag": 5.602836608886719e-05, "ewc_loss_parallel": 6.4849853515625e-05, "grad_norm": 4.812684535980225, "learning_rate": 1e-05, "loss": 0.555, "mean_token_accuracy": 0.8398300409317017, "num_tokens": 101884508.0, "step": 2681 }, { "epoch": 0.34117796717974813, "ewc_loss": 0.0791015625, "ewc_loss_diag": 5.602836608886719e-05, "ewc_loss_parallel": 7.390975952148438e-05, "grad_norm": 5.475593090057373, "learning_rate": 1e-05, "loss": 0.4484, "mean_token_accuracy": 0.8741617202758789, "num_tokens": 101925351.0, "step": 2682 }, { "epoch": 0.34130517745833866, "ewc_loss": 0.068359375, "ewc_loss_diag": 5.602836608886719e-05, "ewc_loss_parallel": 6.29425048828125e-05, "grad_norm": 4.690922260284424, "learning_rate": 1e-05, "loss": 0.512, "mean_token_accuracy": 0.8536242246627808, "num_tokens": 101966086.0, "step": 2683 }, { "epoch": 0.34143238773692913, "ewc_loss": 0.0830078125, "ewc_loss_diag": 5.602836608886719e-05, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.727109909057617, "learning_rate": 1e-05, "loss": 0.5164, "mean_token_accuracy": 0.8563166856765747, "num_tokens": 102004935.0, "step": 2684 }, { "epoch": 0.34155959801551966, "ewc_loss": 0.068359375, "ewc_loss_diag": 5.602836608886719e-05, "ewc_loss_parallel": 6.29425048828125e-05, "grad_norm": 4.634085655212402, "learning_rate": 1e-05, "loss": 0.5328, "mean_token_accuracy": 0.8492488861083984, "num_tokens": 102039026.0, "step": 2685 }, { "epoch": 0.3416868082941102, "ewc_loss": 0.08984375, "ewc_loss_diag": 5.602836608886719e-05, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 6.053924083709717, "learning_rate": 1e-05, "loss": 0.5432, "mean_token_accuracy": 0.8480614423751831, "num_tokens": 102078990.0, "step": 2686 }, { "epoch": 0.34181401857270066, "ewc_loss": 0.072265625, "ewc_loss_diag": 5.602836608886719e-05, "ewc_loss_parallel": 6.628036499023438e-05, "grad_norm": 4.742849349975586, "learning_rate": 1e-05, "loss": 0.5268, "mean_token_accuracy": 0.8552308678627014, "num_tokens": 102114631.0, "step": 2687 }, { "epoch": 0.3419412288512912, "ewc_loss": 0.09765625, "ewc_loss_diag": 5.602836608886719e-05, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.303744316101074, "learning_rate": 1e-05, "loss": 0.5507, "mean_token_accuracy": 0.8517231941223145, "num_tokens": 102161583.0, "step": 2688 }, { "epoch": 0.3420684391298817, "ewc_loss": 0.0791015625, "ewc_loss_diag": 5.602836608886719e-05, "ewc_loss_parallel": 7.343292236328125e-05, "grad_norm": 5.066499710083008, "learning_rate": 1e-05, "loss": 0.5059, "mean_token_accuracy": 0.8561995029449463, "num_tokens": 102202183.0, "step": 2689 }, { "epoch": 0.3421956494084722, "ewc_loss": 0.0947265625, "ewc_loss_diag": 5.602836608886719e-05, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.240882396697998, "learning_rate": 1e-05, "loss": 0.5138, "mean_token_accuracy": 0.8573060035705566, "num_tokens": 102235288.0, "step": 2690 }, { "epoch": 0.3423228596870627, "ewc_loss": 0.0830078125, "ewc_loss_diag": 5.602836608886719e-05, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.368785381317139, "learning_rate": 1e-05, "loss": 0.5512, "mean_token_accuracy": 0.8453490734100342, "num_tokens": 102272814.0, "step": 2691 }, { "epoch": 0.34245006996565325, "ewc_loss": 0.0849609375, "ewc_loss_diag": 5.602836608886719e-05, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.8777337074279785, "learning_rate": 1e-05, "loss": 0.5302, "mean_token_accuracy": 0.8491498231887817, "num_tokens": 102312763.0, "step": 2692 }, { "epoch": 0.3425772802442437, "ewc_loss": 0.0830078125, "ewc_loss_diag": 5.602836608886719e-05, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.432338714599609, "learning_rate": 1e-05, "loss": 0.523, "mean_token_accuracy": 0.8542096614837646, "num_tokens": 102351490.0, "step": 2693 }, { "epoch": 0.34270449052283425, "ewc_loss": 0.078125, "ewc_loss_diag": 5.602836608886719e-05, "ewc_loss_parallel": 7.2479248046875e-05, "grad_norm": 5.465429782867432, "learning_rate": 1e-05, "loss": 0.5364, "mean_token_accuracy": 0.8484039902687073, "num_tokens": 102394495.0, "step": 2694 }, { "epoch": 0.3428317008014248, "ewc_loss": 0.08203125, "ewc_loss_diag": 5.602836608886719e-05, "ewc_loss_parallel": 7.62939453125e-05, "grad_norm": 5.4453911781311035, "learning_rate": 1e-05, "loss": 0.5549, "mean_token_accuracy": 0.8446904420852661, "num_tokens": 102431689.0, "step": 2695 }, { "epoch": 0.34295891108001525, "ewc_loss": 0.07421875, "ewc_loss_diag": 5.626678466796875e-05, "ewc_loss_parallel": 6.866455078125e-05, "grad_norm": 5.1155009269714355, "learning_rate": 1e-05, "loss": 0.5115, "mean_token_accuracy": 0.8591729998588562, "num_tokens": 102468653.0, "step": 2696 }, { "epoch": 0.3430861213586058, "ewc_loss": 0.0810546875, "ewc_loss_diag": 5.626678466796875e-05, "ewc_loss_parallel": 7.534027099609375e-05, "grad_norm": 5.494553089141846, "learning_rate": 1e-05, "loss": 0.483, "mean_token_accuracy": 0.8647079467773438, "num_tokens": 102504442.0, "step": 2697 }, { "epoch": 0.3432133316371963, "ewc_loss": 0.0712890625, "ewc_loss_diag": 5.626678466796875e-05, "ewc_loss_parallel": 6.580352783203125e-05, "grad_norm": 4.888792991638184, "learning_rate": 1e-05, "loss": 0.4997, "mean_token_accuracy": 0.8581492304801941, "num_tokens": 102539401.0, "step": 2698 }, { "epoch": 0.3433405419157868, "ewc_loss": 0.08203125, "ewc_loss_diag": 5.626678466796875e-05, "ewc_loss_parallel": 7.62939453125e-05, "grad_norm": 5.627585411071777, "learning_rate": 1e-05, "loss": 0.5152, "mean_token_accuracy": 0.8569836616516113, "num_tokens": 102579874.0, "step": 2699 }, { "epoch": 0.3434677521943773, "ewc_loss": 0.0703125, "ewc_loss_diag": 5.626678466796875e-05, "ewc_loss_parallel": 6.4849853515625e-05, "grad_norm": 4.760344505310059, "learning_rate": 1e-05, "loss": 0.546, "mean_token_accuracy": 0.8463121652603149, "num_tokens": 102617906.0, "step": 2700 }, { "epoch": 0.34359496247296784, "ewc_loss": 0.0849609375, "ewc_loss_diag": 5.626678466796875e-05, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.828197479248047, "learning_rate": 1e-05, "loss": 0.5246, "mean_token_accuracy": 0.8551679849624634, "num_tokens": 102657744.0, "step": 2701 }, { "epoch": 0.3437221727515583, "ewc_loss": 0.0703125, "ewc_loss_diag": 5.626678466796875e-05, "ewc_loss_parallel": 6.4849853515625e-05, "grad_norm": 4.741493225097656, "learning_rate": 1e-05, "loss": 0.5393, "mean_token_accuracy": 0.8458660840988159, "num_tokens": 102696650.0, "step": 2702 }, { "epoch": 0.34384938303014884, "ewc_loss": 0.08984375, "ewc_loss_diag": 5.626678466796875e-05, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 6.050156116485596, "learning_rate": 1e-05, "loss": 0.5374, "mean_token_accuracy": 0.8498371839523315, "num_tokens": 102729114.0, "step": 2703 }, { "epoch": 0.34397659330873936, "ewc_loss": 0.07421875, "ewc_loss_diag": 5.650520324707031e-05, "ewc_loss_parallel": 6.818771362304688e-05, "grad_norm": 4.877328872680664, "learning_rate": 1e-05, "loss": 0.5154, "mean_token_accuracy": 0.8510332703590393, "num_tokens": 102763562.0, "step": 2704 }, { "epoch": 0.34410380358732984, "ewc_loss": 0.09375, "ewc_loss_diag": 5.650520324707031e-05, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 6.1517863273620605, "learning_rate": 1e-05, "loss": 0.5644, "mean_token_accuracy": 0.8422407507896423, "num_tokens": 102803680.0, "step": 2705 }, { "epoch": 0.34423101386592037, "ewc_loss": 0.078125, "ewc_loss_diag": 5.650520324707031e-05, "ewc_loss_parallel": 7.2479248046875e-05, "grad_norm": 5.080244541168213, "learning_rate": 1e-05, "loss": 0.5759, "mean_token_accuracy": 0.8389833569526672, "num_tokens": 102847932.0, "step": 2706 }, { "epoch": 0.3443582241445109, "ewc_loss": 0.0908203125, "ewc_loss_diag": 5.650520324707031e-05, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 6.049511909484863, "learning_rate": 1e-05, "loss": 0.5254, "mean_token_accuracy": 0.8568789958953857, "num_tokens": 102888371.0, "step": 2707 }, { "epoch": 0.34448543442310137, "ewc_loss": 0.0810546875, "ewc_loss_diag": 5.650520324707031e-05, "ewc_loss_parallel": 7.534027099609375e-05, "grad_norm": 5.229473114013672, "learning_rate": 1e-05, "loss": 0.5843, "mean_token_accuracy": 0.8330217003822327, "num_tokens": 102928943.0, "step": 2708 }, { "epoch": 0.3446126447016919, "ewc_loss": 0.08349609375, "ewc_loss_diag": 5.6743621826171875e-05, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.820491790771484, "learning_rate": 1e-05, "loss": 0.5415, "mean_token_accuracy": 0.8500477075576782, "num_tokens": 102969773.0, "step": 2709 }, { "epoch": 0.3447398549802824, "ewc_loss": 0.080078125, "ewc_loss_diag": 5.6743621826171875e-05, "ewc_loss_parallel": 7.43865966796875e-05, "grad_norm": 5.261584281921387, "learning_rate": 1e-05, "loss": 0.5725, "mean_token_accuracy": 0.8385592699050903, "num_tokens": 103012496.0, "step": 2710 }, { "epoch": 0.3448670652588729, "ewc_loss": 0.0791015625, "ewc_loss_diag": 5.6743621826171875e-05, "ewc_loss_parallel": 7.343292236328125e-05, "grad_norm": 5.562841415405273, "learning_rate": 1e-05, "loss": 0.5532, "mean_token_accuracy": 0.8451958894729614, "num_tokens": 103047259.0, "step": 2711 }, { "epoch": 0.3449942755374634, "ewc_loss": 0.078125, "ewc_loss_diag": 5.698204040527344e-05, "ewc_loss_parallel": 7.2479248046875e-05, "grad_norm": 5.261958599090576, "learning_rate": 1e-05, "loss": 0.4966, "mean_token_accuracy": 0.8599710464477539, "num_tokens": 103087534.0, "step": 2712 }, { "epoch": 0.34512148581605395, "ewc_loss": 0.0771484375, "ewc_loss_diag": 5.698204040527344e-05, "ewc_loss_parallel": 7.104873657226562e-05, "grad_norm": 5.331103324890137, "learning_rate": 1e-05, "loss": 0.5083, "mean_token_accuracy": 0.8592151403427124, "num_tokens": 103128782.0, "step": 2713 }, { "epoch": 0.3452486960946444, "ewc_loss": 0.07666015625, "ewc_loss_diag": 5.698204040527344e-05, "ewc_loss_parallel": 7.05718994140625e-05, "grad_norm": 5.1980109214782715, "learning_rate": 1e-05, "loss": 0.5152, "mean_token_accuracy": 0.853482723236084, "num_tokens": 103161252.0, "step": 2714 }, { "epoch": 0.34537590637323495, "ewc_loss": 0.0751953125, "ewc_loss_diag": 5.698204040527344e-05, "ewc_loss_parallel": 6.914138793945312e-05, "grad_norm": 5.19597864151001, "learning_rate": 1e-05, "loss": 0.5095, "mean_token_accuracy": 0.8577629327774048, "num_tokens": 103201524.0, "step": 2715 }, { "epoch": 0.3455031166518255, "ewc_loss": 0.07470703125, "ewc_loss_diag": 5.698204040527344e-05, "ewc_loss_parallel": 6.866455078125e-05, "grad_norm": 5.139528274536133, "learning_rate": 1e-05, "loss": 0.545, "mean_token_accuracy": 0.8424498438835144, "num_tokens": 103235444.0, "step": 2716 }, { "epoch": 0.34563032693041595, "ewc_loss": 0.07470703125, "ewc_loss_diag": 5.698204040527344e-05, "ewc_loss_parallel": 6.866455078125e-05, "grad_norm": 5.142536163330078, "learning_rate": 1e-05, "loss": 0.5662, "mean_token_accuracy": 0.8366476893424988, "num_tokens": 103270927.0, "step": 2717 }, { "epoch": 0.3457575372090065, "ewc_loss": 0.07421875, "ewc_loss_diag": 5.698204040527344e-05, "ewc_loss_parallel": 6.818771362304688e-05, "grad_norm": 5.058379173278809, "learning_rate": 1e-05, "loss": 0.4976, "mean_token_accuracy": 0.8604908585548401, "num_tokens": 103310518.0, "step": 2718 }, { "epoch": 0.345884747487597, "ewc_loss": 0.07470703125, "ewc_loss_diag": 5.698204040527344e-05, "ewc_loss_parallel": 6.866455078125e-05, "grad_norm": 5.1404619216918945, "learning_rate": 1e-05, "loss": 0.4598, "mean_token_accuracy": 0.8708502650260925, "num_tokens": 103347940.0, "step": 2719 }, { "epoch": 0.3460119577661875, "ewc_loss": 0.072265625, "ewc_loss_diag": 5.698204040527344e-05, "ewc_loss_parallel": 6.628036499023438e-05, "grad_norm": 4.987015247344971, "learning_rate": 1e-05, "loss": 0.4886, "mean_token_accuracy": 0.8618274331092834, "num_tokens": 103383995.0, "step": 2720 }, { "epoch": 0.346139168044778, "ewc_loss": 0.07568359375, "ewc_loss_diag": 5.698204040527344e-05, "ewc_loss_parallel": 6.961822509765625e-05, "grad_norm": 5.181421279907227, "learning_rate": 1e-05, "loss": 0.4598, "mean_token_accuracy": 0.8691876530647278, "num_tokens": 103416411.0, "step": 2721 }, { "epoch": 0.34626637832336854, "ewc_loss": 0.0712890625, "ewc_loss_diag": 5.698204040527344e-05, "ewc_loss_parallel": 6.532669067382812e-05, "grad_norm": 4.904781818389893, "learning_rate": 1e-05, "loss": 0.5529, "mean_token_accuracy": 0.843380331993103, "num_tokens": 103454199.0, "step": 2722 }, { "epoch": 0.346393588601959, "ewc_loss": 0.07666015625, "ewc_loss_diag": 5.698204040527344e-05, "ewc_loss_parallel": 7.05718994140625e-05, "grad_norm": 5.28560733795166, "learning_rate": 1e-05, "loss": 0.5582, "mean_token_accuracy": 0.8408137559890747, "num_tokens": 103490996.0, "step": 2723 }, { "epoch": 0.34652079888054954, "ewc_loss": 0.0703125, "ewc_loss_diag": 5.698204040527344e-05, "ewc_loss_parallel": 6.437301635742188e-05, "grad_norm": 4.768623352050781, "learning_rate": 1e-05, "loss": 0.486, "mean_token_accuracy": 0.8569258451461792, "num_tokens": 103525987.0, "step": 2724 }, { "epoch": 0.34664800915914007, "ewc_loss": 0.07958984375, "ewc_loss_diag": 5.698204040527344e-05, "ewc_loss_parallel": 7.390975952148438e-05, "grad_norm": 5.48603630065918, "learning_rate": 1e-05, "loss": 0.5501, "mean_token_accuracy": 0.8393322229385376, "num_tokens": 103565702.0, "step": 2725 }, { "epoch": 0.34677521943773054, "ewc_loss": 0.0673828125, "ewc_loss_diag": 5.698204040527344e-05, "ewc_loss_parallel": 6.151199340820312e-05, "grad_norm": 4.6182541847229, "learning_rate": 1e-05, "loss": 0.5222, "mean_token_accuracy": 0.8517371416091919, "num_tokens": 103601764.0, "step": 2726 }, { "epoch": 0.34690242971632107, "ewc_loss": 0.083984375, "ewc_loss_diag": 5.7220458984375e-05, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.828150749206543, "learning_rate": 1e-05, "loss": 0.5439, "mean_token_accuracy": 0.8473209142684937, "num_tokens": 103640293.0, "step": 2727 }, { "epoch": 0.3470296399949116, "ewc_loss": 0.06787109375, "ewc_loss_diag": 5.7220458984375e-05, "ewc_loss_parallel": 6.198883056640625e-05, "grad_norm": 4.492809295654297, "learning_rate": 1e-05, "loss": 0.5199, "mean_token_accuracy": 0.8474881649017334, "num_tokens": 103680464.0, "step": 2728 }, { "epoch": 0.3471568502735021, "ewc_loss": 0.09765625, "ewc_loss_diag": 5.7220458984375e-05, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.364689350128174, "learning_rate": 1e-05, "loss": 0.5383, "mean_token_accuracy": 0.8520339131355286, "num_tokens": 103717434.0, "step": 2729 }, { "epoch": 0.3472840605520926, "ewc_loss": 0.07373046875, "ewc_loss_diag": 5.7220458984375e-05, "ewc_loss_parallel": 6.771087646484375e-05, "grad_norm": 4.730560302734375, "learning_rate": 1e-05, "loss": 0.5031, "mean_token_accuracy": 0.8536908626556396, "num_tokens": 103755186.0, "step": 2730 }, { "epoch": 0.3474112708306831, "ewc_loss": 0.1083984375, "ewc_loss_diag": 5.7220458984375e-05, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.795994758605957, "learning_rate": 1e-05, "loss": 0.5223, "mean_token_accuracy": 0.8651083707809448, "num_tokens": 103789516.0, "step": 2731 }, { "epoch": 0.34753848110927366, "ewc_loss": 0.08837890625, "ewc_loss_diag": 5.7220458984375e-05, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.446871757507324, "learning_rate": 1e-05, "loss": 0.556, "mean_token_accuracy": 0.8444661498069763, "num_tokens": 103826103.0, "step": 2732 }, { "epoch": 0.34766569138786413, "ewc_loss": 0.1015625, "ewc_loss_diag": 5.7220458984375e-05, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.556585788726807, "learning_rate": 1e-05, "loss": 0.5647, "mean_token_accuracy": 0.8517473936080933, "num_tokens": 103863059.0, "step": 2733 }, { "epoch": 0.34779290166645466, "ewc_loss": 0.09375, "ewc_loss_diag": 5.745887756347656e-05, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 5.768317222595215, "learning_rate": 1e-05, "loss": 0.5174, "mean_token_accuracy": 0.8571757078170776, "num_tokens": 103897976.0, "step": 2734 }, { "epoch": 0.3479201119450452, "ewc_loss": 0.08984375, "ewc_loss_diag": 5.745887756347656e-05, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 6.119631290435791, "learning_rate": 1e-05, "loss": 0.5947, "mean_token_accuracy": 0.8356576561927795, "num_tokens": 103935485.0, "step": 2735 }, { "epoch": 0.34804732222363566, "ewc_loss": 0.087890625, "ewc_loss_diag": 5.745887756347656e-05, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.625032424926758, "learning_rate": 1e-05, "loss": 0.5231, "mean_token_accuracy": 0.8545352816581726, "num_tokens": 103971917.0, "step": 2736 }, { "epoch": 0.3481745325022262, "ewc_loss": 0.08251953125, "ewc_loss_diag": 5.745887756347656e-05, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.801737308502197, "learning_rate": 1e-05, "loss": 0.5756, "mean_token_accuracy": 0.8433789014816284, "num_tokens": 104008797.0, "step": 2737 }, { "epoch": 0.3483017427808167, "ewc_loss": 0.08203125, "ewc_loss_diag": 5.745887756347656e-05, "ewc_loss_parallel": 7.62939453125e-05, "grad_norm": 5.419656753540039, "learning_rate": 1e-05, "loss": 0.5544, "mean_token_accuracy": 0.8437842726707458, "num_tokens": 104049551.0, "step": 2738 }, { "epoch": 0.3484289530594072, "ewc_loss": 0.0791015625, "ewc_loss_diag": 5.745887756347656e-05, "ewc_loss_parallel": 7.343292236328125e-05, "grad_norm": 5.584896087646484, "learning_rate": 1e-05, "loss": 0.5801, "mean_token_accuracy": 0.8341744542121887, "num_tokens": 104089341.0, "step": 2739 }, { "epoch": 0.3485561633379977, "ewc_loss": 0.076171875, "ewc_loss_diag": 5.745887756347656e-05, "ewc_loss_parallel": 7.009506225585938e-05, "grad_norm": 5.237857818603516, "learning_rate": 1e-05, "loss": 0.5451, "mean_token_accuracy": 0.8470677137374878, "num_tokens": 104133445.0, "step": 2740 }, { "epoch": 0.34868337361658824, "ewc_loss": 0.078125, "ewc_loss_diag": 5.745887756347656e-05, "ewc_loss_parallel": 7.2479248046875e-05, "grad_norm": 5.488961219787598, "learning_rate": 1e-05, "loss": 0.5203, "mean_token_accuracy": 0.852039098739624, "num_tokens": 104170435.0, "step": 2741 }, { "epoch": 0.3488105838951787, "ewc_loss": 0.07421875, "ewc_loss_diag": 5.745887756347656e-05, "ewc_loss_parallel": 6.818771362304688e-05, "grad_norm": 5.074525833129883, "learning_rate": 1e-05, "loss": 0.4836, "mean_token_accuracy": 0.8610979318618774, "num_tokens": 104205660.0, "step": 2742 }, { "epoch": 0.34893779417376924, "ewc_loss": 0.07861328125, "ewc_loss_diag": 5.745887756347656e-05, "ewc_loss_parallel": 7.295608520507812e-05, "grad_norm": 5.49728536605835, "learning_rate": 1e-05, "loss": 0.4925, "mean_token_accuracy": 0.8610225319862366, "num_tokens": 104243749.0, "step": 2743 }, { "epoch": 0.3490650044523598, "ewc_loss": 0.0712890625, "ewc_loss_diag": 5.745887756347656e-05, "ewc_loss_parallel": 6.532669067382812e-05, "grad_norm": 4.92227840423584, "learning_rate": 1e-05, "loss": 0.522, "mean_token_accuracy": 0.8508053421974182, "num_tokens": 104283411.0, "step": 2744 }, { "epoch": 0.34919221473095025, "ewc_loss": 0.0810546875, "ewc_loss_diag": 5.745887756347656e-05, "ewc_loss_parallel": 7.534027099609375e-05, "grad_norm": 5.624599456787109, "learning_rate": 1e-05, "loss": 0.5168, "mean_token_accuracy": 0.854010820388794, "num_tokens": 104324570.0, "step": 2745 }, { "epoch": 0.3493194250095408, "ewc_loss": 0.0703125, "ewc_loss_diag": 5.745887756347656e-05, "ewc_loss_parallel": 6.437301635742188e-05, "grad_norm": 4.772155284881592, "learning_rate": 1e-05, "loss": 0.5478, "mean_token_accuracy": 0.8459630012512207, "num_tokens": 104363619.0, "step": 2746 }, { "epoch": 0.3494466352881313, "ewc_loss": 0.08544921875, "ewc_loss_diag": 5.745887756347656e-05, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.887938976287842, "learning_rate": 1e-05, "loss": 0.5153, "mean_token_accuracy": 0.8558751344680786, "num_tokens": 104405572.0, "step": 2747 }, { "epoch": 0.3495738455667218, "ewc_loss": 0.0703125, "ewc_loss_diag": 5.745887756347656e-05, "ewc_loss_parallel": 6.437301635742188e-05, "grad_norm": 4.733586311340332, "learning_rate": 1e-05, "loss": 0.5315, "mean_token_accuracy": 0.845553994178772, "num_tokens": 104446772.0, "step": 2748 }, { "epoch": 0.3497010558453123, "ewc_loss": 0.09375, "ewc_loss_diag": 5.745887756347656e-05, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 6.210414409637451, "learning_rate": 1e-05, "loss": 0.5252, "mean_token_accuracy": 0.8559515476226807, "num_tokens": 104490912.0, "step": 2749 }, { "epoch": 0.34982826612390283, "ewc_loss": 0.0751953125, "ewc_loss_diag": 5.745887756347656e-05, "ewc_loss_parallel": 6.914138793945312e-05, "grad_norm": 4.958057403564453, "learning_rate": 1e-05, "loss": 0.4851, "mean_token_accuracy": 0.8589966893196106, "num_tokens": 104525307.0, "step": 2750 }, { "epoch": 0.3499554764024933, "ewc_loss": 0.0966796875, "ewc_loss_diag": 5.745887756347656e-05, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.333233833312988, "learning_rate": 1e-05, "loss": 0.5599, "mean_token_accuracy": 0.8465591073036194, "num_tokens": 104562431.0, "step": 2751 }, { "epoch": 0.35008268668108383, "ewc_loss": 0.0810546875, "ewc_loss_diag": 5.745887756347656e-05, "ewc_loss_parallel": 7.534027099609375e-05, "grad_norm": 5.332905292510986, "learning_rate": 1e-05, "loss": 0.4803, "mean_token_accuracy": 0.8676525354385376, "num_tokens": 104596440.0, "step": 2752 }, { "epoch": 0.35020989695967436, "ewc_loss": 0.087890625, "ewc_loss_diag": 5.745887756347656e-05, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 6.030650615692139, "learning_rate": 1e-05, "loss": 0.4878, "mean_token_accuracy": 0.8634253740310669, "num_tokens": 104630804.0, "step": 2753 }, { "epoch": 0.35033710723826483, "ewc_loss": 0.0830078125, "ewc_loss_diag": 5.745887756347656e-05, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.508151531219482, "learning_rate": 1e-05, "loss": 0.5099, "mean_token_accuracy": 0.8594796657562256, "num_tokens": 104667997.0, "step": 2754 }, { "epoch": 0.35046431751685536, "ewc_loss": 0.07763671875, "ewc_loss_diag": 5.745887756347656e-05, "ewc_loss_parallel": 7.200241088867188e-05, "grad_norm": 5.551297664642334, "learning_rate": 1e-05, "loss": 0.5061, "mean_token_accuracy": 0.8566581010818481, "num_tokens": 104711605.0, "step": 2755 }, { "epoch": 0.3505915277954459, "ewc_loss": 0.08251953125, "ewc_loss_diag": 5.745887756347656e-05, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.547182559967041, "learning_rate": 1e-05, "loss": 0.5494, "mean_token_accuracy": 0.846747875213623, "num_tokens": 104748427.0, "step": 2756 }, { "epoch": 0.35071873807403636, "ewc_loss": 0.07373046875, "ewc_loss_diag": 5.7697296142578125e-05, "ewc_loss_parallel": 6.771087646484375e-05, "grad_norm": 5.114014148712158, "learning_rate": 1e-05, "loss": 0.5063, "mean_token_accuracy": 0.8539265990257263, "num_tokens": 104784536.0, "step": 2757 }, { "epoch": 0.3508459483526269, "ewc_loss": 0.08349609375, "ewc_loss_diag": 5.7697296142578125e-05, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.662997245788574, "learning_rate": 1e-05, "loss": 0.5217, "mean_token_accuracy": 0.8548064827919006, "num_tokens": 104820096.0, "step": 2758 }, { "epoch": 0.3509731586312174, "ewc_loss": 0.07080078125, "ewc_loss_diag": 5.7697296142578125e-05, "ewc_loss_parallel": 6.4849853515625e-05, "grad_norm": 4.810655117034912, "learning_rate": 1e-05, "loss": 0.5031, "mean_token_accuracy": 0.8566762804985046, "num_tokens": 104857016.0, "step": 2759 }, { "epoch": 0.3511003689098079, "ewc_loss": 0.08642578125, "ewc_loss_diag": 5.7697296142578125e-05, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.91063117980957, "learning_rate": 1e-05, "loss": 0.5312, "mean_token_accuracy": 0.8530809879302979, "num_tokens": 104903999.0, "step": 2760 }, { "epoch": 0.3512275791883984, "ewc_loss": 0.06982421875, "ewc_loss_diag": 5.7697296142578125e-05, "ewc_loss_parallel": 6.389617919921875e-05, "grad_norm": 4.72723913192749, "learning_rate": 1e-05, "loss": 0.5892, "mean_token_accuracy": 0.833325982093811, "num_tokens": 104943954.0, "step": 2761 }, { "epoch": 0.35135478946698895, "ewc_loss": 0.09375, "ewc_loss_diag": 5.7697296142578125e-05, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 6.240049362182617, "learning_rate": 1e-05, "loss": 0.5559, "mean_token_accuracy": 0.8415882587432861, "num_tokens": 104980028.0, "step": 2762 }, { "epoch": 0.3514819997455794, "ewc_loss": 0.07470703125, "ewc_loss_diag": 5.7697296142578125e-05, "ewc_loss_parallel": 6.866455078125e-05, "grad_norm": 4.955952167510986, "learning_rate": 1e-05, "loss": 0.5113, "mean_token_accuracy": 0.8556849956512451, "num_tokens": 105020471.0, "step": 2763 }, { "epoch": 0.35160921002416995, "ewc_loss": 0.09716796875, "ewc_loss_diag": 5.7697296142578125e-05, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.357756614685059, "learning_rate": 1e-05, "loss": 0.5417, "mean_token_accuracy": 0.8535725474357605, "num_tokens": 105050419.0, "step": 2764 }, { "epoch": 0.3517364203027605, "ewc_loss": 0.080078125, "ewc_loss_diag": 5.7697296142578125e-05, "ewc_loss_parallel": 7.43865966796875e-05, "grad_norm": 5.312934875488281, "learning_rate": 1e-05, "loss": 0.4998, "mean_token_accuracy": 0.8609521389007568, "num_tokens": 105079656.0, "step": 2765 }, { "epoch": 0.35186363058135095, "ewc_loss": 0.08935546875, "ewc_loss_diag": 5.7697296142578125e-05, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 6.119334697723389, "learning_rate": 1e-05, "loss": 0.5191, "mean_token_accuracy": 0.8549665808677673, "num_tokens": 105119715.0, "step": 2766 }, { "epoch": 0.3519908408599415, "ewc_loss": 0.08203125, "ewc_loss_diag": 5.7697296142578125e-05, "ewc_loss_parallel": 7.62939453125e-05, "grad_norm": 5.426299095153809, "learning_rate": 1e-05, "loss": 0.5403, "mean_token_accuracy": 0.8446465134620667, "num_tokens": 105158281.0, "step": 2767 }, { "epoch": 0.352118051138532, "ewc_loss": 0.08203125, "ewc_loss_diag": 5.7697296142578125e-05, "ewc_loss_parallel": 7.62939453125e-05, "grad_norm": 5.799283981323242, "learning_rate": 1e-05, "loss": 0.5502, "mean_token_accuracy": 0.8445469737052917, "num_tokens": 105201372.0, "step": 2768 }, { "epoch": 0.3522452614171225, "ewc_loss": 0.0810546875, "ewc_loss_diag": 5.7697296142578125e-05, "ewc_loss_parallel": 7.534027099609375e-05, "grad_norm": 5.362181186676025, "learning_rate": 1e-05, "loss": 0.5224, "mean_token_accuracy": 0.8458458781242371, "num_tokens": 105240557.0, "step": 2769 }, { "epoch": 0.352372471695713, "ewc_loss": 0.078125, "ewc_loss_diag": 5.7697296142578125e-05, "ewc_loss_parallel": 7.2479248046875e-05, "grad_norm": 5.560641288757324, "learning_rate": 1e-05, "loss": 0.5468, "mean_token_accuracy": 0.8503308296203613, "num_tokens": 105276701.0, "step": 2770 }, { "epoch": 0.35249968197430354, "ewc_loss": 0.07666015625, "ewc_loss_diag": 5.7697296142578125e-05, "ewc_loss_parallel": 7.05718994140625e-05, "grad_norm": 5.2447919845581055, "learning_rate": 1e-05, "loss": 0.5836, "mean_token_accuracy": 0.8402271270751953, "num_tokens": 105319185.0, "step": 2771 }, { "epoch": 0.352626892252894, "ewc_loss": 0.0771484375, "ewc_loss_diag": 5.7697296142578125e-05, "ewc_loss_parallel": 7.104873657226562e-05, "grad_norm": 5.414052963256836, "learning_rate": 1e-05, "loss": 0.4693, "mean_token_accuracy": 0.8681995272636414, "num_tokens": 105362800.0, "step": 2772 }, { "epoch": 0.35275410253148454, "ewc_loss": 0.0751953125, "ewc_loss_diag": 5.7697296142578125e-05, "ewc_loss_parallel": 6.914138793945312e-05, "grad_norm": 5.112401008605957, "learning_rate": 1e-05, "loss": 0.5434, "mean_token_accuracy": 0.8464246988296509, "num_tokens": 105400731.0, "step": 2773 }, { "epoch": 0.35288131281007507, "ewc_loss": 0.0771484375, "ewc_loss_diag": 5.7697296142578125e-05, "ewc_loss_parallel": 7.104873657226562e-05, "grad_norm": 5.415116310119629, "learning_rate": 1e-05, "loss": 0.567, "mean_token_accuracy": 0.8437948226928711, "num_tokens": 105439712.0, "step": 2774 }, { "epoch": 0.35300852308866554, "ewc_loss": 0.072265625, "ewc_loss_diag": 5.7697296142578125e-05, "ewc_loss_parallel": 6.628036499023438e-05, "grad_norm": 4.966531276702881, "learning_rate": 1e-05, "loss": 0.5077, "mean_token_accuracy": 0.8559452295303345, "num_tokens": 105475159.0, "step": 2775 }, { "epoch": 0.35313573336725607, "ewc_loss": 0.07958984375, "ewc_loss_diag": 5.7697296142578125e-05, "ewc_loss_parallel": 7.390975952148438e-05, "grad_norm": 5.500690460205078, "learning_rate": 1e-05, "loss": 0.558, "mean_token_accuracy": 0.8408147096633911, "num_tokens": 105511323.0, "step": 2776 }, { "epoch": 0.3532629436458466, "ewc_loss": 0.06982421875, "ewc_loss_diag": 5.7697296142578125e-05, "ewc_loss_parallel": 6.389617919921875e-05, "grad_norm": 4.832907676696777, "learning_rate": 1e-05, "loss": 0.52, "mean_token_accuracy": 0.8533661365509033, "num_tokens": 105544056.0, "step": 2777 }, { "epoch": 0.35339015392443707, "ewc_loss": 0.08154296875, "ewc_loss_diag": 5.7697296142578125e-05, "ewc_loss_parallel": 7.581710815429688e-05, "grad_norm": 5.6776838302612305, "learning_rate": 1e-05, "loss": 0.4806, "mean_token_accuracy": 0.865404486656189, "num_tokens": 105582335.0, "step": 2778 }, { "epoch": 0.3535173642030276, "ewc_loss": 0.0703125, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 6.437301635742188e-05, "grad_norm": 4.729743957519531, "learning_rate": 1e-05, "loss": 0.5054, "mean_token_accuracy": 0.8581235408782959, "num_tokens": 105614046.0, "step": 2779 }, { "epoch": 0.3536445744816181, "ewc_loss": 0.0869140625, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.951905727386475, "learning_rate": 1e-05, "loss": 0.5092, "mean_token_accuracy": 0.8594608306884766, "num_tokens": 105648294.0, "step": 2780 }, { "epoch": 0.35377178476020865, "ewc_loss": 0.07177734375, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 6.580352783203125e-05, "grad_norm": 4.7734551429748535, "learning_rate": 1e-05, "loss": 0.5345, "mean_token_accuracy": 0.8478397727012634, "num_tokens": 105685707.0, "step": 2781 }, { "epoch": 0.3538989950387991, "ewc_loss": 0.0927734375, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 6.179721832275391, "learning_rate": 1e-05, "loss": 0.5435, "mean_token_accuracy": 0.8532083034515381, "num_tokens": 105724662.0, "step": 2782 }, { "epoch": 0.35402620531738965, "ewc_loss": 0.07666015625, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 7.05718994140625e-05, "grad_norm": 5.04028844833374, "learning_rate": 1e-05, "loss": 0.5798, "mean_token_accuracy": 0.8394588828086853, "num_tokens": 105756781.0, "step": 2783 }, { "epoch": 0.3541534155959802, "ewc_loss": 0.09228515625, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 6.132615566253662, "learning_rate": 1e-05, "loss": 0.5148, "mean_token_accuracy": 0.8599139451980591, "num_tokens": 105798466.0, "step": 2784 }, { "epoch": 0.35428062587457065, "ewc_loss": 0.0810546875, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 7.534027099609375e-05, "grad_norm": 5.300782680511475, "learning_rate": 1e-05, "loss": 0.5149, "mean_token_accuracy": 0.8559295535087585, "num_tokens": 105835085.0, "step": 2785 }, { "epoch": 0.3544078361531612, "ewc_loss": 0.08349609375, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.797616481781006, "learning_rate": 1e-05, "loss": 0.5016, "mean_token_accuracy": 0.8615067005157471, "num_tokens": 105872268.0, "step": 2786 }, { "epoch": 0.3545350464317517, "ewc_loss": 0.08154296875, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 7.581710815429688e-05, "grad_norm": 5.424607753753662, "learning_rate": 1e-05, "loss": 0.5275, "mean_token_accuracy": 0.8501082062721252, "num_tokens": 105905721.0, "step": 2787 }, { "epoch": 0.3546622567103422, "ewc_loss": 0.07666015625, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 7.05718994140625e-05, "grad_norm": 5.349212169647217, "learning_rate": 1e-05, "loss": 0.5771, "mean_token_accuracy": 0.8366043567657471, "num_tokens": 105945573.0, "step": 2788 }, { "epoch": 0.3547894669889327, "ewc_loss": 0.08251953125, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.514437675476074, "learning_rate": 1e-05, "loss": 0.5277, "mean_token_accuracy": 0.8497601747512817, "num_tokens": 105978800.0, "step": 2789 }, { "epoch": 0.35491667726752324, "ewc_loss": 0.072265625, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 6.628036499023438e-05, "grad_norm": 4.950821399688721, "learning_rate": 1e-05, "loss": 0.5233, "mean_token_accuracy": 0.8519471287727356, "num_tokens": 106014035.0, "step": 2790 }, { "epoch": 0.3550438875461137, "ewc_loss": 0.083984375, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.719675064086914, "learning_rate": 1e-05, "loss": 0.6325, "mean_token_accuracy": 0.8259836435317993, "num_tokens": 106050721.0, "step": 2791 }, { "epoch": 0.35517109782470424, "ewc_loss": 0.06982421875, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 6.389617919921875e-05, "grad_norm": 4.714105606079102, "learning_rate": 1e-05, "loss": 0.4966, "mean_token_accuracy": 0.8546609878540039, "num_tokens": 106086802.0, "step": 2792 }, { "epoch": 0.35529830810329477, "ewc_loss": 0.08935546875, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 6.0441484451293945, "learning_rate": 1e-05, "loss": 0.4977, "mean_token_accuracy": 0.8625864386558533, "num_tokens": 106124174.0, "step": 2793 }, { "epoch": 0.35542551838188524, "ewc_loss": 0.07177734375, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 6.580352783203125e-05, "grad_norm": 4.730500221252441, "learning_rate": 1e-05, "loss": 0.5051, "mean_token_accuracy": 0.8556885719299316, "num_tokens": 106166147.0, "step": 2794 }, { "epoch": 0.35555272866047577, "ewc_loss": 0.09765625, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.3704328536987305, "learning_rate": 1e-05, "loss": 0.5724, "mean_token_accuracy": 0.843595027923584, "num_tokens": 106206758.0, "step": 2795 }, { "epoch": 0.3556799389390663, "ewc_loss": 0.07861328125, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 7.295608520507812e-05, "grad_norm": 5.090862274169922, "learning_rate": 1e-05, "loss": 0.5371, "mean_token_accuracy": 0.8484396934509277, "num_tokens": 106241823.0, "step": 2796 }, { "epoch": 0.35580714921765677, "ewc_loss": 0.09814453125, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.346817970275879, "learning_rate": 1e-05, "loss": 0.5149, "mean_token_accuracy": 0.8630318641662598, "num_tokens": 106275716.0, "step": 2797 }, { "epoch": 0.3559343594962473, "ewc_loss": 0.083984375, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.416138648986816, "learning_rate": 1e-05, "loss": 0.5628, "mean_token_accuracy": 0.842477023601532, "num_tokens": 106312595.0, "step": 2798 }, { "epoch": 0.3560615697748378, "ewc_loss": 0.08837890625, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 6.038887977600098, "learning_rate": 1e-05, "loss": 0.5691, "mean_token_accuracy": 0.8433015942573547, "num_tokens": 106346009.0, "step": 2799 }, { "epoch": 0.3561887800534283, "ewc_loss": 0.08251953125, "ewc_loss_diag": 5.841255187988281e-05, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.447916507720947, "learning_rate": 1e-05, "loss": 0.4951, "mean_token_accuracy": 0.862476110458374, "num_tokens": 106384501.0, "step": 2800 }, { "epoch": 0.35631599033201883, "ewc_loss": 0.080078125, "ewc_loss_diag": 5.8650970458984375e-05, "ewc_loss_parallel": 7.43865966796875e-05, "grad_norm": 5.712851524353027, "learning_rate": 1e-05, "loss": 0.5323, "mean_token_accuracy": 0.8536757826805115, "num_tokens": 106420099.0, "step": 2801 }, { "epoch": 0.35644320061060936, "ewc_loss": 0.080078125, "ewc_loss_diag": 5.8650970458984375e-05, "ewc_loss_parallel": 7.43865966796875e-05, "grad_norm": 5.354370594024658, "learning_rate": 1e-05, "loss": 0.4899, "mean_token_accuracy": 0.8656041622161865, "num_tokens": 106459697.0, "step": 2802 }, { "epoch": 0.35657041088919983, "ewc_loss": 0.0771484375, "ewc_loss_diag": 5.8650970458984375e-05, "ewc_loss_parallel": 7.152557373046875e-05, "grad_norm": 5.493304252624512, "learning_rate": 1e-05, "loss": 0.6305, "mean_token_accuracy": 0.8268665075302124, "num_tokens": 106499054.0, "step": 2803 }, { "epoch": 0.35669762116779036, "ewc_loss": 0.07666015625, "ewc_loss_diag": 5.8650970458984375e-05, "ewc_loss_parallel": 7.05718994140625e-05, "grad_norm": 5.229574680328369, "learning_rate": 1e-05, "loss": 0.5136, "mean_token_accuracy": 0.8545999526977539, "num_tokens": 106538704.0, "step": 2804 }, { "epoch": 0.3568248314463809, "ewc_loss": 0.07763671875, "ewc_loss_diag": 5.8650970458984375e-05, "ewc_loss_parallel": 7.200241088867188e-05, "grad_norm": 5.396606922149658, "learning_rate": 1e-05, "loss": 0.5545, "mean_token_accuracy": 0.8418446779251099, "num_tokens": 106580012.0, "step": 2805 }, { "epoch": 0.35695204172497136, "ewc_loss": 0.07421875, "ewc_loss_diag": 5.8650970458984375e-05, "ewc_loss_parallel": 6.818771362304688e-05, "grad_norm": 5.089089870452881, "learning_rate": 1e-05, "loss": 0.5094, "mean_token_accuracy": 0.852670431137085, "num_tokens": 106618160.0, "step": 2806 }, { "epoch": 0.3570792520035619, "ewc_loss": 0.078125, "ewc_loss_diag": 5.8650970458984375e-05, "ewc_loss_parallel": 7.2479248046875e-05, "grad_norm": 5.412130832672119, "learning_rate": 1e-05, "loss": 0.6037, "mean_token_accuracy": 0.827414870262146, "num_tokens": 106654387.0, "step": 2807 }, { "epoch": 0.3572064622821524, "ewc_loss": 0.072265625, "ewc_loss_diag": 5.8650970458984375e-05, "ewc_loss_parallel": 6.628036499023438e-05, "grad_norm": 4.945173263549805, "learning_rate": 1e-05, "loss": 0.4824, "mean_token_accuracy": 0.8624200820922852, "num_tokens": 106693318.0, "step": 2808 }, { "epoch": 0.3573336725607429, "ewc_loss": 0.07861328125, "ewc_loss_diag": 5.8650970458984375e-05, "ewc_loss_parallel": 7.295608520507812e-05, "grad_norm": 5.5202836990356445, "learning_rate": 1e-05, "loss": 0.5417, "mean_token_accuracy": 0.8437777161598206, "num_tokens": 106728779.0, "step": 2809 }, { "epoch": 0.3574608828393334, "ewc_loss": 0.07080078125, "ewc_loss_diag": 5.8650970458984375e-05, "ewc_loss_parallel": 6.4849853515625e-05, "grad_norm": 4.806110382080078, "learning_rate": 1e-05, "loss": 0.5221, "mean_token_accuracy": 0.8513681888580322, "num_tokens": 106771149.0, "step": 2810 }, { "epoch": 0.35758809311792394, "ewc_loss": 0.08251953125, "ewc_loss_diag": 5.888938903808594e-05, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.7369160652160645, "learning_rate": 1e-05, "loss": 0.5732, "mean_token_accuracy": 0.8398929834365845, "num_tokens": 106809132.0, "step": 2811 }, { "epoch": 0.3577153033965144, "ewc_loss": 0.0703125, "ewc_loss_diag": 5.888938903808594e-05, "ewc_loss_parallel": 6.437301635742188e-05, "grad_norm": 4.728855609893799, "learning_rate": 1e-05, "loss": 0.4539, "mean_token_accuracy": 0.8696610927581787, "num_tokens": 106842747.0, "step": 2812 }, { "epoch": 0.35784251367510495, "ewc_loss": 0.0888671875, "ewc_loss_diag": 5.888938903808594e-05, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 6.013956546783447, "learning_rate": 1e-05, "loss": 0.5805, "mean_token_accuracy": 0.8384699821472168, "num_tokens": 106883519.0, "step": 2813 }, { "epoch": 0.3579697239536955, "ewc_loss": 0.072265625, "ewc_loss_diag": 5.888938903808594e-05, "ewc_loss_parallel": 6.628036499023438e-05, "grad_norm": 4.80853271484375, "learning_rate": 1e-05, "loss": 0.5369, "mean_token_accuracy": 0.8393523097038269, "num_tokens": 106914818.0, "step": 2814 }, { "epoch": 0.35809693423228595, "ewc_loss": 0.0947265625, "ewc_loss_diag": 5.888938903808594e-05, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.220739841461182, "learning_rate": 1e-05, "loss": 0.5212, "mean_token_accuracy": 0.8615075349807739, "num_tokens": 106954387.0, "step": 2815 }, { "epoch": 0.3582241445108765, "ewc_loss": 0.0771484375, "ewc_loss_diag": 5.888938903808594e-05, "ewc_loss_parallel": 7.104873657226562e-05, "grad_norm": 5.081727981567383, "learning_rate": 1e-05, "loss": 0.5549, "mean_token_accuracy": 0.8427920341491699, "num_tokens": 106994264.0, "step": 2816 }, { "epoch": 0.358351354789467, "ewc_loss": 0.09326171875, "ewc_loss_diag": 5.888938903808594e-05, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 6.1379852294921875, "learning_rate": 1e-05, "loss": 0.5173, "mean_token_accuracy": 0.8578554391860962, "num_tokens": 107035575.0, "step": 2817 }, { "epoch": 0.3584785650680575, "ewc_loss": 0.08203125, "ewc_loss_diag": 5.888938903808594e-05, "ewc_loss_parallel": 7.62939453125e-05, "grad_norm": 5.353880405426025, "learning_rate": 1e-05, "loss": 0.5642, "mean_token_accuracy": 0.8423362374305725, "num_tokens": 107074644.0, "step": 2818 }, { "epoch": 0.358605775346648, "ewc_loss": 0.08251953125, "ewc_loss_diag": 5.91278076171875e-05, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.771940231323242, "learning_rate": 1e-05, "loss": 0.515, "mean_token_accuracy": 0.8549836277961731, "num_tokens": 107108758.0, "step": 2819 }, { "epoch": 0.35873298562523853, "ewc_loss": 0.08251953125, "ewc_loss_diag": 5.91278076171875e-05, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.436624526977539, "learning_rate": 1e-05, "loss": 0.4688, "mean_token_accuracy": 0.8678759932518005, "num_tokens": 107147572.0, "step": 2820 }, { "epoch": 0.358860195903829, "ewc_loss": 0.076171875, "ewc_loss_diag": 5.91278076171875e-05, "ewc_loss_parallel": 7.009506225585938e-05, "grad_norm": 5.353113174438477, "learning_rate": 1e-05, "loss": 0.5508, "mean_token_accuracy": 0.843292236328125, "num_tokens": 107184928.0, "step": 2821 }, { "epoch": 0.35898740618241953, "ewc_loss": 0.0830078125, "ewc_loss_diag": 5.91278076171875e-05, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.5036468505859375, "learning_rate": 1e-05, "loss": 0.4964, "mean_token_accuracy": 0.8643611669540405, "num_tokens": 107223462.0, "step": 2822 }, { "epoch": 0.35911461646101006, "ewc_loss": 0.072265625, "ewc_loss_diag": 5.91278076171875e-05, "ewc_loss_parallel": 6.628036499023438e-05, "grad_norm": 4.980498790740967, "learning_rate": 1e-05, "loss": 0.5059, "mean_token_accuracy": 0.8551403284072876, "num_tokens": 107263114.0, "step": 2823 }, { "epoch": 0.35924182673960053, "ewc_loss": 0.0830078125, "ewc_loss_diag": 5.91278076171875e-05, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.681044101715088, "learning_rate": 1e-05, "loss": 0.4663, "mean_token_accuracy": 0.8713675737380981, "num_tokens": 107301234.0, "step": 2824 }, { "epoch": 0.35936903701819106, "ewc_loss": 0.06982421875, "ewc_loss_diag": 5.91278076171875e-05, "ewc_loss_parallel": 6.389617919921875e-05, "grad_norm": 4.754488468170166, "learning_rate": 1e-05, "loss": 0.5081, "mean_token_accuracy": 0.8549966216087341, "num_tokens": 107334903.0, "step": 2825 }, { "epoch": 0.3594962472967816, "ewc_loss": 0.0869140625, "ewc_loss_diag": 5.91278076171875e-05, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.961636066436768, "learning_rate": 1e-05, "loss": 0.4618, "mean_token_accuracy": 0.8712230920791626, "num_tokens": 107374989.0, "step": 2826 }, { "epoch": 0.35962345757537206, "ewc_loss": 0.07080078125, "ewc_loss_diag": 5.91278076171875e-05, "ewc_loss_parallel": 6.4849853515625e-05, "grad_norm": 4.727485656738281, "learning_rate": 1e-05, "loss": 0.4787, "mean_token_accuracy": 0.8616197109222412, "num_tokens": 107412545.0, "step": 2827 }, { "epoch": 0.3597506678539626, "ewc_loss": 0.09521484375, "ewc_loss_diag": 5.91278076171875e-05, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 6.29503870010376, "learning_rate": 1e-05, "loss": 0.6029, "mean_token_accuracy": 0.8330301642417908, "num_tokens": 107447642.0, "step": 2828 }, { "epoch": 0.3598778781325531, "ewc_loss": 0.07568359375, "ewc_loss_diag": 5.91278076171875e-05, "ewc_loss_parallel": 6.961822509765625e-05, "grad_norm": 4.991615295410156, "learning_rate": 1e-05, "loss": 0.5125, "mean_token_accuracy": 0.8545927405357361, "num_tokens": 107486514.0, "step": 2829 }, { "epoch": 0.36000508841114365, "ewc_loss": 0.09716796875, "ewc_loss_diag": 5.91278076171875e-05, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.365394592285156, "learning_rate": 1e-05, "loss": 0.5555, "mean_token_accuracy": 0.8451440334320068, "num_tokens": 107527351.0, "step": 2830 }, { "epoch": 0.3601322986897341, "ewc_loss": 0.08203125, "ewc_loss_diag": 5.91278076171875e-05, "ewc_loss_parallel": 7.62939453125e-05, "grad_norm": 5.332038879394531, "learning_rate": 1e-05, "loss": 0.5176, "mean_token_accuracy": 0.8550477027893066, "num_tokens": 107558973.0, "step": 2831 }, { "epoch": 0.36025950896832465, "ewc_loss": 0.08984375, "ewc_loss_diag": 5.91278076171875e-05, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 6.123923301696777, "learning_rate": 1e-05, "loss": 0.5511, "mean_token_accuracy": 0.8483125567436218, "num_tokens": 107597659.0, "step": 2832 }, { "epoch": 0.3603867192469152, "ewc_loss": 0.08251953125, "ewc_loss_diag": 5.91278076171875e-05, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.43159294128418, "learning_rate": 1e-05, "loss": 0.5416, "mean_token_accuracy": 0.8489803075790405, "num_tokens": 107635833.0, "step": 2833 }, { "epoch": 0.36051392952550565, "ewc_loss": 0.08251953125, "ewc_loss_diag": 5.91278076171875e-05, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.798154354095459, "learning_rate": 1e-05, "loss": 0.5, "mean_token_accuracy": 0.8604351878166199, "num_tokens": 107673347.0, "step": 2834 }, { "epoch": 0.3606411398040962, "ewc_loss": 0.08056640625, "ewc_loss_diag": 5.91278076171875e-05, "ewc_loss_parallel": 7.486343383789062e-05, "grad_norm": 5.375969886779785, "learning_rate": 1e-05, "loss": 0.538, "mean_token_accuracy": 0.8490864038467407, "num_tokens": 107713820.0, "step": 2835 }, { "epoch": 0.3607683500826867, "ewc_loss": 0.078125, "ewc_loss_diag": 5.91278076171875e-05, "ewc_loss_parallel": 7.2479248046875e-05, "grad_norm": 5.543831825256348, "learning_rate": 1e-05, "loss": 0.5052, "mean_token_accuracy": 0.8569282293319702, "num_tokens": 107745586.0, "step": 2836 }, { "epoch": 0.3608955603612772, "ewc_loss": 0.076171875, "ewc_loss_diag": 5.91278076171875e-05, "ewc_loss_parallel": 7.009506225585938e-05, "grad_norm": 5.26263427734375, "learning_rate": 1e-05, "loss": 0.4796, "mean_token_accuracy": 0.8645621538162231, "num_tokens": 107787649.0, "step": 2837 }, { "epoch": 0.3610227706398677, "ewc_loss": 0.0771484375, "ewc_loss_diag": 5.91278076171875e-05, "ewc_loss_parallel": 7.152557373046875e-05, "grad_norm": 5.40228796005249, "learning_rate": 1e-05, "loss": 0.5292, "mean_token_accuracy": 0.8480600118637085, "num_tokens": 107831545.0, "step": 2838 }, { "epoch": 0.36114998091845824, "ewc_loss": 0.0751953125, "ewc_loss_diag": 5.91278076171875e-05, "ewc_loss_parallel": 6.914138793945312e-05, "grad_norm": 5.139602184295654, "learning_rate": 1e-05, "loss": 0.5738, "mean_token_accuracy": 0.837613046169281, "num_tokens": 107875252.0, "step": 2839 }, { "epoch": 0.3612771911970487, "ewc_loss": 0.0771484375, "ewc_loss_diag": 5.91278076171875e-05, "ewc_loss_parallel": 7.104873657226562e-05, "grad_norm": 5.388752460479736, "learning_rate": 1e-05, "loss": 0.5947, "mean_token_accuracy": 0.8310612440109253, "num_tokens": 107912060.0, "step": 2840 }, { "epoch": 0.36140440147563924, "ewc_loss": 0.072265625, "ewc_loss_diag": 5.91278076171875e-05, "ewc_loss_parallel": 6.628036499023438e-05, "grad_norm": 5.008489608764648, "learning_rate": 1e-05, "loss": 0.5346, "mean_token_accuracy": 0.8468798995018005, "num_tokens": 107949778.0, "step": 2841 }, { "epoch": 0.36153161175422976, "ewc_loss": 0.07763671875, "ewc_loss_diag": 5.91278076171875e-05, "ewc_loss_parallel": 7.200241088867188e-05, "grad_norm": 5.443769454956055, "learning_rate": 1e-05, "loss": 0.5277, "mean_token_accuracy": 0.8508086204528809, "num_tokens": 107991160.0, "step": 2842 }, { "epoch": 0.36165882203282024, "ewc_loss": 0.0712890625, "ewc_loss_diag": 5.91278076171875e-05, "ewc_loss_parallel": 6.532669067382812e-05, "grad_norm": 4.88269567489624, "learning_rate": 1e-05, "loss": 0.5528, "mean_token_accuracy": 0.8389520645141602, "num_tokens": 108027059.0, "step": 2843 }, { "epoch": 0.36178603231141077, "ewc_loss": 0.08056640625, "ewc_loss_diag": 5.91278076171875e-05, "ewc_loss_parallel": 7.486343383789062e-05, "grad_norm": 5.589445114135742, "learning_rate": 1e-05, "loss": 0.5434, "mean_token_accuracy": 0.8486934304237366, "num_tokens": 108066738.0, "step": 2844 }, { "epoch": 0.3619132425900013, "ewc_loss": 0.06982421875, "ewc_loss_diag": 5.91278076171875e-05, "ewc_loss_parallel": 6.389617919921875e-05, "grad_norm": 4.7603631019592285, "learning_rate": 1e-05, "loss": 0.5483, "mean_token_accuracy": 0.839790940284729, "num_tokens": 108106498.0, "step": 2845 }, { "epoch": 0.36204045286859177, "ewc_loss": 0.0849609375, "ewc_loss_diag": 5.91278076171875e-05, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.849159240722656, "learning_rate": 1e-05, "loss": 0.5362, "mean_token_accuracy": 0.8491054773330688, "num_tokens": 108139146.0, "step": 2846 }, { "epoch": 0.3621676631471823, "ewc_loss": 0.0703125, "ewc_loss_diag": 5.91278076171875e-05, "ewc_loss_parallel": 6.437301635742188e-05, "grad_norm": 4.72050142288208, "learning_rate": 1e-05, "loss": 0.458, "mean_token_accuracy": 0.8660168647766113, "num_tokens": 108176070.0, "step": 2847 }, { "epoch": 0.3622948734257728, "ewc_loss": 0.0927734375, "ewc_loss_diag": 5.91278076171875e-05, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 6.141555309295654, "learning_rate": 1e-05, "loss": 0.5247, "mean_token_accuracy": 0.8586481809616089, "num_tokens": 108214859.0, "step": 2848 }, { "epoch": 0.3624220837043633, "ewc_loss": 0.07421875, "ewc_loss_diag": 5.936622619628906e-05, "ewc_loss_parallel": 6.818771362304688e-05, "grad_norm": 4.900849342346191, "learning_rate": 1e-05, "loss": 0.5264, "mean_token_accuracy": 0.8510248064994812, "num_tokens": 108250201.0, "step": 2849 }, { "epoch": 0.3625492939829538, "ewc_loss": 0.095703125, "ewc_loss_diag": 5.91278076171875e-05, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.274016857147217, "learning_rate": 1e-05, "loss": 0.5204, "mean_token_accuracy": 0.8566874861717224, "num_tokens": 108282111.0, "step": 2850 }, { "epoch": 0.36267650426154435, "ewc_loss": 0.0791015625, "ewc_loss_diag": 5.936622619628906e-05, "ewc_loss_parallel": 7.343292236328125e-05, "grad_norm": 5.2294511795043945, "learning_rate": 1e-05, "loss": 0.5649, "mean_token_accuracy": 0.8366450071334839, "num_tokens": 108317873.0, "step": 2851 }, { "epoch": 0.3628037145401348, "ewc_loss": 0.08837890625, "ewc_loss_diag": 5.936622619628906e-05, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 6.056144714355469, "learning_rate": 1e-05, "loss": 0.5047, "mean_token_accuracy": 0.8654094934463501, "num_tokens": 108352874.0, "step": 2852 }, { "epoch": 0.36293092481872535, "ewc_loss": 0.0830078125, "ewc_loss_diag": 5.936622619628906e-05, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.416555881500244, "learning_rate": 1e-05, "loss": 0.5589, "mean_token_accuracy": 0.8427602052688599, "num_tokens": 108387620.0, "step": 2853 }, { "epoch": 0.3630581350973159, "ewc_loss": 0.080078125, "ewc_loss_diag": 5.936622619628906e-05, "ewc_loss_parallel": 7.43865966796875e-05, "grad_norm": 5.6391921043396, "learning_rate": 1e-05, "loss": 0.5721, "mean_token_accuracy": 0.8371595740318298, "num_tokens": 108428961.0, "step": 2854 }, { "epoch": 0.36318534537590635, "ewc_loss": 0.08251953125, "ewc_loss_diag": 5.936622619628906e-05, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.483556747436523, "learning_rate": 1e-05, "loss": 0.615, "mean_token_accuracy": 0.8295421600341797, "num_tokens": 108471505.0, "step": 2855 }, { "epoch": 0.3633125556544969, "ewc_loss": 0.0751953125, "ewc_loss_diag": 5.936622619628906e-05, "ewc_loss_parallel": 6.914138793945312e-05, "grad_norm": 5.199094772338867, "learning_rate": 1e-05, "loss": 0.5223, "mean_token_accuracy": 0.8503137826919556, "num_tokens": 108515874.0, "step": 2856 }, { "epoch": 0.3634397659330874, "ewc_loss": 0.08349609375, "ewc_loss_diag": 5.936622619628906e-05, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.587830066680908, "learning_rate": 1e-05, "loss": 0.5315, "mean_token_accuracy": 0.853367030620575, "num_tokens": 108558302.0, "step": 2857 }, { "epoch": 0.3635669762116779, "ewc_loss": 0.07177734375, "ewc_loss_diag": 5.936622619628906e-05, "ewc_loss_parallel": 6.580352783203125e-05, "grad_norm": 4.884439468383789, "learning_rate": 1e-05, "loss": 0.5421, "mean_token_accuracy": 0.8437860012054443, "num_tokens": 108597668.0, "step": 2858 }, { "epoch": 0.3636941864902684, "ewc_loss": 0.08544921875, "ewc_loss_diag": 5.936622619628906e-05, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.799605846405029, "learning_rate": 1e-05, "loss": 0.5354, "mean_token_accuracy": 0.8497064113616943, "num_tokens": 108636623.0, "step": 2859 }, { "epoch": 0.36382139676885894, "ewc_loss": 0.06982421875, "ewc_loss_diag": 5.936622619628906e-05, "ewc_loss_parallel": 6.389617919921875e-05, "grad_norm": 4.739503860473633, "learning_rate": 1e-05, "loss": 0.4405, "mean_token_accuracy": 0.8713374733924866, "num_tokens": 108673660.0, "step": 2860 }, { "epoch": 0.3639486070474494, "ewc_loss": 0.09130859375, "ewc_loss_diag": 5.936622619628906e-05, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 6.102207183837891, "learning_rate": 1e-05, "loss": 0.5118, "mean_token_accuracy": 0.8631714582443237, "num_tokens": 108710350.0, "step": 2861 }, { "epoch": 0.36407581732603994, "ewc_loss": 0.07275390625, "ewc_loss_diag": 5.936622619628906e-05, "ewc_loss_parallel": 6.67572021484375e-05, "grad_norm": 4.846084117889404, "learning_rate": 1e-05, "loss": 0.5374, "mean_token_accuracy": 0.8457037210464478, "num_tokens": 108747028.0, "step": 2862 }, { "epoch": 0.36420302760463047, "ewc_loss": 0.09619140625, "ewc_loss_diag": 5.936622619628906e-05, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.30694055557251, "learning_rate": 1e-05, "loss": 0.5867, "mean_token_accuracy": 0.8406835198402405, "num_tokens": 108780315.0, "step": 2863 }, { "epoch": 0.36433023788322094, "ewc_loss": 0.0791015625, "ewc_loss_diag": 5.936622619628906e-05, "ewc_loss_parallel": 7.343292236328125e-05, "grad_norm": 5.142910957336426, "learning_rate": 1e-05, "loss": 0.5416, "mean_token_accuracy": 0.8416726589202881, "num_tokens": 108811706.0, "step": 2864 }, { "epoch": 0.36445744816181147, "ewc_loss": 0.09375, "ewc_loss_diag": 5.936622619628906e-05, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 6.230249404907227, "learning_rate": 1e-05, "loss": 0.5825, "mean_token_accuracy": 0.8420257568359375, "num_tokens": 108851545.0, "step": 2865 }, { "epoch": 0.364584658440402, "ewc_loss": 0.08203125, "ewc_loss_diag": 5.936622619628906e-05, "ewc_loss_parallel": 7.62939453125e-05, "grad_norm": 5.340895175933838, "learning_rate": 1e-05, "loss": 0.4852, "mean_token_accuracy": 0.8643836379051208, "num_tokens": 108887013.0, "step": 2866 }, { "epoch": 0.36471186871899247, "ewc_loss": 0.08642578125, "ewc_loss_diag": 5.936622619628906e-05, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.9615159034729, "learning_rate": 1e-05, "loss": 0.5665, "mean_token_accuracy": 0.8408487439155579, "num_tokens": 108934289.0, "step": 2867 }, { "epoch": 0.364839078997583, "ewc_loss": 0.0810546875, "ewc_loss_diag": 6.008148193359375e-05, "ewc_loss_parallel": 7.534027099609375e-05, "grad_norm": 5.365126132965088, "learning_rate": 1e-05, "loss": 0.4987, "mean_token_accuracy": 0.8607252836227417, "num_tokens": 108973823.0, "step": 2868 }, { "epoch": 0.36496628927617353, "ewc_loss": 0.07958984375, "ewc_loss_diag": 6.008148193359375e-05, "ewc_loss_parallel": 7.390975952148438e-05, "grad_norm": 5.700164318084717, "learning_rate": 1e-05, "loss": 0.5004, "mean_token_accuracy": 0.8593326807022095, "num_tokens": 109006947.0, "step": 2869 }, { "epoch": 0.365093499554764, "ewc_loss": 0.07958984375, "ewc_loss_diag": 6.008148193359375e-05, "ewc_loss_parallel": 7.390975952148438e-05, "grad_norm": 5.306578636169434, "learning_rate": 1e-05, "loss": 0.5644, "mean_token_accuracy": 0.8397603034973145, "num_tokens": 109039839.0, "step": 2870 }, { "epoch": 0.36522070983335453, "ewc_loss": 0.078125, "ewc_loss_diag": 6.008148193359375e-05, "ewc_loss_parallel": 7.2479248046875e-05, "grad_norm": 5.501262187957764, "learning_rate": 1e-05, "loss": 0.4964, "mean_token_accuracy": 0.8613373041152954, "num_tokens": 109076348.0, "step": 2871 }, { "epoch": 0.36534792011194506, "ewc_loss": 0.076171875, "ewc_loss_diag": 6.008148193359375e-05, "ewc_loss_parallel": 7.009506225585938e-05, "grad_norm": 5.193161487579346, "learning_rate": 1e-05, "loss": 0.4782, "mean_token_accuracy": 0.8661242127418518, "num_tokens": 109117050.0, "step": 2872 }, { "epoch": 0.36547513039053553, "ewc_loss": 0.07763671875, "ewc_loss_diag": 6.008148193359375e-05, "ewc_loss_parallel": 7.200241088867188e-05, "grad_norm": 5.432055473327637, "learning_rate": 1e-05, "loss": 0.5322, "mean_token_accuracy": 0.8491609692573547, "num_tokens": 109151131.0, "step": 2873 }, { "epoch": 0.36560234066912606, "ewc_loss": 0.07421875, "ewc_loss_diag": 6.008148193359375e-05, "ewc_loss_parallel": 6.818771362304688e-05, "grad_norm": 5.058652877807617, "learning_rate": 1e-05, "loss": 0.5399, "mean_token_accuracy": 0.8473902940750122, "num_tokens": 109194542.0, "step": 2874 }, { "epoch": 0.3657295509477166, "ewc_loss": 0.07861328125, "ewc_loss_diag": 6.008148193359375e-05, "ewc_loss_parallel": 7.295608520507812e-05, "grad_norm": 5.469381332397461, "learning_rate": 1e-05, "loss": 0.5226, "mean_token_accuracy": 0.8498685359954834, "num_tokens": 109229636.0, "step": 2875 }, { "epoch": 0.36585676122630706, "ewc_loss": 0.07177734375, "ewc_loss_diag": 6.008148193359375e-05, "ewc_loss_parallel": 6.580352783203125e-05, "grad_norm": 4.934997081756592, "learning_rate": 1e-05, "loss": 0.541, "mean_token_accuracy": 0.844475269317627, "num_tokens": 109264705.0, "step": 2876 }, { "epoch": 0.3659839715048976, "ewc_loss": 0.08056640625, "ewc_loss_diag": 6.008148193359375e-05, "ewc_loss_parallel": 7.486343383789062e-05, "grad_norm": 5.570679187774658, "learning_rate": 1e-05, "loss": 0.5056, "mean_token_accuracy": 0.8601976633071899, "num_tokens": 109304220.0, "step": 2877 }, { "epoch": 0.3661111817834881, "ewc_loss": 0.0712890625, "ewc_loss_diag": 6.008148193359375e-05, "ewc_loss_parallel": 6.532669067382812e-05, "grad_norm": 4.822048664093018, "learning_rate": 1e-05, "loss": 0.5345, "mean_token_accuracy": 0.8505829572677612, "num_tokens": 109347315.0, "step": 2878 }, { "epoch": 0.3662383920620786, "ewc_loss": 0.083984375, "ewc_loss_diag": 6.008148193359375e-05, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.767758846282959, "learning_rate": 1e-05, "loss": 0.57, "mean_token_accuracy": 0.8401884436607361, "num_tokens": 109382679.0, "step": 2879 }, { "epoch": 0.3663656023406691, "ewc_loss": 0.0712890625, "ewc_loss_diag": 6.008148193359375e-05, "ewc_loss_parallel": 6.532669067382812e-05, "grad_norm": 4.773096084594727, "learning_rate": 1e-05, "loss": 0.593, "mean_token_accuracy": 0.8351008892059326, "num_tokens": 109428491.0, "step": 2880 }, { "epoch": 0.36649281261925964, "ewc_loss": 0.08935546875, "ewc_loss_diag": 6.008148193359375e-05, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 6.012407302856445, "learning_rate": 1e-05, "loss": 0.5653, "mean_token_accuracy": 0.841152548789978, "num_tokens": 109469700.0, "step": 2881 }, { "epoch": 0.3666200228978502, "ewc_loss": 0.07275390625, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 6.67572021484375e-05, "grad_norm": 4.860307693481445, "learning_rate": 1e-05, "loss": 0.4917, "mean_token_accuracy": 0.8601268529891968, "num_tokens": 109510245.0, "step": 2882 }, { "epoch": 0.36674723317644065, "ewc_loss": 0.09326171875, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 6.174746036529541, "learning_rate": 1e-05, "loss": 0.5621, "mean_token_accuracy": 0.8424723148345947, "num_tokens": 109544932.0, "step": 2883 }, { "epoch": 0.3668744434550312, "ewc_loss": 0.078125, "ewc_loss_diag": 6.079673767089844e-05, "ewc_loss_parallel": 7.2479248046875e-05, "grad_norm": 5.111408710479736, "learning_rate": 1e-05, "loss": 0.5044, "mean_token_accuracy": 0.8587266802787781, "num_tokens": 109579519.0, "step": 2884 }, { "epoch": 0.3670016537336217, "ewc_loss": 0.08935546875, "ewc_loss_diag": 6.079673767089844e-05, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 6.049118995666504, "learning_rate": 1e-05, "loss": 0.5261, "mean_token_accuracy": 0.8588584661483765, "num_tokens": 109620313.0, "step": 2885 }, { "epoch": 0.3671288640122122, "ewc_loss": 0.08154296875, "ewc_loss_diag": 6.079673767089844e-05, "ewc_loss_parallel": 7.581710815429688e-05, "grad_norm": 5.3309478759765625, "learning_rate": 1e-05, "loss": 0.5776, "mean_token_accuracy": 0.837641716003418, "num_tokens": 109663800.0, "step": 2886 }, { "epoch": 0.3672560742908027, "ewc_loss": 0.0810546875, "ewc_loss_diag": 6.079673767089844e-05, "ewc_loss_parallel": 7.534027099609375e-05, "grad_norm": 5.6796040534973145, "learning_rate": 1e-05, "loss": 0.4912, "mean_token_accuracy": 0.8608629107475281, "num_tokens": 109701118.0, "step": 2887 }, { "epoch": 0.36738328456939323, "ewc_loss": 0.08251953125, "ewc_loss_diag": 6.079673767089844e-05, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.4281744956970215, "learning_rate": 1e-05, "loss": 0.4845, "mean_token_accuracy": 0.8682088851928711, "num_tokens": 109739486.0, "step": 2888 }, { "epoch": 0.3675104948479837, "ewc_loss": 0.0751953125, "ewc_loss_diag": 6.079673767089844e-05, "ewc_loss_parallel": 6.914138793945312e-05, "grad_norm": 5.234147071838379, "learning_rate": 1e-05, "loss": 0.4835, "mean_token_accuracy": 0.8650165796279907, "num_tokens": 109781097.0, "step": 2889 }, { "epoch": 0.36763770512657423, "ewc_loss": 0.083984375, "ewc_loss_diag": 6.079673767089844e-05, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.569212436676025, "learning_rate": 1e-05, "loss": 0.5061, "mean_token_accuracy": 0.856738805770874, "num_tokens": 109813562.0, "step": 2890 }, { "epoch": 0.36776491540516476, "ewc_loss": 0.07177734375, "ewc_loss_diag": 6.079673767089844e-05, "ewc_loss_parallel": 6.580352783203125e-05, "grad_norm": 4.869834899902344, "learning_rate": 1e-05, "loss": 0.5786, "mean_token_accuracy": 0.8333456516265869, "num_tokens": 109855326.0, "step": 2891 }, { "epoch": 0.36789212568375523, "ewc_loss": 0.0869140625, "ewc_loss_diag": 6.079673767089844e-05, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.823373317718506, "learning_rate": 1e-05, "loss": 0.546, "mean_token_accuracy": 0.8492164611816406, "num_tokens": 109890787.0, "step": 2892 }, { "epoch": 0.36801933596234576, "ewc_loss": 0.06982421875, "ewc_loss_diag": 6.079673767089844e-05, "ewc_loss_parallel": 6.389617919921875e-05, "grad_norm": 4.680233478546143, "learning_rate": 1e-05, "loss": 0.533, "mean_token_accuracy": 0.8482544422149658, "num_tokens": 109932199.0, "step": 2893 }, { "epoch": 0.3681465462409363, "ewc_loss": 0.09375, "ewc_loss_diag": 6.079673767089844e-05, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 6.196568489074707, "learning_rate": 1e-05, "loss": 0.5212, "mean_token_accuracy": 0.8562893271446228, "num_tokens": 109967388.0, "step": 2894 }, { "epoch": 0.36827375651952676, "ewc_loss": 0.0732421875, "ewc_loss_diag": 6.079673767089844e-05, "ewc_loss_parallel": 6.723403930664062e-05, "grad_norm": 4.811548709869385, "learning_rate": 1e-05, "loss": 0.4722, "mean_token_accuracy": 0.8630125522613525, "num_tokens": 110003437.0, "step": 2895 }, { "epoch": 0.3684009667981173, "ewc_loss": 0.099609375, "ewc_loss_diag": 6.079673767089844e-05, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.464593887329102, "learning_rate": 1e-05, "loss": 0.5552, "mean_token_accuracy": 0.8471606373786926, "num_tokens": 110040822.0, "step": 2896 }, { "epoch": 0.3685281770767078, "ewc_loss": 0.08154296875, "ewc_loss_diag": 6.079673767089844e-05, "ewc_loss_parallel": 7.581710815429688e-05, "grad_norm": 5.238771915435791, "learning_rate": 1e-05, "loss": 0.507, "mean_token_accuracy": 0.8567379713058472, "num_tokens": 110079628.0, "step": 2897 }, { "epoch": 0.3686553873552983, "ewc_loss": 0.09716796875, "ewc_loss_diag": 6.079673767089844e-05, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.320590972900391, "learning_rate": 1e-05, "loss": 0.5078, "mean_token_accuracy": 0.8612879514694214, "num_tokens": 110115794.0, "step": 2898 }, { "epoch": 0.3687825976338888, "ewc_loss": 0.08544921875, "ewc_loss_diag": 6.079673767089844e-05, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.499208927154541, "learning_rate": 1e-05, "loss": 0.5552, "mean_token_accuracy": 0.846355140209198, "num_tokens": 110153227.0, "step": 2899 }, { "epoch": 0.36890980791247935, "ewc_loss": 0.08642578125, "ewc_loss_diag": 6.079673767089844e-05, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.973334312438965, "learning_rate": 1e-05, "loss": 0.5906, "mean_token_accuracy": 0.8336923718452454, "num_tokens": 110193186.0, "step": 2900 }, { "epoch": 0.3690370181910698, "ewc_loss": 0.08203125, "ewc_loss_diag": 6.079673767089844e-05, "ewc_loss_parallel": 7.62939453125e-05, "grad_norm": 5.468583583831787, "learning_rate": 1e-05, "loss": 0.5693, "mean_token_accuracy": 0.84339439868927, "num_tokens": 110232130.0, "step": 2901 }, { "epoch": 0.36916422846966035, "ewc_loss": 0.0791015625, "ewc_loss_diag": 6.079673767089844e-05, "ewc_loss_parallel": 7.343292236328125e-05, "grad_norm": 5.660110950469971, "learning_rate": 1e-05, "loss": 0.4778, "mean_token_accuracy": 0.8643776178359985, "num_tokens": 110271376.0, "step": 2902 }, { "epoch": 0.3692914387482509, "ewc_loss": 0.080078125, "ewc_loss_diag": 6.079673767089844e-05, "ewc_loss_parallel": 7.43865966796875e-05, "grad_norm": 5.353361129760742, "learning_rate": 1e-05, "loss": 0.55, "mean_token_accuracy": 0.8458601236343384, "num_tokens": 110314220.0, "step": 2903 }, { "epoch": 0.36941864902684135, "ewc_loss": 0.07763671875, "ewc_loss_diag": 6.079673767089844e-05, "ewc_loss_parallel": 7.200241088867188e-05, "grad_norm": 5.481588840484619, "learning_rate": 1e-05, "loss": 0.5413, "mean_token_accuracy": 0.8477648496627808, "num_tokens": 110357292.0, "step": 2904 }, { "epoch": 0.3695458593054319, "ewc_loss": 0.076171875, "ewc_loss_diag": 6.079673767089844e-05, "ewc_loss_parallel": 7.009506225585938e-05, "grad_norm": 5.21820592880249, "learning_rate": 1e-05, "loss": 0.583, "mean_token_accuracy": 0.8328518271446228, "num_tokens": 110397813.0, "step": 2905 }, { "epoch": 0.3696730695840224, "ewc_loss": 0.0771484375, "ewc_loss_diag": 6.079673767089844e-05, "ewc_loss_parallel": 7.152557373046875e-05, "grad_norm": 5.423126697540283, "learning_rate": 1e-05, "loss": 0.4864, "mean_token_accuracy": 0.8624240159988403, "num_tokens": 110435773.0, "step": 2906 }, { "epoch": 0.3698002798626129, "ewc_loss": 0.07421875, "ewc_loss_diag": 6.079673767089844e-05, "ewc_loss_parallel": 6.818771362304688e-05, "grad_norm": 5.067737102508545, "learning_rate": 1e-05, "loss": 0.4405, "mean_token_accuracy": 0.8741812705993652, "num_tokens": 110469518.0, "step": 2907 }, { "epoch": 0.3699274901412034, "ewc_loss": 0.078125, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 7.200241088867188e-05, "grad_norm": 5.472022533416748, "learning_rate": 1e-05, "loss": 0.5301, "mean_token_accuracy": 0.848710834980011, "num_tokens": 110503156.0, "step": 2908 }, { "epoch": 0.37005470041979394, "ewc_loss": 0.07275390625, "ewc_loss_diag": 6.079673767089844e-05, "ewc_loss_parallel": 6.67572021484375e-05, "grad_norm": 4.925530910491943, "learning_rate": 1e-05, "loss": 0.5473, "mean_token_accuracy": 0.8431364893913269, "num_tokens": 110543960.0, "step": 2909 }, { "epoch": 0.3701819106983844, "ewc_loss": 0.0810546875, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 7.534027099609375e-05, "grad_norm": 5.610262393951416, "learning_rate": 1e-05, "loss": 0.5737, "mean_token_accuracy": 0.8387213945388794, "num_tokens": 110584811.0, "step": 2910 }, { "epoch": 0.37030912097697494, "ewc_loss": 0.0712890625, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 6.4849853515625e-05, "grad_norm": 4.797543048858643, "learning_rate": 1e-05, "loss": 0.5571, "mean_token_accuracy": 0.8413823843002319, "num_tokens": 110622433.0, "step": 2911 }, { "epoch": 0.37043633125556547, "ewc_loss": 0.0859375, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.869450092315674, "learning_rate": 1e-05, "loss": 0.5465, "mean_token_accuracy": 0.8468831777572632, "num_tokens": 110660994.0, "step": 2912 }, { "epoch": 0.37056354153415594, "ewc_loss": 0.0712890625, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 6.532669067382812e-05, "grad_norm": 4.756028652191162, "learning_rate": 1e-05, "loss": 0.6022, "mean_token_accuracy": 0.8306740522384644, "num_tokens": 110703022.0, "step": 2913 }, { "epoch": 0.37069075181274647, "ewc_loss": 0.0927734375, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 6.150247573852539, "learning_rate": 1e-05, "loss": 0.5053, "mean_token_accuracy": 0.8620660901069641, "num_tokens": 110743864.0, "step": 2914 }, { "epoch": 0.370817962091337, "ewc_loss": 0.07421875, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 6.818771362304688e-05, "grad_norm": 4.909804821014404, "learning_rate": 1e-05, "loss": 0.5299, "mean_token_accuracy": 0.8464862108230591, "num_tokens": 110778591.0, "step": 2915 }, { "epoch": 0.37094517236992747, "ewc_loss": 0.0966796875, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.288822650909424, "learning_rate": 1e-05, "loss": 0.5165, "mean_token_accuracy": 0.8605977296829224, "num_tokens": 110818402.0, "step": 2916 }, { "epoch": 0.371072382648518, "ewc_loss": 0.080078125, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 7.390975952148438e-05, "grad_norm": 5.226221084594727, "learning_rate": 1e-05, "loss": 0.5276, "mean_token_accuracy": 0.8527629375457764, "num_tokens": 110855129.0, "step": 2917 }, { "epoch": 0.3711995929271085, "ewc_loss": 0.0908203125, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 6.099697113037109, "learning_rate": 1e-05, "loss": 0.5243, "mean_token_accuracy": 0.8545887470245361, "num_tokens": 110896307.0, "step": 2918 }, { "epoch": 0.371326803205699, "ewc_loss": 0.083984375, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.431463241577148, "learning_rate": 1e-05, "loss": 0.5149, "mean_token_accuracy": 0.8560303449630737, "num_tokens": 110940639.0, "step": 2919 }, { "epoch": 0.3714540134842895, "ewc_loss": 0.0810546875, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 7.486343383789062e-05, "grad_norm": 5.693181991577148, "learning_rate": 1e-05, "loss": 0.5293, "mean_token_accuracy": 0.8523509502410889, "num_tokens": 110979134.0, "step": 2920 }, { "epoch": 0.37158122376288005, "ewc_loss": 0.083984375, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.5082526206970215, "learning_rate": 1e-05, "loss": 0.4916, "mean_token_accuracy": 0.8641622066497803, "num_tokens": 111013799.0, "step": 2921 }, { "epoch": 0.3717084340414705, "ewc_loss": 0.076171875, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 6.961822509765625e-05, "grad_norm": 5.2512640953063965, "learning_rate": 1e-05, "loss": 0.5253, "mean_token_accuracy": 0.8518970012664795, "num_tokens": 111055486.0, "step": 2922 }, { "epoch": 0.37183564432006105, "ewc_loss": 0.0830078125, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.602669715881348, "learning_rate": 1e-05, "loss": 0.5254, "mean_token_accuracy": 0.854669451713562, "num_tokens": 111101332.0, "step": 2923 }, { "epoch": 0.3719628545986516, "ewc_loss": 0.072265625, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 6.580352783203125e-05, "grad_norm": 4.920334339141846, "learning_rate": 1e-05, "loss": 0.517, "mean_token_accuracy": 0.8537002205848694, "num_tokens": 111137285.0, "step": 2924 }, { "epoch": 0.37209006487724205, "ewc_loss": 0.0859375, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.826179027557373, "learning_rate": 1e-05, "loss": 0.5571, "mean_token_accuracy": 0.8447811603546143, "num_tokens": 111177254.0, "step": 2925 }, { "epoch": 0.3722172751558326, "ewc_loss": 0.0712890625, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 6.4849853515625e-05, "grad_norm": 4.740542411804199, "learning_rate": 1e-05, "loss": 0.5173, "mean_token_accuracy": 0.8527057766914368, "num_tokens": 111217633.0, "step": 2926 }, { "epoch": 0.3723444854344231, "ewc_loss": 0.09375, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 6.153563976287842, "learning_rate": 1e-05, "loss": 0.6368, "mean_token_accuracy": 0.8221123218536377, "num_tokens": 111259273.0, "step": 2927 }, { "epoch": 0.3724716957130136, "ewc_loss": 0.0732421875, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 6.67572021484375e-05, "grad_norm": 4.849146842956543, "learning_rate": 1e-05, "loss": 0.5611, "mean_token_accuracy": 0.8369183540344238, "num_tokens": 111295068.0, "step": 2928 }, { "epoch": 0.3725989059916041, "ewc_loss": 0.09765625, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.373180866241455, "learning_rate": 1e-05, "loss": 0.582, "mean_token_accuracy": 0.8383642435073853, "num_tokens": 111336863.0, "step": 2929 }, { "epoch": 0.37272611627019464, "ewc_loss": 0.080078125, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 7.43865966796875e-05, "grad_norm": 5.184212684631348, "learning_rate": 1e-05, "loss": 0.5788, "mean_token_accuracy": 0.8380706310272217, "num_tokens": 111378432.0, "step": 2930 }, { "epoch": 0.37285332654878517, "ewc_loss": 0.095703125, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.276580333709717, "learning_rate": 1e-05, "loss": 0.5654, "mean_token_accuracy": 0.8442239761352539, "num_tokens": 111412481.0, "step": 2931 }, { "epoch": 0.37298053682737564, "ewc_loss": 0.0830078125, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.419506549835205, "learning_rate": 1e-05, "loss": 0.5807, "mean_token_accuracy": 0.8351156711578369, "num_tokens": 111450534.0, "step": 2932 }, { "epoch": 0.37310774710596617, "ewc_loss": 0.0859375, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.974350452423096, "learning_rate": 1e-05, "loss": 0.5043, "mean_token_accuracy": 0.8603249788284302, "num_tokens": 111484278.0, "step": 2933 }, { "epoch": 0.3732349573845567, "ewc_loss": 0.08203125, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 7.62939453125e-05, "grad_norm": 5.4245829582214355, "learning_rate": 1e-05, "loss": 0.5374, "mean_token_accuracy": 0.8454364538192749, "num_tokens": 111526797.0, "step": 2934 }, { "epoch": 0.37336216766314717, "ewc_loss": 0.080078125, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 7.43865966796875e-05, "grad_norm": 5.676746845245361, "learning_rate": 1e-05, "loss": 0.5667, "mean_token_accuracy": 0.839891791343689, "num_tokens": 111568139.0, "step": 2935 }, { "epoch": 0.3734893779417377, "ewc_loss": 0.0791015625, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 7.343292236328125e-05, "grad_norm": 5.347311019897461, "learning_rate": 1e-05, "loss": 0.5974, "mean_token_accuracy": 0.8282724022865295, "num_tokens": 111614519.0, "step": 2936 }, { "epoch": 0.3736165882203282, "ewc_loss": 0.078125, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 7.200241088867188e-05, "grad_norm": 5.481536865234375, "learning_rate": 1e-05, "loss": 0.5031, "mean_token_accuracy": 0.8565324544906616, "num_tokens": 111647965.0, "step": 2937 }, { "epoch": 0.3737437984989187, "ewc_loss": 0.0771484375, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 7.05718994140625e-05, "grad_norm": 5.23594856262207, "learning_rate": 1e-05, "loss": 0.5264, "mean_token_accuracy": 0.8512938618659973, "num_tokens": 111682492.0, "step": 2938 }, { "epoch": 0.37387100877750923, "ewc_loss": 0.0771484375, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 7.152557373046875e-05, "grad_norm": 5.394574165344238, "learning_rate": 1e-05, "loss": 0.5604, "mean_token_accuracy": 0.8413119316101074, "num_tokens": 111717550.0, "step": 2939 }, { "epoch": 0.37399821905609976, "ewc_loss": 0.0751953125, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 6.866455078125e-05, "grad_norm": 5.1057515144348145, "learning_rate": 1e-05, "loss": 0.4897, "mean_token_accuracy": 0.8619656562805176, "num_tokens": 111752173.0, "step": 2940 }, { "epoch": 0.37412542933469023, "ewc_loss": 0.078125, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 7.200241088867188e-05, "grad_norm": 5.404499530792236, "learning_rate": 1e-05, "loss": 0.5193, "mean_token_accuracy": 0.8501968383789062, "num_tokens": 111790845.0, "step": 2941 }, { "epoch": 0.37425263961328076, "ewc_loss": 0.0732421875, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 6.723403930664062e-05, "grad_norm": 4.973315715789795, "learning_rate": 1e-05, "loss": 0.5282, "mean_token_accuracy": 0.8477611541748047, "num_tokens": 111827837.0, "step": 2942 }, { "epoch": 0.3743798498918713, "ewc_loss": 0.078125, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 7.2479248046875e-05, "grad_norm": 5.504717826843262, "learning_rate": 1e-05, "loss": 0.5552, "mean_token_accuracy": 0.8406407833099365, "num_tokens": 111868785.0, "step": 2943 }, { "epoch": 0.37450706017046176, "ewc_loss": 0.0712890625, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 6.532669067382812e-05, "grad_norm": 4.839818000793457, "learning_rate": 1e-05, "loss": 0.5293, "mean_token_accuracy": 0.8471118807792664, "num_tokens": 111907296.0, "step": 2944 }, { "epoch": 0.3746342704490523, "ewc_loss": 0.08203125, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 7.62939453125e-05, "grad_norm": 5.68995475769043, "learning_rate": 1e-05, "loss": 0.4827, "mean_token_accuracy": 0.8657071590423584, "num_tokens": 111947619.0, "step": 2945 }, { "epoch": 0.3747614807276428, "ewc_loss": 0.0703125, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 6.437301635742188e-05, "grad_norm": 4.746450901031494, "learning_rate": 1e-05, "loss": 0.5777, "mean_token_accuracy": 0.8339998126029968, "num_tokens": 111987255.0, "step": 2946 }, { "epoch": 0.3748886910062333, "ewc_loss": 0.0888671875, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.988801002502441, "learning_rate": 1e-05, "loss": 0.5554, "mean_token_accuracy": 0.8469124436378479, "num_tokens": 112023003.0, "step": 2947 }, { "epoch": 0.3750159012848238, "ewc_loss": 0.072265625, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 6.580352783203125e-05, "grad_norm": 4.768383026123047, "learning_rate": 1e-05, "loss": 0.544, "mean_token_accuracy": 0.8419710397720337, "num_tokens": 112056292.0, "step": 2948 }, { "epoch": 0.37514311156341434, "ewc_loss": 0.095703125, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.276968002319336, "learning_rate": 1e-05, "loss": 0.5509, "mean_token_accuracy": 0.8490985631942749, "num_tokens": 112089947.0, "step": 2949 }, { "epoch": 0.3752703218420048, "ewc_loss": 0.0771484375, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 7.05718994140625e-05, "grad_norm": 5.033992767333984, "learning_rate": 1e-05, "loss": 0.6107, "mean_token_accuracy": 0.8264232873916626, "num_tokens": 112125415.0, "step": 2950 }, { "epoch": 0.37539753212059535, "ewc_loss": 0.09765625, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.292011737823486, "learning_rate": 1e-05, "loss": 0.5705, "mean_token_accuracy": 0.844663143157959, "num_tokens": 112164364.0, "step": 2951 }, { "epoch": 0.3755247423991859, "ewc_loss": 0.0830078125, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.346743583679199, "learning_rate": 1e-05, "loss": 0.5, "mean_token_accuracy": 0.8590736389160156, "num_tokens": 112208994.0, "step": 2952 }, { "epoch": 0.37565195267777635, "ewc_loss": 0.0869140625, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.973779201507568, "learning_rate": 1e-05, "loss": 0.5311, "mean_token_accuracy": 0.8527593016624451, "num_tokens": 112249902.0, "step": 2953 }, { "epoch": 0.3757791629563669, "ewc_loss": 0.0830078125, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.484216213226318, "learning_rate": 1e-05, "loss": 0.5074, "mean_token_accuracy": 0.8594051599502563, "num_tokens": 112290446.0, "step": 2954 }, { "epoch": 0.3759063732349574, "ewc_loss": 0.0791015625, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 7.343292236328125e-05, "grad_norm": 5.5406494140625, "learning_rate": 1e-05, "loss": 0.5325, "mean_token_accuracy": 0.8484888672828674, "num_tokens": 112323217.0, "step": 2955 }, { "epoch": 0.3760335835135479, "ewc_loss": 0.083984375, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.528381824493408, "learning_rate": 1e-05, "loss": 0.5292, "mean_token_accuracy": 0.8536475896835327, "num_tokens": 112365863.0, "step": 2956 }, { "epoch": 0.3761607937921384, "ewc_loss": 0.0751953125, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 6.866455078125e-05, "grad_norm": 5.1382269859313965, "learning_rate": 1e-05, "loss": 0.5192, "mean_token_accuracy": 0.850927472114563, "num_tokens": 112400708.0, "step": 2957 }, { "epoch": 0.37628800407072893, "ewc_loss": 0.083984375, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.641418933868408, "learning_rate": 1e-05, "loss": 0.4978, "mean_token_accuracy": 0.8633845448493958, "num_tokens": 112440063.0, "step": 2958 }, { "epoch": 0.3764152143493194, "ewc_loss": 0.072265625, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 6.580352783203125e-05, "grad_norm": 4.8604230880737305, "learning_rate": 1e-05, "loss": 0.4446, "mean_token_accuracy": 0.8759487867355347, "num_tokens": 112478856.0, "step": 2959 }, { "epoch": 0.37654242462790993, "ewc_loss": 0.0869140625, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.877180576324463, "learning_rate": 1e-05, "loss": 0.5425, "mean_token_accuracy": 0.8507903814315796, "num_tokens": 112515707.0, "step": 2960 }, { "epoch": 0.37666963490650046, "ewc_loss": 0.0712890625, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 6.532669067382812e-05, "grad_norm": 4.78005838394165, "learning_rate": 1e-05, "loss": 0.5407, "mean_token_accuracy": 0.8427800536155701, "num_tokens": 112551302.0, "step": 2961 }, { "epoch": 0.37679684518509093, "ewc_loss": 0.09375, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 6.148947715759277, "learning_rate": 1e-05, "loss": 0.5058, "mean_token_accuracy": 0.8623281121253967, "num_tokens": 112585821.0, "step": 2962 }, { "epoch": 0.37692405546368146, "ewc_loss": 0.07470703125, "ewc_loss_diag": 6.151199340820312e-05, "ewc_loss_parallel": 6.818771362304688e-05, "grad_norm": 4.910206317901611, "learning_rate": 1e-05, "loss": 0.5152, "mean_token_accuracy": 0.8525875806808472, "num_tokens": 112624261.0, "step": 2963 }, { "epoch": 0.377051265742272, "ewc_loss": 0.0966796875, "ewc_loss_diag": 6.151199340820312e-05, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.291121006011963, "learning_rate": 1e-05, "loss": 0.4879, "mean_token_accuracy": 0.867475152015686, "num_tokens": 112656944.0, "step": 2964 }, { "epoch": 0.37717847602086246, "ewc_loss": 0.080078125, "ewc_loss_diag": 6.151199340820312e-05, "ewc_loss_parallel": 7.390975952148438e-05, "grad_norm": 5.183639049530029, "learning_rate": 1e-05, "loss": 0.5002, "mean_token_accuracy": 0.858574628829956, "num_tokens": 112694386.0, "step": 2965 }, { "epoch": 0.377305686299453, "ewc_loss": 0.09326171875, "ewc_loss_diag": 6.151199340820312e-05, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 6.177447319030762, "learning_rate": 1e-05, "loss": 0.5519, "mean_token_accuracy": 0.8516850471496582, "num_tokens": 112731597.0, "step": 2966 }, { "epoch": 0.3774328965780435, "ewc_loss": 0.08251953125, "ewc_loss_diag": 6.151199340820312e-05, "ewc_loss_parallel": 7.62939453125e-05, "grad_norm": 5.341350555419922, "learning_rate": 1e-05, "loss": 0.4492, "mean_token_accuracy": 0.8772926330566406, "num_tokens": 112768012.0, "step": 2967 }, { "epoch": 0.377560106856634, "ewc_loss": 0.08642578125, "ewc_loss_diag": 6.151199340820312e-05, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.938503265380859, "learning_rate": 1e-05, "loss": 0.5706, "mean_token_accuracy": 0.838585376739502, "num_tokens": 112798089.0, "step": 2968 }, { "epoch": 0.3776873171352245, "ewc_loss": 0.08154296875, "ewc_loss_diag": 6.151199340820312e-05, "ewc_loss_parallel": 7.534027099609375e-05, "grad_norm": 5.350839138031006, "learning_rate": 1e-05, "loss": 0.5172, "mean_token_accuracy": 0.855494499206543, "num_tokens": 112842332.0, "step": 2969 }, { "epoch": 0.37781452741381505, "ewc_loss": 0.08056640625, "ewc_loss_diag": 6.151199340820312e-05, "ewc_loss_parallel": 7.43865966796875e-05, "grad_norm": 5.676687240600586, "learning_rate": 1e-05, "loss": 0.5439, "mean_token_accuracy": 0.8464314937591553, "num_tokens": 112882382.0, "step": 2970 }, { "epoch": 0.3779417376924055, "ewc_loss": 0.0791015625, "ewc_loss_diag": 6.151199340820312e-05, "ewc_loss_parallel": 7.295608520507812e-05, "grad_norm": 5.284313201904297, "learning_rate": 1e-05, "loss": 0.5992, "mean_token_accuracy": 0.8285456895828247, "num_tokens": 112922627.0, "step": 2971 }, { "epoch": 0.37806894797099605, "ewc_loss": 0.0791015625, "ewc_loss_diag": 6.151199340820312e-05, "ewc_loss_parallel": 7.295608520507812e-05, "grad_norm": 5.515600204467773, "learning_rate": 1e-05, "loss": 0.5902, "mean_token_accuracy": 0.8300535082817078, "num_tokens": 112957571.0, "step": 2972 }, { "epoch": 0.3781961582495866, "ewc_loss": 0.0771484375, "ewc_loss_diag": 6.151199340820312e-05, "ewc_loss_parallel": 7.05718994140625e-05, "grad_norm": 5.18848180770874, "learning_rate": 1e-05, "loss": 0.5249, "mean_token_accuracy": 0.8497059941291809, "num_tokens": 113001128.0, "step": 2973 }, { "epoch": 0.37832336852817705, "ewc_loss": 0.078125, "ewc_loss_diag": 6.151199340820312e-05, "ewc_loss_parallel": 7.200241088867188e-05, "grad_norm": 5.429818153381348, "learning_rate": 1e-05, "loss": 0.5052, "mean_token_accuracy": 0.856109619140625, "num_tokens": 113034515.0, "step": 2974 }, { "epoch": 0.3784505788067676, "ewc_loss": 0.0751953125, "ewc_loss_diag": 6.151199340820312e-05, "ewc_loss_parallel": 6.866455078125e-05, "grad_norm": 5.081949234008789, "learning_rate": 1e-05, "loss": 0.4675, "mean_token_accuracy": 0.8674935102462769, "num_tokens": 113071221.0, "step": 2975 }, { "epoch": 0.3785777890853581, "ewc_loss": 0.07861328125, "ewc_loss_diag": 6.151199340820312e-05, "ewc_loss_parallel": 7.2479248046875e-05, "grad_norm": 5.447579383850098, "learning_rate": 1e-05, "loss": 0.5944, "mean_token_accuracy": 0.8323150873184204, "num_tokens": 113105804.0, "step": 2976 }, { "epoch": 0.3787049993639486, "ewc_loss": 0.07373046875, "ewc_loss_diag": 6.151199340820312e-05, "ewc_loss_parallel": 6.723403930664062e-05, "grad_norm": 4.956618785858154, "learning_rate": 1e-05, "loss": 0.4983, "mean_token_accuracy": 0.8590521812438965, "num_tokens": 113146051.0, "step": 2977 }, { "epoch": 0.3788322096425391, "ewc_loss": 0.0810546875, "ewc_loss_diag": 6.151199340820312e-05, "ewc_loss_parallel": 7.486343383789062e-05, "grad_norm": 5.546780109405518, "learning_rate": 1e-05, "loss": 0.486, "mean_token_accuracy": 0.8652840852737427, "num_tokens": 113186696.0, "step": 2978 }, { "epoch": 0.37895941992112964, "ewc_loss": 0.07177734375, "ewc_loss_diag": 6.151199340820312e-05, "ewc_loss_parallel": 6.532669067382812e-05, "grad_norm": 4.852226257324219, "learning_rate": 1e-05, "loss": 0.5455, "mean_token_accuracy": 0.8429550528526306, "num_tokens": 113228402.0, "step": 2979 }, { "epoch": 0.37908663019972016, "ewc_loss": 0.0830078125, "ewc_loss_diag": 6.151199340820312e-05, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.720276355743408, "learning_rate": 1e-05, "loss": 0.5649, "mean_token_accuracy": 0.8400765061378479, "num_tokens": 113270407.0, "step": 2980 }, { "epoch": 0.37921384047831064, "ewc_loss": 0.07177734375, "ewc_loss_diag": 6.151199340820312e-05, "ewc_loss_parallel": 6.532669067382812e-05, "grad_norm": 4.791535377502441, "learning_rate": 1e-05, "loss": 0.5392, "mean_token_accuracy": 0.8438411951065063, "num_tokens": 113304523.0, "step": 2981 }, { "epoch": 0.37934105075690117, "ewc_loss": 0.08837890625, "ewc_loss_diag": 6.151199340820312e-05, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.944753170013428, "learning_rate": 1e-05, "loss": 0.5387, "mean_token_accuracy": 0.8547558784484863, "num_tokens": 113345224.0, "step": 2982 }, { "epoch": 0.3794682610354917, "ewc_loss": 0.0732421875, "ewc_loss_diag": 6.151199340820312e-05, "ewc_loss_parallel": 6.67572021484375e-05, "grad_norm": 4.821811676025391, "learning_rate": 1e-05, "loss": 0.4886, "mean_token_accuracy": 0.8615114092826843, "num_tokens": 113382287.0, "step": 2983 }, { "epoch": 0.37959547131408217, "ewc_loss": 0.09326171875, "ewc_loss_diag": 6.151199340820312e-05, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 6.147183418273926, "learning_rate": 1e-05, "loss": 0.5775, "mean_token_accuracy": 0.8396626710891724, "num_tokens": 113427308.0, "step": 2984 }, { "epoch": 0.3797226815926727, "ewc_loss": 0.076171875, "ewc_loss_diag": 6.151199340820312e-05, "ewc_loss_parallel": 6.961822509765625e-05, "grad_norm": 5.0351057052612305, "learning_rate": 1e-05, "loss": 0.5227, "mean_token_accuracy": 0.8506519794464111, "num_tokens": 113461272.0, "step": 2985 }, { "epoch": 0.3798498918712632, "ewc_loss": 0.09326171875, "ewc_loss_diag": 6.151199340820312e-05, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 6.107008457183838, "learning_rate": 1e-05, "loss": 0.4759, "mean_token_accuracy": 0.8706964254379272, "num_tokens": 113499942.0, "step": 2986 }, { "epoch": 0.3799771021498537, "ewc_loss": 0.0810546875, "ewc_loss_diag": 6.151199340820312e-05, "ewc_loss_parallel": 7.486343383789062e-05, "grad_norm": 5.267038345336914, "learning_rate": 1e-05, "loss": 0.4692, "mean_token_accuracy": 0.8694278001785278, "num_tokens": 113531801.0, "step": 2987 }, { "epoch": 0.3801043124284442, "ewc_loss": 0.08544921875, "ewc_loss_diag": 6.151199340820312e-05, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.813174724578857, "learning_rate": 1e-05, "loss": 0.4846, "mean_token_accuracy": 0.8678936958312988, "num_tokens": 113567475.0, "step": 2988 }, { "epoch": 0.38023152270703475, "ewc_loss": 0.0830078125, "ewc_loss_diag": 6.151199340820312e-05, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.39031982421875, "learning_rate": 1e-05, "loss": 0.5666, "mean_token_accuracy": 0.8392492532730103, "num_tokens": 113603767.0, "step": 2989 }, { "epoch": 0.3803587329856252, "ewc_loss": 0.078125, "ewc_loss_diag": 6.151199340820312e-05, "ewc_loss_parallel": 7.200241088867188e-05, "grad_norm": 5.399256229400635, "learning_rate": 1e-05, "loss": 0.5387, "mean_token_accuracy": 0.8469765782356262, "num_tokens": 113641213.0, "step": 2990 }, { "epoch": 0.38048594326421575, "ewc_loss": 0.083984375, "ewc_loss_diag": 6.151199340820312e-05, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.503583908081055, "learning_rate": 1e-05, "loss": 0.6104, "mean_token_accuracy": 0.8265458941459656, "num_tokens": 113683364.0, "step": 2991 }, { "epoch": 0.3806131535428063, "ewc_loss": 0.0732421875, "ewc_loss_diag": 6.151199340820312e-05, "ewc_loss_parallel": 6.67572021484375e-05, "grad_norm": 5.010347366333008, "learning_rate": 1e-05, "loss": 0.5302, "mean_token_accuracy": 0.8475652933120728, "num_tokens": 113718488.0, "step": 2992 }, { "epoch": 0.38074036382139675, "ewc_loss": 0.08642578125, "ewc_loss_diag": 6.151199340820312e-05, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.700976848602295, "learning_rate": 1e-05, "loss": 0.496, "mean_token_accuracy": 0.8642219305038452, "num_tokens": 113756664.0, "step": 2993 }, { "epoch": 0.3808675740999873, "ewc_loss": 0.0712890625, "ewc_loss_diag": 6.151199340820312e-05, "ewc_loss_parallel": 6.4849853515625e-05, "grad_norm": 4.754619121551514, "learning_rate": 1e-05, "loss": 0.4862, "mean_token_accuracy": 0.8579182624816895, "num_tokens": 113790711.0, "step": 2994 }, { "epoch": 0.3809947843785778, "ewc_loss": 0.08984375, "ewc_loss_diag": 6.151199340820312e-05, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 6.020843982696533, "learning_rate": 1e-05, "loss": 0.531, "mean_token_accuracy": 0.8612154126167297, "num_tokens": 113827091.0, "step": 2995 }, { "epoch": 0.3811219946571683, "ewc_loss": 0.07275390625, "ewc_loss_diag": 6.151199340820312e-05, "ewc_loss_parallel": 6.628036499023438e-05, "grad_norm": 4.741235256195068, "learning_rate": 1e-05, "loss": 0.5605, "mean_token_accuracy": 0.8376813530921936, "num_tokens": 113862971.0, "step": 2996 }, { "epoch": 0.3812492049357588, "ewc_loss": 0.09814453125, "ewc_loss_diag": 6.151199340820312e-05, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.347530841827393, "learning_rate": 1e-05, "loss": 0.5675, "mean_token_accuracy": 0.8469492793083191, "num_tokens": 113897430.0, "step": 2997 }, { "epoch": 0.38137641521434934, "ewc_loss": 0.07763671875, "ewc_loss_diag": 6.151199340820312e-05, "ewc_loss_parallel": 7.152557373046875e-05, "grad_norm": 5.026121139526367, "learning_rate": 1e-05, "loss": 0.5249, "mean_token_accuracy": 0.8498923778533936, "num_tokens": 113926626.0, "step": 2998 }, { "epoch": 0.3815036254929398, "ewc_loss": 0.099609375, "ewc_loss_diag": 6.151199340820312e-05, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.395116806030273, "learning_rate": 1e-05, "loss": 0.5165, "mean_token_accuracy": 0.8595374226570129, "num_tokens": 113969656.0, "step": 2999 }, { "epoch": 0.38163083577153034, "ewc_loss": 0.083984375, "ewc_loss_diag": 6.151199340820312e-05, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.355810165405273, "learning_rate": 1e-05, "loss": 0.4826, "mean_token_accuracy": 0.8651111125946045, "num_tokens": 114006140.0, "step": 3000 }, { "epoch": 0.38175804605012087, "ewc_loss": 0.09375, "ewc_loss_diag": 6.151199340820312e-05, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 6.149930000305176, "learning_rate": 1e-05, "loss": 0.5174, "mean_token_accuracy": 0.8579840064048767, "num_tokens": 114047300.0, "step": 3001 }, { "epoch": 0.38188525632871134, "ewc_loss": 0.083984375, "ewc_loss_diag": 6.151199340820312e-05, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.459493637084961, "learning_rate": 1e-05, "loss": 0.5574, "mean_token_accuracy": 0.8465240001678467, "num_tokens": 114086129.0, "step": 3002 }, { "epoch": 0.38201246660730187, "ewc_loss": 0.08447265625, "ewc_loss_diag": 6.151199340820312e-05, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.818619728088379, "learning_rate": 1e-05, "loss": 0.5259, "mean_token_accuracy": 0.8511690497398376, "num_tokens": 114123345.0, "step": 3003 }, { "epoch": 0.3821396768858924, "ewc_loss": 0.0830078125, "ewc_loss_diag": 6.151199340820312e-05, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.409848213195801, "learning_rate": 1e-05, "loss": 0.5291, "mean_token_accuracy": 0.8502854108810425, "num_tokens": 114158983.0, "step": 3004 }, { "epoch": 0.38226688716448287, "ewc_loss": 0.07958984375, "ewc_loss_diag": 6.151199340820312e-05, "ewc_loss_parallel": 7.343292236328125e-05, "grad_norm": 5.566068172454834, "learning_rate": 1e-05, "loss": 0.5975, "mean_token_accuracy": 0.8284703493118286, "num_tokens": 114199055.0, "step": 3005 }, { "epoch": 0.3823940974430734, "ewc_loss": 0.0791015625, "ewc_loss_diag": 6.151199340820312e-05, "ewc_loss_parallel": 7.295608520507812e-05, "grad_norm": 5.30198860168457, "learning_rate": 1e-05, "loss": 0.5119, "mean_token_accuracy": 0.8586462736129761, "num_tokens": 114236355.0, "step": 3006 }, { "epoch": 0.38252130772166393, "ewc_loss": 0.078125, "ewc_loss_diag": 6.151199340820312e-05, "ewc_loss_parallel": 7.200241088867188e-05, "grad_norm": 5.4146833419799805, "learning_rate": 1e-05, "loss": 0.5483, "mean_token_accuracy": 0.8447495698928833, "num_tokens": 114285111.0, "step": 3007 }, { "epoch": 0.3826485180002544, "ewc_loss": 0.0771484375, "ewc_loss_diag": 6.151199340820312e-05, "ewc_loss_parallel": 7.05718994140625e-05, "grad_norm": 5.2024922370910645, "learning_rate": 1e-05, "loss": 0.5836, "mean_token_accuracy": 0.8287389874458313, "num_tokens": 114319306.0, "step": 3008 }, { "epoch": 0.38277572827884493, "ewc_loss": 0.07763671875, "ewc_loss_diag": 6.151199340820312e-05, "ewc_loss_parallel": 7.104873657226562e-05, "grad_norm": 5.367754936218262, "learning_rate": 1e-05, "loss": 0.4717, "mean_token_accuracy": 0.8643163442611694, "num_tokens": 114354117.0, "step": 3009 }, { "epoch": 0.38290293855743546, "ewc_loss": 0.07568359375, "ewc_loss_diag": 6.151199340820312e-05, "ewc_loss_parallel": 6.914138793945312e-05, "grad_norm": 5.076140403747559, "learning_rate": 1e-05, "loss": 0.5587, "mean_token_accuracy": 0.8418624401092529, "num_tokens": 114395865.0, "step": 3010 }, { "epoch": 0.38303014883602593, "ewc_loss": 0.07763671875, "ewc_loss_diag": 6.151199340820312e-05, "ewc_loss_parallel": 7.152557373046875e-05, "grad_norm": 5.403285980224609, "learning_rate": 1e-05, "loss": 0.5032, "mean_token_accuracy": 0.8571064472198486, "num_tokens": 114427862.0, "step": 3011 }, { "epoch": 0.38315735911461646, "ewc_loss": 0.07421875, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 6.771087646484375e-05, "grad_norm": 4.957334518432617, "learning_rate": 1e-05, "loss": 0.4934, "mean_token_accuracy": 0.8610769510269165, "num_tokens": 114462892.0, "step": 3012 }, { "epoch": 0.383284569393207, "ewc_loss": 0.080078125, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 7.390975952148438e-05, "grad_norm": 5.516880989074707, "learning_rate": 1e-05, "loss": 0.4806, "mean_token_accuracy": 0.8623915910720825, "num_tokens": 114500477.0, "step": 3013 }, { "epoch": 0.38341177967179746, "ewc_loss": 0.072265625, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 6.580352783203125e-05, "grad_norm": 4.833464622497559, "learning_rate": 1e-05, "loss": 0.519, "mean_token_accuracy": 0.8487471342086792, "num_tokens": 114534695.0, "step": 3014 }, { "epoch": 0.383538989950388, "ewc_loss": 0.08349609375, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.717730522155762, "learning_rate": 1e-05, "loss": 0.5809, "mean_token_accuracy": 0.8343721628189087, "num_tokens": 114579969.0, "step": 3015 }, { "epoch": 0.3836662002289785, "ewc_loss": 0.07177734375, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 6.532669067382812e-05, "grad_norm": 4.752708911895752, "learning_rate": 1e-05, "loss": 0.5816, "mean_token_accuracy": 0.8349874019622803, "num_tokens": 114614904.0, "step": 3016 }, { "epoch": 0.383793410507569, "ewc_loss": 0.08935546875, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 6.004563808441162, "learning_rate": 1e-05, "loss": 0.4978, "mean_token_accuracy": 0.8628946542739868, "num_tokens": 114647126.0, "step": 3017 }, { "epoch": 0.3839206207861595, "ewc_loss": 0.072265625, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 6.580352783203125e-05, "grad_norm": 4.760456562042236, "learning_rate": 1e-05, "loss": 0.5149, "mean_token_accuracy": 0.8505851030349731, "num_tokens": 114683529.0, "step": 3018 }, { "epoch": 0.38404783106475004, "ewc_loss": 0.09765625, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.315680980682373, "learning_rate": 1e-05, "loss": 0.5269, "mean_token_accuracy": 0.8585116863250732, "num_tokens": 114722471.0, "step": 3019 }, { "epoch": 0.3841750413433405, "ewc_loss": 0.07763671875, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 7.152557373046875e-05, "grad_norm": 5.026884078979492, "learning_rate": 1e-05, "loss": 0.5507, "mean_token_accuracy": 0.8453454971313477, "num_tokens": 114765243.0, "step": 3020 }, { "epoch": 0.38430225162193105, "ewc_loss": 0.09912109375, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.372120380401611, "learning_rate": 1e-05, "loss": 0.5518, "mean_token_accuracy": 0.8542327880859375, "num_tokens": 114806341.0, "step": 3021 }, { "epoch": 0.3844294619005216, "ewc_loss": 0.0849609375, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.386198997497559, "learning_rate": 1e-05, "loss": 0.6282, "mean_token_accuracy": 0.826977014541626, "num_tokens": 114850014.0, "step": 3022 }, { "epoch": 0.38455667217911205, "ewc_loss": 0.08984375, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 6.063527584075928, "learning_rate": 1e-05, "loss": 0.5639, "mean_token_accuracy": 0.8465030193328857, "num_tokens": 114884771.0, "step": 3023 }, { "epoch": 0.3846838824577026, "ewc_loss": 0.08642578125, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.5144195556640625, "learning_rate": 1e-05, "loss": 0.5123, "mean_token_accuracy": 0.8576252460479736, "num_tokens": 114926319.0, "step": 3024 }, { "epoch": 0.3848110927362931, "ewc_loss": 0.080078125, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 7.390975952148438e-05, "grad_norm": 5.61712646484375, "learning_rate": 1e-05, "loss": 0.5593, "mean_token_accuracy": 0.8434129953384399, "num_tokens": 114971008.0, "step": 3025 }, { "epoch": 0.3849383030148836, "ewc_loss": 0.08544921875, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.536962509155273, "learning_rate": 1e-05, "loss": 0.5436, "mean_token_accuracy": 0.8491197824478149, "num_tokens": 115007294.0, "step": 3026 }, { "epoch": 0.3850655132934741, "ewc_loss": 0.07666015625, "ewc_loss_diag": 6.246566772460938e-05, "ewc_loss_parallel": 7.009506225585938e-05, "grad_norm": 5.231951713562012, "learning_rate": 1e-05, "loss": 0.5063, "mean_token_accuracy": 0.8556520938873291, "num_tokens": 115043066.0, "step": 3027 }, { "epoch": 0.38519272357206463, "ewc_loss": 0.083984375, "ewc_loss_diag": 6.246566772460938e-05, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.617519855499268, "learning_rate": 1e-05, "loss": 0.5109, "mean_token_accuracy": 0.8556461334228516, "num_tokens": 115077404.0, "step": 3028 }, { "epoch": 0.3853199338506551, "ewc_loss": 0.0732421875, "ewc_loss_diag": 6.246566772460938e-05, "ewc_loss_parallel": 6.67572021484375e-05, "grad_norm": 4.951335906982422, "learning_rate": 1e-05, "loss": 0.4827, "mean_token_accuracy": 0.8631125688552856, "num_tokens": 115114092.0, "step": 3029 }, { "epoch": 0.38544714412924563, "ewc_loss": 0.0859375, "ewc_loss_diag": 6.246566772460938e-05, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.791630744934082, "learning_rate": 1e-05, "loss": 0.4899, "mean_token_accuracy": 0.8640559911727905, "num_tokens": 115157095.0, "step": 3030 }, { "epoch": 0.38557435440783616, "ewc_loss": 0.072265625, "ewc_loss_diag": 6.246566772460938e-05, "ewc_loss_parallel": 6.580352783203125e-05, "grad_norm": 4.817887306213379, "learning_rate": 1e-05, "loss": 0.4774, "mean_token_accuracy": 0.8654764294624329, "num_tokens": 115189752.0, "step": 3031 }, { "epoch": 0.3857015646864267, "ewc_loss": 0.09033203125, "ewc_loss_diag": 6.246566772460938e-05, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 6.027996063232422, "learning_rate": 1e-05, "loss": 0.611, "mean_token_accuracy": 0.8304269313812256, "num_tokens": 115232650.0, "step": 3032 }, { "epoch": 0.38582877496501716, "ewc_loss": 0.0732421875, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 6.67572021484375e-05, "grad_norm": 4.851404190063477, "learning_rate": 1e-05, "loss": 0.4852, "mean_token_accuracy": 0.8644452095031738, "num_tokens": 115267812.0, "step": 3033 }, { "epoch": 0.3859559852436077, "ewc_loss": 0.09521484375, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.23825216293335, "learning_rate": 1e-05, "loss": 0.5104, "mean_token_accuracy": 0.8586643934249878, "num_tokens": 115300905.0, "step": 3034 }, { "epoch": 0.3860831955221982, "ewc_loss": 0.07763671875, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 7.152557373046875e-05, "grad_norm": 5.046939849853516, "learning_rate": 1e-05, "loss": 0.4927, "mean_token_accuracy": 0.8623148202896118, "num_tokens": 115341869.0, "step": 3035 }, { "epoch": 0.3862104058007887, "ewc_loss": 0.0966796875, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.259977340698242, "learning_rate": 1e-05, "loss": 0.6163, "mean_token_accuracy": 0.8300389051437378, "num_tokens": 115380499.0, "step": 3036 }, { "epoch": 0.3863376160793792, "ewc_loss": 0.08251953125, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 7.62939453125e-05, "grad_norm": 5.284844875335693, "learning_rate": 1e-05, "loss": 0.4835, "mean_token_accuracy": 0.861257016658783, "num_tokens": 115413573.0, "step": 3037 }, { "epoch": 0.38646482635796975, "ewc_loss": 0.09033203125, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 6.0707597732543945, "learning_rate": 1e-05, "loss": 0.5107, "mean_token_accuracy": 0.8608862161636353, "num_tokens": 115452634.0, "step": 3038 }, { "epoch": 0.3865920366365602, "ewc_loss": 0.0830078125, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.37404727935791, "learning_rate": 1e-05, "loss": 0.6196, "mean_token_accuracy": 0.8278477787971497, "num_tokens": 115492509.0, "step": 3039 }, { "epoch": 0.38671924691515075, "ewc_loss": 0.08447265625, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.79125452041626, "learning_rate": 1e-05, "loss": 0.4916, "mean_token_accuracy": 0.8640053272247314, "num_tokens": 115531256.0, "step": 3040 }, { "epoch": 0.3868464571937413, "ewc_loss": 0.08203125, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 7.581710815429688e-05, "grad_norm": 5.353687286376953, "learning_rate": 1e-05, "loss": 0.5357, "mean_token_accuracy": 0.8440170288085938, "num_tokens": 115564173.0, "step": 3041 }, { "epoch": 0.38697366747233175, "ewc_loss": 0.07958984375, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 7.343292236328125e-05, "grad_norm": 5.547562599182129, "learning_rate": 1e-05, "loss": 0.5167, "mean_token_accuracy": 0.8576441407203674, "num_tokens": 115598359.0, "step": 3042 }, { "epoch": 0.3871008777509223, "ewc_loss": 0.0791015625, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 7.295608520507812e-05, "grad_norm": 5.285994529724121, "learning_rate": 1e-05, "loss": 0.5429, "mean_token_accuracy": 0.8506151437759399, "num_tokens": 115637782.0, "step": 3043 }, { "epoch": 0.3872280880295128, "ewc_loss": 0.07763671875, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 7.152557373046875e-05, "grad_norm": 5.390904903411865, "learning_rate": 1e-05, "loss": 0.5817, "mean_token_accuracy": 0.8368508815765381, "num_tokens": 115680809.0, "step": 3044 }, { "epoch": 0.3873552983081033, "ewc_loss": 0.0771484375, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 7.05718994140625e-05, "grad_norm": 5.206631183624268, "learning_rate": 1e-05, "loss": 0.5277, "mean_token_accuracy": 0.8492093682289124, "num_tokens": 115712092.0, "step": 3045 }, { "epoch": 0.3874825085866938, "ewc_loss": 0.0771484375, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 7.05718994140625e-05, "grad_norm": 5.331400394439697, "learning_rate": 1e-05, "loss": 0.5279, "mean_token_accuracy": 0.8489503860473633, "num_tokens": 115749400.0, "step": 3046 }, { "epoch": 0.38760971886528434, "ewc_loss": 0.07568359375, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 6.914138793945312e-05, "grad_norm": 5.124701023101807, "learning_rate": 1e-05, "loss": 0.5391, "mean_token_accuracy": 0.8461002111434937, "num_tokens": 115783932.0, "step": 3047 }, { "epoch": 0.3877369291438748, "ewc_loss": 0.0771484375, "ewc_loss_diag": 6.389617919921875e-05, "ewc_loss_parallel": 7.05718994140625e-05, "grad_norm": 5.327774524688721, "learning_rate": 1e-05, "loss": 0.526, "mean_token_accuracy": 0.852114200592041, "num_tokens": 115819358.0, "step": 3048 }, { "epoch": 0.38786413942246534, "ewc_loss": 0.07373046875, "ewc_loss_diag": 6.389617919921875e-05, "ewc_loss_parallel": 6.723403930664062e-05, "grad_norm": 5.020650863647461, "learning_rate": 1e-05, "loss": 0.4752, "mean_token_accuracy": 0.8625549674034119, "num_tokens": 115855370.0, "step": 3049 }, { "epoch": 0.38799134970105587, "ewc_loss": 0.07763671875, "ewc_loss_diag": 6.389617919921875e-05, "ewc_loss_parallel": 7.152557373046875e-05, "grad_norm": 5.3916144371032715, "learning_rate": 1e-05, "loss": 0.5788, "mean_token_accuracy": 0.8346683979034424, "num_tokens": 115894722.0, "step": 3050 }, { "epoch": 0.38811855997964634, "ewc_loss": 0.0732421875, "ewc_loss_diag": 6.389617919921875e-05, "ewc_loss_parallel": 6.67572021484375e-05, "grad_norm": 4.903982162475586, "learning_rate": 1e-05, "loss": 0.4933, "mean_token_accuracy": 0.8580518364906311, "num_tokens": 115935276.0, "step": 3051 }, { "epoch": 0.38824577025823687, "ewc_loss": 0.0810546875, "ewc_loss_diag": 6.389617919921875e-05, "ewc_loss_parallel": 7.486343383789062e-05, "grad_norm": 5.55285120010376, "learning_rate": 1e-05, "loss": 0.5315, "mean_token_accuracy": 0.850872814655304, "num_tokens": 115971255.0, "step": 3052 }, { "epoch": 0.3883729805368274, "ewc_loss": 0.0712890625, "ewc_loss_diag": 6.389617919921875e-05, "ewc_loss_parallel": 6.4849853515625e-05, "grad_norm": 4.772159099578857, "learning_rate": 1e-05, "loss": 0.4787, "mean_token_accuracy": 0.863303542137146, "num_tokens": 116003818.0, "step": 3053 }, { "epoch": 0.38850019081541787, "ewc_loss": 0.08642578125, "ewc_loss_diag": 6.389617919921875e-05, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.825428485870361, "learning_rate": 1e-05, "loss": 0.5164, "mean_token_accuracy": 0.8567854166030884, "num_tokens": 116040720.0, "step": 3054 }, { "epoch": 0.3886274010940084, "ewc_loss": 0.07080078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 6.437301635742188e-05, "grad_norm": 4.671916961669922, "learning_rate": 1e-05, "loss": 0.4758, "mean_token_accuracy": 0.8657573461532593, "num_tokens": 116076485.0, "step": 3055 }, { "epoch": 0.3887546113725989, "ewc_loss": 0.0947265625, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 6.208216190338135, "learning_rate": 1e-05, "loss": 0.5, "mean_token_accuracy": 0.8643808960914612, "num_tokens": 116109802.0, "step": 3056 }, { "epoch": 0.3888818216511894, "ewc_loss": 0.07421875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 6.771087646484375e-05, "grad_norm": 4.788302898406982, "learning_rate": 1e-05, "loss": 0.5242, "mean_token_accuracy": 0.8510944247245789, "num_tokens": 116149865.0, "step": 3057 }, { "epoch": 0.3890090319297799, "ewc_loss": 0.10205078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.537350177764893, "learning_rate": 1e-05, "loss": 0.5466, "mean_token_accuracy": 0.8543120622634888, "num_tokens": 116185887.0, "step": 3058 }, { "epoch": 0.38913624220837045, "ewc_loss": 0.08349609375, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.221715450286865, "learning_rate": 1e-05, "loss": 0.5711, "mean_token_accuracy": 0.844070315361023, "num_tokens": 116233090.0, "step": 3059 }, { "epoch": 0.3892634524869609, "ewc_loss": 0.1005859375, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.442220211029053, "learning_rate": 1e-05, "loss": 0.562, "mean_token_accuracy": 0.848442792892456, "num_tokens": 116274433.0, "step": 3060 }, { "epoch": 0.38939066276555145, "ewc_loss": 0.08935546875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.578060150146484, "learning_rate": 1e-05, "loss": 0.6091, "mean_token_accuracy": 0.8338048458099365, "num_tokens": 116309104.0, "step": 3061 }, { "epoch": 0.389517873044142, "ewc_loss": 0.08837890625, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 6.034514427185059, "learning_rate": 1e-05, "loss": 0.5308, "mean_token_accuracy": 0.8586251735687256, "num_tokens": 116350248.0, "step": 3062 }, { "epoch": 0.38964508332273246, "ewc_loss": 0.08740234375, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.579895973205566, "learning_rate": 1e-05, "loss": 0.5654, "mean_token_accuracy": 0.8405880928039551, "num_tokens": 116391078.0, "step": 3063 }, { "epoch": 0.389772293601323, "ewc_loss": 0.080078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 7.390975952148438e-05, "grad_norm": 5.634279727935791, "learning_rate": 1e-05, "loss": 0.5236, "mean_token_accuracy": 0.849164605140686, "num_tokens": 116430652.0, "step": 3064 }, { "epoch": 0.3898995038799135, "ewc_loss": 0.0849609375, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.521462440490723, "learning_rate": 1e-05, "loss": 0.4982, "mean_token_accuracy": 0.8640756607055664, "num_tokens": 116467024.0, "step": 3065 }, { "epoch": 0.390026714158504, "ewc_loss": 0.0771484375, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 7.05718994140625e-05, "grad_norm": 5.320802211761475, "learning_rate": 1e-05, "loss": 0.4945, "mean_token_accuracy": 0.8609259128570557, "num_tokens": 116504304.0, "step": 3066 }, { "epoch": 0.3901539244370945, "ewc_loss": 0.08154296875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 7.534027099609375e-05, "grad_norm": 5.505566596984863, "learning_rate": 1e-05, "loss": 0.5387, "mean_token_accuracy": 0.8465456366539001, "num_tokens": 116541079.0, "step": 3067 }, { "epoch": 0.39028113471568504, "ewc_loss": 0.07568359375, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 6.914138793945312e-05, "grad_norm": 5.117391586303711, "learning_rate": 1e-05, "loss": 0.5653, "mean_token_accuracy": 0.8430051207542419, "num_tokens": 116578690.0, "step": 3068 }, { "epoch": 0.3904083449942755, "ewc_loss": 0.080078125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 7.390975952148438e-05, "grad_norm": 5.538730621337891, "learning_rate": 1e-05, "loss": 0.5216, "mean_token_accuracy": 0.8510199189186096, "num_tokens": 116614945.0, "step": 3069 }, { "epoch": 0.39053555527286604, "ewc_loss": 0.0732421875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 6.67572021484375e-05, "grad_norm": 4.9755120277404785, "learning_rate": 1e-05, "loss": 0.4741, "mean_token_accuracy": 0.8663306832313538, "num_tokens": 116646795.0, "step": 3070 }, { "epoch": 0.39066276555145657, "ewc_loss": 0.08251953125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 7.62939453125e-05, "grad_norm": 5.6146464347839355, "learning_rate": 1e-05, "loss": 0.5506, "mean_token_accuracy": 0.8453156352043152, "num_tokens": 116685731.0, "step": 3071 }, { "epoch": 0.39078997583004704, "ewc_loss": 0.07275390625, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 6.628036499023438e-05, "grad_norm": 4.859371662139893, "learning_rate": 1e-05, "loss": 0.4378, "mean_token_accuracy": 0.8721715211868286, "num_tokens": 116722559.0, "step": 3072 }, { "epoch": 0.39091718610863757, "ewc_loss": 0.08447265625, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.790205001831055, "learning_rate": 1e-05, "loss": 0.588, "mean_token_accuracy": 0.8356837034225464, "num_tokens": 116762954.0, "step": 3073 }, { "epoch": 0.3910443963872281, "ewc_loss": 0.072265625, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 6.580352783203125e-05, "grad_norm": 4.778357982635498, "learning_rate": 1e-05, "loss": 0.4888, "mean_token_accuracy": 0.8612440824508667, "num_tokens": 116801745.0, "step": 3074 }, { "epoch": 0.39117160666581857, "ewc_loss": 0.0908203125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 6.043780326843262, "learning_rate": 1e-05, "loss": 0.5022, "mean_token_accuracy": 0.8656158447265625, "num_tokens": 116837650.0, "step": 3075 }, { "epoch": 0.3912988169444091, "ewc_loss": 0.0732421875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 6.67572021484375e-05, "grad_norm": 4.8201212882995605, "learning_rate": 1e-05, "loss": 0.5513, "mean_token_accuracy": 0.8418987393379211, "num_tokens": 116872722.0, "step": 3076 }, { "epoch": 0.39142602722299963, "ewc_loss": 0.09716796875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.283580780029297, "learning_rate": 1e-05, "loss": 0.5291, "mean_token_accuracy": 0.855617105960846, "num_tokens": 116911175.0, "step": 3077 }, { "epoch": 0.3915532375015901, "ewc_loss": 0.0791015625, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 7.295608520507812e-05, "grad_norm": 5.034143924713135, "learning_rate": 1e-05, "loss": 0.5208, "mean_token_accuracy": 0.8502988815307617, "num_tokens": 116946317.0, "step": 3078 }, { "epoch": 0.39168044778018063, "ewc_loss": 0.09814453125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.321014881134033, "learning_rate": 1e-05, "loss": 0.6025, "mean_token_accuracy": 0.8356673717498779, "num_tokens": 116988661.0, "step": 3079 }, { "epoch": 0.39180765805877116, "ewc_loss": 0.08447265625, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.30298376083374, "learning_rate": 1e-05, "loss": 0.5013, "mean_token_accuracy": 0.85792076587677, "num_tokens": 117027309.0, "step": 3080 }, { "epoch": 0.3919348683373617, "ewc_loss": 0.091796875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 6.100845813751221, "learning_rate": 1e-05, "loss": 0.5293, "mean_token_accuracy": 0.8558464050292969, "num_tokens": 117065608.0, "step": 3081 }, { "epoch": 0.39206207861595216, "ewc_loss": 0.08447265625, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.429636478424072, "learning_rate": 1e-05, "loss": 0.5461, "mean_token_accuracy": 0.8486060500144958, "num_tokens": 117101824.0, "step": 3082 }, { "epoch": 0.3921892888945427, "ewc_loss": 0.08349609375, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.767026424407959, "learning_rate": 1e-05, "loss": 0.5507, "mean_token_accuracy": 0.8473239541053772, "num_tokens": 117141985.0, "step": 3083 }, { "epoch": 0.3923164991731332, "ewc_loss": 0.08349609375, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.431141376495361, "learning_rate": 1e-05, "loss": 0.5036, "mean_token_accuracy": 0.8592954874038696, "num_tokens": 117177606.0, "step": 3084 }, { "epoch": 0.3924437094517237, "ewc_loss": 0.0791015625, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 7.295608520507812e-05, "grad_norm": 5.449741363525391, "learning_rate": 1e-05, "loss": 0.5169, "mean_token_accuracy": 0.853064775466919, "num_tokens": 117220214.0, "step": 3085 }, { "epoch": 0.3925709197303142, "ewc_loss": 0.08154296875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 7.534027099609375e-05, "grad_norm": 5.414709568023682, "learning_rate": 1e-05, "loss": 0.5448, "mean_token_accuracy": 0.8484057188034058, "num_tokens": 117259188.0, "step": 3086 }, { "epoch": 0.39269813000890474, "ewc_loss": 0.076171875, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 6.961822509765625e-05, "grad_norm": 5.204240322113037, "learning_rate": 1e-05, "loss": 0.5337, "mean_token_accuracy": 0.8508628606796265, "num_tokens": 117299146.0, "step": 3087 }, { "epoch": 0.3928253402874952, "ewc_loss": 0.07861328125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 7.2479248046875e-05, "grad_norm": 5.420749664306641, "learning_rate": 1e-05, "loss": 0.5283, "mean_token_accuracy": 0.8510396480560303, "num_tokens": 117337560.0, "step": 3088 }, { "epoch": 0.39295255056608575, "ewc_loss": 0.07470703125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 6.818771362304688e-05, "grad_norm": 5.048837661743164, "learning_rate": 1e-05, "loss": 0.5702, "mean_token_accuracy": 0.8391516208648682, "num_tokens": 117376935.0, "step": 3089 }, { "epoch": 0.3930797608446763, "ewc_loss": 0.0791015625, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 7.295608520507812e-05, "grad_norm": 5.444973468780518, "learning_rate": 1e-05, "loss": 0.5368, "mean_token_accuracy": 0.8536189794540405, "num_tokens": 117420273.0, "step": 3090 }, { "epoch": 0.39320697112326675, "ewc_loss": 0.07373046875, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 6.723403930664062e-05, "grad_norm": 4.93060302734375, "learning_rate": 1e-05, "loss": 0.4643, "mean_token_accuracy": 0.8671623468399048, "num_tokens": 117457338.0, "step": 3091 }, { "epoch": 0.3933341814018573, "ewc_loss": 0.0810546875, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 7.486343383789062e-05, "grad_norm": 5.552751064300537, "learning_rate": 1e-05, "loss": 0.4885, "mean_token_accuracy": 0.8636714220046997, "num_tokens": 117491082.0, "step": 3092 }, { "epoch": 0.3934613916804478, "ewc_loss": 0.0732421875, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 6.67572021484375e-05, "grad_norm": 4.821465015411377, "learning_rate": 1e-05, "loss": 0.4193, "mean_token_accuracy": 0.8814050555229187, "num_tokens": 117520425.0, "step": 3093 }, { "epoch": 0.3935886019590383, "ewc_loss": 0.08447265625, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.7435712814331055, "learning_rate": 1e-05, "loss": 0.4942, "mean_token_accuracy": 0.8639613389968872, "num_tokens": 117559021.0, "step": 3094 }, { "epoch": 0.3937158122376288, "ewc_loss": 0.0712890625, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 6.4849853515625e-05, "grad_norm": 4.726312637329102, "learning_rate": 1e-05, "loss": 0.4895, "mean_token_accuracy": 0.8615994453430176, "num_tokens": 117598705.0, "step": 3095 }, { "epoch": 0.39384302251621933, "ewc_loss": 0.0908203125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 6.0442681312561035, "learning_rate": 1e-05, "loss": 0.516, "mean_token_accuracy": 0.8650552034378052, "num_tokens": 117636185.0, "step": 3096 }, { "epoch": 0.3939702327948098, "ewc_loss": 0.0732421875, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 6.67572021484375e-05, "grad_norm": 4.73423957824707, "learning_rate": 1e-05, "loss": 0.4724, "mean_token_accuracy": 0.8655053377151489, "num_tokens": 117674290.0, "step": 3097 }, { "epoch": 0.39409744307340033, "ewc_loss": 0.09912109375, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.400360107421875, "learning_rate": 1e-05, "loss": 0.5291, "mean_token_accuracy": 0.8566914200782776, "num_tokens": 117708708.0, "step": 3098 }, { "epoch": 0.39422465335199086, "ewc_loss": 0.080078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 7.390975952148438e-05, "grad_norm": 5.006027698516846, "learning_rate": 1e-05, "loss": 0.5314, "mean_token_accuracy": 0.8521435856819153, "num_tokens": 117742528.0, "step": 3099 }, { "epoch": 0.39435186363058133, "ewc_loss": 0.10205078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.524288177490234, "learning_rate": 1e-05, "loss": 0.5804, "mean_token_accuracy": 0.8407058119773865, "num_tokens": 117779050.0, "step": 3100 }, { "epoch": 0.39447907390917186, "ewc_loss": 0.087890625, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.401453018188477, "learning_rate": 1e-05, "loss": 0.5685, "mean_token_accuracy": 0.8414509296417236, "num_tokens": 117824004.0, "step": 3101 }, { "epoch": 0.3946062841877624, "ewc_loss": 0.0966796875, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.266216278076172, "learning_rate": 1e-05, "loss": 0.4945, "mean_token_accuracy": 0.8640098571777344, "num_tokens": 117864837.0, "step": 3102 }, { "epoch": 0.39473349446635286, "ewc_loss": 0.08935546875, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.56359338760376, "learning_rate": 1e-05, "loss": 0.5081, "mean_token_accuracy": 0.8569063544273376, "num_tokens": 117903675.0, "step": 3103 }, { "epoch": 0.3948607047449434, "ewc_loss": 0.08544921875, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.88559103012085, "learning_rate": 1e-05, "loss": 0.5385, "mean_token_accuracy": 0.8489726185798645, "num_tokens": 117942561.0, "step": 3104 }, { "epoch": 0.3949879150235339, "ewc_loss": 0.0859375, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.528034210205078, "learning_rate": 1e-05, "loss": 0.5595, "mean_token_accuracy": 0.8473365902900696, "num_tokens": 117982439.0, "step": 3105 }, { "epoch": 0.3951151253021244, "ewc_loss": 0.07958984375, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 7.343292236328125e-05, "grad_norm": 5.519136905670166, "learning_rate": 1e-05, "loss": 0.5371, "mean_token_accuracy": 0.8500546813011169, "num_tokens": 118025153.0, "step": 3106 }, { "epoch": 0.3952423355807149, "ewc_loss": 0.0830078125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.478204250335693, "learning_rate": 1e-05, "loss": 0.5421, "mean_token_accuracy": 0.8473242521286011, "num_tokens": 118064508.0, "step": 3107 }, { "epoch": 0.39536954585930545, "ewc_loss": 0.0771484375, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 7.05718994140625e-05, "grad_norm": 5.26990270614624, "learning_rate": 1e-05, "loss": 0.4928, "mean_token_accuracy": 0.8578391671180725, "num_tokens": 118093368.0, "step": 3108 }, { "epoch": 0.3954967561378959, "ewc_loss": 0.07958984375, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 7.343292236328125e-05, "grad_norm": 5.46589469909668, "learning_rate": 1e-05, "loss": 0.4748, "mean_token_accuracy": 0.8652846813201904, "num_tokens": 118129164.0, "step": 3109 }, { "epoch": 0.39562396641648645, "ewc_loss": 0.07568359375, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 6.914138793945312e-05, "grad_norm": 5.073045253753662, "learning_rate": 1e-05, "loss": 0.4897, "mean_token_accuracy": 0.8602323532104492, "num_tokens": 118169092.0, "step": 3110 }, { "epoch": 0.395751176695077, "ewc_loss": 0.07958984375, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 7.343292236328125e-05, "grad_norm": 5.497676849365234, "learning_rate": 1e-05, "loss": 0.519, "mean_token_accuracy": 0.8546280264854431, "num_tokens": 118211616.0, "step": 3111 }, { "epoch": 0.39587838697366745, "ewc_loss": 0.07373046875, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 6.723403930664062e-05, "grad_norm": 4.941631317138672, "learning_rate": 1e-05, "loss": 0.5275, "mean_token_accuracy": 0.8469427824020386, "num_tokens": 118248483.0, "step": 3112 }, { "epoch": 0.396005597252258, "ewc_loss": 0.08203125, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 7.581710815429688e-05, "grad_norm": 5.61000394821167, "learning_rate": 1e-05, "loss": 0.5324, "mean_token_accuracy": 0.8536186814308167, "num_tokens": 118288540.0, "step": 3113 }, { "epoch": 0.3961328075308485, "ewc_loss": 0.072265625, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 6.580352783203125e-05, "grad_norm": 4.820995330810547, "learning_rate": 1e-05, "loss": 0.5203, "mean_token_accuracy": 0.8541000485420227, "num_tokens": 118326988.0, "step": 3114 }, { "epoch": 0.396260017809439, "ewc_loss": 0.0859375, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.8157830238342285, "learning_rate": 1e-05, "loss": 0.5287, "mean_token_accuracy": 0.8549957275390625, "num_tokens": 118366100.0, "step": 3115 }, { "epoch": 0.3963872280880295, "ewc_loss": 0.072265625, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 6.580352783203125e-05, "grad_norm": 4.736359119415283, "learning_rate": 1e-05, "loss": 0.5151, "mean_token_accuracy": 0.8512094020843506, "num_tokens": 118397243.0, "step": 3116 }, { "epoch": 0.39651443836662004, "ewc_loss": 0.0927734375, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 6.133594989776611, "learning_rate": 1e-05, "loss": 0.5972, "mean_token_accuracy": 0.8369624614715576, "num_tokens": 118438198.0, "step": 3117 }, { "epoch": 0.3966416486452105, "ewc_loss": 0.07421875, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 6.771087646484375e-05, "grad_norm": 4.790711402893066, "learning_rate": 1e-05, "loss": 0.5233, "mean_token_accuracy": 0.8503392934799194, "num_tokens": 118473496.0, "step": 3118 }, { "epoch": 0.39676885892380104, "ewc_loss": 0.1005859375, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.445350646972656, "learning_rate": 1e-05, "loss": 0.5221, "mean_token_accuracy": 0.8631000518798828, "num_tokens": 118511988.0, "step": 3119 }, { "epoch": 0.39689606920239157, "ewc_loss": 0.08056640625, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 7.43865966796875e-05, "grad_norm": 5.126845359802246, "learning_rate": 1e-05, "loss": 0.4855, "mean_token_accuracy": 0.8617596626281738, "num_tokens": 118548287.0, "step": 3120 }, { "epoch": 0.39702327948098204, "ewc_loss": 0.1005859375, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.452480792999268, "learning_rate": 1e-05, "loss": 0.5212, "mean_token_accuracy": 0.8614612817764282, "num_tokens": 118580637.0, "step": 3121 }, { "epoch": 0.39715048975957257, "ewc_loss": 0.087890625, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.458426475524902, "learning_rate": 1e-05, "loss": 0.555, "mean_token_accuracy": 0.8443132638931274, "num_tokens": 118622316.0, "step": 3122 }, { "epoch": 0.3972777000381631, "ewc_loss": 0.0927734375, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 6.148349285125732, "learning_rate": 1e-05, "loss": 0.569, "mean_token_accuracy": 0.8418763875961304, "num_tokens": 118660497.0, "step": 3123 }, { "epoch": 0.39740491031675357, "ewc_loss": 0.0888671875, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.559386730194092, "learning_rate": 1e-05, "loss": 0.5401, "mean_token_accuracy": 0.8480703830718994, "num_tokens": 118691733.0, "step": 3124 }, { "epoch": 0.3975321205953441, "ewc_loss": 0.0830078125, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.764395713806152, "learning_rate": 1e-05, "loss": 0.5469, "mean_token_accuracy": 0.8417509198188782, "num_tokens": 118728576.0, "step": 3125 }, { "epoch": 0.3976593308739346, "ewc_loss": 0.083984375, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.508819103240967, "learning_rate": 1e-05, "loss": 0.5706, "mean_token_accuracy": 0.8417503833770752, "num_tokens": 118765394.0, "step": 3126 }, { "epoch": 0.3977865411525251, "ewc_loss": 0.078125, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 7.200241088867188e-05, "grad_norm": 5.432075500488281, "learning_rate": 1e-05, "loss": 0.5483, "mean_token_accuracy": 0.8403623104095459, "num_tokens": 118804856.0, "step": 3127 }, { "epoch": 0.3979137514311156, "ewc_loss": 0.08251953125, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 7.62939453125e-05, "grad_norm": 5.483281135559082, "learning_rate": 1e-05, "loss": 0.5258, "mean_token_accuracy": 0.8531191349029541, "num_tokens": 118839274.0, "step": 3128 }, { "epoch": 0.39804096170970615, "ewc_loss": 0.0771484375, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 7.05718994140625e-05, "grad_norm": 5.197373390197754, "learning_rate": 1e-05, "loss": 0.5513, "mean_token_accuracy": 0.8427245616912842, "num_tokens": 118877866.0, "step": 3129 }, { "epoch": 0.3981681719882967, "ewc_loss": 0.080078125, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 7.390975952148438e-05, "grad_norm": 5.468301773071289, "learning_rate": 1e-05, "loss": 0.4777, "mean_token_accuracy": 0.8675696849822998, "num_tokens": 118908942.0, "step": 3130 }, { "epoch": 0.39829538226688715, "ewc_loss": 0.0751953125, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 6.866455078125e-05, "grad_norm": 5.105275630950928, "learning_rate": 1e-05, "loss": 0.5641, "mean_token_accuracy": 0.8381938934326172, "num_tokens": 118948957.0, "step": 3131 }, { "epoch": 0.3984225925454777, "ewc_loss": 0.07958984375, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 7.343292236328125e-05, "grad_norm": 5.44949197769165, "learning_rate": 1e-05, "loss": 0.5082, "mean_token_accuracy": 0.858771562576294, "num_tokens": 118991246.0, "step": 3132 }, { "epoch": 0.3985498028240682, "ewc_loss": 0.0751953125, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 6.866455078125e-05, "grad_norm": 4.973724842071533, "learning_rate": 1e-05, "loss": 0.5613, "mean_token_accuracy": 0.8373799324035645, "num_tokens": 119025661.0, "step": 3133 }, { "epoch": 0.3986770131026587, "ewc_loss": 0.08056640625, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 7.43865966796875e-05, "grad_norm": 5.519426345825195, "learning_rate": 1e-05, "loss": 0.5249, "mean_token_accuracy": 0.8485535979270935, "num_tokens": 119059524.0, "step": 3134 }, { "epoch": 0.3988042233812492, "ewc_loss": 0.0732421875, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 6.67572021484375e-05, "grad_norm": 4.8560791015625, "learning_rate": 1e-05, "loss": 0.5552, "mean_token_accuracy": 0.8431401252746582, "num_tokens": 119101636.0, "step": 3135 }, { "epoch": 0.39893143365983974, "ewc_loss": 0.08349609375, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.688307285308838, "learning_rate": 1e-05, "loss": 0.5436, "mean_token_accuracy": 0.8454240560531616, "num_tokens": 119138448.0, "step": 3136 }, { "epoch": 0.3990586439384302, "ewc_loss": 0.07275390625, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 6.628036499023438e-05, "grad_norm": 4.734712600708008, "learning_rate": 1e-05, "loss": 0.544, "mean_token_accuracy": 0.8407135009765625, "num_tokens": 119183337.0, "step": 3137 }, { "epoch": 0.39918585421702074, "ewc_loss": 0.09033203125, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 6.002991199493408, "learning_rate": 1e-05, "loss": 0.4567, "mean_token_accuracy": 0.8780407905578613, "num_tokens": 119217503.0, "step": 3138 }, { "epoch": 0.39931306449561127, "ewc_loss": 0.0732421875, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 6.67572021484375e-05, "grad_norm": 4.709651470184326, "learning_rate": 1e-05, "loss": 0.5729, "mean_token_accuracy": 0.8397613763809204, "num_tokens": 119255824.0, "step": 3139 }, { "epoch": 0.39944027477420174, "ewc_loss": 0.10009765625, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.389842510223389, "learning_rate": 1e-05, "loss": 0.5594, "mean_token_accuracy": 0.8490345478057861, "num_tokens": 119292677.0, "step": 3140 }, { "epoch": 0.39956748505279227, "ewc_loss": 0.07861328125, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 7.2479248046875e-05, "grad_norm": 4.945178508758545, "learning_rate": 1e-05, "loss": 0.5165, "mean_token_accuracy": 0.8554484248161316, "num_tokens": 119330961.0, "step": 3141 }, { "epoch": 0.3996946953313828, "ewc_loss": 0.10400390625, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.616062641143799, "learning_rate": 1e-05, "loss": 0.5882, "mean_token_accuracy": 0.8417755961418152, "num_tokens": 119362122.0, "step": 3142 }, { "epoch": 0.39982190560997327, "ewc_loss": 0.08837890625, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.416677474975586, "learning_rate": 1e-05, "loss": 0.5161, "mean_token_accuracy": 0.8546401262283325, "num_tokens": 119398018.0, "step": 3143 }, { "epoch": 0.3999491158885638, "ewc_loss": 0.10009765625, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.37968111038208, "learning_rate": 1e-05, "loss": 0.5788, "mean_token_accuracy": 0.8430354595184326, "num_tokens": 119438552.0, "step": 3144 }, { "epoch": 0.40007632616715433, "ewc_loss": 0.09130859375, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.635354995727539, "learning_rate": 1e-05, "loss": 0.5461, "mean_token_accuracy": 0.8511389493942261, "num_tokens": 119470777.0, "step": 3145 }, { "epoch": 0.4002035364457448, "ewc_loss": 0.08837890625, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.9674224853515625, "learning_rate": 1e-05, "loss": 0.5047, "mean_token_accuracy": 0.8647810220718384, "num_tokens": 119505062.0, "step": 3146 }, { "epoch": 0.40033074672433533, "ewc_loss": 0.087890625, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.5922441482543945, "learning_rate": 1e-05, "loss": 0.5419, "mean_token_accuracy": 0.8475354313850403, "num_tokens": 119540394.0, "step": 3147 }, { "epoch": 0.40045795700292586, "ewc_loss": 0.080078125, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 7.390975952148438e-05, "grad_norm": 5.579368591308594, "learning_rate": 1e-05, "loss": 0.4573, "mean_token_accuracy": 0.8743009567260742, "num_tokens": 119581657.0, "step": 3148 }, { "epoch": 0.40058516728151633, "ewc_loss": 0.08447265625, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.526394844055176, "learning_rate": 1e-05, "loss": 0.5004, "mean_token_accuracy": 0.8593153357505798, "num_tokens": 119617140.0, "step": 3149 }, { "epoch": 0.40071237756010686, "ewc_loss": 0.076171875, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 6.961822509765625e-05, "grad_norm": 5.310901641845703, "learning_rate": 1e-05, "loss": 0.5755, "mean_token_accuracy": 0.837533175945282, "num_tokens": 119660155.0, "step": 3150 }, { "epoch": 0.4008395878386974, "ewc_loss": 0.0810546875, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 7.486343383789062e-05, "grad_norm": 5.482112407684326, "learning_rate": 1e-05, "loss": 0.5644, "mean_token_accuracy": 0.8403122425079346, "num_tokens": 119706258.0, "step": 3151 }, { "epoch": 0.40096679811728786, "ewc_loss": 0.076171875, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 6.961822509765625e-05, "grad_norm": 5.125463008880615, "learning_rate": 1e-05, "loss": 0.5085, "mean_token_accuracy": 0.8553849458694458, "num_tokens": 119742564.0, "step": 3152 }, { "epoch": 0.4010940083958784, "ewc_loss": 0.080078125, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 7.390975952148438e-05, "grad_norm": 5.491057872772217, "learning_rate": 1e-05, "loss": 0.5244, "mean_token_accuracy": 0.8539756536483765, "num_tokens": 119781301.0, "step": 3153 }, { "epoch": 0.4012212186744689, "ewc_loss": 0.07470703125, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 6.818771362304688e-05, "grad_norm": 4.997459888458252, "learning_rate": 1e-05, "loss": 0.5335, "mean_token_accuracy": 0.847106397151947, "num_tokens": 119816513.0, "step": 3154 }, { "epoch": 0.4013484289530594, "ewc_loss": 0.08251953125, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 7.62939453125e-05, "grad_norm": 5.565361976623535, "learning_rate": 1e-05, "loss": 0.5186, "mean_token_accuracy": 0.857041597366333, "num_tokens": 119858510.0, "step": 3155 }, { "epoch": 0.4014756392316499, "ewc_loss": 0.0732421875, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 6.67572021484375e-05, "grad_norm": 4.8576579093933105, "learning_rate": 1e-05, "loss": 0.4975, "mean_token_accuracy": 0.855942964553833, "num_tokens": 119902243.0, "step": 3156 }, { "epoch": 0.40160284951024044, "ewc_loss": 0.08447265625, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.759790897369385, "learning_rate": 1e-05, "loss": 0.5031, "mean_token_accuracy": 0.8604354858398438, "num_tokens": 119938956.0, "step": 3157 }, { "epoch": 0.4017300597888309, "ewc_loss": 0.072265625, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 6.532669067382812e-05, "grad_norm": 4.750028133392334, "learning_rate": 1e-05, "loss": 0.4954, "mean_token_accuracy": 0.864001452922821, "num_tokens": 119976070.0, "step": 3158 }, { "epoch": 0.40185727006742145, "ewc_loss": 0.0908203125, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 6.067800998687744, "learning_rate": 1e-05, "loss": 0.5115, "mean_token_accuracy": 0.8583074808120728, "num_tokens": 120011811.0, "step": 3159 }, { "epoch": 0.401984480346012, "ewc_loss": 0.0732421875, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 6.67572021484375e-05, "grad_norm": 4.7646803855896, "learning_rate": 1e-05, "loss": 0.5257, "mean_token_accuracy": 0.8470151424407959, "num_tokens": 120042278.0, "step": 3160 }, { "epoch": 0.40211169062460245, "ewc_loss": 0.099609375, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.399218559265137, "learning_rate": 1e-05, "loss": 0.5411, "mean_token_accuracy": 0.8513869047164917, "num_tokens": 120078379.0, "step": 3161 }, { "epoch": 0.402238900903193, "ewc_loss": 0.0791015625, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 7.295608520507812e-05, "grad_norm": 5.011209487915039, "learning_rate": 1e-05, "loss": 0.4779, "mean_token_accuracy": 0.8648574352264404, "num_tokens": 120118852.0, "step": 3162 }, { "epoch": 0.4023661111817835, "ewc_loss": 0.1025390625, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.536378860473633, "learning_rate": 1e-05, "loss": 0.5234, "mean_token_accuracy": 0.863802433013916, "num_tokens": 120153825.0, "step": 3163 }, { "epoch": 0.402493321460374, "ewc_loss": 0.087890625, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.429824352264404, "learning_rate": 1e-05, "loss": 0.5365, "mean_token_accuracy": 0.8501722812652588, "num_tokens": 120190945.0, "step": 3164 }, { "epoch": 0.4026205317389645, "ewc_loss": 0.09765625, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.270238399505615, "learning_rate": 1e-05, "loss": 0.5115, "mean_token_accuracy": 0.8604881763458252, "num_tokens": 120221435.0, "step": 3165 }, { "epoch": 0.40274774201755503, "ewc_loss": 0.0888671875, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.5929059982299805, "learning_rate": 1e-05, "loss": 0.5146, "mean_token_accuracy": 0.8595449924468994, "num_tokens": 120253553.0, "step": 3166 }, { "epoch": 0.4028749522961455, "ewc_loss": 0.0849609375, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.86570930480957, "learning_rate": 1e-05, "loss": 0.5409, "mean_token_accuracy": 0.8487241268157959, "num_tokens": 120291438.0, "step": 3167 }, { "epoch": 0.40300216257473603, "ewc_loss": 0.0859375, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.550688743591309, "learning_rate": 1e-05, "loss": 0.5416, "mean_token_accuracy": 0.8483726978302002, "num_tokens": 120332005.0, "step": 3168 }, { "epoch": 0.40312937285332656, "ewc_loss": 0.080078125, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 7.343292236328125e-05, "grad_norm": 5.510065078735352, "learning_rate": 1e-05, "loss": 0.5642, "mean_token_accuracy": 0.8371056318283081, "num_tokens": 120369280.0, "step": 3169 }, { "epoch": 0.40325658313191703, "ewc_loss": 0.0830078125, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 7.62939453125e-05, "grad_norm": 5.506494522094727, "learning_rate": 1e-05, "loss": 0.5269, "mean_token_accuracy": 0.8542580604553223, "num_tokens": 120402350.0, "step": 3170 }, { "epoch": 0.40338379341050756, "ewc_loss": 0.0771484375, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 7.009506225585938e-05, "grad_norm": 5.248894691467285, "learning_rate": 1e-05, "loss": 0.515, "mean_token_accuracy": 0.8593541979789734, "num_tokens": 120441791.0, "step": 3171 }, { "epoch": 0.4035110036890981, "ewc_loss": 0.08203125, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 7.534027099609375e-05, "grad_norm": 5.4841837882995605, "learning_rate": 1e-05, "loss": 0.4765, "mean_token_accuracy": 0.8684955835342407, "num_tokens": 120476895.0, "step": 3172 }, { "epoch": 0.40363821396768856, "ewc_loss": 0.0751953125, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 6.866455078125e-05, "grad_norm": 5.082211971282959, "learning_rate": 1e-05, "loss": 0.5377, "mean_token_accuracy": 0.8478226661682129, "num_tokens": 120518016.0, "step": 3173 }, { "epoch": 0.4037654242462791, "ewc_loss": 0.08203125, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 7.534027099609375e-05, "grad_norm": 5.51476526260376, "learning_rate": 1e-05, "loss": 0.4821, "mean_token_accuracy": 0.866837739944458, "num_tokens": 120551500.0, "step": 3174 }, { "epoch": 0.4038926345248696, "ewc_loss": 0.07421875, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 6.771087646484375e-05, "grad_norm": 4.955273151397705, "learning_rate": 1e-05, "loss": 0.4757, "mean_token_accuracy": 0.8616064786911011, "num_tokens": 120588543.0, "step": 3175 }, { "epoch": 0.4040198448034601, "ewc_loss": 0.08349609375, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.629654407501221, "learning_rate": 1e-05, "loss": 0.557, "mean_token_accuracy": 0.8519430160522461, "num_tokens": 120634394.0, "step": 3176 }, { "epoch": 0.4041470550820506, "ewc_loss": 0.0732421875, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 6.628036499023438e-05, "grad_norm": 4.836040496826172, "learning_rate": 1e-05, "loss": 0.5356, "mean_token_accuracy": 0.8487407565116882, "num_tokens": 120672982.0, "step": 3177 }, { "epoch": 0.40427426536064115, "ewc_loss": 0.08740234375, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.842175483703613, "learning_rate": 1e-05, "loss": 0.5983, "mean_token_accuracy": 0.8296271562576294, "num_tokens": 120712091.0, "step": 3178 }, { "epoch": 0.4044014756392316, "ewc_loss": 0.07275390625, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 6.580352783203125e-05, "grad_norm": 4.753148555755615, "learning_rate": 1e-05, "loss": 0.4723, "mean_token_accuracy": 0.8651413917541504, "num_tokens": 120743530.0, "step": 3179 }, { "epoch": 0.40452868591782215, "ewc_loss": 0.09326171875, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 6.143215656280518, "learning_rate": 1e-05, "loss": 0.4943, "mean_token_accuracy": 0.8655441999435425, "num_tokens": 120778884.0, "step": 3180 }, { "epoch": 0.4046558961964127, "ewc_loss": 0.07470703125, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 6.771087646484375e-05, "grad_norm": 4.811951637268066, "learning_rate": 1e-05, "loss": 0.5269, "mean_token_accuracy": 0.8484587073326111, "num_tokens": 120815951.0, "step": 3181 }, { "epoch": 0.4047831064750032, "ewc_loss": 0.1005859375, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.435153007507324, "learning_rate": 1e-05, "loss": 0.5412, "mean_token_accuracy": 0.8540355563163757, "num_tokens": 120853111.0, "step": 3182 }, { "epoch": 0.4049103167535937, "ewc_loss": 0.08203125, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 7.534027099609375e-05, "grad_norm": 5.129530906677246, "learning_rate": 1e-05, "loss": 0.5377, "mean_token_accuracy": 0.8484315872192383, "num_tokens": 120894041.0, "step": 3183 }, { "epoch": 0.4050375270321842, "ewc_loss": 0.10107421875, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.440873622894287, "learning_rate": 1e-05, "loss": 0.5203, "mean_token_accuracy": 0.8589596152305603, "num_tokens": 120930730.0, "step": 3184 }, { "epoch": 0.40516473731077474, "ewc_loss": 0.0888671875, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.467176914215088, "learning_rate": 1e-05, "loss": 0.5425, "mean_token_accuracy": 0.8498141169548035, "num_tokens": 120965044.0, "step": 3185 }, { "epoch": 0.4052919475893652, "ewc_loss": 0.09375, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 6.13179874420166, "learning_rate": 1e-05, "loss": 0.5474, "mean_token_accuracy": 0.8506110906600952, "num_tokens": 120999777.0, "step": 3186 }, { "epoch": 0.40541915786795574, "ewc_loss": 0.0888671875, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.557040691375732, "learning_rate": 1e-05, "loss": 0.5843, "mean_token_accuracy": 0.8364596366882324, "num_tokens": 121031910.0, "step": 3187 }, { "epoch": 0.40554636814654627, "ewc_loss": 0.08447265625, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.740602493286133, "learning_rate": 1e-05, "loss": 0.5116, "mean_token_accuracy": 0.858121931552887, "num_tokens": 121069600.0, "step": 3188 }, { "epoch": 0.40567357842513674, "ewc_loss": 0.0849609375, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.516957759857178, "learning_rate": 1e-05, "loss": 0.5697, "mean_token_accuracy": 0.8395339250564575, "num_tokens": 121103790.0, "step": 3189 }, { "epoch": 0.40580078870372727, "ewc_loss": 0.07861328125, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 7.200241088867188e-05, "grad_norm": 5.411686420440674, "learning_rate": 1e-05, "loss": 0.5292, "mean_token_accuracy": 0.8466824293136597, "num_tokens": 121135199.0, "step": 3190 }, { "epoch": 0.4059279989823178, "ewc_loss": 0.0830078125, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 7.62939453125e-05, "grad_norm": 5.487590312957764, "learning_rate": 1e-05, "loss": 0.5349, "mean_token_accuracy": 0.8521913886070251, "num_tokens": 121176040.0, "step": 3191 }, { "epoch": 0.40605520926090827, "ewc_loss": 0.07666015625, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 6.961822509765625e-05, "grad_norm": 5.1807026863098145, "learning_rate": 1e-05, "loss": 0.5606, "mean_token_accuracy": 0.8414653539657593, "num_tokens": 121214675.0, "step": 3192 }, { "epoch": 0.4061824195394988, "ewc_loss": 0.08154296875, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 7.486343383789062e-05, "grad_norm": 5.4891037940979, "learning_rate": 1e-05, "loss": 0.5734, "mean_token_accuracy": 0.8387596607208252, "num_tokens": 121252834.0, "step": 3193 }, { "epoch": 0.4063096298180893, "ewc_loss": 0.0751953125, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 6.818771362304688e-05, "grad_norm": 5.035593032836914, "learning_rate": 1e-05, "loss": 0.5428, "mean_token_accuracy": 0.8493607044219971, "num_tokens": 121292837.0, "step": 3194 }, { "epoch": 0.4064368400966798, "ewc_loss": 0.08203125, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 7.534027099609375e-05, "grad_norm": 5.5310869216918945, "learning_rate": 1e-05, "loss": 0.5074, "mean_token_accuracy": 0.8573680520057678, "num_tokens": 121332353.0, "step": 3195 }, { "epoch": 0.4065640503752703, "ewc_loss": 0.07421875, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 6.723403930664062e-05, "grad_norm": 4.914924621582031, "learning_rate": 1e-05, "loss": 0.4928, "mean_token_accuracy": 0.8614821434020996, "num_tokens": 121371095.0, "step": 3196 }, { "epoch": 0.40669126065386085, "ewc_loss": 0.0830078125, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 7.62939453125e-05, "grad_norm": 5.66043758392334, "learning_rate": 1e-05, "loss": 0.5822, "mean_token_accuracy": 0.8363195657730103, "num_tokens": 121409653.0, "step": 3197 }, { "epoch": 0.4068184709324513, "ewc_loss": 0.0732421875, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 6.628036499023438e-05, "grad_norm": 4.803387641906738, "learning_rate": 1e-05, "loss": 0.4785, "mean_token_accuracy": 0.8643562197685242, "num_tokens": 121449309.0, "step": 3198 }, { "epoch": 0.40694568121104185, "ewc_loss": 0.0888671875, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.912127494812012, "learning_rate": 1e-05, "loss": 0.5104, "mean_token_accuracy": 0.8598865270614624, "num_tokens": 121495004.0, "step": 3199 }, { "epoch": 0.4070728914896324, "ewc_loss": 0.07275390625, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 6.580352783203125e-05, "grad_norm": 4.736193656921387, "learning_rate": 1e-05, "loss": 0.4883, "mean_token_accuracy": 0.8614251613616943, "num_tokens": 121537679.0, "step": 3200 }, { "epoch": 0.40720010176822286, "ewc_loss": 0.09716796875, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.261231899261475, "learning_rate": 1e-05, "loss": 0.5421, "mean_token_accuracy": 0.850714921951294, "num_tokens": 121579156.0, "step": 3201 }, { "epoch": 0.4073273120468134, "ewc_loss": 0.07666015625, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 6.961822509765625e-05, "grad_norm": 4.907800197601318, "learning_rate": 1e-05, "loss": 0.5418, "mean_token_accuracy": 0.8502424955368042, "num_tokens": 121611056.0, "step": 3202 }, { "epoch": 0.4074545223254039, "ewc_loss": 0.1015625, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.4800028800964355, "learning_rate": 1e-05, "loss": 0.5096, "mean_token_accuracy": 0.8632501363754272, "num_tokens": 121642821.0, "step": 3203 }, { "epoch": 0.4075817326039944, "ewc_loss": 0.08447265625, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.243403434753418, "learning_rate": 1e-05, "loss": 0.5386, "mean_token_accuracy": 0.8501022458076477, "num_tokens": 121674021.0, "step": 3204 }, { "epoch": 0.4077089428825849, "ewc_loss": 0.10009765625, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.373182773590088, "learning_rate": 1e-05, "loss": 0.5289, "mean_token_accuracy": 0.8560762405395508, "num_tokens": 121711434.0, "step": 3205 }, { "epoch": 0.40783615316117544, "ewc_loss": 0.0888671875, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.4965009689331055, "learning_rate": 1e-05, "loss": 0.5734, "mean_token_accuracy": 0.839331865310669, "num_tokens": 121755357.0, "step": 3206 }, { "epoch": 0.4079633634397659, "ewc_loss": 0.08984375, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 6.0339789390563965, "learning_rate": 1e-05, "loss": 0.5095, "mean_token_accuracy": 0.8613876104354858, "num_tokens": 121791703.0, "step": 3207 }, { "epoch": 0.40809057371835644, "ewc_loss": 0.0888671875, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.533651351928711, "learning_rate": 1e-05, "loss": 0.497, "mean_token_accuracy": 0.8650187253952026, "num_tokens": 121824940.0, "step": 3208 }, { "epoch": 0.40821778399694697, "ewc_loss": 0.08203125, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 7.534027099609375e-05, "grad_norm": 5.670556545257568, "learning_rate": 1e-05, "loss": 0.5097, "mean_token_accuracy": 0.8569806814193726, "num_tokens": 121860434.0, "step": 3209 }, { "epoch": 0.40834499427553744, "ewc_loss": 0.08544921875, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.495444297790527, "learning_rate": 1e-05, "loss": 0.5617, "mean_token_accuracy": 0.8410665392875671, "num_tokens": 121897536.0, "step": 3210 }, { "epoch": 0.40847220455412797, "ewc_loss": 0.078125, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 7.104873657226562e-05, "grad_norm": 5.3671417236328125, "learning_rate": 1e-05, "loss": 0.5759, "mean_token_accuracy": 0.8364346623420715, "num_tokens": 121936721.0, "step": 3211 }, { "epoch": 0.4085994148327185, "ewc_loss": 0.08154296875, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 7.486343383789062e-05, "grad_norm": 5.467337131500244, "learning_rate": 1e-05, "loss": 0.5431, "mean_token_accuracy": 0.8462374806404114, "num_tokens": 121977407.0, "step": 3212 }, { "epoch": 0.40872662511130897, "ewc_loss": 0.07666015625, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 6.961822509765625e-05, "grad_norm": 5.174647331237793, "learning_rate": 1e-05, "loss": 0.5773, "mean_token_accuracy": 0.8365497589111328, "num_tokens": 122015068.0, "step": 3213 }, { "epoch": 0.4088538353898995, "ewc_loss": 0.08056640625, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 7.390975952148438e-05, "grad_norm": 5.478729248046875, "learning_rate": 1e-05, "loss": 0.5297, "mean_token_accuracy": 0.8509834408760071, "num_tokens": 122058899.0, "step": 3214 }, { "epoch": 0.40898104566849003, "ewc_loss": 0.0751953125, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 6.818771362304688e-05, "grad_norm": 5.031559467315674, "learning_rate": 1e-05, "loss": 0.4957, "mean_token_accuracy": 0.8569542765617371, "num_tokens": 122095407.0, "step": 3215 }, { "epoch": 0.4091082559470805, "ewc_loss": 0.0810546875, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 7.43865966796875e-05, "grad_norm": 5.513422966003418, "learning_rate": 1e-05, "loss": 0.503, "mean_token_accuracy": 0.8577934503555298, "num_tokens": 122134836.0, "step": 3216 }, { "epoch": 0.40923546622567103, "ewc_loss": 0.07421875, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 6.723403930664062e-05, "grad_norm": 4.936304092407227, "learning_rate": 1e-05, "loss": 0.5174, "mean_token_accuracy": 0.8526761531829834, "num_tokens": 122165959.0, "step": 3217 }, { "epoch": 0.40936267650426156, "ewc_loss": 0.0830078125, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 7.62939453125e-05, "grad_norm": 5.628961086273193, "learning_rate": 1e-05, "loss": 0.5318, "mean_token_accuracy": 0.8486210107803345, "num_tokens": 122206223.0, "step": 3218 }, { "epoch": 0.40948988678285203, "ewc_loss": 0.0732421875, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 6.628036499023438e-05, "grad_norm": 4.8016133308410645, "learning_rate": 1e-05, "loss": 0.4922, "mean_token_accuracy": 0.8602854013442993, "num_tokens": 122245416.0, "step": 3219 }, { "epoch": 0.40961709706144256, "ewc_loss": 0.087890625, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.875077247619629, "learning_rate": 1e-05, "loss": 0.5693, "mean_token_accuracy": 0.8378956317901611, "num_tokens": 122285264.0, "step": 3220 }, { "epoch": 0.4097443073400331, "ewc_loss": 0.07275390625, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 6.580352783203125e-05, "grad_norm": 4.712792873382568, "learning_rate": 1e-05, "loss": 0.5158, "mean_token_accuracy": 0.8506491780281067, "num_tokens": 122326956.0, "step": 3221 }, { "epoch": 0.40987151761862356, "ewc_loss": 0.09765625, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.234954833984375, "learning_rate": 1e-05, "loss": 0.5166, "mean_token_accuracy": 0.8612871170043945, "num_tokens": 122362409.0, "step": 3222 }, { "epoch": 0.4099987278972141, "ewc_loss": 0.07568359375, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 6.866455078125e-05, "grad_norm": 4.8232855796813965, "learning_rate": 1e-05, "loss": 0.5812, "mean_token_accuracy": 0.8315727710723877, "num_tokens": 122405011.0, "step": 3223 }, { "epoch": 0.4101259381758046, "ewc_loss": 0.10302734375, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.551586627960205, "learning_rate": 1e-05, "loss": 0.5288, "mean_token_accuracy": 0.8585120439529419, "num_tokens": 122445482.0, "step": 3224 }, { "epoch": 0.4102531484543951, "ewc_loss": 0.0849609375, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.221012592315674, "learning_rate": 1e-05, "loss": 0.5133, "mean_token_accuracy": 0.8583368062973022, "num_tokens": 122485024.0, "step": 3225 }, { "epoch": 0.4103803587329856, "ewc_loss": 0.10205078125, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.490847587585449, "learning_rate": 1e-05, "loss": 0.53, "mean_token_accuracy": 0.8558847904205322, "num_tokens": 122524766.0, "step": 3226 }, { "epoch": 0.41050756901157615, "ewc_loss": 0.09033203125, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.563513278961182, "learning_rate": 1e-05, "loss": 0.5233, "mean_token_accuracy": 0.8544341325759888, "num_tokens": 122566664.0, "step": 3227 }, { "epoch": 0.4106347792901666, "ewc_loss": 0.09228515625, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 6.139603137969971, "learning_rate": 1e-05, "loss": 0.6329, "mean_token_accuracy": 0.8197413682937622, "num_tokens": 122606724.0, "step": 3228 }, { "epoch": 0.41076198956875715, "ewc_loss": 0.08984375, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.613420009613037, "learning_rate": 1e-05, "loss": 0.5522, "mean_token_accuracy": 0.8466006517410278, "num_tokens": 122641664.0, "step": 3229 }, { "epoch": 0.4108891998473477, "ewc_loss": 0.08349609375, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.7292656898498535, "learning_rate": 1e-05, "loss": 0.4709, "mean_token_accuracy": 0.8720378875732422, "num_tokens": 122673370.0, "step": 3230 }, { "epoch": 0.4110164101259382, "ewc_loss": 0.08544921875, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.548445224761963, "learning_rate": 1e-05, "loss": 0.5329, "mean_token_accuracy": 0.8515779972076416, "num_tokens": 122711545.0, "step": 3231 }, { "epoch": 0.4111436204045287, "ewc_loss": 0.078125, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 7.104873657226562e-05, "grad_norm": 5.407643795013428, "learning_rate": 1e-05, "loss": 0.5081, "mean_token_accuracy": 0.8561984300613403, "num_tokens": 122743656.0, "step": 3232 }, { "epoch": 0.4112708306831192, "ewc_loss": 0.08251953125, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 7.581710815429688e-05, "grad_norm": 5.5080366134643555, "learning_rate": 1e-05, "loss": 0.5413, "mean_token_accuracy": 0.8464858531951904, "num_tokens": 122778233.0, "step": 3233 }, { "epoch": 0.41139804096170973, "ewc_loss": 0.0771484375, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 7.009506225585938e-05, "grad_norm": 5.17134428024292, "learning_rate": 1e-05, "loss": 0.4958, "mean_token_accuracy": 0.8630188703536987, "num_tokens": 122821063.0, "step": 3234 }, { "epoch": 0.4115252512403002, "ewc_loss": 0.0810546875, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 7.43865966796875e-05, "grad_norm": 5.522397518157959, "learning_rate": 1e-05, "loss": 0.5441, "mean_token_accuracy": 0.8498907685279846, "num_tokens": 122855871.0, "step": 3235 }, { "epoch": 0.41165246151889073, "ewc_loss": 0.0751953125, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 6.818771362304688e-05, "grad_norm": 5.017678737640381, "learning_rate": 1e-05, "loss": 0.5132, "mean_token_accuracy": 0.854694664478302, "num_tokens": 122893990.0, "step": 3236 }, { "epoch": 0.41177967179748126, "ewc_loss": 0.0830078125, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 7.62939453125e-05, "grad_norm": 5.600112438201904, "learning_rate": 1e-05, "loss": 0.5206, "mean_token_accuracy": 0.8568038940429688, "num_tokens": 122932017.0, "step": 3237 }, { "epoch": 0.41190688207607173, "ewc_loss": 0.07421875, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 6.723403930664062e-05, "grad_norm": 4.882380485534668, "learning_rate": 1e-05, "loss": 0.4979, "mean_token_accuracy": 0.8588173985481262, "num_tokens": 122975581.0, "step": 3238 }, { "epoch": 0.41203409235466226, "ewc_loss": 0.0859375, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.780433177947998, "learning_rate": 1e-05, "loss": 0.5337, "mean_token_accuracy": 0.8495759963989258, "num_tokens": 123006051.0, "step": 3239 }, { "epoch": 0.4121613026332528, "ewc_loss": 0.07275390625, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 6.580352783203125e-05, "grad_norm": 4.773964881896973, "learning_rate": 1e-05, "loss": 0.5228, "mean_token_accuracy": 0.8498084545135498, "num_tokens": 123049756.0, "step": 3240 }, { "epoch": 0.41228851291184326, "ewc_loss": 0.09228515625, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 6.066098213195801, "learning_rate": 1e-05, "loss": 0.5609, "mean_token_accuracy": 0.8465855121612549, "num_tokens": 123091059.0, "step": 3241 }, { "epoch": 0.4124157231904338, "ewc_loss": 0.07421875, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 6.723403930664062e-05, "grad_norm": 4.77708625793457, "learning_rate": 1e-05, "loss": 0.4943, "mean_token_accuracy": 0.8610427975654602, "num_tokens": 123130369.0, "step": 3242 }, { "epoch": 0.4125429334690243, "ewc_loss": 0.10009765625, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.3994340896606445, "learning_rate": 1e-05, "loss": 0.5417, "mean_token_accuracy": 0.8541566133499146, "num_tokens": 123173507.0, "step": 3243 }, { "epoch": 0.4126701437476148, "ewc_loss": 0.0810546875, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 7.43865966796875e-05, "grad_norm": 5.0366716384887695, "learning_rate": 1e-05, "loss": 0.4855, "mean_token_accuracy": 0.8634625673294067, "num_tokens": 123211428.0, "step": 3244 }, { "epoch": 0.4127973540262053, "ewc_loss": 0.10302734375, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.5235595703125, "learning_rate": 1e-05, "loss": 0.5766, "mean_token_accuracy": 0.8472954034805298, "num_tokens": 123256115.0, "step": 3245 }, { "epoch": 0.41292456430479585, "ewc_loss": 0.08837890625, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.420360088348389, "learning_rate": 1e-05, "loss": 0.5224, "mean_token_accuracy": 0.8546279668807983, "num_tokens": 123297319.0, "step": 3246 }, { "epoch": 0.4130517745833863, "ewc_loss": 0.0986328125, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.267337799072266, "learning_rate": 1e-05, "loss": 0.5855, "mean_token_accuracy": 0.8384998440742493, "num_tokens": 123341241.0, "step": 3247 }, { "epoch": 0.41317898486197685, "ewc_loss": 0.08935546875, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.590984344482422, "learning_rate": 1e-05, "loss": 0.5615, "mean_token_accuracy": 0.842869222164154, "num_tokens": 123381832.0, "step": 3248 }, { "epoch": 0.4133061951405674, "ewc_loss": 0.08544921875, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.868059158325195, "learning_rate": 1e-05, "loss": 0.5221, "mean_token_accuracy": 0.8540587425231934, "num_tokens": 123418393.0, "step": 3249 }, { "epoch": 0.41343340541915785, "ewc_loss": 0.0869140625, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.560007572174072, "learning_rate": 1e-05, "loss": 0.4697, "mean_token_accuracy": 0.8632413744926453, "num_tokens": 123449916.0, "step": 3250 }, { "epoch": 0.4135606156977484, "ewc_loss": 0.080078125, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 7.343292236328125e-05, "grad_norm": 5.510439395904541, "learning_rate": 1e-05, "loss": 0.4616, "mean_token_accuracy": 0.8750842809677124, "num_tokens": 123484025.0, "step": 3251 }, { "epoch": 0.4136878259763389, "ewc_loss": 0.08349609375, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.502627372741699, "learning_rate": 1e-05, "loss": 0.511, "mean_token_accuracy": 0.856704592704773, "num_tokens": 123522345.0, "step": 3252 }, { "epoch": 0.4138150362549294, "ewc_loss": 0.0771484375, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 7.009506225585938e-05, "grad_norm": 5.256481170654297, "learning_rate": 1e-05, "loss": 0.5372, "mean_token_accuracy": 0.8525181412696838, "num_tokens": 123562123.0, "step": 3253 }, { "epoch": 0.4139422465335199, "ewc_loss": 0.08056640625, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 7.390975952148438e-05, "grad_norm": 5.494428634643555, "learning_rate": 1e-05, "loss": 0.5442, "mean_token_accuracy": 0.846290111541748, "num_tokens": 123606014.0, "step": 3254 }, { "epoch": 0.41406945681211044, "ewc_loss": 0.076171875, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 6.914138793945312e-05, "grad_norm": 5.085855007171631, "learning_rate": 1e-05, "loss": 0.4845, "mean_token_accuracy": 0.8638290166854858, "num_tokens": 123643483.0, "step": 3255 }, { "epoch": 0.4141966670907009, "ewc_loss": 0.08154296875, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 7.486343383789062e-05, "grad_norm": 5.519350528717041, "learning_rate": 1e-05, "loss": 0.5252, "mean_token_accuracy": 0.8567800521850586, "num_tokens": 123689303.0, "step": 3256 }, { "epoch": 0.41432387736929144, "ewc_loss": 0.0751953125, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 6.818771362304688e-05, "grad_norm": 4.972951412200928, "learning_rate": 1e-05, "loss": 0.488, "mean_token_accuracy": 0.8599900007247925, "num_tokens": 123722388.0, "step": 3257 }, { "epoch": 0.41445108764788197, "ewc_loss": 0.08251953125, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 7.581710815429688e-05, "grad_norm": 5.606246471405029, "learning_rate": 1e-05, "loss": 0.5249, "mean_token_accuracy": 0.8520696759223938, "num_tokens": 123761761.0, "step": 3258 }, { "epoch": 0.41457829792647244, "ewc_loss": 0.07373046875, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 6.67572021484375e-05, "grad_norm": 4.852719783782959, "learning_rate": 1e-05, "loss": 0.441, "mean_token_accuracy": 0.8771408200263977, "num_tokens": 123799138.0, "step": 3259 }, { "epoch": 0.41470550820506297, "ewc_loss": 0.08642578125, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.806572437286377, "learning_rate": 1e-05, "loss": 0.4923, "mean_token_accuracy": 0.859686017036438, "num_tokens": 123840803.0, "step": 3260 }, { "epoch": 0.4148327184836535, "ewc_loss": 0.0732421875, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 6.628036499023438e-05, "grad_norm": 4.760990142822266, "learning_rate": 1e-05, "loss": 0.5394, "mean_token_accuracy": 0.8500304818153381, "num_tokens": 123875705.0, "step": 3261 }, { "epoch": 0.41495992876224397, "ewc_loss": 0.091796875, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 6.107263565063477, "learning_rate": 1e-05, "loss": 0.5798, "mean_token_accuracy": 0.8394559025764465, "num_tokens": 123915830.0, "step": 3262 }, { "epoch": 0.4150871390408345, "ewc_loss": 0.0751953125, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 6.818771362304688e-05, "grad_norm": 4.788412094116211, "learning_rate": 1e-05, "loss": 0.5432, "mean_token_accuracy": 0.8437058925628662, "num_tokens": 123953855.0, "step": 3263 }, { "epoch": 0.415214349319425, "ewc_loss": 0.10009765625, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.43489933013916, "learning_rate": 1e-05, "loss": 0.5353, "mean_token_accuracy": 0.8540250062942505, "num_tokens": 123985697.0, "step": 3264 }, { "epoch": 0.4153415595980155, "ewc_loss": 0.0810546875, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 7.43865966796875e-05, "grad_norm": 5.101576328277588, "learning_rate": 1e-05, "loss": 0.5296, "mean_token_accuracy": 0.8510217070579529, "num_tokens": 124021682.0, "step": 3265 }, { "epoch": 0.415468769876606, "ewc_loss": 0.1015625, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.483742713928223, "learning_rate": 1e-05, "loss": 0.5884, "mean_token_accuracy": 0.8398324251174927, "num_tokens": 124057472.0, "step": 3266 }, { "epoch": 0.41559598015519655, "ewc_loss": 0.08837890625, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.434619903564453, "learning_rate": 1e-05, "loss": 0.5265, "mean_token_accuracy": 0.8556603193283081, "num_tokens": 124100160.0, "step": 3267 }, { "epoch": 0.415723190433787, "ewc_loss": 0.095703125, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.203190803527832, "learning_rate": 1e-05, "loss": 0.5263, "mean_token_accuracy": 0.8527135848999023, "num_tokens": 124138503.0, "step": 3268 }, { "epoch": 0.41585040071237755, "ewc_loss": 0.0888671875, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.729173183441162, "learning_rate": 1e-05, "loss": 0.5785, "mean_token_accuracy": 0.8429195880889893, "num_tokens": 124176082.0, "step": 3269 }, { "epoch": 0.4159776109909681, "ewc_loss": 0.0830078125, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 7.62939453125e-05, "grad_norm": 5.704837799072266, "learning_rate": 1e-05, "loss": 0.5504, "mean_token_accuracy": 0.8481289744377136, "num_tokens": 124213853.0, "step": 3270 }, { "epoch": 0.41610482126955856, "ewc_loss": 0.0869140625, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.559724807739258, "learning_rate": 1e-05, "loss": 0.4759, "mean_token_accuracy": 0.8715651631355286, "num_tokens": 124250562.0, "step": 3271 }, { "epoch": 0.4162320315481491, "ewc_loss": 0.07861328125, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 7.200241088867188e-05, "grad_norm": 5.364162445068359, "learning_rate": 1e-05, "loss": 0.5687, "mean_token_accuracy": 0.8363757729530334, "num_tokens": 124282321.0, "step": 3272 }, { "epoch": 0.4163592418267396, "ewc_loss": 0.08447265625, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.55159854888916, "learning_rate": 1e-05, "loss": 0.5178, "mean_token_accuracy": 0.8579174280166626, "num_tokens": 124323172.0, "step": 3273 }, { "epoch": 0.4164864521053301, "ewc_loss": 0.0771484375, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 7.009506225585938e-05, "grad_norm": 5.129563331604004, "learning_rate": 1e-05, "loss": 0.5277, "mean_token_accuracy": 0.8520049452781677, "num_tokens": 124361585.0, "step": 3274 }, { "epoch": 0.4166136623839206, "ewc_loss": 0.083984375, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.583216667175293, "learning_rate": 1e-05, "loss": 0.6079, "mean_token_accuracy": 0.8314942717552185, "num_tokens": 124398926.0, "step": 3275 }, { "epoch": 0.41674087266251114, "ewc_loss": 0.07666015625, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 6.961822509765625e-05, "grad_norm": 4.988545894622803, "learning_rate": 1e-05, "loss": 0.4676, "mean_token_accuracy": 0.8673043251037598, "num_tokens": 124437914.0, "step": 3276 }, { "epoch": 0.4168680829411016, "ewc_loss": 0.08544921875, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.669197082519531, "learning_rate": 1e-05, "loss": 0.5313, "mean_token_accuracy": 0.8556824922561646, "num_tokens": 124478850.0, "step": 3277 }, { "epoch": 0.41699529321969214, "ewc_loss": 0.07568359375, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 6.866455078125e-05, "grad_norm": 4.869672775268555, "learning_rate": 1e-05, "loss": 0.4742, "mean_token_accuracy": 0.8672388792037964, "num_tokens": 124519361.0, "step": 3278 }, { "epoch": 0.41712250349828267, "ewc_loss": 0.08935546875, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.859847068786621, "learning_rate": 1e-05, "loss": 0.5466, "mean_token_accuracy": 0.8509488105773926, "num_tokens": 124555616.0, "step": 3279 }, { "epoch": 0.4172497137768732, "ewc_loss": 0.0751953125, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 6.818771362304688e-05, "grad_norm": 4.816344261169434, "learning_rate": 1e-05, "loss": 0.568, "mean_token_accuracy": 0.8367871046066284, "num_tokens": 124593045.0, "step": 3280 }, { "epoch": 0.41737692405546367, "ewc_loss": 0.09521484375, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 6.107518196105957, "learning_rate": 1e-05, "loss": 0.6023, "mean_token_accuracy": 0.831009030342102, "num_tokens": 124631842.0, "step": 3281 }, { "epoch": 0.4175041343340542, "ewc_loss": 0.076171875, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 6.914138793945312e-05, "grad_norm": 4.850962162017822, "learning_rate": 1e-05, "loss": 0.536, "mean_token_accuracy": 0.8493659496307373, "num_tokens": 124674466.0, "step": 3282 }, { "epoch": 0.41763134461264473, "ewc_loss": 0.10009765625, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.357173442840576, "learning_rate": 1e-05, "loss": 0.5201, "mean_token_accuracy": 0.8564141392707825, "num_tokens": 124707893.0, "step": 3283 }, { "epoch": 0.4177585548912352, "ewc_loss": 0.0830078125, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 7.62939453125e-05, "grad_norm": 5.0999250411987305, "learning_rate": 1e-05, "loss": 0.5255, "mean_token_accuracy": 0.8506945967674255, "num_tokens": 124746230.0, "step": 3284 }, { "epoch": 0.41788576516982573, "ewc_loss": 0.1015625, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.382858753204346, "learning_rate": 1e-05, "loss": 0.5397, "mean_token_accuracy": 0.8522204756736755, "num_tokens": 124786988.0, "step": 3285 }, { "epoch": 0.41801297544841626, "ewc_loss": 0.08740234375, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.381646156311035, "learning_rate": 1e-05, "loss": 0.5566, "mean_token_accuracy": 0.8448410630226135, "num_tokens": 124831061.0, "step": 3286 }, { "epoch": 0.41814018572700673, "ewc_loss": 0.09521484375, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 6.134103775024414, "learning_rate": 1e-05, "loss": 0.5448, "mean_token_accuracy": 0.8493506908416748, "num_tokens": 124870486.0, "step": 3287 }, { "epoch": 0.41826739600559726, "ewc_loss": 0.0888671875, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.48251485824585, "learning_rate": 1e-05, "loss": 0.4867, "mean_token_accuracy": 0.8658015727996826, "num_tokens": 124909675.0, "step": 3288 }, { "epoch": 0.4183946062841878, "ewc_loss": 0.0859375, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.787354469299316, "learning_rate": 1e-05, "loss": 0.5654, "mean_token_accuracy": 0.8419182896614075, "num_tokens": 124953854.0, "step": 3289 }, { "epoch": 0.41852181656277826, "ewc_loss": 0.08544921875, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.4860992431640625, "learning_rate": 1e-05, "loss": 0.543, "mean_token_accuracy": 0.8510335683822632, "num_tokens": 124991117.0, "step": 3290 }, { "epoch": 0.4186490268413688, "ewc_loss": 0.08056640625, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 7.390975952148438e-05, "grad_norm": 5.4601149559021, "learning_rate": 1e-05, "loss": 0.5315, "mean_token_accuracy": 0.851851224899292, "num_tokens": 125027160.0, "step": 3291 }, { "epoch": 0.4187762371199593, "ewc_loss": 0.08447265625, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.46846342086792, "learning_rate": 1e-05, "loss": 0.4965, "mean_token_accuracy": 0.8591504693031311, "num_tokens": 125061324.0, "step": 3292 }, { "epoch": 0.4189034473985498, "ewc_loss": 0.07763671875, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 7.05718994140625e-05, "grad_norm": 5.221550941467285, "learning_rate": 1e-05, "loss": 0.5296, "mean_token_accuracy": 0.8504735231399536, "num_tokens": 125099616.0, "step": 3293 }, { "epoch": 0.4190306576771403, "ewc_loss": 0.0810546875, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 7.43865966796875e-05, "grad_norm": 5.472943305969238, "learning_rate": 1e-05, "loss": 0.5369, "mean_token_accuracy": 0.849200963973999, "num_tokens": 125132855.0, "step": 3294 }, { "epoch": 0.41915786795573085, "ewc_loss": 0.07666015625, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 6.961822509765625e-05, "grad_norm": 5.07004976272583, "learning_rate": 1e-05, "loss": 0.5619, "mean_token_accuracy": 0.8453415632247925, "num_tokens": 125168040.0, "step": 3295 }, { "epoch": 0.4192850782343213, "ewc_loss": 0.08203125, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 7.534027099609375e-05, "grad_norm": 5.51505708694458, "learning_rate": 1e-05, "loss": 0.4916, "mean_token_accuracy": 0.8575868606567383, "num_tokens": 125198788.0, "step": 3296 }, { "epoch": 0.41941228851291185, "ewc_loss": 0.07568359375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 6.866455078125e-05, "grad_norm": 4.939269065856934, "learning_rate": 1e-05, "loss": 0.4998, "mean_token_accuracy": 0.8594317436218262, "num_tokens": 125238387.0, "step": 3297 }, { "epoch": 0.4195394987915024, "ewc_loss": 0.08349609375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.634469509124756, "learning_rate": 1e-05, "loss": 0.4895, "mean_token_accuracy": 0.8658397197723389, "num_tokens": 125273666.0, "step": 3298 }, { "epoch": 0.41966670907009285, "ewc_loss": 0.07421875, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 6.723403930664062e-05, "grad_norm": 4.827978610992432, "learning_rate": 1e-05, "loss": 0.5486, "mean_token_accuracy": 0.8445620536804199, "num_tokens": 125315961.0, "step": 3299 }, { "epoch": 0.4197939193486834, "ewc_loss": 0.08837890625, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.857994079589844, "learning_rate": 1e-05, "loss": 0.5689, "mean_token_accuracy": 0.8408331274986267, "num_tokens": 125358830.0, "step": 3300 }, { "epoch": 0.4199211296272739, "ewc_loss": 0.07373046875, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 6.67572021484375e-05, "grad_norm": 4.7470831871032715, "learning_rate": 1e-05, "loss": 0.5311, "mean_token_accuracy": 0.8458791971206665, "num_tokens": 125397497.0, "step": 3301 }, { "epoch": 0.4200483399058644, "ewc_loss": 0.095703125, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.1962056159973145, "learning_rate": 1e-05, "loss": 0.5996, "mean_token_accuracy": 0.8326195478439331, "num_tokens": 125437337.0, "step": 3302 }, { "epoch": 0.4201755501844549, "ewc_loss": 0.07568359375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 6.866455078125e-05, "grad_norm": 4.807134628295898, "learning_rate": 1e-05, "loss": 0.4833, "mean_token_accuracy": 0.861943244934082, "num_tokens": 125470535.0, "step": 3303 }, { "epoch": 0.42030276046304543, "ewc_loss": 0.10302734375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.521878242492676, "learning_rate": 1e-05, "loss": 0.5878, "mean_token_accuracy": 0.8396822214126587, "num_tokens": 125510692.0, "step": 3304 }, { "epoch": 0.4204299707416359, "ewc_loss": 0.083984375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.1600823402404785, "learning_rate": 1e-05, "loss": 0.5701, "mean_token_accuracy": 0.83750319480896, "num_tokens": 125557328.0, "step": 3305 }, { "epoch": 0.42055718102022643, "ewc_loss": 0.103515625, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.53806734085083, "learning_rate": 1e-05, "loss": 0.5737, "mean_token_accuracy": 0.8484058380126953, "num_tokens": 125594006.0, "step": 3306 }, { "epoch": 0.42068439129881696, "ewc_loss": 0.091796875, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.5487518310546875, "learning_rate": 1e-05, "loss": 0.537, "mean_token_accuracy": 0.8484535813331604, "num_tokens": 125627692.0, "step": 3307 }, { "epoch": 0.42081160157740743, "ewc_loss": 0.095703125, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.197753429412842, "learning_rate": 1e-05, "loss": 0.5221, "mean_token_accuracy": 0.8553844690322876, "num_tokens": 125660218.0, "step": 3308 }, { "epoch": 0.42093881185599796, "ewc_loss": 0.09033203125, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.615473747253418, "learning_rate": 1e-05, "loss": 0.502, "mean_token_accuracy": 0.8615227341651917, "num_tokens": 125695793.0, "step": 3309 }, { "epoch": 0.4210660221345885, "ewc_loss": 0.08447265625, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.797908306121826, "learning_rate": 1e-05, "loss": 0.5271, "mean_token_accuracy": 0.8551750779151917, "num_tokens": 125732438.0, "step": 3310 }, { "epoch": 0.42119323241317896, "ewc_loss": 0.0869140625, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.561148166656494, "learning_rate": 1e-05, "loss": 0.4948, "mean_token_accuracy": 0.8599022626876831, "num_tokens": 125764919.0, "step": 3311 }, { "epoch": 0.4213204426917695, "ewc_loss": 0.080078125, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 7.343292236328125e-05, "grad_norm": 5.470623970031738, "learning_rate": 1e-05, "loss": 0.5325, "mean_token_accuracy": 0.8507359623908997, "num_tokens": 125797172.0, "step": 3312 }, { "epoch": 0.42144765297036, "ewc_loss": 0.083984375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.500102996826172, "learning_rate": 1e-05, "loss": 0.5019, "mean_token_accuracy": 0.8553277254104614, "num_tokens": 125838144.0, "step": 3313 }, { "epoch": 0.4215748632489505, "ewc_loss": 0.07763671875, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 7.05718994140625e-05, "grad_norm": 5.239776134490967, "learning_rate": 1e-05, "loss": 0.4847, "mean_token_accuracy": 0.8638848066329956, "num_tokens": 125873897.0, "step": 3314 }, { "epoch": 0.421702073527541, "ewc_loss": 0.08154296875, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 7.486343383789062e-05, "grad_norm": 5.483553886413574, "learning_rate": 1e-05, "loss": 0.5628, "mean_token_accuracy": 0.8386154770851135, "num_tokens": 125909773.0, "step": 3315 }, { "epoch": 0.42182928380613155, "ewc_loss": 0.07666015625, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 6.961822509765625e-05, "grad_norm": 5.101630687713623, "learning_rate": 1e-05, "loss": 0.5087, "mean_token_accuracy": 0.8562458753585815, "num_tokens": 125947661.0, "step": 3316 }, { "epoch": 0.421956494084722, "ewc_loss": 0.08056640625, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 7.390975952148438e-05, "grad_norm": 5.499683856964111, "learning_rate": 1e-05, "loss": 0.5132, "mean_token_accuracy": 0.8570315837860107, "num_tokens": 125979483.0, "step": 3317 }, { "epoch": 0.42208370436331255, "ewc_loss": 0.07568359375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 6.866455078125e-05, "grad_norm": 4.977206707000732, "learning_rate": 1e-05, "loss": 0.4748, "mean_token_accuracy": 0.863766074180603, "num_tokens": 126016907.0, "step": 3318 }, { "epoch": 0.4222109146419031, "ewc_loss": 0.0830078125, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 7.62939453125e-05, "grad_norm": 5.5896124839782715, "learning_rate": 1e-05, "loss": 0.5369, "mean_token_accuracy": 0.8487777709960938, "num_tokens": 126059475.0, "step": 3319 }, { "epoch": 0.42233812492049355, "ewc_loss": 0.07470703125, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 6.771087646484375e-05, "grad_norm": 4.879621982574463, "learning_rate": 1e-05, "loss": 0.4618, "mean_token_accuracy": 0.8704432249069214, "num_tokens": 126097373.0, "step": 3320 }, { "epoch": 0.4224653351990841, "ewc_loss": 0.08544921875, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.773966312408447, "learning_rate": 1e-05, "loss": 0.4945, "mean_token_accuracy": 0.8656457662582397, "num_tokens": 126138884.0, "step": 3321 }, { "epoch": 0.4225925454776746, "ewc_loss": 0.07275390625, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 6.580352783203125e-05, "grad_norm": 4.761593341827393, "learning_rate": 1e-05, "loss": 0.5046, "mean_token_accuracy": 0.8591574430465698, "num_tokens": 126178453.0, "step": 3322 }, { "epoch": 0.4227197557562651, "ewc_loss": 0.091796875, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 6.09610652923584, "learning_rate": 1e-05, "loss": 0.5471, "mean_token_accuracy": 0.8482468128204346, "num_tokens": 126208979.0, "step": 3323 }, { "epoch": 0.4228469660348556, "ewc_loss": 0.07421875, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 6.723403930664062e-05, "grad_norm": 4.745162010192871, "learning_rate": 1e-05, "loss": 0.4994, "mean_token_accuracy": 0.8577251434326172, "num_tokens": 126252195.0, "step": 3324 }, { "epoch": 0.42297417631344614, "ewc_loss": 0.1015625, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.470745086669922, "learning_rate": 1e-05, "loss": 0.5192, "mean_token_accuracy": 0.8586801886558533, "num_tokens": 126287526.0, "step": 3325 }, { "epoch": 0.4231013865920366, "ewc_loss": 0.08056640625, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 7.390975952148438e-05, "grad_norm": 5.049405097961426, "learning_rate": 1e-05, "loss": 0.4869, "mean_token_accuracy": 0.8638688325881958, "num_tokens": 126325868.0, "step": 3326 }, { "epoch": 0.42322859687062714, "ewc_loss": 0.10546875, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.600625514984131, "learning_rate": 1e-05, "loss": 0.5334, "mean_token_accuracy": 0.8577653169631958, "num_tokens": 126364613.0, "step": 3327 }, { "epoch": 0.42335580714921767, "ewc_loss": 0.08984375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.4816575050354, "learning_rate": 1e-05, "loss": 0.4937, "mean_token_accuracy": 0.8655284643173218, "num_tokens": 126404373.0, "step": 3328 }, { "epoch": 0.42348301742780814, "ewc_loss": 0.0986328125, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.333777904510498, "learning_rate": 1e-05, "loss": 0.6047, "mean_token_accuracy": 0.8327155709266663, "num_tokens": 126444557.0, "step": 3329 }, { "epoch": 0.42361022770639867, "ewc_loss": 0.091796875, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.648950099945068, "learning_rate": 1e-05, "loss": 0.5404, "mean_token_accuracy": 0.8489888310432434, "num_tokens": 126479657.0, "step": 3330 }, { "epoch": 0.4237374379849892, "ewc_loss": 0.08642578125, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.915375232696533, "learning_rate": 1e-05, "loss": 0.5584, "mean_token_accuracy": 0.846247673034668, "num_tokens": 126518004.0, "step": 3331 }, { "epoch": 0.4238646482635797, "ewc_loss": 0.08740234375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.604725360870361, "learning_rate": 1e-05, "loss": 0.5622, "mean_token_accuracy": 0.8465110063552856, "num_tokens": 126557854.0, "step": 3332 }, { "epoch": 0.4239918585421702, "ewc_loss": 0.080078125, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 7.343292236328125e-05, "grad_norm": 5.545313358306885, "learning_rate": 1e-05, "loss": 0.5602, "mean_token_accuracy": 0.8489752411842346, "num_tokens": 126593657.0, "step": 3333 }, { "epoch": 0.4241190688207607, "ewc_loss": 0.0859375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.541123390197754, "learning_rate": 1e-05, "loss": 0.462, "mean_token_accuracy": 0.8769739866256714, "num_tokens": 126635913.0, "step": 3334 }, { "epoch": 0.42424627909935125, "ewc_loss": 0.0771484375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 7.009506225585938e-05, "grad_norm": 5.299345970153809, "learning_rate": 1e-05, "loss": 0.5605, "mean_token_accuracy": 0.8403946161270142, "num_tokens": 126668703.0, "step": 3335 }, { "epoch": 0.4243734893779417, "ewc_loss": 0.08251953125, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 7.581710815429688e-05, "grad_norm": 5.5138397216796875, "learning_rate": 1e-05, "loss": 0.4687, "mean_token_accuracy": 0.8686569333076477, "num_tokens": 126707952.0, "step": 3336 }, { "epoch": 0.42450069965653225, "ewc_loss": 0.076171875, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 6.914138793945312e-05, "grad_norm": 5.125658988952637, "learning_rate": 1e-05, "loss": 0.5179, "mean_token_accuracy": 0.8527436256408691, "num_tokens": 126748756.0, "step": 3337 }, { "epoch": 0.4246279099351228, "ewc_loss": 0.08251953125, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 7.581710815429688e-05, "grad_norm": 5.541468143463135, "learning_rate": 1e-05, "loss": 0.489, "mean_token_accuracy": 0.8641340732574463, "num_tokens": 126784780.0, "step": 3338 }, { "epoch": 0.42475512021371326, "ewc_loss": 0.0751953125, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 6.818771362304688e-05, "grad_norm": 4.982562065124512, "learning_rate": 1e-05, "loss": 0.4954, "mean_token_accuracy": 0.8579256534576416, "num_tokens": 126827371.0, "step": 3339 }, { "epoch": 0.4248823304923038, "ewc_loss": 0.0830078125, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 7.62939453125e-05, "grad_norm": 5.660681247711182, "learning_rate": 1e-05, "loss": 0.5434, "mean_token_accuracy": 0.8483843207359314, "num_tokens": 126862357.0, "step": 3340 }, { "epoch": 0.4250095407708943, "ewc_loss": 0.07373046875, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 6.67572021484375e-05, "grad_norm": 4.851443767547607, "learning_rate": 1e-05, "loss": 0.4543, "mean_token_accuracy": 0.8741605281829834, "num_tokens": 126902433.0, "step": 3341 }, { "epoch": 0.4251367510494848, "ewc_loss": 0.087890625, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.884500980377197, "learning_rate": 1e-05, "loss": 0.5543, "mean_token_accuracy": 0.8482444286346436, "num_tokens": 126946958.0, "step": 3342 }, { "epoch": 0.4252639613280753, "ewc_loss": 0.0732421875, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 6.628036499023438e-05, "grad_norm": 4.763997554779053, "learning_rate": 1e-05, "loss": 0.4667, "mean_token_accuracy": 0.8645595908164978, "num_tokens": 126980267.0, "step": 3343 }, { "epoch": 0.42539117160666584, "ewc_loss": 0.0966796875, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.217451095581055, "learning_rate": 1e-05, "loss": 0.4764, "mean_token_accuracy": 0.8726080656051636, "num_tokens": 127018215.0, "step": 3344 }, { "epoch": 0.4255183818852563, "ewc_loss": 0.0751953125, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 6.818771362304688e-05, "grad_norm": 4.842919826507568, "learning_rate": 1e-05, "loss": 0.5492, "mean_token_accuracy": 0.8449024558067322, "num_tokens": 127055844.0, "step": 3345 }, { "epoch": 0.42564559216384684, "ewc_loss": 0.10205078125, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.522642612457275, "learning_rate": 1e-05, "loss": 0.526, "mean_token_accuracy": 0.8581631779670715, "num_tokens": 127099373.0, "step": 3346 }, { "epoch": 0.42577280244243737, "ewc_loss": 0.083984375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.186853885650635, "learning_rate": 1e-05, "loss": 0.5127, "mean_token_accuracy": 0.8567842245101929, "num_tokens": 127143755.0, "step": 3347 }, { "epoch": 0.42590001272102784, "ewc_loss": 0.10302734375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.527194976806641, "learning_rate": 1e-05, "loss": 0.5194, "mean_token_accuracy": 0.8638612031936646, "num_tokens": 127180543.0, "step": 3348 }, { "epoch": 0.42602722299961837, "ewc_loss": 0.09033203125, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.54871129989624, "learning_rate": 1e-05, "loss": 0.5433, "mean_token_accuracy": 0.848698616027832, "num_tokens": 127225203.0, "step": 3349 }, { "epoch": 0.4261544332782089, "ewc_loss": 0.09521484375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 6.195660591125488, "learning_rate": 1e-05, "loss": 0.6019, "mean_token_accuracy": 0.834979772567749, "num_tokens": 127260949.0, "step": 3350 }, { "epoch": 0.4262816435567994, "ewc_loss": 0.0908203125, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.623627662658691, "learning_rate": 1e-05, "loss": 0.4779, "mean_token_accuracy": 0.8692290782928467, "num_tokens": 127297487.0, "step": 3351 }, { "epoch": 0.4264088538353899, "ewc_loss": 0.08447265625, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.793699264526367, "learning_rate": 1e-05, "loss": 0.5746, "mean_token_accuracy": 0.8387569189071655, "num_tokens": 127342612.0, "step": 3352 }, { "epoch": 0.42653606411398043, "ewc_loss": 0.08642578125, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.560672760009766, "learning_rate": 1e-05, "loss": 0.5038, "mean_token_accuracy": 0.8634273409843445, "num_tokens": 127383915.0, "step": 3353 }, { "epoch": 0.4266632743925709, "ewc_loss": 0.07958984375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 7.295608520507812e-05, "grad_norm": 5.463809013366699, "learning_rate": 1e-05, "loss": 0.5497, "mean_token_accuracy": 0.8480344414710999, "num_tokens": 127421564.0, "step": 3354 }, { "epoch": 0.42679048467116143, "ewc_loss": 0.0830078125, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 7.62939453125e-05, "grad_norm": 5.529225826263428, "learning_rate": 1e-05, "loss": 0.5755, "mean_token_accuracy": 0.8363910913467407, "num_tokens": 127463750.0, "step": 3355 }, { "epoch": 0.42691769494975196, "ewc_loss": 0.0771484375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 7.009506225585938e-05, "grad_norm": 5.229416847229004, "learning_rate": 1e-05, "loss": 0.5493, "mean_token_accuracy": 0.8475228548049927, "num_tokens": 127502793.0, "step": 3356 }, { "epoch": 0.42704490522834243, "ewc_loss": 0.08349609375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.523138999938965, "learning_rate": 1e-05, "loss": 0.5324, "mean_token_accuracy": 0.8523541688919067, "num_tokens": 127542650.0, "step": 3357 }, { "epoch": 0.42717211550693296, "ewc_loss": 0.0751953125, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 6.818771362304688e-05, "grad_norm": 5.070140838623047, "learning_rate": 1e-05, "loss": 0.4903, "mean_token_accuracy": 0.8635423183441162, "num_tokens": 127584753.0, "step": 3358 }, { "epoch": 0.4272993257855235, "ewc_loss": 0.08251953125, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 7.581710815429688e-05, "grad_norm": 5.581801891326904, "learning_rate": 1e-05, "loss": 0.5436, "mean_token_accuracy": 0.8465696573257446, "num_tokens": 127624427.0, "step": 3359 }, { "epoch": 0.42742653606411396, "ewc_loss": 0.07421875, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 6.723403930664062e-05, "grad_norm": 4.940604209899902, "learning_rate": 1e-05, "loss": 0.4914, "mean_token_accuracy": 0.8630274534225464, "num_tokens": 127665636.0, "step": 3360 }, { "epoch": 0.4275537463427045, "ewc_loss": 0.083984375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.719616413116455, "learning_rate": 1e-05, "loss": 0.545, "mean_token_accuracy": 0.8487551212310791, "num_tokens": 127706066.0, "step": 3361 }, { "epoch": 0.427680956621295, "ewc_loss": 0.0732421875, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 6.628036499023438e-05, "grad_norm": 4.828743934631348, "learning_rate": 1e-05, "loss": 0.4815, "mean_token_accuracy": 0.8636693358421326, "num_tokens": 127741680.0, "step": 3362 }, { "epoch": 0.4278081668998855, "ewc_loss": 0.0888671875, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.979254245758057, "learning_rate": 1e-05, "loss": 0.5487, "mean_token_accuracy": 0.8490535020828247, "num_tokens": 127781147.0, "step": 3363 }, { "epoch": 0.427935377178476, "ewc_loss": 0.0732421875, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 6.628036499023438e-05, "grad_norm": 4.779466152191162, "learning_rate": 1e-05, "loss": 0.5809, "mean_token_accuracy": 0.8327522277832031, "num_tokens": 127823787.0, "step": 3364 }, { "epoch": 0.42806258745706655, "ewc_loss": 0.0986328125, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.305861949920654, "learning_rate": 1e-05, "loss": 0.6108, "mean_token_accuracy": 0.8316411972045898, "num_tokens": 127862239.0, "step": 3365 }, { "epoch": 0.428189797735657, "ewc_loss": 0.07763671875, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 7.05718994140625e-05, "grad_norm": 4.927002429962158, "learning_rate": 1e-05, "loss": 0.5719, "mean_token_accuracy": 0.8420217037200928, "num_tokens": 127904376.0, "step": 3366 }, { "epoch": 0.42831700801424755, "ewc_loss": 0.1025390625, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.539034843444824, "learning_rate": 1e-05, "loss": 0.5811, "mean_token_accuracy": 0.8413662910461426, "num_tokens": 127944132.0, "step": 3367 }, { "epoch": 0.4284442182928381, "ewc_loss": 0.0859375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.3102850914001465, "learning_rate": 1e-05, "loss": 0.5577, "mean_token_accuracy": 0.8427197933197021, "num_tokens": 127983512.0, "step": 3368 }, { "epoch": 0.42857142857142855, "ewc_loss": 0.10009765625, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.41459846496582, "learning_rate": 1e-05, "loss": 0.5047, "mean_token_accuracy": 0.8605233430862427, "num_tokens": 128018864.0, "step": 3369 }, { "epoch": 0.4286986388500191, "ewc_loss": 0.08935546875, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.558475971221924, "learning_rate": 1e-05, "loss": 0.4793, "mean_token_accuracy": 0.8680354356765747, "num_tokens": 128056586.0, "step": 3370 }, { "epoch": 0.4288258491286096, "ewc_loss": 0.08935546875, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 6.0587592124938965, "learning_rate": 1e-05, "loss": 0.5351, "mean_token_accuracy": 0.849101185798645, "num_tokens": 128091131.0, "step": 3371 }, { "epoch": 0.4289530594072001, "ewc_loss": 0.0888671875, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.594932556152344, "learning_rate": 1e-05, "loss": 0.5462, "mean_token_accuracy": 0.8473719358444214, "num_tokens": 128125455.0, "step": 3372 }, { "epoch": 0.4290802696857906, "ewc_loss": 0.08154296875, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 7.486343383789062e-05, "grad_norm": 5.667942047119141, "learning_rate": 1e-05, "loss": 0.5284, "mean_token_accuracy": 0.8496342897415161, "num_tokens": 128164357.0, "step": 3373 }, { "epoch": 0.42920747996438113, "ewc_loss": 0.08544921875, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.5431809425354, "learning_rate": 1e-05, "loss": 0.5456, "mean_token_accuracy": 0.8481688499450684, "num_tokens": 128203590.0, "step": 3374 }, { "epoch": 0.4293346902429716, "ewc_loss": 0.078125, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 7.152557373046875e-05, "grad_norm": 5.373078346252441, "learning_rate": 1e-05, "loss": 0.5295, "mean_token_accuracy": 0.849511444568634, "num_tokens": 128238819.0, "step": 3375 }, { "epoch": 0.42946190052156213, "ewc_loss": 0.0830078125, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 7.62939453125e-05, "grad_norm": 5.506878852844238, "learning_rate": 1e-05, "loss": 0.4731, "mean_token_accuracy": 0.8697148561477661, "num_tokens": 128273879.0, "step": 3376 }, { "epoch": 0.42958911080015266, "ewc_loss": 0.0771484375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 7.009506225585938e-05, "grad_norm": 5.1756792068481445, "learning_rate": 1e-05, "loss": 0.5098, "mean_token_accuracy": 0.8559489250183105, "num_tokens": 128309688.0, "step": 3377 }, { "epoch": 0.42971632107874314, "ewc_loss": 0.08154296875, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 7.486343383789062e-05, "grad_norm": 5.523984432220459, "learning_rate": 1e-05, "loss": 0.5153, "mean_token_accuracy": 0.854421854019165, "num_tokens": 128341515.0, "step": 3378 }, { "epoch": 0.42984353135733366, "ewc_loss": 0.0751953125, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 6.818771362304688e-05, "grad_norm": 5.028051376342773, "learning_rate": 1e-05, "loss": 0.507, "mean_token_accuracy": 0.8594234585762024, "num_tokens": 128382676.0, "step": 3379 }, { "epoch": 0.4299707416359242, "ewc_loss": 0.08251953125, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 7.581710815429688e-05, "grad_norm": 5.587759971618652, "learning_rate": 1e-05, "loss": 0.4582, "mean_token_accuracy": 0.8716919422149658, "num_tokens": 128417512.0, "step": 3380 }, { "epoch": 0.4300979519145147, "ewc_loss": 0.07421875, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 6.723403930664062e-05, "grad_norm": 4.901008605957031, "learning_rate": 1e-05, "loss": 0.4884, "mean_token_accuracy": 0.8582090735435486, "num_tokens": 128457695.0, "step": 3381 }, { "epoch": 0.4302251621931052, "ewc_loss": 0.0859375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.779380798339844, "learning_rate": 1e-05, "loss": 0.5221, "mean_token_accuracy": 0.8523776531219482, "num_tokens": 128487459.0, "step": 3382 }, { "epoch": 0.4303523724716957, "ewc_loss": 0.07275390625, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 6.580352783203125e-05, "grad_norm": 4.801213264465332, "learning_rate": 1e-05, "loss": 0.575, "mean_token_accuracy": 0.8307082653045654, "num_tokens": 128526823.0, "step": 3383 }, { "epoch": 0.43047958275028625, "ewc_loss": 0.0908203125, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 6.045782089233398, "learning_rate": 1e-05, "loss": 0.5308, "mean_token_accuracy": 0.8506184220314026, "num_tokens": 128566841.0, "step": 3384 }, { "epoch": 0.4306067930288767, "ewc_loss": 0.07373046875, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 6.67572021484375e-05, "grad_norm": 4.777064323425293, "learning_rate": 1e-05, "loss": 0.5376, "mean_token_accuracy": 0.8425015211105347, "num_tokens": 128608147.0, "step": 3385 }, { "epoch": 0.43073400330746725, "ewc_loss": 0.099609375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.396574020385742, "learning_rate": 1e-05, "loss": 0.5183, "mean_token_accuracy": 0.8579029440879822, "num_tokens": 128648277.0, "step": 3386 }, { "epoch": 0.4308612135860578, "ewc_loss": 0.07861328125, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 7.200241088867188e-05, "grad_norm": 5.007275581359863, "learning_rate": 1e-05, "loss": 0.577, "mean_token_accuracy": 0.8379427194595337, "num_tokens": 128682547.0, "step": 3387 }, { "epoch": 0.43098842386464825, "ewc_loss": 0.1044921875, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.56744384765625, "learning_rate": 1e-05, "loss": 0.5367, "mean_token_accuracy": 0.8564362525939941, "num_tokens": 128715103.0, "step": 3388 }, { "epoch": 0.4311156341432388, "ewc_loss": 0.087890625, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.412160873413086, "learning_rate": 1e-05, "loss": 0.5376, "mean_token_accuracy": 0.8484889268875122, "num_tokens": 128748924.0, "step": 3389 }, { "epoch": 0.4312428444218293, "ewc_loss": 0.10009765625, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.326169967651367, "learning_rate": 1e-05, "loss": 0.5564, "mean_token_accuracy": 0.8501595258712769, "num_tokens": 128786879.0, "step": 3390 }, { "epoch": 0.4313700547004198, "ewc_loss": 0.0908203125, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.595715522766113, "learning_rate": 1e-05, "loss": 0.5309, "mean_token_accuracy": 0.849540114402771, "num_tokens": 128822121.0, "step": 3391 }, { "epoch": 0.4314972649790103, "ewc_loss": 0.087890625, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.940863132476807, "learning_rate": 1e-05, "loss": 0.5918, "mean_token_accuracy": 0.8335568308830261, "num_tokens": 128858594.0, "step": 3392 }, { "epoch": 0.43162447525760084, "ewc_loss": 0.08837890625, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.58627986907959, "learning_rate": 1e-05, "loss": 0.5405, "mean_token_accuracy": 0.8522864580154419, "num_tokens": 128896147.0, "step": 3393 }, { "epoch": 0.4317516855361913, "ewc_loss": 0.07958984375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 7.295608520507812e-05, "grad_norm": 5.564632415771484, "learning_rate": 1e-05, "loss": 0.5556, "mean_token_accuracy": 0.8402808904647827, "num_tokens": 128930055.0, "step": 3394 }, { "epoch": 0.43187889581478184, "ewc_loss": 0.08642578125, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.54130220413208, "learning_rate": 1e-05, "loss": 0.5323, "mean_token_accuracy": 0.8508425354957581, "num_tokens": 128968577.0, "step": 3395 }, { "epoch": 0.43200610609337237, "ewc_loss": 0.078125, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 7.104873657226562e-05, "grad_norm": 5.2885661125183105, "learning_rate": 1e-05, "loss": 0.5397, "mean_token_accuracy": 0.8466272950172424, "num_tokens": 129007074.0, "step": 3396 }, { "epoch": 0.43213331637196284, "ewc_loss": 0.083984375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.540604114532471, "learning_rate": 1e-05, "loss": 0.5512, "mean_token_accuracy": 0.843031644821167, "num_tokens": 129044722.0, "step": 3397 }, { "epoch": 0.43226052665055337, "ewc_loss": 0.076171875, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 6.914138793945312e-05, "grad_norm": 5.099917411804199, "learning_rate": 1e-05, "loss": 0.5131, "mean_token_accuracy": 0.8544845581054688, "num_tokens": 129081567.0, "step": 3398 }, { "epoch": 0.4323877369291439, "ewc_loss": 0.08349609375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.583842754364014, "learning_rate": 1e-05, "loss": 0.5475, "mean_token_accuracy": 0.8480899930000305, "num_tokens": 129119354.0, "step": 3399 }, { "epoch": 0.43251494720773437, "ewc_loss": 0.07470703125, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 6.771087646484375e-05, "grad_norm": 4.967060089111328, "learning_rate": 1e-05, "loss": 0.4828, "mean_token_accuracy": 0.8656750321388245, "num_tokens": 129157871.0, "step": 3400 }, { "epoch": 0.4326421574863249, "ewc_loss": 0.0849609375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.703575611114502, "learning_rate": 1e-05, "loss": 0.5055, "mean_token_accuracy": 0.8621737360954285, "num_tokens": 129200164.0, "step": 3401 }, { "epoch": 0.4327693677649154, "ewc_loss": 0.07470703125, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 6.771087646484375e-05, "grad_norm": 4.8427581787109375, "learning_rate": 1e-05, "loss": 0.5283, "mean_token_accuracy": 0.8509495258331299, "num_tokens": 129239382.0, "step": 3402 }, { "epoch": 0.4328965780435059, "ewc_loss": 0.08935546875, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.9385786056518555, "learning_rate": 1e-05, "loss": 0.5504, "mean_token_accuracy": 0.8455471992492676, "num_tokens": 129273201.0, "step": 3403 }, { "epoch": 0.4330237883220964, "ewc_loss": 0.07373046875, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 6.67572021484375e-05, "grad_norm": 4.775954723358154, "learning_rate": 1e-05, "loss": 0.4909, "mean_token_accuracy": 0.8619651794433594, "num_tokens": 129309636.0, "step": 3404 }, { "epoch": 0.43315099860068695, "ewc_loss": 0.0986328125, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.2631306648254395, "learning_rate": 1e-05, "loss": 0.5092, "mean_token_accuracy": 0.8618740439414978, "num_tokens": 129345803.0, "step": 3405 }, { "epoch": 0.4332782088792774, "ewc_loss": 0.07763671875, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 7.05718994140625e-05, "grad_norm": 4.900223255157471, "learning_rate": 1e-05, "loss": 0.4949, "mean_token_accuracy": 0.857952892780304, "num_tokens": 129384955.0, "step": 3406 }, { "epoch": 0.43340541915786795, "ewc_loss": 0.10302734375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.504330158233643, "learning_rate": 1e-05, "loss": 0.4952, "mean_token_accuracy": 0.8700916171073914, "num_tokens": 129425129.0, "step": 3407 }, { "epoch": 0.4335326294364585, "ewc_loss": 0.0849609375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.252627372741699, "learning_rate": 1e-05, "loss": 0.5342, "mean_token_accuracy": 0.8504644632339478, "num_tokens": 129465931.0, "step": 3408 }, { "epoch": 0.43365983971504896, "ewc_loss": 0.10107421875, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.436559677124023, "learning_rate": 1e-05, "loss": 0.499, "mean_token_accuracy": 0.8632198572158813, "num_tokens": 129497296.0, "step": 3409 }, { "epoch": 0.4337870499936395, "ewc_loss": 0.09033203125, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.534060001373291, "learning_rate": 1e-05, "loss": 0.5151, "mean_token_accuracy": 0.8582452535629272, "num_tokens": 129539037.0, "step": 3410 }, { "epoch": 0.43391426027223, "ewc_loss": 0.0927734375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 6.091553211212158, "learning_rate": 1e-05, "loss": 0.5583, "mean_token_accuracy": 0.8479172587394714, "num_tokens": 129581036.0, "step": 3411 }, { "epoch": 0.4340414705508205, "ewc_loss": 0.0888671875, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.587099075317383, "learning_rate": 1e-05, "loss": 0.5492, "mean_token_accuracy": 0.8470300436019897, "num_tokens": 129617515.0, "step": 3412 }, { "epoch": 0.434168680829411, "ewc_loss": 0.08349609375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.706889629364014, "learning_rate": 1e-05, "loss": 0.4561, "mean_token_accuracy": 0.8734353184700012, "num_tokens": 129659348.0, "step": 3413 }, { "epoch": 0.43429589110800154, "ewc_loss": 0.0859375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.5579118728637695, "learning_rate": 1e-05, "loss": 0.5866, "mean_token_accuracy": 0.8335650563240051, "num_tokens": 129695000.0, "step": 3414 }, { "epoch": 0.434423101386592, "ewc_loss": 0.078125, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 7.104873657226562e-05, "grad_norm": 5.392110824584961, "learning_rate": 1e-05, "loss": 0.5348, "mean_token_accuracy": 0.846695065498352, "num_tokens": 129734125.0, "step": 3415 }, { "epoch": 0.43455031166518254, "ewc_loss": 0.083984375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.523670196533203, "learning_rate": 1e-05, "loss": 0.5322, "mean_token_accuracy": 0.8515406847000122, "num_tokens": 129774799.0, "step": 3416 }, { "epoch": 0.43467752194377307, "ewc_loss": 0.07666015625, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 6.961822509765625e-05, "grad_norm": 5.176896572113037, "learning_rate": 1e-05, "loss": 0.4894, "mean_token_accuracy": 0.8584408760070801, "num_tokens": 129814390.0, "step": 3417 }, { "epoch": 0.43480473222236354, "ewc_loss": 0.0830078125, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 7.62939453125e-05, "grad_norm": 5.538631439208984, "learning_rate": 1e-05, "loss": 0.4967, "mean_token_accuracy": 0.8610910177230835, "num_tokens": 129848172.0, "step": 3418 }, { "epoch": 0.43493194250095407, "ewc_loss": 0.0751953125, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 6.818771362304688e-05, "grad_norm": 5.050012111663818, "learning_rate": 1e-05, "loss": 0.4606, "mean_token_accuracy": 0.8673843145370483, "num_tokens": 129878728.0, "step": 3419 }, { "epoch": 0.4350591527795446, "ewc_loss": 0.08251953125, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 7.581710815429688e-05, "grad_norm": 5.596131324768066, "learning_rate": 1e-05, "loss": 0.5197, "mean_token_accuracy": 0.8521372079849243, "num_tokens": 129916843.0, "step": 3420 }, { "epoch": 0.4351863630581351, "ewc_loss": 0.0751953125, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 6.818771362304688e-05, "grad_norm": 4.926846027374268, "learning_rate": 1e-05, "loss": 0.5502, "mean_token_accuracy": 0.8416488170623779, "num_tokens": 129954614.0, "step": 3421 }, { "epoch": 0.4353135733367256, "ewc_loss": 0.0849609375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.739743709564209, "learning_rate": 1e-05, "loss": 0.5989, "mean_token_accuracy": 0.8344698548316956, "num_tokens": 129993084.0, "step": 3422 }, { "epoch": 0.43544078361531613, "ewc_loss": 0.07275390625, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 6.580352783203125e-05, "grad_norm": 4.826064586639404, "learning_rate": 1e-05, "loss": 0.4865, "mean_token_accuracy": 0.8599854707717896, "num_tokens": 130028313.0, "step": 3423 }, { "epoch": 0.4355679938939066, "ewc_loss": 0.0908203125, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.9904069900512695, "learning_rate": 1e-05, "loss": 0.5362, "mean_token_accuracy": 0.850472092628479, "num_tokens": 130066722.0, "step": 3424 }, { "epoch": 0.43569520417249713, "ewc_loss": 0.07373046875, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 6.67572021484375e-05, "grad_norm": 4.773194313049316, "learning_rate": 1e-05, "loss": 0.5208, "mean_token_accuracy": 0.8497700691223145, "num_tokens": 130108628.0, "step": 3425 }, { "epoch": 0.43582241445108766, "ewc_loss": 0.0986328125, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.323981285095215, "learning_rate": 1e-05, "loss": 0.4995, "mean_token_accuracy": 0.8660151958465576, "num_tokens": 130145770.0, "step": 3426 }, { "epoch": 0.43594962472967813, "ewc_loss": 0.07861328125, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 7.200241088867188e-05, "grad_norm": 4.937604904174805, "learning_rate": 1e-05, "loss": 0.5171, "mean_token_accuracy": 0.8538252711296082, "num_tokens": 130191539.0, "step": 3427 }, { "epoch": 0.43607683500826866, "ewc_loss": 0.103515625, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.541447639465332, "learning_rate": 1e-05, "loss": 0.5179, "mean_token_accuracy": 0.8644245862960815, "num_tokens": 130229358.0, "step": 3428 }, { "epoch": 0.4362040452868592, "ewc_loss": 0.0869140625, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.362569808959961, "learning_rate": 1e-05, "loss": 0.5305, "mean_token_accuracy": 0.8540554642677307, "num_tokens": 130272173.0, "step": 3429 }, { "epoch": 0.4363312555654497, "ewc_loss": 0.1005859375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.3561296463012695, "learning_rate": 1e-05, "loss": 0.5508, "mean_token_accuracy": 0.852250874042511, "num_tokens": 130313135.0, "step": 3430 }, { "epoch": 0.4364584658440402, "ewc_loss": 0.09033203125, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.543381690979004, "learning_rate": 1e-05, "loss": 0.4932, "mean_token_accuracy": 0.8635686635971069, "num_tokens": 130350595.0, "step": 3431 }, { "epoch": 0.4365856761226307, "ewc_loss": 0.09033203125, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 6.003981590270996, "learning_rate": 1e-05, "loss": 0.5093, "mean_token_accuracy": 0.8606020212173462, "num_tokens": 130390194.0, "step": 3432 }, { "epoch": 0.43671288640122125, "ewc_loss": 0.0888671875, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.564672946929932, "learning_rate": 1e-05, "loss": 0.5244, "mean_token_accuracy": 0.8558077216148376, "num_tokens": 130424768.0, "step": 3433 }, { "epoch": 0.4368400966798117, "ewc_loss": 0.08154296875, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 7.486343383789062e-05, "grad_norm": 5.636476993560791, "learning_rate": 1e-05, "loss": 0.5728, "mean_token_accuracy": 0.8396336436271667, "num_tokens": 130469457.0, "step": 3434 }, { "epoch": 0.43696730695840225, "ewc_loss": 0.0859375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.532639503479004, "learning_rate": 1e-05, "loss": 0.4851, "mean_token_accuracy": 0.8688229322433472, "num_tokens": 130506250.0, "step": 3435 }, { "epoch": 0.4370945172369928, "ewc_loss": 0.078125, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 7.152557373046875e-05, "grad_norm": 5.351987838745117, "learning_rate": 1e-05, "loss": 0.516, "mean_token_accuracy": 0.8539793491363525, "num_tokens": 130538605.0, "step": 3436 }, { "epoch": 0.43722172751558325, "ewc_loss": 0.083984375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.505232334136963, "learning_rate": 1e-05, "loss": 0.4763, "mean_token_accuracy": 0.8678894639015198, "num_tokens": 130577172.0, "step": 3437 }, { "epoch": 0.4373489377941738, "ewc_loss": 0.07763671875, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 7.05718994140625e-05, "grad_norm": 5.16406774520874, "learning_rate": 1e-05, "loss": 0.5232, "mean_token_accuracy": 0.8516682386398315, "num_tokens": 130613067.0, "step": 3438 }, { "epoch": 0.4374761480727643, "ewc_loss": 0.08251953125, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 7.581710815429688e-05, "grad_norm": 5.517925262451172, "learning_rate": 1e-05, "loss": 0.4675, "mean_token_accuracy": 0.8716517686843872, "num_tokens": 130649176.0, "step": 3439 }, { "epoch": 0.4376033583513548, "ewc_loss": 0.0751953125, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 6.818771362304688e-05, "grad_norm": 5.037925720214844, "learning_rate": 1e-05, "loss": 0.5124, "mean_token_accuracy": 0.8559329509735107, "num_tokens": 130685904.0, "step": 3440 }, { "epoch": 0.4377305686299453, "ewc_loss": 0.08349609375, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.5900797843933105, "learning_rate": 1e-05, "loss": 0.4796, "mean_token_accuracy": 0.8666399717330933, "num_tokens": 130727502.0, "step": 3441 }, { "epoch": 0.43785777890853583, "ewc_loss": 0.0751953125, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 6.818771362304688e-05, "grad_norm": 4.910103797912598, "learning_rate": 1e-05, "loss": 0.5362, "mean_token_accuracy": 0.8458899855613708, "num_tokens": 130765038.0, "step": 3442 }, { "epoch": 0.4379849891871263, "ewc_loss": 0.08544921875, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.762320518493652, "learning_rate": 1e-05, "loss": 0.4979, "mean_token_accuracy": 0.8605982661247253, "num_tokens": 130810349.0, "step": 3443 }, { "epoch": 0.43811219946571683, "ewc_loss": 0.07275390625, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 6.580352783203125e-05, "grad_norm": 4.809067726135254, "learning_rate": 1e-05, "loss": 0.5385, "mean_token_accuracy": 0.8466442823410034, "num_tokens": 130843383.0, "step": 3444 }, { "epoch": 0.43823940974430736, "ewc_loss": 0.091796875, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 6.039305210113525, "learning_rate": 1e-05, "loss": 0.5484, "mean_token_accuracy": 0.8499161005020142, "num_tokens": 130880399.0, "step": 3445 }, { "epoch": 0.43836662002289783, "ewc_loss": 0.07373046875, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 6.67572021484375e-05, "grad_norm": 4.770018577575684, "learning_rate": 1e-05, "loss": 0.5726, "mean_token_accuracy": 0.8378778696060181, "num_tokens": 130915537.0, "step": 3446 }, { "epoch": 0.43849383030148836, "ewc_loss": 0.10009765625, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.396958827972412, "learning_rate": 1e-05, "loss": 0.5123, "mean_token_accuracy": 0.8633038997650146, "num_tokens": 130948560.0, "step": 3447 }, { "epoch": 0.4386210405800789, "ewc_loss": 0.07958984375, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 7.295608520507812e-05, "grad_norm": 4.981729984283447, "learning_rate": 1e-05, "loss": 0.4875, "mean_token_accuracy": 0.8621832132339478, "num_tokens": 130982405.0, "step": 3448 }, { "epoch": 0.43874825085866936, "ewc_loss": 0.10498046875, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.592703342437744, "learning_rate": 1e-05, "loss": 0.5203, "mean_token_accuracy": 0.8595514297485352, "num_tokens": 131022673.0, "step": 3449 }, { "epoch": 0.4388754611372599, "ewc_loss": 0.08837890625, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.398662090301514, "learning_rate": 1e-05, "loss": 0.5225, "mean_token_accuracy": 0.8516850471496582, "num_tokens": 131054647.0, "step": 3450 }, { "epoch": 0.4390026714158504, "ewc_loss": 0.10205078125, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.396697998046875, "learning_rate": 1e-05, "loss": 0.52, "mean_token_accuracy": 0.8599478602409363, "num_tokens": 131091616.0, "step": 3451 }, { "epoch": 0.4391298816944409, "ewc_loss": 0.091796875, "ewc_loss_diag": 6.914138793945312e-05, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.624485015869141, "learning_rate": 1e-05, "loss": 0.6082, "mean_token_accuracy": 0.8274463415145874, "num_tokens": 131129663.0, "step": 3452 }, { "epoch": 0.4392570919730314, "ewc_loss": 0.0888671875, "ewc_loss_diag": 6.914138793945312e-05, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.992369651794434, "learning_rate": 1e-05, "loss": 0.507, "mean_token_accuracy": 0.8582617044448853, "num_tokens": 131170858.0, "step": 3453 }, { "epoch": 0.43938430225162195, "ewc_loss": 0.08984375, "ewc_loss_diag": 6.914138793945312e-05, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.606208324432373, "learning_rate": 1e-05, "loss": 0.5236, "mean_token_accuracy": 0.8528896570205688, "num_tokens": 131210892.0, "step": 3454 }, { "epoch": 0.4395115125302124, "ewc_loss": 0.0810546875, "ewc_loss_diag": 6.914138793945312e-05, "ewc_loss_parallel": 7.43865966796875e-05, "grad_norm": 5.62492561340332, "learning_rate": 1e-05, "loss": 0.5317, "mean_token_accuracy": 0.8526366353034973, "num_tokens": 131246616.0, "step": 3455 }, { "epoch": 0.43963872280880295, "ewc_loss": 0.08642578125, "ewc_loss_diag": 6.914138793945312e-05, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.555943489074707, "learning_rate": 1e-05, "loss": 0.5138, "mean_token_accuracy": 0.8570495843887329, "num_tokens": 131286515.0, "step": 3456 }, { "epoch": 0.4397659330873935, "ewc_loss": 0.078125, "ewc_loss_diag": 6.914138793945312e-05, "ewc_loss_parallel": 7.152557373046875e-05, "grad_norm": 5.340838432312012, "learning_rate": 1e-05, "loss": 0.5237, "mean_token_accuracy": 0.852863073348999, "num_tokens": 131321715.0, "step": 3457 }, { "epoch": 0.43989314336598395, "ewc_loss": 0.08349609375, "ewc_loss_diag": 6.914138793945312e-05, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.543548583984375, "learning_rate": 1e-05, "loss": 0.565, "mean_token_accuracy": 0.8454007506370544, "num_tokens": 131359235.0, "step": 3458 }, { "epoch": 0.4400203536445745, "ewc_loss": 0.0771484375, "ewc_loss_diag": 6.914138793945312e-05, "ewc_loss_parallel": 7.009506225585938e-05, "grad_norm": 5.151271820068359, "learning_rate": 1e-05, "loss": 0.4563, "mean_token_accuracy": 0.8736845254898071, "num_tokens": 131390551.0, "step": 3459 }, { "epoch": 0.440147563923165, "ewc_loss": 0.08349609375, "ewc_loss_diag": 6.914138793945312e-05, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.5748796463012695, "learning_rate": 1e-05, "loss": 0.5203, "mean_token_accuracy": 0.8560992479324341, "num_tokens": 131424107.0, "step": 3460 }, { "epoch": 0.4402747742017555, "ewc_loss": 0.07568359375, "ewc_loss_diag": 6.914138793945312e-05, "ewc_loss_parallel": 6.866455078125e-05, "grad_norm": 5.011514186859131, "learning_rate": 1e-05, "loss": 0.4911, "mean_token_accuracy": 0.861918032169342, "num_tokens": 131459797.0, "step": 3461 }, { "epoch": 0.440401984480346, "ewc_loss": 0.08349609375, "ewc_loss_diag": 6.914138793945312e-05, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.691047191619873, "learning_rate": 1e-05, "loss": 0.5206, "mean_token_accuracy": 0.8543107509613037, "num_tokens": 131496877.0, "step": 3462 }, { "epoch": 0.44052919475893654, "ewc_loss": 0.07421875, "ewc_loss_diag": 6.914138793945312e-05, "ewc_loss_parallel": 6.723403930664062e-05, "grad_norm": 4.891326904296875, "learning_rate": 1e-05, "loss": 0.4848, "mean_token_accuracy": 0.8635611534118652, "num_tokens": 131528958.0, "step": 3463 }, { "epoch": 0.440656405037527, "ewc_loss": 0.08740234375, "ewc_loss_diag": 6.914138793945312e-05, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.878724575042725, "learning_rate": 1e-05, "loss": 0.5869, "mean_token_accuracy": 0.8344981670379639, "num_tokens": 131572157.0, "step": 3464 }, { "epoch": 0.44078361531611754, "ewc_loss": 0.0732421875, "ewc_loss_diag": 6.914138793945312e-05, "ewc_loss_parallel": 6.628036499023438e-05, "grad_norm": 4.794666767120361, "learning_rate": 1e-05, "loss": 0.529, "mean_token_accuracy": 0.848726749420166, "num_tokens": 131613855.0, "step": 3465 }, { "epoch": 0.44091082559470807, "ewc_loss": 0.09619140625, "ewc_loss_diag": 6.914138793945312e-05, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 6.190083026885986, "learning_rate": 1e-05, "loss": 0.5399, "mean_token_accuracy": 0.8531767129898071, "num_tokens": 131654056.0, "step": 3466 }, { "epoch": 0.44103803587329854, "ewc_loss": 0.07568359375, "ewc_loss_diag": 6.914138793945312e-05, "ewc_loss_parallel": 6.866455078125e-05, "grad_norm": 4.857540607452393, "learning_rate": 1e-05, "loss": 0.575, "mean_token_accuracy": 0.8328020572662354, "num_tokens": 131694865.0, "step": 3467 }, { "epoch": 0.44116524615188907, "ewc_loss": 0.10205078125, "ewc_loss_diag": 6.914138793945312e-05, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.481683731079102, "learning_rate": 1e-05, "loss": 0.545, "mean_token_accuracy": 0.8546825051307678, "num_tokens": 131732883.0, "step": 3468 }, { "epoch": 0.4412924564304796, "ewc_loss": 0.08349609375, "ewc_loss_diag": 6.914138793945312e-05, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.170371055603027, "learning_rate": 1e-05, "loss": 0.5451, "mean_token_accuracy": 0.8475061655044556, "num_tokens": 131770384.0, "step": 3469 }, { "epoch": 0.44141966670907007, "ewc_loss": 0.1025390625, "ewc_loss_diag": 6.961822509765625e-05, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.50040864944458, "learning_rate": 1e-05, "loss": 0.5463, "mean_token_accuracy": 0.849932074546814, "num_tokens": 131801611.0, "step": 3470 }, { "epoch": 0.4415468769876606, "ewc_loss": 0.08984375, "ewc_loss_diag": 6.961822509765625e-05, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.507730960845947, "learning_rate": 1e-05, "loss": 0.6345, "mean_token_accuracy": 0.8301821947097778, "num_tokens": 131842522.0, "step": 3471 }, { "epoch": 0.4416740872662511, "ewc_loss": 0.0966796875, "ewc_loss_diag": 6.961822509765625e-05, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.1930060386657715, "learning_rate": 1e-05, "loss": 0.6039, "mean_token_accuracy": 0.8387033939361572, "num_tokens": 131880041.0, "step": 3472 }, { "epoch": 0.4418012975448416, "ewc_loss": 0.08984375, "ewc_loss_diag": 6.961822509765625e-05, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.597971439361572, "learning_rate": 1e-05, "loss": 0.5199, "mean_token_accuracy": 0.8571392893791199, "num_tokens": 131917682.0, "step": 3473 }, { "epoch": 0.4419285078234321, "ewc_loss": 0.08447265625, "ewc_loss_diag": 6.961822509765625e-05, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.798090934753418, "learning_rate": 1e-05, "loss": 0.5307, "mean_token_accuracy": 0.8519328236579895, "num_tokens": 131958149.0, "step": 3474 }, { "epoch": 0.44205571810202265, "ewc_loss": 0.08740234375, "ewc_loss_diag": 6.961822509765625e-05, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.568982124328613, "learning_rate": 1e-05, "loss": 0.5623, "mean_token_accuracy": 0.8404236435890198, "num_tokens": 131996648.0, "step": 3475 }, { "epoch": 0.4421829283806131, "ewc_loss": 0.080078125, "ewc_loss_diag": 6.961822509765625e-05, "ewc_loss_parallel": 7.343292236328125e-05, "grad_norm": 5.442787170410156, "learning_rate": 1e-05, "loss": 0.5163, "mean_token_accuracy": 0.8554362654685974, "num_tokens": 132036524.0, "step": 3476 }, { "epoch": 0.44231013865920366, "ewc_loss": 0.0849609375, "ewc_loss_diag": 6.961822509765625e-05, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.570436477661133, "learning_rate": 1e-05, "loss": 0.5175, "mean_token_accuracy": 0.8560116291046143, "num_tokens": 132077474.0, "step": 3477 }, { "epoch": 0.4424373489377942, "ewc_loss": 0.0771484375, "ewc_loss_diag": 6.961822509765625e-05, "ewc_loss_parallel": 7.009506225585938e-05, "grad_norm": 5.200636386871338, "learning_rate": 1e-05, "loss": 0.5542, "mean_token_accuracy": 0.8388268351554871, "num_tokens": 132117153.0, "step": 3478 }, { "epoch": 0.44256455921638466, "ewc_loss": 0.0830078125, "ewc_loss_diag": 6.961822509765625e-05, "ewc_loss_parallel": 7.62939453125e-05, "grad_norm": 5.580759525299072, "learning_rate": 1e-05, "loss": 0.4758, "mean_token_accuracy": 0.8665729761123657, "num_tokens": 132156832.0, "step": 3479 }, { "epoch": 0.4426917694949752, "ewc_loss": 0.076171875, "ewc_loss_diag": 6.961822509765625e-05, "ewc_loss_parallel": 6.914138793945312e-05, "grad_norm": 5.049437046051025, "learning_rate": 1e-05, "loss": 0.5188, "mean_token_accuracy": 0.8554689884185791, "num_tokens": 132189245.0, "step": 3480 }, { "epoch": 0.4428189797735657, "ewc_loss": 0.083984375, "ewc_loss_diag": 6.961822509765625e-05, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.6449809074401855, "learning_rate": 1e-05, "loss": 0.5225, "mean_token_accuracy": 0.8539462089538574, "num_tokens": 132230572.0, "step": 3481 }, { "epoch": 0.44294619005215624, "ewc_loss": 0.07470703125, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 6.771087646484375e-05, "grad_norm": 4.922539710998535, "learning_rate": 1e-05, "loss": 0.5076, "mean_token_accuracy": 0.8542632460594177, "num_tokens": 132271850.0, "step": 3482 }, { "epoch": 0.4430734003307467, "ewc_loss": 0.08642578125, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.8060688972473145, "learning_rate": 1e-05, "loss": 0.5113, "mean_token_accuracy": 0.8574517965316772, "num_tokens": 132312461.0, "step": 3483 }, { "epoch": 0.44320061060933724, "ewc_loss": 0.07373046875, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 6.67572021484375e-05, "grad_norm": 4.824979305267334, "learning_rate": 1e-05, "loss": 0.4874, "mean_token_accuracy": 0.8606675863265991, "num_tokens": 132350628.0, "step": 3484 }, { "epoch": 0.44332782088792777, "ewc_loss": 0.091796875, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 6.0742292404174805, "learning_rate": 1e-05, "loss": 0.5428, "mean_token_accuracy": 0.8525177240371704, "num_tokens": 132394040.0, "step": 3485 }, { "epoch": 0.44345503116651824, "ewc_loss": 0.07470703125, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 6.771087646484375e-05, "grad_norm": 4.823233127593994, "learning_rate": 1e-05, "loss": 0.5051, "mean_token_accuracy": 0.8543015718460083, "num_tokens": 132435727.0, "step": 3486 }, { "epoch": 0.44358224144510877, "ewc_loss": 0.10009765625, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.380453586578369, "learning_rate": 1e-05, "loss": 0.5471, "mean_token_accuracy": 0.8499755859375, "num_tokens": 132475769.0, "step": 3487 }, { "epoch": 0.4437094517236993, "ewc_loss": 0.0810546875, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 7.43865966796875e-05, "grad_norm": 5.0443572998046875, "learning_rate": 1e-05, "loss": 0.5291, "mean_token_accuracy": 0.8510960340499878, "num_tokens": 132520944.0, "step": 3488 }, { "epoch": 0.4438366620022898, "ewc_loss": 0.1025390625, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.555490493774414, "learning_rate": 1e-05, "loss": 0.5376, "mean_token_accuracy": 0.8512928485870361, "num_tokens": 132558138.0, "step": 3489 }, { "epoch": 0.4439638722808803, "ewc_loss": 0.0888671875, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.422366142272949, "learning_rate": 1e-05, "loss": 0.5993, "mean_token_accuracy": 0.8296734094619751, "num_tokens": 132594937.0, "step": 3490 }, { "epoch": 0.44409108255947083, "ewc_loss": 0.0986328125, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.297921657562256, "learning_rate": 1e-05, "loss": 0.5969, "mean_token_accuracy": 0.8363286852836609, "num_tokens": 132629168.0, "step": 3491 }, { "epoch": 0.4442182928380613, "ewc_loss": 0.08935546875, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.572731018066406, "learning_rate": 1e-05, "loss": 0.6219, "mean_token_accuracy": 0.8261669874191284, "num_tokens": 132673239.0, "step": 3492 }, { "epoch": 0.44434550311665183, "ewc_loss": 0.0869140625, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.934509754180908, "learning_rate": 1e-05, "loss": 0.5142, "mean_token_accuracy": 0.8609303832054138, "num_tokens": 132707429.0, "step": 3493 }, { "epoch": 0.44447271339524236, "ewc_loss": 0.0888671875, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.564719200134277, "learning_rate": 1e-05, "loss": 0.524, "mean_token_accuracy": 0.8553056716918945, "num_tokens": 132745592.0, "step": 3494 }, { "epoch": 0.44459992367383283, "ewc_loss": 0.08154296875, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 7.486343383789062e-05, "grad_norm": 5.557610988616943, "learning_rate": 1e-05, "loss": 0.4934, "mean_token_accuracy": 0.8612834811210632, "num_tokens": 132778260.0, "step": 3495 }, { "epoch": 0.44472713395242336, "ewc_loss": 0.0869140625, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.555471420288086, "learning_rate": 1e-05, "loss": 0.4266, "mean_token_accuracy": 0.8823947906494141, "num_tokens": 132813156.0, "step": 3496 }, { "epoch": 0.4448543442310139, "ewc_loss": 0.078125, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 7.152557373046875e-05, "grad_norm": 5.2809648513793945, "learning_rate": 1e-05, "loss": 0.5229, "mean_token_accuracy": 0.8511797189712524, "num_tokens": 132847654.0, "step": 3497 }, { "epoch": 0.44498155450960436, "ewc_loss": 0.08349609375, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.553480625152588, "learning_rate": 1e-05, "loss": 0.4946, "mean_token_accuracy": 0.8617026805877686, "num_tokens": 132889258.0, "step": 3498 }, { "epoch": 0.4451087647881949, "ewc_loss": 0.0771484375, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 6.961822509765625e-05, "grad_norm": 5.106026649475098, "learning_rate": 1e-05, "loss": 0.5838, "mean_token_accuracy": 0.8319559097290039, "num_tokens": 132919962.0, "step": 3499 }, { "epoch": 0.4452359750667854, "ewc_loss": 0.083984375, "ewc_loss_diag": 7.152557373046875e-05, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.610332489013672, "learning_rate": 1e-05, "loss": 0.5584, "mean_token_accuracy": 0.846979558467865, "num_tokens": 132957333.0, "step": 3500 }, { "epoch": 0.4453631853453759, "ewc_loss": 0.07666015625, "ewc_loss_diag": 7.152557373046875e-05, "ewc_loss_parallel": 6.914138793945312e-05, "grad_norm": 4.972036361694336, "learning_rate": 1e-05, "loss": 0.5045, "mean_token_accuracy": 0.8538695573806763, "num_tokens": 132990645.0, "step": 3501 }, { "epoch": 0.4454903956239664, "ewc_loss": 0.0869140625, "ewc_loss_diag": 7.152557373046875e-05, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.7290239334106445, "learning_rate": 1e-05, "loss": 0.4738, "mean_token_accuracy": 0.8644935488700867, "num_tokens": 133026234.0, "step": 3502 }, { "epoch": 0.44561760590255695, "ewc_loss": 0.07470703125, "ewc_loss_diag": 7.152557373046875e-05, "ewc_loss_parallel": 6.723403930664062e-05, "grad_norm": 4.854035377502441, "learning_rate": 1e-05, "loss": 0.5058, "mean_token_accuracy": 0.8558134436607361, "num_tokens": 133062833.0, "step": 3503 }, { "epoch": 0.4457448161811474, "ewc_loss": 0.09130859375, "ewc_loss_diag": 7.152557373046875e-05, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.9757184982299805, "learning_rate": 1e-05, "loss": 0.4864, "mean_token_accuracy": 0.8676419258117676, "num_tokens": 133097868.0, "step": 3504 }, { "epoch": 0.44587202645973795, "ewc_loss": 0.07470703125, "ewc_loss_diag": 7.152557373046875e-05, "ewc_loss_parallel": 6.723403930664062e-05, "grad_norm": 4.791730880737305, "learning_rate": 1e-05, "loss": 0.4945, "mean_token_accuracy": 0.8602171540260315, "num_tokens": 133142062.0, "step": 3505 }, { "epoch": 0.4459992367383285, "ewc_loss": 0.10009765625, "ewc_loss_diag": 7.152557373046875e-05, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.303720951080322, "learning_rate": 1e-05, "loss": 0.5537, "mean_token_accuracy": 0.8499996066093445, "num_tokens": 133181415.0, "step": 3506 }, { "epoch": 0.44612644701691895, "ewc_loss": 0.0791015625, "ewc_loss_diag": 7.152557373046875e-05, "ewc_loss_parallel": 7.200241088867188e-05, "grad_norm": 4.931419849395752, "learning_rate": 1e-05, "loss": 0.5158, "mean_token_accuracy": 0.8561550378799438, "num_tokens": 133217849.0, "step": 3507 }, { "epoch": 0.4462536572955095, "ewc_loss": 0.10498046875, "ewc_loss_diag": 7.152557373046875e-05, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.553658962249756, "learning_rate": 1e-05, "loss": 0.5491, "mean_token_accuracy": 0.8523790240287781, "num_tokens": 133257639.0, "step": 3508 }, { "epoch": 0.4463808675741, "ewc_loss": 0.08740234375, "ewc_loss_diag": 7.152557373046875e-05, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.298230171203613, "learning_rate": 1e-05, "loss": 0.5431, "mean_token_accuracy": 0.8493029475212097, "num_tokens": 133299692.0, "step": 3509 }, { "epoch": 0.4465080778526905, "ewc_loss": 0.10302734375, "ewc_loss_diag": 7.152557373046875e-05, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.460653305053711, "learning_rate": 1e-05, "loss": 0.594, "mean_token_accuracy": 0.8387644290924072, "num_tokens": 133339315.0, "step": 3510 }, { "epoch": 0.446635288131281, "ewc_loss": 0.09130859375, "ewc_loss_diag": 7.152557373046875e-05, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.5683770179748535, "learning_rate": 1e-05, "loss": 0.5221, "mean_token_accuracy": 0.8570515513420105, "num_tokens": 133378477.0, "step": 3511 }, { "epoch": 0.44676249840987153, "ewc_loss": 0.09375, "ewc_loss_diag": 7.152557373046875e-05, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 6.115294456481934, "learning_rate": 1e-05, "loss": 0.6149, "mean_token_accuracy": 0.8298392295837402, "num_tokens": 133416041.0, "step": 3512 }, { "epoch": 0.446889708688462, "ewc_loss": 0.0908203125, "ewc_loss_diag": 7.152557373046875e-05, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.599535942077637, "learning_rate": 1e-05, "loss": 0.5192, "mean_token_accuracy": 0.8587193489074707, "num_tokens": 133458201.0, "step": 3513 }, { "epoch": 0.44701691896705253, "ewc_loss": 0.0849609375, "ewc_loss_diag": 7.152557373046875e-05, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.736486434936523, "learning_rate": 1e-05, "loss": 0.5401, "mean_token_accuracy": 0.8557519912719727, "num_tokens": 133495645.0, "step": 3514 }, { "epoch": 0.44714412924564306, "ewc_loss": 0.08642578125, "ewc_loss_diag": 7.152557373046875e-05, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.558006763458252, "learning_rate": 1e-05, "loss": 0.5514, "mean_token_accuracy": 0.8480590581893921, "num_tokens": 133535385.0, "step": 3515 }, { "epoch": 0.44727133952423354, "ewc_loss": 0.0810546875, "ewc_loss_diag": 7.152557373046875e-05, "ewc_loss_parallel": 7.390975952148438e-05, "grad_norm": 5.425468921661377, "learning_rate": 1e-05, "loss": 0.5106, "mean_token_accuracy": 0.8588314056396484, "num_tokens": 133575093.0, "step": 3516 }, { "epoch": 0.44739854980282406, "ewc_loss": 0.083984375, "ewc_loss_diag": 7.152557373046875e-05, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.5315423011779785, "learning_rate": 1e-05, "loss": 0.5341, "mean_token_accuracy": 0.8450301885604858, "num_tokens": 133610447.0, "step": 3517 }, { "epoch": 0.4475257600814146, "ewc_loss": 0.07763671875, "ewc_loss_diag": 7.152557373046875e-05, "ewc_loss_parallel": 7.009506225585938e-05, "grad_norm": 5.214746952056885, "learning_rate": 1e-05, "loss": 0.4911, "mean_token_accuracy": 0.8618428707122803, "num_tokens": 133643506.0, "step": 3518 }, { "epoch": 0.44765297036000506, "ewc_loss": 0.0830078125, "ewc_loss_diag": 7.152557373046875e-05, "ewc_loss_parallel": 7.581710815429688e-05, "grad_norm": 5.544379711151123, "learning_rate": 1e-05, "loss": 0.4469, "mean_token_accuracy": 0.8763155937194824, "num_tokens": 133683081.0, "step": 3519 }, { "epoch": 0.4477801806385956, "ewc_loss": 0.0771484375, "ewc_loss_diag": 7.152557373046875e-05, "ewc_loss_parallel": 6.961822509765625e-05, "grad_norm": 5.068676948547363, "learning_rate": 1e-05, "loss": 0.5483, "mean_token_accuracy": 0.8453634977340698, "num_tokens": 133721836.0, "step": 3520 }, { "epoch": 0.4479073909171861, "ewc_loss": 0.08349609375, "ewc_loss_diag": 7.152557373046875e-05, "ewc_loss_parallel": 7.62939453125e-05, "grad_norm": 5.598847389221191, "learning_rate": 1e-05, "loss": 0.5104, "mean_token_accuracy": 0.8590162992477417, "num_tokens": 133756243.0, "step": 3521 }, { "epoch": 0.4480346011957766, "ewc_loss": 0.07568359375, "ewc_loss_diag": 7.152557373046875e-05, "ewc_loss_parallel": 6.818771362304688e-05, "grad_norm": 4.935513019561768, "learning_rate": 1e-05, "loss": 0.5487, "mean_token_accuracy": 0.8413872718811035, "num_tokens": 133807126.0, "step": 3522 }, { "epoch": 0.4481618114743671, "ewc_loss": 0.0869140625, "ewc_loss_diag": 7.200241088867188e-05, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.766719341278076, "learning_rate": 1e-05, "loss": 0.538, "mean_token_accuracy": 0.8504430055618286, "num_tokens": 133842036.0, "step": 3523 }, { "epoch": 0.44828902175295765, "ewc_loss": 0.0751953125, "ewc_loss_diag": 7.2479248046875e-05, "ewc_loss_parallel": 6.771087646484375e-05, "grad_norm": 4.832939147949219, "learning_rate": 1e-05, "loss": 0.5638, "mean_token_accuracy": 0.839340090751648, "num_tokens": 133881948.0, "step": 3524 }, { "epoch": 0.4484162320315481, "ewc_loss": 0.091796875, "ewc_loss_diag": 7.2479248046875e-05, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 6.030071258544922, "learning_rate": 1e-05, "loss": 0.5017, "mean_token_accuracy": 0.8613386750221252, "num_tokens": 133917765.0, "step": 3525 }, { "epoch": 0.44854344231013865, "ewc_loss": 0.07470703125, "ewc_loss_diag": 7.2479248046875e-05, "ewc_loss_parallel": 6.723403930664062e-05, "grad_norm": 4.778676509857178, "learning_rate": 1e-05, "loss": 0.5157, "mean_token_accuracy": 0.8512094616889954, "num_tokens": 133959199.0, "step": 3526 }, { "epoch": 0.4486706525887292, "ewc_loss": 0.10009765625, "ewc_loss_diag": 7.2479248046875e-05, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.390082836151123, "learning_rate": 1e-05, "loss": 0.5759, "mean_token_accuracy": 0.8422075510025024, "num_tokens": 133991484.0, "step": 3527 }, { "epoch": 0.44879786286731965, "ewc_loss": 0.08056640625, "ewc_loss_diag": 7.2479248046875e-05, "ewc_loss_parallel": 7.343292236328125e-05, "grad_norm": 4.982041358947754, "learning_rate": 1e-05, "loss": 0.4798, "mean_token_accuracy": 0.8635563850402832, "num_tokens": 134025817.0, "step": 3528 }, { "epoch": 0.4489250731459102, "ewc_loss": 0.10595703125, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.602048397064209, "learning_rate": 1e-05, "loss": 0.5333, "mean_token_accuracy": 0.8537172675132751, "num_tokens": 134063956.0, "step": 3529 }, { "epoch": 0.4490522834245007, "ewc_loss": 0.0888671875, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.378976821899414, "learning_rate": 1e-05, "loss": 0.5265, "mean_token_accuracy": 0.8526721000671387, "num_tokens": 134101959.0, "step": 3530 }, { "epoch": 0.44917949370309124, "ewc_loss": 0.10302734375, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.439203262329102, "learning_rate": 1e-05, "loss": 0.5502, "mean_token_accuracy": 0.8479259014129639, "num_tokens": 134139961.0, "step": 3531 }, { "epoch": 0.4493067039816817, "ewc_loss": 0.0927734375, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.617669105529785, "learning_rate": 1e-05, "loss": 0.5533, "mean_token_accuracy": 0.8465164303779602, "num_tokens": 134180368.0, "step": 3532 }, { "epoch": 0.44943391426027224, "ewc_loss": 0.0908203125, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 6.05657434463501, "learning_rate": 1e-05, "loss": 0.4949, "mean_token_accuracy": 0.8643604516983032, "num_tokens": 134216842.0, "step": 3533 }, { "epoch": 0.44956112453886277, "ewc_loss": 0.09033203125, "ewc_loss_diag": 7.343292236328125e-05, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.619649410247803, "learning_rate": 1e-05, "loss": 0.5152, "mean_token_accuracy": 0.8572033643722534, "num_tokens": 134254089.0, "step": 3534 }, { "epoch": 0.44968833481745324, "ewc_loss": 0.08349609375, "ewc_loss_diag": 7.343292236328125e-05, "ewc_loss_parallel": 7.62939453125e-05, "grad_norm": 5.665635108947754, "learning_rate": 1e-05, "loss": 0.4851, "mean_token_accuracy": 0.8609657287597656, "num_tokens": 134289559.0, "step": 3535 }, { "epoch": 0.44981554509604377, "ewc_loss": 0.08642578125, "ewc_loss_diag": 7.343292236328125e-05, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.580332279205322, "learning_rate": 1e-05, "loss": 0.536, "mean_token_accuracy": 0.8551589250564575, "num_tokens": 134323029.0, "step": 3536 }, { "epoch": 0.4499427553746343, "ewc_loss": 0.07958984375, "ewc_loss_diag": 7.343292236328125e-05, "ewc_loss_parallel": 7.2479248046875e-05, "grad_norm": 5.364587306976318, "learning_rate": 1e-05, "loss": 0.5238, "mean_token_accuracy": 0.851516604423523, "num_tokens": 134362657.0, "step": 3537 }, { "epoch": 0.45006996565322477, "ewc_loss": 0.083984375, "ewc_loss_diag": 7.343292236328125e-05, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.566143989562988, "learning_rate": 1e-05, "loss": 0.506, "mean_token_accuracy": 0.8570944666862488, "num_tokens": 134403555.0, "step": 3538 }, { "epoch": 0.4501971759318153, "ewc_loss": 0.07763671875, "ewc_loss_diag": 7.343292236328125e-05, "ewc_loss_parallel": 7.009506225585938e-05, "grad_norm": 5.162985324859619, "learning_rate": 1e-05, "loss": 0.5079, "mean_token_accuracy": 0.8574325442314148, "num_tokens": 134438780.0, "step": 3539 }, { "epoch": 0.4503243862104058, "ewc_loss": 0.08349609375, "ewc_loss_diag": 7.343292236328125e-05, "ewc_loss_parallel": 7.62939453125e-05, "grad_norm": 5.597966194152832, "learning_rate": 1e-05, "loss": 0.5408, "mean_token_accuracy": 0.8491599559783936, "num_tokens": 134477331.0, "step": 3540 }, { "epoch": 0.4504515964889963, "ewc_loss": 0.07666015625, "ewc_loss_diag": 7.343292236328125e-05, "ewc_loss_parallel": 6.914138793945312e-05, "grad_norm": 5.030981063842773, "learning_rate": 1e-05, "loss": 0.4652, "mean_token_accuracy": 0.8674727082252502, "num_tokens": 134513898.0, "step": 3541 }, { "epoch": 0.4505788067675868, "ewc_loss": 0.08447265625, "ewc_loss_diag": 7.343292236328125e-05, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.690694808959961, "learning_rate": 1e-05, "loss": 0.5881, "mean_token_accuracy": 0.8390288949012756, "num_tokens": 134548507.0, "step": 3542 }, { "epoch": 0.45070601704617735, "ewc_loss": 0.0751953125, "ewc_loss_diag": 7.343292236328125e-05, "ewc_loss_parallel": 6.771087646484375e-05, "grad_norm": 4.908848762512207, "learning_rate": 1e-05, "loss": 0.5371, "mean_token_accuracy": 0.8479433655738831, "num_tokens": 134590243.0, "step": 3543 }, { "epoch": 0.4508332273247678, "ewc_loss": 0.0888671875, "ewc_loss_diag": 7.343292236328125e-05, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.884120941162109, "learning_rate": 1e-05, "loss": 0.512, "mean_token_accuracy": 0.8557994961738586, "num_tokens": 134624693.0, "step": 3544 }, { "epoch": 0.45096043760335836, "ewc_loss": 0.07421875, "ewc_loss_diag": 7.343292236328125e-05, "ewc_loss_parallel": 6.67572021484375e-05, "grad_norm": 4.81191349029541, "learning_rate": 1e-05, "loss": 0.4544, "mean_token_accuracy": 0.8754133582115173, "num_tokens": 134661365.0, "step": 3545 }, { "epoch": 0.4510876478819489, "ewc_loss": 0.0966796875, "ewc_loss_diag": 7.343292236328125e-05, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 6.190227508544922, "learning_rate": 1e-05, "loss": 0.5308, "mean_token_accuracy": 0.858198881149292, "num_tokens": 134698218.0, "step": 3546 }, { "epoch": 0.45121485816053936, "ewc_loss": 0.07666015625, "ewc_loss_diag": 7.343292236328125e-05, "ewc_loss_parallel": 6.914138793945312e-05, "grad_norm": 4.864167213439941, "learning_rate": 1e-05, "loss": 0.5372, "mean_token_accuracy": 0.8492810726165771, "num_tokens": 134736011.0, "step": 3547 }, { "epoch": 0.4513420684391299, "ewc_loss": 0.1025390625, "ewc_loss_diag": 7.343292236328125e-05, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.4780049324035645, "learning_rate": 1e-05, "loss": 0.5244, "mean_token_accuracy": 0.8580448627471924, "num_tokens": 134774457.0, "step": 3548 }, { "epoch": 0.4514692787177204, "ewc_loss": 0.08447265625, "ewc_loss_diag": 7.343292236328125e-05, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.162230968475342, "learning_rate": 1e-05, "loss": 0.5864, "mean_token_accuracy": 0.8321963548660278, "num_tokens": 134813788.0, "step": 3549 }, { "epoch": 0.4515964889963109, "ewc_loss": 0.103515625, "ewc_loss_diag": 7.343292236328125e-05, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.508695602416992, "learning_rate": 1e-05, "loss": 0.5732, "mean_token_accuracy": 0.8422831892967224, "num_tokens": 134851916.0, "step": 3550 }, { "epoch": 0.4517236992749014, "ewc_loss": 0.08935546875, "ewc_loss_diag": 7.343292236328125e-05, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.494931221008301, "learning_rate": 1e-05, "loss": 0.5006, "mean_token_accuracy": 0.8598589897155762, "num_tokens": 134892655.0, "step": 3551 }, { "epoch": 0.45185090955349194, "ewc_loss": 0.09716796875, "ewc_loss_diag": 7.343292236328125e-05, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.214775085449219, "learning_rate": 1e-05, "loss": 0.5386, "mean_token_accuracy": 0.8532516956329346, "num_tokens": 134931391.0, "step": 3552 }, { "epoch": 0.4519781198320824, "ewc_loss": 0.08984375, "ewc_loss_diag": 7.343292236328125e-05, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.61269998550415, "learning_rate": 1e-05, "loss": 0.5446, "mean_token_accuracy": 0.850175142288208, "num_tokens": 134966285.0, "step": 3553 }, { "epoch": 0.45210533011067294, "ewc_loss": 0.0869140625, "ewc_loss_diag": 7.343292236328125e-05, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.821410655975342, "learning_rate": 1e-05, "loss": 0.524, "mean_token_accuracy": 0.8554223775863647, "num_tokens": 135003893.0, "step": 3554 }, { "epoch": 0.45223254038926347, "ewc_loss": 0.08740234375, "ewc_loss_diag": 7.343292236328125e-05, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.579881191253662, "learning_rate": 1e-05, "loss": 0.5203, "mean_token_accuracy": 0.8553484678268433, "num_tokens": 135043245.0, "step": 3555 }, { "epoch": 0.45235975066785394, "ewc_loss": 0.08056640625, "ewc_loss_diag": 7.343292236328125e-05, "ewc_loss_parallel": 7.343292236328125e-05, "grad_norm": 5.469764232635498, "learning_rate": 1e-05, "loss": 0.5337, "mean_token_accuracy": 0.8518645763397217, "num_tokens": 135082867.0, "step": 3556 }, { "epoch": 0.45248696094644447, "ewc_loss": 0.0849609375, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.561828136444092, "learning_rate": 1e-05, "loss": 0.4502, "mean_token_accuracy": 0.8762458562850952, "num_tokens": 135121153.0, "step": 3557 }, { "epoch": 0.452614171225035, "ewc_loss": 0.07763671875, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 7.009506225585938e-05, "grad_norm": 5.2403740882873535, "learning_rate": 1e-05, "loss": 0.5072, "mean_token_accuracy": 0.8511985540390015, "num_tokens": 135154221.0, "step": 3558 }, { "epoch": 0.4527413815036255, "ewc_loss": 0.08447265625, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.566039562225342, "learning_rate": 1e-05, "loss": 0.5011, "mean_token_accuracy": 0.8554736375808716, "num_tokens": 135196065.0, "step": 3559 }, { "epoch": 0.452868591782216, "ewc_loss": 0.07666015625, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 6.914138793945312e-05, "grad_norm": 5.088578224182129, "learning_rate": 1e-05, "loss": 0.4913, "mean_token_accuracy": 0.8637490272521973, "num_tokens": 135233692.0, "step": 3560 }, { "epoch": 0.45299580206080653, "ewc_loss": 0.083984375, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.6170196533203125, "learning_rate": 1e-05, "loss": 0.4647, "mean_token_accuracy": 0.8718104362487793, "num_tokens": 135267538.0, "step": 3561 }, { "epoch": 0.453123012339397, "ewc_loss": 0.0751953125, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 6.771087646484375e-05, "grad_norm": 4.977928638458252, "learning_rate": 1e-05, "loss": 0.4805, "mean_token_accuracy": 0.8633466958999634, "num_tokens": 135303946.0, "step": 3562 }, { "epoch": 0.45325022261798753, "ewc_loss": 0.08642578125, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.740192413330078, "learning_rate": 1e-05, "loss": 0.5497, "mean_token_accuracy": 0.8443435430526733, "num_tokens": 135340940.0, "step": 3563 }, { "epoch": 0.45337743289657806, "ewc_loss": 0.0751953125, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 6.771087646484375e-05, "grad_norm": 4.875115394592285, "learning_rate": 1e-05, "loss": 0.5472, "mean_token_accuracy": 0.845666766166687, "num_tokens": 135377472.0, "step": 3564 }, { "epoch": 0.45350464317516853, "ewc_loss": 0.0908203125, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.969280242919922, "learning_rate": 1e-05, "loss": 0.5834, "mean_token_accuracy": 0.8360125422477722, "num_tokens": 135411396.0, "step": 3565 }, { "epoch": 0.45363185345375906, "ewc_loss": 0.07421875, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 6.67572021484375e-05, "grad_norm": 4.807701587677002, "learning_rate": 1e-05, "loss": 0.5396, "mean_token_accuracy": 0.8441381454467773, "num_tokens": 135445323.0, "step": 3566 }, { "epoch": 0.4537590637323496, "ewc_loss": 0.0986328125, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.278838157653809, "learning_rate": 1e-05, "loss": 0.5203, "mean_token_accuracy": 0.8586575388908386, "num_tokens": 135487364.0, "step": 3567 }, { "epoch": 0.45388627401094006, "ewc_loss": 0.07861328125, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 7.104873657226562e-05, "grad_norm": 4.915897846221924, "learning_rate": 1e-05, "loss": 0.5168, "mean_token_accuracy": 0.8559095859527588, "num_tokens": 135529879.0, "step": 3568 }, { "epoch": 0.4540134842895306, "ewc_loss": 0.1044921875, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.524821758270264, "learning_rate": 1e-05, "loss": 0.518, "mean_token_accuracy": 0.8602801561355591, "num_tokens": 135568812.0, "step": 3569 }, { "epoch": 0.4541406945681211, "ewc_loss": 0.0859375, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.268062591552734, "learning_rate": 1e-05, "loss": 0.5383, "mean_token_accuracy": 0.8497048616409302, "num_tokens": 135608096.0, "step": 3570 }, { "epoch": 0.4542679048467116, "ewc_loss": 0.1025390625, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.451358795166016, "learning_rate": 1e-05, "loss": 0.6275, "mean_token_accuracy": 0.8288657665252686, "num_tokens": 135645796.0, "step": 3571 }, { "epoch": 0.4543951151253021, "ewc_loss": 0.0908203125, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.553621292114258, "learning_rate": 1e-05, "loss": 0.5471, "mean_token_accuracy": 0.8517472743988037, "num_tokens": 135686535.0, "step": 3572 }, { "epoch": 0.45452232540389265, "ewc_loss": 0.09326171875, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 6.09689474105835, "learning_rate": 1e-05, "loss": 0.531, "mean_token_accuracy": 0.8547516465187073, "num_tokens": 135719555.0, "step": 3573 }, { "epoch": 0.4546495356824831, "ewc_loss": 0.09033203125, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.607553482055664, "learning_rate": 1e-05, "loss": 0.4778, "mean_token_accuracy": 0.8667747378349304, "num_tokens": 135756498.0, "step": 3574 }, { "epoch": 0.45477674596107365, "ewc_loss": 0.08349609375, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 7.62939453125e-05, "grad_norm": 5.704226970672607, "learning_rate": 1e-05, "loss": 0.5844, "mean_token_accuracy": 0.8357263803482056, "num_tokens": 135797438.0, "step": 3575 }, { "epoch": 0.4549039562396642, "ewc_loss": 0.08740234375, "ewc_loss_diag": 7.390975952148438e-05, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.574287414550781, "learning_rate": 1e-05, "loss": 0.5419, "mean_token_accuracy": 0.851078987121582, "num_tokens": 135836085.0, "step": 3576 }, { "epoch": 0.45503116651825465, "ewc_loss": 0.080078125, "ewc_loss_diag": 7.43865966796875e-05, "ewc_loss_parallel": 7.295608520507812e-05, "grad_norm": 5.371296405792236, "learning_rate": 1e-05, "loss": 0.5092, "mean_token_accuracy": 0.8593642711639404, "num_tokens": 135872272.0, "step": 3577 }, { "epoch": 0.4551583767968452, "ewc_loss": 0.08544921875, "ewc_loss_diag": 7.43865966796875e-05, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.586834907531738, "learning_rate": 1e-05, "loss": 0.5879, "mean_token_accuracy": 0.835310697555542, "num_tokens": 135908667.0, "step": 3578 }, { "epoch": 0.4552855870754357, "ewc_loss": 0.0771484375, "ewc_loss_diag": 7.43865966796875e-05, "ewc_loss_parallel": 6.961822509765625e-05, "grad_norm": 5.164072036743164, "learning_rate": 1e-05, "loss": 0.4805, "mean_token_accuracy": 0.8648635149002075, "num_tokens": 135939205.0, "step": 3579 }, { "epoch": 0.4554127973540262, "ewc_loss": 0.08447265625, "ewc_loss_diag": 7.43865966796875e-05, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.623754978179932, "learning_rate": 1e-05, "loss": 0.5059, "mean_token_accuracy": 0.8577179312705994, "num_tokens": 135977084.0, "step": 3580 }, { "epoch": 0.4555400076326167, "ewc_loss": 0.07666015625, "ewc_loss_diag": 7.43865966796875e-05, "ewc_loss_parallel": 6.914138793945312e-05, "grad_norm": 5.014461994171143, "learning_rate": 1e-05, "loss": 0.5337, "mean_token_accuracy": 0.8467003107070923, "num_tokens": 136012712.0, "step": 3581 }, { "epoch": 0.45566721791120723, "ewc_loss": 0.08544921875, "ewc_loss_diag": 7.43865966796875e-05, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.71708869934082, "learning_rate": 1e-05, "loss": 0.5175, "mean_token_accuracy": 0.8601152300834656, "num_tokens": 136050253.0, "step": 3582 }, { "epoch": 0.45579442818979776, "ewc_loss": 0.0751953125, "ewc_loss_diag": 7.43865966796875e-05, "ewc_loss_parallel": 6.771087646484375e-05, "grad_norm": 4.903526306152344, "learning_rate": 1e-05, "loss": 0.5107, "mean_token_accuracy": 0.853472888469696, "num_tokens": 136088072.0, "step": 3583 }, { "epoch": 0.45592163846838824, "ewc_loss": 0.08984375, "ewc_loss_diag": 7.43865966796875e-05, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.934329986572266, "learning_rate": 1e-05, "loss": 0.5529, "mean_token_accuracy": 0.8499406576156616, "num_tokens": 136122434.0, "step": 3584 }, { "epoch": 0.45604884874697876, "ewc_loss": 0.07470703125, "ewc_loss_diag": 7.43865966796875e-05, "ewc_loss_parallel": 6.723403930664062e-05, "grad_norm": 4.818236827850342, "learning_rate": 1e-05, "loss": 0.4972, "mean_token_accuracy": 0.861899197101593, "num_tokens": 136164489.0, "step": 3585 }, { "epoch": 0.4561760590255693, "ewc_loss": 0.09814453125, "ewc_loss_diag": 7.43865966796875e-05, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.247001647949219, "learning_rate": 1e-05, "loss": 0.5443, "mean_token_accuracy": 0.8510538339614868, "num_tokens": 136198534.0, "step": 3586 }, { "epoch": 0.45630326930415976, "ewc_loss": 0.07763671875, "ewc_loss_diag": 7.43865966796875e-05, "ewc_loss_parallel": 7.009506225585938e-05, "grad_norm": 4.906810283660889, "learning_rate": 1e-05, "loss": 0.5121, "mean_token_accuracy": 0.8556467890739441, "num_tokens": 136242261.0, "step": 3587 }, { "epoch": 0.4564304795827503, "ewc_loss": 0.10302734375, "ewc_loss_diag": 7.43865966796875e-05, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.518584728240967, "learning_rate": 1e-05, "loss": 0.5768, "mean_token_accuracy": 0.8457551598548889, "num_tokens": 136277106.0, "step": 3588 }, { "epoch": 0.4565576898613408, "ewc_loss": 0.08544921875, "ewc_loss_diag": 7.43865966796875e-05, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.238615036010742, "learning_rate": 1e-05, "loss": 0.5097, "mean_token_accuracy": 0.8566312789916992, "num_tokens": 136315872.0, "step": 3589 }, { "epoch": 0.4566849001399313, "ewc_loss": 0.10205078125, "ewc_loss_diag": 7.43865966796875e-05, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.561923503875732, "learning_rate": 1e-05, "loss": 0.5452, "mean_token_accuracy": 0.854568362236023, "num_tokens": 136359513.0, "step": 3590 }, { "epoch": 0.4568121104185218, "ewc_loss": 0.0908203125, "ewc_loss_diag": 7.43865966796875e-05, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.562646389007568, "learning_rate": 1e-05, "loss": 0.5573, "mean_token_accuracy": 0.8487412333488464, "num_tokens": 136398575.0, "step": 3591 }, { "epoch": 0.45693932069711235, "ewc_loss": 0.0966796875, "ewc_loss_diag": 7.43865966796875e-05, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 6.168315410614014, "learning_rate": 1e-05, "loss": 0.4969, "mean_token_accuracy": 0.8652358055114746, "num_tokens": 136439062.0, "step": 3592 }, { "epoch": 0.4570665309757028, "ewc_loss": 0.09033203125, "ewc_loss_diag": 7.43865966796875e-05, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.613068580627441, "learning_rate": 1e-05, "loss": 0.5333, "mean_token_accuracy": 0.853306770324707, "num_tokens": 136475044.0, "step": 3593 }, { "epoch": 0.45719374125429335, "ewc_loss": 0.0849609375, "ewc_loss_diag": 7.43865966796875e-05, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.801461696624756, "learning_rate": 1e-05, "loss": 0.5635, "mean_token_accuracy": 0.8422250151634216, "num_tokens": 136513674.0, "step": 3594 }, { "epoch": 0.4573209515328839, "ewc_loss": 0.08740234375, "ewc_loss_diag": 7.43865966796875e-05, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.5689849853515625, "learning_rate": 1e-05, "loss": 0.5125, "mean_token_accuracy": 0.858960747718811, "num_tokens": 136551218.0, "step": 3595 }, { "epoch": 0.45744816181147435, "ewc_loss": 0.08056640625, "ewc_loss_diag": 7.43865966796875e-05, "ewc_loss_parallel": 7.343292236328125e-05, "grad_norm": 5.476750373840332, "learning_rate": 1e-05, "loss": 0.4723, "mean_token_accuracy": 0.8676515221595764, "num_tokens": 136589088.0, "step": 3596 }, { "epoch": 0.4575753720900649, "ewc_loss": 0.0859375, "ewc_loss_diag": 7.43865966796875e-05, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.539629936218262, "learning_rate": 1e-05, "loss": 0.5047, "mean_token_accuracy": 0.8574239015579224, "num_tokens": 136625269.0, "step": 3597 }, { "epoch": 0.4577025823686554, "ewc_loss": 0.078125, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 7.05718994140625e-05, "grad_norm": 5.271255970001221, "learning_rate": 1e-05, "loss": 0.5032, "mean_token_accuracy": 0.8592171669006348, "num_tokens": 136660212.0, "step": 3598 }, { "epoch": 0.4578297926472459, "ewc_loss": 0.08349609375, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 7.62939453125e-05, "grad_norm": 5.513913154602051, "learning_rate": 1e-05, "loss": 0.5237, "mean_token_accuracy": 0.8482873439788818, "num_tokens": 136703372.0, "step": 3599 }, { "epoch": 0.4579570029258364, "ewc_loss": 0.0771484375, "ewc_loss_diag": 7.534027099609375e-05, "ewc_loss_parallel": 6.961822509765625e-05, "grad_norm": 5.131384372711182, "learning_rate": 1e-05, "loss": 0.5338, "mean_token_accuracy": 0.8496742248535156, "num_tokens": 136741241.0, "step": 3600 }, { "epoch": 0.45808421320442694, "ewc_loss": 0.08349609375, "ewc_loss_diag": 7.534027099609375e-05, "ewc_loss_parallel": 7.62939453125e-05, "grad_norm": 5.544212818145752, "learning_rate": 1e-05, "loss": 0.5163, "mean_token_accuracy": 0.8542354106903076, "num_tokens": 136780329.0, "step": 3601 }, { "epoch": 0.4582114234830174, "ewc_loss": 0.076171875, "ewc_loss_diag": 7.534027099609375e-05, "ewc_loss_parallel": 6.866455078125e-05, "grad_norm": 5.033981800079346, "learning_rate": 1e-05, "loss": 0.5635, "mean_token_accuracy": 0.8386658430099487, "num_tokens": 136812567.0, "step": 3602 }, { "epoch": 0.45833863376160794, "ewc_loss": 0.083984375, "ewc_loss_diag": 7.534027099609375e-05, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.636264801025391, "learning_rate": 1e-05, "loss": 0.544, "mean_token_accuracy": 0.8456646203994751, "num_tokens": 136854370.0, "step": 3603 }, { "epoch": 0.45846584404019847, "ewc_loss": 0.0751953125, "ewc_loss_diag": 7.534027099609375e-05, "ewc_loss_parallel": 6.771087646484375e-05, "grad_norm": 4.914168357849121, "learning_rate": 1e-05, "loss": 0.5285, "mean_token_accuracy": 0.8526376485824585, "num_tokens": 136890536.0, "step": 3604 }, { "epoch": 0.45859305431878894, "ewc_loss": 0.087890625, "ewc_loss_diag": 7.534027099609375e-05, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.8442702293396, "learning_rate": 1e-05, "loss": 0.6145, "mean_token_accuracy": 0.8285492658615112, "num_tokens": 136929376.0, "step": 3605 }, { "epoch": 0.45872026459737947, "ewc_loss": 0.07421875, "ewc_loss_diag": 7.534027099609375e-05, "ewc_loss_parallel": 6.67572021484375e-05, "grad_norm": 4.786382675170898, "learning_rate": 1e-05, "loss": 0.4893, "mean_token_accuracy": 0.8599683046340942, "num_tokens": 136967857.0, "step": 3606 }, { "epoch": 0.45884747487597, "ewc_loss": 0.095703125, "ewc_loss_diag": 7.534027099609375e-05, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 6.160201549530029, "learning_rate": 1e-05, "loss": 0.5384, "mean_token_accuracy": 0.8541906476020813, "num_tokens": 137007853.0, "step": 3607 }, { "epoch": 0.45897468515456047, "ewc_loss": 0.0751953125, "ewc_loss_diag": 7.534027099609375e-05, "ewc_loss_parallel": 6.771087646484375e-05, "grad_norm": 4.782360553741455, "learning_rate": 1e-05, "loss": 0.5114, "mean_token_accuracy": 0.8566997051239014, "num_tokens": 137048172.0, "step": 3608 }, { "epoch": 0.459101895433151, "ewc_loss": 0.1044921875, "ewc_loss_diag": 7.534027099609375e-05, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.570911407470703, "learning_rate": 1e-05, "loss": 0.5965, "mean_token_accuracy": 0.8373351097106934, "num_tokens": 137083598.0, "step": 3609 }, { "epoch": 0.4592291057117415, "ewc_loss": 0.083984375, "ewc_loss_diag": 7.534027099609375e-05, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.114370346069336, "learning_rate": 1e-05, "loss": 0.5455, "mean_token_accuracy": 0.8439608812332153, "num_tokens": 137117626.0, "step": 3610 }, { "epoch": 0.459356315990332, "ewc_loss": 0.10791015625, "ewc_loss_diag": 7.534027099609375e-05, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.696882724761963, "learning_rate": 1e-05, "loss": 0.5802, "mean_token_accuracy": 0.8435215950012207, "num_tokens": 137161536.0, "step": 3611 }, { "epoch": 0.4594835262689225, "ewc_loss": 0.09423828125, "ewc_loss_diag": 7.534027099609375e-05, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.570353984832764, "learning_rate": 1e-05, "loss": 0.4998, "mean_token_accuracy": 0.8622446060180664, "num_tokens": 137195932.0, "step": 3612 }, { "epoch": 0.45961073654751305, "ewc_loss": 0.10205078125, "ewc_loss_diag": 7.534027099609375e-05, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.397573947906494, "learning_rate": 1e-05, "loss": 0.5125, "mean_token_accuracy": 0.8635247945785522, "num_tokens": 137229263.0, "step": 3613 }, { "epoch": 0.4597379468261035, "ewc_loss": 0.09521484375, "ewc_loss_diag": 7.534027099609375e-05, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 5.719622611999512, "learning_rate": 1e-05, "loss": 0.5352, "mean_token_accuracy": 0.8552889823913574, "num_tokens": 137265250.0, "step": 3614 }, { "epoch": 0.45986515710469406, "ewc_loss": 0.08935546875, "ewc_loss_diag": 7.534027099609375e-05, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.975114822387695, "learning_rate": 1e-05, "loss": 0.4776, "mean_token_accuracy": 0.8656525611877441, "num_tokens": 137307612.0, "step": 3615 }, { "epoch": 0.4599923673832846, "ewc_loss": 0.09130859375, "ewc_loss_diag": 7.534027099609375e-05, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.665207386016846, "learning_rate": 1e-05, "loss": 0.477, "mean_token_accuracy": 0.8675527572631836, "num_tokens": 137336075.0, "step": 3616 }, { "epoch": 0.46011957766187506, "ewc_loss": 0.08154296875, "ewc_loss_diag": 7.534027099609375e-05, "ewc_loss_parallel": 7.43865966796875e-05, "grad_norm": 5.609661102294922, "learning_rate": 1e-05, "loss": 0.5276, "mean_token_accuracy": 0.8536476492881775, "num_tokens": 137374571.0, "step": 3617 }, { "epoch": 0.4602467879404656, "ewc_loss": 0.08740234375, "ewc_loss_diag": 7.534027099609375e-05, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.598996162414551, "learning_rate": 1e-05, "loss": 0.5454, "mean_token_accuracy": 0.8471469879150391, "num_tokens": 137409275.0, "step": 3618 }, { "epoch": 0.4603739982190561, "ewc_loss": 0.07861328125, "ewc_loss_diag": 7.534027099609375e-05, "ewc_loss_parallel": 7.152557373046875e-05, "grad_norm": 5.345955848693848, "learning_rate": 1e-05, "loss": 0.5087, "mean_token_accuracy": 0.8558934926986694, "num_tokens": 137447665.0, "step": 3619 }, { "epoch": 0.4605012084976466, "ewc_loss": 0.083984375, "ewc_loss_diag": 7.534027099609375e-05, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.566741466522217, "learning_rate": 1e-05, "loss": 0.5075, "mean_token_accuracy": 0.8639276027679443, "num_tokens": 137484256.0, "step": 3620 }, { "epoch": 0.4606284187762371, "ewc_loss": 0.07763671875, "ewc_loss_diag": 7.534027099609375e-05, "ewc_loss_parallel": 7.009506225585938e-05, "grad_norm": 5.177803039550781, "learning_rate": 1e-05, "loss": 0.5037, "mean_token_accuracy": 0.8575479984283447, "num_tokens": 137523868.0, "step": 3621 }, { "epoch": 0.46075562905482764, "ewc_loss": 0.083984375, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.582086086273193, "learning_rate": 1e-05, "loss": 0.5919, "mean_token_accuracy": 0.8294200897216797, "num_tokens": 137564707.0, "step": 3622 }, { "epoch": 0.4608828393334181, "ewc_loss": 0.0771484375, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 6.914138793945312e-05, "grad_norm": 5.052923202514648, "learning_rate": 1e-05, "loss": 0.4815, "mean_token_accuracy": 0.86163330078125, "num_tokens": 137600347.0, "step": 3623 }, { "epoch": 0.46101004961200864, "ewc_loss": 0.083984375, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.669981956481934, "learning_rate": 1e-05, "loss": 0.5433, "mean_token_accuracy": 0.849394679069519, "num_tokens": 137642874.0, "step": 3624 }, { "epoch": 0.46113725989059917, "ewc_loss": 0.076171875, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 6.818771362304688e-05, "grad_norm": 4.927661895751953, "learning_rate": 1e-05, "loss": 0.5616, "mean_token_accuracy": 0.8476472496986389, "num_tokens": 137685253.0, "step": 3625 }, { "epoch": 0.46126447016918964, "ewc_loss": 0.087890625, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.8549418449401855, "learning_rate": 1e-05, "loss": 0.4624, "mean_token_accuracy": 0.8707289695739746, "num_tokens": 137722180.0, "step": 3626 }, { "epoch": 0.4613916804477802, "ewc_loss": 0.0751953125, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 6.723403930664062e-05, "grad_norm": 4.83318567276001, "learning_rate": 1e-05, "loss": 0.4962, "mean_token_accuracy": 0.860489010810852, "num_tokens": 137756665.0, "step": 3627 }, { "epoch": 0.4615188907263707, "ewc_loss": 0.0966796875, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.149683952331543, "learning_rate": 1e-05, "loss": 0.5153, "mean_token_accuracy": 0.8616938591003418, "num_tokens": 137795928.0, "step": 3628 }, { "epoch": 0.4616461010049612, "ewc_loss": 0.076171875, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 6.866455078125e-05, "grad_norm": 4.839423656463623, "learning_rate": 1e-05, "loss": 0.5469, "mean_token_accuracy": 0.8411768078804016, "num_tokens": 137833695.0, "step": 3629 }, { "epoch": 0.4617733112835517, "ewc_loss": 0.103515625, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.489070415496826, "learning_rate": 1e-05, "loss": 0.5223, "mean_token_accuracy": 0.8586044311523438, "num_tokens": 137871894.0, "step": 3630 }, { "epoch": 0.46190052156214223, "ewc_loss": 0.08203125, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 7.486343383789062e-05, "grad_norm": 5.103594779968262, "learning_rate": 1e-05, "loss": 0.5308, "mean_token_accuracy": 0.848054051399231, "num_tokens": 137907818.0, "step": 3631 }, { "epoch": 0.46202773184073276, "ewc_loss": 0.1064453125, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.6062846183776855, "learning_rate": 1e-05, "loss": 0.4974, "mean_token_accuracy": 0.8676210641860962, "num_tokens": 137943506.0, "step": 3632 }, { "epoch": 0.46215494211932323, "ewc_loss": 0.08984375, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.510277271270752, "learning_rate": 1e-05, "loss": 0.5716, "mean_token_accuracy": 0.847488522529602, "num_tokens": 137975113.0, "step": 3633 }, { "epoch": 0.46228215239791376, "ewc_loss": 0.1005859375, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.343321323394775, "learning_rate": 1e-05, "loss": 0.5326, "mean_token_accuracy": 0.8611595034599304, "num_tokens": 138006503.0, "step": 3634 }, { "epoch": 0.4624093626765043, "ewc_loss": 0.0927734375, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.649659633636475, "learning_rate": 1e-05, "loss": 0.495, "mean_token_accuracy": 0.8643621206283569, "num_tokens": 138046667.0, "step": 3635 }, { "epoch": 0.46253657295509476, "ewc_loss": 0.0888671875, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.940891742706299, "learning_rate": 1e-05, "loss": 0.5694, "mean_token_accuracy": 0.8428477048873901, "num_tokens": 138084821.0, "step": 3636 }, { "epoch": 0.4626637832336853, "ewc_loss": 0.08984375, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.628554821014404, "learning_rate": 1e-05, "loss": 0.5259, "mean_token_accuracy": 0.8529438376426697, "num_tokens": 138123994.0, "step": 3637 }, { "epoch": 0.4627909935122758, "ewc_loss": 0.0810546875, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 7.390975952148438e-05, "grad_norm": 5.572746276855469, "learning_rate": 1e-05, "loss": 0.5346, "mean_token_accuracy": 0.8504129648208618, "num_tokens": 138161225.0, "step": 3638 }, { "epoch": 0.4629182037908663, "ewc_loss": 0.087890625, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.589925765991211, "learning_rate": 1e-05, "loss": 0.4853, "mean_token_accuracy": 0.862343430519104, "num_tokens": 138197441.0, "step": 3639 }, { "epoch": 0.4630454140694568, "ewc_loss": 0.0791015625, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 7.152557373046875e-05, "grad_norm": 5.310432434082031, "learning_rate": 1e-05, "loss": 0.5125, "mean_token_accuracy": 0.8593037128448486, "num_tokens": 138229663.0, "step": 3640 }, { "epoch": 0.46317262434804735, "ewc_loss": 0.083984375, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 7.62939453125e-05, "grad_norm": 5.5792388916015625, "learning_rate": 1e-05, "loss": 0.5377, "mean_token_accuracy": 0.8520835638046265, "num_tokens": 138267955.0, "step": 3641 }, { "epoch": 0.4632998346266378, "ewc_loss": 0.078125, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 7.009506225585938e-05, "grad_norm": 5.141353130340576, "learning_rate": 1e-05, "loss": 0.4774, "mean_token_accuracy": 0.8661412000656128, "num_tokens": 138303968.0, "step": 3642 }, { "epoch": 0.46342704490522835, "ewc_loss": 0.08447265625, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.62149715423584, "learning_rate": 1e-05, "loss": 0.5235, "mean_token_accuracy": 0.8566488027572632, "num_tokens": 138344667.0, "step": 3643 }, { "epoch": 0.4635542551838189, "ewc_loss": 0.07666015625, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 6.866455078125e-05, "grad_norm": 4.99338960647583, "learning_rate": 1e-05, "loss": 0.4624, "mean_token_accuracy": 0.8663597106933594, "num_tokens": 138387513.0, "step": 3644 }, { "epoch": 0.46368146546240935, "ewc_loss": 0.0869140625, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.764518737792969, "learning_rate": 1e-05, "loss": 0.4612, "mean_token_accuracy": 0.8728048205375671, "num_tokens": 138428070.0, "step": 3645 }, { "epoch": 0.4638086757409999, "ewc_loss": 0.07568359375, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 6.771087646484375e-05, "grad_norm": 4.877016067504883, "learning_rate": 1e-05, "loss": 0.4966, "mean_token_accuracy": 0.8577209711074829, "num_tokens": 138462303.0, "step": 3646 }, { "epoch": 0.4639358860195904, "ewc_loss": 0.091796875, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 6.0106706619262695, "learning_rate": 1e-05, "loss": 0.6006, "mean_token_accuracy": 0.8352169990539551, "num_tokens": 138502099.0, "step": 3647 }, { "epoch": 0.4640630962981809, "ewc_loss": 0.0751953125, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 6.723403930664062e-05, "grad_norm": 4.8388261795043945, "learning_rate": 1e-05, "loss": 0.5278, "mean_token_accuracy": 0.848438024520874, "num_tokens": 138539400.0, "step": 3648 }, { "epoch": 0.4641903065767714, "ewc_loss": 0.099609375, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.319037437438965, "learning_rate": 1e-05, "loss": 0.5415, "mean_token_accuracy": 0.8561081886291504, "num_tokens": 138573082.0, "step": 3649 }, { "epoch": 0.46431751685536193, "ewc_loss": 0.0791015625, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 7.104873657226562e-05, "grad_norm": 4.950282096862793, "learning_rate": 1e-05, "loss": 0.4865, "mean_token_accuracy": 0.8597549796104431, "num_tokens": 138613693.0, "step": 3650 }, { "epoch": 0.4644447271339524, "ewc_loss": 0.10498046875, "ewc_loss_diag": 7.724761962890625e-05, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.557400226593018, "learning_rate": 1e-05, "loss": 0.492, "mean_token_accuracy": 0.8685964941978455, "num_tokens": 138655138.0, "step": 3651 }, { "epoch": 0.46457193741254293, "ewc_loss": 0.08642578125, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.310653209686279, "learning_rate": 1e-05, "loss": 0.6201, "mean_token_accuracy": 0.8236976861953735, "num_tokens": 138697117.0, "step": 3652 }, { "epoch": 0.46469914769113346, "ewc_loss": 0.10302734375, "ewc_loss_diag": 7.724761962890625e-05, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.468557357788086, "learning_rate": 1e-05, "loss": 0.553, "mean_token_accuracy": 0.8526939153671265, "num_tokens": 138732723.0, "step": 3653 }, { "epoch": 0.46482635796972394, "ewc_loss": 0.09228515625, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.583200931549072, "learning_rate": 1e-05, "loss": 0.5073, "mean_token_accuracy": 0.8568578958511353, "num_tokens": 138767680.0, "step": 3654 }, { "epoch": 0.46495356824831446, "ewc_loss": 0.09521484375, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 6.119918346405029, "learning_rate": 1e-05, "loss": 0.5101, "mean_token_accuracy": 0.8622772693634033, "num_tokens": 138803119.0, "step": 3655 }, { "epoch": 0.465080778526905, "ewc_loss": 0.09130859375, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.618002891540527, "learning_rate": 1e-05, "loss": 0.4989, "mean_token_accuracy": 0.8637045621871948, "num_tokens": 138845032.0, "step": 3656 }, { "epoch": 0.46520798880549546, "ewc_loss": 0.08544921875, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.728273391723633, "learning_rate": 1e-05, "loss": 0.5049, "mean_token_accuracy": 0.8611104488372803, "num_tokens": 138886521.0, "step": 3657 }, { "epoch": 0.465335199084086, "ewc_loss": 0.08837890625, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.583011627197266, "learning_rate": 1e-05, "loss": 0.5145, "mean_token_accuracy": 0.8552792072296143, "num_tokens": 138927586.0, "step": 3658 }, { "epoch": 0.4654624093626765, "ewc_loss": 0.080078125, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 7.2479248046875e-05, "grad_norm": 5.43091344833374, "learning_rate": 1e-05, "loss": 0.5244, "mean_token_accuracy": 0.8507258296012878, "num_tokens": 138967382.0, "step": 3659 }, { "epoch": 0.465589619641267, "ewc_loss": 0.08642578125, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.569413661956787, "learning_rate": 1e-05, "loss": 0.5643, "mean_token_accuracy": 0.8396217226982117, "num_tokens": 139003498.0, "step": 3660 }, { "epoch": 0.4657168299198575, "ewc_loss": 0.078125, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 7.009506225585938e-05, "grad_norm": 5.214148044586182, "learning_rate": 1e-05, "loss": 0.5416, "mean_token_accuracy": 0.8448054194450378, "num_tokens": 139046724.0, "step": 3661 }, { "epoch": 0.46584404019844805, "ewc_loss": 0.083984375, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 7.62939453125e-05, "grad_norm": 5.574219703674316, "learning_rate": 1e-05, "loss": 0.5247, "mean_token_accuracy": 0.8538559675216675, "num_tokens": 139091513.0, "step": 3662 }, { "epoch": 0.4659712504770385, "ewc_loss": 0.0771484375, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 6.914138793945312e-05, "grad_norm": 5.088807582855225, "learning_rate": 1e-05, "loss": 0.5862, "mean_token_accuracy": 0.8328707218170166, "num_tokens": 139130082.0, "step": 3663 }, { "epoch": 0.46609846075562905, "ewc_loss": 0.08544921875, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.635558128356934, "learning_rate": 1e-05, "loss": 0.5109, "mean_token_accuracy": 0.8572129011154175, "num_tokens": 139172974.0, "step": 3664 }, { "epoch": 0.4662256710342196, "ewc_loss": 0.0751953125, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 6.723403930664062e-05, "grad_norm": 4.972537517547607, "learning_rate": 1e-05, "loss": 0.4819, "mean_token_accuracy": 0.8627830743789673, "num_tokens": 139212164.0, "step": 3665 }, { "epoch": 0.46635288131281005, "ewc_loss": 0.0869140625, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.790219783782959, "learning_rate": 1e-05, "loss": 0.6088, "mean_token_accuracy": 0.8272382616996765, "num_tokens": 139252960.0, "step": 3666 }, { "epoch": 0.4664800915914006, "ewc_loss": 0.0751953125, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 6.723403930664062e-05, "grad_norm": 4.864133358001709, "learning_rate": 1e-05, "loss": 0.5504, "mean_token_accuracy": 0.8468070030212402, "num_tokens": 139287078.0, "step": 3667 }, { "epoch": 0.4666073018699911, "ewc_loss": 0.09326171875, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 6.053834915161133, "learning_rate": 1e-05, "loss": 0.5944, "mean_token_accuracy": 0.8445490598678589, "num_tokens": 139321545.0, "step": 3668 }, { "epoch": 0.4667345121485816, "ewc_loss": 0.07568359375, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 6.771087646484375e-05, "grad_norm": 4.818379878997803, "learning_rate": 1e-05, "loss": 0.5649, "mean_token_accuracy": 0.8373927474021912, "num_tokens": 139360829.0, "step": 3669 }, { "epoch": 0.4668617224271721, "ewc_loss": 0.1005859375, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.388262748718262, "learning_rate": 1e-05, "loss": 0.5486, "mean_token_accuracy": 0.8558862805366516, "num_tokens": 139396347.0, "step": 3670 }, { "epoch": 0.46698893270576264, "ewc_loss": 0.080078125, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 7.2479248046875e-05, "grad_norm": 5.014863967895508, "learning_rate": 1e-05, "loss": 0.5162, "mean_token_accuracy": 0.8575571775436401, "num_tokens": 139430367.0, "step": 3671 }, { "epoch": 0.4671161429843531, "ewc_loss": 0.10546875, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.576477527618408, "learning_rate": 1e-05, "loss": 0.5538, "mean_token_accuracy": 0.8465224504470825, "num_tokens": 139468233.0, "step": 3672 }, { "epoch": 0.46724335326294364, "ewc_loss": 0.08935546875, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.362965106964111, "learning_rate": 1e-05, "loss": 0.532, "mean_token_accuracy": 0.8514171838760376, "num_tokens": 139502377.0, "step": 3673 }, { "epoch": 0.46737056354153417, "ewc_loss": 0.1025390625, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.423039436340332, "learning_rate": 1e-05, "loss": 0.566, "mean_token_accuracy": 0.8460364937782288, "num_tokens": 139540503.0, "step": 3674 }, { "epoch": 0.46749777382012464, "ewc_loss": 0.09228515625, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.6029157638549805, "learning_rate": 1e-05, "loss": 0.562, "mean_token_accuracy": 0.8421850204467773, "num_tokens": 139578063.0, "step": 3675 }, { "epoch": 0.46762498409871517, "ewc_loss": 0.09130859375, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 6.045193672180176, "learning_rate": 1e-05, "loss": 0.4696, "mean_token_accuracy": 0.8697957992553711, "num_tokens": 139615451.0, "step": 3676 }, { "epoch": 0.4677521943773057, "ewc_loss": 0.0908203125, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.616279125213623, "learning_rate": 1e-05, "loss": 0.5724, "mean_token_accuracy": 0.8446030616760254, "num_tokens": 139657779.0, "step": 3677 }, { "epoch": 0.46787940465589617, "ewc_loss": 0.0830078125, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 7.534027099609375e-05, "grad_norm": 5.66279935836792, "learning_rate": 1e-05, "loss": 0.5156, "mean_token_accuracy": 0.8565267324447632, "num_tokens": 139698166.0, "step": 3678 }, { "epoch": 0.4680066149344867, "ewc_loss": 0.087890625, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.588229179382324, "learning_rate": 1e-05, "loss": 0.5786, "mean_token_accuracy": 0.8388659358024597, "num_tokens": 139736970.0, "step": 3679 }, { "epoch": 0.4681338252130772, "ewc_loss": 0.0791015625, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 7.152557373046875e-05, "grad_norm": 5.364530086517334, "learning_rate": 1e-05, "loss": 0.5087, "mean_token_accuracy": 0.854878306388855, "num_tokens": 139771570.0, "step": 3680 }, { "epoch": 0.46826103549166775, "ewc_loss": 0.0859375, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.585675239562988, "learning_rate": 1e-05, "loss": 0.5997, "mean_token_accuracy": 0.8324607610702515, "num_tokens": 139811927.0, "step": 3681 }, { "epoch": 0.4683882457702582, "ewc_loss": 0.07861328125, "ewc_loss_diag": 7.724761962890625e-05, "ewc_loss_parallel": 7.05718994140625e-05, "grad_norm": 5.151991844177246, "learning_rate": 1e-05, "loss": 0.5477, "mean_token_accuracy": 0.8424839973449707, "num_tokens": 139852903.0, "step": 3682 }, { "epoch": 0.46851545604884876, "ewc_loss": 0.0859375, "ewc_loss_diag": 7.724761962890625e-05, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.6399312019348145, "learning_rate": 1e-05, "loss": 0.5528, "mean_token_accuracy": 0.8432648181915283, "num_tokens": 139895375.0, "step": 3683 }, { "epoch": 0.4686426663274393, "ewc_loss": 0.07666015625, "ewc_loss_diag": 7.724761962890625e-05, "ewc_loss_parallel": 6.866455078125e-05, "grad_norm": 5.088831424713135, "learning_rate": 1e-05, "loss": 0.6118, "mean_token_accuracy": 0.8314241170883179, "num_tokens": 139927878.0, "step": 3684 }, { "epoch": 0.46876987660602976, "ewc_loss": 0.08544921875, "ewc_loss_diag": 7.724761962890625e-05, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.681990623474121, "learning_rate": 1e-05, "loss": 0.5196, "mean_token_accuracy": 0.8517546653747559, "num_tokens": 139959864.0, "step": 3685 }, { "epoch": 0.4688970868846203, "ewc_loss": 0.07666015625, "ewc_loss_diag": 7.724761962890625e-05, "ewc_loss_parallel": 6.866455078125e-05, "grad_norm": 4.927334785461426, "learning_rate": 1e-05, "loss": 0.5582, "mean_token_accuracy": 0.8418847322463989, "num_tokens": 140003815.0, "step": 3686 }, { "epoch": 0.4690242971632108, "ewc_loss": 0.08935546875, "ewc_loss_diag": 7.724761962890625e-05, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.851569175720215, "learning_rate": 1e-05, "loss": 0.5761, "mean_token_accuracy": 0.8388416767120361, "num_tokens": 140043126.0, "step": 3687 }, { "epoch": 0.4691515074418013, "ewc_loss": 0.07568359375, "ewc_loss_diag": 7.724761962890625e-05, "ewc_loss_parallel": 6.771087646484375e-05, "grad_norm": 4.834529876708984, "learning_rate": 1e-05, "loss": 0.5144, "mean_token_accuracy": 0.8505088686943054, "num_tokens": 140078466.0, "step": 3688 }, { "epoch": 0.4692787177203918, "ewc_loss": 0.0986328125, "ewc_loss_diag": 7.724761962890625e-05, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.15175724029541, "learning_rate": 1e-05, "loss": 0.5636, "mean_token_accuracy": 0.8430343866348267, "num_tokens": 140114733.0, "step": 3689 }, { "epoch": 0.46940592799898234, "ewc_loss": 0.07763671875, "ewc_loss_diag": 7.724761962890625e-05, "ewc_loss_parallel": 6.961822509765625e-05, "grad_norm": 4.840826511383057, "learning_rate": 1e-05, "loss": 0.534, "mean_token_accuracy": 0.8508838415145874, "num_tokens": 140157017.0, "step": 3690 }, { "epoch": 0.4695331382775728, "ewc_loss": 0.10400390625, "ewc_loss_diag": 7.724761962890625e-05, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.483031749725342, "learning_rate": 1e-05, "loss": 0.5597, "mean_token_accuracy": 0.8483490943908691, "num_tokens": 140197408.0, "step": 3691 }, { "epoch": 0.46966034855616334, "ewc_loss": 0.0849609375, "ewc_loss_diag": 7.724761962890625e-05, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.120151519775391, "learning_rate": 1e-05, "loss": 0.5608, "mean_token_accuracy": 0.8459289073944092, "num_tokens": 140232458.0, "step": 3692 }, { "epoch": 0.46978755883475387, "ewc_loss": 0.10693359375, "ewc_loss_diag": 7.724761962890625e-05, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.579850673675537, "learning_rate": 1e-05, "loss": 0.5632, "mean_token_accuracy": 0.8506160974502563, "num_tokens": 140273790.0, "step": 3693 }, { "epoch": 0.46991476911334434, "ewc_loss": 0.09228515625, "ewc_loss_diag": 7.724761962890625e-05, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.476508140563965, "learning_rate": 1e-05, "loss": 0.4813, "mean_token_accuracy": 0.8694146871566772, "num_tokens": 140313339.0, "step": 3694 }, { "epoch": 0.47004197939193487, "ewc_loss": 0.10205078125, "ewc_loss_diag": 7.724761962890625e-05, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.326607704162598, "learning_rate": 1e-05, "loss": 0.5508, "mean_token_accuracy": 0.8510751724243164, "num_tokens": 140352572.0, "step": 3695 }, { "epoch": 0.4701691896705254, "ewc_loss": 0.09375, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.646890640258789, "learning_rate": 1e-05, "loss": 0.5672, "mean_token_accuracy": 0.8451740145683289, "num_tokens": 140385221.0, "step": 3696 }, { "epoch": 0.4702963999491159, "ewc_loss": 0.08984375, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.925743103027344, "learning_rate": 1e-05, "loss": 0.4785, "mean_token_accuracy": 0.8691180944442749, "num_tokens": 140427041.0, "step": 3697 }, { "epoch": 0.4704236102277064, "ewc_loss": 0.09033203125, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.616703510284424, "learning_rate": 1e-05, "loss": 0.4777, "mean_token_accuracy": 0.8692747354507446, "num_tokens": 140460956.0, "step": 3698 }, { "epoch": 0.47055082050629693, "ewc_loss": 0.08251953125, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 7.486343383789062e-05, "grad_norm": 5.567920207977295, "learning_rate": 1e-05, "loss": 0.4848, "mean_token_accuracy": 0.8680310249328613, "num_tokens": 140497644.0, "step": 3699 }, { "epoch": 0.4706780307848874, "ewc_loss": 0.087890625, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.578366279602051, "learning_rate": 1e-05, "loss": 0.5173, "mean_token_accuracy": 0.8543218374252319, "num_tokens": 140532764.0, "step": 3700 }, { "epoch": 0.47080524106347793, "ewc_loss": 0.0791015625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 7.152557373046875e-05, "grad_norm": 5.321709632873535, "learning_rate": 1e-05, "loss": 0.5172, "mean_token_accuracy": 0.8524439334869385, "num_tokens": 140564705.0, "step": 3701 }, { "epoch": 0.47093245134206846, "ewc_loss": 0.0859375, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.5645294189453125, "learning_rate": 1e-05, "loss": 0.5329, "mean_token_accuracy": 0.8510191440582275, "num_tokens": 140606959.0, "step": 3702 }, { "epoch": 0.47105966162065893, "ewc_loss": 0.078125, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 7.009506225585938e-05, "grad_norm": 5.152813911437988, "learning_rate": 1e-05, "loss": 0.4875, "mean_token_accuracy": 0.8609861135482788, "num_tokens": 140646034.0, "step": 3703 }, { "epoch": 0.47118687189924946, "ewc_loss": 0.0849609375, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.605731964111328, "learning_rate": 1e-05, "loss": 0.5163, "mean_token_accuracy": 0.851847767829895, "num_tokens": 140679948.0, "step": 3704 }, { "epoch": 0.47131408217784, "ewc_loss": 0.0771484375, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 6.914138793945312e-05, "grad_norm": 5.025796890258789, "learning_rate": 1e-05, "loss": 0.5436, "mean_token_accuracy": 0.8452938199043274, "num_tokens": 140713071.0, "step": 3705 }, { "epoch": 0.47144129245643046, "ewc_loss": 0.08642578125, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.71552848815918, "learning_rate": 1e-05, "loss": 0.4977, "mean_token_accuracy": 0.8618097305297852, "num_tokens": 140744243.0, "step": 3706 }, { "epoch": 0.471568502735021, "ewc_loss": 0.07568359375, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 6.771087646484375e-05, "grad_norm": 4.906271934509277, "learning_rate": 1e-05, "loss": 0.4477, "mean_token_accuracy": 0.8718980550765991, "num_tokens": 140776716.0, "step": 3707 }, { "epoch": 0.4716957130136115, "ewc_loss": 0.09033203125, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.933485984802246, "learning_rate": 1e-05, "loss": 0.4694, "mean_token_accuracy": 0.8735671639442444, "num_tokens": 140808874.0, "step": 3708 }, { "epoch": 0.471822923292202, "ewc_loss": 0.0751953125, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 6.723403930664062e-05, "grad_norm": 4.810941696166992, "learning_rate": 1e-05, "loss": 0.4862, "mean_token_accuracy": 0.8647346496582031, "num_tokens": 140846198.0, "step": 3709 }, { "epoch": 0.4719501335707925, "ewc_loss": 0.099609375, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.2599897384643555, "learning_rate": 1e-05, "loss": 0.5256, "mean_token_accuracy": 0.8564015626907349, "num_tokens": 140889228.0, "step": 3710 }, { "epoch": 0.47207734384938305, "ewc_loss": 0.07763671875, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 6.961822509765625e-05, "grad_norm": 4.902346134185791, "learning_rate": 1e-05, "loss": 0.4924, "mean_token_accuracy": 0.8569048047065735, "num_tokens": 140924549.0, "step": 3711 }, { "epoch": 0.4722045541279735, "ewc_loss": 0.10546875, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.555747032165527, "learning_rate": 1e-05, "loss": 0.5095, "mean_token_accuracy": 0.8654893636703491, "num_tokens": 140960995.0, "step": 3712 }, { "epoch": 0.47233176440656405, "ewc_loss": 0.08642578125, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.227880954742432, "learning_rate": 1e-05, "loss": 0.5267, "mean_token_accuracy": 0.8556135296821594, "num_tokens": 141000899.0, "step": 3713 }, { "epoch": 0.4724589746851546, "ewc_loss": 0.10498046875, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.5604448318481445, "learning_rate": 1e-05, "loss": 0.5116, "mean_token_accuracy": 0.8622928857803345, "num_tokens": 141037736.0, "step": 3714 }, { "epoch": 0.47258618496374505, "ewc_loss": 0.0927734375, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.575727462768555, "learning_rate": 1e-05, "loss": 0.5806, "mean_token_accuracy": 0.8378773927688599, "num_tokens": 141068916.0, "step": 3715 }, { "epoch": 0.4727133952423356, "ewc_loss": 0.0986328125, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.240095615386963, "learning_rate": 1e-05, "loss": 0.5876, "mean_token_accuracy": 0.8368861675262451, "num_tokens": 141102353.0, "step": 3716 }, { "epoch": 0.4728406055209261, "ewc_loss": 0.09228515625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.638905048370361, "learning_rate": 1e-05, "loss": 0.4974, "mean_token_accuracy": 0.8644553422927856, "num_tokens": 141142653.0, "step": 3717 }, { "epoch": 0.4729678157995166, "ewc_loss": 0.0869140625, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.846284866333008, "learning_rate": 1e-05, "loss": 0.5196, "mean_token_accuracy": 0.8582990169525146, "num_tokens": 141180091.0, "step": 3718 }, { "epoch": 0.4730950260781071, "ewc_loss": 0.08837890625, "ewc_loss_diag": 7.915496826171875e-05, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.607991695404053, "learning_rate": 1e-05, "loss": 0.4847, "mean_token_accuracy": 0.8716595768928528, "num_tokens": 141218674.0, "step": 3719 }, { "epoch": 0.47322223635669763, "ewc_loss": 0.08154296875, "ewc_loss_diag": 7.915496826171875e-05, "ewc_loss_parallel": 7.390975952148438e-05, "grad_norm": 5.504916191101074, "learning_rate": 1e-05, "loss": 0.4743, "mean_token_accuracy": 0.8657695651054382, "num_tokens": 141258680.0, "step": 3720 }, { "epoch": 0.4733494466352881, "ewc_loss": 0.08740234375, "ewc_loss_diag": 7.915496826171875e-05, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.597858905792236, "learning_rate": 1e-05, "loss": 0.5413, "mean_token_accuracy": 0.8494867086410522, "num_tokens": 141297883.0, "step": 3721 }, { "epoch": 0.47347665691387864, "ewc_loss": 0.0791015625, "ewc_loss_diag": 7.915496826171875e-05, "ewc_loss_parallel": 7.152557373046875e-05, "grad_norm": 5.2585062980651855, "learning_rate": 1e-05, "loss": 0.5112, "mean_token_accuracy": 0.8557388782501221, "num_tokens": 141336549.0, "step": 3722 }, { "epoch": 0.47360386719246916, "ewc_loss": 0.08447265625, "ewc_loss_diag": 7.915496826171875e-05, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.613807678222656, "learning_rate": 1e-05, "loss": 0.5179, "mean_token_accuracy": 0.8557793498039246, "num_tokens": 141368697.0, "step": 3723 }, { "epoch": 0.47373107747105964, "ewc_loss": 0.078125, "ewc_loss_diag": 7.915496826171875e-05, "ewc_loss_parallel": 7.009506225585938e-05, "grad_norm": 5.120263576507568, "learning_rate": 1e-05, "loss": 0.4884, "mean_token_accuracy": 0.8602825999259949, "num_tokens": 141399349.0, "step": 3724 }, { "epoch": 0.47385828774965016, "ewc_loss": 0.08544921875, "ewc_loss_diag": 7.915496826171875e-05, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.6545844078063965, "learning_rate": 1e-05, "loss": 0.4752, "mean_token_accuracy": 0.8700430989265442, "num_tokens": 141435198.0, "step": 3725 }, { "epoch": 0.4739854980282407, "ewc_loss": 0.07666015625, "ewc_loss_diag": 7.915496826171875e-05, "ewc_loss_parallel": 6.866455078125e-05, "grad_norm": 4.988657474517822, "learning_rate": 1e-05, "loss": 0.5788, "mean_token_accuracy": 0.8400788307189941, "num_tokens": 141470659.0, "step": 3726 }, { "epoch": 0.47411270830683117, "ewc_loss": 0.0869140625, "ewc_loss_diag": 7.915496826171875e-05, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.794356822967529, "learning_rate": 1e-05, "loss": 0.5269, "mean_token_accuracy": 0.8527297377586365, "num_tokens": 141514774.0, "step": 3727 }, { "epoch": 0.4742399185854217, "ewc_loss": 0.076171875, "ewc_loss_diag": 7.915496826171875e-05, "ewc_loss_parallel": 6.818771362304688e-05, "grad_norm": 4.87874698638916, "learning_rate": 1e-05, "loss": 0.4897, "mean_token_accuracy": 0.8604761958122253, "num_tokens": 141551071.0, "step": 3728 }, { "epoch": 0.4743671288640122, "ewc_loss": 0.0927734375, "ewc_loss_diag": 7.915496826171875e-05, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 6.049183368682861, "learning_rate": 1e-05, "loss": 0.5261, "mean_token_accuracy": 0.8553194403648376, "num_tokens": 141590201.0, "step": 3729 }, { "epoch": 0.4744943391426027, "ewc_loss": 0.076171875, "ewc_loss_diag": 7.915496826171875e-05, "ewc_loss_parallel": 6.818771362304688e-05, "grad_norm": 4.826931953430176, "learning_rate": 1e-05, "loss": 0.488, "mean_token_accuracy": 0.8626376390457153, "num_tokens": 141637580.0, "step": 3730 }, { "epoch": 0.4746215494211932, "ewc_loss": 0.10009765625, "ewc_loss_diag": 7.915496826171875e-05, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.385974884033203, "learning_rate": 1e-05, "loss": 0.5964, "mean_token_accuracy": 0.8419296145439148, "num_tokens": 141674092.0, "step": 3731 }, { "epoch": 0.47474875969978375, "ewc_loss": 0.0810546875, "ewc_loss_diag": 7.915496826171875e-05, "ewc_loss_parallel": 7.343292236328125e-05, "grad_norm": 5.014781475067139, "learning_rate": 1e-05, "loss": 0.575, "mean_token_accuracy": 0.8366221189498901, "num_tokens": 141709416.0, "step": 3732 }, { "epoch": 0.4748759699783743, "ewc_loss": 0.10595703125, "ewc_loss_diag": 7.915496826171875e-05, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.597455024719238, "learning_rate": 1e-05, "loss": 0.5372, "mean_token_accuracy": 0.8535999655723572, "num_tokens": 141742445.0, "step": 3733 }, { "epoch": 0.47500318025696475, "ewc_loss": 0.0888671875, "ewc_loss_diag": 7.915496826171875e-05, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.370978832244873, "learning_rate": 1e-05, "loss": 0.4898, "mean_token_accuracy": 0.864923894405365, "num_tokens": 141778224.0, "step": 3734 }, { "epoch": 0.4751303905355553, "ewc_loss": 0.10302734375, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.441310405731201, "learning_rate": 1e-05, "loss": 0.5745, "mean_token_accuracy": 0.8453295826911926, "num_tokens": 141818239.0, "step": 3735 }, { "epoch": 0.4752576008141458, "ewc_loss": 0.0927734375, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.6151123046875, "learning_rate": 1e-05, "loss": 0.5808, "mean_token_accuracy": 0.8399366140365601, "num_tokens": 141853624.0, "step": 3736 }, { "epoch": 0.4753848110927363, "ewc_loss": 0.09326171875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 6.081943035125732, "learning_rate": 1e-05, "loss": 0.5264, "mean_token_accuracy": 0.8555203676223755, "num_tokens": 141893620.0, "step": 3737 }, { "epoch": 0.4755120213713268, "ewc_loss": 0.0908203125, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.631568431854248, "learning_rate": 1e-05, "loss": 0.4982, "mean_token_accuracy": 0.8629189729690552, "num_tokens": 141930504.0, "step": 3738 }, { "epoch": 0.47563923164991734, "ewc_loss": 0.08447265625, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.704788684844971, "learning_rate": 1e-05, "loss": 0.5345, "mean_token_accuracy": 0.8461402058601379, "num_tokens": 141968195.0, "step": 3739 }, { "epoch": 0.4757664419285078, "ewc_loss": 0.087890625, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.594870567321777, "learning_rate": 1e-05, "loss": 0.5212, "mean_token_accuracy": 0.857690155506134, "num_tokens": 142010143.0, "step": 3740 }, { "epoch": 0.47589365220709834, "ewc_loss": 0.080078125, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 7.2479248046875e-05, "grad_norm": 5.406869888305664, "learning_rate": 1e-05, "loss": 0.5558, "mean_token_accuracy": 0.8437625169754028, "num_tokens": 142051170.0, "step": 3741 }, { "epoch": 0.47602086248568887, "ewc_loss": 0.0859375, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.58290958404541, "learning_rate": 1e-05, "loss": 0.5751, "mean_token_accuracy": 0.8357040882110596, "num_tokens": 142084327.0, "step": 3742 }, { "epoch": 0.47614807276427934, "ewc_loss": 0.07861328125, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 7.05718994140625e-05, "grad_norm": 5.21229887008667, "learning_rate": 1e-05, "loss": 0.5507, "mean_token_accuracy": 0.8445044755935669, "num_tokens": 142125512.0, "step": 3743 }, { "epoch": 0.47627528304286987, "ewc_loss": 0.08447265625, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.603043079376221, "learning_rate": 1e-05, "loss": 0.5057, "mean_token_accuracy": 0.8567571043968201, "num_tokens": 142159117.0, "step": 3744 }, { "epoch": 0.4764024933214604, "ewc_loss": 0.0771484375, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 6.914138793945312e-05, "grad_norm": 5.061488151550293, "learning_rate": 1e-05, "loss": 0.5108, "mean_token_accuracy": 0.8550740480422974, "num_tokens": 142194580.0, "step": 3745 }, { "epoch": 0.47652970360005087, "ewc_loss": 0.0859375, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.693911552429199, "learning_rate": 1e-05, "loss": 0.4769, "mean_token_accuracy": 0.8702532052993774, "num_tokens": 142230945.0, "step": 3746 }, { "epoch": 0.4766569138786414, "ewc_loss": 0.076171875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 6.818771362304688e-05, "grad_norm": 4.942723274230957, "learning_rate": 1e-05, "loss": 0.476, "mean_token_accuracy": 0.8638769388198853, "num_tokens": 142264792.0, "step": 3747 }, { "epoch": 0.4767841241572319, "ewc_loss": 0.0888671875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.857046604156494, "learning_rate": 1e-05, "loss": 0.4846, "mean_token_accuracy": 0.8649812936782837, "num_tokens": 142302210.0, "step": 3748 }, { "epoch": 0.4769113344358224, "ewc_loss": 0.076171875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 6.818771362304688e-05, "grad_norm": 4.855135440826416, "learning_rate": 1e-05, "loss": 0.5409, "mean_token_accuracy": 0.8477095365524292, "num_tokens": 142345646.0, "step": 3749 }, { "epoch": 0.4770385447144129, "ewc_loss": 0.09716796875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 6.146636486053467, "learning_rate": 1e-05, "loss": 0.6095, "mean_token_accuracy": 0.8315931558609009, "num_tokens": 142387097.0, "step": 3750 }, { "epoch": 0.47716575499300345, "ewc_loss": 0.07666015625, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 6.866455078125e-05, "grad_norm": 4.8621416091918945, "learning_rate": 1e-05, "loss": 0.5767, "mean_token_accuracy": 0.8332222104072571, "num_tokens": 142424087.0, "step": 3751 }, { "epoch": 0.4772929652715939, "ewc_loss": 0.10205078125, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.474056243896484, "learning_rate": 1e-05, "loss": 0.5187, "mean_token_accuracy": 0.8642960786819458, "num_tokens": 142458837.0, "step": 3752 }, { "epoch": 0.47742017555018446, "ewc_loss": 0.08251953125, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 7.486343383789062e-05, "grad_norm": 5.094485759735107, "learning_rate": 1e-05, "loss": 0.4768, "mean_token_accuracy": 0.8632104992866516, "num_tokens": 142495661.0, "step": 3753 }, { "epoch": 0.477547385828775, "ewc_loss": 0.10595703125, "ewc_loss_diag": 8.0108642578125e-05, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.59963321685791, "learning_rate": 1e-05, "loss": 0.571, "mean_token_accuracy": 0.8445958495140076, "num_tokens": 142532439.0, "step": 3754 }, { "epoch": 0.47767459610736546, "ewc_loss": 0.09033203125, "ewc_loss_diag": 8.0108642578125e-05, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.46956729888916, "learning_rate": 1e-05, "loss": 0.561, "mean_token_accuracy": 0.8428590297698975, "num_tokens": 142574032.0, "step": 3755 }, { "epoch": 0.477801806385956, "ewc_loss": 0.10107421875, "ewc_loss_diag": 8.0108642578125e-05, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.373642444610596, "learning_rate": 1e-05, "loss": 0.5188, "mean_token_accuracy": 0.8600912094116211, "num_tokens": 142607438.0, "step": 3756 }, { "epoch": 0.4779290166645465, "ewc_loss": 0.09326171875, "ewc_loss_diag": 8.0108642578125e-05, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.655258655548096, "learning_rate": 1e-05, "loss": 0.5598, "mean_token_accuracy": 0.8439942598342896, "num_tokens": 142642527.0, "step": 3757 }, { "epoch": 0.478056226943137, "ewc_loss": 0.08984375, "ewc_loss_diag": 8.0108642578125e-05, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.981502532958984, "learning_rate": 1e-05, "loss": 0.5319, "mean_token_accuracy": 0.8548714518547058, "num_tokens": 142677447.0, "step": 3758 }, { "epoch": 0.4781834372217275, "ewc_loss": 0.08935546875, "ewc_loss_diag": 8.0108642578125e-05, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.678317070007324, "learning_rate": 1e-05, "loss": 0.551, "mean_token_accuracy": 0.8463783264160156, "num_tokens": 142714429.0, "step": 3759 }, { "epoch": 0.47831064750031804, "ewc_loss": 0.08251953125, "ewc_loss_diag": 8.0108642578125e-05, "ewc_loss_parallel": 7.486343383789062e-05, "grad_norm": 5.5874714851379395, "learning_rate": 1e-05, "loss": 0.5299, "mean_token_accuracy": 0.8507362008094788, "num_tokens": 142752742.0, "step": 3760 }, { "epoch": 0.4784378577789085, "ewc_loss": 0.087890625, "ewc_loss_diag": 8.0108642578125e-05, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.612594127655029, "learning_rate": 1e-05, "loss": 0.5211, "mean_token_accuracy": 0.8553096055984497, "num_tokens": 142788925.0, "step": 3761 }, { "epoch": 0.47856506805749904, "ewc_loss": 0.07958984375, "ewc_loss_diag": 8.0108642578125e-05, "ewc_loss_parallel": 7.200241088867188e-05, "grad_norm": 5.302846908569336, "learning_rate": 1e-05, "loss": 0.4618, "mean_token_accuracy": 0.8638923168182373, "num_tokens": 142823215.0, "step": 3762 }, { "epoch": 0.47869227833608957, "ewc_loss": 0.0849609375, "ewc_loss_diag": 8.0108642578125e-05, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.631099700927734, "learning_rate": 1e-05, "loss": 0.5504, "mean_token_accuracy": 0.8443112373352051, "num_tokens": 142860346.0, "step": 3763 }, { "epoch": 0.47881948861468004, "ewc_loss": 0.078125, "ewc_loss_diag": 8.0108642578125e-05, "ewc_loss_parallel": 7.009506225585938e-05, "grad_norm": 5.1185994148254395, "learning_rate": 1e-05, "loss": 0.5297, "mean_token_accuracy": 0.8525354862213135, "num_tokens": 142895814.0, "step": 3764 }, { "epoch": 0.4789466988932706, "ewc_loss": 0.08642578125, "ewc_loss_diag": 8.0108642578125e-05, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.710461616516113, "learning_rate": 1e-05, "loss": 0.5929, "mean_token_accuracy": 0.8317716717720032, "num_tokens": 142934379.0, "step": 3765 }, { "epoch": 0.4790739091718611, "ewc_loss": 0.07666015625, "ewc_loss_diag": 8.0108642578125e-05, "ewc_loss_parallel": 6.866455078125e-05, "grad_norm": 4.99386739730835, "learning_rate": 1e-05, "loss": 0.5712, "mean_token_accuracy": 0.8342312574386597, "num_tokens": 142970512.0, "step": 3766 }, { "epoch": 0.4792011194504516, "ewc_loss": 0.08984375, "ewc_loss_diag": 8.0108642578125e-05, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.863848686218262, "learning_rate": 1e-05, "loss": 0.5701, "mean_token_accuracy": 0.8393816947937012, "num_tokens": 143008635.0, "step": 3767 }, { "epoch": 0.4793283297290421, "ewc_loss": 0.076171875, "ewc_loss_diag": 8.0108642578125e-05, "ewc_loss_parallel": 6.818771362304688e-05, "grad_norm": 4.892339706420898, "learning_rate": 1e-05, "loss": 0.4961, "mean_token_accuracy": 0.861110270023346, "num_tokens": 143047638.0, "step": 3768 }, { "epoch": 0.47945554000763263, "ewc_loss": 0.095703125, "ewc_loss_diag": 8.0108642578125e-05, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 6.107401371002197, "learning_rate": 1e-05, "loss": 0.5587, "mean_token_accuracy": 0.8462856411933899, "num_tokens": 143082470.0, "step": 3769 }, { "epoch": 0.4795827502862231, "ewc_loss": 0.07666015625, "ewc_loss_diag": 8.0108642578125e-05, "ewc_loss_parallel": 6.866455078125e-05, "grad_norm": 4.875903129577637, "learning_rate": 1e-05, "loss": 0.4957, "mean_token_accuracy": 0.8614194989204407, "num_tokens": 143126869.0, "step": 3770 }, { "epoch": 0.47970996056481363, "ewc_loss": 0.1015625, "ewc_loss_diag": 8.0108642578125e-05, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.405334949493408, "learning_rate": 1e-05, "loss": 0.4772, "mean_token_accuracy": 0.8737519979476929, "num_tokens": 143161371.0, "step": 3771 }, { "epoch": 0.47983717084340416, "ewc_loss": 0.08154296875, "ewc_loss_diag": 8.0108642578125e-05, "ewc_loss_parallel": 7.390975952148438e-05, "grad_norm": 5.067178249359131, "learning_rate": 1e-05, "loss": 0.4675, "mean_token_accuracy": 0.8696849942207336, "num_tokens": 143199896.0, "step": 3772 }, { "epoch": 0.47996438112199463, "ewc_loss": 0.10546875, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.55405855178833, "learning_rate": 1e-05, "loss": 0.5626, "mean_token_accuracy": 0.8495534062385559, "num_tokens": 143236577.0, "step": 3773 }, { "epoch": 0.48009159140058516, "ewc_loss": 0.09033203125, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.415914535522461, "learning_rate": 1e-05, "loss": 0.5299, "mean_token_accuracy": 0.8547367453575134, "num_tokens": 143279780.0, "step": 3774 }, { "epoch": 0.4802188016791757, "ewc_loss": 0.10205078125, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.3651885986328125, "learning_rate": 1e-05, "loss": 0.5229, "mean_token_accuracy": 0.8601166009902954, "num_tokens": 143321946.0, "step": 3775 }, { "epoch": 0.48034601195776616, "ewc_loss": 0.091796875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.591152667999268, "learning_rate": 1e-05, "loss": 0.5013, "mean_token_accuracy": 0.8641122579574585, "num_tokens": 143359278.0, "step": 3776 }, { "epoch": 0.4804732222363567, "ewc_loss": 0.0908203125, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 6.019450664520264, "learning_rate": 1e-05, "loss": 0.5473, "mean_token_accuracy": 0.8484873175621033, "num_tokens": 143398968.0, "step": 3777 }, { "epoch": 0.4806004325149472, "ewc_loss": 0.0908203125, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.595360279083252, "learning_rate": 1e-05, "loss": 0.5456, "mean_token_accuracy": 0.849157452583313, "num_tokens": 143438020.0, "step": 3778 }, { "epoch": 0.4807276427935377, "ewc_loss": 0.08349609375, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 7.534027099609375e-05, "grad_norm": 5.659854412078857, "learning_rate": 1e-05, "loss": 0.5021, "mean_token_accuracy": 0.861092209815979, "num_tokens": 143478813.0, "step": 3779 }, { "epoch": 0.4808548530721282, "ewc_loss": 0.087890625, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.560336589813232, "learning_rate": 1e-05, "loss": 0.557, "mean_token_accuracy": 0.846888542175293, "num_tokens": 143519766.0, "step": 3780 }, { "epoch": 0.48098206335071875, "ewc_loss": 0.0810546875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 7.295608520507812e-05, "grad_norm": 5.382325649261475, "learning_rate": 1e-05, "loss": 0.4946, "mean_token_accuracy": 0.8608899712562561, "num_tokens": 143562570.0, "step": 3781 }, { "epoch": 0.4811092736293093, "ewc_loss": 0.0849609375, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.563816070556641, "learning_rate": 1e-05, "loss": 0.5809, "mean_token_accuracy": 0.8322486281394958, "num_tokens": 143597203.0, "step": 3782 }, { "epoch": 0.48123648390789975, "ewc_loss": 0.0791015625, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 7.05718994140625e-05, "grad_norm": 5.203480243682861, "learning_rate": 1e-05, "loss": 0.5307, "mean_token_accuracy": 0.8535894155502319, "num_tokens": 143635979.0, "step": 3783 }, { "epoch": 0.4813636941864903, "ewc_loss": 0.08447265625, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 7.62939453125e-05, "grad_norm": 5.577272415161133, "learning_rate": 1e-05, "loss": 0.56, "mean_token_accuracy": 0.8445819616317749, "num_tokens": 143679915.0, "step": 3784 }, { "epoch": 0.4814909044650808, "ewc_loss": 0.078125, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 6.961822509765625e-05, "grad_norm": 5.0872039794921875, "learning_rate": 1e-05, "loss": 0.5636, "mean_token_accuracy": 0.8393030166625977, "num_tokens": 143716275.0, "step": 3785 }, { "epoch": 0.4816181147436713, "ewc_loss": 0.0849609375, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.634915351867676, "learning_rate": 1e-05, "loss": 0.5344, "mean_token_accuracy": 0.8472059965133667, "num_tokens": 143752029.0, "step": 3786 }, { "epoch": 0.4817453250222618, "ewc_loss": 0.076171875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 6.771087646484375e-05, "grad_norm": 4.966951847076416, "learning_rate": 1e-05, "loss": 0.4956, "mean_token_accuracy": 0.8609135746955872, "num_tokens": 143793139.0, "step": 3787 }, { "epoch": 0.48187253530085233, "ewc_loss": 0.087890625, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.794713973999023, "learning_rate": 1e-05, "loss": 0.4834, "mean_token_accuracy": 0.8672876954078674, "num_tokens": 143832790.0, "step": 3788 }, { "epoch": 0.4819997455794428, "ewc_loss": 0.0751953125, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 6.723403930664062e-05, "grad_norm": 4.852657794952393, "learning_rate": 1e-05, "loss": 0.5232, "mean_token_accuracy": 0.8493112921714783, "num_tokens": 143876046.0, "step": 3789 }, { "epoch": 0.48212695585803333, "ewc_loss": 0.09375, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 6.082688808441162, "learning_rate": 1e-05, "loss": 0.5485, "mean_token_accuracy": 0.8466092348098755, "num_tokens": 143917019.0, "step": 3790 }, { "epoch": 0.48225416613662386, "ewc_loss": 0.076171875, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 6.818771362304688e-05, "grad_norm": 4.811277389526367, "learning_rate": 1e-05, "loss": 0.4995, "mean_token_accuracy": 0.8606797456741333, "num_tokens": 143957789.0, "step": 3791 }, { "epoch": 0.48238137641521434, "ewc_loss": 0.1025390625, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.443737506866455, "learning_rate": 1e-05, "loss": 0.5204, "mean_token_accuracy": 0.8595912456512451, "num_tokens": 143993797.0, "step": 3792 }, { "epoch": 0.48250858669380486, "ewc_loss": 0.08203125, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 7.390975952148438e-05, "grad_norm": 5.040069580078125, "learning_rate": 1e-05, "loss": 0.5851, "mean_token_accuracy": 0.8367879390716553, "num_tokens": 144033583.0, "step": 3793 }, { "epoch": 0.4826357969723954, "ewc_loss": 0.10791015625, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.632077217102051, "learning_rate": 1e-05, "loss": 0.5315, "mean_token_accuracy": 0.8561128377914429, "num_tokens": 144074358.0, "step": 3794 }, { "epoch": 0.48276300725098586, "ewc_loss": 0.09033203125, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.402012825012207, "learning_rate": 1e-05, "loss": 0.5654, "mean_token_accuracy": 0.846147358417511, "num_tokens": 144114480.0, "step": 3795 }, { "epoch": 0.4828902175295764, "ewc_loss": 0.10400390625, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.470852851867676, "learning_rate": 1e-05, "loss": 0.5303, "mean_token_accuracy": 0.8596634864807129, "num_tokens": 144152306.0, "step": 3796 }, { "epoch": 0.4830174278081669, "ewc_loss": 0.09326171875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.6397705078125, "learning_rate": 1e-05, "loss": 0.5177, "mean_token_accuracy": 0.8563884496688843, "num_tokens": 144191134.0, "step": 3797 }, { "epoch": 0.4831446380867574, "ewc_loss": 0.0947265625, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 6.101889610290527, "learning_rate": 1e-05, "loss": 0.5487, "mean_token_accuracy": 0.8476082682609558, "num_tokens": 144230165.0, "step": 3798 }, { "epoch": 0.4832718483653479, "ewc_loss": 0.09130859375, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.641991138458252, "learning_rate": 1e-05, "loss": 0.5236, "mean_token_accuracy": 0.8583969473838806, "num_tokens": 144265035.0, "step": 3799 }, { "epoch": 0.48339905864393845, "ewc_loss": 0.08544921875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.7262163162231445, "learning_rate": 1e-05, "loss": 0.4682, "mean_token_accuracy": 0.8719483613967896, "num_tokens": 144304692.0, "step": 3800 }, { "epoch": 0.4835262689225289, "ewc_loss": 0.087890625, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.603699684143066, "learning_rate": 1e-05, "loss": 0.5335, "mean_token_accuracy": 0.8494049906730652, "num_tokens": 144345288.0, "step": 3801 }, { "epoch": 0.48365347920111945, "ewc_loss": 0.08203125, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 7.390975952148438e-05, "grad_norm": 5.445196628570557, "learning_rate": 1e-05, "loss": 0.5463, "mean_token_accuracy": 0.8453480005264282, "num_tokens": 144377427.0, "step": 3802 }, { "epoch": 0.48378068947971, "ewc_loss": 0.08642578125, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.5792460441589355, "learning_rate": 1e-05, "loss": 0.4919, "mean_token_accuracy": 0.8644589185714722, "num_tokens": 144412405.0, "step": 3803 }, { "epoch": 0.48390789975830045, "ewc_loss": 0.0791015625, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 7.05718994140625e-05, "grad_norm": 5.234692573547363, "learning_rate": 1e-05, "loss": 0.4763, "mean_token_accuracy": 0.866424560546875, "num_tokens": 144448093.0, "step": 3804 }, { "epoch": 0.484035110036891, "ewc_loss": 0.08544921875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.595493793487549, "learning_rate": 1e-05, "loss": 0.5299, "mean_token_accuracy": 0.8517704606056213, "num_tokens": 144483280.0, "step": 3805 }, { "epoch": 0.4841623203154815, "ewc_loss": 0.078125, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 6.961822509765625e-05, "grad_norm": 5.090397834777832, "learning_rate": 1e-05, "loss": 0.448, "mean_token_accuracy": 0.8744224309921265, "num_tokens": 144520827.0, "step": 3806 }, { "epoch": 0.484289530594072, "ewc_loss": 0.0859375, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.662955284118652, "learning_rate": 1e-05, "loss": 0.4912, "mean_token_accuracy": 0.8626090288162231, "num_tokens": 144562314.0, "step": 3807 }, { "epoch": 0.4844167408726625, "ewc_loss": 0.07763671875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 6.914138793945312e-05, "grad_norm": 4.971446514129639, "learning_rate": 1e-05, "loss": 0.4894, "mean_token_accuracy": 0.8607757687568665, "num_tokens": 144599101.0, "step": 3808 }, { "epoch": 0.48454395115125304, "ewc_loss": 0.0888671875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.831233978271484, "learning_rate": 1e-05, "loss": 0.5137, "mean_token_accuracy": 0.8579344749450684, "num_tokens": 144641308.0, "step": 3809 }, { "epoch": 0.4846711614298435, "ewc_loss": 0.076171875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 6.771087646484375e-05, "grad_norm": 4.861392021179199, "learning_rate": 1e-05, "loss": 0.5593, "mean_token_accuracy": 0.8440710306167603, "num_tokens": 144677267.0, "step": 3810 }, { "epoch": 0.48479837170843404, "ewc_loss": 0.09521484375, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 6.11992073059082, "learning_rate": 1e-05, "loss": 0.5702, "mean_token_accuracy": 0.8434128761291504, "num_tokens": 144713904.0, "step": 3811 }, { "epoch": 0.48492558198702457, "ewc_loss": 0.0771484375, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 6.866455078125e-05, "grad_norm": 4.830538749694824, "learning_rate": 1e-05, "loss": 0.5283, "mean_token_accuracy": 0.8536175489425659, "num_tokens": 144750523.0, "step": 3812 }, { "epoch": 0.48505279226561504, "ewc_loss": 0.10302734375, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.476043701171875, "learning_rate": 1e-05, "loss": 0.5427, "mean_token_accuracy": 0.8541818857192993, "num_tokens": 144791569.0, "step": 3813 }, { "epoch": 0.48518000254420557, "ewc_loss": 0.08251953125, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 7.43865966796875e-05, "grad_norm": 5.0514726638793945, "learning_rate": 1e-05, "loss": 0.4995, "mean_token_accuracy": 0.8625246286392212, "num_tokens": 144827919.0, "step": 3814 }, { "epoch": 0.4853072128227961, "ewc_loss": 0.1083984375, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.669386386871338, "learning_rate": 1e-05, "loss": 0.4863, "mean_token_accuracy": 0.8715658783912659, "num_tokens": 144866859.0, "step": 3815 }, { "epoch": 0.48543442310138657, "ewc_loss": 0.09130859375, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.47482967376709, "learning_rate": 1e-05, "loss": 0.4874, "mean_token_accuracy": 0.8699771165847778, "num_tokens": 144903088.0, "step": 3816 }, { "epoch": 0.4855616333799771, "ewc_loss": 0.103515625, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.469377040863037, "learning_rate": 1e-05, "loss": 0.4933, "mean_token_accuracy": 0.8682961463928223, "num_tokens": 144938485.0, "step": 3817 }, { "epoch": 0.4856888436585676, "ewc_loss": 0.09521484375, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.708752155303955, "learning_rate": 1e-05, "loss": 0.6129, "mean_token_accuracy": 0.8382292985916138, "num_tokens": 144974296.0, "step": 3818 }, { "epoch": 0.4858160539371581, "ewc_loss": 0.091796875, "ewc_loss_diag": 8.106231689453125e-05, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 6.050959587097168, "learning_rate": 1e-05, "loss": 0.5273, "mean_token_accuracy": 0.8568210601806641, "num_tokens": 145012391.0, "step": 3819 }, { "epoch": 0.4859432642157486, "ewc_loss": 0.091796875, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.692458152770996, "learning_rate": 1e-05, "loss": 0.5714, "mean_token_accuracy": 0.8397712707519531, "num_tokens": 145045409.0, "step": 3820 }, { "epoch": 0.48607047449433916, "ewc_loss": 0.0830078125, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 7.486343383789062e-05, "grad_norm": 5.6618218421936035, "learning_rate": 1e-05, "loss": 0.5011, "mean_token_accuracy": 0.8626425862312317, "num_tokens": 145081162.0, "step": 3821 }, { "epoch": 0.48619768477292963, "ewc_loss": 0.0888671875, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.635595798492432, "learning_rate": 1e-05, "loss": 0.5063, "mean_token_accuracy": 0.8599861264228821, "num_tokens": 145117045.0, "step": 3822 }, { "epoch": 0.48632489505152016, "ewc_loss": 0.0810546875, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 7.295608520507812e-05, "grad_norm": 5.383828163146973, "learning_rate": 1e-05, "loss": 0.4885, "mean_token_accuracy": 0.8633192777633667, "num_tokens": 145153931.0, "step": 3823 }, { "epoch": 0.4864521053301107, "ewc_loss": 0.08642578125, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.620996475219727, "learning_rate": 1e-05, "loss": 0.5077, "mean_token_accuracy": 0.8601709604263306, "num_tokens": 145193812.0, "step": 3824 }, { "epoch": 0.48657931560870116, "ewc_loss": 0.07861328125, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 7.009506225585938e-05, "grad_norm": 5.188714504241943, "learning_rate": 1e-05, "loss": 0.5562, "mean_token_accuracy": 0.8478537797927856, "num_tokens": 145231814.0, "step": 3825 }, { "epoch": 0.4867065258872917, "ewc_loss": 0.0849609375, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.654578685760498, "learning_rate": 1e-05, "loss": 0.5316, "mean_token_accuracy": 0.8499008417129517, "num_tokens": 145279821.0, "step": 3826 }, { "epoch": 0.4868337361658822, "ewc_loss": 0.078125, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 6.961822509765625e-05, "grad_norm": 5.057418346405029, "learning_rate": 1e-05, "loss": 0.5365, "mean_token_accuracy": 0.849538266658783, "num_tokens": 145318972.0, "step": 3827 }, { "epoch": 0.4869609464444727, "ewc_loss": 0.08740234375, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.753225803375244, "learning_rate": 1e-05, "loss": 0.5304, "mean_token_accuracy": 0.8582831025123596, "num_tokens": 145359479.0, "step": 3828 }, { "epoch": 0.4870881567230632, "ewc_loss": 0.07666015625, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 6.818771362304688e-05, "grad_norm": 4.927195072174072, "learning_rate": 1e-05, "loss": 0.5429, "mean_token_accuracy": 0.8461386561393738, "num_tokens": 145403119.0, "step": 3829 }, { "epoch": 0.48721536700165374, "ewc_loss": 0.09130859375, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.970661640167236, "learning_rate": 1e-05, "loss": 0.5429, "mean_token_accuracy": 0.846208930015564, "num_tokens": 145439715.0, "step": 3830 }, { "epoch": 0.48734257728024427, "ewc_loss": 0.076171875, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 6.771087646484375e-05, "grad_norm": 4.85077428817749, "learning_rate": 1e-05, "loss": 0.485, "mean_token_accuracy": 0.8635274171829224, "num_tokens": 145480715.0, "step": 3831 }, { "epoch": 0.48746978755883474, "ewc_loss": 0.09912109375, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.285459518432617, "learning_rate": 1e-05, "loss": 0.5457, "mean_token_accuracy": 0.8540551662445068, "num_tokens": 145523501.0, "step": 3832 }, { "epoch": 0.48759699783742527, "ewc_loss": 0.07958984375, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 7.104873657226562e-05, "grad_norm": 4.932252407073975, "learning_rate": 1e-05, "loss": 0.5832, "mean_token_accuracy": 0.831571638584137, "num_tokens": 145563387.0, "step": 3833 }, { "epoch": 0.4877242081160158, "ewc_loss": 0.10546875, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.569146633148193, "learning_rate": 1e-05, "loss": 0.5884, "mean_token_accuracy": 0.8404930830001831, "num_tokens": 145607818.0, "step": 3834 }, { "epoch": 0.4878514183946063, "ewc_loss": 0.0869140625, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.261733531951904, "learning_rate": 1e-05, "loss": 0.5582, "mean_token_accuracy": 0.8427075147628784, "num_tokens": 145646115.0, "step": 3835 }, { "epoch": 0.4879786286731968, "ewc_loss": 0.1044921875, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.551755905151367, "learning_rate": 1e-05, "loss": 0.5059, "mean_token_accuracy": 0.8613515496253967, "num_tokens": 145681555.0, "step": 3836 }, { "epoch": 0.48810583895178733, "ewc_loss": 0.09228515625, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.57684326171875, "learning_rate": 1e-05, "loss": 0.4928, "mean_token_accuracy": 0.8651701211929321, "num_tokens": 145717459.0, "step": 3837 }, { "epoch": 0.4882330492303778, "ewc_loss": 0.0986328125, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.236088275909424, "learning_rate": 1e-05, "loss": 0.5572, "mean_token_accuracy": 0.8465872406959534, "num_tokens": 145753143.0, "step": 3838 }, { "epoch": 0.48836025950896833, "ewc_loss": 0.091796875, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.657748222351074, "learning_rate": 1e-05, "loss": 0.5546, "mean_token_accuracy": 0.8466641902923584, "num_tokens": 145790345.0, "step": 3839 }, { "epoch": 0.48848746978755886, "ewc_loss": 0.0869140625, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.830135822296143, "learning_rate": 1e-05, "loss": 0.547, "mean_token_accuracy": 0.8489677309989929, "num_tokens": 145832900.0, "step": 3840 }, { "epoch": 0.48861468006614933, "ewc_loss": 0.08935546875, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.627295017242432, "learning_rate": 1e-05, "loss": 0.5068, "mean_token_accuracy": 0.8598936796188354, "num_tokens": 145870309.0, "step": 3841 }, { "epoch": 0.48874189034473986, "ewc_loss": 0.08154296875, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 7.343292236328125e-05, "grad_norm": 5.492350101470947, "learning_rate": 1e-05, "loss": 0.5041, "mean_token_accuracy": 0.8613398671150208, "num_tokens": 145910276.0, "step": 3842 }, { "epoch": 0.4888691006233304, "ewc_loss": 0.08837890625, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.611399173736572, "learning_rate": 1e-05, "loss": 0.4946, "mean_token_accuracy": 0.863670289516449, "num_tokens": 145952945.0, "step": 3843 }, { "epoch": 0.48899631090192086, "ewc_loss": 0.07958984375, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 7.104873657226562e-05, "grad_norm": 5.262831687927246, "learning_rate": 1e-05, "loss": 0.5093, "mean_token_accuracy": 0.8508670926094055, "num_tokens": 145987339.0, "step": 3844 }, { "epoch": 0.4891235211805114, "ewc_loss": 0.08544921875, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.626413822174072, "learning_rate": 1e-05, "loss": 0.5142, "mean_token_accuracy": 0.8564162254333496, "num_tokens": 146027447.0, "step": 3845 }, { "epoch": 0.4892507314591019, "ewc_loss": 0.078125, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 6.961822509765625e-05, "grad_norm": 5.11560583114624, "learning_rate": 1e-05, "loss": 0.5018, "mean_token_accuracy": 0.85847008228302, "num_tokens": 146064192.0, "step": 3846 }, { "epoch": 0.4893779417376924, "ewc_loss": 0.08642578125, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.6901116371154785, "learning_rate": 1e-05, "loss": 0.5171, "mean_token_accuracy": 0.8570421934127808, "num_tokens": 146103468.0, "step": 3847 }, { "epoch": 0.4895051520162829, "ewc_loss": 0.0771484375, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 6.866455078125e-05, "grad_norm": 5.027221202850342, "learning_rate": 1e-05, "loss": 0.4798, "mean_token_accuracy": 0.8634265661239624, "num_tokens": 146142499.0, "step": 3848 }, { "epoch": 0.48963236229487345, "ewc_loss": 0.0888671875, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.802493572235107, "learning_rate": 1e-05, "loss": 0.531, "mean_token_accuracy": 0.8499748110771179, "num_tokens": 146177538.0, "step": 3849 }, { "epoch": 0.4897595725734639, "ewc_loss": 0.07666015625, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 6.818771362304688e-05, "grad_norm": 4.9034318923950195, "learning_rate": 1e-05, "loss": 0.4923, "mean_token_accuracy": 0.8641064167022705, "num_tokens": 146218259.0, "step": 3850 }, { "epoch": 0.48988678285205445, "ewc_loss": 0.09228515625, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 6.02667236328125, "learning_rate": 1e-05, "loss": 0.5267, "mean_token_accuracy": 0.8562444448471069, "num_tokens": 146256630.0, "step": 3851 }, { "epoch": 0.490013993130645, "ewc_loss": 0.076171875, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 6.771087646484375e-05, "grad_norm": 4.87534761428833, "learning_rate": 1e-05, "loss": 0.5765, "mean_token_accuracy": 0.8363968133926392, "num_tokens": 146290891.0, "step": 3852 }, { "epoch": 0.49014120340923545, "ewc_loss": 0.10009765625, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.324031352996826, "learning_rate": 1e-05, "loss": 0.5465, "mean_token_accuracy": 0.8490548133850098, "num_tokens": 146327148.0, "step": 3853 }, { "epoch": 0.490268413687826, "ewc_loss": 0.080078125, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 7.200241088867188e-05, "grad_norm": 4.97987699508667, "learning_rate": 1e-05, "loss": 0.4994, "mean_token_accuracy": 0.8577320575714111, "num_tokens": 146367932.0, "step": 3854 }, { "epoch": 0.4903956239664165, "ewc_loss": 0.10546875, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.552041053771973, "learning_rate": 1e-05, "loss": 0.5189, "mean_token_accuracy": 0.8608682155609131, "num_tokens": 146407639.0, "step": 3855 }, { "epoch": 0.490522834245007, "ewc_loss": 0.087890625, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.301537990570068, "learning_rate": 1e-05, "loss": 0.5434, "mean_token_accuracy": 0.847545862197876, "num_tokens": 146442974.0, "step": 3856 }, { "epoch": 0.4906500445235975, "ewc_loss": 0.10400390625, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.482608795166016, "learning_rate": 1e-05, "loss": 0.5199, "mean_token_accuracy": 0.8621517419815063, "num_tokens": 146482510.0, "step": 3857 }, { "epoch": 0.49077725480218803, "ewc_loss": 0.091796875, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.562923908233643, "learning_rate": 1e-05, "loss": 0.4852, "mean_token_accuracy": 0.8681852221488953, "num_tokens": 146522616.0, "step": 3858 }, { "epoch": 0.4909044650807785, "ewc_loss": 0.09814453125, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.166316509246826, "learning_rate": 1e-05, "loss": 0.5002, "mean_token_accuracy": 0.863349437713623, "num_tokens": 146558226.0, "step": 3859 }, { "epoch": 0.49103167535936904, "ewc_loss": 0.0927734375, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.622770309448242, "learning_rate": 1e-05, "loss": 0.5434, "mean_token_accuracy": 0.8501812815666199, "num_tokens": 146601873.0, "step": 3860 }, { "epoch": 0.49115888563795956, "ewc_loss": 0.0869140625, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.790235996246338, "learning_rate": 1e-05, "loss": 0.519, "mean_token_accuracy": 0.8577579259872437, "num_tokens": 146645284.0, "step": 3861 }, { "epoch": 0.49128609591655004, "ewc_loss": 0.08935546875, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.605926513671875, "learning_rate": 1e-05, "loss": 0.5224, "mean_token_accuracy": 0.8581175804138184, "num_tokens": 146680374.0, "step": 3862 }, { "epoch": 0.49141330619514056, "ewc_loss": 0.0810546875, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 7.295608520507812e-05, "grad_norm": 5.472292900085449, "learning_rate": 1e-05, "loss": 0.5325, "mean_token_accuracy": 0.850159764289856, "num_tokens": 146721915.0, "step": 3863 }, { "epoch": 0.4915405164737311, "ewc_loss": 0.087890625, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.596439361572266, "learning_rate": 1e-05, "loss": 0.4865, "mean_token_accuracy": 0.8659316301345825, "num_tokens": 146758981.0, "step": 3864 }, { "epoch": 0.49166772675232157, "ewc_loss": 0.0791015625, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 7.05718994140625e-05, "grad_norm": 5.252364635467529, "learning_rate": 1e-05, "loss": 0.5172, "mean_token_accuracy": 0.8502348065376282, "num_tokens": 146795606.0, "step": 3865 }, { "epoch": 0.4917949370309121, "ewc_loss": 0.08544921875, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.623505115509033, "learning_rate": 1e-05, "loss": 0.4761, "mean_token_accuracy": 0.8646588325500488, "num_tokens": 146829681.0, "step": 3866 }, { "epoch": 0.4919221473095026, "ewc_loss": 0.078125, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 6.961822509765625e-05, "grad_norm": 5.090583801269531, "learning_rate": 1e-05, "loss": 0.4616, "mean_token_accuracy": 0.8711955547332764, "num_tokens": 146874142.0, "step": 3867 }, { "epoch": 0.4920493575880931, "ewc_loss": 0.0869140625, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.712726593017578, "learning_rate": 1e-05, "loss": 0.6291, "mean_token_accuracy": 0.8249850273132324, "num_tokens": 146911414.0, "step": 3868 }, { "epoch": 0.4921765678666836, "ewc_loss": 0.0771484375, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 6.866455078125e-05, "grad_norm": 4.971226215362549, "learning_rate": 1e-05, "loss": 0.5649, "mean_token_accuracy": 0.8372235298156738, "num_tokens": 146956376.0, "step": 3869 }, { "epoch": 0.49230377814527415, "ewc_loss": 0.08984375, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.8801045417785645, "learning_rate": 1e-05, "loss": 0.5622, "mean_token_accuracy": 0.8439359068870544, "num_tokens": 147001539.0, "step": 3870 }, { "epoch": 0.4924309884238646, "ewc_loss": 0.076171875, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 6.771087646484375e-05, "grad_norm": 4.875004768371582, "learning_rate": 1e-05, "loss": 0.5453, "mean_token_accuracy": 0.8436100482940674, "num_tokens": 147043782.0, "step": 3871 }, { "epoch": 0.49255819870245515, "ewc_loss": 0.09716796875, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.18019962310791, "learning_rate": 1e-05, "loss": 0.5863, "mean_token_accuracy": 0.8349367380142212, "num_tokens": 147077431.0, "step": 3872 }, { "epoch": 0.4926854089810457, "ewc_loss": 0.07763671875, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 6.914138793945312e-05, "grad_norm": 4.874675273895264, "learning_rate": 1e-05, "loss": 0.4795, "mean_token_accuracy": 0.8645188808441162, "num_tokens": 147113325.0, "step": 3873 }, { "epoch": 0.49281261925963615, "ewc_loss": 0.103515625, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.50070858001709, "learning_rate": 1e-05, "loss": 0.5956, "mean_token_accuracy": 0.8369107842445374, "num_tokens": 147158363.0, "step": 3874 }, { "epoch": 0.4929398295382267, "ewc_loss": 0.0849609375, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.123782634735107, "learning_rate": 1e-05, "loss": 0.5656, "mean_token_accuracy": 0.8448226451873779, "num_tokens": 147195868.0, "step": 3875 }, { "epoch": 0.4930670398168172, "ewc_loss": 0.10595703125, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.612184524536133, "learning_rate": 1e-05, "loss": 0.6146, "mean_token_accuracy": 0.8328912258148193, "num_tokens": 147237782.0, "step": 3876 }, { "epoch": 0.4931942500954077, "ewc_loss": 0.09033203125, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.500720977783203, "learning_rate": 1e-05, "loss": 0.5479, "mean_token_accuracy": 0.8458305597305298, "num_tokens": 147274709.0, "step": 3877 }, { "epoch": 0.4933214603739982, "ewc_loss": 0.1025390625, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.366994380950928, "learning_rate": 1e-05, "loss": 0.5256, "mean_token_accuracy": 0.8611075282096863, "num_tokens": 147312261.0, "step": 3878 }, { "epoch": 0.49344867065258874, "ewc_loss": 0.09423828125, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.653034687042236, "learning_rate": 1e-05, "loss": 0.5295, "mean_token_accuracy": 0.8559476137161255, "num_tokens": 147352290.0, "step": 3879 }, { "epoch": 0.4935758809311792, "ewc_loss": 0.08984375, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.982937335968018, "learning_rate": 1e-05, "loss": 0.5352, "mean_token_accuracy": 0.8574643135070801, "num_tokens": 147390558.0, "step": 3880 }, { "epoch": 0.49370309120976974, "ewc_loss": 0.0908203125, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.642609119415283, "learning_rate": 1e-05, "loss": 0.5463, "mean_token_accuracy": 0.8493012189865112, "num_tokens": 147429597.0, "step": 3881 }, { "epoch": 0.49383030148836027, "ewc_loss": 0.08251953125, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 7.43865966796875e-05, "grad_norm": 5.61575174331665, "learning_rate": 1e-05, "loss": 0.5418, "mean_token_accuracy": 0.8473926782608032, "num_tokens": 147472302.0, "step": 3882 }, { "epoch": 0.4939575117669508, "ewc_loss": 0.0888671875, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.618223190307617, "learning_rate": 1e-05, "loss": 0.4797, "mean_token_accuracy": 0.8680315613746643, "num_tokens": 147502846.0, "step": 3883 }, { "epoch": 0.49408472204554127, "ewc_loss": 0.08056640625, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 7.2479248046875e-05, "grad_norm": 5.350293159484863, "learning_rate": 1e-05, "loss": 0.4978, "mean_token_accuracy": 0.8626524209976196, "num_tokens": 147542351.0, "step": 3884 }, { "epoch": 0.4942119323241318, "ewc_loss": 0.0859375, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.608582019805908, "learning_rate": 1e-05, "loss": 0.4985, "mean_token_accuracy": 0.8606044054031372, "num_tokens": 147576912.0, "step": 3885 }, { "epoch": 0.4943391426027223, "ewc_loss": 0.07861328125, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 7.009506225585938e-05, "grad_norm": 5.181553363800049, "learning_rate": 1e-05, "loss": 0.536, "mean_token_accuracy": 0.8518441915512085, "num_tokens": 147614687.0, "step": 3886 }, { "epoch": 0.4944663528813128, "ewc_loss": 0.0859375, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.64654541015625, "learning_rate": 1e-05, "loss": 0.4945, "mean_token_accuracy": 0.863604724407196, "num_tokens": 147649048.0, "step": 3887 }, { "epoch": 0.4945935631599033, "ewc_loss": 0.07763671875, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 6.914138793945312e-05, "grad_norm": 5.060985565185547, "learning_rate": 1e-05, "loss": 0.5451, "mean_token_accuracy": 0.8454768657684326, "num_tokens": 147689337.0, "step": 3888 }, { "epoch": 0.49472077343849385, "ewc_loss": 0.08740234375, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.734976291656494, "learning_rate": 1e-05, "loss": 0.524, "mean_token_accuracy": 0.8561185598373413, "num_tokens": 147725040.0, "step": 3889 }, { "epoch": 0.4948479837170843, "ewc_loss": 0.0771484375, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 6.866455078125e-05, "grad_norm": 4.954909324645996, "learning_rate": 1e-05, "loss": 0.5129, "mean_token_accuracy": 0.8537963628768921, "num_tokens": 147761442.0, "step": 3890 }, { "epoch": 0.49497519399567486, "ewc_loss": 0.09033203125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.926100254058838, "learning_rate": 1e-05, "loss": 0.5337, "mean_token_accuracy": 0.8505786657333374, "num_tokens": 147795421.0, "step": 3891 }, { "epoch": 0.4951024042742654, "ewc_loss": 0.07666015625, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 6.818771362304688e-05, "grad_norm": 4.858320713043213, "learning_rate": 1e-05, "loss": 0.5068, "mean_token_accuracy": 0.8542446494102478, "num_tokens": 147833591.0, "step": 3892 }, { "epoch": 0.49522961455285586, "ewc_loss": 0.0986328125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.236225128173828, "learning_rate": 1e-05, "loss": 0.5608, "mean_token_accuracy": 0.844272255897522, "num_tokens": 147870707.0, "step": 3893 }, { "epoch": 0.4953568248314464, "ewc_loss": 0.0771484375, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 6.866455078125e-05, "grad_norm": 4.904145240783691, "learning_rate": 1e-05, "loss": 0.5032, "mean_token_accuracy": 0.8546757698059082, "num_tokens": 147905231.0, "step": 3894 }, { "epoch": 0.4954840351100369, "ewc_loss": 0.10498046875, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.546565055847168, "learning_rate": 1e-05, "loss": 0.5251, "mean_token_accuracy": 0.8570176362991333, "num_tokens": 147945890.0, "step": 3895 }, { "epoch": 0.4956112453886274, "ewc_loss": 0.08544921875, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.184699535369873, "learning_rate": 1e-05, "loss": 0.5153, "mean_token_accuracy": 0.8575066924095154, "num_tokens": 147984572.0, "step": 3896 }, { "epoch": 0.4957384556672179, "ewc_loss": 0.10546875, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.6086506843566895, "learning_rate": 1e-05, "loss": 0.5059, "mean_token_accuracy": 0.8644900321960449, "num_tokens": 148020059.0, "step": 3897 }, { "epoch": 0.49586566594580844, "ewc_loss": 0.091796875, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.551581382751465, "learning_rate": 1e-05, "loss": 0.4749, "mean_token_accuracy": 0.8688153028488159, "num_tokens": 148058442.0, "step": 3898 }, { "epoch": 0.4959928762243989, "ewc_loss": 0.1015625, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.327269554138184, "learning_rate": 1e-05, "loss": 0.5071, "mean_token_accuracy": 0.863371729850769, "num_tokens": 148093710.0, "step": 3899 }, { "epoch": 0.49612008650298944, "ewc_loss": 0.09375, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.671464920043945, "learning_rate": 1e-05, "loss": 0.4518, "mean_token_accuracy": 0.8760617971420288, "num_tokens": 148135763.0, "step": 3900 }, { "epoch": 0.49624729678157997, "ewc_loss": 0.08837890625, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.927298545837402, "learning_rate": 1e-05, "loss": 0.4954, "mean_token_accuracy": 0.8640997409820557, "num_tokens": 148169065.0, "step": 3901 }, { "epoch": 0.49637450706017044, "ewc_loss": 0.0908203125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.659851551055908, "learning_rate": 1e-05, "loss": 0.572, "mean_token_accuracy": 0.8440454602241516, "num_tokens": 148205779.0, "step": 3902 }, { "epoch": 0.496501717338761, "ewc_loss": 0.08349609375, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 7.534027099609375e-05, "grad_norm": 5.569063663482666, "learning_rate": 1e-05, "loss": 0.5473, "mean_token_accuracy": 0.8411528468132019, "num_tokens": 148241893.0, "step": 3903 }, { "epoch": 0.4966289276173515, "ewc_loss": 0.08837890625, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.636514186859131, "learning_rate": 1e-05, "loss": 0.499, "mean_token_accuracy": 0.8586740493774414, "num_tokens": 148275384.0, "step": 3904 }, { "epoch": 0.496756137895942, "ewc_loss": 0.07958984375, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 7.152557373046875e-05, "grad_norm": 5.2971038818359375, "learning_rate": 1e-05, "loss": 0.5172, "mean_token_accuracy": 0.855854332447052, "num_tokens": 148311248.0, "step": 3905 }, { "epoch": 0.4968833481745325, "ewc_loss": 0.0869140625, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.653934001922607, "learning_rate": 1e-05, "loss": 0.524, "mean_token_accuracy": 0.8555940389633179, "num_tokens": 148347717.0, "step": 3906 }, { "epoch": 0.49701055845312303, "ewc_loss": 0.07861328125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 7.009506225585938e-05, "grad_norm": 5.124106407165527, "learning_rate": 1e-05, "loss": 0.5743, "mean_token_accuracy": 0.837895929813385, "num_tokens": 148393395.0, "step": 3907 }, { "epoch": 0.4971377687317135, "ewc_loss": 0.0869140625, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.73118257522583, "learning_rate": 1e-05, "loss": 0.4983, "mean_token_accuracy": 0.8656208515167236, "num_tokens": 148427843.0, "step": 3908 }, { "epoch": 0.49726497901030403, "ewc_loss": 0.0771484375, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 6.866455078125e-05, "grad_norm": 4.9879350662231445, "learning_rate": 1e-05, "loss": 0.5456, "mean_token_accuracy": 0.8452903628349304, "num_tokens": 148470813.0, "step": 3909 }, { "epoch": 0.49739218928889456, "ewc_loss": 0.08984375, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.906684875488281, "learning_rate": 1e-05, "loss": 0.5302, "mean_token_accuracy": 0.8542821407318115, "num_tokens": 148504491.0, "step": 3910 }, { "epoch": 0.49751939956748503, "ewc_loss": 0.07666015625, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 6.818771362304688e-05, "grad_norm": 4.8932390213012695, "learning_rate": 1e-05, "loss": 0.5261, "mean_token_accuracy": 0.8491832613945007, "num_tokens": 148545934.0, "step": 3911 }, { "epoch": 0.49764660984607556, "ewc_loss": 0.0966796875, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 6.1753644943237305, "learning_rate": 1e-05, "loss": 0.5076, "mean_token_accuracy": 0.8626443147659302, "num_tokens": 148583521.0, "step": 3912 }, { "epoch": 0.4977738201246661, "ewc_loss": 0.07763671875, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 6.914138793945312e-05, "grad_norm": 4.908710956573486, "learning_rate": 1e-05, "loss": 0.5186, "mean_token_accuracy": 0.8509485125541687, "num_tokens": 148621484.0, "step": 3913 }, { "epoch": 0.49790103040325656, "ewc_loss": 0.10302734375, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.470133304595947, "learning_rate": 1e-05, "loss": 0.5751, "mean_token_accuracy": 0.8448190689086914, "num_tokens": 148661198.0, "step": 3914 }, { "epoch": 0.4980282406818471, "ewc_loss": 0.083984375, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 7.581710815429688e-05, "grad_norm": 5.129355430603027, "learning_rate": 1e-05, "loss": 0.5233, "mean_token_accuracy": 0.8543426990509033, "num_tokens": 148702620.0, "step": 3915 }, { "epoch": 0.4981554509604376, "ewc_loss": 0.1064453125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.576495170593262, "learning_rate": 1e-05, "loss": 0.554, "mean_token_accuracy": 0.851780891418457, "num_tokens": 148739991.0, "step": 3916 }, { "epoch": 0.4982826612390281, "ewc_loss": 0.09033203125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.472804069519043, "learning_rate": 1e-05, "loss": 0.4721, "mean_token_accuracy": 0.8725391030311584, "num_tokens": 148780520.0, "step": 3917 }, { "epoch": 0.4984098715176186, "ewc_loss": 0.1015625, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.362744331359863, "learning_rate": 1e-05, "loss": 0.5554, "mean_token_accuracy": 0.846989631652832, "num_tokens": 148817604.0, "step": 3918 }, { "epoch": 0.49853708179620915, "ewc_loss": 0.09228515625, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.643836975097656, "learning_rate": 1e-05, "loss": 0.5314, "mean_token_accuracy": 0.8513402342796326, "num_tokens": 148858028.0, "step": 3919 }, { "epoch": 0.4986642920747996, "ewc_loss": 0.08984375, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.976396083831787, "learning_rate": 1e-05, "loss": 0.5079, "mean_token_accuracy": 0.8593149781227112, "num_tokens": 148895458.0, "step": 3920 }, { "epoch": 0.49879150235339015, "ewc_loss": 0.09033203125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.642272472381592, "learning_rate": 1e-05, "loss": 0.5357, "mean_token_accuracy": 0.8497499823570251, "num_tokens": 148933305.0, "step": 3921 }, { "epoch": 0.4989187126319807, "ewc_loss": 0.08349609375, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 7.534027099609375e-05, "grad_norm": 5.608198165893555, "learning_rate": 1e-05, "loss": 0.481, "mean_token_accuracy": 0.8675903677940369, "num_tokens": 148970587.0, "step": 3922 }, { "epoch": 0.49904592291057115, "ewc_loss": 0.08837890625, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.612876892089844, "learning_rate": 1e-05, "loss": 0.541, "mean_token_accuracy": 0.8493458032608032, "num_tokens": 149004874.0, "step": 3923 }, { "epoch": 0.4991731331891617, "ewc_loss": 0.07958984375, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 7.152557373046875e-05, "grad_norm": 5.340017795562744, "learning_rate": 1e-05, "loss": 0.4875, "mean_token_accuracy": 0.8603681325912476, "num_tokens": 149045169.0, "step": 3924 }, { "epoch": 0.4993003434677522, "ewc_loss": 0.0869140625, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.625143527984619, "learning_rate": 1e-05, "loss": 0.6178, "mean_token_accuracy": 0.8239957690238953, "num_tokens": 149081554.0, "step": 3925 }, { "epoch": 0.4994275537463427, "ewc_loss": 0.07861328125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 7.009506225585938e-05, "grad_norm": 5.167652606964111, "learning_rate": 1e-05, "loss": 0.5436, "mean_token_accuracy": 0.8463793992996216, "num_tokens": 149124500.0, "step": 3926 }, { "epoch": 0.4995547640249332, "ewc_loss": 0.0859375, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.658623218536377, "learning_rate": 1e-05, "loss": 0.5418, "mean_token_accuracy": 0.848471462726593, "num_tokens": 149162522.0, "step": 3927 }, { "epoch": 0.49968197430352373, "ewc_loss": 0.07763671875, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 6.914138793945312e-05, "grad_norm": 5.049482822418213, "learning_rate": 1e-05, "loss": 0.5036, "mean_token_accuracy": 0.8575884103775024, "num_tokens": 149198083.0, "step": 3928 }, { "epoch": 0.4998091845821142, "ewc_loss": 0.087890625, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.76937198638916, "learning_rate": 1e-05, "loss": 0.5548, "mean_token_accuracy": 0.8419704437255859, "num_tokens": 149234786.0, "step": 3929 }, { "epoch": 0.49993639486070474, "ewc_loss": 0.07666015625, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 6.818771362304688e-05, "grad_norm": 4.9380784034729, "learning_rate": 1e-05, "loss": 0.4533, "mean_token_accuracy": 0.8709115982055664, "num_tokens": 149269859.0, "step": 3930 }, { "epoch": 0.5000636051392953, "ewc_loss": 0.091796875, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.980714797973633, "learning_rate": 1e-05, "loss": 0.5913, "mean_token_accuracy": 0.834475040435791, "num_tokens": 149311209.0, "step": 3931 }, { "epoch": 0.5001908154178858, "ewc_loss": 0.07666015625, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 6.818771362304688e-05, "grad_norm": 4.866064071655273, "learning_rate": 1e-05, "loss": 0.5227, "mean_token_accuracy": 0.849254846572876, "num_tokens": 149351283.0, "step": 3932 }, { "epoch": 0.5003180256964763, "ewc_loss": 0.09912109375, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.2822585105896, "learning_rate": 1e-05, "loss": 0.5336, "mean_token_accuracy": 0.8549541234970093, "num_tokens": 149391068.0, "step": 3933 }, { "epoch": 0.5004452359750667, "ewc_loss": 0.07861328125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 7.009506225585938e-05, "grad_norm": 4.933285713195801, "learning_rate": 1e-05, "loss": 0.5385, "mean_token_accuracy": 0.8465840816497803, "num_tokens": 149435078.0, "step": 3934 }, { "epoch": 0.5005724462536573, "ewc_loss": 0.10595703125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.56166934967041, "learning_rate": 1e-05, "loss": 0.589, "mean_token_accuracy": 0.842988908290863, "num_tokens": 149481336.0, "step": 3935 }, { "epoch": 0.5006996565322478, "ewc_loss": 0.08642578125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.251070976257324, "learning_rate": 1e-05, "loss": 0.5131, "mean_token_accuracy": 0.8553154468536377, "num_tokens": 149523278.0, "step": 3936 }, { "epoch": 0.5008268668108383, "ewc_loss": 0.1044921875, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.563223838806152, "learning_rate": 1e-05, "loss": 0.4902, "mean_token_accuracy": 0.8732081651687622, "num_tokens": 149563011.0, "step": 3937 }, { "epoch": 0.5009540770894289, "ewc_loss": 0.091796875, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.569633960723877, "learning_rate": 1e-05, "loss": 0.5299, "mean_token_accuracy": 0.8522878885269165, "num_tokens": 149606609.0, "step": 3938 }, { "epoch": 0.5010812873680194, "ewc_loss": 0.0986328125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.250291347503662, "learning_rate": 1e-05, "loss": 0.4923, "mean_token_accuracy": 0.8665615320205688, "num_tokens": 149647012.0, "step": 3939 }, { "epoch": 0.5012084976466098, "ewc_loss": 0.09228515625, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.663458824157715, "learning_rate": 1e-05, "loss": 0.5078, "mean_token_accuracy": 0.8586611151695251, "num_tokens": 149680153.0, "step": 3940 }, { "epoch": 0.5013357079252003, "ewc_loss": 0.0869140625, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.854989051818848, "learning_rate": 1e-05, "loss": 0.5004, "mean_token_accuracy": 0.8612987995147705, "num_tokens": 149717080.0, "step": 3941 }, { "epoch": 0.5014629182037909, "ewc_loss": 0.09033203125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.634395599365234, "learning_rate": 1e-05, "loss": 0.5173, "mean_token_accuracy": 0.8582639694213867, "num_tokens": 149755934.0, "step": 3942 }, { "epoch": 0.5015901284823814, "ewc_loss": 0.08251953125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 7.43865966796875e-05, "grad_norm": 5.525477886199951, "learning_rate": 1e-05, "loss": 0.5058, "mean_token_accuracy": 0.8551716804504395, "num_tokens": 149792714.0, "step": 3943 }, { "epoch": 0.5017173387609719, "ewc_loss": 0.0888671875, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.623775482177734, "learning_rate": 1e-05, "loss": 0.5482, "mean_token_accuracy": 0.8456435799598694, "num_tokens": 149828647.0, "step": 3944 }, { "epoch": 0.5018445490395624, "ewc_loss": 0.07958984375, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 7.152557373046875e-05, "grad_norm": 5.272580146789551, "learning_rate": 1e-05, "loss": 0.5172, "mean_token_accuracy": 0.8556647300720215, "num_tokens": 149864303.0, "step": 3945 }, { "epoch": 0.5019717593181529, "ewc_loss": 0.08740234375, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.6559014320373535, "learning_rate": 1e-05, "loss": 0.5655, "mean_token_accuracy": 0.8405646085739136, "num_tokens": 149902154.0, "step": 3946 }, { "epoch": 0.5020989695967434, "ewc_loss": 0.078125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 6.961822509765625e-05, "grad_norm": 5.115069389343262, "learning_rate": 1e-05, "loss": 0.4762, "mean_token_accuracy": 0.8615320324897766, "num_tokens": 149930902.0, "step": 3947 }, { "epoch": 0.5022261798753339, "ewc_loss": 0.08740234375, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.72367525100708, "learning_rate": 1e-05, "loss": 0.5276, "mean_token_accuracy": 0.8535354733467102, "num_tokens": 149972033.0, "step": 3948 }, { "epoch": 0.5023533901539244, "ewc_loss": 0.07763671875, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 6.914138793945312e-05, "grad_norm": 4.980408191680908, "learning_rate": 1e-05, "loss": 0.4823, "mean_token_accuracy": 0.8648250699043274, "num_tokens": 150007278.0, "step": 3949 }, { "epoch": 0.502480600432515, "ewc_loss": 0.08935546875, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.890243053436279, "learning_rate": 1e-05, "loss": 0.5354, "mean_token_accuracy": 0.8510074615478516, "num_tokens": 150054512.0, "step": 3950 }, { "epoch": 0.5026078107111055, "ewc_loss": 0.07666015625, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 6.818771362304688e-05, "grad_norm": 4.885854244232178, "learning_rate": 1e-05, "loss": 0.5149, "mean_token_accuracy": 0.8534412980079651, "num_tokens": 150094614.0, "step": 3951 }, { "epoch": 0.5027350209896959, "ewc_loss": 0.09716796875, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.174564361572266, "learning_rate": 1e-05, "loss": 0.5262, "mean_token_accuracy": 0.8560906052589417, "num_tokens": 150133011.0, "step": 3952 }, { "epoch": 0.5028622312682864, "ewc_loss": 0.0771484375, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 6.866455078125e-05, "grad_norm": 4.889842510223389, "learning_rate": 1e-05, "loss": 0.5156, "mean_token_accuracy": 0.8513167500495911, "num_tokens": 150170041.0, "step": 3953 }, { "epoch": 0.502989441546877, "ewc_loss": 0.10400390625, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.499618053436279, "learning_rate": 1e-05, "loss": 0.5431, "mean_token_accuracy": 0.8515787720680237, "num_tokens": 150206034.0, "step": 3954 }, { "epoch": 0.5031166518254675, "ewc_loss": 0.0849609375, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.132608413696289, "learning_rate": 1e-05, "loss": 0.5185, "mean_token_accuracy": 0.8544553518295288, "num_tokens": 150241601.0, "step": 3955 }, { "epoch": 0.503243862104058, "ewc_loss": 0.10595703125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.614838600158691, "learning_rate": 1e-05, "loss": 0.5098, "mean_token_accuracy": 0.8652544617652893, "num_tokens": 150279219.0, "step": 3956 }, { "epoch": 0.5033710723826486, "ewc_loss": 0.09033203125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.488126754760742, "learning_rate": 1e-05, "loss": 0.5306, "mean_token_accuracy": 0.8514125347137451, "num_tokens": 150323160.0, "step": 3957 }, { "epoch": 0.5034982826612391, "ewc_loss": 0.10205078125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.374279975891113, "learning_rate": 1e-05, "loss": 0.5243, "mean_token_accuracy": 0.8588017225265503, "num_tokens": 150363788.0, "step": 3958 }, { "epoch": 0.5036254929398295, "ewc_loss": 0.09375, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.669025421142578, "learning_rate": 1e-05, "loss": 0.5348, "mean_token_accuracy": 0.8498427271842957, "num_tokens": 150395481.0, "step": 3959 }, { "epoch": 0.50375270321842, "ewc_loss": 0.08935546875, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.988739490509033, "learning_rate": 1e-05, "loss": 0.5022, "mean_token_accuracy": 0.8597102165222168, "num_tokens": 150425726.0, "step": 3960 }, { "epoch": 0.5038799134970106, "ewc_loss": 0.0908203125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.659823894500732, "learning_rate": 1e-05, "loss": 0.5558, "mean_token_accuracy": 0.8495469689369202, "num_tokens": 150463073.0, "step": 3961 }, { "epoch": 0.5040071237756011, "ewc_loss": 0.08349609375, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 7.534027099609375e-05, "grad_norm": 5.6145806312561035, "learning_rate": 1e-05, "loss": 0.5533, "mean_token_accuracy": 0.8408524990081787, "num_tokens": 150502938.0, "step": 3962 }, { "epoch": 0.5041343340541916, "ewc_loss": 0.08837890625, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.637864589691162, "learning_rate": 1e-05, "loss": 0.5272, "mean_token_accuracy": 0.8541079163551331, "num_tokens": 150538601.0, "step": 3963 }, { "epoch": 0.5042615443327821, "ewc_loss": 0.07958984375, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 7.152557373046875e-05, "grad_norm": 5.3424906730651855, "learning_rate": 1e-05, "loss": 0.5261, "mean_token_accuracy": 0.8515263795852661, "num_tokens": 150573745.0, "step": 3964 }, { "epoch": 0.5043887546113726, "ewc_loss": 0.08642578125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.639309883117676, "learning_rate": 1e-05, "loss": 0.5565, "mean_token_accuracy": 0.8426921367645264, "num_tokens": 150613029.0, "step": 3965 }, { "epoch": 0.5045159648899631, "ewc_loss": 0.07861328125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 7.009506225585938e-05, "grad_norm": 5.170322895050049, "learning_rate": 1e-05, "loss": 0.4911, "mean_token_accuracy": 0.8613801002502441, "num_tokens": 150649773.0, "step": 3966 }, { "epoch": 0.5046431751685536, "ewc_loss": 0.0869140625, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.6769304275512695, "learning_rate": 1e-05, "loss": 0.5077, "mean_token_accuracy": 0.8575991988182068, "num_tokens": 150694764.0, "step": 3967 }, { "epoch": 0.5047703854471441, "ewc_loss": 0.078125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 6.961822509765625e-05, "grad_norm": 5.0419230461120605, "learning_rate": 1e-05, "loss": 0.5325, "mean_token_accuracy": 0.8478506207466125, "num_tokens": 150735233.0, "step": 3968 }, { "epoch": 0.5048975957257347, "ewc_loss": 0.08837890625, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.806543350219727, "learning_rate": 1e-05, "loss": 0.4619, "mean_token_accuracy": 0.8754013180732727, "num_tokens": 150769596.0, "step": 3969 }, { "epoch": 0.5050248060043252, "ewc_loss": 0.0771484375, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 6.866455078125e-05, "grad_norm": 4.950239658355713, "learning_rate": 1e-05, "loss": 0.5335, "mean_token_accuracy": 0.8480792045593262, "num_tokens": 150806096.0, "step": 3970 }, { "epoch": 0.5051520162829156, "ewc_loss": 0.09228515625, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 6.009735107421875, "learning_rate": 1e-05, "loss": 0.5204, "mean_token_accuracy": 0.8548405766487122, "num_tokens": 150844721.0, "step": 3971 }, { "epoch": 0.5052792265615061, "ewc_loss": 0.07666015625, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 6.818771362304688e-05, "grad_norm": 4.880165100097656, "learning_rate": 1e-05, "loss": 0.4973, "mean_token_accuracy": 0.8571300506591797, "num_tokens": 150885711.0, "step": 3972 }, { "epoch": 0.5054064368400967, "ewc_loss": 0.10009765625, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.312839508056641, "learning_rate": 1e-05, "loss": 0.5417, "mean_token_accuracy": 0.8520027995109558, "num_tokens": 150924993.0, "step": 3973 }, { "epoch": 0.5055336471186872, "ewc_loss": 0.0791015625, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 7.05718994140625e-05, "grad_norm": 4.975335597991943, "learning_rate": 1e-05, "loss": 0.481, "mean_token_accuracy": 0.8662693500518799, "num_tokens": 150956139.0, "step": 3974 }, { "epoch": 0.5056608573972777, "ewc_loss": 0.10546875, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.564958572387695, "learning_rate": 1e-05, "loss": 0.6096, "mean_token_accuracy": 0.8357991576194763, "num_tokens": 150995816.0, "step": 3975 }, { "epoch": 0.5057880676758683, "ewc_loss": 0.08740234375, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.280776023864746, "learning_rate": 1e-05, "loss": 0.5667, "mean_token_accuracy": 0.8383787870407104, "num_tokens": 151035191.0, "step": 3976 }, { "epoch": 0.5059152779544587, "ewc_loss": 0.1044921875, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.534480094909668, "learning_rate": 1e-05, "loss": 0.5902, "mean_token_accuracy": 0.8347018361091614, "num_tokens": 151069625.0, "step": 3977 }, { "epoch": 0.5060424882330492, "ewc_loss": 0.091796875, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.571018695831299, "learning_rate": 1e-05, "loss": 0.5649, "mean_token_accuracy": 0.8447317481040955, "num_tokens": 151105999.0, "step": 3978 }, { "epoch": 0.5061696985116397, "ewc_loss": 0.0986328125, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.212846279144287, "learning_rate": 1e-05, "loss": 0.5283, "mean_token_accuracy": 0.8547146320343018, "num_tokens": 151148712.0, "step": 3979 }, { "epoch": 0.5062969087902303, "ewc_loss": 0.09130859375, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.644581317901611, "learning_rate": 1e-05, "loss": 0.4797, "mean_token_accuracy": 0.8684005737304688, "num_tokens": 151182842.0, "step": 3980 }, { "epoch": 0.5064241190688208, "ewc_loss": 0.08740234375, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.83314847946167, "learning_rate": 1e-05, "loss": 0.5262, "mean_token_accuracy": 0.8526920676231384, "num_tokens": 151223731.0, "step": 3981 }, { "epoch": 0.5065513293474113, "ewc_loss": 0.08935546875, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.6263556480407715, "learning_rate": 1e-05, "loss": 0.5332, "mean_token_accuracy": 0.8506117463111877, "num_tokens": 151259711.0, "step": 3982 }, { "epoch": 0.5066785396260017, "ewc_loss": 0.08251953125, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 7.43865966796875e-05, "grad_norm": 5.503900527954102, "learning_rate": 1e-05, "loss": 0.5427, "mean_token_accuracy": 0.8479167819023132, "num_tokens": 151301486.0, "step": 3983 }, { "epoch": 0.5068057499045923, "ewc_loss": 0.0888671875, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.623938083648682, "learning_rate": 1e-05, "loss": 0.5184, "mean_token_accuracy": 0.8536298274993896, "num_tokens": 151339068.0, "step": 3984 }, { "epoch": 0.5069329601831828, "ewc_loss": 0.07958984375, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 7.152557373046875e-05, "grad_norm": 5.264984130859375, "learning_rate": 1e-05, "loss": 0.546, "mean_token_accuracy": 0.844623327255249, "num_tokens": 151380694.0, "step": 3985 }, { "epoch": 0.5070601704617733, "ewc_loss": 0.08740234375, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.650257587432861, "learning_rate": 1e-05, "loss": 0.5197, "mean_token_accuracy": 0.8573901653289795, "num_tokens": 151416522.0, "step": 3986 }, { "epoch": 0.5071873807403638, "ewc_loss": 0.0791015625, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 7.05718994140625e-05, "grad_norm": 5.113279342651367, "learning_rate": 1e-05, "loss": 0.5669, "mean_token_accuracy": 0.8385618329048157, "num_tokens": 151453786.0, "step": 3987 }, { "epoch": 0.5073145910189544, "ewc_loss": 0.087890625, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.747968673706055, "learning_rate": 1e-05, "loss": 0.5344, "mean_token_accuracy": 0.8536044359207153, "num_tokens": 151492155.0, "step": 3988 }, { "epoch": 0.5074418012975448, "ewc_loss": 0.07763671875, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 6.914138793945312e-05, "grad_norm": 4.977928638458252, "learning_rate": 1e-05, "loss": 0.4788, "mean_token_accuracy": 0.861426830291748, "num_tokens": 151534979.0, "step": 3989 }, { "epoch": 0.5075690115761353, "ewc_loss": 0.08984375, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.93008279800415, "learning_rate": 1e-05, "loss": 0.5548, "mean_token_accuracy": 0.8498245477676392, "num_tokens": 151573670.0, "step": 3990 }, { "epoch": 0.5076962218547258, "ewc_loss": 0.07666015625, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 6.818771362304688e-05, "grad_norm": 4.883454322814941, "learning_rate": 1e-05, "loss": 0.4295, "mean_token_accuracy": 0.879546582698822, "num_tokens": 151605232.0, "step": 3991 }, { "epoch": 0.5078234321333164, "ewc_loss": 0.09912109375, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.234367847442627, "learning_rate": 1e-05, "loss": 0.5884, "mean_token_accuracy": 0.8365919589996338, "num_tokens": 151648696.0, "step": 3992 }, { "epoch": 0.5079506424119069, "ewc_loss": 0.07763671875, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 6.914138793945312e-05, "grad_norm": 4.91262149810791, "learning_rate": 1e-05, "loss": 0.4902, "mean_token_accuracy": 0.8638147115707397, "num_tokens": 151684203.0, "step": 3993 }, { "epoch": 0.5080778526904974, "ewc_loss": 0.1044921875, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.544319152832031, "learning_rate": 1e-05, "loss": 0.5618, "mean_token_accuracy": 0.8475674390792847, "num_tokens": 151716749.0, "step": 3994 }, { "epoch": 0.5082050629690879, "ewc_loss": 0.0859375, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.181955814361572, "learning_rate": 1e-05, "loss": 0.5591, "mean_token_accuracy": 0.8433749675750732, "num_tokens": 151752567.0, "step": 3995 }, { "epoch": 0.5083322732476784, "ewc_loss": 0.1064453125, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.602911949157715, "learning_rate": 1e-05, "loss": 0.4796, "mean_token_accuracy": 0.8763157725334167, "num_tokens": 151794201.0, "step": 3996 }, { "epoch": 0.5084594835262689, "ewc_loss": 0.09130859375, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.526211738586426, "learning_rate": 1e-05, "loss": 0.5149, "mean_token_accuracy": 0.8594216704368591, "num_tokens": 151833731.0, "step": 3997 }, { "epoch": 0.5085866938048594, "ewc_loss": 0.10107421875, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.344882011413574, "learning_rate": 1e-05, "loss": 0.5327, "mean_token_accuracy": 0.8567585945129395, "num_tokens": 151872738.0, "step": 3998 }, { "epoch": 0.50871390408345, "ewc_loss": 0.0927734375, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.667115688323975, "learning_rate": 1e-05, "loss": 0.4955, "mean_token_accuracy": 0.8631090521812439, "num_tokens": 151916801.0, "step": 3999 }, { "epoch": 0.5088411143620405, "ewc_loss": 0.0888671875, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.968830108642578, "learning_rate": 1e-05, "loss": 0.5369, "mean_token_accuracy": 0.8515501022338867, "num_tokens": 151949661.0, "step": 4000 }, { "epoch": 0.5089683246406309, "ewc_loss": 0.0908203125, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.654013633728027, "learning_rate": 1e-05, "loss": 0.5432, "mean_token_accuracy": 0.8473640084266663, "num_tokens": 151986719.0, "step": 4001 }, { "epoch": 0.5090955349192214, "ewc_loss": 0.08349609375, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 7.534027099609375e-05, "grad_norm": 5.626763343811035, "learning_rate": 1e-05, "loss": 0.5323, "mean_token_accuracy": 0.8505339622497559, "num_tokens": 152023921.0, "step": 4002 }, { "epoch": 0.509222745197812, "ewc_loss": 0.08837890625, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.623780250549316, "learning_rate": 1e-05, "loss": 0.4999, "mean_token_accuracy": 0.8625248670578003, "num_tokens": 152061246.0, "step": 4003 }, { "epoch": 0.5093499554764025, "ewc_loss": 0.08056640625, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 7.2479248046875e-05, "grad_norm": 5.3606181144714355, "learning_rate": 1e-05, "loss": 0.5376, "mean_token_accuracy": 0.8449516892433167, "num_tokens": 152103299.0, "step": 4004 }, { "epoch": 0.509477165754993, "ewc_loss": 0.0869140625, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.61263370513916, "learning_rate": 1e-05, "loss": 0.5589, "mean_token_accuracy": 0.8445457220077515, "num_tokens": 152148576.0, "step": 4005 }, { "epoch": 0.5096043760335836, "ewc_loss": 0.07861328125, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 7.009506225585938e-05, "grad_norm": 5.204277038574219, "learning_rate": 1e-05, "loss": 0.5141, "mean_token_accuracy": 0.8539014458656311, "num_tokens": 152189221.0, "step": 4006 }, { "epoch": 0.5097315863121741, "ewc_loss": 0.0859375, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.6386590003967285, "learning_rate": 1e-05, "loss": 0.5332, "mean_token_accuracy": 0.8504608869552612, "num_tokens": 152232732.0, "step": 4007 }, { "epoch": 0.5098587965907645, "ewc_loss": 0.078125, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 6.961822509765625e-05, "grad_norm": 5.092210292816162, "learning_rate": 1e-05, "loss": 0.4972, "mean_token_accuracy": 0.8599080443382263, "num_tokens": 152268715.0, "step": 4008 }, { "epoch": 0.509986006869355, "ewc_loss": 0.087890625, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.72452449798584, "learning_rate": 1e-05, "loss": 0.5443, "mean_token_accuracy": 0.8474030494689941, "num_tokens": 152311141.0, "step": 4009 }, { "epoch": 0.5101132171479456, "ewc_loss": 0.0771484375, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 6.866455078125e-05, "grad_norm": 4.97810173034668, "learning_rate": 1e-05, "loss": 0.4705, "mean_token_accuracy": 0.8697516322135925, "num_tokens": 152348892.0, "step": 4010 }, { "epoch": 0.5102404274265361, "ewc_loss": 0.08984375, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.91391134262085, "learning_rate": 1e-05, "loss": 0.5173, "mean_token_accuracy": 0.8582033514976501, "num_tokens": 152383300.0, "step": 4011 }, { "epoch": 0.5103676377051266, "ewc_loss": 0.076171875, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 6.771087646484375e-05, "grad_norm": 4.867910861968994, "learning_rate": 1e-05, "loss": 0.5273, "mean_token_accuracy": 0.8506032824516296, "num_tokens": 152424404.0, "step": 4012 }, { "epoch": 0.5104948479837171, "ewc_loss": 0.0986328125, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.232010841369629, "learning_rate": 1e-05, "loss": 0.4815, "mean_token_accuracy": 0.871242880821228, "num_tokens": 152465226.0, "step": 4013 }, { "epoch": 0.5106220582623076, "ewc_loss": 0.07763671875, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 6.914138793945312e-05, "grad_norm": 4.873594284057617, "learning_rate": 1e-05, "loss": 0.5439, "mean_token_accuracy": 0.8477990627288818, "num_tokens": 152507819.0, "step": 4014 }, { "epoch": 0.5107492685408981, "ewc_loss": 0.10498046875, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.589463233947754, "learning_rate": 1e-05, "loss": 0.4747, "mean_token_accuracy": 0.8772971630096436, "num_tokens": 152548194.0, "step": 4015 }, { "epoch": 0.5108764788194886, "ewc_loss": 0.0859375, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.177471160888672, "learning_rate": 1e-05, "loss": 0.4982, "mean_token_accuracy": 0.8572502136230469, "num_tokens": 152580255.0, "step": 4016 }, { "epoch": 0.5110036890980791, "ewc_loss": 0.1083984375, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.699536323547363, "learning_rate": 1e-05, "loss": 0.5447, "mean_token_accuracy": 0.8535083532333374, "num_tokens": 152619813.0, "step": 4017 }, { "epoch": 0.5111308993766697, "ewc_loss": 0.0927734375, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.584016799926758, "learning_rate": 1e-05, "loss": 0.5619, "mean_token_accuracy": 0.8467658758163452, "num_tokens": 152655552.0, "step": 4018 }, { "epoch": 0.5112581096552602, "ewc_loss": 0.1015625, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.417661666870117, "learning_rate": 1e-05, "loss": 0.5813, "mean_token_accuracy": 0.8420203924179077, "num_tokens": 152693695.0, "step": 4019 }, { "epoch": 0.5113853199338506, "ewc_loss": 0.09521484375, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.727863788604736, "learning_rate": 1e-05, "loss": 0.5349, "mean_token_accuracy": 0.8526769876480103, "num_tokens": 152732917.0, "step": 4020 }, { "epoch": 0.5115125302124411, "ewc_loss": 0.08984375, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 6.015261173248291, "learning_rate": 1e-05, "loss": 0.493, "mean_token_accuracy": 0.8626254200935364, "num_tokens": 152768052.0, "step": 4021 }, { "epoch": 0.5116397404910317, "ewc_loss": 0.0908203125, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.701125144958496, "learning_rate": 1e-05, "loss": 0.5737, "mean_token_accuracy": 0.8408389091491699, "num_tokens": 152808106.0, "step": 4022 }, { "epoch": 0.5117669507696222, "ewc_loss": 0.08447265625, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 7.62939453125e-05, "grad_norm": 5.637417316436768, "learning_rate": 1e-05, "loss": 0.4481, "mean_token_accuracy": 0.8779563903808594, "num_tokens": 152849989.0, "step": 4023 }, { "epoch": 0.5118941610482127, "ewc_loss": 0.08837890625, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.653829574584961, "learning_rate": 1e-05, "loss": 0.5028, "mean_token_accuracy": 0.8606312870979309, "num_tokens": 152886599.0, "step": 4024 }, { "epoch": 0.5120213713268033, "ewc_loss": 0.08154296875, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 7.343292236328125e-05, "grad_norm": 5.376913547515869, "learning_rate": 1e-05, "loss": 0.5472, "mean_token_accuracy": 0.8478345274925232, "num_tokens": 152926383.0, "step": 4025 }, { "epoch": 0.5121485816053937, "ewc_loss": 0.0869140625, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.654850482940674, "learning_rate": 1e-05, "loss": 0.5436, "mean_token_accuracy": 0.8486146330833435, "num_tokens": 152966141.0, "step": 4026 }, { "epoch": 0.5122757918839842, "ewc_loss": 0.0791015625, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 7.05718994140625e-05, "grad_norm": 5.202362060546875, "learning_rate": 1e-05, "loss": 0.5027, "mean_token_accuracy": 0.8604060411453247, "num_tokens": 152998217.0, "step": 4027 }, { "epoch": 0.5124030021625747, "ewc_loss": 0.0869140625, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.713616847991943, "learning_rate": 1e-05, "loss": 0.5872, "mean_token_accuracy": 0.8401632308959961, "num_tokens": 153032258.0, "step": 4028 }, { "epoch": 0.5125302124411653, "ewc_loss": 0.078125, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 6.961822509765625e-05, "grad_norm": 5.0538434982299805, "learning_rate": 1e-05, "loss": 0.4996, "mean_token_accuracy": 0.8578091859817505, "num_tokens": 153075762.0, "step": 4029 }, { "epoch": 0.5126574227197558, "ewc_loss": 0.08837890625, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.829797267913818, "learning_rate": 1e-05, "loss": 0.5461, "mean_token_accuracy": 0.8502436876296997, "num_tokens": 153112571.0, "step": 4030 }, { "epoch": 0.5127846329983463, "ewc_loss": 0.0771484375, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 6.866455078125e-05, "grad_norm": 4.935258865356445, "learning_rate": 1e-05, "loss": 0.4834, "mean_token_accuracy": 0.8640601634979248, "num_tokens": 153153348.0, "step": 4031 }, { "epoch": 0.5129118432769367, "ewc_loss": 0.09423828125, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 6.070817470550537, "learning_rate": 1e-05, "loss": 0.4944, "mean_token_accuracy": 0.8647000193595886, "num_tokens": 153189616.0, "step": 4032 }, { "epoch": 0.5130390535555273, "ewc_loss": 0.07666015625, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 6.818771362304688e-05, "grad_norm": 4.906257629394531, "learning_rate": 1e-05, "loss": 0.5721, "mean_token_accuracy": 0.839940071105957, "num_tokens": 153230879.0, "step": 4033 }, { "epoch": 0.5131662638341178, "ewc_loss": 0.1005859375, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.359428882598877, "learning_rate": 1e-05, "loss": 0.5531, "mean_token_accuracy": 0.8496062755584717, "num_tokens": 153275234.0, "step": 4034 }, { "epoch": 0.5132934741127083, "ewc_loss": 0.08203125, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 7.390975952148438e-05, "grad_norm": 5.004389762878418, "learning_rate": 1e-05, "loss": 0.4775, "mean_token_accuracy": 0.8657063245773315, "num_tokens": 153312606.0, "step": 4035 }, { "epoch": 0.5134206843912988, "ewc_loss": 0.1064453125, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.595507621765137, "learning_rate": 1e-05, "loss": 0.571, "mean_token_accuracy": 0.8445396423339844, "num_tokens": 153352370.0, "step": 4036 }, { "epoch": 0.5135478946698894, "ewc_loss": 0.08837890625, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.337185382843018, "learning_rate": 1e-05, "loss": 0.517, "mean_token_accuracy": 0.8570517897605896, "num_tokens": 153393030.0, "step": 4037 }, { "epoch": 0.5136751049484798, "ewc_loss": 0.1044921875, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.535800457000732, "learning_rate": 1e-05, "loss": 0.5536, "mean_token_accuracy": 0.8522964715957642, "num_tokens": 153423831.0, "step": 4038 }, { "epoch": 0.5138023152270703, "ewc_loss": 0.09326171875, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.61814546585083, "learning_rate": 1e-05, "loss": 0.4895, "mean_token_accuracy": 0.8664606809616089, "num_tokens": 153458388.0, "step": 4039 }, { "epoch": 0.5139295255056608, "ewc_loss": 0.09765625, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 6.190773963928223, "learning_rate": 1e-05, "loss": 0.5598, "mean_token_accuracy": 0.847312331199646, "num_tokens": 153503486.0, "step": 4040 }, { "epoch": 0.5140567357842514, "ewc_loss": 0.09228515625, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.675319671630859, "learning_rate": 1e-05, "loss": 0.519, "mean_token_accuracy": 0.8589856624603271, "num_tokens": 153541474.0, "step": 4041 }, { "epoch": 0.5141839460628419, "ewc_loss": 0.0869140625, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.796440124511719, "learning_rate": 1e-05, "loss": 0.5621, "mean_token_accuracy": 0.8453176021575928, "num_tokens": 153583518.0, "step": 4042 }, { "epoch": 0.5143111563414324, "ewc_loss": 0.08935546875, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.657220840454102, "learning_rate": 1e-05, "loss": 0.5332, "mean_token_accuracy": 0.8532750010490417, "num_tokens": 153617552.0, "step": 4043 }, { "epoch": 0.5144383666200228, "ewc_loss": 0.0830078125, "ewc_loss_diag": 8.535385131835938e-05, "ewc_loss_parallel": 7.43865966796875e-05, "grad_norm": 5.492482662200928, "learning_rate": 1e-05, "loss": 0.5687, "mean_token_accuracy": 0.8424422740936279, "num_tokens": 153655616.0, "step": 4044 }, { "epoch": 0.5145655768986134, "ewc_loss": 0.087890625, "ewc_loss_diag": 8.535385131835938e-05, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.620153903961182, "learning_rate": 1e-05, "loss": 0.5008, "mean_token_accuracy": 0.8609168529510498, "num_tokens": 153692835.0, "step": 4045 }, { "epoch": 0.5146927871772039, "ewc_loss": 0.080078125, "ewc_loss_diag": 8.535385131835938e-05, "ewc_loss_parallel": 7.200241088867188e-05, "grad_norm": 5.29185152053833, "learning_rate": 1e-05, "loss": 0.5298, "mean_token_accuracy": 0.8506210446357727, "num_tokens": 153720867.0, "step": 4046 }, { "epoch": 0.5148199974557944, "ewc_loss": 0.0859375, "ewc_loss_diag": 8.535385131835938e-05, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.625024318695068, "learning_rate": 1e-05, "loss": 0.505, "mean_token_accuracy": 0.8600205183029175, "num_tokens": 153758798.0, "step": 4047 }, { "epoch": 0.514947207734385, "ewc_loss": 0.0791015625, "ewc_loss_diag": 8.535385131835938e-05, "ewc_loss_parallel": 7.009506225585938e-05, "grad_norm": 5.148989200592041, "learning_rate": 1e-05, "loss": 0.5754, "mean_token_accuracy": 0.8396885395050049, "num_tokens": 153798257.0, "step": 4048 }, { "epoch": 0.5150744180129755, "ewc_loss": 0.0869140625, "ewc_loss_diag": 8.535385131835938e-05, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.677075386047363, "learning_rate": 1e-05, "loss": 0.4793, "mean_token_accuracy": 0.8664301633834839, "num_tokens": 153839241.0, "step": 4049 }, { "epoch": 0.5152016282915659, "ewc_loss": 0.078125, "ewc_loss_diag": 8.535385131835938e-05, "ewc_loss_parallel": 6.914138793945312e-05, "grad_norm": 5.027943134307861, "learning_rate": 1e-05, "loss": 0.5119, "mean_token_accuracy": 0.8543423414230347, "num_tokens": 153882522.0, "step": 4050 }, { "epoch": 0.5153288385701564, "ewc_loss": 0.0888671875, "ewc_loss_diag": 8.535385131835938e-05, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.820231914520264, "learning_rate": 1e-05, "loss": 0.4686, "mean_token_accuracy": 0.8666831254959106, "num_tokens": 153912173.0, "step": 4051 }, { "epoch": 0.515456048848747, "ewc_loss": 0.0771484375, "ewc_loss_diag": 8.535385131835938e-05, "ewc_loss_parallel": 6.818771362304688e-05, "grad_norm": 4.92455530166626, "learning_rate": 1e-05, "loss": 0.6215, "mean_token_accuracy": 0.8200375437736511, "num_tokens": 153955070.0, "step": 4052 }, { "epoch": 0.5155832591273375, "ewc_loss": 0.09375, "ewc_loss_diag": 8.535385131835938e-05, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 6.06687068939209, "learning_rate": 1e-05, "loss": 0.4888, "mean_token_accuracy": 0.8598505258560181, "num_tokens": 153995526.0, "step": 4053 }, { "epoch": 0.515710469405928, "ewc_loss": 0.0771484375, "ewc_loss_diag": 8.535385131835938e-05, "ewc_loss_parallel": 6.818771362304688e-05, "grad_norm": 4.861621379852295, "learning_rate": 1e-05, "loss": 0.5172, "mean_token_accuracy": 0.8530640602111816, "num_tokens": 154035476.0, "step": 4054 }, { "epoch": 0.5158376796845185, "ewc_loss": 0.1025390625, "ewc_loss_diag": 8.535385131835938e-05, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.409450054168701, "learning_rate": 1e-05, "loss": 0.5627, "mean_token_accuracy": 0.8473508358001709, "num_tokens": 154079108.0, "step": 4055 }, { "epoch": 0.5159648899631091, "ewc_loss": 0.08203125, "ewc_loss_diag": 8.535385131835938e-05, "ewc_loss_parallel": 7.343292236328125e-05, "grad_norm": 5.002906322479248, "learning_rate": 1e-05, "loss": 0.6087, "mean_token_accuracy": 0.8278835415840149, "num_tokens": 154118484.0, "step": 4056 }, { "epoch": 0.5160921002416995, "ewc_loss": 0.109375, "ewc_loss_diag": 8.535385131835938e-05, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.655370235443115, "learning_rate": 1e-05, "loss": 0.54, "mean_token_accuracy": 0.8556033372879028, "num_tokens": 154163981.0, "step": 4057 }, { "epoch": 0.51621931052029, "ewc_loss": 0.08984375, "ewc_loss_diag": 8.535385131835938e-05, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.371023178100586, "learning_rate": 1e-05, "loss": 0.5723, "mean_token_accuracy": 0.8411399126052856, "num_tokens": 154205035.0, "step": 4058 }, { "epoch": 0.5163465207988805, "ewc_loss": 0.10595703125, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.568354606628418, "learning_rate": 1e-05, "loss": 0.506, "mean_token_accuracy": 0.8642066121101379, "num_tokens": 154241243.0, "step": 4059 }, { "epoch": 0.5164737310774711, "ewc_loss": 0.09521484375, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.6542840003967285, "learning_rate": 1e-05, "loss": 0.5779, "mean_token_accuracy": 0.8402454257011414, "num_tokens": 154289182.0, "step": 4060 }, { "epoch": 0.5166009413560616, "ewc_loss": 0.099609375, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.222902297973633, "learning_rate": 1e-05, "loss": 0.5664, "mean_token_accuracy": 0.8443095088005066, "num_tokens": 154326735.0, "step": 4061 }, { "epoch": 0.5167281516346521, "ewc_loss": 0.0947265625, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.696957588195801, "learning_rate": 1e-05, "loss": 0.5336, "mean_token_accuracy": 0.8553998470306396, "num_tokens": 154367795.0, "step": 4062 }, { "epoch": 0.5168553619132426, "ewc_loss": 0.0869140625, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.831574440002441, "learning_rate": 1e-05, "loss": 0.4689, "mean_token_accuracy": 0.8697793483734131, "num_tokens": 154401949.0, "step": 4063 }, { "epoch": 0.5169825721918331, "ewc_loss": 0.08984375, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.657848358154297, "learning_rate": 1e-05, "loss": 0.4929, "mean_token_accuracy": 0.8650771379470825, "num_tokens": 154442693.0, "step": 4064 }, { "epoch": 0.5171097824704236, "ewc_loss": 0.08251953125, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 7.390975952148438e-05, "grad_norm": 5.500275135040283, "learning_rate": 1e-05, "loss": 0.5094, "mean_token_accuracy": 0.8558001518249512, "num_tokens": 154489219.0, "step": 4065 }, { "epoch": 0.5172369927490141, "ewc_loss": 0.08935546875, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.640381813049316, "learning_rate": 1e-05, "loss": 0.4802, "mean_token_accuracy": 0.8696929812431335, "num_tokens": 154526639.0, "step": 4066 }, { "epoch": 0.5173642030276047, "ewc_loss": 0.080078125, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 7.104873657226562e-05, "grad_norm": 5.299081325531006, "learning_rate": 1e-05, "loss": 0.4804, "mean_token_accuracy": 0.8674788475036621, "num_tokens": 154557339.0, "step": 4067 }, { "epoch": 0.5174914133061952, "ewc_loss": 0.08642578125, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.648869514465332, "learning_rate": 1e-05, "loss": 0.5382, "mean_token_accuracy": 0.8497635126113892, "num_tokens": 154592738.0, "step": 4068 }, { "epoch": 0.5176186235847856, "ewc_loss": 0.0791015625, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 7.009506225585938e-05, "grad_norm": 5.13772439956665, "learning_rate": 1e-05, "loss": 0.5172, "mean_token_accuracy": 0.8539791107177734, "num_tokens": 154631633.0, "step": 4069 }, { "epoch": 0.5177458338633761, "ewc_loss": 0.087890625, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.720459461212158, "learning_rate": 1e-05, "loss": 0.6018, "mean_token_accuracy": 0.831311821937561, "num_tokens": 154676274.0, "step": 4070 }, { "epoch": 0.5178730441419667, "ewc_loss": 0.07763671875, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 6.866455078125e-05, "grad_norm": 5.005577564239502, "learning_rate": 1e-05, "loss": 0.5526, "mean_token_accuracy": 0.8389967679977417, "num_tokens": 154716379.0, "step": 4071 }, { "epoch": 0.5180002544205572, "ewc_loss": 0.0908203125, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.9046173095703125, "learning_rate": 1e-05, "loss": 0.5419, "mean_token_accuracy": 0.8493148684501648, "num_tokens": 154757038.0, "step": 4072 }, { "epoch": 0.5181274646991477, "ewc_loss": 0.07666015625, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 6.771087646484375e-05, "grad_norm": 4.888316631317139, "learning_rate": 1e-05, "loss": 0.4363, "mean_token_accuracy": 0.8795782923698425, "num_tokens": 154795684.0, "step": 4073 }, { "epoch": 0.5182546749777382, "ewc_loss": 0.0986328125, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.206231117248535, "learning_rate": 1e-05, "loss": 0.5136, "mean_token_accuracy": 0.8575341701507568, "num_tokens": 154834858.0, "step": 4074 }, { "epoch": 0.5183818852563287, "ewc_loss": 0.07763671875, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 6.866455078125e-05, "grad_norm": 4.9024577140808105, "learning_rate": 1e-05, "loss": 0.5007, "mean_token_accuracy": 0.8578550815582275, "num_tokens": 154868919.0, "step": 4075 }, { "epoch": 0.5185090955349192, "ewc_loss": 0.10595703125, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.544508934020996, "learning_rate": 1e-05, "loss": 0.558, "mean_token_accuracy": 0.8516495227813721, "num_tokens": 154910511.0, "step": 4076 }, { "epoch": 0.5186363058135097, "ewc_loss": 0.08544921875, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.1399617195129395, "learning_rate": 1e-05, "loss": 0.5688, "mean_token_accuracy": 0.8400148749351501, "num_tokens": 154955345.0, "step": 4077 }, { "epoch": 0.5187635160921003, "ewc_loss": 0.109375, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.6848883628845215, "learning_rate": 1e-05, "loss": 0.5355, "mean_token_accuracy": 0.860108494758606, "num_tokens": 154992541.0, "step": 4078 }, { "epoch": 0.5188907263706908, "ewc_loss": 0.0927734375, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.535953998565674, "learning_rate": 1e-05, "loss": 0.5233, "mean_token_accuracy": 0.8543837666511536, "num_tokens": 155033592.0, "step": 4079 }, { "epoch": 0.5190179366492813, "ewc_loss": 0.10400390625, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.444578647613525, "learning_rate": 1e-05, "loss": 0.5321, "mean_token_accuracy": 0.8563933372497559, "num_tokens": 155080034.0, "step": 4080 }, { "epoch": 0.5191451469278717, "ewc_loss": 0.095703125, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.7235798835754395, "learning_rate": 1e-05, "loss": 0.5615, "mean_token_accuracy": 0.8428044319152832, "num_tokens": 155116207.0, "step": 4081 }, { "epoch": 0.5192723572064623, "ewc_loss": 0.091796875, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 6.045016765594482, "learning_rate": 1e-05, "loss": 0.5266, "mean_token_accuracy": 0.8513350486755371, "num_tokens": 155155787.0, "step": 4082 }, { "epoch": 0.5193995674850528, "ewc_loss": 0.09130859375, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.706319332122803, "learning_rate": 1e-05, "loss": 0.5134, "mean_token_accuracy": 0.861060380935669, "num_tokens": 155194021.0, "step": 4083 }, { "epoch": 0.5195267777636433, "ewc_loss": 0.08544921875, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.665952205657959, "learning_rate": 1e-05, "loss": 0.5031, "mean_token_accuracy": 0.8589754700660706, "num_tokens": 155228685.0, "step": 4084 }, { "epoch": 0.5196539880422338, "ewc_loss": 0.08984375, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.661392688751221, "learning_rate": 1e-05, "loss": 0.535, "mean_token_accuracy": 0.8534740805625916, "num_tokens": 155269146.0, "step": 4085 }, { "epoch": 0.5197811983208244, "ewc_loss": 0.08154296875, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 7.295608520507812e-05, "grad_norm": 5.382941246032715, "learning_rate": 1e-05, "loss": 0.5196, "mean_token_accuracy": 0.8546124696731567, "num_tokens": 155309080.0, "step": 4086 }, { "epoch": 0.5199084085994148, "ewc_loss": 0.087890625, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.656743049621582, "learning_rate": 1e-05, "loss": 0.5106, "mean_token_accuracy": 0.8564919233322144, "num_tokens": 155348661.0, "step": 4087 }, { "epoch": 0.5200356188780053, "ewc_loss": 0.0791015625, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 7.009506225585938e-05, "grad_norm": 5.2083964347839355, "learning_rate": 1e-05, "loss": 0.5326, "mean_token_accuracy": 0.8479114174842834, "num_tokens": 155384984.0, "step": 4088 }, { "epoch": 0.5201628291565958, "ewc_loss": 0.08837890625, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.7074432373046875, "learning_rate": 1e-05, "loss": 0.5439, "mean_token_accuracy": 0.8473691940307617, "num_tokens": 155421914.0, "step": 4089 }, { "epoch": 0.5202900394351864, "ewc_loss": 0.07861328125, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 6.961822509765625e-05, "grad_norm": 5.069125652313232, "learning_rate": 1e-05, "loss": 0.5021, "mean_token_accuracy": 0.8601625561714172, "num_tokens": 155466460.0, "step": 4090 }, { "epoch": 0.5204172497137769, "ewc_loss": 0.0888671875, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.82703161239624, "learning_rate": 1e-05, "loss": 0.518, "mean_token_accuracy": 0.8594400882720947, "num_tokens": 155504034.0, "step": 4091 }, { "epoch": 0.5205444599923674, "ewc_loss": 0.07763671875, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 6.866455078125e-05, "grad_norm": 4.96193265914917, "learning_rate": 1e-05, "loss": 0.5584, "mean_token_accuracy": 0.8427962064743042, "num_tokens": 155542521.0, "step": 4092 }, { "epoch": 0.5206716702709578, "ewc_loss": 0.09375, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 6.043832302093506, "learning_rate": 1e-05, "loss": 0.5096, "mean_token_accuracy": 0.8600531816482544, "num_tokens": 155574170.0, "step": 4093 }, { "epoch": 0.5207988805495484, "ewc_loss": 0.0771484375, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 6.818771362304688e-05, "grad_norm": 4.909340858459473, "learning_rate": 1e-05, "loss": 0.4958, "mean_token_accuracy": 0.8580187559127808, "num_tokens": 155604314.0, "step": 4094 }, { "epoch": 0.5209260908281389, "ewc_loss": 0.1015625, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.337405681610107, "learning_rate": 1e-05, "loss": 0.5233, "mean_token_accuracy": 0.8572418689727783, "num_tokens": 155643259.0, "step": 4095 }, { "epoch": 0.5210533011067294, "ewc_loss": 0.08056640625, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 7.200241088867188e-05, "grad_norm": 4.997561454772949, "learning_rate": 1e-05, "loss": 0.5197, "mean_token_accuracy": 0.8515927791595459, "num_tokens": 155682522.0, "step": 4096 }, { "epoch": 0.52118051138532, "ewc_loss": 0.10595703125, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.578271865844727, "learning_rate": 1e-05, "loss": 0.5021, "mean_token_accuracy": 0.8640784621238708, "num_tokens": 155714021.0, "step": 4097 }, { "epoch": 0.5213077216639105, "ewc_loss": 0.08837890625, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.3072309494018555, "learning_rate": 1e-05, "loss": 0.582, "mean_token_accuracy": 0.8367041945457458, "num_tokens": 155754064.0, "step": 4098 }, { "epoch": 0.5214349319425009, "ewc_loss": 0.10498046875, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.5475263595581055, "learning_rate": 1e-05, "loss": 0.5307, "mean_token_accuracy": 0.8564661741256714, "num_tokens": 155788428.0, "step": 4099 }, { "epoch": 0.5215621422210914, "ewc_loss": 0.09326171875, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.595889091491699, "learning_rate": 1e-05, "loss": 0.5322, "mean_token_accuracy": 0.8502075672149658, "num_tokens": 155826608.0, "step": 4100 }, { "epoch": 0.521689352499682, "ewc_loss": 0.10009765625, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.227540969848633, "learning_rate": 1e-05, "loss": 0.5376, "mean_token_accuracy": 0.8541470766067505, "num_tokens": 155870314.0, "step": 4101 }, { "epoch": 0.5218165627782725, "ewc_loss": 0.0927734375, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.671924591064453, "learning_rate": 1e-05, "loss": 0.5617, "mean_token_accuracy": 0.8431872129440308, "num_tokens": 155907742.0, "step": 4102 }, { "epoch": 0.521943773056863, "ewc_loss": 0.087890625, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.842849254608154, "learning_rate": 1e-05, "loss": 0.4684, "mean_token_accuracy": 0.8701195120811462, "num_tokens": 155949837.0, "step": 4103 }, { "epoch": 0.5220709833354535, "ewc_loss": 0.09033203125, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.649616241455078, "learning_rate": 1e-05, "loss": 0.54, "mean_token_accuracy": 0.8528962731361389, "num_tokens": 155990407.0, "step": 4104 }, { "epoch": 0.522198193614044, "ewc_loss": 0.08203125, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 7.343292236328125e-05, "grad_norm": 5.521522045135498, "learning_rate": 1e-05, "loss": 0.525, "mean_token_accuracy": 0.8512483239173889, "num_tokens": 156027157.0, "step": 4105 }, { "epoch": 0.5223254038926345, "ewc_loss": 0.0888671875, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.644460678100586, "learning_rate": 1e-05, "loss": 0.5334, "mean_token_accuracy": 0.8524637222290039, "num_tokens": 156065074.0, "step": 4106 }, { "epoch": 0.522452614171225, "ewc_loss": 0.080078125, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 7.104873657226562e-05, "grad_norm": 5.288407802581787, "learning_rate": 1e-05, "loss": 0.4898, "mean_token_accuracy": 0.8636530637741089, "num_tokens": 156103399.0, "step": 4107 }, { "epoch": 0.5225798244498155, "ewc_loss": 0.08740234375, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.652626037597656, "learning_rate": 1e-05, "loss": 0.499, "mean_token_accuracy": 0.8587737679481506, "num_tokens": 156143825.0, "step": 4108 }, { "epoch": 0.5227070347284061, "ewc_loss": 0.0791015625, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 7.009506225585938e-05, "grad_norm": 5.138049602508545, "learning_rate": 1e-05, "loss": 0.4876, "mean_token_accuracy": 0.8583803176879883, "num_tokens": 156183841.0, "step": 4109 }, { "epoch": 0.5228342450069966, "ewc_loss": 0.08837890625, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.736252784729004, "learning_rate": 1e-05, "loss": 0.5033, "mean_token_accuracy": 0.8597919940948486, "num_tokens": 156225288.0, "step": 4110 }, { "epoch": 0.5229614552855871, "ewc_loss": 0.07763671875, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 6.866455078125e-05, "grad_norm": 5.020644664764404, "learning_rate": 1e-05, "loss": 0.4629, "mean_token_accuracy": 0.8698246479034424, "num_tokens": 156257463.0, "step": 4111 }, { "epoch": 0.5230886655641775, "ewc_loss": 0.09033203125, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.904511451721191, "learning_rate": 1e-05, "loss": 0.527, "mean_token_accuracy": 0.854815661907196, "num_tokens": 156293059.0, "step": 4112 }, { "epoch": 0.5232158758427681, "ewc_loss": 0.07763671875, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 6.866455078125e-05, "grad_norm": 4.915383338928223, "learning_rate": 1e-05, "loss": 0.5885, "mean_token_accuracy": 0.8324706554412842, "num_tokens": 156338605.0, "step": 4113 }, { "epoch": 0.5233430861213586, "ewc_loss": 0.09765625, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.180028438568115, "learning_rate": 1e-05, "loss": 0.5443, "mean_token_accuracy": 0.8525025844573975, "num_tokens": 156375472.0, "step": 4114 }, { "epoch": 0.5234702963999491, "ewc_loss": 0.07763671875, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 6.866455078125e-05, "grad_norm": 4.908511161804199, "learning_rate": 1e-05, "loss": 0.5175, "mean_token_accuracy": 0.8567161560058594, "num_tokens": 156412290.0, "step": 4115 }, { "epoch": 0.5235975066785397, "ewc_loss": 0.10498046875, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.502981662750244, "learning_rate": 1e-05, "loss": 0.5481, "mean_token_accuracy": 0.8534550666809082, "num_tokens": 156452268.0, "step": 4116 }, { "epoch": 0.5237247169571302, "ewc_loss": 0.08447265625, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 7.581710815429688e-05, "grad_norm": 5.103019714355469, "learning_rate": 1e-05, "loss": 0.5443, "mean_token_accuracy": 0.8486231565475464, "num_tokens": 156500659.0, "step": 4117 }, { "epoch": 0.5238519272357206, "ewc_loss": 0.10791015625, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.67368745803833, "learning_rate": 1e-05, "loss": 0.53, "mean_token_accuracy": 0.8604030609130859, "num_tokens": 156534529.0, "step": 4118 }, { "epoch": 0.5239791375143111, "ewc_loss": 0.09228515625, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.496519565582275, "learning_rate": 1e-05, "loss": 0.4786, "mean_token_accuracy": 0.8680081367492676, "num_tokens": 156574459.0, "step": 4119 }, { "epoch": 0.5241063477929017, "ewc_loss": 0.10400390625, "ewc_loss_diag": 8.7738037109375e-05, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.47708797454834, "learning_rate": 1e-05, "loss": 0.5336, "mean_token_accuracy": 0.8561677932739258, "num_tokens": 156611626.0, "step": 4120 }, { "epoch": 0.5242335580714922, "ewc_loss": 0.09619140625, "ewc_loss_diag": 8.7738037109375e-05, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 5.701633453369141, "learning_rate": 1e-05, "loss": 0.5134, "mean_token_accuracy": 0.8598918914794922, "num_tokens": 156649454.0, "step": 4121 }, { "epoch": 0.5243607683500827, "ewc_loss": 0.09521484375, "ewc_loss_diag": 8.7738037109375e-05, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 6.089169502258301, "learning_rate": 1e-05, "loss": 0.5342, "mean_token_accuracy": 0.8538331389427185, "num_tokens": 156687263.0, "step": 4122 }, { "epoch": 0.5244879786286732, "ewc_loss": 0.091796875, "ewc_loss_diag": 8.7738037109375e-05, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.706010818481445, "learning_rate": 1e-05, "loss": 0.5339, "mean_token_accuracy": 0.8522235155105591, "num_tokens": 156720384.0, "step": 4123 }, { "epoch": 0.5246151889072637, "ewc_loss": 0.08642578125, "ewc_loss_diag": 8.7738037109375e-05, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.708603858947754, "learning_rate": 1e-05, "loss": 0.5086, "mean_token_accuracy": 0.8590559363365173, "num_tokens": 156758365.0, "step": 4124 }, { "epoch": 0.5247423991858542, "ewc_loss": 0.08984375, "ewc_loss_diag": 8.7738037109375e-05, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.663548946380615, "learning_rate": 1e-05, "loss": 0.5147, "mean_token_accuracy": 0.8545528054237366, "num_tokens": 156790411.0, "step": 4125 }, { "epoch": 0.5248696094644447, "ewc_loss": 0.08203125, "ewc_loss_diag": 8.7738037109375e-05, "ewc_loss_parallel": 7.343292236328125e-05, "grad_norm": 5.4146647453308105, "learning_rate": 1e-05, "loss": 0.5145, "mean_token_accuracy": 0.8558533787727356, "num_tokens": 156829176.0, "step": 4126 }, { "epoch": 0.5249968197430352, "ewc_loss": 0.087890625, "ewc_loss_diag": 8.7738037109375e-05, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.6363348960876465, "learning_rate": 1e-05, "loss": 0.4917, "mean_token_accuracy": 0.8634750843048096, "num_tokens": 156871932.0, "step": 4127 }, { "epoch": 0.5251240300216258, "ewc_loss": 0.080078125, "ewc_loss_diag": 8.7738037109375e-05, "ewc_loss_parallel": 7.104873657226562e-05, "grad_norm": 5.236594200134277, "learning_rate": 1e-05, "loss": 0.4654, "mean_token_accuracy": 0.8672244548797607, "num_tokens": 156905810.0, "step": 4128 }, { "epoch": 0.5252512403002163, "ewc_loss": 0.0869140625, "ewc_loss_diag": 8.7738037109375e-05, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.665779113769531, "learning_rate": 1e-05, "loss": 0.4861, "mean_token_accuracy": 0.8634995222091675, "num_tokens": 156947218.0, "step": 4129 }, { "epoch": 0.5253784505788067, "ewc_loss": 0.0791015625, "ewc_loss_diag": 8.7738037109375e-05, "ewc_loss_parallel": 7.009506225585938e-05, "grad_norm": 5.113372802734375, "learning_rate": 1e-05, "loss": 0.541, "mean_token_accuracy": 0.8436074256896973, "num_tokens": 156978943.0, "step": 4130 }, { "epoch": 0.5255056608573972, "ewc_loss": 0.087890625, "ewc_loss_diag": 8.7738037109375e-05, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.7561869621276855, "learning_rate": 1e-05, "loss": 0.4737, "mean_token_accuracy": 0.8653534054756165, "num_tokens": 157015015.0, "step": 4131 }, { "epoch": 0.5256328711359878, "ewc_loss": 0.078125, "ewc_loss_diag": 8.7738037109375e-05, "ewc_loss_parallel": 6.914138793945312e-05, "grad_norm": 4.994593620300293, "learning_rate": 1e-05, "loss": 0.5163, "mean_token_accuracy": 0.8542811870574951, "num_tokens": 157051789.0, "step": 4132 }, { "epoch": 0.5257600814145783, "ewc_loss": 0.091796875, "ewc_loss_diag": 8.7738037109375e-05, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.937830448150635, "learning_rate": 1e-05, "loss": 0.4868, "mean_token_accuracy": 0.8657901287078857, "num_tokens": 157090568.0, "step": 4133 }, { "epoch": 0.5258872916931688, "ewc_loss": 0.07763671875, "ewc_loss_diag": 8.7738037109375e-05, "ewc_loss_parallel": 6.866455078125e-05, "grad_norm": 4.906356334686279, "learning_rate": 1e-05, "loss": 0.4831, "mean_token_accuracy": 0.8590213656425476, "num_tokens": 157125255.0, "step": 4134 }, { "epoch": 0.5260145019717594, "ewc_loss": 0.0986328125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.218438148498535, "learning_rate": 1e-05, "loss": 0.5277, "mean_token_accuracy": 0.8540441989898682, "num_tokens": 157164917.0, "step": 4135 }, { "epoch": 0.5261417122503498, "ewc_loss": 0.07861328125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 6.961822509765625e-05, "grad_norm": 4.917929172515869, "learning_rate": 1e-05, "loss": 0.5286, "mean_token_accuracy": 0.8503108620643616, "num_tokens": 157205186.0, "step": 4136 }, { "epoch": 0.5262689225289403, "ewc_loss": 0.10498046875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.543477535247803, "learning_rate": 1e-05, "loss": 0.5746, "mean_token_accuracy": 0.8455789089202881, "num_tokens": 157241511.0, "step": 4137 }, { "epoch": 0.5263961328075308, "ewc_loss": 0.08544921875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.165225982666016, "learning_rate": 1e-05, "loss": 0.4982, "mean_token_accuracy": 0.8622381687164307, "num_tokens": 157283381.0, "step": 4138 }, { "epoch": 0.5265233430861214, "ewc_loss": 0.1083984375, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.66088342666626, "learning_rate": 1e-05, "loss": 0.5454, "mean_token_accuracy": 0.8538464903831482, "num_tokens": 157321315.0, "step": 4139 }, { "epoch": 0.5266505533647119, "ewc_loss": 0.09326171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.543614387512207, "learning_rate": 1e-05, "loss": 0.5643, "mean_token_accuracy": 0.8426177501678467, "num_tokens": 157359205.0, "step": 4140 }, { "epoch": 0.5267777636433024, "ewc_loss": 0.10302734375, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.4176225662231445, "learning_rate": 1e-05, "loss": 0.5049, "mean_token_accuracy": 0.8631374835968018, "num_tokens": 157398889.0, "step": 4141 }, { "epoch": 0.5269049739218928, "ewc_loss": 0.09521484375, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.700098991394043, "learning_rate": 1e-05, "loss": 0.5734, "mean_token_accuracy": 0.8458744287490845, "num_tokens": 157440831.0, "step": 4142 }, { "epoch": 0.5270321842004834, "ewc_loss": 0.09228515625, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 6.029922962188721, "learning_rate": 1e-05, "loss": 0.5029, "mean_token_accuracy": 0.8570778965950012, "num_tokens": 157476275.0, "step": 4143 }, { "epoch": 0.5271593944790739, "ewc_loss": 0.091796875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.691384315490723, "learning_rate": 1e-05, "loss": 0.4783, "mean_token_accuracy": 0.8690716028213501, "num_tokens": 157508830.0, "step": 4144 }, { "epoch": 0.5272866047576644, "ewc_loss": 0.08544921875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.663692951202393, "learning_rate": 1e-05, "loss": 0.4925, "mean_token_accuracy": 0.8627251386642456, "num_tokens": 157539077.0, "step": 4145 }, { "epoch": 0.527413815036255, "ewc_loss": 0.09033203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.65036153793335, "learning_rate": 1e-05, "loss": 0.4817, "mean_token_accuracy": 0.8672370314598083, "num_tokens": 157583501.0, "step": 4146 }, { "epoch": 0.5275410253148455, "ewc_loss": 0.08203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 7.343292236328125e-05, "grad_norm": 5.380488872528076, "learning_rate": 1e-05, "loss": 0.52, "mean_token_accuracy": 0.8547930717468262, "num_tokens": 157629778.0, "step": 4147 }, { "epoch": 0.5276682355934359, "ewc_loss": 0.08837890625, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.654430389404297, "learning_rate": 1e-05, "loss": 0.5163, "mean_token_accuracy": 0.8596994876861572, "num_tokens": 157669375.0, "step": 4148 }, { "epoch": 0.5277954458720264, "ewc_loss": 0.080078125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 7.104873657226562e-05, "grad_norm": 5.205946922302246, "learning_rate": 1e-05, "loss": 0.4666, "mean_token_accuracy": 0.8663985133171082, "num_tokens": 157707400.0, "step": 4149 }, { "epoch": 0.527922656150617, "ewc_loss": 0.087890625, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.705177307128906, "learning_rate": 1e-05, "loss": 0.53, "mean_token_accuracy": 0.8543897867202759, "num_tokens": 157747039.0, "step": 4150 }, { "epoch": 0.5280498664292075, "ewc_loss": 0.07861328125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 6.961822509765625e-05, "grad_norm": 5.079575061798096, "learning_rate": 1e-05, "loss": 0.5007, "mean_token_accuracy": 0.8568243980407715, "num_tokens": 157782790.0, "step": 4151 }, { "epoch": 0.528177076707798, "ewc_loss": 0.08935546875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.808631896972656, "learning_rate": 1e-05, "loss": 0.5142, "mean_token_accuracy": 0.8601940870285034, "num_tokens": 157832002.0, "step": 4152 }, { "epoch": 0.5283042869863885, "ewc_loss": 0.07763671875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 6.866455078125e-05, "grad_norm": 4.970291614532471, "learning_rate": 1e-05, "loss": 0.5467, "mean_token_accuracy": 0.8458932042121887, "num_tokens": 157869875.0, "step": 4153 }, { "epoch": 0.528431497264979, "ewc_loss": 0.09326171875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 6.027923107147217, "learning_rate": 1e-05, "loss": 0.5393, "mean_token_accuracy": 0.8488390445709229, "num_tokens": 157908910.0, "step": 4154 }, { "epoch": 0.5285587075435695, "ewc_loss": 0.078125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 6.914138793945312e-05, "grad_norm": 4.896223545074463, "learning_rate": 1e-05, "loss": 0.5325, "mean_token_accuracy": 0.8413201570510864, "num_tokens": 157944366.0, "step": 4155 }, { "epoch": 0.52868591782216, "ewc_loss": 0.10107421875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.345736980438232, "learning_rate": 1e-05, "loss": 0.5315, "mean_token_accuracy": 0.8577722311019897, "num_tokens": 157982053.0, "step": 4156 }, { "epoch": 0.5288131281007505, "ewc_loss": 0.080078125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 7.152557373046875e-05, "grad_norm": 4.971746444702148, "learning_rate": 1e-05, "loss": 0.5123, "mean_token_accuracy": 0.8550093173980713, "num_tokens": 158022654.0, "step": 4157 }, { "epoch": 0.5289403383793411, "ewc_loss": 0.10693359375, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.639571189880371, "learning_rate": 1e-05, "loss": 0.6143, "mean_token_accuracy": 0.8320720195770264, "num_tokens": 158060508.0, "step": 4158 }, { "epoch": 0.5290675486579316, "ewc_loss": 0.08837890625, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.311404705047607, "learning_rate": 1e-05, "loss": 0.5592, "mean_token_accuracy": 0.8440520763397217, "num_tokens": 158098956.0, "step": 4159 }, { "epoch": 0.5291947589365221, "ewc_loss": 0.1064453125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.608924865722656, "learning_rate": 1e-05, "loss": 0.5571, "mean_token_accuracy": 0.8521329164505005, "num_tokens": 158134781.0, "step": 4160 }, { "epoch": 0.5293219692151125, "ewc_loss": 0.095703125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.624039173126221, "learning_rate": 1e-05, "loss": 0.5014, "mean_token_accuracy": 0.86473548412323, "num_tokens": 158177495.0, "step": 4161 }, { "epoch": 0.5294491794937031, "ewc_loss": 0.1005859375, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.2858052253723145, "learning_rate": 1e-05, "loss": 0.5367, "mean_token_accuracy": 0.8535230159759521, "num_tokens": 158216672.0, "step": 4162 }, { "epoch": 0.5295763897722936, "ewc_loss": 0.095703125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.708189487457275, "learning_rate": 1e-05, "loss": 0.5479, "mean_token_accuracy": 0.84749436378479, "num_tokens": 158255397.0, "step": 4163 }, { "epoch": 0.5297036000508841, "ewc_loss": 0.08837890625, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.898002624511719, "learning_rate": 1e-05, "loss": 0.5309, "mean_token_accuracy": 0.8486448526382446, "num_tokens": 158292230.0, "step": 4164 }, { "epoch": 0.5298308103294747, "ewc_loss": 0.09130859375, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.676713943481445, "learning_rate": 1e-05, "loss": 0.5179, "mean_token_accuracy": 0.8564417362213135, "num_tokens": 158329884.0, "step": 4165 }, { "epoch": 0.5299580206080652, "ewc_loss": 0.083984375, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 7.534027099609375e-05, "grad_norm": 5.57172155380249, "learning_rate": 1e-05, "loss": 0.6162, "mean_token_accuracy": 0.8298689126968384, "num_tokens": 158367290.0, "step": 4166 }, { "epoch": 0.5300852308866556, "ewc_loss": 0.0888671875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.638848304748535, "learning_rate": 1e-05, "loss": 0.4531, "mean_token_accuracy": 0.8778160214424133, "num_tokens": 158407999.0, "step": 4167 }, { "epoch": 0.5302124411652461, "ewc_loss": 0.080078125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 7.152557373046875e-05, "grad_norm": 5.342704772949219, "learning_rate": 1e-05, "loss": 0.4639, "mean_token_accuracy": 0.8698607087135315, "num_tokens": 158444283.0, "step": 4168 }, { "epoch": 0.5303396514438367, "ewc_loss": 0.08740234375, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.641769886016846, "learning_rate": 1e-05, "loss": 0.5238, "mean_token_accuracy": 0.8534387946128845, "num_tokens": 158484807.0, "step": 4169 }, { "epoch": 0.5304668617224272, "ewc_loss": 0.0791015625, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 7.009506225585938e-05, "grad_norm": 5.186763763427734, "learning_rate": 1e-05, "loss": 0.4915, "mean_token_accuracy": 0.861095666885376, "num_tokens": 158518863.0, "step": 4170 }, { "epoch": 0.5305940720010177, "ewc_loss": 0.0869140625, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.700467109680176, "learning_rate": 1e-05, "loss": 0.4934, "mean_token_accuracy": 0.8639612793922424, "num_tokens": 158555057.0, "step": 4171 }, { "epoch": 0.5307212822796082, "ewc_loss": 0.078125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 6.914138793945312e-05, "grad_norm": 5.059928894042969, "learning_rate": 1e-05, "loss": 0.5048, "mean_token_accuracy": 0.858818531036377, "num_tokens": 158596032.0, "step": 4172 }, { "epoch": 0.5308484925581987, "ewc_loss": 0.0888671875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.8358893394470215, "learning_rate": 1e-05, "loss": 0.5547, "mean_token_accuracy": 0.8460574746131897, "num_tokens": 158631098.0, "step": 4173 }, { "epoch": 0.5309757028367892, "ewc_loss": 0.078125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 6.914138793945312e-05, "grad_norm": 4.954795837402344, "learning_rate": 1e-05, "loss": 0.5376, "mean_token_accuracy": 0.8473198413848877, "num_tokens": 158664579.0, "step": 4174 }, { "epoch": 0.5311029131153797, "ewc_loss": 0.095703125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 6.068927764892578, "learning_rate": 1e-05, "loss": 0.5591, "mean_token_accuracy": 0.8469349145889282, "num_tokens": 158701220.0, "step": 4175 }, { "epoch": 0.5312301233939702, "ewc_loss": 0.07763671875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 6.866455078125e-05, "grad_norm": 4.872811794281006, "learning_rate": 1e-05, "loss": 0.4824, "mean_token_accuracy": 0.8627648949623108, "num_tokens": 158744235.0, "step": 4176 }, { "epoch": 0.5313573336725608, "ewc_loss": 0.10302734375, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.405479431152344, "learning_rate": 1e-05, "loss": 0.5151, "mean_token_accuracy": 0.8605681657791138, "num_tokens": 158782602.0, "step": 4177 }, { "epoch": 0.5314845439511513, "ewc_loss": 0.08203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 7.343292236328125e-05, "grad_norm": 5.012792110443115, "learning_rate": 1e-05, "loss": 0.5051, "mean_token_accuracy": 0.8596509695053101, "num_tokens": 158813345.0, "step": 4178 }, { "epoch": 0.5316117542297417, "ewc_loss": 0.10791015625, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.667257785797119, "learning_rate": 1e-05, "loss": 0.5761, "mean_token_accuracy": 0.8420610427856445, "num_tokens": 158847204.0, "step": 4179 }, { "epoch": 0.5317389645083322, "ewc_loss": 0.08984375, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.37746000289917, "learning_rate": 1e-05, "loss": 0.5294, "mean_token_accuracy": 0.8526347875595093, "num_tokens": 158889481.0, "step": 4180 }, { "epoch": 0.5318661747869228, "ewc_loss": 0.1064453125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.572154521942139, "learning_rate": 1e-05, "loss": 0.5097, "mean_token_accuracy": 0.8641694784164429, "num_tokens": 158929267.0, "step": 4181 }, { "epoch": 0.5319933850655133, "ewc_loss": 0.09521484375, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.666187763214111, "learning_rate": 1e-05, "loss": 0.562, "mean_token_accuracy": 0.8440207242965698, "num_tokens": 158974526.0, "step": 4182 }, { "epoch": 0.5321205953441038, "ewc_loss": 0.09912109375, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.218576908111572, "learning_rate": 1e-05, "loss": 0.526, "mean_token_accuracy": 0.8598735332489014, "num_tokens": 159015428.0, "step": 4183 }, { "epoch": 0.5322478056226944, "ewc_loss": 0.09375, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.719874382019043, "learning_rate": 1e-05, "loss": 0.5669, "mean_token_accuracy": 0.8446763753890991, "num_tokens": 159050492.0, "step": 4184 }, { "epoch": 0.5323750159012848, "ewc_loss": 0.08837890625, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.838597774505615, "learning_rate": 1e-05, "loss": 0.554, "mean_token_accuracy": 0.8457617163658142, "num_tokens": 159085644.0, "step": 4185 }, { "epoch": 0.5325022261798753, "ewc_loss": 0.0908203125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.670892238616943, "learning_rate": 1e-05, "loss": 0.5807, "mean_token_accuracy": 0.8431359529495239, "num_tokens": 159123093.0, "step": 4186 }, { "epoch": 0.5326294364584658, "ewc_loss": 0.0830078125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 7.43865966796875e-05, "grad_norm": 5.512893199920654, "learning_rate": 1e-05, "loss": 0.5076, "mean_token_accuracy": 0.8545045256614685, "num_tokens": 159157688.0, "step": 4187 }, { "epoch": 0.5327566467370564, "ewc_loss": 0.08935546875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.652555465698242, "learning_rate": 1e-05, "loss": 0.5392, "mean_token_accuracy": 0.8490228652954102, "num_tokens": 159194473.0, "step": 4188 }, { "epoch": 0.5328838570156469, "ewc_loss": 0.08056640625, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 7.200241088867188e-05, "grad_norm": 5.297158718109131, "learning_rate": 1e-05, "loss": 0.5336, "mean_token_accuracy": 0.8488354682922363, "num_tokens": 159230314.0, "step": 4189 }, { "epoch": 0.5330110672942374, "ewc_loss": 0.08837890625, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.656951427459717, "learning_rate": 1e-05, "loss": 0.4973, "mean_token_accuracy": 0.8635623455047607, "num_tokens": 159266579.0, "step": 4190 }, { "epoch": 0.5331382775728278, "ewc_loss": 0.0791015625, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 7.009506225585938e-05, "grad_norm": 5.157253265380859, "learning_rate": 1e-05, "loss": 0.4822, "mean_token_accuracy": 0.8653907179832458, "num_tokens": 159306502.0, "step": 4191 }, { "epoch": 0.5332654878514184, "ewc_loss": 0.08837890625, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.728793144226074, "learning_rate": 1e-05, "loss": 0.5543, "mean_token_accuracy": 0.8393265008926392, "num_tokens": 159342512.0, "step": 4192 }, { "epoch": 0.5333926981300089, "ewc_loss": 0.0791015625, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 7.009506225585938e-05, "grad_norm": 5.042539596557617, "learning_rate": 1e-05, "loss": 0.4955, "mean_token_accuracy": 0.8614799976348877, "num_tokens": 159378408.0, "step": 4193 }, { "epoch": 0.5335199084085994, "ewc_loss": 0.0908203125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.88284158706665, "learning_rate": 1e-05, "loss": 0.5933, "mean_token_accuracy": 0.835519552230835, "num_tokens": 159415843.0, "step": 4194 }, { "epoch": 0.53364711868719, "ewc_loss": 0.078125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 6.914138793945312e-05, "grad_norm": 4.938655853271484, "learning_rate": 1e-05, "loss": 0.4999, "mean_token_accuracy": 0.8553797006607056, "num_tokens": 159446880.0, "step": 4195 }, { "epoch": 0.5337743289657805, "ewc_loss": 0.0966796875, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 6.116455078125, "learning_rate": 1e-05, "loss": 0.536, "mean_token_accuracy": 0.8582772612571716, "num_tokens": 159491960.0, "step": 4196 }, { "epoch": 0.5339015392443709, "ewc_loss": 0.078125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 6.914138793945312e-05, "grad_norm": 4.8950114250183105, "learning_rate": 1e-05, "loss": 0.4834, "mean_token_accuracy": 0.8590025901794434, "num_tokens": 159529157.0, "step": 4197 }, { "epoch": 0.5340287495229614, "ewc_loss": 0.10400390625, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.45121431350708, "learning_rate": 1e-05, "loss": 0.5513, "mean_token_accuracy": 0.8505696654319763, "num_tokens": 159563589.0, "step": 4198 }, { "epoch": 0.534155959801552, "ewc_loss": 0.08349609375, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 7.486343383789062e-05, "grad_norm": 5.052725791931152, "learning_rate": 1e-05, "loss": 0.4468, "mean_token_accuracy": 0.8748196363449097, "num_tokens": 159601789.0, "step": 4199 }, { "epoch": 0.5342831700801425, "ewc_loss": 0.10986328125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.6688432693481445, "learning_rate": 1e-05, "loss": 0.605, "mean_token_accuracy": 0.8338313102722168, "num_tokens": 159637357.0, "step": 4200 }, { "epoch": 0.534410380358733, "ewc_loss": 0.09130859375, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.4280219078063965, "learning_rate": 1e-05, "loss": 0.4762, "mean_token_accuracy": 0.8692536354064941, "num_tokens": 159678919.0, "step": 4201 }, { "epoch": 0.5345375906373235, "ewc_loss": 0.1064453125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.535035610198975, "learning_rate": 1e-05, "loss": 0.5534, "mean_token_accuracy": 0.8506714105606079, "num_tokens": 159714109.0, "step": 4202 }, { "epoch": 0.534664800915914, "ewc_loss": 0.095703125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.6822285652160645, "learning_rate": 1e-05, "loss": 0.5051, "mean_token_accuracy": 0.8636692762374878, "num_tokens": 159752171.0, "step": 4203 }, { "epoch": 0.5347920111945045, "ewc_loss": 0.0986328125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.162909984588623, "learning_rate": 1e-05, "loss": 0.5619, "mean_token_accuracy": 0.8458320498466492, "num_tokens": 159791251.0, "step": 4204 }, { "epoch": 0.534919221473095, "ewc_loss": 0.09423828125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.715480804443359, "learning_rate": 1e-05, "loss": 0.4844, "mean_token_accuracy": 0.8657395839691162, "num_tokens": 159819925.0, "step": 4205 }, { "epoch": 0.5350464317516855, "ewc_loss": 0.087890625, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.76807975769043, "learning_rate": 1e-05, "loss": 0.5455, "mean_token_accuracy": 0.8496230840682983, "num_tokens": 159859788.0, "step": 4206 }, { "epoch": 0.5351736420302761, "ewc_loss": 0.0908203125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.677988052368164, "learning_rate": 1e-05, "loss": 0.549, "mean_token_accuracy": 0.8482502698898315, "num_tokens": 159897202.0, "step": 4207 }, { "epoch": 0.5353008523088666, "ewc_loss": 0.08251953125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 7.390975952148438e-05, "grad_norm": 5.45147705078125, "learning_rate": 1e-05, "loss": 0.5489, "mean_token_accuracy": 0.8440523147583008, "num_tokens": 159935840.0, "step": 4208 }, { "epoch": 0.5354280625874571, "ewc_loss": 0.0888671875, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.674070835113525, "learning_rate": 1e-05, "loss": 0.4947, "mean_token_accuracy": 0.8647279739379883, "num_tokens": 159975120.0, "step": 4209 }, { "epoch": 0.5355552728660475, "ewc_loss": 0.080078125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 7.152557373046875e-05, "grad_norm": 5.241500377655029, "learning_rate": 1e-05, "loss": 0.4873, "mean_token_accuracy": 0.8624448180198669, "num_tokens": 160020935.0, "step": 4210 }, { "epoch": 0.5356824831446381, "ewc_loss": 0.087890625, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.716279983520508, "learning_rate": 1e-05, "loss": 0.4759, "mean_token_accuracy": 0.8696364164352417, "num_tokens": 160065029.0, "step": 4211 }, { "epoch": 0.5358096934232286, "ewc_loss": 0.07958984375, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 7.05718994140625e-05, "grad_norm": 5.115671157836914, "learning_rate": 1e-05, "loss": 0.5367, "mean_token_accuracy": 0.8463183045387268, "num_tokens": 160104865.0, "step": 4212 }, { "epoch": 0.5359369037018191, "ewc_loss": 0.08935546875, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.81757116317749, "learning_rate": 1e-05, "loss": 0.4972, "mean_token_accuracy": 0.8649361729621887, "num_tokens": 160145477.0, "step": 4213 }, { "epoch": 0.5360641139804097, "ewc_loss": 0.07763671875, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 6.866455078125e-05, "grad_norm": 4.998363494873047, "learning_rate": 1e-05, "loss": 0.5372, "mean_token_accuracy": 0.8475492000579834, "num_tokens": 160184906.0, "step": 4214 }, { "epoch": 0.5361913242590002, "ewc_loss": 0.09326171875, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 6.005911350250244, "learning_rate": 1e-05, "loss": 0.5394, "mean_token_accuracy": 0.8518158197402954, "num_tokens": 160225011.0, "step": 4215 }, { "epoch": 0.5363185345375906, "ewc_loss": 0.07763671875, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 6.866455078125e-05, "grad_norm": 4.927466869354248, "learning_rate": 1e-05, "loss": 0.5026, "mean_token_accuracy": 0.8565046787261963, "num_tokens": 160263193.0, "step": 4216 }, { "epoch": 0.5364457448161811, "ewc_loss": 0.099609375, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.284212112426758, "learning_rate": 1e-05, "loss": 0.4589, "mean_token_accuracy": 0.8799734115600586, "num_tokens": 160299552.0, "step": 4217 }, { "epoch": 0.5365729550947717, "ewc_loss": 0.080078125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 7.104873657226562e-05, "grad_norm": 4.978434085845947, "learning_rate": 1e-05, "loss": 0.4702, "mean_token_accuracy": 0.8686199188232422, "num_tokens": 160334609.0, "step": 4218 }, { "epoch": 0.5367001653733622, "ewc_loss": 0.1064453125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.567775249481201, "learning_rate": 1e-05, "loss": 0.5213, "mean_token_accuracy": 0.8604156374931335, "num_tokens": 160374295.0, "step": 4219 }, { "epoch": 0.5368273756519527, "ewc_loss": 0.0869140625, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.24884033203125, "learning_rate": 1e-05, "loss": 0.5657, "mean_token_accuracy": 0.839339017868042, "num_tokens": 160404420.0, "step": 4220 }, { "epoch": 0.5369545859305432, "ewc_loss": 0.1064453125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.590860366821289, "learning_rate": 1e-05, "loss": 0.5309, "mean_token_accuracy": 0.8536868691444397, "num_tokens": 160439278.0, "step": 4221 }, { "epoch": 0.5370817962091337, "ewc_loss": 0.09326171875, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.552099227905273, "learning_rate": 1e-05, "loss": 0.4682, "mean_token_accuracy": 0.8731646537780762, "num_tokens": 160475225.0, "step": 4222 }, { "epoch": 0.5372090064877242, "ewc_loss": 0.10107421875, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.322028160095215, "learning_rate": 1e-05, "loss": 0.5359, "mean_token_accuracy": 0.8592947721481323, "num_tokens": 160516073.0, "step": 4223 }, { "epoch": 0.5373362167663147, "ewc_loss": 0.09423828125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.684383392333984, "learning_rate": 1e-05, "loss": 0.5471, "mean_token_accuracy": 0.8513982892036438, "num_tokens": 160550938.0, "step": 4224 }, { "epoch": 0.5374634270449052, "ewc_loss": 0.0908203125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.943802833557129, "learning_rate": 1e-05, "loss": 0.4986, "mean_token_accuracy": 0.8622403144836426, "num_tokens": 160590382.0, "step": 4225 }, { "epoch": 0.5375906373234958, "ewc_loss": 0.0908203125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.660763263702393, "learning_rate": 1e-05, "loss": 0.4816, "mean_token_accuracy": 0.8695584535598755, "num_tokens": 160624830.0, "step": 4226 }, { "epoch": 0.5377178476020863, "ewc_loss": 0.08447265625, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 7.581710815429688e-05, "grad_norm": 5.600106716156006, "learning_rate": 1e-05, "loss": 0.5199, "mean_token_accuracy": 0.8535536527633667, "num_tokens": 160663042.0, "step": 4227 }, { "epoch": 0.5378450578806767, "ewc_loss": 0.08984375, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.651115417480469, "learning_rate": 1e-05, "loss": 0.5472, "mean_token_accuracy": 0.8467342853546143, "num_tokens": 160700502.0, "step": 4228 }, { "epoch": 0.5379722681592672, "ewc_loss": 0.0810546875, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 7.2479248046875e-05, "grad_norm": 5.353446960449219, "learning_rate": 1e-05, "loss": 0.5644, "mean_token_accuracy": 0.8396388292312622, "num_tokens": 160736746.0, "step": 4229 }, { "epoch": 0.5380994784378578, "ewc_loss": 0.08837890625, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.658374309539795, "learning_rate": 1e-05, "loss": 0.4802, "mean_token_accuracy": 0.8645197153091431, "num_tokens": 160771430.0, "step": 4230 }, { "epoch": 0.5382266887164483, "ewc_loss": 0.080078125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 7.104873657226562e-05, "grad_norm": 5.19263219833374, "learning_rate": 1e-05, "loss": 0.5142, "mean_token_accuracy": 0.8552384376525879, "num_tokens": 160807047.0, "step": 4231 }, { "epoch": 0.5383538989950388, "ewc_loss": 0.08837890625, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.713445663452148, "learning_rate": 1e-05, "loss": 0.5789, "mean_token_accuracy": 0.8376404047012329, "num_tokens": 160844600.0, "step": 4232 }, { "epoch": 0.5384811092736294, "ewc_loss": 0.0791015625, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 7.009506225585938e-05, "grad_norm": 5.061807632446289, "learning_rate": 1e-05, "loss": 0.5051, "mean_token_accuracy": 0.8595672249794006, "num_tokens": 160883132.0, "step": 4233 }, { "epoch": 0.5386083195522198, "ewc_loss": 0.08935546875, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.839477062225342, "learning_rate": 1e-05, "loss": 0.5155, "mean_token_accuracy": 0.8565762042999268, "num_tokens": 160922408.0, "step": 4234 }, { "epoch": 0.5387355298308103, "ewc_loss": 0.07763671875, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 6.866455078125e-05, "grad_norm": 4.951290607452393, "learning_rate": 1e-05, "loss": 0.497, "mean_token_accuracy": 0.8607441782951355, "num_tokens": 160960563.0, "step": 4235 }, { "epoch": 0.5388627401094008, "ewc_loss": 0.0947265625, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 6.077188014984131, "learning_rate": 1e-05, "loss": 0.5289, "mean_token_accuracy": 0.8562740683555603, "num_tokens": 160995594.0, "step": 4236 }, { "epoch": 0.5389899503879914, "ewc_loss": 0.078125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 6.914138793945312e-05, "grad_norm": 4.9109063148498535, "learning_rate": 1e-05, "loss": 0.5197, "mean_token_accuracy": 0.853020429611206, "num_tokens": 161028341.0, "step": 4237 }, { "epoch": 0.5391171606665819, "ewc_loss": 0.10205078125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.39230489730835, "learning_rate": 1e-05, "loss": 0.5195, "mean_token_accuracy": 0.8577630519866943, "num_tokens": 161064549.0, "step": 4238 }, { "epoch": 0.5392443709451724, "ewc_loss": 0.08203125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 7.343292236328125e-05, "grad_norm": 5.009845733642578, "learning_rate": 1e-05, "loss": 0.5591, "mean_token_accuracy": 0.8412691354751587, "num_tokens": 161105170.0, "step": 4239 }, { "epoch": 0.5393715812237628, "ewc_loss": 0.1083984375, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.666529655456543, "learning_rate": 1e-05, "loss": 0.6013, "mean_token_accuracy": 0.8365468382835388, "num_tokens": 161141018.0, "step": 4240 }, { "epoch": 0.5394987915023534, "ewc_loss": 0.08984375, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.373833656311035, "learning_rate": 1e-05, "loss": 0.5696, "mean_token_accuracy": 0.8410508632659912, "num_tokens": 161175363.0, "step": 4241 }, { "epoch": 0.5396260017809439, "ewc_loss": 0.107421875, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.5910725593566895, "learning_rate": 1e-05, "loss": 0.5243, "mean_token_accuracy": 0.8617148995399475, "num_tokens": 161210920.0, "step": 4242 }, { "epoch": 0.5397532120595344, "ewc_loss": 0.095703125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.650174140930176, "learning_rate": 1e-05, "loss": 0.4865, "mean_token_accuracy": 0.8684237599372864, "num_tokens": 161248743.0, "step": 4243 }, { "epoch": 0.5398804223381249, "ewc_loss": 0.1005859375, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.2565693855285645, "learning_rate": 1e-05, "loss": 0.5095, "mean_token_accuracy": 0.8594588041305542, "num_tokens": 161282897.0, "step": 4244 }, { "epoch": 0.5400076326167155, "ewc_loss": 0.095703125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.717743873596191, "learning_rate": 1e-05, "loss": 0.528, "mean_token_accuracy": 0.8568937182426453, "num_tokens": 161323103.0, "step": 4245 }, { "epoch": 0.5401348428953059, "ewc_loss": 0.08984375, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.868175506591797, "learning_rate": 1e-05, "loss": 0.5191, "mean_token_accuracy": 0.855380654335022, "num_tokens": 161359612.0, "step": 4246 }, { "epoch": 0.5402620531738964, "ewc_loss": 0.0908203125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.670942306518555, "learning_rate": 1e-05, "loss": 0.4874, "mean_token_accuracy": 0.8682608604431152, "num_tokens": 161396961.0, "step": 4247 }, { "epoch": 0.540389263452487, "ewc_loss": 0.0849609375, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 7.581710815429688e-05, "grad_norm": 5.538951396942139, "learning_rate": 1e-05, "loss": 0.5275, "mean_token_accuracy": 0.8502620458602905, "num_tokens": 161437380.0, "step": 4248 }, { "epoch": 0.5405164737310775, "ewc_loss": 0.08984375, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.660032749176025, "learning_rate": 1e-05, "loss": 0.523, "mean_token_accuracy": 0.8535884022712708, "num_tokens": 161480237.0, "step": 4249 }, { "epoch": 0.540643684009668, "ewc_loss": 0.08203125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 7.295608520507812e-05, "grad_norm": 5.303269386291504, "learning_rate": 1e-05, "loss": 0.4546, "mean_token_accuracy": 0.8707800507545471, "num_tokens": 161519447.0, "step": 4250 }, { "epoch": 0.5407708942882585, "ewc_loss": 0.0888671875, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.685301303863525, "learning_rate": 1e-05, "loss": 0.5416, "mean_token_accuracy": 0.8478650450706482, "num_tokens": 161556588.0, "step": 4251 }, { "epoch": 0.540898104566849, "ewc_loss": 0.080078125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 7.05718994140625e-05, "grad_norm": 5.156252861022949, "learning_rate": 1e-05, "loss": 0.4925, "mean_token_accuracy": 0.8594403266906738, "num_tokens": 161591542.0, "step": 4252 }, { "epoch": 0.5410253148454395, "ewc_loss": 0.0888671875, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.766587734222412, "learning_rate": 1e-05, "loss": 0.524, "mean_token_accuracy": 0.8569799661636353, "num_tokens": 161630453.0, "step": 4253 }, { "epoch": 0.54115252512403, "ewc_loss": 0.0791015625, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 7.009506225585938e-05, "grad_norm": 5.036952972412109, "learning_rate": 1e-05, "loss": 0.4298, "mean_token_accuracy": 0.880447506904602, "num_tokens": 161660927.0, "step": 4254 }, { "epoch": 0.5412797354026205, "ewc_loss": 0.091796875, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.9285807609558105, "learning_rate": 1e-05, "loss": 0.526, "mean_token_accuracy": 0.8589363098144531, "num_tokens": 161703284.0, "step": 4255 }, { "epoch": 0.5414069456812111, "ewc_loss": 0.078125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 6.866455078125e-05, "grad_norm": 4.9267706871032715, "learning_rate": 1e-05, "loss": 0.5469, "mean_token_accuracy": 0.84527188539505, "num_tokens": 161739212.0, "step": 4256 }, { "epoch": 0.5415341559598016, "ewc_loss": 0.0986328125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.218501091003418, "learning_rate": 1e-05, "loss": 0.5637, "mean_token_accuracy": 0.8550946712493896, "num_tokens": 161777635.0, "step": 4257 }, { "epoch": 0.5416613662383921, "ewc_loss": 0.078125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 6.914138793945312e-05, "grad_norm": 4.921200752258301, "learning_rate": 1e-05, "loss": 0.5333, "mean_token_accuracy": 0.8508487939834595, "num_tokens": 161816019.0, "step": 4258 }, { "epoch": 0.5417885765169825, "ewc_loss": 0.1064453125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.542049407958984, "learning_rate": 1e-05, "loss": 0.5227, "mean_token_accuracy": 0.8583842515945435, "num_tokens": 161849690.0, "step": 4259 }, { "epoch": 0.5419157867955731, "ewc_loss": 0.0859375, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.147247314453125, "learning_rate": 1e-05, "loss": 0.4634, "mean_token_accuracy": 0.8720927834510803, "num_tokens": 161889596.0, "step": 4260 }, { "epoch": 0.5420429970741636, "ewc_loss": 0.1103515625, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.695363998413086, "learning_rate": 1e-05, "loss": 0.5978, "mean_token_accuracy": 0.8385856747627258, "num_tokens": 161929451.0, "step": 4261 }, { "epoch": 0.5421702073527541, "ewc_loss": 0.0927734375, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.5358076095581055, "learning_rate": 1e-05, "loss": 0.5555, "mean_token_accuracy": 0.8463644981384277, "num_tokens": 161966422.0, "step": 4262 }, { "epoch": 0.5422974176313446, "ewc_loss": 0.1044921875, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.480647563934326, "learning_rate": 1e-05, "loss": 0.54, "mean_token_accuracy": 0.8526229858398438, "num_tokens": 162005192.0, "step": 4263 }, { "epoch": 0.5424246279099352, "ewc_loss": 0.0966796875, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 5.718614101409912, "learning_rate": 1e-05, "loss": 0.5523, "mean_token_accuracy": 0.853378415107727, "num_tokens": 162051039.0, "step": 4264 }, { "epoch": 0.5425518381885256, "ewc_loss": 0.0947265625, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 6.095990180969238, "learning_rate": 1e-05, "loss": 0.495, "mean_token_accuracy": 0.8610880374908447, "num_tokens": 162086869.0, "step": 4265 }, { "epoch": 0.5426790484671161, "ewc_loss": 0.0927734375, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.7125067710876465, "learning_rate": 1e-05, "loss": 0.5091, "mean_token_accuracy": 0.8570699095726013, "num_tokens": 162122442.0, "step": 4266 }, { "epoch": 0.5428062587457066, "ewc_loss": 0.0869140625, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.740478992462158, "learning_rate": 1e-05, "loss": 0.51, "mean_token_accuracy": 0.8585497140884399, "num_tokens": 162153057.0, "step": 4267 }, { "epoch": 0.5429334690242972, "ewc_loss": 0.08984375, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.667588233947754, "learning_rate": 1e-05, "loss": 0.5817, "mean_token_accuracy": 0.8405719995498657, "num_tokens": 162192086.0, "step": 4268 }, { "epoch": 0.5430606793028877, "ewc_loss": 0.0830078125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 7.390975952148438e-05, "grad_norm": 5.457778453826904, "learning_rate": 1e-05, "loss": 0.4815, "mean_token_accuracy": 0.8655054569244385, "num_tokens": 162227206.0, "step": 4269 }, { "epoch": 0.5431878895814782, "ewc_loss": 0.087890625, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.650106430053711, "learning_rate": 1e-05, "loss": 0.5413, "mean_token_accuracy": 0.8495216369628906, "num_tokens": 162263490.0, "step": 4270 }, { "epoch": 0.5433150998600687, "ewc_loss": 0.0810546875, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 7.200241088867188e-05, "grad_norm": 5.269792556762695, "learning_rate": 1e-05, "loss": 0.4819, "mean_token_accuracy": 0.8689680695533752, "num_tokens": 162302750.0, "step": 4271 }, { "epoch": 0.5434423101386592, "ewc_loss": 0.087890625, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.662480354309082, "learning_rate": 1e-05, "loss": 0.4656, "mean_token_accuracy": 0.8727046251296997, "num_tokens": 162339613.0, "step": 4272 }, { "epoch": 0.5435695204172497, "ewc_loss": 0.080078125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 7.05718994140625e-05, "grad_norm": 5.1479620933532715, "learning_rate": 1e-05, "loss": 0.532, "mean_token_accuracy": 0.8503830432891846, "num_tokens": 162380163.0, "step": 4273 }, { "epoch": 0.5436967306958402, "ewc_loss": 0.0888671875, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.755136489868164, "learning_rate": 1e-05, "loss": 0.5357, "mean_token_accuracy": 0.851308286190033, "num_tokens": 162413306.0, "step": 4274 }, { "epoch": 0.5438239409744308, "ewc_loss": 0.0791015625, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 6.961822509765625e-05, "grad_norm": 5.033254146575928, "learning_rate": 1e-05, "loss": 0.4821, "mean_token_accuracy": 0.8640759587287903, "num_tokens": 162446861.0, "step": 4275 }, { "epoch": 0.5439511512530213, "ewc_loss": 0.091796875, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.922329425811768, "learning_rate": 1e-05, "loss": 0.5444, "mean_token_accuracy": 0.8491275906562805, "num_tokens": 162491414.0, "step": 4276 }, { "epoch": 0.5440783615316117, "ewc_loss": 0.078125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 6.866455078125e-05, "grad_norm": 4.916189670562744, "learning_rate": 1e-05, "loss": 0.4851, "mean_token_accuracy": 0.8626441955566406, "num_tokens": 162535265.0, "step": 4277 }, { "epoch": 0.5442055718102022, "ewc_loss": 0.099609375, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.225009918212891, "learning_rate": 1e-05, "loss": 0.5668, "mean_token_accuracy": 0.8438822627067566, "num_tokens": 162577198.0, "step": 4278 }, { "epoch": 0.5443327820887928, "ewc_loss": 0.078125, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 6.866455078125e-05, "grad_norm": 4.911351680755615, "learning_rate": 1e-05, "loss": 0.4901, "mean_token_accuracy": 0.8598877787590027, "num_tokens": 162609557.0, "step": 4279 }, { "epoch": 0.5444599923673833, "ewc_loss": 0.10546875, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.581843852996826, "learning_rate": 1e-05, "loss": 0.5216, "mean_token_accuracy": 0.8602817058563232, "num_tokens": 162642943.0, "step": 4280 }, { "epoch": 0.5445872026459738, "ewc_loss": 0.0859375, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.1599249839782715, "learning_rate": 1e-05, "loss": 0.5156, "mean_token_accuracy": 0.8588390946388245, "num_tokens": 162676444.0, "step": 4281 }, { "epoch": 0.5447144129245644, "ewc_loss": 0.1103515625, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.73422908782959, "learning_rate": 1e-05, "loss": 0.5543, "mean_token_accuracy": 0.8527606725692749, "num_tokens": 162715445.0, "step": 4282 }, { "epoch": 0.5448416232031548, "ewc_loss": 0.09375, "ewc_loss_diag": 9.012222290039062e-05, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.562252998352051, "learning_rate": 1e-05, "loss": 0.5814, "mean_token_accuracy": 0.8400592803955078, "num_tokens": 162751306.0, "step": 4283 }, { "epoch": 0.5449688334817453, "ewc_loss": 0.1044921875, "ewc_loss_diag": 9.059906005859375e-05, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.48891544342041, "learning_rate": 1e-05, "loss": 0.6265, "mean_token_accuracy": 0.8252518773078918, "num_tokens": 162793587.0, "step": 4284 }, { "epoch": 0.5450960437603358, "ewc_loss": 0.09716796875, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 5.751683712005615, "learning_rate": 1e-05, "loss": 0.5709, "mean_token_accuracy": 0.8409897685050964, "num_tokens": 162831000.0, "step": 4285 }, { "epoch": 0.5452232540389264, "ewc_loss": 0.095703125, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 6.086031436920166, "learning_rate": 1e-05, "loss": 0.441, "mean_token_accuracy": 0.8807506561279297, "num_tokens": 162863913.0, "step": 4286 }, { "epoch": 0.5453504643175169, "ewc_loss": 0.09326171875, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.73186731338501, "learning_rate": 1e-05, "loss": 0.5035, "mean_token_accuracy": 0.8606216311454773, "num_tokens": 162899135.0, "step": 4287 }, { "epoch": 0.5454776745961074, "ewc_loss": 0.0869140625, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.711398124694824, "learning_rate": 1e-05, "loss": 0.4836, "mean_token_accuracy": 0.8679071664810181, "num_tokens": 162932534.0, "step": 4288 }, { "epoch": 0.5456048848746978, "ewc_loss": 0.09033203125, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.677126407623291, "learning_rate": 1e-05, "loss": 0.5159, "mean_token_accuracy": 0.8592817187309265, "num_tokens": 162973199.0, "step": 4289 }, { "epoch": 0.5457320951532884, "ewc_loss": 0.08349609375, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 7.43865966796875e-05, "grad_norm": 5.450681686401367, "learning_rate": 1e-05, "loss": 0.537, "mean_token_accuracy": 0.8531931638717651, "num_tokens": 163007018.0, "step": 4290 }, { "epoch": 0.5458593054318789, "ewc_loss": 0.08837890625, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.663624286651611, "learning_rate": 1e-05, "loss": 0.5386, "mean_token_accuracy": 0.8521242737770081, "num_tokens": 163044381.0, "step": 4291 }, { "epoch": 0.5459865157104694, "ewc_loss": 0.08056640625, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 7.152557373046875e-05, "grad_norm": 5.265242099761963, "learning_rate": 1e-05, "loss": 0.519, "mean_token_accuracy": 0.8558564186096191, "num_tokens": 163079606.0, "step": 4292 }, { "epoch": 0.5461137259890599, "ewc_loss": 0.0888671875, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.679227352142334, "learning_rate": 1e-05, "loss": 0.5304, "mean_token_accuracy": 0.8516916632652283, "num_tokens": 163120421.0, "step": 4293 }, { "epoch": 0.5462409362676505, "ewc_loss": 0.080078125, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 7.05718994140625e-05, "grad_norm": 5.134158611297607, "learning_rate": 1e-05, "loss": 0.5683, "mean_token_accuracy": 0.8370697498321533, "num_tokens": 163164761.0, "step": 4294 }, { "epoch": 0.5463681465462409, "ewc_loss": 0.0888671875, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.779615879058838, "learning_rate": 1e-05, "loss": 0.6655, "mean_token_accuracy": 0.8135879039764404, "num_tokens": 163212432.0, "step": 4295 }, { "epoch": 0.5464953568248314, "ewc_loss": 0.0791015625, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 6.961822509765625e-05, "grad_norm": 5.0134687423706055, "learning_rate": 1e-05, "loss": 0.4545, "mean_token_accuracy": 0.8736844658851624, "num_tokens": 163251333.0, "step": 4296 }, { "epoch": 0.5466225671034219, "ewc_loss": 0.09326171875, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.989709377288818, "learning_rate": 1e-05, "loss": 0.5315, "mean_token_accuracy": 0.8538563847541809, "num_tokens": 163286428.0, "step": 4297 }, { "epoch": 0.5467497773820125, "ewc_loss": 0.07861328125, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 6.914138793945312e-05, "grad_norm": 4.9207658767700195, "learning_rate": 1e-05, "loss": 0.4842, "mean_token_accuracy": 0.8673366904258728, "num_tokens": 163318974.0, "step": 4298 }, { "epoch": 0.546876987660603, "ewc_loss": 0.10009765625, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.295563697814941, "learning_rate": 1e-05, "loss": 0.5515, "mean_token_accuracy": 0.8491030335426331, "num_tokens": 163358621.0, "step": 4299 }, { "epoch": 0.5470041979391935, "ewc_loss": 0.08056640625, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 7.104873657226562e-05, "grad_norm": 4.933786869049072, "learning_rate": 1e-05, "loss": 0.4837, "mean_token_accuracy": 0.8629153966903687, "num_tokens": 163401062.0, "step": 4300 }, { "epoch": 0.5471314082177839, "ewc_loss": 0.10791015625, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.650441646575928, "learning_rate": 1e-05, "loss": 0.5584, "mean_token_accuracy": 0.8502399921417236, "num_tokens": 163436194.0, "step": 4301 }, { "epoch": 0.5472586184963745, "ewc_loss": 0.08837890625, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.246743679046631, "learning_rate": 1e-05, "loss": 0.5295, "mean_token_accuracy": 0.8495062589645386, "num_tokens": 163476403.0, "step": 4302 }, { "epoch": 0.547385828774965, "ewc_loss": 0.10986328125, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.731714248657227, "learning_rate": 1e-05, "loss": 0.5446, "mean_token_accuracy": 0.8528589010238647, "num_tokens": 163515822.0, "step": 4303 }, { "epoch": 0.5475130390535555, "ewc_loss": 0.09716796875, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 5.642760276794434, "learning_rate": 1e-05, "loss": 0.5628, "mean_token_accuracy": 0.8445218801498413, "num_tokens": 163553611.0, "step": 4304 }, { "epoch": 0.5476402493321461, "ewc_loss": 0.103515625, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.425499439239502, "learning_rate": 1e-05, "loss": 0.5142, "mean_token_accuracy": 0.8611131310462952, "num_tokens": 163588591.0, "step": 4305 }, { "epoch": 0.5477674596107366, "ewc_loss": 0.09716796875, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 5.75984001159668, "learning_rate": 1e-05, "loss": 0.5197, "mean_token_accuracy": 0.8571780323982239, "num_tokens": 163630857.0, "step": 4306 }, { "epoch": 0.5478946698893271, "ewc_loss": 0.09228515625, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 6.02646017074585, "learning_rate": 1e-05, "loss": 0.5445, "mean_token_accuracy": 0.8463061451911926, "num_tokens": 163666684.0, "step": 4307 }, { "epoch": 0.5480218801679175, "ewc_loss": 0.0927734375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.717629432678223, "learning_rate": 1e-05, "loss": 0.5174, "mean_token_accuracy": 0.8527161478996277, "num_tokens": 163707409.0, "step": 4308 }, { "epoch": 0.5481490904465081, "ewc_loss": 0.08544921875, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 7.62939453125e-05, "grad_norm": 5.660855770111084, "learning_rate": 1e-05, "loss": 0.494, "mean_token_accuracy": 0.8611841201782227, "num_tokens": 163745190.0, "step": 4309 }, { "epoch": 0.5482763007250986, "ewc_loss": 0.09130859375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.692133903503418, "learning_rate": 1e-05, "loss": 0.5292, "mean_token_accuracy": 0.8542640209197998, "num_tokens": 163781538.0, "step": 4310 }, { "epoch": 0.5484035110036891, "ewc_loss": 0.08251953125, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 7.343292236328125e-05, "grad_norm": 5.405288219451904, "learning_rate": 1e-05, "loss": 0.5086, "mean_token_accuracy": 0.8535565137863159, "num_tokens": 163816412.0, "step": 4311 }, { "epoch": 0.5485307212822796, "ewc_loss": 0.087890625, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.678184986114502, "learning_rate": 1e-05, "loss": 0.4991, "mean_token_accuracy": 0.8631327748298645, "num_tokens": 163853202.0, "step": 4312 }, { "epoch": 0.5486579315608702, "ewc_loss": 0.08056640625, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 7.152557373046875e-05, "grad_norm": 5.232797145843506, "learning_rate": 1e-05, "loss": 0.5267, "mean_token_accuracy": 0.8532437086105347, "num_tokens": 163893669.0, "step": 4313 }, { "epoch": 0.5487851418394606, "ewc_loss": 0.0888671875, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.719161510467529, "learning_rate": 1e-05, "loss": 0.5114, "mean_token_accuracy": 0.8568676114082336, "num_tokens": 163932372.0, "step": 4314 }, { "epoch": 0.5489123521180511, "ewc_loss": 0.07958984375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 7.009506225585938e-05, "grad_norm": 5.099289417266846, "learning_rate": 1e-05, "loss": 0.4765, "mean_token_accuracy": 0.867828369140625, "num_tokens": 163969431.0, "step": 4315 }, { "epoch": 0.5490395623966416, "ewc_loss": 0.08984375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.855791091918945, "learning_rate": 1e-05, "loss": 0.5325, "mean_token_accuracy": 0.8552742004394531, "num_tokens": 164006776.0, "step": 4316 }, { "epoch": 0.5491667726752322, "ewc_loss": 0.0791015625, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 6.961822509765625e-05, "grad_norm": 4.973422527313232, "learning_rate": 1e-05, "loss": 0.4661, "mean_token_accuracy": 0.8693701028823853, "num_tokens": 164042178.0, "step": 4317 }, { "epoch": 0.5492939829538227, "ewc_loss": 0.09521484375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 6.0930867195129395, "learning_rate": 1e-05, "loss": 0.5537, "mean_token_accuracy": 0.8483801484107971, "num_tokens": 164089028.0, "step": 4318 }, { "epoch": 0.5494211932324132, "ewc_loss": 0.07861328125, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 6.914138793945312e-05, "grad_norm": 4.910682678222656, "learning_rate": 1e-05, "loss": 0.5375, "mean_token_accuracy": 0.8545821309089661, "num_tokens": 164130311.0, "step": 4319 }, { "epoch": 0.5495484035110036, "ewc_loss": 0.10302734375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.423521041870117, "learning_rate": 1e-05, "loss": 0.5427, "mean_token_accuracy": 0.8527127504348755, "num_tokens": 164165398.0, "step": 4320 }, { "epoch": 0.5496756137895942, "ewc_loss": 0.08251953125, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 7.343292236328125e-05, "grad_norm": 5.023846626281738, "learning_rate": 1e-05, "loss": 0.4878, "mean_token_accuracy": 0.8663712739944458, "num_tokens": 164202724.0, "step": 4321 }, { "epoch": 0.5498028240681847, "ewc_loss": 0.10888671875, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.692640781402588, "learning_rate": 1e-05, "loss": 0.5914, "mean_token_accuracy": 0.8400594592094421, "num_tokens": 164247666.0, "step": 4322 }, { "epoch": 0.5499300343467752, "ewc_loss": 0.09033203125, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.388728141784668, "learning_rate": 1e-05, "loss": 0.5262, "mean_token_accuracy": 0.8536214828491211, "num_tokens": 164292323.0, "step": 4323 }, { "epoch": 0.5500572446253658, "ewc_loss": 0.107421875, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.617700099945068, "learning_rate": 1e-05, "loss": 0.4869, "mean_token_accuracy": 0.8708277344703674, "num_tokens": 164331160.0, "step": 4324 }, { "epoch": 0.5501844549039563, "ewc_loss": 0.09716796875, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 5.685501575469971, "learning_rate": 1e-05, "loss": 0.4743, "mean_token_accuracy": 0.8712666034698486, "num_tokens": 164365457.0, "step": 4325 }, { "epoch": 0.5503116651825467, "ewc_loss": 0.10107421875, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.275825023651123, "learning_rate": 1e-05, "loss": 0.5542, "mean_token_accuracy": 0.8494310975074768, "num_tokens": 164403731.0, "step": 4326 }, { "epoch": 0.5504388754611372, "ewc_loss": 0.0947265625, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.740920066833496, "learning_rate": 1e-05, "loss": 0.5072, "mean_token_accuracy": 0.8608658313751221, "num_tokens": 164436616.0, "step": 4327 }, { "epoch": 0.5505660857397278, "ewc_loss": 0.0888671875, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.87398624420166, "learning_rate": 1e-05, "loss": 0.4934, "mean_token_accuracy": 0.859648585319519, "num_tokens": 164472453.0, "step": 4328 }, { "epoch": 0.5506932960183183, "ewc_loss": 0.09130859375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.698111057281494, "learning_rate": 1e-05, "loss": 0.5351, "mean_token_accuracy": 0.8479295372962952, "num_tokens": 164506551.0, "step": 4329 }, { "epoch": 0.5508205062969088, "ewc_loss": 0.083984375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 7.486343383789062e-05, "grad_norm": 5.552391052246094, "learning_rate": 1e-05, "loss": 0.4597, "mean_token_accuracy": 0.8709956407546997, "num_tokens": 164542638.0, "step": 4330 }, { "epoch": 0.5509477165754993, "ewc_loss": 0.08984375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.662804126739502, "learning_rate": 1e-05, "loss": 0.523, "mean_token_accuracy": 0.857728123664856, "num_tokens": 164580071.0, "step": 4331 }, { "epoch": 0.5510749268540898, "ewc_loss": 0.0810546875, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 7.200241088867188e-05, "grad_norm": 5.354699611663818, "learning_rate": 1e-05, "loss": 0.5572, "mean_token_accuracy": 0.8418318033218384, "num_tokens": 164617894.0, "step": 4332 }, { "epoch": 0.5512021371326803, "ewc_loss": 0.087890625, "ewc_loss_diag": 9.250640869140625e-05, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.6514458656311035, "learning_rate": 1e-05, "loss": 0.5937, "mean_token_accuracy": 0.8338767290115356, "num_tokens": 164660509.0, "step": 4333 }, { "epoch": 0.5513293474112708, "ewc_loss": 0.08056640625, "ewc_loss_diag": 9.250640869140625e-05, "ewc_loss_parallel": 7.104873657226562e-05, "grad_norm": 5.209312438964844, "learning_rate": 1e-05, "loss": 0.4664, "mean_token_accuracy": 0.8671476244926453, "num_tokens": 164699683.0, "step": 4334 }, { "epoch": 0.5514565576898613, "ewc_loss": 0.0888671875, "ewc_loss_diag": 9.250640869140625e-05, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.710239887237549, "learning_rate": 1e-05, "loss": 0.5268, "mean_token_accuracy": 0.8548330664634705, "num_tokens": 164733134.0, "step": 4335 }, { "epoch": 0.5515837679684519, "ewc_loss": 0.07958984375, "ewc_loss_diag": 9.250640869140625e-05, "ewc_loss_parallel": 7.009506225585938e-05, "grad_norm": 5.093115329742432, "learning_rate": 1e-05, "loss": 0.5273, "mean_token_accuracy": 0.8497284054756165, "num_tokens": 164771911.0, "step": 4336 }, { "epoch": 0.5517109782470424, "ewc_loss": 0.08984375, "ewc_loss_diag": 9.250640869140625e-05, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.837819576263428, "learning_rate": 1e-05, "loss": 0.5447, "mean_token_accuracy": 0.8466901779174805, "num_tokens": 164810579.0, "step": 4337 }, { "epoch": 0.5518381885256328, "ewc_loss": 0.078125, "ewc_loss_diag": 9.250640869140625e-05, "ewc_loss_parallel": 6.866455078125e-05, "grad_norm": 4.963930130004883, "learning_rate": 1e-05, "loss": 0.463, "mean_token_accuracy": 0.8662216663360596, "num_tokens": 164850215.0, "step": 4338 }, { "epoch": 0.5519653988042234, "ewc_loss": 0.09521484375, "ewc_loss_diag": 9.250640869140625e-05, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 6.088682174682617, "learning_rate": 1e-05, "loss": 0.5362, "mean_token_accuracy": 0.8502187728881836, "num_tokens": 164889523.0, "step": 4339 }, { "epoch": 0.5520926090828139, "ewc_loss": 0.078125, "ewc_loss_diag": 9.250640869140625e-05, "ewc_loss_parallel": 6.866455078125e-05, "grad_norm": 4.889826774597168, "learning_rate": 1e-05, "loss": 0.4572, "mean_token_accuracy": 0.8722853660583496, "num_tokens": 164925946.0, "step": 4340 }, { "epoch": 0.5522198193614044, "ewc_loss": 0.103515625, "ewc_loss_diag": 9.250640869140625e-05, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.443105220794678, "learning_rate": 1e-05, "loss": 0.5162, "mean_token_accuracy": 0.8633109331130981, "num_tokens": 164961954.0, "step": 4341 }, { "epoch": 0.5523470296399949, "ewc_loss": 0.08154296875, "ewc_loss_diag": 9.250640869140625e-05, "ewc_loss_parallel": 7.2479248046875e-05, "grad_norm": 5.001123905181885, "learning_rate": 1e-05, "loss": 0.5004, "mean_token_accuracy": 0.8593125343322754, "num_tokens": 165008013.0, "step": 4342 }, { "epoch": 0.5524742399185855, "ewc_loss": 0.11083984375, "ewc_loss_diag": 9.250640869140625e-05, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.743644714355469, "learning_rate": 1e-05, "loss": 0.5216, "mean_token_accuracy": 0.8615124821662903, "num_tokens": 165037723.0, "step": 4343 }, { "epoch": 0.5526014501971759, "ewc_loss": 0.09130859375, "ewc_loss_diag": 9.250640869140625e-05, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.420290470123291, "learning_rate": 1e-05, "loss": 0.5086, "mean_token_accuracy": 0.862564206123352, "num_tokens": 165071432.0, "step": 4344 }, { "epoch": 0.5527286604757664, "ewc_loss": 0.1083984375, "ewc_loss_diag": 9.250640869140625e-05, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.654978275299072, "learning_rate": 1e-05, "loss": 0.5124, "mean_token_accuracy": 0.8643330335617065, "num_tokens": 165118784.0, "step": 4345 }, { "epoch": 0.5528558707543569, "ewc_loss": 0.09814453125, "ewc_loss_diag": 9.250640869140625e-05, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 5.7272114753723145, "learning_rate": 1e-05, "loss": 0.5852, "mean_token_accuracy": 0.8402488827705383, "num_tokens": 165163281.0, "step": 4346 }, { "epoch": 0.5529830810329475, "ewc_loss": 0.10107421875, "ewc_loss_diag": 9.250640869140625e-05, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.284496307373047, "learning_rate": 1e-05, "loss": 0.5436, "mean_token_accuracy": 0.8485926985740662, "num_tokens": 165195615.0, "step": 4347 }, { "epoch": 0.553110291311538, "ewc_loss": 0.09619140625, "ewc_loss_diag": 9.250640869140625e-05, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.773552417755127, "learning_rate": 1e-05, "loss": 0.531, "mean_token_accuracy": 0.8529706001281738, "num_tokens": 165233710.0, "step": 4348 }, { "epoch": 0.5532375015901285, "ewc_loss": 0.08935546875, "ewc_loss_diag": 9.250640869140625e-05, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.866718292236328, "learning_rate": 1e-05, "loss": 0.5174, "mean_token_accuracy": 0.8536308407783508, "num_tokens": 165275788.0, "step": 4349 }, { "epoch": 0.5533647118687189, "ewc_loss": 0.091796875, "ewc_loss_diag": 9.250640869140625e-05, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.724057197570801, "learning_rate": 1e-05, "loss": 0.4909, "mean_token_accuracy": 0.8655706644058228, "num_tokens": 165316627.0, "step": 4350 }, { "epoch": 0.5534919221473095, "ewc_loss": 0.08447265625, "ewc_loss_diag": 9.250640869140625e-05, "ewc_loss_parallel": 7.534027099609375e-05, "grad_norm": 5.53781270980835, "learning_rate": 1e-05, "loss": 0.5139, "mean_token_accuracy": 0.8571009039878845, "num_tokens": 165356643.0, "step": 4351 }, { "epoch": 0.5536191324259, "ewc_loss": 0.09033203125, "ewc_loss_diag": 9.250640869140625e-05, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.705519676208496, "learning_rate": 1e-05, "loss": 0.5343, "mean_token_accuracy": 0.8508484959602356, "num_tokens": 165396897.0, "step": 4352 }, { "epoch": 0.5537463427044905, "ewc_loss": 0.08154296875, "ewc_loss_diag": 9.250640869140625e-05, "ewc_loss_parallel": 7.2479248046875e-05, "grad_norm": 5.301980018615723, "learning_rate": 1e-05, "loss": 0.4912, "mean_token_accuracy": 0.8606736660003662, "num_tokens": 165434409.0, "step": 4353 }, { "epoch": 0.553873552983081, "ewc_loss": 0.08935546875, "ewc_loss_diag": 9.250640869140625e-05, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.741624355316162, "learning_rate": 1e-05, "loss": 0.5101, "mean_token_accuracy": 0.8569148778915405, "num_tokens": 165474674.0, "step": 4354 }, { "epoch": 0.5540007632616716, "ewc_loss": 0.07958984375, "ewc_loss_diag": 9.250640869140625e-05, "ewc_loss_parallel": 7.009506225585938e-05, "grad_norm": 5.164858341217041, "learning_rate": 1e-05, "loss": 0.5497, "mean_token_accuracy": 0.8442662954330444, "num_tokens": 165511865.0, "step": 4355 }, { "epoch": 0.554127973540262, "ewc_loss": 0.08984375, "ewc_loss_diag": 9.250640869140625e-05, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.812569618225098, "learning_rate": 1e-05, "loss": 0.5364, "mean_token_accuracy": 0.8545852899551392, "num_tokens": 165548589.0, "step": 4356 }, { "epoch": 0.5542551838188525, "ewc_loss": 0.0791015625, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 6.961822509765625e-05, "grad_norm": 5.034230709075928, "learning_rate": 1e-05, "loss": 0.5033, "mean_token_accuracy": 0.8608754873275757, "num_tokens": 165589553.0, "step": 4357 }, { "epoch": 0.554382394097443, "ewc_loss": 0.09326171875, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 6.013623237609863, "learning_rate": 1e-05, "loss": 0.6137, "mean_token_accuracy": 0.8314615488052368, "num_tokens": 165625499.0, "step": 4358 }, { "epoch": 0.5545096043760336, "ewc_loss": 0.07861328125, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 6.914138793945312e-05, "grad_norm": 4.931924343109131, "learning_rate": 1e-05, "loss": 0.5045, "mean_token_accuracy": 0.8569453954696655, "num_tokens": 165663967.0, "step": 4359 }, { "epoch": 0.5546368146546241, "ewc_loss": 0.10107421875, "ewc_loss_diag": 9.250640869140625e-05, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.329815864562988, "learning_rate": 1e-05, "loss": 0.5603, "mean_token_accuracy": 0.8504183292388916, "num_tokens": 165706382.0, "step": 4360 }, { "epoch": 0.5547640249332146, "ewc_loss": 0.08056640625, "ewc_loss_diag": 9.250640869140625e-05, "ewc_loss_parallel": 7.152557373046875e-05, "grad_norm": 4.9759521484375, "learning_rate": 1e-05, "loss": 0.4998, "mean_token_accuracy": 0.858593225479126, "num_tokens": 165742473.0, "step": 4361 }, { "epoch": 0.5548912352118052, "ewc_loss": 0.10888671875, "ewc_loss_diag": 9.250640869140625e-05, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.651897430419922, "learning_rate": 1e-05, "loss": 0.5487, "mean_token_accuracy": 0.8505123853683472, "num_tokens": 165776329.0, "step": 4362 }, { "epoch": 0.5550184454903956, "ewc_loss": 0.0888671875, "ewc_loss_diag": 9.250640869140625e-05, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.290587902069092, "learning_rate": 1e-05, "loss": 0.5371, "mean_token_accuracy": 0.8519763350486755, "num_tokens": 165812887.0, "step": 4363 }, { "epoch": 0.5551456557689861, "ewc_loss": 0.10888671875, "ewc_loss_diag": 9.250640869140625e-05, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.675780773162842, "learning_rate": 1e-05, "loss": 0.5126, "mean_token_accuracy": 0.8664180040359497, "num_tokens": 165848933.0, "step": 4364 }, { "epoch": 0.5552728660475766, "ewc_loss": 0.09521484375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.643271446228027, "learning_rate": 1e-05, "loss": 0.5736, "mean_token_accuracy": 0.841184139251709, "num_tokens": 165887158.0, "step": 4365 }, { "epoch": 0.5554000763261672, "ewc_loss": 0.1025390625, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.377830982208252, "learning_rate": 1e-05, "loss": 0.5463, "mean_token_accuracy": 0.8537015914916992, "num_tokens": 165928214.0, "step": 4366 }, { "epoch": 0.5555272866047577, "ewc_loss": 0.095703125, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.73866605758667, "learning_rate": 1e-05, "loss": 0.5316, "mean_token_accuracy": 0.8534673452377319, "num_tokens": 165964838.0, "step": 4367 }, { "epoch": 0.5556544968833482, "ewc_loss": 0.09130859375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.994195461273193, "learning_rate": 1e-05, "loss": 0.5553, "mean_token_accuracy": 0.8467740416526794, "num_tokens": 165996097.0, "step": 4368 }, { "epoch": 0.5557817071619386, "ewc_loss": 0.0927734375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.706514358520508, "learning_rate": 1e-05, "loss": 0.5428, "mean_token_accuracy": 0.8539088368415833, "num_tokens": 166028660.0, "step": 4369 }, { "epoch": 0.5559089174405292, "ewc_loss": 0.08544921875, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 7.62939453125e-05, "grad_norm": 5.646689414978027, "learning_rate": 1e-05, "loss": 0.5245, "mean_token_accuracy": 0.8560070991516113, "num_tokens": 166065259.0, "step": 4370 }, { "epoch": 0.5560361277191197, "ewc_loss": 0.09033203125, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.658794403076172, "learning_rate": 1e-05, "loss": 0.5248, "mean_token_accuracy": 0.8525662422180176, "num_tokens": 166105757.0, "step": 4371 }, { "epoch": 0.5561633379977102, "ewc_loss": 0.08251953125, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 7.343292236328125e-05, "grad_norm": 5.424644947052002, "learning_rate": 1e-05, "loss": 0.5503, "mean_token_accuracy": 0.8400270938873291, "num_tokens": 166142362.0, "step": 4372 }, { "epoch": 0.5562905482763008, "ewc_loss": 0.087890625, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.630918025970459, "learning_rate": 1e-05, "loss": 0.5193, "mean_token_accuracy": 0.8559409379959106, "num_tokens": 166179437.0, "step": 4373 }, { "epoch": 0.5564177585548913, "ewc_loss": 0.08056640625, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 7.152557373046875e-05, "grad_norm": 5.281700611114502, "learning_rate": 1e-05, "loss": 0.5179, "mean_token_accuracy": 0.8490195870399475, "num_tokens": 166218555.0, "step": 4374 }, { "epoch": 0.5565449688334817, "ewc_loss": 0.08740234375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.641168594360352, "learning_rate": 1e-05, "loss": 0.4884, "mean_token_accuracy": 0.8650078773498535, "num_tokens": 166252297.0, "step": 4375 }, { "epoch": 0.5566721791120722, "ewc_loss": 0.07958984375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 7.009506225585938e-05, "grad_norm": 5.173190116882324, "learning_rate": 1e-05, "loss": 0.4976, "mean_token_accuracy": 0.8598741292953491, "num_tokens": 166289597.0, "step": 4376 }, { "epoch": 0.5567993893906628, "ewc_loss": 0.08837890625, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.701639652252197, "learning_rate": 1e-05, "loss": 0.5494, "mean_token_accuracy": 0.845864474773407, "num_tokens": 166328655.0, "step": 4377 }, { "epoch": 0.5569265996692533, "ewc_loss": 0.0791015625, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 6.961822509765625e-05, "grad_norm": 5.057477951049805, "learning_rate": 1e-05, "loss": 0.5128, "mean_token_accuracy": 0.8528243899345398, "num_tokens": 166371825.0, "step": 4378 }, { "epoch": 0.5570538099478438, "ewc_loss": 0.09033203125, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.873587131500244, "learning_rate": 1e-05, "loss": 0.5141, "mean_token_accuracy": 0.8549611568450928, "num_tokens": 166407084.0, "step": 4379 }, { "epoch": 0.5571810202264343, "ewc_loss": 0.078125, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 6.866455078125e-05, "grad_norm": 4.925075531005859, "learning_rate": 1e-05, "loss": 0.4816, "mean_token_accuracy": 0.8623501658439636, "num_tokens": 166442272.0, "step": 4380 }, { "epoch": 0.5573082305050248, "ewc_loss": 0.09765625, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 6.1721601486206055, "learning_rate": 1e-05, "loss": 0.5567, "mean_token_accuracy": 0.8469924330711365, "num_tokens": 166475015.0, "step": 4381 }, { "epoch": 0.5574354407836153, "ewc_loss": 0.07861328125, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 6.914138793945312e-05, "grad_norm": 4.862576484680176, "learning_rate": 1e-05, "loss": 0.5019, "mean_token_accuracy": 0.8571680784225464, "num_tokens": 166512421.0, "step": 4382 }, { "epoch": 0.5575626510622058, "ewc_loss": 0.10693359375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.580226421356201, "learning_rate": 1e-05, "loss": 0.5572, "mean_token_accuracy": 0.8474609851837158, "num_tokens": 166554248.0, "step": 4383 }, { "epoch": 0.5576898613407963, "ewc_loss": 0.0849609375, "ewc_loss_diag": 9.393692016601562e-05, "ewc_loss_parallel": 7.581710815429688e-05, "grad_norm": 5.073015213012695, "learning_rate": 1e-05, "loss": 0.4589, "mean_token_accuracy": 0.8718898296356201, "num_tokens": 166590664.0, "step": 4384 }, { "epoch": 0.5578170716193869, "ewc_loss": 0.11474609375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.8485307693481445, "learning_rate": 1e-05, "loss": 0.5517, "mean_token_accuracy": 0.8542698621749878, "num_tokens": 166625615.0, "step": 4385 }, { "epoch": 0.5579442818979774, "ewc_loss": 0.09521484375, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.561331748962402, "learning_rate": 1e-05, "loss": 0.5244, "mean_token_accuracy": 0.8607538938522339, "num_tokens": 166662699.0, "step": 4386 }, { "epoch": 0.5580714921765678, "ewc_loss": 0.10888671875, "ewc_loss_diag": 9.34600830078125e-05, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.655323028564453, "learning_rate": 1e-05, "loss": 0.5151, "mean_token_accuracy": 0.8623707294464111, "num_tokens": 166701218.0, "step": 4387 }, { "epoch": 0.5581987024551583, "ewc_loss": 0.1005859375, "ewc_loss_diag": 9.393692016601562e-05, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 5.817807197570801, "learning_rate": 1e-05, "loss": 0.5379, "mean_token_accuracy": 0.8542234897613525, "num_tokens": 166739699.0, "step": 4388 }, { "epoch": 0.5583259127337489, "ewc_loss": 0.1005859375, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.255162239074707, "learning_rate": 1e-05, "loss": 0.5128, "mean_token_accuracy": 0.863952100276947, "num_tokens": 166782275.0, "step": 4389 }, { "epoch": 0.5584531230123394, "ewc_loss": 0.0966796875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 5.778129577636719, "learning_rate": 1e-05, "loss": 0.5246, "mean_token_accuracy": 0.8601733446121216, "num_tokens": 166820535.0, "step": 4390 }, { "epoch": 0.5585803332909299, "ewc_loss": 0.08935546875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.884608268737793, "learning_rate": 1e-05, "loss": 0.5606, "mean_token_accuracy": 0.8410179615020752, "num_tokens": 166865450.0, "step": 4391 }, { "epoch": 0.5587075435695205, "ewc_loss": 0.09130859375, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.702263832092285, "learning_rate": 1e-05, "loss": 0.5089, "mean_token_accuracy": 0.8592230081558228, "num_tokens": 166905838.0, "step": 4392 }, { "epoch": 0.5588347538481109, "ewc_loss": 0.08447265625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 7.534027099609375e-05, "grad_norm": 5.6068220138549805, "learning_rate": 1e-05, "loss": 0.5612, "mean_token_accuracy": 0.8423203825950623, "num_tokens": 166943437.0, "step": 4393 }, { "epoch": 0.5589619641267014, "ewc_loss": 0.087890625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.635476112365723, "learning_rate": 1e-05, "loss": 0.4993, "mean_token_accuracy": 0.8617465496063232, "num_tokens": 166980863.0, "step": 4394 }, { "epoch": 0.5590891744052919, "ewc_loss": 0.083984375, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 7.486343383789062e-05, "grad_norm": 5.422138214111328, "learning_rate": 1e-05, "loss": 0.4627, "mean_token_accuracy": 0.8721214532852173, "num_tokens": 167017861.0, "step": 4395 }, { "epoch": 0.5592163846838825, "ewc_loss": 0.0859375, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.5961079597473145, "learning_rate": 1e-05, "loss": 0.473, "mean_token_accuracy": 0.8662816882133484, "num_tokens": 167051779.0, "step": 4396 }, { "epoch": 0.559343594962473, "ewc_loss": 0.08056640625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 7.152557373046875e-05, "grad_norm": 5.290031433105469, "learning_rate": 1e-05, "loss": 0.5117, "mean_token_accuracy": 0.8560145497322083, "num_tokens": 167095205.0, "step": 4397 }, { "epoch": 0.5594708052410635, "ewc_loss": 0.0859375, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.61489200592041, "learning_rate": 1e-05, "loss": 0.5311, "mean_token_accuracy": 0.8503556251525879, "num_tokens": 167130594.0, "step": 4398 }, { "epoch": 0.5595980155196539, "ewc_loss": 0.080078125, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 7.05718994140625e-05, "grad_norm": 5.16717004776001, "learning_rate": 1e-05, "loss": 0.5188, "mean_token_accuracy": 0.8554620146751404, "num_tokens": 167170816.0, "step": 4399 }, { "epoch": 0.5597252257982445, "ewc_loss": 0.0869140625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.702080249786377, "learning_rate": 1e-05, "loss": 0.4804, "mean_token_accuracy": 0.8660013675689697, "num_tokens": 167205301.0, "step": 4400 }, { "epoch": 0.559852436076835, "ewc_loss": 0.07861328125, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 6.914138793945312e-05, "grad_norm": 5.0460896492004395, "learning_rate": 1e-05, "loss": 0.5497, "mean_token_accuracy": 0.8414177894592285, "num_tokens": 167240356.0, "step": 4401 }, { "epoch": 0.5599796463554255, "ewc_loss": 0.0908203125, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.8808465003967285, "learning_rate": 1e-05, "loss": 0.5032, "mean_token_accuracy": 0.8621690273284912, "num_tokens": 167269853.0, "step": 4402 }, { "epoch": 0.560106856634016, "ewc_loss": 0.07763671875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 6.818771362304688e-05, "grad_norm": 4.898162841796875, "learning_rate": 1e-05, "loss": 0.4968, "mean_token_accuracy": 0.859675407409668, "num_tokens": 167315398.0, "step": 4403 }, { "epoch": 0.5602340669126066, "ewc_loss": 0.10009765625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.229650497436523, "learning_rate": 1e-05, "loss": 0.5191, "mean_token_accuracy": 0.8590660095214844, "num_tokens": 167352715.0, "step": 4404 }, { "epoch": 0.560361277191197, "ewc_loss": 0.07861328125, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 6.914138793945312e-05, "grad_norm": 4.837011337280273, "learning_rate": 1e-05, "loss": 0.562, "mean_token_accuracy": 0.8424108624458313, "num_tokens": 167391646.0, "step": 4405 }, { "epoch": 0.5604884874697875, "ewc_loss": 0.109375, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.706264972686768, "learning_rate": 1e-05, "loss": 0.551, "mean_token_accuracy": 0.853929877281189, "num_tokens": 167431676.0, "step": 4406 }, { "epoch": 0.560615697748378, "ewc_loss": 0.0869140625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.120028972625732, "learning_rate": 1e-05, "loss": 0.5378, "mean_token_accuracy": 0.8494259119033813, "num_tokens": 167469121.0, "step": 4407 }, { "epoch": 0.5607429080269686, "ewc_loss": 0.1181640625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 6.984330654144287, "learning_rate": 1e-05, "loss": 0.5362, "mean_token_accuracy": 0.8601405620574951, "num_tokens": 167511721.0, "step": 4408 }, { "epoch": 0.5608701183055591, "ewc_loss": 0.10107421875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 5.711796760559082, "learning_rate": 1e-05, "loss": 0.5117, "mean_token_accuracy": 0.864598274230957, "num_tokens": 167546965.0, "step": 4409 }, { "epoch": 0.5609973285841496, "ewc_loss": 0.109375, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.722989082336426, "learning_rate": 1e-05, "loss": 0.5856, "mean_token_accuracy": 0.8428268432617188, "num_tokens": 167584968.0, "step": 4410 }, { "epoch": 0.5611245388627402, "ewc_loss": 0.103515625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 5.949914932250977, "learning_rate": 1e-05, "loss": 0.5826, "mean_token_accuracy": 0.8439831137657166, "num_tokens": 167623124.0, "step": 4411 }, { "epoch": 0.5612517491413306, "ewc_loss": 0.10009765625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.272328853607178, "learning_rate": 1e-05, "loss": 0.5375, "mean_token_accuracy": 0.852527379989624, "num_tokens": 167664877.0, "step": 4412 }, { "epoch": 0.5613789594199211, "ewc_loss": 0.09765625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 5.856210231781006, "learning_rate": 1e-05, "loss": 0.5418, "mean_token_accuracy": 0.8505493402481079, "num_tokens": 167706154.0, "step": 4413 }, { "epoch": 0.5615061696985116, "ewc_loss": 0.08837890625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.913712978363037, "learning_rate": 1e-05, "loss": 0.5224, "mean_token_accuracy": 0.852965235710144, "num_tokens": 167740081.0, "step": 4414 }, { "epoch": 0.5616333799771022, "ewc_loss": 0.09130859375, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.729766845703125, "learning_rate": 1e-05, "loss": 0.5875, "mean_token_accuracy": 0.8371289968490601, "num_tokens": 167783324.0, "step": 4415 }, { "epoch": 0.5617605902556927, "ewc_loss": 0.08544921875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 7.62939453125e-05, "grad_norm": 5.644023895263672, "learning_rate": 1e-05, "loss": 0.4978, "mean_token_accuracy": 0.8589739799499512, "num_tokens": 167823216.0, "step": 4416 }, { "epoch": 0.5618878005342832, "ewc_loss": 0.087890625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.63585090637207, "learning_rate": 1e-05, "loss": 0.5694, "mean_token_accuracy": 0.8370522260665894, "num_tokens": 167857452.0, "step": 4417 }, { "epoch": 0.5620150108128736, "ewc_loss": 0.0830078125, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 7.390975952148438e-05, "grad_norm": 5.48289155960083, "learning_rate": 1e-05, "loss": 0.4904, "mean_token_accuracy": 0.8625504970550537, "num_tokens": 167896402.0, "step": 4418 }, { "epoch": 0.5621422210914642, "ewc_loss": 0.0849609375, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 7.581710815429688e-05, "grad_norm": 5.541975975036621, "learning_rate": 1e-05, "loss": 0.5039, "mean_token_accuracy": 0.8583027720451355, "num_tokens": 167936441.0, "step": 4419 }, { "epoch": 0.5622694313700547, "ewc_loss": 0.08154296875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 7.2479248046875e-05, "grad_norm": 5.395539283752441, "learning_rate": 1e-05, "loss": 0.4864, "mean_token_accuracy": 0.8662829399108887, "num_tokens": 167970786.0, "step": 4420 }, { "epoch": 0.5623966416486452, "ewc_loss": 0.08203125, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 7.295608520507812e-05, "grad_norm": 5.487508773803711, "learning_rate": 1e-05, "loss": 0.5303, "mean_token_accuracy": 0.8544028997421265, "num_tokens": 168011004.0, "step": 4421 }, { "epoch": 0.5625238519272358, "ewc_loss": 0.08056640625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 7.152557373046875e-05, "grad_norm": 5.331375598907471, "learning_rate": 1e-05, "loss": 0.5072, "mean_token_accuracy": 0.853549599647522, "num_tokens": 168053418.0, "step": 4422 }, { "epoch": 0.5626510622058263, "ewc_loss": 0.08251953125, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 7.343292236328125e-05, "grad_norm": 5.46168327331543, "learning_rate": 1e-05, "loss": 0.5093, "mean_token_accuracy": 0.8623669147491455, "num_tokens": 168088156.0, "step": 4423 }, { "epoch": 0.5627782724844167, "ewc_loss": 0.08056640625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 7.152557373046875e-05, "grad_norm": 5.277639389038086, "learning_rate": 1e-05, "loss": 0.5425, "mean_token_accuracy": 0.8451347351074219, "num_tokens": 168132329.0, "step": 4424 }, { "epoch": 0.5629054827630072, "ewc_loss": 0.08349609375, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 7.43865966796875e-05, "grad_norm": 5.469374656677246, "learning_rate": 1e-05, "loss": 0.4926, "mean_token_accuracy": 0.8600599765777588, "num_tokens": 168167882.0, "step": 4425 }, { "epoch": 0.5630326930415978, "ewc_loss": 0.08056640625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 7.104873657226562e-05, "grad_norm": 5.2109293937683105, "learning_rate": 1e-05, "loss": 0.506, "mean_token_accuracy": 0.8568372130393982, "num_tokens": 168203265.0, "step": 4426 }, { "epoch": 0.5631599033201883, "ewc_loss": 0.0849609375, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 7.581710815429688e-05, "grad_norm": 5.524764537811279, "learning_rate": 1e-05, "loss": 0.473, "mean_token_accuracy": 0.8686162829399109, "num_tokens": 168241410.0, "step": 4427 }, { "epoch": 0.5632871135987788, "ewc_loss": 0.0791015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 6.961822509765625e-05, "grad_norm": 5.099390983581543, "learning_rate": 1e-05, "loss": 0.5027, "mean_token_accuracy": 0.8618127107620239, "num_tokens": 168285412.0, "step": 4428 }, { "epoch": 0.5634143238773693, "ewc_loss": 0.0869140625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.678112983703613, "learning_rate": 1e-05, "loss": 0.5262, "mean_token_accuracy": 0.8498134613037109, "num_tokens": 168324628.0, "step": 4429 }, { "epoch": 0.5635415341559598, "ewc_loss": 0.07763671875, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 6.818771362304688e-05, "grad_norm": 4.966049671173096, "learning_rate": 1e-05, "loss": 0.5414, "mean_token_accuracy": 0.8485806584358215, "num_tokens": 168359759.0, "step": 4430 }, { "epoch": 0.5636687444345503, "ewc_loss": 0.091796875, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.954592704772949, "learning_rate": 1e-05, "loss": 0.5297, "mean_token_accuracy": 0.8514320850372314, "num_tokens": 168393904.0, "step": 4431 }, { "epoch": 0.5637959547131408, "ewc_loss": 0.0771484375, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 6.771087646484375e-05, "grad_norm": 4.776290416717529, "learning_rate": 1e-05, "loss": 0.5268, "mean_token_accuracy": 0.8500932455062866, "num_tokens": 168429718.0, "step": 4432 }, { "epoch": 0.5639231649917313, "ewc_loss": 0.10400390625, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.470669269561768, "learning_rate": 1e-05, "loss": 0.4968, "mean_token_accuracy": 0.8658007383346558, "num_tokens": 168462950.0, "step": 4433 }, { "epoch": 0.5640503752703219, "ewc_loss": 0.07861328125, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 6.914138793945312e-05, "grad_norm": 4.744180202484131, "learning_rate": 1e-05, "loss": 0.52, "mean_token_accuracy": 0.8557376861572266, "num_tokens": 168504095.0, "step": 4434 }, { "epoch": 0.5641775855489124, "ewc_loss": 0.12158203125, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 0.0001125335693359375, "grad_norm": 7.177929878234863, "learning_rate": 1e-05, "loss": 0.5565, "mean_token_accuracy": 0.8518354892730713, "num_tokens": 168539931.0, "step": 4435 }, { "epoch": 0.5643047958275028, "ewc_loss": 0.0947265625, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.4357380867004395, "learning_rate": 1e-05, "loss": 0.5133, "mean_token_accuracy": 0.8608399629592896, "num_tokens": 168573064.0, "step": 4436 }, { "epoch": 0.5644320061060933, "ewc_loss": 0.12451171875, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 0.00011539459228515625, "grad_norm": 7.327329635620117, "learning_rate": 1e-05, "loss": 0.5569, "mean_token_accuracy": 0.8549712300300598, "num_tokens": 168607997.0, "step": 4437 }, { "epoch": 0.5645592163846839, "ewc_loss": 0.1142578125, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.220785617828369, "learning_rate": 1e-05, "loss": 0.4807, "mean_token_accuracy": 0.8758504986763, "num_tokens": 168647590.0, "step": 4438 }, { "epoch": 0.5646864266632744, "ewc_loss": 0.11083984375, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.766859531402588, "learning_rate": 1e-05, "loss": 0.5246, "mean_token_accuracy": 0.8601834774017334, "num_tokens": 168681347.0, "step": 4439 }, { "epoch": 0.5648136369418649, "ewc_loss": 0.10595703125, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.184383869171143, "learning_rate": 1e-05, "loss": 0.5346, "mean_token_accuracy": 0.8547519445419312, "num_tokens": 168722830.0, "step": 4440 }, { "epoch": 0.5649408472204555, "ewc_loss": 0.09814453125, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.370793342590332, "learning_rate": 1e-05, "loss": 0.584, "mean_token_accuracy": 0.8401820063591003, "num_tokens": 168762381.0, "step": 4441 }, { "epoch": 0.5650680574990459, "ewc_loss": 0.09619140625, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.898586273193359, "learning_rate": 1e-05, "loss": 0.5147, "mean_token_accuracy": 0.8598393201828003, "num_tokens": 168802832.0, "step": 4442 }, { "epoch": 0.5651952677776364, "ewc_loss": 0.09375, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 6.1140642166137695, "learning_rate": 1e-05, "loss": 0.53, "mean_token_accuracy": 0.8541038036346436, "num_tokens": 168837780.0, "step": 4443 }, { "epoch": 0.5653224780562269, "ewc_loss": 0.08935546875, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.667787075042725, "learning_rate": 1e-05, "loss": 0.5663, "mean_token_accuracy": 0.8460259437561035, "num_tokens": 168871830.0, "step": 4444 }, { "epoch": 0.5654496883348175, "ewc_loss": 0.09228515625, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.974826812744141, "learning_rate": 1e-05, "loss": 0.4986, "mean_token_accuracy": 0.8612262606620789, "num_tokens": 168912477.0, "step": 4445 }, { "epoch": 0.565576898613408, "ewc_loss": 0.0830078125, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 7.390975952148438e-05, "grad_norm": 5.451789855957031, "learning_rate": 1e-05, "loss": 0.5458, "mean_token_accuracy": 0.8466284871101379, "num_tokens": 168948976.0, "step": 4446 }, { "epoch": 0.5657041088919985, "ewc_loss": 0.09228515625, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.975351333618164, "learning_rate": 1e-05, "loss": 0.5403, "mean_token_accuracy": 0.8559730052947998, "num_tokens": 168992454.0, "step": 4447 }, { "epoch": 0.5658313191705889, "ewc_loss": 0.08056640625, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 7.104873657226562e-05, "grad_norm": 5.270195960998535, "learning_rate": 1e-05, "loss": 0.519, "mean_token_accuracy": 0.85194993019104, "num_tokens": 169023135.0, "step": 4448 }, { "epoch": 0.5659585294491795, "ewc_loss": 0.0947265625, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 6.1134467124938965, "learning_rate": 1e-05, "loss": 0.5228, "mean_token_accuracy": 0.8558865785598755, "num_tokens": 169062066.0, "step": 4449 }, { "epoch": 0.56608573972777, "ewc_loss": 0.0791015625, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 6.961822509765625e-05, "grad_norm": 5.170376777648926, "learning_rate": 1e-05, "loss": 0.5714, "mean_token_accuracy": 0.8379320502281189, "num_tokens": 169100317.0, "step": 4450 }, { "epoch": 0.5662129500063605, "ewc_loss": 0.10107421875, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.352554798126221, "learning_rate": 1e-05, "loss": 0.5071, "mean_token_accuracy": 0.8628539443016052, "num_tokens": 169128474.0, "step": 4451 }, { "epoch": 0.566340160284951, "ewc_loss": 0.08056640625, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 7.152557373046875e-05, "grad_norm": 5.150126934051514, "learning_rate": 1e-05, "loss": 0.4682, "mean_token_accuracy": 0.8731808066368103, "num_tokens": 169164905.0, "step": 4452 }, { "epoch": 0.5664673705635416, "ewc_loss": 0.10595703125, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.6140899658203125, "learning_rate": 1e-05, "loss": 0.5442, "mean_token_accuracy": 0.8514981269836426, "num_tokens": 169201153.0, "step": 4453 }, { "epoch": 0.566594580842132, "ewc_loss": 0.08642578125, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.430480480194092, "learning_rate": 1e-05, "loss": 0.473, "mean_token_accuracy": 0.8700814247131348, "num_tokens": 169235325.0, "step": 4454 }, { "epoch": 0.5667217911207225, "ewc_loss": 0.10546875, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.567070960998535, "learning_rate": 1e-05, "loss": 0.4791, "mean_token_accuracy": 0.8738335371017456, "num_tokens": 169276690.0, "step": 4455 }, { "epoch": 0.566849001399313, "ewc_loss": 0.09228515625, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.75505256652832, "learning_rate": 1e-05, "loss": 0.5568, "mean_token_accuracy": 0.8418742418289185, "num_tokens": 169315751.0, "step": 4456 }, { "epoch": 0.5669762116779036, "ewc_loss": 0.0947265625, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 6.158596992492676, "learning_rate": 1e-05, "loss": 0.5151, "mean_token_accuracy": 0.8578490614891052, "num_tokens": 169354626.0, "step": 4457 }, { "epoch": 0.5671034219564941, "ewc_loss": 0.091796875, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.848371982574463, "learning_rate": 1e-05, "loss": 0.522, "mean_token_accuracy": 0.8550689816474915, "num_tokens": 169390255.0, "step": 4458 }, { "epoch": 0.5672306322350846, "ewc_loss": 0.083984375, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 7.486343383789062e-05, "grad_norm": 5.643138408660889, "learning_rate": 1e-05, "loss": 0.5354, "mean_token_accuracy": 0.8469513654708862, "num_tokens": 169431912.0, "step": 4459 }, { "epoch": 0.5673578425136752, "ewc_loss": 0.09375, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.915695667266846, "learning_rate": 1e-05, "loss": 0.4377, "mean_token_accuracy": 0.8798581957817078, "num_tokens": 169467803.0, "step": 4460 }, { "epoch": 0.5674850527922656, "ewc_loss": 0.0791015625, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 6.961822509765625e-05, "grad_norm": 5.1930928230285645, "learning_rate": 1e-05, "loss": 0.5264, "mean_token_accuracy": 0.84675133228302, "num_tokens": 169512252.0, "step": 4461 }, { "epoch": 0.5676122630708561, "ewc_loss": 0.09619140625, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 6.133508682250977, "learning_rate": 1e-05, "loss": 0.5038, "mean_token_accuracy": 0.8609175682067871, "num_tokens": 169547632.0, "step": 4462 }, { "epoch": 0.5677394733494466, "ewc_loss": 0.07763671875, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 6.818771362304688e-05, "grad_norm": 4.944047927856445, "learning_rate": 1e-05, "loss": 0.5242, "mean_token_accuracy": 0.8550953269004822, "num_tokens": 169589355.0, "step": 4463 }, { "epoch": 0.5678666836280372, "ewc_loss": 0.10400390625, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.52800989151001, "learning_rate": 1e-05, "loss": 0.506, "mean_token_accuracy": 0.8631464838981628, "num_tokens": 169627441.0, "step": 4464 }, { "epoch": 0.5679938939066277, "ewc_loss": 0.080078125, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 7.05718994140625e-05, "grad_norm": 5.043004512786865, "learning_rate": 1e-05, "loss": 0.5102, "mean_token_accuracy": 0.8573817014694214, "num_tokens": 169659752.0, "step": 4465 }, { "epoch": 0.5681211041852182, "ewc_loss": 0.11328125, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.885810375213623, "learning_rate": 1e-05, "loss": 0.4939, "mean_token_accuracy": 0.8732154369354248, "num_tokens": 169697197.0, "step": 4466 }, { "epoch": 0.5682483144638086, "ewc_loss": 0.09130859375, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.554897308349609, "learning_rate": 1e-05, "loss": 0.5824, "mean_token_accuracy": 0.8369570970535278, "num_tokens": 169732523.0, "step": 4467 }, { "epoch": 0.5683755247423992, "ewc_loss": 0.109375, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.736283779144287, "learning_rate": 1e-05, "loss": 0.5535, "mean_token_accuracy": 0.8537290692329407, "num_tokens": 169771494.0, "step": 4468 }, { "epoch": 0.5685027350209897, "ewc_loss": 0.099609375, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 5.893343925476074, "learning_rate": 1e-05, "loss": 0.4753, "mean_token_accuracy": 0.8729969263076782, "num_tokens": 169809835.0, "step": 4469 }, { "epoch": 0.5686299452995802, "ewc_loss": 0.10009765625, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.308645248413086, "learning_rate": 1e-05, "loss": 0.5503, "mean_token_accuracy": 0.8498423099517822, "num_tokens": 169853276.0, "step": 4470 }, { "epoch": 0.5687571555781707, "ewc_loss": 0.09375, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.860171318054199, "learning_rate": 1e-05, "loss": 0.5606, "mean_token_accuracy": 0.83990478515625, "num_tokens": 169892106.0, "step": 4471 }, { "epoch": 0.5688843658567613, "ewc_loss": 0.087890625, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.946613311767578, "learning_rate": 1e-05, "loss": 0.5404, "mean_token_accuracy": 0.8474442362785339, "num_tokens": 169933635.0, "step": 4472 }, { "epoch": 0.5690115761353517, "ewc_loss": 0.08935546875, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.712094783782959, "learning_rate": 1e-05, "loss": 0.5118, "mean_token_accuracy": 0.8544409871101379, "num_tokens": 169971694.0, "step": 4473 }, { "epoch": 0.5691387864139422, "ewc_loss": 0.0869140625, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.72588586807251, "learning_rate": 1e-05, "loss": 0.5661, "mean_token_accuracy": 0.8390984535217285, "num_tokens": 170013285.0, "step": 4474 }, { "epoch": 0.5692659966925327, "ewc_loss": 0.08447265625, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 7.534027099609375e-05, "grad_norm": 5.542387008666992, "learning_rate": 1e-05, "loss": 0.5106, "mean_token_accuracy": 0.8567276000976562, "num_tokens": 170052831.0, "step": 4475 }, { "epoch": 0.5693932069711233, "ewc_loss": 0.08642578125, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.65493106842041, "learning_rate": 1e-05, "loss": 0.5616, "mean_token_accuracy": 0.8425549268722534, "num_tokens": 170089859.0, "step": 4476 }, { "epoch": 0.5695204172497138, "ewc_loss": 0.0810546875, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 7.152557373046875e-05, "grad_norm": 5.374317646026611, "learning_rate": 1e-05, "loss": 0.5252, "mean_token_accuracy": 0.8514300584793091, "num_tokens": 170126494.0, "step": 4477 }, { "epoch": 0.5696476275283043, "ewc_loss": 0.08642578125, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.6924262046813965, "learning_rate": 1e-05, "loss": 0.5412, "mean_token_accuracy": 0.8477756977081299, "num_tokens": 170167744.0, "step": 4478 }, { "epoch": 0.5697748378068948, "ewc_loss": 0.080078125, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 7.009506225585938e-05, "grad_norm": 5.197659015655518, "learning_rate": 1e-05, "loss": 0.5147, "mean_token_accuracy": 0.8550317287445068, "num_tokens": 170203988.0, "step": 4479 }, { "epoch": 0.5699020480854853, "ewc_loss": 0.08984375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.839695930480957, "learning_rate": 1e-05, "loss": 0.5899, "mean_token_accuracy": 0.8330286741256714, "num_tokens": 170243231.0, "step": 4480 }, { "epoch": 0.5700292583640758, "ewc_loss": 0.078125, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 6.866455078125e-05, "grad_norm": 5.0401082038879395, "learning_rate": 1e-05, "loss": 0.5704, "mean_token_accuracy": 0.8367242813110352, "num_tokens": 170280637.0, "step": 4481 }, { "epoch": 0.5701564686426663, "ewc_loss": 0.095703125, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 6.1217803955078125, "learning_rate": 1e-05, "loss": 0.5621, "mean_token_accuracy": 0.8471456170082092, "num_tokens": 170317106.0, "step": 4482 }, { "epoch": 0.5702836789212569, "ewc_loss": 0.0771484375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 6.771087646484375e-05, "grad_norm": 4.939529895782471, "learning_rate": 1e-05, "loss": 0.5158, "mean_token_accuracy": 0.8513664603233337, "num_tokens": 170353374.0, "step": 4483 }, { "epoch": 0.5704108891998474, "ewc_loss": 0.103515625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.512068271636963, "learning_rate": 1e-05, "loss": 0.5073, "mean_token_accuracy": 0.8611056208610535, "num_tokens": 170386968.0, "step": 4484 }, { "epoch": 0.5705380994784378, "ewc_loss": 0.0810546875, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 7.152557373046875e-05, "grad_norm": 5.0958781242370605, "learning_rate": 1e-05, "loss": 0.5133, "mean_token_accuracy": 0.8615739345550537, "num_tokens": 170425386.0, "step": 4485 }, { "epoch": 0.5706653097570283, "ewc_loss": 0.111328125, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.775285720825195, "learning_rate": 1e-05, "loss": 0.5233, "mean_token_accuracy": 0.8584024906158447, "num_tokens": 170467828.0, "step": 4486 }, { "epoch": 0.5707925200356189, "ewc_loss": 0.091796875, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.565439224243164, "learning_rate": 1e-05, "loss": 0.5301, "mean_token_accuracy": 0.8538340926170349, "num_tokens": 170506333.0, "step": 4487 }, { "epoch": 0.5709197303142094, "ewc_loss": 0.1044921875, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.565800666809082, "learning_rate": 1e-05, "loss": 0.5675, "mean_token_accuracy": 0.8508383631706238, "num_tokens": 170542352.0, "step": 4488 }, { "epoch": 0.5710469405927999, "ewc_loss": 0.095703125, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.847044944763184, "learning_rate": 1e-05, "loss": 0.4701, "mean_token_accuracy": 0.8701722621917725, "num_tokens": 170574251.0, "step": 4489 }, { "epoch": 0.5711741508713905, "ewc_loss": 0.09375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 6.079483985900879, "learning_rate": 1e-05, "loss": 0.4811, "mean_token_accuracy": 0.8664531707763672, "num_tokens": 170610294.0, "step": 4490 }, { "epoch": 0.5713013611499809, "ewc_loss": 0.09375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.854075908660889, "learning_rate": 1e-05, "loss": 0.5094, "mean_token_accuracy": 0.860657811164856, "num_tokens": 170649245.0, "step": 4491 }, { "epoch": 0.5714285714285714, "ewc_loss": 0.083984375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 7.486343383789062e-05, "grad_norm": 5.60585355758667, "learning_rate": 1e-05, "loss": 0.5179, "mean_token_accuracy": 0.8514814376831055, "num_tokens": 170692106.0, "step": 4492 }, { "epoch": 0.5715557817071619, "ewc_loss": 0.0927734375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.887106418609619, "learning_rate": 1e-05, "loss": 0.4981, "mean_token_accuracy": 0.8625147342681885, "num_tokens": 170730319.0, "step": 4493 }, { "epoch": 0.5716829919857525, "ewc_loss": 0.0810546875, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 7.104873657226562e-05, "grad_norm": 5.270896911621094, "learning_rate": 1e-05, "loss": 0.4871, "mean_token_accuracy": 0.8629082441329956, "num_tokens": 170763805.0, "step": 4494 }, { "epoch": 0.571810202264343, "ewc_loss": 0.09375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 6.02478551864624, "learning_rate": 1e-05, "loss": 0.5786, "mean_token_accuracy": 0.8394588232040405, "num_tokens": 170804966.0, "step": 4495 }, { "epoch": 0.5719374125429335, "ewc_loss": 0.078125, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 6.866455078125e-05, "grad_norm": 5.085330963134766, "learning_rate": 1e-05, "loss": 0.4982, "mean_token_accuracy": 0.8603997230529785, "num_tokens": 170845652.0, "step": 4496 }, { "epoch": 0.5720646228215239, "ewc_loss": 0.099609375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.249342918395996, "learning_rate": 1e-05, "loss": 0.5042, "mean_token_accuracy": 0.8602814674377441, "num_tokens": 170883449.0, "step": 4497 }, { "epoch": 0.5721918331001145, "ewc_loss": 0.080078125, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 7.009506225585938e-05, "grad_norm": 5.075216293334961, "learning_rate": 1e-05, "loss": 0.5412, "mean_token_accuracy": 0.8476330041885376, "num_tokens": 170920187.0, "step": 4498 }, { "epoch": 0.572319043378705, "ewc_loss": 0.103515625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.490286827087402, "learning_rate": 1e-05, "loss": 0.5581, "mean_token_accuracy": 0.8501982092857361, "num_tokens": 170952650.0, "step": 4499 }, { "epoch": 0.5724462536572955, "ewc_loss": 0.0849609375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 7.581710815429688e-05, "grad_norm": 5.249757766723633, "learning_rate": 1e-05, "loss": 0.5325, "mean_token_accuracy": 0.8505862951278687, "num_tokens": 170989130.0, "step": 4500 }, { "epoch": 0.572573463935886, "ewc_loss": 0.10498046875, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.548709392547607, "learning_rate": 1e-05, "loss": 0.5232, "mean_token_accuracy": 0.8586701154708862, "num_tokens": 171030713.0, "step": 4501 }, { "epoch": 0.5727006742144766, "ewc_loss": 0.09033203125, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.494768142700195, "learning_rate": 1e-05, "loss": 0.5176, "mean_token_accuracy": 0.8562456369400024, "num_tokens": 171066363.0, "step": 4502 }, { "epoch": 0.572827884493067, "ewc_loss": 0.10205078125, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.378697395324707, "learning_rate": 1e-05, "loss": 0.5676, "mean_token_accuracy": 0.8496559262275696, "num_tokens": 171108290.0, "step": 4503 }, { "epoch": 0.5729550947716575, "ewc_loss": 0.09130859375, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.609545707702637, "learning_rate": 1e-05, "loss": 0.4408, "mean_token_accuracy": 0.8810612559318542, "num_tokens": 171141990.0, "step": 4504 }, { "epoch": 0.573082305050248, "ewc_loss": 0.09521484375, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 6.108129501342773, "learning_rate": 1e-05, "loss": 0.5252, "mean_token_accuracy": 0.8573155999183655, "num_tokens": 171180032.0, "step": 4505 }, { "epoch": 0.5732095153288386, "ewc_loss": 0.08935546875, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.605264186859131, "learning_rate": 1e-05, "loss": 0.5805, "mean_token_accuracy": 0.8379350900650024, "num_tokens": 171218918.0, "step": 4506 }, { "epoch": 0.5733367256074291, "ewc_loss": 0.087890625, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.847867488861084, "learning_rate": 1e-05, "loss": 0.5373, "mean_token_accuracy": 0.8550945520401001, "num_tokens": 171261028.0, "step": 4507 }, { "epoch": 0.5734639358860196, "ewc_loss": 0.0869140625, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.528868675231934, "learning_rate": 1e-05, "loss": 0.5075, "mean_token_accuracy": 0.8577263951301575, "num_tokens": 171301477.0, "step": 4508 }, { "epoch": 0.5735911461646102, "ewc_loss": 0.08642578125, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.687978267669678, "learning_rate": 1e-05, "loss": 0.4757, "mean_token_accuracy": 0.8682849407196045, "num_tokens": 171338937.0, "step": 4509 }, { "epoch": 0.5737183564432006, "ewc_loss": 0.08447265625, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 7.486343383789062e-05, "grad_norm": 5.4300665855407715, "learning_rate": 1e-05, "loss": 0.5209, "mean_token_accuracy": 0.853287398815155, "num_tokens": 171375058.0, "step": 4510 }, { "epoch": 0.5738455667217911, "ewc_loss": 0.0869140625, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.645776748657227, "learning_rate": 1e-05, "loss": 0.5457, "mean_token_accuracy": 0.8473255038261414, "num_tokens": 171410282.0, "step": 4511 }, { "epoch": 0.5739727770003816, "ewc_loss": 0.08154296875, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 7.200241088867188e-05, "grad_norm": 5.307433605194092, "learning_rate": 1e-05, "loss": 0.5205, "mean_token_accuracy": 0.8531085252761841, "num_tokens": 171442864.0, "step": 4512 }, { "epoch": 0.5740999872789722, "ewc_loss": 0.087890625, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.689274787902832, "learning_rate": 1e-05, "loss": 0.4952, "mean_token_accuracy": 0.8673804998397827, "num_tokens": 171476267.0, "step": 4513 }, { "epoch": 0.5742271975575627, "ewc_loss": 0.080078125, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 7.009506225585938e-05, "grad_norm": 5.165970802307129, "learning_rate": 1e-05, "loss": 0.5173, "mean_token_accuracy": 0.8537136912345886, "num_tokens": 171513173.0, "step": 4514 }, { "epoch": 0.5743544078361532, "ewc_loss": 0.09033203125, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.8317155838012695, "learning_rate": 1e-05, "loss": 0.6061, "mean_token_accuracy": 0.8331851363182068, "num_tokens": 171555071.0, "step": 4515 }, { "epoch": 0.5744816181147436, "ewc_loss": 0.07861328125, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 6.866455078125e-05, "grad_norm": 5.036962985992432, "learning_rate": 1e-05, "loss": 0.5141, "mean_token_accuracy": 0.8542444705963135, "num_tokens": 171594263.0, "step": 4516 }, { "epoch": 0.5746088283933342, "ewc_loss": 0.0947265625, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 6.064979553222656, "learning_rate": 1e-05, "loss": 0.5542, "mean_token_accuracy": 0.847819447517395, "num_tokens": 171635238.0, "step": 4517 }, { "epoch": 0.5747360386719247, "ewc_loss": 0.078125, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 6.818771362304688e-05, "grad_norm": 4.964694023132324, "learning_rate": 1e-05, "loss": 0.5294, "mean_token_accuracy": 0.8508949875831604, "num_tokens": 171674537.0, "step": 4518 }, { "epoch": 0.5748632489505152, "ewc_loss": 0.10205078125, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.385197162628174, "learning_rate": 1e-05, "loss": 0.4953, "mean_token_accuracy": 0.8679346442222595, "num_tokens": 171718606.0, "step": 4519 }, { "epoch": 0.5749904592291057, "ewc_loss": 0.08154296875, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 7.200241088867188e-05, "grad_norm": 5.073702812194824, "learning_rate": 1e-05, "loss": 0.5069, "mean_token_accuracy": 0.8601256608963013, "num_tokens": 171757092.0, "step": 4520 }, { "epoch": 0.5751176695076963, "ewc_loss": 0.1083984375, "ewc_loss_diag": 9.632110595703125e-05, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.603484153747559, "learning_rate": 1e-05, "loss": 0.5354, "mean_token_accuracy": 0.8583415746688843, "num_tokens": 171799606.0, "step": 4521 }, { "epoch": 0.5752448797862867, "ewc_loss": 0.09033203125, "ewc_loss_diag": 9.632110595703125e-05, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.419764518737793, "learning_rate": 1e-05, "loss": 0.4956, "mean_token_accuracy": 0.862431526184082, "num_tokens": 171834686.0, "step": 4522 }, { "epoch": 0.5753720900648772, "ewc_loss": 0.10400390625, "ewc_loss_diag": 9.632110595703125e-05, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.476423263549805, "learning_rate": 1e-05, "loss": 0.6038, "mean_token_accuracy": 0.8372021317481995, "num_tokens": 171870443.0, "step": 4523 }, { "epoch": 0.5754993003434677, "ewc_loss": 0.09326171875, "ewc_loss_diag": 9.679794311523438e-05, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.677657604217529, "learning_rate": 1e-05, "loss": 0.5083, "mean_token_accuracy": 0.8603129982948303, "num_tokens": 171907615.0, "step": 4524 }, { "epoch": 0.5756265106220583, "ewc_loss": 0.09521484375, "ewc_loss_diag": 9.679794311523438e-05, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 6.0722455978393555, "learning_rate": 1e-05, "loss": 0.5599, "mean_token_accuracy": 0.8431766033172607, "num_tokens": 171943120.0, "step": 4525 }, { "epoch": 0.5757537209006488, "ewc_loss": 0.09375, "ewc_loss_diag": 9.679794311523438e-05, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.761314392089844, "learning_rate": 1e-05, "loss": 0.5077, "mean_token_accuracy": 0.8621557950973511, "num_tokens": 171980637.0, "step": 4526 }, { "epoch": 0.5758809311792393, "ewc_loss": 0.0849609375, "ewc_loss_diag": 9.679794311523438e-05, "ewc_loss_parallel": 7.534027099609375e-05, "grad_norm": 5.617933750152588, "learning_rate": 1e-05, "loss": 0.5169, "mean_token_accuracy": 0.8569453954696655, "num_tokens": 172014232.0, "step": 4527 }, { "epoch": 0.5760081414578297, "ewc_loss": 0.09326171875, "ewc_loss_diag": 9.679794311523438e-05, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.836838722229004, "learning_rate": 1e-05, "loss": 0.5188, "mean_token_accuracy": 0.8546634912490845, "num_tokens": 172049639.0, "step": 4528 }, { "epoch": 0.5761353517364203, "ewc_loss": 0.08056640625, "ewc_loss_diag": 9.679794311523438e-05, "ewc_loss_parallel": 7.05718994140625e-05, "grad_norm": 5.242033958435059, "learning_rate": 1e-05, "loss": 0.5132, "mean_token_accuracy": 0.856510579586029, "num_tokens": 172086482.0, "step": 4529 }, { "epoch": 0.5762625620150108, "ewc_loss": 0.09619140625, "ewc_loss_diag": 9.679794311523438e-05, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 6.016008377075195, "learning_rate": 1e-05, "loss": 0.5037, "mean_token_accuracy": 0.8623642325401306, "num_tokens": 172117220.0, "step": 4530 }, { "epoch": 0.5763897722936013, "ewc_loss": 0.078125, "ewc_loss_diag": 9.679794311523438e-05, "ewc_loss_parallel": 6.818771362304688e-05, "grad_norm": 5.037600994110107, "learning_rate": 1e-05, "loss": 0.4916, "mean_token_accuracy": 0.8594374656677246, "num_tokens": 172149925.0, "step": 4531 }, { "epoch": 0.5765169825721919, "ewc_loss": 0.099609375, "ewc_loss_diag": 9.679794311523438e-05, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.293458938598633, "learning_rate": 1e-05, "loss": 0.4861, "mean_token_accuracy": 0.8666667938232422, "num_tokens": 172186903.0, "step": 4532 }, { "epoch": 0.5766441928507824, "ewc_loss": 0.080078125, "ewc_loss_diag": 9.679794311523438e-05, "ewc_loss_parallel": 7.009506225585938e-05, "grad_norm": 5.029691696166992, "learning_rate": 1e-05, "loss": 0.519, "mean_token_accuracy": 0.8517534136772156, "num_tokens": 172226573.0, "step": 4533 }, { "epoch": 0.5767714031293728, "ewc_loss": 0.10595703125, "ewc_loss_diag": 9.679794311523438e-05, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.57626485824585, "learning_rate": 1e-05, "loss": 0.5301, "mean_token_accuracy": 0.8609974384307861, "num_tokens": 172263639.0, "step": 4534 }, { "epoch": 0.5768986134079633, "ewc_loss": 0.0869140625, "ewc_loss_diag": 9.679794311523438e-05, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.283529758453369, "learning_rate": 1e-05, "loss": 0.4794, "mean_token_accuracy": 0.8640347123146057, "num_tokens": 172297740.0, "step": 4535 }, { "epoch": 0.5770258236865539, "ewc_loss": 0.107421875, "ewc_loss_diag": 9.679794311523438e-05, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.600157737731934, "learning_rate": 1e-05, "loss": 0.5656, "mean_token_accuracy": 0.847417414188385, "num_tokens": 172338243.0, "step": 4536 }, { "epoch": 0.5771530339651444, "ewc_loss": 0.09130859375, "ewc_loss_diag": 9.679794311523438e-05, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.564798831939697, "learning_rate": 1e-05, "loss": 0.5213, "mean_token_accuracy": 0.8568418622016907, "num_tokens": 172383201.0, "step": 4537 }, { "epoch": 0.5772802442437349, "ewc_loss": 0.10205078125, "ewc_loss_diag": 9.679794311523438e-05, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.369632244110107, "learning_rate": 1e-05, "loss": 0.5083, "mean_token_accuracy": 0.8672235608100891, "num_tokens": 172420587.0, "step": 4538 }, { "epoch": 0.5774074545223254, "ewc_loss": 0.0927734375, "ewc_loss_diag": 9.679794311523438e-05, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.670321941375732, "learning_rate": 1e-05, "loss": 0.5693, "mean_token_accuracy": 0.8404700756072998, "num_tokens": 172459194.0, "step": 4539 }, { "epoch": 0.5775346648009159, "ewc_loss": 0.09423828125, "ewc_loss_diag": 9.679794311523438e-05, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 6.064286708831787, "learning_rate": 1e-05, "loss": 0.5696, "mean_token_accuracy": 0.8376634120941162, "num_tokens": 172492463.0, "step": 4540 }, { "epoch": 0.5776618750795064, "ewc_loss": 0.0908203125, "ewc_loss_diag": 9.679794311523438e-05, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.628499507904053, "learning_rate": 1e-05, "loss": 0.5593, "mean_token_accuracy": 0.8450192213058472, "num_tokens": 172531675.0, "step": 4541 }, { "epoch": 0.5777890853580969, "ewc_loss": 0.0888671875, "ewc_loss_diag": 9.679794311523438e-05, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.808052062988281, "learning_rate": 1e-05, "loss": 0.5118, "mean_token_accuracy": 0.8563761115074158, "num_tokens": 172571666.0, "step": 4542 }, { "epoch": 0.5779162956366874, "ewc_loss": 0.087890625, "ewc_loss_diag": 9.679794311523438e-05, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.537313461303711, "learning_rate": 1e-05, "loss": 0.5131, "mean_token_accuracy": 0.8570356369018555, "num_tokens": 172608132.0, "step": 4543 }, { "epoch": 0.578043505915278, "ewc_loss": 0.08740234375, "ewc_loss_diag": 9.679794311523438e-05, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.655787467956543, "learning_rate": 1e-05, "loss": 0.5069, "mean_token_accuracy": 0.8589169383049011, "num_tokens": 172650881.0, "step": 4544 }, { "epoch": 0.5781707161938685, "ewc_loss": 0.08447265625, "ewc_loss_diag": 9.679794311523438e-05, "ewc_loss_parallel": 7.486343383789062e-05, "grad_norm": 5.423798084259033, "learning_rate": 1e-05, "loss": 0.5036, "mean_token_accuracy": 0.8594763278961182, "num_tokens": 172684632.0, "step": 4545 }, { "epoch": 0.5782979264724589, "ewc_loss": 0.0869140625, "ewc_loss_diag": 9.679794311523438e-05, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.629705905914307, "learning_rate": 1e-05, "loss": 0.4836, "mean_token_accuracy": 0.8634852766990662, "num_tokens": 172724804.0, "step": 4546 }, { "epoch": 0.5784251367510495, "ewc_loss": 0.08154296875, "ewc_loss_diag": 9.679794311523438e-05, "ewc_loss_parallel": 7.200241088867188e-05, "grad_norm": 5.292316436767578, "learning_rate": 1e-05, "loss": 0.5103, "mean_token_accuracy": 0.8527032136917114, "num_tokens": 172762687.0, "step": 4547 }, { "epoch": 0.57855234702964, "ewc_loss": 0.08837890625, "ewc_loss_diag": 9.679794311523438e-05, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.681735992431641, "learning_rate": 1e-05, "loss": 0.5143, "mean_token_accuracy": 0.865218997001648, "num_tokens": 172802885.0, "step": 4548 }, { "epoch": 0.5786795573082305, "ewc_loss": 0.080078125, "ewc_loss_diag": 9.679794311523438e-05, "ewc_loss_parallel": 7.009506225585938e-05, "grad_norm": 5.172975063323975, "learning_rate": 1e-05, "loss": 0.4941, "mean_token_accuracy": 0.8591207265853882, "num_tokens": 172836158.0, "step": 4549 }, { "epoch": 0.578806767586821, "ewc_loss": 0.09033203125, "ewc_loss_diag": 9.679794311523438e-05, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.823451519012451, "learning_rate": 1e-05, "loss": 0.5254, "mean_token_accuracy": 0.8563909530639648, "num_tokens": 172872319.0, "step": 4550 }, { "epoch": 0.5789339778654116, "ewc_loss": 0.07861328125, "ewc_loss_diag": 9.679794311523438e-05, "ewc_loss_parallel": 6.866455078125e-05, "grad_norm": 5.0398736000061035, "learning_rate": 1e-05, "loss": 0.4912, "mean_token_accuracy": 0.861890435218811, "num_tokens": 172906803.0, "step": 4551 }, { "epoch": 0.579061188144002, "ewc_loss": 0.09521484375, "ewc_loss_diag": 9.679794311523438e-05, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 6.055962085723877, "learning_rate": 1e-05, "loss": 0.5188, "mean_token_accuracy": 0.8584720492362976, "num_tokens": 172944095.0, "step": 4552 }, { "epoch": 0.5791883984225925, "ewc_loss": 0.0791015625, "ewc_loss_diag": 9.679794311523438e-05, "ewc_loss_parallel": 6.914138793945312e-05, "grad_norm": 4.960284233093262, "learning_rate": 1e-05, "loss": 0.4758, "mean_token_accuracy": 0.8628569841384888, "num_tokens": 172980026.0, "step": 4553 }, { "epoch": 0.579315608701183, "ewc_loss": 0.1025390625, "ewc_loss_diag": 9.679794311523438e-05, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.379631996154785, "learning_rate": 1e-05, "loss": 0.548, "mean_token_accuracy": 0.8488978743553162, "num_tokens": 173019558.0, "step": 4554 }, { "epoch": 0.5794428189797736, "ewc_loss": 0.0810546875, "ewc_loss_diag": 9.679794311523438e-05, "ewc_loss_parallel": 7.104873657226562e-05, "grad_norm": 5.049040794372559, "learning_rate": 1e-05, "loss": 0.4871, "mean_token_accuracy": 0.8596977591514587, "num_tokens": 173059067.0, "step": 4555 }, { "epoch": 0.5795700292583641, "ewc_loss": 0.1083984375, "ewc_loss_diag": 9.679794311523438e-05, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.659271717071533, "learning_rate": 1e-05, "loss": 0.5098, "mean_token_accuracy": 0.8654477000236511, "num_tokens": 173096277.0, "step": 4556 }, { "epoch": 0.5796972395369546, "ewc_loss": 0.0908203125, "ewc_loss_diag": 9.679794311523438e-05, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.405840873718262, "learning_rate": 1e-05, "loss": 0.562, "mean_token_accuracy": 0.8436225652694702, "num_tokens": 173134746.0, "step": 4557 }, { "epoch": 0.5798244498155452, "ewc_loss": 0.10693359375, "ewc_loss_diag": 9.679794311523438e-05, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.572976589202881, "learning_rate": 1e-05, "loss": 0.605, "mean_token_accuracy": 0.8346782922744751, "num_tokens": 173164786.0, "step": 4558 }, { "epoch": 0.5799516600941356, "ewc_loss": 0.0947265625, "ewc_loss_diag": 9.679794311523438e-05, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.697377681732178, "learning_rate": 1e-05, "loss": 0.5502, "mean_token_accuracy": 0.84848552942276, "num_tokens": 173202948.0, "step": 4559 }, { "epoch": 0.5800788703727261, "ewc_loss": 0.0986328125, "ewc_loss_diag": 9.679794311523438e-05, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.173617839813232, "learning_rate": 1e-05, "loss": 0.5384, "mean_token_accuracy": 0.8502210974693298, "num_tokens": 173235253.0, "step": 4560 }, { "epoch": 0.5802060806513166, "ewc_loss": 0.09423828125, "ewc_loss_diag": 9.679794311523438e-05, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.771172523498535, "learning_rate": 1e-05, "loss": 0.5367, "mean_token_accuracy": 0.8516559600830078, "num_tokens": 173279878.0, "step": 4561 }, { "epoch": 0.5803332909299072, "ewc_loss": 0.087890625, "ewc_loss_diag": 9.679794311523438e-05, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.7166972160339355, "learning_rate": 1e-05, "loss": 0.5078, "mean_token_accuracy": 0.857829213142395, "num_tokens": 173316719.0, "step": 4562 }, { "epoch": 0.5804605012084977, "ewc_loss": 0.09375, "ewc_loss_diag": 9.679794311523438e-05, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.828227996826172, "learning_rate": 1e-05, "loss": 0.5047, "mean_token_accuracy": 0.8623899817466736, "num_tokens": 173347077.0, "step": 4563 }, { "epoch": 0.5805877114870882, "ewc_loss": 0.08154296875, "ewc_loss_diag": 9.679794311523438e-05, "ewc_loss_parallel": 7.200241088867188e-05, "grad_norm": 5.33650541305542, "learning_rate": 1e-05, "loss": 0.5467, "mean_token_accuracy": 0.8473684787750244, "num_tokens": 173386891.0, "step": 4564 }, { "epoch": 0.5807149217656786, "ewc_loss": 0.0947265625, "ewc_loss_diag": 9.679794311523438e-05, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.953648090362549, "learning_rate": 1e-05, "loss": 0.5216, "mean_token_accuracy": 0.8570168018341064, "num_tokens": 173426919.0, "step": 4565 }, { "epoch": 0.5808421320442692, "ewc_loss": 0.0791015625, "ewc_loss_diag": 9.679794311523438e-05, "ewc_loss_parallel": 6.914138793945312e-05, "grad_norm": 5.106168746948242, "learning_rate": 1e-05, "loss": 0.4919, "mean_token_accuracy": 0.8626258969306946, "num_tokens": 173461380.0, "step": 4566 }, { "epoch": 0.5809693423228597, "ewc_loss": 0.09912109375, "ewc_loss_diag": 9.679794311523438e-05, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 6.16771125793457, "learning_rate": 1e-05, "loss": 0.5381, "mean_token_accuracy": 0.8575537204742432, "num_tokens": 173504680.0, "step": 4567 }, { "epoch": 0.5810965526014502, "ewc_loss": 0.0791015625, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 6.914138793945312e-05, "grad_norm": 5.043025016784668, "learning_rate": 1e-05, "loss": 0.5102, "mean_token_accuracy": 0.8570225238800049, "num_tokens": 173541229.0, "step": 4568 }, { "epoch": 0.5812237628800407, "ewc_loss": 0.1025390625, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.421059608459473, "learning_rate": 1e-05, "loss": 0.5359, "mean_token_accuracy": 0.853602409362793, "num_tokens": 173573805.0, "step": 4569 }, { "epoch": 0.5813509731586313, "ewc_loss": 0.08349609375, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 7.390975952148438e-05, "grad_norm": 5.1751580238342285, "learning_rate": 1e-05, "loss": 0.5549, "mean_token_accuracy": 0.8451943397521973, "num_tokens": 173613899.0, "step": 4570 }, { "epoch": 0.5814781834372217, "ewc_loss": 0.10595703125, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.54612398147583, "learning_rate": 1e-05, "loss": 0.5227, "mean_token_accuracy": 0.8593865036964417, "num_tokens": 173650590.0, "step": 4571 }, { "epoch": 0.5816053937158122, "ewc_loss": 0.08984375, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.411798477172852, "learning_rate": 1e-05, "loss": 0.5172, "mean_token_accuracy": 0.8575426936149597, "num_tokens": 173684067.0, "step": 4572 }, { "epoch": 0.5817326039944027, "ewc_loss": 0.1044921875, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.4398627281188965, "learning_rate": 1e-05, "loss": 0.5042, "mean_token_accuracy": 0.8623489141464233, "num_tokens": 173721846.0, "step": 4573 }, { "epoch": 0.5818598142729933, "ewc_loss": 0.09130859375, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.577175140380859, "learning_rate": 1e-05, "loss": 0.5671, "mean_token_accuracy": 0.84014892578125, "num_tokens": 173754616.0, "step": 4574 }, { "epoch": 0.5819870245515838, "ewc_loss": 0.09912109375, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 6.1942877769470215, "learning_rate": 1e-05, "loss": 0.5299, "mean_token_accuracy": 0.8519635796546936, "num_tokens": 173790208.0, "step": 4575 }, { "epoch": 0.5821142348301743, "ewc_loss": 0.091796875, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.5864973068237305, "learning_rate": 1e-05, "loss": 0.565, "mean_token_accuracy": 0.8461284637451172, "num_tokens": 173833132.0, "step": 4576 }, { "epoch": 0.5822414451087647, "ewc_loss": 0.091796875, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.94069242477417, "learning_rate": 1e-05, "loss": 0.537, "mean_token_accuracy": 0.8480900526046753, "num_tokens": 173870526.0, "step": 4577 }, { "epoch": 0.5823686553873553, "ewc_loss": 0.0888671875, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.547852039337158, "learning_rate": 1e-05, "loss": 0.529, "mean_token_accuracy": 0.8512804508209229, "num_tokens": 173901561.0, "step": 4578 }, { "epoch": 0.5824958656659458, "ewc_loss": 0.08837890625, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.740416049957275, "learning_rate": 1e-05, "loss": 0.5572, "mean_token_accuracy": 0.8461126089096069, "num_tokens": 173941197.0, "step": 4579 }, { "epoch": 0.5826230759445363, "ewc_loss": 0.08642578125, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.45981502532959, "learning_rate": 1e-05, "loss": 0.5278, "mean_token_accuracy": 0.853239893913269, "num_tokens": 173981755.0, "step": 4580 }, { "epoch": 0.5827502862231269, "ewc_loss": 0.0869140625, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.657143592834473, "learning_rate": 1e-05, "loss": 0.519, "mean_token_accuracy": 0.8508917689323425, "num_tokens": 174015831.0, "step": 4581 }, { "epoch": 0.5828774965017174, "ewc_loss": 0.083984375, "ewc_loss_diag": 9.775161743164062e-05, "ewc_loss_parallel": 7.43865966796875e-05, "grad_norm": 5.352977275848389, "learning_rate": 1e-05, "loss": 0.4894, "mean_token_accuracy": 0.8614175319671631, "num_tokens": 174051623.0, "step": 4582 }, { "epoch": 0.5830047067803078, "ewc_loss": 0.087890625, "ewc_loss_diag": 9.775161743164062e-05, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.657187461853027, "learning_rate": 1e-05, "loss": 0.5611, "mean_token_accuracy": 0.8431515097618103, "num_tokens": 174086422.0, "step": 4583 }, { "epoch": 0.5831319170588983, "ewc_loss": 0.0810546875, "ewc_loss_diag": 9.775161743164062e-05, "ewc_loss_parallel": 7.152557373046875e-05, "grad_norm": 5.237037658691406, "learning_rate": 1e-05, "loss": 0.5327, "mean_token_accuracy": 0.849956750869751, "num_tokens": 174128557.0, "step": 4584 }, { "epoch": 0.5832591273374889, "ewc_loss": 0.08984375, "ewc_loss_diag": 9.775161743164062e-05, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.73766565322876, "learning_rate": 1e-05, "loss": 0.5502, "mean_token_accuracy": 0.8461631536483765, "num_tokens": 174169704.0, "step": 4585 }, { "epoch": 0.5833863376160794, "ewc_loss": 0.080078125, "ewc_loss_diag": 9.775161743164062e-05, "ewc_loss_parallel": 7.009506225585938e-05, "grad_norm": 5.112727165222168, "learning_rate": 1e-05, "loss": 0.548, "mean_token_accuracy": 0.8443470001220703, "num_tokens": 174213975.0, "step": 4586 }, { "epoch": 0.5835135478946699, "ewc_loss": 0.09228515625, "ewc_loss_diag": 9.775161743164062e-05, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.900995254516602, "learning_rate": 1e-05, "loss": 0.5337, "mean_token_accuracy": 0.853425145149231, "num_tokens": 174254425.0, "step": 4587 }, { "epoch": 0.5836407581732604, "ewc_loss": 0.07958984375, "ewc_loss_diag": 9.775161743164062e-05, "ewc_loss_parallel": 6.961822509765625e-05, "grad_norm": 5.017268657684326, "learning_rate": 1e-05, "loss": 0.5435, "mean_token_accuracy": 0.8475554585456848, "num_tokens": 174292411.0, "step": 4588 }, { "epoch": 0.5837679684518509, "ewc_loss": 0.09912109375, "ewc_loss_diag": 9.775161743164062e-05, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 6.1621785163879395, "learning_rate": 1e-05, "loss": 0.5906, "mean_token_accuracy": 0.8375859260559082, "num_tokens": 174335103.0, "step": 4589 }, { "epoch": 0.5838951787304414, "ewc_loss": 0.080078125, "ewc_loss_diag": 9.775161743164062e-05, "ewc_loss_parallel": 7.009506225585938e-05, "grad_norm": 4.987321853637695, "learning_rate": 1e-05, "loss": 0.5533, "mean_token_accuracy": 0.8458577394485474, "num_tokens": 174377984.0, "step": 4590 }, { "epoch": 0.5840223890090319, "ewc_loss": 0.10400390625, "ewc_loss_diag": 9.775161743164062e-05, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.459842681884766, "learning_rate": 1e-05, "loss": 0.5392, "mean_token_accuracy": 0.8531900644302368, "num_tokens": 174408211.0, "step": 4591 }, { "epoch": 0.5841495992876224, "ewc_loss": 0.0869140625, "ewc_loss_diag": 9.775161743164062e-05, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.178479194641113, "learning_rate": 1e-05, "loss": 0.5534, "mean_token_accuracy": 0.840511679649353, "num_tokens": 174444137.0, "step": 4592 }, { "epoch": 0.584276809566213, "ewc_loss": 0.107421875, "ewc_loss_diag": 9.775161743164062e-05, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.562459468841553, "learning_rate": 1e-05, "loss": 0.5603, "mean_token_accuracy": 0.8481414318084717, "num_tokens": 174488382.0, "step": 4593 }, { "epoch": 0.5844040198448035, "ewc_loss": 0.0908203125, "ewc_loss_diag": 9.775161743164062e-05, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.482247352600098, "learning_rate": 1e-05, "loss": 0.4846, "mean_token_accuracy": 0.8694726228713989, "num_tokens": 174524436.0, "step": 4594 }, { "epoch": 0.5845312301233939, "ewc_loss": 0.10302734375, "ewc_loss_diag": 9.775161743164062e-05, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.350032806396484, "learning_rate": 1e-05, "loss": 0.5859, "mean_token_accuracy": 0.8445127010345459, "num_tokens": 174563830.0, "step": 4595 }, { "epoch": 0.5846584404019844, "ewc_loss": 0.09375, "ewc_loss_diag": 9.775161743164062e-05, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.682693958282471, "learning_rate": 1e-05, "loss": 0.5271, "mean_token_accuracy": 0.8545407056808472, "num_tokens": 174598600.0, "step": 4596 }, { "epoch": 0.584785650680575, "ewc_loss": 0.09130859375, "ewc_loss_diag": 9.775161743164062e-05, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.93326473236084, "learning_rate": 1e-05, "loss": 0.4936, "mean_token_accuracy": 0.8636102676391602, "num_tokens": 174631234.0, "step": 4597 }, { "epoch": 0.5849128609591655, "ewc_loss": 0.09423828125, "ewc_loss_diag": 9.775161743164062e-05, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.7476372718811035, "learning_rate": 1e-05, "loss": 0.5275, "mean_token_accuracy": 0.8542209267616272, "num_tokens": 174673933.0, "step": 4598 }, { "epoch": 0.585040071237756, "ewc_loss": 0.08349609375, "ewc_loss_diag": 9.775161743164062e-05, "ewc_loss_parallel": 7.390975952148438e-05, "grad_norm": 5.4934492111206055, "learning_rate": 1e-05, "loss": 0.5045, "mean_token_accuracy": 0.8567734956741333, "num_tokens": 174715801.0, "step": 4599 }, { "epoch": 0.5851672815163466, "ewc_loss": 0.09375, "ewc_loss_diag": 9.775161743164062e-05, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.860124111175537, "learning_rate": 1e-05, "loss": 0.5233, "mean_token_accuracy": 0.8538737297058105, "num_tokens": 174757309.0, "step": 4600 }, { "epoch": 0.585294491794937, "ewc_loss": 0.080078125, "ewc_loss_diag": 9.775161743164062e-05, "ewc_loss_parallel": 7.009506225585938e-05, "grad_norm": 5.169183254241943, "learning_rate": 1e-05, "loss": 0.4589, "mean_token_accuracy": 0.8701987266540527, "num_tokens": 174787942.0, "step": 4601 }, { "epoch": 0.5854217020735275, "ewc_loss": 0.09716796875, "ewc_loss_diag": 9.775161743164062e-05, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 6.082616329193115, "learning_rate": 1e-05, "loss": 0.5249, "mean_token_accuracy": 0.8576952815055847, "num_tokens": 174827483.0, "step": 4602 }, { "epoch": 0.585548912352118, "ewc_loss": 0.07861328125, "ewc_loss_diag": 9.775161743164062e-05, "ewc_loss_parallel": 6.866455078125e-05, "grad_norm": 5.003076076507568, "learning_rate": 1e-05, "loss": 0.4613, "mean_token_accuracy": 0.8665832281112671, "num_tokens": 174862847.0, "step": 4603 }, { "epoch": 0.5856761226307086, "ewc_loss": 0.10205078125, "ewc_loss_diag": 9.775161743164062e-05, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.380398750305176, "learning_rate": 1e-05, "loss": 0.5542, "mean_token_accuracy": 0.8471558690071106, "num_tokens": 174903327.0, "step": 4604 }, { "epoch": 0.5858033329092991, "ewc_loss": 0.08154296875, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 7.200241088867188e-05, "grad_norm": 5.075319290161133, "learning_rate": 1e-05, "loss": 0.5566, "mean_token_accuracy": 0.840692400932312, "num_tokens": 174939278.0, "step": 4605 }, { "epoch": 0.5859305431878896, "ewc_loss": 0.107421875, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.628325939178467, "learning_rate": 1e-05, "loss": 0.538, "mean_token_accuracy": 0.8544257879257202, "num_tokens": 174972671.0, "step": 4606 }, { "epoch": 0.5860577534664801, "ewc_loss": 0.0888671875, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.361400127410889, "learning_rate": 1e-05, "loss": 0.5375, "mean_token_accuracy": 0.8466391563415527, "num_tokens": 175009753.0, "step": 4607 }, { "epoch": 0.5861849637450706, "ewc_loss": 0.10791015625, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.590413570404053, "learning_rate": 1e-05, "loss": 0.4882, "mean_token_accuracy": 0.8713609576225281, "num_tokens": 175044271.0, "step": 4608 }, { "epoch": 0.5863121740236611, "ewc_loss": 0.0927734375, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.6084184646606445, "learning_rate": 1e-05, "loss": 0.5527, "mean_token_accuracy": 0.8503222465515137, "num_tokens": 175086132.0, "step": 4609 }, { "epoch": 0.5864393843022516, "ewc_loss": 0.10205078125, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.331576347351074, "learning_rate": 1e-05, "loss": 0.4781, "mean_token_accuracy": 0.8691767454147339, "num_tokens": 175114902.0, "step": 4610 }, { "epoch": 0.5865665945808421, "ewc_loss": 0.09326171875, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.681886196136475, "learning_rate": 1e-05, "loss": 0.5337, "mean_token_accuracy": 0.8484144806861877, "num_tokens": 175152274.0, "step": 4611 }, { "epoch": 0.5866938048594327, "ewc_loss": 0.09423828125, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 6.0026021003723145, "learning_rate": 1e-05, "loss": 0.4979, "mean_token_accuracy": 0.861505389213562, "num_tokens": 175198449.0, "step": 4612 }, { "epoch": 0.5868210151380232, "ewc_loss": 0.0908203125, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.629105091094971, "learning_rate": 1e-05, "loss": 0.476, "mean_token_accuracy": 0.8692749738693237, "num_tokens": 175236851.0, "step": 4613 }, { "epoch": 0.5869482254166136, "ewc_loss": 0.0888671875, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.768894672393799, "learning_rate": 1e-05, "loss": 0.5274, "mean_token_accuracy": 0.8579776287078857, "num_tokens": 175275694.0, "step": 4614 }, { "epoch": 0.5870754356952042, "ewc_loss": 0.087890625, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.533569812774658, "learning_rate": 1e-05, "loss": 0.5704, "mean_token_accuracy": 0.8409081101417542, "num_tokens": 175321484.0, "step": 4615 }, { "epoch": 0.5872026459737947, "ewc_loss": 0.08642578125, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.646757125854492, "learning_rate": 1e-05, "loss": 0.5071, "mean_token_accuracy": 0.8589522838592529, "num_tokens": 175358387.0, "step": 4616 }, { "epoch": 0.5873298562523852, "ewc_loss": 0.08447265625, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 7.486343383789062e-05, "grad_norm": 5.4131550788879395, "learning_rate": 1e-05, "loss": 0.53, "mean_token_accuracy": 0.8476519584655762, "num_tokens": 175402363.0, "step": 4617 }, { "epoch": 0.5874570665309757, "ewc_loss": 0.08642578125, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.633950710296631, "learning_rate": 1e-05, "loss": 0.5324, "mean_token_accuracy": 0.8526521921157837, "num_tokens": 175440655.0, "step": 4618 }, { "epoch": 0.5875842768095663, "ewc_loss": 0.08154296875, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 7.200241088867188e-05, "grad_norm": 5.29611349105835, "learning_rate": 1e-05, "loss": 0.5232, "mean_token_accuracy": 0.8552814722061157, "num_tokens": 175476041.0, "step": 4619 }, { "epoch": 0.5877114870881567, "ewc_loss": 0.08837890625, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.683516025543213, "learning_rate": 1e-05, "loss": 0.4619, "mean_token_accuracy": 0.8746105432510376, "num_tokens": 175514673.0, "step": 4620 }, { "epoch": 0.5878386973667472, "ewc_loss": 0.08056640625, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 7.05718994140625e-05, "grad_norm": 5.1646904945373535, "learning_rate": 1e-05, "loss": 0.4774, "mean_token_accuracy": 0.8662326335906982, "num_tokens": 175555396.0, "step": 4621 }, { "epoch": 0.5879659076453377, "ewc_loss": 0.09033203125, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.8336100578308105, "learning_rate": 1e-05, "loss": 0.4676, "mean_token_accuracy": 0.8721575736999512, "num_tokens": 175597588.0, "step": 4622 }, { "epoch": 0.5880931179239283, "ewc_loss": 0.0791015625, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 6.914138793945312e-05, "grad_norm": 5.028912544250488, "learning_rate": 1e-05, "loss": 0.4951, "mean_token_accuracy": 0.8626196384429932, "num_tokens": 175636911.0, "step": 4623 }, { "epoch": 0.5882203282025188, "ewc_loss": 0.09619140625, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 6.101191520690918, "learning_rate": 1e-05, "loss": 0.5744, "mean_token_accuracy": 0.8409299254417419, "num_tokens": 175671778.0, "step": 4624 }, { "epoch": 0.5883475384811093, "ewc_loss": 0.07861328125, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 6.866455078125e-05, "grad_norm": 4.932590007781982, "learning_rate": 1e-05, "loss": 0.4481, "mean_token_accuracy": 0.8739427328109741, "num_tokens": 175711218.0, "step": 4625 }, { "epoch": 0.5884747487596997, "ewc_loss": 0.10400390625, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.481853008270264, "learning_rate": 1e-05, "loss": 0.5419, "mean_token_accuracy": 0.8571200370788574, "num_tokens": 175744630.0, "step": 4626 }, { "epoch": 0.5886019590382903, "ewc_loss": 0.08251953125, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 7.295608520507812e-05, "grad_norm": 5.062572002410889, "learning_rate": 1e-05, "loss": 0.5101, "mean_token_accuracy": 0.8569855690002441, "num_tokens": 175783342.0, "step": 4627 }, { "epoch": 0.5887291693168808, "ewc_loss": 0.111328125, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.771349906921387, "learning_rate": 1e-05, "loss": 0.563, "mean_token_accuracy": 0.8459683656692505, "num_tokens": 175820361.0, "step": 4628 }, { "epoch": 0.5888563795954713, "ewc_loss": 0.09130859375, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.474626541137695, "learning_rate": 1e-05, "loss": 0.5335, "mean_token_accuracy": 0.8516039848327637, "num_tokens": 175860388.0, "step": 4629 }, { "epoch": 0.5889835898740619, "ewc_loss": 0.1083984375, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.636982440948486, "learning_rate": 1e-05, "loss": 0.5594, "mean_token_accuracy": 0.85488361120224, "num_tokens": 175895725.0, "step": 4630 }, { "epoch": 0.5891108001526524, "ewc_loss": 0.09912109375, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 5.78460168838501, "learning_rate": 1e-05, "loss": 0.5234, "mean_token_accuracy": 0.8613648414611816, "num_tokens": 175933535.0, "step": 4631 }, { "epoch": 0.5892380104312428, "ewc_loss": 0.099609375, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.221774101257324, "learning_rate": 1e-05, "loss": 0.5582, "mean_token_accuracy": 0.8489674925804138, "num_tokens": 175968681.0, "step": 4632 }, { "epoch": 0.5893652207098333, "ewc_loss": 0.09619140625, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.827524662017822, "learning_rate": 1e-05, "loss": 0.5415, "mean_token_accuracy": 0.8502906560897827, "num_tokens": 176006088.0, "step": 4633 }, { "epoch": 0.5894924309884239, "ewc_loss": 0.0888671875, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.7581400871276855, "learning_rate": 1e-05, "loss": 0.46, "mean_token_accuracy": 0.871817946434021, "num_tokens": 176040855.0, "step": 4634 }, { "epoch": 0.5896196412670144, "ewc_loss": 0.09375, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.8366289138793945, "learning_rate": 1e-05, "loss": 0.5158, "mean_token_accuracy": 0.8565943241119385, "num_tokens": 176080015.0, "step": 4635 }, { "epoch": 0.5897468515456049, "ewc_loss": 0.08203125, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 7.2479248046875e-05, "grad_norm": 5.394421577453613, "learning_rate": 1e-05, "loss": 0.5353, "mean_token_accuracy": 0.8456693291664124, "num_tokens": 176120266.0, "step": 4636 }, { "epoch": 0.5898740618241954, "ewc_loss": 0.09521484375, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.932931900024414, "learning_rate": 1e-05, "loss": 0.5059, "mean_token_accuracy": 0.8619419932365417, "num_tokens": 176150436.0, "step": 4637 }, { "epoch": 0.5900012721027859, "ewc_loss": 0.0810546875, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 7.104873657226562e-05, "grad_norm": 5.179058074951172, "learning_rate": 1e-05, "loss": 0.5387, "mean_token_accuracy": 0.8471992015838623, "num_tokens": 176185340.0, "step": 4638 }, { "epoch": 0.5901284823813764, "ewc_loss": 0.09765625, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 6.08504056930542, "learning_rate": 1e-05, "loss": 0.5557, "mean_token_accuracy": 0.8502811193466187, "num_tokens": 176225372.0, "step": 4639 }, { "epoch": 0.5902556926599669, "ewc_loss": 0.080078125, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 7.009506225585938e-05, "grad_norm": 5.06874942779541, "learning_rate": 1e-05, "loss": 0.4982, "mean_token_accuracy": 0.8574919104576111, "num_tokens": 176267114.0, "step": 4640 }, { "epoch": 0.5903829029385574, "ewc_loss": 0.10107421875, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.322936058044434, "learning_rate": 1e-05, "loss": 0.4903, "mean_token_accuracy": 0.8717986941337585, "num_tokens": 176302937.0, "step": 4641 }, { "epoch": 0.590510113217148, "ewc_loss": 0.08203125, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 7.2479248046875e-05, "grad_norm": 5.121869087219238, "learning_rate": 1e-05, "loss": 0.4686, "mean_token_accuracy": 0.8679232001304626, "num_tokens": 176340307.0, "step": 4642 }, { "epoch": 0.5906373234957385, "ewc_loss": 0.10498046875, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.506760597229004, "learning_rate": 1e-05, "loss": 0.5647, "mean_token_accuracy": 0.8487260341644287, "num_tokens": 176385420.0, "step": 4643 }, { "epoch": 0.5907645337743289, "ewc_loss": 0.08740234375, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.312151908874512, "learning_rate": 1e-05, "loss": 0.5195, "mean_token_accuracy": 0.8533535003662109, "num_tokens": 176424975.0, "step": 4644 }, { "epoch": 0.5908917440529194, "ewc_loss": 0.10546875, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.512659072875977, "learning_rate": 1e-05, "loss": 0.5885, "mean_token_accuracy": 0.8396924734115601, "num_tokens": 176465471.0, "step": 4645 }, { "epoch": 0.59101895433151, "ewc_loss": 0.09130859375, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.527195453643799, "learning_rate": 1e-05, "loss": 0.5029, "mean_token_accuracy": 0.8592469692230225, "num_tokens": 176493717.0, "step": 4646 }, { "epoch": 0.5911461646101005, "ewc_loss": 0.10107421875, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.310641288757324, "learning_rate": 1e-05, "loss": 0.532, "mean_token_accuracy": 0.8547111749649048, "num_tokens": 176533636.0, "step": 4647 }, { "epoch": 0.591273374888691, "ewc_loss": 0.091796875, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.591071605682373, "learning_rate": 1e-05, "loss": 0.5319, "mean_token_accuracy": 0.8543448448181152, "num_tokens": 176573248.0, "step": 4648 }, { "epoch": 0.5914005851672816, "ewc_loss": 0.095703125, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 6.032480716705322, "learning_rate": 1e-05, "loss": 0.4776, "mean_token_accuracy": 0.8686317205429077, "num_tokens": 176614607.0, "step": 4649 }, { "epoch": 0.591527795445872, "ewc_loss": 0.08935546875, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.575590133666992, "learning_rate": 1e-05, "loss": 0.5377, "mean_token_accuracy": 0.8508103489875793, "num_tokens": 176657482.0, "step": 4650 }, { "epoch": 0.5916550057244625, "ewc_loss": 0.08837890625, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.793850898742676, "learning_rate": 1e-05, "loss": 0.4978, "mean_token_accuracy": 0.8656034469604492, "num_tokens": 176694448.0, "step": 4651 }, { "epoch": 0.591782216003053, "ewc_loss": 0.08837890625, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.534492492675781, "learning_rate": 1e-05, "loss": 0.5479, "mean_token_accuracy": 0.8523309230804443, "num_tokens": 176726899.0, "step": 4652 }, { "epoch": 0.5919094262816436, "ewc_loss": 0.08642578125, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.629541873931885, "learning_rate": 1e-05, "loss": 0.5139, "mean_token_accuracy": 0.8515874147415161, "num_tokens": 176767495.0, "step": 4653 }, { "epoch": 0.5920366365602341, "ewc_loss": 0.08544921875, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 7.581710815429688e-05, "grad_norm": 5.458554744720459, "learning_rate": 1e-05, "loss": 0.523, "mean_token_accuracy": 0.8536155223846436, "num_tokens": 176802804.0, "step": 4654 }, { "epoch": 0.5921638468388246, "ewc_loss": 0.0849609375, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 7.534027099609375e-05, "grad_norm": 5.551316738128662, "learning_rate": 1e-05, "loss": 0.4809, "mean_token_accuracy": 0.8665295839309692, "num_tokens": 176847224.0, "step": 4655 }, { "epoch": 0.592291057117415, "ewc_loss": 0.0830078125, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 7.343292236328125e-05, "grad_norm": 5.380203723907471, "learning_rate": 1e-05, "loss": 0.527, "mean_token_accuracy": 0.854461669921875, "num_tokens": 176884423.0, "step": 4656 }, { "epoch": 0.5924182673960056, "ewc_loss": 0.0859375, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 7.62939453125e-05, "grad_norm": 5.553310871124268, "learning_rate": 1e-05, "loss": 0.5025, "mean_token_accuracy": 0.8651804327964783, "num_tokens": 176917776.0, "step": 4657 }, { "epoch": 0.5925454776745961, "ewc_loss": 0.0810546875, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 7.152557373046875e-05, "grad_norm": 5.282966136932373, "learning_rate": 1e-05, "loss": 0.477, "mean_token_accuracy": 0.8653365969657898, "num_tokens": 176951387.0, "step": 4658 }, { "epoch": 0.5926726879531866, "ewc_loss": 0.0869140625, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.611584663391113, "learning_rate": 1e-05, "loss": 0.5364, "mean_token_accuracy": 0.8504406213760376, "num_tokens": 176986533.0, "step": 4659 }, { "epoch": 0.5927998982317771, "ewc_loss": 0.080078125, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 7.009506225585938e-05, "grad_norm": 5.168643474578857, "learning_rate": 1e-05, "loss": 0.4746, "mean_token_accuracy": 0.8676316142082214, "num_tokens": 177021541.0, "step": 4660 }, { "epoch": 0.5929271085103677, "ewc_loss": 0.08984375, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.7399797439575195, "learning_rate": 1e-05, "loss": 0.5106, "mean_token_accuracy": 0.8590986728668213, "num_tokens": 177062933.0, "step": 4661 }, { "epoch": 0.5930543187889582, "ewc_loss": 0.0791015625, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 6.914138793945312e-05, "grad_norm": 5.036908149719238, "learning_rate": 1e-05, "loss": 0.5631, "mean_token_accuracy": 0.8378193974494934, "num_tokens": 177102549.0, "step": 4662 }, { "epoch": 0.5931815290675486, "ewc_loss": 0.09326171875, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.9853644371032715, "learning_rate": 1e-05, "loss": 0.5155, "mean_token_accuracy": 0.8557279109954834, "num_tokens": 177139618.0, "step": 4663 }, { "epoch": 0.5933087393461391, "ewc_loss": 0.07861328125, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 6.866455078125e-05, "grad_norm": 4.905707359313965, "learning_rate": 1e-05, "loss": 0.5172, "mean_token_accuracy": 0.8488821983337402, "num_tokens": 177171912.0, "step": 4664 }, { "epoch": 0.5934359496247297, "ewc_loss": 0.1025390625, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.381330966949463, "learning_rate": 1e-05, "loss": 0.535, "mean_token_accuracy": 0.8573404550552368, "num_tokens": 177214279.0, "step": 4665 }, { "epoch": 0.5935631599033202, "ewc_loss": 0.07958984375, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 6.961822509765625e-05, "grad_norm": 4.925057411193848, "learning_rate": 1e-05, "loss": 0.5327, "mean_token_accuracy": 0.8530868291854858, "num_tokens": 177247277.0, "step": 4666 }, { "epoch": 0.5936903701819107, "ewc_loss": 0.11376953125, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.816540718078613, "learning_rate": 1e-05, "loss": 0.5369, "mean_token_accuracy": 0.8563472032546997, "num_tokens": 177287097.0, "step": 4667 }, { "epoch": 0.5938175804605013, "ewc_loss": 0.09033203125, "ewc_loss_diag": 9.870529174804688e-05, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.307627201080322, "learning_rate": 1e-05, "loss": 0.4811, "mean_token_accuracy": 0.8691043257713318, "num_tokens": 177324400.0, "step": 4668 }, { "epoch": 0.5939447907390917, "ewc_loss": 0.1171875, "ewc_loss_diag": 9.870529174804688e-05, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.904013156890869, "learning_rate": 1e-05, "loss": 0.5334, "mean_token_accuracy": 0.863417387008667, "num_tokens": 177362962.0, "step": 4669 }, { "epoch": 0.5940720010176822, "ewc_loss": 0.1015625, "ewc_loss_diag": 9.870529174804688e-05, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 5.804070472717285, "learning_rate": 1e-05, "loss": 0.5188, "mean_token_accuracy": 0.8566071391105652, "num_tokens": 177402708.0, "step": 4670 }, { "epoch": 0.5941992112962727, "ewc_loss": 0.10595703125, "ewc_loss_diag": 9.870529174804688e-05, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.539192199707031, "learning_rate": 1e-05, "loss": 0.4958, "mean_token_accuracy": 0.8663233518600464, "num_tokens": 177433957.0, "step": 4671 }, { "epoch": 0.5943264215748633, "ewc_loss": 0.1015625, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 5.904212474822998, "learning_rate": 1e-05, "loss": 0.5034, "mean_token_accuracy": 0.8615024089813232, "num_tokens": 177465265.0, "step": 4672 }, { "epoch": 0.5944536318534538, "ewc_loss": 0.09521484375, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 6.115264415740967, "learning_rate": 1e-05, "loss": 0.5421, "mean_token_accuracy": 0.8477672338485718, "num_tokens": 177495448.0, "step": 4673 }, { "epoch": 0.5945808421320443, "ewc_loss": 0.09619140625, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.811180591583252, "learning_rate": 1e-05, "loss": 0.5059, "mean_token_accuracy": 0.8616673946380615, "num_tokens": 177531582.0, "step": 4674 }, { "epoch": 0.5947080524106347, "ewc_loss": 0.087890625, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.773183345794678, "learning_rate": 1e-05, "loss": 0.5143, "mean_token_accuracy": 0.8554784059524536, "num_tokens": 177565039.0, "step": 4675 }, { "epoch": 0.5948352626892253, "ewc_loss": 0.091796875, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.722914218902588, "learning_rate": 1e-05, "loss": 0.5232, "mean_token_accuracy": 0.855743944644928, "num_tokens": 177601872.0, "step": 4676 }, { "epoch": 0.5949624729678158, "ewc_loss": 0.0849609375, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 7.534027099609375e-05, "grad_norm": 5.534949779510498, "learning_rate": 1e-05, "loss": 0.5551, "mean_token_accuracy": 0.8437303304672241, "num_tokens": 177645693.0, "step": 4677 }, { "epoch": 0.5950896832464063, "ewc_loss": 0.08984375, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.669898509979248, "learning_rate": 1e-05, "loss": 0.4974, "mean_token_accuracy": 0.8615453243255615, "num_tokens": 177685334.0, "step": 4678 }, { "epoch": 0.5952168935249968, "ewc_loss": 0.0830078125, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 7.343292236328125e-05, "grad_norm": 5.391965866088867, "learning_rate": 1e-05, "loss": 0.6438, "mean_token_accuracy": 0.8175368309020996, "num_tokens": 177727851.0, "step": 4679 }, { "epoch": 0.5953441038035874, "ewc_loss": 0.08837890625, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.644079208374023, "learning_rate": 1e-05, "loss": 0.5295, "mean_token_accuracy": 0.8527246713638306, "num_tokens": 177769115.0, "step": 4680 }, { "epoch": 0.5954713140821778, "ewc_loss": 0.08154296875, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 7.200241088867188e-05, "grad_norm": 5.287625789642334, "learning_rate": 1e-05, "loss": 0.5216, "mean_token_accuracy": 0.8506481647491455, "num_tokens": 177809094.0, "step": 4681 }, { "epoch": 0.5955985243607683, "ewc_loss": 0.08935546875, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.661904335021973, "learning_rate": 1e-05, "loss": 0.5139, "mean_token_accuracy": 0.8567367792129517, "num_tokens": 177842409.0, "step": 4682 }, { "epoch": 0.5957257346393589, "ewc_loss": 0.0810546875, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 7.104873657226562e-05, "grad_norm": 5.19305419921875, "learning_rate": 1e-05, "loss": 0.5322, "mean_token_accuracy": 0.8514861464500427, "num_tokens": 177882525.0, "step": 4683 }, { "epoch": 0.5958529449179494, "ewc_loss": 0.08984375, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.753955364227295, "learning_rate": 1e-05, "loss": 0.5732, "mean_token_accuracy": 0.8405715227127075, "num_tokens": 177924293.0, "step": 4684 }, { "epoch": 0.5959801551965399, "ewc_loss": 0.080078125, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 7.009506225585938e-05, "grad_norm": 5.070826530456543, "learning_rate": 1e-05, "loss": 0.5002, "mean_token_accuracy": 0.8597722053527832, "num_tokens": 177964716.0, "step": 4685 }, { "epoch": 0.5961073654751304, "ewc_loss": 0.09326171875, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.942017555236816, "learning_rate": 1e-05, "loss": 0.4547, "mean_token_accuracy": 0.8777792453765869, "num_tokens": 177999029.0, "step": 4686 }, { "epoch": 0.5962345757537209, "ewc_loss": 0.07861328125, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 6.866455078125e-05, "grad_norm": 4.946645259857178, "learning_rate": 1e-05, "loss": 0.4694, "mean_token_accuracy": 0.8694912195205688, "num_tokens": 178033872.0, "step": 4687 }, { "epoch": 0.5963617860323114, "ewc_loss": 0.1005859375, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.258111953735352, "learning_rate": 1e-05, "loss": 0.5108, "mean_token_accuracy": 0.8608442544937134, "num_tokens": 178077835.0, "step": 4688 }, { "epoch": 0.5964889963109019, "ewc_loss": 0.07958984375, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 6.961822509765625e-05, "grad_norm": 4.887967586517334, "learning_rate": 1e-05, "loss": 0.475, "mean_token_accuracy": 0.8638492822647095, "num_tokens": 178119205.0, "step": 4689 }, { "epoch": 0.5966162065894924, "ewc_loss": 0.109375, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.701076984405518, "learning_rate": 1e-05, "loss": 0.572, "mean_token_accuracy": 0.8463367223739624, "num_tokens": 178157923.0, "step": 4690 }, { "epoch": 0.596743416868083, "ewc_loss": 0.0869140625, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.1506266593933105, "learning_rate": 1e-05, "loss": 0.456, "mean_token_accuracy": 0.8729884028434753, "num_tokens": 178200465.0, "step": 4691 }, { "epoch": 0.5968706271466735, "ewc_loss": 0.1171875, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.950262069702148, "learning_rate": 1e-05, "loss": 0.6285, "mean_token_accuracy": 0.828980565071106, "num_tokens": 178234767.0, "step": 4692 }, { "epoch": 0.5969978374252639, "ewc_loss": 0.10009765625, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 5.678199768066406, "learning_rate": 1e-05, "loss": 0.558, "mean_token_accuracy": 0.8477946519851685, "num_tokens": 178269569.0, "step": 4693 }, { "epoch": 0.5971250477038544, "ewc_loss": 0.1103515625, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.7140936851501465, "learning_rate": 1e-05, "loss": 0.5471, "mean_token_accuracy": 0.8559402823448181, "num_tokens": 178300789.0, "step": 4694 }, { "epoch": 0.597252257982445, "ewc_loss": 0.10400390625, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 5.908816814422607, "learning_rate": 1e-05, "loss": 0.5007, "mean_token_accuracy": 0.865617036819458, "num_tokens": 178333341.0, "step": 4695 }, { "epoch": 0.5973794682610355, "ewc_loss": 0.1015625, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.286996364593506, "learning_rate": 1e-05, "loss": 0.5742, "mean_token_accuracy": 0.8433762788772583, "num_tokens": 178368743.0, "step": 4696 }, { "epoch": 0.597506678539626, "ewc_loss": 0.09765625, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 5.844522953033447, "learning_rate": 1e-05, "loss": 0.4467, "mean_token_accuracy": 0.8765590190887451, "num_tokens": 178407432.0, "step": 4697 }, { "epoch": 0.5976338888182166, "ewc_loss": 0.09033203125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.930103302001953, "learning_rate": 1e-05, "loss": 0.5358, "mean_token_accuracy": 0.8530123829841614, "num_tokens": 178440245.0, "step": 4698 }, { "epoch": 0.597761099096807, "ewc_loss": 0.09375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.738230228424072, "learning_rate": 1e-05, "loss": 0.5018, "mean_token_accuracy": 0.862085223197937, "num_tokens": 178484446.0, "step": 4699 }, { "epoch": 0.5978883093753975, "ewc_loss": 0.0869140625, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.650866508483887, "learning_rate": 1e-05, "loss": 0.5155, "mean_token_accuracy": 0.8594210743904114, "num_tokens": 178522390.0, "step": 4700 }, { "epoch": 0.598015519653988, "ewc_loss": 0.08984375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.659983158111572, "learning_rate": 1e-05, "loss": 0.4982, "mean_token_accuracy": 0.8629319667816162, "num_tokens": 178561671.0, "step": 4701 }, { "epoch": 0.5981427299325786, "ewc_loss": 0.0849609375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 7.534027099609375e-05, "grad_norm": 5.481741428375244, "learning_rate": 1e-05, "loss": 0.5761, "mean_token_accuracy": 0.8390191793441772, "num_tokens": 178605289.0, "step": 4702 }, { "epoch": 0.5982699402111691, "ewc_loss": 0.0859375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 7.62939453125e-05, "grad_norm": 5.601407051086426, "learning_rate": 1e-05, "loss": 0.499, "mean_token_accuracy": 0.8651920557022095, "num_tokens": 178647947.0, "step": 4703 }, { "epoch": 0.5983971504897596, "ewc_loss": 0.08251953125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 7.295608520507812e-05, "grad_norm": 5.378363609313965, "learning_rate": 1e-05, "loss": 0.4681, "mean_token_accuracy": 0.869601845741272, "num_tokens": 178686779.0, "step": 4704 }, { "epoch": 0.59852436076835, "ewc_loss": 0.0859375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 7.62939453125e-05, "grad_norm": 5.573339462280273, "learning_rate": 1e-05, "loss": 0.5478, "mean_token_accuracy": 0.8472229242324829, "num_tokens": 178730517.0, "step": 4705 }, { "epoch": 0.5986515710469406, "ewc_loss": 0.0810546875, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 7.152557373046875e-05, "grad_norm": 5.306890964508057, "learning_rate": 1e-05, "loss": 0.5052, "mean_token_accuracy": 0.8556450009346008, "num_tokens": 178762521.0, "step": 4706 }, { "epoch": 0.5987787813255311, "ewc_loss": 0.08642578125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.595791816711426, "learning_rate": 1e-05, "loss": 0.5443, "mean_token_accuracy": 0.8485575914382935, "num_tokens": 178804816.0, "step": 4707 }, { "epoch": 0.5989059916041216, "ewc_loss": 0.0810546875, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 7.104873657226562e-05, "grad_norm": 5.209000587463379, "learning_rate": 1e-05, "loss": 0.5676, "mean_token_accuracy": 0.8455635905265808, "num_tokens": 178845413.0, "step": 4708 }, { "epoch": 0.5990332018827121, "ewc_loss": 0.0869140625, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.672671794891357, "learning_rate": 1e-05, "loss": 0.5232, "mean_token_accuracy": 0.8548678755760193, "num_tokens": 178886887.0, "step": 4709 }, { "epoch": 0.5991604121613027, "ewc_loss": 0.07958984375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 6.961822509765625e-05, "grad_norm": 5.080537796020508, "learning_rate": 1e-05, "loss": 0.4821, "mean_token_accuracy": 0.8609327077865601, "num_tokens": 178922067.0, "step": 4710 }, { "epoch": 0.5992876224398932, "ewc_loss": 0.0908203125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.87485408782959, "learning_rate": 1e-05, "loss": 0.5871, "mean_token_accuracy": 0.8357182145118713, "num_tokens": 178955645.0, "step": 4711 }, { "epoch": 0.5994148327184836, "ewc_loss": 0.07861328125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 6.866455078125e-05, "grad_norm": 4.923775672912598, "learning_rate": 1e-05, "loss": 0.4927, "mean_token_accuracy": 0.8590013384819031, "num_tokens": 178992389.0, "step": 4712 }, { "epoch": 0.5995420429970741, "ewc_loss": 0.09912109375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 6.229818344116211, "learning_rate": 1e-05, "loss": 0.5522, "mean_token_accuracy": 0.8453827500343323, "num_tokens": 179031140.0, "step": 4713 }, { "epoch": 0.5996692532756647, "ewc_loss": 0.07861328125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 6.866455078125e-05, "grad_norm": 4.813162803649902, "learning_rate": 1e-05, "loss": 0.4526, "mean_token_accuracy": 0.873206377029419, "num_tokens": 179065486.0, "step": 4714 }, { "epoch": 0.5997964635542552, "ewc_loss": 0.11181640625, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.778258800506592, "learning_rate": 1e-05, "loss": 0.5517, "mean_token_accuracy": 0.8541811108589172, "num_tokens": 179104957.0, "step": 4715 }, { "epoch": 0.5999236738328457, "ewc_loss": 0.0859375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 7.62939453125e-05, "grad_norm": 5.071217060089111, "learning_rate": 1e-05, "loss": 0.526, "mean_token_accuracy": 0.8565458059310913, "num_tokens": 179146022.0, "step": 4716 }, { "epoch": 0.6000508841114363, "ewc_loss": 0.1240234375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 0.00011444091796875, "grad_norm": 7.19670295715332, "learning_rate": 1e-05, "loss": 0.5288, "mean_token_accuracy": 0.8652428388595581, "num_tokens": 179184691.0, "step": 4717 }, { "epoch": 0.6001780943900267, "ewc_loss": 0.10546875, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 5.8118672370910645, "learning_rate": 1e-05, "loss": 0.5131, "mean_token_accuracy": 0.8589223623275757, "num_tokens": 179223734.0, "step": 4718 }, { "epoch": 0.6003053046686172, "ewc_loss": 0.1162109375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.931046962738037, "learning_rate": 1e-05, "loss": 0.5214, "mean_token_accuracy": 0.8649910688400269, "num_tokens": 179265997.0, "step": 4719 }, { "epoch": 0.6004325149472077, "ewc_loss": 0.10888671875, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.120765686035156, "learning_rate": 1e-05, "loss": 0.5513, "mean_token_accuracy": 0.8504859209060669, "num_tokens": 179302734.0, "step": 4720 }, { "epoch": 0.6005597252257983, "ewc_loss": 0.1025390625, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.463994026184082, "learning_rate": 1e-05, "loss": 0.5534, "mean_token_accuracy": 0.8512399792671204, "num_tokens": 179343779.0, "step": 4721 }, { "epoch": 0.6006869355043888, "ewc_loss": 0.1015625, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 5.968243598937988, "learning_rate": 1e-05, "loss": 0.5176, "mean_token_accuracy": 0.8637573719024658, "num_tokens": 179384713.0, "step": 4722 }, { "epoch": 0.6008141457829793, "ewc_loss": 0.095703125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 6.1232590675354, "learning_rate": 1e-05, "loss": 0.4989, "mean_token_accuracy": 0.8633589148521423, "num_tokens": 179418005.0, "step": 4723 }, { "epoch": 0.6009413560615697, "ewc_loss": 0.09375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.773809909820557, "learning_rate": 1e-05, "loss": 0.5413, "mean_token_accuracy": 0.8499650955200195, "num_tokens": 179459437.0, "step": 4724 }, { "epoch": 0.6010685663401603, "ewc_loss": 0.08935546875, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.900699138641357, "learning_rate": 1e-05, "loss": 0.5142, "mean_token_accuracy": 0.8599497675895691, "num_tokens": 179490534.0, "step": 4725 }, { "epoch": 0.6011957766187508, "ewc_loss": 0.0888671875, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.601174354553223, "learning_rate": 1e-05, "loss": 0.5279, "mean_token_accuracy": 0.8557735681533813, "num_tokens": 179533141.0, "step": 4726 }, { "epoch": 0.6013229868973413, "ewc_loss": 0.0888671875, "ewc_loss_diag": 0.00010013580322265625, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.789834499359131, "learning_rate": 1e-05, "loss": 0.5492, "mean_token_accuracy": 0.8468095064163208, "num_tokens": 179565438.0, "step": 4727 }, { "epoch": 0.6014501971759318, "ewc_loss": 0.08447265625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 7.43865966796875e-05, "grad_norm": 5.430693626403809, "learning_rate": 1e-05, "loss": 0.5689, "mean_token_accuracy": 0.8386662602424622, "num_tokens": 179605184.0, "step": 4728 }, { "epoch": 0.6015774074545224, "ewc_loss": 0.08984375, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.775407791137695, "learning_rate": 1e-05, "loss": 0.5139, "mean_token_accuracy": 0.8565659523010254, "num_tokens": 179643355.0, "step": 4729 }, { "epoch": 0.6017046177331128, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 7.152557373046875e-05, "grad_norm": 5.274177074432373, "learning_rate": 1e-05, "loss": 0.5011, "mean_token_accuracy": 0.86284339427948, "num_tokens": 179682471.0, "step": 4730 }, { "epoch": 0.6018318280117033, "ewc_loss": 0.0908203125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.8658270835876465, "learning_rate": 1e-05, "loss": 0.5528, "mean_token_accuracy": 0.8472841382026672, "num_tokens": 179722290.0, "step": 4731 }, { "epoch": 0.6019590382902938, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 6.961822509765625e-05, "grad_norm": 5.110215663909912, "learning_rate": 1e-05, "loss": 0.55, "mean_token_accuracy": 0.8422167301177979, "num_tokens": 179762409.0, "step": 4732 }, { "epoch": 0.6020862485688844, "ewc_loss": 0.09716796875, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 6.104569435119629, "learning_rate": 1e-05, "loss": 0.553, "mean_token_accuracy": 0.8461934924125671, "num_tokens": 179802561.0, "step": 4733 }, { "epoch": 0.6022134588474749, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 6.961822509765625e-05, "grad_norm": 4.989530563354492, "learning_rate": 1e-05, "loss": 0.507, "mean_token_accuracy": 0.8579665422439575, "num_tokens": 179842923.0, "step": 4734 }, { "epoch": 0.6023406691260654, "ewc_loss": 0.10498046875, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.4856343269348145, "learning_rate": 1e-05, "loss": 0.5475, "mean_token_accuracy": 0.8528592586517334, "num_tokens": 179882430.0, "step": 4735 }, { "epoch": 0.6024678794046558, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 7.295608520507812e-05, "grad_norm": 5.100757122039795, "learning_rate": 1e-05, "loss": 0.5365, "mean_token_accuracy": 0.8491793274879456, "num_tokens": 179923974.0, "step": 4736 }, { "epoch": 0.6025950896832464, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.783722877502441, "learning_rate": 1e-05, "loss": 0.5115, "mean_token_accuracy": 0.8651537895202637, "num_tokens": 179966468.0, "step": 4737 }, { "epoch": 0.6027222999618369, "ewc_loss": 0.09228515625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.523350238800049, "learning_rate": 1e-05, "loss": 0.4715, "mean_token_accuracy": 0.870959460735321, "num_tokens": 180005356.0, "step": 4738 }, { "epoch": 0.6028495102404274, "ewc_loss": 0.10986328125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.666440486907959, "learning_rate": 1e-05, "loss": 0.5402, "mean_token_accuracy": 0.8554060459136963, "num_tokens": 180044765.0, "step": 4739 }, { "epoch": 0.602976720519018, "ewc_loss": 0.099609375, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 5.874368667602539, "learning_rate": 1e-05, "loss": 0.5572, "mean_token_accuracy": 0.848799467086792, "num_tokens": 180082311.0, "step": 4740 }, { "epoch": 0.6031039307976085, "ewc_loss": 0.09912109375, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.1678924560546875, "learning_rate": 1e-05, "loss": 0.5045, "mean_token_accuracy": 0.861275851726532, "num_tokens": 180122889.0, "step": 4741 }, { "epoch": 0.6032311410761989, "ewc_loss": 0.09521484375, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.922884941101074, "learning_rate": 1e-05, "loss": 0.5008, "mean_token_accuracy": 0.8629410266876221, "num_tokens": 180157425.0, "step": 4742 }, { "epoch": 0.6033583513547894, "ewc_loss": 0.0849609375, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 7.486343383789062e-05, "grad_norm": 5.666537761688232, "learning_rate": 1e-05, "loss": 0.4951, "mean_token_accuracy": 0.8637324571609497, "num_tokens": 180196099.0, "step": 4743 }, { "epoch": 0.60348556163338, "ewc_loss": 0.095703125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.966212272644043, "learning_rate": 1e-05, "loss": 0.5788, "mean_token_accuracy": 0.8397301435470581, "num_tokens": 180238036.0, "step": 4744 }, { "epoch": 0.6036127719119705, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 7.009506225585938e-05, "grad_norm": 5.279849052429199, "learning_rate": 1e-05, "loss": 0.4952, "mean_token_accuracy": 0.859548032283783, "num_tokens": 180276193.0, "step": 4745 }, { "epoch": 0.603739982190561, "ewc_loss": 0.0986328125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 6.160515308380127, "learning_rate": 1e-05, "loss": 0.5401, "mean_token_accuracy": 0.8545416593551636, "num_tokens": 180313470.0, "step": 4746 }, { "epoch": 0.6038671924691515, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 6.914138793945312e-05, "grad_norm": 5.080101013183594, "learning_rate": 1e-05, "loss": 0.504, "mean_token_accuracy": 0.8600000739097595, "num_tokens": 180354588.0, "step": 4747 }, { "epoch": 0.603994402747742, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.441503524780273, "learning_rate": 1e-05, "loss": 0.5324, "mean_token_accuracy": 0.8565573692321777, "num_tokens": 180395792.0, "step": 4748 }, { "epoch": 0.6041216130263325, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 7.2479248046875e-05, "grad_norm": 5.1479620933532715, "learning_rate": 1e-05, "loss": 0.5358, "mean_token_accuracy": 0.8490831851959229, "num_tokens": 180436823.0, "step": 4749 }, { "epoch": 0.604248823304923, "ewc_loss": 0.10888671875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.675599098205566, "learning_rate": 1e-05, "loss": 0.5047, "mean_token_accuracy": 0.865506112575531, "num_tokens": 180472643.0, "step": 4750 }, { "epoch": 0.6043760335835135, "ewc_loss": 0.0908203125, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.455620765686035, "learning_rate": 1e-05, "loss": 0.5408, "mean_token_accuracy": 0.847836434841156, "num_tokens": 180511788.0, "step": 4751 }, { "epoch": 0.6045032438621041, "ewc_loss": 0.10791015625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.602223873138428, "learning_rate": 1e-05, "loss": 0.5025, "mean_token_accuracy": 0.8704460859298706, "num_tokens": 180547871.0, "step": 4752 }, { "epoch": 0.6046304541406946, "ewc_loss": 0.09326171875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.690516471862793, "learning_rate": 1e-05, "loss": 0.5507, "mean_token_accuracy": 0.8450133800506592, "num_tokens": 180580521.0, "step": 4753 }, { "epoch": 0.604757664419285, "ewc_loss": 0.10205078125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.335989475250244, "learning_rate": 1e-05, "loss": 0.5532, "mean_token_accuracy": 0.8509356379508972, "num_tokens": 180617726.0, "step": 4754 }, { "epoch": 0.6048848746978756, "ewc_loss": 0.09228515625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.703326225280762, "learning_rate": 1e-05, "loss": 0.5223, "mean_token_accuracy": 0.8557236194610596, "num_tokens": 180657705.0, "step": 4755 }, { "epoch": 0.6050120849764661, "ewc_loss": 0.09326171875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 6.063018321990967, "learning_rate": 1e-05, "loss": 0.4989, "mean_token_accuracy": 0.8657035231590271, "num_tokens": 180693582.0, "step": 4756 }, { "epoch": 0.6051392952550566, "ewc_loss": 0.0888671875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.6107892990112305, "learning_rate": 1e-05, "loss": 0.5343, "mean_token_accuracy": 0.8492069244384766, "num_tokens": 180731159.0, "step": 4757 }, { "epoch": 0.6052665055336471, "ewc_loss": 0.08984375, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.877556324005127, "learning_rate": 1e-05, "loss": 0.5425, "mean_token_accuracy": 0.8462781310081482, "num_tokens": 180769901.0, "step": 4758 }, { "epoch": 0.6053937158122377, "ewc_loss": 0.0859375, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 7.581710815429688e-05, "grad_norm": 5.484817028045654, "learning_rate": 1e-05, "loss": 0.5092, "mean_token_accuracy": 0.8601698875427246, "num_tokens": 180811763.0, "step": 4759 }, { "epoch": 0.6055209260908282, "ewc_loss": 0.08935546875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.821012496948242, "learning_rate": 1e-05, "loss": 0.511, "mean_token_accuracy": 0.8600279092788696, "num_tokens": 180846073.0, "step": 4760 }, { "epoch": 0.6056481363694186, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 7.390975952148438e-05, "grad_norm": 5.339855194091797, "learning_rate": 1e-05, "loss": 0.4806, "mean_token_accuracy": 0.8659871816635132, "num_tokens": 180883609.0, "step": 4761 }, { "epoch": 0.6057753466480091, "ewc_loss": 0.0908203125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.879230499267578, "learning_rate": 1e-05, "loss": 0.5234, "mean_token_accuracy": 0.8548980951309204, "num_tokens": 180921492.0, "step": 4762 }, { "epoch": 0.6059025569265997, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 7.009506225585938e-05, "grad_norm": 5.212880611419678, "learning_rate": 1e-05, "loss": 0.5504, "mean_token_accuracy": 0.8425579071044922, "num_tokens": 180960298.0, "step": 4763 }, { "epoch": 0.6060297672051902, "ewc_loss": 0.09423828125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 6.01846981048584, "learning_rate": 1e-05, "loss": 0.542, "mean_token_accuracy": 0.8506273031234741, "num_tokens": 180997002.0, "step": 4764 }, { "epoch": 0.6061569774837807, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 7.05718994140625e-05, "grad_norm": 5.131284713745117, "learning_rate": 1e-05, "loss": 0.5693, "mean_token_accuracy": 0.8374639749526978, "num_tokens": 181033575.0, "step": 4765 }, { "epoch": 0.6062841877623713, "ewc_loss": 0.099609375, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 6.20564079284668, "learning_rate": 1e-05, "loss": 0.5355, "mean_token_accuracy": 0.8558576703071594, "num_tokens": 181068584.0, "step": 4766 }, { "epoch": 0.6064113980409617, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 7.152557373046875e-05, "grad_norm": 5.158872604370117, "learning_rate": 1e-05, "loss": 0.4976, "mean_token_accuracy": 0.8615673780441284, "num_tokens": 181103017.0, "step": 4767 }, { "epoch": 0.6065386083195522, "ewc_loss": 0.10205078125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.362064361572266, "learning_rate": 1e-05, "loss": 0.5201, "mean_token_accuracy": 0.8567293882369995, "num_tokens": 181137590.0, "step": 4768 }, { "epoch": 0.6066658185981427, "ewc_loss": 0.0869140625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.297140121459961, "learning_rate": 1e-05, "loss": 0.5117, "mean_token_accuracy": 0.8573389053344727, "num_tokens": 181177581.0, "step": 4769 }, { "epoch": 0.6067930288767333, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.335640907287598, "learning_rate": 1e-05, "loss": 0.5036, "mean_token_accuracy": 0.8671576976776123, "num_tokens": 181212582.0, "step": 4770 }, { "epoch": 0.6069202391553238, "ewc_loss": 0.0908203125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.522937297821045, "learning_rate": 1e-05, "loss": 0.5466, "mean_token_accuracy": 0.8475832939147949, "num_tokens": 181246373.0, "step": 4771 }, { "epoch": 0.6070474494339143, "ewc_loss": 0.095703125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 6.0626983642578125, "learning_rate": 1e-05, "loss": 0.4753, "mean_token_accuracy": 0.871609628200531, "num_tokens": 181280657.0, "step": 4772 }, { "epoch": 0.6071746597125047, "ewc_loss": 0.09130859375, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.641140937805176, "learning_rate": 1e-05, "loss": 0.4866, "mean_token_accuracy": 0.8657553195953369, "num_tokens": 181322165.0, "step": 4773 }, { "epoch": 0.6073018699910953, "ewc_loss": 0.08642578125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 7.62939453125e-05, "grad_norm": 5.666326522827148, "learning_rate": 1e-05, "loss": 0.5007, "mean_token_accuracy": 0.8556455373764038, "num_tokens": 181358612.0, "step": 4774 }, { "epoch": 0.6074290802696858, "ewc_loss": 0.0927734375, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.769130706787109, "learning_rate": 1e-05, "loss": 0.478, "mean_token_accuracy": 0.8700177669525146, "num_tokens": 181392928.0, "step": 4775 }, { "epoch": 0.6075562905482763, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 7.104873657226562e-05, "grad_norm": 5.256617546081543, "learning_rate": 1e-05, "loss": 0.5741, "mean_token_accuracy": 0.8396540284156799, "num_tokens": 181431127.0, "step": 4776 }, { "epoch": 0.6076835008268668, "ewc_loss": 0.09521484375, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.983907699584961, "learning_rate": 1e-05, "loss": 0.5193, "mean_token_accuracy": 0.8583449125289917, "num_tokens": 181469830.0, "step": 4777 }, { "epoch": 0.6078107111054574, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 6.818771362304688e-05, "grad_norm": 4.957566261291504, "learning_rate": 1e-05, "loss": 0.4979, "mean_token_accuracy": 0.8564397096633911, "num_tokens": 181505590.0, "step": 4778 }, { "epoch": 0.6079379213840478, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.373404026031494, "learning_rate": 1e-05, "loss": 0.4861, "mean_token_accuracy": 0.8723338842391968, "num_tokens": 181541675.0, "step": 4779 }, { "epoch": 0.6080651316626383, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 6.961822509765625e-05, "grad_norm": 4.912662506103516, "learning_rate": 1e-05, "loss": 0.5445, "mean_token_accuracy": 0.8443557024002075, "num_tokens": 181579890.0, "step": 4780 }, { "epoch": 0.6081923419412288, "ewc_loss": 0.11376953125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.834630012512207, "learning_rate": 1e-05, "loss": 0.5513, "mean_token_accuracy": 0.8554755449295044, "num_tokens": 181613821.0, "step": 4781 }, { "epoch": 0.6083195522198194, "ewc_loss": 0.0908203125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.318420886993408, "learning_rate": 1e-05, "loss": 0.5614, "mean_token_accuracy": 0.8438458442687988, "num_tokens": 181656899.0, "step": 4782 }, { "epoch": 0.6084467624984099, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.9257588386535645, "learning_rate": 1e-05, "loss": 0.5819, "mean_token_accuracy": 0.8402949571609497, "num_tokens": 181690044.0, "step": 4783 }, { "epoch": 0.6085739727770004, "ewc_loss": 0.10107421875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 5.822054862976074, "learning_rate": 1e-05, "loss": 0.5122, "mean_token_accuracy": 0.8606619834899902, "num_tokens": 181725038.0, "step": 4784 }, { "epoch": 0.6087011830555908, "ewc_loss": 0.10791015625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.554773807525635, "learning_rate": 1e-05, "loss": 0.534, "mean_token_accuracy": 0.8583406805992126, "num_tokens": 181767914.0, "step": 4785 }, { "epoch": 0.6088283933341814, "ewc_loss": 0.10205078125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 5.92196798324585, "learning_rate": 1e-05, "loss": 0.4891, "mean_token_accuracy": 0.8676267862319946, "num_tokens": 181798955.0, "step": 4786 }, { "epoch": 0.6089556036127719, "ewc_loss": 0.09716796875, "ewc_loss_diag": 0.000102996826171875, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 6.155881881713867, "learning_rate": 1e-05, "loss": 0.5685, "mean_token_accuracy": 0.8441984057426453, "num_tokens": 181831967.0, "step": 4787 }, { "epoch": 0.6090828138913624, "ewc_loss": 0.0947265625, "ewc_loss_diag": 0.000102996826171875, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.803808212280273, "learning_rate": 1e-05, "loss": 0.5237, "mean_token_accuracy": 0.8570109009742737, "num_tokens": 181865717.0, "step": 4788 }, { "epoch": 0.609210024169953, "ewc_loss": 0.08984375, "ewc_loss_diag": 0.000102996826171875, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.856324195861816, "learning_rate": 1e-05, "loss": 0.5726, "mean_token_accuracy": 0.8399934768676758, "num_tokens": 181902847.0, "step": 4789 }, { "epoch": 0.6093372344485435, "ewc_loss": 0.0908203125, "ewc_loss_diag": 0.000102996826171875, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.656246662139893, "learning_rate": 1e-05, "loss": 0.4744, "mean_token_accuracy": 0.8675695061683655, "num_tokens": 181937129.0, "step": 4790 }, { "epoch": 0.6094644447271339, "ewc_loss": 0.08837890625, "ewc_loss_diag": 0.000102996826171875, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.690196514129639, "learning_rate": 1e-05, "loss": 0.5164, "mean_token_accuracy": 0.8565666079521179, "num_tokens": 181979576.0, "step": 4791 }, { "epoch": 0.6095916550057244, "ewc_loss": 0.0859375, "ewc_loss_diag": 0.000102996826171875, "ewc_loss_parallel": 7.581710815429688e-05, "grad_norm": 5.520184516906738, "learning_rate": 1e-05, "loss": 0.513, "mean_token_accuracy": 0.8565683960914612, "num_tokens": 182021510.0, "step": 4792 }, { "epoch": 0.609718865284315, "ewc_loss": 0.08642578125, "ewc_loss_diag": 0.000102996826171875, "ewc_loss_parallel": 7.62939453125e-05, "grad_norm": 5.634114742279053, "learning_rate": 1e-05, "loss": 0.5214, "mean_token_accuracy": 0.8581490516662598, "num_tokens": 182061178.0, "step": 4793 }, { "epoch": 0.6098460755629055, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.000102996826171875, "ewc_loss_parallel": 7.390975952148438e-05, "grad_norm": 5.392507553100586, "learning_rate": 1e-05, "loss": 0.565, "mean_token_accuracy": 0.842788577079773, "num_tokens": 182100575.0, "step": 4794 }, { "epoch": 0.609973285841496, "ewc_loss": 0.08837890625, "ewc_loss_diag": 0.000102996826171875, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.670873165130615, "learning_rate": 1e-05, "loss": 0.5209, "mean_token_accuracy": 0.8544522523880005, "num_tokens": 182136142.0, "step": 4795 }, { "epoch": 0.6101004961200865, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.000102996826171875, "ewc_loss_parallel": 7.104873657226562e-05, "grad_norm": 5.244009017944336, "learning_rate": 1e-05, "loss": 0.5435, "mean_token_accuracy": 0.8478816747665405, "num_tokens": 182173108.0, "step": 4796 }, { "epoch": 0.610227706398677, "ewc_loss": 0.09033203125, "ewc_loss_diag": 0.000102996826171875, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.80518102645874, "learning_rate": 1e-05, "loss": 0.5594, "mean_token_accuracy": 0.8444972634315491, "num_tokens": 182209931.0, "step": 4797 }, { "epoch": 0.6103549166772675, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.000102996826171875, "ewc_loss_parallel": 7.009506225585938e-05, "grad_norm": 5.092663764953613, "learning_rate": 1e-05, "loss": 0.4987, "mean_token_accuracy": 0.8615416288375854, "num_tokens": 182244128.0, "step": 4798 }, { "epoch": 0.610482126955858, "ewc_loss": 0.095703125, "ewc_loss_diag": 0.000102996826171875, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 6.036968231201172, "learning_rate": 1e-05, "loss": 0.5223, "mean_token_accuracy": 0.8555826544761658, "num_tokens": 182278112.0, "step": 4799 }, { "epoch": 0.6106093372344485, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.000102996826171875, "ewc_loss_parallel": 6.914138793945312e-05, "grad_norm": 4.96381139755249, "learning_rate": 1e-05, "loss": 0.4929, "mean_token_accuracy": 0.8601714968681335, "num_tokens": 182313753.0, "step": 4800 }, { "epoch": 0.6107365475130391, "ewc_loss": 0.10400390625, "ewc_loss_diag": 0.000102996826171875, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.409115791320801, "learning_rate": 1e-05, "loss": 0.4654, "mean_token_accuracy": 0.876741349697113, "num_tokens": 182348758.0, "step": 4801 }, { "epoch": 0.6108637577916296, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.000102996826171875, "ewc_loss_parallel": 7.104873657226562e-05, "grad_norm": 5.008107662200928, "learning_rate": 1e-05, "loss": 0.5187, "mean_token_accuracy": 0.8560472726821899, "num_tokens": 182381084.0, "step": 4802 }, { "epoch": 0.61099096807022, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.000102996826171875, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.797713756561279, "learning_rate": 1e-05, "loss": 0.5481, "mean_token_accuracy": 0.8575273752212524, "num_tokens": 182420530.0, "step": 4803 }, { "epoch": 0.6111181783488105, "ewc_loss": 0.09130859375, "ewc_loss_diag": 0.000102996826171875, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.390620231628418, "learning_rate": 1e-05, "loss": 0.5218, "mean_token_accuracy": 0.8571759462356567, "num_tokens": 182462286.0, "step": 4804 }, { "epoch": 0.6112453886274011, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.000102996826171875, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.817775726318359, "learning_rate": 1e-05, "loss": 0.4842, "mean_token_accuracy": 0.8717600107192993, "num_tokens": 182497209.0, "step": 4805 }, { "epoch": 0.6113725989059916, "ewc_loss": 0.10009765625, "ewc_loss_diag": 0.000102996826171875, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 5.8131232261657715, "learning_rate": 1e-05, "loss": 0.4839, "mean_token_accuracy": 0.869111180305481, "num_tokens": 182530609.0, "step": 4806 }, { "epoch": 0.6114998091845821, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.000102996826171875, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.43749475479126, "learning_rate": 1e-05, "loss": 0.5094, "mean_token_accuracy": 0.8638904094696045, "num_tokens": 182566698.0, "step": 4807 }, { "epoch": 0.6116270194631727, "ewc_loss": 0.10107421875, "ewc_loss_diag": 0.000102996826171875, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 5.892886638641357, "learning_rate": 1e-05, "loss": 0.5202, "mean_token_accuracy": 0.863400399684906, "num_tokens": 182609524.0, "step": 4808 }, { "epoch": 0.6117542297417632, "ewc_loss": 0.09130859375, "ewc_loss_diag": 0.000102996826171875, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.986813545227051, "learning_rate": 1e-05, "loss": 0.5177, "mean_token_accuracy": 0.8591806292533875, "num_tokens": 182649339.0, "step": 4809 }, { "epoch": 0.6118814400203536, "ewc_loss": 0.09521484375, "ewc_loss_diag": 0.000102996826171875, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.848837375640869, "learning_rate": 1e-05, "loss": 0.5274, "mean_token_accuracy": 0.8593355417251587, "num_tokens": 182684142.0, "step": 4810 }, { "epoch": 0.6120086502989441, "ewc_loss": 0.08642578125, "ewc_loss_diag": 0.000102996826171875, "ewc_loss_parallel": 7.62939453125e-05, "grad_norm": 5.609694004058838, "learning_rate": 1e-05, "loss": 0.5107, "mean_token_accuracy": 0.8554893732070923, "num_tokens": 182722857.0, "step": 4811 }, { "epoch": 0.6121358605775347, "ewc_loss": 0.09326171875, "ewc_loss_diag": 0.000102996826171875, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.8473358154296875, "learning_rate": 1e-05, "loss": 0.5328, "mean_token_accuracy": 0.8510525226593018, "num_tokens": 182760196.0, "step": 4812 }, { "epoch": 0.6122630708561252, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.000102996826171875, "ewc_loss_parallel": 7.295608520507812e-05, "grad_norm": 5.36106538772583, "learning_rate": 1e-05, "loss": 0.5117, "mean_token_accuracy": 0.8538124561309814, "num_tokens": 182795807.0, "step": 4813 }, { "epoch": 0.6123902811347157, "ewc_loss": 0.0927734375, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.8922624588012695, "learning_rate": 1e-05, "loss": 0.5466, "mean_token_accuracy": 0.8481944799423218, "num_tokens": 182837956.0, "step": 4814 }, { "epoch": 0.6125174914133062, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 7.152557373046875e-05, "grad_norm": 5.196489334106445, "learning_rate": 1e-05, "loss": 0.4876, "mean_token_accuracy": 0.8633347749710083, "num_tokens": 182873374.0, "step": 4815 }, { "epoch": 0.6126447016918967, "ewc_loss": 0.095703125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 6.024472713470459, "learning_rate": 1e-05, "loss": 0.4963, "mean_token_accuracy": 0.8700680732727051, "num_tokens": 182908018.0, "step": 4816 }, { "epoch": 0.6127719119704872, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 7.104873657226562e-05, "grad_norm": 5.096940994262695, "learning_rate": 1e-05, "loss": 0.5162, "mean_token_accuracy": 0.8547469973564148, "num_tokens": 182950314.0, "step": 4817 }, { "epoch": 0.6128991222490777, "ewc_loss": 0.099609375, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 6.230693340301514, "learning_rate": 1e-05, "loss": 0.5382, "mean_token_accuracy": 0.8557987213134766, "num_tokens": 182988434.0, "step": 4818 }, { "epoch": 0.6130263325276682, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 7.200241088867188e-05, "grad_norm": 5.077215194702148, "learning_rate": 1e-05, "loss": 0.4841, "mean_token_accuracy": 0.8634483814239502, "num_tokens": 183026355.0, "step": 4819 }, { "epoch": 0.6131535428062588, "ewc_loss": 0.10400390625, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.469805717468262, "learning_rate": 1e-05, "loss": 0.5314, "mean_token_accuracy": 0.8555096387863159, "num_tokens": 183063031.0, "step": 4820 }, { "epoch": 0.6132807530848493, "ewc_loss": 0.0869140625, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.209734916687012, "learning_rate": 1e-05, "loss": 0.5646, "mean_token_accuracy": 0.8457621335983276, "num_tokens": 183102082.0, "step": 4821 }, { "epoch": 0.6134079633634397, "ewc_loss": 0.10791015625, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.601230621337891, "learning_rate": 1e-05, "loss": 0.5302, "mean_token_accuracy": 0.8556938171386719, "num_tokens": 183137961.0, "step": 4822 }, { "epoch": 0.6135351736420303, "ewc_loss": 0.091796875, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.451178073883057, "learning_rate": 1e-05, "loss": 0.477, "mean_token_accuracy": 0.8695324063301086, "num_tokens": 183173779.0, "step": 4823 }, { "epoch": 0.6136623839206208, "ewc_loss": 0.10498046875, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.503269195556641, "learning_rate": 1e-05, "loss": 0.5541, "mean_token_accuracy": 0.8495075702667236, "num_tokens": 183215044.0, "step": 4824 }, { "epoch": 0.6137895941992113, "ewc_loss": 0.0927734375, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.640383243560791, "learning_rate": 1e-05, "loss": 0.4956, "mean_token_accuracy": 0.8630474209785461, "num_tokens": 183248793.0, "step": 4825 }, { "epoch": 0.6139168044778018, "ewc_loss": 0.099609375, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 6.215153217315674, "learning_rate": 1e-05, "loss": 0.5311, "mean_token_accuracy": 0.855351448059082, "num_tokens": 183291180.0, "step": 4826 }, { "epoch": 0.6140440147563924, "ewc_loss": 0.0927734375, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.6814141273498535, "learning_rate": 1e-05, "loss": 0.5122, "mean_token_accuracy": 0.8572373986244202, "num_tokens": 183326214.0, "step": 4827 }, { "epoch": 0.6141712250349828, "ewc_loss": 0.0908203125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.878962516784668, "learning_rate": 1e-05, "loss": 0.5042, "mean_token_accuracy": 0.8611539602279663, "num_tokens": 183362403.0, "step": 4828 }, { "epoch": 0.6142984353135733, "ewc_loss": 0.09130859375, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.661080360412598, "learning_rate": 1e-05, "loss": 0.5035, "mean_token_accuracy": 0.8626975417137146, "num_tokens": 183407419.0, "step": 4829 }, { "epoch": 0.6144256455921638, "ewc_loss": 0.0859375, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 7.581710815429688e-05, "grad_norm": 5.611368656158447, "learning_rate": 1e-05, "loss": 0.5109, "mean_token_accuracy": 0.8562207818031311, "num_tokens": 183443140.0, "step": 4830 }, { "epoch": 0.6145528558707544, "ewc_loss": 0.08935546875, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.638487339019775, "learning_rate": 1e-05, "loss": 0.576, "mean_token_accuracy": 0.8415676951408386, "num_tokens": 183485986.0, "step": 4831 }, { "epoch": 0.6146800661493449, "ewc_loss": 0.08447265625, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 7.43865966796875e-05, "grad_norm": 5.441840171813965, "learning_rate": 1e-05, "loss": 0.5752, "mean_token_accuracy": 0.8418104648590088, "num_tokens": 183525302.0, "step": 4832 }, { "epoch": 0.6148072764279354, "ewc_loss": 0.08837890625, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.6388115882873535, "learning_rate": 1e-05, "loss": 0.5292, "mean_token_accuracy": 0.8511884212493896, "num_tokens": 183561575.0, "step": 4833 }, { "epoch": 0.6149344867065258, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 7.200241088867188e-05, "grad_norm": 5.326115131378174, "learning_rate": 1e-05, "loss": 0.4931, "mean_token_accuracy": 0.8648074269294739, "num_tokens": 183596465.0, "step": 4834 }, { "epoch": 0.6150616969851164, "ewc_loss": 0.0888671875, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.628205299377441, "learning_rate": 1e-05, "loss": 0.5696, "mean_token_accuracy": 0.8360534906387329, "num_tokens": 183631733.0, "step": 4835 }, { "epoch": 0.6151889072637069, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 7.152557373046875e-05, "grad_norm": 5.225773811340332, "learning_rate": 1e-05, "loss": 0.526, "mean_token_accuracy": 0.8534166812896729, "num_tokens": 183676138.0, "step": 4836 }, { "epoch": 0.6153161175422974, "ewc_loss": 0.08984375, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.7071533203125, "learning_rate": 1e-05, "loss": 0.5073, "mean_token_accuracy": 0.8570245504379272, "num_tokens": 183711004.0, "step": 4837 }, { "epoch": 0.615443327820888, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 7.009506225585938e-05, "grad_norm": 5.125174522399902, "learning_rate": 1e-05, "loss": 0.448, "mean_token_accuracy": 0.8781670331954956, "num_tokens": 183744945.0, "step": 4838 }, { "epoch": 0.6155705380994785, "ewc_loss": 0.091796875, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.854790687561035, "learning_rate": 1e-05, "loss": 0.498, "mean_token_accuracy": 0.8664776086807251, "num_tokens": 183779119.0, "step": 4839 }, { "epoch": 0.6156977483780689, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 6.914138793945312e-05, "grad_norm": 5.000940799713135, "learning_rate": 1e-05, "loss": 0.477, "mean_token_accuracy": 0.8640699982643127, "num_tokens": 183814136.0, "step": 4840 }, { "epoch": 0.6158249586566594, "ewc_loss": 0.0986328125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 6.124513149261475, "learning_rate": 1e-05, "loss": 0.5365, "mean_token_accuracy": 0.8521715998649597, "num_tokens": 183850957.0, "step": 4841 }, { "epoch": 0.61595216893525, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 6.961822509765625e-05, "grad_norm": 4.878889560699463, "learning_rate": 1e-05, "loss": 0.4782, "mean_token_accuracy": 0.8672771453857422, "num_tokens": 183890049.0, "step": 4842 }, { "epoch": 0.6160793792138405, "ewc_loss": 0.10791015625, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.5749006271362305, "learning_rate": 1e-05, "loss": 0.5279, "mean_token_accuracy": 0.8580085039138794, "num_tokens": 183933055.0, "step": 4843 }, { "epoch": 0.616206589492431, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 7.295608520507812e-05, "grad_norm": 4.9901227951049805, "learning_rate": 1e-05, "loss": 0.5097, "mean_token_accuracy": 0.8570494651794434, "num_tokens": 183975059.0, "step": 4844 }, { "epoch": 0.6163337997710215, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 0.00010967254638671875, "grad_norm": 6.999492645263672, "learning_rate": 1e-05, "loss": 0.5719, "mean_token_accuracy": 0.8511210083961487, "num_tokens": 184015130.0, "step": 4845 }, { "epoch": 0.616461010049612, "ewc_loss": 0.09716796875, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.538054466247559, "learning_rate": 1e-05, "loss": 0.5675, "mean_token_accuracy": 0.8467348217964172, "num_tokens": 184048416.0, "step": 4846 }, { "epoch": 0.6165882203282025, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.937641143798828, "learning_rate": 1e-05, "loss": 0.5778, "mean_token_accuracy": 0.847625732421875, "num_tokens": 184081591.0, "step": 4847 }, { "epoch": 0.616715430606793, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.00010395050048828125, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 5.9485249519348145, "learning_rate": 1e-05, "loss": 0.4647, "mean_token_accuracy": 0.877324104309082, "num_tokens": 184118078.0, "step": 4848 }, { "epoch": 0.6168426408853835, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.00010395050048828125, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.50512170791626, "learning_rate": 1e-05, "loss": 0.5152, "mean_token_accuracy": 0.8623675107955933, "num_tokens": 184154860.0, "step": 4849 }, { "epoch": 0.6169698511639741, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.00010395050048828125, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 5.959543704986572, "learning_rate": 1e-05, "loss": 0.4933, "mean_token_accuracy": 0.8704448342323303, "num_tokens": 184186425.0, "step": 4850 }, { "epoch": 0.6170970614425646, "ewc_loss": 0.09716796875, "ewc_loss_diag": 0.00010395050048828125, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 6.105288505554199, "learning_rate": 1e-05, "loss": 0.531, "mean_token_accuracy": 0.8505908250808716, "num_tokens": 184224891.0, "step": 4851 }, { "epoch": 0.617224271721155, "ewc_loss": 0.095703125, "ewc_loss_diag": 0.00010395050048828125, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.825984954833984, "learning_rate": 1e-05, "loss": 0.4828, "mean_token_accuracy": 0.8689953684806824, "num_tokens": 184266648.0, "step": 4852 }, { "epoch": 0.6173514819997455, "ewc_loss": 0.09033203125, "ewc_loss_diag": 0.00010395050048828125, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.794098854064941, "learning_rate": 1e-05, "loss": 0.5769, "mean_token_accuracy": 0.8451725840568542, "num_tokens": 184301453.0, "step": 4853 }, { "epoch": 0.6174786922783361, "ewc_loss": 0.091796875, "ewc_loss_diag": 0.00010395050048828125, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.734116554260254, "learning_rate": 1e-05, "loss": 0.5321, "mean_token_accuracy": 0.8549627065658569, "num_tokens": 184344271.0, "step": 4854 }, { "epoch": 0.6176059025569266, "ewc_loss": 0.087890625, "ewc_loss_diag": 0.00010395050048828125, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.569954872131348, "learning_rate": 1e-05, "loss": 0.492, "mean_token_accuracy": 0.862399697303772, "num_tokens": 184378212.0, "step": 4855 }, { "epoch": 0.6177331128355171, "ewc_loss": 0.08984375, "ewc_loss_diag": 0.00010395050048828125, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.669610500335693, "learning_rate": 1e-05, "loss": 0.4952, "mean_token_accuracy": 0.863647997379303, "num_tokens": 184414177.0, "step": 4856 }, { "epoch": 0.6178603231141077, "ewc_loss": 0.0849609375, "ewc_loss_diag": 0.00010395050048828125, "ewc_loss_parallel": 7.486343383789062e-05, "grad_norm": 5.4331841468811035, "learning_rate": 1e-05, "loss": 0.5494, "mean_token_accuracy": 0.8463464975357056, "num_tokens": 184455088.0, "step": 4857 }, { "epoch": 0.6179875333926982, "ewc_loss": 0.08837890625, "ewc_loss_diag": 0.00010395050048828125, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.648735046386719, "learning_rate": 1e-05, "loss": 0.5591, "mean_token_accuracy": 0.8471493721008301, "num_tokens": 184496574.0, "step": 4858 }, { "epoch": 0.6181147436712886, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.00010395050048828125, "ewc_loss_parallel": 7.295608520507812e-05, "grad_norm": 5.327735424041748, "learning_rate": 1e-05, "loss": 0.5373, "mean_token_accuracy": 0.8517372608184814, "num_tokens": 184542303.0, "step": 4859 }, { "epoch": 0.6182419539498791, "ewc_loss": 0.0888671875, "ewc_loss_diag": 0.00010395050048828125, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.665605068206787, "learning_rate": 1e-05, "loss": 0.5286, "mean_token_accuracy": 0.856352686882019, "num_tokens": 184583337.0, "step": 4860 }, { "epoch": 0.6183691642284697, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.00010395050048828125, "ewc_loss_parallel": 7.200241088867188e-05, "grad_norm": 5.219322681427002, "learning_rate": 1e-05, "loss": 0.5289, "mean_token_accuracy": 0.8528134822845459, "num_tokens": 184622424.0, "step": 4861 }, { "epoch": 0.6184963745070602, "ewc_loss": 0.08984375, "ewc_loss_diag": 0.00010395050048828125, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.765151500701904, "learning_rate": 1e-05, "loss": 0.5535, "mean_token_accuracy": 0.8422461748123169, "num_tokens": 184660137.0, "step": 4862 }, { "epoch": 0.6186235847856507, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.00010395050048828125, "ewc_loss_parallel": 6.961822509765625e-05, "grad_norm": 5.070774555206299, "learning_rate": 1e-05, "loss": 0.5108, "mean_token_accuracy": 0.8552258014678955, "num_tokens": 184698353.0, "step": 4863 }, { "epoch": 0.6187507950642412, "ewc_loss": 0.09423828125, "ewc_loss_diag": 0.00010395050048828125, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.996160507202148, "learning_rate": 1e-05, "loss": 0.5495, "mean_token_accuracy": 0.8484259843826294, "num_tokens": 184738734.0, "step": 4864 }, { "epoch": 0.6188780053428317, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.00010395050048828125, "ewc_loss_parallel": 6.914138793945312e-05, "grad_norm": 4.909630298614502, "learning_rate": 1e-05, "loss": 0.5291, "mean_token_accuracy": 0.8472157716751099, "num_tokens": 184774525.0, "step": 4865 }, { "epoch": 0.6190052156214222, "ewc_loss": 0.10400390625, "ewc_loss_diag": 0.00010395050048828125, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.422396659851074, "learning_rate": 1e-05, "loss": 0.5107, "mean_token_accuracy": 0.8619654774665833, "num_tokens": 184809028.0, "step": 4866 }, { "epoch": 0.6191324259000127, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.00010395050048828125, "ewc_loss_parallel": 6.961822509765625e-05, "grad_norm": 4.884214878082275, "learning_rate": 1e-05, "loss": 0.4394, "mean_token_accuracy": 0.8759198188781738, "num_tokens": 184845479.0, "step": 4867 }, { "epoch": 0.6192596361786032, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.00010395050048828125, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 6.957079887390137, "learning_rate": 1e-05, "loss": 0.6001, "mean_token_accuracy": 0.8415279984474182, "num_tokens": 184886596.0, "step": 4868 }, { "epoch": 0.6193868464571938, "ewc_loss": 0.091796875, "ewc_loss_diag": 0.00010395050048828125, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.344923496246338, "learning_rate": 1e-05, "loss": 0.5297, "mean_token_accuracy": 0.8540635108947754, "num_tokens": 184923763.0, "step": 4869 }, { "epoch": 0.6195140567357843, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.00010395050048828125, "ewc_loss_parallel": 0.00011157989501953125, "grad_norm": 7.117282390594482, "learning_rate": 1e-05, "loss": 0.5391, "mean_token_accuracy": 0.8634376525878906, "num_tokens": 184959882.0, "step": 4870 }, { "epoch": 0.6196412670143747, "ewc_loss": 0.10595703125, "ewc_loss_diag": 0.00010395050048828125, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 5.9609694480896, "learning_rate": 1e-05, "loss": 0.4798, "mean_token_accuracy": 0.8721151351928711, "num_tokens": 184999063.0, "step": 4871 }, { "epoch": 0.6197684772929652, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.709876537322998, "learning_rate": 1e-05, "loss": 0.4877, "mean_token_accuracy": 0.8736572265625, "num_tokens": 185040016.0, "step": 4872 }, { "epoch": 0.6198956875715558, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.067457675933838, "learning_rate": 1e-05, "loss": 0.5319, "mean_token_accuracy": 0.854644238948822, "num_tokens": 185075085.0, "step": 4873 }, { "epoch": 0.6200228978501463, "ewc_loss": 0.10009765625, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.278645992279053, "learning_rate": 1e-05, "loss": 0.537, "mean_token_accuracy": 0.8569468855857849, "num_tokens": 185111621.0, "step": 4874 }, { "epoch": 0.6201501081287368, "ewc_loss": 0.0986328125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 5.913153171539307, "learning_rate": 1e-05, "loss": 0.551, "mean_token_accuracy": 0.8507273197174072, "num_tokens": 185149681.0, "step": 4875 }, { "epoch": 0.6202773184073274, "ewc_loss": 0.0927734375, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.944158554077148, "learning_rate": 1e-05, "loss": 0.501, "mean_token_accuracy": 0.8634335994720459, "num_tokens": 185186449.0, "step": 4876 }, { "epoch": 0.6204045286859178, "ewc_loss": 0.09375, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.753615856170654, "learning_rate": 1e-05, "loss": 0.4972, "mean_token_accuracy": 0.8633943796157837, "num_tokens": 185219324.0, "step": 4877 }, { "epoch": 0.6205317389645083, "ewc_loss": 0.0888671875, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.73892068862915, "learning_rate": 1e-05, "loss": 0.5226, "mean_token_accuracy": 0.8544108867645264, "num_tokens": 185256049.0, "step": 4878 }, { "epoch": 0.6206589492430988, "ewc_loss": 0.0888671875, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.612151145935059, "learning_rate": 1e-05, "loss": 0.4926, "mean_token_accuracy": 0.8615309000015259, "num_tokens": 185293304.0, "step": 4879 }, { "epoch": 0.6207861595216894, "ewc_loss": 0.0888671875, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.621337413787842, "learning_rate": 1e-05, "loss": 0.5017, "mean_token_accuracy": 0.8602737188339233, "num_tokens": 185334007.0, "step": 4880 }, { "epoch": 0.6209133698002799, "ewc_loss": 0.0859375, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 7.581710815429688e-05, "grad_norm": 5.50111722946167, "learning_rate": 1e-05, "loss": 0.5391, "mean_token_accuracy": 0.8491169810295105, "num_tokens": 185372932.0, "step": 4881 }, { "epoch": 0.6210405800788704, "ewc_loss": 0.0869140625, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 7.62939453125e-05, "grad_norm": 5.573032379150391, "learning_rate": 1e-05, "loss": 0.4861, "mean_token_accuracy": 0.8637557625770569, "num_tokens": 185410267.0, "step": 4882 }, { "epoch": 0.6211677903574608, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 7.2479248046875e-05, "grad_norm": 5.403216361999512, "learning_rate": 1e-05, "loss": 0.5334, "mean_token_accuracy": 0.8506355881690979, "num_tokens": 185445838.0, "step": 4883 }, { "epoch": 0.6212950006360514, "ewc_loss": 0.087890625, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.588055610656738, "learning_rate": 1e-05, "loss": 0.5063, "mean_token_accuracy": 0.8604491949081421, "num_tokens": 185486002.0, "step": 4884 }, { "epoch": 0.6214222109146419, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 7.104873657226562e-05, "grad_norm": 5.2852020263671875, "learning_rate": 1e-05, "loss": 0.515, "mean_token_accuracy": 0.8533253073692322, "num_tokens": 185524501.0, "step": 4885 }, { "epoch": 0.6215494211932324, "ewc_loss": 0.087890625, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.655372142791748, "learning_rate": 1e-05, "loss": 0.5098, "mean_token_accuracy": 0.8549414873123169, "num_tokens": 185562105.0, "step": 4886 }, { "epoch": 0.621676631471823, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 7.009506225585938e-05, "grad_norm": 5.1432342529296875, "learning_rate": 1e-05, "loss": 0.4847, "mean_token_accuracy": 0.8671828508377075, "num_tokens": 185602821.0, "step": 4887 }, { "epoch": 0.6218038417504135, "ewc_loss": 0.091796875, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.863705635070801, "learning_rate": 1e-05, "loss": 0.5475, "mean_token_accuracy": 0.8484472036361694, "num_tokens": 185643188.0, "step": 4888 }, { "epoch": 0.6219310520290039, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 6.818771362304688e-05, "grad_norm": 4.955235481262207, "learning_rate": 1e-05, "loss": 0.5581, "mean_token_accuracy": 0.8419148921966553, "num_tokens": 185685770.0, "step": 4889 }, { "epoch": 0.6220582623075944, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.254274845123291, "learning_rate": 1e-05, "loss": 0.5092, "mean_token_accuracy": 0.8605703115463257, "num_tokens": 185718981.0, "step": 4890 }, { "epoch": 0.622185472586185, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 6.818771362304688e-05, "grad_norm": 4.836266040802002, "learning_rate": 1e-05, "loss": 0.4883, "mean_token_accuracy": 0.8616886734962463, "num_tokens": 185752184.0, "step": 4891 }, { "epoch": 0.6223126828647755, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.8264851570129395, "learning_rate": 1e-05, "loss": 0.5512, "mean_token_accuracy": 0.8525741100311279, "num_tokens": 185791405.0, "step": 4892 }, { "epoch": 0.622439893143366, "ewc_loss": 0.0869140625, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 7.62939453125e-05, "grad_norm": 5.115646839141846, "learning_rate": 1e-05, "loss": 0.5777, "mean_token_accuracy": 0.8372460603713989, "num_tokens": 185832227.0, "step": 4893 }, { "epoch": 0.6225671034219565, "ewc_loss": 0.125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 0.00011491775512695312, "grad_norm": 7.250916481018066, "learning_rate": 1e-05, "loss": 0.5442, "mean_token_accuracy": 0.861141562461853, "num_tokens": 185867175.0, "step": 4894 }, { "epoch": 0.622694313700547, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 5.887730598449707, "learning_rate": 1e-05, "loss": 0.5423, "mean_token_accuracy": 0.8536957502365112, "num_tokens": 185907582.0, "step": 4895 }, { "epoch": 0.6228215239791375, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 6.925704002380371, "learning_rate": 1e-05, "loss": 0.572, "mean_token_accuracy": 0.8448244333267212, "num_tokens": 185944625.0, "step": 4896 }, { "epoch": 0.622948734257728, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.1973419189453125, "learning_rate": 1e-05, "loss": 0.5053, "mean_token_accuracy": 0.8686728477478027, "num_tokens": 185975579.0, "step": 4897 }, { "epoch": 0.6230759445363185, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.402196884155273, "learning_rate": 1e-05, "loss": 0.519, "mean_token_accuracy": 0.8593035340309143, "num_tokens": 186016948.0, "step": 4898 }, { "epoch": 0.6232031548149091, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.019739151000977, "learning_rate": 1e-05, "loss": 0.5525, "mean_token_accuracy": 0.8501998782157898, "num_tokens": 186060700.0, "step": 4899 }, { "epoch": 0.6233303650934996, "ewc_loss": 0.0947265625, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 6.038945198059082, "learning_rate": 1e-05, "loss": 0.4742, "mean_token_accuracy": 0.8720642924308777, "num_tokens": 186099624.0, "step": 4900 }, { "epoch": 0.62345757537209, "ewc_loss": 0.09375, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.8219146728515625, "learning_rate": 1e-05, "loss": 0.5194, "mean_token_accuracy": 0.8558675050735474, "num_tokens": 186142360.0, "step": 4901 }, { "epoch": 0.6235847856506805, "ewc_loss": 0.08984375, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.80385684967041, "learning_rate": 1e-05, "loss": 0.4857, "mean_token_accuracy": 0.8684180378913879, "num_tokens": 186182642.0, "step": 4902 }, { "epoch": 0.6237119959292711, "ewc_loss": 0.08984375, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.678733825683594, "learning_rate": 1e-05, "loss": 0.53, "mean_token_accuracy": 0.8528324365615845, "num_tokens": 186218867.0, "step": 4903 }, { "epoch": 0.6238392062078616, "ewc_loss": 0.087890625, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.652299880981445, "learning_rate": 1e-05, "loss": 0.4721, "mean_token_accuracy": 0.8689945936203003, "num_tokens": 186257424.0, "step": 4904 }, { "epoch": 0.6239664164864521, "ewc_loss": 0.0869140625, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 7.62939453125e-05, "grad_norm": 5.540436267852783, "learning_rate": 1e-05, "loss": 0.5071, "mean_token_accuracy": 0.8576645851135254, "num_tokens": 186298588.0, "step": 4905 }, { "epoch": 0.6240936267650427, "ewc_loss": 0.0869140625, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.588961124420166, "learning_rate": 1e-05, "loss": 0.4806, "mean_token_accuracy": 0.8656139373779297, "num_tokens": 186334638.0, "step": 4906 }, { "epoch": 0.6242208370436332, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 7.295608520507812e-05, "grad_norm": 5.438260555267334, "learning_rate": 1e-05, "loss": 0.531, "mean_token_accuracy": 0.8508287668228149, "num_tokens": 186368303.0, "step": 4907 }, { "epoch": 0.6243480473222236, "ewc_loss": 0.087890625, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.582254886627197, "learning_rate": 1e-05, "loss": 0.46, "mean_token_accuracy": 0.8716470003128052, "num_tokens": 186406133.0, "step": 4908 }, { "epoch": 0.6244752576008141, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 7.152557373046875e-05, "grad_norm": 5.310597896575928, "learning_rate": 1e-05, "loss": 0.4418, "mean_token_accuracy": 0.8758906722068787, "num_tokens": 186440763.0, "step": 4909 }, { "epoch": 0.6246024678794047, "ewc_loss": 0.087890625, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.66122579574585, "learning_rate": 1e-05, "loss": 0.536, "mean_token_accuracy": 0.8565511703491211, "num_tokens": 186479662.0, "step": 4910 }, { "epoch": 0.6247296781579952, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 7.009506225585938e-05, "grad_norm": 5.163968086242676, "learning_rate": 1e-05, "loss": 0.5498, "mean_token_accuracy": 0.8468215465545654, "num_tokens": 186518944.0, "step": 4911 }, { "epoch": 0.6248568884365857, "ewc_loss": 0.091796875, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.835067272186279, "learning_rate": 1e-05, "loss": 0.4795, "mean_token_accuracy": 0.866841197013855, "num_tokens": 186555902.0, "step": 4912 }, { "epoch": 0.6249840987151762, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 6.818771362304688e-05, "grad_norm": 4.990303039550781, "learning_rate": 1e-05, "loss": 0.4507, "mean_token_accuracy": 0.874511182308197, "num_tokens": 186591962.0, "step": 4913 }, { "epoch": 0.6251113089937667, "ewc_loss": 0.09814453125, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 6.192983150482178, "learning_rate": 1e-05, "loss": 0.554, "mean_token_accuracy": 0.8507289886474609, "num_tokens": 186631841.0, "step": 4914 }, { "epoch": 0.6252385192723572, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 6.818771362304688e-05, "grad_norm": 4.826173782348633, "learning_rate": 1e-05, "loss": 0.4701, "mean_token_accuracy": 0.8678185343742371, "num_tokens": 186677963.0, "step": 4915 }, { "epoch": 0.6253657295509477, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.787106990814209, "learning_rate": 1e-05, "loss": 0.5551, "mean_token_accuracy": 0.8553177118301392, "num_tokens": 186715658.0, "step": 4916 }, { "epoch": 0.6254929398295382, "ewc_loss": 0.08544921875, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 7.486343383789062e-05, "grad_norm": 5.0614776611328125, "learning_rate": 1e-05, "loss": 0.4841, "mean_token_accuracy": 0.8642711639404297, "num_tokens": 186749405.0, "step": 4917 }, { "epoch": 0.6256201501081288, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 0.00011587142944335938, "grad_norm": 7.284192085266113, "learning_rate": 1e-05, "loss": 0.5792, "mean_token_accuracy": 0.8480257987976074, "num_tokens": 186785667.0, "step": 4918 }, { "epoch": 0.6257473603867193, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 5.863026142120361, "learning_rate": 1e-05, "loss": 0.4863, "mean_token_accuracy": 0.8703088164329529, "num_tokens": 186817425.0, "step": 4919 }, { "epoch": 0.6258745706653097, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 6.993696212768555, "learning_rate": 1e-05, "loss": 0.5213, "mean_token_accuracy": 0.866154670715332, "num_tokens": 186859693.0, "step": 4920 }, { "epoch": 0.6260017809439002, "ewc_loss": 0.10986328125, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.216488361358643, "learning_rate": 1e-05, "loss": 0.5214, "mean_token_accuracy": 0.8663938045501709, "num_tokens": 186897774.0, "step": 4921 }, { "epoch": 0.6261289912224908, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.477315425872803, "learning_rate": 1e-05, "loss": 0.5222, "mean_token_accuracy": 0.8578934073448181, "num_tokens": 186935530.0, "step": 4922 }, { "epoch": 0.6262562015010813, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.056901454925537, "learning_rate": 1e-05, "loss": 0.5652, "mean_token_accuracy": 0.8464646339416504, "num_tokens": 186970778.0, "step": 4923 }, { "epoch": 0.6263834117796718, "ewc_loss": 0.09423828125, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 6.1044440269470215, "learning_rate": 1e-05, "loss": 0.5135, "mean_token_accuracy": 0.8577753305435181, "num_tokens": 187007338.0, "step": 4924 }, { "epoch": 0.6265106220582624, "ewc_loss": 0.0947265625, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.8614606857299805, "learning_rate": 1e-05, "loss": 0.5269, "mean_token_accuracy": 0.8583773970603943, "num_tokens": 187042614.0, "step": 4925 }, { "epoch": 0.6266378323368528, "ewc_loss": 0.09033203125, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.864628314971924, "learning_rate": 1e-05, "loss": 0.5343, "mean_token_accuracy": 0.8516131043434143, "num_tokens": 187081753.0, "step": 4926 }, { "epoch": 0.6267650426154433, "ewc_loss": 0.0908203125, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.673403739929199, "learning_rate": 1e-05, "loss": 0.5408, "mean_token_accuracy": 0.8505589962005615, "num_tokens": 187116758.0, "step": 4927 }, { "epoch": 0.6268922528940338, "ewc_loss": 0.08837890625, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.718989372253418, "learning_rate": 1e-05, "loss": 0.5432, "mean_token_accuracy": 0.8495213389396667, "num_tokens": 187152296.0, "step": 4928 }, { "epoch": 0.6270194631726244, "ewc_loss": 0.0859375, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 7.534027099609375e-05, "grad_norm": 5.522705554962158, "learning_rate": 1e-05, "loss": 0.5232, "mean_token_accuracy": 0.8514479398727417, "num_tokens": 187187619.0, "step": 4929 }, { "epoch": 0.6271466734512149, "ewc_loss": 0.08837890625, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.6649346351623535, "learning_rate": 1e-05, "loss": 0.5078, "mean_token_accuracy": 0.858069658279419, "num_tokens": 187227872.0, "step": 4930 }, { "epoch": 0.6272738837298054, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 7.2479248046875e-05, "grad_norm": 5.386719226837158, "learning_rate": 1e-05, "loss": 0.5162, "mean_token_accuracy": 0.8566269278526306, "num_tokens": 187265720.0, "step": 4931 }, { "epoch": 0.6274010940083958, "ewc_loss": 0.08935546875, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.693413734436035, "learning_rate": 1e-05, "loss": 0.601, "mean_token_accuracy": 0.8335795402526855, "num_tokens": 187307119.0, "step": 4932 }, { "epoch": 0.6275283042869864, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 7.05718994140625e-05, "grad_norm": 5.234766006469727, "learning_rate": 1e-05, "loss": 0.4946, "mean_token_accuracy": 0.8616679906845093, "num_tokens": 187342804.0, "step": 4933 }, { "epoch": 0.6276555145655769, "ewc_loss": 0.09228515625, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.827608108520508, "learning_rate": 1e-05, "loss": 0.5166, "mean_token_accuracy": 0.8582936525344849, "num_tokens": 187381540.0, "step": 4934 }, { "epoch": 0.6277827248441674, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 6.961822509765625e-05, "grad_norm": 5.0618391036987305, "learning_rate": 1e-05, "loss": 0.5763, "mean_token_accuracy": 0.8412987589836121, "num_tokens": 187424952.0, "step": 4935 }, { "epoch": 0.627909935122758, "ewc_loss": 0.09619140625, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 6.114121437072754, "learning_rate": 1e-05, "loss": 0.5522, "mean_token_accuracy": 0.8471125364303589, "num_tokens": 187463668.0, "step": 4936 }, { "epoch": 0.6280371454013485, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 6.866455078125e-05, "grad_norm": 4.924169063568115, "learning_rate": 1e-05, "loss": 0.5428, "mean_token_accuracy": 0.8529162406921387, "num_tokens": 187505962.0, "step": 4937 }, { "epoch": 0.6281643556799389, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.583316802978516, "learning_rate": 1e-05, "loss": 0.546, "mean_token_accuracy": 0.8514701128005981, "num_tokens": 187546342.0, "step": 4938 }, { "epoch": 0.6282915659585294, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 7.152557373046875e-05, "grad_norm": 5.007837295532227, "learning_rate": 1e-05, "loss": 0.5091, "mean_token_accuracy": 0.8561524152755737, "num_tokens": 187587378.0, "step": 4939 }, { "epoch": 0.62841877623712, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 7.070488452911377, "learning_rate": 1e-05, "loss": 0.5345, "mean_token_accuracy": 0.8594729900360107, "num_tokens": 187621972.0, "step": 4940 }, { "epoch": 0.6285459865157105, "ewc_loss": 0.09619140625, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.6033148765563965, "learning_rate": 1e-05, "loss": 0.5524, "mean_token_accuracy": 0.845479428768158, "num_tokens": 187657205.0, "step": 4941 }, { "epoch": 0.628673196794301, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 6.979449272155762, "learning_rate": 1e-05, "loss": 0.5623, "mean_token_accuracy": 0.8508399724960327, "num_tokens": 187698201.0, "step": 4942 }, { "epoch": 0.6288004070728915, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.067490100860596, "learning_rate": 1e-05, "loss": 0.4937, "mean_token_accuracy": 0.8669513463973999, "num_tokens": 187734883.0, "step": 4943 }, { "epoch": 0.628927617351482, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.478909015655518, "learning_rate": 1e-05, "loss": 0.4843, "mean_token_accuracy": 0.8712266087532043, "num_tokens": 187777459.0, "step": 4944 }, { "epoch": 0.6290548276300725, "ewc_loss": 0.10302734375, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.038211822509766, "learning_rate": 1e-05, "loss": 0.5475, "mean_token_accuracy": 0.8508840799331665, "num_tokens": 187819661.0, "step": 4945 }, { "epoch": 0.629182037908663, "ewc_loss": 0.09375, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 6.034910202026367, "learning_rate": 1e-05, "loss": 0.4993, "mean_token_accuracy": 0.8641178011894226, "num_tokens": 187857685.0, "step": 4946 }, { "epoch": 0.6293092481872535, "ewc_loss": 0.09619140625, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.916189193725586, "learning_rate": 1e-05, "loss": 0.5977, "mean_token_accuracy": 0.8349692821502686, "num_tokens": 187896614.0, "step": 4947 }, { "epoch": 0.6294364584658441, "ewc_loss": 0.08837890625, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.71795129776001, "learning_rate": 1e-05, "loss": 0.5388, "mean_token_accuracy": 0.8501538038253784, "num_tokens": 187925480.0, "step": 4948 }, { "epoch": 0.6295636687444346, "ewc_loss": 0.0927734375, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.820597171783447, "learning_rate": 1e-05, "loss": 0.5138, "mean_token_accuracy": 0.8585686683654785, "num_tokens": 187965045.0, "step": 4949 }, { "epoch": 0.629690879023025, "ewc_loss": 0.0849609375, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 7.43865966796875e-05, "grad_norm": 5.4861345291137695, "learning_rate": 1e-05, "loss": 0.5095, "mean_token_accuracy": 0.8571575880050659, "num_tokens": 188001633.0, "step": 4950 }, { "epoch": 0.6298180893016155, "ewc_loss": 0.08984375, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.790660381317139, "learning_rate": 1e-05, "loss": 0.5339, "mean_token_accuracy": 0.8544601798057556, "num_tokens": 188039931.0, "step": 4951 }, { "epoch": 0.6299452995802061, "ewc_loss": 0.08349609375, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 7.295608520507812e-05, "grad_norm": 5.318159580230713, "learning_rate": 1e-05, "loss": 0.5689, "mean_token_accuracy": 0.840509295463562, "num_tokens": 188085353.0, "step": 4952 }, { "epoch": 0.6300725098587966, "ewc_loss": 0.09130859375, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.8440327644348145, "learning_rate": 1e-05, "loss": 0.4568, "mean_token_accuracy": 0.876509428024292, "num_tokens": 188118966.0, "step": 4953 }, { "epoch": 0.6301997201373871, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 7.104873657226562e-05, "grad_norm": 5.183969497680664, "learning_rate": 1e-05, "loss": 0.4234, "mean_token_accuracy": 0.8810396194458008, "num_tokens": 188153680.0, "step": 4954 }, { "epoch": 0.6303269304159776, "ewc_loss": 0.09375, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.989499568939209, "learning_rate": 1e-05, "loss": 0.4761, "mean_token_accuracy": 0.8709179162979126, "num_tokens": 188189549.0, "step": 4955 }, { "epoch": 0.6304541406945681, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 6.961822509765625e-05, "grad_norm": 5.0575480461120605, "learning_rate": 1e-05, "loss": 0.529, "mean_token_accuracy": 0.8504638075828552, "num_tokens": 188230469.0, "step": 4956 }, { "epoch": 0.6305813509731586, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.256316184997559, "learning_rate": 1e-05, "loss": 0.5182, "mean_token_accuracy": 0.8595765233039856, "num_tokens": 188265074.0, "step": 4957 }, { "epoch": 0.6307085612517491, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 6.961822509765625e-05, "grad_norm": 5.003328323364258, "learning_rate": 1e-05, "loss": 0.5309, "mean_token_accuracy": 0.8503490686416626, "num_tokens": 188302759.0, "step": 4958 }, { "epoch": 0.6308357715303397, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.617766857147217, "learning_rate": 1e-05, "loss": 0.6315, "mean_token_accuracy": 0.8298168182373047, "num_tokens": 188337627.0, "step": 4959 }, { "epoch": 0.6309629818089302, "ewc_loss": 0.0869140625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 7.62939453125e-05, "grad_norm": 5.201023578643799, "learning_rate": 1e-05, "loss": 0.5026, "mean_token_accuracy": 0.8558995723724365, "num_tokens": 188365094.0, "step": 4960 }, { "epoch": 0.6310901920875207, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.813881874084473, "learning_rate": 1e-05, "loss": 0.5471, "mean_token_accuracy": 0.860295832157135, "num_tokens": 188403244.0, "step": 4961 }, { "epoch": 0.6312174023661112, "ewc_loss": 0.0947265625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.575481414794922, "learning_rate": 1e-05, "loss": 0.4894, "mean_token_accuracy": 0.8659746646881104, "num_tokens": 188442785.0, "step": 4962 }, { "epoch": 0.6313446126447017, "ewc_loss": 0.109375, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.675774097442627, "learning_rate": 1e-05, "loss": 0.5637, "mean_token_accuracy": 0.8503532409667969, "num_tokens": 188479298.0, "step": 4963 }, { "epoch": 0.6314718229232922, "ewc_loss": 0.10009765625, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 5.828114986419678, "learning_rate": 1e-05, "loss": 0.5771, "mean_token_accuracy": 0.8408210277557373, "num_tokens": 188519304.0, "step": 4964 }, { "epoch": 0.6315990332018827, "ewc_loss": 0.10302734375, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.286037921905518, "learning_rate": 1e-05, "loss": 0.5728, "mean_token_accuracy": 0.8409188985824585, "num_tokens": 188559372.0, "step": 4965 }, { "epoch": 0.6317262434804732, "ewc_loss": 0.09716796875, "ewc_loss_diag": 0.0001087188720703125, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.827640533447266, "learning_rate": 1e-05, "loss": 0.4744, "mean_token_accuracy": 0.8721285462379456, "num_tokens": 188595337.0, "step": 4966 }, { "epoch": 0.6318534537590638, "ewc_loss": 0.0908203125, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.900787353515625, "learning_rate": 1e-05, "loss": 0.5055, "mean_token_accuracy": 0.8597772717475891, "num_tokens": 188632988.0, "step": 4967 }, { "epoch": 0.6319806640376543, "ewc_loss": 0.095703125, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.779045581817627, "learning_rate": 1e-05, "loss": 0.4824, "mean_token_accuracy": 0.8714269399642944, "num_tokens": 188667205.0, "step": 4968 }, { "epoch": 0.6321078743162447, "ewc_loss": 0.0869140625, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 7.62939453125e-05, "grad_norm": 5.60331916809082, "learning_rate": 1e-05, "loss": 0.5344, "mean_token_accuracy": 0.8521323800086975, "num_tokens": 188707717.0, "step": 4969 }, { "epoch": 0.6322350845948352, "ewc_loss": 0.0927734375, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.743740558624268, "learning_rate": 1e-05, "loss": 0.4839, "mean_token_accuracy": 0.8667173981666565, "num_tokens": 188742811.0, "step": 4970 }, { "epoch": 0.6323622948734258, "ewc_loss": 0.0849609375, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 7.43865966796875e-05, "grad_norm": 5.41145658493042, "learning_rate": 1e-05, "loss": 0.502, "mean_token_accuracy": 0.8577185273170471, "num_tokens": 188781004.0, "step": 4971 }, { "epoch": 0.6324895051520163, "ewc_loss": 0.0908203125, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.744584560394287, "learning_rate": 1e-05, "loss": 0.5201, "mean_token_accuracy": 0.8531609177589417, "num_tokens": 188813923.0, "step": 4972 }, { "epoch": 0.6326167154306068, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 7.152557373046875e-05, "grad_norm": 5.272656440734863, "learning_rate": 1e-05, "loss": 0.4848, "mean_token_accuracy": 0.8608505725860596, "num_tokens": 188855375.0, "step": 4973 }, { "epoch": 0.6327439257091974, "ewc_loss": 0.0908203125, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.784532070159912, "learning_rate": 1e-05, "loss": 0.4618, "mean_token_accuracy": 0.873634934425354, "num_tokens": 188892807.0, "step": 4974 }, { "epoch": 0.6328711359877878, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 7.104873657226562e-05, "grad_norm": 5.165709018707275, "learning_rate": 1e-05, "loss": 0.4794, "mean_token_accuracy": 0.8657248616218567, "num_tokens": 188932942.0, "step": 4975 }, { "epoch": 0.6329983462663783, "ewc_loss": 0.0927734375, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.918426990509033, "learning_rate": 1e-05, "loss": 0.5141, "mean_token_accuracy": 0.8573368787765503, "num_tokens": 188968206.0, "step": 4976 }, { "epoch": 0.6331255565449688, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 6.961822509765625e-05, "grad_norm": 5.054327487945557, "learning_rate": 1e-05, "loss": 0.5265, "mean_token_accuracy": 0.8520402908325195, "num_tokens": 189001315.0, "step": 4977 }, { "epoch": 0.6332527668235594, "ewc_loss": 0.099609375, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.1649932861328125, "learning_rate": 1e-05, "loss": 0.5786, "mean_token_accuracy": 0.8403224945068359, "num_tokens": 189039729.0, "step": 4978 }, { "epoch": 0.6333799771021499, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 6.961822509765625e-05, "grad_norm": 4.964353561401367, "learning_rate": 1e-05, "loss": 0.4585, "mean_token_accuracy": 0.8733713626861572, "num_tokens": 189077455.0, "step": 4979 }, { "epoch": 0.6335071873807404, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.00011014938354492188, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.523491859436035, "learning_rate": 1e-05, "loss": 0.5501, "mean_token_accuracy": 0.8489751815795898, "num_tokens": 189118417.0, "step": 4980 }, { "epoch": 0.6336343976593308, "ewc_loss": 0.0869140625, "ewc_loss_diag": 0.00011014938354492188, "ewc_loss_parallel": 7.581710815429688e-05, "grad_norm": 5.080108642578125, "learning_rate": 1e-05, "loss": 0.4731, "mean_token_accuracy": 0.8696424961090088, "num_tokens": 189155581.0, "step": 4981 }, { "epoch": 0.6337616079379214, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.00011014938354492188, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.832632064819336, "learning_rate": 1e-05, "loss": 0.5443, "mean_token_accuracy": 0.8559134006500244, "num_tokens": 189191192.0, "step": 4982 }, { "epoch": 0.6338888182165119, "ewc_loss": 0.0947265625, "ewc_loss_diag": 0.00011014938354492188, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.487627029418945, "learning_rate": 1e-05, "loss": 0.5005, "mean_token_accuracy": 0.8641655445098877, "num_tokens": 189230475.0, "step": 4983 }, { "epoch": 0.6340160284951024, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.00011014938354492188, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.757535457611084, "learning_rate": 1e-05, "loss": 0.5374, "mean_token_accuracy": 0.8575050830841064, "num_tokens": 189265949.0, "step": 4984 }, { "epoch": 0.6341432387736929, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.00011014938354492188, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 5.824775695800781, "learning_rate": 1e-05, "loss": 0.5334, "mean_token_accuracy": 0.8563140034675598, "num_tokens": 189302914.0, "step": 4985 }, { "epoch": 0.6342704490522835, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.00011014938354492188, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.36726188659668, "learning_rate": 1e-05, "loss": 0.4981, "mean_token_accuracy": 0.8682586550712585, "num_tokens": 189347741.0, "step": 4986 }, { "epoch": 0.6343976593308739, "ewc_loss": 0.099609375, "ewc_loss_diag": 0.00011014938354492188, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 5.8547563552856445, "learning_rate": 1e-05, "loss": 0.5061, "mean_token_accuracy": 0.8666130304336548, "num_tokens": 189389672.0, "step": 4987 }, { "epoch": 0.6345248696094644, "ewc_loss": 0.09326171875, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.960442066192627, "learning_rate": 1e-05, "loss": 0.4872, "mean_token_accuracy": 0.8682661056518555, "num_tokens": 189425418.0, "step": 4988 }, { "epoch": 0.6346520798880549, "ewc_loss": 0.0966796875, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.8138556480407715, "learning_rate": 1e-05, "loss": 0.4913, "mean_token_accuracy": 0.8670972585678101, "num_tokens": 189457042.0, "step": 4989 }, { "epoch": 0.6347792901666455, "ewc_loss": 0.08837890625, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.623942852020264, "learning_rate": 1e-05, "loss": 0.5504, "mean_token_accuracy": 0.8470427393913269, "num_tokens": 189497135.0, "step": 4990 }, { "epoch": 0.634906500445236, "ewc_loss": 0.09375, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.772303104400635, "learning_rate": 1e-05, "loss": 0.4576, "mean_token_accuracy": 0.8768860101699829, "num_tokens": 189534829.0, "step": 4991 }, { "epoch": 0.6350337107238265, "ewc_loss": 0.08447265625, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 7.343292236328125e-05, "grad_norm": 5.412086486816406, "learning_rate": 1e-05, "loss": 0.5062, "mean_token_accuracy": 0.85187166929245, "num_tokens": 189570535.0, "step": 4992 }, { "epoch": 0.635160921002417, "ewc_loss": 0.091796875, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.779664993286133, "learning_rate": 1e-05, "loss": 0.4968, "mean_token_accuracy": 0.8658824563026428, "num_tokens": 189606448.0, "step": 4993 }, { "epoch": 0.6352881312810075, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 7.152557373046875e-05, "grad_norm": 5.263418197631836, "learning_rate": 1e-05, "loss": 0.5061, "mean_token_accuracy": 0.8590046167373657, "num_tokens": 189650880.0, "step": 4994 }, { "epoch": 0.635415341559598, "ewc_loss": 0.091796875, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.843671798706055, "learning_rate": 1e-05, "loss": 0.566, "mean_token_accuracy": 0.843206524848938, "num_tokens": 189690866.0, "step": 4995 }, { "epoch": 0.6355425518381885, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 7.05718994140625e-05, "grad_norm": 5.158374786376953, "learning_rate": 1e-05, "loss": 0.5956, "mean_token_accuracy": 0.829443633556366, "num_tokens": 189726891.0, "step": 4996 }, { "epoch": 0.6356697621167791, "ewc_loss": 0.0947265625, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.984554290771484, "learning_rate": 1e-05, "loss": 0.5128, "mean_token_accuracy": 0.8585128784179688, "num_tokens": 189763893.0, "step": 4997 }, { "epoch": 0.6357969723953696, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 6.961822509765625e-05, "grad_norm": 5.04600191116333, "learning_rate": 1e-05, "loss": 0.557, "mean_token_accuracy": 0.8427078723907471, "num_tokens": 189804122.0, "step": 4998 }, { "epoch": 0.63592418267396, "ewc_loss": 0.10107421875, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.246362686157227, "learning_rate": 1e-05, "loss": 0.5368, "mean_token_accuracy": 0.8543633222579956, "num_tokens": 189836355.0, "step": 4999 }, { "epoch": 0.6360513929525505, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 7.152557373046875e-05, "grad_norm": 4.998106956481934, "learning_rate": 1e-05, "loss": 0.4934, "mean_token_accuracy": 0.8611516952514648, "num_tokens": 189871839.0, "step": 5000 }, { "epoch": 0.6361786032311411, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.582053184509277, "learning_rate": 1e-05, "loss": 0.494, "mean_token_accuracy": 0.8684481382369995, "num_tokens": 189915631.0, "step": 5001 }, { "epoch": 0.6363058135097316, "ewc_loss": 0.0888671875, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.1723713874816895, "learning_rate": 1e-05, "loss": 0.529, "mean_token_accuracy": 0.850557804107666, "num_tokens": 189953635.0, "step": 5002 }, { "epoch": 0.6364330237883221, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.796055316925049, "learning_rate": 1e-05, "loss": 0.5292, "mean_token_accuracy": 0.8606134653091431, "num_tokens": 189999338.0, "step": 5003 }, { "epoch": 0.6365602340669126, "ewc_loss": 0.09521484375, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.54574728012085, "learning_rate": 1e-05, "loss": 0.5204, "mean_token_accuracy": 0.8591102361679077, "num_tokens": 190045804.0, "step": 5004 }, { "epoch": 0.6366874443455031, "ewc_loss": 0.11181640625, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.660261154174805, "learning_rate": 1e-05, "loss": 0.5478, "mean_token_accuracy": 0.854332685470581, "num_tokens": 190083350.0, "step": 5005 }, { "epoch": 0.6368146546240936, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 5.8028974533081055, "learning_rate": 1e-05, "loss": 0.5505, "mean_token_accuracy": 0.8495253920555115, "num_tokens": 190121910.0, "step": 5006 }, { "epoch": 0.6369418649026841, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.283552169799805, "learning_rate": 1e-05, "loss": 0.5098, "mean_token_accuracy": 0.8612840175628662, "num_tokens": 190163857.0, "step": 5007 }, { "epoch": 0.6370690751812746, "ewc_loss": 0.09912109375, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 5.823730945587158, "learning_rate": 1e-05, "loss": 0.5458, "mean_token_accuracy": 0.8511601686477661, "num_tokens": 190204337.0, "step": 5008 }, { "epoch": 0.6371962854598652, "ewc_loss": 0.09326171875, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.891181945800781, "learning_rate": 1e-05, "loss": 0.4777, "mean_token_accuracy": 0.8703864216804504, "num_tokens": 190240646.0, "step": 5009 }, { "epoch": 0.6373234957384557, "ewc_loss": 0.09619140625, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.787320137023926, "learning_rate": 1e-05, "loss": 0.5277, "mean_token_accuracy": 0.8542273640632629, "num_tokens": 190273507.0, "step": 5010 }, { "epoch": 0.6374507060170462, "ewc_loss": 0.08740234375, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 7.62939453125e-05, "grad_norm": 5.579246520996094, "learning_rate": 1e-05, "loss": 0.4972, "mean_token_accuracy": 0.8574924468994141, "num_tokens": 190306876.0, "step": 5011 }, { "epoch": 0.6375779162956366, "ewc_loss": 0.09423828125, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.778710842132568, "learning_rate": 1e-05, "loss": 0.5434, "mean_token_accuracy": 0.8474292159080505, "num_tokens": 190343126.0, "step": 5012 }, { "epoch": 0.6377051265742272, "ewc_loss": 0.0849609375, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 7.390975952148438e-05, "grad_norm": 5.37403678894043, "learning_rate": 1e-05, "loss": 0.5408, "mean_token_accuracy": 0.8482834100723267, "num_tokens": 190382089.0, "step": 5013 }, { "epoch": 0.6378323368528177, "ewc_loss": 0.09326171875, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.798209190368652, "learning_rate": 1e-05, "loss": 0.5527, "mean_token_accuracy": 0.8448995351791382, "num_tokens": 190419577.0, "step": 5014 }, { "epoch": 0.6379595471314082, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 7.200241088867188e-05, "grad_norm": 5.237918376922607, "learning_rate": 1e-05, "loss": 0.5634, "mean_token_accuracy": 0.8450673818588257, "num_tokens": 190459725.0, "step": 5015 }, { "epoch": 0.6380867574099988, "ewc_loss": 0.09375, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.893194198608398, "learning_rate": 1e-05, "loss": 0.565, "mean_token_accuracy": 0.8464102745056152, "num_tokens": 190493575.0, "step": 5016 }, { "epoch": 0.6382139676885893, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 7.104873657226562e-05, "grad_norm": 5.1146697998046875, "learning_rate": 1e-05, "loss": 0.5041, "mean_token_accuracy": 0.8593335151672363, "num_tokens": 190530326.0, "step": 5017 }, { "epoch": 0.6383411779671797, "ewc_loss": 0.09814453125, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 6.294607162475586, "learning_rate": 1e-05, "loss": 0.5274, "mean_token_accuracy": 0.8564492464065552, "num_tokens": 190565727.0, "step": 5018 }, { "epoch": 0.6384683882457702, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 7.009506225585938e-05, "grad_norm": 5.0011701583862305, "learning_rate": 1e-05, "loss": 0.5444, "mean_token_accuracy": 0.8465924263000488, "num_tokens": 190600163.0, "step": 5019 }, { "epoch": 0.6385955985243608, "ewc_loss": 0.10888671875, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.544154167175293, "learning_rate": 1e-05, "loss": 0.567, "mean_token_accuracy": 0.8476263880729675, "num_tokens": 190640801.0, "step": 5020 }, { "epoch": 0.6387228088029513, "ewc_loss": 0.08740234375, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 7.62939453125e-05, "grad_norm": 5.103768825531006, "learning_rate": 1e-05, "loss": 0.4435, "mean_token_accuracy": 0.8761647343635559, "num_tokens": 190679886.0, "step": 5021 }, { "epoch": 0.6388500190815418, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.841825008392334, "learning_rate": 1e-05, "loss": 0.5574, "mean_token_accuracy": 0.8566139936447144, "num_tokens": 190726246.0, "step": 5022 }, { "epoch": 0.6389772293601323, "ewc_loss": 0.09521484375, "ewc_loss_diag": 0.00011157989501953125, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.519033908843994, "learning_rate": 1e-05, "loss": 0.5195, "mean_token_accuracy": 0.8569493293762207, "num_tokens": 190765496.0, "step": 5023 }, { "epoch": 0.6391044396387228, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.00011157989501953125, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.7654595375061035, "learning_rate": 1e-05, "loss": 0.5623, "mean_token_accuracy": 0.8531206846237183, "num_tokens": 190808770.0, "step": 5024 }, { "epoch": 0.6392316499173133, "ewc_loss": 0.10498046875, "ewc_loss_diag": 0.00011157989501953125, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 5.840510845184326, "learning_rate": 1e-05, "loss": 0.497, "mean_token_accuracy": 0.8663239479064941, "num_tokens": 190844242.0, "step": 5025 }, { "epoch": 0.6393588601959038, "ewc_loss": 0.10595703125, "ewc_loss_diag": 0.00011157989501953125, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.3962883949279785, "learning_rate": 1e-05, "loss": 0.5942, "mean_token_accuracy": 0.8380321264266968, "num_tokens": 190887658.0, "step": 5026 }, { "epoch": 0.6394860704744944, "ewc_loss": 0.10302734375, "ewc_loss_diag": 0.00011157989501953125, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 5.870851516723633, "learning_rate": 1e-05, "loss": 0.5201, "mean_token_accuracy": 0.8615690469741821, "num_tokens": 190922322.0, "step": 5027 }, { "epoch": 0.6396132807530849, "ewc_loss": 0.09619140625, "ewc_loss_diag": 0.00011157989501953125, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 6.012090682983398, "learning_rate": 1e-05, "loss": 0.5298, "mean_token_accuracy": 0.8516026735305786, "num_tokens": 190959085.0, "step": 5028 }, { "epoch": 0.6397404910316754, "ewc_loss": 0.09619140625, "ewc_loss_diag": 0.00011157989501953125, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.804416179656982, "learning_rate": 1e-05, "loss": 0.4918, "mean_token_accuracy": 0.871191143989563, "num_tokens": 190991646.0, "step": 5029 }, { "epoch": 0.6398677013102658, "ewc_loss": 0.0908203125, "ewc_loss_diag": 0.00011157989501953125, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.682814121246338, "learning_rate": 1e-05, "loss": 0.5171, "mean_token_accuracy": 0.8575723171234131, "num_tokens": 191039489.0, "step": 5030 }, { "epoch": 0.6399949115888564, "ewc_loss": 0.0947265625, "ewc_loss_diag": 0.00011157989501953125, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.755950927734375, "learning_rate": 1e-05, "loss": 0.5337, "mean_token_accuracy": 0.8527687191963196, "num_tokens": 191078166.0, "step": 5031 }, { "epoch": 0.6401221218674469, "ewc_loss": 0.08740234375, "ewc_loss_diag": 0.00011157989501953125, "ewc_loss_parallel": 7.62939453125e-05, "grad_norm": 5.477467060089111, "learning_rate": 1e-05, "loss": 0.5726, "mean_token_accuracy": 0.8362870216369629, "num_tokens": 191115981.0, "step": 5032 }, { "epoch": 0.6402493321460374, "ewc_loss": 0.09228515625, "ewc_loss_diag": 0.00011157989501953125, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.729184627532959, "learning_rate": 1e-05, "loss": 0.5158, "mean_token_accuracy": 0.8580901622772217, "num_tokens": 191153989.0, "step": 5033 }, { "epoch": 0.6403765424246279, "ewc_loss": 0.0849609375, "ewc_loss_diag": 0.00011157989501953125, "ewc_loss_parallel": 7.390975952148438e-05, "grad_norm": 5.3319315910339355, "learning_rate": 1e-05, "loss": 0.5102, "mean_token_accuracy": 0.858102560043335, "num_tokens": 191192799.0, "step": 5034 }, { "epoch": 0.6405037527032185, "ewc_loss": 0.09130859375, "ewc_loss_diag": 0.00011157989501953125, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.7584228515625, "learning_rate": 1e-05, "loss": 0.5178, "mean_token_accuracy": 0.856450080871582, "num_tokens": 191231923.0, "step": 5035 }, { "epoch": 0.6406309629818089, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.00011157989501953125, "ewc_loss_parallel": 7.295608520507812e-05, "grad_norm": 5.207467079162598, "learning_rate": 1e-05, "loss": 0.5128, "mean_token_accuracy": 0.8564389944076538, "num_tokens": 191271650.0, "step": 5036 }, { "epoch": 0.6407581732603994, "ewc_loss": 0.09423828125, "ewc_loss_diag": 0.00011157989501953125, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.8681321144104, "learning_rate": 1e-05, "loss": 0.5068, "mean_token_accuracy": 0.8611692786216736, "num_tokens": 191300212.0, "step": 5037 }, { "epoch": 0.6408853835389899, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.00011157989501953125, "ewc_loss_parallel": 7.104873657226562e-05, "grad_norm": 5.080587863922119, "learning_rate": 1e-05, "loss": 0.5152, "mean_token_accuracy": 0.8525694012641907, "num_tokens": 191336163.0, "step": 5038 }, { "epoch": 0.6410125938175805, "ewc_loss": 0.10009765625, "ewc_loss_diag": 0.00011157989501953125, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.087690353393555, "learning_rate": 1e-05, "loss": 0.5313, "mean_token_accuracy": 0.8560925126075745, "num_tokens": 191374749.0, "step": 5039 }, { "epoch": 0.641139804096171, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.00011157989501953125, "ewc_loss_parallel": 7.104873657226562e-05, "grad_norm": 4.965529441833496, "learning_rate": 1e-05, "loss": 0.5027, "mean_token_accuracy": 0.8595556020736694, "num_tokens": 191413342.0, "step": 5040 }, { "epoch": 0.6412670143747615, "ewc_loss": 0.10595703125, "ewc_loss_diag": 0.00011157989501953125, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.446888446807861, "learning_rate": 1e-05, "loss": 0.5121, "mean_token_accuracy": 0.8631401062011719, "num_tokens": 191450332.0, "step": 5041 }, { "epoch": 0.6413942246533519, "ewc_loss": 0.08447265625, "ewc_loss_diag": 0.00011157989501953125, "ewc_loss_parallel": 7.343292236328125e-05, "grad_norm": 5.005486488342285, "learning_rate": 1e-05, "loss": 0.4949, "mean_token_accuracy": 0.8603992462158203, "num_tokens": 191486969.0, "step": 5042 }, { "epoch": 0.6415214349319425, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.00011157989501953125, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.843897819519043, "learning_rate": 1e-05, "loss": 0.5304, "mean_token_accuracy": 0.8615472316741943, "num_tokens": 191525074.0, "step": 5043 }, { "epoch": 0.641648645210533, "ewc_loss": 0.09326171875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.381468296051025, "learning_rate": 1e-05, "loss": 0.5176, "mean_token_accuracy": 0.8553828001022339, "num_tokens": 191564324.0, "step": 5044 }, { "epoch": 0.6417758554891235, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.917022228240967, "learning_rate": 1e-05, "loss": 0.5682, "mean_token_accuracy": 0.8502600789070129, "num_tokens": 191604081.0, "step": 5045 }, { "epoch": 0.641903065767714, "ewc_loss": 0.10498046875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 5.849030017852783, "learning_rate": 1e-05, "loss": 0.5607, "mean_token_accuracy": 0.8430130481719971, "num_tokens": 191639811.0, "step": 5046 }, { "epoch": 0.6420302760463046, "ewc_loss": 0.10888671875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.53793478012085, "learning_rate": 1e-05, "loss": 0.5106, "mean_token_accuracy": 0.8661655187606812, "num_tokens": 191680941.0, "step": 5047 }, { "epoch": 0.642157486324895, "ewc_loss": 0.10498046875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 5.935172080993652, "learning_rate": 1e-05, "loss": 0.5463, "mean_token_accuracy": 0.8501012921333313, "num_tokens": 191718039.0, "step": 5048 }, { "epoch": 0.6422846966034855, "ewc_loss": 0.09814453125, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 6.109110355377197, "learning_rate": 1e-05, "loss": 0.5275, "mean_token_accuracy": 0.8571053743362427, "num_tokens": 191759837.0, "step": 5049 }, { "epoch": 0.6424119068820761, "ewc_loss": 0.09716796875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.855056285858154, "learning_rate": 1e-05, "loss": 0.5354, "mean_token_accuracy": 0.8523221015930176, "num_tokens": 191801791.0, "step": 5050 }, { "epoch": 0.6425391171606666, "ewc_loss": 0.08984375, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.755276203155518, "learning_rate": 1e-05, "loss": 0.4849, "mean_token_accuracy": 0.8675265312194824, "num_tokens": 191838193.0, "step": 5051 }, { "epoch": 0.6426663274392571, "ewc_loss": 0.09521484375, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.805264949798584, "learning_rate": 1e-05, "loss": 0.6101, "mean_token_accuracy": 0.8307923674583435, "num_tokens": 191874811.0, "step": 5052 }, { "epoch": 0.6427935377178476, "ewc_loss": 0.08642578125, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 7.534027099609375e-05, "grad_norm": 5.5003743171691895, "learning_rate": 1e-05, "loss": 0.535, "mean_token_accuracy": 0.8513514995574951, "num_tokens": 191921289.0, "step": 5053 }, { "epoch": 0.6429207479964381, "ewc_loss": 0.09423828125, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.780089378356934, "learning_rate": 1e-05, "loss": 0.5211, "mean_token_accuracy": 0.8591325283050537, "num_tokens": 191961796.0, "step": 5054 }, { "epoch": 0.6430479582750286, "ewc_loss": 0.08544921875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 7.43865966796875e-05, "grad_norm": 5.334671974182129, "learning_rate": 1e-05, "loss": 0.4455, "mean_token_accuracy": 0.8758478760719299, "num_tokens": 192000853.0, "step": 5055 }, { "epoch": 0.6431751685536191, "ewc_loss": 0.09326171875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.828587055206299, "learning_rate": 1e-05, "loss": 0.5119, "mean_token_accuracy": 0.8608920574188232, "num_tokens": 192029126.0, "step": 5056 }, { "epoch": 0.6433023788322096, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 7.152557373046875e-05, "grad_norm": 5.2001872062683105, "learning_rate": 1e-05, "loss": 0.5285, "mean_token_accuracy": 0.8498678803443909, "num_tokens": 192070472.0, "step": 5057 }, { "epoch": 0.6434295891108002, "ewc_loss": 0.09619140625, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.955781936645508, "learning_rate": 1e-05, "loss": 0.5642, "mean_token_accuracy": 0.845750629901886, "num_tokens": 192107844.0, "step": 5058 }, { "epoch": 0.6435567993893907, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 7.009506225585938e-05, "grad_norm": 5.072795391082764, "learning_rate": 1e-05, "loss": 0.4927, "mean_token_accuracy": 0.8580597639083862, "num_tokens": 192140637.0, "step": 5059 }, { "epoch": 0.6436840096679812, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 6.209177017211914, "learning_rate": 1e-05, "loss": 0.5086, "mean_token_accuracy": 0.8654678463935852, "num_tokens": 192176177.0, "step": 5060 }, { "epoch": 0.6438112199465716, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 7.009506225585938e-05, "grad_norm": 4.98900032043457, "learning_rate": 1e-05, "loss": 0.5016, "mean_token_accuracy": 0.8619786500930786, "num_tokens": 192218131.0, "step": 5061 }, { "epoch": 0.6439384302251622, "ewc_loss": 0.10791015625, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.567508697509766, "learning_rate": 1e-05, "loss": 0.5169, "mean_token_accuracy": 0.8610847592353821, "num_tokens": 192258720.0, "step": 5062 }, { "epoch": 0.6440656405037527, "ewc_loss": 0.087890625, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.127992630004883, "learning_rate": 1e-05, "loss": 0.5583, "mean_token_accuracy": 0.8466211557388306, "num_tokens": 192303197.0, "step": 5063 }, { "epoch": 0.6441928507823432, "ewc_loss": 0.11474609375, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.8494648933410645, "learning_rate": 1e-05, "loss": 0.6207, "mean_token_accuracy": 0.8341204524040222, "num_tokens": 192340882.0, "step": 5064 }, { "epoch": 0.6443200610609338, "ewc_loss": 0.09619140625, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.535595417022705, "learning_rate": 1e-05, "loss": 0.5095, "mean_token_accuracy": 0.8616669774055481, "num_tokens": 192383596.0, "step": 5065 }, { "epoch": 0.6444472713395243, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.747657775878906, "learning_rate": 1e-05, "loss": 0.5431, "mean_token_accuracy": 0.8569324016571045, "num_tokens": 192423861.0, "step": 5066 }, { "epoch": 0.6445744816181147, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 5.851546764373779, "learning_rate": 1e-05, "loss": 0.4912, "mean_token_accuracy": 0.864171028137207, "num_tokens": 192456564.0, "step": 5067 }, { "epoch": 0.6447016918967052, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.348801136016846, "learning_rate": 1e-05, "loss": 0.5497, "mean_token_accuracy": 0.8531480431556702, "num_tokens": 192496917.0, "step": 5068 }, { "epoch": 0.6448289021752958, "ewc_loss": 0.099609375, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 5.860311985015869, "learning_rate": 1e-05, "loss": 0.5284, "mean_token_accuracy": 0.8553483486175537, "num_tokens": 192539548.0, "step": 5069 }, { "epoch": 0.6449561124538863, "ewc_loss": 0.09375, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.950588703155518, "learning_rate": 1e-05, "loss": 0.567, "mean_token_accuracy": 0.8462823629379272, "num_tokens": 192577430.0, "step": 5070 }, { "epoch": 0.6450833227324768, "ewc_loss": 0.09619140625, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.809933185577393, "learning_rate": 1e-05, "loss": 0.4814, "mean_token_accuracy": 0.8682758212089539, "num_tokens": 192617343.0, "step": 5071 }, { "epoch": 0.6452105330110673, "ewc_loss": 0.08837890625, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.622879505157471, "learning_rate": 1e-05, "loss": 0.5114, "mean_token_accuracy": 0.8599963188171387, "num_tokens": 192662274.0, "step": 5072 }, { "epoch": 0.6453377432896578, "ewc_loss": 0.09326171875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.780538558959961, "learning_rate": 1e-05, "loss": 0.5065, "mean_token_accuracy": 0.8644084930419922, "num_tokens": 192703305.0, "step": 5073 }, { "epoch": 0.6454649535682483, "ewc_loss": 0.0849609375, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 7.390975952148438e-05, "grad_norm": 5.423785209655762, "learning_rate": 1e-05, "loss": 0.5277, "mean_token_accuracy": 0.8543997406959534, "num_tokens": 192738325.0, "step": 5074 }, { "epoch": 0.6455921638468388, "ewc_loss": 0.09228515625, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.778387546539307, "learning_rate": 1e-05, "loss": 0.4816, "mean_token_accuracy": 0.8718671202659607, "num_tokens": 192778285.0, "step": 5075 }, { "epoch": 0.6457193741254293, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 7.152557373046875e-05, "grad_norm": 5.281181812286377, "learning_rate": 1e-05, "loss": 0.5, "mean_token_accuracy": 0.8615195751190186, "num_tokens": 192822822.0, "step": 5076 }, { "epoch": 0.6458465844040199, "ewc_loss": 0.09326171875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.855865001678467, "learning_rate": 1e-05, "loss": 0.5432, "mean_token_accuracy": 0.8535505533218384, "num_tokens": 192857312.0, "step": 5077 }, { "epoch": 0.6459737946826104, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 7.05718994140625e-05, "grad_norm": 5.16152811050415, "learning_rate": 1e-05, "loss": 0.4769, "mean_token_accuracy": 0.8634340763092041, "num_tokens": 192892363.0, "step": 5078 }, { "epoch": 0.6461010049612008, "ewc_loss": 0.09716796875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 6.001922130584717, "learning_rate": 1e-05, "loss": 0.5462, "mean_token_accuracy": 0.8483101725578308, "num_tokens": 192931774.0, "step": 5079 }, { "epoch": 0.6462282152397913, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 7.009506225585938e-05, "grad_norm": 5.04744815826416, "learning_rate": 1e-05, "loss": 0.4831, "mean_token_accuracy": 0.8649156093597412, "num_tokens": 192964463.0, "step": 5080 }, { "epoch": 0.6463554255183819, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.266970157623291, "learning_rate": 1e-05, "loss": 0.4895, "mean_token_accuracy": 0.8689409494400024, "num_tokens": 192999807.0, "step": 5081 }, { "epoch": 0.6464826357969724, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 7.200241088867188e-05, "grad_norm": 4.998575210571289, "learning_rate": 1e-05, "loss": 0.5225, "mean_token_accuracy": 0.8525422811508179, "num_tokens": 193041408.0, "step": 5082 }, { "epoch": 0.6466098460755629, "ewc_loss": 0.10888671875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.623053073883057, "learning_rate": 1e-05, "loss": 0.6202, "mean_token_accuracy": 0.8322439193725586, "num_tokens": 193078104.0, "step": 5083 }, { "epoch": 0.6467370563541535, "ewc_loss": 0.08837890625, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.196878910064697, "learning_rate": 1e-05, "loss": 0.5649, "mean_token_accuracy": 0.8413999080657959, "num_tokens": 193118821.0, "step": 5084 }, { "epoch": 0.6468642666327439, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.834519863128662, "learning_rate": 1e-05, "loss": 0.5613, "mean_token_accuracy": 0.8532410860061646, "num_tokens": 193152633.0, "step": 5085 }, { "epoch": 0.6469914769113344, "ewc_loss": 0.09716796875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.597322940826416, "learning_rate": 1e-05, "loss": 0.5096, "mean_token_accuracy": 0.8606228828430176, "num_tokens": 193191398.0, "step": 5086 }, { "epoch": 0.6471186871899249, "ewc_loss": 0.11083984375, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.655829429626465, "learning_rate": 1e-05, "loss": 0.5667, "mean_token_accuracy": 0.8495581150054932, "num_tokens": 193229612.0, "step": 5087 }, { "epoch": 0.6472458974685155, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 5.834958553314209, "learning_rate": 1e-05, "loss": 0.5661, "mean_token_accuracy": 0.8432743549346924, "num_tokens": 193271993.0, "step": 5088 }, { "epoch": 0.647373107747106, "ewc_loss": 0.10302734375, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.265572547912598, "learning_rate": 1e-05, "loss": 0.5661, "mean_token_accuracy": 0.8476648926734924, "num_tokens": 193308146.0, "step": 5089 }, { "epoch": 0.6475003180256965, "ewc_loss": 0.099609375, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 5.829598426818848, "learning_rate": 1e-05, "loss": 0.5048, "mean_token_accuracy": 0.8632002472877502, "num_tokens": 193348448.0, "step": 5090 }, { "epoch": 0.6476275283042869, "ewc_loss": 0.091796875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.881192684173584, "learning_rate": 1e-05, "loss": 0.5137, "mean_token_accuracy": 0.857763409614563, "num_tokens": 193379279.0, "step": 5091 }, { "epoch": 0.6477547385828775, "ewc_loss": 0.09619140625, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.792790412902832, "learning_rate": 1e-05, "loss": 0.5055, "mean_token_accuracy": 0.8626909255981445, "num_tokens": 193415999.0, "step": 5092 }, { "epoch": 0.647881948861468, "ewc_loss": 0.0869140625, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 7.581710815429688e-05, "grad_norm": 5.5681376457214355, "learning_rate": 1e-05, "loss": 0.51, "mean_token_accuracy": 0.8607500791549683, "num_tokens": 193450287.0, "step": 5093 }, { "epoch": 0.6480091591400585, "ewc_loss": 0.09375, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.77541971206665, "learning_rate": 1e-05, "loss": 0.5109, "mean_token_accuracy": 0.8602343797683716, "num_tokens": 193485980.0, "step": 5094 }, { "epoch": 0.648136369418649, "ewc_loss": 0.0849609375, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 7.390975952148438e-05, "grad_norm": 5.380312919616699, "learning_rate": 1e-05, "loss": 0.5356, "mean_token_accuracy": 0.8526880741119385, "num_tokens": 193525179.0, "step": 5095 }, { "epoch": 0.6482635796972396, "ewc_loss": 0.09228515625, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.783396244049072, "learning_rate": 1e-05, "loss": 0.5239, "mean_token_accuracy": 0.8558257818222046, "num_tokens": 193561580.0, "step": 5096 }, { "epoch": 0.64839078997583, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 7.200241088867188e-05, "grad_norm": 5.258324146270752, "learning_rate": 1e-05, "loss": 0.5548, "mean_token_accuracy": 0.8463871479034424, "num_tokens": 193598218.0, "step": 5097 }, { "epoch": 0.6485180002544205, "ewc_loss": 0.09326171875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.858820915222168, "learning_rate": 1e-05, "loss": 0.5609, "mean_token_accuracy": 0.8460861444473267, "num_tokens": 193635256.0, "step": 5098 }, { "epoch": 0.648645210533011, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 7.104873657226562e-05, "grad_norm": 5.123781204223633, "learning_rate": 1e-05, "loss": 0.5254, "mean_token_accuracy": 0.8483415842056274, "num_tokens": 193672957.0, "step": 5099 }, { "epoch": 0.6487724208116016, "ewc_loss": 0.09716796875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 6.037328720092773, "learning_rate": 1e-05, "loss": 0.4967, "mean_token_accuracy": 0.8678101301193237, "num_tokens": 193710575.0, "step": 5100 }, { "epoch": 0.6488996310901921, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 7.009506225585938e-05, "grad_norm": 5.016901969909668, "learning_rate": 1e-05, "loss": 0.4975, "mean_token_accuracy": 0.8660826683044434, "num_tokens": 193749006.0, "step": 5101 }, { "epoch": 0.6490268413687826, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.325241565704346, "learning_rate": 1e-05, "loss": 0.5929, "mean_token_accuracy": 0.8416886329650879, "num_tokens": 193788975.0, "step": 5102 }, { "epoch": 0.649154051647373, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 7.200241088867188e-05, "grad_norm": 5.017400741577148, "learning_rate": 1e-05, "loss": 0.5031, "mean_token_accuracy": 0.8569221496582031, "num_tokens": 193822898.0, "step": 5103 }, { "epoch": 0.6492812619259636, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.659026145935059, "learning_rate": 1e-05, "loss": 0.4854, "mean_token_accuracy": 0.8728885650634766, "num_tokens": 193862439.0, "step": 5104 }, { "epoch": 0.6494084722045541, "ewc_loss": 0.08984375, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.250044822692871, "learning_rate": 1e-05, "loss": 0.5334, "mean_token_accuracy": 0.8551405072212219, "num_tokens": 193906721.0, "step": 5105 }, { "epoch": 0.6495356824831446, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.810415744781494, "learning_rate": 1e-05, "loss": 0.5254, "mean_token_accuracy": 0.8605656623840332, "num_tokens": 193940521.0, "step": 5106 }, { "epoch": 0.6496628927617352, "ewc_loss": 0.099609375, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 5.646754264831543, "learning_rate": 1e-05, "loss": 0.544, "mean_token_accuracy": 0.8488801121711731, "num_tokens": 193979340.0, "step": 5107 }, { "epoch": 0.6497901030403257, "ewc_loss": 0.10986328125, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.575150966644287, "learning_rate": 1e-05, "loss": 0.5287, "mean_token_accuracy": 0.8571189641952515, "num_tokens": 194018562.0, "step": 5108 }, { "epoch": 0.6499173133189162, "ewc_loss": 0.10302734375, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 5.839704990386963, "learning_rate": 1e-05, "loss": 0.5988, "mean_token_accuracy": 0.8296340703964233, "num_tokens": 194056804.0, "step": 5109 }, { "epoch": 0.6500445235975066, "ewc_loss": 0.10107421875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.17003870010376, "learning_rate": 1e-05, "loss": 0.5187, "mean_token_accuracy": 0.8582465052604675, "num_tokens": 194095194.0, "step": 5110 }, { "epoch": 0.6501717338760972, "ewc_loss": 0.09814453125, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.816410064697266, "learning_rate": 1e-05, "loss": 0.5345, "mean_token_accuracy": 0.8558658361434937, "num_tokens": 194135061.0, "step": 5111 }, { "epoch": 0.6502989441546877, "ewc_loss": 0.09228515625, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.849149703979492, "learning_rate": 1e-05, "loss": 0.5069, "mean_token_accuracy": 0.8601641654968262, "num_tokens": 194169017.0, "step": 5112 }, { "epoch": 0.6504261544332782, "ewc_loss": 0.0947265625, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.7691497802734375, "learning_rate": 1e-05, "loss": 0.5441, "mean_token_accuracy": 0.847802460193634, "num_tokens": 194213424.0, "step": 5113 }, { "epoch": 0.6505533647118688, "ewc_loss": 0.0869140625, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 7.581710815429688e-05, "grad_norm": 5.540764808654785, "learning_rate": 1e-05, "loss": 0.4813, "mean_token_accuracy": 0.8678306937217712, "num_tokens": 194247465.0, "step": 5114 }, { "epoch": 0.6506805749904593, "ewc_loss": 0.09423828125, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.761239528656006, "learning_rate": 1e-05, "loss": 0.4878, "mean_token_accuracy": 0.8644461631774902, "num_tokens": 194282575.0, "step": 5115 }, { "epoch": 0.6508077852690497, "ewc_loss": 0.0849609375, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 7.390975952148438e-05, "grad_norm": 5.362581253051758, "learning_rate": 1e-05, "loss": 0.5048, "mean_token_accuracy": 0.858698844909668, "num_tokens": 194311822.0, "step": 5116 }, { "epoch": 0.6509349955476402, "ewc_loss": 0.0927734375, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.784609794616699, "learning_rate": 1e-05, "loss": 0.5423, "mean_token_accuracy": 0.8471617102622986, "num_tokens": 194346990.0, "step": 5117 }, { "epoch": 0.6510622058262308, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 7.200241088867188e-05, "grad_norm": 5.214410781860352, "learning_rate": 1e-05, "loss": 0.4924, "mean_token_accuracy": 0.8581870794296265, "num_tokens": 194384836.0, "step": 5118 }, { "epoch": 0.6511894161048213, "ewc_loss": 0.09375, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.8917999267578125, "learning_rate": 1e-05, "loss": 0.4767, "mean_token_accuracy": 0.8698363900184631, "num_tokens": 194418080.0, "step": 5119 }, { "epoch": 0.6513166263834118, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 7.05718994140625e-05, "grad_norm": 5.096014499664307, "learning_rate": 1e-05, "loss": 0.552, "mean_token_accuracy": 0.8454240560531616, "num_tokens": 194460111.0, "step": 5120 }, { "epoch": 0.6514438366620023, "ewc_loss": 0.09912109375, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 6.102272987365723, "learning_rate": 1e-05, "loss": 0.4845, "mean_token_accuracy": 0.8726675510406494, "num_tokens": 194498157.0, "step": 5121 }, { "epoch": 0.6515710469405928, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 7.009506225585938e-05, "grad_norm": 4.997880458831787, "learning_rate": 1e-05, "loss": 0.5017, "mean_token_accuracy": 0.8597792983055115, "num_tokens": 194538928.0, "step": 5122 }, { "epoch": 0.6516982572191833, "ewc_loss": 0.10595703125, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.428779602050781, "learning_rate": 1e-05, "loss": 0.5253, "mean_token_accuracy": 0.8597375750541687, "num_tokens": 194582276.0, "step": 5123 }, { "epoch": 0.6518254674977738, "ewc_loss": 0.0849609375, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 7.390975952148438e-05, "grad_norm": 5.040722846984863, "learning_rate": 1e-05, "loss": 0.5047, "mean_token_accuracy": 0.857244610786438, "num_tokens": 194620197.0, "step": 5124 }, { "epoch": 0.6519526777763643, "ewc_loss": 0.11376953125, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.768718719482422, "learning_rate": 1e-05, "loss": 0.511, "mean_token_accuracy": 0.8651180267333984, "num_tokens": 194654634.0, "step": 5125 }, { "epoch": 0.6520798880549549, "ewc_loss": 0.091796875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.364181041717529, "learning_rate": 1e-05, "loss": 0.5372, "mean_token_accuracy": 0.8515238761901855, "num_tokens": 194696400.0, "step": 5126 }, { "epoch": 0.6522070983335454, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.817824840545654, "learning_rate": 1e-05, "loss": 0.521, "mean_token_accuracy": 0.864543080329895, "num_tokens": 194738326.0, "step": 5127 }, { "epoch": 0.6523343086121358, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 5.7462944984436035, "learning_rate": 1e-05, "loss": 0.5165, "mean_token_accuracy": 0.8606613278388977, "num_tokens": 194778550.0, "step": 5128 }, { "epoch": 0.6524615188907263, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.508308410644531, "learning_rate": 1e-05, "loss": 0.5464, "mean_token_accuracy": 0.8544524908065796, "num_tokens": 194815124.0, "step": 5129 }, { "epoch": 0.6525887291693169, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 5.871294021606445, "learning_rate": 1e-05, "loss": 0.5277, "mean_token_accuracy": 0.8580528497695923, "num_tokens": 194851372.0, "step": 5130 }, { "epoch": 0.6527159394479074, "ewc_loss": 0.099609375, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 6.098402976989746, "learning_rate": 1e-05, "loss": 0.4881, "mean_token_accuracy": 0.8704255819320679, "num_tokens": 194886705.0, "step": 5131 }, { "epoch": 0.6528431497264979, "ewc_loss": 0.09765625, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.8208537101745605, "learning_rate": 1e-05, "loss": 0.5135, "mean_token_accuracy": 0.8584296703338623, "num_tokens": 194930372.0, "step": 5132 }, { "epoch": 0.6529703600050885, "ewc_loss": 0.0908203125, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.732779026031494, "learning_rate": 1e-05, "loss": 0.501, "mean_token_accuracy": 0.8606839179992676, "num_tokens": 194971869.0, "step": 5133 }, { "epoch": 0.6530975702836789, "ewc_loss": 0.095703125, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.794041156768799, "learning_rate": 1e-05, "loss": 0.5411, "mean_token_accuracy": 0.8510441780090332, "num_tokens": 195011271.0, "step": 5134 }, { "epoch": 0.6532247805622694, "ewc_loss": 0.0869140625, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 7.581710815429688e-05, "grad_norm": 5.482576370239258, "learning_rate": 1e-05, "loss": 0.5013, "mean_token_accuracy": 0.8598843812942505, "num_tokens": 195046328.0, "step": 5135 }, { "epoch": 0.6533519908408599, "ewc_loss": 0.09423828125, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.790044784545898, "learning_rate": 1e-05, "loss": 0.4976, "mean_token_accuracy": 0.8635959625244141, "num_tokens": 195086424.0, "step": 5136 }, { "epoch": 0.6534792011194505, "ewc_loss": 0.0849609375, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 7.390975952148438e-05, "grad_norm": 5.31854248046875, "learning_rate": 1e-05, "loss": 0.5389, "mean_token_accuracy": 0.8508123159408569, "num_tokens": 195121751.0, "step": 5137 }, { "epoch": 0.653606411398041, "ewc_loss": 0.09326171875, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.843888282775879, "learning_rate": 1e-05, "loss": 0.5138, "mean_token_accuracy": 0.8604769706726074, "num_tokens": 195158231.0, "step": 5138 }, { "epoch": 0.6537336216766315, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 7.200241088867188e-05, "grad_norm": 5.181367874145508, "learning_rate": 1e-05, "loss": 0.5123, "mean_token_accuracy": 0.8573852181434631, "num_tokens": 195195009.0, "step": 5139 }, { "epoch": 0.6538608319552219, "ewc_loss": 0.0966796875, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.964175224304199, "learning_rate": 1e-05, "loss": 0.4859, "mean_token_accuracy": 0.8681090474128723, "num_tokens": 195232892.0, "step": 5140 }, { "epoch": 0.6539880422338125, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 7.05718994140625e-05, "grad_norm": 5.0599045753479, "learning_rate": 1e-05, "loss": 0.5109, "mean_token_accuracy": 0.8572502136230469, "num_tokens": 195283233.0, "step": 5141 }, { "epoch": 0.654115252512403, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.206414699554443, "learning_rate": 1e-05, "loss": 0.4614, "mean_token_accuracy": 0.8763225078582764, "num_tokens": 195319402.0, "step": 5142 }, { "epoch": 0.6542424627909935, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 7.152557373046875e-05, "grad_norm": 5.0036797523498535, "learning_rate": 1e-05, "loss": 0.4838, "mean_token_accuracy": 0.8634147644042969, "num_tokens": 195358330.0, "step": 5143 }, { "epoch": 0.654369673069584, "ewc_loss": 0.10791015625, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.550745487213135, "learning_rate": 1e-05, "loss": 0.482, "mean_token_accuracy": 0.8727525472640991, "num_tokens": 195398074.0, "step": 5144 }, { "epoch": 0.6544968833481746, "ewc_loss": 0.08837890625, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.134967803955078, "learning_rate": 1e-05, "loss": 0.4817, "mean_token_accuracy": 0.8720648288726807, "num_tokens": 195435684.0, "step": 5145 }, { "epoch": 0.654624093626765, "ewc_loss": 0.11474609375, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.80202054977417, "learning_rate": 1e-05, "loss": 0.5019, "mean_token_accuracy": 0.8699941039085388, "num_tokens": 195475126.0, "step": 5146 }, { "epoch": 0.6547513039053555, "ewc_loss": 0.0966796875, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.507664203643799, "learning_rate": 1e-05, "loss": 0.5413, "mean_token_accuracy": 0.8540290594100952, "num_tokens": 195516398.0, "step": 5147 }, { "epoch": 0.654878514183946, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.71074914932251, "learning_rate": 1e-05, "loss": 0.5252, "mean_token_accuracy": 0.8619786500930786, "num_tokens": 195558777.0, "step": 5148 }, { "epoch": 0.6550057244625366, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 5.794032573699951, "learning_rate": 1e-05, "loss": 0.542, "mean_token_accuracy": 0.8528711795806885, "num_tokens": 195598663.0, "step": 5149 }, { "epoch": 0.6551329347411271, "ewc_loss": 0.10302734375, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.372610569000244, "learning_rate": 1e-05, "loss": 0.5552, "mean_token_accuracy": 0.8433687090873718, "num_tokens": 195628594.0, "step": 5150 }, { "epoch": 0.6552601450197176, "ewc_loss": 0.099609375, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 5.843906402587891, "learning_rate": 1e-05, "loss": 0.5466, "mean_token_accuracy": 0.8505732417106628, "num_tokens": 195662484.0, "step": 5151 }, { "epoch": 0.655387355298308, "ewc_loss": 0.0947265625, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.962457656860352, "learning_rate": 1e-05, "loss": 0.5482, "mean_token_accuracy": 0.8488953113555908, "num_tokens": 195697753.0, "step": 5152 }, { "epoch": 0.6555145655768986, "ewc_loss": 0.095703125, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.792306900024414, "learning_rate": 1e-05, "loss": 0.4881, "mean_token_accuracy": 0.8686583638191223, "num_tokens": 195740931.0, "step": 5153 }, { "epoch": 0.6556417758554891, "ewc_loss": 0.08984375, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.648823261260986, "learning_rate": 1e-05, "loss": 0.5601, "mean_token_accuracy": 0.8442115187644958, "num_tokens": 195783056.0, "step": 5154 }, { "epoch": 0.6557689861340796, "ewc_loss": 0.09375, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.763679027557373, "learning_rate": 1e-05, "loss": 0.5612, "mean_token_accuracy": 0.8466932773590088, "num_tokens": 195819515.0, "step": 5155 }, { "epoch": 0.6558961964126702, "ewc_loss": 0.08544921875, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 7.43865966796875e-05, "grad_norm": 5.442079067230225, "learning_rate": 1e-05, "loss": 0.56, "mean_token_accuracy": 0.8435156345367432, "num_tokens": 195861039.0, "step": 5156 }, { "epoch": 0.6560234066912607, "ewc_loss": 0.0927734375, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.761849880218506, "learning_rate": 1e-05, "loss": 0.5683, "mean_token_accuracy": 0.8424909710884094, "num_tokens": 195897818.0, "step": 5157 }, { "epoch": 0.6561506169698512, "ewc_loss": 0.08349609375, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 7.2479248046875e-05, "grad_norm": 5.295876502990723, "learning_rate": 1e-05, "loss": 0.4471, "mean_token_accuracy": 0.8731863498687744, "num_tokens": 195930752.0, "step": 5158 }, { "epoch": 0.6562778272484416, "ewc_loss": 0.09228515625, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.824255466461182, "learning_rate": 1e-05, "loss": 0.53, "mean_token_accuracy": 0.8526875376701355, "num_tokens": 195965036.0, "step": 5159 }, { "epoch": 0.6564050375270322, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 7.152557373046875e-05, "grad_norm": 5.175891399383545, "learning_rate": 1e-05, "loss": 0.5286, "mean_token_accuracy": 0.8563793897628784, "num_tokens": 195999293.0, "step": 5160 }, { "epoch": 0.6565322478056227, "ewc_loss": 0.09619140625, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.957993507385254, "learning_rate": 1e-05, "loss": 0.5497, "mean_token_accuracy": 0.8489699363708496, "num_tokens": 196039066.0, "step": 5161 }, { "epoch": 0.6566594580842132, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 7.009506225585938e-05, "grad_norm": 5.053414344787598, "learning_rate": 1e-05, "loss": 0.5362, "mean_token_accuracy": 0.8499084115028381, "num_tokens": 196077805.0, "step": 5162 }, { "epoch": 0.6567866683628037, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.2211689949035645, "learning_rate": 1e-05, "loss": 0.5933, "mean_token_accuracy": 0.8353815078735352, "num_tokens": 196117460.0, "step": 5163 }, { "epoch": 0.6569138786413943, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 7.104873657226562e-05, "grad_norm": 4.9858832359313965, "learning_rate": 1e-05, "loss": 0.4986, "mean_token_accuracy": 0.8594149351119995, "num_tokens": 196154855.0, "step": 5164 }, { "epoch": 0.6570410889199847, "ewc_loss": 0.10888671875, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.58301305770874, "learning_rate": 1e-05, "loss": 0.5859, "mean_token_accuracy": 0.8417683839797974, "num_tokens": 196197360.0, "step": 5165 }, { "epoch": 0.6571682991985752, "ewc_loss": 0.08837890625, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.127095699310303, "learning_rate": 1e-05, "loss": 0.5865, "mean_token_accuracy": 0.8360934257507324, "num_tokens": 196237419.0, "step": 5166 }, { "epoch": 0.6572955094771658, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.859889507293701, "learning_rate": 1e-05, "loss": 0.5728, "mean_token_accuracy": 0.8470046520233154, "num_tokens": 196275533.0, "step": 5167 }, { "epoch": 0.6574227197557563, "ewc_loss": 0.09814453125, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.537517070770264, "learning_rate": 1e-05, "loss": 0.5133, "mean_token_accuracy": 0.8635985851287842, "num_tokens": 196312269.0, "step": 5168 }, { "epoch": 0.6575499300343468, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.751309394836426, "learning_rate": 1e-05, "loss": 0.5522, "mean_token_accuracy": 0.8532208800315857, "num_tokens": 196349338.0, "step": 5169 }, { "epoch": 0.6576771403129373, "ewc_loss": 0.10400390625, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 5.83780574798584, "learning_rate": 1e-05, "loss": 0.5781, "mean_token_accuracy": 0.8449393510818481, "num_tokens": 196396000.0, "step": 5170 }, { "epoch": 0.6578043505915278, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.377567291259766, "learning_rate": 1e-05, "loss": 0.5464, "mean_token_accuracy": 0.8501056432723999, "num_tokens": 196428119.0, "step": 5171 }, { "epoch": 0.6579315608701183, "ewc_loss": 0.10205078125, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 5.90660285949707, "learning_rate": 1e-05, "loss": 0.5772, "mean_token_accuracy": 0.8435741066932678, "num_tokens": 196470466.0, "step": 5172 }, { "epoch": 0.6580587711487088, "ewc_loss": 0.09375, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.948024272918701, "learning_rate": 1e-05, "loss": 0.5207, "mean_token_accuracy": 0.8580455780029297, "num_tokens": 196508883.0, "step": 5173 }, { "epoch": 0.6581859814272993, "ewc_loss": 0.09716796875, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.819522380828857, "learning_rate": 1e-05, "loss": 0.5522, "mean_token_accuracy": 0.8455920815467834, "num_tokens": 196552762.0, "step": 5174 }, { "epoch": 0.6583131917058899, "ewc_loss": 0.08837890625, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.637543201446533, "learning_rate": 1e-05, "loss": 0.4481, "mean_token_accuracy": 0.8775955438613892, "num_tokens": 196581394.0, "step": 5175 }, { "epoch": 0.6584404019844804, "ewc_loss": 0.09423828125, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.786356449127197, "learning_rate": 1e-05, "loss": 0.4964, "mean_token_accuracy": 0.8613762855529785, "num_tokens": 196623237.0, "step": 5176 }, { "epoch": 0.6585676122630708, "ewc_loss": 0.08642578125, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 7.534027099609375e-05, "grad_norm": 5.438729763031006, "learning_rate": 1e-05, "loss": 0.464, "mean_token_accuracy": 0.8694919943809509, "num_tokens": 196659219.0, "step": 5177 }, { "epoch": 0.6586948225416613, "ewc_loss": 0.09326171875, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.788300037384033, "learning_rate": 1e-05, "loss": 0.535, "mean_token_accuracy": 0.8521376252174377, "num_tokens": 196700102.0, "step": 5178 }, { "epoch": 0.6588220328202519, "ewc_loss": 0.08447265625, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 7.343292236328125e-05, "grad_norm": 5.294315814971924, "learning_rate": 1e-05, "loss": 0.4255, "mean_token_accuracy": 0.8843652009963989, "num_tokens": 196740569.0, "step": 5179 }, { "epoch": 0.6589492430988424, "ewc_loss": 0.09375, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.850712776184082, "learning_rate": 1e-05, "loss": 0.5365, "mean_token_accuracy": 0.8525324463844299, "num_tokens": 196778634.0, "step": 5180 }, { "epoch": 0.6590764533774329, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 7.200241088867188e-05, "grad_norm": 5.196472644805908, "learning_rate": 1e-05, "loss": 0.5406, "mean_token_accuracy": 0.8467743992805481, "num_tokens": 196805021.0, "step": 5181 }, { "epoch": 0.6592036636560235, "ewc_loss": 0.09619140625, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.97252082824707, "learning_rate": 1e-05, "loss": 0.5165, "mean_token_accuracy": 0.860135555267334, "num_tokens": 196839676.0, "step": 5182 }, { "epoch": 0.6593308739346139, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 7.104873657226562e-05, "grad_norm": 5.075005054473877, "learning_rate": 1e-05, "loss": 0.4841, "mean_token_accuracy": 0.8654572367668152, "num_tokens": 196877564.0, "step": 5183 }, { "epoch": 0.6594580842132044, "ewc_loss": 0.10205078125, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.209014892578125, "learning_rate": 1e-05, "loss": 0.5472, "mean_token_accuracy": 0.8509052991867065, "num_tokens": 196913819.0, "step": 5184 }, { "epoch": 0.6595852944917949, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 7.200241088867188e-05, "grad_norm": 5.0019450187683105, "learning_rate": 1e-05, "loss": 0.539, "mean_token_accuracy": 0.8430501222610474, "num_tokens": 196951019.0, "step": 5185 }, { "epoch": 0.6597125047703855, "ewc_loss": 0.10986328125, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.571974277496338, "learning_rate": 1e-05, "loss": 0.5383, "mean_token_accuracy": 0.8565812110900879, "num_tokens": 196995112.0, "step": 5186 }, { "epoch": 0.659839715048976, "ewc_loss": 0.0888671875, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.125805377960205, "learning_rate": 1e-05, "loss": 0.4946, "mean_token_accuracy": 0.8620510101318359, "num_tokens": 197035334.0, "step": 5187 }, { "epoch": 0.6599669253275665, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.870820045471191, "learning_rate": 1e-05, "loss": 0.5753, "mean_token_accuracy": 0.8480674624443054, "num_tokens": 197074234.0, "step": 5188 }, { "epoch": 0.6600941356061569, "ewc_loss": 0.09716796875, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.540112495422363, "learning_rate": 1e-05, "loss": 0.5442, "mean_token_accuracy": 0.8516645431518555, "num_tokens": 197110977.0, "step": 5189 }, { "epoch": 0.6602213458847475, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.766568183898926, "learning_rate": 1e-05, "loss": 0.578, "mean_token_accuracy": 0.8464325070381165, "num_tokens": 197148634.0, "step": 5190 }, { "epoch": 0.660348556163338, "ewc_loss": 0.10400390625, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 5.844961166381836, "learning_rate": 1e-05, "loss": 0.5203, "mean_token_accuracy": 0.8613921403884888, "num_tokens": 197186361.0, "step": 5191 }, { "epoch": 0.6604757664419285, "ewc_loss": 0.10595703125, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.399949073791504, "learning_rate": 1e-05, "loss": 0.5316, "mean_token_accuracy": 0.8561602234840393, "num_tokens": 197222642.0, "step": 5192 }, { "epoch": 0.660602976720519, "ewc_loss": 0.10107421875, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 5.871676445007324, "learning_rate": 1e-05, "loss": 0.4854, "mean_token_accuracy": 0.8731545209884644, "num_tokens": 197263320.0, "step": 5193 }, { "epoch": 0.6607301869991096, "ewc_loss": 0.0947265625, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 6.008153438568115, "learning_rate": 1e-05, "loss": 0.55, "mean_token_accuracy": 0.8455886840820312, "num_tokens": 197294041.0, "step": 5194 }, { "epoch": 0.6608573972777, "ewc_loss": 0.09716796875, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.823840618133545, "learning_rate": 1e-05, "loss": 0.4861, "mean_token_accuracy": 0.8635538816452026, "num_tokens": 197324878.0, "step": 5195 }, { "epoch": 0.6609846075562905, "ewc_loss": 0.09033203125, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.670787811279297, "learning_rate": 1e-05, "loss": 0.4753, "mean_token_accuracy": 0.8675745725631714, "num_tokens": 197368318.0, "step": 5196 }, { "epoch": 0.661111817834881, "ewc_loss": 0.0947265625, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.7796854972839355, "learning_rate": 1e-05, "loss": 0.517, "mean_token_accuracy": 0.8603993654251099, "num_tokens": 197407665.0, "step": 5197 }, { "epoch": 0.6612390281134716, "ewc_loss": 0.0859375, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 7.486343383789062e-05, "grad_norm": 5.4556097984313965, "learning_rate": 1e-05, "loss": 0.5214, "mean_token_accuracy": 0.8573956489562988, "num_tokens": 197445812.0, "step": 5198 }, { "epoch": 0.6613662383920621, "ewc_loss": 0.09326171875, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.782861232757568, "learning_rate": 1e-05, "loss": 0.4912, "mean_token_accuracy": 0.8659344911575317, "num_tokens": 197482733.0, "step": 5199 }, { "epoch": 0.6614934486706526, "ewc_loss": 0.08349609375, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 7.2479248046875e-05, "grad_norm": 5.3047709465026855, "learning_rate": 1e-05, "loss": 0.5029, "mean_token_accuracy": 0.8576445579528809, "num_tokens": 197520874.0, "step": 5200 }, { "epoch": 0.661620658949243, "ewc_loss": 0.0927734375, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.837315082550049, "learning_rate": 1e-05, "loss": 0.5241, "mean_token_accuracy": 0.8545525670051575, "num_tokens": 197563422.0, "step": 5201 }, { "epoch": 0.6617478692278336, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 7.200241088867188e-05, "grad_norm": 5.181589126586914, "learning_rate": 1e-05, "loss": 0.5237, "mean_token_accuracy": 0.8505609035491943, "num_tokens": 197603632.0, "step": 5202 }, { "epoch": 0.6618750795064241, "ewc_loss": 0.09619140625, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.979824542999268, "learning_rate": 1e-05, "loss": 0.5269, "mean_token_accuracy": 0.8596706390380859, "num_tokens": 197642090.0, "step": 5203 }, { "epoch": 0.6620022897850146, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 7.104873657226562e-05, "grad_norm": 5.065379619598389, "learning_rate": 1e-05, "loss": 0.6023, "mean_token_accuracy": 0.8293169736862183, "num_tokens": 197679294.0, "step": 5204 }, { "epoch": 0.6621295000636052, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.241300582885742, "learning_rate": 1e-05, "loss": 0.6046, "mean_token_accuracy": 0.8368297815322876, "num_tokens": 197721254.0, "step": 5205 }, { "epoch": 0.6622567103421957, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 7.152557373046875e-05, "grad_norm": 4.992290496826172, "learning_rate": 1e-05, "loss": 0.4503, "mean_token_accuracy": 0.8750780820846558, "num_tokens": 197764656.0, "step": 5206 }, { "epoch": 0.6623839206207861, "ewc_loss": 0.11083984375, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.619231224060059, "learning_rate": 1e-05, "loss": 0.5396, "mean_token_accuracy": 0.8601353168487549, "num_tokens": 197805900.0, "step": 5207 }, { "epoch": 0.6625111308993766, "ewc_loss": 0.087890625, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.156454563140869, "learning_rate": 1e-05, "loss": 0.498, "mean_token_accuracy": 0.8610656261444092, "num_tokens": 197842752.0, "step": 5208 }, { "epoch": 0.6626383411779672, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.883492946624756, "learning_rate": 1e-05, "loss": 0.5029, "mean_token_accuracy": 0.8674526214599609, "num_tokens": 197883531.0, "step": 5209 }, { "epoch": 0.6627655514565577, "ewc_loss": 0.09814453125, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.5787835121154785, "learning_rate": 1e-05, "loss": 0.5044, "mean_token_accuracy": 0.8618218302726746, "num_tokens": 197920875.0, "step": 5210 }, { "epoch": 0.6628927617351482, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.75056266784668, "learning_rate": 1e-05, "loss": 0.5356, "mean_token_accuracy": 0.8595486879348755, "num_tokens": 197955194.0, "step": 5211 }, { "epoch": 0.6630199720137387, "ewc_loss": 0.10400390625, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 5.867358207702637, "learning_rate": 1e-05, "loss": 0.4921, "mean_token_accuracy": 0.8696694374084473, "num_tokens": 197991730.0, "step": 5212 }, { "epoch": 0.6631471822923293, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.358218193054199, "learning_rate": 1e-05, "loss": 0.51, "mean_token_accuracy": 0.86344975233078, "num_tokens": 198030032.0, "step": 5213 }, { "epoch": 0.6632743925709197, "ewc_loss": 0.10009765625, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 5.876575469970703, "learning_rate": 1e-05, "loss": 0.5076, "mean_token_accuracy": 0.8611661195755005, "num_tokens": 198068274.0, "step": 5214 }, { "epoch": 0.6634016028495102, "ewc_loss": 0.09423828125, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.965019702911377, "learning_rate": 1e-05, "loss": 0.5161, "mean_token_accuracy": 0.8539158701896667, "num_tokens": 198105326.0, "step": 5215 }, { "epoch": 0.6635288131281007, "ewc_loss": 0.09716796875, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.8159871101379395, "learning_rate": 1e-05, "loss": 0.4782, "mean_token_accuracy": 0.8708821535110474, "num_tokens": 198147485.0, "step": 5216 }, { "epoch": 0.6636560234066913, "ewc_loss": 0.0888671875, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.647375583648682, "learning_rate": 1e-05, "loss": 0.5048, "mean_token_accuracy": 0.8633044362068176, "num_tokens": 198184928.0, "step": 5217 }, { "epoch": 0.6637832336852818, "ewc_loss": 0.09423828125, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.794188499450684, "learning_rate": 1e-05, "loss": 0.5154, "mean_token_accuracy": 0.8606064319610596, "num_tokens": 198224641.0, "step": 5218 }, { "epoch": 0.6639104439638723, "ewc_loss": 0.0859375, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 7.486343383789062e-05, "grad_norm": 5.440075874328613, "learning_rate": 1e-05, "loss": 0.4708, "mean_token_accuracy": 0.8694233894348145, "num_tokens": 198263821.0, "step": 5219 }, { "epoch": 0.6640376542424627, "ewc_loss": 0.09326171875, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.7904229164123535, "learning_rate": 1e-05, "loss": 0.5097, "mean_token_accuracy": 0.8613777160644531, "num_tokens": 198300315.0, "step": 5220 }, { "epoch": 0.6641648645210533, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 7.295608520507812e-05, "grad_norm": 5.302493095397949, "learning_rate": 1e-05, "loss": 0.4885, "mean_token_accuracy": 0.8637048006057739, "num_tokens": 198336543.0, "step": 5221 }, { "epoch": 0.6642920747996438, "ewc_loss": 0.0927734375, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.846714973449707, "learning_rate": 1e-05, "loss": 0.4998, "mean_token_accuracy": 0.8614680767059326, "num_tokens": 198374728.0, "step": 5222 }, { "epoch": 0.6644192850782343, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 7.152557373046875e-05, "grad_norm": 5.183047771453857, "learning_rate": 1e-05, "loss": 0.4865, "mean_token_accuracy": 0.8620688915252686, "num_tokens": 198414504.0, "step": 5223 }, { "epoch": 0.6645464953568249, "ewc_loss": 0.09619140625, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.983366966247559, "learning_rate": 1e-05, "loss": 0.4865, "mean_token_accuracy": 0.8707932233810425, "num_tokens": 198449976.0, "step": 5224 }, { "epoch": 0.6646737056354154, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 7.009506225585938e-05, "grad_norm": 5.065711975097656, "learning_rate": 1e-05, "loss": 0.4844, "mean_token_accuracy": 0.8652319312095642, "num_tokens": 198489522.0, "step": 5225 }, { "epoch": 0.6648009159140058, "ewc_loss": 0.10107421875, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.236245632171631, "learning_rate": 1e-05, "loss": 0.4922, "mean_token_accuracy": 0.868524432182312, "num_tokens": 198524460.0, "step": 5226 }, { "epoch": 0.6649281261925963, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 7.152557373046875e-05, "grad_norm": 4.990241050720215, "learning_rate": 1e-05, "loss": 0.4729, "mean_token_accuracy": 0.8643220067024231, "num_tokens": 198569139.0, "step": 5227 }, { "epoch": 0.6650553364711869, "ewc_loss": 0.10888671875, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.616533279418945, "learning_rate": 1e-05, "loss": 0.4601, "mean_token_accuracy": 0.8778613209724426, "num_tokens": 198597166.0, "step": 5228 }, { "epoch": 0.6651825467497774, "ewc_loss": 0.0888671875, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.161358833312988, "learning_rate": 1e-05, "loss": 0.4935, "mean_token_accuracy": 0.8643679022789001, "num_tokens": 198631646.0, "step": 5229 }, { "epoch": 0.6653097570283679, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.864553451538086, "learning_rate": 1e-05, "loss": 0.5782, "mean_token_accuracy": 0.848041296005249, "num_tokens": 198675743.0, "step": 5230 }, { "epoch": 0.6654369673069584, "ewc_loss": 0.09765625, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.569892883300781, "learning_rate": 1e-05, "loss": 0.5348, "mean_token_accuracy": 0.8562011122703552, "num_tokens": 198715826.0, "step": 5231 }, { "epoch": 0.6655641775855489, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.737722873687744, "learning_rate": 1e-05, "loss": 0.5324, "mean_token_accuracy": 0.8580935001373291, "num_tokens": 198757320.0, "step": 5232 }, { "epoch": 0.6656913878641394, "ewc_loss": 0.10302734375, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 5.856142044067383, "learning_rate": 1e-05, "loss": 0.4989, "mean_token_accuracy": 0.8663549423217773, "num_tokens": 198791110.0, "step": 5233 }, { "epoch": 0.6658185981427299, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.348511219024658, "learning_rate": 1e-05, "loss": 0.5039, "mean_token_accuracy": 0.8636519908905029, "num_tokens": 198828506.0, "step": 5234 }, { "epoch": 0.6659458084213205, "ewc_loss": 0.10009765625, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 5.869588851928711, "learning_rate": 1e-05, "loss": 0.5645, "mean_token_accuracy": 0.8434226512908936, "num_tokens": 198870759.0, "step": 5235 }, { "epoch": 0.666073018699911, "ewc_loss": 0.09423828125, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.949889659881592, "learning_rate": 1e-05, "loss": 0.4819, "mean_token_accuracy": 0.868611752986908, "num_tokens": 198908178.0, "step": 5236 }, { "epoch": 0.6662002289785015, "ewc_loss": 0.09716796875, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.836082458496094, "learning_rate": 1e-05, "loss": 0.5343, "mean_token_accuracy": 0.8588463068008423, "num_tokens": 198942847.0, "step": 5237 }, { "epoch": 0.6663274392570919, "ewc_loss": 0.0888671875, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.622672080993652, "learning_rate": 1e-05, "loss": 0.5085, "mean_token_accuracy": 0.8579180836677551, "num_tokens": 198981444.0, "step": 5238 }, { "epoch": 0.6664546495356825, "ewc_loss": 0.0947265625, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.804993152618408, "learning_rate": 1e-05, "loss": 0.5001, "mean_token_accuracy": 0.8676740527153015, "num_tokens": 199019159.0, "step": 5239 }, { "epoch": 0.666581859814273, "ewc_loss": 0.0859375, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 7.486343383789062e-05, "grad_norm": 5.423938274383545, "learning_rate": 1e-05, "loss": 0.5303, "mean_token_accuracy": 0.8515058159828186, "num_tokens": 199056696.0, "step": 5240 }, { "epoch": 0.6667090700928635, "ewc_loss": 0.0927734375, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.8018269538879395, "learning_rate": 1e-05, "loss": 0.467, "mean_token_accuracy": 0.8707672357559204, "num_tokens": 199092358.0, "step": 5241 }, { "epoch": 0.666836280371454, "ewc_loss": 0.08349609375, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 7.2479248046875e-05, "grad_norm": 5.289584159851074, "learning_rate": 1e-05, "loss": 0.5992, "mean_token_accuracy": 0.8295977711677551, "num_tokens": 199135527.0, "step": 5242 }, { "epoch": 0.6669634906500446, "ewc_loss": 0.0927734375, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.8671674728393555, "learning_rate": 1e-05, "loss": 0.5037, "mean_token_accuracy": 0.860558807849884, "num_tokens": 199171886.0, "step": 5243 }, { "epoch": 0.667090700928635, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 7.152557373046875e-05, "grad_norm": 5.173191547393799, "learning_rate": 1e-05, "loss": 0.5127, "mean_token_accuracy": 0.855100154876709, "num_tokens": 199209807.0, "step": 5244 }, { "epoch": 0.6672179112072255, "ewc_loss": 0.09619140625, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 6.014273643493652, "learning_rate": 1e-05, "loss": 0.5336, "mean_token_accuracy": 0.8571717143058777, "num_tokens": 199244878.0, "step": 5245 }, { "epoch": 0.667345121485816, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 7.05718994140625e-05, "grad_norm": 5.064657211303711, "learning_rate": 1e-05, "loss": 0.5271, "mean_token_accuracy": 0.8546514511108398, "num_tokens": 199279445.0, "step": 5246 }, { "epoch": 0.6674723317644066, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.274806499481201, "learning_rate": 1e-05, "loss": 0.4674, "mean_token_accuracy": 0.8746747970581055, "num_tokens": 199315504.0, "step": 5247 }, { "epoch": 0.6675995420429971, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 7.152557373046875e-05, "grad_norm": 5.0262064933776855, "learning_rate": 1e-05, "loss": 0.552, "mean_token_accuracy": 0.8416926860809326, "num_tokens": 199354345.0, "step": 5248 }, { "epoch": 0.6677267523215876, "ewc_loss": 0.109375, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.616844177246094, "learning_rate": 1e-05, "loss": 0.5861, "mean_token_accuracy": 0.8409315347671509, "num_tokens": 199401530.0, "step": 5249 }, { "epoch": 0.667853962600178, "ewc_loss": 0.08935546875, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.1910552978515625, "learning_rate": 1e-05, "loss": 0.5221, "mean_token_accuracy": 0.854806661605835, "num_tokens": 199440428.0, "step": 5250 }, { "epoch": 0.6679811728787686, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.8364152908325195, "learning_rate": 1e-05, "loss": 0.4905, "mean_token_accuracy": 0.8751391172409058, "num_tokens": 199478875.0, "step": 5251 }, { "epoch": 0.6681083831573591, "ewc_loss": 0.0986328125, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 5.588729381561279, "learning_rate": 1e-05, "loss": 0.5952, "mean_token_accuracy": 0.837645411491394, "num_tokens": 199517287.0, "step": 5252 }, { "epoch": 0.6682355934359496, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.695114612579346, "learning_rate": 1e-05, "loss": 0.5529, "mean_token_accuracy": 0.8545560836791992, "num_tokens": 199553723.0, "step": 5253 }, { "epoch": 0.6683628037145402, "ewc_loss": 0.10302734375, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 5.839381694793701, "learning_rate": 1e-05, "loss": 0.5152, "mean_token_accuracy": 0.8589995503425598, "num_tokens": 199592059.0, "step": 5254 }, { "epoch": 0.6684900139931307, "ewc_loss": 0.10302734375, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.303586959838867, "learning_rate": 1e-05, "loss": 0.5429, "mean_token_accuracy": 0.8538925051689148, "num_tokens": 199633005.0, "step": 5255 }, { "epoch": 0.6686172242717211, "ewc_loss": 0.10009765625, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 5.8527021408081055, "learning_rate": 1e-05, "loss": 0.6106, "mean_token_accuracy": 0.8308510184288025, "num_tokens": 199677507.0, "step": 5256 }, { "epoch": 0.6687444345503116, "ewc_loss": 0.0927734375, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.902720928192139, "learning_rate": 1e-05, "loss": 0.5072, "mean_token_accuracy": 0.8601047992706299, "num_tokens": 199716040.0, "step": 5257 }, { "epoch": 0.6688716448289022, "ewc_loss": 0.09619140625, "ewc_loss_diag": 0.0001125335693359375, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.828341484069824, "learning_rate": 1e-05, "loss": 0.4816, "mean_token_accuracy": 0.863599419593811, "num_tokens": 199750438.0, "step": 5258 }, { "epoch": 0.6689988551074927, "ewc_loss": 0.0888671875, "ewc_loss_diag": 0.00011301040649414062, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.584082126617432, "learning_rate": 1e-05, "loss": 0.5227, "mean_token_accuracy": 0.8586310744285583, "num_tokens": 199786889.0, "step": 5259 }, { "epoch": 0.6691260653860832, "ewc_loss": 0.0947265625, "ewc_loss_diag": 0.00011301040649414062, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.820659637451172, "learning_rate": 1e-05, "loss": 0.5117, "mean_token_accuracy": 0.8614222407341003, "num_tokens": 199821673.0, "step": 5260 }, { "epoch": 0.6692532756646737, "ewc_loss": 0.08447265625, "ewc_loss_diag": 0.00011301040649414062, "ewc_loss_parallel": 7.343292236328125e-05, "grad_norm": 5.3818488121032715, "learning_rate": 1e-05, "loss": 0.5488, "mean_token_accuracy": 0.845344066619873, "num_tokens": 199854313.0, "step": 5261 }, { "epoch": 0.6693804859432643, "ewc_loss": 0.0947265625, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.857318878173828, "learning_rate": 1e-05, "loss": 0.5162, "mean_token_accuracy": 0.8625366687774658, "num_tokens": 199892853.0, "step": 5262 }, { "epoch": 0.6695076962218547, "ewc_loss": 0.08349609375, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 7.2479248046875e-05, "grad_norm": 5.235431671142578, "learning_rate": 1e-05, "loss": 0.5014, "mean_token_accuracy": 0.855777382850647, "num_tokens": 199931754.0, "step": 5263 }, { "epoch": 0.6696349065004452, "ewc_loss": 0.0966796875, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.9701924324035645, "learning_rate": 1e-05, "loss": 0.455, "mean_token_accuracy": 0.8707273006439209, "num_tokens": 199965570.0, "step": 5264 }, { "epoch": 0.6697621167790357, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 7.009506225585938e-05, "grad_norm": 5.111230373382568, "learning_rate": 1e-05, "loss": 0.4621, "mean_token_accuracy": 0.8718428611755371, "num_tokens": 200004931.0, "step": 5265 }, { "epoch": 0.6698893270576263, "ewc_loss": 0.10107421875, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.2046074867248535, "learning_rate": 1e-05, "loss": 0.5064, "mean_token_accuracy": 0.867742657661438, "num_tokens": 200040315.0, "step": 5266 }, { "epoch": 0.6700165373362168, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 7.152557373046875e-05, "grad_norm": 5.033212184906006, "learning_rate": 1e-05, "loss": 0.5048, "mean_token_accuracy": 0.8545361161231995, "num_tokens": 200081206.0, "step": 5267 }, { "epoch": 0.6701437476148073, "ewc_loss": 0.10693359375, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.532302379608154, "learning_rate": 1e-05, "loss": 0.5108, "mean_token_accuracy": 0.8632829189300537, "num_tokens": 200119975.0, "step": 5268 }, { "epoch": 0.6702709578933977, "ewc_loss": 0.087890625, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.1533522605896, "learning_rate": 1e-05, "loss": 0.5775, "mean_token_accuracy": 0.8377227187156677, "num_tokens": 200154956.0, "step": 5269 }, { "epoch": 0.6703981681719883, "ewc_loss": 0.11376953125, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.780157566070557, "learning_rate": 1e-05, "loss": 0.5628, "mean_token_accuracy": 0.8494091629981995, "num_tokens": 200194499.0, "step": 5270 }, { "epoch": 0.6705253784505788, "ewc_loss": 0.09326171875, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.473110675811768, "learning_rate": 1e-05, "loss": 0.4979, "mean_token_accuracy": 0.8633084297180176, "num_tokens": 200227142.0, "step": 5271 }, { "epoch": 0.6706525887291693, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.734259605407715, "learning_rate": 1e-05, "loss": 0.5893, "mean_token_accuracy": 0.8444778323173523, "num_tokens": 200262123.0, "step": 5272 }, { "epoch": 0.6707797990077599, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 5.761736869812012, "learning_rate": 1e-05, "loss": 0.5365, "mean_token_accuracy": 0.8532757759094238, "num_tokens": 200299648.0, "step": 5273 }, { "epoch": 0.6709070092863504, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.403469562530518, "learning_rate": 1e-05, "loss": 0.5814, "mean_token_accuracy": 0.8409011363983154, "num_tokens": 200337191.0, "step": 5274 }, { "epoch": 0.6710342195649408, "ewc_loss": 0.10009765625, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 5.83646297454834, "learning_rate": 1e-05, "loss": 0.4946, "mean_token_accuracy": 0.8650274872779846, "num_tokens": 200371487.0, "step": 5275 }, { "epoch": 0.6711614298435313, "ewc_loss": 0.095703125, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 6.008063793182373, "learning_rate": 1e-05, "loss": 0.5118, "mean_token_accuracy": 0.8609246015548706, "num_tokens": 200412634.0, "step": 5276 }, { "epoch": 0.6712886401221219, "ewc_loss": 0.0966796875, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.804449558258057, "learning_rate": 1e-05, "loss": 0.501, "mean_token_accuracy": 0.8624603152275085, "num_tokens": 200449663.0, "step": 5277 }, { "epoch": 0.6714158504007124, "ewc_loss": 0.09033203125, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.680877685546875, "learning_rate": 1e-05, "loss": 0.5679, "mean_token_accuracy": 0.8443889021873474, "num_tokens": 200494962.0, "step": 5278 }, { "epoch": 0.6715430606793029, "ewc_loss": 0.09423828125, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.778504371643066, "learning_rate": 1e-05, "loss": 0.4676, "mean_token_accuracy": 0.8705648183822632, "num_tokens": 200531101.0, "step": 5279 }, { "epoch": 0.6716702709578934, "ewc_loss": 0.08642578125, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 7.534027099609375e-05, "grad_norm": 5.462514400482178, "learning_rate": 1e-05, "loss": 0.489, "mean_token_accuracy": 0.8651160597801208, "num_tokens": 200568725.0, "step": 5280 }, { "epoch": 0.6717974812364839, "ewc_loss": 0.09326171875, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.775294303894043, "learning_rate": 1e-05, "loss": 0.5179, "mean_token_accuracy": 0.853761613368988, "num_tokens": 200607319.0, "step": 5281 }, { "epoch": 0.6719246915150744, "ewc_loss": 0.08447265625, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 7.343292236328125e-05, "grad_norm": 5.312639236450195, "learning_rate": 1e-05, "loss": 0.5119, "mean_token_accuracy": 0.8555603623390198, "num_tokens": 200654791.0, "step": 5282 }, { "epoch": 0.6720519017936649, "ewc_loss": 0.09228515625, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.836215496063232, "learning_rate": 1e-05, "loss": 0.5023, "mean_token_accuracy": 0.8624030947685242, "num_tokens": 200694175.0, "step": 5283 }, { "epoch": 0.6721791120722554, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 7.152557373046875e-05, "grad_norm": 5.190093517303467, "learning_rate": 1e-05, "loss": 0.5026, "mean_token_accuracy": 0.8574573993682861, "num_tokens": 200726675.0, "step": 5284 }, { "epoch": 0.672306322350846, "ewc_loss": 0.095703125, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.984378337860107, "learning_rate": 1e-05, "loss": 0.5525, "mean_token_accuracy": 0.8482178449630737, "num_tokens": 200759358.0, "step": 5285 }, { "epoch": 0.6724335326294365, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 7.05718994140625e-05, "grad_norm": 5.068755149841309, "learning_rate": 1e-05, "loss": 0.5032, "mean_token_accuracy": 0.8603885173797607, "num_tokens": 200797735.0, "step": 5286 }, { "epoch": 0.6725607429080269, "ewc_loss": 0.10205078125, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.232359886169434, "learning_rate": 1e-05, "loss": 0.5337, "mean_token_accuracy": 0.8541186451911926, "num_tokens": 200835446.0, "step": 5287 }, { "epoch": 0.6726879531866174, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 7.104873657226562e-05, "grad_norm": 4.999843597412109, "learning_rate": 1e-05, "loss": 0.4975, "mean_token_accuracy": 0.8629699945449829, "num_tokens": 200881020.0, "step": 5288 }, { "epoch": 0.672815163465208, "ewc_loss": 0.10888671875, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.597697734832764, "learning_rate": 1e-05, "loss": 0.516, "mean_token_accuracy": 0.862076997756958, "num_tokens": 200921549.0, "step": 5289 }, { "epoch": 0.6729423737437985, "ewc_loss": 0.08837890625, "ewc_loss_diag": 0.00011396408081054688, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.149535655975342, "learning_rate": 1e-05, "loss": 0.5387, "mean_token_accuracy": 0.8455479741096497, "num_tokens": 200953991.0, "step": 5290 }, { "epoch": 0.673069584022389, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.881230354309082, "learning_rate": 1e-05, "loss": 0.5218, "mean_token_accuracy": 0.860991358757019, "num_tokens": 200990240.0, "step": 5291 }, { "epoch": 0.6731967943009796, "ewc_loss": 0.09716796875, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.55095100402832, "learning_rate": 1e-05, "loss": 0.4722, "mean_token_accuracy": 0.8724160194396973, "num_tokens": 201022433.0, "step": 5292 }, { "epoch": 0.67332400457957, "ewc_loss": 0.11376953125, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.789162635803223, "learning_rate": 1e-05, "loss": 0.531, "mean_token_accuracy": 0.8586090803146362, "num_tokens": 201055801.0, "step": 5293 }, { "epoch": 0.6734512148581605, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 5.848043918609619, "learning_rate": 1e-05, "loss": 0.5519, "mean_token_accuracy": 0.850121021270752, "num_tokens": 201102141.0, "step": 5294 }, { "epoch": 0.673578425136751, "ewc_loss": 0.10595703125, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.41144323348999, "learning_rate": 1e-05, "loss": 0.5975, "mean_token_accuracy": 0.8483654260635376, "num_tokens": 201139834.0, "step": 5295 }, { "epoch": 0.6737056354153416, "ewc_loss": 0.10205078125, "ewc_loss_diag": 0.00011396408081054688, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 5.902432441711426, "learning_rate": 1e-05, "loss": 0.493, "mean_token_accuracy": 0.8679783940315247, "num_tokens": 201169650.0, "step": 5296 }, { "epoch": 0.6738328456939321, "ewc_loss": 0.095703125, "ewc_loss_diag": 0.00011396408081054688, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.990293025970459, "learning_rate": 1e-05, "loss": 0.507, "mean_token_accuracy": 0.8615058660507202, "num_tokens": 201208716.0, "step": 5297 }, { "epoch": 0.6739600559725226, "ewc_loss": 0.09765625, "ewc_loss_diag": 0.00011396408081054688, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.831302642822266, "learning_rate": 1e-05, "loss": 0.5187, "mean_token_accuracy": 0.8604567646980286, "num_tokens": 201253976.0, "step": 5298 }, { "epoch": 0.674087266251113, "ewc_loss": 0.08984375, "ewc_loss_diag": 0.00011396408081054688, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.663236618041992, "learning_rate": 1e-05, "loss": 0.5409, "mean_token_accuracy": 0.8510624170303345, "num_tokens": 201291478.0, "step": 5299 }, { "epoch": 0.6742144765297036, "ewc_loss": 0.095703125, "ewc_loss_diag": 0.00011396408081054688, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.803822040557861, "learning_rate": 1e-05, "loss": 0.5113, "mean_token_accuracy": 0.8566805720329285, "num_tokens": 201328824.0, "step": 5300 }, { "epoch": 0.6743416868082941, "ewc_loss": 0.0859375, "ewc_loss_diag": 0.00011396408081054688, "ewc_loss_parallel": 7.486343383789062e-05, "grad_norm": 5.459774494171143, "learning_rate": 1e-05, "loss": 0.5523, "mean_token_accuracy": 0.8450684547424316, "num_tokens": 201363859.0, "step": 5301 }, { "epoch": 0.6744688970868846, "ewc_loss": 0.0927734375, "ewc_loss_diag": 0.00011396408081054688, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.794314384460449, "learning_rate": 1e-05, "loss": 0.4904, "mean_token_accuracy": 0.8660727739334106, "num_tokens": 201402635.0, "step": 5302 }, { "epoch": 0.6745961073654752, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.00011396408081054688, "ewc_loss_parallel": 7.295608520507812e-05, "grad_norm": 5.316749095916748, "learning_rate": 1e-05, "loss": 0.5305, "mean_token_accuracy": 0.8506878018379211, "num_tokens": 201444203.0, "step": 5303 }, { "epoch": 0.6747233176440657, "ewc_loss": 0.09326171875, "ewc_loss_diag": 0.00011396408081054688, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.849734783172607, "learning_rate": 1e-05, "loss": 0.5748, "mean_token_accuracy": 0.8429633378982544, "num_tokens": 201486397.0, "step": 5304 }, { "epoch": 0.6748505279226561, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.00011396408081054688, "ewc_loss_parallel": 7.152557373046875e-05, "grad_norm": 5.19355583190918, "learning_rate": 1e-05, "loss": 0.5259, "mean_token_accuracy": 0.8516452312469482, "num_tokens": 201528262.0, "step": 5305 }, { "epoch": 0.6749777382012466, "ewc_loss": 0.09619140625, "ewc_loss_diag": 0.00011396408081054688, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.983263969421387, "learning_rate": 1e-05, "loss": 0.5354, "mean_token_accuracy": 0.8581737875938416, "num_tokens": 201566274.0, "step": 5306 }, { "epoch": 0.6751049484798372, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.00011396408081054688, "ewc_loss_parallel": 7.104873657226562e-05, "grad_norm": 5.082651615142822, "learning_rate": 1e-05, "loss": 0.5288, "mean_token_accuracy": 0.8530263900756836, "num_tokens": 201610196.0, "step": 5307 }, { "epoch": 0.6752321587584277, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.00011396408081054688, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.234254837036133, "learning_rate": 1e-05, "loss": 0.538, "mean_token_accuracy": 0.852201521396637, "num_tokens": 201650144.0, "step": 5308 }, { "epoch": 0.6753593690370182, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.00011396408081054688, "ewc_loss_parallel": 7.200241088867188e-05, "grad_norm": 5.012463569641113, "learning_rate": 1e-05, "loss": 0.5088, "mean_token_accuracy": 0.8579522371292114, "num_tokens": 201689550.0, "step": 5309 }, { "epoch": 0.6754865793156087, "ewc_loss": 0.109375, "ewc_loss_diag": 0.00011396408081054688, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.5843024253845215, "learning_rate": 1e-05, "loss": 0.4853, "mean_token_accuracy": 0.8724796175956726, "num_tokens": 201729807.0, "step": 5310 }, { "epoch": 0.6756137895941993, "ewc_loss": 0.08837890625, "ewc_loss_diag": 0.00011396408081054688, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.158278942108154, "learning_rate": 1e-05, "loss": 0.489, "mean_token_accuracy": 0.8632348775863647, "num_tokens": 201759540.0, "step": 5311 }, { "epoch": 0.6757409998727897, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.00011396408081054688, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.847904205322266, "learning_rate": 1e-05, "loss": 0.5625, "mean_token_accuracy": 0.8541977405548096, "num_tokens": 201800067.0, "step": 5312 }, { "epoch": 0.6758682101513802, "ewc_loss": 0.095703125, "ewc_loss_diag": 0.00011396408081054688, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.5404791831970215, "learning_rate": 1e-05, "loss": 0.4991, "mean_token_accuracy": 0.8607376217842102, "num_tokens": 201832688.0, "step": 5313 }, { "epoch": 0.6759954204299707, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.00011396408081054688, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.747309684753418, "learning_rate": 1e-05, "loss": 0.5483, "mean_token_accuracy": 0.8540331125259399, "num_tokens": 201876090.0, "step": 5314 }, { "epoch": 0.6761226307085613, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.00011396408081054688, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 5.823002815246582, "learning_rate": 1e-05, "loss": 0.5764, "mean_token_accuracy": 0.8419017195701599, "num_tokens": 201914339.0, "step": 5315 }, { "epoch": 0.6762498409871518, "ewc_loss": 0.10498046875, "ewc_loss_diag": 0.00011396408081054688, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.396341323852539, "learning_rate": 1e-05, "loss": 0.5029, "mean_token_accuracy": 0.8691251277923584, "num_tokens": 201945746.0, "step": 5316 }, { "epoch": 0.6763770512657423, "ewc_loss": 0.09912109375, "ewc_loss_diag": 0.00011396408081054688, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 5.8709540367126465, "learning_rate": 1e-05, "loss": 0.4888, "mean_token_accuracy": 0.8635398745536804, "num_tokens": 201978773.0, "step": 5317 }, { "epoch": 0.6765042615443327, "ewc_loss": 0.095703125, "ewc_loss_diag": 0.00011396408081054688, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.9944963455200195, "learning_rate": 1e-05, "loss": 0.5216, "mean_token_accuracy": 0.8584101796150208, "num_tokens": 202019838.0, "step": 5318 }, { "epoch": 0.6766314718229233, "ewc_loss": 0.0966796875, "ewc_loss_diag": 0.00011396408081054688, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.823101043701172, "learning_rate": 1e-05, "loss": 0.5326, "mean_token_accuracy": 0.8601127862930298, "num_tokens": 202060825.0, "step": 5319 }, { "epoch": 0.6767586821015138, "ewc_loss": 0.0888671875, "ewc_loss_diag": 0.00011396408081054688, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.675185203552246, "learning_rate": 1e-05, "loss": 0.4547, "mean_token_accuracy": 0.874372661113739, "num_tokens": 202092656.0, "step": 5320 }, { "epoch": 0.6768858923801043, "ewc_loss": 0.0947265625, "ewc_loss_diag": 0.00011396408081054688, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.801588535308838, "learning_rate": 1e-05, "loss": 0.5601, "mean_token_accuracy": 0.8432846665382385, "num_tokens": 202126982.0, "step": 5321 }, { "epoch": 0.6770131026586949, "ewc_loss": 0.0859375, "ewc_loss_diag": 0.00011396408081054688, "ewc_loss_parallel": 7.486343383789062e-05, "grad_norm": 5.457849979400635, "learning_rate": 1e-05, "loss": 0.4925, "mean_token_accuracy": 0.8620844483375549, "num_tokens": 202157550.0, "step": 5322 }, { "epoch": 0.6771403129372854, "ewc_loss": 0.09423828125, "ewc_loss_diag": 0.00011396408081054688, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.820432662963867, "learning_rate": 1e-05, "loss": 0.5531, "mean_token_accuracy": 0.8458794355392456, "num_tokens": 202192168.0, "step": 5323 }, { "epoch": 0.6772675232158758, "ewc_loss": 0.08447265625, "ewc_loss_diag": 0.00011396408081054688, "ewc_loss_parallel": 7.343292236328125e-05, "grad_norm": 5.295149326324463, "learning_rate": 1e-05, "loss": 0.4657, "mean_token_accuracy": 0.8684595823287964, "num_tokens": 202226887.0, "step": 5324 }, { "epoch": 0.6773947334944663, "ewc_loss": 0.095703125, "ewc_loss_diag": 0.00011491775512695312, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.897035121917725, "learning_rate": 1e-05, "loss": 0.5522, "mean_token_accuracy": 0.8449686765670776, "num_tokens": 202269631.0, "step": 5325 }, { "epoch": 0.6775219437730569, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.00011491775512695312, "ewc_loss_parallel": 7.200241088867188e-05, "grad_norm": 5.165600299835205, "learning_rate": 1e-05, "loss": 0.5138, "mean_token_accuracy": 0.8574391603469849, "num_tokens": 202306951.0, "step": 5326 }, { "epoch": 0.6776491540516474, "ewc_loss": 0.0986328125, "ewc_loss_diag": 0.00011491775512695312, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 6.060702323913574, "learning_rate": 1e-05, "loss": 0.4843, "mean_token_accuracy": 0.8684042096138, "num_tokens": 202342922.0, "step": 5327 }, { "epoch": 0.6777763643302379, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.00011491775512695312, "ewc_loss_parallel": 7.05718994140625e-05, "grad_norm": 5.05093240737915, "learning_rate": 1e-05, "loss": 0.5043, "mean_token_accuracy": 0.8561899662017822, "num_tokens": 202382125.0, "step": 5328 }, { "epoch": 0.6779035746088284, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.00011491775512695312, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.349013805389404, "learning_rate": 1e-05, "loss": 0.52, "mean_token_accuracy": 0.8598612546920776, "num_tokens": 202424914.0, "step": 5329 }, { "epoch": 0.6780307848874189, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.00011491775512695312, "ewc_loss_parallel": 7.2479248046875e-05, "grad_norm": 5.039023399353027, "learning_rate": 1e-05, "loss": 0.4966, "mean_token_accuracy": 0.8576356172561646, "num_tokens": 202460828.0, "step": 5330 }, { "epoch": 0.6781579951660094, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.00011491775512695312, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.711959362030029, "learning_rate": 1e-05, "loss": 0.5072, "mean_token_accuracy": 0.8640127778053284, "num_tokens": 202492484.0, "step": 5331 }, { "epoch": 0.6782852054445999, "ewc_loss": 0.091796875, "ewc_loss_diag": 0.00011491775512695312, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.277234077453613, "learning_rate": 1e-05, "loss": 0.6061, "mean_token_accuracy": 0.8289239406585693, "num_tokens": 202529079.0, "step": 5332 }, { "epoch": 0.6784124157231904, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.00011491775512695312, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.889946937561035, "learning_rate": 1e-05, "loss": 0.5799, "mean_token_accuracy": 0.845306396484375, "num_tokens": 202563024.0, "step": 5333 }, { "epoch": 0.678539626001781, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.00011491775512695312, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 5.688559532165527, "learning_rate": 1e-05, "loss": 0.5187, "mean_token_accuracy": 0.8588167428970337, "num_tokens": 202601264.0, "step": 5334 }, { "epoch": 0.6786668362803715, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.00011491775512695312, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.65932035446167, "learning_rate": 1e-05, "loss": 0.5426, "mean_token_accuracy": 0.8534661531448364, "num_tokens": 202642933.0, "step": 5335 }, { "epoch": 0.6787940465589619, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.00011491775512695312, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 5.911284923553467, "learning_rate": 1e-05, "loss": 0.4987, "mean_token_accuracy": 0.866452157497406, "num_tokens": 202674851.0, "step": 5336 }, { "epoch": 0.6789212568375524, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.00011491775512695312, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.23647928237915, "learning_rate": 1e-05, "loss": 0.5078, "mean_token_accuracy": 0.8690400123596191, "num_tokens": 202711117.0, "step": 5337 }, { "epoch": 0.679048467116143, "ewc_loss": 0.099609375, "ewc_loss_diag": 0.00011491775512695312, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 5.87282657623291, "learning_rate": 1e-05, "loss": 0.4936, "mean_token_accuracy": 0.8684582710266113, "num_tokens": 202749025.0, "step": 5338 }, { "epoch": 0.6791756773947335, "ewc_loss": 0.0927734375, "ewc_loss_diag": 0.00011491775512695312, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.855221271514893, "learning_rate": 1e-05, "loss": 0.5718, "mean_token_accuracy": 0.8403915762901306, "num_tokens": 202790733.0, "step": 5339 }, { "epoch": 0.679302887673324, "ewc_loss": 0.0966796875, "ewc_loss_diag": 0.00011491775512695312, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.826707363128662, "learning_rate": 1e-05, "loss": 0.5311, "mean_token_accuracy": 0.8545728921890259, "num_tokens": 202825753.0, "step": 5340 }, { "epoch": 0.6794300979519146, "ewc_loss": 0.0888671875, "ewc_loss_diag": 0.00011491775512695312, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.562314987182617, "learning_rate": 1e-05, "loss": 0.4677, "mean_token_accuracy": 0.8723993301391602, "num_tokens": 202862069.0, "step": 5341 }, { "epoch": 0.679557308230505, "ewc_loss": 0.0947265625, "ewc_loss_diag": 0.00011491775512695312, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.811643600463867, "learning_rate": 1e-05, "loss": 0.5059, "mean_token_accuracy": 0.8606709837913513, "num_tokens": 202904526.0, "step": 5342 }, { "epoch": 0.6796845185090955, "ewc_loss": 0.0849609375, "ewc_loss_diag": 0.00011491775512695312, "ewc_loss_parallel": 7.390975952148438e-05, "grad_norm": 5.380148887634277, "learning_rate": 1e-05, "loss": 0.5513, "mean_token_accuracy": 0.8461298942565918, "num_tokens": 202942342.0, "step": 5343 }, { "epoch": 0.679811728787686, "ewc_loss": 0.0947265625, "ewc_loss_diag": 0.00011491775512695312, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.860605716705322, "learning_rate": 1e-05, "loss": 0.5225, "mean_token_accuracy": 0.8501073122024536, "num_tokens": 202974644.0, "step": 5344 }, { "epoch": 0.6799389390662766, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.00011491775512695312, "ewc_loss_parallel": 7.200241088867188e-05, "grad_norm": 5.233414173126221, "learning_rate": 1e-05, "loss": 0.4631, "mean_token_accuracy": 0.8714728355407715, "num_tokens": 203011728.0, "step": 5345 }, { "epoch": 0.6800661493448671, "ewc_loss": 0.09765625, "ewc_loss_diag": 0.00011491775512695312, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.9784111976623535, "learning_rate": 1e-05, "loss": 0.53, "mean_token_accuracy": 0.8541991710662842, "num_tokens": 203051768.0, "step": 5346 }, { "epoch": 0.6801933596234576, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.00011491775512695312, "ewc_loss_parallel": 7.152557373046875e-05, "grad_norm": 5.1144304275512695, "learning_rate": 1e-05, "loss": 0.5527, "mean_token_accuracy": 0.8486518263816833, "num_tokens": 203094943.0, "step": 5347 }, { "epoch": 0.680320569902048, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.00011491775512695312, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.212297439575195, "learning_rate": 1e-05, "loss": 0.5266, "mean_token_accuracy": 0.860171377658844, "num_tokens": 203130587.0, "step": 5348 }, { "epoch": 0.6804477801806386, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.00011491775512695312, "ewc_loss_parallel": 7.152557373046875e-05, "grad_norm": 5.039633274078369, "learning_rate": 1e-05, "loss": 0.5156, "mean_token_accuracy": 0.8542052507400513, "num_tokens": 203170030.0, "step": 5349 }, { "epoch": 0.6805749904592291, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.00011491775512695312, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.536056995391846, "learning_rate": 1e-05, "loss": 0.5406, "mean_token_accuracy": 0.8549258708953857, "num_tokens": 203209382.0, "step": 5350 }, { "epoch": 0.6807022007378196, "ewc_loss": 0.087890625, "ewc_loss_diag": 0.00011491775512695312, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.135542869567871, "learning_rate": 1e-05, "loss": 0.514, "mean_token_accuracy": 0.8560261726379395, "num_tokens": 203247750.0, "step": 5351 }, { "epoch": 0.6808294110164101, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.00011491775512695312, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.817331790924072, "learning_rate": 1e-05, "loss": 0.5269, "mean_token_accuracy": 0.8627943992614746, "num_tokens": 203284443.0, "step": 5352 }, { "epoch": 0.6809566212950007, "ewc_loss": 0.0947265625, "ewc_loss_diag": 0.00011491775512695312, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.478464603424072, "learning_rate": 1e-05, "loss": 0.4815, "mean_token_accuracy": 0.8682284355163574, "num_tokens": 203324525.0, "step": 5353 }, { "epoch": 0.6810838315735911, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.00011491775512695312, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.7845306396484375, "learning_rate": 1e-05, "loss": 0.5303, "mean_token_accuracy": 0.8619226217269897, "num_tokens": 203369633.0, "step": 5354 }, { "epoch": 0.6812110418521816, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.00011491775512695312, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 5.804527759552002, "learning_rate": 1e-05, "loss": 0.5242, "mean_token_accuracy": 0.8578211665153503, "num_tokens": 203409549.0, "step": 5355 }, { "epoch": 0.6813382521307721, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.00011491775512695312, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.441856384277344, "learning_rate": 1e-05, "loss": 0.5548, "mean_token_accuracy": 0.8498423099517822, "num_tokens": 203446856.0, "step": 5356 }, { "epoch": 0.6814654624093627, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.00011491775512695312, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 5.882040023803711, "learning_rate": 1e-05, "loss": 0.4559, "mean_token_accuracy": 0.8816423416137695, "num_tokens": 203478575.0, "step": 5357 }, { "epoch": 0.6815926726879532, "ewc_loss": 0.0966796875, "ewc_loss_diag": 0.00011491775512695312, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 6.029261112213135, "learning_rate": 1e-05, "loss": 0.4807, "mean_token_accuracy": 0.8713228702545166, "num_tokens": 203515617.0, "step": 5358 }, { "epoch": 0.6817198829665437, "ewc_loss": 0.09765625, "ewc_loss_diag": 0.00011491775512695312, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.857811450958252, "learning_rate": 1e-05, "loss": 0.5469, "mean_token_accuracy": 0.8518418073654175, "num_tokens": 203548636.0, "step": 5359 }, { "epoch": 0.6818470932451343, "ewc_loss": 0.0888671875, "ewc_loss_diag": 0.00011491775512695312, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.674193859100342, "learning_rate": 1e-05, "loss": 0.4918, "mean_token_accuracy": 0.8662705421447754, "num_tokens": 203581358.0, "step": 5360 }, { "epoch": 0.6819743035237247, "ewc_loss": 0.095703125, "ewc_loss_diag": 0.00011491775512695312, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.824483871459961, "learning_rate": 1e-05, "loss": 0.4703, "mean_token_accuracy": 0.8761888742446899, "num_tokens": 203615052.0, "step": 5361 }, { "epoch": 0.6821015138023152, "ewc_loss": 0.0859375, "ewc_loss_diag": 0.00011491775512695312, "ewc_loss_parallel": 7.43865966796875e-05, "grad_norm": 5.454099178314209, "learning_rate": 1e-05, "loss": 0.5373, "mean_token_accuracy": 0.8537598848342896, "num_tokens": 203655743.0, "step": 5362 }, { "epoch": 0.6822287240809057, "ewc_loss": 0.09375, "ewc_loss_diag": 0.00011491775512695312, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.826642036437988, "learning_rate": 1e-05, "loss": 0.5525, "mean_token_accuracy": 0.8469635844230652, "num_tokens": 203689370.0, "step": 5363 }, { "epoch": 0.6823559343594963, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.00011491775512695312, "ewc_loss_parallel": 7.2479248046875e-05, "grad_norm": 5.302615642547607, "learning_rate": 1e-05, "loss": 0.514, "mean_token_accuracy": 0.8550933599472046, "num_tokens": 203723594.0, "step": 5364 }, { "epoch": 0.6824831446380868, "ewc_loss": 0.09375, "ewc_loss_diag": 0.00011491775512695312, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.884892463684082, "learning_rate": 1e-05, "loss": 0.5636, "mean_token_accuracy": 0.8425357937812805, "num_tokens": 203761746.0, "step": 5365 }, { "epoch": 0.6826103549166773, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.00011491775512695312, "ewc_loss_parallel": 7.152557373046875e-05, "grad_norm": 5.185221195220947, "learning_rate": 1e-05, "loss": 0.504, "mean_token_accuracy": 0.8585103154182434, "num_tokens": 203801904.0, "step": 5366 }, { "epoch": 0.6827375651952677, "ewc_loss": 0.0966796875, "ewc_loss_diag": 0.00011491775512695312, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 6.029881954193115, "learning_rate": 1e-05, "loss": 0.5119, "mean_token_accuracy": 0.8610367774963379, "num_tokens": 203837775.0, "step": 5367 }, { "epoch": 0.6828647754738583, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.00011491775512695312, "ewc_loss_parallel": 7.104873657226562e-05, "grad_norm": 5.072497844696045, "learning_rate": 1e-05, "loss": 0.4478, "mean_token_accuracy": 0.8769159317016602, "num_tokens": 203876456.0, "step": 5368 }, { "epoch": 0.6829919857524488, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.00011491775512695312, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.3082170486450195, "learning_rate": 1e-05, "loss": 0.548, "mean_token_accuracy": 0.8553720712661743, "num_tokens": 203917946.0, "step": 5369 }, { "epoch": 0.6831191960310393, "ewc_loss": 0.08447265625, "ewc_loss_diag": 0.00011539459228515625, "ewc_loss_parallel": 7.295608520507812e-05, "grad_norm": 5.036687850952148, "learning_rate": 1e-05, "loss": 0.4957, "mean_token_accuracy": 0.8628888130187988, "num_tokens": 203959544.0, "step": 5370 }, { "epoch": 0.6832464063096299, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.00011539459228515625, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.656332492828369, "learning_rate": 1e-05, "loss": 0.5611, "mean_token_accuracy": 0.8500419855117798, "num_tokens": 204000882.0, "step": 5371 }, { "epoch": 0.6833736165882204, "ewc_loss": 0.0908203125, "ewc_loss_diag": 0.00011539459228515625, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.233451843261719, "learning_rate": 1e-05, "loss": 0.4957, "mean_token_accuracy": 0.8640698194503784, "num_tokens": 204037715.0, "step": 5372 }, { "epoch": 0.6835008268668108, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.8583598136901855, "learning_rate": 1e-05, "loss": 0.543, "mean_token_accuracy": 0.8542425632476807, "num_tokens": 204073145.0, "step": 5373 }, { "epoch": 0.6836280371454013, "ewc_loss": 0.0986328125, "ewc_loss_diag": 0.00011539459228515625, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.6239166259765625, "learning_rate": 1e-05, "loss": 0.5377, "mean_token_accuracy": 0.853357195854187, "num_tokens": 204112781.0, "step": 5374 }, { "epoch": 0.6837552474239919, "ewc_loss": 0.11181640625, "ewc_loss_diag": 0.00011539459228515625, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.684898376464844, "learning_rate": 1e-05, "loss": 0.534, "mean_token_accuracy": 0.8578953742980957, "num_tokens": 204151128.0, "step": 5375 }, { "epoch": 0.6838824577025824, "ewc_loss": 0.10302734375, "ewc_loss_diag": 0.00011539459228515625, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 5.853515148162842, "learning_rate": 1e-05, "loss": 0.5057, "mean_token_accuracy": 0.8656919598579407, "num_tokens": 204188666.0, "step": 5376 }, { "epoch": 0.6840096679811729, "ewc_loss": 0.10400390625, "ewc_loss_diag": 0.00011539459228515625, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.301807880401611, "learning_rate": 1e-05, "loss": 0.463, "mean_token_accuracy": 0.8787928819656372, "num_tokens": 204225410.0, "step": 5377 }, { "epoch": 0.6841368782597634, "ewc_loss": 0.099609375, "ewc_loss_diag": 0.00011539459228515625, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 5.8606343269348145, "learning_rate": 1e-05, "loss": 0.5889, "mean_token_accuracy": 0.84161776304245, "num_tokens": 204271506.0, "step": 5378 }, { "epoch": 0.6842640885383539, "ewc_loss": 0.0947265625, "ewc_loss_diag": 0.00011539459228515625, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.921813011169434, "learning_rate": 1e-05, "loss": 0.5203, "mean_token_accuracy": 0.85511314868927, "num_tokens": 204310569.0, "step": 5379 }, { "epoch": 0.6843912988169444, "ewc_loss": 0.09619140625, "ewc_loss_diag": 0.00011539459228515625, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.819076061248779, "learning_rate": 1e-05, "loss": 0.5067, "mean_token_accuracy": 0.8615453243255615, "num_tokens": 204349614.0, "step": 5380 }, { "epoch": 0.6845185090955349, "ewc_loss": 0.08984375, "ewc_loss_diag": 0.00011539459228515625, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.722102165222168, "learning_rate": 1e-05, "loss": 0.5546, "mean_token_accuracy": 0.844946563243866, "num_tokens": 204381985.0, "step": 5381 }, { "epoch": 0.6846457193741254, "ewc_loss": 0.09423828125, "ewc_loss_diag": 0.00011539459228515625, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.739494800567627, "learning_rate": 1e-05, "loss": 0.5187, "mean_token_accuracy": 0.8530942797660828, "num_tokens": 204418433.0, "step": 5382 }, { "epoch": 0.684772929652716, "ewc_loss": 0.0869140625, "ewc_loss_diag": 0.00011539459228515625, "ewc_loss_parallel": 7.534027099609375e-05, "grad_norm": 5.4810967445373535, "learning_rate": 1e-05, "loss": 0.5013, "mean_token_accuracy": 0.8570268154144287, "num_tokens": 204457544.0, "step": 5383 }, { "epoch": 0.6849001399313065, "ewc_loss": 0.0927734375, "ewc_loss_diag": 0.00011539459228515625, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.705505847930908, "learning_rate": 1e-05, "loss": 0.5108, "mean_token_accuracy": 0.8599340915679932, "num_tokens": 204500626.0, "step": 5384 }, { "epoch": 0.6850273502098969, "ewc_loss": 0.08642578125, "ewc_loss_diag": 0.00011539459228515625, "ewc_loss_parallel": 7.486343383789062e-05, "grad_norm": 5.369982719421387, "learning_rate": 1e-05, "loss": 0.5424, "mean_token_accuracy": 0.845858633518219, "num_tokens": 204540861.0, "step": 5385 }, { "epoch": 0.6851545604884874, "ewc_loss": 0.091796875, "ewc_loss_diag": 0.00011539459228515625, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.730661392211914, "learning_rate": 1e-05, "loss": 0.4843, "mean_token_accuracy": 0.8703367114067078, "num_tokens": 204581555.0, "step": 5386 }, { "epoch": 0.685281770767078, "ewc_loss": 0.0849609375, "ewc_loss_diag": 0.00011539459228515625, "ewc_loss_parallel": 7.343292236328125e-05, "grad_norm": 5.274158477783203, "learning_rate": 1e-05, "loss": 0.4938, "mean_token_accuracy": 0.859886646270752, "num_tokens": 204614931.0, "step": 5387 }, { "epoch": 0.6854089810456685, "ewc_loss": 0.09326171875, "ewc_loss_diag": 0.00011539459228515625, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.824103355407715, "learning_rate": 1e-05, "loss": 0.5278, "mean_token_accuracy": 0.8553460836410522, "num_tokens": 204649678.0, "step": 5388 }, { "epoch": 0.685536191324259, "ewc_loss": 0.08349609375, "ewc_loss_diag": 0.00011539459228515625, "ewc_loss_parallel": 7.200241088867188e-05, "grad_norm": 5.141401767730713, "learning_rate": 1e-05, "loss": 0.531, "mean_token_accuracy": 0.8516958355903625, "num_tokens": 204690617.0, "step": 5389 }, { "epoch": 0.6856634016028496, "ewc_loss": 0.09765625, "ewc_loss_diag": 0.00011539459228515625, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 6.020816802978516, "learning_rate": 1e-05, "loss": 0.5303, "mean_token_accuracy": 0.8510658740997314, "num_tokens": 204725685.0, "step": 5390 }, { "epoch": 0.68579061188144, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.00011539459228515625, "ewc_loss_parallel": 7.05718994140625e-05, "grad_norm": 5.006673336029053, "learning_rate": 1e-05, "loss": 0.5714, "mean_token_accuracy": 0.8376470804214478, "num_tokens": 204768265.0, "step": 5391 }, { "epoch": 0.6859178221600305, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.00011539459228515625, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.354738235473633, "learning_rate": 1e-05, "loss": 0.5159, "mean_token_accuracy": 0.8614272475242615, "num_tokens": 204806211.0, "step": 5392 }, { "epoch": 0.686045032438621, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.00011539459228515625, "ewc_loss_parallel": 7.104873657226562e-05, "grad_norm": 4.946559906005859, "learning_rate": 1e-05, "loss": 0.5119, "mean_token_accuracy": 0.8588972091674805, "num_tokens": 204844982.0, "step": 5393 }, { "epoch": 0.6861722427172116, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.00011539459228515625, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.815731525421143, "learning_rate": 1e-05, "loss": 0.5033, "mean_token_accuracy": 0.871275782585144, "num_tokens": 204881676.0, "step": 5394 }, { "epoch": 0.6862994529958021, "ewc_loss": 0.09228515625, "ewc_loss_diag": 0.00011539459228515625, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.2603936195373535, "learning_rate": 1e-05, "loss": 0.5229, "mean_token_accuracy": 0.8564465045928955, "num_tokens": 204920316.0, "step": 5395 }, { "epoch": 0.6864266632743926, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 0.00011110305786132812, "grad_norm": 7.042026996612549, "learning_rate": 1e-05, "loss": 0.5745, "mean_token_accuracy": 0.8472375273704529, "num_tokens": 204960269.0, "step": 5396 }, { "epoch": 0.686553873552983, "ewc_loss": 0.10595703125, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 5.767488956451416, "learning_rate": 1e-05, "loss": 0.5291, "mean_token_accuracy": 0.8573086857795715, "num_tokens": 204993238.0, "step": 5397 }, { "epoch": 0.6866810838315736, "ewc_loss": 0.11474609375, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.791553974151611, "learning_rate": 1e-05, "loss": 0.5675, "mean_token_accuracy": 0.8485147953033447, "num_tokens": 205029550.0, "step": 5398 }, { "epoch": 0.6868082941101641, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 5.9902777671813965, "learning_rate": 1e-05, "loss": 0.5518, "mean_token_accuracy": 0.8510777354240417, "num_tokens": 205073198.0, "step": 5399 }, { "epoch": 0.6869355043887546, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.370913982391357, "learning_rate": 1e-05, "loss": 0.5079, "mean_token_accuracy": 0.8619256019592285, "num_tokens": 205107581.0, "step": 5400 }, { "epoch": 0.6870627146673451, "ewc_loss": 0.10302734375, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 5.938563346862793, "learning_rate": 1e-05, "loss": 0.5274, "mean_token_accuracy": 0.8557486534118652, "num_tokens": 205144030.0, "step": 5401 }, { "epoch": 0.6871899249459357, "ewc_loss": 0.095703125, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.987418174743652, "learning_rate": 1e-05, "loss": 0.5333, "mean_token_accuracy": 0.8552844524383545, "num_tokens": 205180583.0, "step": 5402 }, { "epoch": 0.6873171352245261, "ewc_loss": 0.09814453125, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.850491523742676, "learning_rate": 1e-05, "loss": 0.5219, "mean_token_accuracy": 0.858342170715332, "num_tokens": 205215889.0, "step": 5403 }, { "epoch": 0.6874443455031166, "ewc_loss": 0.09130859375, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.708343982696533, "learning_rate": 1e-05, "loss": 0.5637, "mean_token_accuracy": 0.8389092087745667, "num_tokens": 205254725.0, "step": 5404 }, { "epoch": 0.6875715557817071, "ewc_loss": 0.0947265625, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.769857406616211, "learning_rate": 1e-05, "loss": 0.485, "mean_token_accuracy": 0.8676151037216187, "num_tokens": 205289533.0, "step": 5405 }, { "epoch": 0.6876987660602977, "ewc_loss": 0.08837890625, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.527075290679932, "learning_rate": 1e-05, "loss": 0.593, "mean_token_accuracy": 0.836239218711853, "num_tokens": 205330462.0, "step": 5406 }, { "epoch": 0.6878259763388882, "ewc_loss": 0.0927734375, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.742411136627197, "learning_rate": 1e-05, "loss": 0.5684, "mean_token_accuracy": 0.8440433740615845, "num_tokens": 205371462.0, "step": 5407 }, { "epoch": 0.6879531866174787, "ewc_loss": 0.0859375, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 7.43865966796875e-05, "grad_norm": 5.38999605178833, "learning_rate": 1e-05, "loss": 0.4387, "mean_token_accuracy": 0.8787994384765625, "num_tokens": 205404629.0, "step": 5408 }, { "epoch": 0.6880803968960693, "ewc_loss": 0.091796875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.759371280670166, "learning_rate": 1e-05, "loss": 0.545, "mean_token_accuracy": 0.8464130163192749, "num_tokens": 205440600.0, "step": 5409 }, { "epoch": 0.6882076071746597, "ewc_loss": 0.08349609375, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 7.200241088867188e-05, "grad_norm": 5.272056579589844, "learning_rate": 1e-05, "loss": 0.4984, "mean_token_accuracy": 0.860853910446167, "num_tokens": 205473696.0, "step": 5410 }, { "epoch": 0.6883348174532502, "ewc_loss": 0.09375, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.85218620300293, "learning_rate": 1e-05, "loss": 0.5584, "mean_token_accuracy": 0.8469504117965698, "num_tokens": 205519889.0, "step": 5411 }, { "epoch": 0.6884620277318407, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 7.152557373046875e-05, "grad_norm": 5.147615909576416, "learning_rate": 1e-05, "loss": 0.4552, "mean_token_accuracy": 0.8734042644500732, "num_tokens": 205552303.0, "step": 5412 }, { "epoch": 0.6885892380104313, "ewc_loss": 0.099609375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 6.063581943511963, "learning_rate": 1e-05, "loss": 0.5611, "mean_token_accuracy": 0.8434779644012451, "num_tokens": 205588918.0, "step": 5413 }, { "epoch": 0.6887164482890218, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 6.961822509765625e-05, "grad_norm": 5.009806156158447, "learning_rate": 1e-05, "loss": 0.525, "mean_token_accuracy": 0.8571337461471558, "num_tokens": 205624634.0, "step": 5414 }, { "epoch": 0.6888436585676123, "ewc_loss": 0.10595703125, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.423994541168213, "learning_rate": 1e-05, "loss": 0.5147, "mean_token_accuracy": 0.863034725189209, "num_tokens": 205668700.0, "step": 5415 }, { "epoch": 0.6889708688462027, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 7.152557373046875e-05, "grad_norm": 4.972809314727783, "learning_rate": 1e-05, "loss": 0.5224, "mean_token_accuracy": 0.8524623513221741, "num_tokens": 205711051.0, "step": 5416 }, { "epoch": 0.6890980791247933, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.892073154449463, "learning_rate": 1e-05, "loss": 0.5281, "mean_token_accuracy": 0.8620117902755737, "num_tokens": 205750349.0, "step": 5417 }, { "epoch": 0.6892252894033838, "ewc_loss": 0.0927734375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.316498756408691, "learning_rate": 1e-05, "loss": 0.4767, "mean_token_accuracy": 0.8700115084648132, "num_tokens": 205789525.0, "step": 5418 }, { "epoch": 0.6893524996819743, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 0.00011157989501953125, "grad_norm": 7.102208614349365, "learning_rate": 1e-05, "loss": 0.5477, "mean_token_accuracy": 0.8568185567855835, "num_tokens": 205827266.0, "step": 5419 }, { "epoch": 0.6894797099605648, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 5.890360355377197, "learning_rate": 1e-05, "loss": 0.5125, "mean_token_accuracy": 0.8638448715209961, "num_tokens": 205866170.0, "step": 5420 }, { "epoch": 0.6896069202391554, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.764575958251953, "learning_rate": 1e-05, "loss": 0.5065, "mean_token_accuracy": 0.8657119274139404, "num_tokens": 205905613.0, "step": 5421 }, { "epoch": 0.6897341305177458, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.06155252456665, "learning_rate": 1e-05, "loss": 0.5935, "mean_token_accuracy": 0.836463212966919, "num_tokens": 205943142.0, "step": 5422 }, { "epoch": 0.6898613407963363, "ewc_loss": 0.10400390625, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.319238185882568, "learning_rate": 1e-05, "loss": 0.5202, "mean_token_accuracy": 0.8564260005950928, "num_tokens": 205983321.0, "step": 5423 }, { "epoch": 0.6899885510749268, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 5.961499214172363, "learning_rate": 1e-05, "loss": 0.5674, "mean_token_accuracy": 0.846337080001831, "num_tokens": 206019013.0, "step": 5424 }, { "epoch": 0.6901157613535174, "ewc_loss": 0.09423828125, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.9438982009887695, "learning_rate": 1e-05, "loss": 0.5096, "mean_token_accuracy": 0.8575095534324646, "num_tokens": 206059958.0, "step": 5425 }, { "epoch": 0.6902429716321079, "ewc_loss": 0.095703125, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.840090751647949, "learning_rate": 1e-05, "loss": 0.5092, "mean_token_accuracy": 0.8607512712478638, "num_tokens": 206096536.0, "step": 5426 }, { "epoch": 0.6903701819106984, "ewc_loss": 0.0908203125, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.693633079528809, "learning_rate": 1e-05, "loss": 0.5394, "mean_token_accuracy": 0.851810872554779, "num_tokens": 206138399.0, "step": 5427 }, { "epoch": 0.6904973921892888, "ewc_loss": 0.09375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.763483047485352, "learning_rate": 1e-05, "loss": 0.4746, "mean_token_accuracy": 0.8730188012123108, "num_tokens": 206171666.0, "step": 5428 }, { "epoch": 0.6906246024678794, "ewc_loss": 0.08740234375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 7.581710815429688e-05, "grad_norm": 5.528226375579834, "learning_rate": 1e-05, "loss": 0.4929, "mean_token_accuracy": 0.8640471696853638, "num_tokens": 206212439.0, "step": 5429 }, { "epoch": 0.6907518127464699, "ewc_loss": 0.091796875, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.735750198364258, "learning_rate": 1e-05, "loss": 0.5152, "mean_token_accuracy": 0.8570801019668579, "num_tokens": 206245213.0, "step": 5430 }, { "epoch": 0.6908790230250604, "ewc_loss": 0.0859375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 7.43865966796875e-05, "grad_norm": 5.4093241691589355, "learning_rate": 1e-05, "loss": 0.5018, "mean_token_accuracy": 0.8597580790519714, "num_tokens": 206283875.0, "step": 5431 }, { "epoch": 0.691006233303651, "ewc_loss": 0.091796875, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.748063564300537, "learning_rate": 1e-05, "loss": 0.5255, "mean_token_accuracy": 0.8528802394866943, "num_tokens": 206319442.0, "step": 5432 }, { "epoch": 0.6911334435822415, "ewc_loss": 0.08349609375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 7.200241088867188e-05, "grad_norm": 5.292140007019043, "learning_rate": 1e-05, "loss": 0.5537, "mean_token_accuracy": 0.841584324836731, "num_tokens": 206358210.0, "step": 5433 }, { "epoch": 0.6912606538608319, "ewc_loss": 0.09228515625, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.8404951095581055, "learning_rate": 1e-05, "loss": 0.4755, "mean_token_accuracy": 0.8691598176956177, "num_tokens": 206390900.0, "step": 5434 }, { "epoch": 0.6913878641394224, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 7.05718994140625e-05, "grad_norm": 5.155025482177734, "learning_rate": 1e-05, "loss": 0.566, "mean_token_accuracy": 0.8395615220069885, "num_tokens": 206426296.0, "step": 5435 }, { "epoch": 0.691515074418013, "ewc_loss": 0.09765625, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 6.0531768798828125, "learning_rate": 1e-05, "loss": 0.516, "mean_token_accuracy": 0.8580628633499146, "num_tokens": 206457010.0, "step": 5436 }, { "epoch": 0.6916422846966035, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 7.009506225585938e-05, "grad_norm": 4.985357761383057, "learning_rate": 1e-05, "loss": 0.4845, "mean_token_accuracy": 0.8648031949996948, "num_tokens": 206496324.0, "step": 5437 }, { "epoch": 0.691769494975194, "ewc_loss": 0.10595703125, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.452672481536865, "learning_rate": 1e-05, "loss": 0.5617, "mean_token_accuracy": 0.850937008857727, "num_tokens": 206530507.0, "step": 5438 }, { "epoch": 0.6918967052537845, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 7.104873657226562e-05, "grad_norm": 4.96344518661499, "learning_rate": 1e-05, "loss": 0.4954, "mean_token_accuracy": 0.8643500208854675, "num_tokens": 206565123.0, "step": 5439 }, { "epoch": 0.692023915532375, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 6.957231521606445, "learning_rate": 1e-05, "loss": 0.531, "mean_token_accuracy": 0.8611323833465576, "num_tokens": 206602026.0, "step": 5440 }, { "epoch": 0.6921511258109655, "ewc_loss": 0.0927734375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.3388776779174805, "learning_rate": 1e-05, "loss": 0.5272, "mean_token_accuracy": 0.853905200958252, "num_tokens": 206643371.0, "step": 5441 }, { "epoch": 0.692278336089556, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 0.00011157989501953125, "grad_norm": 7.164401531219482, "learning_rate": 1e-05, "loss": 0.5173, "mean_token_accuracy": 0.8688218593597412, "num_tokens": 206675822.0, "step": 5442 }, { "epoch": 0.6924055463681466, "ewc_loss": 0.10791015625, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 5.957268238067627, "learning_rate": 1e-05, "loss": 0.4188, "mean_token_accuracy": 0.8941577672958374, "num_tokens": 206703973.0, "step": 5443 }, { "epoch": 0.6925327566467371, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.8060455322265625, "learning_rate": 1e-05, "loss": 0.5661, "mean_token_accuracy": 0.8523557186126709, "num_tokens": 206740645.0, "step": 5444 }, { "epoch": 0.6926599669253276, "ewc_loss": 0.10888671875, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.1010332107543945, "learning_rate": 1e-05, "loss": 0.5001, "mean_token_accuracy": 0.8667950630187988, "num_tokens": 206785620.0, "step": 5445 }, { "epoch": 0.692787177203918, "ewc_loss": 0.10400390625, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.357889175415039, "learning_rate": 1e-05, "loss": 0.5526, "mean_token_accuracy": 0.8513126373291016, "num_tokens": 206818761.0, "step": 5446 }, { "epoch": 0.6929143874825086, "ewc_loss": 0.10302734375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 5.973174571990967, "learning_rate": 1e-05, "loss": 0.5008, "mean_token_accuracy": 0.867889940738678, "num_tokens": 206857523.0, "step": 5447 }, { "epoch": 0.6930415977610991, "ewc_loss": 0.09716796875, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 6.006472110748291, "learning_rate": 1e-05, "loss": 0.5434, "mean_token_accuracy": 0.8511426448822021, "num_tokens": 206893389.0, "step": 5448 }, { "epoch": 0.6931688080396896, "ewc_loss": 0.0966796875, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.839646816253662, "learning_rate": 1e-05, "loss": 0.5453, "mean_token_accuracy": 0.8517305254936218, "num_tokens": 206924997.0, "step": 5449 }, { "epoch": 0.6932960183182801, "ewc_loss": 0.09130859375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.766384601593018, "learning_rate": 1e-05, "loss": 0.5163, "mean_token_accuracy": 0.8540270924568176, "num_tokens": 206957032.0, "step": 5450 }, { "epoch": 0.6934232285968707, "ewc_loss": 0.09228515625, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.714743137359619, "learning_rate": 1e-05, "loss": 0.5043, "mean_token_accuracy": 0.8621679544448853, "num_tokens": 206995407.0, "step": 5451 }, { "epoch": 0.6935504388754611, "ewc_loss": 0.08935546875, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.616990566253662, "learning_rate": 1e-05, "loss": 0.503, "mean_token_accuracy": 0.8640641570091248, "num_tokens": 207031563.0, "step": 5452 }, { "epoch": 0.6936776491540516, "ewc_loss": 0.08984375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.632726669311523, "learning_rate": 1e-05, "loss": 0.4883, "mean_token_accuracy": 0.8639400005340576, "num_tokens": 207064641.0, "step": 5453 }, { "epoch": 0.6938048594326421, "ewc_loss": 0.08740234375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 7.581710815429688e-05, "grad_norm": 5.546341419219971, "learning_rate": 1e-05, "loss": 0.5794, "mean_token_accuracy": 0.8411598205566406, "num_tokens": 207099474.0, "step": 5454 }, { "epoch": 0.6939320697112327, "ewc_loss": 0.08837890625, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.555258274078369, "learning_rate": 1e-05, "loss": 0.4722, "mean_token_accuracy": 0.868409276008606, "num_tokens": 207139250.0, "step": 5455 }, { "epoch": 0.6940592799898232, "ewc_loss": 0.08837890625, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.499011516571045, "learning_rate": 1e-05, "loss": 0.5037, "mean_token_accuracy": 0.8578968048095703, "num_tokens": 207179523.0, "step": 5456 }, { "epoch": 0.6941864902684137, "ewc_loss": 0.08740234375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 7.581710815429688e-05, "grad_norm": 5.501891613006592, "learning_rate": 1e-05, "loss": 0.4814, "mean_token_accuracy": 0.864032506942749, "num_tokens": 207218143.0, "step": 5457 }, { "epoch": 0.6943137005470043, "ewc_loss": 0.08642578125, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 7.486343383789062e-05, "grad_norm": 5.466022968292236, "learning_rate": 1e-05, "loss": 0.5034, "mean_token_accuracy": 0.8589143753051758, "num_tokens": 207256946.0, "step": 5458 }, { "epoch": 0.6944409108255947, "ewc_loss": 0.0869140625, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 7.534027099609375e-05, "grad_norm": 5.468910217285156, "learning_rate": 1e-05, "loss": 0.4691, "mean_token_accuracy": 0.871907114982605, "num_tokens": 207296793.0, "step": 5459 }, { "epoch": 0.6945681211041852, "ewc_loss": 0.08642578125, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 7.486343383789062e-05, "grad_norm": 5.456107139587402, "learning_rate": 1e-05, "loss": 0.539, "mean_token_accuracy": 0.8480024337768555, "num_tokens": 207333602.0, "step": 5460 }, { "epoch": 0.6946953313827757, "ewc_loss": 0.08740234375, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 7.581710815429688e-05, "grad_norm": 5.454250335693359, "learning_rate": 1e-05, "loss": 0.5329, "mean_token_accuracy": 0.8473895788192749, "num_tokens": 207375806.0, "step": 5461 }, { "epoch": 0.6948225416613663, "ewc_loss": 0.08642578125, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 7.486343383789062e-05, "grad_norm": 5.446340084075928, "learning_rate": 1e-05, "loss": 0.4503, "mean_token_accuracy": 0.8781099319458008, "num_tokens": 207407856.0, "step": 5462 }, { "epoch": 0.6949497519399568, "ewc_loss": 0.087890625, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 7.62939453125e-05, "grad_norm": 5.4593706130981445, "learning_rate": 1e-05, "loss": 0.5504, "mean_token_accuracy": 0.8433680534362793, "num_tokens": 207447000.0, "step": 5463 }, { "epoch": 0.6950769622185473, "ewc_loss": 0.08642578125, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 7.486343383789062e-05, "grad_norm": 5.43721866607666, "learning_rate": 1e-05, "loss": 0.4786, "mean_token_accuracy": 0.8679753541946411, "num_tokens": 207479860.0, "step": 5464 }, { "epoch": 0.6952041724971377, "ewc_loss": 0.0869140625, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 7.534027099609375e-05, "grad_norm": 5.446654319763184, "learning_rate": 1e-05, "loss": 0.4545, "mean_token_accuracy": 0.8725342154502869, "num_tokens": 207515454.0, "step": 5465 }, { "epoch": 0.6953313827757283, "ewc_loss": 0.08642578125, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 7.486343383789062e-05, "grad_norm": 5.407993316650391, "learning_rate": 1e-05, "loss": 0.5201, "mean_token_accuracy": 0.8537355661392212, "num_tokens": 207559259.0, "step": 5466 }, { "epoch": 0.6954585930543188, "ewc_loss": 0.08740234375, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 7.581710815429688e-05, "grad_norm": 5.467280387878418, "learning_rate": 1e-05, "loss": 0.5283, "mean_token_accuracy": 0.8488478660583496, "num_tokens": 207595909.0, "step": 5467 }, { "epoch": 0.6955858033329093, "ewc_loss": 0.0859375, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 7.43865966796875e-05, "grad_norm": 5.375601768493652, "learning_rate": 1e-05, "loss": 0.6051, "mean_token_accuracy": 0.83185875415802, "num_tokens": 207636063.0, "step": 5468 }, { "epoch": 0.6957130136114998, "ewc_loss": 0.087890625, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 7.62939453125e-05, "grad_norm": 5.496042251586914, "learning_rate": 1e-05, "loss": 0.4916, "mean_token_accuracy": 0.862138032913208, "num_tokens": 207675990.0, "step": 5469 }, { "epoch": 0.6958402238900904, "ewc_loss": 0.0859375, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 7.43865966796875e-05, "grad_norm": 5.327023506164551, "learning_rate": 1e-05, "loss": 0.5094, "mean_token_accuracy": 0.8609125018119812, "num_tokens": 207708996.0, "step": 5470 }, { "epoch": 0.6959674341686808, "ewc_loss": 0.0888671875, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.568243026733398, "learning_rate": 1e-05, "loss": 0.5738, "mean_token_accuracy": 0.8419820070266724, "num_tokens": 207750433.0, "step": 5471 }, { "epoch": 0.6960946444472713, "ewc_loss": 0.08349609375, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 7.200241088867188e-05, "grad_norm": 5.238875865936279, "learning_rate": 1e-05, "loss": 0.5274, "mean_token_accuracy": 0.8515166640281677, "num_tokens": 207792479.0, "step": 5472 }, { "epoch": 0.6962218547258618, "ewc_loss": 0.0908203125, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.702871322631836, "learning_rate": 1e-05, "loss": 0.5034, "mean_token_accuracy": 0.8594035506248474, "num_tokens": 207833420.0, "step": 5473 }, { "epoch": 0.6963490650044524, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 7.104873657226562e-05, "grad_norm": 5.069315433502197, "learning_rate": 1e-05, "loss": 0.4728, "mean_token_accuracy": 0.8695621490478516, "num_tokens": 207875933.0, "step": 5474 }, { "epoch": 0.6964762752830429, "ewc_loss": 0.095703125, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 6.021702766418457, "learning_rate": 1e-05, "loss": 0.5254, "mean_token_accuracy": 0.856282114982605, "num_tokens": 207915481.0, "step": 5475 }, { "epoch": 0.6966034855616334, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 6.866455078125e-05, "grad_norm": 4.8168816566467285, "learning_rate": 1e-05, "loss": 0.5137, "mean_token_accuracy": 0.855607271194458, "num_tokens": 207952024.0, "step": 5476 }, { "epoch": 0.6967306958402238, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.647926330566406, "learning_rate": 1e-05, "loss": 0.5745, "mean_token_accuracy": 0.8487268090248108, "num_tokens": 207990938.0, "step": 5477 }, { "epoch": 0.6968579061188144, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 6.866455078125e-05, "grad_norm": 4.717867851257324, "learning_rate": 1e-05, "loss": 0.4999, "mean_token_accuracy": 0.8564854264259338, "num_tokens": 208032492.0, "step": 5478 }, { "epoch": 0.6969851163974049, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 0.00012683868408203125, "grad_norm": 7.610788345336914, "learning_rate": 1e-05, "loss": 0.594, "mean_token_accuracy": 0.8469334840774536, "num_tokens": 208068817.0, "step": 5479 }, { "epoch": 0.6971123266759954, "ewc_loss": 0.10498046875, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 5.655506134033203, "learning_rate": 1e-05, "loss": 0.5138, "mean_token_accuracy": 0.8634288907051086, "num_tokens": 208105090.0, "step": 5480 }, { "epoch": 0.697239536954586, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 0.0001277923583984375, "grad_norm": 7.82000732421875, "learning_rate": 1e-05, "loss": 0.5807, "mean_token_accuracy": 0.8557644486427307, "num_tokens": 208136653.0, "step": 5481 }, { "epoch": 0.6973667472331765, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 0.000118255615234375, "grad_norm": 6.777585983276367, "learning_rate": 1e-05, "loss": 0.5488, "mean_token_accuracy": 0.8590992093086243, "num_tokens": 208175556.0, "step": 5482 }, { "epoch": 0.6974939575117669, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.941686630249023, "learning_rate": 1e-05, "loss": 0.5137, "mean_token_accuracy": 0.8662124276161194, "num_tokens": 208214549.0, "step": 5483 }, { "epoch": 0.6976211677903574, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.551096439361572, "learning_rate": 1e-05, "loss": 0.4831, "mean_token_accuracy": 0.8738592863082886, "num_tokens": 208242582.0, "step": 5484 }, { "epoch": 0.697748378068948, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.535701751708984, "learning_rate": 1e-05, "loss": 0.5535, "mean_token_accuracy": 0.8514554500579834, "num_tokens": 208281639.0, "step": 5485 }, { "epoch": 0.6978755883475385, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.00011777877807617188, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.105215549468994, "learning_rate": 1e-05, "loss": 0.472, "mean_token_accuracy": 0.8744106292724609, "num_tokens": 208323543.0, "step": 5486 }, { "epoch": 0.698002798626129, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.000118255615234375, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.316067218780518, "learning_rate": 1e-05, "loss": 0.5417, "mean_token_accuracy": 0.8580769300460815, "num_tokens": 208359881.0, "step": 5487 }, { "epoch": 0.6981300089047195, "ewc_loss": 0.09619140625, "ewc_loss_diag": 0.000118255615234375, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.808063507080078, "learning_rate": 1e-05, "loss": 0.546, "mean_token_accuracy": 0.8490320444107056, "num_tokens": 208404541.0, "step": 5488 }, { "epoch": 0.69825721918331, "ewc_loss": 0.10009765625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 6.252041339874268, "learning_rate": 1e-05, "loss": 0.5048, "mean_token_accuracy": 0.8628808259963989, "num_tokens": 208437731.0, "step": 5489 }, { "epoch": 0.6983844294619005, "ewc_loss": 0.08984375, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.566303253173828, "learning_rate": 1e-05, "loss": 0.5341, "mean_token_accuracy": 0.8574250936508179, "num_tokens": 208475171.0, "step": 5490 }, { "epoch": 0.698511639740491, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.300146102905273, "learning_rate": 1e-05, "loss": 0.5151, "mean_token_accuracy": 0.858558714389801, "num_tokens": 208511410.0, "step": 5491 }, { "epoch": 0.6986388500190815, "ewc_loss": 0.08544921875, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 7.390975952148438e-05, "grad_norm": 5.373476982116699, "learning_rate": 1e-05, "loss": 0.5065, "mean_token_accuracy": 0.8630161285400391, "num_tokens": 208551133.0, "step": 5492 }, { "epoch": 0.6987660602976721, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.536578178405762, "learning_rate": 1e-05, "loss": 0.5161, "mean_token_accuracy": 0.8623046875, "num_tokens": 208593394.0, "step": 5493 }, { "epoch": 0.6988932705762626, "ewc_loss": 0.08642578125, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 7.486343383789062e-05, "grad_norm": 5.338990211486816, "learning_rate": 1e-05, "loss": 0.5443, "mean_token_accuracy": 0.8477945923805237, "num_tokens": 208627160.0, "step": 5494 }, { "epoch": 0.699020480854853, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.843276500701904, "learning_rate": 1e-05, "loss": 0.5325, "mean_token_accuracy": 0.8564350605010986, "num_tokens": 208664896.0, "step": 5495 }, { "epoch": 0.6991476911334435, "ewc_loss": 0.09130859375, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.5892767906188965, "learning_rate": 1e-05, "loss": 0.5301, "mean_token_accuracy": 0.8530099391937256, "num_tokens": 208702118.0, "step": 5496 }, { "epoch": 0.6992749014120341, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.900672912597656, "learning_rate": 1e-05, "loss": 0.5681, "mean_token_accuracy": 0.8492133617401123, "num_tokens": 208741243.0, "step": 5497 }, { "epoch": 0.6994021116906246, "ewc_loss": 0.09814453125, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 6.100061416625977, "learning_rate": 1e-05, "loss": 0.496, "mean_token_accuracy": 0.8666602969169617, "num_tokens": 208778914.0, "step": 5498 }, { "epoch": 0.6995293219692151, "ewc_loss": 0.10302734375, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.384427547454834, "learning_rate": 1e-05, "loss": 0.5412, "mean_token_accuracy": 0.8568591475486755, "num_tokens": 208815683.0, "step": 5499 }, { "epoch": 0.6996565322478057, "ewc_loss": 0.09765625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 6.024111747741699, "learning_rate": 1e-05, "loss": 0.5228, "mean_token_accuracy": 0.8592017889022827, "num_tokens": 208853812.0, "step": 5500 }, { "epoch": 0.6997837425263961, "ewc_loss": 0.0908203125, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.908865451812744, "learning_rate": 1e-05, "loss": 0.482, "mean_token_accuracy": 0.8686298131942749, "num_tokens": 208894326.0, "step": 5501 }, { "epoch": 0.6999109528049866, "ewc_loss": 0.09716796875, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 6.005676746368408, "learning_rate": 1e-05, "loss": 0.4987, "mean_token_accuracy": 0.8660120964050293, "num_tokens": 208929644.0, "step": 5502 }, { "epoch": 0.7000381630835771, "ewc_loss": 0.0849609375, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 7.343292236328125e-05, "grad_norm": 5.485170364379883, "learning_rate": 1e-05, "loss": 0.5354, "mean_token_accuracy": 0.8558762073516846, "num_tokens": 208972836.0, "step": 5503 }, { "epoch": 0.7001653733621677, "ewc_loss": 0.09912109375, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 6.067504405975342, "learning_rate": 1e-05, "loss": 0.5206, "mean_token_accuracy": 0.8586808443069458, "num_tokens": 209016255.0, "step": 5504 }, { "epoch": 0.7002925836407582, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 7.05718994140625e-05, "grad_norm": 5.2111430168151855, "learning_rate": 1e-05, "loss": 0.5526, "mean_token_accuracy": 0.8458602428436279, "num_tokens": 209054576.0, "step": 5505 }, { "epoch": 0.7004197939193487, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.254890441894531, "learning_rate": 1e-05, "loss": 0.4889, "mean_token_accuracy": 0.8680558800697327, "num_tokens": 209097881.0, "step": 5506 }, { "epoch": 0.7005470041979391, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 6.914138793945312e-05, "grad_norm": 5.040671348571777, "learning_rate": 1e-05, "loss": 0.5424, "mean_token_accuracy": 0.8488000631332397, "num_tokens": 209137354.0, "step": 5507 }, { "epoch": 0.7006742144765297, "ewc_loss": 0.10888671875, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.657504081726074, "learning_rate": 1e-05, "loss": 0.5338, "mean_token_accuracy": 0.8548245429992676, "num_tokens": 209170505.0, "step": 5508 }, { "epoch": 0.7008014247551202, "ewc_loss": 0.08447265625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 7.295608520507812e-05, "grad_norm": 5.141276836395264, "learning_rate": 1e-05, "loss": 0.5255, "mean_token_accuracy": 0.8502294421195984, "num_tokens": 209207745.0, "step": 5509 }, { "epoch": 0.7009286350337107, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 7.049341678619385, "learning_rate": 1e-05, "loss": 0.5596, "mean_token_accuracy": 0.8520956635475159, "num_tokens": 209246674.0, "step": 5510 }, { "epoch": 0.7010558453123013, "ewc_loss": 0.0966796875, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.654250144958496, "learning_rate": 1e-05, "loss": 0.5113, "mean_token_accuracy": 0.8603638410568237, "num_tokens": 209289019.0, "step": 5511 }, { "epoch": 0.7011830555908918, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 6.957186222076416, "learning_rate": 1e-05, "loss": 0.5361, "mean_token_accuracy": 0.8580552339553833, "num_tokens": 209323053.0, "step": 5512 }, { "epoch": 0.7013102658694823, "ewc_loss": 0.10498046875, "ewc_loss_diag": 0.00011920928955078125, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.079197406768799, "learning_rate": 1e-05, "loss": 0.5299, "mean_token_accuracy": 0.8582574129104614, "num_tokens": 209358810.0, "step": 5513 }, { "epoch": 0.7014374761480727, "ewc_loss": 0.10693359375, "ewc_loss_diag": 0.00011920928955078125, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.483182907104492, "learning_rate": 1e-05, "loss": 0.5415, "mean_token_accuracy": 0.8603076934814453, "num_tokens": 209401868.0, "step": 5514 }, { "epoch": 0.7015646864266633, "ewc_loss": 0.10302734375, "ewc_loss_diag": 0.00011920928955078125, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.0726094245910645, "learning_rate": 1e-05, "loss": 0.5485, "mean_token_accuracy": 0.8469257950782776, "num_tokens": 209435212.0, "step": 5515 }, { "epoch": 0.7016918967052538, "ewc_loss": 0.09521484375, "ewc_loss_diag": 0.00011920928955078125, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 6.0371809005737305, "learning_rate": 1e-05, "loss": 0.4968, "mean_token_accuracy": 0.8651901483535767, "num_tokens": 209472357.0, "step": 5516 }, { "epoch": 0.7018191069838443, "ewc_loss": 0.0966796875, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.924911022186279, "learning_rate": 1e-05, "loss": 0.5667, "mean_token_accuracy": 0.8440815210342407, "num_tokens": 209512687.0, "step": 5517 }, { "epoch": 0.7019463172624348, "ewc_loss": 0.0908203125, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.752201557159424, "learning_rate": 1e-05, "loss": 0.495, "mean_token_accuracy": 0.8632144331932068, "num_tokens": 209547223.0, "step": 5518 }, { "epoch": 0.7020735275410254, "ewc_loss": 0.0927734375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.794271945953369, "learning_rate": 1e-05, "loss": 0.5079, "mean_token_accuracy": 0.8614757657051086, "num_tokens": 209583299.0, "step": 5519 }, { "epoch": 0.7022007378196158, "ewc_loss": 0.0888671875, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.606510639190674, "learning_rate": 1e-05, "loss": 0.5058, "mean_token_accuracy": 0.8589786887168884, "num_tokens": 209618792.0, "step": 5520 }, { "epoch": 0.7023279480982063, "ewc_loss": 0.08984375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.674376010894775, "learning_rate": 1e-05, "loss": 0.5284, "mean_token_accuracy": 0.8538978099822998, "num_tokens": 209654324.0, "step": 5521 }, { "epoch": 0.7024551583767968, "ewc_loss": 0.0869140625, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 7.486343383789062e-05, "grad_norm": 5.543274879455566, "learning_rate": 1e-05, "loss": 0.4606, "mean_token_accuracy": 0.871505081653595, "num_tokens": 209684963.0, "step": 5522 }, { "epoch": 0.7025823686553874, "ewc_loss": 0.0869140625, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 7.486343383789062e-05, "grad_norm": 5.564144611358643, "learning_rate": 1e-05, "loss": 0.4612, "mean_token_accuracy": 0.8738059401512146, "num_tokens": 209724197.0, "step": 5523 }, { "epoch": 0.7027095789339779, "ewc_loss": 0.087890625, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 7.581710815429688e-05, "grad_norm": 5.520678997039795, "learning_rate": 1e-05, "loss": 0.4817, "mean_token_accuracy": 0.8653766512870789, "num_tokens": 209759397.0, "step": 5524 }, { "epoch": 0.7028367892125684, "ewc_loss": 0.0869140625, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 7.486343383789062e-05, "grad_norm": 5.489304542541504, "learning_rate": 1e-05, "loss": 0.5021, "mean_token_accuracy": 0.8608782887458801, "num_tokens": 209802504.0, "step": 5525 }, { "epoch": 0.7029639994911588, "ewc_loss": 0.0869140625, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 7.534027099609375e-05, "grad_norm": 5.525824069976807, "learning_rate": 1e-05, "loss": 0.51, "mean_token_accuracy": 0.8617364168167114, "num_tokens": 209838798.0, "step": 5526 }, { "epoch": 0.7030912097697494, "ewc_loss": 0.0849609375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 7.295608520507812e-05, "grad_norm": 5.418350696563721, "learning_rate": 1e-05, "loss": 0.5282, "mean_token_accuracy": 0.8490380048751831, "num_tokens": 209880935.0, "step": 5527 }, { "epoch": 0.7032184200483399, "ewc_loss": 0.0888671875, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.552639007568359, "learning_rate": 1e-05, "loss": 0.5199, "mean_token_accuracy": 0.8522871136665344, "num_tokens": 209921514.0, "step": 5528 }, { "epoch": 0.7033456303269304, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 7.2479248046875e-05, "grad_norm": 5.335413455963135, "learning_rate": 1e-05, "loss": 0.479, "mean_token_accuracy": 0.8669500350952148, "num_tokens": 209960082.0, "step": 5529 }, { "epoch": 0.703472840605521, "ewc_loss": 0.0908203125, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.64304256439209, "learning_rate": 1e-05, "loss": 0.5047, "mean_token_accuracy": 0.8593937754631042, "num_tokens": 209996338.0, "step": 5530 }, { "epoch": 0.7036000508841115, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 7.152557373046875e-05, "grad_norm": 5.215078353881836, "learning_rate": 1e-05, "loss": 0.5443, "mean_token_accuracy": 0.8479205369949341, "num_tokens": 210039140.0, "step": 5531 }, { "epoch": 0.7037272611627019, "ewc_loss": 0.091796875, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.802911758422852, "learning_rate": 1e-05, "loss": 0.4949, "mean_token_accuracy": 0.8622395396232605, "num_tokens": 210074166.0, "step": 5532 }, { "epoch": 0.7038544714412924, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 7.009506225585938e-05, "grad_norm": 5.0449957847595215, "learning_rate": 1e-05, "loss": 0.4605, "mean_token_accuracy": 0.8721214532852173, "num_tokens": 210110295.0, "step": 5533 }, { "epoch": 0.703981681719883, "ewc_loss": 0.0986328125, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 6.135956287384033, "learning_rate": 1e-05, "loss": 0.546, "mean_token_accuracy": 0.8508726358413696, "num_tokens": 210151994.0, "step": 5534 }, { "epoch": 0.7041088919984735, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 6.914138793945312e-05, "grad_norm": 4.848921775817871, "learning_rate": 1e-05, "loss": 0.4914, "mean_token_accuracy": 0.8620628118515015, "num_tokens": 210184769.0, "step": 5535 }, { "epoch": 0.704236102277064, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.7241387367248535, "learning_rate": 1e-05, "loss": 0.5564, "mean_token_accuracy": 0.8525954484939575, "num_tokens": 210225561.0, "step": 5536 }, { "epoch": 0.7043633125556545, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 7.200241088867188e-05, "grad_norm": 4.8970627784729, "learning_rate": 1e-05, "loss": 0.466, "mean_token_accuracy": 0.8717504739761353, "num_tokens": 210265544.0, "step": 5537 }, { "epoch": 0.704490522834245, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 0.00012111663818359375, "grad_norm": 7.45520544052124, "learning_rate": 1e-05, "loss": 0.5111, "mean_token_accuracy": 0.8707952499389648, "num_tokens": 210298178.0, "step": 5538 }, { "epoch": 0.7046177331128355, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 5.752464294433594, "learning_rate": 1e-05, "loss": 0.5523, "mean_token_accuracy": 0.8502770662307739, "num_tokens": 210342391.0, "step": 5539 }, { "epoch": 0.704744943391426, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 0.00011968612670898438, "grad_norm": 7.42076301574707, "learning_rate": 1e-05, "loss": 0.5357, "mean_token_accuracy": 0.8645676374435425, "num_tokens": 210386673.0, "step": 5540 }, { "epoch": 0.7048721536700165, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 6.464804649353027, "learning_rate": 1e-05, "loss": 0.5744, "mean_token_accuracy": 0.8499680161476135, "num_tokens": 210424050.0, "step": 5541 }, { "epoch": 0.7049993639486071, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.783105373382568, "learning_rate": 1e-05, "loss": 0.5059, "mean_token_accuracy": 0.8674652576446533, "num_tokens": 210462593.0, "step": 5542 }, { "epoch": 0.7051265742271976, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.313478469848633, "learning_rate": 1e-05, "loss": 0.5302, "mean_token_accuracy": 0.8589405417442322, "num_tokens": 210496534.0, "step": 5543 }, { "epoch": 0.705253784505788, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.389201641082764, "learning_rate": 1e-05, "loss": 0.5467, "mean_token_accuracy": 0.8507438898086548, "num_tokens": 210534945.0, "step": 5544 }, { "epoch": 0.7053809947843785, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 6.0023722648620605, "learning_rate": 1e-05, "loss": 0.4891, "mean_token_accuracy": 0.8695427179336548, "num_tokens": 210576742.0, "step": 5545 }, { "epoch": 0.7055082050629691, "ewc_loss": 0.099609375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 6.15125846862793, "learning_rate": 1e-05, "loss": 0.5713, "mean_token_accuracy": 0.8404767513275146, "num_tokens": 210616248.0, "step": 5546 }, { "epoch": 0.7056354153415596, "ewc_loss": 0.0947265625, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.760052680969238, "learning_rate": 1e-05, "loss": 0.5496, "mean_token_accuracy": 0.8526613712310791, "num_tokens": 210658229.0, "step": 5547 }, { "epoch": 0.7057626256201501, "ewc_loss": 0.09765625, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 6.027304649353027, "learning_rate": 1e-05, "loss": 0.4725, "mean_token_accuracy": 0.8709421157836914, "num_tokens": 210696312.0, "step": 5548 }, { "epoch": 0.7058898358987407, "ewc_loss": 0.0888671875, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.547830581665039, "learning_rate": 1e-05, "loss": 0.5235, "mean_token_accuracy": 0.8531873226165771, "num_tokens": 210736916.0, "step": 5549 }, { "epoch": 0.7060170461773311, "ewc_loss": 0.0966796875, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 6.040028095245361, "learning_rate": 1e-05, "loss": 0.5894, "mean_token_accuracy": 0.8402586579322815, "num_tokens": 210777779.0, "step": 5550 }, { "epoch": 0.7061442564559216, "ewc_loss": 0.0849609375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 7.343292236328125e-05, "grad_norm": 5.363461494445801, "learning_rate": 1e-05, "loss": 0.5716, "mean_token_accuracy": 0.8392233848571777, "num_tokens": 210823134.0, "step": 5551 }, { "epoch": 0.7062714667345121, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.179249286651611, "learning_rate": 1e-05, "loss": 0.5322, "mean_token_accuracy": 0.8549179434776306, "num_tokens": 210862792.0, "step": 5552 }, { "epoch": 0.7063986770131027, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 7.152557373046875e-05, "grad_norm": 5.2005205154418945, "learning_rate": 1e-05, "loss": 0.4869, "mean_token_accuracy": 0.863265872001648, "num_tokens": 210905411.0, "step": 5553 }, { "epoch": 0.7065258872916932, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.513523578643799, "learning_rate": 1e-05, "loss": 0.5568, "mean_token_accuracy": 0.848957896232605, "num_tokens": 210944090.0, "step": 5554 }, { "epoch": 0.7066530975702837, "ewc_loss": 0.0859375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 7.43865966796875e-05, "grad_norm": 5.216403484344482, "learning_rate": 1e-05, "loss": 0.4885, "mean_token_accuracy": 0.8564679622650146, "num_tokens": 210977372.0, "step": 5555 }, { "epoch": 0.7067803078488741, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.845312595367432, "learning_rate": 1e-05, "loss": 0.5834, "mean_token_accuracy": 0.8486807346343994, "num_tokens": 211021800.0, "step": 5556 }, { "epoch": 0.7069075181274647, "ewc_loss": 0.0927734375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.549861431121826, "learning_rate": 1e-05, "loss": 0.5377, "mean_token_accuracy": 0.8543666005134583, "num_tokens": 211063568.0, "step": 5557 }, { "epoch": 0.7070347284060552, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.854207992553711, "learning_rate": 1e-05, "loss": 0.5236, "mean_token_accuracy": 0.8605188131332397, "num_tokens": 211100422.0, "step": 5558 }, { "epoch": 0.7071619386846457, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 5.956179618835449, "learning_rate": 1e-05, "loss": 0.5517, "mean_token_accuracy": 0.8512792587280273, "num_tokens": 211138644.0, "step": 5559 }, { "epoch": 0.7072891489632362, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.409450531005859, "learning_rate": 1e-05, "loss": 0.5325, "mean_token_accuracy": 0.8548178672790527, "num_tokens": 211173033.0, "step": 5560 }, { "epoch": 0.7074163592418268, "ewc_loss": 0.099609375, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 6.055586338043213, "learning_rate": 1e-05, "loss": 0.4963, "mean_token_accuracy": 0.8662905693054199, "num_tokens": 211212902.0, "step": 5561 }, { "epoch": 0.7075435695204173, "ewc_loss": 0.0908203125, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.8608317375183105, "learning_rate": 1e-05, "loss": 0.5778, "mean_token_accuracy": 0.8369007110595703, "num_tokens": 211258812.0, "step": 5562 }, { "epoch": 0.7076707797990077, "ewc_loss": 0.0986328125, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 6.071282863616943, "learning_rate": 1e-05, "loss": 0.526, "mean_token_accuracy": 0.8520317077636719, "num_tokens": 211296496.0, "step": 5563 }, { "epoch": 0.7077979900775982, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.0001201629638671875, "ewc_loss_parallel": 7.200241088867188e-05, "grad_norm": 5.407336235046387, "learning_rate": 1e-05, "loss": 0.511, "mean_token_accuracy": 0.8541339635848999, "num_tokens": 211337081.0, "step": 5564 }, { "epoch": 0.7079252003561888, "ewc_loss": 0.10107421875, "ewc_loss_diag": 0.0001201629638671875, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.237502574920654, "learning_rate": 1e-05, "loss": 0.5624, "mean_token_accuracy": 0.8476166129112244, "num_tokens": 211373845.0, "step": 5565 }, { "epoch": 0.7080524106347793, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0001201629638671875, "ewc_loss_parallel": 6.961822509765625e-05, "grad_norm": 5.134446620941162, "learning_rate": 1e-05, "loss": 0.512, "mean_token_accuracy": 0.8554456233978271, "num_tokens": 211412994.0, "step": 5566 }, { "epoch": 0.7081796209133698, "ewc_loss": 0.10791015625, "ewc_loss_diag": 0.0001201629638671875, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.5800886154174805, "learning_rate": 1e-05, "loss": 0.5657, "mean_token_accuracy": 0.8480185270309448, "num_tokens": 211456678.0, "step": 5567 }, { "epoch": 0.7083068311919604, "ewc_loss": 0.08447265625, "ewc_loss_diag": 0.0001201629638671875, "ewc_loss_parallel": 7.2479248046875e-05, "grad_norm": 5.150777816772461, "learning_rate": 1e-05, "loss": 0.5332, "mean_token_accuracy": 0.8500812649726868, "num_tokens": 211497040.0, "step": 5568 }, { "epoch": 0.7084340414705508, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.0001201629638671875, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.944403648376465, "learning_rate": 1e-05, "loss": 0.5619, "mean_token_accuracy": 0.8479138612747192, "num_tokens": 211534113.0, "step": 5569 }, { "epoch": 0.7085612517491413, "ewc_loss": 0.0947265625, "ewc_loss_diag": 0.0001201629638671875, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.571641445159912, "learning_rate": 1e-05, "loss": 0.4931, "mean_token_accuracy": 0.8642663955688477, "num_tokens": 211569518.0, "step": 5570 }, { "epoch": 0.7086884620277318, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.0001201629638671875, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.904377460479736, "learning_rate": 1e-05, "loss": 0.5084, "mean_token_accuracy": 0.8693875074386597, "num_tokens": 211607918.0, "step": 5571 }, { "epoch": 0.7088156723063224, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.0001201629638671875, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 5.954342842102051, "learning_rate": 1e-05, "loss": 0.5727, "mean_token_accuracy": 0.8461002111434937, "num_tokens": 211647222.0, "step": 5572 }, { "epoch": 0.7089428825849129, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.0001201629638671875, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.52908182144165, "learning_rate": 1e-05, "loss": 0.5027, "mean_token_accuracy": 0.8659752607345581, "num_tokens": 211684584.0, "step": 5573 }, { "epoch": 0.7090700928635034, "ewc_loss": 0.09814453125, "ewc_loss_diag": 0.0001201629638671875, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.975862503051758, "learning_rate": 1e-05, "loss": 0.5163, "mean_token_accuracy": 0.8614320158958435, "num_tokens": 211718968.0, "step": 5574 }, { "epoch": 0.7091973031420938, "ewc_loss": 0.0986328125, "ewc_loss_diag": 0.0001201629638671875, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 6.168379306793213, "learning_rate": 1e-05, "loss": 0.5015, "mean_token_accuracy": 0.8607537746429443, "num_tokens": 211756230.0, "step": 5575 }, { "epoch": 0.7093245134206844, "ewc_loss": 0.09423828125, "ewc_loss_diag": 0.0001201629638671875, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.831255912780762, "learning_rate": 1e-05, "loss": 0.4531, "mean_token_accuracy": 0.8756442070007324, "num_tokens": 211794155.0, "step": 5576 }, { "epoch": 0.7094517236992749, "ewc_loss": 0.09326171875, "ewc_loss_diag": 0.0001201629638671875, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.9379119873046875, "learning_rate": 1e-05, "loss": 0.5252, "mean_token_accuracy": 0.8569815158843994, "num_tokens": 211834483.0, "step": 5577 }, { "epoch": 0.7095789339778654, "ewc_loss": 0.09033203125, "ewc_loss_diag": 0.0001201629638671875, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.6809539794921875, "learning_rate": 1e-05, "loss": 0.5487, "mean_token_accuracy": 0.8477660417556763, "num_tokens": 211873369.0, "step": 5578 }, { "epoch": 0.709706144256456, "ewc_loss": 0.091796875, "ewc_loss_diag": 0.0001201629638671875, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.856645107269287, "learning_rate": 1e-05, "loss": 0.5621, "mean_token_accuracy": 0.8474298715591431, "num_tokens": 211911985.0, "step": 5579 }, { "epoch": 0.7098333545350465, "ewc_loss": 0.0869140625, "ewc_loss_diag": 0.0001201629638671875, "ewc_loss_parallel": 7.486343383789062e-05, "grad_norm": 5.480196952819824, "learning_rate": 1e-05, "loss": 0.4908, "mean_token_accuracy": 0.8616278171539307, "num_tokens": 211953739.0, "step": 5580 }, { "epoch": 0.7099605648136369, "ewc_loss": 0.09375, "ewc_loss_diag": 0.0001201629638671875, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.909717559814453, "learning_rate": 1e-05, "loss": 0.4756, "mean_token_accuracy": 0.8679262399673462, "num_tokens": 211988817.0, "step": 5581 }, { "epoch": 0.7100877750922274, "ewc_loss": 0.08349609375, "ewc_loss_diag": 0.0001201629638671875, "ewc_loss_parallel": 7.152557373046875e-05, "grad_norm": 5.300075531005859, "learning_rate": 1e-05, "loss": 0.5537, "mean_token_accuracy": 0.8473471403121948, "num_tokens": 212030891.0, "step": 5582 }, { "epoch": 0.710214985370818, "ewc_loss": 0.09912109375, "ewc_loss_diag": 0.0001201629638671875, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 6.080599784851074, "learning_rate": 1e-05, "loss": 0.493, "mean_token_accuracy": 0.8699036836624146, "num_tokens": 212065727.0, "step": 5583 }, { "epoch": 0.7103421956494085, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.0001201629638671875, "ewc_loss_parallel": 7.009506225585938e-05, "grad_norm": 5.165934085845947, "learning_rate": 1e-05, "loss": 0.5241, "mean_token_accuracy": 0.8556373119354248, "num_tokens": 212098718.0, "step": 5584 }, { "epoch": 0.710469405927999, "ewc_loss": 0.10400390625, "ewc_loss_diag": 0.0001201629638671875, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.361140727996826, "learning_rate": 1e-05, "loss": 0.5306, "mean_token_accuracy": 0.8556609749794006, "num_tokens": 212142597.0, "step": 5585 }, { "epoch": 0.7105966162065895, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0001201629638671875, "ewc_loss_parallel": 7.05718994140625e-05, "grad_norm": 5.141783237457275, "learning_rate": 1e-05, "loss": 0.5707, "mean_token_accuracy": 0.8336151242256165, "num_tokens": 212182878.0, "step": 5586 }, { "epoch": 0.71072382648518, "ewc_loss": 0.10986328125, "ewc_loss_diag": 0.0001201629638671875, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.685354709625244, "learning_rate": 1e-05, "loss": 0.5017, "mean_token_accuracy": 0.8654224276542664, "num_tokens": 212216759.0, "step": 5587 }, { "epoch": 0.7108510367637705, "ewc_loss": 0.08935546875, "ewc_loss_diag": 0.0001201629638671875, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.393682956695557, "learning_rate": 1e-05, "loss": 0.5351, "mean_token_accuracy": 0.8505695462226868, "num_tokens": 212253559.0, "step": 5588 }, { "epoch": 0.710978247042361, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.0001201629638671875, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.745731830596924, "learning_rate": 1e-05, "loss": 0.5165, "mean_token_accuracy": 0.8612352013587952, "num_tokens": 212296821.0, "step": 5589 }, { "epoch": 0.7111054573209515, "ewc_loss": 0.095703125, "ewc_loss_diag": 0.0001201629638671875, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.755825996398926, "learning_rate": 1e-05, "loss": 0.4995, "mean_token_accuracy": 0.8643662333488464, "num_tokens": 212334872.0, "step": 5590 }, { "epoch": 0.7112326675995421, "ewc_loss": 0.10595703125, "ewc_loss_diag": 0.0001201629638671875, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.428037166595459, "learning_rate": 1e-05, "loss": 0.5108, "mean_token_accuracy": 0.8622584342956543, "num_tokens": 212375692.0, "step": 5591 }, { "epoch": 0.7113598778781326, "ewc_loss": 0.09912109375, "ewc_loss_diag": 0.0001201629638671875, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.9287261962890625, "learning_rate": 1e-05, "loss": 0.5378, "mean_token_accuracy": 0.8494418859481812, "num_tokens": 212407301.0, "step": 5592 }, { "epoch": 0.711487088156723, "ewc_loss": 0.0927734375, "ewc_loss_diag": 0.0001201629638671875, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.92210578918457, "learning_rate": 1e-05, "loss": 0.5044, "mean_token_accuracy": 0.8595958948135376, "num_tokens": 212443831.0, "step": 5593 }, { "epoch": 0.7116142984353135, "ewc_loss": 0.09814453125, "ewc_loss_diag": 0.0001201629638671875, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.973647117614746, "learning_rate": 1e-05, "loss": 0.4808, "mean_token_accuracy": 0.8669356107711792, "num_tokens": 212477595.0, "step": 5594 }, { "epoch": 0.7117415087139041, "ewc_loss": 0.08642578125, "ewc_loss_diag": 0.0001201629638671875, "ewc_loss_parallel": 7.43865966796875e-05, "grad_norm": 5.470407962799072, "learning_rate": 1e-05, "loss": 0.5729, "mean_token_accuracy": 0.8392724990844727, "num_tokens": 212513786.0, "step": 5595 }, { "epoch": 0.7118687189924946, "ewc_loss": 0.10009765625, "ewc_loss_diag": 0.0001201629638671875, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 6.13034725189209, "learning_rate": 1e-05, "loss": 0.5224, "mean_token_accuracy": 0.8547407388687134, "num_tokens": 212548111.0, "step": 5596 }, { "epoch": 0.7119959292710851, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0001201629638671875, "ewc_loss_parallel": 7.05718994140625e-05, "grad_norm": 5.161717414855957, "learning_rate": 1e-05, "loss": 0.5291, "mean_token_accuracy": 0.8530020117759705, "num_tokens": 212585395.0, "step": 5597 }, { "epoch": 0.7121231395496757, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.435266017913818, "learning_rate": 1e-05, "loss": 0.5204, "mean_token_accuracy": 0.8687576055526733, "num_tokens": 212625351.0, "step": 5598 }, { "epoch": 0.7122503498282661, "ewc_loss": 0.08349609375, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 7.152557373046875e-05, "grad_norm": 5.116783618927002, "learning_rate": 1e-05, "loss": 0.516, "mean_token_accuracy": 0.8509048223495483, "num_tokens": 212657689.0, "step": 5599 }, { "epoch": 0.7123775601068566, "ewc_loss": 0.11376953125, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.7643141746521, "learning_rate": 1e-05, "loss": 0.4787, "mean_token_accuracy": 0.8728014230728149, "num_tokens": 212694741.0, "step": 5600 }, { "epoch": 0.7125047703854471, "ewc_loss": 0.0908203125, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.395530700683594, "learning_rate": 1e-05, "loss": 0.5253, "mean_token_accuracy": 0.8541461825370789, "num_tokens": 212727109.0, "step": 5601 }, { "epoch": 0.7126319806640377, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.851447105407715, "learning_rate": 1e-05, "loss": 0.5804, "mean_token_accuracy": 0.8457591533660889, "num_tokens": 212763031.0, "step": 5602 }, { "epoch": 0.7127591909426282, "ewc_loss": 0.0986328125, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.763631343841553, "learning_rate": 1e-05, "loss": 0.5021, "mean_token_accuracy": 0.8616260290145874, "num_tokens": 212803601.0, "step": 5603 }, { "epoch": 0.7128864012212187, "ewc_loss": 0.10888671875, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.58205509185791, "learning_rate": 1e-05, "loss": 0.4795, "mean_token_accuracy": 0.8731884360313416, "num_tokens": 212848331.0, "step": 5604 }, { "epoch": 0.7130136114998091, "ewc_loss": 0.09912109375, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.880557537078857, "learning_rate": 1e-05, "loss": 0.5375, "mean_token_accuracy": 0.8550190329551697, "num_tokens": 212890364.0, "step": 5605 }, { "epoch": 0.7131408217783997, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 6.236605167388916, "learning_rate": 1e-05, "loss": 0.5028, "mean_token_accuracy": 0.8614919185638428, "num_tokens": 212931234.0, "step": 5606 }, { "epoch": 0.7132680320569902, "ewc_loss": 0.095703125, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.808417320251465, "learning_rate": 1e-05, "loss": 0.5109, "mean_token_accuracy": 0.8596454858779907, "num_tokens": 212970103.0, "step": 5607 }, { "epoch": 0.7133952423355807, "ewc_loss": 0.09521484375, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.981626510620117, "learning_rate": 1e-05, "loss": 0.5367, "mean_token_accuracy": 0.8523765802383423, "num_tokens": 213004895.0, "step": 5608 }, { "epoch": 0.7135224526141712, "ewc_loss": 0.091796875, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.686864852905273, "learning_rate": 1e-05, "loss": 0.5016, "mean_token_accuracy": 0.8608189821243286, "num_tokens": 213044263.0, "step": 5609 }, { "epoch": 0.7136496628927618, "ewc_loss": 0.09375, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.858434200286865, "learning_rate": 1e-05, "loss": 0.5252, "mean_token_accuracy": 0.8540768623352051, "num_tokens": 213081685.0, "step": 5610 }, { "epoch": 0.7137768731713523, "ewc_loss": 0.0888671875, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.535507678985596, "learning_rate": 1e-05, "loss": 0.5067, "mean_token_accuracy": 0.8609011173248291, "num_tokens": 213116986.0, "step": 5611 }, { "epoch": 0.7139040834499427, "ewc_loss": 0.09375, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.8442158699035645, "learning_rate": 1e-05, "loss": 0.4797, "mean_token_accuracy": 0.8677929043769836, "num_tokens": 213156313.0, "step": 5612 }, { "epoch": 0.7140312937285332, "ewc_loss": 0.0859375, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 7.390975952148438e-05, "grad_norm": 5.3922119140625, "learning_rate": 1e-05, "loss": 0.5525, "mean_token_accuracy": 0.8461521863937378, "num_tokens": 213187216.0, "step": 5613 }, { "epoch": 0.7141585040071238, "ewc_loss": 0.09619140625, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.922675132751465, "learning_rate": 1e-05, "loss": 0.5248, "mean_token_accuracy": 0.853110671043396, "num_tokens": 213223380.0, "step": 5614 }, { "epoch": 0.7142857142857143, "ewc_loss": 0.08349609375, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 7.152557373046875e-05, "grad_norm": 5.241433143615723, "learning_rate": 1e-05, "loss": 0.4634, "mean_token_accuracy": 0.8722211122512817, "num_tokens": 213259100.0, "step": 5615 }, { "epoch": 0.7144129245643048, "ewc_loss": 0.10009765625, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 6.105066299438477, "learning_rate": 1e-05, "loss": 0.5504, "mean_token_accuracy": 0.8492389917373657, "num_tokens": 213300103.0, "step": 5616 }, { "epoch": 0.7145401348428954, "ewc_loss": 0.08349609375, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 7.104873657226562e-05, "grad_norm": 5.140977382659912, "learning_rate": 1e-05, "loss": 0.4733, "mean_token_accuracy": 0.8681955337524414, "num_tokens": 213338283.0, "step": 5617 }, { "epoch": 0.7146673451214858, "ewc_loss": 0.10498046875, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.3834333419799805, "learning_rate": 1e-05, "loss": 0.5118, "mean_token_accuracy": 0.8640034198760986, "num_tokens": 213382211.0, "step": 5618 }, { "epoch": 0.7147945554000763, "ewc_loss": 0.0859375, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 7.390975952148438e-05, "grad_norm": 5.160128116607666, "learning_rate": 1e-05, "loss": 0.5114, "mean_token_accuracy": 0.8594400882720947, "num_tokens": 213418790.0, "step": 5619 }, { "epoch": 0.7149217656786668, "ewc_loss": 0.11083984375, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.6685638427734375, "learning_rate": 1e-05, "loss": 0.5219, "mean_token_accuracy": 0.858951210975647, "num_tokens": 213457843.0, "step": 5620 }, { "epoch": 0.7150489759572574, "ewc_loss": 0.091796875, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.41715145111084, "learning_rate": 1e-05, "loss": 0.5189, "mean_token_accuracy": 0.8563182353973389, "num_tokens": 213496182.0, "step": 5621 }, { "epoch": 0.7151761862358479, "ewc_loss": 0.11181640625, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.669048309326172, "learning_rate": 1e-05, "loss": 0.4641, "mean_token_accuracy": 0.8753383755683899, "num_tokens": 213535254.0, "step": 5622 }, { "epoch": 0.7153033965144384, "ewc_loss": 0.09716796875, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.737457275390625, "learning_rate": 1e-05, "loss": 0.4949, "mean_token_accuracy": 0.8609246611595154, "num_tokens": 213574331.0, "step": 5623 }, { "epoch": 0.7154306067930288, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.328113079071045, "learning_rate": 1e-05, "loss": 0.5122, "mean_token_accuracy": 0.8568768501281738, "num_tokens": 213608185.0, "step": 5624 }, { "epoch": 0.7155578170716194, "ewc_loss": 0.0986328125, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.882378578186035, "learning_rate": 1e-05, "loss": 0.4998, "mean_token_accuracy": 0.866076648235321, "num_tokens": 213642032.0, "step": 5625 }, { "epoch": 0.7156850273502099, "ewc_loss": 0.09228515625, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.847084999084473, "learning_rate": 1e-05, "loss": 0.485, "mean_token_accuracy": 0.8635628819465637, "num_tokens": 213674652.0, "step": 5626 }, { "epoch": 0.7158122376288004, "ewc_loss": 0.0986328125, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.949770927429199, "learning_rate": 1e-05, "loss": 0.5018, "mean_token_accuracy": 0.8656082153320312, "num_tokens": 213711112.0, "step": 5627 }, { "epoch": 0.715939447907391, "ewc_loss": 0.0849609375, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 7.295608520507812e-05, "grad_norm": 5.418667793273926, "learning_rate": 1e-05, "loss": 0.531, "mean_token_accuracy": 0.8495590686798096, "num_tokens": 213750319.0, "step": 5628 }, { "epoch": 0.7160666581859815, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 6.133552074432373, "learning_rate": 1e-05, "loss": 0.4953, "mean_token_accuracy": 0.864484965801239, "num_tokens": 213788408.0, "step": 5629 }, { "epoch": 0.7161938684645719, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 7.009506225585938e-05, "grad_norm": 5.139458179473877, "learning_rate": 1e-05, "loss": 0.5709, "mean_token_accuracy": 0.8381856083869934, "num_tokens": 213826822.0, "step": 5630 }, { "epoch": 0.7163210787431624, "ewc_loss": 0.10595703125, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.467830181121826, "learning_rate": 1e-05, "loss": 0.5996, "mean_token_accuracy": 0.8421099781990051, "num_tokens": 213864046.0, "step": 5631 }, { "epoch": 0.716448289021753, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 7.200241088867188e-05, "grad_norm": 5.107737064361572, "learning_rate": 1e-05, "loss": 0.4927, "mean_token_accuracy": 0.8633488416671753, "num_tokens": 213907235.0, "step": 5632 }, { "epoch": 0.7165754993003435, "ewc_loss": 0.11474609375, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.814791679382324, "learning_rate": 1e-05, "loss": 0.5271, "mean_token_accuracy": 0.8623179197311401, "num_tokens": 213942126.0, "step": 5633 }, { "epoch": 0.716702709578934, "ewc_loss": 0.0927734375, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.433560848236084, "learning_rate": 1e-05, "loss": 0.5066, "mean_token_accuracy": 0.8581753373146057, "num_tokens": 213980486.0, "step": 5634 }, { "epoch": 0.7168299198575245, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.8536810874938965, "learning_rate": 1e-05, "loss": 0.5084, "mean_token_accuracy": 0.8697521686553955, "num_tokens": 214016991.0, "step": 5635 }, { "epoch": 0.716957130136115, "ewc_loss": 0.10107421875, "ewc_loss_diag": 0.00012111663818359375, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 5.8028764724731445, "learning_rate": 1e-05, "loss": 0.5134, "mean_token_accuracy": 0.863503634929657, "num_tokens": 214049424.0, "step": 5636 }, { "epoch": 0.7170843404147055, "ewc_loss": 0.109375, "ewc_loss_diag": 0.00012111663818359375, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.557042598724365, "learning_rate": 1e-05, "loss": 0.5362, "mean_token_accuracy": 0.8591010570526123, "num_tokens": 214090422.0, "step": 5637 }, { "epoch": 0.717211550693296, "ewc_loss": 0.10009765625, "ewc_loss_diag": 0.00012111663818359375, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 5.889064311981201, "learning_rate": 1e-05, "loss": 0.4781, "mean_token_accuracy": 0.8674896359443665, "num_tokens": 214123075.0, "step": 5638 }, { "epoch": 0.7173387609718865, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.00012111663818359375, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 6.214752674102783, "learning_rate": 1e-05, "loss": 0.6113, "mean_token_accuracy": 0.8387746214866638, "num_tokens": 214158962.0, "step": 5639 }, { "epoch": 0.7174659712504771, "ewc_loss": 0.0966796875, "ewc_loss_diag": 0.00012159347534179688, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.804305553436279, "learning_rate": 1e-05, "loss": 0.4934, "mean_token_accuracy": 0.8656518459320068, "num_tokens": 214201867.0, "step": 5640 }, { "epoch": 0.7175931815290676, "ewc_loss": 0.09619140625, "ewc_loss_diag": 0.00012159347534179688, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.947967529296875, "learning_rate": 1e-05, "loss": 0.4866, "mean_token_accuracy": 0.8670369982719421, "num_tokens": 214241300.0, "step": 5641 }, { "epoch": 0.717720391807658, "ewc_loss": 0.091796875, "ewc_loss_diag": 0.00012159347534179688, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.687772274017334, "learning_rate": 1e-05, "loss": 0.5272, "mean_token_accuracy": 0.8535915017127991, "num_tokens": 214276455.0, "step": 5642 }, { "epoch": 0.7178476020862485, "ewc_loss": 0.09326171875, "ewc_loss_diag": 0.00012159347534179688, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.81825065612793, "learning_rate": 1e-05, "loss": 0.4932, "mean_token_accuracy": 0.8624826073646545, "num_tokens": 214313737.0, "step": 5643 }, { "epoch": 0.7179748123648391, "ewc_loss": 0.0888671875, "ewc_loss_diag": 0.00012111663818359375, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.543779373168945, "learning_rate": 1e-05, "loss": 0.4948, "mean_token_accuracy": 0.8605832457542419, "num_tokens": 214351179.0, "step": 5644 }, { "epoch": 0.7181020226434296, "ewc_loss": 0.09326171875, "ewc_loss_diag": 0.00012111663818359375, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.794938087463379, "learning_rate": 1e-05, "loss": 0.471, "mean_token_accuracy": 0.8727641105651855, "num_tokens": 214387606.0, "step": 5645 }, { "epoch": 0.7182292329220201, "ewc_loss": 0.0869140625, "ewc_loss_diag": 0.00012111663818359375, "ewc_loss_parallel": 7.486343383789062e-05, "grad_norm": 5.3929901123046875, "learning_rate": 1e-05, "loss": 0.543, "mean_token_accuracy": 0.8507663607597351, "num_tokens": 214428149.0, "step": 5646 }, { "epoch": 0.7183564432006107, "ewc_loss": 0.0947265625, "ewc_loss_diag": 0.00012111663818359375, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.884198188781738, "learning_rate": 1e-05, "loss": 0.4908, "mean_token_accuracy": 0.8654847145080566, "num_tokens": 214468878.0, "step": 5647 }, { "epoch": 0.7184836534792011, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.00012111663818359375, "ewc_loss_parallel": 7.200241088867188e-05, "grad_norm": 5.251669406890869, "learning_rate": 1e-05, "loss": 0.5352, "mean_token_accuracy": 0.850121259689331, "num_tokens": 214508813.0, "step": 5648 }, { "epoch": 0.7186108637577916, "ewc_loss": 0.0986328125, "ewc_loss_diag": 0.00012111663818359375, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 6.053431987762451, "learning_rate": 1e-05, "loss": 0.5157, "mean_token_accuracy": 0.8602768778800964, "num_tokens": 214553896.0, "step": 5649 }, { "epoch": 0.7187380740363821, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.00012159347534179688, "ewc_loss_parallel": 7.05718994140625e-05, "grad_norm": 5.130837440490723, "learning_rate": 1e-05, "loss": 0.4826, "mean_token_accuracy": 0.8638297319412231, "num_tokens": 214593727.0, "step": 5650 }, { "epoch": 0.7188652843149727, "ewc_loss": 0.10498046875, "ewc_loss_diag": 0.00012159347534179688, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.35194206237793, "learning_rate": 1e-05, "loss": 0.5235, "mean_token_accuracy": 0.8605438470840454, "num_tokens": 214627029.0, "step": 5651 }, { "epoch": 0.7189924945935632, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.00012159347534179688, "ewc_loss_parallel": 7.200241088867188e-05, "grad_norm": 5.123815059661865, "learning_rate": 1e-05, "loss": 0.5581, "mean_token_accuracy": 0.8439765572547913, "num_tokens": 214665484.0, "step": 5652 }, { "epoch": 0.7191197048721537, "ewc_loss": 0.11181640625, "ewc_loss_diag": 0.00012159347534179688, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.6652913093566895, "learning_rate": 1e-05, "loss": 0.5084, "mean_token_accuracy": 0.8668887615203857, "num_tokens": 214708654.0, "step": 5653 }, { "epoch": 0.7192469151507441, "ewc_loss": 0.0908203125, "ewc_loss_diag": 0.00012111663818359375, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.337288856506348, "learning_rate": 1e-05, "loss": 0.5259, "mean_token_accuracy": 0.8539973497390747, "num_tokens": 214746962.0, "step": 5654 }, { "epoch": 0.7193741254293347, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.00012111663818359375, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.777608394622803, "learning_rate": 1e-05, "loss": 0.5302, "mean_token_accuracy": 0.8576996326446533, "num_tokens": 214784691.0, "step": 5655 }, { "epoch": 0.7195013357079252, "ewc_loss": 0.0966796875, "ewc_loss_diag": 0.00012159347534179688, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.709682464599609, "learning_rate": 1e-05, "loss": 0.486, "mean_token_accuracy": 0.8680837750434875, "num_tokens": 214817182.0, "step": 5656 }, { "epoch": 0.7196285459865157, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.00012159347534179688, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.492520809173584, "learning_rate": 1e-05, "loss": 0.5679, "mean_token_accuracy": 0.8437641859054565, "num_tokens": 214857053.0, "step": 5657 }, { "epoch": 0.7197557562651062, "ewc_loss": 0.10009765625, "ewc_loss_diag": 0.00012159347534179688, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 5.8895463943481445, "learning_rate": 1e-05, "loss": 0.6265, "mean_token_accuracy": 0.8344272971153259, "num_tokens": 214894908.0, "step": 5658 }, { "epoch": 0.7198829665436968, "ewc_loss": 0.09765625, "ewc_loss_diag": 0.00012159347534179688, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 6.0303168296813965, "learning_rate": 1e-05, "loss": 0.5218, "mean_token_accuracy": 0.8562507629394531, "num_tokens": 214930235.0, "step": 5659 }, { "epoch": 0.7200101768222873, "ewc_loss": 0.09912109375, "ewc_loss_diag": 0.00012159347534179688, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.921074867248535, "learning_rate": 1e-05, "loss": 0.4941, "mean_token_accuracy": 0.8670646548271179, "num_tokens": 214967140.0, "step": 5660 }, { "epoch": 0.7201373871008777, "ewc_loss": 0.08935546875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.585793972015381, "learning_rate": 1e-05, "loss": 0.5123, "mean_token_accuracy": 0.8578369617462158, "num_tokens": 215004039.0, "step": 5661 }, { "epoch": 0.7202645973794682, "ewc_loss": 0.10009765625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 6.031120777130127, "learning_rate": 1e-05, "loss": 0.4807, "mean_token_accuracy": 0.8685213327407837, "num_tokens": 215039305.0, "step": 5662 }, { "epoch": 0.7203918076580588, "ewc_loss": 0.08447265625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 7.2479248046875e-05, "grad_norm": 5.260735988616943, "learning_rate": 1e-05, "loss": 0.4752, "mean_token_accuracy": 0.8686069250106812, "num_tokens": 215076455.0, "step": 5663 }, { "epoch": 0.7205190179366493, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.2609381675720215, "learning_rate": 1e-05, "loss": 0.5579, "mean_token_accuracy": 0.8495728969573975, "num_tokens": 215116826.0, "step": 5664 }, { "epoch": 0.7206462282152398, "ewc_loss": 0.08349609375, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 7.104873657226562e-05, "grad_norm": 5.113327980041504, "learning_rate": 1e-05, "loss": 0.531, "mean_token_accuracy": 0.849217414855957, "num_tokens": 215162302.0, "step": 5665 }, { "epoch": 0.7207734384938304, "ewc_loss": 0.109375, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.585360050201416, "learning_rate": 1e-05, "loss": 0.5447, "mean_token_accuracy": 0.8525069952011108, "num_tokens": 215192402.0, "step": 5666 }, { "epoch": 0.7209006487724208, "ewc_loss": 0.087890625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 7.581710815429688e-05, "grad_norm": 5.229451656341553, "learning_rate": 1e-05, "loss": 0.5465, "mean_token_accuracy": 0.8529750108718872, "num_tokens": 215228521.0, "step": 5667 }, { "epoch": 0.7210278590510113, "ewc_loss": 0.11474609375, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.806475639343262, "learning_rate": 1e-05, "loss": 0.5301, "mean_token_accuracy": 0.8598914742469788, "num_tokens": 215264997.0, "step": 5668 }, { "epoch": 0.7211550693296018, "ewc_loss": 0.0947265625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.5489606857299805, "learning_rate": 1e-05, "loss": 0.4445, "mean_token_accuracy": 0.879449725151062, "num_tokens": 215301997.0, "step": 5669 }, { "epoch": 0.7212822796081924, "ewc_loss": 0.11376953125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.720086574554443, "learning_rate": 1e-05, "loss": 0.5617, "mean_token_accuracy": 0.851540207862854, "num_tokens": 215343637.0, "step": 5670 }, { "epoch": 0.7214094898867829, "ewc_loss": 0.099609375, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 5.794978618621826, "learning_rate": 1e-05, "loss": 0.5049, "mean_token_accuracy": 0.8609008193016052, "num_tokens": 215381966.0, "step": 5671 }, { "epoch": 0.7215367001653734, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.428812503814697, "learning_rate": 1e-05, "loss": 0.539, "mean_token_accuracy": 0.8558078408241272, "num_tokens": 215422930.0, "step": 5672 }, { "epoch": 0.7216639104439638, "ewc_loss": 0.09814453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.8273515701293945, "learning_rate": 1e-05, "loss": 0.4792, "mean_token_accuracy": 0.8711814284324646, "num_tokens": 215461204.0, "step": 5673 }, { "epoch": 0.7217911207225544, "ewc_loss": 0.09912109375, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 6.110921382904053, "learning_rate": 1e-05, "loss": 0.5159, "mean_token_accuracy": 0.857648491859436, "num_tokens": 215493771.0, "step": 5674 }, { "epoch": 0.7219183310011449, "ewc_loss": 0.09521484375, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.7606682777404785, "learning_rate": 1e-05, "loss": 0.5149, "mean_token_accuracy": 0.858580470085144, "num_tokens": 215529398.0, "step": 5675 }, { "epoch": 0.7220455412797354, "ewc_loss": 0.0947265625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.867506504058838, "learning_rate": 1e-05, "loss": 0.4939, "mean_token_accuracy": 0.8654295206069946, "num_tokens": 215565794.0, "step": 5676 }, { "epoch": 0.7221727515583259, "ewc_loss": 0.091796875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.65330171585083, "learning_rate": 1e-05, "loss": 0.5385, "mean_token_accuracy": 0.849850058555603, "num_tokens": 215604005.0, "step": 5677 }, { "epoch": 0.7222999618369165, "ewc_loss": 0.09228515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.75855827331543, "learning_rate": 1e-05, "loss": 0.4651, "mean_token_accuracy": 0.8703041076660156, "num_tokens": 215635325.0, "step": 5678 }, { "epoch": 0.7224271721155069, "ewc_loss": 0.0888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.538219928741455, "learning_rate": 1e-05, "loss": 0.4919, "mean_token_accuracy": 0.862290620803833, "num_tokens": 215667477.0, "step": 5679 }, { "epoch": 0.7225543823940974, "ewc_loss": 0.09326171875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.745217800140381, "learning_rate": 1e-05, "loss": 0.5371, "mean_token_accuracy": 0.850317120552063, "num_tokens": 215707087.0, "step": 5680 }, { "epoch": 0.722681592672688, "ewc_loss": 0.0869140625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 7.486343383789062e-05, "grad_norm": 5.3945794105529785, "learning_rate": 1e-05, "loss": 0.4963, "mean_token_accuracy": 0.8627536296844482, "num_tokens": 215746182.0, "step": 5681 }, { "epoch": 0.7228088029512785, "ewc_loss": 0.09423828125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.824971675872803, "learning_rate": 1e-05, "loss": 0.5259, "mean_token_accuracy": 0.854583740234375, "num_tokens": 215785941.0, "step": 5682 }, { "epoch": 0.722936013229869, "ewc_loss": 0.08349609375, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 7.152557373046875e-05, "grad_norm": 5.253407955169678, "learning_rate": 1e-05, "loss": 0.4901, "mean_token_accuracy": 0.8622180223464966, "num_tokens": 215823562.0, "step": 5683 }, { "epoch": 0.7230632235084595, "ewc_loss": 0.09765625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 6.0042219161987305, "learning_rate": 1e-05, "loss": 0.5791, "mean_token_accuracy": 0.8371515274047852, "num_tokens": 215860454.0, "step": 5684 }, { "epoch": 0.72319043378705, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 7.009506225585938e-05, "grad_norm": 5.122734546661377, "learning_rate": 1e-05, "loss": 0.4759, "mean_token_accuracy": 0.8669778108596802, "num_tokens": 215895088.0, "step": 5685 }, { "epoch": 0.7233176440656405, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.28476095199585, "learning_rate": 1e-05, "loss": 0.5422, "mean_token_accuracy": 0.8518699407577515, "num_tokens": 215935170.0, "step": 5686 }, { "epoch": 0.723444854344231, "ewc_loss": 0.08349609375, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 7.104873657226562e-05, "grad_norm": 5.054040908813477, "learning_rate": 1e-05, "loss": 0.4701, "mean_token_accuracy": 0.8713403940200806, "num_tokens": 215977323.0, "step": 5687 }, { "epoch": 0.7235720646228215, "ewc_loss": 0.11181640625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.668167591094971, "learning_rate": 1e-05, "loss": 0.4954, "mean_token_accuracy": 0.8702808618545532, "num_tokens": 216016729.0, "step": 5688 }, { "epoch": 0.7236992749014121, "ewc_loss": 0.0888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.249232292175293, "learning_rate": 1e-05, "loss": 0.5135, "mean_token_accuracy": 0.857625424861908, "num_tokens": 216058514.0, "step": 5689 }, { "epoch": 0.7238264851800026, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.890347957611084, "learning_rate": 1e-05, "loss": 0.6053, "mean_token_accuracy": 0.8382331132888794, "num_tokens": 216098633.0, "step": 5690 }, { "epoch": 0.723953695458593, "ewc_loss": 0.099609375, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 5.673197269439697, "learning_rate": 1e-05, "loss": 0.4501, "mean_token_accuracy": 0.8789872527122498, "num_tokens": 216136124.0, "step": 5691 }, { "epoch": 0.7240809057371835, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.664748668670654, "learning_rate": 1e-05, "loss": 0.5517, "mean_token_accuracy": 0.8562299013137817, "num_tokens": 216178799.0, "step": 5692 }, { "epoch": 0.7242081160157741, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 5.926433086395264, "learning_rate": 1e-05, "loss": 0.5727, "mean_token_accuracy": 0.8446794748306274, "num_tokens": 216215233.0, "step": 5693 }, { "epoch": 0.7243353262943646, "ewc_loss": 0.10205078125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.219570159912109, "learning_rate": 1e-05, "loss": 0.4995, "mean_token_accuracy": 0.8642408847808838, "num_tokens": 216252258.0, "step": 5694 }, { "epoch": 0.7244625365729551, "ewc_loss": 0.09912109375, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.945455074310303, "learning_rate": 1e-05, "loss": 0.5557, "mean_token_accuracy": 0.8468478918075562, "num_tokens": 216287845.0, "step": 5695 }, { "epoch": 0.7245897468515456, "ewc_loss": 0.0908203125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.777469635009766, "learning_rate": 1e-05, "loss": 0.4942, "mean_token_accuracy": 0.8665911555290222, "num_tokens": 216325490.0, "step": 5696 }, { "epoch": 0.7247169571301361, "ewc_loss": 0.09912109375, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.965099334716797, "learning_rate": 1e-05, "loss": 0.5383, "mean_token_accuracy": 0.8507981300354004, "num_tokens": 216359353.0, "step": 5697 }, { "epoch": 0.7248441674087266, "ewc_loss": 0.0849609375, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 7.295608520507812e-05, "grad_norm": 5.4432692527771, "learning_rate": 1e-05, "loss": 0.5546, "mean_token_accuracy": 0.8502942323684692, "num_tokens": 216393620.0, "step": 5698 }, { "epoch": 0.7249713776873171, "ewc_loss": 0.099609375, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 6.079192638397217, "learning_rate": 1e-05, "loss": 0.506, "mean_token_accuracy": 0.8629442453384399, "num_tokens": 216427400.0, "step": 5699 }, { "epoch": 0.7250985879659076, "ewc_loss": 0.08447265625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 7.2479248046875e-05, "grad_norm": 5.235137939453125, "learning_rate": 1e-05, "loss": 0.5102, "mean_token_accuracy": 0.8561403155326843, "num_tokens": 216463817.0, "step": 5700 }, { "epoch": 0.7252257982444982, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.273042678833008, "learning_rate": 1e-05, "loss": 0.5247, "mean_token_accuracy": 0.8561216592788696, "num_tokens": 216495596.0, "step": 5701 }, { "epoch": 0.7253530085230887, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 7.200241088867188e-05, "grad_norm": 5.158963680267334, "learning_rate": 1e-05, "loss": 0.4924, "mean_token_accuracy": 0.864026665687561, "num_tokens": 216533003.0, "step": 5702 }, { "epoch": 0.7254802188016791, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.525698184967041, "learning_rate": 1e-05, "loss": 0.5209, "mean_token_accuracy": 0.8622393608093262, "num_tokens": 216569280.0, "step": 5703 }, { "epoch": 0.7256074290802697, "ewc_loss": 0.087890625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 7.581710815429688e-05, "grad_norm": 5.262749671936035, "learning_rate": 1e-05, "loss": 0.4669, "mean_token_accuracy": 0.8702518343925476, "num_tokens": 216603073.0, "step": 5704 }, { "epoch": 0.7257346393588602, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.694133281707764, "learning_rate": 1e-05, "loss": 0.5331, "mean_token_accuracy": 0.8537257313728333, "num_tokens": 216641971.0, "step": 5705 }, { "epoch": 0.7258618496374507, "ewc_loss": 0.09375, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.503356456756592, "learning_rate": 1e-05, "loss": 0.5738, "mean_token_accuracy": 0.8424860239028931, "num_tokens": 216683119.0, "step": 5706 }, { "epoch": 0.7259890599160412, "ewc_loss": 0.11181640625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.627939224243164, "learning_rate": 1e-05, "loss": 0.6064, "mean_token_accuracy": 0.8350593447685242, "num_tokens": 216722139.0, "step": 5707 }, { "epoch": 0.7261162701946318, "ewc_loss": 0.0966796875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.703305721282959, "learning_rate": 1e-05, "loss": 0.4789, "mean_token_accuracy": 0.8694710731506348, "num_tokens": 216757712.0, "step": 5708 }, { "epoch": 0.7262434804732223, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.373149394989014, "learning_rate": 1e-05, "loss": 0.5364, "mean_token_accuracy": 0.8501648902893066, "num_tokens": 216794132.0, "step": 5709 }, { "epoch": 0.7263706907518127, "ewc_loss": 0.09619140625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.746963977813721, "learning_rate": 1e-05, "loss": 0.5175, "mean_token_accuracy": 0.8547967672348022, "num_tokens": 216833937.0, "step": 5710 }, { "epoch": 0.7264979010304032, "ewc_loss": 0.09814453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 6.067171573638916, "learning_rate": 1e-05, "loss": 0.5228, "mean_token_accuracy": 0.8567554950714111, "num_tokens": 216872979.0, "step": 5711 }, { "epoch": 0.7266251113089938, "ewc_loss": 0.0947265625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.7032694816589355, "learning_rate": 1e-05, "loss": 0.4813, "mean_token_accuracy": 0.8682419061660767, "num_tokens": 216915369.0, "step": 5712 }, { "epoch": 0.7267523215875843, "ewc_loss": 0.09423828125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.852367401123047, "learning_rate": 1e-05, "loss": 0.4931, "mean_token_accuracy": 0.8635501861572266, "num_tokens": 216948876.0, "step": 5713 }, { "epoch": 0.7268795318661748, "ewc_loss": 0.09130859375, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.625738143920898, "learning_rate": 1e-05, "loss": 0.5096, "mean_token_accuracy": 0.8594926595687866, "num_tokens": 216988254.0, "step": 5714 }, { "epoch": 0.7270067421447653, "ewc_loss": 0.09228515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.738483905792236, "learning_rate": 1e-05, "loss": 0.5108, "mean_token_accuracy": 0.8561604022979736, "num_tokens": 217024792.0, "step": 5715 }, { "epoch": 0.7271339524233558, "ewc_loss": 0.08837890625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 7.62939453125e-05, "grad_norm": 5.540040493011475, "learning_rate": 1e-05, "loss": 0.5175, "mean_token_accuracy": 0.8579761385917664, "num_tokens": 217063228.0, "step": 5716 }, { "epoch": 0.7272611627019463, "ewc_loss": 0.09228515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.703369140625, "learning_rate": 1e-05, "loss": 0.5887, "mean_token_accuracy": 0.8324217200279236, "num_tokens": 217102601.0, "step": 5717 }, { "epoch": 0.7273883729805368, "ewc_loss": 0.08740234375, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 7.534027099609375e-05, "grad_norm": 5.426419258117676, "learning_rate": 1e-05, "loss": 0.493, "mean_token_accuracy": 0.8634771108627319, "num_tokens": 217137839.0, "step": 5718 }, { "epoch": 0.7275155832591274, "ewc_loss": 0.09228515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.751437664031982, "learning_rate": 1e-05, "loss": 0.5127, "mean_token_accuracy": 0.8632386326789856, "num_tokens": 217179862.0, "step": 5719 }, { "epoch": 0.7276427935377179, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 7.200241088867188e-05, "grad_norm": 5.301676273345947, "learning_rate": 1e-05, "loss": 0.5833, "mean_token_accuracy": 0.8354120254516602, "num_tokens": 217217728.0, "step": 5720 }, { "epoch": 0.7277700038163084, "ewc_loss": 0.09423828125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.877650260925293, "learning_rate": 1e-05, "loss": 0.519, "mean_token_accuracy": 0.8582452535629272, "num_tokens": 217255410.0, "step": 5721 }, { "epoch": 0.7278972140948988, "ewc_loss": 0.08349609375, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 7.152557373046875e-05, "grad_norm": 5.158504009246826, "learning_rate": 1e-05, "loss": 0.5292, "mean_token_accuracy": 0.850421130657196, "num_tokens": 217294365.0, "step": 5722 }, { "epoch": 0.7280244243734894, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 6.126972198486328, "learning_rate": 1e-05, "loss": 0.5036, "mean_token_accuracy": 0.8623713850975037, "num_tokens": 217333451.0, "step": 5723 }, { "epoch": 0.7281516346520799, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 7.05718994140625e-05, "grad_norm": 5.058860778808594, "learning_rate": 1e-05, "loss": 0.511, "mean_token_accuracy": 0.8536473512649536, "num_tokens": 217368737.0, "step": 5724 }, { "epoch": 0.7282788449306704, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.481095790863037, "learning_rate": 1e-05, "loss": 0.5786, "mean_token_accuracy": 0.8403536081314087, "num_tokens": 217409476.0, "step": 5725 }, { "epoch": 0.7284060552092609, "ewc_loss": 0.0859375, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 7.390975952148438e-05, "grad_norm": 5.068675994873047, "learning_rate": 1e-05, "loss": 0.4796, "mean_token_accuracy": 0.8674708008766174, "num_tokens": 217446481.0, "step": 5726 }, { "epoch": 0.7285332654878515, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.873489856719971, "learning_rate": 1e-05, "loss": 0.5168, "mean_token_accuracy": 0.8678998947143555, "num_tokens": 217482647.0, "step": 5727 }, { "epoch": 0.7286604757664419, "ewc_loss": 0.0947265625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.439431190490723, "learning_rate": 1e-05, "loss": 0.5363, "mean_token_accuracy": 0.8548348546028137, "num_tokens": 217520783.0, "step": 5728 }, { "epoch": 0.7287876860450324, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.91898250579834, "learning_rate": 1e-05, "loss": 0.5455, "mean_token_accuracy": 0.8568239808082581, "num_tokens": 217557061.0, "step": 5729 }, { "epoch": 0.7289148963236229, "ewc_loss": 0.10400390625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 5.870001316070557, "learning_rate": 1e-05, "loss": 0.4846, "mean_token_accuracy": 0.8684242367744446, "num_tokens": 217586833.0, "step": 5730 }, { "epoch": 0.7290421066022135, "ewc_loss": 0.10986328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.545528888702393, "learning_rate": 1e-05, "loss": 0.559, "mean_token_accuracy": 0.8486599922180176, "num_tokens": 217622644.0, "step": 5731 }, { "epoch": 0.729169316880804, "ewc_loss": 0.10498046875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 5.9791259765625, "learning_rate": 1e-05, "loss": 0.5172, "mean_token_accuracy": 0.862982988357544, "num_tokens": 217659345.0, "step": 5732 }, { "epoch": 0.7292965271593945, "ewc_loss": 0.10107421875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.085341930389404, "learning_rate": 1e-05, "loss": 0.499, "mean_token_accuracy": 0.8665570020675659, "num_tokens": 217698134.0, "step": 5733 }, { "epoch": 0.7294237374379849, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 5.9541826248168945, "learning_rate": 1e-05, "loss": 0.5466, "mean_token_accuracy": 0.8473285436630249, "num_tokens": 217730578.0, "step": 5734 }, { "epoch": 0.7295509477165755, "ewc_loss": 0.09033203125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.687538146972656, "learning_rate": 1e-05, "loss": 0.4518, "mean_token_accuracy": 0.8735028505325317, "num_tokens": 217768090.0, "step": 5735 }, { "epoch": 0.729678157995166, "ewc_loss": 0.09912109375, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.964054107666016, "learning_rate": 1e-05, "loss": 0.5169, "mean_token_accuracy": 0.8583616018295288, "num_tokens": 217809350.0, "step": 5736 }, { "epoch": 0.7298053682737565, "ewc_loss": 0.08642578125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 7.43865966796875e-05, "grad_norm": 5.415198802947998, "learning_rate": 1e-05, "loss": 0.5267, "mean_token_accuracy": 0.8543223142623901, "num_tokens": 217852478.0, "step": 5737 }, { "epoch": 0.7299325785523471, "ewc_loss": 0.10009765625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 6.054019451141357, "learning_rate": 1e-05, "loss": 0.4932, "mean_token_accuracy": 0.8664352297782898, "num_tokens": 217885463.0, "step": 5738 }, { "epoch": 0.7300597888309376, "ewc_loss": 0.08544921875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 7.343292236328125e-05, "grad_norm": 5.261511325836182, "learning_rate": 1e-05, "loss": 0.5403, "mean_token_accuracy": 0.8502415418624878, "num_tokens": 217925712.0, "step": 5739 }, { "epoch": 0.730186999109528, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.210330963134766, "learning_rate": 1e-05, "loss": 0.5214, "mean_token_accuracy": 0.8564196825027466, "num_tokens": 217969628.0, "step": 5740 }, { "epoch": 0.7303142093881185, "ewc_loss": 0.0869140625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 7.486343383789062e-05, "grad_norm": 5.190558433532715, "learning_rate": 1e-05, "loss": 0.4999, "mean_token_accuracy": 0.8596303462982178, "num_tokens": 218012027.0, "step": 5741 }, { "epoch": 0.7304414196667091, "ewc_loss": 0.10595703125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.423733234405518, "learning_rate": 1e-05, "loss": 0.553, "mean_token_accuracy": 0.8528939485549927, "num_tokens": 218054018.0, "step": 5742 }, { "epoch": 0.7305686299452996, "ewc_loss": 0.08837890625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 7.62939453125e-05, "grad_norm": 5.227595806121826, "learning_rate": 1e-05, "loss": 0.5578, "mean_token_accuracy": 0.8448055982589722, "num_tokens": 218099181.0, "step": 5743 }, { "epoch": 0.7306958402238901, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.610519886016846, "learning_rate": 1e-05, "loss": 0.5045, "mean_token_accuracy": 0.8654502034187317, "num_tokens": 218135330.0, "step": 5744 }, { "epoch": 0.7308230505024806, "ewc_loss": 0.09228515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.4092020988464355, "learning_rate": 1e-05, "loss": 0.5719, "mean_token_accuracy": 0.8431868553161621, "num_tokens": 218171785.0, "step": 5745 }, { "epoch": 0.7309502607810711, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.645073890686035, "learning_rate": 1e-05, "loss": 0.6078, "mean_token_accuracy": 0.834834098815918, "num_tokens": 218204230.0, "step": 5746 }, { "epoch": 0.7310774710596616, "ewc_loss": 0.09521484375, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.603339195251465, "learning_rate": 1e-05, "loss": 0.4842, "mean_token_accuracy": 0.871995210647583, "num_tokens": 218237755.0, "step": 5747 }, { "epoch": 0.7312046813382521, "ewc_loss": 0.10791015625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.4643874168396, "learning_rate": 1e-05, "loss": 0.5737, "mean_token_accuracy": 0.8437304496765137, "num_tokens": 218271850.0, "step": 5748 }, { "epoch": 0.7313318916168426, "ewc_loss": 0.09619140625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.702165603637695, "learning_rate": 1e-05, "loss": 0.5143, "mean_token_accuracy": 0.8634334206581116, "num_tokens": 218308466.0, "step": 5749 }, { "epoch": 0.7314591018954332, "ewc_loss": 0.10205078125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.186151027679443, "learning_rate": 1e-05, "loss": 0.5636, "mean_token_accuracy": 0.8440558910369873, "num_tokens": 218346133.0, "step": 5750 }, { "epoch": 0.7315863121740237, "ewc_loss": 0.095703125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.715178966522217, "learning_rate": 1e-05, "loss": 0.5267, "mean_token_accuracy": 0.8583784699440002, "num_tokens": 218384341.0, "step": 5751 }, { "epoch": 0.7317135224526141, "ewc_loss": 0.095703125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.9134907722473145, "learning_rate": 1e-05, "loss": 0.4997, "mean_token_accuracy": 0.8604961633682251, "num_tokens": 218418314.0, "step": 5752 }, { "epoch": 0.7318407327312046, "ewc_loss": 0.09375, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.680897235870361, "learning_rate": 1e-05, "loss": 0.5407, "mean_token_accuracy": 0.8492305278778076, "num_tokens": 218455258.0, "step": 5753 }, { "epoch": 0.7319679430097952, "ewc_loss": 0.09228515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.711174488067627, "learning_rate": 1e-05, "loss": 0.5211, "mean_token_accuracy": 0.8582038879394531, "num_tokens": 218498733.0, "step": 5754 }, { "epoch": 0.7320951532883857, "ewc_loss": 0.09228515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.629404067993164, "learning_rate": 1e-05, "loss": 0.5334, "mean_token_accuracy": 0.8517137169837952, "num_tokens": 218541365.0, "step": 5755 }, { "epoch": 0.7322223635669762, "ewc_loss": 0.08984375, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.6059746742248535, "learning_rate": 1e-05, "loss": 0.5474, "mean_token_accuracy": 0.8485419750213623, "num_tokens": 218578292.0, "step": 5756 }, { "epoch": 0.7323495738455668, "ewc_loss": 0.08935546875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.559876918792725, "learning_rate": 1e-05, "loss": 0.5146, "mean_token_accuracy": 0.8574701547622681, "num_tokens": 218617988.0, "step": 5757 }, { "epoch": 0.7324767841241572, "ewc_loss": 0.08935546875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.568398475646973, "learning_rate": 1e-05, "loss": 0.5381, "mean_token_accuracy": 0.8544140458106995, "num_tokens": 218657759.0, "step": 5758 }, { "epoch": 0.7326039944027477, "ewc_loss": 0.08837890625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 7.62939453125e-05, "grad_norm": 5.504742622375488, "learning_rate": 1e-05, "loss": 0.5068, "mean_token_accuracy": 0.8579288125038147, "num_tokens": 218693619.0, "step": 5759 }, { "epoch": 0.7327312046813382, "ewc_loss": 0.08984375, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.570341110229492, "learning_rate": 1e-05, "loss": 0.4813, "mean_token_accuracy": 0.864313542842865, "num_tokens": 218725978.0, "step": 5760 }, { "epoch": 0.7328584149599288, "ewc_loss": 0.08740234375, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 7.534027099609375e-05, "grad_norm": 5.432615756988525, "learning_rate": 1e-05, "loss": 0.4968, "mean_token_accuracy": 0.8618103861808777, "num_tokens": 218769616.0, "step": 5761 }, { "epoch": 0.7329856252385193, "ewc_loss": 0.09130859375, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.60687255859375, "learning_rate": 1e-05, "loss": 0.5035, "mean_token_accuracy": 0.8632959127426147, "num_tokens": 218806053.0, "step": 5762 }, { "epoch": 0.7331128355171098, "ewc_loss": 0.08544921875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 7.343292236328125e-05, "grad_norm": 5.345770359039307, "learning_rate": 1e-05, "loss": 0.5292, "mean_token_accuracy": 0.8492785096168518, "num_tokens": 218844650.0, "step": 5763 }, { "epoch": 0.7332400457957003, "ewc_loss": 0.0927734375, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.709929466247559, "learning_rate": 1e-05, "loss": 0.5376, "mean_token_accuracy": 0.8503279685974121, "num_tokens": 218885943.0, "step": 5764 }, { "epoch": 0.7333672560742908, "ewc_loss": 0.08349609375, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 7.152557373046875e-05, "grad_norm": 5.2167134284973145, "learning_rate": 1e-05, "loss": 0.5002, "mean_token_accuracy": 0.8579575419425964, "num_tokens": 218923939.0, "step": 5765 }, { "epoch": 0.7334944663528813, "ewc_loss": 0.095703125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.89055061340332, "learning_rate": 1e-05, "loss": 0.5078, "mean_token_accuracy": 0.8590980768203735, "num_tokens": 218964252.0, "step": 5766 }, { "epoch": 0.7336216766314718, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 6.961822509765625e-05, "grad_norm": 5.060054779052734, "learning_rate": 1e-05, "loss": 0.5168, "mean_token_accuracy": 0.8529763221740723, "num_tokens": 219009815.0, "step": 5767 }, { "epoch": 0.7337488869100623, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.230719089508057, "learning_rate": 1e-05, "loss": 0.5141, "mean_token_accuracy": 0.8638635277748108, "num_tokens": 219047998.0, "step": 5768 }, { "epoch": 0.7338760971886529, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 6.961822509765625e-05, "grad_norm": 4.904159069061279, "learning_rate": 1e-05, "loss": 0.5611, "mean_token_accuracy": 0.842824399471283, "num_tokens": 219082447.0, "step": 5769 }, { "epoch": 0.7340033074672434, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.803063869476318, "learning_rate": 1e-05, "loss": 0.576, "mean_token_accuracy": 0.847107470035553, "num_tokens": 219119393.0, "step": 5770 }, { "epoch": 0.7341305177458338, "ewc_loss": 0.087890625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 7.581710815429688e-05, "grad_norm": 5.042901992797852, "learning_rate": 1e-05, "loss": 0.5695, "mean_token_accuracy": 0.8467104434967041, "num_tokens": 219155954.0, "step": 5771 }, { "epoch": 0.7342577280244243, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 0.000118255615234375, "grad_norm": 7.363348960876465, "learning_rate": 1e-05, "loss": 0.5719, "mean_token_accuracy": 0.8565980195999146, "num_tokens": 219195960.0, "step": 5772 }, { "epoch": 0.7343849383030149, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 5.780839443206787, "learning_rate": 1e-05, "loss": 0.5294, "mean_token_accuracy": 0.8570375442504883, "num_tokens": 219238323.0, "step": 5773 }, { "epoch": 0.7345121485816054, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 0.00011444091796875, "grad_norm": 7.250477313995361, "learning_rate": 1e-05, "loss": 0.5314, "mean_token_accuracy": 0.8612855672836304, "num_tokens": 219279044.0, "step": 5774 }, { "epoch": 0.7346393588601959, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.300158977508545, "learning_rate": 1e-05, "loss": 0.5087, "mean_token_accuracy": 0.8649969100952148, "num_tokens": 219312635.0, "step": 5775 }, { "epoch": 0.7347665691387865, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.741222381591797, "learning_rate": 1e-05, "loss": 0.4846, "mean_token_accuracy": 0.8728320598602295, "num_tokens": 219349424.0, "step": 5776 }, { "epoch": 0.7348937794173769, "ewc_loss": 0.109375, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.184213638305664, "learning_rate": 1e-05, "loss": 0.5417, "mean_token_accuracy": 0.8555773496627808, "num_tokens": 219389230.0, "step": 5777 }, { "epoch": 0.7350209896959674, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.394554138183594, "learning_rate": 1e-05, "loss": 0.54, "mean_token_accuracy": 0.8575196266174316, "num_tokens": 219424082.0, "step": 5778 }, { "epoch": 0.7351481999745579, "ewc_loss": 0.10009765625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 5.938383102416992, "learning_rate": 1e-05, "loss": 0.5171, "mean_token_accuracy": 0.860199511051178, "num_tokens": 219463324.0, "step": 5779 }, { "epoch": 0.7352754102531485, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 6.160707950592041, "learning_rate": 1e-05, "loss": 0.4595, "mean_token_accuracy": 0.8770712614059448, "num_tokens": 219500653.0, "step": 5780 }, { "epoch": 0.735402620531739, "ewc_loss": 0.09423828125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.725464820861816, "learning_rate": 1e-05, "loss": 0.4842, "mean_token_accuracy": 0.8682674765586853, "num_tokens": 219538906.0, "step": 5781 }, { "epoch": 0.7355298308103295, "ewc_loss": 0.099609375, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 6.072809219360352, "learning_rate": 1e-05, "loss": 0.5846, "mean_token_accuracy": 0.8351665735244751, "num_tokens": 219578950.0, "step": 5782 }, { "epoch": 0.7356570410889199, "ewc_loss": 0.08935546875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.5331830978393555, "learning_rate": 1e-05, "loss": 0.4662, "mean_token_accuracy": 0.8720306158065796, "num_tokens": 219614298.0, "step": 5783 }, { "epoch": 0.7357842513675105, "ewc_loss": 0.10009765625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 6.098546504974365, "learning_rate": 1e-05, "loss": 0.5458, "mean_token_accuracy": 0.8508495092391968, "num_tokens": 219655521.0, "step": 5784 }, { "epoch": 0.735911461646101, "ewc_loss": 0.0869140625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 7.486343383789062e-05, "grad_norm": 5.370280742645264, "learning_rate": 1e-05, "loss": 0.4966, "mean_token_accuracy": 0.8621387481689453, "num_tokens": 219689112.0, "step": 5785 }, { "epoch": 0.7360386719246915, "ewc_loss": 0.10302734375, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.2651591300964355, "learning_rate": 1e-05, "loss": 0.6068, "mean_token_accuracy": 0.8337414860725403, "num_tokens": 219727432.0, "step": 5786 }, { "epoch": 0.736165882203282, "ewc_loss": 0.08544921875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 7.343292236328125e-05, "grad_norm": 5.239738464355469, "learning_rate": 1e-05, "loss": 0.5166, "mean_token_accuracy": 0.8560117483139038, "num_tokens": 219767231.0, "step": 5787 }, { "epoch": 0.7362930924818726, "ewc_loss": 0.10791015625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.5489020347595215, "learning_rate": 1e-05, "loss": 0.5041, "mean_token_accuracy": 0.8646955490112305, "num_tokens": 219798999.0, "step": 5788 }, { "epoch": 0.736420302760463, "ewc_loss": 0.087890625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 7.581710815429688e-05, "grad_norm": 5.300352573394775, "learning_rate": 1e-05, "loss": 0.4753, "mean_token_accuracy": 0.8678539395332336, "num_tokens": 219844867.0, "step": 5789 }, { "epoch": 0.7365475130390535, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.794856071472168, "learning_rate": 1e-05, "loss": 0.5319, "mean_token_accuracy": 0.8599758744239807, "num_tokens": 219882817.0, "step": 5790 }, { "epoch": 0.736674723317644, "ewc_loss": 0.09423828125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.628184795379639, "learning_rate": 1e-05, "loss": 0.5507, "mean_token_accuracy": 0.8436226844787598, "num_tokens": 219916776.0, "step": 5791 }, { "epoch": 0.7368019335962346, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.668153285980225, "learning_rate": 1e-05, "loss": 0.4715, "mean_token_accuracy": 0.8768839240074158, "num_tokens": 219953471.0, "step": 5792 }, { "epoch": 0.7369291438748251, "ewc_loss": 0.09814453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.919259548187256, "learning_rate": 1e-05, "loss": 0.5485, "mean_token_accuracy": 0.8496895432472229, "num_tokens": 219983313.0, "step": 5793 }, { "epoch": 0.7370563541534156, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.222049236297607, "learning_rate": 1e-05, "loss": 0.593, "mean_token_accuracy": 0.8387418985366821, "num_tokens": 220028222.0, "step": 5794 }, { "epoch": 0.737183564432006, "ewc_loss": 0.09912109375, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.985208511352539, "learning_rate": 1e-05, "loss": 0.5197, "mean_token_accuracy": 0.8580094575881958, "num_tokens": 220069005.0, "step": 5795 }, { "epoch": 0.7373107747105966, "ewc_loss": 0.08984375, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.723463535308838, "learning_rate": 1e-05, "loss": 0.5404, "mean_token_accuracy": 0.8509293794631958, "num_tokens": 220103557.0, "step": 5796 }, { "epoch": 0.7374379849891871, "ewc_loss": 0.09912109375, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 6.074210166931152, "learning_rate": 1e-05, "loss": 0.5139, "mean_token_accuracy": 0.8571714162826538, "num_tokens": 220141405.0, "step": 5797 }, { "epoch": 0.7375651952677776, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 7.200241088867188e-05, "grad_norm": 5.293044090270996, "learning_rate": 1e-05, "loss": 0.5097, "mean_token_accuracy": 0.8565690517425537, "num_tokens": 220184913.0, "step": 5798 }, { "epoch": 0.7376924055463682, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.305456638336182, "learning_rate": 1e-05, "loss": 0.5323, "mean_token_accuracy": 0.8542150259017944, "num_tokens": 220221886.0, "step": 5799 }, { "epoch": 0.7378196158249587, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 7.009506225585938e-05, "grad_norm": 5.06622838973999, "learning_rate": 1e-05, "loss": 0.4826, "mean_token_accuracy": 0.8642622232437134, "num_tokens": 220260945.0, "step": 5800 }, { "epoch": 0.7379468261035491, "ewc_loss": 0.11181640625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.724338531494141, "learning_rate": 1e-05, "loss": 0.5231, "mean_token_accuracy": 0.8613501787185669, "num_tokens": 220299580.0, "step": 5801 }, { "epoch": 0.7380740363821396, "ewc_loss": 0.0888671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.204501152038574, "learning_rate": 1e-05, "loss": 0.465, "mean_token_accuracy": 0.869401216506958, "num_tokens": 220338063.0, "step": 5802 }, { "epoch": 0.7382012466607302, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 7.05070161819458, "learning_rate": 1e-05, "loss": 0.5118, "mean_token_accuracy": 0.8657629489898682, "num_tokens": 220381212.0, "step": 5803 }, { "epoch": 0.7383284569393207, "ewc_loss": 0.09912109375, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.727042198181152, "learning_rate": 1e-05, "loss": 0.5421, "mean_token_accuracy": 0.8556632995605469, "num_tokens": 220422665.0, "step": 5804 }, { "epoch": 0.7384556672179112, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.875305652618408, "learning_rate": 1e-05, "loss": 0.5091, "mean_token_accuracy": 0.867186963558197, "num_tokens": 220462044.0, "step": 5805 }, { "epoch": 0.7385828774965018, "ewc_loss": 0.10498046875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.047215461730957, "learning_rate": 1e-05, "loss": 0.5913, "mean_token_accuracy": 0.8362457156181335, "num_tokens": 220497373.0, "step": 5806 }, { "epoch": 0.7387100877750922, "ewc_loss": 0.10595703125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.433592319488525, "learning_rate": 1e-05, "loss": 0.4845, "mean_token_accuracy": 0.8706094026565552, "num_tokens": 220537612.0, "step": 5807 }, { "epoch": 0.7388372980536827, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 5.997596740722656, "learning_rate": 1e-05, "loss": 0.5456, "mean_token_accuracy": 0.8511984944343567, "num_tokens": 220574874.0, "step": 5808 }, { "epoch": 0.7389645083322732, "ewc_loss": 0.09716796875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 6.067476272583008, "learning_rate": 1e-05, "loss": 0.4851, "mean_token_accuracy": 0.8700025081634521, "num_tokens": 220615699.0, "step": 5809 }, { "epoch": 0.7390917186108638, "ewc_loss": 0.09521484375, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.870974063873291, "learning_rate": 1e-05, "loss": 0.5409, "mean_token_accuracy": 0.8482969999313354, "num_tokens": 220654517.0, "step": 5810 }, { "epoch": 0.7392189288894543, "ewc_loss": 0.09228515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.850734233856201, "learning_rate": 1e-05, "loss": 0.6088, "mean_token_accuracy": 0.8306924104690552, "num_tokens": 220695915.0, "step": 5811 }, { "epoch": 0.7393461391680448, "ewc_loss": 0.091796875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.701656341552734, "learning_rate": 1e-05, "loss": 0.5197, "mean_token_accuracy": 0.8564778566360474, "num_tokens": 220734880.0, "step": 5812 }, { "epoch": 0.7394733494466353, "ewc_loss": 0.0927734375, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.779079914093018, "learning_rate": 1e-05, "loss": 0.533, "mean_token_accuracy": 0.8540282249450684, "num_tokens": 220767800.0, "step": 5813 }, { "epoch": 0.7396005597252258, "ewc_loss": 0.087890625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 7.581710815429688e-05, "grad_norm": 5.556694030761719, "learning_rate": 1e-05, "loss": 0.5545, "mean_token_accuracy": 0.8473814725875854, "num_tokens": 220812253.0, "step": 5814 }, { "epoch": 0.7397277700038163, "ewc_loss": 0.0927734375, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.773621082305908, "learning_rate": 1e-05, "loss": 0.4655, "mean_token_accuracy": 0.8726080656051636, "num_tokens": 220847777.0, "step": 5815 }, { "epoch": 0.7398549802824068, "ewc_loss": 0.08642578125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 7.43865966796875e-05, "grad_norm": 5.403069972991943, "learning_rate": 1e-05, "loss": 0.5015, "mean_token_accuracy": 0.858269989490509, "num_tokens": 220884974.0, "step": 5816 }, { "epoch": 0.7399821905609973, "ewc_loss": 0.09375, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.869150638580322, "learning_rate": 1e-05, "loss": 0.5153, "mean_token_accuracy": 0.8557397127151489, "num_tokens": 220926236.0, "step": 5817 }, { "epoch": 0.7401094008395879, "ewc_loss": 0.08349609375, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 7.104873657226562e-05, "grad_norm": 5.246585369110107, "learning_rate": 1e-05, "loss": 0.553, "mean_token_accuracy": 0.8436422348022461, "num_tokens": 220967985.0, "step": 5818 }, { "epoch": 0.7402366111181784, "ewc_loss": 0.10009765625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 6.093041896820068, "learning_rate": 1e-05, "loss": 0.5319, "mean_token_accuracy": 0.8550808429718018, "num_tokens": 221004308.0, "step": 5819 }, { "epoch": 0.7403638213967688, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 6.961822509765625e-05, "grad_norm": 5.086665153503418, "learning_rate": 1e-05, "loss": 0.445, "mean_token_accuracy": 0.8759328126907349, "num_tokens": 221043455.0, "step": 5820 }, { "epoch": 0.7404910316753593, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.460427284240723, "learning_rate": 1e-05, "loss": 0.5613, "mean_token_accuracy": 0.8465114235877991, "num_tokens": 221084634.0, "step": 5821 }, { "epoch": 0.7406182419539499, "ewc_loss": 0.08349609375, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 7.104873657226562e-05, "grad_norm": 5.080903053283691, "learning_rate": 1e-05, "loss": 0.5043, "mean_token_accuracy": 0.8587506413459778, "num_tokens": 221117350.0, "step": 5822 }, { "epoch": 0.7407454522325404, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.896210193634033, "learning_rate": 1e-05, "loss": 0.524, "mean_token_accuracy": 0.8586317300796509, "num_tokens": 221150864.0, "step": 5823 }, { "epoch": 0.7408726625111309, "ewc_loss": 0.09326171875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.404477119445801, "learning_rate": 1e-05, "loss": 0.4738, "mean_token_accuracy": 0.8719728589057922, "num_tokens": 221186501.0, "step": 5824 }, { "epoch": 0.7409998727897215, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 7.033855438232422, "learning_rate": 1e-05, "loss": 0.6153, "mean_token_accuracy": 0.8383499383926392, "num_tokens": 221227734.0, "step": 5825 }, { "epoch": 0.7411270830683119, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 5.920546054840088, "learning_rate": 1e-05, "loss": 0.5281, "mean_token_accuracy": 0.856529176235199, "num_tokens": 221260615.0, "step": 5826 }, { "epoch": 0.7412542933469024, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.681827068328857, "learning_rate": 1e-05, "loss": 0.5659, "mean_token_accuracy": 0.8485881090164185, "num_tokens": 221298553.0, "step": 5827 }, { "epoch": 0.7413815036254929, "ewc_loss": 0.10498046875, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.068153381347656, "learning_rate": 1e-05, "loss": 0.5185, "mean_token_accuracy": 0.859398365020752, "num_tokens": 221338803.0, "step": 5828 }, { "epoch": 0.7415087139040835, "ewc_loss": 0.10107421875, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.199288368225098, "learning_rate": 1e-05, "loss": 0.5129, "mean_token_accuracy": 0.8604092001914978, "num_tokens": 221379557.0, "step": 5829 }, { "epoch": 0.741635924182674, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 6.013542652130127, "learning_rate": 1e-05, "loss": 0.5127, "mean_token_accuracy": 0.858250617980957, "num_tokens": 221412242.0, "step": 5830 }, { "epoch": 0.7417631344612645, "ewc_loss": 0.0908203125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.78634786605835, "learning_rate": 1e-05, "loss": 0.4522, "mean_token_accuracy": 0.8744518756866455, "num_tokens": 221449382.0, "step": 5831 }, { "epoch": 0.7418903447398549, "ewc_loss": 0.099609375, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 5.989438056945801, "learning_rate": 1e-05, "loss": 0.5539, "mean_token_accuracy": 0.8480738401412964, "num_tokens": 221489480.0, "step": 5832 }, { "epoch": 0.7420175550184455, "ewc_loss": 0.087890625, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 7.581710815429688e-05, "grad_norm": 5.505442142486572, "learning_rate": 1e-05, "loss": 0.5104, "mean_token_accuracy": 0.8547824025154114, "num_tokens": 221525264.0, "step": 5833 }, { "epoch": 0.742144765297036, "ewc_loss": 0.0986328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 6.042435646057129, "learning_rate": 1e-05, "loss": 0.4991, "mean_token_accuracy": 0.8639354109764099, "num_tokens": 221568447.0, "step": 5834 }, { "epoch": 0.7422719755756265, "ewc_loss": 0.0849609375, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 7.295608520507812e-05, "grad_norm": 5.315800189971924, "learning_rate": 1e-05, "loss": 0.587, "mean_token_accuracy": 0.8335987329483032, "num_tokens": 221612943.0, "step": 5835 }, { "epoch": 0.742399185854217, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 6.204455852508545, "learning_rate": 1e-05, "loss": 0.6199, "mean_token_accuracy": 0.8275150060653687, "num_tokens": 221645449.0, "step": 5836 }, { "epoch": 0.7425263961328076, "ewc_loss": 0.0849609375, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 7.295608520507812e-05, "grad_norm": 5.197265625, "learning_rate": 1e-05, "loss": 0.4642, "mean_token_accuracy": 0.8713489770889282, "num_tokens": 221684041.0, "step": 5837 }, { "epoch": 0.742653606411398, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.442245006561279, "learning_rate": 1e-05, "loss": 0.4888, "mean_token_accuracy": 0.8670721650123596, "num_tokens": 221720200.0, "step": 5838 }, { "epoch": 0.7427808166899885, "ewc_loss": 0.08740234375, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 7.534027099609375e-05, "grad_norm": 5.213860988616943, "learning_rate": 1e-05, "loss": 0.511, "mean_token_accuracy": 0.85683673620224, "num_tokens": 221758997.0, "step": 5839 }, { "epoch": 0.742908026968579, "ewc_loss": 0.11083984375, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.6728105545043945, "learning_rate": 1e-05, "loss": 0.5306, "mean_token_accuracy": 0.8591740131378174, "num_tokens": 221800713.0, "step": 5840 }, { "epoch": 0.7430352372471696, "ewc_loss": 0.09228515625, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.407308578491211, "learning_rate": 1e-05, "loss": 0.5615, "mean_token_accuracy": 0.8461499214172363, "num_tokens": 221840336.0, "step": 5841 }, { "epoch": 0.7431624475257601, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.725045204162598, "learning_rate": 1e-05, "loss": 0.4762, "mean_token_accuracy": 0.8783459663391113, "num_tokens": 221876286.0, "step": 5842 }, { "epoch": 0.7432896578043506, "ewc_loss": 0.095703125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.666080474853516, "learning_rate": 1e-05, "loss": 0.5409, "mean_token_accuracy": 0.8517030477523804, "num_tokens": 221917596.0, "step": 5843 }, { "epoch": 0.743416868082941, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.534333229064941, "learning_rate": 1e-05, "loss": 0.5075, "mean_token_accuracy": 0.8647089004516602, "num_tokens": 221954045.0, "step": 5844 }, { "epoch": 0.7435440783615316, "ewc_loss": 0.09814453125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.781179904937744, "learning_rate": 1e-05, "loss": 0.5078, "mean_token_accuracy": 0.8601111173629761, "num_tokens": 221996493.0, "step": 5845 }, { "epoch": 0.7436712886401221, "ewc_loss": 0.10205078125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.227940559387207, "learning_rate": 1e-05, "loss": 0.5322, "mean_token_accuracy": 0.8554353713989258, "num_tokens": 222036608.0, "step": 5846 }, { "epoch": 0.7437984989187126, "ewc_loss": 0.09521484375, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.784032821655273, "learning_rate": 1e-05, "loss": 0.5056, "mean_token_accuracy": 0.8620952367782593, "num_tokens": 222080647.0, "step": 5847 }, { "epoch": 0.7439257091973032, "ewc_loss": 0.095703125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.935855865478516, "learning_rate": 1e-05, "loss": 0.5497, "mean_token_accuracy": 0.8497955203056335, "num_tokens": 222114039.0, "step": 5848 }, { "epoch": 0.7440529194758937, "ewc_loss": 0.0947265625, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.737781524658203, "learning_rate": 1e-05, "loss": 0.5454, "mean_token_accuracy": 0.8511278033256531, "num_tokens": 222149233.0, "step": 5849 }, { "epoch": 0.7441801297544841, "ewc_loss": 0.0908203125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.713266849517822, "learning_rate": 1e-05, "loss": 0.4692, "mean_token_accuracy": 0.8711204528808594, "num_tokens": 222185842.0, "step": 5850 }, { "epoch": 0.7443073400330746, "ewc_loss": 0.09326171875, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.679189205169678, "learning_rate": 1e-05, "loss": 0.4965, "mean_token_accuracy": 0.8643016815185547, "num_tokens": 222233576.0, "step": 5851 }, { "epoch": 0.7444345503116652, "ewc_loss": 0.0888671875, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.600728988647461, "learning_rate": 1e-05, "loss": 0.485, "mean_token_accuracy": 0.8603194952011108, "num_tokens": 222267822.0, "step": 5852 }, { "epoch": 0.7445617605902557, "ewc_loss": 0.08984375, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.616360187530518, "learning_rate": 1e-05, "loss": 0.4806, "mean_token_accuracy": 0.8688154220581055, "num_tokens": 222304694.0, "step": 5853 }, { "epoch": 0.7446889708688462, "ewc_loss": 0.08984375, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.560777187347412, "learning_rate": 1e-05, "loss": 0.5359, "mean_token_accuracy": 0.8510506749153137, "num_tokens": 222340793.0, "step": 5854 }, { "epoch": 0.7448161811474368, "ewc_loss": 0.08837890625, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 7.62939453125e-05, "grad_norm": 5.5618367195129395, "learning_rate": 1e-05, "loss": 0.5264, "mean_token_accuracy": 0.8554060459136963, "num_tokens": 222380316.0, "step": 5855 }, { "epoch": 0.7449433914260272, "ewc_loss": 0.0888671875, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.543154716491699, "learning_rate": 1e-05, "loss": 0.4976, "mean_token_accuracy": 0.8540513515472412, "num_tokens": 222418867.0, "step": 5856 }, { "epoch": 0.7450706017046177, "ewc_loss": 0.0888671875, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.504073143005371, "learning_rate": 1e-05, "loss": 0.4915, "mean_token_accuracy": 0.8635706901550293, "num_tokens": 222456585.0, "step": 5857 }, { "epoch": 0.7451978119832082, "ewc_loss": 0.08935546875, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.556298732757568, "learning_rate": 1e-05, "loss": 0.4757, "mean_token_accuracy": 0.8690401315689087, "num_tokens": 222492894.0, "step": 5858 }, { "epoch": 0.7453250222617988, "ewc_loss": 0.08740234375, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 7.534027099609375e-05, "grad_norm": 5.440070152282715, "learning_rate": 1e-05, "loss": 0.5054, "mean_token_accuracy": 0.8608430624008179, "num_tokens": 222538214.0, "step": 5859 }, { "epoch": 0.7454522325403893, "ewc_loss": 0.08984375, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.588705539703369, "learning_rate": 1e-05, "loss": 0.5185, "mean_token_accuracy": 0.8585866689682007, "num_tokens": 222577643.0, "step": 5860 }, { "epoch": 0.7455794428189798, "ewc_loss": 0.08544921875, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 7.343292236328125e-05, "grad_norm": 5.379897594451904, "learning_rate": 1e-05, "loss": 0.5155, "mean_token_accuracy": 0.8559983968734741, "num_tokens": 222615080.0, "step": 5861 }, { "epoch": 0.7457066530975703, "ewc_loss": 0.0908203125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.664104461669922, "learning_rate": 1e-05, "loss": 0.4743, "mean_token_accuracy": 0.8697806596755981, "num_tokens": 222650358.0, "step": 5862 }, { "epoch": 0.7458338633761608, "ewc_loss": 0.0849609375, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 7.295608520507812e-05, "grad_norm": 5.268444538116455, "learning_rate": 1e-05, "loss": 0.5563, "mean_token_accuracy": 0.8449892401695251, "num_tokens": 222691558.0, "step": 5863 }, { "epoch": 0.7459610736547513, "ewc_loss": 0.09375, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.816565990447998, "learning_rate": 1e-05, "loss": 0.5175, "mean_token_accuracy": 0.8577276468276978, "num_tokens": 222734632.0, "step": 5864 }, { "epoch": 0.7460882839333418, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 7.009506225585938e-05, "grad_norm": 5.122371673583984, "learning_rate": 1e-05, "loss": 0.5157, "mean_token_accuracy": 0.8551195859909058, "num_tokens": 222769422.0, "step": 5865 }, { "epoch": 0.7462154942119323, "ewc_loss": 0.09814453125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 6.101995468139648, "learning_rate": 1e-05, "loss": 0.4592, "mean_token_accuracy": 0.8786420822143555, "num_tokens": 222808667.0, "step": 5866 }, { "epoch": 0.7463427044905229, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 6.866455078125e-05, "grad_norm": 4.920685768127441, "learning_rate": 1e-05, "loss": 0.4847, "mean_token_accuracy": 0.8629828095436096, "num_tokens": 222844291.0, "step": 5867 }, { "epoch": 0.7464699147691134, "ewc_loss": 0.11083984375, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.640400409698486, "learning_rate": 1e-05, "loss": 0.5889, "mean_token_accuracy": 0.8419774770736694, "num_tokens": 222885812.0, "step": 5868 }, { "epoch": 0.7465971250477038, "ewc_loss": 0.08349609375, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 7.104873657226562e-05, "grad_norm": 4.882086277008057, "learning_rate": 1e-05, "loss": 0.4891, "mean_token_accuracy": 0.8604826331138611, "num_tokens": 222920768.0, "step": 5869 }, { "epoch": 0.7467243353262943, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 0.00011873245239257812, "grad_norm": 7.3736653327941895, "learning_rate": 1e-05, "loss": 0.5316, "mean_token_accuracy": 0.8632378578186035, "num_tokens": 222961973.0, "step": 5870 }, { "epoch": 0.7468515456048849, "ewc_loss": 0.10009765625, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 5.56532096862793, "learning_rate": 1e-05, "loss": 0.4937, "mean_token_accuracy": 0.8630502223968506, "num_tokens": 222997356.0, "step": 5871 }, { "epoch": 0.7469787558834754, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.570561408996582, "learning_rate": 1e-05, "loss": 0.5438, "mean_token_accuracy": 0.8655659556388855, "num_tokens": 223035933.0, "step": 5872 }, { "epoch": 0.7471059661620659, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 0.00011014938354492188, "grad_norm": 6.423541069030762, "learning_rate": 1e-05, "loss": 0.5605, "mean_token_accuracy": 0.8512697815895081, "num_tokens": 223074597.0, "step": 5873 }, { "epoch": 0.7472331764406565, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 6.974227428436279, "learning_rate": 1e-05, "loss": 0.5009, "mean_token_accuracy": 0.8711934089660645, "num_tokens": 223113736.0, "step": 5874 }, { "epoch": 0.7473603867192469, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.384395599365234, "learning_rate": 1e-05, "loss": 0.5475, "mean_token_accuracy": 0.8517627716064453, "num_tokens": 223153974.0, "step": 5875 }, { "epoch": 0.7474875969978374, "ewc_loss": 0.11083984375, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.572195053100586, "learning_rate": 1e-05, "loss": 0.5118, "mean_token_accuracy": 0.8642666339874268, "num_tokens": 223187043.0, "step": 5876 }, { "epoch": 0.7476148072764279, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.089625835418701, "learning_rate": 1e-05, "loss": 0.4825, "mean_token_accuracy": 0.8707680106163025, "num_tokens": 223221265.0, "step": 5877 }, { "epoch": 0.7477420175550185, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.305150508880615, "learning_rate": 1e-05, "loss": 0.5238, "mean_token_accuracy": 0.8563871383666992, "num_tokens": 223262167.0, "step": 5878 }, { "epoch": 0.747869227833609, "ewc_loss": 0.09619140625, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.8370361328125, "learning_rate": 1e-05, "loss": 0.5574, "mean_token_accuracy": 0.8490074872970581, "num_tokens": 223302333.0, "step": 5879 }, { "epoch": 0.7479964381121995, "ewc_loss": 0.10107421875, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.1867828369140625, "learning_rate": 1e-05, "loss": 0.4858, "mean_token_accuracy": 0.868362545967102, "num_tokens": 223343337.0, "step": 5880 }, { "epoch": 0.7481236483907899, "ewc_loss": 0.09130859375, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.603521823883057, "learning_rate": 1e-05, "loss": 0.4611, "mean_token_accuracy": 0.8729937076568604, "num_tokens": 223382905.0, "step": 5881 }, { "epoch": 0.7482508586693805, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 6.190393447875977, "learning_rate": 1e-05, "loss": 0.4954, "mean_token_accuracy": 0.8618063926696777, "num_tokens": 223422177.0, "step": 5882 }, { "epoch": 0.748378068947971, "ewc_loss": 0.087890625, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 7.581710815429688e-05, "grad_norm": 5.4273576736450195, "learning_rate": 1e-05, "loss": 0.508, "mean_token_accuracy": 0.8580960631370544, "num_tokens": 223465725.0, "step": 5883 }, { "epoch": 0.7485052792265615, "ewc_loss": 0.10400390625, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.345208168029785, "learning_rate": 1e-05, "loss": 0.5434, "mean_token_accuracy": 0.8510456681251526, "num_tokens": 223507246.0, "step": 5884 }, { "epoch": 0.748632489505152, "ewc_loss": 0.0859375, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 7.390975952148438e-05, "grad_norm": 5.30328893661499, "learning_rate": 1e-05, "loss": 0.4778, "mean_token_accuracy": 0.8673387169837952, "num_tokens": 223550736.0, "step": 5885 }, { "epoch": 0.7487596997837426, "ewc_loss": 0.10888671875, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.639223098754883, "learning_rate": 1e-05, "loss": 0.5754, "mean_token_accuracy": 0.8455774784088135, "num_tokens": 223588096.0, "step": 5886 }, { "epoch": 0.748886910062333, "ewc_loss": 0.0888671875, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.388716697692871, "learning_rate": 1e-05, "loss": 0.5109, "mean_token_accuracy": 0.8598232865333557, "num_tokens": 223629414.0, "step": 5887 }, { "epoch": 0.7490141203409235, "ewc_loss": 0.11474609375, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.85178804397583, "learning_rate": 1e-05, "loss": 0.5415, "mean_token_accuracy": 0.85770583152771, "num_tokens": 223664468.0, "step": 5888 }, { "epoch": 0.749141330619514, "ewc_loss": 0.095703125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.729572296142578, "learning_rate": 1e-05, "loss": 0.477, "mean_token_accuracy": 0.8688066005706787, "num_tokens": 223699832.0, "step": 5889 }, { "epoch": 0.7492685408981046, "ewc_loss": 0.11083984375, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.674593448638916, "learning_rate": 1e-05, "loss": 0.5355, "mean_token_accuracy": 0.8567557334899902, "num_tokens": 223736774.0, "step": 5890 }, { "epoch": 0.7493957511766951, "ewc_loss": 0.10009765625, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 5.983850479125977, "learning_rate": 1e-05, "loss": 0.5026, "mean_token_accuracy": 0.8611170053482056, "num_tokens": 223774622.0, "step": 5891 }, { "epoch": 0.7495229614552856, "ewc_loss": 0.099609375, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 6.202622413635254, "learning_rate": 1e-05, "loss": 0.5335, "mean_token_accuracy": 0.8548921346664429, "num_tokens": 223816845.0, "step": 5892 }, { "epoch": 0.749650171733876, "ewc_loss": 0.099609375, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 6.032843589782715, "learning_rate": 1e-05, "loss": 0.4807, "mean_token_accuracy": 0.8750246167182922, "num_tokens": 223849964.0, "step": 5893 }, { "epoch": 0.7497773820124666, "ewc_loss": 0.08935546875, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.701725006103516, "learning_rate": 1e-05, "loss": 0.5356, "mean_token_accuracy": 0.8504494428634644, "num_tokens": 223887821.0, "step": 5894 }, { "epoch": 0.7499045922910571, "ewc_loss": 0.10009765625, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 6.0981292724609375, "learning_rate": 1e-05, "loss": 0.4672, "mean_token_accuracy": 0.8742374181747437, "num_tokens": 223929420.0, "step": 5895 }, { "epoch": 0.7500318025696476, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 7.05718994140625e-05, "grad_norm": 5.302887439727783, "learning_rate": 1e-05, "loss": 0.4723, "mean_token_accuracy": 0.866790771484375, "num_tokens": 223964910.0, "step": 5896 }, { "epoch": 0.7501590128482382, "ewc_loss": 0.10302734375, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.307437419891357, "learning_rate": 1e-05, "loss": 0.5693, "mean_token_accuracy": 0.8411388397216797, "num_tokens": 224001131.0, "step": 5897 }, { "epoch": 0.7502862231268287, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 6.914138793945312e-05, "grad_norm": 5.096104145050049, "learning_rate": 1e-05, "loss": 0.5576, "mean_token_accuracy": 0.8432070016860962, "num_tokens": 224038969.0, "step": 5898 }, { "epoch": 0.7504134334054191, "ewc_loss": 0.11083984375, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.674659729003906, "learning_rate": 1e-05, "loss": 0.5638, "mean_token_accuracy": 0.8482640385627747, "num_tokens": 224081044.0, "step": 5899 }, { "epoch": 0.7505406436840096, "ewc_loss": 0.0859375, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 7.390975952148438e-05, "grad_norm": 5.186941146850586, "learning_rate": 1e-05, "loss": 0.596, "mean_token_accuracy": 0.83375084400177, "num_tokens": 224117967.0, "step": 5900 }, { "epoch": 0.7506678539626002, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 7.03363037109375, "learning_rate": 1e-05, "loss": 0.5883, "mean_token_accuracy": 0.8443859815597534, "num_tokens": 224148890.0, "step": 5901 }, { "epoch": 0.7507950642411907, "ewc_loss": 0.0966796875, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.673144817352295, "learning_rate": 1e-05, "loss": 0.5725, "mean_token_accuracy": 0.8420950174331665, "num_tokens": 224183592.0, "step": 5902 }, { "epoch": 0.7509222745197812, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.909217357635498, "learning_rate": 1e-05, "loss": 0.5224, "mean_token_accuracy": 0.8608882427215576, "num_tokens": 224225983.0, "step": 5903 }, { "epoch": 0.7510494847983717, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.015329360961914, "learning_rate": 1e-05, "loss": 0.5764, "mean_token_accuracy": 0.8466183543205261, "num_tokens": 224267068.0, "step": 5904 }, { "epoch": 0.7511766950769622, "ewc_loss": 0.10693359375, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.489222526550293, "learning_rate": 1e-05, "loss": 0.4635, "mean_token_accuracy": 0.8776061534881592, "num_tokens": 224313060.0, "step": 5905 }, { "epoch": 0.7513039053555527, "ewc_loss": 0.10302734375, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.014363765716553, "learning_rate": 1e-05, "loss": 0.5348, "mean_token_accuracy": 0.8521402478218079, "num_tokens": 224355066.0, "step": 5906 }, { "epoch": 0.7514311156341432, "ewc_loss": 0.0986328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 6.132745742797852, "learning_rate": 1e-05, "loss": 0.492, "mean_token_accuracy": 0.8707298040390015, "num_tokens": 224383930.0, "step": 5907 }, { "epoch": 0.7515583259127337, "ewc_loss": 0.09716796875, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.891782760620117, "learning_rate": 1e-05, "loss": 0.5717, "mean_token_accuracy": 0.8411242961883545, "num_tokens": 224418462.0, "step": 5908 }, { "epoch": 0.7516855361913243, "ewc_loss": 0.09375, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.872340679168701, "learning_rate": 1e-05, "loss": 0.5026, "mean_token_accuracy": 0.8618668913841248, "num_tokens": 224453672.0, "step": 5909 }, { "epoch": 0.7518127464699148, "ewc_loss": 0.09423828125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.740649700164795, "learning_rate": 1e-05, "loss": 0.4887, "mean_token_accuracy": 0.8632022142410278, "num_tokens": 224494093.0, "step": 5910 }, { "epoch": 0.7519399567485053, "ewc_loss": 0.0908203125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.766047954559326, "learning_rate": 1e-05, "loss": 0.5022, "mean_token_accuracy": 0.8641159534454346, "num_tokens": 224531829.0, "step": 5911 }, { "epoch": 0.7520671670270958, "ewc_loss": 0.08935546875, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.585750579833984, "learning_rate": 1e-05, "loss": 0.5363, "mean_token_accuracy": 0.8490586876869202, "num_tokens": 224569403.0, "step": 5912 }, { "epoch": 0.7521943773056863, "ewc_loss": 0.09130859375, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.774611949920654, "learning_rate": 1e-05, "loss": 0.5386, "mean_token_accuracy": 0.8534582853317261, "num_tokens": 224605189.0, "step": 5913 }, { "epoch": 0.7523215875842768, "ewc_loss": 0.08642578125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 7.43865966796875e-05, "grad_norm": 5.434051513671875, "learning_rate": 1e-05, "loss": 0.5138, "mean_token_accuracy": 0.8573621511459351, "num_tokens": 224638729.0, "step": 5914 }, { "epoch": 0.7524487978628673, "ewc_loss": 0.0927734375, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.8460893630981445, "learning_rate": 1e-05, "loss": 0.4888, "mean_token_accuracy": 0.8618873357772827, "num_tokens": 224673442.0, "step": 5915 }, { "epoch": 0.7525760081414579, "ewc_loss": 0.08349609375, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 7.152557373046875e-05, "grad_norm": 5.284291744232178, "learning_rate": 1e-05, "loss": 0.5511, "mean_token_accuracy": 0.844700276851654, "num_tokens": 224705945.0, "step": 5916 }, { "epoch": 0.7527032184200484, "ewc_loss": 0.09814453125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 6.036957263946533, "learning_rate": 1e-05, "loss": 0.5198, "mean_token_accuracy": 0.8607101440429688, "num_tokens": 224740035.0, "step": 5917 }, { "epoch": 0.7528304286986388, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 7.009506225585938e-05, "grad_norm": 5.114554405212402, "learning_rate": 1e-05, "loss": 0.4652, "mean_token_accuracy": 0.8714868426322937, "num_tokens": 224782047.0, "step": 5918 }, { "epoch": 0.7529576389772293, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.360282897949219, "learning_rate": 1e-05, "loss": 0.5421, "mean_token_accuracy": 0.8513537645339966, "num_tokens": 224821755.0, "step": 5919 }, { "epoch": 0.7530848492558199, "ewc_loss": 0.08349609375, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 7.104873657226562e-05, "grad_norm": 5.0610175132751465, "learning_rate": 1e-05, "loss": 0.4706, "mean_token_accuracy": 0.8672308921813965, "num_tokens": 224854822.0, "step": 5920 }, { "epoch": 0.7532120595344104, "ewc_loss": 0.11474609375, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.7810959815979, "learning_rate": 1e-05, "loss": 0.525, "mean_token_accuracy": 0.8642141222953796, "num_tokens": 224899570.0, "step": 5921 }, { "epoch": 0.7533392698130009, "ewc_loss": 0.09033203125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.286390781402588, "learning_rate": 1e-05, "loss": 0.4952, "mean_token_accuracy": 0.862157940864563, "num_tokens": 224934897.0, "step": 5922 }, { "epoch": 0.7534664800915915, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 7.047063827514648, "learning_rate": 1e-05, "loss": 0.5027, "mean_token_accuracy": 0.8702112436294556, "num_tokens": 224972866.0, "step": 5923 }, { "epoch": 0.7535936903701819, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 5.813535690307617, "learning_rate": 1e-05, "loss": 0.5748, "mean_token_accuracy": 0.8442352414131165, "num_tokens": 225015075.0, "step": 5924 }, { "epoch": 0.7537209006487724, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.772723197937012, "learning_rate": 1e-05, "loss": 0.5261, "mean_token_accuracy": 0.8615490198135376, "num_tokens": 225046729.0, "step": 5925 }, { "epoch": 0.7538481109273629, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.070552349090576, "learning_rate": 1e-05, "loss": 0.5543, "mean_token_accuracy": 0.854084849357605, "num_tokens": 225080117.0, "step": 5926 }, { "epoch": 0.7539753212059535, "ewc_loss": 0.10400390625, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.270345211029053, "learning_rate": 1e-05, "loss": 0.5198, "mean_token_accuracy": 0.8584856390953064, "num_tokens": 225116854.0, "step": 5927 }, { "epoch": 0.754102531484544, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 6.025888919830322, "learning_rate": 1e-05, "loss": 0.4963, "mean_token_accuracy": 0.8659405708312988, "num_tokens": 225154044.0, "step": 5928 }, { "epoch": 0.7542297417631345, "ewc_loss": 0.091796875, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.832773208618164, "learning_rate": 1e-05, "loss": 0.4569, "mean_token_accuracy": 0.8801166415214539, "num_tokens": 225191393.0, "step": 5929 }, { "epoch": 0.7543569520417249, "ewc_loss": 0.0986328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 6.007230758666992, "learning_rate": 1e-05, "loss": 0.5492, "mean_token_accuracy": 0.8497304916381836, "num_tokens": 225232577.0, "step": 5930 }, { "epoch": 0.7544841623203155, "ewc_loss": 0.08837890625, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 7.62939453125e-05, "grad_norm": 5.519586563110352, "learning_rate": 1e-05, "loss": 0.5489, "mean_token_accuracy": 0.8462738394737244, "num_tokens": 225270458.0, "step": 5931 }, { "epoch": 0.754611372598906, "ewc_loss": 0.099609375, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 6.070917129516602, "learning_rate": 1e-05, "loss": 0.5244, "mean_token_accuracy": 0.8574994802474976, "num_tokens": 225306777.0, "step": 5932 }, { "epoch": 0.7547385828774965, "ewc_loss": 0.0849609375, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 7.295608520507812e-05, "grad_norm": 5.314784049987793, "learning_rate": 1e-05, "loss": 0.482, "mean_token_accuracy": 0.8664863109588623, "num_tokens": 225345602.0, "step": 5933 }, { "epoch": 0.754865793156087, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.222256660461426, "learning_rate": 1e-05, "loss": 0.4959, "mean_token_accuracy": 0.8659535646438599, "num_tokens": 225378885.0, "step": 5934 }, { "epoch": 0.7549930034346776, "ewc_loss": 0.0859375, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 7.390975952148438e-05, "grad_norm": 5.236310958862305, "learning_rate": 1e-05, "loss": 0.4973, "mean_token_accuracy": 0.8608605265617371, "num_tokens": 225408421.0, "step": 5935 }, { "epoch": 0.755120213713268, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.420609474182129, "learning_rate": 1e-05, "loss": 0.5157, "mean_token_accuracy": 0.8609527349472046, "num_tokens": 225441428.0, "step": 5936 }, { "epoch": 0.7552474239918585, "ewc_loss": 0.08837890625, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 7.62939453125e-05, "grad_norm": 5.2520318031311035, "learning_rate": 1e-05, "loss": 0.5386, "mean_token_accuracy": 0.8500949144363403, "num_tokens": 225474912.0, "step": 5937 }, { "epoch": 0.755374634270449, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.616817474365234, "learning_rate": 1e-05, "loss": 0.5484, "mean_token_accuracy": 0.8558378219604492, "num_tokens": 225511876.0, "step": 5938 }, { "epoch": 0.7555018445490396, "ewc_loss": 0.091796875, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.405096530914307, "learning_rate": 1e-05, "loss": 0.5183, "mean_token_accuracy": 0.8574569225311279, "num_tokens": 225548754.0, "step": 5939 }, { "epoch": 0.7556290548276301, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.666541576385498, "learning_rate": 1e-05, "loss": 0.5625, "mean_token_accuracy": 0.848789393901825, "num_tokens": 225585554.0, "step": 5940 }, { "epoch": 0.7557562651062206, "ewc_loss": 0.09619140625, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.614743232727051, "learning_rate": 1e-05, "loss": 0.4948, "mean_token_accuracy": 0.8638710975646973, "num_tokens": 225622727.0, "step": 5941 }, { "epoch": 0.755883475384811, "ewc_loss": 0.10791015625, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.511162281036377, "learning_rate": 1e-05, "loss": 0.4867, "mean_token_accuracy": 0.8696057796478271, "num_tokens": 225662163.0, "step": 5942 }, { "epoch": 0.7560106856634016, "ewc_loss": 0.09765625, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.7262372970581055, "learning_rate": 1e-05, "loss": 0.5476, "mean_token_accuracy": 0.8458758592605591, "num_tokens": 225698932.0, "step": 5943 }, { "epoch": 0.7561378959419921, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.246643543243408, "learning_rate": 1e-05, "loss": 0.4924, "mean_token_accuracy": 0.8695115447044373, "num_tokens": 225735587.0, "step": 5944 }, { "epoch": 0.7562651062205826, "ewc_loss": 0.0947265625, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.7304253578186035, "learning_rate": 1e-05, "loss": 0.5019, "mean_token_accuracy": 0.8716047406196594, "num_tokens": 225777013.0, "step": 5945 }, { "epoch": 0.7563923164991732, "ewc_loss": 0.0966796875, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.986846923828125, "learning_rate": 1e-05, "loss": 0.4986, "mean_token_accuracy": 0.8631622791290283, "num_tokens": 225808496.0, "step": 5946 }, { "epoch": 0.7565195267777637, "ewc_loss": 0.0947265625, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.701387882232666, "learning_rate": 1e-05, "loss": 0.495, "mean_token_accuracy": 0.8657505512237549, "num_tokens": 225848629.0, "step": 5947 }, { "epoch": 0.7566467370563541, "ewc_loss": 0.09326171875, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.780889511108398, "learning_rate": 1e-05, "loss": 0.4805, "mean_token_accuracy": 0.8687601089477539, "num_tokens": 225882284.0, "step": 5948 }, { "epoch": 0.7567739473349446, "ewc_loss": 0.09326171875, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.642129898071289, "learning_rate": 1e-05, "loss": 0.5433, "mean_token_accuracy": 0.8494716286659241, "num_tokens": 225919294.0, "step": 5949 }, { "epoch": 0.7569011576135352, "ewc_loss": 0.0908203125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.679370403289795, "learning_rate": 1e-05, "loss": 0.5513, "mean_token_accuracy": 0.8459062576293945, "num_tokens": 225954742.0, "step": 5950 }, { "epoch": 0.7570283678921257, "ewc_loss": 0.09033203125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.564455986022949, "learning_rate": 1e-05, "loss": 0.4363, "mean_token_accuracy": 0.8838750123977661, "num_tokens": 225992532.0, "step": 5951 }, { "epoch": 0.7571555781707162, "ewc_loss": 0.09130859375, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.631446838378906, "learning_rate": 1e-05, "loss": 0.512, "mean_token_accuracy": 0.86033695936203, "num_tokens": 226034272.0, "step": 5952 }, { "epoch": 0.7572827884493067, "ewc_loss": 0.08837890625, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 7.62939453125e-05, "grad_norm": 5.498103618621826, "learning_rate": 1e-05, "loss": 0.5181, "mean_token_accuracy": 0.8528025150299072, "num_tokens": 226066607.0, "step": 5953 }, { "epoch": 0.7574099987278972, "ewc_loss": 0.091796875, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.655928611755371, "learning_rate": 1e-05, "loss": 0.5333, "mean_token_accuracy": 0.8515781760215759, "num_tokens": 226103077.0, "step": 5954 }, { "epoch": 0.7575372090064877, "ewc_loss": 0.08642578125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 7.43865966796875e-05, "grad_norm": 5.3995513916015625, "learning_rate": 1e-05, "loss": 0.4418, "mean_token_accuracy": 0.8747780323028564, "num_tokens": 226140488.0, "step": 5955 }, { "epoch": 0.7576644192850782, "ewc_loss": 0.0927734375, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.730916976928711, "learning_rate": 1e-05, "loss": 0.4506, "mean_token_accuracy": 0.8760006427764893, "num_tokens": 226177139.0, "step": 5956 }, { "epoch": 0.7577916295636687, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 7.200241088867188e-05, "grad_norm": 5.273886203765869, "learning_rate": 1e-05, "loss": 0.4853, "mean_token_accuracy": 0.8639602661132812, "num_tokens": 226216340.0, "step": 5957 }, { "epoch": 0.7579188398422593, "ewc_loss": 0.0947265625, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.89195442199707, "learning_rate": 1e-05, "loss": 0.4535, "mean_token_accuracy": 0.8769500851631165, "num_tokens": 226253507.0, "step": 5958 }, { "epoch": 0.7580460501208498, "ewc_loss": 0.08349609375, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 7.104873657226562e-05, "grad_norm": 5.126751899719238, "learning_rate": 1e-05, "loss": 0.516, "mean_token_accuracy": 0.8546011447906494, "num_tokens": 226296617.0, "step": 5959 }, { "epoch": 0.7581732603994403, "ewc_loss": 0.10205078125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.186176300048828, "learning_rate": 1e-05, "loss": 0.5318, "mean_token_accuracy": 0.8547597527503967, "num_tokens": 226335483.0, "step": 5960 }, { "epoch": 0.7583004706780307, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 6.961822509765625e-05, "grad_norm": 4.9973320960998535, "learning_rate": 1e-05, "loss": 0.5278, "mean_token_accuracy": 0.8486174941062927, "num_tokens": 226373773.0, "step": 5961 }, { "epoch": 0.7584276809566213, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.645336151123047, "learning_rate": 1e-05, "loss": 0.5167, "mean_token_accuracy": 0.8630721569061279, "num_tokens": 226410333.0, "step": 5962 }, { "epoch": 0.7585548912352118, "ewc_loss": 0.0859375, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 7.390975952148438e-05, "grad_norm": 5.02515172958374, "learning_rate": 1e-05, "loss": 0.4957, "mean_token_accuracy": 0.8614240884780884, "num_tokens": 226451058.0, "step": 5963 }, { "epoch": 0.7586821015138023, "ewc_loss": 0.125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 0.00011301040649414062, "grad_norm": 7.165030002593994, "learning_rate": 1e-05, "loss": 0.5175, "mean_token_accuracy": 0.8638379573822021, "num_tokens": 226487099.0, "step": 5964 }, { "epoch": 0.7588093117923929, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 5.584036827087402, "learning_rate": 1e-05, "loss": 0.4957, "mean_token_accuracy": 0.8650966882705688, "num_tokens": 226521462.0, "step": 5965 }, { "epoch": 0.7589365220709834, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 0.00011396408081054688, "grad_norm": 7.1903157234191895, "learning_rate": 1e-05, "loss": 0.5466, "mean_token_accuracy": 0.8591452240943909, "num_tokens": 226560573.0, "step": 5966 }, { "epoch": 0.7590637323495738, "ewc_loss": 0.11181640625, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.130474090576172, "learning_rate": 1e-05, "loss": 0.5343, "mean_token_accuracy": 0.8563526272773743, "num_tokens": 226603717.0, "step": 5967 }, { "epoch": 0.7591909426281643, "ewc_loss": 0.11376953125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.736551284790039, "learning_rate": 1e-05, "loss": 0.5193, "mean_token_accuracy": 0.8645337224006653, "num_tokens": 226647049.0, "step": 5968 }, { "epoch": 0.7593181529067549, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.14543342590332, "learning_rate": 1e-05, "loss": 0.4918, "mean_token_accuracy": 0.8686370253562927, "num_tokens": 226682459.0, "step": 5969 }, { "epoch": 0.7594453631853454, "ewc_loss": 0.10498046875, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.326915740966797, "learning_rate": 1e-05, "loss": 0.5359, "mean_token_accuracy": 0.8549032807350159, "num_tokens": 226727381.0, "step": 5970 }, { "epoch": 0.7595725734639359, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 5.973643779754639, "learning_rate": 1e-05, "loss": 0.5742, "mean_token_accuracy": 0.8408869504928589, "num_tokens": 226766732.0, "step": 5971 }, { "epoch": 0.7596997837425264, "ewc_loss": 0.09765625, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 6.024717807769775, "learning_rate": 1e-05, "loss": 0.5445, "mean_token_accuracy": 0.8557816743850708, "num_tokens": 226808165.0, "step": 5972 }, { "epoch": 0.7598269940211169, "ewc_loss": 0.09619140625, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.8228325843811035, "learning_rate": 1e-05, "loss": 0.5446, "mean_token_accuracy": 0.8534530401229858, "num_tokens": 226848113.0, "step": 5973 }, { "epoch": 0.7599542042997074, "ewc_loss": 0.0947265625, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.8457865715026855, "learning_rate": 1e-05, "loss": 0.5826, "mean_token_accuracy": 0.8391937017440796, "num_tokens": 226889390.0, "step": 5974 }, { "epoch": 0.7600814145782979, "ewc_loss": 0.09228515625, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.679027557373047, "learning_rate": 1e-05, "loss": 0.5483, "mean_token_accuracy": 0.8494476079940796, "num_tokens": 226929111.0, "step": 5975 }, { "epoch": 0.7602086248568884, "ewc_loss": 0.091796875, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.753181457519531, "learning_rate": 1e-05, "loss": 0.4589, "mean_token_accuracy": 0.8735687136650085, "num_tokens": 226966677.0, "step": 5976 }, { "epoch": 0.760335835135479, "ewc_loss": 0.0888671875, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.548863410949707, "learning_rate": 1e-05, "loss": 0.5143, "mean_token_accuracy": 0.8576947450637817, "num_tokens": 227004083.0, "step": 5977 }, { "epoch": 0.7604630454140695, "ewc_loss": 0.091796875, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.750269889831543, "learning_rate": 1e-05, "loss": 0.5284, "mean_token_accuracy": 0.856590747833252, "num_tokens": 227049954.0, "step": 5978 }, { "epoch": 0.7605902556926599, "ewc_loss": 0.0859375, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 7.390975952148438e-05, "grad_norm": 5.409778118133545, "learning_rate": 1e-05, "loss": 0.494, "mean_token_accuracy": 0.8590196371078491, "num_tokens": 227088532.0, "step": 5979 }, { "epoch": 0.7607174659712505, "ewc_loss": 0.09326171875, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.833646774291992, "learning_rate": 1e-05, "loss": 0.4694, "mean_token_accuracy": 0.8747097253799438, "num_tokens": 227129471.0, "step": 5980 }, { "epoch": 0.760844676249841, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 7.200241088867188e-05, "grad_norm": 5.258618354797363, "learning_rate": 1e-05, "loss": 0.5096, "mean_token_accuracy": 0.8557006120681763, "num_tokens": 227167443.0, "step": 5981 }, { "epoch": 0.7609718865284315, "ewc_loss": 0.09912109375, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 6.0281758308410645, "learning_rate": 1e-05, "loss": 0.4902, "mean_token_accuracy": 0.8695924878120422, "num_tokens": 227207577.0, "step": 5982 }, { "epoch": 0.761099096807022, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 6.961822509765625e-05, "grad_norm": 5.098459720611572, "learning_rate": 1e-05, "loss": 0.4979, "mean_token_accuracy": 0.8598183393478394, "num_tokens": 227247424.0, "step": 5983 }, { "epoch": 0.7612263070856126, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.38494348526001, "learning_rate": 1e-05, "loss": 0.5235, "mean_token_accuracy": 0.8595410585403442, "num_tokens": 227286361.0, "step": 5984 }, { "epoch": 0.761353517364203, "ewc_loss": 0.08349609375, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 7.104873657226562e-05, "grad_norm": 5.000519752502441, "learning_rate": 1e-05, "loss": 0.4761, "mean_token_accuracy": 0.8670328259468079, "num_tokens": 227326786.0, "step": 5985 }, { "epoch": 0.7614807276427935, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.903076171875, "learning_rate": 1e-05, "loss": 0.5623, "mean_token_accuracy": 0.8492423295974731, "num_tokens": 227370237.0, "step": 5986 }, { "epoch": 0.761607937921384, "ewc_loss": 0.091796875, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.291417121887207, "learning_rate": 1e-05, "loss": 0.4444, "mean_token_accuracy": 0.8790247440338135, "num_tokens": 227399711.0, "step": 5987 }, { "epoch": 0.7617351481999746, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 0.00011444091796875, "grad_norm": 7.212810516357422, "learning_rate": 1e-05, "loss": 0.5278, "mean_token_accuracy": 0.8667590022087097, "num_tokens": 227436468.0, "step": 5988 }, { "epoch": 0.7618623584785651, "ewc_loss": 0.10595703125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 5.921074867248535, "learning_rate": 1e-05, "loss": 0.5334, "mean_token_accuracy": 0.8584161400794983, "num_tokens": 227476128.0, "step": 5989 }, { "epoch": 0.7619895687571556, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.8989973068237305, "learning_rate": 1e-05, "loss": 0.5231, "mean_token_accuracy": 0.8617426156997681, "num_tokens": 227510499.0, "step": 5990 }, { "epoch": 0.762116779035746, "ewc_loss": 0.109375, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.184576988220215, "learning_rate": 1e-05, "loss": 0.5326, "mean_token_accuracy": 0.856557309627533, "num_tokens": 227542886.0, "step": 5991 }, { "epoch": 0.7622439893143366, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.399803161621094, "learning_rate": 1e-05, "loss": 0.4886, "mean_token_accuracy": 0.8688098788261414, "num_tokens": 227583244.0, "step": 5992 }, { "epoch": 0.7623711995929271, "ewc_loss": 0.10400390625, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.075100898742676, "learning_rate": 1e-05, "loss": 0.5009, "mean_token_accuracy": 0.8665938377380371, "num_tokens": 227622690.0, "step": 5993 }, { "epoch": 0.7624984098715176, "ewc_loss": 0.095703125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 6.00259256362915, "learning_rate": 1e-05, "loss": 0.5213, "mean_token_accuracy": 0.8577130436897278, "num_tokens": 227663537.0, "step": 5994 }, { "epoch": 0.7626256201501082, "ewc_loss": 0.09912109375, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.960015773773193, "learning_rate": 1e-05, "loss": 0.5095, "mean_token_accuracy": 0.8620377779006958, "num_tokens": 227700679.0, "step": 5995 }, { "epoch": 0.7627528304286987, "ewc_loss": 0.091796875, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.7328782081604, "learning_rate": 1e-05, "loss": 0.5454, "mean_token_accuracy": 0.8515924215316772, "num_tokens": 227738410.0, "step": 5996 }, { "epoch": 0.7628800407072891, "ewc_loss": 0.0966796875, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.888621807098389, "learning_rate": 1e-05, "loss": 0.5043, "mean_token_accuracy": 0.864744246006012, "num_tokens": 227780849.0, "step": 5997 }, { "epoch": 0.7630072509858796, "ewc_loss": 0.0888671875, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.551877975463867, "learning_rate": 1e-05, "loss": 0.4791, "mean_token_accuracy": 0.8682349920272827, "num_tokens": 227815488.0, "step": 5998 }, { "epoch": 0.7631344612644702, "ewc_loss": 0.0947265625, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.8688483238220215, "learning_rate": 1e-05, "loss": 0.4884, "mean_token_accuracy": 0.8668832778930664, "num_tokens": 227848627.0, "step": 5999 }, { "epoch": 0.7632616715430607, "ewc_loss": 0.0869140625, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 7.486343383789062e-05, "grad_norm": 5.410447120666504, "learning_rate": 1e-05, "loss": 0.4827, "mean_token_accuracy": 0.8610302209854126, "num_tokens": 227888394.0, "step": 6000 }, { "epoch": 0.7633888818216512, "ewc_loss": 0.095703125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.9221930503845215, "learning_rate": 1e-05, "loss": 0.4889, "mean_token_accuracy": 0.867547333240509, "num_tokens": 227923785.0, "step": 6001 }, { "epoch": 0.7635160921002417, "ewc_loss": 0.0849609375, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 7.295608520507812e-05, "grad_norm": 5.286863327026367, "learning_rate": 1e-05, "loss": 0.5232, "mean_token_accuracy": 0.8529118299484253, "num_tokens": 227958714.0, "step": 6002 }, { "epoch": 0.7636433023788322, "ewc_loss": 0.09912109375, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 6.06559944152832, "learning_rate": 1e-05, "loss": 0.5919, "mean_token_accuracy": 0.8342573642730713, "num_tokens": 228004414.0, "step": 6003 }, { "epoch": 0.7637705126574227, "ewc_loss": 0.08349609375, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 7.152557373046875e-05, "grad_norm": 5.1618146896362305, "learning_rate": 1e-05, "loss": 0.5503, "mean_token_accuracy": 0.8461576700210571, "num_tokens": 228044291.0, "step": 6004 }, { "epoch": 0.7638977229360132, "ewc_loss": 0.10302734375, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.312071323394775, "learning_rate": 1e-05, "loss": 0.4999, "mean_token_accuracy": 0.8675587177276611, "num_tokens": 228080656.0, "step": 6005 }, { "epoch": 0.7640249332146037, "ewc_loss": 0.0849609375, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 7.295608520507812e-05, "grad_norm": 5.088606357574463, "learning_rate": 1e-05, "loss": 0.4973, "mean_token_accuracy": 0.8615944981575012, "num_tokens": 228117787.0, "step": 6006 }, { "epoch": 0.7641521434931943, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.678675651550293, "learning_rate": 1e-05, "loss": 0.5489, "mean_token_accuracy": 0.8540104031562805, "num_tokens": 228150620.0, "step": 6007 }, { "epoch": 0.7642793537717848, "ewc_loss": 0.08935546875, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.212355136871338, "learning_rate": 1e-05, "loss": 0.4784, "mean_token_accuracy": 0.8708537817001343, "num_tokens": 228188330.0, "step": 6008 }, { "epoch": 0.7644065640503753, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 6.9599690437316895, "learning_rate": 1e-05, "loss": 0.514, "mean_token_accuracy": 0.8670188188552856, "num_tokens": 228231369.0, "step": 6009 }, { "epoch": 0.7645337743289657, "ewc_loss": 0.09912109375, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.6184258460998535, "learning_rate": 1e-05, "loss": 0.4889, "mean_token_accuracy": 0.8680530190467834, "num_tokens": 228269321.0, "step": 6010 }, { "epoch": 0.7646609846075563, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.882545471191406, "learning_rate": 1e-05, "loss": 0.5121, "mean_token_accuracy": 0.86761873960495, "num_tokens": 228305240.0, "step": 6011 }, { "epoch": 0.7647881948861468, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 5.9376959800720215, "learning_rate": 1e-05, "loss": 0.5227, "mean_token_accuracy": 0.8618806600570679, "num_tokens": 228345100.0, "step": 6012 }, { "epoch": 0.7649154051647373, "ewc_loss": 0.10693359375, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.494758605957031, "learning_rate": 1e-05, "loss": 0.499, "mean_token_accuracy": 0.8645003437995911, "num_tokens": 228379215.0, "step": 6013 }, { "epoch": 0.7650426154433279, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 5.97857666015625, "learning_rate": 1e-05, "loss": 0.5269, "mean_token_accuracy": 0.8573222160339355, "num_tokens": 228415954.0, "step": 6014 }, { "epoch": 0.7651698257219184, "ewc_loss": 0.09814453125, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 6.086677551269531, "learning_rate": 1e-05, "loss": 0.5549, "mean_token_accuracy": 0.8453832864761353, "num_tokens": 228453569.0, "step": 6015 }, { "epoch": 0.7652970360005088, "ewc_loss": 0.09912109375, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.916783809661865, "learning_rate": 1e-05, "loss": 0.5819, "mean_token_accuracy": 0.8440210819244385, "num_tokens": 228491384.0, "step": 6016 }, { "epoch": 0.7654242462790993, "ewc_loss": 0.09228515625, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.753053188323975, "learning_rate": 1e-05, "loss": 0.5581, "mean_token_accuracy": 0.8434618711471558, "num_tokens": 228528118.0, "step": 6017 }, { "epoch": 0.7655514565576899, "ewc_loss": 0.09716796875, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.879793167114258, "learning_rate": 1e-05, "loss": 0.5303, "mean_token_accuracy": 0.8543054461479187, "num_tokens": 228568207.0, "step": 6018 }, { "epoch": 0.7656786668362804, "ewc_loss": 0.08837890625, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 7.62939453125e-05, "grad_norm": 5.525594711303711, "learning_rate": 1e-05, "loss": 0.5035, "mean_token_accuracy": 0.8589653372764587, "num_tokens": 228604313.0, "step": 6019 }, { "epoch": 0.7658058771148709, "ewc_loss": 0.0966796875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.875237464904785, "learning_rate": 1e-05, "loss": 0.4632, "mean_token_accuracy": 0.874014139175415, "num_tokens": 228645615.0, "step": 6020 }, { "epoch": 0.7659330873934614, "ewc_loss": 0.0869140625, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 7.486343383789062e-05, "grad_norm": 5.386693477630615, "learning_rate": 1e-05, "loss": 0.5104, "mean_token_accuracy": 0.8556208610534668, "num_tokens": 228679137.0, "step": 6021 }, { "epoch": 0.7660602976720519, "ewc_loss": 0.09716796875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.930788993835449, "learning_rate": 1e-05, "loss": 0.4847, "mean_token_accuracy": 0.8665732741355896, "num_tokens": 228717383.0, "step": 6022 }, { "epoch": 0.7661875079506424, "ewc_loss": 0.08544921875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 7.343292236328125e-05, "grad_norm": 5.272913455963135, "learning_rate": 1e-05, "loss": 0.4592, "mean_token_accuracy": 0.8712838888168335, "num_tokens": 228747260.0, "step": 6023 }, { "epoch": 0.7663147182292329, "ewc_loss": 0.10009765625, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 6.061728000640869, "learning_rate": 1e-05, "loss": 0.5314, "mean_token_accuracy": 0.8535237908363342, "num_tokens": 228780852.0, "step": 6024 }, { "epoch": 0.7664419285078234, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 7.200241088867188e-05, "grad_norm": 5.148684501647949, "learning_rate": 1e-05, "loss": 0.4919, "mean_token_accuracy": 0.8649924993515015, "num_tokens": 228820220.0, "step": 6025 }, { "epoch": 0.766569138786414, "ewc_loss": 0.10400390625, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.317248344421387, "learning_rate": 1e-05, "loss": 0.4967, "mean_token_accuracy": 0.8650296926498413, "num_tokens": 228858903.0, "step": 6026 }, { "epoch": 0.7666963490650045, "ewc_loss": 0.08642578125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 7.43865966796875e-05, "grad_norm": 5.085103511810303, "learning_rate": 1e-05, "loss": 0.464, "mean_token_accuracy": 0.8727375268936157, "num_tokens": 228894998.0, "step": 6027 }, { "epoch": 0.7668235593435949, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.6682047843933105, "learning_rate": 1e-05, "loss": 0.5038, "mean_token_accuracy": 0.8657475113868713, "num_tokens": 228933333.0, "step": 6028 }, { "epoch": 0.7669507696221854, "ewc_loss": 0.0908203125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.231541156768799, "learning_rate": 1e-05, "loss": 0.4952, "mean_token_accuracy": 0.8627768754959106, "num_tokens": 228972960.0, "step": 6029 }, { "epoch": 0.767077979900776, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.922659873962402, "learning_rate": 1e-05, "loss": 0.5666, "mean_token_accuracy": 0.850597620010376, "num_tokens": 229010967.0, "step": 6030 }, { "epoch": 0.7672051901793665, "ewc_loss": 0.099609375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 5.617702960968018, "learning_rate": 1e-05, "loss": 0.4935, "mean_token_accuracy": 0.8666919469833374, "num_tokens": 229052424.0, "step": 6031 }, { "epoch": 0.767332400457957, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.821643352508545, "learning_rate": 1e-05, "loss": 0.5363, "mean_token_accuracy": 0.8590378165245056, "num_tokens": 229088106.0, "step": 6032 }, { "epoch": 0.7674596107365476, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 5.912732124328613, "learning_rate": 1e-05, "loss": 0.5098, "mean_token_accuracy": 0.864044189453125, "num_tokens": 229122340.0, "step": 6033 }, { "epoch": 0.767586821015138, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.439915657043457, "learning_rate": 1e-05, "loss": 0.5648, "mean_token_accuracy": 0.8513956069946289, "num_tokens": 229158299.0, "step": 6034 }, { "epoch": 0.7677140312937285, "ewc_loss": 0.10400390625, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 5.948644638061523, "learning_rate": 1e-05, "loss": 0.5131, "mean_token_accuracy": 0.8592445850372314, "num_tokens": 229187870.0, "step": 6035 }, { "epoch": 0.767841241572319, "ewc_loss": 0.0986328125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 6.02440881729126, "learning_rate": 1e-05, "loss": 0.5489, "mean_token_accuracy": 0.8476340770721436, "num_tokens": 229226459.0, "step": 6036 }, { "epoch": 0.7679684518509096, "ewc_loss": 0.09912109375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.904872417449951, "learning_rate": 1e-05, "loss": 0.5084, "mean_token_accuracy": 0.8613331913948059, "num_tokens": 229257973.0, "step": 6037 }, { "epoch": 0.7680956621295001, "ewc_loss": 0.09130859375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.69938325881958, "learning_rate": 1e-05, "loss": 0.4968, "mean_token_accuracy": 0.8646485805511475, "num_tokens": 229297870.0, "step": 6038 }, { "epoch": 0.7682228724080906, "ewc_loss": 0.09765625, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.889528751373291, "learning_rate": 1e-05, "loss": 0.5423, "mean_token_accuracy": 0.8535850048065186, "num_tokens": 229331295.0, "step": 6039 }, { "epoch": 0.768350082686681, "ewc_loss": 0.08837890625, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 7.62939453125e-05, "grad_norm": 5.478564739227295, "learning_rate": 1e-05, "loss": 0.5096, "mean_token_accuracy": 0.8581322431564331, "num_tokens": 229369491.0, "step": 6040 }, { "epoch": 0.7684772929652716, "ewc_loss": 0.09814453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.910655498504639, "learning_rate": 1e-05, "loss": 0.5183, "mean_token_accuracy": 0.8544254302978516, "num_tokens": 229407836.0, "step": 6041 }, { "epoch": 0.7686045032438621, "ewc_loss": 0.0869140625, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 7.486343383789062e-05, "grad_norm": 5.340912818908691, "learning_rate": 1e-05, "loss": 0.482, "mean_token_accuracy": 0.864944577217102, "num_tokens": 229444806.0, "step": 6042 }, { "epoch": 0.7687317135224526, "ewc_loss": 0.099609375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 5.995689868927002, "learning_rate": 1e-05, "loss": 0.5032, "mean_token_accuracy": 0.865708589553833, "num_tokens": 229485588.0, "step": 6043 }, { "epoch": 0.7688589238010431, "ewc_loss": 0.0849609375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 7.295608520507812e-05, "grad_norm": 5.218028545379639, "learning_rate": 1e-05, "loss": 0.4816, "mean_token_accuracy": 0.8675589561462402, "num_tokens": 229521924.0, "step": 6044 }, { "epoch": 0.7689861340796337, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 6.175945281982422, "learning_rate": 1e-05, "loss": 0.498, "mean_token_accuracy": 0.8646892309188843, "num_tokens": 229561511.0, "step": 6045 }, { "epoch": 0.7691133443582241, "ewc_loss": 0.0859375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 7.390975952148438e-05, "grad_norm": 5.114384174346924, "learning_rate": 1e-05, "loss": 0.499, "mean_token_accuracy": 0.8604669570922852, "num_tokens": 229607193.0, "step": 6046 }, { "epoch": 0.7692405546368146, "ewc_loss": 0.10791015625, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.485710620880127, "learning_rate": 1e-05, "loss": 0.4936, "mean_token_accuracy": 0.8686914443969727, "num_tokens": 229642775.0, "step": 6047 }, { "epoch": 0.7693677649154052, "ewc_loss": 0.087890625, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 7.581710815429688e-05, "grad_norm": 5.140710830688477, "learning_rate": 1e-05, "loss": 0.545, "mean_token_accuracy": 0.8470407724380493, "num_tokens": 229680960.0, "step": 6048 }, { "epoch": 0.7694949751939957, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.8116278648376465, "learning_rate": 1e-05, "loss": 0.4866, "mean_token_accuracy": 0.8741815686225891, "num_tokens": 229713473.0, "step": 6049 }, { "epoch": 0.7696221854725862, "ewc_loss": 0.09375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.39957332611084, "learning_rate": 1e-05, "loss": 0.4733, "mean_token_accuracy": 0.8725610971450806, "num_tokens": 229753919.0, "step": 6050 }, { "epoch": 0.7697493957511767, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.916516304016113, "learning_rate": 1e-05, "loss": 0.5935, "mean_token_accuracy": 0.8450034260749817, "num_tokens": 229788635.0, "step": 6051 }, { "epoch": 0.7698766060297672, "ewc_loss": 0.10400390625, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 5.783489227294922, "learning_rate": 1e-05, "loss": 0.5094, "mean_token_accuracy": 0.8660460710525513, "num_tokens": 229831221.0, "step": 6052 }, { "epoch": 0.7700038163083577, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.651379585266113, "learning_rate": 1e-05, "loss": 0.5593, "mean_token_accuracy": 0.8492473363876343, "num_tokens": 229866762.0, "step": 6053 }, { "epoch": 0.7701310265869482, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 5.935755729675293, "learning_rate": 1e-05, "loss": 0.5015, "mean_token_accuracy": 0.86651611328125, "num_tokens": 229904715.0, "step": 6054 }, { "epoch": 0.7702582368655387, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.235994815826416, "learning_rate": 1e-05, "loss": 0.4965, "mean_token_accuracy": 0.8634161949157715, "num_tokens": 229942938.0, "step": 6055 }, { "epoch": 0.7703854471441293, "ewc_loss": 0.10205078125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 5.914344787597656, "learning_rate": 1e-05, "loss": 0.4926, "mean_token_accuracy": 0.8668686151504517, "num_tokens": 229982468.0, "step": 6056 }, { "epoch": 0.7705126574227198, "ewc_loss": 0.0947265625, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.8709564208984375, "learning_rate": 1e-05, "loss": 0.5273, "mean_token_accuracy": 0.8537037968635559, "num_tokens": 230022386.0, "step": 6057 }, { "epoch": 0.7706398677013102, "ewc_loss": 0.09814453125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.889157295227051, "learning_rate": 1e-05, "loss": 0.5609, "mean_token_accuracy": 0.8489944934844971, "num_tokens": 230054871.0, "step": 6058 }, { "epoch": 0.7707670779799007, "ewc_loss": 0.09130859375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.603367805480957, "learning_rate": 1e-05, "loss": 0.4632, "mean_token_accuracy": 0.8694368600845337, "num_tokens": 230090996.0, "step": 6059 }, { "epoch": 0.7708942882584913, "ewc_loss": 0.09716796875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.878818035125732, "learning_rate": 1e-05, "loss": 0.4818, "mean_token_accuracy": 0.865717351436615, "num_tokens": 230124575.0, "step": 6060 }, { "epoch": 0.7710214985370818, "ewc_loss": 0.08740234375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 7.534027099609375e-05, "grad_norm": 5.429000377655029, "learning_rate": 1e-05, "loss": 0.5697, "mean_token_accuracy": 0.841175377368927, "num_tokens": 230168225.0, "step": 6061 }, { "epoch": 0.7711487088156723, "ewc_loss": 0.0966796875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.91047477722168, "learning_rate": 1e-05, "loss": 0.5189, "mean_token_accuracy": 0.8590458631515503, "num_tokens": 230208004.0, "step": 6062 }, { "epoch": 0.7712759190942629, "ewc_loss": 0.0869140625, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 7.486343383789062e-05, "grad_norm": 5.31014347076416, "learning_rate": 1e-05, "loss": 0.5353, "mean_token_accuracy": 0.8534220457077026, "num_tokens": 230243247.0, "step": 6063 }, { "epoch": 0.7714031293728534, "ewc_loss": 0.10009765625, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 6.013672351837158, "learning_rate": 1e-05, "loss": 0.4851, "mean_token_accuracy": 0.8672728538513184, "num_tokens": 230286646.0, "step": 6064 }, { "epoch": 0.7715303396514438, "ewc_loss": 0.08447265625, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 7.2479248046875e-05, "grad_norm": 5.191619873046875, "learning_rate": 1e-05, "loss": 0.5555, "mean_token_accuracy": 0.8400082588195801, "num_tokens": 230317705.0, "step": 6065 }, { "epoch": 0.7716575499300343, "ewc_loss": 0.10205078125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.212701320648193, "learning_rate": 1e-05, "loss": 0.5414, "mean_token_accuracy": 0.8523504734039307, "num_tokens": 230362523.0, "step": 6066 }, { "epoch": 0.7717847602086249, "ewc_loss": 0.0849609375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 7.295608520507812e-05, "grad_norm": 5.656907558441162, "learning_rate": 1e-05, "loss": 0.5266, "mean_token_accuracy": 0.8531732559204102, "num_tokens": 230402678.0, "step": 6067 }, { "epoch": 0.7719119704872154, "ewc_loss": 0.099609375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 5.976808071136475, "learning_rate": 1e-05, "loss": 0.5119, "mean_token_accuracy": 0.8604042530059814, "num_tokens": 230440557.0, "step": 6068 }, { "epoch": 0.7720391807658059, "ewc_loss": 0.08984375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.2753143310546875, "learning_rate": 1e-05, "loss": 0.5044, "mean_token_accuracy": 0.8601928353309631, "num_tokens": 230477602.0, "step": 6069 }, { "epoch": 0.7721663910443964, "ewc_loss": 0.10107421875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 5.9903564453125, "learning_rate": 1e-05, "loss": 0.4623, "mean_token_accuracy": 0.8790236711502075, "num_tokens": 230518381.0, "step": 6070 }, { "epoch": 0.7722936013229869, "ewc_loss": 0.0888671875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.205051898956299, "learning_rate": 1e-05, "loss": 0.5417, "mean_token_accuracy": 0.8528547883033752, "num_tokens": 230559615.0, "step": 6071 }, { "epoch": 0.7724208116015774, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.178422451019287, "learning_rate": 1e-05, "loss": 0.5227, "mean_token_accuracy": 0.8576664328575134, "num_tokens": 230595270.0, "step": 6072 }, { "epoch": 0.7725480218801679, "ewc_loss": 0.08984375, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.1451416015625, "learning_rate": 1e-05, "loss": 0.5411, "mean_token_accuracy": 0.8485924005508423, "num_tokens": 230630405.0, "step": 6073 }, { "epoch": 0.7726752321587584, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.440334320068359, "learning_rate": 1e-05, "loss": 0.5941, "mean_token_accuracy": 0.8406596779823303, "num_tokens": 230670836.0, "step": 6074 }, { "epoch": 0.772802442437349, "ewc_loss": 0.09228515625, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.15881872177124, "learning_rate": 1e-05, "loss": 0.5559, "mean_token_accuracy": 0.8523908853530884, "num_tokens": 230709150.0, "step": 6075 }, { "epoch": 0.7729296527159395, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.708104610443115, "learning_rate": 1e-05, "loss": 0.5434, "mean_token_accuracy": 0.8566052913665771, "num_tokens": 230740306.0, "step": 6076 }, { "epoch": 0.7730568629945299, "ewc_loss": 0.0966796875, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.375085830688477, "learning_rate": 1e-05, "loss": 0.5178, "mean_token_accuracy": 0.8593188524246216, "num_tokens": 230776499.0, "step": 6077 }, { "epoch": 0.7731840732731204, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 6.789852142333984, "learning_rate": 1e-05, "loss": 0.5227, "mean_token_accuracy": 0.8618971109390259, "num_tokens": 230816614.0, "step": 6078 }, { "epoch": 0.773311283551711, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 5.672025203704834, "learning_rate": 1e-05, "loss": 0.5377, "mean_token_accuracy": 0.85313481092453, "num_tokens": 230865271.0, "step": 6079 }, { "epoch": 0.7734384938303015, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.566349983215332, "learning_rate": 1e-05, "loss": 0.5491, "mean_token_accuracy": 0.8556846380233765, "num_tokens": 230903092.0, "step": 6080 }, { "epoch": 0.773565704108892, "ewc_loss": 0.10791015625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 5.838964462280273, "learning_rate": 1e-05, "loss": 0.5159, "mean_token_accuracy": 0.860383152961731, "num_tokens": 230942726.0, "step": 6081 }, { "epoch": 0.7736929143874826, "ewc_loss": 0.10693359375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.1767473220825195, "learning_rate": 1e-05, "loss": 0.525, "mean_token_accuracy": 0.8555617928504944, "num_tokens": 230978694.0, "step": 6082 }, { "epoch": 0.773820124666073, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 5.8737616539001465, "learning_rate": 1e-05, "loss": 0.5511, "mean_token_accuracy": 0.8564739227294922, "num_tokens": 231015466.0, "step": 6083 }, { "epoch": 0.7739473349446635, "ewc_loss": 0.0966796875, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.77382230758667, "learning_rate": 1e-05, "loss": 0.5053, "mean_token_accuracy": 0.8599810600280762, "num_tokens": 231054823.0, "step": 6084 }, { "epoch": 0.774074545223254, "ewc_loss": 0.10400390625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 5.923591136932373, "learning_rate": 1e-05, "loss": 0.4851, "mean_token_accuracy": 0.8682551383972168, "num_tokens": 231087893.0, "step": 6085 }, { "epoch": 0.7742017555018446, "ewc_loss": 0.09033203125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.459352970123291, "learning_rate": 1e-05, "loss": 0.5131, "mean_token_accuracy": 0.8590104579925537, "num_tokens": 231123446.0, "step": 6086 }, { "epoch": 0.7743289657804351, "ewc_loss": 0.10302734375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.046627521514893, "learning_rate": 1e-05, "loss": 0.506, "mean_token_accuracy": 0.8630635142326355, "num_tokens": 231151558.0, "step": 6087 }, { "epoch": 0.7744561760590256, "ewc_loss": 0.08837890625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 7.581710815429688e-05, "grad_norm": 5.235233783721924, "learning_rate": 1e-05, "loss": 0.4873, "mean_token_accuracy": 0.863469123840332, "num_tokens": 231190715.0, "step": 6088 }, { "epoch": 0.774583386337616, "ewc_loss": 0.10595703125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.267394542694092, "learning_rate": 1e-05, "loss": 0.4847, "mean_token_accuracy": 0.8693891763687134, "num_tokens": 231224634.0, "step": 6089 }, { "epoch": 0.7747105966162066, "ewc_loss": 0.08837890625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 7.581710815429688e-05, "grad_norm": 5.126934051513672, "learning_rate": 1e-05, "loss": 0.5491, "mean_token_accuracy": 0.8494137525558472, "num_tokens": 231265497.0, "step": 6090 }, { "epoch": 0.7748378068947971, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.578636646270752, "learning_rate": 1e-05, "loss": 0.5115, "mean_token_accuracy": 0.862416684627533, "num_tokens": 231301886.0, "step": 6091 }, { "epoch": 0.7749650171733876, "ewc_loss": 0.0908203125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.222501277923584, "learning_rate": 1e-05, "loss": 0.5493, "mean_token_accuracy": 0.8478981256484985, "num_tokens": 231339580.0, "step": 6092 }, { "epoch": 0.7750922274519781, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.840884685516357, "learning_rate": 1e-05, "loss": 0.5337, "mean_token_accuracy": 0.8602376580238342, "num_tokens": 231376176.0, "step": 6093 }, { "epoch": 0.7752194377305687, "ewc_loss": 0.09814453125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.528958797454834, "learning_rate": 1e-05, "loss": 0.5102, "mean_token_accuracy": 0.8597004413604736, "num_tokens": 231409975.0, "step": 6094 }, { "epoch": 0.7753466480091591, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.813031196594238, "learning_rate": 1e-05, "loss": 0.5601, "mean_token_accuracy": 0.8545071482658386, "num_tokens": 231447998.0, "step": 6095 }, { "epoch": 0.7754738582877496, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 5.810030937194824, "learning_rate": 1e-05, "loss": 0.4917, "mean_token_accuracy": 0.8667351007461548, "num_tokens": 231481810.0, "step": 6096 }, { "epoch": 0.7756010685663401, "ewc_loss": 0.11181640625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.521368980407715, "learning_rate": 1e-05, "loss": 0.4983, "mean_token_accuracy": 0.8679560422897339, "num_tokens": 231521866.0, "step": 6097 }, { "epoch": 0.7757282788449307, "ewc_loss": 0.10400390625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 5.878503799438477, "learning_rate": 1e-05, "loss": 0.5672, "mean_token_accuracy": 0.8494232892990112, "num_tokens": 231558247.0, "step": 6098 }, { "epoch": 0.7758554891235212, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.1842193603515625, "learning_rate": 1e-05, "loss": 0.5149, "mean_token_accuracy": 0.8634153008460999, "num_tokens": 231594183.0, "step": 6099 }, { "epoch": 0.7759826994021117, "ewc_loss": 0.099609375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.837011814117432, "learning_rate": 1e-05, "loss": 0.5175, "mean_token_accuracy": 0.8590596914291382, "num_tokens": 231628991.0, "step": 6100 }, { "epoch": 0.7761099096807021, "ewc_loss": 0.09619140625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.8919548988342285, "learning_rate": 1e-05, "loss": 0.5582, "mean_token_accuracy": 0.8456790447235107, "num_tokens": 231669413.0, "step": 6101 }, { "epoch": 0.7762371199592927, "ewc_loss": 0.09716796875, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.763476848602295, "learning_rate": 1e-05, "loss": 0.5038, "mean_token_accuracy": 0.8612010478973389, "num_tokens": 231708765.0, "step": 6102 }, { "epoch": 0.7763643302378832, "ewc_loss": 0.09228515625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.718970775604248, "learning_rate": 1e-05, "loss": 0.513, "mean_token_accuracy": 0.8577972650527954, "num_tokens": 231747577.0, "step": 6103 }, { "epoch": 0.7764915405164737, "ewc_loss": 0.09423828125, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.6817097663879395, "learning_rate": 1e-05, "loss": 0.5199, "mean_token_accuracy": 0.8533774614334106, "num_tokens": 231788256.0, "step": 6104 }, { "epoch": 0.7766187507950643, "ewc_loss": 0.091796875, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.63957405090332, "learning_rate": 1e-05, "loss": 0.5203, "mean_token_accuracy": 0.8581079244613647, "num_tokens": 231825709.0, "step": 6105 }, { "epoch": 0.7767459610736548, "ewc_loss": 0.0927734375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.597184181213379, "learning_rate": 1e-05, "loss": 0.4917, "mean_token_accuracy": 0.8644016981124878, "num_tokens": 231862240.0, "step": 6106 }, { "epoch": 0.7768731713522452, "ewc_loss": 0.09228515625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.616194725036621, "learning_rate": 1e-05, "loss": 0.4785, "mean_token_accuracy": 0.8712639808654785, "num_tokens": 231897711.0, "step": 6107 }, { "epoch": 0.7770003816308357, "ewc_loss": 0.0888671875, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 7.62939453125e-05, "grad_norm": 5.52085018157959, "learning_rate": 1e-05, "loss": 0.53, "mean_token_accuracy": 0.8541542291641235, "num_tokens": 231932383.0, "step": 6108 }, { "epoch": 0.7771275919094263, "ewc_loss": 0.091796875, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.650832176208496, "learning_rate": 1e-05, "loss": 0.5193, "mean_token_accuracy": 0.853449285030365, "num_tokens": 231960919.0, "step": 6109 }, { "epoch": 0.7772548021880168, "ewc_loss": 0.08837890625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 7.581710815429688e-05, "grad_norm": 5.412182807922363, "learning_rate": 1e-05, "loss": 0.4448, "mean_token_accuracy": 0.8788191080093384, "num_tokens": 231996435.0, "step": 6110 }, { "epoch": 0.7773820124666073, "ewc_loss": 0.09326171875, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.712632656097412, "learning_rate": 1e-05, "loss": 0.5303, "mean_token_accuracy": 0.8516498804092407, "num_tokens": 232036146.0, "step": 6111 }, { "epoch": 0.7775092227451978, "ewc_loss": 0.08544921875, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 7.295608520507812e-05, "grad_norm": 5.289491653442383, "learning_rate": 1e-05, "loss": 0.5179, "mean_token_accuracy": 0.8519628047943115, "num_tokens": 232077656.0, "step": 6112 }, { "epoch": 0.7776364330237884, "ewc_loss": 0.09521484375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.878933429718018, "learning_rate": 1e-05, "loss": 0.5005, "mean_token_accuracy": 0.8657822608947754, "num_tokens": 232113370.0, "step": 6113 }, { "epoch": 0.7777636433023788, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 7.104873657226562e-05, "grad_norm": 5.125790119171143, "learning_rate": 1e-05, "loss": 0.4734, "mean_token_accuracy": 0.8669025301933289, "num_tokens": 232155450.0, "step": 6114 }, { "epoch": 0.7778908535809693, "ewc_loss": 0.10302734375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.186761379241943, "learning_rate": 1e-05, "loss": 0.4888, "mean_token_accuracy": 0.8659288883209229, "num_tokens": 232198076.0, "step": 6115 }, { "epoch": 0.7780180638595598, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 6.961822509765625e-05, "grad_norm": 4.949579238891602, "learning_rate": 1e-05, "loss": 0.5231, "mean_token_accuracy": 0.8532786965370178, "num_tokens": 232246472.0, "step": 6116 }, { "epoch": 0.7781452741381504, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.740463733673096, "learning_rate": 1e-05, "loss": 0.5216, "mean_token_accuracy": 0.8642082214355469, "num_tokens": 232280576.0, "step": 6117 }, { "epoch": 0.7782724844167409, "ewc_loss": 0.08642578125, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 7.390975952148438e-05, "grad_norm": 5.014091491699219, "learning_rate": 1e-05, "loss": 0.485, "mean_token_accuracy": 0.8657439947128296, "num_tokens": 232322172.0, "step": 6118 }, { "epoch": 0.7783996946953314, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 0.00011920928955078125, "grad_norm": 7.355093955993652, "learning_rate": 1e-05, "loss": 0.5399, "mean_token_accuracy": 0.8607271909713745, "num_tokens": 232359317.0, "step": 6119 }, { "epoch": 0.7785269049739219, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 5.712254047393799, "learning_rate": 1e-05, "loss": 0.5707, "mean_token_accuracy": 0.8434532880783081, "num_tokens": 232395209.0, "step": 6120 }, { "epoch": 0.7786541152525124, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 0.00011682510375976562, "grad_norm": 7.340892791748047, "learning_rate": 1e-05, "loss": 0.5752, "mean_token_accuracy": 0.8521552085876465, "num_tokens": 232433303.0, "step": 6121 }, { "epoch": 0.7787813255311029, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.3200225830078125, "learning_rate": 1e-05, "loss": 0.4772, "mean_token_accuracy": 0.8733036518096924, "num_tokens": 232464465.0, "step": 6122 }, { "epoch": 0.7789085358096934, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.811631202697754, "learning_rate": 1e-05, "loss": 0.4812, "mean_token_accuracy": 0.8757637739181519, "num_tokens": 232499846.0, "step": 6123 }, { "epoch": 0.779035746088284, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.252870082855225, "learning_rate": 1e-05, "loss": 0.4816, "mean_token_accuracy": 0.8717361688613892, "num_tokens": 232528714.0, "step": 6124 }, { "epoch": 0.7791629563668745, "ewc_loss": 0.10693359375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.420942783355713, "learning_rate": 1e-05, "loss": 0.5529, "mean_token_accuracy": 0.8521692156791687, "num_tokens": 232561155.0, "step": 6125 }, { "epoch": 0.7792901666454649, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.007489204406738, "learning_rate": 1e-05, "loss": 0.472, "mean_token_accuracy": 0.8737561702728271, "num_tokens": 232595685.0, "step": 6126 }, { "epoch": 0.7794173769240554, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 6.155283451080322, "learning_rate": 1e-05, "loss": 0.582, "mean_token_accuracy": 0.8462722301483154, "num_tokens": 232638683.0, "step": 6127 }, { "epoch": 0.779544587202646, "ewc_loss": 0.0966796875, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.797048568725586, "learning_rate": 1e-05, "loss": 0.5019, "mean_token_accuracy": 0.861839234828949, "num_tokens": 232677492.0, "step": 6128 }, { "epoch": 0.7796717974812365, "ewc_loss": 0.09765625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 6.036294460296631, "learning_rate": 1e-05, "loss": 0.5458, "mean_token_accuracy": 0.8468842506408691, "num_tokens": 232717997.0, "step": 6129 }, { "epoch": 0.779799007759827, "ewc_loss": 0.091796875, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.608994007110596, "learning_rate": 1e-05, "loss": 0.5187, "mean_token_accuracy": 0.8527278304100037, "num_tokens": 232756784.0, "step": 6130 }, { "epoch": 0.7799262180384176, "ewc_loss": 0.0986328125, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 6.0155110359191895, "learning_rate": 1e-05, "loss": 0.5176, "mean_token_accuracy": 0.8574422001838684, "num_tokens": 232795194.0, "step": 6131 }, { "epoch": 0.780053428317008, "ewc_loss": 0.08837890625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 7.581710815429688e-05, "grad_norm": 5.437673091888428, "learning_rate": 1e-05, "loss": 0.5008, "mean_token_accuracy": 0.8638477921485901, "num_tokens": 232826505.0, "step": 6132 }, { "epoch": 0.7801806385955985, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 6.116267204284668, "learning_rate": 1e-05, "loss": 0.4907, "mean_token_accuracy": 0.8701109290122986, "num_tokens": 232860111.0, "step": 6133 }, { "epoch": 0.780307848874189, "ewc_loss": 0.08642578125, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 7.390975952148438e-05, "grad_norm": 5.272429943084717, "learning_rate": 1e-05, "loss": 0.5348, "mean_token_accuracy": 0.85165935754776, "num_tokens": 232895558.0, "step": 6134 }, { "epoch": 0.7804350591527796, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.355731010437012, "learning_rate": 1e-05, "loss": 0.4949, "mean_token_accuracy": 0.8666000962257385, "num_tokens": 232935737.0, "step": 6135 }, { "epoch": 0.7805622694313701, "ewc_loss": 0.0859375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 7.343292236328125e-05, "grad_norm": 5.202436447143555, "learning_rate": 1e-05, "loss": 0.5601, "mean_token_accuracy": 0.8421458601951599, "num_tokens": 232977917.0, "step": 6136 }, { "epoch": 0.7806894797099606, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.693269729614258, "learning_rate": 1e-05, "loss": 0.4847, "mean_token_accuracy": 0.8730170726776123, "num_tokens": 233016748.0, "step": 6137 }, { "epoch": 0.780816689988551, "ewc_loss": 0.09033203125, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.3573431968688965, "learning_rate": 1e-05, "loss": 0.5163, "mean_token_accuracy": 0.8625121712684631, "num_tokens": 233052814.0, "step": 6138 }, { "epoch": 0.7809439002671416, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.893709659576416, "learning_rate": 1e-05, "loss": 0.533, "mean_token_accuracy": 0.8588385581970215, "num_tokens": 233088727.0, "step": 6139 }, { "epoch": 0.7810711105457321, "ewc_loss": 0.09912109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.7580885887146, "learning_rate": 1e-05, "loss": 0.5113, "mean_token_accuracy": 0.8579217791557312, "num_tokens": 233126402.0, "step": 6140 }, { "epoch": 0.7811983208243226, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.651752471923828, "learning_rate": 1e-05, "loss": 0.5808, "mean_token_accuracy": 0.8443952798843384, "num_tokens": 233161204.0, "step": 6141 }, { "epoch": 0.7813255311029131, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.0059332847595215, "learning_rate": 1e-05, "loss": 0.5325, "mean_token_accuracy": 0.8593311309814453, "num_tokens": 233196527.0, "step": 6142 }, { "epoch": 0.7814527413815037, "ewc_loss": 0.10107421875, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 6.150696754455566, "learning_rate": 1e-05, "loss": 0.5383, "mean_token_accuracy": 0.8524461984634399, "num_tokens": 233236164.0, "step": 6143 }, { "epoch": 0.7815799516600941, "ewc_loss": 0.10107421875, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 6.050967693328857, "learning_rate": 1e-05, "loss": 0.5451, "mean_token_accuracy": 0.8530796766281128, "num_tokens": 233268762.0, "step": 6144 }, { "epoch": 0.7817071619386846, "ewc_loss": 0.08935546875, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.657808780670166, "learning_rate": 1e-05, "loss": 0.5268, "mean_token_accuracy": 0.8539701700210571, "num_tokens": 233307388.0, "step": 6145 }, { "epoch": 0.7818343722172751, "ewc_loss": 0.10205078125, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 6.1449666023254395, "learning_rate": 1e-05, "loss": 0.5024, "mean_token_accuracy": 0.8636890053749084, "num_tokens": 233347902.0, "step": 6146 }, { "epoch": 0.7819615824958657, "ewc_loss": 0.0849609375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 7.2479248046875e-05, "grad_norm": 5.316561222076416, "learning_rate": 1e-05, "loss": 0.5499, "mean_token_accuracy": 0.8452404737472534, "num_tokens": 233378959.0, "step": 6147 }, { "epoch": 0.7820887927744562, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.378427982330322, "learning_rate": 1e-05, "loss": 0.5417, "mean_token_accuracy": 0.8546043634414673, "num_tokens": 233420580.0, "step": 6148 }, { "epoch": 0.7822160030530467, "ewc_loss": 0.08447265625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 7.200241088867188e-05, "grad_norm": 5.175764083862305, "learning_rate": 1e-05, "loss": 0.473, "mean_token_accuracy": 0.8667775392532349, "num_tokens": 233451628.0, "step": 6149 }, { "epoch": 0.7823432133316371, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.707507610321045, "learning_rate": 1e-05, "loss": 0.5572, "mean_token_accuracy": 0.852249264717102, "num_tokens": 233493994.0, "step": 6150 }, { "epoch": 0.7824704236102277, "ewc_loss": 0.08984375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.315186500549316, "learning_rate": 1e-05, "loss": 0.4827, "mean_token_accuracy": 0.8643619418144226, "num_tokens": 233534156.0, "step": 6151 }, { "epoch": 0.7825976338888182, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.914435386657715, "learning_rate": 1e-05, "loss": 0.5406, "mean_token_accuracy": 0.8586653470993042, "num_tokens": 233573179.0, "step": 6152 }, { "epoch": 0.7827248441674087, "ewc_loss": 0.09814453125, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.6922125816345215, "learning_rate": 1e-05, "loss": 0.4958, "mean_token_accuracy": 0.8665704131126404, "num_tokens": 233606424.0, "step": 6153 }, { "epoch": 0.7828520544459993, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.784026145935059, "learning_rate": 1e-05, "loss": 0.5456, "mean_token_accuracy": 0.854611337184906, "num_tokens": 233649155.0, "step": 6154 }, { "epoch": 0.7829792647245898, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 5.927793025970459, "learning_rate": 1e-05, "loss": 0.5551, "mean_token_accuracy": 0.846207857131958, "num_tokens": 233677163.0, "step": 6155 }, { "epoch": 0.7831064750031802, "ewc_loss": 0.10791015625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.454400062561035, "learning_rate": 1e-05, "loss": 0.5541, "mean_token_accuracy": 0.8482660055160522, "num_tokens": 233713120.0, "step": 6156 }, { "epoch": 0.7832336852817707, "ewc_loss": 0.099609375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.907004356384277, "learning_rate": 1e-05, "loss": 0.5263, "mean_token_accuracy": 0.8566208481788635, "num_tokens": 233751160.0, "step": 6157 }, { "epoch": 0.7833608955603613, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 6.151230335235596, "learning_rate": 1e-05, "loss": 0.5191, "mean_token_accuracy": 0.8579442501068115, "num_tokens": 233786474.0, "step": 6158 }, { "epoch": 0.7834881058389518, "ewc_loss": 0.095703125, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.789357662200928, "learning_rate": 1e-05, "loss": 0.4969, "mean_token_accuracy": 0.8639204502105713, "num_tokens": 233818256.0, "step": 6159 }, { "epoch": 0.7836153161175423, "ewc_loss": 0.095703125, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.969580173492432, "learning_rate": 1e-05, "loss": 0.5432, "mean_token_accuracy": 0.849465012550354, "num_tokens": 233856519.0, "step": 6160 }, { "epoch": 0.7837425263961328, "ewc_loss": 0.09130859375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.634711265563965, "learning_rate": 1e-05, "loss": 0.4979, "mean_token_accuracy": 0.8596188426017761, "num_tokens": 233896132.0, "step": 6161 }, { "epoch": 0.7838697366747234, "ewc_loss": 0.09716796875, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.934658050537109, "learning_rate": 1e-05, "loss": 0.5188, "mean_token_accuracy": 0.8592756986618042, "num_tokens": 233931668.0, "step": 6162 }, { "epoch": 0.7839969469533138, "ewc_loss": 0.0888671875, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 7.62939453125e-05, "grad_norm": 5.477322578430176, "learning_rate": 1e-05, "loss": 0.4623, "mean_token_accuracy": 0.868804395198822, "num_tokens": 233965821.0, "step": 6163 }, { "epoch": 0.7841241572319043, "ewc_loss": 0.09912109375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 6.004256725311279, "learning_rate": 1e-05, "loss": 0.5058, "mean_token_accuracy": 0.8605632185935974, "num_tokens": 234003488.0, "step": 6164 }, { "epoch": 0.7842513675104948, "ewc_loss": 0.08740234375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 7.486343383789062e-05, "grad_norm": 5.330078601837158, "learning_rate": 1e-05, "loss": 0.5217, "mean_token_accuracy": 0.8544172048568726, "num_tokens": 234044270.0, "step": 6165 }, { "epoch": 0.7843785777890854, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.180429935455322, "learning_rate": 1e-05, "loss": 0.4554, "mean_token_accuracy": 0.879718542098999, "num_tokens": 234081754.0, "step": 6166 }, { "epoch": 0.7845057880676759, "ewc_loss": 0.0859375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 7.343292236328125e-05, "grad_norm": 5.220570087432861, "learning_rate": 1e-05, "loss": 0.4577, "mean_token_accuracy": 0.8761187791824341, "num_tokens": 234118961.0, "step": 6167 }, { "epoch": 0.7846329983462664, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.460687160491943, "learning_rate": 1e-05, "loss": 0.5751, "mean_token_accuracy": 0.8436263799667358, "num_tokens": 234159084.0, "step": 6168 }, { "epoch": 0.7847602086248568, "ewc_loss": 0.087890625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 7.534027099609375e-05, "grad_norm": 5.249833106994629, "learning_rate": 1e-05, "loss": 0.487, "mean_token_accuracy": 0.8642591238021851, "num_tokens": 234201296.0, "step": 6169 }, { "epoch": 0.7848874189034474, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.696610450744629, "learning_rate": 1e-05, "loss": 0.5156, "mean_token_accuracy": 0.8659378290176392, "num_tokens": 234240661.0, "step": 6170 }, { "epoch": 0.7850146291820379, "ewc_loss": 0.09326171875, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.508132457733154, "learning_rate": 1e-05, "loss": 0.5385, "mean_token_accuracy": 0.8494831323623657, "num_tokens": 234277192.0, "step": 6171 }, { "epoch": 0.7851418394606284, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.655848503112793, "learning_rate": 1e-05, "loss": 0.5388, "mean_token_accuracy": 0.8581153154373169, "num_tokens": 234321527.0, "step": 6172 }, { "epoch": 0.785269049739219, "ewc_loss": 0.10009765625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 5.796229362487793, "learning_rate": 1e-05, "loss": 0.5503, "mean_token_accuracy": 0.8486219048500061, "num_tokens": 234355578.0, "step": 6173 }, { "epoch": 0.7853962600178095, "ewc_loss": 0.10400390625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.286596775054932, "learning_rate": 1e-05, "loss": 0.5044, "mean_token_accuracy": 0.8682233095169067, "num_tokens": 234395240.0, "step": 6174 }, { "epoch": 0.7855234702963999, "ewc_loss": 0.09912109375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.920526504516602, "learning_rate": 1e-05, "loss": 0.573, "mean_token_accuracy": 0.845558226108551, "num_tokens": 234436914.0, "step": 6175 }, { "epoch": 0.7856506805749904, "ewc_loss": 0.09228515625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.80304479598999, "learning_rate": 1e-05, "loss": 0.5193, "mean_token_accuracy": 0.8584372401237488, "num_tokens": 234477811.0, "step": 6176 }, { "epoch": 0.785777890853581, "ewc_loss": 0.099609375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 6.030356407165527, "learning_rate": 1e-05, "loss": 0.4987, "mean_token_accuracy": 0.864660382270813, "num_tokens": 234515489.0, "step": 6177 }, { "epoch": 0.7859051011321715, "ewc_loss": 0.08544921875, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 7.295608520507812e-05, "grad_norm": 5.378730773925781, "learning_rate": 1e-05, "loss": 0.5387, "mean_token_accuracy": 0.8564214706420898, "num_tokens": 234544846.0, "step": 6178 }, { "epoch": 0.786032311410762, "ewc_loss": 0.10302734375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.263453006744385, "learning_rate": 1e-05, "loss": 0.4855, "mean_token_accuracy": 0.8695380687713623, "num_tokens": 234581497.0, "step": 6179 }, { "epoch": 0.7861595216893525, "ewc_loss": 0.08349609375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 7.05718994140625e-05, "grad_norm": 5.115469932556152, "learning_rate": 1e-05, "loss": 0.5287, "mean_token_accuracy": 0.8526899218559265, "num_tokens": 234621837.0, "step": 6180 }, { "epoch": 0.786286731967943, "ewc_loss": 0.11181640625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.66203498840332, "learning_rate": 1e-05, "loss": 0.5084, "mean_token_accuracy": 0.8644775152206421, "num_tokens": 234655049.0, "step": 6181 }, { "epoch": 0.7864139422465335, "ewc_loss": 0.087890625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 7.534027099609375e-05, "grad_norm": 5.190639972686768, "learning_rate": 1e-05, "loss": 0.5286, "mean_token_accuracy": 0.8521213531494141, "num_tokens": 234689177.0, "step": 6182 }, { "epoch": 0.786541152525124, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 7.01128625869751, "learning_rate": 1e-05, "loss": 0.4878, "mean_token_accuracy": 0.8729901313781738, "num_tokens": 234722744.0, "step": 6183 }, { "epoch": 0.7866683628037145, "ewc_loss": 0.09814453125, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.6254963874816895, "learning_rate": 1e-05, "loss": 0.4904, "mean_token_accuracy": 0.8647806644439697, "num_tokens": 234762998.0, "step": 6184 }, { "epoch": 0.7867955730823051, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.9446210861206055, "learning_rate": 1e-05, "loss": 0.5863, "mean_token_accuracy": 0.8440135717391968, "num_tokens": 234799283.0, "step": 6185 }, { "epoch": 0.7869227833608956, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 5.988121509552002, "learning_rate": 1e-05, "loss": 0.5498, "mean_token_accuracy": 0.8511318564414978, "num_tokens": 234838019.0, "step": 6186 }, { "epoch": 0.787049993639486, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.564504146575928, "learning_rate": 1e-05, "loss": 0.5525, "mean_token_accuracy": 0.8525078892707825, "num_tokens": 234874449.0, "step": 6187 }, { "epoch": 0.7871772039180766, "ewc_loss": 0.10302734375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 5.994085311889648, "learning_rate": 1e-05, "loss": 0.523, "mean_token_accuracy": 0.8615812063217163, "num_tokens": 234913868.0, "step": 6188 }, { "epoch": 0.7873044141966671, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 6.200150489807129, "learning_rate": 1e-05, "loss": 0.4631, "mean_token_accuracy": 0.8779362440109253, "num_tokens": 234949060.0, "step": 6189 }, { "epoch": 0.7874316244752576, "ewc_loss": 0.09765625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.869455337524414, "learning_rate": 1e-05, "loss": 0.5044, "mean_token_accuracy": 0.858471155166626, "num_tokens": 234987732.0, "step": 6190 }, { "epoch": 0.7875588347538481, "ewc_loss": 0.0966796875, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.97392463684082, "learning_rate": 1e-05, "loss": 0.4776, "mean_token_accuracy": 0.8693023920059204, "num_tokens": 235023030.0, "step": 6191 }, { "epoch": 0.7876860450324387, "ewc_loss": 0.09228515625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.721498012542725, "learning_rate": 1e-05, "loss": 0.5027, "mean_token_accuracy": 0.8636007905006409, "num_tokens": 235061560.0, "step": 6192 }, { "epoch": 0.7878132553110291, "ewc_loss": 0.0947265625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.876645088195801, "learning_rate": 1e-05, "loss": 0.4901, "mean_token_accuracy": 0.8649269342422485, "num_tokens": 235098826.0, "step": 6193 }, { "epoch": 0.7879404655896196, "ewc_loss": 0.08935546875, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.555008411407471, "learning_rate": 1e-05, "loss": 0.5112, "mean_token_accuracy": 0.8562771081924438, "num_tokens": 235142116.0, "step": 6194 }, { "epoch": 0.7880676758682101, "ewc_loss": 0.095703125, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.905280113220215, "learning_rate": 1e-05, "loss": 0.5294, "mean_token_accuracy": 0.8548195958137512, "num_tokens": 235182920.0, "step": 6195 }, { "epoch": 0.7881948861468007, "ewc_loss": 0.087890625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 7.534027099609375e-05, "grad_norm": 5.392680644989014, "learning_rate": 1e-05, "loss": 0.5133, "mean_token_accuracy": 0.85548335313797, "num_tokens": 235214669.0, "step": 6196 }, { "epoch": 0.7883220964253912, "ewc_loss": 0.09912109375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 6.047910690307617, "learning_rate": 1e-05, "loss": 0.5067, "mean_token_accuracy": 0.8626099824905396, "num_tokens": 235251723.0, "step": 6197 }, { "epoch": 0.7884493067039817, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 7.104873657226562e-05, "grad_norm": 5.235723495483398, "learning_rate": 1e-05, "loss": 0.4627, "mean_token_accuracy": 0.8663564324378967, "num_tokens": 235285931.0, "step": 6198 }, { "epoch": 0.7885765169825721, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.300780773162842, "learning_rate": 1e-05, "loss": 0.4889, "mean_token_accuracy": 0.8682286739349365, "num_tokens": 235324571.0, "step": 6199 }, { "epoch": 0.7887037272611627, "ewc_loss": 0.0849609375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 7.2479248046875e-05, "grad_norm": 5.157960891723633, "learning_rate": 1e-05, "loss": 0.4697, "mean_token_accuracy": 0.8716387748718262, "num_tokens": 235359348.0, "step": 6200 }, { "epoch": 0.7888309375397532, "ewc_loss": 0.11083984375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.62792444229126, "learning_rate": 1e-05, "loss": 0.4941, "mean_token_accuracy": 0.8713163137435913, "num_tokens": 235397998.0, "step": 6201 }, { "epoch": 0.7889581478183437, "ewc_loss": 0.08935546875, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.293975830078125, "learning_rate": 1e-05, "loss": 0.502, "mean_token_accuracy": 0.8571733236312866, "num_tokens": 235436991.0, "step": 6202 }, { "epoch": 0.7890853580969343, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.839260578155518, "learning_rate": 1e-05, "loss": 0.4725, "mean_token_accuracy": 0.8764564990997314, "num_tokens": 235478129.0, "step": 6203 }, { "epoch": 0.7892125683755248, "ewc_loss": 0.0966796875, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.6574859619140625, "learning_rate": 1e-05, "loss": 0.4897, "mean_token_accuracy": 0.8676192760467529, "num_tokens": 235514630.0, "step": 6204 }, { "epoch": 0.7893397786541152, "ewc_loss": 0.11181640625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.675245761871338, "learning_rate": 1e-05, "loss": 0.4964, "mean_token_accuracy": 0.870317816734314, "num_tokens": 235550212.0, "step": 6205 }, { "epoch": 0.7894669889327057, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 5.909189224243164, "learning_rate": 1e-05, "loss": 0.4929, "mean_token_accuracy": 0.8703147768974304, "num_tokens": 235592602.0, "step": 6206 }, { "epoch": 0.7895941992112963, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.231718063354492, "learning_rate": 1e-05, "loss": 0.4825, "mean_token_accuracy": 0.8705726861953735, "num_tokens": 235634199.0, "step": 6207 }, { "epoch": 0.7897214094898868, "ewc_loss": 0.099609375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.995852470397949, "learning_rate": 1e-05, "loss": 0.4779, "mean_token_accuracy": 0.8696953654289246, "num_tokens": 235670086.0, "step": 6208 }, { "epoch": 0.7898486197684773, "ewc_loss": 0.0908203125, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.739041328430176, "learning_rate": 1e-05, "loss": 0.57, "mean_token_accuracy": 0.8449779748916626, "num_tokens": 235709749.0, "step": 6209 }, { "epoch": 0.7899758300470678, "ewc_loss": 0.10009765625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 6.089560508728027, "learning_rate": 1e-05, "loss": 0.5057, "mean_token_accuracy": 0.8630131483078003, "num_tokens": 235750574.0, "step": 6210 }, { "epoch": 0.7901030403256584, "ewc_loss": 0.08544921875, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 7.295608520507812e-05, "grad_norm": 5.379354953765869, "learning_rate": 1e-05, "loss": 0.4853, "mean_token_accuracy": 0.8647545576095581, "num_tokens": 235785234.0, "step": 6211 }, { "epoch": 0.7902302506042488, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.299232006072998, "learning_rate": 1e-05, "loss": 0.557, "mean_token_accuracy": 0.8479169607162476, "num_tokens": 235822931.0, "step": 6212 }, { "epoch": 0.7903574608828393, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 7.152557373046875e-05, "grad_norm": 5.164447784423828, "learning_rate": 1e-05, "loss": 0.5233, "mean_token_accuracy": 0.8496643304824829, "num_tokens": 235863022.0, "step": 6213 }, { "epoch": 0.7904846711614298, "ewc_loss": 0.11083984375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.627711296081543, "learning_rate": 1e-05, "loss": 0.5109, "mean_token_accuracy": 0.8613938093185425, "num_tokens": 235893275.0, "step": 6214 }, { "epoch": 0.7906118814400204, "ewc_loss": 0.0888671875, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 7.62939453125e-05, "grad_norm": 5.247520923614502, "learning_rate": 1e-05, "loss": 0.4382, "mean_token_accuracy": 0.880617618560791, "num_tokens": 235933908.0, "step": 6215 }, { "epoch": 0.7907390917186109, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.898808479309082, "learning_rate": 1e-05, "loss": 0.5627, "mean_token_accuracy": 0.8522875308990479, "num_tokens": 235975296.0, "step": 6216 }, { "epoch": 0.7908663019972014, "ewc_loss": 0.09619140625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.60126256942749, "learning_rate": 1e-05, "loss": 0.4798, "mean_token_accuracy": 0.8698634505271912, "num_tokens": 236012872.0, "step": 6217 }, { "epoch": 0.7909935122757918, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.821341514587402, "learning_rate": 1e-05, "loss": 0.5274, "mean_token_accuracy": 0.8640453815460205, "num_tokens": 236055762.0, "step": 6218 }, { "epoch": 0.7911207225543824, "ewc_loss": 0.10302734375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 5.8845906257629395, "learning_rate": 1e-05, "loss": 0.5034, "mean_token_accuracy": 0.8645881414413452, "num_tokens": 236099968.0, "step": 6219 }, { "epoch": 0.7912479328329729, "ewc_loss": 0.10791015625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.500843524932861, "learning_rate": 1e-05, "loss": 0.5431, "mean_token_accuracy": 0.8513213396072388, "num_tokens": 236134000.0, "step": 6220 }, { "epoch": 0.7913751431115634, "ewc_loss": 0.10009765625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 5.912261009216309, "learning_rate": 1e-05, "loss": 0.4911, "mean_token_accuracy": 0.8667610883712769, "num_tokens": 236174427.0, "step": 6221 }, { "epoch": 0.791502353390154, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 6.183713912963867, "learning_rate": 1e-05, "loss": 0.5474, "mean_token_accuracy": 0.8518927097320557, "num_tokens": 236205643.0, "step": 6222 }, { "epoch": 0.7916295636687445, "ewc_loss": 0.09716796875, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.815201759338379, "learning_rate": 1e-05, "loss": 0.5668, "mean_token_accuracy": 0.8442132472991943, "num_tokens": 236252232.0, "step": 6223 }, { "epoch": 0.7917567739473349, "ewc_loss": 0.0966796875, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.9530510902404785, "learning_rate": 1e-05, "loss": 0.5229, "mean_token_accuracy": 0.8564261198043823, "num_tokens": 236293694.0, "step": 6224 }, { "epoch": 0.7918839842259254, "ewc_loss": 0.091796875, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.686561107635498, "learning_rate": 1e-05, "loss": 0.4865, "mean_token_accuracy": 0.8664861917495728, "num_tokens": 236335129.0, "step": 6225 }, { "epoch": 0.792011194504516, "ewc_loss": 0.09423828125, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.873491287231445, "learning_rate": 1e-05, "loss": 0.4747, "mean_token_accuracy": 0.8715156316757202, "num_tokens": 236370294.0, "step": 6226 }, { "epoch": 0.7921384047831065, "ewc_loss": 0.08984375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.537245273590088, "learning_rate": 1e-05, "loss": 0.4785, "mean_token_accuracy": 0.8697390556335449, "num_tokens": 236410728.0, "step": 6227 }, { "epoch": 0.792265615061697, "ewc_loss": 0.095703125, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.904428482055664, "learning_rate": 1e-05, "loss": 0.5154, "mean_token_accuracy": 0.8604928255081177, "num_tokens": 236453053.0, "step": 6228 }, { "epoch": 0.7923928253402875, "ewc_loss": 0.0869140625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 7.43865966796875e-05, "grad_norm": 5.376499652862549, "learning_rate": 1e-05, "loss": 0.4733, "mean_token_accuracy": 0.869485080242157, "num_tokens": 236491746.0, "step": 6229 }, { "epoch": 0.792520035618878, "ewc_loss": 0.099609375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 6.049529552459717, "learning_rate": 1e-05, "loss": 0.5481, "mean_token_accuracy": 0.8478479385375977, "num_tokens": 236530938.0, "step": 6230 }, { "epoch": 0.7926472458974685, "ewc_loss": 0.08447265625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 7.200241088867188e-05, "grad_norm": 5.234987258911133, "learning_rate": 1e-05, "loss": 0.4192, "mean_token_accuracy": 0.8861982822418213, "num_tokens": 236567765.0, "step": 6231 }, { "epoch": 0.792774456176059, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.310960292816162, "learning_rate": 1e-05, "loss": 0.5107, "mean_token_accuracy": 0.8596277832984924, "num_tokens": 236596522.0, "step": 6232 }, { "epoch": 0.7929016664546495, "ewc_loss": 0.0859375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 7.343292236328125e-05, "grad_norm": 5.156643867492676, "learning_rate": 1e-05, "loss": 0.5131, "mean_token_accuracy": 0.8531842827796936, "num_tokens": 236637493.0, "step": 6233 }, { "epoch": 0.7930288767332401, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.655750751495361, "learning_rate": 1e-05, "loss": 0.5251, "mean_token_accuracy": 0.8582841157913208, "num_tokens": 236668405.0, "step": 6234 }, { "epoch": 0.7931560870118306, "ewc_loss": 0.08984375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.2964396476745605, "learning_rate": 1e-05, "loss": 0.4844, "mean_token_accuracy": 0.8642927408218384, "num_tokens": 236703105.0, "step": 6235 }, { "epoch": 0.793283297290421, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.863687992095947, "learning_rate": 1e-05, "loss": 0.5274, "mean_token_accuracy": 0.8625525236129761, "num_tokens": 236747017.0, "step": 6236 }, { "epoch": 0.7934105075690115, "ewc_loss": 0.09814453125, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.6639909744262695, "learning_rate": 1e-05, "loss": 0.4815, "mean_token_accuracy": 0.867514967918396, "num_tokens": 236785659.0, "step": 6237 }, { "epoch": 0.7935377178476021, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.686443328857422, "learning_rate": 1e-05, "loss": 0.5279, "mean_token_accuracy": 0.8531537055969238, "num_tokens": 236819603.0, "step": 6238 }, { "epoch": 0.7936649281261926, "ewc_loss": 0.10302734375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 5.9416728019714355, "learning_rate": 1e-05, "loss": 0.4936, "mean_token_accuracy": 0.8659568428993225, "num_tokens": 236859440.0, "step": 6239 }, { "epoch": 0.7937921384047831, "ewc_loss": 0.10302734375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.216388702392578, "learning_rate": 1e-05, "loss": 0.5174, "mean_token_accuracy": 0.8613405227661133, "num_tokens": 236892138.0, "step": 6240 }, { "epoch": 0.7939193486833737, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.012118339538574, "learning_rate": 1e-05, "loss": 0.5654, "mean_token_accuracy": 0.8437696695327759, "num_tokens": 236930559.0, "step": 6241 }, { "epoch": 0.7940465589619641, "ewc_loss": 0.09130859375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.72100305557251, "learning_rate": 1e-05, "loss": 0.5152, "mean_token_accuracy": 0.8580679893493652, "num_tokens": 236969275.0, "step": 6242 }, { "epoch": 0.7941737692405546, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.100834846496582, "learning_rate": 1e-05, "loss": 0.4796, "mean_token_accuracy": 0.8686148524284363, "num_tokens": 237007662.0, "step": 6243 }, { "epoch": 0.7943009795191451, "ewc_loss": 0.08544921875, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 7.295608520507812e-05, "grad_norm": 5.351846218109131, "learning_rate": 1e-05, "loss": 0.5448, "mean_token_accuracy": 0.84406578540802, "num_tokens": 237046553.0, "step": 6244 }, { "epoch": 0.7944281897977357, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.326016902923584, "learning_rate": 1e-05, "loss": 0.523, "mean_token_accuracy": 0.8581278920173645, "num_tokens": 237082309.0, "step": 6245 }, { "epoch": 0.7945554000763262, "ewc_loss": 0.08544921875, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 7.295608520507812e-05, "grad_norm": 5.183027267456055, "learning_rate": 1e-05, "loss": 0.4758, "mean_token_accuracy": 0.8692876100540161, "num_tokens": 237117997.0, "step": 6246 }, { "epoch": 0.7946826103549167, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.645129680633545, "learning_rate": 1e-05, "loss": 0.5695, "mean_token_accuracy": 0.8472575545310974, "num_tokens": 237159169.0, "step": 6247 }, { "epoch": 0.7948098206335071, "ewc_loss": 0.08935546875, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.2848100662231445, "learning_rate": 1e-05, "loss": 0.519, "mean_token_accuracy": 0.8573590517044067, "num_tokens": 237200863.0, "step": 6248 }, { "epoch": 0.7949370309120977, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.8772807121276855, "learning_rate": 1e-05, "loss": 0.569, "mean_token_accuracy": 0.8471884727478027, "num_tokens": 237237318.0, "step": 6249 }, { "epoch": 0.7950642411906882, "ewc_loss": 0.0966796875, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.624067783355713, "learning_rate": 1e-05, "loss": 0.561, "mean_token_accuracy": 0.8424903154373169, "num_tokens": 237278724.0, "step": 6250 }, { "epoch": 0.7951914514692787, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.774364471435547, "learning_rate": 1e-05, "loss": 0.5191, "mean_token_accuracy": 0.8644884824752808, "num_tokens": 237314190.0, "step": 6251 }, { "epoch": 0.7953186617478692, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 5.867188930511475, "learning_rate": 1e-05, "loss": 0.4705, "mean_token_accuracy": 0.8719111680984497, "num_tokens": 237348560.0, "step": 6252 }, { "epoch": 0.7954458720264598, "ewc_loss": 0.10693359375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.456089019775391, "learning_rate": 1e-05, "loss": 0.519, "mean_token_accuracy": 0.8620190620422363, "num_tokens": 237388421.0, "step": 6253 }, { "epoch": 0.7955730823050502, "ewc_loss": 0.09912109375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.883174419403076, "learning_rate": 1e-05, "loss": 0.5604, "mean_token_accuracy": 0.8451001048088074, "num_tokens": 237424517.0, "step": 6254 }, { "epoch": 0.7957002925836407, "ewc_loss": 0.099609375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 6.138123512268066, "learning_rate": 1e-05, "loss": 0.5062, "mean_token_accuracy": 0.8634864091873169, "num_tokens": 237466195.0, "step": 6255 }, { "epoch": 0.7958275028622313, "ewc_loss": 0.09716796875, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.800044536590576, "learning_rate": 1e-05, "loss": 0.5091, "mean_token_accuracy": 0.8646263480186462, "num_tokens": 237500189.0, "step": 6256 }, { "epoch": 0.7959547131408218, "ewc_loss": 0.0966796875, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.936459541320801, "learning_rate": 1e-05, "loss": 0.5324, "mean_token_accuracy": 0.8548434972763062, "num_tokens": 237534186.0, "step": 6257 }, { "epoch": 0.7960819234194123, "ewc_loss": 0.09228515625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.669668674468994, "learning_rate": 1e-05, "loss": 0.5153, "mean_token_accuracy": 0.8570711612701416, "num_tokens": 237572382.0, "step": 6258 }, { "epoch": 0.7962091336980028, "ewc_loss": 0.09521484375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.86555814743042, "learning_rate": 1e-05, "loss": 0.5113, "mean_token_accuracy": 0.8578876852989197, "num_tokens": 237614300.0, "step": 6259 }, { "epoch": 0.7963363439765934, "ewc_loss": 0.08935546875, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.5197553634643555, "learning_rate": 1e-05, "loss": 0.4896, "mean_token_accuracy": 0.862449586391449, "num_tokens": 237650952.0, "step": 6260 }, { "epoch": 0.7964635542551838, "ewc_loss": 0.09619140625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.899860858917236, "learning_rate": 1e-05, "loss": 0.5356, "mean_token_accuracy": 0.8518093824386597, "num_tokens": 237689901.0, "step": 6261 }, { "epoch": 0.7965907645337743, "ewc_loss": 0.0869140625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 7.43865966796875e-05, "grad_norm": 5.3793439865112305, "learning_rate": 1e-05, "loss": 0.5476, "mean_token_accuracy": 0.8451387286186218, "num_tokens": 237724889.0, "step": 6262 }, { "epoch": 0.7967179748123648, "ewc_loss": 0.09912109375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 6.026575088500977, "learning_rate": 1e-05, "loss": 0.4909, "mean_token_accuracy": 0.8646655678749084, "num_tokens": 237758853.0, "step": 6263 }, { "epoch": 0.7968451850909554, "ewc_loss": 0.08544921875, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 7.295608520507812e-05, "grad_norm": 5.240746974945068, "learning_rate": 1e-05, "loss": 0.5097, "mean_token_accuracy": 0.8589463829994202, "num_tokens": 237795646.0, "step": 6264 }, { "epoch": 0.7969723953695459, "ewc_loss": 0.10302734375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.25319242477417, "learning_rate": 1e-05, "loss": 0.5616, "mean_token_accuracy": 0.848219633102417, "num_tokens": 237832229.0, "step": 6265 }, { "epoch": 0.7970996056481364, "ewc_loss": 0.0849609375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 7.2479248046875e-05, "grad_norm": 5.156721591949463, "learning_rate": 1e-05, "loss": 0.5286, "mean_token_accuracy": 0.847465991973877, "num_tokens": 237870459.0, "step": 6266 }, { "epoch": 0.7972268159267268, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.578052520751953, "learning_rate": 1e-05, "loss": 0.5411, "mean_token_accuracy": 0.8552700877189636, "num_tokens": 237912367.0, "step": 6267 }, { "epoch": 0.7973540262053174, "ewc_loss": 0.08984375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.249208927154541, "learning_rate": 1e-05, "loss": 0.5109, "mean_token_accuracy": 0.8588271737098694, "num_tokens": 237950847.0, "step": 6268 }, { "epoch": 0.7974812364839079, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.826748371124268, "learning_rate": 1e-05, "loss": 0.5105, "mean_token_accuracy": 0.8675428628921509, "num_tokens": 237985325.0, "step": 6269 }, { "epoch": 0.7976084467624984, "ewc_loss": 0.0947265625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.595621109008789, "learning_rate": 1e-05, "loss": 0.5295, "mean_token_accuracy": 0.8500107526779175, "num_tokens": 238021710.0, "step": 6270 }, { "epoch": 0.797735657041089, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.706851959228516, "learning_rate": 1e-05, "loss": 0.5089, "mean_token_accuracy": 0.8657547235488892, "num_tokens": 238053024.0, "step": 6271 }, { "epoch": 0.7978628673196795, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 5.882506370544434, "learning_rate": 1e-05, "loss": 0.4994, "mean_token_accuracy": 0.8665866851806641, "num_tokens": 238089732.0, "step": 6272 }, { "epoch": 0.7979900775982699, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.281315803527832, "learning_rate": 1e-05, "loss": 0.5101, "mean_token_accuracy": 0.8608732223510742, "num_tokens": 238127583.0, "step": 6273 }, { "epoch": 0.7981172878768604, "ewc_loss": 0.10205078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 5.9643144607543945, "learning_rate": 1e-05, "loss": 0.5255, "mean_token_accuracy": 0.8585821390151978, "num_tokens": 238166328.0, "step": 6274 }, { "epoch": 0.798244498155451, "ewc_loss": 0.09326171875, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.804311275482178, "learning_rate": 1e-05, "loss": 0.5401, "mean_token_accuracy": 0.8500695824623108, "num_tokens": 238206111.0, "step": 6275 }, { "epoch": 0.7983717084340415, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.049217700958252, "learning_rate": 1e-05, "loss": 0.499, "mean_token_accuracy": 0.8615763783454895, "num_tokens": 238243793.0, "step": 6276 }, { "epoch": 0.798498918712632, "ewc_loss": 0.08642578125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 7.390975952148438e-05, "grad_norm": 5.407378673553467, "learning_rate": 1e-05, "loss": 0.47, "mean_token_accuracy": 0.8762627243995667, "num_tokens": 238284005.0, "step": 6277 }, { "epoch": 0.7986261289912225, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.262424468994141, "learning_rate": 1e-05, "loss": 0.5285, "mean_token_accuracy": 0.8550187349319458, "num_tokens": 238321180.0, "step": 6278 }, { "epoch": 0.798753339269813, "ewc_loss": 0.0859375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 7.343292236328125e-05, "grad_norm": 5.179587364196777, "learning_rate": 1e-05, "loss": 0.4859, "mean_token_accuracy": 0.8655773401260376, "num_tokens": 238364383.0, "step": 6279 }, { "epoch": 0.7988805495484035, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.593605995178223, "learning_rate": 1e-05, "loss": 0.5833, "mean_token_accuracy": 0.8472354412078857, "num_tokens": 238405547.0, "step": 6280 }, { "epoch": 0.799007759826994, "ewc_loss": 0.08935546875, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.229693412780762, "learning_rate": 1e-05, "loss": 0.4862, "mean_token_accuracy": 0.8622539043426514, "num_tokens": 238436810.0, "step": 6281 }, { "epoch": 0.7991349701055845, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.8731160163879395, "learning_rate": 1e-05, "loss": 0.5576, "mean_token_accuracy": 0.8529057502746582, "num_tokens": 238470062.0, "step": 6282 }, { "epoch": 0.7992621803841751, "ewc_loss": 0.095703125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.541690826416016, "learning_rate": 1e-05, "loss": 0.4598, "mean_token_accuracy": 0.8764440417289734, "num_tokens": 238502680.0, "step": 6283 }, { "epoch": 0.7993893906627656, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.852610111236572, "learning_rate": 1e-05, "loss": 0.5576, "mean_token_accuracy": 0.8524646759033203, "num_tokens": 238534355.0, "step": 6284 }, { "epoch": 0.799516600941356, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 5.840954780578613, "learning_rate": 1e-05, "loss": 0.534, "mean_token_accuracy": 0.8571436405181885, "num_tokens": 238572225.0, "step": 6285 }, { "epoch": 0.7996438112199465, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.560147762298584, "learning_rate": 1e-05, "loss": 0.5035, "mean_token_accuracy": 0.8644946813583374, "num_tokens": 238608313.0, "step": 6286 }, { "epoch": 0.7997710214985371, "ewc_loss": 0.10205078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 5.895941257476807, "learning_rate": 1e-05, "loss": 0.5659, "mean_token_accuracy": 0.8440467715263367, "num_tokens": 238652973.0, "step": 6287 }, { "epoch": 0.7998982317771276, "ewc_loss": 0.10302734375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.242892265319824, "learning_rate": 1e-05, "loss": 0.5297, "mean_token_accuracy": 0.8560895323753357, "num_tokens": 238688490.0, "step": 6288 }, { "epoch": 0.8000254420557181, "ewc_loss": 0.09765625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.823169231414795, "learning_rate": 1e-05, "loss": 0.4908, "mean_token_accuracy": 0.8688514232635498, "num_tokens": 238728641.0, "step": 6289 }, { "epoch": 0.8001526523343087, "ewc_loss": 0.09765625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 6.013738632202148, "learning_rate": 1e-05, "loss": 0.5507, "mean_token_accuracy": 0.8468986749649048, "num_tokens": 238764134.0, "step": 6290 }, { "epoch": 0.8002798626128991, "ewc_loss": 0.09375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.7039995193481445, "learning_rate": 1e-05, "loss": 0.4865, "mean_token_accuracy": 0.8652165532112122, "num_tokens": 238803705.0, "step": 6291 }, { "epoch": 0.8004070728914896, "ewc_loss": 0.0966796875, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.939619541168213, "learning_rate": 1e-05, "loss": 0.5326, "mean_token_accuracy": 0.8521464467048645, "num_tokens": 238836816.0, "step": 6292 }, { "epoch": 0.8005342831700801, "ewc_loss": 0.09033203125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.557753086090088, "learning_rate": 1e-05, "loss": 0.4802, "mean_token_accuracy": 0.8677100539207458, "num_tokens": 238870006.0, "step": 6293 }, { "epoch": 0.8006614934486707, "ewc_loss": 0.09814453125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.943629741668701, "learning_rate": 1e-05, "loss": 0.509, "mean_token_accuracy": 0.8579928278923035, "num_tokens": 238907018.0, "step": 6294 }, { "epoch": 0.8007887037272612, "ewc_loss": 0.087890625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 7.534027099609375e-05, "grad_norm": 5.405282020568848, "learning_rate": 1e-05, "loss": 0.5531, "mean_token_accuracy": 0.8463051319122314, "num_tokens": 238943319.0, "step": 6295 }, { "epoch": 0.8009159140058517, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 6.057334899902344, "learning_rate": 1e-05, "loss": 0.5604, "mean_token_accuracy": 0.8404541611671448, "num_tokens": 238978546.0, "step": 6296 }, { "epoch": 0.8010431242844421, "ewc_loss": 0.08642578125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 7.390975952148438e-05, "grad_norm": 5.280974388122559, "learning_rate": 1e-05, "loss": 0.5322, "mean_token_accuracy": 0.8561434745788574, "num_tokens": 239014785.0, "step": 6297 }, { "epoch": 0.8011703345630327, "ewc_loss": 0.10302734375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.265440464019775, "learning_rate": 1e-05, "loss": 0.5747, "mean_token_accuracy": 0.8387994766235352, "num_tokens": 239048212.0, "step": 6298 }, { "epoch": 0.8012975448416232, "ewc_loss": 0.0859375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 7.343292236328125e-05, "grad_norm": 5.216638565063477, "learning_rate": 1e-05, "loss": 0.5444, "mean_token_accuracy": 0.8475210666656494, "num_tokens": 239086946.0, "step": 6299 }, { "epoch": 0.8014247551202137, "ewc_loss": 0.109375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.533091068267822, "learning_rate": 1e-05, "loss": 0.4693, "mean_token_accuracy": 0.8753229379653931, "num_tokens": 239125100.0, "step": 6300 }, { "epoch": 0.8015519653988042, "ewc_loss": 0.09033203125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.29084587097168, "learning_rate": 1e-05, "loss": 0.5253, "mean_token_accuracy": 0.8513621687889099, "num_tokens": 239163118.0, "step": 6301 }, { "epoch": 0.8016791756773948, "ewc_loss": 0.11376953125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.730723857879639, "learning_rate": 1e-05, "loss": 0.5153, "mean_token_accuracy": 0.8654570579528809, "num_tokens": 239196914.0, "step": 6302 }, { "epoch": 0.8018063859559852, "ewc_loss": 0.09521484375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.564968585968018, "learning_rate": 1e-05, "loss": 0.509, "mean_token_accuracy": 0.8606210947036743, "num_tokens": 239229609.0, "step": 6303 }, { "epoch": 0.8019335962345757, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.6160149574279785, "learning_rate": 1e-05, "loss": 0.475, "mean_token_accuracy": 0.8784923553466797, "num_tokens": 239269829.0, "step": 6304 }, { "epoch": 0.8020608065131662, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 5.809865951538086, "learning_rate": 1e-05, "loss": 0.4612, "mean_token_accuracy": 0.8746665716171265, "num_tokens": 239308563.0, "step": 6305 }, { "epoch": 0.8021880167917568, "ewc_loss": 0.10400390625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.23175573348999, "learning_rate": 1e-05, "loss": 0.5151, "mean_token_accuracy": 0.8630698919296265, "num_tokens": 239348620.0, "step": 6306 }, { "epoch": 0.8023152270703473, "ewc_loss": 0.10107421875, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 5.935812473297119, "learning_rate": 1e-05, "loss": 0.5173, "mean_token_accuracy": 0.8586710691452026, "num_tokens": 239380706.0, "step": 6307 }, { "epoch": 0.8024424373489378, "ewc_loss": 0.09228515625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.753864288330078, "learning_rate": 1e-05, "loss": 0.5439, "mean_token_accuracy": 0.8481212854385376, "num_tokens": 239417365.0, "step": 6308 }, { "epoch": 0.8025696476275284, "ewc_loss": 0.10205078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 6.055037975311279, "learning_rate": 1e-05, "loss": 0.5047, "mean_token_accuracy": 0.8584332466125488, "num_tokens": 239449926.0, "step": 6309 }, { "epoch": 0.8026968579061188, "ewc_loss": 0.08544921875, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 7.295608520507812e-05, "grad_norm": 5.342441558837891, "learning_rate": 1e-05, "loss": 0.508, "mean_token_accuracy": 0.8580870032310486, "num_tokens": 239485134.0, "step": 6310 }, { "epoch": 0.8028240681847093, "ewc_loss": 0.10498046875, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.313604354858398, "learning_rate": 1e-05, "loss": 0.4988, "mean_token_accuracy": 0.8685853481292725, "num_tokens": 239521353.0, "step": 6311 }, { "epoch": 0.8029512784632998, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 7.152557373046875e-05, "grad_norm": 5.136258125305176, "learning_rate": 1e-05, "loss": 0.5135, "mean_token_accuracy": 0.8565876483917236, "num_tokens": 239554902.0, "step": 6312 }, { "epoch": 0.8030784887418904, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.688144207000732, "learning_rate": 1e-05, "loss": 0.565, "mean_token_accuracy": 0.8490233421325684, "num_tokens": 239597436.0, "step": 6313 }, { "epoch": 0.8032056990204809, "ewc_loss": 0.08935546875, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.222349166870117, "learning_rate": 1e-05, "loss": 0.4931, "mean_token_accuracy": 0.8639099597930908, "num_tokens": 239635543.0, "step": 6314 }, { "epoch": 0.8033329092990714, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 6.992156982421875, "learning_rate": 1e-05, "loss": 0.5025, "mean_token_accuracy": 0.87050861120224, "num_tokens": 239673956.0, "step": 6315 }, { "epoch": 0.8034601195776618, "ewc_loss": 0.099609375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.653764247894287, "learning_rate": 1e-05, "loss": 0.5269, "mean_token_accuracy": 0.8553773164749146, "num_tokens": 239712556.0, "step": 6316 }, { "epoch": 0.8035873298562524, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.890015602111816, "learning_rate": 1e-05, "loss": 0.4886, "mean_token_accuracy": 0.8728989362716675, "num_tokens": 239747034.0, "step": 6317 }, { "epoch": 0.8037145401348429, "ewc_loss": 0.10693359375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 5.962104320526123, "learning_rate": 1e-05, "loss": 0.5158, "mean_token_accuracy": 0.8621041178703308, "num_tokens": 239787054.0, "step": 6318 }, { "epoch": 0.8038417504134334, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.535555362701416, "learning_rate": 1e-05, "loss": 0.577, "mean_token_accuracy": 0.8461357355117798, "num_tokens": 239824796.0, "step": 6319 }, { "epoch": 0.803968960692024, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 5.9697957038879395, "learning_rate": 1e-05, "loss": 0.5461, "mean_token_accuracy": 0.8520351052284241, "num_tokens": 239862526.0, "step": 6320 }, { "epoch": 0.8040961709706145, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.196384906768799, "learning_rate": 1e-05, "loss": 0.5394, "mean_token_accuracy": 0.8560343980789185, "num_tokens": 239897409.0, "step": 6321 }, { "epoch": 0.8042233812492049, "ewc_loss": 0.0986328125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.857004642486572, "learning_rate": 1e-05, "loss": 0.4932, "mean_token_accuracy": 0.8662490844726562, "num_tokens": 239935202.0, "step": 6322 }, { "epoch": 0.8043505915277954, "ewc_loss": 0.09716796875, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.970006465911865, "learning_rate": 1e-05, "loss": 0.525, "mean_token_accuracy": 0.8616935014724731, "num_tokens": 239967603.0, "step": 6323 }, { "epoch": 0.804477801806386, "ewc_loss": 0.0947265625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.722992420196533, "learning_rate": 1e-05, "loss": 0.5323, "mean_token_accuracy": 0.8522623181343079, "num_tokens": 240000757.0, "step": 6324 }, { "epoch": 0.8046050120849765, "ewc_loss": 0.09716796875, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.872053623199463, "learning_rate": 1e-05, "loss": 0.5392, "mean_token_accuracy": 0.8529437184333801, "num_tokens": 240039747.0, "step": 6325 }, { "epoch": 0.804732222363567, "ewc_loss": 0.09130859375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.576514720916748, "learning_rate": 1e-05, "loss": 0.5217, "mean_token_accuracy": 0.8548974990844727, "num_tokens": 240078951.0, "step": 6326 }, { "epoch": 0.8048594326421575, "ewc_loss": 0.0966796875, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.872304916381836, "learning_rate": 1e-05, "loss": 0.5699, "mean_token_accuracy": 0.843033492565155, "num_tokens": 240120097.0, "step": 6327 }, { "epoch": 0.804986642920748, "ewc_loss": 0.087890625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 7.534027099609375e-05, "grad_norm": 5.417495250701904, "learning_rate": 1e-05, "loss": 0.5764, "mean_token_accuracy": 0.8413420915603638, "num_tokens": 240166510.0, "step": 6328 }, { "epoch": 0.8051138531993385, "ewc_loss": 0.099609375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.992534637451172, "learning_rate": 1e-05, "loss": 0.5139, "mean_token_accuracy": 0.8594043254852295, "num_tokens": 240207961.0, "step": 6329 }, { "epoch": 0.805241063477929, "ewc_loss": 0.0859375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 7.343292236328125e-05, "grad_norm": 5.274481296539307, "learning_rate": 1e-05, "loss": 0.4831, "mean_token_accuracy": 0.8630190491676331, "num_tokens": 240243257.0, "step": 6330 }, { "epoch": 0.8053682737565195, "ewc_loss": 0.10302734375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.226004123687744, "learning_rate": 1e-05, "loss": 0.4977, "mean_token_accuracy": 0.8651770353317261, "num_tokens": 240279844.0, "step": 6331 }, { "epoch": 0.8054954840351101, "ewc_loss": 0.0849609375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 7.2479248046875e-05, "grad_norm": 5.166256427764893, "learning_rate": 1e-05, "loss": 0.4888, "mean_token_accuracy": 0.862481951713562, "num_tokens": 240316836.0, "step": 6332 }, { "epoch": 0.8056226943137006, "ewc_loss": 0.10888671875, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.564692497253418, "learning_rate": 1e-05, "loss": 0.5771, "mean_token_accuracy": 0.8420164585113525, "num_tokens": 240357794.0, "step": 6333 }, { "epoch": 0.805749904592291, "ewc_loss": 0.0888671875, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 7.62939453125e-05, "grad_norm": 5.214350700378418, "learning_rate": 1e-05, "loss": 0.5507, "mean_token_accuracy": 0.8416533470153809, "num_tokens": 240397929.0, "step": 6334 }, { "epoch": 0.8058771148708815, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.867275714874268, "learning_rate": 1e-05, "loss": 0.4961, "mean_token_accuracy": 0.8710917234420776, "num_tokens": 240434319.0, "step": 6335 }, { "epoch": 0.8060043251494721, "ewc_loss": 0.095703125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.5483880043029785, "learning_rate": 1e-05, "loss": 0.5119, "mean_token_accuracy": 0.8590669631958008, "num_tokens": 240476755.0, "step": 6336 }, { "epoch": 0.8061315354280626, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.826785087585449, "learning_rate": 1e-05, "loss": 0.5189, "mean_token_accuracy": 0.8655481934547424, "num_tokens": 240513042.0, "step": 6337 }, { "epoch": 0.8062587457066531, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 5.893082141876221, "learning_rate": 1e-05, "loss": 0.5145, "mean_token_accuracy": 0.863402783870697, "num_tokens": 240554805.0, "step": 6338 }, { "epoch": 0.8063859559852437, "ewc_loss": 0.10693359375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.440552711486816, "learning_rate": 1e-05, "loss": 0.4817, "mean_token_accuracy": 0.8732280135154724, "num_tokens": 240593196.0, "step": 6339 }, { "epoch": 0.8065131662638341, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.005489349365234, "learning_rate": 1e-05, "loss": 0.5235, "mean_token_accuracy": 0.8617225885391235, "num_tokens": 240632640.0, "step": 6340 }, { "epoch": 0.8066403765424246, "ewc_loss": 0.09716796875, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.959218502044678, "learning_rate": 1e-05, "loss": 0.5276, "mean_token_accuracy": 0.86396723985672, "num_tokens": 240672868.0, "step": 6341 }, { "epoch": 0.8067675868210151, "ewc_loss": 0.10107421875, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 6.046103477478027, "learning_rate": 1e-05, "loss": 0.5603, "mean_token_accuracy": 0.8458244800567627, "num_tokens": 240713499.0, "step": 6342 }, { "epoch": 0.8068947970996057, "ewc_loss": 0.08984375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.532382011413574, "learning_rate": 1e-05, "loss": 0.4656, "mean_token_accuracy": 0.8703597784042358, "num_tokens": 240759684.0, "step": 6343 }, { "epoch": 0.8070220073781962, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.19621467590332, "learning_rate": 1e-05, "loss": 0.5554, "mean_token_accuracy": 0.8507299423217773, "num_tokens": 240799439.0, "step": 6344 }, { "epoch": 0.8071492176567867, "ewc_loss": 0.0859375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 7.343292236328125e-05, "grad_norm": 5.284726142883301, "learning_rate": 1e-05, "loss": 0.5265, "mean_token_accuracy": 0.8531454205513, "num_tokens": 240845569.0, "step": 6345 }, { "epoch": 0.8072764279353771, "ewc_loss": 0.10693359375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.438700199127197, "learning_rate": 1e-05, "loss": 0.467, "mean_token_accuracy": 0.8754720687866211, "num_tokens": 240879721.0, "step": 6346 }, { "epoch": 0.8074036382139677, "ewc_loss": 0.0869140625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 7.390975952148438e-05, "grad_norm": 5.220566272735596, "learning_rate": 1e-05, "loss": 0.5026, "mean_token_accuracy": 0.8627272844314575, "num_tokens": 240918433.0, "step": 6347 }, { "epoch": 0.8075308484925582, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.738380432128906, "learning_rate": 1e-05, "loss": 0.5348, "mean_token_accuracy": 0.8549680113792419, "num_tokens": 240949650.0, "step": 6348 }, { "epoch": 0.8076580587711487, "ewc_loss": 0.091796875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.404170513153076, "learning_rate": 1e-05, "loss": 0.4838, "mean_token_accuracy": 0.8665529489517212, "num_tokens": 240984803.0, "step": 6349 }, { "epoch": 0.8077852690497392, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.848444938659668, "learning_rate": 1e-05, "loss": 0.5454, "mean_token_accuracy": 0.8555505275726318, "num_tokens": 241023173.0, "step": 6350 }, { "epoch": 0.8079124793283298, "ewc_loss": 0.0986328125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.705420970916748, "learning_rate": 1e-05, "loss": 0.5097, "mean_token_accuracy": 0.8644053936004639, "num_tokens": 241065235.0, "step": 6351 }, { "epoch": 0.8080396896069202, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.676506996154785, "learning_rate": 1e-05, "loss": 0.5217, "mean_token_accuracy": 0.8611987829208374, "num_tokens": 241106283.0, "step": 6352 }, { "epoch": 0.8081668998855107, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 5.867676258087158, "learning_rate": 1e-05, "loss": 0.5542, "mean_token_accuracy": 0.8494526147842407, "num_tokens": 241145544.0, "step": 6353 }, { "epoch": 0.8082941101641012, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.377833843231201, "learning_rate": 1e-05, "loss": 0.5516, "mean_token_accuracy": 0.8532644510269165, "num_tokens": 241183888.0, "step": 6354 }, { "epoch": 0.8084213204426918, "ewc_loss": 0.0986328125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.847424507141113, "learning_rate": 1e-05, "loss": 0.4669, "mean_token_accuracy": 0.8764135837554932, "num_tokens": 241224650.0, "step": 6355 }, { "epoch": 0.8085485307212823, "ewc_loss": 0.099609375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 6.100802421569824, "learning_rate": 1e-05, "loss": 0.5192, "mean_token_accuracy": 0.8582648038864136, "num_tokens": 241263907.0, "step": 6356 }, { "epoch": 0.8086757409998728, "ewc_loss": 0.095703125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.752315044403076, "learning_rate": 1e-05, "loss": 0.5669, "mean_token_accuracy": 0.8391602039337158, "num_tokens": 241305009.0, "step": 6357 }, { "epoch": 0.8088029512784632, "ewc_loss": 0.0966796875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.953306198120117, "learning_rate": 1e-05, "loss": 0.4964, "mean_token_accuracy": 0.8667378425598145, "num_tokens": 241344283.0, "step": 6358 }, { "epoch": 0.8089301615570538, "ewc_loss": 0.0927734375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.620306968688965, "learning_rate": 1e-05, "loss": 0.4994, "mean_token_accuracy": 0.8622105121612549, "num_tokens": 241380243.0, "step": 6359 }, { "epoch": 0.8090573718356443, "ewc_loss": 0.09765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.920937538146973, "learning_rate": 1e-05, "loss": 0.5062, "mean_token_accuracy": 0.8572301268577576, "num_tokens": 241414373.0, "step": 6360 }, { "epoch": 0.8091845821142348, "ewc_loss": 0.08984375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.47720193862915, "learning_rate": 1e-05, "loss": 0.4952, "mean_token_accuracy": 0.8645703196525574, "num_tokens": 241456393.0, "step": 6361 }, { "epoch": 0.8093117923928254, "ewc_loss": 0.0986328125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.993344783782959, "learning_rate": 1e-05, "loss": 0.5449, "mean_token_accuracy": 0.8497523069381714, "num_tokens": 241496086.0, "step": 6362 }, { "epoch": 0.8094390026714159, "ewc_loss": 0.0869140625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 7.43865966796875e-05, "grad_norm": 5.336171627044678, "learning_rate": 1e-05, "loss": 0.4839, "mean_token_accuracy": 0.8639810085296631, "num_tokens": 241537814.0, "step": 6363 }, { "epoch": 0.8095662129500064, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.160519599914551, "learning_rate": 1e-05, "loss": 0.5344, "mean_token_accuracy": 0.8533028960227966, "num_tokens": 241576762.0, "step": 6364 }, { "epoch": 0.8096934232285968, "ewc_loss": 0.0859375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 7.295608520507812e-05, "grad_norm": 5.22349214553833, "learning_rate": 1e-05, "loss": 0.4621, "mean_token_accuracy": 0.8732083439826965, "num_tokens": 241614667.0, "step": 6365 }, { "epoch": 0.8098206335071874, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.4345197677612305, "learning_rate": 1e-05, "loss": 0.4722, "mean_token_accuracy": 0.8749671578407288, "num_tokens": 241647216.0, "step": 6366 }, { "epoch": 0.8099478437857779, "ewc_loss": 0.087890625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 7.534027099609375e-05, "grad_norm": 5.237203598022461, "learning_rate": 1e-05, "loss": 0.4765, "mean_token_accuracy": 0.866531491279602, "num_tokens": 241684278.0, "step": 6367 }, { "epoch": 0.8100750540643684, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.684742450714111, "learning_rate": 1e-05, "loss": 0.4773, "mean_token_accuracy": 0.8755658864974976, "num_tokens": 241723326.0, "step": 6368 }, { "epoch": 0.8102022643429589, "ewc_loss": 0.09375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.459046363830566, "learning_rate": 1e-05, "loss": 0.5028, "mean_token_accuracy": 0.8637148141860962, "num_tokens": 241765251.0, "step": 6369 }, { "epoch": 0.8103294746215495, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.693752765655518, "learning_rate": 1e-05, "loss": 0.4912, "mean_token_accuracy": 0.869705080986023, "num_tokens": 241801862.0, "step": 6370 }, { "epoch": 0.8104566849001399, "ewc_loss": 0.0986328125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.7449235916137695, "learning_rate": 1e-05, "loss": 0.5109, "mean_token_accuracy": 0.8604199886322021, "num_tokens": 241841448.0, "step": 6371 }, { "epoch": 0.8105838951787304, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.384982109069824, "learning_rate": 1e-05, "loss": 0.5283, "mean_token_accuracy": 0.856401801109314, "num_tokens": 241883151.0, "step": 6372 }, { "epoch": 0.810711105457321, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 5.899214267730713, "learning_rate": 1e-05, "loss": 0.5468, "mean_token_accuracy": 0.8516607284545898, "num_tokens": 241924643.0, "step": 6373 }, { "epoch": 0.8108383157359115, "ewc_loss": 0.0966796875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.93043327331543, "learning_rate": 1e-05, "loss": 0.5227, "mean_token_accuracy": 0.8557894825935364, "num_tokens": 241960361.0, "step": 6374 }, { "epoch": 0.810965526014502, "ewc_loss": 0.099609375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.996831893920898, "learning_rate": 1e-05, "loss": 0.564, "mean_token_accuracy": 0.8460410833358765, "num_tokens": 241999616.0, "step": 6375 }, { "epoch": 0.8110927362930925, "ewc_loss": 0.0869140625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 7.43865966796875e-05, "grad_norm": 5.48441219329834, "learning_rate": 1e-05, "loss": 0.5134, "mean_token_accuracy": 0.8567054271697998, "num_tokens": 242035824.0, "step": 6376 }, { "epoch": 0.811219946571683, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.204461574554443, "learning_rate": 1e-05, "loss": 0.5083, "mean_token_accuracy": 0.8639644384384155, "num_tokens": 242072680.0, "step": 6377 }, { "epoch": 0.8113471568502735, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 7.104873657226562e-05, "grad_norm": 5.178757190704346, "learning_rate": 1e-05, "loss": 0.4846, "mean_token_accuracy": 0.8673669099807739, "num_tokens": 242111135.0, "step": 6378 }, { "epoch": 0.811474367128864, "ewc_loss": 0.109375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.565508842468262, "learning_rate": 1e-05, "loss": 0.5016, "mean_token_accuracy": 0.8674206733703613, "num_tokens": 242148993.0, "step": 6379 }, { "epoch": 0.8116015774074545, "ewc_loss": 0.0869140625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 7.390975952148438e-05, "grad_norm": 5.152912616729736, "learning_rate": 1e-05, "loss": 0.5327, "mean_token_accuracy": 0.8569777011871338, "num_tokens": 242186230.0, "step": 6380 }, { "epoch": 0.8117287876860451, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.940990924835205, "learning_rate": 1e-05, "loss": 0.5316, "mean_token_accuracy": 0.8635799884796143, "num_tokens": 242222341.0, "step": 6381 }, { "epoch": 0.8118559979646356, "ewc_loss": 0.095703125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.508640766143799, "learning_rate": 1e-05, "loss": 0.5341, "mean_token_accuracy": 0.85113525390625, "num_tokens": 242255408.0, "step": 6382 }, { "epoch": 0.811983208243226, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.996957778930664, "learning_rate": 1e-05, "loss": 0.6077, "mean_token_accuracy": 0.8385292887687683, "num_tokens": 242294664.0, "step": 6383 }, { "epoch": 0.8121104185218165, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 5.902833938598633, "learning_rate": 1e-05, "loss": 0.4532, "mean_token_accuracy": 0.8776153922080994, "num_tokens": 242328982.0, "step": 6384 }, { "epoch": 0.8122376288004071, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.691533088684082, "learning_rate": 1e-05, "loss": 0.5803, "mean_token_accuracy": 0.8492131233215332, "num_tokens": 242369542.0, "step": 6385 }, { "epoch": 0.8123648390789976, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.004724025726318, "learning_rate": 1e-05, "loss": 0.539, "mean_token_accuracy": 0.8542347550392151, "num_tokens": 242406699.0, "step": 6386 }, { "epoch": 0.8124920493575881, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.311273574829102, "learning_rate": 1e-05, "loss": 0.5419, "mean_token_accuracy": 0.8522900938987732, "num_tokens": 242448057.0, "step": 6387 }, { "epoch": 0.8126192596361786, "ewc_loss": 0.099609375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.9085493087768555, "learning_rate": 1e-05, "loss": 0.557, "mean_token_accuracy": 0.8472863435745239, "num_tokens": 242489792.0, "step": 6388 }, { "epoch": 0.8127464699147691, "ewc_loss": 0.0986328125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 6.037530422210693, "learning_rate": 1e-05, "loss": 0.5323, "mean_token_accuracy": 0.8553648591041565, "num_tokens": 242535458.0, "step": 6389 }, { "epoch": 0.8128736801933596, "ewc_loss": 0.0966796875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.779947757720947, "learning_rate": 1e-05, "loss": 0.521, "mean_token_accuracy": 0.8599241971969604, "num_tokens": 242577380.0, "step": 6390 }, { "epoch": 0.8130008904719501, "ewc_loss": 0.0966796875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.905889987945557, "learning_rate": 1e-05, "loss": 0.5241, "mean_token_accuracy": 0.855675220489502, "num_tokens": 242614353.0, "step": 6391 }, { "epoch": 0.8131281007505406, "ewc_loss": 0.0927734375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.635933876037598, "learning_rate": 1e-05, "loss": 0.4832, "mean_token_accuracy": 0.8701366782188416, "num_tokens": 242648444.0, "step": 6392 }, { "epoch": 0.8132553110291312, "ewc_loss": 0.0966796875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.88601541519165, "learning_rate": 1e-05, "loss": 0.5241, "mean_token_accuracy": 0.8576525449752808, "num_tokens": 242687658.0, "step": 6393 }, { "epoch": 0.8133825213077217, "ewc_loss": 0.0888671875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 7.62939453125e-05, "grad_norm": 5.476099014282227, "learning_rate": 1e-05, "loss": 0.4891, "mean_token_accuracy": 0.8610516786575317, "num_tokens": 242726097.0, "step": 6394 }, { "epoch": 0.8135097315863121, "ewc_loss": 0.09765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.9691972732543945, "learning_rate": 1e-05, "loss": 0.5416, "mean_token_accuracy": 0.8539283871650696, "num_tokens": 242769229.0, "step": 6395 }, { "epoch": 0.8136369418649027, "ewc_loss": 0.0869140625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 7.390975952148438e-05, "grad_norm": 5.32735013961792, "learning_rate": 1e-05, "loss": 0.5018, "mean_token_accuracy": 0.85757976770401, "num_tokens": 242805815.0, "step": 6396 }, { "epoch": 0.8137641521434932, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 6.172773838043213, "learning_rate": 1e-05, "loss": 0.5588, "mean_token_accuracy": 0.8443666696548462, "num_tokens": 242835929.0, "step": 6397 }, { "epoch": 0.8138913624220837, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 7.152557373046875e-05, "grad_norm": 5.19297456741333, "learning_rate": 1e-05, "loss": 0.469, "mean_token_accuracy": 0.8691518902778625, "num_tokens": 242875295.0, "step": 6398 }, { "epoch": 0.8140185727006742, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.493690490722656, "learning_rate": 1e-05, "loss": 0.5188, "mean_token_accuracy": 0.857833743095398, "num_tokens": 242911825.0, "step": 6399 }, { "epoch": 0.8141457829792648, "ewc_loss": 0.087890625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 7.486343383789062e-05, "grad_norm": 5.190536022186279, "learning_rate": 1e-05, "loss": 0.4918, "mean_token_accuracy": 0.8621417880058289, "num_tokens": 242950568.0, "step": 6400 }, { "epoch": 0.8142729932578552, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.824221611022949, "learning_rate": 1e-05, "loss": 0.5094, "mean_token_accuracy": 0.8662261962890625, "num_tokens": 242981223.0, "step": 6401 }, { "epoch": 0.8144002035364457, "ewc_loss": 0.0947265625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.473766326904297, "learning_rate": 1e-05, "loss": 0.4406, "mean_token_accuracy": 0.8799237012863159, "num_tokens": 243016261.0, "step": 6402 }, { "epoch": 0.8145274138150362, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.86985969543457, "learning_rate": 1e-05, "loss": 0.5793, "mean_token_accuracy": 0.8456028699874878, "num_tokens": 243056213.0, "step": 6403 }, { "epoch": 0.8146546240936268, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 5.848796367645264, "learning_rate": 1e-05, "loss": 0.5131, "mean_token_accuracy": 0.8597931861877441, "num_tokens": 243097184.0, "step": 6404 }, { "epoch": 0.8147818343722173, "ewc_loss": 0.109375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.527095317840576, "learning_rate": 1e-05, "loss": 0.5312, "mean_token_accuracy": 0.858970582485199, "num_tokens": 243136995.0, "step": 6405 }, { "epoch": 0.8149090446508078, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 5.999874591827393, "learning_rate": 1e-05, "loss": 0.4841, "mean_token_accuracy": 0.8747850656509399, "num_tokens": 243173495.0, "step": 6406 }, { "epoch": 0.8150362549293982, "ewc_loss": 0.099609375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 6.043091773986816, "learning_rate": 1e-05, "loss": 0.5173, "mean_token_accuracy": 0.8572409152984619, "num_tokens": 243209198.0, "step": 6407 }, { "epoch": 0.8151634652079888, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 6.036405563354492, "learning_rate": 1e-05, "loss": 0.4876, "mean_token_accuracy": 0.8653204441070557, "num_tokens": 243249777.0, "step": 6408 }, { "epoch": 0.8152906754865793, "ewc_loss": 0.08984375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.593344211578369, "learning_rate": 1e-05, "loss": 0.5401, "mean_token_accuracy": 0.8498797416687012, "num_tokens": 243293094.0, "step": 6409 }, { "epoch": 0.8154178857651698, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.176380157470703, "learning_rate": 1e-05, "loss": 0.5799, "mean_token_accuracy": 0.8442977666854858, "num_tokens": 243330881.0, "step": 6410 }, { "epoch": 0.8155450960437604, "ewc_loss": 0.0859375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 7.295608520507812e-05, "grad_norm": 5.288287162780762, "learning_rate": 1e-05, "loss": 0.5288, "mean_token_accuracy": 0.8513802886009216, "num_tokens": 243371292.0, "step": 6411 }, { "epoch": 0.8156723063223509, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.445935249328613, "learning_rate": 1e-05, "loss": 0.5536, "mean_token_accuracy": 0.8455867767333984, "num_tokens": 243404925.0, "step": 6412 }, { "epoch": 0.8157995166009414, "ewc_loss": 0.0869140625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 7.390975952148438e-05, "grad_norm": 5.2004241943359375, "learning_rate": 1e-05, "loss": 0.4753, "mean_token_accuracy": 0.8682438135147095, "num_tokens": 243442377.0, "step": 6413 }, { "epoch": 0.8159267268795318, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.75091028213501, "learning_rate": 1e-05, "loss": 0.5043, "mean_token_accuracy": 0.8660157918930054, "num_tokens": 243476857.0, "step": 6414 }, { "epoch": 0.8160539371581224, "ewc_loss": 0.0927734375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.404141902923584, "learning_rate": 1e-05, "loss": 0.4502, "mean_token_accuracy": 0.8779227137565613, "num_tokens": 243506641.0, "step": 6415 }, { "epoch": 0.8161811474367129, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.880661964416504, "learning_rate": 1e-05, "loss": 0.5504, "mean_token_accuracy": 0.8546214699745178, "num_tokens": 243547555.0, "step": 6416 }, { "epoch": 0.8163083577153034, "ewc_loss": 0.099609375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.724457263946533, "learning_rate": 1e-05, "loss": 0.5616, "mean_token_accuracy": 0.8500449657440186, "num_tokens": 243588210.0, "step": 6417 }, { "epoch": 0.8164355679938939, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.6932692527771, "learning_rate": 1e-05, "loss": 0.5268, "mean_token_accuracy": 0.8578715920448303, "num_tokens": 243626913.0, "step": 6418 }, { "epoch": 0.8165627782724845, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 5.881201267242432, "learning_rate": 1e-05, "loss": 0.4833, "mean_token_accuracy": 0.8702861666679382, "num_tokens": 243662645.0, "step": 6419 }, { "epoch": 0.8166899885510749, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.376914024353027, "learning_rate": 1e-05, "loss": 0.5158, "mean_token_accuracy": 0.862747311592102, "num_tokens": 243704603.0, "step": 6420 }, { "epoch": 0.8168171988296654, "ewc_loss": 0.099609375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.857903003692627, "learning_rate": 1e-05, "loss": 0.5281, "mean_token_accuracy": 0.8527125716209412, "num_tokens": 243739571.0, "step": 6421 }, { "epoch": 0.8169444091082559, "ewc_loss": 0.0986328125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 6.095833778381348, "learning_rate": 1e-05, "loss": 0.4688, "mean_token_accuracy": 0.8757420182228088, "num_tokens": 243782047.0, "step": 6422 }, { "epoch": 0.8170716193868465, "ewc_loss": 0.0966796875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.770320415496826, "learning_rate": 1e-05, "loss": 0.5574, "mean_token_accuracy": 0.8434226512908936, "num_tokens": 243825159.0, "step": 6423 }, { "epoch": 0.817198829665437, "ewc_loss": 0.0966796875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.93572473526001, "learning_rate": 1e-05, "loss": 0.5249, "mean_token_accuracy": 0.8580174446105957, "num_tokens": 243865627.0, "step": 6424 }, { "epoch": 0.8173260399440275, "ewc_loss": 0.0927734375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.639129161834717, "learning_rate": 1e-05, "loss": 0.5219, "mean_token_accuracy": 0.853975772857666, "num_tokens": 243905750.0, "step": 6425 }, { "epoch": 0.8174532502226179, "ewc_loss": 0.09765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.915462970733643, "learning_rate": 1e-05, "loss": 0.5287, "mean_token_accuracy": 0.8515346050262451, "num_tokens": 243936176.0, "step": 6426 }, { "epoch": 0.8175804605012085, "ewc_loss": 0.08984375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.478199005126953, "learning_rate": 1e-05, "loss": 0.4673, "mean_token_accuracy": 0.8698998689651489, "num_tokens": 243979191.0, "step": 6427 }, { "epoch": 0.817707670779799, "ewc_loss": 0.099609375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.993214130401611, "learning_rate": 1e-05, "loss": 0.5474, "mean_token_accuracy": 0.8496264219284058, "num_tokens": 244025074.0, "step": 6428 }, { "epoch": 0.8178348810583895, "ewc_loss": 0.0869140625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 7.43865966796875e-05, "grad_norm": 5.331009387969971, "learning_rate": 1e-05, "loss": 0.49, "mean_token_accuracy": 0.8643056154251099, "num_tokens": 244062033.0, "step": 6429 }, { "epoch": 0.8179620913369801, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 6.174635410308838, "learning_rate": 1e-05, "loss": 0.4575, "mean_token_accuracy": 0.8767562508583069, "num_tokens": 244100025.0, "step": 6430 }, { "epoch": 0.8180893016155706, "ewc_loss": 0.0869140625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 7.390975952148438e-05, "grad_norm": 5.212174892425537, "learning_rate": 1e-05, "loss": 0.4404, "mean_token_accuracy": 0.8792014122009277, "num_tokens": 244136253.0, "step": 6431 }, { "epoch": 0.818216511894161, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.4669928550720215, "learning_rate": 1e-05, "loss": 0.517, "mean_token_accuracy": 0.8625321388244629, "num_tokens": 244174179.0, "step": 6432 }, { "epoch": 0.8183437221727515, "ewc_loss": 0.087890625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 7.534027099609375e-05, "grad_norm": 5.22669792175293, "learning_rate": 1e-05, "loss": 0.4657, "mean_token_accuracy": 0.8725248575210571, "num_tokens": 244209096.0, "step": 6433 }, { "epoch": 0.8184709324513421, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.7424726486206055, "learning_rate": 1e-05, "loss": 0.5016, "mean_token_accuracy": 0.866531491279602, "num_tokens": 244248920.0, "step": 6434 }, { "epoch": 0.8185981427299326, "ewc_loss": 0.0947265625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.4748854637146, "learning_rate": 1e-05, "loss": 0.5141, "mean_token_accuracy": 0.8601763248443604, "num_tokens": 244289338.0, "step": 6435 }, { "epoch": 0.8187253530085231, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.7583818435668945, "learning_rate": 1e-05, "loss": 0.4995, "mean_token_accuracy": 0.8704575300216675, "num_tokens": 244324844.0, "step": 6436 }, { "epoch": 0.8188525632871136, "ewc_loss": 0.099609375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.796959400177002, "learning_rate": 1e-05, "loss": 0.5714, "mean_token_accuracy": 0.8441476821899414, "num_tokens": 244364815.0, "step": 6437 }, { "epoch": 0.8189797735657041, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.402703762054443, "learning_rate": 1e-05, "loss": 0.5166, "mean_token_accuracy": 0.8617953062057495, "num_tokens": 244407509.0, "step": 6438 }, { "epoch": 0.8191069838442946, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 5.9529242515563965, "learning_rate": 1e-05, "loss": 0.4841, "mean_token_accuracy": 0.8742046356201172, "num_tokens": 244446663.0, "step": 6439 }, { "epoch": 0.8192341941228851, "ewc_loss": 0.095703125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.907792568206787, "learning_rate": 1e-05, "loss": 0.5249, "mean_token_accuracy": 0.8558316230773926, "num_tokens": 244492292.0, "step": 6440 }, { "epoch": 0.8193614044014756, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.045904159545898, "learning_rate": 1e-05, "loss": 0.5064, "mean_token_accuracy": 0.8611376285552979, "num_tokens": 244531390.0, "step": 6441 }, { "epoch": 0.8194886146800662, "ewc_loss": 0.0869140625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 7.43865966796875e-05, "grad_norm": 5.471379280090332, "learning_rate": 1e-05, "loss": 0.4836, "mean_token_accuracy": 0.865151047706604, "num_tokens": 244567377.0, "step": 6442 }, { "epoch": 0.8196158249586567, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.254844665527344, "learning_rate": 1e-05, "loss": 0.4654, "mean_token_accuracy": 0.8766031265258789, "num_tokens": 244605690.0, "step": 6443 }, { "epoch": 0.8197430352372471, "ewc_loss": 0.0849609375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 7.200241088867188e-05, "grad_norm": 5.199347496032715, "learning_rate": 1e-05, "loss": 0.4905, "mean_token_accuracy": 0.8625310659408569, "num_tokens": 244643512.0, "step": 6444 }, { "epoch": 0.8198702455158376, "ewc_loss": 0.109375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.606767654418945, "learning_rate": 1e-05, "loss": 0.5607, "mean_token_accuracy": 0.849234938621521, "num_tokens": 244691048.0, "step": 6445 }, { "epoch": 0.8199974557944282, "ewc_loss": 0.0888671875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 7.581710815429688e-05, "grad_norm": 5.196954250335693, "learning_rate": 1e-05, "loss": 0.5687, "mean_token_accuracy": 0.8389145135879517, "num_tokens": 244734205.0, "step": 6446 }, { "epoch": 0.8201246660730187, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.977221965789795, "learning_rate": 1e-05, "loss": 0.5646, "mean_token_accuracy": 0.8585331439971924, "num_tokens": 244768302.0, "step": 6447 }, { "epoch": 0.8202518763516092, "ewc_loss": 0.095703125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.567706108093262, "learning_rate": 1e-05, "loss": 0.508, "mean_token_accuracy": 0.8612529039382935, "num_tokens": 244808880.0, "step": 6448 }, { "epoch": 0.8203790866301998, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.958043575286865, "learning_rate": 1e-05, "loss": 0.5091, "mean_token_accuracy": 0.8701583743095398, "num_tokens": 244845507.0, "step": 6449 }, { "epoch": 0.8205062969087902, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 5.92664909362793, "learning_rate": 1e-05, "loss": 0.5555, "mean_token_accuracy": 0.8487275242805481, "num_tokens": 244888142.0, "step": 6450 }, { "epoch": 0.8206335071873807, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.613259315490723, "learning_rate": 1e-05, "loss": 0.4699, "mean_token_accuracy": 0.8768577575683594, "num_tokens": 244928612.0, "step": 6451 }, { "epoch": 0.8207607174659712, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 5.982538223266602, "learning_rate": 1e-05, "loss": 0.5114, "mean_token_accuracy": 0.8628048300743103, "num_tokens": 244969879.0, "step": 6452 }, { "epoch": 0.8208879277445618, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.281620025634766, "learning_rate": 1e-05, "loss": 0.5241, "mean_token_accuracy": 0.8604853749275208, "num_tokens": 245004597.0, "step": 6453 }, { "epoch": 0.8210151380231523, "ewc_loss": 0.0986328125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.88062858581543, "learning_rate": 1e-05, "loss": 0.5491, "mean_token_accuracy": 0.854145884513855, "num_tokens": 245050087.0, "step": 6454 }, { "epoch": 0.8211423483017428, "ewc_loss": 0.09765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 6.040031909942627, "learning_rate": 1e-05, "loss": 0.5465, "mean_token_accuracy": 0.8530902862548828, "num_tokens": 245091626.0, "step": 6455 }, { "epoch": 0.8212695585803332, "ewc_loss": 0.095703125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.741072177886963, "learning_rate": 1e-05, "loss": 0.5085, "mean_token_accuracy": 0.8580521941184998, "num_tokens": 245126780.0, "step": 6456 }, { "epoch": 0.8213967688589238, "ewc_loss": 0.0966796875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.950541973114014, "learning_rate": 1e-05, "loss": 0.5672, "mean_token_accuracy": 0.8429149389266968, "num_tokens": 245169226.0, "step": 6457 }, { "epoch": 0.8215239791375143, "ewc_loss": 0.08984375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.588552474975586, "learning_rate": 1e-05, "loss": 0.5492, "mean_token_accuracy": 0.8485044836997986, "num_tokens": 245204983.0, "step": 6458 }, { "epoch": 0.8216511894161048, "ewc_loss": 0.0986328125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.944509983062744, "learning_rate": 1e-05, "loss": 0.4961, "mean_token_accuracy": 0.8676103949546814, "num_tokens": 245249162.0, "step": 6459 }, { "epoch": 0.8217783996946953, "ewc_loss": 0.087890625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 7.486343383789062e-05, "grad_norm": 5.441554546356201, "learning_rate": 1e-05, "loss": 0.5537, "mean_token_accuracy": 0.845909059047699, "num_tokens": 245288883.0, "step": 6460 }, { "epoch": 0.8219056099732859, "ewc_loss": 0.099609375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 6.05715274810791, "learning_rate": 1e-05, "loss": 0.5188, "mean_token_accuracy": 0.8574646711349487, "num_tokens": 245328578.0, "step": 6461 }, { "epoch": 0.8220328202518764, "ewc_loss": 0.0869140625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 7.390975952148438e-05, "grad_norm": 5.2971882820129395, "learning_rate": 1e-05, "loss": 0.5282, "mean_token_accuracy": 0.8562686443328857, "num_tokens": 245368086.0, "step": 6462 }, { "epoch": 0.8221600305304668, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.2865519523620605, "learning_rate": 1e-05, "loss": 0.5332, "mean_token_accuracy": 0.8591764569282532, "num_tokens": 245405234.0, "step": 6463 }, { "epoch": 0.8222872408090574, "ewc_loss": 0.0869140625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 7.390975952148438e-05, "grad_norm": 5.2143940925598145, "learning_rate": 1e-05, "loss": 0.4715, "mean_token_accuracy": 0.8691304922103882, "num_tokens": 245439759.0, "step": 6464 }, { "epoch": 0.8224144510876479, "ewc_loss": 0.109375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.590117454528809, "learning_rate": 1e-05, "loss": 0.5322, "mean_token_accuracy": 0.8554983735084534, "num_tokens": 245474689.0, "step": 6465 }, { "epoch": 0.8225416613662384, "ewc_loss": 0.0908203125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.297660827636719, "learning_rate": 1e-05, "loss": 0.4991, "mean_token_accuracy": 0.861519455909729, "num_tokens": 245509494.0, "step": 6466 }, { "epoch": 0.8226688716448289, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.808160305023193, "learning_rate": 1e-05, "loss": 0.5152, "mean_token_accuracy": 0.8641211986541748, "num_tokens": 245550798.0, "step": 6467 }, { "epoch": 0.8227960819234195, "ewc_loss": 0.095703125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.600261688232422, "learning_rate": 1e-05, "loss": 0.5072, "mean_token_accuracy": 0.8621273040771484, "num_tokens": 245588929.0, "step": 6468 }, { "epoch": 0.8229232922020099, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.704766273498535, "learning_rate": 1e-05, "loss": 0.4977, "mean_token_accuracy": 0.8624287843704224, "num_tokens": 245625952.0, "step": 6469 }, { "epoch": 0.8230505024806004, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 5.8818039894104, "learning_rate": 1e-05, "loss": 0.5392, "mean_token_accuracy": 0.851188063621521, "num_tokens": 245668156.0, "step": 6470 }, { "epoch": 0.8231777127591909, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.298022747039795, "learning_rate": 1e-05, "loss": 0.5279, "mean_token_accuracy": 0.8606494069099426, "num_tokens": 245707117.0, "step": 6471 }, { "epoch": 0.8233049230377815, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 5.987618923187256, "learning_rate": 1e-05, "loss": 0.5313, "mean_token_accuracy": 0.8555630445480347, "num_tokens": 245747248.0, "step": 6472 }, { "epoch": 0.823432133316372, "ewc_loss": 0.0927734375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.796695709228516, "learning_rate": 1e-05, "loss": 0.5153, "mean_token_accuracy": 0.8558813333511353, "num_tokens": 245795760.0, "step": 6473 }, { "epoch": 0.8235593435949625, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 6.096295356750488, "learning_rate": 1e-05, "loss": 0.5, "mean_token_accuracy": 0.8656169772148132, "num_tokens": 245827778.0, "step": 6474 }, { "epoch": 0.8236865538735529, "ewc_loss": 0.0859375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 7.343292236328125e-05, "grad_norm": 5.387549877166748, "learning_rate": 1e-05, "loss": 0.4848, "mean_token_accuracy": 0.8653767704963684, "num_tokens": 245866869.0, "step": 6475 }, { "epoch": 0.8238137641521435, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.348476409912109, "learning_rate": 1e-05, "loss": 0.4714, "mean_token_accuracy": 0.8739585876464844, "num_tokens": 245904686.0, "step": 6476 }, { "epoch": 0.823940974430734, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 7.104873657226562e-05, "grad_norm": 5.180521011352539, "learning_rate": 1e-05, "loss": 0.5044, "mean_token_accuracy": 0.8591176271438599, "num_tokens": 245942169.0, "step": 6477 }, { "epoch": 0.8240681847093245, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.717986583709717, "learning_rate": 1e-05, "loss": 0.5353, "mean_token_accuracy": 0.8546541929244995, "num_tokens": 245979957.0, "step": 6478 }, { "epoch": 0.824195394987915, "ewc_loss": 0.0888671875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 7.62939453125e-05, "grad_norm": 5.273026466369629, "learning_rate": 1e-05, "loss": 0.4865, "mean_token_accuracy": 0.8681585788726807, "num_tokens": 246019433.0, "step": 6479 }, { "epoch": 0.8243226052665056, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 7.006631851196289, "learning_rate": 1e-05, "loss": 0.5402, "mean_token_accuracy": 0.8592613935470581, "num_tokens": 246057595.0, "step": 6480 }, { "epoch": 0.824449815545096, "ewc_loss": 0.099609375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.686779499053955, "learning_rate": 1e-05, "loss": 0.4751, "mean_token_accuracy": 0.8708570599555969, "num_tokens": 246100141.0, "step": 6481 }, { "epoch": 0.8245770258236865, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.91122579574585, "learning_rate": 1e-05, "loss": 0.5974, "mean_token_accuracy": 0.8406223058700562, "num_tokens": 246141642.0, "step": 6482 }, { "epoch": 0.824704236102277, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 5.994266510009766, "learning_rate": 1e-05, "loss": 0.5366, "mean_token_accuracy": 0.8592382669448853, "num_tokens": 246177844.0, "step": 6483 }, { "epoch": 0.8248314463808676, "ewc_loss": 0.109375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.539087295532227, "learning_rate": 1e-05, "loss": 0.5312, "mean_token_accuracy": 0.8574639558792114, "num_tokens": 246217470.0, "step": 6484 }, { "epoch": 0.8249586566594581, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 5.985361576080322, "learning_rate": 1e-05, "loss": 0.4998, "mean_token_accuracy": 0.8664464950561523, "num_tokens": 246256637.0, "step": 6485 }, { "epoch": 0.8250858669380486, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 6.196025371551514, "learning_rate": 1e-05, "loss": 0.4907, "mean_token_accuracy": 0.8689132928848267, "num_tokens": 246298894.0, "step": 6486 }, { "epoch": 0.8252130772166391, "ewc_loss": 0.09765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.8615522384643555, "learning_rate": 1e-05, "loss": 0.5374, "mean_token_accuracy": 0.8543948531150818, "num_tokens": 246340986.0, "step": 6487 }, { "epoch": 0.8253402874952296, "ewc_loss": 0.09765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.993005275726318, "learning_rate": 1e-05, "loss": 0.5962, "mean_token_accuracy": 0.8332529067993164, "num_tokens": 246379201.0, "step": 6488 }, { "epoch": 0.8254674977738201, "ewc_loss": 0.095703125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.728135585784912, "learning_rate": 1e-05, "loss": 0.4852, "mean_token_accuracy": 0.8670337200164795, "num_tokens": 246411605.0, "step": 6489 }, { "epoch": 0.8255947080524106, "ewc_loss": 0.09765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.914755821228027, "learning_rate": 1e-05, "loss": 0.5065, "mean_token_accuracy": 0.8620650768280029, "num_tokens": 246450908.0, "step": 6490 }, { "epoch": 0.8257219183310012, "ewc_loss": 0.0908203125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.556025981903076, "learning_rate": 1e-05, "loss": 0.5139, "mean_token_accuracy": 0.8558411598205566, "num_tokens": 246493132.0, "step": 6491 }, { "epoch": 0.8258491286095917, "ewc_loss": 0.09765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.9837846755981445, "learning_rate": 1e-05, "loss": 0.5459, "mean_token_accuracy": 0.8496802449226379, "num_tokens": 246528532.0, "step": 6492 }, { "epoch": 0.8259763388881821, "ewc_loss": 0.0869140625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 7.43865966796875e-05, "grad_norm": 5.387380123138428, "learning_rate": 1e-05, "loss": 0.4936, "mean_token_accuracy": 0.8622256517410278, "num_tokens": 246567328.0, "step": 6493 }, { "epoch": 0.8261035491667726, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.146735668182373, "learning_rate": 1e-05, "loss": 0.6041, "mean_token_accuracy": 0.8373851776123047, "num_tokens": 246613798.0, "step": 6494 }, { "epoch": 0.8262307594453632, "ewc_loss": 0.0859375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 7.295608520507812e-05, "grad_norm": 5.248387336730957, "learning_rate": 1e-05, "loss": 0.4819, "mean_token_accuracy": 0.867362380027771, "num_tokens": 246649117.0, "step": 6495 }, { "epoch": 0.8263579697239537, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.4167256355285645, "learning_rate": 1e-05, "loss": 0.5083, "mean_token_accuracy": 0.8638442158699036, "num_tokens": 246690304.0, "step": 6496 }, { "epoch": 0.8264851800025442, "ewc_loss": 0.0869140625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 7.43865966796875e-05, "grad_norm": 5.211557865142822, "learning_rate": 1e-05, "loss": 0.4687, "mean_token_accuracy": 0.8655736446380615, "num_tokens": 246728103.0, "step": 6497 }, { "epoch": 0.8266123902811348, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.743106365203857, "learning_rate": 1e-05, "loss": 0.5037, "mean_token_accuracy": 0.8703715801239014, "num_tokens": 246763902.0, "step": 6498 }, { "epoch": 0.8267396005597252, "ewc_loss": 0.0927734375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.41670036315918, "learning_rate": 1e-05, "loss": 0.5628, "mean_token_accuracy": 0.8451946973800659, "num_tokens": 246803899.0, "step": 6499 }, { "epoch": 0.8268668108383157, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.856971263885498, "learning_rate": 1e-05, "loss": 0.5367, "mean_token_accuracy": 0.853348970413208, "num_tokens": 246840464.0, "step": 6500 }, { "epoch": 0.8269940211169062, "ewc_loss": 0.099609375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.770105361938477, "learning_rate": 1e-05, "loss": 0.5074, "mean_token_accuracy": 0.8609280586242676, "num_tokens": 246886743.0, "step": 6501 }, { "epoch": 0.8271212313954968, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.591916084289551, "learning_rate": 1e-05, "loss": 0.5078, "mean_token_accuracy": 0.8623181581497192, "num_tokens": 246925260.0, "step": 6502 }, { "epoch": 0.8272484416740873, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 5.970005989074707, "learning_rate": 1e-05, "loss": 0.5433, "mean_token_accuracy": 0.8528164625167847, "num_tokens": 246961523.0, "step": 6503 }, { "epoch": 0.8273756519526778, "ewc_loss": 0.099609375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 6.1129279136657715, "learning_rate": 1e-05, "loss": 0.5032, "mean_token_accuracy": 0.8647062182426453, "num_tokens": 246996881.0, "step": 6504 }, { "epoch": 0.8275028622312682, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.029979705810547, "learning_rate": 1e-05, "loss": 0.5348, "mean_token_accuracy": 0.8535515069961548, "num_tokens": 247033492.0, "step": 6505 }, { "epoch": 0.8276300725098588, "ewc_loss": 0.0908203125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.646003246307373, "learning_rate": 1e-05, "loss": 0.486, "mean_token_accuracy": 0.864108681678772, "num_tokens": 247065078.0, "step": 6506 }, { "epoch": 0.8277572827884493, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.1670708656311035, "learning_rate": 1e-05, "loss": 0.5489, "mean_token_accuracy": 0.8493064641952515, "num_tokens": 247103876.0, "step": 6507 }, { "epoch": 0.8278844930670398, "ewc_loss": 0.0859375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 7.295608520507812e-05, "grad_norm": 5.311063289642334, "learning_rate": 1e-05, "loss": 0.5463, "mean_token_accuracy": 0.852095901966095, "num_tokens": 247142132.0, "step": 6508 }, { "epoch": 0.8280117033456303, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.431970596313477, "learning_rate": 1e-05, "loss": 0.5188, "mean_token_accuracy": 0.8604382872581482, "num_tokens": 247182141.0, "step": 6509 }, { "epoch": 0.8281389136242209, "ewc_loss": 0.0859375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 7.295608520507812e-05, "grad_norm": 5.191661357879639, "learning_rate": 1e-05, "loss": 0.5722, "mean_token_accuracy": 0.8423089981079102, "num_tokens": 247218240.0, "step": 6510 }, { "epoch": 0.8282661239028114, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.764354228973389, "learning_rate": 1e-05, "loss": 0.5321, "mean_token_accuracy": 0.8567715883255005, "num_tokens": 247257370.0, "step": 6511 }, { "epoch": 0.8283933341814018, "ewc_loss": 0.091796875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.371013164520264, "learning_rate": 1e-05, "loss": 0.5088, "mean_token_accuracy": 0.8612881302833557, "num_tokens": 247294205.0, "step": 6512 }, { "epoch": 0.8285205444599923, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.9502410888671875, "learning_rate": 1e-05, "loss": 0.5774, "mean_token_accuracy": 0.8449038863182068, "num_tokens": 247329974.0, "step": 6513 }, { "epoch": 0.8286477547385829, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 5.7407917976379395, "learning_rate": 1e-05, "loss": 0.5298, "mean_token_accuracy": 0.8565191030502319, "num_tokens": 247365279.0, "step": 6514 }, { "epoch": 0.8287749650171734, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.760675430297852, "learning_rate": 1e-05, "loss": 0.4831, "mean_token_accuracy": 0.8741750121116638, "num_tokens": 247405319.0, "step": 6515 }, { "epoch": 0.8289021752957639, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 5.928379058837891, "learning_rate": 1e-05, "loss": 0.468, "mean_token_accuracy": 0.8734304308891296, "num_tokens": 247444399.0, "step": 6516 }, { "epoch": 0.8290293855743545, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.443000316619873, "learning_rate": 1e-05, "loss": 0.5077, "mean_token_accuracy": 0.8644869327545166, "num_tokens": 247475637.0, "step": 6517 }, { "epoch": 0.8291565958529449, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 5.91358757019043, "learning_rate": 1e-05, "loss": 0.4959, "mean_token_accuracy": 0.864949643611908, "num_tokens": 247511009.0, "step": 6518 }, { "epoch": 0.8292838061315354, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 6.140038013458252, "learning_rate": 1e-05, "loss": 0.505, "mean_token_accuracy": 0.8640496730804443, "num_tokens": 247549188.0, "step": 6519 }, { "epoch": 0.8294110164101259, "ewc_loss": 0.09765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.804788589477539, "learning_rate": 1e-05, "loss": 0.5217, "mean_token_accuracy": 0.8554100394248962, "num_tokens": 247591914.0, "step": 6520 }, { "epoch": 0.8295382266887165, "ewc_loss": 0.0966796875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.965559005737305, "learning_rate": 1e-05, "loss": 0.4993, "mean_token_accuracy": 0.8615427017211914, "num_tokens": 247626324.0, "step": 6521 }, { "epoch": 0.829665436967307, "ewc_loss": 0.09375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.675583839416504, "learning_rate": 1e-05, "loss": 0.5149, "mean_token_accuracy": 0.8594704866409302, "num_tokens": 247662846.0, "step": 6522 }, { "epoch": 0.8297926472458975, "ewc_loss": 0.09765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.914815902709961, "learning_rate": 1e-05, "loss": 0.486, "mean_token_accuracy": 0.8639513254165649, "num_tokens": 247700831.0, "step": 6523 }, { "epoch": 0.8299198575244879, "ewc_loss": 0.08984375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.519368648529053, "learning_rate": 1e-05, "loss": 0.5099, "mean_token_accuracy": 0.860471785068512, "num_tokens": 247740535.0, "step": 6524 }, { "epoch": 0.8300470678030785, "ewc_loss": 0.0986328125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.9691572189331055, "learning_rate": 1e-05, "loss": 0.5233, "mean_token_accuracy": 0.8567832112312317, "num_tokens": 247779490.0, "step": 6525 }, { "epoch": 0.830174278081669, "ewc_loss": 0.087890625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 7.486343383789062e-05, "grad_norm": 5.372980117797852, "learning_rate": 1e-05, "loss": 0.5256, "mean_token_accuracy": 0.8511204123497009, "num_tokens": 247815867.0, "step": 6526 }, { "epoch": 0.8303014883602595, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 6.129404544830322, "learning_rate": 1e-05, "loss": 0.6182, "mean_token_accuracy": 0.8305971622467041, "num_tokens": 247854871.0, "step": 6527 }, { "epoch": 0.83042869863885, "ewc_loss": 0.0869140625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 7.390975952148438e-05, "grad_norm": 5.228126049041748, "learning_rate": 1e-05, "loss": 0.5301, "mean_token_accuracy": 0.8508356213569641, "num_tokens": 247897711.0, "step": 6528 }, { "epoch": 0.8305559089174406, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.415115833282471, "learning_rate": 1e-05, "loss": 0.5197, "mean_token_accuracy": 0.8581052422523499, "num_tokens": 247933792.0, "step": 6529 }, { "epoch": 0.830683119196031, "ewc_loss": 0.087890625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 7.534027099609375e-05, "grad_norm": 5.204556465148926, "learning_rate": 1e-05, "loss": 0.4882, "mean_token_accuracy": 0.8633732199668884, "num_tokens": 247964183.0, "step": 6530 }, { "epoch": 0.8308103294746215, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.724499702453613, "learning_rate": 1e-05, "loss": 0.5553, "mean_token_accuracy": 0.850381076335907, "num_tokens": 248003064.0, "step": 6531 }, { "epoch": 0.830937539753212, "ewc_loss": 0.0927734375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.399587154388428, "learning_rate": 1e-05, "loss": 0.4916, "mean_token_accuracy": 0.8663674592971802, "num_tokens": 248041025.0, "step": 6532 }, { "epoch": 0.8310647500318026, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.845626354217529, "learning_rate": 1e-05, "loss": 0.5631, "mean_token_accuracy": 0.8536591529846191, "num_tokens": 248079503.0, "step": 6533 }, { "epoch": 0.8311919603103931, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 5.753511428833008, "learning_rate": 1e-05, "loss": 0.5194, "mean_token_accuracy": 0.8594106435775757, "num_tokens": 248119541.0, "step": 6534 }, { "epoch": 0.8313191705889836, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.5818190574646, "learning_rate": 1e-05, "loss": 0.5567, "mean_token_accuracy": 0.8534219861030579, "num_tokens": 248163359.0, "step": 6535 }, { "epoch": 0.831446380867574, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 5.9465532302856445, "learning_rate": 1e-05, "loss": 0.4826, "mean_token_accuracy": 0.8715082406997681, "num_tokens": 248204384.0, "step": 6536 }, { "epoch": 0.8315735911461646, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 6.11706018447876, "learning_rate": 1e-05, "loss": 0.4964, "mean_token_accuracy": 0.8665478825569153, "num_tokens": 248238812.0, "step": 6537 }, { "epoch": 0.8317008014247551, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 6.017312049865723, "learning_rate": 1e-05, "loss": 0.5145, "mean_token_accuracy": 0.8631396293640137, "num_tokens": 248277725.0, "step": 6538 }, { "epoch": 0.8318280117033456, "ewc_loss": 0.0908203125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.6548333168029785, "learning_rate": 1e-05, "loss": 0.5524, "mean_token_accuracy": 0.853506326675415, "num_tokens": 248317591.0, "step": 6539 }, { "epoch": 0.8319552219819362, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.154040336608887, "learning_rate": 1e-05, "loss": 0.4914, "mean_token_accuracy": 0.8684919476509094, "num_tokens": 248356330.0, "step": 6540 }, { "epoch": 0.8320824322605267, "ewc_loss": 0.0859375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 7.295608520507812e-05, "grad_norm": 5.3135247230529785, "learning_rate": 1e-05, "loss": 0.4837, "mean_token_accuracy": 0.8625410795211792, "num_tokens": 248390854.0, "step": 6541 }, { "epoch": 0.8322096425391171, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.4109272956848145, "learning_rate": 1e-05, "loss": 0.4742, "mean_token_accuracy": 0.8728326559066772, "num_tokens": 248431581.0, "step": 6542 }, { "epoch": 0.8323368528177076, "ewc_loss": 0.0859375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 7.295608520507812e-05, "grad_norm": 5.176322937011719, "learning_rate": 1e-05, "loss": 0.4738, "mean_token_accuracy": 0.868553876876831, "num_tokens": 248474009.0, "step": 6543 }, { "epoch": 0.8324640630962982, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.773792743682861, "learning_rate": 1e-05, "loss": 0.4949, "mean_token_accuracy": 0.8705846071243286, "num_tokens": 248512552.0, "step": 6544 }, { "epoch": 0.8325912733748887, "ewc_loss": 0.091796875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.3643574714660645, "learning_rate": 1e-05, "loss": 0.4988, "mean_token_accuracy": 0.8654815554618835, "num_tokens": 248549729.0, "step": 6545 }, { "epoch": 0.8327184836534792, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.954399108886719, "learning_rate": 1e-05, "loss": 0.5332, "mean_token_accuracy": 0.8601781129837036, "num_tokens": 248589481.0, "step": 6546 }, { "epoch": 0.8328456939320698, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 5.753809452056885, "learning_rate": 1e-05, "loss": 0.471, "mean_token_accuracy": 0.8720076084136963, "num_tokens": 248621871.0, "step": 6547 }, { "epoch": 0.8329729042106602, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.770862579345703, "learning_rate": 1e-05, "loss": 0.5452, "mean_token_accuracy": 0.8589335680007935, "num_tokens": 248659416.0, "step": 6548 }, { "epoch": 0.8331001144892507, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 5.926604747772217, "learning_rate": 1e-05, "loss": 0.5179, "mean_token_accuracy": 0.8596997857093811, "num_tokens": 248700052.0, "step": 6549 }, { "epoch": 0.8332273247678412, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.445039749145508, "learning_rate": 1e-05, "loss": 0.5402, "mean_token_accuracy": 0.8524233102798462, "num_tokens": 248731965.0, "step": 6550 }, { "epoch": 0.8333545350464318, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 5.9080424308776855, "learning_rate": 1e-05, "loss": 0.5269, "mean_token_accuracy": 0.8555386662483215, "num_tokens": 248770030.0, "step": 6551 }, { "epoch": 0.8334817453250223, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 6.150302410125732, "learning_rate": 1e-05, "loss": 0.5452, "mean_token_accuracy": 0.8519390821456909, "num_tokens": 248811990.0, "step": 6552 }, { "epoch": 0.8336089556036128, "ewc_loss": 0.09765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.827033996582031, "learning_rate": 1e-05, "loss": 0.5482, "mean_token_accuracy": 0.8464391231536865, "num_tokens": 248842728.0, "step": 6553 }, { "epoch": 0.8337361658822032, "ewc_loss": 0.0966796875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.943647861480713, "learning_rate": 1e-05, "loss": 0.5497, "mean_token_accuracy": 0.8444476127624512, "num_tokens": 248885084.0, "step": 6554 }, { "epoch": 0.8338633761607938, "ewc_loss": 0.0947265625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.699681758880615, "learning_rate": 1e-05, "loss": 0.478, "mean_token_accuracy": 0.8692396879196167, "num_tokens": 248923701.0, "step": 6555 }, { "epoch": 0.8339905864393843, "ewc_loss": 0.0966796875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.874704360961914, "learning_rate": 1e-05, "loss": 0.4734, "mean_token_accuracy": 0.8711197972297668, "num_tokens": 248960796.0, "step": 6556 }, { "epoch": 0.8341177967179748, "ewc_loss": 0.091796875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.55946683883667, "learning_rate": 1e-05, "loss": 0.4954, "mean_token_accuracy": 0.8627055287361145, "num_tokens": 249000628.0, "step": 6557 }, { "epoch": 0.8342450069965653, "ewc_loss": 0.0966796875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.917347431182861, "learning_rate": 1e-05, "loss": 0.4906, "mean_token_accuracy": 0.8647215366363525, "num_tokens": 249034409.0, "step": 6558 }, { "epoch": 0.8343722172751559, "ewc_loss": 0.087890625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 7.486343383789062e-05, "grad_norm": 5.400305271148682, "learning_rate": 1e-05, "loss": 0.4777, "mean_token_accuracy": 0.8693826794624329, "num_tokens": 249071227.0, "step": 6559 }, { "epoch": 0.8344994275537464, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 6.075107097625732, "learning_rate": 1e-05, "loss": 0.5059, "mean_token_accuracy": 0.8640556931495667, "num_tokens": 249105859.0, "step": 6560 }, { "epoch": 0.8346266378323368, "ewc_loss": 0.0859375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 7.343292236328125e-05, "grad_norm": 5.247683525085449, "learning_rate": 1e-05, "loss": 0.4944, "mean_token_accuracy": 0.8644075989723206, "num_tokens": 249148237.0, "step": 6561 }, { "epoch": 0.8347538481109273, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.347409725189209, "learning_rate": 1e-05, "loss": 0.5511, "mean_token_accuracy": 0.8519771099090576, "num_tokens": 249181745.0, "step": 6562 }, { "epoch": 0.8348810583895179, "ewc_loss": 0.0859375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 7.343292236328125e-05, "grad_norm": 5.176180839538574, "learning_rate": 1e-05, "loss": 0.5167, "mean_token_accuracy": 0.8547905683517456, "num_tokens": 249221839.0, "step": 6563 }, { "epoch": 0.8350082686681084, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.6882758140563965, "learning_rate": 1e-05, "loss": 0.5522, "mean_token_accuracy": 0.8551236987113953, "num_tokens": 249264338.0, "step": 6564 }, { "epoch": 0.8351354789466989, "ewc_loss": 0.0908203125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.306385517120361, "learning_rate": 1e-05, "loss": 0.5246, "mean_token_accuracy": 0.859126091003418, "num_tokens": 249304624.0, "step": 6565 }, { "epoch": 0.8352626892252895, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.931198596954346, "learning_rate": 1e-05, "loss": 0.525, "mean_token_accuracy": 0.860987663269043, "num_tokens": 249342996.0, "step": 6566 }, { "epoch": 0.8353898995038799, "ewc_loss": 0.099609375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.687921047210693, "learning_rate": 1e-05, "loss": 0.5239, "mean_token_accuracy": 0.8567421436309814, "num_tokens": 249384795.0, "step": 6567 }, { "epoch": 0.8355171097824704, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.767984390258789, "learning_rate": 1e-05, "loss": 0.5097, "mean_token_accuracy": 0.8690141439437866, "num_tokens": 249418492.0, "step": 6568 }, { "epoch": 0.8356443200610609, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 5.973509311676025, "learning_rate": 1e-05, "loss": 0.5222, "mean_token_accuracy": 0.8621819019317627, "num_tokens": 249455714.0, "step": 6569 }, { "epoch": 0.8357715303396515, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.311279296875, "learning_rate": 1e-05, "loss": 0.5826, "mean_token_accuracy": 0.843637228012085, "num_tokens": 249493162.0, "step": 6570 }, { "epoch": 0.835898740618242, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.041868209838867, "learning_rate": 1e-05, "loss": 0.5575, "mean_token_accuracy": 0.8506935238838196, "num_tokens": 249538154.0, "step": 6571 }, { "epoch": 0.8360259508968325, "ewc_loss": 0.09375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.822681427001953, "learning_rate": 1e-05, "loss": 0.4835, "mean_token_accuracy": 0.8668407797813416, "num_tokens": 249576457.0, "step": 6572 }, { "epoch": 0.8361531611754229, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.120215892791748, "learning_rate": 1e-05, "loss": 0.5287, "mean_token_accuracy": 0.8578819632530212, "num_tokens": 249611408.0, "step": 6573 }, { "epoch": 0.8362803714540135, "ewc_loss": 0.087890625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 7.534027099609375e-05, "grad_norm": 5.4408979415893555, "learning_rate": 1e-05, "loss": 0.5185, "mean_token_accuracy": 0.8564720749855042, "num_tokens": 249651371.0, "step": 6574 }, { "epoch": 0.836407581732604, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.309813499450684, "learning_rate": 1e-05, "loss": 0.5283, "mean_token_accuracy": 0.8589843511581421, "num_tokens": 249697669.0, "step": 6575 }, { "epoch": 0.8365347920111945, "ewc_loss": 0.0859375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 7.343292236328125e-05, "grad_norm": 5.247786521911621, "learning_rate": 1e-05, "loss": 0.5266, "mean_token_accuracy": 0.8519076108932495, "num_tokens": 249739640.0, "step": 6576 }, { "epoch": 0.836662002289785, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.608175754547119, "learning_rate": 1e-05, "loss": 0.5516, "mean_token_accuracy": 0.8511779308319092, "num_tokens": 249778430.0, "step": 6577 }, { "epoch": 0.8367892125683756, "ewc_loss": 0.08984375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.287115573883057, "learning_rate": 1e-05, "loss": 0.4698, "mean_token_accuracy": 0.8702816963195801, "num_tokens": 249817875.0, "step": 6578 }, { "epoch": 0.836916422846966, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.86856746673584, "learning_rate": 1e-05, "loss": 0.4756, "mean_token_accuracy": 0.8802621960639954, "num_tokens": 249861987.0, "step": 6579 }, { "epoch": 0.8370436331255565, "ewc_loss": 0.095703125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.570872783660889, "learning_rate": 1e-05, "loss": 0.4989, "mean_token_accuracy": 0.863893985748291, "num_tokens": 249903137.0, "step": 6580 }, { "epoch": 0.837170843404147, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.842682838439941, "learning_rate": 1e-05, "loss": 0.5434, "mean_token_accuracy": 0.8533743023872375, "num_tokens": 249937760.0, "step": 6581 }, { "epoch": 0.8372980536827376, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 5.849097728729248, "learning_rate": 1e-05, "loss": 0.5211, "mean_token_accuracy": 0.855262041091919, "num_tokens": 249977280.0, "step": 6582 }, { "epoch": 0.8374252639613281, "ewc_loss": 0.109375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.565456390380859, "learning_rate": 1e-05, "loss": 0.5597, "mean_token_accuracy": 0.8526559472084045, "num_tokens": 250022160.0, "step": 6583 }, { "epoch": 0.8375524742399186, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 5.902111053466797, "learning_rate": 1e-05, "loss": 0.4993, "mean_token_accuracy": 0.8652104139328003, "num_tokens": 250063328.0, "step": 6584 }, { "epoch": 0.837679684518509, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.252465724945068, "learning_rate": 1e-05, "loss": 0.4801, "mean_token_accuracy": 0.869966447353363, "num_tokens": 250096994.0, "step": 6585 }, { "epoch": 0.8378068947970996, "ewc_loss": 0.0966796875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.843553066253662, "learning_rate": 1e-05, "loss": 0.4939, "mean_token_accuracy": 0.8652079105377197, "num_tokens": 250126058.0, "step": 6586 }, { "epoch": 0.8379341050756901, "ewc_loss": 0.09765625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 6.01148796081543, "learning_rate": 1e-05, "loss": 0.5574, "mean_token_accuracy": 0.846613883972168, "num_tokens": 250171965.0, "step": 6587 }, { "epoch": 0.8380613153542806, "ewc_loss": 0.095703125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.741569519042969, "learning_rate": 1e-05, "loss": 0.5305, "mean_token_accuracy": 0.8567870855331421, "num_tokens": 250209862.0, "step": 6588 }, { "epoch": 0.8381885256328712, "ewc_loss": 0.095703125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.898057460784912, "learning_rate": 1e-05, "loss": 0.5149, "mean_token_accuracy": 0.8601061105728149, "num_tokens": 250244895.0, "step": 6589 }, { "epoch": 0.8383157359114617, "ewc_loss": 0.091796875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.603844165802002, "learning_rate": 1e-05, "loss": 0.4951, "mean_token_accuracy": 0.8614165782928467, "num_tokens": 250283340.0, "step": 6590 }, { "epoch": 0.8384429461900521, "ewc_loss": 0.095703125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.894420623779297, "learning_rate": 1e-05, "loss": 0.501, "mean_token_accuracy": 0.8601911067962646, "num_tokens": 250318671.0, "step": 6591 }, { "epoch": 0.8385701564686426, "ewc_loss": 0.0888671875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 7.62939453125e-05, "grad_norm": 5.465956211090088, "learning_rate": 1e-05, "loss": 0.4776, "mean_token_accuracy": 0.8661264777183533, "num_tokens": 250357718.0, "step": 6592 }, { "epoch": 0.8386973667472332, "ewc_loss": 0.0986328125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 6.00102424621582, "learning_rate": 1e-05, "loss": 0.5147, "mean_token_accuracy": 0.8597956895828247, "num_tokens": 250391583.0, "step": 6593 }, { "epoch": 0.8388245770258237, "ewc_loss": 0.0869140625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 7.390975952148438e-05, "grad_norm": 5.347928047180176, "learning_rate": 1e-05, "loss": 0.5164, "mean_token_accuracy": 0.8595627546310425, "num_tokens": 250420938.0, "step": 6594 }, { "epoch": 0.8389517873044142, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.16957426071167, "learning_rate": 1e-05, "loss": 0.4893, "mean_token_accuracy": 0.8711575269699097, "num_tokens": 250455608.0, "step": 6595 }, { "epoch": 0.8390789975830047, "ewc_loss": 0.0859375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 7.295608520507812e-05, "grad_norm": 5.229084491729736, "learning_rate": 1e-05, "loss": 0.5215, "mean_token_accuracy": 0.8553093671798706, "num_tokens": 250489745.0, "step": 6596 }, { "epoch": 0.8392062078615952, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.43991231918335, "learning_rate": 1e-05, "loss": 0.5256, "mean_token_accuracy": 0.8597862720489502, "num_tokens": 250523711.0, "step": 6597 }, { "epoch": 0.8393334181401857, "ewc_loss": 0.0869140625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 7.43865966796875e-05, "grad_norm": 5.206128120422363, "learning_rate": 1e-05, "loss": 0.5526, "mean_token_accuracy": 0.8493571281433105, "num_tokens": 250562044.0, "step": 6598 }, { "epoch": 0.8394606284187762, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.746452331542969, "learning_rate": 1e-05, "loss": 0.5452, "mean_token_accuracy": 0.8567444086074829, "num_tokens": 250605930.0, "step": 6599 }, { "epoch": 0.8395878386973668, "ewc_loss": 0.09326171875, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.414294719696045, "learning_rate": 1e-05, "loss": 0.5045, "mean_token_accuracy": 0.8589590787887573, "num_tokens": 250640621.0, "step": 6600 }, { "epoch": 0.8397150489759573, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.849800109863281, "learning_rate": 1e-05, "loss": 0.5307, "mean_token_accuracy": 0.8615096807479858, "num_tokens": 250678056.0, "step": 6601 }, { "epoch": 0.8398422592545478, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 5.761152744293213, "learning_rate": 1e-05, "loss": 0.5074, "mean_token_accuracy": 0.8666445016860962, "num_tokens": 250719642.0, "step": 6602 }, { "epoch": 0.8399694695331382, "ewc_loss": 0.11181640625, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.590428829193115, "learning_rate": 1e-05, "loss": 0.5973, "mean_token_accuracy": 0.839309811592102, "num_tokens": 250757643.0, "step": 6603 }, { "epoch": 0.8400966798117288, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 5.951517581939697, "learning_rate": 1e-05, "loss": 0.554, "mean_token_accuracy": 0.8530066013336182, "num_tokens": 250795277.0, "step": 6604 }, { "epoch": 0.8402238900903193, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 6.1184163093566895, "learning_rate": 1e-05, "loss": 0.5246, "mean_token_accuracy": 0.8567458391189575, "num_tokens": 250834161.0, "step": 6605 }, { "epoch": 0.8403511003689098, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 6.008196830749512, "learning_rate": 1e-05, "loss": 0.5375, "mean_token_accuracy": 0.8551758527755737, "num_tokens": 250872518.0, "step": 6606 }, { "epoch": 0.8404783106475003, "ewc_loss": 0.091796875, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.663496971130371, "learning_rate": 1e-05, "loss": 0.4979, "mean_token_accuracy": 0.8652962446212769, "num_tokens": 250908516.0, "step": 6607 }, { "epoch": 0.8406055209260909, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 6.131227493286133, "learning_rate": 1e-05, "loss": 0.5201, "mean_token_accuracy": 0.8555779457092285, "num_tokens": 250949351.0, "step": 6608 }, { "epoch": 0.8407327312046813, "ewc_loss": 0.0859375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 7.295608520507812e-05, "grad_norm": 5.344950199127197, "learning_rate": 1e-05, "loss": 0.495, "mean_token_accuracy": 0.8598247170448303, "num_tokens": 250987145.0, "step": 6609 }, { "epoch": 0.8408599414832718, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.371540069580078, "learning_rate": 1e-05, "loss": 0.4991, "mean_token_accuracy": 0.8684791922569275, "num_tokens": 251024073.0, "step": 6610 }, { "epoch": 0.8409871517618623, "ewc_loss": 0.08642578125, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 7.343292236328125e-05, "grad_norm": 5.205428600311279, "learning_rate": 1e-05, "loss": 0.5301, "mean_token_accuracy": 0.8510817885398865, "num_tokens": 251069537.0, "step": 6611 }, { "epoch": 0.8411143620404529, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.688746929168701, "learning_rate": 1e-05, "loss": 0.553, "mean_token_accuracy": 0.8530058860778809, "num_tokens": 251103343.0, "step": 6612 }, { "epoch": 0.8412415723190434, "ewc_loss": 0.08984375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.32519006729126, "learning_rate": 1e-05, "loss": 0.5226, "mean_token_accuracy": 0.853793740272522, "num_tokens": 251140437.0, "step": 6613 }, { "epoch": 0.8413687825976339, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.904426097869873, "learning_rate": 1e-05, "loss": 0.5357, "mean_token_accuracy": 0.8610639572143555, "num_tokens": 251178002.0, "step": 6614 }, { "epoch": 0.8414959928762245, "ewc_loss": 0.09814453125, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.64932107925415, "learning_rate": 1e-05, "loss": 0.4783, "mean_token_accuracy": 0.8715543746948242, "num_tokens": 251219694.0, "step": 6615 }, { "epoch": 0.8416232031548149, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.801417827606201, "learning_rate": 1e-05, "loss": 0.5142, "mean_token_accuracy": 0.8648577332496643, "num_tokens": 251255576.0, "step": 6616 }, { "epoch": 0.8417504134334054, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 5.888260364532471, "learning_rate": 1e-05, "loss": 0.536, "mean_token_accuracy": 0.8550402522087097, "num_tokens": 251287245.0, "step": 6617 }, { "epoch": 0.8418776237119959, "ewc_loss": 0.10986328125, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.498744487762451, "learning_rate": 1e-05, "loss": 0.5497, "mean_token_accuracy": 0.8595011830329895, "num_tokens": 251320435.0, "step": 6618 }, { "epoch": 0.8420048339905865, "ewc_loss": 0.10205078125, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 5.896915435791016, "learning_rate": 1e-05, "loss": 0.4881, "mean_token_accuracy": 0.867536187171936, "num_tokens": 251358893.0, "step": 6619 }, { "epoch": 0.842132044269177, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 6.224948406219482, "learning_rate": 1e-05, "loss": 0.4798, "mean_token_accuracy": 0.8700816631317139, "num_tokens": 251392805.0, "step": 6620 }, { "epoch": 0.8422592545477675, "ewc_loss": 0.0986328125, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.806639194488525, "learning_rate": 1e-05, "loss": 0.4904, "mean_token_accuracy": 0.8682234287261963, "num_tokens": 251436848.0, "step": 6621 }, { "epoch": 0.8423864648263579, "ewc_loss": 0.0986328125, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.994121074676514, "learning_rate": 1e-05, "loss": 0.4739, "mean_token_accuracy": 0.8688820600509644, "num_tokens": 251471590.0, "step": 6622 }, { "epoch": 0.8425136751049485, "ewc_loss": 0.09521484375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.693371295928955, "learning_rate": 1e-05, "loss": 0.524, "mean_token_accuracy": 0.8543732166290283, "num_tokens": 251509251.0, "step": 6623 }, { "epoch": 0.842640885383539, "ewc_loss": 0.09765625, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.911186695098877, "learning_rate": 1e-05, "loss": 0.556, "mean_token_accuracy": 0.848049521446228, "num_tokens": 251549337.0, "step": 6624 }, { "epoch": 0.8427680956621295, "ewc_loss": 0.0927734375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.56213903427124, "learning_rate": 1e-05, "loss": 0.5805, "mean_token_accuracy": 0.838834822177887, "num_tokens": 251593953.0, "step": 6625 }, { "epoch": 0.84289530594072, "ewc_loss": 0.09912109375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.937613487243652, "learning_rate": 1e-05, "loss": 0.5298, "mean_token_accuracy": 0.8580046892166138, "num_tokens": 251631797.0, "step": 6626 }, { "epoch": 0.8430225162193106, "ewc_loss": 0.0888671875, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 7.581710815429688e-05, "grad_norm": 5.429625988006592, "learning_rate": 1e-05, "loss": 0.5686, "mean_token_accuracy": 0.8475809097290039, "num_tokens": 251668378.0, "step": 6627 }, { "epoch": 0.843149726497901, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 6.047263145446777, "learning_rate": 1e-05, "loss": 0.5524, "mean_token_accuracy": 0.8466595411300659, "num_tokens": 251712248.0, "step": 6628 }, { "epoch": 0.8432769367764915, "ewc_loss": 0.0869140625, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 7.390975952148438e-05, "grad_norm": 5.305819988250732, "learning_rate": 1e-05, "loss": 0.5139, "mean_token_accuracy": 0.8587570190429688, "num_tokens": 251748770.0, "step": 6629 }, { "epoch": 0.843404147055082, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.250882625579834, "learning_rate": 1e-05, "loss": 0.5006, "mean_token_accuracy": 0.8644059896469116, "num_tokens": 251787262.0, "step": 6630 }, { "epoch": 0.8435313573336726, "ewc_loss": 0.0869140625, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 7.390975952148438e-05, "grad_norm": 5.225096225738525, "learning_rate": 1e-05, "loss": 0.5438, "mean_token_accuracy": 0.8467391133308411, "num_tokens": 251820584.0, "step": 6631 }, { "epoch": 0.8436585676122631, "ewc_loss": 0.109375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.535166263580322, "learning_rate": 1e-05, "loss": 0.5533, "mean_token_accuracy": 0.8516486883163452, "num_tokens": 251861394.0, "step": 6632 }, { "epoch": 0.8437857778908536, "ewc_loss": 0.08984375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.271923542022705, "learning_rate": 1e-05, "loss": 0.5076, "mean_token_accuracy": 0.859717845916748, "num_tokens": 251902144.0, "step": 6633 }, { "epoch": 0.843912988169444, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.763489723205566, "learning_rate": 1e-05, "loss": 0.4516, "mean_token_accuracy": 0.8817486763000488, "num_tokens": 251936162.0, "step": 6634 }, { "epoch": 0.8440401984480346, "ewc_loss": 0.09619140625, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.536759376525879, "learning_rate": 1e-05, "loss": 0.5213, "mean_token_accuracy": 0.85428786277771, "num_tokens": 251977223.0, "step": 6635 }, { "epoch": 0.8441674087266251, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.722960948944092, "learning_rate": 1e-05, "loss": 0.52, "mean_token_accuracy": 0.8629791736602783, "num_tokens": 252013413.0, "step": 6636 }, { "epoch": 0.8442946190052156, "ewc_loss": 0.10205078125, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 5.821663856506348, "learning_rate": 1e-05, "loss": 0.5465, "mean_token_accuracy": 0.8508318662643433, "num_tokens": 252051395.0, "step": 6637 }, { "epoch": 0.8444218292838062, "ewc_loss": 0.10595703125, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.349559307098389, "learning_rate": 1e-05, "loss": 0.5455, "mean_token_accuracy": 0.8540048599243164, "num_tokens": 252092982.0, "step": 6638 }, { "epoch": 0.8445490395623967, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 5.947158336639404, "learning_rate": 1e-05, "loss": 0.4654, "mean_token_accuracy": 0.8788866996765137, "num_tokens": 252128953.0, "step": 6639 }, { "epoch": 0.8446762498409871, "ewc_loss": 0.09619140625, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.861903190612793, "learning_rate": 1e-05, "loss": 0.5447, "mean_token_accuracy": 0.8482984900474548, "num_tokens": 252171438.0, "step": 6640 }, { "epoch": 0.8448034601195776, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.072812557220459, "learning_rate": 1e-05, "loss": 0.5159, "mean_token_accuracy": 0.8584458231925964, "num_tokens": 252210310.0, "step": 6641 }, { "epoch": 0.8449306703981682, "ewc_loss": 0.087890625, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 7.486343383789062e-05, "grad_norm": 5.434518814086914, "learning_rate": 1e-05, "loss": 0.4608, "mean_token_accuracy": 0.8743222951889038, "num_tokens": 252242690.0, "step": 6642 }, { "epoch": 0.8450578806767587, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.310859203338623, "learning_rate": 1e-05, "loss": 0.5133, "mean_token_accuracy": 0.8606024980545044, "num_tokens": 252281570.0, "step": 6643 }, { "epoch": 0.8451850909553492, "ewc_loss": 0.0859375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 7.295608520507812e-05, "grad_norm": 5.180581092834473, "learning_rate": 1e-05, "loss": 0.5446, "mean_token_accuracy": 0.8466147184371948, "num_tokens": 252323478.0, "step": 6644 }, { "epoch": 0.8453123012339397, "ewc_loss": 0.11376953125, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.694214820861816, "learning_rate": 1e-05, "loss": 0.5787, "mean_token_accuracy": 0.8447133302688599, "num_tokens": 252362620.0, "step": 6645 }, { "epoch": 0.8454395115125302, "ewc_loss": 0.08935546875, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 7.62939453125e-05, "grad_norm": 5.244744777679443, "learning_rate": 1e-05, "loss": 0.4731, "mean_token_accuracy": 0.8684179782867432, "num_tokens": 252397549.0, "step": 6646 }, { "epoch": 0.8455667217911207, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 7.005671977996826, "learning_rate": 1e-05, "loss": 0.5546, "mean_token_accuracy": 0.8525840044021606, "num_tokens": 252433021.0, "step": 6647 }, { "epoch": 0.8456939320697112, "ewc_loss": 0.09814453125, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.644488334655762, "learning_rate": 1e-05, "loss": 0.5103, "mean_token_accuracy": 0.8596166968345642, "num_tokens": 252466455.0, "step": 6648 }, { "epoch": 0.8458211423483017, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.936634540557861, "learning_rate": 1e-05, "loss": 0.5531, "mean_token_accuracy": 0.854266345500946, "num_tokens": 252503547.0, "step": 6649 }, { "epoch": 0.8459483526268923, "ewc_loss": 0.10595703125, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 5.971184730529785, "learning_rate": 1e-05, "loss": 0.5335, "mean_token_accuracy": 0.8546332716941833, "num_tokens": 252535621.0, "step": 6650 }, { "epoch": 0.8460755629054828, "ewc_loss": 0.11181640625, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.584590911865234, "learning_rate": 1e-05, "loss": 0.5555, "mean_token_accuracy": 0.8532003164291382, "num_tokens": 252573699.0, "step": 6651 }, { "epoch": 0.8462027731840732, "ewc_loss": 0.10400390625, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 5.972205638885498, "learning_rate": 1e-05, "loss": 0.494, "mean_token_accuracy": 0.8673080205917358, "num_tokens": 252619816.0, "step": 6652 }, { "epoch": 0.8463299834626637, "ewc_loss": 0.10400390625, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.2517547607421875, "learning_rate": 1e-05, "loss": 0.4893, "mean_token_accuracy": 0.8700518012046814, "num_tokens": 252652631.0, "step": 6653 }, { "epoch": 0.8464571937412543, "ewc_loss": 0.099609375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.874503135681152, "learning_rate": 1e-05, "loss": 0.4971, "mean_token_accuracy": 0.8654425144195557, "num_tokens": 252690534.0, "step": 6654 }, { "epoch": 0.8465844040198448, "ewc_loss": 0.0986328125, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 6.038035869598389, "learning_rate": 1e-05, "loss": 0.5512, "mean_token_accuracy": 0.849366307258606, "num_tokens": 252723090.0, "step": 6655 }, { "epoch": 0.8467116142984353, "ewc_loss": 0.09619140625, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 6.729493141174316, "learning_rate": 1e-05, "loss": 0.4856, "mean_token_accuracy": 0.878350019454956, "num_tokens": 252755879.0, "step": 6656 }, { "epoch": 0.8468388245770259, "ewc_loss": 0.091796875, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.539732933044434, "learning_rate": 1e-05, "loss": 0.5629, "mean_token_accuracy": 0.8420246839523315, "num_tokens": 252799137.0, "step": 6657 }, { "epoch": 0.8469660348556163, "ewc_loss": 0.10498046875, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.161077499389648, "learning_rate": 1e-05, "loss": 0.5314, "mean_token_accuracy": 0.8572626709938049, "num_tokens": 252840871.0, "step": 6658 }, { "epoch": 0.8470932451342068, "ewc_loss": 0.0888671875, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 7.581710815429688e-05, "grad_norm": 5.312886714935303, "learning_rate": 1e-05, "loss": 0.5276, "mean_token_accuracy": 0.8540071249008179, "num_tokens": 252879708.0, "step": 6659 }, { "epoch": 0.8472204554127973, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.2917962074279785, "learning_rate": 1e-05, "loss": 0.5102, "mean_token_accuracy": 0.8637945652008057, "num_tokens": 252914223.0, "step": 6660 }, { "epoch": 0.8473476656913879, "ewc_loss": 0.091796875, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.285735607147217, "learning_rate": 1e-05, "loss": 0.5107, "mean_token_accuracy": 0.8609557151794434, "num_tokens": 252956666.0, "step": 6661 }, { "epoch": 0.8474748759699784, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.428316593170166, "learning_rate": 1e-05, "loss": 0.5212, "mean_token_accuracy": 0.859630823135376, "num_tokens": 252997566.0, "step": 6662 }, { "epoch": 0.8476020862485689, "ewc_loss": 0.09423828125, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.315911769866943, "learning_rate": 1e-05, "loss": 0.4893, "mean_token_accuracy": 0.8662623763084412, "num_tokens": 253038545.0, "step": 6663 }, { "epoch": 0.8477292965271594, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.6129679679870605, "learning_rate": 1e-05, "loss": 0.5346, "mean_token_accuracy": 0.862114667892456, "num_tokens": 253078123.0, "step": 6664 }, { "epoch": 0.8478565068057499, "ewc_loss": 0.0986328125, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.411779880523682, "learning_rate": 1e-05, "loss": 0.4779, "mean_token_accuracy": 0.8728623986244202, "num_tokens": 253118273.0, "step": 6665 }, { "epoch": 0.8479837170843404, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.732525825500488, "learning_rate": 1e-05, "loss": 0.5766, "mean_token_accuracy": 0.8489776849746704, "num_tokens": 253152994.0, "step": 6666 }, { "epoch": 0.8481109273629309, "ewc_loss": 0.10107421875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 5.594837665557861, "learning_rate": 1e-05, "loss": 0.489, "mean_token_accuracy": 0.8730209469795227, "num_tokens": 253189736.0, "step": 6667 }, { "epoch": 0.8482381376415215, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.6825737953186035, "learning_rate": 1e-05, "loss": 0.5714, "mean_token_accuracy": 0.8490029573440552, "num_tokens": 253225872.0, "step": 6668 }, { "epoch": 0.848365347920112, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 5.749343395233154, "learning_rate": 1e-05, "loss": 0.478, "mean_token_accuracy": 0.8737425804138184, "num_tokens": 253263902.0, "step": 6669 }, { "epoch": 0.8484925581987025, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.481137752532959, "learning_rate": 1e-05, "loss": 0.5007, "mean_token_accuracy": 0.8702535629272461, "num_tokens": 253297536.0, "step": 6670 }, { "epoch": 0.8486197684772929, "ewc_loss": 0.10498046875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 5.821815013885498, "learning_rate": 1e-05, "loss": 0.4812, "mean_token_accuracy": 0.8713470697402954, "num_tokens": 253334624.0, "step": 6671 }, { "epoch": 0.8487469787558835, "ewc_loss": 0.10595703125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.186744689941406, "learning_rate": 1e-05, "loss": 0.5482, "mean_token_accuracy": 0.8546377420425415, "num_tokens": 253379203.0, "step": 6672 }, { "epoch": 0.848874189034474, "ewc_loss": 0.10302734375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 5.817405700683594, "learning_rate": 1e-05, "loss": 0.4889, "mean_token_accuracy": 0.870022714138031, "num_tokens": 253417209.0, "step": 6673 }, { "epoch": 0.8490013993130645, "ewc_loss": 0.10009765625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.939712047576904, "learning_rate": 1e-05, "loss": 0.5462, "mean_token_accuracy": 0.8480451703071594, "num_tokens": 253453909.0, "step": 6674 }, { "epoch": 0.849128609591655, "ewc_loss": 0.10009765625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.789756774902344, "learning_rate": 1e-05, "loss": 0.6332, "mean_token_accuracy": 0.8274657130241394, "num_tokens": 253489448.0, "step": 6675 }, { "epoch": 0.8492558198702456, "ewc_loss": 0.09521484375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.745406150817871, "learning_rate": 1e-05, "loss": 0.5003, "mean_token_accuracy": 0.8579809069633484, "num_tokens": 253517544.0, "step": 6676 }, { "epoch": 0.849383030148836, "ewc_loss": 0.09765625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.73626184463501, "learning_rate": 1e-05, "loss": 0.506, "mean_token_accuracy": 0.8665584325790405, "num_tokens": 253558829.0, "step": 6677 }, { "epoch": 0.8495102404274265, "ewc_loss": 0.0927734375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.6356706619262695, "learning_rate": 1e-05, "loss": 0.514, "mean_token_accuracy": 0.863095760345459, "num_tokens": 253593944.0, "step": 6678 }, { "epoch": 0.849637450706017, "ewc_loss": 0.095703125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.700973033905029, "learning_rate": 1e-05, "loss": 0.5243, "mean_token_accuracy": 0.8553861379623413, "num_tokens": 253633579.0, "step": 6679 }, { "epoch": 0.8497646609846076, "ewc_loss": 0.09228515625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.56557035446167, "learning_rate": 1e-05, "loss": 0.5297, "mean_token_accuracy": 0.8566782474517822, "num_tokens": 253680247.0, "step": 6680 }, { "epoch": 0.8498918712631981, "ewc_loss": 0.0947265625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.701944828033447, "learning_rate": 1e-05, "loss": 0.5064, "mean_token_accuracy": 0.8620918989181519, "num_tokens": 253714766.0, "step": 6681 }, { "epoch": 0.8500190815417886, "ewc_loss": 0.091796875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.514826774597168, "learning_rate": 1e-05, "loss": 0.5293, "mean_token_accuracy": 0.8517778515815735, "num_tokens": 253747643.0, "step": 6682 }, { "epoch": 0.850146291820379, "ewc_loss": 0.09375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.701588153839111, "learning_rate": 1e-05, "loss": 0.562, "mean_token_accuracy": 0.8401980996131897, "num_tokens": 253791252.0, "step": 6683 }, { "epoch": 0.8502735020989696, "ewc_loss": 0.08984375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.459928512573242, "learning_rate": 1e-05, "loss": 0.4871, "mean_token_accuracy": 0.8663917779922485, "num_tokens": 253822244.0, "step": 6684 }, { "epoch": 0.8504007123775601, "ewc_loss": 0.0947265625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.751647472381592, "learning_rate": 1e-05, "loss": 0.5312, "mean_token_accuracy": 0.8540072441101074, "num_tokens": 253858680.0, "step": 6685 }, { "epoch": 0.8505279226561506, "ewc_loss": 0.0888671875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 7.581710815429688e-05, "grad_norm": 5.381050109863281, "learning_rate": 1e-05, "loss": 0.4712, "mean_token_accuracy": 0.8706288933753967, "num_tokens": 253896266.0, "step": 6686 }, { "epoch": 0.8506551329347412, "ewc_loss": 0.095703125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.864358901977539, "learning_rate": 1e-05, "loss": 0.5085, "mean_token_accuracy": 0.8617540597915649, "num_tokens": 253934059.0, "step": 6687 }, { "epoch": 0.8507823432133317, "ewc_loss": 0.0869140625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 7.390975952148438e-05, "grad_norm": 5.246985912322998, "learning_rate": 1e-05, "loss": 0.4917, "mean_token_accuracy": 0.8626616597175598, "num_tokens": 253967862.0, "step": 6688 }, { "epoch": 0.8509095534919221, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 6.089559555053711, "learning_rate": 1e-05, "loss": 0.5362, "mean_token_accuracy": 0.8529249429702759, "num_tokens": 254008277.0, "step": 6689 }, { "epoch": 0.8510367637705126, "ewc_loss": 0.08447265625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 7.104873657226562e-05, "grad_norm": 5.060457229614258, "learning_rate": 1e-05, "loss": 0.495, "mean_token_accuracy": 0.8624717593193054, "num_tokens": 254039939.0, "step": 6690 }, { "epoch": 0.8511639740491032, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.525528430938721, "learning_rate": 1e-05, "loss": 0.5157, "mean_token_accuracy": 0.8654285669326782, "num_tokens": 254080497.0, "step": 6691 }, { "epoch": 0.8512911843276937, "ewc_loss": 0.0849609375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 7.200241088867188e-05, "grad_norm": 4.940487384796143, "learning_rate": 1e-05, "loss": 0.4923, "mean_token_accuracy": 0.862245500087738, "num_tokens": 254111917.0, "step": 6692 }, { "epoch": 0.8514183946062842, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 0.00011539459228515625, "grad_norm": 7.196541786193848, "learning_rate": 1e-05, "loss": 0.5156, "mean_token_accuracy": 0.8688410520553589, "num_tokens": 254150245.0, "step": 6693 }, { "epoch": 0.8515456048848747, "ewc_loss": 0.09716796875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.3105316162109375, "learning_rate": 1e-05, "loss": 0.5216, "mean_token_accuracy": 0.8598574995994568, "num_tokens": 254187106.0, "step": 6694 }, { "epoch": 0.8516728151634652, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 0.00012969970703125, "grad_norm": 7.677978992462158, "learning_rate": 1e-05, "loss": 0.5776, "mean_token_accuracy": 0.8512186408042908, "num_tokens": 254219877.0, "step": 6695 }, { "epoch": 0.8518000254420557, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 6.2668657302856445, "learning_rate": 1e-05, "loss": 0.5539, "mean_token_accuracy": 0.8521997332572937, "num_tokens": 254264234.0, "step": 6696 }, { "epoch": 0.8519272357206462, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 0.00011491775512695312, "grad_norm": 7.246899604797363, "learning_rate": 1e-05, "loss": 0.5434, "mean_token_accuracy": 0.8667325377464294, "num_tokens": 254299762.0, "step": 6697 }, { "epoch": 0.8520544459992367, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 6.514064788818359, "learning_rate": 1e-05, "loss": 0.5728, "mean_token_accuracy": 0.8461195826530457, "num_tokens": 254336403.0, "step": 6698 }, { "epoch": 0.8521816562778273, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.72898006439209, "learning_rate": 1e-05, "loss": 0.5172, "mean_token_accuracy": 0.8655202984809875, "num_tokens": 254375971.0, "step": 6699 }, { "epoch": 0.8523088665564178, "ewc_loss": 0.11181640625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.252832889556885, "learning_rate": 1e-05, "loss": 0.547, "mean_token_accuracy": 0.8511694073677063, "num_tokens": 254418430.0, "step": 6700 }, { "epoch": 0.8524360768350082, "ewc_loss": 0.10693359375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.412665367126465, "learning_rate": 1e-05, "loss": 0.523, "mean_token_accuracy": 0.8603925704956055, "num_tokens": 254457275.0, "step": 6701 }, { "epoch": 0.8525632871135987, "ewc_loss": 0.10498046875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 5.998543739318848, "learning_rate": 1e-05, "loss": 0.5353, "mean_token_accuracy": 0.8549972176551819, "num_tokens": 254499280.0, "step": 6702 }, { "epoch": 0.8526904973921893, "ewc_loss": 0.10302734375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.225109100341797, "learning_rate": 1e-05, "loss": 0.5347, "mean_token_accuracy": 0.854872465133667, "num_tokens": 254536571.0, "step": 6703 }, { "epoch": 0.8528177076707798, "ewc_loss": 0.095703125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.771549701690674, "learning_rate": 1e-05, "loss": 0.5338, "mean_token_accuracy": 0.8498904705047607, "num_tokens": 254576900.0, "step": 6704 }, { "epoch": 0.8529449179493703, "ewc_loss": 0.10205078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.162998676300049, "learning_rate": 1e-05, "loss": 0.5932, "mean_token_accuracy": 0.8373165726661682, "num_tokens": 254617791.0, "step": 6705 }, { "epoch": 0.8530721282279609, "ewc_loss": 0.0927734375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.571633815765381, "learning_rate": 1e-05, "loss": 0.4872, "mean_token_accuracy": 0.8680872917175293, "num_tokens": 254658300.0, "step": 6706 }, { "epoch": 0.8531993385065513, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.230339527130127, "learning_rate": 1e-05, "loss": 0.5354, "mean_token_accuracy": 0.8583077192306519, "num_tokens": 254698133.0, "step": 6707 }, { "epoch": 0.8533265487851418, "ewc_loss": 0.087890625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 7.486343383789062e-05, "grad_norm": 5.40242338180542, "learning_rate": 1e-05, "loss": 0.4835, "mean_token_accuracy": 0.8642572164535522, "num_tokens": 254736099.0, "step": 6708 }, { "epoch": 0.8534537590637323, "ewc_loss": 0.10693359375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.453230857849121, "learning_rate": 1e-05, "loss": 0.4906, "mean_token_accuracy": 0.8718482255935669, "num_tokens": 254773460.0, "step": 6709 }, { "epoch": 0.8535809693423229, "ewc_loss": 0.08837890625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 7.534027099609375e-05, "grad_norm": 5.297478675842285, "learning_rate": 1e-05, "loss": 0.4796, "mean_token_accuracy": 0.8658385276794434, "num_tokens": 254812722.0, "step": 6710 }, { "epoch": 0.8537081796209134, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.8105058670043945, "learning_rate": 1e-05, "loss": 0.5272, "mean_token_accuracy": 0.8649144768714905, "num_tokens": 254851994.0, "step": 6711 }, { "epoch": 0.8538353898995039, "ewc_loss": 0.09228515625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.470749855041504, "learning_rate": 1e-05, "loss": 0.5393, "mean_token_accuracy": 0.852429986000061, "num_tokens": 254889111.0, "step": 6712 }, { "epoch": 0.8539626001780944, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.990163803100586, "learning_rate": 1e-05, "loss": 0.5552, "mean_token_accuracy": 0.8568893671035767, "num_tokens": 254931779.0, "step": 6713 }, { "epoch": 0.8540898104566849, "ewc_loss": 0.10107421875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 5.864999294281006, "learning_rate": 1e-05, "loss": 0.5101, "mean_token_accuracy": 0.8634175062179565, "num_tokens": 254968709.0, "step": 6714 }, { "epoch": 0.8542170207352754, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.731532573699951, "learning_rate": 1e-05, "loss": 0.5186, "mean_token_accuracy": 0.8638862371444702, "num_tokens": 255000235.0, "step": 6715 }, { "epoch": 0.8543442310138659, "ewc_loss": 0.10595703125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.1076459884643555, "learning_rate": 1e-05, "loss": 0.529, "mean_token_accuracy": 0.8550964593887329, "num_tokens": 255041157.0, "step": 6716 }, { "epoch": 0.8544714412924564, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.211602210998535, "learning_rate": 1e-05, "loss": 0.5226, "mean_token_accuracy": 0.8573750257492065, "num_tokens": 255077001.0, "step": 6717 }, { "epoch": 0.854598651571047, "ewc_loss": 0.10302734375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.153957366943359, "learning_rate": 1e-05, "loss": 0.4844, "mean_token_accuracy": 0.8672553300857544, "num_tokens": 255113581.0, "step": 6718 }, { "epoch": 0.8547258618496375, "ewc_loss": 0.0908203125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.681117057800293, "learning_rate": 1e-05, "loss": 0.4605, "mean_token_accuracy": 0.8739352822303772, "num_tokens": 255153183.0, "step": 6719 }, { "epoch": 0.8548530721282279, "ewc_loss": 0.10498046875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.28010368347168, "learning_rate": 1e-05, "loss": 0.4933, "mean_token_accuracy": 0.8678155541419983, "num_tokens": 255192216.0, "step": 6720 }, { "epoch": 0.8549802824068184, "ewc_loss": 0.0859375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 7.295608520507812e-05, "grad_norm": 5.299534320831299, "learning_rate": 1e-05, "loss": 0.4676, "mean_token_accuracy": 0.8665107488632202, "num_tokens": 255229969.0, "step": 6721 }, { "epoch": 0.855107492685409, "ewc_loss": 0.10986328125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.603647232055664, "learning_rate": 1e-05, "loss": 0.5454, "mean_token_accuracy": 0.8501393795013428, "num_tokens": 255267232.0, "step": 6722 }, { "epoch": 0.8552347029639995, "ewc_loss": 0.08642578125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 7.343292236328125e-05, "grad_norm": 5.19406795501709, "learning_rate": 1e-05, "loss": 0.4874, "mean_token_accuracy": 0.8629685640335083, "num_tokens": 255304923.0, "step": 6723 }, { "epoch": 0.85536191324259, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 7.014481067657471, "learning_rate": 1e-05, "loss": 0.5931, "mean_token_accuracy": 0.8408955335617065, "num_tokens": 255342422.0, "step": 6724 }, { "epoch": 0.8554891235211806, "ewc_loss": 0.09423828125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.51223087310791, "learning_rate": 1e-05, "loss": 0.4911, "mean_token_accuracy": 0.8688787221908569, "num_tokens": 255380218.0, "step": 6725 }, { "epoch": 0.855616333799771, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 0.00011014938354492188, "grad_norm": 7.155551433563232, "learning_rate": 1e-05, "loss": 0.5374, "mean_token_accuracy": 0.8630152940750122, "num_tokens": 255413964.0, "step": 6726 }, { "epoch": 0.8557435440783615, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.015295505523682, "learning_rate": 1e-05, "loss": 0.4946, "mean_token_accuracy": 0.8674888610839844, "num_tokens": 255450421.0, "step": 6727 }, { "epoch": 0.855870754356952, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.813270092010498, "learning_rate": 1e-05, "loss": 0.5428, "mean_token_accuracy": 0.8603917360305786, "num_tokens": 255487239.0, "step": 6728 }, { "epoch": 0.8559979646355426, "ewc_loss": 0.10791015625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.147725582122803, "learning_rate": 1e-05, "loss": 0.532, "mean_token_accuracy": 0.8595339059829712, "num_tokens": 255529363.0, "step": 6729 }, { "epoch": 0.8561251749141331, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.393832206726074, "learning_rate": 1e-05, "loss": 0.433, "mean_token_accuracy": 0.8846656084060669, "num_tokens": 255570790.0, "step": 6730 }, { "epoch": 0.8562523851927236, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 6.023478031158447, "learning_rate": 1e-05, "loss": 0.5392, "mean_token_accuracy": 0.8567208051681519, "num_tokens": 255608188.0, "step": 6731 }, { "epoch": 0.856379595471314, "ewc_loss": 0.09912109375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 6.114158630371094, "learning_rate": 1e-05, "loss": 0.5155, "mean_token_accuracy": 0.8610604405403137, "num_tokens": 255640929.0, "step": 6732 }, { "epoch": 0.8565068057499046, "ewc_loss": 0.09765625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.855621337890625, "learning_rate": 1e-05, "loss": 0.4997, "mean_token_accuracy": 0.8646209836006165, "num_tokens": 255684579.0, "step": 6733 }, { "epoch": 0.8566340160284951, "ewc_loss": 0.0966796875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.984113693237305, "learning_rate": 1e-05, "loss": 0.5212, "mean_token_accuracy": 0.8605483770370483, "num_tokens": 255721195.0, "step": 6734 }, { "epoch": 0.8567612263070856, "ewc_loss": 0.0927734375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.674684047698975, "learning_rate": 1e-05, "loss": 0.5156, "mean_token_accuracy": 0.8594483137130737, "num_tokens": 255761697.0, "step": 6735 }, { "epoch": 0.8568884365856761, "ewc_loss": 0.09814453125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.980952262878418, "learning_rate": 1e-05, "loss": 0.5151, "mean_token_accuracy": 0.8585083484649658, "num_tokens": 255800492.0, "step": 6736 }, { "epoch": 0.8570156468642667, "ewc_loss": 0.0888671875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 7.581710815429688e-05, "grad_norm": 5.495915412902832, "learning_rate": 1e-05, "loss": 0.4823, "mean_token_accuracy": 0.862387478351593, "num_tokens": 255837074.0, "step": 6737 }, { "epoch": 0.8571428571428571, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 6.096578121185303, "learning_rate": 1e-05, "loss": 0.5038, "mean_token_accuracy": 0.8610299825668335, "num_tokens": 255876238.0, "step": 6738 }, { "epoch": 0.8572700674214476, "ewc_loss": 0.08642578125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 7.343292236328125e-05, "grad_norm": 5.323945045471191, "learning_rate": 1e-05, "loss": 0.5462, "mean_token_accuracy": 0.8464920520782471, "num_tokens": 255914622.0, "step": 6739 }, { "epoch": 0.8573972777000382, "ewc_loss": 0.10498046875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.339592456817627, "learning_rate": 1e-05, "loss": 0.5169, "mean_token_accuracy": 0.8627499341964722, "num_tokens": 255954027.0, "step": 6740 }, { "epoch": 0.8575244879786287, "ewc_loss": 0.08642578125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 7.343292236328125e-05, "grad_norm": 5.226711750030518, "learning_rate": 1e-05, "loss": 0.5266, "mean_token_accuracy": 0.8530189394950867, "num_tokens": 255997416.0, "step": 6741 }, { "epoch": 0.8576516982572192, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.681433200836182, "learning_rate": 1e-05, "loss": 0.5243, "mean_token_accuracy": 0.8583368062973022, "num_tokens": 256030447.0, "step": 6742 }, { "epoch": 0.8577789085358097, "ewc_loss": 0.08984375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.33981990814209, "learning_rate": 1e-05, "loss": 0.4848, "mean_token_accuracy": 0.8645961284637451, "num_tokens": 256068533.0, "step": 6743 }, { "epoch": 0.8579061188144002, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.913838863372803, "learning_rate": 1e-05, "loss": 0.4838, "mean_token_accuracy": 0.8740029335021973, "num_tokens": 256104783.0, "step": 6744 }, { "epoch": 0.8580333290929907, "ewc_loss": 0.09716796875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.701233863830566, "learning_rate": 1e-05, "loss": 0.5043, "mean_token_accuracy": 0.8647741079330444, "num_tokens": 256138272.0, "step": 6745 }, { "epoch": 0.8581605393715812, "ewc_loss": 0.11474609375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.777745246887207, "learning_rate": 1e-05, "loss": 0.5544, "mean_token_accuracy": 0.8525339365005493, "num_tokens": 256171102.0, "step": 6746 }, { "epoch": 0.8582877496501717, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 5.9854817390441895, "learning_rate": 1e-05, "loss": 0.5105, "mean_token_accuracy": 0.8617609143257141, "num_tokens": 256210593.0, "step": 6747 }, { "epoch": 0.8584149599287623, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.325632095336914, "learning_rate": 1e-05, "loss": 0.4912, "mean_token_accuracy": 0.8697491884231567, "num_tokens": 256248693.0, "step": 6748 }, { "epoch": 0.8585421702073528, "ewc_loss": 0.10302734375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.061008930206299, "learning_rate": 1e-05, "loss": 0.4857, "mean_token_accuracy": 0.8664191365242004, "num_tokens": 256284875.0, "step": 6749 }, { "epoch": 0.8586693804859432, "ewc_loss": 0.0927734375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.832438945770264, "learning_rate": 1e-05, "loss": 0.486, "mean_token_accuracy": 0.8646502494812012, "num_tokens": 256319121.0, "step": 6750 }, { "epoch": 0.8587965907645337, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 6.147367000579834, "learning_rate": 1e-05, "loss": 0.4814, "mean_token_accuracy": 0.8710020780563354, "num_tokens": 256354112.0, "step": 6751 }, { "epoch": 0.8589238010431243, "ewc_loss": 0.0869140625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 7.390975952148438e-05, "grad_norm": 5.426252365112305, "learning_rate": 1e-05, "loss": 0.4623, "mean_token_accuracy": 0.8738394379615784, "num_tokens": 256387612.0, "step": 6752 }, { "epoch": 0.8590510113217148, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.373835563659668, "learning_rate": 1e-05, "loss": 0.5222, "mean_token_accuracy": 0.859387993812561, "num_tokens": 256419720.0, "step": 6753 }, { "epoch": 0.8591782216003053, "ewc_loss": 0.08544921875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 7.2479248046875e-05, "grad_norm": 5.217949390411377, "learning_rate": 1e-05, "loss": 0.5196, "mean_token_accuracy": 0.85190349817276, "num_tokens": 256456717.0, "step": 6754 }, { "epoch": 0.8593054318788959, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.718896389007568, "learning_rate": 1e-05, "loss": 0.5843, "mean_token_accuracy": 0.8412089347839355, "num_tokens": 256496480.0, "step": 6755 }, { "epoch": 0.8594326421574863, "ewc_loss": 0.08984375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.305310249328613, "learning_rate": 1e-05, "loss": 0.4854, "mean_token_accuracy": 0.8644695281982422, "num_tokens": 256531974.0, "step": 6756 }, { "epoch": 0.8595598524360768, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 6.9919114112854, "learning_rate": 1e-05, "loss": 0.5548, "mean_token_accuracy": 0.854954719543457, "num_tokens": 256565983.0, "step": 6757 }, { "epoch": 0.8596870627146673, "ewc_loss": 0.09765625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.672711372375488, "learning_rate": 1e-05, "loss": 0.483, "mean_token_accuracy": 0.8683104515075684, "num_tokens": 256607215.0, "step": 6758 }, { "epoch": 0.8598142729932579, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.905075550079346, "learning_rate": 1e-05, "loss": 0.4759, "mean_token_accuracy": 0.8737420439720154, "num_tokens": 256646213.0, "step": 6759 }, { "epoch": 0.8599414832718484, "ewc_loss": 0.10498046875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 5.9636149406433105, "learning_rate": 1e-05, "loss": 0.4916, "mean_token_accuracy": 0.8678581118583679, "num_tokens": 256687638.0, "step": 6760 }, { "epoch": 0.8600686935504389, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.574068546295166, "learning_rate": 1e-05, "loss": 0.5297, "mean_token_accuracy": 0.8573579788208008, "num_tokens": 256716852.0, "step": 6761 }, { "epoch": 0.8601959038290294, "ewc_loss": 0.10302734375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 5.993221282958984, "learning_rate": 1e-05, "loss": 0.5032, "mean_token_accuracy": 0.8601672649383545, "num_tokens": 256748260.0, "step": 6762 }, { "epoch": 0.8603231141076199, "ewc_loss": 0.10205078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.234638214111328, "learning_rate": 1e-05, "loss": 0.5804, "mean_token_accuracy": 0.8419474959373474, "num_tokens": 256786400.0, "step": 6763 }, { "epoch": 0.8604503243862104, "ewc_loss": 0.0986328125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.883595943450928, "learning_rate": 1e-05, "loss": 0.4733, "mean_token_accuracy": 0.8716965317726135, "num_tokens": 256820186.0, "step": 6764 }, { "epoch": 0.8605775346648009, "ewc_loss": 0.09814453125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 6.012275695800781, "learning_rate": 1e-05, "loss": 0.5104, "mean_token_accuracy": 0.8629541993141174, "num_tokens": 256855735.0, "step": 6765 }, { "epoch": 0.8607047449433914, "ewc_loss": 0.0947265625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.7507476806640625, "learning_rate": 1e-05, "loss": 0.4717, "mean_token_accuracy": 0.8704006671905518, "num_tokens": 256887809.0, "step": 6766 }, { "epoch": 0.860831955221982, "ewc_loss": 0.09765625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.921492099761963, "learning_rate": 1e-05, "loss": 0.4936, "mean_token_accuracy": 0.8640743494033813, "num_tokens": 256921801.0, "step": 6767 }, { "epoch": 0.8609591655005725, "ewc_loss": 0.091796875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.598287582397461, "learning_rate": 1e-05, "loss": 0.4568, "mean_token_accuracy": 0.8751755952835083, "num_tokens": 256960399.0, "step": 6768 }, { "epoch": 0.8610863757791629, "ewc_loss": 0.09814453125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.950347900390625, "learning_rate": 1e-05, "loss": 0.4905, "mean_token_accuracy": 0.8633322715759277, "num_tokens": 256995094.0, "step": 6769 }, { "epoch": 0.8612135860577534, "ewc_loss": 0.08837890625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 7.534027099609375e-05, "grad_norm": 5.445623874664307, "learning_rate": 1e-05, "loss": 0.5805, "mean_token_accuracy": 0.8400473594665527, "num_tokens": 257037792.0, "step": 6770 }, { "epoch": 0.861340796336344, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 6.078700542449951, "learning_rate": 1e-05, "loss": 0.4948, "mean_token_accuracy": 0.8650150895118713, "num_tokens": 257075376.0, "step": 6771 }, { "epoch": 0.8614680066149345, "ewc_loss": 0.08642578125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 7.343292236328125e-05, "grad_norm": 5.299831867218018, "learning_rate": 1e-05, "loss": 0.5305, "mean_token_accuracy": 0.8539532423019409, "num_tokens": 257115878.0, "step": 6772 }, { "epoch": 0.861595216893525, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.326645851135254, "learning_rate": 1e-05, "loss": 0.5162, "mean_token_accuracy": 0.8601484298706055, "num_tokens": 257150780.0, "step": 6773 }, { "epoch": 0.8617224271721156, "ewc_loss": 0.08642578125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 7.343292236328125e-05, "grad_norm": 5.209169387817383, "learning_rate": 1e-05, "loss": 0.4904, "mean_token_accuracy": 0.8633847236633301, "num_tokens": 257194704.0, "step": 6774 }, { "epoch": 0.861849637450706, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.6699371337890625, "learning_rate": 1e-05, "loss": 0.5672, "mean_token_accuracy": 0.8487913608551025, "num_tokens": 257235337.0, "step": 6775 }, { "epoch": 0.8619768477292965, "ewc_loss": 0.08984375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.334020614624023, "learning_rate": 1e-05, "loss": 0.5351, "mean_token_accuracy": 0.8484063744544983, "num_tokens": 257268976.0, "step": 6776 }, { "epoch": 0.862104058007887, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.893993377685547, "learning_rate": 1e-05, "loss": 0.5152, "mean_token_accuracy": 0.8626554012298584, "num_tokens": 257304638.0, "step": 6777 }, { "epoch": 0.8622312682864776, "ewc_loss": 0.0966796875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.6724934577941895, "learning_rate": 1e-05, "loss": 0.5325, "mean_token_accuracy": 0.8525814414024353, "num_tokens": 257341773.0, "step": 6778 }, { "epoch": 0.8623584785650681, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.750871658325195, "learning_rate": 1e-05, "loss": 0.5203, "mean_token_accuracy": 0.8643462657928467, "num_tokens": 257375145.0, "step": 6779 }, { "epoch": 0.8624856888436586, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 5.942568778991699, "learning_rate": 1e-05, "loss": 0.4931, "mean_token_accuracy": 0.8673044443130493, "num_tokens": 257411070.0, "step": 6780 }, { "epoch": 0.862612899122249, "ewc_loss": 0.10498046875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.321823596954346, "learning_rate": 1e-05, "loss": 0.5403, "mean_token_accuracy": 0.8507658243179321, "num_tokens": 257447283.0, "step": 6781 }, { "epoch": 0.8627401094008396, "ewc_loss": 0.10400390625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.031876564025879, "learning_rate": 1e-05, "loss": 0.5078, "mean_token_accuracy": 0.861553430557251, "num_tokens": 257487370.0, "step": 6782 }, { "epoch": 0.8628673196794301, "ewc_loss": 0.0947265625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.8321919441223145, "learning_rate": 1e-05, "loss": 0.5472, "mean_token_accuracy": 0.8478503227233887, "num_tokens": 257525028.0, "step": 6783 }, { "epoch": 0.8629945299580206, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.126985549926758, "learning_rate": 1e-05, "loss": 0.5274, "mean_token_accuracy": 0.8580276370048523, "num_tokens": 257563884.0, "step": 6784 }, { "epoch": 0.8631217402366111, "ewc_loss": 0.08740234375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 7.43865966796875e-05, "grad_norm": 5.4327850341796875, "learning_rate": 1e-05, "loss": 0.4852, "mean_token_accuracy": 0.8656705617904663, "num_tokens": 257599431.0, "step": 6785 }, { "epoch": 0.8632489505152017, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.34830379486084, "learning_rate": 1e-05, "loss": 0.515, "mean_token_accuracy": 0.8599696159362793, "num_tokens": 257636060.0, "step": 6786 }, { "epoch": 0.8633761607937921, "ewc_loss": 0.08642578125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 7.343292236328125e-05, "grad_norm": 5.227059364318848, "learning_rate": 1e-05, "loss": 0.4701, "mean_token_accuracy": 0.8688932657241821, "num_tokens": 257672374.0, "step": 6787 }, { "epoch": 0.8635033710723826, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.685349941253662, "learning_rate": 1e-05, "loss": 0.5491, "mean_token_accuracy": 0.8539287447929382, "num_tokens": 257710875.0, "step": 6788 }, { "epoch": 0.8636305813509731, "ewc_loss": 0.08984375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.290166854858398, "learning_rate": 1e-05, "loss": 0.5475, "mean_token_accuracy": 0.8493664860725403, "num_tokens": 257756697.0, "step": 6789 }, { "epoch": 0.8637577916295637, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.964133262634277, "learning_rate": 1e-05, "loss": 0.5404, "mean_token_accuracy": 0.8548458814620972, "num_tokens": 257795281.0, "step": 6790 }, { "epoch": 0.8638850019081542, "ewc_loss": 0.09716796875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.651928424835205, "learning_rate": 1e-05, "loss": 0.4774, "mean_token_accuracy": 0.8728299140930176, "num_tokens": 257835743.0, "step": 6791 }, { "epoch": 0.8640122121867447, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.883368492126465, "learning_rate": 1e-05, "loss": 0.4884, "mean_token_accuracy": 0.8739811182022095, "num_tokens": 257878546.0, "step": 6792 }, { "epoch": 0.8641394224653351, "ewc_loss": 0.10302734375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 5.939586162567139, "learning_rate": 1e-05, "loss": 0.4986, "mean_token_accuracy": 0.8654149770736694, "num_tokens": 257915480.0, "step": 6793 }, { "epoch": 0.8642666327439257, "ewc_loss": 0.11083984375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.559258460998535, "learning_rate": 1e-05, "loss": 0.5259, "mean_token_accuracy": 0.8596843481063843, "num_tokens": 257951215.0, "step": 6794 }, { "epoch": 0.8643938430225162, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 5.966831207275391, "learning_rate": 1e-05, "loss": 0.5223, "mean_token_accuracy": 0.856303870677948, "num_tokens": 257987623.0, "step": 6795 }, { "epoch": 0.8645210533011067, "ewc_loss": 0.10107421875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 6.233133792877197, "learning_rate": 1e-05, "loss": 0.5533, "mean_token_accuracy": 0.8489125967025757, "num_tokens": 258030078.0, "step": 6796 }, { "epoch": 0.8646482635796973, "ewc_loss": 0.0986328125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.922250747680664, "learning_rate": 1e-05, "loss": 0.477, "mean_token_accuracy": 0.8719780445098877, "num_tokens": 258064598.0, "step": 6797 }, { "epoch": 0.8647754738582878, "ewc_loss": 0.09765625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.984189510345459, "learning_rate": 1e-05, "loss": 0.4944, "mean_token_accuracy": 0.8649576306343079, "num_tokens": 258100121.0, "step": 6798 }, { "epoch": 0.8649026841368782, "ewc_loss": 0.095703125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.773370742797852, "learning_rate": 1e-05, "loss": 0.5116, "mean_token_accuracy": 0.8572123646736145, "num_tokens": 258138398.0, "step": 6799 }, { "epoch": 0.8650298944154687, "ewc_loss": 0.0966796875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.865723609924316, "learning_rate": 1e-05, "loss": 0.5142, "mean_token_accuracy": 0.8596041202545166, "num_tokens": 258178730.0, "step": 6800 }, { "epoch": 0.8651571046940593, "ewc_loss": 0.0927734375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.642861843109131, "learning_rate": 1e-05, "loss": 0.4963, "mean_token_accuracy": 0.8647098541259766, "num_tokens": 258211122.0, "step": 6801 }, { "epoch": 0.8652843149726498, "ewc_loss": 0.09716796875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.875580310821533, "learning_rate": 1e-05, "loss": 0.4857, "mean_token_accuracy": 0.8679664134979248, "num_tokens": 258249753.0, "step": 6802 }, { "epoch": 0.8654115252512403, "ewc_loss": 0.08935546875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 7.62939453125e-05, "grad_norm": 5.48891019821167, "learning_rate": 1e-05, "loss": 0.5024, "mean_token_accuracy": 0.8590461015701294, "num_tokens": 258287951.0, "step": 6803 }, { "epoch": 0.8655387355298308, "ewc_loss": 0.09912109375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.969438076019287, "learning_rate": 1e-05, "loss": 0.5481, "mean_token_accuracy": 0.8515511751174927, "num_tokens": 258333216.0, "step": 6804 }, { "epoch": 0.8656659458084213, "ewc_loss": 0.0869140625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 7.390975952148438e-05, "grad_norm": 5.336593151092529, "learning_rate": 1e-05, "loss": 0.4945, "mean_token_accuracy": 0.8624390363693237, "num_tokens": 258370380.0, "step": 6805 }, { "epoch": 0.8657931560870118, "ewc_loss": 0.10205078125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.176263332366943, "learning_rate": 1e-05, "loss": 0.5661, "mean_token_accuracy": 0.8484506607055664, "num_tokens": 258413730.0, "step": 6806 }, { "epoch": 0.8659203663656023, "ewc_loss": 0.08544921875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 7.2479248046875e-05, "grad_norm": 5.196636199951172, "learning_rate": 1e-05, "loss": 0.4944, "mean_token_accuracy": 0.8638768196105957, "num_tokens": 258451644.0, "step": 6807 }, { "epoch": 0.8660475766441929, "ewc_loss": 0.10791015625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.528389930725098, "learning_rate": 1e-05, "loss": 0.4842, "mean_token_accuracy": 0.8741563558578491, "num_tokens": 258490084.0, "step": 6808 }, { "epoch": 0.8661747869227834, "ewc_loss": 0.08740234375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 7.43865966796875e-05, "grad_norm": 5.1705241203308105, "learning_rate": 1e-05, "loss": 0.5621, "mean_token_accuracy": 0.8437926173210144, "num_tokens": 258530769.0, "step": 6809 }, { "epoch": 0.8663019972013739, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.924367427825928, "learning_rate": 1e-05, "loss": 0.5612, "mean_token_accuracy": 0.8536694645881653, "num_tokens": 258567578.0, "step": 6810 }, { "epoch": 0.8664292074799644, "ewc_loss": 0.095703125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.469174861907959, "learning_rate": 1e-05, "loss": 0.5002, "mean_token_accuracy": 0.8609486818313599, "num_tokens": 258605498.0, "step": 6811 }, { "epoch": 0.8665564177585549, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 7.040808200836182, "learning_rate": 1e-05, "loss": 0.4863, "mean_token_accuracy": 0.878389298915863, "num_tokens": 258646931.0, "step": 6812 }, { "epoch": 0.8666836280371454, "ewc_loss": 0.10498046875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 5.911321640014648, "learning_rate": 1e-05, "loss": 0.4953, "mean_token_accuracy": 0.8674699664115906, "num_tokens": 258683853.0, "step": 6813 }, { "epoch": 0.8668108383157359, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.705938339233398, "learning_rate": 1e-05, "loss": 0.5184, "mean_token_accuracy": 0.862245500087738, "num_tokens": 258723706.0, "step": 6814 }, { "epoch": 0.8669380485943264, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.086226940155029, "learning_rate": 1e-05, "loss": 0.5085, "mean_token_accuracy": 0.8638651967048645, "num_tokens": 258762404.0, "step": 6815 }, { "epoch": 0.867065258872917, "ewc_loss": 0.10400390625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.22399377822876, "learning_rate": 1e-05, "loss": 0.5583, "mean_token_accuracy": 0.8494789600372314, "num_tokens": 258803209.0, "step": 6816 }, { "epoch": 0.8671924691515075, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.0899658203125, "learning_rate": 1e-05, "loss": 0.524, "mean_token_accuracy": 0.8607022762298584, "num_tokens": 258840001.0, "step": 6817 }, { "epoch": 0.8673196794300979, "ewc_loss": 0.09375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.799578666687012, "learning_rate": 1e-05, "loss": 0.4435, "mean_token_accuracy": 0.8796484470367432, "num_tokens": 258871401.0, "step": 6818 }, { "epoch": 0.8674468897086884, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 6.138614177703857, "learning_rate": 1e-05, "loss": 0.5642, "mean_token_accuracy": 0.8453353643417358, "num_tokens": 258902898.0, "step": 6819 }, { "epoch": 0.867574099987279, "ewc_loss": 0.0908203125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.487935543060303, "learning_rate": 1e-05, "loss": 0.4931, "mean_token_accuracy": 0.865659236907959, "num_tokens": 258942136.0, "step": 6820 }, { "epoch": 0.8677013102658695, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.2679362297058105, "learning_rate": 1e-05, "loss": 0.5148, "mean_token_accuracy": 0.860237181186676, "num_tokens": 258976599.0, "step": 6821 }, { "epoch": 0.86782852054446, "ewc_loss": 0.08837890625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 7.534027099609375e-05, "grad_norm": 5.3344926834106445, "learning_rate": 1e-05, "loss": 0.5034, "mean_token_accuracy": 0.8582478165626526, "num_tokens": 259010457.0, "step": 6822 }, { "epoch": 0.8679557308230506, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.466000556945801, "learning_rate": 1e-05, "loss": 0.5146, "mean_token_accuracy": 0.8606671094894409, "num_tokens": 259046368.0, "step": 6823 }, { "epoch": 0.868082941101641, "ewc_loss": 0.08984375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.30854606628418, "learning_rate": 1e-05, "loss": 0.4796, "mean_token_accuracy": 0.8689035177230835, "num_tokens": 259087240.0, "step": 6824 }, { "epoch": 0.8682101513802315, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.678187847137451, "learning_rate": 1e-05, "loss": 0.53, "mean_token_accuracy": 0.8623010516166687, "num_tokens": 259128297.0, "step": 6825 }, { "epoch": 0.868337361658822, "ewc_loss": 0.09326171875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.435701370239258, "learning_rate": 1e-05, "loss": 0.5154, "mean_token_accuracy": 0.8589925765991211, "num_tokens": 259170242.0, "step": 6826 }, { "epoch": 0.8684645719374126, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.782406806945801, "learning_rate": 1e-05, "loss": 0.5231, "mean_token_accuracy": 0.8584264516830444, "num_tokens": 259208580.0, "step": 6827 }, { "epoch": 0.8685917822160031, "ewc_loss": 0.09814453125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.6561760902404785, "learning_rate": 1e-05, "loss": 0.5792, "mean_token_accuracy": 0.8413742184638977, "num_tokens": 259249721.0, "step": 6828 }, { "epoch": 0.8687189924945936, "ewc_loss": 0.11376953125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.674036979675293, "learning_rate": 1e-05, "loss": 0.5026, "mean_token_accuracy": 0.8677548170089722, "num_tokens": 259292056.0, "step": 6829 }, { "epoch": 0.868846202773184, "ewc_loss": 0.09912109375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.80598258972168, "learning_rate": 1e-05, "loss": 0.4957, "mean_token_accuracy": 0.8681047558784485, "num_tokens": 259323390.0, "step": 6830 }, { "epoch": 0.8689734130517746, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.4253106117248535, "learning_rate": 1e-05, "loss": 0.4809, "mean_token_accuracy": 0.8717482089996338, "num_tokens": 259360728.0, "step": 6831 }, { "epoch": 0.8691006233303651, "ewc_loss": 0.09814453125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.817235946655273, "learning_rate": 1e-05, "loss": 0.5079, "mean_token_accuracy": 0.8587661981582642, "num_tokens": 259397619.0, "step": 6832 }, { "epoch": 0.8692278336089556, "ewc_loss": 0.10107421875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 6.166757106781006, "learning_rate": 1e-05, "loss": 0.5566, "mean_token_accuracy": 0.847535252571106, "num_tokens": 259440049.0, "step": 6833 }, { "epoch": 0.8693550438875461, "ewc_loss": 0.09765625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.778797626495361, "learning_rate": 1e-05, "loss": 0.4961, "mean_token_accuracy": 0.8670931458473206, "num_tokens": 259475657.0, "step": 6834 }, { "epoch": 0.8694822541661367, "ewc_loss": 0.0986328125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.973853588104248, "learning_rate": 1e-05, "loss": 0.5804, "mean_token_accuracy": 0.8388596773147583, "num_tokens": 259511289.0, "step": 6835 }, { "epoch": 0.8696094644447271, "ewc_loss": 0.09423828125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.713226318359375, "learning_rate": 1e-05, "loss": 0.472, "mean_token_accuracy": 0.8736559152603149, "num_tokens": 259546704.0, "step": 6836 }, { "epoch": 0.8697366747233176, "ewc_loss": 0.09619140625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.862189769744873, "learning_rate": 1e-05, "loss": 0.48, "mean_token_accuracy": 0.8724231719970703, "num_tokens": 259588490.0, "step": 6837 }, { "epoch": 0.8698638850019081, "ewc_loss": 0.09326171875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.631328105926514, "learning_rate": 1e-05, "loss": 0.5186, "mean_token_accuracy": 0.8608291149139404, "num_tokens": 259619838.0, "step": 6838 }, { "epoch": 0.8699910952804987, "ewc_loss": 0.09619140625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.8407464027404785, "learning_rate": 1e-05, "loss": 0.5407, "mean_token_accuracy": 0.8522056937217712, "num_tokens": 259662232.0, "step": 6839 }, { "epoch": 0.8701183055590892, "ewc_loss": 0.09130859375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.524059772491455, "learning_rate": 1e-05, "loss": 0.5755, "mean_token_accuracy": 0.838432788848877, "num_tokens": 259699824.0, "step": 6840 }, { "epoch": 0.8702455158376797, "ewc_loss": 0.09619140625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.896743297576904, "learning_rate": 1e-05, "loss": 0.5153, "mean_token_accuracy": 0.8589165210723877, "num_tokens": 259728217.0, "step": 6841 }, { "epoch": 0.8703727261162701, "ewc_loss": 0.08837890625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 7.534027099609375e-05, "grad_norm": 5.404361724853516, "learning_rate": 1e-05, "loss": 0.4545, "mean_token_accuracy": 0.8743760585784912, "num_tokens": 259763522.0, "step": 6842 }, { "epoch": 0.8704999363948607, "ewc_loss": 0.099609375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 6.023035049438477, "learning_rate": 1e-05, "loss": 0.5433, "mean_token_accuracy": 0.8508611917495728, "num_tokens": 259797650.0, "step": 6843 }, { "epoch": 0.8706271466734512, "ewc_loss": 0.08642578125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 7.343292236328125e-05, "grad_norm": 5.272729873657227, "learning_rate": 1e-05, "loss": 0.5447, "mean_token_accuracy": 0.851947546005249, "num_tokens": 259837265.0, "step": 6844 }, { "epoch": 0.8707543569520417, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.26529598236084, "learning_rate": 1e-05, "loss": 0.5111, "mean_token_accuracy": 0.8634021282196045, "num_tokens": 259873040.0, "step": 6845 }, { "epoch": 0.8708815672306323, "ewc_loss": 0.08544921875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 7.2479248046875e-05, "grad_norm": 5.1693115234375, "learning_rate": 1e-05, "loss": 0.5443, "mean_token_accuracy": 0.8452777862548828, "num_tokens": 259907627.0, "step": 6846 }, { "epoch": 0.8710087775092228, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.615004062652588, "learning_rate": 1e-05, "loss": 0.591, "mean_token_accuracy": 0.8411675691604614, "num_tokens": 259948213.0, "step": 6847 }, { "epoch": 0.8711359877878132, "ewc_loss": 0.08984375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.215588092803955, "learning_rate": 1e-05, "loss": 0.5191, "mean_token_accuracy": 0.8545453548431396, "num_tokens": 259985239.0, "step": 6848 }, { "epoch": 0.8712631980664037, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.95005464553833, "learning_rate": 1e-05, "loss": 0.6286, "mean_token_accuracy": 0.8326305747032166, "num_tokens": 260023046.0, "step": 6849 }, { "epoch": 0.8713904083449943, "ewc_loss": 0.09716796875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.565222263336182, "learning_rate": 1e-05, "loss": 0.4762, "mean_token_accuracy": 0.871004581451416, "num_tokens": 260060716.0, "step": 6850 }, { "epoch": 0.8715176186235848, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.931550979614258, "learning_rate": 1e-05, "loss": 0.4874, "mean_token_accuracy": 0.8790339231491089, "num_tokens": 260093939.0, "step": 6851 }, { "epoch": 0.8716448289021753, "ewc_loss": 0.10595703125, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 5.937054634094238, "learning_rate": 1e-05, "loss": 0.5184, "mean_token_accuracy": 0.8611775040626526, "num_tokens": 260129282.0, "step": 6852 }, { "epoch": 0.8717720391807658, "ewc_loss": 0.11083984375, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.535271167755127, "learning_rate": 1e-05, "loss": 0.4802, "mean_token_accuracy": 0.8752607107162476, "num_tokens": 260172167.0, "step": 6853 }, { "epoch": 0.8718992494593563, "ewc_loss": 0.10693359375, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.034304618835449, "learning_rate": 1e-05, "loss": 0.4992, "mean_token_accuracy": 0.8670665621757507, "num_tokens": 260212613.0, "step": 6854 }, { "epoch": 0.8720264597379468, "ewc_loss": 0.10009765625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 6.073759078979492, "learning_rate": 1e-05, "loss": 0.4982, "mean_token_accuracy": 0.8668165802955627, "num_tokens": 260248774.0, "step": 6855 }, { "epoch": 0.8721536700165373, "ewc_loss": 0.10302734375, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.061192512512207, "learning_rate": 1e-05, "loss": 0.5648, "mean_token_accuracy": 0.8427259922027588, "num_tokens": 260291105.0, "step": 6856 }, { "epoch": 0.8722808802951278, "ewc_loss": 0.09228515625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.654882431030273, "learning_rate": 1e-05, "loss": 0.4797, "mean_token_accuracy": 0.8668307065963745, "num_tokens": 260332316.0, "step": 6857 }, { "epoch": 0.8724080905737184, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.1698784828186035, "learning_rate": 1e-05, "loss": 0.4746, "mean_token_accuracy": 0.8710027933120728, "num_tokens": 260365016.0, "step": 6858 }, { "epoch": 0.8725353008523089, "ewc_loss": 0.087890625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 7.486343383789062e-05, "grad_norm": 5.404400825500488, "learning_rate": 1e-05, "loss": 0.4362, "mean_token_accuracy": 0.879172682762146, "num_tokens": 260396131.0, "step": 6859 }, { "epoch": 0.8726625111308994, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.338475704193115, "learning_rate": 1e-05, "loss": 0.5719, "mean_token_accuracy": 0.8469794988632202, "num_tokens": 260438053.0, "step": 6860 }, { "epoch": 0.8727897214094898, "ewc_loss": 0.0888671875, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 7.581710815429688e-05, "grad_norm": 5.284305572509766, "learning_rate": 1e-05, "loss": 0.506, "mean_token_accuracy": 0.8581651449203491, "num_tokens": 260479523.0, "step": 6861 }, { "epoch": 0.8729169316880804, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.592635154724121, "learning_rate": 1e-05, "loss": 0.5157, "mean_token_accuracy": 0.8679730296134949, "num_tokens": 260516623.0, "step": 6862 }, { "epoch": 0.8730441419666709, "ewc_loss": 0.09130859375, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.351824760437012, "learning_rate": 1e-05, "loss": 0.4795, "mean_token_accuracy": 0.867246150970459, "num_tokens": 260552499.0, "step": 6863 }, { "epoch": 0.8731713522452614, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.7528204917907715, "learning_rate": 1e-05, "loss": 0.5129, "mean_token_accuracy": 0.8629138469696045, "num_tokens": 260595636.0, "step": 6864 }, { "epoch": 0.873298562523852, "ewc_loss": 0.09521484375, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.545009613037109, "learning_rate": 1e-05, "loss": 0.492, "mean_token_accuracy": 0.867354154586792, "num_tokens": 260639787.0, "step": 6865 }, { "epoch": 0.8734257728024425, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.731183052062988, "learning_rate": 1e-05, "loss": 0.4929, "mean_token_accuracy": 0.8742988109588623, "num_tokens": 260674117.0, "step": 6866 }, { "epoch": 0.8735529830810329, "ewc_loss": 0.099609375, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.739993095397949, "learning_rate": 1e-05, "loss": 0.5089, "mean_token_accuracy": 0.8617745637893677, "num_tokens": 260712757.0, "step": 6867 }, { "epoch": 0.8736801933596234, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.533843994140625, "learning_rate": 1e-05, "loss": 0.5812, "mean_token_accuracy": 0.8423106670379639, "num_tokens": 260750756.0, "step": 6868 }, { "epoch": 0.873807403638214, "ewc_loss": 0.10009765625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.809162616729736, "learning_rate": 1e-05, "loss": 0.5171, "mean_token_accuracy": 0.8575133085250854, "num_tokens": 260782929.0, "step": 6869 }, { "epoch": 0.8739346139168045, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.26625919342041, "learning_rate": 1e-05, "loss": 0.51, "mean_token_accuracy": 0.861200749874115, "num_tokens": 260819762.0, "step": 6870 }, { "epoch": 0.874061824195395, "ewc_loss": 0.09912109375, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.790319919586182, "learning_rate": 1e-05, "loss": 0.5287, "mean_token_accuracy": 0.8550887703895569, "num_tokens": 260862543.0, "step": 6871 }, { "epoch": 0.8741890344739855, "ewc_loss": 0.09912109375, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 6.045919895172119, "learning_rate": 1e-05, "loss": 0.5548, "mean_token_accuracy": 0.8486228585243225, "num_tokens": 260899835.0, "step": 6872 }, { "epoch": 0.874316244752576, "ewc_loss": 0.09619140625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.724277019500732, "learning_rate": 1e-05, "loss": 0.5518, "mean_token_accuracy": 0.8474992513656616, "num_tokens": 260940593.0, "step": 6873 }, { "epoch": 0.8744434550311665, "ewc_loss": 0.0986328125, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.920558929443359, "learning_rate": 1e-05, "loss": 0.5128, "mean_token_accuracy": 0.86017245054245, "num_tokens": 260977454.0, "step": 6874 }, { "epoch": 0.874570665309757, "ewc_loss": 0.09375, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.631381034851074, "learning_rate": 1e-05, "loss": 0.5058, "mean_token_accuracy": 0.8612818121910095, "num_tokens": 261019095.0, "step": 6875 }, { "epoch": 0.8746978755883476, "ewc_loss": 0.09716796875, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.876718521118164, "learning_rate": 1e-05, "loss": 0.5328, "mean_token_accuracy": 0.8526662588119507, "num_tokens": 261056428.0, "step": 6876 }, { "epoch": 0.8748250858669381, "ewc_loss": 0.09130859375, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.534381866455078, "learning_rate": 1e-05, "loss": 0.4672, "mean_token_accuracy": 0.873255729675293, "num_tokens": 261090679.0, "step": 6877 }, { "epoch": 0.8749522961455286, "ewc_loss": 0.09814453125, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.925859451293945, "learning_rate": 1e-05, "loss": 0.5207, "mean_token_accuracy": 0.8572039604187012, "num_tokens": 261130640.0, "step": 6878 }, { "epoch": 0.875079506424119, "ewc_loss": 0.08837890625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 7.534027099609375e-05, "grad_norm": 9.863900184631348, "learning_rate": 1e-05, "loss": 0.4562, "mean_token_accuracy": 0.8731474876403809, "num_tokens": 261169064.0, "step": 6879 }, { "epoch": 0.8752067167027096, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.047494411468506, "learning_rate": 1e-05, "loss": 0.5136, "mean_token_accuracy": 0.863031804561615, "num_tokens": 261205625.0, "step": 6880 }, { "epoch": 0.8753339269813001, "ewc_loss": 0.11083984375, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.525940418243408, "learning_rate": 1e-05, "loss": 0.4912, "mean_token_accuracy": 0.8695729970932007, "num_tokens": 261251859.0, "step": 6881 }, { "epoch": 0.8754611372598906, "ewc_loss": 0.09130859375, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.268354892730713, "learning_rate": 1e-05, "loss": 0.5021, "mean_token_accuracy": 0.8623591065406799, "num_tokens": 261289604.0, "step": 6882 }, { "epoch": 0.8755883475384811, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 0.00011777877807617188, "grad_norm": 7.300558090209961, "learning_rate": 1e-05, "loss": 0.4702, "mean_token_accuracy": 0.8840300440788269, "num_tokens": 261330053.0, "step": 6883 }, { "epoch": 0.8757155578170717, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 5.612596035003662, "learning_rate": 1e-05, "loss": 0.5381, "mean_token_accuracy": 0.8510509729385376, "num_tokens": 261371042.0, "step": 6884 }, { "epoch": 0.8758427680956621, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 0.00012493133544921875, "grad_norm": 7.553001403808594, "learning_rate": 1e-05, "loss": 0.5821, "mean_token_accuracy": 0.8534183502197266, "num_tokens": 261406911.0, "step": 6885 }, { "epoch": 0.8759699783742526, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.2638654708862305, "learning_rate": 1e-05, "loss": 0.517, "mean_token_accuracy": 0.863370418548584, "num_tokens": 261443856.0, "step": 6886 }, { "epoch": 0.8760971886528431, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 0.00011777877807617188, "grad_norm": 7.2386016845703125, "learning_rate": 1e-05, "loss": 0.5673, "mean_token_accuracy": 0.8552377223968506, "num_tokens": 261485987.0, "step": 6887 }, { "epoch": 0.8762243989314337, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.466505527496338, "learning_rate": 1e-05, "loss": 0.5568, "mean_token_accuracy": 0.8540492057800293, "num_tokens": 261519888.0, "step": 6888 }, { "epoch": 0.8763516092100242, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.768749713897705, "learning_rate": 1e-05, "loss": 0.5364, "mean_token_accuracy": 0.8612582683563232, "num_tokens": 261557435.0, "step": 6889 }, { "epoch": 0.8764788194886147, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.288188457489014, "learning_rate": 1e-05, "loss": 0.5326, "mean_token_accuracy": 0.8594146966934204, "num_tokens": 261598209.0, "step": 6890 }, { "epoch": 0.8766060297672051, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.442263603210449, "learning_rate": 1e-05, "loss": 0.5639, "mean_token_accuracy": 0.8486043214797974, "num_tokens": 261637972.0, "step": 6891 }, { "epoch": 0.8767332400457957, "ewc_loss": 0.10400390625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.0422210693359375, "learning_rate": 1e-05, "loss": 0.5575, "mean_token_accuracy": 0.8473854064941406, "num_tokens": 261676717.0, "step": 6892 }, { "epoch": 0.8768604503243862, "ewc_loss": 0.10595703125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.2768659591674805, "learning_rate": 1e-05, "loss": 0.5795, "mean_token_accuracy": 0.8427301645278931, "num_tokens": 261711434.0, "step": 6893 }, { "epoch": 0.8769876606029767, "ewc_loss": 0.09912109375, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.817221641540527, "learning_rate": 1e-05, "loss": 0.5039, "mean_token_accuracy": 0.8630610704421997, "num_tokens": 261749527.0, "step": 6894 }, { "epoch": 0.8771148708815673, "ewc_loss": 0.10595703125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.2250590324401855, "learning_rate": 1e-05, "loss": 0.4962, "mean_token_accuracy": 0.8700263500213623, "num_tokens": 261792031.0, "step": 6895 }, { "epoch": 0.8772420811601578, "ewc_loss": 0.09326171875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.608126640319824, "learning_rate": 1e-05, "loss": 0.5709, "mean_token_accuracy": 0.8431035280227661, "num_tokens": 261831862.0, "step": 6896 }, { "epoch": 0.8773692914387482, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.347537040710449, "learning_rate": 1e-05, "loss": 0.5046, "mean_token_accuracy": 0.8662819266319275, "num_tokens": 261869949.0, "step": 6897 }, { "epoch": 0.8774965017173387, "ewc_loss": 0.0908203125, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.441557884216309, "learning_rate": 1e-05, "loss": 0.4899, "mean_token_accuracy": 0.8639621138572693, "num_tokens": 261904080.0, "step": 6898 }, { "epoch": 0.8776237119959293, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.6324543952941895, "learning_rate": 1e-05, "loss": 0.5469, "mean_token_accuracy": 0.8482431173324585, "num_tokens": 261932809.0, "step": 6899 }, { "epoch": 0.8777509222745198, "ewc_loss": 0.09033203125, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.433236598968506, "learning_rate": 1e-05, "loss": 0.4995, "mean_token_accuracy": 0.8604116439819336, "num_tokens": 261974313.0, "step": 6900 }, { "epoch": 0.8778781325531103, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.9205546379089355, "learning_rate": 1e-05, "loss": 0.4814, "mean_token_accuracy": 0.8742204308509827, "num_tokens": 262011505.0, "step": 6901 }, { "epoch": 0.8780053428317008, "ewc_loss": 0.09619140625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.684411525726318, "learning_rate": 1e-05, "loss": 0.5084, "mean_token_accuracy": 0.864495575428009, "num_tokens": 262042571.0, "step": 6902 }, { "epoch": 0.8781325531102913, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.931271553039551, "learning_rate": 1e-05, "loss": 0.5655, "mean_token_accuracy": 0.8517768383026123, "num_tokens": 262088038.0, "step": 6903 }, { "epoch": 0.8782597633888818, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.036383628845215, "learning_rate": 1e-05, "loss": 0.4824, "mean_token_accuracy": 0.8701260685920715, "num_tokens": 262126662.0, "step": 6904 }, { "epoch": 0.8783869736674723, "ewc_loss": 0.10888671875, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.51388692855835, "learning_rate": 1e-05, "loss": 0.5212, "mean_token_accuracy": 0.863252580165863, "num_tokens": 262170525.0, "step": 6905 }, { "epoch": 0.8785141839460628, "ewc_loss": 0.10498046875, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.170803546905518, "learning_rate": 1e-05, "loss": 0.5784, "mean_token_accuracy": 0.8468544483184814, "num_tokens": 262210524.0, "step": 6906 }, { "epoch": 0.8786413942246534, "ewc_loss": 0.095703125, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.968108177185059, "learning_rate": 1e-05, "loss": 0.5846, "mean_token_accuracy": 0.8421967029571533, "num_tokens": 262248636.0, "step": 6907 }, { "epoch": 0.8787686045032439, "ewc_loss": 0.10498046875, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.2182722091674805, "learning_rate": 1e-05, "loss": 0.553, "mean_token_accuracy": 0.8493489623069763, "num_tokens": 262291795.0, "step": 6908 }, { "epoch": 0.8788958147818343, "ewc_loss": 0.087890625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 7.43865966796875e-05, "grad_norm": 5.479121685028076, "learning_rate": 1e-05, "loss": 0.4867, "mean_token_accuracy": 0.8647592663764954, "num_tokens": 262328484.0, "step": 6909 }, { "epoch": 0.8790230250604248, "ewc_loss": 0.10693359375, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.422188758850098, "learning_rate": 1e-05, "loss": 0.5063, "mean_token_accuracy": 0.8656530380249023, "num_tokens": 262362262.0, "step": 6910 }, { "epoch": 0.8791502353390154, "ewc_loss": 0.0859375, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 7.2479248046875e-05, "grad_norm": 5.198606014251709, "learning_rate": 1e-05, "loss": 0.5284, "mean_token_accuracy": 0.8535131812095642, "num_tokens": 262399082.0, "step": 6911 }, { "epoch": 0.8792774456176059, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.8256940841674805, "learning_rate": 1e-05, "loss": 0.5249, "mean_token_accuracy": 0.8621032238006592, "num_tokens": 262434968.0, "step": 6912 }, { "epoch": 0.8794046558961964, "ewc_loss": 0.08984375, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 7.62939453125e-05, "grad_norm": 5.265090465545654, "learning_rate": 1e-05, "loss": 0.4791, "mean_token_accuracy": 0.8658354878425598, "num_tokens": 262472989.0, "step": 6913 }, { "epoch": 0.879531866174787, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 0.00011205673217773438, "grad_norm": 7.221867084503174, "learning_rate": 1e-05, "loss": 0.5391, "mean_token_accuracy": 0.861539363861084, "num_tokens": 262512526.0, "step": 6914 }, { "epoch": 0.8796590764533775, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 5.802587509155273, "learning_rate": 1e-05, "loss": 0.4855, "mean_token_accuracy": 0.8673151731491089, "num_tokens": 262548082.0, "step": 6915 }, { "epoch": 0.8797862867319679, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 0.00011014938354492188, "grad_norm": 7.101174354553223, "learning_rate": 1e-05, "loss": 0.5892, "mean_token_accuracy": 0.8442147970199585, "num_tokens": 262584786.0, "step": 6916 }, { "epoch": 0.8799134970105584, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.21405553817749, "learning_rate": 1e-05, "loss": 0.5148, "mean_token_accuracy": 0.8646830916404724, "num_tokens": 262625898.0, "step": 6917 }, { "epoch": 0.880040707289149, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.622849941253662, "learning_rate": 1e-05, "loss": 0.5136, "mean_token_accuracy": 0.8617859482765198, "num_tokens": 262664041.0, "step": 6918 }, { "epoch": 0.8801679175677395, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.1928935050964355, "learning_rate": 1e-05, "loss": 0.5628, "mean_token_accuracy": 0.8506481647491455, "num_tokens": 262702597.0, "step": 6919 }, { "epoch": 0.88029512784633, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.214033603668213, "learning_rate": 1e-05, "loss": 0.5798, "mean_token_accuracy": 0.8415130376815796, "num_tokens": 262734951.0, "step": 6920 }, { "epoch": 0.8804223381249205, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 6.034265041351318, "learning_rate": 1e-05, "loss": 0.4945, "mean_token_accuracy": 0.8677753210067749, "num_tokens": 262771513.0, "step": 6921 }, { "epoch": 0.880549548403511, "ewc_loss": 0.0966796875, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.955888748168945, "learning_rate": 1e-05, "loss": 0.5268, "mean_token_accuracy": 0.8523687124252319, "num_tokens": 262813045.0, "step": 6922 }, { "epoch": 0.8806767586821015, "ewc_loss": 0.0966796875, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.873774528503418, "learning_rate": 1e-05, "loss": 0.4713, "mean_token_accuracy": 0.8691527843475342, "num_tokens": 262846663.0, "step": 6923 }, { "epoch": 0.880803968960692, "ewc_loss": 0.0947265625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.864038944244385, "learning_rate": 1e-05, "loss": 0.5069, "mean_token_accuracy": 0.8624175786972046, "num_tokens": 262880056.0, "step": 6924 }, { "epoch": 0.8809311792392825, "ewc_loss": 0.0927734375, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.693833827972412, "learning_rate": 1e-05, "loss": 0.4785, "mean_token_accuracy": 0.8687970042228699, "num_tokens": 262914819.0, "step": 6925 }, { "epoch": 0.8810583895178731, "ewc_loss": 0.0966796875, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.875733852386475, "learning_rate": 1e-05, "loss": 0.5238, "mean_token_accuracy": 0.8567342758178711, "num_tokens": 262953167.0, "step": 6926 }, { "epoch": 0.8811855997964636, "ewc_loss": 0.0888671875, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 7.534027099609375e-05, "grad_norm": 5.515427112579346, "learning_rate": 1e-05, "loss": 0.5269, "mean_token_accuracy": 0.8548868894577026, "num_tokens": 262991628.0, "step": 6927 }, { "epoch": 0.881312810075054, "ewc_loss": 0.0986328125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.993618965148926, "learning_rate": 1e-05, "loss": 0.494, "mean_token_accuracy": 0.8651663661003113, "num_tokens": 263029287.0, "step": 6928 }, { "epoch": 0.8814400203536445, "ewc_loss": 0.0869140625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 7.343292236328125e-05, "grad_norm": 5.33240270614624, "learning_rate": 1e-05, "loss": 0.4875, "mean_token_accuracy": 0.8606338500976562, "num_tokens": 263068982.0, "step": 6929 }, { "epoch": 0.8815672306322351, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.239826679229736, "learning_rate": 1e-05, "loss": 0.6067, "mean_token_accuracy": 0.8373781442642212, "num_tokens": 263111260.0, "step": 6930 }, { "epoch": 0.8816944409108256, "ewc_loss": 0.0849609375, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 7.152557373046875e-05, "grad_norm": 5.228270053863525, "learning_rate": 1e-05, "loss": 0.475, "mean_token_accuracy": 0.8628256320953369, "num_tokens": 263149310.0, "step": 6931 }, { "epoch": 0.8818216511894161, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.570322513580322, "learning_rate": 1e-05, "loss": 0.4902, "mean_token_accuracy": 0.8723098039627075, "num_tokens": 263188440.0, "step": 6932 }, { "epoch": 0.8819488614680067, "ewc_loss": 0.0869140625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 7.343292236328125e-05, "grad_norm": 5.167899131774902, "learning_rate": 1e-05, "loss": 0.5273, "mean_token_accuracy": 0.8523377180099487, "num_tokens": 263229079.0, "step": 6933 }, { "epoch": 0.8820760717465971, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 7.010921478271484, "learning_rate": 1e-05, "loss": 0.5576, "mean_token_accuracy": 0.8539508581161499, "num_tokens": 263260591.0, "step": 6934 }, { "epoch": 0.8822032820251876, "ewc_loss": 0.09521484375, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.497627258300781, "learning_rate": 1e-05, "loss": 0.544, "mean_token_accuracy": 0.8486360311508179, "num_tokens": 263301256.0, "step": 6935 }, { "epoch": 0.8823304923037781, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 0.00011157989501953125, "grad_norm": 7.172935962677002, "learning_rate": 1e-05, "loss": 0.5157, "mean_token_accuracy": 0.8656150102615356, "num_tokens": 263337600.0, "step": 6936 }, { "epoch": 0.8824577025823687, "ewc_loss": 0.10693359375, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.015815258026123, "learning_rate": 1e-05, "loss": 0.5346, "mean_token_accuracy": 0.8558855056762695, "num_tokens": 263383124.0, "step": 6937 }, { "epoch": 0.8825849128609592, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.811007022857666, "learning_rate": 1e-05, "loss": 0.5062, "mean_token_accuracy": 0.8682184219360352, "num_tokens": 263416047.0, "step": 6938 }, { "epoch": 0.8827121231395497, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.191418647766113, "learning_rate": 1e-05, "loss": 0.4787, "mean_token_accuracy": 0.872728705406189, "num_tokens": 263458515.0, "step": 6939 }, { "epoch": 0.8828393334181401, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.309038162231445, "learning_rate": 1e-05, "loss": 0.553, "mean_token_accuracy": 0.8512486815452576, "num_tokens": 263499833.0, "step": 6940 }, { "epoch": 0.8829665436967307, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.156523704528809, "learning_rate": 1e-05, "loss": 0.5266, "mean_token_accuracy": 0.8568769693374634, "num_tokens": 263531680.0, "step": 6941 }, { "epoch": 0.8830937539753212, "ewc_loss": 0.0947265625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.875232219696045, "learning_rate": 1e-05, "loss": 0.5438, "mean_token_accuracy": 0.8474578857421875, "num_tokens": 263565889.0, "step": 6942 }, { "epoch": 0.8832209642539117, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.151284217834473, "learning_rate": 1e-05, "loss": 0.5391, "mean_token_accuracy": 0.8533940315246582, "num_tokens": 263604974.0, "step": 6943 }, { "epoch": 0.8833481745325023, "ewc_loss": 0.09130859375, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.57263708114624, "learning_rate": 1e-05, "loss": 0.5053, "mean_token_accuracy": 0.8612948656082153, "num_tokens": 263646960.0, "step": 6944 }, { "epoch": 0.8834753848110928, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.217211723327637, "learning_rate": 1e-05, "loss": 0.5265, "mean_token_accuracy": 0.858803927898407, "num_tokens": 263686427.0, "step": 6945 }, { "epoch": 0.8836025950896832, "ewc_loss": 0.08837890625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 7.486343383789062e-05, "grad_norm": 5.401289463043213, "learning_rate": 1e-05, "loss": 0.4591, "mean_token_accuracy": 0.8704001307487488, "num_tokens": 263717520.0, "step": 6946 }, { "epoch": 0.8837298053682737, "ewc_loss": 0.10595703125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.369388103485107, "learning_rate": 1e-05, "loss": 0.534, "mean_token_accuracy": 0.8578693866729736, "num_tokens": 263756226.0, "step": 6947 }, { "epoch": 0.8838570156468643, "ewc_loss": 0.08935546875, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 7.581710815429688e-05, "grad_norm": 5.319457530975342, "learning_rate": 1e-05, "loss": 0.471, "mean_token_accuracy": 0.872262179851532, "num_tokens": 263793961.0, "step": 6948 }, { "epoch": 0.8839842259254548, "ewc_loss": 0.11083984375, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.588094711303711, "learning_rate": 1e-05, "loss": 0.5323, "mean_token_accuracy": 0.8567363023757935, "num_tokens": 263827325.0, "step": 6949 }, { "epoch": 0.8841114362040453, "ewc_loss": 0.091796875, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.367717742919922, "learning_rate": 1e-05, "loss": 0.5678, "mean_token_accuracy": 0.8385109305381775, "num_tokens": 263862854.0, "step": 6950 }, { "epoch": 0.8842386464826358, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.762350559234619, "learning_rate": 1e-05, "loss": 0.6153, "mean_token_accuracy": 0.834037721157074, "num_tokens": 263907035.0, "step": 6951 }, { "epoch": 0.8843658567612263, "ewc_loss": 0.09521484375, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.554790019989014, "learning_rate": 1e-05, "loss": 0.5448, "mean_token_accuracy": 0.8508166074752808, "num_tokens": 263949550.0, "step": 6952 }, { "epoch": 0.8844930670398168, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.775968551635742, "learning_rate": 1e-05, "loss": 0.5254, "mean_token_accuracy": 0.8600959777832031, "num_tokens": 263988966.0, "step": 6953 }, { "epoch": 0.8846202773184073, "ewc_loss": 0.10009765625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.758195877075195, "learning_rate": 1e-05, "loss": 0.47, "mean_token_accuracy": 0.8744520545005798, "num_tokens": 264023442.0, "step": 6954 }, { "epoch": 0.8847474875969978, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.566898345947266, "learning_rate": 1e-05, "loss": 0.4963, "mean_token_accuracy": 0.8694002032279968, "num_tokens": 264064185.0, "step": 6955 }, { "epoch": 0.8848746978755884, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.849429130554199, "learning_rate": 1e-05, "loss": 0.541, "mean_token_accuracy": 0.8513752818107605, "num_tokens": 264107455.0, "step": 6956 }, { "epoch": 0.8850019081541789, "ewc_loss": 0.10498046875, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.274551868438721, "learning_rate": 1e-05, "loss": 0.5262, "mean_token_accuracy": 0.8534358143806458, "num_tokens": 264143090.0, "step": 6957 }, { "epoch": 0.8851291184327693, "ewc_loss": 0.09912109375, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.853741645812988, "learning_rate": 1e-05, "loss": 0.529, "mean_token_accuracy": 0.8561598062515259, "num_tokens": 264180002.0, "step": 6958 }, { "epoch": 0.8852563287113598, "ewc_loss": 0.09765625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 6.008884906768799, "learning_rate": 1e-05, "loss": 0.4869, "mean_token_accuracy": 0.8686200380325317, "num_tokens": 264215444.0, "step": 6959 }, { "epoch": 0.8853835389899504, "ewc_loss": 0.0966796875, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.809086799621582, "learning_rate": 1e-05, "loss": 0.4983, "mean_token_accuracy": 0.8596360683441162, "num_tokens": 264250758.0, "step": 6960 }, { "epoch": 0.8855107492685409, "ewc_loss": 0.0947265625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.825772762298584, "learning_rate": 1e-05, "loss": 0.5571, "mean_token_accuracy": 0.8469175696372986, "num_tokens": 264290181.0, "step": 6961 }, { "epoch": 0.8856379595471314, "ewc_loss": 0.09521484375, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.748355388641357, "learning_rate": 1e-05, "loss": 0.5161, "mean_token_accuracy": 0.8613208532333374, "num_tokens": 264327480.0, "step": 6962 }, { "epoch": 0.885765169825722, "ewc_loss": 0.09228515625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.725213050842285, "learning_rate": 1e-05, "loss": 0.5064, "mean_token_accuracy": 0.8615502119064331, "num_tokens": 264368762.0, "step": 6963 }, { "epoch": 0.8858923801043125, "ewc_loss": 0.09326171875, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.662532329559326, "learning_rate": 1e-05, "loss": 0.5051, "mean_token_accuracy": 0.8604874014854431, "num_tokens": 264410261.0, "step": 6964 }, { "epoch": 0.8860195903829029, "ewc_loss": 0.09375, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.7092719078063965, "learning_rate": 1e-05, "loss": 0.4895, "mean_token_accuracy": 0.8659989833831787, "num_tokens": 264449111.0, "step": 6965 }, { "epoch": 0.8861468006614934, "ewc_loss": 0.09228515625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.591774940490723, "learning_rate": 1e-05, "loss": 0.5175, "mean_token_accuracy": 0.8550657629966736, "num_tokens": 264487045.0, "step": 6966 }, { "epoch": 0.886274010940084, "ewc_loss": 0.09423828125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.740161418914795, "learning_rate": 1e-05, "loss": 0.5003, "mean_token_accuracy": 0.8633021116256714, "num_tokens": 264524025.0, "step": 6967 }, { "epoch": 0.8864012212186745, "ewc_loss": 0.0908203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.491217613220215, "learning_rate": 1e-05, "loss": 0.5142, "mean_token_accuracy": 0.8530325889587402, "num_tokens": 264561361.0, "step": 6968 }, { "epoch": 0.886528431497265, "ewc_loss": 0.095703125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.819498062133789, "learning_rate": 1e-05, "loss": 0.5889, "mean_token_accuracy": 0.8349425792694092, "num_tokens": 264603557.0, "step": 6969 }, { "epoch": 0.8866556417758555, "ewc_loss": 0.087890625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 7.43865966796875e-05, "grad_norm": 5.363658428192139, "learning_rate": 1e-05, "loss": 0.4679, "mean_token_accuracy": 0.8738031387329102, "num_tokens": 264638539.0, "step": 6970 }, { "epoch": 0.886782852054446, "ewc_loss": 0.09814453125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 6.006363391876221, "learning_rate": 1e-05, "loss": 0.5016, "mean_token_accuracy": 0.8615967035293579, "num_tokens": 264676592.0, "step": 6971 }, { "epoch": 0.8869100623330365, "ewc_loss": 0.0849609375, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 7.104873657226562e-05, "grad_norm": 5.195188522338867, "learning_rate": 1e-05, "loss": 0.4635, "mean_token_accuracy": 0.8713582754135132, "num_tokens": 264708339.0, "step": 6972 }, { "epoch": 0.887037272611627, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.34112024307251, "learning_rate": 1e-05, "loss": 0.5282, "mean_token_accuracy": 0.8590169548988342, "num_tokens": 264741705.0, "step": 6973 }, { "epoch": 0.8871644828902175, "ewc_loss": 0.0849609375, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 7.104873657226562e-05, "grad_norm": 5.052249431610107, "learning_rate": 1e-05, "loss": 0.49, "mean_token_accuracy": 0.8588520288467407, "num_tokens": 264773095.0, "step": 6974 }, { "epoch": 0.8872916931688081, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.861316204071045, "learning_rate": 1e-05, "loss": 0.4901, "mean_token_accuracy": 0.8769741058349609, "num_tokens": 264813499.0, "step": 6975 }, { "epoch": 0.8874189034473986, "ewc_loss": 0.09033203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.138664722442627, "learning_rate": 1e-05, "loss": 0.4901, "mean_token_accuracy": 0.8668283820152283, "num_tokens": 264851314.0, "step": 6976 }, { "epoch": 0.887546113725989, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 0.00011968612670898438, "grad_norm": 7.413783550262451, "learning_rate": 1e-05, "loss": 0.5335, "mean_token_accuracy": 0.8624252080917358, "num_tokens": 264886908.0, "step": 6977 }, { "epoch": 0.8876733240045795, "ewc_loss": 0.10595703125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 5.829396724700928, "learning_rate": 1e-05, "loss": 0.6029, "mean_token_accuracy": 0.8333913087844849, "num_tokens": 264927516.0, "step": 6978 }, { "epoch": 0.8878005342831701, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 0.00011539459228515625, "grad_norm": 7.319483280181885, "learning_rate": 1e-05, "loss": 0.5223, "mean_token_accuracy": 0.8656310439109802, "num_tokens": 264966758.0, "step": 6979 }, { "epoch": 0.8879277445617606, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.34395170211792, "learning_rate": 1e-05, "loss": 0.5588, "mean_token_accuracy": 0.8503931760787964, "num_tokens": 265002267.0, "step": 6980 }, { "epoch": 0.8880549548403511, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.813135147094727, "learning_rate": 1e-05, "loss": 0.5062, "mean_token_accuracy": 0.86758953332901, "num_tokens": 265034957.0, "step": 6981 }, { "epoch": 0.8881821651189417, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.258425712585449, "learning_rate": 1e-05, "loss": 0.5191, "mean_token_accuracy": 0.8609464764595032, "num_tokens": 265075587.0, "step": 6982 }, { "epoch": 0.8883093753975321, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.41576623916626, "learning_rate": 1e-05, "loss": 0.4798, "mean_token_accuracy": 0.8719493746757507, "num_tokens": 265111385.0, "step": 6983 }, { "epoch": 0.8884365856761226, "ewc_loss": 0.10400390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.058342933654785, "learning_rate": 1e-05, "loss": 0.4942, "mean_token_accuracy": 0.8700152635574341, "num_tokens": 265153674.0, "step": 6984 }, { "epoch": 0.8885637959547131, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.148985862731934, "learning_rate": 1e-05, "loss": 0.4803, "mean_token_accuracy": 0.8728413581848145, "num_tokens": 265190168.0, "step": 6985 }, { "epoch": 0.8886910062333037, "ewc_loss": 0.099609375, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.875367641448975, "learning_rate": 1e-05, "loss": 0.4737, "mean_token_accuracy": 0.8724825382232666, "num_tokens": 265230145.0, "step": 6986 }, { "epoch": 0.8888182165118942, "ewc_loss": 0.10009765625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 6.019619941711426, "learning_rate": 1e-05, "loss": 0.5505, "mean_token_accuracy": 0.8489671945571899, "num_tokens": 265263229.0, "step": 6987 }, { "epoch": 0.8889454267904847, "ewc_loss": 0.09423828125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.704112529754639, "learning_rate": 1e-05, "loss": 0.5243, "mean_token_accuracy": 0.8527297377586365, "num_tokens": 265305426.0, "step": 6988 }, { "epoch": 0.8890726370690751, "ewc_loss": 0.099609375, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.97918701171875, "learning_rate": 1e-05, "loss": 0.54, "mean_token_accuracy": 0.8526150584220886, "num_tokens": 265345677.0, "step": 6989 }, { "epoch": 0.8891998473476657, "ewc_loss": 0.09130859375, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.524170398712158, "learning_rate": 1e-05, "loss": 0.4614, "mean_token_accuracy": 0.8693138360977173, "num_tokens": 265383987.0, "step": 6990 }, { "epoch": 0.8893270576262562, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 6.07677698135376, "learning_rate": 1e-05, "loss": 0.5681, "mean_token_accuracy": 0.8431977033615112, "num_tokens": 265423209.0, "step": 6991 }, { "epoch": 0.8894542679048467, "ewc_loss": 0.08740234375, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 7.390975952148438e-05, "grad_norm": 5.344279766082764, "learning_rate": 1e-05, "loss": 0.5275, "mean_token_accuracy": 0.8498632907867432, "num_tokens": 265463304.0, "step": 6992 }, { "epoch": 0.8895814781834372, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.301193714141846, "learning_rate": 1e-05, "loss": 0.5357, "mean_token_accuracy": 0.8507295250892639, "num_tokens": 265506991.0, "step": 6993 }, { "epoch": 0.8897086884620278, "ewc_loss": 0.0859375, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 7.2479248046875e-05, "grad_norm": 5.2148003578186035, "learning_rate": 1e-05, "loss": 0.5052, "mean_token_accuracy": 0.8619014620780945, "num_tokens": 265539050.0, "step": 6994 }, { "epoch": 0.8898358987406182, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.6693501472473145, "learning_rate": 1e-05, "loss": 0.5341, "mean_token_accuracy": 0.8573392629623413, "num_tokens": 265583593.0, "step": 6995 }, { "epoch": 0.8899631090192087, "ewc_loss": 0.08935546875, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 7.581710815429688e-05, "grad_norm": 5.251269817352295, "learning_rate": 1e-05, "loss": 0.5598, "mean_token_accuracy": 0.842799961566925, "num_tokens": 265623873.0, "step": 6996 }, { "epoch": 0.8900903192977992, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 7.052245616912842, "learning_rate": 1e-05, "loss": 0.5456, "mean_token_accuracy": 0.8550280332565308, "num_tokens": 265666191.0, "step": 6997 }, { "epoch": 0.8902175295763898, "ewc_loss": 0.09814453125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.657476902008057, "learning_rate": 1e-05, "loss": 0.5102, "mean_token_accuracy": 0.857581615447998, "num_tokens": 265703929.0, "step": 6998 }, { "epoch": 0.8903447398549803, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 7.030133247375488, "learning_rate": 1e-05, "loss": 0.5332, "mean_token_accuracy": 0.8630986213684082, "num_tokens": 265737246.0, "step": 6999 }, { "epoch": 0.8904719501335708, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.091041564941406, "learning_rate": 1e-05, "loss": 0.5691, "mean_token_accuracy": 0.8511221408843994, "num_tokens": 265774549.0, "step": 7000 }, { "epoch": 0.8905991604121613, "ewc_loss": 0.10986328125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.576272010803223, "learning_rate": 1e-05, "loss": 0.561, "mean_token_accuracy": 0.8539259433746338, "num_tokens": 265806972.0, "step": 7001 }, { "epoch": 0.8907263706907518, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.175145149230957, "learning_rate": 1e-05, "loss": 0.5324, "mean_token_accuracy": 0.8516852855682373, "num_tokens": 265844505.0, "step": 7002 }, { "epoch": 0.8908535809693423, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 6.042163372039795, "learning_rate": 1e-05, "loss": 0.5099, "mean_token_accuracy": 0.8608587980270386, "num_tokens": 265885350.0, "step": 7003 }, { "epoch": 0.8909807912479328, "ewc_loss": 0.10498046875, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.174774646759033, "learning_rate": 1e-05, "loss": 0.4977, "mean_token_accuracy": 0.8643794059753418, "num_tokens": 265925345.0, "step": 7004 }, { "epoch": 0.8911080015265234, "ewc_loss": 0.09130859375, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.626664638519287, "learning_rate": 1e-05, "loss": 0.4951, "mean_token_accuracy": 0.8658708930015564, "num_tokens": 265961329.0, "step": 7005 }, { "epoch": 0.8912352118051139, "ewc_loss": 0.10498046875, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.313213348388672, "learning_rate": 1e-05, "loss": 0.6108, "mean_token_accuracy": 0.833031415939331, "num_tokens": 266000836.0, "step": 7006 }, { "epoch": 0.8913624220837043, "ewc_loss": 0.087890625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 7.43865966796875e-05, "grad_norm": 5.380395412445068, "learning_rate": 1e-05, "loss": 0.5181, "mean_token_accuracy": 0.8553081154823303, "num_tokens": 266037178.0, "step": 7007 }, { "epoch": 0.8914896323622948, "ewc_loss": 0.10986328125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.513402938842773, "learning_rate": 1e-05, "loss": 0.4942, "mean_token_accuracy": 0.867253303527832, "num_tokens": 266077125.0, "step": 7008 }, { "epoch": 0.8916168426408854, "ewc_loss": 0.09033203125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.314873218536377, "learning_rate": 1e-05, "loss": 0.5476, "mean_token_accuracy": 0.8435006141662598, "num_tokens": 266117313.0, "step": 7009 }, { "epoch": 0.8917440529194759, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.785653114318848, "learning_rate": 1e-05, "loss": 0.5278, "mean_token_accuracy": 0.8606650829315186, "num_tokens": 266157432.0, "step": 7010 }, { "epoch": 0.8918712631980664, "ewc_loss": 0.09423828125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.474083423614502, "learning_rate": 1e-05, "loss": 0.5021, "mean_token_accuracy": 0.8593010902404785, "num_tokens": 266197170.0, "step": 7011 }, { "epoch": 0.891998473476657, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.917549133300781, "learning_rate": 1e-05, "loss": 0.5568, "mean_token_accuracy": 0.8521433472633362, "num_tokens": 266230383.0, "step": 7012 }, { "epoch": 0.8921256837552475, "ewc_loss": 0.10009765625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.760947227478027, "learning_rate": 1e-05, "loss": 0.5087, "mean_token_accuracy": 0.8625286817550659, "num_tokens": 266268685.0, "step": 7013 }, { "epoch": 0.8922528940338379, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.763291835784912, "learning_rate": 1e-05, "loss": 0.535, "mean_token_accuracy": 0.8557489514350891, "num_tokens": 266308875.0, "step": 7014 }, { "epoch": 0.8923801043124284, "ewc_loss": 0.10302734375, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 5.919065952301025, "learning_rate": 1e-05, "loss": 0.6019, "mean_token_accuracy": 0.8335767388343811, "num_tokens": 266346540.0, "step": 7015 }, { "epoch": 0.892507314591019, "ewc_loss": 0.10791015625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.4729743003845215, "learning_rate": 1e-05, "loss": 0.5193, "mean_token_accuracy": 0.8601402640342712, "num_tokens": 266383359.0, "step": 7016 }, { "epoch": 0.8926345248696095, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.9133734703063965, "learning_rate": 1e-05, "loss": 0.5392, "mean_token_accuracy": 0.8543219566345215, "num_tokens": 266420235.0, "step": 7017 }, { "epoch": 0.8927617351482, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.202968120574951, "learning_rate": 1e-05, "loss": 0.5271, "mean_token_accuracy": 0.856304407119751, "num_tokens": 266455097.0, "step": 7018 }, { "epoch": 0.8928889454267905, "ewc_loss": 0.0986328125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.81797981262207, "learning_rate": 1e-05, "loss": 0.4985, "mean_token_accuracy": 0.8639791011810303, "num_tokens": 266488471.0, "step": 7019 }, { "epoch": 0.893016155705381, "ewc_loss": 0.09814453125, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 6.0262837409973145, "learning_rate": 1e-05, "loss": 0.5235, "mean_token_accuracy": 0.8549492359161377, "num_tokens": 266528107.0, "step": 7020 }, { "epoch": 0.8931433659839715, "ewc_loss": 0.0947265625, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.69039249420166, "learning_rate": 1e-05, "loss": 0.4856, "mean_token_accuracy": 0.8658894300460815, "num_tokens": 266567859.0, "step": 7021 }, { "epoch": 0.893270576262562, "ewc_loss": 0.099609375, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.983370780944824, "learning_rate": 1e-05, "loss": 0.4904, "mean_token_accuracy": 0.8674834966659546, "num_tokens": 266609722.0, "step": 7022 }, { "epoch": 0.8933977865411525, "ewc_loss": 0.09130859375, "ewc_loss_diag": 0.00013828277587890625, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.551161766052246, "learning_rate": 1e-05, "loss": 0.466, "mean_token_accuracy": 0.8728964328765869, "num_tokens": 266650546.0, "step": 7023 }, { "epoch": 0.8935249968197431, "ewc_loss": 0.10009765625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 6.038191318511963, "learning_rate": 1e-05, "loss": 0.5375, "mean_token_accuracy": 0.8574488162994385, "num_tokens": 266686508.0, "step": 7024 }, { "epoch": 0.8936522070983336, "ewc_loss": 0.08984375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 7.581710815429688e-05, "grad_norm": 5.422330856323242, "learning_rate": 1e-05, "loss": 0.4548, "mean_token_accuracy": 0.8742755651473999, "num_tokens": 266725528.0, "step": 7025 }, { "epoch": 0.893779417376924, "ewc_loss": 0.10302734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.201828479766846, "learning_rate": 1e-05, "loss": 0.4548, "mean_token_accuracy": 0.8753061294555664, "num_tokens": 266754818.0, "step": 7026 }, { "epoch": 0.8939066276555145, "ewc_loss": 0.08935546875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 7.534027099609375e-05, "grad_norm": 5.320644378662109, "learning_rate": 1e-05, "loss": 0.5164, "mean_token_accuracy": 0.8575499653816223, "num_tokens": 266790723.0, "step": 7027 }, { "epoch": 0.8940338379341051, "ewc_loss": 0.109375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.4203338623046875, "learning_rate": 1e-05, "loss": 0.5053, "mean_token_accuracy": 0.8626502752304077, "num_tokens": 266824889.0, "step": 7028 }, { "epoch": 0.8941610482126956, "ewc_loss": 0.08984375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 7.581710815429688e-05, "grad_norm": 5.297011375427246, "learning_rate": 1e-05, "loss": 0.5079, "mean_token_accuracy": 0.8595092296600342, "num_tokens": 266866129.0, "step": 7029 }, { "epoch": 0.8942882584912861, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.660558223724365, "learning_rate": 1e-05, "loss": 0.5408, "mean_token_accuracy": 0.8568108677864075, "num_tokens": 266910005.0, "step": 7030 }, { "epoch": 0.8944154687698767, "ewc_loss": 0.09326171875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.446620464324951, "learning_rate": 1e-05, "loss": 0.4819, "mean_token_accuracy": 0.8678907155990601, "num_tokens": 266947366.0, "step": 7031 }, { "epoch": 0.8945426790484671, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.751651763916016, "learning_rate": 1e-05, "loss": 0.5325, "mean_token_accuracy": 0.8553647994995117, "num_tokens": 266982402.0, "step": 7032 }, { "epoch": 0.8946698893270576, "ewc_loss": 0.09912109375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.722418308258057, "learning_rate": 1e-05, "loss": 0.5572, "mean_token_accuracy": 0.8493289351463318, "num_tokens": 267015397.0, "step": 7033 }, { "epoch": 0.8947970996056481, "ewc_loss": 0.11181640625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.518091201782227, "learning_rate": 1e-05, "loss": 0.4766, "mean_token_accuracy": 0.8732163310050964, "num_tokens": 267050204.0, "step": 7034 }, { "epoch": 0.8949243098842387, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 5.896229267120361, "learning_rate": 1e-05, "loss": 0.4748, "mean_token_accuracy": 0.8749279975891113, "num_tokens": 267085127.0, "step": 7035 }, { "epoch": 0.8950515201628292, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 6.0951642990112305, "learning_rate": 1e-05, "loss": 0.5525, "mean_token_accuracy": 0.8505723476409912, "num_tokens": 267122210.0, "step": 7036 }, { "epoch": 0.8951787304414197, "ewc_loss": 0.10205078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 6.001482963562012, "learning_rate": 1e-05, "loss": 0.5131, "mean_token_accuracy": 0.8583682775497437, "num_tokens": 267161124.0, "step": 7037 }, { "epoch": 0.8953059407200101, "ewc_loss": 0.0927734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.634200096130371, "learning_rate": 1e-05, "loss": 0.4858, "mean_token_accuracy": 0.8703210353851318, "num_tokens": 267195339.0, "step": 7038 }, { "epoch": 0.8954331509986007, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.188253879547119, "learning_rate": 1e-05, "loss": 0.5312, "mean_token_accuracy": 0.8548293709754944, "num_tokens": 267228712.0, "step": 7039 }, { "epoch": 0.8955603612771912, "ewc_loss": 0.0869140625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 7.295608520507812e-05, "grad_norm": 5.282129287719727, "learning_rate": 1e-05, "loss": 0.5252, "mean_token_accuracy": 0.8506174087524414, "num_tokens": 267262796.0, "step": 7040 }, { "epoch": 0.8956875715557817, "ewc_loss": 0.11083984375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.53187894821167, "learning_rate": 1e-05, "loss": 0.5375, "mean_token_accuracy": 0.8532072901725769, "num_tokens": 267304101.0, "step": 7041 }, { "epoch": 0.8958147818343722, "ewc_loss": 0.08740234375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 7.343292236328125e-05, "grad_norm": 5.159506320953369, "learning_rate": 1e-05, "loss": 0.5589, "mean_token_accuracy": 0.8444386720657349, "num_tokens": 267348943.0, "step": 7042 }, { "epoch": 0.8959419921129628, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.955169677734375, "learning_rate": 1e-05, "loss": 0.5196, "mean_token_accuracy": 0.8661165237426758, "num_tokens": 267390436.0, "step": 7043 }, { "epoch": 0.8960692023915532, "ewc_loss": 0.09619140625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.441522121429443, "learning_rate": 1e-05, "loss": 0.4727, "mean_token_accuracy": 0.8718452453613281, "num_tokens": 267428069.0, "step": 7044 }, { "epoch": 0.8961964126701437, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 0.00011157989501953125, "grad_norm": 7.150722980499268, "learning_rate": 1e-05, "loss": 0.5762, "mean_token_accuracy": 0.8465725183486938, "num_tokens": 267468029.0, "step": 7045 }, { "epoch": 0.8963236229487342, "ewc_loss": 0.10693359375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 5.933481693267822, "learning_rate": 1e-05, "loss": 0.5471, "mean_token_accuracy": 0.8530559539794922, "num_tokens": 267504013.0, "step": 7046 }, { "epoch": 0.8964508332273248, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.878231048583984, "learning_rate": 1e-05, "loss": 0.5074, "mean_token_accuracy": 0.8689097166061401, "num_tokens": 267537696.0, "step": 7047 }, { "epoch": 0.8965780435059153, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.118313789367676, "learning_rate": 1e-05, "loss": 0.4978, "mean_token_accuracy": 0.8686987161636353, "num_tokens": 267568971.0, "step": 7048 }, { "epoch": 0.8967052537845058, "ewc_loss": 0.109375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.456833362579346, "learning_rate": 1e-05, "loss": 0.5329, "mean_token_accuracy": 0.8572111129760742, "num_tokens": 267607829.0, "step": 7049 }, { "epoch": 0.8968324640630962, "ewc_loss": 0.10498046875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.023390769958496, "learning_rate": 1e-05, "loss": 0.5092, "mean_token_accuracy": 0.8637818694114685, "num_tokens": 267649338.0, "step": 7050 }, { "epoch": 0.8969596743416868, "ewc_loss": 0.10205078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 6.148226261138916, "learning_rate": 1e-05, "loss": 0.488, "mean_token_accuracy": 0.86603182554245, "num_tokens": 267685053.0, "step": 7051 }, { "epoch": 0.8970868846202773, "ewc_loss": 0.09912109375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.900965690612793, "learning_rate": 1e-05, "loss": 0.4994, "mean_token_accuracy": 0.866887092590332, "num_tokens": 267722291.0, "step": 7052 }, { "epoch": 0.8972140948988678, "ewc_loss": 0.09814453125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.959651947021484, "learning_rate": 1e-05, "loss": 0.5079, "mean_token_accuracy": 0.8614847660064697, "num_tokens": 267760533.0, "step": 7053 }, { "epoch": 0.8973413051774584, "ewc_loss": 0.09521484375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.750344753265381, "learning_rate": 1e-05, "loss": 0.4937, "mean_token_accuracy": 0.8623312711715698, "num_tokens": 267796655.0, "step": 7054 }, { "epoch": 0.8974685154560489, "ewc_loss": 0.0986328125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.9043426513671875, "learning_rate": 1e-05, "loss": 0.5028, "mean_token_accuracy": 0.8640013337135315, "num_tokens": 267835950.0, "step": 7055 }, { "epoch": 0.8975957257346393, "ewc_loss": 0.0927734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.6005659103393555, "learning_rate": 1e-05, "loss": 0.4936, "mean_token_accuracy": 0.8652960658073425, "num_tokens": 267871796.0, "step": 7056 }, { "epoch": 0.8977229360132298, "ewc_loss": 0.099609375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.955001354217529, "learning_rate": 1e-05, "loss": 0.492, "mean_token_accuracy": 0.8643344640731812, "num_tokens": 267910562.0, "step": 7057 }, { "epoch": 0.8978501462918204, "ewc_loss": 0.08935546875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 7.534027099609375e-05, "grad_norm": 5.43531608581543, "learning_rate": 1e-05, "loss": 0.4945, "mean_token_accuracy": 0.8622841835021973, "num_tokens": 267948962.0, "step": 7058 }, { "epoch": 0.8979773565704109, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 6.101724147796631, "learning_rate": 1e-05, "loss": 0.548, "mean_token_accuracy": 0.8535077571868896, "num_tokens": 267993158.0, "step": 7059 }, { "epoch": 0.8981045668490014, "ewc_loss": 0.08740234375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 7.343292236328125e-05, "grad_norm": 5.287309646606445, "learning_rate": 1e-05, "loss": 0.5337, "mean_token_accuracy": 0.8514983654022217, "num_tokens": 268033366.0, "step": 7060 }, { "epoch": 0.898231777127592, "ewc_loss": 0.10693359375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.3849358558654785, "learning_rate": 1e-05, "loss": 0.5171, "mean_token_accuracy": 0.8618505001068115, "num_tokens": 268074125.0, "step": 7061 }, { "epoch": 0.8983589874061825, "ewc_loss": 0.08740234375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 7.343292236328125e-05, "grad_norm": 5.190288066864014, "learning_rate": 1e-05, "loss": 0.5224, "mean_token_accuracy": 0.8564413785934448, "num_tokens": 268112132.0, "step": 7062 }, { "epoch": 0.8984861976847729, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.7724409103393555, "learning_rate": 1e-05, "loss": 0.5988, "mean_token_accuracy": 0.8378874063491821, "num_tokens": 268148161.0, "step": 7063 }, { "epoch": 0.8986134079633634, "ewc_loss": 0.09130859375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.32127046585083, "learning_rate": 1e-05, "loss": 0.4281, "mean_token_accuracy": 0.8848159313201904, "num_tokens": 268184977.0, "step": 7064 }, { "epoch": 0.898740618241954, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 0.00011014938354492188, "grad_norm": 7.079814910888672, "learning_rate": 1e-05, "loss": 0.5421, "mean_token_accuracy": 0.8617144823074341, "num_tokens": 268219355.0, "step": 7065 }, { "epoch": 0.8988678285205445, "ewc_loss": 0.10302734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 5.761880397796631, "learning_rate": 1e-05, "loss": 0.5016, "mean_token_accuracy": 0.863905668258667, "num_tokens": 268260533.0, "step": 7066 }, { "epoch": 0.898995038799135, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.934573173522949, "learning_rate": 1e-05, "loss": 0.5511, "mean_token_accuracy": 0.8552482724189758, "num_tokens": 268294415.0, "step": 7067 }, { "epoch": 0.8991222490777255, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.1038103103637695, "learning_rate": 1e-05, "loss": 0.5002, "mean_token_accuracy": 0.8682065606117249, "num_tokens": 268331038.0, "step": 7068 }, { "epoch": 0.899249459356316, "ewc_loss": 0.10888671875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.446828842163086, "learning_rate": 1e-05, "loss": 0.5713, "mean_token_accuracy": 0.8434699773788452, "num_tokens": 268369438.0, "step": 7069 }, { "epoch": 0.8993766696349065, "ewc_loss": 0.10888671875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.145772933959961, "learning_rate": 1e-05, "loss": 0.5104, "mean_token_accuracy": 0.863227367401123, "num_tokens": 268403808.0, "step": 7070 }, { "epoch": 0.899503879913497, "ewc_loss": 0.09765625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.957016468048096, "learning_rate": 1e-05, "loss": 0.4687, "mean_token_accuracy": 0.8735407590866089, "num_tokens": 268442575.0, "step": 7071 }, { "epoch": 0.8996310901920875, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.159852027893066, "learning_rate": 1e-05, "loss": 0.5227, "mean_token_accuracy": 0.8577148914337158, "num_tokens": 268475387.0, "step": 7072 }, { "epoch": 0.8997583004706781, "ewc_loss": 0.09375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.595879554748535, "learning_rate": 1e-05, "loss": 0.5115, "mean_token_accuracy": 0.8596258163452148, "num_tokens": 268513514.0, "step": 7073 }, { "epoch": 0.8998855107492686, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.260959148406982, "learning_rate": 1e-05, "loss": 0.5035, "mean_token_accuracy": 0.8649431467056274, "num_tokens": 268553690.0, "step": 7074 }, { "epoch": 0.900012721027859, "ewc_loss": 0.08984375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 7.581710815429688e-05, "grad_norm": 5.371311664581299, "learning_rate": 1e-05, "loss": 0.4741, "mean_token_accuracy": 0.868523359298706, "num_tokens": 268593285.0, "step": 7075 }, { "epoch": 0.9001399313064495, "ewc_loss": 0.10888671875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.476840972900391, "learning_rate": 1e-05, "loss": 0.5166, "mean_token_accuracy": 0.8621872067451477, "num_tokens": 268631929.0, "step": 7076 }, { "epoch": 0.9002671415850401, "ewc_loss": 0.08984375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 7.581710815429688e-05, "grad_norm": 5.334262371063232, "learning_rate": 1e-05, "loss": 0.5746, "mean_token_accuracy": 0.832973837852478, "num_tokens": 268669882.0, "step": 7077 }, { "epoch": 0.9003943518636306, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.705153942108154, "learning_rate": 1e-05, "loss": 0.572, "mean_token_accuracy": 0.8465838432312012, "num_tokens": 268712709.0, "step": 7078 }, { "epoch": 0.9005215621422211, "ewc_loss": 0.09375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.447937488555908, "learning_rate": 1e-05, "loss": 0.5025, "mean_token_accuracy": 0.8582154512405396, "num_tokens": 268751514.0, "step": 7079 }, { "epoch": 0.9006487724208116, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.819916725158691, "learning_rate": 1e-05, "loss": 0.5727, "mean_token_accuracy": 0.8463538885116577, "num_tokens": 268796841.0, "step": 7080 }, { "epoch": 0.9007759826994021, "ewc_loss": 0.099609375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.66994571685791, "learning_rate": 1e-05, "loss": 0.4781, "mean_token_accuracy": 0.8701589107513428, "num_tokens": 268841536.0, "step": 7081 }, { "epoch": 0.9009031929779926, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.72788143157959, "learning_rate": 1e-05, "loss": 0.5196, "mean_token_accuracy": 0.8599709272384644, "num_tokens": 268884636.0, "step": 7082 }, { "epoch": 0.9010304032565831, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 5.83505392074585, "learning_rate": 1e-05, "loss": 0.5265, "mean_token_accuracy": 0.8564085960388184, "num_tokens": 268924114.0, "step": 7083 }, { "epoch": 0.9011576135351737, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.478865146636963, "learning_rate": 1e-05, "loss": 0.4987, "mean_token_accuracy": 0.8673739433288574, "num_tokens": 268966476.0, "step": 7084 }, { "epoch": 0.9012848238137642, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 5.870561122894287, "learning_rate": 1e-05, "loss": 0.4907, "mean_token_accuracy": 0.8711077570915222, "num_tokens": 268998224.0, "step": 7085 }, { "epoch": 0.9014120340923547, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 6.210202693939209, "learning_rate": 1e-05, "loss": 0.539, "mean_token_accuracy": 0.8543388247489929, "num_tokens": 269043552.0, "step": 7086 }, { "epoch": 0.9015392443709451, "ewc_loss": 0.0986328125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.813600063323975, "learning_rate": 1e-05, "loss": 0.549, "mean_token_accuracy": 0.8478531837463379, "num_tokens": 269085180.0, "step": 7087 }, { "epoch": 0.9016664546495357, "ewc_loss": 0.10009765625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 6.009773254394531, "learning_rate": 1e-05, "loss": 0.5545, "mean_token_accuracy": 0.8485252261161804, "num_tokens": 269121694.0, "step": 7088 }, { "epoch": 0.9017936649281262, "ewc_loss": 0.09619140625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.72300386428833, "learning_rate": 1e-05, "loss": 0.5757, "mean_token_accuracy": 0.8429083824157715, "num_tokens": 269164896.0, "step": 7089 }, { "epoch": 0.9019208752067167, "ewc_loss": 0.09912109375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.928549766540527, "learning_rate": 1e-05, "loss": 0.4366, "mean_token_accuracy": 0.8811582326889038, "num_tokens": 269201095.0, "step": 7090 }, { "epoch": 0.9020480854853072, "ewc_loss": 0.09326171875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.618648052215576, "learning_rate": 1e-05, "loss": 0.5025, "mean_token_accuracy": 0.862908124923706, "num_tokens": 269243055.0, "step": 7091 }, { "epoch": 0.9021752957638978, "ewc_loss": 0.09912109375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.925290584564209, "learning_rate": 1e-05, "loss": 0.4837, "mean_token_accuracy": 0.8686859011650085, "num_tokens": 269279902.0, "step": 7092 }, { "epoch": 0.9023025060424882, "ewc_loss": 0.09130859375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.501614093780518, "learning_rate": 1e-05, "loss": 0.4624, "mean_token_accuracy": 0.8760182857513428, "num_tokens": 269315773.0, "step": 7093 }, { "epoch": 0.9024297163210787, "ewc_loss": 0.099609375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 6.023801803588867, "learning_rate": 1e-05, "loss": 0.5759, "mean_token_accuracy": 0.8424491286277771, "num_tokens": 269353127.0, "step": 7094 }, { "epoch": 0.9025569265996692, "ewc_loss": 0.08837890625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 7.43865966796875e-05, "grad_norm": 5.373080253601074, "learning_rate": 1e-05, "loss": 0.5018, "mean_token_accuracy": 0.8634207248687744, "num_tokens": 269387023.0, "step": 7095 }, { "epoch": 0.9026841368782598, "ewc_loss": 0.10302734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.20891809463501, "learning_rate": 1e-05, "loss": 0.5857, "mean_token_accuracy": 0.84163498878479, "num_tokens": 269419983.0, "step": 7096 }, { "epoch": 0.9028113471568503, "ewc_loss": 0.087890625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 7.390975952148438e-05, "grad_norm": 5.259555816650391, "learning_rate": 1e-05, "loss": 0.4835, "mean_token_accuracy": 0.8653699159622192, "num_tokens": 269460580.0, "step": 7097 }, { "epoch": 0.9029385574354408, "ewc_loss": 0.10888671875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.486480712890625, "learning_rate": 1e-05, "loss": 0.4699, "mean_token_accuracy": 0.8756487369537354, "num_tokens": 269497711.0, "step": 7098 }, { "epoch": 0.9030657677140312, "ewc_loss": 0.08984375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 7.581710815429688e-05, "grad_norm": 5.230718612670898, "learning_rate": 1e-05, "loss": 0.498, "mean_token_accuracy": 0.8607171773910522, "num_tokens": 269538111.0, "step": 7099 }, { "epoch": 0.9031929779926218, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.810057640075684, "learning_rate": 1e-05, "loss": 0.4693, "mean_token_accuracy": 0.8801939487457275, "num_tokens": 269570695.0, "step": 7100 }, { "epoch": 0.9033201882712123, "ewc_loss": 0.09619140625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.469972133636475, "learning_rate": 1e-05, "loss": 0.4977, "mean_token_accuracy": 0.8634541034698486, "num_tokens": 269605849.0, "step": 7101 }, { "epoch": 0.9034473985498028, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.888221263885498, "learning_rate": 1e-05, "loss": 0.5406, "mean_token_accuracy": 0.855774998664856, "num_tokens": 269646445.0, "step": 7102 }, { "epoch": 0.9035746088283934, "ewc_loss": 0.10400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 5.821357727050781, "learning_rate": 1e-05, "loss": 0.5689, "mean_token_accuracy": 0.8455096483230591, "num_tokens": 269686806.0, "step": 7103 }, { "epoch": 0.9037018191069839, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.592372894287109, "learning_rate": 1e-05, "loss": 0.555, "mean_token_accuracy": 0.8504072427749634, "num_tokens": 269727311.0, "step": 7104 }, { "epoch": 0.9038290293855743, "ewc_loss": 0.10595703125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.009682655334473, "learning_rate": 1e-05, "loss": 0.5921, "mean_token_accuracy": 0.841288149356842, "num_tokens": 269765066.0, "step": 7105 }, { "epoch": 0.9039562396641648, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 6.099395275115967, "learning_rate": 1e-05, "loss": 0.5395, "mean_token_accuracy": 0.8526746034622192, "num_tokens": 269807065.0, "step": 7106 }, { "epoch": 0.9040834499427554, "ewc_loss": 0.10498046875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.076136589050293, "learning_rate": 1e-05, "loss": 0.4705, "mean_token_accuracy": 0.8754000067710876, "num_tokens": 269839874.0, "step": 7107 }, { "epoch": 0.9042106602213459, "ewc_loss": 0.0927734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.643912315368652, "learning_rate": 1e-05, "loss": 0.465, "mean_token_accuracy": 0.869404673576355, "num_tokens": 269877647.0, "step": 7108 }, { "epoch": 0.9043378704999364, "ewc_loss": 0.10595703125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.236292839050293, "learning_rate": 1e-05, "loss": 0.5088, "mean_token_accuracy": 0.8608055710792542, "num_tokens": 269912403.0, "step": 7109 }, { "epoch": 0.9044650807785269, "ewc_loss": 0.08837890625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 7.43865966796875e-05, "grad_norm": 5.3405375480651855, "learning_rate": 1e-05, "loss": 0.4671, "mean_token_accuracy": 0.8678240776062012, "num_tokens": 269946086.0, "step": 7110 }, { "epoch": 0.9045922910571175, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.518105983734131, "learning_rate": 1e-05, "loss": 0.4881, "mean_token_accuracy": 0.87180495262146, "num_tokens": 269985820.0, "step": 7111 }, { "epoch": 0.9047195013357079, "ewc_loss": 0.09033203125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 7.62939453125e-05, "grad_norm": 5.26862907409668, "learning_rate": 1e-05, "loss": 0.4722, "mean_token_accuracy": 0.8691550493240356, "num_tokens": 270021377.0, "step": 7112 }, { "epoch": 0.9048467116142984, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.840439319610596, "learning_rate": 1e-05, "loss": 0.586, "mean_token_accuracy": 0.8468831181526184, "num_tokens": 270056928.0, "step": 7113 }, { "epoch": 0.9049739218928889, "ewc_loss": 0.09619140625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.4664387702941895, "learning_rate": 1e-05, "loss": 0.4743, "mean_token_accuracy": 0.8715777397155762, "num_tokens": 270093720.0, "step": 7114 }, { "epoch": 0.9051011321714795, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.953605651855469, "learning_rate": 1e-05, "loss": 0.5638, "mean_token_accuracy": 0.8523168563842773, "num_tokens": 270140151.0, "step": 7115 }, { "epoch": 0.90522834245007, "ewc_loss": 0.10302734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 5.799165725708008, "learning_rate": 1e-05, "loss": 0.5153, "mean_token_accuracy": 0.8645014762878418, "num_tokens": 270173970.0, "step": 7116 }, { "epoch": 0.9053555527286605, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.767333507537842, "learning_rate": 1e-05, "loss": 0.5144, "mean_token_accuracy": 0.8621208667755127, "num_tokens": 270207014.0, "step": 7117 }, { "epoch": 0.905482763007251, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 5.943317890167236, "learning_rate": 1e-05, "loss": 0.5697, "mean_token_accuracy": 0.8464359641075134, "num_tokens": 270246635.0, "step": 7118 }, { "epoch": 0.9056099732858415, "ewc_loss": 0.109375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.442728042602539, "learning_rate": 1e-05, "loss": 0.5351, "mean_token_accuracy": 0.8581097722053528, "num_tokens": 270289902.0, "step": 7119 }, { "epoch": 0.905737183564432, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 5.919588088989258, "learning_rate": 1e-05, "loss": 0.5484, "mean_token_accuracy": 0.8463559150695801, "num_tokens": 270321173.0, "step": 7120 }, { "epoch": 0.9058643938430225, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 6.161365985870361, "learning_rate": 1e-05, "loss": 0.4781, "mean_token_accuracy": 0.8707614541053772, "num_tokens": 270360281.0, "step": 7121 }, { "epoch": 0.9059916041216131, "ewc_loss": 0.0986328125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.822115898132324, "learning_rate": 1e-05, "loss": 0.4766, "mean_token_accuracy": 0.8692492246627808, "num_tokens": 270396609.0, "step": 7122 }, { "epoch": 0.9061188144002036, "ewc_loss": 0.09912109375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 6.003966808319092, "learning_rate": 1e-05, "loss": 0.5545, "mean_token_accuracy": 0.8486098647117615, "num_tokens": 270436365.0, "step": 7123 }, { "epoch": 0.906246024678794, "ewc_loss": 0.09521484375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.6974639892578125, "learning_rate": 1e-05, "loss": 0.4876, "mean_token_accuracy": 0.8645521402359009, "num_tokens": 270471314.0, "step": 7124 }, { "epoch": 0.9063732349573845, "ewc_loss": 0.10009765625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.961045265197754, "learning_rate": 1e-05, "loss": 0.5473, "mean_token_accuracy": 0.8503844141960144, "num_tokens": 270510169.0, "step": 7125 }, { "epoch": 0.9065004452359751, "ewc_loss": 0.091796875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.545815944671631, "learning_rate": 1e-05, "loss": 0.4975, "mean_token_accuracy": 0.8672577142715454, "num_tokens": 270549787.0, "step": 7126 }, { "epoch": 0.9066276555145656, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 6.0390400886535645, "learning_rate": 1e-05, "loss": 0.4874, "mean_token_accuracy": 0.8682528734207153, "num_tokens": 270584059.0, "step": 7127 }, { "epoch": 0.9067548657931561, "ewc_loss": 0.08935546875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 7.534027099609375e-05, "grad_norm": 5.401214599609375, "learning_rate": 1e-05, "loss": 0.4784, "mean_token_accuracy": 0.8676279783248901, "num_tokens": 270618059.0, "step": 7128 }, { "epoch": 0.9068820760717466, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 6.211874961853027, "learning_rate": 1e-05, "loss": 0.5094, "mean_token_accuracy": 0.8611650466918945, "num_tokens": 270656968.0, "step": 7129 }, { "epoch": 0.9070092863503371, "ewc_loss": 0.0888671875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 7.486343383789062e-05, "grad_norm": 5.287477016448975, "learning_rate": 1e-05, "loss": 0.5058, "mean_token_accuracy": 0.8613089323043823, "num_tokens": 270692088.0, "step": 7130 }, { "epoch": 0.9071364966289276, "ewc_loss": 0.11083984375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.482812881469727, "learning_rate": 1e-05, "loss": 0.512, "mean_token_accuracy": 0.8678523302078247, "num_tokens": 270729480.0, "step": 7131 }, { "epoch": 0.9072637069075181, "ewc_loss": 0.08984375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 7.581710815429688e-05, "grad_norm": 5.2702107429504395, "learning_rate": 1e-05, "loss": 0.4669, "mean_token_accuracy": 0.8686535954475403, "num_tokens": 270769062.0, "step": 7132 }, { "epoch": 0.9073909171861086, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.758118629455566, "learning_rate": 1e-05, "loss": 0.5326, "mean_token_accuracy": 0.8571323156356812, "num_tokens": 270808142.0, "step": 7133 }, { "epoch": 0.9075181274646992, "ewc_loss": 0.09619140625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.475464344024658, "learning_rate": 1e-05, "loss": 0.5161, "mean_token_accuracy": 0.8557933568954468, "num_tokens": 270842232.0, "step": 7134 }, { "epoch": 0.9076453377432897, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.83272123336792, "learning_rate": 1e-05, "loss": 0.5775, "mean_token_accuracy": 0.8487425446510315, "num_tokens": 270881588.0, "step": 7135 }, { "epoch": 0.9077725480218801, "ewc_loss": 0.10205078125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 5.792819976806641, "learning_rate": 1e-05, "loss": 0.5301, "mean_token_accuracy": 0.855333685874939, "num_tokens": 270921122.0, "step": 7136 }, { "epoch": 0.9078997583004706, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.552273273468018, "learning_rate": 1e-05, "loss": 0.5438, "mean_token_accuracy": 0.8558412194252014, "num_tokens": 270959037.0, "step": 7137 }, { "epoch": 0.9080269685790612, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 5.969357013702393, "learning_rate": 1e-05, "loss": 0.5079, "mean_token_accuracy": 0.8628010749816895, "num_tokens": 271002731.0, "step": 7138 }, { "epoch": 0.9081541788576517, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 6.087355613708496, "learning_rate": 1e-05, "loss": 0.5019, "mean_token_accuracy": 0.8628652095794678, "num_tokens": 271038487.0, "step": 7139 }, { "epoch": 0.9082813891362422, "ewc_loss": 0.10400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.072494029998779, "learning_rate": 1e-05, "loss": 0.5197, "mean_token_accuracy": 0.858244776725769, "num_tokens": 271068811.0, "step": 7140 }, { "epoch": 0.9084085994148328, "ewc_loss": 0.0927734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.60971212387085, "learning_rate": 1e-05, "loss": 0.4717, "mean_token_accuracy": 0.872829794883728, "num_tokens": 271111019.0, "step": 7141 }, { "epoch": 0.9085358096934232, "ewc_loss": 0.10693359375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.248953342437744, "learning_rate": 1e-05, "loss": 0.5241, "mean_token_accuracy": 0.8581960201263428, "num_tokens": 271149317.0, "step": 7142 }, { "epoch": 0.9086630199720137, "ewc_loss": 0.08740234375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 7.343292236328125e-05, "grad_norm": 5.286323070526123, "learning_rate": 1e-05, "loss": 0.4353, "mean_token_accuracy": 0.8788650035858154, "num_tokens": 271190664.0, "step": 7143 }, { "epoch": 0.9087902302506042, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.582414150238037, "learning_rate": 1e-05, "loss": 0.5047, "mean_token_accuracy": 0.8655896186828613, "num_tokens": 271225071.0, "step": 7144 }, { "epoch": 0.9089174405291948, "ewc_loss": 0.0888671875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 7.486343383789062e-05, "grad_norm": 5.222429275512695, "learning_rate": 1e-05, "loss": 0.5071, "mean_token_accuracy": 0.8616104125976562, "num_tokens": 271264228.0, "step": 7145 }, { "epoch": 0.9090446508077853, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.965914249420166, "learning_rate": 1e-05, "loss": 0.5291, "mean_token_accuracy": 0.8611707091331482, "num_tokens": 271300888.0, "step": 7146 }, { "epoch": 0.9091718610863758, "ewc_loss": 0.095703125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.51457405090332, "learning_rate": 1e-05, "loss": 0.5122, "mean_token_accuracy": 0.8607649207115173, "num_tokens": 271339629.0, "step": 7147 }, { "epoch": 0.9092990713649662, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 7.064691066741943, "learning_rate": 1e-05, "loss": 0.5357, "mean_token_accuracy": 0.862250566482544, "num_tokens": 271379116.0, "step": 7148 }, { "epoch": 0.9094262816435568, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 5.928955078125, "learning_rate": 1e-05, "loss": 0.5572, "mean_token_accuracy": 0.8494799733161926, "num_tokens": 271420829.0, "step": 7149 }, { "epoch": 0.9095534919221473, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.779837131500244, "learning_rate": 1e-05, "loss": 0.532, "mean_token_accuracy": 0.8586544990539551, "num_tokens": 271457934.0, "step": 7150 }, { "epoch": 0.9096807022007378, "ewc_loss": 0.10693359375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.052692890167236, "learning_rate": 1e-05, "loss": 0.5195, "mean_token_accuracy": 0.8619438409805298, "num_tokens": 271497441.0, "step": 7151 }, { "epoch": 0.9098079124793284, "ewc_loss": 0.10791015625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.414516925811768, "learning_rate": 1e-05, "loss": 0.5253, "mean_token_accuracy": 0.8546453714370728, "num_tokens": 271532358.0, "step": 7152 }, { "epoch": 0.9099351227579189, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 5.978055000305176, "learning_rate": 1e-05, "loss": 0.4983, "mean_token_accuracy": 0.8661261796951294, "num_tokens": 271570103.0, "step": 7153 }, { "epoch": 0.9100623330365093, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 6.120152950286865, "learning_rate": 1e-05, "loss": 0.5085, "mean_token_accuracy": 0.8652752637863159, "num_tokens": 271604542.0, "step": 7154 }, { "epoch": 0.9101895433150998, "ewc_loss": 0.09912109375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.856436252593994, "learning_rate": 1e-05, "loss": 0.5544, "mean_token_accuracy": 0.8509281277656555, "num_tokens": 271646901.0, "step": 7155 }, { "epoch": 0.9103167535936904, "ewc_loss": 0.09814453125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.9800262451171875, "learning_rate": 1e-05, "loss": 0.5636, "mean_token_accuracy": 0.8424303531646729, "num_tokens": 271684922.0, "step": 7156 }, { "epoch": 0.9104439638722809, "ewc_loss": 0.09423828125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.698520660400391, "learning_rate": 1e-05, "loss": 0.4504, "mean_token_accuracy": 0.8782039880752563, "num_tokens": 271721206.0, "step": 7157 }, { "epoch": 0.9105711741508714, "ewc_loss": 0.09912109375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.964245319366455, "learning_rate": 1e-05, "loss": 0.4775, "mean_token_accuracy": 0.8733363151550293, "num_tokens": 271757978.0, "step": 7158 }, { "epoch": 0.9106983844294619, "ewc_loss": 0.091796875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.5502753257751465, "learning_rate": 1e-05, "loss": 0.594, "mean_token_accuracy": 0.8351255655288696, "num_tokens": 271792208.0, "step": 7159 }, { "epoch": 0.9108255947080524, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 6.05702018737793, "learning_rate": 1e-05, "loss": 0.544, "mean_token_accuracy": 0.8553309440612793, "num_tokens": 271828853.0, "step": 7160 }, { "epoch": 0.9109528049866429, "ewc_loss": 0.08935546875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 7.534027099609375e-05, "grad_norm": 5.381109237670898, "learning_rate": 1e-05, "loss": 0.5384, "mean_token_accuracy": 0.8457807302474976, "num_tokens": 271868596.0, "step": 7161 }, { "epoch": 0.9110800152652334, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.264468193054199, "learning_rate": 1e-05, "loss": 0.5045, "mean_token_accuracy": 0.8669416308403015, "num_tokens": 271911292.0, "step": 7162 }, { "epoch": 0.9112072255438239, "ewc_loss": 0.087890625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 7.390975952148438e-05, "grad_norm": 5.249863147735596, "learning_rate": 1e-05, "loss": 0.4869, "mean_token_accuracy": 0.8686375617980957, "num_tokens": 271954512.0, "step": 7163 }, { "epoch": 0.9113344358224145, "ewc_loss": 0.11181640625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.595322608947754, "learning_rate": 1e-05, "loss": 0.534, "mean_token_accuracy": 0.8554203510284424, "num_tokens": 271994229.0, "step": 7164 }, { "epoch": 0.911461646101005, "ewc_loss": 0.09033203125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 7.62939453125e-05, "grad_norm": 5.2762532234191895, "learning_rate": 1e-05, "loss": 0.4826, "mean_token_accuracy": 0.8671289682388306, "num_tokens": 272031712.0, "step": 7165 }, { "epoch": 0.9115888563795955, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.931568145751953, "learning_rate": 1e-05, "loss": 0.5485, "mean_token_accuracy": 0.8565074801445007, "num_tokens": 272070269.0, "step": 7166 }, { "epoch": 0.9117160666581859, "ewc_loss": 0.0966796875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.585494041442871, "learning_rate": 1e-05, "loss": 0.5117, "mean_token_accuracy": 0.8597463965415955, "num_tokens": 272115407.0, "step": 7167 }, { "epoch": 0.9118432769367765, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.942400932312012, "learning_rate": 1e-05, "loss": 0.5384, "mean_token_accuracy": 0.8609710931777954, "num_tokens": 272161463.0, "step": 7168 }, { "epoch": 0.911970487215367, "ewc_loss": 0.10595703125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 5.962118625640869, "learning_rate": 1e-05, "loss": 0.5074, "mean_token_accuracy": 0.8658625483512878, "num_tokens": 272196712.0, "step": 7169 }, { "epoch": 0.9120976974939575, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.559256076812744, "learning_rate": 1e-05, "loss": 0.5474, "mean_token_accuracy": 0.8560422658920288, "num_tokens": 272235617.0, "step": 7170 }, { "epoch": 0.912224907772548, "ewc_loss": 0.10693359375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.094770431518555, "learning_rate": 1e-05, "loss": 0.503, "mean_token_accuracy": 0.8652549982070923, "num_tokens": 272266464.0, "step": 7171 }, { "epoch": 0.9123521180511386, "ewc_loss": 0.099609375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 6.0538458824157715, "learning_rate": 1e-05, "loss": 0.5509, "mean_token_accuracy": 0.8480365872383118, "num_tokens": 272301069.0, "step": 7172 }, { "epoch": 0.912479328329729, "ewc_loss": 0.10595703125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.147151470184326, "learning_rate": 1e-05, "loss": 0.5758, "mean_token_accuracy": 0.8440525531768799, "num_tokens": 272333119.0, "step": 7173 }, { "epoch": 0.9126065386083195, "ewc_loss": 0.091796875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.610690593719482, "learning_rate": 1e-05, "loss": 0.5397, "mean_token_accuracy": 0.8488917350769043, "num_tokens": 272365560.0, "step": 7174 }, { "epoch": 0.9127337488869101, "ewc_loss": 0.10693359375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.303653717041016, "learning_rate": 1e-05, "loss": 0.4547, "mean_token_accuracy": 0.8787554502487183, "num_tokens": 272401290.0, "step": 7175 }, { "epoch": 0.9128609591655006, "ewc_loss": 0.08837890625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 7.43865966796875e-05, "grad_norm": 5.324671745300293, "learning_rate": 1e-05, "loss": 0.4911, "mean_token_accuracy": 0.8652423620223999, "num_tokens": 272442925.0, "step": 7176 }, { "epoch": 0.9129881694440911, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.594324588775635, "learning_rate": 1e-05, "loss": 0.5469, "mean_token_accuracy": 0.854157567024231, "num_tokens": 272477192.0, "step": 7177 }, { "epoch": 0.9131153797226816, "ewc_loss": 0.0908203125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.3172502517700195, "learning_rate": 1e-05, "loss": 0.5561, "mean_token_accuracy": 0.8445180654525757, "num_tokens": 272506987.0, "step": 7178 }, { "epoch": 0.9132425900012721, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.868333339691162, "learning_rate": 1e-05, "loss": 0.5163, "mean_token_accuracy": 0.8665593266487122, "num_tokens": 272544399.0, "step": 7179 }, { "epoch": 0.9133698002798626, "ewc_loss": 0.095703125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.537217140197754, "learning_rate": 1e-05, "loss": 0.5039, "mean_token_accuracy": 0.8596338033676147, "num_tokens": 272577639.0, "step": 7180 }, { "epoch": 0.9134970105584531, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.927967071533203, "learning_rate": 1e-05, "loss": 0.516, "mean_token_accuracy": 0.8660369515419006, "num_tokens": 272615358.0, "step": 7181 }, { "epoch": 0.9136242208370436, "ewc_loss": 0.10400390625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 5.838214874267578, "learning_rate": 1e-05, "loss": 0.5275, "mean_token_accuracy": 0.8589761257171631, "num_tokens": 272652927.0, "step": 7182 }, { "epoch": 0.9137514311156342, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.716424465179443, "learning_rate": 1e-05, "loss": 0.5466, "mean_token_accuracy": 0.8554764986038208, "num_tokens": 272682509.0, "step": 7183 }, { "epoch": 0.9138786413942247, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 5.953466415405273, "learning_rate": 1e-05, "loss": 0.5669, "mean_token_accuracy": 0.8454990386962891, "num_tokens": 272723295.0, "step": 7184 }, { "epoch": 0.9140058516728151, "ewc_loss": 0.10791015625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.39066219329834, "learning_rate": 1e-05, "loss": 0.5497, "mean_token_accuracy": 0.8470818996429443, "num_tokens": 272764985.0, "step": 7185 }, { "epoch": 0.9141330619514056, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.911862373352051, "learning_rate": 1e-05, "loss": 0.5074, "mean_token_accuracy": 0.8630411624908447, "num_tokens": 272807926.0, "step": 7186 }, { "epoch": 0.9142602722299962, "ewc_loss": 0.10107421875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 6.131865978240967, "learning_rate": 1e-05, "loss": 0.4897, "mean_token_accuracy": 0.8650786280632019, "num_tokens": 272844249.0, "step": 7187 }, { "epoch": 0.9143874825085867, "ewc_loss": 0.0986328125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.825005531311035, "learning_rate": 1e-05, "loss": 0.5161, "mean_token_accuracy": 0.8587099313735962, "num_tokens": 272878029.0, "step": 7188 }, { "epoch": 0.9145146927871772, "ewc_loss": 0.0986328125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.9889655113220215, "learning_rate": 1e-05, "loss": 0.5478, "mean_token_accuracy": 0.8481377363204956, "num_tokens": 272913185.0, "step": 7189 }, { "epoch": 0.9146419030657678, "ewc_loss": 0.0947265625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.693713665008545, "learning_rate": 1e-05, "loss": 0.4956, "mean_token_accuracy": 0.8654533624649048, "num_tokens": 272946835.0, "step": 7190 }, { "epoch": 0.9147691133443582, "ewc_loss": 0.09912109375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.9520039558410645, "learning_rate": 1e-05, "loss": 0.5335, "mean_token_accuracy": 0.8537205457687378, "num_tokens": 272983467.0, "step": 7191 }, { "epoch": 0.9148963236229487, "ewc_loss": 0.09228515625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.5550007820129395, "learning_rate": 1e-05, "loss": 0.5003, "mean_token_accuracy": 0.8613279461860657, "num_tokens": 273022464.0, "step": 7192 }, { "epoch": 0.9150235339015392, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 6.046330451965332, "learning_rate": 1e-05, "loss": 0.5639, "mean_token_accuracy": 0.84363853931427, "num_tokens": 273059746.0, "step": 7193 }, { "epoch": 0.9151507441801298, "ewc_loss": 0.09033203125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 7.62939453125e-05, "grad_norm": 5.407613277435303, "learning_rate": 1e-05, "loss": 0.4747, "mean_token_accuracy": 0.8699138164520264, "num_tokens": 273100815.0, "step": 7194 }, { "epoch": 0.9152779544587203, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 6.193005084991455, "learning_rate": 1e-05, "loss": 0.5475, "mean_token_accuracy": 0.8497785329818726, "num_tokens": 273139429.0, "step": 7195 }, { "epoch": 0.9154051647373108, "ewc_loss": 0.08935546875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 7.534027099609375e-05, "grad_norm": 5.282475471496582, "learning_rate": 1e-05, "loss": 0.4872, "mean_token_accuracy": 0.8676630258560181, "num_tokens": 273174159.0, "step": 7196 }, { "epoch": 0.9155323750159012, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.472778797149658, "learning_rate": 1e-05, "loss": 0.4843, "mean_token_accuracy": 0.868803858757019, "num_tokens": 273211434.0, "step": 7197 }, { "epoch": 0.9156595852944918, "ewc_loss": 0.0908203125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.2571539878845215, "learning_rate": 1e-05, "loss": 0.5186, "mean_token_accuracy": 0.853396475315094, "num_tokens": 273243463.0, "step": 7198 }, { "epoch": 0.9157867955730823, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.77957010269165, "learning_rate": 1e-05, "loss": 0.5157, "mean_token_accuracy": 0.8639698028564453, "num_tokens": 273283515.0, "step": 7199 }, { "epoch": 0.9159140058516728, "ewc_loss": 0.09619140625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.528860569000244, "learning_rate": 1e-05, "loss": 0.5394, "mean_token_accuracy": 0.853385329246521, "num_tokens": 273322927.0, "step": 7200 }, { "epoch": 0.9160412161302633, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.7955803871154785, "learning_rate": 1e-05, "loss": 0.5169, "mean_token_accuracy": 0.8692136406898499, "num_tokens": 273359157.0, "step": 7201 }, { "epoch": 0.9161684264088539, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 5.743783473968506, "learning_rate": 1e-05, "loss": 0.5099, "mean_token_accuracy": 0.8648632764816284, "num_tokens": 273394385.0, "step": 7202 }, { "epoch": 0.9162956366874443, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.557315826416016, "learning_rate": 1e-05, "loss": 0.5321, "mean_token_accuracy": 0.8528696298599243, "num_tokens": 273430286.0, "step": 7203 }, { "epoch": 0.9164228469660348, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.000141143798828125, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 5.902768135070801, "learning_rate": 1e-05, "loss": 0.4852, "mean_token_accuracy": 0.8691664338111877, "num_tokens": 273467821.0, "step": 7204 }, { "epoch": 0.9165500572446253, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.000141143798828125, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.160543441772461, "learning_rate": 1e-05, "loss": 0.5411, "mean_token_accuracy": 0.855211615562439, "num_tokens": 273507659.0, "step": 7205 }, { "epoch": 0.9166772675232159, "ewc_loss": 0.10498046875, "ewc_loss_diag": 0.000141143798828125, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 5.9864606857299805, "learning_rate": 1e-05, "loss": 0.5613, "mean_token_accuracy": 0.847579836845398, "num_tokens": 273546445.0, "step": 7206 }, { "epoch": 0.9168044778018064, "ewc_loss": 0.0947265625, "ewc_loss_diag": 0.000141143798828125, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.7371296882629395, "learning_rate": 1e-05, "loss": 0.5179, "mean_token_accuracy": 0.8555947542190552, "num_tokens": 273580354.0, "step": 7207 }, { "epoch": 0.9169316880803969, "ewc_loss": 0.10400390625, "ewc_loss_diag": 0.000141143798828125, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.110812664031982, "learning_rate": 1e-05, "loss": 0.5266, "mean_token_accuracy": 0.858462929725647, "num_tokens": 273619194.0, "step": 7208 }, { "epoch": 0.9170588983589874, "ewc_loss": 0.08935546875, "ewc_loss_diag": 0.000141143798828125, "ewc_loss_parallel": 7.534027099609375e-05, "grad_norm": 5.391439437866211, "learning_rate": 1e-05, "loss": 0.487, "mean_token_accuracy": 0.866284966468811, "num_tokens": 273662888.0, "step": 7209 }, { "epoch": 0.9171861086375779, "ewc_loss": 0.10888671875, "ewc_loss_diag": 0.000141143798828125, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.371367454528809, "learning_rate": 1e-05, "loss": 0.5046, "mean_token_accuracy": 0.8650710582733154, "num_tokens": 273697823.0, "step": 7210 }, { "epoch": 0.9173133189161684, "ewc_loss": 0.087890625, "ewc_loss_diag": 0.000141143798828125, "ewc_loss_parallel": 7.390975952148438e-05, "grad_norm": 5.2043609619140625, "learning_rate": 1e-05, "loss": 0.5421, "mean_token_accuracy": 0.8487085700035095, "num_tokens": 273737541.0, "step": 7211 }, { "epoch": 0.9174405291947589, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.000141143798828125, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.7383012771606445, "learning_rate": 1e-05, "loss": 0.5532, "mean_token_accuracy": 0.8534551858901978, "num_tokens": 273777565.0, "step": 7212 }, { "epoch": 0.9175677394733495, "ewc_loss": 0.091796875, "ewc_loss_diag": 0.000141143798828125, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.288575172424316, "learning_rate": 1e-05, "loss": 0.5423, "mean_token_accuracy": 0.84852135181427, "num_tokens": 273814269.0, "step": 7213 }, { "epoch": 0.91769494975194, "ewc_loss": 0.125, "ewc_loss_diag": 0.000141143798828125, "ewc_loss_parallel": 0.00011110305786132812, "grad_norm": 7.049253940582275, "learning_rate": 1e-05, "loss": 0.5417, "mean_token_accuracy": 0.8608676195144653, "num_tokens": 273855828.0, "step": 7214 }, { "epoch": 0.9178221600305305, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.000141143798828125, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.676932334899902, "learning_rate": 1e-05, "loss": 0.5212, "mean_token_accuracy": 0.8588677048683167, "num_tokens": 273887514.0, "step": 7215 }, { "epoch": 0.9179493703091209, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.000141143798828125, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 7.005456924438477, "learning_rate": 1e-05, "loss": 0.5618, "mean_token_accuracy": 0.8509272336959839, "num_tokens": 273923134.0, "step": 7216 }, { "epoch": 0.9180765805877115, "ewc_loss": 0.10888671875, "ewc_loss_diag": 0.000141143798828125, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 5.992919445037842, "learning_rate": 1e-05, "loss": 0.4976, "mean_token_accuracy": 0.863876461982727, "num_tokens": 273962089.0, "step": 7217 }, { "epoch": 0.918203790866302, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.000141143798828125, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.668148040771484, "learning_rate": 1e-05, "loss": 0.5222, "mean_token_accuracy": 0.8619211912155151, "num_tokens": 273997150.0, "step": 7218 }, { "epoch": 0.9183310011448925, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.000141143798828125, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.040196418762207, "learning_rate": 1e-05, "loss": 0.4686, "mean_token_accuracy": 0.8771929144859314, "num_tokens": 274032144.0, "step": 7219 }, { "epoch": 0.918458211423483, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.000141143798828125, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.320063591003418, "learning_rate": 1e-05, "loss": 0.4697, "mean_token_accuracy": 0.8754007816314697, "num_tokens": 274063344.0, "step": 7220 }, { "epoch": 0.9185854217020736, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.000141143798828125, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 5.962861061096191, "learning_rate": 1e-05, "loss": 0.5053, "mean_token_accuracy": 0.8659654259681702, "num_tokens": 274102242.0, "step": 7221 }, { "epoch": 0.918712631980664, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.000141143798828125, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 6.042010307312012, "learning_rate": 1e-05, "loss": 0.5128, "mean_token_accuracy": 0.8607034087181091, "num_tokens": 274146560.0, "step": 7222 }, { "epoch": 0.9188398422592545, "ewc_loss": 0.09912109375, "ewc_loss_diag": 0.00014209747314453125, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.864573955535889, "learning_rate": 1e-05, "loss": 0.5482, "mean_token_accuracy": 0.8479442596435547, "num_tokens": 274180793.0, "step": 7223 }, { "epoch": 0.918967052537845, "ewc_loss": 0.09765625, "ewc_loss_diag": 0.00014209747314453125, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.901933193206787, "learning_rate": 1e-05, "loss": 0.5045, "mean_token_accuracy": 0.8636141419410706, "num_tokens": 274221207.0, "step": 7224 }, { "epoch": 0.9190942628164356, "ewc_loss": 0.0966796875, "ewc_loss_diag": 0.00014209747314453125, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.723717212677002, "learning_rate": 1e-05, "loss": 0.5328, "mean_token_accuracy": 0.8535504937171936, "num_tokens": 274265413.0, "step": 7225 }, { "epoch": 0.9192214730950261, "ewc_loss": 0.09814453125, "ewc_loss_diag": 0.00014209747314453125, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.880993843078613, "learning_rate": 1e-05, "loss": 0.5892, "mean_token_accuracy": 0.8360258936882019, "num_tokens": 274305875.0, "step": 7226 }, { "epoch": 0.9193486833736166, "ewc_loss": 0.09326171875, "ewc_loss_diag": 0.00014209747314453125, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.59697961807251, "learning_rate": 1e-05, "loss": 0.5079, "mean_token_accuracy": 0.8554338216781616, "num_tokens": 274344223.0, "step": 7227 }, { "epoch": 0.919475893652207, "ewc_loss": 0.09912109375, "ewc_loss_diag": 0.00014209747314453125, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.921303749084473, "learning_rate": 1e-05, "loss": 0.5315, "mean_token_accuracy": 0.8574784398078918, "num_tokens": 274382671.0, "step": 7228 }, { "epoch": 0.9196031039307976, "ewc_loss": 0.08984375, "ewc_loss_diag": 0.00014209747314453125, "ewc_loss_parallel": 7.581710815429688e-05, "grad_norm": 5.467365741729736, "learning_rate": 1e-05, "loss": 0.5006, "mean_token_accuracy": 0.8596833944320679, "num_tokens": 274413475.0, "step": 7229 }, { "epoch": 0.9197303142093881, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.00014209747314453125, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 6.044254779815674, "learning_rate": 1e-05, "loss": 0.4616, "mean_token_accuracy": 0.8768625259399414, "num_tokens": 274453714.0, "step": 7230 }, { "epoch": 0.9198575244879786, "ewc_loss": 0.08740234375, "ewc_loss_diag": 0.00014209747314453125, "ewc_loss_parallel": 7.343292236328125e-05, "grad_norm": 5.317491054534912, "learning_rate": 1e-05, "loss": 0.5183, "mean_token_accuracy": 0.8525581359863281, "num_tokens": 274490388.0, "step": 7231 }, { "epoch": 0.9199847347665692, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.00014209747314453125, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.285304069519043, "learning_rate": 1e-05, "loss": 0.5031, "mean_token_accuracy": 0.864934504032135, "num_tokens": 274529939.0, "step": 7232 }, { "epoch": 0.9201119450451597, "ewc_loss": 0.0869140625, "ewc_loss_diag": 0.00014209747314453125, "ewc_loss_parallel": 7.295608520507812e-05, "grad_norm": 5.188528537750244, "learning_rate": 1e-05, "loss": 0.5222, "mean_token_accuracy": 0.8580889701843262, "num_tokens": 274570967.0, "step": 7233 }, { "epoch": 0.9202391553237501, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.00014209747314453125, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.669942855834961, "learning_rate": 1e-05, "loss": 0.5283, "mean_token_accuracy": 0.8638889789581299, "num_tokens": 274601977.0, "step": 7234 }, { "epoch": 0.9203663656023406, "ewc_loss": 0.09033203125, "ewc_loss_diag": 0.00014209747314453125, "ewc_loss_parallel": 7.62939453125e-05, "grad_norm": 5.227808952331543, "learning_rate": 1e-05, "loss": 0.4903, "mean_token_accuracy": 0.8663793802261353, "num_tokens": 274638509.0, "step": 7235 }, { "epoch": 0.9204935758809312, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.00014209747314453125, "ewc_loss_parallel": 0.00011014938354492188, "grad_norm": 7.048797607421875, "learning_rate": 1e-05, "loss": 0.4616, "mean_token_accuracy": 0.8811632990837097, "num_tokens": 274674325.0, "step": 7236 }, { "epoch": 0.9206207861595217, "ewc_loss": 0.099609375, "ewc_loss_diag": 0.00014209747314453125, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.63171911239624, "learning_rate": 1e-05, "loss": 0.5066, "mean_token_accuracy": 0.8610630631446838, "num_tokens": 274709686.0, "step": 7237 }, { "epoch": 0.9207479964381122, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.00014209747314453125, "ewc_loss_parallel": 0.00010967254638671875, "grad_norm": 7.045494079589844, "learning_rate": 1e-05, "loss": 0.5625, "mean_token_accuracy": 0.8555777072906494, "num_tokens": 274748917.0, "step": 7238 }, { "epoch": 0.9208752067167028, "ewc_loss": 0.10888671875, "ewc_loss_diag": 0.00014209747314453125, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.0187506675720215, "learning_rate": 1e-05, "loss": 0.5553, "mean_token_accuracy": 0.8489388823509216, "num_tokens": 274794119.0, "step": 7239 }, { "epoch": 0.9210024169952932, "ewc_loss": 0.11376953125, "ewc_loss_diag": 0.00014209747314453125, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.663686752319336, "learning_rate": 1e-05, "loss": 0.5551, "mean_token_accuracy": 0.8508298993110657, "num_tokens": 274827907.0, "step": 7240 }, { "epoch": 0.9211296272738837, "ewc_loss": 0.109375, "ewc_loss_diag": 0.00014209747314453125, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.122922897338867, "learning_rate": 1e-05, "loss": 0.4494, "mean_token_accuracy": 0.8804866075515747, "num_tokens": 274865821.0, "step": 7241 }, { "epoch": 0.9212568375524742, "ewc_loss": 0.10400390625, "ewc_loss_diag": 0.00014209747314453125, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.187011241912842, "learning_rate": 1e-05, "loss": 0.5156, "mean_token_accuracy": 0.8633500933647156, "num_tokens": 274903550.0, "step": 7242 }, { "epoch": 0.9213840478310648, "ewc_loss": 0.10498046875, "ewc_loss_diag": 0.00014209747314453125, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.104883670806885, "learning_rate": 1e-05, "loss": 0.5053, "mean_token_accuracy": 0.8642522692680359, "num_tokens": 274944212.0, "step": 7243 }, { "epoch": 0.9215112581096553, "ewc_loss": 0.09521484375, "ewc_loss_diag": 0.00014209747314453125, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.786866664886475, "learning_rate": 1e-05, "loss": 0.4901, "mean_token_accuracy": 0.8622088432312012, "num_tokens": 274980175.0, "step": 7244 }, { "epoch": 0.9216384683882458, "ewc_loss": 0.10400390625, "ewc_loss_diag": 0.00014209747314453125, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.148281097412109, "learning_rate": 1e-05, "loss": 0.5269, "mean_token_accuracy": 0.8546319007873535, "num_tokens": 275020018.0, "step": 7245 }, { "epoch": 0.9217656786668362, "ewc_loss": 0.0908203125, "ewc_loss_diag": 0.00014209747314453125, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.514622211456299, "learning_rate": 1e-05, "loss": 0.5175, "mean_token_accuracy": 0.8559014201164246, "num_tokens": 275063538.0, "step": 7246 }, { "epoch": 0.9218928889454268, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.00014209747314453125, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.2729339599609375, "learning_rate": 1e-05, "loss": 0.5161, "mean_token_accuracy": 0.8568184971809387, "num_tokens": 275099884.0, "step": 7247 }, { "epoch": 0.9220200992240173, "ewc_loss": 0.08935546875, "ewc_loss_diag": 0.00014209747314453125, "ewc_loss_parallel": 7.534027099609375e-05, "grad_norm": 5.355233669281006, "learning_rate": 1e-05, "loss": 0.5195, "mean_token_accuracy": 0.8552061319351196, "num_tokens": 275140292.0, "step": 7248 }, { "epoch": 0.9221473095026078, "ewc_loss": 0.10888671875, "ewc_loss_diag": 0.00014209747314453125, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.4863057136535645, "learning_rate": 1e-05, "loss": 0.5202, "mean_token_accuracy": 0.8626114130020142, "num_tokens": 275178480.0, "step": 7249 }, { "epoch": 0.9222745197811983, "ewc_loss": 0.09130859375, "ewc_loss_diag": 0.00014209747314453125, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.3152265548706055, "learning_rate": 1e-05, "loss": 0.5229, "mean_token_accuracy": 0.8541022539138794, "num_tokens": 275215855.0, "step": 7250 }, { "epoch": 0.9224017300597889, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.00014209747314453125, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.7254180908203125, "learning_rate": 1e-05, "loss": 0.5805, "mean_token_accuracy": 0.8458994030952454, "num_tokens": 275254895.0, "step": 7251 }, { "epoch": 0.9225289403383793, "ewc_loss": 0.09521484375, "ewc_loss_diag": 0.00014209747314453125, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.447005748748779, "learning_rate": 1e-05, "loss": 0.5081, "mean_token_accuracy": 0.8600113391876221, "num_tokens": 275292352.0, "step": 7252 }, { "epoch": 0.9226561506169698, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.00014209747314453125, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.8457512855529785, "learning_rate": 1e-05, "loss": 0.565, "mean_token_accuracy": 0.8537107110023499, "num_tokens": 275335867.0, "step": 7253 }, { "epoch": 0.9227833608955603, "ewc_loss": 0.09912109375, "ewc_loss_diag": 0.00014209747314453125, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.67902135848999, "learning_rate": 1e-05, "loss": 0.5414, "mean_token_accuracy": 0.8534901142120361, "num_tokens": 275376551.0, "step": 7254 }, { "epoch": 0.9229105711741509, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.00014209747314453125, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.7704997062683105, "learning_rate": 1e-05, "loss": 0.5435, "mean_token_accuracy": 0.8540078401565552, "num_tokens": 275412568.0, "step": 7255 }, { "epoch": 0.9230377814527414, "ewc_loss": 0.10205078125, "ewc_loss_diag": 0.00014209747314453125, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 5.850581169128418, "learning_rate": 1e-05, "loss": 0.5172, "mean_token_accuracy": 0.860286295413971, "num_tokens": 275450534.0, "step": 7256 }, { "epoch": 0.9231649917313319, "ewc_loss": 0.11083984375, "ewc_loss_diag": 0.00014209747314453125, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.508084774017334, "learning_rate": 1e-05, "loss": 0.5413, "mean_token_accuracy": 0.8555153012275696, "num_tokens": 275493165.0, "step": 7257 }, { "epoch": 0.9232922020099223, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.00014209747314453125, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 5.882813453674316, "learning_rate": 1e-05, "loss": 0.5167, "mean_token_accuracy": 0.8615847826004028, "num_tokens": 275533935.0, "step": 7258 }, { "epoch": 0.9234194122885129, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.00014209747314453125, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 6.228401184082031, "learning_rate": 1e-05, "loss": 0.5329, "mean_token_accuracy": 0.8554274439811707, "num_tokens": 275570937.0, "step": 7259 }, { "epoch": 0.9235466225671034, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.00014209747314453125, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.851452827453613, "learning_rate": 1e-05, "loss": 0.4948, "mean_token_accuracy": 0.8656718134880066, "num_tokens": 275606036.0, "step": 7260 }, { "epoch": 0.9236738328456939, "ewc_loss": 0.099609375, "ewc_loss_diag": 0.00014209747314453125, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.991973876953125, "learning_rate": 1e-05, "loss": 0.5077, "mean_token_accuracy": 0.863286018371582, "num_tokens": 275645849.0, "step": 7261 }, { "epoch": 0.9238010431242845, "ewc_loss": 0.0986328125, "ewc_loss_diag": 0.00014209747314453125, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.787369251251221, "learning_rate": 1e-05, "loss": 0.4709, "mean_token_accuracy": 0.8727430105209351, "num_tokens": 275682192.0, "step": 7262 }, { "epoch": 0.923928253402875, "ewc_loss": 0.09716796875, "ewc_loss_diag": 0.00014209747314453125, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.851058006286621, "learning_rate": 1e-05, "loss": 0.5333, "mean_token_accuracy": 0.8518545031547546, "num_tokens": 275722074.0, "step": 7263 }, { "epoch": 0.9240554636814655, "ewc_loss": 0.095703125, "ewc_loss_diag": 0.00014209747314453125, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.712710380554199, "learning_rate": 1e-05, "loss": 0.5239, "mean_token_accuracy": 0.8538537621498108, "num_tokens": 275764966.0, "step": 7264 }, { "epoch": 0.9241826739600559, "ewc_loss": 0.095703125, "ewc_loss_diag": 0.00014209747314453125, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.794991970062256, "learning_rate": 1e-05, "loss": 0.4702, "mean_token_accuracy": 0.8725409507751465, "num_tokens": 275805085.0, "step": 7265 }, { "epoch": 0.9243098842386465, "ewc_loss": 0.09423828125, "ewc_loss_diag": 0.00014209747314453125, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.630069732666016, "learning_rate": 1e-05, "loss": 0.5039, "mean_token_accuracy": 0.865167498588562, "num_tokens": 275837533.0, "step": 7266 }, { "epoch": 0.924437094517237, "ewc_loss": 0.0966796875, "ewc_loss_diag": 0.00014209747314453125, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.814614295959473, "learning_rate": 1e-05, "loss": 0.5144, "mean_token_accuracy": 0.8601430654525757, "num_tokens": 275877168.0, "step": 7267 }, { "epoch": 0.9245643047958275, "ewc_loss": 0.0927734375, "ewc_loss_diag": 0.00014209747314453125, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.524264335632324, "learning_rate": 1e-05, "loss": 0.5263, "mean_token_accuracy": 0.8512974977493286, "num_tokens": 275917861.0, "step": 7268 }, { "epoch": 0.924691515074418, "ewc_loss": 0.09765625, "ewc_loss_diag": 0.00014209747314453125, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.896541118621826, "learning_rate": 1e-05, "loss": 0.5367, "mean_token_accuracy": 0.8482521772384644, "num_tokens": 275952433.0, "step": 7269 }, { "epoch": 0.9248187253530086, "ewc_loss": 0.08935546875, "ewc_loss_diag": 0.00014209747314453125, "ewc_loss_parallel": 7.534027099609375e-05, "grad_norm": 5.399897575378418, "learning_rate": 1e-05, "loss": 0.4939, "mean_token_accuracy": 0.8614476919174194, "num_tokens": 275993124.0, "step": 7270 }, { "epoch": 0.924945935631599, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.00014209747314453125, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 6.0612311363220215, "learning_rate": 1e-05, "loss": 0.4896, "mean_token_accuracy": 0.8673406839370728, "num_tokens": 276032967.0, "step": 7271 }, { "epoch": 0.9250731459101895, "ewc_loss": 0.08740234375, "ewc_loss_diag": 0.00014209747314453125, "ewc_loss_parallel": 7.343292236328125e-05, "grad_norm": 5.247556686401367, "learning_rate": 1e-05, "loss": 0.4835, "mean_token_accuracy": 0.8642510175704956, "num_tokens": 276068840.0, "step": 7272 }, { "epoch": 0.92520035618878, "ewc_loss": 0.10693359375, "ewc_loss_diag": 0.00014209747314453125, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.377239227294922, "learning_rate": 1e-05, "loss": 0.5305, "mean_token_accuracy": 0.8556283712387085, "num_tokens": 276109196.0, "step": 7273 }, { "epoch": 0.9253275664673706, "ewc_loss": 0.08544921875, "ewc_loss_diag": 0.00014209747314453125, "ewc_loss_parallel": 7.152557373046875e-05, "grad_norm": 5.108458995819092, "learning_rate": 1e-05, "loss": 0.4885, "mean_token_accuracy": 0.867946207523346, "num_tokens": 276151544.0, "step": 7274 }, { "epoch": 0.9254547767459611, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.00014209747314453125, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.861630439758301, "learning_rate": 1e-05, "loss": 0.5352, "mean_token_accuracy": 0.8568525314331055, "num_tokens": 276192849.0, "step": 7275 }, { "epoch": 0.9255819870245516, "ewc_loss": 0.091796875, "ewc_loss_diag": 0.00014209747314453125, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.254518032073975, "learning_rate": 1e-05, "loss": 0.4679, "mean_token_accuracy": 0.8727782964706421, "num_tokens": 276226543.0, "step": 7276 }, { "epoch": 0.925709197303142, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.00014209747314453125, "ewc_loss_parallel": 0.00011682510375976562, "grad_norm": 7.2710981369018555, "learning_rate": 1e-05, "loss": 0.5193, "mean_token_accuracy": 0.8669962286949158, "num_tokens": 276267199.0, "step": 7277 }, { "epoch": 0.9258364075817326, "ewc_loss": 0.10498046875, "ewc_loss_diag": 0.00014209747314453125, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 5.820507049560547, "learning_rate": 1e-05, "loss": 0.5326, "mean_token_accuracy": 0.8549737930297852, "num_tokens": 276306786.0, "step": 7278 }, { "epoch": 0.9259636178603231, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0001430511474609375, "ewc_loss_parallel": 0.00011157989501953125, "grad_norm": 7.149391174316406, "learning_rate": 1e-05, "loss": 0.5518, "mean_token_accuracy": 0.8602946996688843, "num_tokens": 276343479.0, "step": 7279 }, { "epoch": 0.9260908281389136, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.0001430511474609375, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.207667350769043, "learning_rate": 1e-05, "loss": 0.5384, "mean_token_accuracy": 0.857697606086731, "num_tokens": 276378067.0, "step": 7280 }, { "epoch": 0.9262180384175042, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.0001430511474609375, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.679531574249268, "learning_rate": 1e-05, "loss": 0.5219, "mean_token_accuracy": 0.8608719706535339, "num_tokens": 276407610.0, "step": 7281 }, { "epoch": 0.9263452486960947, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.0001430511474609375, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.183097839355469, "learning_rate": 1e-05, "loss": 0.5497, "mean_token_accuracy": 0.8501338958740234, "num_tokens": 276443664.0, "step": 7282 }, { "epoch": 0.9264724589746851, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.0001430511474609375, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.251004695892334, "learning_rate": 1e-05, "loss": 0.5027, "mean_token_accuracy": 0.8648152947425842, "num_tokens": 276480446.0, "step": 7283 }, { "epoch": 0.9265996692532756, "ewc_loss": 0.10400390625, "ewc_loss_diag": 0.0001430511474609375, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.079171180725098, "learning_rate": 1e-05, "loss": 0.4916, "mean_token_accuracy": 0.8676306009292603, "num_tokens": 276515012.0, "step": 7284 }, { "epoch": 0.9267268795318662, "ewc_loss": 0.09765625, "ewc_loss_diag": 0.0001430511474609375, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.929191589355469, "learning_rate": 1e-05, "loss": 0.4608, "mean_token_accuracy": 0.8796824216842651, "num_tokens": 276555804.0, "step": 7285 }, { "epoch": 0.9268540898104567, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.0001430511474609375, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.9977922439575195, "learning_rate": 1e-05, "loss": 0.5522, "mean_token_accuracy": 0.8470481038093567, "num_tokens": 276596290.0, "step": 7286 }, { "epoch": 0.9269813000890472, "ewc_loss": 0.0947265625, "ewc_loss_diag": 0.0001430511474609375, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.7213311195373535, "learning_rate": 1e-05, "loss": 0.5403, "mean_token_accuracy": 0.8511752486228943, "num_tokens": 276641767.0, "step": 7287 }, { "epoch": 0.9271085103676378, "ewc_loss": 0.09912109375, "ewc_loss_diag": 0.0001430511474609375, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.952406406402588, "learning_rate": 1e-05, "loss": 0.4607, "mean_token_accuracy": 0.8748319149017334, "num_tokens": 276678375.0, "step": 7288 }, { "epoch": 0.9272357206462282, "ewc_loss": 0.09326171875, "ewc_loss_diag": 0.0001430511474609375, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.604482650756836, "learning_rate": 1e-05, "loss": 0.518, "mean_token_accuracy": 0.8601261973381042, "num_tokens": 276715835.0, "step": 7289 }, { "epoch": 0.9273629309248187, "ewc_loss": 0.10009765625, "ewc_loss_diag": 0.0001430511474609375, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.947278022766113, "learning_rate": 1e-05, "loss": 0.5174, "mean_token_accuracy": 0.857985258102417, "num_tokens": 276750323.0, "step": 7290 }, { "epoch": 0.9274901412034092, "ewc_loss": 0.091796875, "ewc_loss_diag": 0.0001430511474609375, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.502430438995361, "learning_rate": 1e-05, "loss": 0.4896, "mean_token_accuracy": 0.8614038825035095, "num_tokens": 276786867.0, "step": 7291 }, { "epoch": 0.9276173514819998, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.0001430511474609375, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.9968109130859375, "learning_rate": 1e-05, "loss": 0.5908, "mean_token_accuracy": 0.8380914926528931, "num_tokens": 276831263.0, "step": 7292 }, { "epoch": 0.9277445617605903, "ewc_loss": 0.08935546875, "ewc_loss_diag": 0.0001430511474609375, "ewc_loss_parallel": 7.534027099609375e-05, "grad_norm": 5.389502048492432, "learning_rate": 1e-05, "loss": 0.4912, "mean_token_accuracy": 0.865804135799408, "num_tokens": 276867573.0, "step": 7293 }, { "epoch": 0.9278717720391808, "ewc_loss": 0.10302734375, "ewc_loss_diag": 0.0001430511474609375, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.144940376281738, "learning_rate": 1e-05, "loss": 0.4472, "mean_token_accuracy": 0.8819083571434021, "num_tokens": 276900414.0, "step": 7294 }, { "epoch": 0.9279989823177712, "ewc_loss": 0.08837890625, "ewc_loss_diag": 0.0001430511474609375, "ewc_loss_parallel": 7.43865966796875e-05, "grad_norm": 5.2539591789245605, "learning_rate": 1e-05, "loss": 0.5178, "mean_token_accuracy": 0.853652834892273, "num_tokens": 276938336.0, "step": 7295 }, { "epoch": 0.9281261925963618, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.0001430511474609375, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.413701057434082, "learning_rate": 1e-05, "loss": 0.4951, "mean_token_accuracy": 0.8688668608665466, "num_tokens": 276981277.0, "step": 7296 }, { "epoch": 0.9282534028749523, "ewc_loss": 0.0888671875, "ewc_loss_diag": 0.0001430511474609375, "ewc_loss_parallel": 7.486343383789062e-05, "grad_norm": 5.178893566131592, "learning_rate": 1e-05, "loss": 0.5511, "mean_token_accuracy": 0.8539448976516724, "num_tokens": 277023346.0, "step": 7297 }, { "epoch": 0.9283806131535428, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.0001430511474609375, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.797610759735107, "learning_rate": 1e-05, "loss": 0.5088, "mean_token_accuracy": 0.8633950352668762, "num_tokens": 277054821.0, "step": 7298 }, { "epoch": 0.9285078234321333, "ewc_loss": 0.09228515625, "ewc_loss_diag": 0.0001430511474609375, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.2590155601501465, "learning_rate": 1e-05, "loss": 0.5025, "mean_token_accuracy": 0.8634523153305054, "num_tokens": 277092124.0, "step": 7299 }, { "epoch": 0.9286350337107239, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0001430511474609375, "ewc_loss_parallel": 0.0001125335693359375, "grad_norm": 7.162538051605225, "learning_rate": 1e-05, "loss": 0.5395, "mean_token_accuracy": 0.8601014614105225, "num_tokens": 277130119.0, "step": 7300 }, { "epoch": 0.9287622439893143, "ewc_loss": 0.10302734375, "ewc_loss_diag": 0.0001430511474609375, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 5.694943904876709, "learning_rate": 1e-05, "loss": 0.4561, "mean_token_accuracy": 0.8784717321395874, "num_tokens": 277167777.0, "step": 7301 }, { "epoch": 0.9288894542679048, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0001430511474609375, "ewc_loss_parallel": 0.00011157989501953125, "grad_norm": 7.141045570373535, "learning_rate": 1e-05, "loss": 0.5208, "mean_token_accuracy": 0.8689072132110596, "num_tokens": 277207273.0, "step": 7302 }, { "epoch": 0.9290166645464953, "ewc_loss": 0.11181640625, "ewc_loss_diag": 0.0001430511474609375, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.0994038581848145, "learning_rate": 1e-05, "loss": 0.4876, "mean_token_accuracy": 0.8698203563690186, "num_tokens": 277248996.0, "step": 7303 }, { "epoch": 0.9291438748250859, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.0001430511474609375, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.750925540924072, "learning_rate": 1e-05, "loss": 0.5799, "mean_token_accuracy": 0.8458744287490845, "num_tokens": 277286212.0, "step": 7304 }, { "epoch": 0.9292710851036764, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.0001430511474609375, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.154648303985596, "learning_rate": 1e-05, "loss": 0.581, "mean_token_accuracy": 0.8429244756698608, "num_tokens": 277330482.0, "step": 7305 }, { "epoch": 0.9293982953822669, "ewc_loss": 0.10595703125, "ewc_loss_diag": 0.0001430511474609375, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.31674337387085, "learning_rate": 1e-05, "loss": 0.4924, "mean_token_accuracy": 0.8673272132873535, "num_tokens": 277364965.0, "step": 7306 }, { "epoch": 0.9295255056608573, "ewc_loss": 0.10595703125, "ewc_loss_diag": 0.0001430511474609375, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.073947906494141, "learning_rate": 1e-05, "loss": 0.5048, "mean_token_accuracy": 0.8614450097084045, "num_tokens": 277392227.0, "step": 7307 }, { "epoch": 0.9296527159394479, "ewc_loss": 0.09814453125, "ewc_loss_diag": 0.0001430511474609375, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.954870223999023, "learning_rate": 1e-05, "loss": 0.4812, "mean_token_accuracy": 0.87052321434021, "num_tokens": 277435658.0, "step": 7308 }, { "epoch": 0.9297799262180384, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.0001430511474609375, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 6.0135369300842285, "learning_rate": 1e-05, "loss": 0.5294, "mean_token_accuracy": 0.8579109907150269, "num_tokens": 277472513.0, "step": 7309 }, { "epoch": 0.9299071364966289, "ewc_loss": 0.0947265625, "ewc_loss_diag": 0.0001430511474609375, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.716310977935791, "learning_rate": 1e-05, "loss": 0.5407, "mean_token_accuracy": 0.8473585247993469, "num_tokens": 277517740.0, "step": 7310 }, { "epoch": 0.9300343467752195, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.0001430511474609375, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.998132228851318, "learning_rate": 1e-05, "loss": 0.4605, "mean_token_accuracy": 0.8754416108131409, "num_tokens": 277547304.0, "step": 7311 }, { "epoch": 0.93016155705381, "ewc_loss": 0.09375, "ewc_loss_diag": 0.0001430511474609375, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.5729146003723145, "learning_rate": 1e-05, "loss": 0.5542, "mean_token_accuracy": 0.847506582736969, "num_tokens": 277581380.0, "step": 7312 }, { "epoch": 0.9302887673324005, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.0001430511474609375, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 6.014636516571045, "learning_rate": 1e-05, "loss": 0.4859, "mean_token_accuracy": 0.8691197633743286, "num_tokens": 277621615.0, "step": 7313 }, { "epoch": 0.9304159776109909, "ewc_loss": 0.09130859375, "ewc_loss_diag": 0.0001430511474609375, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.4427361488342285, "learning_rate": 1e-05, "loss": 0.5383, "mean_token_accuracy": 0.8496413230895996, "num_tokens": 277655157.0, "step": 7314 }, { "epoch": 0.9305431878895815, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.0001430511474609375, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 6.131319046020508, "learning_rate": 1e-05, "loss": 0.5628, "mean_token_accuracy": 0.8473437428474426, "num_tokens": 277688634.0, "step": 7315 }, { "epoch": 0.930670398168172, "ewc_loss": 0.08935546875, "ewc_loss_diag": 0.0001430511474609375, "ewc_loss_parallel": 7.534027099609375e-05, "grad_norm": 5.305845260620117, "learning_rate": 1e-05, "loss": 0.5563, "mean_token_accuracy": 0.8482813239097595, "num_tokens": 277736123.0, "step": 7316 }, { "epoch": 0.9307976084467625, "ewc_loss": 0.10595703125, "ewc_loss_diag": 0.0001430511474609375, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.343183517456055, "learning_rate": 1e-05, "loss": 0.5286, "mean_token_accuracy": 0.8595963716506958, "num_tokens": 277776197.0, "step": 7317 }, { "epoch": 0.930924818725353, "ewc_loss": 0.08935546875, "ewc_loss_diag": 0.0001430511474609375, "ewc_loss_parallel": 7.534027099609375e-05, "grad_norm": 5.193048000335693, "learning_rate": 1e-05, "loss": 0.4636, "mean_token_accuracy": 0.8725385665893555, "num_tokens": 277820477.0, "step": 7318 }, { "epoch": 0.9310520290039436, "ewc_loss": 0.11474609375, "ewc_loss_diag": 0.0001430511474609375, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.691158771514893, "learning_rate": 1e-05, "loss": 0.5259, "mean_token_accuracy": 0.8623545169830322, "num_tokens": 277864660.0, "step": 7319 }, { "epoch": 0.931179239282534, "ewc_loss": 0.09228515625, "ewc_loss_diag": 0.0001430511474609375, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.243664264678955, "learning_rate": 1e-05, "loss": 0.5479, "mean_token_accuracy": 0.8460277318954468, "num_tokens": 277905702.0, "step": 7320 }, { "epoch": 0.9313064495611245, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.0001430511474609375, "ewc_loss_parallel": 0.00010967254638671875, "grad_norm": 7.048977851867676, "learning_rate": 1e-05, "loss": 0.5231, "mean_token_accuracy": 0.8650716543197632, "num_tokens": 277946033.0, "step": 7321 }, { "epoch": 0.931433659839715, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.0001430511474609375, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 5.603476524353027, "learning_rate": 1e-05, "loss": 0.5117, "mean_token_accuracy": 0.8615409135818481, "num_tokens": 277981556.0, "step": 7322 }, { "epoch": 0.9315608701183056, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.0001430511474609375, "ewc_loss_parallel": 0.00011014938354492188, "grad_norm": 7.086243152618408, "learning_rate": 1e-05, "loss": 0.4935, "mean_token_accuracy": 0.8741892576217651, "num_tokens": 278021455.0, "step": 7323 }, { "epoch": 0.9316880803968961, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.0001430511474609375, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 5.993979454040527, "learning_rate": 1e-05, "loss": 0.4957, "mean_token_accuracy": 0.8665369749069214, "num_tokens": 278057840.0, "step": 7324 }, { "epoch": 0.9318152906754866, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.746784210205078, "learning_rate": 1e-05, "loss": 0.55, "mean_token_accuracy": 0.8562700748443604, "num_tokens": 278093186.0, "step": 7325 }, { "epoch": 0.931942500954077, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.10678768157959, "learning_rate": 1e-05, "loss": 0.531, "mean_token_accuracy": 0.8593942523002625, "num_tokens": 278127235.0, "step": 7326 }, { "epoch": 0.9320697112326676, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.331120014190674, "learning_rate": 1e-05, "loss": 0.5153, "mean_token_accuracy": 0.8631958961486816, "num_tokens": 278165236.0, "step": 7327 }, { "epoch": 0.9321969215112581, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.051996231079102, "learning_rate": 1e-05, "loss": 0.5482, "mean_token_accuracy": 0.8481417894363403, "num_tokens": 278205153.0, "step": 7328 }, { "epoch": 0.9323241317898486, "ewc_loss": 0.0986328125, "ewc_loss_diag": 0.0001430511474609375, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.955575466156006, "learning_rate": 1e-05, "loss": 0.4662, "mean_token_accuracy": 0.874129056930542, "num_tokens": 278240808.0, "step": 7329 }, { "epoch": 0.9324513420684392, "ewc_loss": 0.10205078125, "ewc_loss_diag": 0.0001430511474609375, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 6.015414237976074, "learning_rate": 1e-05, "loss": 0.5355, "mean_token_accuracy": 0.8529016375541687, "num_tokens": 278281364.0, "step": 7330 }, { "epoch": 0.9325785523470297, "ewc_loss": 0.0947265625, "ewc_loss_diag": 0.0001430511474609375, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.691184043884277, "learning_rate": 1e-05, "loss": 0.4855, "mean_token_accuracy": 0.867209792137146, "num_tokens": 278316021.0, "step": 7331 }, { "epoch": 0.9327057626256201, "ewc_loss": 0.10107421875, "ewc_loss_diag": 0.0001430511474609375, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 6.01596212387085, "learning_rate": 1e-05, "loss": 0.5341, "mean_token_accuracy": 0.8553886413574219, "num_tokens": 278357744.0, "step": 7332 }, { "epoch": 0.9328329729042106, "ewc_loss": 0.09228515625, "ewc_loss_diag": 0.0001430511474609375, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.5346784591674805, "learning_rate": 1e-05, "loss": 0.4943, "mean_token_accuracy": 0.8600331544876099, "num_tokens": 278390826.0, "step": 7333 }, { "epoch": 0.9329601831828012, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 6.063121318817139, "learning_rate": 1e-05, "loss": 0.5666, "mean_token_accuracy": 0.8436236381530762, "num_tokens": 278424395.0, "step": 7334 }, { "epoch": 0.9330873934613917, "ewc_loss": 0.0908203125, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.394991397857666, "learning_rate": 1e-05, "loss": 0.4608, "mean_token_accuracy": 0.8737099170684814, "num_tokens": 278462180.0, "step": 7335 }, { "epoch": 0.9332146037399822, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.195744514465332, "learning_rate": 1e-05, "loss": 0.5388, "mean_token_accuracy": 0.8561882972717285, "num_tokens": 278499349.0, "step": 7336 }, { "epoch": 0.9333418140185727, "ewc_loss": 0.08984375, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 7.581710815429688e-05, "grad_norm": 5.283544063568115, "learning_rate": 1e-05, "loss": 0.533, "mean_token_accuracy": 0.849555492401123, "num_tokens": 278530782.0, "step": 7337 }, { "epoch": 0.9334690242971632, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.0001430511474609375, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.439866542816162, "learning_rate": 1e-05, "loss": 0.5311, "mean_token_accuracy": 0.8561203479766846, "num_tokens": 278563262.0, "step": 7338 }, { "epoch": 0.9335962345757537, "ewc_loss": 0.08935546875, "ewc_loss_diag": 0.0001430511474609375, "ewc_loss_parallel": 7.534027099609375e-05, "grad_norm": 5.214777946472168, "learning_rate": 1e-05, "loss": 0.5048, "mean_token_accuracy": 0.8608965873718262, "num_tokens": 278598882.0, "step": 7339 }, { "epoch": 0.9337234448543442, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.0001430511474609375, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.786304473876953, "learning_rate": 1e-05, "loss": 0.5193, "mean_token_accuracy": 0.8622934818267822, "num_tokens": 278628535.0, "step": 7340 }, { "epoch": 0.9338506551329347, "ewc_loss": 0.0947265625, "ewc_loss_diag": 0.0001430511474609375, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.331814289093018, "learning_rate": 1e-05, "loss": 0.5226, "mean_token_accuracy": 0.8565002679824829, "num_tokens": 278667732.0, "step": 7341 }, { "epoch": 0.9339778654115253, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.0001430511474609375, "ewc_loss_parallel": 0.00011014938354492188, "grad_norm": 7.061817169189453, "learning_rate": 1e-05, "loss": 0.5161, "mean_token_accuracy": 0.8631874322891235, "num_tokens": 278706925.0, "step": 7342 }, { "epoch": 0.9341050756901158, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 5.7180657386779785, "learning_rate": 1e-05, "loss": 0.4755, "mean_token_accuracy": 0.8734946846961975, "num_tokens": 278740491.0, "step": 7343 }, { "epoch": 0.9342322859687062, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.0001430511474609375, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 6.9758992195129395, "learning_rate": 1e-05, "loss": 0.57, "mean_token_accuracy": 0.8498232364654541, "num_tokens": 278776739.0, "step": 7344 }, { "epoch": 0.9343594962472968, "ewc_loss": 0.109375, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.020356178283691, "learning_rate": 1e-05, "loss": 0.4927, "mean_token_accuracy": 0.8703872561454773, "num_tokens": 278813019.0, "step": 7345 }, { "epoch": 0.9344867065258873, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.60609769821167, "learning_rate": 1e-05, "loss": 0.5357, "mean_token_accuracy": 0.860196590423584, "num_tokens": 278853807.0, "step": 7346 }, { "epoch": 0.9346139168044778, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.067989349365234, "learning_rate": 1e-05, "loss": 0.4688, "mean_token_accuracy": 0.8786939382553101, "num_tokens": 278891643.0, "step": 7347 }, { "epoch": 0.9347411270830683, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 6.20744514465332, "learning_rate": 1e-05, "loss": 0.6208, "mean_token_accuracy": 0.840783417224884, "num_tokens": 278929915.0, "step": 7348 }, { "epoch": 0.9348683373616589, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.0292277336120605, "learning_rate": 1e-05, "loss": 0.5684, "mean_token_accuracy": 0.8431849479675293, "num_tokens": 278964400.0, "step": 7349 }, { "epoch": 0.9349955476402493, "ewc_loss": 0.09765625, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.86326265335083, "learning_rate": 1e-05, "loss": 0.5037, "mean_token_accuracy": 0.8626953959465027, "num_tokens": 279003322.0, "step": 7350 }, { "epoch": 0.9351227579188398, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 5.9882001876831055, "learning_rate": 1e-05, "loss": 0.5521, "mean_token_accuracy": 0.8490167856216431, "num_tokens": 279045630.0, "step": 7351 }, { "epoch": 0.9352499681974303, "ewc_loss": 0.0947265625, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.658020973205566, "learning_rate": 1e-05, "loss": 0.5223, "mean_token_accuracy": 0.8609590530395508, "num_tokens": 279085330.0, "step": 7352 }, { "epoch": 0.9353771784760209, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 5.9613142013549805, "learning_rate": 1e-05, "loss": 0.4941, "mean_token_accuracy": 0.8673422932624817, "num_tokens": 279125499.0, "step": 7353 }, { "epoch": 0.9355043887546114, "ewc_loss": 0.0927734375, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.543900489807129, "learning_rate": 1e-05, "loss": 0.4918, "mean_token_accuracy": 0.8636269569396973, "num_tokens": 279154772.0, "step": 7354 }, { "epoch": 0.9356315990332019, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 5.979199409484863, "learning_rate": 1e-05, "loss": 0.5402, "mean_token_accuracy": 0.8561404943466187, "num_tokens": 279196947.0, "step": 7355 }, { "epoch": 0.9357588093117923, "ewc_loss": 0.0908203125, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 7.62939453125e-05, "grad_norm": 5.425843238830566, "learning_rate": 1e-05, "loss": 0.4712, "mean_token_accuracy": 0.8766602873802185, "num_tokens": 279236872.0, "step": 7356 }, { "epoch": 0.9358860195903829, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 6.089934825897217, "learning_rate": 1e-05, "loss": 0.4908, "mean_token_accuracy": 0.868213415145874, "num_tokens": 279275456.0, "step": 7357 }, { "epoch": 0.9360132298689734, "ewc_loss": 0.08984375, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 7.581710815429688e-05, "grad_norm": 5.307929515838623, "learning_rate": 1e-05, "loss": 0.5127, "mean_token_accuracy": 0.8571794629096985, "num_tokens": 279308406.0, "step": 7358 }, { "epoch": 0.9361404401475639, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.294346809387207, "learning_rate": 1e-05, "loss": 0.5213, "mean_token_accuracy": 0.8577199578285217, "num_tokens": 279349522.0, "step": 7359 }, { "epoch": 0.9362676504261545, "ewc_loss": 0.0888671875, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 7.486343383789062e-05, "grad_norm": 5.184886455535889, "learning_rate": 1e-05, "loss": 0.4625, "mean_token_accuracy": 0.8714503645896912, "num_tokens": 279388242.0, "step": 7360 }, { "epoch": 0.936394860704745, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.636886119842529, "learning_rate": 1e-05, "loss": 0.4827, "mean_token_accuracy": 0.8742280006408691, "num_tokens": 279421239.0, "step": 7361 }, { "epoch": 0.9365220709833355, "ewc_loss": 0.091796875, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.202478408813477, "learning_rate": 1e-05, "loss": 0.5803, "mean_token_accuracy": 0.8373513221740723, "num_tokens": 279464823.0, "step": 7362 }, { "epoch": 0.9366492812619259, "ewc_loss": 0.125, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 0.00011014938354492188, "grad_norm": 7.021518230438232, "learning_rate": 1e-05, "loss": 0.578, "mean_token_accuracy": 0.8485283851623535, "num_tokens": 279507645.0, "step": 7363 }, { "epoch": 0.9367764915405165, "ewc_loss": 0.099609375, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.5172224044799805, "learning_rate": 1e-05, "loss": 0.5156, "mean_token_accuracy": 0.8588423132896423, "num_tokens": 279539905.0, "step": 7364 }, { "epoch": 0.936903701819107, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 0.00011205673217773438, "grad_norm": 7.140132904052734, "learning_rate": 1e-05, "loss": 0.5097, "mean_token_accuracy": 0.869523286819458, "num_tokens": 279577929.0, "step": 7365 }, { "epoch": 0.9370309120976975, "ewc_loss": 0.109375, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 5.961101055145264, "learning_rate": 1e-05, "loss": 0.5113, "mean_token_accuracy": 0.8636507987976074, "num_tokens": 279618421.0, "step": 7366 }, { "epoch": 0.937158122376288, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.8445868492126465, "learning_rate": 1e-05, "loss": 0.5169, "mean_token_accuracy": 0.8673219680786133, "num_tokens": 279656563.0, "step": 7367 }, { "epoch": 0.9372853326548786, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.125764846801758, "learning_rate": 1e-05, "loss": 0.5489, "mean_token_accuracy": 0.8520087003707886, "num_tokens": 279693159.0, "step": 7368 }, { "epoch": 0.937412542933469, "ewc_loss": 0.109375, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.407698631286621, "learning_rate": 1e-05, "loss": 0.4763, "mean_token_accuracy": 0.8737484216690063, "num_tokens": 279731492.0, "step": 7369 }, { "epoch": 0.9375397532120595, "ewc_loss": 0.109375, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.077194690704346, "learning_rate": 1e-05, "loss": 0.5127, "mean_token_accuracy": 0.8660364151000977, "num_tokens": 279774906.0, "step": 7370 }, { "epoch": 0.93766696349065, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 6.039836406707764, "learning_rate": 1e-05, "loss": 0.5836, "mean_token_accuracy": 0.8399460315704346, "num_tokens": 279809797.0, "step": 7371 }, { "epoch": 0.9377941737692406, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.024847507476807, "learning_rate": 1e-05, "loss": 0.4924, "mean_token_accuracy": 0.8662048578262329, "num_tokens": 279850107.0, "step": 7372 }, { "epoch": 0.9379213840478311, "ewc_loss": 0.0966796875, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.757904529571533, "learning_rate": 1e-05, "loss": 0.4331, "mean_token_accuracy": 0.8848649263381958, "num_tokens": 279885115.0, "step": 7373 }, { "epoch": 0.9380485943264216, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 6.019205093383789, "learning_rate": 1e-05, "loss": 0.4936, "mean_token_accuracy": 0.8687021732330322, "num_tokens": 279921343.0, "step": 7374 }, { "epoch": 0.938175804605012, "ewc_loss": 0.0927734375, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.565651893615723, "learning_rate": 1e-05, "loss": 0.4566, "mean_token_accuracy": 0.8714055418968201, "num_tokens": 279953059.0, "step": 7375 }, { "epoch": 0.9383030148836026, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 6.057645320892334, "learning_rate": 1e-05, "loss": 0.4889, "mean_token_accuracy": 0.8689036965370178, "num_tokens": 279987993.0, "step": 7376 }, { "epoch": 0.9384302251621931, "ewc_loss": 0.091796875, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.429708957672119, "learning_rate": 1e-05, "loss": 0.5316, "mean_token_accuracy": 0.8535997867584229, "num_tokens": 280024620.0, "step": 7377 }, { "epoch": 0.9385574354407836, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 6.163057327270508, "learning_rate": 1e-05, "loss": 0.5774, "mean_token_accuracy": 0.8419965505599976, "num_tokens": 280069011.0, "step": 7378 }, { "epoch": 0.9386846457193742, "ewc_loss": 0.08984375, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 7.534027099609375e-05, "grad_norm": 5.300324440002441, "learning_rate": 1e-05, "loss": 0.445, "mean_token_accuracy": 0.880772590637207, "num_tokens": 280105877.0, "step": 7379 }, { "epoch": 0.9388118559979647, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.390291690826416, "learning_rate": 1e-05, "loss": 0.4653, "mean_token_accuracy": 0.8727595210075378, "num_tokens": 280143155.0, "step": 7380 }, { "epoch": 0.9389390662765551, "ewc_loss": 0.08984375, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 7.534027099609375e-05, "grad_norm": 5.214231967926025, "learning_rate": 1e-05, "loss": 0.4672, "mean_token_accuracy": 0.8706552386283875, "num_tokens": 280181490.0, "step": 7381 }, { "epoch": 0.9390662765551456, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.744677543640137, "learning_rate": 1e-05, "loss": 0.5634, "mean_token_accuracy": 0.8511825799942017, "num_tokens": 280221959.0, "step": 7382 }, { "epoch": 0.9391934868337362, "ewc_loss": 0.0927734375, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.290227890014648, "learning_rate": 1e-05, "loss": 0.4392, "mean_token_accuracy": 0.8795210719108582, "num_tokens": 280265224.0, "step": 7383 }, { "epoch": 0.9393206971123267, "ewc_loss": 0.125, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 0.00011014938354492188, "grad_norm": 7.069648742675781, "learning_rate": 1e-05, "loss": 0.5341, "mean_token_accuracy": 0.8665785789489746, "num_tokens": 280302039.0, "step": 7384 }, { "epoch": 0.9394479073909172, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 5.662725448608398, "learning_rate": 1e-05, "loss": 0.5101, "mean_token_accuracy": 0.8633727431297302, "num_tokens": 280338559.0, "step": 7385 }, { "epoch": 0.9395751176695077, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 0.00010967254638671875, "grad_norm": 7.050375938415527, "learning_rate": 1e-05, "loss": 0.4957, "mean_token_accuracy": 0.8719487190246582, "num_tokens": 280376731.0, "step": 7386 }, { "epoch": 0.9397023279480982, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.020864963531494, "learning_rate": 1e-05, "loss": 0.5254, "mean_token_accuracy": 0.8597884178161621, "num_tokens": 280414201.0, "step": 7387 }, { "epoch": 0.9398295382266887, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.6763691902160645, "learning_rate": 1e-05, "loss": 0.5184, "mean_token_accuracy": 0.8630275130271912, "num_tokens": 280456490.0, "step": 7388 }, { "epoch": 0.9399567485052792, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.090446472167969, "learning_rate": 1e-05, "loss": 0.5237, "mean_token_accuracy": 0.8594781160354614, "num_tokens": 280499593.0, "step": 7389 }, { "epoch": 0.9400839587838697, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.263979911804199, "learning_rate": 1e-05, "loss": 0.5336, "mean_token_accuracy": 0.853745698928833, "num_tokens": 280535722.0, "step": 7390 }, { "epoch": 0.9402111690624603, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.058420181274414, "learning_rate": 1e-05, "loss": 0.4868, "mean_token_accuracy": 0.8669068813323975, "num_tokens": 280569666.0, "step": 7391 }, { "epoch": 0.9403383793410508, "ewc_loss": 0.09765625, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.889532566070557, "learning_rate": 1e-05, "loss": 0.4189, "mean_token_accuracy": 0.8883016705513, "num_tokens": 280607359.0, "step": 7392 }, { "epoch": 0.9404655896196412, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.033494472503662, "learning_rate": 1e-05, "loss": 0.5065, "mean_token_accuracy": 0.8591643571853638, "num_tokens": 280641054.0, "step": 7393 }, { "epoch": 0.9405927998982317, "ewc_loss": 0.09375, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.642974376678467, "learning_rate": 1e-05, "loss": 0.5251, "mean_token_accuracy": 0.8514364957809448, "num_tokens": 280683798.0, "step": 7394 }, { "epoch": 0.9407200101768223, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 6.03753662109375, "learning_rate": 1e-05, "loss": 0.5415, "mean_token_accuracy": 0.8505903482437134, "num_tokens": 280729407.0, "step": 7395 }, { "epoch": 0.9408472204554128, "ewc_loss": 0.091796875, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.491177082061768, "learning_rate": 1e-05, "loss": 0.4928, "mean_token_accuracy": 0.8645786046981812, "num_tokens": 280764327.0, "step": 7396 }, { "epoch": 0.9409744307340033, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.1304931640625, "learning_rate": 1e-05, "loss": 0.5275, "mean_token_accuracy": 0.8583746552467346, "num_tokens": 280802056.0, "step": 7397 }, { "epoch": 0.9411016410125939, "ewc_loss": 0.08984375, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 7.534027099609375e-05, "grad_norm": 5.355166912078857, "learning_rate": 1e-05, "loss": 0.4155, "mean_token_accuracy": 0.883974552154541, "num_tokens": 280837269.0, "step": 7398 }, { "epoch": 0.9412288512911843, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.310405254364014, "learning_rate": 1e-05, "loss": 0.5844, "mean_token_accuracy": 0.8400951027870178, "num_tokens": 280873914.0, "step": 7399 }, { "epoch": 0.9413560615697748, "ewc_loss": 0.08984375, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 7.534027099609375e-05, "grad_norm": 5.2574052810668945, "learning_rate": 1e-05, "loss": 0.488, "mean_token_accuracy": 0.8635058999061584, "num_tokens": 280911436.0, "step": 7400 }, { "epoch": 0.9414832718483653, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.578230381011963, "learning_rate": 1e-05, "loss": 0.4952, "mean_token_accuracy": 0.8719849586486816, "num_tokens": 280952395.0, "step": 7401 }, { "epoch": 0.9416104821269559, "ewc_loss": 0.0908203125, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 7.62939453125e-05, "grad_norm": 5.2533440589904785, "learning_rate": 1e-05, "loss": 0.5484, "mean_token_accuracy": 0.8460827469825745, "num_tokens": 280990455.0, "step": 7402 }, { "epoch": 0.9417376924055464, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.914392471313477, "learning_rate": 1e-05, "loss": 0.5323, "mean_token_accuracy": 0.8627018332481384, "num_tokens": 281020584.0, "step": 7403 }, { "epoch": 0.9418649026841369, "ewc_loss": 0.095703125, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.47676420211792, "learning_rate": 1e-05, "loss": 0.4764, "mean_token_accuracy": 0.8737713694572449, "num_tokens": 281059907.0, "step": 7404 }, { "epoch": 0.9419921129627273, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 7.045109272003174, "learning_rate": 1e-05, "loss": 0.5319, "mean_token_accuracy": 0.864913284778595, "num_tokens": 281097425.0, "step": 7405 }, { "epoch": 0.9421193232413179, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 5.852487087249756, "learning_rate": 1e-05, "loss": 0.5333, "mean_token_accuracy": 0.8593348264694214, "num_tokens": 281146696.0, "step": 7406 }, { "epoch": 0.9422465335199084, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.823449611663818, "learning_rate": 1e-05, "loss": 0.5646, "mean_token_accuracy": 0.8481935858726501, "num_tokens": 281186166.0, "step": 7407 }, { "epoch": 0.9423737437984989, "ewc_loss": 0.109375, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.051621913909912, "learning_rate": 1e-05, "loss": 0.6249, "mean_token_accuracy": 0.8300992250442505, "num_tokens": 281230372.0, "step": 7408 }, { "epoch": 0.9425009540770894, "ewc_loss": 0.109375, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.4364094734191895, "learning_rate": 1e-05, "loss": 0.5086, "mean_token_accuracy": 0.8641348481178284, "num_tokens": 281256035.0, "step": 7409 }, { "epoch": 0.94262816435568, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.053704261779785, "learning_rate": 1e-05, "loss": 0.4681, "mean_token_accuracy": 0.8752078413963318, "num_tokens": 281290792.0, "step": 7410 }, { "epoch": 0.9427553746342705, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 6.0294575691223145, "learning_rate": 1e-05, "loss": 0.5061, "mean_token_accuracy": 0.8669221997261047, "num_tokens": 281329949.0, "step": 7411 }, { "epoch": 0.9428825849128609, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 6.017267227172852, "learning_rate": 1e-05, "loss": 0.4909, "mean_token_accuracy": 0.8700105547904968, "num_tokens": 281366383.0, "step": 7412 }, { "epoch": 0.9430097951914514, "ewc_loss": 0.095703125, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.741426944732666, "learning_rate": 1e-05, "loss": 0.5315, "mean_token_accuracy": 0.8523966670036316, "num_tokens": 281408537.0, "step": 7413 }, { "epoch": 0.943137005470042, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 6.005014896392822, "learning_rate": 1e-05, "loss": 0.5285, "mean_token_accuracy": 0.8566330075263977, "num_tokens": 281447745.0, "step": 7414 }, { "epoch": 0.9432642157486325, "ewc_loss": 0.0927734375, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.562427043914795, "learning_rate": 1e-05, "loss": 0.4835, "mean_token_accuracy": 0.8683258891105652, "num_tokens": 281484430.0, "step": 7415 }, { "epoch": 0.943391426027223, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 6.032415866851807, "learning_rate": 1e-05, "loss": 0.4605, "mean_token_accuracy": 0.881039023399353, "num_tokens": 281519720.0, "step": 7416 }, { "epoch": 0.9435186363058136, "ewc_loss": 0.091796875, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.434330463409424, "learning_rate": 1e-05, "loss": 0.5426, "mean_token_accuracy": 0.8465989828109741, "num_tokens": 281556039.0, "step": 7417 }, { "epoch": 0.943645846584404, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 6.134238243103027, "learning_rate": 1e-05, "loss": 0.5378, "mean_token_accuracy": 0.8534067869186401, "num_tokens": 281594781.0, "step": 7418 }, { "epoch": 0.9437730568629945, "ewc_loss": 0.08984375, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 7.534027099609375e-05, "grad_norm": 5.311129570007324, "learning_rate": 1e-05, "loss": 0.5386, "mean_token_accuracy": 0.8489574193954468, "num_tokens": 281632743.0, "step": 7419 }, { "epoch": 0.943900267141585, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.348837852478027, "learning_rate": 1e-05, "loss": 0.5246, "mean_token_accuracy": 0.8624419569969177, "num_tokens": 281669706.0, "step": 7420 }, { "epoch": 0.9440274774201756, "ewc_loss": 0.08984375, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 7.534027099609375e-05, "grad_norm": 5.213390350341797, "learning_rate": 1e-05, "loss": 0.483, "mean_token_accuracy": 0.8669811487197876, "num_tokens": 281706108.0, "step": 7421 }, { "epoch": 0.9441546876987661, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.672643184661865, "learning_rate": 1e-05, "loss": 0.5497, "mean_token_accuracy": 0.8562073111534119, "num_tokens": 281743437.0, "step": 7422 }, { "epoch": 0.9442818979773566, "ewc_loss": 0.091796875, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.25984001159668, "learning_rate": 1e-05, "loss": 0.4484, "mean_token_accuracy": 0.8754185438156128, "num_tokens": 281779652.0, "step": 7423 }, { "epoch": 0.944409108255947, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 7.009237289428711, "learning_rate": 1e-05, "loss": 0.5223, "mean_token_accuracy": 0.8660153746604919, "num_tokens": 281815841.0, "step": 7424 }, { "epoch": 0.9445363185345376, "ewc_loss": 0.099609375, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.577265739440918, "learning_rate": 1e-05, "loss": 0.5071, "mean_token_accuracy": 0.861199140548706, "num_tokens": 281853991.0, "step": 7425 }, { "epoch": 0.9446635288131281, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 7.082039833068848, "learning_rate": 1e-05, "loss": 0.5472, "mean_token_accuracy": 0.8568347692489624, "num_tokens": 281883264.0, "step": 7426 }, { "epoch": 0.9447907390917186, "ewc_loss": 0.109375, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 5.962211608886719, "learning_rate": 1e-05, "loss": 0.4807, "mean_token_accuracy": 0.876606822013855, "num_tokens": 281920266.0, "step": 7427 }, { "epoch": 0.9449179493703092, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.753800392150879, "learning_rate": 1e-05, "loss": 0.5226, "mean_token_accuracy": 0.862238347530365, "num_tokens": 281959949.0, "step": 7428 }, { "epoch": 0.9450451596488997, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.092406749725342, "learning_rate": 1e-05, "loss": 0.5024, "mean_token_accuracy": 0.8673139214515686, "num_tokens": 281992255.0, "step": 7429 }, { "epoch": 0.9451723699274901, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.350099086761475, "learning_rate": 1e-05, "loss": 0.5119, "mean_token_accuracy": 0.8618746399879456, "num_tokens": 282028000.0, "step": 7430 }, { "epoch": 0.9452995802060806, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.058072090148926, "learning_rate": 1e-05, "loss": 0.5104, "mean_token_accuracy": 0.8649458885192871, "num_tokens": 282064332.0, "step": 7431 }, { "epoch": 0.9454267904846712, "ewc_loss": 0.0986328125, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.975468158721924, "learning_rate": 1e-05, "loss": 0.5351, "mean_token_accuracy": 0.850515604019165, "num_tokens": 282102185.0, "step": 7432 }, { "epoch": 0.9455540007632617, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.038717269897461, "learning_rate": 1e-05, "loss": 0.6213, "mean_token_accuracy": 0.8249496221542358, "num_tokens": 282138708.0, "step": 7433 }, { "epoch": 0.9456812110418522, "ewc_loss": 0.0947265625, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.700299263000488, "learning_rate": 1e-05, "loss": 0.4698, "mean_token_accuracy": 0.8724747896194458, "num_tokens": 282168223.0, "step": 7434 }, { "epoch": 0.9458084213204427, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 6.019918918609619, "learning_rate": 1e-05, "loss": 0.4746, "mean_token_accuracy": 0.8725785613059998, "num_tokens": 282205226.0, "step": 7435 }, { "epoch": 0.9459356315990332, "ewc_loss": 0.0927734375, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.537714004516602, "learning_rate": 1e-05, "loss": 0.544, "mean_token_accuracy": 0.8484115600585938, "num_tokens": 282239131.0, "step": 7436 }, { "epoch": 0.9460628418776237, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 6.063169956207275, "learning_rate": 1e-05, "loss": 0.5127, "mean_token_accuracy": 0.8634871244430542, "num_tokens": 282276175.0, "step": 7437 }, { "epoch": 0.9461900521562142, "ewc_loss": 0.08984375, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 7.581710815429688e-05, "grad_norm": 5.409214496612549, "learning_rate": 1e-05, "loss": 0.5345, "mean_token_accuracy": 0.8541250228881836, "num_tokens": 282308598.0, "step": 7438 }, { "epoch": 0.9463172624348047, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 6.181023120880127, "learning_rate": 1e-05, "loss": 0.4494, "mean_token_accuracy": 0.8807937502861023, "num_tokens": 282342193.0, "step": 7439 }, { "epoch": 0.9464444727133953, "ewc_loss": 0.08984375, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 7.534027099609375e-05, "grad_norm": 5.28725004196167, "learning_rate": 1e-05, "loss": 0.4907, "mean_token_accuracy": 0.8630819320678711, "num_tokens": 282380531.0, "step": 7440 }, { "epoch": 0.9465716829919858, "ewc_loss": 0.109375, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.428240776062012, "learning_rate": 1e-05, "loss": 0.5142, "mean_token_accuracy": 0.8581881523132324, "num_tokens": 282410713.0, "step": 7441 }, { "epoch": 0.9466988932705762, "ewc_loss": 0.08984375, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 7.581710815429688e-05, "grad_norm": 5.211965560913086, "learning_rate": 1e-05, "loss": 0.4943, "mean_token_accuracy": 0.8648281097412109, "num_tokens": 282451881.0, "step": 7442 }, { "epoch": 0.9468261035491667, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.778550624847412, "learning_rate": 1e-05, "loss": 0.5682, "mean_token_accuracy": 0.8462036848068237, "num_tokens": 282486031.0, "step": 7443 }, { "epoch": 0.9469533138277573, "ewc_loss": 0.0947265625, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.329248905181885, "learning_rate": 1e-05, "loss": 0.4867, "mean_token_accuracy": 0.8739337921142578, "num_tokens": 282523753.0, "step": 7444 }, { "epoch": 0.9470805241063478, "ewc_loss": 0.125, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 7.058468818664551, "learning_rate": 1e-05, "loss": 0.5382, "mean_token_accuracy": 0.8568350076675415, "num_tokens": 282563322.0, "step": 7445 }, { "epoch": 0.9472077343849383, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 5.694876670837402, "learning_rate": 1e-05, "loss": 0.5186, "mean_token_accuracy": 0.8591339588165283, "num_tokens": 282602574.0, "step": 7446 }, { "epoch": 0.9473349446635289, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 7.002154350280762, "learning_rate": 1e-05, "loss": 0.5049, "mean_token_accuracy": 0.8677452802658081, "num_tokens": 282636608.0, "step": 7447 }, { "epoch": 0.9474621549421193, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.01861047744751, "learning_rate": 1e-05, "loss": 0.4672, "mean_token_accuracy": 0.8767278790473938, "num_tokens": 282675320.0, "step": 7448 }, { "epoch": 0.9475893652207098, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.641323089599609, "learning_rate": 1e-05, "loss": 0.5573, "mean_token_accuracy": 0.8577315211296082, "num_tokens": 282712537.0, "step": 7449 }, { "epoch": 0.9477165754993003, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.084647178649902, "learning_rate": 1e-05, "loss": 0.503, "mean_token_accuracy": 0.8664599657058716, "num_tokens": 282744620.0, "step": 7450 }, { "epoch": 0.9478437857778909, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.219313621520996, "learning_rate": 1e-05, "loss": 0.495, "mean_token_accuracy": 0.8700644373893738, "num_tokens": 282786480.0, "step": 7451 }, { "epoch": 0.9479709960564814, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.054818153381348, "learning_rate": 1e-05, "loss": 0.4874, "mean_token_accuracy": 0.8703829646110535, "num_tokens": 282819143.0, "step": 7452 }, { "epoch": 0.9480982063350719, "ewc_loss": 0.09765625, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.859555244445801, "learning_rate": 1e-05, "loss": 0.501, "mean_token_accuracy": 0.8642455339431763, "num_tokens": 282859658.0, "step": 7453 }, { "epoch": 0.9482254166136623, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 6.030887126922607, "learning_rate": 1e-05, "loss": 0.5179, "mean_token_accuracy": 0.8603512048721313, "num_tokens": 282896190.0, "step": 7454 }, { "epoch": 0.9483526268922529, "ewc_loss": 0.09375, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.629608631134033, "learning_rate": 1e-05, "loss": 0.5349, "mean_token_accuracy": 0.8520030975341797, "num_tokens": 282937458.0, "step": 7455 }, { "epoch": 0.9484798371708434, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 6.055264472961426, "learning_rate": 1e-05, "loss": 0.5067, "mean_token_accuracy": 0.8624346256256104, "num_tokens": 282970669.0, "step": 7456 }, { "epoch": 0.9486070474494339, "ewc_loss": 0.091796875, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.478711128234863, "learning_rate": 1e-05, "loss": 0.5072, "mean_token_accuracy": 0.858474850654602, "num_tokens": 283012372.0, "step": 7457 }, { "epoch": 0.9487342577280244, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.133926868438721, "learning_rate": 1e-05, "loss": 0.5411, "mean_token_accuracy": 0.8498153686523438, "num_tokens": 283046165.0, "step": 7458 }, { "epoch": 0.948861468006615, "ewc_loss": 0.08984375, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 7.534027099609375e-05, "grad_norm": 5.359139442443848, "learning_rate": 1e-05, "loss": 0.5089, "mean_token_accuracy": 0.8597530722618103, "num_tokens": 283086539.0, "step": 7459 }, { "epoch": 0.9489886782852054, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.301477909088135, "learning_rate": 1e-05, "loss": 0.5252, "mean_token_accuracy": 0.8588398694992065, "num_tokens": 283123643.0, "step": 7460 }, { "epoch": 0.9491158885637959, "ewc_loss": 0.08984375, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 7.534027099609375e-05, "grad_norm": 5.256969451904297, "learning_rate": 1e-05, "loss": 0.439, "mean_token_accuracy": 0.8797013163566589, "num_tokens": 283162354.0, "step": 7461 }, { "epoch": 0.9492430988423864, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.594603538513184, "learning_rate": 1e-05, "loss": 0.4856, "mean_token_accuracy": 0.8686556816101074, "num_tokens": 283196814.0, "step": 7462 }, { "epoch": 0.949370309120977, "ewc_loss": 0.091796875, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.2507100105285645, "learning_rate": 1e-05, "loss": 0.5249, "mean_token_accuracy": 0.8539463877677917, "num_tokens": 283238051.0, "step": 7463 }, { "epoch": 0.9494975193995675, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.92990255355835, "learning_rate": 1e-05, "loss": 0.5646, "mean_token_accuracy": 0.8524143695831299, "num_tokens": 283278025.0, "step": 7464 }, { "epoch": 0.949624729678158, "ewc_loss": 0.09765625, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.496067047119141, "learning_rate": 1e-05, "loss": 0.4877, "mean_token_accuracy": 0.8666819334030151, "num_tokens": 283311279.0, "step": 7465 }, { "epoch": 0.9497519399567486, "ewc_loss": 0.125, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 7.06680965423584, "learning_rate": 1e-05, "loss": 0.5527, "mean_token_accuracy": 0.853789210319519, "num_tokens": 283351350.0, "step": 7466 }, { "epoch": 0.949879150235339, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 5.8693623542785645, "learning_rate": 1e-05, "loss": 0.4993, "mean_token_accuracy": 0.8673571348190308, "num_tokens": 283387374.0, "step": 7467 }, { "epoch": 0.9500063605139295, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.830506801605225, "learning_rate": 1e-05, "loss": 0.5491, "mean_token_accuracy": 0.8563892245292664, "num_tokens": 283431305.0, "step": 7468 }, { "epoch": 0.95013357079252, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.066171646118164, "learning_rate": 1e-05, "loss": 0.5142, "mean_token_accuracy": 0.8625338673591614, "num_tokens": 283464941.0, "step": 7469 }, { "epoch": 0.9502607810711106, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.422685623168945, "learning_rate": 1e-05, "loss": 0.5914, "mean_token_accuracy": 0.8434428572654724, "num_tokens": 283507254.0, "step": 7470 }, { "epoch": 0.9503879913497011, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.069931507110596, "learning_rate": 1e-05, "loss": 0.5544, "mean_token_accuracy": 0.8494822978973389, "num_tokens": 283545117.0, "step": 7471 }, { "epoch": 0.9505152016282916, "ewc_loss": 0.099609375, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 6.022748947143555, "learning_rate": 1e-05, "loss": 0.5493, "mean_token_accuracy": 0.8494579792022705, "num_tokens": 283577786.0, "step": 7472 }, { "epoch": 0.950642411906882, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.03808069229126, "learning_rate": 1e-05, "loss": 0.4903, "mean_token_accuracy": 0.8679227828979492, "num_tokens": 283613486.0, "step": 7473 }, { "epoch": 0.9507696221854726, "ewc_loss": 0.0966796875, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.721678733825684, "learning_rate": 1e-05, "loss": 0.5088, "mean_token_accuracy": 0.8608724474906921, "num_tokens": 283654480.0, "step": 7474 }, { "epoch": 0.9508968324640631, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 6.050565242767334, "learning_rate": 1e-05, "loss": 0.5234, "mean_token_accuracy": 0.8553898334503174, "num_tokens": 283686089.0, "step": 7475 }, { "epoch": 0.9510240427426536, "ewc_loss": 0.09423828125, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.526101112365723, "learning_rate": 1e-05, "loss": 0.5013, "mean_token_accuracy": 0.8613066673278809, "num_tokens": 283726483.0, "step": 7476 }, { "epoch": 0.9511512530212441, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.120211124420166, "learning_rate": 1e-05, "loss": 0.499, "mean_token_accuracy": 0.8661430478096008, "num_tokens": 283756170.0, "step": 7477 }, { "epoch": 0.9512784632998347, "ewc_loss": 0.09130859375, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.387923717498779, "learning_rate": 1e-05, "loss": 0.4919, "mean_token_accuracy": 0.8634964227676392, "num_tokens": 283794950.0, "step": 7478 }, { "epoch": 0.9514056735784251, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.348174571990967, "learning_rate": 1e-05, "loss": 0.5969, "mean_token_accuracy": 0.8392657041549683, "num_tokens": 283835596.0, "step": 7479 }, { "epoch": 0.9515328838570156, "ewc_loss": 0.09033203125, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 7.581710815429688e-05, "grad_norm": 5.2548723220825195, "learning_rate": 1e-05, "loss": 0.5307, "mean_token_accuracy": 0.8525104522705078, "num_tokens": 283875380.0, "step": 7480 }, { "epoch": 0.9516600941356061, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.591762065887451, "learning_rate": 1e-05, "loss": 0.5331, "mean_token_accuracy": 0.8555600643157959, "num_tokens": 283910026.0, "step": 7481 }, { "epoch": 0.9517873044141967, "ewc_loss": 0.09228515625, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.273173809051514, "learning_rate": 1e-05, "loss": 0.4852, "mean_token_accuracy": 0.8654693365097046, "num_tokens": 283950533.0, "step": 7482 }, { "epoch": 0.9519145146927872, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 6.899315357208252, "learning_rate": 1e-05, "loss": 0.5373, "mean_token_accuracy": 0.8626906275749207, "num_tokens": 283980939.0, "step": 7483 }, { "epoch": 0.9520417249713777, "ewc_loss": 0.09716796875, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.490837574005127, "learning_rate": 1e-05, "loss": 0.5033, "mean_token_accuracy": 0.8623137474060059, "num_tokens": 284018923.0, "step": 7484 }, { "epoch": 0.9521689352499682, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 7.002573013305664, "learning_rate": 1e-05, "loss": 0.479, "mean_token_accuracy": 0.8784960508346558, "num_tokens": 284059752.0, "step": 7485 }, { "epoch": 0.9522961455285587, "ewc_loss": 0.10791015625, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 5.833216190338135, "learning_rate": 1e-05, "loss": 0.4955, "mean_token_accuracy": 0.8698400855064392, "num_tokens": 284103783.0, "step": 7486 }, { "epoch": 0.9524233558071492, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.804631233215332, "learning_rate": 1e-05, "loss": 0.6086, "mean_token_accuracy": 0.8403735160827637, "num_tokens": 284144595.0, "step": 7487 }, { "epoch": 0.9525505660857397, "ewc_loss": 0.11083984375, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.022552490234375, "learning_rate": 1e-05, "loss": 0.5331, "mean_token_accuracy": 0.8580948114395142, "num_tokens": 284182373.0, "step": 7488 }, { "epoch": 0.9526777763643303, "ewc_loss": 0.10986328125, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.434789180755615, "learning_rate": 1e-05, "loss": 0.5897, "mean_token_accuracy": 0.8394196033477783, "num_tokens": 284222537.0, "step": 7489 }, { "epoch": 0.9528049866429208, "ewc_loss": 0.10888671875, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.020813465118408, "learning_rate": 1e-05, "loss": 0.5353, "mean_token_accuracy": 0.8545385599136353, "num_tokens": 284264780.0, "step": 7490 }, { "epoch": 0.9529321969215112, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 6.062497615814209, "learning_rate": 1e-05, "loss": 0.4756, "mean_token_accuracy": 0.8749369382858276, "num_tokens": 284298663.0, "step": 7491 }, { "epoch": 0.9530594072001017, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 5.991586208343506, "learning_rate": 1e-05, "loss": 0.4825, "mean_token_accuracy": 0.8666051030158997, "num_tokens": 284330516.0, "step": 7492 }, { "epoch": 0.9531866174786923, "ewc_loss": 0.09814453125, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.784423828125, "learning_rate": 1e-05, "loss": 0.5371, "mean_token_accuracy": 0.8496789932250977, "num_tokens": 284370414.0, "step": 7493 }, { "epoch": 0.9533138277572828, "ewc_loss": 0.10205078125, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 5.949468612670898, "learning_rate": 1e-05, "loss": 0.4391, "mean_token_accuracy": 0.8832386136054993, "num_tokens": 284409079.0, "step": 7494 }, { "epoch": 0.9534410380358733, "ewc_loss": 0.09423828125, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.611461162567139, "learning_rate": 1e-05, "loss": 0.4738, "mean_token_accuracy": 0.8701673746109009, "num_tokens": 284451937.0, "step": 7495 }, { "epoch": 0.9535682483144639, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.957144737243652, "learning_rate": 1e-05, "loss": 0.5305, "mean_token_accuracy": 0.8568464517593384, "num_tokens": 284490174.0, "step": 7496 }, { "epoch": 0.9536954585930543, "ewc_loss": 0.0927734375, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.490259647369385, "learning_rate": 1e-05, "loss": 0.4776, "mean_token_accuracy": 0.8711074590682983, "num_tokens": 284530578.0, "step": 7497 }, { "epoch": 0.9538226688716448, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 6.0312981605529785, "learning_rate": 1e-05, "loss": 0.552, "mean_token_accuracy": 0.8479218482971191, "num_tokens": 284562412.0, "step": 7498 }, { "epoch": 0.9539498791502353, "ewc_loss": 0.09130859375, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.366301536560059, "learning_rate": 1e-05, "loss": 0.506, "mean_token_accuracy": 0.862921953201294, "num_tokens": 284601846.0, "step": 7499 }, { "epoch": 0.9540770894288259, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.166864395141602, "learning_rate": 1e-05, "loss": 0.5213, "mean_token_accuracy": 0.8564125299453735, "num_tokens": 284640918.0, "step": 7500 }, { "epoch": 0.9542042997074164, "ewc_loss": 0.08984375, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 7.534027099609375e-05, "grad_norm": 5.252736568450928, "learning_rate": 1e-05, "loss": 0.4736, "mean_token_accuracy": 0.868898868560791, "num_tokens": 284672268.0, "step": 7501 }, { "epoch": 0.9543315099860069, "ewc_loss": 0.10986328125, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.440649032592773, "learning_rate": 1e-05, "loss": 0.5087, "mean_token_accuracy": 0.8695355653762817, "num_tokens": 284708638.0, "step": 7502 }, { "epoch": 0.9544587202645973, "ewc_loss": 0.09033203125, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 7.581710815429688e-05, "grad_norm": 5.173251628875732, "learning_rate": 1e-05, "loss": 0.5774, "mean_token_accuracy": 0.8387395739555359, "num_tokens": 284747670.0, "step": 7503 }, { "epoch": 0.9545859305431879, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.843319416046143, "learning_rate": 1e-05, "loss": 0.5465, "mean_token_accuracy": 0.8549902439117432, "num_tokens": 284786561.0, "step": 7504 }, { "epoch": 0.9547131408217784, "ewc_loss": 0.09521484375, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.298485279083252, "learning_rate": 1e-05, "loss": 0.4937, "mean_token_accuracy": 0.8655544519424438, "num_tokens": 284828631.0, "step": 7505 }, { "epoch": 0.9548403511003689, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 0.00011396408081054688, "grad_norm": 7.1759209632873535, "learning_rate": 1e-05, "loss": 0.5287, "mean_token_accuracy": 0.8619378209114075, "num_tokens": 284865782.0, "step": 7506 }, { "epoch": 0.9549675613789594, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 5.758296012878418, "learning_rate": 1e-05, "loss": 0.4909, "mean_token_accuracy": 0.8677124977111816, "num_tokens": 284904000.0, "step": 7507 }, { "epoch": 0.95509477165755, "ewc_loss": 0.125, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 7.08853816986084, "learning_rate": 1e-05, "loss": 0.5323, "mean_token_accuracy": 0.8657487630844116, "num_tokens": 284946432.0, "step": 7508 }, { "epoch": 0.9552219819361404, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.117430686950684, "learning_rate": 1e-05, "loss": 0.524, "mean_token_accuracy": 0.8626339435577393, "num_tokens": 284985534.0, "step": 7509 }, { "epoch": 0.9553491922147309, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.6811628341674805, "learning_rate": 1e-05, "loss": 0.4939, "mean_token_accuracy": 0.8738255500793457, "num_tokens": 285023710.0, "step": 7510 }, { "epoch": 0.9554764024933214, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.138862609863281, "learning_rate": 1e-05, "loss": 0.5524, "mean_token_accuracy": 0.8493432998657227, "num_tokens": 285064182.0, "step": 7511 }, { "epoch": 0.955603612771912, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.268633842468262, "learning_rate": 1e-05, "loss": 0.4983, "mean_token_accuracy": 0.8657292127609253, "num_tokens": 285100896.0, "step": 7512 }, { "epoch": 0.9557308230505025, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.063965320587158, "learning_rate": 1e-05, "loss": 0.5156, "mean_token_accuracy": 0.8606155514717102, "num_tokens": 285139347.0, "step": 7513 }, { "epoch": 0.955858033329093, "ewc_loss": 0.09814453125, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.930513381958008, "learning_rate": 1e-05, "loss": 0.5123, "mean_token_accuracy": 0.8607681393623352, "num_tokens": 285183047.0, "step": 7514 }, { "epoch": 0.9559852436076836, "ewc_loss": 0.10205078125, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 6.002156734466553, "learning_rate": 1e-05, "loss": 0.5076, "mean_token_accuracy": 0.8614460229873657, "num_tokens": 285225407.0, "step": 7515 }, { "epoch": 0.956112453886274, "ewc_loss": 0.0947265625, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.705352783203125, "learning_rate": 1e-05, "loss": 0.4776, "mean_token_accuracy": 0.8674957752227783, "num_tokens": 285258465.0, "step": 7516 }, { "epoch": 0.9562396641648645, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.985173225402832, "learning_rate": 1e-05, "loss": 0.5103, "mean_token_accuracy": 0.8619109392166138, "num_tokens": 285297221.0, "step": 7517 }, { "epoch": 0.956366874443455, "ewc_loss": 0.09228515625, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.556341171264648, "learning_rate": 1e-05, "loss": 0.4914, "mean_token_accuracy": 0.8669556379318237, "num_tokens": 285338145.0, "step": 7518 }, { "epoch": 0.9564940847220456, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 6.014782905578613, "learning_rate": 1e-05, "loss": 0.497, "mean_token_accuracy": 0.8642117381095886, "num_tokens": 285381647.0, "step": 7519 }, { "epoch": 0.9566212950006361, "ewc_loss": 0.09130859375, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.464843273162842, "learning_rate": 1e-05, "loss": 0.556, "mean_token_accuracy": 0.841605544090271, "num_tokens": 285424468.0, "step": 7520 }, { "epoch": 0.9567485052792266, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.101741313934326, "learning_rate": 1e-05, "loss": 0.5024, "mean_token_accuracy": 0.8655067086219788, "num_tokens": 285470415.0, "step": 7521 }, { "epoch": 0.956875715557817, "ewc_loss": 0.09033203125, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 7.581710815429688e-05, "grad_norm": 5.318679332733154, "learning_rate": 1e-05, "loss": 0.5293, "mean_token_accuracy": 0.8494071960449219, "num_tokens": 285513299.0, "step": 7522 }, { "epoch": 0.9570029258364076, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.332417011260986, "learning_rate": 1e-05, "loss": 0.5676, "mean_token_accuracy": 0.8459630012512207, "num_tokens": 285555855.0, "step": 7523 }, { "epoch": 0.9571301361149981, "ewc_loss": 0.09033203125, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 7.581710815429688e-05, "grad_norm": 5.197336673736572, "learning_rate": 1e-05, "loss": 0.4842, "mean_token_accuracy": 0.8650429248809814, "num_tokens": 285589670.0, "step": 7524 }, { "epoch": 0.9572573463935886, "ewc_loss": 0.11474609375, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.694526195526123, "learning_rate": 1e-05, "loss": 0.5678, "mean_token_accuracy": 0.8505939245223999, "num_tokens": 285627257.0, "step": 7525 }, { "epoch": 0.9573845566721791, "ewc_loss": 0.09228515625, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.211409568786621, "learning_rate": 1e-05, "loss": 0.4899, "mean_token_accuracy": 0.8665184378623962, "num_tokens": 285670100.0, "step": 7526 }, { "epoch": 0.9575117669507697, "ewc_loss": 0.125, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 7.107707500457764, "learning_rate": 1e-05, "loss": 0.5193, "mean_token_accuracy": 0.8657504320144653, "num_tokens": 285706902.0, "step": 7527 }, { "epoch": 0.9576389772293601, "ewc_loss": 0.10107421875, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.591553688049316, "learning_rate": 1e-05, "loss": 0.5325, "mean_token_accuracy": 0.8576944470405579, "num_tokens": 285746952.0, "step": 7528 }, { "epoch": 0.9577661875079506, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 0.00011348724365234375, "grad_norm": 7.204444885253906, "learning_rate": 1e-05, "loss": 0.5022, "mean_token_accuracy": 0.8744657635688782, "num_tokens": 285781532.0, "step": 7529 }, { "epoch": 0.9578933977865411, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.057065486907959, "learning_rate": 1e-05, "loss": 0.5386, "mean_token_accuracy": 0.8564685583114624, "num_tokens": 285818725.0, "step": 7530 }, { "epoch": 0.9580206080651317, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.825173377990723, "learning_rate": 1e-05, "loss": 0.5629, "mean_token_accuracy": 0.8535253405570984, "num_tokens": 285855502.0, "step": 7531 }, { "epoch": 0.9581478183437222, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.177944183349609, "learning_rate": 1e-05, "loss": 0.5519, "mean_token_accuracy": 0.858896017074585, "num_tokens": 285893815.0, "step": 7532 }, { "epoch": 0.9582750286223127, "ewc_loss": 0.10791015625, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.3793044090271, "learning_rate": 1e-05, "loss": 0.5362, "mean_token_accuracy": 0.8582468032836914, "num_tokens": 285928709.0, "step": 7533 }, { "epoch": 0.9584022389009031, "ewc_loss": 0.10888671875, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.102409362792969, "learning_rate": 1e-05, "loss": 0.4601, "mean_token_accuracy": 0.8772721290588379, "num_tokens": 285959358.0, "step": 7534 }, { "epoch": 0.9585294491794937, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 6.002078533172607, "learning_rate": 1e-05, "loss": 0.5677, "mean_token_accuracy": 0.8429127931594849, "num_tokens": 285997643.0, "step": 7535 }, { "epoch": 0.9586566594580842, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.043138027191162, "learning_rate": 1e-05, "loss": 0.5051, "mean_token_accuracy": 0.8645159602165222, "num_tokens": 286032866.0, "step": 7536 }, { "epoch": 0.9587838697366747, "ewc_loss": 0.09619140625, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.747061252593994, "learning_rate": 1e-05, "loss": 0.5057, "mean_token_accuracy": 0.8611972332000732, "num_tokens": 286066585.0, "step": 7537 }, { "epoch": 0.9589110800152653, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 6.0184736251831055, "learning_rate": 1e-05, "loss": 0.5614, "mean_token_accuracy": 0.8469853401184082, "num_tokens": 286106151.0, "step": 7538 }, { "epoch": 0.9590382902938558, "ewc_loss": 0.0947265625, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.573812961578369, "learning_rate": 1e-05, "loss": 0.4837, "mean_token_accuracy": 0.8674789071083069, "num_tokens": 286142409.0, "step": 7539 }, { "epoch": 0.9591655005724462, "ewc_loss": 0.10205078125, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 6.061547756195068, "learning_rate": 1e-05, "loss": 0.5607, "mean_token_accuracy": 0.8501545786857605, "num_tokens": 286179408.0, "step": 7540 }, { "epoch": 0.9592927108510367, "ewc_loss": 0.091796875, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.435403823852539, "learning_rate": 1e-05, "loss": 0.5191, "mean_token_accuracy": 0.8596819043159485, "num_tokens": 286212767.0, "step": 7541 }, { "epoch": 0.9594199211296273, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.178329944610596, "learning_rate": 1e-05, "loss": 0.5116, "mean_token_accuracy": 0.860487699508667, "num_tokens": 286250176.0, "step": 7542 }, { "epoch": 0.9595471314082178, "ewc_loss": 0.0908203125, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 7.62939453125e-05, "grad_norm": 5.3068528175354, "learning_rate": 1e-05, "loss": 0.5073, "mean_token_accuracy": 0.8592032194137573, "num_tokens": 286292060.0, "step": 7543 }, { "epoch": 0.9596743416868083, "ewc_loss": 0.10986328125, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.4107890129089355, "learning_rate": 1e-05, "loss": 0.5006, "mean_token_accuracy": 0.8686218857765198, "num_tokens": 286333047.0, "step": 7544 }, { "epoch": 0.9598015519653988, "ewc_loss": 0.09033203125, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 7.581710815429688e-05, "grad_norm": 5.21613883972168, "learning_rate": 1e-05, "loss": 0.4809, "mean_token_accuracy": 0.8648516535758972, "num_tokens": 286373465.0, "step": 7545 }, { "epoch": 0.9599287622439893, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.758588790893555, "learning_rate": 1e-05, "loss": 0.577, "mean_token_accuracy": 0.8499524593353271, "num_tokens": 286414431.0, "step": 7546 }, { "epoch": 0.9600559725225798, "ewc_loss": 0.09375, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.315136909484863, "learning_rate": 1e-05, "loss": 0.5088, "mean_token_accuracy": 0.8581099510192871, "num_tokens": 286449431.0, "step": 7547 }, { "epoch": 0.9601831828011703, "ewc_loss": 0.125, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 7.059061527252197, "learning_rate": 1e-05, "loss": 0.5247, "mean_token_accuracy": 0.8631144762039185, "num_tokens": 286486563.0, "step": 7548 }, { "epoch": 0.9603103930797608, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 5.674465179443359, "learning_rate": 1e-05, "loss": 0.5445, "mean_token_accuracy": 0.8535019755363464, "num_tokens": 286528691.0, "step": 7549 }, { "epoch": 0.9604376033583514, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 0.00010967254638671875, "grad_norm": 7.034539222717285, "learning_rate": 1e-05, "loss": 0.5513, "mean_token_accuracy": 0.8577821850776672, "num_tokens": 286568726.0, "step": 7550 }, { "epoch": 0.9605648136369419, "ewc_loss": 0.11083984375, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.015279293060303, "learning_rate": 1e-05, "loss": 0.4965, "mean_token_accuracy": 0.8707785606384277, "num_tokens": 286608159.0, "step": 7551 }, { "epoch": 0.9606920239155323, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.684167861938477, "learning_rate": 1e-05, "loss": 0.5527, "mean_token_accuracy": 0.8577930331230164, "num_tokens": 286645494.0, "step": 7552 }, { "epoch": 0.9608192341941229, "ewc_loss": 0.11083984375, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.092559337615967, "learning_rate": 1e-05, "loss": 0.4873, "mean_token_accuracy": 0.8697925806045532, "num_tokens": 286679819.0, "step": 7553 }, { "epoch": 0.9609464444727134, "ewc_loss": 0.10595703125, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.265794277191162, "learning_rate": 1e-05, "loss": 0.5594, "mean_token_accuracy": 0.8454887866973877, "num_tokens": 286714116.0, "step": 7554 }, { "epoch": 0.9610736547513039, "ewc_loss": 0.10595703125, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.053233623504639, "learning_rate": 1e-05, "loss": 0.4783, "mean_token_accuracy": 0.8729900121688843, "num_tokens": 286746905.0, "step": 7555 }, { "epoch": 0.9612008650298944, "ewc_loss": 0.0986328125, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.904010772705078, "learning_rate": 1e-05, "loss": 0.4418, "mean_token_accuracy": 0.8826711177825928, "num_tokens": 286784724.0, "step": 7556 }, { "epoch": 0.961328075308485, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 6.030223846435547, "learning_rate": 1e-05, "loss": 0.4938, "mean_token_accuracy": 0.8658751249313354, "num_tokens": 286825672.0, "step": 7557 }, { "epoch": 0.9614552855870754, "ewc_loss": 0.09375, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.670836925506592, "learning_rate": 1e-05, "loss": 0.557, "mean_token_accuracy": 0.8492763638496399, "num_tokens": 286862072.0, "step": 7558 }, { "epoch": 0.9615824958656659, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 6.0313401222229, "learning_rate": 1e-05, "loss": 0.5144, "mean_token_accuracy": 0.8631615042686462, "num_tokens": 286903111.0, "step": 7559 }, { "epoch": 0.9617097061442564, "ewc_loss": 0.0927734375, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.504575252532959, "learning_rate": 1e-05, "loss": 0.4863, "mean_token_accuracy": 0.8659197092056274, "num_tokens": 286947957.0, "step": 7560 }, { "epoch": 0.961836916422847, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.108733177185059, "learning_rate": 1e-05, "loss": 0.5049, "mean_token_accuracy": 0.862159252166748, "num_tokens": 286983254.0, "step": 7561 }, { "epoch": 0.9619641267014375, "ewc_loss": 0.0908203125, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 7.62939453125e-05, "grad_norm": 5.374945640563965, "learning_rate": 1e-05, "loss": 0.432, "mean_token_accuracy": 0.8814979195594788, "num_tokens": 287015294.0, "step": 7562 }, { "epoch": 0.962091336980028, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.271758079528809, "learning_rate": 1e-05, "loss": 0.4999, "mean_token_accuracy": 0.8659407496452332, "num_tokens": 287053439.0, "step": 7563 }, { "epoch": 0.9622185472586186, "ewc_loss": 0.08984375, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 7.534027099609375e-05, "grad_norm": 5.262936592102051, "learning_rate": 1e-05, "loss": 0.5747, "mean_token_accuracy": 0.8378564119338989, "num_tokens": 287096970.0, "step": 7564 }, { "epoch": 0.962345757537209, "ewc_loss": 0.11181640625, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.541127681732178, "learning_rate": 1e-05, "loss": 0.5437, "mean_token_accuracy": 0.8541942238807678, "num_tokens": 287139179.0, "step": 7565 }, { "epoch": 0.9624729678157995, "ewc_loss": 0.091796875, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.2401251792907715, "learning_rate": 1e-05, "loss": 0.5138, "mean_token_accuracy": 0.8551450967788696, "num_tokens": 287177375.0, "step": 7566 }, { "epoch": 0.96260017809439, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.883432865142822, "learning_rate": 1e-05, "loss": 0.5132, "mean_token_accuracy": 0.863726019859314, "num_tokens": 287211296.0, "step": 7567 }, { "epoch": 0.9627273883729806, "ewc_loss": 0.09619140625, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.440737724304199, "learning_rate": 1e-05, "loss": 0.4924, "mean_token_accuracy": 0.8679372668266296, "num_tokens": 287246913.0, "step": 7568 }, { "epoch": 0.9628545986515711, "ewc_loss": 0.125, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 7.0759100914001465, "learning_rate": 1e-05, "loss": 0.5552, "mean_token_accuracy": 0.8578455448150635, "num_tokens": 287283639.0, "step": 7569 }, { "epoch": 0.9629818089301616, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 5.8245978355407715, "learning_rate": 1e-05, "loss": 0.5099, "mean_token_accuracy": 0.8675464391708374, "num_tokens": 287322694.0, "step": 7570 }, { "epoch": 0.963109019208752, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.963391304016113, "learning_rate": 1e-05, "loss": 0.5515, "mean_token_accuracy": 0.8523665070533752, "num_tokens": 287361103.0, "step": 7571 }, { "epoch": 0.9632362294873426, "ewc_loss": 0.11181640625, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.081748008728027, "learning_rate": 1e-05, "loss": 0.5702, "mean_token_accuracy": 0.8469447493553162, "num_tokens": 287400666.0, "step": 7572 }, { "epoch": 0.9633634397659331, "ewc_loss": 0.11083984375, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.494392395019531, "learning_rate": 1e-05, "loss": 0.4677, "mean_token_accuracy": 0.8774118423461914, "num_tokens": 287434729.0, "step": 7573 }, { "epoch": 0.9634906500445236, "ewc_loss": 0.109375, "ewc_loss_diag": 0.00014591217041015625, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.0755391120910645, "learning_rate": 1e-05, "loss": 0.4793, "mean_token_accuracy": 0.8726199865341187, "num_tokens": 287472194.0, "step": 7574 }, { "epoch": 0.9636178603231141, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.00014591217041015625, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.1009745597839355, "learning_rate": 1e-05, "loss": 0.5346, "mean_token_accuracy": 0.8561254143714905, "num_tokens": 287510139.0, "step": 7575 }, { "epoch": 0.9637450706017047, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.00014591217041015625, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.039569854736328, "learning_rate": 1e-05, "loss": 0.55, "mean_token_accuracy": 0.8500844240188599, "num_tokens": 287549529.0, "step": 7576 }, { "epoch": 0.9638722808802951, "ewc_loss": 0.0966796875, "ewc_loss_diag": 0.00014591217041015625, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.7971320152282715, "learning_rate": 1e-05, "loss": 0.5481, "mean_token_accuracy": 0.844840943813324, "num_tokens": 287596525.0, "step": 7577 }, { "epoch": 0.9639994911588856, "ewc_loss": 0.10302734375, "ewc_loss_diag": 0.00014591217041015625, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 6.018407344818115, "learning_rate": 1e-05, "loss": 0.461, "mean_token_accuracy": 0.8779184222221375, "num_tokens": 287632284.0, "step": 7578 }, { "epoch": 0.9641267014374761, "ewc_loss": 0.09423828125, "ewc_loss_diag": 0.00014591217041015625, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.600581169128418, "learning_rate": 1e-05, "loss": 0.4936, "mean_token_accuracy": 0.8663922548294067, "num_tokens": 287671867.0, "step": 7579 }, { "epoch": 0.9642539117160667, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.00014591217041015625, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 6.040193557739258, "learning_rate": 1e-05, "loss": 0.543, "mean_token_accuracy": 0.8514277935028076, "num_tokens": 287706715.0, "step": 7580 }, { "epoch": 0.9643811219946572, "ewc_loss": 0.0927734375, "ewc_loss_diag": 0.00014591217041015625, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.469162464141846, "learning_rate": 1e-05, "loss": 0.4942, "mean_token_accuracy": 0.8644112348556519, "num_tokens": 287746269.0, "step": 7581 }, { "epoch": 0.9645083322732477, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.00014591217041015625, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.13223934173584, "learning_rate": 1e-05, "loss": 0.5456, "mean_token_accuracy": 0.8529723286628723, "num_tokens": 287786194.0, "step": 7582 }, { "epoch": 0.9646355425518381, "ewc_loss": 0.08984375, "ewc_loss_diag": 0.00014591217041015625, "ewc_loss_parallel": 7.534027099609375e-05, "grad_norm": 5.349971294403076, "learning_rate": 1e-05, "loss": 0.5238, "mean_token_accuracy": 0.8539830446243286, "num_tokens": 287820873.0, "step": 7583 }, { "epoch": 0.9647627528304287, "ewc_loss": 0.10693359375, "ewc_loss_diag": 0.00014591217041015625, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.320741176605225, "learning_rate": 1e-05, "loss": 0.5159, "mean_token_accuracy": 0.8608974814414978, "num_tokens": 287847356.0, "step": 7584 }, { "epoch": 0.9648899631090192, "ewc_loss": 0.09033203125, "ewc_loss_diag": 0.00014591217041015625, "ewc_loss_parallel": 7.581710815429688e-05, "grad_norm": 5.24852180480957, "learning_rate": 1e-05, "loss": 0.5255, "mean_token_accuracy": 0.8553681373596191, "num_tokens": 287879611.0, "step": 7585 }, { "epoch": 0.9650171733876097, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.00014591217041015625, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.607785701751709, "learning_rate": 1e-05, "loss": 0.5775, "mean_token_accuracy": 0.8459104895591736, "num_tokens": 287920865.0, "step": 7586 }, { "epoch": 0.9651443836662003, "ewc_loss": 0.09228515625, "ewc_loss_diag": 0.00014591217041015625, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.234403133392334, "learning_rate": 1e-05, "loss": 0.5261, "mean_token_accuracy": 0.8548825979232788, "num_tokens": 287963997.0, "step": 7587 }, { "epoch": 0.9652715939447908, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.00014591217041015625, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 6.971364498138428, "learning_rate": 1e-05, "loss": 0.5575, "mean_token_accuracy": 0.8525445461273193, "num_tokens": 287997283.0, "step": 7588 }, { "epoch": 0.9653988042233812, "ewc_loss": 0.09912109375, "ewc_loss_diag": 0.00014591217041015625, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.531647205352783, "learning_rate": 1e-05, "loss": 0.589, "mean_token_accuracy": 0.834877610206604, "num_tokens": 288037876.0, "step": 7589 }, { "epoch": 0.9655260145019717, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.00014591217041015625, "ewc_loss_parallel": 0.00011110305786132812, "grad_norm": 7.073480606079102, "learning_rate": 1e-05, "loss": 0.5955, "mean_token_accuracy": 0.8457766771316528, "num_tokens": 288078835.0, "step": 7590 }, { "epoch": 0.9656532247805623, "ewc_loss": 0.10986328125, "ewc_loss_diag": 0.00014591217041015625, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 5.88422155380249, "learning_rate": 1e-05, "loss": 0.5397, "mean_token_accuracy": 0.8540468811988831, "num_tokens": 288118342.0, "step": 7591 }, { "epoch": 0.9657804350591528, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.00014591217041015625, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.830427646636963, "learning_rate": 1e-05, "loss": 0.5276, "mean_token_accuracy": 0.8640788793563843, "num_tokens": 288159629.0, "step": 7592 }, { "epoch": 0.9659076453377433, "ewc_loss": 0.11181640625, "ewc_loss_diag": 0.00014591217041015625, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.067553520202637, "learning_rate": 1e-05, "loss": 0.5449, "mean_token_accuracy": 0.8544823527336121, "num_tokens": 288199887.0, "step": 7593 }, { "epoch": 0.9660348556163338, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.00014591217041015625, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.423160552978516, "learning_rate": 1e-05, "loss": 0.5457, "mean_token_accuracy": 0.851111888885498, "num_tokens": 288233958.0, "step": 7594 }, { "epoch": 0.9661620658949243, "ewc_loss": 0.109375, "ewc_loss_diag": 0.00014591217041015625, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.058883190155029, "learning_rate": 1e-05, "loss": 0.5283, "mean_token_accuracy": 0.8567045331001282, "num_tokens": 288274735.0, "step": 7595 }, { "epoch": 0.9662892761735148, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.00014591217041015625, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 6.046610355377197, "learning_rate": 1e-05, "loss": 0.4942, "mean_token_accuracy": 0.8637886047363281, "num_tokens": 288309699.0, "step": 7596 }, { "epoch": 0.9664164864521053, "ewc_loss": 0.10498046875, "ewc_loss_diag": 0.00014591217041015625, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.023029804229736, "learning_rate": 1e-05, "loss": 0.5238, "mean_token_accuracy": 0.859927237033844, "num_tokens": 288344397.0, "step": 7597 }, { "epoch": 0.9665436967306958, "ewc_loss": 0.09716796875, "ewc_loss_diag": 0.00014591217041015625, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.753342151641846, "learning_rate": 1e-05, "loss": 0.5009, "mean_token_accuracy": 0.8665417432785034, "num_tokens": 288385612.0, "step": 7598 }, { "epoch": 0.9666709070092864, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.00014591217041015625, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 6.030447483062744, "learning_rate": 1e-05, "loss": 0.5655, "mean_token_accuracy": 0.8462126851081848, "num_tokens": 288420570.0, "step": 7599 }, { "epoch": 0.9667981172878769, "ewc_loss": 0.09375, "ewc_loss_diag": 0.00014591217041015625, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.565280437469482, "learning_rate": 1e-05, "loss": 0.5628, "mean_token_accuracy": 0.8480571508407593, "num_tokens": 288456917.0, "step": 7600 }, { "epoch": 0.9669253275664673, "ewc_loss": 0.10205078125, "ewc_loss_diag": 0.00014591217041015625, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 6.124331951141357, "learning_rate": 1e-05, "loss": 0.5505, "mean_token_accuracy": 0.8459718227386475, "num_tokens": 288498228.0, "step": 7601 }, { "epoch": 0.9670525378450578, "ewc_loss": 0.09228515625, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.405911445617676, "learning_rate": 1e-05, "loss": 0.5707, "mean_token_accuracy": 0.8400936722755432, "num_tokens": 288534082.0, "step": 7602 }, { "epoch": 0.9671797481236484, "ewc_loss": 0.10595703125, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.2325544357299805, "learning_rate": 1e-05, "loss": 0.5327, "mean_token_accuracy": 0.8554128408432007, "num_tokens": 288576142.0, "step": 7603 }, { "epoch": 0.9673069584022389, "ewc_loss": 0.0908203125, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 7.62939453125e-05, "grad_norm": 5.284440040588379, "learning_rate": 1e-05, "loss": 0.5257, "mean_token_accuracy": 0.8538678884506226, "num_tokens": 288609707.0, "step": 7604 }, { "epoch": 0.9674341686808294, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.491635799407959, "learning_rate": 1e-05, "loss": 0.5863, "mean_token_accuracy": 0.844964861869812, "num_tokens": 288648228.0, "step": 7605 }, { "epoch": 0.96756137895942, "ewc_loss": 0.091796875, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.224756717681885, "learning_rate": 1e-05, "loss": 0.4893, "mean_token_accuracy": 0.8651348948478699, "num_tokens": 288689845.0, "step": 7606 }, { "epoch": 0.9676885892380104, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.845872402191162, "learning_rate": 1e-05, "loss": 0.4878, "mean_token_accuracy": 0.8732710480690002, "num_tokens": 288729571.0, "step": 7607 }, { "epoch": 0.9678157995166009, "ewc_loss": 0.0966796875, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.391262054443359, "learning_rate": 1e-05, "loss": 0.4672, "mean_token_accuracy": 0.8736910223960876, "num_tokens": 288764070.0, "step": 7608 }, { "epoch": 0.9679430097951914, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 0.00011205673217773438, "grad_norm": 7.078114986419678, "learning_rate": 1e-05, "loss": 0.5317, "mean_token_accuracy": 0.8626871705055237, "num_tokens": 288805557.0, "step": 7609 }, { "epoch": 0.968070220073782, "ewc_loss": 0.10693359375, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 5.7834601402282715, "learning_rate": 1e-05, "loss": 0.4947, "mean_token_accuracy": 0.8686238527297974, "num_tokens": 288840567.0, "step": 7610 }, { "epoch": 0.9681974303523725, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 6.926141738891602, "learning_rate": 1e-05, "loss": 0.5275, "mean_token_accuracy": 0.8595261573791504, "num_tokens": 288881952.0, "step": 7611 }, { "epoch": 0.968324640630963, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.049960136413574, "learning_rate": 1e-05, "loss": 0.5374, "mean_token_accuracy": 0.8592998385429382, "num_tokens": 288919208.0, "step": 7612 }, { "epoch": 0.9684518509095535, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.537406921386719, "learning_rate": 1e-05, "loss": 0.5346, "mean_token_accuracy": 0.8584831953048706, "num_tokens": 288961622.0, "step": 7613 }, { "epoch": 0.968579061188144, "ewc_loss": 0.10986328125, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.071990966796875, "learning_rate": 1e-05, "loss": 0.4922, "mean_token_accuracy": 0.8694426417350769, "num_tokens": 288999799.0, "step": 7614 }, { "epoch": 0.9687062714667345, "ewc_loss": 0.10205078125, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 6.137818336486816, "learning_rate": 1e-05, "loss": 0.4767, "mean_token_accuracy": 0.870856761932373, "num_tokens": 289040434.0, "step": 7615 }, { "epoch": 0.968833481745325, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.039966583251953, "learning_rate": 1e-05, "loss": 0.5211, "mean_token_accuracy": 0.8572680950164795, "num_tokens": 289072847.0, "step": 7616 }, { "epoch": 0.9689606920239155, "ewc_loss": 0.09765625, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.8247857093811035, "learning_rate": 1e-05, "loss": 0.569, "mean_token_accuracy": 0.8462255001068115, "num_tokens": 289114838.0, "step": 7617 }, { "epoch": 0.9690879023025061, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 6.047728538513184, "learning_rate": 1e-05, "loss": 0.5345, "mean_token_accuracy": 0.8533543944358826, "num_tokens": 289151937.0, "step": 7618 }, { "epoch": 0.9692151125810966, "ewc_loss": 0.09375, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.612724304199219, "learning_rate": 1e-05, "loss": 0.4969, "mean_token_accuracy": 0.860893189907074, "num_tokens": 289188630.0, "step": 7619 }, { "epoch": 0.969342322859687, "ewc_loss": 0.10302734375, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 6.048102855682373, "learning_rate": 1e-05, "loss": 0.5098, "mean_token_accuracy": 0.8633404970169067, "num_tokens": 289229412.0, "step": 7620 }, { "epoch": 0.9694695331382776, "ewc_loss": 0.0927734375, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.466989994049072, "learning_rate": 1e-05, "loss": 0.4706, "mean_token_accuracy": 0.8704447150230408, "num_tokens": 289265195.0, "step": 7621 }, { "epoch": 0.9695967434168681, "ewc_loss": 0.10400390625, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 6.13095235824585, "learning_rate": 1e-05, "loss": 0.5413, "mean_token_accuracy": 0.8504475355148315, "num_tokens": 289306203.0, "step": 7622 }, { "epoch": 0.9697239536954586, "ewc_loss": 0.09130859375, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.358449459075928, "learning_rate": 1e-05, "loss": 0.5283, "mean_token_accuracy": 0.8556692600250244, "num_tokens": 289345196.0, "step": 7623 }, { "epoch": 0.9698511639740491, "ewc_loss": 0.10693359375, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.311294078826904, "learning_rate": 1e-05, "loss": 0.4918, "mean_token_accuracy": 0.8697912693023682, "num_tokens": 289377221.0, "step": 7624 }, { "epoch": 0.9699783742526397, "ewc_loss": 0.09033203125, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 7.581710815429688e-05, "grad_norm": 5.257923603057861, "learning_rate": 1e-05, "loss": 0.5262, "mean_token_accuracy": 0.8563804626464844, "num_tokens": 289413196.0, "step": 7625 }, { "epoch": 0.9701055845312301, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.598071575164795, "learning_rate": 1e-05, "loss": 0.5382, "mean_token_accuracy": 0.8540806770324707, "num_tokens": 289452774.0, "step": 7626 }, { "epoch": 0.9702327948098206, "ewc_loss": 0.0927734375, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.243625640869141, "learning_rate": 1e-05, "loss": 0.5011, "mean_token_accuracy": 0.8613557815551758, "num_tokens": 289489806.0, "step": 7627 }, { "epoch": 0.9703600050884111, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 6.944844722747803, "learning_rate": 1e-05, "loss": 0.5414, "mean_token_accuracy": 0.8590438365936279, "num_tokens": 289528362.0, "step": 7628 }, { "epoch": 0.9704872153670017, "ewc_loss": 0.0986328125, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.4991631507873535, "learning_rate": 1e-05, "loss": 0.5012, "mean_token_accuracy": 0.8605833649635315, "num_tokens": 289562725.0, "step": 7629 }, { "epoch": 0.9706144256455922, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 0.00011205673217773438, "grad_norm": 7.094231605529785, "learning_rate": 1e-05, "loss": 0.564, "mean_token_accuracy": 0.8553025126457214, "num_tokens": 289605968.0, "step": 7630 }, { "epoch": 0.9707416359241827, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 5.892921447753906, "learning_rate": 1e-05, "loss": 0.5342, "mean_token_accuracy": 0.8527789115905762, "num_tokens": 289639065.0, "step": 7631 }, { "epoch": 0.9708688462027731, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.853728294372559, "learning_rate": 1e-05, "loss": 0.6056, "mean_token_accuracy": 0.8387788534164429, "num_tokens": 289679616.0, "step": 7632 }, { "epoch": 0.9709960564813637, "ewc_loss": 0.11181640625, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.0830888748168945, "learning_rate": 1e-05, "loss": 0.5249, "mean_token_accuracy": 0.8618515729904175, "num_tokens": 289713189.0, "step": 7633 }, { "epoch": 0.9711232667599542, "ewc_loss": 0.109375, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.440948009490967, "learning_rate": 1e-05, "loss": 0.5035, "mean_token_accuracy": 0.8652809858322144, "num_tokens": 289751406.0, "step": 7634 }, { "epoch": 0.9712504770385447, "ewc_loss": 0.10986328125, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.0677103996276855, "learning_rate": 1e-05, "loss": 0.5169, "mean_token_accuracy": 0.8610713481903076, "num_tokens": 289795735.0, "step": 7635 }, { "epoch": 0.9713776873171353, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 6.0485334396362305, "learning_rate": 1e-05, "loss": 0.5557, "mean_token_accuracy": 0.8478351831436157, "num_tokens": 289833574.0, "step": 7636 }, { "epoch": 0.9715048975957258, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.041712760925293, "learning_rate": 1e-05, "loss": 0.4816, "mean_token_accuracy": 0.8714054822921753, "num_tokens": 289871922.0, "step": 7637 }, { "epoch": 0.9716321078743162, "ewc_loss": 0.09619140625, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.74850606918335, "learning_rate": 1e-05, "loss": 0.5421, "mean_token_accuracy": 0.8538458943367004, "num_tokens": 289913517.0, "step": 7638 }, { "epoch": 0.9717593181529067, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 6.043430805206299, "learning_rate": 1e-05, "loss": 0.5219, "mean_token_accuracy": 0.8581982254981995, "num_tokens": 289957046.0, "step": 7639 }, { "epoch": 0.9718865284314973, "ewc_loss": 0.09375, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.551578521728516, "learning_rate": 1e-05, "loss": 0.512, "mean_token_accuracy": 0.859380841255188, "num_tokens": 289997983.0, "step": 7640 }, { "epoch": 0.9720137387100878, "ewc_loss": 0.10302734375, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 6.093447208404541, "learning_rate": 1e-05, "loss": 0.4993, "mean_token_accuracy": 0.8626269102096558, "num_tokens": 290035923.0, "step": 7641 }, { "epoch": 0.9721409489886783, "ewc_loss": 0.0908203125, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 7.62939453125e-05, "grad_norm": 5.4069013595581055, "learning_rate": 1e-05, "loss": 0.4706, "mean_token_accuracy": 0.8688226938247681, "num_tokens": 290075152.0, "step": 7642 }, { "epoch": 0.9722681592672688, "ewc_loss": 0.10595703125, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.249549388885498, "learning_rate": 1e-05, "loss": 0.4698, "mean_token_accuracy": 0.875034749507904, "num_tokens": 290108360.0, "step": 7643 }, { "epoch": 0.9723953695458593, "ewc_loss": 0.08935546875, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 7.486343383789062e-05, "grad_norm": 5.276636123657227, "learning_rate": 1e-05, "loss": 0.47, "mean_token_accuracy": 0.8684782981872559, "num_tokens": 290148698.0, "step": 7644 }, { "epoch": 0.9725225798244498, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.515038013458252, "learning_rate": 1e-05, "loss": 0.5002, "mean_token_accuracy": 0.8675504922866821, "num_tokens": 290185847.0, "step": 7645 }, { "epoch": 0.9726497901030403, "ewc_loss": 0.0908203125, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 7.62939453125e-05, "grad_norm": 5.224177837371826, "learning_rate": 1e-05, "loss": 0.4957, "mean_token_accuracy": 0.8618903160095215, "num_tokens": 290228763.0, "step": 7646 }, { "epoch": 0.9727770003816308, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.878321647644043, "learning_rate": 1e-05, "loss": 0.5688, "mean_token_accuracy": 0.857299268245697, "num_tokens": 290262040.0, "step": 7647 }, { "epoch": 0.9729042106602214, "ewc_loss": 0.09619140625, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.404896259307861, "learning_rate": 1e-05, "loss": 0.4871, "mean_token_accuracy": 0.8658303022384644, "num_tokens": 290302205.0, "step": 7648 }, { "epoch": 0.9730314209388119, "ewc_loss": 0.125, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 0.00011014938354492188, "grad_norm": 7.104669094085693, "learning_rate": 1e-05, "loss": 0.5217, "mean_token_accuracy": 0.8646056056022644, "num_tokens": 290340132.0, "step": 7649 }, { "epoch": 0.9731586312174023, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 5.799577236175537, "learning_rate": 1e-05, "loss": 0.4349, "mean_token_accuracy": 0.8861150741577148, "num_tokens": 290375808.0, "step": 7650 }, { "epoch": 0.9732858414959928, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 6.964931011199951, "learning_rate": 1e-05, "loss": 0.5617, "mean_token_accuracy": 0.8566290140151978, "num_tokens": 290415538.0, "step": 7651 }, { "epoch": 0.9734130517745834, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.070620536804199, "learning_rate": 1e-05, "loss": 0.5108, "mean_token_accuracy": 0.868488609790802, "num_tokens": 290453363.0, "step": 7652 }, { "epoch": 0.9735402620531739, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.5975775718688965, "learning_rate": 1e-05, "loss": 0.5425, "mean_token_accuracy": 0.8526142239570618, "num_tokens": 290483869.0, "step": 7653 }, { "epoch": 0.9736674723317644, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.095884323120117, "learning_rate": 1e-05, "loss": 0.4924, "mean_token_accuracy": 0.8692922592163086, "num_tokens": 290521772.0, "step": 7654 }, { "epoch": 0.973794682610355, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.172295093536377, "learning_rate": 1e-05, "loss": 0.5633, "mean_token_accuracy": 0.8453794717788696, "num_tokens": 290563226.0, "step": 7655 }, { "epoch": 0.9739218928889454, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.055653095245361, "learning_rate": 1e-05, "loss": 0.5068, "mean_token_accuracy": 0.8598009943962097, "num_tokens": 290602209.0, "step": 7656 }, { "epoch": 0.9740491031675359, "ewc_loss": 0.09716796875, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.8486738204956055, "learning_rate": 1e-05, "loss": 0.5361, "mean_token_accuracy": 0.8505577445030212, "num_tokens": 290645554.0, "step": 7657 }, { "epoch": 0.9741763134461264, "ewc_loss": 0.10205078125, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 6.0176682472229, "learning_rate": 1e-05, "loss": 0.4934, "mean_token_accuracy": 0.8684819340705872, "num_tokens": 290683982.0, "step": 7658 }, { "epoch": 0.974303523724717, "ewc_loss": 0.09423828125, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.654988765716553, "learning_rate": 1e-05, "loss": 0.5304, "mean_token_accuracy": 0.8517554998397827, "num_tokens": 290717720.0, "step": 7659 }, { "epoch": 0.9744307340033075, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 6.003188133239746, "learning_rate": 1e-05, "loss": 0.5168, "mean_token_accuracy": 0.8578422665596008, "num_tokens": 290759711.0, "step": 7660 }, { "epoch": 0.974557944281898, "ewc_loss": 0.0927734375, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.521387577056885, "learning_rate": 1e-05, "loss": 0.4937, "mean_token_accuracy": 0.8627469539642334, "num_tokens": 290795767.0, "step": 7661 }, { "epoch": 0.9746851545604885, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 6.053344249725342, "learning_rate": 1e-05, "loss": 0.4909, "mean_token_accuracy": 0.8667216300964355, "num_tokens": 290839148.0, "step": 7662 }, { "epoch": 0.974812364839079, "ewc_loss": 0.09033203125, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 7.581710815429688e-05, "grad_norm": 5.394922733306885, "learning_rate": 1e-05, "loss": 0.4977, "mean_token_accuracy": 0.8636021018028259, "num_tokens": 290878325.0, "step": 7663 }, { "epoch": 0.9749395751176695, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.195462226867676, "learning_rate": 1e-05, "loss": 0.5391, "mean_token_accuracy": 0.8575437664985657, "num_tokens": 290913053.0, "step": 7664 }, { "epoch": 0.97506678539626, "ewc_loss": 0.08935546875, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 7.486343383789062e-05, "grad_norm": 5.263009548187256, "learning_rate": 1e-05, "loss": 0.474, "mean_token_accuracy": 0.8683146238327026, "num_tokens": 290954984.0, "step": 7665 }, { "epoch": 0.9751939956748505, "ewc_loss": 0.11083984375, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.474183559417725, "learning_rate": 1e-05, "loss": 0.5489, "mean_token_accuracy": 0.8527752757072449, "num_tokens": 290996052.0, "step": 7666 }, { "epoch": 0.9753212059534411, "ewc_loss": 0.09033203125, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 7.581710815429688e-05, "grad_norm": 5.192525386810303, "learning_rate": 1e-05, "loss": 0.5287, "mean_token_accuracy": 0.8490620851516724, "num_tokens": 291034021.0, "step": 7667 }, { "epoch": 0.9754484162320316, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.85259485244751, "learning_rate": 1e-05, "loss": 0.4873, "mean_token_accuracy": 0.8725154995918274, "num_tokens": 291064216.0, "step": 7668 }, { "epoch": 0.975575626510622, "ewc_loss": 0.0947265625, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.342244625091553, "learning_rate": 1e-05, "loss": 0.4895, "mean_token_accuracy": 0.868016242980957, "num_tokens": 291100119.0, "step": 7669 }, { "epoch": 0.9757028367892125, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 0.0001125335693359375, "grad_norm": 7.147241115570068, "learning_rate": 1e-05, "loss": 0.5432, "mean_token_accuracy": 0.8599780201911926, "num_tokens": 291136709.0, "step": 7670 }, { "epoch": 0.9758300470678031, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 5.780332565307617, "learning_rate": 1e-05, "loss": 0.5739, "mean_token_accuracy": 0.844014585018158, "num_tokens": 291175038.0, "step": 7671 }, { "epoch": 0.9759572573463936, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 7.031915187835693, "learning_rate": 1e-05, "loss": 0.5257, "mean_token_accuracy": 0.8621509075164795, "num_tokens": 291206950.0, "step": 7672 }, { "epoch": 0.9760844676249841, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.09413480758667, "learning_rate": 1e-05, "loss": 0.5516, "mean_token_accuracy": 0.8531444072723389, "num_tokens": 291247847.0, "step": 7673 }, { "epoch": 0.9762116779035747, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.624835968017578, "learning_rate": 1e-05, "loss": 0.4766, "mean_token_accuracy": 0.8745647668838501, "num_tokens": 291282262.0, "step": 7674 }, { "epoch": 0.9763388881821651, "ewc_loss": 0.11083984375, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.12057638168335, "learning_rate": 1e-05, "loss": 0.5131, "mean_token_accuracy": 0.8610179424285889, "num_tokens": 291315090.0, "step": 7675 }, { "epoch": 0.9764660984607556, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.2009782791137695, "learning_rate": 1e-05, "loss": 0.4879, "mean_token_accuracy": 0.8682976961135864, "num_tokens": 291356380.0, "step": 7676 }, { "epoch": 0.9765933087393461, "ewc_loss": 0.10693359375, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.058511734008789, "learning_rate": 1e-05, "loss": 0.5186, "mean_token_accuracy": 0.860875129699707, "num_tokens": 291401133.0, "step": 7677 }, { "epoch": 0.9767205190179367, "ewc_loss": 0.09765625, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.875789642333984, "learning_rate": 1e-05, "loss": 0.5057, "mean_token_accuracy": 0.8627305030822754, "num_tokens": 291439999.0, "step": 7678 }, { "epoch": 0.9768477292965272, "ewc_loss": 0.10302734375, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 6.029590606689453, "learning_rate": 1e-05, "loss": 0.4985, "mean_token_accuracy": 0.8658410310745239, "num_tokens": 291476784.0, "step": 7679 }, { "epoch": 0.9769749395751177, "ewc_loss": 0.09375, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.669089317321777, "learning_rate": 1e-05, "loss": 0.4722, "mean_token_accuracy": 0.8719562292098999, "num_tokens": 291517677.0, "step": 7680 }, { "epoch": 0.9771021498537081, "ewc_loss": 0.10205078125, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 6.033615589141846, "learning_rate": 1e-05, "loss": 0.4862, "mean_token_accuracy": 0.8657245635986328, "num_tokens": 291551838.0, "step": 7681 }, { "epoch": 0.9772293601322987, "ewc_loss": 0.0927734375, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.522771835327148, "learning_rate": 1e-05, "loss": 0.4875, "mean_token_accuracy": 0.8707680702209473, "num_tokens": 291591192.0, "step": 7682 }, { "epoch": 0.9773565704108892, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.095610618591309, "learning_rate": 1e-05, "loss": 0.5273, "mean_token_accuracy": 0.8615620732307434, "num_tokens": 291626885.0, "step": 7683 }, { "epoch": 0.9774837806894797, "ewc_loss": 0.0908203125, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 7.62939453125e-05, "grad_norm": 5.391514778137207, "learning_rate": 1e-05, "loss": 0.4897, "mean_token_accuracy": 0.8633813261985779, "num_tokens": 291666421.0, "step": 7684 }, { "epoch": 0.9776109909680702, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.263772487640381, "learning_rate": 1e-05, "loss": 0.5231, "mean_token_accuracy": 0.858417272567749, "num_tokens": 291704973.0, "step": 7685 }, { "epoch": 0.9777382012466608, "ewc_loss": 0.08984375, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 7.534027099609375e-05, "grad_norm": 5.263954162597656, "learning_rate": 1e-05, "loss": 0.5157, "mean_token_accuracy": 0.8557998538017273, "num_tokens": 291750279.0, "step": 7686 }, { "epoch": 0.9778654115252512, "ewc_loss": 0.11181640625, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.55114221572876, "learning_rate": 1e-05, "loss": 0.5428, "mean_token_accuracy": 0.8582101464271545, "num_tokens": 291784390.0, "step": 7687 }, { "epoch": 0.9779926218038417, "ewc_loss": 0.09130859375, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.218914031982422, "learning_rate": 1e-05, "loss": 0.4769, "mean_token_accuracy": 0.8699315786361694, "num_tokens": 291826385.0, "step": 7688 }, { "epoch": 0.9781198320824323, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.937806606292725, "learning_rate": 1e-05, "loss": 0.5606, "mean_token_accuracy": 0.8517350554466248, "num_tokens": 291858129.0, "step": 7689 }, { "epoch": 0.9782470423610228, "ewc_loss": 0.09619140625, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.428454399108887, "learning_rate": 1e-05, "loss": 0.5786, "mean_token_accuracy": 0.8473502993583679, "num_tokens": 291902763.0, "step": 7690 }, { "epoch": 0.9783742526396133, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 0.00011301040649414062, "grad_norm": 7.147783279418945, "learning_rate": 1e-05, "loss": 0.5016, "mean_token_accuracy": 0.8748823404312134, "num_tokens": 291940686.0, "step": 7691 }, { "epoch": 0.9785014629182038, "ewc_loss": 0.10791015625, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 5.862889766693115, "learning_rate": 1e-05, "loss": 0.514, "mean_token_accuracy": 0.8618465662002563, "num_tokens": 291976849.0, "step": 7692 }, { "epoch": 0.9786286731967943, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.955470085144043, "learning_rate": 1e-05, "loss": 0.485, "mean_token_accuracy": 0.8751804232597351, "num_tokens": 292014105.0, "step": 7693 }, { "epoch": 0.9787558834753848, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.105955600738525, "learning_rate": 1e-05, "loss": 0.5176, "mean_token_accuracy": 0.864205002784729, "num_tokens": 292055083.0, "step": 7694 }, { "epoch": 0.9788830937539753, "ewc_loss": 0.11083984375, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.540894508361816, "learning_rate": 1e-05, "loss": 0.5104, "mean_token_accuracy": 0.8635186553001404, "num_tokens": 292089713.0, "step": 7695 }, { "epoch": 0.9790103040325658, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.1060967445373535, "learning_rate": 1e-05, "loss": 0.5486, "mean_token_accuracy": 0.8510672450065613, "num_tokens": 292128139.0, "step": 7696 }, { "epoch": 0.9791375143111564, "ewc_loss": 0.10205078125, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 6.148486614227295, "learning_rate": 1e-05, "loss": 0.5224, "mean_token_accuracy": 0.8588153123855591, "num_tokens": 292164674.0, "step": 7697 }, { "epoch": 0.9792647245897469, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.055535316467285, "learning_rate": 1e-05, "loss": 0.4961, "mean_token_accuracy": 0.8636132478713989, "num_tokens": 292197175.0, "step": 7698 }, { "epoch": 0.9793919348683373, "ewc_loss": 0.0966796875, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.830312252044678, "learning_rate": 1e-05, "loss": 0.5478, "mean_token_accuracy": 0.847277820110321, "num_tokens": 292237051.0, "step": 7699 }, { "epoch": 0.9795191451469278, "ewc_loss": 0.10205078125, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 6.025276184082031, "learning_rate": 1e-05, "loss": 0.472, "mean_token_accuracy": 0.8713575005531311, "num_tokens": 292272817.0, "step": 7700 }, { "epoch": 0.9796463554255184, "ewc_loss": 0.09423828125, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.6263580322265625, "learning_rate": 1e-05, "loss": 0.5164, "mean_token_accuracy": 0.8598805069923401, "num_tokens": 292310239.0, "step": 7701 }, { "epoch": 0.9797735657041089, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 6.042938232421875, "learning_rate": 1e-05, "loss": 0.5244, "mean_token_accuracy": 0.8596084713935852, "num_tokens": 292351090.0, "step": 7702 }, { "epoch": 0.9799007759826994, "ewc_loss": 0.091796875, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.4834160804748535, "learning_rate": 1e-05, "loss": 0.5665, "mean_token_accuracy": 0.8430850505828857, "num_tokens": 292390500.0, "step": 7703 }, { "epoch": 0.98002798626129, "ewc_loss": 0.10400390625, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 6.12908411026001, "learning_rate": 1e-05, "loss": 0.5435, "mean_token_accuracy": 0.8534349203109741, "num_tokens": 292423483.0, "step": 7704 }, { "epoch": 0.9801551965398804, "ewc_loss": 0.09033203125, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 7.581710815429688e-05, "grad_norm": 5.357734680175781, "learning_rate": 1e-05, "loss": 0.4554, "mean_token_accuracy": 0.8758907914161682, "num_tokens": 292460791.0, "step": 7705 }, { "epoch": 0.9802824068184709, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.317917823791504, "learning_rate": 1e-05, "loss": 0.471, "mean_token_accuracy": 0.8767762780189514, "num_tokens": 292502214.0, "step": 7706 }, { "epoch": 0.9804096170970614, "ewc_loss": 0.08984375, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 7.534027099609375e-05, "grad_norm": 5.252335548400879, "learning_rate": 1e-05, "loss": 0.5349, "mean_token_accuracy": 0.8482221961021423, "num_tokens": 292541845.0, "step": 7707 }, { "epoch": 0.980536827375652, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.640128135681152, "learning_rate": 1e-05, "loss": 0.4861, "mean_token_accuracy": 0.8730391263961792, "num_tokens": 292573932.0, "step": 7708 }, { "epoch": 0.9806640376542425, "ewc_loss": 0.09228515625, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.237478256225586, "learning_rate": 1e-05, "loss": 0.4757, "mean_token_accuracy": 0.8682824373245239, "num_tokens": 292615932.0, "step": 7709 }, { "epoch": 0.980791247932833, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 7.024651527404785, "learning_rate": 1e-05, "loss": 0.5682, "mean_token_accuracy": 0.8492723703384399, "num_tokens": 292648680.0, "step": 7710 }, { "epoch": 0.9809184582114235, "ewc_loss": 0.0986328125, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.5386738777160645, "learning_rate": 1e-05, "loss": 0.4691, "mean_token_accuracy": 0.8740912675857544, "num_tokens": 292687987.0, "step": 7711 }, { "epoch": 0.981045668490014, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 7.134012222290039, "learning_rate": 1e-05, "loss": 0.5431, "mean_token_accuracy": 0.8584828972816467, "num_tokens": 292732243.0, "step": 7712 }, { "epoch": 0.9811728787686045, "ewc_loss": 0.10986328125, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 5.963653564453125, "learning_rate": 1e-05, "loss": 0.4944, "mean_token_accuracy": 0.8689759373664856, "num_tokens": 292764869.0, "step": 7713 }, { "epoch": 0.981300089047195, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.8418965339660645, "learning_rate": 1e-05, "loss": 0.5038, "mean_token_accuracy": 0.8707525730133057, "num_tokens": 292806401.0, "step": 7714 }, { "epoch": 0.9814272993257855, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.120696067810059, "learning_rate": 1e-05, "loss": 0.5005, "mean_token_accuracy": 0.8690288066864014, "num_tokens": 292843165.0, "step": 7715 }, { "epoch": 0.9815545096043761, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.406747817993164, "learning_rate": 1e-05, "loss": 0.5184, "mean_token_accuracy": 0.8604026436805725, "num_tokens": 292885092.0, "step": 7716 }, { "epoch": 0.9816817198829666, "ewc_loss": 0.10791015625, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.092440128326416, "learning_rate": 1e-05, "loss": 0.5431, "mean_token_accuracy": 0.853381335735321, "num_tokens": 292920595.0, "step": 7717 }, { "epoch": 0.981808930161557, "ewc_loss": 0.09912109375, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 6.031979084014893, "learning_rate": 1e-05, "loss": 0.5328, "mean_token_accuracy": 0.8527184128761292, "num_tokens": 292954079.0, "step": 7718 }, { "epoch": 0.9819361404401475, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.04149055480957, "learning_rate": 1e-05, "loss": 0.4918, "mean_token_accuracy": 0.8666530847549438, "num_tokens": 292985554.0, "step": 7719 }, { "epoch": 0.9820633507187381, "ewc_loss": 0.09619140625, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.759347915649414, "learning_rate": 1e-05, "loss": 0.4982, "mean_token_accuracy": 0.8629289865493774, "num_tokens": 293021832.0, "step": 7720 }, { "epoch": 0.9821905609973286, "ewc_loss": 0.10205078125, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 6.026979446411133, "learning_rate": 1e-05, "loss": 0.5159, "mean_token_accuracy": 0.8605113625526428, "num_tokens": 293054220.0, "step": 7721 }, { "epoch": 0.9823177712759191, "ewc_loss": 0.09375, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.5796356201171875, "learning_rate": 1e-05, "loss": 0.5406, "mean_token_accuracy": 0.8511865735054016, "num_tokens": 293099348.0, "step": 7722 }, { "epoch": 0.9824449815545097, "ewc_loss": 0.10302734375, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 6.060269832611084, "learning_rate": 1e-05, "loss": 0.5136, "mean_token_accuracy": 0.8588826656341553, "num_tokens": 293134801.0, "step": 7723 }, { "epoch": 0.9825721918331001, "ewc_loss": 0.091796875, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.440490245819092, "learning_rate": 1e-05, "loss": 0.567, "mean_token_accuracy": 0.8391422033309937, "num_tokens": 293172023.0, "step": 7724 }, { "epoch": 0.9826994021116906, "ewc_loss": 0.10400390625, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 6.172543525695801, "learning_rate": 1e-05, "loss": 0.5071, "mean_token_accuracy": 0.8618630170822144, "num_tokens": 293210019.0, "step": 7725 }, { "epoch": 0.9828266123902811, "ewc_loss": 0.09033203125, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 7.581710815429688e-05, "grad_norm": 5.321819305419922, "learning_rate": 1e-05, "loss": 0.5003, "mean_token_accuracy": 0.8622533082962036, "num_tokens": 293249614.0, "step": 7726 }, { "epoch": 0.9829538226688717, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.379962921142578, "learning_rate": 1e-05, "loss": 0.487, "mean_token_accuracy": 0.8709354996681213, "num_tokens": 293289860.0, "step": 7727 }, { "epoch": 0.9830810329474622, "ewc_loss": 0.08984375, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 7.534027099609375e-05, "grad_norm": 5.227894306182861, "learning_rate": 1e-05, "loss": 0.5503, "mean_token_accuracy": 0.8493402004241943, "num_tokens": 293330337.0, "step": 7728 }, { "epoch": 0.9832082432260527, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.7301836013793945, "learning_rate": 1e-05, "loss": 0.5339, "mean_token_accuracy": 0.8620519042015076, "num_tokens": 293370062.0, "step": 7729 }, { "epoch": 0.9833354535046431, "ewc_loss": 0.09326171875, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.287924289703369, "learning_rate": 1e-05, "loss": 0.4989, "mean_token_accuracy": 0.8600885272026062, "num_tokens": 293409544.0, "step": 7730 }, { "epoch": 0.9834626637832337, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 0.00011110305786132812, "grad_norm": 7.0656023025512695, "learning_rate": 1e-05, "loss": 0.5912, "mean_token_accuracy": 0.8478829860687256, "num_tokens": 293453924.0, "step": 7731 }, { "epoch": 0.9835898740618242, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.628777027130127, "learning_rate": 1e-05, "loss": 0.4929, "mean_token_accuracy": 0.8672540187835693, "num_tokens": 293493759.0, "step": 7732 }, { "epoch": 0.9837170843404147, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 0.00010967254638671875, "grad_norm": 7.099642753601074, "learning_rate": 1e-05, "loss": 0.6036, "mean_token_accuracy": 0.8401991128921509, "num_tokens": 293530155.0, "step": 7733 }, { "epoch": 0.9838442946190052, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.022772312164307, "learning_rate": 1e-05, "loss": 0.5333, "mean_token_accuracy": 0.8586418032646179, "num_tokens": 293573083.0, "step": 7734 }, { "epoch": 0.9839715048975958, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.753345489501953, "learning_rate": 1e-05, "loss": 0.511, "mean_token_accuracy": 0.870310366153717, "num_tokens": 293608312.0, "step": 7735 }, { "epoch": 0.9840987151761862, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.120818138122559, "learning_rate": 1e-05, "loss": 0.5213, "mean_token_accuracy": 0.8596576452255249, "num_tokens": 293645036.0, "step": 7736 }, { "epoch": 0.9842259254547767, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.335164546966553, "learning_rate": 1e-05, "loss": 0.6067, "mean_token_accuracy": 0.836769700050354, "num_tokens": 293679063.0, "step": 7737 }, { "epoch": 0.9843531357333672, "ewc_loss": 0.10693359375, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.07231330871582, "learning_rate": 1e-05, "loss": 0.5019, "mean_token_accuracy": 0.8631910085678101, "num_tokens": 293720239.0, "step": 7738 }, { "epoch": 0.9844803460119578, "ewc_loss": 0.09912109375, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.962973594665527, "learning_rate": 1e-05, "loss": 0.4562, "mean_token_accuracy": 0.8782258033752441, "num_tokens": 293757154.0, "step": 7739 }, { "epoch": 0.9846075562905483, "ewc_loss": 0.10400390625, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 6.047893524169922, "learning_rate": 1e-05, "loss": 0.5521, "mean_token_accuracy": 0.849585771560669, "num_tokens": 293797181.0, "step": 7740 }, { "epoch": 0.9847347665691388, "ewc_loss": 0.09521484375, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.700061321258545, "learning_rate": 1e-05, "loss": 0.5112, "mean_token_accuracy": 0.8594785332679749, "num_tokens": 293838130.0, "step": 7741 }, { "epoch": 0.9848619768477292, "ewc_loss": 0.10302734375, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 6.0529985427856445, "learning_rate": 1e-05, "loss": 0.5272, "mean_token_accuracy": 0.8569640517234802, "num_tokens": 293877341.0, "step": 7742 }, { "epoch": 0.9849891871263198, "ewc_loss": 0.0927734375, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.531609535217285, "learning_rate": 1e-05, "loss": 0.49, "mean_token_accuracy": 0.8659296035766602, "num_tokens": 293913764.0, "step": 7743 }, { "epoch": 0.9851163974049103, "ewc_loss": 0.10400390625, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 6.13083553314209, "learning_rate": 1e-05, "loss": 0.5301, "mean_token_accuracy": 0.8556399345397949, "num_tokens": 293954142.0, "step": 7744 }, { "epoch": 0.9852436076835008, "ewc_loss": 0.09130859375, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.395129680633545, "learning_rate": 1e-05, "loss": 0.5243, "mean_token_accuracy": 0.8547263145446777, "num_tokens": 293999016.0, "step": 7745 }, { "epoch": 0.9853708179620914, "ewc_loss": 0.10595703125, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.2897562980651855, "learning_rate": 1e-05, "loss": 0.5122, "mean_token_accuracy": 0.8619779348373413, "num_tokens": 294033545.0, "step": 7746 }, { "epoch": 0.9854980282406819, "ewc_loss": 0.08984375, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 7.534027099609375e-05, "grad_norm": 5.274509429931641, "learning_rate": 1e-05, "loss": 0.4501, "mean_token_accuracy": 0.8755488395690918, "num_tokens": 294076763.0, "step": 7747 }, { "epoch": 0.9856252385192723, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.554518222808838, "learning_rate": 1e-05, "loss": 0.4525, "mean_token_accuracy": 0.8824559450149536, "num_tokens": 294114661.0, "step": 7748 }, { "epoch": 0.9857524487978628, "ewc_loss": 0.091796875, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.260968208312988, "learning_rate": 1e-05, "loss": 0.5171, "mean_token_accuracy": 0.8573452234268188, "num_tokens": 294148819.0, "step": 7749 }, { "epoch": 0.9858796590764534, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.8916707038879395, "learning_rate": 1e-05, "loss": 0.5167, "mean_token_accuracy": 0.8660616874694824, "num_tokens": 294190820.0, "step": 7750 }, { "epoch": 0.9860068693550439, "ewc_loss": 0.09619140625, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.441267967224121, "learning_rate": 1e-05, "loss": 0.5028, "mean_token_accuracy": 0.8612110018730164, "num_tokens": 294230814.0, "step": 7751 }, { "epoch": 0.9861340796336344, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 7.0753302574157715, "learning_rate": 1e-05, "loss": 0.555, "mean_token_accuracy": 0.8592352867126465, "num_tokens": 294271811.0, "step": 7752 }, { "epoch": 0.986261289912225, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 5.818325519561768, "learning_rate": 1e-05, "loss": 0.5264, "mean_token_accuracy": 0.8594495058059692, "num_tokens": 294313901.0, "step": 7753 }, { "epoch": 0.9863885001908154, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.916168689727783, "learning_rate": 1e-05, "loss": 0.5587, "mean_token_accuracy": 0.8508049249649048, "num_tokens": 294353275.0, "step": 7754 }, { "epoch": 0.9865157104694059, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.059223175048828, "learning_rate": 1e-05, "loss": 0.4666, "mean_token_accuracy": 0.8783448934555054, "num_tokens": 294390569.0, "step": 7755 }, { "epoch": 0.9866429207479964, "ewc_loss": 0.11181640625, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.5209736824035645, "learning_rate": 1e-05, "loss": 0.5547, "mean_token_accuracy": 0.8493213057518005, "num_tokens": 294433050.0, "step": 7756 }, { "epoch": 0.986770131026587, "ewc_loss": 0.10791015625, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.082388877868652, "learning_rate": 1e-05, "loss": 0.4973, "mean_token_accuracy": 0.8694840669631958, "num_tokens": 294471198.0, "step": 7757 }, { "epoch": 0.9868973413051775, "ewc_loss": 0.10107421875, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 6.106375694274902, "learning_rate": 1e-05, "loss": 0.5012, "mean_token_accuracy": 0.8654685020446777, "num_tokens": 294511855.0, "step": 7758 }, { "epoch": 0.987024551583768, "ewc_loss": 0.10302734375, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 6.060689926147461, "learning_rate": 1e-05, "loss": 0.4962, "mean_token_accuracy": 0.8668740391731262, "num_tokens": 294547056.0, "step": 7759 }, { "epoch": 0.9871517618623584, "ewc_loss": 0.0966796875, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.796476364135742, "learning_rate": 1e-05, "loss": 0.5341, "mean_token_accuracy": 0.8560692667961121, "num_tokens": 294584516.0, "step": 7760 }, { "epoch": 0.987278972140949, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 6.046649932861328, "learning_rate": 1e-05, "loss": 0.5152, "mean_token_accuracy": 0.8589916229248047, "num_tokens": 294622253.0, "step": 7761 }, { "epoch": 0.9874061824195395, "ewc_loss": 0.09423828125, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.588009357452393, "learning_rate": 1e-05, "loss": 0.4771, "mean_token_accuracy": 0.8690731525421143, "num_tokens": 294664842.0, "step": 7762 }, { "epoch": 0.98753339269813, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 6.087274074554443, "learning_rate": 1e-05, "loss": 0.5332, "mean_token_accuracy": 0.8547192811965942, "num_tokens": 294707135.0, "step": 7763 }, { "epoch": 0.9876606029767205, "ewc_loss": 0.09130859375, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.450963973999023, "learning_rate": 1e-05, "loss": 0.5383, "mean_token_accuracy": 0.8476216197013855, "num_tokens": 294751356.0, "step": 7764 }, { "epoch": 0.9877878132553111, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.208247661590576, "learning_rate": 1e-05, "loss": 0.5264, "mean_token_accuracy": 0.8540558218955994, "num_tokens": 294789444.0, "step": 7765 }, { "epoch": 0.9879150235339016, "ewc_loss": 0.09033203125, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 7.581710815429688e-05, "grad_norm": 5.345642566680908, "learning_rate": 1e-05, "loss": 0.5273, "mean_token_accuracy": 0.8515520691871643, "num_tokens": 294823247.0, "step": 7766 }, { "epoch": 0.988042233812492, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.404198169708252, "learning_rate": 1e-05, "loss": 0.56, "mean_token_accuracy": 0.8495835661888123, "num_tokens": 294862932.0, "step": 7767 }, { "epoch": 0.9881694440910825, "ewc_loss": 0.08984375, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 7.534027099609375e-05, "grad_norm": 5.255331516265869, "learning_rate": 1e-05, "loss": 0.4991, "mean_token_accuracy": 0.8604525923728943, "num_tokens": 294902686.0, "step": 7768 }, { "epoch": 0.9882966543696731, "ewc_loss": 0.11474609375, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.707024574279785, "learning_rate": 1e-05, "loss": 0.5144, "mean_token_accuracy": 0.8684518337249756, "num_tokens": 294942028.0, "step": 7769 }, { "epoch": 0.9884238646482636, "ewc_loss": 0.09375, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.327300548553467, "learning_rate": 1e-05, "loss": 0.5136, "mean_token_accuracy": 0.8621581792831421, "num_tokens": 294977699.0, "step": 7770 }, { "epoch": 0.9885510749268541, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 6.990853786468506, "learning_rate": 1e-05, "loss": 0.5521, "mean_token_accuracy": 0.8590566515922546, "num_tokens": 295014441.0, "step": 7771 }, { "epoch": 0.9886782852054447, "ewc_loss": 0.10107421875, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.624053001403809, "learning_rate": 1e-05, "loss": 0.4881, "mean_token_accuracy": 0.8689255118370056, "num_tokens": 295053569.0, "step": 7772 }, { "epoch": 0.9888054954840351, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 7.0112690925598145, "learning_rate": 1e-05, "loss": 0.5391, "mean_token_accuracy": 0.8585816621780396, "num_tokens": 295083607.0, "step": 7773 }, { "epoch": 0.9889327057626256, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 5.951111793518066, "learning_rate": 1e-05, "loss": 0.5442, "mean_token_accuracy": 0.8506778478622437, "num_tokens": 295118074.0, "step": 7774 }, { "epoch": 0.9890599160412161, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.705641269683838, "learning_rate": 1e-05, "loss": 0.5242, "mean_token_accuracy": 0.8627861142158508, "num_tokens": 295157926.0, "step": 7775 }, { "epoch": 0.9891871263198067, "ewc_loss": 0.10888671875, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.051021099090576, "learning_rate": 1e-05, "loss": 0.5324, "mean_token_accuracy": 0.8591076135635376, "num_tokens": 295201297.0, "step": 7776 }, { "epoch": 0.9893143365983972, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.308263301849365, "learning_rate": 1e-05, "loss": 0.5498, "mean_token_accuracy": 0.8551197052001953, "num_tokens": 295238178.0, "step": 7777 }, { "epoch": 0.9894415468769877, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.035373210906982, "learning_rate": 1e-05, "loss": 0.5112, "mean_token_accuracy": 0.8610343933105469, "num_tokens": 295273843.0, "step": 7778 }, { "epoch": 0.9895687571555781, "ewc_loss": 0.0986328125, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.957449436187744, "learning_rate": 1e-05, "loss": 0.4889, "mean_token_accuracy": 0.8650048971176147, "num_tokens": 295308703.0, "step": 7779 }, { "epoch": 0.9896959674341687, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 6.008368968963623, "learning_rate": 1e-05, "loss": 0.5158, "mean_token_accuracy": 0.8597414493560791, "num_tokens": 295347710.0, "step": 7780 }, { "epoch": 0.9898231777127592, "ewc_loss": 0.09521484375, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.7090277671813965, "learning_rate": 1e-05, "loss": 0.4811, "mean_token_accuracy": 0.8702511787414551, "num_tokens": 295382255.0, "step": 7781 }, { "epoch": 0.9899503879913497, "ewc_loss": 0.10205078125, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 6.0120625495910645, "learning_rate": 1e-05, "loss": 0.4883, "mean_token_accuracy": 0.8648786544799805, "num_tokens": 295418185.0, "step": 7782 }, { "epoch": 0.9900775982699402, "ewc_loss": 0.09326171875, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.558379173278809, "learning_rate": 1e-05, "loss": 0.5681, "mean_token_accuracy": 0.8426254391670227, "num_tokens": 295455124.0, "step": 7783 }, { "epoch": 0.9902048085485308, "ewc_loss": 0.10205078125, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 6.04816198348999, "learning_rate": 1e-05, "loss": 0.5521, "mean_token_accuracy": 0.8495734930038452, "num_tokens": 295498058.0, "step": 7784 }, { "epoch": 0.9903320188271212, "ewc_loss": 0.09130859375, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.428484916687012, "learning_rate": 1e-05, "loss": 0.4579, "mean_token_accuracy": 0.8757003545761108, "num_tokens": 295534623.0, "step": 7785 }, { "epoch": 0.9904592291057117, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.163774490356445, "learning_rate": 1e-05, "loss": 0.5459, "mean_token_accuracy": 0.8517251014709473, "num_tokens": 295580064.0, "step": 7786 }, { "epoch": 0.9905864393843022, "ewc_loss": 0.09033203125, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 7.581710815429688e-05, "grad_norm": 5.315521240234375, "learning_rate": 1e-05, "loss": 0.5479, "mean_token_accuracy": 0.8473712801933289, "num_tokens": 295620720.0, "step": 7787 }, { "epoch": 0.9907136496628928, "ewc_loss": 0.10791015625, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.387744426727295, "learning_rate": 1e-05, "loss": 0.5093, "mean_token_accuracy": 0.867286205291748, "num_tokens": 295662157.0, "step": 7788 }, { "epoch": 0.9908408599414833, "ewc_loss": 0.09033203125, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 7.581710815429688e-05, "grad_norm": 5.223626136779785, "learning_rate": 1e-05, "loss": 0.506, "mean_token_accuracy": 0.8592612743377686, "num_tokens": 295700215.0, "step": 7789 }, { "epoch": 0.9909680702200738, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.729705333709717, "learning_rate": 1e-05, "loss": 0.4575, "mean_token_accuracy": 0.8831421136856079, "num_tokens": 295738363.0, "step": 7790 }, { "epoch": 0.9910952804986642, "ewc_loss": 0.09228515625, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.275657653808594, "learning_rate": 1e-05, "loss": 0.4929, "mean_token_accuracy": 0.8731704950332642, "num_tokens": 295773911.0, "step": 7791 }, { "epoch": 0.9912224907772548, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 7.098406791687012, "learning_rate": 1e-05, "loss": 0.5297, "mean_token_accuracy": 0.8594753742218018, "num_tokens": 295809341.0, "step": 7792 }, { "epoch": 0.9913497010558453, "ewc_loss": 0.10302734375, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 5.62139892578125, "learning_rate": 1e-05, "loss": 0.538, "mean_token_accuracy": 0.8503307104110718, "num_tokens": 295855952.0, "step": 7793 }, { "epoch": 0.9914769113344358, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 7.144150733947754, "learning_rate": 1e-05, "loss": 0.5193, "mean_token_accuracy": 0.8644285798072815, "num_tokens": 295892047.0, "step": 7794 }, { "epoch": 0.9916041216130264, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.047820091247559, "learning_rate": 1e-05, "loss": 0.5793, "mean_token_accuracy": 0.8419761657714844, "num_tokens": 295926259.0, "step": 7795 }, { "epoch": 0.9917313318916169, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.791543960571289, "learning_rate": 1e-05, "loss": 0.5425, "mean_token_accuracy": 0.855589747428894, "num_tokens": 295965584.0, "step": 7796 }, { "epoch": 0.9918585421702073, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.140763759613037, "learning_rate": 1e-05, "loss": 0.5023, "mean_token_accuracy": 0.8691002130508423, "num_tokens": 296002754.0, "step": 7797 }, { "epoch": 0.9919857524487978, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.370109558105469, "learning_rate": 1e-05, "loss": 0.5262, "mean_token_accuracy": 0.8592567443847656, "num_tokens": 296042161.0, "step": 7798 }, { "epoch": 0.9921129627273884, "ewc_loss": 0.10693359375, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.101764678955078, "learning_rate": 1e-05, "loss": 0.5512, "mean_token_accuracy": 0.85150146484375, "num_tokens": 296079153.0, "step": 7799 }, { "epoch": 0.9922401730059789, "ewc_loss": 0.099609375, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.985777378082275, "learning_rate": 1e-05, "loss": 0.5421, "mean_token_accuracy": 0.8513684272766113, "num_tokens": 296124064.0, "step": 7800 }, { "epoch": 0.9923673832845694, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.055151462554932, "learning_rate": 1e-05, "loss": 0.4958, "mean_token_accuracy": 0.8669304847717285, "num_tokens": 296166035.0, "step": 7801 }, { "epoch": 0.9924945935631599, "ewc_loss": 0.09619140625, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.731298923492432, "learning_rate": 1e-05, "loss": 0.4602, "mean_token_accuracy": 0.8768488168716431, "num_tokens": 296206566.0, "step": 7802 }, { "epoch": 0.9926218038417504, "ewc_loss": 0.10205078125, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 6.057179927825928, "learning_rate": 1e-05, "loss": 0.5016, "mean_token_accuracy": 0.8660118579864502, "num_tokens": 296243587.0, "step": 7803 }, { "epoch": 0.9927490141203409, "ewc_loss": 0.0927734375, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.563564777374268, "learning_rate": 1e-05, "loss": 0.5313, "mean_token_accuracy": 0.8580483198165894, "num_tokens": 296282704.0, "step": 7804 }, { "epoch": 0.9928762243989314, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 6.108392238616943, "learning_rate": 1e-05, "loss": 0.5478, "mean_token_accuracy": 0.8448401689529419, "num_tokens": 296317964.0, "step": 7805 }, { "epoch": 0.993003434677522, "ewc_loss": 0.0908203125, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 7.62939453125e-05, "grad_norm": 5.42094087600708, "learning_rate": 1e-05, "loss": 0.534, "mean_token_accuracy": 0.852695107460022, "num_tokens": 296352861.0, "step": 7806 }, { "epoch": 0.9931306449561125, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.251509189605713, "learning_rate": 1e-05, "loss": 0.5116, "mean_token_accuracy": 0.8606575727462769, "num_tokens": 296391302.0, "step": 7807 }, { "epoch": 0.993257855234703, "ewc_loss": 0.09033203125, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 7.581710815429688e-05, "grad_norm": 5.29344367980957, "learning_rate": 1e-05, "loss": 0.5163, "mean_token_accuracy": 0.8550390005111694, "num_tokens": 296428026.0, "step": 7808 }, { "epoch": 0.9933850655132934, "ewc_loss": 0.11083984375, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.526221752166748, "learning_rate": 1e-05, "loss": 0.533, "mean_token_accuracy": 0.860193133354187, "num_tokens": 296468866.0, "step": 7809 }, { "epoch": 0.993512275791884, "ewc_loss": 0.09130859375, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.220162391662598, "learning_rate": 1e-05, "loss": 0.5078, "mean_token_accuracy": 0.8613154888153076, "num_tokens": 296514806.0, "step": 7810 }, { "epoch": 0.9936394860704745, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.912611961364746, "learning_rate": 1e-05, "loss": 0.5108, "mean_token_accuracy": 0.8659554719924927, "num_tokens": 296552328.0, "step": 7811 }, { "epoch": 0.993766696349065, "ewc_loss": 0.09619140625, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.419612407684326, "learning_rate": 1e-05, "loss": 0.5296, "mean_token_accuracy": 0.8545345067977905, "num_tokens": 296589908.0, "step": 7812 }, { "epoch": 0.9938939066276555, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 0.00011205673217773438, "grad_norm": 7.151515007019043, "learning_rate": 1e-05, "loss": 0.5713, "mean_token_accuracy": 0.849612295627594, "num_tokens": 296631147.0, "step": 7813 }, { "epoch": 0.9940211169062461, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 5.851299285888672, "learning_rate": 1e-05, "loss": 0.5469, "mean_token_accuracy": 0.8524545431137085, "num_tokens": 296673633.0, "step": 7814 }, { "epoch": 0.9941483271848366, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 6.9788103103637695, "learning_rate": 1e-05, "loss": 0.5312, "mean_token_accuracy": 0.8593065738677979, "num_tokens": 296714782.0, "step": 7815 }, { "epoch": 0.994275537463427, "ewc_loss": 0.11181640625, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.129615306854248, "learning_rate": 1e-05, "loss": 0.5433, "mean_token_accuracy": 0.8559480905532837, "num_tokens": 296751556.0, "step": 7816 }, { "epoch": 0.9944027477420175, "ewc_loss": 0.11181640625, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.557158946990967, "learning_rate": 1e-05, "loss": 0.5861, "mean_token_accuracy": 0.8395115733146667, "num_tokens": 296788319.0, "step": 7817 }, { "epoch": 0.9945299580206081, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.116014003753662, "learning_rate": 1e-05, "loss": 0.5481, "mean_token_accuracy": 0.8536233901977539, "num_tokens": 296823513.0, "step": 7818 }, { "epoch": 0.9946571682991986, "ewc_loss": 0.10107421875, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 6.134217262268066, "learning_rate": 1e-05, "loss": 0.5344, "mean_token_accuracy": 0.8533657193183899, "num_tokens": 296862059.0, "step": 7819 }, { "epoch": 0.9947843785777891, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.078701972961426, "learning_rate": 1e-05, "loss": 0.5057, "mean_token_accuracy": 0.8639340400695801, "num_tokens": 296898051.0, "step": 7820 }, { "epoch": 0.9949115888563796, "ewc_loss": 0.0966796875, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.8132829666137695, "learning_rate": 1e-05, "loss": 0.4933, "mean_token_accuracy": 0.8651604652404785, "num_tokens": 296935266.0, "step": 7821 }, { "epoch": 0.9950387991349701, "ewc_loss": 0.10302734375, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 6.061238765716553, "learning_rate": 1e-05, "loss": 0.5652, "mean_token_accuracy": 0.8461275100708008, "num_tokens": 296971653.0, "step": 7822 }, { "epoch": 0.9951660094135606, "ewc_loss": 0.09423828125, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.610466480255127, "learning_rate": 1e-05, "loss": 0.5914, "mean_token_accuracy": 0.83539879322052, "num_tokens": 297015825.0, "step": 7823 }, { "epoch": 0.9952932196921511, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 6.08394718170166, "learning_rate": 1e-05, "loss": 0.5585, "mean_token_accuracy": 0.8471654057502747, "num_tokens": 297059115.0, "step": 7824 }, { "epoch": 0.9954204299707416, "ewc_loss": 0.0927734375, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.474051475524902, "learning_rate": 1e-05, "loss": 0.5155, "mean_token_accuracy": 0.8542465567588806, "num_tokens": 297092634.0, "step": 7825 }, { "epoch": 0.9955476402493322, "ewc_loss": 0.10400390625, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 6.188363075256348, "learning_rate": 1e-05, "loss": 0.5213, "mean_token_accuracy": 0.860787570476532, "num_tokens": 297137154.0, "step": 7826 }, { "epoch": 0.9956748505279227, "ewc_loss": 0.09130859375, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.34927225112915, "learning_rate": 1e-05, "loss": 0.5452, "mean_token_accuracy": 0.8484537601470947, "num_tokens": 297175029.0, "step": 7827 }, { "epoch": 0.9958020608065131, "ewc_loss": 0.10888671875, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.403501033782959, "learning_rate": 1e-05, "loss": 0.513, "mean_token_accuracy": 0.862625777721405, "num_tokens": 297207397.0, "step": 7828 }, { "epoch": 0.9959292710851037, "ewc_loss": 0.09033203125, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 7.581710815429688e-05, "grad_norm": 5.262118339538574, "learning_rate": 1e-05, "loss": 0.5273, "mean_token_accuracy": 0.8559770584106445, "num_tokens": 297250114.0, "step": 7829 }, { "epoch": 0.9960564813636942, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.70092248916626, "learning_rate": 1e-05, "loss": 0.5273, "mean_token_accuracy": 0.8644614219665527, "num_tokens": 297289765.0, "step": 7830 }, { "epoch": 0.9961836916422847, "ewc_loss": 0.09228515625, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.309258937835693, "learning_rate": 1e-05, "loss": 0.4383, "mean_token_accuracy": 0.8807721734046936, "num_tokens": 297327359.0, "step": 7831 }, { "epoch": 0.9963109019208752, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 7.014695644378662, "learning_rate": 1e-05, "loss": 0.5265, "mean_token_accuracy": 0.8655893802642822, "num_tokens": 297360851.0, "step": 7832 }, { "epoch": 0.9964381121994658, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.614380836486816, "learning_rate": 1e-05, "loss": 0.5176, "mean_token_accuracy": 0.8581985235214233, "num_tokens": 297396957.0, "step": 7833 }, { "epoch": 0.9965653224780562, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 7.065739631652832, "learning_rate": 1e-05, "loss": 0.5363, "mean_token_accuracy": 0.8607316017150879, "num_tokens": 297427900.0, "step": 7834 }, { "epoch": 0.9966925327566467, "ewc_loss": 0.10888671875, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 5.962798595428467, "learning_rate": 1e-05, "loss": 0.5591, "mean_token_accuracy": 0.8487258553504944, "num_tokens": 297464910.0, "step": 7835 }, { "epoch": 0.9968197430352372, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.7752509117126465, "learning_rate": 1e-05, "loss": 0.482, "mean_token_accuracy": 0.8723487854003906, "num_tokens": 297497585.0, "step": 7836 }, { "epoch": 0.9969469533138278, "ewc_loss": 0.109375, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.083179950714111, "learning_rate": 1e-05, "loss": 0.5483, "mean_token_accuracy": 0.8506778478622437, "num_tokens": 297537074.0, "step": 7837 }, { "epoch": 0.9970741635924183, "ewc_loss": 0.10791015625, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.3788909912109375, "learning_rate": 1e-05, "loss": 0.4892, "mean_token_accuracy": 0.8698042631149292, "num_tokens": 297570848.0, "step": 7838 }, { "epoch": 0.9972013738710088, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.055385112762451, "learning_rate": 1e-05, "loss": 0.5462, "mean_token_accuracy": 0.8498554229736328, "num_tokens": 297607520.0, "step": 7839 }, { "epoch": 0.9973285841495992, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 6.007457256317139, "learning_rate": 1e-05, "loss": 0.5092, "mean_token_accuracy": 0.8644312620162964, "num_tokens": 297648442.0, "step": 7840 }, { "epoch": 0.9974557944281898, "ewc_loss": 0.10205078125, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 6.018402099609375, "learning_rate": 1e-05, "loss": 0.573, "mean_token_accuracy": 0.8404049277305603, "num_tokens": 297686163.0, "step": 7841 }, { "epoch": 0.9975830047067803, "ewc_loss": 0.09619140625, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.847531795501709, "learning_rate": 1e-05, "loss": 0.4682, "mean_token_accuracy": 0.8706423044204712, "num_tokens": 297724191.0, "step": 7842 }, { "epoch": 0.9977102149853708, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.943958282470703, "learning_rate": 1e-05, "loss": 0.5402, "mean_token_accuracy": 0.8531678915023804, "num_tokens": 297766205.0, "step": 7843 }, { "epoch": 0.9978374252639614, "ewc_loss": 0.09423828125, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.652415752410889, "learning_rate": 1e-05, "loss": 0.4856, "mean_token_accuracy": 0.8691769242286682, "num_tokens": 297802153.0, "step": 7844 }, { "epoch": 0.9979646355425519, "ewc_loss": 0.10009765625, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.92558479309082, "learning_rate": 1e-05, "loss": 0.4973, "mean_token_accuracy": 0.8679600954055786, "num_tokens": 297841252.0, "step": 7845 }, { "epoch": 0.9980918458211423, "ewc_loss": 0.09326171875, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.54658317565918, "learning_rate": 1e-05, "loss": 0.411, "mean_token_accuracy": 0.8889422416687012, "num_tokens": 297875460.0, "step": 7846 }, { "epoch": 0.9982190560997328, "ewc_loss": 0.10107421875, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.981158256530762, "learning_rate": 1e-05, "loss": 0.5224, "mean_token_accuracy": 0.8581427335739136, "num_tokens": 297908534.0, "step": 7847 }, { "epoch": 0.9983462663783234, "ewc_loss": 0.09228515625, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.443719863891602, "learning_rate": 1e-05, "loss": 0.5653, "mean_token_accuracy": 0.8374704718589783, "num_tokens": 297950798.0, "step": 7848 }, { "epoch": 0.9984734766569139, "ewc_loss": 0.10302734375, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 6.084741592407227, "learning_rate": 1e-05, "loss": 0.4928, "mean_token_accuracy": 0.8665977716445923, "num_tokens": 297989696.0, "step": 7849 }, { "epoch": 0.9986006869355044, "ewc_loss": 0.08935546875, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 7.486343383789062e-05, "grad_norm": 5.316098690032959, "learning_rate": 1e-05, "loss": 0.511, "mean_token_accuracy": 0.8565139770507812, "num_tokens": 298033702.0, "step": 7850 }, { "epoch": 0.9987278972140949, "ewc_loss": 0.10693359375, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.313169956207275, "learning_rate": 1e-05, "loss": 0.477, "mean_token_accuracy": 0.875472903251648, "num_tokens": 298071867.0, "step": 7851 }, { "epoch": 0.9988551074926854, "ewc_loss": 0.0888671875, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 7.43865966796875e-05, "grad_norm": 5.191390514373779, "learning_rate": 1e-05, "loss": 0.541, "mean_token_accuracy": 0.8491641879081726, "num_tokens": 298108726.0, "step": 7852 }, { "epoch": 0.9989823177712759, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.711350917816162, "learning_rate": 1e-05, "loss": 0.5581, "mean_token_accuracy": 0.8541710376739502, "num_tokens": 298145174.0, "step": 7853 }, { "epoch": 0.9991095280498664, "ewc_loss": 0.091796875, "ewc_loss_diag": 0.00014781951904296875, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.1731648445129395, "learning_rate": 1e-05, "loss": 0.4783, "mean_token_accuracy": 0.8712978363037109, "num_tokens": 298181300.0, "step": 7854 }, { "epoch": 0.9992367383284569, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.00014781951904296875, "ewc_loss_parallel": 0.00011348724365234375, "grad_norm": 7.195256233215332, "learning_rate": 1e-05, "loss": 0.5425, "mean_token_accuracy": 0.8600469827651978, "num_tokens": 298220592.0, "step": 7855 }, { "epoch": 0.9993639486070475, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.00014781951904296875, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 5.563032627105713, "learning_rate": 1e-05, "loss": 0.5442, "mean_token_accuracy": 0.8556958436965942, "num_tokens": 298260796.0, "step": 7856 }, { "epoch": 0.999491158885638, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.000148773193359375, "ewc_loss_parallel": 0.00011777877807617188, "grad_norm": 7.370594024658203, "learning_rate": 1e-05, "loss": 0.5901, "mean_token_accuracy": 0.8423680067062378, "num_tokens": 298300640.0, "step": 7857 }, { "epoch": 0.9996183691642284, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.000148773193359375, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.140270709991455, "learning_rate": 1e-05, "loss": 0.4847, "mean_token_accuracy": 0.8756589889526367, "num_tokens": 298336433.0, "step": 7858 }, { "epoch": 0.9997455794428189, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.000148773193359375, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 7.005757808685303, "learning_rate": 1e-05, "loss": 0.5002, "mean_token_accuracy": 0.8735153675079346, "num_tokens": 298370355.0, "step": 7859 }, { "epoch": 0.9998727897214095, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.000148773193359375, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.287898540496826, "learning_rate": 1e-05, "loss": 0.5307, "mean_token_accuracy": 0.8600241541862488, "num_tokens": 298406262.0, "step": 7860 }, { "epoch": 1.0, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.000148773193359375, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.566700458526611, "learning_rate": 1e-05, "loss": 0.6031, "mean_token_accuracy": 0.8417243957519531, "num_tokens": 298443310.0, "step": 7861 }, { "epoch": 1.0001272102785905, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.000148773193359375, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.163806915283203, "learning_rate": 1e-05, "loss": 0.5196, "mean_token_accuracy": 0.8600247502326965, "num_tokens": 298482309.0, "step": 7862 }, { "epoch": 1.000254420557181, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.000148773193359375, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 6.207028388977051, "learning_rate": 1e-05, "loss": 0.508, "mean_token_accuracy": 0.8621256351470947, "num_tokens": 298522555.0, "step": 7863 }, { "epoch": 1.0003816308357716, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.000148773193359375, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.038023471832275, "learning_rate": 1e-05, "loss": 0.4541, "mean_token_accuracy": 0.879447340965271, "num_tokens": 298559559.0, "step": 7864 }, { "epoch": 1.0005088411143621, "ewc_loss": 0.0986328125, "ewc_loss_diag": 0.000148773193359375, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.963556289672852, "learning_rate": 1e-05, "loss": 0.4966, "mean_token_accuracy": 0.8655425310134888, "num_tokens": 298596604.0, "step": 7865 }, { "epoch": 1.0006360513929526, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.000148773193359375, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.933669567108154, "learning_rate": 1e-05, "loss": 0.4389, "mean_token_accuracy": 0.8813595771789551, "num_tokens": 298634590.0, "step": 7866 }, { "epoch": 1.0007632616715432, "ewc_loss": 0.0986328125, "ewc_loss_diag": 0.000148773193359375, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.824092864990234, "learning_rate": 1e-05, "loss": 0.4568, "mean_token_accuracy": 0.8767603635787964, "num_tokens": 298670714.0, "step": 7867 }, { "epoch": 1.0008904719501335, "ewc_loss": 0.0986328125, "ewc_loss_diag": 0.000148773193359375, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.838135719299316, "learning_rate": 1e-05, "loss": 0.4899, "mean_token_accuracy": 0.8671201467514038, "num_tokens": 298713924.0, "step": 7868 }, { "epoch": 1.001017682228724, "ewc_loss": 0.0966796875, "ewc_loss_diag": 0.000148773193359375, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.759812355041504, "learning_rate": 1e-05, "loss": 0.4789, "mean_token_accuracy": 0.870417594909668, "num_tokens": 298749537.0, "step": 7869 }, { "epoch": 1.0011448925073145, "ewc_loss": 0.09765625, "ewc_loss_diag": 0.0001506805419921875, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.771399974822998, "learning_rate": 1e-05, "loss": 0.4998, "mean_token_accuracy": 0.8597695231437683, "num_tokens": 298789887.0, "step": 7870 }, { "epoch": 1.001272102785905, "ewc_loss": 0.095703125, "ewc_loss_diag": 0.000148773193359375, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.72000789642334, "learning_rate": 1e-05, "loss": 0.4886, "mean_token_accuracy": 0.8686927556991577, "num_tokens": 298826790.0, "step": 7871 }, { "epoch": 1.0013993130644956, "ewc_loss": 0.095703125, "ewc_loss_diag": 0.000148773193359375, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.717777252197266, "learning_rate": 1e-05, "loss": 0.5138, "mean_token_accuracy": 0.8597812652587891, "num_tokens": 298860427.0, "step": 7872 }, { "epoch": 1.0015265233430861, "ewc_loss": 0.095703125, "ewc_loss_diag": 0.000148773193359375, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.710061550140381, "learning_rate": 1e-05, "loss": 0.4943, "mean_token_accuracy": 0.8659932613372803, "num_tokens": 298892627.0, "step": 7873 }, { "epoch": 1.0016537336216766, "ewc_loss": 0.0947265625, "ewc_loss_diag": 0.000148773193359375, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.672259330749512, "learning_rate": 1e-05, "loss": 0.5254, "mean_token_accuracy": 0.8555936813354492, "num_tokens": 298930513.0, "step": 7874 }, { "epoch": 1.0017809439002672, "ewc_loss": 0.09521484375, "ewc_loss_diag": 0.0001506805419921875, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.685816764831543, "learning_rate": 1e-05, "loss": 0.4264, "mean_token_accuracy": 0.8862978219985962, "num_tokens": 298971234.0, "step": 7875 }, { "epoch": 1.0019081541788577, "ewc_loss": 0.0947265625, "ewc_loss_diag": 0.0001506805419921875, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.621396541595459, "learning_rate": 1e-05, "loss": 0.4702, "mean_token_accuracy": 0.8712690472602844, "num_tokens": 299013224.0, "step": 7876 }, { "epoch": 1.0020353644574482, "ewc_loss": 0.095703125, "ewc_loss_diag": 0.0001506805419921875, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.707230091094971, "learning_rate": 1e-05, "loss": 0.5561, "mean_token_accuracy": 0.842958927154541, "num_tokens": 299055093.0, "step": 7877 }, { "epoch": 1.0021625747360388, "ewc_loss": 0.09375, "ewc_loss_diag": 0.0001506805419921875, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.588779926300049, "learning_rate": 1e-05, "loss": 0.4343, "mean_token_accuracy": 0.8797358870506287, "num_tokens": 299085368.0, "step": 7878 }, { "epoch": 1.0022897850146293, "ewc_loss": 0.09619140625, "ewc_loss_diag": 0.0001506805419921875, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.730399131774902, "learning_rate": 1e-05, "loss": 0.532, "mean_token_accuracy": 0.8594084978103638, "num_tokens": 299128279.0, "step": 7879 }, { "epoch": 1.0024169952932196, "ewc_loss": 0.0927734375, "ewc_loss_diag": 0.0001506805419921875, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.542914867401123, "learning_rate": 1e-05, "loss": 0.4708, "mean_token_accuracy": 0.8680697679519653, "num_tokens": 299161714.0, "step": 7880 }, { "epoch": 1.0025442055718101, "ewc_loss": 0.09619140625, "ewc_loss_diag": 0.0001506805419921875, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.791879653930664, "learning_rate": 1e-05, "loss": 0.5149, "mean_token_accuracy": 0.8569760918617249, "num_tokens": 299200565.0, "step": 7881 }, { "epoch": 1.0026714158504006, "ewc_loss": 0.091796875, "ewc_loss_diag": 0.0001506805419921875, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.455591678619385, "learning_rate": 1e-05, "loss": 0.4864, "mean_token_accuracy": 0.8661861419677734, "num_tokens": 299242242.0, "step": 7882 }, { "epoch": 1.0027986261289912, "ewc_loss": 0.09912109375, "ewc_loss_diag": 0.0001506805419921875, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.92270565032959, "learning_rate": 1e-05, "loss": 0.4513, "mean_token_accuracy": 0.879217267036438, "num_tokens": 299278141.0, "step": 7883 }, { "epoch": 1.0029258364075817, "ewc_loss": 0.08984375, "ewc_loss_diag": 0.0001506805419921875, "ewc_loss_parallel": 7.486343383789062e-05, "grad_norm": 5.307629585266113, "learning_rate": 1e-05, "loss": 0.455, "mean_token_accuracy": 0.8724436163902283, "num_tokens": 299312413.0, "step": 7884 }, { "epoch": 1.0030530466861722, "ewc_loss": 0.10498046875, "ewc_loss_diag": 0.0001506805419921875, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.1934380531311035, "learning_rate": 1e-05, "loss": 0.4681, "mean_token_accuracy": 0.8758671879768372, "num_tokens": 299348158.0, "step": 7885 }, { "epoch": 1.0031802569647628, "ewc_loss": 0.0869140625, "ewc_loss_diag": 0.0001506805419921875, "ewc_loss_parallel": 7.200241088867188e-05, "grad_norm": 5.074301719665527, "learning_rate": 1e-05, "loss": 0.4567, "mean_token_accuracy": 0.8733205795288086, "num_tokens": 299386963.0, "step": 7886 }, { "epoch": 1.0033074672433533, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.0001506805419921875, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.746546745300293, "learning_rate": 1e-05, "loss": 0.4806, "mean_token_accuracy": 0.8744013905525208, "num_tokens": 299424929.0, "step": 7887 }, { "epoch": 1.0034346775219438, "ewc_loss": 0.08740234375, "ewc_loss_diag": 0.0001506805419921875, "ewc_loss_parallel": 7.2479248046875e-05, "grad_norm": 4.9531965255737305, "learning_rate": 1e-05, "loss": 0.4939, "mean_token_accuracy": 0.8614113330841064, "num_tokens": 299461588.0, "step": 7888 }, { "epoch": 1.0035618878005343, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.0001506805419921875, "ewc_loss_parallel": 0.00012493133544921875, "grad_norm": 7.6072306632995605, "learning_rate": 1e-05, "loss": 0.4996, "mean_token_accuracy": 0.8752990961074829, "num_tokens": 299498226.0, "step": 7889 }, { "epoch": 1.0036890980791249, "ewc_loss": 0.10498046875, "ewc_loss_diag": 0.0001506805419921875, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 5.58920431137085, "learning_rate": 1e-05, "loss": 0.4735, "mean_token_accuracy": 0.8700327277183533, "num_tokens": 299536806.0, "step": 7890 }, { "epoch": 1.0038163083577154, "ewc_loss": 0.1494140625, "ewc_loss_diag": 0.0001506805419921875, "ewc_loss_parallel": 0.00013446807861328125, "grad_norm": 8.030301094055176, "learning_rate": 1e-05, "loss": 0.5404, "mean_token_accuracy": 0.8627753853797913, "num_tokens": 299583658.0, "step": 7891 }, { "epoch": 1.0039435186363057, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.0001506805419921875, "ewc_loss_parallel": 0.00012111663818359375, "grad_norm": 6.75405740737915, "learning_rate": 1e-05, "loss": 0.5361, "mean_token_accuracy": 0.8599438667297363, "num_tokens": 299621381.0, "step": 7892 }, { "epoch": 1.0040707289148962, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.0001506805419921875, "ewc_loss_parallel": 0.00011730194091796875, "grad_norm": 7.3493475914001465, "learning_rate": 1e-05, "loss": 0.539, "mean_token_accuracy": 0.8616921901702881, "num_tokens": 299661055.0, "step": 7893 }, { "epoch": 1.0041979391934868, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0001506805419921875, "ewc_loss_parallel": 0.00011157989501953125, "grad_norm": 6.772768974304199, "learning_rate": 1e-05, "loss": 0.4954, "mean_token_accuracy": 0.8706574440002441, "num_tokens": 299702500.0, "step": 7894 }, { "epoch": 1.0043251494720773, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.0001506805419921875, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.833044528961182, "learning_rate": 1e-05, "loss": 0.5067, "mean_token_accuracy": 0.8663328886032104, "num_tokens": 299742590.0, "step": 7895 }, { "epoch": 1.0044523597506678, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.0001506805419921875, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.38570499420166, "learning_rate": 1e-05, "loss": 0.5153, "mean_token_accuracy": 0.8641738891601562, "num_tokens": 299777716.0, "step": 7896 }, { "epoch": 1.0045795700292584, "ewc_loss": 0.11083984375, "ewc_loss_diag": 0.0001506805419921875, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.58165979385376, "learning_rate": 1e-05, "loss": 0.4768, "mean_token_accuracy": 0.8790605664253235, "num_tokens": 299808829.0, "step": 7897 }, { "epoch": 1.0047067803078489, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.0001506805419921875, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.078874111175537, "learning_rate": 1e-05, "loss": 0.4935, "mean_token_accuracy": 0.8675661087036133, "num_tokens": 299845582.0, "step": 7898 }, { "epoch": 1.0048339905864394, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.0001506805419921875, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.452549934387207, "learning_rate": 1e-05, "loss": 0.4792, "mean_token_accuracy": 0.8706997632980347, "num_tokens": 299886373.0, "step": 7899 }, { "epoch": 1.00496120086503, "ewc_loss": 0.0986328125, "ewc_loss_diag": 0.0001506805419921875, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.833085536956787, "learning_rate": 1e-05, "loss": 0.4644, "mean_token_accuracy": 0.8701176643371582, "num_tokens": 299917464.0, "step": 7900 }, { "epoch": 1.0050884111436205, "ewc_loss": 0.10888671875, "ewc_loss_diag": 0.0001506805419921875, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.457022190093994, "learning_rate": 1e-05, "loss": 0.5167, "mean_token_accuracy": 0.8673301935195923, "num_tokens": 299953455.0, "step": 7901 }, { "epoch": 1.005215621422211, "ewc_loss": 0.095703125, "ewc_loss_diag": 0.0001506805419921875, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.626171588897705, "learning_rate": 1e-05, "loss": 0.5301, "mean_token_accuracy": 0.8581974506378174, "num_tokens": 299993707.0, "step": 7902 }, { "epoch": 1.0053428317008015, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.0001506805419921875, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.625763416290283, "learning_rate": 1e-05, "loss": 0.5172, "mean_token_accuracy": 0.8635275363922119, "num_tokens": 300033021.0, "step": 7903 }, { "epoch": 1.0054700419793918, "ewc_loss": 0.0927734375, "ewc_loss_diag": 0.0001506805419921875, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.510873794555664, "learning_rate": 1e-05, "loss": 0.4799, "mean_token_accuracy": 0.8658797740936279, "num_tokens": 300072497.0, "step": 7904 }, { "epoch": 1.0055972522579824, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.0001506805419921875, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.930862903594971, "learning_rate": 1e-05, "loss": 0.4813, "mean_token_accuracy": 0.8750354647636414, "num_tokens": 300113510.0, "step": 7905 }, { "epoch": 1.0057244625365729, "ewc_loss": 0.095703125, "ewc_loss_diag": 0.0001506805419921875, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.6429123878479, "learning_rate": 1e-05, "loss": 0.4992, "mean_token_accuracy": 0.8636118173599243, "num_tokens": 300151524.0, "step": 7906 }, { "epoch": 1.0058516728151634, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.0001506805419921875, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 7.130985260009766, "learning_rate": 1e-05, "loss": 0.5095, "mean_token_accuracy": 0.8664348125457764, "num_tokens": 300184281.0, "step": 7907 }, { "epoch": 1.005978883093754, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.0001506805419921875, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 6.019380569458008, "learning_rate": 1e-05, "loss": 0.5426, "mean_token_accuracy": 0.8558211326599121, "num_tokens": 300222772.0, "step": 7908 }, { "epoch": 1.0061060933723445, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.0001506805419921875, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.895493030548096, "learning_rate": 1e-05, "loss": 0.5239, "mean_token_accuracy": 0.8601106405258179, "num_tokens": 300257347.0, "step": 7909 }, { "epoch": 1.006233303650935, "ewc_loss": 0.10791015625, "ewc_loss_diag": 0.0001506805419921875, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.250648021697998, "learning_rate": 1e-05, "loss": 0.5157, "mean_token_accuracy": 0.8609231114387512, "num_tokens": 300300678.0, "step": 7910 }, { "epoch": 1.0063605139295255, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.0001506805419921875, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.368436336517334, "learning_rate": 1e-05, "loss": 0.4718, "mean_token_accuracy": 0.8729970455169678, "num_tokens": 300339851.0, "step": 7911 }, { "epoch": 1.006487724208116, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.0001506805419921875, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.260519504547119, "learning_rate": 1e-05, "loss": 0.4656, "mean_token_accuracy": 0.8768559694290161, "num_tokens": 300379753.0, "step": 7912 }, { "epoch": 1.0066149344867066, "ewc_loss": 0.09375, "ewc_loss_diag": 0.0001506805419921875, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.859448432922363, "learning_rate": 1e-05, "loss": 0.4761, "mean_token_accuracy": 0.8687189817428589, "num_tokens": 300418235.0, "step": 7913 }, { "epoch": 1.006742144765297, "ewc_loss": 0.10693359375, "ewc_loss_diag": 0.0001506805419921875, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.298398017883301, "learning_rate": 1e-05, "loss": 0.448, "mean_token_accuracy": 0.8802419900894165, "num_tokens": 300450237.0, "step": 7914 }, { "epoch": 1.0068693550438876, "ewc_loss": 0.08935546875, "ewc_loss_diag": 0.0001506805419921875, "ewc_loss_parallel": 7.43865966796875e-05, "grad_norm": 5.471660137176514, "learning_rate": 1e-05, "loss": 0.47, "mean_token_accuracy": 0.8669121861457825, "num_tokens": 300489416.0, "step": 7915 }, { "epoch": 1.0069965653224782, "ewc_loss": 0.10791015625, "ewc_loss_diag": 0.0001506805419921875, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.5073676109313965, "learning_rate": 1e-05, "loss": 0.478, "mean_token_accuracy": 0.8745360374450684, "num_tokens": 300526180.0, "step": 7916 }, { "epoch": 1.0071237756010685, "ewc_loss": 0.08740234375, "ewc_loss_diag": 0.0001506805419921875, "ewc_loss_parallel": 7.2479248046875e-05, "grad_norm": 5.265596389770508, "learning_rate": 1e-05, "loss": 0.4765, "mean_token_accuracy": 0.8679359555244446, "num_tokens": 300566327.0, "step": 7917 }, { "epoch": 1.007250985879659, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.0001506805419921875, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.902353286743164, "learning_rate": 1e-05, "loss": 0.5391, "mean_token_accuracy": 0.8595308065414429, "num_tokens": 300610136.0, "step": 7918 }, { "epoch": 1.0073781961582495, "ewc_loss": 0.09228515625, "ewc_loss_diag": 0.0001506805419921875, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.363183975219727, "learning_rate": 1e-05, "loss": 0.4877, "mean_token_accuracy": 0.8668080568313599, "num_tokens": 300647238.0, "step": 7919 }, { "epoch": 1.00750540643684, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0001506805419921875, "ewc_loss_parallel": 0.00011205673217773438, "grad_norm": 7.276888847351074, "learning_rate": 1e-05, "loss": 0.509, "mean_token_accuracy": 0.8698764443397522, "num_tokens": 300684291.0, "step": 7920 }, { "epoch": 1.0076326167154306, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.0001506805419921875, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 5.887408256530762, "learning_rate": 1e-05, "loss": 0.5055, "mean_token_accuracy": 0.8627977967262268, "num_tokens": 300718919.0, "step": 7921 }, { "epoch": 1.0077598269940211, "ewc_loss": 0.125, "ewc_loss_diag": 0.0001506805419921875, "ewc_loss_parallel": 0.00011014938354492188, "grad_norm": 7.148988246917725, "learning_rate": 1e-05, "loss": 0.5223, "mean_token_accuracy": 0.8649219870567322, "num_tokens": 300752353.0, "step": 7922 }, { "epoch": 1.0078870372726116, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.0001506805419921875, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.24869441986084, "learning_rate": 1e-05, "loss": 0.4751, "mean_token_accuracy": 0.871357798576355, "num_tokens": 300789214.0, "step": 7923 }, { "epoch": 1.0080142475512022, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.0001506805419921875, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.708570957183838, "learning_rate": 1e-05, "loss": 0.5264, "mean_token_accuracy": 0.8605059385299683, "num_tokens": 300825629.0, "step": 7924 }, { "epoch": 1.0081414578297927, "ewc_loss": 0.10986328125, "ewc_loss_diag": 0.0001506805419921875, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.223813056945801, "learning_rate": 1e-05, "loss": 0.4555, "mean_token_accuracy": 0.8819789290428162, "num_tokens": 300863908.0, "step": 7925 }, { "epoch": 1.0082686681083832, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.00015163421630859375, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.31865119934082, "learning_rate": 1e-05, "loss": 0.5165, "mean_token_accuracy": 0.8619345426559448, "num_tokens": 300905458.0, "step": 7926 }, { "epoch": 1.0083958783869738, "ewc_loss": 0.10400390625, "ewc_loss_diag": 0.0001506805419921875, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.083497047424316, "learning_rate": 1e-05, "loss": 0.4519, "mean_token_accuracy": 0.879941463470459, "num_tokens": 300943906.0, "step": 7927 }, { "epoch": 1.0085230886655643, "ewc_loss": 0.10009765625, "ewc_loss_diag": 0.0001506805419921875, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 6.079702854156494, "learning_rate": 1e-05, "loss": 0.4826, "mean_token_accuracy": 0.87034010887146, "num_tokens": 300979371.0, "step": 7928 }, { "epoch": 1.0086502989441546, "ewc_loss": 0.0986328125, "ewc_loss_diag": 0.0001506805419921875, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.916712284088135, "learning_rate": 1e-05, "loss": 0.4861, "mean_token_accuracy": 0.8661628365516663, "num_tokens": 301017575.0, "step": 7929 }, { "epoch": 1.0087775092227451, "ewc_loss": 0.09912109375, "ewc_loss_diag": 0.0001506805419921875, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.992597579956055, "learning_rate": 1e-05, "loss": 0.4759, "mean_token_accuracy": 0.869663417339325, "num_tokens": 301053130.0, "step": 7930 }, { "epoch": 1.0089047195013356, "ewc_loss": 0.0947265625, "ewc_loss_diag": 0.0001506805419921875, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.738139629364014, "learning_rate": 1e-05, "loss": 0.4324, "mean_token_accuracy": 0.8813008069992065, "num_tokens": 301090740.0, "step": 7931 }, { "epoch": 1.0090319297799262, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.0001506805419921875, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 6.018926620483398, "learning_rate": 1e-05, "loss": 0.4852, "mean_token_accuracy": 0.8676358461380005, "num_tokens": 301133539.0, "step": 7932 }, { "epoch": 1.0091591400585167, "ewc_loss": 0.09130859375, "ewc_loss_diag": 0.0001506805419921875, "ewc_loss_parallel": 7.62939453125e-05, "grad_norm": 5.56365966796875, "learning_rate": 1e-05, "loss": 0.5504, "mean_token_accuracy": 0.8457993268966675, "num_tokens": 301168243.0, "step": 7933 }, { "epoch": 1.0092863503371072, "ewc_loss": 0.10302734375, "ewc_loss_diag": 0.0001506805419921875, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 6.1535162925720215, "learning_rate": 1e-05, "loss": 0.4745, "mean_token_accuracy": 0.8724942803382874, "num_tokens": 301204444.0, "step": 7934 }, { "epoch": 1.0094135606156978, "ewc_loss": 0.08935546875, "ewc_loss_diag": 0.0001506805419921875, "ewc_loss_parallel": 7.43865966796875e-05, "grad_norm": 5.364266872406006, "learning_rate": 1e-05, "loss": 0.4277, "mean_token_accuracy": 0.8846242427825928, "num_tokens": 301245768.0, "step": 7935 }, { "epoch": 1.0095407708942883, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.0001506805419921875, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.437465667724609, "learning_rate": 1e-05, "loss": 0.5313, "mean_token_accuracy": 0.8546806573867798, "num_tokens": 301282381.0, "step": 7936 }, { "epoch": 1.0096679811728788, "ewc_loss": 0.08837890625, "ewc_loss_diag": 0.0001506805419921875, "ewc_loss_parallel": 7.343292236328125e-05, "grad_norm": 5.24625301361084, "learning_rate": 1e-05, "loss": 0.4429, "mean_token_accuracy": 0.8784120082855225, "num_tokens": 301322538.0, "step": 7937 }, { "epoch": 1.0097951914514693, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.0001506805419921875, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.864674091339111, "learning_rate": 1e-05, "loss": 0.4972, "mean_token_accuracy": 0.8718000650405884, "num_tokens": 301357527.0, "step": 7938 }, { "epoch": 1.0099224017300599, "ewc_loss": 0.09326171875, "ewc_loss_diag": 0.0001506805419921875, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.3830885887146, "learning_rate": 1e-05, "loss": 0.544, "mean_token_accuracy": 0.8496342897415161, "num_tokens": 301399609.0, "step": 7939 }, { "epoch": 1.0100496120086504, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0001506805419921875, "ewc_loss_parallel": 0.00011157989501953125, "grad_norm": 7.197264671325684, "learning_rate": 1e-05, "loss": 0.5121, "mean_token_accuracy": 0.8681002259254456, "num_tokens": 301439271.0, "step": 7940 }, { "epoch": 1.0101768222872407, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.0001506805419921875, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 5.864521026611328, "learning_rate": 1e-05, "loss": 0.4974, "mean_token_accuracy": 0.8671337962150574, "num_tokens": 301476895.0, "step": 7941 }, { "epoch": 1.0103040325658312, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.0001506805419921875, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 7.030322074890137, "learning_rate": 1e-05, "loss": 0.5337, "mean_token_accuracy": 0.8600314259529114, "num_tokens": 301519649.0, "step": 7942 }, { "epoch": 1.0104312428444218, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.0001506805419921875, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.2107768058776855, "learning_rate": 1e-05, "loss": 0.5338, "mean_token_accuracy": 0.8592953681945801, "num_tokens": 301557132.0, "step": 7943 }, { "epoch": 1.0105584531230123, "ewc_loss": 0.11181640625, "ewc_loss_diag": 0.0001506805419921875, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.517822742462158, "learning_rate": 1e-05, "loss": 0.5061, "mean_token_accuracy": 0.8642318844795227, "num_tokens": 301603604.0, "step": 7944 }, { "epoch": 1.0106856634016028, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.0001506805419921875, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.239838600158691, "learning_rate": 1e-05, "loss": 0.5302, "mean_token_accuracy": 0.856348991394043, "num_tokens": 301643797.0, "step": 7945 }, { "epoch": 1.0108128736801933, "ewc_loss": 0.09912109375, "ewc_loss_diag": 0.0001506805419921875, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 6.0372185707092285, "learning_rate": 1e-05, "loss": 0.4611, "mean_token_accuracy": 0.8732983469963074, "num_tokens": 301676658.0, "step": 7946 }, { "epoch": 1.0109400839587839, "ewc_loss": 0.10791015625, "ewc_loss_diag": 0.0001506805419921875, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.237697601318359, "learning_rate": 1e-05, "loss": 0.4779, "mean_token_accuracy": 0.8763142824172974, "num_tokens": 301714749.0, "step": 7947 }, { "epoch": 1.0110672942373744, "ewc_loss": 0.09326171875, "ewc_loss_diag": 0.0001506805419921875, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.648001670837402, "learning_rate": 1e-05, "loss": 0.4678, "mean_token_accuracy": 0.8707839250564575, "num_tokens": 301754116.0, "step": 7948 }, { "epoch": 1.011194504515965, "ewc_loss": 0.10888671875, "ewc_loss_diag": 0.0001506805419921875, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.359219551086426, "learning_rate": 1e-05, "loss": 0.457, "mean_token_accuracy": 0.8796659708023071, "num_tokens": 301791441.0, "step": 7949 }, { "epoch": 1.0113217147945555, "ewc_loss": 0.09033203125, "ewc_loss_diag": 0.0001506805419921875, "ewc_loss_parallel": 7.534027099609375e-05, "grad_norm": 5.424850940704346, "learning_rate": 1e-05, "loss": 0.5139, "mean_token_accuracy": 0.8576399683952332, "num_tokens": 301830971.0, "step": 7950 }, { "epoch": 1.011448925073146, "ewc_loss": 0.11181640625, "ewc_loss_diag": 0.0001506805419921875, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.5818562507629395, "learning_rate": 1e-05, "loss": 0.5136, "mean_token_accuracy": 0.8623644709587097, "num_tokens": 301867848.0, "step": 7951 }, { "epoch": 1.0115761353517365, "ewc_loss": 0.0927734375, "ewc_loss_diag": 0.0001506805419921875, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.372629165649414, "learning_rate": 1e-05, "loss": 0.4864, "mean_token_accuracy": 0.8651812672615051, "num_tokens": 301912700.0, "step": 7952 }, { "epoch": 1.0117033456303268, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0001506805419921875, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.858153820037842, "learning_rate": 1e-05, "loss": 0.4911, "mean_token_accuracy": 0.8726546764373779, "num_tokens": 301945890.0, "step": 7953 }, { "epoch": 1.0118305559089174, "ewc_loss": 0.0966796875, "ewc_loss_diag": 0.0001506805419921875, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.554187297821045, "learning_rate": 1e-05, "loss": 0.4494, "mean_token_accuracy": 0.8785715103149414, "num_tokens": 301980095.0, "step": 7954 }, { "epoch": 1.0119577661875079, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.0001506805419921875, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.950068950653076, "learning_rate": 1e-05, "loss": 0.5091, "mean_token_accuracy": 0.8681349754333496, "num_tokens": 302011855.0, "step": 7955 }, { "epoch": 1.0120849764660984, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.0001506805419921875, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 5.821792125701904, "learning_rate": 1e-05, "loss": 0.4962, "mean_token_accuracy": 0.8653501272201538, "num_tokens": 302048299.0, "step": 7956 }, { "epoch": 1.012212186744689, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.0001506805419921875, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.789301872253418, "learning_rate": 1e-05, "loss": 0.5073, "mean_token_accuracy": 0.8662627339363098, "num_tokens": 302082628.0, "step": 7957 }, { "epoch": 1.0123393970232795, "ewc_loss": 0.10400390625, "ewc_loss_diag": 0.0001506805419921875, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 5.961688041687012, "learning_rate": 1e-05, "loss": 0.456, "mean_token_accuracy": 0.8783075213432312, "num_tokens": 302119832.0, "step": 7958 }, { "epoch": 1.01246660730187, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.0001506805419921875, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.4943695068359375, "learning_rate": 1e-05, "loss": 0.4875, "mean_token_accuracy": 0.8709625601768494, "num_tokens": 302162930.0, "step": 7959 }, { "epoch": 1.0125938175804605, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.0001506805419921875, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 5.950230121612549, "learning_rate": 1e-05, "loss": 0.4846, "mean_token_accuracy": 0.8701207637786865, "num_tokens": 302202250.0, "step": 7960 }, { "epoch": 1.012721027859051, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.0001506805419921875, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 6.241329669952393, "learning_rate": 1e-05, "loss": 0.5112, "mean_token_accuracy": 0.8627671003341675, "num_tokens": 302239446.0, "step": 7961 }, { "epoch": 1.0128482381376416, "ewc_loss": 0.10009765625, "ewc_loss_diag": 0.0001506805419921875, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.860555171966553, "learning_rate": 1e-05, "loss": 0.439, "mean_token_accuracy": 0.8836206793785095, "num_tokens": 302279949.0, "step": 7962 }, { "epoch": 1.012975448416232, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.0001506805419921875, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 6.092652320861816, "learning_rate": 1e-05, "loss": 0.5078, "mean_token_accuracy": 0.8602021932601929, "num_tokens": 302318128.0, "step": 7963 }, { "epoch": 1.0131026586948226, "ewc_loss": 0.09716796875, "ewc_loss_diag": 0.0001506805419921875, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.7526044845581055, "learning_rate": 1e-05, "loss": 0.4922, "mean_token_accuracy": 0.8678646087646484, "num_tokens": 302357353.0, "step": 7964 }, { "epoch": 1.0132298689734132, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.0001506805419921875, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 6.038054466247559, "learning_rate": 1e-05, "loss": 0.4681, "mean_token_accuracy": 0.871117353439331, "num_tokens": 302399843.0, "step": 7965 }, { "epoch": 1.0133570792520035, "ewc_loss": 0.09423828125, "ewc_loss_diag": 0.0001506805419921875, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.626648902893066, "learning_rate": 1e-05, "loss": 0.4911, "mean_token_accuracy": 0.8661240339279175, "num_tokens": 302440081.0, "step": 7966 }, { "epoch": 1.013484289530594, "ewc_loss": 0.10302734375, "ewc_loss_diag": 0.0001506805419921875, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 6.087625503540039, "learning_rate": 1e-05, "loss": 0.5228, "mean_token_accuracy": 0.858105480670929, "num_tokens": 302479061.0, "step": 7967 }, { "epoch": 1.0136114998091845, "ewc_loss": 0.091796875, "ewc_loss_diag": 0.0001506805419921875, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.4867939949035645, "learning_rate": 1e-05, "loss": 0.4482, "mean_token_accuracy": 0.8761698007583618, "num_tokens": 302514490.0, "step": 7968 }, { "epoch": 1.013738710087775, "ewc_loss": 0.10498046875, "ewc_loss_diag": 0.0001506805419921875, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.231348037719727, "learning_rate": 1e-05, "loss": 0.4769, "mean_token_accuracy": 0.8739193677902222, "num_tokens": 302552555.0, "step": 7969 }, { "epoch": 1.0138659203663656, "ewc_loss": 0.09130859375, "ewc_loss_diag": 0.0001506805419921875, "ewc_loss_parallel": 7.62939453125e-05, "grad_norm": 5.3756327629089355, "learning_rate": 1e-05, "loss": 0.4962, "mean_token_accuracy": 0.857419490814209, "num_tokens": 302589296.0, "step": 7970 }, { "epoch": 1.013993130644956, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.0001506805419921875, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.463920593261719, "learning_rate": 1e-05, "loss": 0.4408, "mean_token_accuracy": 0.884501576423645, "num_tokens": 302626085.0, "step": 7971 }, { "epoch": 1.0141203409235466, "ewc_loss": 0.0908203125, "ewc_loss_diag": 0.0001506805419921875, "ewc_loss_parallel": 7.581710815429688e-05, "grad_norm": 5.333600997924805, "learning_rate": 1e-05, "loss": 0.4587, "mean_token_accuracy": 0.8703805804252625, "num_tokens": 302662671.0, "step": 7972 }, { "epoch": 1.0142475512021372, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.0001506805419921875, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.732922077178955, "learning_rate": 1e-05, "loss": 0.4864, "mean_token_accuracy": 0.8700504302978516, "num_tokens": 302706112.0, "step": 7973 }, { "epoch": 1.0143747614807277, "ewc_loss": 0.09521484375, "ewc_loss_diag": 0.0001506805419921875, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.477069854736328, "learning_rate": 1e-05, "loss": 0.5096, "mean_token_accuracy": 0.8577218055725098, "num_tokens": 302744942.0, "step": 7974 }, { "epoch": 1.0145019717593182, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.00015163421630859375, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.8464860916137695, "learning_rate": 1e-05, "loss": 0.4568, "mean_token_accuracy": 0.8822591304779053, "num_tokens": 302784085.0, "step": 7975 }, { "epoch": 1.0146291820379088, "ewc_loss": 0.10205078125, "ewc_loss_diag": 0.00015163421630859375, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.758299350738525, "learning_rate": 1e-05, "loss": 0.4682, "mean_token_accuracy": 0.8735690116882324, "num_tokens": 302821336.0, "step": 7976 }, { "epoch": 1.0147563923164993, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.00015163421630859375, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.648561000823975, "learning_rate": 1e-05, "loss": 0.4874, "mean_token_accuracy": 0.8702927827835083, "num_tokens": 302857245.0, "step": 7977 }, { "epoch": 1.0148836025950896, "ewc_loss": 0.10595703125, "ewc_loss_diag": 0.00015163421630859375, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 5.970649719238281, "learning_rate": 1e-05, "loss": 0.4819, "mean_token_accuracy": 0.8707144260406494, "num_tokens": 302894115.0, "step": 7978 }, { "epoch": 1.0150108128736801, "ewc_loss": 0.10693359375, "ewc_loss_diag": 0.00015163421630859375, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.207434177398682, "learning_rate": 1e-05, "loss": 0.42, "mean_token_accuracy": 0.8887370824813843, "num_tokens": 302930162.0, "step": 7979 }, { "epoch": 1.0151380231522706, "ewc_loss": 0.10595703125, "ewc_loss_diag": 0.00015163421630859375, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.073938846588135, "learning_rate": 1e-05, "loss": 0.4818, "mean_token_accuracy": 0.8690778017044067, "num_tokens": 302965184.0, "step": 7980 }, { "epoch": 1.0152652334308612, "ewc_loss": 0.09521484375, "ewc_loss_diag": 0.00015163421630859375, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.7392897605896, "learning_rate": 1e-05, "loss": 0.5153, "mean_token_accuracy": 0.8621199131011963, "num_tokens": 303003895.0, "step": 7981 }, { "epoch": 1.0153924437094517, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.00015163421630859375, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.257525444030762, "learning_rate": 1e-05, "loss": 0.4775, "mean_token_accuracy": 0.8685253858566284, "num_tokens": 303038328.0, "step": 7982 }, { "epoch": 1.0155196539880422, "ewc_loss": 0.08984375, "ewc_loss_diag": 0.00015163421630859375, "ewc_loss_parallel": 7.486343383789062e-05, "grad_norm": 5.3749895095825195, "learning_rate": 1e-05, "loss": 0.5542, "mean_token_accuracy": 0.8442724943161011, "num_tokens": 303075546.0, "step": 7983 }, { "epoch": 1.0156468642666328, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.00015163421630859375, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.579944610595703, "learning_rate": 1e-05, "loss": 0.5048, "mean_token_accuracy": 0.8664315938949585, "num_tokens": 303111825.0, "step": 7984 }, { "epoch": 1.0157740745452233, "ewc_loss": 0.0888671875, "ewc_loss_diag": 0.00015163421630859375, "ewc_loss_parallel": 7.390975952148438e-05, "grad_norm": 5.227361679077148, "learning_rate": 1e-05, "loss": 0.44, "mean_token_accuracy": 0.8745710849761963, "num_tokens": 303147607.0, "step": 7985 }, { "epoch": 1.0159012848238138, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.00015163421630859375, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 7.015454292297363, "learning_rate": 1e-05, "loss": 0.5201, "mean_token_accuracy": 0.8642430901527405, "num_tokens": 303185646.0, "step": 7986 }, { "epoch": 1.0160284951024043, "ewc_loss": 0.0966796875, "ewc_loss_diag": 0.00015163421630859375, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.491127967834473, "learning_rate": 1e-05, "loss": 0.4683, "mean_token_accuracy": 0.8709521293640137, "num_tokens": 303218723.0, "step": 7987 }, { "epoch": 1.0161557053809949, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.00015163421630859375, "ewc_loss_parallel": 0.00011157989501953125, "grad_norm": 7.206700801849365, "learning_rate": 1e-05, "loss": 0.5213, "mean_token_accuracy": 0.8642332553863525, "num_tokens": 303259012.0, "step": 7988 }, { "epoch": 1.0162829156595854, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.00015163421630859375, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 5.977655410766602, "learning_rate": 1e-05, "loss": 0.5056, "mean_token_accuracy": 0.8678284883499146, "num_tokens": 303294868.0, "step": 7989 }, { "epoch": 1.0164101259381757, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.00015163421630859375, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.9340667724609375, "learning_rate": 1e-05, "loss": 0.4956, "mean_token_accuracy": 0.8697444200515747, "num_tokens": 303331987.0, "step": 7990 }, { "epoch": 1.0165373362167662, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.00015163421630859375, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.165112018585205, "learning_rate": 1e-05, "loss": 0.4274, "mean_token_accuracy": 0.8886449337005615, "num_tokens": 303364230.0, "step": 7991 }, { "epoch": 1.0166645464953568, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.00015163421630859375, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.537984848022461, "learning_rate": 1e-05, "loss": 0.4672, "mean_token_accuracy": 0.8768130540847778, "num_tokens": 303404169.0, "step": 7992 }, { "epoch": 1.0167917567739473, "ewc_loss": 0.10693359375, "ewc_loss_diag": 0.00015163421630859375, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.0715179443359375, "learning_rate": 1e-05, "loss": 0.4819, "mean_token_accuracy": 0.8719956278800964, "num_tokens": 303445137.0, "step": 7993 }, { "epoch": 1.0169189670525378, "ewc_loss": 0.10498046875, "ewc_loss_diag": 0.00015163421630859375, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.230108261108398, "learning_rate": 1e-05, "loss": 0.4455, "mean_token_accuracy": 0.8832846879959106, "num_tokens": 303486515.0, "step": 7994 }, { "epoch": 1.0170461773311283, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.00015163421630859375, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.959555149078369, "learning_rate": 1e-05, "loss": 0.4953, "mean_token_accuracy": 0.8664209246635437, "num_tokens": 303521514.0, "step": 7995 }, { "epoch": 1.0171733876097189, "ewc_loss": 0.10107421875, "ewc_loss_diag": 0.00015163421630859375, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 6.067767143249512, "learning_rate": 1e-05, "loss": 0.5006, "mean_token_accuracy": 0.8658716678619385, "num_tokens": 303555606.0, "step": 7996 }, { "epoch": 1.0173005978883094, "ewc_loss": 0.09912109375, "ewc_loss_diag": 0.00015163421630859375, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.795119285583496, "learning_rate": 1e-05, "loss": 0.4776, "mean_token_accuracy": 0.8684685826301575, "num_tokens": 303592075.0, "step": 7997 }, { "epoch": 1.0174278081669, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.00015163421630859375, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 6.0208353996276855, "learning_rate": 1e-05, "loss": 0.5405, "mean_token_accuracy": 0.8498364686965942, "num_tokens": 303632529.0, "step": 7998 }, { "epoch": 1.0175550184454905, "ewc_loss": 0.09423828125, "ewc_loss_diag": 0.00015163421630859375, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.643309116363525, "learning_rate": 1e-05, "loss": 0.4636, "mean_token_accuracy": 0.8736507892608643, "num_tokens": 303665419.0, "step": 7999 }, { "epoch": 1.017682228724081, "ewc_loss": 0.10205078125, "ewc_loss_diag": 0.00015163421630859375, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 6.078227519989014, "learning_rate": 1e-05, "loss": 0.487, "mean_token_accuracy": 0.8675920963287354, "num_tokens": 303698778.0, "step": 8000 }, { "epoch": 1.0178094390026715, "ewc_loss": 0.091796875, "ewc_loss_diag": 0.00015163421630859375, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.4756364822387695, "learning_rate": 1e-05, "loss": 0.4837, "mean_token_accuracy": 0.8664920926094055, "num_tokens": 303738713.0, "step": 8001 }, { "epoch": 1.0179366492812618, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.00015163421630859375, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.258543491363525, "learning_rate": 1e-05, "loss": 0.483, "mean_token_accuracy": 0.8678804039955139, "num_tokens": 303767821.0, "step": 8002 }, { "epoch": 1.0180638595598523, "ewc_loss": 0.08984375, "ewc_loss_diag": 0.00015163421630859375, "ewc_loss_parallel": 7.486343383789062e-05, "grad_norm": 5.332732677459717, "learning_rate": 1e-05, "loss": 0.4713, "mean_token_accuracy": 0.8708454966545105, "num_tokens": 303813088.0, "step": 8003 }, { "epoch": 1.0181910698384429, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.00015163421630859375, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.556108474731445, "learning_rate": 1e-05, "loss": 0.5504, "mean_token_accuracy": 0.8524625301361084, "num_tokens": 303849926.0, "step": 8004 }, { "epoch": 1.0183182801170334, "ewc_loss": 0.091796875, "ewc_loss_diag": 0.00015163421630859375, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.320487976074219, "learning_rate": 1e-05, "loss": 0.46, "mean_token_accuracy": 0.8738899230957031, "num_tokens": 303888681.0, "step": 8005 }, { "epoch": 1.018445490395624, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.00015163421630859375, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.882444381713867, "learning_rate": 1e-05, "loss": 0.4936, "mean_token_accuracy": 0.871299147605896, "num_tokens": 303926042.0, "step": 8006 }, { "epoch": 1.0185727006742145, "ewc_loss": 0.09716796875, "ewc_loss_diag": 0.00015163421630859375, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.497605800628662, "learning_rate": 1e-05, "loss": 0.4472, "mean_token_accuracy": 0.879679799079895, "num_tokens": 303962014.0, "step": 8007 }, { "epoch": 1.018699910952805, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.00015163421630859375, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 7.036734104156494, "learning_rate": 1e-05, "loss": 0.5086, "mean_token_accuracy": 0.870872974395752, "num_tokens": 303998856.0, "step": 8008 }, { "epoch": 1.0188271212313955, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.00015163421630859375, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 5.883683681488037, "learning_rate": 1e-05, "loss": 0.4661, "mean_token_accuracy": 0.8744872808456421, "num_tokens": 304034228.0, "step": 8009 }, { "epoch": 1.018954331509986, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.00015163421630859375, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.796576023101807, "learning_rate": 1e-05, "loss": 0.5854, "mean_token_accuracy": 0.8489421606063843, "num_tokens": 304070477.0, "step": 8010 }, { "epoch": 1.0190815417885766, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.00015163421630859375, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.096733093261719, "learning_rate": 1e-05, "loss": 0.4795, "mean_token_accuracy": 0.8742101192474365, "num_tokens": 304116073.0, "step": 8011 }, { "epoch": 1.019208752067167, "ewc_loss": 0.10791015625, "ewc_loss_diag": 0.00015163421630859375, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.31510066986084, "learning_rate": 1e-05, "loss": 0.5274, "mean_token_accuracy": 0.8610489964485168, "num_tokens": 304149950.0, "step": 8012 }, { "epoch": 1.0193359623457576, "ewc_loss": 0.109375, "ewc_loss_diag": 0.00015163421630859375, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.143327236175537, "learning_rate": 1e-05, "loss": 0.4384, "mean_token_accuracy": 0.8860834240913391, "num_tokens": 304191824.0, "step": 8013 }, { "epoch": 1.0194631726243482, "ewc_loss": 0.09765625, "ewc_loss_diag": 0.00015163421630859375, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.85571813583374, "learning_rate": 1e-05, "loss": 0.4336, "mean_token_accuracy": 0.8818800449371338, "num_tokens": 304223616.0, "step": 8014 }, { "epoch": 1.0195903829029385, "ewc_loss": 0.10888671875, "ewc_loss_diag": 0.00015163421630859375, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.232186317443848, "learning_rate": 1e-05, "loss": 0.5022, "mean_token_accuracy": 0.8669032454490662, "num_tokens": 304261120.0, "step": 8015 }, { "epoch": 1.019717593181529, "ewc_loss": 0.09375, "ewc_loss_diag": 0.00015163421630859375, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.515198230743408, "learning_rate": 1e-05, "loss": 0.5328, "mean_token_accuracy": 0.8558657169342041, "num_tokens": 304294796.0, "step": 8016 }, { "epoch": 1.0198448034601195, "ewc_loss": 0.11083984375, "ewc_loss_diag": 0.00015163421630859375, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.4343132972717285, "learning_rate": 1e-05, "loss": 0.5327, "mean_token_accuracy": 0.8592352271080017, "num_tokens": 304333396.0, "step": 8017 }, { "epoch": 1.01997201373871, "ewc_loss": 0.09130859375, "ewc_loss_diag": 0.00015163421630859375, "ewc_loss_parallel": 7.62939453125e-05, "grad_norm": 5.323386192321777, "learning_rate": 1e-05, "loss": 0.4761, "mean_token_accuracy": 0.8704996705055237, "num_tokens": 304371246.0, "step": 8018 }, { "epoch": 1.0200992240173006, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.000152587890625, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.741843223571777, "learning_rate": 1e-05, "loss": 0.4601, "mean_token_accuracy": 0.882319450378418, "num_tokens": 304408179.0, "step": 8019 }, { "epoch": 1.020226434295891, "ewc_loss": 0.09326171875, "ewc_loss_diag": 0.000152587890625, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.383608341217041, "learning_rate": 1e-05, "loss": 0.4575, "mean_token_accuracy": 0.8720897436141968, "num_tokens": 304442758.0, "step": 8020 }, { "epoch": 1.0203536445744816, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.000152587890625, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 7.011562824249268, "learning_rate": 1e-05, "loss": 0.4721, "mean_token_accuracy": 0.880391001701355, "num_tokens": 304479568.0, "step": 8021 }, { "epoch": 1.0204808548530722, "ewc_loss": 0.10009765625, "ewc_loss_diag": 0.000152587890625, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.67819881439209, "learning_rate": 1e-05, "loss": 0.4872, "mean_token_accuracy": 0.868699312210083, "num_tokens": 304521255.0, "step": 8022 }, { "epoch": 1.0206080651316627, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.000152587890625, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 7.010489463806152, "learning_rate": 1e-05, "loss": 0.4862, "mean_token_accuracy": 0.870502233505249, "num_tokens": 304553569.0, "step": 8023 }, { "epoch": 1.0207352754102532, "ewc_loss": 0.109375, "ewc_loss_diag": 0.000152587890625, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 5.9857964515686035, "learning_rate": 1e-05, "loss": 0.5026, "mean_token_accuracy": 0.8653970956802368, "num_tokens": 304594632.0, "step": 8024 }, { "epoch": 1.0208624856888437, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.000152587890625, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.7255859375, "learning_rate": 1e-05, "loss": 0.5163, "mean_token_accuracy": 0.8672704696655273, "num_tokens": 304635255.0, "step": 8025 }, { "epoch": 1.0209896959674343, "ewc_loss": 0.10791015625, "ewc_loss_diag": 0.000152587890625, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.049959182739258, "learning_rate": 1e-05, "loss": 0.4853, "mean_token_accuracy": 0.870064377784729, "num_tokens": 304674472.0, "step": 8026 }, { "epoch": 1.0211169062460246, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.00015354156494140625, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.387980937957764, "learning_rate": 1e-05, "loss": 0.4706, "mean_token_accuracy": 0.8762966394424438, "num_tokens": 304708950.0, "step": 8027 }, { "epoch": 1.021244116524615, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.00015354156494140625, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 5.97501277923584, "learning_rate": 1e-05, "loss": 0.4984, "mean_token_accuracy": 0.866982102394104, "num_tokens": 304752849.0, "step": 8028 }, { "epoch": 1.0213713268032056, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.00015354156494140625, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 6.13470458984375, "learning_rate": 1e-05, "loss": 0.4952, "mean_token_accuracy": 0.8654508590698242, "num_tokens": 304791936.0, "step": 8029 }, { "epoch": 1.0214985370817962, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.00015354156494140625, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.875147819519043, "learning_rate": 1e-05, "loss": 0.4825, "mean_token_accuracy": 0.8709717392921448, "num_tokens": 304832228.0, "step": 8030 }, { "epoch": 1.0216257473603867, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.00015354156494140625, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 6.006039619445801, "learning_rate": 1e-05, "loss": 0.4253, "mean_token_accuracy": 0.8858927488327026, "num_tokens": 304863169.0, "step": 8031 }, { "epoch": 1.0217529576389772, "ewc_loss": 0.09765625, "ewc_loss_diag": 0.00015354156494140625, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.755110740661621, "learning_rate": 1e-05, "loss": 0.5737, "mean_token_accuracy": 0.8377511501312256, "num_tokens": 304904938.0, "step": 8032 }, { "epoch": 1.0218801679175677, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.00015354156494140625, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.982150077819824, "learning_rate": 1e-05, "loss": 0.486, "mean_token_accuracy": 0.8694881200790405, "num_tokens": 304940590.0, "step": 8033 }, { "epoch": 1.0220073781961583, "ewc_loss": 0.09375, "ewc_loss_diag": 0.00015354156494140625, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.606443881988525, "learning_rate": 1e-05, "loss": 0.4937, "mean_token_accuracy": 0.8619061708450317, "num_tokens": 304976945.0, "step": 8034 }, { "epoch": 1.0221345884747488, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.00015354156494140625, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 6.053981304168701, "learning_rate": 1e-05, "loss": 0.4886, "mean_token_accuracy": 0.8697682023048401, "num_tokens": 305021407.0, "step": 8035 }, { "epoch": 1.0222617987533393, "ewc_loss": 0.0927734375, "ewc_loss_diag": 0.00015354156494140625, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.462011814117432, "learning_rate": 1e-05, "loss": 0.4564, "mean_token_accuracy": 0.8767437934875488, "num_tokens": 305053141.0, "step": 8036 }, { "epoch": 1.0223890090319299, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.00015354156494140625, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.224924564361572, "learning_rate": 1e-05, "loss": 0.4738, "mean_token_accuracy": 0.8755253553390503, "num_tokens": 305086795.0, "step": 8037 }, { "epoch": 1.0225162193105204, "ewc_loss": 0.0908203125, "ewc_loss_diag": 0.00015354156494140625, "ewc_loss_parallel": 7.534027099609375e-05, "grad_norm": 5.334457874298096, "learning_rate": 1e-05, "loss": 0.5227, "mean_token_accuracy": 0.8556778430938721, "num_tokens": 305125759.0, "step": 8038 }, { "epoch": 1.0226434295891107, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.00015354156494140625, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.503087043762207, "learning_rate": 1e-05, "loss": 0.487, "mean_token_accuracy": 0.8700125217437744, "num_tokens": 305165169.0, "step": 8039 }, { "epoch": 1.0227706398677012, "ewc_loss": 0.0908203125, "ewc_loss_diag": 0.00015354156494140625, "ewc_loss_parallel": 7.581710815429688e-05, "grad_norm": 5.269323348999023, "learning_rate": 1e-05, "loss": 0.4253, "mean_token_accuracy": 0.8872502446174622, "num_tokens": 305203457.0, "step": 8040 }, { "epoch": 1.0228978501462918, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.00015354156494140625, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.858943939208984, "learning_rate": 1e-05, "loss": 0.5322, "mean_token_accuracy": 0.8590238094329834, "num_tokens": 305241021.0, "step": 8041 }, { "epoch": 1.0230250604248823, "ewc_loss": 0.09765625, "ewc_loss_diag": 0.00015354156494140625, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.451764106750488, "learning_rate": 1e-05, "loss": 0.5382, "mean_token_accuracy": 0.8473500609397888, "num_tokens": 305277579.0, "step": 8042 }, { "epoch": 1.0231522707034728, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.00015354156494140625, "ewc_loss_parallel": 0.00011014938354492188, "grad_norm": 7.065879821777344, "learning_rate": 1e-05, "loss": 0.5424, "mean_token_accuracy": 0.8605360388755798, "num_tokens": 305318027.0, "step": 8043 }, { "epoch": 1.0232794809820633, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.00015354156494140625, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 5.847786903381348, "learning_rate": 1e-05, "loss": 0.4967, "mean_token_accuracy": 0.8702718615531921, "num_tokens": 305351249.0, "step": 8044 }, { "epoch": 1.0234066912606539, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.00015354156494140625, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.858392715454102, "learning_rate": 1e-05, "loss": 0.4931, "mean_token_accuracy": 0.8711143732070923, "num_tokens": 305387906.0, "step": 8045 }, { "epoch": 1.0235339015392444, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.00015354156494140625, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.0958757400512695, "learning_rate": 1e-05, "loss": 0.4617, "mean_token_accuracy": 0.8775656819343567, "num_tokens": 305419511.0, "step": 8046 }, { "epoch": 1.023661111817835, "ewc_loss": 0.109375, "ewc_loss_diag": 0.00015354156494140625, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.39723539352417, "learning_rate": 1e-05, "loss": 0.4984, "mean_token_accuracy": 0.8678260445594788, "num_tokens": 305457483.0, "step": 8047 }, { "epoch": 1.0237883220964255, "ewc_loss": 0.109375, "ewc_loss_diag": 0.00015354156494140625, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.144267559051514, "learning_rate": 1e-05, "loss": 0.474, "mean_token_accuracy": 0.8728018403053284, "num_tokens": 305486498.0, "step": 8048 }, { "epoch": 1.023915532375016, "ewc_loss": 0.0986328125, "ewc_loss_diag": 0.00015354156494140625, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.927578926086426, "learning_rate": 1e-05, "loss": 0.4869, "mean_token_accuracy": 0.868057131767273, "num_tokens": 305525250.0, "step": 8049 }, { "epoch": 1.0240427426536065, "ewc_loss": 0.109375, "ewc_loss_diag": 0.00015354156494140625, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.207683086395264, "learning_rate": 1e-05, "loss": 0.4997, "mean_token_accuracy": 0.8660078048706055, "num_tokens": 305566229.0, "step": 8050 }, { "epoch": 1.0241699529321968, "ewc_loss": 0.09375, "ewc_loss_diag": 0.00015354156494140625, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.568337440490723, "learning_rate": 1e-05, "loss": 0.4521, "mean_token_accuracy": 0.8749591112136841, "num_tokens": 305601884.0, "step": 8051 }, { "epoch": 1.0242971632107873, "ewc_loss": 0.109375, "ewc_loss_diag": 0.00015354156494140625, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.398195266723633, "learning_rate": 1e-05, "loss": 0.5031, "mean_token_accuracy": 0.8661607503890991, "num_tokens": 305644495.0, "step": 8052 }, { "epoch": 1.0244243734893779, "ewc_loss": 0.091796875, "ewc_loss_diag": 0.00015354156494140625, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.377222537994385, "learning_rate": 1e-05, "loss": 0.4638, "mean_token_accuracy": 0.8738820552825928, "num_tokens": 305680662.0, "step": 8053 }, { "epoch": 1.0245515837679684, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.00015354156494140625, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.664255619049072, "learning_rate": 1e-05, "loss": 0.4822, "mean_token_accuracy": 0.8730050325393677, "num_tokens": 305720821.0, "step": 8054 }, { "epoch": 1.024678794046559, "ewc_loss": 0.09375, "ewc_loss_diag": 0.00015354156494140625, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.393462657928467, "learning_rate": 1e-05, "loss": 0.4964, "mean_token_accuracy": 0.8599880933761597, "num_tokens": 305757312.0, "step": 8055 }, { "epoch": 1.0248060043251495, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.00015354156494140625, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.912838935852051, "learning_rate": 1e-05, "loss": 0.5426, "mean_token_accuracy": 0.8588863611221313, "num_tokens": 305794372.0, "step": 8056 }, { "epoch": 1.02493321460374, "ewc_loss": 0.099609375, "ewc_loss_diag": 0.00015354156494140625, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.605074882507324, "learning_rate": 1e-05, "loss": 0.4885, "mean_token_accuracy": 0.8692733645439148, "num_tokens": 305839975.0, "step": 8057 }, { "epoch": 1.0250604248823305, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.00015354156494140625, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.955511569976807, "learning_rate": 1e-05, "loss": 0.479, "mean_token_accuracy": 0.8769190311431885, "num_tokens": 305877964.0, "step": 8058 }, { "epoch": 1.025187635160921, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.00015354156494140625, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 5.872727394104004, "learning_rate": 1e-05, "loss": 0.497, "mean_token_accuracy": 0.8678035140037537, "num_tokens": 305911077.0, "step": 8059 }, { "epoch": 1.0253148454395116, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.00015354156494140625, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.754801273345947, "learning_rate": 1e-05, "loss": 0.4781, "mean_token_accuracy": 0.8739479184150696, "num_tokens": 305947741.0, "step": 8060 }, { "epoch": 1.025442055718102, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.00015354156494140625, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 5.968376636505127, "learning_rate": 1e-05, "loss": 0.5054, "mean_token_accuracy": 0.8657466769218445, "num_tokens": 305986872.0, "step": 8061 }, { "epoch": 1.0255692659966926, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.00015354156494140625, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.479337692260742, "learning_rate": 1e-05, "loss": 0.508, "mean_token_accuracy": 0.8614867925643921, "num_tokens": 306023437.0, "step": 8062 }, { "epoch": 1.0256964762752832, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.00015354156494140625, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 5.938344478607178, "learning_rate": 1e-05, "loss": 0.4598, "mean_token_accuracy": 0.8798781037330627, "num_tokens": 306063026.0, "step": 8063 }, { "epoch": 1.0258236865538735, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.00015354156494140625, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.215737819671631, "learning_rate": 1e-05, "loss": 0.5074, "mean_token_accuracy": 0.8626505732536316, "num_tokens": 306102574.0, "step": 8064 }, { "epoch": 1.025950896832464, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.00015354156494140625, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.867005348205566, "learning_rate": 1e-05, "loss": 0.5164, "mean_token_accuracy": 0.8609788417816162, "num_tokens": 306137194.0, "step": 8065 }, { "epoch": 1.0260781071110545, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.00015354156494140625, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 6.063065528869629, "learning_rate": 1e-05, "loss": 0.4904, "mean_token_accuracy": 0.8665648102760315, "num_tokens": 306171539.0, "step": 8066 }, { "epoch": 1.026205317389645, "ewc_loss": 0.0986328125, "ewc_loss_diag": 0.00015354156494140625, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.752457141876221, "learning_rate": 1e-05, "loss": 0.507, "mean_token_accuracy": 0.8647236227989197, "num_tokens": 306212385.0, "step": 8067 }, { "epoch": 1.0263325276682356, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.0001544952392578125, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 6.007508754730225, "learning_rate": 1e-05, "loss": 0.4778, "mean_token_accuracy": 0.8692755699157715, "num_tokens": 306247624.0, "step": 8068 }, { "epoch": 1.026459737946826, "ewc_loss": 0.095703125, "ewc_loss_diag": 0.0001544952392578125, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.643832683563232, "learning_rate": 1e-05, "loss": 0.5288, "mean_token_accuracy": 0.8540956974029541, "num_tokens": 306286445.0, "step": 8069 }, { "epoch": 1.0265869482254166, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.0001544952392578125, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 6.038863658905029, "learning_rate": 1e-05, "loss": 0.4623, "mean_token_accuracy": 0.8748847246170044, "num_tokens": 306327792.0, "step": 8070 }, { "epoch": 1.0267141585040072, "ewc_loss": 0.09423828125, "ewc_loss_diag": 0.0001544952392578125, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.5244364738464355, "learning_rate": 1e-05, "loss": 0.5409, "mean_token_accuracy": 0.8500853180885315, "num_tokens": 306366811.0, "step": 8071 }, { "epoch": 1.0268413687825977, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.0001544952392578125, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.160085201263428, "learning_rate": 1e-05, "loss": 0.5036, "mean_token_accuracy": 0.8641213178634644, "num_tokens": 306403938.0, "step": 8072 }, { "epoch": 1.0269685790611882, "ewc_loss": 0.09228515625, "ewc_loss_diag": 0.0001544952392578125, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.387239456176758, "learning_rate": 1e-05, "loss": 0.4298, "mean_token_accuracy": 0.8819177746772766, "num_tokens": 306440623.0, "step": 8073 }, { "epoch": 1.0270957893397787, "ewc_loss": 0.10888671875, "ewc_loss_diag": 0.0001544952392578125, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.3921051025390625, "learning_rate": 1e-05, "loss": 0.4955, "mean_token_accuracy": 0.8659732937812805, "num_tokens": 306477623.0, "step": 8074 }, { "epoch": 1.0272229996183693, "ewc_loss": 0.091796875, "ewc_loss_diag": 0.0001544952392578125, "ewc_loss_parallel": 7.62939453125e-05, "grad_norm": 5.307200908660889, "learning_rate": 1e-05, "loss": 0.478, "mean_token_accuracy": 0.8677431344985962, "num_tokens": 306513586.0, "step": 8075 }, { "epoch": 1.0273502098969596, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.0001544952392578125, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.707562446594238, "learning_rate": 1e-05, "loss": 0.4539, "mean_token_accuracy": 0.8848376274108887, "num_tokens": 306548121.0, "step": 8076 }, { "epoch": 1.02747742017555, "ewc_loss": 0.0947265625, "ewc_loss_diag": 0.0001544952392578125, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.374091148376465, "learning_rate": 1e-05, "loss": 0.4575, "mean_token_accuracy": 0.8770620226860046, "num_tokens": 306596776.0, "step": 8077 }, { "epoch": 1.0276046304541406, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.0001544952392578125, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.964903354644775, "learning_rate": 1e-05, "loss": 0.5282, "mean_token_accuracy": 0.862247109413147, "num_tokens": 306635497.0, "step": 8078 }, { "epoch": 1.0277318407327312, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.0001544952392578125, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.711854934692383, "learning_rate": 1e-05, "loss": 0.5626, "mean_token_accuracy": 0.85489821434021, "num_tokens": 306671933.0, "step": 8079 }, { "epoch": 1.0278590510113217, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.0001544952392578125, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.876025199890137, "learning_rate": 1e-05, "loss": 0.5006, "mean_token_accuracy": 0.8692276477813721, "num_tokens": 306709313.0, "step": 8080 }, { "epoch": 1.0279862612899122, "ewc_loss": 0.10791015625, "ewc_loss_diag": 0.0001544952392578125, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 5.975470542907715, "learning_rate": 1e-05, "loss": 0.5384, "mean_token_accuracy": 0.8510263562202454, "num_tokens": 306748181.0, "step": 8081 }, { "epoch": 1.0281134715685027, "ewc_loss": 0.11181640625, "ewc_loss_diag": 0.0001544952392578125, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.486173152923584, "learning_rate": 1e-05, "loss": 0.4695, "mean_token_accuracy": 0.8792711496353149, "num_tokens": 306786353.0, "step": 8082 }, { "epoch": 1.0282406818470933, "ewc_loss": 0.109375, "ewc_loss_diag": 0.0001544952392578125, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.081305027008057, "learning_rate": 1e-05, "loss": 0.5034, "mean_token_accuracy": 0.8672360181808472, "num_tokens": 306821660.0, "step": 8083 }, { "epoch": 1.0283678921256838, "ewc_loss": 0.10205078125, "ewc_loss_diag": 0.0001544952392578125, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 6.0245585441589355, "learning_rate": 1e-05, "loss": 0.4188, "mean_token_accuracy": 0.8908185958862305, "num_tokens": 306858479.0, "step": 8084 }, { "epoch": 1.0284951024042743, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.0001544952392578125, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.149636745452881, "learning_rate": 1e-05, "loss": 0.4927, "mean_token_accuracy": 0.8671512007713318, "num_tokens": 306890483.0, "step": 8085 }, { "epoch": 1.0286223126828649, "ewc_loss": 0.0947265625, "ewc_loss_diag": 0.0001544952392578125, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.609053611755371, "learning_rate": 1e-05, "loss": 0.4701, "mean_token_accuracy": 0.871798574924469, "num_tokens": 306931865.0, "step": 8086 }, { "epoch": 1.0287495229614554, "ewc_loss": 0.10888671875, "ewc_loss_diag": 0.0001544952392578125, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.3357977867126465, "learning_rate": 1e-05, "loss": 0.5126, "mean_token_accuracy": 0.8643736243247986, "num_tokens": 306966907.0, "step": 8087 }, { "epoch": 1.0288767332400457, "ewc_loss": 0.09130859375, "ewc_loss_diag": 0.0001544952392578125, "ewc_loss_parallel": 7.581710815429688e-05, "grad_norm": 5.348672866821289, "learning_rate": 1e-05, "loss": 0.4625, "mean_token_accuracy": 0.8748146295547485, "num_tokens": 307006032.0, "step": 8088 }, { "epoch": 1.0290039435186362, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.0001544952392578125, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.6380486488342285, "learning_rate": 1e-05, "loss": 0.5118, "mean_token_accuracy": 0.8615415692329407, "num_tokens": 307045621.0, "step": 8089 }, { "epoch": 1.0291311537972267, "ewc_loss": 0.0927734375, "ewc_loss_diag": 0.0001544952392578125, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.32228946685791, "learning_rate": 1e-05, "loss": 0.5131, "mean_token_accuracy": 0.8589608669281006, "num_tokens": 307078768.0, "step": 8090 }, { "epoch": 1.0292583640758173, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.0001544952392578125, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.952186107635498, "learning_rate": 1e-05, "loss": 0.503, "mean_token_accuracy": 0.8709446787834167, "num_tokens": 307120089.0, "step": 8091 }, { "epoch": 1.0293855743544078, "ewc_loss": 0.09912109375, "ewc_loss_diag": 0.0001544952392578125, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.562407493591309, "learning_rate": 1e-05, "loss": 0.4758, "mean_token_accuracy": 0.8680092692375183, "num_tokens": 307157478.0, "step": 8092 }, { "epoch": 1.0295127846329983, "ewc_loss": 0.125, "ewc_loss_diag": 0.0001544952392578125, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 7.050317287445068, "learning_rate": 1e-05, "loss": 0.5063, "mean_token_accuracy": 0.8687196969985962, "num_tokens": 307199842.0, "step": 8093 }, { "epoch": 1.0296399949115889, "ewc_loss": 0.10693359375, "ewc_loss_diag": 0.0001544952392578125, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 5.907785892486572, "learning_rate": 1e-05, "loss": 0.4889, "mean_token_accuracy": 0.8675692081451416, "num_tokens": 307240609.0, "step": 8094 }, { "epoch": 1.0297672051901794, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.0001544952392578125, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.821929931640625, "learning_rate": 1e-05, "loss": 0.5322, "mean_token_accuracy": 0.8594542145729065, "num_tokens": 307279112.0, "step": 8095 }, { "epoch": 1.02989441546877, "ewc_loss": 0.109375, "ewc_loss_diag": 0.0001544952392578125, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.049153804779053, "learning_rate": 1e-05, "loss": 0.4728, "mean_token_accuracy": 0.8756004571914673, "num_tokens": 307313718.0, "step": 8096 }, { "epoch": 1.0300216257473604, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.0001544952392578125, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.4874420166015625, "learning_rate": 1e-05, "loss": 0.4949, "mean_token_accuracy": 0.8688064217567444, "num_tokens": 307353660.0, "step": 8097 }, { "epoch": 1.030148836025951, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.0001544952392578125, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.000948429107666, "learning_rate": 1e-05, "loss": 0.5216, "mean_token_accuracy": 0.8595079183578491, "num_tokens": 307393974.0, "step": 8098 }, { "epoch": 1.0302760463045415, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.0001544952392578125, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.208823204040527, "learning_rate": 1e-05, "loss": 0.458, "mean_token_accuracy": 0.8792787194252014, "num_tokens": 307434877.0, "step": 8099 }, { "epoch": 1.0304032565831318, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.0001544952392578125, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.914348125457764, "learning_rate": 1e-05, "loss": 0.4535, "mean_token_accuracy": 0.8772711753845215, "num_tokens": 307473167.0, "step": 8100 }, { "epoch": 1.0305304668617223, "ewc_loss": 0.10205078125, "ewc_loss_diag": 0.0001544952392578125, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 6.0297017097473145, "learning_rate": 1e-05, "loss": 0.4428, "mean_token_accuracy": 0.8799077272415161, "num_tokens": 307513781.0, "step": 8101 }, { "epoch": 1.0306576771403129, "ewc_loss": 0.10009765625, "ewc_loss_diag": 0.0001544952392578125, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.809258460998535, "learning_rate": 1e-05, "loss": 0.4234, "mean_token_accuracy": 0.8854400515556335, "num_tokens": 307549524.0, "step": 8102 }, { "epoch": 1.0307848874189034, "ewc_loss": 0.10107421875, "ewc_loss_diag": 0.0001544952392578125, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.970959186553955, "learning_rate": 1e-05, "loss": 0.4799, "mean_token_accuracy": 0.8672213554382324, "num_tokens": 307585196.0, "step": 8103 }, { "epoch": 1.030912097697494, "ewc_loss": 0.095703125, "ewc_loss_diag": 0.0001544952392578125, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.685521125793457, "learning_rate": 1e-05, "loss": 0.5342, "mean_token_accuracy": 0.8557448387145996, "num_tokens": 307623002.0, "step": 8104 }, { "epoch": 1.0310393079760845, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.0001544952392578125, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.994271278381348, "learning_rate": 1e-05, "loss": 0.4777, "mean_token_accuracy": 0.8722290396690369, "num_tokens": 307667736.0, "step": 8105 }, { "epoch": 1.031166518254675, "ewc_loss": 0.09423828125, "ewc_loss_diag": 0.0001544952392578125, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.542741298675537, "learning_rate": 1e-05, "loss": 0.4461, "mean_token_accuracy": 0.8797118663787842, "num_tokens": 307706031.0, "step": 8106 }, { "epoch": 1.0312937285332655, "ewc_loss": 0.10400390625, "ewc_loss_diag": 0.0001544952392578125, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 6.117332935333252, "learning_rate": 1e-05, "loss": 0.531, "mean_token_accuracy": 0.8565610647201538, "num_tokens": 307746333.0, "step": 8107 }, { "epoch": 1.031420938811856, "ewc_loss": 0.091796875, "ewc_loss_diag": 0.0001544952392578125, "ewc_loss_parallel": 7.62939453125e-05, "grad_norm": 5.3963212966918945, "learning_rate": 1e-05, "loss": 0.5146, "mean_token_accuracy": 0.8546174764633179, "num_tokens": 307784479.0, "step": 8108 }, { "epoch": 1.0315481490904466, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.0001544952392578125, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.346696376800537, "learning_rate": 1e-05, "loss": 0.4952, "mean_token_accuracy": 0.8676813840866089, "num_tokens": 307818245.0, "step": 8109 }, { "epoch": 1.031675359369037, "ewc_loss": 0.09033203125, "ewc_loss_diag": 0.0001544952392578125, "ewc_loss_parallel": 7.486343383789062e-05, "grad_norm": 5.276534080505371, "learning_rate": 1e-05, "loss": 0.4529, "mean_token_accuracy": 0.8740128874778748, "num_tokens": 307858730.0, "step": 8110 }, { "epoch": 1.0318025696476276, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.0001544952392578125, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.689517974853516, "learning_rate": 1e-05, "loss": 0.4598, "mean_token_accuracy": 0.8820073008537292, "num_tokens": 307897067.0, "step": 8111 }, { "epoch": 1.0319297799262181, "ewc_loss": 0.09228515625, "ewc_loss_diag": 0.0001544952392578125, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.309577465057373, "learning_rate": 1e-05, "loss": 0.4722, "mean_token_accuracy": 0.866979718208313, "num_tokens": 307935896.0, "step": 8112 }, { "epoch": 1.0320569902048085, "ewc_loss": 0.125, "ewc_loss_diag": 0.0001544952392578125, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 7.049023151397705, "learning_rate": 1e-05, "loss": 0.5719, "mean_token_accuracy": 0.8486689329147339, "num_tokens": 307970930.0, "step": 8113 }, { "epoch": 1.032184200483399, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.0001544952392578125, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.628810405731201, "learning_rate": 1e-05, "loss": 0.5223, "mean_token_accuracy": 0.8568954467773438, "num_tokens": 308015857.0, "step": 8114 }, { "epoch": 1.0323114107619895, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0001544952392578125, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 7.086574554443359, "learning_rate": 1e-05, "loss": 0.5111, "mean_token_accuracy": 0.8696219325065613, "num_tokens": 308055432.0, "step": 8115 }, { "epoch": 1.03243862104058, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.0001544952392578125, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.030704021453857, "learning_rate": 1e-05, "loss": 0.5247, "mean_token_accuracy": 0.8574842214584351, "num_tokens": 308094922.0, "step": 8116 }, { "epoch": 1.0325658313191706, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.0001544952392578125, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.726895809173584, "learning_rate": 1e-05, "loss": 0.4868, "mean_token_accuracy": 0.8708720803260803, "num_tokens": 308132941.0, "step": 8117 }, { "epoch": 1.032693041597761, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.0001544952392578125, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.15950870513916, "learning_rate": 1e-05, "loss": 0.5527, "mean_token_accuracy": 0.8508168458938599, "num_tokens": 308176749.0, "step": 8118 }, { "epoch": 1.0328202518763516, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.0001544952392578125, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.243201732635498, "learning_rate": 1e-05, "loss": 0.5116, "mean_token_accuracy": 0.8594611883163452, "num_tokens": 308215845.0, "step": 8119 }, { "epoch": 1.0329474621549422, "ewc_loss": 0.109375, "ewc_loss_diag": 0.0001544952392578125, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.168997764587402, "learning_rate": 1e-05, "loss": 0.4671, "mean_token_accuracy": 0.8760913014411926, "num_tokens": 308254220.0, "step": 8120 }, { "epoch": 1.0330746724335327, "ewc_loss": 0.09814453125, "ewc_loss_diag": 0.0001544952392578125, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.826954364776611, "learning_rate": 1e-05, "loss": 0.5255, "mean_token_accuracy": 0.8569958806037903, "num_tokens": 308290059.0, "step": 8121 }, { "epoch": 1.0332018827121232, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.0001544952392578125, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.247037410736084, "learning_rate": 1e-05, "loss": 0.5753, "mean_token_accuracy": 0.8408803343772888, "num_tokens": 308320478.0, "step": 8122 }, { "epoch": 1.0333290929907137, "ewc_loss": 0.09375, "ewc_loss_diag": 0.0001544952392578125, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.527957916259766, "learning_rate": 1e-05, "loss": 0.459, "mean_token_accuracy": 0.8725600242614746, "num_tokens": 308356094.0, "step": 8123 }, { "epoch": 1.0334563032693043, "ewc_loss": 0.10986328125, "ewc_loss_diag": 0.00015544891357421875, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.448620319366455, "learning_rate": 1e-05, "loss": 0.4947, "mean_token_accuracy": 0.8710454702377319, "num_tokens": 308393120.0, "step": 8124 }, { "epoch": 1.0335835135478946, "ewc_loss": 0.0927734375, "ewc_loss_diag": 0.00015544891357421875, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.382005214691162, "learning_rate": 1e-05, "loss": 0.446, "mean_token_accuracy": 0.8797920942306519, "num_tokens": 308424153.0, "step": 8125 }, { "epoch": 1.033710723826485, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.00015544891357421875, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.6640849113464355, "learning_rate": 1e-05, "loss": 0.488, "mean_token_accuracy": 0.8720072507858276, "num_tokens": 308462884.0, "step": 8126 }, { "epoch": 1.0338379341050756, "ewc_loss": 0.09423828125, "ewc_loss_diag": 0.00015544891357421875, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.412927627563477, "learning_rate": 1e-05, "loss": 0.4132, "mean_token_accuracy": 0.8894420862197876, "num_tokens": 308501185.0, "step": 8127 }, { "epoch": 1.0339651443836662, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.00015544891357421875, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.883544921875, "learning_rate": 1e-05, "loss": 0.5416, "mean_token_accuracy": 0.8588927984237671, "num_tokens": 308538770.0, "step": 8128 }, { "epoch": 1.0340923546622567, "ewc_loss": 0.10009765625, "ewc_loss_diag": 0.00015544891357421875, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.62619161605835, "learning_rate": 1e-05, "loss": 0.4724, "mean_token_accuracy": 0.8695529699325562, "num_tokens": 308575224.0, "step": 8129 }, { "epoch": 1.0342195649408472, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.00015544891357421875, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.892149448394775, "learning_rate": 1e-05, "loss": 0.4768, "mean_token_accuracy": 0.8805949091911316, "num_tokens": 308611859.0, "step": 8130 }, { "epoch": 1.0343467752194377, "ewc_loss": 0.10595703125, "ewc_loss_diag": 0.00015544891357421875, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 5.8500566482543945, "learning_rate": 1e-05, "loss": 0.5279, "mean_token_accuracy": 0.8536897897720337, "num_tokens": 308648212.0, "step": 8131 }, { "epoch": 1.0344739854980283, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.00015544891357421875, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.690550327301025, "learning_rate": 1e-05, "loss": 0.5162, "mean_token_accuracy": 0.8658818006515503, "num_tokens": 308689830.0, "step": 8132 }, { "epoch": 1.0346011957766188, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.00015544891357421875, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 5.943641185760498, "learning_rate": 1e-05, "loss": 0.5422, "mean_token_accuracy": 0.8529317378997803, "num_tokens": 308726238.0, "step": 8133 }, { "epoch": 1.0347284060552093, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.00015544891357421875, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.422452449798584, "learning_rate": 1e-05, "loss": 0.5132, "mean_token_accuracy": 0.8581142425537109, "num_tokens": 308755090.0, "step": 8134 }, { "epoch": 1.0348556163337999, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.00015544891357421875, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 5.921173572540283, "learning_rate": 1e-05, "loss": 0.5282, "mean_token_accuracy": 0.8529393672943115, "num_tokens": 308794025.0, "step": 8135 }, { "epoch": 1.0349828266123904, "ewc_loss": 0.10595703125, "ewc_loss_diag": 0.00015544891357421875, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.1706719398498535, "learning_rate": 1e-05, "loss": 0.505, "mean_token_accuracy": 0.8681409955024719, "num_tokens": 308831521.0, "step": 8136 }, { "epoch": 1.0351100368909807, "ewc_loss": 0.10205078125, "ewc_loss_diag": 0.00015544891357421875, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.841343879699707, "learning_rate": 1e-05, "loss": 0.4844, "mean_token_accuracy": 0.8733636736869812, "num_tokens": 308868113.0, "step": 8137 }, { "epoch": 1.0352372471695712, "ewc_loss": 0.10107421875, "ewc_loss_diag": 0.00015544891357421875, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 6.022639274597168, "learning_rate": 1e-05, "loss": 0.4893, "mean_token_accuracy": 0.8668972849845886, "num_tokens": 308905070.0, "step": 8138 }, { "epoch": 1.0353644574481617, "ewc_loss": 0.099609375, "ewc_loss_diag": 0.00015544891357421875, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.761032581329346, "learning_rate": 1e-05, "loss": 0.5046, "mean_token_accuracy": 0.8604115843772888, "num_tokens": 308942703.0, "step": 8139 }, { "epoch": 1.0354916677267523, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.00015544891357421875, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.963836193084717, "learning_rate": 1e-05, "loss": 0.4771, "mean_token_accuracy": 0.8703947067260742, "num_tokens": 308981328.0, "step": 8140 }, { "epoch": 1.0356188780053428, "ewc_loss": 0.0966796875, "ewc_loss_diag": 0.00015544891357421875, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.649872779846191, "learning_rate": 1e-05, "loss": 0.4362, "mean_token_accuracy": 0.8813320398330688, "num_tokens": 309017751.0, "step": 8141 }, { "epoch": 1.0357460882839333, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.00015544891357421875, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.992468357086182, "learning_rate": 1e-05, "loss": 0.4843, "mean_token_accuracy": 0.8687236905097961, "num_tokens": 309052216.0, "step": 8142 }, { "epoch": 1.0358732985625239, "ewc_loss": 0.0947265625, "ewc_loss_diag": 0.00015544891357421875, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.536191940307617, "learning_rate": 1e-05, "loss": 0.5093, "mean_token_accuracy": 0.8670275807380676, "num_tokens": 309091323.0, "step": 8143 }, { "epoch": 1.0360005088411144, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.00015544891357421875, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.105279445648193, "learning_rate": 1e-05, "loss": 0.497, "mean_token_accuracy": 0.8682506084442139, "num_tokens": 309125324.0, "step": 8144 }, { "epoch": 1.036127719119705, "ewc_loss": 0.0927734375, "ewc_loss_diag": 0.00015544891357421875, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.402588844299316, "learning_rate": 1e-05, "loss": 0.4877, "mean_token_accuracy": 0.8681607246398926, "num_tokens": 309166009.0, "step": 8145 }, { "epoch": 1.0362549293982954, "ewc_loss": 0.10791015625, "ewc_loss_diag": 0.00015544891357421875, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.30488395690918, "learning_rate": 1e-05, "loss": 0.4579, "mean_token_accuracy": 0.879606306552887, "num_tokens": 309205822.0, "step": 8146 }, { "epoch": 1.036382139676886, "ewc_loss": 0.09130859375, "ewc_loss_diag": 0.00015544891357421875, "ewc_loss_parallel": 7.581710815429688e-05, "grad_norm": 5.297650337219238, "learning_rate": 1e-05, "loss": 0.4694, "mean_token_accuracy": 0.8712178468704224, "num_tokens": 309243463.0, "step": 8147 }, { "epoch": 1.0365093499554765, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.00015544891357421875, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.620073318481445, "learning_rate": 1e-05, "loss": 0.5153, "mean_token_accuracy": 0.8643386363983154, "num_tokens": 309276904.0, "step": 8148 }, { "epoch": 1.0366365602340668, "ewc_loss": 0.09375, "ewc_loss_diag": 0.00015544891357421875, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.303487300872803, "learning_rate": 1e-05, "loss": 0.4813, "mean_token_accuracy": 0.8670063614845276, "num_tokens": 309314292.0, "step": 8149 }, { "epoch": 1.0367637705126573, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.00015544891357421875, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 6.95365571975708, "learning_rate": 1e-05, "loss": 0.543, "mean_token_accuracy": 0.8608639240264893, "num_tokens": 309349231.0, "step": 8150 }, { "epoch": 1.0368909807912479, "ewc_loss": 0.099609375, "ewc_loss_diag": 0.00015544891357421875, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.564635753631592, "learning_rate": 1e-05, "loss": 0.5664, "mean_token_accuracy": 0.8429399728775024, "num_tokens": 309387901.0, "step": 8151 }, { "epoch": 1.0370181910698384, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.00015544891357421875, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 7.01425838470459, "learning_rate": 1e-05, "loss": 0.4915, "mean_token_accuracy": 0.8710452318191528, "num_tokens": 309427167.0, "step": 8152 }, { "epoch": 1.037145401348429, "ewc_loss": 0.109375, "ewc_loss_diag": 0.00015544891357421875, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 5.920103073120117, "learning_rate": 1e-05, "loss": 0.4933, "mean_token_accuracy": 0.8697463870048523, "num_tokens": 309464121.0, "step": 8153 }, { "epoch": 1.0372726116270194, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.00015544891357421875, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.7049150466918945, "learning_rate": 1e-05, "loss": 0.5033, "mean_token_accuracy": 0.8707168102264404, "num_tokens": 309506759.0, "step": 8154 }, { "epoch": 1.03739982190561, "ewc_loss": 0.11083984375, "ewc_loss_diag": 0.00015544891357421875, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.076096057891846, "learning_rate": 1e-05, "loss": 0.4714, "mean_token_accuracy": 0.8762534856796265, "num_tokens": 309546059.0, "step": 8155 }, { "epoch": 1.0375270321842005, "ewc_loss": 0.10791015625, "ewc_loss_diag": 0.00015544891357421875, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.25982141494751, "learning_rate": 1e-05, "loss": 0.5051, "mean_token_accuracy": 0.8627400398254395, "num_tokens": 309585076.0, "step": 8156 }, { "epoch": 1.037654242462791, "ewc_loss": 0.109375, "ewc_loss_diag": 0.00015544891357421875, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.126145362854004, "learning_rate": 1e-05, "loss": 0.4689, "mean_token_accuracy": 0.8774963617324829, "num_tokens": 309623354.0, "step": 8157 }, { "epoch": 1.0377814527413816, "ewc_loss": 0.0986328125, "ewc_loss_diag": 0.00015544891357421875, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.8142194747924805, "learning_rate": 1e-05, "loss": 0.5245, "mean_token_accuracy": 0.8533895015716553, "num_tokens": 309668420.0, "step": 8158 }, { "epoch": 1.037908663019972, "ewc_loss": 0.10888671875, "ewc_loss_diag": 0.00015544891357421875, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.235976219177246, "learning_rate": 1e-05, "loss": 0.4961, "mean_token_accuracy": 0.8702414035797119, "num_tokens": 309711417.0, "step": 8159 }, { "epoch": 1.0380358732985626, "ewc_loss": 0.0947265625, "ewc_loss_diag": 0.00015544891357421875, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.503685474395752, "learning_rate": 1e-05, "loss": 0.4693, "mean_token_accuracy": 0.8722249269485474, "num_tokens": 309749938.0, "step": 8160 }, { "epoch": 1.0381630835771531, "ewc_loss": 0.11083984375, "ewc_loss_diag": 0.00015544891357421875, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.459625244140625, "learning_rate": 1e-05, "loss": 0.4802, "mean_token_accuracy": 0.8729493618011475, "num_tokens": 309783328.0, "step": 8161 }, { "epoch": 1.0382902938557435, "ewc_loss": 0.09228515625, "ewc_loss_diag": 0.00015544891357421875, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.337218284606934, "learning_rate": 1e-05, "loss": 0.4645, "mean_token_accuracy": 0.8751325607299805, "num_tokens": 309821347.0, "step": 8162 }, { "epoch": 1.038417504134334, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.00015544891357421875, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.7643327713012695, "learning_rate": 1e-05, "loss": 0.5531, "mean_token_accuracy": 0.8506525754928589, "num_tokens": 309858430.0, "step": 8163 }, { "epoch": 1.0385447144129245, "ewc_loss": 0.09521484375, "ewc_loss_diag": 0.00015544891357421875, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.421803951263428, "learning_rate": 1e-05, "loss": 0.4764, "mean_token_accuracy": 0.8680360317230225, "num_tokens": 309896096.0, "step": 8164 }, { "epoch": 1.038671924691515, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.00015544891357421875, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 6.985941410064697, "learning_rate": 1e-05, "loss": 0.5096, "mean_token_accuracy": 0.86722731590271, "num_tokens": 309929069.0, "step": 8165 }, { "epoch": 1.0387991349701056, "ewc_loss": 0.10205078125, "ewc_loss_diag": 0.00015544891357421875, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.693873405456543, "learning_rate": 1e-05, "loss": 0.5205, "mean_token_accuracy": 0.8555260300636292, "num_tokens": 309969706.0, "step": 8166 }, { "epoch": 1.038926345248696, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.00015544891357421875, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.945046901702881, "learning_rate": 1e-05, "loss": 0.5149, "mean_token_accuracy": 0.8630973100662231, "num_tokens": 310003990.0, "step": 8167 }, { "epoch": 1.0390535555272866, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.00015544891357421875, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 5.937949180603027, "learning_rate": 1e-05, "loss": 0.5478, "mean_token_accuracy": 0.8479174375534058, "num_tokens": 310036253.0, "step": 8168 }, { "epoch": 1.0391807658058771, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.00015544891357421875, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.690742492675781, "learning_rate": 1e-05, "loss": 0.4454, "mean_token_accuracy": 0.8829818964004517, "num_tokens": 310073189.0, "step": 8169 }, { "epoch": 1.0393079760844677, "ewc_loss": 0.10595703125, "ewc_loss_diag": 0.00015544891357421875, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 5.973364353179932, "learning_rate": 1e-05, "loss": 0.5103, "mean_token_accuracy": 0.8603777885437012, "num_tokens": 310110432.0, "step": 8170 }, { "epoch": 1.0394351863630582, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.00015544891357421875, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.406237602233887, "learning_rate": 1e-05, "loss": 0.5299, "mean_token_accuracy": 0.86124587059021, "num_tokens": 310154830.0, "step": 8171 }, { "epoch": 1.0395623966416487, "ewc_loss": 0.10302734375, "ewc_loss_diag": 0.00015544891357421875, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 5.925261974334717, "learning_rate": 1e-05, "loss": 0.513, "mean_token_accuracy": 0.8540545701980591, "num_tokens": 310187270.0, "step": 8172 }, { "epoch": 1.0396896069202393, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.00015544891357421875, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 6.183496475219727, "learning_rate": 1e-05, "loss": 0.4853, "mean_token_accuracy": 0.8698263168334961, "num_tokens": 310223736.0, "step": 8173 }, { "epoch": 1.0398168171988296, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.00015544891357421875, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.841675281524658, "learning_rate": 1e-05, "loss": 0.5597, "mean_token_accuracy": 0.8502989411354065, "num_tokens": 310259262.0, "step": 8174 }, { "epoch": 1.03994402747742, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.00015544891357421875, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 6.0543389320373535, "learning_rate": 1e-05, "loss": 0.4964, "mean_token_accuracy": 0.8673818111419678, "num_tokens": 310300438.0, "step": 8175 }, { "epoch": 1.0400712377560106, "ewc_loss": 0.09814453125, "ewc_loss_diag": 0.00015544891357421875, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.726848125457764, "learning_rate": 1e-05, "loss": 0.4644, "mean_token_accuracy": 0.8770182132720947, "num_tokens": 310336630.0, "step": 8176 }, { "epoch": 1.0401984480346012, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.00015544891357421875, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 6.043338775634766, "learning_rate": 1e-05, "loss": 0.5043, "mean_token_accuracy": 0.8623880743980408, "num_tokens": 310368755.0, "step": 8177 }, { "epoch": 1.0403256583131917, "ewc_loss": 0.0947265625, "ewc_loss_diag": 0.00015544891357421875, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.590493202209473, "learning_rate": 1e-05, "loss": 0.4363, "mean_token_accuracy": 0.8785284757614136, "num_tokens": 310403520.0, "step": 8178 }, { "epoch": 1.0404528685917822, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.00015544891357421875, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 6.1363396644592285, "learning_rate": 1e-05, "loss": 0.4875, "mean_token_accuracy": 0.8652018904685974, "num_tokens": 310433805.0, "step": 8179 }, { "epoch": 1.0405800788703727, "ewc_loss": 0.09228515625, "ewc_loss_diag": 0.00015544891357421875, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.463413238525391, "learning_rate": 1e-05, "loss": 0.4924, "mean_token_accuracy": 0.8619036674499512, "num_tokens": 310469801.0, "step": 8180 }, { "epoch": 1.0407072891489633, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.00015544891357421875, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.310179233551025, "learning_rate": 1e-05, "loss": 0.4391, "mean_token_accuracy": 0.8836740851402283, "num_tokens": 310504972.0, "step": 8181 }, { "epoch": 1.0408344994275538, "ewc_loss": 0.09228515625, "ewc_loss_diag": 0.00015544891357421875, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.361725807189941, "learning_rate": 1e-05, "loss": 0.4849, "mean_token_accuracy": 0.8624473214149475, "num_tokens": 310537851.0, "step": 8182 }, { "epoch": 1.0409617097061443, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.00015544891357421875, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.556685447692871, "learning_rate": 1e-05, "loss": 0.4918, "mean_token_accuracy": 0.8677878379821777, "num_tokens": 310572647.0, "step": 8183 }, { "epoch": 1.0410889199847349, "ewc_loss": 0.09326171875, "ewc_loss_diag": 0.00015544891357421875, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.36196231842041, "learning_rate": 1e-05, "loss": 0.4982, "mean_token_accuracy": 0.8630567193031311, "num_tokens": 310612239.0, "step": 8184 }, { "epoch": 1.0412161302633254, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.00015544891357421875, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.82829475402832, "learning_rate": 1e-05, "loss": 0.5311, "mean_token_accuracy": 0.859713077545166, "num_tokens": 310650442.0, "step": 8185 }, { "epoch": 1.0413433405419157, "ewc_loss": 0.09912109375, "ewc_loss_diag": 0.00015544891357421875, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.539768695831299, "learning_rate": 1e-05, "loss": 0.502, "mean_token_accuracy": 0.8642469644546509, "num_tokens": 310687687.0, "step": 8186 }, { "epoch": 1.0414705508205062, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.00015544891357421875, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.905894756317139, "learning_rate": 1e-05, "loss": 0.5447, "mean_token_accuracy": 0.858729362487793, "num_tokens": 310725165.0, "step": 8187 }, { "epoch": 1.0415977610990967, "ewc_loss": 0.10498046875, "ewc_loss_diag": 0.00015544891357421875, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 5.835570812225342, "learning_rate": 1e-05, "loss": 0.4425, "mean_token_accuracy": 0.8828208446502686, "num_tokens": 310762532.0, "step": 8188 }, { "epoch": 1.0417249713776873, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.00015544891357421875, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.65625524520874, "learning_rate": 1e-05, "loss": 0.5278, "mean_token_accuracy": 0.8588665127754211, "num_tokens": 310800903.0, "step": 8189 }, { "epoch": 1.0418521816562778, "ewc_loss": 0.109375, "ewc_loss_diag": 0.00015544891357421875, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.027261257171631, "learning_rate": 1e-05, "loss": 0.5458, "mean_token_accuracy": 0.8468797206878662, "num_tokens": 310841391.0, "step": 8190 }, { "epoch": 1.0419793919348683, "ewc_loss": 0.10693359375, "ewc_loss_diag": 0.00015544891357421875, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.210940837860107, "learning_rate": 1e-05, "loss": 0.513, "mean_token_accuracy": 0.8588200807571411, "num_tokens": 310879027.0, "step": 8191 }, { "epoch": 1.0421066022134589, "ewc_loss": 0.109375, "ewc_loss_diag": 0.00015544891357421875, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.106684684753418, "learning_rate": 1e-05, "loss": 0.4884, "mean_token_accuracy": 0.8709205389022827, "num_tokens": 310918448.0, "step": 8192 }, { "epoch": 1.0422338124920494, "ewc_loss": 0.0966796875, "ewc_loss_diag": 0.00015544891357421875, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.754047393798828, "learning_rate": 1e-05, "loss": 0.4952, "mean_token_accuracy": 0.8625322580337524, "num_tokens": 310955641.0, "step": 8193 }, { "epoch": 1.04236102277064, "ewc_loss": 0.109375, "ewc_loss_diag": 0.00015544891357421875, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.266874313354492, "learning_rate": 1e-05, "loss": 0.574, "mean_token_accuracy": 0.8445009589195251, "num_tokens": 310996487.0, "step": 8194 }, { "epoch": 1.0424882330492304, "ewc_loss": 0.09228515625, "ewc_loss_diag": 0.00015544891357421875, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.40257453918457, "learning_rate": 1e-05, "loss": 0.5107, "mean_token_accuracy": 0.8594078421592712, "num_tokens": 311038208.0, "step": 8195 }, { "epoch": 1.042615443327821, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.00015544891357421875, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.567892074584961, "learning_rate": 1e-05, "loss": 0.4982, "mean_token_accuracy": 0.867993950843811, "num_tokens": 311073720.0, "step": 8196 }, { "epoch": 1.0427426536064115, "ewc_loss": 0.0927734375, "ewc_loss_diag": 0.00015544891357421875, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.268314361572266, "learning_rate": 1e-05, "loss": 0.4341, "mean_token_accuracy": 0.88066166639328, "num_tokens": 311115488.0, "step": 8197 }, { "epoch": 1.0428698638850018, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.00015544891357421875, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.947526931762695, "learning_rate": 1e-05, "loss": 0.5064, "mean_token_accuracy": 0.8683403730392456, "num_tokens": 311159217.0, "step": 8198 }, { "epoch": 1.0429970741635923, "ewc_loss": 0.09716796875, "ewc_loss_diag": 0.000156402587890625, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.472287654876709, "learning_rate": 1e-05, "loss": 0.5202, "mean_token_accuracy": 0.8564455509185791, "num_tokens": 311196058.0, "step": 8199 }, { "epoch": 1.0431242844421829, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.000156402587890625, "ewc_loss_parallel": 0.0001125335693359375, "grad_norm": 7.174759864807129, "learning_rate": 1e-05, "loss": 0.4965, "mean_token_accuracy": 0.8733648657798767, "num_tokens": 311231024.0, "step": 8200 }, { "epoch": 1.0432514947207734, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.000156402587890625, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 5.896096706390381, "learning_rate": 1e-05, "loss": 0.4882, "mean_token_accuracy": 0.867828369140625, "num_tokens": 311266883.0, "step": 8201 }, { "epoch": 1.043378704999364, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.000156402587890625, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.989950180053711, "learning_rate": 1e-05, "loss": 0.5583, "mean_token_accuracy": 0.8500746488571167, "num_tokens": 311302594.0, "step": 8202 }, { "epoch": 1.0435059152779544, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.00015544891357421875, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.128387928009033, "learning_rate": 1e-05, "loss": 0.4373, "mean_token_accuracy": 0.885948896408081, "num_tokens": 311334384.0, "step": 8203 }, { "epoch": 1.043633125556545, "ewc_loss": 0.11474609375, "ewc_loss_diag": 0.00015544891357421875, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.615097522735596, "learning_rate": 1e-05, "loss": 0.4937, "mean_token_accuracy": 0.8696930408477783, "num_tokens": 311381350.0, "step": 8204 }, { "epoch": 1.0437603358351355, "ewc_loss": 0.10986328125, "ewc_loss_diag": 0.00015544891357421875, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.081066608428955, "learning_rate": 1e-05, "loss": 0.4822, "mean_token_accuracy": 0.8731331825256348, "num_tokens": 311418107.0, "step": 8205 }, { "epoch": 1.043887546113726, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.00015544891357421875, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.304225444793701, "learning_rate": 1e-05, "loss": 0.4939, "mean_token_accuracy": 0.8672281503677368, "num_tokens": 311457881.0, "step": 8206 }, { "epoch": 1.0440147563923166, "ewc_loss": 0.10400390625, "ewc_loss_diag": 0.00015544891357421875, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 5.986328601837158, "learning_rate": 1e-05, "loss": 0.5012, "mean_token_accuracy": 0.8607947826385498, "num_tokens": 311492821.0, "step": 8207 }, { "epoch": 1.044141966670907, "ewc_loss": 0.10205078125, "ewc_loss_diag": 0.00015544891357421875, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 6.10153865814209, "learning_rate": 1e-05, "loss": 0.4968, "mean_token_accuracy": 0.8664005994796753, "num_tokens": 311527981.0, "step": 8208 }, { "epoch": 1.0442691769494976, "ewc_loss": 0.099609375, "ewc_loss_diag": 0.00015544891357421875, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.845128059387207, "learning_rate": 1e-05, "loss": 0.5268, "mean_token_accuracy": 0.8573906421661377, "num_tokens": 311568853.0, "step": 8209 }, { "epoch": 1.0443963872280881, "ewc_loss": 0.10205078125, "ewc_loss_diag": 0.00015544891357421875, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 6.014288902282715, "learning_rate": 1e-05, "loss": 0.46, "mean_token_accuracy": 0.8782411813735962, "num_tokens": 311613321.0, "step": 8210 }, { "epoch": 1.0445235975066784, "ewc_loss": 0.0966796875, "ewc_loss_diag": 0.00015544891357421875, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.697203636169434, "learning_rate": 1e-05, "loss": 0.496, "mean_token_accuracy": 0.8619644641876221, "num_tokens": 311656849.0, "step": 8211 }, { "epoch": 1.044650807785269, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.00015544891357421875, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 6.066943645477295, "learning_rate": 1e-05, "loss": 0.5356, "mean_token_accuracy": 0.8541256785392761, "num_tokens": 311692231.0, "step": 8212 }, { "epoch": 1.0447780180638595, "ewc_loss": 0.09375, "ewc_loss_diag": 0.00015544891357421875, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.534952640533447, "learning_rate": 1e-05, "loss": 0.4459, "mean_token_accuracy": 0.8782665133476257, "num_tokens": 311732793.0, "step": 8213 }, { "epoch": 1.04490522834245, "ewc_loss": 0.10498046875, "ewc_loss_diag": 0.00015544891357421875, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 6.207632541656494, "learning_rate": 1e-05, "loss": 0.5188, "mean_token_accuracy": 0.8634176254272461, "num_tokens": 311773320.0, "step": 8214 }, { "epoch": 1.0450324386210406, "ewc_loss": 0.0927734375, "ewc_loss_diag": 0.00015544891357421875, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.389080047607422, "learning_rate": 1e-05, "loss": 0.4892, "mean_token_accuracy": 0.863723874092102, "num_tokens": 311822187.0, "step": 8215 }, { "epoch": 1.045159648899631, "ewc_loss": 0.11083984375, "ewc_loss_diag": 0.000156402587890625, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.471598148345947, "learning_rate": 1e-05, "loss": 0.4953, "mean_token_accuracy": 0.8649622201919556, "num_tokens": 311860272.0, "step": 8216 }, { "epoch": 1.0452868591782216, "ewc_loss": 0.091796875, "ewc_loss_diag": 0.000156402587890625, "ewc_loss_parallel": 7.62939453125e-05, "grad_norm": 5.306262493133545, "learning_rate": 1e-05, "loss": 0.4702, "mean_token_accuracy": 0.8666977882385254, "num_tokens": 311894479.0, "step": 8217 }, { "epoch": 1.0454140694568121, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.000156402587890625, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.826347827911377, "learning_rate": 1e-05, "loss": 0.469, "mean_token_accuracy": 0.8774521350860596, "num_tokens": 311929401.0, "step": 8218 }, { "epoch": 1.0455412797354027, "ewc_loss": 0.095703125, "ewc_loss_diag": 0.000156402587890625, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.426121711730957, "learning_rate": 1e-05, "loss": 0.5013, "mean_token_accuracy": 0.859473705291748, "num_tokens": 311969401.0, "step": 8219 }, { "epoch": 1.0456684900139932, "ewc_loss": 0.125, "ewc_loss_diag": 0.000156402587890625, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 7.072367191314697, "learning_rate": 1e-05, "loss": 0.5592, "mean_token_accuracy": 0.8567274808883667, "num_tokens": 312004295.0, "step": 8220 }, { "epoch": 1.0457957002925837, "ewc_loss": 0.10595703125, "ewc_loss_diag": 0.000156402587890625, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 5.803397178649902, "learning_rate": 1e-05, "loss": 0.4877, "mean_token_accuracy": 0.8650485873222351, "num_tokens": 312038487.0, "step": 8221 }, { "epoch": 1.0459229105711743, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.000156402587890625, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.917840480804443, "learning_rate": 1e-05, "loss": 0.5019, "mean_token_accuracy": 0.8670551180839539, "num_tokens": 312076011.0, "step": 8222 }, { "epoch": 1.0460501208497646, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.000156402587890625, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.0965375900268555, "learning_rate": 1e-05, "loss": 0.5221, "mean_token_accuracy": 0.8591487407684326, "num_tokens": 312114489.0, "step": 8223 }, { "epoch": 1.046177331128355, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.000156402587890625, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.468310356140137, "learning_rate": 1e-05, "loss": 0.5051, "mean_token_accuracy": 0.8618801832199097, "num_tokens": 312152388.0, "step": 8224 }, { "epoch": 1.0463045414069456, "ewc_loss": 0.11083984375, "ewc_loss_diag": 0.000156402587890625, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.159451007843018, "learning_rate": 1e-05, "loss": 0.4826, "mean_token_accuracy": 0.8722946643829346, "num_tokens": 312190269.0, "step": 8225 }, { "epoch": 1.0464317516855361, "ewc_loss": 0.10009765625, "ewc_loss_diag": 0.000156402587890625, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.986487865447998, "learning_rate": 1e-05, "loss": 0.4628, "mean_token_accuracy": 0.8762402534484863, "num_tokens": 312229909.0, "step": 8226 }, { "epoch": 1.0465589619641267, "ewc_loss": 0.10986328125, "ewc_loss_diag": 0.000156402587890625, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.255801677703857, "learning_rate": 1e-05, "loss": 0.5668, "mean_token_accuracy": 0.8510302901268005, "num_tokens": 312267527.0, "step": 8227 }, { "epoch": 1.0466861722427172, "ewc_loss": 0.09423828125, "ewc_loss_diag": 0.000156402587890625, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.583755970001221, "learning_rate": 1e-05, "loss": 0.4742, "mean_token_accuracy": 0.8692835569381714, "num_tokens": 312310632.0, "step": 8228 }, { "epoch": 1.0468133825213077, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.000156402587890625, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.443663120269775, "learning_rate": 1e-05, "loss": 0.5341, "mean_token_accuracy": 0.85756915807724, "num_tokens": 312351570.0, "step": 8229 }, { "epoch": 1.0469405927998983, "ewc_loss": 0.09228515625, "ewc_loss_diag": 0.000156402587890625, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.375746726989746, "learning_rate": 1e-05, "loss": 0.4162, "mean_token_accuracy": 0.8889509439468384, "num_tokens": 312382765.0, "step": 8230 }, { "epoch": 1.0470678030784888, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.000156402587890625, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.731009483337402, "learning_rate": 1e-05, "loss": 0.4786, "mean_token_accuracy": 0.875908613204956, "num_tokens": 312415517.0, "step": 8231 }, { "epoch": 1.0471950133570793, "ewc_loss": 0.09326171875, "ewc_loss_diag": 0.000156402587890625, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.400184154510498, "learning_rate": 1e-05, "loss": 0.456, "mean_token_accuracy": 0.8735061287879944, "num_tokens": 312450574.0, "step": 8232 }, { "epoch": 1.0473222236356698, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.000156402587890625, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 7.009702205657959, "learning_rate": 1e-05, "loss": 0.5422, "mean_token_accuracy": 0.8561574816703796, "num_tokens": 312483054.0, "step": 8233 }, { "epoch": 1.0474494339142604, "ewc_loss": 0.10107421875, "ewc_loss_diag": 0.000156402587890625, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.666435718536377, "learning_rate": 1e-05, "loss": 0.4862, "mean_token_accuracy": 0.8667758703231812, "num_tokens": 312527825.0, "step": 8234 }, { "epoch": 1.0475766441928507, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.000156402587890625, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 7.021698474884033, "learning_rate": 1e-05, "loss": 0.5386, "mean_token_accuracy": 0.8602737188339233, "num_tokens": 312566971.0, "step": 8235 }, { "epoch": 1.0477038544714412, "ewc_loss": 0.10791015625, "ewc_loss_diag": 0.000156402587890625, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 5.971286296844482, "learning_rate": 1e-05, "loss": 0.4967, "mean_token_accuracy": 0.8669780492782593, "num_tokens": 312603527.0, "step": 8236 }, { "epoch": 1.0478310647500317, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.000156402587890625, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.7571892738342285, "learning_rate": 1e-05, "loss": 0.5124, "mean_token_accuracy": 0.8666903972625732, "num_tokens": 312634281.0, "step": 8237 }, { "epoch": 1.0479582750286223, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.000156402587890625, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.047588348388672, "learning_rate": 1e-05, "loss": 0.5209, "mean_token_accuracy": 0.8622078895568848, "num_tokens": 312668036.0, "step": 8238 }, { "epoch": 1.0480854853072128, "ewc_loss": 0.11181640625, "ewc_loss_diag": 0.000156402587890625, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.435856819152832, "learning_rate": 1e-05, "loss": 0.522, "mean_token_accuracy": 0.8574224710464478, "num_tokens": 312707422.0, "step": 8239 }, { "epoch": 1.0482126955858033, "ewc_loss": 0.10400390625, "ewc_loss_diag": 0.000156402587890625, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 5.981228828430176, "learning_rate": 1e-05, "loss": 0.4991, "mean_token_accuracy": 0.8629592657089233, "num_tokens": 312740715.0, "step": 8240 }, { "epoch": 1.0483399058643939, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.000156402587890625, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 6.184230804443359, "learning_rate": 1e-05, "loss": 0.4852, "mean_token_accuracy": 0.8691146373748779, "num_tokens": 312773196.0, "step": 8241 }, { "epoch": 1.0484671161429844, "ewc_loss": 0.10107421875, "ewc_loss_diag": 0.000156402587890625, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.882753372192383, "learning_rate": 1e-05, "loss": 0.5041, "mean_token_accuracy": 0.8593919277191162, "num_tokens": 312817502.0, "step": 8242 }, { "epoch": 1.048594326421575, "ewc_loss": 0.10205078125, "ewc_loss_diag": 0.000156402587890625, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 6.049625396728516, "learning_rate": 1e-05, "loss": 0.5011, "mean_token_accuracy": 0.8626546859741211, "num_tokens": 312852810.0, "step": 8243 }, { "epoch": 1.0487215367001654, "ewc_loss": 0.09814453125, "ewc_loss_diag": 0.000156402587890625, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 6.036145210266113, "learning_rate": 1e-05, "loss": 0.4509, "mean_token_accuracy": 0.8754532933235168, "num_tokens": 312889069.0, "step": 8244 }, { "epoch": 1.048848746978756, "ewc_loss": 0.09912109375, "ewc_loss_diag": 0.000156402587890625, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.833003997802734, "learning_rate": 1e-05, "loss": 0.5051, "mean_token_accuracy": 0.8644915819168091, "num_tokens": 312929032.0, "step": 8245 }, { "epoch": 1.0489759572573465, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.00015735626220703125, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.810958385467529, "learning_rate": 1e-05, "loss": 0.4751, "mean_token_accuracy": 0.8684282302856445, "num_tokens": 312963792.0, "step": 8246 }, { "epoch": 1.0491031675359368, "ewc_loss": 0.0986328125, "ewc_loss_diag": 0.00015735626220703125, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.762305736541748, "learning_rate": 1e-05, "loss": 0.436, "mean_token_accuracy": 0.8849332332611084, "num_tokens": 312999532.0, "step": 8247 }, { "epoch": 1.0492303778145273, "ewc_loss": 0.09814453125, "ewc_loss_diag": 0.00015735626220703125, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.728590965270996, "learning_rate": 1e-05, "loss": 0.5069, "mean_token_accuracy": 0.8659585118293762, "num_tokens": 313040807.0, "step": 8248 }, { "epoch": 1.0493575880931179, "ewc_loss": 0.099609375, "ewc_loss_diag": 0.00015735626220703125, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.795912742614746, "learning_rate": 1e-05, "loss": 0.4524, "mean_token_accuracy": 0.8759692311286926, "num_tokens": 313074895.0, "step": 8249 }, { "epoch": 1.0494847983717084, "ewc_loss": 0.0966796875, "ewc_loss_diag": 0.00015735626220703125, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.62626314163208, "learning_rate": 1e-05, "loss": 0.4776, "mean_token_accuracy": 0.870769739151001, "num_tokens": 313115299.0, "step": 8250 }, { "epoch": 1.049612008650299, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.00015735626220703125, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.865950107574463, "learning_rate": 1e-05, "loss": 0.4963, "mean_token_accuracy": 0.8652581572532654, "num_tokens": 313157578.0, "step": 8251 }, { "epoch": 1.0497392189288894, "ewc_loss": 0.095703125, "ewc_loss_diag": 0.00015735626220703125, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.520753383636475, "learning_rate": 1e-05, "loss": 0.4937, "mean_token_accuracy": 0.8669699430465698, "num_tokens": 313202219.0, "step": 8252 }, { "epoch": 1.04986642920748, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.00015735626220703125, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 5.997211456298828, "learning_rate": 1e-05, "loss": 0.4796, "mean_token_accuracy": 0.8724257946014404, "num_tokens": 313242345.0, "step": 8253 }, { "epoch": 1.0499936394860705, "ewc_loss": 0.0927734375, "ewc_loss_diag": 0.00015735626220703125, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.364160060882568, "learning_rate": 1e-05, "loss": 0.4684, "mean_token_accuracy": 0.8722358345985413, "num_tokens": 313277052.0, "step": 8254 }, { "epoch": 1.050120849764661, "ewc_loss": 0.10791015625, "ewc_loss_diag": 0.00015735626220703125, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.267676830291748, "learning_rate": 1e-05, "loss": 0.5083, "mean_token_accuracy": 0.8631773591041565, "num_tokens": 313311291.0, "step": 8255 }, { "epoch": 1.0502480600432516, "ewc_loss": 0.08935546875, "ewc_loss_diag": 0.00015735626220703125, "ewc_loss_parallel": 7.390975952148438e-05, "grad_norm": 5.171547889709473, "learning_rate": 1e-05, "loss": 0.4701, "mean_token_accuracy": 0.8698240518569946, "num_tokens": 313349141.0, "step": 8256 }, { "epoch": 1.050375270321842, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.00015735626220703125, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.743273735046387, "learning_rate": 1e-05, "loss": 0.553, "mean_token_accuracy": 0.8553158044815063, "num_tokens": 313387575.0, "step": 8257 }, { "epoch": 1.0505024806004326, "ewc_loss": 0.09228515625, "ewc_loss_diag": 0.00015735626220703125, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.115216255187988, "learning_rate": 1e-05, "loss": 0.496, "mean_token_accuracy": 0.8598416447639465, "num_tokens": 313424087.0, "step": 8258 }, { "epoch": 1.0506296908790231, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.00015735626220703125, "ewc_loss_parallel": 0.00012111663818359375, "grad_norm": 7.3681793212890625, "learning_rate": 1e-05, "loss": 0.5199, "mean_token_accuracy": 0.8717056512832642, "num_tokens": 313464517.0, "step": 8259 }, { "epoch": 1.0507569011576134, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.00015735626220703125, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 5.568207740783691, "learning_rate": 1e-05, "loss": 0.5235, "mean_token_accuracy": 0.8567776679992676, "num_tokens": 313504146.0, "step": 8260 }, { "epoch": 1.050884111436204, "ewc_loss": 0.1455078125, "ewc_loss_diag": 0.00015735626220703125, "ewc_loss_parallel": 0.00012969970703125, "grad_norm": 7.677236557006836, "learning_rate": 1e-05, "loss": 0.5454, "mean_token_accuracy": 0.8631455302238464, "num_tokens": 313541842.0, "step": 8261 }, { "epoch": 1.0510113217147945, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.00015735626220703125, "ewc_loss_parallel": 0.00011014938354492188, "grad_norm": 6.379550933837891, "learning_rate": 1e-05, "loss": 0.5327, "mean_token_accuracy": 0.8617933392524719, "num_tokens": 313580695.0, "step": 8262 }, { "epoch": 1.051138531993385, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.00015735626220703125, "ewc_loss_parallel": 0.0001163482666015625, "grad_norm": 7.228351593017578, "learning_rate": 1e-05, "loss": 0.56, "mean_token_accuracy": 0.855941116809845, "num_tokens": 313619494.0, "step": 8263 }, { "epoch": 1.0512657422719756, "ewc_loss": 0.125, "ewc_loss_diag": 0.00015735626220703125, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 6.52892541885376, "learning_rate": 1e-05, "loss": 0.4852, "mean_token_accuracy": 0.8744572401046753, "num_tokens": 313653234.0, "step": 8264 }, { "epoch": 1.051392952550566, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.00015735626220703125, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.753581523895264, "learning_rate": 1e-05, "loss": 0.5288, "mean_token_accuracy": 0.8654955625534058, "num_tokens": 313694395.0, "step": 8265 }, { "epoch": 1.0515201628291566, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.00015735626220703125, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.301363468170166, "learning_rate": 1e-05, "loss": 0.5489, "mean_token_accuracy": 0.852745532989502, "num_tokens": 313733370.0, "step": 8266 }, { "epoch": 1.0516473731077471, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.00015735626220703125, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.441770076751709, "learning_rate": 1e-05, "loss": 0.5498, "mean_token_accuracy": 0.8530546426773071, "num_tokens": 313770611.0, "step": 8267 }, { "epoch": 1.0517745833863377, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.073163986206055, "learning_rate": 1e-05, "loss": 0.4265, "mean_token_accuracy": 0.8879973888397217, "num_tokens": 313807262.0, "step": 8268 }, { "epoch": 1.0519017936649282, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.00015735626220703125, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.2585577964782715, "learning_rate": 1e-05, "loss": 0.5123, "mean_token_accuracy": 0.860878586769104, "num_tokens": 313844294.0, "step": 8269 }, { "epoch": 1.0520290039435187, "ewc_loss": 0.10009765625, "ewc_loss_diag": 0.00015735626220703125, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.883034706115723, "learning_rate": 1e-05, "loss": 0.4483, "mean_token_accuracy": 0.8784207701683044, "num_tokens": 313878701.0, "step": 8270 }, { "epoch": 1.0521562142221093, "ewc_loss": 0.10595703125, "ewc_loss_diag": 0.00015735626220703125, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.1869683265686035, "learning_rate": 1e-05, "loss": 0.4564, "mean_token_accuracy": 0.8769704103469849, "num_tokens": 313917780.0, "step": 8271 }, { "epoch": 1.0522834245006996, "ewc_loss": 0.095703125, "ewc_loss_diag": 0.00015735626220703125, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.6959452629089355, "learning_rate": 1e-05, "loss": 0.4493, "mean_token_accuracy": 0.8796902894973755, "num_tokens": 313948611.0, "step": 8272 }, { "epoch": 1.05241063477929, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.24662971496582, "learning_rate": 1e-05, "loss": 0.4783, "mean_token_accuracy": 0.8732162714004517, "num_tokens": 313981999.0, "step": 8273 }, { "epoch": 1.0525378450578806, "ewc_loss": 0.09375, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.523419380187988, "learning_rate": 1e-05, "loss": 0.4574, "mean_token_accuracy": 0.8738627433776855, "num_tokens": 314021199.0, "step": 8274 }, { "epoch": 1.0526650553364711, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.43386697769165, "learning_rate": 1e-05, "loss": 0.4555, "mean_token_accuracy": 0.8784185647964478, "num_tokens": 314058187.0, "step": 8275 }, { "epoch": 1.0527922656150617, "ewc_loss": 0.0927734375, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.375177383422852, "learning_rate": 1e-05, "loss": 0.4693, "mean_token_accuracy": 0.8684453368186951, "num_tokens": 314095612.0, "step": 8276 }, { "epoch": 1.0529194758936522, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.795782566070557, "learning_rate": 1e-05, "loss": 0.5142, "mean_token_accuracy": 0.8610937595367432, "num_tokens": 314124895.0, "step": 8277 }, { "epoch": 1.0530466861722427, "ewc_loss": 0.0947265625, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.4161696434021, "learning_rate": 1e-05, "loss": 0.4614, "mean_token_accuracy": 0.8773813247680664, "num_tokens": 314167787.0, "step": 8278 }, { "epoch": 1.0531738964508333, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 0.00011014938354492188, "grad_norm": 7.139703750610352, "learning_rate": 1e-05, "loss": 0.5121, "mean_token_accuracy": 0.8691458106040955, "num_tokens": 314203389.0, "step": 8279 }, { "epoch": 1.0533011067294238, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.799405097961426, "learning_rate": 1e-05, "loss": 0.4573, "mean_token_accuracy": 0.8778457641601562, "num_tokens": 314245161.0, "step": 8280 }, { "epoch": 1.0534283170080143, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 7.0882487297058105, "learning_rate": 1e-05, "loss": 0.5798, "mean_token_accuracy": 0.844343364238739, "num_tokens": 314288097.0, "step": 8281 }, { "epoch": 1.0535555272866048, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.19196891784668, "learning_rate": 1e-05, "loss": 0.5126, "mean_token_accuracy": 0.865341305732727, "num_tokens": 314324066.0, "step": 8282 }, { "epoch": 1.0536827375651954, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.61322021484375, "learning_rate": 1e-05, "loss": 0.4594, "mean_token_accuracy": 0.8765067458152771, "num_tokens": 314358430.0, "step": 8283 }, { "epoch": 1.0538099478437857, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.2716965675354, "learning_rate": 1e-05, "loss": 0.5043, "mean_token_accuracy": 0.8640346527099609, "num_tokens": 314394199.0, "step": 8284 }, { "epoch": 1.0539371581223762, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 6.082484722137451, "learning_rate": 1e-05, "loss": 0.4826, "mean_token_accuracy": 0.86955726146698, "num_tokens": 314432851.0, "step": 8285 }, { "epoch": 1.0540643684009667, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.300481796264648, "learning_rate": 1e-05, "loss": 0.5043, "mean_token_accuracy": 0.8669883608818054, "num_tokens": 314473843.0, "step": 8286 }, { "epoch": 1.0541915786795573, "ewc_loss": 0.0947265625, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.637258529663086, "learning_rate": 1e-05, "loss": 0.4385, "mean_token_accuracy": 0.8821598887443542, "num_tokens": 314518782.0, "step": 8287 }, { "epoch": 1.0543187889581478, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.4667439460754395, "learning_rate": 1e-05, "loss": 0.5138, "mean_token_accuracy": 0.8653668761253357, "num_tokens": 314561140.0, "step": 8288 }, { "epoch": 1.0544459992367383, "ewc_loss": 0.0908203125, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 7.534027099609375e-05, "grad_norm": 5.392478942871094, "learning_rate": 1e-05, "loss": 0.4711, "mean_token_accuracy": 0.866747260093689, "num_tokens": 314602605.0, "step": 8289 }, { "epoch": 1.0545732095153288, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.777543067932129, "learning_rate": 1e-05, "loss": 0.5503, "mean_token_accuracy": 0.8522613644599915, "num_tokens": 314647719.0, "step": 8290 }, { "epoch": 1.0547004197939194, "ewc_loss": 0.0927734375, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.392680644989014, "learning_rate": 1e-05, "loss": 0.5286, "mean_token_accuracy": 0.8544912338256836, "num_tokens": 314690016.0, "step": 8291 }, { "epoch": 1.05482763007251, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 7.092325687408447, "learning_rate": 1e-05, "loss": 0.5129, "mean_token_accuracy": 0.8697988986968994, "num_tokens": 314730432.0, "step": 8292 }, { "epoch": 1.0549548403511004, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.7049736976623535, "learning_rate": 1e-05, "loss": 0.4574, "mean_token_accuracy": 0.8757163882255554, "num_tokens": 314764957.0, "step": 8293 }, { "epoch": 1.055082050629691, "ewc_loss": 0.125, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 7.113137722015381, "learning_rate": 1e-05, "loss": 0.5018, "mean_token_accuracy": 0.870978593826294, "num_tokens": 314803787.0, "step": 8294 }, { "epoch": 1.0552092609082815, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.049887657165527, "learning_rate": 1e-05, "loss": 0.4779, "mean_token_accuracy": 0.8735513091087341, "num_tokens": 314841732.0, "step": 8295 }, { "epoch": 1.0553364711868718, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.822537422180176, "learning_rate": 1e-05, "loss": 0.4808, "mean_token_accuracy": 0.8722987174987793, "num_tokens": 314878212.0, "step": 8296 }, { "epoch": 1.0554636814654623, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.136342525482178, "learning_rate": 1e-05, "loss": 0.5121, "mean_token_accuracy": 0.862076997756958, "num_tokens": 314917130.0, "step": 8297 }, { "epoch": 1.0555908917440529, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.473294258117676, "learning_rate": 1e-05, "loss": 0.4787, "mean_token_accuracy": 0.8751835227012634, "num_tokens": 314961804.0, "step": 8298 }, { "epoch": 1.0557181020226434, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 6.044241428375244, "learning_rate": 1e-05, "loss": 0.4341, "mean_token_accuracy": 0.8846001625061035, "num_tokens": 314997157.0, "step": 8299 }, { "epoch": 1.055845312301234, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 6.229404449462891, "learning_rate": 1e-05, "loss": 0.5408, "mean_token_accuracy": 0.8500968217849731, "num_tokens": 315038370.0, "step": 8300 }, { "epoch": 1.0559725225798244, "ewc_loss": 0.099609375, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.912917137145996, "learning_rate": 1e-05, "loss": 0.4439, "mean_token_accuracy": 0.8812370300292969, "num_tokens": 315081716.0, "step": 8301 }, { "epoch": 1.056099732858415, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 6.100639820098877, "learning_rate": 1e-05, "loss": 0.5545, "mean_token_accuracy": 0.8476218581199646, "num_tokens": 315124528.0, "step": 8302 }, { "epoch": 1.0562269431370055, "ewc_loss": 0.09765625, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.765942096710205, "learning_rate": 1e-05, "loss": 0.4343, "mean_token_accuracy": 0.881804883480072, "num_tokens": 315163893.0, "step": 8303 }, { "epoch": 1.056354153415596, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 6.1043620109558105, "learning_rate": 1e-05, "loss": 0.5621, "mean_token_accuracy": 0.8456608057022095, "num_tokens": 315199462.0, "step": 8304 }, { "epoch": 1.0564813636941865, "ewc_loss": 0.09375, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.598897457122803, "learning_rate": 1e-05, "loss": 0.4805, "mean_token_accuracy": 0.8671164512634277, "num_tokens": 315245075.0, "step": 8305 }, { "epoch": 1.056608573972777, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.225349426269531, "learning_rate": 1e-05, "loss": 0.4796, "mean_token_accuracy": 0.8714269399642944, "num_tokens": 315289291.0, "step": 8306 }, { "epoch": 1.0567357842513676, "ewc_loss": 0.091796875, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 7.581710815429688e-05, "grad_norm": 5.441902160644531, "learning_rate": 1e-05, "loss": 0.4579, "mean_token_accuracy": 0.8753722310066223, "num_tokens": 315327750.0, "step": 8307 }, { "epoch": 1.0568629945299581, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.466801166534424, "learning_rate": 1e-05, "loss": 0.5009, "mean_token_accuracy": 0.8637340068817139, "num_tokens": 315366863.0, "step": 8308 }, { "epoch": 1.0569902048085484, "ewc_loss": 0.091796875, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 7.581710815429688e-05, "grad_norm": 5.356520652770996, "learning_rate": 1e-05, "loss": 0.4332, "mean_token_accuracy": 0.8816078901290894, "num_tokens": 315406935.0, "step": 8309 }, { "epoch": 1.057117415087139, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.7936601638793945, "learning_rate": 1e-05, "loss": 0.5175, "mean_token_accuracy": 0.8614985942840576, "num_tokens": 315446483.0, "step": 8310 }, { "epoch": 1.0572446253657295, "ewc_loss": 0.09375, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.477072238922119, "learning_rate": 1e-05, "loss": 0.493, "mean_token_accuracy": 0.8624976873397827, "num_tokens": 315482862.0, "step": 8311 }, { "epoch": 1.05737183564432, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.992308616638184, "learning_rate": 1e-05, "loss": 0.4935, "mean_token_accuracy": 0.8713631629943848, "num_tokens": 315522649.0, "step": 8312 }, { "epoch": 1.0574990459229106, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.795261383056641, "learning_rate": 1e-05, "loss": 0.5122, "mean_token_accuracy": 0.8613510727882385, "num_tokens": 315562642.0, "step": 8313 }, { "epoch": 1.057626256201501, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.842587947845459, "learning_rate": 1e-05, "loss": 0.4311, "mean_token_accuracy": 0.892581582069397, "num_tokens": 315594056.0, "step": 8314 }, { "epoch": 1.0577534664800916, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.066532135009766, "learning_rate": 1e-05, "loss": 0.5148, "mean_token_accuracy": 0.8631430864334106, "num_tokens": 315632474.0, "step": 8315 }, { "epoch": 1.0578806767586821, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.408524036407471, "learning_rate": 1e-05, "loss": 0.4882, "mean_token_accuracy": 0.8686374425888062, "num_tokens": 315662036.0, "step": 8316 }, { "epoch": 1.0580078870372727, "ewc_loss": 0.109375, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.156798839569092, "learning_rate": 1e-05, "loss": 0.4837, "mean_token_accuracy": 0.8713581562042236, "num_tokens": 315701974.0, "step": 8317 }, { "epoch": 1.0581350973158632, "ewc_loss": 0.0986328125, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.906185626983643, "learning_rate": 1e-05, "loss": 0.4882, "mean_token_accuracy": 0.8649073839187622, "num_tokens": 315745876.0, "step": 8318 }, { "epoch": 1.0582623075944537, "ewc_loss": 0.109375, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.259649276733398, "learning_rate": 1e-05, "loss": 0.4904, "mean_token_accuracy": 0.8668036460876465, "num_tokens": 315787386.0, "step": 8319 }, { "epoch": 1.058389517873044, "ewc_loss": 0.0927734375, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.514461994171143, "learning_rate": 1e-05, "loss": 0.4419, "mean_token_accuracy": 0.8765109777450562, "num_tokens": 315824047.0, "step": 8320 }, { "epoch": 1.0585167281516346, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.525747299194336, "learning_rate": 1e-05, "loss": 0.5445, "mean_token_accuracy": 0.8498502969741821, "num_tokens": 315863890.0, "step": 8321 }, { "epoch": 1.058643938430225, "ewc_loss": 0.0908203125, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 7.534027099609375e-05, "grad_norm": 5.310317039489746, "learning_rate": 1e-05, "loss": 0.501, "mean_token_accuracy": 0.8614721298217773, "num_tokens": 315902855.0, "step": 8322 }, { "epoch": 1.0587711487088156, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.90483283996582, "learning_rate": 1e-05, "loss": 0.5066, "mean_token_accuracy": 0.8695827722549438, "num_tokens": 315942112.0, "step": 8323 }, { "epoch": 1.0588983589874061, "ewc_loss": 0.0947265625, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.441870212554932, "learning_rate": 1e-05, "loss": 0.4723, "mean_token_accuracy": 0.8710724115371704, "num_tokens": 315981256.0, "step": 8324 }, { "epoch": 1.0590255692659967, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 0.00011110305786132812, "grad_norm": 7.170365810394287, "learning_rate": 1e-05, "loss": 0.5279, "mean_token_accuracy": 0.8620600700378418, "num_tokens": 316021192.0, "step": 8325 }, { "epoch": 1.0591527795445872, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 5.855526924133301, "learning_rate": 1e-05, "loss": 0.5008, "mean_token_accuracy": 0.8641607165336609, "num_tokens": 316056261.0, "step": 8326 }, { "epoch": 1.0592799898231777, "ewc_loss": 0.125, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 7.032623291015625, "learning_rate": 1e-05, "loss": 0.4544, "mean_token_accuracy": 0.8861201405525208, "num_tokens": 316093390.0, "step": 8327 }, { "epoch": 1.0594072001017683, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.117218971252441, "learning_rate": 1e-05, "loss": 0.5356, "mean_token_accuracy": 0.8546575903892517, "num_tokens": 316133771.0, "step": 8328 }, { "epoch": 1.0595344103803588, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.673635959625244, "learning_rate": 1e-05, "loss": 0.5042, "mean_token_accuracy": 0.8671772480010986, "num_tokens": 316172123.0, "step": 8329 }, { "epoch": 1.0596616206589493, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.114284038543701, "learning_rate": 1e-05, "loss": 0.4519, "mean_token_accuracy": 0.8805885910987854, "num_tokens": 316211866.0, "step": 8330 }, { "epoch": 1.0597888309375398, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.351110458374023, "learning_rate": 1e-05, "loss": 0.4853, "mean_token_accuracy": 0.8729071617126465, "num_tokens": 316248838.0, "step": 8331 }, { "epoch": 1.0599160412161304, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 6.012763023376465, "learning_rate": 1e-05, "loss": 0.479, "mean_token_accuracy": 0.8713786602020264, "num_tokens": 316286551.0, "step": 8332 }, { "epoch": 1.0600432514947207, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 6.13800048828125, "learning_rate": 1e-05, "loss": 0.5128, "mean_token_accuracy": 0.8661276698112488, "num_tokens": 316321278.0, "step": 8333 }, { "epoch": 1.0601704617733112, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.86375093460083, "learning_rate": 1e-05, "loss": 0.4812, "mean_token_accuracy": 0.8672414422035217, "num_tokens": 316356282.0, "step": 8334 }, { "epoch": 1.0602976720519017, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 6.07523250579834, "learning_rate": 1e-05, "loss": 0.4965, "mean_token_accuracy": 0.8645858764648438, "num_tokens": 316398858.0, "step": 8335 }, { "epoch": 1.0604248823304923, "ewc_loss": 0.095703125, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.696902275085449, "learning_rate": 1e-05, "loss": 0.4828, "mean_token_accuracy": 0.8660659790039062, "num_tokens": 316432883.0, "step": 8336 }, { "epoch": 1.0605520926090828, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 6.121720790863037, "learning_rate": 1e-05, "loss": 0.452, "mean_token_accuracy": 0.8829166293144226, "num_tokens": 316469305.0, "step": 8337 }, { "epoch": 1.0606793028876733, "ewc_loss": 0.09375, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.525717735290527, "learning_rate": 1e-05, "loss": 0.4239, "mean_token_accuracy": 0.8857458829879761, "num_tokens": 316509072.0, "step": 8338 }, { "epoch": 1.0608065131662638, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.276767730712891, "learning_rate": 1e-05, "loss": 0.4942, "mean_token_accuracy": 0.863746166229248, "num_tokens": 316549391.0, "step": 8339 }, { "epoch": 1.0609337234448544, "ewc_loss": 0.091796875, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 7.62939453125e-05, "grad_norm": 5.389784812927246, "learning_rate": 1e-05, "loss": 0.4229, "mean_token_accuracy": 0.88502037525177, "num_tokens": 316583696.0, "step": 8340 }, { "epoch": 1.061060933723445, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.558541774749756, "learning_rate": 1e-05, "loss": 0.5139, "mean_token_accuracy": 0.8607387542724609, "num_tokens": 316618421.0, "step": 8341 }, { "epoch": 1.0611881440020354, "ewc_loss": 0.09375, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.334010601043701, "learning_rate": 1e-05, "loss": 0.4773, "mean_token_accuracy": 0.8675429821014404, "num_tokens": 316660284.0, "step": 8342 }, { "epoch": 1.061315354280626, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.908262729644775, "learning_rate": 1e-05, "loss": 0.5673, "mean_token_accuracy": 0.8530329465866089, "num_tokens": 316699753.0, "step": 8343 }, { "epoch": 1.0614425645592165, "ewc_loss": 0.0986328125, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.5355730056762695, "learning_rate": 1e-05, "loss": 0.4883, "mean_token_accuracy": 0.8662657737731934, "num_tokens": 316735336.0, "step": 8344 }, { "epoch": 1.0615697748378068, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 7.0545783042907715, "learning_rate": 1e-05, "loss": 0.4955, "mean_token_accuracy": 0.8725932836532593, "num_tokens": 316775327.0, "step": 8345 }, { "epoch": 1.0616969851163973, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 5.913605690002441, "learning_rate": 1e-05, "loss": 0.5351, "mean_token_accuracy": 0.851578950881958, "num_tokens": 316817027.0, "step": 8346 }, { "epoch": 1.0618241953949878, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.806110858917236, "learning_rate": 1e-05, "loss": 0.5047, "mean_token_accuracy": 0.869051456451416, "num_tokens": 316850980.0, "step": 8347 }, { "epoch": 1.0619514056735784, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.1229095458984375, "learning_rate": 1e-05, "loss": 0.4467, "mean_token_accuracy": 0.8858850002288818, "num_tokens": 316886533.0, "step": 8348 }, { "epoch": 1.062078615952169, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.330935955047607, "learning_rate": 1e-05, "loss": 0.4665, "mean_token_accuracy": 0.8772086501121521, "num_tokens": 316927447.0, "step": 8349 }, { "epoch": 1.0622058262307594, "ewc_loss": 0.109375, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.18895149230957, "learning_rate": 1e-05, "loss": 0.5263, "mean_token_accuracy": 0.8587021827697754, "num_tokens": 316965983.0, "step": 8350 }, { "epoch": 1.06233303650935, "ewc_loss": 0.0986328125, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.8547444343566895, "learning_rate": 1e-05, "loss": 0.4712, "mean_token_accuracy": 0.8694555759429932, "num_tokens": 316999678.0, "step": 8351 }, { "epoch": 1.0624602467879405, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.301032066345215, "learning_rate": 1e-05, "loss": 0.5241, "mean_token_accuracy": 0.8583718538284302, "num_tokens": 317038650.0, "step": 8352 }, { "epoch": 1.062587457066531, "ewc_loss": 0.091796875, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 7.62939453125e-05, "grad_norm": 5.49415922164917, "learning_rate": 1e-05, "loss": 0.4635, "mean_token_accuracy": 0.8708304166793823, "num_tokens": 317075507.0, "step": 8353 }, { "epoch": 1.0627146673451215, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.539854049682617, "learning_rate": 1e-05, "loss": 0.4522, "mean_token_accuracy": 0.8807146549224854, "num_tokens": 317109364.0, "step": 8354 }, { "epoch": 1.062841877623712, "ewc_loss": 0.0927734375, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.335127830505371, "learning_rate": 1e-05, "loss": 0.4628, "mean_token_accuracy": 0.87183678150177, "num_tokens": 317149959.0, "step": 8355 }, { "epoch": 1.0629690879023026, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.888291358947754, "learning_rate": 1e-05, "loss": 0.5074, "mean_token_accuracy": 0.8674653768539429, "num_tokens": 317182604.0, "step": 8356 }, { "epoch": 1.0630962981808931, "ewc_loss": 0.0966796875, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.471732139587402, "learning_rate": 1e-05, "loss": 0.4549, "mean_token_accuracy": 0.8774136900901794, "num_tokens": 317224190.0, "step": 8357 }, { "epoch": 1.0632235084594834, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 7.1206135749816895, "learning_rate": 1e-05, "loss": 0.5319, "mean_token_accuracy": 0.8643320202827454, "num_tokens": 317261607.0, "step": 8358 }, { "epoch": 1.063350718738074, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 5.847583293914795, "learning_rate": 1e-05, "loss": 0.5308, "mean_token_accuracy": 0.8556900024414062, "num_tokens": 317300772.0, "step": 8359 }, { "epoch": 1.0634779290166645, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 6.989911079406738, "learning_rate": 1e-05, "loss": 0.512, "mean_token_accuracy": 0.866982102394104, "num_tokens": 317336981.0, "step": 8360 }, { "epoch": 1.063605139295255, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.076840400695801, "learning_rate": 1e-05, "loss": 0.531, "mean_token_accuracy": 0.8571151494979858, "num_tokens": 317376522.0, "step": 8361 }, { "epoch": 1.0637323495738455, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.6631574630737305, "learning_rate": 1e-05, "loss": 0.4989, "mean_token_accuracy": 0.8696243762969971, "num_tokens": 317409924.0, "step": 8362 }, { "epoch": 1.063859559852436, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.00016021728515625, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.081763744354248, "learning_rate": 1e-05, "loss": 0.455, "mean_token_accuracy": 0.8811713457107544, "num_tokens": 317440762.0, "step": 8363 }, { "epoch": 1.0639867701310266, "ewc_loss": 0.109375, "ewc_loss_diag": 0.00016021728515625, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.341676235198975, "learning_rate": 1e-05, "loss": 0.4987, "mean_token_accuracy": 0.8698683381080627, "num_tokens": 317471606.0, "step": 8364 }, { "epoch": 1.0641139804096171, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.00016021728515625, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 5.99330472946167, "learning_rate": 1e-05, "loss": 0.5266, "mean_token_accuracy": 0.855605959892273, "num_tokens": 317508383.0, "step": 8365 }, { "epoch": 1.0642411906882077, "ewc_loss": 0.10302734375, "ewc_loss_diag": 0.00016021728515625, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 6.111076831817627, "learning_rate": 1e-05, "loss": 0.4339, "mean_token_accuracy": 0.8882706165313721, "num_tokens": 317546690.0, "step": 8366 }, { "epoch": 1.0643684009667982, "ewc_loss": 0.10107421875, "ewc_loss_diag": 0.00016021728515625, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.872468948364258, "learning_rate": 1e-05, "loss": 0.4554, "mean_token_accuracy": 0.8808724880218506, "num_tokens": 317584926.0, "step": 8367 }, { "epoch": 1.0644956112453887, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.00016021728515625, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 6.032578468322754, "learning_rate": 1e-05, "loss": 0.5147, "mean_token_accuracy": 0.8572050929069519, "num_tokens": 317619941.0, "step": 8368 }, { "epoch": 1.064622821523979, "ewc_loss": 0.09814453125, "ewc_loss_diag": 0.00016021728515625, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.734267234802246, "learning_rate": 1e-05, "loss": 0.4494, "mean_token_accuracy": 0.8796825408935547, "num_tokens": 317651006.0, "step": 8369 }, { "epoch": 1.0647500318025696, "ewc_loss": 0.10302734375, "ewc_loss_diag": 0.00016021728515625, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 6.05258321762085, "learning_rate": 1e-05, "loss": 0.4832, "mean_token_accuracy": 0.8701053261756897, "num_tokens": 317688051.0, "step": 8370 }, { "epoch": 1.06487724208116, "ewc_loss": 0.09521484375, "ewc_loss_diag": 0.00016021728515625, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.567240238189697, "learning_rate": 1e-05, "loss": 0.4546, "mean_token_accuracy": 0.8752680420875549, "num_tokens": 317727447.0, "step": 8371 }, { "epoch": 1.0650044523597506, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.00016021728515625, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.174630641937256, "learning_rate": 1e-05, "loss": 0.4638, "mean_token_accuracy": 0.8783872127532959, "num_tokens": 317771495.0, "step": 8372 }, { "epoch": 1.0651316626383411, "ewc_loss": 0.091796875, "ewc_loss_diag": 0.00016021728515625, "ewc_loss_parallel": 7.581710815429688e-05, "grad_norm": 5.413194179534912, "learning_rate": 1e-05, "loss": 0.4394, "mean_token_accuracy": 0.8796385526657104, "num_tokens": 317808166.0, "step": 8373 }, { "epoch": 1.0652588729169317, "ewc_loss": 0.11083984375, "ewc_loss_diag": 0.00016021728515625, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.427229404449463, "learning_rate": 1e-05, "loss": 0.4823, "mean_token_accuracy": 0.874094545841217, "num_tokens": 317850928.0, "step": 8374 }, { "epoch": 1.0653860831955222, "ewc_loss": 0.09130859375, "ewc_loss_diag": 0.00016021728515625, "ewc_loss_parallel": 7.534027099609375e-05, "grad_norm": 5.302743434906006, "learning_rate": 1e-05, "loss": 0.4528, "mean_token_accuracy": 0.8753489255905151, "num_tokens": 317893156.0, "step": 8375 }, { "epoch": 1.0655132934741127, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.00016021728515625, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.93011474609375, "learning_rate": 1e-05, "loss": 0.5811, "mean_token_accuracy": 0.8446010947227478, "num_tokens": 317937146.0, "step": 8376 }, { "epoch": 1.0656405037527032, "ewc_loss": 0.09521484375, "ewc_loss_diag": 0.00016021728515625, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.4114484786987305, "learning_rate": 1e-05, "loss": 0.4458, "mean_token_accuracy": 0.8781173229217529, "num_tokens": 317977429.0, "step": 8377 }, { "epoch": 1.0657677140312938, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.00016021728515625, "ewc_loss_parallel": 0.00011301040649414062, "grad_norm": 7.173956871032715, "learning_rate": 1e-05, "loss": 0.5255, "mean_token_accuracy": 0.8646103739738464, "num_tokens": 318013445.0, "step": 8378 }, { "epoch": 1.0658949243098843, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.00016117095947265625, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 5.83610200881958, "learning_rate": 1e-05, "loss": 0.5197, "mean_token_accuracy": 0.8598387241363525, "num_tokens": 318054082.0, "step": 8379 }, { "epoch": 1.0660221345884748, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.00016021728515625, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 7.045963287353516, "learning_rate": 1e-05, "loss": 0.5532, "mean_token_accuracy": 0.8550716638565063, "num_tokens": 318096125.0, "step": 8380 }, { "epoch": 1.0661493448670654, "ewc_loss": 0.11376953125, "ewc_loss_diag": 0.00016021728515625, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.170240879058838, "learning_rate": 1e-05, "loss": 0.4898, "mean_token_accuracy": 0.8704957365989685, "num_tokens": 318130859.0, "step": 8381 }, { "epoch": 1.0662765551456557, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.00016117095947265625, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.574952125549316, "learning_rate": 1e-05, "loss": 0.4768, "mean_token_accuracy": 0.8747975826263428, "num_tokens": 318171616.0, "step": 8382 }, { "epoch": 1.0664037654242462, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.00016117095947265625, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.2313103675842285, "learning_rate": 1e-05, "loss": 0.5111, "mean_token_accuracy": 0.8636278510093689, "num_tokens": 318206741.0, "step": 8383 }, { "epoch": 1.0665309757028367, "ewc_loss": 0.10400390625, "ewc_loss_diag": 0.00016117095947265625, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 6.074173450469971, "learning_rate": 1e-05, "loss": 0.5065, "mean_token_accuracy": 0.8634072542190552, "num_tokens": 318252513.0, "step": 8384 }, { "epoch": 1.0666581859814273, "ewc_loss": 0.11083984375, "ewc_loss_diag": 0.00016117095947265625, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.264235973358154, "learning_rate": 1e-05, "loss": 0.5124, "mean_token_accuracy": 0.8582817912101746, "num_tokens": 318286632.0, "step": 8385 }, { "epoch": 1.0667853962600178, "ewc_loss": 0.0966796875, "ewc_loss_diag": 0.00016117095947265625, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.68088436126709, "learning_rate": 1e-05, "loss": 0.472, "mean_token_accuracy": 0.8707431554794312, "num_tokens": 318326923.0, "step": 8386 }, { "epoch": 1.0669126065386083, "ewc_loss": 0.11181640625, "ewc_loss_diag": 0.00016117095947265625, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.407221794128418, "learning_rate": 1e-05, "loss": 0.5501, "mean_token_accuracy": 0.8488694429397583, "num_tokens": 318364159.0, "step": 8387 }, { "epoch": 1.0670398168171988, "ewc_loss": 0.09375, "ewc_loss_diag": 0.00016117095947265625, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.451257228851318, "learning_rate": 1e-05, "loss": 0.552, "mean_token_accuracy": 0.8459086418151855, "num_tokens": 318402715.0, "step": 8388 }, { "epoch": 1.0671670270957894, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.00016117095947265625, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.630760192871094, "learning_rate": 1e-05, "loss": 0.5241, "mean_token_accuracy": 0.8635961413383484, "num_tokens": 318445925.0, "step": 8389 }, { "epoch": 1.06729423737438, "ewc_loss": 0.09423828125, "ewc_loss_diag": 0.00016117095947265625, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.419017314910889, "learning_rate": 1e-05, "loss": 0.5015, "mean_token_accuracy": 0.8591774702072144, "num_tokens": 318481756.0, "step": 8390 }, { "epoch": 1.0674214476529704, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.00016117095947265625, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.870972633361816, "learning_rate": 1e-05, "loss": 0.5123, "mean_token_accuracy": 0.8674169778823853, "num_tokens": 318526885.0, "step": 8391 }, { "epoch": 1.067548657931561, "ewc_loss": 0.0986328125, "ewc_loss_diag": 0.00016117095947265625, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.58390998840332, "learning_rate": 1e-05, "loss": 0.5066, "mean_token_accuracy": 0.8635721206665039, "num_tokens": 318569396.0, "step": 8392 }, { "epoch": 1.0676758682101515, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.00016117095947265625, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.964984893798828, "learning_rate": 1e-05, "loss": 0.473, "mean_token_accuracy": 0.8834011554718018, "num_tokens": 318606404.0, "step": 8393 }, { "epoch": 1.0678030784887418, "ewc_loss": 0.10498046875, "ewc_loss_diag": 0.00016117095947265625, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 5.836279392242432, "learning_rate": 1e-05, "loss": 0.5343, "mean_token_accuracy": 0.853614866733551, "num_tokens": 318642457.0, "step": 8394 }, { "epoch": 1.0679302887673323, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.00016117095947265625, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.819426536560059, "learning_rate": 1e-05, "loss": 0.5367, "mean_token_accuracy": 0.8577579259872437, "num_tokens": 318680498.0, "step": 8395 }, { "epoch": 1.0680574990459228, "ewc_loss": 0.10693359375, "ewc_loss_diag": 0.00016117095947265625, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 5.975105285644531, "learning_rate": 1e-05, "loss": 0.5201, "mean_token_accuracy": 0.8608201742172241, "num_tokens": 318720493.0, "step": 8396 }, { "epoch": 1.0681847093245134, "ewc_loss": 0.11474609375, "ewc_loss_diag": 0.00016117095947265625, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.549020290374756, "learning_rate": 1e-05, "loss": 0.524, "mean_token_accuracy": 0.8596020340919495, "num_tokens": 318755297.0, "step": 8397 }, { "epoch": 1.068311919603104, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.00016117095947265625, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 5.961371898651123, "learning_rate": 1e-05, "loss": 0.4821, "mean_token_accuracy": 0.8743487596511841, "num_tokens": 318788716.0, "step": 8398 }, { "epoch": 1.0684391298816944, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.00016117095947265625, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.306580543518066, "learning_rate": 1e-05, "loss": 0.4887, "mean_token_accuracy": 0.8671971559524536, "num_tokens": 318825130.0, "step": 8399 }, { "epoch": 1.068566340160285, "ewc_loss": 0.10302734375, "ewc_loss_diag": 0.00016117095947265625, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.883449077606201, "learning_rate": 1e-05, "loss": 0.4885, "mean_token_accuracy": 0.869056761264801, "num_tokens": 318861216.0, "step": 8400 }, { "epoch": 1.0686935504388755, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.00016117095947265625, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 6.146181583404541, "learning_rate": 1e-05, "loss": 0.4856, "mean_token_accuracy": 0.8708655834197998, "num_tokens": 318896410.0, "step": 8401 }, { "epoch": 1.068820760717466, "ewc_loss": 0.099609375, "ewc_loss_diag": 0.00016117095947265625, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.786022186279297, "learning_rate": 1e-05, "loss": 0.444, "mean_token_accuracy": 0.8816331028938293, "num_tokens": 318933834.0, "step": 8402 }, { "epoch": 1.0689479709960565, "ewc_loss": 0.10400390625, "ewc_loss_diag": 0.00016117095947265625, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 6.094054698944092, "learning_rate": 1e-05, "loss": 0.5388, "mean_token_accuracy": 0.8502864837646484, "num_tokens": 318973752.0, "step": 8403 }, { "epoch": 1.069075181274647, "ewc_loss": 0.09619140625, "ewc_loss_diag": 0.00016117095947265625, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.65562105178833, "learning_rate": 1e-05, "loss": 0.4274, "mean_token_accuracy": 0.8853460550308228, "num_tokens": 319009376.0, "step": 8404 }, { "epoch": 1.0692023915532376, "ewc_loss": 0.10498046875, "ewc_loss_diag": 0.00016117095947265625, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.1478400230407715, "learning_rate": 1e-05, "loss": 0.4883, "mean_token_accuracy": 0.8718850612640381, "num_tokens": 319050175.0, "step": 8405 }, { "epoch": 1.0693296018318281, "ewc_loss": 0.0947265625, "ewc_loss_diag": 0.00016117095947265625, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.528486728668213, "learning_rate": 1e-05, "loss": 0.4663, "mean_token_accuracy": 0.8714175820350647, "num_tokens": 319080808.0, "step": 8406 }, { "epoch": 1.0694568121104184, "ewc_loss": 0.10791015625, "ewc_loss_diag": 0.00016117095947265625, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.288898944854736, "learning_rate": 1e-05, "loss": 0.4673, "mean_token_accuracy": 0.8740959167480469, "num_tokens": 319123123.0, "step": 8407 }, { "epoch": 1.069584022389009, "ewc_loss": 0.09326171875, "ewc_loss_diag": 0.00016117095947265625, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.420035362243652, "learning_rate": 1e-05, "loss": 0.4587, "mean_token_accuracy": 0.8748984336853027, "num_tokens": 319155116.0, "step": 8408 }, { "epoch": 1.0697112326675995, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.00016117095947265625, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.513432502746582, "learning_rate": 1e-05, "loss": 0.4774, "mean_token_accuracy": 0.8754544258117676, "num_tokens": 319188087.0, "step": 8409 }, { "epoch": 1.06983844294619, "ewc_loss": 0.09423828125, "ewc_loss_diag": 0.00016117095947265625, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.395038604736328, "learning_rate": 1e-05, "loss": 0.5184, "mean_token_accuracy": 0.8577226400375366, "num_tokens": 319228820.0, "step": 8410 }, { "epoch": 1.0699656532247805, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.00016117095947265625, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.752136707305908, "learning_rate": 1e-05, "loss": 0.5847, "mean_token_accuracy": 0.8455923199653625, "num_tokens": 319266046.0, "step": 8411 }, { "epoch": 1.070092863503371, "ewc_loss": 0.09765625, "ewc_loss_diag": 0.00016117095947265625, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.513227939605713, "learning_rate": 1e-05, "loss": 0.4683, "mean_token_accuracy": 0.8722804188728333, "num_tokens": 319305209.0, "step": 8412 }, { "epoch": 1.0702200737819616, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.00016117095947265625, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.849851131439209, "learning_rate": 1e-05, "loss": 0.5282, "mean_token_accuracy": 0.8645994663238525, "num_tokens": 319347463.0, "step": 8413 }, { "epoch": 1.0703472840605521, "ewc_loss": 0.10302734375, "ewc_loss_diag": 0.00016117095947265625, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.775854587554932, "learning_rate": 1e-05, "loss": 0.448, "mean_token_accuracy": 0.8780197501182556, "num_tokens": 319383965.0, "step": 8414 }, { "epoch": 1.0704744943391427, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.00016117095947265625, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.675498008728027, "learning_rate": 1e-05, "loss": 0.469, "mean_token_accuracy": 0.8767306804656982, "num_tokens": 319418330.0, "step": 8415 }, { "epoch": 1.0706017046177332, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.00016117095947265625, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 5.969667434692383, "learning_rate": 1e-05, "loss": 0.5056, "mean_token_accuracy": 0.866182804107666, "num_tokens": 319460518.0, "step": 8416 }, { "epoch": 1.0707289148963237, "ewc_loss": 0.109375, "ewc_loss_diag": 0.00016117095947265625, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.269764423370361, "learning_rate": 1e-05, "loss": 0.5177, "mean_token_accuracy": 0.8611856698989868, "num_tokens": 319504044.0, "step": 8417 }, { "epoch": 1.070856125174914, "ewc_loss": 0.10888671875, "ewc_loss_diag": 0.00016117095947265625, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.087753772735596, "learning_rate": 1e-05, "loss": 0.4939, "mean_token_accuracy": 0.8692245483398438, "num_tokens": 319540707.0, "step": 8418 }, { "epoch": 1.0709833354535045, "ewc_loss": 0.09716796875, "ewc_loss_diag": 0.00016117095947265625, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.795008659362793, "learning_rate": 1e-05, "loss": 0.4595, "mean_token_accuracy": 0.8727426528930664, "num_tokens": 319580968.0, "step": 8419 }, { "epoch": 1.071110545732095, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.00016117095947265625, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.272851943969727, "learning_rate": 1e-05, "loss": 0.5367, "mean_token_accuracy": 0.8581293821334839, "num_tokens": 319619271.0, "step": 8420 }, { "epoch": 1.0712377560106856, "ewc_loss": 0.091796875, "ewc_loss_diag": 0.00016117095947265625, "ewc_loss_parallel": 7.581710815429688e-05, "grad_norm": 5.409903526306152, "learning_rate": 1e-05, "loss": 0.5024, "mean_token_accuracy": 0.8614932298660278, "num_tokens": 319657728.0, "step": 8421 }, { "epoch": 1.0713649662892761, "ewc_loss": 0.11474609375, "ewc_loss_diag": 0.00016117095947265625, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.595151424407959, "learning_rate": 1e-05, "loss": 0.5171, "mean_token_accuracy": 0.860255241394043, "num_tokens": 319696807.0, "step": 8422 }, { "epoch": 1.0714921765678667, "ewc_loss": 0.091796875, "ewc_loss_diag": 0.00016117095947265625, "ewc_loss_parallel": 7.581710815429688e-05, "grad_norm": 5.2523417472839355, "learning_rate": 1e-05, "loss": 0.5038, "mean_token_accuracy": 0.8583589196205139, "num_tokens": 319734397.0, "step": 8423 }, { "epoch": 1.0716193868464572, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.00016117095947265625, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 7.022204875946045, "learning_rate": 1e-05, "loss": 0.5502, "mean_token_accuracy": 0.8528231382369995, "num_tokens": 319769247.0, "step": 8424 }, { "epoch": 1.0717465971250477, "ewc_loss": 0.09814453125, "ewc_loss_diag": 0.00016117095947265625, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.483582496643066, "learning_rate": 1e-05, "loss": 0.489, "mean_token_accuracy": 0.8648779392242432, "num_tokens": 319801147.0, "step": 8425 }, { "epoch": 1.0718738074036382, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.00016117095947265625, "ewc_loss_parallel": 0.00011587142944335938, "grad_norm": 7.254992485046387, "learning_rate": 1e-05, "loss": 0.4944, "mean_token_accuracy": 0.8741624355316162, "num_tokens": 319834083.0, "step": 8426 }, { "epoch": 1.0720010176822288, "ewc_loss": 0.11083984375, "ewc_loss_diag": 0.00016117095947265625, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 5.958595275878906, "learning_rate": 1e-05, "loss": 0.4463, "mean_token_accuracy": 0.8817540407180786, "num_tokens": 319872586.0, "step": 8427 }, { "epoch": 1.0721282279608193, "ewc_loss": 0.125, "ewc_loss_diag": 0.00016117095947265625, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 7.036888599395752, "learning_rate": 1e-05, "loss": 0.5129, "mean_token_accuracy": 0.8675110340118408, "num_tokens": 319902756.0, "step": 8428 }, { "epoch": 1.0722554382394098, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.00016117095947265625, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.186779022216797, "learning_rate": 1e-05, "loss": 0.4878, "mean_token_accuracy": 0.8718883991241455, "num_tokens": 319943610.0, "step": 8429 }, { "epoch": 1.0723826485180004, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.00016117095947265625, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.638361930847168, "learning_rate": 1e-05, "loss": 0.5206, "mean_token_accuracy": 0.8627249002456665, "num_tokens": 319983762.0, "step": 8430 }, { "epoch": 1.0725098587965907, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.00016117095947265625, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.145369529724121, "learning_rate": 1e-05, "loss": 0.4958, "mean_token_accuracy": 0.8682803511619568, "num_tokens": 320019560.0, "step": 8431 }, { "epoch": 1.0726370690751812, "ewc_loss": 0.10791015625, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.309521198272705, "learning_rate": 1e-05, "loss": 0.4994, "mean_token_accuracy": 0.8617321848869324, "num_tokens": 320052907.0, "step": 8432 }, { "epoch": 1.0727642793537717, "ewc_loss": 0.10498046875, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.015895366668701, "learning_rate": 1e-05, "loss": 0.5421, "mean_token_accuracy": 0.8515381217002869, "num_tokens": 320087670.0, "step": 8433 }, { "epoch": 1.0728914896323622, "ewc_loss": 0.10400390625, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 6.110180377960205, "learning_rate": 1e-05, "loss": 0.4967, "mean_token_accuracy": 0.8681766986846924, "num_tokens": 320125768.0, "step": 8434 }, { "epoch": 1.0730186999109528, "ewc_loss": 0.10107421875, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.868553638458252, "learning_rate": 1e-05, "loss": 0.4599, "mean_token_accuracy": 0.8758065104484558, "num_tokens": 320163498.0, "step": 8435 }, { "epoch": 1.0731459101895433, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 6.037688255310059, "learning_rate": 1e-05, "loss": 0.5054, "mean_token_accuracy": 0.8641374111175537, "num_tokens": 320200951.0, "step": 8436 }, { "epoch": 1.0732731204681338, "ewc_loss": 0.09814453125, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.723539352416992, "learning_rate": 1e-05, "loss": 0.4631, "mean_token_accuracy": 0.8743178844451904, "num_tokens": 320239258.0, "step": 8437 }, { "epoch": 1.0734003307467244, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 6.065281391143799, "learning_rate": 1e-05, "loss": 0.487, "mean_token_accuracy": 0.8641693592071533, "num_tokens": 320276872.0, "step": 8438 }, { "epoch": 1.073527541025315, "ewc_loss": 0.09521484375, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.569677829742432, "learning_rate": 1e-05, "loss": 0.4948, "mean_token_accuracy": 0.8649118542671204, "num_tokens": 320315340.0, "step": 8439 }, { "epoch": 1.0736547513039054, "ewc_loss": 0.10595703125, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.190327167510986, "learning_rate": 1e-05, "loss": 0.4745, "mean_token_accuracy": 0.8741915225982666, "num_tokens": 320351213.0, "step": 8440 }, { "epoch": 1.073781961582496, "ewc_loss": 0.0927734375, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.419260025024414, "learning_rate": 1e-05, "loss": 0.4929, "mean_token_accuracy": 0.8632968664169312, "num_tokens": 320391456.0, "step": 8441 }, { "epoch": 1.0739091718610865, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.441679000854492, "learning_rate": 1e-05, "loss": 0.4825, "mean_token_accuracy": 0.8723790645599365, "num_tokens": 320428193.0, "step": 8442 }, { "epoch": 1.0740363821396768, "ewc_loss": 0.091796875, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 7.581710815429688e-05, "grad_norm": 5.305782318115234, "learning_rate": 1e-05, "loss": 0.4611, "mean_token_accuracy": 0.8725350499153137, "num_tokens": 320466685.0, "step": 8443 }, { "epoch": 1.0741635924182673, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.805312156677246, "learning_rate": 1e-05, "loss": 0.5333, "mean_token_accuracy": 0.8601172566413879, "num_tokens": 320508936.0, "step": 8444 }, { "epoch": 1.0742908026968578, "ewc_loss": 0.09423828125, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.3990159034729, "learning_rate": 1e-05, "loss": 0.547, "mean_token_accuracy": 0.8468889594078064, "num_tokens": 320546644.0, "step": 8445 }, { "epoch": 1.0744180129754484, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 0.00011205673217773438, "grad_norm": 7.110340118408203, "learning_rate": 1e-05, "loss": 0.435, "mean_token_accuracy": 0.8892138004302979, "num_tokens": 320571074.0, "step": 8446 }, { "epoch": 1.074545223254039, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 5.767570972442627, "learning_rate": 1e-05, "loss": 0.4933, "mean_token_accuracy": 0.8631993532180786, "num_tokens": 320605412.0, "step": 8447 }, { "epoch": 1.0746724335326294, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 0.00010967254638671875, "grad_norm": 7.029421806335449, "learning_rate": 1e-05, "loss": 0.5334, "mean_token_accuracy": 0.8616945743560791, "num_tokens": 320643077.0, "step": 8448 }, { "epoch": 1.07479964381122, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.0970048904418945, "learning_rate": 1e-05, "loss": 0.4733, "mean_token_accuracy": 0.8745870590209961, "num_tokens": 320680630.0, "step": 8449 }, { "epoch": 1.0749268540898105, "ewc_loss": 0.11474609375, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.604477882385254, "learning_rate": 1e-05, "loss": 0.4728, "mean_token_accuracy": 0.8772438764572144, "num_tokens": 320720405.0, "step": 8450 }, { "epoch": 1.075054064368401, "ewc_loss": 0.11376953125, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.203070163726807, "learning_rate": 1e-05, "loss": 0.4825, "mean_token_accuracy": 0.8708764314651489, "num_tokens": 320753453.0, "step": 8451 }, { "epoch": 1.0751812746469915, "ewc_loss": 0.10400390625, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 6.102073669433594, "learning_rate": 1e-05, "loss": 0.4767, "mean_token_accuracy": 0.8718996047973633, "num_tokens": 320795316.0, "step": 8452 }, { "epoch": 1.075308484925582, "ewc_loss": 0.11181640625, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.232285499572754, "learning_rate": 1e-05, "loss": 0.4829, "mean_token_accuracy": 0.8710302114486694, "num_tokens": 320834022.0, "step": 8453 }, { "epoch": 1.0754356952041726, "ewc_loss": 0.0966796875, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.70927095413208, "learning_rate": 1e-05, "loss": 0.5926, "mean_token_accuracy": 0.8323261141777039, "num_tokens": 320878410.0, "step": 8454 }, { "epoch": 1.0755629054827631, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.364569187164307, "learning_rate": 1e-05, "loss": 0.4364, "mean_token_accuracy": 0.886366605758667, "num_tokens": 320918767.0, "step": 8455 }, { "epoch": 1.0756901157613534, "ewc_loss": 0.09375, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.465079307556152, "learning_rate": 1e-05, "loss": 0.4812, "mean_token_accuracy": 0.869423508644104, "num_tokens": 320963845.0, "step": 8456 }, { "epoch": 1.075817326039944, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.592785358428955, "learning_rate": 1e-05, "loss": 0.486, "mean_token_accuracy": 0.872819185256958, "num_tokens": 321004811.0, "step": 8457 }, { "epoch": 1.0759445363185345, "ewc_loss": 0.09375, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.402015209197998, "learning_rate": 1e-05, "loss": 0.489, "mean_token_accuracy": 0.8632200956344604, "num_tokens": 321040894.0, "step": 8458 }, { "epoch": 1.076071746597125, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.864653587341309, "learning_rate": 1e-05, "loss": 0.565, "mean_token_accuracy": 0.8502069115638733, "num_tokens": 321083602.0, "step": 8459 }, { "epoch": 1.0761989568757155, "ewc_loss": 0.09814453125, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.542958736419678, "learning_rate": 1e-05, "loss": 0.5077, "mean_token_accuracy": 0.8668535947799683, "num_tokens": 321119444.0, "step": 8460 }, { "epoch": 1.076326167154306, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 7.000485420227051, "learning_rate": 1e-05, "loss": 0.4839, "mean_token_accuracy": 0.8723897933959961, "num_tokens": 321155125.0, "step": 8461 }, { "epoch": 1.0764533774328966, "ewc_loss": 0.10400390625, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 5.794139385223389, "learning_rate": 1e-05, "loss": 0.46, "mean_token_accuracy": 0.8765193223953247, "num_tokens": 321193645.0, "step": 8462 }, { "epoch": 1.0765805877114871, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.913321495056152, "learning_rate": 1e-05, "loss": 0.5608, "mean_token_accuracy": 0.852806806564331, "num_tokens": 321227118.0, "step": 8463 }, { "epoch": 1.0767077979900777, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 5.98681640625, "learning_rate": 1e-05, "loss": 0.4952, "mean_token_accuracy": 0.8667170405387878, "num_tokens": 321263245.0, "step": 8464 }, { "epoch": 1.0768350082686682, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.632897853851318, "learning_rate": 1e-05, "loss": 0.5683, "mean_token_accuracy": 0.8438221216201782, "num_tokens": 321301476.0, "step": 8465 }, { "epoch": 1.0769622185472587, "ewc_loss": 0.10791015625, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.0143280029296875, "learning_rate": 1e-05, "loss": 0.4881, "mean_token_accuracy": 0.8711999654769897, "num_tokens": 321333464.0, "step": 8466 }, { "epoch": 1.077089428825849, "ewc_loss": 0.10791015625, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.341330528259277, "learning_rate": 1e-05, "loss": 0.4924, "mean_token_accuracy": 0.8673827052116394, "num_tokens": 321367807.0, "step": 8467 }, { "epoch": 1.0772166391044395, "ewc_loss": 0.10400390625, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 5.933008193969727, "learning_rate": 1e-05, "loss": 0.4615, "mean_token_accuracy": 0.8771967887878418, "num_tokens": 321408638.0, "step": 8468 }, { "epoch": 1.07734384938303, "ewc_loss": 0.10302734375, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 6.134671688079834, "learning_rate": 1e-05, "loss": 0.5051, "mean_token_accuracy": 0.8630886077880859, "num_tokens": 321451036.0, "step": 8469 }, { "epoch": 1.0774710596616206, "ewc_loss": 0.10107421875, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.835832118988037, "learning_rate": 1e-05, "loss": 0.4426, "mean_token_accuracy": 0.8791835904121399, "num_tokens": 321491423.0, "step": 8470 }, { "epoch": 1.0775982699402111, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 6.052768707275391, "learning_rate": 1e-05, "loss": 0.5199, "mean_token_accuracy": 0.8574842214584351, "num_tokens": 321527956.0, "step": 8471 }, { "epoch": 1.0777254802188017, "ewc_loss": 0.09814453125, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.726561546325684, "learning_rate": 1e-05, "loss": 0.4486, "mean_token_accuracy": 0.8780652284622192, "num_tokens": 321560185.0, "step": 8472 }, { "epoch": 1.0778526904973922, "ewc_loss": 0.10400390625, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 6.058972358703613, "learning_rate": 1e-05, "loss": 0.4734, "mean_token_accuracy": 0.8725055456161499, "num_tokens": 321602073.0, "step": 8473 }, { "epoch": 1.0779799007759827, "ewc_loss": 0.095703125, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.608808994293213, "learning_rate": 1e-05, "loss": 0.4782, "mean_token_accuracy": 0.8647831082344055, "num_tokens": 321636722.0, "step": 8474 }, { "epoch": 1.0781071110545732, "ewc_loss": 0.10498046875, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.135749816894531, "learning_rate": 1e-05, "loss": 0.4958, "mean_token_accuracy": 0.8682008385658264, "num_tokens": 321675348.0, "step": 8475 }, { "epoch": 1.0782343213331638, "ewc_loss": 0.0947265625, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.478177070617676, "learning_rate": 1e-05, "loss": 0.4802, "mean_token_accuracy": 0.8669601678848267, "num_tokens": 321711983.0, "step": 8476 }, { "epoch": 1.0783615316117543, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.319876670837402, "learning_rate": 1e-05, "loss": 0.4477, "mean_token_accuracy": 0.8822412490844727, "num_tokens": 321748268.0, "step": 8477 }, { "epoch": 1.0784887418903448, "ewc_loss": 0.0927734375, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.3677496910095215, "learning_rate": 1e-05, "loss": 0.4901, "mean_token_accuracy": 0.8642928600311279, "num_tokens": 321783082.0, "step": 8478 }, { "epoch": 1.0786159521689354, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.591490745544434, "learning_rate": 1e-05, "loss": 0.5176, "mean_token_accuracy": 0.8611566424369812, "num_tokens": 321823634.0, "step": 8479 }, { "epoch": 1.0787431624475257, "ewc_loss": 0.09423828125, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.361938953399658, "learning_rate": 1e-05, "loss": 0.4621, "mean_token_accuracy": 0.874660849571228, "num_tokens": 321859033.0, "step": 8480 }, { "epoch": 1.0788703727261162, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.874343395233154, "learning_rate": 1e-05, "loss": 0.5173, "mean_token_accuracy": 0.8654626607894897, "num_tokens": 321896227.0, "step": 8481 }, { "epoch": 1.0789975830047067, "ewc_loss": 0.0986328125, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.5569586753845215, "learning_rate": 1e-05, "loss": 0.4853, "mean_token_accuracy": 0.8677371144294739, "num_tokens": 321934329.0, "step": 8482 }, { "epoch": 1.0791247932832972, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 6.968138217926025, "learning_rate": 1e-05, "loss": 0.5216, "mean_token_accuracy": 0.8614490032196045, "num_tokens": 321977426.0, "step": 8483 }, { "epoch": 1.0792520035618878, "ewc_loss": 0.10693359375, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 5.880638599395752, "learning_rate": 1e-05, "loss": 0.4883, "mean_token_accuracy": 0.8688188791275024, "num_tokens": 322014603.0, "step": 8484 }, { "epoch": 1.0793792138404783, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.693692684173584, "learning_rate": 1e-05, "loss": 0.4794, "mean_token_accuracy": 0.8772026896476746, "num_tokens": 322051809.0, "step": 8485 }, { "epoch": 1.0795064241190688, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.0726637840271, "learning_rate": 1e-05, "loss": 0.5055, "mean_token_accuracy": 0.8667806386947632, "num_tokens": 322091628.0, "step": 8486 }, { "epoch": 1.0796336343976594, "ewc_loss": 0.10791015625, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.2406325340271, "learning_rate": 1e-05, "loss": 0.5264, "mean_token_accuracy": 0.8592289090156555, "num_tokens": 322130616.0, "step": 8487 }, { "epoch": 1.0797608446762499, "ewc_loss": 0.11083984375, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.149735450744629, "learning_rate": 1e-05, "loss": 0.4571, "mean_token_accuracy": 0.8805373907089233, "num_tokens": 322172525.0, "step": 8488 }, { "epoch": 1.0798880549548404, "ewc_loss": 0.09765625, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.767456531524658, "learning_rate": 1e-05, "loss": 0.4709, "mean_token_accuracy": 0.874221682548523, "num_tokens": 322213438.0, "step": 8489 }, { "epoch": 1.080015265233431, "ewc_loss": 0.11083984375, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.318027973175049, "learning_rate": 1e-05, "loss": 0.478, "mean_token_accuracy": 0.872995138168335, "num_tokens": 322252492.0, "step": 8490 }, { "epoch": 1.0801424755120215, "ewc_loss": 0.09423828125, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.424505710601807, "learning_rate": 1e-05, "loss": 0.4756, "mean_token_accuracy": 0.8687458038330078, "num_tokens": 322295320.0, "step": 8491 }, { "epoch": 1.0802696857906118, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.63029146194458, "learning_rate": 1e-05, "loss": 0.4593, "mean_token_accuracy": 0.8816059231758118, "num_tokens": 322335475.0, "step": 8492 }, { "epoch": 1.0803968960692023, "ewc_loss": 0.09326171875, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.325693607330322, "learning_rate": 1e-05, "loss": 0.4543, "mean_token_accuracy": 0.8722027540206909, "num_tokens": 322373559.0, "step": 8493 }, { "epoch": 1.0805241063477928, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 6.994594573974609, "learning_rate": 1e-05, "loss": 0.5356, "mean_token_accuracy": 0.858637809753418, "num_tokens": 322407978.0, "step": 8494 }, { "epoch": 1.0806513166263834, "ewc_loss": 0.0986328125, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.547847747802734, "learning_rate": 1e-05, "loss": 0.4856, "mean_token_accuracy": 0.8680148124694824, "num_tokens": 322443657.0, "step": 8495 }, { "epoch": 1.080778526904974, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 0.0001125335693359375, "grad_norm": 7.14998722076416, "learning_rate": 1e-05, "loss": 0.4574, "mean_token_accuracy": 0.8868857026100159, "num_tokens": 322479424.0, "step": 8496 }, { "epoch": 1.0809057371835644, "ewc_loss": 0.10791015625, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 5.937495231628418, "learning_rate": 1e-05, "loss": 0.5352, "mean_token_accuracy": 0.8541662096977234, "num_tokens": 322523112.0, "step": 8497 }, { "epoch": 1.081032947462155, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.938180446624756, "learning_rate": 1e-05, "loss": 0.4941, "mean_token_accuracy": 0.8722811937332153, "num_tokens": 322560444.0, "step": 8498 }, { "epoch": 1.0811601577407455, "ewc_loss": 0.11181640625, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.11789608001709, "learning_rate": 1e-05, "loss": 0.4825, "mean_token_accuracy": 0.8692523241043091, "num_tokens": 322591211.0, "step": 8499 }, { "epoch": 1.081287368019336, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.591848373413086, "learning_rate": 1e-05, "loss": 0.5713, "mean_token_accuracy": 0.8492370843887329, "num_tokens": 322625129.0, "step": 8500 }, { "epoch": 1.0814145782979265, "ewc_loss": 0.109375, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.070638656616211, "learning_rate": 1e-05, "loss": 0.4858, "mean_token_accuracy": 0.8714015483856201, "num_tokens": 322663428.0, "step": 8501 }, { "epoch": 1.081541788576517, "ewc_loss": 0.10791015625, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.303315162658691, "learning_rate": 1e-05, "loss": 0.466, "mean_token_accuracy": 0.8753928542137146, "num_tokens": 322707936.0, "step": 8502 }, { "epoch": 1.0816689988551076, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 5.954921245574951, "learning_rate": 1e-05, "loss": 0.4587, "mean_token_accuracy": 0.8785704970359802, "num_tokens": 322748605.0, "step": 8503 }, { "epoch": 1.0817962091336981, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 6.135651588439941, "learning_rate": 1e-05, "loss": 0.5373, "mean_token_accuracy": 0.8493958711624146, "num_tokens": 322788009.0, "step": 8504 }, { "epoch": 1.0819234194122884, "ewc_loss": 0.099609375, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.814347267150879, "learning_rate": 1e-05, "loss": 0.4941, "mean_token_accuracy": 0.8637821674346924, "num_tokens": 322825159.0, "step": 8505 }, { "epoch": 1.082050629690879, "ewc_loss": 0.10302734375, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 6.108397960662842, "learning_rate": 1e-05, "loss": 0.5655, "mean_token_accuracy": 0.8421250581741333, "num_tokens": 322866769.0, "step": 8506 }, { "epoch": 1.0821778399694695, "ewc_loss": 0.09765625, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.661200523376465, "learning_rate": 1e-05, "loss": 0.4376, "mean_token_accuracy": 0.8770010471343994, "num_tokens": 322901851.0, "step": 8507 }, { "epoch": 1.08230505024806, "ewc_loss": 0.10498046875, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.18654203414917, "learning_rate": 1e-05, "loss": 0.485, "mean_token_accuracy": 0.8724395632743835, "num_tokens": 322938527.0, "step": 8508 }, { "epoch": 1.0824322605266505, "ewc_loss": 0.0947265625, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.50758171081543, "learning_rate": 1e-05, "loss": 0.4635, "mean_token_accuracy": 0.8714278936386108, "num_tokens": 322972135.0, "step": 8509 }, { "epoch": 1.082559470805241, "ewc_loss": 0.10888671875, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.373985767364502, "learning_rate": 1e-05, "loss": 0.5259, "mean_token_accuracy": 0.8563694953918457, "num_tokens": 323010451.0, "step": 8510 }, { "epoch": 1.0826866810838316, "ewc_loss": 0.09326171875, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.387032985687256, "learning_rate": 1e-05, "loss": 0.4733, "mean_token_accuracy": 0.8686168193817139, "num_tokens": 323047583.0, "step": 8511 }, { "epoch": 1.0828138913624221, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.654827117919922, "learning_rate": 1e-05, "loss": 0.5158, "mean_token_accuracy": 0.8595867156982422, "num_tokens": 323088754.0, "step": 8512 }, { "epoch": 1.0829411016410126, "ewc_loss": 0.09375, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.3919358253479, "learning_rate": 1e-05, "loss": 0.4634, "mean_token_accuracy": 0.8726588487625122, "num_tokens": 323128510.0, "step": 8513 }, { "epoch": 1.0830683119196032, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.938231468200684, "learning_rate": 1e-05, "loss": 0.5287, "mean_token_accuracy": 0.8616151809692383, "num_tokens": 323172557.0, "step": 8514 }, { "epoch": 1.0831955221981937, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.646113395690918, "learning_rate": 1e-05, "loss": 0.4413, "mean_token_accuracy": 0.877831220626831, "num_tokens": 323208298.0, "step": 8515 }, { "epoch": 1.083322732476784, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.958888530731201, "learning_rate": 1e-05, "loss": 0.54, "mean_token_accuracy": 0.8581479787826538, "num_tokens": 323250824.0, "step": 8516 }, { "epoch": 1.0834499427553745, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 5.964296817779541, "learning_rate": 1e-05, "loss": 0.4961, "mean_token_accuracy": 0.8683172464370728, "num_tokens": 323290792.0, "step": 8517 }, { "epoch": 1.083577153033965, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.630827903747559, "learning_rate": 1e-05, "loss": 0.5011, "mean_token_accuracy": 0.8638186454772949, "num_tokens": 323327806.0, "step": 8518 }, { "epoch": 1.0837043633125556, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.1289777755737305, "learning_rate": 1e-05, "loss": 0.5063, "mean_token_accuracy": 0.8672702312469482, "num_tokens": 323364275.0, "step": 8519 }, { "epoch": 1.0838315735911461, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.151810169219971, "learning_rate": 1e-05, "loss": 0.4707, "mean_token_accuracy": 0.8751273155212402, "num_tokens": 323395671.0, "step": 8520 }, { "epoch": 1.0839587838697367, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.193327903747559, "learning_rate": 1e-05, "loss": 0.4929, "mean_token_accuracy": 0.8687126040458679, "num_tokens": 323432719.0, "step": 8521 }, { "epoch": 1.0840859941483272, "ewc_loss": 0.0966796875, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.698115825653076, "learning_rate": 1e-05, "loss": 0.5123, "mean_token_accuracy": 0.8602043390274048, "num_tokens": 323474431.0, "step": 8522 }, { "epoch": 1.0842132044269177, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.35808801651001, "learning_rate": 1e-05, "loss": 0.4944, "mean_token_accuracy": 0.8714720010757446, "num_tokens": 323515439.0, "step": 8523 }, { "epoch": 1.0843404147055082, "ewc_loss": 0.09375, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.401215553283691, "learning_rate": 1e-05, "loss": 0.5132, "mean_token_accuracy": 0.8577053546905518, "num_tokens": 323552613.0, "step": 8524 }, { "epoch": 1.0844676249840988, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.6715192794799805, "learning_rate": 1e-05, "loss": 0.44, "mean_token_accuracy": 0.8869385719299316, "num_tokens": 323590037.0, "step": 8525 }, { "epoch": 1.0845948352626893, "ewc_loss": 0.09375, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.327570915222168, "learning_rate": 1e-05, "loss": 0.445, "mean_token_accuracy": 0.8789567947387695, "num_tokens": 323634540.0, "step": 8526 }, { "epoch": 1.0847220455412798, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 7.039360523223877, "learning_rate": 1e-05, "loss": 0.4954, "mean_token_accuracy": 0.8707786798477173, "num_tokens": 323666772.0, "step": 8527 }, { "epoch": 1.0848492558198704, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.596488952636719, "learning_rate": 1e-05, "loss": 0.4934, "mean_token_accuracy": 0.867207407951355, "num_tokens": 323707847.0, "step": 8528 }, { "epoch": 1.0849764660984607, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 0.00011157989501953125, "grad_norm": 7.14555025100708, "learning_rate": 1e-05, "loss": 0.4611, "mean_token_accuracy": 0.885242760181427, "num_tokens": 323737648.0, "step": 8529 }, { "epoch": 1.0851036763770512, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 5.972020626068115, "learning_rate": 1e-05, "loss": 0.4775, "mean_token_accuracy": 0.8728933334350586, "num_tokens": 323772621.0, "step": 8530 }, { "epoch": 1.0852308866556417, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.914577960968018, "learning_rate": 1e-05, "loss": 0.5516, "mean_token_accuracy": 0.8569601774215698, "num_tokens": 323806384.0, "step": 8531 }, { "epoch": 1.0853580969342322, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.116901397705078, "learning_rate": 1e-05, "loss": 0.4314, "mean_token_accuracy": 0.8887481093406677, "num_tokens": 323845287.0, "step": 8532 }, { "epoch": 1.0854853072128228, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.565194606781006, "learning_rate": 1e-05, "loss": 0.4761, "mean_token_accuracy": 0.8752287030220032, "num_tokens": 323884182.0, "step": 8533 }, { "epoch": 1.0856125174914133, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.076143264770508, "learning_rate": 1e-05, "loss": 0.5546, "mean_token_accuracy": 0.849754810333252, "num_tokens": 323919357.0, "step": 8534 }, { "epoch": 1.0857397277700038, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.2821149826049805, "learning_rate": 1e-05, "loss": 0.4666, "mean_token_accuracy": 0.8767890930175781, "num_tokens": 323957993.0, "step": 8535 }, { "epoch": 1.0858669380485944, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 5.9571027755737305, "learning_rate": 1e-05, "loss": 0.4435, "mean_token_accuracy": 0.8797461986541748, "num_tokens": 323991726.0, "step": 8536 }, { "epoch": 1.0859941483271849, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 6.125282287597656, "learning_rate": 1e-05, "loss": 0.4841, "mean_token_accuracy": 0.8677440881729126, "num_tokens": 324031639.0, "step": 8537 }, { "epoch": 1.0861213586057754, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.823582172393799, "learning_rate": 1e-05, "loss": 0.5126, "mean_token_accuracy": 0.857886552810669, "num_tokens": 324063359.0, "step": 8538 }, { "epoch": 1.086248568884366, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 6.0803117752075195, "learning_rate": 1e-05, "loss": 0.4987, "mean_token_accuracy": 0.8644761443138123, "num_tokens": 324105058.0, "step": 8539 }, { "epoch": 1.0863757791629565, "ewc_loss": 0.09765625, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.678844451904297, "learning_rate": 1e-05, "loss": 0.5228, "mean_token_accuracy": 0.857755184173584, "num_tokens": 324144209.0, "step": 8540 }, { "epoch": 1.0865029894415468, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.146697044372559, "learning_rate": 1e-05, "loss": 0.4681, "mean_token_accuracy": 0.8725465536117554, "num_tokens": 324179194.0, "step": 8541 }, { "epoch": 1.0866301997201373, "ewc_loss": 0.0947265625, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.517683506011963, "learning_rate": 1e-05, "loss": 0.4552, "mean_token_accuracy": 0.8759111762046814, "num_tokens": 324218405.0, "step": 8542 }, { "epoch": 1.0867574099987278, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.326245307922363, "learning_rate": 1e-05, "loss": 0.571, "mean_token_accuracy": 0.8442274332046509, "num_tokens": 324258298.0, "step": 8543 }, { "epoch": 1.0868846202773184, "ewc_loss": 0.0927734375, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.396307468414307, "learning_rate": 1e-05, "loss": 0.4497, "mean_token_accuracy": 0.875992476940155, "num_tokens": 324296518.0, "step": 8544 }, { "epoch": 1.0870118305559089, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.616479396820068, "learning_rate": 1e-05, "loss": 0.5101, "mean_token_accuracy": 0.866478681564331, "num_tokens": 324336724.0, "step": 8545 }, { "epoch": 1.0871390408344994, "ewc_loss": 0.09375, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.374161720275879, "learning_rate": 1e-05, "loss": 0.4967, "mean_token_accuracy": 0.862582802772522, "num_tokens": 324373464.0, "step": 8546 }, { "epoch": 1.08726625111309, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.936809539794922, "learning_rate": 1e-05, "loss": 0.4899, "mean_token_accuracy": 0.8726352453231812, "num_tokens": 324410893.0, "step": 8547 }, { "epoch": 1.0873934613916805, "ewc_loss": 0.099609375, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.586390495300293, "learning_rate": 1e-05, "loss": 0.4802, "mean_token_accuracy": 0.8742365837097168, "num_tokens": 324446765.0, "step": 8548 }, { "epoch": 1.087520671670271, "ewc_loss": 0.125, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 7.035051345825195, "learning_rate": 1e-05, "loss": 0.5012, "mean_token_accuracy": 0.8699630498886108, "num_tokens": 324482745.0, "step": 8549 }, { "epoch": 1.0876478819488615, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 5.945344924926758, "learning_rate": 1e-05, "loss": 0.4835, "mean_token_accuracy": 0.8739857077598572, "num_tokens": 324513764.0, "step": 8550 }, { "epoch": 1.087775092227452, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.745286464691162, "learning_rate": 1e-05, "loss": 0.5058, "mean_token_accuracy": 0.866644561290741, "num_tokens": 324549502.0, "step": 8551 }, { "epoch": 1.0879023025060426, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.127875804901123, "learning_rate": 1e-05, "loss": 0.4975, "mean_token_accuracy": 0.8666501045227051, "num_tokens": 324584456.0, "step": 8552 }, { "epoch": 1.0880295127846331, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.274456024169922, "learning_rate": 1e-05, "loss": 0.5091, "mean_token_accuracy": 0.8611173629760742, "num_tokens": 324624070.0, "step": 8553 }, { "epoch": 1.0881567230632234, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.194571018218994, "learning_rate": 1e-05, "loss": 0.4974, "mean_token_accuracy": 0.8674036860466003, "num_tokens": 324665971.0, "step": 8554 }, { "epoch": 1.088283933341814, "ewc_loss": 0.0986328125, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.793190956115723, "learning_rate": 1e-05, "loss": 0.4377, "mean_token_accuracy": 0.8813869953155518, "num_tokens": 324706705.0, "step": 8555 }, { "epoch": 1.0884111436204045, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.35213041305542, "learning_rate": 1e-05, "loss": 0.5185, "mean_token_accuracy": 0.858746349811554, "num_tokens": 324744701.0, "step": 8556 }, { "epoch": 1.088538353898995, "ewc_loss": 0.0927734375, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.460394859313965, "learning_rate": 1e-05, "loss": 0.4867, "mean_token_accuracy": 0.8667311668395996, "num_tokens": 324781447.0, "step": 8557 }, { "epoch": 1.0886655641775855, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.6305131912231445, "learning_rate": 1e-05, "loss": 0.498, "mean_token_accuracy": 0.8708153963088989, "num_tokens": 324816753.0, "step": 8558 }, { "epoch": 1.088792774456176, "ewc_loss": 0.09375, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.344329833984375, "learning_rate": 1e-05, "loss": 0.4974, "mean_token_accuracy": 0.8620650768280029, "num_tokens": 324855093.0, "step": 8559 }, { "epoch": 1.0889199847347666, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 6.984797477722168, "learning_rate": 1e-05, "loss": 0.5196, "mean_token_accuracy": 0.8692854642868042, "num_tokens": 324893107.0, "step": 8560 }, { "epoch": 1.0890471950133571, "ewc_loss": 0.0986328125, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.54095458984375, "learning_rate": 1e-05, "loss": 0.4846, "mean_token_accuracy": 0.8649134635925293, "num_tokens": 324927589.0, "step": 8561 }, { "epoch": 1.0891744052919476, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 7.161990642547607, "learning_rate": 1e-05, "loss": 0.4935, "mean_token_accuracy": 0.8761616945266724, "num_tokens": 324958801.0, "step": 8562 }, { "epoch": 1.0893016155705382, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 5.927079677581787, "learning_rate": 1e-05, "loss": 0.4785, "mean_token_accuracy": 0.8721041679382324, "num_tokens": 324995617.0, "step": 8563 }, { "epoch": 1.0894288258491287, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.977721214294434, "learning_rate": 1e-05, "loss": 0.5193, "mean_token_accuracy": 0.8632726073265076, "num_tokens": 325035702.0, "step": 8564 }, { "epoch": 1.089556036127719, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.1229448318481445, "learning_rate": 1e-05, "loss": 0.5336, "mean_token_accuracy": 0.8577803373336792, "num_tokens": 325078063.0, "step": 8565 }, { "epoch": 1.0896832464063095, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.630542755126953, "learning_rate": 1e-05, "loss": 0.5082, "mean_token_accuracy": 0.8640167713165283, "num_tokens": 325113127.0, "step": 8566 }, { "epoch": 1.0898104566849, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.099386692047119, "learning_rate": 1e-05, "loss": 0.5141, "mean_token_accuracy": 0.8639400005340576, "num_tokens": 325149451.0, "step": 8567 }, { "epoch": 1.0899376669634906, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.311705112457275, "learning_rate": 1e-05, "loss": 0.4991, "mean_token_accuracy": 0.8655818700790405, "num_tokens": 325193265.0, "step": 8568 }, { "epoch": 1.0900648772420811, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 5.9876179695129395, "learning_rate": 1e-05, "loss": 0.4794, "mean_token_accuracy": 0.8708442449569702, "num_tokens": 325235160.0, "step": 8569 }, { "epoch": 1.0901920875206716, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 6.12412691116333, "learning_rate": 1e-05, "loss": 0.5047, "mean_token_accuracy": 0.8645452857017517, "num_tokens": 325276299.0, "step": 8570 }, { "epoch": 1.0903192977992622, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.86159610748291, "learning_rate": 1e-05, "loss": 0.523, "mean_token_accuracy": 0.8536441922187805, "num_tokens": 325313509.0, "step": 8571 }, { "epoch": 1.0904465080778527, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 6.06256628036499, "learning_rate": 1e-05, "loss": 0.5226, "mean_token_accuracy": 0.8572126626968384, "num_tokens": 325355883.0, "step": 8572 }, { "epoch": 1.0905737183564432, "ewc_loss": 0.09765625, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.710830211639404, "learning_rate": 1e-05, "loss": 0.4737, "mean_token_accuracy": 0.8744948506355286, "num_tokens": 325392937.0, "step": 8573 }, { "epoch": 1.0907009286350338, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 6.120601177215576, "learning_rate": 1e-05, "loss": 0.4888, "mean_token_accuracy": 0.8690454959869385, "num_tokens": 325422132.0, "step": 8574 }, { "epoch": 1.0908281389136243, "ewc_loss": 0.0947265625, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.551011562347412, "learning_rate": 1e-05, "loss": 0.4899, "mean_token_accuracy": 0.8648403286933899, "num_tokens": 325460383.0, "step": 8575 }, { "epoch": 1.0909553491922148, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.270759105682373, "learning_rate": 1e-05, "loss": 0.5143, "mean_token_accuracy": 0.8602503538131714, "num_tokens": 325499242.0, "step": 8576 }, { "epoch": 1.0910825594708053, "ewc_loss": 0.0927734375, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.405638694763184, "learning_rate": 1e-05, "loss": 0.4636, "mean_token_accuracy": 0.8737205862998962, "num_tokens": 325535389.0, "step": 8577 }, { "epoch": 1.0912097697493957, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.546826362609863, "learning_rate": 1e-05, "loss": 0.4659, "mean_token_accuracy": 0.8758769631385803, "num_tokens": 325574459.0, "step": 8578 }, { "epoch": 1.0913369800279862, "ewc_loss": 0.0927734375, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.349334716796875, "learning_rate": 1e-05, "loss": 0.533, "mean_token_accuracy": 0.8514139652252197, "num_tokens": 325616296.0, "step": 8579 }, { "epoch": 1.0914641903065767, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.889598369598389, "learning_rate": 1e-05, "loss": 0.5253, "mean_token_accuracy": 0.8579931259155273, "num_tokens": 325652707.0, "step": 8580 }, { "epoch": 1.0915914005851672, "ewc_loss": 0.09765625, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.506321430206299, "learning_rate": 1e-05, "loss": 0.4503, "mean_token_accuracy": 0.8796108961105347, "num_tokens": 325692423.0, "step": 8581 }, { "epoch": 1.0917186108637578, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 0.00010967254638671875, "grad_norm": 7.074544906616211, "learning_rate": 1e-05, "loss": 0.5105, "mean_token_accuracy": 0.8677331209182739, "num_tokens": 325731638.0, "step": 8582 }, { "epoch": 1.0918458211423483, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 5.867126941680908, "learning_rate": 1e-05, "loss": 0.475, "mean_token_accuracy": 0.8740049004554749, "num_tokens": 325771482.0, "step": 8583 }, { "epoch": 1.0919730314209388, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.877180099487305, "learning_rate": 1e-05, "loss": 0.5312, "mean_token_accuracy": 0.8605291843414307, "num_tokens": 325807447.0, "step": 8584 }, { "epoch": 1.0921002416995294, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.132034778594971, "learning_rate": 1e-05, "loss": 0.4927, "mean_token_accuracy": 0.8690508604049683, "num_tokens": 325848070.0, "step": 8585 }, { "epoch": 1.0922274519781199, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.409714221954346, "learning_rate": 1e-05, "loss": 0.4709, "mean_token_accuracy": 0.8737760782241821, "num_tokens": 325889027.0, "step": 8586 }, { "epoch": 1.0923546622567104, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.204096794128418, "learning_rate": 1e-05, "loss": 0.5121, "mean_token_accuracy": 0.864484429359436, "num_tokens": 325926547.0, "step": 8587 }, { "epoch": 1.092481872535301, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.908984184265137, "learning_rate": 1e-05, "loss": 0.4745, "mean_token_accuracy": 0.8716407418251038, "num_tokens": 325968968.0, "step": 8588 }, { "epoch": 1.0926090828138915, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.307427406311035, "learning_rate": 1e-05, "loss": 0.5035, "mean_token_accuracy": 0.8695130348205566, "num_tokens": 326005060.0, "step": 8589 }, { "epoch": 1.0927362930924818, "ewc_loss": 0.095703125, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.5312676429748535, "learning_rate": 1e-05, "loss": 0.5169, "mean_token_accuracy": 0.856587290763855, "num_tokens": 326049895.0, "step": 8590 }, { "epoch": 1.0928635033710723, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.564323425292969, "learning_rate": 1e-05, "loss": 0.5396, "mean_token_accuracy": 0.8587368726730347, "num_tokens": 326087772.0, "step": 8591 }, { "epoch": 1.0929907136496628, "ewc_loss": 0.09375, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.362995147705078, "learning_rate": 1e-05, "loss": 0.4612, "mean_token_accuracy": 0.8727583289146423, "num_tokens": 326126518.0, "step": 8592 }, { "epoch": 1.0931179239282534, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.904022693634033, "learning_rate": 1e-05, "loss": 0.5359, "mean_token_accuracy": 0.8581844568252563, "num_tokens": 326166672.0, "step": 8593 }, { "epoch": 1.0932451342068439, "ewc_loss": 0.09765625, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.481540203094482, "learning_rate": 1e-05, "loss": 0.486, "mean_token_accuracy": 0.8634603023529053, "num_tokens": 326209099.0, "step": 8594 }, { "epoch": 1.0933723444854344, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 0.00010967254638671875, "grad_norm": 7.139626979827881, "learning_rate": 1e-05, "loss": 0.5499, "mean_token_accuracy": 0.8593942523002625, "num_tokens": 326244101.0, "step": 8595 }, { "epoch": 1.093499554764025, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 5.842140197753906, "learning_rate": 1e-05, "loss": 0.489, "mean_token_accuracy": 0.8662809133529663, "num_tokens": 326279997.0, "step": 8596 }, { "epoch": 1.0936267650426155, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 7.025242328643799, "learning_rate": 1e-05, "loss": 0.5279, "mean_token_accuracy": 0.8586935997009277, "num_tokens": 326321241.0, "step": 8597 }, { "epoch": 1.093753975321206, "ewc_loss": 0.11181640625, "ewc_loss_diag": 0.00016498565673828125, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.085241794586182, "learning_rate": 1e-05, "loss": 0.4986, "mean_token_accuracy": 0.8668520450592041, "num_tokens": 326359662.0, "step": 8598 }, { "epoch": 1.0938811855997965, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.702425479888916, "learning_rate": 1e-05, "loss": 0.4902, "mean_token_accuracy": 0.8753260374069214, "num_tokens": 326397072.0, "step": 8599 }, { "epoch": 1.094008395878387, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.093375205993652, "learning_rate": 1e-05, "loss": 0.5029, "mean_token_accuracy": 0.8634451031684875, "num_tokens": 326440029.0, "step": 8600 }, { "epoch": 1.0941356061569776, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.389970302581787, "learning_rate": 1e-05, "loss": 0.4807, "mean_token_accuracy": 0.8683452606201172, "num_tokens": 326479586.0, "step": 8601 }, { "epoch": 1.094262816435568, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 6.002941608428955, "learning_rate": 1e-05, "loss": 0.4956, "mean_token_accuracy": 0.8653910160064697, "num_tokens": 326518235.0, "step": 8602 }, { "epoch": 1.0943900267141584, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.180662631988525, "learning_rate": 1e-05, "loss": 0.4656, "mean_token_accuracy": 0.8724052309989929, "num_tokens": 326559420.0, "step": 8603 }, { "epoch": 1.094517236992749, "ewc_loss": 0.10009765625, "ewc_loss_diag": 0.00016498565673828125, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.892792224884033, "learning_rate": 1e-05, "loss": 0.5093, "mean_token_accuracy": 0.861860990524292, "num_tokens": 326597020.0, "step": 8604 }, { "epoch": 1.0946444472713395, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 6.092984676361084, "learning_rate": 1e-05, "loss": 0.5715, "mean_token_accuracy": 0.8412565588951111, "num_tokens": 326632248.0, "step": 8605 }, { "epoch": 1.09477165754993, "ewc_loss": 0.09765625, "ewc_loss_diag": 0.00016498565673828125, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.741851806640625, "learning_rate": 1e-05, "loss": 0.4794, "mean_token_accuracy": 0.8707382082939148, "num_tokens": 326666185.0, "step": 8606 }, { "epoch": 1.0948988678285205, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.00016498565673828125, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 6.13162899017334, "learning_rate": 1e-05, "loss": 0.5404, "mean_token_accuracy": 0.8490105271339417, "num_tokens": 326700356.0, "step": 8607 }, { "epoch": 1.095026078107111, "ewc_loss": 0.095703125, "ewc_loss_diag": 0.00016498565673828125, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.5869903564453125, "learning_rate": 1e-05, "loss": 0.4767, "mean_token_accuracy": 0.8700213432312012, "num_tokens": 326739448.0, "step": 8608 }, { "epoch": 1.0951532883857016, "ewc_loss": 0.10693359375, "ewc_loss_diag": 0.00016498565673828125, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.264834880828857, "learning_rate": 1e-05, "loss": 0.4885, "mean_token_accuracy": 0.8673261404037476, "num_tokens": 326776290.0, "step": 8609 }, { "epoch": 1.0952804986642921, "ewc_loss": 0.09228515625, "ewc_loss_diag": 0.00016498565673828125, "ewc_loss_parallel": 7.581710815429688e-05, "grad_norm": 5.440271377563477, "learning_rate": 1e-05, "loss": 0.4666, "mean_token_accuracy": 0.8702816963195801, "num_tokens": 326817975.0, "step": 8610 }, { "epoch": 1.0954077089428826, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.00016498565673828125, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.5081658363342285, "learning_rate": 1e-05, "loss": 0.4878, "mean_token_accuracy": 0.871407151222229, "num_tokens": 326852404.0, "step": 8611 }, { "epoch": 1.0955349192214732, "ewc_loss": 0.09375, "ewc_loss_diag": 0.00016498565673828125, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.38098669052124, "learning_rate": 1e-05, "loss": 0.5083, "mean_token_accuracy": 0.8566162586212158, "num_tokens": 326888519.0, "step": 8612 }, { "epoch": 1.0956621295000637, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.00016498565673828125, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.806628227233887, "learning_rate": 1e-05, "loss": 0.5137, "mean_token_accuracy": 0.8602156043052673, "num_tokens": 326925644.0, "step": 8613 }, { "epoch": 1.095789339778654, "ewc_loss": 0.0966796875, "ewc_loss_diag": 0.00016498565673828125, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.486571788787842, "learning_rate": 1e-05, "loss": 0.4881, "mean_token_accuracy": 0.8663330674171448, "num_tokens": 326961490.0, "step": 8614 }, { "epoch": 1.0959165500572445, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.00016498565673828125, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 7.003344535827637, "learning_rate": 1e-05, "loss": 0.5487, "mean_token_accuracy": 0.8561394214630127, "num_tokens": 326998726.0, "step": 8615 }, { "epoch": 1.096043760335835, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.00016498565673828125, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 5.795727729797363, "learning_rate": 1e-05, "loss": 0.5432, "mean_token_accuracy": 0.8520441651344299, "num_tokens": 327040717.0, "step": 8616 }, { "epoch": 1.0961709706144256, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.00016498565673828125, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.870307922363281, "learning_rate": 1e-05, "loss": 0.5199, "mean_token_accuracy": 0.8626480102539062, "num_tokens": 327070597.0, "step": 8617 }, { "epoch": 1.0962981808930161, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.00016498565673828125, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.059723377227783, "learning_rate": 1e-05, "loss": 0.5538, "mean_token_accuracy": 0.8508374094963074, "num_tokens": 327108598.0, "step": 8618 }, { "epoch": 1.0964253911716066, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.00016498565673828125, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.436319351196289, "learning_rate": 1e-05, "loss": 0.4627, "mean_token_accuracy": 0.8778167963027954, "num_tokens": 327142750.0, "step": 8619 }, { "epoch": 1.0965526014501972, "ewc_loss": 0.11083984375, "ewc_loss_diag": 0.00016498565673828125, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.1578216552734375, "learning_rate": 1e-05, "loss": 0.5295, "mean_token_accuracy": 0.8552969694137573, "num_tokens": 327182536.0, "step": 8620 }, { "epoch": 1.0966798117287877, "ewc_loss": 0.10107421875, "ewc_loss_diag": 0.00016498565673828125, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.957454681396484, "learning_rate": 1e-05, "loss": 0.4882, "mean_token_accuracy": 0.8650122284889221, "num_tokens": 327218135.0, "step": 8621 }, { "epoch": 1.0968070220073782, "ewc_loss": 0.11083984375, "ewc_loss_diag": 0.00016498565673828125, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.264721393585205, "learning_rate": 1e-05, "loss": 0.4666, "mean_token_accuracy": 0.8790141940116882, "num_tokens": 327251581.0, "step": 8622 }, { "epoch": 1.0969342322859688, "ewc_loss": 0.095703125, "ewc_loss_diag": 0.00016498565673828125, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.559937477111816, "learning_rate": 1e-05, "loss": 0.4735, "mean_token_accuracy": 0.8676583766937256, "num_tokens": 327288466.0, "step": 8623 }, { "epoch": 1.0970614425645593, "ewc_loss": 0.11376953125, "ewc_loss_diag": 0.00016498565673828125, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.497217655181885, "learning_rate": 1e-05, "loss": 0.4891, "mean_token_accuracy": 0.8723807334899902, "num_tokens": 327325646.0, "step": 8624 }, { "epoch": 1.0971886528431498, "ewc_loss": 0.09375, "ewc_loss_diag": 0.00016498565673828125, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.346357822418213, "learning_rate": 1e-05, "loss": 0.4979, "mean_token_accuracy": 0.8613331913948059, "num_tokens": 327358244.0, "step": 8625 }, { "epoch": 1.0973158631217403, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.00016498565673828125, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.847908973693848, "learning_rate": 1e-05, "loss": 0.4776, "mean_token_accuracy": 0.8738905191421509, "num_tokens": 327397315.0, "step": 8626 }, { "epoch": 1.0974430734003306, "ewc_loss": 0.095703125, "ewc_loss_diag": 0.00016498565673828125, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.974546909332275, "learning_rate": 1e-05, "loss": 0.471, "mean_token_accuracy": 0.8666975498199463, "num_tokens": 327432642.0, "step": 8627 }, { "epoch": 1.0975702836789212, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.00016498565673828125, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.571991443634033, "learning_rate": 1e-05, "loss": 0.5521, "mean_token_accuracy": 0.8555805683135986, "num_tokens": 327478563.0, "step": 8628 }, { "epoch": 1.0976974939575117, "ewc_loss": 0.10009765625, "ewc_loss_diag": 0.00016498565673828125, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.578590393066406, "learning_rate": 1e-05, "loss": 0.468, "mean_token_accuracy": 0.8746548891067505, "num_tokens": 327514295.0, "step": 8629 }, { "epoch": 1.0978247042361022, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.00016498565673828125, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.538332462310791, "learning_rate": 1e-05, "loss": 0.4813, "mean_token_accuracy": 0.8733851909637451, "num_tokens": 327550496.0, "step": 8630 }, { "epoch": 1.0979519145146928, "ewc_loss": 0.099609375, "ewc_loss_diag": 0.00016498565673828125, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.592388153076172, "learning_rate": 1e-05, "loss": 0.4419, "mean_token_accuracy": 0.8839764595031738, "num_tokens": 327584692.0, "step": 8631 }, { "epoch": 1.0980791247932833, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.00016498565673828125, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.568477630615234, "learning_rate": 1e-05, "loss": 0.4777, "mean_token_accuracy": 0.8756973743438721, "num_tokens": 327627710.0, "step": 8632 }, { "epoch": 1.0982063350718738, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.00016498565673828125, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.619477272033691, "learning_rate": 1e-05, "loss": 0.5074, "mean_token_accuracy": 0.8593810796737671, "num_tokens": 327666958.0, "step": 8633 }, { "epoch": 1.0983335453504643, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.00016498565673828125, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.60481071472168, "learning_rate": 1e-05, "loss": 0.488, "mean_token_accuracy": 0.8718546032905579, "num_tokens": 327700829.0, "step": 8634 }, { "epoch": 1.0984607556290549, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.00016498565673828125, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.665171146392822, "learning_rate": 1e-05, "loss": 0.5267, "mean_token_accuracy": 0.857903778553009, "num_tokens": 327741410.0, "step": 8635 }, { "epoch": 1.0985879659076454, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.00016498565673828125, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.5735015869140625, "learning_rate": 1e-05, "loss": 0.5477, "mean_token_accuracy": 0.8520933389663696, "num_tokens": 327778305.0, "step": 8636 }, { "epoch": 1.098715176186236, "ewc_loss": 0.10302734375, "ewc_loss_diag": 0.00016498565673828125, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.716450214385986, "learning_rate": 1e-05, "loss": 0.5156, "mean_token_accuracy": 0.8576310276985168, "num_tokens": 327813384.0, "step": 8637 }, { "epoch": 1.0988423864648265, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.487400054931641, "learning_rate": 1e-05, "loss": 0.4993, "mean_token_accuracy": 0.8704131245613098, "num_tokens": 327851928.0, "step": 8638 }, { "epoch": 1.0989695967434168, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 5.759245872497559, "learning_rate": 1e-05, "loss": 0.4543, "mean_token_accuracy": 0.8781331777572632, "num_tokens": 327890053.0, "step": 8639 }, { "epoch": 1.0990968070220073, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.333717346191406, "learning_rate": 1e-05, "loss": 0.503, "mean_token_accuracy": 0.8658787608146667, "num_tokens": 327928947.0, "step": 8640 }, { "epoch": 1.0992240173005978, "ewc_loss": 0.10498046875, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 5.799924850463867, "learning_rate": 1e-05, "loss": 0.4751, "mean_token_accuracy": 0.8717167377471924, "num_tokens": 327966286.0, "step": 8641 }, { "epoch": 1.0993512275791884, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.150223255157471, "learning_rate": 1e-05, "loss": 0.503, "mean_token_accuracy": 0.8639268279075623, "num_tokens": 328005480.0, "step": 8642 }, { "epoch": 1.0994784378577789, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 5.819235801696777, "learning_rate": 1e-05, "loss": 0.4779, "mean_token_accuracy": 0.8710170984268188, "num_tokens": 328040587.0, "step": 8643 }, { "epoch": 1.0996056481363694, "ewc_loss": 0.10595703125, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 5.9771199226379395, "learning_rate": 1e-05, "loss": 0.5091, "mean_token_accuracy": 0.859978199005127, "num_tokens": 328075856.0, "step": 8644 }, { "epoch": 1.09973285841496, "ewc_loss": 0.10302734375, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.828855037689209, "learning_rate": 1e-05, "loss": 0.5162, "mean_token_accuracy": 0.8565967082977295, "num_tokens": 328116243.0, "step": 8645 }, { "epoch": 1.0998600686935505, "ewc_loss": 0.10205078125, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.823603630065918, "learning_rate": 1e-05, "loss": 0.482, "mean_token_accuracy": 0.8693938255310059, "num_tokens": 328151854.0, "step": 8646 }, { "epoch": 1.099987278972141, "ewc_loss": 0.10302734375, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.852851390838623, "learning_rate": 1e-05, "loss": 0.5106, "mean_token_accuracy": 0.8592979907989502, "num_tokens": 328191455.0, "step": 8647 }, { "epoch": 1.1001144892507315, "ewc_loss": 0.099609375, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.699251651763916, "learning_rate": 1e-05, "loss": 0.4623, "mean_token_accuracy": 0.8719438910484314, "num_tokens": 328232034.0, "step": 8648 }, { "epoch": 1.100241699529322, "ewc_loss": 0.10302734375, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.892730712890625, "learning_rate": 1e-05, "loss": 0.5041, "mean_token_accuracy": 0.8677839040756226, "num_tokens": 328265581.0, "step": 8649 }, { "epoch": 1.1003689098079126, "ewc_loss": 0.09765625, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.6009602546691895, "learning_rate": 1e-05, "loss": 0.5588, "mean_token_accuracy": 0.8501339554786682, "num_tokens": 328304354.0, "step": 8650 }, { "epoch": 1.100496120086503, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.95650053024292, "learning_rate": 1e-05, "loss": 0.517, "mean_token_accuracy": 0.8588750958442688, "num_tokens": 328347924.0, "step": 8651 }, { "epoch": 1.1006233303650934, "ewc_loss": 0.09619140625, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.478303909301758, "learning_rate": 1e-05, "loss": 0.4383, "mean_token_accuracy": 0.8822472095489502, "num_tokens": 328385603.0, "step": 8652 }, { "epoch": 1.100750540643684, "ewc_loss": 0.10693359375, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.111433029174805, "learning_rate": 1e-05, "loss": 0.4747, "mean_token_accuracy": 0.8728345632553101, "num_tokens": 328422881.0, "step": 8653 }, { "epoch": 1.1008777509222745, "ewc_loss": 0.0927734375, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 7.62939453125e-05, "grad_norm": 5.331856727600098, "learning_rate": 1e-05, "loss": 0.4897, "mean_token_accuracy": 0.8646618127822876, "num_tokens": 328461191.0, "step": 8654 }, { "epoch": 1.101004961200865, "ewc_loss": 0.11083984375, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.389847278594971, "learning_rate": 1e-05, "loss": 0.5058, "mean_token_accuracy": 0.8653590083122253, "num_tokens": 328499824.0, "step": 8655 }, { "epoch": 1.1011321714794555, "ewc_loss": 0.09033203125, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 7.390975952148438e-05, "grad_norm": 5.16653299331665, "learning_rate": 1e-05, "loss": 0.472, "mean_token_accuracy": 0.8732844591140747, "num_tokens": 328538456.0, "step": 8656 }, { "epoch": 1.101259381758046, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.894486904144287, "learning_rate": 1e-05, "loss": 0.5499, "mean_token_accuracy": 0.8523578643798828, "num_tokens": 328574722.0, "step": 8657 }, { "epoch": 1.1013865920366366, "ewc_loss": 0.09423828125, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.168844699859619, "learning_rate": 1e-05, "loss": 0.4776, "mean_token_accuracy": 0.8690035343170166, "num_tokens": 328608248.0, "step": 8658 }, { "epoch": 1.101513802315227, "ewc_loss": 0.140625, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 0.0001239776611328125, "grad_norm": 7.5099663734436035, "learning_rate": 1e-05, "loss": 0.5402, "mean_token_accuracy": 0.867228090763092, "num_tokens": 328651011.0, "step": 8659 }, { "epoch": 1.1016410125938176, "ewc_loss": 0.10888671875, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 5.725036144256592, "learning_rate": 1e-05, "loss": 0.5151, "mean_token_accuracy": 0.861029863357544, "num_tokens": 328694438.0, "step": 8660 }, { "epoch": 1.1017682228724082, "ewc_loss": 0.1455078125, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 0.00012874603271484375, "grad_norm": 7.69277811050415, "learning_rate": 1e-05, "loss": 0.5447, "mean_token_accuracy": 0.8671013116836548, "num_tokens": 328731851.0, "step": 8661 }, { "epoch": 1.1018954331509987, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 0.0001125335693359375, "grad_norm": 6.474905490875244, "learning_rate": 1e-05, "loss": 0.5239, "mean_token_accuracy": 0.8661078810691833, "num_tokens": 328772159.0, "step": 8662 }, { "epoch": 1.102022643429589, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 0.00011444091796875, "grad_norm": 7.199787139892578, "learning_rate": 1e-05, "loss": 0.4676, "mean_token_accuracy": 0.8832383155822754, "num_tokens": 328814712.0, "step": 8663 }, { "epoch": 1.1021498537081795, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 0.00010967254638671875, "grad_norm": 6.529299736022949, "learning_rate": 1e-05, "loss": 0.53, "mean_token_accuracy": 0.8628052473068237, "num_tokens": 328854243.0, "step": 8664 }, { "epoch": 1.10227706398677, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.765322685241699, "learning_rate": 1e-05, "loss": 0.5367, "mean_token_accuracy": 0.8557353019714355, "num_tokens": 328894109.0, "step": 8665 }, { "epoch": 1.1024042742653606, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.289662837982178, "learning_rate": 1e-05, "loss": 0.4723, "mean_token_accuracy": 0.8783116936683655, "num_tokens": 328937275.0, "step": 8666 }, { "epoch": 1.1025314845439511, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.4674530029296875, "learning_rate": 1e-05, "loss": 0.4766, "mean_token_accuracy": 0.8757139444351196, "num_tokens": 328974572.0, "step": 8667 }, { "epoch": 1.1026586948225416, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.084075450897217, "learning_rate": 1e-05, "loss": 0.5135, "mean_token_accuracy": 0.8631279468536377, "num_tokens": 329013871.0, "step": 8668 }, { "epoch": 1.1027859051011322, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.2964324951171875, "learning_rate": 1e-05, "loss": 0.5383, "mean_token_accuracy": 0.8510764837265015, "num_tokens": 329050333.0, "step": 8669 }, { "epoch": 1.1029131153797227, "ewc_loss": 0.10107421875, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.892818927764893, "learning_rate": 1e-05, "loss": 0.5031, "mean_token_accuracy": 0.861266553401947, "num_tokens": 329091817.0, "step": 8670 }, { "epoch": 1.1030403256583132, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.22554349899292, "learning_rate": 1e-05, "loss": 0.4778, "mean_token_accuracy": 0.873297929763794, "num_tokens": 329128523.0, "step": 8671 }, { "epoch": 1.1031675359369038, "ewc_loss": 0.09765625, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.71340799331665, "learning_rate": 1e-05, "loss": 0.436, "mean_token_accuracy": 0.8834792375564575, "num_tokens": 329166320.0, "step": 8672 }, { "epoch": 1.1032947462154943, "ewc_loss": 0.10888671875, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.298160076141357, "learning_rate": 1e-05, "loss": 0.4795, "mean_token_accuracy": 0.8704681396484375, "num_tokens": 329195140.0, "step": 8673 }, { "epoch": 1.1034219564940848, "ewc_loss": 0.0947265625, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.532712459564209, "learning_rate": 1e-05, "loss": 0.4523, "mean_token_accuracy": 0.8768402338027954, "num_tokens": 329230814.0, "step": 8674 }, { "epoch": 1.1035491667726753, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.499495983123779, "learning_rate": 1e-05, "loss": 0.4745, "mean_token_accuracy": 0.8732994198799133, "num_tokens": 329270695.0, "step": 8675 }, { "epoch": 1.1036763770512656, "ewc_loss": 0.09423828125, "ewc_loss_diag": 0.00016689300537109375, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.412533283233643, "learning_rate": 1e-05, "loss": 0.5326, "mean_token_accuracy": 0.8530945777893066, "num_tokens": 329308061.0, "step": 8676 }, { "epoch": 1.1038035873298562, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.00016689300537109375, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.83134126663208, "learning_rate": 1e-05, "loss": 0.4993, "mean_token_accuracy": 0.864600419998169, "num_tokens": 329347093.0, "step": 8677 }, { "epoch": 1.1039307976084467, "ewc_loss": 0.09619140625, "ewc_loss_diag": 0.00016689300537109375, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.474687099456787, "learning_rate": 1e-05, "loss": 0.497, "mean_token_accuracy": 0.8654272556304932, "num_tokens": 329385009.0, "step": 8678 }, { "epoch": 1.1040580078870372, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.00016689300537109375, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 7.139910697937012, "learning_rate": 1e-05, "loss": 0.5323, "mean_token_accuracy": 0.865208625793457, "num_tokens": 329425317.0, "step": 8679 }, { "epoch": 1.1041852181656278, "ewc_loss": 0.10498046875, "ewc_loss_diag": 0.00016689300537109375, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 5.8377604484558105, "learning_rate": 1e-05, "loss": 0.4696, "mean_token_accuracy": 0.874138355255127, "num_tokens": 329461581.0, "step": 8680 }, { "epoch": 1.1043124284442183, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.00016689300537109375, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 7.047182559967041, "learning_rate": 1e-05, "loss": 0.4811, "mean_token_accuracy": 0.879666805267334, "num_tokens": 329494064.0, "step": 8681 }, { "epoch": 1.1044396387228088, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.00016689300537109375, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.194703578948975, "learning_rate": 1e-05, "loss": 0.4963, "mean_token_accuracy": 0.8722647428512573, "num_tokens": 329530807.0, "step": 8682 }, { "epoch": 1.1045668490013993, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.00016689300537109375, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.581019401550293, "learning_rate": 1e-05, "loss": 0.5731, "mean_token_accuracy": 0.8454303741455078, "num_tokens": 329575689.0, "step": 8683 }, { "epoch": 1.1046940592799899, "ewc_loss": 0.11376953125, "ewc_loss_diag": 0.00016689300537109375, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.2830047607421875, "learning_rate": 1e-05, "loss": 0.5264, "mean_token_accuracy": 0.8607913255691528, "num_tokens": 329616855.0, "step": 8684 }, { "epoch": 1.1048212695585804, "ewc_loss": 0.10205078125, "ewc_loss_diag": 0.00016689300537109375, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 6.039426803588867, "learning_rate": 1e-05, "loss": 0.4984, "mean_token_accuracy": 0.8672531843185425, "num_tokens": 329654085.0, "step": 8685 }, { "epoch": 1.104948479837171, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.00016689300537109375, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.361414909362793, "learning_rate": 1e-05, "loss": 0.5112, "mean_token_accuracy": 0.8679303526878357, "num_tokens": 329688375.0, "step": 8686 }, { "epoch": 1.1050756901157615, "ewc_loss": 0.09521484375, "ewc_loss_diag": 0.00016689300537109375, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.60257625579834, "learning_rate": 1e-05, "loss": 0.5142, "mean_token_accuracy": 0.8575685620307922, "num_tokens": 329734258.0, "step": 8687 }, { "epoch": 1.1052029003943518, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.577774524688721, "learning_rate": 1e-05, "loss": 0.5159, "mean_token_accuracy": 0.869247317314148, "num_tokens": 329772080.0, "step": 8688 }, { "epoch": 1.1053301106729423, "ewc_loss": 0.09326171875, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.366004943847656, "learning_rate": 1e-05, "loss": 0.4542, "mean_token_accuracy": 0.8747130632400513, "num_tokens": 329817809.0, "step": 8689 }, { "epoch": 1.1054573209515328, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.954763889312744, "learning_rate": 1e-05, "loss": 0.4837, "mean_token_accuracy": 0.8738842606544495, "num_tokens": 329856063.0, "step": 8690 }, { "epoch": 1.1055845312301233, "ewc_loss": 0.09619140625, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.4834465980529785, "learning_rate": 1e-05, "loss": 0.5133, "mean_token_accuracy": 0.8567279577255249, "num_tokens": 329892548.0, "step": 8691 }, { "epoch": 1.1057117415087139, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 0.00011348724365234375, "grad_norm": 7.2359209060668945, "learning_rate": 1e-05, "loss": 0.4638, "mean_token_accuracy": 0.8836389780044556, "num_tokens": 329931897.0, "step": 8692 }, { "epoch": 1.1058389517873044, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 5.907826900482178, "learning_rate": 1e-05, "loss": 0.5158, "mean_token_accuracy": 0.8617264032363892, "num_tokens": 329967477.0, "step": 8693 }, { "epoch": 1.105966162065895, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 7.112649917602539, "learning_rate": 1e-05, "loss": 0.5246, "mean_token_accuracy": 0.8660672903060913, "num_tokens": 330004840.0, "step": 8694 }, { "epoch": 1.1060933723444855, "ewc_loss": 0.11376953125, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.1867146492004395, "learning_rate": 1e-05, "loss": 0.4957, "mean_token_accuracy": 0.8685310482978821, "num_tokens": 330048928.0, "step": 8695 }, { "epoch": 1.106220582623076, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.746046543121338, "learning_rate": 1e-05, "loss": 0.5129, "mean_token_accuracy": 0.8639310598373413, "num_tokens": 330084112.0, "step": 8696 }, { "epoch": 1.1063477929016665, "ewc_loss": 0.11083984375, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.168072700500488, "learning_rate": 1e-05, "loss": 0.5298, "mean_token_accuracy": 0.8584885597229004, "num_tokens": 330127533.0, "step": 8697 }, { "epoch": 1.106475003180257, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.424931049346924, "learning_rate": 1e-05, "loss": 0.5683, "mean_token_accuracy": 0.8481401205062866, "num_tokens": 330169379.0, "step": 8698 }, { "epoch": 1.1066022134588476, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.040555477142334, "learning_rate": 1e-05, "loss": 0.4588, "mean_token_accuracy": 0.8761610984802246, "num_tokens": 330205553.0, "step": 8699 }, { "epoch": 1.106729423737438, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.223478317260742, "learning_rate": 1e-05, "loss": 0.4813, "mean_token_accuracy": 0.8706977367401123, "num_tokens": 330241088.0, "step": 8700 }, { "epoch": 1.1068566340160284, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.871142387390137, "learning_rate": 1e-05, "loss": 0.4834, "mean_token_accuracy": 0.8685131072998047, "num_tokens": 330282520.0, "step": 8701 }, { "epoch": 1.106983844294619, "ewc_loss": 0.10595703125, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 6.172879695892334, "learning_rate": 1e-05, "loss": 0.4997, "mean_token_accuracy": 0.8624695539474487, "num_tokens": 330326231.0, "step": 8702 }, { "epoch": 1.1071110545732095, "ewc_loss": 0.09765625, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.69589900970459, "learning_rate": 1e-05, "loss": 0.4515, "mean_token_accuracy": 0.8773735165596008, "num_tokens": 330363674.0, "step": 8703 }, { "epoch": 1.1072382648518, "ewc_loss": 0.10791015625, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.253259658813477, "learning_rate": 1e-05, "loss": 0.4194, "mean_token_accuracy": 0.8909150958061218, "num_tokens": 330398890.0, "step": 8704 }, { "epoch": 1.1073654751303905, "ewc_loss": 0.09375, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.550501823425293, "learning_rate": 1e-05, "loss": 0.455, "mean_token_accuracy": 0.8737399578094482, "num_tokens": 330434564.0, "step": 8705 }, { "epoch": 1.107492685408981, "ewc_loss": 0.11083984375, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.425611972808838, "learning_rate": 1e-05, "loss": 0.4652, "mean_token_accuracy": 0.8740708827972412, "num_tokens": 330475580.0, "step": 8706 }, { "epoch": 1.1076198956875716, "ewc_loss": 0.0927734375, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 7.62939453125e-05, "grad_norm": 5.443312644958496, "learning_rate": 1e-05, "loss": 0.4455, "mean_token_accuracy": 0.8789331912994385, "num_tokens": 330513255.0, "step": 8707 }, { "epoch": 1.107747105966162, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.698594570159912, "learning_rate": 1e-05, "loss": 0.5363, "mean_token_accuracy": 0.8578919172286987, "num_tokens": 330554101.0, "step": 8708 }, { "epoch": 1.1078743162447526, "ewc_loss": 0.09423828125, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.471048355102539, "learning_rate": 1e-05, "loss": 0.5331, "mean_token_accuracy": 0.8543940782546997, "num_tokens": 330595312.0, "step": 8709 }, { "epoch": 1.1080015265233432, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.937203884124756, "learning_rate": 1e-05, "loss": 0.4863, "mean_token_accuracy": 0.8746379017829895, "num_tokens": 330633510.0, "step": 8710 }, { "epoch": 1.1081287368019337, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.698161602020264, "learning_rate": 1e-05, "loss": 0.4789, "mean_token_accuracy": 0.8711231350898743, "num_tokens": 330673626.0, "step": 8711 }, { "epoch": 1.108255947080524, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.923617839813232, "learning_rate": 1e-05, "loss": 0.5319, "mean_token_accuracy": 0.8639676570892334, "num_tokens": 330715051.0, "step": 8712 }, { "epoch": 1.1083831573591145, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 5.986307621002197, "learning_rate": 1e-05, "loss": 0.496, "mean_token_accuracy": 0.8664581775665283, "num_tokens": 330765603.0, "step": 8713 }, { "epoch": 1.108510367637705, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.595516204833984, "learning_rate": 1e-05, "loss": 0.508, "mean_token_accuracy": 0.8643381595611572, "num_tokens": 330799427.0, "step": 8714 }, { "epoch": 1.1086375779162956, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.1414384841918945, "learning_rate": 1e-05, "loss": 0.4852, "mean_token_accuracy": 0.8707301020622253, "num_tokens": 330840810.0, "step": 8715 }, { "epoch": 1.108764788194886, "ewc_loss": 0.10400390625, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 6.098357677459717, "learning_rate": 1e-05, "loss": 0.5507, "mean_token_accuracy": 0.8570640683174133, "num_tokens": 330878325.0, "step": 8716 }, { "epoch": 1.1088919984734766, "ewc_loss": 0.10986328125, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.238255977630615, "learning_rate": 1e-05, "loss": 0.4551, "mean_token_accuracy": 0.8797223567962646, "num_tokens": 330914817.0, "step": 8717 }, { "epoch": 1.1090192087520672, "ewc_loss": 0.09521484375, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.631498336791992, "learning_rate": 1e-05, "loss": 0.4553, "mean_token_accuracy": 0.8763413429260254, "num_tokens": 330951890.0, "step": 8718 }, { "epoch": 1.1091464190306577, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.476901531219482, "learning_rate": 1e-05, "loss": 0.5303, "mean_token_accuracy": 0.857146143913269, "num_tokens": 330986807.0, "step": 8719 }, { "epoch": 1.1092736293092482, "ewc_loss": 0.091796875, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 7.534027099609375e-05, "grad_norm": 5.33795690536499, "learning_rate": 1e-05, "loss": 0.5072, "mean_token_accuracy": 0.8592529892921448, "num_tokens": 331027132.0, "step": 8720 }, { "epoch": 1.1094008395878387, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.86124849319458, "learning_rate": 1e-05, "loss": 0.4826, "mean_token_accuracy": 0.8727956414222717, "num_tokens": 331062669.0, "step": 8721 }, { "epoch": 1.1095280498664293, "ewc_loss": 0.09375, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.362548351287842, "learning_rate": 1e-05, "loss": 0.464, "mean_token_accuracy": 0.8700565695762634, "num_tokens": 331097376.0, "step": 8722 }, { "epoch": 1.1096552601450198, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 0.00011444091796875, "grad_norm": 7.25199031829834, "learning_rate": 1e-05, "loss": 0.4852, "mean_token_accuracy": 0.8762238025665283, "num_tokens": 331130930.0, "step": 8723 }, { "epoch": 1.1097824704236103, "ewc_loss": 0.10498046875, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 5.786902904510498, "learning_rate": 1e-05, "loss": 0.4809, "mean_token_accuracy": 0.8652248382568359, "num_tokens": 331172012.0, "step": 8724 }, { "epoch": 1.1099096807022006, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 0.00011157989501953125, "grad_norm": 7.23344087600708, "learning_rate": 1e-05, "loss": 0.5874, "mean_token_accuracy": 0.8446483612060547, "num_tokens": 331210016.0, "step": 8725 }, { "epoch": 1.1100368909807912, "ewc_loss": 0.11474609375, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.195946216583252, "learning_rate": 1e-05, "loss": 0.5237, "mean_token_accuracy": 0.86036217212677, "num_tokens": 331251818.0, "step": 8726 }, { "epoch": 1.1101641012593817, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.859363555908203, "learning_rate": 1e-05, "loss": 0.5346, "mean_token_accuracy": 0.8557416200637817, "num_tokens": 331283034.0, "step": 8727 }, { "epoch": 1.1102913115379722, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.235419273376465, "learning_rate": 1e-05, "loss": 0.515, "mean_token_accuracy": 0.8627265691757202, "num_tokens": 331324985.0, "step": 8728 }, { "epoch": 1.1104185218165628, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.462847709655762, "learning_rate": 1e-05, "loss": 0.4916, "mean_token_accuracy": 0.8701809644699097, "num_tokens": 331358595.0, "step": 8729 }, { "epoch": 1.1105457320951533, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.131414413452148, "learning_rate": 1e-05, "loss": 0.5002, "mean_token_accuracy": 0.8615675568580627, "num_tokens": 331387388.0, "step": 8730 }, { "epoch": 1.1106729423737438, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.195696830749512, "learning_rate": 1e-05, "loss": 0.521, "mean_token_accuracy": 0.8608773350715637, "num_tokens": 331423562.0, "step": 8731 }, { "epoch": 1.1108001526523343, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.980945587158203, "learning_rate": 1e-05, "loss": 0.4744, "mean_token_accuracy": 0.8712707757949829, "num_tokens": 331458835.0, "step": 8732 }, { "epoch": 1.1109273629309249, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 6.0794758796691895, "learning_rate": 1e-05, "loss": 0.5044, "mean_token_accuracy": 0.8599288463592529, "num_tokens": 331489889.0, "step": 8733 }, { "epoch": 1.1110545732095154, "ewc_loss": 0.099609375, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.809811115264893, "learning_rate": 1e-05, "loss": 0.4293, "mean_token_accuracy": 0.880763053894043, "num_tokens": 331529103.0, "step": 8734 }, { "epoch": 1.111181783488106, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 6.08742618560791, "learning_rate": 1e-05, "loss": 0.4693, "mean_token_accuracy": 0.8736836910247803, "num_tokens": 331566100.0, "step": 8735 }, { "epoch": 1.1113089937666965, "ewc_loss": 0.09619140625, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.637516975402832, "learning_rate": 1e-05, "loss": 0.4748, "mean_token_accuracy": 0.8743172287940979, "num_tokens": 331606704.0, "step": 8736 }, { "epoch": 1.1114362040452868, "ewc_loss": 0.10693359375, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.2108378410339355, "learning_rate": 1e-05, "loss": 0.4664, "mean_token_accuracy": 0.8760240077972412, "num_tokens": 331642019.0, "step": 8737 }, { "epoch": 1.1115634143238773, "ewc_loss": 0.09423828125, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.475155353546143, "learning_rate": 1e-05, "loss": 0.4703, "mean_token_accuracy": 0.8685771822929382, "num_tokens": 331672888.0, "step": 8738 }, { "epoch": 1.1116906246024678, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.45478630065918, "learning_rate": 1e-05, "loss": 0.4986, "mean_token_accuracy": 0.8702906966209412, "num_tokens": 331710383.0, "step": 8739 }, { "epoch": 1.1118178348810583, "ewc_loss": 0.0927734375, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 7.62939453125e-05, "grad_norm": 5.334621906280518, "learning_rate": 1e-05, "loss": 0.4704, "mean_token_accuracy": 0.8713979721069336, "num_tokens": 331754299.0, "step": 8740 }, { "epoch": 1.1119450451596489, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.831014633178711, "learning_rate": 1e-05, "loss": 0.5005, "mean_token_accuracy": 0.8703345060348511, "num_tokens": 331789329.0, "step": 8741 }, { "epoch": 1.1120722554382394, "ewc_loss": 0.0947265625, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.401359558105469, "learning_rate": 1e-05, "loss": 0.4614, "mean_token_accuracy": 0.8737136125564575, "num_tokens": 331822727.0, "step": 8742 }, { "epoch": 1.11219946571683, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 0.00011348724365234375, "grad_norm": 7.175235748291016, "learning_rate": 1e-05, "loss": 0.5011, "mean_token_accuracy": 0.8710860013961792, "num_tokens": 331863091.0, "step": 8743 }, { "epoch": 1.1123266759954205, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 5.782042503356934, "learning_rate": 1e-05, "loss": 0.4855, "mean_token_accuracy": 0.8709021806716919, "num_tokens": 331900866.0, "step": 8744 }, { "epoch": 1.112453886274011, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 7.121643543243408, "learning_rate": 1e-05, "loss": 0.476, "mean_token_accuracy": 0.8804726004600525, "num_tokens": 331936774.0, "step": 8745 }, { "epoch": 1.1125810965526015, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.172978401184082, "learning_rate": 1e-05, "loss": 0.4886, "mean_token_accuracy": 0.8710771799087524, "num_tokens": 331974629.0, "step": 8746 }, { "epoch": 1.112708306831192, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.688624382019043, "learning_rate": 1e-05, "loss": 0.4777, "mean_token_accuracy": 0.8756921887397766, "num_tokens": 332008707.0, "step": 8747 }, { "epoch": 1.1128355171097826, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.25490665435791, "learning_rate": 1e-05, "loss": 0.462, "mean_token_accuracy": 0.8800573348999023, "num_tokens": 332046973.0, "step": 8748 }, { "epoch": 1.112962727388373, "ewc_loss": 0.10595703125, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 6.181698322296143, "learning_rate": 1e-05, "loss": 0.5213, "mean_token_accuracy": 0.8590819239616394, "num_tokens": 332092658.0, "step": 8749 }, { "epoch": 1.1130899376669634, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.286135196685791, "learning_rate": 1e-05, "loss": 0.4812, "mean_token_accuracy": 0.8703622221946716, "num_tokens": 332129705.0, "step": 8750 }, { "epoch": 1.113217147945554, "ewc_loss": 0.09716796875, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.765352249145508, "learning_rate": 1e-05, "loss": 0.4784, "mean_token_accuracy": 0.8682486414909363, "num_tokens": 332163129.0, "step": 8751 }, { "epoch": 1.1133443582241445, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.406689167022705, "learning_rate": 1e-05, "loss": 0.5152, "mean_token_accuracy": 0.8645493984222412, "num_tokens": 332198148.0, "step": 8752 }, { "epoch": 1.113471568502735, "ewc_loss": 0.0947265625, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.4894866943359375, "learning_rate": 1e-05, "loss": 0.4775, "mean_token_accuracy": 0.8695950508117676, "num_tokens": 332238068.0, "step": 8753 }, { "epoch": 1.1135987787813255, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.647670745849609, "learning_rate": 1e-05, "loss": 0.5382, "mean_token_accuracy": 0.857169508934021, "num_tokens": 332277455.0, "step": 8754 }, { "epoch": 1.113725989059916, "ewc_loss": 0.0947265625, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.41984224319458, "learning_rate": 1e-05, "loss": 0.4928, "mean_token_accuracy": 0.8635920286178589, "num_tokens": 332314778.0, "step": 8755 }, { "epoch": 1.1138531993385066, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.929996967315674, "learning_rate": 1e-05, "loss": 0.4955, "mean_token_accuracy": 0.8731415867805481, "num_tokens": 332347863.0, "step": 8756 }, { "epoch": 1.113980409617097, "ewc_loss": 0.0986328125, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.573907375335693, "learning_rate": 1e-05, "loss": 0.4331, "mean_token_accuracy": 0.8807299733161926, "num_tokens": 332389577.0, "step": 8757 }, { "epoch": 1.1141076198956876, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 0.00010967254638671875, "grad_norm": 7.079813480377197, "learning_rate": 1e-05, "loss": 0.5289, "mean_token_accuracy": 0.8618409633636475, "num_tokens": 332434140.0, "step": 8758 }, { "epoch": 1.1142348301742782, "ewc_loss": 0.10791015625, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 5.8889994621276855, "learning_rate": 1e-05, "loss": 0.4737, "mean_token_accuracy": 0.8706789612770081, "num_tokens": 332468843.0, "step": 8759 }, { "epoch": 1.1143620404528687, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.00016689300537109375, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.92678165435791, "learning_rate": 1e-05, "loss": 0.4791, "mean_token_accuracy": 0.8758642077445984, "num_tokens": 332509863.0, "step": 8760 }, { "epoch": 1.114489250731459, "ewc_loss": 0.10986328125, "ewc_loss_diag": 0.00016689300537109375, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.062107563018799, "learning_rate": 1e-05, "loss": 0.5051, "mean_token_accuracy": 0.8648275136947632, "num_tokens": 332550242.0, "step": 8761 }, { "epoch": 1.1146164610100495, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.00016689300537109375, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.624274730682373, "learning_rate": 1e-05, "loss": 0.4913, "mean_token_accuracy": 0.8679729700088501, "num_tokens": 332588606.0, "step": 8762 }, { "epoch": 1.11474367128864, "ewc_loss": 0.109375, "ewc_loss_diag": 0.00016689300537109375, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.072019577026367, "learning_rate": 1e-05, "loss": 0.4799, "mean_token_accuracy": 0.870246171951294, "num_tokens": 332623787.0, "step": 8763 }, { "epoch": 1.1148708815672306, "ewc_loss": 0.109375, "ewc_loss_diag": 0.00016689300537109375, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.314180374145508, "learning_rate": 1e-05, "loss": 0.5249, "mean_token_accuracy": 0.8595010042190552, "num_tokens": 332665358.0, "step": 8764 }, { "epoch": 1.114998091845821, "ewc_loss": 0.10498046875, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 5.997307777404785, "learning_rate": 1e-05, "loss": 0.4521, "mean_token_accuracy": 0.87865149974823, "num_tokens": 332700541.0, "step": 8765 }, { "epoch": 1.1151253021244116, "ewc_loss": 0.10400390625, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 6.11773157119751, "learning_rate": 1e-05, "loss": 0.5168, "mean_token_accuracy": 0.8601632714271545, "num_tokens": 332737555.0, "step": 8766 }, { "epoch": 1.1152525124030022, "ewc_loss": 0.10107421875, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.885958671569824, "learning_rate": 1e-05, "loss": 0.5164, "mean_token_accuracy": 0.8583427667617798, "num_tokens": 332775957.0, "step": 8767 }, { "epoch": 1.1153797226815927, "ewc_loss": 0.10302734375, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 6.0357985496521, "learning_rate": 1e-05, "loss": 0.4734, "mean_token_accuracy": 0.8710899353027344, "num_tokens": 332812000.0, "step": 8768 }, { "epoch": 1.1155069329601832, "ewc_loss": 0.09765625, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.7538557052612305, "learning_rate": 1e-05, "loss": 0.436, "mean_token_accuracy": 0.8835513591766357, "num_tokens": 332848974.0, "step": 8769 }, { "epoch": 1.1156341432387737, "ewc_loss": 0.10400390625, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 6.059489727020264, "learning_rate": 1e-05, "loss": 0.4869, "mean_token_accuracy": 0.8718439340591431, "num_tokens": 332890362.0, "step": 8770 }, { "epoch": 1.1157613535173643, "ewc_loss": 0.095703125, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.617008686065674, "learning_rate": 1e-05, "loss": 0.4923, "mean_token_accuracy": 0.8664725422859192, "num_tokens": 332924547.0, "step": 8771 }, { "epoch": 1.1158885637959548, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.1703362464904785, "learning_rate": 1e-05, "loss": 0.4467, "mean_token_accuracy": 0.8809839487075806, "num_tokens": 332957697.0, "step": 8772 }, { "epoch": 1.1160157740745453, "ewc_loss": 0.09375, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.465365886688232, "learning_rate": 1e-05, "loss": 0.47, "mean_token_accuracy": 0.8695479035377502, "num_tokens": 332992671.0, "step": 8773 }, { "epoch": 1.1161429843531356, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.397964954376221, "learning_rate": 1e-05, "loss": 0.4757, "mean_token_accuracy": 0.8734988570213318, "num_tokens": 333033092.0, "step": 8774 }, { "epoch": 1.1162701946317262, "ewc_loss": 0.0927734375, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 7.62939453125e-05, "grad_norm": 5.330857276916504, "learning_rate": 1e-05, "loss": 0.4986, "mean_token_accuracy": 0.8626296520233154, "num_tokens": 333071631.0, "step": 8775 }, { "epoch": 1.1163974049103167, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.751021862030029, "learning_rate": 1e-05, "loss": 0.4716, "mean_token_accuracy": 0.8759112358093262, "num_tokens": 333100931.0, "step": 8776 }, { "epoch": 1.1165246151889072, "ewc_loss": 0.0947265625, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.349921226501465, "learning_rate": 1e-05, "loss": 0.4248, "mean_token_accuracy": 0.8890523910522461, "num_tokens": 333141063.0, "step": 8777 }, { "epoch": 1.1166518254674977, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 0.00011157989501953125, "grad_norm": 7.125516891479492, "learning_rate": 1e-05, "loss": 0.5194, "mean_token_accuracy": 0.8686995506286621, "num_tokens": 333180322.0, "step": 8778 }, { "epoch": 1.1167790357460883, "ewc_loss": 0.10205078125, "ewc_loss_diag": 0.00016689300537109375, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.705805778503418, "learning_rate": 1e-05, "loss": 0.5087, "mean_token_accuracy": 0.8630292415618896, "num_tokens": 333219836.0, "step": 8779 }, { "epoch": 1.1169062460246788, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.00016689300537109375, "ewc_loss_parallel": 0.00011205673217773438, "grad_norm": 7.1497650146484375, "learning_rate": 1e-05, "loss": 0.5364, "mean_token_accuracy": 0.8614922761917114, "num_tokens": 333263748.0, "step": 8780 }, { "epoch": 1.1170334563032693, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.00016689300537109375, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.110397815704346, "learning_rate": 1e-05, "loss": 0.4926, "mean_token_accuracy": 0.8690662384033203, "num_tokens": 333296110.0, "step": 8781 }, { "epoch": 1.1171606665818599, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.00016689300537109375, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.7602858543396, "learning_rate": 1e-05, "loss": 0.5089, "mean_token_accuracy": 0.8643131256103516, "num_tokens": 333328372.0, "step": 8782 }, { "epoch": 1.1172878768604504, "ewc_loss": 0.11474609375, "ewc_loss_diag": 0.00016689300537109375, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.231834888458252, "learning_rate": 1e-05, "loss": 0.4729, "mean_token_accuracy": 0.8756742477416992, "num_tokens": 333367680.0, "step": 8783 }, { "epoch": 1.117415087139041, "ewc_loss": 0.10791015625, "ewc_loss_diag": 0.00016689300537109375, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.2533488273620605, "learning_rate": 1e-05, "loss": 0.4858, "mean_token_accuracy": 0.868242621421814, "num_tokens": 333406761.0, "step": 8784 }, { "epoch": 1.1175422974176314, "ewc_loss": 0.11181640625, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.254029273986816, "learning_rate": 1e-05, "loss": 0.484, "mean_token_accuracy": 0.8720722198486328, "num_tokens": 333443268.0, "step": 8785 }, { "epoch": 1.1176695076962218, "ewc_loss": 0.099609375, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.821686744689941, "learning_rate": 1e-05, "loss": 0.5144, "mean_token_accuracy": 0.8573879599571228, "num_tokens": 333478770.0, "step": 8786 }, { "epoch": 1.1177967179748123, "ewc_loss": 0.11181640625, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.361626625061035, "learning_rate": 1e-05, "loss": 0.4463, "mean_token_accuracy": 0.8843711614608765, "num_tokens": 333514971.0, "step": 8787 }, { "epoch": 1.1179239282534028, "ewc_loss": 0.095703125, "ewc_loss_diag": 0.00016689300537109375, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.541929721832275, "learning_rate": 1e-05, "loss": 0.4791, "mean_token_accuracy": 0.8675244450569153, "num_tokens": 333553610.0, "step": 8788 }, { "epoch": 1.1180511385319933, "ewc_loss": 0.11376953125, "ewc_loss_diag": 0.00016689300537109375, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.560328960418701, "learning_rate": 1e-05, "loss": 0.4961, "mean_token_accuracy": 0.8689290285110474, "num_tokens": 333604283.0, "step": 8789 }, { "epoch": 1.1181783488105839, "ewc_loss": 0.09521484375, "ewc_loss_diag": 0.00016689300537109375, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.422509670257568, "learning_rate": 1e-05, "loss": 0.4432, "mean_token_accuracy": 0.8787232637405396, "num_tokens": 333641621.0, "step": 8790 }, { "epoch": 1.1183055590891744, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.00016689300537109375, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.84241247177124, "learning_rate": 1e-05, "loss": 0.4673, "mean_token_accuracy": 0.8784440755844116, "num_tokens": 333676435.0, "step": 8791 }, { "epoch": 1.118432769367765, "ewc_loss": 0.09765625, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.5247883796691895, "learning_rate": 1e-05, "loss": 0.4547, "mean_token_accuracy": 0.8804811239242554, "num_tokens": 333715308.0, "step": 8792 }, { "epoch": 1.1185599796463555, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 7.026921272277832, "learning_rate": 1e-05, "loss": 0.5505, "mean_token_accuracy": 0.8542672395706177, "num_tokens": 333752086.0, "step": 8793 }, { "epoch": 1.118687189924946, "ewc_loss": 0.10400390625, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 5.784869194030762, "learning_rate": 1e-05, "loss": 0.4841, "mean_token_accuracy": 0.867571234703064, "num_tokens": 333789726.0, "step": 8794 }, { "epoch": 1.1188144002035365, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.94987154006958, "learning_rate": 1e-05, "loss": 0.555, "mean_token_accuracy": 0.8574490547180176, "num_tokens": 333828518.0, "step": 8795 }, { "epoch": 1.118941610482127, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 5.9866814613342285, "learning_rate": 1e-05, "loss": 0.5186, "mean_token_accuracy": 0.8625191450119019, "num_tokens": 333870520.0, "step": 8796 }, { "epoch": 1.1190688207607176, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.694592475891113, "learning_rate": 1e-05, "loss": 0.4836, "mean_token_accuracy": 0.8724411129951477, "num_tokens": 333914676.0, "step": 8797 }, { "epoch": 1.119196031039308, "ewc_loss": 0.10888671875, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.018741130828857, "learning_rate": 1e-05, "loss": 0.4642, "mean_token_accuracy": 0.8750498294830322, "num_tokens": 333953381.0, "step": 8798 }, { "epoch": 1.1193232413178984, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.425042629241943, "learning_rate": 1e-05, "loss": 0.545, "mean_token_accuracy": 0.8566586971282959, "num_tokens": 333983275.0, "step": 8799 }, { "epoch": 1.119450451596489, "ewc_loss": 0.10595703125, "ewc_loss_diag": 0.00016689300537109375, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 5.9724507331848145, "learning_rate": 1e-05, "loss": 0.4748, "mean_token_accuracy": 0.8745418787002563, "num_tokens": 334018530.0, "step": 8800 }, { "epoch": 1.1195776618750795, "ewc_loss": 0.10498046875, "ewc_loss_diag": 0.00016689300537109375, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 6.203900337219238, "learning_rate": 1e-05, "loss": 0.4705, "mean_token_accuracy": 0.8794638514518738, "num_tokens": 334056606.0, "step": 8801 }, { "epoch": 1.11970487215367, "ewc_loss": 0.10205078125, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.879340171813965, "learning_rate": 1e-05, "loss": 0.4795, "mean_token_accuracy": 0.8721176385879517, "num_tokens": 334098541.0, "step": 8802 }, { "epoch": 1.1198320824322605, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 6.090732574462891, "learning_rate": 1e-05, "loss": 0.5069, "mean_token_accuracy": 0.859908938407898, "num_tokens": 334140243.0, "step": 8803 }, { "epoch": 1.119959292710851, "ewc_loss": 0.0986328125, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.764515399932861, "learning_rate": 1e-05, "loss": 0.452, "mean_token_accuracy": 0.8769063949584961, "num_tokens": 334177126.0, "step": 8804 }, { "epoch": 1.1200865029894416, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 6.082397937774658, "learning_rate": 1e-05, "loss": 0.4677, "mean_token_accuracy": 0.8772068023681641, "num_tokens": 334217058.0, "step": 8805 }, { "epoch": 1.120213713268032, "ewc_loss": 0.09619140625, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.652724266052246, "learning_rate": 1e-05, "loss": 0.4272, "mean_token_accuracy": 0.8864938020706177, "num_tokens": 334255060.0, "step": 8806 }, { "epoch": 1.1203409235466226, "ewc_loss": 0.10595703125, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 6.174032211303711, "learning_rate": 1e-05, "loss": 0.4918, "mean_token_accuracy": 0.8657347559928894, "num_tokens": 334290661.0, "step": 8807 }, { "epoch": 1.1204681338252132, "ewc_loss": 0.095703125, "ewc_loss_diag": 0.00016689300537109375, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.51489782333374, "learning_rate": 1e-05, "loss": 0.4487, "mean_token_accuracy": 0.8796888589859009, "num_tokens": 334328661.0, "step": 8808 }, { "epoch": 1.1205953441038037, "ewc_loss": 0.109375, "ewc_loss_diag": 0.00016689300537109375, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.357690334320068, "learning_rate": 1e-05, "loss": 0.4912, "mean_token_accuracy": 0.8695458173751831, "num_tokens": 334367535.0, "step": 8809 }, { "epoch": 1.120722554382394, "ewc_loss": 0.09423828125, "ewc_loss_diag": 0.00016689300537109375, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.3945841789245605, "learning_rate": 1e-05, "loss": 0.4596, "mean_token_accuracy": 0.8725913763046265, "num_tokens": 334407366.0, "step": 8810 }, { "epoch": 1.1208497646609845, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.00016689300537109375, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.649475574493408, "learning_rate": 1e-05, "loss": 0.5007, "mean_token_accuracy": 0.8716970086097717, "num_tokens": 334444049.0, "step": 8811 }, { "epoch": 1.120976974939575, "ewc_loss": 0.0947265625, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.383759498596191, "learning_rate": 1e-05, "loss": 0.5014, "mean_token_accuracy": 0.8620045781135559, "num_tokens": 334485405.0, "step": 8812 }, { "epoch": 1.1211041852181656, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.00016689300537109375, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.947268009185791, "learning_rate": 1e-05, "loss": 0.4275, "mean_token_accuracy": 0.892899751663208, "num_tokens": 334525209.0, "step": 8813 }, { "epoch": 1.121231395496756, "ewc_loss": 0.10009765625, "ewc_loss_diag": 0.00016689300537109375, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.5948944091796875, "learning_rate": 1e-05, "loss": 0.4774, "mean_token_accuracy": 0.8710511922836304, "num_tokens": 334569156.0, "step": 8814 }, { "epoch": 1.1213586057753466, "ewc_loss": 0.125, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 7.035673141479492, "learning_rate": 1e-05, "loss": 0.4787, "mean_token_accuracy": 0.8766665458679199, "num_tokens": 334602692.0, "step": 8815 }, { "epoch": 1.1214858160539372, "ewc_loss": 0.109375, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 5.957133769989014, "learning_rate": 1e-05, "loss": 0.5076, "mean_token_accuracy": 0.8646122217178345, "num_tokens": 334644926.0, "step": 8816 }, { "epoch": 1.1216130263325277, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.7309160232543945, "learning_rate": 1e-05, "loss": 0.4923, "mean_token_accuracy": 0.872719407081604, "num_tokens": 334690116.0, "step": 8817 }, { "epoch": 1.1217402366111182, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.139131546020508, "learning_rate": 1e-05, "loss": 0.4583, "mean_token_accuracy": 0.8827637434005737, "num_tokens": 334728403.0, "step": 8818 }, { "epoch": 1.1218674468897087, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.260310173034668, "learning_rate": 1e-05, "loss": 0.4969, "mean_token_accuracy": 0.8686550855636597, "num_tokens": 334763394.0, "step": 8819 }, { "epoch": 1.1219946571682993, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.335083961486816, "learning_rate": 1e-05, "loss": 0.4802, "mean_token_accuracy": 0.8760502934455872, "num_tokens": 334795029.0, "step": 8820 }, { "epoch": 1.1221218674468898, "ewc_loss": 0.09765625, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.730983734130859, "learning_rate": 1e-05, "loss": 0.4848, "mean_token_accuracy": 0.8664607405662537, "num_tokens": 334833298.0, "step": 8821 }, { "epoch": 1.1222490777254803, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.470704078674316, "learning_rate": 1e-05, "loss": 0.4484, "mean_token_accuracy": 0.8823677897453308, "num_tokens": 334865146.0, "step": 8822 }, { "epoch": 1.1223762880040706, "ewc_loss": 0.09375, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.41933012008667, "learning_rate": 1e-05, "loss": 0.5264, "mean_token_accuracy": 0.8554292917251587, "num_tokens": 334899392.0, "step": 8823 }, { "epoch": 1.1225034982826612, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.824331760406494, "learning_rate": 1e-05, "loss": 0.5459, "mean_token_accuracy": 0.8558979034423828, "num_tokens": 334936501.0, "step": 8824 }, { "epoch": 1.1226307085612517, "ewc_loss": 0.0966796875, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.4209794998168945, "learning_rate": 1e-05, "loss": 0.5053, "mean_token_accuracy": 0.8603965640068054, "num_tokens": 334975920.0, "step": 8825 }, { "epoch": 1.1227579188398422, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 0.0001125335693359375, "grad_norm": 7.132636547088623, "learning_rate": 1e-05, "loss": 0.5076, "mean_token_accuracy": 0.8685059547424316, "num_tokens": 335012104.0, "step": 8826 }, { "epoch": 1.1228851291184327, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 5.7603840827941895, "learning_rate": 1e-05, "loss": 0.5311, "mean_token_accuracy": 0.8543128371238708, "num_tokens": 335051302.0, "step": 8827 }, { "epoch": 1.1230123393970233, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 7.1137237548828125, "learning_rate": 1e-05, "loss": 0.5039, "mean_token_accuracy": 0.8685541749000549, "num_tokens": 335091778.0, "step": 8828 }, { "epoch": 1.1231395496756138, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.094062805175781, "learning_rate": 1e-05, "loss": 0.5362, "mean_token_accuracy": 0.8547174334526062, "num_tokens": 335120687.0, "step": 8829 }, { "epoch": 1.1232667599542043, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.798028469085693, "learning_rate": 1e-05, "loss": 0.5234, "mean_token_accuracy": 0.8660444021224976, "num_tokens": 335158166.0, "step": 8830 }, { "epoch": 1.1233939702327949, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.1278395652771, "learning_rate": 1e-05, "loss": 0.4758, "mean_token_accuracy": 0.876071035861969, "num_tokens": 335194674.0, "step": 8831 }, { "epoch": 1.1235211805113854, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.473865985870361, "learning_rate": 1e-05, "loss": 0.4792, "mean_token_accuracy": 0.8744025230407715, "num_tokens": 335229768.0, "step": 8832 }, { "epoch": 1.123648390789976, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.023140907287598, "learning_rate": 1e-05, "loss": 0.476, "mean_token_accuracy": 0.8722804188728333, "num_tokens": 335270500.0, "step": 8833 }, { "epoch": 1.1237756010685664, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.281756401062012, "learning_rate": 1e-05, "loss": 0.453, "mean_token_accuracy": 0.8793429732322693, "num_tokens": 335309672.0, "step": 8834 }, { "epoch": 1.1239028113471567, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.886810779571533, "learning_rate": 1e-05, "loss": 0.443, "mean_token_accuracy": 0.8791937232017517, "num_tokens": 335341505.0, "step": 8835 }, { "epoch": 1.1240300216257473, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.212822437286377, "learning_rate": 1e-05, "loss": 0.5275, "mean_token_accuracy": 0.8557521104812622, "num_tokens": 335379013.0, "step": 8836 }, { "epoch": 1.1241572319043378, "ewc_loss": 0.0986328125, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.73529052734375, "learning_rate": 1e-05, "loss": 0.5007, "mean_token_accuracy": 0.8605196475982666, "num_tokens": 335415170.0, "step": 8837 }, { "epoch": 1.1242844421829283, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.249086380004883, "learning_rate": 1e-05, "loss": 0.4708, "mean_token_accuracy": 0.8773536086082458, "num_tokens": 335455476.0, "step": 8838 }, { "epoch": 1.1244116524615189, "ewc_loss": 0.095703125, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.567800521850586, "learning_rate": 1e-05, "loss": 0.4832, "mean_token_accuracy": 0.865649938583374, "num_tokens": 335497185.0, "step": 8839 }, { "epoch": 1.1245388627401094, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.4232683181762695, "learning_rate": 1e-05, "loss": 0.5573, "mean_token_accuracy": 0.8534534573554993, "num_tokens": 335531754.0, "step": 8840 }, { "epoch": 1.1246660730187, "ewc_loss": 0.095703125, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.456806659698486, "learning_rate": 1e-05, "loss": 0.4528, "mean_token_accuracy": 0.8766327500343323, "num_tokens": 335570486.0, "step": 8841 }, { "epoch": 1.1247932832972904, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.673999309539795, "learning_rate": 1e-05, "loss": 0.5309, "mean_token_accuracy": 0.859932005405426, "num_tokens": 335610179.0, "step": 8842 }, { "epoch": 1.124920493575881, "ewc_loss": 0.0966796875, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.499276161193848, "learning_rate": 1e-05, "loss": 0.5289, "mean_token_accuracy": 0.8549277782440186, "num_tokens": 335639785.0, "step": 8843 }, { "epoch": 1.1250477038544715, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.861851215362549, "learning_rate": 1e-05, "loss": 0.5083, "mean_token_accuracy": 0.8684682250022888, "num_tokens": 335679012.0, "step": 8844 }, { "epoch": 1.125174914133062, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.690548419952393, "learning_rate": 1e-05, "loss": 0.4475, "mean_token_accuracy": 0.8788671493530273, "num_tokens": 335717495.0, "step": 8845 }, { "epoch": 1.1253021244116526, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.829763412475586, "learning_rate": 1e-05, "loss": 0.5044, "mean_token_accuracy": 0.8693196177482605, "num_tokens": 335751617.0, "step": 8846 }, { "epoch": 1.125429334690243, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 5.955029010772705, "learning_rate": 1e-05, "loss": 0.516, "mean_token_accuracy": 0.8596742153167725, "num_tokens": 335782698.0, "step": 8847 }, { "epoch": 1.1255565449688334, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.493710517883301, "learning_rate": 1e-05, "loss": 0.5165, "mean_token_accuracy": 0.8630543947219849, "num_tokens": 335823091.0, "step": 8848 }, { "epoch": 1.125683755247424, "ewc_loss": 0.109375, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.092551231384277, "learning_rate": 1e-05, "loss": 0.4943, "mean_token_accuracy": 0.8690573573112488, "num_tokens": 335863318.0, "step": 8849 }, { "epoch": 1.1258109655260145, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 6.025949954986572, "learning_rate": 1e-05, "loss": 0.4712, "mean_token_accuracy": 0.8726401329040527, "num_tokens": 335900467.0, "step": 8850 }, { "epoch": 1.125938175804605, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.210312843322754, "learning_rate": 1e-05, "loss": 0.5168, "mean_token_accuracy": 0.8602649569511414, "num_tokens": 335942274.0, "step": 8851 }, { "epoch": 1.1260653860831955, "ewc_loss": 0.095703125, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.594130039215088, "learning_rate": 1e-05, "loss": 0.4883, "mean_token_accuracy": 0.865161657333374, "num_tokens": 335975073.0, "step": 8852 }, { "epoch": 1.126192596361786, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.463670253753662, "learning_rate": 1e-05, "loss": 0.5439, "mean_token_accuracy": 0.8539201021194458, "num_tokens": 336011846.0, "step": 8853 }, { "epoch": 1.1263198066403766, "ewc_loss": 0.0927734375, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 7.581710815429688e-05, "grad_norm": 5.302105903625488, "learning_rate": 1e-05, "loss": 0.49, "mean_token_accuracy": 0.8635982871055603, "num_tokens": 336050969.0, "step": 8854 }, { "epoch": 1.126447016918967, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.878687858581543, "learning_rate": 1e-05, "loss": 0.4583, "mean_token_accuracy": 0.8830187320709229, "num_tokens": 336087714.0, "step": 8855 }, { "epoch": 1.1265742271975576, "ewc_loss": 0.0947265625, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.335049152374268, "learning_rate": 1e-05, "loss": 0.5162, "mean_token_accuracy": 0.8570623397827148, "num_tokens": 336124937.0, "step": 8856 }, { "epoch": 1.1267014374761481, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 0.00011730194091796875, "grad_norm": 7.2832255363464355, "learning_rate": 1e-05, "loss": 0.4733, "mean_token_accuracy": 0.8835598230361938, "num_tokens": 336167666.0, "step": 8857 }, { "epoch": 1.1268286477547387, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 5.777155876159668, "learning_rate": 1e-05, "loss": 0.4473, "mean_token_accuracy": 0.8826186656951904, "num_tokens": 336202170.0, "step": 8858 }, { "epoch": 1.126955858033329, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 0.00011587142944335938, "grad_norm": 7.288271427154541, "learning_rate": 1e-05, "loss": 0.476, "mean_token_accuracy": 0.878388524055481, "num_tokens": 336235613.0, "step": 8859 }, { "epoch": 1.1270830683119195, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.22431755065918, "learning_rate": 1e-05, "loss": 0.5079, "mean_token_accuracy": 0.864769697189331, "num_tokens": 336272027.0, "step": 8860 }, { "epoch": 1.12721027859051, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.8841423988342285, "learning_rate": 1e-05, "loss": 0.5578, "mean_token_accuracy": 0.8544815182685852, "num_tokens": 336312554.0, "step": 8861 }, { "epoch": 1.1273374888691006, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.266286849975586, "learning_rate": 1e-05, "loss": 0.5025, "mean_token_accuracy": 0.8689779043197632, "num_tokens": 336356719.0, "step": 8862 }, { "epoch": 1.127464699147691, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.493285179138184, "learning_rate": 1e-05, "loss": 0.4715, "mean_token_accuracy": 0.8782421946525574, "num_tokens": 336395242.0, "step": 8863 }, { "epoch": 1.1275919094262816, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.151548385620117, "learning_rate": 1e-05, "loss": 0.5053, "mean_token_accuracy": 0.8642924427986145, "num_tokens": 336431738.0, "step": 8864 }, { "epoch": 1.1277191197048722, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.211648464202881, "learning_rate": 1e-05, "loss": 0.4965, "mean_token_accuracy": 0.8648971319198608, "num_tokens": 336467799.0, "step": 8865 }, { "epoch": 1.1278463299834627, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 6.005756378173828, "learning_rate": 1e-05, "loss": 0.5583, "mean_token_accuracy": 0.8486505150794983, "num_tokens": 336506628.0, "step": 8866 }, { "epoch": 1.1279735402620532, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 6.070407867431641, "learning_rate": 1e-05, "loss": 0.4628, "mean_token_accuracy": 0.8765324950218201, "num_tokens": 336546034.0, "step": 8867 }, { "epoch": 1.1281007505406437, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.857902526855469, "learning_rate": 1e-05, "loss": 0.4784, "mean_token_accuracy": 0.8677055835723877, "num_tokens": 336578561.0, "step": 8868 }, { "epoch": 1.1282279608192343, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 6.052542686462402, "learning_rate": 1e-05, "loss": 0.4387, "mean_token_accuracy": 0.8824241161346436, "num_tokens": 336610500.0, "step": 8869 }, { "epoch": 1.1283551710978248, "ewc_loss": 0.09765625, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.693157196044922, "learning_rate": 1e-05, "loss": 0.46, "mean_token_accuracy": 0.8769824504852295, "num_tokens": 336653889.0, "step": 8870 }, { "epoch": 1.1284823813764153, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.130571365356445, "learning_rate": 1e-05, "loss": 0.5035, "mean_token_accuracy": 0.8655012845993042, "num_tokens": 336690605.0, "step": 8871 }, { "epoch": 1.1286095916550056, "ewc_loss": 0.095703125, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.532693862915039, "learning_rate": 1e-05, "loss": 0.4966, "mean_token_accuracy": 0.8643252849578857, "num_tokens": 336729159.0, "step": 8872 }, { "epoch": 1.1287368019335962, "ewc_loss": 0.109375, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.328178405761719, "learning_rate": 1e-05, "loss": 0.4703, "mean_token_accuracy": 0.8756096959114075, "num_tokens": 336762376.0, "step": 8873 }, { "epoch": 1.1288640122121867, "ewc_loss": 0.0927734375, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 7.581710815429688e-05, "grad_norm": 5.376829624176025, "learning_rate": 1e-05, "loss": 0.4657, "mean_token_accuracy": 0.8752503395080566, "num_tokens": 336797060.0, "step": 8874 }, { "epoch": 1.1289912224907772, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.662675857543945, "learning_rate": 1e-05, "loss": 0.5481, "mean_token_accuracy": 0.8498145937919617, "num_tokens": 336838686.0, "step": 8875 }, { "epoch": 1.1291184327693677, "ewc_loss": 0.0947265625, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.319821357727051, "learning_rate": 1e-05, "loss": 0.5327, "mean_token_accuracy": 0.852247953414917, "num_tokens": 336887535.0, "step": 8876 }, { "epoch": 1.1292456430479583, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 7.092577934265137, "learning_rate": 1e-05, "loss": 0.4851, "mean_token_accuracy": 0.8747580051422119, "num_tokens": 336920167.0, "step": 8877 }, { "epoch": 1.1293728533265488, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.588150978088379, "learning_rate": 1e-05, "loss": 0.4712, "mean_token_accuracy": 0.8712390661239624, "num_tokens": 336953925.0, "step": 8878 }, { "epoch": 1.1295000636051393, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 0.00011682510375976562, "grad_norm": 7.278353214263916, "learning_rate": 1e-05, "loss": 0.4996, "mean_token_accuracy": 0.8734809160232544, "num_tokens": 336992407.0, "step": 8879 }, { "epoch": 1.1296272738837299, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.068840980529785, "learning_rate": 1e-05, "loss": 0.537, "mean_token_accuracy": 0.8549215793609619, "num_tokens": 337029495.0, "step": 8880 }, { "epoch": 1.1297544841623204, "ewc_loss": 0.125, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 6.970748424530029, "learning_rate": 1e-05, "loss": 0.5193, "mean_token_accuracy": 0.8651545643806458, "num_tokens": 337067533.0, "step": 8881 }, { "epoch": 1.129881694440911, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.278872966766357, "learning_rate": 1e-05, "loss": 0.5424, "mean_token_accuracy": 0.8556876182556152, "num_tokens": 337105112.0, "step": 8882 }, { "epoch": 1.1300089047195012, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.489021301269531, "learning_rate": 1e-05, "loss": 0.5307, "mean_token_accuracy": 0.8581599593162537, "num_tokens": 337140807.0, "step": 8883 }, { "epoch": 1.1301361149980917, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.255778789520264, "learning_rate": 1e-05, "loss": 0.4401, "mean_token_accuracy": 0.885657548904419, "num_tokens": 337182644.0, "step": 8884 }, { "epoch": 1.1302633252766823, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 6.036182880401611, "learning_rate": 1e-05, "loss": 0.4781, "mean_token_accuracy": 0.8711190223693848, "num_tokens": 337226868.0, "step": 8885 }, { "epoch": 1.1303905355552728, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.283130645751953, "learning_rate": 1e-05, "loss": 0.5317, "mean_token_accuracy": 0.8585110306739807, "num_tokens": 337262118.0, "step": 8886 }, { "epoch": 1.1305177458338633, "ewc_loss": 0.09765625, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.705865383148193, "learning_rate": 1e-05, "loss": 0.5401, "mean_token_accuracy": 0.8493406772613525, "num_tokens": 337299575.0, "step": 8887 }, { "epoch": 1.1306449561124539, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.384666442871094, "learning_rate": 1e-05, "loss": 0.4852, "mean_token_accuracy": 0.8708053827285767, "num_tokens": 337335178.0, "step": 8888 }, { "epoch": 1.1307721663910444, "ewc_loss": 0.095703125, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.530438423156738, "learning_rate": 1e-05, "loss": 0.5537, "mean_token_accuracy": 0.8530584573745728, "num_tokens": 337368982.0, "step": 8889 }, { "epoch": 1.130899376669635, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.551844596862793, "learning_rate": 1e-05, "loss": 0.5221, "mean_token_accuracy": 0.8617863655090332, "num_tokens": 337403708.0, "step": 8890 }, { "epoch": 1.1310265869482254, "ewc_loss": 0.0947265625, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.44052791595459, "learning_rate": 1e-05, "loss": 0.4758, "mean_token_accuracy": 0.8722343444824219, "num_tokens": 337441265.0, "step": 8891 }, { "epoch": 1.131153797226816, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.8037614822387695, "learning_rate": 1e-05, "loss": 0.4916, "mean_token_accuracy": 0.8715322017669678, "num_tokens": 337475600.0, "step": 8892 }, { "epoch": 1.1312810075054065, "ewc_loss": 0.09765625, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.509396553039551, "learning_rate": 1e-05, "loss": 0.4405, "mean_token_accuracy": 0.8793998956680298, "num_tokens": 337513051.0, "step": 8893 }, { "epoch": 1.131408217783997, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 7.002248287200928, "learning_rate": 1e-05, "loss": 0.5653, "mean_token_accuracy": 0.8579387664794922, "num_tokens": 337552187.0, "step": 8894 }, { "epoch": 1.1315354280625876, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.7440571784973145, "learning_rate": 1e-05, "loss": 0.4425, "mean_token_accuracy": 0.8825405836105347, "num_tokens": 337586877.0, "step": 8895 }, { "epoch": 1.131662638341178, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.956765651702881, "learning_rate": 1e-05, "loss": 0.4979, "mean_token_accuracy": 0.8685096502304077, "num_tokens": 337627979.0, "step": 8896 }, { "epoch": 1.1317898486197684, "ewc_loss": 0.109375, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 5.960785865783691, "learning_rate": 1e-05, "loss": 0.4884, "mean_token_accuracy": 0.8707838654518127, "num_tokens": 337670064.0, "step": 8897 }, { "epoch": 1.131917058898359, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.733401298522949, "learning_rate": 1e-05, "loss": 0.6062, "mean_token_accuracy": 0.8401584029197693, "num_tokens": 337709670.0, "step": 8898 }, { "epoch": 1.1320442691769494, "ewc_loss": 0.109375, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.038110733032227, "learning_rate": 1e-05, "loss": 0.4983, "mean_token_accuracy": 0.866390585899353, "num_tokens": 337747507.0, "step": 8899 }, { "epoch": 1.13217147945554, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.421848297119141, "learning_rate": 1e-05, "loss": 0.5104, "mean_token_accuracy": 0.8641207218170166, "num_tokens": 337789542.0, "step": 8900 }, { "epoch": 1.1322986897341305, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.00016880035400390625, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.014094829559326, "learning_rate": 1e-05, "loss": 0.4435, "mean_token_accuracy": 0.8811583518981934, "num_tokens": 337832100.0, "step": 8901 }, { "epoch": 1.132425900012721, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.00016880035400390625, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 6.152384281158447, "learning_rate": 1e-05, "loss": 0.4747, "mean_token_accuracy": 0.8691209554672241, "num_tokens": 337869395.0, "step": 8902 }, { "epoch": 1.1325531102913116, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.00016880035400390625, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.962864875793457, "learning_rate": 1e-05, "loss": 0.5306, "mean_token_accuracy": 0.8526023030281067, "num_tokens": 337909644.0, "step": 8903 }, { "epoch": 1.132680320569902, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.00016880035400390625, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.998488426208496, "learning_rate": 1e-05, "loss": 0.5222, "mean_token_accuracy": 0.8546823263168335, "num_tokens": 337947255.0, "step": 8904 }, { "epoch": 1.1328075308484926, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.00016880035400390625, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.876857280731201, "learning_rate": 1e-05, "loss": 0.4532, "mean_token_accuracy": 0.8805829286575317, "num_tokens": 337986377.0, "step": 8905 }, { "epoch": 1.1329347411270831, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.00016880035400390625, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.9248881340026855, "learning_rate": 1e-05, "loss": 0.4551, "mean_token_accuracy": 0.8746792078018188, "num_tokens": 338025234.0, "step": 8906 }, { "epoch": 1.1330619514056737, "ewc_loss": 0.0986328125, "ewc_loss_diag": 0.00016880035400390625, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.790421485900879, "learning_rate": 1e-05, "loss": 0.5178, "mean_token_accuracy": 0.8571414947509766, "num_tokens": 338069070.0, "step": 8907 }, { "epoch": 1.133189161684264, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.00016880035400390625, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.917279243469238, "learning_rate": 1e-05, "loss": 0.4696, "mean_token_accuracy": 0.8724068403244019, "num_tokens": 338111529.0, "step": 8908 }, { "epoch": 1.1333163719628545, "ewc_loss": 0.09765625, "ewc_loss_diag": 0.00016880035400390625, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.702569007873535, "learning_rate": 1e-05, "loss": 0.5145, "mean_token_accuracy": 0.8569834232330322, "num_tokens": 338155881.0, "step": 8909 }, { "epoch": 1.133443582241445, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.00016880035400390625, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.977915287017822, "learning_rate": 1e-05, "loss": 0.506, "mean_token_accuracy": 0.8612669110298157, "num_tokens": 338192570.0, "step": 8910 }, { "epoch": 1.1335707925200356, "ewc_loss": 0.0966796875, "ewc_loss_diag": 0.00016880035400390625, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.5840983390808105, "learning_rate": 1e-05, "loss": 0.4868, "mean_token_accuracy": 0.8639113903045654, "num_tokens": 338225088.0, "step": 8911 }, { "epoch": 1.133698002798626, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.00016880035400390625, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.114648818969727, "learning_rate": 1e-05, "loss": 0.5366, "mean_token_accuracy": 0.8552601933479309, "num_tokens": 338267301.0, "step": 8912 }, { "epoch": 1.1338252130772166, "ewc_loss": 0.0927734375, "ewc_loss_diag": 0.00016880035400390625, "ewc_loss_parallel": 7.62939453125e-05, "grad_norm": 5.432567596435547, "learning_rate": 1e-05, "loss": 0.4507, "mean_token_accuracy": 0.8743464946746826, "num_tokens": 338305130.0, "step": 8913 }, { "epoch": 1.1339524233558071, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.00016880035400390625, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.3516130447387695, "learning_rate": 1e-05, "loss": 0.5391, "mean_token_accuracy": 0.8504809141159058, "num_tokens": 338347556.0, "step": 8914 }, { "epoch": 1.1340796336343977, "ewc_loss": 0.0927734375, "ewc_loss_diag": 0.00016880035400390625, "ewc_loss_parallel": 7.581710815429688e-05, "grad_norm": 5.275869846343994, "learning_rate": 1e-05, "loss": 0.492, "mean_token_accuracy": 0.867906928062439, "num_tokens": 338384994.0, "step": 8915 }, { "epoch": 1.1342068439129882, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.00016880035400390625, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.775669097900391, "learning_rate": 1e-05, "loss": 0.5401, "mean_token_accuracy": 0.858121395111084, "num_tokens": 338427311.0, "step": 8916 }, { "epoch": 1.1343340541915787, "ewc_loss": 0.09375, "ewc_loss_diag": 0.00016880035400390625, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.223172187805176, "learning_rate": 1e-05, "loss": 0.4493, "mean_token_accuracy": 0.8759474754333496, "num_tokens": 338464091.0, "step": 8917 }, { "epoch": 1.1344612644701693, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.00016880035400390625, "ewc_loss_parallel": 0.00011682510375976562, "grad_norm": 7.309182167053223, "learning_rate": 1e-05, "loss": 0.5157, "mean_token_accuracy": 0.8661279678344727, "num_tokens": 338507961.0, "step": 8918 }, { "epoch": 1.1345884747487598, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.0001697540283203125, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 5.634551048278809, "learning_rate": 1e-05, "loss": 0.4837, "mean_token_accuracy": 0.8688145875930786, "num_tokens": 338544028.0, "step": 8919 }, { "epoch": 1.1347156850273503, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.0001697540283203125, "ewc_loss_parallel": 0.00012111663818359375, "grad_norm": 7.519643783569336, "learning_rate": 1e-05, "loss": 0.5823, "mean_token_accuracy": 0.853023886680603, "num_tokens": 338587827.0, "step": 8920 }, { "epoch": 1.1348428953059406, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.0001697540283203125, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.267001628875732, "learning_rate": 1e-05, "loss": 0.496, "mean_token_accuracy": 0.871030330657959, "num_tokens": 338627729.0, "step": 8921 }, { "epoch": 1.1349701055845312, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0001697540283203125, "ewc_loss_parallel": 0.0001125335693359375, "grad_norm": 7.126448154449463, "learning_rate": 1e-05, "loss": 0.4711, "mean_token_accuracy": 0.8802966475486755, "num_tokens": 338663805.0, "step": 8922 }, { "epoch": 1.1350973158631217, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.0001697540283203125, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.418063640594482, "learning_rate": 1e-05, "loss": 0.5193, "mean_token_accuracy": 0.8674938678741455, "num_tokens": 338702204.0, "step": 8923 }, { "epoch": 1.1352245261417122, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.0001697540283203125, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.654351234436035, "learning_rate": 1e-05, "loss": 0.4907, "mean_token_accuracy": 0.8713464736938477, "num_tokens": 338748769.0, "step": 8924 }, { "epoch": 1.1353517364203027, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.0001697540283203125, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.270955562591553, "learning_rate": 1e-05, "loss": 0.478, "mean_token_accuracy": 0.8754652738571167, "num_tokens": 338793617.0, "step": 8925 }, { "epoch": 1.1354789466988933, "ewc_loss": 0.10888671875, "ewc_loss_diag": 0.0001697540283203125, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.30439567565918, "learning_rate": 1e-05, "loss": 0.4825, "mean_token_accuracy": 0.8731217384338379, "num_tokens": 338827893.0, "step": 8926 }, { "epoch": 1.1356061569774838, "ewc_loss": 0.10986328125, "ewc_loss_diag": 0.0001697540283203125, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.142346382141113, "learning_rate": 1e-05, "loss": 0.4881, "mean_token_accuracy": 0.8717074394226074, "num_tokens": 338866036.0, "step": 8927 }, { "epoch": 1.1357333672560743, "ewc_loss": 0.10400390625, "ewc_loss_diag": 0.0001697540283203125, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 6.066897392272949, "learning_rate": 1e-05, "loss": 0.5079, "mean_token_accuracy": 0.86211758852005, "num_tokens": 338900595.0, "step": 8928 }, { "epoch": 1.1358605775346649, "ewc_loss": 0.10400390625, "ewc_loss_diag": 0.0001697540283203125, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 6.030267715454102, "learning_rate": 1e-05, "loss": 0.4342, "mean_token_accuracy": 0.883872389793396, "num_tokens": 338930268.0, "step": 8929 }, { "epoch": 1.1359877878132554, "ewc_loss": 0.10205078125, "ewc_loss_diag": 0.0001697540283203125, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.923830986022949, "learning_rate": 1e-05, "loss": 0.487, "mean_token_accuracy": 0.868211567401886, "num_tokens": 338974633.0, "step": 8930 }, { "epoch": 1.136114998091846, "ewc_loss": 0.10205078125, "ewc_loss_diag": 0.0001697540283203125, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.927929401397705, "learning_rate": 1e-05, "loss": 0.5168, "mean_token_accuracy": 0.8570214509963989, "num_tokens": 339008884.0, "step": 8931 }, { "epoch": 1.1362422083704362, "ewc_loss": 0.10205078125, "ewc_loss_diag": 0.0001697540283203125, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.863337516784668, "learning_rate": 1e-05, "loss": 0.5651, "mean_token_accuracy": 0.842951774597168, "num_tokens": 339043923.0, "step": 8932 }, { "epoch": 1.1363694186490267, "ewc_loss": 0.10107421875, "ewc_loss_diag": 0.0001697540283203125, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.842584133148193, "learning_rate": 1e-05, "loss": 0.4762, "mean_token_accuracy": 0.8728986978530884, "num_tokens": 339080545.0, "step": 8933 }, { "epoch": 1.1364966289276173, "ewc_loss": 0.10009765625, "ewc_loss_diag": 0.0001697540283203125, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.834564208984375, "learning_rate": 1e-05, "loss": 0.4945, "mean_token_accuracy": 0.8675370812416077, "num_tokens": 339114243.0, "step": 8934 }, { "epoch": 1.1366238392062078, "ewc_loss": 0.09912109375, "ewc_loss_diag": 0.0001697540283203125, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.765143871307373, "learning_rate": 1e-05, "loss": 0.4668, "mean_token_accuracy": 0.8753683567047119, "num_tokens": 339147609.0, "step": 8935 }, { "epoch": 1.1367510494847983, "ewc_loss": 0.099609375, "ewc_loss_diag": 0.0001697540283203125, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.843430042266846, "learning_rate": 1e-05, "loss": 0.4579, "mean_token_accuracy": 0.8762078881263733, "num_tokens": 339186295.0, "step": 8936 }, { "epoch": 1.1368782597633889, "ewc_loss": 0.09765625, "ewc_loss_diag": 0.0001697540283203125, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.691427230834961, "learning_rate": 1e-05, "loss": 0.469, "mean_token_accuracy": 0.8723915219306946, "num_tokens": 339217572.0, "step": 8937 }, { "epoch": 1.1370054700419794, "ewc_loss": 0.10107421875, "ewc_loss_diag": 0.0001697540283203125, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.892404079437256, "learning_rate": 1e-05, "loss": 0.4729, "mean_token_accuracy": 0.8715771436691284, "num_tokens": 339257305.0, "step": 8938 }, { "epoch": 1.13713268032057, "ewc_loss": 0.0966796875, "ewc_loss_diag": 0.0001697540283203125, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.592166423797607, "learning_rate": 1e-05, "loss": 0.4918, "mean_token_accuracy": 0.8631606101989746, "num_tokens": 339295117.0, "step": 8939 }, { "epoch": 1.1372598905991604, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.0001697540283203125, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 6.008086204528809, "learning_rate": 1e-05, "loss": 0.486, "mean_token_accuracy": 0.8699659109115601, "num_tokens": 339330543.0, "step": 8940 }, { "epoch": 1.137387100877751, "ewc_loss": 0.09375, "ewc_loss_diag": 0.0001697540283203125, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.444250583648682, "learning_rate": 1e-05, "loss": 0.4806, "mean_token_accuracy": 0.8685263395309448, "num_tokens": 339375040.0, "step": 8941 }, { "epoch": 1.1375143111563415, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.0001697540283203125, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.242003440856934, "learning_rate": 1e-05, "loss": 0.5004, "mean_token_accuracy": 0.8621929883956909, "num_tokens": 339410862.0, "step": 8942 }, { "epoch": 1.137641521434932, "ewc_loss": 0.091796875, "ewc_loss_diag": 0.0001697540283203125, "ewc_loss_parallel": 7.486343383789062e-05, "grad_norm": 5.250322341918945, "learning_rate": 1e-05, "loss": 0.4589, "mean_token_accuracy": 0.8717581033706665, "num_tokens": 339441929.0, "step": 8943 }, { "epoch": 1.1377687317135226, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.0001697540283203125, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.684967041015625, "learning_rate": 1e-05, "loss": 0.4996, "mean_token_accuracy": 0.8701639175415039, "num_tokens": 339486156.0, "step": 8944 }, { "epoch": 1.137895941992113, "ewc_loss": 0.09033203125, "ewc_loss_diag": 0.0001697540283203125, "ewc_loss_parallel": 7.343292236328125e-05, "grad_norm": 5.075816631317139, "learning_rate": 1e-05, "loss": 0.5155, "mean_token_accuracy": 0.8544771671295166, "num_tokens": 339531676.0, "step": 8945 }, { "epoch": 1.1380231522707034, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0001697540283203125, "ewc_loss_parallel": 0.00011968612670898438, "grad_norm": 7.42986536026001, "learning_rate": 1e-05, "loss": 0.56, "mean_token_accuracy": 0.8573601245880127, "num_tokens": 339571003.0, "step": 8946 }, { "epoch": 1.138150362549294, "ewc_loss": 0.10205078125, "ewc_loss_diag": 0.0001697540283203125, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.461354732513428, "learning_rate": 1e-05, "loss": 0.4812, "mean_token_accuracy": 0.863820493221283, "num_tokens": 339604389.0, "step": 8947 }, { "epoch": 1.1382775728278844, "ewc_loss": 0.1513671875, "ewc_loss_diag": 0.0001697540283203125, "ewc_loss_parallel": 0.00013446807861328125, "grad_norm": 7.969017028808594, "learning_rate": 1e-05, "loss": 0.5796, "mean_token_accuracy": 0.8551089763641357, "num_tokens": 339649267.0, "step": 8948 }, { "epoch": 1.138404783106475, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.0001697540283203125, "ewc_loss_parallel": 0.00011348724365234375, "grad_norm": 6.500438213348389, "learning_rate": 1e-05, "loss": 0.5317, "mean_token_accuracy": 0.8588979840278625, "num_tokens": 339685564.0, "step": 8949 }, { "epoch": 1.1385319933850655, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.0001697540283203125, "ewc_loss_parallel": 0.00012063980102539062, "grad_norm": 7.505774021148682, "learning_rate": 1e-05, "loss": 0.474, "mean_token_accuracy": 0.8824029564857483, "num_tokens": 339716312.0, "step": 8950 }, { "epoch": 1.138659203663656, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0001697540283203125, "ewc_loss_parallel": 0.00011396408081054688, "grad_norm": 6.753975868225098, "learning_rate": 1e-05, "loss": 0.5217, "mean_token_accuracy": 0.8688809275627136, "num_tokens": 339760666.0, "step": 8951 }, { "epoch": 1.1387864139422466, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.0001697540283203125, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.964603900909424, "learning_rate": 1e-05, "loss": 0.4867, "mean_token_accuracy": 0.8753258585929871, "num_tokens": 339799432.0, "step": 8952 }, { "epoch": 1.138913624220837, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.0001697540283203125, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.468194007873535, "learning_rate": 1e-05, "loss": 0.5195, "mean_token_accuracy": 0.8639304637908936, "num_tokens": 339839787.0, "step": 8953 }, { "epoch": 1.1390408344994276, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.0001697540283203125, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.658862590789795, "learning_rate": 1e-05, "loss": 0.4955, "mean_token_accuracy": 0.8685788512229919, "num_tokens": 339876193.0, "step": 8954 }, { "epoch": 1.1391680447780181, "ewc_loss": 0.10888671875, "ewc_loss_diag": 0.0001697540283203125, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.181676864624023, "learning_rate": 1e-05, "loss": 0.4987, "mean_token_accuracy": 0.866612434387207, "num_tokens": 339919791.0, "step": 8955 }, { "epoch": 1.1392952550566087, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.0001697540283203125, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.491822719573975, "learning_rate": 1e-05, "loss": 0.5189, "mean_token_accuracy": 0.8579099178314209, "num_tokens": 339954432.0, "step": 8956 }, { "epoch": 1.139422465335199, "ewc_loss": 0.10302734375, "ewc_loss_diag": 0.0001697540283203125, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.9399213790893555, "learning_rate": 1e-05, "loss": 0.4829, "mean_token_accuracy": 0.8722551465034485, "num_tokens": 339996327.0, "step": 8957 }, { "epoch": 1.1395496756137895, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.0001697540283203125, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.441190242767334, "learning_rate": 1e-05, "loss": 0.4163, "mean_token_accuracy": 0.8973987102508545, "num_tokens": 340031604.0, "step": 8958 }, { "epoch": 1.13967688589238, "ewc_loss": 0.099609375, "ewc_loss_diag": 0.0001697540283203125, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.761014461517334, "learning_rate": 1e-05, "loss": 0.479, "mean_token_accuracy": 0.8683425188064575, "num_tokens": 340072433.0, "step": 8959 }, { "epoch": 1.1398040961709706, "ewc_loss": 0.11181640625, "ewc_loss_diag": 0.0001697540283203125, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.524735927581787, "learning_rate": 1e-05, "loss": 0.5112, "mean_token_accuracy": 0.8670741319656372, "num_tokens": 340114306.0, "step": 8960 }, { "epoch": 1.139931306449561, "ewc_loss": 0.095703125, "ewc_loss_diag": 0.0001697540283203125, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.596806526184082, "learning_rate": 1e-05, "loss": 0.5481, "mean_token_accuracy": 0.8455806374549866, "num_tokens": 340152826.0, "step": 8961 }, { "epoch": 1.1400585167281516, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.0001697540283203125, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.742593288421631, "learning_rate": 1e-05, "loss": 0.5303, "mean_token_accuracy": 0.8610957264900208, "num_tokens": 340191110.0, "step": 8962 }, { "epoch": 1.1401857270067421, "ewc_loss": 0.09619140625, "ewc_loss_diag": 0.0001697540283203125, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.553116321563721, "learning_rate": 1e-05, "loss": 0.4804, "mean_token_accuracy": 0.8675899505615234, "num_tokens": 340230606.0, "step": 8963 }, { "epoch": 1.1403129372853327, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.0001697540283203125, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 7.031946659088135, "learning_rate": 1e-05, "loss": 0.5307, "mean_token_accuracy": 0.8602569103240967, "num_tokens": 340274831.0, "step": 8964 }, { "epoch": 1.1404401475639232, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.0001697540283203125, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.766503810882568, "learning_rate": 1e-05, "loss": 0.4754, "mean_token_accuracy": 0.8692289590835571, "num_tokens": 340308245.0, "step": 8965 }, { "epoch": 1.1405673578425137, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0001697540283203125, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 7.1002655029296875, "learning_rate": 1e-05, "loss": 0.4556, "mean_token_accuracy": 0.888716459274292, "num_tokens": 340343091.0, "step": 8966 }, { "epoch": 1.1406945681211043, "ewc_loss": 0.10888671875, "ewc_loss_diag": 0.0001697540283203125, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.108630180358887, "learning_rate": 1e-05, "loss": 0.4952, "mean_token_accuracy": 0.8711900115013123, "num_tokens": 340382968.0, "step": 8967 }, { "epoch": 1.1408217783996948, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.0001697540283203125, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.764206886291504, "learning_rate": 1e-05, "loss": 0.4737, "mean_token_accuracy": 0.8758949041366577, "num_tokens": 340421740.0, "step": 8968 }, { "epoch": 1.1409489886782853, "ewc_loss": 0.11083984375, "ewc_loss_diag": 0.0001697540283203125, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.271065711975098, "learning_rate": 1e-05, "loss": 0.5007, "mean_token_accuracy": 0.8671494722366333, "num_tokens": 340460156.0, "step": 8969 }, { "epoch": 1.1410761989568756, "ewc_loss": 0.10693359375, "ewc_loss_diag": 0.0001697540283203125, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.2581400871276855, "learning_rate": 1e-05, "loss": 0.4935, "mean_token_accuracy": 0.8696479797363281, "num_tokens": 340496106.0, "step": 8970 }, { "epoch": 1.1412034092354661, "ewc_loss": 0.11083984375, "ewc_loss_diag": 0.0001697540283203125, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.314399242401123, "learning_rate": 1e-05, "loss": 0.4819, "mean_token_accuracy": 0.8711005449295044, "num_tokens": 340530338.0, "step": 8971 }, { "epoch": 1.1413306195140567, "ewc_loss": 0.09619140625, "ewc_loss_diag": 0.0001697540283203125, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.760871410369873, "learning_rate": 1e-05, "loss": 0.4744, "mean_token_accuracy": 0.8703889846801758, "num_tokens": 340565590.0, "step": 8972 }, { "epoch": 1.1414578297926472, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.0001697540283203125, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.459768295288086, "learning_rate": 1e-05, "loss": 0.4904, "mean_token_accuracy": 0.8707031011581421, "num_tokens": 340609651.0, "step": 8973 }, { "epoch": 1.1415850400712377, "ewc_loss": 0.091796875, "ewc_loss_diag": 0.0001697540283203125, "ewc_loss_parallel": 7.486343383789062e-05, "grad_norm": 5.413540840148926, "learning_rate": 1e-05, "loss": 0.4749, "mean_token_accuracy": 0.8680749535560608, "num_tokens": 340650119.0, "step": 8974 }, { "epoch": 1.1417122503498283, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.0001697540283203125, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.7931413650512695, "learning_rate": 1e-05, "loss": 0.4934, "mean_token_accuracy": 0.8682312369346619, "num_tokens": 340686734.0, "step": 8975 }, { "epoch": 1.1418394606284188, "ewc_loss": 0.09326171875, "ewc_loss_diag": 0.0001697540283203125, "ewc_loss_parallel": 7.62939453125e-05, "grad_norm": 5.323939800262451, "learning_rate": 1e-05, "loss": 0.4474, "mean_token_accuracy": 0.8766912221908569, "num_tokens": 340722405.0, "step": 8976 }, { "epoch": 1.1419666709070093, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0001697540283203125, "ewc_loss_parallel": 0.0001125335693359375, "grad_norm": 7.237942695617676, "learning_rate": 1e-05, "loss": 0.4882, "mean_token_accuracy": 0.8774895668029785, "num_tokens": 340763402.0, "step": 8977 }, { "epoch": 1.1420938811855998, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.0001697540283203125, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.6659159660339355, "learning_rate": 1e-05, "loss": 0.4062, "mean_token_accuracy": 0.8948082327842712, "num_tokens": 340799410.0, "step": 8978 }, { "epoch": 1.1422210914641904, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0001697540283203125, "ewc_loss_parallel": 0.000118255615234375, "grad_norm": 7.419734001159668, "learning_rate": 1e-05, "loss": 0.5446, "mean_token_accuracy": 0.8588300943374634, "num_tokens": 340832970.0, "step": 8979 }, { "epoch": 1.142348301742781, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0001697540283203125, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.231447219848633, "learning_rate": 1e-05, "loss": 0.4813, "mean_token_accuracy": 0.8749840259552002, "num_tokens": 340872500.0, "step": 8980 }, { "epoch": 1.1424755120213712, "ewc_loss": 0.125, "ewc_loss_diag": 0.0001697540283203125, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 7.023784160614014, "learning_rate": 1e-05, "loss": 0.5214, "mean_token_accuracy": 0.8632850646972656, "num_tokens": 340913784.0, "step": 8981 }, { "epoch": 1.1426027222999617, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0001697540283203125, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.38359260559082, "learning_rate": 1e-05, "loss": 0.4574, "mean_token_accuracy": 0.8798651695251465, "num_tokens": 340945660.0, "step": 8982 }, { "epoch": 1.1427299325785523, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.0001697540283203125, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.571276664733887, "learning_rate": 1e-05, "loss": 0.5071, "mean_token_accuracy": 0.8659674525260925, "num_tokens": 340979528.0, "step": 8983 }, { "epoch": 1.1428571428571428, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.0001697540283203125, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.25987434387207, "learning_rate": 1e-05, "loss": 0.4989, "mean_token_accuracy": 0.8667492866516113, "num_tokens": 341016256.0, "step": 8984 }, { "epoch": 1.1429843531357333, "ewc_loss": 0.10693359375, "ewc_loss_diag": 0.0001697540283203125, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.225760459899902, "learning_rate": 1e-05, "loss": 0.528, "mean_token_accuracy": 0.8577994108200073, "num_tokens": 341056339.0, "step": 8985 }, { "epoch": 1.1431115634143239, "ewc_loss": 0.10595703125, "ewc_loss_diag": 0.0001697540283203125, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.10791015625, "learning_rate": 1e-05, "loss": 0.473, "mean_token_accuracy": 0.8753606081008911, "num_tokens": 341093060.0, "step": 8986 }, { "epoch": 1.1432387736929144, "ewc_loss": 0.10302734375, "ewc_loss_diag": 0.0001697540283203125, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 6.056579113006592, "learning_rate": 1e-05, "loss": 0.515, "mean_token_accuracy": 0.8617396354675293, "num_tokens": 341128634.0, "step": 8987 }, { "epoch": 1.143365983971505, "ewc_loss": 0.10205078125, "ewc_loss_diag": 0.0001697540283203125, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.964476108551025, "learning_rate": 1e-05, "loss": 0.5362, "mean_token_accuracy": 0.8528586626052856, "num_tokens": 341160649.0, "step": 8988 }, { "epoch": 1.1434931942500954, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.0001697540283203125, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.983935832977295, "learning_rate": 1e-05, "loss": 0.4823, "mean_token_accuracy": 0.8672055006027222, "num_tokens": 341196660.0, "step": 8989 }, { "epoch": 1.143620404528686, "ewc_loss": 0.099609375, "ewc_loss_diag": 0.0001697540283203125, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.801212787628174, "learning_rate": 1e-05, "loss": 0.5222, "mean_token_accuracy": 0.8560574054718018, "num_tokens": 341236704.0, "step": 8990 }, { "epoch": 1.1437476148072765, "ewc_loss": 0.10400390625, "ewc_loss_diag": 0.0001697540283203125, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 6.0103535652160645, "learning_rate": 1e-05, "loss": 0.4972, "mean_token_accuracy": 0.8639398813247681, "num_tokens": 341272874.0, "step": 8991 }, { "epoch": 1.143874825085867, "ewc_loss": 0.095703125, "ewc_loss_diag": 0.0001697540283203125, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.648094654083252, "learning_rate": 1e-05, "loss": 0.4999, "mean_token_accuracy": 0.861713171005249, "num_tokens": 341310209.0, "step": 8992 }, { "epoch": 1.1440020353644575, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.0001697540283203125, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 6.127824306488037, "learning_rate": 1e-05, "loss": 0.4475, "mean_token_accuracy": 0.8803009986877441, "num_tokens": 341341717.0, "step": 8993 }, { "epoch": 1.144129245643048, "ewc_loss": 0.09423828125, "ewc_loss_diag": 0.0001697540283203125, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.477567672729492, "learning_rate": 1e-05, "loss": 0.4197, "mean_token_accuracy": 0.8860755562782288, "num_tokens": 341374298.0, "step": 8994 }, { "epoch": 1.1442564559216384, "ewc_loss": 0.10888671875, "ewc_loss_diag": 0.0001697540283203125, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.353092193603516, "learning_rate": 1e-05, "loss": 0.5014, "mean_token_accuracy": 0.8670501708984375, "num_tokens": 341416214.0, "step": 8995 }, { "epoch": 1.144383666200229, "ewc_loss": 0.0908203125, "ewc_loss_diag": 0.0001697540283203125, "ewc_loss_parallel": 7.390975952148438e-05, "grad_norm": 5.306440353393555, "learning_rate": 1e-05, "loss": 0.4597, "mean_token_accuracy": 0.8703327178955078, "num_tokens": 341457419.0, "step": 8996 }, { "epoch": 1.1445108764788194, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.000171661376953125, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.764181613922119, "learning_rate": 1e-05, "loss": 0.4839, "mean_token_accuracy": 0.8746026754379272, "num_tokens": 341493251.0, "step": 8997 }, { "epoch": 1.14463808675741, "ewc_loss": 0.09375, "ewc_loss_diag": 0.000171661376953125, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.27444314956665, "learning_rate": 1e-05, "loss": 0.5298, "mean_token_accuracy": 0.8512803316116333, "num_tokens": 341531501.0, "step": 8998 }, { "epoch": 1.1447652970360005, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.000171661376953125, "ewc_loss_parallel": 0.00011682510375976562, "grad_norm": 7.2735090255737305, "learning_rate": 1e-05, "loss": 0.4607, "mean_token_accuracy": 0.8836138844490051, "num_tokens": 341572114.0, "step": 8999 }, { "epoch": 1.144892507314591, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.000171661376953125, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.6376848220825195, "learning_rate": 1e-05, "loss": 0.5026, "mean_token_accuracy": 0.8626258373260498, "num_tokens": 341617446.0, "step": 9000 }, { "epoch": 1.1450197175931816, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.000171661376953125, "ewc_loss_parallel": 0.0001220703125, "grad_norm": 7.494638919830322, "learning_rate": 1e-05, "loss": 0.5354, "mean_token_accuracy": 0.8623831272125244, "num_tokens": 341653636.0, "step": 9001 }, { "epoch": 1.145146927871772, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.00017070770263671875, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.27545690536499, "learning_rate": 1e-05, "loss": 0.6338, "mean_token_accuracy": 0.8309019804000854, "num_tokens": 341692174.0, "step": 9002 }, { "epoch": 1.1452741381503626, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.00017070770263671875, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 7.07089900970459, "learning_rate": 1e-05, "loss": 0.5302, "mean_token_accuracy": 0.8610350489616394, "num_tokens": 341737215.0, "step": 9003 }, { "epoch": 1.1454013484289531, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.00017070770263671875, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.432114124298096, "learning_rate": 1e-05, "loss": 0.5227, "mean_token_accuracy": 0.8639657497406006, "num_tokens": 341778571.0, "step": 9004 }, { "epoch": 1.1455285587075437, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.00017070770263671875, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.575007915496826, "learning_rate": 1e-05, "loss": 0.5477, "mean_token_accuracy": 0.8549328446388245, "num_tokens": 341814725.0, "step": 9005 }, { "epoch": 1.145655768986134, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.00017070770263671875, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.3195085525512695, "learning_rate": 1e-05, "loss": 0.4577, "mean_token_accuracy": 0.8813936710357666, "num_tokens": 341854982.0, "step": 9006 }, { "epoch": 1.1457829792647245, "ewc_loss": 0.10498046875, "ewc_loss_diag": 0.00017070770263671875, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 6.1739606857299805, "learning_rate": 1e-05, "loss": 0.4841, "mean_token_accuracy": 0.8761999607086182, "num_tokens": 341888952.0, "step": 9007 }, { "epoch": 1.145910189543315, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.00017070770263671875, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.227227687835693, "learning_rate": 1e-05, "loss": 0.4812, "mean_token_accuracy": 0.874514102935791, "num_tokens": 341927985.0, "step": 9008 }, { "epoch": 1.1460373998219056, "ewc_loss": 0.10107421875, "ewc_loss_diag": 0.00017070770263671875, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.8998517990112305, "learning_rate": 1e-05, "loss": 0.4827, "mean_token_accuracy": 0.8682659268379211, "num_tokens": 341959813.0, "step": 9009 }, { "epoch": 1.146164610100496, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.00017070770263671875, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.1898908615112305, "learning_rate": 1e-05, "loss": 0.4858, "mean_token_accuracy": 0.8742498159408569, "num_tokens": 341997231.0, "step": 9010 }, { "epoch": 1.1462918203790866, "ewc_loss": 0.09716796875, "ewc_loss_diag": 0.00017070770263671875, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.717677116394043, "learning_rate": 1e-05, "loss": 0.4456, "mean_token_accuracy": 0.880452036857605, "num_tokens": 342029277.0, "step": 9011 }, { "epoch": 1.1464190306576771, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.00017070770263671875, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 6.205356121063232, "learning_rate": 1e-05, "loss": 0.4662, "mean_token_accuracy": 0.8784319758415222, "num_tokens": 342069297.0, "step": 9012 }, { "epoch": 1.1465462409362677, "ewc_loss": 0.0966796875, "ewc_loss_diag": 0.00017070770263671875, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.572779655456543, "learning_rate": 1e-05, "loss": 0.4929, "mean_token_accuracy": 0.8651710748672485, "num_tokens": 342111525.0, "step": 9013 }, { "epoch": 1.1466734512148582, "ewc_loss": 0.109375, "ewc_loss_diag": 0.00017070770263671875, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.319631099700928, "learning_rate": 1e-05, "loss": 0.5008, "mean_token_accuracy": 0.8683267831802368, "num_tokens": 342148445.0, "step": 9014 }, { "epoch": 1.1468006614934487, "ewc_loss": 0.09521484375, "ewc_loss_diag": 0.00017070770263671875, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.455169200897217, "learning_rate": 1e-05, "loss": 0.4786, "mean_token_accuracy": 0.8671015501022339, "num_tokens": 342186855.0, "step": 9015 }, { "epoch": 1.1469278717720393, "ewc_loss": 0.11376953125, "ewc_loss_diag": 0.00017070770263671875, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.535991668701172, "learning_rate": 1e-05, "loss": 0.503, "mean_token_accuracy": 0.8670003414154053, "num_tokens": 342224619.0, "step": 9016 }, { "epoch": 1.1470550820506298, "ewc_loss": 0.09423828125, "ewc_loss_diag": 0.00017070770263671875, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.370393753051758, "learning_rate": 1e-05, "loss": 0.4461, "mean_token_accuracy": 0.878086268901825, "num_tokens": 342261775.0, "step": 9017 }, { "epoch": 1.1471822923292203, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.00017070770263671875, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.858491897583008, "learning_rate": 1e-05, "loss": 0.4783, "mean_token_accuracy": 0.8755292892456055, "num_tokens": 342296355.0, "step": 9018 }, { "epoch": 1.1473095026078106, "ewc_loss": 0.09716796875, "ewc_loss_diag": 0.000171661376953125, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.445815086364746, "learning_rate": 1e-05, "loss": 0.5425, "mean_token_accuracy": 0.8500241041183472, "num_tokens": 342337363.0, "step": 9019 }, { "epoch": 1.1474367128864011, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.000171661376953125, "ewc_loss_parallel": 0.00011205673217773438, "grad_norm": 7.148820877075195, "learning_rate": 1e-05, "loss": 0.5156, "mean_token_accuracy": 0.8685146570205688, "num_tokens": 342376462.0, "step": 9020 }, { "epoch": 1.1475639231649917, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.000171661376953125, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 5.77421760559082, "learning_rate": 1e-05, "loss": 0.4765, "mean_token_accuracy": 0.8715298175811768, "num_tokens": 342417341.0, "step": 9021 }, { "epoch": 1.1476911334435822, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.00017070770263671875, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 7.13199520111084, "learning_rate": 1e-05, "loss": 0.4829, "mean_token_accuracy": 0.8772351741790771, "num_tokens": 342455489.0, "step": 9022 }, { "epoch": 1.1478183437221727, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.00017070770263671875, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.094801425933838, "learning_rate": 1e-05, "loss": 0.4638, "mean_token_accuracy": 0.8753010630607605, "num_tokens": 342489594.0, "step": 9023 }, { "epoch": 1.1479455540007633, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.00017070770263671875, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.811540126800537, "learning_rate": 1e-05, "loss": 0.5096, "mean_token_accuracy": 0.8605901002883911, "num_tokens": 342525473.0, "step": 9024 }, { "epoch": 1.1480727642793538, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.000171661376953125, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.192018985748291, "learning_rate": 1e-05, "loss": 0.4625, "mean_token_accuracy": 0.8809813857078552, "num_tokens": 342557472.0, "step": 9025 }, { "epoch": 1.1481999745579443, "ewc_loss": 0.11181640625, "ewc_loss_diag": 0.000171661376953125, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.385491371154785, "learning_rate": 1e-05, "loss": 0.4741, "mean_token_accuracy": 0.8741341829299927, "num_tokens": 342593386.0, "step": 9026 }, { "epoch": 1.1483271848365348, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.000171661376953125, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.167690277099609, "learning_rate": 1e-05, "loss": 0.518, "mean_token_accuracy": 0.8646590709686279, "num_tokens": 342631370.0, "step": 9027 }, { "epoch": 1.1484543951151254, "ewc_loss": 0.10302734375, "ewc_loss_diag": 0.000171661376953125, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 6.012400150299072, "learning_rate": 1e-05, "loss": 0.4845, "mean_token_accuracy": 0.8694782853126526, "num_tokens": 342675570.0, "step": 9028 }, { "epoch": 1.148581605393716, "ewc_loss": 0.10888671875, "ewc_loss_diag": 0.000171661376953125, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.153655529022217, "learning_rate": 1e-05, "loss": 0.4943, "mean_token_accuracy": 0.8641108274459839, "num_tokens": 342717519.0, "step": 9029 }, { "epoch": 1.1487088156723062, "ewc_loss": 0.09912109375, "ewc_loss_diag": 0.000171661376953125, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.765475273132324, "learning_rate": 1e-05, "loss": 0.4354, "mean_token_accuracy": 0.8818253874778748, "num_tokens": 342754104.0, "step": 9030 }, { "epoch": 1.1488360259508967, "ewc_loss": 0.10693359375, "ewc_loss_diag": 0.000171661376953125, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.175509929656982, "learning_rate": 1e-05, "loss": 0.5306, "mean_token_accuracy": 0.858405590057373, "num_tokens": 342796059.0, "step": 9031 }, { "epoch": 1.1489632362294873, "ewc_loss": 0.0966796875, "ewc_loss_diag": 0.000171661376953125, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.611617088317871, "learning_rate": 1e-05, "loss": 0.5025, "mean_token_accuracy": 0.863421618938446, "num_tokens": 342833731.0, "step": 9032 }, { "epoch": 1.1490904465080778, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.000171661376953125, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.25238561630249, "learning_rate": 1e-05, "loss": 0.4633, "mean_token_accuracy": 0.8771756291389465, "num_tokens": 342871918.0, "step": 9033 }, { "epoch": 1.1492176567866683, "ewc_loss": 0.095703125, "ewc_loss_diag": 0.000171661376953125, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.496248722076416, "learning_rate": 1e-05, "loss": 0.514, "mean_token_accuracy": 0.8543398380279541, "num_tokens": 342903129.0, "step": 9034 }, { "epoch": 1.1493448670652588, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.000171661376953125, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.398603439331055, "learning_rate": 1e-05, "loss": 0.5262, "mean_token_accuracy": 0.8591271042823792, "num_tokens": 342941503.0, "step": 9035 }, { "epoch": 1.1494720773438494, "ewc_loss": 0.09521484375, "ewc_loss_diag": 0.000171661376953125, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.397102355957031, "learning_rate": 1e-05, "loss": 0.4868, "mean_token_accuracy": 0.8633884191513062, "num_tokens": 342977766.0, "step": 9036 }, { "epoch": 1.14959928762244, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.000171661376953125, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.639312267303467, "learning_rate": 1e-05, "loss": 0.4902, "mean_token_accuracy": 0.8699383735656738, "num_tokens": 343019865.0, "step": 9037 }, { "epoch": 1.1497264979010304, "ewc_loss": 0.09521484375, "ewc_loss_diag": 0.000171661376953125, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.362274646759033, "learning_rate": 1e-05, "loss": 0.5097, "mean_token_accuracy": 0.8580923080444336, "num_tokens": 343056500.0, "step": 9038 }, { "epoch": 1.149853708179621, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.000171661376953125, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.969377517700195, "learning_rate": 1e-05, "loss": 0.5521, "mean_token_accuracy": 0.8552235960960388, "num_tokens": 343098404.0, "step": 9039 }, { "epoch": 1.1499809184582115, "ewc_loss": 0.099609375, "ewc_loss_diag": 0.000171661376953125, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.520955562591553, "learning_rate": 1e-05, "loss": 0.4352, "mean_token_accuracy": 0.8834429383277893, "num_tokens": 343137360.0, "step": 9040 }, { "epoch": 1.150108128736802, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.000171661376953125, "ewc_loss_parallel": 0.00011301040649414062, "grad_norm": 7.161269187927246, "learning_rate": 1e-05, "loss": 0.4831, "mean_token_accuracy": 0.8777568340301514, "num_tokens": 343176038.0, "step": 9041 }, { "epoch": 1.1502353390153925, "ewc_loss": 0.10986328125, "ewc_loss_diag": 0.000171661376953125, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 5.884003162384033, "learning_rate": 1e-05, "loss": 0.4855, "mean_token_accuracy": 0.8709874153137207, "num_tokens": 343217888.0, "step": 9042 }, { "epoch": 1.150362549293983, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.000171661376953125, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 7.015509128570557, "learning_rate": 1e-05, "loss": 0.4983, "mean_token_accuracy": 0.872907280921936, "num_tokens": 343257217.0, "step": 9043 }, { "epoch": 1.1504897595725734, "ewc_loss": 0.11474609375, "ewc_loss_diag": 0.000171661376953125, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.144047737121582, "learning_rate": 1e-05, "loss": 0.4708, "mean_token_accuracy": 0.8744678497314453, "num_tokens": 343290607.0, "step": 9044 }, { "epoch": 1.150616969851164, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.000171661376953125, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.613028049468994, "learning_rate": 1e-05, "loss": 0.4856, "mean_token_accuracy": 0.8711175918579102, "num_tokens": 343329608.0, "step": 9045 }, { "epoch": 1.1507441801297544, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.000171661376953125, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.173809051513672, "learning_rate": 1e-05, "loss": 0.4745, "mean_token_accuracy": 0.8807073831558228, "num_tokens": 343365459.0, "step": 9046 }, { "epoch": 1.150871390408345, "ewc_loss": 0.10791015625, "ewc_loss_diag": 0.000171661376953125, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.19720983505249, "learning_rate": 1e-05, "loss": 0.4717, "mean_token_accuracy": 0.8731622099876404, "num_tokens": 343403909.0, "step": 9047 }, { "epoch": 1.1509986006869355, "ewc_loss": 0.11083984375, "ewc_loss_diag": 0.000171661376953125, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.1643171310424805, "learning_rate": 1e-05, "loss": 0.429, "mean_token_accuracy": 0.8906972408294678, "num_tokens": 343438506.0, "step": 9048 }, { "epoch": 1.151125810965526, "ewc_loss": 0.10009765625, "ewc_loss_diag": 0.000171661376953125, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.861920356750488, "learning_rate": 1e-05, "loss": 0.4987, "mean_token_accuracy": 0.8636151552200317, "num_tokens": 343475059.0, "step": 9049 }, { "epoch": 1.1512530212441165, "ewc_loss": 0.10888671875, "ewc_loss_diag": 0.000171661376953125, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.1964030265808105, "learning_rate": 1e-05, "loss": 0.4863, "mean_token_accuracy": 0.8714331984519958, "num_tokens": 343511286.0, "step": 9050 }, { "epoch": 1.151380231522707, "ewc_loss": 0.09814453125, "ewc_loss_diag": 0.000171661376953125, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.639279842376709, "learning_rate": 1e-05, "loss": 0.4451, "mean_token_accuracy": 0.879918098449707, "num_tokens": 343548092.0, "step": 9051 }, { "epoch": 1.1515074418012976, "ewc_loss": 0.10888671875, "ewc_loss_diag": 0.000171661376953125, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.260336399078369, "learning_rate": 1e-05, "loss": 0.4392, "mean_token_accuracy": 0.8856189250946045, "num_tokens": 343586487.0, "step": 9052 }, { "epoch": 1.1516346520798881, "ewc_loss": 0.09619140625, "ewc_loss_diag": 0.000171661376953125, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.492307662963867, "learning_rate": 1e-05, "loss": 0.4425, "mean_token_accuracy": 0.883208155632019, "num_tokens": 343627747.0, "step": 9053 }, { "epoch": 1.1517618623584787, "ewc_loss": 0.11181640625, "ewc_loss_diag": 0.000171661376953125, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.409387111663818, "learning_rate": 1e-05, "loss": 0.5259, "mean_token_accuracy": 0.8562468886375427, "num_tokens": 343672725.0, "step": 9054 }, { "epoch": 1.151889072637069, "ewc_loss": 0.09521484375, "ewc_loss_diag": 0.000171661376953125, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.394640922546387, "learning_rate": 1e-05, "loss": 0.4893, "mean_token_accuracy": 0.8662375807762146, "num_tokens": 343714426.0, "step": 9055 }, { "epoch": 1.1520162829156595, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.000171661376953125, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.668832778930664, "learning_rate": 1e-05, "loss": 0.5228, "mean_token_accuracy": 0.8627288341522217, "num_tokens": 343754613.0, "step": 9056 }, { "epoch": 1.15214349319425, "ewc_loss": 0.09521484375, "ewc_loss_diag": 0.00017261505126953125, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.373153209686279, "learning_rate": 1e-05, "loss": 0.4815, "mean_token_accuracy": 0.8701773881912231, "num_tokens": 343796187.0, "step": 9057 }, { "epoch": 1.1522707034728406, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.00017261505126953125, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 6.984633445739746, "learning_rate": 1e-05, "loss": 0.5227, "mean_token_accuracy": 0.8644390106201172, "num_tokens": 343833323.0, "step": 9058 }, { "epoch": 1.152397913751431, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.00017261505126953125, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.563798904418945, "learning_rate": 1e-05, "loss": 0.519, "mean_token_accuracy": 0.8584759831428528, "num_tokens": 343872642.0, "step": 9059 }, { "epoch": 1.1525251240300216, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.00017261505126953125, "ewc_loss_parallel": 0.00011348724365234375, "grad_norm": 7.145612716674805, "learning_rate": 1e-05, "loss": 0.5164, "mean_token_accuracy": 0.8668438196182251, "num_tokens": 343909104.0, "step": 9060 }, { "epoch": 1.1526523343086121, "ewc_loss": 0.11083984375, "ewc_loss_diag": 0.00017261505126953125, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 5.91289758682251, "learning_rate": 1e-05, "loss": 0.5107, "mean_token_accuracy": 0.8640954494476318, "num_tokens": 343947102.0, "step": 9061 }, { "epoch": 1.1527795445872027, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.00017261505126953125, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 6.9641289710998535, "learning_rate": 1e-05, "loss": 0.4771, "mean_token_accuracy": 0.8763749599456787, "num_tokens": 343985887.0, "step": 9062 }, { "epoch": 1.1529067548657932, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.00017261505126953125, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.121196269989014, "learning_rate": 1e-05, "loss": 0.532, "mean_token_accuracy": 0.8591165542602539, "num_tokens": 344019427.0, "step": 9063 }, { "epoch": 1.1530339651443837, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.00017261505126953125, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.587172985076904, "learning_rate": 1e-05, "loss": 0.5184, "mean_token_accuracy": 0.8606400489807129, "num_tokens": 344061706.0, "step": 9064 }, { "epoch": 1.1531611754229742, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.00017261505126953125, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.1417975425720215, "learning_rate": 1e-05, "loss": 0.4508, "mean_token_accuracy": 0.8851934671401978, "num_tokens": 344100912.0, "step": 9065 }, { "epoch": 1.1532883857015648, "ewc_loss": 0.10791015625, "ewc_loss_diag": 0.00017261505126953125, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.215049743652344, "learning_rate": 1e-05, "loss": 0.5091, "mean_token_accuracy": 0.8635783195495605, "num_tokens": 344135645.0, "step": 9066 }, { "epoch": 1.1534155959801553, "ewc_loss": 0.10986328125, "ewc_loss_diag": 0.00017261505126953125, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.118785858154297, "learning_rate": 1e-05, "loss": 0.4996, "mean_token_accuracy": 0.8684533834457397, "num_tokens": 344176467.0, "step": 9067 }, { "epoch": 1.1535428062587456, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.00017261505126953125, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.9222941398620605, "learning_rate": 1e-05, "loss": 0.4993, "mean_token_accuracy": 0.8642100691795349, "num_tokens": 344213498.0, "step": 9068 }, { "epoch": 1.1536700165373361, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.00017261505126953125, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.092559814453125, "learning_rate": 1e-05, "loss": 0.4449, "mean_token_accuracy": 0.8855059146881104, "num_tokens": 344250105.0, "step": 9069 }, { "epoch": 1.1537972268159267, "ewc_loss": 0.09912109375, "ewc_loss_diag": 0.00017261505126953125, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.735190391540527, "learning_rate": 1e-05, "loss": 0.4926, "mean_token_accuracy": 0.8662813305854797, "num_tokens": 344284017.0, "step": 9070 }, { "epoch": 1.1539244370945172, "ewc_loss": 0.10595703125, "ewc_loss_diag": 0.00017261505126953125, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.093440532684326, "learning_rate": 1e-05, "loss": 0.4722, "mean_token_accuracy": 0.8720654249191284, "num_tokens": 344317729.0, "step": 9071 }, { "epoch": 1.1540516473731077, "ewc_loss": 0.09814453125, "ewc_loss_diag": 0.00017261505126953125, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.626829624176025, "learning_rate": 1e-05, "loss": 0.5272, "mean_token_accuracy": 0.8515651226043701, "num_tokens": 344356032.0, "step": 9072 }, { "epoch": 1.1541788576516983, "ewc_loss": 0.10595703125, "ewc_loss_diag": 0.00017261505126953125, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.140210151672363, "learning_rate": 1e-05, "loss": 0.5204, "mean_token_accuracy": 0.8589482307434082, "num_tokens": 344385694.0, "step": 9073 }, { "epoch": 1.1543060679302888, "ewc_loss": 0.09619140625, "ewc_loss_diag": 0.00017261505126953125, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.511188983917236, "learning_rate": 1e-05, "loss": 0.4797, "mean_token_accuracy": 0.8648031949996948, "num_tokens": 344420435.0, "step": 9074 }, { "epoch": 1.1544332782088793, "ewc_loss": 0.10888671875, "ewc_loss_diag": 0.00017261505126953125, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.271744728088379, "learning_rate": 1e-05, "loss": 0.4522, "mean_token_accuracy": 0.8794053792953491, "num_tokens": 344461065.0, "step": 9075 }, { "epoch": 1.1545604884874698, "ewc_loss": 0.095703125, "ewc_loss_diag": 0.00017261505126953125, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.386078357696533, "learning_rate": 1e-05, "loss": 0.4419, "mean_token_accuracy": 0.8755471706390381, "num_tokens": 344495419.0, "step": 9076 }, { "epoch": 1.1546876987660604, "ewc_loss": 0.11474609375, "ewc_loss_diag": 0.00017261505126953125, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.53035831451416, "learning_rate": 1e-05, "loss": 0.5359, "mean_token_accuracy": 0.8565418720245361, "num_tokens": 344536304.0, "step": 9077 }, { "epoch": 1.154814909044651, "ewc_loss": 0.0947265625, "ewc_loss_diag": 0.00017261505126953125, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.285298824310303, "learning_rate": 1e-05, "loss": 0.4881, "mean_token_accuracy": 0.8708584308624268, "num_tokens": 344577118.0, "step": 9078 }, { "epoch": 1.1549421193232412, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.00017261505126953125, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.942510604858398, "learning_rate": 1e-05, "loss": 0.5393, "mean_token_accuracy": 0.8585978746414185, "num_tokens": 344609919.0, "step": 9079 }, { "epoch": 1.1550693296018317, "ewc_loss": 0.09814453125, "ewc_loss_diag": 0.00017261505126953125, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.381563663482666, "learning_rate": 1e-05, "loss": 0.4714, "mean_token_accuracy": 0.8691372871398926, "num_tokens": 344651543.0, "step": 9080 }, { "epoch": 1.1551965398804223, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.00017261505126953125, "ewc_loss_parallel": 0.00011968612670898438, "grad_norm": 7.31687593460083, "learning_rate": 1e-05, "loss": 0.5013, "mean_token_accuracy": 0.8739534616470337, "num_tokens": 344691477.0, "step": 9081 }, { "epoch": 1.1553237501590128, "ewc_loss": 0.11083984375, "ewc_loss_diag": 0.00017261505126953125, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 5.820550918579102, "learning_rate": 1e-05, "loss": 0.4971, "mean_token_accuracy": 0.8673462867736816, "num_tokens": 344731533.0, "step": 9082 }, { "epoch": 1.1554509604376033, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.00017261505126953125, "ewc_loss_parallel": 0.00011587142944335938, "grad_norm": 7.2959136962890625, "learning_rate": 1e-05, "loss": 0.5018, "mean_token_accuracy": 0.8804584741592407, "num_tokens": 344771628.0, "step": 9083 }, { "epoch": 1.1555781707161938, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.00017261505126953125, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.2367119789123535, "learning_rate": 1e-05, "loss": 0.5368, "mean_token_accuracy": 0.8602429628372192, "num_tokens": 344812079.0, "step": 9084 }, { "epoch": 1.1557053809947844, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.00017261505126953125, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.893538475036621, "learning_rate": 1e-05, "loss": 0.4874, "mean_token_accuracy": 0.8735309839248657, "num_tokens": 344856395.0, "step": 9085 }, { "epoch": 1.155832591273375, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.00017261505126953125, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.2921013832092285, "learning_rate": 1e-05, "loss": 0.4873, "mean_token_accuracy": 0.8766375780105591, "num_tokens": 344894578.0, "step": 9086 }, { "epoch": 1.1559598015519654, "ewc_loss": 0.11376953125, "ewc_loss_diag": 0.00017261505126953125, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.453663349151611, "learning_rate": 1e-05, "loss": 0.5493, "mean_token_accuracy": 0.8543568253517151, "num_tokens": 344937896.0, "step": 9087 }, { "epoch": 1.156087011830556, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.00017261505126953125, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.217769145965576, "learning_rate": 1e-05, "loss": 0.4574, "mean_token_accuracy": 0.8816559314727783, "num_tokens": 344972852.0, "step": 9088 }, { "epoch": 1.1562142221091465, "ewc_loss": 0.10498046875, "ewc_loss_diag": 0.00017261505126953125, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 6.093222618103027, "learning_rate": 1e-05, "loss": 0.5048, "mean_token_accuracy": 0.8637354969978333, "num_tokens": 345010322.0, "step": 9089 }, { "epoch": 1.156341432387737, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.00017261505126953125, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.164767742156982, "learning_rate": 1e-05, "loss": 0.4372, "mean_token_accuracy": 0.8864269256591797, "num_tokens": 345046268.0, "step": 9090 }, { "epoch": 1.1564686426663275, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.00017261505126953125, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.833312511444092, "learning_rate": 1e-05, "loss": 0.4285, "mean_token_accuracy": 0.8846778869628906, "num_tokens": 345078601.0, "step": 9091 }, { "epoch": 1.156595852944918, "ewc_loss": 0.10693359375, "ewc_loss_diag": 0.00017261505126953125, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.148950576782227, "learning_rate": 1e-05, "loss": 0.4959, "mean_token_accuracy": 0.869011640548706, "num_tokens": 345121863.0, "step": 9092 }, { "epoch": 1.1567230632235084, "ewc_loss": 0.09716796875, "ewc_loss_diag": 0.00017261505126953125, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.665129661560059, "learning_rate": 1e-05, "loss": 0.4083, "mean_token_accuracy": 0.8908894062042236, "num_tokens": 345153884.0, "step": 9093 }, { "epoch": 1.156850273502099, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.00017261505126953125, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.205160617828369, "learning_rate": 1e-05, "loss": 0.4724, "mean_token_accuracy": 0.8763166666030884, "num_tokens": 345188793.0, "step": 9094 }, { "epoch": 1.1569774837806894, "ewc_loss": 0.0966796875, "ewc_loss_diag": 0.00017261505126953125, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.524565696716309, "learning_rate": 1e-05, "loss": 0.456, "mean_token_accuracy": 0.8767318725585938, "num_tokens": 345222537.0, "step": 9095 }, { "epoch": 1.15710469405928, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.00017261505126953125, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.353296756744385, "learning_rate": 1e-05, "loss": 0.4997, "mean_token_accuracy": 0.8664157390594482, "num_tokens": 345262006.0, "step": 9096 }, { "epoch": 1.1572319043378705, "ewc_loss": 0.0947265625, "ewc_loss_diag": 0.00017261505126953125, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.399993896484375, "learning_rate": 1e-05, "loss": 0.5282, "mean_token_accuracy": 0.8542259931564331, "num_tokens": 345295388.0, "step": 9097 }, { "epoch": 1.157359114616461, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.00017261505126953125, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.624914169311523, "learning_rate": 1e-05, "loss": 0.4672, "mean_token_accuracy": 0.8813665509223938, "num_tokens": 345331144.0, "step": 9098 }, { "epoch": 1.1574863248950515, "ewc_loss": 0.09521484375, "ewc_loss_diag": 0.00017261505126953125, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.322701454162598, "learning_rate": 1e-05, "loss": 0.5307, "mean_token_accuracy": 0.850446879863739, "num_tokens": 345373009.0, "step": 9099 }, { "epoch": 1.157613535173642, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.00017261505126953125, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 7.008450031280518, "learning_rate": 1e-05, "loss": 0.4731, "mean_token_accuracy": 0.8811852931976318, "num_tokens": 345411931.0, "step": 9100 }, { "epoch": 1.1577407454522326, "ewc_loss": 0.10107421875, "ewc_loss_diag": 0.00017261505126953125, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.488479137420654, "learning_rate": 1e-05, "loss": 0.4786, "mean_token_accuracy": 0.868820071220398, "num_tokens": 345448818.0, "step": 9101 }, { "epoch": 1.1578679557308231, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.00017261505126953125, "ewc_loss_parallel": 0.00011730194091796875, "grad_norm": 7.299356937408447, "learning_rate": 1e-05, "loss": 0.4768, "mean_token_accuracy": 0.8794986009597778, "num_tokens": 345484867.0, "step": 9102 }, { "epoch": 1.1579951660094137, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.00017261505126953125, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 5.926743984222412, "learning_rate": 1e-05, "loss": 0.4882, "mean_token_accuracy": 0.87138831615448, "num_tokens": 345525949.0, "step": 9103 }, { "epoch": 1.158122376288004, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.00017261505126953125, "ewc_loss_parallel": 0.00011205673217773438, "grad_norm": 7.162861347198486, "learning_rate": 1e-05, "loss": 0.4642, "mean_token_accuracy": 0.8824174404144287, "num_tokens": 345563027.0, "step": 9104 }, { "epoch": 1.1582495865665945, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.00017261505126953125, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.241314888000488, "learning_rate": 1e-05, "loss": 0.4696, "mean_token_accuracy": 0.8818610906600952, "num_tokens": 345605920.0, "step": 9105 }, { "epoch": 1.158376796845185, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.00017261505126953125, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.752554416656494, "learning_rate": 1e-05, "loss": 0.5351, "mean_token_accuracy": 0.863186776638031, "num_tokens": 345645843.0, "step": 9106 }, { "epoch": 1.1585040071237755, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.00017261505126953125, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.258916854858398, "learning_rate": 1e-05, "loss": 0.4953, "mean_token_accuracy": 0.8681389093399048, "num_tokens": 345684207.0, "step": 9107 }, { "epoch": 1.158631217402366, "ewc_loss": 0.11083984375, "ewc_loss_diag": 0.00017261505126953125, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.3230767250061035, "learning_rate": 1e-05, "loss": 0.5212, "mean_token_accuracy": 0.8603425621986389, "num_tokens": 345721628.0, "step": 9108 }, { "epoch": 1.1587584276809566, "ewc_loss": 0.11181640625, "ewc_loss_diag": 0.00017261505126953125, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.195869445800781, "learning_rate": 1e-05, "loss": 0.4731, "mean_token_accuracy": 0.8754067420959473, "num_tokens": 345761864.0, "step": 9109 }, { "epoch": 1.1588856379595471, "ewc_loss": 0.10205078125, "ewc_loss_diag": 0.00017261505126953125, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.977232933044434, "learning_rate": 1e-05, "loss": 0.4771, "mean_token_accuracy": 0.8708603382110596, "num_tokens": 345802478.0, "step": 9110 }, { "epoch": 1.1590128482381377, "ewc_loss": 0.10986328125, "ewc_loss_diag": 0.00017261505126953125, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.171807289123535, "learning_rate": 1e-05, "loss": 0.478, "mean_token_accuracy": 0.8724924325942993, "num_tokens": 345846598.0, "step": 9111 }, { "epoch": 1.1591400585167282, "ewc_loss": 0.09912109375, "ewc_loss_diag": 0.00017261505126953125, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.760478973388672, "learning_rate": 1e-05, "loss": 0.5242, "mean_token_accuracy": 0.8575440645217896, "num_tokens": 345885050.0, "step": 9112 }, { "epoch": 1.1592672687953187, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.00017261505126953125, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.1951189041137695, "learning_rate": 1e-05, "loss": 0.579, "mean_token_accuracy": 0.8393003940582275, "num_tokens": 345927460.0, "step": 9113 }, { "epoch": 1.1593944790739092, "ewc_loss": 0.0966796875, "ewc_loss_diag": 0.00017261505126953125, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.609316825866699, "learning_rate": 1e-05, "loss": 0.4698, "mean_token_accuracy": 0.8681504130363464, "num_tokens": 345960855.0, "step": 9114 }, { "epoch": 1.1595216893524998, "ewc_loss": 0.109375, "ewc_loss_diag": 0.00017261505126953125, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.284343719482422, "learning_rate": 1e-05, "loss": 0.4665, "mean_token_accuracy": 0.8752690553665161, "num_tokens": 345997086.0, "step": 9115 }, { "epoch": 1.1596488996310903, "ewc_loss": 0.09521484375, "ewc_loss_diag": 0.00017261505126953125, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.478066921234131, "learning_rate": 1e-05, "loss": 0.4567, "mean_token_accuracy": 0.876686155796051, "num_tokens": 346034317.0, "step": 9116 }, { "epoch": 1.1597761099096806, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.00017261505126953125, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.463351249694824, "learning_rate": 1e-05, "loss": 0.508, "mean_token_accuracy": 0.864397406578064, "num_tokens": 346072626.0, "step": 9117 }, { "epoch": 1.1599033201882711, "ewc_loss": 0.0947265625, "ewc_loss_diag": 0.00017261505126953125, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.37579870223999, "learning_rate": 1e-05, "loss": 0.5025, "mean_token_accuracy": 0.8626797199249268, "num_tokens": 346109792.0, "step": 9118 }, { "epoch": 1.1600305304668617, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.00017261505126953125, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.7626471519470215, "learning_rate": 1e-05, "loss": 0.473, "mean_token_accuracy": 0.8765996694564819, "num_tokens": 346147696.0, "step": 9119 }, { "epoch": 1.1601577407454522, "ewc_loss": 0.09619140625, "ewc_loss_diag": 0.00017261505126953125, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.383760452270508, "learning_rate": 1e-05, "loss": 0.433, "mean_token_accuracy": 0.880501389503479, "num_tokens": 346183944.0, "step": 9120 }, { "epoch": 1.1602849510240427, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.00017261505126953125, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 7.094799518585205, "learning_rate": 1e-05, "loss": 0.472, "mean_token_accuracy": 0.8775010108947754, "num_tokens": 346223352.0, "step": 9121 }, { "epoch": 1.1604121613026332, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.00017261505126953125, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 5.660588264465332, "learning_rate": 1e-05, "loss": 0.4909, "mean_token_accuracy": 0.8663926124572754, "num_tokens": 346255008.0, "step": 9122 }, { "epoch": 1.1605393715812238, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.00017261505126953125, "ewc_loss_parallel": 0.00011348724365234375, "grad_norm": 7.204535484313965, "learning_rate": 1e-05, "loss": 0.6127, "mean_token_accuracy": 0.842423677444458, "num_tokens": 346285188.0, "step": 9123 }, { "epoch": 1.1606665818598143, "ewc_loss": 0.11376953125, "ewc_loss_diag": 0.00017261505126953125, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.031762599945068, "learning_rate": 1e-05, "loss": 0.4795, "mean_token_accuracy": 0.875921368598938, "num_tokens": 346318691.0, "step": 9124 }, { "epoch": 1.1607937921384048, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.00017261505126953125, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.937768459320068, "learning_rate": 1e-05, "loss": 0.4627, "mean_token_accuracy": 0.8823986053466797, "num_tokens": 346353800.0, "step": 9125 }, { "epoch": 1.1609210024169954, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.00017261505126953125, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.184927463531494, "learning_rate": 1e-05, "loss": 0.55, "mean_token_accuracy": 0.8522460460662842, "num_tokens": 346396616.0, "step": 9126 }, { "epoch": 1.161048212695586, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.00017261505126953125, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.535857200622559, "learning_rate": 1e-05, "loss": 0.4939, "mean_token_accuracy": 0.8704306483268738, "num_tokens": 346438244.0, "step": 9127 }, { "epoch": 1.1611754229741762, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.00017261505126953125, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.173403739929199, "learning_rate": 1e-05, "loss": 0.4976, "mean_token_accuracy": 0.868954062461853, "num_tokens": 346475081.0, "step": 9128 }, { "epoch": 1.1613026332527667, "ewc_loss": 0.10595703125, "ewc_loss_diag": 0.00017261505126953125, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.149489402770996, "learning_rate": 1e-05, "loss": 0.4604, "mean_token_accuracy": 0.8751927018165588, "num_tokens": 346513435.0, "step": 9129 }, { "epoch": 1.1614298435313573, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.15402364730835, "learning_rate": 1e-05, "loss": 0.5155, "mean_token_accuracy": 0.8632835149765015, "num_tokens": 346553216.0, "step": 9130 }, { "epoch": 1.1615570538099478, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.86557674407959, "learning_rate": 1e-05, "loss": 0.4877, "mean_token_accuracy": 0.865959107875824, "num_tokens": 346589271.0, "step": 9131 }, { "epoch": 1.1616842640885383, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.15108585357666, "learning_rate": 1e-05, "loss": 0.4806, "mean_token_accuracy": 0.8699875473976135, "num_tokens": 346631806.0, "step": 9132 }, { "epoch": 1.1618114743671288, "ewc_loss": 0.0986328125, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.685223579406738, "learning_rate": 1e-05, "loss": 0.4479, "mean_token_accuracy": 0.8819096088409424, "num_tokens": 346670805.0, "step": 9133 }, { "epoch": 1.1619386846457194, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.1852264404296875, "learning_rate": 1e-05, "loss": 0.453, "mean_token_accuracy": 0.8815871477127075, "num_tokens": 346710506.0, "step": 9134 }, { "epoch": 1.16206589492431, "ewc_loss": 0.0966796875, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.542155742645264, "learning_rate": 1e-05, "loss": 0.5067, "mean_token_accuracy": 0.8609458208084106, "num_tokens": 346753410.0, "step": 9135 }, { "epoch": 1.1621931052029004, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.32061767578125, "learning_rate": 1e-05, "loss": 0.5341, "mean_token_accuracy": 0.8575595617294312, "num_tokens": 346795104.0, "step": 9136 }, { "epoch": 1.162320315481491, "ewc_loss": 0.0947265625, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.418243885040283, "learning_rate": 1e-05, "loss": 0.5235, "mean_token_accuracy": 0.8538196682929993, "num_tokens": 346836888.0, "step": 9137 }, { "epoch": 1.1624475257600815, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.541822910308838, "learning_rate": 1e-05, "loss": 0.4912, "mean_token_accuracy": 0.8665237426757812, "num_tokens": 346874449.0, "step": 9138 }, { "epoch": 1.162574736038672, "ewc_loss": 0.0947265625, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.336093425750732, "learning_rate": 1e-05, "loss": 0.3978, "mean_token_accuracy": 0.8928210735321045, "num_tokens": 346906706.0, "step": 9139 }, { "epoch": 1.1627019463172625, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.8910369873046875, "learning_rate": 1e-05, "loss": 0.5003, "mean_token_accuracy": 0.8697051405906677, "num_tokens": 346947250.0, "step": 9140 }, { "epoch": 1.162829156595853, "ewc_loss": 0.0986328125, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.417617321014404, "learning_rate": 1e-05, "loss": 0.4932, "mean_token_accuracy": 0.8661022186279297, "num_tokens": 346984159.0, "step": 9141 }, { "epoch": 1.1629563668744434, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 0.00011396408081054688, "grad_norm": 7.215508460998535, "learning_rate": 1e-05, "loss": 0.5685, "mean_token_accuracy": 0.8487804532051086, "num_tokens": 347026449.0, "step": 9142 }, { "epoch": 1.163083577153034, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 5.785244464874268, "learning_rate": 1e-05, "loss": 0.4486, "mean_token_accuracy": 0.8810845613479614, "num_tokens": 347057903.0, "step": 9143 }, { "epoch": 1.1632107874316244, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 0.00011348724365234375, "grad_norm": 7.185672760009766, "learning_rate": 1e-05, "loss": 0.5177, "mean_token_accuracy": 0.8682734966278076, "num_tokens": 347095805.0, "step": 9144 }, { "epoch": 1.163337997710215, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.148009300231934, "learning_rate": 1e-05, "loss": 0.5007, "mean_token_accuracy": 0.8690482974052429, "num_tokens": 347133357.0, "step": 9145 }, { "epoch": 1.1634652079888055, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.821835517883301, "learning_rate": 1e-05, "loss": 0.5374, "mean_token_accuracy": 0.8576807975769043, "num_tokens": 347169679.0, "step": 9146 }, { "epoch": 1.163592418267396, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.230526447296143, "learning_rate": 1e-05, "loss": 0.5237, "mean_token_accuracy": 0.8612598776817322, "num_tokens": 347207990.0, "step": 9147 }, { "epoch": 1.1637196285459865, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.39097785949707, "learning_rate": 1e-05, "loss": 0.4624, "mean_token_accuracy": 0.8788414001464844, "num_tokens": 347244427.0, "step": 9148 }, { "epoch": 1.163846838824577, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.198330879211426, "learning_rate": 1e-05, "loss": 0.4301, "mean_token_accuracy": 0.890756368637085, "num_tokens": 347276303.0, "step": 9149 }, { "epoch": 1.1639740491031676, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 6.0028557777404785, "learning_rate": 1e-05, "loss": 0.4765, "mean_token_accuracy": 0.8744246363639832, "num_tokens": 347318682.0, "step": 9150 }, { "epoch": 1.1641012593817581, "ewc_loss": 0.109375, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.181059837341309, "learning_rate": 1e-05, "loss": 0.5067, "mean_token_accuracy": 0.8593372106552124, "num_tokens": 347359355.0, "step": 9151 }, { "epoch": 1.1642284696603487, "ewc_loss": 0.0986328125, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.744377136230469, "learning_rate": 1e-05, "loss": 0.5001, "mean_token_accuracy": 0.8655993938446045, "num_tokens": 347397617.0, "step": 9152 }, { "epoch": 1.164355679938939, "ewc_loss": 0.109375, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.233291149139404, "learning_rate": 1e-05, "loss": 0.5561, "mean_token_accuracy": 0.848957896232605, "num_tokens": 347433067.0, "step": 9153 }, { "epoch": 1.1644828902175295, "ewc_loss": 0.0966796875, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.588936805725098, "learning_rate": 1e-05, "loss": 0.4297, "mean_token_accuracy": 0.882276713848114, "num_tokens": 347467091.0, "step": 9154 }, { "epoch": 1.16461010049612, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.312324523925781, "learning_rate": 1e-05, "loss": 0.4744, "mean_token_accuracy": 0.8771278262138367, "num_tokens": 347503921.0, "step": 9155 }, { "epoch": 1.1647373107747105, "ewc_loss": 0.095703125, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.465971946716309, "learning_rate": 1e-05, "loss": 0.4709, "mean_token_accuracy": 0.8722370266914368, "num_tokens": 347547061.0, "step": 9156 }, { "epoch": 1.164864521053301, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.527418613433838, "learning_rate": 1e-05, "loss": 0.5122, "mean_token_accuracy": 0.8627078533172607, "num_tokens": 347583845.0, "step": 9157 }, { "epoch": 1.1649917313318916, "ewc_loss": 0.0947265625, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.384091854095459, "learning_rate": 1e-05, "loss": 0.4822, "mean_token_accuracy": 0.8688235878944397, "num_tokens": 347624146.0, "step": 9158 }, { "epoch": 1.1651189416104821, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.817309379577637, "learning_rate": 1e-05, "loss": 0.5062, "mean_token_accuracy": 0.8687554597854614, "num_tokens": 347666171.0, "step": 9159 }, { "epoch": 1.1652461518890727, "ewc_loss": 0.09765625, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.432017803192139, "learning_rate": 1e-05, "loss": 0.4878, "mean_token_accuracy": 0.867171585559845, "num_tokens": 347706960.0, "step": 9160 }, { "epoch": 1.1653733621676632, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 0.0001125335693359375, "grad_norm": 7.109979629516602, "learning_rate": 1e-05, "loss": 0.5161, "mean_token_accuracy": 0.8661521673202515, "num_tokens": 347744830.0, "step": 9161 }, { "epoch": 1.1655005724462537, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 5.704987049102783, "learning_rate": 1e-05, "loss": 0.4787, "mean_token_accuracy": 0.8753678202629089, "num_tokens": 347791033.0, "step": 9162 }, { "epoch": 1.1656277827248442, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.00017547607421875, "ewc_loss_parallel": 0.00011205673217773438, "grad_norm": 7.158393859863281, "learning_rate": 1e-05, "loss": 0.5195, "mean_token_accuracy": 0.8663152456283569, "num_tokens": 347832579.0, "step": 9163 }, { "epoch": 1.1657549930034348, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.00017547607421875, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.057572841644287, "learning_rate": 1e-05, "loss": 0.4367, "mean_token_accuracy": 0.887436032295227, "num_tokens": 347863404.0, "step": 9164 }, { "epoch": 1.1658822032820253, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.00017547607421875, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.867312431335449, "learning_rate": 1e-05, "loss": 0.5163, "mean_token_accuracy": 0.8634735345840454, "num_tokens": 347900882.0, "step": 9165 }, { "epoch": 1.1660094135606156, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.00017547607421875, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.181822299957275, "learning_rate": 1e-05, "loss": 0.5197, "mean_token_accuracy": 0.8614134192466736, "num_tokens": 347943105.0, "step": 9166 }, { "epoch": 1.1661366238392061, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.00017547607421875, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.485448360443115, "learning_rate": 1e-05, "loss": 0.5969, "mean_token_accuracy": 0.8344823122024536, "num_tokens": 347978783.0, "step": 9167 }, { "epoch": 1.1662638341177967, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.00017547607421875, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.158616542816162, "learning_rate": 1e-05, "loss": 0.4624, "mean_token_accuracy": 0.8783469200134277, "num_tokens": 348016169.0, "step": 9168 }, { "epoch": 1.1663910443963872, "ewc_loss": 0.10498046875, "ewc_loss_diag": 0.00017547607421875, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 6.112017631530762, "learning_rate": 1e-05, "loss": 0.5102, "mean_token_accuracy": 0.8618011474609375, "num_tokens": 348052431.0, "step": 9169 }, { "epoch": 1.1665182546749777, "ewc_loss": 0.10888671875, "ewc_loss_diag": 0.00017547607421875, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.136828899383545, "learning_rate": 1e-05, "loss": 0.523, "mean_token_accuracy": 0.8652793169021606, "num_tokens": 348091412.0, "step": 9170 }, { "epoch": 1.1666454649535682, "ewc_loss": 0.10009765625, "ewc_loss_diag": 0.00017547607421875, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.853147506713867, "learning_rate": 1e-05, "loss": 0.4841, "mean_token_accuracy": 0.864856481552124, "num_tokens": 348128370.0, "step": 9171 }, { "epoch": 1.1667726752321588, "ewc_loss": 0.10693359375, "ewc_loss_diag": 0.00017547607421875, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 6.117724895477295, "learning_rate": 1e-05, "loss": 0.5306, "mean_token_accuracy": 0.8611941337585449, "num_tokens": 348167076.0, "step": 9172 }, { "epoch": 1.1668998855107493, "ewc_loss": 0.0986328125, "ewc_loss_diag": 0.00017547607421875, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.690622329711914, "learning_rate": 1e-05, "loss": 0.4584, "mean_token_accuracy": 0.8737442493438721, "num_tokens": 348206926.0, "step": 9173 }, { "epoch": 1.1670270957893398, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.00017547607421875, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.148292541503906, "learning_rate": 1e-05, "loss": 0.5094, "mean_token_accuracy": 0.8625481128692627, "num_tokens": 348247209.0, "step": 9174 }, { "epoch": 1.1671543060679304, "ewc_loss": 0.09716796875, "ewc_loss_diag": 0.00017547607421875, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.572399616241455, "learning_rate": 1e-05, "loss": 0.4835, "mean_token_accuracy": 0.8656530380249023, "num_tokens": 348286613.0, "step": 9175 }, { "epoch": 1.1672815163465209, "ewc_loss": 0.10888671875, "ewc_loss_diag": 0.00017547607421875, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.242142677307129, "learning_rate": 1e-05, "loss": 0.5026, "mean_token_accuracy": 0.8654159307479858, "num_tokens": 348322246.0, "step": 9176 }, { "epoch": 1.1674087266251112, "ewc_loss": 0.095703125, "ewc_loss_diag": 0.00017547607421875, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.446009635925293, "learning_rate": 1e-05, "loss": 0.4517, "mean_token_accuracy": 0.8752961158752441, "num_tokens": 348359435.0, "step": 9177 }, { "epoch": 1.1675359369037017, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.00017547607421875, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.449887275695801, "learning_rate": 1e-05, "loss": 0.4441, "mean_token_accuracy": 0.8861908912658691, "num_tokens": 348397330.0, "step": 9178 }, { "epoch": 1.1676631471822922, "ewc_loss": 0.0947265625, "ewc_loss_diag": 0.00017547607421875, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.341621398925781, "learning_rate": 1e-05, "loss": 0.524, "mean_token_accuracy": 0.8558690547943115, "num_tokens": 348435413.0, "step": 9179 }, { "epoch": 1.1677903574608828, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.00017547607421875, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.777961254119873, "learning_rate": 1e-05, "loss": 0.5406, "mean_token_accuracy": 0.8569287061691284, "num_tokens": 348474963.0, "step": 9180 }, { "epoch": 1.1679175677394733, "ewc_loss": 0.09619140625, "ewc_loss_diag": 0.00017547607421875, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.344995975494385, "learning_rate": 1e-05, "loss": 0.4911, "mean_token_accuracy": 0.8670377731323242, "num_tokens": 348511306.0, "step": 9181 }, { "epoch": 1.1680447780180638, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.00017547607421875, "ewc_loss_parallel": 0.00011396408081054688, "grad_norm": 7.16718864440918, "learning_rate": 1e-05, "loss": 0.4935, "mean_token_accuracy": 0.876477837562561, "num_tokens": 348552257.0, "step": 9182 }, { "epoch": 1.1681719882966544, "ewc_loss": 0.10595703125, "ewc_loss_diag": 0.00017547607421875, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 5.643271446228027, "learning_rate": 1e-05, "loss": 0.4952, "mean_token_accuracy": 0.8644624948501587, "num_tokens": 348588143.0, "step": 9183 }, { "epoch": 1.168299198575245, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.00017547607421875, "ewc_loss_parallel": 0.00011777877807617188, "grad_norm": 7.302321910858154, "learning_rate": 1e-05, "loss": 0.5246, "mean_token_accuracy": 0.8668822050094604, "num_tokens": 348633525.0, "step": 9184 }, { "epoch": 1.1684264088538354, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.00017547607421875, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.089759826660156, "learning_rate": 1e-05, "loss": 0.5188, "mean_token_accuracy": 0.8658066987991333, "num_tokens": 348676024.0, "step": 9185 }, { "epoch": 1.168553619132426, "ewc_loss": 0.125, "ewc_loss_diag": 0.00017547607421875, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 7.016750335693359, "learning_rate": 1e-05, "loss": 0.4909, "mean_token_accuracy": 0.8771647214889526, "num_tokens": 348715704.0, "step": 9186 }, { "epoch": 1.1686808294110165, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.00017547607421875, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.2729620933532715, "learning_rate": 1e-05, "loss": 0.5132, "mean_token_accuracy": 0.8627538681030273, "num_tokens": 348753160.0, "step": 9187 }, { "epoch": 1.168808039689607, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.00017547607421875, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.5704026222229, "learning_rate": 1e-05, "loss": 0.4879, "mean_token_accuracy": 0.8711020946502686, "num_tokens": 348796423.0, "step": 9188 }, { "epoch": 1.1689352499681975, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.00017547607421875, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.2319722175598145, "learning_rate": 1e-05, "loss": 0.4401, "mean_token_accuracy": 0.8889553546905518, "num_tokens": 348832813.0, "step": 9189 }, { "epoch": 1.169062460246788, "ewc_loss": 0.10791015625, "ewc_loss_diag": 0.00017547607421875, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.172667026519775, "learning_rate": 1e-05, "loss": 0.486, "mean_token_accuracy": 0.8683099150657654, "num_tokens": 348868273.0, "step": 9190 }, { "epoch": 1.1691896705253784, "ewc_loss": 0.11083984375, "ewc_loss_diag": 0.00017547607421875, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.182070255279541, "learning_rate": 1e-05, "loss": 0.4753, "mean_token_accuracy": 0.8757373094558716, "num_tokens": 348908008.0, "step": 9191 }, { "epoch": 1.169316880803969, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.00017547607421875, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.880087852478027, "learning_rate": 1e-05, "loss": 0.4868, "mean_token_accuracy": 0.8686262965202332, "num_tokens": 348947701.0, "step": 9192 }, { "epoch": 1.1694440910825594, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.00017547607421875, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.176509380340576, "learning_rate": 1e-05, "loss": 0.4936, "mean_token_accuracy": 0.867713212966919, "num_tokens": 348992221.0, "step": 9193 }, { "epoch": 1.16957130136115, "ewc_loss": 0.0986328125, "ewc_loss_diag": 0.00017547607421875, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.694573402404785, "learning_rate": 1e-05, "loss": 0.4505, "mean_token_accuracy": 0.8784463405609131, "num_tokens": 349031982.0, "step": 9194 }, { "epoch": 1.1696985116397405, "ewc_loss": 0.10888671875, "ewc_loss_diag": 0.00017547607421875, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.210333824157715, "learning_rate": 1e-05, "loss": 0.4597, "mean_token_accuracy": 0.8808616399765015, "num_tokens": 349070905.0, "step": 9195 }, { "epoch": 1.169825721918331, "ewc_loss": 0.0966796875, "ewc_loss_diag": 0.00017547607421875, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.55708646774292, "learning_rate": 1e-05, "loss": 0.4804, "mean_token_accuracy": 0.8693574666976929, "num_tokens": 349114736.0, "step": 9196 }, { "epoch": 1.1699529321969215, "ewc_loss": 0.10986328125, "ewc_loss_diag": 0.00017547607421875, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.339773654937744, "learning_rate": 1e-05, "loss": 0.5225, "mean_token_accuracy": 0.861238956451416, "num_tokens": 349153760.0, "step": 9197 }, { "epoch": 1.170080142475512, "ewc_loss": 0.09521484375, "ewc_loss_diag": 0.00017547607421875, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.43374490737915, "learning_rate": 1e-05, "loss": 0.4295, "mean_token_accuracy": 0.8831983208656311, "num_tokens": 349192464.0, "step": 9198 }, { "epoch": 1.1702073527541026, "ewc_loss": 0.11474609375, "ewc_loss_diag": 0.00017547607421875, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.569289684295654, "learning_rate": 1e-05, "loss": 0.4973, "mean_token_accuracy": 0.8683856725692749, "num_tokens": 349234444.0, "step": 9199 }, { "epoch": 1.1703345630326931, "ewc_loss": 0.09521484375, "ewc_loss_diag": 0.00017547607421875, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.344550609588623, "learning_rate": 1e-05, "loss": 0.5108, "mean_token_accuracy": 0.8561766147613525, "num_tokens": 349276777.0, "step": 9200 }, { "epoch": 1.1704617733112836, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.00017547607421875, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.92189359664917, "learning_rate": 1e-05, "loss": 0.4747, "mean_token_accuracy": 0.8798067569732666, "num_tokens": 349317158.0, "step": 9201 }, { "epoch": 1.170588983589874, "ewc_loss": 0.099609375, "ewc_loss_diag": 0.00017547607421875, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.446552276611328, "learning_rate": 1e-05, "loss": 0.5137, "mean_token_accuracy": 0.8640332221984863, "num_tokens": 349357043.0, "step": 9202 }, { "epoch": 1.1707161938684645, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.00017547607421875, "ewc_loss_parallel": 0.00011491775512695312, "grad_norm": 7.235729217529297, "learning_rate": 1e-05, "loss": 0.5729, "mean_token_accuracy": 0.85137939453125, "num_tokens": 349395843.0, "step": 9203 }, { "epoch": 1.170843404147055, "ewc_loss": 0.10986328125, "ewc_loss_diag": 0.00017547607421875, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 5.834267616271973, "learning_rate": 1e-05, "loss": 0.5114, "mean_token_accuracy": 0.8627514839172363, "num_tokens": 349429410.0, "step": 9204 }, { "epoch": 1.1709706144256455, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.00017547607421875, "ewc_loss_parallel": 0.00011301040649414062, "grad_norm": 7.177549362182617, "learning_rate": 1e-05, "loss": 0.5576, "mean_token_accuracy": 0.8535144329071045, "num_tokens": 349466507.0, "step": 9205 }, { "epoch": 1.171097824704236, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.00017547607421875, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.169169902801514, "learning_rate": 1e-05, "loss": 0.4571, "mean_token_accuracy": 0.8822476863861084, "num_tokens": 349502918.0, "step": 9206 }, { "epoch": 1.1712250349828266, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.00017547607421875, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.802750110626221, "learning_rate": 1e-05, "loss": 0.5062, "mean_token_accuracy": 0.8697166442871094, "num_tokens": 349546740.0, "step": 9207 }, { "epoch": 1.1713522452614171, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.00017547607421875, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.232860565185547, "learning_rate": 1e-05, "loss": 0.4861, "mean_token_accuracy": 0.872031569480896, "num_tokens": 349590479.0, "step": 9208 }, { "epoch": 1.1714794555400077, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.00017547607421875, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.377411365509033, "learning_rate": 1e-05, "loss": 0.4367, "mean_token_accuracy": 0.8860033750534058, "num_tokens": 349627726.0, "step": 9209 }, { "epoch": 1.1716066658185982, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.00017547607421875, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.203042507171631, "learning_rate": 1e-05, "loss": 0.5046, "mean_token_accuracy": 0.869114100933075, "num_tokens": 349663547.0, "step": 9210 }, { "epoch": 1.1717338760971887, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.00017547607421875, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 6.020968914031982, "learning_rate": 1e-05, "loss": 0.5045, "mean_token_accuracy": 0.8642320036888123, "num_tokens": 349696927.0, "step": 9211 }, { "epoch": 1.1718610863757792, "ewc_loss": 0.10986328125, "ewc_loss_diag": 0.00017547607421875, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.179295539855957, "learning_rate": 1e-05, "loss": 0.4765, "mean_token_accuracy": 0.8774077892303467, "num_tokens": 349736027.0, "step": 9212 }, { "epoch": 1.1719882966543698, "ewc_loss": 0.099609375, "ewc_loss_diag": 0.00017547607421875, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.764763355255127, "learning_rate": 1e-05, "loss": 0.5028, "mean_token_accuracy": 0.8634556531906128, "num_tokens": 349778225.0, "step": 9213 }, { "epoch": 1.1721155069329603, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.00017547607421875, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.191455364227295, "learning_rate": 1e-05, "loss": 0.4957, "mean_token_accuracy": 0.867901086807251, "num_tokens": 349818093.0, "step": 9214 }, { "epoch": 1.1722427172115506, "ewc_loss": 0.09716796875, "ewc_loss_diag": 0.00017547607421875, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.618768215179443, "learning_rate": 1e-05, "loss": 0.4909, "mean_token_accuracy": 0.8662664294242859, "num_tokens": 349856433.0, "step": 9215 }, { "epoch": 1.1723699274901411, "ewc_loss": 0.109375, "ewc_loss_diag": 0.00017547607421875, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.273758411407471, "learning_rate": 1e-05, "loss": 0.5138, "mean_token_accuracy": 0.8636184930801392, "num_tokens": 349898829.0, "step": 9216 }, { "epoch": 1.1724971377687317, "ewc_loss": 0.095703125, "ewc_loss_diag": 0.00017547607421875, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.483854293823242, "learning_rate": 1e-05, "loss": 0.4677, "mean_token_accuracy": 0.8711993098258972, "num_tokens": 349937059.0, "step": 9217 }, { "epoch": 1.1726243480473222, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.00017547607421875, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.460533142089844, "learning_rate": 1e-05, "loss": 0.5211, "mean_token_accuracy": 0.8602637052536011, "num_tokens": 349973433.0, "step": 9218 }, { "epoch": 1.1727515583259127, "ewc_loss": 0.0947265625, "ewc_loss_diag": 0.00017547607421875, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.3812174797058105, "learning_rate": 1e-05, "loss": 0.5381, "mean_token_accuracy": 0.8501360416412354, "num_tokens": 350012859.0, "step": 9219 }, { "epoch": 1.1728787686045032, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.00017547607421875, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.729207515716553, "learning_rate": 1e-05, "loss": 0.5385, "mean_token_accuracy": 0.8567580580711365, "num_tokens": 350055128.0, "step": 9220 }, { "epoch": 1.1730059788830938, "ewc_loss": 0.0966796875, "ewc_loss_diag": 0.00017547607421875, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.3906660079956055, "learning_rate": 1e-05, "loss": 0.5158, "mean_token_accuracy": 0.8545557856559753, "num_tokens": 350090058.0, "step": 9221 }, { "epoch": 1.1731331891616843, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.00017547607421875, "ewc_loss_parallel": 0.00010967254638671875, "grad_norm": 7.044379234313965, "learning_rate": 1e-05, "loss": 0.4776, "mean_token_accuracy": 0.8766629099845886, "num_tokens": 350129441.0, "step": 9222 }, { "epoch": 1.1732603994402748, "ewc_loss": 0.10205078125, "ewc_loss_diag": 0.00017547607421875, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.598285675048828, "learning_rate": 1e-05, "loss": 0.4575, "mean_token_accuracy": 0.8748683929443359, "num_tokens": 350168866.0, "step": 9223 }, { "epoch": 1.1733876097188654, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.00017547607421875, "ewc_loss_parallel": 0.00011396408081054688, "grad_norm": 7.191822528839111, "learning_rate": 1e-05, "loss": 0.5204, "mean_token_accuracy": 0.8651067018508911, "num_tokens": 350207498.0, "step": 9224 }, { "epoch": 1.1735148199974559, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.00017547607421875, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 5.975127696990967, "learning_rate": 1e-05, "loss": 0.5307, "mean_token_accuracy": 0.8573626279830933, "num_tokens": 350248729.0, "step": 9225 }, { "epoch": 1.1736420302760462, "ewc_loss": 0.125, "ewc_loss_diag": 0.00017547607421875, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.9800896644592285, "learning_rate": 1e-05, "loss": 0.5091, "mean_token_accuracy": 0.8679909706115723, "num_tokens": 350279399.0, "step": 9226 }, { "epoch": 1.1737692405546367, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.00017547607421875, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.17779541015625, "learning_rate": 1e-05, "loss": 0.5573, "mean_token_accuracy": 0.8508753180503845, "num_tokens": 350321135.0, "step": 9227 }, { "epoch": 1.1738964508332272, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.00017547607421875, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.57534646987915, "learning_rate": 1e-05, "loss": 0.5103, "mean_token_accuracy": 0.8647460341453552, "num_tokens": 350359348.0, "step": 9228 }, { "epoch": 1.1740236611118178, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.00017547607421875, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.1823554039001465, "learning_rate": 1e-05, "loss": 0.4773, "mean_token_accuracy": 0.8727283477783203, "num_tokens": 350401140.0, "step": 9229 }, { "epoch": 1.1741508713904083, "ewc_loss": 0.10791015625, "ewc_loss_diag": 0.00017547607421875, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.1753764152526855, "learning_rate": 1e-05, "loss": 0.5055, "mean_token_accuracy": 0.8641871809959412, "num_tokens": 350443936.0, "step": 9230 }, { "epoch": 1.1742780816689988, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.00017547607421875, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.160530090332031, "learning_rate": 1e-05, "loss": 0.4585, "mean_token_accuracy": 0.8780078887939453, "num_tokens": 350486110.0, "step": 9231 }, { "epoch": 1.1744052919475894, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.00017547607421875, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.888185501098633, "learning_rate": 1e-05, "loss": 0.5563, "mean_token_accuracy": 0.8488271832466125, "num_tokens": 350521964.0, "step": 9232 }, { "epoch": 1.1745325022261799, "ewc_loss": 0.109375, "ewc_loss_diag": 0.00017547607421875, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.162562370300293, "learning_rate": 1e-05, "loss": 0.4573, "mean_token_accuracy": 0.8820915222167969, "num_tokens": 350557578.0, "step": 9233 }, { "epoch": 1.1746597125047704, "ewc_loss": 0.09912109375, "ewc_loss_diag": 0.00017547607421875, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.6906046867370605, "learning_rate": 1e-05, "loss": 0.4794, "mean_token_accuracy": 0.8670550584793091, "num_tokens": 350595693.0, "step": 9234 }, { "epoch": 1.174786922783361, "ewc_loss": 0.10791015625, "ewc_loss_diag": 0.00017547607421875, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.204726219177246, "learning_rate": 1e-05, "loss": 0.4691, "mean_token_accuracy": 0.8766050338745117, "num_tokens": 350633247.0, "step": 9235 }, { "epoch": 1.1749141330619515, "ewc_loss": 0.0966796875, "ewc_loss_diag": 0.00017547607421875, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.5452728271484375, "learning_rate": 1e-05, "loss": 0.4561, "mean_token_accuracy": 0.8789088726043701, "num_tokens": 350673048.0, "step": 9236 }, { "epoch": 1.175041343340542, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.00017547607421875, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.339956760406494, "learning_rate": 1e-05, "loss": 0.5365, "mean_token_accuracy": 0.8571134805679321, "num_tokens": 350708840.0, "step": 9237 }, { "epoch": 1.1751685536191325, "ewc_loss": 0.095703125, "ewc_loss_diag": 0.00017547607421875, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.414114475250244, "learning_rate": 1e-05, "loss": 0.4901, "mean_token_accuracy": 0.8682580590248108, "num_tokens": 350748875.0, "step": 9238 }, { "epoch": 1.175295763897723, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.00017547607421875, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.594858169555664, "learning_rate": 1e-05, "loss": 0.4932, "mean_token_accuracy": 0.8665577173233032, "num_tokens": 350783200.0, "step": 9239 }, { "epoch": 1.1754229741763134, "ewc_loss": 0.095703125, "ewc_loss_diag": 0.00017547607421875, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.339928150177002, "learning_rate": 1e-05, "loss": 0.4722, "mean_token_accuracy": 0.8690569400787354, "num_tokens": 350819514.0, "step": 9240 }, { "epoch": 1.175550184454904, "ewc_loss": 0.125, "ewc_loss_diag": 0.00017547607421875, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.938061714172363, "learning_rate": 1e-05, "loss": 0.4466, "mean_token_accuracy": 0.8888673782348633, "num_tokens": 350854261.0, "step": 9241 }, { "epoch": 1.1756773947334944, "ewc_loss": 0.099609375, "ewc_loss_diag": 0.00017547607421875, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.45224666595459, "learning_rate": 1e-05, "loss": 0.4814, "mean_token_accuracy": 0.8697326183319092, "num_tokens": 350897847.0, "step": 9242 }, { "epoch": 1.175804605012085, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.00017547607421875, "ewc_loss_parallel": 0.0001163482666015625, "grad_norm": 7.241927623748779, "learning_rate": 1e-05, "loss": 0.5347, "mean_token_accuracy": 0.8605673313140869, "num_tokens": 350931926.0, "step": 9243 }, { "epoch": 1.1759318152906755, "ewc_loss": 0.11083984375, "ewc_loss_diag": 0.00017547607421875, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 5.845260143280029, "learning_rate": 1e-05, "loss": 0.459, "mean_token_accuracy": 0.8798658847808838, "num_tokens": 350972352.0, "step": 9244 }, { "epoch": 1.176059025569266, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.00017547607421875, "ewc_loss_parallel": 0.00011301040649414062, "grad_norm": 7.174672603607178, "learning_rate": 1e-05, "loss": 0.5152, "mean_token_accuracy": 0.8661606311798096, "num_tokens": 351012245.0, "step": 9245 }, { "epoch": 1.1761862358478565, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.00017547607421875, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.191459655761719, "learning_rate": 1e-05, "loss": 0.4706, "mean_token_accuracy": 0.8751105070114136, "num_tokens": 351043005.0, "step": 9246 }, { "epoch": 1.176313446126447, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.00017547607421875, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.786646366119385, "learning_rate": 1e-05, "loss": 0.5563, "mean_token_accuracy": 0.8543184995651245, "num_tokens": 351084810.0, "step": 9247 }, { "epoch": 1.1764406564050376, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.00017547607421875, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.2353057861328125, "learning_rate": 1e-05, "loss": 0.4776, "mean_token_accuracy": 0.8742532730102539, "num_tokens": 351124243.0, "step": 9248 }, { "epoch": 1.1765678666836281, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.00017547607421875, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.357910633087158, "learning_rate": 1e-05, "loss": 0.4666, "mean_token_accuracy": 0.8783141374588013, "num_tokens": 351163650.0, "step": 9249 }, { "epoch": 1.1766950769622184, "ewc_loss": 0.11181640625, "ewc_loss_diag": 0.00017547607421875, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.201202869415283, "learning_rate": 1e-05, "loss": 0.5992, "mean_token_accuracy": 0.8429037928581238, "num_tokens": 351205912.0, "step": 9250 }, { "epoch": 1.176822287240809, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.00017547607421875, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 6.004282474517822, "learning_rate": 1e-05, "loss": 0.4947, "mean_token_accuracy": 0.8678529262542725, "num_tokens": 351250515.0, "step": 9251 }, { "epoch": 1.1769494975193995, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.00017547607421875, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.184865951538086, "learning_rate": 1e-05, "loss": 0.4995, "mean_token_accuracy": 0.865944504737854, "num_tokens": 351284222.0, "step": 9252 }, { "epoch": 1.17707670779799, "ewc_loss": 0.10009765625, "ewc_loss_diag": 0.00017547607421875, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.763766288757324, "learning_rate": 1e-05, "loss": 0.5205, "mean_token_accuracy": 0.8582582473754883, "num_tokens": 351327799.0, "step": 9253 }, { "epoch": 1.1772039180765805, "ewc_loss": 0.10791015625, "ewc_loss_diag": 0.00017547607421875, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.20111608505249, "learning_rate": 1e-05, "loss": 0.4874, "mean_token_accuracy": 0.8700599074363708, "num_tokens": 351362194.0, "step": 9254 }, { "epoch": 1.177331128355171, "ewc_loss": 0.0966796875, "ewc_loss_diag": 0.00017547607421875, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.625837802886963, "learning_rate": 1e-05, "loss": 0.4633, "mean_token_accuracy": 0.8762282729148865, "num_tokens": 351395186.0, "step": 9255 }, { "epoch": 1.1774583386337616, "ewc_loss": 0.109375, "ewc_loss_diag": 0.00017547607421875, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.257011890411377, "learning_rate": 1e-05, "loss": 0.503, "mean_token_accuracy": 0.8700158596038818, "num_tokens": 351435484.0, "step": 9256 }, { "epoch": 1.1775855489123521, "ewc_loss": 0.0966796875, "ewc_loss_diag": 0.00017547607421875, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.497312068939209, "learning_rate": 1e-05, "loss": 0.4988, "mean_token_accuracy": 0.8595731854438782, "num_tokens": 351474613.0, "step": 9257 }, { "epoch": 1.1777127591909426, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.00017547607421875, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.429224491119385, "learning_rate": 1e-05, "loss": 0.4864, "mean_token_accuracy": 0.8757045865058899, "num_tokens": 351508612.0, "step": 9258 }, { "epoch": 1.1778399694695332, "ewc_loss": 0.095703125, "ewc_loss_diag": 0.00017547607421875, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.3868536949157715, "learning_rate": 1e-05, "loss": 0.5395, "mean_token_accuracy": 0.8558189868927002, "num_tokens": 351554811.0, "step": 9259 }, { "epoch": 1.1779671797481237, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.00017547607421875, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.70103120803833, "learning_rate": 1e-05, "loss": 0.4817, "mean_token_accuracy": 0.8712575435638428, "num_tokens": 351591540.0, "step": 9260 }, { "epoch": 1.1780943900267142, "ewc_loss": 0.09619140625, "ewc_loss_diag": 0.00017547607421875, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.370273590087891, "learning_rate": 1e-05, "loss": 0.4835, "mean_token_accuracy": 0.8677669167518616, "num_tokens": 351630529.0, "step": 9261 }, { "epoch": 1.1782216003053048, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.00017547607421875, "ewc_loss_parallel": 0.00011014938354492188, "grad_norm": 7.04790735244751, "learning_rate": 1e-05, "loss": 0.5514, "mean_token_accuracy": 0.856623113155365, "num_tokens": 351669412.0, "step": 9262 }, { "epoch": 1.1783488105838953, "ewc_loss": 0.10302734375, "ewc_loss_diag": 0.00017642974853515625, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.578380107879639, "learning_rate": 1e-05, "loss": 0.5094, "mean_token_accuracy": 0.8626243472099304, "num_tokens": 351711825.0, "step": 9263 }, { "epoch": 1.1784760208624856, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.00017642974853515625, "ewc_loss_parallel": 0.00011587142944335938, "grad_norm": 7.235493183135986, "learning_rate": 1e-05, "loss": 0.5159, "mean_token_accuracy": 0.8749934434890747, "num_tokens": 351747967.0, "step": 9264 }, { "epoch": 1.1786032311410761, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.00017642974853515625, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 5.973567485809326, "learning_rate": 1e-05, "loss": 0.4686, "mean_token_accuracy": 0.8753045797348022, "num_tokens": 351786208.0, "step": 9265 }, { "epoch": 1.1787304414196667, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.00017642974853515625, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 7.040663242340088, "learning_rate": 1e-05, "loss": 0.5168, "mean_token_accuracy": 0.8653756976127625, "num_tokens": 351825711.0, "step": 9266 }, { "epoch": 1.1788576516982572, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.00017642974853515625, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.200004577636719, "learning_rate": 1e-05, "loss": 0.4617, "mean_token_accuracy": 0.8799434900283813, "num_tokens": 351866772.0, "step": 9267 }, { "epoch": 1.1789848619768477, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.00017642974853515625, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.637241840362549, "learning_rate": 1e-05, "loss": 0.496, "mean_token_accuracy": 0.8711193799972534, "num_tokens": 351905812.0, "step": 9268 }, { "epoch": 1.1791120722554382, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.00017642974853515625, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.203524112701416, "learning_rate": 1e-05, "loss": 0.4784, "mean_token_accuracy": 0.8752188086509705, "num_tokens": 351950295.0, "step": 9269 }, { "epoch": 1.1792392825340288, "ewc_loss": 0.10888671875, "ewc_loss_diag": 0.00017642974853515625, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.239014625549316, "learning_rate": 1e-05, "loss": 0.5416, "mean_token_accuracy": 0.8554791212081909, "num_tokens": 351990980.0, "step": 9270 }, { "epoch": 1.1793664928126193, "ewc_loss": 0.11181640625, "ewc_loss_diag": 0.00017642974853515625, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.22121000289917, "learning_rate": 1e-05, "loss": 0.532, "mean_token_accuracy": 0.8552055358886719, "num_tokens": 352034483.0, "step": 9271 }, { "epoch": 1.1794937030912098, "ewc_loss": 0.10107421875, "ewc_loss_diag": 0.00017642974853515625, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.899726390838623, "learning_rate": 1e-05, "loss": 0.4461, "mean_token_accuracy": 0.8779420256614685, "num_tokens": 352069195.0, "step": 9272 }, { "epoch": 1.1796209133698004, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.00017642974853515625, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.193764686584473, "learning_rate": 1e-05, "loss": 0.4395, "mean_token_accuracy": 0.8802422285079956, "num_tokens": 352103393.0, "step": 9273 }, { "epoch": 1.1797481236483909, "ewc_loss": 0.09912109375, "ewc_loss_diag": 0.00017642974853515625, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.699926376342773, "learning_rate": 1e-05, "loss": 0.4528, "mean_token_accuracy": 0.8763784170150757, "num_tokens": 352138032.0, "step": 9274 }, { "epoch": 1.1798753339269812, "ewc_loss": 0.10888671875, "ewc_loss_diag": 0.00017642974853515625, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.540524482727051, "learning_rate": 1e-05, "loss": 0.5153, "mean_token_accuracy": 0.8595499396324158, "num_tokens": 352177243.0, "step": 9275 }, { "epoch": 1.1800025442055717, "ewc_loss": 0.09619140625, "ewc_loss_diag": 0.00017642974853515625, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.471071243286133, "learning_rate": 1e-05, "loss": 0.532, "mean_token_accuracy": 0.8518770933151245, "num_tokens": 352212461.0, "step": 9276 }, { "epoch": 1.1801297544841622, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.00017642974853515625, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.60826301574707, "learning_rate": 1e-05, "loss": 0.5308, "mean_token_accuracy": 0.8587007522583008, "num_tokens": 352250001.0, "step": 9277 }, { "epoch": 1.1802569647627528, "ewc_loss": 0.0966796875, "ewc_loss_diag": 0.00017642974853515625, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.381155490875244, "learning_rate": 1e-05, "loss": 0.5185, "mean_token_accuracy": 0.8550921678543091, "num_tokens": 352288893.0, "step": 9278 }, { "epoch": 1.1803841750413433, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.00017642974853515625, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 6.939165115356445, "learning_rate": 1e-05, "loss": 0.4985, "mean_token_accuracy": 0.8702629804611206, "num_tokens": 352328255.0, "step": 9279 }, { "epoch": 1.1805113853199338, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.00017642974853515625, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.520071029663086, "learning_rate": 1e-05, "loss": 0.4828, "mean_token_accuracy": 0.8677204847335815, "num_tokens": 352364871.0, "step": 9280 }, { "epoch": 1.1806385955985244, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.00017833709716796875, "ewc_loss_parallel": 0.00011396408081054688, "grad_norm": 7.157823085784912, "learning_rate": 1e-05, "loss": 0.4863, "mean_token_accuracy": 0.8794254064559937, "num_tokens": 352398894.0, "step": 9281 }, { "epoch": 1.1807658058771149, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.00017833709716796875, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 5.86150598526001, "learning_rate": 1e-05, "loss": 0.5181, "mean_token_accuracy": 0.8630340099334717, "num_tokens": 352438327.0, "step": 9282 }, { "epoch": 1.1808930161557054, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.00017833709716796875, "ewc_loss_parallel": 0.00011205673217773438, "grad_norm": 7.072567462921143, "learning_rate": 1e-05, "loss": 0.5492, "mean_token_accuracy": 0.8570998907089233, "num_tokens": 352475732.0, "step": 9283 }, { "epoch": 1.181020226434296, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.00017833709716796875, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.132815837860107, "learning_rate": 1e-05, "loss": 0.5613, "mean_token_accuracy": 0.8525824546813965, "num_tokens": 352509405.0, "step": 9284 }, { "epoch": 1.1811474367128865, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.00017833709716796875, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.718750476837158, "learning_rate": 1e-05, "loss": 0.5391, "mean_token_accuracy": 0.8588086366653442, "num_tokens": 352547172.0, "step": 9285 }, { "epoch": 1.181274646991477, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.00017833709716796875, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.175196170806885, "learning_rate": 1e-05, "loss": 0.5139, "mean_token_accuracy": 0.8671183586120605, "num_tokens": 352585705.0, "step": 9286 }, { "epoch": 1.1814018572700675, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.00017833709716796875, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.335585594177246, "learning_rate": 1e-05, "loss": 0.4703, "mean_token_accuracy": 0.8754018545150757, "num_tokens": 352629246.0, "step": 9287 }, { "epoch": 1.181529067548658, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.00017833709716796875, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.121626377105713, "learning_rate": 1e-05, "loss": 0.53, "mean_token_accuracy": 0.8599898815155029, "num_tokens": 352669504.0, "step": 9288 }, { "epoch": 1.1816562778272484, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.00017833709716796875, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 6.0510149002075195, "learning_rate": 1e-05, "loss": 0.508, "mean_token_accuracy": 0.8628376722335815, "num_tokens": 352708981.0, "step": 9289 }, { "epoch": 1.1817834881058389, "ewc_loss": 0.109375, "ewc_loss_diag": 0.00017833709716796875, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.053974151611328, "learning_rate": 1e-05, "loss": 0.5062, "mean_token_accuracy": 0.8644387125968933, "num_tokens": 352751646.0, "step": 9290 }, { "epoch": 1.1819106983844294, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.00017833709716796875, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.885527610778809, "learning_rate": 1e-05, "loss": 0.4958, "mean_token_accuracy": 0.8680069446563721, "num_tokens": 352789959.0, "step": 9291 }, { "epoch": 1.18203790866302, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.00017833709716796875, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 5.997671127319336, "learning_rate": 1e-05, "loss": 0.5374, "mean_token_accuracy": 0.8539719581604004, "num_tokens": 352828725.0, "step": 9292 }, { "epoch": 1.1821651189416105, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.00017833709716796875, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.794489860534668, "learning_rate": 1e-05, "loss": 0.4265, "mean_token_accuracy": 0.8871569037437439, "num_tokens": 352866373.0, "step": 9293 }, { "epoch": 1.182292329220201, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.00017833709716796875, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.953419208526611, "learning_rate": 1e-05, "loss": 0.4923, "mean_token_accuracy": 0.867424488067627, "num_tokens": 352908964.0, "step": 9294 }, { "epoch": 1.1824195394987915, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.00017833709716796875, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.737310409545898, "learning_rate": 1e-05, "loss": 0.4638, "mean_token_accuracy": 0.8761887550354004, "num_tokens": 352946292.0, "step": 9295 }, { "epoch": 1.182546749777382, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.00017833709716796875, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.931238651275635, "learning_rate": 1e-05, "loss": 0.4755, "mean_token_accuracy": 0.8707103729248047, "num_tokens": 352985875.0, "step": 9296 }, { "epoch": 1.1826739600559726, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.00017833709716796875, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.681145668029785, "learning_rate": 1e-05, "loss": 0.5073, "mean_token_accuracy": 0.8636716604232788, "num_tokens": 353032019.0, "step": 9297 }, { "epoch": 1.1828011703345631, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.000179290771484375, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 5.96748685836792, "learning_rate": 1e-05, "loss": 0.4963, "mean_token_accuracy": 0.8649789094924927, "num_tokens": 353062624.0, "step": 9298 }, { "epoch": 1.1829283806131534, "ewc_loss": 0.09814453125, "ewc_loss_diag": 0.000179290771484375, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.592740535736084, "learning_rate": 1e-05, "loss": 0.4727, "mean_token_accuracy": 0.8712442517280579, "num_tokens": 353105401.0, "step": 9299 }, { "epoch": 1.183055590891744, "ewc_loss": 0.10693359375, "ewc_loss_diag": 0.000179290771484375, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.0953288078308105, "learning_rate": 1e-05, "loss": 0.5462, "mean_token_accuracy": 0.851137638092041, "num_tokens": 353143180.0, "step": 9300 }, { "epoch": 1.1831828011703345, "ewc_loss": 0.09619140625, "ewc_loss_diag": 0.000179290771484375, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.4550909996032715, "learning_rate": 1e-05, "loss": 0.4746, "mean_token_accuracy": 0.8690110445022583, "num_tokens": 353177800.0, "step": 9301 }, { "epoch": 1.183310011448925, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.000179290771484375, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.2931952476501465, "learning_rate": 1e-05, "loss": 0.4873, "mean_token_accuracy": 0.8671745657920837, "num_tokens": 353221491.0, "step": 9302 }, { "epoch": 1.1834372217275155, "ewc_loss": 0.09423828125, "ewc_loss_diag": 0.000179290771484375, "ewc_loss_parallel": 7.62939453125e-05, "grad_norm": 5.274993419647217, "learning_rate": 1e-05, "loss": 0.5085, "mean_token_accuracy": 0.854505181312561, "num_tokens": 353263851.0, "step": 9303 }, { "epoch": 1.183564432006106, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.000179290771484375, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.724630355834961, "learning_rate": 1e-05, "loss": 0.5563, "mean_token_accuracy": 0.8546728491783142, "num_tokens": 353299197.0, "step": 9304 }, { "epoch": 1.1836916422846966, "ewc_loss": 0.09423828125, "ewc_loss_diag": 0.000179290771484375, "ewc_loss_parallel": 7.62939453125e-05, "grad_norm": 5.149156093597412, "learning_rate": 1e-05, "loss": 0.5492, "mean_token_accuracy": 0.8453887701034546, "num_tokens": 353339008.0, "step": 9305 }, { "epoch": 1.1838188525632871, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.000179290771484375, "ewc_loss_parallel": 0.0001201629638671875, "grad_norm": 7.3866868019104, "learning_rate": 1e-05, "loss": 0.524, "mean_token_accuracy": 0.8668773174285889, "num_tokens": 353378143.0, "step": 9306 }, { "epoch": 1.1839460628418776, "ewc_loss": 0.10498046875, "ewc_loss_diag": 0.000179290771484375, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.486417293548584, "learning_rate": 1e-05, "loss": 0.4309, "mean_token_accuracy": 0.8878430724143982, "num_tokens": 353412659.0, "step": 9307 }, { "epoch": 1.1840732731204682, "ewc_loss": 0.1484375, "ewc_loss_diag": 0.000179290771484375, "ewc_loss_parallel": 0.00013065338134765625, "grad_norm": 7.8695173263549805, "learning_rate": 1e-05, "loss": 0.5756, "mean_token_accuracy": 0.8544062972068787, "num_tokens": 353454459.0, "step": 9308 }, { "epoch": 1.1842004833990587, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.000179290771484375, "ewc_loss_parallel": 0.00011157989501953125, "grad_norm": 6.397595405578613, "learning_rate": 1e-05, "loss": 0.5367, "mean_token_accuracy": 0.8615090847015381, "num_tokens": 353501446.0, "step": 9309 }, { "epoch": 1.1843276936776492, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.000179290771484375, "ewc_loss_parallel": 0.0001201629638671875, "grad_norm": 7.485657691955566, "learning_rate": 1e-05, "loss": 0.5515, "mean_token_accuracy": 0.8615542054176331, "num_tokens": 353538129.0, "step": 9310 }, { "epoch": 1.1844549039562398, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.000179290771484375, "ewc_loss_parallel": 0.00011539459228515625, "grad_norm": 6.6878228187561035, "learning_rate": 1e-05, "loss": 0.4982, "mean_token_accuracy": 0.8776155710220337, "num_tokens": 353581488.0, "step": 9311 }, { "epoch": 1.1845821142348303, "ewc_loss": 0.125, "ewc_loss_diag": 0.000179290771484375, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.959658145904541, "learning_rate": 1e-05, "loss": 0.509, "mean_token_accuracy": 0.8687455058097839, "num_tokens": 353622729.0, "step": 9312 }, { "epoch": 1.1847093245134206, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.000179290771484375, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.460897445678711, "learning_rate": 1e-05, "loss": 0.4813, "mean_token_accuracy": 0.8759801983833313, "num_tokens": 353663696.0, "step": 9313 }, { "epoch": 1.1848365347920111, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.000179290771484375, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.6318039894104, "learning_rate": 1e-05, "loss": 0.5411, "mean_token_accuracy": 0.8565882444381714, "num_tokens": 353705934.0, "step": 9314 }, { "epoch": 1.1849637450706016, "ewc_loss": 0.11376953125, "ewc_loss_diag": 0.000179290771484375, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.221539497375488, "learning_rate": 1e-05, "loss": 0.4548, "mean_token_accuracy": 0.8848968744277954, "num_tokens": 353741535.0, "step": 9315 }, { "epoch": 1.1850909553491922, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.000179290771484375, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.400462627410889, "learning_rate": 1e-05, "loss": 0.4941, "mean_token_accuracy": 0.867427408695221, "num_tokens": 353784446.0, "step": 9316 }, { "epoch": 1.1852181656277827, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.000179290771484375, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 6.018032550811768, "learning_rate": 1e-05, "loss": 0.497, "mean_token_accuracy": 0.8656699657440186, "num_tokens": 353821955.0, "step": 9317 }, { "epoch": 1.1853453759063732, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.000179290771484375, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.313401222229004, "learning_rate": 1e-05, "loss": 0.5114, "mean_token_accuracy": 0.8668656349182129, "num_tokens": 353859994.0, "step": 9318 }, { "epoch": 1.1854725861849638, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.000179290771484375, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.807925701141357, "learning_rate": 1e-05, "loss": 0.4797, "mean_token_accuracy": 0.8732309341430664, "num_tokens": 353903301.0, "step": 9319 }, { "epoch": 1.1855997964635543, "ewc_loss": 0.11083984375, "ewc_loss_diag": 0.000179290771484375, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.351683139801025, "learning_rate": 1e-05, "loss": 0.5729, "mean_token_accuracy": 0.8438693284988403, "num_tokens": 353944527.0, "step": 9320 }, { "epoch": 1.1857270067421448, "ewc_loss": 0.09814453125, "ewc_loss_diag": 0.000179290771484375, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.637294292449951, "learning_rate": 1e-05, "loss": 0.4502, "mean_token_accuracy": 0.8760735988616943, "num_tokens": 353977402.0, "step": 9321 }, { "epoch": 1.1858542170207353, "ewc_loss": 0.11474609375, "ewc_loss_diag": 0.000179290771484375, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.51885461807251, "learning_rate": 1e-05, "loss": 0.4804, "mean_token_accuracy": 0.8750436902046204, "num_tokens": 354016681.0, "step": 9322 }, { "epoch": 1.1859814272993259, "ewc_loss": 0.09619140625, "ewc_loss_diag": 0.000179290771484375, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.4915547370910645, "learning_rate": 1e-05, "loss": 0.5262, "mean_token_accuracy": 0.8561490774154663, "num_tokens": 354057923.0, "step": 9323 }, { "epoch": 1.1861086375779162, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.000179290771484375, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.841944217681885, "learning_rate": 1e-05, "loss": 0.5085, "mean_token_accuracy": 0.8664911389350891, "num_tokens": 354097787.0, "step": 9324 }, { "epoch": 1.1862358478565067, "ewc_loss": 0.09765625, "ewc_loss_diag": 0.000179290771484375, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.499566555023193, "learning_rate": 1e-05, "loss": 0.466, "mean_token_accuracy": 0.8748845458030701, "num_tokens": 354136694.0, "step": 9325 }, { "epoch": 1.1863630581350972, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.000179290771484375, "ewc_loss_parallel": 0.00011157989501953125, "grad_norm": 7.1884989738464355, "learning_rate": 1e-05, "loss": 0.4628, "mean_token_accuracy": 0.8857754468917847, "num_tokens": 354176326.0, "step": 9326 }, { "epoch": 1.1864902684136878, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.000179290771484375, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 5.842103958129883, "learning_rate": 1e-05, "loss": 0.5068, "mean_token_accuracy": 0.8625295162200928, "num_tokens": 354214350.0, "step": 9327 }, { "epoch": 1.1866174786922783, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.000179290771484375, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 7.175790786743164, "learning_rate": 1e-05, "loss": 0.5105, "mean_token_accuracy": 0.8663088083267212, "num_tokens": 354252910.0, "step": 9328 }, { "epoch": 1.1867446889708688, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.000179290771484375, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.250854969024658, "learning_rate": 1e-05, "loss": 0.487, "mean_token_accuracy": 0.8681643009185791, "num_tokens": 354291105.0, "step": 9329 }, { "epoch": 1.1868718992494594, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.000179290771484375, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.719126224517822, "learning_rate": 1e-05, "loss": 0.4926, "mean_token_accuracy": 0.8727656602859497, "num_tokens": 354329767.0, "step": 9330 }, { "epoch": 1.1869991095280499, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.000179290771484375, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.36346960067749, "learning_rate": 1e-05, "loss": 0.4684, "mean_token_accuracy": 0.8785766363143921, "num_tokens": 354369826.0, "step": 9331 }, { "epoch": 1.1871263198066404, "ewc_loss": 0.10498046875, "ewc_loss_diag": 0.000179290771484375, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 6.158597469329834, "learning_rate": 1e-05, "loss": 0.4536, "mean_token_accuracy": 0.8796075582504272, "num_tokens": 354406088.0, "step": 9332 }, { "epoch": 1.187253530085231, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.000179290771484375, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.415412902832031, "learning_rate": 1e-05, "loss": 0.4606, "mean_token_accuracy": 0.8804975152015686, "num_tokens": 354445675.0, "step": 9333 }, { "epoch": 1.1873807403638215, "ewc_loss": 0.0966796875, "ewc_loss_diag": 0.000179290771484375, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.683782577514648, "learning_rate": 1e-05, "loss": 0.5292, "mean_token_accuracy": 0.8492913246154785, "num_tokens": 354482733.0, "step": 9334 }, { "epoch": 1.187507950642412, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.000179290771484375, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.631254196166992, "learning_rate": 1e-05, "loss": 0.5191, "mean_token_accuracy": 0.8606823682785034, "num_tokens": 354515483.0, "step": 9335 }, { "epoch": 1.1876351609210025, "ewc_loss": 0.0947265625, "ewc_loss_diag": 0.000179290771484375, "ewc_loss_parallel": 7.677078247070312e-05, "grad_norm": 5.409203052520752, "learning_rate": 1e-05, "loss": 0.4504, "mean_token_accuracy": 0.8756129145622253, "num_tokens": 354551972.0, "step": 9336 }, { "epoch": 1.187762371199593, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.000179290771484375, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 7.018377780914307, "learning_rate": 1e-05, "loss": 0.5079, "mean_token_accuracy": 0.8683135509490967, "num_tokens": 354592089.0, "step": 9337 }, { "epoch": 1.1878895814781834, "ewc_loss": 0.0986328125, "ewc_loss_diag": 0.00018024444580078125, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.49359655380249, "learning_rate": 1e-05, "loss": 0.4573, "mean_token_accuracy": 0.8778014183044434, "num_tokens": 354629582.0, "step": 9338 }, { "epoch": 1.1880167917567739, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.000179290771484375, "ewc_loss_parallel": 0.00011777877807617188, "grad_norm": 7.36042594909668, "learning_rate": 1e-05, "loss": 0.5988, "mean_token_accuracy": 0.8452649116516113, "num_tokens": 354677114.0, "step": 9339 }, { "epoch": 1.1881440020353644, "ewc_loss": 0.10986328125, "ewc_loss_diag": 0.00018024444580078125, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 5.967358112335205, "learning_rate": 1e-05, "loss": 0.5132, "mean_token_accuracy": 0.8645431995391846, "num_tokens": 354723125.0, "step": 9340 }, { "epoch": 1.188271212313955, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.00018024444580078125, "ewc_loss_parallel": 0.00011205673217773438, "grad_norm": 7.240233898162842, "learning_rate": 1e-05, "loss": 0.555, "mean_token_accuracy": 0.858134925365448, "num_tokens": 354761032.0, "step": 9341 }, { "epoch": 1.1883984225925455, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.00018024444580078125, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.31465482711792, "learning_rate": 1e-05, "loss": 0.4795, "mean_token_accuracy": 0.8775532245635986, "num_tokens": 354798522.0, "step": 9342 }, { "epoch": 1.188525632871136, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.00018024444580078125, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.835575580596924, "learning_rate": 1e-05, "loss": 0.5058, "mean_token_accuracy": 0.8701374530792236, "num_tokens": 354833752.0, "step": 9343 }, { "epoch": 1.1886528431497265, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.00018024444580078125, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.298944473266602, "learning_rate": 1e-05, "loss": 0.4889, "mean_token_accuracy": 0.8727114200592041, "num_tokens": 354869901.0, "step": 9344 }, { "epoch": 1.188780053428317, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.00018024444580078125, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.476859092712402, "learning_rate": 1e-05, "loss": 0.4916, "mean_token_accuracy": 0.8677811026573181, "num_tokens": 354902487.0, "step": 9345 }, { "epoch": 1.1889072637069076, "ewc_loss": 0.109375, "ewc_loss_diag": 0.00018024444580078125, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.148409843444824, "learning_rate": 1e-05, "loss": 0.5594, "mean_token_accuracy": 0.8459091186523438, "num_tokens": 354948778.0, "step": 9346 }, { "epoch": 1.189034473985498, "ewc_loss": 0.10791015625, "ewc_loss_diag": 0.00018024444580078125, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.249812126159668, "learning_rate": 1e-05, "loss": 0.4867, "mean_token_accuracy": 0.8680005073547363, "num_tokens": 354990242.0, "step": 9347 }, { "epoch": 1.1891616842640884, "ewc_loss": 0.10302734375, "ewc_loss_diag": 0.00018024444580078125, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.958633899688721, "learning_rate": 1e-05, "loss": 0.4392, "mean_token_accuracy": 0.8806725740432739, "num_tokens": 355025462.0, "step": 9348 }, { "epoch": 1.189288894542679, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.00018024444580078125, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 6.191011428833008, "learning_rate": 1e-05, "loss": 0.535, "mean_token_accuracy": 0.8578473925590515, "num_tokens": 355067488.0, "step": 9349 }, { "epoch": 1.1894161048212695, "ewc_loss": 0.09912109375, "ewc_loss_diag": 0.00018024444580078125, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.773902416229248, "learning_rate": 1e-05, "loss": 0.5014, "mean_token_accuracy": 0.8634284138679504, "num_tokens": 355101734.0, "step": 9350 }, { "epoch": 1.18954331509986, "ewc_loss": 0.10986328125, "ewc_loss_diag": 0.00018024444580078125, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.476097583770752, "learning_rate": 1e-05, "loss": 0.5276, "mean_token_accuracy": 0.8575266003608704, "num_tokens": 355140209.0, "step": 9351 }, { "epoch": 1.1896705253784505, "ewc_loss": 0.09619140625, "ewc_loss_diag": 0.00018024444580078125, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.541968822479248, "learning_rate": 1e-05, "loss": 0.4883, "mean_token_accuracy": 0.8660213351249695, "num_tokens": 355177424.0, "step": 9352 }, { "epoch": 1.189797735657041, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.00018024444580078125, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.625671863555908, "learning_rate": 1e-05, "loss": 0.4871, "mean_token_accuracy": 0.8693997859954834, "num_tokens": 355215112.0, "step": 9353 }, { "epoch": 1.1899249459356316, "ewc_loss": 0.0966796875, "ewc_loss_diag": 0.00018024444580078125, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.442745208740234, "learning_rate": 1e-05, "loss": 0.4508, "mean_token_accuracy": 0.8798338174819946, "num_tokens": 355250402.0, "step": 9354 }, { "epoch": 1.1900521562142221, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.00018024444580078125, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.9639177322387695, "learning_rate": 1e-05, "loss": 0.4912, "mean_token_accuracy": 0.8712223172187805, "num_tokens": 355289887.0, "step": 9355 }, { "epoch": 1.1901793664928126, "ewc_loss": 0.10009765625, "ewc_loss_diag": 0.00018024444580078125, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.604890823364258, "learning_rate": 1e-05, "loss": 0.4773, "mean_token_accuracy": 0.8690444827079773, "num_tokens": 355330946.0, "step": 9356 }, { "epoch": 1.1903065767714032, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.00018024444580078125, "ewc_loss_parallel": 0.00011157989501953125, "grad_norm": 7.144959926605225, "learning_rate": 1e-05, "loss": 0.5081, "mean_token_accuracy": 0.8685367107391357, "num_tokens": 355369029.0, "step": 9357 }, { "epoch": 1.1904337870499937, "ewc_loss": 0.11181640625, "ewc_loss_diag": 0.00018024444580078125, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 5.9970245361328125, "learning_rate": 1e-05, "loss": 0.5711, "mean_token_accuracy": 0.8471682071685791, "num_tokens": 355406145.0, "step": 9358 }, { "epoch": 1.1905609973285842, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.00018024444580078125, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.899137496948242, "learning_rate": 1e-05, "loss": 0.4431, "mean_token_accuracy": 0.8893048763275146, "num_tokens": 355445702.0, "step": 9359 }, { "epoch": 1.1906882076071748, "ewc_loss": 0.11376953125, "ewc_loss_diag": 0.00018024444580078125, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.221286296844482, "learning_rate": 1e-05, "loss": 0.5039, "mean_token_accuracy": 0.8640302419662476, "num_tokens": 355485592.0, "step": 9360 }, { "epoch": 1.1908154178857653, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.00018024444580078125, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.413938999176025, "learning_rate": 1e-05, "loss": 0.4633, "mean_token_accuracy": 0.8828949332237244, "num_tokens": 355523208.0, "step": 9361 }, { "epoch": 1.1909426281643556, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.00018024444580078125, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.287696838378906, "learning_rate": 1e-05, "loss": 0.5003, "mean_token_accuracy": 0.8689053058624268, "num_tokens": 355557729.0, "step": 9362 }, { "epoch": 1.1910698384429461, "ewc_loss": 0.10107421875, "ewc_loss_diag": 0.00018024444580078125, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.9039506912231445, "learning_rate": 1e-05, "loss": 0.4482, "mean_token_accuracy": 0.8798646926879883, "num_tokens": 355597222.0, "step": 9363 }, { "epoch": 1.1911970487215366, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.00018024444580078125, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.411110877990723, "learning_rate": 1e-05, "loss": 0.4547, "mean_token_accuracy": 0.8842753171920776, "num_tokens": 355630477.0, "step": 9364 }, { "epoch": 1.1913242590001272, "ewc_loss": 0.0966796875, "ewc_loss_diag": 0.00018024444580078125, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.534291744232178, "learning_rate": 1e-05, "loss": 0.4807, "mean_token_accuracy": 0.8670133352279663, "num_tokens": 355664506.0, "step": 9365 }, { "epoch": 1.1914514692787177, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.00018024444580078125, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.6873908042907715, "learning_rate": 1e-05, "loss": 0.5044, "mean_token_accuracy": 0.8682142496109009, "num_tokens": 355710268.0, "step": 9366 }, { "epoch": 1.1915786795573082, "ewc_loss": 0.09521484375, "ewc_loss_diag": 0.00018024444580078125, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.379563331604004, "learning_rate": 1e-05, "loss": 0.5435, "mean_token_accuracy": 0.8522164821624756, "num_tokens": 355746396.0, "step": 9367 }, { "epoch": 1.1917058898358988, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.00018024444580078125, "ewc_loss_parallel": 0.00010967254638671875, "grad_norm": 7.079308032989502, "learning_rate": 1e-05, "loss": 0.5394, "mean_token_accuracy": 0.8581485748291016, "num_tokens": 355787483.0, "step": 9368 }, { "epoch": 1.1918331001144893, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.00018024444580078125, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.555955410003662, "learning_rate": 1e-05, "loss": 0.5103, "mean_token_accuracy": 0.861065149307251, "num_tokens": 355824138.0, "step": 9369 }, { "epoch": 1.1919603103930798, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.00018024444580078125, "ewc_loss_parallel": 0.00011730194091796875, "grad_norm": 7.341155052185059, "learning_rate": 1e-05, "loss": 0.5414, "mean_token_accuracy": 0.857880175113678, "num_tokens": 355854258.0, "step": 9370 }, { "epoch": 1.1920875206716703, "ewc_loss": 0.11376953125, "ewc_loss_diag": 0.00018024444580078125, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.042718410491943, "learning_rate": 1e-05, "loss": 0.4961, "mean_token_accuracy": 0.8661718368530273, "num_tokens": 355889750.0, "step": 9371 }, { "epoch": 1.1922147309502609, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.00018024444580078125, "ewc_loss_parallel": 0.00011014938354492188, "grad_norm": 7.118516445159912, "learning_rate": 1e-05, "loss": 0.4832, "mean_token_accuracy": 0.8775752782821655, "num_tokens": 355923950.0, "step": 9372 }, { "epoch": 1.1923419412288512, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.00018024444580078125, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.281536102294922, "learning_rate": 1e-05, "loss": 0.4833, "mean_token_accuracy": 0.8750576972961426, "num_tokens": 355959807.0, "step": 9373 }, { "epoch": 1.1924691515074417, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.00018024444580078125, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.720338821411133, "learning_rate": 1e-05, "loss": 0.4712, "mean_token_accuracy": 0.8756150603294373, "num_tokens": 355992574.0, "step": 9374 }, { "epoch": 1.1925963617860322, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.00018024444580078125, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.235189914703369, "learning_rate": 1e-05, "loss": 0.5304, "mean_token_accuracy": 0.8595929741859436, "num_tokens": 356033726.0, "step": 9375 }, { "epoch": 1.1927235720646228, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.00018024444580078125, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.383045673370361, "learning_rate": 1e-05, "loss": 0.4647, "mean_token_accuracy": 0.8786097764968872, "num_tokens": 356066959.0, "step": 9376 }, { "epoch": 1.1928507823432133, "ewc_loss": 0.10986328125, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.107548713684082, "learning_rate": 1e-05, "loss": 0.4822, "mean_token_accuracy": 0.8701421022415161, "num_tokens": 356102316.0, "step": 9377 }, { "epoch": 1.1929779926218038, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.00018024444580078125, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 6.157529830932617, "learning_rate": 1e-05, "loss": 0.4518, "mean_token_accuracy": 0.8794633150100708, "num_tokens": 356146203.0, "step": 9378 }, { "epoch": 1.1931052029003943, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.00018024444580078125, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.965591907501221, "learning_rate": 1e-05, "loss": 0.5201, "mean_token_accuracy": 0.8568601608276367, "num_tokens": 356184673.0, "step": 9379 }, { "epoch": 1.1932324131789849, "ewc_loss": 0.10595703125, "ewc_loss_diag": 0.00018024444580078125, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 6.107196807861328, "learning_rate": 1e-05, "loss": 0.4929, "mean_token_accuracy": 0.8656182289123535, "num_tokens": 356224693.0, "step": 9380 }, { "epoch": 1.1933596234575754, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.00018024444580078125, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.793121337890625, "learning_rate": 1e-05, "loss": 0.5026, "mean_token_accuracy": 0.8612574338912964, "num_tokens": 356265454.0, "step": 9381 }, { "epoch": 1.193486833736166, "ewc_loss": 0.10791015625, "ewc_loss_diag": 0.00018024444580078125, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.154305934906006, "learning_rate": 1e-05, "loss": 0.4884, "mean_token_accuracy": 0.868443489074707, "num_tokens": 356302866.0, "step": 9382 }, { "epoch": 1.1936140440147565, "ewc_loss": 0.09765625, "ewc_loss_diag": 0.00018024444580078125, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.622890472412109, "learning_rate": 1e-05, "loss": 0.4673, "mean_token_accuracy": 0.8712215423583984, "num_tokens": 356341897.0, "step": 9383 }, { "epoch": 1.193741254293347, "ewc_loss": 0.11083984375, "ewc_loss_diag": 0.00018024444580078125, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.310637474060059, "learning_rate": 1e-05, "loss": 0.4826, "mean_token_accuracy": 0.8745633363723755, "num_tokens": 356380637.0, "step": 9384 }, { "epoch": 1.1938684645719375, "ewc_loss": 0.095703125, "ewc_loss_diag": 0.00018024444580078125, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.4674601554870605, "learning_rate": 1e-05, "loss": 0.5088, "mean_token_accuracy": 0.8578082919120789, "num_tokens": 356422584.0, "step": 9385 }, { "epoch": 1.193995674850528, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.00018024444580078125, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.595433712005615, "learning_rate": 1e-05, "loss": 0.4801, "mean_token_accuracy": 0.8757288455963135, "num_tokens": 356457007.0, "step": 9386 }, { "epoch": 1.1941228851291183, "ewc_loss": 0.095703125, "ewc_loss_diag": 0.00018024444580078125, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.382472991943359, "learning_rate": 1e-05, "loss": 0.475, "mean_token_accuracy": 0.8709039688110352, "num_tokens": 356491837.0, "step": 9387 }, { "epoch": 1.1942500954077089, "ewc_loss": 0.125, "ewc_loss_diag": 0.00018024444580078125, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.975154876708984, "learning_rate": 1e-05, "loss": 0.5125, "mean_token_accuracy": 0.8693755865097046, "num_tokens": 356527892.0, "step": 9388 }, { "epoch": 1.1943773056862994, "ewc_loss": 0.099609375, "ewc_loss_diag": 0.00018024444580078125, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.523348808288574, "learning_rate": 1e-05, "loss": 0.4863, "mean_token_accuracy": 0.8626308441162109, "num_tokens": 356564310.0, "step": 9389 }, { "epoch": 1.19450451596489, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.00018024444580078125, "ewc_loss_parallel": 0.00011539459228515625, "grad_norm": 7.2343950271606445, "learning_rate": 1e-05, "loss": 0.4899, "mean_token_accuracy": 0.8760423064231873, "num_tokens": 356606476.0, "step": 9390 }, { "epoch": 1.1946317262434805, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.00018024444580078125, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 5.9428277015686035, "learning_rate": 1e-05, "loss": 0.4826, "mean_token_accuracy": 0.8668491840362549, "num_tokens": 356646104.0, "step": 9391 }, { "epoch": 1.194758936522071, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.00018024444580078125, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 7.050628662109375, "learning_rate": 1e-05, "loss": 0.4538, "mean_token_accuracy": 0.8845112323760986, "num_tokens": 356687196.0, "step": 9392 }, { "epoch": 1.1948861468006615, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.00018024444580078125, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.246766567230225, "learning_rate": 1e-05, "loss": 0.5089, "mean_token_accuracy": 0.8672625422477722, "num_tokens": 356717634.0, "step": 9393 }, { "epoch": 1.195013357079252, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.00018024444580078125, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.583404541015625, "learning_rate": 1e-05, "loss": 0.5481, "mean_token_accuracy": 0.8550561666488647, "num_tokens": 356755494.0, "step": 9394 }, { "epoch": 1.1951405673578426, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.00018024444580078125, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.281617641448975, "learning_rate": 1e-05, "loss": 0.4786, "mean_token_accuracy": 0.8768166303634644, "num_tokens": 356800798.0, "step": 9395 }, { "epoch": 1.195267777636433, "ewc_loss": 0.10498046875, "ewc_loss_diag": 0.00018024444580078125, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 6.095938682556152, "learning_rate": 1e-05, "loss": 0.431, "mean_token_accuracy": 0.8896624445915222, "num_tokens": 356837819.0, "step": 9396 }, { "epoch": 1.1953949879150234, "ewc_loss": 0.11376953125, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.33922815322876, "learning_rate": 1e-05, "loss": 0.4793, "mean_token_accuracy": 0.8745659589767456, "num_tokens": 356876371.0, "step": 9397 }, { "epoch": 1.195522198193614, "ewc_loss": 0.0986328125, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.70897912979126, "learning_rate": 1e-05, "loss": 0.5047, "mean_token_accuracy": 0.8690602779388428, "num_tokens": 356911427.0, "step": 9398 }, { "epoch": 1.1956494084722045, "ewc_loss": 0.11474609375, "ewc_loss_diag": 0.00018024444580078125, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.490623950958252, "learning_rate": 1e-05, "loss": 0.4658, "mean_token_accuracy": 0.8764244318008423, "num_tokens": 356951229.0, "step": 9399 }, { "epoch": 1.195776618750795, "ewc_loss": 0.0966796875, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.485701084136963, "learning_rate": 1e-05, "loss": 0.448, "mean_token_accuracy": 0.8759623765945435, "num_tokens": 356988954.0, "step": 9400 }, { "epoch": 1.1959038290293855, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.766712188720703, "learning_rate": 1e-05, "loss": 0.5424, "mean_token_accuracy": 0.8532335758209229, "num_tokens": 357027982.0, "step": 9401 }, { "epoch": 1.196031039307976, "ewc_loss": 0.09814453125, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.47190523147583, "learning_rate": 1e-05, "loss": 0.4889, "mean_token_accuracy": 0.8667966723442078, "num_tokens": 357061434.0, "step": 9402 }, { "epoch": 1.1961582495865666, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 7.04888391494751, "learning_rate": 1e-05, "loss": 0.5085, "mean_token_accuracy": 0.8660175800323486, "num_tokens": 357099255.0, "step": 9403 }, { "epoch": 1.196285459865157, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.69443416595459, "learning_rate": 1e-05, "loss": 0.4585, "mean_token_accuracy": 0.8762500882148743, "num_tokens": 357131917.0, "step": 9404 }, { "epoch": 1.1964126701437476, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 7.128363609313965, "learning_rate": 1e-05, "loss": 0.4862, "mean_token_accuracy": 0.8774582743644714, "num_tokens": 357171520.0, "step": 9405 }, { "epoch": 1.1965398804223382, "ewc_loss": 0.11083984375, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 5.989193916320801, "learning_rate": 1e-05, "loss": 0.4935, "mean_token_accuracy": 0.8690452575683594, "num_tokens": 357208517.0, "step": 9406 }, { "epoch": 1.1966670907009287, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.9282379150390625, "learning_rate": 1e-05, "loss": 0.4688, "mean_token_accuracy": 0.8788323402404785, "num_tokens": 357245300.0, "step": 9407 }, { "epoch": 1.1967943009795192, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.116183757781982, "learning_rate": 1e-05, "loss": 0.4597, "mean_token_accuracy": 0.8830652832984924, "num_tokens": 357284765.0, "step": 9408 }, { "epoch": 1.1969215112581097, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.621921062469482, "learning_rate": 1e-05, "loss": 0.5363, "mean_token_accuracy": 0.8593487739562988, "num_tokens": 357322968.0, "step": 9409 }, { "epoch": 1.1970487215367003, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.095017433166504, "learning_rate": 1e-05, "loss": 0.4826, "mean_token_accuracy": 0.8722225427627563, "num_tokens": 357360546.0, "step": 9410 }, { "epoch": 1.1971759318152906, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.3339972496032715, "learning_rate": 1e-05, "loss": 0.4431, "mean_token_accuracy": 0.8828150033950806, "num_tokens": 357394433.0, "step": 9411 }, { "epoch": 1.1973031420938811, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 6.0202155113220215, "learning_rate": 1e-05, "loss": 0.4823, "mean_token_accuracy": 0.8666212558746338, "num_tokens": 357431498.0, "step": 9412 }, { "epoch": 1.1974303523724716, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 6.147655487060547, "learning_rate": 1e-05, "loss": 0.4977, "mean_token_accuracy": 0.8635820150375366, "num_tokens": 357466930.0, "step": 9413 }, { "epoch": 1.1975575626510622, "ewc_loss": 0.10302734375, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.909728527069092, "learning_rate": 1e-05, "loss": 0.5355, "mean_token_accuracy": 0.8584641814231873, "num_tokens": 357509343.0, "step": 9414 }, { "epoch": 1.1976847729296527, "ewc_loss": 0.10498046875, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 6.084722518920898, "learning_rate": 1e-05, "loss": 0.4835, "mean_token_accuracy": 0.8706874251365662, "num_tokens": 357541980.0, "step": 9415 }, { "epoch": 1.1978119832082432, "ewc_loss": 0.10009765625, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.763638973236084, "learning_rate": 1e-05, "loss": 0.4615, "mean_token_accuracy": 0.8765623569488525, "num_tokens": 357579380.0, "step": 9416 }, { "epoch": 1.1979391934868338, "ewc_loss": 0.10693359375, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.132882118225098, "learning_rate": 1e-05, "loss": 0.4508, "mean_token_accuracy": 0.8795070052146912, "num_tokens": 357614666.0, "step": 9417 }, { "epoch": 1.1980664037654243, "ewc_loss": 0.09765625, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.625122547149658, "learning_rate": 1e-05, "loss": 0.4605, "mean_token_accuracy": 0.8754333257675171, "num_tokens": 357648532.0, "step": 9418 }, { "epoch": 1.1981936140440148, "ewc_loss": 0.10986328125, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.272054195404053, "learning_rate": 1e-05, "loss": 0.4688, "mean_token_accuracy": 0.8734397888183594, "num_tokens": 357683585.0, "step": 9419 }, { "epoch": 1.1983208243226053, "ewc_loss": 0.095703125, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.468583106994629, "learning_rate": 1e-05, "loss": 0.5395, "mean_token_accuracy": 0.8457419872283936, "num_tokens": 357725580.0, "step": 9420 }, { "epoch": 1.1984480346011959, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.536863803863525, "learning_rate": 1e-05, "loss": 0.4915, "mean_token_accuracy": 0.8670291900634766, "num_tokens": 357763017.0, "step": 9421 }, { "epoch": 1.1985752448797862, "ewc_loss": 0.09521484375, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.373542785644531, "learning_rate": 1e-05, "loss": 0.4462, "mean_token_accuracy": 0.8793908357620239, "num_tokens": 357797433.0, "step": 9422 }, { "epoch": 1.1987024551583767, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.894053936004639, "learning_rate": 1e-05, "loss": 0.5117, "mean_token_accuracy": 0.8665231466293335, "num_tokens": 357835348.0, "step": 9423 }, { "epoch": 1.1988296654369672, "ewc_loss": 0.09912109375, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.472893714904785, "learning_rate": 1e-05, "loss": 0.5002, "mean_token_accuracy": 0.8627158403396606, "num_tokens": 357877376.0, "step": 9424 }, { "epoch": 1.1989568757155578, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 0.00011444091796875, "grad_norm": 7.184537410736084, "learning_rate": 1e-05, "loss": 0.5147, "mean_token_accuracy": 0.8677107095718384, "num_tokens": 357912983.0, "step": 9425 }, { "epoch": 1.1990840859941483, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 5.859114646911621, "learning_rate": 1e-05, "loss": 0.4591, "mean_token_accuracy": 0.8768982291221619, "num_tokens": 357948068.0, "step": 9426 }, { "epoch": 1.1992112962727388, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 0.00010967254638671875, "grad_norm": 7.069095134735107, "learning_rate": 1e-05, "loss": 0.4951, "mean_token_accuracy": 0.874200165271759, "num_tokens": 357984019.0, "step": 9427 }, { "epoch": 1.1993385065513293, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.176865577697754, "learning_rate": 1e-05, "loss": 0.4788, "mean_token_accuracy": 0.875880777835846, "num_tokens": 358022440.0, "step": 9428 }, { "epoch": 1.1994657168299199, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.635815620422363, "learning_rate": 1e-05, "loss": 0.5071, "mean_token_accuracy": 0.8670957684516907, "num_tokens": 358061615.0, "step": 9429 }, { "epoch": 1.1995929271085104, "ewc_loss": 0.11474609375, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.2646965980529785, "learning_rate": 1e-05, "loss": 0.5132, "mean_token_accuracy": 0.8651341199874878, "num_tokens": 358097042.0, "step": 9430 }, { "epoch": 1.199720137387101, "ewc_loss": 0.10693359375, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.1361260414123535, "learning_rate": 1e-05, "loss": 0.5148, "mean_token_accuracy": 0.8638879060745239, "num_tokens": 358134067.0, "step": 9431 }, { "epoch": 1.1998473476656915, "ewc_loss": 0.11376953125, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.320825099945068, "learning_rate": 1e-05, "loss": 0.4982, "mean_token_accuracy": 0.8682605028152466, "num_tokens": 358178975.0, "step": 9432 }, { "epoch": 1.199974557944282, "ewc_loss": 0.099609375, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.7342095375061035, "learning_rate": 1e-05, "loss": 0.4687, "mean_token_accuracy": 0.8716742992401123, "num_tokens": 358212879.0, "step": 9433 }, { "epoch": 1.2001017682228725, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.490940093994141, "learning_rate": 1e-05, "loss": 0.4624, "mean_token_accuracy": 0.8817548751831055, "num_tokens": 358245711.0, "step": 9434 }, { "epoch": 1.200228978501463, "ewc_loss": 0.09619140625, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.484457492828369, "learning_rate": 1e-05, "loss": 0.4382, "mean_token_accuracy": 0.8816787600517273, "num_tokens": 358283073.0, "step": 9435 }, { "epoch": 1.2003561887800533, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.757303237915039, "learning_rate": 1e-05, "loss": 0.4721, "mean_token_accuracy": 0.8771144151687622, "num_tokens": 358320338.0, "step": 9436 }, { "epoch": 1.2004833990586439, "ewc_loss": 0.0986328125, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.478538990020752, "learning_rate": 1e-05, "loss": 0.488, "mean_token_accuracy": 0.8674952983856201, "num_tokens": 358358704.0, "step": 9437 }, { "epoch": 1.2006106093372344, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 7.019314765930176, "learning_rate": 1e-05, "loss": 0.4905, "mean_token_accuracy": 0.87603759765625, "num_tokens": 358393512.0, "step": 9438 }, { "epoch": 1.200737819615825, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.682201385498047, "learning_rate": 1e-05, "loss": 0.5214, "mean_token_accuracy": 0.8596730828285217, "num_tokens": 358435843.0, "step": 9439 }, { "epoch": 1.2008650298944155, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 7.089045524597168, "learning_rate": 1e-05, "loss": 0.4477, "mean_token_accuracy": 0.8852005004882812, "num_tokens": 358468918.0, "step": 9440 }, { "epoch": 1.200992240173006, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 5.96467924118042, "learning_rate": 1e-05, "loss": 0.5654, "mean_token_accuracy": 0.8488686084747314, "num_tokens": 358511275.0, "step": 9441 }, { "epoch": 1.2011194504515965, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.89533805847168, "learning_rate": 1e-05, "loss": 0.5348, "mean_token_accuracy": 0.8610157370567322, "num_tokens": 358552665.0, "step": 9442 }, { "epoch": 1.201246660730187, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.082643985748291, "learning_rate": 1e-05, "loss": 0.4621, "mean_token_accuracy": 0.876090407371521, "num_tokens": 358584725.0, "step": 9443 }, { "epoch": 1.2013738710087776, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.6153717041015625, "learning_rate": 1e-05, "loss": 0.4834, "mean_token_accuracy": 0.8733124732971191, "num_tokens": 358620026.0, "step": 9444 }, { "epoch": 1.201501081287368, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.050793647766113, "learning_rate": 1e-05, "loss": 0.4989, "mean_token_accuracy": 0.8669052124023438, "num_tokens": 358657386.0, "step": 9445 }, { "epoch": 1.2016282915659584, "ewc_loss": 0.11181640625, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.372128963470459, "learning_rate": 1e-05, "loss": 0.4913, "mean_token_accuracy": 0.871823787689209, "num_tokens": 358695069.0, "step": 9446 }, { "epoch": 1.201755501844549, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 5.9689812660217285, "learning_rate": 1e-05, "loss": 0.4613, "mean_token_accuracy": 0.8766371607780457, "num_tokens": 358728091.0, "step": 9447 }, { "epoch": 1.2018827121231395, "ewc_loss": 0.109375, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.223729133605957, "learning_rate": 1e-05, "loss": 0.5328, "mean_token_accuracy": 0.8554012775421143, "num_tokens": 358767930.0, "step": 9448 }, { "epoch": 1.20200992240173, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.8518147468566895, "learning_rate": 1e-05, "loss": 0.4959, "mean_token_accuracy": 0.8657475709915161, "num_tokens": 358803705.0, "step": 9449 }, { "epoch": 1.2021371326803205, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 6.185458183288574, "learning_rate": 1e-05, "loss": 0.4857, "mean_token_accuracy": 0.8704220056533813, "num_tokens": 358840730.0, "step": 9450 }, { "epoch": 1.202264342958911, "ewc_loss": 0.10009765625, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.712388038635254, "learning_rate": 1e-05, "loss": 0.5002, "mean_token_accuracy": 0.8632411360740662, "num_tokens": 358881457.0, "step": 9451 }, { "epoch": 1.2023915532375016, "ewc_loss": 0.109375, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.265051364898682, "learning_rate": 1e-05, "loss": 0.5103, "mean_token_accuracy": 0.8615608811378479, "num_tokens": 358917034.0, "step": 9452 }, { "epoch": 1.202518763516092, "ewc_loss": 0.0966796875, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.5782647132873535, "learning_rate": 1e-05, "loss": 0.5177, "mean_token_accuracy": 0.8607286214828491, "num_tokens": 358957851.0, "step": 9453 }, { "epoch": 1.2026459737946826, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.41747522354126, "learning_rate": 1e-05, "loss": 0.5194, "mean_token_accuracy": 0.8608464598655701, "num_tokens": 358999816.0, "step": 9454 }, { "epoch": 1.2027731840732732, "ewc_loss": 0.09716796875, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.484255790710449, "learning_rate": 1e-05, "loss": 0.5394, "mean_token_accuracy": 0.8492139577865601, "num_tokens": 359041067.0, "step": 9455 }, { "epoch": 1.2029003943518637, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.721519947052002, "learning_rate": 1e-05, "loss": 0.5029, "mean_token_accuracy": 0.867303192615509, "num_tokens": 359080495.0, "step": 9456 }, { "epoch": 1.2030276046304542, "ewc_loss": 0.09814453125, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.484167575836182, "learning_rate": 1e-05, "loss": 0.4444, "mean_token_accuracy": 0.8805381059646606, "num_tokens": 359118245.0, "step": 9457 }, { "epoch": 1.2031548149090447, "ewc_loss": 0.125, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.944695472717285, "learning_rate": 1e-05, "loss": 0.5503, "mean_token_accuracy": 0.8579182624816895, "num_tokens": 359159840.0, "step": 9458 }, { "epoch": 1.2032820251876353, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.687474250793457, "learning_rate": 1e-05, "loss": 0.4923, "mean_token_accuracy": 0.8646606206893921, "num_tokens": 359198917.0, "step": 9459 }, { "epoch": 1.2034092354662256, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 6.976102352142334, "learning_rate": 1e-05, "loss": 0.5089, "mean_token_accuracy": 0.8653711080551147, "num_tokens": 359231904.0, "step": 9460 }, { "epoch": 1.203536445744816, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 5.983344078063965, "learning_rate": 1e-05, "loss": 0.5339, "mean_token_accuracy": 0.8557296991348267, "num_tokens": 359271713.0, "step": 9461 }, { "epoch": 1.2036636560234066, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.650662422180176, "learning_rate": 1e-05, "loss": 0.5191, "mean_token_accuracy": 0.8663787841796875, "num_tokens": 359312548.0, "step": 9462 }, { "epoch": 1.2037908663019972, "ewc_loss": 0.11376953125, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.150602340698242, "learning_rate": 1e-05, "loss": 0.4679, "mean_token_accuracy": 0.8769487142562866, "num_tokens": 359348171.0, "step": 9463 }, { "epoch": 1.2039180765805877, "ewc_loss": 0.10791015625, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.173867702484131, "learning_rate": 1e-05, "loss": 0.5446, "mean_token_accuracy": 0.858765721321106, "num_tokens": 359383863.0, "step": 9464 }, { "epoch": 1.2040452868591782, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.264307498931885, "learning_rate": 1e-05, "loss": 0.5058, "mean_token_accuracy": 0.8629552125930786, "num_tokens": 359421595.0, "step": 9465 }, { "epoch": 1.2041724971377687, "ewc_loss": 0.09912109375, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.703675746917725, "learning_rate": 1e-05, "loss": 0.4489, "mean_token_accuracy": 0.8799827098846436, "num_tokens": 359457889.0, "step": 9466 }, { "epoch": 1.2042997074163593, "ewc_loss": 0.11474609375, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.483556747436523, "learning_rate": 1e-05, "loss": 0.5198, "mean_token_accuracy": 0.8629710674285889, "num_tokens": 359496938.0, "step": 9467 }, { "epoch": 1.2044269176949498, "ewc_loss": 0.095703125, "ewc_loss_diag": 0.00018310546875, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.410983085632324, "learning_rate": 1e-05, "loss": 0.5117, "mean_token_accuracy": 0.862352192401886, "num_tokens": 359536124.0, "step": 9468 }, { "epoch": 1.2045541279735403, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.00018310546875, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.867042064666748, "learning_rate": 1e-05, "loss": 0.5413, "mean_token_accuracy": 0.8601373434066772, "num_tokens": 359573525.0, "step": 9469 }, { "epoch": 1.2046813382521309, "ewc_loss": 0.095703125, "ewc_loss_diag": 0.00018310546875, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.390586853027344, "learning_rate": 1e-05, "loss": 0.4703, "mean_token_accuracy": 0.8715676665306091, "num_tokens": 359614960.0, "step": 9470 }, { "epoch": 1.2048085485307212, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.00018310546875, "ewc_loss_parallel": 0.00011539459228515625, "grad_norm": 7.25762414932251, "learning_rate": 1e-05, "loss": 0.5421, "mean_token_accuracy": 0.8621169328689575, "num_tokens": 359653367.0, "step": 9471 }, { "epoch": 1.2049357588093117, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.00018310546875, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 5.77789831161499, "learning_rate": 1e-05, "loss": 0.5547, "mean_token_accuracy": 0.8479593992233276, "num_tokens": 359692983.0, "step": 9472 }, { "epoch": 1.2050629690879022, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.00018310546875, "ewc_loss_parallel": 0.00011396408081054688, "grad_norm": 7.2898149490356445, "learning_rate": 1e-05, "loss": 0.5555, "mean_token_accuracy": 0.8579214811325073, "num_tokens": 359729487.0, "step": 9473 }, { "epoch": 1.2051901793664928, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.00018310546875, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.1885528564453125, "learning_rate": 1e-05, "loss": 0.47, "mean_token_accuracy": 0.8761944770812988, "num_tokens": 359764805.0, "step": 9474 }, { "epoch": 1.2053173896450833, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.00018310546875, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.936321258544922, "learning_rate": 1e-05, "loss": 0.5413, "mean_token_accuracy": 0.8575406074523926, "num_tokens": 359803014.0, "step": 9475 }, { "epoch": 1.2054445999236738, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.00018310546875, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.263515472412109, "learning_rate": 1e-05, "loss": 0.4951, "mean_token_accuracy": 0.8689292669296265, "num_tokens": 359837866.0, "step": 9476 }, { "epoch": 1.2055718102022643, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.00018310546875, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.56013298034668, "learning_rate": 1e-05, "loss": 0.4687, "mean_token_accuracy": 0.8819620609283447, "num_tokens": 359874052.0, "step": 9477 }, { "epoch": 1.2056990204808549, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.00018310546875, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.149017810821533, "learning_rate": 1e-05, "loss": 0.4439, "mean_token_accuracy": 0.8860639333724976, "num_tokens": 359915746.0, "step": 9478 }, { "epoch": 1.2058262307594454, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.00018310546875, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.296915054321289, "learning_rate": 1e-05, "loss": 0.4523, "mean_token_accuracy": 0.8813241720199585, "num_tokens": 359945857.0, "step": 9479 }, { "epoch": 1.205953441038036, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.00018310546875, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.011734962463379, "learning_rate": 1e-05, "loss": 0.4775, "mean_token_accuracy": 0.873820424079895, "num_tokens": 359988455.0, "step": 9480 }, { "epoch": 1.2060806513166265, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.00018310546875, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 6.162562370300293, "learning_rate": 1e-05, "loss": 0.4702, "mean_token_accuracy": 0.8743109107017517, "num_tokens": 360030610.0, "step": 9481 }, { "epoch": 1.206207861595217, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.00018310546875, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.862771034240723, "learning_rate": 1e-05, "loss": 0.5041, "mean_token_accuracy": 0.8607242703437805, "num_tokens": 360068919.0, "step": 9482 }, { "epoch": 1.2063350718738075, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.00018310546875, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.170583248138428, "learning_rate": 1e-05, "loss": 0.4849, "mean_token_accuracy": 0.8672529458999634, "num_tokens": 360099073.0, "step": 9483 }, { "epoch": 1.206462282152398, "ewc_loss": 0.099609375, "ewc_loss_diag": 0.00018310546875, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.689676761627197, "learning_rate": 1e-05, "loss": 0.4951, "mean_token_accuracy": 0.8688446879386902, "num_tokens": 360141947.0, "step": 9484 }, { "epoch": 1.2065894924309883, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.00018310546875, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.288041114807129, "learning_rate": 1e-05, "loss": 0.52, "mean_token_accuracy": 0.8654272556304932, "num_tokens": 360179435.0, "step": 9485 }, { "epoch": 1.2067167027095789, "ewc_loss": 0.095703125, "ewc_loss_diag": 0.00018310546875, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.51497745513916, "learning_rate": 1e-05, "loss": 0.4602, "mean_token_accuracy": 0.8756006956100464, "num_tokens": 360221254.0, "step": 9486 }, { "epoch": 1.2068439129881694, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.00018310546875, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.535658836364746, "learning_rate": 1e-05, "loss": 0.4858, "mean_token_accuracy": 0.8759598135948181, "num_tokens": 360261657.0, "step": 9487 }, { "epoch": 1.20697112326676, "ewc_loss": 0.095703125, "ewc_loss_diag": 0.00018310546875, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.415050029754639, "learning_rate": 1e-05, "loss": 0.469, "mean_token_accuracy": 0.8709861040115356, "num_tokens": 360294408.0, "step": 9488 }, { "epoch": 1.2070983335453505, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.00018310546875, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.879439353942871, "learning_rate": 1e-05, "loss": 0.4896, "mean_token_accuracy": 0.8763527870178223, "num_tokens": 360332275.0, "step": 9489 }, { "epoch": 1.207225543823941, "ewc_loss": 0.0986328125, "ewc_loss_diag": 0.00018310546875, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.50053071975708, "learning_rate": 1e-05, "loss": 0.4541, "mean_token_accuracy": 0.8742905855178833, "num_tokens": 360372381.0, "step": 9490 }, { "epoch": 1.2073527541025315, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.00018310546875, "ewc_loss_parallel": 0.00011110305786132812, "grad_norm": 7.176509380340576, "learning_rate": 1e-05, "loss": 0.5288, "mean_token_accuracy": 0.8607608079910278, "num_tokens": 360404066.0, "step": 9491 }, { "epoch": 1.207479964381122, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.00018310546875, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 5.862881183624268, "learning_rate": 1e-05, "loss": 0.4576, "mean_token_accuracy": 0.8770995140075684, "num_tokens": 360434457.0, "step": 9492 }, { "epoch": 1.2076071746597126, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.00018310546875, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 7.086076259613037, "learning_rate": 1e-05, "loss": 0.5158, "mean_token_accuracy": 0.8690961599349976, "num_tokens": 360467905.0, "step": 9493 }, { "epoch": 1.207734384938303, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.00018310546875, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.18054723739624, "learning_rate": 1e-05, "loss": 0.4908, "mean_token_accuracy": 0.8727232217788696, "num_tokens": 360505190.0, "step": 9494 }, { "epoch": 1.2078615952168934, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.00018310546875, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.6537675857543945, "learning_rate": 1e-05, "loss": 0.4746, "mean_token_accuracy": 0.8722594380378723, "num_tokens": 360539191.0, "step": 9495 }, { "epoch": 1.207988805495484, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.00018310546875, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.290986061096191, "learning_rate": 1e-05, "loss": 0.544, "mean_token_accuracy": 0.857217013835907, "num_tokens": 360569508.0, "step": 9496 }, { "epoch": 1.2081160157740745, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.00018310546875, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 6.135546684265137, "learning_rate": 1e-05, "loss": 0.4634, "mean_token_accuracy": 0.8785512447357178, "num_tokens": 360604669.0, "step": 9497 }, { "epoch": 1.208243226052665, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.00018310546875, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.346022129058838, "learning_rate": 1e-05, "loss": 0.421, "mean_token_accuracy": 0.8911207914352417, "num_tokens": 360639009.0, "step": 9498 }, { "epoch": 1.2083704363312555, "ewc_loss": 0.0986328125, "ewc_loss_diag": 0.00018310546875, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.696182727813721, "learning_rate": 1e-05, "loss": 0.4204, "mean_token_accuracy": 0.8841277956962585, "num_tokens": 360673857.0, "step": 9499 }, { "epoch": 1.208497646609846, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.00018310546875, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.540045261383057, "learning_rate": 1e-05, "loss": 0.5293, "mean_token_accuracy": 0.8609649538993835, "num_tokens": 360717553.0, "step": 9500 }, { "epoch": 1.2086248568884366, "ewc_loss": 0.095703125, "ewc_loss_diag": 0.00018310546875, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.457423686981201, "learning_rate": 1e-05, "loss": 0.5043, "mean_token_accuracy": 0.8713769912719727, "num_tokens": 360761013.0, "step": 9501 }, { "epoch": 1.208752067167027, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.00018310546875, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.843647480010986, "learning_rate": 1e-05, "loss": 0.5059, "mean_token_accuracy": 0.8675541281700134, "num_tokens": 360804302.0, "step": 9502 }, { "epoch": 1.2088792774456176, "ewc_loss": 0.099609375, "ewc_loss_diag": 0.00018310546875, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.480599880218506, "learning_rate": 1e-05, "loss": 0.4001, "mean_token_accuracy": 0.8907380700111389, "num_tokens": 360837045.0, "step": 9503 }, { "epoch": 1.2090064877242082, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.00018310546875, "ewc_loss_parallel": 0.00011110305786132812, "grad_norm": 7.137062072753906, "learning_rate": 1e-05, "loss": 0.5052, "mean_token_accuracy": 0.8698992133140564, "num_tokens": 360876243.0, "step": 9504 }, { "epoch": 1.2091336980027987, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.00018310546875, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 5.779107570648193, "learning_rate": 1e-05, "loss": 0.5012, "mean_token_accuracy": 0.8644452691078186, "num_tokens": 360915112.0, "step": 9505 }, { "epoch": 1.2092609082813892, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.00018310546875, "ewc_loss_parallel": 0.00011014938354492188, "grad_norm": 7.152871608734131, "learning_rate": 1e-05, "loss": 0.5007, "mean_token_accuracy": 0.8709656000137329, "num_tokens": 360951856.0, "step": 9506 }, { "epoch": 1.2093881185599797, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.00018310546875, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.073831081390381, "learning_rate": 1e-05, "loss": 0.4739, "mean_token_accuracy": 0.877026379108429, "num_tokens": 360987243.0, "step": 9507 }, { "epoch": 1.2095153288385703, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.00018310546875, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.895650863647461, "learning_rate": 1e-05, "loss": 0.4949, "mean_token_accuracy": 0.8729599118232727, "num_tokens": 361025099.0, "step": 9508 }, { "epoch": 1.2096425391171606, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.00018310546875, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.17716646194458, "learning_rate": 1e-05, "loss": 0.5423, "mean_token_accuracy": 0.8519235849380493, "num_tokens": 361055232.0, "step": 9509 }, { "epoch": 1.209769749395751, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.00018310546875, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.5600152015686035, "learning_rate": 1e-05, "loss": 0.5081, "mean_token_accuracy": 0.8682291507720947, "num_tokens": 361092500.0, "step": 9510 }, { "epoch": 1.2098969596743416, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.00018310546875, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.103823661804199, "learning_rate": 1e-05, "loss": 0.4713, "mean_token_accuracy": 0.8748456835746765, "num_tokens": 361127044.0, "step": 9511 }, { "epoch": 1.2100241699529322, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.00018310546875, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.300843715667725, "learning_rate": 1e-05, "loss": 0.4531, "mean_token_accuracy": 0.8891937136650085, "num_tokens": 361163065.0, "step": 9512 }, { "epoch": 1.2101513802315227, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.00018310546875, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.985162258148193, "learning_rate": 1e-05, "loss": 0.5071, "mean_token_accuracy": 0.8616925477981567, "num_tokens": 361204465.0, "step": 9513 }, { "epoch": 1.2102785905101132, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.00018310546875, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.182236194610596, "learning_rate": 1e-05, "loss": 0.5519, "mean_token_accuracy": 0.8483190536499023, "num_tokens": 361243788.0, "step": 9514 }, { "epoch": 1.2104058007887037, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.00018310546875, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.846370220184326, "learning_rate": 1e-05, "loss": 0.4743, "mean_token_accuracy": 0.8738188743591309, "num_tokens": 361281427.0, "step": 9515 }, { "epoch": 1.2105330110672943, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.00018310546875, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 6.186511516571045, "learning_rate": 1e-05, "loss": 0.5406, "mean_token_accuracy": 0.8540104031562805, "num_tokens": 361324695.0, "step": 9516 }, { "epoch": 1.2106602213458848, "ewc_loss": 0.0986328125, "ewc_loss_diag": 0.00018310546875, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.679196357727051, "learning_rate": 1e-05, "loss": 0.4903, "mean_token_accuracy": 0.8677765727043152, "num_tokens": 361370446.0, "step": 9517 }, { "epoch": 1.2107874316244753, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.28997278213501, "learning_rate": 1e-05, "loss": 0.496, "mean_token_accuracy": 0.867725670337677, "num_tokens": 361414622.0, "step": 9518 }, { "epoch": 1.2109146419030659, "ewc_loss": 0.095703125, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.5363030433654785, "learning_rate": 1e-05, "loss": 0.447, "mean_token_accuracy": 0.8800297975540161, "num_tokens": 361452363.0, "step": 9519 }, { "epoch": 1.2110418521816562, "ewc_loss": 0.11474609375, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.528268337249756, "learning_rate": 1e-05, "loss": 0.5562, "mean_token_accuracy": 0.8527486324310303, "num_tokens": 361484195.0, "step": 9520 }, { "epoch": 1.2111690624602467, "ewc_loss": 0.09619140625, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.440703868865967, "learning_rate": 1e-05, "loss": 0.4743, "mean_token_accuracy": 0.8672668933868408, "num_tokens": 361523604.0, "step": 9521 }, { "epoch": 1.2112962727388372, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.821280002593994, "learning_rate": 1e-05, "loss": 0.5538, "mean_token_accuracy": 0.8527489900588989, "num_tokens": 361563723.0, "step": 9522 }, { "epoch": 1.2114234830174277, "ewc_loss": 0.09814453125, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.5101213455200195, "learning_rate": 1e-05, "loss": 0.4639, "mean_token_accuracy": 0.871169924736023, "num_tokens": 361602578.0, "step": 9523 }, { "epoch": 1.2115506932960183, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 0.00010967254638671875, "grad_norm": 7.074295520782471, "learning_rate": 1e-05, "loss": 0.5244, "mean_token_accuracy": 0.8678814172744751, "num_tokens": 361644255.0, "step": 9524 }, { "epoch": 1.2116779035746088, "ewc_loss": 0.10595703125, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 5.81341552734375, "learning_rate": 1e-05, "loss": 0.4743, "mean_token_accuracy": 0.8698233366012573, "num_tokens": 361683424.0, "step": 9525 }, { "epoch": 1.2118051138531993, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 6.990942478179932, "learning_rate": 1e-05, "loss": 0.4625, "mean_token_accuracy": 0.8807751536369324, "num_tokens": 361720965.0, "step": 9526 }, { "epoch": 1.2119323241317899, "ewc_loss": 0.11376953125, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.123097896575928, "learning_rate": 1e-05, "loss": 0.5453, "mean_token_accuracy": 0.8559837341308594, "num_tokens": 361753900.0, "step": 9527 }, { "epoch": 1.2120595344103804, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.568143367767334, "learning_rate": 1e-05, "loss": 0.4993, "mean_token_accuracy": 0.8680455684661865, "num_tokens": 361793072.0, "step": 9528 }, { "epoch": 1.212186744688971, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.232024192810059, "learning_rate": 1e-05, "loss": 0.5502, "mean_token_accuracy": 0.8526200652122498, "num_tokens": 361831082.0, "step": 9529 }, { "epoch": 1.2123139549675614, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 6.067969799041748, "learning_rate": 1e-05, "loss": 0.4616, "mean_token_accuracy": 0.8719809651374817, "num_tokens": 361864834.0, "step": 9530 }, { "epoch": 1.212441165246152, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.328183650970459, "learning_rate": 1e-05, "loss": 0.4896, "mean_token_accuracy": 0.8701662421226501, "num_tokens": 361905567.0, "step": 9531 }, { "epoch": 1.2125683755247425, "ewc_loss": 0.09716796875, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.642772197723389, "learning_rate": 1e-05, "loss": 0.4626, "mean_token_accuracy": 0.8785218000411987, "num_tokens": 361945718.0, "step": 9532 }, { "epoch": 1.212695585803333, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.582131385803223, "learning_rate": 1e-05, "loss": 0.5071, "mean_token_accuracy": 0.8627294898033142, "num_tokens": 361985317.0, "step": 9533 }, { "epoch": 1.2128227960819233, "ewc_loss": 0.095703125, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.407098293304443, "learning_rate": 1e-05, "loss": 0.4805, "mean_token_accuracy": 0.8705475330352783, "num_tokens": 362020725.0, "step": 9534 }, { "epoch": 1.2129500063605139, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.959089756011963, "learning_rate": 1e-05, "loss": 0.5338, "mean_token_accuracy": 0.8577209115028381, "num_tokens": 362054398.0, "step": 9535 }, { "epoch": 1.2130772166391044, "ewc_loss": 0.09814453125, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.491869926452637, "learning_rate": 1e-05, "loss": 0.5199, "mean_token_accuracy": 0.8536456823348999, "num_tokens": 362092075.0, "step": 9536 }, { "epoch": 1.213204426917695, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 0.00011444091796875, "grad_norm": 7.253553867340088, "learning_rate": 1e-05, "loss": 0.4579, "mean_token_accuracy": 0.8899824619293213, "num_tokens": 362128286.0, "step": 9537 }, { "epoch": 1.2133316371962855, "ewc_loss": 0.10986328125, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 5.880669593811035, "learning_rate": 1e-05, "loss": 0.5292, "mean_token_accuracy": 0.8607040643692017, "num_tokens": 362169965.0, "step": 9538 }, { "epoch": 1.213458847474876, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 0.00011014938354492188, "grad_norm": 7.175660133361816, "learning_rate": 1e-05, "loss": 0.514, "mean_token_accuracy": 0.8678780198097229, "num_tokens": 362211638.0, "step": 9539 }, { "epoch": 1.2135860577534665, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.191856861114502, "learning_rate": 1e-05, "loss": 0.4431, "mean_token_accuracy": 0.8862833380699158, "num_tokens": 362241597.0, "step": 9540 }, { "epoch": 1.213713268032057, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.827433109283447, "learning_rate": 1e-05, "loss": 0.5145, "mean_token_accuracy": 0.8668699860572815, "num_tokens": 362280028.0, "step": 9541 }, { "epoch": 1.2138404783106476, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.211442470550537, "learning_rate": 1e-05, "loss": 0.5027, "mean_token_accuracy": 0.8677983283996582, "num_tokens": 362323611.0, "step": 9542 }, { "epoch": 1.213967688589238, "ewc_loss": 0.11376953125, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.483741760253906, "learning_rate": 1e-05, "loss": 0.4902, "mean_token_accuracy": 0.8688209056854248, "num_tokens": 362359097.0, "step": 9543 }, { "epoch": 1.2140948988678284, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.106339454650879, "learning_rate": 1e-05, "loss": 0.4952, "mean_token_accuracy": 0.8671913146972656, "num_tokens": 362402175.0, "step": 9544 }, { "epoch": 1.214222109146419, "ewc_loss": 0.109375, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.255682468414307, "learning_rate": 1e-05, "loss": 0.4992, "mean_token_accuracy": 0.8700024485588074, "num_tokens": 362442581.0, "step": 9545 }, { "epoch": 1.2143493194250095, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.963624000549316, "learning_rate": 1e-05, "loss": 0.4541, "mean_token_accuracy": 0.877687394618988, "num_tokens": 362480691.0, "step": 9546 }, { "epoch": 1.2144765297036, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.176982402801514, "learning_rate": 1e-05, "loss": 0.4852, "mean_token_accuracy": 0.8698074817657471, "num_tokens": 362522864.0, "step": 9547 }, { "epoch": 1.2146037399821905, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.792880535125732, "learning_rate": 1e-05, "loss": 0.4749, "mean_token_accuracy": 0.8717666864395142, "num_tokens": 362564431.0, "step": 9548 }, { "epoch": 1.214730950260781, "ewc_loss": 0.109375, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.2233452796936035, "learning_rate": 1e-05, "loss": 0.454, "mean_token_accuracy": 0.8815934062004089, "num_tokens": 362599670.0, "step": 9549 }, { "epoch": 1.2148581605393716, "ewc_loss": 0.0986328125, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.632755279541016, "learning_rate": 1e-05, "loss": 0.4986, "mean_token_accuracy": 0.8640561699867249, "num_tokens": 362637588.0, "step": 9550 }, { "epoch": 1.214985370817962, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.394476890563965, "learning_rate": 1e-05, "loss": 0.4875, "mean_token_accuracy": 0.8700207471847534, "num_tokens": 362674679.0, "step": 9551 }, { "epoch": 1.2151125810965526, "ewc_loss": 0.09619140625, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.4928154945373535, "learning_rate": 1e-05, "loss": 0.4748, "mean_token_accuracy": 0.8640826940536499, "num_tokens": 362705840.0, "step": 9552 }, { "epoch": 1.2152397913751432, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.672874450683594, "learning_rate": 1e-05, "loss": 0.505, "mean_token_accuracy": 0.8668205142021179, "num_tokens": 362739880.0, "step": 9553 }, { "epoch": 1.2153670016537337, "ewc_loss": 0.0966796875, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.438992977142334, "learning_rate": 1e-05, "loss": 0.4744, "mean_token_accuracy": 0.8763947486877441, "num_tokens": 362779101.0, "step": 9554 }, { "epoch": 1.2154942119323242, "ewc_loss": 0.125, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.988834857940674, "learning_rate": 1e-05, "loss": 0.4864, "mean_token_accuracy": 0.8724666833877563, "num_tokens": 362821034.0, "step": 9555 }, { "epoch": 1.2156214222109147, "ewc_loss": 0.10107421875, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.635929107666016, "learning_rate": 1e-05, "loss": 0.454, "mean_token_accuracy": 0.8789857029914856, "num_tokens": 362856895.0, "step": 9556 }, { "epoch": 1.2157486324895053, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 7.111559867858887, "learning_rate": 1e-05, "loss": 0.5167, "mean_token_accuracy": 0.8671281337738037, "num_tokens": 362894955.0, "step": 9557 }, { "epoch": 1.2158758427680956, "ewc_loss": 0.11083984375, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 5.987880706787109, "learning_rate": 1e-05, "loss": 0.5122, "mean_token_accuracy": 0.863679051399231, "num_tokens": 362930397.0, "step": 9558 }, { "epoch": 1.216003053046686, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.848444938659668, "learning_rate": 1e-05, "loss": 0.4548, "mean_token_accuracy": 0.8855974674224854, "num_tokens": 362970728.0, "step": 9559 }, { "epoch": 1.2161302633252766, "ewc_loss": 0.11376953125, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.194357395172119, "learning_rate": 1e-05, "loss": 0.4536, "mean_token_accuracy": 0.8827928304672241, "num_tokens": 363007803.0, "step": 9560 }, { "epoch": 1.2162574736038672, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.369344711303711, "learning_rate": 1e-05, "loss": 0.502, "mean_token_accuracy": 0.8668603301048279, "num_tokens": 363047815.0, "step": 9561 }, { "epoch": 1.2163846838824577, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.264682292938232, "learning_rate": 1e-05, "loss": 0.4681, "mean_token_accuracy": 0.8779219388961792, "num_tokens": 363084156.0, "step": 9562 }, { "epoch": 1.2165118941610482, "ewc_loss": 0.10107421875, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.885470390319824, "learning_rate": 1e-05, "loss": 0.4916, "mean_token_accuracy": 0.8668012619018555, "num_tokens": 363125390.0, "step": 9563 }, { "epoch": 1.2166391044396387, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.400707244873047, "learning_rate": 1e-05, "loss": 0.4921, "mean_token_accuracy": 0.8718286752700806, "num_tokens": 363164210.0, "step": 9564 }, { "epoch": 1.2167663147182293, "ewc_loss": 0.09716796875, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.540006637573242, "learning_rate": 1e-05, "loss": 0.5253, "mean_token_accuracy": 0.8511415123939514, "num_tokens": 363200356.0, "step": 9565 }, { "epoch": 1.2168935249968198, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.679778575897217, "learning_rate": 1e-05, "loss": 0.5558, "mean_token_accuracy": 0.8502753973007202, "num_tokens": 363239622.0, "step": 9566 }, { "epoch": 1.2170207352754103, "ewc_loss": 0.0966796875, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.394684791564941, "learning_rate": 1e-05, "loss": 0.4836, "mean_token_accuracy": 0.8722549676895142, "num_tokens": 363277348.0, "step": 9567 }, { "epoch": 1.2171479455540009, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 7.068500518798828, "learning_rate": 1e-05, "loss": 0.5062, "mean_token_accuracy": 0.8687410354614258, "num_tokens": 363310077.0, "step": 9568 }, { "epoch": 1.2172751558325912, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.585708141326904, "learning_rate": 1e-05, "loss": 0.5893, "mean_token_accuracy": 0.8341890573501587, "num_tokens": 363347919.0, "step": 9569 }, { "epoch": 1.2174023661111817, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 0.00011587142944335938, "grad_norm": 7.274212837219238, "learning_rate": 1e-05, "loss": 0.5035, "mean_token_accuracy": 0.871192455291748, "num_tokens": 363390818.0, "step": 9570 }, { "epoch": 1.2175295763897722, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.006216526031494, "learning_rate": 1e-05, "loss": 0.5843, "mean_token_accuracy": 0.8441304564476013, "num_tokens": 363424028.0, "step": 9571 }, { "epoch": 1.2176567866683627, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 7.072279930114746, "learning_rate": 1e-05, "loss": 0.5067, "mean_token_accuracy": 0.8664820194244385, "num_tokens": 363456862.0, "step": 9572 }, { "epoch": 1.2177839969469533, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.207980155944824, "learning_rate": 1e-05, "loss": 0.4921, "mean_token_accuracy": 0.8712545037269592, "num_tokens": 363498538.0, "step": 9573 }, { "epoch": 1.2179112072255438, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.717087745666504, "learning_rate": 1e-05, "loss": 0.4877, "mean_token_accuracy": 0.8719220161437988, "num_tokens": 363533050.0, "step": 9574 }, { "epoch": 1.2180384175041343, "ewc_loss": 0.11376953125, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.1839599609375, "learning_rate": 1e-05, "loss": 0.5099, "mean_token_accuracy": 0.867815375328064, "num_tokens": 363574151.0, "step": 9575 }, { "epoch": 1.2181656277827249, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.393391132354736, "learning_rate": 1e-05, "loss": 0.501, "mean_token_accuracy": 0.8636000156402588, "num_tokens": 363615829.0, "step": 9576 }, { "epoch": 1.2182928380613154, "ewc_loss": 0.109375, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.090163707733154, "learning_rate": 1e-05, "loss": 0.4868, "mean_token_accuracy": 0.8674973845481873, "num_tokens": 363649974.0, "step": 9577 }, { "epoch": 1.218420048339906, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.195215225219727, "learning_rate": 1e-05, "loss": 0.4923, "mean_token_accuracy": 0.8707761168479919, "num_tokens": 363691196.0, "step": 9578 }, { "epoch": 1.2185472586184964, "ewc_loss": 0.10400390625, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.941044807434082, "learning_rate": 1e-05, "loss": 0.4473, "mean_token_accuracy": 0.8806506395339966, "num_tokens": 363726496.0, "step": 9579 }, { "epoch": 1.218674468897087, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.141887664794922, "learning_rate": 1e-05, "loss": 0.5292, "mean_token_accuracy": 0.8588612675666809, "num_tokens": 363769119.0, "step": 9580 }, { "epoch": 1.2188016791756775, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.799831390380859, "learning_rate": 1e-05, "loss": 0.5465, "mean_token_accuracy": 0.8535164594650269, "num_tokens": 363803363.0, "step": 9581 }, { "epoch": 1.218928889454268, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.179409027099609, "learning_rate": 1e-05, "loss": 0.4717, "mean_token_accuracy": 0.8731019496917725, "num_tokens": 363842602.0, "step": 9582 }, { "epoch": 1.2190560997328583, "ewc_loss": 0.0986328125, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.6455464363098145, "learning_rate": 1e-05, "loss": 0.4912, "mean_token_accuracy": 0.8666850328445435, "num_tokens": 363882758.0, "step": 9583 }, { "epoch": 1.2191833100114489, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.316244602203369, "learning_rate": 1e-05, "loss": 0.4488, "mean_token_accuracy": 0.8831263780593872, "num_tokens": 363923738.0, "step": 9584 }, { "epoch": 1.2193105202900394, "ewc_loss": 0.0966796875, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.493311405181885, "learning_rate": 1e-05, "loss": 0.4492, "mean_token_accuracy": 0.8771816492080688, "num_tokens": 363957999.0, "step": 9585 }, { "epoch": 1.21943773056863, "ewc_loss": 0.11474609375, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.57589054107666, "learning_rate": 1e-05, "loss": 0.4875, "mean_token_accuracy": 0.8711692094802856, "num_tokens": 363992653.0, "step": 9586 }, { "epoch": 1.2195649408472204, "ewc_loss": 0.0966796875, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.405916213989258, "learning_rate": 1e-05, "loss": 0.4301, "mean_token_accuracy": 0.8840595483779907, "num_tokens": 364027376.0, "step": 9587 }, { "epoch": 1.219692151125811, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.93751335144043, "learning_rate": 1e-05, "loss": 0.5005, "mean_token_accuracy": 0.8676358461380005, "num_tokens": 364063299.0, "step": 9588 }, { "epoch": 1.2198193614044015, "ewc_loss": 0.10009765625, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.524397373199463, "learning_rate": 1e-05, "loss": 0.5639, "mean_token_accuracy": 0.8411068916320801, "num_tokens": 364105724.0, "step": 9589 }, { "epoch": 1.219946571682992, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 0.00011348724365234375, "grad_norm": 7.203110218048096, "learning_rate": 1e-05, "loss": 0.5433, "mean_token_accuracy": 0.8581838011741638, "num_tokens": 364140427.0, "step": 9590 }, { "epoch": 1.2200737819615826, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 5.902565956115723, "learning_rate": 1e-05, "loss": 0.517, "mean_token_accuracy": 0.857967734336853, "num_tokens": 364185262.0, "step": 9591 }, { "epoch": 1.220200992240173, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 0.00011014938354492188, "grad_norm": 7.077909469604492, "learning_rate": 1e-05, "loss": 0.5036, "mean_token_accuracy": 0.869484543800354, "num_tokens": 364219006.0, "step": 9592 }, { "epoch": 1.2203282025187634, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.230735778808594, "learning_rate": 1e-05, "loss": 0.4799, "mean_token_accuracy": 0.8739224076271057, "num_tokens": 364259699.0, "step": 9593 }, { "epoch": 1.220455412797354, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.613504409790039, "learning_rate": 1e-05, "loss": 0.4807, "mean_token_accuracy": 0.8745487332344055, "num_tokens": 364297759.0, "step": 9594 }, { "epoch": 1.2205826230759445, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.2943596839904785, "learning_rate": 1e-05, "loss": 0.5189, "mean_token_accuracy": 0.8656684160232544, "num_tokens": 364338683.0, "step": 9595 }, { "epoch": 1.220709833354535, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 6.112103462219238, "learning_rate": 1e-05, "loss": 0.4908, "mean_token_accuracy": 0.8694150447845459, "num_tokens": 364377587.0, "step": 9596 }, { "epoch": 1.2208370436331255, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.348058223724365, "learning_rate": 1e-05, "loss": 0.5172, "mean_token_accuracy": 0.862450122833252, "num_tokens": 364416740.0, "step": 9597 }, { "epoch": 1.220964253911716, "ewc_loss": 0.099609375, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.719496250152588, "learning_rate": 1e-05, "loss": 0.5108, "mean_token_accuracy": 0.8588206768035889, "num_tokens": 364458695.0, "step": 9598 }, { "epoch": 1.2210914641903066, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.515104293823242, "learning_rate": 1e-05, "loss": 0.5199, "mean_token_accuracy": 0.860590934753418, "num_tokens": 364501395.0, "step": 9599 }, { "epoch": 1.221218674468897, "ewc_loss": 0.0966796875, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.488668918609619, "learning_rate": 1e-05, "loss": 0.4597, "mean_token_accuracy": 0.8756029009819031, "num_tokens": 364534126.0, "step": 9600 }, { "epoch": 1.2213458847474876, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.807656764984131, "learning_rate": 1e-05, "loss": 0.4606, "mean_token_accuracy": 0.8820242881774902, "num_tokens": 364573776.0, "step": 9601 }, { "epoch": 1.2214730950260781, "ewc_loss": 0.0986328125, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.484860420227051, "learning_rate": 1e-05, "loss": 0.4894, "mean_token_accuracy": 0.8647992610931396, "num_tokens": 364617900.0, "step": 9602 }, { "epoch": 1.2216003053046687, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 0.00010967254638671875, "grad_norm": 7.090860366821289, "learning_rate": 1e-05, "loss": 0.4837, "mean_token_accuracy": 0.876522958278656, "num_tokens": 364658992.0, "step": 9603 }, { "epoch": 1.2217275155832592, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.732769966125488, "learning_rate": 1e-05, "loss": 0.4498, "mean_token_accuracy": 0.8834227323532104, "num_tokens": 364695224.0, "step": 9604 }, { "epoch": 1.2218547258618497, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 0.00011110305786132812, "grad_norm": 7.143673896789551, "learning_rate": 1e-05, "loss": 0.5038, "mean_token_accuracy": 0.8713980913162231, "num_tokens": 364731476.0, "step": 9605 }, { "epoch": 1.2219819361404403, "ewc_loss": 0.11181640625, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.039603233337402, "learning_rate": 1e-05, "loss": 0.4867, "mean_token_accuracy": 0.8690431118011475, "num_tokens": 364774129.0, "step": 9606 }, { "epoch": 1.2221091464190306, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.906836986541748, "learning_rate": 1e-05, "loss": 0.508, "mean_token_accuracy": 0.8702043294906616, "num_tokens": 364808097.0, "step": 9607 }, { "epoch": 1.222236356697621, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.137393474578857, "learning_rate": 1e-05, "loss": 0.4678, "mean_token_accuracy": 0.8765654563903809, "num_tokens": 364847592.0, "step": 9608 }, { "epoch": 1.2223635669762116, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.5877156257629395, "learning_rate": 1e-05, "loss": 0.4888, "mean_token_accuracy": 0.8728421926498413, "num_tokens": 364885846.0, "step": 9609 }, { "epoch": 1.2224907772548022, "ewc_loss": 0.11083984375, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.110494136810303, "learning_rate": 1e-05, "loss": 0.5499, "mean_token_accuracy": 0.8547627925872803, "num_tokens": 364927536.0, "step": 9610 }, { "epoch": 1.2226179875333927, "ewc_loss": 0.11083984375, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.317600250244141, "learning_rate": 1e-05, "loss": 0.4877, "mean_token_accuracy": 0.8676964640617371, "num_tokens": 364961050.0, "step": 9611 }, { "epoch": 1.2227451978119832, "ewc_loss": 0.10498046875, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 6.004215240478516, "learning_rate": 1e-05, "loss": 0.5012, "mean_token_accuracy": 0.8638724684715271, "num_tokens": 365001058.0, "step": 9612 }, { "epoch": 1.2228724080905737, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.1700944900512695, "learning_rate": 1e-05, "loss": 0.4684, "mean_token_accuracy": 0.8766974210739136, "num_tokens": 365041345.0, "step": 9613 }, { "epoch": 1.2229996183691643, "ewc_loss": 0.10400390625, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.872375965118408, "learning_rate": 1e-05, "loss": 0.4924, "mean_token_accuracy": 0.866500735282898, "num_tokens": 365085669.0, "step": 9614 }, { "epoch": 1.2231268286477548, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.139660358428955, "learning_rate": 1e-05, "loss": 0.446, "mean_token_accuracy": 0.88405442237854, "num_tokens": 365125630.0, "step": 9615 }, { "epoch": 1.2232540389263453, "ewc_loss": 0.10107421875, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.742923736572266, "learning_rate": 1e-05, "loss": 0.4688, "mean_token_accuracy": 0.8731341361999512, "num_tokens": 365162904.0, "step": 9616 }, { "epoch": 1.2233812492049359, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.218029975891113, "learning_rate": 1e-05, "loss": 0.484, "mean_token_accuracy": 0.8705690503120422, "num_tokens": 365196509.0, "step": 9617 }, { "epoch": 1.2235084594835262, "ewc_loss": 0.09814453125, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.597018718719482, "learning_rate": 1e-05, "loss": 0.5268, "mean_token_accuracy": 0.8554173707962036, "num_tokens": 365232947.0, "step": 9618 }, { "epoch": 1.2236356697621167, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.391139984130859, "learning_rate": 1e-05, "loss": 0.4963, "mean_token_accuracy": 0.8709034323692322, "num_tokens": 365273434.0, "step": 9619 }, { "epoch": 1.2237628800407072, "ewc_loss": 0.09619140625, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.453830242156982, "learning_rate": 1e-05, "loss": 0.5198, "mean_token_accuracy": 0.8630993962287903, "num_tokens": 365316638.0, "step": 9620 }, { "epoch": 1.2238900903192977, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.687085151672363, "learning_rate": 1e-05, "loss": 0.5207, "mean_token_accuracy": 0.8642099499702454, "num_tokens": 365348519.0, "step": 9621 }, { "epoch": 1.2240173005978883, "ewc_loss": 0.09716796875, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.416365146636963, "learning_rate": 1e-05, "loss": 0.4419, "mean_token_accuracy": 0.8823103904724121, "num_tokens": 365383782.0, "step": 9622 }, { "epoch": 1.2241445108764788, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 7.030076026916504, "learning_rate": 1e-05, "loss": 0.5412, "mean_token_accuracy": 0.8578449487686157, "num_tokens": 365421363.0, "step": 9623 }, { "epoch": 1.2242717211550693, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.619044303894043, "learning_rate": 1e-05, "loss": 0.4818, "mean_token_accuracy": 0.868532657623291, "num_tokens": 365468207.0, "step": 9624 }, { "epoch": 1.2243989314336599, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 0.00011444091796875, "grad_norm": 7.1782402992248535, "learning_rate": 1e-05, "loss": 0.4791, "mean_token_accuracy": 0.8756327629089355, "num_tokens": 365498318.0, "step": 9625 }, { "epoch": 1.2245261417122504, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.016026496887207, "learning_rate": 1e-05, "loss": 0.486, "mean_token_accuracy": 0.8707465529441833, "num_tokens": 365536140.0, "step": 9626 }, { "epoch": 1.224653351990841, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.90819787979126, "learning_rate": 1e-05, "loss": 0.5482, "mean_token_accuracy": 0.8542048931121826, "num_tokens": 365576490.0, "step": 9627 }, { "epoch": 1.2247805622694314, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0001850128173828125, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.227865219116211, "learning_rate": 1e-05, "loss": 0.5204, "mean_token_accuracy": 0.8648993372917175, "num_tokens": 365618043.0, "step": 9628 }, { "epoch": 1.224907772548022, "ewc_loss": 0.11376953125, "ewc_loss_diag": 0.0001850128173828125, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.440001964569092, "learning_rate": 1e-05, "loss": 0.5218, "mean_token_accuracy": 0.8615490794181824, "num_tokens": 365648303.0, "step": 9629 }, { "epoch": 1.2250349828266125, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.281408309936523, "learning_rate": 1e-05, "loss": 0.507, "mean_token_accuracy": 0.868184506893158, "num_tokens": 365684739.0, "step": 9630 }, { "epoch": 1.225162193105203, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.945950031280518, "learning_rate": 1e-05, "loss": 0.5289, "mean_token_accuracy": 0.8543248176574707, "num_tokens": 365726542.0, "step": 9631 }, { "epoch": 1.2252894033837933, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.41085958480835, "learning_rate": 1e-05, "loss": 0.4626, "mean_token_accuracy": 0.8790541291236877, "num_tokens": 365769949.0, "step": 9632 }, { "epoch": 1.2254166136623839, "ewc_loss": 0.09765625, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.593016624450684, "learning_rate": 1e-05, "loss": 0.4753, "mean_token_accuracy": 0.8717513084411621, "num_tokens": 365805891.0, "step": 9633 }, { "epoch": 1.2255438239409744, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.65810489654541, "learning_rate": 1e-05, "loss": 0.569, "mean_token_accuracy": 0.849765419960022, "num_tokens": 365847011.0, "step": 9634 }, { "epoch": 1.225671034219565, "ewc_loss": 0.09716796875, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.450453281402588, "learning_rate": 1e-05, "loss": 0.4676, "mean_token_accuracy": 0.8710727691650391, "num_tokens": 365884546.0, "step": 9635 }, { "epoch": 1.2257982444981554, "ewc_loss": 0.125, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.97916316986084, "learning_rate": 1e-05, "loss": 0.4765, "mean_token_accuracy": 0.8783464431762695, "num_tokens": 365919619.0, "step": 9636 }, { "epoch": 1.225925454776746, "ewc_loss": 0.10205078125, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.581578731536865, "learning_rate": 1e-05, "loss": 0.4796, "mean_token_accuracy": 0.8756996393203735, "num_tokens": 365956020.0, "step": 9637 }, { "epoch": 1.2260526650553365, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 0.00011348724365234375, "grad_norm": 7.192977428436279, "learning_rate": 1e-05, "loss": 0.5761, "mean_token_accuracy": 0.8465104103088379, "num_tokens": 365994945.0, "step": 9638 }, { "epoch": 1.226179875333927, "ewc_loss": 0.109375, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 5.916142463684082, "learning_rate": 1e-05, "loss": 0.5182, "mean_token_accuracy": 0.8603894710540771, "num_tokens": 366036476.0, "step": 9639 }, { "epoch": 1.2263070856125176, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 0.00010967254638671875, "grad_norm": 7.058197975158691, "learning_rate": 1e-05, "loss": 0.492, "mean_token_accuracy": 0.8751530647277832, "num_tokens": 366072966.0, "step": 9640 }, { "epoch": 1.226434295891108, "ewc_loss": 0.11376953125, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.13526725769043, "learning_rate": 1e-05, "loss": 0.4858, "mean_token_accuracy": 0.8704841136932373, "num_tokens": 366111535.0, "step": 9641 }, { "epoch": 1.2265615061696984, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.751200199127197, "learning_rate": 1e-05, "loss": 0.5261, "mean_token_accuracy": 0.8636717796325684, "num_tokens": 366151104.0, "step": 9642 }, { "epoch": 1.226688716448289, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.148595809936523, "learning_rate": 1e-05, "loss": 0.5347, "mean_token_accuracy": 0.8602808713912964, "num_tokens": 366184062.0, "step": 9643 }, { "epoch": 1.2268159267268794, "ewc_loss": 0.11376953125, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.449091911315918, "learning_rate": 1e-05, "loss": 0.4797, "mean_token_accuracy": 0.8724101781845093, "num_tokens": 366217774.0, "step": 9644 }, { "epoch": 1.22694313700547, "ewc_loss": 0.10986328125, "ewc_loss_diag": 0.0001850128173828125, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.057215213775635, "learning_rate": 1e-05, "loss": 0.4666, "mean_token_accuracy": 0.8748869895935059, "num_tokens": 366252776.0, "step": 9645 }, { "epoch": 1.2270703472840605, "ewc_loss": 0.109375, "ewc_loss_diag": 0.0001850128173828125, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.253772258758545, "learning_rate": 1e-05, "loss": 0.499, "mean_token_accuracy": 0.8656981587409973, "num_tokens": 366289031.0, "step": 9646 }, { "epoch": 1.227197557562651, "ewc_loss": 0.10400390625, "ewc_loss_diag": 0.0001850128173828125, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.930098056793213, "learning_rate": 1e-05, "loss": 0.4544, "mean_token_accuracy": 0.8802114725112915, "num_tokens": 366327919.0, "step": 9647 }, { "epoch": 1.2273247678412416, "ewc_loss": 0.10888671875, "ewc_loss_diag": 0.00018596649169921875, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.19875431060791, "learning_rate": 1e-05, "loss": 0.5236, "mean_token_accuracy": 0.8549266457557678, "num_tokens": 366368655.0, "step": 9648 }, { "epoch": 1.227451978119832, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.00018596649169921875, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.780818939208984, "learning_rate": 1e-05, "loss": 0.4876, "mean_token_accuracy": 0.867885947227478, "num_tokens": 366406987.0, "step": 9649 }, { "epoch": 1.2275791883984226, "ewc_loss": 0.10888671875, "ewc_loss_diag": 0.00018596649169921875, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.250567436218262, "learning_rate": 1e-05, "loss": 0.5342, "mean_token_accuracy": 0.8548084497451782, "num_tokens": 366447360.0, "step": 9650 }, { "epoch": 1.2277063986770131, "ewc_loss": 0.0986328125, "ewc_loss_diag": 0.00018596649169921875, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.627720832824707, "learning_rate": 1e-05, "loss": 0.4557, "mean_token_accuracy": 0.8759331703186035, "num_tokens": 366487640.0, "step": 9651 }, { "epoch": 1.2278336089556037, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.00018596649169921875, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.414453029632568, "learning_rate": 1e-05, "loss": 0.5369, "mean_token_accuracy": 0.8592643737792969, "num_tokens": 366526569.0, "step": 9652 }, { "epoch": 1.2279608192341942, "ewc_loss": 0.0966796875, "ewc_loss_diag": 0.00018596649169921875, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.503044128417969, "learning_rate": 1e-05, "loss": 0.495, "mean_token_accuracy": 0.8639309406280518, "num_tokens": 366567503.0, "step": 9653 }, { "epoch": 1.2280880295127847, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.00018596649169921875, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.714375972747803, "learning_rate": 1e-05, "loss": 0.5012, "mean_token_accuracy": 0.8674008846282959, "num_tokens": 366609230.0, "step": 9654 }, { "epoch": 1.2282152397913753, "ewc_loss": 0.09814453125, "ewc_loss_diag": 0.00018596649169921875, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.476559162139893, "learning_rate": 1e-05, "loss": 0.4834, "mean_token_accuracy": 0.8709262609481812, "num_tokens": 366646982.0, "step": 9655 }, { "epoch": 1.2283424500699656, "ewc_loss": 0.125, "ewc_loss_diag": 0.00018596649169921875, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.974513053894043, "learning_rate": 1e-05, "loss": 0.5591, "mean_token_accuracy": 0.8479063510894775, "num_tokens": 366684051.0, "step": 9656 }, { "epoch": 1.228469660348556, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.00018596649169921875, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.663869380950928, "learning_rate": 1e-05, "loss": 0.5322, "mean_token_accuracy": 0.8532031774520874, "num_tokens": 366724617.0, "step": 9657 }, { "epoch": 1.2285968706271466, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.00018596649169921875, "ewc_loss_parallel": 0.00010967254638671875, "grad_norm": 7.066109657287598, "learning_rate": 1e-05, "loss": 0.4758, "mean_token_accuracy": 0.8834503889083862, "num_tokens": 366756642.0, "step": 9658 }, { "epoch": 1.2287240809057371, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.00018596649169921875, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 5.998847007751465, "learning_rate": 1e-05, "loss": 0.5122, "mean_token_accuracy": 0.8682858943939209, "num_tokens": 366789197.0, "step": 9659 }, { "epoch": 1.2288512911843277, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.00018596649169921875, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.790793418884277, "learning_rate": 1e-05, "loss": 0.5232, "mean_token_accuracy": 0.8643162250518799, "num_tokens": 366825684.0, "step": 9660 }, { "epoch": 1.2289785014629182, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.00018596649169921875, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.174572944641113, "learning_rate": 1e-05, "loss": 0.5343, "mean_token_accuracy": 0.8566479682922363, "num_tokens": 366871597.0, "step": 9661 }, { "epoch": 1.2291057117415087, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.00018596649169921875, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.32478141784668, "learning_rate": 1e-05, "loss": 0.4443, "mean_token_accuracy": 0.8851314783096313, "num_tokens": 366910961.0, "step": 9662 }, { "epoch": 1.2292329220200993, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.00018596649169921875, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.253326892852783, "learning_rate": 1e-05, "loss": 0.4607, "mean_token_accuracy": 0.8826492428779602, "num_tokens": 366951701.0, "step": 9663 }, { "epoch": 1.2293601322986898, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.00018596649169921875, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.8516621589660645, "learning_rate": 1e-05, "loss": 0.5204, "mean_token_accuracy": 0.858048677444458, "num_tokens": 366992472.0, "step": 9664 }, { "epoch": 1.2294873425772803, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.00018596649169921875, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.428544044494629, "learning_rate": 1e-05, "loss": 0.537, "mean_token_accuracy": 0.861206591129303, "num_tokens": 367031220.0, "step": 9665 }, { "epoch": 1.2296145528558708, "ewc_loss": 0.09716796875, "ewc_loss_diag": 0.00018596649169921875, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.515899658203125, "learning_rate": 1e-05, "loss": 0.5132, "mean_token_accuracy": 0.8624599575996399, "num_tokens": 367067275.0, "step": 9666 }, { "epoch": 1.2297417631344612, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.00018596649169921875, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.732921123504639, "learning_rate": 1e-05, "loss": 0.5548, "mean_token_accuracy": 0.8512973785400391, "num_tokens": 367111851.0, "step": 9667 }, { "epoch": 1.2298689734130517, "ewc_loss": 0.09716796875, "ewc_loss_diag": 0.00018596649169921875, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.392323970794678, "learning_rate": 1e-05, "loss": 0.5014, "mean_token_accuracy": 0.8620855808258057, "num_tokens": 367150895.0, "step": 9668 }, { "epoch": 1.2299961836916422, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.00018596649169921875, "ewc_loss_parallel": 0.00011157989501953125, "grad_norm": 7.127273082733154, "learning_rate": 1e-05, "loss": 0.5537, "mean_token_accuracy": 0.8547425866127014, "num_tokens": 367187874.0, "step": 9669 }, { "epoch": 1.2301233939702327, "ewc_loss": 0.10302734375, "ewc_loss_diag": 0.00018596649169921875, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.621910095214844, "learning_rate": 1e-05, "loss": 0.4728, "mean_token_accuracy": 0.8726786375045776, "num_tokens": 367228859.0, "step": 9670 }, { "epoch": 1.2302506042488233, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.00018596649169921875, "ewc_loss_parallel": 0.00011682510375976562, "grad_norm": 7.32301664352417, "learning_rate": 1e-05, "loss": 0.4725, "mean_token_accuracy": 0.8833104968070984, "num_tokens": 367264686.0, "step": 9671 }, { "epoch": 1.2303778145274138, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.00018596649169921875, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.0540056228637695, "learning_rate": 1e-05, "loss": 0.4828, "mean_token_accuracy": 0.8732630014419556, "num_tokens": 367309909.0, "step": 9672 }, { "epoch": 1.2305050248060043, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.00018596649169921875, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 7.088740348815918, "learning_rate": 1e-05, "loss": 0.4843, "mean_token_accuracy": 0.8771835565567017, "num_tokens": 367343996.0, "step": 9673 }, { "epoch": 1.2306322350845948, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.00018596649169921875, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.256308078765869, "learning_rate": 1e-05, "loss": 0.5158, "mean_token_accuracy": 0.864646315574646, "num_tokens": 367386267.0, "step": 9674 }, { "epoch": 1.2307594453631854, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.00018596649169921875, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.704957485198975, "learning_rate": 1e-05, "loss": 0.5276, "mean_token_accuracy": 0.8596197962760925, "num_tokens": 367430256.0, "step": 9675 }, { "epoch": 1.230886655641776, "ewc_loss": 0.11474609375, "ewc_loss_diag": 0.00018596649169921875, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.209889888763428, "learning_rate": 1e-05, "loss": 0.5488, "mean_token_accuracy": 0.853415846824646, "num_tokens": 367469189.0, "step": 9676 }, { "epoch": 1.2310138659203664, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.00018596649169921875, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.399734020233154, "learning_rate": 1e-05, "loss": 0.55, "mean_token_accuracy": 0.8537992238998413, "num_tokens": 367501323.0, "step": 9677 }, { "epoch": 1.231141076198957, "ewc_loss": 0.10888671875, "ewc_loss_diag": 0.00018596649169921875, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.087207794189453, "learning_rate": 1e-05, "loss": 0.5195, "mean_token_accuracy": 0.8629715442657471, "num_tokens": 367542361.0, "step": 9678 }, { "epoch": 1.2312682864775475, "ewc_loss": 0.10888671875, "ewc_loss_diag": 0.00018596649169921875, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.222131252288818, "learning_rate": 1e-05, "loss": 0.546, "mean_token_accuracy": 0.8513457775115967, "num_tokens": 367575859.0, "step": 9679 }, { "epoch": 1.231395496756138, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.00018596649169921875, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.928520202636719, "learning_rate": 1e-05, "loss": 0.4887, "mean_token_accuracy": 0.8677241802215576, "num_tokens": 367611330.0, "step": 9680 }, { "epoch": 1.2315227070347283, "ewc_loss": 0.10791015625, "ewc_loss_diag": 0.00018596649169921875, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 6.170459270477295, "learning_rate": 1e-05, "loss": 0.4588, "mean_token_accuracy": 0.8787215352058411, "num_tokens": 367653110.0, "step": 9681 }, { "epoch": 1.2316499173133189, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.00018596649169921875, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.7695770263671875, "learning_rate": 1e-05, "loss": 0.4596, "mean_token_accuracy": 0.8748623132705688, "num_tokens": 367687799.0, "step": 9682 }, { "epoch": 1.2317771275919094, "ewc_loss": 0.109375, "ewc_loss_diag": 0.00018596649169921875, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.232380390167236, "learning_rate": 1e-05, "loss": 0.4902, "mean_token_accuracy": 0.8696039915084839, "num_tokens": 367728253.0, "step": 9683 }, { "epoch": 1.2319043378705, "ewc_loss": 0.0986328125, "ewc_loss_diag": 0.00018596649169921875, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.60141658782959, "learning_rate": 1e-05, "loss": 0.3969, "mean_token_accuracy": 0.8966647386550903, "num_tokens": 367770397.0, "step": 9684 }, { "epoch": 1.2320315481490904, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.00018596649169921875, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.426703929901123, "learning_rate": 1e-05, "loss": 0.4991, "mean_token_accuracy": 0.8674987554550171, "num_tokens": 367810630.0, "step": 9685 }, { "epoch": 1.232158758427681, "ewc_loss": 0.095703125, "ewc_loss_diag": 0.00018596649169921875, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.460775852203369, "learning_rate": 1e-05, "loss": 0.5065, "mean_token_accuracy": 0.8610912561416626, "num_tokens": 367853563.0, "step": 9686 }, { "epoch": 1.2322859687062715, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.00018596649169921875, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.751975059509277, "learning_rate": 1e-05, "loss": 0.5512, "mean_token_accuracy": 0.855306088924408, "num_tokens": 367888113.0, "step": 9687 }, { "epoch": 1.232413178984862, "ewc_loss": 0.09765625, "ewc_loss_diag": 0.00018596649169921875, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.444752216339111, "learning_rate": 1e-05, "loss": 0.4984, "mean_token_accuracy": 0.8626756072044373, "num_tokens": 367925839.0, "step": 9688 }, { "epoch": 1.2325403892634526, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.00018596649169921875, "ewc_loss_parallel": 0.00010967254638671875, "grad_norm": 7.082884788513184, "learning_rate": 1e-05, "loss": 0.5508, "mean_token_accuracy": 0.8580582141876221, "num_tokens": 367966944.0, "step": 9689 }, { "epoch": 1.232667599542043, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.00018596649169921875, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.699612140655518, "learning_rate": 1e-05, "loss": 0.4679, "mean_token_accuracy": 0.8693801760673523, "num_tokens": 367999470.0, "step": 9690 }, { "epoch": 1.2327948098206334, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.00018596649169921875, "ewc_loss_parallel": 0.00011157989501953125, "grad_norm": 7.155259609222412, "learning_rate": 1e-05, "loss": 0.4565, "mean_token_accuracy": 0.8867992162704468, "num_tokens": 368042780.0, "step": 9691 }, { "epoch": 1.232922020099224, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.00018596649169921875, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.066769599914551, "learning_rate": 1e-05, "loss": 0.4747, "mean_token_accuracy": 0.8725999593734741, "num_tokens": 368078062.0, "step": 9692 }, { "epoch": 1.2330492303778144, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.00018596649169921875, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.841711044311523, "learning_rate": 1e-05, "loss": 0.5149, "mean_token_accuracy": 0.8690077662467957, "num_tokens": 368121570.0, "step": 9693 }, { "epoch": 1.233176440656405, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.00018596649169921875, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.258459091186523, "learning_rate": 1e-05, "loss": 0.4761, "mean_token_accuracy": 0.8752243518829346, "num_tokens": 368158576.0, "step": 9694 }, { "epoch": 1.2333036509349955, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.00018596649169921875, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.338134288787842, "learning_rate": 1e-05, "loss": 0.4406, "mean_token_accuracy": 0.885907769203186, "num_tokens": 368199939.0, "step": 9695 }, { "epoch": 1.233430861213586, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.00018596649169921875, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.313966274261475, "learning_rate": 1e-05, "loss": 0.5181, "mean_token_accuracy": 0.8612635135650635, "num_tokens": 368240156.0, "step": 9696 }, { "epoch": 1.2335580714921766, "ewc_loss": 0.10107421875, "ewc_loss_diag": 0.00018596649169921875, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.875918865203857, "learning_rate": 1e-05, "loss": 0.5318, "mean_token_accuracy": 0.8491829633712769, "num_tokens": 368270852.0, "step": 9697 }, { "epoch": 1.233685281770767, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.00018596649169921875, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.458339691162109, "learning_rate": 1e-05, "loss": 0.5201, "mean_token_accuracy": 0.8634041547775269, "num_tokens": 368311986.0, "step": 9698 }, { "epoch": 1.2338124920493576, "ewc_loss": 0.09716796875, "ewc_loss_diag": 0.00018596649169921875, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.53654670715332, "learning_rate": 1e-05, "loss": 0.506, "mean_token_accuracy": 0.8624904155731201, "num_tokens": 368354330.0, "step": 9699 }, { "epoch": 1.2339397023279481, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.00018596649169921875, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.729175567626953, "learning_rate": 1e-05, "loss": 0.4692, "mean_token_accuracy": 0.87734454870224, "num_tokens": 368392712.0, "step": 9700 }, { "epoch": 1.2340669126065387, "ewc_loss": 0.09619140625, "ewc_loss_diag": 0.00018596649169921875, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.446651458740234, "learning_rate": 1e-05, "loss": 0.5074, "mean_token_accuracy": 0.8605730533599854, "num_tokens": 368430913.0, "step": 9701 }, { "epoch": 1.2341941228851292, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.000186920166015625, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 7.06209135055542, "learning_rate": 1e-05, "loss": 0.4743, "mean_token_accuracy": 0.883149266242981, "num_tokens": 368462353.0, "step": 9702 }, { "epoch": 1.2343213331637197, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.000186920166015625, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.636080741882324, "learning_rate": 1e-05, "loss": 0.5058, "mean_token_accuracy": 0.8603662252426147, "num_tokens": 368506061.0, "step": 9703 }, { "epoch": 1.2344485434423103, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.000186920166015625, "ewc_loss_parallel": 0.00011348724365234375, "grad_norm": 7.221898078918457, "learning_rate": 1e-05, "loss": 0.4857, "mean_token_accuracy": 0.8739861249923706, "num_tokens": 368539435.0, "step": 9704 }, { "epoch": 1.2345757537209006, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.000186920166015625, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 5.995241165161133, "learning_rate": 1e-05, "loss": 0.5032, "mean_token_accuracy": 0.8664643168449402, "num_tokens": 368578249.0, "step": 9705 }, { "epoch": 1.234702963999491, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.000186920166015625, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 7.029428958892822, "learning_rate": 1e-05, "loss": 0.4811, "mean_token_accuracy": 0.8820034265518188, "num_tokens": 368617766.0, "step": 9706 }, { "epoch": 1.2348301742780816, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.000186920166015625, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.174105644226074, "learning_rate": 1e-05, "loss": 0.4648, "mean_token_accuracy": 0.8769278526306152, "num_tokens": 368651458.0, "step": 9707 }, { "epoch": 1.2349573845566721, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.000186920166015625, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.6955485343933105, "learning_rate": 1e-05, "loss": 0.5662, "mean_token_accuracy": 0.8495205044746399, "num_tokens": 368688689.0, "step": 9708 }, { "epoch": 1.2350845948352627, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.000186920166015625, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.1538801193237305, "learning_rate": 1e-05, "loss": 0.5529, "mean_token_accuracy": 0.8479655385017395, "num_tokens": 368731996.0, "step": 9709 }, { "epoch": 1.2352118051138532, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.000186920166015625, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.379922389984131, "learning_rate": 1e-05, "loss": 0.4648, "mean_token_accuracy": 0.8789991140365601, "num_tokens": 368772712.0, "step": 9710 }, { "epoch": 1.2353390153924437, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.000186920166015625, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.072500228881836, "learning_rate": 1e-05, "loss": 0.4793, "mean_token_accuracy": 0.8726022243499756, "num_tokens": 368816582.0, "step": 9711 }, { "epoch": 1.2354662256710343, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.000186920166015625, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.190035820007324, "learning_rate": 1e-05, "loss": 0.4891, "mean_token_accuracy": 0.8720300197601318, "num_tokens": 368854284.0, "step": 9712 }, { "epoch": 1.2355934359496248, "ewc_loss": 0.10400390625, "ewc_loss_diag": 0.000186920166015625, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.949641704559326, "learning_rate": 1e-05, "loss": 0.4901, "mean_token_accuracy": 0.8694686889648438, "num_tokens": 368891564.0, "step": 9713 }, { "epoch": 1.2357206462282153, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.000186920166015625, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 6.131737232208252, "learning_rate": 1e-05, "loss": 0.5298, "mean_token_accuracy": 0.8525602221488953, "num_tokens": 368931176.0, "step": 9714 }, { "epoch": 1.2358478565068058, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.000186920166015625, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.814207077026367, "learning_rate": 1e-05, "loss": 0.4698, "mean_token_accuracy": 0.8734723329544067, "num_tokens": 368967730.0, "step": 9715 }, { "epoch": 1.2359750667853961, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.000186920166015625, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.153697490692139, "learning_rate": 1e-05, "loss": 0.4798, "mean_token_accuracy": 0.8707985877990723, "num_tokens": 369009378.0, "step": 9716 }, { "epoch": 1.2361022770639867, "ewc_loss": 0.0986328125, "ewc_loss_diag": 0.000186920166015625, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.670479774475098, "learning_rate": 1e-05, "loss": 0.5007, "mean_token_accuracy": 0.8633583188056946, "num_tokens": 369045716.0, "step": 9717 }, { "epoch": 1.2362294873425772, "ewc_loss": 0.11083984375, "ewc_loss_diag": 0.000186920166015625, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.274523735046387, "learning_rate": 1e-05, "loss": 0.4597, "mean_token_accuracy": 0.878570556640625, "num_tokens": 369082022.0, "step": 9718 }, { "epoch": 1.2363566976211677, "ewc_loss": 0.0966796875, "ewc_loss_diag": 0.000186920166015625, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.527590274810791, "learning_rate": 1e-05, "loss": 0.5104, "mean_token_accuracy": 0.8592195510864258, "num_tokens": 369122456.0, "step": 9719 }, { "epoch": 1.2364839078997583, "ewc_loss": 0.11474609375, "ewc_loss_diag": 0.000186920166015625, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.521120071411133, "learning_rate": 1e-05, "loss": 0.5431, "mean_token_accuracy": 0.8558362722396851, "num_tokens": 369156977.0, "step": 9720 }, { "epoch": 1.2366111181783488, "ewc_loss": 0.0966796875, "ewc_loss_diag": 0.000186920166015625, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.418487548828125, "learning_rate": 1e-05, "loss": 0.5147, "mean_token_accuracy": 0.8574270009994507, "num_tokens": 369195113.0, "step": 9721 }, { "epoch": 1.2367383284569393, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.000186920166015625, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.848541259765625, "learning_rate": 1e-05, "loss": 0.4707, "mean_token_accuracy": 0.8792877197265625, "num_tokens": 369238650.0, "step": 9722 }, { "epoch": 1.2368655387355298, "ewc_loss": 0.0986328125, "ewc_loss_diag": 0.000186920166015625, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.467256546020508, "learning_rate": 1e-05, "loss": 0.5418, "mean_token_accuracy": 0.8480955362319946, "num_tokens": 369275753.0, "step": 9723 }, { "epoch": 1.2369927490141204, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.000186920166015625, "ewc_loss_parallel": 0.0001125335693359375, "grad_norm": 7.168515682220459, "learning_rate": 1e-05, "loss": 0.5813, "mean_token_accuracy": 0.8434258699417114, "num_tokens": 369317364.0, "step": 9724 }, { "epoch": 1.237119959292711, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.000186920166015625, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 5.791940689086914, "learning_rate": 1e-05, "loss": 0.4991, "mean_token_accuracy": 0.8718419671058655, "num_tokens": 369355889.0, "step": 9725 }, { "epoch": 1.2372471695713014, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.000186920166015625, "ewc_loss_parallel": 0.00011110305786132812, "grad_norm": 7.160157680511475, "learning_rate": 1e-05, "loss": 0.4618, "mean_token_accuracy": 0.8812896013259888, "num_tokens": 369393749.0, "step": 9726 }, { "epoch": 1.237374379849892, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.000186920166015625, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.156238079071045, "learning_rate": 1e-05, "loss": 0.5108, "mean_token_accuracy": 0.8676178455352783, "num_tokens": 369432442.0, "step": 9727 }, { "epoch": 1.2375015901284825, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.000186920166015625, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.778514862060547, "learning_rate": 1e-05, "loss": 0.5224, "mean_token_accuracy": 0.86653733253479, "num_tokens": 369469513.0, "step": 9728 }, { "epoch": 1.237628800407073, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.000186920166015625, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.2717509269714355, "learning_rate": 1e-05, "loss": 0.4899, "mean_token_accuracy": 0.8693894743919373, "num_tokens": 369505720.0, "step": 9729 }, { "epoch": 1.2377560106856633, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.000186920166015625, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.2898688316345215, "learning_rate": 1e-05, "loss": 0.4569, "mean_token_accuracy": 0.8759177923202515, "num_tokens": 369535225.0, "step": 9730 }, { "epoch": 1.2378832209642538, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.000186920166015625, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.3037590980529785, "learning_rate": 1e-05, "loss": 0.488, "mean_token_accuracy": 0.8727874755859375, "num_tokens": 369572580.0, "step": 9731 }, { "epoch": 1.2380104312428444, "ewc_loss": 0.10205078125, "ewc_loss_diag": 0.000186920166015625, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.849148750305176, "learning_rate": 1e-05, "loss": 0.4561, "mean_token_accuracy": 0.8744761943817139, "num_tokens": 369611566.0, "step": 9732 }, { "epoch": 1.238137641521435, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.000186920166015625, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.42592716217041, "learning_rate": 1e-05, "loss": 0.5126, "mean_token_accuracy": 0.86934894323349, "num_tokens": 369652541.0, "step": 9733 }, { "epoch": 1.2382648518000254, "ewc_loss": 0.0986328125, "ewc_loss_diag": 0.000186920166015625, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.585108757019043, "learning_rate": 1e-05, "loss": 0.5272, "mean_token_accuracy": 0.85451740026474, "num_tokens": 369686744.0, "step": 9734 }, { "epoch": 1.238392062078616, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.000186920166015625, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.632291793823242, "learning_rate": 1e-05, "loss": 0.4644, "mean_token_accuracy": 0.8776706457138062, "num_tokens": 369725090.0, "step": 9735 }, { "epoch": 1.2385192723572065, "ewc_loss": 0.09765625, "ewc_loss_diag": 0.000186920166015625, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.469838619232178, "learning_rate": 1e-05, "loss": 0.4216, "mean_token_accuracy": 0.887197732925415, "num_tokens": 369764255.0, "step": 9736 }, { "epoch": 1.238646482635797, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.000186920166015625, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.923460960388184, "learning_rate": 1e-05, "loss": 0.4897, "mean_token_accuracy": 0.8738595843315125, "num_tokens": 369803589.0, "step": 9737 }, { "epoch": 1.2387736929143875, "ewc_loss": 0.10205078125, "ewc_loss_diag": 0.000186920166015625, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.574366092681885, "learning_rate": 1e-05, "loss": 0.4533, "mean_token_accuracy": 0.8803269863128662, "num_tokens": 369837984.0, "step": 9738 }, { "epoch": 1.238900903192978, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.000186920166015625, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 7.11811637878418, "learning_rate": 1e-05, "loss": 0.5113, "mean_token_accuracy": 0.8690680861473083, "num_tokens": 369881398.0, "step": 9739 }, { "epoch": 1.2390281134715684, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.000186920166015625, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 5.848235607147217, "learning_rate": 1e-05, "loss": 0.501, "mean_token_accuracy": 0.8674548864364624, "num_tokens": 369915230.0, "step": 9740 }, { "epoch": 1.239155323750159, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.000186920166015625, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 7.053755283355713, "learning_rate": 1e-05, "loss": 0.5098, "mean_token_accuracy": 0.8682750463485718, "num_tokens": 369955237.0, "step": 9741 }, { "epoch": 1.2392825340287494, "ewc_loss": 0.11474609375, "ewc_loss_diag": 0.000186920166015625, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.063340663909912, "learning_rate": 1e-05, "loss": 0.5022, "mean_token_accuracy": 0.864247739315033, "num_tokens": 369995938.0, "step": 9742 }, { "epoch": 1.23940974430734, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.000186920166015625, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.786318778991699, "learning_rate": 1e-05, "loss": 0.531, "mean_token_accuracy": 0.8585003614425659, "num_tokens": 370037346.0, "step": 9743 }, { "epoch": 1.2395369545859305, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.000186920166015625, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.1151885986328125, "learning_rate": 1e-05, "loss": 0.4976, "mean_token_accuracy": 0.8676499724388123, "num_tokens": 370067222.0, "step": 9744 }, { "epoch": 1.239664164864521, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.000186920166015625, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.4895453453063965, "learning_rate": 1e-05, "loss": 0.4637, "mean_token_accuracy": 0.8811078071594238, "num_tokens": 370101438.0, "step": 9745 }, { "epoch": 1.2397913751431116, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.000186920166015625, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.058177471160889, "learning_rate": 1e-05, "loss": 0.5012, "mean_token_accuracy": 0.8649238348007202, "num_tokens": 370136887.0, "step": 9746 }, { "epoch": 1.239918585421702, "ewc_loss": 0.109375, "ewc_loss_diag": 0.000186920166015625, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.259468078613281, "learning_rate": 1e-05, "loss": 0.4735, "mean_token_accuracy": 0.876807689666748, "num_tokens": 370176797.0, "step": 9747 }, { "epoch": 1.2400457957002926, "ewc_loss": 0.10498046875, "ewc_loss_diag": 0.000186920166015625, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.9811625480651855, "learning_rate": 1e-05, "loss": 0.5035, "mean_token_accuracy": 0.8627687096595764, "num_tokens": 370206246.0, "step": 9748 }, { "epoch": 1.2401730059788831, "ewc_loss": 0.109375, "ewc_loss_diag": 0.000186920166015625, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.129589080810547, "learning_rate": 1e-05, "loss": 0.5019, "mean_token_accuracy": 0.8689025640487671, "num_tokens": 370246132.0, "step": 9749 }, { "epoch": 1.2403002162574737, "ewc_loss": 0.10205078125, "ewc_loss_diag": 0.000186920166015625, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.860550880432129, "learning_rate": 1e-05, "loss": 0.47, "mean_token_accuracy": 0.8734469413757324, "num_tokens": 370280039.0, "step": 9750 }, { "epoch": 1.2404274265360642, "ewc_loss": 0.10595703125, "ewc_loss_diag": 0.000186920166015625, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 6.108595848083496, "learning_rate": 1e-05, "loss": 0.4705, "mean_token_accuracy": 0.8744919896125793, "num_tokens": 370317522.0, "step": 9751 }, { "epoch": 1.2405546368146547, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.000186920166015625, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.730820655822754, "learning_rate": 1e-05, "loss": 0.4695, "mean_token_accuracy": 0.8715444803237915, "num_tokens": 370352350.0, "step": 9752 }, { "epoch": 1.2406818470932452, "ewc_loss": 0.10791015625, "ewc_loss_diag": 0.000186920166015625, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 6.195284366607666, "learning_rate": 1e-05, "loss": 0.4711, "mean_token_accuracy": 0.8780770301818848, "num_tokens": 370382645.0, "step": 9753 }, { "epoch": 1.2408090573718356, "ewc_loss": 0.09814453125, "ewc_loss_diag": 0.000186920166015625, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.585784435272217, "learning_rate": 1e-05, "loss": 0.4892, "mean_token_accuracy": 0.8729407787322998, "num_tokens": 370422679.0, "step": 9754 }, { "epoch": 1.240936267650426, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.00018787384033203125, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.370236873626709, "learning_rate": 1e-05, "loss": 0.4899, "mean_token_accuracy": 0.8668506145477295, "num_tokens": 370458475.0, "step": 9755 }, { "epoch": 1.2410634779290166, "ewc_loss": 0.0966796875, "ewc_loss_diag": 0.00018787384033203125, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.45711088180542, "learning_rate": 1e-05, "loss": 0.4928, "mean_token_accuracy": 0.8642066717147827, "num_tokens": 370489641.0, "step": 9756 }, { "epoch": 1.2411906882076071, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.000186920166015625, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.675756454467773, "learning_rate": 1e-05, "loss": 0.496, "mean_token_accuracy": 0.8694159388542175, "num_tokens": 370526649.0, "step": 9757 }, { "epoch": 1.2413178984861977, "ewc_loss": 0.09765625, "ewc_loss_diag": 0.00018787384033203125, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.402417182922363, "learning_rate": 1e-05, "loss": 0.4966, "mean_token_accuracy": 0.8637301921844482, "num_tokens": 370564664.0, "step": 9758 }, { "epoch": 1.2414451087647882, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.00018787384033203125, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 7.034931659698486, "learning_rate": 1e-05, "loss": 0.4868, "mean_token_accuracy": 0.8784583806991577, "num_tokens": 370598000.0, "step": 9759 }, { "epoch": 1.2415723190433787, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.00018787384033203125, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.590467929840088, "learning_rate": 1e-05, "loss": 0.5226, "mean_token_accuracy": 0.8577352166175842, "num_tokens": 370639318.0, "step": 9760 }, { "epoch": 1.2416995293219693, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.00018787384033203125, "ewc_loss_parallel": 0.00011444091796875, "grad_norm": 7.217533111572266, "learning_rate": 1e-05, "loss": 0.5594, "mean_token_accuracy": 0.8593704104423523, "num_tokens": 370676064.0, "step": 9761 }, { "epoch": 1.2418267396005598, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.00018787384033203125, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 5.991962909698486, "learning_rate": 1e-05, "loss": 0.4832, "mean_token_accuracy": 0.8720298409461975, "num_tokens": 370713190.0, "step": 9762 }, { "epoch": 1.2419539498791503, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.00018787384033203125, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 6.973275661468506, "learning_rate": 1e-05, "loss": 0.4741, "mean_token_accuracy": 0.8821496963500977, "num_tokens": 370748433.0, "step": 9763 }, { "epoch": 1.2420811601577408, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.00018787384033203125, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.2342143058776855, "learning_rate": 1e-05, "loss": 0.4769, "mean_token_accuracy": 0.8756126761436462, "num_tokens": 370781522.0, "step": 9764 }, { "epoch": 1.2422083704363311, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.000186920166015625, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.499629020690918, "learning_rate": 1e-05, "loss": 0.4886, "mean_token_accuracy": 0.8709395527839661, "num_tokens": 370821878.0, "step": 9765 }, { "epoch": 1.2423355807149217, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.00018787384033203125, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.287036418914795, "learning_rate": 1e-05, "loss": 0.5194, "mean_token_accuracy": 0.8645904064178467, "num_tokens": 370857754.0, "step": 9766 }, { "epoch": 1.2424627909935122, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.00018787384033203125, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 6.022510051727295, "learning_rate": 1e-05, "loss": 0.4934, "mean_token_accuracy": 0.8624899983406067, "num_tokens": 370893672.0, "step": 9767 }, { "epoch": 1.2425900012721027, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.00018787384033203125, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.376219272613525, "learning_rate": 1e-05, "loss": 0.5105, "mean_token_accuracy": 0.8630562424659729, "num_tokens": 370928118.0, "step": 9768 }, { "epoch": 1.2427172115506933, "ewc_loss": 0.099609375, "ewc_loss_diag": 0.00018787384033203125, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.680721282958984, "learning_rate": 1e-05, "loss": 0.539, "mean_token_accuracy": 0.8502120971679688, "num_tokens": 370960748.0, "step": 9769 }, { "epoch": 1.2428444218292838, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.00018787384033203125, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.554940223693848, "learning_rate": 1e-05, "loss": 0.4675, "mean_token_accuracy": 0.8808953762054443, "num_tokens": 370992408.0, "step": 9770 }, { "epoch": 1.2429716321078743, "ewc_loss": 0.09765625, "ewc_loss_diag": 0.00018787384033203125, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.494030475616455, "learning_rate": 1e-05, "loss": 0.5371, "mean_token_accuracy": 0.8542931079864502, "num_tokens": 371035020.0, "step": 9771 }, { "epoch": 1.2430988423864648, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.00018787384033203125, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.806604862213135, "learning_rate": 1e-05, "loss": 0.4815, "mean_token_accuracy": 0.8753683567047119, "num_tokens": 371072738.0, "step": 9772 }, { "epoch": 1.2432260526650554, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.00018787384033203125, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.52031135559082, "learning_rate": 1e-05, "loss": 0.455, "mean_token_accuracy": 0.8764384388923645, "num_tokens": 371109246.0, "step": 9773 }, { "epoch": 1.243353262943646, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.00018787384033203125, "ewc_loss_parallel": 0.00010967254638671875, "grad_norm": 7.0431108474731445, "learning_rate": 1e-05, "loss": 0.474, "mean_token_accuracy": 0.877062976360321, "num_tokens": 371148322.0, "step": 9774 }, { "epoch": 1.2434804732222364, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.00018787384033203125, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 5.751372814178467, "learning_rate": 1e-05, "loss": 0.5275, "mean_token_accuracy": 0.857526421546936, "num_tokens": 371186534.0, "step": 9775 }, { "epoch": 1.243607683500827, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.00018787384033203125, "ewc_loss_parallel": 0.00010967254638671875, "grad_norm": 7.059738636016846, "learning_rate": 1e-05, "loss": 0.5058, "mean_token_accuracy": 0.8700511455535889, "num_tokens": 371224973.0, "step": 9776 }, { "epoch": 1.2437348937794175, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.00018787384033203125, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 5.989762783050537, "learning_rate": 1e-05, "loss": 0.4878, "mean_token_accuracy": 0.8702482581138611, "num_tokens": 371262629.0, "step": 9777 }, { "epoch": 1.243862104058008, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.00018787384033203125, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.856090068817139, "learning_rate": 1e-05, "loss": 0.5179, "mean_token_accuracy": 0.8677691221237183, "num_tokens": 371298421.0, "step": 9778 }, { "epoch": 1.2439893143365983, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.00018787384033203125, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.093044281005859, "learning_rate": 1e-05, "loss": 0.5609, "mean_token_accuracy": 0.8460845351219177, "num_tokens": 371337393.0, "step": 9779 }, { "epoch": 1.2441165246151888, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.00018787384033203125, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.558566093444824, "learning_rate": 1e-05, "loss": 0.4805, "mean_token_accuracy": 0.8729605674743652, "num_tokens": 371372436.0, "step": 9780 }, { "epoch": 1.2442437348937794, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.00018787384033203125, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.064366817474365, "learning_rate": 1e-05, "loss": 0.484, "mean_token_accuracy": 0.8712694644927979, "num_tokens": 371404269.0, "step": 9781 }, { "epoch": 1.24437094517237, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.00018787384033203125, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.297471523284912, "learning_rate": 1e-05, "loss": 0.5031, "mean_token_accuracy": 0.8704684972763062, "num_tokens": 371446494.0, "step": 9782 }, { "epoch": 1.2444981554509604, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.00018787384033203125, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 5.989902019500732, "learning_rate": 1e-05, "loss": 0.5205, "mean_token_accuracy": 0.8598912358283997, "num_tokens": 371482524.0, "step": 9783 }, { "epoch": 1.244625365729551, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.00018787384033203125, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.149467945098877, "learning_rate": 1e-05, "loss": 0.5728, "mean_token_accuracy": 0.8475387096405029, "num_tokens": 371519745.0, "step": 9784 }, { "epoch": 1.2447525760081415, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.00018787384033203125, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.892681121826172, "learning_rate": 1e-05, "loss": 0.4891, "mean_token_accuracy": 0.872195839881897, "num_tokens": 371563593.0, "step": 9785 }, { "epoch": 1.244879786286732, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.00018787384033203125, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 6.094451904296875, "learning_rate": 1e-05, "loss": 0.4712, "mean_token_accuracy": 0.873349666595459, "num_tokens": 371600646.0, "step": 9786 }, { "epoch": 1.2450069965653225, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.00018787384033203125, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.791153430938721, "learning_rate": 1e-05, "loss": 0.4603, "mean_token_accuracy": 0.8766080141067505, "num_tokens": 371635071.0, "step": 9787 }, { "epoch": 1.245134206843913, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.00018787384033203125, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 6.1251630783081055, "learning_rate": 1e-05, "loss": 0.486, "mean_token_accuracy": 0.8683720231056213, "num_tokens": 371672441.0, "step": 9788 }, { "epoch": 1.2452614171225034, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.00018787384033203125, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.649168968200684, "learning_rate": 1e-05, "loss": 0.4567, "mean_token_accuracy": 0.8736737370491028, "num_tokens": 371712751.0, "step": 9789 }, { "epoch": 1.245388627401094, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.00018787384033203125, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.258121013641357, "learning_rate": 1e-05, "loss": 0.4769, "mean_token_accuracy": 0.870146632194519, "num_tokens": 371748863.0, "step": 9790 }, { "epoch": 1.2455158376796844, "ewc_loss": 0.09765625, "ewc_loss_diag": 0.00018787384033203125, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.520791053771973, "learning_rate": 1e-05, "loss": 0.5111, "mean_token_accuracy": 0.8582568168640137, "num_tokens": 371784763.0, "step": 9791 }, { "epoch": 1.245643047958275, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.00018787384033203125, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.492457866668701, "learning_rate": 1e-05, "loss": 0.5592, "mean_token_accuracy": 0.8475427627563477, "num_tokens": 371822722.0, "step": 9792 }, { "epoch": 1.2457702582368655, "ewc_loss": 0.09765625, "ewc_loss_diag": 0.00018787384033203125, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.401302337646484, "learning_rate": 1e-05, "loss": 0.4138, "mean_token_accuracy": 0.8887306451797485, "num_tokens": 371855423.0, "step": 9793 }, { "epoch": 1.245897468515456, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.00018787384033203125, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.843197345733643, "learning_rate": 1e-05, "loss": 0.528, "mean_token_accuracy": 0.8633356094360352, "num_tokens": 371897900.0, "step": 9794 }, { "epoch": 1.2460246787940465, "ewc_loss": 0.0986328125, "ewc_loss_diag": 0.00018787384033203125, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.429161071777344, "learning_rate": 1e-05, "loss": 0.47, "mean_token_accuracy": 0.8706364631652832, "num_tokens": 371942394.0, "step": 9795 }, { "epoch": 1.246151889072637, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.00018787384033203125, "ewc_loss_parallel": 0.00011491775512695312, "grad_norm": 7.200412273406982, "learning_rate": 1e-05, "loss": 0.581, "mean_token_accuracy": 0.8476762771606445, "num_tokens": 371982391.0, "step": 9796 }, { "epoch": 1.2462790993512276, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.00018787384033203125, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 5.7562971115112305, "learning_rate": 1e-05, "loss": 0.4894, "mean_token_accuracy": 0.8691315054893494, "num_tokens": 372018599.0, "step": 9797 }, { "epoch": 1.2464063096298181, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.00018787384033203125, "ewc_loss_parallel": 0.00011491775512695312, "grad_norm": 7.237847328186035, "learning_rate": 1e-05, "loss": 0.5248, "mean_token_accuracy": 0.8642168045043945, "num_tokens": 372058650.0, "step": 9798 }, { "epoch": 1.2465335199084087, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.00018787384033203125, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.1597819328308105, "learning_rate": 1e-05, "loss": 0.5091, "mean_token_accuracy": 0.8642096519470215, "num_tokens": 372102272.0, "step": 9799 }, { "epoch": 1.2466607301869992, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.00018787384033203125, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.865482330322266, "learning_rate": 1e-05, "loss": 0.4489, "mean_token_accuracy": 0.8868262767791748, "num_tokens": 372139086.0, "step": 9800 }, { "epoch": 1.2467879404655897, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.00018787384033203125, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.290379524230957, "learning_rate": 1e-05, "loss": 0.4888, "mean_token_accuracy": 0.8708485960960388, "num_tokens": 372178418.0, "step": 9801 }, { "epoch": 1.2469151507441802, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.00018787384033203125, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.379684925079346, "learning_rate": 1e-05, "loss": 0.4578, "mean_token_accuracy": 0.8802411556243896, "num_tokens": 372218837.0, "step": 9802 }, { "epoch": 1.2470423610227706, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.00018787384033203125, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.306239604949951, "learning_rate": 1e-05, "loss": 0.5395, "mean_token_accuracy": 0.8525331020355225, "num_tokens": 372256620.0, "step": 9803 }, { "epoch": 1.247169571301361, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.00018787384033203125, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.966362953186035, "learning_rate": 1e-05, "loss": 0.502, "mean_token_accuracy": 0.863452136516571, "num_tokens": 372292379.0, "step": 9804 }, { "epoch": 1.2472967815799516, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.00018787384033203125, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.355409622192383, "learning_rate": 1e-05, "loss": 0.4613, "mean_token_accuracy": 0.8772904276847839, "num_tokens": 372336319.0, "step": 9805 }, { "epoch": 1.2474239918585421, "ewc_loss": 0.099609375, "ewc_loss_diag": 0.00018787384033203125, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.677601337432861, "learning_rate": 1e-05, "loss": 0.5023, "mean_token_accuracy": 0.8618892431259155, "num_tokens": 372372965.0, "step": 9806 }, { "epoch": 1.2475512021371327, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.00018787384033203125, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.506871700286865, "learning_rate": 1e-05, "loss": 0.4825, "mean_token_accuracy": 0.871656060218811, "num_tokens": 372415453.0, "step": 9807 }, { "epoch": 1.2476784124157232, "ewc_loss": 0.0966796875, "ewc_loss_diag": 0.00018787384033203125, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.518054008483887, "learning_rate": 1e-05, "loss": 0.4993, "mean_token_accuracy": 0.8629915714263916, "num_tokens": 372452822.0, "step": 9808 }, { "epoch": 1.2478056226943137, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.00018787384033203125, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.740149021148682, "learning_rate": 1e-05, "loss": 0.5051, "mean_token_accuracy": 0.8657684326171875, "num_tokens": 372494316.0, "step": 9809 }, { "epoch": 1.2479328329729042, "ewc_loss": 0.099609375, "ewc_loss_diag": 0.00018787384033203125, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.506382942199707, "learning_rate": 1e-05, "loss": 0.4437, "mean_token_accuracy": 0.8792532086372375, "num_tokens": 372531334.0, "step": 9810 }, { "epoch": 1.2480600432514948, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.00018787384033203125, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 6.970939636230469, "learning_rate": 1e-05, "loss": 0.4948, "mean_token_accuracy": 0.875767707824707, "num_tokens": 372568303.0, "step": 9811 }, { "epoch": 1.2481872535300853, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.00018787384033203125, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.677063941955566, "learning_rate": 1e-05, "loss": 0.4461, "mean_token_accuracy": 0.8807109594345093, "num_tokens": 372602783.0, "step": 9812 }, { "epoch": 1.2483144638086758, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.00018787384033203125, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 7.054437160491943, "learning_rate": 1e-05, "loss": 0.5083, "mean_token_accuracy": 0.8714215755462646, "num_tokens": 372643883.0, "step": 9813 }, { "epoch": 1.2484416740872661, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.00018787384033203125, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 5.908965110778809, "learning_rate": 1e-05, "loss": 0.4926, "mean_token_accuracy": 0.865725576877594, "num_tokens": 372681864.0, "step": 9814 }, { "epoch": 1.2485688843658567, "ewc_loss": 0.125, "ewc_loss_diag": 0.00018787384033203125, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.924689292907715, "learning_rate": 1e-05, "loss": 0.5245, "mean_token_accuracy": 0.864820659160614, "num_tokens": 372717657.0, "step": 9815 }, { "epoch": 1.2486960946444472, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.00018787384033203125, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.034765720367432, "learning_rate": 1e-05, "loss": 0.4422, "mean_token_accuracy": 0.8845587968826294, "num_tokens": 372756702.0, "step": 9816 }, { "epoch": 1.2488233049230377, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.00018787384033203125, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.656786918640137, "learning_rate": 1e-05, "loss": 0.5036, "mean_token_accuracy": 0.8658456206321716, "num_tokens": 372795620.0, "step": 9817 }, { "epoch": 1.2489505152016283, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.00018787384033203125, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.0676727294921875, "learning_rate": 1e-05, "loss": 0.507, "mean_token_accuracy": 0.8631874322891235, "num_tokens": 372833803.0, "step": 9818 }, { "epoch": 1.2490777254802188, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.00018787384033203125, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.374618053436279, "learning_rate": 1e-05, "loss": 0.4828, "mean_token_accuracy": 0.8738977313041687, "num_tokens": 372864229.0, "step": 9819 }, { "epoch": 1.2492049357588093, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.00018787384033203125, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.033098220825195, "learning_rate": 1e-05, "loss": 0.4799, "mean_token_accuracy": 0.8717930316925049, "num_tokens": 372900321.0, "step": 9820 }, { "epoch": 1.2493321460373998, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.00018787384033203125, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.156590938568115, "learning_rate": 1e-05, "loss": 0.4902, "mean_token_accuracy": 0.867287814617157, "num_tokens": 372941237.0, "step": 9821 }, { "epoch": 1.2494593563159904, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.00018787384033203125, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.962702751159668, "learning_rate": 1e-05, "loss": 0.4893, "mean_token_accuracy": 0.8646962642669678, "num_tokens": 372980057.0, "step": 9822 }, { "epoch": 1.249586566594581, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 6.046260356903076, "learning_rate": 1e-05, "loss": 0.4945, "mean_token_accuracy": 0.8695467114448547, "num_tokens": 373014746.0, "step": 9823 }, { "epoch": 1.2497137768731714, "ewc_loss": 0.10302734375, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.876608371734619, "learning_rate": 1e-05, "loss": 0.5079, "mean_token_accuracy": 0.8594743609428406, "num_tokens": 373053298.0, "step": 9824 }, { "epoch": 1.249840987151762, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.00018787384033203125, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 6.006529331207275, "learning_rate": 1e-05, "loss": 0.4362, "mean_token_accuracy": 0.8845769166946411, "num_tokens": 373100155.0, "step": 9825 }, { "epoch": 1.2499681974303525, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.00018787384033203125, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.782916069030762, "learning_rate": 1e-05, "loss": 0.5092, "mean_token_accuracy": 0.8615710735321045, "num_tokens": 373138554.0, "step": 9826 }, { "epoch": 1.250095407708943, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.00018787384033203125, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 6.059818744659424, "learning_rate": 1e-05, "loss": 0.5256, "mean_token_accuracy": 0.8559353351593018, "num_tokens": 373173023.0, "step": 9827 }, { "epoch": 1.2502226179875333, "ewc_loss": 0.10009765625, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.669039249420166, "learning_rate": 1e-05, "loss": 0.4945, "mean_token_accuracy": 0.8640446662902832, "num_tokens": 373210630.0, "step": 9828 }, { "epoch": 1.2503498282661238, "ewc_loss": 0.10791015625, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.167435646057129, "learning_rate": 1e-05, "loss": 0.5235, "mean_token_accuracy": 0.8603461980819702, "num_tokens": 373247859.0, "step": 9829 }, { "epoch": 1.2504770385447144, "ewc_loss": 0.09765625, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.543107032775879, "learning_rate": 1e-05, "loss": 0.4743, "mean_token_accuracy": 0.8698945641517639, "num_tokens": 373285663.0, "step": 9830 }, { "epoch": 1.250604248823305, "ewc_loss": 0.11181640625, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.366661548614502, "learning_rate": 1e-05, "loss": 0.5132, "mean_token_accuracy": 0.8610612154006958, "num_tokens": 373321743.0, "step": 9831 }, { "epoch": 1.2507314591018954, "ewc_loss": 0.0966796875, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.399819374084473, "learning_rate": 1e-05, "loss": 0.5519, "mean_token_accuracy": 0.8452699184417725, "num_tokens": 373355768.0, "step": 9832 }, { "epoch": 1.250858669380486, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.696996212005615, "learning_rate": 1e-05, "loss": 0.5153, "mean_token_accuracy": 0.863303542137146, "num_tokens": 373399335.0, "step": 9833 }, { "epoch": 1.2509858796590765, "ewc_loss": 0.09765625, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.327922344207764, "learning_rate": 1e-05, "loss": 0.4928, "mean_token_accuracy": 0.8614479899406433, "num_tokens": 373433708.0, "step": 9834 }, { "epoch": 1.251113089937667, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 0.0001125335693359375, "grad_norm": 7.139810085296631, "learning_rate": 1e-05, "loss": 0.4981, "mean_token_accuracy": 0.8726247549057007, "num_tokens": 373472145.0, "step": 9835 }, { "epoch": 1.2512403002162575, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.545382499694824, "learning_rate": 1e-05, "loss": 0.4741, "mean_token_accuracy": 0.8743774890899658, "num_tokens": 373506578.0, "step": 9836 }, { "epoch": 1.2513675104948478, "ewc_loss": 0.140625, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 0.00012159347534179688, "grad_norm": 7.416998386383057, "learning_rate": 1e-05, "loss": 0.5018, "mean_token_accuracy": 0.8765391111373901, "num_tokens": 373545988.0, "step": 9837 }, { "epoch": 1.2514947207734384, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.0707902908325195, "learning_rate": 1e-05, "loss": 0.4757, "mean_token_accuracy": 0.8742451667785645, "num_tokens": 373581001.0, "step": 9838 }, { "epoch": 1.251621931052029, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 0.0001125335693359375, "grad_norm": 7.198113918304443, "learning_rate": 1e-05, "loss": 0.5281, "mean_token_accuracy": 0.8634844422340393, "num_tokens": 373614961.0, "step": 9839 }, { "epoch": 1.2517491413306194, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.338082790374756, "learning_rate": 1e-05, "loss": 0.4587, "mean_token_accuracy": 0.8813146948814392, "num_tokens": 373652385.0, "step": 9840 }, { "epoch": 1.25187635160921, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.743160724639893, "learning_rate": 1e-05, "loss": 0.5147, "mean_token_accuracy": 0.8636727333068848, "num_tokens": 373692468.0, "step": 9841 }, { "epoch": 1.2520035618878005, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.314317226409912, "learning_rate": 1e-05, "loss": 0.5008, "mean_token_accuracy": 0.872577428817749, "num_tokens": 373736885.0, "step": 9842 }, { "epoch": 1.252130772166391, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.314513683319092, "learning_rate": 1e-05, "loss": 0.5157, "mean_token_accuracy": 0.863526463508606, "num_tokens": 373777354.0, "step": 9843 }, { "epoch": 1.2522579824449815, "ewc_loss": 0.11474609375, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.274744510650635, "learning_rate": 1e-05, "loss": 0.5217, "mean_token_accuracy": 0.8590996265411377, "num_tokens": 373812400.0, "step": 9844 }, { "epoch": 1.252385192723572, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.995030403137207, "learning_rate": 1e-05, "loss": 0.4479, "mean_token_accuracy": 0.882015585899353, "num_tokens": 373846033.0, "step": 9845 }, { "epoch": 1.2525124030021626, "ewc_loss": 0.11376953125, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.248698711395264, "learning_rate": 1e-05, "loss": 0.4866, "mean_token_accuracy": 0.8753514885902405, "num_tokens": 373887888.0, "step": 9846 }, { "epoch": 1.2526396132807531, "ewc_loss": 0.10205078125, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.784879207611084, "learning_rate": 1e-05, "loss": 0.4544, "mean_token_accuracy": 0.8813271522521973, "num_tokens": 373928082.0, "step": 9847 }, { "epoch": 1.2527668235593437, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.274503707885742, "learning_rate": 1e-05, "loss": 0.5256, "mean_token_accuracy": 0.8578951358795166, "num_tokens": 373964078.0, "step": 9848 }, { "epoch": 1.2528940338379342, "ewc_loss": 0.10009765625, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.660538673400879, "learning_rate": 1e-05, "loss": 0.5319, "mean_token_accuracy": 0.8505246639251709, "num_tokens": 373998163.0, "step": 9849 }, { "epoch": 1.2530212441165247, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.356415748596191, "learning_rate": 1e-05, "loss": 0.4919, "mean_token_accuracy": 0.8663395643234253, "num_tokens": 374036322.0, "step": 9850 }, { "epoch": 1.2531484543951152, "ewc_loss": 0.0986328125, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.541408061981201, "learning_rate": 1e-05, "loss": 0.4939, "mean_token_accuracy": 0.863213300704956, "num_tokens": 374079144.0, "step": 9851 }, { "epoch": 1.2532756646737058, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.526124477386475, "learning_rate": 1e-05, "loss": 0.5065, "mean_token_accuracy": 0.86636883020401, "num_tokens": 374116200.0, "step": 9852 }, { "epoch": 1.253402874952296, "ewc_loss": 0.09814453125, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.470832347869873, "learning_rate": 1e-05, "loss": 0.5008, "mean_token_accuracy": 0.8622563481330872, "num_tokens": 374151009.0, "step": 9853 }, { "epoch": 1.2535300852308866, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.778552532196045, "learning_rate": 1e-05, "loss": 0.5798, "mean_token_accuracy": 0.8457413911819458, "num_tokens": 374187974.0, "step": 9854 }, { "epoch": 1.2536572955094771, "ewc_loss": 0.09912109375, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.47362756729126, "learning_rate": 1e-05, "loss": 0.5217, "mean_token_accuracy": 0.8567838668823242, "num_tokens": 374229203.0, "step": 9855 }, { "epoch": 1.2537845057880677, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 7.0620622634887695, "learning_rate": 1e-05, "loss": 0.47, "mean_token_accuracy": 0.8789196014404297, "num_tokens": 374261606.0, "step": 9856 }, { "epoch": 1.2539117160666582, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.662219524383545, "learning_rate": 1e-05, "loss": 0.5174, "mean_token_accuracy": 0.8610339164733887, "num_tokens": 374300274.0, "step": 9857 }, { "epoch": 1.2540389263452487, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 0.00011205673217773438, "grad_norm": 7.180696964263916, "learning_rate": 1e-05, "loss": 0.5847, "mean_token_accuracy": 0.8473694324493408, "num_tokens": 374336942.0, "step": 9858 }, { "epoch": 1.2541661366238392, "ewc_loss": 0.11376953125, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 5.974748134613037, "learning_rate": 1e-05, "loss": 0.5298, "mean_token_accuracy": 0.8583238124847412, "num_tokens": 374370213.0, "step": 9859 }, { "epoch": 1.2542933469024298, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 6.990057468414307, "learning_rate": 1e-05, "loss": 0.5373, "mean_token_accuracy": 0.859316885471344, "num_tokens": 374410653.0, "step": 9860 }, { "epoch": 1.2544205571810203, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.1481404304504395, "learning_rate": 1e-05, "loss": 0.539, "mean_token_accuracy": 0.854526162147522, "num_tokens": 374454449.0, "step": 9861 }, { "epoch": 1.2545477674596106, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.617694854736328, "learning_rate": 1e-05, "loss": 0.4815, "mean_token_accuracy": 0.8722886443138123, "num_tokens": 374497494.0, "step": 9862 }, { "epoch": 1.2546749777382011, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.181420803070068, "learning_rate": 1e-05, "loss": 0.4917, "mean_token_accuracy": 0.8727339506149292, "num_tokens": 374534739.0, "step": 9863 }, { "epoch": 1.2548021880167917, "ewc_loss": 0.11083984375, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.232942581176758, "learning_rate": 1e-05, "loss": 0.525, "mean_token_accuracy": 0.8619874715805054, "num_tokens": 374573609.0, "step": 9864 }, { "epoch": 1.2549293982953822, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.180636882781982, "learning_rate": 1e-05, "loss": 0.4876, "mean_token_accuracy": 0.874786913394928, "num_tokens": 374610269.0, "step": 9865 }, { "epoch": 1.2550566085739727, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.923196315765381, "learning_rate": 1e-05, "loss": 0.5107, "mean_token_accuracy": 0.8620269298553467, "num_tokens": 374644888.0, "step": 9866 }, { "epoch": 1.2551838188525632, "ewc_loss": 0.11083984375, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.194907188415527, "learning_rate": 1e-05, "loss": 0.4753, "mean_token_accuracy": 0.8699480295181274, "num_tokens": 374685845.0, "step": 9867 }, { "epoch": 1.2553110291311538, "ewc_loss": 0.10107421875, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.724488258361816, "learning_rate": 1e-05, "loss": 0.4733, "mean_token_accuracy": 0.8716073036193848, "num_tokens": 374723501.0, "step": 9868 }, { "epoch": 1.2554382394097443, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.252409934997559, "learning_rate": 1e-05, "loss": 0.5227, "mean_token_accuracy": 0.8630139827728271, "num_tokens": 374763082.0, "step": 9869 }, { "epoch": 1.2555654496883348, "ewc_loss": 0.0986328125, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.5782999992370605, "learning_rate": 1e-05, "loss": 0.471, "mean_token_accuracy": 0.8749567866325378, "num_tokens": 374802829.0, "step": 9870 }, { "epoch": 1.2556926599669254, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.391634941101074, "learning_rate": 1e-05, "loss": 0.4827, "mean_token_accuracy": 0.8748648166656494, "num_tokens": 374838802.0, "step": 9871 }, { "epoch": 1.255819870245516, "ewc_loss": 0.09814453125, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.463897228240967, "learning_rate": 1e-05, "loss": 0.424, "mean_token_accuracy": 0.8873844146728516, "num_tokens": 374876831.0, "step": 9872 }, { "epoch": 1.2559470805241064, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.634897708892822, "learning_rate": 1e-05, "loss": 0.4485, "mean_token_accuracy": 0.8847795724868774, "num_tokens": 374918120.0, "step": 9873 }, { "epoch": 1.256074290802697, "ewc_loss": 0.09716796875, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.392245769500732, "learning_rate": 1e-05, "loss": 0.4851, "mean_token_accuracy": 0.8622620701789856, "num_tokens": 374958247.0, "step": 9874 }, { "epoch": 1.2562015010812875, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.972050189971924, "learning_rate": 1e-05, "loss": 0.5361, "mean_token_accuracy": 0.8620898127555847, "num_tokens": 374998461.0, "step": 9875 }, { "epoch": 1.256328711359878, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.512904644012451, "learning_rate": 1e-05, "loss": 0.4597, "mean_token_accuracy": 0.8759968280792236, "num_tokens": 375029813.0, "step": 9876 }, { "epoch": 1.2564559216384683, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 0.00011444091796875, "grad_norm": 7.238377094268799, "learning_rate": 1e-05, "loss": 0.5083, "mean_token_accuracy": 0.8721534013748169, "num_tokens": 375070724.0, "step": 9877 }, { "epoch": 1.2565831319170588, "ewc_loss": 0.11083984375, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 5.847231388092041, "learning_rate": 1e-05, "loss": 0.4971, "mean_token_accuracy": 0.8659581542015076, "num_tokens": 375110634.0, "step": 9878 }, { "epoch": 1.2567103421956494, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 0.00011396408081054688, "grad_norm": 7.204770088195801, "learning_rate": 1e-05, "loss": 0.4968, "mean_token_accuracy": 0.8756324052810669, "num_tokens": 375150054.0, "step": 9879 }, { "epoch": 1.25683755247424, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.176295280456543, "learning_rate": 1e-05, "loss": 0.5493, "mean_token_accuracy": 0.8567043542861938, "num_tokens": 375185406.0, "step": 9880 }, { "epoch": 1.2569647627528304, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.853556156158447, "learning_rate": 1e-05, "loss": 0.4934, "mean_token_accuracy": 0.874879002571106, "num_tokens": 375227544.0, "step": 9881 }, { "epoch": 1.257091973031421, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.243797779083252, "learning_rate": 1e-05, "loss": 0.4781, "mean_token_accuracy": 0.874269962310791, "num_tokens": 375263523.0, "step": 9882 }, { "epoch": 1.2572191833100115, "ewc_loss": 0.11474609375, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.438693046569824, "learning_rate": 1e-05, "loss": 0.4814, "mean_token_accuracy": 0.8753422498703003, "num_tokens": 375304764.0, "step": 9883 }, { "epoch": 1.257346393588602, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.22420072555542, "learning_rate": 1e-05, "loss": 0.484, "mean_token_accuracy": 0.8726624250411987, "num_tokens": 375340989.0, "step": 9884 }, { "epoch": 1.2574736038671925, "ewc_loss": 0.10498046875, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 6.077630043029785, "learning_rate": 1e-05, "loss": 0.4973, "mean_token_accuracy": 0.8659756779670715, "num_tokens": 375374987.0, "step": 9885 }, { "epoch": 1.2576008141457828, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.214252948760986, "learning_rate": 1e-05, "loss": 0.5177, "mean_token_accuracy": 0.8613088726997375, "num_tokens": 375415847.0, "step": 9886 }, { "epoch": 1.2577280244243734, "ewc_loss": 0.10205078125, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.818045139312744, "learning_rate": 1e-05, "loss": 0.4472, "mean_token_accuracy": 0.8812447786331177, "num_tokens": 375455761.0, "step": 9887 }, { "epoch": 1.257855234702964, "ewc_loss": 0.11083984375, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.239855766296387, "learning_rate": 1e-05, "loss": 0.4443, "mean_token_accuracy": 0.8853766918182373, "num_tokens": 375493774.0, "step": 9888 }, { "epoch": 1.2579824449815544, "ewc_loss": 0.099609375, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.6596174240112305, "learning_rate": 1e-05, "loss": 0.5171, "mean_token_accuracy": 0.8619129061698914, "num_tokens": 375531947.0, "step": 9889 }, { "epoch": 1.258109655260145, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.322943687438965, "learning_rate": 1e-05, "loss": 0.4673, "mean_token_accuracy": 0.8778452277183533, "num_tokens": 375573546.0, "step": 9890 }, { "epoch": 1.2582368655387355, "ewc_loss": 0.0986328125, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.522804260253906, "learning_rate": 1e-05, "loss": 0.5507, "mean_token_accuracy": 0.846123218536377, "num_tokens": 375617351.0, "step": 9891 }, { "epoch": 1.258364075817326, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.507967472076416, "learning_rate": 1e-05, "loss": 0.5055, "mean_token_accuracy": 0.8661035895347595, "num_tokens": 375653939.0, "step": 9892 }, { "epoch": 1.2584912860959165, "ewc_loss": 0.09814453125, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.417654037475586, "learning_rate": 1e-05, "loss": 0.5044, "mean_token_accuracy": 0.8597919940948486, "num_tokens": 375695345.0, "step": 9893 }, { "epoch": 1.258618496374507, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.802870273590088, "learning_rate": 1e-05, "loss": 0.4483, "mean_token_accuracy": 0.8871976733207703, "num_tokens": 375731786.0, "step": 9894 }, { "epoch": 1.2587457066530976, "ewc_loss": 0.09912109375, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.42115592956543, "learning_rate": 1e-05, "loss": 0.4769, "mean_token_accuracy": 0.8678560853004456, "num_tokens": 375773258.0, "step": 9895 }, { "epoch": 1.2588729169316881, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 0.00011205673217773438, "grad_norm": 7.149420738220215, "learning_rate": 1e-05, "loss": 0.4688, "mean_token_accuracy": 0.8813758492469788, "num_tokens": 375805712.0, "step": 9896 }, { "epoch": 1.2590001272102787, "ewc_loss": 0.10693359375, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 5.689420223236084, "learning_rate": 1e-05, "loss": 0.4882, "mean_token_accuracy": 0.8661360144615173, "num_tokens": 375835927.0, "step": 9897 }, { "epoch": 1.2591273374888692, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 0.00011491775512695312, "grad_norm": 7.26691198348999, "learning_rate": 1e-05, "loss": 0.4954, "mean_token_accuracy": 0.8779304623603821, "num_tokens": 375870568.0, "step": 9898 }, { "epoch": 1.2592545477674597, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.062854766845703, "learning_rate": 1e-05, "loss": 0.503, "mean_token_accuracy": 0.8674256205558777, "num_tokens": 375904941.0, "step": 9899 }, { "epoch": 1.2593817580460502, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 7.022563934326172, "learning_rate": 1e-05, "loss": 0.5583, "mean_token_accuracy": 0.8518579006195068, "num_tokens": 375941193.0, "step": 9900 }, { "epoch": 1.2595089683246408, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.238956451416016, "learning_rate": 1e-05, "loss": 0.4892, "mean_token_accuracy": 0.8735027313232422, "num_tokens": 375977952.0, "step": 9901 }, { "epoch": 1.259636178603231, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.604733943939209, "learning_rate": 1e-05, "loss": 0.5291, "mean_token_accuracy": 0.8574098348617554, "num_tokens": 376016711.0, "step": 9902 }, { "epoch": 1.2597633888818216, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.228522777557373, "learning_rate": 1e-05, "loss": 0.5358, "mean_token_accuracy": 0.8573874235153198, "num_tokens": 376061753.0, "step": 9903 }, { "epoch": 1.2598905991604121, "ewc_loss": 0.11083984375, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.250447750091553, "learning_rate": 1e-05, "loss": 0.5314, "mean_token_accuracy": 0.8588119745254517, "num_tokens": 376100148.0, "step": 9904 }, { "epoch": 1.2600178094390027, "ewc_loss": 0.11474609375, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.188558101654053, "learning_rate": 1e-05, "loss": 0.4847, "mean_token_accuracy": 0.8729655742645264, "num_tokens": 376140663.0, "step": 9905 }, { "epoch": 1.2601450197175932, "ewc_loss": 0.10302734375, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.95009708404541, "learning_rate": 1e-05, "loss": 0.5534, "mean_token_accuracy": 0.8501463532447815, "num_tokens": 376186702.0, "step": 9906 }, { "epoch": 1.2602722299961837, "ewc_loss": 0.11083984375, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.172791481018066, "learning_rate": 1e-05, "loss": 0.5047, "mean_token_accuracy": 0.8695827722549438, "num_tokens": 376222736.0, "step": 9907 }, { "epoch": 1.2603994402747742, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.777428150177002, "learning_rate": 1e-05, "loss": 0.5281, "mean_token_accuracy": 0.8549463748931885, "num_tokens": 376258871.0, "step": 9908 }, { "epoch": 1.2605266505533648, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.195119380950928, "learning_rate": 1e-05, "loss": 0.5117, "mean_token_accuracy": 0.8657206296920776, "num_tokens": 376295486.0, "step": 9909 }, { "epoch": 1.2606538608319553, "ewc_loss": 0.10009765625, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.638714790344238, "learning_rate": 1e-05, "loss": 0.4654, "mean_token_accuracy": 0.8727024793624878, "num_tokens": 376333138.0, "step": 9910 }, { "epoch": 1.2607810711105456, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.281850337982178, "learning_rate": 1e-05, "loss": 0.4588, "mean_token_accuracy": 0.8775774240493774, "num_tokens": 376367284.0, "step": 9911 }, { "epoch": 1.2609082813891361, "ewc_loss": 0.0986328125, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.522932529449463, "learning_rate": 1e-05, "loss": 0.5168, "mean_token_accuracy": 0.8573481440544128, "num_tokens": 376398318.0, "step": 9912 }, { "epoch": 1.2610354916677267, "ewc_loss": 0.11474609375, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.473752498626709, "learning_rate": 1e-05, "loss": 0.4738, "mean_token_accuracy": 0.8770550489425659, "num_tokens": 376430599.0, "step": 9913 }, { "epoch": 1.2611627019463172, "ewc_loss": 0.09765625, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.384824275970459, "learning_rate": 1e-05, "loss": 0.4646, "mean_token_accuracy": 0.8709796667098999, "num_tokens": 376467238.0, "step": 9914 }, { "epoch": 1.2612899122249077, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.801934242248535, "learning_rate": 1e-05, "loss": 0.5216, "mean_token_accuracy": 0.8624516725540161, "num_tokens": 376505728.0, "step": 9915 }, { "epoch": 1.2614171225034982, "ewc_loss": 0.0986328125, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.365403652191162, "learning_rate": 1e-05, "loss": 0.5454, "mean_token_accuracy": 0.8488190174102783, "num_tokens": 376550387.0, "step": 9916 }, { "epoch": 1.2615443327820888, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 0.00011444091796875, "grad_norm": 7.214799880981445, "learning_rate": 1e-05, "loss": 0.463, "mean_token_accuracy": 0.8854238986968994, "num_tokens": 376583365.0, "step": 9917 }, { "epoch": 1.2616715430606793, "ewc_loss": 0.10693359375, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 5.645176410675049, "learning_rate": 1e-05, "loss": 0.4617, "mean_token_accuracy": 0.8760885000228882, "num_tokens": 376620625.0, "step": 9918 }, { "epoch": 1.2617987533392698, "ewc_loss": 0.140625, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 0.00012159347534179688, "grad_norm": 7.407312393188477, "learning_rate": 1e-05, "loss": 0.5124, "mean_token_accuracy": 0.8711627125740051, "num_tokens": 376657048.0, "step": 9919 }, { "epoch": 1.2619259636178604, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.130505561828613, "learning_rate": 1e-05, "loss": 0.5264, "mean_token_accuracy": 0.8620643615722656, "num_tokens": 376697000.0, "step": 9920 }, { "epoch": 1.2620531738964509, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 0.00011205673217773438, "grad_norm": 7.134188175201416, "learning_rate": 1e-05, "loss": 0.5177, "mean_token_accuracy": 0.8679061532020569, "num_tokens": 376739306.0, "step": 9921 }, { "epoch": 1.2621803841750414, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.34356689453125, "learning_rate": 1e-05, "loss": 0.4888, "mean_token_accuracy": 0.8738290071487427, "num_tokens": 376777136.0, "step": 9922 }, { "epoch": 1.262307594453632, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.689676284790039, "learning_rate": 1e-05, "loss": 0.4719, "mean_token_accuracy": 0.8776928186416626, "num_tokens": 376823875.0, "step": 9923 }, { "epoch": 1.2624348047322225, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.305788516998291, "learning_rate": 1e-05, "loss": 0.5086, "mean_token_accuracy": 0.8657057285308838, "num_tokens": 376861335.0, "step": 9924 }, { "epoch": 1.262562015010813, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.3001909255981445, "learning_rate": 1e-05, "loss": 0.5425, "mean_token_accuracy": 0.8549354076385498, "num_tokens": 376896076.0, "step": 9925 }, { "epoch": 1.2626892252894033, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.232868671417236, "learning_rate": 1e-05, "loss": 0.5248, "mean_token_accuracy": 0.8601875901222229, "num_tokens": 376943761.0, "step": 9926 }, { "epoch": 1.2628164355679938, "ewc_loss": 0.10400390625, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.998051166534424, "learning_rate": 1e-05, "loss": 0.5063, "mean_token_accuracy": 0.8626307845115662, "num_tokens": 376983526.0, "step": 9927 }, { "epoch": 1.2629436458465844, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.1965227127075195, "learning_rate": 1e-05, "loss": 0.4513, "mean_token_accuracy": 0.881758987903595, "num_tokens": 377025293.0, "step": 9928 }, { "epoch": 1.263070856125175, "ewc_loss": 0.10205078125, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.796430587768555, "learning_rate": 1e-05, "loss": 0.4818, "mean_token_accuracy": 0.8672695159912109, "num_tokens": 377071380.0, "step": 9929 }, { "epoch": 1.2631980664037654, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.208267688751221, "learning_rate": 1e-05, "loss": 0.5276, "mean_token_accuracy": 0.8582602143287659, "num_tokens": 377110215.0, "step": 9930 }, { "epoch": 1.263325276682356, "ewc_loss": 0.10009765625, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.6669721603393555, "learning_rate": 1e-05, "loss": 0.4538, "mean_token_accuracy": 0.878447949886322, "num_tokens": 377150191.0, "step": 9931 }, { "epoch": 1.2634524869609465, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.295359134674072, "learning_rate": 1e-05, "loss": 0.5639, "mean_token_accuracy": 0.8474150896072388, "num_tokens": 377190636.0, "step": 9932 }, { "epoch": 1.263579697239537, "ewc_loss": 0.0986328125, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.53381872177124, "learning_rate": 1e-05, "loss": 0.4813, "mean_token_accuracy": 0.8676905632019043, "num_tokens": 377230314.0, "step": 9933 }, { "epoch": 1.2637069075181275, "ewc_loss": 0.11474609375, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.4816460609436035, "learning_rate": 1e-05, "loss": 0.5215, "mean_token_accuracy": 0.8621295094490051, "num_tokens": 377270830.0, "step": 9934 }, { "epoch": 1.2638341177967178, "ewc_loss": 0.0966796875, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.409441947937012, "learning_rate": 1e-05, "loss": 0.5426, "mean_token_accuracy": 0.8496354222297668, "num_tokens": 377313862.0, "step": 9935 }, { "epoch": 1.2639613280753084, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.809257507324219, "learning_rate": 1e-05, "loss": 0.5148, "mean_token_accuracy": 0.8655834197998047, "num_tokens": 377355451.0, "step": 9936 }, { "epoch": 1.264088538353899, "ewc_loss": 0.09765625, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.3821282386779785, "learning_rate": 1e-05, "loss": 0.4937, "mean_token_accuracy": 0.8629588484764099, "num_tokens": 377397608.0, "step": 9937 }, { "epoch": 1.2642157486324894, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 0.00011491775512695312, "grad_norm": 7.206353187561035, "learning_rate": 1e-05, "loss": 0.5172, "mean_token_accuracy": 0.8694562911987305, "num_tokens": 377436268.0, "step": 9938 }, { "epoch": 1.26434295891108, "ewc_loss": 0.10693359375, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 5.66981840133667, "learning_rate": 1e-05, "loss": 0.4984, "mean_token_accuracy": 0.8662281036376953, "num_tokens": 377470896.0, "step": 9939 }, { "epoch": 1.2644701691896705, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 0.00011873245239257812, "grad_norm": 7.3763909339904785, "learning_rate": 1e-05, "loss": 0.5045, "mean_token_accuracy": 0.8738555908203125, "num_tokens": 377508828.0, "step": 9940 }, { "epoch": 1.264597379468261, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.131955623626709, "learning_rate": 1e-05, "loss": 0.5402, "mean_token_accuracy": 0.8532799482345581, "num_tokens": 377542007.0, "step": 9941 }, { "epoch": 1.2647245897468515, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 7.101746559143066, "learning_rate": 1e-05, "loss": 0.5314, "mean_token_accuracy": 0.8636676073074341, "num_tokens": 377582671.0, "step": 9942 }, { "epoch": 1.264851800025442, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.311348915100098, "learning_rate": 1e-05, "loss": 0.4925, "mean_token_accuracy": 0.8705533742904663, "num_tokens": 377624840.0, "step": 9943 }, { "epoch": 1.2649790103040326, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.6699419021606445, "learning_rate": 1e-05, "loss": 0.5065, "mean_token_accuracy": 0.8657711744308472, "num_tokens": 377661955.0, "step": 9944 }, { "epoch": 1.2651062205826231, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.287290573120117, "learning_rate": 1e-05, "loss": 0.4845, "mean_token_accuracy": 0.8736255168914795, "num_tokens": 377698574.0, "step": 9945 }, { "epoch": 1.2652334308612136, "ewc_loss": 0.11083984375, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.261216640472412, "learning_rate": 1e-05, "loss": 0.5214, "mean_token_accuracy": 0.8591291308403015, "num_tokens": 377739874.0, "step": 9946 }, { "epoch": 1.2653606411398042, "ewc_loss": 0.11376953125, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.24017858505249, "learning_rate": 1e-05, "loss": 0.5598, "mean_token_accuracy": 0.8497511148452759, "num_tokens": 377775644.0, "step": 9947 }, { "epoch": 1.2654878514183947, "ewc_loss": 0.10400390625, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.960443019866943, "learning_rate": 1e-05, "loss": 0.5222, "mean_token_accuracy": 0.861515998840332, "num_tokens": 377812648.0, "step": 9948 }, { "epoch": 1.2656150616969852, "ewc_loss": 0.11181640625, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.227989673614502, "learning_rate": 1e-05, "loss": 0.537, "mean_token_accuracy": 0.8559408187866211, "num_tokens": 377849928.0, "step": 9949 }, { "epoch": 1.2657422719755758, "ewc_loss": 0.10107421875, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.748907566070557, "learning_rate": 1e-05, "loss": 0.4686, "mean_token_accuracy": 0.8733144402503967, "num_tokens": 377891154.0, "step": 9950 }, { "epoch": 1.265869482254166, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.275062561035156, "learning_rate": 1e-05, "loss": 0.5149, "mean_token_accuracy": 0.8645994663238525, "num_tokens": 377926858.0, "step": 9951 }, { "epoch": 1.2659966925327566, "ewc_loss": 0.099609375, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.600528717041016, "learning_rate": 1e-05, "loss": 0.4628, "mean_token_accuracy": 0.8748154640197754, "num_tokens": 377976895.0, "step": 9952 }, { "epoch": 1.2661239028113471, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.405460834503174, "learning_rate": 1e-05, "loss": 0.4924, "mean_token_accuracy": 0.8680084943771362, "num_tokens": 378010300.0, "step": 9953 }, { "epoch": 1.2662511130899377, "ewc_loss": 0.09814453125, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.4853515625, "learning_rate": 1e-05, "loss": 0.4834, "mean_token_accuracy": 0.8670752644538879, "num_tokens": 378051764.0, "step": 9954 }, { "epoch": 1.2663783233685282, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.65060567855835, "learning_rate": 1e-05, "loss": 0.5118, "mean_token_accuracy": 0.8612855672836304, "num_tokens": 378087976.0, "step": 9955 }, { "epoch": 1.2665055336471187, "ewc_loss": 0.09716796875, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.407909870147705, "learning_rate": 1e-05, "loss": 0.4729, "mean_token_accuracy": 0.8690342903137207, "num_tokens": 378124062.0, "step": 9956 }, { "epoch": 1.2666327439257092, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.988173484802246, "learning_rate": 1e-05, "loss": 0.5, "mean_token_accuracy": 0.8657313585281372, "num_tokens": 378161248.0, "step": 9957 }, { "epoch": 1.2667599542042998, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.515904426574707, "learning_rate": 1e-05, "loss": 0.4986, "mean_token_accuracy": 0.8597502708435059, "num_tokens": 378197959.0, "step": 9958 }, { "epoch": 1.2668871644828903, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 0.00011730194091796875, "grad_norm": 7.277710437774658, "learning_rate": 1e-05, "loss": 0.5288, "mean_token_accuracy": 0.8615432381629944, "num_tokens": 378233030.0, "step": 9959 }, { "epoch": 1.2670143747614806, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 5.900050163269043, "learning_rate": 1e-05, "loss": 0.5318, "mean_token_accuracy": 0.860645055770874, "num_tokens": 378265147.0, "step": 9960 }, { "epoch": 1.2671415850400711, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 0.00011301040649414062, "grad_norm": 7.20143985748291, "learning_rate": 1e-05, "loss": 0.5212, "mean_token_accuracy": 0.8623507022857666, "num_tokens": 378298759.0, "step": 9961 }, { "epoch": 1.2672687953186617, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.193953514099121, "learning_rate": 1e-05, "loss": 0.5485, "mean_token_accuracy": 0.8505118489265442, "num_tokens": 378338456.0, "step": 9962 }, { "epoch": 1.2673960055972522, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.840211391448975, "learning_rate": 1e-05, "loss": 0.5134, "mean_token_accuracy": 0.870059072971344, "num_tokens": 378374755.0, "step": 9963 }, { "epoch": 1.2675232158758427, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.257468223571777, "learning_rate": 1e-05, "loss": 0.4535, "mean_token_accuracy": 0.88325035572052, "num_tokens": 378415302.0, "step": 9964 }, { "epoch": 1.2676504261544332, "ewc_loss": 0.11474609375, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.42838191986084, "learning_rate": 1e-05, "loss": 0.5035, "mean_token_accuracy": 0.8634746074676514, "num_tokens": 378454400.0, "step": 9965 }, { "epoch": 1.2677776364330238, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.225955486297607, "learning_rate": 1e-05, "loss": 0.4929, "mean_token_accuracy": 0.8694936037063599, "num_tokens": 378496986.0, "step": 9966 }, { "epoch": 1.2679048467116143, "ewc_loss": 0.10595703125, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 6.0835700035095215, "learning_rate": 1e-05, "loss": 0.4659, "mean_token_accuracy": 0.8750475645065308, "num_tokens": 378533937.0, "step": 9967 }, { "epoch": 1.2680320569902048, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.206750392913818, "learning_rate": 1e-05, "loss": 0.4725, "mean_token_accuracy": 0.8762577772140503, "num_tokens": 378567420.0, "step": 9968 }, { "epoch": 1.2681592672687954, "ewc_loss": 0.10205078125, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.8549604415893555, "learning_rate": 1e-05, "loss": 0.4764, "mean_token_accuracy": 0.8736333847045898, "num_tokens": 378600939.0, "step": 9969 }, { "epoch": 1.2682864775473859, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.191162109375, "learning_rate": 1e-05, "loss": 0.52, "mean_token_accuracy": 0.8604631423950195, "num_tokens": 378647858.0, "step": 9970 }, { "epoch": 1.2684136878259764, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.700870990753174, "learning_rate": 1e-05, "loss": 0.4827, "mean_token_accuracy": 0.8653337359428406, "num_tokens": 378686695.0, "step": 9971 }, { "epoch": 1.268540898104567, "ewc_loss": 0.11083984375, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.2413859367370605, "learning_rate": 1e-05, "loss": 0.4846, "mean_token_accuracy": 0.8725608587265015, "num_tokens": 378732583.0, "step": 9972 }, { "epoch": 1.2686681083831575, "ewc_loss": 0.09912109375, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.579299449920654, "learning_rate": 1e-05, "loss": 0.4535, "mean_token_accuracy": 0.8777070045471191, "num_tokens": 378768134.0, "step": 9973 }, { "epoch": 1.268795318661748, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.363759994506836, "learning_rate": 1e-05, "loss": 0.4541, "mean_token_accuracy": 0.8823665380477905, "num_tokens": 378799684.0, "step": 9974 }, { "epoch": 1.2689225289403383, "ewc_loss": 0.0986328125, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.463173866271973, "learning_rate": 1e-05, "loss": 0.4983, "mean_token_accuracy": 0.8617468476295471, "num_tokens": 378837831.0, "step": 9975 }, { "epoch": 1.2690497392189288, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.6091766357421875, "learning_rate": 1e-05, "loss": 0.5, "mean_token_accuracy": 0.8672462105751038, "num_tokens": 378876697.0, "step": 9976 }, { "epoch": 1.2691769494975194, "ewc_loss": 0.09716796875, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.374083995819092, "learning_rate": 1e-05, "loss": 0.4498, "mean_token_accuracy": 0.8803557753562927, "num_tokens": 378916301.0, "step": 9977 }, { "epoch": 1.2693041597761099, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.969631671905518, "learning_rate": 1e-05, "loss": 0.5325, "mean_token_accuracy": 0.8608031868934631, "num_tokens": 378956294.0, "step": 9978 }, { "epoch": 1.2694313700547004, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.472318172454834, "learning_rate": 1e-05, "loss": 0.5119, "mean_token_accuracy": 0.8600559234619141, "num_tokens": 378992704.0, "step": 9979 }, { "epoch": 1.269558580333291, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 0.000118255615234375, "grad_norm": 7.297354698181152, "learning_rate": 1e-05, "loss": 0.5025, "mean_token_accuracy": 0.8748797178268433, "num_tokens": 379034955.0, "step": 9980 }, { "epoch": 1.2696857906118815, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 5.842893123626709, "learning_rate": 1e-05, "loss": 0.4623, "mean_token_accuracy": 0.8787997364997864, "num_tokens": 379075026.0, "step": 9981 }, { "epoch": 1.269813000890472, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 0.0001163482666015625, "grad_norm": 7.281360149383545, "learning_rate": 1e-05, "loss": 0.5221, "mean_token_accuracy": 0.8660452365875244, "num_tokens": 379113452.0, "step": 9982 }, { "epoch": 1.2699402111690625, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.222610950469971, "learning_rate": 1e-05, "loss": 0.5798, "mean_token_accuracy": 0.8429459929466248, "num_tokens": 379155632.0, "step": 9983 }, { "epoch": 1.2700674214476528, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.925740718841553, "learning_rate": 1e-05, "loss": 0.4914, "mean_token_accuracy": 0.8743693828582764, "num_tokens": 379188987.0, "step": 9984 }, { "epoch": 1.2701946317262434, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.30070686340332, "learning_rate": 1e-05, "loss": 0.4837, "mean_token_accuracy": 0.8739280700683594, "num_tokens": 379229099.0, "step": 9985 }, { "epoch": 1.270321842004834, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.487548351287842, "learning_rate": 1e-05, "loss": 0.472, "mean_token_accuracy": 0.8793125152587891, "num_tokens": 379265365.0, "step": 9986 }, { "epoch": 1.2704490522834244, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.268042087554932, "learning_rate": 1e-05, "loss": 0.4879, "mean_token_accuracy": 0.8754153847694397, "num_tokens": 379301170.0, "step": 9987 }, { "epoch": 1.270576262562015, "ewc_loss": 0.10693359375, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 6.109412670135498, "learning_rate": 1e-05, "loss": 0.4864, "mean_token_accuracy": 0.8645354509353638, "num_tokens": 379334983.0, "step": 9988 }, { "epoch": 1.2707034728406055, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.245826721191406, "learning_rate": 1e-05, "loss": 0.5223, "mean_token_accuracy": 0.8577570915222168, "num_tokens": 379371584.0, "step": 9989 }, { "epoch": 1.270830683119196, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.845345973968506, "learning_rate": 1e-05, "loss": 0.4882, "mean_token_accuracy": 0.8689329624176025, "num_tokens": 379415363.0, "step": 9990 }, { "epoch": 1.2709578933977865, "ewc_loss": 0.11083984375, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.2614336013793945, "learning_rate": 1e-05, "loss": 0.4561, "mean_token_accuracy": 0.8796455264091492, "num_tokens": 379447019.0, "step": 9991 }, { "epoch": 1.271085103676377, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.677630424499512, "learning_rate": 1e-05, "loss": 0.4624, "mean_token_accuracy": 0.8752667903900146, "num_tokens": 379490229.0, "step": 9992 }, { "epoch": 1.2712123139549676, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.351314544677734, "learning_rate": 1e-05, "loss": 0.4906, "mean_token_accuracy": 0.867742657661438, "num_tokens": 379528608.0, "step": 9993 }, { "epoch": 1.2713395242335581, "ewc_loss": 0.0986328125, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.540281772613525, "learning_rate": 1e-05, "loss": 0.5207, "mean_token_accuracy": 0.8597849607467651, "num_tokens": 379566916.0, "step": 9994 }, { "epoch": 1.2714667345121486, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.5248637199401855, "learning_rate": 1e-05, "loss": 0.4512, "mean_token_accuracy": 0.8820583820343018, "num_tokens": 379600648.0, "step": 9995 }, { "epoch": 1.2715939447907392, "ewc_loss": 0.09716796875, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.439733982086182, "learning_rate": 1e-05, "loss": 0.4984, "mean_token_accuracy": 0.8621047735214233, "num_tokens": 379639519.0, "step": 9996 }, { "epoch": 1.2717211550693297, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.815618515014648, "learning_rate": 1e-05, "loss": 0.4852, "mean_token_accuracy": 0.8757857084274292, "num_tokens": 379674426.0, "step": 9997 }, { "epoch": 1.2718483653479202, "ewc_loss": 0.10009765625, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.4385504722595215, "learning_rate": 1e-05, "loss": 0.4728, "mean_token_accuracy": 0.8758703470230103, "num_tokens": 379713121.0, "step": 9998 }, { "epoch": 1.2719755756265108, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 0.00011348724365234375, "grad_norm": 7.147001266479492, "learning_rate": 1e-05, "loss": 0.4864, "mean_token_accuracy": 0.8754087686538696, "num_tokens": 379749952.0, "step": 9999 }, { "epoch": 1.272102785905101, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 5.684389591217041, "learning_rate": 1e-05, "loss": 0.5405, "mean_token_accuracy": 0.858323335647583, "num_tokens": 379786057.0, "step": 10000 }, { "epoch": 1.2722299961836916, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 0.0001163482666015625, "grad_norm": 7.285366535186768, "learning_rate": 1e-05, "loss": 0.5463, "mean_token_accuracy": 0.8596875071525574, "num_tokens": 379820034.0, "step": 10001 }, { "epoch": 1.2723572064622821, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.067479133605957, "learning_rate": 1e-05, "loss": 0.4965, "mean_token_accuracy": 0.8723759651184082, "num_tokens": 379863969.0, "step": 10002 }, { "epoch": 1.2724844167408726, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 7.031928062438965, "learning_rate": 1e-05, "loss": 0.4848, "mean_token_accuracy": 0.8761321306228638, "num_tokens": 379903360.0, "step": 10003 }, { "epoch": 1.2726116270194632, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.252943515777588, "learning_rate": 1e-05, "loss": 0.4995, "mean_token_accuracy": 0.868970513343811, "num_tokens": 379935874.0, "step": 10004 }, { "epoch": 1.2727388372980537, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.629641532897949, "learning_rate": 1e-05, "loss": 0.5038, "mean_token_accuracy": 0.8629617691040039, "num_tokens": 379972525.0, "step": 10005 }, { "epoch": 1.2728660475766442, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.229147911071777, "learning_rate": 1e-05, "loss": 0.4918, "mean_token_accuracy": 0.8711963295936584, "num_tokens": 380012925.0, "step": 10006 }, { "epoch": 1.2729932578552348, "ewc_loss": 0.11083984375, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.257118225097656, "learning_rate": 1e-05, "loss": 0.5316, "mean_token_accuracy": 0.8569852113723755, "num_tokens": 380048235.0, "step": 10007 }, { "epoch": 1.2731204681338253, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.2016401290893555, "learning_rate": 1e-05, "loss": 0.569, "mean_token_accuracy": 0.8472754955291748, "num_tokens": 380084607.0, "step": 10008 }, { "epoch": 1.2732476784124156, "ewc_loss": 0.10302734375, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.962297439575195, "learning_rate": 1e-05, "loss": 0.5219, "mean_token_accuracy": 0.8563673496246338, "num_tokens": 380121514.0, "step": 10009 }, { "epoch": 1.2733748886910061, "ewc_loss": 0.11083984375, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.182659149169922, "learning_rate": 1e-05, "loss": 0.5027, "mean_token_accuracy": 0.8623079061508179, "num_tokens": 380158640.0, "step": 10010 }, { "epoch": 1.2735020989695967, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.775301933288574, "learning_rate": 1e-05, "loss": 0.4863, "mean_token_accuracy": 0.8669910430908203, "num_tokens": 380196195.0, "step": 10011 }, { "epoch": 1.2736293092481872, "ewc_loss": 0.109375, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.198540687561035, "learning_rate": 1e-05, "loss": 0.5037, "mean_token_accuracy": 0.8654858469963074, "num_tokens": 380234777.0, "step": 10012 }, { "epoch": 1.2737565195267777, "ewc_loss": 0.10009765625, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.648263454437256, "learning_rate": 1e-05, "loss": 0.5075, "mean_token_accuracy": 0.8616592884063721, "num_tokens": 380276494.0, "step": 10013 }, { "epoch": 1.2738837298053682, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.287813663482666, "learning_rate": 1e-05, "loss": 0.5049, "mean_token_accuracy": 0.8646365404129028, "num_tokens": 380314791.0, "step": 10014 }, { "epoch": 1.2740109400839588, "ewc_loss": 0.0986328125, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.521643161773682, "learning_rate": 1e-05, "loss": 0.5157, "mean_token_accuracy": 0.8643207550048828, "num_tokens": 380354372.0, "step": 10015 }, { "epoch": 1.2741381503625493, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.482547760009766, "learning_rate": 1e-05, "loss": 0.4606, "mean_token_accuracy": 0.8795604109764099, "num_tokens": 380388022.0, "step": 10016 }, { "epoch": 1.2742653606411398, "ewc_loss": 0.0966796875, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.400065898895264, "learning_rate": 1e-05, "loss": 0.4993, "mean_token_accuracy": 0.8606032133102417, "num_tokens": 380426665.0, "step": 10017 }, { "epoch": 1.2743925709197303, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.8102192878723145, "learning_rate": 1e-05, "loss": 0.4748, "mean_token_accuracy": 0.8764088153839111, "num_tokens": 380466068.0, "step": 10018 }, { "epoch": 1.2745197811983209, "ewc_loss": 0.09765625, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.375337600708008, "learning_rate": 1e-05, "loss": 0.4349, "mean_token_accuracy": 0.883047878742218, "num_tokens": 380500370.0, "step": 10019 }, { "epoch": 1.2746469914769114, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 0.00011587142944335938, "grad_norm": 7.218991756439209, "learning_rate": 1e-05, "loss": 0.5194, "mean_token_accuracy": 0.8694957494735718, "num_tokens": 380538886.0, "step": 10020 }, { "epoch": 1.274774201755502, "ewc_loss": 0.10693359375, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 5.661788463592529, "learning_rate": 1e-05, "loss": 0.4942, "mean_token_accuracy": 0.8699011206626892, "num_tokens": 380577425.0, "step": 10021 }, { "epoch": 1.2749014120340925, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 0.00011968612670898438, "grad_norm": 7.40352201461792, "learning_rate": 1e-05, "loss": 0.5612, "mean_token_accuracy": 0.854033887386322, "num_tokens": 380616999.0, "step": 10022 }, { "epoch": 1.275028622312683, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.1313371658325195, "learning_rate": 1e-05, "loss": 0.5222, "mean_token_accuracy": 0.865109384059906, "num_tokens": 380654065.0, "step": 10023 }, { "epoch": 1.2751558325912733, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 0.00011157989501953125, "grad_norm": 7.13039493560791, "learning_rate": 1e-05, "loss": 0.5276, "mean_token_accuracy": 0.8614597916603088, "num_tokens": 380687947.0, "step": 10024 }, { "epoch": 1.2752830428698638, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.33831787109375, "learning_rate": 1e-05, "loss": 0.4861, "mean_token_accuracy": 0.8762688636779785, "num_tokens": 380728387.0, "step": 10025 }, { "epoch": 1.2754102531484544, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.6847052574157715, "learning_rate": 1e-05, "loss": 0.5247, "mean_token_accuracy": 0.8622007369995117, "num_tokens": 380763906.0, "step": 10026 }, { "epoch": 1.2755374634270449, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.305390357971191, "learning_rate": 1e-05, "loss": 0.5351, "mean_token_accuracy": 0.8561674356460571, "num_tokens": 380806904.0, "step": 10027 }, { "epoch": 1.2756646737056354, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.271235466003418, "learning_rate": 1e-05, "loss": 0.4642, "mean_token_accuracy": 0.877298891544342, "num_tokens": 380842694.0, "step": 10028 }, { "epoch": 1.275791883984226, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.239859104156494, "learning_rate": 1e-05, "loss": 0.4759, "mean_token_accuracy": 0.8788020610809326, "num_tokens": 380879624.0, "step": 10029 }, { "epoch": 1.2759190942628165, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.97588586807251, "learning_rate": 1e-05, "loss": 0.4855, "mean_token_accuracy": 0.8709880113601685, "num_tokens": 380916358.0, "step": 10030 }, { "epoch": 1.276046304541407, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.221497058868408, "learning_rate": 1e-05, "loss": 0.4691, "mean_token_accuracy": 0.8783589601516724, "num_tokens": 380957858.0, "step": 10031 }, { "epoch": 1.2761735148199975, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.768561840057373, "learning_rate": 1e-05, "loss": 0.4595, "mean_token_accuracy": 0.8749831914901733, "num_tokens": 380997301.0, "step": 10032 }, { "epoch": 1.2763007250985878, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.26625919342041, "learning_rate": 1e-05, "loss": 0.4644, "mean_token_accuracy": 0.8752549886703491, "num_tokens": 381034023.0, "step": 10033 }, { "epoch": 1.2764279353771784, "ewc_loss": 0.10009765625, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.639896392822266, "learning_rate": 1e-05, "loss": 0.5293, "mean_token_accuracy": 0.8528741002082825, "num_tokens": 381073402.0, "step": 10034 }, { "epoch": 1.2765551456557689, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.3592095375061035, "learning_rate": 1e-05, "loss": 0.4783, "mean_token_accuracy": 0.8742795586585999, "num_tokens": 381111182.0, "step": 10035 }, { "epoch": 1.2766823559343594, "ewc_loss": 0.0986328125, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.503836154937744, "learning_rate": 1e-05, "loss": 0.4724, "mean_token_accuracy": 0.8690595030784607, "num_tokens": 381153527.0, "step": 10036 }, { "epoch": 1.27680956621295, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.575024604797363, "learning_rate": 1e-05, "loss": 0.5123, "mean_token_accuracy": 0.8670414686203003, "num_tokens": 381196337.0, "step": 10037 }, { "epoch": 1.2769367764915405, "ewc_loss": 0.09716796875, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.403531551361084, "learning_rate": 1e-05, "loss": 0.4845, "mean_token_accuracy": 0.8650981187820435, "num_tokens": 381239065.0, "step": 10038 }, { "epoch": 1.277063986770131, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.917309284210205, "learning_rate": 1e-05, "loss": 0.4552, "mean_token_accuracy": 0.8853849172592163, "num_tokens": 381274826.0, "step": 10039 }, { "epoch": 1.2771911970487215, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.460257530212402, "learning_rate": 1e-05, "loss": 0.483, "mean_token_accuracy": 0.8675816655158997, "num_tokens": 381318043.0, "step": 10040 }, { "epoch": 1.277318407327312, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 0.00011587142944335938, "grad_norm": 7.265544414520264, "learning_rate": 1e-05, "loss": 0.5295, "mean_token_accuracy": 0.8671728372573853, "num_tokens": 381352583.0, "step": 10041 }, { "epoch": 1.2774456176059026, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 5.799649715423584, "learning_rate": 1e-05, "loss": 0.5195, "mean_token_accuracy": 0.8602718710899353, "num_tokens": 381387271.0, "step": 10042 }, { "epoch": 1.2775728278844931, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 0.00011539459228515625, "grad_norm": 7.291517734527588, "learning_rate": 1e-05, "loss": 0.5454, "mean_token_accuracy": 0.8617352247238159, "num_tokens": 381429749.0, "step": 10043 }, { "epoch": 1.2777000381630836, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.1732378005981445, "learning_rate": 1e-05, "loss": 0.4949, "mean_token_accuracy": 0.879956841468811, "num_tokens": 381468763.0, "step": 10044 }, { "epoch": 1.2778272484416742, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.959924697875977, "learning_rate": 1e-05, "loss": 0.4948, "mean_token_accuracy": 0.8716048002243042, "num_tokens": 381505789.0, "step": 10045 }, { "epoch": 1.2779544587202647, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.284365177154541, "learning_rate": 1e-05, "loss": 0.4276, "mean_token_accuracy": 0.8906078934669495, "num_tokens": 381545861.0, "step": 10046 }, { "epoch": 1.2780816689988552, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.5504913330078125, "learning_rate": 1e-05, "loss": 0.5685, "mean_token_accuracy": 0.8509966731071472, "num_tokens": 381585373.0, "step": 10047 }, { "epoch": 1.2782088792774458, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.253226280212402, "learning_rate": 1e-05, "loss": 0.49, "mean_token_accuracy": 0.8727782368659973, "num_tokens": 381623029.0, "step": 10048 }, { "epoch": 1.278336089556036, "ewc_loss": 0.10791015625, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.16609001159668, "learning_rate": 1e-05, "loss": 0.5316, "mean_token_accuracy": 0.8592756390571594, "num_tokens": 381663481.0, "step": 10049 }, { "epoch": 1.2784632998346266, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.219440460205078, "learning_rate": 1e-05, "loss": 0.4449, "mean_token_accuracy": 0.8881338238716125, "num_tokens": 381703977.0, "step": 10050 }, { "epoch": 1.2785905101132171, "ewc_loss": 0.10205078125, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.899764060974121, "learning_rate": 1e-05, "loss": 0.5033, "mean_token_accuracy": 0.8632243871688843, "num_tokens": 381741409.0, "step": 10051 }, { "epoch": 1.2787177203918076, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.230468273162842, "learning_rate": 1e-05, "loss": 0.5143, "mean_token_accuracy": 0.8659263253211975, "num_tokens": 381779013.0, "step": 10052 }, { "epoch": 1.2788449306703982, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.726232528686523, "learning_rate": 1e-05, "loss": 0.4904, "mean_token_accuracy": 0.8649874925613403, "num_tokens": 381813460.0, "step": 10053 }, { "epoch": 1.2789721409489887, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.271499156951904, "learning_rate": 1e-05, "loss": 0.5091, "mean_token_accuracy": 0.8628107309341431, "num_tokens": 381853821.0, "step": 10054 }, { "epoch": 1.2790993512275792, "ewc_loss": 0.0986328125, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.592424392700195, "learning_rate": 1e-05, "loss": 0.5124, "mean_token_accuracy": 0.860572874546051, "num_tokens": 381895411.0, "step": 10055 }, { "epoch": 1.2792265615061698, "ewc_loss": 0.11376953125, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.402678966522217, "learning_rate": 1e-05, "loss": 0.5172, "mean_token_accuracy": 0.8610237240791321, "num_tokens": 381929067.0, "step": 10056 }, { "epoch": 1.2793537717847603, "ewc_loss": 0.09814453125, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.474118232727051, "learning_rate": 1e-05, "loss": 0.4728, "mean_token_accuracy": 0.8702310919761658, "num_tokens": 381964002.0, "step": 10057 }, { "epoch": 1.2794809820633506, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.640054225921631, "learning_rate": 1e-05, "loss": 0.4457, "mean_token_accuracy": 0.8829249143600464, "num_tokens": 382002913.0, "step": 10058 }, { "epoch": 1.2796081923419411, "ewc_loss": 0.09716796875, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.390077590942383, "learning_rate": 1e-05, "loss": 0.4613, "mean_token_accuracy": 0.8778135776519775, "num_tokens": 382039026.0, "step": 10059 }, { "epoch": 1.2797354026205316, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 7.011264801025391, "learning_rate": 1e-05, "loss": 0.509, "mean_token_accuracy": 0.8694517612457275, "num_tokens": 382077413.0, "step": 10060 }, { "epoch": 1.2798626128991222, "ewc_loss": 0.10205078125, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.5047197341918945, "learning_rate": 1e-05, "loss": 0.5123, "mean_token_accuracy": 0.8612584471702576, "num_tokens": 382112686.0, "step": 10061 }, { "epoch": 1.2799898231777127, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 0.00011777877807617188, "grad_norm": 7.312011241912842, "learning_rate": 1e-05, "loss": 0.5341, "mean_token_accuracy": 0.8647133111953735, "num_tokens": 382152184.0, "step": 10062 }, { "epoch": 1.2801170334563032, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 5.900664806365967, "learning_rate": 1e-05, "loss": 0.5574, "mean_token_accuracy": 0.8597551584243774, "num_tokens": 382190916.0, "step": 10063 }, { "epoch": 1.2802442437348938, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 0.00011396408081054688, "grad_norm": 7.25747013092041, "learning_rate": 1e-05, "loss": 0.5587, "mean_token_accuracy": 0.8558955788612366, "num_tokens": 382222986.0, "step": 10064 }, { "epoch": 1.2803714540134843, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.232007026672363, "learning_rate": 1e-05, "loss": 0.4762, "mean_token_accuracy": 0.8779242634773254, "num_tokens": 382266039.0, "step": 10065 }, { "epoch": 1.2804986642920748, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.875308513641357, "learning_rate": 1e-05, "loss": 0.4879, "mean_token_accuracy": 0.8717160224914551, "num_tokens": 382302071.0, "step": 10066 }, { "epoch": 1.2806258745706653, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.3004841804504395, "learning_rate": 1e-05, "loss": 0.5048, "mean_token_accuracy": 0.8680646419525146, "num_tokens": 382340240.0, "step": 10067 }, { "epoch": 1.2807530848492559, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.454652786254883, "learning_rate": 1e-05, "loss": 0.475, "mean_token_accuracy": 0.8713849782943726, "num_tokens": 382368308.0, "step": 10068 }, { "epoch": 1.2808802951278464, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.25147008895874, "learning_rate": 1e-05, "loss": 0.4893, "mean_token_accuracy": 0.8736808896064758, "num_tokens": 382406365.0, "step": 10069 }, { "epoch": 1.281007505406437, "ewc_loss": 0.10693359375, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 6.081356525421143, "learning_rate": 1e-05, "loss": 0.5249, "mean_token_accuracy": 0.8580065369606018, "num_tokens": 382449005.0, "step": 10070 }, { "epoch": 1.2811347156850275, "ewc_loss": 0.11376953125, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.2437286376953125, "learning_rate": 1e-05, "loss": 0.5198, "mean_token_accuracy": 0.8624471426010132, "num_tokens": 382487744.0, "step": 10071 }, { "epoch": 1.281261925963618, "ewc_loss": 0.10205078125, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.835874080657959, "learning_rate": 1e-05, "loss": 0.4883, "mean_token_accuracy": 0.8665140271186829, "num_tokens": 382526954.0, "step": 10072 }, { "epoch": 1.2813891362422083, "ewc_loss": 0.11181640625, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.253537654876709, "learning_rate": 1e-05, "loss": 0.4651, "mean_token_accuracy": 0.8780683875083923, "num_tokens": 382565257.0, "step": 10073 }, { "epoch": 1.2815163465207988, "ewc_loss": 0.10107421875, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.684311389923096, "learning_rate": 1e-05, "loss": 0.4979, "mean_token_accuracy": 0.8648292422294617, "num_tokens": 382601992.0, "step": 10074 }, { "epoch": 1.2816435567993893, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.3293914794921875, "learning_rate": 1e-05, "loss": 0.5022, "mean_token_accuracy": 0.8652399778366089, "num_tokens": 382637181.0, "step": 10075 }, { "epoch": 1.2817707670779799, "ewc_loss": 0.0986328125, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.544213771820068, "learning_rate": 1e-05, "loss": 0.518, "mean_token_accuracy": 0.8556770086288452, "num_tokens": 382679886.0, "step": 10076 }, { "epoch": 1.2818979773565704, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.510060787200928, "learning_rate": 1e-05, "loss": 0.5244, "mean_token_accuracy": 0.8626085519790649, "num_tokens": 382717062.0, "step": 10077 }, { "epoch": 1.282025187635161, "ewc_loss": 0.09765625, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.438892841339111, "learning_rate": 1e-05, "loss": 0.4862, "mean_token_accuracy": 0.8644667267799377, "num_tokens": 382752188.0, "step": 10078 }, { "epoch": 1.2821523979137515, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.797723770141602, "learning_rate": 1e-05, "loss": 0.4992, "mean_token_accuracy": 0.8664216995239258, "num_tokens": 382785229.0, "step": 10079 }, { "epoch": 1.282279608192342, "ewc_loss": 0.09912109375, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.429835796356201, "learning_rate": 1e-05, "loss": 0.4934, "mean_token_accuracy": 0.8645285367965698, "num_tokens": 382823432.0, "step": 10080 }, { "epoch": 1.2824068184709325, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 0.00011301040649414062, "grad_norm": 7.138101100921631, "learning_rate": 1e-05, "loss": 0.5397, "mean_token_accuracy": 0.8589895963668823, "num_tokens": 382862763.0, "step": 10081 }, { "epoch": 1.2825340287495228, "ewc_loss": 0.10595703125, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.6696672439575195, "learning_rate": 1e-05, "loss": 0.487, "mean_token_accuracy": 0.8686697483062744, "num_tokens": 382898312.0, "step": 10082 }, { "epoch": 1.2826612390281134, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 0.0001163482666015625, "grad_norm": 7.290205001831055, "learning_rate": 1e-05, "loss": 0.5111, "mean_token_accuracy": 0.8676445484161377, "num_tokens": 382933494.0, "step": 10083 }, { "epoch": 1.2827884493067039, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.053580284118652, "learning_rate": 1e-05, "loss": 0.4633, "mean_token_accuracy": 0.880255937576294, "num_tokens": 382971029.0, "step": 10084 }, { "epoch": 1.2829156595852944, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 7.075355052947998, "learning_rate": 1e-05, "loss": 0.536, "mean_token_accuracy": 0.8602042198181152, "num_tokens": 383003911.0, "step": 10085 }, { "epoch": 1.283042869863885, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.254312992095947, "learning_rate": 1e-05, "loss": 0.5365, "mean_token_accuracy": 0.8617653846740723, "num_tokens": 383042945.0, "step": 10086 }, { "epoch": 1.2831700801424755, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.668183326721191, "learning_rate": 1e-05, "loss": 0.4897, "mean_token_accuracy": 0.8717167377471924, "num_tokens": 383085944.0, "step": 10087 }, { "epoch": 1.283297290421066, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.2570390701293945, "learning_rate": 1e-05, "loss": 0.531, "mean_token_accuracy": 0.8625301718711853, "num_tokens": 383124854.0, "step": 10088 }, { "epoch": 1.2834245006996565, "ewc_loss": 0.11083984375, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.2769646644592285, "learning_rate": 1e-05, "loss": 0.475, "mean_token_accuracy": 0.873123288154602, "num_tokens": 383159427.0, "step": 10089 }, { "epoch": 1.283551710978247, "ewc_loss": 0.11376953125, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.220699310302734, "learning_rate": 1e-05, "loss": 0.4782, "mean_token_accuracy": 0.8704239130020142, "num_tokens": 383199320.0, "step": 10090 }, { "epoch": 1.2836789212568376, "ewc_loss": 0.10498046875, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.981342315673828, "learning_rate": 1e-05, "loss": 0.4392, "mean_token_accuracy": 0.8803495168685913, "num_tokens": 383231367.0, "step": 10091 }, { "epoch": 1.283806131535428, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.208376407623291, "learning_rate": 1e-05, "loss": 0.4797, "mean_token_accuracy": 0.8707572221755981, "num_tokens": 383268238.0, "step": 10092 }, { "epoch": 1.2839333418140186, "ewc_loss": 0.10205078125, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.789829254150391, "learning_rate": 1e-05, "loss": 0.4673, "mean_token_accuracy": 0.874744176864624, "num_tokens": 383303445.0, "step": 10093 }, { "epoch": 1.2840605520926092, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.230008125305176, "learning_rate": 1e-05, "loss": 0.5491, "mean_token_accuracy": 0.8505856394767761, "num_tokens": 383341106.0, "step": 10094 }, { "epoch": 1.2841877623711997, "ewc_loss": 0.099609375, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.648611545562744, "learning_rate": 1e-05, "loss": 0.5227, "mean_token_accuracy": 0.8605577349662781, "num_tokens": 383379225.0, "step": 10095 }, { "epoch": 1.2843149726497902, "ewc_loss": 0.11181640625, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.2987542152404785, "learning_rate": 1e-05, "loss": 0.4911, "mean_token_accuracy": 0.8719415664672852, "num_tokens": 383420798.0, "step": 10096 }, { "epoch": 1.2844421829283807, "ewc_loss": 0.0986328125, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.528290271759033, "learning_rate": 1e-05, "loss": 0.5275, "mean_token_accuracy": 0.8541315793991089, "num_tokens": 383467145.0, "step": 10097 }, { "epoch": 1.284569393206971, "ewc_loss": 0.11474609375, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.481150150299072, "learning_rate": 1e-05, "loss": 0.5396, "mean_token_accuracy": 0.8547016382217407, "num_tokens": 383505143.0, "step": 10098 }, { "epoch": 1.2846966034855616, "ewc_loss": 0.09814453125, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.423046588897705, "learning_rate": 1e-05, "loss": 0.5016, "mean_token_accuracy": 0.8602584600448608, "num_tokens": 383545987.0, "step": 10099 }, { "epoch": 1.2848238137641521, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.785930633544922, "learning_rate": 1e-05, "loss": 0.4974, "mean_token_accuracy": 0.871492862701416, "num_tokens": 383578546.0, "step": 10100 }, { "epoch": 1.2849510240427426, "ewc_loss": 0.09814453125, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.396674633026123, "learning_rate": 1e-05, "loss": 0.4722, "mean_token_accuracy": 0.8705841302871704, "num_tokens": 383619603.0, "step": 10101 }, { "epoch": 1.2850782343213332, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 0.00011348724365234375, "grad_norm": 7.154904842376709, "learning_rate": 1e-05, "loss": 0.5521, "mean_token_accuracy": 0.8559074401855469, "num_tokens": 383660716.0, "step": 10102 }, { "epoch": 1.2852054445999237, "ewc_loss": 0.10595703125, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.6349310874938965, "learning_rate": 1e-05, "loss": 0.475, "mean_token_accuracy": 0.8724251389503479, "num_tokens": 383704344.0, "step": 10103 }, { "epoch": 1.2853326548785142, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 0.00011873245239257812, "grad_norm": 7.352354526519775, "learning_rate": 1e-05, "loss": 0.5622, "mean_token_accuracy": 0.8575826287269592, "num_tokens": 383738141.0, "step": 10104 }, { "epoch": 1.2854598651571048, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.080443382263184, "learning_rate": 1e-05, "loss": 0.5342, "mean_token_accuracy": 0.857214093208313, "num_tokens": 383774682.0, "step": 10105 }, { "epoch": 1.2855870754356953, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 0.00010967254638671875, "grad_norm": 7.1187520027160645, "learning_rate": 1e-05, "loss": 0.4911, "mean_token_accuracy": 0.8763077855110168, "num_tokens": 383814056.0, "step": 10106 }, { "epoch": 1.2857142857142856, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.296814441680908, "learning_rate": 1e-05, "loss": 0.537, "mean_token_accuracy": 0.8562678098678589, "num_tokens": 383858151.0, "step": 10107 }, { "epoch": 1.2858414959928761, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.691320896148682, "learning_rate": 1e-05, "loss": 0.5052, "mean_token_accuracy": 0.8658080697059631, "num_tokens": 383895121.0, "step": 10108 }, { "epoch": 1.2859687062714666, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.290552616119385, "learning_rate": 1e-05, "loss": 0.5562, "mean_token_accuracy": 0.8534811735153198, "num_tokens": 383929273.0, "step": 10109 }, { "epoch": 1.2860959165500572, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.00019073486328125, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.276762008666992, "learning_rate": 1e-05, "loss": 0.4901, "mean_token_accuracy": 0.868276059627533, "num_tokens": 383970374.0, "step": 10110 }, { "epoch": 1.2862231268286477, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.00019073486328125, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.2418036460876465, "learning_rate": 1e-05, "loss": 0.5041, "mean_token_accuracy": 0.8672754168510437, "num_tokens": 384010375.0, "step": 10111 }, { "epoch": 1.2863503371072382, "ewc_loss": 0.10400390625, "ewc_loss_diag": 0.00019073486328125, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.973250389099121, "learning_rate": 1e-05, "loss": 0.5181, "mean_token_accuracy": 0.8600549101829529, "num_tokens": 384040547.0, "step": 10112 }, { "epoch": 1.2864775473858288, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.2465105056762695, "learning_rate": 1e-05, "loss": 0.513, "mean_token_accuracy": 0.8609930276870728, "num_tokens": 384076005.0, "step": 10113 }, { "epoch": 1.2866047576644193, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.759766101837158, "learning_rate": 1e-05, "loss": 0.4551, "mean_token_accuracy": 0.8768885731697083, "num_tokens": 384112568.0, "step": 10114 }, { "epoch": 1.2867319679430098, "ewc_loss": 0.11181640625, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.278168201446533, "learning_rate": 1e-05, "loss": 0.4803, "mean_token_accuracy": 0.8761850595474243, "num_tokens": 384151942.0, "step": 10115 }, { "epoch": 1.2868591782216003, "ewc_loss": 0.09912109375, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.610082149505615, "learning_rate": 1e-05, "loss": 0.4832, "mean_token_accuracy": 0.8664383888244629, "num_tokens": 384192534.0, "step": 10116 }, { "epoch": 1.2869863885001909, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.399961948394775, "learning_rate": 1e-05, "loss": 0.4819, "mean_token_accuracy": 0.8734608888626099, "num_tokens": 384230600.0, "step": 10117 }, { "epoch": 1.2871135987787814, "ewc_loss": 0.0986328125, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.49139404296875, "learning_rate": 1e-05, "loss": 0.425, "mean_token_accuracy": 0.8866860866546631, "num_tokens": 384266566.0, "step": 10118 }, { "epoch": 1.287240809057372, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.663268089294434, "learning_rate": 1e-05, "loss": 0.5268, "mean_token_accuracy": 0.8600724935531616, "num_tokens": 384304536.0, "step": 10119 }, { "epoch": 1.2873680193359625, "ewc_loss": 0.09814453125, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.418237686157227, "learning_rate": 1e-05, "loss": 0.4826, "mean_token_accuracy": 0.8660929799079895, "num_tokens": 384347865.0, "step": 10120 }, { "epoch": 1.287495229614553, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 6.998404026031494, "learning_rate": 1e-05, "loss": 0.5208, "mean_token_accuracy": 0.8634089827537537, "num_tokens": 384380060.0, "step": 10121 }, { "epoch": 1.2876224398931433, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.531158447265625, "learning_rate": 1e-05, "loss": 0.5398, "mean_token_accuracy": 0.8522511124610901, "num_tokens": 384417574.0, "step": 10122 }, { "epoch": 1.2877496501717338, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 0.0001163482666015625, "grad_norm": 7.282780647277832, "learning_rate": 1e-05, "loss": 0.5119, "mean_token_accuracy": 0.872727632522583, "num_tokens": 384454591.0, "step": 10123 }, { "epoch": 1.2878768604503243, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 5.88607931137085, "learning_rate": 1e-05, "loss": 0.4915, "mean_token_accuracy": 0.8686329126358032, "num_tokens": 384489642.0, "step": 10124 }, { "epoch": 1.2880040707289149, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.00018978118896484375, "ewc_loss_parallel": 0.00011348724365234375, "grad_norm": 7.231194496154785, "learning_rate": 1e-05, "loss": 0.5272, "mean_token_accuracy": 0.864680290222168, "num_tokens": 384528993.0, "step": 10125 }, { "epoch": 1.2881312810075054, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.00019073486328125, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.212757110595703, "learning_rate": 1e-05, "loss": 0.5124, "mean_token_accuracy": 0.8642395734786987, "num_tokens": 384569321.0, "step": 10126 }, { "epoch": 1.288258491286096, "ewc_loss": 0.125, "ewc_loss_diag": 0.00019073486328125, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.8788228034973145, "learning_rate": 1e-05, "loss": 0.5328, "mean_token_accuracy": 0.8625056147575378, "num_tokens": 384614821.0, "step": 10127 }, { "epoch": 1.2883857015646865, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.00019073486328125, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.287318706512451, "learning_rate": 1e-05, "loss": 0.4838, "mean_token_accuracy": 0.8774769306182861, "num_tokens": 384658557.0, "step": 10128 }, { "epoch": 1.288512911843277, "ewc_loss": 0.11474609375, "ewc_loss_diag": 0.00019073486328125, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.465806007385254, "learning_rate": 1e-05, "loss": 0.5462, "mean_token_accuracy": 0.8554107546806335, "num_tokens": 384697121.0, "step": 10129 }, { "epoch": 1.2886401221218675, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.00019073486328125, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.247149467468262, "learning_rate": 1e-05, "loss": 0.5595, "mean_token_accuracy": 0.8505174517631531, "num_tokens": 384735410.0, "step": 10130 }, { "epoch": 1.2887673324004578, "ewc_loss": 0.10693359375, "ewc_loss_diag": 0.00019073486328125, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 6.112424850463867, "learning_rate": 1e-05, "loss": 0.5046, "mean_token_accuracy": 0.8653357028961182, "num_tokens": 384769394.0, "step": 10131 }, { "epoch": 1.2888945426790483, "ewc_loss": 0.11376953125, "ewc_loss_diag": 0.00019073486328125, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.2107696533203125, "learning_rate": 1e-05, "loss": 0.4955, "mean_token_accuracy": 0.8686767816543579, "num_tokens": 384808845.0, "step": 10132 }, { "epoch": 1.2890217529576389, "ewc_loss": 0.10205078125, "ewc_loss_diag": 0.00019073486328125, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.8611273765563965, "learning_rate": 1e-05, "loss": 0.4887, "mean_token_accuracy": 0.8675037622451782, "num_tokens": 384848129.0, "step": 10133 }, { "epoch": 1.2891489632362294, "ewc_loss": 0.11083984375, "ewc_loss_diag": 0.00019073486328125, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.206963062286377, "learning_rate": 1e-05, "loss": 0.4527, "mean_token_accuracy": 0.8826340436935425, "num_tokens": 384885181.0, "step": 10134 }, { "epoch": 1.28927617351482, "ewc_loss": 0.10107421875, "ewc_loss_diag": 0.00019073486328125, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.7253031730651855, "learning_rate": 1e-05, "loss": 0.4531, "mean_token_accuracy": 0.8762032389640808, "num_tokens": 384923499.0, "step": 10135 }, { "epoch": 1.2894033837934105, "ewc_loss": 0.11083984375, "ewc_loss_diag": 0.00019073486328125, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.250385284423828, "learning_rate": 1e-05, "loss": 0.4592, "mean_token_accuracy": 0.8783327341079712, "num_tokens": 384957367.0, "step": 10136 }, { "epoch": 1.289530594072001, "ewc_loss": 0.09912109375, "ewc_loss_diag": 0.00019073486328125, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.594126224517822, "learning_rate": 1e-05, "loss": 0.4725, "mean_token_accuracy": 0.8727425932884216, "num_tokens": 385000409.0, "step": 10137 }, { "epoch": 1.2896578043505915, "ewc_loss": 0.11376953125, "ewc_loss_diag": 0.00019073486328125, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.3773298263549805, "learning_rate": 1e-05, "loss": 0.5095, "mean_token_accuracy": 0.8627961874008179, "num_tokens": 385038439.0, "step": 10138 }, { "epoch": 1.289785014629182, "ewc_loss": 0.0986328125, "ewc_loss_diag": 0.00019073486328125, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.473221778869629, "learning_rate": 1e-05, "loss": 0.4796, "mean_token_accuracy": 0.8699508309364319, "num_tokens": 385073950.0, "step": 10139 }, { "epoch": 1.2899122249077726, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.00019073486328125, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.638272762298584, "learning_rate": 1e-05, "loss": 0.5053, "mean_token_accuracy": 0.8601173162460327, "num_tokens": 385112394.0, "step": 10140 }, { "epoch": 1.290039435186363, "ewc_loss": 0.09716796875, "ewc_loss_diag": 0.00019073486328125, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.385565280914307, "learning_rate": 1e-05, "loss": 0.4954, "mean_token_accuracy": 0.8650377988815308, "num_tokens": 385149096.0, "step": 10141 }, { "epoch": 1.2901666454649536, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.00019073486328125, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 7.004748344421387, "learning_rate": 1e-05, "loss": 0.5202, "mean_token_accuracy": 0.8620371222496033, "num_tokens": 385181986.0, "step": 10142 }, { "epoch": 1.2902938557435442, "ewc_loss": 0.10107421875, "ewc_loss_diag": 0.00019073486328125, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.487069129943848, "learning_rate": 1e-05, "loss": 0.5761, "mean_token_accuracy": 0.8402101397514343, "num_tokens": 385220730.0, "step": 10143 }, { "epoch": 1.2904210660221347, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.00019073486328125, "ewc_loss_parallel": 0.00011873245239257812, "grad_norm": 7.327383518218994, "learning_rate": 1e-05, "loss": 0.5821, "mean_token_accuracy": 0.8477848768234253, "num_tokens": 385262250.0, "step": 10144 }, { "epoch": 1.2905482763007252, "ewc_loss": 0.11376953125, "ewc_loss_diag": 0.00019168853759765625, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 5.877373695373535, "learning_rate": 1e-05, "loss": 0.5179, "mean_token_accuracy": 0.8604207038879395, "num_tokens": 385303518.0, "step": 10145 }, { "epoch": 1.2906754865793157, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.00019168853759765625, "ewc_loss_parallel": 0.0001163482666015625, "grad_norm": 7.311613082885742, "learning_rate": 1e-05, "loss": 0.5298, "mean_token_accuracy": 0.8648930788040161, "num_tokens": 385342303.0, "step": 10146 }, { "epoch": 1.290802696857906, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.00019168853759765625, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.246458053588867, "learning_rate": 1e-05, "loss": 0.4878, "mean_token_accuracy": 0.870944619178772, "num_tokens": 385381581.0, "step": 10147 }, { "epoch": 1.2909299071364966, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.00019168853759765625, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.943005084991455, "learning_rate": 1e-05, "loss": 0.4856, "mean_token_accuracy": 0.8757240772247314, "num_tokens": 385414835.0, "step": 10148 }, { "epoch": 1.291057117415087, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.00019168853759765625, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.331263065338135, "learning_rate": 1e-05, "loss": 0.5078, "mean_token_accuracy": 0.8660487532615662, "num_tokens": 385454022.0, "step": 10149 }, { "epoch": 1.2911843276936776, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.00019168853759765625, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.50283670425415, "learning_rate": 1e-05, "loss": 0.5389, "mean_token_accuracy": 0.8565495014190674, "num_tokens": 385486479.0, "step": 10150 }, { "epoch": 1.2913115379722682, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.00019168853759765625, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.282003879547119, "learning_rate": 1e-05, "loss": 0.5296, "mean_token_accuracy": 0.8576582670211792, "num_tokens": 385527942.0, "step": 10151 }, { "epoch": 1.2914387482508587, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.00019168853759765625, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 6.12506103515625, "learning_rate": 1e-05, "loss": 0.5028, "mean_token_accuracy": 0.8651516437530518, "num_tokens": 385567994.0, "step": 10152 }, { "epoch": 1.2915659585294492, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.00019168853759765625, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.249058246612549, "learning_rate": 1e-05, "loss": 0.4727, "mean_token_accuracy": 0.8730437755584717, "num_tokens": 385603908.0, "step": 10153 }, { "epoch": 1.2916931688080397, "ewc_loss": 0.10205078125, "ewc_loss_diag": 0.00019168853759765625, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.877774238586426, "learning_rate": 1e-05, "loss": 0.5005, "mean_token_accuracy": 0.8663125038146973, "num_tokens": 385641461.0, "step": 10154 }, { "epoch": 1.2918203790866303, "ewc_loss": 0.11181640625, "ewc_loss_diag": 0.00019168853759765625, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.249027252197266, "learning_rate": 1e-05, "loss": 0.4797, "mean_token_accuracy": 0.8745090365409851, "num_tokens": 385676677.0, "step": 10155 }, { "epoch": 1.2919475893652206, "ewc_loss": 0.10107421875, "ewc_loss_diag": 0.00019168853759765625, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.719804763793945, "learning_rate": 1e-05, "loss": 0.4954, "mean_token_accuracy": 0.8667834401130676, "num_tokens": 385715674.0, "step": 10156 }, { "epoch": 1.2920747996438111, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.00019168853759765625, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.3158392906188965, "learning_rate": 1e-05, "loss": 0.5079, "mean_token_accuracy": 0.8631131649017334, "num_tokens": 385747947.0, "step": 10157 }, { "epoch": 1.2922020099224016, "ewc_loss": 0.09814453125, "ewc_loss_diag": 0.00019168853759765625, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.580427646636963, "learning_rate": 1e-05, "loss": 0.5024, "mean_token_accuracy": 0.8585640788078308, "num_tokens": 385787395.0, "step": 10158 }, { "epoch": 1.2923292202009922, "ewc_loss": 0.11474609375, "ewc_loss_diag": 0.00019168853759765625, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.456637382507324, "learning_rate": 1e-05, "loss": 0.471, "mean_token_accuracy": 0.8783879280090332, "num_tokens": 385821724.0, "step": 10159 }, { "epoch": 1.2924564304795827, "ewc_loss": 0.09814453125, "ewc_loss_diag": 0.00019168853759765625, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.455264091491699, "learning_rate": 1e-05, "loss": 0.4265, "mean_token_accuracy": 0.888149082660675, "num_tokens": 385856281.0, "step": 10160 }, { "epoch": 1.2925836407581732, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.00019168853759765625, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.739721775054932, "learning_rate": 1e-05, "loss": 0.526, "mean_token_accuracy": 0.8616927862167358, "num_tokens": 385895979.0, "step": 10161 }, { "epoch": 1.2927108510367638, "ewc_loss": 0.09814453125, "ewc_loss_diag": 0.00019168853759765625, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.424261569976807, "learning_rate": 1e-05, "loss": 0.4591, "mean_token_accuracy": 0.8762834072113037, "num_tokens": 385933134.0, "step": 10162 }, { "epoch": 1.2928380613153543, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.00019168853759765625, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 7.075767517089844, "learning_rate": 1e-05, "loss": 0.523, "mean_token_accuracy": 0.8636187314987183, "num_tokens": 385974487.0, "step": 10163 }, { "epoch": 1.2929652715939448, "ewc_loss": 0.10400390625, "ewc_loss_diag": 0.00019168853759765625, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.599478721618652, "learning_rate": 1e-05, "loss": 0.5337, "mean_token_accuracy": 0.8576160669326782, "num_tokens": 386014594.0, "step": 10164 }, { "epoch": 1.2930924818725353, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.00019168853759765625, "ewc_loss_parallel": 0.00011730194091796875, "grad_norm": 7.3108649253845215, "learning_rate": 1e-05, "loss": 0.5562, "mean_token_accuracy": 0.8602123260498047, "num_tokens": 386044982.0, "step": 10165 }, { "epoch": 1.2932196921511259, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.00019168853759765625, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 5.990823745727539, "learning_rate": 1e-05, "loss": 0.4906, "mean_token_accuracy": 0.8751540184020996, "num_tokens": 386084517.0, "step": 10166 }, { "epoch": 1.2933469024297164, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.00019168853759765625, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 7.1536054611206055, "learning_rate": 1e-05, "loss": 0.5126, "mean_token_accuracy": 0.8672565817832947, "num_tokens": 386121841.0, "step": 10167 }, { "epoch": 1.293474112708307, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.00019168853759765625, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.249837398529053, "learning_rate": 1e-05, "loss": 0.4892, "mean_token_accuracy": 0.8693400025367737, "num_tokens": 386157426.0, "step": 10168 }, { "epoch": 1.2936013229868975, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.00019168853759765625, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.745537281036377, "learning_rate": 1e-05, "loss": 0.5191, "mean_token_accuracy": 0.8619852662086487, "num_tokens": 386196739.0, "step": 10169 }, { "epoch": 1.293728533265488, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.00019168853759765625, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.2646355628967285, "learning_rate": 1e-05, "loss": 0.4542, "mean_token_accuracy": 0.8828409910202026, "num_tokens": 386234335.0, "step": 10170 }, { "epoch": 1.2938557435440783, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.00019168853759765625, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.340251445770264, "learning_rate": 1e-05, "loss": 0.5055, "mean_token_accuracy": 0.8658885955810547, "num_tokens": 386274166.0, "step": 10171 }, { "epoch": 1.2939829538226688, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.00019168853759765625, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.234618663787842, "learning_rate": 1e-05, "loss": 0.5317, "mean_token_accuracy": 0.8571089506149292, "num_tokens": 386313709.0, "step": 10172 }, { "epoch": 1.2941101641012593, "ewc_loss": 0.10595703125, "ewc_loss_diag": 0.00019168853759765625, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 6.007090091705322, "learning_rate": 1e-05, "loss": 0.537, "mean_token_accuracy": 0.8493785858154297, "num_tokens": 386345516.0, "step": 10173 }, { "epoch": 1.2942373743798499, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.00019168853759765625, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.2357707023620605, "learning_rate": 1e-05, "loss": 0.4539, "mean_token_accuracy": 0.881395697593689, "num_tokens": 386383284.0, "step": 10174 }, { "epoch": 1.2943645846584404, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.00019168853759765625, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.777344226837158, "learning_rate": 1e-05, "loss": 0.511, "mean_token_accuracy": 0.860368013381958, "num_tokens": 386423141.0, "step": 10175 }, { "epoch": 1.294491794937031, "ewc_loss": 0.11083984375, "ewc_loss_diag": 0.00019168853759765625, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.269207954406738, "learning_rate": 1e-05, "loss": 0.475, "mean_token_accuracy": 0.8734267950057983, "num_tokens": 386461888.0, "step": 10176 }, { "epoch": 1.2946190052156215, "ewc_loss": 0.10009765625, "ewc_loss_diag": 0.00019168853759765625, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.634247303009033, "learning_rate": 1e-05, "loss": 0.4948, "mean_token_accuracy": 0.8638318777084351, "num_tokens": 386500061.0, "step": 10177 }, { "epoch": 1.294746215494212, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.00019168853759765625, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.367315769195557, "learning_rate": 1e-05, "loss": 0.44, "mean_token_accuracy": 0.8868108987808228, "num_tokens": 386538843.0, "step": 10178 }, { "epoch": 1.2948734257728025, "ewc_loss": 0.09912109375, "ewc_loss_diag": 0.00019168853759765625, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.5103936195373535, "learning_rate": 1e-05, "loss": 0.484, "mean_token_accuracy": 0.8668217658996582, "num_tokens": 386584346.0, "step": 10179 }, { "epoch": 1.2950006360513928, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.00019168853759765625, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.584333419799805, "learning_rate": 1e-05, "loss": 0.4614, "mean_token_accuracy": 0.880326509475708, "num_tokens": 386619574.0, "step": 10180 }, { "epoch": 1.2951278463299833, "ewc_loss": 0.09814453125, "ewc_loss_diag": 0.00019168853759765625, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.427525997161865, "learning_rate": 1e-05, "loss": 0.4786, "mean_token_accuracy": 0.8689494729042053, "num_tokens": 386655111.0, "step": 10181 }, { "epoch": 1.2952550566085739, "ewc_loss": 0.125, "ewc_loss_diag": 0.00019168853759765625, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.900358200073242, "learning_rate": 1e-05, "loss": 0.5072, "mean_token_accuracy": 0.8693166971206665, "num_tokens": 386701102.0, "step": 10182 }, { "epoch": 1.2953822668871644, "ewc_loss": 0.10107421875, "ewc_loss_diag": 0.00019168853759765625, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.479251861572266, "learning_rate": 1e-05, "loss": 0.4785, "mean_token_accuracy": 0.8685691356658936, "num_tokens": 386740259.0, "step": 10183 }, { "epoch": 1.295509477165755, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.00019168853759765625, "ewc_loss_parallel": 0.00011396408081054688, "grad_norm": 7.2157416343688965, "learning_rate": 1e-05, "loss": 0.4909, "mean_token_accuracy": 0.8751686811447144, "num_tokens": 386780345.0, "step": 10184 }, { "epoch": 1.2956366874443455, "ewc_loss": 0.10888671875, "ewc_loss_diag": 0.00019168853759765625, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 5.772942066192627, "learning_rate": 1e-05, "loss": 0.4773, "mean_token_accuracy": 0.871307909488678, "num_tokens": 386819401.0, "step": 10185 }, { "epoch": 1.295763897722936, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.00019168853759765625, "ewc_loss_parallel": 0.00011539459228515625, "grad_norm": 7.272017002105713, "learning_rate": 1e-05, "loss": 0.5075, "mean_token_accuracy": 0.8733632564544678, "num_tokens": 386856617.0, "step": 10186 }, { "epoch": 1.2958911080015265, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.00019168853759765625, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.145359992980957, "learning_rate": 1e-05, "loss": 0.5426, "mean_token_accuracy": 0.8564698696136475, "num_tokens": 386895426.0, "step": 10187 }, { "epoch": 1.296018318280117, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.00019168853759765625, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.960466384887695, "learning_rate": 1e-05, "loss": 0.5284, "mean_token_accuracy": 0.862836480140686, "num_tokens": 386931577.0, "step": 10188 }, { "epoch": 1.2961455285587076, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.00019168853759765625, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.271607398986816, "learning_rate": 1e-05, "loss": 0.4833, "mean_token_accuracy": 0.8734968900680542, "num_tokens": 386968658.0, "step": 10189 }, { "epoch": 1.296272738837298, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.00019168853759765625, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.542757034301758, "learning_rate": 1e-05, "loss": 0.4949, "mean_token_accuracy": 0.8715415000915527, "num_tokens": 387005364.0, "step": 10190 }, { "epoch": 1.2963999491158886, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.00019168853759765625, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.234738349914551, "learning_rate": 1e-05, "loss": 0.4889, "mean_token_accuracy": 0.871956467628479, "num_tokens": 387048767.0, "step": 10191 }, { "epoch": 1.2965271593944792, "ewc_loss": 0.10888671875, "ewc_loss_diag": 0.00019168853759765625, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.166666030883789, "learning_rate": 1e-05, "loss": 0.4579, "mean_token_accuracy": 0.8798760175704956, "num_tokens": 387083456.0, "step": 10192 }, { "epoch": 1.2966543696730697, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.00019168853759765625, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.206104278564453, "learning_rate": 1e-05, "loss": 0.4857, "mean_token_accuracy": 0.8728954792022705, "num_tokens": 387124220.0, "step": 10193 }, { "epoch": 1.2967815799516602, "ewc_loss": 0.10302734375, "ewc_loss_diag": 0.00019168853759765625, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.9110002517700195, "learning_rate": 1e-05, "loss": 0.5267, "mean_token_accuracy": 0.8576767444610596, "num_tokens": 387163825.0, "step": 10194 }, { "epoch": 1.2969087902302507, "ewc_loss": 0.11083984375, "ewc_loss_diag": 0.00019168853759765625, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.201596260070801, "learning_rate": 1e-05, "loss": 0.5015, "mean_token_accuracy": 0.8691544532775879, "num_tokens": 387201230.0, "step": 10195 }, { "epoch": 1.297036000508841, "ewc_loss": 0.10107421875, "ewc_loss_diag": 0.00019168853759765625, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.750598907470703, "learning_rate": 1e-05, "loss": 0.4865, "mean_token_accuracy": 0.8651549816131592, "num_tokens": 387238426.0, "step": 10196 }, { "epoch": 1.2971632107874316, "ewc_loss": 0.11083984375, "ewc_loss_diag": 0.00019168853759765625, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.24090576171875, "learning_rate": 1e-05, "loss": 0.4917, "mean_token_accuracy": 0.8730874061584473, "num_tokens": 387275883.0, "step": 10197 }, { "epoch": 1.297290421066022, "ewc_loss": 0.099609375, "ewc_loss_diag": 0.00019168853759765625, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.613352298736572, "learning_rate": 1e-05, "loss": 0.4594, "mean_token_accuracy": 0.8775933384895325, "num_tokens": 387319053.0, "step": 10198 }, { "epoch": 1.2974176313446126, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.00019168853759765625, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.364388942718506, "learning_rate": 1e-05, "loss": 0.4617, "mean_token_accuracy": 0.877531886100769, "num_tokens": 387355013.0, "step": 10199 }, { "epoch": 1.2975448416232032, "ewc_loss": 0.0986328125, "ewc_loss_diag": 0.00019168853759765625, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.4771809577941895, "learning_rate": 1e-05, "loss": 0.4476, "mean_token_accuracy": 0.8785276412963867, "num_tokens": 387392736.0, "step": 10200 }, { "epoch": 1.2976720519017937, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.00019168853759765625, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.609381198883057, "learning_rate": 1e-05, "loss": 0.536, "mean_token_accuracy": 0.855473518371582, "num_tokens": 387431938.0, "step": 10201 }, { "epoch": 1.2977992621803842, "ewc_loss": 0.09716796875, "ewc_loss_diag": 0.00019168853759765625, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.380504608154297, "learning_rate": 1e-05, "loss": 0.4779, "mean_token_accuracy": 0.8692290186882019, "num_tokens": 387472724.0, "step": 10202 }, { "epoch": 1.2979264724589747, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.00019168853759765625, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 6.979145050048828, "learning_rate": 1e-05, "loss": 0.5152, "mean_token_accuracy": 0.8689800500869751, "num_tokens": 387513666.0, "step": 10203 }, { "epoch": 1.2980536827375653, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.00019168853759765625, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.464868545532227, "learning_rate": 1e-05, "loss": 0.4551, "mean_token_accuracy": 0.875188946723938, "num_tokens": 387544691.0, "step": 10204 }, { "epoch": 1.2981808930161556, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.00019168853759765625, "ewc_loss_parallel": 0.00011873245239257812, "grad_norm": 7.336198329925537, "learning_rate": 1e-05, "loss": 0.4897, "mean_token_accuracy": 0.8786531090736389, "num_tokens": 387580839.0, "step": 10205 }, { "epoch": 1.298308103294746, "ewc_loss": 0.11181640625, "ewc_loss_diag": 0.00019168853759765625, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 5.851998805999756, "learning_rate": 1e-05, "loss": 0.5099, "mean_token_accuracy": 0.862836480140686, "num_tokens": 387616258.0, "step": 10206 }, { "epoch": 1.2984353135733366, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.00019168853759765625, "ewc_loss_parallel": 0.00011682510375976562, "grad_norm": 7.3451104164123535, "learning_rate": 1e-05, "loss": 0.5473, "mean_token_accuracy": 0.8604097366333008, "num_tokens": 387654058.0, "step": 10207 }, { "epoch": 1.2985625238519272, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.00019168853759765625, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.254688262939453, "learning_rate": 1e-05, "loss": 0.533, "mean_token_accuracy": 0.8613998889923096, "num_tokens": 387690036.0, "step": 10208 }, { "epoch": 1.2986897341305177, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.00019168853759765625, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 6.967714786529541, "learning_rate": 1e-05, "loss": 0.5077, "mean_token_accuracy": 0.8680849075317383, "num_tokens": 387729114.0, "step": 10209 }, { "epoch": 1.2988169444091082, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.00019168853759765625, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.337018013000488, "learning_rate": 1e-05, "loss": 0.4927, "mean_token_accuracy": 0.8696478009223938, "num_tokens": 387768225.0, "step": 10210 }, { "epoch": 1.2989441546876987, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.00019168853759765625, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.547215461730957, "learning_rate": 1e-05, "loss": 0.5249, "mean_token_accuracy": 0.860154390335083, "num_tokens": 387802093.0, "step": 10211 }, { "epoch": 1.2990713649662893, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.00019168853759765625, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.268404483795166, "learning_rate": 1e-05, "loss": 0.4896, "mean_token_accuracy": 0.8708591461181641, "num_tokens": 387838857.0, "step": 10212 }, { "epoch": 1.2991985752448798, "ewc_loss": 0.10888671875, "ewc_loss_diag": 0.00019168853759765625, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.190054893493652, "learning_rate": 1e-05, "loss": 0.4946, "mean_token_accuracy": 0.8692827224731445, "num_tokens": 387878974.0, "step": 10213 }, { "epoch": 1.2993257855234703, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.00019168853759765625, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.2114057540893555, "learning_rate": 1e-05, "loss": 0.5264, "mean_token_accuracy": 0.8612970113754272, "num_tokens": 387920157.0, "step": 10214 }, { "epoch": 1.2994529958020609, "ewc_loss": 0.10400390625, "ewc_loss_diag": 0.00019168853759765625, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.944172382354736, "learning_rate": 1e-05, "loss": 0.4695, "mean_token_accuracy": 0.8726392388343811, "num_tokens": 387953345.0, "step": 10215 }, { "epoch": 1.2995802060806514, "ewc_loss": 0.11083984375, "ewc_loss_diag": 0.00019168853759765625, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.190036296844482, "learning_rate": 1e-05, "loss": 0.4718, "mean_token_accuracy": 0.8776814937591553, "num_tokens": 387990696.0, "step": 10216 }, { "epoch": 1.299707416359242, "ewc_loss": 0.10205078125, "ewc_loss_diag": 0.00019168853759765625, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.787553310394287, "learning_rate": 1e-05, "loss": 0.5177, "mean_token_accuracy": 0.8591883778572083, "num_tokens": 388028142.0, "step": 10217 }, { "epoch": 1.2998346266378324, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.00019168853759765625, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.203542232513428, "learning_rate": 1e-05, "loss": 0.5563, "mean_token_accuracy": 0.8428680896759033, "num_tokens": 388058095.0, "step": 10218 }, { "epoch": 1.299961836916423, "ewc_loss": 0.10009765625, "ewc_loss_diag": 0.00019168853759765625, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.6536383628845215, "learning_rate": 1e-05, "loss": 0.5038, "mean_token_accuracy": 0.8677618503570557, "num_tokens": 388097141.0, "step": 10219 }, { "epoch": 1.3000890471950133, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.00019168853759765625, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.286334991455078, "learning_rate": 1e-05, "loss": 0.4963, "mean_token_accuracy": 0.8703106045722961, "num_tokens": 388132138.0, "step": 10220 }, { "epoch": 1.3002162574736038, "ewc_loss": 0.09814453125, "ewc_loss_diag": 0.00019168853759765625, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.515667915344238, "learning_rate": 1e-05, "loss": 0.4834, "mean_token_accuracy": 0.8701907396316528, "num_tokens": 388172837.0, "step": 10221 }, { "epoch": 1.3003434677521943, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.00019168853759765625, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.487437725067139, "learning_rate": 1e-05, "loss": 0.5232, "mean_token_accuracy": 0.8572574853897095, "num_tokens": 388214797.0, "step": 10222 }, { "epoch": 1.3004706780307849, "ewc_loss": 0.0966796875, "ewc_loss_diag": 0.00019168853759765625, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.384644985198975, "learning_rate": 1e-05, "loss": 0.4941, "mean_token_accuracy": 0.8603188395500183, "num_tokens": 388256374.0, "step": 10223 }, { "epoch": 1.3005978883093754, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.00019168853759765625, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.85089635848999, "learning_rate": 1e-05, "loss": 0.4894, "mean_token_accuracy": 0.8720861673355103, "num_tokens": 388291227.0, "step": 10224 }, { "epoch": 1.300725098587966, "ewc_loss": 0.0986328125, "ewc_loss_diag": 0.00019168853759765625, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.368016242980957, "learning_rate": 1e-05, "loss": 0.4533, "mean_token_accuracy": 0.8777778744697571, "num_tokens": 388332963.0, "step": 10225 }, { "epoch": 1.3008523088665565, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.00019168853759765625, "ewc_loss_parallel": 0.000118255615234375, "grad_norm": 7.284167766571045, "learning_rate": 1e-05, "loss": 0.4659, "mean_token_accuracy": 0.8833314180374146, "num_tokens": 388372825.0, "step": 10226 }, { "epoch": 1.300979519145147, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.00019168853759765625, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 5.71678352355957, "learning_rate": 1e-05, "loss": 0.4822, "mean_token_accuracy": 0.8699072599411011, "num_tokens": 388411850.0, "step": 10227 }, { "epoch": 1.3011067294237375, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.00019168853759765625, "ewc_loss_parallel": 0.0001220703125, "grad_norm": 7.44326114654541, "learning_rate": 1e-05, "loss": 0.5339, "mean_token_accuracy": 0.8643491268157959, "num_tokens": 388449509.0, "step": 10228 }, { "epoch": 1.3012339397023278, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.00019168853759765625, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.19832706451416, "learning_rate": 1e-05, "loss": 0.5404, "mean_token_accuracy": 0.8624401092529297, "num_tokens": 388483698.0, "step": 10229 }, { "epoch": 1.3013611499809183, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.00019168853759765625, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 7.13654899597168, "learning_rate": 1e-05, "loss": 0.5175, "mean_token_accuracy": 0.8672910332679749, "num_tokens": 388516832.0, "step": 10230 }, { "epoch": 1.3014883602595089, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.00019168853759765625, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.36419677734375, "learning_rate": 1e-05, "loss": 0.4791, "mean_token_accuracy": 0.8782460689544678, "num_tokens": 388553272.0, "step": 10231 }, { "epoch": 1.3016155705380994, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.00019168853759765625, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.701510906219482, "learning_rate": 1e-05, "loss": 0.4511, "mean_token_accuracy": 0.8866263031959534, "num_tokens": 388589163.0, "step": 10232 }, { "epoch": 1.30174278081669, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.00019168853759765625, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.307793617248535, "learning_rate": 1e-05, "loss": 0.5285, "mean_token_accuracy": 0.8591770529747009, "num_tokens": 388628305.0, "step": 10233 }, { "epoch": 1.3018699910952805, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.00019168853759765625, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.316945552825928, "learning_rate": 1e-05, "loss": 0.5217, "mean_token_accuracy": 0.8608052134513855, "num_tokens": 388673462.0, "step": 10234 }, { "epoch": 1.301997201373871, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.00019168853759765625, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.24350643157959, "learning_rate": 1e-05, "loss": 0.5258, "mean_token_accuracy": 0.8608963489532471, "num_tokens": 388709357.0, "step": 10235 }, { "epoch": 1.3021244116524615, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.00019168853759765625, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 6.02813196182251, "learning_rate": 1e-05, "loss": 0.4402, "mean_token_accuracy": 0.8863474130630493, "num_tokens": 388746550.0, "step": 10236 }, { "epoch": 1.302251621931052, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.00019168853759765625, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.195034980773926, "learning_rate": 1e-05, "loss": 0.5505, "mean_token_accuracy": 0.8526581525802612, "num_tokens": 388795139.0, "step": 10237 }, { "epoch": 1.3023788322096426, "ewc_loss": 0.10302734375, "ewc_loss_diag": 0.00019168853759765625, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.852137565612793, "learning_rate": 1e-05, "loss": 0.47, "mean_token_accuracy": 0.8725835084915161, "num_tokens": 388832282.0, "step": 10238 }, { "epoch": 1.302506042488233, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.00019168853759765625, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.1757941246032715, "learning_rate": 1e-05, "loss": 0.4655, "mean_token_accuracy": 0.8771361112594604, "num_tokens": 388873974.0, "step": 10239 }, { "epoch": 1.3026332527668236, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.00019168853759765625, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.717601776123047, "learning_rate": 1e-05, "loss": 0.4364, "mean_token_accuracy": 0.881833553314209, "num_tokens": 388914562.0, "step": 10240 }, { "epoch": 1.3027604630454142, "ewc_loss": 0.10986328125, "ewc_loss_diag": 0.00019168853759765625, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.2443461418151855, "learning_rate": 1e-05, "loss": 0.5247, "mean_token_accuracy": 0.859797477722168, "num_tokens": 388950293.0, "step": 10241 }, { "epoch": 1.3028876733240047, "ewc_loss": 0.0986328125, "ewc_loss_diag": 0.00019168853759765625, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.586941719055176, "learning_rate": 1e-05, "loss": 0.4899, "mean_token_accuracy": 0.8672580122947693, "num_tokens": 388986649.0, "step": 10242 }, { "epoch": 1.3030148836025952, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.00019168853759765625, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.380537509918213, "learning_rate": 1e-05, "loss": 0.4604, "mean_token_accuracy": 0.8812557458877563, "num_tokens": 389027779.0, "step": 10243 }, { "epoch": 1.3031420938811857, "ewc_loss": 0.0986328125, "ewc_loss_diag": 0.00019168853759765625, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.451387882232666, "learning_rate": 1e-05, "loss": 0.5192, "mean_token_accuracy": 0.8525636196136475, "num_tokens": 389062760.0, "step": 10244 }, { "epoch": 1.303269304159776, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.00019168853759765625, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.676933765411377, "learning_rate": 1e-05, "loss": 0.5556, "mean_token_accuracy": 0.8548904061317444, "num_tokens": 389101892.0, "step": 10245 }, { "epoch": 1.3033965144383666, "ewc_loss": 0.09716796875, "ewc_loss_diag": 0.00019168853759765625, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.350577354431152, "learning_rate": 1e-05, "loss": 0.4908, "mean_token_accuracy": 0.8688318729400635, "num_tokens": 389140155.0, "step": 10246 }, { "epoch": 1.303523724716957, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.00019168853759765625, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 7.106697082519531, "learning_rate": 1e-05, "loss": 0.5376, "mean_token_accuracy": 0.860751211643219, "num_tokens": 389181563.0, "step": 10247 }, { "epoch": 1.3036509349955476, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.00019168853759765625, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.498570442199707, "learning_rate": 1e-05, "loss": 0.4727, "mean_token_accuracy": 0.8732125163078308, "num_tokens": 389217415.0, "step": 10248 }, { "epoch": 1.3037781452741382, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.00019168853759765625, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.468443870544434, "learning_rate": 1e-05, "loss": 0.4872, "mean_token_accuracy": 0.8829352855682373, "num_tokens": 389255889.0, "step": 10249 }, { "epoch": 1.3039053555527287, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.00019168853759765625, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.002854824066162, "learning_rate": 1e-05, "loss": 0.4694, "mean_token_accuracy": 0.874101996421814, "num_tokens": 389297650.0, "step": 10250 }, { "epoch": 1.3040325658313192, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.00019168853759765625, "ewc_loss_parallel": 0.000118255615234375, "grad_norm": 7.3683295249938965, "learning_rate": 1e-05, "loss": 0.566, "mean_token_accuracy": 0.85215163230896, "num_tokens": 389339647.0, "step": 10251 }, { "epoch": 1.3041597761099097, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.00019168853759765625, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.384476661682129, "learning_rate": 1e-05, "loss": 0.5034, "mean_token_accuracy": 0.8726117610931396, "num_tokens": 389376135.0, "step": 10252 }, { "epoch": 1.3042869863885003, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.00019168853759765625, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.929250240325928, "learning_rate": 1e-05, "loss": 0.5388, "mean_token_accuracy": 0.8590102195739746, "num_tokens": 389410516.0, "step": 10253 }, { "epoch": 1.3044141966670906, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.00019168853759765625, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.3812408447265625, "learning_rate": 1e-05, "loss": 0.5, "mean_token_accuracy": 0.8704294562339783, "num_tokens": 389452712.0, "step": 10254 }, { "epoch": 1.304541406945681, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.00019168853759765625, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.514859199523926, "learning_rate": 1e-05, "loss": 0.4958, "mean_token_accuracy": 0.8704010844230652, "num_tokens": 389489257.0, "step": 10255 }, { "epoch": 1.3046686172242716, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.00019168853759765625, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.2949347496032715, "learning_rate": 1e-05, "loss": 0.5375, "mean_token_accuracy": 0.85340416431427, "num_tokens": 389524344.0, "step": 10256 }, { "epoch": 1.3047958275028622, "ewc_loss": 0.109375, "ewc_loss_diag": 0.00019168853759765625, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.164243221282959, "learning_rate": 1e-05, "loss": 0.5432, "mean_token_accuracy": 0.8498274087905884, "num_tokens": 389564351.0, "step": 10257 }, { "epoch": 1.3049230377814527, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.00019168853759765625, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.225907325744629, "learning_rate": 1e-05, "loss": 0.4844, "mean_token_accuracy": 0.869452178478241, "num_tokens": 389604606.0, "step": 10258 }, { "epoch": 1.3050502480600432, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.00019168853759765625, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.942033290863037, "learning_rate": 1e-05, "loss": 0.4872, "mean_token_accuracy": 0.8701512813568115, "num_tokens": 389639600.0, "step": 10259 }, { "epoch": 1.3051774583386337, "ewc_loss": 0.10888671875, "ewc_loss_diag": 0.00019168853759765625, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.188142776489258, "learning_rate": 1e-05, "loss": 0.5237, "mean_token_accuracy": 0.8570601940155029, "num_tokens": 389677770.0, "step": 10260 }, { "epoch": 1.3053046686172243, "ewc_loss": 0.10205078125, "ewc_loss_diag": 0.00019168853759765625, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.794136047363281, "learning_rate": 1e-05, "loss": 0.5287, "mean_token_accuracy": 0.855192244052887, "num_tokens": 389715366.0, "step": 10261 }, { "epoch": 1.3054318788958148, "ewc_loss": 0.10888671875, "ewc_loss_diag": 0.00019168853759765625, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.229300022125244, "learning_rate": 1e-05, "loss": 0.477, "mean_token_accuracy": 0.8744471669197083, "num_tokens": 389754745.0, "step": 10262 }, { "epoch": 1.3055590891744053, "ewc_loss": 0.10009765625, "ewc_loss_diag": 0.00019168853759765625, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.651532173156738, "learning_rate": 1e-05, "loss": 0.5121, "mean_token_accuracy": 0.8591993451118469, "num_tokens": 389798599.0, "step": 10263 }, { "epoch": 1.3056862994529959, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.00019168853759765625, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.316580772399902, "learning_rate": 1e-05, "loss": 0.4969, "mean_token_accuracy": 0.8664800524711609, "num_tokens": 389831861.0, "step": 10264 }, { "epoch": 1.3058135097315864, "ewc_loss": 0.09814453125, "ewc_loss_diag": 0.00019168853759765625, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.512566566467285, "learning_rate": 1e-05, "loss": 0.5204, "mean_token_accuracy": 0.8573041558265686, "num_tokens": 389877785.0, "step": 10265 }, { "epoch": 1.305940720010177, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.00019168853759765625, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.538841247558594, "learning_rate": 1e-05, "loss": 0.4586, "mean_token_accuracy": 0.8830482959747314, "num_tokens": 389921969.0, "step": 10266 }, { "epoch": 1.3060679302887674, "ewc_loss": 0.09765625, "ewc_loss_diag": 0.00019168853759765625, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.381000518798828, "learning_rate": 1e-05, "loss": 0.4661, "mean_token_accuracy": 0.8709879517555237, "num_tokens": 389959281.0, "step": 10267 }, { "epoch": 1.306195140567358, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.00019168853759765625, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.923285961151123, "learning_rate": 1e-05, "loss": 0.47, "mean_token_accuracy": 0.8804593086242676, "num_tokens": 389992580.0, "step": 10268 }, { "epoch": 1.3063223508459483, "ewc_loss": 0.099609375, "ewc_loss_diag": 0.0001926422119140625, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.403496265411377, "learning_rate": 1e-05, "loss": 0.517, "mean_token_accuracy": 0.8591181039810181, "num_tokens": 390029377.0, "step": 10269 }, { "epoch": 1.3064495611245388, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.0001926422119140625, "ewc_loss_parallel": 0.000118255615234375, "grad_norm": 7.349575042724609, "learning_rate": 1e-05, "loss": 0.501, "mean_token_accuracy": 0.8733969330787659, "num_tokens": 390059679.0, "step": 10270 }, { "epoch": 1.3065767714031293, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.0001926422119140625, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 5.7651591300964355, "learning_rate": 1e-05, "loss": 0.4936, "mean_token_accuracy": 0.8664991855621338, "num_tokens": 390096751.0, "step": 10271 }, { "epoch": 1.3067039816817199, "ewc_loss": 0.140625, "ewc_loss_diag": 0.0001926422119140625, "ewc_loss_parallel": 0.00012111663818359375, "grad_norm": 7.464784622192383, "learning_rate": 1e-05, "loss": 0.5542, "mean_token_accuracy": 0.8582786321640015, "num_tokens": 390138648.0, "step": 10272 }, { "epoch": 1.3068311919603104, "ewc_loss": 0.125, "ewc_loss_diag": 0.0001926422119140625, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.275636672973633, "learning_rate": 1e-05, "loss": 0.5109, "mean_token_accuracy": 0.8669751882553101, "num_tokens": 390176008.0, "step": 10273 }, { "epoch": 1.306958402238901, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.0001926422119140625, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 7.108123779296875, "learning_rate": 1e-05, "loss": 0.518, "mean_token_accuracy": 0.865940272808075, "num_tokens": 390218120.0, "step": 10274 }, { "epoch": 1.3070856125174914, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.0001926422119140625, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.407708644866943, "learning_rate": 1e-05, "loss": 0.4989, "mean_token_accuracy": 0.8721342086791992, "num_tokens": 390264776.0, "step": 10275 }, { "epoch": 1.307212822796082, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.0001926422119140625, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.652597904205322, "learning_rate": 1e-05, "loss": 0.4812, "mean_token_accuracy": 0.8780083060264587, "num_tokens": 390298798.0, "step": 10276 }, { "epoch": 1.3073400330746725, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.0001926422119140625, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.340581893920898, "learning_rate": 1e-05, "loss": 0.5243, "mean_token_accuracy": 0.8636606335639954, "num_tokens": 390333332.0, "step": 10277 }, { "epoch": 1.3074672433532628, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.0001926422119140625, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.2644734382629395, "learning_rate": 1e-05, "loss": 0.5243, "mean_token_accuracy": 0.861112117767334, "num_tokens": 390374128.0, "step": 10278 }, { "epoch": 1.3075944536318533, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.0001926422119140625, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.256985664367676, "learning_rate": 1e-05, "loss": 0.4951, "mean_token_accuracy": 0.8665040731430054, "num_tokens": 390416509.0, "step": 10279 }, { "epoch": 1.3077216639104439, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.0001926422119140625, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.999467372894287, "learning_rate": 1e-05, "loss": 0.4865, "mean_token_accuracy": 0.8690738677978516, "num_tokens": 390448129.0, "step": 10280 }, { "epoch": 1.3078488741890344, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.0001926422119140625, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.217191219329834, "learning_rate": 1e-05, "loss": 0.5029, "mean_token_accuracy": 0.8676648736000061, "num_tokens": 390489329.0, "step": 10281 }, { "epoch": 1.307976084467625, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.0001926422119140625, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.814338207244873, "learning_rate": 1e-05, "loss": 0.4809, "mean_token_accuracy": 0.8703707456588745, "num_tokens": 390529839.0, "step": 10282 }, { "epoch": 1.3081032947462155, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.0001926422119140625, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.216154098510742, "learning_rate": 1e-05, "loss": 0.5077, "mean_token_accuracy": 0.8685773611068726, "num_tokens": 390568888.0, "step": 10283 }, { "epoch": 1.308230505024806, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.0001926422119140625, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.694437503814697, "learning_rate": 1e-05, "loss": 0.4658, "mean_token_accuracy": 0.8744391798973083, "num_tokens": 390606709.0, "step": 10284 }, { "epoch": 1.3083577153033965, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.0001926422119140625, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.278141975402832, "learning_rate": 1e-05, "loss": 0.51, "mean_token_accuracy": 0.8662889003753662, "num_tokens": 390645027.0, "step": 10285 }, { "epoch": 1.308484925581987, "ewc_loss": 0.0986328125, "ewc_loss_diag": 0.0001926422119140625, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.562844753265381, "learning_rate": 1e-05, "loss": 0.4896, "mean_token_accuracy": 0.8672679662704468, "num_tokens": 390685923.0, "step": 10286 }, { "epoch": 1.3086121358605776, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.0001926422119140625, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.455356597900391, "learning_rate": 1e-05, "loss": 0.4563, "mean_token_accuracy": 0.882335364818573, "num_tokens": 390728410.0, "step": 10287 }, { "epoch": 1.308739346139168, "ewc_loss": 0.0986328125, "ewc_loss_diag": 0.0001926422119140625, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.432891845703125, "learning_rate": 1e-05, "loss": 0.4985, "mean_token_accuracy": 0.8607651591300964, "num_tokens": 390764146.0, "step": 10288 }, { "epoch": 1.3088665564177586, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.0001926422119140625, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.779539108276367, "learning_rate": 1e-05, "loss": 0.5059, "mean_token_accuracy": 0.8669623136520386, "num_tokens": 390800363.0, "step": 10289 }, { "epoch": 1.3089937666963491, "ewc_loss": 0.0986328125, "ewc_loss_diag": 0.0001926422119140625, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.376680374145508, "learning_rate": 1e-05, "loss": 0.476, "mean_token_accuracy": 0.8697380423545837, "num_tokens": 390840213.0, "step": 10290 }, { "epoch": 1.3091209769749397, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.0001926422119140625, "ewc_loss_parallel": 0.00011491775512695312, "grad_norm": 7.198947906494141, "learning_rate": 1e-05, "loss": 0.4652, "mean_token_accuracy": 0.8836010098457336, "num_tokens": 390879087.0, "step": 10291 }, { "epoch": 1.3092481872535302, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.0001926422119140625, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.603311061859131, "learning_rate": 1e-05, "loss": 0.4234, "mean_token_accuracy": 0.8857628107070923, "num_tokens": 390914074.0, "step": 10292 }, { "epoch": 1.3093753975321207, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.0001926422119140625, "ewc_loss_parallel": 0.0001220703125, "grad_norm": 7.461278438568115, "learning_rate": 1e-05, "loss": 0.494, "mean_token_accuracy": 0.8784170150756836, "num_tokens": 390956283.0, "step": 10293 }, { "epoch": 1.309502607810711, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.0001926422119140625, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.10893440246582, "learning_rate": 1e-05, "loss": 0.5133, "mean_token_accuracy": 0.8694490194320679, "num_tokens": 390996898.0, "step": 10294 }, { "epoch": 1.3096298180893016, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.0001926422119140625, "ewc_loss_parallel": 0.0001125335693359375, "grad_norm": 7.256256580352783, "learning_rate": 1e-05, "loss": 0.4777, "mean_token_accuracy": 0.879166841506958, "num_tokens": 391033280.0, "step": 10295 }, { "epoch": 1.309757028367892, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0001926422119140625, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.3743977546691895, "learning_rate": 1e-05, "loss": 0.4551, "mean_token_accuracy": 0.8863356113433838, "num_tokens": 391074159.0, "step": 10296 }, { "epoch": 1.3098842386464826, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.0001926422119140625, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.811529636383057, "learning_rate": 1e-05, "loss": 0.4858, "mean_token_accuracy": 0.873204231262207, "num_tokens": 391115649.0, "step": 10297 }, { "epoch": 1.3100114489250732, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.0001926422119140625, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.351751804351807, "learning_rate": 1e-05, "loss": 0.4397, "mean_token_accuracy": 0.8904916048049927, "num_tokens": 391147601.0, "step": 10298 }, { "epoch": 1.3101386592036637, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.0001926422119140625, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.403350830078125, "learning_rate": 1e-05, "loss": 0.4927, "mean_token_accuracy": 0.8687146902084351, "num_tokens": 391183483.0, "step": 10299 }, { "epoch": 1.3102658694822542, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.0001926422119140625, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.278553009033203, "learning_rate": 1e-05, "loss": 0.4539, "mean_token_accuracy": 0.8862473964691162, "num_tokens": 391217346.0, "step": 10300 }, { "epoch": 1.3103930797608447, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.0001926422119140625, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 6.0760602951049805, "learning_rate": 1e-05, "loss": 0.4731, "mean_token_accuracy": 0.8764694929122925, "num_tokens": 391255575.0, "step": 10301 }, { "epoch": 1.3105202900394353, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.0001926422119140625, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.239289283752441, "learning_rate": 1e-05, "loss": 0.4768, "mean_token_accuracy": 0.8765992522239685, "num_tokens": 391291680.0, "step": 10302 }, { "epoch": 1.3106475003180256, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.0001926422119140625, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.858347415924072, "learning_rate": 1e-05, "loss": 0.488, "mean_token_accuracy": 0.8718493580818176, "num_tokens": 391330034.0, "step": 10303 }, { "epoch": 1.310774710596616, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.0001926422119140625, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.219865322113037, "learning_rate": 1e-05, "loss": 0.4837, "mean_token_accuracy": 0.8789058923721313, "num_tokens": 391370558.0, "step": 10304 }, { "epoch": 1.3109019208752066, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.0001926422119140625, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.708837985992432, "learning_rate": 1e-05, "loss": 0.4886, "mean_token_accuracy": 0.8666411638259888, "num_tokens": 391414091.0, "step": 10305 }, { "epoch": 1.3110291311537972, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.0001926422119140625, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.281211853027344, "learning_rate": 1e-05, "loss": 0.5297, "mean_token_accuracy": 0.8573915958404541, "num_tokens": 391458057.0, "step": 10306 }, { "epoch": 1.3111563414323877, "ewc_loss": 0.0986328125, "ewc_loss_diag": 0.0001926422119140625, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.5792765617370605, "learning_rate": 1e-05, "loss": 0.4653, "mean_token_accuracy": 0.8712040185928345, "num_tokens": 391495419.0, "step": 10307 }, { "epoch": 1.3112835517109782, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.0001926422119140625, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.45694637298584, "learning_rate": 1e-05, "loss": 0.5035, "mean_token_accuracy": 0.8674564957618713, "num_tokens": 391529353.0, "step": 10308 }, { "epoch": 1.3114107619895687, "ewc_loss": 0.09765625, "ewc_loss_diag": 0.0001926422119140625, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.4542999267578125, "learning_rate": 1e-05, "loss": 0.5652, "mean_token_accuracy": 0.8445858955383301, "num_tokens": 391571111.0, "step": 10309 }, { "epoch": 1.3115379722681593, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.0001926422119140625, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.745020389556885, "learning_rate": 1e-05, "loss": 0.5128, "mean_token_accuracy": 0.8662660121917725, "num_tokens": 391612284.0, "step": 10310 }, { "epoch": 1.3116651825467498, "ewc_loss": 0.09765625, "ewc_loss_diag": 0.0001926422119140625, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.39564323425293, "learning_rate": 1e-05, "loss": 0.51, "mean_token_accuracy": 0.8604031205177307, "num_tokens": 391646934.0, "step": 10311 }, { "epoch": 1.3117923928253403, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.0001926422119140625, "ewc_loss_parallel": 0.0001125335693359375, "grad_norm": 7.139422416687012, "learning_rate": 1e-05, "loss": 0.4888, "mean_token_accuracy": 0.8768481016159058, "num_tokens": 391686871.0, "step": 10312 }, { "epoch": 1.3119196031039309, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.0001926422119140625, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.584512233734131, "learning_rate": 1e-05, "loss": 0.481, "mean_token_accuracy": 0.8685141801834106, "num_tokens": 391728188.0, "step": 10313 }, { "epoch": 1.3120468133825214, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.0001926422119140625, "ewc_loss_parallel": 0.00012063980102539062, "grad_norm": 7.41071081161499, "learning_rate": 1e-05, "loss": 0.5374, "mean_token_accuracy": 0.8613228797912598, "num_tokens": 391767434.0, "step": 10314 }, { "epoch": 1.312174023661112, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.0001926422119140625, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.050046443939209, "learning_rate": 1e-05, "loss": 0.4821, "mean_token_accuracy": 0.8746138215065002, "num_tokens": 391809581.0, "step": 10315 }, { "epoch": 1.3123012339397024, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0001926422119140625, "ewc_loss_parallel": 0.00011348724365234375, "grad_norm": 7.231363773345947, "learning_rate": 1e-05, "loss": 0.5431, "mean_token_accuracy": 0.8637442588806152, "num_tokens": 391849877.0, "step": 10316 }, { "epoch": 1.312428444218293, "ewc_loss": 0.125, "ewc_loss_diag": 0.0001926422119140625, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.341038227081299, "learning_rate": 1e-05, "loss": 0.5344, "mean_token_accuracy": 0.8593438863754272, "num_tokens": 391885995.0, "step": 10317 }, { "epoch": 1.3125556544968833, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.0001926422119140625, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.8006720542907715, "learning_rate": 1e-05, "loss": 0.4624, "mean_token_accuracy": 0.8798564672470093, "num_tokens": 391919100.0, "step": 10318 }, { "epoch": 1.3126828647754738, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.0001926422119140625, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.344374656677246, "learning_rate": 1e-05, "loss": 0.514, "mean_token_accuracy": 0.8652304410934448, "num_tokens": 391953246.0, "step": 10319 }, { "epoch": 1.3128100750540643, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.0001926422119140625, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.370446681976318, "learning_rate": 1e-05, "loss": 0.4821, "mean_token_accuracy": 0.8744040131568909, "num_tokens": 391986618.0, "step": 10320 }, { "epoch": 1.3129372853326549, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.0001926422119140625, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.272891521453857, "learning_rate": 1e-05, "loss": 0.5023, "mean_token_accuracy": 0.8647369146347046, "num_tokens": 392029663.0, "step": 10321 }, { "epoch": 1.3130644956112454, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.0001926422119140625, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 6.053037643432617, "learning_rate": 1e-05, "loss": 0.4953, "mean_token_accuracy": 0.8674248456954956, "num_tokens": 392063329.0, "step": 10322 }, { "epoch": 1.313191705889836, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.0001926422119140625, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.2361531257629395, "learning_rate": 1e-05, "loss": 0.4782, "mean_token_accuracy": 0.874092698097229, "num_tokens": 392101400.0, "step": 10323 }, { "epoch": 1.3133189161684264, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.0001926422119140625, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.870272636413574, "learning_rate": 1e-05, "loss": 0.4893, "mean_token_accuracy": 0.8658355474472046, "num_tokens": 392140146.0, "step": 10324 }, { "epoch": 1.313446126447017, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.0001926422119140625, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.212558269500732, "learning_rate": 1e-05, "loss": 0.4727, "mean_token_accuracy": 0.87911057472229, "num_tokens": 392176427.0, "step": 10325 }, { "epoch": 1.3135733367256075, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.0001926422119140625, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.722675800323486, "learning_rate": 1e-05, "loss": 0.5055, "mean_token_accuracy": 0.8633385896682739, "num_tokens": 392221105.0, "step": 10326 }, { "epoch": 1.3137005470041978, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.0001926422119140625, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.2576422691345215, "learning_rate": 1e-05, "loss": 0.5469, "mean_token_accuracy": 0.8524152636528015, "num_tokens": 392257301.0, "step": 10327 }, { "epoch": 1.3138277572827883, "ewc_loss": 0.099609375, "ewc_loss_diag": 0.0001926422119140625, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.608078479766846, "learning_rate": 1e-05, "loss": 0.4791, "mean_token_accuracy": 0.8711847066879272, "num_tokens": 392295849.0, "step": 10328 }, { "epoch": 1.3139549675613789, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.0001926422119140625, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.395209312438965, "learning_rate": 1e-05, "loss": 0.517, "mean_token_accuracy": 0.8582706451416016, "num_tokens": 392336480.0, "step": 10329 }, { "epoch": 1.3140821778399694, "ewc_loss": 0.0986328125, "ewc_loss_diag": 0.0001926422119140625, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.489860534667969, "learning_rate": 1e-05, "loss": 0.5016, "mean_token_accuracy": 0.8612971305847168, "num_tokens": 392372562.0, "step": 10330 }, { "epoch": 1.31420938811856, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.0001926422119140625, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.640872478485107, "learning_rate": 1e-05, "loss": 0.4702, "mean_token_accuracy": 0.8800952434539795, "num_tokens": 392408843.0, "step": 10331 }, { "epoch": 1.3143365983971504, "ewc_loss": 0.09765625, "ewc_loss_diag": 0.0001926422119140625, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.390450477600098, "learning_rate": 1e-05, "loss": 0.4755, "mean_token_accuracy": 0.8712043166160583, "num_tokens": 392440725.0, "step": 10332 }, { "epoch": 1.314463808675741, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0001926422119140625, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 7.015759468078613, "learning_rate": 1e-05, "loss": 0.5537, "mean_token_accuracy": 0.855907142162323, "num_tokens": 392484643.0, "step": 10333 }, { "epoch": 1.3145910189543315, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.0001926422119140625, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.481843948364258, "learning_rate": 1e-05, "loss": 0.4688, "mean_token_accuracy": 0.8724459409713745, "num_tokens": 392523849.0, "step": 10334 }, { "epoch": 1.314718229232922, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0001926422119140625, "ewc_loss_parallel": 0.00011920928955078125, "grad_norm": 7.374202728271484, "learning_rate": 1e-05, "loss": 0.5286, "mean_token_accuracy": 0.8696571588516235, "num_tokens": 392559598.0, "step": 10335 }, { "epoch": 1.3148454395115126, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.0001926422119140625, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 5.90049409866333, "learning_rate": 1e-05, "loss": 0.5524, "mean_token_accuracy": 0.8502620458602905, "num_tokens": 392603856.0, "step": 10336 }, { "epoch": 1.314972649790103, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.0001926422119140625, "ewc_loss_parallel": 0.000118255615234375, "grad_norm": 7.35805606842041, "learning_rate": 1e-05, "loss": 0.4925, "mean_token_accuracy": 0.8771049976348877, "num_tokens": 392645317.0, "step": 10337 }, { "epoch": 1.3150998600686936, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.0001926422119140625, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.29451322555542, "learning_rate": 1e-05, "loss": 0.5172, "mean_token_accuracy": 0.8661888837814331, "num_tokens": 392681650.0, "step": 10338 }, { "epoch": 1.3152270703472841, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0001926422119140625, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.966154098510742, "learning_rate": 1e-05, "loss": 0.5137, "mean_token_accuracy": 0.8608624935150146, "num_tokens": 392711588.0, "step": 10339 }, { "epoch": 1.3153542806258747, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.0001926422119140625, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.356692790985107, "learning_rate": 1e-05, "loss": 0.4886, "mean_token_accuracy": 0.8735114932060242, "num_tokens": 392758504.0, "step": 10340 }, { "epoch": 1.3154814909044652, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.0001926422119140625, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.538491725921631, "learning_rate": 1e-05, "loss": 0.4776, "mean_token_accuracy": 0.8724293112754822, "num_tokens": 392792153.0, "step": 10341 }, { "epoch": 1.3156087011830557, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0001926422119140625, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.279628276824951, "learning_rate": 1e-05, "loss": 0.4587, "mean_token_accuracy": 0.881589412689209, "num_tokens": 392840971.0, "step": 10342 }, { "epoch": 1.315735911461646, "ewc_loss": 0.109375, "ewc_loss_diag": 0.0001926422119140625, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.180316925048828, "learning_rate": 1e-05, "loss": 0.5159, "mean_token_accuracy": 0.8633816242218018, "num_tokens": 392882760.0, "step": 10343 }, { "epoch": 1.3158631217402366, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.0001926422119140625, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.243616104125977, "learning_rate": 1e-05, "loss": 0.5348, "mean_token_accuracy": 0.8563711643218994, "num_tokens": 392922165.0, "step": 10344 }, { "epoch": 1.315990332018827, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.0001926422119140625, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.928347110748291, "learning_rate": 1e-05, "loss": 0.512, "mean_token_accuracy": 0.8601260185241699, "num_tokens": 392957181.0, "step": 10345 }, { "epoch": 1.3161175422974176, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.0001926422119140625, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.217379570007324, "learning_rate": 1e-05, "loss": 0.5152, "mean_token_accuracy": 0.8624216318130493, "num_tokens": 392998405.0, "step": 10346 }, { "epoch": 1.3162447525760081, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.0001926422119140625, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.77951192855835, "learning_rate": 1e-05, "loss": 0.4755, "mean_token_accuracy": 0.8730833530426025, "num_tokens": 393036437.0, "step": 10347 }, { "epoch": 1.3163719628545987, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.0001926422119140625, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.238156318664551, "learning_rate": 1e-05, "loss": 0.4687, "mean_token_accuracy": 0.8755614757537842, "num_tokens": 393070711.0, "step": 10348 }, { "epoch": 1.3164991731331892, "ewc_loss": 0.099609375, "ewc_loss_diag": 0.0001926422119140625, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.6598076820373535, "learning_rate": 1e-05, "loss": 0.5224, "mean_token_accuracy": 0.8525439500808716, "num_tokens": 393109052.0, "step": 10349 }, { "epoch": 1.3166263834117797, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.0001926422119140625, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.33392858505249, "learning_rate": 1e-05, "loss": 0.5315, "mean_token_accuracy": 0.8601527214050293, "num_tokens": 393146244.0, "step": 10350 }, { "epoch": 1.3167535936903703, "ewc_loss": 0.09765625, "ewc_loss_diag": 0.0001926422119140625, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.5308403968811035, "learning_rate": 1e-05, "loss": 0.4732, "mean_token_accuracy": 0.8720730543136597, "num_tokens": 393180630.0, "step": 10351 }, { "epoch": 1.3168808039689606, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.0001926422119140625, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.534182071685791, "learning_rate": 1e-05, "loss": 0.5025, "mean_token_accuracy": 0.8691762089729309, "num_tokens": 393222687.0, "step": 10352 }, { "epoch": 1.317008014247551, "ewc_loss": 0.09765625, "ewc_loss_diag": 0.0001926422119140625, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.4110918045043945, "learning_rate": 1e-05, "loss": 0.4972, "mean_token_accuracy": 0.8632992506027222, "num_tokens": 393267383.0, "step": 10353 }, { "epoch": 1.3171352245261416, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.0001926422119140625, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.874640464782715, "learning_rate": 1e-05, "loss": 0.4892, "mean_token_accuracy": 0.8754382133483887, "num_tokens": 393302055.0, "step": 10354 }, { "epoch": 1.3172624348047322, "ewc_loss": 0.099609375, "ewc_loss_diag": 0.0001926422119140625, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.416676044464111, "learning_rate": 1e-05, "loss": 0.4394, "mean_token_accuracy": 0.8802969455718994, "num_tokens": 393336863.0, "step": 10355 }, { "epoch": 1.3173896450833227, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0001926422119140625, "ewc_loss_parallel": 0.00011730194091796875, "grad_norm": 7.291508197784424, "learning_rate": 1e-05, "loss": 0.5829, "mean_token_accuracy": 0.846700131893158, "num_tokens": 393373602.0, "step": 10356 }, { "epoch": 1.3175168553619132, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.0001926422119140625, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 5.739609241485596, "learning_rate": 1e-05, "loss": 0.5189, "mean_token_accuracy": 0.8579760789871216, "num_tokens": 393409858.0, "step": 10357 }, { "epoch": 1.3176440656405037, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.0001926422119140625, "ewc_loss_parallel": 0.00012063980102539062, "grad_norm": 7.427793979644775, "learning_rate": 1e-05, "loss": 0.5042, "mean_token_accuracy": 0.8747537732124329, "num_tokens": 393451198.0, "step": 10358 }, { "epoch": 1.3177712759190943, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.0001926422119140625, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.195790767669678, "learning_rate": 1e-05, "loss": 0.4716, "mean_token_accuracy": 0.8789035677909851, "num_tokens": 393486916.0, "step": 10359 }, { "epoch": 1.3178984861976848, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.0001926422119140625, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 7.1236186027526855, "learning_rate": 1e-05, "loss": 0.4754, "mean_token_accuracy": 0.881851315498352, "num_tokens": 393520039.0, "step": 10360 }, { "epoch": 1.3180256964762753, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.0001926422119140625, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.370518207550049, "learning_rate": 1e-05, "loss": 0.5418, "mean_token_accuracy": 0.8576615452766418, "num_tokens": 393557960.0, "step": 10361 }, { "epoch": 1.3181529067548658, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0001926422119140625, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.684469699859619, "learning_rate": 1e-05, "loss": 0.5389, "mean_token_accuracy": 0.8567742109298706, "num_tokens": 393592697.0, "step": 10362 }, { "epoch": 1.3182801170334564, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.0001926422119140625, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.307553768157959, "learning_rate": 1e-05, "loss": 0.5181, "mean_token_accuracy": 0.8603646755218506, "num_tokens": 393635423.0, "step": 10363 }, { "epoch": 1.318407327312047, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.0001926422119140625, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.291435718536377, "learning_rate": 1e-05, "loss": 0.4721, "mean_token_accuracy": 0.8748380541801453, "num_tokens": 393673508.0, "step": 10364 }, { "epoch": 1.3185345375906374, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.0001926422119140625, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.232683181762695, "learning_rate": 1e-05, "loss": 0.4521, "mean_token_accuracy": 0.8852179050445557, "num_tokens": 393711974.0, "step": 10365 }, { "epoch": 1.318661747869228, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.0001926422119140625, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 6.022312164306641, "learning_rate": 1e-05, "loss": 0.5593, "mean_token_accuracy": 0.8469736576080322, "num_tokens": 393752670.0, "step": 10366 }, { "epoch": 1.3187889581478183, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.0001926422119140625, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.192480564117432, "learning_rate": 1e-05, "loss": 0.4991, "mean_token_accuracy": 0.8647562265396118, "num_tokens": 393794312.0, "step": 10367 }, { "epoch": 1.3189161684264088, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.8557868003845215, "learning_rate": 1e-05, "loss": 0.5306, "mean_token_accuracy": 0.8561069965362549, "num_tokens": 393833830.0, "step": 10368 }, { "epoch": 1.3190433787049993, "ewc_loss": 0.10986328125, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.18432092666626, "learning_rate": 1e-05, "loss": 0.4816, "mean_token_accuracy": 0.8715559244155884, "num_tokens": 393874844.0, "step": 10369 }, { "epoch": 1.3191705889835899, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.717009544372559, "learning_rate": 1e-05, "loss": 0.5353, "mean_token_accuracy": 0.8519014716148376, "num_tokens": 393913861.0, "step": 10370 }, { "epoch": 1.3192977992621804, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.2604780197143555, "learning_rate": 1e-05, "loss": 0.5229, "mean_token_accuracy": 0.8628174066543579, "num_tokens": 393953014.0, "step": 10371 }, { "epoch": 1.319425009540771, "ewc_loss": 0.0986328125, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.571935176849365, "learning_rate": 1e-05, "loss": 0.4874, "mean_token_accuracy": 0.8685535192489624, "num_tokens": 393995603.0, "step": 10372 }, { "epoch": 1.3195522198193614, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.4469685554504395, "learning_rate": 1e-05, "loss": 0.4741, "mean_token_accuracy": 0.8759899139404297, "num_tokens": 394032127.0, "step": 10373 }, { "epoch": 1.319679430097952, "ewc_loss": 0.0986328125, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.428903579711914, "learning_rate": 1e-05, "loss": 0.4276, "mean_token_accuracy": 0.8839924335479736, "num_tokens": 394067153.0, "step": 10374 }, { "epoch": 1.3198066403765425, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.776519775390625, "learning_rate": 1e-05, "loss": 0.5012, "mean_token_accuracy": 0.8695060014724731, "num_tokens": 394103484.0, "step": 10375 }, { "epoch": 1.3199338506551328, "ewc_loss": 0.09814453125, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.350401878356934, "learning_rate": 1e-05, "loss": 0.4657, "mean_token_accuracy": 0.8749434947967529, "num_tokens": 394144302.0, "step": 10376 }, { "epoch": 1.3200610609337233, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 0.00011539459228515625, "grad_norm": 7.231921195983887, "learning_rate": 1e-05, "loss": 0.5162, "mean_token_accuracy": 0.8654999136924744, "num_tokens": 394179258.0, "step": 10377 }, { "epoch": 1.3201882712123139, "ewc_loss": 0.10595703125, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.586949825286865, "learning_rate": 1e-05, "loss": 0.5293, "mean_token_accuracy": 0.8588805198669434, "num_tokens": 394221498.0, "step": 10378 }, { "epoch": 1.3203154814909044, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.540130615234375, "learning_rate": 1e-05, "loss": 0.5019, "mean_token_accuracy": 0.8796485066413879, "num_tokens": 394255985.0, "step": 10379 }, { "epoch": 1.320442691769495, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.166232109069824, "learning_rate": 1e-05, "loss": 0.5404, "mean_token_accuracy": 0.855103611946106, "num_tokens": 394291020.0, "step": 10380 }, { "epoch": 1.3205699020480854, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 0.00011444091796875, "grad_norm": 7.293346405029297, "learning_rate": 1e-05, "loss": 0.5221, "mean_token_accuracy": 0.8650394082069397, "num_tokens": 394329415.0, "step": 10381 }, { "epoch": 1.320697112326676, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.433801174163818, "learning_rate": 1e-05, "loss": 0.5123, "mean_token_accuracy": 0.8677979111671448, "num_tokens": 394365287.0, "step": 10382 }, { "epoch": 1.3208243226052665, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.834201812744141, "learning_rate": 1e-05, "loss": 0.4992, "mean_token_accuracy": 0.8720695376396179, "num_tokens": 394403749.0, "step": 10383 }, { "epoch": 1.320951532883857, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.378901958465576, "learning_rate": 1e-05, "loss": 0.4431, "mean_token_accuracy": 0.8861396312713623, "num_tokens": 394445443.0, "step": 10384 }, { "epoch": 1.3210787431624476, "ewc_loss": 0.11474609375, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.412923336029053, "learning_rate": 1e-05, "loss": 0.4424, "mean_token_accuracy": 0.8847808837890625, "num_tokens": 394487650.0, "step": 10385 }, { "epoch": 1.321205953441038, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.285553932189941, "learning_rate": 1e-05, "loss": 0.4674, "mean_token_accuracy": 0.8781517744064331, "num_tokens": 394524606.0, "step": 10386 }, { "epoch": 1.3213331637196286, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 6.124035358428955, "learning_rate": 1e-05, "loss": 0.5123, "mean_token_accuracy": 0.8599787950515747, "num_tokens": 394561177.0, "step": 10387 }, { "epoch": 1.3214603739982191, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.214718818664551, "learning_rate": 1e-05, "loss": 0.467, "mean_token_accuracy": 0.8782951235771179, "num_tokens": 394605212.0, "step": 10388 }, { "epoch": 1.3215875842768097, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.931690692901611, "learning_rate": 1e-05, "loss": 0.494, "mean_token_accuracy": 0.8705604672431946, "num_tokens": 394637662.0, "step": 10389 }, { "epoch": 1.3217147945554002, "ewc_loss": 0.10888671875, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 6.180466651916504, "learning_rate": 1e-05, "loss": 0.464, "mean_token_accuracy": 0.8758963346481323, "num_tokens": 394675551.0, "step": 10390 }, { "epoch": 1.3218420048339907, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.797251224517822, "learning_rate": 1e-05, "loss": 0.5341, "mean_token_accuracy": 0.8521639108657837, "num_tokens": 394715225.0, "step": 10391 }, { "epoch": 1.321969215112581, "ewc_loss": 0.109375, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.172297954559326, "learning_rate": 1e-05, "loss": 0.4497, "mean_token_accuracy": 0.8782892823219299, "num_tokens": 394749872.0, "step": 10392 }, { "epoch": 1.3220964253911716, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.671072006225586, "learning_rate": 1e-05, "loss": 0.5302, "mean_token_accuracy": 0.8549195528030396, "num_tokens": 394791674.0, "step": 10393 }, { "epoch": 1.322223635669762, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.2617950439453125, "learning_rate": 1e-05, "loss": 0.5229, "mean_token_accuracy": 0.858987033367157, "num_tokens": 394829556.0, "step": 10394 }, { "epoch": 1.3223508459483526, "ewc_loss": 0.0986328125, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.533822536468506, "learning_rate": 1e-05, "loss": 0.47, "mean_token_accuracy": 0.8716493844985962, "num_tokens": 394864421.0, "step": 10395 }, { "epoch": 1.3224780562269431, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.467705726623535, "learning_rate": 1e-05, "loss": 0.4883, "mean_token_accuracy": 0.8725140690803528, "num_tokens": 394901008.0, "step": 10396 }, { "epoch": 1.3226052665055337, "ewc_loss": 0.09765625, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.391982555389404, "learning_rate": 1e-05, "loss": 0.58, "mean_token_accuracy": 0.8419009447097778, "num_tokens": 394947172.0, "step": 10397 }, { "epoch": 1.3227324767841242, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.846381664276123, "learning_rate": 1e-05, "loss": 0.5301, "mean_token_accuracy": 0.8607751727104187, "num_tokens": 394984515.0, "step": 10398 }, { "epoch": 1.3228596870627147, "ewc_loss": 0.09814453125, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.347390174865723, "learning_rate": 1e-05, "loss": 0.5162, "mean_token_accuracy": 0.8577775955200195, "num_tokens": 395019819.0, "step": 10399 }, { "epoch": 1.3229868973413053, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 0.00011873245239257812, "grad_norm": 7.331162929534912, "learning_rate": 1e-05, "loss": 0.5292, "mean_token_accuracy": 0.8660876750946045, "num_tokens": 395058606.0, "step": 10400 }, { "epoch": 1.3231141076198956, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 5.665741920471191, "learning_rate": 1e-05, "loss": 0.4541, "mean_token_accuracy": 0.8774643540382385, "num_tokens": 395094354.0, "step": 10401 }, { "epoch": 1.323241317898486, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 0.0001239776611328125, "grad_norm": 7.579576015472412, "learning_rate": 1e-05, "loss": 0.5275, "mean_token_accuracy": 0.8685321807861328, "num_tokens": 395136034.0, "step": 10402 }, { "epoch": 1.3233685281770766, "ewc_loss": 0.125, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.255702018737793, "learning_rate": 1e-05, "loss": 0.4612, "mean_token_accuracy": 0.8848028182983398, "num_tokens": 395176335.0, "step": 10403 }, { "epoch": 1.3234957384556671, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 0.00011396408081054688, "grad_norm": 7.265506267547607, "learning_rate": 1e-05, "loss": 0.4789, "mean_token_accuracy": 0.8798772096633911, "num_tokens": 395216297.0, "step": 10404 }, { "epoch": 1.3236229487342577, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.485877513885498, "learning_rate": 1e-05, "loss": 0.5079, "mean_token_accuracy": 0.8708953261375427, "num_tokens": 395255685.0, "step": 10405 }, { "epoch": 1.3237501590128482, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.784121990203857, "learning_rate": 1e-05, "loss": 0.5234, "mean_token_accuracy": 0.8606206178665161, "num_tokens": 395300484.0, "step": 10406 }, { "epoch": 1.3238773692914387, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.408902645111084, "learning_rate": 1e-05, "loss": 0.5346, "mean_token_accuracy": 0.8638712167739868, "num_tokens": 395333760.0, "step": 10407 }, { "epoch": 1.3240045795700293, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.37053108215332, "learning_rate": 1e-05, "loss": 0.5102, "mean_token_accuracy": 0.8668603897094727, "num_tokens": 395376264.0, "step": 10408 }, { "epoch": 1.3241317898486198, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.292207717895508, "learning_rate": 1e-05, "loss": 0.5405, "mean_token_accuracy": 0.8554103374481201, "num_tokens": 395417758.0, "step": 10409 }, { "epoch": 1.3242590001272103, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 6.081201076507568, "learning_rate": 1e-05, "loss": 0.5108, "mean_token_accuracy": 0.8626063466072083, "num_tokens": 395458657.0, "step": 10410 }, { "epoch": 1.3243862104058008, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.233569622039795, "learning_rate": 1e-05, "loss": 0.5419, "mean_token_accuracy": 0.8547320365905762, "num_tokens": 395500838.0, "step": 10411 }, { "epoch": 1.3245134206843914, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.891197204589844, "learning_rate": 1e-05, "loss": 0.5055, "mean_token_accuracy": 0.8663457632064819, "num_tokens": 395544912.0, "step": 10412 }, { "epoch": 1.324640630962982, "ewc_loss": 0.10986328125, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.197310924530029, "learning_rate": 1e-05, "loss": 0.493, "mean_token_accuracy": 0.8688949346542358, "num_tokens": 395584661.0, "step": 10413 }, { "epoch": 1.3247678412415724, "ewc_loss": 0.10205078125, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.76512336730957, "learning_rate": 1e-05, "loss": 0.4761, "mean_token_accuracy": 0.8673915863037109, "num_tokens": 395623374.0, "step": 10414 }, { "epoch": 1.324895051520163, "ewc_loss": 0.10986328125, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.2308149337768555, "learning_rate": 1e-05, "loss": 0.4558, "mean_token_accuracy": 0.878292977809906, "num_tokens": 395658854.0, "step": 10415 }, { "epoch": 1.3250222617987533, "ewc_loss": 0.10009765625, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.6371541023254395, "learning_rate": 1e-05, "loss": 0.4664, "mean_token_accuracy": 0.8742400407791138, "num_tokens": 395695298.0, "step": 10416 }, { "epoch": 1.3251494720773438, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.370955467224121, "learning_rate": 1e-05, "loss": 0.5657, "mean_token_accuracy": 0.8445793390274048, "num_tokens": 395731800.0, "step": 10417 }, { "epoch": 1.3252766823559343, "ewc_loss": 0.0986328125, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.487068176269531, "learning_rate": 1e-05, "loss": 0.5006, "mean_token_accuracy": 0.8582005500793457, "num_tokens": 395773811.0, "step": 10418 }, { "epoch": 1.3254038926345248, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.629940986633301, "learning_rate": 1e-05, "loss": 0.499, "mean_token_accuracy": 0.8724764585494995, "num_tokens": 395817280.0, "step": 10419 }, { "epoch": 1.3255311029131154, "ewc_loss": 0.09765625, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.374037265777588, "learning_rate": 1e-05, "loss": 0.5061, "mean_token_accuracy": 0.8603953719139099, "num_tokens": 395857950.0, "step": 10420 }, { "epoch": 1.325658313191706, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 7.063697338104248, "learning_rate": 1e-05, "loss": 0.5363, "mean_token_accuracy": 0.8583434820175171, "num_tokens": 395893048.0, "step": 10421 }, { "epoch": 1.3257855234702964, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.465810298919678, "learning_rate": 1e-05, "loss": 0.4765, "mean_token_accuracy": 0.873566746711731, "num_tokens": 395932678.0, "step": 10422 }, { "epoch": 1.325912733748887, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.475366115570068, "learning_rate": 1e-05, "loss": 0.5117, "mean_token_accuracy": 0.8718535304069519, "num_tokens": 395967253.0, "step": 10423 }, { "epoch": 1.3260399440274775, "ewc_loss": 0.11474609375, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 5.945777416229248, "learning_rate": 1e-05, "loss": 0.4451, "mean_token_accuracy": 0.8848080635070801, "num_tokens": 396000935.0, "step": 10424 }, { "epoch": 1.3261671543060678, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 0.00012063980102539062, "grad_norm": 7.452455997467041, "learning_rate": 1e-05, "loss": 0.5672, "mean_token_accuracy": 0.8532856702804565, "num_tokens": 396039148.0, "step": 10425 }, { "epoch": 1.3262943645846583, "ewc_loss": 0.125, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.383759498596191, "learning_rate": 1e-05, "loss": 0.4613, "mean_token_accuracy": 0.881783664226532, "num_tokens": 396072035.0, "step": 10426 }, { "epoch": 1.3264215748632489, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 7.0299072265625, "learning_rate": 1e-05, "loss": 0.5316, "mean_token_accuracy": 0.864628255367279, "num_tokens": 396112047.0, "step": 10427 }, { "epoch": 1.3265487851418394, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.420485973358154, "learning_rate": 1e-05, "loss": 0.5072, "mean_token_accuracy": 0.8673584461212158, "num_tokens": 396149844.0, "step": 10428 }, { "epoch": 1.32667599542043, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.591712951660156, "learning_rate": 1e-05, "loss": 0.4333, "mean_token_accuracy": 0.8905879259109497, "num_tokens": 396187493.0, "step": 10429 }, { "epoch": 1.3268032056990204, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.331237316131592, "learning_rate": 1e-05, "loss": 0.4998, "mean_token_accuracy": 0.8692871332168579, "num_tokens": 396220590.0, "step": 10430 }, { "epoch": 1.326930415977611, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.236839294433594, "learning_rate": 1e-05, "loss": 0.5033, "mean_token_accuracy": 0.8647345900535583, "num_tokens": 396262590.0, "step": 10431 }, { "epoch": 1.3270576262562015, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.239355087280273, "learning_rate": 1e-05, "loss": 0.4849, "mean_token_accuracy": 0.8720399141311646, "num_tokens": 396296491.0, "step": 10432 }, { "epoch": 1.327184836534792, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.995915412902832, "learning_rate": 1e-05, "loss": 0.4785, "mean_token_accuracy": 0.8711388111114502, "num_tokens": 396334798.0, "step": 10433 }, { "epoch": 1.3273120468133826, "ewc_loss": 0.109375, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.1880598068237305, "learning_rate": 1e-05, "loss": 0.4802, "mean_token_accuracy": 0.8658024668693542, "num_tokens": 396368278.0, "step": 10434 }, { "epoch": 1.327439257091973, "ewc_loss": 0.10302734375, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.846404075622559, "learning_rate": 1e-05, "loss": 0.4737, "mean_token_accuracy": 0.8716425895690918, "num_tokens": 396405034.0, "step": 10435 }, { "epoch": 1.3275664673705636, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.16495943069458, "learning_rate": 1e-05, "loss": 0.5407, "mean_token_accuracy": 0.8530747294425964, "num_tokens": 396445503.0, "step": 10436 }, { "epoch": 1.3276936776491541, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.743444442749023, "learning_rate": 1e-05, "loss": 0.4665, "mean_token_accuracy": 0.8739232420921326, "num_tokens": 396482462.0, "step": 10437 }, { "epoch": 1.3278208879277447, "ewc_loss": 0.10888671875, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 6.19378137588501, "learning_rate": 1e-05, "loss": 0.4277, "mean_token_accuracy": 0.8869783878326416, "num_tokens": 396519186.0, "step": 10438 }, { "epoch": 1.3279480982063352, "ewc_loss": 0.10009765625, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.639591693878174, "learning_rate": 1e-05, "loss": 0.4773, "mean_token_accuracy": 0.8677495121955872, "num_tokens": 396558996.0, "step": 10439 }, { "epoch": 1.3280753084849257, "ewc_loss": 0.11181640625, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.312569618225098, "learning_rate": 1e-05, "loss": 0.5504, "mean_token_accuracy": 0.84951251745224, "num_tokens": 396593158.0, "step": 10440 }, { "epoch": 1.328202518763516, "ewc_loss": 0.0986328125, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.493409156799316, "learning_rate": 1e-05, "loss": 0.5027, "mean_token_accuracy": 0.8597090244293213, "num_tokens": 396629189.0, "step": 10441 }, { "epoch": 1.3283297290421066, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.584911823272705, "learning_rate": 1e-05, "loss": 0.5538, "mean_token_accuracy": 0.8497153520584106, "num_tokens": 396665932.0, "step": 10442 }, { "epoch": 1.328456939320697, "ewc_loss": 0.0966796875, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.3544535636901855, "learning_rate": 1e-05, "loss": 0.5218, "mean_token_accuracy": 0.85426926612854, "num_tokens": 396706638.0, "step": 10443 }, { "epoch": 1.3285841495992876, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 6.99774169921875, "learning_rate": 1e-05, "loss": 0.4962, "mean_token_accuracy": 0.8728461861610413, "num_tokens": 396744354.0, "step": 10444 }, { "epoch": 1.3287113598778781, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.409923553466797, "learning_rate": 1e-05, "loss": 0.4773, "mean_token_accuracy": 0.8714537024497986, "num_tokens": 396783806.0, "step": 10445 }, { "epoch": 1.3288385701564687, "ewc_loss": 0.140625, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 0.00012111663818359375, "grad_norm": 7.439303874969482, "learning_rate": 1e-05, "loss": 0.521, "mean_token_accuracy": 0.8684971928596497, "num_tokens": 396822337.0, "step": 10446 }, { "epoch": 1.3289657804350592, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 5.859868049621582, "learning_rate": 1e-05, "loss": 0.551, "mean_token_accuracy": 0.8525646924972534, "num_tokens": 396861674.0, "step": 10447 }, { "epoch": 1.3290929907136497, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 0.0001220703125, "grad_norm": 7.511747360229492, "learning_rate": 1e-05, "loss": 0.5478, "mean_token_accuracy": 0.8630342483520508, "num_tokens": 396897686.0, "step": 10448 }, { "epoch": 1.3292202009922403, "ewc_loss": 0.125, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.361294746398926, "learning_rate": 1e-05, "loss": 0.5192, "mean_token_accuracy": 0.8601876497268677, "num_tokens": 396936261.0, "step": 10449 }, { "epoch": 1.3293474112708306, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 0.00011014938354492188, "grad_norm": 7.122215747833252, "learning_rate": 1e-05, "loss": 0.516, "mean_token_accuracy": 0.8625340461730957, "num_tokens": 396967452.0, "step": 10450 }, { "epoch": 1.329474621549421, "ewc_loss": 0.125, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.447803020477295, "learning_rate": 1e-05, "loss": 0.4617, "mean_token_accuracy": 0.8823602199554443, "num_tokens": 397004672.0, "step": 10451 }, { "epoch": 1.3296018318280116, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.683419227600098, "learning_rate": 1e-05, "loss": 0.5068, "mean_token_accuracy": 0.866291344165802, "num_tokens": 397039310.0, "step": 10452 }, { "epoch": 1.3297290421066021, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.342616558074951, "learning_rate": 1e-05, "loss": 0.4812, "mean_token_accuracy": 0.8765125870704651, "num_tokens": 397077465.0, "step": 10453 }, { "epoch": 1.3298562523851927, "ewc_loss": 0.11181640625, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.318140029907227, "learning_rate": 1e-05, "loss": 0.4799, "mean_token_accuracy": 0.8747085928916931, "num_tokens": 397108606.0, "step": 10454 }, { "epoch": 1.3299834626637832, "ewc_loss": 0.11474609375, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.2415452003479, "learning_rate": 1e-05, "loss": 0.5144, "mean_token_accuracy": 0.8639293909072876, "num_tokens": 397146628.0, "step": 10455 }, { "epoch": 1.3301106729423737, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 6.064763069152832, "learning_rate": 1e-05, "loss": 0.4864, "mean_token_accuracy": 0.868144154548645, "num_tokens": 397182345.0, "step": 10456 }, { "epoch": 1.3302378832209643, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.169265270233154, "learning_rate": 1e-05, "loss": 0.503, "mean_token_accuracy": 0.8690741062164307, "num_tokens": 397218673.0, "step": 10457 }, { "epoch": 1.3303650934995548, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.9069504737854, "learning_rate": 1e-05, "loss": 0.4719, "mean_token_accuracy": 0.8778655529022217, "num_tokens": 397250186.0, "step": 10458 }, { "epoch": 1.3304923037781453, "ewc_loss": 0.10986328125, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.125535488128662, "learning_rate": 1e-05, "loss": 0.4976, "mean_token_accuracy": 0.8670898675918579, "num_tokens": 397287947.0, "step": 10459 }, { "epoch": 1.3306195140567358, "ewc_loss": 0.10302734375, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.797660827636719, "learning_rate": 1e-05, "loss": 0.487, "mean_token_accuracy": 0.865963876247406, "num_tokens": 397328523.0, "step": 10460 }, { "epoch": 1.3307467243353264, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.139970779418945, "learning_rate": 1e-05, "loss": 0.47, "mean_token_accuracy": 0.8709543943405151, "num_tokens": 397366238.0, "step": 10461 }, { "epoch": 1.330873934613917, "ewc_loss": 0.099609375, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.68295955657959, "learning_rate": 1e-05, "loss": 0.4392, "mean_token_accuracy": 0.8831509351730347, "num_tokens": 397404324.0, "step": 10462 }, { "epoch": 1.3310011448925074, "ewc_loss": 0.109375, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.22839879989624, "learning_rate": 1e-05, "loss": 0.4842, "mean_token_accuracy": 0.8678387999534607, "num_tokens": 397441074.0, "step": 10463 }, { "epoch": 1.331128355171098, "ewc_loss": 0.0986328125, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.5475006103515625, "learning_rate": 1e-05, "loss": 0.4968, "mean_token_accuracy": 0.8565100431442261, "num_tokens": 397476913.0, "step": 10464 }, { "epoch": 1.3312555654496883, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.441898822784424, "learning_rate": 1e-05, "loss": 0.4704, "mean_token_accuracy": 0.8723243474960327, "num_tokens": 397507569.0, "step": 10465 }, { "epoch": 1.3313827757282788, "ewc_loss": 0.09765625, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.376857757568359, "learning_rate": 1e-05, "loss": 0.4586, "mean_token_accuracy": 0.8734480738639832, "num_tokens": 397544131.0, "step": 10466 }, { "epoch": 1.3315099860068693, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.8206706047058105, "learning_rate": 1e-05, "loss": 0.4675, "mean_token_accuracy": 0.8851779699325562, "num_tokens": 397585042.0, "step": 10467 }, { "epoch": 1.3316371962854598, "ewc_loss": 0.09765625, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.308409214019775, "learning_rate": 1e-05, "loss": 0.4929, "mean_token_accuracy": 0.8634271025657654, "num_tokens": 397620169.0, "step": 10468 }, { "epoch": 1.3317644065640504, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 0.00011873245239257812, "grad_norm": 7.3697190284729, "learning_rate": 1e-05, "loss": 0.5313, "mean_token_accuracy": 0.8640990853309631, "num_tokens": 397657349.0, "step": 10469 }, { "epoch": 1.331891616842641, "ewc_loss": 0.10791015625, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 5.634750843048096, "learning_rate": 1e-05, "loss": 0.5548, "mean_token_accuracy": 0.8517106175422668, "num_tokens": 397696309.0, "step": 10470 }, { "epoch": 1.3320188271212314, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 0.00012683868408203125, "grad_norm": 7.684443473815918, "learning_rate": 1e-05, "loss": 0.5142, "mean_token_accuracy": 0.8736982345581055, "num_tokens": 397734520.0, "step": 10471 }, { "epoch": 1.332146037399822, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.315563201904297, "learning_rate": 1e-05, "loss": 0.5337, "mean_token_accuracy": 0.860005795955658, "num_tokens": 397773078.0, "step": 10472 }, { "epoch": 1.3322732476784125, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 0.0001163482666015625, "grad_norm": 7.351629734039307, "learning_rate": 1e-05, "loss": 0.4817, "mean_token_accuracy": 0.8790097236633301, "num_tokens": 397814180.0, "step": 10473 }, { "epoch": 1.3324004579570028, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 0.00010967254638671875, "grad_norm": 6.54426383972168, "learning_rate": 1e-05, "loss": 0.476, "mean_token_accuracy": 0.877502977848053, "num_tokens": 397850618.0, "step": 10474 }, { "epoch": 1.3325276682355933, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.8783464431762695, "learning_rate": 1e-05, "loss": 0.5616, "mean_token_accuracy": 0.8529416918754578, "num_tokens": 397888538.0, "step": 10475 }, { "epoch": 1.3326548785141838, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.430668830871582, "learning_rate": 1e-05, "loss": 0.4831, "mean_token_accuracy": 0.8790342211723328, "num_tokens": 397917861.0, "step": 10476 }, { "epoch": 1.3327820887927744, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.489687442779541, "learning_rate": 1e-05, "loss": 0.4456, "mean_token_accuracy": 0.8852951526641846, "num_tokens": 397954052.0, "step": 10477 }, { "epoch": 1.332909299071365, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.271291255950928, "learning_rate": 1e-05, "loss": 0.4973, "mean_token_accuracy": 0.8675076961517334, "num_tokens": 397996458.0, "step": 10478 }, { "epoch": 1.3330365093499554, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.222400188446045, "learning_rate": 1e-05, "loss": 0.4682, "mean_token_accuracy": 0.8783053755760193, "num_tokens": 398034164.0, "step": 10479 }, { "epoch": 1.333163719628546, "ewc_loss": 0.11083984375, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.141477584838867, "learning_rate": 1e-05, "loss": 0.4682, "mean_token_accuracy": 0.878196656703949, "num_tokens": 398075828.0, "step": 10480 }, { "epoch": 1.3332909299071365, "ewc_loss": 0.10595703125, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 6.06854772567749, "learning_rate": 1e-05, "loss": 0.4856, "mean_token_accuracy": 0.8694733381271362, "num_tokens": 398112941.0, "step": 10481 }, { "epoch": 1.333418140185727, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 6.039146423339844, "learning_rate": 1e-05, "loss": 0.4619, "mean_token_accuracy": 0.8788648843765259, "num_tokens": 398146255.0, "step": 10482 }, { "epoch": 1.3335453504643175, "ewc_loss": 0.10595703125, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.98648738861084, "learning_rate": 1e-05, "loss": 0.4583, "mean_token_accuracy": 0.88069087266922, "num_tokens": 398178552.0, "step": 10483 }, { "epoch": 1.333672560742908, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.96419620513916, "learning_rate": 1e-05, "loss": 0.4614, "mean_token_accuracy": 0.8793690800666809, "num_tokens": 398216310.0, "step": 10484 }, { "epoch": 1.3337997710214986, "ewc_loss": 0.10498046875, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.93342924118042, "learning_rate": 1e-05, "loss": 0.4909, "mean_token_accuracy": 0.8684713840484619, "num_tokens": 398249756.0, "step": 10485 }, { "epoch": 1.3339269813000891, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.901841640472412, "learning_rate": 1e-05, "loss": 0.4913, "mean_token_accuracy": 0.8677653074264526, "num_tokens": 398285264.0, "step": 10486 }, { "epoch": 1.3340541915786797, "ewc_loss": 0.10498046875, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.915534019470215, "learning_rate": 1e-05, "loss": 0.5438, "mean_token_accuracy": 0.8515880703926086, "num_tokens": 398320964.0, "step": 10487 }, { "epoch": 1.3341814018572702, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.840637683868408, "learning_rate": 1e-05, "loss": 0.5453, "mean_token_accuracy": 0.8506578207015991, "num_tokens": 398355392.0, "step": 10488 }, { "epoch": 1.3343086121358605, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.920661449432373, "learning_rate": 1e-05, "loss": 0.4919, "mean_token_accuracy": 0.8671855330467224, "num_tokens": 398393551.0, "step": 10489 }, { "epoch": 1.334435822414451, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.771080493927002, "learning_rate": 1e-05, "loss": 0.5329, "mean_token_accuracy": 0.8536291718482971, "num_tokens": 398433812.0, "step": 10490 }, { "epoch": 1.3345630326930416, "ewc_loss": 0.10595703125, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.960385799407959, "learning_rate": 1e-05, "loss": 0.4884, "mean_token_accuracy": 0.8684595823287964, "num_tokens": 398469790.0, "step": 10491 }, { "epoch": 1.334690242971632, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.695912837982178, "learning_rate": 1e-05, "loss": 0.4731, "mean_token_accuracy": 0.8744435906410217, "num_tokens": 398512253.0, "step": 10492 }, { "epoch": 1.3348174532502226, "ewc_loss": 0.10693359375, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 6.066385746002197, "learning_rate": 1e-05, "loss": 0.5527, "mean_token_accuracy": 0.8480395078659058, "num_tokens": 398549790.0, "step": 10493 }, { "epoch": 1.3349446635288131, "ewc_loss": 0.0986328125, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.555894374847412, "learning_rate": 1e-05, "loss": 0.4293, "mean_token_accuracy": 0.8825511932373047, "num_tokens": 398586844.0, "step": 10494 }, { "epoch": 1.3350718738074037, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.277573585510254, "learning_rate": 1e-05, "loss": 0.5082, "mean_token_accuracy": 0.864986777305603, "num_tokens": 398624177.0, "step": 10495 }, { "epoch": 1.3351990840859942, "ewc_loss": 0.0966796875, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.368587970733643, "learning_rate": 1e-05, "loss": 0.4383, "mean_token_accuracy": 0.879926860332489, "num_tokens": 398661840.0, "step": 10496 }, { "epoch": 1.3353262943645847, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.683884143829346, "learning_rate": 1e-05, "loss": 0.465, "mean_token_accuracy": 0.8801127672195435, "num_tokens": 398699535.0, "step": 10497 }, { "epoch": 1.3354535046431752, "ewc_loss": 0.09326171875, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 7.390975952148438e-05, "grad_norm": 5.167121887207031, "learning_rate": 1e-05, "loss": 0.5022, "mean_token_accuracy": 0.859049916267395, "num_tokens": 398736294.0, "step": 10498 }, { "epoch": 1.3355807149217656, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 0.00011968612670898438, "grad_norm": 7.408328533172607, "learning_rate": 1e-05, "loss": 0.4988, "mean_token_accuracy": 0.8750532865524292, "num_tokens": 398776372.0, "step": 10499 }, { "epoch": 1.335707925200356, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.383339881896973, "learning_rate": 1e-05, "loss": 0.5086, "mean_token_accuracy": 0.8624734282493591, "num_tokens": 398816000.0, "step": 10500 }, { "epoch": 1.3358351354789466, "ewc_loss": 0.1572265625, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 0.0001373291015625, "grad_norm": 8.132237434387207, "learning_rate": 1e-05, "loss": 0.5182, "mean_token_accuracy": 0.8784252405166626, "num_tokens": 398851960.0, "step": 10501 }, { "epoch": 1.3359623457575371, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 0.00011301040649414062, "grad_norm": 6.486237049102783, "learning_rate": 1e-05, "loss": 0.533, "mean_token_accuracy": 0.865589439868927, "num_tokens": 398886485.0, "step": 10502 }, { "epoch": 1.3360895560361277, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 0.00012683868408203125, "grad_norm": 7.7961249351501465, "learning_rate": 1e-05, "loss": 0.5559, "mean_token_accuracy": 0.860590934753418, "num_tokens": 398923800.0, "step": 10503 }, { "epoch": 1.3362167663147182, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 0.00012063980102539062, "grad_norm": 6.959124565124512, "learning_rate": 1e-05, "loss": 0.5092, "mean_token_accuracy": 0.8720460534095764, "num_tokens": 398963578.0, "step": 10504 }, { "epoch": 1.3363439765933087, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 0.00011157989501953125, "grad_norm": 7.162717342376709, "learning_rate": 1e-05, "loss": 0.5074, "mean_token_accuracy": 0.870895266532898, "num_tokens": 399009791.0, "step": 10505 }, { "epoch": 1.3364711868718993, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 6.673568248748779, "learning_rate": 1e-05, "loss": 0.4784, "mean_token_accuracy": 0.8778283596038818, "num_tokens": 399048701.0, "step": 10506 }, { "epoch": 1.3365983971504898, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.819836139678955, "learning_rate": 1e-05, "loss": 0.4706, "mean_token_accuracy": 0.8783082962036133, "num_tokens": 399083441.0, "step": 10507 }, { "epoch": 1.3367256074290803, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.368528842926025, "learning_rate": 1e-05, "loss": 0.4881, "mean_token_accuracy": 0.8754448890686035, "num_tokens": 399113762.0, "step": 10508 }, { "epoch": 1.3368528177076708, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.611623764038086, "learning_rate": 1e-05, "loss": 0.4696, "mean_token_accuracy": 0.8774454593658447, "num_tokens": 399149966.0, "step": 10509 }, { "epoch": 1.3369800279862614, "ewc_loss": 0.10986328125, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.113503456115723, "learning_rate": 1e-05, "loss": 0.4978, "mean_token_accuracy": 0.8666289448738098, "num_tokens": 399188777.0, "step": 10510 }, { "epoch": 1.337107238264852, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.53297233581543, "learning_rate": 1e-05, "loss": 0.4673, "mean_token_accuracy": 0.8778994679450989, "num_tokens": 399229618.0, "step": 10511 }, { "epoch": 1.3372344485434424, "ewc_loss": 0.10400390625, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.887918472290039, "learning_rate": 1e-05, "loss": 0.4678, "mean_token_accuracy": 0.8759063482284546, "num_tokens": 399270558.0, "step": 10512 }, { "epoch": 1.337361658822033, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.6178364753723145, "learning_rate": 1e-05, "loss": 0.6015, "mean_token_accuracy": 0.8385205268859863, "num_tokens": 399310660.0, "step": 10513 }, { "epoch": 1.3374888691006233, "ewc_loss": 0.10107421875, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.704976558685303, "learning_rate": 1e-05, "loss": 0.5008, "mean_token_accuracy": 0.8657419681549072, "num_tokens": 399347838.0, "step": 10514 }, { "epoch": 1.3376160793792138, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.851468086242676, "learning_rate": 1e-05, "loss": 0.4913, "mean_token_accuracy": 0.8745154142379761, "num_tokens": 399384790.0, "step": 10515 }, { "epoch": 1.3377432896578043, "ewc_loss": 0.099609375, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.621244430541992, "learning_rate": 1e-05, "loss": 0.4528, "mean_token_accuracy": 0.8769651651382446, "num_tokens": 399425479.0, "step": 10516 }, { "epoch": 1.3378704999363948, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 7.196554183959961, "learning_rate": 1e-05, "loss": 0.5048, "mean_token_accuracy": 0.8706606030464172, "num_tokens": 399467161.0, "step": 10517 }, { "epoch": 1.3379977102149854, "ewc_loss": 0.10498046875, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.858569622039795, "learning_rate": 1e-05, "loss": 0.5291, "mean_token_accuracy": 0.8591426610946655, "num_tokens": 399500204.0, "step": 10518 }, { "epoch": 1.338124920493576, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 0.00011348724365234375, "grad_norm": 7.30574893951416, "learning_rate": 1e-05, "loss": 0.559, "mean_token_accuracy": 0.8533445000648499, "num_tokens": 399546142.0, "step": 10519 }, { "epoch": 1.3382521307721664, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.248964309692383, "learning_rate": 1e-05, "loss": 0.5583, "mean_token_accuracy": 0.8510076403617859, "num_tokens": 399582446.0, "step": 10520 }, { "epoch": 1.338379341050757, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.984710693359375, "learning_rate": 1e-05, "loss": 0.6038, "mean_token_accuracy": 0.840313196182251, "num_tokens": 399616998.0, "step": 10521 }, { "epoch": 1.3385065513293475, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.445096969604492, "learning_rate": 1e-05, "loss": 0.4915, "mean_token_accuracy": 0.8720373511314392, "num_tokens": 399648649.0, "step": 10522 }, { "epoch": 1.3386337616079378, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.417700290679932, "learning_rate": 1e-05, "loss": 0.4474, "mean_token_accuracy": 0.8828791975975037, "num_tokens": 399690417.0, "step": 10523 }, { "epoch": 1.3387609718865283, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.466985702514648, "learning_rate": 1e-05, "loss": 0.5189, "mean_token_accuracy": 0.8618802428245544, "num_tokens": 399732346.0, "step": 10524 }, { "epoch": 1.3388881821651188, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.9125165939331055, "learning_rate": 1e-05, "loss": 0.5132, "mean_token_accuracy": 0.8584802746772766, "num_tokens": 399770516.0, "step": 10525 }, { "epoch": 1.3390153924437094, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.583191871643066, "learning_rate": 1e-05, "loss": 0.5053, "mean_token_accuracy": 0.8670445680618286, "num_tokens": 399803266.0, "step": 10526 }, { "epoch": 1.3391426027223, "ewc_loss": 0.09765625, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.545769214630127, "learning_rate": 1e-05, "loss": 0.4979, "mean_token_accuracy": 0.866727352142334, "num_tokens": 399841940.0, "step": 10527 }, { "epoch": 1.3392698130008904, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.908255577087402, "learning_rate": 1e-05, "loss": 0.5389, "mean_token_accuracy": 0.8571118116378784, "num_tokens": 399880207.0, "step": 10528 }, { "epoch": 1.339397023279481, "ewc_loss": 0.0966796875, "ewc_loss_diag": 0.000194549560546875, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.435216426849365, "learning_rate": 1e-05, "loss": 0.5117, "mean_token_accuracy": 0.8603878617286682, "num_tokens": 399915731.0, "step": 10529 }, { "epoch": 1.3395242335580715, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 0.00011730194091796875, "grad_norm": 7.3572773933410645, "learning_rate": 1e-05, "loss": 0.4426, "mean_token_accuracy": 0.8927996158599854, "num_tokens": 399953398.0, "step": 10530 }, { "epoch": 1.339651443836662, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.7727742195129395, "learning_rate": 1e-05, "loss": 0.5397, "mean_token_accuracy": 0.8510749340057373, "num_tokens": 399991132.0, "step": 10531 }, { "epoch": 1.3397786541152525, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 0.00012063980102539062, "grad_norm": 7.528411865234375, "learning_rate": 1e-05, "loss": 0.5537, "mean_token_accuracy": 0.8610564470291138, "num_tokens": 400031615.0, "step": 10532 }, { "epoch": 1.339905864393843, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.328396797180176, "learning_rate": 1e-05, "loss": 0.548, "mean_token_accuracy": 0.8501391410827637, "num_tokens": 400070804.0, "step": 10533 }, { "epoch": 1.3400330746724336, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 7.155856132507324, "learning_rate": 1e-05, "loss": 0.5464, "mean_token_accuracy": 0.8613383769989014, "num_tokens": 400114021.0, "step": 10534 }, { "epoch": 1.3401602849510241, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.471307277679443, "learning_rate": 1e-05, "loss": 0.5041, "mean_token_accuracy": 0.8720096349716187, "num_tokens": 400156526.0, "step": 10535 }, { "epoch": 1.3402874952296147, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.714128494262695, "learning_rate": 1e-05, "loss": 0.4415, "mean_token_accuracy": 0.8874894976615906, "num_tokens": 400188487.0, "step": 10536 }, { "epoch": 1.3404147055082052, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.357853889465332, "learning_rate": 1e-05, "loss": 0.5788, "mean_token_accuracy": 0.8461586236953735, "num_tokens": 400222223.0, "step": 10537 }, { "epoch": 1.3405419157867955, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.3824968338012695, "learning_rate": 1e-05, "loss": 0.4452, "mean_token_accuracy": 0.8844956159591675, "num_tokens": 400259181.0, "step": 10538 }, { "epoch": 1.340669126065386, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.170010089874268, "learning_rate": 1e-05, "loss": 0.4497, "mean_token_accuracy": 0.8795897960662842, "num_tokens": 400295202.0, "step": 10539 }, { "epoch": 1.3407963363439765, "ewc_loss": 0.109375, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.237431526184082, "learning_rate": 1e-05, "loss": 0.4582, "mean_token_accuracy": 0.8808868527412415, "num_tokens": 400330936.0, "step": 10540 }, { "epoch": 1.340923546622567, "ewc_loss": 0.10498046875, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.99311637878418, "learning_rate": 1e-05, "loss": 0.4871, "mean_token_accuracy": 0.870021641254425, "num_tokens": 400367971.0, "step": 10541 }, { "epoch": 1.3410507569011576, "ewc_loss": 0.10986328125, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.2215895652771, "learning_rate": 1e-05, "loss": 0.5161, "mean_token_accuracy": 0.8639200925827026, "num_tokens": 400406868.0, "step": 10542 }, { "epoch": 1.3411779671797481, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.80234432220459, "learning_rate": 1e-05, "loss": 0.5329, "mean_token_accuracy": 0.8533327579498291, "num_tokens": 400442907.0, "step": 10543 }, { "epoch": 1.3413051774583387, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.325087070465088, "learning_rate": 1e-05, "loss": 0.5366, "mean_token_accuracy": 0.856075644493103, "num_tokens": 400479485.0, "step": 10544 }, { "epoch": 1.3414323877369292, "ewc_loss": 0.0986328125, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.627225875854492, "learning_rate": 1e-05, "loss": 0.4912, "mean_token_accuracy": 0.8646281957626343, "num_tokens": 400513790.0, "step": 10545 }, { "epoch": 1.3415595980155197, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.541220188140869, "learning_rate": 1e-05, "loss": 0.4973, "mean_token_accuracy": 0.8675620555877686, "num_tokens": 400557638.0, "step": 10546 }, { "epoch": 1.3416868082941102, "ewc_loss": 0.09765625, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.472957134246826, "learning_rate": 1e-05, "loss": 0.4482, "mean_token_accuracy": 0.8797445297241211, "num_tokens": 400594121.0, "step": 10547 }, { "epoch": 1.3418140185727006, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.929019927978516, "learning_rate": 1e-05, "loss": 0.5076, "mean_token_accuracy": 0.8657246828079224, "num_tokens": 400632427.0, "step": 10548 }, { "epoch": 1.341941228851291, "ewc_loss": 0.099609375, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.502394199371338, "learning_rate": 1e-05, "loss": 0.4655, "mean_token_accuracy": 0.8724969029426575, "num_tokens": 400666938.0, "step": 10549 }, { "epoch": 1.3420684391298816, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 0.0001163482666015625, "grad_norm": 7.314881801605225, "learning_rate": 1e-05, "loss": 0.4785, "mean_token_accuracy": 0.8813944458961487, "num_tokens": 400703510.0, "step": 10550 }, { "epoch": 1.3421956494084721, "ewc_loss": 0.10888671875, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 5.878347873687744, "learning_rate": 1e-05, "loss": 0.4754, "mean_token_accuracy": 0.8726033568382263, "num_tokens": 400737082.0, "step": 10551 }, { "epoch": 1.3423228596870627, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 0.00011539459228515625, "grad_norm": 7.305830955505371, "learning_rate": 1e-05, "loss": 0.5249, "mean_token_accuracy": 0.8682233691215515, "num_tokens": 400787254.0, "step": 10552 }, { "epoch": 1.3424500699656532, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.308856010437012, "learning_rate": 1e-05, "loss": 0.5033, "mean_token_accuracy": 0.8712693452835083, "num_tokens": 400824384.0, "step": 10553 }, { "epoch": 1.3425772802442437, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.8836774826049805, "learning_rate": 1e-05, "loss": 0.5098, "mean_token_accuracy": 0.8628000020980835, "num_tokens": 400859536.0, "step": 10554 }, { "epoch": 1.3427044905228342, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.411901950836182, "learning_rate": 1e-05, "loss": 0.4621, "mean_token_accuracy": 0.8801442384719849, "num_tokens": 400896022.0, "step": 10555 }, { "epoch": 1.3428317008014248, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.357431411743164, "learning_rate": 1e-05, "loss": 0.4652, "mean_token_accuracy": 0.8764336705207825, "num_tokens": 400936291.0, "step": 10556 }, { "epoch": 1.3429589110800153, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.415821075439453, "learning_rate": 1e-05, "loss": 0.4612, "mean_token_accuracy": 0.878160834312439, "num_tokens": 400980749.0, "step": 10557 }, { "epoch": 1.3430861213586058, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.91686487197876, "learning_rate": 1e-05, "loss": 0.5467, "mean_token_accuracy": 0.8510111570358276, "num_tokens": 401016591.0, "step": 10558 }, { "epoch": 1.3432133316371964, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.52588415145874, "learning_rate": 1e-05, "loss": 0.5183, "mean_token_accuracy": 0.8671547174453735, "num_tokens": 401062520.0, "step": 10559 }, { "epoch": 1.343340541915787, "ewc_loss": 0.09912109375, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.614555835723877, "learning_rate": 1e-05, "loss": 0.4721, "mean_token_accuracy": 0.8734554648399353, "num_tokens": 401105968.0, "step": 10560 }, { "epoch": 1.3434677521943774, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.778996467590332, "learning_rate": 1e-05, "loss": 0.4753, "mean_token_accuracy": 0.8832470178604126, "num_tokens": 401145290.0, "step": 10561 }, { "epoch": 1.343594962472968, "ewc_loss": 0.0986328125, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.517086029052734, "learning_rate": 1e-05, "loss": 0.4572, "mean_token_accuracy": 0.8783968687057495, "num_tokens": 401181253.0, "step": 10562 }, { "epoch": 1.3437221727515583, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 7.108598709106445, "learning_rate": 1e-05, "loss": 0.4991, "mean_token_accuracy": 0.8703420162200928, "num_tokens": 401218191.0, "step": 10563 }, { "epoch": 1.3438493830301488, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.694250106811523, "learning_rate": 1e-05, "loss": 0.4697, "mean_token_accuracy": 0.8684090971946716, "num_tokens": 401252049.0, "step": 10564 }, { "epoch": 1.3439765933087393, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 0.00011396408081054688, "grad_norm": 7.266988754272461, "learning_rate": 1e-05, "loss": 0.5781, "mean_token_accuracy": 0.8470816612243652, "num_tokens": 401294819.0, "step": 10565 }, { "epoch": 1.3441038035873298, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.043692588806152, "learning_rate": 1e-05, "loss": 0.4798, "mean_token_accuracy": 0.8722066283226013, "num_tokens": 401332007.0, "step": 10566 }, { "epoch": 1.3442310138659204, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 7.093881130218506, "learning_rate": 1e-05, "loss": 0.4817, "mean_token_accuracy": 0.8784530162811279, "num_tokens": 401371442.0, "step": 10567 }, { "epoch": 1.344358224144511, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.231738567352295, "learning_rate": 1e-05, "loss": 0.4747, "mean_token_accuracy": 0.8739211559295654, "num_tokens": 401409956.0, "step": 10568 }, { "epoch": 1.3444854344231014, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.767834186553955, "learning_rate": 1e-05, "loss": 0.4759, "mean_token_accuracy": 0.8772096037864685, "num_tokens": 401446445.0, "step": 10569 }, { "epoch": 1.344612644701692, "ewc_loss": 0.11474609375, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.214335918426514, "learning_rate": 1e-05, "loss": 0.534, "mean_token_accuracy": 0.8575692772865295, "num_tokens": 401489366.0, "step": 10570 }, { "epoch": 1.3447398549802825, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.459993362426758, "learning_rate": 1e-05, "loss": 0.5163, "mean_token_accuracy": 0.8633459210395813, "num_tokens": 401533326.0, "step": 10571 }, { "epoch": 1.3448670652588728, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.120750904083252, "learning_rate": 1e-05, "loss": 0.4763, "mean_token_accuracy": 0.8742259740829468, "num_tokens": 401571607.0, "step": 10572 }, { "epoch": 1.3449942755374633, "ewc_loss": 0.11083984375, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.282970428466797, "learning_rate": 1e-05, "loss": 0.5563, "mean_token_accuracy": 0.8491659164428711, "num_tokens": 401607930.0, "step": 10573 }, { "epoch": 1.3451214858160538, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.984978199005127, "learning_rate": 1e-05, "loss": 0.4437, "mean_token_accuracy": 0.8819061517715454, "num_tokens": 401645841.0, "step": 10574 }, { "epoch": 1.3452486960946444, "ewc_loss": 0.10986328125, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.221513271331787, "learning_rate": 1e-05, "loss": 0.5178, "mean_token_accuracy": 0.8603612780570984, "num_tokens": 401680795.0, "step": 10575 }, { "epoch": 1.345375906373235, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.832855224609375, "learning_rate": 1e-05, "loss": 0.4845, "mean_token_accuracy": 0.8692044019699097, "num_tokens": 401719710.0, "step": 10576 }, { "epoch": 1.3455031166518254, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.2775115966796875, "learning_rate": 1e-05, "loss": 0.4899, "mean_token_accuracy": 0.8710399270057678, "num_tokens": 401758956.0, "step": 10577 }, { "epoch": 1.345630326930416, "ewc_loss": 0.09912109375, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.663151741027832, "learning_rate": 1e-05, "loss": 0.5119, "mean_token_accuracy": 0.859285831451416, "num_tokens": 401798108.0, "step": 10578 }, { "epoch": 1.3457575372090065, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.460886001586914, "learning_rate": 1e-05, "loss": 0.4692, "mean_token_accuracy": 0.8761695027351379, "num_tokens": 401833495.0, "step": 10579 }, { "epoch": 1.345884747487597, "ewc_loss": 0.09765625, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.513560771942139, "learning_rate": 1e-05, "loss": 0.5016, "mean_token_accuracy": 0.8656173944473267, "num_tokens": 401875399.0, "step": 10580 }, { "epoch": 1.3460119577661875, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.767922401428223, "learning_rate": 1e-05, "loss": 0.5031, "mean_token_accuracy": 0.870449423789978, "num_tokens": 401915540.0, "step": 10581 }, { "epoch": 1.346139168044778, "ewc_loss": 0.0986328125, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.484258651733398, "learning_rate": 1e-05, "loss": 0.5262, "mean_token_accuracy": 0.8536906242370605, "num_tokens": 401960013.0, "step": 10582 }, { "epoch": 1.3462663783233686, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 0.00011110305786132812, "grad_norm": 7.132961273193359, "learning_rate": 1e-05, "loss": 0.5501, "mean_token_accuracy": 0.8571164011955261, "num_tokens": 401990709.0, "step": 10583 }, { "epoch": 1.3463935886019591, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.735728740692139, "learning_rate": 1e-05, "loss": 0.4671, "mean_token_accuracy": 0.8741194605827332, "num_tokens": 402031277.0, "step": 10584 }, { "epoch": 1.3465207988805497, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 0.00011348724365234375, "grad_norm": 7.231433868408203, "learning_rate": 1e-05, "loss": 0.51, "mean_token_accuracy": 0.872123122215271, "num_tokens": 402073248.0, "step": 10585 }, { "epoch": 1.3466480091591402, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.131076335906982, "learning_rate": 1e-05, "loss": 0.5296, "mean_token_accuracy": 0.8647850751876831, "num_tokens": 402116072.0, "step": 10586 }, { "epoch": 1.3467752194377305, "ewc_loss": 0.125, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.9117279052734375, "learning_rate": 1e-05, "loss": 0.4989, "mean_token_accuracy": 0.8691971898078918, "num_tokens": 402153092.0, "step": 10587 }, { "epoch": 1.346902429716321, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.3095598220825195, "learning_rate": 1e-05, "loss": 0.4876, "mean_token_accuracy": 0.8752868175506592, "num_tokens": 402192959.0, "step": 10588 }, { "epoch": 1.3470296399949115, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.409253120422363, "learning_rate": 1e-05, "loss": 0.4983, "mean_token_accuracy": 0.8676759004592896, "num_tokens": 402232245.0, "step": 10589 }, { "epoch": 1.347156850273502, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.370271682739258, "learning_rate": 1e-05, "loss": 0.4604, "mean_token_accuracy": 0.8804083466529846, "num_tokens": 402265862.0, "step": 10590 }, { "epoch": 1.3472840605520926, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.928997993469238, "learning_rate": 1e-05, "loss": 0.4625, "mean_token_accuracy": 0.8757919073104858, "num_tokens": 402303747.0, "step": 10591 }, { "epoch": 1.3474112708306831, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.5270280838012695, "learning_rate": 1e-05, "loss": 0.4858, "mean_token_accuracy": 0.8722784519195557, "num_tokens": 402342952.0, "step": 10592 }, { "epoch": 1.3475384811092737, "ewc_loss": 0.099609375, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.595613479614258, "learning_rate": 1e-05, "loss": 0.5639, "mean_token_accuracy": 0.8436239361763, "num_tokens": 402382388.0, "step": 10593 }, { "epoch": 1.3476656913878642, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.811645984649658, "learning_rate": 1e-05, "loss": 0.4853, "mean_token_accuracy": 0.874224066734314, "num_tokens": 402418937.0, "step": 10594 }, { "epoch": 1.3477929016664547, "ewc_loss": 0.09912109375, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.496364116668701, "learning_rate": 1e-05, "loss": 0.5119, "mean_token_accuracy": 0.858842134475708, "num_tokens": 402454347.0, "step": 10595 }, { "epoch": 1.3479201119450452, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 7.160500526428223, "learning_rate": 1e-05, "loss": 0.5795, "mean_token_accuracy": 0.8523283004760742, "num_tokens": 402490887.0, "step": 10596 }, { "epoch": 1.3480473222236355, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.726531982421875, "learning_rate": 1e-05, "loss": 0.5318, "mean_token_accuracy": 0.8619683384895325, "num_tokens": 402531293.0, "step": 10597 }, { "epoch": 1.348174532502226, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 0.00011301040649414062, "grad_norm": 7.284963607788086, "learning_rate": 1e-05, "loss": 0.5651, "mean_token_accuracy": 0.8572225570678711, "num_tokens": 402574396.0, "step": 10598 }, { "epoch": 1.3483017427808166, "ewc_loss": 0.11376953125, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.081424236297607, "learning_rate": 1e-05, "loss": 0.4931, "mean_token_accuracy": 0.8674129247665405, "num_tokens": 402615609.0, "step": 10599 }, { "epoch": 1.3484289530594071, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 7.088830471038818, "learning_rate": 1e-05, "loss": 0.5242, "mean_token_accuracy": 0.8639638423919678, "num_tokens": 402650627.0, "step": 10600 }, { "epoch": 1.3485561633379977, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.24555778503418, "learning_rate": 1e-05, "loss": 0.5148, "mean_token_accuracy": 0.8643684387207031, "num_tokens": 402693036.0, "step": 10601 }, { "epoch": 1.3486833736165882, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.761228561401367, "learning_rate": 1e-05, "loss": 0.4696, "mean_token_accuracy": 0.8794431090354919, "num_tokens": 402730748.0, "step": 10602 }, { "epoch": 1.3488105838951787, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.226768493652344, "learning_rate": 1e-05, "loss": 0.5363, "mean_token_accuracy": 0.8573845624923706, "num_tokens": 402762642.0, "step": 10603 }, { "epoch": 1.3489377941737692, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.00019550323486328125, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.463509559631348, "learning_rate": 1e-05, "loss": 0.5218, "mean_token_accuracy": 0.8612425327301025, "num_tokens": 402794265.0, "step": 10604 }, { "epoch": 1.3490650044523598, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.113498210906982, "learning_rate": 1e-05, "loss": 0.5673, "mean_token_accuracy": 0.848711371421814, "num_tokens": 402830609.0, "step": 10605 }, { "epoch": 1.3491922147309503, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.285277843475342, "learning_rate": 1e-05, "loss": 0.4608, "mean_token_accuracy": 0.874868631362915, "num_tokens": 402866608.0, "step": 10606 }, { "epoch": 1.3493194250095408, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.9671406745910645, "learning_rate": 1e-05, "loss": 0.5146, "mean_token_accuracy": 0.8613015413284302, "num_tokens": 402904030.0, "step": 10607 }, { "epoch": 1.3494466352881314, "ewc_loss": 0.10986328125, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.248664855957031, "learning_rate": 1e-05, "loss": 0.492, "mean_token_accuracy": 0.8675469160079956, "num_tokens": 402939552.0, "step": 10608 }, { "epoch": 1.3495738455667219, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.7999958992004395, "learning_rate": 1e-05, "loss": 0.4586, "mean_token_accuracy": 0.8753408789634705, "num_tokens": 402981683.0, "step": 10609 }, { "epoch": 1.3497010558453124, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.338840484619141, "learning_rate": 1e-05, "loss": 0.5553, "mean_token_accuracy": 0.8555721044540405, "num_tokens": 403019497.0, "step": 10610 }, { "epoch": 1.349828266123903, "ewc_loss": 0.099609375, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.647021770477295, "learning_rate": 1e-05, "loss": 0.4759, "mean_token_accuracy": 0.8670535683631897, "num_tokens": 403053415.0, "step": 10611 }, { "epoch": 1.3499554764024932, "ewc_loss": 0.11474609375, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.516139507293701, "learning_rate": 1e-05, "loss": 0.4975, "mean_token_accuracy": 0.8696590662002563, "num_tokens": 403089676.0, "step": 10612 }, { "epoch": 1.3500826866810838, "ewc_loss": 0.0986328125, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.530446529388428, "learning_rate": 1e-05, "loss": 0.4911, "mean_token_accuracy": 0.8649330139160156, "num_tokens": 403130429.0, "step": 10613 }, { "epoch": 1.3502098969596743, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.8144989013671875, "learning_rate": 1e-05, "loss": 0.4664, "mean_token_accuracy": 0.8818597793579102, "num_tokens": 403171420.0, "step": 10614 }, { "epoch": 1.3503371072382648, "ewc_loss": 0.10009765625, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.537996768951416, "learning_rate": 1e-05, "loss": 0.4919, "mean_token_accuracy": 0.8661633729934692, "num_tokens": 403210700.0, "step": 10615 }, { "epoch": 1.3504643175168554, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 7.0984601974487305, "learning_rate": 1e-05, "loss": 0.5106, "mean_token_accuracy": 0.8685455322265625, "num_tokens": 403249572.0, "step": 10616 }, { "epoch": 1.350591527795446, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 5.788411617279053, "learning_rate": 1e-05, "loss": 0.5037, "mean_token_accuracy": 0.8620661497116089, "num_tokens": 403288176.0, "step": 10617 }, { "epoch": 1.3507187380740364, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 0.00011157989501953125, "grad_norm": 7.119592666625977, "learning_rate": 1e-05, "loss": 0.4752, "mean_token_accuracy": 0.881548285484314, "num_tokens": 403329403.0, "step": 10618 }, { "epoch": 1.350845948352627, "ewc_loss": 0.11474609375, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.120515823364258, "learning_rate": 1e-05, "loss": 0.515, "mean_token_accuracy": 0.8657693266868591, "num_tokens": 403362276.0, "step": 10619 }, { "epoch": 1.3509731586312175, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.00019359588623046875, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.784223556518555, "learning_rate": 1e-05, "loss": 0.509, "mean_token_accuracy": 0.8676732182502747, "num_tokens": 403397797.0, "step": 10620 }, { "epoch": 1.3511003689098078, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.000194549560546875, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.270106792449951, "learning_rate": 1e-05, "loss": 0.4956, "mean_token_accuracy": 0.8674927949905396, "num_tokens": 403437943.0, "step": 10621 }, { "epoch": 1.3512275791883983, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.000194549560546875, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.294336318969727, "learning_rate": 1e-05, "loss": 0.4855, "mean_token_accuracy": 0.8742722868919373, "num_tokens": 403473575.0, "step": 10622 }, { "epoch": 1.3513547894669888, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.000194549560546875, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.369070053100586, "learning_rate": 1e-05, "loss": 0.5162, "mean_token_accuracy": 0.8632604479789734, "num_tokens": 403504720.0, "step": 10623 }, { "epoch": 1.3514819997455794, "ewc_loss": 0.10205078125, "ewc_loss_diag": 0.000194549560546875, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.820767402648926, "learning_rate": 1e-05, "loss": 0.5145, "mean_token_accuracy": 0.8580600619316101, "num_tokens": 403549244.0, "step": 10624 }, { "epoch": 1.35160921002417, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.000194549560546875, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.556503772735596, "learning_rate": 1e-05, "loss": 0.562, "mean_token_accuracy": 0.8512846231460571, "num_tokens": 403582881.0, "step": 10625 }, { "epoch": 1.3517364203027604, "ewc_loss": 0.0986328125, "ewc_loss_diag": 0.000194549560546875, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.512471675872803, "learning_rate": 1e-05, "loss": 0.4485, "mean_token_accuracy": 0.8794695138931274, "num_tokens": 403623477.0, "step": 10626 }, { "epoch": 1.351863630581351, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.000194549560546875, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.915024757385254, "learning_rate": 1e-05, "loss": 0.553, "mean_token_accuracy": 0.8539639711380005, "num_tokens": 403662399.0, "step": 10627 }, { "epoch": 1.3519908408599415, "ewc_loss": 0.099609375, "ewc_loss_diag": 0.000194549560546875, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.46617317199707, "learning_rate": 1e-05, "loss": 0.4809, "mean_token_accuracy": 0.8680218458175659, "num_tokens": 403703908.0, "step": 10628 }, { "epoch": 1.352118051138532, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.000194549560546875, "ewc_loss_parallel": 0.00011587142944335938, "grad_norm": 7.303552150726318, "learning_rate": 1e-05, "loss": 0.5117, "mean_token_accuracy": 0.8709237575531006, "num_tokens": 403746385.0, "step": 10629 }, { "epoch": 1.3522452614171225, "ewc_loss": 0.10986328125, "ewc_loss_diag": 0.000194549560546875, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 5.809568405151367, "learning_rate": 1e-05, "loss": 0.4797, "mean_token_accuracy": 0.8720006942749023, "num_tokens": 403785728.0, "step": 10630 }, { "epoch": 1.352372471695713, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.000194549560546875, "ewc_loss_parallel": 0.00011730194091796875, "grad_norm": 7.377498626708984, "learning_rate": 1e-05, "loss": 0.5046, "mean_token_accuracy": 0.8755609393119812, "num_tokens": 403827991.0, "step": 10631 }, { "epoch": 1.3524996819743036, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.000194549560546875, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.240041732788086, "learning_rate": 1e-05, "loss": 0.5094, "mean_token_accuracy": 0.8641690015792847, "num_tokens": 403861963.0, "step": 10632 }, { "epoch": 1.3526268922528941, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.000194549560546875, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 7.047056198120117, "learning_rate": 1e-05, "loss": 0.4566, "mean_token_accuracy": 0.8827473521232605, "num_tokens": 403899744.0, "step": 10633 }, { "epoch": 1.3527541025314846, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.000194549560546875, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.334519863128662, "learning_rate": 1e-05, "loss": 0.4463, "mean_token_accuracy": 0.8858727812767029, "num_tokens": 403939482.0, "step": 10634 }, { "epoch": 1.3528813128100752, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.000194549560546875, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.669737815856934, "learning_rate": 1e-05, "loss": 0.4578, "mean_token_accuracy": 0.8821466565132141, "num_tokens": 403976233.0, "step": 10635 }, { "epoch": 1.3530085230886655, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.000194549560546875, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.260533332824707, "learning_rate": 1e-05, "loss": 0.4721, "mean_token_accuracy": 0.8788747787475586, "num_tokens": 404011181.0, "step": 10636 }, { "epoch": 1.353135733367256, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.000194549560546875, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.369785785675049, "learning_rate": 1e-05, "loss": 0.4523, "mean_token_accuracy": 0.8814312219619751, "num_tokens": 404050519.0, "step": 10637 }, { "epoch": 1.3532629436458465, "ewc_loss": 0.11083984375, "ewc_loss_diag": 0.000194549560546875, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.111570835113525, "learning_rate": 1e-05, "loss": 0.4636, "mean_token_accuracy": 0.8768543004989624, "num_tokens": 404089407.0, "step": 10638 }, { "epoch": 1.353390153924437, "ewc_loss": 0.10986328125, "ewc_loss_diag": 0.000194549560546875, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.237066745758057, "learning_rate": 1e-05, "loss": 0.42, "mean_token_accuracy": 0.8914052844047546, "num_tokens": 404126564.0, "step": 10639 }, { "epoch": 1.3535173642030276, "ewc_loss": 0.10400390625, "ewc_loss_diag": 0.000194549560546875, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.953857898712158, "learning_rate": 1e-05, "loss": 0.4827, "mean_token_accuracy": 0.866773247718811, "num_tokens": 404166113.0, "step": 10640 }, { "epoch": 1.3536445744816181, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.000194549560546875, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.231151580810547, "learning_rate": 1e-05, "loss": 0.4956, "mean_token_accuracy": 0.8678890466690063, "num_tokens": 404204513.0, "step": 10641 }, { "epoch": 1.3537717847602087, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.000194549560546875, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.795602798461914, "learning_rate": 1e-05, "loss": 0.5675, "mean_token_accuracy": 0.8515273332595825, "num_tokens": 404244475.0, "step": 10642 }, { "epoch": 1.3538989950387992, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.000194549560546875, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.323408603668213, "learning_rate": 1e-05, "loss": 0.5192, "mean_token_accuracy": 0.8604495525360107, "num_tokens": 404278974.0, "step": 10643 }, { "epoch": 1.3540262053173897, "ewc_loss": 0.099609375, "ewc_loss_diag": 0.000194549560546875, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.775150775909424, "learning_rate": 1e-05, "loss": 0.5022, "mean_token_accuracy": 0.8606592416763306, "num_tokens": 404313123.0, "step": 10644 }, { "epoch": 1.3541534155959802, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.0001964569091796875, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.37614631652832, "learning_rate": 1e-05, "loss": 0.4745, "mean_token_accuracy": 0.8731502294540405, "num_tokens": 404350195.0, "step": 10645 }, { "epoch": 1.3542806258745705, "ewc_loss": 0.09912109375, "ewc_loss_diag": 0.000194549560546875, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.556579113006592, "learning_rate": 1e-05, "loss": 0.4866, "mean_token_accuracy": 0.87073814868927, "num_tokens": 404387044.0, "step": 10646 }, { "epoch": 1.354407836153161, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.0001964569091796875, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.602133274078369, "learning_rate": 1e-05, "loss": 0.5003, "mean_token_accuracy": 0.8700947761535645, "num_tokens": 404423877.0, "step": 10647 }, { "epoch": 1.3545350464317516, "ewc_loss": 0.09814453125, "ewc_loss_diag": 0.0001964569091796875, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.461946964263916, "learning_rate": 1e-05, "loss": 0.462, "mean_token_accuracy": 0.874884843826294, "num_tokens": 404466428.0, "step": 10648 }, { "epoch": 1.3546622567103421, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0001964569091796875, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.962264537811279, "learning_rate": 1e-05, "loss": 0.515, "mean_token_accuracy": 0.8658270835876465, "num_tokens": 404503341.0, "step": 10649 }, { "epoch": 1.3547894669889327, "ewc_loss": 0.10205078125, "ewc_loss_diag": 0.0001964569091796875, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.559688091278076, "learning_rate": 1e-05, "loss": 0.4918, "mean_token_accuracy": 0.8665763139724731, "num_tokens": 404543650.0, "step": 10650 }, { "epoch": 1.3549166772675232, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0001964569091796875, "ewc_loss_parallel": 0.00011491775512695312, "grad_norm": 7.233497142791748, "learning_rate": 1e-05, "loss": 0.5307, "mean_token_accuracy": 0.8619762063026428, "num_tokens": 404588834.0, "step": 10651 }, { "epoch": 1.3550438875461137, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.0001964569091796875, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 5.905524730682373, "learning_rate": 1e-05, "loss": 0.5028, "mean_token_accuracy": 0.8637248873710632, "num_tokens": 404625903.0, "step": 10652 }, { "epoch": 1.3551710978247042, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0001964569091796875, "ewc_loss_parallel": 0.00011301040649414062, "grad_norm": 7.155818939208984, "learning_rate": 1e-05, "loss": 0.5031, "mean_token_accuracy": 0.8724985718727112, "num_tokens": 404666299.0, "step": 10653 }, { "epoch": 1.3552983081032948, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.0001964569091796875, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.22423791885376, "learning_rate": 1e-05, "loss": 0.4367, "mean_token_accuracy": 0.8888493776321411, "num_tokens": 404702743.0, "step": 10654 }, { "epoch": 1.3554255183818853, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.0001964569091796875, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.742316246032715, "learning_rate": 1e-05, "loss": 0.5598, "mean_token_accuracy": 0.8503028750419617, "num_tokens": 404748010.0, "step": 10655 }, { "epoch": 1.3555527286604758, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.0001964569091796875, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.329843997955322, "learning_rate": 1e-05, "loss": 0.5042, "mean_token_accuracy": 0.8679994344711304, "num_tokens": 404785886.0, "step": 10656 }, { "epoch": 1.3556799389390664, "ewc_loss": 0.11083984375, "ewc_loss_diag": 0.0001964569091796875, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.23951530456543, "learning_rate": 1e-05, "loss": 0.4301, "mean_token_accuracy": 0.8875386118888855, "num_tokens": 404820982.0, "step": 10657 }, { "epoch": 1.3558071492176569, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.0001964569091796875, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.394534587860107, "learning_rate": 1e-05, "loss": 0.5106, "mean_token_accuracy": 0.8662661910057068, "num_tokens": 404863943.0, "step": 10658 }, { "epoch": 1.3559343594962474, "ewc_loss": 0.10205078125, "ewc_loss_diag": 0.0001964569091796875, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.8151726722717285, "learning_rate": 1e-05, "loss": 0.5005, "mean_token_accuracy": 0.862808346748352, "num_tokens": 404905168.0, "step": 10659 }, { "epoch": 1.356061569774838, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0001964569091796875, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.574413776397705, "learning_rate": 1e-05, "loss": 0.5035, "mean_token_accuracy": 0.8684401512145996, "num_tokens": 404941863.0, "step": 10660 }, { "epoch": 1.3561887800534282, "ewc_loss": 0.09912109375, "ewc_loss_diag": 0.0001964569091796875, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.556938171386719, "learning_rate": 1e-05, "loss": 0.499, "mean_token_accuracy": 0.8637780547142029, "num_tokens": 404979780.0, "step": 10661 }, { "epoch": 1.3563159903320188, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.0001964569091796875, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.8617401123046875, "learning_rate": 1e-05, "loss": 0.5267, "mean_token_accuracy": 0.8615754246711731, "num_tokens": 405019194.0, "step": 10662 }, { "epoch": 1.3564432006106093, "ewc_loss": 0.099609375, "ewc_loss_diag": 0.0001964569091796875, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.51165771484375, "learning_rate": 1e-05, "loss": 0.4285, "mean_token_accuracy": 0.8863846063613892, "num_tokens": 405056021.0, "step": 10663 }, { "epoch": 1.3565704108891998, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0001964569091796875, "ewc_loss_parallel": 0.00011301040649414062, "grad_norm": 7.193377494812012, "learning_rate": 1e-05, "loss": 0.5495, "mean_token_accuracy": 0.8590399622917175, "num_tokens": 405099195.0, "step": 10664 }, { "epoch": 1.3566976211677904, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.0001964569091796875, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 5.766414165496826, "learning_rate": 1e-05, "loss": 0.4991, "mean_token_accuracy": 0.8643506765365601, "num_tokens": 405142821.0, "step": 10665 }, { "epoch": 1.3568248314463809, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.0001964569091796875, "ewc_loss_parallel": 0.00011444091796875, "grad_norm": 7.290992259979248, "learning_rate": 1e-05, "loss": 0.5376, "mean_token_accuracy": 0.8636217713356018, "num_tokens": 405178434.0, "step": 10666 }, { "epoch": 1.3569520417249714, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.0001964569091796875, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.134850025177002, "learning_rate": 1e-05, "loss": 0.54, "mean_token_accuracy": 0.8535850644111633, "num_tokens": 405211197.0, "step": 10667 }, { "epoch": 1.357079252003562, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0001964569091796875, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 7.026429176330566, "learning_rate": 1e-05, "loss": 0.4978, "mean_token_accuracy": 0.8723647594451904, "num_tokens": 405250330.0, "step": 10668 }, { "epoch": 1.3572064622821525, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0001964569091796875, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.24953556060791, "learning_rate": 1e-05, "loss": 0.4969, "mean_token_accuracy": 0.8700729012489319, "num_tokens": 405291694.0, "step": 10669 }, { "epoch": 1.3573336725607428, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.0001964569091796875, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.676846981048584, "learning_rate": 1e-05, "loss": 0.4709, "mean_token_accuracy": 0.8787119388580322, "num_tokens": 405326162.0, "step": 10670 }, { "epoch": 1.3574608828393333, "ewc_loss": 0.11474609375, "ewc_loss_diag": 0.0001964569091796875, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.203141689300537, "learning_rate": 1e-05, "loss": 0.4724, "mean_token_accuracy": 0.8768958449363708, "num_tokens": 405361862.0, "step": 10671 }, { "epoch": 1.3575880931179238, "ewc_loss": 0.11474609375, "ewc_loss_diag": 0.0001964569091796875, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.399200916290283, "learning_rate": 1e-05, "loss": 0.4861, "mean_token_accuracy": 0.8688576221466064, "num_tokens": 405402934.0, "step": 10672 }, { "epoch": 1.3577153033965144, "ewc_loss": 0.11083984375, "ewc_loss_diag": 0.0001964569091796875, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.088738918304443, "learning_rate": 1e-05, "loss": 0.51, "mean_token_accuracy": 0.861172616481781, "num_tokens": 405442206.0, "step": 10673 }, { "epoch": 1.357842513675105, "ewc_loss": 0.10986328125, "ewc_loss_diag": 0.0001964569091796875, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.242272853851318, "learning_rate": 1e-05, "loss": 0.5789, "mean_token_accuracy": 0.8424311280250549, "num_tokens": 405483351.0, "step": 10674 }, { "epoch": 1.3579697239536954, "ewc_loss": 0.10498046875, "ewc_loss_diag": 0.0001964569091796875, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.962706565856934, "learning_rate": 1e-05, "loss": 0.5043, "mean_token_accuracy": 0.8621734380722046, "num_tokens": 405523763.0, "step": 10675 }, { "epoch": 1.358096934232286, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.0001964569091796875, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.2161760330200195, "learning_rate": 1e-05, "loss": 0.4507, "mean_token_accuracy": 0.8815808296203613, "num_tokens": 405559640.0, "step": 10676 }, { "epoch": 1.3582241445108765, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.0001964569091796875, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.809588432312012, "learning_rate": 1e-05, "loss": 0.4609, "mean_token_accuracy": 0.8741487860679626, "num_tokens": 405594907.0, "step": 10677 }, { "epoch": 1.358351354789467, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.0001964569091796875, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.281691074371338, "learning_rate": 1e-05, "loss": 0.4475, "mean_token_accuracy": 0.8836270570755005, "num_tokens": 405630346.0, "step": 10678 }, { "epoch": 1.3584785650680575, "ewc_loss": 0.10009765625, "ewc_loss_diag": 0.0001964569091796875, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.668313503265381, "learning_rate": 1e-05, "loss": 0.4883, "mean_token_accuracy": 0.8599877953529358, "num_tokens": 405661603.0, "step": 10679 }, { "epoch": 1.358605775346648, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.0001964569091796875, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.446080207824707, "learning_rate": 1e-05, "loss": 0.5106, "mean_token_accuracy": 0.8681618571281433, "num_tokens": 405698907.0, "step": 10680 }, { "epoch": 1.3587329856252386, "ewc_loss": 0.0986328125, "ewc_loss_diag": 0.0001964569091796875, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.516467094421387, "learning_rate": 1e-05, "loss": 0.5125, "mean_token_accuracy": 0.8587564826011658, "num_tokens": 405744780.0, "step": 10681 }, { "epoch": 1.3588601959038291, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.0001964569091796875, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.754011631011963, "learning_rate": 1e-05, "loss": 0.4667, "mean_token_accuracy": 0.8771463632583618, "num_tokens": 405777883.0, "step": 10682 }, { "epoch": 1.3589874061824196, "ewc_loss": 0.09814453125, "ewc_loss_diag": 0.0001964569091796875, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.47227144241333, "learning_rate": 1e-05, "loss": 0.4898, "mean_token_accuracy": 0.865667462348938, "num_tokens": 405814680.0, "step": 10683 }, { "epoch": 1.3591146164610102, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0001964569091796875, "ewc_loss_parallel": 0.00011110305786132812, "grad_norm": 7.125199317932129, "learning_rate": 1e-05, "loss": 0.5599, "mean_token_accuracy": 0.8508756160736084, "num_tokens": 405847734.0, "step": 10684 }, { "epoch": 1.3592418267396005, "ewc_loss": 0.10498046875, "ewc_loss_diag": 0.0001964569091796875, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.687640190124512, "learning_rate": 1e-05, "loss": 0.493, "mean_token_accuracy": 0.8659172654151917, "num_tokens": 405882365.0, "step": 10685 }, { "epoch": 1.359369037018191, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.0001964569091796875, "ewc_loss_parallel": 0.0001163482666015625, "grad_norm": 7.280932426452637, "learning_rate": 1e-05, "loss": 0.5095, "mean_token_accuracy": 0.8716121315956116, "num_tokens": 405920641.0, "step": 10686 }, { "epoch": 1.3594962472967815, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.0001964569091796875, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.101686000823975, "learning_rate": 1e-05, "loss": 0.5081, "mean_token_accuracy": 0.8636075258255005, "num_tokens": 405958523.0, "step": 10687 }, { "epoch": 1.359623457575372, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0001964569091796875, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 7.010132789611816, "learning_rate": 1e-05, "loss": 0.4954, "mean_token_accuracy": 0.8734833002090454, "num_tokens": 405994338.0, "step": 10688 }, { "epoch": 1.3597506678539626, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.0001964569091796875, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.321810245513916, "learning_rate": 1e-05, "loss": 0.5506, "mean_token_accuracy": 0.8540617227554321, "num_tokens": 406033285.0, "step": 10689 }, { "epoch": 1.3598778781325531, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.0001964569091796875, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.4997992515563965, "learning_rate": 1e-05, "loss": 0.4683, "mean_token_accuracy": 0.8774763941764832, "num_tokens": 406076933.0, "step": 10690 }, { "epoch": 1.3600050884111436, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.0001964569091796875, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.365429401397705, "learning_rate": 1e-05, "loss": 0.5101, "mean_token_accuracy": 0.8695781230926514, "num_tokens": 406118258.0, "step": 10691 }, { "epoch": 1.3601322986897342, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.0001964569091796875, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 6.023256778717041, "learning_rate": 1e-05, "loss": 0.4475, "mean_token_accuracy": 0.8773290514945984, "num_tokens": 406155445.0, "step": 10692 }, { "epoch": 1.3602595089683247, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.0001964569091796875, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.465742588043213, "learning_rate": 1e-05, "loss": 0.5081, "mean_token_accuracy": 0.8655774593353271, "num_tokens": 406195128.0, "step": 10693 }, { "epoch": 1.3603867192469152, "ewc_loss": 0.10009765625, "ewc_loss_diag": 0.0001964569091796875, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.6753249168396, "learning_rate": 1e-05, "loss": 0.4511, "mean_token_accuracy": 0.8753738403320312, "num_tokens": 406229311.0, "step": 10694 }, { "epoch": 1.3605139295255055, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.0001964569091796875, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.697057247161865, "learning_rate": 1e-05, "loss": 0.5114, "mean_token_accuracy": 0.8685023188591003, "num_tokens": 406266252.0, "step": 10695 }, { "epoch": 1.360641139804096, "ewc_loss": 0.0986328125, "ewc_loss_diag": 0.0001964569091796875, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.510775566101074, "learning_rate": 1e-05, "loss": 0.4653, "mean_token_accuracy": 0.8752128481864929, "num_tokens": 406310555.0, "step": 10696 }, { "epoch": 1.3607683500826866, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0001964569091796875, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 7.017149448394775, "learning_rate": 1e-05, "loss": 0.4696, "mean_token_accuracy": 0.8800756931304932, "num_tokens": 406349620.0, "step": 10697 }, { "epoch": 1.3608955603612771, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.0001964569091796875, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.616043567657471, "learning_rate": 1e-05, "loss": 0.4857, "mean_token_accuracy": 0.8705163598060608, "num_tokens": 406380956.0, "step": 10698 }, { "epoch": 1.3610227706398677, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0001964569091796875, "ewc_loss_parallel": 0.00011348724365234375, "grad_norm": 7.238805770874023, "learning_rate": 1e-05, "loss": 0.4772, "mean_token_accuracy": 0.882179319858551, "num_tokens": 406417274.0, "step": 10699 }, { "epoch": 1.3611499809184582, "ewc_loss": 0.11181640625, "ewc_loss_diag": 0.0001964569091796875, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 5.945262432098389, "learning_rate": 1e-05, "loss": 0.5347, "mean_token_accuracy": 0.855985164642334, "num_tokens": 406453809.0, "step": 10700 }, { "epoch": 1.3612771911970487, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.0001964569091796875, "ewc_loss_parallel": 0.00011157989501953125, "grad_norm": 7.146182060241699, "learning_rate": 1e-05, "loss": 0.5046, "mean_token_accuracy": 0.8731662034988403, "num_tokens": 406489227.0, "step": 10701 }, { "epoch": 1.3614044014756392, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.0001964569091796875, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.164360523223877, "learning_rate": 1e-05, "loss": 0.4872, "mean_token_accuracy": 0.8736611008644104, "num_tokens": 406532736.0, "step": 10702 }, { "epoch": 1.3615316117542298, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.0001964569091796875, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.865529537200928, "learning_rate": 1e-05, "loss": 0.5139, "mean_token_accuracy": 0.867821991443634, "num_tokens": 406574409.0, "step": 10703 }, { "epoch": 1.3616588220328203, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.0001964569091796875, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.198671817779541, "learning_rate": 1e-05, "loss": 0.5335, "mean_token_accuracy": 0.856480598449707, "num_tokens": 406612519.0, "step": 10704 }, { "epoch": 1.3617860323114108, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.0001964569091796875, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.561638355255127, "learning_rate": 1e-05, "loss": 0.5147, "mean_token_accuracy": 0.8610647916793823, "num_tokens": 406652578.0, "step": 10705 }, { "epoch": 1.3619132425900013, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.0001964569091796875, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.118033409118652, "learning_rate": 1e-05, "loss": 0.5283, "mean_token_accuracy": 0.8565976619720459, "num_tokens": 406688739.0, "step": 10706 }, { "epoch": 1.3620404528685919, "ewc_loss": 0.11376953125, "ewc_loss_diag": 0.0001964569091796875, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.3570637702941895, "learning_rate": 1e-05, "loss": 0.4776, "mean_token_accuracy": 0.8754904270172119, "num_tokens": 406727721.0, "step": 10707 }, { "epoch": 1.3621676631471824, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.0001964569091796875, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 6.00611686706543, "learning_rate": 1e-05, "loss": 0.5007, "mean_token_accuracy": 0.863904595375061, "num_tokens": 406762864.0, "step": 10708 }, { "epoch": 1.362294873425773, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.0001964569091796875, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.254329681396484, "learning_rate": 1e-05, "loss": 0.4658, "mean_token_accuracy": 0.8781353235244751, "num_tokens": 406803333.0, "step": 10709 }, { "epoch": 1.3624220837043632, "ewc_loss": 0.10498046875, "ewc_loss_diag": 0.0001964569091796875, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.879239559173584, "learning_rate": 1e-05, "loss": 0.4714, "mean_token_accuracy": 0.8748055100440979, "num_tokens": 406840942.0, "step": 10710 }, { "epoch": 1.3625492939829538, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.0001964569091796875, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.267955780029297, "learning_rate": 1e-05, "loss": 0.4926, "mean_token_accuracy": 0.8702901005744934, "num_tokens": 406877293.0, "step": 10711 }, { "epoch": 1.3626765042615443, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.0001964569091796875, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.740288734436035, "learning_rate": 1e-05, "loss": 0.4775, "mean_token_accuracy": 0.8707482218742371, "num_tokens": 406915622.0, "step": 10712 }, { "epoch": 1.3628037145401348, "ewc_loss": 0.11376953125, "ewc_loss_diag": 0.0001964569091796875, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.381350040435791, "learning_rate": 1e-05, "loss": 0.4614, "mean_token_accuracy": 0.8799526691436768, "num_tokens": 406955378.0, "step": 10713 }, { "epoch": 1.3629309248187254, "ewc_loss": 0.099609375, "ewc_loss_diag": 0.0001964569091796875, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.653095722198486, "learning_rate": 1e-05, "loss": 0.5071, "mean_token_accuracy": 0.8608192205429077, "num_tokens": 406997305.0, "step": 10714 }, { "epoch": 1.3630581350973159, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.0001964569091796875, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.537428855895996, "learning_rate": 1e-05, "loss": 0.4518, "mean_token_accuracy": 0.8826894760131836, "num_tokens": 407039121.0, "step": 10715 }, { "epoch": 1.3631853453759064, "ewc_loss": 0.09765625, "ewc_loss_diag": 0.0001964569091796875, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.543909072875977, "learning_rate": 1e-05, "loss": 0.5032, "mean_token_accuracy": 0.8609167337417603, "num_tokens": 407081007.0, "step": 10716 }, { "epoch": 1.363312555654497, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.0001964569091796875, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.787620544433594, "learning_rate": 1e-05, "loss": 0.4857, "mean_token_accuracy": 0.8709085583686829, "num_tokens": 407122498.0, "step": 10717 }, { "epoch": 1.3634397659330875, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.0001964569091796875, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.56130313873291, "learning_rate": 1e-05, "loss": 0.4617, "mean_token_accuracy": 0.8785891532897949, "num_tokens": 407169304.0, "step": 10718 }, { "epoch": 1.3635669762116778, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0001964569091796875, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 7.039054870605469, "learning_rate": 1e-05, "loss": 0.5432, "mean_token_accuracy": 0.8560492396354675, "num_tokens": 407204364.0, "step": 10719 }, { "epoch": 1.3636941864902683, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.0001964569091796875, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.788235664367676, "learning_rate": 1e-05, "loss": 0.5613, "mean_token_accuracy": 0.8450490236282349, "num_tokens": 407242467.0, "step": 10720 }, { "epoch": 1.3638213967688588, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0001964569091796875, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 7.040645122528076, "learning_rate": 1e-05, "loss": 0.5153, "mean_token_accuracy": 0.87154620885849, "num_tokens": 407279965.0, "step": 10721 }, { "epoch": 1.3639486070474494, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.0001964569091796875, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.067594051361084, "learning_rate": 1e-05, "loss": 0.4582, "mean_token_accuracy": 0.8786343336105347, "num_tokens": 407316086.0, "step": 10722 }, { "epoch": 1.3640758173260399, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.0001964569091796875, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.707696437835693, "learning_rate": 1e-05, "loss": 0.5168, "mean_token_accuracy": 0.8574736714363098, "num_tokens": 407355433.0, "step": 10723 }, { "epoch": 1.3642030276046304, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.0001964569091796875, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.21842622756958, "learning_rate": 1e-05, "loss": 0.4711, "mean_token_accuracy": 0.8750224709510803, "num_tokens": 407392919.0, "step": 10724 }, { "epoch": 1.364330237883221, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.0001964569091796875, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.238885402679443, "learning_rate": 1e-05, "loss": 0.5028, "mean_token_accuracy": 0.8642377257347107, "num_tokens": 407428089.0, "step": 10725 }, { "epoch": 1.3644574481618115, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.0001964569091796875, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.330474853515625, "learning_rate": 1e-05, "loss": 0.4795, "mean_token_accuracy": 0.876576840877533, "num_tokens": 407467232.0, "step": 10726 }, { "epoch": 1.364584658440402, "ewc_loss": 0.10107421875, "ewc_loss_diag": 0.0001964569091796875, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.771785736083984, "learning_rate": 1e-05, "loss": 0.512, "mean_token_accuracy": 0.8605374693870544, "num_tokens": 407504975.0, "step": 10727 }, { "epoch": 1.3647118687189925, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.0001964569091796875, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.571405410766602, "learning_rate": 1e-05, "loss": 0.4287, "mean_token_accuracy": 0.8900406360626221, "num_tokens": 407535308.0, "step": 10728 }, { "epoch": 1.364839078997583, "ewc_loss": 0.09716796875, "ewc_loss_diag": 0.0001964569091796875, "ewc_loss_parallel": 7.772445678710938e-05, "grad_norm": 5.456532001495361, "learning_rate": 1e-05, "loss": 0.4706, "mean_token_accuracy": 0.8713173866271973, "num_tokens": 407578956.0, "step": 10729 }, { "epoch": 1.3649662892761736, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0001964569091796875, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.975143909454346, "learning_rate": 1e-05, "loss": 0.4961, "mean_token_accuracy": 0.8713169097900391, "num_tokens": 407617633.0, "step": 10730 }, { "epoch": 1.3650934995547641, "ewc_loss": 0.099609375, "ewc_loss_diag": 0.00019741058349609375, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.4662017822265625, "learning_rate": 1e-05, "loss": 0.463, "mean_token_accuracy": 0.8742791414260864, "num_tokens": 407646322.0, "step": 10731 }, { "epoch": 1.3652207098333546, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.0001964569091796875, "ewc_loss_parallel": 0.00011777877807617188, "grad_norm": 7.37438440322876, "learning_rate": 1e-05, "loss": 0.5413, "mean_token_accuracy": 0.8598583936691284, "num_tokens": 407684957.0, "step": 10732 }, { "epoch": 1.3653479201119452, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.0001964569091796875, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 5.866865158081055, "learning_rate": 1e-05, "loss": 0.4684, "mean_token_accuracy": 0.8740994930267334, "num_tokens": 407722917.0, "step": 10733 }, { "epoch": 1.3654751303905355, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.0001964569091796875, "ewc_loss_parallel": 0.00011968612670898438, "grad_norm": 7.399232387542725, "learning_rate": 1e-05, "loss": 0.51, "mean_token_accuracy": 0.8686012625694275, "num_tokens": 407759599.0, "step": 10734 }, { "epoch": 1.365602340669126, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.0001964569091796875, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.296189308166504, "learning_rate": 1e-05, "loss": 0.5107, "mean_token_accuracy": 0.8657232522964478, "num_tokens": 407796416.0, "step": 10735 }, { "epoch": 1.3657295509477165, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0001964569091796875, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 7.044118404388428, "learning_rate": 1e-05, "loss": 0.5954, "mean_token_accuracy": 0.8456528186798096, "num_tokens": 407841211.0, "step": 10736 }, { "epoch": 1.365856761226307, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.0001964569091796875, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.364596366882324, "learning_rate": 1e-05, "loss": 0.5036, "mean_token_accuracy": 0.8679828643798828, "num_tokens": 407879409.0, "step": 10737 }, { "epoch": 1.3659839715048976, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.0001964569091796875, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.657464981079102, "learning_rate": 1e-05, "loss": 0.5108, "mean_token_accuracy": 0.864289402961731, "num_tokens": 407915275.0, "step": 10738 }, { "epoch": 1.3661111817834881, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.0001964569091796875, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.255885601043701, "learning_rate": 1e-05, "loss": 0.5035, "mean_token_accuracy": 0.8715282678604126, "num_tokens": 407953641.0, "step": 10739 }, { "epoch": 1.3662383920620786, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.00019741058349609375, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.378018379211426, "learning_rate": 1e-05, "loss": 0.4788, "mean_token_accuracy": 0.8742800951004028, "num_tokens": 407992662.0, "step": 10740 }, { "epoch": 1.3663656023406692, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.00019741058349609375, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.119449615478516, "learning_rate": 1e-05, "loss": 0.4685, "mean_token_accuracy": 0.8763135075569153, "num_tokens": 408026005.0, "step": 10741 }, { "epoch": 1.3664928126192597, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.00019741058349609375, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.242425441741943, "learning_rate": 1e-05, "loss": 0.4897, "mean_token_accuracy": 0.8678182363510132, "num_tokens": 408065250.0, "step": 10742 }, { "epoch": 1.3666200228978502, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.00019741058349609375, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.950131416320801, "learning_rate": 1e-05, "loss": 0.489, "mean_token_accuracy": 0.8675880432128906, "num_tokens": 408109333.0, "step": 10743 }, { "epoch": 1.3667472331764405, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.00019741058349609375, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.235465049743652, "learning_rate": 1e-05, "loss": 0.4843, "mean_token_accuracy": 0.8750507831573486, "num_tokens": 408142416.0, "step": 10744 }, { "epoch": 1.366874443455031, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.00019741058349609375, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.785370349884033, "learning_rate": 1e-05, "loss": 0.4319, "mean_token_accuracy": 0.885093092918396, "num_tokens": 408184335.0, "step": 10745 }, { "epoch": 1.3670016537336216, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.00019741058349609375, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.359479904174805, "learning_rate": 1e-05, "loss": 0.5031, "mean_token_accuracy": 0.8689815998077393, "num_tokens": 408221585.0, "step": 10746 }, { "epoch": 1.3671288640122121, "ewc_loss": 0.0986328125, "ewc_loss_diag": 0.00019741058349609375, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.617269992828369, "learning_rate": 1e-05, "loss": 0.5402, "mean_token_accuracy": 0.84641432762146, "num_tokens": 408262105.0, "step": 10747 }, { "epoch": 1.3672560742908026, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.00019741058349609375, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.583171844482422, "learning_rate": 1e-05, "loss": 0.4853, "mean_token_accuracy": 0.8706755638122559, "num_tokens": 408302365.0, "step": 10748 }, { "epoch": 1.3673832845693932, "ewc_loss": 0.09765625, "ewc_loss_diag": 0.00019741058349609375, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.4987359046936035, "learning_rate": 1e-05, "loss": 0.4907, "mean_token_accuracy": 0.8631613850593567, "num_tokens": 408342033.0, "step": 10749 }, { "epoch": 1.3675104948479837, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.00019741058349609375, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.923077583312988, "learning_rate": 1e-05, "loss": 0.4922, "mean_token_accuracy": 0.8722397089004517, "num_tokens": 408381512.0, "step": 10750 }, { "epoch": 1.3676377051265742, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.00019741058349609375, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.548059940338135, "learning_rate": 1e-05, "loss": 0.4883, "mean_token_accuracy": 0.8684208393096924, "num_tokens": 408423867.0, "step": 10751 }, { "epoch": 1.3677649154051648, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.00019741058349609375, "ewc_loss_parallel": 0.00011348724365234375, "grad_norm": 7.227367401123047, "learning_rate": 1e-05, "loss": 0.5088, "mean_token_accuracy": 0.8713892698287964, "num_tokens": 408468417.0, "step": 10752 }, { "epoch": 1.3678921256837553, "ewc_loss": 0.109375, "ewc_loss_diag": 0.00019741058349609375, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 5.878950119018555, "learning_rate": 1e-05, "loss": 0.5615, "mean_token_accuracy": 0.8489251136779785, "num_tokens": 408514628.0, "step": 10753 }, { "epoch": 1.3680193359623458, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.00019741058349609375, "ewc_loss_parallel": 0.00011396408081054688, "grad_norm": 7.199529647827148, "learning_rate": 1e-05, "loss": 0.5575, "mean_token_accuracy": 0.8566309213638306, "num_tokens": 408551936.0, "step": 10754 }, { "epoch": 1.3681465462409363, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.00019741058349609375, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.247241020202637, "learning_rate": 1e-05, "loss": 0.516, "mean_token_accuracy": 0.8648667335510254, "num_tokens": 408581713.0, "step": 10755 }, { "epoch": 1.3682737565195269, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.00019741058349609375, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.7795000076293945, "learning_rate": 1e-05, "loss": 0.4709, "mean_token_accuracy": 0.8768462538719177, "num_tokens": 408618265.0, "step": 10756 }, { "epoch": 1.3684009667981174, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.00019741058349609375, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.346460342407227, "learning_rate": 1e-05, "loss": 0.5176, "mean_token_accuracy": 0.8692115545272827, "num_tokens": 408655701.0, "step": 10757 }, { "epoch": 1.368528177076708, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.00019741058349609375, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.277085304260254, "learning_rate": 1e-05, "loss": 0.5221, "mean_token_accuracy": 0.8587125539779663, "num_tokens": 408690281.0, "step": 10758 }, { "epoch": 1.3686553873552982, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.00019741058349609375, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.394191265106201, "learning_rate": 1e-05, "loss": 0.5678, "mean_token_accuracy": 0.8500699996948242, "num_tokens": 408733483.0, "step": 10759 }, { "epoch": 1.3687825976338888, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.00019741058349609375, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.844570636749268, "learning_rate": 1e-05, "loss": 0.4822, "mean_token_accuracy": 0.8698772192001343, "num_tokens": 408770596.0, "step": 10760 }, { "epoch": 1.3689098079124793, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.00019741058349609375, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.552579879760742, "learning_rate": 1e-05, "loss": 0.4662, "mean_token_accuracy": 0.878789484500885, "num_tokens": 408808265.0, "step": 10761 }, { "epoch": 1.3690370181910698, "ewc_loss": 0.099609375, "ewc_loss_diag": 0.00019741058349609375, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.57015323638916, "learning_rate": 1e-05, "loss": 0.4424, "mean_token_accuracy": 0.8791477680206299, "num_tokens": 408847205.0, "step": 10762 }, { "epoch": 1.3691642284696603, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.00019741058349609375, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.8541412353515625, "learning_rate": 1e-05, "loss": 0.5243, "mean_token_accuracy": 0.8642180562019348, "num_tokens": 408885744.0, "step": 10763 }, { "epoch": 1.3692914387482509, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.00019741058349609375, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.506203651428223, "learning_rate": 1e-05, "loss": 0.4494, "mean_token_accuracy": 0.8756225109100342, "num_tokens": 408925393.0, "step": 10764 }, { "epoch": 1.3694186490268414, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.00019741058349609375, "ewc_loss_parallel": 0.0001125335693359375, "grad_norm": 7.198922634124756, "learning_rate": 1e-05, "loss": 0.5267, "mean_token_accuracy": 0.8653762936592102, "num_tokens": 408964995.0, "step": 10765 }, { "epoch": 1.369545859305432, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.00019741058349609375, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.758988857269287, "learning_rate": 1e-05, "loss": 0.4634, "mean_token_accuracy": 0.8773224353790283, "num_tokens": 408996336.0, "step": 10766 }, { "epoch": 1.3696730695840225, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.00019741058349609375, "ewc_loss_parallel": 0.00011539459228515625, "grad_norm": 7.31953763961792, "learning_rate": 1e-05, "loss": 0.5714, "mean_token_accuracy": 0.8520413637161255, "num_tokens": 409034972.0, "step": 10767 }, { "epoch": 1.3698002798626128, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.00019741058349609375, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.126864433288574, "learning_rate": 1e-05, "loss": 0.4579, "mean_token_accuracy": 0.8820996880531311, "num_tokens": 409069421.0, "step": 10768 }, { "epoch": 1.3699274901412033, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.00019741058349609375, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 7.082183837890625, "learning_rate": 1e-05, "loss": 0.5258, "mean_token_accuracy": 0.8628939986228943, "num_tokens": 409104763.0, "step": 10769 }, { "epoch": 1.3700547004197938, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.00019741058349609375, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.280892372131348, "learning_rate": 1e-05, "loss": 0.4862, "mean_token_accuracy": 0.872389018535614, "num_tokens": 409142799.0, "step": 10770 }, { "epoch": 1.3701819106983844, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.00019741058349609375, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.733535289764404, "learning_rate": 1e-05, "loss": 0.5228, "mean_token_accuracy": 0.8610336780548096, "num_tokens": 409178861.0, "step": 10771 }, { "epoch": 1.3703091209769749, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.00019741058349609375, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.236785888671875, "learning_rate": 1e-05, "loss": 0.4903, "mean_token_accuracy": 0.8709641098976135, "num_tokens": 409218628.0, "step": 10772 }, { "epoch": 1.3704363312555654, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.00019741058349609375, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.425663471221924, "learning_rate": 1e-05, "loss": 0.5535, "mean_token_accuracy": 0.8512945771217346, "num_tokens": 409255209.0, "step": 10773 }, { "epoch": 1.370563541534156, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.00019741058349609375, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.136934280395508, "learning_rate": 1e-05, "loss": 0.5038, "mean_token_accuracy": 0.8670953512191772, "num_tokens": 409295807.0, "step": 10774 }, { "epoch": 1.3706907518127465, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.00019741058349609375, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.245639801025391, "learning_rate": 1e-05, "loss": 0.4826, "mean_token_accuracy": 0.8721389770507812, "num_tokens": 409329263.0, "step": 10775 }, { "epoch": 1.370817962091337, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.00019741058349609375, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 6.0069379806518555, "learning_rate": 1e-05, "loss": 0.4349, "mean_token_accuracy": 0.8867272734642029, "num_tokens": 409365571.0, "step": 10776 }, { "epoch": 1.3709451723699275, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.00019741058349609375, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.180145740509033, "learning_rate": 1e-05, "loss": 0.4932, "mean_token_accuracy": 0.8650377988815308, "num_tokens": 409406096.0, "step": 10777 }, { "epoch": 1.371072382648518, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.00019741058349609375, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.854187965393066, "learning_rate": 1e-05, "loss": 0.5266, "mean_token_accuracy": 0.8575337529182434, "num_tokens": 409450436.0, "step": 10778 }, { "epoch": 1.3711995929271086, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.00019741058349609375, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.247776508331299, "learning_rate": 1e-05, "loss": 0.5006, "mean_token_accuracy": 0.868266761302948, "num_tokens": 409485134.0, "step": 10779 }, { "epoch": 1.371326803205699, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.00019741058349609375, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.694611072540283, "learning_rate": 1e-05, "loss": 0.4735, "mean_token_accuracy": 0.8703677654266357, "num_tokens": 409522890.0, "step": 10780 }, { "epoch": 1.3714540134842896, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.00019741058349609375, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.4089484214782715, "learning_rate": 1e-05, "loss": 0.5085, "mean_token_accuracy": 0.8656699657440186, "num_tokens": 409561537.0, "step": 10781 }, { "epoch": 1.3715812237628802, "ewc_loss": 0.09765625, "ewc_loss_diag": 0.00019741058349609375, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.5472869873046875, "learning_rate": 1e-05, "loss": 0.4424, "mean_token_accuracy": 0.8796707391738892, "num_tokens": 409595259.0, "step": 10782 }, { "epoch": 1.3717084340414705, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.00019741058349609375, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.681371688842773, "learning_rate": 1e-05, "loss": 0.4773, "mean_token_accuracy": 0.8763394355773926, "num_tokens": 409633978.0, "step": 10783 }, { "epoch": 1.371835644320061, "ewc_loss": 0.09765625, "ewc_loss_diag": 0.00019741058349609375, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.470573902130127, "learning_rate": 1e-05, "loss": 0.4878, "mean_token_accuracy": 0.8690191507339478, "num_tokens": 409671320.0, "step": 10784 }, { "epoch": 1.3719628545986515, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.00019741058349609375, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 7.046688556671143, "learning_rate": 1e-05, "loss": 0.4954, "mean_token_accuracy": 0.8732770681381226, "num_tokens": 409712816.0, "step": 10785 }, { "epoch": 1.372090064877242, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.00019741058349609375, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.606773853302002, "learning_rate": 1e-05, "loss": 0.4898, "mean_token_accuracy": 0.8676280975341797, "num_tokens": 409751918.0, "step": 10786 }, { "epoch": 1.3722172751558326, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.00019741058349609375, "ewc_loss_parallel": 0.00011587142944335938, "grad_norm": 7.300848960876465, "learning_rate": 1e-05, "loss": 0.4735, "mean_token_accuracy": 0.8843674063682556, "num_tokens": 409786905.0, "step": 10787 }, { "epoch": 1.3723444854344231, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.00019741058349609375, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 5.99902868270874, "learning_rate": 1e-05, "loss": 0.4902, "mean_token_accuracy": 0.8680427670478821, "num_tokens": 409827634.0, "step": 10788 }, { "epoch": 1.3724716957130136, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.00019741058349609375, "ewc_loss_parallel": 0.00011014938354492188, "grad_norm": 7.137118339538574, "learning_rate": 1e-05, "loss": 0.4932, "mean_token_accuracy": 0.8730522990226746, "num_tokens": 409863060.0, "step": 10789 }, { "epoch": 1.3725989059916042, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.00019741058349609375, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.299864768981934, "learning_rate": 1e-05, "loss": 0.4755, "mean_token_accuracy": 0.8755744099617004, "num_tokens": 409896723.0, "step": 10790 }, { "epoch": 1.3727261162701947, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.00019741058349609375, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.66959810256958, "learning_rate": 1e-05, "loss": 0.4988, "mean_token_accuracy": 0.8689261674880981, "num_tokens": 409935912.0, "step": 10791 }, { "epoch": 1.3728533265487852, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.00019741058349609375, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.364449501037598, "learning_rate": 1e-05, "loss": 0.4967, "mean_token_accuracy": 0.8723543286323547, "num_tokens": 409973381.0, "step": 10792 }, { "epoch": 1.3729805368273755, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.00019741058349609375, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 6.172755241394043, "learning_rate": 1e-05, "loss": 0.5037, "mean_token_accuracy": 0.8625226020812988, "num_tokens": 410010548.0, "step": 10793 }, { "epoch": 1.373107747105966, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.00019741058349609375, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.41502571105957, "learning_rate": 1e-05, "loss": 0.4536, "mean_token_accuracy": 0.8835967779159546, "num_tokens": 410046456.0, "step": 10794 }, { "epoch": 1.3732349573845566, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.00019741058349609375, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.786146640777588, "learning_rate": 1e-05, "loss": 0.4296, "mean_token_accuracy": 0.8852417469024658, "num_tokens": 410080568.0, "step": 10795 }, { "epoch": 1.3733621676631471, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.00019741058349609375, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.578171730041504, "learning_rate": 1e-05, "loss": 0.4693, "mean_token_accuracy": 0.8752541542053223, "num_tokens": 410121008.0, "step": 10796 }, { "epoch": 1.3734893779417376, "ewc_loss": 0.099609375, "ewc_loss_diag": 0.00019741058349609375, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.571104049682617, "learning_rate": 1e-05, "loss": 0.4682, "mean_token_accuracy": 0.872395396232605, "num_tokens": 410165872.0, "step": 10797 }, { "epoch": 1.3736165882203282, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.00019741058349609375, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.851407527923584, "learning_rate": 1e-05, "loss": 0.4959, "mean_token_accuracy": 0.8743366003036499, "num_tokens": 410207379.0, "step": 10798 }, { "epoch": 1.3737437984989187, "ewc_loss": 0.099609375, "ewc_loss_diag": 0.00019741058349609375, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.560355186462402, "learning_rate": 1e-05, "loss": 0.5222, "mean_token_accuracy": 0.856730580329895, "num_tokens": 410246729.0, "step": 10799 }, { "epoch": 1.3738710087775092, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.00019741058349609375, "ewc_loss_parallel": 0.00010967254638671875, "grad_norm": 7.133269786834717, "learning_rate": 1e-05, "loss": 0.4853, "mean_token_accuracy": 0.8760076761245728, "num_tokens": 410281237.0, "step": 10800 }, { "epoch": 1.3739982190560998, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.00019741058349609375, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.775537490844727, "learning_rate": 1e-05, "loss": 0.4979, "mean_token_accuracy": 0.8629528284072876, "num_tokens": 410318485.0, "step": 10801 }, { "epoch": 1.3741254293346903, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.00019741058349609375, "ewc_loss_parallel": 0.00011157989501953125, "grad_norm": 7.204382419586182, "learning_rate": 1e-05, "loss": 0.5196, "mean_token_accuracy": 0.8691864609718323, "num_tokens": 410357995.0, "step": 10802 }, { "epoch": 1.3742526396132808, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.00019931793212890625, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.060186862945557, "learning_rate": 1e-05, "loss": 0.5023, "mean_token_accuracy": 0.8678773641586304, "num_tokens": 410395672.0, "step": 10803 }, { "epoch": 1.3743798498918713, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.00019931793212890625, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 7.018365383148193, "learning_rate": 1e-05, "loss": 0.5735, "mean_token_accuracy": 0.8468716144561768, "num_tokens": 410433566.0, "step": 10804 }, { "epoch": 1.3745070601704619, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.00019931793212890625, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.177999019622803, "learning_rate": 1e-05, "loss": 0.4994, "mean_token_accuracy": 0.8678786754608154, "num_tokens": 410473622.0, "step": 10805 }, { "epoch": 1.3746342704490524, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.00019931793212890625, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.7204060554504395, "learning_rate": 1e-05, "loss": 0.4755, "mean_token_accuracy": 0.8765707015991211, "num_tokens": 410507155.0, "step": 10806 }, { "epoch": 1.374761480727643, "ewc_loss": 0.11376953125, "ewc_loss_diag": 0.00019931793212890625, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.168613910675049, "learning_rate": 1e-05, "loss": 0.5098, "mean_token_accuracy": 0.8654226064682007, "num_tokens": 410543352.0, "step": 10807 }, { "epoch": 1.3748886910062332, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.00019931793212890625, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.442509651184082, "learning_rate": 1e-05, "loss": 0.4745, "mean_token_accuracy": 0.8752589225769043, "num_tokens": 410584560.0, "step": 10808 }, { "epoch": 1.3750159012848238, "ewc_loss": 0.11181640625, "ewc_loss_diag": 0.00019931793212890625, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.097110271453857, "learning_rate": 1e-05, "loss": 0.4666, "mean_token_accuracy": 0.8777583837509155, "num_tokens": 410615948.0, "step": 10809 }, { "epoch": 1.3751431115634143, "ewc_loss": 0.11083984375, "ewc_loss_diag": 0.00019931793212890625, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.261613368988037, "learning_rate": 1e-05, "loss": 0.4787, "mean_token_accuracy": 0.8704375624656677, "num_tokens": 410649513.0, "step": 10810 }, { "epoch": 1.3752703218420048, "ewc_loss": 0.10595703125, "ewc_loss_diag": 0.00019931793212890625, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.973069667816162, "learning_rate": 1e-05, "loss": 0.4926, "mean_token_accuracy": 0.8669884204864502, "num_tokens": 410693087.0, "step": 10811 }, { "epoch": 1.3753975321205953, "ewc_loss": 0.11083984375, "ewc_loss_diag": 0.00019931793212890625, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.217601299285889, "learning_rate": 1e-05, "loss": 0.4566, "mean_token_accuracy": 0.8811986446380615, "num_tokens": 410729200.0, "step": 10812 }, { "epoch": 1.3755247423991859, "ewc_loss": 0.10205078125, "ewc_loss_diag": 0.00019931793212890625, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.846301078796387, "learning_rate": 1e-05, "loss": 0.5071, "mean_token_accuracy": 0.8636459708213806, "num_tokens": 410769830.0, "step": 10813 }, { "epoch": 1.3756519526777764, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.00019931793212890625, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.262953281402588, "learning_rate": 1e-05, "loss": 0.4989, "mean_token_accuracy": 0.8660588264465332, "num_tokens": 410815570.0, "step": 10814 }, { "epoch": 1.375779162956367, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.00019931793212890625, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.69460391998291, "learning_rate": 1e-05, "loss": 0.4119, "mean_token_accuracy": 0.8938674330711365, "num_tokens": 410852777.0, "step": 10815 }, { "epoch": 1.3759063732349575, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.00019931793212890625, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.424715995788574, "learning_rate": 1e-05, "loss": 0.497, "mean_token_accuracy": 0.8731216788291931, "num_tokens": 410886809.0, "step": 10816 }, { "epoch": 1.3760335835135478, "ewc_loss": 0.09912109375, "ewc_loss_diag": 0.00019931793212890625, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.5605034828186035, "learning_rate": 1e-05, "loss": 0.4721, "mean_token_accuracy": 0.8728646636009216, "num_tokens": 410924631.0, "step": 10817 }, { "epoch": 1.3761607937921383, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.00019931793212890625, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.686677932739258, "learning_rate": 1e-05, "loss": 0.4969, "mean_token_accuracy": 0.8699923753738403, "num_tokens": 410965212.0, "step": 10818 }, { "epoch": 1.3762880040707288, "ewc_loss": 0.0986328125, "ewc_loss_diag": 0.00019931793212890625, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.605421543121338, "learning_rate": 1e-05, "loss": 0.5208, "mean_token_accuracy": 0.8572462797164917, "num_tokens": 411003429.0, "step": 10819 }, { "epoch": 1.3764152143493193, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.00019931793212890625, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.893406391143799, "learning_rate": 1e-05, "loss": 0.5542, "mean_token_accuracy": 0.8549959659576416, "num_tokens": 411043121.0, "step": 10820 }, { "epoch": 1.3765424246279099, "ewc_loss": 0.10205078125, "ewc_loss_diag": 0.00019931793212890625, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.596627712249756, "learning_rate": 1e-05, "loss": 0.5199, "mean_token_accuracy": 0.858626127243042, "num_tokens": 411082146.0, "step": 10821 }, { "epoch": 1.3766696349065004, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.00019931793212890625, "ewc_loss_parallel": 0.00011014938354492188, "grad_norm": 7.082341194152832, "learning_rate": 1e-05, "loss": 0.5195, "mean_token_accuracy": 0.8651289939880371, "num_tokens": 411116070.0, "step": 10822 }, { "epoch": 1.376796845185091, "ewc_loss": 0.10888671875, "ewc_loss_diag": 0.00019931793212890625, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 5.861728191375732, "learning_rate": 1e-05, "loss": 0.5066, "mean_token_accuracy": 0.8635258674621582, "num_tokens": 411162066.0, "step": 10823 }, { "epoch": 1.3769240554636815, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.00019931793212890625, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 6.996760368347168, "learning_rate": 1e-05, "loss": 0.494, "mean_token_accuracy": 0.8739273548126221, "num_tokens": 411200939.0, "step": 10824 }, { "epoch": 1.377051265742272, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.00019931793212890625, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.124046802520752, "learning_rate": 1e-05, "loss": 0.5141, "mean_token_accuracy": 0.8635973930358887, "num_tokens": 411241161.0, "step": 10825 }, { "epoch": 1.3771784760208625, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.00019931793212890625, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.615970134735107, "learning_rate": 1e-05, "loss": 0.5052, "mean_token_accuracy": 0.8667126893997192, "num_tokens": 411284316.0, "step": 10826 }, { "epoch": 1.377305686299453, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.00019931793212890625, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.258411884307861, "learning_rate": 1e-05, "loss": 0.5431, "mean_token_accuracy": 0.8529424667358398, "num_tokens": 411320795.0, "step": 10827 }, { "epoch": 1.3774328965780436, "ewc_loss": 0.10888671875, "ewc_loss_diag": 0.00019931793212890625, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.136500358581543, "learning_rate": 1e-05, "loss": 0.4767, "mean_token_accuracy": 0.8726520538330078, "num_tokens": 411357513.0, "step": 10828 }, { "epoch": 1.377560106856634, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.00019931793212890625, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.373591423034668, "learning_rate": 1e-05, "loss": 0.4375, "mean_token_accuracy": 0.8885633945465088, "num_tokens": 411395310.0, "step": 10829 }, { "epoch": 1.3776873171352246, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.00019931793212890625, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.703433513641357, "learning_rate": 1e-05, "loss": 0.4811, "mean_token_accuracy": 0.8673678636550903, "num_tokens": 411437774.0, "step": 10830 }, { "epoch": 1.3778145274138152, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.00019931793212890625, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.653437614440918, "learning_rate": 1e-05, "loss": 0.4735, "mean_token_accuracy": 0.8787609338760376, "num_tokens": 411471399.0, "step": 10831 }, { "epoch": 1.3779417376924055, "ewc_loss": 0.0986328125, "ewc_loss_diag": 0.00019931793212890625, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.4653239250183105, "learning_rate": 1e-05, "loss": 0.5134, "mean_token_accuracy": 0.8574366569519043, "num_tokens": 411505289.0, "step": 10832 }, { "epoch": 1.378068947970996, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.00019931793212890625, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 7.0410027503967285, "learning_rate": 1e-05, "loss": 0.4797, "mean_token_accuracy": 0.8795820474624634, "num_tokens": 411541463.0, "step": 10833 }, { "epoch": 1.3781961582495865, "ewc_loss": 0.10107421875, "ewc_loss_diag": 0.00019931793212890625, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.534724235534668, "learning_rate": 1e-05, "loss": 0.4871, "mean_token_accuracy": 0.8666974306106567, "num_tokens": 411577276.0, "step": 10834 }, { "epoch": 1.378323368528177, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.00019931793212890625, "ewc_loss_parallel": 0.0001201629638671875, "grad_norm": 7.399120807647705, "learning_rate": 1e-05, "loss": 0.6012, "mean_token_accuracy": 0.8400470614433289, "num_tokens": 411610417.0, "step": 10835 }, { "epoch": 1.3784505788067676, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.00019931793212890625, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 5.961024761199951, "learning_rate": 1e-05, "loss": 0.5462, "mean_token_accuracy": 0.8496552109718323, "num_tokens": 411646539.0, "step": 10836 }, { "epoch": 1.378577789085358, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.00019931793212890625, "ewc_loss_parallel": 0.0001163482666015625, "grad_norm": 7.324620723724365, "learning_rate": 1e-05, "loss": 0.5351, "mean_token_accuracy": 0.8615739345550537, "num_tokens": 411685890.0, "step": 10837 }, { "epoch": 1.3787049993639486, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.00019931793212890625, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.282155513763428, "learning_rate": 1e-05, "loss": 0.4746, "mean_token_accuracy": 0.8778489828109741, "num_tokens": 411725566.0, "step": 10838 }, { "epoch": 1.3788322096425392, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.00019931793212890625, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 6.9821319580078125, "learning_rate": 1e-05, "loss": 0.4948, "mean_token_accuracy": 0.8746777772903442, "num_tokens": 411767256.0, "step": 10839 }, { "epoch": 1.3789594199211297, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.00019931793212890625, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.328169822692871, "learning_rate": 1e-05, "loss": 0.4927, "mean_token_accuracy": 0.8676345348358154, "num_tokens": 411802210.0, "step": 10840 }, { "epoch": 1.3790866301997202, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.00019931793212890625, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.626283645629883, "learning_rate": 1e-05, "loss": 0.4372, "mean_token_accuracy": 0.8879309296607971, "num_tokens": 411842011.0, "step": 10841 }, { "epoch": 1.3792138404783105, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.00019931793212890625, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.222577095031738, "learning_rate": 1e-05, "loss": 0.4839, "mean_token_accuracy": 0.8749901056289673, "num_tokens": 411877137.0, "step": 10842 }, { "epoch": 1.379341050756901, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.00019931793212890625, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.378475666046143, "learning_rate": 1e-05, "loss": 0.469, "mean_token_accuracy": 0.876736044883728, "num_tokens": 411909130.0, "step": 10843 }, { "epoch": 1.3794682610354916, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.00019931793212890625, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.082177639007568, "learning_rate": 1e-05, "loss": 0.4973, "mean_token_accuracy": 0.8651838898658752, "num_tokens": 411942542.0, "step": 10844 }, { "epoch": 1.3795954713140821, "ewc_loss": 0.11181640625, "ewc_loss_diag": 0.00019931793212890625, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.27169132232666, "learning_rate": 1e-05, "loss": 0.5142, "mean_token_accuracy": 0.8610725998878479, "num_tokens": 411974272.0, "step": 10845 }, { "epoch": 1.3797226815926726, "ewc_loss": 0.10595703125, "ewc_loss_diag": 0.00019931793212890625, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.916022300720215, "learning_rate": 1e-05, "loss": 0.46, "mean_token_accuracy": 0.8786804676055908, "num_tokens": 412012748.0, "step": 10846 }, { "epoch": 1.3798498918712632, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.00019931793212890625, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.276032447814941, "learning_rate": 1e-05, "loss": 0.501, "mean_token_accuracy": 0.8630584478378296, "num_tokens": 412051171.0, "step": 10847 }, { "epoch": 1.3799771021498537, "ewc_loss": 0.10205078125, "ewc_loss_diag": 0.00019931793212890625, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.753530502319336, "learning_rate": 1e-05, "loss": 0.54, "mean_token_accuracy": 0.8561794757843018, "num_tokens": 412090267.0, "step": 10848 }, { "epoch": 1.3801043124284442, "ewc_loss": 0.11376953125, "ewc_loss_diag": 0.00019931793212890625, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.4036688804626465, "learning_rate": 1e-05, "loss": 0.5229, "mean_token_accuracy": 0.8617879152297974, "num_tokens": 412118471.0, "step": 10849 }, { "epoch": 1.3802315227070348, "ewc_loss": 0.099609375, "ewc_loss_diag": 0.00019931793212890625, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.597716808319092, "learning_rate": 1e-05, "loss": 0.473, "mean_token_accuracy": 0.8732621669769287, "num_tokens": 412163639.0, "step": 10850 }, { "epoch": 1.3803587329856253, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.00019931793212890625, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.646477699279785, "learning_rate": 1e-05, "loss": 0.5452, "mean_token_accuracy": 0.8572969436645508, "num_tokens": 412203685.0, "step": 10851 }, { "epoch": 1.3804859432642158, "ewc_loss": 0.09912109375, "ewc_loss_diag": 0.00019931793212890625, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.511830806732178, "learning_rate": 1e-05, "loss": 0.4674, "mean_token_accuracy": 0.8706878423690796, "num_tokens": 412238810.0, "step": 10852 }, { "epoch": 1.3806131535428063, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.00019931793212890625, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 6.990628719329834, "learning_rate": 1e-05, "loss": 0.5225, "mean_token_accuracy": 0.8653425574302673, "num_tokens": 412273840.0, "step": 10853 }, { "epoch": 1.3807403638213969, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.00019931793212890625, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.618245601654053, "learning_rate": 1e-05, "loss": 0.4822, "mean_token_accuracy": 0.8670505285263062, "num_tokens": 412309931.0, "step": 10854 }, { "epoch": 1.3808675740999874, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.00019931793212890625, "ewc_loss_parallel": 0.00011348724365234375, "grad_norm": 7.237370491027832, "learning_rate": 1e-05, "loss": 0.606, "mean_token_accuracy": 0.838084876537323, "num_tokens": 412346934.0, "step": 10855 }, { "epoch": 1.380994784378578, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.00019931793212890625, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 5.969799518585205, "learning_rate": 1e-05, "loss": 0.5016, "mean_token_accuracy": 0.8663922548294067, "num_tokens": 412387942.0, "step": 10856 }, { "epoch": 1.3811219946571682, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.00019931793212890625, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 7.099239349365234, "learning_rate": 1e-05, "loss": 0.5109, "mean_token_accuracy": 0.8695487976074219, "num_tokens": 412429726.0, "step": 10857 }, { "epoch": 1.3812492049357588, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.00019931793212890625, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.258561611175537, "learning_rate": 1e-05, "loss": 0.485, "mean_token_accuracy": 0.8719584941864014, "num_tokens": 412472711.0, "step": 10858 }, { "epoch": 1.3813764152143493, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.00019931793212890625, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.646633148193359, "learning_rate": 1e-05, "loss": 0.4667, "mean_token_accuracy": 0.8783797025680542, "num_tokens": 412511384.0, "step": 10859 }, { "epoch": 1.3815036254929398, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.00019931793212890625, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.35413932800293, "learning_rate": 1e-05, "loss": 0.5616, "mean_token_accuracy": 0.8530066013336182, "num_tokens": 412552991.0, "step": 10860 }, { "epoch": 1.3816308357715303, "ewc_loss": 0.10693359375, "ewc_loss_diag": 0.00019931793212890625, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 6.138765811920166, "learning_rate": 1e-05, "loss": 0.5966, "mean_token_accuracy": 0.8338646292686462, "num_tokens": 412593794.0, "step": 10861 }, { "epoch": 1.3817580460501209, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.00019931793212890625, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.451727867126465, "learning_rate": 1e-05, "loss": 0.4715, "mean_token_accuracy": 0.8779357671737671, "num_tokens": 412625569.0, "step": 10862 }, { "epoch": 1.3818852563287114, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.00019931793212890625, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.713050365447998, "learning_rate": 1e-05, "loss": 0.4339, "mean_token_accuracy": 0.883284330368042, "num_tokens": 412658086.0, "step": 10863 }, { "epoch": 1.382012466607302, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.00019931793212890625, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.691755294799805, "learning_rate": 1e-05, "loss": 0.5113, "mean_token_accuracy": 0.8681353330612183, "num_tokens": 412698642.0, "step": 10864 }, { "epoch": 1.3821396768858925, "ewc_loss": 0.09814453125, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.493265151977539, "learning_rate": 1e-05, "loss": 0.4691, "mean_token_accuracy": 0.8715680241584778, "num_tokens": 412738327.0, "step": 10865 }, { "epoch": 1.3822668871644828, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.00019931793212890625, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 7.0507283210754395, "learning_rate": 1e-05, "loss": 0.5023, "mean_token_accuracy": 0.8691498637199402, "num_tokens": 412773420.0, "step": 10866 }, { "epoch": 1.3823940974430733, "ewc_loss": 0.10205078125, "ewc_loss_diag": 0.00019931793212890625, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.5948944091796875, "learning_rate": 1e-05, "loss": 0.4687, "mean_token_accuracy": 0.8749063014984131, "num_tokens": 412805383.0, "step": 10867 }, { "epoch": 1.3825213077216638, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.00019931793212890625, "ewc_loss_parallel": 0.0001163482666015625, "grad_norm": 7.325183868408203, "learning_rate": 1e-05, "loss": 0.4976, "mean_token_accuracy": 0.875205397605896, "num_tokens": 412846924.0, "step": 10868 }, { "epoch": 1.3826485180002543, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.00019931793212890625, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 5.95959997177124, "learning_rate": 1e-05, "loss": 0.5186, "mean_token_accuracy": 0.859306812286377, "num_tokens": 412883232.0, "step": 10869 }, { "epoch": 1.3827757282788449, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.00019931793212890625, "ewc_loss_parallel": 0.00011396408081054688, "grad_norm": 7.252882480621338, "learning_rate": 1e-05, "loss": 0.4909, "mean_token_accuracy": 0.8759075999259949, "num_tokens": 412922110.0, "step": 10870 }, { "epoch": 1.3829029385574354, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.00019931793212890625, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.2526535987854, "learning_rate": 1e-05, "loss": 0.5074, "mean_token_accuracy": 0.8646097183227539, "num_tokens": 412963661.0, "step": 10871 }, { "epoch": 1.383030148836026, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.00019931793212890625, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.902620315551758, "learning_rate": 1e-05, "loss": 0.4295, "mean_token_accuracy": 0.8940819501876831, "num_tokens": 413002468.0, "step": 10872 }, { "epoch": 1.3831573591146165, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.00019931793212890625, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.2784199714660645, "learning_rate": 1e-05, "loss": 0.5529, "mean_token_accuracy": 0.8546935319900513, "num_tokens": 413048135.0, "step": 10873 }, { "epoch": 1.383284569393207, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.00019931793212890625, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.583893775939941, "learning_rate": 1e-05, "loss": 0.4997, "mean_token_accuracy": 0.8692830801010132, "num_tokens": 413082552.0, "step": 10874 }, { "epoch": 1.3834117796717975, "ewc_loss": 0.11376953125, "ewc_loss_diag": 0.00019931793212890625, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.179304599761963, "learning_rate": 1e-05, "loss": 0.5247, "mean_token_accuracy": 0.8625357151031494, "num_tokens": 413129230.0, "step": 10875 }, { "epoch": 1.383538989950388, "ewc_loss": 0.11376953125, "ewc_loss_diag": 0.00019931793212890625, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.344484329223633, "learning_rate": 1e-05, "loss": 0.4656, "mean_token_accuracy": 0.8816662430763245, "num_tokens": 413168830.0, "step": 10876 }, { "epoch": 1.3836662002289786, "ewc_loss": 0.10986328125, "ewc_loss_diag": 0.00019931793212890625, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.058091163635254, "learning_rate": 1e-05, "loss": 0.488, "mean_token_accuracy": 0.8701179027557373, "num_tokens": 413208423.0, "step": 10877 }, { "epoch": 1.383793410507569, "ewc_loss": 0.11181640625, "ewc_loss_diag": 0.00019931793212890625, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.268579483032227, "learning_rate": 1e-05, "loss": 0.5, "mean_token_accuracy": 0.8688735961914062, "num_tokens": 413238133.0, "step": 10878 }, { "epoch": 1.3839206207861596, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.00019931793212890625, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.9015398025512695, "learning_rate": 1e-05, "loss": 0.4626, "mean_token_accuracy": 0.8752801418304443, "num_tokens": 413273653.0, "step": 10879 }, { "epoch": 1.3840478310647502, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.00019931793212890625, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.28236722946167, "learning_rate": 1e-05, "loss": 0.4833, "mean_token_accuracy": 0.8707897663116455, "num_tokens": 413316492.0, "step": 10880 }, { "epoch": 1.3841750413433405, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.00019931793212890625, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.729807376861572, "learning_rate": 1e-05, "loss": 0.4578, "mean_token_accuracy": 0.8785161972045898, "num_tokens": 413359101.0, "step": 10881 }, { "epoch": 1.384302251621931, "ewc_loss": 0.11474609375, "ewc_loss_diag": 0.00019931793212890625, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.427318572998047, "learning_rate": 1e-05, "loss": 0.4539, "mean_token_accuracy": 0.8804222345352173, "num_tokens": 413396618.0, "step": 10882 }, { "epoch": 1.3844294619005215, "ewc_loss": 0.10009765625, "ewc_loss_diag": 0.00019931793212890625, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.581642150878906, "learning_rate": 1e-05, "loss": 0.4514, "mean_token_accuracy": 0.8797860145568848, "num_tokens": 413431863.0, "step": 10883 }, { "epoch": 1.384556672179112, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.00019931793212890625, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.698388576507568, "learning_rate": 1e-05, "loss": 0.5049, "mean_token_accuracy": 0.865350604057312, "num_tokens": 413470137.0, "step": 10884 }, { "epoch": 1.3846838824577026, "ewc_loss": 0.099609375, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.507643222808838, "learning_rate": 1e-05, "loss": 0.4484, "mean_token_accuracy": 0.8797262907028198, "num_tokens": 413508475.0, "step": 10885 }, { "epoch": 1.384811092736293, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 7.020813465118408, "learning_rate": 1e-05, "loss": 0.5109, "mean_token_accuracy": 0.8674907088279724, "num_tokens": 413548846.0, "step": 10886 }, { "epoch": 1.3849383030148836, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.651400089263916, "learning_rate": 1e-05, "loss": 0.4858, "mean_token_accuracy": 0.8670973777770996, "num_tokens": 413587375.0, "step": 10887 }, { "epoch": 1.3850655132934742, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 0.00011348724365234375, "grad_norm": 7.212822437286377, "learning_rate": 1e-05, "loss": 0.5125, "mean_token_accuracy": 0.8709012269973755, "num_tokens": 413626701.0, "step": 10888 }, { "epoch": 1.3851927235720647, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.002749919891357, "learning_rate": 1e-05, "loss": 0.5464, "mean_token_accuracy": 0.8545934557914734, "num_tokens": 413665207.0, "step": 10889 }, { "epoch": 1.385319933850655, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 7.022331237792969, "learning_rate": 1e-05, "loss": 0.4604, "mean_token_accuracy": 0.8835562467575073, "num_tokens": 413700549.0, "step": 10890 }, { "epoch": 1.3854471441292455, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.25968074798584, "learning_rate": 1e-05, "loss": 0.5887, "mean_token_accuracy": 0.8403060436248779, "num_tokens": 413738308.0, "step": 10891 }, { "epoch": 1.385574354407836, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.554632663726807, "learning_rate": 1e-05, "loss": 0.4907, "mean_token_accuracy": 0.8709432482719421, "num_tokens": 413779540.0, "step": 10892 }, { "epoch": 1.3857015646864266, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.335170269012451, "learning_rate": 1e-05, "loss": 0.4403, "mean_token_accuracy": 0.8871159553527832, "num_tokens": 413818677.0, "step": 10893 }, { "epoch": 1.385828774965017, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 6.049375057220459, "learning_rate": 1e-05, "loss": 0.5055, "mean_token_accuracy": 0.8649513721466064, "num_tokens": 413861397.0, "step": 10894 }, { "epoch": 1.3859559852436076, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.461432456970215, "learning_rate": 1e-05, "loss": 0.4721, "mean_token_accuracy": 0.8747891187667847, "num_tokens": 413897629.0, "step": 10895 }, { "epoch": 1.3860831955221982, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.6724724769592285, "learning_rate": 1e-05, "loss": 0.5371, "mean_token_accuracy": 0.8520337343215942, "num_tokens": 413935711.0, "step": 10896 }, { "epoch": 1.3862104058007887, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.732641220092773, "learning_rate": 1e-05, "loss": 0.483, "mean_token_accuracy": 0.8779281973838806, "num_tokens": 413977340.0, "step": 10897 }, { "epoch": 1.3863376160793792, "ewc_loss": 0.0986328125, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.472412586212158, "learning_rate": 1e-05, "loss": 0.4662, "mean_token_accuracy": 0.8783402442932129, "num_tokens": 414014429.0, "step": 10898 }, { "epoch": 1.3864648263579697, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 0.00011014938354492188, "grad_norm": 7.124046325683594, "learning_rate": 1e-05, "loss": 0.5558, "mean_token_accuracy": 0.857574462890625, "num_tokens": 414055426.0, "step": 10899 }, { "epoch": 1.3865920366365603, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.626444339752197, "learning_rate": 1e-05, "loss": 0.5055, "mean_token_accuracy": 0.8605388402938843, "num_tokens": 414086549.0, "step": 10900 }, { "epoch": 1.3867192469151508, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 0.00011968612670898438, "grad_norm": 7.3695292472839355, "learning_rate": 1e-05, "loss": 0.5002, "mean_token_accuracy": 0.875670313835144, "num_tokens": 414121767.0, "step": 10901 }, { "epoch": 1.3868464571937413, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.039715766906738, "learning_rate": 1e-05, "loss": 0.515, "mean_token_accuracy": 0.8609400987625122, "num_tokens": 414154201.0, "step": 10902 }, { "epoch": 1.3869736674723319, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 0.00011301040649414062, "grad_norm": 7.223275184631348, "learning_rate": 1e-05, "loss": 0.554, "mean_token_accuracy": 0.8567426204681396, "num_tokens": 414189898.0, "step": 10903 }, { "epoch": 1.3871008777509224, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.294231414794922, "learning_rate": 1e-05, "loss": 0.503, "mean_token_accuracy": 0.8719240427017212, "num_tokens": 414226507.0, "step": 10904 }, { "epoch": 1.387228088029513, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.856506824493408, "learning_rate": 1e-05, "loss": 0.5109, "mean_token_accuracy": 0.8648122549057007, "num_tokens": 414261909.0, "step": 10905 }, { "epoch": 1.3873552983081032, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.285327434539795, "learning_rate": 1e-05, "loss": 0.5167, "mean_token_accuracy": 0.8612436652183533, "num_tokens": 414303296.0, "step": 10906 }, { "epoch": 1.3874825085866938, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.536721706390381, "learning_rate": 1e-05, "loss": 0.5155, "mean_token_accuracy": 0.8670141100883484, "num_tokens": 414340638.0, "step": 10907 }, { "epoch": 1.3876097188652843, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.168087959289551, "learning_rate": 1e-05, "loss": 0.4955, "mean_token_accuracy": 0.8674542903900146, "num_tokens": 414376900.0, "step": 10908 }, { "epoch": 1.3877369291438748, "ewc_loss": 0.11376953125, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.340075969696045, "learning_rate": 1e-05, "loss": 0.4914, "mean_token_accuracy": 0.8712028861045837, "num_tokens": 414410657.0, "step": 10909 }, { "epoch": 1.3878641394224653, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 6.0275349617004395, "learning_rate": 1e-05, "loss": 0.5276, "mean_token_accuracy": 0.8519525527954102, "num_tokens": 414443165.0, "step": 10910 }, { "epoch": 1.3879913497010559, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.2805681228637695, "learning_rate": 1e-05, "loss": 0.5362, "mean_token_accuracy": 0.8538318276405334, "num_tokens": 414482185.0, "step": 10911 }, { "epoch": 1.3881185599796464, "ewc_loss": 0.10302734375, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.865245819091797, "learning_rate": 1e-05, "loss": 0.5459, "mean_token_accuracy": 0.8570096492767334, "num_tokens": 414517509.0, "step": 10912 }, { "epoch": 1.388245770258237, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.3270673751831055, "learning_rate": 1e-05, "loss": 0.49, "mean_token_accuracy": 0.8706613183021545, "num_tokens": 414558650.0, "step": 10913 }, { "epoch": 1.3883729805368275, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.704869747161865, "learning_rate": 1e-05, "loss": 0.4759, "mean_token_accuracy": 0.8725340962409973, "num_tokens": 414594301.0, "step": 10914 }, { "epoch": 1.3885001908154178, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.499816417694092, "learning_rate": 1e-05, "loss": 0.4857, "mean_token_accuracy": 0.8757116198539734, "num_tokens": 414626895.0, "step": 10915 }, { "epoch": 1.3886274010940083, "ewc_loss": 0.099609375, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.550270080566406, "learning_rate": 1e-05, "loss": 0.4625, "mean_token_accuracy": 0.8708440065383911, "num_tokens": 414664708.0, "step": 10916 }, { "epoch": 1.3887546113725988, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.785383224487305, "learning_rate": 1e-05, "loss": 0.567, "mean_token_accuracy": 0.8485234975814819, "num_tokens": 414706243.0, "step": 10917 }, { "epoch": 1.3888818216511893, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.547240257263184, "learning_rate": 1e-05, "loss": 0.4753, "mean_token_accuracy": 0.868378221988678, "num_tokens": 414740681.0, "step": 10918 }, { "epoch": 1.3890090319297799, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 0.00011110305786132812, "grad_norm": 7.0855937004089355, "learning_rate": 1e-05, "loss": 0.5016, "mean_token_accuracy": 0.8714226484298706, "num_tokens": 414774942.0, "step": 10919 }, { "epoch": 1.3891362422083704, "ewc_loss": 0.10595703125, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.72792911529541, "learning_rate": 1e-05, "loss": 0.4534, "mean_token_accuracy": 0.8789223432540894, "num_tokens": 414817369.0, "step": 10920 }, { "epoch": 1.389263452486961, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 0.00011396408081054688, "grad_norm": 7.203516960144043, "learning_rate": 1e-05, "loss": 0.5287, "mean_token_accuracy": 0.8639264702796936, "num_tokens": 414855503.0, "step": 10921 }, { "epoch": 1.3893906627655515, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.089454174041748, "learning_rate": 1e-05, "loss": 0.514, "mean_token_accuracy": 0.8623356819152832, "num_tokens": 414891974.0, "step": 10922 }, { "epoch": 1.389517873044142, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.90948486328125, "learning_rate": 1e-05, "loss": 0.4704, "mean_token_accuracy": 0.8795549869537354, "num_tokens": 414926773.0, "step": 10923 }, { "epoch": 1.3896450833227325, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.294066905975342, "learning_rate": 1e-05, "loss": 0.4724, "mean_token_accuracy": 0.8767836093902588, "num_tokens": 414962848.0, "step": 10924 }, { "epoch": 1.389772293601323, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.409813404083252, "learning_rate": 1e-05, "loss": 0.4793, "mean_token_accuracy": 0.8729652166366577, "num_tokens": 415003341.0, "step": 10925 }, { "epoch": 1.3898995038799136, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.375275611877441, "learning_rate": 1e-05, "loss": 0.4679, "mean_token_accuracy": 0.8783963322639465, "num_tokens": 415036836.0, "step": 10926 }, { "epoch": 1.390026714158504, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.916083335876465, "learning_rate": 1e-05, "loss": 0.4902, "mean_token_accuracy": 0.8695627450942993, "num_tokens": 415078589.0, "step": 10927 }, { "epoch": 1.3901539244370946, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.536549091339111, "learning_rate": 1e-05, "loss": 0.4587, "mean_token_accuracy": 0.8833179473876953, "num_tokens": 415114916.0, "step": 10928 }, { "epoch": 1.3902811347156852, "ewc_loss": 0.09912109375, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.576939582824707, "learning_rate": 1e-05, "loss": 0.481, "mean_token_accuracy": 0.8650597333908081, "num_tokens": 415153774.0, "step": 10929 }, { "epoch": 1.3904083449942755, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.86967658996582, "learning_rate": 1e-05, "loss": 0.5383, "mean_token_accuracy": 0.8546777367591858, "num_tokens": 415188359.0, "step": 10930 }, { "epoch": 1.390535555272866, "ewc_loss": 0.10009765625, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.477207183837891, "learning_rate": 1e-05, "loss": 0.4661, "mean_token_accuracy": 0.8737955093383789, "num_tokens": 415224423.0, "step": 10931 }, { "epoch": 1.3906627655514565, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 0.00011444091796875, "grad_norm": 7.26497220993042, "learning_rate": 1e-05, "loss": 0.5306, "mean_token_accuracy": 0.8635144829750061, "num_tokens": 415263219.0, "step": 10932 }, { "epoch": 1.390789975830047, "ewc_loss": 0.10693359375, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.7534685134887695, "learning_rate": 1e-05, "loss": 0.4967, "mean_token_accuracy": 0.8674405813217163, "num_tokens": 415301043.0, "step": 10933 }, { "epoch": 1.3909171861086376, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 0.00011873245239257812, "grad_norm": 7.405664443969727, "learning_rate": 1e-05, "loss": 0.5253, "mean_token_accuracy": 0.8670287728309631, "num_tokens": 415339128.0, "step": 10934 }, { "epoch": 1.391044396387228, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.194512367248535, "learning_rate": 1e-05, "loss": 0.5099, "mean_token_accuracy": 0.8688452243804932, "num_tokens": 415374574.0, "step": 10935 }, { "epoch": 1.3911716066658186, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 0.00011014938354492188, "grad_norm": 7.1417412757873535, "learning_rate": 1e-05, "loss": 0.5958, "mean_token_accuracy": 0.8411105871200562, "num_tokens": 415410133.0, "step": 10936 }, { "epoch": 1.3912988169444092, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.355776786804199, "learning_rate": 1e-05, "loss": 0.5359, "mean_token_accuracy": 0.8561049699783325, "num_tokens": 415451329.0, "step": 10937 }, { "epoch": 1.3914260272229997, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.758915424346924, "learning_rate": 1e-05, "loss": 0.5076, "mean_token_accuracy": 0.8675506114959717, "num_tokens": 415491448.0, "step": 10938 }, { "epoch": 1.39155323750159, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.2806782722473145, "learning_rate": 1e-05, "loss": 0.5006, "mean_token_accuracy": 0.869246244430542, "num_tokens": 415533583.0, "step": 10939 }, { "epoch": 1.3916804477801805, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.450880527496338, "learning_rate": 1e-05, "loss": 0.5275, "mean_token_accuracy": 0.8582000136375427, "num_tokens": 415567985.0, "step": 10940 }, { "epoch": 1.391807658058771, "ewc_loss": 0.11181640625, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.209667682647705, "learning_rate": 1e-05, "loss": 0.4814, "mean_token_accuracy": 0.8702886700630188, "num_tokens": 415603130.0, "step": 10941 }, { "epoch": 1.3919348683373616, "ewc_loss": 0.11181640625, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.2597432136535645, "learning_rate": 1e-05, "loss": 0.5223, "mean_token_accuracy": 0.8591062426567078, "num_tokens": 415641284.0, "step": 10942 }, { "epoch": 1.392062078615952, "ewc_loss": 0.10693359375, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 6.020590782165527, "learning_rate": 1e-05, "loss": 0.5358, "mean_token_accuracy": 0.8545584678649902, "num_tokens": 415682483.0, "step": 10943 }, { "epoch": 1.3921892888945426, "ewc_loss": 0.11083984375, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.2071075439453125, "learning_rate": 1e-05, "loss": 0.5517, "mean_token_accuracy": 0.8525044322013855, "num_tokens": 415723364.0, "step": 10944 }, { "epoch": 1.3923164991731332, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.857550621032715, "learning_rate": 1e-05, "loss": 0.5305, "mean_token_accuracy": 0.8585094809532166, "num_tokens": 415762750.0, "step": 10945 }, { "epoch": 1.3924437094517237, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.270398139953613, "learning_rate": 1e-05, "loss": 0.5234, "mean_token_accuracy": 0.8575966954231262, "num_tokens": 415801363.0, "step": 10946 }, { "epoch": 1.3925709197303142, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.686848163604736, "learning_rate": 1e-05, "loss": 0.4808, "mean_token_accuracy": 0.8723632097244263, "num_tokens": 415836121.0, "step": 10947 }, { "epoch": 1.3926981300089047, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.446479797363281, "learning_rate": 1e-05, "loss": 0.5329, "mean_token_accuracy": 0.8570325970649719, "num_tokens": 415880145.0, "step": 10948 }, { "epoch": 1.3928253402874953, "ewc_loss": 0.099609375, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.533392906188965, "learning_rate": 1e-05, "loss": 0.4902, "mean_token_accuracy": 0.8670188188552856, "num_tokens": 415921244.0, "step": 10949 }, { "epoch": 1.3929525505660858, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.767991542816162, "learning_rate": 1e-05, "loss": 0.4917, "mean_token_accuracy": 0.8687240481376648, "num_tokens": 415957166.0, "step": 10950 }, { "epoch": 1.3930797608446763, "ewc_loss": 0.0986328125, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.459522247314453, "learning_rate": 1e-05, "loss": 0.5012, "mean_token_accuracy": 0.859878659248352, "num_tokens": 415999462.0, "step": 10951 }, { "epoch": 1.3932069711232669, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 0.0001125335693359375, "grad_norm": 7.169167995452881, "learning_rate": 1e-05, "loss": 0.5108, "mean_token_accuracy": 0.8682725429534912, "num_tokens": 416038619.0, "step": 10952 }, { "epoch": 1.3933341814018574, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.6750078201293945, "learning_rate": 1e-05, "loss": 0.4938, "mean_token_accuracy": 0.8652703762054443, "num_tokens": 416084885.0, "step": 10953 }, { "epoch": 1.393461391680448, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 0.000118255615234375, "grad_norm": 7.379774570465088, "learning_rate": 1e-05, "loss": 0.567, "mean_token_accuracy": 0.8575263023376465, "num_tokens": 416126863.0, "step": 10954 }, { "epoch": 1.3935886019590382, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.135680675506592, "learning_rate": 1e-05, "loss": 0.4263, "mean_token_accuracy": 0.891239583492279, "num_tokens": 416160888.0, "step": 10955 }, { "epoch": 1.3937158122376287, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 7.185669422149658, "learning_rate": 1e-05, "loss": 0.4943, "mean_token_accuracy": 0.8736932277679443, "num_tokens": 416199012.0, "step": 10956 }, { "epoch": 1.3938430225162193, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.407279014587402, "learning_rate": 1e-05, "loss": 0.4717, "mean_token_accuracy": 0.8764474391937256, "num_tokens": 416234078.0, "step": 10957 }, { "epoch": 1.3939702327948098, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.628468990325928, "learning_rate": 1e-05, "loss": 0.4749, "mean_token_accuracy": 0.874731719493866, "num_tokens": 416272912.0, "step": 10958 }, { "epoch": 1.3940974430734003, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.412935733795166, "learning_rate": 1e-05, "loss": 0.4783, "mean_token_accuracy": 0.878083348274231, "num_tokens": 416312829.0, "step": 10959 }, { "epoch": 1.3942246533519909, "ewc_loss": 0.10888671875, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.140721797943115, "learning_rate": 1e-05, "loss": 0.487, "mean_token_accuracy": 0.8683947920799255, "num_tokens": 416353337.0, "step": 10960 }, { "epoch": 1.3943518636305814, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.477834224700928, "learning_rate": 1e-05, "loss": 0.5109, "mean_token_accuracy": 0.864871621131897, "num_tokens": 416392425.0, "step": 10961 }, { "epoch": 1.394479073909172, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.78228759765625, "learning_rate": 1e-05, "loss": 0.5184, "mean_token_accuracy": 0.8578615784645081, "num_tokens": 416434641.0, "step": 10962 }, { "epoch": 1.3946062841877624, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.6420063972473145, "learning_rate": 1e-05, "loss": 0.5024, "mean_token_accuracy": 0.8662382960319519, "num_tokens": 416479107.0, "step": 10963 }, { "epoch": 1.3947334944663528, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.6125712394714355, "learning_rate": 1e-05, "loss": 0.5159, "mean_token_accuracy": 0.8635109663009644, "num_tokens": 416512523.0, "step": 10964 }, { "epoch": 1.3948607047449433, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.897576332092285, "learning_rate": 1e-05, "loss": 0.518, "mean_token_accuracy": 0.8681029081344604, "num_tokens": 416550176.0, "step": 10965 }, { "epoch": 1.3949879150235338, "ewc_loss": 0.10205078125, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.601809978485107, "learning_rate": 1e-05, "loss": 0.4864, "mean_token_accuracy": 0.8673956394195557, "num_tokens": 416583878.0, "step": 10966 }, { "epoch": 1.3951151253021243, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 7.169899940490723, "learning_rate": 1e-05, "loss": 0.5192, "mean_token_accuracy": 0.8643956184387207, "num_tokens": 416612863.0, "step": 10967 }, { "epoch": 1.3952423355807149, "ewc_loss": 0.10791015625, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 5.825974941253662, "learning_rate": 1e-05, "loss": 0.5563, "mean_token_accuracy": 0.8537604808807373, "num_tokens": 416650691.0, "step": 10968 }, { "epoch": 1.3953695458593054, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 0.00011301040649414062, "grad_norm": 7.188640594482422, "learning_rate": 1e-05, "loss": 0.4977, "mean_token_accuracy": 0.8726754188537598, "num_tokens": 416687227.0, "step": 10969 }, { "epoch": 1.395496756137896, "ewc_loss": 0.11376953125, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.080052852630615, "learning_rate": 1e-05, "loss": 0.5262, "mean_token_accuracy": 0.8598302602767944, "num_tokens": 416725712.0, "step": 10970 }, { "epoch": 1.3956239664164865, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.993236064910889, "learning_rate": 1e-05, "loss": 0.5087, "mean_token_accuracy": 0.8757618069648743, "num_tokens": 416756169.0, "step": 10971 }, { "epoch": 1.395751176695077, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.196612358093262, "learning_rate": 1e-05, "loss": 0.561, "mean_token_accuracy": 0.8473489284515381, "num_tokens": 416794351.0, "step": 10972 }, { "epoch": 1.3958783869736675, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.676952362060547, "learning_rate": 1e-05, "loss": 0.4628, "mean_token_accuracy": 0.8819886445999146, "num_tokens": 416829201.0, "step": 10973 }, { "epoch": 1.396005597252258, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.172694206237793, "learning_rate": 1e-05, "loss": 0.4778, "mean_token_accuracy": 0.8756831884384155, "num_tokens": 416862171.0, "step": 10974 }, { "epoch": 1.3961328075308486, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.424436569213867, "learning_rate": 1e-05, "loss": 0.53, "mean_token_accuracy": 0.857688307762146, "num_tokens": 416895923.0, "step": 10975 }, { "epoch": 1.396260017809439, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.093882083892822, "learning_rate": 1e-05, "loss": 0.5236, "mean_token_accuracy": 0.8612465858459473, "num_tokens": 416927441.0, "step": 10976 }, { "epoch": 1.3963872280880296, "ewc_loss": 0.11181640625, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.261303424835205, "learning_rate": 1e-05, "loss": 0.4891, "mean_token_accuracy": 0.8710511326789856, "num_tokens": 416962911.0, "step": 10977 }, { "epoch": 1.3965144383666201, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.963895797729492, "learning_rate": 1e-05, "loss": 0.5153, "mean_token_accuracy": 0.8615659475326538, "num_tokens": 417001403.0, "step": 10978 }, { "epoch": 1.3966416486452105, "ewc_loss": 0.11181640625, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.226473808288574, "learning_rate": 1e-05, "loss": 0.5225, "mean_token_accuracy": 0.8589950799942017, "num_tokens": 417038043.0, "step": 10979 }, { "epoch": 1.396768858923801, "ewc_loss": 0.10302734375, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.842666149139404, "learning_rate": 1e-05, "loss": 0.4924, "mean_token_accuracy": 0.8623708486557007, "num_tokens": 417073516.0, "step": 10980 }, { "epoch": 1.3968960692023915, "ewc_loss": 0.11181640625, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.25659704208374, "learning_rate": 1e-05, "loss": 0.4944, "mean_token_accuracy": 0.8663512468338013, "num_tokens": 417112334.0, "step": 10981 }, { "epoch": 1.397023279480982, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.704477310180664, "learning_rate": 1e-05, "loss": 0.4811, "mean_token_accuracy": 0.8691496849060059, "num_tokens": 417155273.0, "step": 10982 }, { "epoch": 1.3971504897595726, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.398933410644531, "learning_rate": 1e-05, "loss": 0.4755, "mean_token_accuracy": 0.8766429424285889, "num_tokens": 417190145.0, "step": 10983 }, { "epoch": 1.397277700038163, "ewc_loss": 0.10009765625, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.573266506195068, "learning_rate": 1e-05, "loss": 0.4324, "mean_token_accuracy": 0.881560206413269, "num_tokens": 417228184.0, "step": 10984 }, { "epoch": 1.3974049103167536, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.647527694702148, "learning_rate": 1e-05, "loss": 0.4655, "mean_token_accuracy": 0.8801203966140747, "num_tokens": 417258590.0, "step": 10985 }, { "epoch": 1.3975321205953442, "ewc_loss": 0.099609375, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.501382350921631, "learning_rate": 1e-05, "loss": 0.4466, "mean_token_accuracy": 0.8794988393783569, "num_tokens": 417295969.0, "step": 10986 }, { "epoch": 1.3976593308739347, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.977768421173096, "learning_rate": 1e-05, "loss": 0.5616, "mean_token_accuracy": 0.8517877459526062, "num_tokens": 417334428.0, "step": 10987 }, { "epoch": 1.397786541152525, "ewc_loss": 0.10302734375, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.607355117797852, "learning_rate": 1e-05, "loss": 0.4801, "mean_token_accuracy": 0.8709673881530762, "num_tokens": 417370074.0, "step": 10988 }, { "epoch": 1.3979137514311155, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 0.00011348724365234375, "grad_norm": 7.1995530128479, "learning_rate": 1e-05, "loss": 0.4893, "mean_token_accuracy": 0.872887134552002, "num_tokens": 417413544.0, "step": 10989 }, { "epoch": 1.398040961709706, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 5.927255630493164, "learning_rate": 1e-05, "loss": 0.5118, "mean_token_accuracy": 0.8638942837715149, "num_tokens": 417445893.0, "step": 10990 }, { "epoch": 1.3981681719882966, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 7.081889629364014, "learning_rate": 1e-05, "loss": 0.4837, "mean_token_accuracy": 0.8785426616668701, "num_tokens": 417481469.0, "step": 10991 }, { "epoch": 1.398295382266887, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.210328102111816, "learning_rate": 1e-05, "loss": 0.5249, "mean_token_accuracy": 0.8605362772941589, "num_tokens": 417517070.0, "step": 10992 }, { "epoch": 1.3984225925454776, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.661879539489746, "learning_rate": 1e-05, "loss": 0.5109, "mean_token_accuracy": 0.8639956116676331, "num_tokens": 417558265.0, "step": 10993 }, { "epoch": 1.3985498028240682, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.312487602233887, "learning_rate": 1e-05, "loss": 0.4788, "mean_token_accuracy": 0.8720386028289795, "num_tokens": 417599083.0, "step": 10994 }, { "epoch": 1.3986770131026587, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.167229175567627, "learning_rate": 1e-05, "loss": 0.4793, "mean_token_accuracy": 0.8795878887176514, "num_tokens": 417632425.0, "step": 10995 }, { "epoch": 1.3988042233812492, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.407263278961182, "learning_rate": 1e-05, "loss": 0.4707, "mean_token_accuracy": 0.8750863075256348, "num_tokens": 417667209.0, "step": 10996 }, { "epoch": 1.3989314336598397, "ewc_loss": 0.10302734375, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.764235019683838, "learning_rate": 1e-05, "loss": 0.4744, "mean_token_accuracy": 0.8727365732192993, "num_tokens": 417706168.0, "step": 10997 }, { "epoch": 1.3990586439384303, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.616523742675781, "learning_rate": 1e-05, "loss": 0.4778, "mean_token_accuracy": 0.8760049343109131, "num_tokens": 417745225.0, "step": 10998 }, { "epoch": 1.3991858542170208, "ewc_loss": 0.09912109375, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.531966686248779, "learning_rate": 1e-05, "loss": 0.4755, "mean_token_accuracy": 0.873171329498291, "num_tokens": 417782688.0, "step": 10999 }, { "epoch": 1.3993130644956113, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.960328578948975, "learning_rate": 1e-05, "loss": 0.5039, "mean_token_accuracy": 0.8673078417778015, "num_tokens": 417814157.0, "step": 11000 }, { "epoch": 1.3994402747742019, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.556678295135498, "learning_rate": 1e-05, "loss": 0.5158, "mean_token_accuracy": 0.8613684177398682, "num_tokens": 417845460.0, "step": 11001 }, { "epoch": 1.3995674850527924, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 0.00011491775512695312, "grad_norm": 7.252786636352539, "learning_rate": 1e-05, "loss": 0.5119, "mean_token_accuracy": 0.8714499473571777, "num_tokens": 417881460.0, "step": 11002 }, { "epoch": 1.399694695331383, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 5.844362735748291, "learning_rate": 1e-05, "loss": 0.4397, "mean_token_accuracy": 0.8855758905410767, "num_tokens": 417919690.0, "step": 11003 }, { "epoch": 1.3998219056099732, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 0.00011539459228515625, "grad_norm": 7.27090311050415, "learning_rate": 1e-05, "loss": 0.459, "mean_token_accuracy": 0.8858546614646912, "num_tokens": 417959474.0, "step": 11004 }, { "epoch": 1.3999491158885637, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.17182731628418, "learning_rate": 1e-05, "loss": 0.5011, "mean_token_accuracy": 0.8673925399780273, "num_tokens": 417992342.0, "step": 11005 }, { "epoch": 1.4000763261671543, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 6.996640205383301, "learning_rate": 1e-05, "loss": 0.5727, "mean_token_accuracy": 0.8512074947357178, "num_tokens": 418032006.0, "step": 11006 }, { "epoch": 1.4002035364457448, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.26287841796875, "learning_rate": 1e-05, "loss": 0.5459, "mean_token_accuracy": 0.8557727932929993, "num_tokens": 418069318.0, "step": 11007 }, { "epoch": 1.4003307467243353, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.655256748199463, "learning_rate": 1e-05, "loss": 0.4577, "mean_token_accuracy": 0.8820902705192566, "num_tokens": 418103605.0, "step": 11008 }, { "epoch": 1.4004579570029259, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.2047576904296875, "learning_rate": 1e-05, "loss": 0.5091, "mean_token_accuracy": 0.862833559513092, "num_tokens": 418136363.0, "step": 11009 }, { "epoch": 1.4005851672815164, "ewc_loss": 0.11376953125, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.38287878036499, "learning_rate": 1e-05, "loss": 0.4947, "mean_token_accuracy": 0.8692469000816345, "num_tokens": 418178096.0, "step": 11010 }, { "epoch": 1.400712377560107, "ewc_loss": 0.11181640625, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.109411716461182, "learning_rate": 1e-05, "loss": 0.4523, "mean_token_accuracy": 0.8805422186851501, "num_tokens": 418211025.0, "step": 11011 }, { "epoch": 1.4008395878386974, "ewc_loss": 0.11083984375, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.233304023742676, "learning_rate": 1e-05, "loss": 0.4874, "mean_token_accuracy": 0.8708023428916931, "num_tokens": 418251210.0, "step": 11012 }, { "epoch": 1.4009667981172877, "ewc_loss": 0.10595703125, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.97429895401001, "learning_rate": 1e-05, "loss": 0.4662, "mean_token_accuracy": 0.8774349689483643, "num_tokens": 418286593.0, "step": 11013 }, { "epoch": 1.4010940083958783, "ewc_loss": 0.10888671875, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.2066731452941895, "learning_rate": 1e-05, "loss": 0.4619, "mean_token_accuracy": 0.880347490310669, "num_tokens": 418324794.0, "step": 11014 }, { "epoch": 1.4012212186744688, "ewc_loss": 0.10400390625, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.824947834014893, "learning_rate": 1e-05, "loss": 0.4686, "mean_token_accuracy": 0.8747178316116333, "num_tokens": 418366067.0, "step": 11015 }, { "epoch": 1.4013484289530593, "ewc_loss": 0.11181640625, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.276779651641846, "learning_rate": 1e-05, "loss": 0.4381, "mean_token_accuracy": 0.8873183131217957, "num_tokens": 418406600.0, "step": 11016 }, { "epoch": 1.4014756392316499, "ewc_loss": 0.10107421875, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.682947635650635, "learning_rate": 1e-05, "loss": 0.4639, "mean_token_accuracy": 0.873252809047699, "num_tokens": 418440168.0, "step": 11017 }, { "epoch": 1.4016028495102404, "ewc_loss": 0.11474609375, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.468364715576172, "learning_rate": 1e-05, "loss": 0.542, "mean_token_accuracy": 0.8506616353988647, "num_tokens": 418482247.0, "step": 11018 }, { "epoch": 1.401730059788831, "ewc_loss": 0.09912109375, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.538806438446045, "learning_rate": 1e-05, "loss": 0.4553, "mean_token_accuracy": 0.8804135918617249, "num_tokens": 418517376.0, "step": 11019 }, { "epoch": 1.4018572700674214, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.765424728393555, "learning_rate": 1e-05, "loss": 0.5367, "mean_token_accuracy": 0.8635432720184326, "num_tokens": 418555841.0, "step": 11020 }, { "epoch": 1.401984480346012, "ewc_loss": 0.10009765625, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.486042022705078, "learning_rate": 1e-05, "loss": 0.4809, "mean_token_accuracy": 0.8696737289428711, "num_tokens": 418594232.0, "step": 11021 }, { "epoch": 1.4021116906246025, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 0.0001125335693359375, "grad_norm": 7.130222320556641, "learning_rate": 1e-05, "loss": 0.4813, "mean_token_accuracy": 0.8764691352844238, "num_tokens": 418631726.0, "step": 11022 }, { "epoch": 1.402238900903193, "ewc_loss": 0.10498046875, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.672403335571289, "learning_rate": 1e-05, "loss": 0.4244, "mean_token_accuracy": 0.8917213678359985, "num_tokens": 418671172.0, "step": 11023 }, { "epoch": 1.4023661111817836, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 0.00011730194091796875, "grad_norm": 7.33636474609375, "learning_rate": 1e-05, "loss": 0.5001, "mean_token_accuracy": 0.8754199743270874, "num_tokens": 418706541.0, "step": 11024 }, { "epoch": 1.402493321460374, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.089168548583984, "learning_rate": 1e-05, "loss": 0.4933, "mean_token_accuracy": 0.8749071955680847, "num_tokens": 418743557.0, "step": 11025 }, { "epoch": 1.4026205317389646, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 0.00010967254638671875, "grad_norm": 7.100064277648926, "learning_rate": 1e-05, "loss": 0.5154, "mean_token_accuracy": 0.8678884506225586, "num_tokens": 418783741.0, "step": 11026 }, { "epoch": 1.4027477420175551, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.328527927398682, "learning_rate": 1e-05, "loss": 0.4737, "mean_token_accuracy": 0.8769688606262207, "num_tokens": 418822208.0, "step": 11027 }, { "epoch": 1.4028749522961454, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.622487545013428, "learning_rate": 1e-05, "loss": 0.4834, "mean_token_accuracy": 0.8751062154769897, "num_tokens": 418862717.0, "step": 11028 }, { "epoch": 1.403002162574736, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.386101245880127, "learning_rate": 1e-05, "loss": 0.4882, "mean_token_accuracy": 0.8731880187988281, "num_tokens": 418892961.0, "step": 11029 }, { "epoch": 1.4031293728533265, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 6.141345024108887, "learning_rate": 1e-05, "loss": 0.5376, "mean_token_accuracy": 0.8525382280349731, "num_tokens": 418933384.0, "step": 11030 }, { "epoch": 1.403256583131917, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.445296287536621, "learning_rate": 1e-05, "loss": 0.4842, "mean_token_accuracy": 0.8745014071464539, "num_tokens": 418971042.0, "step": 11031 }, { "epoch": 1.4033837934105076, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.769155502319336, "learning_rate": 1e-05, "loss": 0.4613, "mean_token_accuracy": 0.8736777305603027, "num_tokens": 419009949.0, "step": 11032 }, { "epoch": 1.403511003689098, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.629920482635498, "learning_rate": 1e-05, "loss": 0.4811, "mean_token_accuracy": 0.8755843639373779, "num_tokens": 419044607.0, "step": 11033 }, { "epoch": 1.4036382139676886, "ewc_loss": 0.10009765625, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.578688621520996, "learning_rate": 1e-05, "loss": 0.4729, "mean_token_accuracy": 0.873027503490448, "num_tokens": 419081407.0, "step": 11034 }, { "epoch": 1.4037654242462791, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.891765594482422, "learning_rate": 1e-05, "loss": 0.469, "mean_token_accuracy": 0.8778724670410156, "num_tokens": 419125007.0, "step": 11035 }, { "epoch": 1.4038926345248697, "ewc_loss": 0.10205078125, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.595322608947754, "learning_rate": 1e-05, "loss": 0.5082, "mean_token_accuracy": 0.8730937838554382, "num_tokens": 419166009.0, "step": 11036 }, { "epoch": 1.40401984480346, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 0.00011110305786132812, "grad_norm": 7.153234004974365, "learning_rate": 1e-05, "loss": 0.5123, "mean_token_accuracy": 0.8681923151016235, "num_tokens": 419194185.0, "step": 11037 }, { "epoch": 1.4041470550820505, "ewc_loss": 0.10791015625, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 5.787337303161621, "learning_rate": 1e-05, "loss": 0.4763, "mean_token_accuracy": 0.8733915090560913, "num_tokens": 419236742.0, "step": 11038 }, { "epoch": 1.404274265360641, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 0.00011205673217773438, "grad_norm": 7.209043979644775, "learning_rate": 1e-05, "loss": 0.4965, "mean_token_accuracy": 0.8698285818099976, "num_tokens": 419272477.0, "step": 11039 }, { "epoch": 1.4044014756392316, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.071604251861572, "learning_rate": 1e-05, "loss": 0.5172, "mean_token_accuracy": 0.8620823621749878, "num_tokens": 419309198.0, "step": 11040 }, { "epoch": 1.404528685917822, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 7.021270275115967, "learning_rate": 1e-05, "loss": 0.473, "mean_token_accuracy": 0.8775840997695923, "num_tokens": 419343065.0, "step": 11041 }, { "epoch": 1.4046558961964126, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.19364070892334, "learning_rate": 1e-05, "loss": 0.5295, "mean_token_accuracy": 0.8592289090156555, "num_tokens": 419382826.0, "step": 11042 }, { "epoch": 1.4047831064750032, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.704048156738281, "learning_rate": 1e-05, "loss": 0.5226, "mean_token_accuracy": 0.8606352806091309, "num_tokens": 419416839.0, "step": 11043 }, { "epoch": 1.4049103167535937, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.180842399597168, "learning_rate": 1e-05, "loss": 0.4806, "mean_token_accuracy": 0.8706547021865845, "num_tokens": 419448041.0, "step": 11044 }, { "epoch": 1.4050375270321842, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.431217193603516, "learning_rate": 1e-05, "loss": 0.5716, "mean_token_accuracy": 0.8456157445907593, "num_tokens": 419484917.0, "step": 11045 }, { "epoch": 1.4051647373107747, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.095458030700684, "learning_rate": 1e-05, "loss": 0.4834, "mean_token_accuracy": 0.8687223196029663, "num_tokens": 419521329.0, "step": 11046 }, { "epoch": 1.4052919475893653, "ewc_loss": 0.11181640625, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.259705066680908, "learning_rate": 1e-05, "loss": 0.4956, "mean_token_accuracy": 0.8671772480010986, "num_tokens": 419556141.0, "step": 11047 }, { "epoch": 1.4054191578679558, "ewc_loss": 0.10693359375, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.980580806732178, "learning_rate": 1e-05, "loss": 0.4768, "mean_token_accuracy": 0.8723973035812378, "num_tokens": 419593027.0, "step": 11048 }, { "epoch": 1.4055463681465463, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.209274768829346, "learning_rate": 1e-05, "loss": 0.4645, "mean_token_accuracy": 0.8789702653884888, "num_tokens": 419632419.0, "step": 11049 }, { "epoch": 1.4056735784251368, "ewc_loss": 0.10400390625, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.847280025482178, "learning_rate": 1e-05, "loss": 0.4952, "mean_token_accuracy": 0.8649888038635254, "num_tokens": 419677625.0, "step": 11050 }, { "epoch": 1.4058007887037274, "ewc_loss": 0.11181640625, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.261438846588135, "learning_rate": 1e-05, "loss": 0.4951, "mean_token_accuracy": 0.8669142127037048, "num_tokens": 419721546.0, "step": 11051 }, { "epoch": 1.405927998982318, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.714632987976074, "learning_rate": 1e-05, "loss": 0.5376, "mean_token_accuracy": 0.8515130281448364, "num_tokens": 419759345.0, "step": 11052 }, { "epoch": 1.4060552092609082, "ewc_loss": 0.11474609375, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.409897327423096, "learning_rate": 1e-05, "loss": 0.511, "mean_token_accuracy": 0.8657971024513245, "num_tokens": 419789819.0, "step": 11053 }, { "epoch": 1.4061824195394987, "ewc_loss": 0.10009765625, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.574176788330078, "learning_rate": 1e-05, "loss": 0.4929, "mean_token_accuracy": 0.8662658929824829, "num_tokens": 419830525.0, "step": 11054 }, { "epoch": 1.4063096298180893, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.651280403137207, "learning_rate": 1e-05, "loss": 0.4995, "mean_token_accuracy": 0.8691661357879639, "num_tokens": 419875592.0, "step": 11055 }, { "epoch": 1.4064368400966798, "ewc_loss": 0.10009765625, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.491776943206787, "learning_rate": 1e-05, "loss": 0.4317, "mean_token_accuracy": 0.8863028287887573, "num_tokens": 419913530.0, "step": 11056 }, { "epoch": 1.4065640503752703, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.995485305786133, "learning_rate": 1e-05, "loss": 0.5156, "mean_token_accuracy": 0.8678072094917297, "num_tokens": 419955947.0, "step": 11057 }, { "epoch": 1.4066912606538609, "ewc_loss": 0.10400390625, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.610649585723877, "learning_rate": 1e-05, "loss": 0.4537, "mean_token_accuracy": 0.8796718120574951, "num_tokens": 419986931.0, "step": 11058 }, { "epoch": 1.4068184709324514, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 0.00011348724365234375, "grad_norm": 7.243445873260498, "learning_rate": 1e-05, "loss": 0.5129, "mean_token_accuracy": 0.8700222373008728, "num_tokens": 420024037.0, "step": 11059 }, { "epoch": 1.406945681211042, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 5.944805145263672, "learning_rate": 1e-05, "loss": 0.5083, "mean_token_accuracy": 0.864557147026062, "num_tokens": 420062843.0, "step": 11060 }, { "epoch": 1.4070728914896324, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 7.131683826446533, "learning_rate": 1e-05, "loss": 0.5621, "mean_token_accuracy": 0.8538358211517334, "num_tokens": 420102311.0, "step": 11061 }, { "epoch": 1.4072001017682227, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.234013557434082, "learning_rate": 1e-05, "loss": 0.5188, "mean_token_accuracy": 0.8622174263000488, "num_tokens": 420143364.0, "step": 11062 }, { "epoch": 1.4073273120468133, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.7173566818237305, "learning_rate": 1e-05, "loss": 0.4856, "mean_token_accuracy": 0.8763138651847839, "num_tokens": 420180930.0, "step": 11063 }, { "epoch": 1.4074545223254038, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.332706451416016, "learning_rate": 1e-05, "loss": 0.4928, "mean_token_accuracy": 0.8703886270523071, "num_tokens": 420216048.0, "step": 11064 }, { "epoch": 1.4075817326039943, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.22370719909668, "learning_rate": 1e-05, "loss": 0.5025, "mean_token_accuracy": 0.8705767393112183, "num_tokens": 420250792.0, "step": 11065 }, { "epoch": 1.4077089428825849, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.428953170776367, "learning_rate": 1e-05, "loss": 0.4813, "mean_token_accuracy": 0.8733392953872681, "num_tokens": 420280518.0, "step": 11066 }, { "epoch": 1.4078361531611754, "ewc_loss": 0.10302734375, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.7961883544921875, "learning_rate": 1e-05, "loss": 0.5285, "mean_token_accuracy": 0.8517417907714844, "num_tokens": 420318274.0, "step": 11067 }, { "epoch": 1.407963363439766, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.631287097930908, "learning_rate": 1e-05, "loss": 0.4799, "mean_token_accuracy": 0.8723073601722717, "num_tokens": 420353277.0, "step": 11068 }, { "epoch": 1.4080905737183564, "ewc_loss": 0.099609375, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.55265998840332, "learning_rate": 1e-05, "loss": 0.472, "mean_token_accuracy": 0.8761154413223267, "num_tokens": 420384457.0, "step": 11069 }, { "epoch": 1.408217783996947, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.945816993713379, "learning_rate": 1e-05, "loss": 0.5345, "mean_token_accuracy": 0.8586604595184326, "num_tokens": 420419229.0, "step": 11070 }, { "epoch": 1.4083449942755375, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.553896427154541, "learning_rate": 1e-05, "loss": 0.5092, "mean_token_accuracy": 0.8602566123008728, "num_tokens": 420457002.0, "step": 11071 }, { "epoch": 1.408472204554128, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 0.0001163482666015625, "grad_norm": 7.257948398590088, "learning_rate": 1e-05, "loss": 0.4807, "mean_token_accuracy": 0.878868043422699, "num_tokens": 420490802.0, "step": 11072 }, { "epoch": 1.4085994148327186, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 5.835063934326172, "learning_rate": 1e-05, "loss": 0.4658, "mean_token_accuracy": 0.8767341375350952, "num_tokens": 420531782.0, "step": 11073 }, { "epoch": 1.408726625111309, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 0.0001163482666015625, "grad_norm": 7.293871879577637, "learning_rate": 1e-05, "loss": 0.5533, "mean_token_accuracy": 0.8586694002151489, "num_tokens": 420575577.0, "step": 11074 }, { "epoch": 1.4088538353898996, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.1842360496521, "learning_rate": 1e-05, "loss": 0.4935, "mean_token_accuracy": 0.8728494644165039, "num_tokens": 420612963.0, "step": 11075 }, { "epoch": 1.4089810456684901, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 7.014384746551514, "learning_rate": 1e-05, "loss": 0.4764, "mean_token_accuracy": 0.8782209157943726, "num_tokens": 420649518.0, "step": 11076 }, { "epoch": 1.4091082559470804, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.29024076461792, "learning_rate": 1e-05, "loss": 0.5342, "mean_token_accuracy": 0.8587232232093811, "num_tokens": 420684377.0, "step": 11077 }, { "epoch": 1.409235466225671, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.671408176422119, "learning_rate": 1e-05, "loss": 0.5381, "mean_token_accuracy": 0.8597983121871948, "num_tokens": 420721566.0, "step": 11078 }, { "epoch": 1.4093626765042615, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.2238898277282715, "learning_rate": 1e-05, "loss": 0.4859, "mean_token_accuracy": 0.8696527481079102, "num_tokens": 420758591.0, "step": 11079 }, { "epoch": 1.409489886782852, "ewc_loss": 0.11376953125, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.391921520233154, "learning_rate": 1e-05, "loss": 0.5192, "mean_token_accuracy": 0.8611788153648376, "num_tokens": 420796864.0, "step": 11080 }, { "epoch": 1.4096170970614426, "ewc_loss": 0.11181640625, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.105120658874512, "learning_rate": 1e-05, "loss": 0.4721, "mean_token_accuracy": 0.8778596520423889, "num_tokens": 420834689.0, "step": 11081 }, { "epoch": 1.409744307340033, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.245210647583008, "learning_rate": 1e-05, "loss": 0.4963, "mean_token_accuracy": 0.869377851486206, "num_tokens": 420877795.0, "step": 11082 }, { "epoch": 1.4098715176186236, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.968272686004639, "learning_rate": 1e-05, "loss": 0.4777, "mean_token_accuracy": 0.8724928498268127, "num_tokens": 420917217.0, "step": 11083 }, { "epoch": 1.4099987278972141, "ewc_loss": 0.11083984375, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.240665435791016, "learning_rate": 1e-05, "loss": 0.5019, "mean_token_accuracy": 0.8677811622619629, "num_tokens": 420952607.0, "step": 11084 }, { "epoch": 1.4101259381758047, "ewc_loss": 0.10400390625, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.8256964683532715, "learning_rate": 1e-05, "loss": 0.5074, "mean_token_accuracy": 0.8604648113250732, "num_tokens": 420986570.0, "step": 11085 }, { "epoch": 1.410253148454395, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.311518669128418, "learning_rate": 1e-05, "loss": 0.4925, "mean_token_accuracy": 0.8702734708786011, "num_tokens": 421024482.0, "step": 11086 }, { "epoch": 1.4103803587329855, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.679827690124512, "learning_rate": 1e-05, "loss": 0.4084, "mean_token_accuracy": 0.8911815881729126, "num_tokens": 421058185.0, "step": 11087 }, { "epoch": 1.410507569011576, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.49256706237793, "learning_rate": 1e-05, "loss": 0.5087, "mean_token_accuracy": 0.8677281141281128, "num_tokens": 421089796.0, "step": 11088 }, { "epoch": 1.4106347792901666, "ewc_loss": 0.099609375, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.545740127563477, "learning_rate": 1e-05, "loss": 0.5017, "mean_token_accuracy": 0.863826334476471, "num_tokens": 421132644.0, "step": 11089 }, { "epoch": 1.410761989568757, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.790661334991455, "learning_rate": 1e-05, "loss": 0.5165, "mean_token_accuracy": 0.8652488589286804, "num_tokens": 421167790.0, "step": 11090 }, { "epoch": 1.4108891998473476, "ewc_loss": 0.099609375, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.508437633514404, "learning_rate": 1e-05, "loss": 0.4992, "mean_token_accuracy": 0.8635177612304688, "num_tokens": 421206765.0, "step": 11091 }, { "epoch": 1.4110164101259381, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 0.00011157989501953125, "grad_norm": 7.128264427185059, "learning_rate": 1e-05, "loss": 0.4871, "mean_token_accuracy": 0.877030611038208, "num_tokens": 421249997.0, "step": 11092 }, { "epoch": 1.4111436204045287, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.71497917175293, "learning_rate": 1e-05, "loss": 0.4747, "mean_token_accuracy": 0.8718695640563965, "num_tokens": 421287674.0, "step": 11093 }, { "epoch": 1.4112708306831192, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 0.00011539459228515625, "grad_norm": 7.267205238342285, "learning_rate": 1e-05, "loss": 0.4989, "mean_token_accuracy": 0.8734769225120544, "num_tokens": 421325494.0, "step": 11094 }, { "epoch": 1.4113980409617097, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.099859714508057, "learning_rate": 1e-05, "loss": 0.4602, "mean_token_accuracy": 0.8791258335113525, "num_tokens": 421359934.0, "step": 11095 }, { "epoch": 1.4115252512403003, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 7.005301475524902, "learning_rate": 1e-05, "loss": 0.5684, "mean_token_accuracy": 0.8549115061759949, "num_tokens": 421399195.0, "step": 11096 }, { "epoch": 1.4116524615188908, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.310729026794434, "learning_rate": 1e-05, "loss": 0.439, "mean_token_accuracy": 0.8873542547225952, "num_tokens": 421428822.0, "step": 11097 }, { "epoch": 1.4117796717974813, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.524235248565674, "learning_rate": 1e-05, "loss": 0.5278, "mean_token_accuracy": 0.8642566204071045, "num_tokens": 421468409.0, "step": 11098 }, { "epoch": 1.4119068820760718, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.376280307769775, "learning_rate": 1e-05, "loss": 0.5763, "mean_token_accuracy": 0.8486264944076538, "num_tokens": 421506905.0, "step": 11099 }, { "epoch": 1.4120340923546624, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 6.038801193237305, "learning_rate": 1e-05, "loss": 0.5372, "mean_token_accuracy": 0.8549484014511108, "num_tokens": 421545347.0, "step": 11100 }, { "epoch": 1.412161302633253, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.487273216247559, "learning_rate": 1e-05, "loss": 0.5003, "mean_token_accuracy": 0.8705984354019165, "num_tokens": 421580819.0, "step": 11101 }, { "epoch": 1.4122885129118432, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.685599327087402, "learning_rate": 1e-05, "loss": 0.4662, "mean_token_accuracy": 0.87152498960495, "num_tokens": 421628474.0, "step": 11102 }, { "epoch": 1.4124157231904337, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.727921962738037, "learning_rate": 1e-05, "loss": 0.5211, "mean_token_accuracy": 0.8609048128128052, "num_tokens": 421670437.0, "step": 11103 }, { "epoch": 1.4125429334690243, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.528786659240723, "learning_rate": 1e-05, "loss": 0.5436, "mean_token_accuracy": 0.8511260747909546, "num_tokens": 421709157.0, "step": 11104 }, { "epoch": 1.4126701437476148, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 7.069366455078125, "learning_rate": 1e-05, "loss": 0.4846, "mean_token_accuracy": 0.8766629695892334, "num_tokens": 421743783.0, "step": 11105 }, { "epoch": 1.4127973540262053, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.632728099822998, "learning_rate": 1e-05, "loss": 0.4373, "mean_token_accuracy": 0.8834489583969116, "num_tokens": 421781376.0, "step": 11106 }, { "epoch": 1.4129245643047958, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 0.00011730194091796875, "grad_norm": 7.304903507232666, "learning_rate": 1e-05, "loss": 0.4904, "mean_token_accuracy": 0.8806369304656982, "num_tokens": 421819367.0, "step": 11107 }, { "epoch": 1.4130517745833864, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 5.966500759124756, "learning_rate": 1e-05, "loss": 0.4543, "mean_token_accuracy": 0.8802969455718994, "num_tokens": 421862830.0, "step": 11108 }, { "epoch": 1.413178984861977, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 0.00011348724365234375, "grad_norm": 7.231174468994141, "learning_rate": 1e-05, "loss": 0.4911, "mean_token_accuracy": 0.8772563934326172, "num_tokens": 421898353.0, "step": 11109 }, { "epoch": 1.4133061951405674, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.244555473327637, "learning_rate": 1e-05, "loss": 0.4986, "mean_token_accuracy": 0.8682190775871277, "num_tokens": 421932911.0, "step": 11110 }, { "epoch": 1.4134334054191577, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.907258987426758, "learning_rate": 1e-05, "loss": 0.5098, "mean_token_accuracy": 0.871492326259613, "num_tokens": 421974409.0, "step": 11111 }, { "epoch": 1.4135606156977483, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.268099784851074, "learning_rate": 1e-05, "loss": 0.5557, "mean_token_accuracy": 0.8490908145904541, "num_tokens": 422011913.0, "step": 11112 }, { "epoch": 1.4136878259763388, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.57360315322876, "learning_rate": 1e-05, "loss": 0.5092, "mean_token_accuracy": 0.8699653148651123, "num_tokens": 422054487.0, "step": 11113 }, { "epoch": 1.4138150362549293, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.180166244506836, "learning_rate": 1e-05, "loss": 0.5057, "mean_token_accuracy": 0.8685311079025269, "num_tokens": 422092980.0, "step": 11114 }, { "epoch": 1.4139422465335199, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.346181869506836, "learning_rate": 1e-05, "loss": 0.5245, "mean_token_accuracy": 0.8600972890853882, "num_tokens": 422139368.0, "step": 11115 }, { "epoch": 1.4140694568121104, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.055157661437988, "learning_rate": 1e-05, "loss": 0.452, "mean_token_accuracy": 0.8811920285224915, "num_tokens": 422177342.0, "step": 11116 }, { "epoch": 1.414196667090701, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.257029056549072, "learning_rate": 1e-05, "loss": 0.4726, "mean_token_accuracy": 0.8748568296432495, "num_tokens": 422221378.0, "step": 11117 }, { "epoch": 1.4143238773692914, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.905444622039795, "learning_rate": 1e-05, "loss": 0.4933, "mean_token_accuracy": 0.8682056665420532, "num_tokens": 422260503.0, "step": 11118 }, { "epoch": 1.414451087647882, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.295440196990967, "learning_rate": 1e-05, "loss": 0.5118, "mean_token_accuracy": 0.8628342151641846, "num_tokens": 422301218.0, "step": 11119 }, { "epoch": 1.4145782979264725, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.746235370635986, "learning_rate": 1e-05, "loss": 0.4609, "mean_token_accuracy": 0.8784189820289612, "num_tokens": 422337366.0, "step": 11120 }, { "epoch": 1.414705508205063, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.42950439453125, "learning_rate": 1e-05, "loss": 0.4586, "mean_token_accuracy": 0.8822913765907288, "num_tokens": 422376087.0, "step": 11121 }, { "epoch": 1.4148327184836536, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.610918998718262, "learning_rate": 1e-05, "loss": 0.4773, "mean_token_accuracy": 0.8692970275878906, "num_tokens": 422413871.0, "step": 11122 }, { "epoch": 1.414959928762244, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.697458744049072, "learning_rate": 1e-05, "loss": 0.5008, "mean_token_accuracy": 0.8707332015037537, "num_tokens": 422450834.0, "step": 11123 }, { "epoch": 1.4150871390408346, "ewc_loss": 0.099609375, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.527266502380371, "learning_rate": 1e-05, "loss": 0.4443, "mean_token_accuracy": 0.8772042989730835, "num_tokens": 422494566.0, "step": 11124 }, { "epoch": 1.4152143493194251, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 7.034107208251953, "learning_rate": 1e-05, "loss": 0.5495, "mean_token_accuracy": 0.8624233603477478, "num_tokens": 422530107.0, "step": 11125 }, { "epoch": 1.4153415595980154, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.653361797332764, "learning_rate": 1e-05, "loss": 0.5428, "mean_token_accuracy": 0.8498555421829224, "num_tokens": 422567598.0, "step": 11126 }, { "epoch": 1.415468769876606, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 0.00011587142944335938, "grad_norm": 7.243227958679199, "learning_rate": 1e-05, "loss": 0.5086, "mean_token_accuracy": 0.8696144819259644, "num_tokens": 422611705.0, "step": 11127 }, { "epoch": 1.4155959801551965, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 5.999040126800537, "learning_rate": 1e-05, "loss": 0.5147, "mean_token_accuracy": 0.8615825176239014, "num_tokens": 422648082.0, "step": 11128 }, { "epoch": 1.415723190433787, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 7.0804572105407715, "learning_rate": 1e-05, "loss": 0.508, "mean_token_accuracy": 0.8707640171051025, "num_tokens": 422691974.0, "step": 11129 }, { "epoch": 1.4158504007123776, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.278228282928467, "learning_rate": 1e-05, "loss": 0.5254, "mean_token_accuracy": 0.8602327108383179, "num_tokens": 422728245.0, "step": 11130 }, { "epoch": 1.415977610990968, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.621855735778809, "learning_rate": 1e-05, "loss": 0.4924, "mean_token_accuracy": 0.8709256052970886, "num_tokens": 422760920.0, "step": 11131 }, { "epoch": 1.4161048212695586, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.3577070236206055, "learning_rate": 1e-05, "loss": 0.5033, "mean_token_accuracy": 0.8667588829994202, "num_tokens": 422796188.0, "step": 11132 }, { "epoch": 1.4162320315481491, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 6.1072869300842285, "learning_rate": 1e-05, "loss": 0.4481, "mean_token_accuracy": 0.8797392249107361, "num_tokens": 422836232.0, "step": 11133 }, { "epoch": 1.4163592418267397, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.476252555847168, "learning_rate": 1e-05, "loss": 0.4805, "mean_token_accuracy": 0.8746262788772583, "num_tokens": 422872855.0, "step": 11134 }, { "epoch": 1.41648645210533, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.696382522583008, "learning_rate": 1e-05, "loss": 0.4836, "mean_token_accuracy": 0.8658604025840759, "num_tokens": 422912826.0, "step": 11135 }, { "epoch": 1.4166136623839205, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.762966632843018, "learning_rate": 1e-05, "loss": 0.5337, "mean_token_accuracy": 0.8608198165893555, "num_tokens": 422950034.0, "step": 11136 }, { "epoch": 1.416740872662511, "ewc_loss": 0.099609375, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.505425453186035, "learning_rate": 1e-05, "loss": 0.4556, "mean_token_accuracy": 0.8736082911491394, "num_tokens": 422987764.0, "step": 11137 }, { "epoch": 1.4168680829411016, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 0.00010967254638671875, "grad_norm": 7.131906509399414, "learning_rate": 1e-05, "loss": 0.4584, "mean_token_accuracy": 0.8857349157333374, "num_tokens": 423024274.0, "step": 11138 }, { "epoch": 1.416995293219692, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.646582126617432, "learning_rate": 1e-05, "loss": 0.5171, "mean_token_accuracy": 0.8582510948181152, "num_tokens": 423058630.0, "step": 11139 }, { "epoch": 1.4171225034982826, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 0.00011873245239257812, "grad_norm": 7.383979797363281, "learning_rate": 1e-05, "loss": 0.4949, "mean_token_accuracy": 0.8788269758224487, "num_tokens": 423099829.0, "step": 11140 }, { "epoch": 1.4172497137768731, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.050714015960693, "learning_rate": 1e-05, "loss": 0.4852, "mean_token_accuracy": 0.8721096515655518, "num_tokens": 423130369.0, "step": 11141 }, { "epoch": 1.4173769240554637, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 0.00011205673217773438, "grad_norm": 7.235214710235596, "learning_rate": 1e-05, "loss": 0.4996, "mean_token_accuracy": 0.8723373413085938, "num_tokens": 423166437.0, "step": 11142 }, { "epoch": 1.4175041343340542, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.304161548614502, "learning_rate": 1e-05, "loss": 0.4624, "mean_token_accuracy": 0.8830840587615967, "num_tokens": 423202220.0, "step": 11143 }, { "epoch": 1.4176313446126447, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.886159420013428, "learning_rate": 1e-05, "loss": 0.516, "mean_token_accuracy": 0.8664652109146118, "num_tokens": 423243417.0, "step": 11144 }, { "epoch": 1.4177585548912353, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.2926554679870605, "learning_rate": 1e-05, "loss": 0.5114, "mean_token_accuracy": 0.8668091297149658, "num_tokens": 423282337.0, "step": 11145 }, { "epoch": 1.4178857651698258, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.56577205657959, "learning_rate": 1e-05, "loss": 0.5201, "mean_token_accuracy": 0.8639987707138062, "num_tokens": 423318666.0, "step": 11146 }, { "epoch": 1.4180129754484163, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.195201396942139, "learning_rate": 1e-05, "loss": 0.5162, "mean_token_accuracy": 0.8636823892593384, "num_tokens": 423357895.0, "step": 11147 }, { "epoch": 1.4181401857270068, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.354724407196045, "learning_rate": 1e-05, "loss": 0.5, "mean_token_accuracy": 0.8698133826255798, "num_tokens": 423395330.0, "step": 11148 }, { "epoch": 1.4182673960055974, "ewc_loss": 0.109375, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 6.051069259643555, "learning_rate": 1e-05, "loss": 0.4932, "mean_token_accuracy": 0.8681056499481201, "num_tokens": 423436824.0, "step": 11149 }, { "epoch": 1.418394606284188, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.2856340408325195, "learning_rate": 1e-05, "loss": 0.5064, "mean_token_accuracy": 0.8671627044677734, "num_tokens": 423471906.0, "step": 11150 }, { "epoch": 1.4185218165627782, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.895529270172119, "learning_rate": 1e-05, "loss": 0.4769, "mean_token_accuracy": 0.8705992698669434, "num_tokens": 423510774.0, "step": 11151 }, { "epoch": 1.4186490268413687, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.324948787689209, "learning_rate": 1e-05, "loss": 0.4934, "mean_token_accuracy": 0.8692753314971924, "num_tokens": 423549825.0, "step": 11152 }, { "epoch": 1.4187762371199593, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.748985767364502, "learning_rate": 1e-05, "loss": 0.4693, "mean_token_accuracy": 0.8730822801589966, "num_tokens": 423577546.0, "step": 11153 }, { "epoch": 1.4189034473985498, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.458588123321533, "learning_rate": 1e-05, "loss": 0.464, "mean_token_accuracy": 0.8793187141418457, "num_tokens": 423613167.0, "step": 11154 }, { "epoch": 1.4190306576771403, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.6022443771362305, "learning_rate": 1e-05, "loss": 0.5084, "mean_token_accuracy": 0.857284426689148, "num_tokens": 423649601.0, "step": 11155 }, { "epoch": 1.4191578679557308, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.710143089294434, "learning_rate": 1e-05, "loss": 0.5261, "mean_token_accuracy": 0.8604229688644409, "num_tokens": 423685429.0, "step": 11156 }, { "epoch": 1.4192850782343214, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.531135082244873, "learning_rate": 1e-05, "loss": 0.4663, "mean_token_accuracy": 0.876997709274292, "num_tokens": 423724270.0, "step": 11157 }, { "epoch": 1.419412288512912, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 7.046726703643799, "learning_rate": 1e-05, "loss": 0.4759, "mean_token_accuracy": 0.8814654350280762, "num_tokens": 423760935.0, "step": 11158 }, { "epoch": 1.4195394987915024, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.674594402313232, "learning_rate": 1e-05, "loss": 0.4907, "mean_token_accuracy": 0.8682574033737183, "num_tokens": 423801949.0, "step": 11159 }, { "epoch": 1.4196667090700927, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 0.00011491775512695312, "grad_norm": 7.23864221572876, "learning_rate": 1e-05, "loss": 0.475, "mean_token_accuracy": 0.8798975944519043, "num_tokens": 423840532.0, "step": 11160 }, { "epoch": 1.4197939193486833, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.033051490783691, "learning_rate": 1e-05, "loss": 0.5234, "mean_token_accuracy": 0.8607831001281738, "num_tokens": 423877912.0, "step": 11161 }, { "epoch": 1.4199211296272738, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 7.035881519317627, "learning_rate": 1e-05, "loss": 0.4876, "mean_token_accuracy": 0.8760647773742676, "num_tokens": 423917836.0, "step": 11162 }, { "epoch": 1.4200483399058643, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.278073787689209, "learning_rate": 1e-05, "loss": 0.5207, "mean_token_accuracy": 0.862987220287323, "num_tokens": 423957307.0, "step": 11163 }, { "epoch": 1.4201755501844548, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.578975200653076, "learning_rate": 1e-05, "loss": 0.4983, "mean_token_accuracy": 0.8693884611129761, "num_tokens": 423994280.0, "step": 11164 }, { "epoch": 1.4203027604630454, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.357639789581299, "learning_rate": 1e-05, "loss": 0.5053, "mean_token_accuracy": 0.8677656054496765, "num_tokens": 424033682.0, "step": 11165 }, { "epoch": 1.420429970741636, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 6.080446243286133, "learning_rate": 1e-05, "loss": 0.4911, "mean_token_accuracy": 0.8657267093658447, "num_tokens": 424072432.0, "step": 11166 }, { "epoch": 1.4205571810202264, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.493946075439453, "learning_rate": 1e-05, "loss": 0.5229, "mean_token_accuracy": 0.8609822988510132, "num_tokens": 424110074.0, "step": 11167 }, { "epoch": 1.420684391298817, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.675881862640381, "learning_rate": 1e-05, "loss": 0.4906, "mean_token_accuracy": 0.8648266792297363, "num_tokens": 424148263.0, "step": 11168 }, { "epoch": 1.4208116015774075, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.769830703735352, "learning_rate": 1e-05, "loss": 0.535, "mean_token_accuracy": 0.8598907589912415, "num_tokens": 424190342.0, "step": 11169 }, { "epoch": 1.420938811855998, "ewc_loss": 0.099609375, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.493741989135742, "learning_rate": 1e-05, "loss": 0.5114, "mean_token_accuracy": 0.8597317934036255, "num_tokens": 424236136.0, "step": 11170 }, { "epoch": 1.4210660221345885, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 7.140196800231934, "learning_rate": 1e-05, "loss": 0.5372, "mean_token_accuracy": 0.8580771684646606, "num_tokens": 424274436.0, "step": 11171 }, { "epoch": 1.421193232413179, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.651882171630859, "learning_rate": 1e-05, "loss": 0.4412, "mean_token_accuracy": 0.8870159387588501, "num_tokens": 424309661.0, "step": 11172 }, { "epoch": 1.4213204426917696, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 0.00011873245239257812, "grad_norm": 7.3916802406311035, "learning_rate": 1e-05, "loss": 0.5474, "mean_token_accuracy": 0.8593679666519165, "num_tokens": 424347127.0, "step": 11173 }, { "epoch": 1.4214476529703601, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.060257434844971, "learning_rate": 1e-05, "loss": 0.4722, "mean_token_accuracy": 0.876641035079956, "num_tokens": 424391287.0, "step": 11174 }, { "epoch": 1.4215748632489504, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 0.00011157989501953125, "grad_norm": 7.24028205871582, "learning_rate": 1e-05, "loss": 0.4631, "mean_token_accuracy": 0.878892719745636, "num_tokens": 424428091.0, "step": 11175 }, { "epoch": 1.421702073527541, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.316493511199951, "learning_rate": 1e-05, "loss": 0.4982, "mean_token_accuracy": 0.8662040829658508, "num_tokens": 424469270.0, "step": 11176 }, { "epoch": 1.4218292838061315, "ewc_loss": 0.125, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.877964019775391, "learning_rate": 1e-05, "loss": 0.5463, "mean_token_accuracy": 0.8556205034255981, "num_tokens": 424509116.0, "step": 11177 }, { "epoch": 1.421956494084722, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.313127040863037, "learning_rate": 1e-05, "loss": 0.4934, "mean_token_accuracy": 0.8726597428321838, "num_tokens": 424542777.0, "step": 11178 }, { "epoch": 1.4220837043633126, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.528616905212402, "learning_rate": 1e-05, "loss": 0.5025, "mean_token_accuracy": 0.8718180656433105, "num_tokens": 424582299.0, "step": 11179 }, { "epoch": 1.422210914641903, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.206674575805664, "learning_rate": 1e-05, "loss": 0.4485, "mean_token_accuracy": 0.8843593597412109, "num_tokens": 424617543.0, "step": 11180 }, { "epoch": 1.4223381249204936, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.307702541351318, "learning_rate": 1e-05, "loss": 0.456, "mean_token_accuracy": 0.8805791139602661, "num_tokens": 424657512.0, "step": 11181 }, { "epoch": 1.4224653351990841, "ewc_loss": 0.109375, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.072463512420654, "learning_rate": 1e-05, "loss": 0.5214, "mean_token_accuracy": 0.8594421148300171, "num_tokens": 424695954.0, "step": 11182 }, { "epoch": 1.4225925454776747, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.230292797088623, "learning_rate": 1e-05, "loss": 0.5305, "mean_token_accuracy": 0.8576140403747559, "num_tokens": 424736174.0, "step": 11183 }, { "epoch": 1.422719755756265, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.909976959228516, "learning_rate": 1e-05, "loss": 0.4946, "mean_token_accuracy": 0.8689505457878113, "num_tokens": 424776371.0, "step": 11184 }, { "epoch": 1.4228469660348555, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.267100811004639, "learning_rate": 1e-05, "loss": 0.5283, "mean_token_accuracy": 0.8603223562240601, "num_tokens": 424816874.0, "step": 11185 }, { "epoch": 1.422974176313446, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.755178928375244, "learning_rate": 1e-05, "loss": 0.4271, "mean_token_accuracy": 0.8868423700332642, "num_tokens": 424847924.0, "step": 11186 }, { "epoch": 1.4231013865920366, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.401607990264893, "learning_rate": 1e-05, "loss": 0.4473, "mean_token_accuracy": 0.8841418027877808, "num_tokens": 424887950.0, "step": 11187 }, { "epoch": 1.423228596870627, "ewc_loss": 0.099609375, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.6053853034973145, "learning_rate": 1e-05, "loss": 0.5001, "mean_token_accuracy": 0.8606778979301453, "num_tokens": 424922608.0, "step": 11188 }, { "epoch": 1.4233558071492176, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.6655144691467285, "learning_rate": 1e-05, "loss": 0.4738, "mean_token_accuracy": 0.8780351877212524, "num_tokens": 424958564.0, "step": 11189 }, { "epoch": 1.4234830174278081, "ewc_loss": 0.0986328125, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.496882915496826, "learning_rate": 1e-05, "loss": 0.4811, "mean_token_accuracy": 0.8655818700790405, "num_tokens": 424993351.0, "step": 11190 }, { "epoch": 1.4236102277063987, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 7.031726360321045, "learning_rate": 1e-05, "loss": 0.5149, "mean_token_accuracy": 0.8691022396087646, "num_tokens": 425028018.0, "step": 11191 }, { "epoch": 1.4237374379849892, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.594980239868164, "learning_rate": 1e-05, "loss": 0.4828, "mean_token_accuracy": 0.870290994644165, "num_tokens": 425065023.0, "step": 11192 }, { "epoch": 1.4238646482635797, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 0.00011777877807617188, "grad_norm": 7.312222957611084, "learning_rate": 1e-05, "loss": 0.4969, "mean_token_accuracy": 0.8746036887168884, "num_tokens": 425101152.0, "step": 11193 }, { "epoch": 1.4239918585421703, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 5.973434925079346, "learning_rate": 1e-05, "loss": 0.5158, "mean_token_accuracy": 0.8619835376739502, "num_tokens": 425137756.0, "step": 11194 }, { "epoch": 1.4241190688207608, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 0.0001125335693359375, "grad_norm": 7.212484836578369, "learning_rate": 1e-05, "loss": 0.4937, "mean_token_accuracy": 0.8732104301452637, "num_tokens": 425171947.0, "step": 11195 }, { "epoch": 1.4242462790993513, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.302062034606934, "learning_rate": 1e-05, "loss": 0.4916, "mean_token_accuracy": 0.8685272932052612, "num_tokens": 425207219.0, "step": 11196 }, { "epoch": 1.4243734893779418, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.7665228843688965, "learning_rate": 1e-05, "loss": 0.5235, "mean_token_accuracy": 0.8609863519668579, "num_tokens": 425242515.0, "step": 11197 }, { "epoch": 1.4245006996565324, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.383234024047852, "learning_rate": 1e-05, "loss": 0.5401, "mean_token_accuracy": 0.856232762336731, "num_tokens": 425283409.0, "step": 11198 }, { "epoch": 1.424627909935123, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.256198883056641, "learning_rate": 1e-05, "loss": 0.5106, "mean_token_accuracy": 0.862687349319458, "num_tokens": 425321862.0, "step": 11199 }, { "epoch": 1.4247551202137132, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.440229892730713, "learning_rate": 1e-05, "loss": 0.469, "mean_token_accuracy": 0.8794434070587158, "num_tokens": 425362034.0, "step": 11200 }, { "epoch": 1.4248823304923037, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.849513053894043, "learning_rate": 1e-05, "loss": 0.5374, "mean_token_accuracy": 0.8535916805267334, "num_tokens": 425397900.0, "step": 11201 }, { "epoch": 1.4250095407708943, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.598400115966797, "learning_rate": 1e-05, "loss": 0.4911, "mean_token_accuracy": 0.8705183863639832, "num_tokens": 425435458.0, "step": 11202 }, { "epoch": 1.4251367510494848, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.603754997253418, "learning_rate": 1e-05, "loss": 0.4475, "mean_token_accuracy": 0.8789374232292175, "num_tokens": 425465687.0, "step": 11203 }, { "epoch": 1.4252639613280753, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.8733439445495605, "learning_rate": 1e-05, "loss": 0.4947, "mean_token_accuracy": 0.8672326803207397, "num_tokens": 425500702.0, "step": 11204 }, { "epoch": 1.4253911716066658, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.566410064697266, "learning_rate": 1e-05, "loss": 0.4781, "mean_token_accuracy": 0.8706411123275757, "num_tokens": 425542827.0, "step": 11205 }, { "epoch": 1.4255183818852564, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 0.00011157989501953125, "grad_norm": 7.166319370269775, "learning_rate": 1e-05, "loss": 0.536, "mean_token_accuracy": 0.8593012690544128, "num_tokens": 425585146.0, "step": 11206 }, { "epoch": 1.425645592163847, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 5.781038284301758, "learning_rate": 1e-05, "loss": 0.4675, "mean_token_accuracy": 0.8764302730560303, "num_tokens": 425623790.0, "step": 11207 }, { "epoch": 1.4257728024424374, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 0.00011491775512695312, "grad_norm": 7.2638115882873535, "learning_rate": 1e-05, "loss": 0.5093, "mean_token_accuracy": 0.870004415512085, "num_tokens": 425663290.0, "step": 11208 }, { "epoch": 1.4259000127210277, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.091196537017822, "learning_rate": 1e-05, "loss": 0.4841, "mean_token_accuracy": 0.8728011846542358, "num_tokens": 425701226.0, "step": 11209 }, { "epoch": 1.4260272229996183, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 7.050522804260254, "learning_rate": 1e-05, "loss": 0.5114, "mean_token_accuracy": 0.8704668879508972, "num_tokens": 425737524.0, "step": 11210 }, { "epoch": 1.4261544332782088, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.2413763999938965, "learning_rate": 1e-05, "loss": 0.4833, "mean_token_accuracy": 0.8763719201087952, "num_tokens": 425771010.0, "step": 11211 }, { "epoch": 1.4262816435567993, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.72036600112915, "learning_rate": 1e-05, "loss": 0.5334, "mean_token_accuracy": 0.8615221977233887, "num_tokens": 425813077.0, "step": 11212 }, { "epoch": 1.4264088538353898, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.204633712768555, "learning_rate": 1e-05, "loss": 0.4583, "mean_token_accuracy": 0.8836022019386292, "num_tokens": 425852718.0, "step": 11213 }, { "epoch": 1.4265360641139804, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.428906440734863, "learning_rate": 1e-05, "loss": 0.501, "mean_token_accuracy": 0.8695383071899414, "num_tokens": 425888196.0, "step": 11214 }, { "epoch": 1.426663274392571, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.111124038696289, "learning_rate": 1e-05, "loss": 0.4747, "mean_token_accuracy": 0.8766244053840637, "num_tokens": 425926774.0, "step": 11215 }, { "epoch": 1.4267904846711614, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.268847465515137, "learning_rate": 1e-05, "loss": 0.4433, "mean_token_accuracy": 0.8847638964653015, "num_tokens": 425966207.0, "step": 11216 }, { "epoch": 1.426917694949752, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 5.9848175048828125, "learning_rate": 1e-05, "loss": 0.5145, "mean_token_accuracy": 0.8592909574508667, "num_tokens": 426008792.0, "step": 11217 }, { "epoch": 1.4270449052283425, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.241130352020264, "learning_rate": 1e-05, "loss": 0.5032, "mean_token_accuracy": 0.8693063259124756, "num_tokens": 426049650.0, "step": 11218 }, { "epoch": 1.427172115506933, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.840137958526611, "learning_rate": 1e-05, "loss": 0.457, "mean_token_accuracy": 0.878555178642273, "num_tokens": 426092143.0, "step": 11219 }, { "epoch": 1.4272993257855235, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.3072686195373535, "learning_rate": 1e-05, "loss": 0.5412, "mean_token_accuracy": 0.8566367030143738, "num_tokens": 426135393.0, "step": 11220 }, { "epoch": 1.427426536064114, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.702920436859131, "learning_rate": 1e-05, "loss": 0.4715, "mean_token_accuracy": 0.8705431222915649, "num_tokens": 426171858.0, "step": 11221 }, { "epoch": 1.4275537463427046, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.475897312164307, "learning_rate": 1e-05, "loss": 0.4607, "mean_token_accuracy": 0.8809382319450378, "num_tokens": 426206635.0, "step": 11222 }, { "epoch": 1.4276809566212951, "ewc_loss": 0.099609375, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.557620525360107, "learning_rate": 1e-05, "loss": 0.4685, "mean_token_accuracy": 0.8731191158294678, "num_tokens": 426240942.0, "step": 11223 }, { "epoch": 1.4278081668998854, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.757887840270996, "learning_rate": 1e-05, "loss": 0.4346, "mean_token_accuracy": 0.8896613717079163, "num_tokens": 426279592.0, "step": 11224 }, { "epoch": 1.427935377178476, "ewc_loss": 0.099609375, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.510560989379883, "learning_rate": 1e-05, "loss": 0.4387, "mean_token_accuracy": 0.8818094730377197, "num_tokens": 426314218.0, "step": 11225 }, { "epoch": 1.4280625874570665, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 7.106769561767578, "learning_rate": 1e-05, "loss": 0.5056, "mean_token_accuracy": 0.8699308633804321, "num_tokens": 426351514.0, "step": 11226 }, { "epoch": 1.428189797735657, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.693037509918213, "learning_rate": 1e-05, "loss": 0.5403, "mean_token_accuracy": 0.8586670756340027, "num_tokens": 426389553.0, "step": 11227 }, { "epoch": 1.4283170080142475, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 0.00011491775512695312, "grad_norm": 7.2620110511779785, "learning_rate": 1e-05, "loss": 0.4927, "mean_token_accuracy": 0.8788589835166931, "num_tokens": 426424026.0, "step": 11228 }, { "epoch": 1.428444218292838, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.059647560119629, "learning_rate": 1e-05, "loss": 0.4862, "mean_token_accuracy": 0.8711456060409546, "num_tokens": 426466711.0, "step": 11229 }, { "epoch": 1.4285714285714286, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 7.055260181427002, "learning_rate": 1e-05, "loss": 0.5268, "mean_token_accuracy": 0.8655014038085938, "num_tokens": 426507636.0, "step": 11230 }, { "epoch": 1.4286986388500191, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.296204566955566, "learning_rate": 1e-05, "loss": 0.5049, "mean_token_accuracy": 0.8658809065818787, "num_tokens": 426551811.0, "step": 11231 }, { "epoch": 1.4288258491286097, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.5614213943481445, "learning_rate": 1e-05, "loss": 0.5016, "mean_token_accuracy": 0.8675773739814758, "num_tokens": 426585814.0, "step": 11232 }, { "epoch": 1.4289530594072, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.366608619689941, "learning_rate": 1e-05, "loss": 0.4729, "mean_token_accuracy": 0.8768788576126099, "num_tokens": 426619327.0, "step": 11233 }, { "epoch": 1.4290802696857905, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 6.074161529541016, "learning_rate": 1e-05, "loss": 0.5032, "mean_token_accuracy": 0.861440896987915, "num_tokens": 426656976.0, "step": 11234 }, { "epoch": 1.429207479964381, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.490838050842285, "learning_rate": 1e-05, "loss": 0.4792, "mean_token_accuracy": 0.8776635527610779, "num_tokens": 426693930.0, "step": 11235 }, { "epoch": 1.4293346902429716, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.694820880889893, "learning_rate": 1e-05, "loss": 0.4272, "mean_token_accuracy": 0.8870404958724976, "num_tokens": 426731655.0, "step": 11236 }, { "epoch": 1.429461900521562, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.7474894523620605, "learning_rate": 1e-05, "loss": 0.5232, "mean_token_accuracy": 0.8650079965591431, "num_tokens": 426768384.0, "step": 11237 }, { "epoch": 1.4295891108001526, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.531167507171631, "learning_rate": 1e-05, "loss": 0.4885, "mean_token_accuracy": 0.8673446178436279, "num_tokens": 426806319.0, "step": 11238 }, { "epoch": 1.4297163210787431, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 0.00011014938354492188, "grad_norm": 7.08674430847168, "learning_rate": 1e-05, "loss": 0.5829, "mean_token_accuracy": 0.8462872505187988, "num_tokens": 426841076.0, "step": 11239 }, { "epoch": 1.4298435313573337, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.649075984954834, "learning_rate": 1e-05, "loss": 0.4541, "mean_token_accuracy": 0.8769773244857788, "num_tokens": 426877647.0, "step": 11240 }, { "epoch": 1.4299707416359242, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 0.00011587142944335938, "grad_norm": 7.300174236297607, "learning_rate": 1e-05, "loss": 0.5111, "mean_token_accuracy": 0.8714433312416077, "num_tokens": 426917975.0, "step": 11241 }, { "epoch": 1.4300979519145147, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 5.985292434692383, "learning_rate": 1e-05, "loss": 0.4925, "mean_token_accuracy": 0.8703858256340027, "num_tokens": 426958975.0, "step": 11242 }, { "epoch": 1.4302251621931052, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 0.00011301040649414062, "grad_norm": 7.211625576019287, "learning_rate": 1e-05, "loss": 0.5139, "mean_token_accuracy": 0.8692747354507446, "num_tokens": 426997205.0, "step": 11243 }, { "epoch": 1.4303523724716958, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.234231948852539, "learning_rate": 1e-05, "loss": 0.5046, "mean_token_accuracy": 0.8703311681747437, "num_tokens": 427037455.0, "step": 11244 }, { "epoch": 1.4304795827502863, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.904233455657959, "learning_rate": 1e-05, "loss": 0.5271, "mean_token_accuracy": 0.8663532733917236, "num_tokens": 427071443.0, "step": 11245 }, { "epoch": 1.4306067930288768, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.253028392791748, "learning_rate": 1e-05, "loss": 0.491, "mean_token_accuracy": 0.8724217414855957, "num_tokens": 427113258.0, "step": 11246 }, { "epoch": 1.4307340033074674, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.589282989501953, "learning_rate": 1e-05, "loss": 0.4753, "mean_token_accuracy": 0.8724000453948975, "num_tokens": 427153523.0, "step": 11247 }, { "epoch": 1.430861213586058, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.185005187988281, "learning_rate": 1e-05, "loss": 0.4885, "mean_token_accuracy": 0.8696010708808899, "num_tokens": 427186818.0, "step": 11248 }, { "epoch": 1.4309884238646482, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.35931396484375, "learning_rate": 1e-05, "loss": 0.5432, "mean_token_accuracy": 0.8545854091644287, "num_tokens": 427226817.0, "step": 11249 }, { "epoch": 1.4311156341432387, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.065762996673584, "learning_rate": 1e-05, "loss": 0.4966, "mean_token_accuracy": 0.8671061396598816, "num_tokens": 427268471.0, "step": 11250 }, { "epoch": 1.4312428444218293, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.26284122467041, "learning_rate": 1e-05, "loss": 0.4485, "mean_token_accuracy": 0.8818222284317017, "num_tokens": 427301242.0, "step": 11251 }, { "epoch": 1.4313700547004198, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.928359508514404, "learning_rate": 1e-05, "loss": 0.5098, "mean_token_accuracy": 0.8613068461418152, "num_tokens": 427335230.0, "step": 11252 }, { "epoch": 1.4314972649790103, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.278456211090088, "learning_rate": 1e-05, "loss": 0.5182, "mean_token_accuracy": 0.8614553809165955, "num_tokens": 427371993.0, "step": 11253 }, { "epoch": 1.4316244752576008, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.7742533683776855, "learning_rate": 1e-05, "loss": 0.5096, "mean_token_accuracy": 0.8613292574882507, "num_tokens": 427408810.0, "step": 11254 }, { "epoch": 1.4317516855361914, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.394410610198975, "learning_rate": 1e-05, "loss": 0.5349, "mean_token_accuracy": 0.8551515340805054, "num_tokens": 427448740.0, "step": 11255 }, { "epoch": 1.431878895814782, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.62060022354126, "learning_rate": 1e-05, "loss": 0.4794, "mean_token_accuracy": 0.8712472915649414, "num_tokens": 427488234.0, "step": 11256 }, { "epoch": 1.4320061060933724, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.629814624786377, "learning_rate": 1e-05, "loss": 0.5075, "mean_token_accuracy": 0.8676939010620117, "num_tokens": 427528053.0, "step": 11257 }, { "epoch": 1.4321333163719627, "ewc_loss": 0.0986328125, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.523951530456543, "learning_rate": 1e-05, "loss": 0.4589, "mean_token_accuracy": 0.8761460185050964, "num_tokens": 427566882.0, "step": 11258 }, { "epoch": 1.4322605266505533, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.956379413604736, "learning_rate": 1e-05, "loss": 0.491, "mean_token_accuracy": 0.8729405999183655, "num_tokens": 427600464.0, "step": 11259 }, { "epoch": 1.4323877369291438, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.586998462677002, "learning_rate": 1e-05, "loss": 0.461, "mean_token_accuracy": 0.8750638365745544, "num_tokens": 427633394.0, "step": 11260 }, { "epoch": 1.4325149472077343, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 0.00011682510375976562, "grad_norm": 7.229940891265869, "learning_rate": 1e-05, "loss": 0.5208, "mean_token_accuracy": 0.8692207336425781, "num_tokens": 427670591.0, "step": 11261 }, { "epoch": 1.4326421574863248, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 5.895824432373047, "learning_rate": 1e-05, "loss": 0.4495, "mean_token_accuracy": 0.881941556930542, "num_tokens": 427711752.0, "step": 11262 }, { "epoch": 1.4327693677649154, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 0.00011444091796875, "grad_norm": 7.1913323402404785, "learning_rate": 1e-05, "loss": 0.5152, "mean_token_accuracy": 0.8688856959342957, "num_tokens": 427744655.0, "step": 11263 }, { "epoch": 1.432896578043506, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.238645553588867, "learning_rate": 1e-05, "loss": 0.4834, "mean_token_accuracy": 0.8750265836715698, "num_tokens": 427784388.0, "step": 11264 }, { "epoch": 1.4330237883220964, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.7678093910217285, "learning_rate": 1e-05, "loss": 0.5262, "mean_token_accuracy": 0.8607141971588135, "num_tokens": 427823498.0, "step": 11265 }, { "epoch": 1.433150998600687, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.36826753616333, "learning_rate": 1e-05, "loss": 0.4911, "mean_token_accuracy": 0.8702825307846069, "num_tokens": 427868185.0, "step": 11266 }, { "epoch": 1.4332782088792775, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.248741149902344, "learning_rate": 1e-05, "loss": 0.5188, "mean_token_accuracy": 0.8617684841156006, "num_tokens": 427903315.0, "step": 11267 }, { "epoch": 1.433405419157868, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.456930637359619, "learning_rate": 1e-05, "loss": 0.5369, "mean_token_accuracy": 0.8565371632575989, "num_tokens": 427950206.0, "step": 11268 }, { "epoch": 1.4335326294364585, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.814449310302734, "learning_rate": 1e-05, "loss": 0.5847, "mean_token_accuracy": 0.8473160266876221, "num_tokens": 427983038.0, "step": 11269 }, { "epoch": 1.433659839715049, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.6775803565979, "learning_rate": 1e-05, "loss": 0.4812, "mean_token_accuracy": 0.8761303424835205, "num_tokens": 428015148.0, "step": 11270 }, { "epoch": 1.4337870499936396, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.532587051391602, "learning_rate": 1e-05, "loss": 0.4818, "mean_token_accuracy": 0.8704508543014526, "num_tokens": 428054300.0, "step": 11271 }, { "epoch": 1.4339142602722301, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 7.039329528808594, "learning_rate": 1e-05, "loss": 0.5582, "mean_token_accuracy": 0.8575447201728821, "num_tokens": 428093802.0, "step": 11272 }, { "epoch": 1.4340414705508204, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.572846412658691, "learning_rate": 1e-05, "loss": 0.4616, "mean_token_accuracy": 0.8792039155960083, "num_tokens": 428133283.0, "step": 11273 }, { "epoch": 1.434168680829411, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 0.000118255615234375, "grad_norm": 7.375364780426025, "learning_rate": 1e-05, "loss": 0.5553, "mean_token_accuracy": 0.8607133626937866, "num_tokens": 428171125.0, "step": 11274 }, { "epoch": 1.4342958911080015, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 5.9572529792785645, "learning_rate": 1e-05, "loss": 0.543, "mean_token_accuracy": 0.8508752584457397, "num_tokens": 428211731.0, "step": 11275 }, { "epoch": 1.434423101386592, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 0.00011730194091796875, "grad_norm": 7.333341598510742, "learning_rate": 1e-05, "loss": 0.5452, "mean_token_accuracy": 0.8652248978614807, "num_tokens": 428243737.0, "step": 11276 }, { "epoch": 1.4345503116651825, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.297846794128418, "learning_rate": 1e-05, "loss": 0.4774, "mean_token_accuracy": 0.8783432245254517, "num_tokens": 428278761.0, "step": 11277 }, { "epoch": 1.434677521943773, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.979013442993164, "learning_rate": 1e-05, "loss": 0.4606, "mean_token_accuracy": 0.886699914932251, "num_tokens": 428316158.0, "step": 11278 }, { "epoch": 1.4348047322223636, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.332527160644531, "learning_rate": 1e-05, "loss": 0.5064, "mean_token_accuracy": 0.8660205006599426, "num_tokens": 428347180.0, "step": 11279 }, { "epoch": 1.4349319425009541, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.637699604034424, "learning_rate": 1e-05, "loss": 0.4763, "mean_token_accuracy": 0.8775687217712402, "num_tokens": 428389962.0, "step": 11280 }, { "epoch": 1.4350591527795447, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.217032432556152, "learning_rate": 1e-05, "loss": 0.5583, "mean_token_accuracy": 0.8468650579452515, "num_tokens": 428432950.0, "step": 11281 }, { "epoch": 1.435186363058135, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.400897979736328, "learning_rate": 1e-05, "loss": 0.5433, "mean_token_accuracy": 0.8534432649612427, "num_tokens": 428470745.0, "step": 11282 }, { "epoch": 1.4353135733367255, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.08401346206665, "learning_rate": 1e-05, "loss": 0.5198, "mean_token_accuracy": 0.8647557497024536, "num_tokens": 428508836.0, "step": 11283 }, { "epoch": 1.435440783615316, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.3068976402282715, "learning_rate": 1e-05, "loss": 0.4773, "mean_token_accuracy": 0.8709960579872131, "num_tokens": 428546371.0, "step": 11284 }, { "epoch": 1.4355679938939065, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.915387153625488, "learning_rate": 1e-05, "loss": 0.4381, "mean_token_accuracy": 0.8818877935409546, "num_tokens": 428578878.0, "step": 11285 }, { "epoch": 1.435695204172497, "ewc_loss": 0.11376953125, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.344115257263184, "learning_rate": 1e-05, "loss": 0.5118, "mean_token_accuracy": 0.8635010719299316, "num_tokens": 428611218.0, "step": 11286 }, { "epoch": 1.4358224144510876, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.794595718383789, "learning_rate": 1e-05, "loss": 0.5717, "mean_token_accuracy": 0.8429738879203796, "num_tokens": 428649381.0, "step": 11287 }, { "epoch": 1.4359496247296781, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.438270568847656, "learning_rate": 1e-05, "loss": 0.5011, "mean_token_accuracy": 0.8682336807250977, "num_tokens": 428685221.0, "step": 11288 }, { "epoch": 1.4360768350082687, "ewc_loss": 0.10107421875, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.620490550994873, "learning_rate": 1e-05, "loss": 0.479, "mean_token_accuracy": 0.8705464601516724, "num_tokens": 428724194.0, "step": 11289 }, { "epoch": 1.4362040452868592, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.676299571990967, "learning_rate": 1e-05, "loss": 0.5105, "mean_token_accuracy": 0.8664815425872803, "num_tokens": 428760501.0, "step": 11290 }, { "epoch": 1.4363312555654497, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.5427069664001465, "learning_rate": 1e-05, "loss": 0.5132, "mean_token_accuracy": 0.8580977320671082, "num_tokens": 428793917.0, "step": 11291 }, { "epoch": 1.4364584658440402, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 6.997958183288574, "learning_rate": 1e-05, "loss": 0.4974, "mean_token_accuracy": 0.8761246800422668, "num_tokens": 428828998.0, "step": 11292 }, { "epoch": 1.4365856761226308, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.637548923492432, "learning_rate": 1e-05, "loss": 0.4526, "mean_token_accuracy": 0.8847046494483948, "num_tokens": 428864739.0, "step": 11293 }, { "epoch": 1.4367128864012213, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 0.00011396408081054688, "grad_norm": 7.226367950439453, "learning_rate": 1e-05, "loss": 0.5422, "mean_token_accuracy": 0.856511116027832, "num_tokens": 428901026.0, "step": 11294 }, { "epoch": 1.4368400966798118, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 5.974491596221924, "learning_rate": 1e-05, "loss": 0.4597, "mean_token_accuracy": 0.8772295713424683, "num_tokens": 428937698.0, "step": 11295 }, { "epoch": 1.4369673069584024, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 0.00011157989501953125, "grad_norm": 7.091899394989014, "learning_rate": 1e-05, "loss": 0.5747, "mean_token_accuracy": 0.8526809215545654, "num_tokens": 428978288.0, "step": 11296 }, { "epoch": 1.4370945172369929, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.252181529998779, "learning_rate": 1e-05, "loss": 0.4413, "mean_token_accuracy": 0.8856656551361084, "num_tokens": 429014299.0, "step": 11297 }, { "epoch": 1.4372217275155832, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.6481614112854, "learning_rate": 1e-05, "loss": 0.5246, "mean_token_accuracy": 0.8611748814582825, "num_tokens": 429052455.0, "step": 11298 }, { "epoch": 1.4373489377941737, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.35455846786499, "learning_rate": 1e-05, "loss": 0.5043, "mean_token_accuracy": 0.8679872155189514, "num_tokens": 429094168.0, "step": 11299 }, { "epoch": 1.4374761480727642, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.131850719451904, "learning_rate": 1e-05, "loss": 0.4998, "mean_token_accuracy": 0.8670544028282166, "num_tokens": 429137147.0, "step": 11300 }, { "epoch": 1.4376033583513548, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.470099925994873, "learning_rate": 1e-05, "loss": 0.4853, "mean_token_accuracy": 0.8760004639625549, "num_tokens": 429181679.0, "step": 11301 }, { "epoch": 1.4377305686299453, "ewc_loss": 0.10205078125, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.70522403717041, "learning_rate": 1e-05, "loss": 0.4766, "mean_token_accuracy": 0.8697420358657837, "num_tokens": 429222378.0, "step": 11302 }, { "epoch": 1.4378577789085358, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.749679088592529, "learning_rate": 1e-05, "loss": 0.5108, "mean_token_accuracy": 0.8669646382331848, "num_tokens": 429262044.0, "step": 11303 }, { "epoch": 1.4379849891871264, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.497519016265869, "learning_rate": 1e-05, "loss": 0.4894, "mean_token_accuracy": 0.8678889274597168, "num_tokens": 429301643.0, "step": 11304 }, { "epoch": 1.438112199465717, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 0.00011157989501953125, "grad_norm": 7.152523517608643, "learning_rate": 1e-05, "loss": 0.5915, "mean_token_accuracy": 0.845399022102356, "num_tokens": 429339190.0, "step": 11305 }, { "epoch": 1.4382394097443074, "ewc_loss": 0.10400390625, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.621311187744141, "learning_rate": 1e-05, "loss": 0.4995, "mean_token_accuracy": 0.866108238697052, "num_tokens": 429378569.0, "step": 11306 }, { "epoch": 1.4383666200228977, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 0.00012159347534179688, "grad_norm": 7.446873664855957, "learning_rate": 1e-05, "loss": 0.5191, "mean_token_accuracy": 0.8688674569129944, "num_tokens": 429416012.0, "step": 11307 }, { "epoch": 1.4384938303014883, "ewc_loss": 0.11474609375, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.069949626922607, "learning_rate": 1e-05, "loss": 0.4274, "mean_token_accuracy": 0.8912754654884338, "num_tokens": 429453242.0, "step": 11308 }, { "epoch": 1.4386210405800788, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 0.00011587142944335938, "grad_norm": 7.3060784339904785, "learning_rate": 1e-05, "loss": 0.5109, "mean_token_accuracy": 0.8742585182189941, "num_tokens": 429493908.0, "step": 11309 }, { "epoch": 1.4387482508586693, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.370162010192871, "learning_rate": 1e-05, "loss": 0.4788, "mean_token_accuracy": 0.8733906745910645, "num_tokens": 429530335.0, "step": 11310 }, { "epoch": 1.4388754611372598, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.926121711730957, "learning_rate": 1e-05, "loss": 0.5165, "mean_token_accuracy": 0.8648302555084229, "num_tokens": 429567523.0, "step": 11311 }, { "epoch": 1.4390026714158504, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.352860450744629, "learning_rate": 1e-05, "loss": 0.4894, "mean_token_accuracy": 0.8775885105133057, "num_tokens": 429605887.0, "step": 11312 }, { "epoch": 1.439129881694441, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.576564788818359, "learning_rate": 1e-05, "loss": 0.4624, "mean_token_accuracy": 0.8807011842727661, "num_tokens": 429644834.0, "step": 11313 }, { "epoch": 1.4392570919730314, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.240335464477539, "learning_rate": 1e-05, "loss": 0.4699, "mean_token_accuracy": 0.8790910243988037, "num_tokens": 429681151.0, "step": 11314 }, { "epoch": 1.439384302251622, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.360714435577393, "learning_rate": 1e-05, "loss": 0.5258, "mean_token_accuracy": 0.8586858510971069, "num_tokens": 429719135.0, "step": 11315 }, { "epoch": 1.4395115125302125, "ewc_loss": 0.11083984375, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.084428787231445, "learning_rate": 1e-05, "loss": 0.4517, "mean_token_accuracy": 0.8786269426345825, "num_tokens": 429753443.0, "step": 11316 }, { "epoch": 1.439638722808803, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.2993669509887695, "learning_rate": 1e-05, "loss": 0.49, "mean_token_accuracy": 0.8734655380249023, "num_tokens": 429788762.0, "step": 11317 }, { "epoch": 1.4397659330873935, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.900108814239502, "learning_rate": 1e-05, "loss": 0.5016, "mean_token_accuracy": 0.8575472831726074, "num_tokens": 429824054.0, "step": 11318 }, { "epoch": 1.439893143365984, "ewc_loss": 0.11376953125, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.360516548156738, "learning_rate": 1e-05, "loss": 0.5111, "mean_token_accuracy": 0.8667550086975098, "num_tokens": 429857772.0, "step": 11319 }, { "epoch": 1.4400203536445746, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.733100414276123, "learning_rate": 1e-05, "loss": 0.4651, "mean_token_accuracy": 0.8742008805274963, "num_tokens": 429891184.0, "step": 11320 }, { "epoch": 1.4401475639231651, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.5201616287231445, "learning_rate": 1e-05, "loss": 0.4737, "mean_token_accuracy": 0.8774562478065491, "num_tokens": 429930322.0, "step": 11321 }, { "epoch": 1.4402747742017554, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.579615592956543, "learning_rate": 1e-05, "loss": 0.4494, "mean_token_accuracy": 0.8729252815246582, "num_tokens": 429965708.0, "step": 11322 }, { "epoch": 1.440401984480346, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.80777645111084, "learning_rate": 1e-05, "loss": 0.5064, "mean_token_accuracy": 0.8688496947288513, "num_tokens": 430007266.0, "step": 11323 }, { "epoch": 1.4405291947589365, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.551556587219238, "learning_rate": 1e-05, "loss": 0.5035, "mean_token_accuracy": 0.8605777621269226, "num_tokens": 430044470.0, "step": 11324 }, { "epoch": 1.440656405037527, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 0.00011110305786132812, "grad_norm": 7.150923252105713, "learning_rate": 1e-05, "loss": 0.5085, "mean_token_accuracy": 0.8715682029724121, "num_tokens": 430079776.0, "step": 11325 }, { "epoch": 1.4407836153161175, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.769241809844971, "learning_rate": 1e-05, "loss": 0.4554, "mean_token_accuracy": 0.8775808215141296, "num_tokens": 430110697.0, "step": 11326 }, { "epoch": 1.440910825594708, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 0.00011444091796875, "grad_norm": 7.260176181793213, "learning_rate": 1e-05, "loss": 0.5138, "mean_token_accuracy": 0.8700271248817444, "num_tokens": 430150350.0, "step": 11327 }, { "epoch": 1.4410380358732986, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.1366190910339355, "learning_rate": 1e-05, "loss": 0.5355, "mean_token_accuracy": 0.8528240323066711, "num_tokens": 430185415.0, "step": 11328 }, { "epoch": 1.4411652461518891, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.9666900634765625, "learning_rate": 1e-05, "loss": 0.5256, "mean_token_accuracy": 0.864027738571167, "num_tokens": 430227865.0, "step": 11329 }, { "epoch": 1.4412924564304797, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.326544284820557, "learning_rate": 1e-05, "loss": 0.4741, "mean_token_accuracy": 0.878031849861145, "num_tokens": 430267821.0, "step": 11330 }, { "epoch": 1.44141966670907, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.4769439697265625, "learning_rate": 1e-05, "loss": 0.4794, "mean_token_accuracy": 0.8728466033935547, "num_tokens": 430307251.0, "step": 11331 }, { "epoch": 1.4415468769876605, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.407805919647217, "learning_rate": 1e-05, "loss": 0.5376, "mean_token_accuracy": 0.8555180430412292, "num_tokens": 430347842.0, "step": 11332 }, { "epoch": 1.441674087266251, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.99486780166626, "learning_rate": 1e-05, "loss": 0.5206, "mean_token_accuracy": 0.8628714084625244, "num_tokens": 430383175.0, "step": 11333 }, { "epoch": 1.4418012975448415, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.5603108406066895, "learning_rate": 1e-05, "loss": 0.5104, "mean_token_accuracy": 0.8651998043060303, "num_tokens": 430414319.0, "step": 11334 }, { "epoch": 1.441928507823432, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.625156402587891, "learning_rate": 1e-05, "loss": 0.4477, "mean_token_accuracy": 0.8818814754486084, "num_tokens": 430454142.0, "step": 11335 }, { "epoch": 1.4420557181020226, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.862667083740234, "learning_rate": 1e-05, "loss": 0.5302, "mean_token_accuracy": 0.858595073223114, "num_tokens": 430496980.0, "step": 11336 }, { "epoch": 1.4421829283806131, "ewc_loss": 0.10107421875, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.508216857910156, "learning_rate": 1e-05, "loss": 0.5118, "mean_token_accuracy": 0.8600320816040039, "num_tokens": 430533645.0, "step": 11337 }, { "epoch": 1.4423101386592037, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 0.00011539459228515625, "grad_norm": 7.2484331130981445, "learning_rate": 1e-05, "loss": 0.522, "mean_token_accuracy": 0.8698146343231201, "num_tokens": 430567744.0, "step": 11338 }, { "epoch": 1.4424373489377942, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.746833801269531, "learning_rate": 1e-05, "loss": 0.5727, "mean_token_accuracy": 0.8426299095153809, "num_tokens": 430603887.0, "step": 11339 }, { "epoch": 1.4425645592163847, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 0.000118255615234375, "grad_norm": 7.414516448974609, "learning_rate": 1e-05, "loss": 0.5188, "mean_token_accuracy": 0.8697725534439087, "num_tokens": 430638748.0, "step": 11340 }, { "epoch": 1.4426917694949752, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.162147521972656, "learning_rate": 1e-05, "loss": 0.5088, "mean_token_accuracy": 0.8695544004440308, "num_tokens": 430677799.0, "step": 11341 }, { "epoch": 1.4428189797735658, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 0.00011157989501953125, "grad_norm": 7.186002254486084, "learning_rate": 1e-05, "loss": 0.5365, "mean_token_accuracy": 0.8637953400611877, "num_tokens": 430720816.0, "step": 11342 }, { "epoch": 1.4429461900521563, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.346800804138184, "learning_rate": 1e-05, "loss": 0.4954, "mean_token_accuracy": 0.8705915212631226, "num_tokens": 430755220.0, "step": 11343 }, { "epoch": 1.4430734003307468, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.823321342468262, "learning_rate": 1e-05, "loss": 0.4952, "mean_token_accuracy": 0.8708189129829407, "num_tokens": 430787737.0, "step": 11344 }, { "epoch": 1.4432006106093374, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.309808254241943, "learning_rate": 1e-05, "loss": 0.4926, "mean_token_accuracy": 0.8701289892196655, "num_tokens": 430823817.0, "step": 11345 }, { "epoch": 1.4433278208879279, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.4974493980407715, "learning_rate": 1e-05, "loss": 0.5197, "mean_token_accuracy": 0.8637721538543701, "num_tokens": 430863416.0, "step": 11346 }, { "epoch": 1.4434550311665182, "ewc_loss": 0.11376953125, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.207253456115723, "learning_rate": 1e-05, "loss": 0.5194, "mean_token_accuracy": 0.861109733581543, "num_tokens": 430894017.0, "step": 11347 }, { "epoch": 1.4435822414451087, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.303195476531982, "learning_rate": 1e-05, "loss": 0.5102, "mean_token_accuracy": 0.8653218150138855, "num_tokens": 430935023.0, "step": 11348 }, { "epoch": 1.4437094517236992, "ewc_loss": 0.109375, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.055322647094727, "learning_rate": 1e-05, "loss": 0.4846, "mean_token_accuracy": 0.869233250617981, "num_tokens": 430971663.0, "step": 11349 }, { "epoch": 1.4438366620022898, "ewc_loss": 0.11181640625, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.244470119476318, "learning_rate": 1e-05, "loss": 0.5077, "mean_token_accuracy": 0.8635721206665039, "num_tokens": 431018919.0, "step": 11350 }, { "epoch": 1.4439638722808803, "ewc_loss": 0.10498046875, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.889151096343994, "learning_rate": 1e-05, "loss": 0.4649, "mean_token_accuracy": 0.8786171674728394, "num_tokens": 431050076.0, "step": 11351 }, { "epoch": 1.4440910825594708, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.324159622192383, "learning_rate": 1e-05, "loss": 0.5361, "mean_token_accuracy": 0.8555797338485718, "num_tokens": 431085535.0, "step": 11352 }, { "epoch": 1.4442182928380614, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.720528602600098, "learning_rate": 1e-05, "loss": 0.4623, "mean_token_accuracy": 0.8741281032562256, "num_tokens": 431123187.0, "step": 11353 }, { "epoch": 1.4443455031166519, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.505742073059082, "learning_rate": 1e-05, "loss": 0.5398, "mean_token_accuracy": 0.8576972484588623, "num_tokens": 431157867.0, "step": 11354 }, { "epoch": 1.4444727133952424, "ewc_loss": 0.10009765625, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.570499897003174, "learning_rate": 1e-05, "loss": 0.5183, "mean_token_accuracy": 0.863219141960144, "num_tokens": 431191344.0, "step": 11355 }, { "epoch": 1.4445999236738327, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.8120832443237305, "learning_rate": 1e-05, "loss": 0.4995, "mean_token_accuracy": 0.8705853819847107, "num_tokens": 431231019.0, "step": 11356 }, { "epoch": 1.4447271339524232, "ewc_loss": 0.10009765625, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.517106533050537, "learning_rate": 1e-05, "loss": 0.4773, "mean_token_accuracy": 0.8689455389976501, "num_tokens": 431265660.0, "step": 11357 }, { "epoch": 1.4448543442310138, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 0.0001125335693359375, "grad_norm": 7.177667617797852, "learning_rate": 1e-05, "loss": 0.4892, "mean_token_accuracy": 0.8775099515914917, "num_tokens": 431298666.0, "step": 11358 }, { "epoch": 1.4449815545096043, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.731899738311768, "learning_rate": 1e-05, "loss": 0.4444, "mean_token_accuracy": 0.8828780651092529, "num_tokens": 431327516.0, "step": 11359 }, { "epoch": 1.4451087647881948, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 0.000118255615234375, "grad_norm": 7.333908557891846, "learning_rate": 1e-05, "loss": 0.5247, "mean_token_accuracy": 0.8664971590042114, "num_tokens": 431366244.0, "step": 11360 }, { "epoch": 1.4452359750667854, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.133181095123291, "learning_rate": 1e-05, "loss": 0.5118, "mean_token_accuracy": 0.8605073690414429, "num_tokens": 431402440.0, "step": 11361 }, { "epoch": 1.445363185345376, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 7.084584712982178, "learning_rate": 1e-05, "loss": 0.4683, "mean_token_accuracy": 0.8851574659347534, "num_tokens": 431436628.0, "step": 11362 }, { "epoch": 1.4454903956239664, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.362135410308838, "learning_rate": 1e-05, "loss": 0.4618, "mean_token_accuracy": 0.8811113238334656, "num_tokens": 431476343.0, "step": 11363 }, { "epoch": 1.445617605902557, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.582824230194092, "learning_rate": 1e-05, "loss": 0.4764, "mean_token_accuracy": 0.8759211301803589, "num_tokens": 431513741.0, "step": 11364 }, { "epoch": 1.4457448161811475, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.418712139129639, "learning_rate": 1e-05, "loss": 0.5382, "mean_token_accuracy": 0.8555734753608704, "num_tokens": 431556227.0, "step": 11365 }, { "epoch": 1.445872026459738, "ewc_loss": 0.10791015625, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 6.089627742767334, "learning_rate": 1e-05, "loss": 0.4819, "mean_token_accuracy": 0.8709409236907959, "num_tokens": 431592004.0, "step": 11366 }, { "epoch": 1.4459992367383285, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.523256301879883, "learning_rate": 1e-05, "loss": 0.5156, "mean_token_accuracy": 0.8646349310874939, "num_tokens": 431631286.0, "step": 11367 }, { "epoch": 1.446126447016919, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.722528457641602, "learning_rate": 1e-05, "loss": 0.5054, "mean_token_accuracy": 0.8665021657943726, "num_tokens": 431669269.0, "step": 11368 }, { "epoch": 1.4462536572955096, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.772825241088867, "learning_rate": 1e-05, "loss": 0.5309, "mean_token_accuracy": 0.860844612121582, "num_tokens": 431708061.0, "step": 11369 }, { "epoch": 1.4463808675741001, "ewc_loss": 0.10107421875, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.562545299530029, "learning_rate": 1e-05, "loss": 0.4748, "mean_token_accuracy": 0.8722996711730957, "num_tokens": 431745855.0, "step": 11370 }, { "epoch": 1.4465080778526904, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 0.00011014938354492188, "grad_norm": 7.10771369934082, "learning_rate": 1e-05, "loss": 0.518, "mean_token_accuracy": 0.8705390095710754, "num_tokens": 431784481.0, "step": 11371 }, { "epoch": 1.446635288131281, "ewc_loss": 0.10400390625, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.675886631011963, "learning_rate": 1e-05, "loss": 0.4957, "mean_token_accuracy": 0.8633328676223755, "num_tokens": 431822724.0, "step": 11372 }, { "epoch": 1.4467624984098715, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 0.00011730194091796875, "grad_norm": 7.345921516418457, "learning_rate": 1e-05, "loss": 0.5211, "mean_token_accuracy": 0.8729171752929688, "num_tokens": 431859415.0, "step": 11373 }, { "epoch": 1.446889708688462, "ewc_loss": 0.11376953125, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.022749423980713, "learning_rate": 1e-05, "loss": 0.4391, "mean_token_accuracy": 0.889167308807373, "num_tokens": 431899432.0, "step": 11374 }, { "epoch": 1.4470169189670525, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 0.00011301040649414062, "grad_norm": 7.228155612945557, "learning_rate": 1e-05, "loss": 0.4969, "mean_token_accuracy": 0.8742960095405579, "num_tokens": 431935546.0, "step": 11375 }, { "epoch": 1.447144129245643, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.275949001312256, "learning_rate": 1e-05, "loss": 0.506, "mean_token_accuracy": 0.8704624176025391, "num_tokens": 431972366.0, "step": 11376 }, { "epoch": 1.4472713395242336, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.902426242828369, "learning_rate": 1e-05, "loss": 0.5365, "mean_token_accuracy": 0.8575615882873535, "num_tokens": 432007795.0, "step": 11377 }, { "epoch": 1.4473985498028241, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.297532081604004, "learning_rate": 1e-05, "loss": 0.5682, "mean_token_accuracy": 0.8478124141693115, "num_tokens": 432047589.0, "step": 11378 }, { "epoch": 1.4475257600814146, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.572460174560547, "learning_rate": 1e-05, "loss": 0.4628, "mean_token_accuracy": 0.8787288665771484, "num_tokens": 432085615.0, "step": 11379 }, { "epoch": 1.447652970360005, "ewc_loss": 0.11474609375, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.200032711029053, "learning_rate": 1e-05, "loss": 0.4913, "mean_token_accuracy": 0.8733594417572021, "num_tokens": 432121112.0, "step": 11380 }, { "epoch": 1.4477801806385955, "ewc_loss": 0.11474609375, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.36415958404541, "learning_rate": 1e-05, "loss": 0.4853, "mean_token_accuracy": 0.8682602643966675, "num_tokens": 432156429.0, "step": 11381 }, { "epoch": 1.447907390917186, "ewc_loss": 0.109375, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.075440883636475, "learning_rate": 1e-05, "loss": 0.5479, "mean_token_accuracy": 0.8540037870407104, "num_tokens": 432195653.0, "step": 11382 }, { "epoch": 1.4480346011957765, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.284609317779541, "learning_rate": 1e-05, "loss": 0.4938, "mean_token_accuracy": 0.8685666918754578, "num_tokens": 432230997.0, "step": 11383 }, { "epoch": 1.448161811474367, "ewc_loss": 0.10595703125, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.914484977722168, "learning_rate": 1e-05, "loss": 0.4971, "mean_token_accuracy": 0.8616604804992676, "num_tokens": 432270868.0, "step": 11384 }, { "epoch": 1.4482890217529576, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.329700469970703, "learning_rate": 1e-05, "loss": 0.4933, "mean_token_accuracy": 0.8711177706718445, "num_tokens": 432309999.0, "step": 11385 }, { "epoch": 1.4484162320315481, "ewc_loss": 0.10205078125, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.748166561126709, "learning_rate": 1e-05, "loss": 0.4545, "mean_token_accuracy": 0.8802286386489868, "num_tokens": 432345240.0, "step": 11386 }, { "epoch": 1.4485434423101387, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.4828410148620605, "learning_rate": 1e-05, "loss": 0.5272, "mean_token_accuracy": 0.8606767654418945, "num_tokens": 432391478.0, "step": 11387 }, { "epoch": 1.4486706525887292, "ewc_loss": 0.10107421875, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.59562349319458, "learning_rate": 1e-05, "loss": 0.432, "mean_token_accuracy": 0.8843367695808411, "num_tokens": 432430601.0, "step": 11388 }, { "epoch": 1.4487978628673197, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.770340442657471, "learning_rate": 1e-05, "loss": 0.5222, "mean_token_accuracy": 0.8663449883460999, "num_tokens": 432467260.0, "step": 11389 }, { "epoch": 1.4489250731459102, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.540604591369629, "learning_rate": 1e-05, "loss": 0.5226, "mean_token_accuracy": 0.8523522615432739, "num_tokens": 432506270.0, "step": 11390 }, { "epoch": 1.4490522834245008, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 7.111732006072998, "learning_rate": 1e-05, "loss": 0.5197, "mean_token_accuracy": 0.865024745464325, "num_tokens": 432543688.0, "step": 11391 }, { "epoch": 1.4491794937030913, "ewc_loss": 0.10595703125, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.716330528259277, "learning_rate": 1e-05, "loss": 0.4674, "mean_token_accuracy": 0.8770548701286316, "num_tokens": 432583522.0, "step": 11392 }, { "epoch": 1.4493067039816818, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 0.00011539459228515625, "grad_norm": 7.271193027496338, "learning_rate": 1e-05, "loss": 0.4736, "mean_token_accuracy": 0.8835740685462952, "num_tokens": 432622471.0, "step": 11393 }, { "epoch": 1.4494339142602723, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.087726593017578, "learning_rate": 1e-05, "loss": 0.4657, "mean_token_accuracy": 0.8807008266448975, "num_tokens": 432666689.0, "step": 11394 }, { "epoch": 1.4495611245388629, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 7.0424723625183105, "learning_rate": 1e-05, "loss": 0.4939, "mean_token_accuracy": 0.8727972507476807, "num_tokens": 432707816.0, "step": 11395 }, { "epoch": 1.4496883348174532, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.311532974243164, "learning_rate": 1e-05, "loss": 0.4584, "mean_token_accuracy": 0.8833615779876709, "num_tokens": 432746873.0, "step": 11396 }, { "epoch": 1.4498155450960437, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.561514854431152, "learning_rate": 1e-05, "loss": 0.454, "mean_token_accuracy": 0.8820544481277466, "num_tokens": 432777825.0, "step": 11397 }, { "epoch": 1.4499427553746342, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.394763469696045, "learning_rate": 1e-05, "loss": 0.5074, "mean_token_accuracy": 0.8670463562011719, "num_tokens": 432816806.0, "step": 11398 }, { "epoch": 1.4500699656532248, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 6.047276973724365, "learning_rate": 1e-05, "loss": 0.4786, "mean_token_accuracy": 0.8787384629249573, "num_tokens": 432855726.0, "step": 11399 }, { "epoch": 1.4501971759318153, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.535799980163574, "learning_rate": 1e-05, "loss": 0.5267, "mean_token_accuracy": 0.8609704375267029, "num_tokens": 432891526.0, "step": 11400 }, { "epoch": 1.4503243862104058, "ewc_loss": 0.10205078125, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.667847156524658, "learning_rate": 1e-05, "loss": 0.4072, "mean_token_accuracy": 0.8910049200057983, "num_tokens": 432921995.0, "step": 11401 }, { "epoch": 1.4504515964889964, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.825441360473633, "learning_rate": 1e-05, "loss": 0.5205, "mean_token_accuracy": 0.864456295967102, "num_tokens": 432962120.0, "step": 11402 }, { "epoch": 1.4505788067675869, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.515452861785889, "learning_rate": 1e-05, "loss": 0.4854, "mean_token_accuracy": 0.8678332567214966, "num_tokens": 433001944.0, "step": 11403 }, { "epoch": 1.4507060170461774, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 0.00011348724365234375, "grad_norm": 7.191478252410889, "learning_rate": 1e-05, "loss": 0.5699, "mean_token_accuracy": 0.8499382734298706, "num_tokens": 433040984.0, "step": 11404 }, { "epoch": 1.4508332273247677, "ewc_loss": 0.10595703125, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.708313941955566, "learning_rate": 1e-05, "loss": 0.5031, "mean_token_accuracy": 0.8649255633354187, "num_tokens": 433077914.0, "step": 11405 }, { "epoch": 1.4509604376033582, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 0.00011873245239257812, "grad_norm": 7.391010761260986, "learning_rate": 1e-05, "loss": 0.4867, "mean_token_accuracy": 0.8782259225845337, "num_tokens": 433111219.0, "step": 11406 }, { "epoch": 1.4510876478819488, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.104256629943848, "learning_rate": 1e-05, "loss": 0.4475, "mean_token_accuracy": 0.8858522772789001, "num_tokens": 433143204.0, "step": 11407 }, { "epoch": 1.4512148581605393, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 0.00011301040649414062, "grad_norm": 7.208877086639404, "learning_rate": 1e-05, "loss": 0.4713, "mean_token_accuracy": 0.8842005729675293, "num_tokens": 433175361.0, "step": 11408 }, { "epoch": 1.4513420684391298, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.315323829650879, "learning_rate": 1e-05, "loss": 0.4732, "mean_token_accuracy": 0.8789803981781006, "num_tokens": 433215303.0, "step": 11409 }, { "epoch": 1.4514692787177204, "ewc_loss": 0.125, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.847286224365234, "learning_rate": 1e-05, "loss": 0.4832, "mean_token_accuracy": 0.8751050233840942, "num_tokens": 433255301.0, "step": 11410 }, { "epoch": 1.4515964889963109, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.30194616317749, "learning_rate": 1e-05, "loss": 0.5279, "mean_token_accuracy": 0.8598102331161499, "num_tokens": 433291110.0, "step": 11411 }, { "epoch": 1.4517236992749014, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.535498142242432, "learning_rate": 1e-05, "loss": 0.5011, "mean_token_accuracy": 0.8702112436294556, "num_tokens": 433330910.0, "step": 11412 }, { "epoch": 1.451850909553492, "ewc_loss": 0.11376953125, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.190149307250977, "learning_rate": 1e-05, "loss": 0.4928, "mean_token_accuracy": 0.8705481290817261, "num_tokens": 433364297.0, "step": 11413 }, { "epoch": 1.4519781198320825, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.338208198547363, "learning_rate": 1e-05, "loss": 0.506, "mean_token_accuracy": 0.8665906190872192, "num_tokens": 433405900.0, "step": 11414 }, { "epoch": 1.452105330110673, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 6.05712366104126, "learning_rate": 1e-05, "loss": 0.5165, "mean_token_accuracy": 0.8605804443359375, "num_tokens": 433442639.0, "step": 11415 }, { "epoch": 1.4522325403892635, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.2912917137146, "learning_rate": 1e-05, "loss": 0.5023, "mean_token_accuracy": 0.8648866415023804, "num_tokens": 433471657.0, "step": 11416 }, { "epoch": 1.452359750667854, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.907236576080322, "learning_rate": 1e-05, "loss": 0.5368, "mean_token_accuracy": 0.8507798910140991, "num_tokens": 433510936.0, "step": 11417 }, { "epoch": 1.4524869609464446, "ewc_loss": 0.11376953125, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.348935604095459, "learning_rate": 1e-05, "loss": 0.501, "mean_token_accuracy": 0.8676670789718628, "num_tokens": 433546940.0, "step": 11418 }, { "epoch": 1.4526141712250351, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.718170166015625, "learning_rate": 1e-05, "loss": 0.4764, "mean_token_accuracy": 0.8710126876831055, "num_tokens": 433588701.0, "step": 11419 }, { "epoch": 1.4527413815036254, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.521669387817383, "learning_rate": 1e-05, "loss": 0.4866, "mean_token_accuracy": 0.8691593408584595, "num_tokens": 433625678.0, "step": 11420 }, { "epoch": 1.452868591782216, "ewc_loss": 0.10107421875, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.5986151695251465, "learning_rate": 1e-05, "loss": 0.5493, "mean_token_accuracy": 0.8524214625358582, "num_tokens": 433660624.0, "step": 11421 }, { "epoch": 1.4529958020608065, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.796043395996094, "learning_rate": 1e-05, "loss": 0.4741, "mean_token_accuracy": 0.8770943284034729, "num_tokens": 433702737.0, "step": 11422 }, { "epoch": 1.453123012339397, "ewc_loss": 0.10205078125, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.55354642868042, "learning_rate": 1e-05, "loss": 0.4865, "mean_token_accuracy": 0.8668627142906189, "num_tokens": 433741866.0, "step": 11423 }, { "epoch": 1.4532502226179875, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 0.00011205673217773438, "grad_norm": 7.132527828216553, "learning_rate": 1e-05, "loss": 0.5452, "mean_token_accuracy": 0.8606624603271484, "num_tokens": 433779869.0, "step": 11424 }, { "epoch": 1.453377432896578, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.7408318519592285, "learning_rate": 1e-05, "loss": 0.4586, "mean_token_accuracy": 0.8806646466255188, "num_tokens": 433819629.0, "step": 11425 }, { "epoch": 1.4535046431751686, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 0.00011444091796875, "grad_norm": 7.270575523376465, "learning_rate": 1e-05, "loss": 0.492, "mean_token_accuracy": 0.8752132654190063, "num_tokens": 433855330.0, "step": 11426 }, { "epoch": 1.4536318534537591, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.112390995025635, "learning_rate": 1e-05, "loss": 0.5162, "mean_token_accuracy": 0.8617952466011047, "num_tokens": 433896130.0, "step": 11427 }, { "epoch": 1.4537590637323496, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 7.000729560852051, "learning_rate": 1e-05, "loss": 0.4842, "mean_token_accuracy": 0.8788686990737915, "num_tokens": 433928514.0, "step": 11428 }, { "epoch": 1.45388627401094, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.3285980224609375, "learning_rate": 1e-05, "loss": 0.506, "mean_token_accuracy": 0.865410327911377, "num_tokens": 433969213.0, "step": 11429 }, { "epoch": 1.4540134842895305, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.499680519104004, "learning_rate": 1e-05, "loss": 0.55, "mean_token_accuracy": 0.8547937273979187, "num_tokens": 434007902.0, "step": 11430 }, { "epoch": 1.454140694568121, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.408004283905029, "learning_rate": 1e-05, "loss": 0.5566, "mean_token_accuracy": 0.8576081991195679, "num_tokens": 434043243.0, "step": 11431 }, { "epoch": 1.4542679048467115, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.9978790283203125, "learning_rate": 1e-05, "loss": 0.4392, "mean_token_accuracy": 0.8821422457695007, "num_tokens": 434081075.0, "step": 11432 }, { "epoch": 1.454395115125302, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.569075107574463, "learning_rate": 1e-05, "loss": 0.503, "mean_token_accuracy": 0.8670533895492554, "num_tokens": 434121230.0, "step": 11433 }, { "epoch": 1.4545223254038926, "ewc_loss": 0.10205078125, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.640986919403076, "learning_rate": 1e-05, "loss": 0.4954, "mean_token_accuracy": 0.8654027581214905, "num_tokens": 434158324.0, "step": 11434 }, { "epoch": 1.4546495356824831, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.865265846252441, "learning_rate": 1e-05, "loss": 0.5014, "mean_token_accuracy": 0.8713375926017761, "num_tokens": 434203910.0, "step": 11435 }, { "epoch": 1.4547767459610736, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.523799896240234, "learning_rate": 1e-05, "loss": 0.4565, "mean_token_accuracy": 0.8799477815628052, "num_tokens": 434238915.0, "step": 11436 }, { "epoch": 1.4549039562396642, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 0.00011491775512695312, "grad_norm": 7.237612247467041, "learning_rate": 1e-05, "loss": 0.5101, "mean_token_accuracy": 0.8720995187759399, "num_tokens": 434280049.0, "step": 11437 }, { "epoch": 1.4550311665182547, "ewc_loss": 0.10693359375, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.746250152587891, "learning_rate": 1e-05, "loss": 0.4632, "mean_token_accuracy": 0.8790467977523804, "num_tokens": 434323980.0, "step": 11438 }, { "epoch": 1.4551583767968452, "ewc_loss": 0.140625, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 0.0001201629638671875, "grad_norm": 7.397531032562256, "learning_rate": 1e-05, "loss": 0.5408, "mean_token_accuracy": 0.8638433218002319, "num_tokens": 434363386.0, "step": 11439 }, { "epoch": 1.4552855870754358, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.1444993019104, "learning_rate": 1e-05, "loss": 0.4642, "mean_token_accuracy": 0.8795976638793945, "num_tokens": 434399480.0, "step": 11440 }, { "epoch": 1.4554127973540263, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 0.00011301040649414062, "grad_norm": 7.17840051651001, "learning_rate": 1e-05, "loss": 0.5233, "mean_token_accuracy": 0.8666138648986816, "num_tokens": 434445662.0, "step": 11441 }, { "epoch": 1.4555400076326168, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.33690071105957, "learning_rate": 1e-05, "loss": 0.4938, "mean_token_accuracy": 0.8708150386810303, "num_tokens": 434480926.0, "step": 11442 }, { "epoch": 1.4556672179112073, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.817619323730469, "learning_rate": 1e-05, "loss": 0.5297, "mean_token_accuracy": 0.8590005040168762, "num_tokens": 434520018.0, "step": 11443 }, { "epoch": 1.4557944281897979, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.298048496246338, "learning_rate": 1e-05, "loss": 0.5378, "mean_token_accuracy": 0.8569502234458923, "num_tokens": 434555163.0, "step": 11444 }, { "epoch": 1.4559216384683882, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.513650894165039, "learning_rate": 1e-05, "loss": 0.4494, "mean_token_accuracy": 0.881756603717804, "num_tokens": 434586353.0, "step": 11445 }, { "epoch": 1.4560488487469787, "ewc_loss": 0.11376953125, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.176705360412598, "learning_rate": 1e-05, "loss": 0.4807, "mean_token_accuracy": 0.8751981258392334, "num_tokens": 434630558.0, "step": 11446 }, { "epoch": 1.4561760590255692, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.333581924438477, "learning_rate": 1e-05, "loss": 0.498, "mean_token_accuracy": 0.8686356544494629, "num_tokens": 434672949.0, "step": 11447 }, { "epoch": 1.4563032693041598, "ewc_loss": 0.10791015625, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 6.0332350730896, "learning_rate": 1e-05, "loss": 0.5025, "mean_token_accuracy": 0.8653476238250732, "num_tokens": 434709207.0, "step": 11448 }, { "epoch": 1.4564304795827503, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.280899524688721, "learning_rate": 1e-05, "loss": 0.4545, "mean_token_accuracy": 0.8812230825424194, "num_tokens": 434745787.0, "step": 11449 }, { "epoch": 1.4565576898613408, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.875502586364746, "learning_rate": 1e-05, "loss": 0.4729, "mean_token_accuracy": 0.8703038692474365, "num_tokens": 434785419.0, "step": 11450 }, { "epoch": 1.4566849001399313, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.366133689880371, "learning_rate": 1e-05, "loss": 0.5065, "mean_token_accuracy": 0.8619133234024048, "num_tokens": 434818217.0, "step": 11451 }, { "epoch": 1.4568121104185219, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.710351467132568, "learning_rate": 1e-05, "loss": 0.5131, "mean_token_accuracy": 0.8605966567993164, "num_tokens": 434855654.0, "step": 11452 }, { "epoch": 1.4569393206971124, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.559730529785156, "learning_rate": 1e-05, "loss": 0.5478, "mean_token_accuracy": 0.8548113703727722, "num_tokens": 434890657.0, "step": 11453 }, { "epoch": 1.4570665309757027, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.57543420791626, "learning_rate": 1e-05, "loss": 0.4529, "mean_token_accuracy": 0.8805457949638367, "num_tokens": 434929164.0, "step": 11454 }, { "epoch": 1.4571937412542932, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.856978893280029, "learning_rate": 1e-05, "loss": 0.5307, "mean_token_accuracy": 0.8626235723495483, "num_tokens": 434966017.0, "step": 11455 }, { "epoch": 1.4573209515328838, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.561336994171143, "learning_rate": 1e-05, "loss": 0.4533, "mean_token_accuracy": 0.8785346746444702, "num_tokens": 435002076.0, "step": 11456 }, { "epoch": 1.4574481618114743, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 0.00011301040649414062, "grad_norm": 7.191391468048096, "learning_rate": 1e-05, "loss": 0.4784, "mean_token_accuracy": 0.8755329847335815, "num_tokens": 435030058.0, "step": 11457 }, { "epoch": 1.4575753720900648, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 5.803898334503174, "learning_rate": 1e-05, "loss": 0.5241, "mean_token_accuracy": 0.8632394075393677, "num_tokens": 435067816.0, "step": 11458 }, { "epoch": 1.4577025823686554, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 0.00011587142944335938, "grad_norm": 7.262269496917725, "learning_rate": 1e-05, "loss": 0.5079, "mean_token_accuracy": 0.8735197186470032, "num_tokens": 435106078.0, "step": 11459 }, { "epoch": 1.4578297926472459, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.169809818267822, "learning_rate": 1e-05, "loss": 0.4685, "mean_token_accuracy": 0.8809795379638672, "num_tokens": 435144558.0, "step": 11460 }, { "epoch": 1.4579570029258364, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.9468159675598145, "learning_rate": 1e-05, "loss": 0.5246, "mean_token_accuracy": 0.8660268783569336, "num_tokens": 435183083.0, "step": 11461 }, { "epoch": 1.458084213204427, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.341957092285156, "learning_rate": 1e-05, "loss": 0.4409, "mean_token_accuracy": 0.8862478137016296, "num_tokens": 435223459.0, "step": 11462 }, { "epoch": 1.4582114234830175, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.448574542999268, "learning_rate": 1e-05, "loss": 0.5, "mean_token_accuracy": 0.8752884864807129, "num_tokens": 435261491.0, "step": 11463 }, { "epoch": 1.458338633761608, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.425467491149902, "learning_rate": 1e-05, "loss": 0.54, "mean_token_accuracy": 0.8586359620094299, "num_tokens": 435299634.0, "step": 11464 }, { "epoch": 1.4584658440401985, "ewc_loss": 0.10595703125, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.962194442749023, "learning_rate": 1e-05, "loss": 0.5294, "mean_token_accuracy": 0.8537498712539673, "num_tokens": 435339645.0, "step": 11465 }, { "epoch": 1.458593054318789, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.5782694816589355, "learning_rate": 1e-05, "loss": 0.5124, "mean_token_accuracy": 0.867588222026825, "num_tokens": 435382635.0, "step": 11466 }, { "epoch": 1.4587202645973796, "ewc_loss": 0.10107421875, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.624539375305176, "learning_rate": 1e-05, "loss": 0.5028, "mean_token_accuracy": 0.860053300857544, "num_tokens": 435421653.0, "step": 11467 }, { "epoch": 1.45884747487597, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.887556076049805, "learning_rate": 1e-05, "loss": 0.5362, "mean_token_accuracy": 0.8591326475143433, "num_tokens": 435460135.0, "step": 11468 }, { "epoch": 1.4589746851545604, "ewc_loss": 0.10205078125, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.531094074249268, "learning_rate": 1e-05, "loss": 0.4456, "mean_token_accuracy": 0.8851524591445923, "num_tokens": 435497717.0, "step": 11469 }, { "epoch": 1.459101895433151, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 0.0001163482666015625, "grad_norm": 7.244132995605469, "learning_rate": 1e-05, "loss": 0.5629, "mean_token_accuracy": 0.8582770824432373, "num_tokens": 435535247.0, "step": 11470 }, { "epoch": 1.4592291057117415, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 5.767653942108154, "learning_rate": 1e-05, "loss": 0.4609, "mean_token_accuracy": 0.8767657279968262, "num_tokens": 435576841.0, "step": 11471 }, { "epoch": 1.459356315990332, "ewc_loss": 0.140625, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 0.00011968612670898438, "grad_norm": 7.396356582641602, "learning_rate": 1e-05, "loss": 0.5173, "mean_token_accuracy": 0.8717368841171265, "num_tokens": 435618197.0, "step": 11472 }, { "epoch": 1.4594835262689225, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.161060810089111, "learning_rate": 1e-05, "loss": 0.585, "mean_token_accuracy": 0.8447267413139343, "num_tokens": 435663327.0, "step": 11473 }, { "epoch": 1.459610736547513, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 0.00011205673217773438, "grad_norm": 7.186366081237793, "learning_rate": 1e-05, "loss": 0.5213, "mean_token_accuracy": 0.8684853315353394, "num_tokens": 435697795.0, "step": 11474 }, { "epoch": 1.4597379468261036, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.463889122009277, "learning_rate": 1e-05, "loss": 0.5455, "mean_token_accuracy": 0.859662652015686, "num_tokens": 435734308.0, "step": 11475 }, { "epoch": 1.4598651571046941, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.730393886566162, "learning_rate": 1e-05, "loss": 0.5146, "mean_token_accuracy": 0.8635172843933105, "num_tokens": 435772575.0, "step": 11476 }, { "epoch": 1.4599923673832846, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.291924476623535, "learning_rate": 1e-05, "loss": 0.458, "mean_token_accuracy": 0.8812448978424072, "num_tokens": 435807578.0, "step": 11477 }, { "epoch": 1.460119577661875, "ewc_loss": 0.11474609375, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.418713569641113, "learning_rate": 1e-05, "loss": 0.4985, "mean_token_accuracy": 0.8674613833427429, "num_tokens": 435849035.0, "step": 11478 }, { "epoch": 1.4602467879404655, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.167027473449707, "learning_rate": 1e-05, "loss": 0.4696, "mean_token_accuracy": 0.8714011907577515, "num_tokens": 435884912.0, "step": 11479 }, { "epoch": 1.460373998219056, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.273569583892822, "learning_rate": 1e-05, "loss": 0.4745, "mean_token_accuracy": 0.8776525259017944, "num_tokens": 435928822.0, "step": 11480 }, { "epoch": 1.4605012084976465, "ewc_loss": 0.109375, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.0208659172058105, "learning_rate": 1e-05, "loss": 0.4721, "mean_token_accuracy": 0.877822995185852, "num_tokens": 435963906.0, "step": 11481 }, { "epoch": 1.460628418776237, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.274053573608398, "learning_rate": 1e-05, "loss": 0.5526, "mean_token_accuracy": 0.8536435961723328, "num_tokens": 436002826.0, "step": 11482 }, { "epoch": 1.4607556290548276, "ewc_loss": 0.10498046875, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.8344621658325195, "learning_rate": 1e-05, "loss": 0.435, "mean_token_accuracy": 0.8854113817214966, "num_tokens": 436039043.0, "step": 11483 }, { "epoch": 1.4608828393334181, "ewc_loss": 0.11474609375, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.3735504150390625, "learning_rate": 1e-05, "loss": 0.4632, "mean_token_accuracy": 0.878261923789978, "num_tokens": 436078081.0, "step": 11484 }, { "epoch": 1.4610100496120086, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.6897969245910645, "learning_rate": 1e-05, "loss": 0.5314, "mean_token_accuracy": 0.8539609909057617, "num_tokens": 436110415.0, "step": 11485 }, { "epoch": 1.4611372598905992, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.576231002807617, "learning_rate": 1e-05, "loss": 0.4849, "mean_token_accuracy": 0.8781220316886902, "num_tokens": 436152473.0, "step": 11486 }, { "epoch": 1.4612644701691897, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.533022880554199, "learning_rate": 1e-05, "loss": 0.5087, "mean_token_accuracy": 0.8664572238922119, "num_tokens": 436192410.0, "step": 11487 }, { "epoch": 1.4613916804477802, "ewc_loss": 0.125, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.929631233215332, "learning_rate": 1e-05, "loss": 0.5405, "mean_token_accuracy": 0.8589626550674438, "num_tokens": 436230605.0, "step": 11488 }, { "epoch": 1.4615188907263708, "ewc_loss": 0.10205078125, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.528058052062988, "learning_rate": 1e-05, "loss": 0.4673, "mean_token_accuracy": 0.8765084743499756, "num_tokens": 436267951.0, "step": 11489 }, { "epoch": 1.4616461010049613, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 0.00011730194091796875, "grad_norm": 7.289938926696777, "learning_rate": 1e-05, "loss": 0.5616, "mean_token_accuracy": 0.8526425361633301, "num_tokens": 436306685.0, "step": 11490 }, { "epoch": 1.4617733112835518, "ewc_loss": 0.11181640625, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 5.838910102844238, "learning_rate": 1e-05, "loss": 0.4903, "mean_token_accuracy": 0.8721863031387329, "num_tokens": 436345568.0, "step": 11491 }, { "epoch": 1.4619005215621423, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 0.00011682510375976562, "grad_norm": 7.349706649780273, "learning_rate": 1e-05, "loss": 0.5098, "mean_token_accuracy": 0.8704261779785156, "num_tokens": 436387580.0, "step": 11492 }, { "epoch": 1.4620277318407329, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.258790969848633, "learning_rate": 1e-05, "loss": 0.517, "mean_token_accuracy": 0.8652973175048828, "num_tokens": 436423000.0, "step": 11493 }, { "epoch": 1.4621549421193232, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.980256080627441, "learning_rate": 1e-05, "loss": 0.506, "mean_token_accuracy": 0.8712810277938843, "num_tokens": 436464366.0, "step": 11494 }, { "epoch": 1.4622821523979137, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.405952453613281, "learning_rate": 1e-05, "loss": 0.5441, "mean_token_accuracy": 0.8575208783149719, "num_tokens": 436505385.0, "step": 11495 }, { "epoch": 1.4624093626765042, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.484830379486084, "learning_rate": 1e-05, "loss": 0.5217, "mean_token_accuracy": 0.8588718175888062, "num_tokens": 436546251.0, "step": 11496 }, { "epoch": 1.4625365729550948, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.434935569763184, "learning_rate": 1e-05, "loss": 0.5393, "mean_token_accuracy": 0.8593191504478455, "num_tokens": 436582136.0, "step": 11497 }, { "epoch": 1.4626637832336853, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 6.029135704040527, "learning_rate": 1e-05, "loss": 0.4836, "mean_token_accuracy": 0.8705313801765442, "num_tokens": 436620027.0, "step": 11498 }, { "epoch": 1.4627909935122758, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.5222344398498535, "learning_rate": 1e-05, "loss": 0.5148, "mean_token_accuracy": 0.866195559501648, "num_tokens": 436656720.0, "step": 11499 }, { "epoch": 1.4629182037908663, "ewc_loss": 0.10302734375, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.728183269500732, "learning_rate": 1e-05, "loss": 0.4707, "mean_token_accuracy": 0.8785676956176758, "num_tokens": 436693813.0, "step": 11500 }, { "epoch": 1.4630454140694569, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.715343952178955, "learning_rate": 1e-05, "loss": 0.491, "mean_token_accuracy": 0.8721526265144348, "num_tokens": 436732023.0, "step": 11501 }, { "epoch": 1.4631726243480474, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.614377498626709, "learning_rate": 1e-05, "loss": 0.499, "mean_token_accuracy": 0.8644287586212158, "num_tokens": 436772330.0, "step": 11502 }, { "epoch": 1.4632998346266377, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.958059787750244, "learning_rate": 1e-05, "loss": 0.5197, "mean_token_accuracy": 0.8654093742370605, "num_tokens": 436814593.0, "step": 11503 }, { "epoch": 1.4634270449052282, "ewc_loss": 0.10302734375, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.644097805023193, "learning_rate": 1e-05, "loss": 0.4962, "mean_token_accuracy": 0.8696906566619873, "num_tokens": 436849951.0, "step": 11504 }, { "epoch": 1.4635542551838188, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 0.0001125335693359375, "grad_norm": 7.176178932189941, "learning_rate": 1e-05, "loss": 0.5391, "mean_token_accuracy": 0.8550411462783813, "num_tokens": 436889113.0, "step": 11505 }, { "epoch": 1.4636814654624093, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 5.873699188232422, "learning_rate": 1e-05, "loss": 0.4623, "mean_token_accuracy": 0.8790071606636047, "num_tokens": 436921009.0, "step": 11506 }, { "epoch": 1.4638086757409998, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 0.00011157989501953125, "grad_norm": 7.192899703979492, "learning_rate": 1e-05, "loss": 0.5098, "mean_token_accuracy": 0.8710484504699707, "num_tokens": 436958425.0, "step": 11507 }, { "epoch": 1.4639358860195903, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.112184047698975, "learning_rate": 1e-05, "loss": 0.4963, "mean_token_accuracy": 0.871044397354126, "num_tokens": 436992995.0, "step": 11508 }, { "epoch": 1.4640630962981809, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.974235534667969, "learning_rate": 1e-05, "loss": 0.4797, "mean_token_accuracy": 0.8766571879386902, "num_tokens": 437028884.0, "step": 11509 }, { "epoch": 1.4641903065767714, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.210472106933594, "learning_rate": 1e-05, "loss": 0.477, "mean_token_accuracy": 0.8743370175361633, "num_tokens": 437071753.0, "step": 11510 }, { "epoch": 1.464317516855362, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.66212272644043, "learning_rate": 1e-05, "loss": 0.4753, "mean_token_accuracy": 0.8768140077590942, "num_tokens": 437108783.0, "step": 11511 }, { "epoch": 1.4644447271339525, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.00020694732666015625, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.185533046722412, "learning_rate": 1e-05, "loss": 0.4822, "mean_token_accuracy": 0.8734370470046997, "num_tokens": 437143234.0, "step": 11512 }, { "epoch": 1.464571937412543, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.00020694732666015625, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.406758785247803, "learning_rate": 1e-05, "loss": 0.5121, "mean_token_accuracy": 0.8642245531082153, "num_tokens": 437182363.0, "step": 11513 }, { "epoch": 1.4646991476911335, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.00020694732666015625, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.111473560333252, "learning_rate": 1e-05, "loss": 0.4879, "mean_token_accuracy": 0.8744058609008789, "num_tokens": 437224680.0, "step": 11514 }, { "epoch": 1.464826357969724, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.250130653381348, "learning_rate": 1e-05, "loss": 0.486, "mean_token_accuracy": 0.8705599308013916, "num_tokens": 437261154.0, "step": 11515 }, { "epoch": 1.4649535682483146, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.00020694732666015625, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 6.006505489349365, "learning_rate": 1e-05, "loss": 0.4871, "mean_token_accuracy": 0.8691198229789734, "num_tokens": 437301654.0, "step": 11516 }, { "epoch": 1.465080778526905, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.00020694732666015625, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.208398818969727, "learning_rate": 1e-05, "loss": 0.5027, "mean_token_accuracy": 0.8680399656295776, "num_tokens": 437338437.0, "step": 11517 }, { "epoch": 1.4652079888054954, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.884467601776123, "learning_rate": 1e-05, "loss": 0.5077, "mean_token_accuracy": 0.86443030834198, "num_tokens": 437384258.0, "step": 11518 }, { "epoch": 1.465335199084086, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.00020694732666015625, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.256806373596191, "learning_rate": 1e-05, "loss": 0.5367, "mean_token_accuracy": 0.8569878339767456, "num_tokens": 437421762.0, "step": 11519 }, { "epoch": 1.4654624093626765, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.00020694732666015625, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.741604328155518, "learning_rate": 1e-05, "loss": 0.4482, "mean_token_accuracy": 0.8781782984733582, "num_tokens": 437456797.0, "step": 11520 }, { "epoch": 1.465589619641267, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.00020694732666015625, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.413140773773193, "learning_rate": 1e-05, "loss": 0.4657, "mean_token_accuracy": 0.879991352558136, "num_tokens": 437484435.0, "step": 11521 }, { "epoch": 1.4657168299198575, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.00020694732666015625, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.604780673980713, "learning_rate": 1e-05, "loss": 0.4903, "mean_token_accuracy": 0.8693380355834961, "num_tokens": 437522573.0, "step": 11522 }, { "epoch": 1.465844040198448, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.00020694732666015625, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.657279014587402, "learning_rate": 1e-05, "loss": 0.531, "mean_token_accuracy": 0.856691300868988, "num_tokens": 437560204.0, "step": 11523 }, { "epoch": 1.4659712504770386, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.00020694732666015625, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.507644176483154, "learning_rate": 1e-05, "loss": 0.4932, "mean_token_accuracy": 0.8656374216079712, "num_tokens": 437595845.0, "step": 11524 }, { "epoch": 1.466098460755629, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.00020694732666015625, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 7.000726222991943, "learning_rate": 1e-05, "loss": 0.481, "mean_token_accuracy": 0.8742791414260864, "num_tokens": 437631985.0, "step": 11525 }, { "epoch": 1.4662256710342196, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.00020694732666015625, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.575559616088867, "learning_rate": 1e-05, "loss": 0.5242, "mean_token_accuracy": 0.8593658804893494, "num_tokens": 437677024.0, "step": 11526 }, { "epoch": 1.46635288131281, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.00020694732666015625, "ewc_loss_parallel": 0.00011730194091796875, "grad_norm": 7.306639194488525, "learning_rate": 1e-05, "loss": 0.4682, "mean_token_accuracy": 0.8850324153900146, "num_tokens": 437714980.0, "step": 11527 }, { "epoch": 1.4664800915914005, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.00020694732666015625, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 5.93541145324707, "learning_rate": 1e-05, "loss": 0.5043, "mean_token_accuracy": 0.8670819997787476, "num_tokens": 437752444.0, "step": 11528 }, { "epoch": 1.466607301869991, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.00020694732666015625, "ewc_loss_parallel": 0.00011444091796875, "grad_norm": 7.249002933502197, "learning_rate": 1e-05, "loss": 0.5314, "mean_token_accuracy": 0.862592339515686, "num_tokens": 437787032.0, "step": 11529 }, { "epoch": 1.4667345121485815, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.00020694732666015625, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.2793965339660645, "learning_rate": 1e-05, "loss": 0.5061, "mean_token_accuracy": 0.8672820925712585, "num_tokens": 437827560.0, "step": 11530 }, { "epoch": 1.466861722427172, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.00020694732666015625, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.819629192352295, "learning_rate": 1e-05, "loss": 0.5187, "mean_token_accuracy": 0.8661845922470093, "num_tokens": 437865339.0, "step": 11531 }, { "epoch": 1.4669889327057626, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.00020694732666015625, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.391636848449707, "learning_rate": 1e-05, "loss": 0.4993, "mean_token_accuracy": 0.8690478205680847, "num_tokens": 437899769.0, "step": 11532 }, { "epoch": 1.467116142984353, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.00020694732666015625, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.310810565948486, "learning_rate": 1e-05, "loss": 0.4869, "mean_token_accuracy": 0.8727155923843384, "num_tokens": 437934738.0, "step": 11533 }, { "epoch": 1.4672433532629436, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.00020694732666015625, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.44571590423584, "learning_rate": 1e-05, "loss": 0.5171, "mean_token_accuracy": 0.8629745244979858, "num_tokens": 437975948.0, "step": 11534 }, { "epoch": 1.4673705635415342, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.00020694732666015625, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.882871627807617, "learning_rate": 1e-05, "loss": 0.468, "mean_token_accuracy": 0.8776094913482666, "num_tokens": 438010081.0, "step": 11535 }, { "epoch": 1.4674977738201247, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.00020694732666015625, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.618898391723633, "learning_rate": 1e-05, "loss": 0.4934, "mean_token_accuracy": 0.872512936592102, "num_tokens": 438044069.0, "step": 11536 }, { "epoch": 1.4676249840987152, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.00020694732666015625, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.62785005569458, "learning_rate": 1e-05, "loss": 0.4711, "mean_token_accuracy": 0.8724502921104431, "num_tokens": 438080866.0, "step": 11537 }, { "epoch": 1.4677521943773058, "ewc_loss": 0.125, "ewc_loss_diag": 0.00020694732666015625, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.890604019165039, "learning_rate": 1e-05, "loss": 0.5049, "mean_token_accuracy": 0.8684234023094177, "num_tokens": 438120550.0, "step": 11538 }, { "epoch": 1.4678794046558963, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.00020694732666015625, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.593847751617432, "learning_rate": 1e-05, "loss": 0.4989, "mean_token_accuracy": 0.8650979995727539, "num_tokens": 438156968.0, "step": 11539 }, { "epoch": 1.4680066149344868, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.00020694732666015625, "ewc_loss_parallel": 0.00011301040649414062, "grad_norm": 7.174790382385254, "learning_rate": 1e-05, "loss": 0.5103, "mean_token_accuracy": 0.87129145860672, "num_tokens": 438191470.0, "step": 11540 }, { "epoch": 1.4681338252130773, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.00020694732666015625, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 5.814669132232666, "learning_rate": 1e-05, "loss": 0.5358, "mean_token_accuracy": 0.8557787537574768, "num_tokens": 438225944.0, "step": 11541 }, { "epoch": 1.4682610354916679, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.00020694732666015625, "ewc_loss_parallel": 0.0001125335693359375, "grad_norm": 7.230949878692627, "learning_rate": 1e-05, "loss": 0.5107, "mean_token_accuracy": 0.8710061311721802, "num_tokens": 438263759.0, "step": 11542 }, { "epoch": 1.4683882457702582, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.00020694732666015625, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.079904556274414, "learning_rate": 1e-05, "loss": 0.4801, "mean_token_accuracy": 0.8706457614898682, "num_tokens": 438302371.0, "step": 11543 }, { "epoch": 1.4685154560488487, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.00020694732666015625, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 7.0369110107421875, "learning_rate": 1e-05, "loss": 0.5648, "mean_token_accuracy": 0.8526272773742676, "num_tokens": 438342006.0, "step": 11544 }, { "epoch": 1.4686426663274392, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.00020694732666015625, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.211716175079346, "learning_rate": 1e-05, "loss": 0.529, "mean_token_accuracy": 0.862749457359314, "num_tokens": 438375649.0, "step": 11545 }, { "epoch": 1.4687698766060298, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.00020694732666015625, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.737269878387451, "learning_rate": 1e-05, "loss": 0.49, "mean_token_accuracy": 0.8763371109962463, "num_tokens": 438417566.0, "step": 11546 }, { "epoch": 1.4688970868846203, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.00020694732666015625, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.191431045532227, "learning_rate": 1e-05, "loss": 0.4683, "mean_token_accuracy": 0.8760416507720947, "num_tokens": 438452219.0, "step": 11547 }, { "epoch": 1.4690242971632108, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.00020694732666015625, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.465762615203857, "learning_rate": 1e-05, "loss": 0.4543, "mean_token_accuracy": 0.8818798065185547, "num_tokens": 438490503.0, "step": 11548 }, { "epoch": 1.4691515074418013, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.00020694732666015625, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.118316173553467, "learning_rate": 1e-05, "loss": 0.4523, "mean_token_accuracy": 0.8823610544204712, "num_tokens": 438521792.0, "step": 11549 }, { "epoch": 1.4692787177203919, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.00020694732666015625, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.306077003479004, "learning_rate": 1e-05, "loss": 0.4748, "mean_token_accuracy": 0.8771064281463623, "num_tokens": 438557408.0, "step": 11550 }, { "epoch": 1.4694059279989824, "ewc_loss": 0.109375, "ewc_loss_diag": 0.00020694732666015625, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 5.988061428070068, "learning_rate": 1e-05, "loss": 0.4448, "mean_token_accuracy": 0.8826976418495178, "num_tokens": 438601424.0, "step": 11551 }, { "epoch": 1.4695331382775727, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.00020694732666015625, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.2691216468811035, "learning_rate": 1e-05, "loss": 0.4817, "mean_token_accuracy": 0.8755869269371033, "num_tokens": 438638772.0, "step": 11552 }, { "epoch": 1.4696603485561632, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.00020694732666015625, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.876391410827637, "learning_rate": 1e-05, "loss": 0.4922, "mean_token_accuracy": 0.8663102388381958, "num_tokens": 438676906.0, "step": 11553 }, { "epoch": 1.4697875588347538, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.00020694732666015625, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.320455074310303, "learning_rate": 1e-05, "loss": 0.5358, "mean_token_accuracy": 0.8545855283737183, "num_tokens": 438715790.0, "step": 11554 }, { "epoch": 1.4699147691133443, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.00020694732666015625, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.738063335418701, "learning_rate": 1e-05, "loss": 0.4664, "mean_token_accuracy": 0.8751075267791748, "num_tokens": 438757739.0, "step": 11555 }, { "epoch": 1.4700419793919348, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.00020694732666015625, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.464443683624268, "learning_rate": 1e-05, "loss": 0.4902, "mean_token_accuracy": 0.8709439039230347, "num_tokens": 438797336.0, "step": 11556 }, { "epoch": 1.4701691896705253, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.00020694732666015625, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.598816871643066, "learning_rate": 1e-05, "loss": 0.4869, "mean_token_accuracy": 0.866507887840271, "num_tokens": 438833452.0, "step": 11557 }, { "epoch": 1.4702963999491159, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.00020694732666015625, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.716027736663818, "learning_rate": 1e-05, "loss": 0.4922, "mean_token_accuracy": 0.8719791173934937, "num_tokens": 438874346.0, "step": 11558 }, { "epoch": 1.4704236102277064, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.00020694732666015625, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.540383338928223, "learning_rate": 1e-05, "loss": 0.5125, "mean_token_accuracy": 0.8578501343727112, "num_tokens": 438908888.0, "step": 11559 }, { "epoch": 1.470550820506297, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.00020694732666015625, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 7.0265583992004395, "learning_rate": 1e-05, "loss": 0.4995, "mean_token_accuracy": 0.8743598461151123, "num_tokens": 438953127.0, "step": 11560 }, { "epoch": 1.4706780307848875, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.00020694732666015625, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.659313201904297, "learning_rate": 1e-05, "loss": 0.487, "mean_token_accuracy": 0.8660902976989746, "num_tokens": 438986485.0, "step": 11561 }, { "epoch": 1.470805241063478, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.00020694732666015625, "ewc_loss_parallel": 0.00011491775512695312, "grad_norm": 7.218013763427734, "learning_rate": 1e-05, "loss": 0.5175, "mean_token_accuracy": 0.8683148622512817, "num_tokens": 439030958.0, "step": 11562 }, { "epoch": 1.4709324513420685, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.00020694732666015625, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 5.975650787353516, "learning_rate": 1e-05, "loss": 0.5073, "mean_token_accuracy": 0.8622178435325623, "num_tokens": 439069139.0, "step": 11563 }, { "epoch": 1.471059661620659, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.00020694732666015625, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 7.081277847290039, "learning_rate": 1e-05, "loss": 0.4876, "mean_token_accuracy": 0.8755983114242554, "num_tokens": 439106526.0, "step": 11564 }, { "epoch": 1.4711868718992496, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.00020694732666015625, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.228202819824219, "learning_rate": 1e-05, "loss": 0.4901, "mean_token_accuracy": 0.8715264201164246, "num_tokens": 439146037.0, "step": 11565 }, { "epoch": 1.47131408217784, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.00020694732666015625, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.671385765075684, "learning_rate": 1e-05, "loss": 0.4709, "mean_token_accuracy": 0.8800144791603088, "num_tokens": 439177239.0, "step": 11566 }, { "epoch": 1.4714412924564304, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.00020694732666015625, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.341851711273193, "learning_rate": 1e-05, "loss": 0.5031, "mean_token_accuracy": 0.8715728521347046, "num_tokens": 439211009.0, "step": 11567 }, { "epoch": 1.471568502735021, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.00020694732666015625, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.171143531799316, "learning_rate": 1e-05, "loss": 0.4749, "mean_token_accuracy": 0.8718575239181519, "num_tokens": 439250656.0, "step": 11568 }, { "epoch": 1.4716957130136115, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.00020694732666015625, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.472520351409912, "learning_rate": 1e-05, "loss": 0.5462, "mean_token_accuracy": 0.8522168397903442, "num_tokens": 439285103.0, "step": 11569 }, { "epoch": 1.471822923292202, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.00020694732666015625, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.746803283691406, "learning_rate": 1e-05, "loss": 0.5253, "mean_token_accuracy": 0.8600434064865112, "num_tokens": 439325104.0, "step": 11570 }, { "epoch": 1.4719501335707925, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.00020694732666015625, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.7250590324401855, "learning_rate": 1e-05, "loss": 0.5358, "mean_token_accuracy": 0.8568407297134399, "num_tokens": 439357760.0, "step": 11571 }, { "epoch": 1.472077343849383, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.00020694732666015625, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.53555154800415, "learning_rate": 1e-05, "loss": 0.4938, "mean_token_accuracy": 0.865576982498169, "num_tokens": 439397823.0, "step": 11572 }, { "epoch": 1.4722045541279736, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.00020694732666015625, "ewc_loss_parallel": 0.00011110305786132812, "grad_norm": 7.068120002746582, "learning_rate": 1e-05, "loss": 0.5273, "mean_token_accuracy": 0.8615483045578003, "num_tokens": 439435989.0, "step": 11573 }, { "epoch": 1.472331764406564, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.00020694732666015625, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.609401226043701, "learning_rate": 1e-05, "loss": 0.4572, "mean_token_accuracy": 0.8785666227340698, "num_tokens": 439474272.0, "step": 11574 }, { "epoch": 1.4724589746851546, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.00020694732666015625, "ewc_loss_parallel": 0.000118255615234375, "grad_norm": 7.357619762420654, "learning_rate": 1e-05, "loss": 0.4988, "mean_token_accuracy": 0.8790692090988159, "num_tokens": 439512729.0, "step": 11575 }, { "epoch": 1.472586184963745, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.00020694732666015625, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 5.981612205505371, "learning_rate": 1e-05, "loss": 0.5167, "mean_token_accuracy": 0.8596720099449158, "num_tokens": 439550368.0, "step": 11576 }, { "epoch": 1.4727133952423355, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.00020694732666015625, "ewc_loss_parallel": 0.00011777877807617188, "grad_norm": 7.294543266296387, "learning_rate": 1e-05, "loss": 0.5304, "mean_token_accuracy": 0.8691074848175049, "num_tokens": 439591244.0, "step": 11577 }, { "epoch": 1.472840605520926, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.00020694732666015625, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.266173839569092, "learning_rate": 1e-05, "loss": 0.5104, "mean_token_accuracy": 0.8659931421279907, "num_tokens": 439632389.0, "step": 11578 }, { "epoch": 1.4729678157995165, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.00020694732666015625, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.978333473205566, "learning_rate": 1e-05, "loss": 0.5429, "mean_token_accuracy": 0.8562333583831787, "num_tokens": 439674543.0, "step": 11579 }, { "epoch": 1.473095026078107, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.00020694732666015625, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.306103706359863, "learning_rate": 1e-05, "loss": 0.5104, "mean_token_accuracy": 0.8678404092788696, "num_tokens": 439709009.0, "step": 11580 }, { "epoch": 1.4732222363566976, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.00020694732666015625, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.652474880218506, "learning_rate": 1e-05, "loss": 0.4901, "mean_token_accuracy": 0.872859001159668, "num_tokens": 439754154.0, "step": 11581 }, { "epoch": 1.473349446635288, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.00020694732666015625, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.202564239501953, "learning_rate": 1e-05, "loss": 0.441, "mean_token_accuracy": 0.8888895511627197, "num_tokens": 439794986.0, "step": 11582 }, { "epoch": 1.4734766569138786, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.4287848472595215, "learning_rate": 1e-05, "loss": 0.5073, "mean_token_accuracy": 0.8627246618270874, "num_tokens": 439833402.0, "step": 11583 }, { "epoch": 1.4736038671924692, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.080920696258545, "learning_rate": 1e-05, "loss": 0.4546, "mean_token_accuracy": 0.8823226690292358, "num_tokens": 439878670.0, "step": 11584 }, { "epoch": 1.4737310774710597, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.330960750579834, "learning_rate": 1e-05, "loss": 0.4737, "mean_token_accuracy": 0.8783705234527588, "num_tokens": 439916505.0, "step": 11585 }, { "epoch": 1.4738582877496502, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.932370185852051, "learning_rate": 1e-05, "loss": 0.4811, "mean_token_accuracy": 0.8712798357009888, "num_tokens": 439955833.0, "step": 11586 }, { "epoch": 1.4739854980282407, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.355229377746582, "learning_rate": 1e-05, "loss": 0.4507, "mean_token_accuracy": 0.8822906017303467, "num_tokens": 439994105.0, "step": 11587 }, { "epoch": 1.4741127083068313, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.785557746887207, "learning_rate": 1e-05, "loss": 0.4616, "mean_token_accuracy": 0.8761610984802246, "num_tokens": 440030298.0, "step": 11588 }, { "epoch": 1.4742399185854218, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.4819111824035645, "learning_rate": 1e-05, "loss": 0.4703, "mean_token_accuracy": 0.8753305673599243, "num_tokens": 440064792.0, "step": 11589 }, { "epoch": 1.4743671288640123, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.654245376586914, "learning_rate": 1e-05, "loss": 0.4879, "mean_token_accuracy": 0.8694543838500977, "num_tokens": 440097886.0, "step": 11590 }, { "epoch": 1.4744943391426026, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.70298433303833, "learning_rate": 1e-05, "loss": 0.4908, "mean_token_accuracy": 0.8746502995491028, "num_tokens": 440133488.0, "step": 11591 }, { "epoch": 1.4746215494211932, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.587704658508301, "learning_rate": 1e-05, "loss": 0.4633, "mean_token_accuracy": 0.8756119012832642, "num_tokens": 440169448.0, "step": 11592 }, { "epoch": 1.4747487596997837, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.981844425201416, "learning_rate": 1e-05, "loss": 0.4673, "mean_token_accuracy": 0.8788576126098633, "num_tokens": 440204755.0, "step": 11593 }, { "epoch": 1.4748759699783742, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.692164897918701, "learning_rate": 1e-05, "loss": 0.4994, "mean_token_accuracy": 0.8664002418518066, "num_tokens": 440239129.0, "step": 11594 }, { "epoch": 1.4750031802569648, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 0.00011301040649414062, "grad_norm": 7.152890205383301, "learning_rate": 1e-05, "loss": 0.5189, "mean_token_accuracy": 0.8708940744400024, "num_tokens": 440280962.0, "step": 11595 }, { "epoch": 1.4751303905355553, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 5.9788689613342285, "learning_rate": 1e-05, "loss": 0.4727, "mean_token_accuracy": 0.8743196725845337, "num_tokens": 440315900.0, "step": 11596 }, { "epoch": 1.4752576008141458, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.991818904876709, "learning_rate": 1e-05, "loss": 0.479, "mean_token_accuracy": 0.877686619758606, "num_tokens": 440356699.0, "step": 11597 }, { "epoch": 1.4753848110927363, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.216182231903076, "learning_rate": 1e-05, "loss": 0.4735, "mean_token_accuracy": 0.8765828609466553, "num_tokens": 440398655.0, "step": 11598 }, { "epoch": 1.4755120213713269, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.578554153442383, "learning_rate": 1e-05, "loss": 0.5344, "mean_token_accuracy": 0.8611840009689331, "num_tokens": 440436204.0, "step": 11599 }, { "epoch": 1.4756392316499174, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.333758354187012, "learning_rate": 1e-05, "loss": 0.4782, "mean_token_accuracy": 0.8724655508995056, "num_tokens": 440478363.0, "step": 11600 }, { "epoch": 1.4757664419285077, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 6.081449031829834, "learning_rate": 1e-05, "loss": 0.4448, "mean_token_accuracy": 0.8825651407241821, "num_tokens": 440515454.0, "step": 11601 }, { "epoch": 1.4758936522070982, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.490921974182129, "learning_rate": 1e-05, "loss": 0.5082, "mean_token_accuracy": 0.8686641454696655, "num_tokens": 440553526.0, "step": 11602 }, { "epoch": 1.4760208624856888, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.653154373168945, "learning_rate": 1e-05, "loss": 0.5061, "mean_token_accuracy": 0.859836220741272, "num_tokens": 440591446.0, "step": 11603 }, { "epoch": 1.4761480727642793, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.8260817527771, "learning_rate": 1e-05, "loss": 0.4716, "mean_token_accuracy": 0.8784083724021912, "num_tokens": 440630505.0, "step": 11604 }, { "epoch": 1.4762752830428698, "ewc_loss": 0.0986328125, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 7.82012939453125e-05, "grad_norm": 5.449582099914551, "learning_rate": 1e-05, "loss": 0.4256, "mean_token_accuracy": 0.8830676078796387, "num_tokens": 440669318.0, "step": 11605 }, { "epoch": 1.4764024933214603, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 0.0001163482666015625, "grad_norm": 7.284418106079102, "learning_rate": 1e-05, "loss": 0.4836, "mean_token_accuracy": 0.878300666809082, "num_tokens": 440711218.0, "step": 11606 }, { "epoch": 1.4765297036000509, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.661147117614746, "learning_rate": 1e-05, "loss": 0.4727, "mean_token_accuracy": 0.8705306649208069, "num_tokens": 440754581.0, "step": 11607 }, { "epoch": 1.4766569138786414, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.59028434753418, "learning_rate": 1e-05, "loss": 0.519, "mean_token_accuracy": 0.8699219226837158, "num_tokens": 440795714.0, "step": 11608 }, { "epoch": 1.476784124157232, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.227567672729492, "learning_rate": 1e-05, "loss": 0.4424, "mean_token_accuracy": 0.8854537010192871, "num_tokens": 440827423.0, "step": 11609 }, { "epoch": 1.4769113344358225, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 0.0001163482666015625, "grad_norm": 7.3489508628845215, "learning_rate": 1e-05, "loss": 0.5421, "mean_token_accuracy": 0.8603371381759644, "num_tokens": 440868857.0, "step": 11610 }, { "epoch": 1.477038544714413, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.495474338531494, "learning_rate": 1e-05, "loss": 0.5484, "mean_token_accuracy": 0.8557652235031128, "num_tokens": 440904788.0, "step": 11611 }, { "epoch": 1.4771657549930035, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.8950276374816895, "learning_rate": 1e-05, "loss": 0.5002, "mean_token_accuracy": 0.8680815696716309, "num_tokens": 440949013.0, "step": 11612 }, { "epoch": 1.477292965271594, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.420564651489258, "learning_rate": 1e-05, "loss": 0.5383, "mean_token_accuracy": 0.8583390116691589, "num_tokens": 440987392.0, "step": 11613 }, { "epoch": 1.4774201755501846, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.5351362228393555, "learning_rate": 1e-05, "loss": 0.4063, "mean_token_accuracy": 0.8968415260314941, "num_tokens": 441024917.0, "step": 11614 }, { "epoch": 1.477547385828775, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.264009952545166, "learning_rate": 1e-05, "loss": 0.4742, "mean_token_accuracy": 0.877291738986969, "num_tokens": 441066355.0, "step": 11615 }, { "epoch": 1.4776745961073654, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.346814155578613, "learning_rate": 1e-05, "loss": 0.5033, "mean_token_accuracy": 0.8634262084960938, "num_tokens": 441100192.0, "step": 11616 }, { "epoch": 1.477801806385956, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.097048282623291, "learning_rate": 1e-05, "loss": 0.5189, "mean_token_accuracy": 0.8597440719604492, "num_tokens": 441144627.0, "step": 11617 }, { "epoch": 1.4779290166645465, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.298398494720459, "learning_rate": 1e-05, "loss": 0.4694, "mean_token_accuracy": 0.8741702437400818, "num_tokens": 441181656.0, "step": 11618 }, { "epoch": 1.478056226943137, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.926774024963379, "learning_rate": 1e-05, "loss": 0.5071, "mean_token_accuracy": 0.8649992346763611, "num_tokens": 441215476.0, "step": 11619 }, { "epoch": 1.4781834372217275, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.361435413360596, "learning_rate": 1e-05, "loss": 0.5089, "mean_token_accuracy": 0.8647414445877075, "num_tokens": 441262919.0, "step": 11620 }, { "epoch": 1.478310647500318, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.729833602905273, "learning_rate": 1e-05, "loss": 0.488, "mean_token_accuracy": 0.8675747513771057, "num_tokens": 441302602.0, "step": 11621 }, { "epoch": 1.4784378577789086, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.566953659057617, "learning_rate": 1e-05, "loss": 0.4926, "mean_token_accuracy": 0.8711020946502686, "num_tokens": 441335432.0, "step": 11622 }, { "epoch": 1.478565068057499, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.571937561035156, "learning_rate": 1e-05, "loss": 0.5127, "mean_token_accuracy": 0.8609501123428345, "num_tokens": 441377849.0, "step": 11623 }, { "epoch": 1.4786922783360896, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.90469217300415, "learning_rate": 1e-05, "loss": 0.5075, "mean_token_accuracy": 0.8717926740646362, "num_tokens": 441413225.0, "step": 11624 }, { "epoch": 1.47881948861468, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.547502517700195, "learning_rate": 1e-05, "loss": 0.4643, "mean_token_accuracy": 0.8755112886428833, "num_tokens": 441448041.0, "step": 11625 }, { "epoch": 1.4789466988932705, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 0.0001163482666015625, "grad_norm": 7.272511005401611, "learning_rate": 1e-05, "loss": 0.4894, "mean_token_accuracy": 0.8825042843818665, "num_tokens": 441483919.0, "step": 11626 }, { "epoch": 1.479073909171861, "ewc_loss": 0.109375, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 5.823046684265137, "learning_rate": 1e-05, "loss": 0.4819, "mean_token_accuracy": 0.8666688203811646, "num_tokens": 441518313.0, "step": 11627 }, { "epoch": 1.4792011194504515, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 0.00011777877807617188, "grad_norm": 7.383993148803711, "learning_rate": 1e-05, "loss": 0.5337, "mean_token_accuracy": 0.8660476207733154, "num_tokens": 441557248.0, "step": 11628 }, { "epoch": 1.479328329729042, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.248656272888184, "learning_rate": 1e-05, "loss": 0.5994, "mean_token_accuracy": 0.8397777676582336, "num_tokens": 441599759.0, "step": 11629 }, { "epoch": 1.4794555400076326, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 7.055063724517822, "learning_rate": 1e-05, "loss": 0.5118, "mean_token_accuracy": 0.8647768497467041, "num_tokens": 441631643.0, "step": 11630 }, { "epoch": 1.479582750286223, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.443711757659912, "learning_rate": 1e-05, "loss": 0.5286, "mean_token_accuracy": 0.8598247766494751, "num_tokens": 441669804.0, "step": 11631 }, { "epoch": 1.4797099605648136, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.533908367156982, "learning_rate": 1e-05, "loss": 0.5322, "mean_token_accuracy": 0.8578773736953735, "num_tokens": 441708661.0, "step": 11632 }, { "epoch": 1.4798371708434042, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.466959476470947, "learning_rate": 1e-05, "loss": 0.4938, "mean_token_accuracy": 0.8701289892196655, "num_tokens": 441745510.0, "step": 11633 }, { "epoch": 1.4799643811219947, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 6.0541605949401855, "learning_rate": 1e-05, "loss": 0.497, "mean_token_accuracy": 0.8696451783180237, "num_tokens": 441781482.0, "step": 11634 }, { "epoch": 1.4800915914005852, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.563770294189453, "learning_rate": 1e-05, "loss": 0.4655, "mean_token_accuracy": 0.8796724677085876, "num_tokens": 441814299.0, "step": 11635 }, { "epoch": 1.4802188016791757, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.716671943664551, "learning_rate": 1e-05, "loss": 0.5243, "mean_token_accuracy": 0.8542215824127197, "num_tokens": 441859742.0, "step": 11636 }, { "epoch": 1.4803460119577663, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.794894695281982, "learning_rate": 1e-05, "loss": 0.4769, "mean_token_accuracy": 0.8816379308700562, "num_tokens": 441896563.0, "step": 11637 }, { "epoch": 1.4804732222363568, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.579629421234131, "learning_rate": 1e-05, "loss": 0.4526, "mean_token_accuracy": 0.8772522211074829, "num_tokens": 441935380.0, "step": 11638 }, { "epoch": 1.4806004325149473, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 7.126819133758545, "learning_rate": 1e-05, "loss": 0.5433, "mean_token_accuracy": 0.8599481582641602, "num_tokens": 441977077.0, "step": 11639 }, { "epoch": 1.4807276427935376, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.703426837921143, "learning_rate": 1e-05, "loss": 0.4822, "mean_token_accuracy": 0.8721364140510559, "num_tokens": 442020538.0, "step": 11640 }, { "epoch": 1.4808548530721282, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 0.00011682510375976562, "grad_norm": 7.347217082977295, "learning_rate": 1e-05, "loss": 0.513, "mean_token_accuracy": 0.8721917271614075, "num_tokens": 442057093.0, "step": 11641 }, { "epoch": 1.4809820633507187, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.054360389709473, "learning_rate": 1e-05, "loss": 0.5089, "mean_token_accuracy": 0.8643409013748169, "num_tokens": 442092364.0, "step": 11642 }, { "epoch": 1.4811092736293092, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 0.00011301040649414062, "grad_norm": 7.220619201660156, "learning_rate": 1e-05, "loss": 0.5259, "mean_token_accuracy": 0.8633089661598206, "num_tokens": 442134181.0, "step": 11643 }, { "epoch": 1.4812364839078997, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.277549743652344, "learning_rate": 1e-05, "loss": 0.5448, "mean_token_accuracy": 0.8554067015647888, "num_tokens": 442174987.0, "step": 11644 }, { "epoch": 1.4813636941864903, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.914739608764648, "learning_rate": 1e-05, "loss": 0.4742, "mean_token_accuracy": 0.8751720190048218, "num_tokens": 442204214.0, "step": 11645 }, { "epoch": 1.4814909044650808, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.290470600128174, "learning_rate": 1e-05, "loss": 0.4673, "mean_token_accuracy": 0.8792266845703125, "num_tokens": 442243227.0, "step": 11646 }, { "epoch": 1.4816181147436713, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.586181163787842, "learning_rate": 1e-05, "loss": 0.5017, "mean_token_accuracy": 0.8709647059440613, "num_tokens": 442282107.0, "step": 11647 }, { "epoch": 1.4817453250222619, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.218514442443848, "learning_rate": 1e-05, "loss": 0.4785, "mean_token_accuracy": 0.8771674633026123, "num_tokens": 442320445.0, "step": 11648 }, { "epoch": 1.4818725353008524, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.366628646850586, "learning_rate": 1e-05, "loss": 0.5072, "mean_token_accuracy": 0.8645921349525452, "num_tokens": 442358626.0, "step": 11649 }, { "epoch": 1.4819997455794427, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.085969924926758, "learning_rate": 1e-05, "loss": 0.4928, "mean_token_accuracy": 0.8682808876037598, "num_tokens": 442405095.0, "step": 11650 }, { "epoch": 1.4821269558580332, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.287057876586914, "learning_rate": 1e-05, "loss": 0.4943, "mean_token_accuracy": 0.868027925491333, "num_tokens": 442442131.0, "step": 11651 }, { "epoch": 1.4822541661366238, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.953640937805176, "learning_rate": 1e-05, "loss": 0.543, "mean_token_accuracy": 0.8555452823638916, "num_tokens": 442479660.0, "step": 11652 }, { "epoch": 1.4823813764152143, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.308509349822998, "learning_rate": 1e-05, "loss": 0.5053, "mean_token_accuracy": 0.8650041818618774, "num_tokens": 442523659.0, "step": 11653 }, { "epoch": 1.4825085866938048, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.794668197631836, "learning_rate": 1e-05, "loss": 0.4951, "mean_token_accuracy": 0.863466739654541, "num_tokens": 442564445.0, "step": 11654 }, { "epoch": 1.4826357969723953, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.448044300079346, "learning_rate": 1e-05, "loss": 0.4898, "mean_token_accuracy": 0.8712327480316162, "num_tokens": 442597662.0, "step": 11655 }, { "epoch": 1.4827630072509859, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.64218282699585, "learning_rate": 1e-05, "loss": 0.4975, "mean_token_accuracy": 0.8631293177604675, "num_tokens": 442633365.0, "step": 11656 }, { "epoch": 1.4828902175295764, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.697240829467773, "learning_rate": 1e-05, "loss": 0.501, "mean_token_accuracy": 0.8678604364395142, "num_tokens": 442668389.0, "step": 11657 }, { "epoch": 1.483017427808167, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.565356254577637, "learning_rate": 1e-05, "loss": 0.4723, "mean_token_accuracy": 0.8720959424972534, "num_tokens": 442695089.0, "step": 11658 }, { "epoch": 1.4831446380867574, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 7.006563186645508, "learning_rate": 1e-05, "loss": 0.4981, "mean_token_accuracy": 0.8699564337730408, "num_tokens": 442738602.0, "step": 11659 }, { "epoch": 1.483271848365348, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.636904716491699, "learning_rate": 1e-05, "loss": 0.4998, "mean_token_accuracy": 0.8656941652297974, "num_tokens": 442777400.0, "step": 11660 }, { "epoch": 1.4833990586439385, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 0.00011539459228515625, "grad_norm": 7.258367538452148, "learning_rate": 1e-05, "loss": 0.4991, "mean_token_accuracy": 0.8723375201225281, "num_tokens": 442815502.0, "step": 11661 }, { "epoch": 1.483526268922529, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 5.96575927734375, "learning_rate": 1e-05, "loss": 0.5259, "mean_token_accuracy": 0.8600805997848511, "num_tokens": 442858418.0, "step": 11662 }, { "epoch": 1.4836534792011196, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 0.00011157989501953125, "grad_norm": 7.158954620361328, "learning_rate": 1e-05, "loss": 0.4663, "mean_token_accuracy": 0.8858172297477722, "num_tokens": 442896073.0, "step": 11663 }, { "epoch": 1.48378068947971, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.269644260406494, "learning_rate": 1e-05, "loss": 0.5003, "mean_token_accuracy": 0.866737961769104, "num_tokens": 442940719.0, "step": 11664 }, { "epoch": 1.4839078997583004, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.741954326629639, "learning_rate": 1e-05, "loss": 0.4643, "mean_token_accuracy": 0.8799169659614563, "num_tokens": 442977724.0, "step": 11665 }, { "epoch": 1.484035110036891, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.385484218597412, "learning_rate": 1e-05, "loss": 0.4704, "mean_token_accuracy": 0.8785348534584045, "num_tokens": 443017754.0, "step": 11666 }, { "epoch": 1.4841623203154815, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.230705261230469, "learning_rate": 1e-05, "loss": 0.4879, "mean_token_accuracy": 0.8707659244537354, "num_tokens": 443048661.0, "step": 11667 }, { "epoch": 1.484289530594072, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.485961437225342, "learning_rate": 1e-05, "loss": 0.5051, "mean_token_accuracy": 0.8683782815933228, "num_tokens": 443094445.0, "step": 11668 }, { "epoch": 1.4844167408726625, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.792201995849609, "learning_rate": 1e-05, "loss": 0.4483, "mean_token_accuracy": 0.8819968700408936, "num_tokens": 443134468.0, "step": 11669 }, { "epoch": 1.484543951151253, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.726705551147461, "learning_rate": 1e-05, "loss": 0.4821, "mean_token_accuracy": 0.8733970522880554, "num_tokens": 443174458.0, "step": 11670 }, { "epoch": 1.4846711614298436, "ewc_loss": 0.099609375, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.55418586730957, "learning_rate": 1e-05, "loss": 0.4548, "mean_token_accuracy": 0.8773739337921143, "num_tokens": 443212421.0, "step": 11671 }, { "epoch": 1.484798371708434, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 0.00010967254638671875, "grad_norm": 7.098718643188477, "learning_rate": 1e-05, "loss": 0.5872, "mean_token_accuracy": 0.844075083732605, "num_tokens": 443252083.0, "step": 11672 }, { "epoch": 1.4849255819870246, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.625125885009766, "learning_rate": 1e-05, "loss": 0.5022, "mean_token_accuracy": 0.8613029718399048, "num_tokens": 443286776.0, "step": 11673 }, { "epoch": 1.485052792265615, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 0.00011873245239257812, "grad_norm": 7.390036582946777, "learning_rate": 1e-05, "loss": 0.5028, "mean_token_accuracy": 0.8767246007919312, "num_tokens": 443325734.0, "step": 11674 }, { "epoch": 1.4851800025442055, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 5.996696949005127, "learning_rate": 1e-05, "loss": 0.4912, "mean_token_accuracy": 0.8684809803962708, "num_tokens": 443362637.0, "step": 11675 }, { "epoch": 1.485307212822796, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 0.0001163482666015625, "grad_norm": 7.318844795227051, "learning_rate": 1e-05, "loss": 0.4891, "mean_token_accuracy": 0.8792836666107178, "num_tokens": 443401629.0, "step": 11676 }, { "epoch": 1.4854344231013865, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.303144454956055, "learning_rate": 1e-05, "loss": 0.449, "mean_token_accuracy": 0.8850456476211548, "num_tokens": 443437619.0, "step": 11677 }, { "epoch": 1.485561633379977, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.997636318206787, "learning_rate": 1e-05, "loss": 0.5508, "mean_token_accuracy": 0.8549772500991821, "num_tokens": 443472175.0, "step": 11678 }, { "epoch": 1.4856888436585676, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.340112209320068, "learning_rate": 1e-05, "loss": 0.5177, "mean_token_accuracy": 0.8636826276779175, "num_tokens": 443506904.0, "step": 11679 }, { "epoch": 1.485816053937158, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.638139724731445, "learning_rate": 1e-05, "loss": 0.5388, "mean_token_accuracy": 0.8564108610153198, "num_tokens": 443548141.0, "step": 11680 }, { "epoch": 1.4859432642157486, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.248000621795654, "learning_rate": 1e-05, "loss": 0.4976, "mean_token_accuracy": 0.8687960505485535, "num_tokens": 443586609.0, "step": 11681 }, { "epoch": 1.4860704744943392, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.395745754241943, "learning_rate": 1e-05, "loss": 0.4906, "mean_token_accuracy": 0.872681736946106, "num_tokens": 443628792.0, "step": 11682 }, { "epoch": 1.4861976847729297, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.116302967071533, "learning_rate": 1e-05, "loss": 0.5549, "mean_token_accuracy": 0.8513646721839905, "num_tokens": 443671336.0, "step": 11683 }, { "epoch": 1.4863248950515202, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.310041427612305, "learning_rate": 1e-05, "loss": 0.4867, "mean_token_accuracy": 0.8684900999069214, "num_tokens": 443708987.0, "step": 11684 }, { "epoch": 1.4864521053301107, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.948322772979736, "learning_rate": 1e-05, "loss": 0.4841, "mean_token_accuracy": 0.8716917037963867, "num_tokens": 443749968.0, "step": 11685 }, { "epoch": 1.4865793156087013, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.343366622924805, "learning_rate": 1e-05, "loss": 0.4912, "mean_token_accuracy": 0.87249755859375, "num_tokens": 443786287.0, "step": 11686 }, { "epoch": 1.4867065258872918, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.790640354156494, "learning_rate": 1e-05, "loss": 0.4407, "mean_token_accuracy": 0.8825621604919434, "num_tokens": 443820729.0, "step": 11687 }, { "epoch": 1.4868337361658823, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.472140789031982, "learning_rate": 1e-05, "loss": 0.4476, "mean_token_accuracy": 0.8820186853408813, "num_tokens": 443859401.0, "step": 11688 }, { "epoch": 1.4869609464444726, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.636997699737549, "learning_rate": 1e-05, "loss": 0.444, "mean_token_accuracy": 0.8856520056724548, "num_tokens": 443895697.0, "step": 11689 }, { "epoch": 1.4870881567230632, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.724332332611084, "learning_rate": 1e-05, "loss": 0.4811, "mean_token_accuracy": 0.8751469850540161, "num_tokens": 443935229.0, "step": 11690 }, { "epoch": 1.4872153670016537, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.561852931976318, "learning_rate": 1e-05, "loss": 0.5044, "mean_token_accuracy": 0.8681407570838928, "num_tokens": 443967544.0, "step": 11691 }, { "epoch": 1.4873425772802442, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 7.047330379486084, "learning_rate": 1e-05, "loss": 0.5207, "mean_token_accuracy": 0.8667279481887817, "num_tokens": 444006973.0, "step": 11692 }, { "epoch": 1.4874697875588347, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.678252220153809, "learning_rate": 1e-05, "loss": 0.4544, "mean_token_accuracy": 0.8777847290039062, "num_tokens": 444041143.0, "step": 11693 }, { "epoch": 1.4875969978374253, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 0.00011730194091796875, "grad_norm": 7.254892349243164, "learning_rate": 1e-05, "loss": 0.5464, "mean_token_accuracy": 0.8614221215248108, "num_tokens": 444086279.0, "step": 11694 }, { "epoch": 1.4877242081160158, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.016378402709961, "learning_rate": 1e-05, "loss": 0.4946, "mean_token_accuracy": 0.8709161281585693, "num_tokens": 444123559.0, "step": 11695 }, { "epoch": 1.4878514183946063, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 0.00011110305786132812, "grad_norm": 7.108777046203613, "learning_rate": 1e-05, "loss": 0.5446, "mean_token_accuracy": 0.8584340810775757, "num_tokens": 444159197.0, "step": 11696 }, { "epoch": 1.4879786286731969, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.292549133300781, "learning_rate": 1e-05, "loss": 0.508, "mean_token_accuracy": 0.8713109493255615, "num_tokens": 444190217.0, "step": 11697 }, { "epoch": 1.4881058389517874, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.650912761688232, "learning_rate": 1e-05, "loss": 0.4799, "mean_token_accuracy": 0.8846734762191772, "num_tokens": 444225241.0, "step": 11698 }, { "epoch": 1.4882330492303777, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.381584644317627, "learning_rate": 1e-05, "loss": 0.5432, "mean_token_accuracy": 0.8555926084518433, "num_tokens": 444263702.0, "step": 11699 }, { "epoch": 1.4883602595089682, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 6.137701988220215, "learning_rate": 1e-05, "loss": 0.48, "mean_token_accuracy": 0.8709961175918579, "num_tokens": 444303911.0, "step": 11700 }, { "epoch": 1.4884874697875587, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.4954023361206055, "learning_rate": 1e-05, "loss": 0.4948, "mean_token_accuracy": 0.8696151375770569, "num_tokens": 444344705.0, "step": 11701 }, { "epoch": 1.4886146800661493, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.724080562591553, "learning_rate": 1e-05, "loss": 0.4415, "mean_token_accuracy": 0.8832354545593262, "num_tokens": 444381598.0, "step": 11702 }, { "epoch": 1.4887418903447398, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.774290561676025, "learning_rate": 1e-05, "loss": 0.4867, "mean_token_accuracy": 0.8740928769111633, "num_tokens": 444413077.0, "step": 11703 }, { "epoch": 1.4888691006233303, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.521536350250244, "learning_rate": 1e-05, "loss": 0.4687, "mean_token_accuracy": 0.8729138374328613, "num_tokens": 444453028.0, "step": 11704 }, { "epoch": 1.4889963109019209, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 0.00011110305786132812, "grad_norm": 7.164615154266357, "learning_rate": 1e-05, "loss": 0.5183, "mean_token_accuracy": 0.8678460121154785, "num_tokens": 444486382.0, "step": 11705 }, { "epoch": 1.4891235211805114, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.652255535125732, "learning_rate": 1e-05, "loss": 0.4495, "mean_token_accuracy": 0.8782774209976196, "num_tokens": 444523299.0, "step": 11706 }, { "epoch": 1.489250731459102, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 0.00012111663818359375, "grad_norm": 7.445563793182373, "learning_rate": 1e-05, "loss": 0.51, "mean_token_accuracy": 0.8719372749328613, "num_tokens": 444561772.0, "step": 11707 }, { "epoch": 1.4893779417376924, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.075157165527344, "learning_rate": 1e-05, "loss": 0.5329, "mean_token_accuracy": 0.8569250106811523, "num_tokens": 444603056.0, "step": 11708 }, { "epoch": 1.489505152016283, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 0.00011587142944335938, "grad_norm": 7.30784273147583, "learning_rate": 1e-05, "loss": 0.4821, "mean_token_accuracy": 0.8818379640579224, "num_tokens": 444640469.0, "step": 11709 }, { "epoch": 1.4896323622948735, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.369228363037109, "learning_rate": 1e-05, "loss": 0.4965, "mean_token_accuracy": 0.8699002265930176, "num_tokens": 444680737.0, "step": 11710 }, { "epoch": 1.489759572573464, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.930273056030273, "learning_rate": 1e-05, "loss": 0.461, "mean_token_accuracy": 0.8833933472633362, "num_tokens": 444713828.0, "step": 11711 }, { "epoch": 1.4898867828520546, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.360832691192627, "learning_rate": 1e-05, "loss": 0.5239, "mean_token_accuracy": 0.8615393042564392, "num_tokens": 444755823.0, "step": 11712 }, { "epoch": 1.490013993130645, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.57426643371582, "learning_rate": 1e-05, "loss": 0.4825, "mean_token_accuracy": 0.872938871383667, "num_tokens": 444791558.0, "step": 11713 }, { "epoch": 1.4901412034092354, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.245082378387451, "learning_rate": 1e-05, "loss": 0.4817, "mean_token_accuracy": 0.875274121761322, "num_tokens": 444832037.0, "step": 11714 }, { "epoch": 1.490268413687826, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.350710391998291, "learning_rate": 1e-05, "loss": 0.5329, "mean_token_accuracy": 0.8586338758468628, "num_tokens": 444870092.0, "step": 11715 }, { "epoch": 1.4903956239664164, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.099578857421875, "learning_rate": 1e-05, "loss": 0.4964, "mean_token_accuracy": 0.8688783645629883, "num_tokens": 444909267.0, "step": 11716 }, { "epoch": 1.490522834245007, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.276419639587402, "learning_rate": 1e-05, "loss": 0.5022, "mean_token_accuracy": 0.8662487268447876, "num_tokens": 444946851.0, "step": 11717 }, { "epoch": 1.4906500445235975, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.9405837059021, "learning_rate": 1e-05, "loss": 0.4363, "mean_token_accuracy": 0.885934591293335, "num_tokens": 444986324.0, "step": 11718 }, { "epoch": 1.490777254802188, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.328716278076172, "learning_rate": 1e-05, "loss": 0.4569, "mean_token_accuracy": 0.8796576261520386, "num_tokens": 445020443.0, "step": 11719 }, { "epoch": 1.4909044650807786, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.759907245635986, "learning_rate": 1e-05, "loss": 0.5247, "mean_token_accuracy": 0.854466438293457, "num_tokens": 445057592.0, "step": 11720 }, { "epoch": 1.491031675359369, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.491224765777588, "learning_rate": 1e-05, "loss": 0.452, "mean_token_accuracy": 0.8839577436447144, "num_tokens": 445091513.0, "step": 11721 }, { "epoch": 1.4911588856379596, "ewc_loss": 0.099609375, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.608270168304443, "learning_rate": 1e-05, "loss": 0.5194, "mean_token_accuracy": 0.8556801080703735, "num_tokens": 445132579.0, "step": 11722 }, { "epoch": 1.49128609591655, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.787840366363525, "learning_rate": 1e-05, "loss": 0.5682, "mean_token_accuracy": 0.8491974472999573, "num_tokens": 445167336.0, "step": 11723 }, { "epoch": 1.4914133061951405, "ewc_loss": 0.099609375, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.532018661499023, "learning_rate": 1e-05, "loss": 0.518, "mean_token_accuracy": 0.8601354956626892, "num_tokens": 445208873.0, "step": 11724 }, { "epoch": 1.491540516473731, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 0.00011301040649414062, "grad_norm": 7.169971942901611, "learning_rate": 1e-05, "loss": 0.5261, "mean_token_accuracy": 0.8678239583969116, "num_tokens": 445245649.0, "step": 11725 }, { "epoch": 1.4916677267523215, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.709322929382324, "learning_rate": 1e-05, "loss": 0.4391, "mean_token_accuracy": 0.8835466504096985, "num_tokens": 445285890.0, "step": 11726 }, { "epoch": 1.491794937030912, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 0.00011920928955078125, "grad_norm": 7.387818813323975, "learning_rate": 1e-05, "loss": 0.5356, "mean_token_accuracy": 0.8615469932556152, "num_tokens": 445323408.0, "step": 11727 }, { "epoch": 1.4919221473095026, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.137813568115234, "learning_rate": 1e-05, "loss": 0.4683, "mean_token_accuracy": 0.8786146640777588, "num_tokens": 445361086.0, "step": 11728 }, { "epoch": 1.492049357588093, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 0.00011110305786132812, "grad_norm": 7.135624408721924, "learning_rate": 1e-05, "loss": 0.4927, "mean_token_accuracy": 0.8738085627555847, "num_tokens": 445404726.0, "step": 11729 }, { "epoch": 1.4921765678666836, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.408838748931885, "learning_rate": 1e-05, "loss": 0.4912, "mean_token_accuracy": 0.8715535402297974, "num_tokens": 445444190.0, "step": 11730 }, { "epoch": 1.4923037781452742, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.640669345855713, "learning_rate": 1e-05, "loss": 0.5225, "mean_token_accuracy": 0.861440896987915, "num_tokens": 445484395.0, "step": 11731 }, { "epoch": 1.4924309884238647, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.461616039276123, "learning_rate": 1e-05, "loss": 0.5051, "mean_token_accuracy": 0.8670682907104492, "num_tokens": 445518104.0, "step": 11732 }, { "epoch": 1.4925581987024552, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 6.132867336273193, "learning_rate": 1e-05, "loss": 0.5115, "mean_token_accuracy": 0.8608389496803284, "num_tokens": 445554761.0, "step": 11733 }, { "epoch": 1.4926854089810457, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.559825420379639, "learning_rate": 1e-05, "loss": 0.5192, "mean_token_accuracy": 0.8654661178588867, "num_tokens": 445595331.0, "step": 11734 }, { "epoch": 1.4928126192596363, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.759325981140137, "learning_rate": 1e-05, "loss": 0.4606, "mean_token_accuracy": 0.8761021494865417, "num_tokens": 445639280.0, "step": 11735 }, { "epoch": 1.4929398295382268, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.785576343536377, "learning_rate": 1e-05, "loss": 0.5027, "mean_token_accuracy": 0.8680084943771362, "num_tokens": 445675437.0, "step": 11736 }, { "epoch": 1.4930670398168173, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.598341941833496, "learning_rate": 1e-05, "loss": 0.4672, "mean_token_accuracy": 0.874218225479126, "num_tokens": 445712419.0, "step": 11737 }, { "epoch": 1.4931942500954076, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 0.00010967254638671875, "grad_norm": 7.097886085510254, "learning_rate": 1e-05, "loss": 0.5573, "mean_token_accuracy": 0.8519802093505859, "num_tokens": 445751465.0, "step": 11738 }, { "epoch": 1.4933214603739982, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.693828105926514, "learning_rate": 1e-05, "loss": 0.4488, "mean_token_accuracy": 0.8809556365013123, "num_tokens": 445788145.0, "step": 11739 }, { "epoch": 1.4934486706525887, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 0.00011682510375976562, "grad_norm": 7.316115379333496, "learning_rate": 1e-05, "loss": 0.5028, "mean_token_accuracy": 0.8751294612884521, "num_tokens": 445825954.0, "step": 11740 }, { "epoch": 1.4935758809311792, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.004210472106934, "learning_rate": 1e-05, "loss": 0.5316, "mean_token_accuracy": 0.8569289445877075, "num_tokens": 445866207.0, "step": 11741 }, { "epoch": 1.4937030912097697, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 0.00011348724365234375, "grad_norm": 7.230824947357178, "learning_rate": 1e-05, "loss": 0.5336, "mean_token_accuracy": 0.8586774468421936, "num_tokens": 445904679.0, "step": 11742 }, { "epoch": 1.4938303014883603, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.259664535522461, "learning_rate": 1e-05, "loss": 0.5222, "mean_token_accuracy": 0.8577074408531189, "num_tokens": 445936001.0, "step": 11743 }, { "epoch": 1.4939575117669508, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.931676864624023, "learning_rate": 1e-05, "loss": 0.5242, "mean_token_accuracy": 0.863888680934906, "num_tokens": 445969212.0, "step": 11744 }, { "epoch": 1.4940847220455413, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.269228458404541, "learning_rate": 1e-05, "loss": 0.5049, "mean_token_accuracy": 0.8682513236999512, "num_tokens": 446013455.0, "step": 11745 }, { "epoch": 1.4942119323241319, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.6454057693481445, "learning_rate": 1e-05, "loss": 0.5154, "mean_token_accuracy": 0.8646367788314819, "num_tokens": 446044152.0, "step": 11746 }, { "epoch": 1.4943391426027224, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.191004753112793, "learning_rate": 1e-05, "loss": 0.5455, "mean_token_accuracy": 0.8547917604446411, "num_tokens": 446087047.0, "step": 11747 }, { "epoch": 1.4944663528813127, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.44008731842041, "learning_rate": 1e-05, "loss": 0.4643, "mean_token_accuracy": 0.8772886991500854, "num_tokens": 446120095.0, "step": 11748 }, { "epoch": 1.4945935631599032, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.0675859451293945, "learning_rate": 1e-05, "loss": 0.4707, "mean_token_accuracy": 0.8752993941307068, "num_tokens": 446156827.0, "step": 11749 }, { "epoch": 1.4947207734384937, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.353917121887207, "learning_rate": 1e-05, "loss": 0.4963, "mean_token_accuracy": 0.8689380288124084, "num_tokens": 446197101.0, "step": 11750 }, { "epoch": 1.4948479837170843, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.932079792022705, "learning_rate": 1e-05, "loss": 0.4893, "mean_token_accuracy": 0.8685554265975952, "num_tokens": 446230442.0, "step": 11751 }, { "epoch": 1.4949751939956748, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.3875603675842285, "learning_rate": 1e-05, "loss": 0.4563, "mean_token_accuracy": 0.8798653483390808, "num_tokens": 446268395.0, "step": 11752 }, { "epoch": 1.4951024042742653, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.773858547210693, "learning_rate": 1e-05, "loss": 0.4681, "mean_token_accuracy": 0.8738595843315125, "num_tokens": 446307855.0, "step": 11753 }, { "epoch": 1.4952296145528559, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.5322442054748535, "learning_rate": 1e-05, "loss": 0.5137, "mean_token_accuracy": 0.8701022863388062, "num_tokens": 446348808.0, "step": 11754 }, { "epoch": 1.4953568248314464, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.658412933349609, "learning_rate": 1e-05, "loss": 0.4745, "mean_token_accuracy": 0.8745465874671936, "num_tokens": 446380580.0, "step": 11755 }, { "epoch": 1.495484035110037, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.775911808013916, "learning_rate": 1e-05, "loss": 0.4848, "mean_token_accuracy": 0.875593900680542, "num_tokens": 446414420.0, "step": 11756 }, { "epoch": 1.4956112453886274, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.607682704925537, "learning_rate": 1e-05, "loss": 0.4253, "mean_token_accuracy": 0.8866003751754761, "num_tokens": 446450220.0, "step": 11757 }, { "epoch": 1.495738455667218, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 7.054460048675537, "learning_rate": 1e-05, "loss": 0.4738, "mean_token_accuracy": 0.882489800453186, "num_tokens": 446483823.0, "step": 11758 }, { "epoch": 1.4958656659458085, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.7988362312316895, "learning_rate": 1e-05, "loss": 0.4811, "mean_token_accuracy": 0.8716313242912292, "num_tokens": 446522317.0, "step": 11759 }, { "epoch": 1.495992876224399, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 0.00011110305786132812, "grad_norm": 7.11395263671875, "learning_rate": 1e-05, "loss": 0.5377, "mean_token_accuracy": 0.8624392151832581, "num_tokens": 446564491.0, "step": 11760 }, { "epoch": 1.4961200865029896, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.03139591217041, "learning_rate": 1e-05, "loss": 0.4775, "mean_token_accuracy": 0.872309684753418, "num_tokens": 446605490.0, "step": 11761 }, { "epoch": 1.49624729678158, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.885464191436768, "learning_rate": 1e-05, "loss": 0.5393, "mean_token_accuracy": 0.8593498468399048, "num_tokens": 446647385.0, "step": 11762 }, { "epoch": 1.4963745070601704, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.229096412658691, "learning_rate": 1e-05, "loss": 0.5126, "mean_token_accuracy": 0.8639938831329346, "num_tokens": 446682895.0, "step": 11763 }, { "epoch": 1.496501717338761, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.451032638549805, "learning_rate": 1e-05, "loss": 0.5277, "mean_token_accuracy": 0.8591282963752747, "num_tokens": 446724724.0, "step": 11764 }, { "epoch": 1.4966289276173514, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.344310760498047, "learning_rate": 1e-05, "loss": 0.4983, "mean_token_accuracy": 0.87024986743927, "num_tokens": 446763433.0, "step": 11765 }, { "epoch": 1.496756137895942, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.968627452850342, "learning_rate": 1e-05, "loss": 0.5593, "mean_token_accuracy": 0.8447290062904358, "num_tokens": 446809275.0, "step": 11766 }, { "epoch": 1.4968833481745325, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.541639804840088, "learning_rate": 1e-05, "loss": 0.4501, "mean_token_accuracy": 0.8829506039619446, "num_tokens": 446840619.0, "step": 11767 }, { "epoch": 1.497010558453123, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.5867204666137695, "learning_rate": 1e-05, "loss": 0.518, "mean_token_accuracy": 0.8567688465118408, "num_tokens": 446879979.0, "step": 11768 }, { "epoch": 1.4971377687317136, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.901615142822266, "learning_rate": 1e-05, "loss": 0.4642, "mean_token_accuracy": 0.8797789812088013, "num_tokens": 446913729.0, "step": 11769 }, { "epoch": 1.497264979010304, "ewc_loss": 0.099609375, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.4608473777771, "learning_rate": 1e-05, "loss": 0.5413, "mean_token_accuracy": 0.8562282919883728, "num_tokens": 446951843.0, "step": 11770 }, { "epoch": 1.4973921892888946, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 0.00011777877807617188, "grad_norm": 7.35581111907959, "learning_rate": 1e-05, "loss": 0.5254, "mean_token_accuracy": 0.8659231662750244, "num_tokens": 446988091.0, "step": 11771 }, { "epoch": 1.497519399567485, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.0002079010009765625, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 5.738153457641602, "learning_rate": 1e-05, "loss": 0.512, "mean_token_accuracy": 0.8606110215187073, "num_tokens": 447023833.0, "step": 11772 }, { "epoch": 1.4976466098460754, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.573365211486816, "learning_rate": 1e-05, "loss": 0.4867, "mean_token_accuracy": 0.8752816319465637, "num_tokens": 447056232.0, "step": 11773 }, { "epoch": 1.497773820124666, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.267996788024902, "learning_rate": 1e-05, "loss": 0.4748, "mean_token_accuracy": 0.8809325695037842, "num_tokens": 447094524.0, "step": 11774 }, { "epoch": 1.4979010304032565, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 0.00011491775512695312, "grad_norm": 7.286982536315918, "learning_rate": 1e-05, "loss": 0.531, "mean_token_accuracy": 0.8641854524612427, "num_tokens": 447129076.0, "step": 11775 }, { "epoch": 1.498028240681847, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.474887847900391, "learning_rate": 1e-05, "loss": 0.5055, "mean_token_accuracy": 0.8681557178497314, "num_tokens": 447168651.0, "step": 11776 }, { "epoch": 1.4981554509604376, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.8599114418029785, "learning_rate": 1e-05, "loss": 0.5166, "mean_token_accuracy": 0.8689647912979126, "num_tokens": 447207491.0, "step": 11777 }, { "epoch": 1.498282661239028, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.407899379730225, "learning_rate": 1e-05, "loss": 0.4833, "mean_token_accuracy": 0.8770902752876282, "num_tokens": 447242955.0, "step": 11778 }, { "epoch": 1.4984098715176186, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.515194892883301, "learning_rate": 1e-05, "loss": 0.4395, "mean_token_accuracy": 0.888856053352356, "num_tokens": 447277634.0, "step": 11779 }, { "epoch": 1.4985370817962091, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.274894714355469, "learning_rate": 1e-05, "loss": 0.5156, "mean_token_accuracy": 0.8639195561408997, "num_tokens": 447315255.0, "step": 11780 }, { "epoch": 1.4986642920747997, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.304220676422119, "learning_rate": 1e-05, "loss": 0.4588, "mean_token_accuracy": 0.8816155195236206, "num_tokens": 447353801.0, "step": 11781 }, { "epoch": 1.4987915023533902, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.125299453735352, "learning_rate": 1e-05, "loss": 0.4865, "mean_token_accuracy": 0.8671594262123108, "num_tokens": 447388098.0, "step": 11782 }, { "epoch": 1.4989187126319807, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.223978042602539, "learning_rate": 1e-05, "loss": 0.5043, "mean_token_accuracy": 0.8689167499542236, "num_tokens": 447424860.0, "step": 11783 }, { "epoch": 1.4990459229105713, "ewc_loss": 0.10693359375, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.984128952026367, "learning_rate": 1e-05, "loss": 0.5495, "mean_token_accuracy": 0.8510695099830627, "num_tokens": 447465179.0, "step": 11784 }, { "epoch": 1.4991731331891618, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.237167835235596, "learning_rate": 1e-05, "loss": 0.5259, "mean_token_accuracy": 0.8566714525222778, "num_tokens": 447505960.0, "step": 11785 }, { "epoch": 1.4993003434677523, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.8186421394348145, "learning_rate": 1e-05, "loss": 0.515, "mean_token_accuracy": 0.8605797290802002, "num_tokens": 447545410.0, "step": 11786 }, { "epoch": 1.4994275537463426, "ewc_loss": 0.11474609375, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.374767780303955, "learning_rate": 1e-05, "loss": 0.5071, "mean_token_accuracy": 0.8649105429649353, "num_tokens": 447586294.0, "step": 11787 }, { "epoch": 1.4995547640249332, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.6391706466674805, "learning_rate": 1e-05, "loss": 0.4715, "mean_token_accuracy": 0.8722389936447144, "num_tokens": 447628841.0, "step": 11788 }, { "epoch": 1.4996819743035237, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.641343593597412, "learning_rate": 1e-05, "loss": 0.5115, "mean_token_accuracy": 0.8674455285072327, "num_tokens": 447669456.0, "step": 11789 }, { "epoch": 1.4998091845821142, "ewc_loss": 0.10009765625, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.501469135284424, "learning_rate": 1e-05, "loss": 0.4762, "mean_token_accuracy": 0.8769275546073914, "num_tokens": 447703297.0, "step": 11790 }, { "epoch": 1.4999363948607047, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 7.04215145111084, "learning_rate": 1e-05, "loss": 0.5631, "mean_token_accuracy": 0.8544654846191406, "num_tokens": 447742497.0, "step": 11791 }, { "epoch": 1.5000636051392953, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.529839992523193, "learning_rate": 1e-05, "loss": 0.4725, "mean_token_accuracy": 0.8703216910362244, "num_tokens": 447783956.0, "step": 11792 }, { "epoch": 1.5001908154178858, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 0.0001220703125, "grad_norm": 7.465879917144775, "learning_rate": 1e-05, "loss": 0.5343, "mean_token_accuracy": 0.8681764602661133, "num_tokens": 447824527.0, "step": 11793 }, { "epoch": 1.5003180256964763, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 5.941465377807617, "learning_rate": 1e-05, "loss": 0.5046, "mean_token_accuracy": 0.8652144074440002, "num_tokens": 447863676.0, "step": 11794 }, { "epoch": 1.5004452359750666, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 0.00012159347534179688, "grad_norm": 7.499960422515869, "learning_rate": 1e-05, "loss": 0.5688, "mean_token_accuracy": 0.8544487953186035, "num_tokens": 447906157.0, "step": 11795 }, { "epoch": 1.5005724462536572, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.418405532836914, "learning_rate": 1e-05, "loss": 0.4631, "mean_token_accuracy": 0.8834906816482544, "num_tokens": 447944016.0, "step": 11796 }, { "epoch": 1.5006996565322477, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 7.056540489196777, "learning_rate": 1e-05, "loss": 0.4812, "mean_token_accuracy": 0.8791606426239014, "num_tokens": 447979500.0, "step": 11797 }, { "epoch": 1.5008268668108382, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.528556823730469, "learning_rate": 1e-05, "loss": 0.5335, "mean_token_accuracy": 0.8553111553192139, "num_tokens": 448012770.0, "step": 11798 }, { "epoch": 1.5009540770894287, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.5388898849487305, "learning_rate": 1e-05, "loss": 0.4953, "mean_token_accuracy": 0.8704283237457275, "num_tokens": 448048227.0, "step": 11799 }, { "epoch": 1.5010812873680193, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.497047424316406, "learning_rate": 1e-05, "loss": 0.5433, "mean_token_accuracy": 0.8582865595817566, "num_tokens": 448081672.0, "step": 11800 }, { "epoch": 1.5012084976466098, "ewc_loss": 0.10888671875, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 6.105006217956543, "learning_rate": 1e-05, "loss": 0.4958, "mean_token_accuracy": 0.8678627014160156, "num_tokens": 448122693.0, "step": 11801 }, { "epoch": 1.5013357079252003, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.527137279510498, "learning_rate": 1e-05, "loss": 0.5447, "mean_token_accuracy": 0.8588197231292725, "num_tokens": 448161246.0, "step": 11802 }, { "epoch": 1.5014629182037909, "ewc_loss": 0.10400390625, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.825894355773926, "learning_rate": 1e-05, "loss": 0.5156, "mean_token_accuracy": 0.8670394420623779, "num_tokens": 448205233.0, "step": 11803 }, { "epoch": 1.5015901284823814, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.655933380126953, "learning_rate": 1e-05, "loss": 0.5718, "mean_token_accuracy": 0.8458449244499207, "num_tokens": 448247404.0, "step": 11804 }, { "epoch": 1.501717338760972, "ewc_loss": 0.10205078125, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.682883262634277, "learning_rate": 1e-05, "loss": 0.4763, "mean_token_accuracy": 0.8716211318969727, "num_tokens": 448286874.0, "step": 11805 }, { "epoch": 1.5018445490395624, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.854909896850586, "learning_rate": 1e-05, "loss": 0.5721, "mean_token_accuracy": 0.8498233556747437, "num_tokens": 448325461.0, "step": 11806 }, { "epoch": 1.501971759318153, "ewc_loss": 0.10302734375, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.653476715087891, "learning_rate": 1e-05, "loss": 0.4582, "mean_token_accuracy": 0.8773958683013916, "num_tokens": 448366871.0, "step": 11807 }, { "epoch": 1.5020989695967435, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 7.092264175415039, "learning_rate": 1e-05, "loss": 0.4844, "mean_token_accuracy": 0.8767340183258057, "num_tokens": 448396798.0, "step": 11808 }, { "epoch": 1.502226179875334, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.785959720611572, "learning_rate": 1e-05, "loss": 0.4792, "mean_token_accuracy": 0.8710660338401794, "num_tokens": 448437782.0, "step": 11809 }, { "epoch": 1.5023533901539246, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 0.00011396408081054688, "grad_norm": 7.202322483062744, "learning_rate": 1e-05, "loss": 0.4971, "mean_token_accuracy": 0.8725605010986328, "num_tokens": 448469892.0, "step": 11810 }, { "epoch": 1.502480600432515, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.016173839569092, "learning_rate": 1e-05, "loss": 0.5122, "mean_token_accuracy": 0.8641226291656494, "num_tokens": 448508233.0, "step": 11811 }, { "epoch": 1.5026078107111056, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 0.00010967254638671875, "grad_norm": 7.097248077392578, "learning_rate": 1e-05, "loss": 0.5367, "mean_token_accuracy": 0.8565851449966431, "num_tokens": 448541071.0, "step": 11812 }, { "epoch": 1.502735020989696, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.178061485290527, "learning_rate": 1e-05, "loss": 0.5318, "mean_token_accuracy": 0.8626148700714111, "num_tokens": 448571348.0, "step": 11813 }, { "epoch": 1.5028622312682864, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.816952228546143, "learning_rate": 1e-05, "loss": 0.5063, "mean_token_accuracy": 0.8696275949478149, "num_tokens": 448611889.0, "step": 11814 }, { "epoch": 1.502989441546877, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.200524806976318, "learning_rate": 1e-05, "loss": 0.4788, "mean_token_accuracy": 0.8756939768791199, "num_tokens": 448646683.0, "step": 11815 }, { "epoch": 1.5031166518254675, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.5227227210998535, "learning_rate": 1e-05, "loss": 0.4875, "mean_token_accuracy": 0.8699877262115479, "num_tokens": 448688154.0, "step": 11816 }, { "epoch": 1.503243862104058, "ewc_loss": 0.11474609375, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.175875663757324, "learning_rate": 1e-05, "loss": 0.4987, "mean_token_accuracy": 0.8653415441513062, "num_tokens": 448726343.0, "step": 11817 }, { "epoch": 1.5033710723826486, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.289079189300537, "learning_rate": 1e-05, "loss": 0.4385, "mean_token_accuracy": 0.8859443068504333, "num_tokens": 448770376.0, "step": 11818 }, { "epoch": 1.503498282661239, "ewc_loss": 0.11181640625, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.110525131225586, "learning_rate": 1e-05, "loss": 0.4612, "mean_token_accuracy": 0.878838062286377, "num_tokens": 448799247.0, "step": 11819 }, { "epoch": 1.5036254929398294, "ewc_loss": 0.11083984375, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.161201000213623, "learning_rate": 1e-05, "loss": 0.5478, "mean_token_accuracy": 0.8595561385154724, "num_tokens": 448839379.0, "step": 11820 }, { "epoch": 1.50375270321842, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 6.0342698097229, "learning_rate": 1e-05, "loss": 0.5513, "mean_token_accuracy": 0.8509777784347534, "num_tokens": 448873924.0, "step": 11821 }, { "epoch": 1.5038799134970104, "ewc_loss": 0.109375, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 6.1117072105407715, "learning_rate": 1e-05, "loss": 0.5071, "mean_token_accuracy": 0.8652737736701965, "num_tokens": 448915765.0, "step": 11822 }, { "epoch": 1.504007123775601, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.933743953704834, "learning_rate": 1e-05, "loss": 0.4737, "mean_token_accuracy": 0.8767982721328735, "num_tokens": 448953116.0, "step": 11823 }, { "epoch": 1.5041343340541915, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 6.1462788581848145, "learning_rate": 1e-05, "loss": 0.5181, "mean_token_accuracy": 0.862882137298584, "num_tokens": 448989065.0, "step": 11824 }, { "epoch": 1.504261544332782, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.81830358505249, "learning_rate": 1e-05, "loss": 0.5066, "mean_token_accuracy": 0.8626681566238403, "num_tokens": 449029608.0, "step": 11825 }, { "epoch": 1.5043887546113726, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.247867584228516, "learning_rate": 1e-05, "loss": 0.5463, "mean_token_accuracy": 0.8589686751365662, "num_tokens": 449068288.0, "step": 11826 }, { "epoch": 1.504515964889963, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.673923015594482, "learning_rate": 1e-05, "loss": 0.487, "mean_token_accuracy": 0.8663792610168457, "num_tokens": 449107754.0, "step": 11827 }, { "epoch": 1.5046431751685536, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.4254679679870605, "learning_rate": 1e-05, "loss": 0.4238, "mean_token_accuracy": 0.8913540244102478, "num_tokens": 449148680.0, "step": 11828 }, { "epoch": 1.5047703854471441, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.521798133850098, "learning_rate": 1e-05, "loss": 0.4916, "mean_token_accuracy": 0.8675141334533691, "num_tokens": 449191616.0, "step": 11829 }, { "epoch": 1.5048975957257347, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.780401229858398, "learning_rate": 1e-05, "loss": 0.4763, "mean_token_accuracy": 0.8779310584068298, "num_tokens": 449227149.0, "step": 11830 }, { "epoch": 1.5050248060043252, "ewc_loss": 0.099609375, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.4257307052612305, "learning_rate": 1e-05, "loss": 0.4891, "mean_token_accuracy": 0.8665065169334412, "num_tokens": 449259186.0, "step": 11831 }, { "epoch": 1.5051520162829157, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 0.00011587142944335938, "grad_norm": 7.261487007141113, "learning_rate": 1e-05, "loss": 0.4879, "mean_token_accuracy": 0.878092885017395, "num_tokens": 449294403.0, "step": 11832 }, { "epoch": 1.5052792265615063, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.585111141204834, "learning_rate": 1e-05, "loss": 0.4331, "mean_token_accuracy": 0.8869351148605347, "num_tokens": 449334205.0, "step": 11833 }, { "epoch": 1.5054064368400968, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.000209808349609375, "ewc_loss_parallel": 0.000125885009765625, "grad_norm": 7.665166854858398, "learning_rate": 1e-05, "loss": 0.6023, "mean_token_accuracy": 0.8520590662956238, "num_tokens": 449372501.0, "step": 11834 }, { "epoch": 1.5055336471186873, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.187565803527832, "learning_rate": 1e-05, "loss": 0.5427, "mean_token_accuracy": 0.857690691947937, "num_tokens": 449410394.0, "step": 11835 }, { "epoch": 1.5056608573972778, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 0.00012159347534179688, "grad_norm": 7.480130672454834, "learning_rate": 1e-05, "loss": 0.5211, "mean_token_accuracy": 0.8715994954109192, "num_tokens": 449449494.0, "step": 11836 }, { "epoch": 1.5057880676758684, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 6.549111366271973, "learning_rate": 1e-05, "loss": 0.521, "mean_token_accuracy": 0.8644887208938599, "num_tokens": 449485086.0, "step": 11837 }, { "epoch": 1.5059152779544587, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 7.005523681640625, "learning_rate": 1e-05, "loss": 0.5187, "mean_token_accuracy": 0.8678122758865356, "num_tokens": 449518681.0, "step": 11838 }, { "epoch": 1.5060424882330492, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.000209808349609375, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.501185417175293, "learning_rate": 1e-05, "loss": 0.4886, "mean_token_accuracy": 0.8740912079811096, "num_tokens": 449557439.0, "step": 11839 }, { "epoch": 1.5061696985116397, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.000209808349609375, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.57158088684082, "learning_rate": 1e-05, "loss": 0.5495, "mean_token_accuracy": 0.8530608415603638, "num_tokens": 449600571.0, "step": 11840 }, { "epoch": 1.5062969087902303, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.000209808349609375, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.393195629119873, "learning_rate": 1e-05, "loss": 0.5396, "mean_token_accuracy": 0.8579792380332947, "num_tokens": 449639323.0, "step": 11841 }, { "epoch": 1.5064241190688208, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.255013465881348, "learning_rate": 1e-05, "loss": 0.495, "mean_token_accuracy": 0.865787923336029, "num_tokens": 449680082.0, "step": 11842 }, { "epoch": 1.5065513293474113, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.319090843200684, "learning_rate": 1e-05, "loss": 0.5462, "mean_token_accuracy": 0.8550663590431213, "num_tokens": 449717655.0, "step": 11843 }, { "epoch": 1.5066785396260016, "ewc_loss": 0.10888671875, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 6.055200099945068, "learning_rate": 1e-05, "loss": 0.4538, "mean_token_accuracy": 0.8801355361938477, "num_tokens": 449756833.0, "step": 11844 }, { "epoch": 1.5068057499045922, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.258882999420166, "learning_rate": 1e-05, "loss": 0.4985, "mean_token_accuracy": 0.8692196011543274, "num_tokens": 449796525.0, "step": 11845 }, { "epoch": 1.5069329601831827, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.933101177215576, "learning_rate": 1e-05, "loss": 0.4937, "mean_token_accuracy": 0.8682307004928589, "num_tokens": 449837731.0, "step": 11846 }, { "epoch": 1.5070601704617732, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.237129211425781, "learning_rate": 1e-05, "loss": 0.4794, "mean_token_accuracy": 0.876004695892334, "num_tokens": 449876235.0, "step": 11847 }, { "epoch": 1.5071873807403637, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.844763278961182, "learning_rate": 1e-05, "loss": 0.4839, "mean_token_accuracy": 0.8690587282180786, "num_tokens": 449916153.0, "step": 11848 }, { "epoch": 1.5073145910189543, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.272485256195068, "learning_rate": 1e-05, "loss": 0.5446, "mean_token_accuracy": 0.8542622327804565, "num_tokens": 449957969.0, "step": 11849 }, { "epoch": 1.5074418012975448, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.724258899688721, "learning_rate": 1e-05, "loss": 0.468, "mean_token_accuracy": 0.8727177381515503, "num_tokens": 449997464.0, "step": 11850 }, { "epoch": 1.5075690115761353, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.401187896728516, "learning_rate": 1e-05, "loss": 0.5261, "mean_token_accuracy": 0.8591910600662231, "num_tokens": 450038102.0, "step": 11851 }, { "epoch": 1.5076962218547258, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.582695960998535, "learning_rate": 1e-05, "loss": 0.4597, "mean_token_accuracy": 0.8776007890701294, "num_tokens": 450073832.0, "step": 11852 }, { "epoch": 1.5078234321333164, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.661628246307373, "learning_rate": 1e-05, "loss": 0.4704, "mean_token_accuracy": 0.8780515193939209, "num_tokens": 450108662.0, "step": 11853 }, { "epoch": 1.507950642411907, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.427147388458252, "learning_rate": 1e-05, "loss": 0.4331, "mean_token_accuracy": 0.8831742405891418, "num_tokens": 450144601.0, "step": 11854 }, { "epoch": 1.5080778526904974, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 0.00011014938354492188, "grad_norm": 7.0909342765808105, "learning_rate": 1e-05, "loss": 0.539, "mean_token_accuracy": 0.8617664575576782, "num_tokens": 450190112.0, "step": 11855 }, { "epoch": 1.508205062969088, "ewc_loss": 0.10205078125, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.460953712463379, "learning_rate": 1e-05, "loss": 0.4431, "mean_token_accuracy": 0.8801236152648926, "num_tokens": 450227660.0, "step": 11856 }, { "epoch": 1.5083322732476785, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 0.00012493133544921875, "grad_norm": 7.572295188903809, "learning_rate": 1e-05, "loss": 0.5074, "mean_token_accuracy": 0.8762494325637817, "num_tokens": 450265243.0, "step": 11857 }, { "epoch": 1.508459483526269, "ewc_loss": 0.11376953125, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 5.920435905456543, "learning_rate": 1e-05, "loss": 0.5149, "mean_token_accuracy": 0.8617712259292603, "num_tokens": 450303382.0, "step": 11858 }, { "epoch": 1.5085866938048595, "ewc_loss": 0.1484375, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 0.00012683868408203125, "grad_norm": 7.670768737792969, "learning_rate": 1e-05, "loss": 0.5272, "mean_token_accuracy": 0.8713414669036865, "num_tokens": 450335458.0, "step": 11859 }, { "epoch": 1.50871390408345, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 6.466159343719482, "learning_rate": 1e-05, "loss": 0.5167, "mean_token_accuracy": 0.8659839630126953, "num_tokens": 450373375.0, "step": 11860 }, { "epoch": 1.5088411143620406, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 0.00011348724365234375, "grad_norm": 7.2563557624816895, "learning_rate": 1e-05, "loss": 0.4867, "mean_token_accuracy": 0.8787645101547241, "num_tokens": 450406258.0, "step": 11861 }, { "epoch": 1.508968324640631, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 6.56036376953125, "learning_rate": 1e-05, "loss": 0.487, "mean_token_accuracy": 0.8763246536254883, "num_tokens": 450441558.0, "step": 11862 }, { "epoch": 1.5090955349192214, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.806951999664307, "learning_rate": 1e-05, "loss": 0.4897, "mean_token_accuracy": 0.874427318572998, "num_tokens": 450477685.0, "step": 11863 }, { "epoch": 1.509222745197812, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.432650566101074, "learning_rate": 1e-05, "loss": 0.5466, "mean_token_accuracy": 0.854743480682373, "num_tokens": 450516661.0, "step": 11864 }, { "epoch": 1.5093499554764025, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.444833755493164, "learning_rate": 1e-05, "loss": 0.4921, "mean_token_accuracy": 0.8700941205024719, "num_tokens": 450555091.0, "step": 11865 }, { "epoch": 1.509477165754993, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.308841705322266, "learning_rate": 1e-05, "loss": 0.5005, "mean_token_accuracy": 0.8673779964447021, "num_tokens": 450596172.0, "step": 11866 }, { "epoch": 1.5096043760335836, "ewc_loss": 0.11181640625, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.21396541595459, "learning_rate": 1e-05, "loss": 0.5084, "mean_token_accuracy": 0.8641357421875, "num_tokens": 450636902.0, "step": 11867 }, { "epoch": 1.509731586312174, "ewc_loss": 0.11376953125, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.207415580749512, "learning_rate": 1e-05, "loss": 0.5672, "mean_token_accuracy": 0.8540337085723877, "num_tokens": 450682826.0, "step": 11868 }, { "epoch": 1.5098587965907644, "ewc_loss": 0.10791015625, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 6.065664768218994, "learning_rate": 1e-05, "loss": 0.5188, "mean_token_accuracy": 0.8587746620178223, "num_tokens": 450722995.0, "step": 11869 }, { "epoch": 1.509986006869355, "ewc_loss": 0.11083984375, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.131448268890381, "learning_rate": 1e-05, "loss": 0.4681, "mean_token_accuracy": 0.8757674694061279, "num_tokens": 450762109.0, "step": 11870 }, { "epoch": 1.5101132171479454, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.991832733154297, "learning_rate": 1e-05, "loss": 0.4755, "mean_token_accuracy": 0.8713890314102173, "num_tokens": 450799543.0, "step": 11871 }, { "epoch": 1.510240427426536, "ewc_loss": 0.10888671875, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 6.086360454559326, "learning_rate": 1e-05, "loss": 0.5116, "mean_token_accuracy": 0.8656414747238159, "num_tokens": 450840424.0, "step": 11872 }, { "epoch": 1.5103676377051265, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.934412002563477, "learning_rate": 1e-05, "loss": 0.4717, "mean_token_accuracy": 0.873563289642334, "num_tokens": 450877652.0, "step": 11873 }, { "epoch": 1.510494847983717, "ewc_loss": 0.109375, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 6.065232276916504, "learning_rate": 1e-05, "loss": 0.5495, "mean_token_accuracy": 0.8486086130142212, "num_tokens": 450916171.0, "step": 11874 }, { "epoch": 1.5106220582623076, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.882785797119141, "learning_rate": 1e-05, "loss": 0.4691, "mean_token_accuracy": 0.873968243598938, "num_tokens": 450950455.0, "step": 11875 }, { "epoch": 1.510749268540898, "ewc_loss": 0.109375, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 6.074221611022949, "learning_rate": 1e-05, "loss": 0.5033, "mean_token_accuracy": 0.8673524260520935, "num_tokens": 450990747.0, "step": 11876 }, { "epoch": 1.5108764788194886, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.806183338165283, "learning_rate": 1e-05, "loss": 0.4732, "mean_token_accuracy": 0.8738187551498413, "num_tokens": 451029396.0, "step": 11877 }, { "epoch": 1.5110036890980791, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 6.140236854553223, "learning_rate": 1e-05, "loss": 0.4878, "mean_token_accuracy": 0.8673824667930603, "num_tokens": 451065071.0, "step": 11878 }, { "epoch": 1.5111308993766697, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.706859588623047, "learning_rate": 1e-05, "loss": 0.4808, "mean_token_accuracy": 0.8712653517723083, "num_tokens": 451105487.0, "step": 11879 }, { "epoch": 1.5112581096552602, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.286778926849365, "learning_rate": 1e-05, "loss": 0.5255, "mean_token_accuracy": 0.8591234683990479, "num_tokens": 451141538.0, "step": 11880 }, { "epoch": 1.5113853199338507, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.5460405349731445, "learning_rate": 1e-05, "loss": 0.4444, "mean_token_accuracy": 0.8823760747909546, "num_tokens": 451175080.0, "step": 11881 }, { "epoch": 1.5115125302124413, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.573649883270264, "learning_rate": 1e-05, "loss": 0.5063, "mean_token_accuracy": 0.8677036762237549, "num_tokens": 451216117.0, "step": 11882 }, { "epoch": 1.5116397404910318, "ewc_loss": 0.09814453125, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 7.724761962890625e-05, "grad_norm": 5.329092502593994, "learning_rate": 1e-05, "loss": 0.4652, "mean_token_accuracy": 0.8719837069511414, "num_tokens": 451251585.0, "step": 11883 }, { "epoch": 1.5117669507696223, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 0.00011301040649414062, "grad_norm": 7.139725208282471, "learning_rate": 1e-05, "loss": 0.5536, "mean_token_accuracy": 0.8564172983169556, "num_tokens": 451290401.0, "step": 11884 }, { "epoch": 1.5118941610482128, "ewc_loss": 0.099609375, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 5.2788472175598145, "learning_rate": 1e-05, "loss": 0.5063, "mean_token_accuracy": 0.8608829975128174, "num_tokens": 451331896.0, "step": 11885 }, { "epoch": 1.5120213713268034, "ewc_loss": 0.15625, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 0.00013446807861328125, "grad_norm": 7.931056499481201, "learning_rate": 1e-05, "loss": 0.5173, "mean_token_accuracy": 0.872221052646637, "num_tokens": 451368802.0, "step": 11886 }, { "epoch": 1.5121485816053937, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.0000319480896, "learning_rate": 1e-05, "loss": 0.4643, "mean_token_accuracy": 0.8814705610275269, "num_tokens": 451400101.0, "step": 11887 }, { "epoch": 1.5122757918839842, "ewc_loss": 0.16015625, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 0.00013828277587890625, "grad_norm": 8.149170875549316, "learning_rate": 1e-05, "loss": 0.5494, "mean_token_accuracy": 0.8672175407409668, "num_tokens": 451444732.0, "step": 11888 }, { "epoch": 1.5124030021625747, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 6.910727500915527, "learning_rate": 1e-05, "loss": 0.461, "mean_token_accuracy": 0.8905618786811829, "num_tokens": 451481467.0, "step": 11889 }, { "epoch": 1.5125302124411653, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 0.00012063980102539062, "grad_norm": 7.510240077972412, "learning_rate": 1e-05, "loss": 0.5256, "mean_token_accuracy": 0.8686548471450806, "num_tokens": 451520048.0, "step": 11890 }, { "epoch": 1.5126574227197558, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 0.0001163482666015625, "grad_norm": 6.907326698303223, "learning_rate": 1e-05, "loss": 0.5103, "mean_token_accuracy": 0.8717243671417236, "num_tokens": 451555121.0, "step": 11891 }, { "epoch": 1.5127846329983463, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 0.00010967254638671875, "grad_norm": 7.05264139175415, "learning_rate": 1e-05, "loss": 0.5264, "mean_token_accuracy": 0.8675332069396973, "num_tokens": 451594894.0, "step": 11892 }, { "epoch": 1.5129118432769366, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.57728385925293, "learning_rate": 1e-05, "loss": 0.492, "mean_token_accuracy": 0.8745168447494507, "num_tokens": 451627175.0, "step": 11893 }, { "epoch": 1.5130390535555271, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.77634334564209, "learning_rate": 1e-05, "loss": 0.4835, "mean_token_accuracy": 0.8755208253860474, "num_tokens": 451664177.0, "step": 11894 }, { "epoch": 1.5131662638341177, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.3130717277526855, "learning_rate": 1e-05, "loss": 0.4761, "mean_token_accuracy": 0.8766675591468811, "num_tokens": 451696964.0, "step": 11895 }, { "epoch": 1.5132934741127082, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.635662078857422, "learning_rate": 1e-05, "loss": 0.45, "mean_token_accuracy": 0.8830301761627197, "num_tokens": 451735349.0, "step": 11896 }, { "epoch": 1.5134206843912987, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 6.074643135070801, "learning_rate": 1e-05, "loss": 0.4542, "mean_token_accuracy": 0.8793031573295593, "num_tokens": 451770410.0, "step": 11897 }, { "epoch": 1.5135478946698893, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.615800857543945, "learning_rate": 1e-05, "loss": 0.4865, "mean_token_accuracy": 0.8740789294242859, "num_tokens": 451812170.0, "step": 11898 }, { "epoch": 1.5136751049484798, "ewc_loss": 0.10498046875, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.870028495788574, "learning_rate": 1e-05, "loss": 0.4772, "mean_token_accuracy": 0.8714646697044373, "num_tokens": 451858079.0, "step": 11899 }, { "epoch": 1.5138023152270703, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.775331974029541, "learning_rate": 1e-05, "loss": 0.5344, "mean_token_accuracy": 0.8618214130401611, "num_tokens": 451901120.0, "step": 11900 }, { "epoch": 1.5139295255056608, "ewc_loss": 0.10205078125, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.732351779937744, "learning_rate": 1e-05, "loss": 0.5426, "mean_token_accuracy": 0.8552138805389404, "num_tokens": 451944990.0, "step": 11901 }, { "epoch": 1.5140567357842514, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 7.064054489135742, "learning_rate": 1e-05, "loss": 0.4913, "mean_token_accuracy": 0.877007007598877, "num_tokens": 451983261.0, "step": 11902 }, { "epoch": 1.514183946062842, "ewc_loss": 0.10400390625, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.766246318817139, "learning_rate": 1e-05, "loss": 0.5173, "mean_token_accuracy": 0.8643099665641785, "num_tokens": 452019438.0, "step": 11903 }, { "epoch": 1.5143111563414324, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 0.00011587142944335938, "grad_norm": 7.335747718811035, "learning_rate": 1e-05, "loss": 0.5635, "mean_token_accuracy": 0.8539533615112305, "num_tokens": 452055941.0, "step": 11904 }, { "epoch": 1.514438366620023, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.124380111694336, "learning_rate": 1e-05, "loss": 0.4766, "mean_token_accuracy": 0.8781405091285706, "num_tokens": 452092625.0, "step": 11905 }, { "epoch": 1.5145655768986135, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 0.0001125335693359375, "grad_norm": 7.212529182434082, "learning_rate": 1e-05, "loss": 0.5297, "mean_token_accuracy": 0.8649094104766846, "num_tokens": 452126279.0, "step": 11906 }, { "epoch": 1.514692787177204, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.358675479888916, "learning_rate": 1e-05, "loss": 0.4907, "mean_token_accuracy": 0.8738738894462585, "num_tokens": 452163506.0, "step": 11907 }, { "epoch": 1.5148199974557945, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.8032097816467285, "learning_rate": 1e-05, "loss": 0.5362, "mean_token_accuracy": 0.8597201704978943, "num_tokens": 452200898.0, "step": 11908 }, { "epoch": 1.514947207734385, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.445799350738525, "learning_rate": 1e-05, "loss": 0.5412, "mean_token_accuracy": 0.8578208088874817, "num_tokens": 452237772.0, "step": 11909 }, { "epoch": 1.5150744180129756, "ewc_loss": 0.11181640625, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.305113315582275, "learning_rate": 1e-05, "loss": 0.496, "mean_token_accuracy": 0.8675168752670288, "num_tokens": 452272741.0, "step": 11910 }, { "epoch": 1.515201628291566, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.495776176452637, "learning_rate": 1e-05, "loss": 0.5115, "mean_token_accuracy": 0.8663011789321899, "num_tokens": 452307674.0, "step": 11911 }, { "epoch": 1.5153288385701564, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.85285758972168, "learning_rate": 1e-05, "loss": 0.5139, "mean_token_accuracy": 0.8590118885040283, "num_tokens": 452352186.0, "step": 11912 }, { "epoch": 1.515456048848747, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.66834020614624, "learning_rate": 1e-05, "loss": 0.4728, "mean_token_accuracy": 0.8771225214004517, "num_tokens": 452390371.0, "step": 11913 }, { "epoch": 1.5155832591273375, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.563105583190918, "learning_rate": 1e-05, "loss": 0.4368, "mean_token_accuracy": 0.8811750411987305, "num_tokens": 452424545.0, "step": 11914 }, { "epoch": 1.515710469405928, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 7.013050079345703, "learning_rate": 1e-05, "loss": 0.5186, "mean_token_accuracy": 0.8633644580841064, "num_tokens": 452456543.0, "step": 11915 }, { "epoch": 1.5158376796845185, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.505312919616699, "learning_rate": 1e-05, "loss": 0.5095, "mean_token_accuracy": 0.8584843873977661, "num_tokens": 452497884.0, "step": 11916 }, { "epoch": 1.515964889963109, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 0.00012063980102539062, "grad_norm": 7.462438106536865, "learning_rate": 1e-05, "loss": 0.5409, "mean_token_accuracy": 0.8627098798751831, "num_tokens": 452542596.0, "step": 11917 }, { "epoch": 1.5160921002416994, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 5.866494178771973, "learning_rate": 1e-05, "loss": 0.4722, "mean_token_accuracy": 0.8776230812072754, "num_tokens": 452581385.0, "step": 11918 }, { "epoch": 1.51621931052029, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.5950493812561035, "learning_rate": 1e-05, "loss": 0.5308, "mean_token_accuracy": 0.865227460861206, "num_tokens": 452621399.0, "step": 11919 }, { "epoch": 1.5163465207988804, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.413926124572754, "learning_rate": 1e-05, "loss": 0.5258, "mean_token_accuracy": 0.8630843162536621, "num_tokens": 452657782.0, "step": 11920 }, { "epoch": 1.516473731077471, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 0.00011301040649414062, "grad_norm": 7.203463554382324, "learning_rate": 1e-05, "loss": 0.4767, "mean_token_accuracy": 0.8806331157684326, "num_tokens": 452696239.0, "step": 11921 }, { "epoch": 1.5166009413560615, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.5545477867126465, "learning_rate": 1e-05, "loss": 0.5502, "mean_token_accuracy": 0.8613359332084656, "num_tokens": 452734235.0, "step": 11922 }, { "epoch": 1.516728151634652, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.741349697113037, "learning_rate": 1e-05, "loss": 0.475, "mean_token_accuracy": 0.8757144808769226, "num_tokens": 452769151.0, "step": 11923 }, { "epoch": 1.5168553619132426, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.435626983642578, "learning_rate": 1e-05, "loss": 0.4696, "mean_token_accuracy": 0.8754865527153015, "num_tokens": 452808944.0, "step": 11924 }, { "epoch": 1.516982572191833, "ewc_loss": 0.11474609375, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.4011030197143555, "learning_rate": 1e-05, "loss": 0.5007, "mean_token_accuracy": 0.8669159412384033, "num_tokens": 452845815.0, "step": 11925 }, { "epoch": 1.5171097824704236, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.299948692321777, "learning_rate": 1e-05, "loss": 0.5, "mean_token_accuracy": 0.870731770992279, "num_tokens": 452888010.0, "step": 11926 }, { "epoch": 1.5172369927490141, "ewc_loss": 0.11181640625, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.211708068847656, "learning_rate": 1e-05, "loss": 0.4591, "mean_token_accuracy": 0.8788726925849915, "num_tokens": 452924149.0, "step": 11927 }, { "epoch": 1.5173642030276047, "ewc_loss": 0.109375, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 6.153818130493164, "learning_rate": 1e-05, "loss": 0.4882, "mean_token_accuracy": 0.8737897872924805, "num_tokens": 452964436.0, "step": 11928 }, { "epoch": 1.5174914133061952, "ewc_loss": 0.11181640625, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.1611504554748535, "learning_rate": 1e-05, "loss": 0.5205, "mean_token_accuracy": 0.862633466720581, "num_tokens": 453000455.0, "step": 11929 }, { "epoch": 1.5176186235847857, "ewc_loss": 0.10693359375, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.990850448608398, "learning_rate": 1e-05, "loss": 0.4406, "mean_token_accuracy": 0.8845862150192261, "num_tokens": 453040419.0, "step": 11930 }, { "epoch": 1.5177458338633762, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 6.186546325683594, "learning_rate": 1e-05, "loss": 0.4906, "mean_token_accuracy": 0.8668794631958008, "num_tokens": 453077668.0, "step": 11931 }, { "epoch": 1.5178730441419668, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.823978900909424, "learning_rate": 1e-05, "loss": 0.4715, "mean_token_accuracy": 0.8760921955108643, "num_tokens": 453115423.0, "step": 11932 }, { "epoch": 1.5180002544205573, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.317211627960205, "learning_rate": 1e-05, "loss": 0.4966, "mean_token_accuracy": 0.8671240210533142, "num_tokens": 453150529.0, "step": 11933 }, { "epoch": 1.5181274646991478, "ewc_loss": 0.10107421875, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.645753383636475, "learning_rate": 1e-05, "loss": 0.492, "mean_token_accuracy": 0.8700703382492065, "num_tokens": 453189123.0, "step": 11934 }, { "epoch": 1.5182546749777384, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.561821460723877, "learning_rate": 1e-05, "loss": 0.4545, "mean_token_accuracy": 0.8828729391098022, "num_tokens": 453231381.0, "step": 11935 }, { "epoch": 1.5183818852563287, "ewc_loss": 0.10009765625, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.472778797149658, "learning_rate": 1e-05, "loss": 0.5419, "mean_token_accuracy": 0.856035590171814, "num_tokens": 453272134.0, "step": 11936 }, { "epoch": 1.5185090955349192, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.996026992797852, "learning_rate": 1e-05, "loss": 0.5282, "mean_token_accuracy": 0.8638795614242554, "num_tokens": 453306950.0, "step": 11937 }, { "epoch": 1.5186363058135097, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.443199157714844, "learning_rate": 1e-05, "loss": 0.5145, "mean_token_accuracy": 0.8574841618537903, "num_tokens": 453354206.0, "step": 11938 }, { "epoch": 1.5187635160921003, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.546621322631836, "learning_rate": 1e-05, "loss": 0.5363, "mean_token_accuracy": 0.864639937877655, "num_tokens": 453388033.0, "step": 11939 }, { "epoch": 1.5188907263706908, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 5.879369258880615, "learning_rate": 1e-05, "loss": 0.5095, "mean_token_accuracy": 0.8636870980262756, "num_tokens": 453421636.0, "step": 11940 }, { "epoch": 1.5190179366492813, "ewc_loss": 0.1474609375, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 0.000125885009765625, "grad_norm": 7.688544273376465, "learning_rate": 1e-05, "loss": 0.5207, "mean_token_accuracy": 0.8720194101333618, "num_tokens": 453467994.0, "step": 11941 }, { "epoch": 1.5191451469278716, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 6.499827861785889, "learning_rate": 1e-05, "loss": 0.5496, "mean_token_accuracy": 0.854872465133667, "num_tokens": 453508627.0, "step": 11942 }, { "epoch": 1.5192723572064621, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 0.00011301040649414062, "grad_norm": 7.246325492858887, "learning_rate": 1e-05, "loss": 0.5155, "mean_token_accuracy": 0.870724618434906, "num_tokens": 453547286.0, "step": 11943 }, { "epoch": 1.5193995674850527, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.638525485992432, "learning_rate": 1e-05, "loss": 0.5779, "mean_token_accuracy": 0.8468073606491089, "num_tokens": 453588485.0, "step": 11944 }, { "epoch": 1.5195267777636432, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.715475559234619, "learning_rate": 1e-05, "loss": 0.521, "mean_token_accuracy": 0.8645455241203308, "num_tokens": 453633103.0, "step": 11945 }, { "epoch": 1.5196539880422337, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.525158882141113, "learning_rate": 1e-05, "loss": 0.4973, "mean_token_accuracy": 0.8731777667999268, "num_tokens": 453671817.0, "step": 11946 }, { "epoch": 1.5197811983208243, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.325634002685547, "learning_rate": 1e-05, "loss": 0.502, "mean_token_accuracy": 0.8687543272972107, "num_tokens": 453704704.0, "step": 11947 }, { "epoch": 1.5199084085994148, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.452586650848389, "learning_rate": 1e-05, "loss": 0.5193, "mean_token_accuracy": 0.8628060221672058, "num_tokens": 453743899.0, "step": 11948 }, { "epoch": 1.5200356188780053, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 6.0424370765686035, "learning_rate": 1e-05, "loss": 0.4668, "mean_token_accuracy": 0.873371958732605, "num_tokens": 453783052.0, "step": 11949 }, { "epoch": 1.5201628291565958, "ewc_loss": 0.11474609375, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.416918754577637, "learning_rate": 1e-05, "loss": 0.5153, "mean_token_accuracy": 0.8630213737487793, "num_tokens": 453825522.0, "step": 11950 }, { "epoch": 1.5202900394351864, "ewc_loss": 0.10498046875, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.868230819702148, "learning_rate": 1e-05, "loss": 0.4711, "mean_token_accuracy": 0.8754200339317322, "num_tokens": 453859353.0, "step": 11951 }, { "epoch": 1.520417249713777, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.456324100494385, "learning_rate": 1e-05, "loss": 0.4327, "mean_token_accuracy": 0.8871144652366638, "num_tokens": 453896792.0, "step": 11952 }, { "epoch": 1.5205444599923674, "ewc_loss": 0.10400390625, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.725917816162109, "learning_rate": 1e-05, "loss": 0.4675, "mean_token_accuracy": 0.8716654777526855, "num_tokens": 453939679.0, "step": 11953 }, { "epoch": 1.520671670270958, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.6225175857543945, "learning_rate": 1e-05, "loss": 0.5411, "mean_token_accuracy": 0.8534740805625916, "num_tokens": 453970171.0, "step": 11954 }, { "epoch": 1.5207988805495485, "ewc_loss": 0.10107421875, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.604719638824463, "learning_rate": 1e-05, "loss": 0.5093, "mean_token_accuracy": 0.8601324558258057, "num_tokens": 454008470.0, "step": 11955 }, { "epoch": 1.520926090828139, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.86113977432251, "learning_rate": 1e-05, "loss": 0.5366, "mean_token_accuracy": 0.8612124919891357, "num_tokens": 454047711.0, "step": 11956 }, { "epoch": 1.5210533011067295, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.579127311706543, "learning_rate": 1e-05, "loss": 0.4955, "mean_token_accuracy": 0.866229772567749, "num_tokens": 454083494.0, "step": 11957 }, { "epoch": 1.52118051138532, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 0.00011205673217773438, "grad_norm": 7.164394378662109, "learning_rate": 1e-05, "loss": 0.5088, "mean_token_accuracy": 0.8728824257850647, "num_tokens": 454115220.0, "step": 11958 }, { "epoch": 1.5213077216639106, "ewc_loss": 0.10791015625, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.735323905944824, "learning_rate": 1e-05, "loss": 0.4642, "mean_token_accuracy": 0.8744732141494751, "num_tokens": 454152754.0, "step": 11959 }, { "epoch": 1.521434931942501, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 0.000118255615234375, "grad_norm": 7.3451972007751465, "learning_rate": 1e-05, "loss": 0.5146, "mean_token_accuracy": 0.8708286285400391, "num_tokens": 454185357.0, "step": 11960 }, { "epoch": 1.5215621422210914, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.074686527252197, "learning_rate": 1e-05, "loss": 0.4632, "mean_token_accuracy": 0.8780581951141357, "num_tokens": 454217277.0, "step": 11961 }, { "epoch": 1.521689352499682, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 0.00011157989501953125, "grad_norm": 7.190045356750488, "learning_rate": 1e-05, "loss": 0.5686, "mean_token_accuracy": 0.8513728380203247, "num_tokens": 454253252.0, "step": 11962 }, { "epoch": 1.5218165627782725, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.2972731590271, "learning_rate": 1e-05, "loss": 0.4771, "mean_token_accuracy": 0.8789082765579224, "num_tokens": 454292779.0, "step": 11963 }, { "epoch": 1.521943773056863, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.82060432434082, "learning_rate": 1e-05, "loss": 0.5116, "mean_token_accuracy": 0.8691443800926208, "num_tokens": 454336077.0, "step": 11964 }, { "epoch": 1.5220709833354535, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.337810039520264, "learning_rate": 1e-05, "loss": 0.4647, "mean_token_accuracy": 0.8799185752868652, "num_tokens": 454369088.0, "step": 11965 }, { "epoch": 1.5221981936140438, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.416624546051025, "learning_rate": 1e-05, "loss": 0.5291, "mean_token_accuracy": 0.8589667677879333, "num_tokens": 454413057.0, "step": 11966 }, { "epoch": 1.5223254038926344, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.303244113922119, "learning_rate": 1e-05, "loss": 0.4559, "mean_token_accuracy": 0.8818358182907104, "num_tokens": 454447458.0, "step": 11967 }, { "epoch": 1.522452614171225, "ewc_loss": 0.10888671875, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 6.11644172668457, "learning_rate": 1e-05, "loss": 0.4534, "mean_token_accuracy": 0.8815335035324097, "num_tokens": 454479057.0, "step": 11968 }, { "epoch": 1.5225798244498154, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.285831928253174, "learning_rate": 1e-05, "loss": 0.5328, "mean_token_accuracy": 0.8577920198440552, "num_tokens": 454520676.0, "step": 11969 }, { "epoch": 1.522707034728406, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.926279067993164, "learning_rate": 1e-05, "loss": 0.5079, "mean_token_accuracy": 0.8628399968147278, "num_tokens": 454558021.0, "step": 11970 }, { "epoch": 1.5228342450069965, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.281364440917969, "learning_rate": 1e-05, "loss": 0.4682, "mean_token_accuracy": 0.874994158744812, "num_tokens": 454594701.0, "step": 11971 }, { "epoch": 1.522961455285587, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.798227310180664, "learning_rate": 1e-05, "loss": 0.5577, "mean_token_accuracy": 0.8445454835891724, "num_tokens": 454638746.0, "step": 11972 }, { "epoch": 1.5230886655641775, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.323672294616699, "learning_rate": 1e-05, "loss": 0.4757, "mean_token_accuracy": 0.874199628829956, "num_tokens": 454676040.0, "step": 11973 }, { "epoch": 1.523215875842768, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.6876044273376465, "learning_rate": 1e-05, "loss": 0.4874, "mean_token_accuracy": 0.8666608333587646, "num_tokens": 454713262.0, "step": 11974 }, { "epoch": 1.5233430861213586, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.464931488037109, "learning_rate": 1e-05, "loss": 0.5065, "mean_token_accuracy": 0.8644189834594727, "num_tokens": 454748613.0, "step": 11975 }, { "epoch": 1.5234702963999491, "ewc_loss": 0.10205078125, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.56929349899292, "learning_rate": 1e-05, "loss": 0.4253, "mean_token_accuracy": 0.8872069120407104, "num_tokens": 454782040.0, "step": 11976 }, { "epoch": 1.5235975066785397, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.701627731323242, "learning_rate": 1e-05, "loss": 0.5263, "mean_token_accuracy": 0.8580226898193359, "num_tokens": 454828394.0, "step": 11977 }, { "epoch": 1.5237247169571302, "ewc_loss": 0.10205078125, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.476715564727783, "learning_rate": 1e-05, "loss": 0.5003, "mean_token_accuracy": 0.8594291806221008, "num_tokens": 454866324.0, "step": 11978 }, { "epoch": 1.5238519272357207, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 0.00011014938354492188, "grad_norm": 7.075625896453857, "learning_rate": 1e-05, "loss": 0.5076, "mean_token_accuracy": 0.8745057582855225, "num_tokens": 454907458.0, "step": 11979 }, { "epoch": 1.5239791375143112, "ewc_loss": 0.10400390625, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.544005870819092, "learning_rate": 1e-05, "loss": 0.4376, "mean_token_accuracy": 0.8844820261001587, "num_tokens": 454948159.0, "step": 11980 }, { "epoch": 1.5241063477929018, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.4599833488464355, "learning_rate": 1e-05, "loss": 0.5087, "mean_token_accuracy": 0.8752250671386719, "num_tokens": 454984760.0, "step": 11981 }, { "epoch": 1.5242335580714923, "ewc_loss": 0.11474609375, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 5.932522773742676, "learning_rate": 1e-05, "loss": 0.4948, "mean_token_accuracy": 0.869577169418335, "num_tokens": 455026670.0, "step": 11982 }, { "epoch": 1.5243607683500828, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 0.00012111663818359375, "grad_norm": 7.491788864135742, "learning_rate": 1e-05, "loss": 0.5389, "mean_token_accuracy": 0.867353618144989, "num_tokens": 455063470.0, "step": 11983 }, { "epoch": 1.5244879786286734, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.359720706939697, "learning_rate": 1e-05, "loss": 0.4642, "mean_token_accuracy": 0.8823310136795044, "num_tokens": 455097787.0, "step": 11984 }, { "epoch": 1.5246151889072637, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 0.00011014938354492188, "grad_norm": 7.12434196472168, "learning_rate": 1e-05, "loss": 0.4691, "mean_token_accuracy": 0.8826061487197876, "num_tokens": 455137987.0, "step": 11985 }, { "epoch": 1.5247423991858542, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.454268932342529, "learning_rate": 1e-05, "loss": 0.5515, "mean_token_accuracy": 0.8537148237228394, "num_tokens": 455184048.0, "step": 11986 }, { "epoch": 1.5248696094644447, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.684420108795166, "learning_rate": 1e-05, "loss": 0.483, "mean_token_accuracy": 0.8732725381851196, "num_tokens": 455219876.0, "step": 11987 }, { "epoch": 1.5249968197430352, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.41101598739624, "learning_rate": 1e-05, "loss": 0.4959, "mean_token_accuracy": 0.8719350099563599, "num_tokens": 455253491.0, "step": 11988 }, { "epoch": 1.5251240300216258, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.294962406158447, "learning_rate": 1e-05, "loss": 0.5109, "mean_token_accuracy": 0.8645817637443542, "num_tokens": 455293183.0, "step": 11989 }, { "epoch": 1.5252512403002163, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.3595476150512695, "learning_rate": 1e-05, "loss": 0.4558, "mean_token_accuracy": 0.8837733268737793, "num_tokens": 455333970.0, "step": 11990 }, { "epoch": 1.5253784505788066, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 6.059751033782959, "learning_rate": 1e-05, "loss": 0.5025, "mean_token_accuracy": 0.867061972618103, "num_tokens": 455371547.0, "step": 11991 }, { "epoch": 1.5255056608573971, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.3275299072265625, "learning_rate": 1e-05, "loss": 0.5096, "mean_token_accuracy": 0.8684214353561401, "num_tokens": 455403427.0, "step": 11992 }, { "epoch": 1.5256328711359877, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.885384559631348, "learning_rate": 1e-05, "loss": 0.4301, "mean_token_accuracy": 0.892343282699585, "num_tokens": 455434707.0, "step": 11993 }, { "epoch": 1.5257600814145782, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.349970817565918, "learning_rate": 1e-05, "loss": 0.4748, "mean_token_accuracy": 0.874713122844696, "num_tokens": 455473255.0, "step": 11994 }, { "epoch": 1.5258872916931687, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.748043060302734, "learning_rate": 1e-05, "loss": 0.4798, "mean_token_accuracy": 0.8712290525436401, "num_tokens": 455504631.0, "step": 11995 }, { "epoch": 1.5260145019717593, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.462183475494385, "learning_rate": 1e-05, "loss": 0.4631, "mean_token_accuracy": 0.8798980712890625, "num_tokens": 455542652.0, "step": 11996 }, { "epoch": 1.5261417122503498, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.617167949676514, "learning_rate": 1e-05, "loss": 0.4716, "mean_token_accuracy": 0.8739434480667114, "num_tokens": 455577912.0, "step": 11997 }, { "epoch": 1.5262689225289403, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.664894104003906, "learning_rate": 1e-05, "loss": 0.4823, "mean_token_accuracy": 0.8748220205307007, "num_tokens": 455617900.0, "step": 11998 }, { "epoch": 1.5263961328075308, "ewc_loss": 0.10107421875, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.49981689453125, "learning_rate": 1e-05, "loss": 0.4657, "mean_token_accuracy": 0.8752229809761047, "num_tokens": 455656779.0, "step": 11999 }, { "epoch": 1.5265233430861214, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 7.021790981292725, "learning_rate": 1e-05, "loss": 0.4952, "mean_token_accuracy": 0.8752357363700867, "num_tokens": 455692620.0, "step": 12000 }, { "epoch": 1.526650553364712, "ewc_loss": 0.10302734375, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.531738758087158, "learning_rate": 1e-05, "loss": 0.4682, "mean_token_accuracy": 0.8742287755012512, "num_tokens": 455729805.0, "step": 12001 }, { "epoch": 1.5267777636433024, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 0.00012111663818359375, "grad_norm": 7.417194366455078, "learning_rate": 1e-05, "loss": 0.5104, "mean_token_accuracy": 0.8747578859329224, "num_tokens": 455767529.0, "step": 12002 }, { "epoch": 1.526904973921893, "ewc_loss": 0.11376953125, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 5.878378391265869, "learning_rate": 1e-05, "loss": 0.4902, "mean_token_accuracy": 0.8689191341400146, "num_tokens": 455807898.0, "step": 12003 }, { "epoch": 1.5270321842004835, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.488566875457764, "learning_rate": 1e-05, "loss": 0.5237, "mean_token_accuracy": 0.8692734241485596, "num_tokens": 455848726.0, "step": 12004 }, { "epoch": 1.527159394479074, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.316259860992432, "learning_rate": 1e-05, "loss": 0.5104, "mean_token_accuracy": 0.8681052327156067, "num_tokens": 455892363.0, "step": 12005 }, { "epoch": 1.5272866047576645, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 0.00011110305786132812, "grad_norm": 7.156160354614258, "learning_rate": 1e-05, "loss": 0.5435, "mean_token_accuracy": 0.8576477766036987, "num_tokens": 455930104.0, "step": 12006 }, { "epoch": 1.527413815036255, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.466043949127197, "learning_rate": 1e-05, "loss": 0.5229, "mean_token_accuracy": 0.8638587594032288, "num_tokens": 455964272.0, "step": 12007 }, { "epoch": 1.5275410253148456, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.691662788391113, "learning_rate": 1e-05, "loss": 0.5071, "mean_token_accuracy": 0.8660235404968262, "num_tokens": 456003020.0, "step": 12008 }, { "epoch": 1.527668235593436, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.409554481506348, "learning_rate": 1e-05, "loss": 0.4988, "mean_token_accuracy": 0.8702466487884521, "num_tokens": 456040305.0, "step": 12009 }, { "epoch": 1.5277954458720264, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.294159889221191, "learning_rate": 1e-05, "loss": 0.5225, "mean_token_accuracy": 0.8630767464637756, "num_tokens": 456080477.0, "step": 12010 }, { "epoch": 1.527922656150617, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.362300872802734, "learning_rate": 1e-05, "loss": 0.472, "mean_token_accuracy": 0.8785319328308105, "num_tokens": 456116941.0, "step": 12011 }, { "epoch": 1.5280498664292075, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 6.02766752243042, "learning_rate": 1e-05, "loss": 0.4982, "mean_token_accuracy": 0.8642358183860779, "num_tokens": 456156184.0, "step": 12012 }, { "epoch": 1.528177076707798, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.339541912078857, "learning_rate": 1e-05, "loss": 0.5432, "mean_token_accuracy": 0.8552892804145813, "num_tokens": 456200920.0, "step": 12013 }, { "epoch": 1.5283042869863885, "ewc_loss": 0.10498046875, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.862774848937988, "learning_rate": 1e-05, "loss": 0.4582, "mean_token_accuracy": 0.8777633905410767, "num_tokens": 456235632.0, "step": 12014 }, { "epoch": 1.5284314972649788, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.387352466583252, "learning_rate": 1e-05, "loss": 0.4777, "mean_token_accuracy": 0.8699442744255066, "num_tokens": 456264898.0, "step": 12015 }, { "epoch": 1.5285587075435694, "ewc_loss": 0.10400390625, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.733212471008301, "learning_rate": 1e-05, "loss": 0.4957, "mean_token_accuracy": 0.8669742345809937, "num_tokens": 456304972.0, "step": 12016 }, { "epoch": 1.52868591782216, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.492406368255615, "learning_rate": 1e-05, "loss": 0.5028, "mean_token_accuracy": 0.8687137365341187, "num_tokens": 456344496.0, "step": 12017 }, { "epoch": 1.5288131281007504, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.609608173370361, "learning_rate": 1e-05, "loss": 0.4687, "mean_token_accuracy": 0.8752198219299316, "num_tokens": 456382408.0, "step": 12018 }, { "epoch": 1.528940338379341, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.72697639465332, "learning_rate": 1e-05, "loss": 0.5276, "mean_token_accuracy": 0.8634182214736938, "num_tokens": 456424946.0, "step": 12019 }, { "epoch": 1.5290675486579315, "ewc_loss": 0.10205078125, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.507021427154541, "learning_rate": 1e-05, "loss": 0.4563, "mean_token_accuracy": 0.8788700103759766, "num_tokens": 456467432.0, "step": 12020 }, { "epoch": 1.529194758936522, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 0.00011110305786132812, "grad_norm": 7.08548641204834, "learning_rate": 1e-05, "loss": 0.4969, "mean_token_accuracy": 0.8759259581565857, "num_tokens": 456501557.0, "step": 12021 }, { "epoch": 1.5293219692151125, "ewc_loss": 0.10498046875, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.587361812591553, "learning_rate": 1e-05, "loss": 0.5735, "mean_token_accuracy": 0.8431345224380493, "num_tokens": 456540500.0, "step": 12022 }, { "epoch": 1.529449179493703, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 0.0001220703125, "grad_norm": 7.418337821960449, "learning_rate": 1e-05, "loss": 0.5031, "mean_token_accuracy": 0.8756417036056519, "num_tokens": 456574403.0, "step": 12023 }, { "epoch": 1.5295763897722936, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 5.955668926239014, "learning_rate": 1e-05, "loss": 0.4928, "mean_token_accuracy": 0.8744105696678162, "num_tokens": 456610044.0, "step": 12024 }, { "epoch": 1.5297036000508841, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 0.0001201629638671875, "grad_norm": 7.415737628936768, "learning_rate": 1e-05, "loss": 0.505, "mean_token_accuracy": 0.8723844289779663, "num_tokens": 456650289.0, "step": 12025 }, { "epoch": 1.5298308103294747, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.32900857925415, "learning_rate": 1e-05, "loss": 0.5294, "mean_token_accuracy": 0.8650389909744263, "num_tokens": 456687193.0, "step": 12026 }, { "epoch": 1.5299580206080652, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 7.06243371963501, "learning_rate": 1e-05, "loss": 0.5197, "mean_token_accuracy": 0.8702341318130493, "num_tokens": 456720553.0, "step": 12027 }, { "epoch": 1.5300852308866557, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.416157245635986, "learning_rate": 1e-05, "loss": 0.5085, "mean_token_accuracy": 0.8642462491989136, "num_tokens": 456754736.0, "step": 12028 }, { "epoch": 1.5302124411652462, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.632343769073486, "learning_rate": 1e-05, "loss": 0.454, "mean_token_accuracy": 0.8838412165641785, "num_tokens": 456791268.0, "step": 12029 }, { "epoch": 1.5303396514438368, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.35905122756958, "learning_rate": 1e-05, "loss": 0.4668, "mean_token_accuracy": 0.8790313005447388, "num_tokens": 456831728.0, "step": 12030 }, { "epoch": 1.5304668617224273, "ewc_loss": 0.11376953125, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.278061866760254, "learning_rate": 1e-05, "loss": 0.4696, "mean_token_accuracy": 0.8758686780929565, "num_tokens": 456869375.0, "step": 12031 }, { "epoch": 1.5305940720010178, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.320837020874023, "learning_rate": 1e-05, "loss": 0.5138, "mean_token_accuracy": 0.8604527711868286, "num_tokens": 456909376.0, "step": 12032 }, { "epoch": 1.5307212822796084, "ewc_loss": 0.10791015625, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 6.032528877258301, "learning_rate": 1e-05, "loss": 0.4536, "mean_token_accuracy": 0.8814014792442322, "num_tokens": 456945851.0, "step": 12033 }, { "epoch": 1.5308484925581987, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.317168235778809, "learning_rate": 1e-05, "loss": 0.504, "mean_token_accuracy": 0.8679592609405518, "num_tokens": 456979833.0, "step": 12034 }, { "epoch": 1.5309757028367892, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.8669610023498535, "learning_rate": 1e-05, "loss": 0.4286, "mean_token_accuracy": 0.8865154981613159, "num_tokens": 457016032.0, "step": 12035 }, { "epoch": 1.5311029131153797, "ewc_loss": 0.11376953125, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.341598033905029, "learning_rate": 1e-05, "loss": 0.4602, "mean_token_accuracy": 0.885195791721344, "num_tokens": 457050015.0, "step": 12036 }, { "epoch": 1.5312301233939702, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.7452392578125, "learning_rate": 1e-05, "loss": 0.4764, "mean_token_accuracy": 0.8714402914047241, "num_tokens": 457087287.0, "step": 12037 }, { "epoch": 1.5313573336725608, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.446451663970947, "learning_rate": 1e-05, "loss": 0.4997, "mean_token_accuracy": 0.8672772645950317, "num_tokens": 457123237.0, "step": 12038 }, { "epoch": 1.5314845439511513, "ewc_loss": 0.10205078125, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.598587989807129, "learning_rate": 1e-05, "loss": 0.4833, "mean_token_accuracy": 0.8664068579673767, "num_tokens": 457161834.0, "step": 12039 }, { "epoch": 1.5316117542297416, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.685056209564209, "learning_rate": 1e-05, "loss": 0.5445, "mean_token_accuracy": 0.8603943586349487, "num_tokens": 457203672.0, "step": 12040 }, { "epoch": 1.5317389645083321, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.483616828918457, "learning_rate": 1e-05, "loss": 0.4478, "mean_token_accuracy": 0.8803684711456299, "num_tokens": 457238486.0, "step": 12041 }, { "epoch": 1.5318661747869227, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 0.00011014938354492188, "grad_norm": 7.051133632659912, "learning_rate": 1e-05, "loss": 0.5487, "mean_token_accuracy": 0.8556884527206421, "num_tokens": 457270817.0, "step": 12042 }, { "epoch": 1.5319933850655132, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.521880626678467, "learning_rate": 1e-05, "loss": 0.4775, "mean_token_accuracy": 0.8697342276573181, "num_tokens": 457306812.0, "step": 12043 }, { "epoch": 1.5321205953441037, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 0.00012159347534179688, "grad_norm": 7.450859546661377, "learning_rate": 1e-05, "loss": 0.4861, "mean_token_accuracy": 0.8782918453216553, "num_tokens": 457340893.0, "step": 12044 }, { "epoch": 1.5322478056226942, "ewc_loss": 0.11474609375, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 5.900341033935547, "learning_rate": 1e-05, "loss": 0.4857, "mean_token_accuracy": 0.8720147013664246, "num_tokens": 457381668.0, "step": 12045 }, { "epoch": 1.5323750159012848, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.525343418121338, "learning_rate": 1e-05, "loss": 0.5512, "mean_token_accuracy": 0.8587472438812256, "num_tokens": 457417498.0, "step": 12046 }, { "epoch": 1.5325022261798753, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.362257480621338, "learning_rate": 1e-05, "loss": 0.5052, "mean_token_accuracy": 0.8702347278594971, "num_tokens": 457460300.0, "step": 12047 }, { "epoch": 1.5326294364584658, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 7.159140110015869, "learning_rate": 1e-05, "loss": 0.551, "mean_token_accuracy": 0.8590254783630371, "num_tokens": 457502030.0, "step": 12048 }, { "epoch": 1.5327566467370564, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.485209941864014, "learning_rate": 1e-05, "loss": 0.4861, "mean_token_accuracy": 0.8793617486953735, "num_tokens": 457540170.0, "step": 12049 }, { "epoch": 1.532883857015647, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.685573577880859, "learning_rate": 1e-05, "loss": 0.4812, "mean_token_accuracy": 0.875804603099823, "num_tokens": 457585834.0, "step": 12050 }, { "epoch": 1.5330110672942374, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.430139064788818, "learning_rate": 1e-05, "loss": 0.4689, "mean_token_accuracy": 0.8806641101837158, "num_tokens": 457622027.0, "step": 12051 }, { "epoch": 1.533138277572828, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.292593479156494, "learning_rate": 1e-05, "loss": 0.5162, "mean_token_accuracy": 0.862576425075531, "num_tokens": 457659129.0, "step": 12052 }, { "epoch": 1.5332654878514185, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.383383274078369, "learning_rate": 1e-05, "loss": 0.443, "mean_token_accuracy": 0.8865221738815308, "num_tokens": 457695046.0, "step": 12053 }, { "epoch": 1.533392698130009, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 6.019249439239502, "learning_rate": 1e-05, "loss": 0.4636, "mean_token_accuracy": 0.8765919804573059, "num_tokens": 457735650.0, "step": 12054 }, { "epoch": 1.5335199084085995, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.377857685089111, "learning_rate": 1e-05, "loss": 0.474, "mean_token_accuracy": 0.8780646920204163, "num_tokens": 457767395.0, "step": 12055 }, { "epoch": 1.53364711868719, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.840337753295898, "learning_rate": 1e-05, "loss": 0.427, "mean_token_accuracy": 0.8831324577331543, "num_tokens": 457800100.0, "step": 12056 }, { "epoch": 1.5337743289657806, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.430732250213623, "learning_rate": 1e-05, "loss": 0.4845, "mean_token_accuracy": 0.8744968771934509, "num_tokens": 457837497.0, "step": 12057 }, { "epoch": 1.533901539244371, "ewc_loss": 0.10302734375, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.692867755889893, "learning_rate": 1e-05, "loss": 0.4685, "mean_token_accuracy": 0.8725663423538208, "num_tokens": 457875711.0, "step": 12058 }, { "epoch": 1.5340287495229614, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.579790115356445, "learning_rate": 1e-05, "loss": 0.523, "mean_token_accuracy": 0.8578094244003296, "num_tokens": 457914888.0, "step": 12059 }, { "epoch": 1.534155959801552, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.577624320983887, "learning_rate": 1e-05, "loss": 0.5229, "mean_token_accuracy": 0.8551226258277893, "num_tokens": 457953315.0, "step": 12060 }, { "epoch": 1.5342831700801425, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.841825485229492, "learning_rate": 1e-05, "loss": 0.5401, "mean_token_accuracy": 0.8578313589096069, "num_tokens": 457990857.0, "step": 12061 }, { "epoch": 1.534410380358733, "ewc_loss": 0.10302734375, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.528667449951172, "learning_rate": 1e-05, "loss": 0.5016, "mean_token_accuracy": 0.8628581762313843, "num_tokens": 458034060.0, "step": 12062 }, { "epoch": 1.5345375906373235, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 0.00011444091796875, "grad_norm": 7.195044040679932, "learning_rate": 1e-05, "loss": 0.5302, "mean_token_accuracy": 0.8634098768234253, "num_tokens": 458074093.0, "step": 12063 }, { "epoch": 1.5346648009159138, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 5.714084625244141, "learning_rate": 1e-05, "loss": 0.459, "mean_token_accuracy": 0.8778966665267944, "num_tokens": 458107374.0, "step": 12064 }, { "epoch": 1.5347920111945044, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 0.0001201629638671875, "grad_norm": 7.397801399230957, "learning_rate": 1e-05, "loss": 0.5193, "mean_token_accuracy": 0.8715934157371521, "num_tokens": 458141873.0, "step": 12065 }, { "epoch": 1.534919221473095, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.088105201721191, "learning_rate": 1e-05, "loss": 0.4947, "mean_token_accuracy": 0.8718099594116211, "num_tokens": 458181483.0, "step": 12066 }, { "epoch": 1.5350464317516854, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 0.00011539459228515625, "grad_norm": 7.25982666015625, "learning_rate": 1e-05, "loss": 0.5508, "mean_token_accuracy": 0.8622196912765503, "num_tokens": 458223644.0, "step": 12067 }, { "epoch": 1.535173642030276, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.347159385681152, "learning_rate": 1e-05, "loss": 0.5058, "mean_token_accuracy": 0.8699718117713928, "num_tokens": 458263518.0, "step": 12068 }, { "epoch": 1.5353008523088665, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.867815971374512, "learning_rate": 1e-05, "loss": 0.471, "mean_token_accuracy": 0.8761817216873169, "num_tokens": 458296621.0, "step": 12069 }, { "epoch": 1.535428062587457, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.3703742027282715, "learning_rate": 1e-05, "loss": 0.5288, "mean_token_accuracy": 0.8634796142578125, "num_tokens": 458335161.0, "step": 12070 }, { "epoch": 1.5355552728660475, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.456615447998047, "learning_rate": 1e-05, "loss": 0.5399, "mean_token_accuracy": 0.8586392998695374, "num_tokens": 458373758.0, "step": 12071 }, { "epoch": 1.535682483144638, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.351269245147705, "learning_rate": 1e-05, "loss": 0.4834, "mean_token_accuracy": 0.8733906745910645, "num_tokens": 458412662.0, "step": 12072 }, { "epoch": 1.5358096934232286, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 6.117718696594238, "learning_rate": 1e-05, "loss": 0.5216, "mean_token_accuracy": 0.8615858554840088, "num_tokens": 458455314.0, "step": 12073 }, { "epoch": 1.5359369037018191, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.339285373687744, "learning_rate": 1e-05, "loss": 0.5654, "mean_token_accuracy": 0.8498363494873047, "num_tokens": 458498633.0, "step": 12074 }, { "epoch": 1.5360641139804097, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.898354530334473, "learning_rate": 1e-05, "loss": 0.4951, "mean_token_accuracy": 0.8672057390213013, "num_tokens": 458538920.0, "step": 12075 }, { "epoch": 1.5361913242590002, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.365024089813232, "learning_rate": 1e-05, "loss": 0.4977, "mean_token_accuracy": 0.8678567409515381, "num_tokens": 458574333.0, "step": 12076 }, { "epoch": 1.5363185345375907, "ewc_loss": 0.10400390625, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.776303291320801, "learning_rate": 1e-05, "loss": 0.5094, "mean_token_accuracy": 0.8607965707778931, "num_tokens": 458612131.0, "step": 12077 }, { "epoch": 1.5364457448161812, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.43697452545166, "learning_rate": 1e-05, "loss": 0.5073, "mean_token_accuracy": 0.8662689328193665, "num_tokens": 458652169.0, "step": 12078 }, { "epoch": 1.5365729550947718, "ewc_loss": 0.10205078125, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 6.1107096672058105, "learning_rate": 1e-05, "loss": 0.5308, "mean_token_accuracy": 0.8628101348876953, "num_tokens": 458687816.0, "step": 12079 }, { "epoch": 1.5367001653733623, "ewc_loss": 0.11181640625, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.186873912811279, "learning_rate": 1e-05, "loss": 0.4624, "mean_token_accuracy": 0.8754162788391113, "num_tokens": 458723600.0, "step": 12080 }, { "epoch": 1.5368273756519528, "ewc_loss": 0.10791015625, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.83739709854126, "learning_rate": 1e-05, "loss": 0.5362, "mean_token_accuracy": 0.8549743294715881, "num_tokens": 458761863.0, "step": 12081 }, { "epoch": 1.5369545859305433, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.118978500366211, "learning_rate": 1e-05, "loss": 0.5289, "mean_token_accuracy": 0.856920063495636, "num_tokens": 458794986.0, "step": 12082 }, { "epoch": 1.5370817962091337, "ewc_loss": 0.10791015625, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.788034915924072, "learning_rate": 1e-05, "loss": 0.5208, "mean_token_accuracy": 0.8608570098876953, "num_tokens": 458833939.0, "step": 12083 }, { "epoch": 1.5372090064877242, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.183990478515625, "learning_rate": 1e-05, "loss": 0.5327, "mean_token_accuracy": 0.8583151698112488, "num_tokens": 458874522.0, "step": 12084 }, { "epoch": 1.5373362167663147, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.694194316864014, "learning_rate": 1e-05, "loss": 0.5548, "mean_token_accuracy": 0.8473106622695923, "num_tokens": 458911794.0, "step": 12085 }, { "epoch": 1.5374634270449052, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.3266215324401855, "learning_rate": 1e-05, "loss": 0.5054, "mean_token_accuracy": 0.8672492504119873, "num_tokens": 458951979.0, "step": 12086 }, { "epoch": 1.5375906373234958, "ewc_loss": 0.10400390625, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.538100242614746, "learning_rate": 1e-05, "loss": 0.5103, "mean_token_accuracy": 0.8601285219192505, "num_tokens": 458993657.0, "step": 12087 }, { "epoch": 1.5377178476020863, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.610751628875732, "learning_rate": 1e-05, "loss": 0.4567, "mean_token_accuracy": 0.8812359571456909, "num_tokens": 459030216.0, "step": 12088 }, { "epoch": 1.5378450578806766, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.380186557769775, "learning_rate": 1e-05, "loss": 0.5224, "mean_token_accuracy": 0.8562358021736145, "num_tokens": 459066451.0, "step": 12089 }, { "epoch": 1.5379722681592671, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 0.00011348724365234375, "grad_norm": 7.107043266296387, "learning_rate": 1e-05, "loss": 0.4907, "mean_token_accuracy": 0.8789731860160828, "num_tokens": 459106928.0, "step": 12090 }, { "epoch": 1.5380994784378577, "ewc_loss": 0.10400390625, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.3747968673706055, "learning_rate": 1e-05, "loss": 0.5196, "mean_token_accuracy": 0.8584685921669006, "num_tokens": 459142769.0, "step": 12091 }, { "epoch": 1.5382266887164482, "ewc_loss": 0.1533203125, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 0.0001316070556640625, "grad_norm": 7.724156379699707, "learning_rate": 1e-05, "loss": 0.5238, "mean_token_accuracy": 0.8733232021331787, "num_tokens": 459178895.0, "step": 12092 }, { "epoch": 1.5383538989950387, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 5.9045329093933105, "learning_rate": 1e-05, "loss": 0.4539, "mean_token_accuracy": 0.8834545612335205, "num_tokens": 459214973.0, "step": 12093 }, { "epoch": 1.5384811092736292, "ewc_loss": 0.15625, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 0.00013446807861328125, "grad_norm": 7.929393768310547, "learning_rate": 1e-05, "loss": 0.5445, "mean_token_accuracy": 0.86637282371521, "num_tokens": 459251279.0, "step": 12094 }, { "epoch": 1.5386083195522198, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 0.0001201629638671875, "grad_norm": 6.654118537902832, "learning_rate": 1e-05, "loss": 0.5759, "mean_token_accuracy": 0.8483269810676575, "num_tokens": 459285587.0, "step": 12095 }, { "epoch": 1.5387355298308103, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 0.00012063980102539062, "grad_norm": 7.451875686645508, "learning_rate": 1e-05, "loss": 0.5716, "mean_token_accuracy": 0.8566220998764038, "num_tokens": 459326103.0, "step": 12096 }, { "epoch": 1.5388627401094008, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 0.00011587142944335938, "grad_norm": 6.72689151763916, "learning_rate": 1e-05, "loss": 0.5284, "mean_token_accuracy": 0.8672547340393066, "num_tokens": 459367075.0, "step": 12097 }, { "epoch": 1.5389899503879914, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 0.00011157989501953125, "grad_norm": 7.014901638031006, "learning_rate": 1e-05, "loss": 0.5475, "mean_token_accuracy": 0.8582703471183777, "num_tokens": 459408164.0, "step": 12098 }, { "epoch": 1.5391171606665819, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.508863925933838, "learning_rate": 1e-05, "loss": 0.533, "mean_token_accuracy": 0.8606554269790649, "num_tokens": 459443081.0, "step": 12099 }, { "epoch": 1.5392443709451724, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.700552940368652, "learning_rate": 1e-05, "loss": 0.5281, "mean_token_accuracy": 0.8644009828567505, "num_tokens": 459481469.0, "step": 12100 }, { "epoch": 1.539371581223763, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.295490741729736, "learning_rate": 1e-05, "loss": 0.5069, "mean_token_accuracy": 0.8665065169334412, "num_tokens": 459519319.0, "step": 12101 }, { "epoch": 1.5394987915023535, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.514232158660889, "learning_rate": 1e-05, "loss": 0.492, "mean_token_accuracy": 0.868911862373352, "num_tokens": 459559998.0, "step": 12102 }, { "epoch": 1.539626001780944, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.126108169555664, "learning_rate": 1e-05, "loss": 0.4549, "mean_token_accuracy": 0.8821830153465271, "num_tokens": 459596765.0, "step": 12103 }, { "epoch": 1.5397532120595345, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.439512729644775, "learning_rate": 1e-05, "loss": 0.5608, "mean_token_accuracy": 0.8505854606628418, "num_tokens": 459634023.0, "step": 12104 }, { "epoch": 1.539880422338125, "ewc_loss": 0.10791015625, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.924673557281494, "learning_rate": 1e-05, "loss": 0.4436, "mean_token_accuracy": 0.8837569952011108, "num_tokens": 459675355.0, "step": 12105 }, { "epoch": 1.5400076326167156, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.5040435791015625, "learning_rate": 1e-05, "loss": 0.5464, "mean_token_accuracy": 0.8575024604797363, "num_tokens": 459716788.0, "step": 12106 }, { "epoch": 1.5401348428953059, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.751067161560059, "learning_rate": 1e-05, "loss": 0.4688, "mean_token_accuracy": 0.8745817542076111, "num_tokens": 459752580.0, "step": 12107 }, { "epoch": 1.5402620531738964, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.711669445037842, "learning_rate": 1e-05, "loss": 0.4849, "mean_token_accuracy": 0.8747220635414124, "num_tokens": 459788594.0, "step": 12108 }, { "epoch": 1.540389263452487, "ewc_loss": 0.10302734375, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.617300987243652, "learning_rate": 1e-05, "loss": 0.4537, "mean_token_accuracy": 0.8801541328430176, "num_tokens": 459826806.0, "step": 12109 }, { "epoch": 1.5405164737310775, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 7.051981449127197, "learning_rate": 1e-05, "loss": 0.4945, "mean_token_accuracy": 0.8742851614952087, "num_tokens": 459870157.0, "step": 12110 }, { "epoch": 1.540643684009668, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.67183780670166, "learning_rate": 1e-05, "loss": 0.473, "mean_token_accuracy": 0.8740507364273071, "num_tokens": 459909081.0, "step": 12111 }, { "epoch": 1.5407708942882585, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 0.00011873245239257812, "grad_norm": 7.368143081665039, "learning_rate": 1e-05, "loss": 0.4986, "mean_token_accuracy": 0.8755741119384766, "num_tokens": 459942886.0, "step": 12112 }, { "epoch": 1.5408981045668488, "ewc_loss": 0.11474609375, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.036310195922852, "learning_rate": 1e-05, "loss": 0.4989, "mean_token_accuracy": 0.8681166172027588, "num_tokens": 459979351.0, "step": 12113 }, { "epoch": 1.5410253148454394, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 0.00011539459228515625, "grad_norm": 7.307200908660889, "learning_rate": 1e-05, "loss": 0.5286, "mean_token_accuracy": 0.8666678071022034, "num_tokens": 460026058.0, "step": 12114 }, { "epoch": 1.54115252512403, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.3919596672058105, "learning_rate": 1e-05, "loss": 0.5158, "mean_token_accuracy": 0.8658660054206848, "num_tokens": 460066414.0, "step": 12115 }, { "epoch": 1.5412797354026204, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.86029577255249, "learning_rate": 1e-05, "loss": 0.528, "mean_token_accuracy": 0.8638396263122559, "num_tokens": 460107747.0, "step": 12116 }, { "epoch": 1.541406945681211, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.494805335998535, "learning_rate": 1e-05, "loss": 0.4992, "mean_token_accuracy": 0.8720511198043823, "num_tokens": 460144418.0, "step": 12117 }, { "epoch": 1.5415341559598015, "ewc_loss": 0.11376953125, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.32564640045166, "learning_rate": 1e-05, "loss": 0.5382, "mean_token_accuracy": 0.8590703010559082, "num_tokens": 460186061.0, "step": 12118 }, { "epoch": 1.541661366238392, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.545835018157959, "learning_rate": 1e-05, "loss": 0.5152, "mean_token_accuracy": 0.8633759021759033, "num_tokens": 460222672.0, "step": 12119 }, { "epoch": 1.5417885765169825, "ewc_loss": 0.10302734375, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.866706371307373, "learning_rate": 1e-05, "loss": 0.4562, "mean_token_accuracy": 0.8784665465354919, "num_tokens": 460265992.0, "step": 12120 }, { "epoch": 1.541915786795573, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.733696460723877, "learning_rate": 1e-05, "loss": 0.4992, "mean_token_accuracy": 0.8701751232147217, "num_tokens": 460305498.0, "step": 12121 }, { "epoch": 1.5420429970741636, "ewc_loss": 0.10205078125, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.610307216644287, "learning_rate": 1e-05, "loss": 0.5035, "mean_token_accuracy": 0.8648242950439453, "num_tokens": 460336013.0, "step": 12122 }, { "epoch": 1.5421702073527541, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 7.067107200622559, "learning_rate": 1e-05, "loss": 0.5029, "mean_token_accuracy": 0.8690934777259827, "num_tokens": 460373452.0, "step": 12123 }, { "epoch": 1.5422974176313446, "ewc_loss": 0.10400390625, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.613481044769287, "learning_rate": 1e-05, "loss": 0.4878, "mean_token_accuracy": 0.8706740140914917, "num_tokens": 460414059.0, "step": 12124 }, { "epoch": 1.5424246279099352, "ewc_loss": 0.140625, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 0.00011920928955078125, "grad_norm": 7.4117865562438965, "learning_rate": 1e-05, "loss": 0.4757, "mean_token_accuracy": 0.8814775943756104, "num_tokens": 460446098.0, "step": 12125 }, { "epoch": 1.5425518381885257, "ewc_loss": 0.11376953125, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 5.964502334594727, "learning_rate": 1e-05, "loss": 0.486, "mean_token_accuracy": 0.8691459894180298, "num_tokens": 460485384.0, "step": 12126 }, { "epoch": 1.5426790484671162, "ewc_loss": 0.140625, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 0.00011920928955078125, "grad_norm": 7.4295148849487305, "learning_rate": 1e-05, "loss": 0.5441, "mean_token_accuracy": 0.861430823802948, "num_tokens": 460517200.0, "step": 12127 }, { "epoch": 1.5428062587457068, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.35469913482666, "learning_rate": 1e-05, "loss": 0.4734, "mean_token_accuracy": 0.8765101432800293, "num_tokens": 460547557.0, "step": 12128 }, { "epoch": 1.5429334690242973, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 7.076783180236816, "learning_rate": 1e-05, "loss": 0.5034, "mean_token_accuracy": 0.8765968084335327, "num_tokens": 460583189.0, "step": 12129 }, { "epoch": 1.5430606793028878, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.411426544189453, "learning_rate": 1e-05, "loss": 0.5012, "mean_token_accuracy": 0.8726128339767456, "num_tokens": 460623619.0, "step": 12130 }, { "epoch": 1.5431878895814783, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.708866119384766, "learning_rate": 1e-05, "loss": 0.5118, "mean_token_accuracy": 0.8647379875183105, "num_tokens": 460666667.0, "step": 12131 }, { "epoch": 1.5433150998600687, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.315768241882324, "learning_rate": 1e-05, "loss": 0.5081, "mean_token_accuracy": 0.8679406642913818, "num_tokens": 460704292.0, "step": 12132 }, { "epoch": 1.5434423101386592, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.435771942138672, "learning_rate": 1e-05, "loss": 0.484, "mean_token_accuracy": 0.8724813461303711, "num_tokens": 460748153.0, "step": 12133 }, { "epoch": 1.5435695204172497, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.171324729919434, "learning_rate": 1e-05, "loss": 0.4681, "mean_token_accuracy": 0.8746448755264282, "num_tokens": 460787663.0, "step": 12134 }, { "epoch": 1.5436967306958402, "ewc_loss": 0.11474609375, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.319260120391846, "learning_rate": 1e-05, "loss": 0.465, "mean_token_accuracy": 0.8774517774581909, "num_tokens": 460821643.0, "step": 12135 }, { "epoch": 1.5438239409744308, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 6.018277645111084, "learning_rate": 1e-05, "loss": 0.516, "mean_token_accuracy": 0.8612931966781616, "num_tokens": 460857180.0, "step": 12136 }, { "epoch": 1.5439511512530213, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.323987007141113, "learning_rate": 1e-05, "loss": 0.4984, "mean_token_accuracy": 0.8660770058631897, "num_tokens": 460891573.0, "step": 12137 }, { "epoch": 1.5440783615316116, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.849483966827393, "learning_rate": 1e-05, "loss": 0.4396, "mean_token_accuracy": 0.8810973763465881, "num_tokens": 460927727.0, "step": 12138 }, { "epoch": 1.5442055718102021, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.432783126831055, "learning_rate": 1e-05, "loss": 0.5356, "mean_token_accuracy": 0.8600648641586304, "num_tokens": 460965531.0, "step": 12139 }, { "epoch": 1.5443327820887927, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.682631492614746, "learning_rate": 1e-05, "loss": 0.4899, "mean_token_accuracy": 0.8693603277206421, "num_tokens": 461002900.0, "step": 12140 }, { "epoch": 1.5444599923673832, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.675937175750732, "learning_rate": 1e-05, "loss": 0.5083, "mean_token_accuracy": 0.870037317276001, "num_tokens": 461041569.0, "step": 12141 }, { "epoch": 1.5445872026459737, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.556765079498291, "learning_rate": 1e-05, "loss": 0.5111, "mean_token_accuracy": 0.8601616024971008, "num_tokens": 461080453.0, "step": 12142 }, { "epoch": 1.5447144129245642, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 7.046765327453613, "learning_rate": 1e-05, "loss": 0.5248, "mean_token_accuracy": 0.8646532893180847, "num_tokens": 461120963.0, "step": 12143 }, { "epoch": 1.5448416232031548, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.621631145477295, "learning_rate": 1e-05, "loss": 0.4658, "mean_token_accuracy": 0.8726009130477905, "num_tokens": 461156441.0, "step": 12144 }, { "epoch": 1.5449688334817453, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 0.00011730194091796875, "grad_norm": 7.376654148101807, "learning_rate": 1e-05, "loss": 0.4826, "mean_token_accuracy": 0.8804547786712646, "num_tokens": 461189058.0, "step": 12145 }, { "epoch": 1.5450960437603358, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 5.994709491729736, "learning_rate": 1e-05, "loss": 0.4645, "mean_token_accuracy": 0.8763905167579651, "num_tokens": 461223962.0, "step": 12146 }, { "epoch": 1.5452232540389264, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 0.00011682510375976562, "grad_norm": 7.311947822570801, "learning_rate": 1e-05, "loss": 0.4949, "mean_token_accuracy": 0.8774843811988831, "num_tokens": 461262114.0, "step": 12147 }, { "epoch": 1.5453504643175169, "ewc_loss": 0.125, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.368231773376465, "learning_rate": 1e-05, "loss": 0.4963, "mean_token_accuracy": 0.8710317611694336, "num_tokens": 461294633.0, "step": 12148 }, { "epoch": 1.5454776745961074, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.867443084716797, "learning_rate": 1e-05, "loss": 0.5212, "mean_token_accuracy": 0.8652231693267822, "num_tokens": 461338549.0, "step": 12149 }, { "epoch": 1.545604884874698, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.466071128845215, "learning_rate": 1e-05, "loss": 0.5168, "mean_token_accuracy": 0.8655158281326294, "num_tokens": 461372612.0, "step": 12150 }, { "epoch": 1.5457320951532885, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.360879898071289, "learning_rate": 1e-05, "loss": 0.5238, "mean_token_accuracy": 0.8586647510528564, "num_tokens": 461404060.0, "step": 12151 }, { "epoch": 1.545859305431879, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.518817901611328, "learning_rate": 1e-05, "loss": 0.5128, "mean_token_accuracy": 0.8664109706878662, "num_tokens": 461436728.0, "step": 12152 }, { "epoch": 1.5459865157104695, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.898210048675537, "learning_rate": 1e-05, "loss": 0.4505, "mean_token_accuracy": 0.8790885210037231, "num_tokens": 461475418.0, "step": 12153 }, { "epoch": 1.54611372598906, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.695117473602295, "learning_rate": 1e-05, "loss": 0.5284, "mean_token_accuracy": 0.861077070236206, "num_tokens": 461512094.0, "step": 12154 }, { "epoch": 1.5462409362676506, "ewc_loss": 0.10302734375, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.630977630615234, "learning_rate": 1e-05, "loss": 0.5409, "mean_token_accuracy": 0.8508100509643555, "num_tokens": 461553454.0, "step": 12155 }, { "epoch": 1.5463681465462409, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 7.005975246429443, "learning_rate": 1e-05, "loss": 0.5948, "mean_token_accuracy": 0.8483905792236328, "num_tokens": 461590818.0, "step": 12156 }, { "epoch": 1.5464953568248314, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.609564781188965, "learning_rate": 1e-05, "loss": 0.4387, "mean_token_accuracy": 0.8819525241851807, "num_tokens": 461638692.0, "step": 12157 }, { "epoch": 1.546622567103422, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 0.00011682510375976562, "grad_norm": 7.333691596984863, "learning_rate": 1e-05, "loss": 0.5215, "mean_token_accuracy": 0.8699339628219604, "num_tokens": 461681040.0, "step": 12158 }, { "epoch": 1.5467497773820125, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 5.912734031677246, "learning_rate": 1e-05, "loss": 0.4861, "mean_token_accuracy": 0.8687726855278015, "num_tokens": 461716999.0, "step": 12159 }, { "epoch": 1.546876987660603, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 0.00011777877807617188, "grad_norm": 7.376924514770508, "learning_rate": 1e-05, "loss": 0.5459, "mean_token_accuracy": 0.860845685005188, "num_tokens": 461755053.0, "step": 12160 }, { "epoch": 1.5470041979391935, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.253652572631836, "learning_rate": 1e-05, "loss": 0.4878, "mean_token_accuracy": 0.874848484992981, "num_tokens": 461795384.0, "step": 12161 }, { "epoch": 1.5471314082177838, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 7.100834369659424, "learning_rate": 1e-05, "loss": 0.4992, "mean_token_accuracy": 0.8700098991394043, "num_tokens": 461843951.0, "step": 12162 }, { "epoch": 1.5472586184963744, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.356592178344727, "learning_rate": 1e-05, "loss": 0.5523, "mean_token_accuracy": 0.8542494177818298, "num_tokens": 461884172.0, "step": 12163 }, { "epoch": 1.5473858287749649, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.765190124511719, "learning_rate": 1e-05, "loss": 0.5221, "mean_token_accuracy": 0.8657435774803162, "num_tokens": 461928493.0, "step": 12164 }, { "epoch": 1.5475130390535554, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.292929172515869, "learning_rate": 1e-05, "loss": 0.4874, "mean_token_accuracy": 0.8741956949234009, "num_tokens": 461964462.0, "step": 12165 }, { "epoch": 1.547640249332146, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.485836029052734, "learning_rate": 1e-05, "loss": 0.4789, "mean_token_accuracy": 0.8751291036605835, "num_tokens": 462004123.0, "step": 12166 }, { "epoch": 1.5477674596107365, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.185726165771484, "learning_rate": 1e-05, "loss": 0.5518, "mean_token_accuracy": 0.8493738770484924, "num_tokens": 462042906.0, "step": 12167 }, { "epoch": 1.547894669889327, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.336503505706787, "learning_rate": 1e-05, "loss": 0.4629, "mean_token_accuracy": 0.8800489902496338, "num_tokens": 462083802.0, "step": 12168 }, { "epoch": 1.5480218801679175, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 6.039705753326416, "learning_rate": 1e-05, "loss": 0.5018, "mean_token_accuracy": 0.8641605973243713, "num_tokens": 462119861.0, "step": 12169 }, { "epoch": 1.548149090446508, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.314632415771484, "learning_rate": 1e-05, "loss": 0.5093, "mean_token_accuracy": 0.8613455891609192, "num_tokens": 462162445.0, "step": 12170 }, { "epoch": 1.5482763007250986, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.878702640533447, "learning_rate": 1e-05, "loss": 0.4847, "mean_token_accuracy": 0.8673572540283203, "num_tokens": 462195176.0, "step": 12171 }, { "epoch": 1.5484035110036891, "ewc_loss": 0.11474609375, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.400508403778076, "learning_rate": 1e-05, "loss": 0.5156, "mean_token_accuracy": 0.8620843887329102, "num_tokens": 462236155.0, "step": 12172 }, { "epoch": 1.5485307212822796, "ewc_loss": 0.10302734375, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.734406471252441, "learning_rate": 1e-05, "loss": 0.4855, "mean_token_accuracy": 0.8682852983474731, "num_tokens": 462267099.0, "step": 12173 }, { "epoch": 1.5486579315608702, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.582247257232666, "learning_rate": 1e-05, "loss": 0.5456, "mean_token_accuracy": 0.8566980361938477, "num_tokens": 462311485.0, "step": 12174 }, { "epoch": 1.5487851418394607, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.60169792175293, "learning_rate": 1e-05, "loss": 0.4785, "mean_token_accuracy": 0.8691037893295288, "num_tokens": 462349078.0, "step": 12175 }, { "epoch": 1.5489123521180512, "ewc_loss": 0.125, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.886611461639404, "learning_rate": 1e-05, "loss": 0.4961, "mean_token_accuracy": 0.868752658367157, "num_tokens": 462381477.0, "step": 12176 }, { "epoch": 1.5490395623966418, "ewc_loss": 0.10302734375, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.586968898773193, "learning_rate": 1e-05, "loss": 0.4859, "mean_token_accuracy": 0.8683936595916748, "num_tokens": 462415996.0, "step": 12177 }, { "epoch": 1.5491667726752323, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 0.00011491775512695312, "grad_norm": 7.230303764343262, "learning_rate": 1e-05, "loss": 0.4952, "mean_token_accuracy": 0.8727594614028931, "num_tokens": 462446737.0, "step": 12178 }, { "epoch": 1.5492939829538228, "ewc_loss": 0.11181640625, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 5.83632755279541, "learning_rate": 1e-05, "loss": 0.5031, "mean_token_accuracy": 0.8621829748153687, "num_tokens": 462487742.0, "step": 12179 }, { "epoch": 1.5494211932324133, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 0.0001163482666015625, "grad_norm": 7.296182632446289, "learning_rate": 1e-05, "loss": 0.5069, "mean_token_accuracy": 0.8728229999542236, "num_tokens": 462529231.0, "step": 12180 }, { "epoch": 1.5495484035110036, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.221457481384277, "learning_rate": 1e-05, "loss": 0.4952, "mean_token_accuracy": 0.8712931275367737, "num_tokens": 462565608.0, "step": 12181 }, { "epoch": 1.5496756137895942, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.961318492889404, "learning_rate": 1e-05, "loss": 0.4941, "mean_token_accuracy": 0.8713803291320801, "num_tokens": 462600919.0, "step": 12182 }, { "epoch": 1.5498028240681847, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.400362968444824, "learning_rate": 1e-05, "loss": 0.5096, "mean_token_accuracy": 0.8647341728210449, "num_tokens": 462635594.0, "step": 12183 }, { "epoch": 1.5499300343467752, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.450272560119629, "learning_rate": 1e-05, "loss": 0.5408, "mean_token_accuracy": 0.8553659915924072, "num_tokens": 462672349.0, "step": 12184 }, { "epoch": 1.5500572446253658, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.4697980880737305, "learning_rate": 1e-05, "loss": 0.4908, "mean_token_accuracy": 0.8725245594978333, "num_tokens": 462710286.0, "step": 12185 }, { "epoch": 1.5501844549039563, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.962048053741455, "learning_rate": 1e-05, "loss": 0.4987, "mean_token_accuracy": 0.8668699860572815, "num_tokens": 462744721.0, "step": 12186 }, { "epoch": 1.5503116651825466, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.642759799957275, "learning_rate": 1e-05, "loss": 0.4655, "mean_token_accuracy": 0.8817410469055176, "num_tokens": 462782812.0, "step": 12187 }, { "epoch": 1.5504388754611371, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.629703521728516, "learning_rate": 1e-05, "loss": 0.4577, "mean_token_accuracy": 0.8782663345336914, "num_tokens": 462827919.0, "step": 12188 }, { "epoch": 1.5505660857397277, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.973455429077148, "learning_rate": 1e-05, "loss": 0.5188, "mean_token_accuracy": 0.864903450012207, "num_tokens": 462864878.0, "step": 12189 }, { "epoch": 1.5506932960183182, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.57265567779541, "learning_rate": 1e-05, "loss": 0.4575, "mean_token_accuracy": 0.8774215579032898, "num_tokens": 462906213.0, "step": 12190 }, { "epoch": 1.5508205062969087, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 0.00011730194091796875, "grad_norm": 7.3431291580200195, "learning_rate": 1e-05, "loss": 0.597, "mean_token_accuracy": 0.848913848400116, "num_tokens": 462943078.0, "step": 12191 }, { "epoch": 1.5509477165754992, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 5.878450870513916, "learning_rate": 1e-05, "loss": 0.4662, "mean_token_accuracy": 0.878605306148529, "num_tokens": 462982444.0, "step": 12192 }, { "epoch": 1.5510749268540898, "ewc_loss": 0.140625, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 0.00011920928955078125, "grad_norm": 7.423112392425537, "learning_rate": 1e-05, "loss": 0.498, "mean_token_accuracy": 0.8779469132423401, "num_tokens": 463024335.0, "step": 12193 }, { "epoch": 1.5512021371326803, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.26524019241333, "learning_rate": 1e-05, "loss": 0.4884, "mean_token_accuracy": 0.8709405660629272, "num_tokens": 463062278.0, "step": 12194 }, { "epoch": 1.5513293474112708, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 0.00011014938354492188, "grad_norm": 7.146228790283203, "learning_rate": 1e-05, "loss": 0.5789, "mean_token_accuracy": 0.8473333120346069, "num_tokens": 463100257.0, "step": 12195 }, { "epoch": 1.5514565576898613, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.3857011795043945, "learning_rate": 1e-05, "loss": 0.5034, "mean_token_accuracy": 0.8736083507537842, "num_tokens": 463139468.0, "step": 12196 }, { "epoch": 1.5515837679684519, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.780834674835205, "learning_rate": 1e-05, "loss": 0.5297, "mean_token_accuracy": 0.86299729347229, "num_tokens": 463176145.0, "step": 12197 }, { "epoch": 1.5517109782470424, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.306511878967285, "learning_rate": 1e-05, "loss": 0.4767, "mean_token_accuracy": 0.8775547742843628, "num_tokens": 463208899.0, "step": 12198 }, { "epoch": 1.551838188525633, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.50554084777832, "learning_rate": 1e-05, "loss": 0.4967, "mean_token_accuracy": 0.872565507888794, "num_tokens": 463250163.0, "step": 12199 }, { "epoch": 1.5519653988042235, "ewc_loss": 0.11376953125, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.179286956787109, "learning_rate": 1e-05, "loss": 0.4979, "mean_token_accuracy": 0.8642767667770386, "num_tokens": 463285447.0, "step": 12200 }, { "epoch": 1.552092609082814, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.364494800567627, "learning_rate": 1e-05, "loss": 0.4528, "mean_token_accuracy": 0.8854133486747742, "num_tokens": 463322601.0, "step": 12201 }, { "epoch": 1.5522198193614045, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 6.017462730407715, "learning_rate": 1e-05, "loss": 0.4624, "mean_token_accuracy": 0.8787307143211365, "num_tokens": 463360624.0, "step": 12202 }, { "epoch": 1.552347029639995, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.362270355224609, "learning_rate": 1e-05, "loss": 0.4851, "mean_token_accuracy": 0.8715143799781799, "num_tokens": 463401662.0, "step": 12203 }, { "epoch": 1.5524742399185856, "ewc_loss": 0.10498046875, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.866302967071533, "learning_rate": 1e-05, "loss": 0.488, "mean_token_accuracy": 0.8669512271881104, "num_tokens": 463433999.0, "step": 12204 }, { "epoch": 1.5526014501971759, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.456178665161133, "learning_rate": 1e-05, "loss": 0.5646, "mean_token_accuracy": 0.8509819507598877, "num_tokens": 463473673.0, "step": 12205 }, { "epoch": 1.5527286604757664, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.712610721588135, "learning_rate": 1e-05, "loss": 0.5249, "mean_token_accuracy": 0.8566324710845947, "num_tokens": 463507197.0, "step": 12206 }, { "epoch": 1.552855870754357, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.684169292449951, "learning_rate": 1e-05, "loss": 0.5035, "mean_token_accuracy": 0.8697418570518494, "num_tokens": 463540777.0, "step": 12207 }, { "epoch": 1.5529830810329475, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.596826553344727, "learning_rate": 1e-05, "loss": 0.4177, "mean_token_accuracy": 0.8869386911392212, "num_tokens": 463573335.0, "step": 12208 }, { "epoch": 1.553110291311538, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 7.002546787261963, "learning_rate": 1e-05, "loss": 0.4967, "mean_token_accuracy": 0.8726129531860352, "num_tokens": 463613898.0, "step": 12209 }, { "epoch": 1.5532375015901285, "ewc_loss": 0.10400390625, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.663567543029785, "learning_rate": 1e-05, "loss": 0.4324, "mean_token_accuracy": 0.8815639019012451, "num_tokens": 463648948.0, "step": 12210 }, { "epoch": 1.5533647118687188, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 0.00011444091796875, "grad_norm": 7.262785911560059, "learning_rate": 1e-05, "loss": 0.494, "mean_token_accuracy": 0.8770642876625061, "num_tokens": 463683417.0, "step": 12211 }, { "epoch": 1.5534919221473094, "ewc_loss": 0.11376953125, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 5.985843658447266, "learning_rate": 1e-05, "loss": 0.5366, "mean_token_accuracy": 0.8604006767272949, "num_tokens": 463719533.0, "step": 12212 }, { "epoch": 1.5536191324258999, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 0.00011301040649414062, "grad_norm": 7.173887729644775, "learning_rate": 1e-05, "loss": 0.4782, "mean_token_accuracy": 0.8792822957038879, "num_tokens": 463757740.0, "step": 12213 }, { "epoch": 1.5537463427044904, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.2760233879089355, "learning_rate": 1e-05, "loss": 0.4956, "mean_token_accuracy": 0.8692391514778137, "num_tokens": 463795619.0, "step": 12214 }, { "epoch": 1.553873552983081, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.768301963806152, "learning_rate": 1e-05, "loss": 0.4904, "mean_token_accuracy": 0.8712432384490967, "num_tokens": 463829934.0, "step": 12215 }, { "epoch": 1.5540007632616715, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.397523880004883, "learning_rate": 1e-05, "loss": 0.5607, "mean_token_accuracy": 0.8470834493637085, "num_tokens": 463869584.0, "step": 12216 }, { "epoch": 1.554127973540262, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.256073951721191, "learning_rate": 1e-05, "loss": 0.5379, "mean_token_accuracy": 0.8547468781471252, "num_tokens": 463909220.0, "step": 12217 }, { "epoch": 1.5542551838188525, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.4994001388549805, "learning_rate": 1e-05, "loss": 0.4789, "mean_token_accuracy": 0.8755226731300354, "num_tokens": 463949367.0, "step": 12218 }, { "epoch": 1.554382394097443, "ewc_loss": 0.10400390625, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.820578098297119, "learning_rate": 1e-05, "loss": 0.5169, "mean_token_accuracy": 0.860925555229187, "num_tokens": 463986037.0, "step": 12219 }, { "epoch": 1.5545096043760336, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.7280073165893555, "learning_rate": 1e-05, "loss": 0.4777, "mean_token_accuracy": 0.8778575658798218, "num_tokens": 464021903.0, "step": 12220 }, { "epoch": 1.554636814654624, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.567955493927002, "learning_rate": 1e-05, "loss": 0.4579, "mean_token_accuracy": 0.8748541474342346, "num_tokens": 464058738.0, "step": 12221 }, { "epoch": 1.5547640249332146, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 7.1031718254089355, "learning_rate": 1e-05, "loss": 0.569, "mean_token_accuracy": 0.8514623045921326, "num_tokens": 464090791.0, "step": 12222 }, { "epoch": 1.5548912352118052, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.608544826507568, "learning_rate": 1e-05, "loss": 0.4882, "mean_token_accuracy": 0.8679153919219971, "num_tokens": 464133897.0, "step": 12223 }, { "epoch": 1.5550184454903957, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 0.00012111663818359375, "grad_norm": 7.437162399291992, "learning_rate": 1e-05, "loss": 0.4914, "mean_token_accuracy": 0.877025842666626, "num_tokens": 464173808.0, "step": 12224 }, { "epoch": 1.5551456557689862, "ewc_loss": 0.11474609375, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.006599426269531, "learning_rate": 1e-05, "loss": 0.4667, "mean_token_accuracy": 0.8729627728462219, "num_tokens": 464209456.0, "step": 12225 }, { "epoch": 1.5552728660475768, "ewc_loss": 0.140625, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 0.00011968612670898438, "grad_norm": 7.407037734985352, "learning_rate": 1e-05, "loss": 0.5469, "mean_token_accuracy": 0.8588742017745972, "num_tokens": 464250192.0, "step": 12226 }, { "epoch": 1.5554000763261673, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.355671405792236, "learning_rate": 1e-05, "loss": 0.5697, "mean_token_accuracy": 0.8504889011383057, "num_tokens": 464291485.0, "step": 12227 }, { "epoch": 1.5555272866047578, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 7.062862873077393, "learning_rate": 1e-05, "loss": 0.5566, "mean_token_accuracy": 0.8607379794120789, "num_tokens": 464329963.0, "step": 12228 }, { "epoch": 1.5556544968833483, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.401782989501953, "learning_rate": 1e-05, "loss": 0.4927, "mean_token_accuracy": 0.8717709183692932, "num_tokens": 464366944.0, "step": 12229 }, { "epoch": 1.5557817071619386, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.693365097045898, "learning_rate": 1e-05, "loss": 0.5194, "mean_token_accuracy": 0.8620049953460693, "num_tokens": 464409070.0, "step": 12230 }, { "epoch": 1.5559089174405292, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.309298038482666, "learning_rate": 1e-05, "loss": 0.4885, "mean_token_accuracy": 0.8744650483131409, "num_tokens": 464448305.0, "step": 12231 }, { "epoch": 1.5560361277191197, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.440741539001465, "learning_rate": 1e-05, "loss": 0.5589, "mean_token_accuracy": 0.8519254922866821, "num_tokens": 464482453.0, "step": 12232 }, { "epoch": 1.5561633379977102, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.175049781799316, "learning_rate": 1e-05, "loss": 0.5234, "mean_token_accuracy": 0.8563588857650757, "num_tokens": 464518692.0, "step": 12233 }, { "epoch": 1.5562905482763008, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.32796573638916, "learning_rate": 1e-05, "loss": 0.5101, "mean_token_accuracy": 0.8623743057250977, "num_tokens": 464554502.0, "step": 12234 }, { "epoch": 1.5564177585548913, "ewc_loss": 0.10791015625, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 6.0133233070373535, "learning_rate": 1e-05, "loss": 0.4657, "mean_token_accuracy": 0.8756735324859619, "num_tokens": 464591414.0, "step": 12235 }, { "epoch": 1.5565449688334816, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.326512336730957, "learning_rate": 1e-05, "loss": 0.5251, "mean_token_accuracy": 0.8621060252189636, "num_tokens": 464627116.0, "step": 12236 }, { "epoch": 1.5566721791120721, "ewc_loss": 0.10498046875, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.865679740905762, "learning_rate": 1e-05, "loss": 0.4913, "mean_token_accuracy": 0.8683815598487854, "num_tokens": 464661783.0, "step": 12237 }, { "epoch": 1.5567993893906626, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.424259662628174, "learning_rate": 1e-05, "loss": 0.5405, "mean_token_accuracy": 0.8580042123794556, "num_tokens": 464704851.0, "step": 12238 }, { "epoch": 1.5569265996692532, "ewc_loss": 0.10302734375, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.70461893081665, "learning_rate": 1e-05, "loss": 0.5218, "mean_token_accuracy": 0.8597098588943481, "num_tokens": 464746583.0, "step": 12239 }, { "epoch": 1.5570538099478437, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.641695976257324, "learning_rate": 1e-05, "loss": 0.5215, "mean_token_accuracy": 0.8628975749015808, "num_tokens": 464783275.0, "step": 12240 }, { "epoch": 1.5571810202264342, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.58943510055542, "learning_rate": 1e-05, "loss": 0.4675, "mean_token_accuracy": 0.8748267292976379, "num_tokens": 464818436.0, "step": 12241 }, { "epoch": 1.5573082305050248, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.962114334106445, "learning_rate": 1e-05, "loss": 0.5122, "mean_token_accuracy": 0.8667476773262024, "num_tokens": 464859708.0, "step": 12242 }, { "epoch": 1.5574354407836153, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.626041889190674, "learning_rate": 1e-05, "loss": 0.4856, "mean_token_accuracy": 0.8728318810462952, "num_tokens": 464894339.0, "step": 12243 }, { "epoch": 1.5575626510622058, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 0.00011491775512695312, "grad_norm": 7.2689948081970215, "learning_rate": 1e-05, "loss": 0.5032, "mean_token_accuracy": 0.8758705854415894, "num_tokens": 464931517.0, "step": 12244 }, { "epoch": 1.5576898613407963, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 5.901932239532471, "learning_rate": 1e-05, "loss": 0.515, "mean_token_accuracy": 0.8663918375968933, "num_tokens": 464971616.0, "step": 12245 }, { "epoch": 1.5578170716193869, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 0.00011587142944335938, "grad_norm": 7.2848687171936035, "learning_rate": 1e-05, "loss": 0.5013, "mean_token_accuracy": 0.8759671449661255, "num_tokens": 465012551.0, "step": 12246 }, { "epoch": 1.5579442818979774, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.259893894195557, "learning_rate": 1e-05, "loss": 0.4786, "mean_token_accuracy": 0.8780919313430786, "num_tokens": 465054266.0, "step": 12247 }, { "epoch": 1.558071492176568, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.923274040222168, "learning_rate": 1e-05, "loss": 0.4757, "mean_token_accuracy": 0.8748384714126587, "num_tokens": 465090057.0, "step": 12248 }, { "epoch": 1.5581987024551585, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.3939008712768555, "learning_rate": 1e-05, "loss": 0.4458, "mean_token_accuracy": 0.8861716389656067, "num_tokens": 465132169.0, "step": 12249 }, { "epoch": 1.558325912733749, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.424002647399902, "learning_rate": 1e-05, "loss": 0.516, "mean_token_accuracy": 0.8662062883377075, "num_tokens": 465169922.0, "step": 12250 }, { "epoch": 1.5584531230123395, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.46389102935791, "learning_rate": 1e-05, "loss": 0.4567, "mean_token_accuracy": 0.8819771409034729, "num_tokens": 465207792.0, "step": 12251 }, { "epoch": 1.55858033329093, "ewc_loss": 0.10595703125, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.961562156677246, "learning_rate": 1e-05, "loss": 0.5093, "mean_token_accuracy": 0.8588758707046509, "num_tokens": 465250498.0, "step": 12252 }, { "epoch": 1.5587075435695206, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.627881050109863, "learning_rate": 1e-05, "loss": 0.5316, "mean_token_accuracy": 0.8641912937164307, "num_tokens": 465288321.0, "step": 12253 }, { "epoch": 1.5588347538481109, "ewc_loss": 0.10302734375, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.64941930770874, "learning_rate": 1e-05, "loss": 0.3941, "mean_token_accuracy": 0.8952492475509644, "num_tokens": 465322241.0, "step": 12254 }, { "epoch": 1.5589619641267014, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.926170349121094, "learning_rate": 1e-05, "loss": 0.4746, "mean_token_accuracy": 0.8830088973045349, "num_tokens": 465359680.0, "step": 12255 }, { "epoch": 1.559089174405292, "ewc_loss": 0.10302734375, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.564029693603516, "learning_rate": 1e-05, "loss": 0.4764, "mean_token_accuracy": 0.8751487731933594, "num_tokens": 465396951.0, "step": 12256 }, { "epoch": 1.5592163846838825, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 0.0001163482666015625, "grad_norm": 7.3078789710998535, "learning_rate": 1e-05, "loss": 0.5383, "mean_token_accuracy": 0.8666421175003052, "num_tokens": 465433087.0, "step": 12257 }, { "epoch": 1.559343594962473, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 5.809042453765869, "learning_rate": 1e-05, "loss": 0.4775, "mean_token_accuracy": 0.8759832382202148, "num_tokens": 465472810.0, "step": 12258 }, { "epoch": 1.5594708052410635, "ewc_loss": 0.140625, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 0.00011920928955078125, "grad_norm": 7.451056957244873, "learning_rate": 1e-05, "loss": 0.556, "mean_token_accuracy": 0.8650646209716797, "num_tokens": 465512714.0, "step": 12259 }, { "epoch": 1.5595980155196538, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.220580577850342, "learning_rate": 1e-05, "loss": 0.5438, "mean_token_accuracy": 0.8541511297225952, "num_tokens": 465547629.0, "step": 12260 }, { "epoch": 1.5597252257982444, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 0.00011301040649414062, "grad_norm": 7.210489273071289, "learning_rate": 1e-05, "loss": 0.5303, "mean_token_accuracy": 0.8637250065803528, "num_tokens": 465584226.0, "step": 12261 }, { "epoch": 1.5598524360768349, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.381818771362305, "learning_rate": 1e-05, "loss": 0.5505, "mean_token_accuracy": 0.8561451435089111, "num_tokens": 465625854.0, "step": 12262 }, { "epoch": 1.5599796463554254, "ewc_loss": 0.125, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.842815399169922, "learning_rate": 1e-05, "loss": 0.5551, "mean_token_accuracy": 0.8503153324127197, "num_tokens": 465660670.0, "step": 12263 }, { "epoch": 1.560106856634016, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.328981876373291, "learning_rate": 1e-05, "loss": 0.5045, "mean_token_accuracy": 0.8674455881118774, "num_tokens": 465700359.0, "step": 12264 }, { "epoch": 1.5602340669126065, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.533987998962402, "learning_rate": 1e-05, "loss": 0.4969, "mean_token_accuracy": 0.872603178024292, "num_tokens": 465733648.0, "step": 12265 }, { "epoch": 1.560361277191197, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.213098526000977, "learning_rate": 1e-05, "loss": 0.5125, "mean_token_accuracy": 0.8619687557220459, "num_tokens": 465769814.0, "step": 12266 }, { "epoch": 1.5604884874697875, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.3735880851745605, "learning_rate": 1e-05, "loss": 0.4696, "mean_token_accuracy": 0.8749836683273315, "num_tokens": 465804637.0, "step": 12267 }, { "epoch": 1.560615697748378, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.058643341064453, "learning_rate": 1e-05, "loss": 0.448, "mean_token_accuracy": 0.8816933631896973, "num_tokens": 465838649.0, "step": 12268 }, { "epoch": 1.5607429080269686, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.351343631744385, "learning_rate": 1e-05, "loss": 0.4621, "mean_token_accuracy": 0.8815522193908691, "num_tokens": 465870628.0, "step": 12269 }, { "epoch": 1.560870118305559, "ewc_loss": 0.10498046875, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.889620780944824, "learning_rate": 1e-05, "loss": 0.4958, "mean_token_accuracy": 0.8629238605499268, "num_tokens": 465905612.0, "step": 12270 }, { "epoch": 1.5609973285841496, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.439499378204346, "learning_rate": 1e-05, "loss": 0.4902, "mean_token_accuracy": 0.867423415184021, "num_tokens": 465945964.0, "step": 12271 }, { "epoch": 1.5611245388627402, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.728421211242676, "learning_rate": 1e-05, "loss": 0.4474, "mean_token_accuracy": 0.8831388354301453, "num_tokens": 465979600.0, "step": 12272 }, { "epoch": 1.5612517491413307, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.6366353034973145, "learning_rate": 1e-05, "loss": 0.5079, "mean_token_accuracy": 0.8634082078933716, "num_tokens": 466017406.0, "step": 12273 }, { "epoch": 1.5613789594199212, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.602381229400635, "learning_rate": 1e-05, "loss": 0.5058, "mean_token_accuracy": 0.8648136258125305, "num_tokens": 466059178.0, "step": 12274 }, { "epoch": 1.5615061696985117, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.97350549697876, "learning_rate": 1e-05, "loss": 0.5197, "mean_token_accuracy": 0.8657435178756714, "num_tokens": 466096385.0, "step": 12275 }, { "epoch": 1.5616333799771023, "ewc_loss": 0.10302734375, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.63270378112793, "learning_rate": 1e-05, "loss": 0.4922, "mean_token_accuracy": 0.8684254288673401, "num_tokens": 466131160.0, "step": 12276 }, { "epoch": 1.5617605902556928, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 0.00011491775512695312, "grad_norm": 7.266688346862793, "learning_rate": 1e-05, "loss": 0.5105, "mean_token_accuracy": 0.8712772130966187, "num_tokens": 466174842.0, "step": 12277 }, { "epoch": 1.5618878005342833, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 5.911008358001709, "learning_rate": 1e-05, "loss": 0.475, "mean_token_accuracy": 0.8753331899642944, "num_tokens": 466212943.0, "step": 12278 }, { "epoch": 1.5620150108128736, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 0.00011587142944335938, "grad_norm": 7.27876091003418, "learning_rate": 1e-05, "loss": 0.4742, "mean_token_accuracy": 0.8838088512420654, "num_tokens": 466251560.0, "step": 12279 }, { "epoch": 1.5621422210914642, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.2691826820373535, "learning_rate": 1e-05, "loss": 0.5309, "mean_token_accuracy": 0.8578230142593384, "num_tokens": 466287150.0, "step": 12280 }, { "epoch": 1.5622694313700547, "ewc_loss": 0.125, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.886938571929932, "learning_rate": 1e-05, "loss": 0.5208, "mean_token_accuracy": 0.8676281571388245, "num_tokens": 466328508.0, "step": 12281 }, { "epoch": 1.5623966416486452, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.428440093994141, "learning_rate": 1e-05, "loss": 0.5112, "mean_token_accuracy": 0.864246129989624, "num_tokens": 466362552.0, "step": 12282 }, { "epoch": 1.5625238519272358, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.345398902893066, "learning_rate": 1e-05, "loss": 0.5093, "mean_token_accuracy": 0.865106999874115, "num_tokens": 466404253.0, "step": 12283 }, { "epoch": 1.5626510622058263, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.511988162994385, "learning_rate": 1e-05, "loss": 0.5292, "mean_token_accuracy": 0.8598839640617371, "num_tokens": 466443258.0, "step": 12284 }, { "epoch": 1.5627782724844166, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.877896785736084, "learning_rate": 1e-05, "loss": 0.4963, "mean_token_accuracy": 0.8678240776062012, "num_tokens": 466477575.0, "step": 12285 }, { "epoch": 1.5629054827630071, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.714869976043701, "learning_rate": 1e-05, "loss": 0.5069, "mean_token_accuracy": 0.8654266595840454, "num_tokens": 466516301.0, "step": 12286 }, { "epoch": 1.5630326930415976, "ewc_loss": 0.10205078125, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.599440574645996, "learning_rate": 1e-05, "loss": 0.5158, "mean_token_accuracy": 0.8577449917793274, "num_tokens": 466553904.0, "step": 12287 }, { "epoch": 1.5631599033201882, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 7.071476459503174, "learning_rate": 1e-05, "loss": 0.5268, "mean_token_accuracy": 0.8638603687286377, "num_tokens": 466591668.0, "step": 12288 }, { "epoch": 1.5632871135987787, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.600607395172119, "learning_rate": 1e-05, "loss": 0.4799, "mean_token_accuracy": 0.8665850162506104, "num_tokens": 466626484.0, "step": 12289 }, { "epoch": 1.5634143238773692, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 0.00012111663818359375, "grad_norm": 7.422126293182373, "learning_rate": 1e-05, "loss": 0.49, "mean_token_accuracy": 0.8777784705162048, "num_tokens": 466663590.0, "step": 12290 }, { "epoch": 1.5635415341559598, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 5.964728832244873, "learning_rate": 1e-05, "loss": 0.4603, "mean_token_accuracy": 0.8822659254074097, "num_tokens": 466699856.0, "step": 12291 }, { "epoch": 1.5636687444345503, "ewc_loss": 0.140625, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 0.00011920928955078125, "grad_norm": 7.419617652893066, "learning_rate": 1e-05, "loss": 0.484, "mean_token_accuracy": 0.8835134506225586, "num_tokens": 466738703.0, "step": 12292 }, { "epoch": 1.5637959547131408, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.337698459625244, "learning_rate": 1e-05, "loss": 0.4829, "mean_token_accuracy": 0.8739882707595825, "num_tokens": 466770948.0, "step": 12293 }, { "epoch": 1.5639231649917313, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 7.092692852020264, "learning_rate": 1e-05, "loss": 0.5027, "mean_token_accuracy": 0.871946394443512, "num_tokens": 466804655.0, "step": 12294 }, { "epoch": 1.5640503752703219, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.400738716125488, "learning_rate": 1e-05, "loss": 0.4854, "mean_token_accuracy": 0.875336766242981, "num_tokens": 466847592.0, "step": 12295 }, { "epoch": 1.5641775855489124, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.715460300445557, "learning_rate": 1e-05, "loss": 0.4827, "mean_token_accuracy": 0.8768742084503174, "num_tokens": 466887855.0, "step": 12296 }, { "epoch": 1.564304795827503, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.313955307006836, "learning_rate": 1e-05, "loss": 0.5096, "mean_token_accuracy": 0.8647080063819885, "num_tokens": 466925749.0, "step": 12297 }, { "epoch": 1.5644320061060935, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.444270610809326, "learning_rate": 1e-05, "loss": 0.4246, "mean_token_accuracy": 0.8927791118621826, "num_tokens": 466966487.0, "step": 12298 }, { "epoch": 1.564559216384684, "ewc_loss": 0.11474609375, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.183349132537842, "learning_rate": 1e-05, "loss": 0.5758, "mean_token_accuracy": 0.8457300662994385, "num_tokens": 467008673.0, "step": 12299 }, { "epoch": 1.5646864266632745, "ewc_loss": 0.11376953125, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.321215629577637, "learning_rate": 1e-05, "loss": 0.4838, "mean_token_accuracy": 0.8758633136749268, "num_tokens": 467048389.0, "step": 12300 }, { "epoch": 1.564813636941865, "ewc_loss": 0.10791015625, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 6.025786399841309, "learning_rate": 1e-05, "loss": 0.4707, "mean_token_accuracy": 0.8745367527008057, "num_tokens": 467087059.0, "step": 12301 }, { "epoch": 1.5649408472204556, "ewc_loss": 0.11376953125, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.323831558227539, "learning_rate": 1e-05, "loss": 0.4405, "mean_token_accuracy": 0.887397289276123, "num_tokens": 467123003.0, "step": 12302 }, { "epoch": 1.5650680574990459, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.871665954589844, "learning_rate": 1e-05, "loss": 0.5254, "mean_token_accuracy": 0.8576297760009766, "num_tokens": 467160526.0, "step": 12303 }, { "epoch": 1.5651952677776364, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.419533729553223, "learning_rate": 1e-05, "loss": 0.4967, "mean_token_accuracy": 0.8683173656463623, "num_tokens": 467195846.0, "step": 12304 }, { "epoch": 1.565322478056227, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.7178120613098145, "learning_rate": 1e-05, "loss": 0.522, "mean_token_accuracy": 0.8583798408508301, "num_tokens": 467232412.0, "step": 12305 }, { "epoch": 1.5654496883348175, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.614111423492432, "learning_rate": 1e-05, "loss": 0.5167, "mean_token_accuracy": 0.8648426532745361, "num_tokens": 467269695.0, "step": 12306 }, { "epoch": 1.565576898613408, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.589460849761963, "learning_rate": 1e-05, "loss": 0.5284, "mean_token_accuracy": 0.8546576499938965, "num_tokens": 467304395.0, "step": 12307 }, { "epoch": 1.5657041088919985, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.937328338623047, "learning_rate": 1e-05, "loss": 0.5014, "mean_token_accuracy": 0.8679492473602295, "num_tokens": 467348606.0, "step": 12308 }, { "epoch": 1.5658313191705888, "ewc_loss": 0.10302734375, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.583435535430908, "learning_rate": 1e-05, "loss": 0.4862, "mean_token_accuracy": 0.868313729763031, "num_tokens": 467382673.0, "step": 12309 }, { "epoch": 1.5659585294491793, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 0.00011587142944335938, "grad_norm": 7.259486675262451, "learning_rate": 1e-05, "loss": 0.4935, "mean_token_accuracy": 0.8760049343109131, "num_tokens": 467426291.0, "step": 12310 }, { "epoch": 1.5660857397277699, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 5.8590545654296875, "learning_rate": 1e-05, "loss": 0.4979, "mean_token_accuracy": 0.8667385578155518, "num_tokens": 467469299.0, "step": 12311 }, { "epoch": 1.5662129500063604, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 0.0001163482666015625, "grad_norm": 7.307341575622559, "learning_rate": 1e-05, "loss": 0.5221, "mean_token_accuracy": 0.8635848760604858, "num_tokens": 467510876.0, "step": 12312 }, { "epoch": 1.566340160284951, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.236825466156006, "learning_rate": 1e-05, "loss": 0.5028, "mean_token_accuracy": 0.8695241212844849, "num_tokens": 467546112.0, "step": 12313 }, { "epoch": 1.5664673705635415, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.965122699737549, "learning_rate": 1e-05, "loss": 0.4867, "mean_token_accuracy": 0.8723492622375488, "num_tokens": 467580009.0, "step": 12314 }, { "epoch": 1.566594580842132, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.394461631774902, "learning_rate": 1e-05, "loss": 0.5219, "mean_token_accuracy": 0.8619788885116577, "num_tokens": 467621009.0, "step": 12315 }, { "epoch": 1.5667217911207225, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.462644100189209, "learning_rate": 1e-05, "loss": 0.4693, "mean_token_accuracy": 0.87993323802948, "num_tokens": 467662304.0, "step": 12316 }, { "epoch": 1.566849001399313, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.458113193511963, "learning_rate": 1e-05, "loss": 0.5531, "mean_token_accuracy": 0.8538229465484619, "num_tokens": 467697608.0, "step": 12317 }, { "epoch": 1.5669762116779036, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.9978132247924805, "learning_rate": 1e-05, "loss": 0.485, "mean_token_accuracy": 0.8717516660690308, "num_tokens": 467739978.0, "step": 12318 }, { "epoch": 1.567103421956494, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.606436729431152, "learning_rate": 1e-05, "loss": 0.5316, "mean_token_accuracy": 0.8584973812103271, "num_tokens": 467775055.0, "step": 12319 }, { "epoch": 1.5672306322350846, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.685304164886475, "learning_rate": 1e-05, "loss": 0.4668, "mean_token_accuracy": 0.8731729984283447, "num_tokens": 467812385.0, "step": 12320 }, { "epoch": 1.5673578425136752, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.897357940673828, "learning_rate": 1e-05, "loss": 0.5545, "mean_token_accuracy": 0.8533153533935547, "num_tokens": 467846784.0, "step": 12321 }, { "epoch": 1.5674850527922657, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.587217807769775, "learning_rate": 1e-05, "loss": 0.5022, "mean_token_accuracy": 0.865886926651001, "num_tokens": 467880259.0, "step": 12322 }, { "epoch": 1.5676122630708562, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 0.00011491775512695312, "grad_norm": 7.2352824211120605, "learning_rate": 1e-05, "loss": 0.5251, "mean_token_accuracy": 0.8697290420532227, "num_tokens": 467914613.0, "step": 12323 }, { "epoch": 1.5677394733494467, "ewc_loss": 0.10986328125, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 5.785862922668457, "learning_rate": 1e-05, "loss": 0.5614, "mean_token_accuracy": 0.8456186056137085, "num_tokens": 467953070.0, "step": 12324 }, { "epoch": 1.5678666836280373, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 0.000118255615234375, "grad_norm": 7.380927085876465, "learning_rate": 1e-05, "loss": 0.5624, "mean_token_accuracy": 0.85665363073349, "num_tokens": 467989404.0, "step": 12325 }, { "epoch": 1.5679938939066278, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.125185489654541, "learning_rate": 1e-05, "loss": 0.4974, "mean_token_accuracy": 0.8672405481338501, "num_tokens": 468030670.0, "step": 12326 }, { "epoch": 1.5681211041852183, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 0.00011301040649414062, "grad_norm": 7.214619159698486, "learning_rate": 1e-05, "loss": 0.5308, "mean_token_accuracy": 0.8671833276748657, "num_tokens": 468064130.0, "step": 12327 }, { "epoch": 1.5682483144638086, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.318819999694824, "learning_rate": 1e-05, "loss": 0.5077, "mean_token_accuracy": 0.8687567114830017, "num_tokens": 468098393.0, "step": 12328 }, { "epoch": 1.5683755247423992, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.882940292358398, "learning_rate": 1e-05, "loss": 0.4781, "mean_token_accuracy": 0.8798721432685852, "num_tokens": 468134961.0, "step": 12329 }, { "epoch": 1.5685027350209897, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.306731224060059, "learning_rate": 1e-05, "loss": 0.4613, "mean_token_accuracy": 0.8796483278274536, "num_tokens": 468169389.0, "step": 12330 }, { "epoch": 1.5686299452995802, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.576656341552734, "learning_rate": 1e-05, "loss": 0.5242, "mean_token_accuracy": 0.8598968982696533, "num_tokens": 468211504.0, "step": 12331 }, { "epoch": 1.5687571555781707, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.224849224090576, "learning_rate": 1e-05, "loss": 0.5307, "mean_token_accuracy": 0.856229305267334, "num_tokens": 468245204.0, "step": 12332 }, { "epoch": 1.5688843658567613, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.384287357330322, "learning_rate": 1e-05, "loss": 0.4956, "mean_token_accuracy": 0.8706651926040649, "num_tokens": 468277054.0, "step": 12333 }, { "epoch": 1.5690115761353516, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.0814208984375, "learning_rate": 1e-05, "loss": 0.4619, "mean_token_accuracy": 0.8827258348464966, "num_tokens": 468316399.0, "step": 12334 }, { "epoch": 1.569138786413942, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.3269805908203125, "learning_rate": 1e-05, "loss": 0.4713, "mean_token_accuracy": 0.8761012554168701, "num_tokens": 468351943.0, "step": 12335 }, { "epoch": 1.5692659966925326, "ewc_loss": 0.10791015625, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.945766925811768, "learning_rate": 1e-05, "loss": 0.4886, "mean_token_accuracy": 0.8676979541778564, "num_tokens": 468387443.0, "step": 12336 }, { "epoch": 1.5693932069711232, "ewc_loss": 0.11474609375, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.369699001312256, "learning_rate": 1e-05, "loss": 0.4836, "mean_token_accuracy": 0.8703875541687012, "num_tokens": 468420503.0, "step": 12337 }, { "epoch": 1.5695204172497137, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.8001389503479, "learning_rate": 1e-05, "loss": 0.5059, "mean_token_accuracy": 0.8633527159690857, "num_tokens": 468456171.0, "step": 12338 }, { "epoch": 1.5696476275283042, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.508511066436768, "learning_rate": 1e-05, "loss": 0.4613, "mean_token_accuracy": 0.8804365992546082, "num_tokens": 468494121.0, "step": 12339 }, { "epoch": 1.5697748378068948, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.647639274597168, "learning_rate": 1e-05, "loss": 0.4816, "mean_token_accuracy": 0.8702317476272583, "num_tokens": 468540960.0, "step": 12340 }, { "epoch": 1.5699020480854853, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.780373573303223, "learning_rate": 1e-05, "loss": 0.5211, "mean_token_accuracy": 0.8658565282821655, "num_tokens": 468577851.0, "step": 12341 }, { "epoch": 1.5700292583640758, "ewc_loss": 0.10205078125, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.586424827575684, "learning_rate": 1e-05, "loss": 0.4915, "mean_token_accuracy": 0.8676284551620483, "num_tokens": 468613210.0, "step": 12342 }, { "epoch": 1.5701564686426663, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 0.00011110305786132812, "grad_norm": 7.113326072692871, "learning_rate": 1e-05, "loss": 0.5224, "mean_token_accuracy": 0.8652360439300537, "num_tokens": 468655871.0, "step": 12343 }, { "epoch": 1.5702836789212569, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.725069046020508, "learning_rate": 1e-05, "loss": 0.4545, "mean_token_accuracy": 0.8775887489318848, "num_tokens": 468692757.0, "step": 12344 }, { "epoch": 1.5704108891998474, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 0.0001163482666015625, "grad_norm": 7.307850360870361, "learning_rate": 1e-05, "loss": 0.4936, "mean_token_accuracy": 0.8773308992385864, "num_tokens": 468730163.0, "step": 12345 }, { "epoch": 1.570538099478438, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.067929267883301, "learning_rate": 1e-05, "loss": 0.4893, "mean_token_accuracy": 0.8701156377792358, "num_tokens": 468768578.0, "step": 12346 }, { "epoch": 1.5706653097570284, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 0.0001125335693359375, "grad_norm": 7.131808280944824, "learning_rate": 1e-05, "loss": 0.4657, "mean_token_accuracy": 0.8818682432174683, "num_tokens": 468802027.0, "step": 12347 }, { "epoch": 1.570792520035619, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.338655948638916, "learning_rate": 1e-05, "loss": 0.5357, "mean_token_accuracy": 0.8578661680221558, "num_tokens": 468845266.0, "step": 12348 }, { "epoch": 1.5709197303142095, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.68092679977417, "learning_rate": 1e-05, "loss": 0.517, "mean_token_accuracy": 0.8730361461639404, "num_tokens": 468879780.0, "step": 12349 }, { "epoch": 1.5710469405928, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.418508529663086, "learning_rate": 1e-05, "loss": 0.4918, "mean_token_accuracy": 0.8726856112480164, "num_tokens": 468916846.0, "step": 12350 }, { "epoch": 1.5711741508713906, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.178134918212891, "learning_rate": 1e-05, "loss": 0.4595, "mean_token_accuracy": 0.8812800645828247, "num_tokens": 468947382.0, "step": 12351 }, { "epoch": 1.5713013611499809, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.518985271453857, "learning_rate": 1e-05, "loss": 0.4504, "mean_token_accuracy": 0.8828057050704956, "num_tokens": 468987258.0, "step": 12352 }, { "epoch": 1.5714285714285714, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.7782206535339355, "learning_rate": 1e-05, "loss": 0.4644, "mean_token_accuracy": 0.875829815864563, "num_tokens": 469024878.0, "step": 12353 }, { "epoch": 1.571555781707162, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.759326934814453, "learning_rate": 1e-05, "loss": 0.5017, "mean_token_accuracy": 0.8719673156738281, "num_tokens": 469065636.0, "step": 12354 }, { "epoch": 1.5716829919857525, "ewc_loss": 0.10205078125, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.573134422302246, "learning_rate": 1e-05, "loss": 0.4761, "mean_token_accuracy": 0.8677468299865723, "num_tokens": 469099817.0, "step": 12355 }, { "epoch": 1.571810202264343, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 0.00011110305786132812, "grad_norm": 7.131049156188965, "learning_rate": 1e-05, "loss": 0.5098, "mean_token_accuracy": 0.8659666180610657, "num_tokens": 469135185.0, "step": 12356 }, { "epoch": 1.5719374125429335, "ewc_loss": 0.10498046875, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.640020370483398, "learning_rate": 1e-05, "loss": 0.4297, "mean_token_accuracy": 0.8853586912155151, "num_tokens": 469173980.0, "step": 12357 }, { "epoch": 1.5720646228215238, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 0.0001201629638671875, "grad_norm": 7.430845737457275, "learning_rate": 1e-05, "loss": 0.5113, "mean_token_accuracy": 0.8726025819778442, "num_tokens": 469206939.0, "step": 12358 }, { "epoch": 1.5721918331001143, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.031726360321045, "learning_rate": 1e-05, "loss": 0.4725, "mean_token_accuracy": 0.8766972422599792, "num_tokens": 469245261.0, "step": 12359 }, { "epoch": 1.5723190433787049, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 0.000118255615234375, "grad_norm": 7.361950397491455, "learning_rate": 1e-05, "loss": 0.52, "mean_token_accuracy": 0.8675850629806519, "num_tokens": 469285569.0, "step": 12360 }, { "epoch": 1.5724462536572954, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.348194122314453, "learning_rate": 1e-05, "loss": 0.509, "mean_token_accuracy": 0.8651580810546875, "num_tokens": 469325249.0, "step": 12361 }, { "epoch": 1.572573463935886, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 7.018029689788818, "learning_rate": 1e-05, "loss": 0.5229, "mean_token_accuracy": 0.8608695268630981, "num_tokens": 469360187.0, "step": 12362 }, { "epoch": 1.5727006742144765, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.3729567527771, "learning_rate": 1e-05, "loss": 0.5137, "mean_token_accuracy": 0.865322470664978, "num_tokens": 469399363.0, "step": 12363 }, { "epoch": 1.572827884493067, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.675267219543457, "learning_rate": 1e-05, "loss": 0.5175, "mean_token_accuracy": 0.8668794631958008, "num_tokens": 469441523.0, "step": 12364 }, { "epoch": 1.5729550947716575, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.2706122398376465, "learning_rate": 1e-05, "loss": 0.4682, "mean_token_accuracy": 0.8786512613296509, "num_tokens": 469475169.0, "step": 12365 }, { "epoch": 1.573082305050248, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.446662902832031, "learning_rate": 1e-05, "loss": 0.5133, "mean_token_accuracy": 0.8654651045799255, "num_tokens": 469509794.0, "step": 12366 }, { "epoch": 1.5732095153288386, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.13664436340332, "learning_rate": 1e-05, "loss": 0.5054, "mean_token_accuracy": 0.8645407557487488, "num_tokens": 469549007.0, "step": 12367 }, { "epoch": 1.573336725607429, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.336759090423584, "learning_rate": 1e-05, "loss": 0.4814, "mean_token_accuracy": 0.872983992099762, "num_tokens": 469586959.0, "step": 12368 }, { "epoch": 1.5734639358860196, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.984164237976074, "learning_rate": 1e-05, "loss": 0.4581, "mean_token_accuracy": 0.8767154216766357, "num_tokens": 469617767.0, "step": 12369 }, { "epoch": 1.5735911461646102, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.365379333496094, "learning_rate": 1e-05, "loss": 0.5012, "mean_token_accuracy": 0.8653191924095154, "num_tokens": 469655300.0, "step": 12370 }, { "epoch": 1.5737183564432007, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.822691917419434, "learning_rate": 1e-05, "loss": 0.4751, "mean_token_accuracy": 0.8717430233955383, "num_tokens": 469693666.0, "step": 12371 }, { "epoch": 1.5738455667217912, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.503359317779541, "learning_rate": 1e-05, "loss": 0.4922, "mean_token_accuracy": 0.8717232942581177, "num_tokens": 469731081.0, "step": 12372 }, { "epoch": 1.5739727770003817, "ewc_loss": 0.10302734375, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.680996894836426, "learning_rate": 1e-05, "loss": 0.4572, "mean_token_accuracy": 0.8753764629364014, "num_tokens": 469768081.0, "step": 12373 }, { "epoch": 1.5740999872789723, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.748532295227051, "learning_rate": 1e-05, "loss": 0.5081, "mean_token_accuracy": 0.869425356388092, "num_tokens": 469806992.0, "step": 12374 }, { "epoch": 1.5742271975575628, "ewc_loss": 0.10302734375, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.600100994110107, "learning_rate": 1e-05, "loss": 0.494, "mean_token_accuracy": 0.8655087351799011, "num_tokens": 469844969.0, "step": 12375 }, { "epoch": 1.5743544078361533, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 7.074840068817139, "learning_rate": 1e-05, "loss": 0.4884, "mean_token_accuracy": 0.876413881778717, "num_tokens": 469881558.0, "step": 12376 }, { "epoch": 1.5744816181147436, "ewc_loss": 0.10498046875, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.712108135223389, "learning_rate": 1e-05, "loss": 0.4856, "mean_token_accuracy": 0.8692876696586609, "num_tokens": 469914659.0, "step": 12377 }, { "epoch": 1.5746088283933342, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 0.00011587142944335938, "grad_norm": 7.289909362792969, "learning_rate": 1e-05, "loss": 0.5508, "mean_token_accuracy": 0.8642470836639404, "num_tokens": 469956012.0, "step": 12378 }, { "epoch": 1.5747360386719247, "ewc_loss": 0.11474609375, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.029052257537842, "learning_rate": 1e-05, "loss": 0.5102, "mean_token_accuracy": 0.8644591569900513, "num_tokens": 469993840.0, "step": 12379 }, { "epoch": 1.5748632489505152, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 0.00011205673217773438, "grad_norm": 7.149866580963135, "learning_rate": 1e-05, "loss": 0.5416, "mean_token_accuracy": 0.8584350943565369, "num_tokens": 470032741.0, "step": 12380 }, { "epoch": 1.5749904592291057, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.311461448669434, "learning_rate": 1e-05, "loss": 0.4576, "mean_token_accuracy": 0.8809396028518677, "num_tokens": 470068969.0, "step": 12381 }, { "epoch": 1.5751176695076963, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.714083671569824, "learning_rate": 1e-05, "loss": 0.4899, "mean_token_accuracy": 0.8754540681838989, "num_tokens": 470108972.0, "step": 12382 }, { "epoch": 1.5752448797862866, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.4021077156066895, "learning_rate": 1e-05, "loss": 0.4479, "mean_token_accuracy": 0.8848472833633423, "num_tokens": 470146655.0, "step": 12383 }, { "epoch": 1.575372090064877, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.212874889373779, "learning_rate": 1e-05, "loss": 0.5043, "mean_token_accuracy": 0.8682528138160706, "num_tokens": 470184068.0, "step": 12384 }, { "epoch": 1.5754993003434676, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.517247200012207, "learning_rate": 1e-05, "loss": 0.5059, "mean_token_accuracy": 0.8674124479293823, "num_tokens": 470221328.0, "step": 12385 }, { "epoch": 1.5756265106220582, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.79190731048584, "learning_rate": 1e-05, "loss": 0.4887, "mean_token_accuracy": 0.8680180311203003, "num_tokens": 470261393.0, "step": 12386 }, { "epoch": 1.5757537209006487, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.757951259613037, "learning_rate": 1e-05, "loss": 0.4972, "mean_token_accuracy": 0.8688188195228577, "num_tokens": 470293705.0, "step": 12387 }, { "epoch": 1.5758809311792392, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.571327209472656, "learning_rate": 1e-05, "loss": 0.4837, "mean_token_accuracy": 0.8684908151626587, "num_tokens": 470329891.0, "step": 12388 }, { "epoch": 1.5760081414578297, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 0.00010967254638671875, "grad_norm": 7.117376804351807, "learning_rate": 1e-05, "loss": 0.4972, "mean_token_accuracy": 0.8710983991622925, "num_tokens": 470368763.0, "step": 12389 }, { "epoch": 1.5761353517364203, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.632578372955322, "learning_rate": 1e-05, "loss": 0.4786, "mean_token_accuracy": 0.870552659034729, "num_tokens": 470407267.0, "step": 12390 }, { "epoch": 1.5762625620150108, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 0.0001201629638671875, "grad_norm": 7.4418511390686035, "learning_rate": 1e-05, "loss": 0.5193, "mean_token_accuracy": 0.8675607442855835, "num_tokens": 470448872.0, "step": 12391 }, { "epoch": 1.5763897722936013, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.020840167999268, "learning_rate": 1e-05, "loss": 0.5008, "mean_token_accuracy": 0.8767917156219482, "num_tokens": 470486460.0, "step": 12392 }, { "epoch": 1.5765169825721919, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 0.000118255615234375, "grad_norm": 7.386956691741943, "learning_rate": 1e-05, "loss": 0.4808, "mean_token_accuracy": 0.8808585405349731, "num_tokens": 470528800.0, "step": 12393 }, { "epoch": 1.5766441928507824, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.347715377807617, "learning_rate": 1e-05, "loss": 0.4908, "mean_token_accuracy": 0.8728464245796204, "num_tokens": 470568185.0, "step": 12394 }, { "epoch": 1.576771403129373, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 7.0609130859375, "learning_rate": 1e-05, "loss": 0.5228, "mean_token_accuracy": 0.8683881759643555, "num_tokens": 470607785.0, "step": 12395 }, { "epoch": 1.5768986134079634, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.394624710083008, "learning_rate": 1e-05, "loss": 0.5105, "mean_token_accuracy": 0.8653151392936707, "num_tokens": 470648536.0, "step": 12396 }, { "epoch": 1.577025823686554, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.695133686065674, "learning_rate": 1e-05, "loss": 0.4771, "mean_token_accuracy": 0.8767614960670471, "num_tokens": 470686071.0, "step": 12397 }, { "epoch": 1.5771530339651445, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.315197467803955, "learning_rate": 1e-05, "loss": 0.468, "mean_token_accuracy": 0.8798195123672485, "num_tokens": 470715165.0, "step": 12398 }, { "epoch": 1.577280244243735, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.435990333557129, "learning_rate": 1e-05, "loss": 0.4708, "mean_token_accuracy": 0.8755280375480652, "num_tokens": 470748950.0, "step": 12399 }, { "epoch": 1.5774074545223256, "ewc_loss": 0.11376953125, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.1715850830078125, "learning_rate": 1e-05, "loss": 0.4245, "mean_token_accuracy": 0.8884122371673584, "num_tokens": 470783856.0, "step": 12400 }, { "epoch": 1.5775346648009159, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.3174662590026855, "learning_rate": 1e-05, "loss": 0.5265, "mean_token_accuracy": 0.8573396801948547, "num_tokens": 470823103.0, "step": 12401 }, { "epoch": 1.5776618750795064, "ewc_loss": 0.10791015625, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 6.02070951461792, "learning_rate": 1e-05, "loss": 0.547, "mean_token_accuracy": 0.8507215976715088, "num_tokens": 470865582.0, "step": 12402 }, { "epoch": 1.577789085358097, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.3312883377075195, "learning_rate": 1e-05, "loss": 0.5214, "mean_token_accuracy": 0.859491229057312, "num_tokens": 470901710.0, "step": 12403 }, { "epoch": 1.5779162956366874, "ewc_loss": 0.10498046875, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.8554277420043945, "learning_rate": 1e-05, "loss": 0.5398, "mean_token_accuracy": 0.8510942459106445, "num_tokens": 470944850.0, "step": 12404 }, { "epoch": 1.578043505915278, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.46807336807251, "learning_rate": 1e-05, "loss": 0.5426, "mean_token_accuracy": 0.8601076006889343, "num_tokens": 470985092.0, "step": 12405 }, { "epoch": 1.5781707161938685, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.681875228881836, "learning_rate": 1e-05, "loss": 0.4595, "mean_token_accuracy": 0.8764159679412842, "num_tokens": 471026018.0, "step": 12406 }, { "epoch": 1.5782979264724588, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.721373558044434, "learning_rate": 1e-05, "loss": 0.5109, "mean_token_accuracy": 0.8651220798492432, "num_tokens": 471061365.0, "step": 12407 }, { "epoch": 1.5784251367510493, "ewc_loss": 0.10107421875, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 5.578855514526367, "learning_rate": 1e-05, "loss": 0.47, "mean_token_accuracy": 0.8723469972610474, "num_tokens": 471098142.0, "step": 12408 }, { "epoch": 1.5785523470296399, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 7.077269077301025, "learning_rate": 1e-05, "loss": 0.5533, "mean_token_accuracy": 0.8546456694602966, "num_tokens": 471138001.0, "step": 12409 }, { "epoch": 1.5786795573082304, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.657812118530273, "learning_rate": 1e-05, "loss": 0.4642, "mean_token_accuracy": 0.8735003471374512, "num_tokens": 471170392.0, "step": 12410 }, { "epoch": 1.578806767586821, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 0.00011730194091796875, "grad_norm": 7.378637790679932, "learning_rate": 1e-05, "loss": 0.5392, "mean_token_accuracy": 0.8620806336402893, "num_tokens": 471215704.0, "step": 12411 }, { "epoch": 1.5789339778654115, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.0232110023498535, "learning_rate": 1e-05, "loss": 0.451, "mean_token_accuracy": 0.8839290738105774, "num_tokens": 471252084.0, "step": 12412 }, { "epoch": 1.579061188144002, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 0.00011491775512695312, "grad_norm": 7.302623748779297, "learning_rate": 1e-05, "loss": 0.525, "mean_token_accuracy": 0.8622038960456848, "num_tokens": 471287096.0, "step": 12413 }, { "epoch": 1.5791883984225925, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.366035461425781, "learning_rate": 1e-05, "loss": 0.4507, "mean_token_accuracy": 0.884575605392456, "num_tokens": 471321732.0, "step": 12414 }, { "epoch": 1.579315608701183, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.866539478302002, "learning_rate": 1e-05, "loss": 0.5254, "mean_token_accuracy": 0.8684691190719604, "num_tokens": 471353680.0, "step": 12415 }, { "epoch": 1.5794428189797736, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.454997539520264, "learning_rate": 1e-05, "loss": 0.5564, "mean_token_accuracy": 0.8539118766784668, "num_tokens": 471395286.0, "step": 12416 }, { "epoch": 1.579570029258364, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.35847282409668, "learning_rate": 1e-05, "loss": 0.4652, "mean_token_accuracy": 0.8809537887573242, "num_tokens": 471433708.0, "step": 12417 }, { "epoch": 1.5796972395369546, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.50344181060791, "learning_rate": 1e-05, "loss": 0.5026, "mean_token_accuracy": 0.8689963817596436, "num_tokens": 471472655.0, "step": 12418 }, { "epoch": 1.5798244498155452, "ewc_loss": 0.10595703125, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.937830924987793, "learning_rate": 1e-05, "loss": 0.4822, "mean_token_accuracy": 0.867563009262085, "num_tokens": 471507013.0, "step": 12419 }, { "epoch": 1.5799516600941357, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.662187099456787, "learning_rate": 1e-05, "loss": 0.4883, "mean_token_accuracy": 0.8719906806945801, "num_tokens": 471539951.0, "step": 12420 }, { "epoch": 1.5800788703727262, "ewc_loss": 0.10302734375, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.670113563537598, "learning_rate": 1e-05, "loss": 0.4947, "mean_token_accuracy": 0.8656315803527832, "num_tokens": 471580505.0, "step": 12421 }, { "epoch": 1.5802060806513167, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.925125598907471, "learning_rate": 1e-05, "loss": 0.463, "mean_token_accuracy": 0.88054358959198, "num_tokens": 471610408.0, "step": 12422 }, { "epoch": 1.5803332909299073, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.618198871612549, "learning_rate": 1e-05, "loss": 0.4925, "mean_token_accuracy": 0.8650237321853638, "num_tokens": 471648529.0, "step": 12423 }, { "epoch": 1.5804605012084978, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 0.00011348724365234375, "grad_norm": 7.236542701721191, "learning_rate": 1e-05, "loss": 0.5114, "mean_token_accuracy": 0.8714792728424072, "num_tokens": 471682182.0, "step": 12424 }, { "epoch": 1.5805877114870883, "ewc_loss": 0.109375, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 5.822441577911377, "learning_rate": 1e-05, "loss": 0.4739, "mean_token_accuracy": 0.8739041090011597, "num_tokens": 471719180.0, "step": 12425 }, { "epoch": 1.5807149217656786, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 0.000118255615234375, "grad_norm": 7.346723556518555, "learning_rate": 1e-05, "loss": 0.5214, "mean_token_accuracy": 0.8661485910415649, "num_tokens": 471759965.0, "step": 12426 }, { "epoch": 1.5808421320442692, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.149464130401611, "learning_rate": 1e-05, "loss": 0.4692, "mean_token_accuracy": 0.8801572918891907, "num_tokens": 471795405.0, "step": 12427 }, { "epoch": 1.5809693423228597, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 0.0001125335693359375, "grad_norm": 7.160050392150879, "learning_rate": 1e-05, "loss": 0.4903, "mean_token_accuracy": 0.8752874732017517, "num_tokens": 471837957.0, "step": 12428 }, { "epoch": 1.5810965526014502, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.312307357788086, "learning_rate": 1e-05, "loss": 0.5156, "mean_token_accuracy": 0.8674090504646301, "num_tokens": 471875422.0, "step": 12429 }, { "epoch": 1.5812237628800407, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.831925392150879, "learning_rate": 1e-05, "loss": 0.4847, "mean_token_accuracy": 0.8743004202842712, "num_tokens": 471913752.0, "step": 12430 }, { "epoch": 1.5813509731586313, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.292747974395752, "learning_rate": 1e-05, "loss": 0.516, "mean_token_accuracy": 0.8625326156616211, "num_tokens": 471952362.0, "step": 12431 }, { "epoch": 1.5814781834372216, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.537920951843262, "learning_rate": 1e-05, "loss": 0.4779, "mean_token_accuracy": 0.8774054646492004, "num_tokens": 471990798.0, "step": 12432 }, { "epoch": 1.581605393715812, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.210830211639404, "learning_rate": 1e-05, "loss": 0.4948, "mean_token_accuracy": 0.8684874773025513, "num_tokens": 472027398.0, "step": 12433 }, { "epoch": 1.5817326039944026, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.357574939727783, "learning_rate": 1e-05, "loss": 0.486, "mean_token_accuracy": 0.8739932775497437, "num_tokens": 472071725.0, "step": 12434 }, { "epoch": 1.5818598142729932, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.088258266448975, "learning_rate": 1e-05, "loss": 0.4465, "mean_token_accuracy": 0.8818325996398926, "num_tokens": 472106913.0, "step": 12435 }, { "epoch": 1.5819870245515837, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.415826797485352, "learning_rate": 1e-05, "loss": 0.5003, "mean_token_accuracy": 0.8691831827163696, "num_tokens": 472138136.0, "step": 12436 }, { "epoch": 1.5821142348301742, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.8917975425720215, "learning_rate": 1e-05, "loss": 0.496, "mean_token_accuracy": 0.8651679158210754, "num_tokens": 472174884.0, "step": 12437 }, { "epoch": 1.5822414451087647, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.442512035369873, "learning_rate": 1e-05, "loss": 0.5086, "mean_token_accuracy": 0.8647680878639221, "num_tokens": 472215862.0, "step": 12438 }, { "epoch": 1.5823686553873553, "ewc_loss": 0.10400390625, "ewc_loss_diag": 0.00021648406982421875, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.732812881469727, "learning_rate": 1e-05, "loss": 0.4615, "mean_token_accuracy": 0.8760079145431519, "num_tokens": 472255770.0, "step": 12439 }, { "epoch": 1.5824958656659458, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.00021648406982421875, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.648169994354248, "learning_rate": 1e-05, "loss": 0.5407, "mean_token_accuracy": 0.861376166343689, "num_tokens": 472290815.0, "step": 12440 }, { "epoch": 1.5826230759445363, "ewc_loss": 0.10302734375, "ewc_loss_diag": 0.00021648406982421875, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.624812602996826, "learning_rate": 1e-05, "loss": 0.5139, "mean_token_accuracy": 0.8604454398155212, "num_tokens": 472328180.0, "step": 12441 }, { "epoch": 1.5827502862231269, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.00021648406982421875, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.940439224243164, "learning_rate": 1e-05, "loss": 0.5309, "mean_token_accuracy": 0.8596433401107788, "num_tokens": 472366519.0, "step": 12442 }, { "epoch": 1.5828774965017174, "ewc_loss": 0.10498046875, "ewc_loss_diag": 0.00021648406982421875, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.651199817657471, "learning_rate": 1e-05, "loss": 0.5253, "mean_token_accuracy": 0.8544960021972656, "num_tokens": 472405270.0, "step": 12443 }, { "epoch": 1.583004706780308, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.00021648406982421875, "ewc_loss_parallel": 0.00011396408081054688, "grad_norm": 7.203397750854492, "learning_rate": 1e-05, "loss": 0.5133, "mean_token_accuracy": 0.8747204542160034, "num_tokens": 472447572.0, "step": 12444 }, { "epoch": 1.5831319170588984, "ewc_loss": 0.11376953125, "ewc_loss_diag": 0.00021648406982421875, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 5.896477222442627, "learning_rate": 1e-05, "loss": 0.4804, "mean_token_accuracy": 0.8730918169021606, "num_tokens": 472487869.0, "step": 12445 }, { "epoch": 1.583259127337489, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.00021648406982421875, "ewc_loss_parallel": 0.00011301040649414062, "grad_norm": 7.200486183166504, "learning_rate": 1e-05, "loss": 0.5088, "mean_token_accuracy": 0.8724498748779297, "num_tokens": 472521856.0, "step": 12446 }, { "epoch": 1.5833863376160795, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.00021648406982421875, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.239665985107422, "learning_rate": 1e-05, "loss": 0.5025, "mean_token_accuracy": 0.8720811009407043, "num_tokens": 472566550.0, "step": 12447 }, { "epoch": 1.58351354789467, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.00021648406982421875, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.829938888549805, "learning_rate": 1e-05, "loss": 0.5143, "mean_token_accuracy": 0.8676358461380005, "num_tokens": 472605054.0, "step": 12448 }, { "epoch": 1.5836407581732606, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.00021648406982421875, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.363455772399902, "learning_rate": 1e-05, "loss": 0.4918, "mean_token_accuracy": 0.8712158203125, "num_tokens": 472649938.0, "step": 12449 }, { "epoch": 1.5837679684518509, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.00021648406982421875, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.31938362121582, "learning_rate": 1e-05, "loss": 0.4753, "mean_token_accuracy": 0.8740876913070679, "num_tokens": 472688399.0, "step": 12450 }, { "epoch": 1.5838951787304414, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.00021648406982421875, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.4782490730285645, "learning_rate": 1e-05, "loss": 0.5161, "mean_token_accuracy": 0.8635610938072205, "num_tokens": 472726383.0, "step": 12451 }, { "epoch": 1.584022389009032, "ewc_loss": 0.10595703125, "ewc_loss_diag": 0.00021648406982421875, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.850550651550293, "learning_rate": 1e-05, "loss": 0.4944, "mean_token_accuracy": 0.8642697334289551, "num_tokens": 472759692.0, "step": 12452 }, { "epoch": 1.5841495992876224, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.00021648406982421875, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.723093032836914, "learning_rate": 1e-05, "loss": 0.4943, "mean_token_accuracy": 0.8754791617393494, "num_tokens": 472799679.0, "step": 12453 }, { "epoch": 1.584276809566213, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.00021648406982421875, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.555587291717529, "learning_rate": 1e-05, "loss": 0.5198, "mean_token_accuracy": 0.855417788028717, "num_tokens": 472842043.0, "step": 12454 }, { "epoch": 1.5844040198448035, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.00021648406982421875, "ewc_loss_parallel": 0.00011205673217773438, "grad_norm": 7.126354217529297, "learning_rate": 1e-05, "loss": 0.5405, "mean_token_accuracy": 0.864922046661377, "num_tokens": 472877594.0, "step": 12455 }, { "epoch": 1.5845312301233938, "ewc_loss": 0.10498046875, "ewc_loss_diag": 0.00021648406982421875, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.584282875061035, "learning_rate": 1e-05, "loss": 0.4745, "mean_token_accuracy": 0.8726680874824524, "num_tokens": 472914052.0, "step": 12456 }, { "epoch": 1.5846584404019843, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.00021648406982421875, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.507865905761719, "learning_rate": 1e-05, "loss": 0.5233, "mean_token_accuracy": 0.8705958127975464, "num_tokens": 472953359.0, "step": 12457 }, { "epoch": 1.5847856506805749, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.00021648406982421875, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.01318359375, "learning_rate": 1e-05, "loss": 0.5048, "mean_token_accuracy": 0.8646703958511353, "num_tokens": 472993898.0, "step": 12458 }, { "epoch": 1.5849128609591654, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.00021648406982421875, "ewc_loss_parallel": 0.00012111663818359375, "grad_norm": 7.469691276550293, "learning_rate": 1e-05, "loss": 0.5124, "mean_token_accuracy": 0.872843861579895, "num_tokens": 473032026.0, "step": 12459 }, { "epoch": 1.585040071237756, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.00021648406982421875, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.41382360458374, "learning_rate": 1e-05, "loss": 0.4839, "mean_token_accuracy": 0.8769299983978271, "num_tokens": 473075282.0, "step": 12460 }, { "epoch": 1.5851672815163464, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.00021648406982421875, "ewc_loss_parallel": 0.00011110305786132812, "grad_norm": 7.089558124542236, "learning_rate": 1e-05, "loss": 0.4893, "mean_token_accuracy": 0.8740806579589844, "num_tokens": 473112834.0, "step": 12461 }, { "epoch": 1.585294491794937, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.00021648406982421875, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.4553046226501465, "learning_rate": 1e-05, "loss": 0.5136, "mean_token_accuracy": 0.8667744994163513, "num_tokens": 473150331.0, "step": 12462 }, { "epoch": 1.5854217020735275, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.00021648406982421875, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.706261157989502, "learning_rate": 1e-05, "loss": 0.4901, "mean_token_accuracy": 0.8750673532485962, "num_tokens": 473185175.0, "step": 12463 }, { "epoch": 1.585548912352118, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.00021648406982421875, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.339347839355469, "learning_rate": 1e-05, "loss": 0.5178, "mean_token_accuracy": 0.8606041669845581, "num_tokens": 473223546.0, "step": 12464 }, { "epoch": 1.5856761226307086, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.00021648406982421875, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.455205917358398, "learning_rate": 1e-05, "loss": 0.5576, "mean_token_accuracy": 0.8556287884712219, "num_tokens": 473260098.0, "step": 12465 }, { "epoch": 1.585803332909299, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.00021648406982421875, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.190166473388672, "learning_rate": 1e-05, "loss": 0.4821, "mean_token_accuracy": 0.8713289499282837, "num_tokens": 473299283.0, "step": 12466 }, { "epoch": 1.5859305431878896, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.00021648406982421875, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.343498706817627, "learning_rate": 1e-05, "loss": 0.5047, "mean_token_accuracy": 0.8654327392578125, "num_tokens": 473332511.0, "step": 12467 }, { "epoch": 1.5860577534664801, "ewc_loss": 0.10791015625, "ewc_loss_diag": 0.00021648406982421875, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 6.0143609046936035, "learning_rate": 1e-05, "loss": 0.472, "mean_token_accuracy": 0.8741968870162964, "num_tokens": 473367259.0, "step": 12468 }, { "epoch": 1.5861849637450707, "ewc_loss": 0.11474609375, "ewc_loss_diag": 0.00021648406982421875, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.358877182006836, "learning_rate": 1e-05, "loss": 0.4893, "mean_token_accuracy": 0.8686827421188354, "num_tokens": 473412421.0, "step": 12469 }, { "epoch": 1.5863121740236612, "ewc_loss": 0.10498046875, "ewc_loss_diag": 0.00021648406982421875, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.843028545379639, "learning_rate": 1e-05, "loss": 0.5573, "mean_token_accuracy": 0.8485193252563477, "num_tokens": 473450444.0, "step": 12470 }, { "epoch": 1.5864393843022517, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.00021648406982421875, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.503100395202637, "learning_rate": 1e-05, "loss": 0.4899, "mean_token_accuracy": 0.8679836392402649, "num_tokens": 473485093.0, "step": 12471 }, { "epoch": 1.5865665945808423, "ewc_loss": 0.10205078125, "ewc_loss_diag": 0.00021648406982421875, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.672634601593018, "learning_rate": 1e-05, "loss": 0.4394, "mean_token_accuracy": 0.8811888098716736, "num_tokens": 473522461.0, "step": 12472 }, { "epoch": 1.5866938048594328, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.00021648406982421875, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.759060859680176, "learning_rate": 1e-05, "loss": 0.5155, "mean_token_accuracy": 0.8639810085296631, "num_tokens": 473559860.0, "step": 12473 }, { "epoch": 1.5868210151380233, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.00021648406982421875, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.576831817626953, "learning_rate": 1e-05, "loss": 0.4541, "mean_token_accuracy": 0.8790558576583862, "num_tokens": 473590144.0, "step": 12474 }, { "epoch": 1.5869482254166136, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.00021648406982421875, "ewc_loss_parallel": 0.0001125335693359375, "grad_norm": 7.137763500213623, "learning_rate": 1e-05, "loss": 0.492, "mean_token_accuracy": 0.8751572370529175, "num_tokens": 473631384.0, "step": 12475 }, { "epoch": 1.5870754356952042, "ewc_loss": 0.10595703125, "ewc_loss_diag": 0.00021648406982421875, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.693512439727783, "learning_rate": 1e-05, "loss": 0.5024, "mean_token_accuracy": 0.8679754734039307, "num_tokens": 473667396.0, "step": 12476 }, { "epoch": 1.5872026459737947, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.00021648406982421875, "ewc_loss_parallel": 0.0001201629638671875, "grad_norm": 7.401177883148193, "learning_rate": 1e-05, "loss": 0.517, "mean_token_accuracy": 0.873132586479187, "num_tokens": 473699856.0, "step": 12477 }, { "epoch": 1.5873298562523852, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.00021648406982421875, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.097934722900391, "learning_rate": 1e-05, "loss": 0.5559, "mean_token_accuracy": 0.8490694761276245, "num_tokens": 473731164.0, "step": 12478 }, { "epoch": 1.5874570665309757, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.00021648406982421875, "ewc_loss_parallel": 0.00011396408081054688, "grad_norm": 7.235762119293213, "learning_rate": 1e-05, "loss": 0.5166, "mean_token_accuracy": 0.8656691312789917, "num_tokens": 473767005.0, "step": 12479 }, { "epoch": 1.5875842768095663, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.00021648406982421875, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.392462253570557, "learning_rate": 1e-05, "loss": 0.5082, "mean_token_accuracy": 0.8680292367935181, "num_tokens": 473805753.0, "step": 12480 }, { "epoch": 1.5877114870881566, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.00021648406982421875, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.75183629989624, "learning_rate": 1e-05, "loss": 0.5128, "mean_token_accuracy": 0.8684231042861938, "num_tokens": 473845375.0, "step": 12481 }, { "epoch": 1.587838697366747, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.00021648406982421875, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.4634928703308105, "learning_rate": 1e-05, "loss": 0.5213, "mean_token_accuracy": 0.8591641187667847, "num_tokens": 473889184.0, "step": 12482 }, { "epoch": 1.5879659076453376, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.00021648406982421875, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.235561370849609, "learning_rate": 1e-05, "loss": 0.4781, "mean_token_accuracy": 0.877411961555481, "num_tokens": 473927917.0, "step": 12483 }, { "epoch": 1.5880931179239282, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.00021648406982421875, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.550845146179199, "learning_rate": 1e-05, "loss": 0.4644, "mean_token_accuracy": 0.8816316723823547, "num_tokens": 473964670.0, "step": 12484 }, { "epoch": 1.5882203282025187, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.00021648406982421875, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.826908588409424, "learning_rate": 1e-05, "loss": 0.4363, "mean_token_accuracy": 0.8850835561752319, "num_tokens": 473998455.0, "step": 12485 }, { "epoch": 1.5883475384811092, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.00021648406982421875, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.774740219116211, "learning_rate": 1e-05, "loss": 0.5148, "mean_token_accuracy": 0.8681326508522034, "num_tokens": 474041623.0, "step": 12486 }, { "epoch": 1.5884747487596997, "ewc_loss": 0.10205078125, "ewc_loss_diag": 0.00021648406982421875, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.598556041717529, "learning_rate": 1e-05, "loss": 0.4806, "mean_token_accuracy": 0.8676372766494751, "num_tokens": 474083294.0, "step": 12487 }, { "epoch": 1.5886019590382903, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.00021648406982421875, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 7.130427360534668, "learning_rate": 1e-05, "loss": 0.5041, "mean_token_accuracy": 0.8698430061340332, "num_tokens": 474117737.0, "step": 12488 }, { "epoch": 1.5887291693168808, "ewc_loss": 0.10498046875, "ewc_loss_diag": 0.00021648406982421875, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.661709785461426, "learning_rate": 1e-05, "loss": 0.4909, "mean_token_accuracy": 0.8663097023963928, "num_tokens": 474154467.0, "step": 12489 }, { "epoch": 1.5888563795954713, "ewc_loss": 0.140625, "ewc_loss_diag": 0.00021648406982421875, "ewc_loss_parallel": 0.00011920928955078125, "grad_norm": 7.4210100173950195, "learning_rate": 1e-05, "loss": 0.5186, "mean_token_accuracy": 0.8658623695373535, "num_tokens": 474191845.0, "step": 12490 }, { "epoch": 1.5889835898740619, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.00021648406982421875, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.022068977355957, "learning_rate": 1e-05, "loss": 0.4658, "mean_token_accuracy": 0.8724079132080078, "num_tokens": 474227787.0, "step": 12491 }, { "epoch": 1.5891108001526524, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.00021648406982421875, "ewc_loss_parallel": 0.00011777877807617188, "grad_norm": 7.367791652679443, "learning_rate": 1e-05, "loss": 0.5362, "mean_token_accuracy": 0.8648221492767334, "num_tokens": 474271890.0, "step": 12492 }, { "epoch": 1.589238010431243, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.00021648406982421875, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.333706855773926, "learning_rate": 1e-05, "loss": 0.5416, "mean_token_accuracy": 0.8567787408828735, "num_tokens": 474314493.0, "step": 12493 }, { "epoch": 1.5893652207098334, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.00021648406982421875, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 7.047350883483887, "learning_rate": 1e-05, "loss": 0.4801, "mean_token_accuracy": 0.8788272142410278, "num_tokens": 474349348.0, "step": 12494 }, { "epoch": 1.589492430988424, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.00021648406982421875, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.388676166534424, "learning_rate": 1e-05, "loss": 0.5214, "mean_token_accuracy": 0.8623371124267578, "num_tokens": 474384145.0, "step": 12495 }, { "epoch": 1.5896196412670145, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.00021648406982421875, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.693012714385986, "learning_rate": 1e-05, "loss": 0.4705, "mean_token_accuracy": 0.8792450428009033, "num_tokens": 474419064.0, "step": 12496 }, { "epoch": 1.589746851545605, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.00021648406982421875, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.291221618652344, "learning_rate": 1e-05, "loss": 0.5092, "mean_token_accuracy": 0.8671055436134338, "num_tokens": 474458716.0, "step": 12497 }, { "epoch": 1.5898740618241956, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.00021648406982421875, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.451888084411621, "learning_rate": 1e-05, "loss": 0.4799, "mean_token_accuracy": 0.8751667737960815, "num_tokens": 474497672.0, "step": 12498 }, { "epoch": 1.5900012721027859, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.00021648406982421875, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.1599225997924805, "learning_rate": 1e-05, "loss": 0.4767, "mean_token_accuracy": 0.874575138092041, "num_tokens": 474530442.0, "step": 12499 }, { "epoch": 1.5901284823813764, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.00021648406982421875, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.344437122344971, "learning_rate": 1e-05, "loss": 0.5491, "mean_token_accuracy": 0.8527041673660278, "num_tokens": 474573103.0, "step": 12500 }, { "epoch": 1.590255692659967, "ewc_loss": 0.10791015625, "ewc_loss_diag": 0.00021648406982421875, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 6.0090837478637695, "learning_rate": 1e-05, "loss": 0.4646, "mean_token_accuracy": 0.8765010237693787, "num_tokens": 474610687.0, "step": 12501 }, { "epoch": 1.5903829029385574, "ewc_loss": 0.11474609375, "ewc_loss_diag": 0.00021648406982421875, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.367213726043701, "learning_rate": 1e-05, "loss": 0.517, "mean_token_accuracy": 0.8642712235450745, "num_tokens": 474650893.0, "step": 12502 }, { "epoch": 1.590510113217148, "ewc_loss": 0.10498046875, "ewc_loss_diag": 0.00021648406982421875, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.846816062927246, "learning_rate": 1e-05, "loss": 0.4861, "mean_token_accuracy": 0.8688380718231201, "num_tokens": 474690808.0, "step": 12503 }, { "epoch": 1.5906373234957385, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.00021648406982421875, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.489378452301025, "learning_rate": 1e-05, "loss": 0.4696, "mean_token_accuracy": 0.8780570030212402, "num_tokens": 474726814.0, "step": 12504 }, { "epoch": 1.5907645337743288, "ewc_loss": 0.10302734375, "ewc_loss_diag": 0.00021648406982421875, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.700422286987305, "learning_rate": 1e-05, "loss": 0.5005, "mean_token_accuracy": 0.8651810884475708, "num_tokens": 474765170.0, "step": 12505 }, { "epoch": 1.5908917440529193, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.00021648406982421875, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.726943492889404, "learning_rate": 1e-05, "loss": 0.5193, "mean_token_accuracy": 0.8658343553543091, "num_tokens": 474808466.0, "step": 12506 }, { "epoch": 1.5910189543315099, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.00021648406982421875, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.596024513244629, "learning_rate": 1e-05, "loss": 0.4651, "mean_token_accuracy": 0.8747279644012451, "num_tokens": 474847334.0, "step": 12507 }, { "epoch": 1.5911461646101004, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.00021648406982421875, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 7.070606708526611, "learning_rate": 1e-05, "loss": 0.5095, "mean_token_accuracy": 0.8684787154197693, "num_tokens": 474885937.0, "step": 12508 }, { "epoch": 1.591273374888691, "ewc_loss": 0.10498046875, "ewc_loss_diag": 0.00021648406982421875, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.688076019287109, "learning_rate": 1e-05, "loss": 0.4675, "mean_token_accuracy": 0.8726155161857605, "num_tokens": 474921089.0, "step": 12509 }, { "epoch": 1.5914005851672814, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.00021648406982421875, "ewc_loss_parallel": 0.0001163482666015625, "grad_norm": 7.31489896774292, "learning_rate": 1e-05, "loss": 0.5006, "mean_token_accuracy": 0.8757356405258179, "num_tokens": 474961660.0, "step": 12510 }, { "epoch": 1.591527795445872, "ewc_loss": 0.11474609375, "ewc_loss_diag": 0.00021648406982421875, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.0141401290893555, "learning_rate": 1e-05, "loss": 0.4851, "mean_token_accuracy": 0.8717941045761108, "num_tokens": 474997815.0, "step": 12511 }, { "epoch": 1.5916550057244625, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.00021648406982421875, "ewc_loss_parallel": 0.00011301040649414062, "grad_norm": 7.218672275543213, "learning_rate": 1e-05, "loss": 0.5871, "mean_token_accuracy": 0.8428347110748291, "num_tokens": 475038334.0, "step": 12512 }, { "epoch": 1.591782216003053, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.00021648406982421875, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.322396755218506, "learning_rate": 1e-05, "loss": 0.483, "mean_token_accuracy": 0.8714069128036499, "num_tokens": 475073522.0, "step": 12513 }, { "epoch": 1.5919094262816436, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.00021648406982421875, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.778207778930664, "learning_rate": 1e-05, "loss": 0.4752, "mean_token_accuracy": 0.8757010698318481, "num_tokens": 475111811.0, "step": 12514 }, { "epoch": 1.592036636560234, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.00021648406982421875, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.424566745758057, "learning_rate": 1e-05, "loss": 0.4531, "mean_token_accuracy": 0.8867651224136353, "num_tokens": 475146496.0, "step": 12515 }, { "epoch": 1.5921638468388246, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.00021648406982421875, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.263088226318359, "learning_rate": 1e-05, "loss": 0.5017, "mean_token_accuracy": 0.8690498471260071, "num_tokens": 475187947.0, "step": 12516 }, { "epoch": 1.5922910571174151, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.00021648406982421875, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.533751010894775, "learning_rate": 1e-05, "loss": 0.4945, "mean_token_accuracy": 0.8699023723602295, "num_tokens": 475225420.0, "step": 12517 }, { "epoch": 1.5924182673960057, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.00021648406982421875, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.826196670532227, "learning_rate": 1e-05, "loss": 0.5119, "mean_token_accuracy": 0.8616496324539185, "num_tokens": 475263913.0, "step": 12518 }, { "epoch": 1.5925454776745962, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.00021648406982421875, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.771078586578369, "learning_rate": 1e-05, "loss": 0.4738, "mean_token_accuracy": 0.8808005452156067, "num_tokens": 475300446.0, "step": 12519 }, { "epoch": 1.5926726879531867, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.00021648406982421875, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.595821857452393, "learning_rate": 1e-05, "loss": 0.5355, "mean_token_accuracy": 0.8584613800048828, "num_tokens": 475338728.0, "step": 12520 }, { "epoch": 1.5927998982317773, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.00021648406982421875, "ewc_loss_parallel": 0.00010967254638671875, "grad_norm": 7.120508193969727, "learning_rate": 1e-05, "loss": 0.5085, "mean_token_accuracy": 0.8678185939788818, "num_tokens": 475369453.0, "step": 12521 }, { "epoch": 1.5929271085103678, "ewc_loss": 0.10400390625, "ewc_loss_diag": 0.00021648406982421875, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.653676986694336, "learning_rate": 1e-05, "loss": 0.4498, "mean_token_accuracy": 0.8825832605361938, "num_tokens": 475407994.0, "step": 12522 }, { "epoch": 1.5930543187889583, "ewc_loss": 0.140625, "ewc_loss_diag": 0.00021648406982421875, "ewc_loss_parallel": 0.00011920928955078125, "grad_norm": 7.4187541007995605, "learning_rate": 1e-05, "loss": 0.5339, "mean_token_accuracy": 0.8662683963775635, "num_tokens": 475445480.0, "step": 12523 }, { "epoch": 1.5931815290675486, "ewc_loss": 0.11474609375, "ewc_loss_diag": 0.00021648406982421875, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.0219340324401855, "learning_rate": 1e-05, "loss": 0.5418, "mean_token_accuracy": 0.8566653728485107, "num_tokens": 475483971.0, "step": 12524 }, { "epoch": 1.5933087393461391, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.00021648406982421875, "ewc_loss_parallel": 0.00011730194091796875, "grad_norm": 7.359271049499512, "learning_rate": 1e-05, "loss": 0.516, "mean_token_accuracy": 0.873013973236084, "num_tokens": 475521332.0, "step": 12525 }, { "epoch": 1.5934359496247297, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.00021648406982421875, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.340939998626709, "learning_rate": 1e-05, "loss": 0.5102, "mean_token_accuracy": 0.8636773824691772, "num_tokens": 475556085.0, "step": 12526 }, { "epoch": 1.5935631599033202, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.00021648406982421875, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 7.01894474029541, "learning_rate": 1e-05, "loss": 0.5152, "mean_token_accuracy": 0.8658468723297119, "num_tokens": 475590966.0, "step": 12527 }, { "epoch": 1.5936903701819107, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.00021648406982421875, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.384409427642822, "learning_rate": 1e-05, "loss": 0.5231, "mean_token_accuracy": 0.8626447916030884, "num_tokens": 475627430.0, "step": 12528 }, { "epoch": 1.5938175804605013, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.00021648406982421875, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.6587066650390625, "learning_rate": 1e-05, "loss": 0.4645, "mean_token_accuracy": 0.8786631226539612, "num_tokens": 475662346.0, "step": 12529 }, { "epoch": 1.5939447907390916, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.00021648406982421875, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.286346435546875, "learning_rate": 1e-05, "loss": 0.5248, "mean_token_accuracy": 0.8615800142288208, "num_tokens": 475702312.0, "step": 12530 }, { "epoch": 1.594072001017682, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.00021648406982421875, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.41015100479126, "learning_rate": 1e-05, "loss": 0.5082, "mean_token_accuracy": 0.870201826095581, "num_tokens": 475743997.0, "step": 12531 }, { "epoch": 1.5941992112962726, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.00021648406982421875, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.163686752319336, "learning_rate": 1e-05, "loss": 0.5253, "mean_token_accuracy": 0.8573209047317505, "num_tokens": 475783369.0, "step": 12532 }, { "epoch": 1.5943264215748632, "ewc_loss": 0.11474609375, "ewc_loss_diag": 0.00021648406982421875, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.31592321395874, "learning_rate": 1e-05, "loss": 0.4489, "mean_token_accuracy": 0.883259654045105, "num_tokens": 475815225.0, "step": 12533 }, { "epoch": 1.5944536318534537, "ewc_loss": 0.10888671875, "ewc_loss_diag": 0.00021648406982421875, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 6.017765998840332, "learning_rate": 1e-05, "loss": 0.4986, "mean_token_accuracy": 0.864089846611023, "num_tokens": 475847917.0, "step": 12534 }, { "epoch": 1.5945808421320442, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.00021648406982421875, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.318310260772705, "learning_rate": 1e-05, "loss": 0.478, "mean_token_accuracy": 0.8744828701019287, "num_tokens": 475889308.0, "step": 12535 }, { "epoch": 1.5947080524106347, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.00021648406982421875, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.859793663024902, "learning_rate": 1e-05, "loss": 0.4572, "mean_token_accuracy": 0.8785502910614014, "num_tokens": 475930603.0, "step": 12536 }, { "epoch": 1.5948352626892253, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.00021648406982421875, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.438538551330566, "learning_rate": 1e-05, "loss": 0.4722, "mean_token_accuracy": 0.8766922950744629, "num_tokens": 475968288.0, "step": 12537 }, { "epoch": 1.5949624729678158, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.00021648406982421875, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.702726364135742, "learning_rate": 1e-05, "loss": 0.4693, "mean_token_accuracy": 0.873677134513855, "num_tokens": 476002967.0, "step": 12538 }, { "epoch": 1.5950896832464063, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.00021648406982421875, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.665243148803711, "learning_rate": 1e-05, "loss": 0.509, "mean_token_accuracy": 0.8699966669082642, "num_tokens": 476037958.0, "step": 12539 }, { "epoch": 1.5952168935249968, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.00021648406982421875, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.587374687194824, "learning_rate": 1e-05, "loss": 0.5446, "mean_token_accuracy": 0.8499077558517456, "num_tokens": 476072036.0, "step": 12540 }, { "epoch": 1.5953441038035874, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.00021648406982421875, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 7.0020670890808105, "learning_rate": 1e-05, "loss": 0.501, "mean_token_accuracy": 0.8739518523216248, "num_tokens": 476115665.0, "step": 12541 }, { "epoch": 1.595471314082178, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.00021648406982421875, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.6132493019104, "learning_rate": 1e-05, "loss": 0.495, "mean_token_accuracy": 0.86371248960495, "num_tokens": 476155380.0, "step": 12542 }, { "epoch": 1.5955985243607684, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.00021648406982421875, "ewc_loss_parallel": 0.0001163482666015625, "grad_norm": 7.329129219055176, "learning_rate": 1e-05, "loss": 0.5302, "mean_token_accuracy": 0.8638597726821899, "num_tokens": 476194833.0, "step": 12543 }, { "epoch": 1.595725734639359, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.00021648406982421875, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 5.927404403686523, "learning_rate": 1e-05, "loss": 0.48, "mean_token_accuracy": 0.8731124401092529, "num_tokens": 476233617.0, "step": 12544 }, { "epoch": 1.5958529449179495, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.00021648406982421875, "ewc_loss_parallel": 0.00011730194091796875, "grad_norm": 7.34876823425293, "learning_rate": 1e-05, "loss": 0.5305, "mean_token_accuracy": 0.864787220954895, "num_tokens": 476270312.0, "step": 12545 }, { "epoch": 1.59598015519654, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.00021648406982421875, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.299098014831543, "learning_rate": 1e-05, "loss": 0.4196, "mean_token_accuracy": 0.8956636190414429, "num_tokens": 476304466.0, "step": 12546 }, { "epoch": 1.5961073654751305, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.00021648406982421875, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.9716997146606445, "learning_rate": 1e-05, "loss": 0.4942, "mean_token_accuracy": 0.8729552626609802, "num_tokens": 476345136.0, "step": 12547 }, { "epoch": 1.5962345757537209, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.00021648406982421875, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.440577507019043, "learning_rate": 1e-05, "loss": 0.4902, "mean_token_accuracy": 0.8746927976608276, "num_tokens": 476381546.0, "step": 12548 }, { "epoch": 1.5963617860323114, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.00021648406982421875, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.476623058319092, "learning_rate": 1e-05, "loss": 0.4978, "mean_token_accuracy": 0.8691327571868896, "num_tokens": 476413990.0, "step": 12549 }, { "epoch": 1.596488996310902, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.00021648406982421875, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.479835033416748, "learning_rate": 1e-05, "loss": 0.4662, "mean_token_accuracy": 0.881898045539856, "num_tokens": 476451487.0, "step": 12550 }, { "epoch": 1.5966162065894924, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.00021648406982421875, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 6.023264408111572, "learning_rate": 1e-05, "loss": 0.5338, "mean_token_accuracy": 0.8567522764205933, "num_tokens": 476493071.0, "step": 12551 }, { "epoch": 1.596743416868083, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.00021648406982421875, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.607311248779297, "learning_rate": 1e-05, "loss": 0.4699, "mean_token_accuracy": 0.8759347200393677, "num_tokens": 476531390.0, "step": 12552 }, { "epoch": 1.5968706271466735, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.00021648406982421875, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.729244232177734, "learning_rate": 1e-05, "loss": 0.4947, "mean_token_accuracy": 0.8666279315948486, "num_tokens": 476572491.0, "step": 12553 }, { "epoch": 1.5969978374252638, "ewc_loss": 0.125, "ewc_loss_diag": 0.00021648406982421875, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.841556549072266, "learning_rate": 1e-05, "loss": 0.5344, "mean_token_accuracy": 0.8617086410522461, "num_tokens": 476613897.0, "step": 12554 }, { "epoch": 1.5971250477038543, "ewc_loss": 0.10302734375, "ewc_loss_diag": 0.00021648406982421875, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.630478858947754, "learning_rate": 1e-05, "loss": 0.5017, "mean_token_accuracy": 0.8622244596481323, "num_tokens": 476650773.0, "step": 12555 }, { "epoch": 1.5972522579824449, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.00021648406982421875, "ewc_loss_parallel": 0.00011157989501953125, "grad_norm": 7.133460521697998, "learning_rate": 1e-05, "loss": 0.5189, "mean_token_accuracy": 0.8699905276298523, "num_tokens": 476685546.0, "step": 12556 }, { "epoch": 1.5973794682610354, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.00021648406982421875, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.7624969482421875, "learning_rate": 1e-05, "loss": 0.5485, "mean_token_accuracy": 0.8524684309959412, "num_tokens": 476720569.0, "step": 12557 }, { "epoch": 1.597506678539626, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.00021648406982421875, "ewc_loss_parallel": 0.00011539459228515625, "grad_norm": 7.299310207366943, "learning_rate": 1e-05, "loss": 0.5015, "mean_token_accuracy": 0.8732227087020874, "num_tokens": 476760162.0, "step": 12558 }, { "epoch": 1.5976338888182164, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.043494701385498, "learning_rate": 1e-05, "loss": 0.4932, "mean_token_accuracy": 0.8688203692436218, "num_tokens": 476793575.0, "step": 12559 }, { "epoch": 1.597761099096807, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 0.00011348724365234375, "grad_norm": 7.202995300292969, "learning_rate": 1e-05, "loss": 0.5124, "mean_token_accuracy": 0.8728508949279785, "num_tokens": 476826714.0, "step": 12560 }, { "epoch": 1.5978883093753975, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.00021648406982421875, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.233756065368652, "learning_rate": 1e-05, "loss": 0.4502, "mean_token_accuracy": 0.8847463130950928, "num_tokens": 476866715.0, "step": 12561 }, { "epoch": 1.598015519653988, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.946319580078125, "learning_rate": 1e-05, "loss": 0.5123, "mean_token_accuracy": 0.8654316067695618, "num_tokens": 476906036.0, "step": 12562 }, { "epoch": 1.5981427299325786, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.254275798797607, "learning_rate": 1e-05, "loss": 0.5005, "mean_token_accuracy": 0.8692479729652405, "num_tokens": 476945005.0, "step": 12563 }, { "epoch": 1.598269940211169, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.659078121185303, "learning_rate": 1e-05, "loss": 0.483, "mean_token_accuracy": 0.8757627606391907, "num_tokens": 476989491.0, "step": 12564 }, { "epoch": 1.5983971504897596, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.198705196380615, "learning_rate": 1e-05, "loss": 0.5077, "mean_token_accuracy": 0.8688355684280396, "num_tokens": 477026062.0, "step": 12565 }, { "epoch": 1.5985243607683501, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.461122989654541, "learning_rate": 1e-05, "loss": 0.5002, "mean_token_accuracy": 0.8661109209060669, "num_tokens": 477060045.0, "step": 12566 }, { "epoch": 1.5986515710469407, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.093906402587891, "learning_rate": 1e-05, "loss": 0.4582, "mean_token_accuracy": 0.8819280862808228, "num_tokens": 477102343.0, "step": 12567 }, { "epoch": 1.5987787813255312, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.375466823577881, "learning_rate": 1e-05, "loss": 0.4677, "mean_token_accuracy": 0.8776100873947144, "num_tokens": 477136239.0, "step": 12568 }, { "epoch": 1.5989059916041217, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.962072849273682, "learning_rate": 1e-05, "loss": 0.4905, "mean_token_accuracy": 0.8673399090766907, "num_tokens": 477178291.0, "step": 12569 }, { "epoch": 1.5990332018827123, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.400786876678467, "learning_rate": 1e-05, "loss": 0.5469, "mean_token_accuracy": 0.8516209721565247, "num_tokens": 477213993.0, "step": 12570 }, { "epoch": 1.5991604121613028, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.829639434814453, "learning_rate": 1e-05, "loss": 0.5194, "mean_token_accuracy": 0.8593632578849792, "num_tokens": 477258384.0, "step": 12571 }, { "epoch": 1.5992876224398933, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.527092933654785, "learning_rate": 1e-05, "loss": 0.4995, "mean_token_accuracy": 0.8698585033416748, "num_tokens": 477297386.0, "step": 12572 }, { "epoch": 1.5994148327184836, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.706845283508301, "learning_rate": 1e-05, "loss": 0.4483, "mean_token_accuracy": 0.8792896866798401, "num_tokens": 477334214.0, "step": 12573 }, { "epoch": 1.5995420429970741, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.744915962219238, "learning_rate": 1e-05, "loss": 0.5411, "mean_token_accuracy": 0.8581000566482544, "num_tokens": 477374777.0, "step": 12574 }, { "epoch": 1.5996692532756647, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.643515586853027, "learning_rate": 1e-05, "loss": 0.4803, "mean_token_accuracy": 0.8685787320137024, "num_tokens": 477415206.0, "step": 12575 }, { "epoch": 1.5997964635542552, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 7.00436544418335, "learning_rate": 1e-05, "loss": 0.5083, "mean_token_accuracy": 0.8679957985877991, "num_tokens": 477453864.0, "step": 12576 }, { "epoch": 1.5999236738328457, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.7470903396606445, "learning_rate": 1e-05, "loss": 0.5349, "mean_token_accuracy": 0.8540400266647339, "num_tokens": 477489236.0, "step": 12577 }, { "epoch": 1.6000508841114363, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 0.00011157989501953125, "grad_norm": 7.162339687347412, "learning_rate": 1e-05, "loss": 0.4838, "mean_token_accuracy": 0.8803458213806152, "num_tokens": 477524459.0, "step": 12578 }, { "epoch": 1.6001780943900266, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 5.997750282287598, "learning_rate": 1e-05, "loss": 0.5053, "mean_token_accuracy": 0.8663454055786133, "num_tokens": 477560045.0, "step": 12579 }, { "epoch": 1.600305304668617, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 7.026756763458252, "learning_rate": 1e-05, "loss": 0.5473, "mean_token_accuracy": 0.8568217754364014, "num_tokens": 477596438.0, "step": 12580 }, { "epoch": 1.6004325149472076, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.233458042144775, "learning_rate": 1e-05, "loss": 0.4945, "mean_token_accuracy": 0.8695234060287476, "num_tokens": 477634370.0, "step": 12581 }, { "epoch": 1.6005597252257981, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.6191606521606445, "learning_rate": 1e-05, "loss": 0.5226, "mean_token_accuracy": 0.863634467124939, "num_tokens": 477669798.0, "step": 12582 }, { "epoch": 1.6006869355043887, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.345622539520264, "learning_rate": 1e-05, "loss": 0.5076, "mean_token_accuracy": 0.8679598569869995, "num_tokens": 477714127.0, "step": 12583 }, { "epoch": 1.6008141457829792, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 6.123908519744873, "learning_rate": 1e-05, "loss": 0.4816, "mean_token_accuracy": 0.8743298053741455, "num_tokens": 477753176.0, "step": 12584 }, { "epoch": 1.6009413560615697, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.520015239715576, "learning_rate": 1e-05, "loss": 0.4836, "mean_token_accuracy": 0.872503399848938, "num_tokens": 477789372.0, "step": 12585 }, { "epoch": 1.6010685663401603, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.701152801513672, "learning_rate": 1e-05, "loss": 0.481, "mean_token_accuracy": 0.8639299869537354, "num_tokens": 477820891.0, "step": 12586 }, { "epoch": 1.6011957766187508, "ewc_loss": 0.125, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.866849899291992, "learning_rate": 1e-05, "loss": 0.5211, "mean_token_accuracy": 0.8679163455963135, "num_tokens": 477853952.0, "step": 12587 }, { "epoch": 1.6013229868973413, "ewc_loss": 0.1005859375, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 7.915496826171875e-05, "grad_norm": 5.486296653747559, "learning_rate": 1e-05, "loss": 0.4386, "mean_token_accuracy": 0.8841766119003296, "num_tokens": 477888692.0, "step": 12588 }, { "epoch": 1.6014501971759318, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 0.00011777877807617188, "grad_norm": 7.342007160186768, "learning_rate": 1e-05, "loss": 0.4838, "mean_token_accuracy": 0.8817166090011597, "num_tokens": 477923769.0, "step": 12589 }, { "epoch": 1.6015774074545224, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.679983615875244, "learning_rate": 1e-05, "loss": 0.4311, "mean_token_accuracy": 0.8863595128059387, "num_tokens": 477959997.0, "step": 12590 }, { "epoch": 1.601704617733113, "ewc_loss": 0.1474609375, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 0.000125885009765625, "grad_norm": 7.682125568389893, "learning_rate": 1e-05, "loss": 0.5699, "mean_token_accuracy": 0.8523786664009094, "num_tokens": 477996211.0, "step": 12591 }, { "epoch": 1.6018318280117034, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.274600028991699, "learning_rate": 1e-05, "loss": 0.4977, "mean_token_accuracy": 0.8706335425376892, "num_tokens": 478029817.0, "step": 12592 }, { "epoch": 1.601959038290294, "ewc_loss": 0.140625, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 0.00011873245239257812, "grad_norm": 7.431265830993652, "learning_rate": 1e-05, "loss": 0.4916, "mean_token_accuracy": 0.8773497939109802, "num_tokens": 478071216.0, "step": 12593 }, { "epoch": 1.6020862485688845, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 6.5521697998046875, "learning_rate": 1e-05, "loss": 0.4903, "mean_token_accuracy": 0.874634861946106, "num_tokens": 478112581.0, "step": 12594 }, { "epoch": 1.602213458847475, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 6.999673843383789, "learning_rate": 1e-05, "loss": 0.5356, "mean_token_accuracy": 0.8582900166511536, "num_tokens": 478148887.0, "step": 12595 }, { "epoch": 1.6023406691260655, "ewc_loss": 0.125, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.482336521148682, "learning_rate": 1e-05, "loss": 0.5374, "mean_token_accuracy": 0.8587883114814758, "num_tokens": 478187580.0, "step": 12596 }, { "epoch": 1.6024678794046558, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.6399078369140625, "learning_rate": 1e-05, "loss": 0.4848, "mean_token_accuracy": 0.8725700378417969, "num_tokens": 478222679.0, "step": 12597 }, { "epoch": 1.6025950896832464, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.330803871154785, "learning_rate": 1e-05, "loss": 0.4818, "mean_token_accuracy": 0.8763788938522339, "num_tokens": 478259464.0, "step": 12598 }, { "epoch": 1.602722299961837, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.421353340148926, "learning_rate": 1e-05, "loss": 0.447, "mean_token_accuracy": 0.8837436437606812, "num_tokens": 478293790.0, "step": 12599 }, { "epoch": 1.6028495102404274, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.16447639465332, "learning_rate": 1e-05, "loss": 0.4917, "mean_token_accuracy": 0.8725714683532715, "num_tokens": 478329311.0, "step": 12600 }, { "epoch": 1.602976720519018, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.352360725402832, "learning_rate": 1e-05, "loss": 0.4897, "mean_token_accuracy": 0.8723564147949219, "num_tokens": 478362818.0, "step": 12601 }, { "epoch": 1.6031039307976085, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.983391761779785, "learning_rate": 1e-05, "loss": 0.4979, "mean_token_accuracy": 0.8664919137954712, "num_tokens": 478402783.0, "step": 12602 }, { "epoch": 1.6032311410761988, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.402045249938965, "learning_rate": 1e-05, "loss": 0.5318, "mean_token_accuracy": 0.8574870824813843, "num_tokens": 478438141.0, "step": 12603 }, { "epoch": 1.6033583513547893, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.806544303894043, "learning_rate": 1e-05, "loss": 0.4601, "mean_token_accuracy": 0.8771079778671265, "num_tokens": 478473176.0, "step": 12604 }, { "epoch": 1.6034855616333799, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.578554630279541, "learning_rate": 1e-05, "loss": 0.4827, "mean_token_accuracy": 0.8742811679840088, "num_tokens": 478511899.0, "step": 12605 }, { "epoch": 1.6036127719119704, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.6389055252075195, "learning_rate": 1e-05, "loss": 0.4361, "mean_token_accuracy": 0.8830009698867798, "num_tokens": 478545018.0, "step": 12606 }, { "epoch": 1.603739982190561, "ewc_loss": 0.125, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.897152900695801, "learning_rate": 1e-05, "loss": 0.5155, "mean_token_accuracy": 0.8636124134063721, "num_tokens": 478582384.0, "step": 12607 }, { "epoch": 1.6038671924691514, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.586347579956055, "learning_rate": 1e-05, "loss": 0.4904, "mean_token_accuracy": 0.868624746799469, "num_tokens": 478621491.0, "step": 12608 }, { "epoch": 1.603994402747742, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 0.00011539459228515625, "grad_norm": 7.271986484527588, "learning_rate": 1e-05, "loss": 0.4529, "mean_token_accuracy": 0.8908637166023254, "num_tokens": 478662018.0, "step": 12609 }, { "epoch": 1.6041216130263325, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.801571846008301, "learning_rate": 1e-05, "loss": 0.4878, "mean_token_accuracy": 0.8687440156936646, "num_tokens": 478702730.0, "step": 12610 }, { "epoch": 1.604248823304923, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 0.0001220703125, "grad_norm": 7.472011566162109, "learning_rate": 1e-05, "loss": 0.5296, "mean_token_accuracy": 0.867218017578125, "num_tokens": 478736696.0, "step": 12611 }, { "epoch": 1.6043760335835135, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.253754138946533, "learning_rate": 1e-05, "loss": 0.5627, "mean_token_accuracy": 0.8509750366210938, "num_tokens": 478775873.0, "step": 12612 }, { "epoch": 1.604503243862104, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 0.00011348724365234375, "grad_norm": 7.175342559814453, "learning_rate": 1e-05, "loss": 0.4815, "mean_token_accuracy": 0.8758453726768494, "num_tokens": 478810571.0, "step": 12613 }, { "epoch": 1.6046304541406946, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.488306999206543, "learning_rate": 1e-05, "loss": 0.5488, "mean_token_accuracy": 0.8570988774299622, "num_tokens": 478845696.0, "step": 12614 }, { "epoch": 1.6047576644192851, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.660333156585693, "learning_rate": 1e-05, "loss": 0.5135, "mean_token_accuracy": 0.8639724850654602, "num_tokens": 478884561.0, "step": 12615 }, { "epoch": 1.6048848746978757, "ewc_loss": 0.125, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.529924392700195, "learning_rate": 1e-05, "loss": 0.5885, "mean_token_accuracy": 0.8463460206985474, "num_tokens": 478921989.0, "step": 12616 }, { "epoch": 1.6050120849764662, "ewc_loss": 0.109375, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 6.151335716247559, "learning_rate": 1e-05, "loss": 0.4923, "mean_token_accuracy": 0.8672038912773132, "num_tokens": 478958346.0, "step": 12617 }, { "epoch": 1.6051392952550567, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.605968475341797, "learning_rate": 1e-05, "loss": 0.498, "mean_token_accuracy": 0.8718193769454956, "num_tokens": 479000141.0, "step": 12618 }, { "epoch": 1.6052665055336472, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.00021839141845703125, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.778926849365234, "learning_rate": 1e-05, "loss": 0.4687, "mean_token_accuracy": 0.8765080571174622, "num_tokens": 479038785.0, "step": 12619 }, { "epoch": 1.6053937158122378, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.00021839141845703125, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.8566789627075195, "learning_rate": 1e-05, "loss": 0.5246, "mean_token_accuracy": 0.860418438911438, "num_tokens": 479074679.0, "step": 12620 }, { "epoch": 1.6055209260908283, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.62825870513916, "learning_rate": 1e-05, "loss": 0.4645, "mean_token_accuracy": 0.8742780685424805, "num_tokens": 479106745.0, "step": 12621 }, { "epoch": 1.6056481363694186, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 0.0001125335693359375, "grad_norm": 7.185697078704834, "learning_rate": 1e-05, "loss": 0.505, "mean_token_accuracy": 0.8702954649925232, "num_tokens": 479145159.0, "step": 12622 }, { "epoch": 1.6057753466480091, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.744999408721924, "learning_rate": 1e-05, "loss": 0.4464, "mean_token_accuracy": 0.8825268745422363, "num_tokens": 479181033.0, "step": 12623 }, { "epoch": 1.6059025569265997, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 0.00012063980102539062, "grad_norm": 7.409628391265869, "learning_rate": 1e-05, "loss": 0.5515, "mean_token_accuracy": 0.8637673258781433, "num_tokens": 479223954.0, "step": 12624 }, { "epoch": 1.6060297672051902, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.104363918304443, "learning_rate": 1e-05, "loss": 0.535, "mean_token_accuracy": 0.855923056602478, "num_tokens": 479268676.0, "step": 12625 }, { "epoch": 1.6061569774837807, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 0.0001163482666015625, "grad_norm": 7.285994052886963, "learning_rate": 1e-05, "loss": 0.4911, "mean_token_accuracy": 0.8792965412139893, "num_tokens": 479305203.0, "step": 12626 }, { "epoch": 1.6062841877623713, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.348176956176758, "learning_rate": 1e-05, "loss": 0.5879, "mean_token_accuracy": 0.8433693647384644, "num_tokens": 479347358.0, "step": 12627 }, { "epoch": 1.6064113980409616, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.948398113250732, "learning_rate": 1e-05, "loss": 0.547, "mean_token_accuracy": 0.8601675629615784, "num_tokens": 479392973.0, "step": 12628 }, { "epoch": 1.606538608319552, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.359978199005127, "learning_rate": 1e-05, "loss": 0.5386, "mean_token_accuracy": 0.8557770252227783, "num_tokens": 479430832.0, "step": 12629 }, { "epoch": 1.6066658185981426, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.614428520202637, "learning_rate": 1e-05, "loss": 0.5379, "mean_token_accuracy": 0.8578509092330933, "num_tokens": 479471363.0, "step": 12630 }, { "epoch": 1.6067930288767331, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.265777587890625, "learning_rate": 1e-05, "loss": 0.47, "mean_token_accuracy": 0.8752020597457886, "num_tokens": 479506041.0, "step": 12631 }, { "epoch": 1.6069202391553237, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.406255722045898, "learning_rate": 1e-05, "loss": 0.4965, "mean_token_accuracy": 0.8700562119483948, "num_tokens": 479537290.0, "step": 12632 }, { "epoch": 1.6070474494339142, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.132526397705078, "learning_rate": 1e-05, "loss": 0.5407, "mean_token_accuracy": 0.8546581864356995, "num_tokens": 479573191.0, "step": 12633 }, { "epoch": 1.6071746597125047, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.336020469665527, "learning_rate": 1e-05, "loss": 0.4452, "mean_token_accuracy": 0.8824267387390137, "num_tokens": 479606210.0, "step": 12634 }, { "epoch": 1.6073018699910953, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.966643333435059, "learning_rate": 1e-05, "loss": 0.4389, "mean_token_accuracy": 0.8841383457183838, "num_tokens": 479645198.0, "step": 12635 }, { "epoch": 1.6074290802696858, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.371060848236084, "learning_rate": 1e-05, "loss": 0.5034, "mean_token_accuracy": 0.864522397518158, "num_tokens": 479686319.0, "step": 12636 }, { "epoch": 1.6075562905482763, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.821383476257324, "learning_rate": 1e-05, "loss": 0.4272, "mean_token_accuracy": 0.8906542062759399, "num_tokens": 479720170.0, "step": 12637 }, { "epoch": 1.6076835008268668, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.519482612609863, "learning_rate": 1e-05, "loss": 0.4794, "mean_token_accuracy": 0.8747868537902832, "num_tokens": 479757263.0, "step": 12638 }, { "epoch": 1.6078107111054574, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.665006160736084, "learning_rate": 1e-05, "loss": 0.4912, "mean_token_accuracy": 0.8668606281280518, "num_tokens": 479793050.0, "step": 12639 }, { "epoch": 1.607937921384048, "ewc_loss": 0.125, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.803067684173584, "learning_rate": 1e-05, "loss": 0.5196, "mean_token_accuracy": 0.8659783601760864, "num_tokens": 479831705.0, "step": 12640 }, { "epoch": 1.6080651316626384, "ewc_loss": 0.1015625, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.579654693603516, "learning_rate": 1e-05, "loss": 0.4286, "mean_token_accuracy": 0.8826358318328857, "num_tokens": 479863888.0, "step": 12641 }, { "epoch": 1.608192341941229, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 0.00011301040649414062, "grad_norm": 7.160534381866455, "learning_rate": 1e-05, "loss": 0.4878, "mean_token_accuracy": 0.8788965940475464, "num_tokens": 479902280.0, "step": 12642 }, { "epoch": 1.6083195522198195, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.726241111755371, "learning_rate": 1e-05, "loss": 0.4754, "mean_token_accuracy": 0.8724824786186218, "num_tokens": 479935649.0, "step": 12643 }, { "epoch": 1.60844676249841, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 0.000118255615234375, "grad_norm": 7.393252849578857, "learning_rate": 1e-05, "loss": 0.5298, "mean_token_accuracy": 0.8667310476303101, "num_tokens": 479971834.0, "step": 12644 }, { "epoch": 1.6085739727770005, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.121649742126465, "learning_rate": 1e-05, "loss": 0.5445, "mean_token_accuracy": 0.855474054813385, "num_tokens": 480011985.0, "step": 12645 }, { "epoch": 1.6087011830555908, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 0.0001125335693359375, "grad_norm": 7.205046653747559, "learning_rate": 1e-05, "loss": 0.5492, "mean_token_accuracy": 0.8567616939544678, "num_tokens": 480047833.0, "step": 12646 }, { "epoch": 1.6088283933341814, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.408535957336426, "learning_rate": 1e-05, "loss": 0.5185, "mean_token_accuracy": 0.860736608505249, "num_tokens": 480080774.0, "step": 12647 }, { "epoch": 1.608955603612772, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.729377269744873, "learning_rate": 1e-05, "loss": 0.4946, "mean_token_accuracy": 0.8713568449020386, "num_tokens": 480119114.0, "step": 12648 }, { "epoch": 1.6090828138913624, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.467316150665283, "learning_rate": 1e-05, "loss": 0.5183, "mean_token_accuracy": 0.8634374141693115, "num_tokens": 480156624.0, "step": 12649 }, { "epoch": 1.609210024169953, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.211188793182373, "learning_rate": 1e-05, "loss": 0.4845, "mean_token_accuracy": 0.8704833388328552, "num_tokens": 480192573.0, "step": 12650 }, { "epoch": 1.6093372344485435, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.577881813049316, "learning_rate": 1e-05, "loss": 0.5504, "mean_token_accuracy": 0.854375958442688, "num_tokens": 480232139.0, "step": 12651 }, { "epoch": 1.6094644447271338, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.803345203399658, "learning_rate": 1e-05, "loss": 0.513, "mean_token_accuracy": 0.8612552881240845, "num_tokens": 480269811.0, "step": 12652 }, { "epoch": 1.6095916550057243, "ewc_loss": 0.125, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.812858581542969, "learning_rate": 1e-05, "loss": 0.5302, "mean_token_accuracy": 0.8622106909751892, "num_tokens": 480307785.0, "step": 12653 }, { "epoch": 1.6097188652843148, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.604799747467041, "learning_rate": 1e-05, "loss": 0.487, "mean_token_accuracy": 0.8673478960990906, "num_tokens": 480347028.0, "step": 12654 }, { "epoch": 1.6098460755629054, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 0.00011157989501953125, "grad_norm": 7.166373252868652, "learning_rate": 1e-05, "loss": 0.4866, "mean_token_accuracy": 0.8766229748725891, "num_tokens": 480383915.0, "step": 12655 }, { "epoch": 1.609973285841496, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.7020769119262695, "learning_rate": 1e-05, "loss": 0.4797, "mean_token_accuracy": 0.8713470697402954, "num_tokens": 480424226.0, "step": 12656 }, { "epoch": 1.6101004961200864, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 0.0001201629638671875, "grad_norm": 7.435337543487549, "learning_rate": 1e-05, "loss": 0.4869, "mean_token_accuracy": 0.8798421025276184, "num_tokens": 480461920.0, "step": 12657 }, { "epoch": 1.610227706398677, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.081876277923584, "learning_rate": 1e-05, "loss": 0.474, "mean_token_accuracy": 0.876844048500061, "num_tokens": 480498509.0, "step": 12658 }, { "epoch": 1.6103549166772675, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 0.00011730194091796875, "grad_norm": 7.331358909606934, "learning_rate": 1e-05, "loss": 0.5427, "mean_token_accuracy": 0.859704852104187, "num_tokens": 480541912.0, "step": 12659 }, { "epoch": 1.610482126955858, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.35697078704834, "learning_rate": 1e-05, "loss": 0.4856, "mean_token_accuracy": 0.8735344409942627, "num_tokens": 480576034.0, "step": 12660 }, { "epoch": 1.6106093372344485, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.997182846069336, "learning_rate": 1e-05, "loss": 0.5307, "mean_token_accuracy": 0.8629846572875977, "num_tokens": 480612265.0, "step": 12661 }, { "epoch": 1.610736547513039, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.3798933029174805, "learning_rate": 1e-05, "loss": 0.5645, "mean_token_accuracy": 0.850036084651947, "num_tokens": 480650667.0, "step": 12662 }, { "epoch": 1.6108637577916296, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.648122310638428, "learning_rate": 1e-05, "loss": 0.5104, "mean_token_accuracy": 0.867158055305481, "num_tokens": 480688931.0, "step": 12663 }, { "epoch": 1.6109909680702201, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.293125629425049, "learning_rate": 1e-05, "loss": 0.5213, "mean_token_accuracy": 0.8624264001846313, "num_tokens": 480719525.0, "step": 12664 }, { "epoch": 1.6111181783488107, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.4095306396484375, "learning_rate": 1e-05, "loss": 0.4592, "mean_token_accuracy": 0.8805963397026062, "num_tokens": 480755872.0, "step": 12665 }, { "epoch": 1.6112453886274012, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.150726318359375, "learning_rate": 1e-05, "loss": 0.4833, "mean_token_accuracy": 0.8740220665931702, "num_tokens": 480794994.0, "step": 12666 }, { "epoch": 1.6113725989059917, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.330339431762695, "learning_rate": 1e-05, "loss": 0.4759, "mean_token_accuracy": 0.8759599924087524, "num_tokens": 480828421.0, "step": 12667 }, { "epoch": 1.6114998091845822, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.995278835296631, "learning_rate": 1e-05, "loss": 0.4433, "mean_token_accuracy": 0.8814064264297485, "num_tokens": 480868571.0, "step": 12668 }, { "epoch": 1.6116270194631728, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.3689351081848145, "learning_rate": 1e-05, "loss": 0.4546, "mean_token_accuracy": 0.8813983201980591, "num_tokens": 480910897.0, "step": 12669 }, { "epoch": 1.6117542297417633, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.83388090133667, "learning_rate": 1e-05, "loss": 0.4873, "mean_token_accuracy": 0.8697194457054138, "num_tokens": 480942243.0, "step": 12670 }, { "epoch": 1.6118814400203536, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.517958164215088, "learning_rate": 1e-05, "loss": 0.5487, "mean_token_accuracy": 0.8572410345077515, "num_tokens": 480980510.0, "step": 12671 }, { "epoch": 1.6120086502989441, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.668207168579102, "learning_rate": 1e-05, "loss": 0.4711, "mean_token_accuracy": 0.8752080798149109, "num_tokens": 481016837.0, "step": 12672 }, { "epoch": 1.6121358605775347, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.787413597106934, "learning_rate": 1e-05, "loss": 0.51, "mean_token_accuracy": 0.8688458204269409, "num_tokens": 481055862.0, "step": 12673 }, { "epoch": 1.6122630708561252, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.581689834594727, "learning_rate": 1e-05, "loss": 0.5109, "mean_token_accuracy": 0.8605665564537048, "num_tokens": 481094632.0, "step": 12674 }, { "epoch": 1.6123902811347157, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 0.0001125335693359375, "grad_norm": 7.157617568969727, "learning_rate": 1e-05, "loss": 0.5172, "mean_token_accuracy": 0.8688867092132568, "num_tokens": 481138468.0, "step": 12675 }, { "epoch": 1.6125174914133062, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.706778526306152, "learning_rate": 1e-05, "loss": 0.4975, "mean_token_accuracy": 0.8649026155471802, "num_tokens": 481182983.0, "step": 12676 }, { "epoch": 1.6126447016918966, "ewc_loss": 0.140625, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 0.00011873245239257812, "grad_norm": 7.41220760345459, "learning_rate": 1e-05, "loss": 0.5116, "mean_token_accuracy": 0.8705930113792419, "num_tokens": 481225805.0, "step": 12677 }, { "epoch": 1.612771911970487, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.113142967224121, "learning_rate": 1e-05, "loss": 0.4905, "mean_token_accuracy": 0.8734851479530334, "num_tokens": 481260532.0, "step": 12678 }, { "epoch": 1.6128991222490776, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 0.00011444091796875, "grad_norm": 7.255882740020752, "learning_rate": 1e-05, "loss": 0.5334, "mean_token_accuracy": 0.8655208945274353, "num_tokens": 481300494.0, "step": 12679 }, { "epoch": 1.6130263325276681, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.403549671173096, "learning_rate": 1e-05, "loss": 0.4643, "mean_token_accuracy": 0.8830293416976929, "num_tokens": 481335244.0, "step": 12680 }, { "epoch": 1.6131535428062587, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.78294563293457, "learning_rate": 1e-05, "loss": 0.4856, "mean_token_accuracy": 0.8746805787086487, "num_tokens": 481370894.0, "step": 12681 }, { "epoch": 1.6132807530848492, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.474979400634766, "learning_rate": 1e-05, "loss": 0.523, "mean_token_accuracy": 0.865697979927063, "num_tokens": 481413570.0, "step": 12682 }, { "epoch": 1.6134079633634397, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.267603874206543, "learning_rate": 1e-05, "loss": 0.4828, "mean_token_accuracy": 0.8724848031997681, "num_tokens": 481461380.0, "step": 12683 }, { "epoch": 1.6135351736420303, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.562256813049316, "learning_rate": 1e-05, "loss": 0.5186, "mean_token_accuracy": 0.8687491416931152, "num_tokens": 481497738.0, "step": 12684 }, { "epoch": 1.6136623839206208, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.858658313751221, "learning_rate": 1e-05, "loss": 0.4524, "mean_token_accuracy": 0.8802636861801147, "num_tokens": 481541000.0, "step": 12685 }, { "epoch": 1.6137895941992113, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.766948699951172, "learning_rate": 1e-05, "loss": 0.4437, "mean_token_accuracy": 0.8871437311172485, "num_tokens": 481578012.0, "step": 12686 }, { "epoch": 1.6139168044778018, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.657320499420166, "learning_rate": 1e-05, "loss": 0.5012, "mean_token_accuracy": 0.8629504442214966, "num_tokens": 481612996.0, "step": 12687 }, { "epoch": 1.6140440147563924, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 7.070964336395264, "learning_rate": 1e-05, "loss": 0.5038, "mean_token_accuracy": 0.8713129162788391, "num_tokens": 481645570.0, "step": 12688 }, { "epoch": 1.614171225034983, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.675413608551025, "learning_rate": 1e-05, "loss": 0.4589, "mean_token_accuracy": 0.876684308052063, "num_tokens": 481687873.0, "step": 12689 }, { "epoch": 1.6142984353135734, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 0.0001163482666015625, "grad_norm": 7.335799217224121, "learning_rate": 1e-05, "loss": 0.5428, "mean_token_accuracy": 0.8601250648498535, "num_tokens": 481728105.0, "step": 12690 }, { "epoch": 1.614425645592164, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 5.961364269256592, "learning_rate": 1e-05, "loss": 0.4531, "mean_token_accuracy": 0.8786212801933289, "num_tokens": 481763558.0, "step": 12691 }, { "epoch": 1.6145528558707545, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 0.00011587142944335938, "grad_norm": 7.337601661682129, "learning_rate": 1e-05, "loss": 0.5592, "mean_token_accuracy": 0.8559930324554443, "num_tokens": 481801153.0, "step": 12692 }, { "epoch": 1.614680066149345, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.256119251251221, "learning_rate": 1e-05, "loss": 0.4398, "mean_token_accuracy": 0.8873748779296875, "num_tokens": 481830523.0, "step": 12693 }, { "epoch": 1.6148072764279355, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 7.0695319175720215, "learning_rate": 1e-05, "loss": 0.5109, "mean_token_accuracy": 0.8673439025878906, "num_tokens": 481870814.0, "step": 12694 }, { "epoch": 1.6149344867065258, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.323603630065918, "learning_rate": 1e-05, "loss": 0.5032, "mean_token_accuracy": 0.8723762035369873, "num_tokens": 481912279.0, "step": 12695 }, { "epoch": 1.6150616969851164, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.755993843078613, "learning_rate": 1e-05, "loss": 0.5276, "mean_token_accuracy": 0.8671468496322632, "num_tokens": 481957506.0, "step": 12696 }, { "epoch": 1.615188907263707, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.263826847076416, "learning_rate": 1e-05, "loss": 0.4697, "mean_token_accuracy": 0.8792263865470886, "num_tokens": 481996718.0, "step": 12697 }, { "epoch": 1.6153161175422974, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.505173683166504, "learning_rate": 1e-05, "loss": 0.51, "mean_token_accuracy": 0.8683005571365356, "num_tokens": 482035891.0, "step": 12698 }, { "epoch": 1.615443327820888, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.161932468414307, "learning_rate": 1e-05, "loss": 0.4921, "mean_token_accuracy": 0.8716176152229309, "num_tokens": 482070381.0, "step": 12699 }, { "epoch": 1.6155705380994785, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.383767127990723, "learning_rate": 1e-05, "loss": 0.4961, "mean_token_accuracy": 0.8732898235321045, "num_tokens": 482114484.0, "step": 12700 }, { "epoch": 1.6156977483780688, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 6.030529975891113, "learning_rate": 1e-05, "loss": 0.4822, "mean_token_accuracy": 0.8705320358276367, "num_tokens": 482147978.0, "step": 12701 }, { "epoch": 1.6158249586566593, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.37822151184082, "learning_rate": 1e-05, "loss": 0.4778, "mean_token_accuracy": 0.8762217164039612, "num_tokens": 482193017.0, "step": 12702 }, { "epoch": 1.6159521689352498, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.886811256408691, "learning_rate": 1e-05, "loss": 0.5305, "mean_token_accuracy": 0.855853796005249, "num_tokens": 482230856.0, "step": 12703 }, { "epoch": 1.6160793792138404, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.471704006195068, "learning_rate": 1e-05, "loss": 0.4639, "mean_token_accuracy": 0.8800682425498962, "num_tokens": 482267540.0, "step": 12704 }, { "epoch": 1.616206589492431, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.741693019866943, "learning_rate": 1e-05, "loss": 0.4434, "mean_token_accuracy": 0.8797111511230469, "num_tokens": 482302256.0, "step": 12705 }, { "epoch": 1.6163337997710214, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.678121566772461, "learning_rate": 1e-05, "loss": 0.4804, "mean_token_accuracy": 0.8763092756271362, "num_tokens": 482340706.0, "step": 12706 }, { "epoch": 1.616461010049612, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.635150909423828, "learning_rate": 1e-05, "loss": 0.4509, "mean_token_accuracy": 0.8784840106964111, "num_tokens": 482377313.0, "step": 12707 }, { "epoch": 1.6165882203282025, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.972527027130127, "learning_rate": 1e-05, "loss": 0.4677, "mean_token_accuracy": 0.8832806944847107, "num_tokens": 482415510.0, "step": 12708 }, { "epoch": 1.616715430606793, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.663703441619873, "learning_rate": 1e-05, "loss": 0.4795, "mean_token_accuracy": 0.8694716691970825, "num_tokens": 482453568.0, "step": 12709 }, { "epoch": 1.6168426408853835, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 0.00011587142944335938, "grad_norm": 7.255411624908447, "learning_rate": 1e-05, "loss": 0.5172, "mean_token_accuracy": 0.8678814768791199, "num_tokens": 482487024.0, "step": 12710 }, { "epoch": 1.616969851163974, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 5.932890892028809, "learning_rate": 1e-05, "loss": 0.4657, "mean_token_accuracy": 0.8797603845596313, "num_tokens": 482524328.0, "step": 12711 }, { "epoch": 1.6170970614425646, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 0.00011539459228515625, "grad_norm": 7.239073276519775, "learning_rate": 1e-05, "loss": 0.4871, "mean_token_accuracy": 0.8791722655296326, "num_tokens": 482557042.0, "step": 12712 }, { "epoch": 1.6172242717211551, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.267028331756592, "learning_rate": 1e-05, "loss": 0.5505, "mean_token_accuracy": 0.8493223190307617, "num_tokens": 482594486.0, "step": 12713 }, { "epoch": 1.6173514819997457, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.839489459991455, "learning_rate": 1e-05, "loss": 0.5039, "mean_token_accuracy": 0.8701421022415161, "num_tokens": 482635335.0, "step": 12714 }, { "epoch": 1.6174786922783362, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.418491840362549, "learning_rate": 1e-05, "loss": 0.4669, "mean_token_accuracy": 0.8807474374771118, "num_tokens": 482678945.0, "step": 12715 }, { "epoch": 1.6176059025569267, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.305302143096924, "learning_rate": 1e-05, "loss": 0.4787, "mean_token_accuracy": 0.8719944953918457, "num_tokens": 482717864.0, "step": 12716 }, { "epoch": 1.6177331128355172, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.524926662445068, "learning_rate": 1e-05, "loss": 0.5027, "mean_token_accuracy": 0.8678761720657349, "num_tokens": 482756361.0, "step": 12717 }, { "epoch": 1.6178603231141078, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.837466716766357, "learning_rate": 1e-05, "loss": 0.4709, "mean_token_accuracy": 0.8744868040084839, "num_tokens": 482794658.0, "step": 12718 }, { "epoch": 1.6179875333926983, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.7878570556640625, "learning_rate": 1e-05, "loss": 0.5098, "mean_token_accuracy": 0.8669655323028564, "num_tokens": 482829426.0, "step": 12719 }, { "epoch": 1.6181147436712886, "ewc_loss": 0.10302734375, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.571496963500977, "learning_rate": 1e-05, "loss": 0.4651, "mean_token_accuracy": 0.8759865760803223, "num_tokens": 482867891.0, "step": 12720 }, { "epoch": 1.6182419539498791, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 0.00011348724365234375, "grad_norm": 7.190430164337158, "learning_rate": 1e-05, "loss": 0.4865, "mean_token_accuracy": 0.8767001628875732, "num_tokens": 482902307.0, "step": 12721 }, { "epoch": 1.6183691642284697, "ewc_loss": 0.10498046875, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.646724224090576, "learning_rate": 1e-05, "loss": 0.4861, "mean_token_accuracy": 0.8677933216094971, "num_tokens": 482939751.0, "step": 12722 }, { "epoch": 1.6184963745070602, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.536257743835449, "learning_rate": 1e-05, "loss": 0.5431, "mean_token_accuracy": 0.8632956743240356, "num_tokens": 482978899.0, "step": 12723 }, { "epoch": 1.6186235847856507, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.104912757873535, "learning_rate": 1e-05, "loss": 0.4697, "mean_token_accuracy": 0.879427433013916, "num_tokens": 483011086.0, "step": 12724 }, { "epoch": 1.6187507950642412, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 0.0001201629638671875, "grad_norm": 7.429779529571533, "learning_rate": 1e-05, "loss": 0.466, "mean_token_accuracy": 0.8847990036010742, "num_tokens": 483054235.0, "step": 12725 }, { "epoch": 1.6188780053428315, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.43745756149292, "learning_rate": 1e-05, "loss": 0.4848, "mean_token_accuracy": 0.8740264177322388, "num_tokens": 483086843.0, "step": 12726 }, { "epoch": 1.619005215621422, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 7.056591033935547, "learning_rate": 1e-05, "loss": 0.4861, "mean_token_accuracy": 0.873415470123291, "num_tokens": 483119198.0, "step": 12727 }, { "epoch": 1.6191324259000126, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.43341064453125, "learning_rate": 1e-05, "loss": 0.4707, "mean_token_accuracy": 0.8801737427711487, "num_tokens": 483160835.0, "step": 12728 }, { "epoch": 1.6192596361786031, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.714351654052734, "learning_rate": 1e-05, "loss": 0.5254, "mean_token_accuracy": 0.863490104675293, "num_tokens": 483194906.0, "step": 12729 }, { "epoch": 1.6193868464571937, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.32081937789917, "learning_rate": 1e-05, "loss": 0.5052, "mean_token_accuracy": 0.8646544218063354, "num_tokens": 483231731.0, "step": 12730 }, { "epoch": 1.6195140567357842, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.472757339477539, "learning_rate": 1e-05, "loss": 0.4919, "mean_token_accuracy": 0.8727969527244568, "num_tokens": 483270058.0, "step": 12731 }, { "epoch": 1.6196412670143747, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.161645889282227, "learning_rate": 1e-05, "loss": 0.4474, "mean_token_accuracy": 0.8842822313308716, "num_tokens": 483303600.0, "step": 12732 }, { "epoch": 1.6197684772929652, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.397315502166748, "learning_rate": 1e-05, "loss": 0.4862, "mean_token_accuracy": 0.8742368221282959, "num_tokens": 483339933.0, "step": 12733 }, { "epoch": 1.6198956875715558, "ewc_loss": 0.10791015625, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.993135929107666, "learning_rate": 1e-05, "loss": 0.4732, "mean_token_accuracy": 0.8771706223487854, "num_tokens": 483377153.0, "step": 12734 }, { "epoch": 1.6200228978501463, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.433516502380371, "learning_rate": 1e-05, "loss": 0.4481, "mean_token_accuracy": 0.8833892345428467, "num_tokens": 483415089.0, "step": 12735 }, { "epoch": 1.6201501081287368, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.830405235290527, "learning_rate": 1e-05, "loss": 0.4808, "mean_token_accuracy": 0.8699659705162048, "num_tokens": 483454023.0, "step": 12736 }, { "epoch": 1.6202773184073274, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.599878787994385, "learning_rate": 1e-05, "loss": 0.4847, "mean_token_accuracy": 0.8720284104347229, "num_tokens": 483493572.0, "step": 12737 }, { "epoch": 1.6204045286859179, "ewc_loss": 0.10302734375, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.677524566650391, "learning_rate": 1e-05, "loss": 0.4897, "mean_token_accuracy": 0.866970419883728, "num_tokens": 483529817.0, "step": 12738 }, { "epoch": 1.6205317389645084, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.888058185577393, "learning_rate": 1e-05, "loss": 0.5208, "mean_token_accuracy": 0.8659685850143433, "num_tokens": 483570905.0, "step": 12739 }, { "epoch": 1.620658949243099, "ewc_loss": 0.10302734375, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.627976894378662, "learning_rate": 1e-05, "loss": 0.4933, "mean_token_accuracy": 0.8666239976882935, "num_tokens": 483618177.0, "step": 12740 }, { "epoch": 1.6207861595216895, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 0.00011348724365234375, "grad_norm": 7.22035026550293, "learning_rate": 1e-05, "loss": 0.5129, "mean_token_accuracy": 0.8738887310028076, "num_tokens": 483651227.0, "step": 12741 }, { "epoch": 1.62091336980028, "ewc_loss": 0.11083984375, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 5.839657783508301, "learning_rate": 1e-05, "loss": 0.468, "mean_token_accuracy": 0.8763843178749084, "num_tokens": 483685577.0, "step": 12742 }, { "epoch": 1.6210405800788705, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 0.00011682510375976562, "grad_norm": 7.340114116668701, "learning_rate": 1e-05, "loss": 0.5571, "mean_token_accuracy": 0.8578628897666931, "num_tokens": 483721842.0, "step": 12743 }, { "epoch": 1.6211677903574608, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.2100348472595215, "learning_rate": 1e-05, "loss": 0.5147, "mean_token_accuracy": 0.8639993667602539, "num_tokens": 483758646.0, "step": 12744 }, { "epoch": 1.6212950006360514, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 7.067058563232422, "learning_rate": 1e-05, "loss": 0.5404, "mean_token_accuracy": 0.858612596988678, "num_tokens": 483789216.0, "step": 12745 }, { "epoch": 1.621422210914642, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.410101890563965, "learning_rate": 1e-05, "loss": 0.4667, "mean_token_accuracy": 0.8805461525917053, "num_tokens": 483832901.0, "step": 12746 }, { "epoch": 1.6215494211932324, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.54567289352417, "learning_rate": 1e-05, "loss": 0.5275, "mean_token_accuracy": 0.8647850155830383, "num_tokens": 483874331.0, "step": 12747 }, { "epoch": 1.621676631471823, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.5019073486328125, "learning_rate": 1e-05, "loss": 0.4935, "mean_token_accuracy": 0.8745640516281128, "num_tokens": 483911608.0, "step": 12748 }, { "epoch": 1.6218038417504135, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 6.039331436157227, "learning_rate": 1e-05, "loss": 0.488, "mean_token_accuracy": 0.8725584745407104, "num_tokens": 483946419.0, "step": 12749 }, { "epoch": 1.6219310520290038, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.66494083404541, "learning_rate": 1e-05, "loss": 0.5269, "mean_token_accuracy": 0.8594521284103394, "num_tokens": 483989740.0, "step": 12750 }, { "epoch": 1.6220582623075943, "ewc_loss": 0.10302734375, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.676516056060791, "learning_rate": 1e-05, "loss": 0.494, "mean_token_accuracy": 0.8634488582611084, "num_tokens": 484025672.0, "step": 12751 }, { "epoch": 1.6221854725861848, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.990026473999023, "learning_rate": 1e-05, "loss": 0.5058, "mean_token_accuracy": 0.8734629154205322, "num_tokens": 484060421.0, "step": 12752 }, { "epoch": 1.6223126828647754, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.574133396148682, "learning_rate": 1e-05, "loss": 0.4757, "mean_token_accuracy": 0.8725433349609375, "num_tokens": 484102238.0, "step": 12753 }, { "epoch": 1.622439893143366, "ewc_loss": 0.140625, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 0.00011873245239257812, "grad_norm": 7.390828609466553, "learning_rate": 1e-05, "loss": 0.4974, "mean_token_accuracy": 0.8751961588859558, "num_tokens": 484138314.0, "step": 12754 }, { "epoch": 1.6225671034219564, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 5.871135234832764, "learning_rate": 1e-05, "loss": 0.5401, "mean_token_accuracy": 0.8508530855178833, "num_tokens": 484175541.0, "step": 12755 }, { "epoch": 1.622694313700547, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 0.00012159347534179688, "grad_norm": 7.528319358825684, "learning_rate": 1e-05, "loss": 0.4921, "mean_token_accuracy": 0.8833421468734741, "num_tokens": 484211276.0, "step": 12756 }, { "epoch": 1.6228215239791375, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.417957305908203, "learning_rate": 1e-05, "loss": 0.5707, "mean_token_accuracy": 0.8509170413017273, "num_tokens": 484250192.0, "step": 12757 }, { "epoch": 1.622948734257728, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 0.00011110305786132812, "grad_norm": 7.131148815155029, "learning_rate": 1e-05, "loss": 0.4798, "mean_token_accuracy": 0.8775604963302612, "num_tokens": 484287330.0, "step": 12758 }, { "epoch": 1.6230759445363185, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.421050548553467, "learning_rate": 1e-05, "loss": 0.5195, "mean_token_accuracy": 0.865388035774231, "num_tokens": 484319937.0, "step": 12759 }, { "epoch": 1.623203154814909, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.00022029876708984375, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.781978607177734, "learning_rate": 1e-05, "loss": 0.5094, "mean_token_accuracy": 0.8655388355255127, "num_tokens": 484353305.0, "step": 12760 }, { "epoch": 1.6233303650934996, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.00022029876708984375, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.328519344329834, "learning_rate": 1e-05, "loss": 0.5444, "mean_token_accuracy": 0.8574429750442505, "num_tokens": 484391970.0, "step": 12761 }, { "epoch": 1.6234575753720901, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.00022029876708984375, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.505814075469971, "learning_rate": 1e-05, "loss": 0.4708, "mean_token_accuracy": 0.8773125410079956, "num_tokens": 484433448.0, "step": 12762 }, { "epoch": 1.6235847856506807, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.00022029876708984375, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.194474697113037, "learning_rate": 1e-05, "loss": 0.4812, "mean_token_accuracy": 0.8765888214111328, "num_tokens": 484469729.0, "step": 12763 }, { "epoch": 1.6237119959292712, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.00022029876708984375, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.404609203338623, "learning_rate": 1e-05, "loss": 0.5319, "mean_token_accuracy": 0.856984555721283, "num_tokens": 484509474.0, "step": 12764 }, { "epoch": 1.6238392062078617, "ewc_loss": 0.109375, "ewc_loss_diag": 0.00022029876708984375, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 6.021751880645752, "learning_rate": 1e-05, "loss": 0.4336, "mean_token_accuracy": 0.8881840109825134, "num_tokens": 484548999.0, "step": 12765 }, { "epoch": 1.6239664164864522, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.431704998016357, "learning_rate": 1e-05, "loss": 0.537, "mean_token_accuracy": 0.8544344902038574, "num_tokens": 484580534.0, "step": 12766 }, { "epoch": 1.6240936267650428, "ewc_loss": 0.10498046875, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.833491802215576, "learning_rate": 1e-05, "loss": 0.4406, "mean_token_accuracy": 0.8856618404388428, "num_tokens": 484616183.0, "step": 12767 }, { "epoch": 1.6242208370436333, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.575654983520508, "learning_rate": 1e-05, "loss": 0.4921, "mean_token_accuracy": 0.8679976463317871, "num_tokens": 484658290.0, "step": 12768 }, { "epoch": 1.6243480473222236, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.00022029876708984375, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.686551570892334, "learning_rate": 1e-05, "loss": 0.4381, "mean_token_accuracy": 0.8815833330154419, "num_tokens": 484690226.0, "step": 12769 }, { "epoch": 1.6244752576008141, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.834968090057373, "learning_rate": 1e-05, "loss": 0.5116, "mean_token_accuracy": 0.8666271567344666, "num_tokens": 484731911.0, "step": 12770 }, { "epoch": 1.6246024678794047, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.00022029876708984375, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.626590728759766, "learning_rate": 1e-05, "loss": 0.4736, "mean_token_accuracy": 0.871639609336853, "num_tokens": 484768648.0, "step": 12771 }, { "epoch": 1.6247296781579952, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 0.00011348724365234375, "grad_norm": 7.155203342437744, "learning_rate": 1e-05, "loss": 0.5486, "mean_token_accuracy": 0.8605200052261353, "num_tokens": 484807266.0, "step": 12772 }, { "epoch": 1.6248568884365857, "ewc_loss": 0.10888671875, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.780571460723877, "learning_rate": 1e-05, "loss": 0.5231, "mean_token_accuracy": 0.8574620485305786, "num_tokens": 484850877.0, "step": 12773 }, { "epoch": 1.6249840987151762, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.00022029876708984375, "ewc_loss_parallel": 0.00011777877807617188, "grad_norm": 7.323033332824707, "learning_rate": 1e-05, "loss": 0.5001, "mean_token_accuracy": 0.8747671842575073, "num_tokens": 484891223.0, "step": 12774 }, { "epoch": 1.6251113089937665, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.00022029876708984375, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.139433860778809, "learning_rate": 1e-05, "loss": 0.4734, "mean_token_accuracy": 0.8746912479400635, "num_tokens": 484922081.0, "step": 12775 }, { "epoch": 1.625238519272357, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.00022029876708984375, "ewc_loss_parallel": 0.00011110305786132812, "grad_norm": 7.100143909454346, "learning_rate": 1e-05, "loss": 0.5386, "mean_token_accuracy": 0.8589687347412109, "num_tokens": 484959016.0, "step": 12776 }, { "epoch": 1.6253657295509476, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.3775177001953125, "learning_rate": 1e-05, "loss": 0.4982, "mean_token_accuracy": 0.8694741725921631, "num_tokens": 484998493.0, "step": 12777 }, { "epoch": 1.6254929398295381, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.00022029876708984375, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.62376070022583, "learning_rate": 1e-05, "loss": 0.4973, "mean_token_accuracy": 0.8709561228752136, "num_tokens": 485035164.0, "step": 12778 }, { "epoch": 1.6256201501081287, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.00022029876708984375, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.461904048919678, "learning_rate": 1e-05, "loss": 0.4769, "mean_token_accuracy": 0.880940854549408, "num_tokens": 485076347.0, "step": 12779 }, { "epoch": 1.6257473603867192, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.00022029876708984375, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 6.118872165679932, "learning_rate": 1e-05, "loss": 0.4715, "mean_token_accuracy": 0.8732356429100037, "num_tokens": 485107940.0, "step": 12780 }, { "epoch": 1.6258745706653097, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.605411052703857, "learning_rate": 1e-05, "loss": 0.5051, "mean_token_accuracy": 0.8670113682746887, "num_tokens": 485145440.0, "step": 12781 }, { "epoch": 1.6260017809439002, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.720090389251709, "learning_rate": 1e-05, "loss": 0.4615, "mean_token_accuracy": 0.874849796295166, "num_tokens": 485182948.0, "step": 12782 }, { "epoch": 1.6261289912224908, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.897034645080566, "learning_rate": 1e-05, "loss": 0.4785, "mean_token_accuracy": 0.8794084787368774, "num_tokens": 485220397.0, "step": 12783 }, { "epoch": 1.6262562015010813, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.559379577636719, "learning_rate": 1e-05, "loss": 0.4705, "mean_token_accuracy": 0.8727517127990723, "num_tokens": 485265382.0, "step": 12784 }, { "epoch": 1.6263834117796718, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 0.00011682510375976562, "grad_norm": 7.31659460067749, "learning_rate": 1e-05, "loss": 0.5189, "mean_token_accuracy": 0.8699663281440735, "num_tokens": 485306558.0, "step": 12785 }, { "epoch": 1.6265106220582624, "ewc_loss": 0.109375, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 5.764903545379639, "learning_rate": 1e-05, "loss": 0.5087, "mean_token_accuracy": 0.8678926229476929, "num_tokens": 485341278.0, "step": 12786 }, { "epoch": 1.6266378323368529, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 0.0001220703125, "grad_norm": 7.557470798492432, "learning_rate": 1e-05, "loss": 0.4985, "mean_token_accuracy": 0.8761093616485596, "num_tokens": 485376803.0, "step": 12787 }, { "epoch": 1.6267650426154434, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.00022029876708984375, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.235641956329346, "learning_rate": 1e-05, "loss": 0.5303, "mean_token_accuracy": 0.858846127986908, "num_tokens": 485411168.0, "step": 12788 }, { "epoch": 1.626892252894034, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.00022029876708984375, "ewc_loss_parallel": 0.00011730194091796875, "grad_norm": 7.331303119659424, "learning_rate": 1e-05, "loss": 0.5459, "mean_token_accuracy": 0.8617147207260132, "num_tokens": 485456320.0, "step": 12789 }, { "epoch": 1.6270194631726245, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.00022029876708984375, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.466787338256836, "learning_rate": 1e-05, "loss": 0.5135, "mean_token_accuracy": 0.8620777130126953, "num_tokens": 485491894.0, "step": 12790 }, { "epoch": 1.627146673451215, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.00022029876708984375, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.93709659576416, "learning_rate": 1e-05, "loss": 0.5576, "mean_token_accuracy": 0.8577079772949219, "num_tokens": 485526232.0, "step": 12791 }, { "epoch": 1.6272738837298055, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.00022029876708984375, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.423868179321289, "learning_rate": 1e-05, "loss": 0.5073, "mean_token_accuracy": 0.8675294518470764, "num_tokens": 485564752.0, "step": 12792 }, { "epoch": 1.6274010940083958, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.00022029876708984375, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.589609622955322, "learning_rate": 1e-05, "loss": 0.4772, "mean_token_accuracy": 0.8757708072662354, "num_tokens": 485600023.0, "step": 12793 }, { "epoch": 1.6275283042869864, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.00022029876708984375, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.285611152648926, "learning_rate": 1e-05, "loss": 0.4787, "mean_token_accuracy": 0.8723684549331665, "num_tokens": 485639195.0, "step": 12794 }, { "epoch": 1.6276555145655769, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.00022029876708984375, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.4052605628967285, "learning_rate": 1e-05, "loss": 0.4704, "mean_token_accuracy": 0.878027617931366, "num_tokens": 485676619.0, "step": 12795 }, { "epoch": 1.6277827248441674, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.00022029876708984375, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.121722221374512, "learning_rate": 1e-05, "loss": 0.5315, "mean_token_accuracy": 0.8555484414100647, "num_tokens": 485713739.0, "step": 12796 }, { "epoch": 1.627909935122758, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.00022029876708984375, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.372218132019043, "learning_rate": 1e-05, "loss": 0.4996, "mean_token_accuracy": 0.8677935600280762, "num_tokens": 485750010.0, "step": 12797 }, { "epoch": 1.6280371454013485, "ewc_loss": 0.10693359375, "ewc_loss_diag": 0.00022029876708984375, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.952358722686768, "learning_rate": 1e-05, "loss": 0.5301, "mean_token_accuracy": 0.8545030355453491, "num_tokens": 485790754.0, "step": 12798 }, { "epoch": 1.6281643556799388, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.00022029876708984375, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.45556640625, "learning_rate": 1e-05, "loss": 0.4796, "mean_token_accuracy": 0.873810350894928, "num_tokens": 485825472.0, "step": 12799 }, { "epoch": 1.6282915659585293, "ewc_loss": 0.10498046875, "ewc_loss_diag": 0.00022029876708984375, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.775290012359619, "learning_rate": 1e-05, "loss": 0.4668, "mean_token_accuracy": 0.8720470666885376, "num_tokens": 485866349.0, "step": 12800 }, { "epoch": 1.6284187762371198, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.00022029876708984375, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.6598029136657715, "learning_rate": 1e-05, "loss": 0.497, "mean_token_accuracy": 0.872447669506073, "num_tokens": 485908323.0, "step": 12801 }, { "epoch": 1.6285459865157104, "ewc_loss": 0.10302734375, "ewc_loss_diag": 0.00022029876708984375, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.6202073097229, "learning_rate": 1e-05, "loss": 0.4491, "mean_token_accuracy": 0.8802903890609741, "num_tokens": 485948637.0, "step": 12802 }, { "epoch": 1.628673196794301, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.00022029876708984375, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 7.015283107757568, "learning_rate": 1e-05, "loss": 0.5162, "mean_token_accuracy": 0.8691625595092773, "num_tokens": 485988824.0, "step": 12803 }, { "epoch": 1.6288004070728914, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.00022029876708984375, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.637157440185547, "learning_rate": 1e-05, "loss": 0.475, "mean_token_accuracy": 0.8743454813957214, "num_tokens": 486026364.0, "step": 12804 }, { "epoch": 1.628927617351482, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.00022029876708984375, "ewc_loss_parallel": 0.00011730194091796875, "grad_norm": 7.364177227020264, "learning_rate": 1e-05, "loss": 0.5292, "mean_token_accuracy": 0.8665010333061218, "num_tokens": 486069578.0, "step": 12805 }, { "epoch": 1.6290548276300725, "ewc_loss": 0.11376953125, "ewc_loss_diag": 0.00022029876708984375, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 5.942442893981934, "learning_rate": 1e-05, "loss": 0.5075, "mean_token_accuracy": 0.8672170639038086, "num_tokens": 486105847.0, "step": 12806 }, { "epoch": 1.629182037908663, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.00022029876708984375, "ewc_loss_parallel": 0.0001201629638671875, "grad_norm": 7.38990592956543, "learning_rate": 1e-05, "loss": 0.4818, "mean_token_accuracy": 0.8837547898292542, "num_tokens": 486146680.0, "step": 12807 }, { "epoch": 1.6293092481872535, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.00022029876708984375, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.337851524353027, "learning_rate": 1e-05, "loss": 0.4785, "mean_token_accuracy": 0.8779506683349609, "num_tokens": 486184461.0, "step": 12808 }, { "epoch": 1.629436458465844, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.00022029876708984375, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 7.020788192749023, "learning_rate": 1e-05, "loss": 0.4691, "mean_token_accuracy": 0.8795154690742493, "num_tokens": 486214084.0, "step": 12809 }, { "epoch": 1.6295636687444346, "ewc_loss": 0.125, "ewc_loss_diag": 0.00022029876708984375, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.483734130859375, "learning_rate": 1e-05, "loss": 0.5057, "mean_token_accuracy": 0.8696569204330444, "num_tokens": 486250932.0, "step": 12810 }, { "epoch": 1.6296908790230251, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.00022029876708984375, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.475239276885986, "learning_rate": 1e-05, "loss": 0.4677, "mean_token_accuracy": 0.8788886070251465, "num_tokens": 486287964.0, "step": 12811 }, { "epoch": 1.6298180893016156, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.00022029876708984375, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.540406703948975, "learning_rate": 1e-05, "loss": 0.5102, "mean_token_accuracy": 0.8679063320159912, "num_tokens": 486321312.0, "step": 12812 }, { "epoch": 1.6299452995802062, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.00022029876708984375, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.993129730224609, "learning_rate": 1e-05, "loss": 0.4851, "mean_token_accuracy": 0.8703805208206177, "num_tokens": 486358992.0, "step": 12813 }, { "epoch": 1.6300725098587967, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.00022029876708984375, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.705002784729004, "learning_rate": 1e-05, "loss": 0.5476, "mean_token_accuracy": 0.8568501472473145, "num_tokens": 486399632.0, "step": 12814 }, { "epoch": 1.6301997201373872, "ewc_loss": 0.10302734375, "ewc_loss_diag": 0.00022029876708984375, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.686883449554443, "learning_rate": 1e-05, "loss": 0.4772, "mean_token_accuracy": 0.8716530799865723, "num_tokens": 486438129.0, "step": 12815 }, { "epoch": 1.6303269304159778, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.00022029876708984375, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 7.003561973571777, "learning_rate": 1e-05, "loss": 0.5046, "mean_token_accuracy": 0.8704774379730225, "num_tokens": 486476167.0, "step": 12816 }, { "epoch": 1.630454140694568, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.00022029876708984375, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.628243923187256, "learning_rate": 1e-05, "loss": 0.4552, "mean_token_accuracy": 0.877295970916748, "num_tokens": 486512065.0, "step": 12817 }, { "epoch": 1.6305813509731586, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.00022029876708984375, "ewc_loss_parallel": 0.00011777877807617188, "grad_norm": 7.348603248596191, "learning_rate": 1e-05, "loss": 0.5652, "mean_token_accuracy": 0.8527606725692749, "num_tokens": 486555723.0, "step": 12818 }, { "epoch": 1.6307085612517491, "ewc_loss": 0.11376953125, "ewc_loss_diag": 0.00022029876708984375, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 5.885880470275879, "learning_rate": 1e-05, "loss": 0.5398, "mean_token_accuracy": 0.8577702641487122, "num_tokens": 486598598.0, "step": 12819 }, { "epoch": 1.6308357715303397, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.00022029876708984375, "ewc_loss_parallel": 0.00011968612670898438, "grad_norm": 7.459527015686035, "learning_rate": 1e-05, "loss": 0.5187, "mean_token_accuracy": 0.8675216436386108, "num_tokens": 486633426.0, "step": 12820 }, { "epoch": 1.6309629818089302, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.00022029876708984375, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.262649059295654, "learning_rate": 1e-05, "loss": 0.4956, "mean_token_accuracy": 0.8736759424209595, "num_tokens": 486675584.0, "step": 12821 }, { "epoch": 1.6310901920875207, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.00022029876708984375, "ewc_loss_parallel": 0.00011205673217773438, "grad_norm": 7.19365930557251, "learning_rate": 1e-05, "loss": 0.4963, "mean_token_accuracy": 0.872341513633728, "num_tokens": 486722083.0, "step": 12822 }, { "epoch": 1.6312174023661112, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.00022029876708984375, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.40640115737915, "learning_rate": 1e-05, "loss": 0.4756, "mean_token_accuracy": 0.8786154985427856, "num_tokens": 486762658.0, "step": 12823 }, { "epoch": 1.6313446126447015, "ewc_loss": 0.125, "ewc_loss_diag": 0.00022029876708984375, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.836156368255615, "learning_rate": 1e-05, "loss": 0.5049, "mean_token_accuracy": 0.8681846857070923, "num_tokens": 486799441.0, "step": 12824 }, { "epoch": 1.631471822923292, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.00022029876708984375, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.35341215133667, "learning_rate": 1e-05, "loss": 0.5207, "mean_token_accuracy": 0.8605954647064209, "num_tokens": 486836742.0, "step": 12825 }, { "epoch": 1.6315990332018826, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.00022029876708984375, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.540668487548828, "learning_rate": 1e-05, "loss": 0.573, "mean_token_accuracy": 0.844595730304718, "num_tokens": 486873460.0, "step": 12826 }, { "epoch": 1.6317262434804731, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.00022029876708984375, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.235693454742432, "learning_rate": 1e-05, "loss": 0.4853, "mean_token_accuracy": 0.8710764050483704, "num_tokens": 486909411.0, "step": 12827 }, { "epoch": 1.6318534537590637, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.00022029876708984375, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.384300708770752, "learning_rate": 1e-05, "loss": 0.5132, "mean_token_accuracy": 0.8615965843200684, "num_tokens": 486948427.0, "step": 12828 }, { "epoch": 1.6319806640376542, "ewc_loss": 0.11083984375, "ewc_loss_diag": 0.00022029876708984375, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.087907791137695, "learning_rate": 1e-05, "loss": 0.5519, "mean_token_accuracy": 0.8514043688774109, "num_tokens": 486984594.0, "step": 12829 }, { "epoch": 1.6321078743162447, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.00022029876708984375, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.3462934494018555, "learning_rate": 1e-05, "loss": 0.4777, "mean_token_accuracy": 0.8771967887878418, "num_tokens": 487028122.0, "step": 12830 }, { "epoch": 1.6322350845948352, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.00022029876708984375, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.925981044769287, "learning_rate": 1e-05, "loss": 0.5122, "mean_token_accuracy": 0.8600425720214844, "num_tokens": 487068731.0, "step": 12831 }, { "epoch": 1.6323622948734258, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.00022029876708984375, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.437653064727783, "learning_rate": 1e-05, "loss": 0.4853, "mean_token_accuracy": 0.8746090531349182, "num_tokens": 487103993.0, "step": 12832 }, { "epoch": 1.6324895051520163, "ewc_loss": 0.10498046875, "ewc_loss_diag": 0.00022029876708984375, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.78433084487915, "learning_rate": 1e-05, "loss": 0.5032, "mean_token_accuracy": 0.86871337890625, "num_tokens": 487139206.0, "step": 12833 }, { "epoch": 1.6326167154306068, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.00022029876708984375, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.612487316131592, "learning_rate": 1e-05, "loss": 0.5954, "mean_token_accuracy": 0.8401768803596497, "num_tokens": 487181553.0, "step": 12834 }, { "epoch": 1.6327439257091974, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.00022029876708984375, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.65226411819458, "learning_rate": 1e-05, "loss": 0.4977, "mean_token_accuracy": 0.8655447959899902, "num_tokens": 487218931.0, "step": 12835 }, { "epoch": 1.6328711359877879, "ewc_loss": 0.125, "ewc_loss_diag": 0.00022029876708984375, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.909992694854736, "learning_rate": 1e-05, "loss": 0.5161, "mean_token_accuracy": 0.8628601431846619, "num_tokens": 487252829.0, "step": 12836 }, { "epoch": 1.6329983462663784, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.00022029876708984375, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.623833179473877, "learning_rate": 1e-05, "loss": 0.4635, "mean_token_accuracy": 0.8759230375289917, "num_tokens": 487292098.0, "step": 12837 }, { "epoch": 1.633125556544969, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.00022029876708984375, "ewc_loss_parallel": 0.00011491775512695312, "grad_norm": 7.246747970581055, "learning_rate": 1e-05, "loss": 0.5199, "mean_token_accuracy": 0.8704299330711365, "num_tokens": 487330639.0, "step": 12838 }, { "epoch": 1.6332527668235595, "ewc_loss": 0.11083984375, "ewc_loss_diag": 0.00022029876708984375, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 5.842297077178955, "learning_rate": 1e-05, "loss": 0.4655, "mean_token_accuracy": 0.8765849471092224, "num_tokens": 487368499.0, "step": 12839 }, { "epoch": 1.63337997710215, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.00022029876708984375, "ewc_loss_parallel": 0.00011777877807617188, "grad_norm": 7.353644847869873, "learning_rate": 1e-05, "loss": 0.487, "mean_token_accuracy": 0.8781331777572632, "num_tokens": 487406626.0, "step": 12840 }, { "epoch": 1.6335071873807405, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.00022029876708984375, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.223102569580078, "learning_rate": 1e-05, "loss": 0.4874, "mean_token_accuracy": 0.8718107342720032, "num_tokens": 487441055.0, "step": 12841 }, { "epoch": 1.6336343976593308, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.00022029876708984375, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 7.046929836273193, "learning_rate": 1e-05, "loss": 0.4431, "mean_token_accuracy": 0.8908414840698242, "num_tokens": 487475973.0, "step": 12842 }, { "epoch": 1.6337616079379214, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.00022029876708984375, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.418809413909912, "learning_rate": 1e-05, "loss": 0.4902, "mean_token_accuracy": 0.8758409023284912, "num_tokens": 487513626.0, "step": 12843 }, { "epoch": 1.6338888182165119, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.00022029876708984375, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.5442705154418945, "learning_rate": 1e-05, "loss": 0.4805, "mean_token_accuracy": 0.8786376714706421, "num_tokens": 487550053.0, "step": 12844 }, { "epoch": 1.6340160284951024, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.00022029876708984375, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.489238262176514, "learning_rate": 1e-05, "loss": 0.4707, "mean_token_accuracy": 0.8781145215034485, "num_tokens": 487586999.0, "step": 12845 }, { "epoch": 1.634143238773693, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.00022029876708984375, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 6.032211780548096, "learning_rate": 1e-05, "loss": 0.4441, "mean_token_accuracy": 0.8812613487243652, "num_tokens": 487631433.0, "step": 12846 }, { "epoch": 1.6342704490522835, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.00022029876708984375, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.6663055419921875, "learning_rate": 1e-05, "loss": 0.4849, "mean_token_accuracy": 0.873502254486084, "num_tokens": 487663259.0, "step": 12847 }, { "epoch": 1.6343976593308738, "ewc_loss": 0.10400390625, "ewc_loss_diag": 0.00022029876708984375, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.699331760406494, "learning_rate": 1e-05, "loss": 0.5129, "mean_token_accuracy": 0.8640918731689453, "num_tokens": 487697616.0, "step": 12848 }, { "epoch": 1.6345248696094643, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.00022029876708984375, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.9586286544799805, "learning_rate": 1e-05, "loss": 0.4792, "mean_token_accuracy": 0.8807408809661865, "num_tokens": 487735693.0, "step": 12849 }, { "epoch": 1.6346520798880548, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.00022029876708984375, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.600802421569824, "learning_rate": 1e-05, "loss": 0.4782, "mean_token_accuracy": 0.8724708557128906, "num_tokens": 487775519.0, "step": 12850 }, { "epoch": 1.6347792901666454, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 0.00011682510375976562, "grad_norm": 7.328064441680908, "learning_rate": 1e-05, "loss": 0.5416, "mean_token_accuracy": 0.862515926361084, "num_tokens": 487814058.0, "step": 12851 }, { "epoch": 1.6349065004452359, "ewc_loss": 0.11181640625, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 5.842578887939453, "learning_rate": 1e-05, "loss": 0.4578, "mean_token_accuracy": 0.8800327181816101, "num_tokens": 487850128.0, "step": 12852 }, { "epoch": 1.6350337107238264, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 0.00012063980102539062, "grad_norm": 7.4697675704956055, "learning_rate": 1e-05, "loss": 0.4788, "mean_token_accuracy": 0.8843742609024048, "num_tokens": 487889465.0, "step": 12853 }, { "epoch": 1.635160921002417, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.2396721839904785, "learning_rate": 1e-05, "loss": 0.5093, "mean_token_accuracy": 0.864831805229187, "num_tokens": 487926264.0, "step": 12854 }, { "epoch": 1.6352881312810075, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 0.00011396408081054688, "grad_norm": 7.240619659423828, "learning_rate": 1e-05, "loss": 0.5069, "mean_token_accuracy": 0.8690131902694702, "num_tokens": 487959872.0, "step": 12855 }, { "epoch": 1.635415341559598, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.00022029876708984375, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.404288291931152, "learning_rate": 1e-05, "loss": 0.5307, "mean_token_accuracy": 0.8614651560783386, "num_tokens": 487997554.0, "step": 12856 }, { "epoch": 1.6355425518381885, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.875484466552734, "learning_rate": 1e-05, "loss": 0.4584, "mean_token_accuracy": 0.886101245880127, "num_tokens": 488029172.0, "step": 12857 }, { "epoch": 1.635669762116779, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.00022029876708984375, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.362680435180664, "learning_rate": 1e-05, "loss": 0.5662, "mean_token_accuracy": 0.8500553369522095, "num_tokens": 488066651.0, "step": 12858 }, { "epoch": 1.6357969723953696, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.577342510223389, "learning_rate": 1e-05, "loss": 0.5311, "mean_token_accuracy": 0.8601934909820557, "num_tokens": 488102207.0, "step": 12859 }, { "epoch": 1.6359241826739601, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.234634876251221, "learning_rate": 1e-05, "loss": 0.4618, "mean_token_accuracy": 0.8806866407394409, "num_tokens": 488142502.0, "step": 12860 }, { "epoch": 1.6360513929525506, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.4088239669799805, "learning_rate": 1e-05, "loss": 0.5417, "mean_token_accuracy": 0.8569773435592651, "num_tokens": 488182334.0, "step": 12861 }, { "epoch": 1.6361786032311412, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.093240261077881, "learning_rate": 1e-05, "loss": 0.4769, "mean_token_accuracy": 0.8739922046661377, "num_tokens": 488219434.0, "step": 12862 }, { "epoch": 1.6363058135097317, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.3811235427856445, "learning_rate": 1e-05, "loss": 0.4743, "mean_token_accuracy": 0.8733007907867432, "num_tokens": 488256961.0, "step": 12863 }, { "epoch": 1.6364330237883222, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.93292236328125, "learning_rate": 1e-05, "loss": 0.5058, "mean_token_accuracy": 0.8640502691268921, "num_tokens": 488297069.0, "step": 12864 }, { "epoch": 1.6365602340669128, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.459616661071777, "learning_rate": 1e-05, "loss": 0.5385, "mean_token_accuracy": 0.8546497821807861, "num_tokens": 488336436.0, "step": 12865 }, { "epoch": 1.636687444345503, "ewc_loss": 0.10498046875, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.7742133140563965, "learning_rate": 1e-05, "loss": 0.485, "mean_token_accuracy": 0.8651833534240723, "num_tokens": 488376145.0, "step": 12866 }, { "epoch": 1.6368146546240936, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.652431011199951, "learning_rate": 1e-05, "loss": 0.4936, "mean_token_accuracy": 0.872193455696106, "num_tokens": 488418319.0, "step": 12867 }, { "epoch": 1.6369418649026841, "ewc_loss": 0.10302734375, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.644568920135498, "learning_rate": 1e-05, "loss": 0.4772, "mean_token_accuracy": 0.8704392910003662, "num_tokens": 488460377.0, "step": 12868 }, { "epoch": 1.6370690751812746, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.98845911026001, "learning_rate": 1e-05, "loss": 0.5336, "mean_token_accuracy": 0.8630393743515015, "num_tokens": 488491794.0, "step": 12869 }, { "epoch": 1.6371962854598652, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.649886608123779, "learning_rate": 1e-05, "loss": 0.5055, "mean_token_accuracy": 0.8627588152885437, "num_tokens": 488534106.0, "step": 12870 }, { "epoch": 1.6373234957384557, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 0.0001163482666015625, "grad_norm": 7.301629066467285, "learning_rate": 1e-05, "loss": 0.5433, "mean_token_accuracy": 0.862467348575592, "num_tokens": 488568198.0, "step": 12871 }, { "epoch": 1.6374507060170462, "ewc_loss": 0.11376953125, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 5.9262189865112305, "learning_rate": 1e-05, "loss": 0.5049, "mean_token_accuracy": 0.8659462928771973, "num_tokens": 488606112.0, "step": 12872 }, { "epoch": 1.6375779162956365, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 0.00011730194091796875, "grad_norm": 7.330422878265381, "learning_rate": 1e-05, "loss": 0.5308, "mean_token_accuracy": 0.8666001558303833, "num_tokens": 488639585.0, "step": 12873 }, { "epoch": 1.637705126574227, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.2867326736450195, "learning_rate": 1e-05, "loss": 0.4497, "mean_token_accuracy": 0.8834700584411621, "num_tokens": 488670789.0, "step": 12874 }, { "epoch": 1.6378323368528176, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.963759422302246, "learning_rate": 1e-05, "loss": 0.5049, "mean_token_accuracy": 0.8691911697387695, "num_tokens": 488707837.0, "step": 12875 }, { "epoch": 1.6379595471314081, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.445346355438232, "learning_rate": 1e-05, "loss": 0.5028, "mean_token_accuracy": 0.8692748546600342, "num_tokens": 488747719.0, "step": 12876 }, { "epoch": 1.6380867574099987, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.4426140785217285, "learning_rate": 1e-05, "loss": 0.5012, "mean_token_accuracy": 0.8639413714408875, "num_tokens": 488786493.0, "step": 12877 }, { "epoch": 1.6382139676885892, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.509426593780518, "learning_rate": 1e-05, "loss": 0.5025, "mean_token_accuracy": 0.8651045560836792, "num_tokens": 488826390.0, "step": 12878 }, { "epoch": 1.6383411779671797, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.975582122802734, "learning_rate": 1e-05, "loss": 0.4933, "mean_token_accuracy": 0.8661473989486694, "num_tokens": 488861458.0, "step": 12879 }, { "epoch": 1.6384683882457702, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.6728997230529785, "learning_rate": 1e-05, "loss": 0.499, "mean_token_accuracy": 0.8720511198043823, "num_tokens": 488902762.0, "step": 12880 }, { "epoch": 1.6385955985243608, "ewc_loss": 0.10302734375, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.673020839691162, "learning_rate": 1e-05, "loss": 0.4331, "mean_token_accuracy": 0.8824526071548462, "num_tokens": 488938139.0, "step": 12881 }, { "epoch": 1.6387228088029513, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.980731010437012, "learning_rate": 1e-05, "loss": 0.4887, "mean_token_accuracy": 0.873249888420105, "num_tokens": 488976036.0, "step": 12882 }, { "epoch": 1.6388500190815418, "ewc_loss": 0.10302734375, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.615496635437012, "learning_rate": 1e-05, "loss": 0.5117, "mean_token_accuracy": 0.8580948710441589, "num_tokens": 489014333.0, "step": 12883 }, { "epoch": 1.6389772293601323, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 0.00011730194091796875, "grad_norm": 7.339411735534668, "learning_rate": 1e-05, "loss": 0.5096, "mean_token_accuracy": 0.8726270794868469, "num_tokens": 489050060.0, "step": 12884 }, { "epoch": 1.6391044396387229, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 5.8585405349731445, "learning_rate": 1e-05, "loss": 0.5033, "mean_token_accuracy": 0.8685489296913147, "num_tokens": 489090053.0, "step": 12885 }, { "epoch": 1.6392316499173134, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 0.00012063980102539062, "grad_norm": 7.4543352127075195, "learning_rate": 1e-05, "loss": 0.5308, "mean_token_accuracy": 0.8675839900970459, "num_tokens": 489129081.0, "step": 12886 }, { "epoch": 1.639358860195904, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.25046968460083, "learning_rate": 1e-05, "loss": 0.5255, "mean_token_accuracy": 0.8599444627761841, "num_tokens": 489174811.0, "step": 12887 }, { "epoch": 1.6394860704744945, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 0.00011301040649414062, "grad_norm": 7.207991123199463, "learning_rate": 1e-05, "loss": 0.5159, "mean_token_accuracy": 0.8690628409385681, "num_tokens": 489214694.0, "step": 12888 }, { "epoch": 1.639613280753085, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.391449928283691, "learning_rate": 1e-05, "loss": 0.496, "mean_token_accuracy": 0.8738042116165161, "num_tokens": 489255299.0, "step": 12889 }, { "epoch": 1.6397404910316755, "ewc_loss": 0.125, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.848808288574219, "learning_rate": 1e-05, "loss": 0.4461, "mean_token_accuracy": 0.8869960308074951, "num_tokens": 489292409.0, "step": 12890 }, { "epoch": 1.6398677013102658, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.354200839996338, "learning_rate": 1e-05, "loss": 0.542, "mean_token_accuracy": 0.8600624203681946, "num_tokens": 489335354.0, "step": 12891 }, { "epoch": 1.6399949115888564, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.559887409210205, "learning_rate": 1e-05, "loss": 0.5729, "mean_token_accuracy": 0.848404049873352, "num_tokens": 489369253.0, "step": 12892 }, { "epoch": 1.6401221218674469, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.237819671630859, "learning_rate": 1e-05, "loss": 0.4978, "mean_token_accuracy": 0.8698667883872986, "num_tokens": 489408453.0, "step": 12893 }, { "epoch": 1.6402493321460374, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.38429069519043, "learning_rate": 1e-05, "loss": 0.4529, "mean_token_accuracy": 0.8812660574913025, "num_tokens": 489441609.0, "step": 12894 }, { "epoch": 1.640376542424628, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.107738971710205, "learning_rate": 1e-05, "loss": 0.4903, "mean_token_accuracy": 0.8708179593086243, "num_tokens": 489477228.0, "step": 12895 }, { "epoch": 1.6405037527032185, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.344329357147217, "learning_rate": 1e-05, "loss": 0.5386, "mean_token_accuracy": 0.8568476438522339, "num_tokens": 489518622.0, "step": 12896 }, { "epoch": 1.6406309629818088, "ewc_loss": 0.10791015625, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.942065715789795, "learning_rate": 1e-05, "loss": 0.5222, "mean_token_accuracy": 0.8584067821502686, "num_tokens": 489557707.0, "step": 12897 }, { "epoch": 1.6407581732603993, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.435810565948486, "learning_rate": 1e-05, "loss": 0.5087, "mean_token_accuracy": 0.8648347854614258, "num_tokens": 489591180.0, "step": 12898 }, { "epoch": 1.6408853835389898, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.7734456062316895, "learning_rate": 1e-05, "loss": 0.4888, "mean_token_accuracy": 0.8687715530395508, "num_tokens": 489631732.0, "step": 12899 }, { "epoch": 1.6410125938175804, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.623234748840332, "learning_rate": 1e-05, "loss": 0.5052, "mean_token_accuracy": 0.8657430410385132, "num_tokens": 489671484.0, "step": 12900 }, { "epoch": 1.6411398040961709, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.635910987854004, "learning_rate": 1e-05, "loss": 0.4844, "mean_token_accuracy": 0.8701172471046448, "num_tokens": 489706929.0, "step": 12901 }, { "epoch": 1.6412670143747614, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.936196804046631, "learning_rate": 1e-05, "loss": 0.515, "mean_token_accuracy": 0.868897020816803, "num_tokens": 489741830.0, "step": 12902 }, { "epoch": 1.641394224653352, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.616348743438721, "learning_rate": 1e-05, "loss": 0.4425, "mean_token_accuracy": 0.8806573152542114, "num_tokens": 489782808.0, "step": 12903 }, { "epoch": 1.6415214349319425, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 0.0001163482666015625, "grad_norm": 7.259388446807861, "learning_rate": 1e-05, "loss": 0.5445, "mean_token_accuracy": 0.8610701560974121, "num_tokens": 489823536.0, "step": 12904 }, { "epoch": 1.641648645210533, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 5.873092174530029, "learning_rate": 1e-05, "loss": 0.4767, "mean_token_accuracy": 0.8734199404716492, "num_tokens": 489854730.0, "step": 12905 }, { "epoch": 1.6417758554891235, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 0.00011682510375976562, "grad_norm": 7.336110591888428, "learning_rate": 1e-05, "loss": 0.5004, "mean_token_accuracy": 0.871164858341217, "num_tokens": 489886836.0, "step": 12906 }, { "epoch": 1.641903065767714, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.241335868835449, "learning_rate": 1e-05, "loss": 0.4927, "mean_token_accuracy": 0.87258380651474, "num_tokens": 489928227.0, "step": 12907 }, { "epoch": 1.6420302760463046, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 7.010631084442139, "learning_rate": 1e-05, "loss": 0.5409, "mean_token_accuracy": 0.8561587929725647, "num_tokens": 489965697.0, "step": 12908 }, { "epoch": 1.642157486324895, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.419067859649658, "learning_rate": 1e-05, "loss": 0.4643, "mean_token_accuracy": 0.8784677386283875, "num_tokens": 490000344.0, "step": 12909 }, { "epoch": 1.6422846966034856, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.512094974517822, "learning_rate": 1e-05, "loss": 0.4334, "mean_token_accuracy": 0.890660285949707, "num_tokens": 490030965.0, "step": 12910 }, { "epoch": 1.6424119068820762, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.489985466003418, "learning_rate": 1e-05, "loss": 0.4797, "mean_token_accuracy": 0.8791761994361877, "num_tokens": 490069386.0, "step": 12911 }, { "epoch": 1.6425391171606667, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 6.0188307762146, "learning_rate": 1e-05, "loss": 0.4578, "mean_token_accuracy": 0.8791882991790771, "num_tokens": 490104040.0, "step": 12912 }, { "epoch": 1.6426663274392572, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.648364543914795, "learning_rate": 1e-05, "loss": 0.525, "mean_token_accuracy": 0.8617398738861084, "num_tokens": 490147883.0, "step": 12913 }, { "epoch": 1.6427935377178478, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.689815521240234, "learning_rate": 1e-05, "loss": 0.4754, "mean_token_accuracy": 0.8702260255813599, "num_tokens": 490183889.0, "step": 12914 }, { "epoch": 1.642920747996438, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.966867923736572, "learning_rate": 1e-05, "loss": 0.4932, "mean_token_accuracy": 0.8717260360717773, "num_tokens": 490224132.0, "step": 12915 }, { "epoch": 1.6430479582750286, "ewc_loss": 0.10400390625, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.601717472076416, "learning_rate": 1e-05, "loss": 0.4892, "mean_token_accuracy": 0.8685678839683533, "num_tokens": 490262745.0, "step": 12916 }, { "epoch": 1.6431751685536191, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 0.00011682510375976562, "grad_norm": 7.331584930419922, "learning_rate": 1e-05, "loss": 0.5312, "mean_token_accuracy": 0.8652585744857788, "num_tokens": 490301449.0, "step": 12917 }, { "epoch": 1.6433023788322096, "ewc_loss": 0.11181640625, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 5.853148937225342, "learning_rate": 1e-05, "loss": 0.5151, "mean_token_accuracy": 0.8576276302337646, "num_tokens": 490328873.0, "step": 12918 }, { "epoch": 1.6434295891108002, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 0.0001201629638671875, "grad_norm": 7.461059093475342, "learning_rate": 1e-05, "loss": 0.511, "mean_token_accuracy": 0.873913586139679, "num_tokens": 490361174.0, "step": 12919 }, { "epoch": 1.6435567993893907, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.224111557006836, "learning_rate": 1e-05, "loss": 0.5287, "mean_token_accuracy": 0.861973226070404, "num_tokens": 490399708.0, "step": 12920 }, { "epoch": 1.6436840096679812, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 0.00011444091796875, "grad_norm": 7.232336044311523, "learning_rate": 1e-05, "loss": 0.4709, "mean_token_accuracy": 0.8823271989822388, "num_tokens": 490433159.0, "step": 12921 }, { "epoch": 1.6438112199465715, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.3924946784973145, "learning_rate": 1e-05, "loss": 0.4891, "mean_token_accuracy": 0.8719133138656616, "num_tokens": 490473335.0, "step": 12922 }, { "epoch": 1.643938430225162, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.8781304359436035, "learning_rate": 1e-05, "loss": 0.54, "mean_token_accuracy": 0.8605471253395081, "num_tokens": 490517091.0, "step": 12923 }, { "epoch": 1.6440656405037526, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.348973274230957, "learning_rate": 1e-05, "loss": 0.5005, "mean_token_accuracy": 0.8670711517333984, "num_tokens": 490550336.0, "step": 12924 }, { "epoch": 1.6441928507823431, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.575018405914307, "learning_rate": 1e-05, "loss": 0.4493, "mean_token_accuracy": 0.8862127065658569, "num_tokens": 490593592.0, "step": 12925 }, { "epoch": 1.6443200610609336, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.2414374351501465, "learning_rate": 1e-05, "loss": 0.4608, "mean_token_accuracy": 0.8802060484886169, "num_tokens": 490629354.0, "step": 12926 }, { "epoch": 1.6444472713395242, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.41206693649292, "learning_rate": 1e-05, "loss": 0.4242, "mean_token_accuracy": 0.8931562304496765, "num_tokens": 490668319.0, "step": 12927 }, { "epoch": 1.6445744816181147, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.086234092712402, "learning_rate": 1e-05, "loss": 0.4732, "mean_token_accuracy": 0.8785111904144287, "num_tokens": 490705978.0, "step": 12928 }, { "epoch": 1.6447016918967052, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.385767459869385, "learning_rate": 1e-05, "loss": 0.5035, "mean_token_accuracy": 0.8713008165359497, "num_tokens": 490742455.0, "step": 12929 }, { "epoch": 1.6448289021752958, "ewc_loss": 0.10693359375, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.937066555023193, "learning_rate": 1e-05, "loss": 0.5292, "mean_token_accuracy": 0.8556442856788635, "num_tokens": 490781733.0, "step": 12930 }, { "epoch": 1.6449561124538863, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.45843505859375, "learning_rate": 1e-05, "loss": 0.4987, "mean_token_accuracy": 0.872280478477478, "num_tokens": 490821077.0, "step": 12931 }, { "epoch": 1.6450833227324768, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.775333404541016, "learning_rate": 1e-05, "loss": 0.4634, "mean_token_accuracy": 0.874631404876709, "num_tokens": 490862725.0, "step": 12932 }, { "epoch": 1.6452105330110673, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.649137020111084, "learning_rate": 1e-05, "loss": 0.5487, "mean_token_accuracy": 0.8565381765365601, "num_tokens": 490909612.0, "step": 12933 }, { "epoch": 1.6453377432896579, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.656161308288574, "learning_rate": 1e-05, "loss": 0.4658, "mean_token_accuracy": 0.8730427026748657, "num_tokens": 490943202.0, "step": 12934 }, { "epoch": 1.6454649535682484, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.966655731201172, "learning_rate": 1e-05, "loss": 0.503, "mean_token_accuracy": 0.8676773905754089, "num_tokens": 490978481.0, "step": 12935 }, { "epoch": 1.645592163846839, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.646028995513916, "learning_rate": 1e-05, "loss": 0.4671, "mean_token_accuracy": 0.8731538653373718, "num_tokens": 491018672.0, "step": 12936 }, { "epoch": 1.6457193741254295, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 0.00011587142944335938, "grad_norm": 7.276989936828613, "learning_rate": 1e-05, "loss": 0.4961, "mean_token_accuracy": 0.8734288215637207, "num_tokens": 491055032.0, "step": 12937 }, { "epoch": 1.64584658440402, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 5.924157619476318, "learning_rate": 1e-05, "loss": 0.5648, "mean_token_accuracy": 0.8448794484138489, "num_tokens": 491089780.0, "step": 12938 }, { "epoch": 1.6459737946826105, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 0.00011587142944335938, "grad_norm": 7.289702415466309, "learning_rate": 1e-05, "loss": 0.4985, "mean_token_accuracy": 0.8737645745277405, "num_tokens": 491124187.0, "step": 12939 }, { "epoch": 1.6461010049612008, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.260900020599365, "learning_rate": 1e-05, "loss": 0.4989, "mean_token_accuracy": 0.8682048320770264, "num_tokens": 491167558.0, "step": 12940 }, { "epoch": 1.6462282152397913, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.936944484710693, "learning_rate": 1e-05, "loss": 0.5331, "mean_token_accuracy": 0.8656011819839478, "num_tokens": 491210054.0, "step": 12941 }, { "epoch": 1.6463554255183819, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.411725997924805, "learning_rate": 1e-05, "loss": 0.4839, "mean_token_accuracy": 0.8754091858863831, "num_tokens": 491256535.0, "step": 12942 }, { "epoch": 1.6464826357969724, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.441127777099609, "learning_rate": 1e-05, "loss": 0.4711, "mean_token_accuracy": 0.8769605159759521, "num_tokens": 491295909.0, "step": 12943 }, { "epoch": 1.646609846075563, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.502251148223877, "learning_rate": 1e-05, "loss": 0.5294, "mean_token_accuracy": 0.8595916032791138, "num_tokens": 491332547.0, "step": 12944 }, { "epoch": 1.6467370563541535, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.962988376617432, "learning_rate": 1e-05, "loss": 0.4321, "mean_token_accuracy": 0.8860807418823242, "num_tokens": 491373187.0, "step": 12945 }, { "epoch": 1.6468642666327438, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.697641372680664, "learning_rate": 1e-05, "loss": 0.499, "mean_token_accuracy": 0.8709971904754639, "num_tokens": 491412476.0, "step": 12946 }, { "epoch": 1.6469914769113343, "ewc_loss": 0.10302734375, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.660285949707031, "learning_rate": 1e-05, "loss": 0.4667, "mean_token_accuracy": 0.8768763542175293, "num_tokens": 491456502.0, "step": 12947 }, { "epoch": 1.6471186871899248, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 7.020137786865234, "learning_rate": 1e-05, "loss": 0.4847, "mean_token_accuracy": 0.8764689564704895, "num_tokens": 491492974.0, "step": 12948 }, { "epoch": 1.6472458974685154, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.612100601196289, "learning_rate": 1e-05, "loss": 0.4776, "mean_token_accuracy": 0.8722387552261353, "num_tokens": 491523621.0, "step": 12949 }, { "epoch": 1.6473731077471059, "ewc_loss": 0.140625, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 0.00011873245239257812, "grad_norm": 7.394832134246826, "learning_rate": 1e-05, "loss": 0.5501, "mean_token_accuracy": 0.8594068288803101, "num_tokens": 491563366.0, "step": 12950 }, { "epoch": 1.6475003180256964, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 5.8953351974487305, "learning_rate": 1e-05, "loss": 0.4833, "mean_token_accuracy": 0.8691490888595581, "num_tokens": 491594221.0, "step": 12951 }, { "epoch": 1.647627528304287, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 0.00012159347534179688, "grad_norm": 7.495805263519287, "learning_rate": 1e-05, "loss": 0.5439, "mean_token_accuracy": 0.8604191541671753, "num_tokens": 491632039.0, "step": 12952 }, { "epoch": 1.6477547385828775, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.287151336669922, "learning_rate": 1e-05, "loss": 0.4837, "mean_token_accuracy": 0.8751827478408813, "num_tokens": 491672282.0, "step": 12953 }, { "epoch": 1.647881948861468, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 0.00011301040649414062, "grad_norm": 7.233331680297852, "learning_rate": 1e-05, "loss": 0.555, "mean_token_accuracy": 0.8575989007949829, "num_tokens": 491707969.0, "step": 12954 }, { "epoch": 1.6480091591400585, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.4459919929504395, "learning_rate": 1e-05, "loss": 0.482, "mean_token_accuracy": 0.87483149766922, "num_tokens": 491743347.0, "step": 12955 }, { "epoch": 1.648136369418649, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.8470563888549805, "learning_rate": 1e-05, "loss": 0.5114, "mean_token_accuracy": 0.8674252033233643, "num_tokens": 491777165.0, "step": 12956 }, { "epoch": 1.6482635796972396, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.382532119750977, "learning_rate": 1e-05, "loss": 0.4527, "mean_token_accuracy": 0.8846421837806702, "num_tokens": 491813859.0, "step": 12957 }, { "epoch": 1.64839078997583, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.532282829284668, "learning_rate": 1e-05, "loss": 0.5177, "mean_token_accuracy": 0.8639874458312988, "num_tokens": 491853373.0, "step": 12958 }, { "epoch": 1.6485180002544206, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.260782718658447, "learning_rate": 1e-05, "loss": 0.4774, "mean_token_accuracy": 0.8736031651496887, "num_tokens": 491889401.0, "step": 12959 }, { "epoch": 1.6486452105330112, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.369277477264404, "learning_rate": 1e-05, "loss": 0.5013, "mean_token_accuracy": 0.8659824132919312, "num_tokens": 491933973.0, "step": 12960 }, { "epoch": 1.6487724208116017, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.108896732330322, "learning_rate": 1e-05, "loss": 0.5155, "mean_token_accuracy": 0.8620705604553223, "num_tokens": 491979198.0, "step": 12961 }, { "epoch": 1.6488996310901922, "ewc_loss": 0.11474609375, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.349937915802002, "learning_rate": 1e-05, "loss": 0.53, "mean_token_accuracy": 0.8583270311355591, "num_tokens": 492009127.0, "step": 12962 }, { "epoch": 1.6490268413687827, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.943750858306885, "learning_rate": 1e-05, "loss": 0.4805, "mean_token_accuracy": 0.8718445301055908, "num_tokens": 492045022.0, "step": 12963 }, { "epoch": 1.649154051647373, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.420956134796143, "learning_rate": 1e-05, "loss": 0.4982, "mean_token_accuracy": 0.8685901165008545, "num_tokens": 492092155.0, "step": 12964 }, { "epoch": 1.6492812619259636, "ewc_loss": 0.10498046875, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.7773895263671875, "learning_rate": 1e-05, "loss": 0.5341, "mean_token_accuracy": 0.8521789312362671, "num_tokens": 492131807.0, "step": 12965 }, { "epoch": 1.649408472204554, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.617269515991211, "learning_rate": 1e-05, "loss": 0.4838, "mean_token_accuracy": 0.87205570936203, "num_tokens": 492166794.0, "step": 12966 }, { "epoch": 1.6495356824831446, "ewc_loss": 0.10302734375, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.6279802322387695, "learning_rate": 1e-05, "loss": 0.476, "mean_token_accuracy": 0.8753308057785034, "num_tokens": 492208375.0, "step": 12967 }, { "epoch": 1.6496628927617352, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.965084552764893, "learning_rate": 1e-05, "loss": 0.5423, "mean_token_accuracy": 0.8571181297302246, "num_tokens": 492241166.0, "step": 12968 }, { "epoch": 1.6497901030403257, "ewc_loss": 0.10302734375, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.604097366333008, "learning_rate": 1e-05, "loss": 0.4399, "mean_token_accuracy": 0.8850459456443787, "num_tokens": 492278887.0, "step": 12969 }, { "epoch": 1.6499173133189162, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 0.00011682510375976562, "grad_norm": 7.334366321563721, "learning_rate": 1e-05, "loss": 0.4597, "mean_token_accuracy": 0.8860272169113159, "num_tokens": 492316523.0, "step": 12970 }, { "epoch": 1.6500445235975065, "ewc_loss": 0.11181640625, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 5.878884315490723, "learning_rate": 1e-05, "loss": 0.5051, "mean_token_accuracy": 0.8639156222343445, "num_tokens": 492354094.0, "step": 12971 }, { "epoch": 1.650171733876097, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 0.0001201629638671875, "grad_norm": 7.444571495056152, "learning_rate": 1e-05, "loss": 0.5451, "mean_token_accuracy": 0.8595851063728333, "num_tokens": 492387250.0, "step": 12972 }, { "epoch": 1.6502989441546876, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.2949299812316895, "learning_rate": 1e-05, "loss": 0.5083, "mean_token_accuracy": 0.867026686668396, "num_tokens": 492423770.0, "step": 12973 }, { "epoch": 1.6504261544332781, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 7.0911946296691895, "learning_rate": 1e-05, "loss": 0.5095, "mean_token_accuracy": 0.8671749830245972, "num_tokens": 492465970.0, "step": 12974 }, { "epoch": 1.6505533647118686, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.488051414489746, "learning_rate": 1e-05, "loss": 0.5231, "mean_token_accuracy": 0.8627995848655701, "num_tokens": 492501668.0, "step": 12975 }, { "epoch": 1.6506805749904592, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.557353973388672, "learning_rate": 1e-05, "loss": 0.5075, "mean_token_accuracy": 0.8651753664016724, "num_tokens": 492541032.0, "step": 12976 }, { "epoch": 1.6508077852690497, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.523443222045898, "learning_rate": 1e-05, "loss": 0.5188, "mean_token_accuracy": 0.8659565448760986, "num_tokens": 492579639.0, "step": 12977 }, { "epoch": 1.6509349955476402, "ewc_loss": 0.10986328125, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 6.081726551055908, "learning_rate": 1e-05, "loss": 0.4962, "mean_token_accuracy": 0.8682660460472107, "num_tokens": 492612324.0, "step": 12978 }, { "epoch": 1.6510622058262308, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.658325672149658, "learning_rate": 1e-05, "loss": 0.5329, "mean_token_accuracy": 0.8630952835083008, "num_tokens": 492648967.0, "step": 12979 }, { "epoch": 1.6511894161048213, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.742758274078369, "learning_rate": 1e-05, "loss": 0.4485, "mean_token_accuracy": 0.8833252191543579, "num_tokens": 492687117.0, "step": 12980 }, { "epoch": 1.6513166263834118, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.924120903015137, "learning_rate": 1e-05, "loss": 0.5259, "mean_token_accuracy": 0.8713937997817993, "num_tokens": 492718760.0, "step": 12981 }, { "epoch": 1.6514438366620023, "ewc_loss": 0.10302734375, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.6332106590271, "learning_rate": 1e-05, "loss": 0.4733, "mean_token_accuracy": 0.8722890615463257, "num_tokens": 492760038.0, "step": 12982 }, { "epoch": 1.6515710469405929, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 0.00011444091796875, "grad_norm": 7.2556304931640625, "learning_rate": 1e-05, "loss": 0.5358, "mean_token_accuracy": 0.8593461513519287, "num_tokens": 492798450.0, "step": 12983 }, { "epoch": 1.6516982572191834, "ewc_loss": 0.109375, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 5.817609786987305, "learning_rate": 1e-05, "loss": 0.4826, "mean_token_accuracy": 0.8727777600288391, "num_tokens": 492835081.0, "step": 12984 }, { "epoch": 1.651825467497774, "ewc_loss": 0.140625, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 0.000118255615234375, "grad_norm": 7.402552127838135, "learning_rate": 1e-05, "loss": 0.4821, "mean_token_accuracy": 0.87804114818573, "num_tokens": 492878255.0, "step": 12985 }, { "epoch": 1.6519526777763645, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.154038906097412, "learning_rate": 1e-05, "loss": 0.5212, "mean_token_accuracy": 0.8582645654678345, "num_tokens": 492920800.0, "step": 12986 }, { "epoch": 1.652079888054955, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 0.00011348724365234375, "grad_norm": 7.2366838455200195, "learning_rate": 1e-05, "loss": 0.5167, "mean_token_accuracy": 0.8681427240371704, "num_tokens": 492957170.0, "step": 12987 }, { "epoch": 1.6522070983335455, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.348326683044434, "learning_rate": 1e-05, "loss": 0.4948, "mean_token_accuracy": 0.8719282150268555, "num_tokens": 492995123.0, "step": 12988 }, { "epoch": 1.6523343086121358, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.931711196899414, "learning_rate": 1e-05, "loss": 0.4907, "mean_token_accuracy": 0.8728638291358948, "num_tokens": 493034538.0, "step": 12989 }, { "epoch": 1.6524615188907263, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.342718124389648, "learning_rate": 1e-05, "loss": 0.5396, "mean_token_accuracy": 0.856670618057251, "num_tokens": 493073218.0, "step": 12990 }, { "epoch": 1.6525887291693169, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.61234188079834, "learning_rate": 1e-05, "loss": 0.497, "mean_token_accuracy": 0.866798996925354, "num_tokens": 493110056.0, "step": 12991 }, { "epoch": 1.6527159394479074, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.240872383117676, "learning_rate": 1e-05, "loss": 0.4669, "mean_token_accuracy": 0.8769659996032715, "num_tokens": 493144324.0, "step": 12992 }, { "epoch": 1.652843149726498, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.429612636566162, "learning_rate": 1e-05, "loss": 0.5229, "mean_token_accuracy": 0.8615762591362, "num_tokens": 493181685.0, "step": 12993 }, { "epoch": 1.6529703600050885, "ewc_loss": 0.11376953125, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.1065545082092285, "learning_rate": 1e-05, "loss": 0.4603, "mean_token_accuracy": 0.8771424293518066, "num_tokens": 493217712.0, "step": 12994 }, { "epoch": 1.6530975702836788, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.378734111785889, "learning_rate": 1e-05, "loss": 0.53, "mean_token_accuracy": 0.8548079133033752, "num_tokens": 493253852.0, "step": 12995 }, { "epoch": 1.6532247805622693, "ewc_loss": 0.10791015625, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.966146469116211, "learning_rate": 1e-05, "loss": 0.4933, "mean_token_accuracy": 0.8708666563034058, "num_tokens": 493287797.0, "step": 12996 }, { "epoch": 1.6533519908408598, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.4254608154296875, "learning_rate": 1e-05, "loss": 0.4842, "mean_token_accuracy": 0.8731642961502075, "num_tokens": 493326072.0, "step": 12997 }, { "epoch": 1.6534792011194503, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.815878391265869, "learning_rate": 1e-05, "loss": 0.5386, "mean_token_accuracy": 0.8528549671173096, "num_tokens": 493365374.0, "step": 12998 }, { "epoch": 1.6536064113980409, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.572650909423828, "learning_rate": 1e-05, "loss": 0.4528, "mean_token_accuracy": 0.8858655691146851, "num_tokens": 493408526.0, "step": 12999 }, { "epoch": 1.6537336216766314, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.6988115310668945, "learning_rate": 1e-05, "loss": 0.5139, "mean_token_accuracy": 0.8592268228530884, "num_tokens": 493445689.0, "step": 13000 }, { "epoch": 1.653860831955222, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.828052520751953, "learning_rate": 1e-05, "loss": 0.5347, "mean_token_accuracy": 0.8598344326019287, "num_tokens": 493480884.0, "step": 13001 }, { "epoch": 1.6539880422338125, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.637280464172363, "learning_rate": 1e-05, "loss": 0.4814, "mean_token_accuracy": 0.8715029954910278, "num_tokens": 493523917.0, "step": 13002 }, { "epoch": 1.654115252512403, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 0.00011110305786132812, "grad_norm": 7.122395992279053, "learning_rate": 1e-05, "loss": 0.5046, "mean_token_accuracy": 0.8686763048171997, "num_tokens": 493560438.0, "step": 13003 }, { "epoch": 1.6542424627909935, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.788787841796875, "learning_rate": 1e-05, "loss": 0.5339, "mean_token_accuracy": 0.8588845133781433, "num_tokens": 493596609.0, "step": 13004 }, { "epoch": 1.654369673069584, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 0.00011587142944335938, "grad_norm": 7.281311511993408, "learning_rate": 1e-05, "loss": 0.5324, "mean_token_accuracy": 0.8640592694282532, "num_tokens": 493634369.0, "step": 13005 }, { "epoch": 1.6544968833481746, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.094658374786377, "learning_rate": 1e-05, "loss": 0.5076, "mean_token_accuracy": 0.8641664385795593, "num_tokens": 493669568.0, "step": 13006 }, { "epoch": 1.654624093626765, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 7.084745407104492, "learning_rate": 1e-05, "loss": 0.4947, "mean_token_accuracy": 0.874442994594574, "num_tokens": 493710136.0, "step": 13007 }, { "epoch": 1.6547513039053556, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.330757141113281, "learning_rate": 1e-05, "loss": 0.5123, "mean_token_accuracy": 0.8654863238334656, "num_tokens": 493750028.0, "step": 13008 }, { "epoch": 1.6548785141839462, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.637844085693359, "learning_rate": 1e-05, "loss": 0.5472, "mean_token_accuracy": 0.8581224083900452, "num_tokens": 493785167.0, "step": 13009 }, { "epoch": 1.6550057244625367, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.44010066986084, "learning_rate": 1e-05, "loss": 0.5363, "mean_token_accuracy": 0.8574294447898865, "num_tokens": 493822724.0, "step": 13010 }, { "epoch": 1.6551329347411272, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 6.119168758392334, "learning_rate": 1e-05, "loss": 0.5192, "mean_token_accuracy": 0.8613529205322266, "num_tokens": 493863910.0, "step": 13011 }, { "epoch": 1.6552601450197177, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.602257251739502, "learning_rate": 1e-05, "loss": 0.4714, "mean_token_accuracy": 0.8783970475196838, "num_tokens": 493901863.0, "step": 13012 }, { "epoch": 1.655387355298308, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.715702056884766, "learning_rate": 1e-05, "loss": 0.4849, "mean_token_accuracy": 0.8711768388748169, "num_tokens": 493938174.0, "step": 13013 }, { "epoch": 1.6555145655768986, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.930831432342529, "learning_rate": 1e-05, "loss": 0.5182, "mean_token_accuracy": 0.8659913539886475, "num_tokens": 493973483.0, "step": 13014 }, { "epoch": 1.655641775855489, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.545551776885986, "learning_rate": 1e-05, "loss": 0.4421, "mean_token_accuracy": 0.8817375898361206, "num_tokens": 494016481.0, "step": 13015 }, { "epoch": 1.6557689861340796, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 0.00011777877807617188, "grad_norm": 7.367758750915527, "learning_rate": 1e-05, "loss": 0.5082, "mean_token_accuracy": 0.8714948892593384, "num_tokens": 494047543.0, "step": 13016 }, { "epoch": 1.6558961964126702, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 5.777431011199951, "learning_rate": 1e-05, "loss": 0.5058, "mean_token_accuracy": 0.8676977753639221, "num_tokens": 494082076.0, "step": 13017 }, { "epoch": 1.6560234066912607, "ewc_loss": 0.1484375, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 0.000125885009765625, "grad_norm": 7.613925933837891, "learning_rate": 1e-05, "loss": 0.5372, "mean_token_accuracy": 0.8662497997283936, "num_tokens": 494123822.0, "step": 13018 }, { "epoch": 1.6561506169698512, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.28591251373291, "learning_rate": 1e-05, "loss": 0.4548, "mean_token_accuracy": 0.8833943605422974, "num_tokens": 494152644.0, "step": 13019 }, { "epoch": 1.6562778272484415, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 0.00011777877807617188, "grad_norm": 7.366504669189453, "learning_rate": 1e-05, "loss": 0.5802, "mean_token_accuracy": 0.8528213500976562, "num_tokens": 494185667.0, "step": 13020 }, { "epoch": 1.656405037527032, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.510164260864258, "learning_rate": 1e-05, "loss": 0.4781, "mean_token_accuracy": 0.8786125779151917, "num_tokens": 494224277.0, "step": 13021 }, { "epoch": 1.6565322478056226, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.958863735198975, "learning_rate": 1e-05, "loss": 0.5262, "mean_token_accuracy": 0.8638755083084106, "num_tokens": 494254475.0, "step": 13022 }, { "epoch": 1.656659458084213, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.4434332847595215, "learning_rate": 1e-05, "loss": 0.5068, "mean_token_accuracy": 0.8684799671173096, "num_tokens": 494291891.0, "step": 13023 }, { "epoch": 1.6567866683628036, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.618116855621338, "learning_rate": 1e-05, "loss": 0.5185, "mean_token_accuracy": 0.8636447191238403, "num_tokens": 494327659.0, "step": 13024 }, { "epoch": 1.6569138786413942, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.309711933135986, "learning_rate": 1e-05, "loss": 0.4467, "mean_token_accuracy": 0.887115478515625, "num_tokens": 494358149.0, "step": 13025 }, { "epoch": 1.6570410889199847, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.4400506019592285, "learning_rate": 1e-05, "loss": 0.4894, "mean_token_accuracy": 0.874417781829834, "num_tokens": 494393673.0, "step": 13026 }, { "epoch": 1.6571682991985752, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.150057315826416, "learning_rate": 1e-05, "loss": 0.4807, "mean_token_accuracy": 0.8710604310035706, "num_tokens": 494422811.0, "step": 13027 }, { "epoch": 1.6572955094771658, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.372712135314941, "learning_rate": 1e-05, "loss": 0.4995, "mean_token_accuracy": 0.8691569566726685, "num_tokens": 494467758.0, "step": 13028 }, { "epoch": 1.6574227197557563, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.972144603729248, "learning_rate": 1e-05, "loss": 0.4351, "mean_token_accuracy": 0.8848456144332886, "num_tokens": 494504698.0, "step": 13029 }, { "epoch": 1.6575499300343468, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.446272373199463, "learning_rate": 1e-05, "loss": 0.5313, "mean_token_accuracy": 0.8601266145706177, "num_tokens": 494541235.0, "step": 13030 }, { "epoch": 1.6576771403129373, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.7989373207092285, "learning_rate": 1e-05, "loss": 0.4793, "mean_token_accuracy": 0.8701362609863281, "num_tokens": 494584499.0, "step": 13031 }, { "epoch": 1.6578043505915279, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.637539386749268, "learning_rate": 1e-05, "loss": 0.5035, "mean_token_accuracy": 0.8674899935722351, "num_tokens": 494625845.0, "step": 13032 }, { "epoch": 1.6579315608701184, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.661179542541504, "learning_rate": 1e-05, "loss": 0.4793, "mean_token_accuracy": 0.8730419874191284, "num_tokens": 494660820.0, "step": 13033 }, { "epoch": 1.658058771148709, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.943103790283203, "learning_rate": 1e-05, "loss": 0.4769, "mean_token_accuracy": 0.8837227821350098, "num_tokens": 494701215.0, "step": 13034 }, { "epoch": 1.6581859814272994, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.639688968658447, "learning_rate": 1e-05, "loss": 0.5751, "mean_token_accuracy": 0.8408738374710083, "num_tokens": 494739062.0, "step": 13035 }, { "epoch": 1.65831319170589, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 0.00011539459228515625, "grad_norm": 7.285440444946289, "learning_rate": 1e-05, "loss": 0.5168, "mean_token_accuracy": 0.8684601783752441, "num_tokens": 494779467.0, "step": 13036 }, { "epoch": 1.6584404019844805, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 5.869255542755127, "learning_rate": 1e-05, "loss": 0.4961, "mean_token_accuracy": 0.868838369846344, "num_tokens": 494817237.0, "step": 13037 }, { "epoch": 1.6585676122630708, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 0.00011730194091796875, "grad_norm": 7.385923385620117, "learning_rate": 1e-05, "loss": 0.463, "mean_token_accuracy": 0.8863073587417603, "num_tokens": 494853485.0, "step": 13038 }, { "epoch": 1.6586948225416613, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.252215385437012, "learning_rate": 1e-05, "loss": 0.4857, "mean_token_accuracy": 0.8748441338539124, "num_tokens": 494890456.0, "step": 13039 }, { "epoch": 1.6588220328202519, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 0.00010967254638671875, "grad_norm": 7.067605495452881, "learning_rate": 1e-05, "loss": 0.5447, "mean_token_accuracy": 0.8573604822158813, "num_tokens": 494929441.0, "step": 13040 }, { "epoch": 1.6589492430988424, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.462603569030762, "learning_rate": 1e-05, "loss": 0.5428, "mean_token_accuracy": 0.8617563247680664, "num_tokens": 494963642.0, "step": 13041 }, { "epoch": 1.659076453377433, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.557110786437988, "learning_rate": 1e-05, "loss": 0.519, "mean_token_accuracy": 0.8608753681182861, "num_tokens": 494998138.0, "step": 13042 }, { "epoch": 1.6592036636560235, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.504887104034424, "learning_rate": 1e-05, "loss": 0.4975, "mean_token_accuracy": 0.8715053200721741, "num_tokens": 495038080.0, "step": 13043 }, { "epoch": 1.6593308739346138, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 6.0767621994018555, "learning_rate": 1e-05, "loss": 0.5123, "mean_token_accuracy": 0.8638543486595154, "num_tokens": 495079103.0, "step": 13044 }, { "epoch": 1.6594580842132043, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.624600887298584, "learning_rate": 1e-05, "loss": 0.4943, "mean_token_accuracy": 0.8709829449653625, "num_tokens": 495119385.0, "step": 13045 }, { "epoch": 1.6595852944917948, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.741065979003906, "learning_rate": 1e-05, "loss": 0.4966, "mean_token_accuracy": 0.8628255128860474, "num_tokens": 495157829.0, "step": 13046 }, { "epoch": 1.6597125047703853, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.9029388427734375, "learning_rate": 1e-05, "loss": 0.5936, "mean_token_accuracy": 0.8446751236915588, "num_tokens": 495198554.0, "step": 13047 }, { "epoch": 1.6598397150489759, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.601449966430664, "learning_rate": 1e-05, "loss": 0.4653, "mean_token_accuracy": 0.8768903613090515, "num_tokens": 495241389.0, "step": 13048 }, { "epoch": 1.6599669253275664, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 0.0001163482666015625, "grad_norm": 7.278079986572266, "learning_rate": 1e-05, "loss": 0.485, "mean_token_accuracy": 0.8805453777313232, "num_tokens": 495274216.0, "step": 13049 }, { "epoch": 1.660094135606157, "ewc_loss": 0.109375, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.789420127868652, "learning_rate": 1e-05, "loss": 0.5106, "mean_token_accuracy": 0.8628369569778442, "num_tokens": 495308941.0, "step": 13050 }, { "epoch": 1.6602213458847475, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 0.00012063980102539062, "grad_norm": 7.47416877746582, "learning_rate": 1e-05, "loss": 0.5127, "mean_token_accuracy": 0.8729425072669983, "num_tokens": 495351668.0, "step": 13051 }, { "epoch": 1.660348556163338, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.182919979095459, "learning_rate": 1e-05, "loss": 0.5246, "mean_token_accuracy": 0.8618068695068359, "num_tokens": 495383402.0, "step": 13052 }, { "epoch": 1.6604757664419285, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 0.00011587142944335938, "grad_norm": 7.296267986297607, "learning_rate": 1e-05, "loss": 0.4776, "mean_token_accuracy": 0.88093101978302, "num_tokens": 495421371.0, "step": 13053 }, { "epoch": 1.660602976720519, "ewc_loss": 0.125, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.390007495880127, "learning_rate": 1e-05, "loss": 0.5347, "mean_token_accuracy": 0.8628610372543335, "num_tokens": 495460692.0, "step": 13054 }, { "epoch": 1.6607301869991096, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.948331832885742, "learning_rate": 1e-05, "loss": 0.4913, "mean_token_accuracy": 0.8734748959541321, "num_tokens": 495493990.0, "step": 13055 }, { "epoch": 1.6608573972777, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.3877668380737305, "learning_rate": 1e-05, "loss": 0.5163, "mean_token_accuracy": 0.8668720722198486, "num_tokens": 495531616.0, "step": 13056 }, { "epoch": 1.6609846075562906, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.619508266448975, "learning_rate": 1e-05, "loss": 0.5216, "mean_token_accuracy": 0.8606475591659546, "num_tokens": 495572065.0, "step": 13057 }, { "epoch": 1.6611118178348812, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.299058437347412, "learning_rate": 1e-05, "loss": 0.5387, "mean_token_accuracy": 0.8542220592498779, "num_tokens": 495612582.0, "step": 13058 }, { "epoch": 1.6612390281134717, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.399588584899902, "learning_rate": 1e-05, "loss": 0.5013, "mean_token_accuracy": 0.8663451671600342, "num_tokens": 495649211.0, "step": 13059 }, { "epoch": 1.6613662383920622, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.1658196449279785, "learning_rate": 1e-05, "loss": 0.4965, "mean_token_accuracy": 0.8677407503128052, "num_tokens": 495689805.0, "step": 13060 }, { "epoch": 1.6614934486706527, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.341686725616455, "learning_rate": 1e-05, "loss": 0.5781, "mean_token_accuracy": 0.8434038758277893, "num_tokens": 495727560.0, "step": 13061 }, { "epoch": 1.661620658949243, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 6.017016887664795, "learning_rate": 1e-05, "loss": 0.4785, "mean_token_accuracy": 0.8702688217163086, "num_tokens": 495759827.0, "step": 13062 }, { "epoch": 1.6617478692278336, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.363199234008789, "learning_rate": 1e-05, "loss": 0.5041, "mean_token_accuracy": 0.8708582520484924, "num_tokens": 495797106.0, "step": 13063 }, { "epoch": 1.661875079506424, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.841968536376953, "learning_rate": 1e-05, "loss": 0.5066, "mean_token_accuracy": 0.8623408079147339, "num_tokens": 495842331.0, "step": 13064 }, { "epoch": 1.6620022897850146, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.504889488220215, "learning_rate": 1e-05, "loss": 0.502, "mean_token_accuracy": 0.8644182682037354, "num_tokens": 495883491.0, "step": 13065 }, { "epoch": 1.6621295000636052, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.692253112792969, "learning_rate": 1e-05, "loss": 0.5143, "mean_token_accuracy": 0.8571388125419617, "num_tokens": 495924536.0, "step": 13066 }, { "epoch": 1.6622567103421957, "ewc_loss": 0.125, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.797904014587402, "learning_rate": 1e-05, "loss": 0.5379, "mean_token_accuracy": 0.8573620319366455, "num_tokens": 495961783.0, "step": 13067 }, { "epoch": 1.662383920620786, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.58447265625, "learning_rate": 1e-05, "loss": 0.4702, "mean_token_accuracy": 0.8694965243339539, "num_tokens": 495994131.0, "step": 13068 }, { "epoch": 1.6625111308993765, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 0.00011110305786132812, "grad_norm": 7.173213958740234, "learning_rate": 1e-05, "loss": 0.5264, "mean_token_accuracy": 0.8668733835220337, "num_tokens": 496035981.0, "step": 13069 }, { "epoch": 1.662638341177967, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.701629638671875, "learning_rate": 1e-05, "loss": 0.5199, "mean_token_accuracy": 0.8577328324317932, "num_tokens": 496075622.0, "step": 13070 }, { "epoch": 1.6627655514565576, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 0.00012063980102539062, "grad_norm": 7.46181583404541, "learning_rate": 1e-05, "loss": 0.5248, "mean_token_accuracy": 0.8677183389663696, "num_tokens": 496117130.0, "step": 13071 }, { "epoch": 1.662892761735148, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.1085920333862305, "learning_rate": 1e-05, "loss": 0.4359, "mean_token_accuracy": 0.8917977213859558, "num_tokens": 496153309.0, "step": 13072 }, { "epoch": 1.6630199720137386, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 0.00011682510375976562, "grad_norm": 7.331874847412109, "learning_rate": 1e-05, "loss": 0.4796, "mean_token_accuracy": 0.8826327919960022, "num_tokens": 496194955.0, "step": 13073 }, { "epoch": 1.6631471822923292, "ewc_loss": 0.125, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.439514636993408, "learning_rate": 1e-05, "loss": 0.5131, "mean_token_accuracy": 0.8665626645088196, "num_tokens": 496238787.0, "step": 13074 }, { "epoch": 1.6632743925709197, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.848360538482666, "learning_rate": 1e-05, "loss": 0.5363, "mean_token_accuracy": 0.8586739301681519, "num_tokens": 496279606.0, "step": 13075 }, { "epoch": 1.6634016028495102, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.515052795410156, "learning_rate": 1e-05, "loss": 0.4846, "mean_token_accuracy": 0.8755929470062256, "num_tokens": 496315724.0, "step": 13076 }, { "epoch": 1.6635288131281007, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.324193000793457, "learning_rate": 1e-05, "loss": 0.4835, "mean_token_accuracy": 0.8730148673057556, "num_tokens": 496357254.0, "step": 13077 }, { "epoch": 1.6636560234066913, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.578362941741943, "learning_rate": 1e-05, "loss": 0.4945, "mean_token_accuracy": 0.8713352084159851, "num_tokens": 496394356.0, "step": 13078 }, { "epoch": 1.6637832336852818, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.904669761657715, "learning_rate": 1e-05, "loss": 0.4935, "mean_token_accuracy": 0.8652949929237366, "num_tokens": 496431569.0, "step": 13079 }, { "epoch": 1.6639104439638723, "ewc_loss": 0.125, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.7779316902160645, "learning_rate": 1e-05, "loss": 0.4957, "mean_token_accuracy": 0.8724795579910278, "num_tokens": 496465854.0, "step": 13080 }, { "epoch": 1.6640376542424629, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.670332431793213, "learning_rate": 1e-05, "loss": 0.4724, "mean_token_accuracy": 0.8716047406196594, "num_tokens": 496506023.0, "step": 13081 }, { "epoch": 1.6641648645210534, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 7.097898960113525, "learning_rate": 1e-05, "loss": 0.5574, "mean_token_accuracy": 0.8575320243835449, "num_tokens": 496545950.0, "step": 13082 }, { "epoch": 1.664292074799644, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.698714256286621, "learning_rate": 1e-05, "loss": 0.4761, "mean_token_accuracy": 0.8722972273826599, "num_tokens": 496585728.0, "step": 13083 }, { "epoch": 1.6644192850782344, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 0.00011777877807617188, "grad_norm": 7.370670318603516, "learning_rate": 1e-05, "loss": 0.5852, "mean_token_accuracy": 0.8507906794548035, "num_tokens": 496628491.0, "step": 13084 }, { "epoch": 1.664546495356825, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 5.998305797576904, "learning_rate": 1e-05, "loss": 0.5186, "mean_token_accuracy": 0.8613096475601196, "num_tokens": 496665952.0, "step": 13085 }, { "epoch": 1.6646737056354155, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 0.00011777877807617188, "grad_norm": 7.35512113571167, "learning_rate": 1e-05, "loss": 0.4913, "mean_token_accuracy": 0.8788658976554871, "num_tokens": 496708188.0, "step": 13086 }, { "epoch": 1.6648009159140058, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.2866740226745605, "learning_rate": 1e-05, "loss": 0.4684, "mean_token_accuracy": 0.8779122233390808, "num_tokens": 496743079.0, "step": 13087 }, { "epoch": 1.6649281261925963, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 7.088135719299316, "learning_rate": 1e-05, "loss": 0.5149, "mean_token_accuracy": 0.8686771988868713, "num_tokens": 496781364.0, "step": 13088 }, { "epoch": 1.6650553364711869, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.355800151824951, "learning_rate": 1e-05, "loss": 0.473, "mean_token_accuracy": 0.8805017471313477, "num_tokens": 496816419.0, "step": 13089 }, { "epoch": 1.6651825467497774, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.756374835968018, "learning_rate": 1e-05, "loss": 0.4821, "mean_token_accuracy": 0.8754367828369141, "num_tokens": 496851941.0, "step": 13090 }, { "epoch": 1.665309757028368, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.289957046508789, "learning_rate": 1e-05, "loss": 0.4813, "mean_token_accuracy": 0.8745264410972595, "num_tokens": 496884974.0, "step": 13091 }, { "epoch": 1.6654369673069584, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.532832145690918, "learning_rate": 1e-05, "loss": 0.4859, "mean_token_accuracy": 0.8734422922134399, "num_tokens": 496920331.0, "step": 13092 }, { "epoch": 1.6655641775855488, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.155954837799072, "learning_rate": 1e-05, "loss": 0.4982, "mean_token_accuracy": 0.8665971159934998, "num_tokens": 496957599.0, "step": 13093 }, { "epoch": 1.6656913878641393, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.436262607574463, "learning_rate": 1e-05, "loss": 0.4506, "mean_token_accuracy": 0.8822424411773682, "num_tokens": 496992804.0, "step": 13094 }, { "epoch": 1.6658185981427298, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 6.010652542114258, "learning_rate": 1e-05, "loss": 0.4464, "mean_token_accuracy": 0.8823102116584778, "num_tokens": 497025376.0, "step": 13095 }, { "epoch": 1.6659458084213203, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.458439826965332, "learning_rate": 1e-05, "loss": 0.4691, "mean_token_accuracy": 0.8778245449066162, "num_tokens": 497067923.0, "step": 13096 }, { "epoch": 1.6660730186999109, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.848298072814941, "learning_rate": 1e-05, "loss": 0.5323, "mean_token_accuracy": 0.8591565489768982, "num_tokens": 497113753.0, "step": 13097 }, { "epoch": 1.6662002289785014, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.600492477416992, "learning_rate": 1e-05, "loss": 0.4839, "mean_token_accuracy": 0.8728305101394653, "num_tokens": 497148001.0, "step": 13098 }, { "epoch": 1.666327439257092, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.728971481323242, "learning_rate": 1e-05, "loss": 0.5414, "mean_token_accuracy": 0.8495060205459595, "num_tokens": 497186702.0, "step": 13099 }, { "epoch": 1.6664546495356825, "ewc_loss": 0.125, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.8377532958984375, "learning_rate": 1e-05, "loss": 0.5124, "mean_token_accuracy": 0.8704133033752441, "num_tokens": 497224550.0, "step": 13100 }, { "epoch": 1.666581859814273, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.67296028137207, "learning_rate": 1e-05, "loss": 0.5352, "mean_token_accuracy": 0.8502082228660583, "num_tokens": 497261922.0, "step": 13101 }, { "epoch": 1.6667090700928635, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 0.00011014938354492188, "grad_norm": 7.128751754760742, "learning_rate": 1e-05, "loss": 0.5888, "mean_token_accuracy": 0.8451995849609375, "num_tokens": 497302297.0, "step": 13102 }, { "epoch": 1.666836280371454, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.811173439025879, "learning_rate": 1e-05, "loss": 0.495, "mean_token_accuracy": 0.8693763017654419, "num_tokens": 497337788.0, "step": 13103 }, { "epoch": 1.6669634906500446, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 0.00011539459228515625, "grad_norm": 7.257823944091797, "learning_rate": 1e-05, "loss": 0.5464, "mean_token_accuracy": 0.8559920191764832, "num_tokens": 497373455.0, "step": 13104 }, { "epoch": 1.667090700928635, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.118619918823242, "learning_rate": 1e-05, "loss": 0.5042, "mean_token_accuracy": 0.8692517280578613, "num_tokens": 497414743.0, "step": 13105 }, { "epoch": 1.6672179112072256, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 7.039003849029541, "learning_rate": 1e-05, "loss": 0.5053, "mean_token_accuracy": 0.8723846673965454, "num_tokens": 497446769.0, "step": 13106 }, { "epoch": 1.6673451214858162, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.339956760406494, "learning_rate": 1e-05, "loss": 0.4476, "mean_token_accuracy": 0.8886126279830933, "num_tokens": 497488693.0, "step": 13107 }, { "epoch": 1.6674723317644067, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.5799031257629395, "learning_rate": 1e-05, "loss": 0.476, "mean_token_accuracy": 0.8764064311981201, "num_tokens": 497522258.0, "step": 13108 }, { "epoch": 1.6675995420429972, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.451943397521973, "learning_rate": 1e-05, "loss": 0.5161, "mean_token_accuracy": 0.8640784025192261, "num_tokens": 497557175.0, "step": 13109 }, { "epoch": 1.6677267523215877, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 6.050029277801514, "learning_rate": 1e-05, "loss": 0.4494, "mean_token_accuracy": 0.8827608227729797, "num_tokens": 497592880.0, "step": 13110 }, { "epoch": 1.667853962600178, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.645506381988525, "learning_rate": 1e-05, "loss": 0.4941, "mean_token_accuracy": 0.872660756111145, "num_tokens": 497628416.0, "step": 13111 }, { "epoch": 1.6679811728787686, "ewc_loss": 0.1025390625, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.64956521987915, "learning_rate": 1e-05, "loss": 0.516, "mean_token_accuracy": 0.8579087257385254, "num_tokens": 497666739.0, "step": 13112 }, { "epoch": 1.668108383157359, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 7.017581939697266, "learning_rate": 1e-05, "loss": 0.5076, "mean_token_accuracy": 0.8686659336090088, "num_tokens": 497706307.0, "step": 13113 }, { "epoch": 1.6682355934359496, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.5327324867248535, "learning_rate": 1e-05, "loss": 0.5227, "mean_token_accuracy": 0.8552117347717285, "num_tokens": 497747623.0, "step": 13114 }, { "epoch": 1.6683628037145402, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 0.0001220703125, "grad_norm": 7.498327255249023, "learning_rate": 1e-05, "loss": 0.5239, "mean_token_accuracy": 0.8670166730880737, "num_tokens": 497783354.0, "step": 13115 }, { "epoch": 1.6684900139931307, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 5.870277404785156, "learning_rate": 1e-05, "loss": 0.4891, "mean_token_accuracy": 0.871145486831665, "num_tokens": 497818178.0, "step": 13116 }, { "epoch": 1.668617224271721, "ewc_loss": 0.1484375, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 0.000125885009765625, "grad_norm": 7.673733234405518, "learning_rate": 1e-05, "loss": 0.5167, "mean_token_accuracy": 0.870324432849884, "num_tokens": 497852899.0, "step": 13117 }, { "epoch": 1.6687444345503115, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.4086222648620605, "learning_rate": 1e-05, "loss": 0.5038, "mean_token_accuracy": 0.8705986142158508, "num_tokens": 497896743.0, "step": 13118 }, { "epoch": 1.668871644828902, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 0.0001163482666015625, "grad_norm": 7.329848766326904, "learning_rate": 1e-05, "loss": 0.5022, "mean_token_accuracy": 0.8745522499084473, "num_tokens": 497938837.0, "step": 13119 }, { "epoch": 1.6689988551074926, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.5804033279418945, "learning_rate": 1e-05, "loss": 0.4835, "mean_token_accuracy": 0.8787713050842285, "num_tokens": 497970335.0, "step": 13120 }, { "epoch": 1.669126065386083, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.901284694671631, "learning_rate": 1e-05, "loss": 0.5287, "mean_token_accuracy": 0.8640583157539368, "num_tokens": 498003484.0, "step": 13121 }, { "epoch": 1.6692532756646736, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.4584059715271, "learning_rate": 1e-05, "loss": 0.4685, "mean_token_accuracy": 0.8815440535545349, "num_tokens": 498040932.0, "step": 13122 }, { "epoch": 1.6693804859432642, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.580216884613037, "learning_rate": 1e-05, "loss": 0.5847, "mean_token_accuracy": 0.8447224497795105, "num_tokens": 498078911.0, "step": 13123 }, { "epoch": 1.6695076962218547, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.3001251220703125, "learning_rate": 1e-05, "loss": 0.4744, "mean_token_accuracy": 0.8789930939674377, "num_tokens": 498119917.0, "step": 13124 }, { "epoch": 1.6696349065004452, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.407151222229004, "learning_rate": 1e-05, "loss": 0.4745, "mean_token_accuracy": 0.8748844861984253, "num_tokens": 498160359.0, "step": 13125 }, { "epoch": 1.6697621167790357, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.1358747482299805, "learning_rate": 1e-05, "loss": 0.5227, "mean_token_accuracy": 0.8626003265380859, "num_tokens": 498197048.0, "step": 13126 }, { "epoch": 1.6698893270576263, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.3860626220703125, "learning_rate": 1e-05, "loss": 0.5575, "mean_token_accuracy": 0.8523578643798828, "num_tokens": 498237007.0, "step": 13127 }, { "epoch": 1.6700165373362168, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.946992874145508, "learning_rate": 1e-05, "loss": 0.4833, "mean_token_accuracy": 0.8688883781433105, "num_tokens": 498275324.0, "step": 13128 }, { "epoch": 1.6701437476148073, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.478510856628418, "learning_rate": 1e-05, "loss": 0.5459, "mean_token_accuracy": 0.8530123829841614, "num_tokens": 498317018.0, "step": 13129 }, { "epoch": 1.6702709578933979, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.793311595916748, "learning_rate": 1e-05, "loss": 0.4616, "mean_token_accuracy": 0.8712506294250488, "num_tokens": 498353903.0, "step": 13130 }, { "epoch": 1.6703981681719884, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.669887542724609, "learning_rate": 1e-05, "loss": 0.5287, "mean_token_accuracy": 0.859431266784668, "num_tokens": 498390959.0, "step": 13131 }, { "epoch": 1.670525378450579, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.632528781890869, "learning_rate": 1e-05, "loss": 0.4552, "mean_token_accuracy": 0.8778280019760132, "num_tokens": 498433067.0, "step": 13132 }, { "epoch": 1.6706525887291694, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 7.015397071838379, "learning_rate": 1e-05, "loss": 0.536, "mean_token_accuracy": 0.8634755611419678, "num_tokens": 498476020.0, "step": 13133 }, { "epoch": 1.67077979900776, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.6385884284973145, "learning_rate": 1e-05, "loss": 0.4626, "mean_token_accuracy": 0.8726875185966492, "num_tokens": 498515644.0, "step": 13134 }, { "epoch": 1.6709070092863505, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 0.00011968612670898438, "grad_norm": 7.379457473754883, "learning_rate": 1e-05, "loss": 0.5045, "mean_token_accuracy": 0.8731377720832825, "num_tokens": 498551063.0, "step": 13135 }, { "epoch": 1.6710342195649408, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 5.95365571975708, "learning_rate": 1e-05, "loss": 0.5104, "mean_token_accuracy": 0.8664732575416565, "num_tokens": 498588849.0, "step": 13136 }, { "epoch": 1.6711614298435313, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 0.0001201629638671875, "grad_norm": 7.425111293792725, "learning_rate": 1e-05, "loss": 0.5394, "mean_token_accuracy": 0.8640077114105225, "num_tokens": 498626073.0, "step": 13137 }, { "epoch": 1.6712886401221219, "ewc_loss": 0.125, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.358054161071777, "learning_rate": 1e-05, "loss": 0.4957, "mean_token_accuracy": 0.8736473321914673, "num_tokens": 498661523.0, "step": 13138 }, { "epoch": 1.6714158504007124, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 7.03005838394165, "learning_rate": 1e-05, "loss": 0.5363, "mean_token_accuracy": 0.8616653680801392, "num_tokens": 498702075.0, "step": 13139 }, { "epoch": 1.671543060679303, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.510927200317383, "learning_rate": 1e-05, "loss": 0.5323, "mean_token_accuracy": 0.8620477914810181, "num_tokens": 498737425.0, "step": 13140 }, { "epoch": 1.6716702709578934, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.495408535003662, "learning_rate": 1e-05, "loss": 0.5684, "mean_token_accuracy": 0.8474754095077515, "num_tokens": 498779715.0, "step": 13141 }, { "epoch": 1.6717974812364838, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.557020664215088, "learning_rate": 1e-05, "loss": 0.4771, "mean_token_accuracy": 0.8775387406349182, "num_tokens": 498817312.0, "step": 13142 }, { "epoch": 1.6719246915150743, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 6.019383430480957, "learning_rate": 1e-05, "loss": 0.5153, "mean_token_accuracy": 0.8602495193481445, "num_tokens": 498857990.0, "step": 13143 }, { "epoch": 1.6720519017936648, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.711142539978027, "learning_rate": 1e-05, "loss": 0.5123, "mean_token_accuracy": 0.8660097122192383, "num_tokens": 498896275.0, "step": 13144 }, { "epoch": 1.6721791120722553, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.721258163452148, "learning_rate": 1e-05, "loss": 0.5315, "mean_token_accuracy": 0.8593454360961914, "num_tokens": 498933662.0, "step": 13145 }, { "epoch": 1.6723063223508459, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.988147735595703, "learning_rate": 1e-05, "loss": 0.4957, "mean_token_accuracy": 0.8703652024269104, "num_tokens": 498974216.0, "step": 13146 }, { "epoch": 1.6724335326294364, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.646233081817627, "learning_rate": 1e-05, "loss": 0.4901, "mean_token_accuracy": 0.867628812789917, "num_tokens": 499015889.0, "step": 13147 }, { "epoch": 1.672560742908027, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 0.00011730194091796875, "grad_norm": 7.313633441925049, "learning_rate": 1e-05, "loss": 0.4912, "mean_token_accuracy": 0.8795684576034546, "num_tokens": 499056889.0, "step": 13148 }, { "epoch": 1.6726879531866174, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 5.871140003204346, "learning_rate": 1e-05, "loss": 0.4854, "mean_token_accuracy": 0.8705590963363647, "num_tokens": 499097440.0, "step": 13149 }, { "epoch": 1.672815163465208, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 0.00011968612670898438, "grad_norm": 7.443981170654297, "learning_rate": 1e-05, "loss": 0.4996, "mean_token_accuracy": 0.8770067095756531, "num_tokens": 499135450.0, "step": 13150 }, { "epoch": 1.6729423737437985, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.227739334106445, "learning_rate": 1e-05, "loss": 0.5331, "mean_token_accuracy": 0.8589006066322327, "num_tokens": 499178487.0, "step": 13151 }, { "epoch": 1.673069584022389, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 0.00011301040649414062, "grad_norm": 7.219611644744873, "learning_rate": 1e-05, "loss": 0.5191, "mean_token_accuracy": 0.864970326423645, "num_tokens": 499213824.0, "step": 13152 }, { "epoch": 1.6731967943009796, "ewc_loss": 0.125, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.392343044281006, "learning_rate": 1e-05, "loss": 0.5174, "mean_token_accuracy": 0.8680374026298523, "num_tokens": 499252406.0, "step": 13153 }, { "epoch": 1.67332400457957, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.861684799194336, "learning_rate": 1e-05, "loss": 0.5376, "mean_token_accuracy": 0.8607099652290344, "num_tokens": 499295249.0, "step": 13154 }, { "epoch": 1.6734512148581606, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.348606109619141, "learning_rate": 1e-05, "loss": 0.4652, "mean_token_accuracy": 0.8794649839401245, "num_tokens": 499334115.0, "step": 13155 }, { "epoch": 1.6735784251367511, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.574081897735596, "learning_rate": 1e-05, "loss": 0.5224, "mean_token_accuracy": 0.8647218346595764, "num_tokens": 499372181.0, "step": 13156 }, { "epoch": 1.6737056354153417, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.249636173248291, "learning_rate": 1e-05, "loss": 0.5158, "mean_token_accuracy": 0.8609998226165771, "num_tokens": 499413617.0, "step": 13157 }, { "epoch": 1.6738328456939322, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.42043399810791, "learning_rate": 1e-05, "loss": 0.4937, "mean_token_accuracy": 0.870468020439148, "num_tokens": 499443976.0, "step": 13158 }, { "epoch": 1.6739600559725227, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.11488676071167, "learning_rate": 1e-05, "loss": 0.5027, "mean_token_accuracy": 0.8649697303771973, "num_tokens": 499477973.0, "step": 13159 }, { "epoch": 1.674087266251113, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.366957187652588, "learning_rate": 1e-05, "loss": 0.5436, "mean_token_accuracy": 0.8563076257705688, "num_tokens": 499523163.0, "step": 13160 }, { "epoch": 1.6742144765297036, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.9596943855285645, "learning_rate": 1e-05, "loss": 0.5355, "mean_token_accuracy": 0.8529629707336426, "num_tokens": 499567583.0, "step": 13161 }, { "epoch": 1.674341686808294, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.439162731170654, "learning_rate": 1e-05, "loss": 0.5183, "mean_token_accuracy": 0.8615687489509583, "num_tokens": 499603257.0, "step": 13162 }, { "epoch": 1.6744688970868846, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.800601959228516, "learning_rate": 1e-05, "loss": 0.4786, "mean_token_accuracy": 0.8723463416099548, "num_tokens": 499648881.0, "step": 13163 }, { "epoch": 1.6745961073654752, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.615167617797852, "learning_rate": 1e-05, "loss": 0.5132, "mean_token_accuracy": 0.8671606779098511, "num_tokens": 499688632.0, "step": 13164 }, { "epoch": 1.6747233176440657, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.675339221954346, "learning_rate": 1e-05, "loss": 0.4815, "mean_token_accuracy": 0.8694272041320801, "num_tokens": 499723285.0, "step": 13165 }, { "epoch": 1.674850527922656, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.8986687660217285, "learning_rate": 1e-05, "loss": 0.5663, "mean_token_accuracy": 0.8504384756088257, "num_tokens": 499761472.0, "step": 13166 }, { "epoch": 1.6749777382012465, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.65191125869751, "learning_rate": 1e-05, "loss": 0.5123, "mean_token_accuracy": 0.8616195917129517, "num_tokens": 499796997.0, "step": 13167 }, { "epoch": 1.675104948479837, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 0.00011348724365234375, "grad_norm": 7.208677768707275, "learning_rate": 1e-05, "loss": 0.5078, "mean_token_accuracy": 0.8692559003829956, "num_tokens": 499834307.0, "step": 13168 }, { "epoch": 1.6752321587584276, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 5.837578773498535, "learning_rate": 1e-05, "loss": 0.4976, "mean_token_accuracy": 0.8680922985076904, "num_tokens": 499876497.0, "step": 13169 }, { "epoch": 1.675359369037018, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 0.0001163482666015625, "grad_norm": 7.317216873168945, "learning_rate": 1e-05, "loss": 0.5578, "mean_token_accuracy": 0.8586817979812622, "num_tokens": 499917620.0, "step": 13170 }, { "epoch": 1.6754865793156086, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.182875156402588, "learning_rate": 1e-05, "loss": 0.4846, "mean_token_accuracy": 0.8741381764411926, "num_tokens": 499957098.0, "step": 13171 }, { "epoch": 1.6756137895941992, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 7.049393177032471, "learning_rate": 1e-05, "loss": 0.5256, "mean_token_accuracy": 0.8662629127502441, "num_tokens": 500001046.0, "step": 13172 }, { "epoch": 1.6757409998727897, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.403201580047607, "learning_rate": 1e-05, "loss": 0.5222, "mean_token_accuracy": 0.8639313578605652, "num_tokens": 500041443.0, "step": 13173 }, { "epoch": 1.6758682101513802, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.537130832672119, "learning_rate": 1e-05, "loss": 0.4531, "mean_token_accuracy": 0.8852072358131409, "num_tokens": 500076624.0, "step": 13174 }, { "epoch": 1.6759954204299707, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.496552467346191, "learning_rate": 1e-05, "loss": 0.5439, "mean_token_accuracy": 0.8619008660316467, "num_tokens": 500118995.0, "step": 13175 }, { "epoch": 1.6761226307085613, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 6.039622783660889, "learning_rate": 1e-05, "loss": 0.5011, "mean_token_accuracy": 0.8645924925804138, "num_tokens": 500155134.0, "step": 13176 }, { "epoch": 1.6762498409871518, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.672192096710205, "learning_rate": 1e-05, "loss": 0.4746, "mean_token_accuracy": 0.8761763572692871, "num_tokens": 500189921.0, "step": 13177 }, { "epoch": 1.6763770512657423, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.685013294219971, "learning_rate": 1e-05, "loss": 0.5381, "mean_token_accuracy": 0.8520415425300598, "num_tokens": 500230655.0, "step": 13178 }, { "epoch": 1.6765042615443329, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 7.016809463500977, "learning_rate": 1e-05, "loss": 0.5126, "mean_token_accuracy": 0.8674995303153992, "num_tokens": 500270827.0, "step": 13179 }, { "epoch": 1.6766314718229234, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.589559555053711, "learning_rate": 1e-05, "loss": 0.5331, "mean_token_accuracy": 0.8576227426528931, "num_tokens": 500312810.0, "step": 13180 }, { "epoch": 1.676758682101514, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 0.00011968612670898438, "grad_norm": 7.418011665344238, "learning_rate": 1e-05, "loss": 0.553, "mean_token_accuracy": 0.859344482421875, "num_tokens": 500345013.0, "step": 13181 }, { "epoch": 1.6768858923801044, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 5.8862996101379395, "learning_rate": 1e-05, "loss": 0.4818, "mean_token_accuracy": 0.8697878122329712, "num_tokens": 500375628.0, "step": 13182 }, { "epoch": 1.677013102658695, "ewc_loss": 0.1455078125, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.53989839553833, "learning_rate": 1e-05, "loss": 0.4754, "mean_token_accuracy": 0.8852463960647583, "num_tokens": 500410084.0, "step": 13183 }, { "epoch": 1.6771403129372855, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.3067450523376465, "learning_rate": 1e-05, "loss": 0.4484, "mean_token_accuracy": 0.8853187561035156, "num_tokens": 500452221.0, "step": 13184 }, { "epoch": 1.6772675232158758, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 0.00011539459228515625, "grad_norm": 7.25899600982666, "learning_rate": 1e-05, "loss": 0.5155, "mean_token_accuracy": 0.868346095085144, "num_tokens": 500491358.0, "step": 13185 }, { "epoch": 1.6773947334944663, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.461349964141846, "learning_rate": 1e-05, "loss": 0.5302, "mean_token_accuracy": 0.8630891442298889, "num_tokens": 500530363.0, "step": 13186 }, { "epoch": 1.6775219437730569, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.8761749267578125, "learning_rate": 1e-05, "loss": 0.4935, "mean_token_accuracy": 0.8754914999008179, "num_tokens": 500568922.0, "step": 13187 }, { "epoch": 1.6776491540516474, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.393133163452148, "learning_rate": 1e-05, "loss": 0.5318, "mean_token_accuracy": 0.859636664390564, "num_tokens": 500616445.0, "step": 13188 }, { "epoch": 1.677776364330238, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.573360919952393, "learning_rate": 1e-05, "loss": 0.4749, "mean_token_accuracy": 0.8775290250778198, "num_tokens": 500650692.0, "step": 13189 }, { "epoch": 1.6779035746088284, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.2668375968933105, "learning_rate": 1e-05, "loss": 0.5428, "mean_token_accuracy": 0.8596055507659912, "num_tokens": 500692465.0, "step": 13190 }, { "epoch": 1.6780307848874187, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.413769721984863, "learning_rate": 1e-05, "loss": 0.5253, "mean_token_accuracy": 0.8572384119033813, "num_tokens": 500728715.0, "step": 13191 }, { "epoch": 1.6781579951660093, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.113918781280518, "learning_rate": 1e-05, "loss": 0.5055, "mean_token_accuracy": 0.8642611503601074, "num_tokens": 500768310.0, "step": 13192 }, { "epoch": 1.6782852054445998, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.378037452697754, "learning_rate": 1e-05, "loss": 0.4915, "mean_token_accuracy": 0.8727713823318481, "num_tokens": 500809777.0, "step": 13193 }, { "epoch": 1.6784124157231903, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.952192783355713, "learning_rate": 1e-05, "loss": 0.5296, "mean_token_accuracy": 0.859612762928009, "num_tokens": 500851412.0, "step": 13194 }, { "epoch": 1.6785396260017809, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.4751691818237305, "learning_rate": 1e-05, "loss": 0.5641, "mean_token_accuracy": 0.8505533933639526, "num_tokens": 500890896.0, "step": 13195 }, { "epoch": 1.6786668362803714, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.787123203277588, "learning_rate": 1e-05, "loss": 0.4591, "mean_token_accuracy": 0.8769874572753906, "num_tokens": 500926826.0, "step": 13196 }, { "epoch": 1.678794046558962, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.685006618499756, "learning_rate": 1e-05, "loss": 0.5302, "mean_token_accuracy": 0.8607853651046753, "num_tokens": 500960755.0, "step": 13197 }, { "epoch": 1.6789212568375524, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.665850639343262, "learning_rate": 1e-05, "loss": 0.4876, "mean_token_accuracy": 0.8675836324691772, "num_tokens": 500993177.0, "step": 13198 }, { "epoch": 1.679048467116143, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.972766399383545, "learning_rate": 1e-05, "loss": 0.5071, "mean_token_accuracy": 0.8683658838272095, "num_tokens": 501031391.0, "step": 13199 }, { "epoch": 1.6791756773947335, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.660965442657471, "learning_rate": 1e-05, "loss": 0.4606, "mean_token_accuracy": 0.8770788908004761, "num_tokens": 501072010.0, "step": 13200 }, { "epoch": 1.679302887673324, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 0.00011730194091796875, "grad_norm": 7.28729248046875, "learning_rate": 1e-05, "loss": 0.5039, "mean_token_accuracy": 0.874874472618103, "num_tokens": 501112677.0, "step": 13201 }, { "epoch": 1.6794300979519146, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 5.906228542327881, "learning_rate": 1e-05, "loss": 0.4815, "mean_token_accuracy": 0.8714336156845093, "num_tokens": 501156634.0, "step": 13202 }, { "epoch": 1.679557308230505, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 0.00011777877807617188, "grad_norm": 7.365992546081543, "learning_rate": 1e-05, "loss": 0.5137, "mean_token_accuracy": 0.8715685606002808, "num_tokens": 501188194.0, "step": 13203 }, { "epoch": 1.6796845185090956, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.288496494293213, "learning_rate": 1e-05, "loss": 0.5179, "mean_token_accuracy": 0.8636179566383362, "num_tokens": 501216962.0, "step": 13204 }, { "epoch": 1.6798117287876861, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 7.002628803253174, "learning_rate": 1e-05, "loss": 0.4807, "mean_token_accuracy": 0.8827981352806091, "num_tokens": 501261964.0, "step": 13205 }, { "epoch": 1.6799389390662767, "ewc_loss": 0.125, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.441072463989258, "learning_rate": 1e-05, "loss": 0.4942, "mean_token_accuracy": 0.8706099987030029, "num_tokens": 501303701.0, "step": 13206 }, { "epoch": 1.6800661493448672, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.497039794921875, "learning_rate": 1e-05, "loss": 0.4768, "mean_token_accuracy": 0.8785095810890198, "num_tokens": 501338531.0, "step": 13207 }, { "epoch": 1.6801933596234577, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.513949394226074, "learning_rate": 1e-05, "loss": 0.513, "mean_token_accuracy": 0.8655204772949219, "num_tokens": 501375968.0, "step": 13208 }, { "epoch": 1.680320569902048, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 6.014998435974121, "learning_rate": 1e-05, "loss": 0.5002, "mean_token_accuracy": 0.864306628704071, "num_tokens": 501414646.0, "step": 13209 }, { "epoch": 1.6804477801806386, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.680568695068359, "learning_rate": 1e-05, "loss": 0.5032, "mean_token_accuracy": 0.8671607375144958, "num_tokens": 501455035.0, "step": 13210 }, { "epoch": 1.680574990459229, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.699934482574463, "learning_rate": 1e-05, "loss": 0.4896, "mean_token_accuracy": 0.8665986657142639, "num_tokens": 501493198.0, "step": 13211 }, { "epoch": 1.6807022007378196, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.9865827560424805, "learning_rate": 1e-05, "loss": 0.4914, "mean_token_accuracy": 0.8736603856086731, "num_tokens": 501534293.0, "step": 13212 }, { "epoch": 1.6808294110164101, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.617180824279785, "learning_rate": 1e-05, "loss": 0.4856, "mean_token_accuracy": 0.8674895763397217, "num_tokens": 501566614.0, "step": 13213 }, { "epoch": 1.6809566212950007, "ewc_loss": 0.140625, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 0.000118255615234375, "grad_norm": 7.348959445953369, "learning_rate": 1e-05, "loss": 0.5555, "mean_token_accuracy": 0.8578383326530457, "num_tokens": 501607535.0, "step": 13214 }, { "epoch": 1.681083831573591, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 5.862133979797363, "learning_rate": 1e-05, "loss": 0.5097, "mean_token_accuracy": 0.8636335730552673, "num_tokens": 501644208.0, "step": 13215 }, { "epoch": 1.6812110418521815, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 0.0001201629638671875, "grad_norm": 7.483097076416016, "learning_rate": 1e-05, "loss": 0.5765, "mean_token_accuracy": 0.8560600280761719, "num_tokens": 501684980.0, "step": 13216 }, { "epoch": 1.681338252130772, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.250591278076172, "learning_rate": 1e-05, "loss": 0.4911, "mean_token_accuracy": 0.8696237802505493, "num_tokens": 501722603.0, "step": 13217 }, { "epoch": 1.6814654624093626, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 0.00011444091796875, "grad_norm": 7.252005100250244, "learning_rate": 1e-05, "loss": 0.5134, "mean_token_accuracy": 0.871012270450592, "num_tokens": 501755650.0, "step": 13218 }, { "epoch": 1.681592672687953, "ewc_loss": 0.125, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.4273905754089355, "learning_rate": 1e-05, "loss": 0.5188, "mean_token_accuracy": 0.865908145904541, "num_tokens": 501788848.0, "step": 13219 }, { "epoch": 1.6817198829665436, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.872991561889648, "learning_rate": 1e-05, "loss": 0.5113, "mean_token_accuracy": 0.8654945492744446, "num_tokens": 501826396.0, "step": 13220 }, { "epoch": 1.6818470932451342, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.385357856750488, "learning_rate": 1e-05, "loss": 0.5382, "mean_token_accuracy": 0.8565788269042969, "num_tokens": 501863780.0, "step": 13221 }, { "epoch": 1.6819743035237247, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.55687141418457, "learning_rate": 1e-05, "loss": 0.5148, "mean_token_accuracy": 0.8648093938827515, "num_tokens": 501903275.0, "step": 13222 }, { "epoch": 1.6821015138023152, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.270262718200684, "learning_rate": 1e-05, "loss": 0.4234, "mean_token_accuracy": 0.8900982141494751, "num_tokens": 501940397.0, "step": 13223 }, { "epoch": 1.6822287240809057, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.383821964263916, "learning_rate": 1e-05, "loss": 0.5127, "mean_token_accuracy": 0.8685355186462402, "num_tokens": 501979814.0, "step": 13224 }, { "epoch": 1.6823559343594963, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.12746000289917, "learning_rate": 1e-05, "loss": 0.4812, "mean_token_accuracy": 0.8716368079185486, "num_tokens": 502017349.0, "step": 13225 }, { "epoch": 1.6824831446380868, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.342254161834717, "learning_rate": 1e-05, "loss": 0.5241, "mean_token_accuracy": 0.8592588901519775, "num_tokens": 502062441.0, "step": 13226 }, { "epoch": 1.6826103549166773, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.969037055969238, "learning_rate": 1e-05, "loss": 0.4817, "mean_token_accuracy": 0.8678199052810669, "num_tokens": 502099897.0, "step": 13227 }, { "epoch": 1.6827375651952678, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.405959129333496, "learning_rate": 1e-05, "loss": 0.4807, "mean_token_accuracy": 0.8722479343414307, "num_tokens": 502139682.0, "step": 13228 }, { "epoch": 1.6828647754738584, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.812096118927002, "learning_rate": 1e-05, "loss": 0.515, "mean_token_accuracy": 0.8612476587295532, "num_tokens": 502179534.0, "step": 13229 }, { "epoch": 1.682991985752449, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.580474853515625, "learning_rate": 1e-05, "loss": 0.4866, "mean_token_accuracy": 0.8746869564056396, "num_tokens": 502218031.0, "step": 13230 }, { "epoch": 1.6831191960310394, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.665978908538818, "learning_rate": 1e-05, "loss": 0.4665, "mean_token_accuracy": 0.8748729825019836, "num_tokens": 502251409.0, "step": 13231 }, { "epoch": 1.68324640630963, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.884017467498779, "learning_rate": 1e-05, "loss": 0.4971, "mean_token_accuracy": 0.8730298280715942, "num_tokens": 502285757.0, "step": 13232 }, { "epoch": 1.6833736165882205, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.597599983215332, "learning_rate": 1e-05, "loss": 0.4663, "mean_token_accuracy": 0.8714612722396851, "num_tokens": 502325247.0, "step": 13233 }, { "epoch": 1.6835008268668108, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 0.00011539459228515625, "grad_norm": 7.246912479400635, "learning_rate": 1e-05, "loss": 0.5296, "mean_token_accuracy": 0.869992733001709, "num_tokens": 502367691.0, "step": 13234 }, { "epoch": 1.6836280371454013, "ewc_loss": 0.109375, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.785809516906738, "learning_rate": 1e-05, "loss": 0.5256, "mean_token_accuracy": 0.8578565120697021, "num_tokens": 502409084.0, "step": 13235 }, { "epoch": 1.6837552474239919, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 0.0001201629638671875, "grad_norm": 7.457211017608643, "learning_rate": 1e-05, "loss": 0.5192, "mean_token_accuracy": 0.8708741664886475, "num_tokens": 502448478.0, "step": 13236 }, { "epoch": 1.6838824577025824, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.205617427825928, "learning_rate": 1e-05, "loss": 0.5192, "mean_token_accuracy": 0.8627976179122925, "num_tokens": 502481459.0, "step": 13237 }, { "epoch": 1.684009667981173, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 0.00011348724365234375, "grad_norm": 7.219587802886963, "learning_rate": 1e-05, "loss": 0.5887, "mean_token_accuracy": 0.8512407541275024, "num_tokens": 502522300.0, "step": 13238 }, { "epoch": 1.6841368782597634, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.458052635192871, "learning_rate": 1e-05, "loss": 0.505, "mean_token_accuracy": 0.871299684047699, "num_tokens": 502560257.0, "step": 13239 }, { "epoch": 1.6842640885383537, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.718735694885254, "learning_rate": 1e-05, "loss": 0.5345, "mean_token_accuracy": 0.8585202693939209, "num_tokens": 502602815.0, "step": 13240 }, { "epoch": 1.6843912988169443, "ewc_loss": 0.125, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.514803886413574, "learning_rate": 1e-05, "loss": 0.4884, "mean_token_accuracy": 0.8713819980621338, "num_tokens": 502640067.0, "step": 13241 }, { "epoch": 1.6845185090955348, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.20735502243042, "learning_rate": 1e-05, "loss": 0.4961, "mean_token_accuracy": 0.8637599945068359, "num_tokens": 502680351.0, "step": 13242 }, { "epoch": 1.6846457193741253, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.616413116455078, "learning_rate": 1e-05, "loss": 0.4883, "mean_token_accuracy": 0.8764721155166626, "num_tokens": 502715279.0, "step": 13243 }, { "epoch": 1.6847729296527159, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.826694488525391, "learning_rate": 1e-05, "loss": 0.5235, "mean_token_accuracy": 0.8573797941207886, "num_tokens": 502762078.0, "step": 13244 }, { "epoch": 1.6849001399313064, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.842756748199463, "learning_rate": 1e-05, "loss": 0.4607, "mean_token_accuracy": 0.8827967047691345, "num_tokens": 502799886.0, "step": 13245 }, { "epoch": 1.685027350209897, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.659115791320801, "learning_rate": 1e-05, "loss": 0.515, "mean_token_accuracy": 0.8583550453186035, "num_tokens": 502838414.0, "step": 13246 }, { "epoch": 1.6851545604884874, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 0.00011157989501953125, "grad_norm": 7.16837739944458, "learning_rate": 1e-05, "loss": 0.536, "mean_token_accuracy": 0.858932375907898, "num_tokens": 502879504.0, "step": 13247 }, { "epoch": 1.685281770767078, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.74517297744751, "learning_rate": 1e-05, "loss": 0.425, "mean_token_accuracy": 0.8886336088180542, "num_tokens": 502919167.0, "step": 13248 }, { "epoch": 1.6854089810456685, "ewc_loss": 0.140625, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 0.000118255615234375, "grad_norm": 7.399270534515381, "learning_rate": 1e-05, "loss": 0.4497, "mean_token_accuracy": 0.8921937346458435, "num_tokens": 502956548.0, "step": 13249 }, { "epoch": 1.685536191324259, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.068708896636963, "learning_rate": 1e-05, "loss": 0.477, "mean_token_accuracy": 0.875390887260437, "num_tokens": 502993001.0, "step": 13250 }, { "epoch": 1.6856634016028496, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 0.0001163482666015625, "grad_norm": 7.341231822967529, "learning_rate": 1e-05, "loss": 0.5168, "mean_token_accuracy": 0.8708178400993347, "num_tokens": 503030690.0, "step": 13251 }, { "epoch": 1.68579061188144, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.330883502960205, "learning_rate": 1e-05, "loss": 0.5159, "mean_token_accuracy": 0.8636291027069092, "num_tokens": 503074010.0, "step": 13252 }, { "epoch": 1.6859178221600306, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 7.032972812652588, "learning_rate": 1e-05, "loss": 0.503, "mean_token_accuracy": 0.8718850612640381, "num_tokens": 503111077.0, "step": 13253 }, { "epoch": 1.6860450324386211, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.366087913513184, "learning_rate": 1e-05, "loss": 0.4349, "mean_token_accuracy": 0.8924578428268433, "num_tokens": 503146594.0, "step": 13254 }, { "epoch": 1.6861722427172117, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.715434551239014, "learning_rate": 1e-05, "loss": 0.4543, "mean_token_accuracy": 0.883766770362854, "num_tokens": 503181470.0, "step": 13255 }, { "epoch": 1.6862994529958022, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.285765171051025, "learning_rate": 1e-05, "loss": 0.4735, "mean_token_accuracy": 0.881837010383606, "num_tokens": 503215839.0, "step": 13256 }, { "epoch": 1.6864266632743927, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.488853931427002, "learning_rate": 1e-05, "loss": 0.4718, "mean_token_accuracy": 0.8781588673591614, "num_tokens": 503253131.0, "step": 13257 }, { "epoch": 1.686553873552983, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.169764995574951, "learning_rate": 1e-05, "loss": 0.5635, "mean_token_accuracy": 0.8468594551086426, "num_tokens": 503290331.0, "step": 13258 }, { "epoch": 1.6866810838315736, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.401927471160889, "learning_rate": 1e-05, "loss": 0.5349, "mean_token_accuracy": 0.8557800650596619, "num_tokens": 503328551.0, "step": 13259 }, { "epoch": 1.686808294110164, "ewc_loss": 0.109375, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 6.025214672088623, "learning_rate": 1e-05, "loss": 0.6284, "mean_token_accuracy": 0.8338456153869629, "num_tokens": 503365288.0, "step": 13260 }, { "epoch": 1.6869355043887546, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.4242401123046875, "learning_rate": 1e-05, "loss": 0.4709, "mean_token_accuracy": 0.8764352798461914, "num_tokens": 503402941.0, "step": 13261 }, { "epoch": 1.6870627146673451, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.863845348358154, "learning_rate": 1e-05, "loss": 0.524, "mean_token_accuracy": 0.8593588471412659, "num_tokens": 503443174.0, "step": 13262 }, { "epoch": 1.6871899249459357, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.558556079864502, "learning_rate": 1e-05, "loss": 0.4899, "mean_token_accuracy": 0.8737802505493164, "num_tokens": 503481790.0, "step": 13263 }, { "epoch": 1.687317135224526, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.71492862701416, "learning_rate": 1e-05, "loss": 0.4737, "mean_token_accuracy": 0.8713217973709106, "num_tokens": 503518386.0, "step": 13264 }, { "epoch": 1.6874443455031165, "ewc_loss": 0.125, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.81657600402832, "learning_rate": 1e-05, "loss": 0.4884, "mean_token_accuracy": 0.8750881552696228, "num_tokens": 503558985.0, "step": 13265 }, { "epoch": 1.687571555781707, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.626138210296631, "learning_rate": 1e-05, "loss": 0.465, "mean_token_accuracy": 0.8761427402496338, "num_tokens": 503602423.0, "step": 13266 }, { "epoch": 1.6876987660602976, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.00022220611572265625, "ewc_loss_parallel": 0.0001125335693359375, "grad_norm": 7.1810126304626465, "learning_rate": 1e-05, "loss": 0.4886, "mean_token_accuracy": 0.8764241933822632, "num_tokens": 503634300.0, "step": 13267 }, { "epoch": 1.687825976338888, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.00022411346435546875, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.8046159744262695, "learning_rate": 1e-05, "loss": 0.5234, "mean_token_accuracy": 0.8590666055679321, "num_tokens": 503668225.0, "step": 13268 }, { "epoch": 1.6879531866174786, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.00022411346435546875, "ewc_loss_parallel": 0.00011730194091796875, "grad_norm": 7.339390754699707, "learning_rate": 1e-05, "loss": 0.5625, "mean_token_accuracy": 0.8537308573722839, "num_tokens": 503705158.0, "step": 13269 }, { "epoch": 1.6880803968960691, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.00022411346435546875, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.121598720550537, "learning_rate": 1e-05, "loss": 0.5011, "mean_token_accuracy": 0.8702502250671387, "num_tokens": 503741322.0, "step": 13270 }, { "epoch": 1.6882076071746597, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.00022411346435546875, "ewc_loss_parallel": 0.00011205673217773438, "grad_norm": 7.140101909637451, "learning_rate": 1e-05, "loss": 0.4792, "mean_token_accuracy": 0.8821303844451904, "num_tokens": 503777932.0, "step": 13271 }, { "epoch": 1.6883348174532502, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.00022411346435546875, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.370180130004883, "learning_rate": 1e-05, "loss": 0.4607, "mean_token_accuracy": 0.8792843818664551, "num_tokens": 503810175.0, "step": 13272 }, { "epoch": 1.6884620277318407, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.00022411346435546875, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.676878452301025, "learning_rate": 1e-05, "loss": 0.5376, "mean_token_accuracy": 0.8577947020530701, "num_tokens": 503844498.0, "step": 13273 }, { "epoch": 1.6885892380104313, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.00022411346435546875, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.464171409606934, "learning_rate": 1e-05, "loss": 0.4427, "mean_token_accuracy": 0.8884947299957275, "num_tokens": 503878898.0, "step": 13274 }, { "epoch": 1.6887164482890218, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.0002231597900390625, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.165600776672363, "learning_rate": 1e-05, "loss": 0.4597, "mean_token_accuracy": 0.8798934817314148, "num_tokens": 503915473.0, "step": 13275 }, { "epoch": 1.6888436585676123, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.00022411346435546875, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.601839065551758, "learning_rate": 1e-05, "loss": 0.4973, "mean_token_accuracy": 0.8719989061355591, "num_tokens": 503954234.0, "step": 13276 }, { "epoch": 1.6889708688462028, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.00022411346435546875, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.770380973815918, "learning_rate": 1e-05, "loss": 0.5625, "mean_token_accuracy": 0.8471002578735352, "num_tokens": 503988740.0, "step": 13277 }, { "epoch": 1.6890980791247934, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.00022411346435546875, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.901269912719727, "learning_rate": 1e-05, "loss": 0.4989, "mean_token_accuracy": 0.8723670244216919, "num_tokens": 504025755.0, "step": 13278 }, { "epoch": 1.689225289403384, "ewc_loss": 0.10400390625, "ewc_loss_diag": 0.00022411346435546875, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.583343505859375, "learning_rate": 1e-05, "loss": 0.5286, "mean_token_accuracy": 0.8532915115356445, "num_tokens": 504062639.0, "step": 13279 }, { "epoch": 1.6893524996819744, "ewc_loss": 0.140625, "ewc_loss_diag": 0.00022411346435546875, "ewc_loss_parallel": 0.00011873245239257812, "grad_norm": 7.3150315284729, "learning_rate": 1e-05, "loss": 0.5806, "mean_token_accuracy": 0.8493810892105103, "num_tokens": 504099645.0, "step": 13280 }, { "epoch": 1.689479709960565, "ewc_loss": 0.109375, "ewc_loss_diag": 0.00022411346435546875, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.75610876083374, "learning_rate": 1e-05, "loss": 0.4547, "mean_token_accuracy": 0.8797773122787476, "num_tokens": 504137678.0, "step": 13281 }, { "epoch": 1.6896069202391555, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.00022411346435546875, "ewc_loss_parallel": 0.0001239776611328125, "grad_norm": 7.575613498687744, "learning_rate": 1e-05, "loss": 0.5287, "mean_token_accuracy": 0.8695511817932129, "num_tokens": 504174037.0, "step": 13282 }, { "epoch": 1.6897341305177458, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.00022411346435546875, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.224218845367432, "learning_rate": 1e-05, "loss": 0.5371, "mean_token_accuracy": 0.8535362482070923, "num_tokens": 504220117.0, "step": 13283 }, { "epoch": 1.6898613407963363, "ewc_loss": 0.140625, "ewc_loss_diag": 0.00022411346435546875, "ewc_loss_parallel": 0.00011873245239257812, "grad_norm": 7.381377696990967, "learning_rate": 1e-05, "loss": 0.5006, "mean_token_accuracy": 0.8765953779220581, "num_tokens": 504259104.0, "step": 13284 }, { "epoch": 1.6899885510749268, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.00022411346435546875, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.475013732910156, "learning_rate": 1e-05, "loss": 0.507, "mean_token_accuracy": 0.869982123374939, "num_tokens": 504302867.0, "step": 13285 }, { "epoch": 1.6901157613535174, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.00022411346435546875, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 6.99672794342041, "learning_rate": 1e-05, "loss": 0.4765, "mean_token_accuracy": 0.8798027038574219, "num_tokens": 504337555.0, "step": 13286 }, { "epoch": 1.690242971632108, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.00022411346435546875, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.449455261230469, "learning_rate": 1e-05, "loss": 0.4626, "mean_token_accuracy": 0.8801516890525818, "num_tokens": 504372334.0, "step": 13287 }, { "epoch": 1.6903701819106984, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.00022411346435546875, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.649090766906738, "learning_rate": 1e-05, "loss": 0.4722, "mean_token_accuracy": 0.8750426769256592, "num_tokens": 504412894.0, "step": 13288 }, { "epoch": 1.6904973921892887, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.00022411346435546875, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.3286285400390625, "learning_rate": 1e-05, "loss": 0.4879, "mean_token_accuracy": 0.8719571828842163, "num_tokens": 504452919.0, "step": 13289 }, { "epoch": 1.6906246024678793, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.00022411346435546875, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.448713779449463, "learning_rate": 1e-05, "loss": 0.4951, "mean_token_accuracy": 0.8730880618095398, "num_tokens": 504490694.0, "step": 13290 }, { "epoch": 1.6907518127464698, "ewc_loss": 0.11474609375, "ewc_loss_diag": 0.00022411346435546875, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.163945198059082, "learning_rate": 1e-05, "loss": 0.5001, "mean_token_accuracy": 0.8656419515609741, "num_tokens": 504526606.0, "step": 13291 }, { "epoch": 1.6908790230250603, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.00022411346435546875, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.387681007385254, "learning_rate": 1e-05, "loss": 0.4781, "mean_token_accuracy": 0.8749076128005981, "num_tokens": 504568885.0, "step": 13292 }, { "epoch": 1.6910062333036509, "ewc_loss": 0.10888671875, "ewc_loss_diag": 0.00022411346435546875, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 6.00071907043457, "learning_rate": 1e-05, "loss": 0.45, "mean_token_accuracy": 0.8834660053253174, "num_tokens": 504604947.0, "step": 13293 }, { "epoch": 1.6911334435822414, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.00022411346435546875, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.457962989807129, "learning_rate": 1e-05, "loss": 0.5177, "mean_token_accuracy": 0.862407386302948, "num_tokens": 504643937.0, "step": 13294 }, { "epoch": 1.691260653860832, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.00022411346435546875, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.822441577911377, "learning_rate": 1e-05, "loss": 0.5007, "mean_token_accuracy": 0.8686410188674927, "num_tokens": 504684298.0, "step": 13295 }, { "epoch": 1.6913878641394224, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.00022411346435546875, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.642650604248047, "learning_rate": 1e-05, "loss": 0.5553, "mean_token_accuracy": 0.8548892736434937, "num_tokens": 504723578.0, "step": 13296 }, { "epoch": 1.691515074418013, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.00022411346435546875, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.676953315734863, "learning_rate": 1e-05, "loss": 0.4991, "mean_token_accuracy": 0.8617073893547058, "num_tokens": 504768190.0, "step": 13297 }, { "epoch": 1.6916422846966035, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.00022411346435546875, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.9485249519348145, "learning_rate": 1e-05, "loss": 0.4723, "mean_token_accuracy": 0.879483163356781, "num_tokens": 504807656.0, "step": 13298 }, { "epoch": 1.691769494975194, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.00022411346435546875, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.64031982421875, "learning_rate": 1e-05, "loss": 0.4774, "mean_token_accuracy": 0.8682833313941956, "num_tokens": 504846631.0, "step": 13299 }, { "epoch": 1.6918967052537845, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.00022411346435546875, "ewc_loss_parallel": 0.0001163482666015625, "grad_norm": 7.299525260925293, "learning_rate": 1e-05, "loss": 0.5109, "mean_token_accuracy": 0.8723841905593872, "num_tokens": 504890451.0, "step": 13300 }, { "epoch": 1.692023915532375, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.00022411346435546875, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 5.876745223999023, "learning_rate": 1e-05, "loss": 0.4512, "mean_token_accuracy": 0.883156418800354, "num_tokens": 504931863.0, "step": 13301 }, { "epoch": 1.6921511258109656, "ewc_loss": 0.140625, "ewc_loss_diag": 0.00022411346435546875, "ewc_loss_parallel": 0.00011873245239257812, "grad_norm": 7.413415431976318, "learning_rate": 1e-05, "loss": 0.4926, "mean_token_accuracy": 0.876910924911499, "num_tokens": 504970247.0, "step": 13302 }, { "epoch": 1.6922783360895561, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.00022411346435546875, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.27712345123291, "learning_rate": 1e-05, "loss": 0.4747, "mean_token_accuracy": 0.8761653304100037, "num_tokens": 505009987.0, "step": 13303 }, { "epoch": 1.6924055463681467, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.00022411346435546875, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 7.091180324554443, "learning_rate": 1e-05, "loss": 0.4886, "mean_token_accuracy": 0.8750753998756409, "num_tokens": 505051933.0, "step": 13304 }, { "epoch": 1.6925327566467372, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.00022411346435546875, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.481805324554443, "learning_rate": 1e-05, "loss": 0.4918, "mean_token_accuracy": 0.8727548122406006, "num_tokens": 505086216.0, "step": 13305 }, { "epoch": 1.6926599669253277, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.00022411346435546875, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.55832052230835, "learning_rate": 1e-05, "loss": 0.4262, "mean_token_accuracy": 0.8911975622177124, "num_tokens": 505123040.0, "step": 13306 }, { "epoch": 1.692787177203918, "ewc_loss": 0.125, "ewc_loss_diag": 0.00022411346435546875, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.557643413543701, "learning_rate": 1e-05, "loss": 0.4991, "mean_token_accuracy": 0.8733985424041748, "num_tokens": 505158546.0, "step": 13307 }, { "epoch": 1.6929143874825086, "ewc_loss": 0.10888671875, "ewc_loss_diag": 0.00022411346435546875, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 6.054594993591309, "learning_rate": 1e-05, "loss": 0.4967, "mean_token_accuracy": 0.8709419965744019, "num_tokens": 505195433.0, "step": 13308 }, { "epoch": 1.693041597761099, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.00022411346435546875, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.717697620391846, "learning_rate": 1e-05, "loss": 0.4779, "mean_token_accuracy": 0.8745697736740112, "num_tokens": 505232411.0, "step": 13309 }, { "epoch": 1.6931688080396896, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.00022411346435546875, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.708491802215576, "learning_rate": 1e-05, "loss": 0.5321, "mean_token_accuracy": 0.855858325958252, "num_tokens": 505270590.0, "step": 13310 }, { "epoch": 1.6932960183182801, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.00022411346435546875, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 7.036881923675537, "learning_rate": 1e-05, "loss": 0.5043, "mean_token_accuracy": 0.8706302642822266, "num_tokens": 505307705.0, "step": 13311 }, { "epoch": 1.6934232285968707, "ewc_loss": 0.10400390625, "ewc_loss_diag": 0.00022411346435546875, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.6260247230529785, "learning_rate": 1e-05, "loss": 0.4756, "mean_token_accuracy": 0.8712559938430786, "num_tokens": 505342676.0, "step": 13312 }, { "epoch": 1.693550438875461, "ewc_loss": 0.140625, "ewc_loss_diag": 0.00022411346435546875, "ewc_loss_parallel": 0.00011873245239257812, "grad_norm": 7.417742729187012, "learning_rate": 1e-05, "loss": 0.5654, "mean_token_accuracy": 0.8581339120864868, "num_tokens": 505387058.0, "step": 13313 }, { "epoch": 1.6936776491540515, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.00022411346435546875, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 5.917372703552246, "learning_rate": 1e-05, "loss": 0.5179, "mean_token_accuracy": 0.8600093722343445, "num_tokens": 505429584.0, "step": 13314 }, { "epoch": 1.693804859432642, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.00022411346435546875, "ewc_loss_parallel": 0.00012111663818359375, "grad_norm": 7.518458366394043, "learning_rate": 1e-05, "loss": 0.4982, "mean_token_accuracy": 0.8800603747367859, "num_tokens": 505464619.0, "step": 13315 }, { "epoch": 1.6939320697112326, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.00022411346435546875, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.331867694854736, "learning_rate": 1e-05, "loss": 0.516, "mean_token_accuracy": 0.8649308681488037, "num_tokens": 505507390.0, "step": 13316 }, { "epoch": 1.694059279989823, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.00022411346435546875, "ewc_loss_parallel": 0.0001125335693359375, "grad_norm": 7.219127655029297, "learning_rate": 1e-05, "loss": 0.516, "mean_token_accuracy": 0.869317889213562, "num_tokens": 505543599.0, "step": 13317 }, { "epoch": 1.6941864902684136, "ewc_loss": 0.125, "ewc_loss_diag": 0.00022411346435546875, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.481379985809326, "learning_rate": 1e-05, "loss": 0.5147, "mean_token_accuracy": 0.8680877685546875, "num_tokens": 505576283.0, "step": 13318 }, { "epoch": 1.6943137005470041, "ewc_loss": 0.125, "ewc_loss_diag": 0.00022411346435546875, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.819808006286621, "learning_rate": 1e-05, "loss": 0.514, "mean_token_accuracy": 0.8645134568214417, "num_tokens": 505615244.0, "step": 13319 }, { "epoch": 1.6944409108255947, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.00022411346435546875, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.394736289978027, "learning_rate": 1e-05, "loss": 0.5273, "mean_token_accuracy": 0.8594110608100891, "num_tokens": 505653505.0, "step": 13320 }, { "epoch": 1.6945681211041852, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.00022411346435546875, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.538401126861572, "learning_rate": 1e-05, "loss": 0.4951, "mean_token_accuracy": 0.8708720207214355, "num_tokens": 505678666.0, "step": 13321 }, { "epoch": 1.6946953313827757, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.00022411346435546875, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.253495216369629, "learning_rate": 1e-05, "loss": 0.524, "mean_token_accuracy": 0.8606807589530945, "num_tokens": 505718191.0, "step": 13322 }, { "epoch": 1.6948225416613663, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.00022411346435546875, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.38653039932251, "learning_rate": 1e-05, "loss": 0.4739, "mean_token_accuracy": 0.8839583396911621, "num_tokens": 505759174.0, "step": 13323 }, { "epoch": 1.6949497519399568, "ewc_loss": 0.11181640625, "ewc_loss_diag": 0.00022411346435546875, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 6.091732501983643, "learning_rate": 1e-05, "loss": 0.5324, "mean_token_accuracy": 0.8547530174255371, "num_tokens": 505804483.0, "step": 13324 }, { "epoch": 1.6950769622185473, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.00022411346435546875, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.393347263336182, "learning_rate": 1e-05, "loss": 0.5213, "mean_token_accuracy": 0.8625887632369995, "num_tokens": 505839767.0, "step": 13325 }, { "epoch": 1.6952041724971378, "ewc_loss": 0.10791015625, "ewc_loss_diag": 0.00022411346435546875, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.9154534339904785, "learning_rate": 1e-05, "loss": 0.5506, "mean_token_accuracy": 0.8507012128829956, "num_tokens": 505880207.0, "step": 13326 }, { "epoch": 1.6953313827757284, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.00022411346435546875, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.52109432220459, "learning_rate": 1e-05, "loss": 0.4893, "mean_token_accuracy": 0.8735342621803284, "num_tokens": 505914686.0, "step": 13327 }, { "epoch": 1.695458593054319, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.00022411346435546875, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.729198455810547, "learning_rate": 1e-05, "loss": 0.4714, "mean_token_accuracy": 0.8721576929092407, "num_tokens": 505959029.0, "step": 13328 }, { "epoch": 1.6955858033329094, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.00022411346435546875, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.781850337982178, "learning_rate": 1e-05, "loss": 0.5321, "mean_token_accuracy": 0.8586971759796143, "num_tokens": 505997501.0, "step": 13329 }, { "epoch": 1.6957130136115, "ewc_loss": 0.10400390625, "ewc_loss_diag": 0.00022411346435546875, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.624627113342285, "learning_rate": 1e-05, "loss": 0.5115, "mean_token_accuracy": 0.8588496446609497, "num_tokens": 506029097.0, "step": 13330 }, { "epoch": 1.6958402238900905, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.00022411346435546875, "ewc_loss_parallel": 0.00011110305786132812, "grad_norm": 7.140214920043945, "learning_rate": 1e-05, "loss": 0.4751, "mean_token_accuracy": 0.8833434581756592, "num_tokens": 506065912.0, "step": 13331 }, { "epoch": 1.6959674341686808, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.00022411346435546875, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.7078375816345215, "learning_rate": 1e-05, "loss": 0.4814, "mean_token_accuracy": 0.8745051622390747, "num_tokens": 506108843.0, "step": 13332 }, { "epoch": 1.6960946444472713, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.00022411346435546875, "ewc_loss_parallel": 0.0001201629638671875, "grad_norm": 7.427862167358398, "learning_rate": 1e-05, "loss": 0.4792, "mean_token_accuracy": 0.882643461227417, "num_tokens": 506147505.0, "step": 13333 }, { "epoch": 1.6962218547258618, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.000225067138671875, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.105778217315674, "learning_rate": 1e-05, "loss": 0.484, "mean_token_accuracy": 0.8755544424057007, "num_tokens": 506186491.0, "step": 13334 }, { "epoch": 1.6963490650044524, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 0.00011682510375976562, "grad_norm": 7.303735733032227, "learning_rate": 1e-05, "loss": 0.4919, "mean_token_accuracy": 0.8763689398765564, "num_tokens": 506223343.0, "step": 13335 }, { "epoch": 1.696476275283043, "ewc_loss": 0.125, "ewc_loss_diag": 0.000225067138671875, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.405302047729492, "learning_rate": 1e-05, "loss": 0.501, "mean_token_accuracy": 0.8759639263153076, "num_tokens": 506262328.0, "step": 13336 }, { "epoch": 1.6966034855616334, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.869009971618652, "learning_rate": 1e-05, "loss": 0.5525, "mean_token_accuracy": 0.8584376573562622, "num_tokens": 506294758.0, "step": 13337 }, { "epoch": 1.6967306958402237, "ewc_loss": 0.125, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.495153427124023, "learning_rate": 1e-05, "loss": 0.4716, "mean_token_accuracy": 0.8788688778877258, "num_tokens": 506325423.0, "step": 13338 }, { "epoch": 1.6968579061188143, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.347813606262207, "learning_rate": 1e-05, "loss": 0.5338, "mean_token_accuracy": 0.8583146333694458, "num_tokens": 506367722.0, "step": 13339 }, { "epoch": 1.6969851163974048, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.576158046722412, "learning_rate": 1e-05, "loss": 0.5281, "mean_token_accuracy": 0.859385073184967, "num_tokens": 506404638.0, "step": 13340 }, { "epoch": 1.6971123266759953, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.917134761810303, "learning_rate": 1e-05, "loss": 0.4807, "mean_token_accuracy": 0.8684876561164856, "num_tokens": 506440976.0, "step": 13341 }, { "epoch": 1.6972395369545858, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.777079105377197, "learning_rate": 1e-05, "loss": 0.5134, "mean_token_accuracy": 0.8652949333190918, "num_tokens": 506475392.0, "step": 13342 }, { "epoch": 1.6973667472331764, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.659267425537109, "learning_rate": 1e-05, "loss": 0.4763, "mean_token_accuracy": 0.8702635765075684, "num_tokens": 506510933.0, "step": 13343 }, { "epoch": 1.697493957511767, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 0.00011014938354492188, "grad_norm": 7.095815181732178, "learning_rate": 1e-05, "loss": 0.5389, "mean_token_accuracy": 0.8632595539093018, "num_tokens": 506550217.0, "step": 13344 }, { "epoch": 1.6976211677903574, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.672529220581055, "learning_rate": 1e-05, "loss": 0.4875, "mean_token_accuracy": 0.8752601146697998, "num_tokens": 506583836.0, "step": 13345 }, { "epoch": 1.697748378068948, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 0.00011873245239257812, "grad_norm": 7.427774429321289, "learning_rate": 1e-05, "loss": 0.5045, "mean_token_accuracy": 0.8766456842422485, "num_tokens": 506616889.0, "step": 13346 }, { "epoch": 1.6978755883475385, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 5.989512920379639, "learning_rate": 1e-05, "loss": 0.4853, "mean_token_accuracy": 0.8759181499481201, "num_tokens": 506657198.0, "step": 13347 }, { "epoch": 1.698002798626129, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 0.00011920928955078125, "grad_norm": 7.441132068634033, "learning_rate": 1e-05, "loss": 0.494, "mean_token_accuracy": 0.8793364763259888, "num_tokens": 506696439.0, "step": 13348 }, { "epoch": 1.6981300089047195, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.349271297454834, "learning_rate": 1e-05, "loss": 0.5382, "mean_token_accuracy": 0.8578846454620361, "num_tokens": 506736140.0, "step": 13349 }, { "epoch": 1.69825721918331, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 0.00010967254638671875, "grad_norm": 7.142436981201172, "learning_rate": 1e-05, "loss": 0.4881, "mean_token_accuracy": 0.8781427145004272, "num_tokens": 506770976.0, "step": 13350 }, { "epoch": 1.6983844294619006, "ewc_loss": 0.125, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.43928861618042, "learning_rate": 1e-05, "loss": 0.4944, "mean_token_accuracy": 0.8711837530136108, "num_tokens": 506804463.0, "step": 13351 }, { "epoch": 1.6985116397404911, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.769355773925781, "learning_rate": 1e-05, "loss": 0.4526, "mean_token_accuracy": 0.8866791725158691, "num_tokens": 506839498.0, "step": 13352 }, { "epoch": 1.6986388500190817, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.363777160644531, "learning_rate": 1e-05, "loss": 0.5121, "mean_token_accuracy": 0.8648725748062134, "num_tokens": 506875899.0, "step": 13353 }, { "epoch": 1.6987660602976722, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.490859508514404, "learning_rate": 1e-05, "loss": 0.5151, "mean_token_accuracy": 0.8634819984436035, "num_tokens": 506914760.0, "step": 13354 }, { "epoch": 1.6988932705762627, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.245551109313965, "learning_rate": 1e-05, "loss": 0.4827, "mean_token_accuracy": 0.8732590675354004, "num_tokens": 506950022.0, "step": 13355 }, { "epoch": 1.699020480854853, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.3483171463012695, "learning_rate": 1e-05, "loss": 0.482, "mean_token_accuracy": 0.8742138147354126, "num_tokens": 506993349.0, "step": 13356 }, { "epoch": 1.6991476911334435, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.098868370056152, "learning_rate": 1e-05, "loss": 0.507, "mean_token_accuracy": 0.8641417026519775, "num_tokens": 507037519.0, "step": 13357 }, { "epoch": 1.699274901412034, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.354925632476807, "learning_rate": 1e-05, "loss": 0.5847, "mean_token_accuracy": 0.8427211046218872, "num_tokens": 507071932.0, "step": 13358 }, { "epoch": 1.6994021116906246, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.934048652648926, "learning_rate": 1e-05, "loss": 0.4733, "mean_token_accuracy": 0.8697608113288879, "num_tokens": 507108140.0, "step": 13359 }, { "epoch": 1.6995293219692151, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.4529805183410645, "learning_rate": 1e-05, "loss": 0.5118, "mean_token_accuracy": 0.8639103770256042, "num_tokens": 507148241.0, "step": 13360 }, { "epoch": 1.6996565322478057, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.772319316864014, "learning_rate": 1e-05, "loss": 0.4814, "mean_token_accuracy": 0.8702906370162964, "num_tokens": 507183200.0, "step": 13361 }, { "epoch": 1.699783742526396, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.655003547668457, "learning_rate": 1e-05, "loss": 0.4965, "mean_token_accuracy": 0.866455614566803, "num_tokens": 507219036.0, "step": 13362 }, { "epoch": 1.6999109528049865, "ewc_loss": 0.10400390625, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.620692253112793, "learning_rate": 1e-05, "loss": 0.4428, "mean_token_accuracy": 0.881289005279541, "num_tokens": 507254767.0, "step": 13363 }, { "epoch": 1.700038163083577, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 7.017159461975098, "learning_rate": 1e-05, "loss": 0.4712, "mean_token_accuracy": 0.8834406733512878, "num_tokens": 507289945.0, "step": 13364 }, { "epoch": 1.7001653733621676, "ewc_loss": 0.10400390625, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.662015914916992, "learning_rate": 1e-05, "loss": 0.483, "mean_token_accuracy": 0.8695348501205444, "num_tokens": 507328350.0, "step": 13365 }, { "epoch": 1.700292583640758, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 0.00011730194091796875, "grad_norm": 7.335697174072266, "learning_rate": 1e-05, "loss": 0.566, "mean_token_accuracy": 0.857067346572876, "num_tokens": 507367151.0, "step": 13366 }, { "epoch": 1.7004197939193486, "ewc_loss": 0.11376953125, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 5.8913893699646, "learning_rate": 1e-05, "loss": 0.4716, "mean_token_accuracy": 0.8757427930831909, "num_tokens": 507404181.0, "step": 13367 }, { "epoch": 1.7005470041979391, "ewc_loss": 0.140625, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 0.000118255615234375, "grad_norm": 7.433964252471924, "learning_rate": 1e-05, "loss": 0.4856, "mean_token_accuracy": 0.881014883518219, "num_tokens": 507438402.0, "step": 13368 }, { "epoch": 1.7006742144765297, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.290722370147705, "learning_rate": 1e-05, "loss": 0.4892, "mean_token_accuracy": 0.8717674016952515, "num_tokens": 507478091.0, "step": 13369 }, { "epoch": 1.7008014247551202, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 0.00011014938354492188, "grad_norm": 7.099520206451416, "learning_rate": 1e-05, "loss": 0.5326, "mean_token_accuracy": 0.8617730736732483, "num_tokens": 507518285.0, "step": 13370 }, { "epoch": 1.7009286350337107, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.469714641571045, "learning_rate": 1e-05, "loss": 0.5005, "mean_token_accuracy": 0.8687393069267273, "num_tokens": 507555250.0, "step": 13371 }, { "epoch": 1.7010558453123013, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.604365825653076, "learning_rate": 1e-05, "loss": 0.4884, "mean_token_accuracy": 0.8756016492843628, "num_tokens": 507591255.0, "step": 13372 }, { "epoch": 1.7011830555908918, "ewc_loss": 0.125, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.519654273986816, "learning_rate": 1e-05, "loss": 0.4819, "mean_token_accuracy": 0.8762350082397461, "num_tokens": 507623805.0, "step": 13373 }, { "epoch": 1.7013102658694823, "ewc_loss": 0.11181640625, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 6.1124067306518555, "learning_rate": 1e-05, "loss": 0.5234, "mean_token_accuracy": 0.8606662750244141, "num_tokens": 507660312.0, "step": 13374 }, { "epoch": 1.7014374761480728, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.642165660858154, "learning_rate": 1e-05, "loss": 0.4969, "mean_token_accuracy": 0.8730649948120117, "num_tokens": 507700132.0, "step": 13375 }, { "epoch": 1.7015646864266634, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.781445503234863, "learning_rate": 1e-05, "loss": 0.4791, "mean_token_accuracy": 0.8707690238952637, "num_tokens": 507737521.0, "step": 13376 }, { "epoch": 1.701691896705254, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.874115467071533, "learning_rate": 1e-05, "loss": 0.4467, "mean_token_accuracy": 0.8895577788352966, "num_tokens": 507779327.0, "step": 13377 }, { "epoch": 1.7018191069838444, "ewc_loss": 0.10400390625, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.648913383483887, "learning_rate": 1e-05, "loss": 0.4676, "mean_token_accuracy": 0.8723453879356384, "num_tokens": 507821899.0, "step": 13378 }, { "epoch": 1.701946317262435, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 0.00011348724365234375, "grad_norm": 7.192033290863037, "learning_rate": 1e-05, "loss": 0.467, "mean_token_accuracy": 0.8844309449195862, "num_tokens": 507862946.0, "step": 13379 }, { "epoch": 1.7020735275410255, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 5.782796859741211, "learning_rate": 1e-05, "loss": 0.5529, "mean_token_accuracy": 0.8493528962135315, "num_tokens": 507898409.0, "step": 13380 }, { "epoch": 1.7022007378196158, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 0.00011873245239257812, "grad_norm": 7.386214256286621, "learning_rate": 1e-05, "loss": 0.4965, "mean_token_accuracy": 0.8760637640953064, "num_tokens": 507936670.0, "step": 13381 }, { "epoch": 1.7023279480982063, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.089914321899414, "learning_rate": 1e-05, "loss": 0.4903, "mean_token_accuracy": 0.8688216805458069, "num_tokens": 507976160.0, "step": 13382 }, { "epoch": 1.7024551583767968, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 0.00011587142944335938, "grad_norm": 7.27994441986084, "learning_rate": 1e-05, "loss": 0.5111, "mean_token_accuracy": 0.8726537227630615, "num_tokens": 508015500.0, "step": 13383 }, { "epoch": 1.7025823686553874, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.313969135284424, "learning_rate": 1e-05, "loss": 0.4553, "mean_token_accuracy": 0.8845263719558716, "num_tokens": 508050860.0, "step": 13384 }, { "epoch": 1.702709578933978, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.980863094329834, "learning_rate": 1e-05, "loss": 0.5139, "mean_token_accuracy": 0.8681928515434265, "num_tokens": 508085794.0, "step": 13385 }, { "epoch": 1.7028367892125684, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.342623710632324, "learning_rate": 1e-05, "loss": 0.4537, "mean_token_accuracy": 0.8825236558914185, "num_tokens": 508122731.0, "step": 13386 }, { "epoch": 1.7029639994911587, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.666744709014893, "learning_rate": 1e-05, "loss": 0.5052, "mean_token_accuracy": 0.8674083352088928, "num_tokens": 508168513.0, "step": 13387 }, { "epoch": 1.7030912097697493, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.263253688812256, "learning_rate": 1e-05, "loss": 0.4414, "mean_token_accuracy": 0.8872793316841125, "num_tokens": 508209758.0, "step": 13388 }, { "epoch": 1.7032184200483398, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.451952934265137, "learning_rate": 1e-05, "loss": 0.4617, "mean_token_accuracy": 0.8824020624160767, "num_tokens": 508248172.0, "step": 13389 }, { "epoch": 1.7033456303269303, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.1624274253845215, "learning_rate": 1e-05, "loss": 0.5114, "mean_token_accuracy": 0.8636837005615234, "num_tokens": 508288694.0, "step": 13390 }, { "epoch": 1.7034728406055208, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.382170677185059, "learning_rate": 1e-05, "loss": 0.4991, "mean_token_accuracy": 0.8633787035942078, "num_tokens": 508329021.0, "step": 13391 }, { "epoch": 1.7036000508841114, "ewc_loss": 0.109375, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 6.0219292640686035, "learning_rate": 1e-05, "loss": 0.5302, "mean_token_accuracy": 0.8533346056938171, "num_tokens": 508363706.0, "step": 13392 }, { "epoch": 1.703727261162702, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.398205757141113, "learning_rate": 1e-05, "loss": 0.4491, "mean_token_accuracy": 0.8811665773391724, "num_tokens": 508399879.0, "step": 13393 }, { "epoch": 1.7038544714412924, "ewc_loss": 0.10595703125, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.87477970123291, "learning_rate": 1e-05, "loss": 0.5037, "mean_token_accuracy": 0.8630645275115967, "num_tokens": 508440172.0, "step": 13394 }, { "epoch": 1.703981681719883, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.526813983917236, "learning_rate": 1e-05, "loss": 0.5084, "mean_token_accuracy": 0.8671383261680603, "num_tokens": 508475531.0, "step": 13395 }, { "epoch": 1.7041088919984735, "ewc_loss": 0.10498046875, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.718900203704834, "learning_rate": 1e-05, "loss": 0.465, "mean_token_accuracy": 0.8782119154930115, "num_tokens": 508509290.0, "step": 13396 }, { "epoch": 1.704236102277064, "ewc_loss": 0.125, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.765816688537598, "learning_rate": 1e-05, "loss": 0.4841, "mean_token_accuracy": 0.8741255402565002, "num_tokens": 508545052.0, "step": 13397 }, { "epoch": 1.7043633125556545, "ewc_loss": 0.10400390625, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.6261115074157715, "learning_rate": 1e-05, "loss": 0.4658, "mean_token_accuracy": 0.8743768930435181, "num_tokens": 508577922.0, "step": 13398 }, { "epoch": 1.704490522834245, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 0.00010967254638671875, "grad_norm": 7.08571195602417, "learning_rate": 1e-05, "loss": 0.4965, "mean_token_accuracy": 0.8751283884048462, "num_tokens": 508618135.0, "step": 13399 }, { "epoch": 1.7046177331128356, "ewc_loss": 0.10595703125, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.7220778465271, "learning_rate": 1e-05, "loss": 0.5229, "mean_token_accuracy": 0.8572217226028442, "num_tokens": 508656665.0, "step": 13400 }, { "epoch": 1.7047449433914261, "ewc_loss": 0.140625, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 0.000118255615234375, "grad_norm": 7.332597255706787, "learning_rate": 1e-05, "loss": 0.5141, "mean_token_accuracy": 0.8708735704421997, "num_tokens": 508693898.0, "step": 13401 }, { "epoch": 1.7048721536700167, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.037497043609619, "learning_rate": 1e-05, "loss": 0.5337, "mean_token_accuracy": 0.8621236085891724, "num_tokens": 508726865.0, "step": 13402 }, { "epoch": 1.7049993639486072, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 0.00011539459228515625, "grad_norm": 7.245490074157715, "learning_rate": 1e-05, "loss": 0.5222, "mean_token_accuracy": 0.872276782989502, "num_tokens": 508762040.0, "step": 13403 }, { "epoch": 1.7051265742271977, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.315494060516357, "learning_rate": 1e-05, "loss": 0.5012, "mean_token_accuracy": 0.8687565326690674, "num_tokens": 508803288.0, "step": 13404 }, { "epoch": 1.705253784505788, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.851004123687744, "learning_rate": 1e-05, "loss": 0.4891, "mean_token_accuracy": 0.875670313835144, "num_tokens": 508840729.0, "step": 13405 }, { "epoch": 1.7053809947843785, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.454041957855225, "learning_rate": 1e-05, "loss": 0.4644, "mean_token_accuracy": 0.8797924518585205, "num_tokens": 508874219.0, "step": 13406 }, { "epoch": 1.705508205062969, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.327801704406738, "learning_rate": 1e-05, "loss": 0.4385, "mean_token_accuracy": 0.8890205025672913, "num_tokens": 508911746.0, "step": 13407 }, { "epoch": 1.7056354153415596, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.557570934295654, "learning_rate": 1e-05, "loss": 0.4769, "mean_token_accuracy": 0.8765238523483276, "num_tokens": 508944996.0, "step": 13408 }, { "epoch": 1.7057626256201501, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.879635334014893, "learning_rate": 1e-05, "loss": 0.503, "mean_token_accuracy": 0.8628649711608887, "num_tokens": 508986416.0, "step": 13409 }, { "epoch": 1.7058898358987407, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.78609037399292, "learning_rate": 1e-05, "loss": 0.5057, "mean_token_accuracy": 0.86771559715271, "num_tokens": 509023690.0, "step": 13410 }, { "epoch": 1.706017046177331, "ewc_loss": 0.10400390625, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.628032207489014, "learning_rate": 1e-05, "loss": 0.4825, "mean_token_accuracy": 0.8725336790084839, "num_tokens": 509065405.0, "step": 13411 }, { "epoch": 1.7061442564559215, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 0.00011301040649414062, "grad_norm": 7.144454002380371, "learning_rate": 1e-05, "loss": 0.5279, "mean_token_accuracy": 0.8668209314346313, "num_tokens": 509104871.0, "step": 13412 }, { "epoch": 1.706271466734512, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.684994697570801, "learning_rate": 1e-05, "loss": 0.5201, "mean_token_accuracy": 0.8534203171730042, "num_tokens": 509140938.0, "step": 13413 }, { "epoch": 1.7063986770131025, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 0.00012111663818359375, "grad_norm": 7.445076942443848, "learning_rate": 1e-05, "loss": 0.5162, "mean_token_accuracy": 0.8746427297592163, "num_tokens": 509175324.0, "step": 13414 }, { "epoch": 1.706525887291693, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.037965297698975, "learning_rate": 1e-05, "loss": 0.4862, "mean_token_accuracy": 0.8723598718643188, "num_tokens": 509214983.0, "step": 13415 }, { "epoch": 1.7066530975702836, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 0.00011873245239257812, "grad_norm": 7.412380218505859, "learning_rate": 1e-05, "loss": 0.5421, "mean_token_accuracy": 0.8631882667541504, "num_tokens": 509249418.0, "step": 13416 }, { "epoch": 1.7067803078488741, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.353359699249268, "learning_rate": 1e-05, "loss": 0.4767, "mean_token_accuracy": 0.8784645795822144, "num_tokens": 509282884.0, "step": 13417 }, { "epoch": 1.7069075181274647, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 0.00010967254638671875, "grad_norm": 7.098043918609619, "learning_rate": 1e-05, "loss": 0.4928, "mean_token_accuracy": 0.8744494915008545, "num_tokens": 509317880.0, "step": 13418 }, { "epoch": 1.7070347284060552, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.412538051605225, "learning_rate": 1e-05, "loss": 0.533, "mean_token_accuracy": 0.8624871969223022, "num_tokens": 509355760.0, "step": 13419 }, { "epoch": 1.7071619386846457, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.745967864990234, "learning_rate": 1e-05, "loss": 0.4926, "mean_token_accuracy": 0.8724334836006165, "num_tokens": 509393088.0, "step": 13420 }, { "epoch": 1.7072891489632362, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.320350170135498, "learning_rate": 1e-05, "loss": 0.52, "mean_token_accuracy": 0.8644135594367981, "num_tokens": 509431568.0, "step": 13421 }, { "epoch": 1.7074163592418268, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.506735801696777, "learning_rate": 1e-05, "loss": 0.5054, "mean_token_accuracy": 0.8673027753829956, "num_tokens": 509471044.0, "step": 13422 }, { "epoch": 1.7075435695204173, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.200286865234375, "learning_rate": 1e-05, "loss": 0.4618, "mean_token_accuracy": 0.8790013790130615, "num_tokens": 509507746.0, "step": 13423 }, { "epoch": 1.7076707797990078, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.398112773895264, "learning_rate": 1e-05, "loss": 0.4551, "mean_token_accuracy": 0.8831883072853088, "num_tokens": 509545753.0, "step": 13424 }, { "epoch": 1.7077979900775984, "ewc_loss": 0.11181640625, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 6.049869537353516, "learning_rate": 1e-05, "loss": 0.4751, "mean_token_accuracy": 0.8728680610656738, "num_tokens": 509581155.0, "step": 13425 }, { "epoch": 1.7079252003561889, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.0002269744873046875, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.415961265563965, "learning_rate": 1e-05, "loss": 0.51, "mean_token_accuracy": 0.8640644550323486, "num_tokens": 509615540.0, "step": 13426 }, { "epoch": 1.7080524106347794, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.0002269744873046875, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.897524356842041, "learning_rate": 1e-05, "loss": 0.4887, "mean_token_accuracy": 0.8674400448799133, "num_tokens": 509651257.0, "step": 13427 }, { "epoch": 1.70817962091337, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.0002269744873046875, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.5426459312438965, "learning_rate": 1e-05, "loss": 0.5486, "mean_token_accuracy": 0.8540084362030029, "num_tokens": 509688088.0, "step": 13428 }, { "epoch": 1.7083068311919605, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.738707542419434, "learning_rate": 1e-05, "loss": 0.4723, "mean_token_accuracy": 0.8715996146202087, "num_tokens": 509721122.0, "step": 13429 }, { "epoch": 1.7084340414705508, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002269744873046875, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.787070274353027, "learning_rate": 1e-05, "loss": 0.5638, "mean_token_accuracy": 0.8500027060508728, "num_tokens": 509756711.0, "step": 13430 }, { "epoch": 1.7085612517491413, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.0002269744873046875, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.6405744552612305, "learning_rate": 1e-05, "loss": 0.4846, "mean_token_accuracy": 0.8695546388626099, "num_tokens": 509793868.0, "step": 13431 }, { "epoch": 1.7086884620277318, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.0002269744873046875, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 7.125251293182373, "learning_rate": 1e-05, "loss": 0.4828, "mean_token_accuracy": 0.8802810311317444, "num_tokens": 509828741.0, "step": 13432 }, { "epoch": 1.7088156723063224, "ewc_loss": 0.10693359375, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.736775875091553, "learning_rate": 1e-05, "loss": 0.4613, "mean_token_accuracy": 0.8790140151977539, "num_tokens": 509866534.0, "step": 13433 }, { "epoch": 1.708942882584913, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.0002269744873046875, "ewc_loss_parallel": 0.00011730194091796875, "grad_norm": 7.377089023590088, "learning_rate": 1e-05, "loss": 0.5075, "mean_token_accuracy": 0.868628740310669, "num_tokens": 509897664.0, "step": 13434 }, { "epoch": 1.7090700928635034, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.0002269744873046875, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.088348865509033, "learning_rate": 1e-05, "loss": 0.5834, "mean_token_accuracy": 0.8455768823623657, "num_tokens": 509933349.0, "step": 13435 }, { "epoch": 1.7091973031420937, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002269744873046875, "ewc_loss_parallel": 0.00011444091796875, "grad_norm": 7.2532501220703125, "learning_rate": 1e-05, "loss": 0.5198, "mean_token_accuracy": 0.8671573400497437, "num_tokens": 509970912.0, "step": 13436 }, { "epoch": 1.7093245134206843, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.0002269744873046875, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.375009536743164, "learning_rate": 1e-05, "loss": 0.5329, "mean_token_accuracy": 0.8637559413909912, "num_tokens": 510007431.0, "step": 13437 }, { "epoch": 1.7094517236992748, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002269744873046875, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.81107234954834, "learning_rate": 1e-05, "loss": 0.5206, "mean_token_accuracy": 0.8663254380226135, "num_tokens": 510044109.0, "step": 13438 }, { "epoch": 1.7095789339778653, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0002269744873046875, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.476693630218506, "learning_rate": 1e-05, "loss": 0.546, "mean_token_accuracy": 0.8528763055801392, "num_tokens": 510084665.0, "step": 13439 }, { "epoch": 1.7097061442564558, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.0002269744873046875, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.293050289154053, "learning_rate": 1e-05, "loss": 0.5413, "mean_token_accuracy": 0.8558458685874939, "num_tokens": 510123108.0, "step": 13440 }, { "epoch": 1.7098333545350464, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.0002269744873046875, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.578585624694824, "learning_rate": 1e-05, "loss": 0.5242, "mean_token_accuracy": 0.864208459854126, "num_tokens": 510167918.0, "step": 13441 }, { "epoch": 1.709960564813637, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.0002269744873046875, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.864992141723633, "learning_rate": 1e-05, "loss": 0.4503, "mean_token_accuracy": 0.8789311647415161, "num_tokens": 510204555.0, "step": 13442 }, { "epoch": 1.7100877750922274, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.816525936126709, "learning_rate": 1e-05, "loss": 0.5643, "mean_token_accuracy": 0.8506022095680237, "num_tokens": 510239508.0, "step": 13443 }, { "epoch": 1.710214985370818, "ewc_loss": 0.10400390625, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.639256477355957, "learning_rate": 1e-05, "loss": 0.476, "mean_token_accuracy": 0.8725017309188843, "num_tokens": 510275246.0, "step": 13444 }, { "epoch": 1.7103421956494085, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 0.00011157989501953125, "grad_norm": 7.168324947357178, "learning_rate": 1e-05, "loss": 0.5092, "mean_token_accuracy": 0.8709319829940796, "num_tokens": 510316558.0, "step": 13445 }, { "epoch": 1.710469405927999, "ewc_loss": 0.10693359375, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.700868606567383, "learning_rate": 1e-05, "loss": 0.4686, "mean_token_accuracy": 0.8751856088638306, "num_tokens": 510351784.0, "step": 13446 }, { "epoch": 1.7105966162065895, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 0.00012063980102539062, "grad_norm": 7.469960689544678, "learning_rate": 1e-05, "loss": 0.5028, "mean_token_accuracy": 0.8766498565673828, "num_tokens": 510391044.0, "step": 13447 }, { "epoch": 1.71072382648518, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.0002269744873046875, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.058873653411865, "learning_rate": 1e-05, "loss": 0.5042, "mean_token_accuracy": 0.8719762563705444, "num_tokens": 510431199.0, "step": 13448 }, { "epoch": 1.7108510367637706, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 0.00011968612670898438, "grad_norm": 7.42941951751709, "learning_rate": 1e-05, "loss": 0.505, "mean_token_accuracy": 0.8742351531982422, "num_tokens": 510471629.0, "step": 13449 }, { "epoch": 1.7109782470423611, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.373786926269531, "learning_rate": 1e-05, "loss": 0.4569, "mean_token_accuracy": 0.8820568323135376, "num_tokens": 510511546.0, "step": 13450 }, { "epoch": 1.7111054573209517, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 0.00010967254638671875, "grad_norm": 7.114657878875732, "learning_rate": 1e-05, "loss": 0.5477, "mean_token_accuracy": 0.8617184162139893, "num_tokens": 510548942.0, "step": 13451 }, { "epoch": 1.7112326675995422, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.4387311935424805, "learning_rate": 1e-05, "loss": 0.5888, "mean_token_accuracy": 0.8435303568840027, "num_tokens": 510589855.0, "step": 13452 }, { "epoch": 1.7113598778781327, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.748985290527344, "learning_rate": 1e-05, "loss": 0.4856, "mean_token_accuracy": 0.8733199238777161, "num_tokens": 510628960.0, "step": 13453 }, { "epoch": 1.711487088156723, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.350710868835449, "learning_rate": 1e-05, "loss": 0.5171, "mean_token_accuracy": 0.8646843433380127, "num_tokens": 510671107.0, "step": 13454 }, { "epoch": 1.7116142984353135, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.502602577209473, "learning_rate": 1e-05, "loss": 0.5104, "mean_token_accuracy": 0.8666977286338806, "num_tokens": 510708957.0, "step": 13455 }, { "epoch": 1.711741508713904, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.0002269744873046875, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.209446907043457, "learning_rate": 1e-05, "loss": 0.4919, "mean_token_accuracy": 0.8709596395492554, "num_tokens": 510746150.0, "step": 13456 }, { "epoch": 1.7118687189924946, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0002269744873046875, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.408797740936279, "learning_rate": 1e-05, "loss": 0.5336, "mean_token_accuracy": 0.8602945804595947, "num_tokens": 510775810.0, "step": 13457 }, { "epoch": 1.7119959292710851, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.0002269744873046875, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 6.0554118156433105, "learning_rate": 1e-05, "loss": 0.5326, "mean_token_accuracy": 0.8573288917541504, "num_tokens": 510817438.0, "step": 13458 }, { "epoch": 1.7121231395496757, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0002269744873046875, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.419427871704102, "learning_rate": 1e-05, "loss": 0.5081, "mean_token_accuracy": 0.865906834602356, "num_tokens": 510854529.0, "step": 13459 }, { "epoch": 1.712250349828266, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.0002269744873046875, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.888675212860107, "learning_rate": 1e-05, "loss": 0.4774, "mean_token_accuracy": 0.868000864982605, "num_tokens": 510891568.0, "step": 13460 }, { "epoch": 1.7123775601068565, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.547476768493652, "learning_rate": 1e-05, "loss": 0.4835, "mean_token_accuracy": 0.8725016117095947, "num_tokens": 510928424.0, "step": 13461 }, { "epoch": 1.712504770385447, "ewc_loss": 0.10400390625, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.7373738288879395, "learning_rate": 1e-05, "loss": 0.5047, "mean_token_accuracy": 0.8642680048942566, "num_tokens": 510971770.0, "step": 13462 }, { "epoch": 1.7126319806640375, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.807275295257568, "learning_rate": 1e-05, "loss": 0.5227, "mean_token_accuracy": 0.8596718311309814, "num_tokens": 511004087.0, "step": 13463 }, { "epoch": 1.712759190942628, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.645212650299072, "learning_rate": 1e-05, "loss": 0.5156, "mean_token_accuracy": 0.8618647456169128, "num_tokens": 511046494.0, "step": 13464 }, { "epoch": 1.7128864012212186, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 7.143360614776611, "learning_rate": 1e-05, "loss": 0.539, "mean_token_accuracy": 0.8626936674118042, "num_tokens": 511088729.0, "step": 13465 }, { "epoch": 1.7130136114998091, "ewc_loss": 0.10791015625, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.767178535461426, "learning_rate": 1e-05, "loss": 0.4633, "mean_token_accuracy": 0.8732158541679382, "num_tokens": 511127581.0, "step": 13466 }, { "epoch": 1.7131408217783997, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 0.00011682510375976562, "grad_norm": 7.361996650695801, "learning_rate": 1e-05, "loss": 0.5325, "mean_token_accuracy": 0.8623172640800476, "num_tokens": 511166072.0, "step": 13467 }, { "epoch": 1.7132680320569902, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.115224838256836, "learning_rate": 1e-05, "loss": 0.5021, "mean_token_accuracy": 0.8688567876815796, "num_tokens": 511203529.0, "step": 13468 }, { "epoch": 1.7133952423355807, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 0.00011301040649414062, "grad_norm": 7.2114715576171875, "learning_rate": 1e-05, "loss": 0.5168, "mean_token_accuracy": 0.8674331903457642, "num_tokens": 511245846.0, "step": 13469 }, { "epoch": 1.7135224526141712, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.384572505950928, "learning_rate": 1e-05, "loss": 0.5189, "mean_token_accuracy": 0.8635937571525574, "num_tokens": 511287280.0, "step": 13470 }, { "epoch": 1.7136496628927618, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.749258041381836, "learning_rate": 1e-05, "loss": 0.4962, "mean_token_accuracy": 0.871049165725708, "num_tokens": 511330905.0, "step": 13471 }, { "epoch": 1.7137768731713523, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.490822792053223, "learning_rate": 1e-05, "loss": 0.5421, "mean_token_accuracy": 0.8580856919288635, "num_tokens": 511369872.0, "step": 13472 }, { "epoch": 1.7139040834499428, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.225369930267334, "learning_rate": 1e-05, "loss": 0.4736, "mean_token_accuracy": 0.8770434856414795, "num_tokens": 511410638.0, "step": 13473 }, { "epoch": 1.7140312937285334, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.614579200744629, "learning_rate": 1e-05, "loss": 0.4953, "mean_token_accuracy": 0.8713353872299194, "num_tokens": 511445517.0, "step": 13474 }, { "epoch": 1.7141585040071239, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.807843208312988, "learning_rate": 1e-05, "loss": 0.525, "mean_token_accuracy": 0.8548314571380615, "num_tokens": 511482749.0, "step": 13475 }, { "epoch": 1.7142857142857144, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.925304889678955, "learning_rate": 1e-05, "loss": 0.5407, "mean_token_accuracy": 0.8592211008071899, "num_tokens": 511517001.0, "step": 13476 }, { "epoch": 1.714412924564305, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.599447727203369, "learning_rate": 1e-05, "loss": 0.5087, "mean_token_accuracy": 0.8631059527397156, "num_tokens": 511559437.0, "step": 13477 }, { "epoch": 1.7145401348428955, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 0.00011587142944335938, "grad_norm": 7.321230888366699, "learning_rate": 1e-05, "loss": 0.5223, "mean_token_accuracy": 0.8698945045471191, "num_tokens": 511593713.0, "step": 13478 }, { "epoch": 1.7146673451214858, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.758519172668457, "learning_rate": 1e-05, "loss": 0.5341, "mean_token_accuracy": 0.8559139966964722, "num_tokens": 511633515.0, "step": 13479 }, { "epoch": 1.7147945554000763, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 0.0001220703125, "grad_norm": 7.580345630645752, "learning_rate": 1e-05, "loss": 0.5204, "mean_token_accuracy": 0.8675285577774048, "num_tokens": 511671539.0, "step": 13480 }, { "epoch": 1.7149217656786668, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.211220741271973, "learning_rate": 1e-05, "loss": 0.4725, "mean_token_accuracy": 0.8768612146377563, "num_tokens": 511706972.0, "step": 13481 }, { "epoch": 1.7150489759572574, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 0.00011920928955078125, "grad_norm": 7.415585041046143, "learning_rate": 1e-05, "loss": 0.5442, "mean_token_accuracy": 0.8636569976806641, "num_tokens": 511740186.0, "step": 13482 }, { "epoch": 1.7151761862358479, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.481747627258301, "learning_rate": 1e-05, "loss": 0.5012, "mean_token_accuracy": 0.8676760792732239, "num_tokens": 511779755.0, "step": 13483 }, { "epoch": 1.7153033965144384, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 7.030447483062744, "learning_rate": 1e-05, "loss": 0.5448, "mean_token_accuracy": 0.8606048822402954, "num_tokens": 511822531.0, "step": 13484 }, { "epoch": 1.7154306067930287, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.4698262214660645, "learning_rate": 1e-05, "loss": 0.5123, "mean_token_accuracy": 0.8665988445281982, "num_tokens": 511852950.0, "step": 13485 }, { "epoch": 1.7155578170716193, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.0002269744873046875, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.6759867668151855, "learning_rate": 1e-05, "loss": 0.4894, "mean_token_accuracy": 0.8748157024383545, "num_tokens": 511890251.0, "step": 13486 }, { "epoch": 1.7156850273502098, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.0002269744873046875, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.34092378616333, "learning_rate": 1e-05, "loss": 0.5526, "mean_token_accuracy": 0.8527436256408691, "num_tokens": 511933103.0, "step": 13487 }, { "epoch": 1.7158122376288003, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.0002269744873046875, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.470208168029785, "learning_rate": 1e-05, "loss": 0.4815, "mean_token_accuracy": 0.8757036924362183, "num_tokens": 511975682.0, "step": 13488 }, { "epoch": 1.7159394479073908, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.0002269744873046875, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.167563438415527, "learning_rate": 1e-05, "loss": 0.4952, "mean_token_accuracy": 0.8686633706092834, "num_tokens": 512014899.0, "step": 13489 }, { "epoch": 1.7160666581859814, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.0002269744873046875, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.409666538238525, "learning_rate": 1e-05, "loss": 0.5283, "mean_token_accuracy": 0.8571518659591675, "num_tokens": 512054951.0, "step": 13490 }, { "epoch": 1.716193868464572, "ewc_loss": 0.10888671875, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 6.0038628578186035, "learning_rate": 1e-05, "loss": 0.5008, "mean_token_accuracy": 0.867639422416687, "num_tokens": 512094273.0, "step": 13491 }, { "epoch": 1.7163210787431624, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0002269744873046875, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.48387336730957, "learning_rate": 1e-05, "loss": 0.5142, "mean_token_accuracy": 0.8643279075622559, "num_tokens": 512133493.0, "step": 13492 }, { "epoch": 1.716448289021753, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.0002269744873046875, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.831352710723877, "learning_rate": 1e-05, "loss": 0.484, "mean_token_accuracy": 0.8692958354949951, "num_tokens": 512171527.0, "step": 13493 }, { "epoch": 1.7165754993003435, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.0002269744873046875, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.655241966247559, "learning_rate": 1e-05, "loss": 0.4856, "mean_token_accuracy": 0.8732871413230896, "num_tokens": 512211106.0, "step": 13494 }, { "epoch": 1.716702709578934, "ewc_loss": 0.10400390625, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.682538986206055, "learning_rate": 1e-05, "loss": 0.456, "mean_token_accuracy": 0.8729344010353088, "num_tokens": 512247444.0, "step": 13495 }, { "epoch": 1.7168299198575245, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.976025581359863, "learning_rate": 1e-05, "loss": 0.5367, "mean_token_accuracy": 0.8606076836585999, "num_tokens": 512293912.0, "step": 13496 }, { "epoch": 1.716957130136115, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.662139415740967, "learning_rate": 1e-05, "loss": 0.452, "mean_token_accuracy": 0.8798494935035706, "num_tokens": 512334149.0, "step": 13497 }, { "epoch": 1.7170843404147056, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 0.0001163482666015625, "grad_norm": 7.317797660827637, "learning_rate": 1e-05, "loss": 0.4995, "mean_token_accuracy": 0.8739126920700073, "num_tokens": 512371161.0, "step": 13498 }, { "epoch": 1.7172115506932961, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 5.914523601531982, "learning_rate": 1e-05, "loss": 0.5315, "mean_token_accuracy": 0.8558719158172607, "num_tokens": 512407217.0, "step": 13499 }, { "epoch": 1.7173387609718866, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 0.00011730194091796875, "grad_norm": 7.3914875984191895, "learning_rate": 1e-05, "loss": 0.4839, "mean_token_accuracy": 0.8799071311950684, "num_tokens": 512441975.0, "step": 13500 }, { "epoch": 1.7174659712504772, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.2950520515441895, "learning_rate": 1e-05, "loss": 0.5258, "mean_token_accuracy": 0.8603885173797607, "num_tokens": 512482956.0, "step": 13501 }, { "epoch": 1.7175931815290677, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 7.056761741638184, "learning_rate": 1e-05, "loss": 0.5135, "mean_token_accuracy": 0.8671447038650513, "num_tokens": 512518514.0, "step": 13502 }, { "epoch": 1.717720391807658, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.478684425354004, "learning_rate": 1e-05, "loss": 0.5375, "mean_token_accuracy": 0.8597736954689026, "num_tokens": 512561620.0, "step": 13503 }, { "epoch": 1.7178476020862485, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.518190383911133, "learning_rate": 1e-05, "loss": 0.4767, "mean_token_accuracy": 0.8728012442588806, "num_tokens": 512598234.0, "step": 13504 }, { "epoch": 1.717974812364839, "ewc_loss": 0.125, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.5522541999816895, "learning_rate": 1e-05, "loss": 0.5095, "mean_token_accuracy": 0.8685349225997925, "num_tokens": 512638219.0, "step": 13505 }, { "epoch": 1.7181020226434296, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.0002269744873046875, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 6.024367332458496, "learning_rate": 1e-05, "loss": 0.5258, "mean_token_accuracy": 0.8582360148429871, "num_tokens": 512672277.0, "step": 13506 }, { "epoch": 1.7182292329220201, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.733843803405762, "learning_rate": 1e-05, "loss": 0.487, "mean_token_accuracy": 0.8719674348831177, "num_tokens": 512709573.0, "step": 13507 }, { "epoch": 1.7183564432006107, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.6942243576049805, "learning_rate": 1e-05, "loss": 0.5119, "mean_token_accuracy": 0.8616183996200562, "num_tokens": 512750243.0, "step": 13508 }, { "epoch": 1.718483653479201, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 7.071417331695557, "learning_rate": 1e-05, "loss": 0.5907, "mean_token_accuracy": 0.8456318378448486, "num_tokens": 512786001.0, "step": 13509 }, { "epoch": 1.7186108637577915, "ewc_loss": 0.10498046875, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.637385368347168, "learning_rate": 1e-05, "loss": 0.4733, "mean_token_accuracy": 0.8720251321792603, "num_tokens": 512823844.0, "step": 13510 }, { "epoch": 1.718738074036382, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 0.00012111663818359375, "grad_norm": 7.435555458068848, "learning_rate": 1e-05, "loss": 0.5205, "mean_token_accuracy": 0.8700927495956421, "num_tokens": 512859474.0, "step": 13511 }, { "epoch": 1.7188652843149725, "ewc_loss": 0.11474609375, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 5.9415059089660645, "learning_rate": 1e-05, "loss": 0.4528, "mean_token_accuracy": 0.8811644315719604, "num_tokens": 512889670.0, "step": 13512 }, { "epoch": 1.718992494593563, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 0.0001220703125, "grad_norm": 7.5200114250183105, "learning_rate": 1e-05, "loss": 0.4944, "mean_token_accuracy": 0.8790425062179565, "num_tokens": 512919749.0, "step": 13513 }, { "epoch": 1.7191197048721536, "ewc_loss": 0.125, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.355508327484131, "learning_rate": 1e-05, "loss": 0.5082, "mean_token_accuracy": 0.8666028380393982, "num_tokens": 512953206.0, "step": 13514 }, { "epoch": 1.7192469151507441, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 0.00011396408081054688, "grad_norm": 7.2183637619018555, "learning_rate": 1e-05, "loss": 0.5382, "mean_token_accuracy": 0.8641279339790344, "num_tokens": 512989066.0, "step": 13515 }, { "epoch": 1.7193741254293347, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.472649574279785, "learning_rate": 1e-05, "loss": 0.45, "mean_token_accuracy": 0.8872951865196228, "num_tokens": 513021975.0, "step": 13516 }, { "epoch": 1.7195013357079252, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0002269744873046875, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.827097415924072, "learning_rate": 1e-05, "loss": 0.5042, "mean_token_accuracy": 0.8697701692581177, "num_tokens": 513061669.0, "step": 13517 }, { "epoch": 1.7196285459865157, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.0002269744873046875, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.40078067779541, "learning_rate": 1e-05, "loss": 0.5059, "mean_token_accuracy": 0.8674839735031128, "num_tokens": 513095714.0, "step": 13518 }, { "epoch": 1.7197557562651062, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.0002269744873046875, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.529677867889404, "learning_rate": 1e-05, "loss": 0.5025, "mean_token_accuracy": 0.8685802221298218, "num_tokens": 513137235.0, "step": 13519 }, { "epoch": 1.7198829665436968, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0002269744873046875, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.260904788970947, "learning_rate": 1e-05, "loss": 0.5001, "mean_token_accuracy": 0.8711525201797485, "num_tokens": 513176339.0, "step": 13520 }, { "epoch": 1.7200101768222873, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0002269744873046875, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.395607948303223, "learning_rate": 1e-05, "loss": 0.4402, "mean_token_accuracy": 0.887633204460144, "num_tokens": 513218984.0, "step": 13521 }, { "epoch": 1.7201373871008778, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.0002269744873046875, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 6.09890604019165, "learning_rate": 1e-05, "loss": 0.5254, "mean_token_accuracy": 0.860833466053009, "num_tokens": 513259487.0, "step": 13522 }, { "epoch": 1.7202645973794684, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.0002269744873046875, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.419499397277832, "learning_rate": 1e-05, "loss": 0.5203, "mean_token_accuracy": 0.8625448346138, "num_tokens": 513296270.0, "step": 13523 }, { "epoch": 1.7203918076580589, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.0002269744873046875, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.914575099945068, "learning_rate": 1e-05, "loss": 0.5123, "mean_token_accuracy": 0.861682653427124, "num_tokens": 513333282.0, "step": 13524 }, { "epoch": 1.7205190179366494, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.0002269744873046875, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.539998531341553, "learning_rate": 1e-05, "loss": 0.4687, "mean_token_accuracy": 0.8727615475654602, "num_tokens": 513365679.0, "step": 13525 }, { "epoch": 1.72064622821524, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.0002269744873046875, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.753046989440918, "learning_rate": 1e-05, "loss": 0.4825, "mean_token_accuracy": 0.8703915476799011, "num_tokens": 513403387.0, "step": 13526 }, { "epoch": 1.7207734384938305, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002269744873046875, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.77286958694458, "learning_rate": 1e-05, "loss": 0.5833, "mean_token_accuracy": 0.8457246422767639, "num_tokens": 513442886.0, "step": 13527 }, { "epoch": 1.7209006487724208, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.0002269744873046875, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.634611129760742, "learning_rate": 1e-05, "loss": 0.451, "mean_token_accuracy": 0.8764901161193848, "num_tokens": 513482401.0, "step": 13528 }, { "epoch": 1.7210278590510113, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002269744873046875, "ewc_loss_parallel": 0.00011014938354492188, "grad_norm": 7.125226020812988, "learning_rate": 1e-05, "loss": 0.493, "mean_token_accuracy": 0.8690038919448853, "num_tokens": 513517898.0, "step": 13529 }, { "epoch": 1.7211550693296018, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.0002269744873046875, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.721421718597412, "learning_rate": 1e-05, "loss": 0.4817, "mean_token_accuracy": 0.8728593587875366, "num_tokens": 513556761.0, "step": 13530 }, { "epoch": 1.7212822796081924, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.0002269744873046875, "ewc_loss_parallel": 0.00012063980102539062, "grad_norm": 7.422706127166748, "learning_rate": 1e-05, "loss": 0.554, "mean_token_accuracy": 0.8592960238456726, "num_tokens": 513590803.0, "step": 13531 }, { "epoch": 1.7214094898867829, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0002269744873046875, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.072271823883057, "learning_rate": 1e-05, "loss": 0.4414, "mean_token_accuracy": 0.8907240629196167, "num_tokens": 513626718.0, "step": 13532 }, { "epoch": 1.7215367001653734, "ewc_loss": 0.140625, "ewc_loss_diag": 0.0002269744873046875, "ewc_loss_parallel": 0.00011777877807617188, "grad_norm": 7.338796138763428, "learning_rate": 1e-05, "loss": 0.5379, "mean_token_accuracy": 0.8682459592819214, "num_tokens": 513663439.0, "step": 13533 }, { "epoch": 1.7216639104439637, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002269744873046875, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.403017997741699, "learning_rate": 1e-05, "loss": 0.501, "mean_token_accuracy": 0.8709825277328491, "num_tokens": 513699916.0, "step": 13534 }, { "epoch": 1.7217911207225542, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002269744873046875, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.896871566772461, "learning_rate": 1e-05, "loss": 0.5025, "mean_token_accuracy": 0.8708210587501526, "num_tokens": 513739411.0, "step": 13535 }, { "epoch": 1.7219183310011448, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002269744873046875, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.515310764312744, "learning_rate": 1e-05, "loss": 0.504, "mean_token_accuracy": 0.8732180595397949, "num_tokens": 513777497.0, "step": 13536 }, { "epoch": 1.7220455412797353, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.0002269744873046875, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.387764930725098, "learning_rate": 1e-05, "loss": 0.5122, "mean_token_accuracy": 0.864163339138031, "num_tokens": 513817658.0, "step": 13537 }, { "epoch": 1.7221727515583258, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002269744873046875, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.589558124542236, "learning_rate": 1e-05, "loss": 0.4697, "mean_token_accuracy": 0.8815199136734009, "num_tokens": 513853011.0, "step": 13538 }, { "epoch": 1.7222999618369164, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.0002269744873046875, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.918603897094727, "learning_rate": 1e-05, "loss": 0.424, "mean_token_accuracy": 0.8933714628219604, "num_tokens": 513892933.0, "step": 13539 }, { "epoch": 1.7224271721155069, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002269744873046875, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.822803020477295, "learning_rate": 1e-05, "loss": 0.5514, "mean_token_accuracy": 0.8559935688972473, "num_tokens": 513932284.0, "step": 13540 }, { "epoch": 1.7225543823940974, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.0002269744873046875, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.655348777770996, "learning_rate": 1e-05, "loss": 0.4622, "mean_token_accuracy": 0.8770157098770142, "num_tokens": 513968151.0, "step": 13541 }, { "epoch": 1.722681592672688, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.0002269744873046875, "ewc_loss_parallel": 0.00011348724365234375, "grad_norm": 7.177090644836426, "learning_rate": 1e-05, "loss": 0.522, "mean_token_accuracy": 0.8683277368545532, "num_tokens": 514005323.0, "step": 13542 }, { "epoch": 1.7228088029512785, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.0002269744873046875, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.7084879875183105, "learning_rate": 1e-05, "loss": 0.5318, "mean_token_accuracy": 0.8553057909011841, "num_tokens": 514044158.0, "step": 13543 }, { "epoch": 1.722936013229869, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.0002269744873046875, "ewc_loss_parallel": 0.00012063980102539062, "grad_norm": 7.488193035125732, "learning_rate": 1e-05, "loss": 0.513, "mean_token_accuracy": 0.8734110593795776, "num_tokens": 514081076.0, "step": 13544 }, { "epoch": 1.7230632235084595, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.0002269744873046875, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.0632243156433105, "learning_rate": 1e-05, "loss": 0.539, "mean_token_accuracy": 0.8562848567962646, "num_tokens": 514122784.0, "step": 13545 }, { "epoch": 1.72319043378705, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.0002269744873046875, "ewc_loss_parallel": 0.00011920928955078125, "grad_norm": 7.4575300216674805, "learning_rate": 1e-05, "loss": 0.5274, "mean_token_accuracy": 0.8683114647865295, "num_tokens": 514158998.0, "step": 13546 }, { "epoch": 1.7233176440656406, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002269744873046875, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.412659168243408, "learning_rate": 1e-05, "loss": 0.5068, "mean_token_accuracy": 0.8650895357131958, "num_tokens": 514195815.0, "step": 13547 }, { "epoch": 1.7234448543442311, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002269744873046875, "ewc_loss_parallel": 0.00011014938354492188, "grad_norm": 7.102994918823242, "learning_rate": 1e-05, "loss": 0.5283, "mean_token_accuracy": 0.8644587397575378, "num_tokens": 514228353.0, "step": 13548 }, { "epoch": 1.7235720646228216, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002269744873046875, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.455099582672119, "learning_rate": 1e-05, "loss": 0.5211, "mean_token_accuracy": 0.864415168762207, "num_tokens": 514267283.0, "step": 13549 }, { "epoch": 1.7236992749014122, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.0002269744873046875, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.725525856018066, "learning_rate": 1e-05, "loss": 0.4888, "mean_token_accuracy": 0.8752761483192444, "num_tokens": 514309950.0, "step": 13550 }, { "epoch": 1.7238264851800027, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.0002269744873046875, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.354988098144531, "learning_rate": 1e-05, "loss": 0.4541, "mean_token_accuracy": 0.881481945514679, "num_tokens": 514345043.0, "step": 13551 }, { "epoch": 1.723953695458593, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.0002269744873046875, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.485673904418945, "learning_rate": 1e-05, "loss": 0.5067, "mean_token_accuracy": 0.8707160949707031, "num_tokens": 514383692.0, "step": 13552 }, { "epoch": 1.7240809057371835, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.0002269744873046875, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.207239627838135, "learning_rate": 1e-05, "loss": 0.5308, "mean_token_accuracy": 0.8564387559890747, "num_tokens": 514423552.0, "step": 13553 }, { "epoch": 1.724208116015774, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.0002269744873046875, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.397772789001465, "learning_rate": 1e-05, "loss": 0.5083, "mean_token_accuracy": 0.8673592805862427, "num_tokens": 514460460.0, "step": 13554 }, { "epoch": 1.7243353262943646, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.0002269744873046875, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 6.039146900177002, "learning_rate": 1e-05, "loss": 0.5031, "mean_token_accuracy": 0.863888144493103, "num_tokens": 514504751.0, "step": 13555 }, { "epoch": 1.7244625365729551, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0002269744873046875, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.441064834594727, "learning_rate": 1e-05, "loss": 0.5099, "mean_token_accuracy": 0.8649678230285645, "num_tokens": 514539967.0, "step": 13556 }, { "epoch": 1.7245897468515456, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.0002269744873046875, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.860566139221191, "learning_rate": 1e-05, "loss": 0.4757, "mean_token_accuracy": 0.8713563680648804, "num_tokens": 514583191.0, "step": 13557 }, { "epoch": 1.724716957130136, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.0002269744873046875, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.605967044830322, "learning_rate": 1e-05, "loss": 0.4927, "mean_token_accuracy": 0.8719395995140076, "num_tokens": 514622832.0, "step": 13558 }, { "epoch": 1.7248441674087265, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.0002269744873046875, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.700863838195801, "learning_rate": 1e-05, "loss": 0.4716, "mean_token_accuracy": 0.8743783831596375, "num_tokens": 514663387.0, "step": 13559 }, { "epoch": 1.724971377687317, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0002269744873046875, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.904758453369141, "learning_rate": 1e-05, "loss": 0.5224, "mean_token_accuracy": 0.863998293876648, "num_tokens": 514702711.0, "step": 13560 }, { "epoch": 1.7250985879659075, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.0002269744873046875, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.652368545532227, "learning_rate": 1e-05, "loss": 0.5302, "mean_token_accuracy": 0.8525936007499695, "num_tokens": 514740341.0, "step": 13561 }, { "epoch": 1.725225798244498, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00011491775512695312, "grad_norm": 7.244990348815918, "learning_rate": 1e-05, "loss": 0.5789, "mean_token_accuracy": 0.8543820977210999, "num_tokens": 514774834.0, "step": 13562 }, { "epoch": 1.7253530085230886, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.0002269744873046875, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 5.833629131317139, "learning_rate": 1e-05, "loss": 0.5328, "mean_token_accuracy": 0.8538721203804016, "num_tokens": 514814988.0, "step": 13563 }, { "epoch": 1.7254802188016791, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.0002269744873046875, "ewc_loss_parallel": 0.00011873245239257812, "grad_norm": 7.412018775939941, "learning_rate": 1e-05, "loss": 0.5148, "mean_token_accuracy": 0.8737297058105469, "num_tokens": 514854936.0, "step": 13564 }, { "epoch": 1.7256074290802697, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.0002269744873046875, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.222513675689697, "learning_rate": 1e-05, "loss": 0.5585, "mean_token_accuracy": 0.8510798215866089, "num_tokens": 514890528.0, "step": 13565 }, { "epoch": 1.7257346393588602, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002269744873046875, "ewc_loss_parallel": 0.00011157989501953125, "grad_norm": 7.143948078155518, "learning_rate": 1e-05, "loss": 0.5318, "mean_token_accuracy": 0.8654502630233765, "num_tokens": 514935455.0, "step": 13566 }, { "epoch": 1.7258618496374507, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.0002269744873046875, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.452581882476807, "learning_rate": 1e-05, "loss": 0.5247, "mean_token_accuracy": 0.8622099757194519, "num_tokens": 514975059.0, "step": 13567 }, { "epoch": 1.7259890599160412, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.0002269744873046875, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.659531593322754, "learning_rate": 1e-05, "loss": 0.477, "mean_token_accuracy": 0.8773689270019531, "num_tokens": 515007820.0, "step": 13568 }, { "epoch": 1.7261162701946318, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002269744873046875, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.522999286651611, "learning_rate": 1e-05, "loss": 0.5443, "mean_token_accuracy": 0.8600879311561584, "num_tokens": 515051826.0, "step": 13569 }, { "epoch": 1.7262434804732223, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.0002269744873046875, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.141633987426758, "learning_rate": 1e-05, "loss": 0.5107, "mean_token_accuracy": 0.8630947470664978, "num_tokens": 515092316.0, "step": 13570 }, { "epoch": 1.7263706907518128, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.0002269744873046875, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.670243263244629, "learning_rate": 1e-05, "loss": 0.4621, "mean_token_accuracy": 0.8825634717941284, "num_tokens": 515124314.0, "step": 13571 }, { "epoch": 1.7264979010304033, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.0002269744873046875, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.76041841506958, "learning_rate": 1e-05, "loss": 0.4809, "mean_token_accuracy": 0.8673030138015747, "num_tokens": 515159725.0, "step": 13572 }, { "epoch": 1.7266251113089939, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.9728922843933105, "learning_rate": 1e-05, "loss": 0.5467, "mean_token_accuracy": 0.8574464321136475, "num_tokens": 515201095.0, "step": 13573 }, { "epoch": 1.7267523215875844, "ewc_loss": 0.10400390625, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.611650466918945, "learning_rate": 1e-05, "loss": 0.4711, "mean_token_accuracy": 0.8720731735229492, "num_tokens": 515241761.0, "step": 13574 }, { "epoch": 1.726879531866175, "ewc_loss": 0.140625, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.000118255615234375, "grad_norm": 7.374457836151123, "learning_rate": 1e-05, "loss": 0.5392, "mean_token_accuracy": 0.8666982650756836, "num_tokens": 515281983.0, "step": 13575 }, { "epoch": 1.7270067421447655, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 5.854004383087158, "learning_rate": 1e-05, "loss": 0.5172, "mean_token_accuracy": 0.8634234666824341, "num_tokens": 515315557.0, "step": 13576 }, { "epoch": 1.7271339524233558, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.0001220703125, "grad_norm": 7.562810897827148, "learning_rate": 1e-05, "loss": 0.5491, "mean_token_accuracy": 0.8617118000984192, "num_tokens": 515345969.0, "step": 13577 }, { "epoch": 1.7272611627019463, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.271843433380127, "learning_rate": 1e-05, "loss": 0.5364, "mean_token_accuracy": 0.8609941005706787, "num_tokens": 515385520.0, "step": 13578 }, { "epoch": 1.7273883729805368, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00011682510375976562, "grad_norm": 7.330428600311279, "learning_rate": 1e-05, "loss": 0.5648, "mean_token_accuracy": 0.8579855561256409, "num_tokens": 515424296.0, "step": 13579 }, { "epoch": 1.7275155832591274, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.470395088195801, "learning_rate": 1e-05, "loss": 0.4772, "mean_token_accuracy": 0.8800919651985168, "num_tokens": 515461602.0, "step": 13580 }, { "epoch": 1.7276427935377179, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 6.945102214813232, "learning_rate": 1e-05, "loss": 0.5486, "mean_token_accuracy": 0.854570746421814, "num_tokens": 515504123.0, "step": 13581 }, { "epoch": 1.7277700038163084, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.433532238006592, "learning_rate": 1e-05, "loss": 0.4936, "mean_token_accuracy": 0.8710877299308777, "num_tokens": 515543056.0, "step": 13582 }, { "epoch": 1.7278972140948987, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.617660999298096, "learning_rate": 1e-05, "loss": 0.5075, "mean_token_accuracy": 0.8653345108032227, "num_tokens": 515580194.0, "step": 13583 }, { "epoch": 1.7280244243734892, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.321413993835449, "learning_rate": 1e-05, "loss": 0.4925, "mean_token_accuracy": 0.8736106157302856, "num_tokens": 515612493.0, "step": 13584 }, { "epoch": 1.7281516346520798, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.422878742218018, "learning_rate": 1e-05, "loss": 0.4606, "mean_token_accuracy": 0.8809714317321777, "num_tokens": 515656285.0, "step": 13585 }, { "epoch": 1.7282788449306703, "ewc_loss": 0.11474609375, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.164247989654541, "learning_rate": 1e-05, "loss": 0.4739, "mean_token_accuracy": 0.8747917413711548, "num_tokens": 515695326.0, "step": 13586 }, { "epoch": 1.7284060552092608, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.382904052734375, "learning_rate": 1e-05, "loss": 0.5481, "mean_token_accuracy": 0.8526102304458618, "num_tokens": 515730865.0, "step": 13587 }, { "epoch": 1.7285332654878514, "ewc_loss": 0.109375, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 6.022626876831055, "learning_rate": 1e-05, "loss": 0.5103, "mean_token_accuracy": 0.8620232343673706, "num_tokens": 515766725.0, "step": 13588 }, { "epoch": 1.7286604757664419, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.427371978759766, "learning_rate": 1e-05, "loss": 0.5272, "mean_token_accuracy": 0.8627710938453674, "num_tokens": 515807660.0, "step": 13589 }, { "epoch": 1.7287876860450324, "ewc_loss": 0.10693359375, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.864587783813477, "learning_rate": 1e-05, "loss": 0.453, "mean_token_accuracy": 0.8785463571548462, "num_tokens": 515844833.0, "step": 13590 }, { "epoch": 1.728914896323623, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.57904577255249, "learning_rate": 1e-05, "loss": 0.5048, "mean_token_accuracy": 0.8679314255714417, "num_tokens": 515876538.0, "step": 13591 }, { "epoch": 1.7290421066022135, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.69655179977417, "learning_rate": 1e-05, "loss": 0.4976, "mean_token_accuracy": 0.868230938911438, "num_tokens": 515916217.0, "step": 13592 }, { "epoch": 1.729169316880804, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.869564533233643, "learning_rate": 1e-05, "loss": 0.5715, "mean_token_accuracy": 0.8485667705535889, "num_tokens": 515955129.0, "step": 13593 }, { "epoch": 1.7292965271593945, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.623383045196533, "learning_rate": 1e-05, "loss": 0.503, "mean_token_accuracy": 0.8615747094154358, "num_tokens": 515993910.0, "step": 13594 }, { "epoch": 1.729423737437985, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00011444091796875, "grad_norm": 7.220507621765137, "learning_rate": 1e-05, "loss": 0.4796, "mean_token_accuracy": 0.8829578161239624, "num_tokens": 516034376.0, "step": 13595 }, { "epoch": 1.7295509477165756, "ewc_loss": 0.109375, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.7702789306640625, "learning_rate": 1e-05, "loss": 0.5159, "mean_token_accuracy": 0.8636831045150757, "num_tokens": 516070383.0, "step": 13596 }, { "epoch": 1.729678157995166, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.0001201629638671875, "grad_norm": 7.453101634979248, "learning_rate": 1e-05, "loss": 0.5318, "mean_token_accuracy": 0.8691381216049194, "num_tokens": 516114527.0, "step": 13597 }, { "epoch": 1.7298053682737566, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.1475419998168945, "learning_rate": 1e-05, "loss": 0.4768, "mean_token_accuracy": 0.8785126209259033, "num_tokens": 516156179.0, "step": 13598 }, { "epoch": 1.7299325785523472, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.0001163482666015625, "grad_norm": 7.289795875549316, "learning_rate": 1e-05, "loss": 0.5027, "mean_token_accuracy": 0.8734430074691772, "num_tokens": 516191000.0, "step": 13599 }, { "epoch": 1.7300597888309377, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.4601287841796875, "learning_rate": 1e-05, "loss": 0.4911, "mean_token_accuracy": 0.8733216524124146, "num_tokens": 516224194.0, "step": 13600 }, { "epoch": 1.730186999109528, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.792293548583984, "learning_rate": 1e-05, "loss": 0.4946, "mean_token_accuracy": 0.873734176158905, "num_tokens": 516262210.0, "step": 13601 }, { "epoch": 1.7303142093881185, "ewc_loss": 0.125, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.538835048675537, "learning_rate": 1e-05, "loss": 0.5275, "mean_token_accuracy": 0.863949716091156, "num_tokens": 516298859.0, "step": 13602 }, { "epoch": 1.730441419666709, "ewc_loss": 0.11474609375, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.265126705169678, "learning_rate": 1e-05, "loss": 0.5228, "mean_token_accuracy": 0.8624130487442017, "num_tokens": 516344100.0, "step": 13603 }, { "epoch": 1.7305686299452996, "ewc_loss": 0.125, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.606367111206055, "learning_rate": 1e-05, "loss": 0.5307, "mean_token_accuracy": 0.8613876700401306, "num_tokens": 516391313.0, "step": 13604 }, { "epoch": 1.7306958402238901, "ewc_loss": 0.10595703125, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.87545108795166, "learning_rate": 1e-05, "loss": 0.5311, "mean_token_accuracy": 0.8565671443939209, "num_tokens": 516430814.0, "step": 13605 }, { "epoch": 1.7308230505024806, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.814153671264648, "learning_rate": 1e-05, "loss": 0.5021, "mean_token_accuracy": 0.8698567748069763, "num_tokens": 516471380.0, "step": 13606 }, { "epoch": 1.730950260781071, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.664861679077148, "learning_rate": 1e-05, "loss": 0.4841, "mean_token_accuracy": 0.8698755502700806, "num_tokens": 516518374.0, "step": 13607 }, { "epoch": 1.7310774710596615, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00011157989501953125, "grad_norm": 7.148049831390381, "learning_rate": 1e-05, "loss": 0.5307, "mean_token_accuracy": 0.864307701587677, "num_tokens": 516549917.0, "step": 13608 }, { "epoch": 1.731204681338252, "ewc_loss": 0.10791015625, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.72312068939209, "learning_rate": 1e-05, "loss": 0.534, "mean_token_accuracy": 0.8552062511444092, "num_tokens": 516592869.0, "step": 13609 }, { "epoch": 1.7313318916168425, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00011968612670898438, "grad_norm": 7.420750141143799, "learning_rate": 1e-05, "loss": 0.5154, "mean_token_accuracy": 0.8719913363456726, "num_tokens": 516630550.0, "step": 13610 }, { "epoch": 1.731459101895433, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.045308589935303, "learning_rate": 1e-05, "loss": 0.4768, "mean_token_accuracy": 0.8767043352127075, "num_tokens": 516669285.0, "step": 13611 }, { "epoch": 1.7315863121740236, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00011730194091796875, "grad_norm": 7.388970851898193, "learning_rate": 1e-05, "loss": 0.4852, "mean_token_accuracy": 0.8827916383743286, "num_tokens": 516704616.0, "step": 13612 }, { "epoch": 1.7317135224526141, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.3305206298828125, "learning_rate": 1e-05, "loss": 0.4863, "mean_token_accuracy": 0.8760524392127991, "num_tokens": 516749566.0, "step": 13613 }, { "epoch": 1.7318407327312046, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00011014938354492188, "grad_norm": 7.097748756408691, "learning_rate": 1e-05, "loss": 0.5064, "mean_token_accuracy": 0.8734591007232666, "num_tokens": 516786264.0, "step": 13614 }, { "epoch": 1.7319679430097952, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.401528835296631, "learning_rate": 1e-05, "loss": 0.5217, "mean_token_accuracy": 0.8623390197753906, "num_tokens": 516822471.0, "step": 13615 }, { "epoch": 1.7320951532883857, "ewc_loss": 0.125, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.764213562011719, "learning_rate": 1e-05, "loss": 0.5164, "mean_token_accuracy": 0.8669474124908447, "num_tokens": 516861840.0, "step": 13616 }, { "epoch": 1.7322223635669762, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.335317611694336, "learning_rate": 1e-05, "loss": 0.5039, "mean_token_accuracy": 0.866314172744751, "num_tokens": 516895451.0, "step": 13617 }, { "epoch": 1.7323495738455668, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.499766826629639, "learning_rate": 1e-05, "loss": 0.4719, "mean_token_accuracy": 0.8758849501609802, "num_tokens": 516935926.0, "step": 13618 }, { "epoch": 1.7324767841241573, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.215797424316406, "learning_rate": 1e-05, "loss": 0.4978, "mean_token_accuracy": 0.8686373829841614, "num_tokens": 516976268.0, "step": 13619 }, { "epoch": 1.7326039944027478, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.383280277252197, "learning_rate": 1e-05, "loss": 0.5058, "mean_token_accuracy": 0.8671978712081909, "num_tokens": 517021459.0, "step": 13620 }, { "epoch": 1.7327312046813383, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 6.075247764587402, "learning_rate": 1e-05, "loss": 0.4729, "mean_token_accuracy": 0.8744703531265259, "num_tokens": 517052441.0, "step": 13621 }, { "epoch": 1.7328584149599289, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.392664909362793, "learning_rate": 1e-05, "loss": 0.5436, "mean_token_accuracy": 0.856795072555542, "num_tokens": 517094575.0, "step": 13622 }, { "epoch": 1.7329856252385194, "ewc_loss": 0.10693359375, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.920007705688477, "learning_rate": 1e-05, "loss": 0.4877, "mean_token_accuracy": 0.8669219017028809, "num_tokens": 517135239.0, "step": 13623 }, { "epoch": 1.73311283551711, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.504459381103516, "learning_rate": 1e-05, "loss": 0.5025, "mean_token_accuracy": 0.8672837018966675, "num_tokens": 517174132.0, "step": 13624 }, { "epoch": 1.7332400457957005, "ewc_loss": 0.10498046875, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.760278701782227, "learning_rate": 1e-05, "loss": 0.4761, "mean_token_accuracy": 0.8718160390853882, "num_tokens": 517208693.0, "step": 13625 }, { "epoch": 1.7333672560742908, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.723482608795166, "learning_rate": 1e-05, "loss": 0.5692, "mean_token_accuracy": 0.8501266241073608, "num_tokens": 517249441.0, "step": 13626 }, { "epoch": 1.7334944663528813, "ewc_loss": 0.10400390625, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.650727272033691, "learning_rate": 1e-05, "loss": 0.498, "mean_token_accuracy": 0.8644426465034485, "num_tokens": 517291368.0, "step": 13627 }, { "epoch": 1.7336216766314718, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 7.043554782867432, "learning_rate": 1e-05, "loss": 0.549, "mean_token_accuracy": 0.8592830896377563, "num_tokens": 517328292.0, "step": 13628 }, { "epoch": 1.7337488869100623, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.6939921379089355, "learning_rate": 1e-05, "loss": 0.4709, "mean_token_accuracy": 0.8755335807800293, "num_tokens": 517361049.0, "step": 13629 }, { "epoch": 1.7338760971886529, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00011730194091796875, "grad_norm": 7.328168869018555, "learning_rate": 1e-05, "loss": 0.5304, "mean_token_accuracy": 0.8685038089752197, "num_tokens": 517395086.0, "step": 13630 }, { "epoch": 1.7340033074672434, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 5.984260082244873, "learning_rate": 1e-05, "loss": 0.4944, "mean_token_accuracy": 0.8670297861099243, "num_tokens": 517432919.0, "step": 13631 }, { "epoch": 1.7341305177458337, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00011587142944335938, "grad_norm": 7.301273345947266, "learning_rate": 1e-05, "loss": 0.5324, "mean_token_accuracy": 0.8687629103660583, "num_tokens": 517474676.0, "step": 13632 }, { "epoch": 1.7342577280244242, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.309978485107422, "learning_rate": 1e-05, "loss": 0.4747, "mean_token_accuracy": 0.8764456510543823, "num_tokens": 517514131.0, "step": 13633 }, { "epoch": 1.7343849383030148, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.930768013000488, "learning_rate": 1e-05, "loss": 0.4977, "mean_token_accuracy": 0.8717072010040283, "num_tokens": 517547704.0, "step": 13634 }, { "epoch": 1.7345121485816053, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.450499534606934, "learning_rate": 1e-05, "loss": 0.5048, "mean_token_accuracy": 0.8672284483909607, "num_tokens": 517586430.0, "step": 13635 }, { "epoch": 1.7346393588601958, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.4174604415893555, "learning_rate": 1e-05, "loss": 0.5528, "mean_token_accuracy": 0.8545178174972534, "num_tokens": 517630303.0, "step": 13636 }, { "epoch": 1.7347665691387864, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.53286600112915, "learning_rate": 1e-05, "loss": 0.4387, "mean_token_accuracy": 0.8880826234817505, "num_tokens": 517663276.0, "step": 13637 }, { "epoch": 1.7348937794173769, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.96278190612793, "learning_rate": 1e-05, "loss": 0.4935, "mean_token_accuracy": 0.8648589849472046, "num_tokens": 517697946.0, "step": 13638 }, { "epoch": 1.7350209896959674, "ewc_loss": 0.125, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.74862003326416, "learning_rate": 1e-05, "loss": 0.5055, "mean_token_accuracy": 0.8667448163032532, "num_tokens": 517730820.0, "step": 13639 }, { "epoch": 1.735148199974558, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.666647434234619, "learning_rate": 1e-05, "loss": 0.4802, "mean_token_accuracy": 0.8736262321472168, "num_tokens": 517774576.0, "step": 13640 }, { "epoch": 1.7352754102531485, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00011014938354492188, "grad_norm": 7.09928035736084, "learning_rate": 1e-05, "loss": 0.5393, "mean_token_accuracy": 0.8590463399887085, "num_tokens": 517814748.0, "step": 13641 }, { "epoch": 1.735402620531739, "ewc_loss": 0.10595703125, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.647366046905518, "learning_rate": 1e-05, "loss": 0.4918, "mean_token_accuracy": 0.8682088851928711, "num_tokens": 517846582.0, "step": 13642 }, { "epoch": 1.7355298308103295, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00011968612670898438, "grad_norm": 7.467682838439941, "learning_rate": 1e-05, "loss": 0.5635, "mean_token_accuracy": 0.8545411229133606, "num_tokens": 517883354.0, "step": 13643 }, { "epoch": 1.73565704108892, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 5.984313011169434, "learning_rate": 1e-05, "loss": 0.545, "mean_token_accuracy": 0.8524325489997864, "num_tokens": 517920920.0, "step": 13644 }, { "epoch": 1.7357842513675106, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00012159347534179688, "grad_norm": 7.510325908660889, "learning_rate": 1e-05, "loss": 0.5557, "mean_token_accuracy": 0.857062816619873, "num_tokens": 517964844.0, "step": 13645 }, { "epoch": 1.735911461646101, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.3636603355407715, "learning_rate": 1e-05, "loss": 0.5149, "mean_token_accuracy": 0.8674166202545166, "num_tokens": 517998803.0, "step": 13646 }, { "epoch": 1.7360386719246916, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00011348724365234375, "grad_norm": 7.208522796630859, "learning_rate": 1e-05, "loss": 0.5183, "mean_token_accuracy": 0.8658161163330078, "num_tokens": 518040012.0, "step": 13647 }, { "epoch": 1.7361658822032822, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.457592010498047, "learning_rate": 1e-05, "loss": 0.5229, "mean_token_accuracy": 0.8600423336029053, "num_tokens": 518082287.0, "step": 13648 }, { "epoch": 1.7362930924818727, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.856233596801758, "learning_rate": 1e-05, "loss": 0.4901, "mean_token_accuracy": 0.8692643642425537, "num_tokens": 518116898.0, "step": 13649 }, { "epoch": 1.736420302760463, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.377365589141846, "learning_rate": 1e-05, "loss": 0.4827, "mean_token_accuracy": 0.8742408752441406, "num_tokens": 518157559.0, "step": 13650 }, { "epoch": 1.7365475130390535, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.5659894943237305, "learning_rate": 1e-05, "loss": 0.5285, "mean_token_accuracy": 0.8598614931106567, "num_tokens": 518200691.0, "step": 13651 }, { "epoch": 1.736674723317644, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.258224964141846, "learning_rate": 1e-05, "loss": 0.5539, "mean_token_accuracy": 0.8533066511154175, "num_tokens": 518239188.0, "step": 13652 }, { "epoch": 1.7368019335962346, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.435647010803223, "learning_rate": 1e-05, "loss": 0.5116, "mean_token_accuracy": 0.8603091835975647, "num_tokens": 518274294.0, "step": 13653 }, { "epoch": 1.736929143874825, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.095470905303955, "learning_rate": 1e-05, "loss": 0.4574, "mean_token_accuracy": 0.8812210559844971, "num_tokens": 518310960.0, "step": 13654 }, { "epoch": 1.7370563541534156, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.425247669219971, "learning_rate": 1e-05, "loss": 0.5784, "mean_token_accuracy": 0.8467093706130981, "num_tokens": 518352133.0, "step": 13655 }, { "epoch": 1.737183564432006, "ewc_loss": 0.10791015625, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.932990074157715, "learning_rate": 1e-05, "loss": 0.524, "mean_token_accuracy": 0.8585099577903748, "num_tokens": 518388793.0, "step": 13656 }, { "epoch": 1.7373107747105965, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.524775981903076, "learning_rate": 1e-05, "loss": 0.4786, "mean_token_accuracy": 0.8771388530731201, "num_tokens": 518425581.0, "step": 13657 }, { "epoch": 1.737437984989187, "ewc_loss": 0.10498046875, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.765781879425049, "learning_rate": 1e-05, "loss": 0.4914, "mean_token_accuracy": 0.8705782890319824, "num_tokens": 518465477.0, "step": 13658 }, { "epoch": 1.7375651952677775, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.749145984649658, "learning_rate": 1e-05, "loss": 0.5226, "mean_token_accuracy": 0.8608007431030273, "num_tokens": 518506790.0, "step": 13659 }, { "epoch": 1.737692405546368, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.66531229019165, "learning_rate": 1e-05, "loss": 0.4961, "mean_token_accuracy": 0.8676779270172119, "num_tokens": 518548240.0, "step": 13660 }, { "epoch": 1.7378196158249586, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00010967254638671875, "grad_norm": 7.083431243896484, "learning_rate": 1e-05, "loss": 0.5618, "mean_token_accuracy": 0.8556356430053711, "num_tokens": 518583865.0, "step": 13661 }, { "epoch": 1.7379468261035491, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.720232009887695, "learning_rate": 1e-05, "loss": 0.4577, "mean_token_accuracy": 0.8788745999336243, "num_tokens": 518623303.0, "step": 13662 }, { "epoch": 1.7380740363821396, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00011730194091796875, "grad_norm": 7.345888614654541, "learning_rate": 1e-05, "loss": 0.5399, "mean_token_accuracy": 0.8659017086029053, "num_tokens": 518662375.0, "step": 13663 }, { "epoch": 1.7382012466607302, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.0368804931640625, "learning_rate": 1e-05, "loss": 0.5271, "mean_token_accuracy": 0.8620932102203369, "num_tokens": 518702007.0, "step": 13664 }, { "epoch": 1.7383284569393207, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00011491775512695312, "grad_norm": 7.285419940948486, "learning_rate": 1e-05, "loss": 0.4884, "mean_token_accuracy": 0.8793310523033142, "num_tokens": 518743676.0, "step": 13665 }, { "epoch": 1.7384556672179112, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.351675033569336, "learning_rate": 1e-05, "loss": 0.4856, "mean_token_accuracy": 0.87431800365448, "num_tokens": 518783550.0, "step": 13666 }, { "epoch": 1.7385828774965018, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.864156723022461, "learning_rate": 1e-05, "loss": 0.457, "mean_token_accuracy": 0.8852618336677551, "num_tokens": 518820788.0, "step": 13667 }, { "epoch": 1.7387100877750923, "ewc_loss": 0.125, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.4890217781066895, "learning_rate": 1e-05, "loss": 0.5442, "mean_token_accuracy": 0.8587939739227295, "num_tokens": 518865195.0, "step": 13668 }, { "epoch": 1.7388372980536828, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.34050178527832, "learning_rate": 1e-05, "loss": 0.5101, "mean_token_accuracy": 0.865003228187561, "num_tokens": 518903683.0, "step": 13669 }, { "epoch": 1.7389645083322733, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.576159477233887, "learning_rate": 1e-05, "loss": 0.4669, "mean_token_accuracy": 0.8795992136001587, "num_tokens": 518939535.0, "step": 13670 }, { "epoch": 1.7390917186108639, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.8998494148254395, "learning_rate": 1e-05, "loss": 0.5421, "mean_token_accuracy": 0.854588508605957, "num_tokens": 518978753.0, "step": 13671 }, { "epoch": 1.7392189288894544, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.8151960372924805, "learning_rate": 1e-05, "loss": 0.5453, "mean_token_accuracy": 0.8552124500274658, "num_tokens": 519014916.0, "step": 13672 }, { "epoch": 1.739346139168045, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.645289421081543, "learning_rate": 1e-05, "loss": 0.5052, "mean_token_accuracy": 0.8640778064727783, "num_tokens": 519043537.0, "step": 13673 }, { "epoch": 1.7394733494466355, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00011110305786132812, "grad_norm": 7.1774749755859375, "learning_rate": 1e-05, "loss": 0.5015, "mean_token_accuracy": 0.8697433471679688, "num_tokens": 519081388.0, "step": 13674 }, { "epoch": 1.7396005597252258, "ewc_loss": 0.10693359375, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.676130294799805, "learning_rate": 1e-05, "loss": 0.4604, "mean_token_accuracy": 0.8780712485313416, "num_tokens": 519119531.0, "step": 13675 }, { "epoch": 1.7397277700038163, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00012111663818359375, "grad_norm": 7.512396812438965, "learning_rate": 1e-05, "loss": 0.5828, "mean_token_accuracy": 0.85288006067276, "num_tokens": 519157982.0, "step": 13676 }, { "epoch": 1.7398549802824068, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.0671820640563965, "learning_rate": 1e-05, "loss": 0.5182, "mean_token_accuracy": 0.8634228110313416, "num_tokens": 519198928.0, "step": 13677 }, { "epoch": 1.7399821905609973, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00012111663818359375, "grad_norm": 7.4832563400268555, "learning_rate": 1e-05, "loss": 0.5503, "mean_token_accuracy": 0.8615284562110901, "num_tokens": 519237760.0, "step": 13678 }, { "epoch": 1.7401094008395879, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.422446250915527, "learning_rate": 1e-05, "loss": 0.5048, "mean_token_accuracy": 0.8700603246688843, "num_tokens": 519273557.0, "step": 13679 }, { "epoch": 1.7402366111181784, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00011014938354492188, "grad_norm": 7.127122402191162, "learning_rate": 1e-05, "loss": 0.5255, "mean_token_accuracy": 0.8630614876747131, "num_tokens": 519316082.0, "step": 13680 }, { "epoch": 1.7403638213967687, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.473366737365723, "learning_rate": 1e-05, "loss": 0.4645, "mean_token_accuracy": 0.8825512528419495, "num_tokens": 519352528.0, "step": 13681 }, { "epoch": 1.7404910316753592, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.75842809677124, "learning_rate": 1e-05, "loss": 0.4445, "mean_token_accuracy": 0.8874403238296509, "num_tokens": 519384931.0, "step": 13682 }, { "epoch": 1.7406182419539498, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.369615077972412, "learning_rate": 1e-05, "loss": 0.5143, "mean_token_accuracy": 0.8693063259124756, "num_tokens": 519421974.0, "step": 13683 }, { "epoch": 1.7407454522325403, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.506009578704834, "learning_rate": 1e-05, "loss": 0.4819, "mean_token_accuracy": 0.8746763467788696, "num_tokens": 519460623.0, "step": 13684 }, { "epoch": 1.7408726625111308, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.232875347137451, "learning_rate": 1e-05, "loss": 0.4614, "mean_token_accuracy": 0.8812629580497742, "num_tokens": 519497079.0, "step": 13685 }, { "epoch": 1.7409998727897213, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.396484851837158, "learning_rate": 1e-05, "loss": 0.548, "mean_token_accuracy": 0.8522303104400635, "num_tokens": 519542680.0, "step": 13686 }, { "epoch": 1.7411270830683119, "ewc_loss": 0.11181640625, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.081276893615723, "learning_rate": 1e-05, "loss": 0.5076, "mean_token_accuracy": 0.8654993772506714, "num_tokens": 519577025.0, "step": 13687 }, { "epoch": 1.7412542933469024, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.4007463455200195, "learning_rate": 1e-05, "loss": 0.5132, "mean_token_accuracy": 0.8652024269104004, "num_tokens": 519618616.0, "step": 13688 }, { "epoch": 1.741381503625493, "ewc_loss": 0.10791015625, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.92567253112793, "learning_rate": 1e-05, "loss": 0.5018, "mean_token_accuracy": 0.8644534945487976, "num_tokens": 519658672.0, "step": 13689 }, { "epoch": 1.7415087139040835, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.5259294509887695, "learning_rate": 1e-05, "loss": 0.4351, "mean_token_accuracy": 0.8910014033317566, "num_tokens": 519692830.0, "step": 13690 }, { "epoch": 1.741635924182674, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.761986255645752, "learning_rate": 1e-05, "loss": 0.4804, "mean_token_accuracy": 0.8721172213554382, "num_tokens": 519729949.0, "step": 13691 }, { "epoch": 1.7417631344612645, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.764065742492676, "learning_rate": 1e-05, "loss": 0.5162, "mean_token_accuracy": 0.8657280206680298, "num_tokens": 519763556.0, "step": 13692 }, { "epoch": 1.741890344739855, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.648662090301514, "learning_rate": 1e-05, "loss": 0.5035, "mean_token_accuracy": 0.8605990409851074, "num_tokens": 519804556.0, "step": 13693 }, { "epoch": 1.7420175550184456, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00011014938354492188, "grad_norm": 7.101785659790039, "learning_rate": 1e-05, "loss": 0.4983, "mean_token_accuracy": 0.874079167842865, "num_tokens": 519845846.0, "step": 13694 }, { "epoch": 1.742144765297036, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.707101821899414, "learning_rate": 1e-05, "loss": 0.514, "mean_token_accuracy": 0.8611915707588196, "num_tokens": 519883333.0, "step": 13695 }, { "epoch": 1.7422719755756266, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00011920928955078125, "grad_norm": 7.415060043334961, "learning_rate": 1e-05, "loss": 0.48, "mean_token_accuracy": 0.8792012333869934, "num_tokens": 519915927.0, "step": 13696 }, { "epoch": 1.7423991858542172, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.0381760597229, "learning_rate": 1e-05, "loss": 0.4916, "mean_token_accuracy": 0.8698375225067139, "num_tokens": 519951555.0, "step": 13697 }, { "epoch": 1.7425263961328077, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00011730194091796875, "grad_norm": 7.359013557434082, "learning_rate": 1e-05, "loss": 0.4854, "mean_token_accuracy": 0.8796243071556091, "num_tokens": 519996616.0, "step": 13698 }, { "epoch": 1.742653606411398, "ewc_loss": 0.125, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.393868923187256, "learning_rate": 1e-05, "loss": 0.5522, "mean_token_accuracy": 0.8526438474655151, "num_tokens": 520032467.0, "step": 13699 }, { "epoch": 1.7427808166899885, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.947784423828125, "learning_rate": 1e-05, "loss": 0.5065, "mean_token_accuracy": 0.8718687295913696, "num_tokens": 520068893.0, "step": 13700 }, { "epoch": 1.742908026968579, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.510827541351318, "learning_rate": 1e-05, "loss": 0.5153, "mean_token_accuracy": 0.8691316843032837, "num_tokens": 520104745.0, "step": 13701 }, { "epoch": 1.7430352372471696, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.4174017906188965, "learning_rate": 1e-05, "loss": 0.5408, "mean_token_accuracy": 0.8553046584129333, "num_tokens": 520146994.0, "step": 13702 }, { "epoch": 1.74316244752576, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.570316791534424, "learning_rate": 1e-05, "loss": 0.4754, "mean_token_accuracy": 0.8791555166244507, "num_tokens": 520188220.0, "step": 13703 }, { "epoch": 1.7432896578043506, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.9701080322265625, "learning_rate": 1e-05, "loss": 0.4358, "mean_token_accuracy": 0.8855987787246704, "num_tokens": 520223410.0, "step": 13704 }, { "epoch": 1.743416868082941, "ewc_loss": 0.125, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.770583629608154, "learning_rate": 1e-05, "loss": 0.5121, "mean_token_accuracy": 0.8666056394577026, "num_tokens": 520260017.0, "step": 13705 }, { "epoch": 1.7435440783615315, "ewc_loss": 0.10498046875, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.70143985748291, "learning_rate": 1e-05, "loss": 0.4686, "mean_token_accuracy": 0.8756972551345825, "num_tokens": 520296232.0, "step": 13706 }, { "epoch": 1.743671288640122, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 7.09197998046875, "learning_rate": 1e-05, "loss": 0.5528, "mean_token_accuracy": 0.8572407960891724, "num_tokens": 520330720.0, "step": 13707 }, { "epoch": 1.7437984989187125, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.687564373016357, "learning_rate": 1e-05, "loss": 0.4711, "mean_token_accuracy": 0.876517653465271, "num_tokens": 520362438.0, "step": 13708 }, { "epoch": 1.743925709197303, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00011920928955078125, "grad_norm": 7.403631210327148, "learning_rate": 1e-05, "loss": 0.474, "mean_token_accuracy": 0.8842898011207581, "num_tokens": 520398170.0, "step": 13709 }, { "epoch": 1.7440529194758936, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 5.968172550201416, "learning_rate": 1e-05, "loss": 0.566, "mean_token_accuracy": 0.8518867492675781, "num_tokens": 520438219.0, "step": 13710 }, { "epoch": 1.744180129754484, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00011968612670898438, "grad_norm": 7.4442267417907715, "learning_rate": 1e-05, "loss": 0.5062, "mean_token_accuracy": 0.8732592463493347, "num_tokens": 520481642.0, "step": 13711 }, { "epoch": 1.7443073400330746, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.312041282653809, "learning_rate": 1e-05, "loss": 0.5007, "mean_token_accuracy": 0.8708882331848145, "num_tokens": 520515945.0, "step": 13712 }, { "epoch": 1.7444345503116652, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00011110305786132812, "grad_norm": 7.180282115936279, "learning_rate": 1e-05, "loss": 0.5451, "mean_token_accuracy": 0.8617801666259766, "num_tokens": 520553259.0, "step": 13713 }, { "epoch": 1.7445617605902557, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.415817737579346, "learning_rate": 1e-05, "loss": 0.4659, "mean_token_accuracy": 0.8801236152648926, "num_tokens": 520590371.0, "step": 13714 }, { "epoch": 1.7446889708688462, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.8399481773376465, "learning_rate": 1e-05, "loss": 0.5019, "mean_token_accuracy": 0.8693124651908875, "num_tokens": 520624738.0, "step": 13715 }, { "epoch": 1.7448161811474368, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.343608856201172, "learning_rate": 1e-05, "loss": 0.4553, "mean_token_accuracy": 0.8820746541023254, "num_tokens": 520666631.0, "step": 13716 }, { "epoch": 1.7449433914260273, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.592230796813965, "learning_rate": 1e-05, "loss": 0.4779, "mean_token_accuracy": 0.8751370906829834, "num_tokens": 520696567.0, "step": 13717 }, { "epoch": 1.7450706017046178, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.223501205444336, "learning_rate": 1e-05, "loss": 0.5064, "mean_token_accuracy": 0.8652052283287048, "num_tokens": 520737975.0, "step": 13718 }, { "epoch": 1.7451978119832083, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.450047016143799, "learning_rate": 1e-05, "loss": 0.4559, "mean_token_accuracy": 0.8819936513900757, "num_tokens": 520776884.0, "step": 13719 }, { "epoch": 1.7453250222617989, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.0912766456604, "learning_rate": 1e-05, "loss": 0.5416, "mean_token_accuracy": 0.856945812702179, "num_tokens": 520817770.0, "step": 13720 }, { "epoch": 1.7454522325403894, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.431597709655762, "learning_rate": 1e-05, "loss": 0.4945, "mean_token_accuracy": 0.8704549074172974, "num_tokens": 520859381.0, "step": 13721 }, { "epoch": 1.74557944281898, "ewc_loss": 0.10791015625, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.932830810546875, "learning_rate": 1e-05, "loss": 0.4215, "mean_token_accuracy": 0.8909178972244263, "num_tokens": 520900007.0, "step": 13722 }, { "epoch": 1.7457066530975704, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.527980804443359, "learning_rate": 1e-05, "loss": 0.4797, "mean_token_accuracy": 0.8770824074745178, "num_tokens": 520940707.0, "step": 13723 }, { "epoch": 1.7458338633761608, "ewc_loss": 0.10595703125, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.811104774475098, "learning_rate": 1e-05, "loss": 0.5152, "mean_token_accuracy": 0.8619783520698547, "num_tokens": 520972565.0, "step": 13724 }, { "epoch": 1.7459610736547513, "ewc_loss": 0.125, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.720279693603516, "learning_rate": 1e-05, "loss": 0.5123, "mean_token_accuracy": 0.864602267742157, "num_tokens": 521012364.0, "step": 13725 }, { "epoch": 1.7460882839333418, "ewc_loss": 0.10498046875, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.6889238357543945, "learning_rate": 1e-05, "loss": 0.4447, "mean_token_accuracy": 0.8815908432006836, "num_tokens": 521058761.0, "step": 13726 }, { "epoch": 1.7462154942119323, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 7.015073299407959, "learning_rate": 1e-05, "loss": 0.5165, "mean_token_accuracy": 0.8692485094070435, "num_tokens": 521096620.0, "step": 13727 }, { "epoch": 1.7463427044905229, "ewc_loss": 0.10693359375, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.726290225982666, "learning_rate": 1e-05, "loss": 0.5029, "mean_token_accuracy": 0.8657682538032532, "num_tokens": 521133223.0, "step": 13728 }, { "epoch": 1.7464699147691134, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00011539459228515625, "grad_norm": 7.265016555786133, "learning_rate": 1e-05, "loss": 0.4776, "mean_token_accuracy": 0.8830848932266235, "num_tokens": 521170334.0, "step": 13729 }, { "epoch": 1.7465971250477037, "ewc_loss": 0.11474609375, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 5.964840888977051, "learning_rate": 1e-05, "loss": 0.4732, "mean_token_accuracy": 0.8762794137001038, "num_tokens": 521207458.0, "step": 13730 }, { "epoch": 1.7467243353262942, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00011396408081054688, "grad_norm": 7.255805969238281, "learning_rate": 1e-05, "loss": 0.5121, "mean_token_accuracy": 0.8705583810806274, "num_tokens": 521245017.0, "step": 13731 }, { "epoch": 1.7468515456048848, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.274666786193848, "learning_rate": 1e-05, "loss": 0.5096, "mean_token_accuracy": 0.8630481958389282, "num_tokens": 521285792.0, "step": 13732 }, { "epoch": 1.7469787558834753, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.893476963043213, "learning_rate": 1e-05, "loss": 0.4754, "mean_token_accuracy": 0.8787155151367188, "num_tokens": 521327107.0, "step": 13733 }, { "epoch": 1.7471059661620658, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.422394275665283, "learning_rate": 1e-05, "loss": 0.4783, "mean_token_accuracy": 0.8799626231193542, "num_tokens": 521364720.0, "step": 13734 }, { "epoch": 1.7472331764406563, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.401628017425537, "learning_rate": 1e-05, "loss": 0.4687, "mean_token_accuracy": 0.8779339790344238, "num_tokens": 521398183.0, "step": 13735 }, { "epoch": 1.7473603867192469, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.529000282287598, "learning_rate": 1e-05, "loss": 0.5042, "mean_token_accuracy": 0.8687973022460938, "num_tokens": 521436837.0, "step": 13736 }, { "epoch": 1.7474875969978374, "ewc_loss": 0.10595703125, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.908691883087158, "learning_rate": 1e-05, "loss": 0.492, "mean_token_accuracy": 0.8696359395980835, "num_tokens": 521477722.0, "step": 13737 }, { "epoch": 1.747614807276428, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.794158935546875, "learning_rate": 1e-05, "loss": 0.4924, "mean_token_accuracy": 0.8749569654464722, "num_tokens": 521513647.0, "step": 13738 }, { "epoch": 1.7477420175550185, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.626273155212402, "learning_rate": 1e-05, "loss": 0.4933, "mean_token_accuracy": 0.8717063069343567, "num_tokens": 521545480.0, "step": 13739 }, { "epoch": 1.747869227833609, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00011205673217773438, "grad_norm": 7.203779697418213, "learning_rate": 1e-05, "loss": 0.5386, "mean_token_accuracy": 0.8590284585952759, "num_tokens": 521587363.0, "step": 13740 }, { "epoch": 1.7479964381121995, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.643543243408203, "learning_rate": 1e-05, "loss": 0.5061, "mean_token_accuracy": 0.8633062243461609, "num_tokens": 521629399.0, "step": 13741 }, { "epoch": 1.74812364839079, "ewc_loss": 0.1455078125, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.592708587646484, "learning_rate": 1e-05, "loss": 0.5319, "mean_token_accuracy": 0.8681334853172302, "num_tokens": 521667065.0, "step": 13742 }, { "epoch": 1.7482508586693806, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.0917463302612305, "learning_rate": 1e-05, "loss": 0.4702, "mean_token_accuracy": 0.8759123682975769, "num_tokens": 521696317.0, "step": 13743 }, { "epoch": 1.748378068947971, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.0001220703125, "grad_norm": 7.556195259094238, "learning_rate": 1e-05, "loss": 0.506, "mean_token_accuracy": 0.8748741745948792, "num_tokens": 521736948.0, "step": 13744 }, { "epoch": 1.7485052792265616, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.492193222045898, "learning_rate": 1e-05, "loss": 0.4666, "mean_token_accuracy": 0.8826526999473572, "num_tokens": 521770320.0, "step": 13745 }, { "epoch": 1.7486324895051522, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00011110305786132812, "grad_norm": 7.1726393699646, "learning_rate": 1e-05, "loss": 0.4685, "mean_token_accuracy": 0.8829699754714966, "num_tokens": 521803836.0, "step": 13746 }, { "epoch": 1.7487596997837427, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.54421854019165, "learning_rate": 1e-05, "loss": 0.5495, "mean_token_accuracy": 0.8563586473464966, "num_tokens": 521838066.0, "step": 13747 }, { "epoch": 1.748886910062333, "ewc_loss": 0.125, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.768698215484619, "learning_rate": 1e-05, "loss": 0.4815, "mean_token_accuracy": 0.8732212781906128, "num_tokens": 521876806.0, "step": 13748 }, { "epoch": 1.7490141203409235, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.413666248321533, "learning_rate": 1e-05, "loss": 0.524, "mean_token_accuracy": 0.863510012626648, "num_tokens": 521916241.0, "step": 13749 }, { "epoch": 1.749141330619514, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.508868217468262, "learning_rate": 1e-05, "loss": 0.4694, "mean_token_accuracy": 0.8751852512359619, "num_tokens": 521949369.0, "step": 13750 }, { "epoch": 1.7492685408981046, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.25832986831665, "learning_rate": 1e-05, "loss": 0.5198, "mean_token_accuracy": 0.8592498302459717, "num_tokens": 521989251.0, "step": 13751 }, { "epoch": 1.749395751176695, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.40815544128418, "learning_rate": 1e-05, "loss": 0.4705, "mean_token_accuracy": 0.8744455575942993, "num_tokens": 522019221.0, "step": 13752 }, { "epoch": 1.7495229614552856, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 6.087094306945801, "learning_rate": 1e-05, "loss": 0.5038, "mean_token_accuracy": 0.867081880569458, "num_tokens": 522058858.0, "step": 13753 }, { "epoch": 1.749650171733876, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.445441722869873, "learning_rate": 1e-05, "loss": 0.4901, "mean_token_accuracy": 0.8686587810516357, "num_tokens": 522089834.0, "step": 13754 }, { "epoch": 1.7497773820124665, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.899489402770996, "learning_rate": 1e-05, "loss": 0.4536, "mean_token_accuracy": 0.8778054118156433, "num_tokens": 522124474.0, "step": 13755 }, { "epoch": 1.749904592291057, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.5880208015441895, "learning_rate": 1e-05, "loss": 0.5235, "mean_token_accuracy": 0.8618365526199341, "num_tokens": 522159046.0, "step": 13756 }, { "epoch": 1.7500318025696475, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.726150989532471, "learning_rate": 1e-05, "loss": 0.4959, "mean_token_accuracy": 0.8680321574211121, "num_tokens": 522200698.0, "step": 13757 }, { "epoch": 1.750159012848238, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.892647743225098, "learning_rate": 1e-05, "loss": 0.5434, "mean_token_accuracy": 0.8558701276779175, "num_tokens": 522233533.0, "step": 13758 }, { "epoch": 1.7502862231268286, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.662961959838867, "learning_rate": 1e-05, "loss": 0.5326, "mean_token_accuracy": 0.8597370386123657, "num_tokens": 522274234.0, "step": 13759 }, { "epoch": 1.750413433405419, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00011444091796875, "grad_norm": 7.247501850128174, "learning_rate": 1e-05, "loss": 0.5146, "mean_token_accuracy": 0.871148943901062, "num_tokens": 522313055.0, "step": 13760 }, { "epoch": 1.7505406436840096, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 5.793095588684082, "learning_rate": 1e-05, "loss": 0.4774, "mean_token_accuracy": 0.8713943958282471, "num_tokens": 522346372.0, "step": 13761 }, { "epoch": 1.7506678539626002, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00012111663818359375, "grad_norm": 7.488193511962891, "learning_rate": 1e-05, "loss": 0.5259, "mean_token_accuracy": 0.8711205720901489, "num_tokens": 522381199.0, "step": 13762 }, { "epoch": 1.7507950642411907, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.194644927978516, "learning_rate": 1e-05, "loss": 0.5607, "mean_token_accuracy": 0.8528311252593994, "num_tokens": 522419509.0, "step": 13763 }, { "epoch": 1.7509222745197812, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00011587142944335938, "grad_norm": 7.296319961547852, "learning_rate": 1e-05, "loss": 0.5107, "mean_token_accuracy": 0.8698348999023438, "num_tokens": 522458433.0, "step": 13764 }, { "epoch": 1.7510494847983717, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.490542411804199, "learning_rate": 1e-05, "loss": 0.5081, "mean_token_accuracy": 0.870303213596344, "num_tokens": 522492010.0, "step": 13765 }, { "epoch": 1.7511766950769623, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.809974670410156, "learning_rate": 1e-05, "loss": 0.4957, "mean_token_accuracy": 0.8731618523597717, "num_tokens": 522524190.0, "step": 13766 }, { "epoch": 1.7513039053555528, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.563936233520508, "learning_rate": 1e-05, "loss": 0.5337, "mean_token_accuracy": 0.8567177057266235, "num_tokens": 522558218.0, "step": 13767 }, { "epoch": 1.7514311156341433, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.287198066711426, "learning_rate": 1e-05, "loss": 0.4889, "mean_token_accuracy": 0.8715765476226807, "num_tokens": 522594683.0, "step": 13768 }, { "epoch": 1.7515583259127339, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.627089500427246, "learning_rate": 1e-05, "loss": 0.4895, "mean_token_accuracy": 0.8746005892753601, "num_tokens": 522633409.0, "step": 13769 }, { "epoch": 1.7516855361913244, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.894896984100342, "learning_rate": 1e-05, "loss": 0.4923, "mean_token_accuracy": 0.866334855556488, "num_tokens": 522672469.0, "step": 13770 }, { "epoch": 1.751812746469915, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.833802700042725, "learning_rate": 1e-05, "loss": 0.4917, "mean_token_accuracy": 0.8735024929046631, "num_tokens": 522714676.0, "step": 13771 }, { "epoch": 1.7519399567485054, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.6780853271484375, "learning_rate": 1e-05, "loss": 0.4428, "mean_token_accuracy": 0.8821566104888916, "num_tokens": 522748726.0, "step": 13772 }, { "epoch": 1.7520671670270958, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.0001125335693359375, "grad_norm": 7.1780476570129395, "learning_rate": 1e-05, "loss": 0.5907, "mean_token_accuracy": 0.8500710725784302, "num_tokens": 522783761.0, "step": 13773 }, { "epoch": 1.7521943773056863, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.726166248321533, "learning_rate": 1e-05, "loss": 0.4714, "mean_token_accuracy": 0.8781407475471497, "num_tokens": 522822027.0, "step": 13774 }, { "epoch": 1.7523215875842768, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00011968612670898438, "grad_norm": 7.461568832397461, "learning_rate": 1e-05, "loss": 0.5024, "mean_token_accuracy": 0.8746178150177002, "num_tokens": 522856952.0, "step": 13775 }, { "epoch": 1.7524487978628673, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.069474220275879, "learning_rate": 1e-05, "loss": 0.5165, "mean_token_accuracy": 0.863312840461731, "num_tokens": 522897450.0, "step": 13776 }, { "epoch": 1.7525760081414579, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00011873245239257812, "grad_norm": 7.433302879333496, "learning_rate": 1e-05, "loss": 0.5056, "mean_token_accuracy": 0.8733558654785156, "num_tokens": 522930547.0, "step": 13777 }, { "epoch": 1.7527032184200484, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.375857353210449, "learning_rate": 1e-05, "loss": 0.5125, "mean_token_accuracy": 0.8671483993530273, "num_tokens": 522970145.0, "step": 13778 }, { "epoch": 1.7528304286986387, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00010967254638671875, "grad_norm": 7.1183905601501465, "learning_rate": 1e-05, "loss": 0.4641, "mean_token_accuracy": 0.8847155570983887, "num_tokens": 523006492.0, "step": 13779 }, { "epoch": 1.7529576389772292, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.432836055755615, "learning_rate": 1e-05, "loss": 0.5391, "mean_token_accuracy": 0.860553503036499, "num_tokens": 523045524.0, "step": 13780 }, { "epoch": 1.7530848492558198, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.771727561950684, "learning_rate": 1e-05, "loss": 0.4956, "mean_token_accuracy": 0.8723974823951721, "num_tokens": 523086091.0, "step": 13781 }, { "epoch": 1.7532120595344103, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.346556186676025, "learning_rate": 1e-05, "loss": 0.4839, "mean_token_accuracy": 0.8758354187011719, "num_tokens": 523127351.0, "step": 13782 }, { "epoch": 1.7533392698130008, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.5233259201049805, "learning_rate": 1e-05, "loss": 0.4981, "mean_token_accuracy": 0.8673179149627686, "num_tokens": 523161775.0, "step": 13783 }, { "epoch": 1.7534664800915913, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.229206562042236, "learning_rate": 1e-05, "loss": 0.5075, "mean_token_accuracy": 0.866299033164978, "num_tokens": 523198977.0, "step": 13784 }, { "epoch": 1.7535936903701819, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.422078609466553, "learning_rate": 1e-05, "loss": 0.5072, "mean_token_accuracy": 0.8639119267463684, "num_tokens": 523233917.0, "step": 13785 }, { "epoch": 1.7537209006487724, "ewc_loss": 0.11181640625, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.073807239532471, "learning_rate": 1e-05, "loss": 0.4779, "mean_token_accuracy": 0.8728845119476318, "num_tokens": 523267679.0, "step": 13786 }, { "epoch": 1.753848110927363, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.439728260040283, "learning_rate": 1e-05, "loss": 0.5309, "mean_token_accuracy": 0.8582059741020203, "num_tokens": 523301512.0, "step": 13787 }, { "epoch": 1.7539753212059535, "ewc_loss": 0.10791015625, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.895843982696533, "learning_rate": 1e-05, "loss": 0.4581, "mean_token_accuracy": 0.8770022988319397, "num_tokens": 523342618.0, "step": 13788 }, { "epoch": 1.754102531484544, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.559295654296875, "learning_rate": 1e-05, "loss": 0.4921, "mean_token_accuracy": 0.8699551820755005, "num_tokens": 523379555.0, "step": 13789 }, { "epoch": 1.7542297417631345, "ewc_loss": 0.10498046875, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.759796619415283, "learning_rate": 1e-05, "loss": 0.4243, "mean_token_accuracy": 0.8890334367752075, "num_tokens": 523414621.0, "step": 13790 }, { "epoch": 1.754356952041725, "ewc_loss": 0.125, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.813157081604004, "learning_rate": 1e-05, "loss": 0.4372, "mean_token_accuracy": 0.8927543759346008, "num_tokens": 523459022.0, "step": 13791 }, { "epoch": 1.7544841623203156, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.661655426025391, "learning_rate": 1e-05, "loss": 0.4452, "mean_token_accuracy": 0.8830583095550537, "num_tokens": 523498718.0, "step": 13792 }, { "epoch": 1.754611372598906, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 0.00011110305786132812, "grad_norm": 7.1425018310546875, "learning_rate": 1e-05, "loss": 0.4872, "mean_token_accuracy": 0.8818258047103882, "num_tokens": 523531926.0, "step": 13793 }, { "epoch": 1.7547385828774966, "ewc_loss": 0.109375, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.767859935760498, "learning_rate": 1e-05, "loss": 0.4766, "mean_token_accuracy": 0.871550440788269, "num_tokens": 523572457.0, "step": 13794 }, { "epoch": 1.7548657931560872, "ewc_loss": 0.140625, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00011777877807617188, "grad_norm": 7.3722004890441895, "learning_rate": 1e-05, "loss": 0.5492, "mean_token_accuracy": 0.8621422052383423, "num_tokens": 523607317.0, "step": 13795 }, { "epoch": 1.7549930034346777, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.1142778396606445, "learning_rate": 1e-05, "loss": 0.5126, "mean_token_accuracy": 0.8641119003295898, "num_tokens": 523638979.0, "step": 13796 }, { "epoch": 1.755120213713268, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00011396408081054688, "grad_norm": 7.214606761932373, "learning_rate": 1e-05, "loss": 0.5007, "mean_token_accuracy": 0.871658444404602, "num_tokens": 523681839.0, "step": 13797 }, { "epoch": 1.7552474239918585, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.378334045410156, "learning_rate": 1e-05, "loss": 0.4835, "mean_token_accuracy": 0.8759061694145203, "num_tokens": 523719797.0, "step": 13798 }, { "epoch": 1.755374634270449, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.766608238220215, "learning_rate": 1e-05, "loss": 0.4695, "mean_token_accuracy": 0.8756099939346313, "num_tokens": 523761774.0, "step": 13799 }, { "epoch": 1.7555018445490396, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.491657257080078, "learning_rate": 1e-05, "loss": 0.4964, "mean_token_accuracy": 0.871290385723114, "num_tokens": 523799522.0, "step": 13800 }, { "epoch": 1.75562905482763, "ewc_loss": 0.11474609375, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.260878086090088, "learning_rate": 1e-05, "loss": 0.4755, "mean_token_accuracy": 0.8721919059753418, "num_tokens": 523831927.0, "step": 13801 }, { "epoch": 1.7557562651062206, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.614612102508545, "learning_rate": 1e-05, "loss": 0.5291, "mean_token_accuracy": 0.8620887398719788, "num_tokens": 523863058.0, "step": 13802 }, { "epoch": 1.755883475384811, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.824137210845947, "learning_rate": 1e-05, "loss": 0.542, "mean_token_accuracy": 0.8495819568634033, "num_tokens": 523907017.0, "step": 13803 }, { "epoch": 1.7560106856634015, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.889845848083496, "learning_rate": 1e-05, "loss": 0.5269, "mean_token_accuracy": 0.8668396472930908, "num_tokens": 523943376.0, "step": 13804 }, { "epoch": 1.756137895941992, "ewc_loss": 0.10400390625, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.612524509429932, "learning_rate": 1e-05, "loss": 0.4564, "mean_token_accuracy": 0.8788172006607056, "num_tokens": 523982902.0, "step": 13805 }, { "epoch": 1.7562651062205825, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00011587142944335938, "grad_norm": 7.305085182189941, "learning_rate": 1e-05, "loss": 0.5334, "mean_token_accuracy": 0.8642850518226624, "num_tokens": 524016423.0, "step": 13806 }, { "epoch": 1.756392316499173, "ewc_loss": 0.109375, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.755768299102783, "learning_rate": 1e-05, "loss": 0.5417, "mean_token_accuracy": 0.8534450531005859, "num_tokens": 524056686.0, "step": 13807 }, { "epoch": 1.7565195267777636, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.0001239776611328125, "grad_norm": 7.586629867553711, "learning_rate": 1e-05, "loss": 0.5253, "mean_token_accuracy": 0.8691966533660889, "num_tokens": 524093731.0, "step": 13808 }, { "epoch": 1.756646737056354, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.19284725189209, "learning_rate": 1e-05, "loss": 0.5105, "mean_token_accuracy": 0.868430495262146, "num_tokens": 524139398.0, "step": 13809 }, { "epoch": 1.7567739473349446, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.0001201629638671875, "grad_norm": 7.448918342590332, "learning_rate": 1e-05, "loss": 0.5473, "mean_token_accuracy": 0.8626766204833984, "num_tokens": 524177012.0, "step": 13810 }, { "epoch": 1.7569011576135352, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.501665115356445, "learning_rate": 1e-05, "loss": 0.4615, "mean_token_accuracy": 0.884415864944458, "num_tokens": 524211785.0, "step": 13811 }, { "epoch": 1.7570283678921257, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 7.035904407501221, "learning_rate": 1e-05, "loss": 0.4705, "mean_token_accuracy": 0.8824236989021301, "num_tokens": 524252263.0, "step": 13812 }, { "epoch": 1.7571555781707162, "ewc_loss": 0.125, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.481314659118652, "learning_rate": 1e-05, "loss": 0.4452, "mean_token_accuracy": 0.8877119421958923, "num_tokens": 524286561.0, "step": 13813 }, { "epoch": 1.7572827884493067, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.6793928146362305, "learning_rate": 1e-05, "loss": 0.4828, "mean_token_accuracy": 0.8744254112243652, "num_tokens": 524320910.0, "step": 13814 }, { "epoch": 1.7574099987278973, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.358931064605713, "learning_rate": 1e-05, "loss": 0.4757, "mean_token_accuracy": 0.878430962562561, "num_tokens": 524362416.0, "step": 13815 }, { "epoch": 1.7575372090064878, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.454319953918457, "learning_rate": 1e-05, "loss": 0.4709, "mean_token_accuracy": 0.8780310153961182, "num_tokens": 524405432.0, "step": 13816 }, { "epoch": 1.7576644192850783, "ewc_loss": 0.11474609375, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.207225799560547, "learning_rate": 1e-05, "loss": 0.4829, "mean_token_accuracy": 0.8732974529266357, "num_tokens": 524440605.0, "step": 13817 }, { "epoch": 1.7577916295636689, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.386181354522705, "learning_rate": 1e-05, "loss": 0.4997, "mean_token_accuracy": 0.8693720102310181, "num_tokens": 524483126.0, "step": 13818 }, { "epoch": 1.7579188398422594, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 6.051158428192139, "learning_rate": 1e-05, "loss": 0.4691, "mean_token_accuracy": 0.8772275447845459, "num_tokens": 524516675.0, "step": 13819 }, { "epoch": 1.75804605012085, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.419775485992432, "learning_rate": 1e-05, "loss": 0.4751, "mean_token_accuracy": 0.8725910186767578, "num_tokens": 524554394.0, "step": 13820 }, { "epoch": 1.7581732603994404, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.874101161956787, "learning_rate": 1e-05, "loss": 0.4932, "mean_token_accuracy": 0.8663939833641052, "num_tokens": 524591195.0, "step": 13821 }, { "epoch": 1.7583004706780307, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.576269626617432, "learning_rate": 1e-05, "loss": 0.513, "mean_token_accuracy": 0.8621646165847778, "num_tokens": 524629077.0, "step": 13822 }, { "epoch": 1.7584276809566213, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.707611083984375, "learning_rate": 1e-05, "loss": 0.5425, "mean_token_accuracy": 0.8536301851272583, "num_tokens": 524672072.0, "step": 13823 }, { "epoch": 1.7585548912352118, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.876843452453613, "learning_rate": 1e-05, "loss": 0.4992, "mean_token_accuracy": 0.8718577027320862, "num_tokens": 524706735.0, "step": 13824 }, { "epoch": 1.7586821015138023, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.6187968254089355, "learning_rate": 1e-05, "loss": 0.4622, "mean_token_accuracy": 0.8759841322898865, "num_tokens": 524741750.0, "step": 13825 }, { "epoch": 1.7588093117923929, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00011444091796875, "grad_norm": 7.269108772277832, "learning_rate": 1e-05, "loss": 0.463, "mean_token_accuracy": 0.8869146108627319, "num_tokens": 524774565.0, "step": 13826 }, { "epoch": 1.7589365220709834, "ewc_loss": 0.10888671875, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.787078857421875, "learning_rate": 1e-05, "loss": 0.4525, "mean_token_accuracy": 0.8807909488677979, "num_tokens": 524803731.0, "step": 13827 }, { "epoch": 1.7590637323495737, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.0001220703125, "grad_norm": 7.521878242492676, "learning_rate": 1e-05, "loss": 0.5348, "mean_token_accuracy": 0.8668928146362305, "num_tokens": 524842570.0, "step": 13828 }, { "epoch": 1.7591909426281642, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.216890811920166, "learning_rate": 1e-05, "loss": 0.5051, "mean_token_accuracy": 0.8707789778709412, "num_tokens": 524880695.0, "step": 13829 }, { "epoch": 1.7593181529067548, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00011587142944335938, "grad_norm": 7.313786506652832, "learning_rate": 1e-05, "loss": 0.5053, "mean_token_accuracy": 0.8722180128097534, "num_tokens": 524921331.0, "step": 13830 }, { "epoch": 1.7594453631853453, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.512174129486084, "learning_rate": 1e-05, "loss": 0.5196, "mean_token_accuracy": 0.8631329536437988, "num_tokens": 524955769.0, "step": 13831 }, { "epoch": 1.7595725734639358, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.820420265197754, "learning_rate": 1e-05, "loss": 0.5504, "mean_token_accuracy": 0.8565661907196045, "num_tokens": 524998299.0, "step": 13832 }, { "epoch": 1.7596997837425263, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.557636260986328, "learning_rate": 1e-05, "loss": 0.5469, "mean_token_accuracy": 0.8587747812271118, "num_tokens": 525038590.0, "step": 13833 }, { "epoch": 1.7598269940211169, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.308624267578125, "learning_rate": 1e-05, "loss": 0.4989, "mean_token_accuracy": 0.8693370223045349, "num_tokens": 525073688.0, "step": 13834 }, { "epoch": 1.7599542042997074, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.615075588226318, "learning_rate": 1e-05, "loss": 0.442, "mean_token_accuracy": 0.8902817368507385, "num_tokens": 525112109.0, "step": 13835 }, { "epoch": 1.760081414578298, "ewc_loss": 0.10693359375, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.91887903213501, "learning_rate": 1e-05, "loss": 0.4684, "mean_token_accuracy": 0.8760372400283813, "num_tokens": 525147554.0, "step": 13836 }, { "epoch": 1.7602086248568884, "ewc_loss": 0.125, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.800755023956299, "learning_rate": 1e-05, "loss": 0.5304, "mean_token_accuracy": 0.8599123954772949, "num_tokens": 525187233.0, "step": 13837 }, { "epoch": 1.760335835135479, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.724704265594482, "learning_rate": 1e-05, "loss": 0.5208, "mean_token_accuracy": 0.8603003621101379, "num_tokens": 525218104.0, "step": 13838 }, { "epoch": 1.7604630454140695, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 7.072839260101318, "learning_rate": 1e-05, "loss": 0.4907, "mean_token_accuracy": 0.8750488758087158, "num_tokens": 525255690.0, "step": 13839 }, { "epoch": 1.76059025569266, "ewc_loss": 0.10693359375, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.729328155517578, "learning_rate": 1e-05, "loss": 0.4952, "mean_token_accuracy": 0.8698068261146545, "num_tokens": 525287972.0, "step": 13840 }, { "epoch": 1.7607174659712506, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00011730194091796875, "grad_norm": 7.341124057769775, "learning_rate": 1e-05, "loss": 0.5439, "mean_token_accuracy": 0.8625410795211792, "num_tokens": 525328003.0, "step": 13841 }, { "epoch": 1.760844676249841, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 5.9600605964660645, "learning_rate": 1e-05, "loss": 0.4737, "mean_token_accuracy": 0.8767257928848267, "num_tokens": 525370642.0, "step": 13842 }, { "epoch": 1.7609718865284316, "ewc_loss": 0.140625, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00011777877807617188, "grad_norm": 7.394686222076416, "learning_rate": 1e-05, "loss": 0.5677, "mean_token_accuracy": 0.8572167754173279, "num_tokens": 525408668.0, "step": 13843 }, { "epoch": 1.7610990968070221, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.250918388366699, "learning_rate": 1e-05, "loss": 0.4892, "mean_token_accuracy": 0.871703028678894, "num_tokens": 525447566.0, "step": 13844 }, { "epoch": 1.7612263070856127, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.0001125335693359375, "grad_norm": 7.173726558685303, "learning_rate": 1e-05, "loss": 0.5315, "mean_token_accuracy": 0.8630993366241455, "num_tokens": 525485456.0, "step": 13845 }, { "epoch": 1.761353517364203, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.375853538513184, "learning_rate": 1e-05, "loss": 0.5508, "mean_token_accuracy": 0.8556740880012512, "num_tokens": 525521190.0, "step": 13846 }, { "epoch": 1.7614807276427935, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.83900260925293, "learning_rate": 1e-05, "loss": 0.5028, "mean_token_accuracy": 0.8724080324172974, "num_tokens": 525559331.0, "step": 13847 }, { "epoch": 1.761607937921384, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.340122699737549, "learning_rate": 1e-05, "loss": 0.5228, "mean_token_accuracy": 0.8625442981719971, "num_tokens": 525598380.0, "step": 13848 }, { "epoch": 1.7617351481999746, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.568329334259033, "learning_rate": 1e-05, "loss": 0.5121, "mean_token_accuracy": 0.8604953289031982, "num_tokens": 525637901.0, "step": 13849 }, { "epoch": 1.761862358478565, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.2567620277404785, "learning_rate": 1e-05, "loss": 0.5125, "mean_token_accuracy": 0.8655480146408081, "num_tokens": 525676072.0, "step": 13850 }, { "epoch": 1.7619895687571556, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.40721321105957, "learning_rate": 1e-05, "loss": 0.4417, "mean_token_accuracy": 0.8859463930130005, "num_tokens": 525710655.0, "step": 13851 }, { "epoch": 1.762116779035746, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.136815071105957, "learning_rate": 1e-05, "loss": 0.4631, "mean_token_accuracy": 0.8757447600364685, "num_tokens": 525745714.0, "step": 13852 }, { "epoch": 1.7622439893143365, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.386045932769775, "learning_rate": 1e-05, "loss": 0.44, "mean_token_accuracy": 0.8852684497833252, "num_tokens": 525782246.0, "step": 13853 }, { "epoch": 1.762371199592927, "ewc_loss": 0.10986328125, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.987834453582764, "learning_rate": 1e-05, "loss": 0.4358, "mean_token_accuracy": 0.8860026597976685, "num_tokens": 525814232.0, "step": 13854 }, { "epoch": 1.7624984098715175, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.442798137664795, "learning_rate": 1e-05, "loss": 0.502, "mean_token_accuracy": 0.8705703616142273, "num_tokens": 525847775.0, "step": 13855 }, { "epoch": 1.762625620150108, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.834563732147217, "learning_rate": 1e-05, "loss": 0.4763, "mean_token_accuracy": 0.8730049729347229, "num_tokens": 525890760.0, "step": 13856 }, { "epoch": 1.7627528304286986, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.6240668296813965, "learning_rate": 1e-05, "loss": 0.4644, "mean_token_accuracy": 0.8808446526527405, "num_tokens": 525926893.0, "step": 13857 }, { "epoch": 1.762880040707289, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.69703483581543, "learning_rate": 1e-05, "loss": 0.5126, "mean_token_accuracy": 0.8613945245742798, "num_tokens": 525968907.0, "step": 13858 }, { "epoch": 1.7630072509858796, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.924374103546143, "learning_rate": 1e-05, "loss": 0.4821, "mean_token_accuracy": 0.8751177787780762, "num_tokens": 526003988.0, "step": 13859 }, { "epoch": 1.7631344612644702, "ewc_loss": 0.10498046875, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.642426013946533, "learning_rate": 1e-05, "loss": 0.4714, "mean_token_accuracy": 0.8738664388656616, "num_tokens": 526044575.0, "step": 13860 }, { "epoch": 1.7632616715430607, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00011539459228515625, "grad_norm": 7.2763190269470215, "learning_rate": 1e-05, "loss": 0.5144, "mean_token_accuracy": 0.8715617656707764, "num_tokens": 526087803.0, "step": 13861 }, { "epoch": 1.7633888818216512, "ewc_loss": 0.11083984375, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 5.837479591369629, "learning_rate": 1e-05, "loss": 0.5244, "mean_token_accuracy": 0.8561593294143677, "num_tokens": 526123432.0, "step": 13862 }, { "epoch": 1.7635160921002417, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.0001201629638671875, "grad_norm": 7.45094108581543, "learning_rate": 1e-05, "loss": 0.4826, "mean_token_accuracy": 0.8816295862197876, "num_tokens": 526161285.0, "step": 13863 }, { "epoch": 1.7636433023788323, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.2297868728637695, "learning_rate": 1e-05, "loss": 0.5619, "mean_token_accuracy": 0.8549031019210815, "num_tokens": 526203096.0, "step": 13864 }, { "epoch": 1.7637705126574228, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00011348724365234375, "grad_norm": 7.1910319328308105, "learning_rate": 1e-05, "loss": 0.4708, "mean_token_accuracy": 0.8825376033782959, "num_tokens": 526237741.0, "step": 13865 }, { "epoch": 1.7638977229360133, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.456974029541016, "learning_rate": 1e-05, "loss": 0.4943, "mean_token_accuracy": 0.8727276921272278, "num_tokens": 526279704.0, "step": 13866 }, { "epoch": 1.7640249332146039, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.712940216064453, "learning_rate": 1e-05, "loss": 0.5017, "mean_token_accuracy": 0.8688641786575317, "num_tokens": 526315708.0, "step": 13867 }, { "epoch": 1.7641521434931944, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.521435260772705, "learning_rate": 1e-05, "loss": 0.5069, "mean_token_accuracy": 0.8669517636299133, "num_tokens": 526360725.0, "step": 13868 }, { "epoch": 1.764279353771785, "ewc_loss": 0.11376953125, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.213131904602051, "learning_rate": 1e-05, "loss": 0.5121, "mean_token_accuracy": 0.8648425340652466, "num_tokens": 526398920.0, "step": 13869 }, { "epoch": 1.7644065640503754, "ewc_loss": 0.125, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.639225006103516, "learning_rate": 1e-05, "loss": 0.4462, "mean_token_accuracy": 0.8852317333221436, "num_tokens": 526433453.0, "step": 13870 }, { "epoch": 1.7645337743289657, "ewc_loss": 0.10595703125, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.827737331390381, "learning_rate": 1e-05, "loss": 0.4559, "mean_token_accuracy": 0.8786846995353699, "num_tokens": 526476256.0, "step": 13871 }, { "epoch": 1.7646609846075563, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.894542217254639, "learning_rate": 1e-05, "loss": 0.4948, "mean_token_accuracy": 0.8734841346740723, "num_tokens": 526517707.0, "step": 13872 }, { "epoch": 1.7647881948861468, "ewc_loss": 0.10400390625, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.663379669189453, "learning_rate": 1e-05, "loss": 0.4803, "mean_token_accuracy": 0.8720817565917969, "num_tokens": 526551696.0, "step": 13873 }, { "epoch": 1.7649154051647373, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00011348724365234375, "grad_norm": 7.247352600097656, "learning_rate": 1e-05, "loss": 0.4756, "mean_token_accuracy": 0.881391167640686, "num_tokens": 526587884.0, "step": 13874 }, { "epoch": 1.7650426154433279, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 5.787975311279297, "learning_rate": 1e-05, "loss": 0.4974, "mean_token_accuracy": 0.8669629096984863, "num_tokens": 526625398.0, "step": 13875 }, { "epoch": 1.7651698257219184, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00012063980102539062, "grad_norm": 7.474311828613281, "learning_rate": 1e-05, "loss": 0.5325, "mean_token_accuracy": 0.8687103986740112, "num_tokens": 526662481.0, "step": 13876 }, { "epoch": 1.7652970360005087, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.142363548278809, "learning_rate": 1e-05, "loss": 0.5209, "mean_token_accuracy": 0.8606211543083191, "num_tokens": 526700719.0, "step": 13877 }, { "epoch": 1.7654242462790992, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00011730194091796875, "grad_norm": 7.361260414123535, "learning_rate": 1e-05, "loss": 0.5102, "mean_token_accuracy": 0.8709061741828918, "num_tokens": 526732941.0, "step": 13878 }, { "epoch": 1.7655514565576897, "ewc_loss": 0.125, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.375429153442383, "learning_rate": 1e-05, "loss": 0.4954, "mean_token_accuracy": 0.8735146522521973, "num_tokens": 526777825.0, "step": 13879 }, { "epoch": 1.7656786668362803, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 7.047536849975586, "learning_rate": 1e-05, "loss": 0.5233, "mean_token_accuracy": 0.8644585609436035, "num_tokens": 526811331.0, "step": 13880 }, { "epoch": 1.7658058771148708, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.404425144195557, "learning_rate": 1e-05, "loss": 0.4993, "mean_token_accuracy": 0.8712713718414307, "num_tokens": 526846788.0, "step": 13881 }, { "epoch": 1.7659330873934613, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.712654113769531, "learning_rate": 1e-05, "loss": 0.4782, "mean_token_accuracy": 0.8776366710662842, "num_tokens": 526880127.0, "step": 13882 }, { "epoch": 1.7660602976720519, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.320003986358643, "learning_rate": 1e-05, "loss": 0.5151, "mean_token_accuracy": 0.8618905544281006, "num_tokens": 526914099.0, "step": 13883 }, { "epoch": 1.7661875079506424, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.487357139587402, "learning_rate": 1e-05, "loss": 0.5233, "mean_token_accuracy": 0.8642522096633911, "num_tokens": 526956409.0, "step": 13884 }, { "epoch": 1.766314718229233, "ewc_loss": 0.11474609375, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.186298370361328, "learning_rate": 1e-05, "loss": 0.4961, "mean_token_accuracy": 0.8706998825073242, "num_tokens": 526995175.0, "step": 13885 }, { "epoch": 1.7664419285078234, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.406155586242676, "learning_rate": 1e-05, "loss": 0.4958, "mean_token_accuracy": 0.8711743354797363, "num_tokens": 527035833.0, "step": 13886 }, { "epoch": 1.766569138786414, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 6.0455427169799805, "learning_rate": 1e-05, "loss": 0.4889, "mean_token_accuracy": 0.8698413372039795, "num_tokens": 527078780.0, "step": 13887 }, { "epoch": 1.7666963490650045, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.4250359535217285, "learning_rate": 1e-05, "loss": 0.4776, "mean_token_accuracy": 0.8787089586257935, "num_tokens": 527119494.0, "step": 13888 }, { "epoch": 1.766823559343595, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.893972873687744, "learning_rate": 1e-05, "loss": 0.4995, "mean_token_accuracy": 0.8685469627380371, "num_tokens": 527158112.0, "step": 13889 }, { "epoch": 1.7669507696221856, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.564451217651367, "learning_rate": 1e-05, "loss": 0.4856, "mean_token_accuracy": 0.8738517165184021, "num_tokens": 527195885.0, "step": 13890 }, { "epoch": 1.767077979900776, "ewc_loss": 0.10498046875, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.747229099273682, "learning_rate": 1e-05, "loss": 0.4791, "mean_token_accuracy": 0.8710103034973145, "num_tokens": 527229681.0, "step": 13891 }, { "epoch": 1.7672051901793666, "ewc_loss": 0.125, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.829645156860352, "learning_rate": 1e-05, "loss": 0.5626, "mean_token_accuracy": 0.8466393351554871, "num_tokens": 527267248.0, "step": 13892 }, { "epoch": 1.7673324004579571, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.6534342765808105, "learning_rate": 1e-05, "loss": 0.4723, "mean_token_accuracy": 0.8713630437850952, "num_tokens": 527299695.0, "step": 13893 }, { "epoch": 1.7674596107365477, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00011396408081054688, "grad_norm": 7.179422378540039, "learning_rate": 1e-05, "loss": 0.5178, "mean_token_accuracy": 0.868266224861145, "num_tokens": 527334398.0, "step": 13894 }, { "epoch": 1.767586821015138, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.756655216217041, "learning_rate": 1e-05, "loss": 0.5368, "mean_token_accuracy": 0.8565776348114014, "num_tokens": 527376954.0, "step": 13895 }, { "epoch": 1.7677140312937285, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00011920928955078125, "grad_norm": 7.434101581573486, "learning_rate": 1e-05, "loss": 0.5267, "mean_token_accuracy": 0.8698917627334595, "num_tokens": 527419737.0, "step": 13896 }, { "epoch": 1.767841241572319, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.13097620010376, "learning_rate": 1e-05, "loss": 0.548, "mean_token_accuracy": 0.8520110845565796, "num_tokens": 527455874.0, "step": 13897 }, { "epoch": 1.7679684518509096, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.0001163482666015625, "grad_norm": 7.292161464691162, "learning_rate": 1e-05, "loss": 0.5899, "mean_token_accuracy": 0.8447396159172058, "num_tokens": 527494796.0, "step": 13898 }, { "epoch": 1.7680956621295, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.425099849700928, "learning_rate": 1e-05, "loss": 0.5571, "mean_token_accuracy": 0.853049099445343, "num_tokens": 527536955.0, "step": 13899 }, { "epoch": 1.7682228724080906, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.840796947479248, "learning_rate": 1e-05, "loss": 0.5048, "mean_token_accuracy": 0.8673954010009766, "num_tokens": 527569472.0, "step": 13900 }, { "epoch": 1.768350082686681, "ewc_loss": 0.125, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.512999057769775, "learning_rate": 1e-05, "loss": 0.4834, "mean_token_accuracy": 0.8725255727767944, "num_tokens": 527603641.0, "step": 13901 }, { "epoch": 1.7684772929652715, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.3176984786987305, "learning_rate": 1e-05, "loss": 0.5123, "mean_token_accuracy": 0.8626981973648071, "num_tokens": 527642942.0, "step": 13902 }, { "epoch": 1.768604503243862, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.600112438201904, "learning_rate": 1e-05, "loss": 0.5151, "mean_token_accuracy": 0.8654829263687134, "num_tokens": 527682687.0, "step": 13903 }, { "epoch": 1.7687317135224525, "ewc_loss": 0.10693359375, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.899003028869629, "learning_rate": 1e-05, "loss": 0.4443, "mean_token_accuracy": 0.8812508583068848, "num_tokens": 527721355.0, "step": 13904 }, { "epoch": 1.768858923801043, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.820542812347412, "learning_rate": 1e-05, "loss": 0.4955, "mean_token_accuracy": 0.8717638850212097, "num_tokens": 527760887.0, "step": 13905 }, { "epoch": 1.7689861340796336, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.671288013458252, "learning_rate": 1e-05, "loss": 0.5199, "mean_token_accuracy": 0.8587175607681274, "num_tokens": 527808669.0, "step": 13906 }, { "epoch": 1.769113344358224, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.0001125335693359375, "grad_norm": 7.1779961585998535, "learning_rate": 1e-05, "loss": 0.5602, "mean_token_accuracy": 0.8586454391479492, "num_tokens": 527840262.0, "step": 13907 }, { "epoch": 1.7692405546368146, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.714694023132324, "learning_rate": 1e-05, "loss": 0.5179, "mean_token_accuracy": 0.859642505645752, "num_tokens": 527880822.0, "step": 13908 }, { "epoch": 1.7693677649154052, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00012111663818359375, "grad_norm": 7.4776177406311035, "learning_rate": 1e-05, "loss": 0.515, "mean_token_accuracy": 0.8745516538619995, "num_tokens": 527923677.0, "step": 13909 }, { "epoch": 1.7694949751939957, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.065171241760254, "learning_rate": 1e-05, "loss": 0.5368, "mean_token_accuracy": 0.8569545745849609, "num_tokens": 527966004.0, "step": 13910 }, { "epoch": 1.7696221854725862, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00012063980102539062, "grad_norm": 7.442891597747803, "learning_rate": 1e-05, "loss": 0.5286, "mean_token_accuracy": 0.8707901239395142, "num_tokens": 528004725.0, "step": 13911 }, { "epoch": 1.7697493957511767, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.38012170791626, "learning_rate": 1e-05, "loss": 0.5066, "mean_token_accuracy": 0.8684185743331909, "num_tokens": 528043262.0, "step": 13912 }, { "epoch": 1.7698766060297673, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 7.132359981536865, "learning_rate": 1e-05, "loss": 0.5327, "mean_token_accuracy": 0.8651721477508545, "num_tokens": 528082810.0, "step": 13913 }, { "epoch": 1.7700038163083578, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.441588878631592, "learning_rate": 1e-05, "loss": 0.5177, "mean_token_accuracy": 0.8629540205001831, "num_tokens": 528117672.0, "step": 13914 }, { "epoch": 1.7701310265869483, "ewc_loss": 0.125, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.782973766326904, "learning_rate": 1e-05, "loss": 0.5088, "mean_token_accuracy": 0.8691710233688354, "num_tokens": 528155756.0, "step": 13915 }, { "epoch": 1.7702582368655388, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.34976053237915, "learning_rate": 1e-05, "loss": 0.4738, "mean_token_accuracy": 0.8773400783538818, "num_tokens": 528194664.0, "step": 13916 }, { "epoch": 1.7703854471441294, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.536828517913818, "learning_rate": 1e-05, "loss": 0.457, "mean_token_accuracy": 0.8813222646713257, "num_tokens": 528234334.0, "step": 13917 }, { "epoch": 1.77051265742272, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.224185466766357, "learning_rate": 1e-05, "loss": 0.4464, "mean_token_accuracy": 0.8842151761054993, "num_tokens": 528270543.0, "step": 13918 }, { "epoch": 1.7706398677013102, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.426438808441162, "learning_rate": 1e-05, "loss": 0.5161, "mean_token_accuracy": 0.862363338470459, "num_tokens": 528309698.0, "step": 13919 }, { "epoch": 1.7707670779799007, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 6.074815273284912, "learning_rate": 1e-05, "loss": 0.5051, "mean_token_accuracy": 0.8645857572555542, "num_tokens": 528356396.0, "step": 13920 }, { "epoch": 1.7708942882584913, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.448873996734619, "learning_rate": 1e-05, "loss": 0.5152, "mean_token_accuracy": 0.862168550491333, "num_tokens": 528394542.0, "step": 13921 }, { "epoch": 1.7710214985370818, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.908212184906006, "learning_rate": 1e-05, "loss": 0.4474, "mean_token_accuracy": 0.8814171552658081, "num_tokens": 528435433.0, "step": 13922 }, { "epoch": 1.7711487088156723, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.584203243255615, "learning_rate": 1e-05, "loss": 0.4521, "mean_token_accuracy": 0.885409951210022, "num_tokens": 528471152.0, "step": 13923 }, { "epoch": 1.7712759190942629, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.763779163360596, "learning_rate": 1e-05, "loss": 0.4552, "mean_token_accuracy": 0.8780720233917236, "num_tokens": 528505084.0, "step": 13924 }, { "epoch": 1.7714031293728534, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.821343421936035, "learning_rate": 1e-05, "loss": 0.5933, "mean_token_accuracy": 0.8426264524459839, "num_tokens": 528545234.0, "step": 13925 }, { "epoch": 1.7715303396514437, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.677030086517334, "learning_rate": 1e-05, "loss": 0.4682, "mean_token_accuracy": 0.8735988140106201, "num_tokens": 528584099.0, "step": 13926 }, { "epoch": 1.7716575499300342, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 0.00011110305786132812, "grad_norm": 7.143542766571045, "learning_rate": 1e-05, "loss": 0.5491, "mean_token_accuracy": 0.8607982397079468, "num_tokens": 528624143.0, "step": 13927 }, { "epoch": 1.7717847602086247, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.778924465179443, "learning_rate": 1e-05, "loss": 0.4852, "mean_token_accuracy": 0.8682857155799866, "num_tokens": 528661521.0, "step": 13928 }, { "epoch": 1.7719119704872153, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00011730194091796875, "grad_norm": 7.361450672149658, "learning_rate": 1e-05, "loss": 0.488, "mean_token_accuracy": 0.8801367282867432, "num_tokens": 528699317.0, "step": 13929 }, { "epoch": 1.7720391807658058, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.120364665985107, "learning_rate": 1e-05, "loss": 0.5402, "mean_token_accuracy": 0.8552836179733276, "num_tokens": 528737277.0, "step": 13930 }, { "epoch": 1.7721663910443963, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00011301040649414062, "grad_norm": 7.207790851593018, "learning_rate": 1e-05, "loss": 0.5529, "mean_token_accuracy": 0.8579428195953369, "num_tokens": 528777076.0, "step": 13931 }, { "epoch": 1.7722936013229869, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.388140678405762, "learning_rate": 1e-05, "loss": 0.4956, "mean_token_accuracy": 0.8709887266159058, "num_tokens": 528813940.0, "step": 13932 }, { "epoch": 1.7724208116015774, "ewc_loss": 0.125, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.758902549743652, "learning_rate": 1e-05, "loss": 0.446, "mean_token_accuracy": 0.8889926671981812, "num_tokens": 528852883.0, "step": 13933 }, { "epoch": 1.772548021880168, "ewc_loss": 0.125, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.4880146980285645, "learning_rate": 1e-05, "loss": 0.5334, "mean_token_accuracy": 0.8602745532989502, "num_tokens": 528891561.0, "step": 13934 }, { "epoch": 1.7726752321587584, "ewc_loss": 0.11474609375, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.2337775230407715, "learning_rate": 1e-05, "loss": 0.4963, "mean_token_accuracy": 0.8664939999580383, "num_tokens": 528930755.0, "step": 13935 }, { "epoch": 1.772802442437349, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.626677989959717, "learning_rate": 1e-05, "loss": 0.5127, "mean_token_accuracy": 0.8650409579277039, "num_tokens": 528969230.0, "step": 13936 }, { "epoch": 1.7729296527159395, "ewc_loss": 0.10595703125, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.811720371246338, "learning_rate": 1e-05, "loss": 0.4627, "mean_token_accuracy": 0.8751891851425171, "num_tokens": 529005157.0, "step": 13937 }, { "epoch": 1.77305686299453, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.906313896179199, "learning_rate": 1e-05, "loss": 0.4762, "mean_token_accuracy": 0.8759222030639648, "num_tokens": 529042934.0, "step": 13938 }, { "epoch": 1.7731840732731206, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.609910011291504, "learning_rate": 1e-05, "loss": 0.4811, "mean_token_accuracy": 0.8734232187271118, "num_tokens": 529087811.0, "step": 13939 }, { "epoch": 1.773311283551711, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00011587142944335938, "grad_norm": 7.319954872131348, "learning_rate": 1e-05, "loss": 0.5554, "mean_token_accuracy": 0.8600556254386902, "num_tokens": 529133143.0, "step": 13940 }, { "epoch": 1.7734384938303016, "ewc_loss": 0.10986328125, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.766307830810547, "learning_rate": 1e-05, "loss": 0.4515, "mean_token_accuracy": 0.8823950290679932, "num_tokens": 529172737.0, "step": 13941 }, { "epoch": 1.7735657041088921, "ewc_loss": 0.1455078125, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.607898712158203, "learning_rate": 1e-05, "loss": 0.5176, "mean_token_accuracy": 0.8701493740081787, "num_tokens": 529208913.0, "step": 13942 }, { "epoch": 1.7736929143874827, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.227525234222412, "learning_rate": 1e-05, "loss": 0.5143, "mean_token_accuracy": 0.8670952320098877, "num_tokens": 529250905.0, "step": 13943 }, { "epoch": 1.773820124666073, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 0.00011920928955078125, "grad_norm": 7.4395222663879395, "learning_rate": 1e-05, "loss": 0.4672, "mean_token_accuracy": 0.8867188096046448, "num_tokens": 529289542.0, "step": 13944 }, { "epoch": 1.7739473349446635, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.507448673248291, "learning_rate": 1e-05, "loss": 0.4791, "mean_token_accuracy": 0.8769432306289673, "num_tokens": 529324759.0, "step": 13945 }, { "epoch": 1.774074545223254, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 7.059359073638916, "learning_rate": 1e-05, "loss": 0.5118, "mean_token_accuracy": 0.8701555132865906, "num_tokens": 529361845.0, "step": 13946 }, { "epoch": 1.7742017555018446, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.470531940460205, "learning_rate": 1e-05, "loss": 0.5038, "mean_token_accuracy": 0.8701414465904236, "num_tokens": 529402220.0, "step": 13947 }, { "epoch": 1.774328965780435, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.7105488777160645, "learning_rate": 1e-05, "loss": 0.5402, "mean_token_accuracy": 0.8573057651519775, "num_tokens": 529440532.0, "step": 13948 }, { "epoch": 1.7744561760590256, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.346174240112305, "learning_rate": 1e-05, "loss": 0.473, "mean_token_accuracy": 0.8748845458030701, "num_tokens": 529480357.0, "step": 13949 }, { "epoch": 1.774583386337616, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.498151779174805, "learning_rate": 1e-05, "loss": 0.5154, "mean_token_accuracy": 0.8633748292922974, "num_tokens": 529514075.0, "step": 13950 }, { "epoch": 1.7747105966162064, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.203685283660889, "learning_rate": 1e-05, "loss": 0.5467, "mean_token_accuracy": 0.8564029932022095, "num_tokens": 529550738.0, "step": 13951 }, { "epoch": 1.774837806894797, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.424839496612549, "learning_rate": 1e-05, "loss": 0.5277, "mean_token_accuracy": 0.862254798412323, "num_tokens": 529594151.0, "step": 13952 }, { "epoch": 1.7749650171733875, "ewc_loss": 0.10888671875, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 6.033142566680908, "learning_rate": 1e-05, "loss": 0.4743, "mean_token_accuracy": 0.8722061514854431, "num_tokens": 529633631.0, "step": 13953 }, { "epoch": 1.775092227451978, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.469524383544922, "learning_rate": 1e-05, "loss": 0.4976, "mean_token_accuracy": 0.8662527799606323, "num_tokens": 529674377.0, "step": 13954 }, { "epoch": 1.7752194377305686, "ewc_loss": 0.10693359375, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.875865459442139, "learning_rate": 1e-05, "loss": 0.5155, "mean_token_accuracy": 0.8597200512886047, "num_tokens": 529707209.0, "step": 13955 }, { "epoch": 1.775346648009159, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.640401363372803, "learning_rate": 1e-05, "loss": 0.509, "mean_token_accuracy": 0.8674459457397461, "num_tokens": 529737546.0, "step": 13956 }, { "epoch": 1.7754738582877496, "ewc_loss": 0.10498046875, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.7270402908325195, "learning_rate": 1e-05, "loss": 0.4886, "mean_token_accuracy": 0.8694047927856445, "num_tokens": 529775532.0, "step": 13957 }, { "epoch": 1.7756010685663401, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.8949432373046875, "learning_rate": 1e-05, "loss": 0.4821, "mean_token_accuracy": 0.8757196068763733, "num_tokens": 529814705.0, "step": 13958 }, { "epoch": 1.7757282788449307, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.681349754333496, "learning_rate": 1e-05, "loss": 0.4833, "mean_token_accuracy": 0.8693675994873047, "num_tokens": 529850912.0, "step": 13959 }, { "epoch": 1.7758554891235212, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00011396408081054688, "grad_norm": 7.216712951660156, "learning_rate": 1e-05, "loss": 0.4792, "mean_token_accuracy": 0.8768293857574463, "num_tokens": 529889042.0, "step": 13960 }, { "epoch": 1.7759826994021117, "ewc_loss": 0.11181640625, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 5.848089694976807, "learning_rate": 1e-05, "loss": 0.4755, "mean_token_accuracy": 0.8725565671920776, "num_tokens": 529926749.0, "step": 13961 }, { "epoch": 1.7761099096807023, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00011730194091796875, "grad_norm": 7.392811298370361, "learning_rate": 1e-05, "loss": 0.4953, "mean_token_accuracy": 0.8740153312683105, "num_tokens": 529951775.0, "step": 13962 }, { "epoch": 1.7762371199592928, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.184773921966553, "learning_rate": 1e-05, "loss": 0.4572, "mean_token_accuracy": 0.884410560131073, "num_tokens": 529996172.0, "step": 13963 }, { "epoch": 1.7763643302378833, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00011301040649414062, "grad_norm": 7.180354118347168, "learning_rate": 1e-05, "loss": 0.5664, "mean_token_accuracy": 0.8517866134643555, "num_tokens": 530039577.0, "step": 13964 }, { "epoch": 1.7764915405164738, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.436570644378662, "learning_rate": 1e-05, "loss": 0.4864, "mean_token_accuracy": 0.87967848777771, "num_tokens": 530076516.0, "step": 13965 }, { "epoch": 1.7766187507950644, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.693094730377197, "learning_rate": 1e-05, "loss": 0.47, "mean_token_accuracy": 0.8797762393951416, "num_tokens": 530118552.0, "step": 13966 }, { "epoch": 1.776745961073655, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.524328708648682, "learning_rate": 1e-05, "loss": 0.5186, "mean_token_accuracy": 0.8630460500717163, "num_tokens": 530154971.0, "step": 13967 }, { "epoch": 1.7768731713522452, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.179682731628418, "learning_rate": 1e-05, "loss": 0.4394, "mean_token_accuracy": 0.8822410702705383, "num_tokens": 530190509.0, "step": 13968 }, { "epoch": 1.7770003816308357, "ewc_loss": 0.125, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.663887977600098, "learning_rate": 1e-05, "loss": 0.4832, "mean_token_accuracy": 0.8763712644577026, "num_tokens": 530229814.0, "step": 13969 }, { "epoch": 1.7771275919094263, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.787419319152832, "learning_rate": 1e-05, "loss": 0.5025, "mean_token_accuracy": 0.8639273643493652, "num_tokens": 530269426.0, "step": 13970 }, { "epoch": 1.7772548021880168, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.968540668487549, "learning_rate": 1e-05, "loss": 0.5245, "mean_token_accuracy": 0.8632251024246216, "num_tokens": 530305036.0, "step": 13971 }, { "epoch": 1.7773820124666073, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.622988700866699, "learning_rate": 1e-05, "loss": 0.4968, "mean_token_accuracy": 0.866867184638977, "num_tokens": 530344995.0, "step": 13972 }, { "epoch": 1.7775092227451978, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 0.00011873245239257812, "grad_norm": 7.371156692504883, "learning_rate": 1e-05, "loss": 0.5104, "mean_token_accuracy": 0.875373363494873, "num_tokens": 530378283.0, "step": 13973 }, { "epoch": 1.7776364330237884, "ewc_loss": 0.11181640625, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 5.8292236328125, "learning_rate": 1e-05, "loss": 0.5444, "mean_token_accuracy": 0.8533094525337219, "num_tokens": 530416147.0, "step": 13974 }, { "epoch": 1.7777636433023787, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.581007957458496, "learning_rate": 1e-05, "loss": 0.5246, "mean_token_accuracy": 0.8687068819999695, "num_tokens": 530452096.0, "step": 13975 }, { "epoch": 1.7778908535809692, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.27838659286499, "learning_rate": 1e-05, "loss": 0.5362, "mean_token_accuracy": 0.858714759349823, "num_tokens": 530489347.0, "step": 13976 }, { "epoch": 1.7780180638595597, "ewc_loss": 0.140625, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 0.000118255615234375, "grad_norm": 7.357666969299316, "learning_rate": 1e-05, "loss": 0.5307, "mean_token_accuracy": 0.8660966157913208, "num_tokens": 530528265.0, "step": 13977 }, { "epoch": 1.7781452741381503, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.495396137237549, "learning_rate": 1e-05, "loss": 0.4715, "mean_token_accuracy": 0.8808200359344482, "num_tokens": 530567501.0, "step": 13978 }, { "epoch": 1.7782724844167408, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 6.963600158691406, "learning_rate": 1e-05, "loss": 0.53, "mean_token_accuracy": 0.8672207593917847, "num_tokens": 530610038.0, "step": 13979 }, { "epoch": 1.7783996946953313, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.453732490539551, "learning_rate": 1e-05, "loss": 0.5147, "mean_token_accuracy": 0.86619633436203, "num_tokens": 530649039.0, "step": 13980 }, { "epoch": 1.7785269049739219, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.639987468719482, "learning_rate": 1e-05, "loss": 0.482, "mean_token_accuracy": 0.87413489818573, "num_tokens": 530687142.0, "step": 13981 }, { "epoch": 1.7786541152525124, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.341386318206787, "learning_rate": 1e-05, "loss": 0.4607, "mean_token_accuracy": 0.8778924942016602, "num_tokens": 530724399.0, "step": 13982 }, { "epoch": 1.778781325531103, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.425759315490723, "learning_rate": 1e-05, "loss": 0.4912, "mean_token_accuracy": 0.8700681924819946, "num_tokens": 530764552.0, "step": 13983 }, { "epoch": 1.7789085358096934, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.190707683563232, "learning_rate": 1e-05, "loss": 0.5105, "mean_token_accuracy": 0.8648683428764343, "num_tokens": 530808587.0, "step": 13984 }, { "epoch": 1.779035746088284, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.391664028167725, "learning_rate": 1e-05, "loss": 0.5091, "mean_token_accuracy": 0.8658274412155151, "num_tokens": 530845887.0, "step": 13985 }, { "epoch": 1.7791629563668745, "ewc_loss": 0.10888671875, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 6.016719341278076, "learning_rate": 1e-05, "loss": 0.5221, "mean_token_accuracy": 0.8581244945526123, "num_tokens": 530887398.0, "step": 13986 }, { "epoch": 1.779290166645465, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.472462177276611, "learning_rate": 1e-05, "loss": 0.4641, "mean_token_accuracy": 0.879963755607605, "num_tokens": 530916377.0, "step": 13987 }, { "epoch": 1.7794173769240555, "ewc_loss": 0.10693359375, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.843163967132568, "learning_rate": 1e-05, "loss": 0.5255, "mean_token_accuracy": 0.859275221824646, "num_tokens": 530954651.0, "step": 13988 }, { "epoch": 1.779544587202646, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.646523475646973, "learning_rate": 1e-05, "loss": 0.5375, "mean_token_accuracy": 0.859502911567688, "num_tokens": 530999406.0, "step": 13989 }, { "epoch": 1.7796717974812366, "ewc_loss": 0.10498046875, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.682016372680664, "learning_rate": 1e-05, "loss": 0.4689, "mean_token_accuracy": 0.8747842311859131, "num_tokens": 531036925.0, "step": 13990 }, { "epoch": 1.7797990077598271, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.975667476654053, "learning_rate": 1e-05, "loss": 0.499, "mean_token_accuracy": 0.872241735458374, "num_tokens": 531076168.0, "step": 13991 }, { "epoch": 1.7799262180384177, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.639662742614746, "learning_rate": 1e-05, "loss": 0.4802, "mean_token_accuracy": 0.8720746636390686, "num_tokens": 531113217.0, "step": 13992 }, { "epoch": 1.780053428317008, "ewc_loss": 0.140625, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00011777877807617188, "grad_norm": 7.358729362487793, "learning_rate": 1e-05, "loss": 0.5197, "mean_token_accuracy": 0.8684678077697754, "num_tokens": 531155938.0, "step": 13993 }, { "epoch": 1.7801806385955985, "ewc_loss": 0.11376953125, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 5.883060932159424, "learning_rate": 1e-05, "loss": 0.4939, "mean_token_accuracy": 0.8719037771224976, "num_tokens": 531194556.0, "step": 13994 }, { "epoch": 1.780307848874189, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00012063980102539062, "grad_norm": 7.505349159240723, "learning_rate": 1e-05, "loss": 0.5557, "mean_token_accuracy": 0.8581938147544861, "num_tokens": 531230704.0, "step": 13995 }, { "epoch": 1.7804350591527796, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.318140029907227, "learning_rate": 1e-05, "loss": 0.4872, "mean_token_accuracy": 0.8723694682121277, "num_tokens": 531270672.0, "step": 13996 }, { "epoch": 1.78056226943137, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00011348724365234375, "grad_norm": 7.197526454925537, "learning_rate": 1e-05, "loss": 0.4682, "mean_token_accuracy": 0.8803418874740601, "num_tokens": 531302043.0, "step": 13997 }, { "epoch": 1.7806894797099606, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.532728672027588, "learning_rate": 1e-05, "loss": 0.4596, "mean_token_accuracy": 0.8840817809104919, "num_tokens": 531337014.0, "step": 13998 }, { "epoch": 1.780816689988551, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.6830902099609375, "learning_rate": 1e-05, "loss": 0.4846, "mean_token_accuracy": 0.8754100203514099, "num_tokens": 531371537.0, "step": 13999 }, { "epoch": 1.7809439002671414, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.562551975250244, "learning_rate": 1e-05, "loss": 0.4739, "mean_token_accuracy": 0.8795394897460938, "num_tokens": 531406171.0, "step": 14000 }, { "epoch": 1.781071110545732, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 6.223108291625977, "learning_rate": 1e-05, "loss": 0.5422, "mean_token_accuracy": 0.857027530670166, "num_tokens": 531445625.0, "step": 14001 }, { "epoch": 1.7811983208243225, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.643984317779541, "learning_rate": 1e-05, "loss": 0.4721, "mean_token_accuracy": 0.8799409866333008, "num_tokens": 531490748.0, "step": 14002 }, { "epoch": 1.781325531102913, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.850061416625977, "learning_rate": 1e-05, "loss": 0.4777, "mean_token_accuracy": 0.8697612881660461, "num_tokens": 531526142.0, "step": 14003 }, { "epoch": 1.7814527413815036, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.859175205230713, "learning_rate": 1e-05, "loss": 0.4978, "mean_token_accuracy": 0.8743108510971069, "num_tokens": 531568325.0, "step": 14004 }, { "epoch": 1.781579951660094, "ewc_loss": 0.10498046875, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.693297863006592, "learning_rate": 1e-05, "loss": 0.4922, "mean_token_accuracy": 0.8670593500137329, "num_tokens": 531608898.0, "step": 14005 }, { "epoch": 1.7817071619386846, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 0.00011110305786132812, "grad_norm": 7.164992332458496, "learning_rate": 1e-05, "loss": 0.5049, "mean_token_accuracy": 0.8710471391677856, "num_tokens": 531640860.0, "step": 14006 }, { "epoch": 1.7818343722172751, "ewc_loss": 0.10791015625, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.771664619445801, "learning_rate": 1e-05, "loss": 0.4797, "mean_token_accuracy": 0.8733980059623718, "num_tokens": 531680121.0, "step": 14007 }, { "epoch": 1.7819615824958657, "ewc_loss": 0.140625, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 0.000118255615234375, "grad_norm": 7.401505947113037, "learning_rate": 1e-05, "loss": 0.5728, "mean_token_accuracy": 0.8526183366775513, "num_tokens": 531713496.0, "step": 14008 }, { "epoch": 1.7820887927744562, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.069596767425537, "learning_rate": 1e-05, "loss": 0.5284, "mean_token_accuracy": 0.8620355129241943, "num_tokens": 531749989.0, "step": 14009 }, { "epoch": 1.7822160030530467, "ewc_loss": 0.140625, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 0.00011730194091796875, "grad_norm": 7.357231140136719, "learning_rate": 1e-05, "loss": 0.5207, "mean_token_accuracy": 0.8673897385597229, "num_tokens": 531788032.0, "step": 14010 }, { "epoch": 1.7823432133316373, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.324702262878418, "learning_rate": 1e-05, "loss": 0.5625, "mean_token_accuracy": 0.8463824391365051, "num_tokens": 531822178.0, "step": 14011 }, { "epoch": 1.7824704236102278, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 7.0762434005737305, "learning_rate": 1e-05, "loss": 0.5351, "mean_token_accuracy": 0.8622923493385315, "num_tokens": 531855713.0, "step": 14012 }, { "epoch": 1.7825976338888183, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.379065990447998, "learning_rate": 1e-05, "loss": 0.5332, "mean_token_accuracy": 0.8576452732086182, "num_tokens": 531898714.0, "step": 14013 }, { "epoch": 1.7827248441674088, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.758439540863037, "learning_rate": 1e-05, "loss": 0.5354, "mean_token_accuracy": 0.8651731014251709, "num_tokens": 531933968.0, "step": 14014 }, { "epoch": 1.7828520544459994, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.3082709312438965, "learning_rate": 1e-05, "loss": 0.507, "mean_token_accuracy": 0.8715660572052002, "num_tokens": 531980672.0, "step": 14015 }, { "epoch": 1.78297926472459, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.517053604125977, "learning_rate": 1e-05, "loss": 0.4979, "mean_token_accuracy": 0.871701180934906, "num_tokens": 532022461.0, "step": 14016 }, { "epoch": 1.7831064750031802, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.205146789550781, "learning_rate": 1e-05, "loss": 0.4407, "mean_token_accuracy": 0.8874208927154541, "num_tokens": 532059090.0, "step": 14017 }, { "epoch": 1.7832336852817707, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.417073726654053, "learning_rate": 1e-05, "loss": 0.5321, "mean_token_accuracy": 0.8604265451431274, "num_tokens": 532095926.0, "step": 14018 }, { "epoch": 1.7833608955603613, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.079226493835449, "learning_rate": 1e-05, "loss": 0.4896, "mean_token_accuracy": 0.8700740933418274, "num_tokens": 532131623.0, "step": 14019 }, { "epoch": 1.7834881058389518, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.414347171783447, "learning_rate": 1e-05, "loss": 0.5459, "mean_token_accuracy": 0.8537259101867676, "num_tokens": 532171594.0, "step": 14020 }, { "epoch": 1.7836153161175423, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.931082248687744, "learning_rate": 1e-05, "loss": 0.4504, "mean_token_accuracy": 0.8791542649269104, "num_tokens": 532209884.0, "step": 14021 }, { "epoch": 1.7837425263961328, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.527406215667725, "learning_rate": 1e-05, "loss": 0.5462, "mean_token_accuracy": 0.8597878217697144, "num_tokens": 532246182.0, "step": 14022 }, { "epoch": 1.7838697366747234, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.79051399230957, "learning_rate": 1e-05, "loss": 0.4884, "mean_token_accuracy": 0.8702242970466614, "num_tokens": 532281133.0, "step": 14023 }, { "epoch": 1.7839969469533137, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.742453098297119, "learning_rate": 1e-05, "loss": 0.506, "mean_token_accuracy": 0.8655490875244141, "num_tokens": 532318593.0, "step": 14024 }, { "epoch": 1.7841241572319042, "ewc_loss": 0.10498046875, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.686031341552734, "learning_rate": 1e-05, "loss": 0.5247, "mean_token_accuracy": 0.8595536947250366, "num_tokens": 532353473.0, "step": 14025 }, { "epoch": 1.7842513675104947, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 7.037057876586914, "learning_rate": 1e-05, "loss": 0.5601, "mean_token_accuracy": 0.8561252951622009, "num_tokens": 532395462.0, "step": 14026 }, { "epoch": 1.7843785777890853, "ewc_loss": 0.10595703125, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.710513114929199, "learning_rate": 1e-05, "loss": 0.4646, "mean_token_accuracy": 0.875267744064331, "num_tokens": 532434677.0, "step": 14027 }, { "epoch": 1.7845057880676758, "ewc_loss": 0.140625, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 0.00011777877807617188, "grad_norm": 7.31010627746582, "learning_rate": 1e-05, "loss": 0.5418, "mean_token_accuracy": 0.8659777641296387, "num_tokens": 532470585.0, "step": 14028 }, { "epoch": 1.7846329983462663, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 5.990374565124512, "learning_rate": 1e-05, "loss": 0.5156, "mean_token_accuracy": 0.8609410524368286, "num_tokens": 532502657.0, "step": 14029 }, { "epoch": 1.7847602086248568, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 0.0001163482666015625, "grad_norm": 7.293152332305908, "learning_rate": 1e-05, "loss": 0.5437, "mean_token_accuracy": 0.8607516288757324, "num_tokens": 532548372.0, "step": 14030 }, { "epoch": 1.7848874189034474, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.297001838684082, "learning_rate": 1e-05, "loss": 0.5362, "mean_token_accuracy": 0.8596974611282349, "num_tokens": 532584707.0, "step": 14031 }, { "epoch": 1.785014629182038, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.929906845092773, "learning_rate": 1e-05, "loss": 0.5076, "mean_token_accuracy": 0.8685011267662048, "num_tokens": 532623179.0, "step": 14032 }, { "epoch": 1.7851418394606284, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.434413909912109, "learning_rate": 1e-05, "loss": 0.5046, "mean_token_accuracy": 0.8718682527542114, "num_tokens": 532663054.0, "step": 14033 }, { "epoch": 1.785269049739219, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.432440280914307, "learning_rate": 1e-05, "loss": 0.5048, "mean_token_accuracy": 0.867027997970581, "num_tokens": 532696456.0, "step": 14034 }, { "epoch": 1.7853962600178095, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.5373125076293945, "learning_rate": 1e-05, "loss": 0.451, "mean_token_accuracy": 0.8825967311859131, "num_tokens": 532730726.0, "step": 14035 }, { "epoch": 1.7855234702964, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.959129810333252, "learning_rate": 1e-05, "loss": 0.4915, "mean_token_accuracy": 0.8662967681884766, "num_tokens": 532765561.0, "step": 14036 }, { "epoch": 1.7856506805749905, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.7688703536987305, "learning_rate": 1e-05, "loss": 0.478, "mean_token_accuracy": 0.8769299983978271, "num_tokens": 532803225.0, "step": 14037 }, { "epoch": 1.785777890853581, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.664852142333984, "learning_rate": 1e-05, "loss": 0.4932, "mean_token_accuracy": 0.8679934144020081, "num_tokens": 532843284.0, "step": 14038 }, { "epoch": 1.7859051011321716, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 0.00011205673217773438, "grad_norm": 7.132044315338135, "learning_rate": 1e-05, "loss": 0.5318, "mean_token_accuracy": 0.8652874231338501, "num_tokens": 532877637.0, "step": 14039 }, { "epoch": 1.7860323114107621, "ewc_loss": 0.10693359375, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.655131816864014, "learning_rate": 1e-05, "loss": 0.4798, "mean_token_accuracy": 0.8703356981277466, "num_tokens": 532914678.0, "step": 14040 }, { "epoch": 1.7861595216893527, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 0.0001220703125, "grad_norm": 7.504146575927734, "learning_rate": 1e-05, "loss": 0.5207, "mean_token_accuracy": 0.8678293228149414, "num_tokens": 532952309.0, "step": 14041 }, { "epoch": 1.786286731967943, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.011142253875732, "learning_rate": 1e-05, "loss": 0.4719, "mean_token_accuracy": 0.8753386735916138, "num_tokens": 532986440.0, "step": 14042 }, { "epoch": 1.7864139422465335, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.540114402770996, "learning_rate": 1e-05, "loss": 0.5359, "mean_token_accuracy": 0.8628280162811279, "num_tokens": 533020488.0, "step": 14043 }, { "epoch": 1.786541152525124, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.404358863830566, "learning_rate": 1e-05, "loss": 0.4827, "mean_token_accuracy": 0.8739029169082642, "num_tokens": 533052535.0, "step": 14044 }, { "epoch": 1.7866683628037145, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 0.0001125335693359375, "grad_norm": 7.198520183563232, "learning_rate": 1e-05, "loss": 0.4899, "mean_token_accuracy": 0.8777775764465332, "num_tokens": 533092924.0, "step": 14045 }, { "epoch": 1.786795573082305, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.492006301879883, "learning_rate": 1e-05, "loss": 0.528, "mean_token_accuracy": 0.8654196262359619, "num_tokens": 533130497.0, "step": 14046 }, { "epoch": 1.7869227833608956, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.8298258781433105, "learning_rate": 1e-05, "loss": 0.5656, "mean_token_accuracy": 0.8487949967384338, "num_tokens": 533170012.0, "step": 14047 }, { "epoch": 1.787049993639486, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.395122528076172, "learning_rate": 1e-05, "loss": 0.508, "mean_token_accuracy": 0.8675215244293213, "num_tokens": 533207433.0, "step": 14048 }, { "epoch": 1.7871772039180764, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.555670738220215, "learning_rate": 1e-05, "loss": 0.4603, "mean_token_accuracy": 0.8839771747589111, "num_tokens": 533248809.0, "step": 14049 }, { "epoch": 1.787304414196667, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.261752605438232, "learning_rate": 1e-05, "loss": 0.47, "mean_token_accuracy": 0.8774756789207458, "num_tokens": 533283381.0, "step": 14050 }, { "epoch": 1.7874316244752575, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.428366184234619, "learning_rate": 1e-05, "loss": 0.4679, "mean_token_accuracy": 0.8801383972167969, "num_tokens": 533320951.0, "step": 14051 }, { "epoch": 1.787558834753848, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.102770805358887, "learning_rate": 1e-05, "loss": 0.487, "mean_token_accuracy": 0.8723986148834229, "num_tokens": 533361080.0, "step": 14052 }, { "epoch": 1.7876860450324386, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.444705963134766, "learning_rate": 1e-05, "loss": 0.4618, "mean_token_accuracy": 0.8820101022720337, "num_tokens": 533396448.0, "step": 14053 }, { "epoch": 1.787813255311029, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.941468715667725, "learning_rate": 1e-05, "loss": 0.4882, "mean_token_accuracy": 0.8701412677764893, "num_tokens": 533434924.0, "step": 14054 }, { "epoch": 1.7879404655896196, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.562976360321045, "learning_rate": 1e-05, "loss": 0.4905, "mean_token_accuracy": 0.8723007440567017, "num_tokens": 533471248.0, "step": 14055 }, { "epoch": 1.7880676758682101, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.776337623596191, "learning_rate": 1e-05, "loss": 0.466, "mean_token_accuracy": 0.8787938356399536, "num_tokens": 533512521.0, "step": 14056 }, { "epoch": 1.7881948861468007, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.805587291717529, "learning_rate": 1e-05, "loss": 0.4817, "mean_token_accuracy": 0.8761717677116394, "num_tokens": 533550385.0, "step": 14057 }, { "epoch": 1.7883220964253912, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.681870937347412, "learning_rate": 1e-05, "loss": 0.478, "mean_token_accuracy": 0.8704138994216919, "num_tokens": 533580048.0, "step": 14058 }, { "epoch": 1.7884493067039817, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 0.00011157989501953125, "grad_norm": 7.130868434906006, "learning_rate": 1e-05, "loss": 0.4823, "mean_token_accuracy": 0.8809620141983032, "num_tokens": 533615115.0, "step": 14059 }, { "epoch": 1.7885765169825723, "ewc_loss": 0.10791015625, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.750942707061768, "learning_rate": 1e-05, "loss": 0.4701, "mean_token_accuracy": 0.8719260692596436, "num_tokens": 533653774.0, "step": 14060 }, { "epoch": 1.7887037272611628, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 0.00011920928955078125, "grad_norm": 7.38611364364624, "learning_rate": 1e-05, "loss": 0.4902, "mean_token_accuracy": 0.8813774585723877, "num_tokens": 533692628.0, "step": 14061 }, { "epoch": 1.7888309375397533, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.068597316741943, "learning_rate": 1e-05, "loss": 0.4486, "mean_token_accuracy": 0.8884952664375305, "num_tokens": 533734935.0, "step": 14062 }, { "epoch": 1.7889581478183438, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 0.0001163482666015625, "grad_norm": 7.304275989532471, "learning_rate": 1e-05, "loss": 0.5241, "mean_token_accuracy": 0.8645588159561157, "num_tokens": 533772541.0, "step": 14063 }, { "epoch": 1.7890853580969344, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.388991832733154, "learning_rate": 1e-05, "loss": 0.4786, "mean_token_accuracy": 0.8756067752838135, "num_tokens": 533816036.0, "step": 14064 }, { "epoch": 1.789212568375525, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.876842498779297, "learning_rate": 1e-05, "loss": 0.5117, "mean_token_accuracy": 0.8696466684341431, "num_tokens": 533848174.0, "step": 14065 }, { "epoch": 1.7893397786541152, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.503574371337891, "learning_rate": 1e-05, "loss": 0.5433, "mean_token_accuracy": 0.8582596778869629, "num_tokens": 533890011.0, "step": 14066 }, { "epoch": 1.7894669889327057, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.365510940551758, "learning_rate": 1e-05, "loss": 0.521, "mean_token_accuracy": 0.8597009181976318, "num_tokens": 533927650.0, "step": 14067 }, { "epoch": 1.7895941992112963, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.610496520996094, "learning_rate": 1e-05, "loss": 0.5205, "mean_token_accuracy": 0.8632844686508179, "num_tokens": 533959117.0, "step": 14068 }, { "epoch": 1.7897214094898868, "ewc_loss": 0.10693359375, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.909084796905518, "learning_rate": 1e-05, "loss": 0.5266, "mean_token_accuracy": 0.8582482933998108, "num_tokens": 533995396.0, "step": 14069 }, { "epoch": 1.7898486197684773, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.846432685852051, "learning_rate": 1e-05, "loss": 0.4704, "mean_token_accuracy": 0.8793460130691528, "num_tokens": 534032498.0, "step": 14070 }, { "epoch": 1.7899758300470678, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.6521477699279785, "learning_rate": 1e-05, "loss": 0.5251, "mean_token_accuracy": 0.8607052564620972, "num_tokens": 534071790.0, "step": 14071 }, { "epoch": 1.7901030403256584, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 0.00011396408081054688, "grad_norm": 7.22512149810791, "learning_rate": 1e-05, "loss": 0.5343, "mean_token_accuracy": 0.8624362945556641, "num_tokens": 534112778.0, "step": 14072 }, { "epoch": 1.7902302506042487, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.713668346405029, "learning_rate": 1e-05, "loss": 0.4659, "mean_token_accuracy": 0.8759269118309021, "num_tokens": 534150815.0, "step": 14073 }, { "epoch": 1.7903574608828392, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 0.0001220703125, "grad_norm": 7.553882598876953, "learning_rate": 1e-05, "loss": 0.5369, "mean_token_accuracy": 0.866135835647583, "num_tokens": 534186148.0, "step": 14074 }, { "epoch": 1.7904846711614297, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.124167442321777, "learning_rate": 1e-05, "loss": 0.5117, "mean_token_accuracy": 0.8642296195030212, "num_tokens": 534219496.0, "step": 14075 }, { "epoch": 1.7906118814400203, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 0.0001201629638671875, "grad_norm": 7.477260589599609, "learning_rate": 1e-05, "loss": 0.5072, "mean_token_accuracy": 0.8760215640068054, "num_tokens": 534263093.0, "step": 14076 }, { "epoch": 1.7907390917186108, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.447449207305908, "learning_rate": 1e-05, "loss": 0.4695, "mean_token_accuracy": 0.8810862302780151, "num_tokens": 534298515.0, "step": 14077 }, { "epoch": 1.7908663019972013, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 0.00011157989501953125, "grad_norm": 7.123748302459717, "learning_rate": 1e-05, "loss": 0.5007, "mean_token_accuracy": 0.8733760118484497, "num_tokens": 534337210.0, "step": 14078 }, { "epoch": 1.7909935122757918, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.4881062507629395, "learning_rate": 1e-05, "loss": 0.5508, "mean_token_accuracy": 0.8603433966636658, "num_tokens": 534373440.0, "step": 14079 }, { "epoch": 1.7911207225543824, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.7526092529296875, "learning_rate": 1e-05, "loss": 0.4336, "mean_token_accuracy": 0.8929649591445923, "num_tokens": 534405560.0, "step": 14080 }, { "epoch": 1.791247932832973, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.384099006652832, "learning_rate": 1e-05, "loss": 0.5031, "mean_token_accuracy": 0.8714176416397095, "num_tokens": 534442824.0, "step": 14081 }, { "epoch": 1.7913751431115634, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.500860691070557, "learning_rate": 1e-05, "loss": 0.4982, "mean_token_accuracy": 0.8712247610092163, "num_tokens": 534482569.0, "step": 14082 }, { "epoch": 1.791502353390154, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.242004871368408, "learning_rate": 1e-05, "loss": 0.5171, "mean_token_accuracy": 0.8631855845451355, "num_tokens": 534518481.0, "step": 14083 }, { "epoch": 1.7916295636687445, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.411289215087891, "learning_rate": 1e-05, "loss": 0.5579, "mean_token_accuracy": 0.8473285436630249, "num_tokens": 534553273.0, "step": 14084 }, { "epoch": 1.791756773947335, "ewc_loss": 0.11181640625, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.078204154968262, "learning_rate": 1e-05, "loss": 0.497, "mean_token_accuracy": 0.8648864030838013, "num_tokens": 534587322.0, "step": 14085 }, { "epoch": 1.7918839842259255, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.441062927246094, "learning_rate": 1e-05, "loss": 0.462, "mean_token_accuracy": 0.8812383413314819, "num_tokens": 534622999.0, "step": 14086 }, { "epoch": 1.792011194504516, "ewc_loss": 0.10693359375, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.904778957366943, "learning_rate": 1e-05, "loss": 0.5208, "mean_token_accuracy": 0.8611901998519897, "num_tokens": 534655953.0, "step": 14087 }, { "epoch": 1.7921384047831066, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.5830512046813965, "learning_rate": 1e-05, "loss": 0.5039, "mean_token_accuracy": 0.8642740249633789, "num_tokens": 534699495.0, "step": 14088 }, { "epoch": 1.7922656150616971, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.728967189788818, "learning_rate": 1e-05, "loss": 0.4632, "mean_token_accuracy": 0.8773858547210693, "num_tokens": 534737650.0, "step": 14089 }, { "epoch": 1.7923928253402877, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.866134166717529, "learning_rate": 1e-05, "loss": 0.5, "mean_token_accuracy": 0.8740373849868774, "num_tokens": 534771169.0, "step": 14090 }, { "epoch": 1.792520035618878, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.6341681480407715, "learning_rate": 1e-05, "loss": 0.4501, "mean_token_accuracy": 0.8787192106246948, "num_tokens": 534802968.0, "step": 14091 }, { "epoch": 1.7926472458974685, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 0.00011587142944335938, "grad_norm": 7.24970817565918, "learning_rate": 1e-05, "loss": 0.5026, "mean_token_accuracy": 0.8733949065208435, "num_tokens": 534846975.0, "step": 14092 }, { "epoch": 1.792774456176059, "ewc_loss": 0.109375, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.782205581665039, "learning_rate": 1e-05, "loss": 0.477, "mean_token_accuracy": 0.8726038336753845, "num_tokens": 534880969.0, "step": 14093 }, { "epoch": 1.7929016664546495, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 0.00012159347534179688, "grad_norm": 7.512179374694824, "learning_rate": 1e-05, "loss": 0.529, "mean_token_accuracy": 0.8645358085632324, "num_tokens": 534918423.0, "step": 14094 }, { "epoch": 1.79302887673324, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.192817211151123, "learning_rate": 1e-05, "loss": 0.4678, "mean_token_accuracy": 0.8763402104377747, "num_tokens": 534952733.0, "step": 14095 }, { "epoch": 1.7931560870118306, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 0.0001163482666015625, "grad_norm": 7.335895538330078, "learning_rate": 1e-05, "loss": 0.4954, "mean_token_accuracy": 0.8752374053001404, "num_tokens": 534981868.0, "step": 14096 }, { "epoch": 1.793283297290421, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.502678394317627, "learning_rate": 1e-05, "loss": 0.5238, "mean_token_accuracy": 0.8678610324859619, "num_tokens": 535020940.0, "step": 14097 }, { "epoch": 1.7934105075690114, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.844568729400635, "learning_rate": 1e-05, "loss": 0.4775, "mean_token_accuracy": 0.8744961619377136, "num_tokens": 535053073.0, "step": 14098 }, { "epoch": 1.793537717847602, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.555118560791016, "learning_rate": 1e-05, "loss": 0.5082, "mean_token_accuracy": 0.8680493235588074, "num_tokens": 535097731.0, "step": 14099 }, { "epoch": 1.7936649281261925, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.313578128814697, "learning_rate": 1e-05, "loss": 0.5161, "mean_token_accuracy": 0.8628672361373901, "num_tokens": 535145571.0, "step": 14100 }, { "epoch": 1.793792138404783, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.638735294342041, "learning_rate": 1e-05, "loss": 0.5118, "mean_token_accuracy": 0.8665964603424072, "num_tokens": 535184226.0, "step": 14101 }, { "epoch": 1.7939193486833735, "ewc_loss": 0.10791015625, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.926120281219482, "learning_rate": 1e-05, "loss": 0.5534, "mean_token_accuracy": 0.8495131731033325, "num_tokens": 535221990.0, "step": 14102 }, { "epoch": 1.794046558961964, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.838809490203857, "learning_rate": 1e-05, "loss": 0.5228, "mean_token_accuracy": 0.8631162047386169, "num_tokens": 535261588.0, "step": 14103 }, { "epoch": 1.7941737692405546, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.705915451049805, "learning_rate": 1e-05, "loss": 0.5412, "mean_token_accuracy": 0.8515486717224121, "num_tokens": 535299023.0, "step": 14104 }, { "epoch": 1.7943009795191451, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 0.00011157989501953125, "grad_norm": 7.14921236038208, "learning_rate": 1e-05, "loss": 0.5225, "mean_token_accuracy": 0.8645749688148499, "num_tokens": 535332174.0, "step": 14105 }, { "epoch": 1.7944281897977357, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.739795207977295, "learning_rate": 1e-05, "loss": 0.4739, "mean_token_accuracy": 0.8742871880531311, "num_tokens": 535367242.0, "step": 14106 }, { "epoch": 1.7945554000763262, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 0.00011873245239257812, "grad_norm": 7.436201095581055, "learning_rate": 1e-05, "loss": 0.5143, "mean_token_accuracy": 0.8707119226455688, "num_tokens": 535401962.0, "step": 14107 }, { "epoch": 1.7946826103549167, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.039036750793457, "learning_rate": 1e-05, "loss": 0.5289, "mean_token_accuracy": 0.8596152067184448, "num_tokens": 535441481.0, "step": 14108 }, { "epoch": 1.7948098206335072, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 0.00011968612670898438, "grad_norm": 7.437800407409668, "learning_rate": 1e-05, "loss": 0.4964, "mean_token_accuracy": 0.8763160705566406, "num_tokens": 535473619.0, "step": 14109 }, { "epoch": 1.7949370309120978, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.341948986053467, "learning_rate": 1e-05, "loss": 0.4665, "mean_token_accuracy": 0.8829479217529297, "num_tokens": 535515340.0, "step": 14110 }, { "epoch": 1.7950642411906883, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 0.00011157989501953125, "grad_norm": 7.160099029541016, "learning_rate": 1e-05, "loss": 0.5602, "mean_token_accuracy": 0.854478657245636, "num_tokens": 535555090.0, "step": 14111 }, { "epoch": 1.7951914514692788, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.425323963165283, "learning_rate": 1e-05, "loss": 0.5156, "mean_token_accuracy": 0.8698476552963257, "num_tokens": 535595484.0, "step": 14112 }, { "epoch": 1.7953186617478694, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.809148788452148, "learning_rate": 1e-05, "loss": 0.5159, "mean_token_accuracy": 0.8668070435523987, "num_tokens": 535634170.0, "step": 14113 }, { "epoch": 1.7954458720264599, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.373335361480713, "learning_rate": 1e-05, "loss": 0.4866, "mean_token_accuracy": 0.8723224401473999, "num_tokens": 535671771.0, "step": 14114 }, { "epoch": 1.7955730823050502, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.530010223388672, "learning_rate": 1e-05, "loss": 0.4753, "mean_token_accuracy": 0.8772655129432678, "num_tokens": 535710280.0, "step": 14115 }, { "epoch": 1.7957002925836407, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.269746780395508, "learning_rate": 1e-05, "loss": 0.5117, "mean_token_accuracy": 0.865311324596405, "num_tokens": 535749509.0, "step": 14116 }, { "epoch": 1.7958275028622313, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.373878002166748, "learning_rate": 1e-05, "loss": 0.4744, "mean_token_accuracy": 0.8762238025665283, "num_tokens": 535794184.0, "step": 14117 }, { "epoch": 1.7959547131408218, "ewc_loss": 0.11376953125, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.1514081954956055, "learning_rate": 1e-05, "loss": 0.5464, "mean_token_accuracy": 0.8501302003860474, "num_tokens": 535828818.0, "step": 14118 }, { "epoch": 1.7960819234194123, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.3383378982543945, "learning_rate": 1e-05, "loss": 0.5189, "mean_token_accuracy": 0.8631325960159302, "num_tokens": 535869461.0, "step": 14119 }, { "epoch": 1.7962091336980028, "ewc_loss": 0.109375, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.990093231201172, "learning_rate": 1e-05, "loss": 0.4782, "mean_token_accuracy": 0.8715918064117432, "num_tokens": 535907287.0, "step": 14120 }, { "epoch": 1.7963363439765934, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.413898944854736, "learning_rate": 1e-05, "loss": 0.4638, "mean_token_accuracy": 0.8786344528198242, "num_tokens": 535949063.0, "step": 14121 }, { "epoch": 1.7964635542551837, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.82835054397583, "learning_rate": 1e-05, "loss": 0.4766, "mean_token_accuracy": 0.8713647127151489, "num_tokens": 535987202.0, "step": 14122 }, { "epoch": 1.7965907645337742, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.595701694488525, "learning_rate": 1e-05, "loss": 0.4812, "mean_token_accuracy": 0.8791061639785767, "num_tokens": 536021790.0, "step": 14123 }, { "epoch": 1.7967179748123647, "ewc_loss": 0.10498046875, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.679178237915039, "learning_rate": 1e-05, "loss": 0.583, "mean_token_accuracy": 0.8406621813774109, "num_tokens": 536069932.0, "step": 14124 }, { "epoch": 1.7968451850909553, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.9023823738098145, "learning_rate": 1e-05, "loss": 0.4736, "mean_token_accuracy": 0.877282977104187, "num_tokens": 536110428.0, "step": 14125 }, { "epoch": 1.7969723953695458, "ewc_loss": 0.10498046875, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.606557846069336, "learning_rate": 1e-05, "loss": 0.5052, "mean_token_accuracy": 0.8665341734886169, "num_tokens": 536147711.0, "step": 14126 }, { "epoch": 1.7970996056481363, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 0.0001163482666015625, "grad_norm": 7.307696342468262, "learning_rate": 1e-05, "loss": 0.5007, "mean_token_accuracy": 0.8755270838737488, "num_tokens": 536184422.0, "step": 14127 }, { "epoch": 1.7972268159267268, "ewc_loss": 0.10986328125, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.778512954711914, "learning_rate": 1e-05, "loss": 0.4429, "mean_token_accuracy": 0.8816883563995361, "num_tokens": 536224802.0, "step": 14128 }, { "epoch": 1.7973540262053174, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 0.0001239776611328125, "grad_norm": 7.556324005126953, "learning_rate": 1e-05, "loss": 0.5377, "mean_token_accuracy": 0.866023600101471, "num_tokens": 536267596.0, "step": 14129 }, { "epoch": 1.797481236483908, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.243893623352051, "learning_rate": 1e-05, "loss": 0.4922, "mean_token_accuracy": 0.8730190396308899, "num_tokens": 536306130.0, "step": 14130 }, { "epoch": 1.7976084467624984, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 0.00011682510375976562, "grad_norm": 7.323394298553467, "learning_rate": 1e-05, "loss": 0.5304, "mean_token_accuracy": 0.867561936378479, "num_tokens": 536345451.0, "step": 14131 }, { "epoch": 1.797735657041089, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.53281307220459, "learning_rate": 1e-05, "loss": 0.5105, "mean_token_accuracy": 0.8630922436714172, "num_tokens": 536377953.0, "step": 14132 }, { "epoch": 1.7978628673196795, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.814650058746338, "learning_rate": 1e-05, "loss": 0.4696, "mean_token_accuracy": 0.8789415955543518, "num_tokens": 536410879.0, "step": 14133 }, { "epoch": 1.79799007759827, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.564652442932129, "learning_rate": 1e-05, "loss": 0.547, "mean_token_accuracy": 0.8564990162849426, "num_tokens": 536448351.0, "step": 14134 }, { "epoch": 1.7981172878768605, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.31316614151001, "learning_rate": 1e-05, "loss": 0.5063, "mean_token_accuracy": 0.8649547100067139, "num_tokens": 536488968.0, "step": 14135 }, { "epoch": 1.798244498155451, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.622358798980713, "learning_rate": 1e-05, "loss": 0.4867, "mean_token_accuracy": 0.8779931664466858, "num_tokens": 536529189.0, "step": 14136 }, { "epoch": 1.7983717084340416, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.935693740844727, "learning_rate": 1e-05, "loss": 0.4376, "mean_token_accuracy": 0.8860536217689514, "num_tokens": 536567778.0, "step": 14137 }, { "epoch": 1.7984989187126321, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.79704475402832, "learning_rate": 1e-05, "loss": 0.5091, "mean_token_accuracy": 0.8688446283340454, "num_tokens": 536609193.0, "step": 14138 }, { "epoch": 1.7986261289912227, "ewc_loss": 0.10498046875, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.7476725578308105, "learning_rate": 1e-05, "loss": 0.5335, "mean_token_accuracy": 0.8522482514381409, "num_tokens": 536646122.0, "step": 14139 }, { "epoch": 1.798753339269813, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 7.070370197296143, "learning_rate": 1e-05, "loss": 0.5142, "mean_token_accuracy": 0.8673481345176697, "num_tokens": 536682146.0, "step": 14140 }, { "epoch": 1.7988805495484035, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.750649929046631, "learning_rate": 1e-05, "loss": 0.5122, "mean_token_accuracy": 0.8596254587173462, "num_tokens": 536715449.0, "step": 14141 }, { "epoch": 1.799007759826994, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 0.0001163482666015625, "grad_norm": 7.320939540863037, "learning_rate": 1e-05, "loss": 0.5351, "mean_token_accuracy": 0.8650093078613281, "num_tokens": 536752430.0, "step": 14142 }, { "epoch": 1.7991349701055845, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.0044074058532715, "learning_rate": 1e-05, "loss": 0.5136, "mean_token_accuracy": 0.8564797043800354, "num_tokens": 536785577.0, "step": 14143 }, { "epoch": 1.799262180384175, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 0.00011587142944335938, "grad_norm": 7.309545040130615, "learning_rate": 1e-05, "loss": 0.4932, "mean_token_accuracy": 0.8809553384780884, "num_tokens": 536824074.0, "step": 14144 }, { "epoch": 1.7993893906627656, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.207816123962402, "learning_rate": 1e-05, "loss": 0.4512, "mean_token_accuracy": 0.8853156566619873, "num_tokens": 536865819.0, "step": 14145 }, { "epoch": 1.799516600941356, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 0.00011157989501953125, "grad_norm": 7.121119022369385, "learning_rate": 1e-05, "loss": 0.5188, "mean_token_accuracy": 0.8710426092147827, "num_tokens": 536905538.0, "step": 14146 }, { "epoch": 1.7996438112199464, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.301765441894531, "learning_rate": 1e-05, "loss": 0.487, "mean_token_accuracy": 0.878635048866272, "num_tokens": 536944895.0, "step": 14147 }, { "epoch": 1.799771021498537, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.867774486541748, "learning_rate": 1e-05, "loss": 0.5247, "mean_token_accuracy": 0.8653445243835449, "num_tokens": 536981094.0, "step": 14148 }, { "epoch": 1.7998982317771275, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.277656078338623, "learning_rate": 1e-05, "loss": 0.5477, "mean_token_accuracy": 0.8557510375976562, "num_tokens": 537020242.0, "step": 14149 }, { "epoch": 1.800025442055718, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.624682903289795, "learning_rate": 1e-05, "loss": 0.5335, "mean_token_accuracy": 0.8586287498474121, "num_tokens": 537062475.0, "step": 14150 }, { "epoch": 1.8001526523343085, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.202523708343506, "learning_rate": 1e-05, "loss": 0.5079, "mean_token_accuracy": 0.8656128644943237, "num_tokens": 537100969.0, "step": 14151 }, { "epoch": 1.800279862612899, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.4814772605896, "learning_rate": 1e-05, "loss": 0.4976, "mean_token_accuracy": 0.8689011931419373, "num_tokens": 537141828.0, "step": 14152 }, { "epoch": 1.8004070728914896, "ewc_loss": 0.11474609375, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.093362331390381, "learning_rate": 1e-05, "loss": 0.4749, "mean_token_accuracy": 0.87530517578125, "num_tokens": 537181324.0, "step": 14153 }, { "epoch": 1.8005342831700801, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.443410873413086, "learning_rate": 1e-05, "loss": 0.4814, "mean_token_accuracy": 0.8742595911026001, "num_tokens": 537224593.0, "step": 14154 }, { "epoch": 1.8006614934486707, "ewc_loss": 0.10986328125, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.979639053344727, "learning_rate": 1e-05, "loss": 0.5097, "mean_token_accuracy": 0.8637410998344421, "num_tokens": 537262835.0, "step": 14155 }, { "epoch": 1.8007887037272612, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.488171100616455, "learning_rate": 1e-05, "loss": 0.4923, "mean_token_accuracy": 0.8662003874778748, "num_tokens": 537299722.0, "step": 14156 }, { "epoch": 1.8009159140058517, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.862623691558838, "learning_rate": 1e-05, "loss": 0.4769, "mean_token_accuracy": 0.8730767965316772, "num_tokens": 537345445.0, "step": 14157 }, { "epoch": 1.8010431242844422, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.623968601226807, "learning_rate": 1e-05, "loss": 0.4908, "mean_token_accuracy": 0.8728833198547363, "num_tokens": 537382327.0, "step": 14158 }, { "epoch": 1.8011703345630328, "ewc_loss": 0.10693359375, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.756891250610352, "learning_rate": 1e-05, "loss": 0.4908, "mean_token_accuracy": 0.8689990043640137, "num_tokens": 537419013.0, "step": 14159 }, { "epoch": 1.8012975448416233, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.846460342407227, "learning_rate": 1e-05, "loss": 0.5092, "mean_token_accuracy": 0.8675127029418945, "num_tokens": 537459841.0, "step": 14160 }, { "epoch": 1.8014247551202138, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.714104652404785, "learning_rate": 1e-05, "loss": 0.4437, "mean_token_accuracy": 0.8807169795036316, "num_tokens": 537494124.0, "step": 14161 }, { "epoch": 1.8015519653988044, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 0.00011014938354492188, "grad_norm": 7.096462249755859, "learning_rate": 1e-05, "loss": 0.4915, "mean_token_accuracy": 0.874798059463501, "num_tokens": 537532798.0, "step": 14162 }, { "epoch": 1.8016791756773949, "ewc_loss": 0.11083984375, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 5.833984851837158, "learning_rate": 1e-05, "loss": 0.4929, "mean_token_accuracy": 0.869784951210022, "num_tokens": 537568548.0, "step": 14163 }, { "epoch": 1.8018063859559852, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 0.00011396408081054688, "grad_norm": 7.200321674346924, "learning_rate": 1e-05, "loss": 0.4739, "mean_token_accuracy": 0.8848297595977783, "num_tokens": 537604413.0, "step": 14164 }, { "epoch": 1.8019335962345757, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.082299709320068, "learning_rate": 1e-05, "loss": 0.499, "mean_token_accuracy": 0.8709728717803955, "num_tokens": 537640588.0, "step": 14165 }, { "epoch": 1.8020608065131662, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 7.044229984283447, "learning_rate": 1e-05, "loss": 0.5358, "mean_token_accuracy": 0.8626935482025146, "num_tokens": 537682664.0, "step": 14166 }, { "epoch": 1.8021880167917568, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.283174514770508, "learning_rate": 1e-05, "loss": 0.5376, "mean_token_accuracy": 0.8528116345405579, "num_tokens": 537716020.0, "step": 14167 }, { "epoch": 1.8023152270703473, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.6449384689331055, "learning_rate": 1e-05, "loss": 0.4864, "mean_token_accuracy": 0.8772117495536804, "num_tokens": 537756895.0, "step": 14168 }, { "epoch": 1.8024424373489378, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.407036781311035, "learning_rate": 1e-05, "loss": 0.5177, "mean_token_accuracy": 0.8655466437339783, "num_tokens": 537794130.0, "step": 14169 }, { "epoch": 1.8025696476275284, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.175848007202148, "learning_rate": 1e-05, "loss": 0.4485, "mean_token_accuracy": 0.8822106122970581, "num_tokens": 537822865.0, "step": 14170 }, { "epoch": 1.8026968579061187, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.581227779388428, "learning_rate": 1e-05, "loss": 0.4826, "mean_token_accuracy": 0.8747433423995972, "num_tokens": 537854594.0, "step": 14171 }, { "epoch": 1.8028240681847092, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.761058330535889, "learning_rate": 1e-05, "loss": 0.4874, "mean_token_accuracy": 0.8704398274421692, "num_tokens": 537889667.0, "step": 14172 }, { "epoch": 1.8029512784632997, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.918200969696045, "learning_rate": 1e-05, "loss": 0.5439, "mean_token_accuracy": 0.8569884896278381, "num_tokens": 537924322.0, "step": 14173 }, { "epoch": 1.8030784887418903, "ewc_loss": 0.10400390625, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.553057670593262, "learning_rate": 1e-05, "loss": 0.4573, "mean_token_accuracy": 0.8791199922561646, "num_tokens": 537962196.0, "step": 14174 }, { "epoch": 1.8032056990204808, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 0.00011873245239257812, "grad_norm": 7.3821282386779785, "learning_rate": 1e-05, "loss": 0.5037, "mean_token_accuracy": 0.8777047395706177, "num_tokens": 537999916.0, "step": 14175 }, { "epoch": 1.8033329092990713, "ewc_loss": 0.10986328125, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.726696014404297, "learning_rate": 1e-05, "loss": 0.454, "mean_token_accuracy": 0.8797567486763, "num_tokens": 538035823.0, "step": 14176 }, { "epoch": 1.8034601195776618, "ewc_loss": 0.150390625, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 0.0001277923583984375, "grad_norm": 7.711388111114502, "learning_rate": 1e-05, "loss": 0.5618, "mean_token_accuracy": 0.8598227500915527, "num_tokens": 538080664.0, "step": 14177 }, { "epoch": 1.8035873298562524, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.276473045349121, "learning_rate": 1e-05, "loss": 0.5143, "mean_token_accuracy": 0.8634424805641174, "num_tokens": 538123178.0, "step": 14178 }, { "epoch": 1.803714540134843, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 0.00012159347534179688, "grad_norm": 7.520144939422607, "learning_rate": 1e-05, "loss": 0.5056, "mean_token_accuracy": 0.8738092184066772, "num_tokens": 538158757.0, "step": 14179 }, { "epoch": 1.8038417504134334, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 6.582571506500244, "learning_rate": 1e-05, "loss": 0.5182, "mean_token_accuracy": 0.8664355278015137, "num_tokens": 538199019.0, "step": 14180 }, { "epoch": 1.803968960692024, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 7.095087051391602, "learning_rate": 1e-05, "loss": 0.5083, "mean_token_accuracy": 0.8700503706932068, "num_tokens": 538239839.0, "step": 14181 }, { "epoch": 1.8040961709706145, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.539381504058838, "learning_rate": 1e-05, "loss": 0.4986, "mean_token_accuracy": 0.8725436925888062, "num_tokens": 538287677.0, "step": 14182 }, { "epoch": 1.804223381249205, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.729211330413818, "learning_rate": 1e-05, "loss": 0.4899, "mean_token_accuracy": 0.8702072501182556, "num_tokens": 538322304.0, "step": 14183 }, { "epoch": 1.8043505915277955, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.398153305053711, "learning_rate": 1e-05, "loss": 0.4775, "mean_token_accuracy": 0.8746203184127808, "num_tokens": 538364087.0, "step": 14184 }, { "epoch": 1.804477801806386, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.507568836212158, "learning_rate": 1e-05, "loss": 0.5047, "mean_token_accuracy": 0.8686055541038513, "num_tokens": 538401507.0, "step": 14185 }, { "epoch": 1.8046050120849766, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.243355751037598, "learning_rate": 1e-05, "loss": 0.4768, "mean_token_accuracy": 0.8739398717880249, "num_tokens": 538440666.0, "step": 14186 }, { "epoch": 1.8047322223635671, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.407754421234131, "learning_rate": 1e-05, "loss": 0.5287, "mean_token_accuracy": 0.8627351522445679, "num_tokens": 538478915.0, "step": 14187 }, { "epoch": 1.8048594326421576, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 6.083024978637695, "learning_rate": 1e-05, "loss": 0.4794, "mean_token_accuracy": 0.8729896545410156, "num_tokens": 538517770.0, "step": 14188 }, { "epoch": 1.804986642920748, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.430295944213867, "learning_rate": 1e-05, "loss": 0.5301, "mean_token_accuracy": 0.8584818243980408, "num_tokens": 538555812.0, "step": 14189 }, { "epoch": 1.8051138531993385, "ewc_loss": 0.10791015625, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.90608549118042, "learning_rate": 1e-05, "loss": 0.4331, "mean_token_accuracy": 0.8862021565437317, "num_tokens": 538596023.0, "step": 14190 }, { "epoch": 1.805241063477929, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.5741753578186035, "learning_rate": 1e-05, "loss": 0.4519, "mean_token_accuracy": 0.8838251829147339, "num_tokens": 538630840.0, "step": 14191 }, { "epoch": 1.8053682737565195, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.749894618988037, "learning_rate": 1e-05, "loss": 0.5242, "mean_token_accuracy": 0.8609588146209717, "num_tokens": 538668515.0, "step": 14192 }, { "epoch": 1.80549548403511, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.838712692260742, "learning_rate": 1e-05, "loss": 0.5304, "mean_token_accuracy": 0.8611146807670593, "num_tokens": 538703996.0, "step": 14193 }, { "epoch": 1.8056226943137006, "ewc_loss": 0.10400390625, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.6602959632873535, "learning_rate": 1e-05, "loss": 0.4732, "mean_token_accuracy": 0.871300458908081, "num_tokens": 538742819.0, "step": 14194 }, { "epoch": 1.805749904592291, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 0.0001125335693359375, "grad_norm": 7.187918186187744, "learning_rate": 1e-05, "loss": 0.4796, "mean_token_accuracy": 0.8799353241920471, "num_tokens": 538776218.0, "step": 14195 }, { "epoch": 1.8058771148708814, "ewc_loss": 0.10888671875, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.7516560554504395, "learning_rate": 1e-05, "loss": 0.4937, "mean_token_accuracy": 0.8697987794876099, "num_tokens": 538813377.0, "step": 14196 }, { "epoch": 1.806004325149472, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 0.0001201629638671875, "grad_norm": 7.479597568511963, "learning_rate": 1e-05, "loss": 0.5315, "mean_token_accuracy": 0.8673434257507324, "num_tokens": 538856755.0, "step": 14197 }, { "epoch": 1.8061315354280625, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.133799076080322, "learning_rate": 1e-05, "loss": 0.4936, "mean_token_accuracy": 0.8685388565063477, "num_tokens": 538898175.0, "step": 14198 }, { "epoch": 1.806258745706653, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 0.00011682510375976562, "grad_norm": 7.368603706359863, "learning_rate": 1e-05, "loss": 0.5454, "mean_token_accuracy": 0.8618342876434326, "num_tokens": 538935539.0, "step": 14199 }, { "epoch": 1.8063859559852435, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.469811916351318, "learning_rate": 1e-05, "loss": 0.5613, "mean_token_accuracy": 0.8535925149917603, "num_tokens": 538978999.0, "step": 14200 }, { "epoch": 1.806513166263834, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.912091255187988, "learning_rate": 1e-05, "loss": 0.4752, "mean_token_accuracy": 0.877877950668335, "num_tokens": 539013480.0, "step": 14201 }, { "epoch": 1.8066403765424246, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.5610833168029785, "learning_rate": 1e-05, "loss": 0.5431, "mean_token_accuracy": 0.8603973984718323, "num_tokens": 539047120.0, "step": 14202 }, { "epoch": 1.8067675868210151, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.387374401092529, "learning_rate": 1e-05, "loss": 0.4777, "mean_token_accuracy": 0.8797165155410767, "num_tokens": 539088243.0, "step": 14203 }, { "epoch": 1.8068947970996057, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.615751266479492, "learning_rate": 1e-05, "loss": 0.5418, "mean_token_accuracy": 0.8608690500259399, "num_tokens": 539128240.0, "step": 14204 }, { "epoch": 1.8070220073781962, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.968962669372559, "learning_rate": 1e-05, "loss": 0.5007, "mean_token_accuracy": 0.8688721656799316, "num_tokens": 539168792.0, "step": 14205 }, { "epoch": 1.8071492176567867, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.797706604003906, "learning_rate": 1e-05, "loss": 0.4846, "mean_token_accuracy": 0.8746802806854248, "num_tokens": 539210510.0, "step": 14206 }, { "epoch": 1.8072764279353772, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.721800804138184, "learning_rate": 1e-05, "loss": 0.4872, "mean_token_accuracy": 0.8722679615020752, "num_tokens": 539248926.0, "step": 14207 }, { "epoch": 1.8074036382139678, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 0.00010967254638671875, "grad_norm": 7.11038064956665, "learning_rate": 1e-05, "loss": 0.5301, "mean_token_accuracy": 0.8649162650108337, "num_tokens": 539285795.0, "step": 14208 }, { "epoch": 1.8075308484925583, "ewc_loss": 0.10693359375, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.705516338348389, "learning_rate": 1e-05, "loss": 0.4654, "mean_token_accuracy": 0.8763819336891174, "num_tokens": 539324891.0, "step": 14209 }, { "epoch": 1.8076580587711488, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 0.00011920928955078125, "grad_norm": 7.437934875488281, "learning_rate": 1e-05, "loss": 0.5471, "mean_token_accuracy": 0.8608262538909912, "num_tokens": 539364684.0, "step": 14210 }, { "epoch": 1.8077852690497394, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 5.993335247039795, "learning_rate": 1e-05, "loss": 0.4671, "mean_token_accuracy": 0.8773568868637085, "num_tokens": 539403312.0, "step": 14211 }, { "epoch": 1.8079124793283299, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 0.0001201629638671875, "grad_norm": 7.485939025878906, "learning_rate": 1e-05, "loss": 0.4827, "mean_token_accuracy": 0.8821685910224915, "num_tokens": 539443567.0, "step": 14212 }, { "epoch": 1.8080396896069202, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.345649242401123, "learning_rate": 1e-05, "loss": 0.509, "mean_token_accuracy": 0.863084614276886, "num_tokens": 539480969.0, "step": 14213 }, { "epoch": 1.8081668998855107, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 0.00011348724365234375, "grad_norm": 7.220479965209961, "learning_rate": 1e-05, "loss": 0.5198, "mean_token_accuracy": 0.870166540145874, "num_tokens": 539518105.0, "step": 14214 }, { "epoch": 1.8082941101641012, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.455559253692627, "learning_rate": 1e-05, "loss": 0.5397, "mean_token_accuracy": 0.8552851676940918, "num_tokens": 539561364.0, "step": 14215 }, { "epoch": 1.8084213204426918, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.858494758605957, "learning_rate": 1e-05, "loss": 0.522, "mean_token_accuracy": 0.8644410967826843, "num_tokens": 539594602.0, "step": 14216 }, { "epoch": 1.8085485307212823, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.397264003753662, "learning_rate": 1e-05, "loss": 0.5378, "mean_token_accuracy": 0.8564549684524536, "num_tokens": 539635985.0, "step": 14217 }, { "epoch": 1.8086757409998728, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.567897796630859, "learning_rate": 1e-05, "loss": 0.4891, "mean_token_accuracy": 0.8721106648445129, "num_tokens": 539675201.0, "step": 14218 }, { "epoch": 1.8088029512784631, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.302562236785889, "learning_rate": 1e-05, "loss": 0.4739, "mean_token_accuracy": 0.8786619305610657, "num_tokens": 539712661.0, "step": 14219 }, { "epoch": 1.8089301615570537, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.404679298400879, "learning_rate": 1e-05, "loss": 0.5323, "mean_token_accuracy": 0.8571902513504028, "num_tokens": 539749941.0, "step": 14220 }, { "epoch": 1.8090573718356442, "ewc_loss": 0.11474609375, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.1508708000183105, "learning_rate": 1e-05, "loss": 0.5076, "mean_token_accuracy": 0.8653610348701477, "num_tokens": 539786906.0, "step": 14221 }, { "epoch": 1.8091845821142347, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.393779277801514, "learning_rate": 1e-05, "loss": 0.4887, "mean_token_accuracy": 0.873942494392395, "num_tokens": 539824235.0, "step": 14222 }, { "epoch": 1.8093117923928252, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.981505393981934, "learning_rate": 1e-05, "loss": 0.5316, "mean_token_accuracy": 0.8578875660896301, "num_tokens": 539864326.0, "step": 14223 }, { "epoch": 1.8094390026714158, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.480100154876709, "learning_rate": 1e-05, "loss": 0.5502, "mean_token_accuracy": 0.8559671640396118, "num_tokens": 539909105.0, "step": 14224 }, { "epoch": 1.8095662129500063, "ewc_loss": 0.10693359375, "ewc_loss_diag": 0.0002307891845703125, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.82550048828125, "learning_rate": 1e-05, "loss": 0.4883, "mean_token_accuracy": 0.8700141310691833, "num_tokens": 539944846.0, "step": 14225 }, { "epoch": 1.8096934232285968, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.0002307891845703125, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.676958084106445, "learning_rate": 1e-05, "loss": 0.4811, "mean_token_accuracy": 0.8725894689559937, "num_tokens": 539984720.0, "step": 14226 }, { "epoch": 1.8098206335071874, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.677590847015381, "learning_rate": 1e-05, "loss": 0.5029, "mean_token_accuracy": 0.8620353937149048, "num_tokens": 540025002.0, "step": 14227 }, { "epoch": 1.8099478437857779, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.0002307891845703125, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 7.022995948791504, "learning_rate": 1e-05, "loss": 0.5417, "mean_token_accuracy": 0.8608645796775818, "num_tokens": 540060256.0, "step": 14228 }, { "epoch": 1.8100750540643684, "ewc_loss": 0.10400390625, "ewc_loss_diag": 0.0002307891845703125, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.658468246459961, "learning_rate": 1e-05, "loss": 0.4947, "mean_token_accuracy": 0.8681185245513916, "num_tokens": 540098073.0, "step": 14229 }, { "epoch": 1.810202264342959, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.0002307891845703125, "ewc_loss_parallel": 0.000118255615234375, "grad_norm": 7.397861480712891, "learning_rate": 1e-05, "loss": 0.5099, "mean_token_accuracy": 0.8696261644363403, "num_tokens": 540128704.0, "step": 14230 }, { "epoch": 1.8103294746215495, "ewc_loss": 0.11474609375, "ewc_loss_diag": 0.0002307891845703125, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 5.91668176651001, "learning_rate": 1e-05, "loss": 0.4558, "mean_token_accuracy": 0.8817799687385559, "num_tokens": 540161293.0, "step": 14231 }, { "epoch": 1.81045668490014, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.0002307891845703125, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.508512496948242, "learning_rate": 1e-05, "loss": 0.4904, "mean_token_accuracy": 0.88152015209198, "num_tokens": 540193685.0, "step": 14232 }, { "epoch": 1.8105838951787305, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0002307891845703125, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.345505714416504, "learning_rate": 1e-05, "loss": 0.4601, "mean_token_accuracy": 0.8813343048095703, "num_tokens": 540232217.0, "step": 14233 }, { "epoch": 1.810711105457321, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.0002307891845703125, "ewc_loss_parallel": 0.0001125335693359375, "grad_norm": 7.16370153427124, "learning_rate": 1e-05, "loss": 0.5182, "mean_token_accuracy": 0.8664556741714478, "num_tokens": 540267561.0, "step": 14234 }, { "epoch": 1.8108383157359116, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002307891845703125, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.5388689041137695, "learning_rate": 1e-05, "loss": 0.4818, "mean_token_accuracy": 0.877180814743042, "num_tokens": 540305788.0, "step": 14235 }, { "epoch": 1.8109655260145021, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.0002307891845703125, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.644466876983643, "learning_rate": 1e-05, "loss": 0.5161, "mean_token_accuracy": 0.8639982342720032, "num_tokens": 540341105.0, "step": 14236 }, { "epoch": 1.8110927362930926, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002307891845703125, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.580508232116699, "learning_rate": 1e-05, "loss": 0.5306, "mean_token_accuracy": 0.8661094307899475, "num_tokens": 540380502.0, "step": 14237 }, { "epoch": 1.811219946571683, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.0002307891845703125, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 6.154069900512695, "learning_rate": 1e-05, "loss": 0.5543, "mean_token_accuracy": 0.8484798669815063, "num_tokens": 540420561.0, "step": 14238 }, { "epoch": 1.8113471568502735, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.0002307891845703125, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.699836254119873, "learning_rate": 1e-05, "loss": 0.5168, "mean_token_accuracy": 0.863914430141449, "num_tokens": 540460561.0, "step": 14239 }, { "epoch": 1.811474367128864, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.0002307891845703125, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.813044548034668, "learning_rate": 1e-05, "loss": 0.496, "mean_token_accuracy": 0.8671115040779114, "num_tokens": 540498939.0, "step": 14240 }, { "epoch": 1.8116015774074545, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002307891845703125, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.947119235992432, "learning_rate": 1e-05, "loss": 0.5262, "mean_token_accuracy": 0.861181914806366, "num_tokens": 540536127.0, "step": 14241 }, { "epoch": 1.811728787686045, "ewc_loss": 0.10498046875, "ewc_loss_diag": 0.0002307891845703125, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.692656517028809, "learning_rate": 1e-05, "loss": 0.4428, "mean_token_accuracy": 0.8812598586082458, "num_tokens": 540567812.0, "step": 14242 }, { "epoch": 1.8118559979646356, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002307891845703125, "ewc_loss_parallel": 0.00011587142944335938, "grad_norm": 7.283411502838135, "learning_rate": 1e-05, "loss": 0.4794, "mean_token_accuracy": 0.8805270195007324, "num_tokens": 540604176.0, "step": 14243 }, { "epoch": 1.811983208243226, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.0002307891845703125, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 5.836604118347168, "learning_rate": 1e-05, "loss": 0.537, "mean_token_accuracy": 0.8528535962104797, "num_tokens": 540640989.0, "step": 14244 }, { "epoch": 1.8121104185218164, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.0002307891845703125, "ewc_loss_parallel": 0.00012159347534179688, "grad_norm": 7.470040798187256, "learning_rate": 1e-05, "loss": 0.5015, "mean_token_accuracy": 0.8742501735687256, "num_tokens": 540680590.0, "step": 14245 }, { "epoch": 1.812237628800407, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.0002307891845703125, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.16605806350708, "learning_rate": 1e-05, "loss": 0.4892, "mean_token_accuracy": 0.8726147413253784, "num_tokens": 540723739.0, "step": 14246 }, { "epoch": 1.8123648390789975, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.0002307891845703125, "ewc_loss_parallel": 0.00011682510375976562, "grad_norm": 7.351293087005615, "learning_rate": 1e-05, "loss": 0.5102, "mean_token_accuracy": 0.8709757924079895, "num_tokens": 540759686.0, "step": 14247 }, { "epoch": 1.812492049357588, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0002307891845703125, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.398489475250244, "learning_rate": 1e-05, "loss": 0.4739, "mean_token_accuracy": 0.8835723400115967, "num_tokens": 540795710.0, "step": 14248 }, { "epoch": 1.8126192596361785, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002307891845703125, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 7.011268615722656, "learning_rate": 1e-05, "loss": 0.5014, "mean_token_accuracy": 0.86855149269104, "num_tokens": 540838558.0, "step": 14249 }, { "epoch": 1.812746469914769, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.0002307891845703125, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.409300804138184, "learning_rate": 1e-05, "loss": 0.4727, "mean_token_accuracy": 0.882493793964386, "num_tokens": 540874543.0, "step": 14250 }, { "epoch": 1.8128736801933596, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.0002307891845703125, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.6880950927734375, "learning_rate": 1e-05, "loss": 0.5112, "mean_token_accuracy": 0.870556116104126, "num_tokens": 540909903.0, "step": 14251 }, { "epoch": 1.8130008904719501, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.0002307891845703125, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.341141700744629, "learning_rate": 1e-05, "loss": 0.5041, "mean_token_accuracy": 0.8691362738609314, "num_tokens": 540947977.0, "step": 14252 }, { "epoch": 1.8131281007505406, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.0002307891845703125, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.468900203704834, "learning_rate": 1e-05, "loss": 0.5342, "mean_token_accuracy": 0.860080361366272, "num_tokens": 540986288.0, "step": 14253 }, { "epoch": 1.8132553110291312, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.0002307891845703125, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.204543113708496, "learning_rate": 1e-05, "loss": 0.4852, "mean_token_accuracy": 0.8712019920349121, "num_tokens": 541027607.0, "step": 14254 }, { "epoch": 1.8133825213077217, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.0002307891845703125, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.387669086456299, "learning_rate": 1e-05, "loss": 0.4771, "mean_token_accuracy": 0.8745046854019165, "num_tokens": 541064424.0, "step": 14255 }, { "epoch": 1.8135097315863122, "ewc_loss": 0.11083984375, "ewc_loss_diag": 0.0002307891845703125, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 6.056428909301758, "learning_rate": 1e-05, "loss": 0.4861, "mean_token_accuracy": 0.863359808921814, "num_tokens": 541097815.0, "step": 14256 }, { "epoch": 1.8136369418649028, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.0002307891845703125, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.423460960388184, "learning_rate": 1e-05, "loss": 0.5221, "mean_token_accuracy": 0.8622511625289917, "num_tokens": 541136567.0, "step": 14257 }, { "epoch": 1.8137641521434933, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.0002307891845703125, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.904420852661133, "learning_rate": 1e-05, "loss": 0.4746, "mean_token_accuracy": 0.8732642531394958, "num_tokens": 541174305.0, "step": 14258 }, { "epoch": 1.8138913624220838, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.0002307891845703125, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.56474494934082, "learning_rate": 1e-05, "loss": 0.4975, "mean_token_accuracy": 0.8698372840881348, "num_tokens": 541207118.0, "step": 14259 }, { "epoch": 1.8140185727006743, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.0002307891845703125, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.738167762756348, "learning_rate": 1e-05, "loss": 0.4479, "mean_token_accuracy": 0.8808579444885254, "num_tokens": 541242208.0, "step": 14260 }, { "epoch": 1.8141457829792649, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002307891845703125, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.822899341583252, "learning_rate": 1e-05, "loss": 0.4997, "mean_token_accuracy": 0.8723578453063965, "num_tokens": 541278553.0, "step": 14261 }, { "epoch": 1.8142729932578552, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.0002307891845703125, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.644163131713867, "learning_rate": 1e-05, "loss": 0.5016, "mean_token_accuracy": 0.8643485307693481, "num_tokens": 541318329.0, "step": 14262 }, { "epoch": 1.8144002035364457, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.0002307891845703125, "ewc_loss_parallel": 0.00011301040649414062, "grad_norm": 7.177371978759766, "learning_rate": 1e-05, "loss": 0.4928, "mean_token_accuracy": 0.8774667382240295, "num_tokens": 541353406.0, "step": 14263 }, { "epoch": 1.8145274138150362, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.0002307891845703125, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.754430294036865, "learning_rate": 1e-05, "loss": 0.4955, "mean_token_accuracy": 0.8663173913955688, "num_tokens": 541385504.0, "step": 14264 }, { "epoch": 1.8146546240936268, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.0002307891845703125, "ewc_loss_parallel": 0.00012063980102539062, "grad_norm": 7.45069694519043, "learning_rate": 1e-05, "loss": 0.5444, "mean_token_accuracy": 0.8630149960517883, "num_tokens": 541426172.0, "step": 14265 }, { "epoch": 1.8147818343722173, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.0002307891845703125, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.109159469604492, "learning_rate": 1e-05, "loss": 0.4944, "mean_token_accuracy": 0.8740373253822327, "num_tokens": 541464785.0, "step": 14266 }, { "epoch": 1.8149090446508078, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.0002307891845703125, "ewc_loss_parallel": 0.00011682510375976562, "grad_norm": 7.344625949859619, "learning_rate": 1e-05, "loss": 0.5358, "mean_token_accuracy": 0.8622130155563354, "num_tokens": 541500271.0, "step": 14267 }, { "epoch": 1.8150362549293981, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0002307891845703125, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.435818672180176, "learning_rate": 1e-05, "loss": 0.5196, "mean_token_accuracy": 0.8663328886032104, "num_tokens": 541537482.0, "step": 14268 }, { "epoch": 1.8151634652079887, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002307891845703125, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.887194633483887, "learning_rate": 1e-05, "loss": 0.4946, "mean_token_accuracy": 0.8726317286491394, "num_tokens": 541576669.0, "step": 14269 }, { "epoch": 1.8152906754865792, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002307891845703125, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.53830623626709, "learning_rate": 1e-05, "loss": 0.5102, "mean_token_accuracy": 0.8699464797973633, "num_tokens": 541616040.0, "step": 14270 }, { "epoch": 1.8154178857651697, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.0002307891845703125, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.356731414794922, "learning_rate": 1e-05, "loss": 0.5415, "mean_token_accuracy": 0.8529348373413086, "num_tokens": 541649994.0, "step": 14271 }, { "epoch": 1.8155450960437602, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002307891845703125, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.626697063446045, "learning_rate": 1e-05, "loss": 0.5509, "mean_token_accuracy": 0.8568460941314697, "num_tokens": 541688468.0, "step": 14272 }, { "epoch": 1.8156723063223508, "ewc_loss": 0.10791015625, "ewc_loss_diag": 0.0002307891845703125, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.932763576507568, "learning_rate": 1e-05, "loss": 0.4993, "mean_token_accuracy": 0.8656872510910034, "num_tokens": 541725173.0, "step": 14273 }, { "epoch": 1.8157995166009413, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002307891845703125, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.82802152633667, "learning_rate": 1e-05, "loss": 0.5194, "mean_token_accuracy": 0.8657748699188232, "num_tokens": 541762523.0, "step": 14274 }, { "epoch": 1.8159267268795318, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.0002307891845703125, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.7075300216674805, "learning_rate": 1e-05, "loss": 0.4768, "mean_token_accuracy": 0.8682928085327148, "num_tokens": 541799731.0, "step": 14275 }, { "epoch": 1.8160539371581224, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002307891845703125, "ewc_loss_parallel": 0.00011110305786132812, "grad_norm": 7.134390354156494, "learning_rate": 1e-05, "loss": 0.4886, "mean_token_accuracy": 0.8792163133621216, "num_tokens": 541844172.0, "step": 14276 }, { "epoch": 1.8161811474367129, "ewc_loss": 0.10693359375, "ewc_loss_diag": 0.0002307891845703125, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.726358890533447, "learning_rate": 1e-05, "loss": 0.4839, "mean_token_accuracy": 0.8704067468643188, "num_tokens": 541887526.0, "step": 14277 }, { "epoch": 1.8163083577153034, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.0002307891845703125, "ewc_loss_parallel": 0.00011920928955078125, "grad_norm": 7.43065071105957, "learning_rate": 1e-05, "loss": 0.5151, "mean_token_accuracy": 0.8702220320701599, "num_tokens": 541932712.0, "step": 14278 }, { "epoch": 1.816435567993894, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.0002307891845703125, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.025506496429443, "learning_rate": 1e-05, "loss": 0.4997, "mean_token_accuracy": 0.869936466217041, "num_tokens": 541968915.0, "step": 14279 }, { "epoch": 1.8165627782724845, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.0002307891845703125, "ewc_loss_parallel": 0.0001201629638671875, "grad_norm": 7.4416117668151855, "learning_rate": 1e-05, "loss": 0.5435, "mean_token_accuracy": 0.863836407661438, "num_tokens": 542010296.0, "step": 14280 }, { "epoch": 1.816689988551075, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.0002307891845703125, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.33319616317749, "learning_rate": 1e-05, "loss": 0.4501, "mean_token_accuracy": 0.8857828974723816, "num_tokens": 542050096.0, "step": 14281 }, { "epoch": 1.8168171988296655, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002307891845703125, "ewc_loss_parallel": 0.00011205673217773438, "grad_norm": 7.171623229980469, "learning_rate": 1e-05, "loss": 0.5256, "mean_token_accuracy": 0.8648051023483276, "num_tokens": 542086746.0, "step": 14282 }, { "epoch": 1.816944409108256, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.0002307891845703125, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.419647216796875, "learning_rate": 1e-05, "loss": 0.4891, "mean_token_accuracy": 0.8733923435211182, "num_tokens": 542123095.0, "step": 14283 }, { "epoch": 1.8170716193868466, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002307891845703125, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.833672523498535, "learning_rate": 1e-05, "loss": 0.4901, "mean_token_accuracy": 0.8743928670883179, "num_tokens": 542157797.0, "step": 14284 }, { "epoch": 1.817198829665437, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.0002307891845703125, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.3540544509887695, "learning_rate": 1e-05, "loss": 0.5088, "mean_token_accuracy": 0.8615260124206543, "num_tokens": 542196091.0, "step": 14285 }, { "epoch": 1.8173260399440276, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.0002307891845703125, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.618035316467285, "learning_rate": 1e-05, "loss": 0.5417, "mean_token_accuracy": 0.8556939363479614, "num_tokens": 542234242.0, "step": 14286 }, { "epoch": 1.817453250222618, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.0002307891845703125, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.233965873718262, "learning_rate": 1e-05, "loss": 0.4822, "mean_token_accuracy": 0.8738765716552734, "num_tokens": 542273187.0, "step": 14287 }, { "epoch": 1.8175804605012085, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.0002307891845703125, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.464865684509277, "learning_rate": 1e-05, "loss": 0.4701, "mean_token_accuracy": 0.8766575455665588, "num_tokens": 542308116.0, "step": 14288 }, { "epoch": 1.817707670779799, "ewc_loss": 0.11474609375, "ewc_loss_diag": 0.0002307891845703125, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.0962395668029785, "learning_rate": 1e-05, "loss": 0.4768, "mean_token_accuracy": 0.8755618929862976, "num_tokens": 542339964.0, "step": 14289 }, { "epoch": 1.8178348810583895, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.0002307891845703125, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.47127103805542, "learning_rate": 1e-05, "loss": 0.5183, "mean_token_accuracy": 0.863650918006897, "num_tokens": 542381923.0, "step": 14290 }, { "epoch": 1.81796209133698, "ewc_loss": 0.109375, "ewc_loss_diag": 0.0002307891845703125, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.945519924163818, "learning_rate": 1e-05, "loss": 0.4589, "mean_token_accuracy": 0.8792788982391357, "num_tokens": 542418468.0, "step": 14291 }, { "epoch": 1.8180893016155706, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.0002307891845703125, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.590300559997559, "learning_rate": 1e-05, "loss": 0.588, "mean_token_accuracy": 0.8498250246047974, "num_tokens": 542448806.0, "step": 14292 }, { "epoch": 1.818216511894161, "ewc_loss": 0.10595703125, "ewc_loss_diag": 0.0002307891845703125, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.807082176208496, "learning_rate": 1e-05, "loss": 0.4806, "mean_token_accuracy": 0.8717124462127686, "num_tokens": 542483991.0, "step": 14293 }, { "epoch": 1.8183437221727514, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0002307891845703125, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.788452625274658, "learning_rate": 1e-05, "loss": 0.4963, "mean_token_accuracy": 0.8739405870437622, "num_tokens": 542528334.0, "step": 14294 }, { "epoch": 1.818470932451342, "ewc_loss": 0.10595703125, "ewc_loss_diag": 0.0002307891845703125, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.712291240692139, "learning_rate": 1e-05, "loss": 0.4467, "mean_token_accuracy": 0.881348729133606, "num_tokens": 542562903.0, "step": 14295 }, { "epoch": 1.8185981427299325, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.0002307891845703125, "ewc_loss_parallel": 0.00011014938354492188, "grad_norm": 7.0848917961120605, "learning_rate": 1e-05, "loss": 0.4971, "mean_token_accuracy": 0.8720846176147461, "num_tokens": 542600150.0, "step": 14296 }, { "epoch": 1.818725353008523, "ewc_loss": 0.10888671875, "ewc_loss_diag": 0.0002307891845703125, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.780074596405029, "learning_rate": 1e-05, "loss": 0.4533, "mean_token_accuracy": 0.8775023818016052, "num_tokens": 542637894.0, "step": 14297 }, { "epoch": 1.8188525632871135, "ewc_loss": 0.140625, "ewc_loss_diag": 0.0002307891845703125, "ewc_loss_parallel": 0.00011730194091796875, "grad_norm": 7.318298816680908, "learning_rate": 1e-05, "loss": 0.5635, "mean_token_accuracy": 0.8558443784713745, "num_tokens": 542667729.0, "step": 14298 }, { "epoch": 1.818979773565704, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.0002307891845703125, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.071697235107422, "learning_rate": 1e-05, "loss": 0.4856, "mean_token_accuracy": 0.8731604814529419, "num_tokens": 542699994.0, "step": 14299 }, { "epoch": 1.8191069838442946, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.0002307891845703125, "ewc_loss_parallel": 0.00011444091796875, "grad_norm": 7.196555137634277, "learning_rate": 1e-05, "loss": 0.4913, "mean_token_accuracy": 0.8763982653617859, "num_tokens": 542743849.0, "step": 14300 }, { "epoch": 1.8192341941228851, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.0002307891845703125, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.334338188171387, "learning_rate": 1e-05, "loss": 0.5205, "mean_token_accuracy": 0.8595054149627686, "num_tokens": 542784788.0, "step": 14301 }, { "epoch": 1.8193614044014756, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0002307891845703125, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.788651943206787, "learning_rate": 1e-05, "loss": 0.5246, "mean_token_accuracy": 0.867944061756134, "num_tokens": 542820942.0, "step": 14302 }, { "epoch": 1.8194886146800662, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.0002307891845703125, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 7.351778984069824, "learning_rate": 1e-05, "loss": 0.472, "mean_token_accuracy": 0.8793065547943115, "num_tokens": 542857345.0, "step": 14303 }, { "epoch": 1.8196158249586567, "ewc_loss": 0.109375, "ewc_loss_diag": 0.0002307891845703125, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.965585708618164, "learning_rate": 1e-05, "loss": 0.4602, "mean_token_accuracy": 0.879015326499939, "num_tokens": 542890952.0, "step": 14304 }, { "epoch": 1.8197430352372472, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.0002307891845703125, "ewc_loss_parallel": 0.0001125335693359375, "grad_norm": 7.092283725738525, "learning_rate": 1e-05, "loss": 0.5611, "mean_token_accuracy": 0.8577983379364014, "num_tokens": 542927487.0, "step": 14305 }, { "epoch": 1.8198702455158378, "ewc_loss": 0.10595703125, "ewc_loss_diag": 0.0002307891845703125, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.6254658699035645, "learning_rate": 1e-05, "loss": 0.4802, "mean_token_accuracy": 0.8682260513305664, "num_tokens": 542967554.0, "step": 14306 }, { "epoch": 1.8199974557944283, "ewc_loss": 0.15234375, "ewc_loss_diag": 0.0002307891845703125, "ewc_loss_parallel": 0.00012874603271484375, "grad_norm": 7.6610894203186035, "learning_rate": 1e-05, "loss": 0.5693, "mean_token_accuracy": 0.8584297299385071, "num_tokens": 543003768.0, "step": 14307 }, { "epoch": 1.8201246660730188, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.0002307891845703125, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 5.9917097091674805, "learning_rate": 1e-05, "loss": 0.4783, "mean_token_accuracy": 0.8739304542541504, "num_tokens": 543041937.0, "step": 14308 }, { "epoch": 1.8202518763516093, "ewc_loss": 0.154296875, "ewc_loss_diag": 0.0002307891845703125, "ewc_loss_parallel": 0.00013065338134765625, "grad_norm": 7.8179755210876465, "learning_rate": 1e-05, "loss": 0.4855, "mean_token_accuracy": 0.8865655660629272, "num_tokens": 543084358.0, "step": 14309 }, { "epoch": 1.8203790866301999, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002307891845703125, "ewc_loss_parallel": 0.00011157989501953125, "grad_norm": 6.594664573669434, "learning_rate": 1e-05, "loss": 0.5088, "mean_token_accuracy": 0.8742092251777649, "num_tokens": 543121749.0, "step": 14310 }, { "epoch": 1.8205062969087902, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.0002307891845703125, "ewc_loss_parallel": 0.00012063980102539062, "grad_norm": 7.450429439544678, "learning_rate": 1e-05, "loss": 0.5462, "mean_token_accuracy": 0.8592322468757629, "num_tokens": 543161160.0, "step": 14311 }, { "epoch": 1.8206335071873807, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002307891845703125, "ewc_loss_parallel": 0.00011110305786132812, "grad_norm": 6.681173801422119, "learning_rate": 1e-05, "loss": 0.5157, "mean_token_accuracy": 0.868297278881073, "num_tokens": 543197364.0, "step": 14312 }, { "epoch": 1.8207607174659712, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.0002307891845703125, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 7.048717498779297, "learning_rate": 1e-05, "loss": 0.471, "mean_token_accuracy": 0.8838037252426147, "num_tokens": 543233381.0, "step": 14313 }, { "epoch": 1.8208879277445618, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0002307891845703125, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.504333019256592, "learning_rate": 1e-05, "loss": 0.5354, "mean_token_accuracy": 0.8617269992828369, "num_tokens": 543266139.0, "step": 14314 }, { "epoch": 1.8210151380231523, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.831586837768555, "learning_rate": 1e-05, "loss": 0.5594, "mean_token_accuracy": 0.8545233011245728, "num_tokens": 543307681.0, "step": 14315 }, { "epoch": 1.8211423483017428, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.246751308441162, "learning_rate": 1e-05, "loss": 0.5421, "mean_token_accuracy": 0.860965371131897, "num_tokens": 543347916.0, "step": 14316 }, { "epoch": 1.8212695585803331, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.809258460998535, "learning_rate": 1e-05, "loss": 0.4691, "mean_token_accuracy": 0.8798113465309143, "num_tokens": 543386326.0, "step": 14317 }, { "epoch": 1.8213967688589237, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.035201549530029, "learning_rate": 1e-05, "loss": 0.5118, "mean_token_accuracy": 0.8649710416793823, "num_tokens": 543423220.0, "step": 14318 }, { "epoch": 1.8215239791375142, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 6.933940410614014, "learning_rate": 1e-05, "loss": 0.4839, "mean_token_accuracy": 0.875293493270874, "num_tokens": 543461672.0, "step": 14319 }, { "epoch": 1.8216511894161047, "ewc_loss": 0.10986328125, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.90598726272583, "learning_rate": 1e-05, "loss": 0.4776, "mean_token_accuracy": 0.8727273941040039, "num_tokens": 543498775.0, "step": 14320 }, { "epoch": 1.8217783996946952, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 0.00011301040649414062, "grad_norm": 7.158288955688477, "learning_rate": 1e-05, "loss": 0.4742, "mean_token_accuracy": 0.8818807601928711, "num_tokens": 543539966.0, "step": 14321 }, { "epoch": 1.8219056099732858, "ewc_loss": 0.11181640625, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 5.944719314575195, "learning_rate": 1e-05, "loss": 0.4818, "mean_token_accuracy": 0.8725075721740723, "num_tokens": 543577176.0, "step": 14322 }, { "epoch": 1.8220328202518763, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 0.00011587142944335938, "grad_norm": 7.322820663452148, "learning_rate": 1e-05, "loss": 0.5625, "mean_token_accuracy": 0.8586951494216919, "num_tokens": 543621328.0, "step": 14323 }, { "epoch": 1.8221600305304668, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.1759748458862305, "learning_rate": 1e-05, "loss": 0.5214, "mean_token_accuracy": 0.8580381870269775, "num_tokens": 543662446.0, "step": 14324 }, { "epoch": 1.8222872408090574, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 0.00011348724365234375, "grad_norm": 7.215815544128418, "learning_rate": 1e-05, "loss": 0.4716, "mean_token_accuracy": 0.8802944421768188, "num_tokens": 543695739.0, "step": 14325 }, { "epoch": 1.8224144510876479, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.4107279777526855, "learning_rate": 1e-05, "loss": 0.4846, "mean_token_accuracy": 0.8753844499588013, "num_tokens": 543735823.0, "step": 14326 }, { "epoch": 1.8225416613662384, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.808985233306885, "learning_rate": 1e-05, "loss": 0.5566, "mean_token_accuracy": 0.8525530099868774, "num_tokens": 543771800.0, "step": 14327 }, { "epoch": 1.822668871644829, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.505475997924805, "learning_rate": 1e-05, "loss": 0.5302, "mean_token_accuracy": 0.8611997365951538, "num_tokens": 543812425.0, "step": 14328 }, { "epoch": 1.8227960819234195, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.315645694732666, "learning_rate": 1e-05, "loss": 0.4846, "mean_token_accuracy": 0.8740522265434265, "num_tokens": 543849867.0, "step": 14329 }, { "epoch": 1.82292329220201, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.578005313873291, "learning_rate": 1e-05, "loss": 0.4746, "mean_token_accuracy": 0.8806877136230469, "num_tokens": 543887289.0, "step": 14330 }, { "epoch": 1.8230505024806005, "ewc_loss": 0.10693359375, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.888762950897217, "learning_rate": 1e-05, "loss": 0.4632, "mean_token_accuracy": 0.8743565082550049, "num_tokens": 543923432.0, "step": 14331 }, { "epoch": 1.823177712759191, "ewc_loss": 0.125, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.7874836921691895, "learning_rate": 1e-05, "loss": 0.5448, "mean_token_accuracy": 0.8557209372520447, "num_tokens": 543966655.0, "step": 14332 }, { "epoch": 1.8233049230377816, "ewc_loss": 0.10400390625, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.602437973022461, "learning_rate": 1e-05, "loss": 0.4521, "mean_token_accuracy": 0.8765205144882202, "num_tokens": 544006920.0, "step": 14333 }, { "epoch": 1.823432133316372, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 0.0001125335693359375, "grad_norm": 7.168827533721924, "learning_rate": 1e-05, "loss": 0.5127, "mean_token_accuracy": 0.8702603578567505, "num_tokens": 544044066.0, "step": 14334 }, { "epoch": 1.8235593435949626, "ewc_loss": 0.10498046875, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.599461555480957, "learning_rate": 1e-05, "loss": 0.508, "mean_token_accuracy": 0.8639200925827026, "num_tokens": 544073795.0, "step": 14335 }, { "epoch": 1.823686553873553, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.614351749420166, "learning_rate": 1e-05, "loss": 0.5141, "mean_token_accuracy": 0.8752263188362122, "num_tokens": 544107378.0, "step": 14336 }, { "epoch": 1.8238137641521435, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 5.993234157562256, "learning_rate": 1e-05, "loss": 0.4853, "mean_token_accuracy": 0.8727370500564575, "num_tokens": 544144377.0, "step": 14337 }, { "epoch": 1.823940974430734, "ewc_loss": 0.150390625, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 0.00012683868408203125, "grad_norm": 7.725133419036865, "learning_rate": 1e-05, "loss": 0.5539, "mean_token_accuracy": 0.8660205602645874, "num_tokens": 544176484.0, "step": 14338 }, { "epoch": 1.8240681847093245, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 6.542243957519531, "learning_rate": 1e-05, "loss": 0.5002, "mean_token_accuracy": 0.8766123056411743, "num_tokens": 544209152.0, "step": 14339 }, { "epoch": 1.824195394987915, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 0.0001163482666015625, "grad_norm": 7.312736511230469, "learning_rate": 1e-05, "loss": 0.5507, "mean_token_accuracy": 0.8653680086135864, "num_tokens": 544249197.0, "step": 14340 }, { "epoch": 1.8243226052665056, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 6.6757893562316895, "learning_rate": 1e-05, "loss": 0.4684, "mean_token_accuracy": 0.8852459192276001, "num_tokens": 544290225.0, "step": 14341 }, { "epoch": 1.8244498155450959, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.816090106964111, "learning_rate": 1e-05, "loss": 0.5109, "mean_token_accuracy": 0.8669968843460083, "num_tokens": 544333015.0, "step": 14342 }, { "epoch": 1.8245770258236864, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.5898308753967285, "learning_rate": 1e-05, "loss": 0.5297, "mean_token_accuracy": 0.8630298972129822, "num_tokens": 544370165.0, "step": 14343 }, { "epoch": 1.824704236102277, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.426346302032471, "learning_rate": 1e-05, "loss": 0.4944, "mean_token_accuracy": 0.8715614080429077, "num_tokens": 544407284.0, "step": 14344 }, { "epoch": 1.8248314463808675, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.486488342285156, "learning_rate": 1e-05, "loss": 0.5508, "mean_token_accuracy": 0.8576998710632324, "num_tokens": 544441558.0, "step": 14345 }, { "epoch": 1.824958656659458, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.209354400634766, "learning_rate": 1e-05, "loss": 0.5659, "mean_token_accuracy": 0.8440005779266357, "num_tokens": 544478468.0, "step": 14346 }, { "epoch": 1.8250858669380485, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.388266563415527, "learning_rate": 1e-05, "loss": 0.4504, "mean_token_accuracy": 0.8874263763427734, "num_tokens": 544516271.0, "step": 14347 }, { "epoch": 1.825213077216639, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 6.0977702140808105, "learning_rate": 1e-05, "loss": 0.4612, "mean_token_accuracy": 0.880493700504303, "num_tokens": 544553198.0, "step": 14348 }, { "epoch": 1.8253402874952296, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.305429458618164, "learning_rate": 1e-05, "loss": 0.4718, "mean_token_accuracy": 0.8767699003219604, "num_tokens": 544586381.0, "step": 14349 }, { "epoch": 1.8254674977738201, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 6.193004131317139, "learning_rate": 1e-05, "loss": 0.498, "mean_token_accuracy": 0.8679921627044678, "num_tokens": 544628342.0, "step": 14350 }, { "epoch": 1.8255947080524106, "ewc_loss": 0.11376953125, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.138730049133301, "learning_rate": 1e-05, "loss": 0.5025, "mean_token_accuracy": 0.8669569492340088, "num_tokens": 544667096.0, "step": 14351 }, { "epoch": 1.8257219183310012, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.1205878257751465, "learning_rate": 1e-05, "loss": 0.4756, "mean_token_accuracy": 0.8754559755325317, "num_tokens": 544711467.0, "step": 14352 }, { "epoch": 1.8258491286095917, "ewc_loss": 0.11181640625, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 6.03578519821167, "learning_rate": 1e-05, "loss": 0.4787, "mean_token_accuracy": 0.8725807666778564, "num_tokens": 544751672.0, "step": 14353 }, { "epoch": 1.8259763388881822, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.157937526702881, "learning_rate": 1e-05, "loss": 0.4236, "mean_token_accuracy": 0.8928759694099426, "num_tokens": 544785479.0, "step": 14354 }, { "epoch": 1.8261035491667728, "ewc_loss": 0.11181640625, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 5.961028575897217, "learning_rate": 1e-05, "loss": 0.4792, "mean_token_accuracy": 0.8720718622207642, "num_tokens": 544823122.0, "step": 14355 }, { "epoch": 1.8262307594453633, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.225330352783203, "learning_rate": 1e-05, "loss": 0.4668, "mean_token_accuracy": 0.8789043426513672, "num_tokens": 544862544.0, "step": 14356 }, { "epoch": 1.8263579697239538, "ewc_loss": 0.109375, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.837801933288574, "learning_rate": 1e-05, "loss": 0.5055, "mean_token_accuracy": 0.864121675491333, "num_tokens": 544899538.0, "step": 14357 }, { "epoch": 1.8264851800025443, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.389103889465332, "learning_rate": 1e-05, "loss": 0.4715, "mean_token_accuracy": 0.8778039813041687, "num_tokens": 544934544.0, "step": 14358 }, { "epoch": 1.8266123902811349, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.666835784912109, "learning_rate": 1e-05, "loss": 0.5116, "mean_token_accuracy": 0.8624696731567383, "num_tokens": 544973304.0, "step": 14359 }, { "epoch": 1.8267396005597252, "ewc_loss": 0.125, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.695620059967041, "learning_rate": 1e-05, "loss": 0.4694, "mean_token_accuracy": 0.8791322708129883, "num_tokens": 545010748.0, "step": 14360 }, { "epoch": 1.8268668108383157, "ewc_loss": 0.103515625, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.441832065582275, "learning_rate": 1e-05, "loss": 0.4915, "mean_token_accuracy": 0.8641947507858276, "num_tokens": 545050742.0, "step": 14361 }, { "epoch": 1.8269940211169062, "ewc_loss": 0.140625, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 0.00011777877807617188, "grad_norm": 7.291023254394531, "learning_rate": 1e-05, "loss": 0.5288, "mean_token_accuracy": 0.8673914074897766, "num_tokens": 545083904.0, "step": 14362 }, { "epoch": 1.8271212313954968, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.406469345092773, "learning_rate": 1e-05, "loss": 0.452, "mean_token_accuracy": 0.8792301416397095, "num_tokens": 545123885.0, "step": 14363 }, { "epoch": 1.8272484416740873, "ewc_loss": 0.16015625, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 0.00013637542724609375, "grad_norm": 8.090121269226074, "learning_rate": 1e-05, "loss": 0.5536, "mean_token_accuracy": 0.8635783791542053, "num_tokens": 545160295.0, "step": 14364 }, { "epoch": 1.8273756519526778, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.14285945892334, "learning_rate": 1e-05, "loss": 0.5505, "mean_token_accuracy": 0.8579614758491516, "num_tokens": 545203469.0, "step": 14365 }, { "epoch": 1.8275028622312681, "ewc_loss": 0.1650390625, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 0.000141143798828125, "grad_norm": 8.281408309936523, "learning_rate": 1e-05, "loss": 0.5231, "mean_token_accuracy": 0.8755319118499756, "num_tokens": 545242173.0, "step": 14366 }, { "epoch": 1.8276300725098586, "ewc_loss": 0.1484375, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 0.00012493133544921875, "grad_norm": 7.0661115646362305, "learning_rate": 1e-05, "loss": 0.5418, "mean_token_accuracy": 0.8634041547775269, "num_tokens": 545284968.0, "step": 14367 }, { "epoch": 1.8277572827884492, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.62509822845459, "learning_rate": 1e-05, "loss": 0.5797, "mean_token_accuracy": 0.8543592691421509, "num_tokens": 545324764.0, "step": 14368 }, { "epoch": 1.8278844930670397, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 0.00011920928955078125, "grad_norm": 7.037957668304443, "learning_rate": 1e-05, "loss": 0.5384, "mean_token_accuracy": 0.8643845915794373, "num_tokens": 545362783.0, "step": 14369 }, { "epoch": 1.8280117033456302, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 0.00011110305786132812, "grad_norm": 7.151239395141602, "learning_rate": 1e-05, "loss": 0.5141, "mean_token_accuracy": 0.8783031105995178, "num_tokens": 545395757.0, "step": 14370 }, { "epoch": 1.8281389136242208, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 6.696738243103027, "learning_rate": 1e-05, "loss": 0.4676, "mean_token_accuracy": 0.8840264081954956, "num_tokens": 545432194.0, "step": 14371 }, { "epoch": 1.8282661239028113, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.86129093170166, "learning_rate": 1e-05, "loss": 0.4886, "mean_token_accuracy": 0.8745990991592407, "num_tokens": 545469743.0, "step": 14372 }, { "epoch": 1.8283933341814018, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.43804407119751, "learning_rate": 1e-05, "loss": 0.5187, "mean_token_accuracy": 0.8640096187591553, "num_tokens": 545508976.0, "step": 14373 }, { "epoch": 1.8285205444599923, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.701995372772217, "learning_rate": 1e-05, "loss": 0.4981, "mean_token_accuracy": 0.8708814978599548, "num_tokens": 545548223.0, "step": 14374 }, { "epoch": 1.8286477547385829, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.214990615844727, "learning_rate": 1e-05, "loss": 0.5437, "mean_token_accuracy": 0.854564905166626, "num_tokens": 545586110.0, "step": 14375 }, { "epoch": 1.8287749650171734, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.668585777282715, "learning_rate": 1e-05, "loss": 0.4983, "mean_token_accuracy": 0.8666149377822876, "num_tokens": 545625779.0, "step": 14376 }, { "epoch": 1.828902175295764, "ewc_loss": 0.109375, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.997956275939941, "learning_rate": 1e-05, "loss": 0.5486, "mean_token_accuracy": 0.8491010665893555, "num_tokens": 545669464.0, "step": 14377 }, { "epoch": 1.8290293855743545, "ewc_loss": 0.125, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.803866863250732, "learning_rate": 1e-05, "loss": 0.4677, "mean_token_accuracy": 0.8784457445144653, "num_tokens": 545702674.0, "step": 14378 }, { "epoch": 1.829156595852945, "ewc_loss": 0.10693359375, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.811728000640869, "learning_rate": 1e-05, "loss": 0.4614, "mean_token_accuracy": 0.8750513792037964, "num_tokens": 545742410.0, "step": 14379 }, { "epoch": 1.8292838061315355, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 7.1075439453125, "learning_rate": 1e-05, "loss": 0.5311, "mean_token_accuracy": 0.8632230162620544, "num_tokens": 545779748.0, "step": 14380 }, { "epoch": 1.829411016410126, "ewc_loss": 0.10693359375, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.787328720092773, "learning_rate": 1e-05, "loss": 0.5037, "mean_token_accuracy": 0.8629336953163147, "num_tokens": 545811401.0, "step": 14381 }, { "epoch": 1.8295382266887166, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 0.00011873245239257812, "grad_norm": 7.4686360359191895, "learning_rate": 1e-05, "loss": 0.4904, "mean_token_accuracy": 0.8775372505187988, "num_tokens": 545848565.0, "step": 14382 }, { "epoch": 1.829665436967307, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.048420429229736, "learning_rate": 1e-05, "loss": 0.5134, "mean_token_accuracy": 0.8664487600326538, "num_tokens": 545891245.0, "step": 14383 }, { "epoch": 1.8297926472458976, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 0.00012111663818359375, "grad_norm": 7.571357250213623, "learning_rate": 1e-05, "loss": 0.4587, "mean_token_accuracy": 0.8881997466087341, "num_tokens": 545924860.0, "step": 14384 }, { "epoch": 1.829919857524488, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.503039360046387, "learning_rate": 1e-05, "loss": 0.4711, "mean_token_accuracy": 0.8804095983505249, "num_tokens": 545958119.0, "step": 14385 }, { "epoch": 1.8300470678030785, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 0.00011014938354492188, "grad_norm": 7.167046070098877, "learning_rate": 1e-05, "loss": 0.5456, "mean_token_accuracy": 0.8642193675041199, "num_tokens": 545994195.0, "step": 14386 }, { "epoch": 1.830174278081669, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.675511360168457, "learning_rate": 1e-05, "loss": 0.5106, "mean_token_accuracy": 0.8701039552688599, "num_tokens": 546031998.0, "step": 14387 }, { "epoch": 1.8303014883602595, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.590146064758301, "learning_rate": 1e-05, "loss": 0.487, "mean_token_accuracy": 0.8701995015144348, "num_tokens": 546067152.0, "step": 14388 }, { "epoch": 1.83042869863885, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.694270133972168, "learning_rate": 1e-05, "loss": 0.4623, "mean_token_accuracy": 0.8834584355354309, "num_tokens": 546104207.0, "step": 14389 }, { "epoch": 1.8305559089174406, "ewc_loss": 0.10986328125, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 6.050230026245117, "learning_rate": 1e-05, "loss": 0.5255, "mean_token_accuracy": 0.8574209213256836, "num_tokens": 546150659.0, "step": 14390 }, { "epoch": 1.8306831191960309, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.866137981414795, "learning_rate": 1e-05, "loss": 0.4834, "mean_token_accuracy": 0.8765825033187866, "num_tokens": 546188427.0, "step": 14391 }, { "epoch": 1.8308103294746214, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.7194504737854, "learning_rate": 1e-05, "loss": 0.5026, "mean_token_accuracy": 0.8627318739891052, "num_tokens": 546227249.0, "step": 14392 }, { "epoch": 1.830937539753212, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 0.00011205673217773438, "grad_norm": 7.233932018280029, "learning_rate": 1e-05, "loss": 0.5522, "mean_token_accuracy": 0.860088586807251, "num_tokens": 546263469.0, "step": 14393 }, { "epoch": 1.8310647500318025, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.696510314941406, "learning_rate": 1e-05, "loss": 0.4998, "mean_token_accuracy": 0.8666601181030273, "num_tokens": 546300820.0, "step": 14394 }, { "epoch": 1.831191960310393, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.631545543670654, "learning_rate": 1e-05, "loss": 0.5017, "mean_token_accuracy": 0.8758891820907593, "num_tokens": 546336639.0, "step": 14395 }, { "epoch": 1.8313191705889835, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.100827693939209, "learning_rate": 1e-05, "loss": 0.5191, "mean_token_accuracy": 0.8637099266052246, "num_tokens": 546377769.0, "step": 14396 }, { "epoch": 1.831446380867574, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.660270690917969, "learning_rate": 1e-05, "loss": 0.4932, "mean_token_accuracy": 0.8780766129493713, "num_tokens": 546419286.0, "step": 14397 }, { "epoch": 1.8315735911461646, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.557551383972168, "learning_rate": 1e-05, "loss": 0.5483, "mean_token_accuracy": 0.8581656217575073, "num_tokens": 546458907.0, "step": 14398 }, { "epoch": 1.831700801424755, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 0.00011348724365234375, "grad_norm": 7.2656707763671875, "learning_rate": 1e-05, "loss": 0.4515, "mean_token_accuracy": 0.8884868025779724, "num_tokens": 546491682.0, "step": 14399 }, { "epoch": 1.8318280117033456, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.625158786773682, "learning_rate": 1e-05, "loss": 0.5727, "mean_token_accuracy": 0.849635124206543, "num_tokens": 546537141.0, "step": 14400 }, { "epoch": 1.8319552219819362, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.857324600219727, "learning_rate": 1e-05, "loss": 0.5084, "mean_token_accuracy": 0.8665002584457397, "num_tokens": 546574318.0, "step": 14401 }, { "epoch": 1.8320824322605267, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.483037948608398, "learning_rate": 1e-05, "loss": 0.5314, "mean_token_accuracy": 0.8605883121490479, "num_tokens": 546615938.0, "step": 14402 }, { "epoch": 1.8322096425391172, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.593491077423096, "learning_rate": 1e-05, "loss": 0.4941, "mean_token_accuracy": 0.8732986450195312, "num_tokens": 546652608.0, "step": 14403 }, { "epoch": 1.8323368528177078, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.302978515625, "learning_rate": 1e-05, "loss": 0.4485, "mean_token_accuracy": 0.8833414316177368, "num_tokens": 546695363.0, "step": 14404 }, { "epoch": 1.8324640630962983, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.497623443603516, "learning_rate": 1e-05, "loss": 0.5423, "mean_token_accuracy": 0.8548169136047363, "num_tokens": 546741174.0, "step": 14405 }, { "epoch": 1.8325912733748888, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 6.093525409698486, "learning_rate": 1e-05, "loss": 0.46, "mean_token_accuracy": 0.8780772686004639, "num_tokens": 546777827.0, "step": 14406 }, { "epoch": 1.8327184836534793, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.5530619621276855, "learning_rate": 1e-05, "loss": 0.4712, "mean_token_accuracy": 0.8775126934051514, "num_tokens": 546821199.0, "step": 14407 }, { "epoch": 1.8328456939320699, "ewc_loss": 0.10693359375, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.90783166885376, "learning_rate": 1e-05, "loss": 0.4862, "mean_token_accuracy": 0.870482325553894, "num_tokens": 546860151.0, "step": 14408 }, { "epoch": 1.8329729042106602, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.732942581176758, "learning_rate": 1e-05, "loss": 0.4746, "mean_token_accuracy": 0.8785799145698547, "num_tokens": 546892456.0, "step": 14409 }, { "epoch": 1.8331001144892507, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.744220733642578, "learning_rate": 1e-05, "loss": 0.5034, "mean_token_accuracy": 0.8625133633613586, "num_tokens": 546931611.0, "step": 14410 }, { "epoch": 1.8332273247678412, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 7.04522180557251, "learning_rate": 1e-05, "loss": 0.5025, "mean_token_accuracy": 0.8752337098121643, "num_tokens": 546968408.0, "step": 14411 }, { "epoch": 1.8333545350464318, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.725914478302002, "learning_rate": 1e-05, "loss": 0.518, "mean_token_accuracy": 0.8575783967971802, "num_tokens": 547007025.0, "step": 14412 }, { "epoch": 1.8334817453250223, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 0.00011682510375976562, "grad_norm": 7.380038261413574, "learning_rate": 1e-05, "loss": 0.5593, "mean_token_accuracy": 0.8566365838050842, "num_tokens": 547051371.0, "step": 14413 }, { "epoch": 1.8336089556036128, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 5.988838195800781, "learning_rate": 1e-05, "loss": 0.521, "mean_token_accuracy": 0.8632004261016846, "num_tokens": 547087476.0, "step": 14414 }, { "epoch": 1.8337361658822031, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 0.00011920928955078125, "grad_norm": 7.4416985511779785, "learning_rate": 1e-05, "loss": 0.5078, "mean_token_accuracy": 0.8746881484985352, "num_tokens": 547126922.0, "step": 14415 }, { "epoch": 1.8338633761607936, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.380620956420898, "learning_rate": 1e-05, "loss": 0.4919, "mean_token_accuracy": 0.8746515512466431, "num_tokens": 547164678.0, "step": 14416 }, { "epoch": 1.8339905864393842, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 7.086050987243652, "learning_rate": 1e-05, "loss": 0.5255, "mean_token_accuracy": 0.8686975240707397, "num_tokens": 547200039.0, "step": 14417 }, { "epoch": 1.8341177967179747, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.560347557067871, "learning_rate": 1e-05, "loss": 0.4426, "mean_token_accuracy": 0.8899964094161987, "num_tokens": 547233325.0, "step": 14418 }, { "epoch": 1.8342450069965652, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.530210971832275, "learning_rate": 1e-05, "loss": 0.4558, "mean_token_accuracy": 0.8835231065750122, "num_tokens": 547266658.0, "step": 14419 }, { "epoch": 1.8343722172751558, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.63374662399292, "learning_rate": 1e-05, "loss": 0.5416, "mean_token_accuracy": 0.8626810312271118, "num_tokens": 547305876.0, "step": 14420 }, { "epoch": 1.8344994275537463, "ewc_loss": 0.10986328125, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 6.024284839630127, "learning_rate": 1e-05, "loss": 0.4721, "mean_token_accuracy": 0.8762310743331909, "num_tokens": 547346847.0, "step": 14421 }, { "epoch": 1.8346266378323368, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.828434467315674, "learning_rate": 1e-05, "loss": 0.464, "mean_token_accuracy": 0.8814435601234436, "num_tokens": 547379896.0, "step": 14422 }, { "epoch": 1.8347538481109273, "ewc_loss": 0.10498046875, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.698271751403809, "learning_rate": 1e-05, "loss": 0.5042, "mean_token_accuracy": 0.864584743976593, "num_tokens": 547425666.0, "step": 14423 }, { "epoch": 1.8348810583895179, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 0.00011157989501953125, "grad_norm": 7.182754993438721, "learning_rate": 1e-05, "loss": 0.4856, "mean_token_accuracy": 0.8772091269493103, "num_tokens": 547467007.0, "step": 14424 }, { "epoch": 1.8350082686681084, "ewc_loss": 0.10693359375, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.691751003265381, "learning_rate": 1e-05, "loss": 0.5494, "mean_token_accuracy": 0.8501603603363037, "num_tokens": 547507718.0, "step": 14425 }, { "epoch": 1.835135478946699, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 0.00012159347534179688, "grad_norm": 7.559970378875732, "learning_rate": 1e-05, "loss": 0.5, "mean_token_accuracy": 0.8760089874267578, "num_tokens": 547543445.0, "step": 14426 }, { "epoch": 1.8352626892252895, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.060333251953125, "learning_rate": 1e-05, "loss": 0.4841, "mean_token_accuracy": 0.8730296492576599, "num_tokens": 547576000.0, "step": 14427 }, { "epoch": 1.83538989950388, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.590637683868408, "learning_rate": 1e-05, "loss": 0.5027, "mean_token_accuracy": 0.8750262260437012, "num_tokens": 547609950.0, "step": 14428 }, { "epoch": 1.8355171097824705, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.478689193725586, "learning_rate": 1e-05, "loss": 0.4788, "mean_token_accuracy": 0.8757582306861877, "num_tokens": 547650133.0, "step": 14429 }, { "epoch": 1.835644320061061, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 0.00011301040649414062, "grad_norm": 7.228306770324707, "learning_rate": 1e-05, "loss": 0.524, "mean_token_accuracy": 0.8665671348571777, "num_tokens": 547687754.0, "step": 14430 }, { "epoch": 1.8357715303396516, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.5550856590271, "learning_rate": 1e-05, "loss": 0.5024, "mean_token_accuracy": 0.873520016670227, "num_tokens": 547732119.0, "step": 14431 }, { "epoch": 1.835898740618242, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.835258483886719, "learning_rate": 1e-05, "loss": 0.4828, "mean_token_accuracy": 0.8769863843917847, "num_tokens": 547770779.0, "step": 14432 }, { "epoch": 1.8360259508968326, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.457529067993164, "learning_rate": 1e-05, "loss": 0.484, "mean_token_accuracy": 0.8772287368774414, "num_tokens": 547812524.0, "step": 14433 }, { "epoch": 1.836153161175423, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.557987689971924, "learning_rate": 1e-05, "loss": 0.5316, "mean_token_accuracy": 0.8642833232879639, "num_tokens": 547850955.0, "step": 14434 }, { "epoch": 1.8362803714540135, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.313802242279053, "learning_rate": 1e-05, "loss": 0.4976, "mean_token_accuracy": 0.8723645210266113, "num_tokens": 547891390.0, "step": 14435 }, { "epoch": 1.836407581732604, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.440122127532959, "learning_rate": 1e-05, "loss": 0.5273, "mean_token_accuracy": 0.8651345372200012, "num_tokens": 547930306.0, "step": 14436 }, { "epoch": 1.8365347920111945, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 6.130709648132324, "learning_rate": 1e-05, "loss": 0.5093, "mean_token_accuracy": 0.8624525666236877, "num_tokens": 547970940.0, "step": 14437 }, { "epoch": 1.836662002289785, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.460010051727295, "learning_rate": 1e-05, "loss": 0.4641, "mean_token_accuracy": 0.8780642747879028, "num_tokens": 548007908.0, "step": 14438 }, { "epoch": 1.8367892125683756, "ewc_loss": 0.10693359375, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.93996000289917, "learning_rate": 1e-05, "loss": 0.493, "mean_token_accuracy": 0.8656055927276611, "num_tokens": 548045040.0, "step": 14439 }, { "epoch": 1.8369164228469659, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.5859503746032715, "learning_rate": 1e-05, "loss": 0.4923, "mean_token_accuracy": 0.8711439371109009, "num_tokens": 548088501.0, "step": 14440 }, { "epoch": 1.8370436331255564, "ewc_loss": 0.10595703125, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.77636194229126, "learning_rate": 1e-05, "loss": 0.5113, "mean_token_accuracy": 0.8585707545280457, "num_tokens": 548121888.0, "step": 14441 }, { "epoch": 1.837170843404147, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.845416069030762, "learning_rate": 1e-05, "loss": 0.4901, "mean_token_accuracy": 0.8729649782180786, "num_tokens": 548161117.0, "step": 14442 }, { "epoch": 1.8372980536827375, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.658563613891602, "learning_rate": 1e-05, "loss": 0.4546, "mean_token_accuracy": 0.8803927898406982, "num_tokens": 548201709.0, "step": 14443 }, { "epoch": 1.837425263961328, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 0.00011348724365234375, "grad_norm": 7.225220680236816, "learning_rate": 1e-05, "loss": 0.512, "mean_token_accuracy": 0.8701766729354858, "num_tokens": 548241843.0, "step": 14444 }, { "epoch": 1.8375524742399185, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.774214267730713, "learning_rate": 1e-05, "loss": 0.4656, "mean_token_accuracy": 0.8764101266860962, "num_tokens": 548278785.0, "step": 14445 }, { "epoch": 1.837679684518509, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 0.00012111663818359375, "grad_norm": 7.516753673553467, "learning_rate": 1e-05, "loss": 0.5146, "mean_token_accuracy": 0.8770654201507568, "num_tokens": 548315766.0, "step": 14446 }, { "epoch": 1.8378068947970996, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.182028770446777, "learning_rate": 1e-05, "loss": 0.5131, "mean_token_accuracy": 0.8664984703063965, "num_tokens": 548353971.0, "step": 14447 }, { "epoch": 1.83793410507569, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 0.00011682510375976562, "grad_norm": 7.368005275726318, "learning_rate": 1e-05, "loss": 0.5163, "mean_token_accuracy": 0.8705724477767944, "num_tokens": 548391261.0, "step": 14448 }, { "epoch": 1.8380613153542806, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.514388561248779, "learning_rate": 1e-05, "loss": 0.4658, "mean_token_accuracy": 0.8817511200904846, "num_tokens": 548424773.0, "step": 14449 }, { "epoch": 1.8381885256328712, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.870231628417969, "learning_rate": 1e-05, "loss": 0.4471, "mean_token_accuracy": 0.8862714171409607, "num_tokens": 548462404.0, "step": 14450 }, { "epoch": 1.8383157359114617, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.59305477142334, "learning_rate": 1e-05, "loss": 0.499, "mean_token_accuracy": 0.8716531991958618, "num_tokens": 548504595.0, "step": 14451 }, { "epoch": 1.8384429461900522, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.333024501800537, "learning_rate": 1e-05, "loss": 0.4636, "mean_token_accuracy": 0.8787664175033569, "num_tokens": 548547945.0, "step": 14452 }, { "epoch": 1.8385701564686427, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.679114818572998, "learning_rate": 1e-05, "loss": 0.5395, "mean_token_accuracy": 0.8594160079956055, "num_tokens": 548590302.0, "step": 14453 }, { "epoch": 1.8386973667472333, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.910826206207275, "learning_rate": 1e-05, "loss": 0.4463, "mean_token_accuracy": 0.8808118104934692, "num_tokens": 548629688.0, "step": 14454 }, { "epoch": 1.8388245770258238, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.915152549743652, "learning_rate": 1e-05, "loss": 0.5698, "mean_token_accuracy": 0.8514596223831177, "num_tokens": 548666294.0, "step": 14455 }, { "epoch": 1.8389517873044143, "ewc_loss": 0.10595703125, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.708961009979248, "learning_rate": 1e-05, "loss": 0.5168, "mean_token_accuracy": 0.8583850860595703, "num_tokens": 548707536.0, "step": 14456 }, { "epoch": 1.8390789975830049, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 0.00011348724365234375, "grad_norm": 7.248035907745361, "learning_rate": 1e-05, "loss": 0.5046, "mean_token_accuracy": 0.8757632374763489, "num_tokens": 548751443.0, "step": 14457 }, { "epoch": 1.8392062078615952, "ewc_loss": 0.10986328125, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.800481796264648, "learning_rate": 1e-05, "loss": 0.4944, "mean_token_accuracy": 0.8682286739349365, "num_tokens": 548789259.0, "step": 14458 }, { "epoch": 1.8393334181401857, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 0.0001201629638671875, "grad_norm": 7.505334854125977, "learning_rate": 1e-05, "loss": 0.5044, "mean_token_accuracy": 0.8731274604797363, "num_tokens": 548822045.0, "step": 14459 }, { "epoch": 1.8394606284187762, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.163797855377197, "learning_rate": 1e-05, "loss": 0.537, "mean_token_accuracy": 0.8588310480117798, "num_tokens": 548854722.0, "step": 14460 }, { "epoch": 1.8395878386973668, "ewc_loss": 0.140625, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 0.00011777877807617188, "grad_norm": 7.410951614379883, "learning_rate": 1e-05, "loss": 0.5337, "mean_token_accuracy": 0.8640528917312622, "num_tokens": 548888903.0, "step": 14461 }, { "epoch": 1.8397150489759573, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.421164035797119, "learning_rate": 1e-05, "loss": 0.4946, "mean_token_accuracy": 0.8751566410064697, "num_tokens": 548925611.0, "step": 14462 }, { "epoch": 1.8398422592545478, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 7.07248067855835, "learning_rate": 1e-05, "loss": 0.4879, "mean_token_accuracy": 0.8776795864105225, "num_tokens": 548967239.0, "step": 14463 }, { "epoch": 1.8399694695331381, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.446615219116211, "learning_rate": 1e-05, "loss": 0.4871, "mean_token_accuracy": 0.8750529289245605, "num_tokens": 549005453.0, "step": 14464 }, { "epoch": 1.8400966798117286, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.751900672912598, "learning_rate": 1e-05, "loss": 0.5204, "mean_token_accuracy": 0.8679251670837402, "num_tokens": 549041418.0, "step": 14465 }, { "epoch": 1.8402238900903192, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.349841117858887, "learning_rate": 1e-05, "loss": 0.4758, "mean_token_accuracy": 0.8787029981613159, "num_tokens": 549079223.0, "step": 14466 }, { "epoch": 1.8403511003689097, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.535175800323486, "learning_rate": 1e-05, "loss": 0.5749, "mean_token_accuracy": 0.8476932048797607, "num_tokens": 549124542.0, "step": 14467 }, { "epoch": 1.8404783106475002, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.226308345794678, "learning_rate": 1e-05, "loss": 0.5192, "mean_token_accuracy": 0.8605504631996155, "num_tokens": 549165374.0, "step": 14468 }, { "epoch": 1.8406055209260908, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.458398342132568, "learning_rate": 1e-05, "loss": 0.5446, "mean_token_accuracy": 0.8578845858573914, "num_tokens": 549203887.0, "step": 14469 }, { "epoch": 1.8407327312046813, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.062565326690674, "learning_rate": 1e-05, "loss": 0.49, "mean_token_accuracy": 0.8684388399124146, "num_tokens": 549247583.0, "step": 14470 }, { "epoch": 1.8408599414832718, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.480388641357422, "learning_rate": 1e-05, "loss": 0.4426, "mean_token_accuracy": 0.8876677751541138, "num_tokens": 549282874.0, "step": 14471 }, { "epoch": 1.8409871517618623, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.910048007965088, "learning_rate": 1e-05, "loss": 0.4544, "mean_token_accuracy": 0.8778353333473206, "num_tokens": 549323133.0, "step": 14472 }, { "epoch": 1.8411143620404529, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.6269731521606445, "learning_rate": 1e-05, "loss": 0.4985, "mean_token_accuracy": 0.8711788654327393, "num_tokens": 549359794.0, "step": 14473 }, { "epoch": 1.8412415723190434, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.777244567871094, "learning_rate": 1e-05, "loss": 0.5498, "mean_token_accuracy": 0.8513134717941284, "num_tokens": 549393281.0, "step": 14474 }, { "epoch": 1.841368782597634, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.885462760925293, "learning_rate": 1e-05, "loss": 0.5353, "mean_token_accuracy": 0.8662348985671997, "num_tokens": 549428181.0, "step": 14475 }, { "epoch": 1.8414959928762245, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.706121921539307, "learning_rate": 1e-05, "loss": 0.4714, "mean_token_accuracy": 0.8767905235290527, "num_tokens": 549465521.0, "step": 14476 }, { "epoch": 1.841623203154815, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 0.0001125335693359375, "grad_norm": 7.185226917266846, "learning_rate": 1e-05, "loss": 0.5212, "mean_token_accuracy": 0.8699382543563843, "num_tokens": 549503800.0, "step": 14477 }, { "epoch": 1.8417504134334055, "ewc_loss": 0.109375, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.83506965637207, "learning_rate": 1e-05, "loss": 0.451, "mean_token_accuracy": 0.8803890943527222, "num_tokens": 549536725.0, "step": 14478 }, { "epoch": 1.841877623711996, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 0.000118255615234375, "grad_norm": 7.3779473304748535, "learning_rate": 1e-05, "loss": 0.5199, "mean_token_accuracy": 0.8715517520904541, "num_tokens": 549575774.0, "step": 14479 }, { "epoch": 1.8420048339905866, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.15394926071167, "learning_rate": 1e-05, "loss": 0.4943, "mean_token_accuracy": 0.8715498447418213, "num_tokens": 549616715.0, "step": 14480 }, { "epoch": 1.842132044269177, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 0.00011301040649414062, "grad_norm": 7.197036266326904, "learning_rate": 1e-05, "loss": 0.5184, "mean_token_accuracy": 0.8697342276573181, "num_tokens": 549652005.0, "step": 14481 }, { "epoch": 1.8422592545477676, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.41716194152832, "learning_rate": 1e-05, "loss": 0.5189, "mean_token_accuracy": 0.8681133985519409, "num_tokens": 549683708.0, "step": 14482 }, { "epoch": 1.842386464826358, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.745688438415527, "learning_rate": 1e-05, "loss": 0.4544, "mean_token_accuracy": 0.8852922916412354, "num_tokens": 549722961.0, "step": 14483 }, { "epoch": 1.8425136751049485, "ewc_loss": 0.125, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.503855228424072, "learning_rate": 1e-05, "loss": 0.5271, "mean_token_accuracy": 0.8628056645393372, "num_tokens": 549758857.0, "step": 14484 }, { "epoch": 1.842640885383539, "ewc_loss": 0.11474609375, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.240382671356201, "learning_rate": 1e-05, "loss": 0.4882, "mean_token_accuracy": 0.8694181442260742, "num_tokens": 549793835.0, "step": 14485 }, { "epoch": 1.8427680956621295, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.632811546325684, "learning_rate": 1e-05, "loss": 0.4937, "mean_token_accuracy": 0.8717201948165894, "num_tokens": 549832691.0, "step": 14486 }, { "epoch": 1.84289530594072, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.842288017272949, "learning_rate": 1e-05, "loss": 0.6215, "mean_token_accuracy": 0.833432674407959, "num_tokens": 549874981.0, "step": 14487 }, { "epoch": 1.8430225162193106, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.898322582244873, "learning_rate": 1e-05, "loss": 0.5063, "mean_token_accuracy": 0.8709274530410767, "num_tokens": 549917904.0, "step": 14488 }, { "epoch": 1.8431497264979009, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.638164043426514, "learning_rate": 1e-05, "loss": 0.4877, "mean_token_accuracy": 0.8686579465866089, "num_tokens": 549955421.0, "step": 14489 }, { "epoch": 1.8432769367764914, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 0.00011539459228515625, "grad_norm": 7.296386241912842, "learning_rate": 1e-05, "loss": 0.4975, "mean_token_accuracy": 0.8746008276939392, "num_tokens": 549986615.0, "step": 14490 }, { "epoch": 1.843404147055082, "ewc_loss": 0.109375, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.753198623657227, "learning_rate": 1e-05, "loss": 0.5138, "mean_token_accuracy": 0.8625329732894897, "num_tokens": 550022747.0, "step": 14491 }, { "epoch": 1.8435313573336725, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.603100776672363, "learning_rate": 1e-05, "loss": 0.5295, "mean_token_accuracy": 0.869697093963623, "num_tokens": 550060423.0, "step": 14492 }, { "epoch": 1.843658567612263, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.175350189208984, "learning_rate": 1e-05, "loss": 0.4979, "mean_token_accuracy": 0.8694186210632324, "num_tokens": 550102086.0, "step": 14493 }, { "epoch": 1.8437857778908535, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 0.00012159347534179688, "grad_norm": 7.489581108093262, "learning_rate": 1e-05, "loss": 0.4831, "mean_token_accuracy": 0.8850312232971191, "num_tokens": 550138281.0, "step": 14494 }, { "epoch": 1.843912988169444, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.499663829803467, "learning_rate": 1e-05, "loss": 0.5584, "mean_token_accuracy": 0.8549718856811523, "num_tokens": 550180219.0, "step": 14495 }, { "epoch": 1.8440401984480346, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002346038818359375, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 7.1174845695495605, "learning_rate": 1e-05, "loss": 0.5226, "mean_token_accuracy": 0.8703985214233398, "num_tokens": 550218095.0, "step": 14496 }, { "epoch": 1.844167408726625, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002346038818359375, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.506301403045654, "learning_rate": 1e-05, "loss": 0.5142, "mean_token_accuracy": 0.8653357028961182, "num_tokens": 550256080.0, "step": 14497 }, { "epoch": 1.8442946190052156, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002346038818359375, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.760643482208252, "learning_rate": 1e-05, "loss": 0.5511, "mean_token_accuracy": 0.8558984994888306, "num_tokens": 550297471.0, "step": 14498 }, { "epoch": 1.8444218292838062, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.397725582122803, "learning_rate": 1e-05, "loss": 0.4454, "mean_token_accuracy": 0.8878616094589233, "num_tokens": 550330495.0, "step": 14499 }, { "epoch": 1.8445490395623967, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.521346092224121, "learning_rate": 1e-05, "loss": 0.5048, "mean_token_accuracy": 0.8689749836921692, "num_tokens": 550362810.0, "step": 14500 }, { "epoch": 1.8446762498409872, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.2525787353515625, "learning_rate": 1e-05, "loss": 0.5009, "mean_token_accuracy": 0.8683196306228638, "num_tokens": 550409408.0, "step": 14501 }, { "epoch": 1.8448034601195777, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.434808731079102, "learning_rate": 1e-05, "loss": 0.5232, "mean_token_accuracy": 0.8633714318275452, "num_tokens": 550441366.0, "step": 14502 }, { "epoch": 1.8449306703981683, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.088327884674072, "learning_rate": 1e-05, "loss": 0.474, "mean_token_accuracy": 0.8763289451599121, "num_tokens": 550478247.0, "step": 14503 }, { "epoch": 1.8450578806767588, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.467539310455322, "learning_rate": 1e-05, "loss": 0.4844, "mean_token_accuracy": 0.876340389251709, "num_tokens": 550511861.0, "step": 14504 }, { "epoch": 1.8451850909553493, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.911131858825684, "learning_rate": 1e-05, "loss": 0.4987, "mean_token_accuracy": 0.8642945289611816, "num_tokens": 550554959.0, "step": 14505 }, { "epoch": 1.8453123012339399, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.614243507385254, "learning_rate": 1e-05, "loss": 0.4781, "mean_token_accuracy": 0.8747596144676208, "num_tokens": 550589960.0, "step": 14506 }, { "epoch": 1.8454395115125302, "ewc_loss": 0.10595703125, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.748260021209717, "learning_rate": 1e-05, "loss": 0.4859, "mean_token_accuracy": 0.86991286277771, "num_tokens": 550630481.0, "step": 14507 }, { "epoch": 1.8455667217911207, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.8944902420043945, "learning_rate": 1e-05, "loss": 0.5326, "mean_token_accuracy": 0.8606688380241394, "num_tokens": 550671819.0, "step": 14508 }, { "epoch": 1.8456939320697112, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.663182258605957, "learning_rate": 1e-05, "loss": 0.4579, "mean_token_accuracy": 0.8783217072486877, "num_tokens": 550712230.0, "step": 14509 }, { "epoch": 1.8458211423483017, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002346038818359375, "ewc_loss_parallel": 0.00011491775512695312, "grad_norm": 7.271005153656006, "learning_rate": 1e-05, "loss": 0.5224, "mean_token_accuracy": 0.8683293461799622, "num_tokens": 550751052.0, "step": 14510 }, { "epoch": 1.8459483526268923, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.0002346038818359375, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 5.808945178985596, "learning_rate": 1e-05, "loss": 0.473, "mean_token_accuracy": 0.8750788569450378, "num_tokens": 550787374.0, "step": 14511 }, { "epoch": 1.8460755629054828, "ewc_loss": 0.1455078125, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 0.00012159347534179688, "grad_norm": 7.522085189819336, "learning_rate": 1e-05, "loss": 0.4972, "mean_token_accuracy": 0.8773815631866455, "num_tokens": 550829398.0, "step": 14512 }, { "epoch": 1.846202773184073, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.226106643676758, "learning_rate": 1e-05, "loss": 0.5331, "mean_token_accuracy": 0.8618685603141785, "num_tokens": 550868818.0, "step": 14513 }, { "epoch": 1.8463299834626636, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 0.0001163482666015625, "grad_norm": 7.319212436676025, "learning_rate": 1e-05, "loss": 0.5103, "mean_token_accuracy": 0.8753571510314941, "num_tokens": 550904136.0, "step": 14514 }, { "epoch": 1.8464571937412542, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.519667625427246, "learning_rate": 1e-05, "loss": 0.6206, "mean_token_accuracy": 0.8363749384880066, "num_tokens": 550943505.0, "step": 14515 }, { "epoch": 1.8465844040198447, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.811835765838623, "learning_rate": 1e-05, "loss": 0.5027, "mean_token_accuracy": 0.869408369064331, "num_tokens": 550983633.0, "step": 14516 }, { "epoch": 1.8467116142984352, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002346038818359375, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.578363418579102, "learning_rate": 1e-05, "loss": 0.4762, "mean_token_accuracy": 0.8794339895248413, "num_tokens": 551017683.0, "step": 14517 }, { "epoch": 1.8468388245770258, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.0002346038818359375, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.303199291229248, "learning_rate": 1e-05, "loss": 0.4954, "mean_token_accuracy": 0.868643045425415, "num_tokens": 551051731.0, "step": 14518 }, { "epoch": 1.8469660348556163, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002346038818359375, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.663146495819092, "learning_rate": 1e-05, "loss": 0.4869, "mean_token_accuracy": 0.8754696249961853, "num_tokens": 551090818.0, "step": 14519 }, { "epoch": 1.8470932451342068, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.0002346038818359375, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.894327640533447, "learning_rate": 1e-05, "loss": 0.4566, "mean_token_accuracy": 0.87764972448349, "num_tokens": 551122032.0, "step": 14520 }, { "epoch": 1.8472204554127973, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002346038818359375, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.896661758422852, "learning_rate": 1e-05, "loss": 0.5335, "mean_token_accuracy": 0.8615903854370117, "num_tokens": 551157708.0, "step": 14521 }, { "epoch": 1.8473476656913879, "ewc_loss": 0.10595703125, "ewc_loss_diag": 0.0002346038818359375, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.690082550048828, "learning_rate": 1e-05, "loss": 0.48, "mean_token_accuracy": 0.8706913590431213, "num_tokens": 551194222.0, "step": 14522 }, { "epoch": 1.8474748759699784, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002346038818359375, "ewc_loss_parallel": 0.00011539459228515625, "grad_norm": 7.237610816955566, "learning_rate": 1e-05, "loss": 0.5292, "mean_token_accuracy": 0.8691341876983643, "num_tokens": 551234387.0, "step": 14523 }, { "epoch": 1.847602086248569, "ewc_loss": 0.10986328125, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.778644561767578, "learning_rate": 1e-05, "loss": 0.4802, "mean_token_accuracy": 0.8748106956481934, "num_tokens": 551273330.0, "step": 14524 }, { "epoch": 1.8477292965271594, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 0.00012111663818359375, "grad_norm": 7.500604629516602, "learning_rate": 1e-05, "loss": 0.5188, "mean_token_accuracy": 0.8709493279457092, "num_tokens": 551309180.0, "step": 14525 }, { "epoch": 1.84785650680575, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.1273417472839355, "learning_rate": 1e-05, "loss": 0.4598, "mean_token_accuracy": 0.881625771522522, "num_tokens": 551344455.0, "step": 14526 }, { "epoch": 1.8479837170843405, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 0.0001201629638671875, "grad_norm": 7.433637619018555, "learning_rate": 1e-05, "loss": 0.5475, "mean_token_accuracy": 0.8621712923049927, "num_tokens": 551380795.0, "step": 14527 }, { "epoch": 1.848110927362931, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.421402931213379, "learning_rate": 1e-05, "loss": 0.5449, "mean_token_accuracy": 0.8575984239578247, "num_tokens": 551421314.0, "step": 14528 }, { "epoch": 1.8482381376415216, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 7.090760231018066, "learning_rate": 1e-05, "loss": 0.5029, "mean_token_accuracy": 0.8724105358123779, "num_tokens": 551459021.0, "step": 14529 }, { "epoch": 1.848365347920112, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.460508346557617, "learning_rate": 1e-05, "loss": 0.5058, "mean_token_accuracy": 0.8691273927688599, "num_tokens": 551498055.0, "step": 14530 }, { "epoch": 1.8484925581987026, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.742696762084961, "learning_rate": 1e-05, "loss": 0.534, "mean_token_accuracy": 0.8646200299263, "num_tokens": 551536746.0, "step": 14531 }, { "epoch": 1.848619768477293, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.364237308502197, "learning_rate": 1e-05, "loss": 0.5223, "mean_token_accuracy": 0.859748899936676, "num_tokens": 551576230.0, "step": 14532 }, { "epoch": 1.8487469787558835, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.513700485229492, "learning_rate": 1e-05, "loss": 0.4956, "mean_token_accuracy": 0.8691703081130981, "num_tokens": 551613476.0, "step": 14533 }, { "epoch": 1.848874189034474, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.229592800140381, "learning_rate": 1e-05, "loss": 0.4906, "mean_token_accuracy": 0.8715459704399109, "num_tokens": 551655177.0, "step": 14534 }, { "epoch": 1.8490013993130645, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.432339191436768, "learning_rate": 1e-05, "loss": 0.5193, "mean_token_accuracy": 0.865302562713623, "num_tokens": 551691236.0, "step": 14535 }, { "epoch": 1.849128609591655, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.069087982177734, "learning_rate": 1e-05, "loss": 0.5223, "mean_token_accuracy": 0.8607648611068726, "num_tokens": 551732152.0, "step": 14536 }, { "epoch": 1.8492558198702456, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.472613334655762, "learning_rate": 1e-05, "loss": 0.475, "mean_token_accuracy": 0.8756694793701172, "num_tokens": 551765221.0, "step": 14537 }, { "epoch": 1.8493830301488359, "ewc_loss": 0.10791015625, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.908624172210693, "learning_rate": 1e-05, "loss": 0.4832, "mean_token_accuracy": 0.8701119422912598, "num_tokens": 551805321.0, "step": 14538 }, { "epoch": 1.8495102404274264, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.609071254730225, "learning_rate": 1e-05, "loss": 0.518, "mean_token_accuracy": 0.8623304963111877, "num_tokens": 551846752.0, "step": 14539 }, { "epoch": 1.849637450706017, "ewc_loss": 0.10693359375, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.764474391937256, "learning_rate": 1e-05, "loss": 0.4987, "mean_token_accuracy": 0.8670562505722046, "num_tokens": 551887877.0, "step": 14540 }, { "epoch": 1.8497646609846075, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.878704071044922, "learning_rate": 1e-05, "loss": 0.5121, "mean_token_accuracy": 0.862850546836853, "num_tokens": 551919669.0, "step": 14541 }, { "epoch": 1.849891871263198, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.673433303833008, "learning_rate": 1e-05, "loss": 0.4783, "mean_token_accuracy": 0.8707901239395142, "num_tokens": 551955459.0, "step": 14542 }, { "epoch": 1.8500190815417885, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 0.00011444091796875, "grad_norm": 7.23539924621582, "learning_rate": 1e-05, "loss": 0.5257, "mean_token_accuracy": 0.8675973415374756, "num_tokens": 551994889.0, "step": 14543 }, { "epoch": 1.850146291820379, "ewc_loss": 0.11083984375, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 5.807719707489014, "learning_rate": 1e-05, "loss": 0.4974, "mean_token_accuracy": 0.8717852830886841, "num_tokens": 552030176.0, "step": 14544 }, { "epoch": 1.8502735020989696, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 0.00012063980102539062, "grad_norm": 7.464605808258057, "learning_rate": 1e-05, "loss": 0.4758, "mean_token_accuracy": 0.882110595703125, "num_tokens": 552063317.0, "step": 14545 }, { "epoch": 1.85040071237756, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.175735950469971, "learning_rate": 1e-05, "loss": 0.4468, "mean_token_accuracy": 0.8874732851982117, "num_tokens": 552100959.0, "step": 14546 }, { "epoch": 1.8505279226561506, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 0.00011587142944335938, "grad_norm": 7.302454948425293, "learning_rate": 1e-05, "loss": 0.4985, "mean_token_accuracy": 0.8758047819137573, "num_tokens": 552139181.0, "step": 14547 }, { "epoch": 1.8506551329347412, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.481715202331543, "learning_rate": 1e-05, "loss": 0.509, "mean_token_accuracy": 0.8656262159347534, "num_tokens": 552172293.0, "step": 14548 }, { "epoch": 1.8507823432133317, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.804305553436279, "learning_rate": 1e-05, "loss": 0.5094, "mean_token_accuracy": 0.8666736483573914, "num_tokens": 552214655.0, "step": 14549 }, { "epoch": 1.8509095534919222, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.562517166137695, "learning_rate": 1e-05, "loss": 0.4487, "mean_token_accuracy": 0.8880332708358765, "num_tokens": 552254188.0, "step": 14550 }, { "epoch": 1.8510367637705127, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.296572208404541, "learning_rate": 1e-05, "loss": 0.557, "mean_token_accuracy": 0.8497453927993774, "num_tokens": 552286712.0, "step": 14551 }, { "epoch": 1.8511639740491033, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.664276123046875, "learning_rate": 1e-05, "loss": 0.5037, "mean_token_accuracy": 0.868513822555542, "num_tokens": 552326846.0, "step": 14552 }, { "epoch": 1.8512911843276938, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.881778240203857, "learning_rate": 1e-05, "loss": 0.4503, "mean_token_accuracy": 0.8807390928268433, "num_tokens": 552356753.0, "step": 14553 }, { "epoch": 1.8514183946062843, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.928923606872559, "learning_rate": 1e-05, "loss": 0.5391, "mean_token_accuracy": 0.8602126836776733, "num_tokens": 552391312.0, "step": 14554 }, { "epoch": 1.8515456048848749, "ewc_loss": 0.10498046875, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.684621334075928, "learning_rate": 1e-05, "loss": 0.5579, "mean_token_accuracy": 0.8455389142036438, "num_tokens": 552428782.0, "step": 14555 }, { "epoch": 1.8516728151634652, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 0.00011396408081054688, "grad_norm": 7.2690558433532715, "learning_rate": 1e-05, "loss": 0.5156, "mean_token_accuracy": 0.8685975074768066, "num_tokens": 552471131.0, "step": 14556 }, { "epoch": 1.8518000254420557, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.786056041717529, "learning_rate": 1e-05, "loss": 0.4797, "mean_token_accuracy": 0.8728765249252319, "num_tokens": 552512242.0, "step": 14557 }, { "epoch": 1.8519272357206462, "ewc_loss": 0.1455078125, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 0.00012159347534179688, "grad_norm": 7.542085647583008, "learning_rate": 1e-05, "loss": 0.5941, "mean_token_accuracy": 0.8490135669708252, "num_tokens": 552545655.0, "step": 14558 }, { "epoch": 1.8520544459992367, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.1615214347839355, "learning_rate": 1e-05, "loss": 0.5163, "mean_token_accuracy": 0.8683012127876282, "num_tokens": 552586340.0, "step": 14559 }, { "epoch": 1.8521816562778273, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 0.0001201629638671875, "grad_norm": 7.438633441925049, "learning_rate": 1e-05, "loss": 0.5186, "mean_token_accuracy": 0.8700995445251465, "num_tokens": 552624461.0, "step": 14560 }, { "epoch": 1.8523088665564178, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.449729919433594, "learning_rate": 1e-05, "loss": 0.4755, "mean_token_accuracy": 0.880862832069397, "num_tokens": 552661566.0, "step": 14561 }, { "epoch": 1.852436076835008, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 7.093738079071045, "learning_rate": 1e-05, "loss": 0.472, "mean_token_accuracy": 0.8813064694404602, "num_tokens": 552698184.0, "step": 14562 }, { "epoch": 1.8525632871135986, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.455471038818359, "learning_rate": 1e-05, "loss": 0.4528, "mean_token_accuracy": 0.8885200023651123, "num_tokens": 552741407.0, "step": 14563 }, { "epoch": 1.8526904973921892, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.766154766082764, "learning_rate": 1e-05, "loss": 0.5302, "mean_token_accuracy": 0.862000048160553, "num_tokens": 552780038.0, "step": 14564 }, { "epoch": 1.8528177076707797, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.365273475646973, "learning_rate": 1e-05, "loss": 0.4766, "mean_token_accuracy": 0.8759785294532776, "num_tokens": 552818479.0, "step": 14565 }, { "epoch": 1.8529449179493702, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.543599605560303, "learning_rate": 1e-05, "loss": 0.4729, "mean_token_accuracy": 0.8765972852706909, "num_tokens": 552853762.0, "step": 14566 }, { "epoch": 1.8530721282279607, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.2274861335754395, "learning_rate": 1e-05, "loss": 0.5648, "mean_token_accuracy": 0.843610942363739, "num_tokens": 552892693.0, "step": 14567 }, { "epoch": 1.8531993385065513, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.466928482055664, "learning_rate": 1e-05, "loss": 0.4771, "mean_token_accuracy": 0.8763792514801025, "num_tokens": 552930127.0, "step": 14568 }, { "epoch": 1.8533265487851418, "ewc_loss": 0.11181640625, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 6.064947605133057, "learning_rate": 1e-05, "loss": 0.4822, "mean_token_accuracy": 0.8708528280258179, "num_tokens": 552969809.0, "step": 14569 }, { "epoch": 1.8534537590637323, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.500885486602783, "learning_rate": 1e-05, "loss": 0.4804, "mean_token_accuracy": 0.8771020174026489, "num_tokens": 553004747.0, "step": 14570 }, { "epoch": 1.8535809693423229, "ewc_loss": 0.10791015625, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.892172336578369, "learning_rate": 1e-05, "loss": 0.5225, "mean_token_accuracy": 0.8578891754150391, "num_tokens": 553049727.0, "step": 14571 }, { "epoch": 1.8537081796209134, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.663353443145752, "learning_rate": 1e-05, "loss": 0.5036, "mean_token_accuracy": 0.8688639998435974, "num_tokens": 553093263.0, "step": 14572 }, { "epoch": 1.853835389899504, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.760456562042236, "learning_rate": 1e-05, "loss": 0.4893, "mean_token_accuracy": 0.8700774908065796, "num_tokens": 553126198.0, "step": 14573 }, { "epoch": 1.8539626001780944, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.933419227600098, "learning_rate": 1e-05, "loss": 0.4847, "mean_token_accuracy": 0.8748214244842529, "num_tokens": 553156798.0, "step": 14574 }, { "epoch": 1.854089810456685, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.713047504425049, "learning_rate": 1e-05, "loss": 0.4682, "mean_token_accuracy": 0.872859537601471, "num_tokens": 553193612.0, "step": 14575 }, { "epoch": 1.8542170207352755, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 0.00011491775512695312, "grad_norm": 7.236398220062256, "learning_rate": 1e-05, "loss": 0.491, "mean_token_accuracy": 0.8749710321426392, "num_tokens": 553227376.0, "step": 14576 }, { "epoch": 1.854344231013866, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 5.875420093536377, "learning_rate": 1e-05, "loss": 0.4978, "mean_token_accuracy": 0.8670745491981506, "num_tokens": 553263045.0, "step": 14577 }, { "epoch": 1.8544714412924566, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 0.00011777877807617188, "grad_norm": 7.37445592880249, "learning_rate": 1e-05, "loss": 0.5423, "mean_token_accuracy": 0.862997829914093, "num_tokens": 553301834.0, "step": 14578 }, { "epoch": 1.854598651571047, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.2223663330078125, "learning_rate": 1e-05, "loss": 0.5365, "mean_token_accuracy": 0.8588196039199829, "num_tokens": 553345980.0, "step": 14579 }, { "epoch": 1.8547258618496376, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 0.00011110305786132812, "grad_norm": 7.133028984069824, "learning_rate": 1e-05, "loss": 0.5356, "mean_token_accuracy": 0.8661187887191772, "num_tokens": 553380645.0, "step": 14580 }, { "epoch": 1.854853072128228, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.452320575714111, "learning_rate": 1e-05, "loss": 0.5232, "mean_token_accuracy": 0.8619048595428467, "num_tokens": 553417952.0, "step": 14581 }, { "epoch": 1.8549802824068184, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.64339017868042, "learning_rate": 1e-05, "loss": 0.4891, "mean_token_accuracy": 0.874805748462677, "num_tokens": 553454575.0, "step": 14582 }, { "epoch": 1.855107492685409, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.546207427978516, "learning_rate": 1e-05, "loss": 0.4361, "mean_token_accuracy": 0.8906401991844177, "num_tokens": 553489846.0, "step": 14583 }, { "epoch": 1.8552347029639995, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.134576797485352, "learning_rate": 1e-05, "loss": 0.4641, "mean_token_accuracy": 0.8773490190505981, "num_tokens": 553524444.0, "step": 14584 }, { "epoch": 1.85536191324259, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.719549655914307, "learning_rate": 1e-05, "loss": 0.4929, "mean_token_accuracy": 0.8750529289245605, "num_tokens": 553562295.0, "step": 14585 }, { "epoch": 1.8554891235211806, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.7559990882873535, "learning_rate": 1e-05, "loss": 0.5345, "mean_token_accuracy": 0.8543238639831543, "num_tokens": 553600240.0, "step": 14586 }, { "epoch": 1.8556163337997709, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 7.042534828186035, "learning_rate": 1e-05, "loss": 0.4576, "mean_token_accuracy": 0.8856759071350098, "num_tokens": 553641215.0, "step": 14587 }, { "epoch": 1.8557435440783614, "ewc_loss": 0.10498046875, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.6322760581970215, "learning_rate": 1e-05, "loss": 0.4379, "mean_token_accuracy": 0.8848264813423157, "num_tokens": 553675599.0, "step": 14588 }, { "epoch": 1.855870754356952, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 0.00011968612670898438, "grad_norm": 7.466454029083252, "learning_rate": 1e-05, "loss": 0.4747, "mean_token_accuracy": 0.8823457956314087, "num_tokens": 553706681.0, "step": 14589 }, { "epoch": 1.8559979646355425, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 5.902674674987793, "learning_rate": 1e-05, "loss": 0.4825, "mean_token_accuracy": 0.8728764057159424, "num_tokens": 553740919.0, "step": 14590 }, { "epoch": 1.856125174914133, "ewc_loss": 0.1484375, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 0.00012493133544921875, "grad_norm": 7.638580799102783, "learning_rate": 1e-05, "loss": 0.5176, "mean_token_accuracy": 0.8748816847801208, "num_tokens": 553775509.0, "step": 14591 }, { "epoch": 1.8562523851927235, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.371992588043213, "learning_rate": 1e-05, "loss": 0.5284, "mean_token_accuracy": 0.8611650466918945, "num_tokens": 553809168.0, "step": 14592 }, { "epoch": 1.856379595471314, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 0.000118255615234375, "grad_norm": 7.3566412925720215, "learning_rate": 1e-05, "loss": 0.4821, "mean_token_accuracy": 0.8824851512908936, "num_tokens": 553848186.0, "step": 14593 }, { "epoch": 1.8565068057499046, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.544912338256836, "learning_rate": 1e-05, "loss": 0.502, "mean_token_accuracy": 0.8738609552383423, "num_tokens": 553888928.0, "step": 14594 }, { "epoch": 1.856634016028495, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.966626167297363, "learning_rate": 1e-05, "loss": 0.5448, "mean_token_accuracy": 0.8581056594848633, "num_tokens": 553929095.0, "step": 14595 }, { "epoch": 1.8567612263070856, "ewc_loss": 0.125, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.476067066192627, "learning_rate": 1e-05, "loss": 0.4803, "mean_token_accuracy": 0.8781826496124268, "num_tokens": 553969610.0, "step": 14596 }, { "epoch": 1.8568884365856761, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.648538589477539, "learning_rate": 1e-05, "loss": 0.5142, "mean_token_accuracy": 0.871222734451294, "num_tokens": 554006707.0, "step": 14597 }, { "epoch": 1.8570156468642667, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.350071907043457, "learning_rate": 1e-05, "loss": 0.5258, "mean_token_accuracy": 0.8605496883392334, "num_tokens": 554049006.0, "step": 14598 }, { "epoch": 1.8571428571428572, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.468075275421143, "learning_rate": 1e-05, "loss": 0.483, "mean_token_accuracy": 0.8732609152793884, "num_tokens": 554086682.0, "step": 14599 }, { "epoch": 1.8572700674214477, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.195021629333496, "learning_rate": 1e-05, "loss": 0.5058, "mean_token_accuracy": 0.8655015826225281, "num_tokens": 554122478.0, "step": 14600 }, { "epoch": 1.8573972777000383, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.430796146392822, "learning_rate": 1e-05, "loss": 0.4718, "mean_token_accuracy": 0.8767755031585693, "num_tokens": 554162640.0, "step": 14601 }, { "epoch": 1.8575244879786288, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 6.024849891662598, "learning_rate": 1e-05, "loss": 0.5067, "mean_token_accuracy": 0.8636679649353027, "num_tokens": 554202161.0, "step": 14602 }, { "epoch": 1.8576516982572193, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.512664318084717, "learning_rate": 1e-05, "loss": 0.4881, "mean_token_accuracy": 0.8748946189880371, "num_tokens": 554235401.0, "step": 14603 }, { "epoch": 1.8577789085358098, "ewc_loss": 0.10693359375, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.8480963706970215, "learning_rate": 1e-05, "loss": 0.5571, "mean_token_accuracy": 0.8507000207901001, "num_tokens": 554276887.0, "step": 14604 }, { "epoch": 1.8579061188144002, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.737479209899902, "learning_rate": 1e-05, "loss": 0.5576, "mean_token_accuracy": 0.8528514504432678, "num_tokens": 554310513.0, "step": 14605 }, { "epoch": 1.8580333290929907, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.691249847412109, "learning_rate": 1e-05, "loss": 0.4686, "mean_token_accuracy": 0.8730158805847168, "num_tokens": 554354225.0, "step": 14606 }, { "epoch": 1.8581605393715812, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 0.00010967254638671875, "grad_norm": 7.087778568267822, "learning_rate": 1e-05, "loss": 0.5353, "mean_token_accuracy": 0.8632031083106995, "num_tokens": 554395198.0, "step": 14607 }, { "epoch": 1.8582877496501717, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.700418949127197, "learning_rate": 1e-05, "loss": 0.514, "mean_token_accuracy": 0.8594423532485962, "num_tokens": 554430234.0, "step": 14608 }, { "epoch": 1.8584149599287623, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 0.00011873245239257812, "grad_norm": 7.43537712097168, "learning_rate": 1e-05, "loss": 0.5271, "mean_token_accuracy": 0.8692502975463867, "num_tokens": 554469381.0, "step": 14609 }, { "epoch": 1.8585421702073528, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.01289701461792, "learning_rate": 1e-05, "loss": 0.4929, "mean_token_accuracy": 0.8656374216079712, "num_tokens": 554500409.0, "step": 14610 }, { "epoch": 1.858669380485943, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 0.00012063980102539062, "grad_norm": 7.458099365234375, "learning_rate": 1e-05, "loss": 0.4967, "mean_token_accuracy": 0.878122866153717, "num_tokens": 554537785.0, "step": 14611 }, { "epoch": 1.8587965907645336, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.395732879638672, "learning_rate": 1e-05, "loss": 0.5377, "mean_token_accuracy": 0.8608781099319458, "num_tokens": 554578420.0, "step": 14612 }, { "epoch": 1.8589238010431242, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 0.00010967254638671875, "grad_norm": 7.080875873565674, "learning_rate": 1e-05, "loss": 0.5169, "mean_token_accuracy": 0.8719969391822815, "num_tokens": 554615264.0, "step": 14613 }, { "epoch": 1.8590510113217147, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.558620929718018, "learning_rate": 1e-05, "loss": 0.5067, "mean_token_accuracy": 0.8699914813041687, "num_tokens": 554647327.0, "step": 14614 }, { "epoch": 1.8591782216003052, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.562839984893799, "learning_rate": 1e-05, "loss": 0.4869, "mean_token_accuracy": 0.8720731735229492, "num_tokens": 554682552.0, "step": 14615 }, { "epoch": 1.8593054318788957, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.597105979919434, "learning_rate": 1e-05, "loss": 0.5333, "mean_token_accuracy": 0.8622918128967285, "num_tokens": 554721348.0, "step": 14616 }, { "epoch": 1.8594326421574863, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 6.083685398101807, "learning_rate": 1e-05, "loss": 0.459, "mean_token_accuracy": 0.879068911075592, "num_tokens": 554760618.0, "step": 14617 }, { "epoch": 1.8595598524360768, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.757486343383789, "learning_rate": 1e-05, "loss": 0.527, "mean_token_accuracy": 0.8626121282577515, "num_tokens": 554801831.0, "step": 14618 }, { "epoch": 1.8596870627146673, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.771765232086182, "learning_rate": 1e-05, "loss": 0.4716, "mean_token_accuracy": 0.8751023411750793, "num_tokens": 554838307.0, "step": 14619 }, { "epoch": 1.8598142729932579, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 7.052648544311523, "learning_rate": 1e-05, "loss": 0.519, "mean_token_accuracy": 0.865608811378479, "num_tokens": 554883833.0, "step": 14620 }, { "epoch": 1.8599414832718484, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.693421840667725, "learning_rate": 1e-05, "loss": 0.4889, "mean_token_accuracy": 0.8684741258621216, "num_tokens": 554920944.0, "step": 14621 }, { "epoch": 1.860068693550439, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 0.000118255615234375, "grad_norm": 7.404862880706787, "learning_rate": 1e-05, "loss": 0.5377, "mean_token_accuracy": 0.8633303642272949, "num_tokens": 554958656.0, "step": 14622 }, { "epoch": 1.8601959038290294, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 5.928370475769043, "learning_rate": 1e-05, "loss": 0.5141, "mean_token_accuracy": 0.8633365631103516, "num_tokens": 555001864.0, "step": 14623 }, { "epoch": 1.86032311410762, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 0.00012111663818359375, "grad_norm": 7.533304691314697, "learning_rate": 1e-05, "loss": 0.5087, "mean_token_accuracy": 0.8743693232536316, "num_tokens": 555034268.0, "step": 14624 }, { "epoch": 1.8604503243862105, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.30608606338501, "learning_rate": 1e-05, "loss": 0.5178, "mean_token_accuracy": 0.8655145168304443, "num_tokens": 555075275.0, "step": 14625 }, { "epoch": 1.860577534664801, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 0.0001163482666015625, "grad_norm": 7.310803413391113, "learning_rate": 1e-05, "loss": 0.6265, "mean_token_accuracy": 0.8399760723114014, "num_tokens": 555110961.0, "step": 14626 }, { "epoch": 1.8607047449433916, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.471175193786621, "learning_rate": 1e-05, "loss": 0.5369, "mean_token_accuracy": 0.8552792072296143, "num_tokens": 555150369.0, "step": 14627 }, { "epoch": 1.860831955221982, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.953392505645752, "learning_rate": 1e-05, "loss": 0.5471, "mean_token_accuracy": 0.8603793382644653, "num_tokens": 555182555.0, "step": 14628 }, { "epoch": 1.8609591655005726, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.422325134277344, "learning_rate": 1e-05, "loss": 0.4672, "mean_token_accuracy": 0.8825509548187256, "num_tokens": 555223267.0, "step": 14629 }, { "epoch": 1.861086375779163, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.65410852432251, "learning_rate": 1e-05, "loss": 0.5078, "mean_token_accuracy": 0.8671876788139343, "num_tokens": 555263461.0, "step": 14630 }, { "epoch": 1.8612135860577534, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.3165812492370605, "learning_rate": 1e-05, "loss": 0.499, "mean_token_accuracy": 0.8687502145767212, "num_tokens": 555306201.0, "step": 14631 }, { "epoch": 1.861340796336344, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.485555171966553, "learning_rate": 1e-05, "loss": 0.4973, "mean_token_accuracy": 0.8704586029052734, "num_tokens": 555344418.0, "step": 14632 }, { "epoch": 1.8614680066149345, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.179107666015625, "learning_rate": 1e-05, "loss": 0.5108, "mean_token_accuracy": 0.867678701877594, "num_tokens": 555386096.0, "step": 14633 }, { "epoch": 1.861595216893525, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.457289218902588, "learning_rate": 1e-05, "loss": 0.4973, "mean_token_accuracy": 0.8744128942489624, "num_tokens": 555423368.0, "step": 14634 }, { "epoch": 1.8617224271721156, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 6.01143741607666, "learning_rate": 1e-05, "loss": 0.4934, "mean_token_accuracy": 0.8641219139099121, "num_tokens": 555458624.0, "step": 14635 }, { "epoch": 1.8618496374507059, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.54469108581543, "learning_rate": 1e-05, "loss": 0.4862, "mean_token_accuracy": 0.877116858959198, "num_tokens": 555491433.0, "step": 14636 }, { "epoch": 1.8619768477292964, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.856932163238525, "learning_rate": 1e-05, "loss": 0.4296, "mean_token_accuracy": 0.8861932754516602, "num_tokens": 555527198.0, "step": 14637 }, { "epoch": 1.862104058007887, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.728715419769287, "learning_rate": 1e-05, "loss": 0.4734, "mean_token_accuracy": 0.8771436810493469, "num_tokens": 555563232.0, "step": 14638 }, { "epoch": 1.8622312682864774, "ewc_loss": 0.10595703125, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.7403974533081055, "learning_rate": 1e-05, "loss": 0.5235, "mean_token_accuracy": 0.8559871912002563, "num_tokens": 555599805.0, "step": 14639 }, { "epoch": 1.862358478565068, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 7.0398383140563965, "learning_rate": 1e-05, "loss": 0.5228, "mean_token_accuracy": 0.8634845018386841, "num_tokens": 555628835.0, "step": 14640 }, { "epoch": 1.8624856888436585, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.7284674644470215, "learning_rate": 1e-05, "loss": 0.4586, "mean_token_accuracy": 0.8785419464111328, "num_tokens": 555668120.0, "step": 14641 }, { "epoch": 1.862612899122249, "ewc_loss": 0.140625, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 0.00011730194091796875, "grad_norm": 7.352502346038818, "learning_rate": 1e-05, "loss": 0.5675, "mean_token_accuracy": 0.8545347452163696, "num_tokens": 555704662.0, "step": 14642 }, { "epoch": 1.8627401094008396, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 5.986586093902588, "learning_rate": 1e-05, "loss": 0.4667, "mean_token_accuracy": 0.8802106976509094, "num_tokens": 555738867.0, "step": 14643 }, { "epoch": 1.86286731967943, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 0.000118255615234375, "grad_norm": 7.37790584564209, "learning_rate": 1e-05, "loss": 0.5391, "mean_token_accuracy": 0.8647884130477905, "num_tokens": 555773500.0, "step": 14644 }, { "epoch": 1.8629945299580206, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.3335981369018555, "learning_rate": 1e-05, "loss": 0.4562, "mean_token_accuracy": 0.8838717341423035, "num_tokens": 555812139.0, "step": 14645 }, { "epoch": 1.8631217402366111, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 7.030988693237305, "learning_rate": 1e-05, "loss": 0.5441, "mean_token_accuracy": 0.8581528067588806, "num_tokens": 555850823.0, "step": 14646 }, { "epoch": 1.8632489505152017, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.496219635009766, "learning_rate": 1e-05, "loss": 0.443, "mean_token_accuracy": 0.888393759727478, "num_tokens": 555888554.0, "step": 14647 }, { "epoch": 1.8633761607937922, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.501860618591309, "learning_rate": 1e-05, "loss": 0.5158, "mean_token_accuracy": 0.8678838014602661, "num_tokens": 555932629.0, "step": 14648 }, { "epoch": 1.8635033710723827, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.600124835968018, "learning_rate": 1e-05, "loss": 0.5344, "mean_token_accuracy": 0.8616761565208435, "num_tokens": 555970722.0, "step": 14649 }, { "epoch": 1.8636305813509733, "ewc_loss": 0.109375, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 6.021511077880859, "learning_rate": 1e-05, "loss": 0.4602, "mean_token_accuracy": 0.8774617910385132, "num_tokens": 556005784.0, "step": 14650 }, { "epoch": 1.8637577916295638, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.811390399932861, "learning_rate": 1e-05, "loss": 0.4807, "mean_token_accuracy": 0.8774346113204956, "num_tokens": 556046053.0, "step": 14651 }, { "epoch": 1.8638850019081543, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.704647064208984, "learning_rate": 1e-05, "loss": 0.4951, "mean_token_accuracy": 0.8644754886627197, "num_tokens": 556084265.0, "step": 14652 }, { "epoch": 1.8640122121867448, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 0.00011205673217773438, "grad_norm": 7.164885997772217, "learning_rate": 1e-05, "loss": 0.4986, "mean_token_accuracy": 0.8755706548690796, "num_tokens": 556125062.0, "step": 14653 }, { "epoch": 1.8641394224653351, "ewc_loss": 0.10693359375, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.69180965423584, "learning_rate": 1e-05, "loss": 0.5304, "mean_token_accuracy": 0.8557661771774292, "num_tokens": 556162590.0, "step": 14654 }, { "epoch": 1.8642666327439257, "ewc_loss": 0.1455078125, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 0.0001220703125, "grad_norm": 7.539837837219238, "learning_rate": 1e-05, "loss": 0.5295, "mean_token_accuracy": 0.866123616695404, "num_tokens": 556200560.0, "step": 14655 }, { "epoch": 1.8643938430225162, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.036684513092041, "learning_rate": 1e-05, "loss": 0.4844, "mean_token_accuracy": 0.8768820762634277, "num_tokens": 556237199.0, "step": 14656 }, { "epoch": 1.8645210533011067, "ewc_loss": 0.1455078125, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 0.0001220703125, "grad_norm": 7.58805513381958, "learning_rate": 1e-05, "loss": 0.5264, "mean_token_accuracy": 0.8672941327095032, "num_tokens": 556278620.0, "step": 14657 }, { "epoch": 1.8646482635796973, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.447519302368164, "learning_rate": 1e-05, "loss": 0.4822, "mean_token_accuracy": 0.87510085105896, "num_tokens": 556312844.0, "step": 14658 }, { "epoch": 1.8647754738582878, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 0.00011348724365234375, "grad_norm": 7.248534679412842, "learning_rate": 1e-05, "loss": 0.4987, "mean_token_accuracy": 0.8770197629928589, "num_tokens": 556351218.0, "step": 14659 }, { "epoch": 1.864902684136878, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.53776741027832, "learning_rate": 1e-05, "loss": 0.4805, "mean_token_accuracy": 0.8801208138465881, "num_tokens": 556388176.0, "step": 14660 }, { "epoch": 1.8650298944154686, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.87158727645874, "learning_rate": 1e-05, "loss": 0.5172, "mean_token_accuracy": 0.8641592264175415, "num_tokens": 556424996.0, "step": 14661 }, { "epoch": 1.8651571046940592, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.453680515289307, "learning_rate": 1e-05, "loss": 0.5722, "mean_token_accuracy": 0.8511229753494263, "num_tokens": 556465961.0, "step": 14662 }, { "epoch": 1.8652843149726497, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.573479652404785, "learning_rate": 1e-05, "loss": 0.4869, "mean_token_accuracy": 0.8717643022537231, "num_tokens": 556504777.0, "step": 14663 }, { "epoch": 1.8654115252512402, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.328813552856445, "learning_rate": 1e-05, "loss": 0.4699, "mean_token_accuracy": 0.875967264175415, "num_tokens": 556539737.0, "step": 14664 }, { "epoch": 1.8655387355298307, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.423407554626465, "learning_rate": 1e-05, "loss": 0.4663, "mean_token_accuracy": 0.883007287979126, "num_tokens": 556578028.0, "step": 14665 }, { "epoch": 1.8656659458084213, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.207608222961426, "learning_rate": 1e-05, "loss": 0.5325, "mean_token_accuracy": 0.8589104413986206, "num_tokens": 556611362.0, "step": 14666 }, { "epoch": 1.8657931560870118, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.400587558746338, "learning_rate": 1e-05, "loss": 0.5196, "mean_token_accuracy": 0.8680110573768616, "num_tokens": 556652329.0, "step": 14667 }, { "epoch": 1.8659203663656023, "ewc_loss": 0.11083984375, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 6.015325546264648, "learning_rate": 1e-05, "loss": 0.4173, "mean_token_accuracy": 0.8933653831481934, "num_tokens": 556689531.0, "step": 14668 }, { "epoch": 1.8660475766441929, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.475790500640869, "learning_rate": 1e-05, "loss": 0.4786, "mean_token_accuracy": 0.879004955291748, "num_tokens": 556722309.0, "step": 14669 }, { "epoch": 1.8661747869227834, "ewc_loss": 0.10693359375, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.854625225067139, "learning_rate": 1e-05, "loss": 0.4525, "mean_token_accuracy": 0.8806567788124084, "num_tokens": 556755682.0, "step": 14670 }, { "epoch": 1.866301997201374, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.6687822341918945, "learning_rate": 1e-05, "loss": 0.4743, "mean_token_accuracy": 0.8791021108627319, "num_tokens": 556792591.0, "step": 14671 }, { "epoch": 1.8664292074799644, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.6800055503845215, "learning_rate": 1e-05, "loss": 0.4688, "mean_token_accuracy": 0.8748881816864014, "num_tokens": 556835168.0, "step": 14672 }, { "epoch": 1.866556417758555, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 7.019758701324463, "learning_rate": 1e-05, "loss": 0.495, "mean_token_accuracy": 0.8728219270706177, "num_tokens": 556876106.0, "step": 14673 }, { "epoch": 1.8666836280371455, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.643283367156982, "learning_rate": 1e-05, "loss": 0.4873, "mean_token_accuracy": 0.8668262958526611, "num_tokens": 556917154.0, "step": 14674 }, { "epoch": 1.866810838315736, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 0.0001201629638671875, "grad_norm": 7.440978527069092, "learning_rate": 1e-05, "loss": 0.5364, "mean_token_accuracy": 0.8656498789787292, "num_tokens": 556953706.0, "step": 14675 }, { "epoch": 1.8669380485943265, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 5.89909553527832, "learning_rate": 1e-05, "loss": 0.4833, "mean_token_accuracy": 0.8725447058677673, "num_tokens": 556992035.0, "step": 14676 }, { "epoch": 1.867065258872917, "ewc_loss": 0.1474609375, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 0.0001239776611328125, "grad_norm": 7.631880283355713, "learning_rate": 1e-05, "loss": 0.5182, "mean_token_accuracy": 0.8728858232498169, "num_tokens": 557034785.0, "step": 14677 }, { "epoch": 1.8671924691515076, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.388790607452393, "learning_rate": 1e-05, "loss": 0.5264, "mean_token_accuracy": 0.8621870875358582, "num_tokens": 557074910.0, "step": 14678 }, { "epoch": 1.867319679430098, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 0.00011539459228515625, "grad_norm": 7.308063507080078, "learning_rate": 1e-05, "loss": 0.5027, "mean_token_accuracy": 0.8713816404342651, "num_tokens": 557105327.0, "step": 14679 }, { "epoch": 1.8674468897086884, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 6.6092424392700195, "learning_rate": 1e-05, "loss": 0.5187, "mean_token_accuracy": 0.8673537969589233, "num_tokens": 557149261.0, "step": 14680 }, { "epoch": 1.867574099987279, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.778743267059326, "learning_rate": 1e-05, "loss": 0.5147, "mean_token_accuracy": 0.8660831451416016, "num_tokens": 557190550.0, "step": 14681 }, { "epoch": 1.8677013102658695, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.625906467437744, "learning_rate": 1e-05, "loss": 0.5316, "mean_token_accuracy": 0.8558287620544434, "num_tokens": 557222828.0, "step": 14682 }, { "epoch": 1.86782852054446, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.286848545074463, "learning_rate": 1e-05, "loss": 0.5175, "mean_token_accuracy": 0.8640989065170288, "num_tokens": 557264228.0, "step": 14683 }, { "epoch": 1.8679557308230506, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.677628040313721, "learning_rate": 1e-05, "loss": 0.4548, "mean_token_accuracy": 0.8882859349250793, "num_tokens": 557301310.0, "step": 14684 }, { "epoch": 1.8680829411016409, "ewc_loss": 0.10888671875, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.945553779602051, "learning_rate": 1e-05, "loss": 0.4832, "mean_token_accuracy": 0.8749285340309143, "num_tokens": 557338596.0, "step": 14685 }, { "epoch": 1.8682101513802314, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.847337245941162, "learning_rate": 1e-05, "loss": 0.5056, "mean_token_accuracy": 0.8686572313308716, "num_tokens": 557374939.0, "step": 14686 }, { "epoch": 1.868337361658822, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.772454738616943, "learning_rate": 1e-05, "loss": 0.4926, "mean_token_accuracy": 0.8697590827941895, "num_tokens": 557411711.0, "step": 14687 }, { "epoch": 1.8684645719374124, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 0.00010967254638671875, "grad_norm": 7.091046333312988, "learning_rate": 1e-05, "loss": 0.5106, "mean_token_accuracy": 0.8716514110565186, "num_tokens": 557448892.0, "step": 14688 }, { "epoch": 1.868591782216003, "ewc_loss": 0.10791015625, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.771760940551758, "learning_rate": 1e-05, "loss": 0.5037, "mean_token_accuracy": 0.8647345304489136, "num_tokens": 557490911.0, "step": 14689 }, { "epoch": 1.8687189924945935, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 0.00011587142944335938, "grad_norm": 7.345130443572998, "learning_rate": 1e-05, "loss": 0.5162, "mean_token_accuracy": 0.8726178407669067, "num_tokens": 557529721.0, "step": 14690 }, { "epoch": 1.868846202773184, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 5.991351127624512, "learning_rate": 1e-05, "loss": 0.4759, "mean_token_accuracy": 0.8746212124824524, "num_tokens": 557567591.0, "step": 14691 }, { "epoch": 1.8689734130517746, "ewc_loss": 0.140625, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 0.00011682510375976562, "grad_norm": 7.391265392303467, "learning_rate": 1e-05, "loss": 0.5358, "mean_token_accuracy": 0.8666322231292725, "num_tokens": 557609105.0, "step": 14692 }, { "epoch": 1.869100623330365, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.264558792114258, "learning_rate": 1e-05, "loss": 0.604, "mean_token_accuracy": 0.8378020524978638, "num_tokens": 557658798.0, "step": 14693 }, { "epoch": 1.8692278336089556, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 0.00011301040649414062, "grad_norm": 7.183322429656982, "learning_rate": 1e-05, "loss": 0.4662, "mean_token_accuracy": 0.8861907720565796, "num_tokens": 557695572.0, "step": 14694 }, { "epoch": 1.8693550438875461, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.382736682891846, "learning_rate": 1e-05, "loss": 0.5093, "mean_token_accuracy": 0.8723402619361877, "num_tokens": 557728796.0, "step": 14695 }, { "epoch": 1.8694822541661367, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.8620500564575195, "learning_rate": 1e-05, "loss": 0.4636, "mean_token_accuracy": 0.8822425603866577, "num_tokens": 557761786.0, "step": 14696 }, { "epoch": 1.8696094644447272, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.351032257080078, "learning_rate": 1e-05, "loss": 0.5271, "mean_token_accuracy": 0.8625084161758423, "num_tokens": 557809454.0, "step": 14697 }, { "epoch": 1.8697366747233177, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.580451488494873, "learning_rate": 1e-05, "loss": 0.4279, "mean_token_accuracy": 0.891264796257019, "num_tokens": 557846959.0, "step": 14698 }, { "epoch": 1.8698638850019083, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.278920650482178, "learning_rate": 1e-05, "loss": 0.4715, "mean_token_accuracy": 0.8811488151550293, "num_tokens": 557879873.0, "step": 14699 }, { "epoch": 1.8699910952804988, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.41356897354126, "learning_rate": 1e-05, "loss": 0.5293, "mean_token_accuracy": 0.8646969795227051, "num_tokens": 557918020.0, "step": 14700 }, { "epoch": 1.8701183055590893, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.169591903686523, "learning_rate": 1e-05, "loss": 0.5039, "mean_token_accuracy": 0.8666445016860962, "num_tokens": 557960633.0, "step": 14701 }, { "epoch": 1.8702455158376798, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.355013370513916, "learning_rate": 1e-05, "loss": 0.4982, "mean_token_accuracy": 0.8695920705795288, "num_tokens": 557998430.0, "step": 14702 }, { "epoch": 1.8703727261162701, "ewc_loss": 0.11181640625, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 6.0474019050598145, "learning_rate": 1e-05, "loss": 0.5127, "mean_token_accuracy": 0.8613216876983643, "num_tokens": 558038283.0, "step": 14703 }, { "epoch": 1.8704999363948607, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.402938365936279, "learning_rate": 1e-05, "loss": 0.5031, "mean_token_accuracy": 0.8693980574607849, "num_tokens": 558079636.0, "step": 14704 }, { "epoch": 1.8706271466734512, "ewc_loss": 0.10791015625, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.907973766326904, "learning_rate": 1e-05, "loss": 0.4613, "mean_token_accuracy": 0.879027247428894, "num_tokens": 558117670.0, "step": 14705 }, { "epoch": 1.8707543569520417, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.54024600982666, "learning_rate": 1e-05, "loss": 0.5447, "mean_token_accuracy": 0.8551082015037537, "num_tokens": 558155732.0, "step": 14706 }, { "epoch": 1.8708815672306323, "ewc_loss": 0.10595703125, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.7485551834106445, "learning_rate": 1e-05, "loss": 0.496, "mean_token_accuracy": 0.8665647506713867, "num_tokens": 558195389.0, "step": 14707 }, { "epoch": 1.8710087775092228, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.793744087219238, "learning_rate": 1e-05, "loss": 0.4767, "mean_token_accuracy": 0.8815921545028687, "num_tokens": 558239535.0, "step": 14708 }, { "epoch": 1.871135987787813, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.631710052490234, "learning_rate": 1e-05, "loss": 0.4873, "mean_token_accuracy": 0.8683651089668274, "num_tokens": 558277441.0, "step": 14709 }, { "epoch": 1.8712631980664036, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 0.00011157989501953125, "grad_norm": 7.1659955978393555, "learning_rate": 1e-05, "loss": 0.54, "mean_token_accuracy": 0.8658225536346436, "num_tokens": 558319331.0, "step": 14710 }, { "epoch": 1.8713904083449941, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.689713954925537, "learning_rate": 1e-05, "loss": 0.4758, "mean_token_accuracy": 0.8731676340103149, "num_tokens": 558365494.0, "step": 14711 }, { "epoch": 1.8715176186235847, "ewc_loss": 0.1455078125, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 0.00012159347534179688, "grad_norm": 7.518812656402588, "learning_rate": 1e-05, "loss": 0.5363, "mean_token_accuracy": 0.8630161881446838, "num_tokens": 558407608.0, "step": 14712 }, { "epoch": 1.8716448289021752, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.092233657836914, "learning_rate": 1e-05, "loss": 0.5703, "mean_token_accuracy": 0.8533704876899719, "num_tokens": 558444848.0, "step": 14713 }, { "epoch": 1.8717720391807657, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 0.0001201629638671875, "grad_norm": 7.462479591369629, "learning_rate": 1e-05, "loss": 0.5899, "mean_token_accuracy": 0.8487623929977417, "num_tokens": 558486630.0, "step": 14714 }, { "epoch": 1.8718992494593563, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.445469379425049, "learning_rate": 1e-05, "loss": 0.4569, "mean_token_accuracy": 0.8817610740661621, "num_tokens": 558523609.0, "step": 14715 }, { "epoch": 1.8720264597379468, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 7.0385212898254395, "learning_rate": 1e-05, "loss": 0.4989, "mean_token_accuracy": 0.8770524263381958, "num_tokens": 558561065.0, "step": 14716 }, { "epoch": 1.8721536700165373, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.55683708190918, "learning_rate": 1e-05, "loss": 0.4862, "mean_token_accuracy": 0.8769747018814087, "num_tokens": 558602007.0, "step": 14717 }, { "epoch": 1.8722808802951278, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.534048080444336, "learning_rate": 1e-05, "loss": 0.5042, "mean_token_accuracy": 0.8700370192527771, "num_tokens": 558637712.0, "step": 14718 }, { "epoch": 1.8724080905737184, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.588794708251953, "learning_rate": 1e-05, "loss": 0.4899, "mean_token_accuracy": 0.8747203946113586, "num_tokens": 558671939.0, "step": 14719 }, { "epoch": 1.872535300852309, "ewc_loss": 0.11083984375, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 6.100643157958984, "learning_rate": 1e-05, "loss": 0.4983, "mean_token_accuracy": 0.8664182424545288, "num_tokens": 558708413.0, "step": 14720 }, { "epoch": 1.8726625111308994, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.71804666519165, "learning_rate": 1e-05, "loss": 0.5859, "mean_token_accuracy": 0.8428255319595337, "num_tokens": 558748993.0, "step": 14721 }, { "epoch": 1.87278972140949, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.820562839508057, "learning_rate": 1e-05, "loss": 0.5034, "mean_token_accuracy": 0.8651667833328247, "num_tokens": 558788058.0, "step": 14722 }, { "epoch": 1.8729169316880805, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.957667827606201, "learning_rate": 1e-05, "loss": 0.5847, "mean_token_accuracy": 0.8453186750411987, "num_tokens": 558828046.0, "step": 14723 }, { "epoch": 1.873044141966671, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.7320122718811035, "learning_rate": 1e-05, "loss": 0.4714, "mean_token_accuracy": 0.8716644048690796, "num_tokens": 558862445.0, "step": 14724 }, { "epoch": 1.8731713522452615, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 0.00011539459228515625, "grad_norm": 7.223568439483643, "learning_rate": 1e-05, "loss": 0.4909, "mean_token_accuracy": 0.8798141479492188, "num_tokens": 558907645.0, "step": 14725 }, { "epoch": 1.873298562523852, "ewc_loss": 0.11181640625, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 5.865114212036133, "learning_rate": 1e-05, "loss": 0.4733, "mean_token_accuracy": 0.8723738789558411, "num_tokens": 558943253.0, "step": 14726 }, { "epoch": 1.8734257728024426, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 0.00011873245239257812, "grad_norm": 7.382477760314941, "learning_rate": 1e-05, "loss": 0.5041, "mean_token_accuracy": 0.8740561008453369, "num_tokens": 558980341.0, "step": 14727 }, { "epoch": 1.873552983081033, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.134731769561768, "learning_rate": 1e-05, "loss": 0.5315, "mean_token_accuracy": 0.8596280217170715, "num_tokens": 559018659.0, "step": 14728 }, { "epoch": 1.8736801933596234, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 0.00011587142944335938, "grad_norm": 7.265522003173828, "learning_rate": 1e-05, "loss": 0.516, "mean_token_accuracy": 0.8728052377700806, "num_tokens": 559060254.0, "step": 14729 }, { "epoch": 1.873807403638214, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.318526744842529, "learning_rate": 1e-05, "loss": 0.4949, "mean_token_accuracy": 0.8722846508026123, "num_tokens": 559099166.0, "step": 14730 }, { "epoch": 1.8739346139168045, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 7.012075424194336, "learning_rate": 1e-05, "loss": 0.4894, "mean_token_accuracy": 0.8780406713485718, "num_tokens": 559135189.0, "step": 14731 }, { "epoch": 1.874061824195395, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.333197116851807, "learning_rate": 1e-05, "loss": 0.4893, "mean_token_accuracy": 0.8716553449630737, "num_tokens": 559178491.0, "step": 14732 }, { "epoch": 1.8741890344739855, "ewc_loss": 0.125, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.737973213195801, "learning_rate": 1e-05, "loss": 0.5348, "mean_token_accuracy": 0.8617257475852966, "num_tokens": 559213977.0, "step": 14733 }, { "epoch": 1.8743162447525759, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.2680535316467285, "learning_rate": 1e-05, "loss": 0.5018, "mean_token_accuracy": 0.8684585094451904, "num_tokens": 559254309.0, "step": 14734 }, { "epoch": 1.8744434550311664, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.552114963531494, "learning_rate": 1e-05, "loss": 0.4587, "mean_token_accuracy": 0.8826752305030823, "num_tokens": 559289756.0, "step": 14735 }, { "epoch": 1.874570665309757, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.166516304016113, "learning_rate": 1e-05, "loss": 0.4886, "mean_token_accuracy": 0.8720092177391052, "num_tokens": 559331143.0, "step": 14736 }, { "epoch": 1.8746978755883474, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.480541229248047, "learning_rate": 1e-05, "loss": 0.5029, "mean_token_accuracy": 0.8674452304840088, "num_tokens": 559368256.0, "step": 14737 }, { "epoch": 1.874825085866938, "ewc_loss": 0.11376953125, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.048396587371826, "learning_rate": 1e-05, "loss": 0.5501, "mean_token_accuracy": 0.8504235744476318, "num_tokens": 559408853.0, "step": 14738 }, { "epoch": 1.8749522961455285, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.510333061218262, "learning_rate": 1e-05, "loss": 0.5349, "mean_token_accuracy": 0.8596475124359131, "num_tokens": 559445606.0, "step": 14739 }, { "epoch": 1.875079506424119, "ewc_loss": 0.109375, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.912458896636963, "learning_rate": 1e-05, "loss": 0.4733, "mean_token_accuracy": 0.8719892501831055, "num_tokens": 559482482.0, "step": 14740 }, { "epoch": 1.8752067167027096, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.627197265625, "learning_rate": 1e-05, "loss": 0.4955, "mean_token_accuracy": 0.8710942268371582, "num_tokens": 559522757.0, "step": 14741 }, { "epoch": 1.8753339269813, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.795039176940918, "learning_rate": 1e-05, "loss": 0.464, "mean_token_accuracy": 0.8784191608428955, "num_tokens": 559556684.0, "step": 14742 }, { "epoch": 1.8754611372598906, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.856185436248779, "learning_rate": 1e-05, "loss": 0.5159, "mean_token_accuracy": 0.8690757155418396, "num_tokens": 559597548.0, "step": 14743 }, { "epoch": 1.8755883475384811, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.725484848022461, "learning_rate": 1e-05, "loss": 0.5094, "mean_token_accuracy": 0.8642992973327637, "num_tokens": 559638440.0, "step": 14744 }, { "epoch": 1.8757155578170717, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 0.00011205673217773438, "grad_norm": 7.143325328826904, "learning_rate": 1e-05, "loss": 0.5837, "mean_token_accuracy": 0.8481762409210205, "num_tokens": 559680819.0, "step": 14745 }, { "epoch": 1.8758427680956622, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 5.833026885986328, "learning_rate": 1e-05, "loss": 0.4997, "mean_token_accuracy": 0.8666183948516846, "num_tokens": 559721861.0, "step": 14746 }, { "epoch": 1.8759699783742527, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 0.0001163482666015625, "grad_norm": 7.305635929107666, "learning_rate": 1e-05, "loss": 0.5268, "mean_token_accuracy": 0.8691248297691345, "num_tokens": 559757370.0, "step": 14747 }, { "epoch": 1.8760971886528433, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.125360012054443, "learning_rate": 1e-05, "loss": 0.5119, "mean_token_accuracy": 0.8629493713378906, "num_tokens": 559795570.0, "step": 14748 }, { "epoch": 1.8762243989314338, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 0.0001125335693359375, "grad_norm": 7.139319896697998, "learning_rate": 1e-05, "loss": 0.4966, "mean_token_accuracy": 0.8741025924682617, "num_tokens": 559835115.0, "step": 14749 }, { "epoch": 1.8763516092100243, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.3641037940979, "learning_rate": 1e-05, "loss": 0.5246, "mean_token_accuracy": 0.8641980290412903, "num_tokens": 559876119.0, "step": 14750 }, { "epoch": 1.8764788194886148, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.6944732666015625, "learning_rate": 1e-05, "loss": 0.4936, "mean_token_accuracy": 0.8739550113677979, "num_tokens": 559908349.0, "step": 14751 }, { "epoch": 1.8766060297672051, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.485466003417969, "learning_rate": 1e-05, "loss": 0.5338, "mean_token_accuracy": 0.8562426567077637, "num_tokens": 559947336.0, "step": 14752 }, { "epoch": 1.8767332400457957, "ewc_loss": 0.11376953125, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.18148136138916, "learning_rate": 1e-05, "loss": 0.518, "mean_token_accuracy": 0.861177384853363, "num_tokens": 559989571.0, "step": 14753 }, { "epoch": 1.8768604503243862, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.6745758056640625, "learning_rate": 1e-05, "loss": 0.5486, "mean_token_accuracy": 0.8504642844200134, "num_tokens": 560022840.0, "step": 14754 }, { "epoch": 1.8769876606029767, "ewc_loss": 0.10595703125, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.763071537017822, "learning_rate": 1e-05, "loss": 0.5126, "mean_token_accuracy": 0.8607989549636841, "num_tokens": 560061939.0, "step": 14755 }, { "epoch": 1.8771148708815673, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 7.017460823059082, "learning_rate": 1e-05, "loss": 0.5174, "mean_token_accuracy": 0.8652832508087158, "num_tokens": 560097480.0, "step": 14756 }, { "epoch": 1.8772420811601578, "ewc_loss": 0.10498046875, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.585940361022949, "learning_rate": 1e-05, "loss": 0.4864, "mean_token_accuracy": 0.867907702922821, "num_tokens": 560137957.0, "step": 14757 }, { "epoch": 1.877369291438748, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 0.00012063980102539062, "grad_norm": 7.494569301605225, "learning_rate": 1e-05, "loss": 0.5318, "mean_token_accuracy": 0.8648297190666199, "num_tokens": 560173717.0, "step": 14758 }, { "epoch": 1.8774965017173386, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.00023651123046875, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 5.855361461639404, "learning_rate": 1e-05, "loss": 0.5039, "mean_token_accuracy": 0.8683853149414062, "num_tokens": 560209833.0, "step": 14759 }, { "epoch": 1.8776237119959291, "ewc_loss": 0.15234375, "ewc_loss_diag": 0.00023651123046875, "ewc_loss_parallel": 0.00012874603271484375, "grad_norm": 7.74048376083374, "learning_rate": 1e-05, "loss": 0.6177, "mean_token_accuracy": 0.8436174392700195, "num_tokens": 560250406.0, "step": 14760 }, { "epoch": 1.8777509222745197, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.00023651123046875, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.405024528503418, "learning_rate": 1e-05, "loss": 0.5132, "mean_token_accuracy": 0.8687241077423096, "num_tokens": 560288828.0, "step": 14761 }, { "epoch": 1.8778781325531102, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 0.0001201629638671875, "grad_norm": 7.4515700340271, "learning_rate": 1e-05, "loss": 0.5027, "mean_token_accuracy": 0.8781836032867432, "num_tokens": 560327431.0, "step": 14762 }, { "epoch": 1.8780053428317007, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.00023651123046875, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 6.617854595184326, "learning_rate": 1e-05, "loss": 0.4719, "mean_token_accuracy": 0.8864271640777588, "num_tokens": 560367767.0, "step": 14763 }, { "epoch": 1.8781325531102913, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.00023651123046875, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 7.020895481109619, "learning_rate": 1e-05, "loss": 0.4825, "mean_token_accuracy": 0.8805634379386902, "num_tokens": 560401179.0, "step": 14764 }, { "epoch": 1.8782597633888818, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.00023651123046875, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.549209117889404, "learning_rate": 1e-05, "loss": 0.5234, "mean_token_accuracy": 0.8645326495170593, "num_tokens": 560431198.0, "step": 14765 }, { "epoch": 1.8783869736674723, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.00023651123046875, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.689306735992432, "learning_rate": 1e-05, "loss": 0.5011, "mean_token_accuracy": 0.8700090646743774, "num_tokens": 560470244.0, "step": 14766 }, { "epoch": 1.8785141839460628, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.00023651123046875, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.381797790527344, "learning_rate": 1e-05, "loss": 0.4851, "mean_token_accuracy": 0.8753844499588013, "num_tokens": 560511696.0, "step": 14767 }, { "epoch": 1.8786413942246534, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.00023651123046875, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.500256061553955, "learning_rate": 1e-05, "loss": 0.4316, "mean_token_accuracy": 0.8931474685668945, "num_tokens": 560547555.0, "step": 14768 }, { "epoch": 1.878768604503244, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.00023651123046875, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.212146282196045, "learning_rate": 1e-05, "loss": 0.5033, "mean_token_accuracy": 0.8669223785400391, "num_tokens": 560584349.0, "step": 14769 }, { "epoch": 1.8788958147818344, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.00023651123046875, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.4638991355896, "learning_rate": 1e-05, "loss": 0.4806, "mean_token_accuracy": 0.8770768046379089, "num_tokens": 560628403.0, "step": 14770 }, { "epoch": 1.879023025060425, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.00023651123046875, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 6.033548831939697, "learning_rate": 1e-05, "loss": 0.491, "mean_token_accuracy": 0.8726611733436584, "num_tokens": 560663454.0, "step": 14771 }, { "epoch": 1.8791502353390155, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.00023651123046875, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.555723190307617, "learning_rate": 1e-05, "loss": 0.501, "mean_token_accuracy": 0.86836177110672, "num_tokens": 560700506.0, "step": 14772 }, { "epoch": 1.879277445617606, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.00023651123046875, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.855566024780273, "learning_rate": 1e-05, "loss": 0.4695, "mean_token_accuracy": 0.8752957582473755, "num_tokens": 560739690.0, "step": 14773 }, { "epoch": 1.8794046558961965, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.00023651123046875, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.763772487640381, "learning_rate": 1e-05, "loss": 0.4998, "mean_token_accuracy": 0.871665358543396, "num_tokens": 560786364.0, "step": 14774 }, { "epoch": 1.879531866174787, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.00023651123046875, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.7126641273498535, "learning_rate": 1e-05, "loss": 0.4931, "mean_token_accuracy": 0.8640502691268921, "num_tokens": 560828850.0, "step": 14775 }, { "epoch": 1.8796590764533776, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.00023651123046875, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 7.104286193847656, "learning_rate": 1e-05, "loss": 0.4963, "mean_token_accuracy": 0.8772860169410706, "num_tokens": 560863401.0, "step": 14776 }, { "epoch": 1.879786286731968, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.00023651123046875, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.744391441345215, "learning_rate": 1e-05, "loss": 0.4863, "mean_token_accuracy": 0.8683397769927979, "num_tokens": 560894347.0, "step": 14777 }, { "epoch": 1.8799134970105584, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.00023651123046875, "ewc_loss_parallel": 0.00011873245239257812, "grad_norm": 7.415441989898682, "learning_rate": 1e-05, "loss": 0.5702, "mean_token_accuracy": 0.8549014329910278, "num_tokens": 560935302.0, "step": 14778 }, { "epoch": 1.880040707289149, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.00023651123046875, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.01664924621582, "learning_rate": 1e-05, "loss": 0.4558, "mean_token_accuracy": 0.8780285120010376, "num_tokens": 560971960.0, "step": 14779 }, { "epoch": 1.8801679175677395, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.00023651123046875, "ewc_loss_parallel": 0.00011920928955078125, "grad_norm": 7.436883926391602, "learning_rate": 1e-05, "loss": 0.5369, "mean_token_accuracy": 0.862816572189331, "num_tokens": 561012267.0, "step": 14780 }, { "epoch": 1.88029512784633, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.00023651123046875, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.386458396911621, "learning_rate": 1e-05, "loss": 0.5073, "mean_token_accuracy": 0.8685749769210815, "num_tokens": 561048861.0, "step": 14781 }, { "epoch": 1.8804223381249205, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.00023651123046875, "ewc_loss_parallel": 0.00011014938354492188, "grad_norm": 7.075463771820068, "learning_rate": 1e-05, "loss": 0.4992, "mean_token_accuracy": 0.8720938563346863, "num_tokens": 561087832.0, "step": 14782 }, { "epoch": 1.8805495484035109, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.543974876403809, "learning_rate": 1e-05, "loss": 0.4816, "mean_token_accuracy": 0.8788601160049438, "num_tokens": 561120657.0, "step": 14783 }, { "epoch": 1.8806767586821014, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.557656764984131, "learning_rate": 1e-05, "loss": 0.4699, "mean_token_accuracy": 0.8796448707580566, "num_tokens": 561161762.0, "step": 14784 }, { "epoch": 1.880803968960692, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.603915214538574, "learning_rate": 1e-05, "loss": 0.4981, "mean_token_accuracy": 0.8711969256401062, "num_tokens": 561199459.0, "step": 14785 }, { "epoch": 1.8809311792392824, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 6.0947265625, "learning_rate": 1e-05, "loss": 0.5476, "mean_token_accuracy": 0.8540773391723633, "num_tokens": 561237958.0, "step": 14786 }, { "epoch": 1.881058389517873, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.763523578643799, "learning_rate": 1e-05, "loss": 0.5155, "mean_token_accuracy": 0.8633169531822205, "num_tokens": 561272086.0, "step": 14787 }, { "epoch": 1.8811855997964635, "ewc_loss": 0.10693359375, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.7687578201293945, "learning_rate": 1e-05, "loss": 0.4621, "mean_token_accuracy": 0.8780196905136108, "num_tokens": 561309821.0, "step": 14788 }, { "epoch": 1.881312810075054, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 7.067333221435547, "learning_rate": 1e-05, "loss": 0.5175, "mean_token_accuracy": 0.8679468631744385, "num_tokens": 561351565.0, "step": 14789 }, { "epoch": 1.8814400203536445, "ewc_loss": 0.10595703125, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.688960075378418, "learning_rate": 1e-05, "loss": 0.5544, "mean_token_accuracy": 0.8481931686401367, "num_tokens": 561387399.0, "step": 14790 }, { "epoch": 1.881567230632235, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 0.00011873245239257812, "grad_norm": 7.429923057556152, "learning_rate": 1e-05, "loss": 0.5402, "mean_token_accuracy": 0.8643147349357605, "num_tokens": 561427917.0, "step": 14791 }, { "epoch": 1.8816944409108256, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 5.9215240478515625, "learning_rate": 1e-05, "loss": 0.5076, "mean_token_accuracy": 0.8647769689559937, "num_tokens": 561465398.0, "step": 14792 }, { "epoch": 1.8818216511894161, "ewc_loss": 0.1455078125, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 0.0001220703125, "grad_norm": 7.574662208557129, "learning_rate": 1e-05, "loss": 0.5069, "mean_token_accuracy": 0.8724983334541321, "num_tokens": 561510264.0, "step": 14793 }, { "epoch": 1.8819488614680067, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.324504852294922, "learning_rate": 1e-05, "loss": 0.5293, "mean_token_accuracy": 0.864586591720581, "num_tokens": 561550942.0, "step": 14794 }, { "epoch": 1.8820760717465972, "ewc_loss": 0.140625, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 0.00011682510375976562, "grad_norm": 7.354822635650635, "learning_rate": 1e-05, "loss": 0.4957, "mean_token_accuracy": 0.8758392333984375, "num_tokens": 561580571.0, "step": 14795 }, { "epoch": 1.8822032820251877, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.511248588562012, "learning_rate": 1e-05, "loss": 0.5071, "mean_token_accuracy": 0.8725124001502991, "num_tokens": 561618993.0, "step": 14796 }, { "epoch": 1.8823304923037782, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.978633880615234, "learning_rate": 1e-05, "loss": 0.5681, "mean_token_accuracy": 0.8516702651977539, "num_tokens": 561659389.0, "step": 14797 }, { "epoch": 1.8824577025823688, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.467216968536377, "learning_rate": 1e-05, "loss": 0.4995, "mean_token_accuracy": 0.8695567846298218, "num_tokens": 561707703.0, "step": 14798 }, { "epoch": 1.8825849128609593, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.642334938049316, "learning_rate": 1e-05, "loss": 0.5007, "mean_token_accuracy": 0.8675901293754578, "num_tokens": 561749367.0, "step": 14799 }, { "epoch": 1.8827121231395498, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.368313312530518, "learning_rate": 1e-05, "loss": 0.5276, "mean_token_accuracy": 0.8603916168212891, "num_tokens": 561785234.0, "step": 14800 }, { "epoch": 1.8828393334181401, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.461767196655273, "learning_rate": 1e-05, "loss": 0.4724, "mean_token_accuracy": 0.8769698739051819, "num_tokens": 561825801.0, "step": 14801 }, { "epoch": 1.8829665436967307, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.211370468139648, "learning_rate": 1e-05, "loss": 0.5253, "mean_token_accuracy": 0.8610405921936035, "num_tokens": 561865030.0, "step": 14802 }, { "epoch": 1.8830937539753212, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.417290687561035, "learning_rate": 1e-05, "loss": 0.4719, "mean_token_accuracy": 0.879827082157135, "num_tokens": 561899767.0, "step": 14803 }, { "epoch": 1.8832209642539117, "ewc_loss": 0.11181640625, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 6.04246711730957, "learning_rate": 1e-05, "loss": 0.5095, "mean_token_accuracy": 0.8623816967010498, "num_tokens": 561936794.0, "step": 14804 }, { "epoch": 1.8833481745325023, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.500160217285156, "learning_rate": 1e-05, "loss": 0.4998, "mean_token_accuracy": 0.8676295280456543, "num_tokens": 561971770.0, "step": 14805 }, { "epoch": 1.8834753848110928, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.851227283477783, "learning_rate": 1e-05, "loss": 0.4973, "mean_token_accuracy": 0.8650674819946289, "num_tokens": 562016400.0, "step": 14806 }, { "epoch": 1.883602595089683, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.701022148132324, "learning_rate": 1e-05, "loss": 0.4753, "mean_token_accuracy": 0.8765894174575806, "num_tokens": 562050732.0, "step": 14807 }, { "epoch": 1.8837298053682736, "ewc_loss": 0.10595703125, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.718080043792725, "learning_rate": 1e-05, "loss": 0.4871, "mean_token_accuracy": 0.8709683418273926, "num_tokens": 562093619.0, "step": 14808 }, { "epoch": 1.8838570156468641, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 7.007131099700928, "learning_rate": 1e-05, "loss": 0.5249, "mean_token_accuracy": 0.8701992034912109, "num_tokens": 562131434.0, "step": 14809 }, { "epoch": 1.8839842259254547, "ewc_loss": 0.10595703125, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.679788112640381, "learning_rate": 1e-05, "loss": 0.5159, "mean_token_accuracy": 0.8588876724243164, "num_tokens": 562169914.0, "step": 14810 }, { "epoch": 1.8841114362040452, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 0.00011777877807617188, "grad_norm": 7.372714042663574, "learning_rate": 1e-05, "loss": 0.5024, "mean_token_accuracy": 0.8742166757583618, "num_tokens": 562207190.0, "step": 14811 }, { "epoch": 1.8842386464826357, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 5.917312145233154, "learning_rate": 1e-05, "loss": 0.5182, "mean_token_accuracy": 0.8624467253684998, "num_tokens": 562246236.0, "step": 14812 }, { "epoch": 1.8843658567612263, "ewc_loss": 0.1455078125, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 0.0001220703125, "grad_norm": 7.491331100463867, "learning_rate": 1e-05, "loss": 0.4949, "mean_token_accuracy": 0.8809346556663513, "num_tokens": 562286875.0, "step": 14813 }, { "epoch": 1.8844930670398168, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.331764221191406, "learning_rate": 1e-05, "loss": 0.4913, "mean_token_accuracy": 0.8743919134140015, "num_tokens": 562325807.0, "step": 14814 }, { "epoch": 1.8846202773184073, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 0.00011205673217773438, "grad_norm": 7.1785888671875, "learning_rate": 1e-05, "loss": 0.5131, "mean_token_accuracy": 0.8706437945365906, "num_tokens": 562363663.0, "step": 14815 }, { "epoch": 1.8847474875969978, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.542575359344482, "learning_rate": 1e-05, "loss": 0.5076, "mean_token_accuracy": 0.8706393837928772, "num_tokens": 562399700.0, "step": 14816 }, { "epoch": 1.8848746978755884, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.650875568389893, "learning_rate": 1e-05, "loss": 0.533, "mean_token_accuracy": 0.8606857061386108, "num_tokens": 562434395.0, "step": 14817 }, { "epoch": 1.885001908154179, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.602048397064209, "learning_rate": 1e-05, "loss": 0.5543, "mean_token_accuracy": 0.853161633014679, "num_tokens": 562472140.0, "step": 14818 }, { "epoch": 1.8851291184327694, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.1485185623168945, "learning_rate": 1e-05, "loss": 0.5024, "mean_token_accuracy": 0.8674670457839966, "num_tokens": 562510556.0, "step": 14819 }, { "epoch": 1.88525632871136, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.7535810470581055, "learning_rate": 1e-05, "loss": 0.5202, "mean_token_accuracy": 0.864304780960083, "num_tokens": 562546679.0, "step": 14820 }, { "epoch": 1.8853835389899505, "ewc_loss": 0.10595703125, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.797144889831543, "learning_rate": 1e-05, "loss": 0.5003, "mean_token_accuracy": 0.8656795024871826, "num_tokens": 562580707.0, "step": 14821 }, { "epoch": 1.885510749268541, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 0.00010967254638671875, "grad_norm": 7.057076454162598, "learning_rate": 1e-05, "loss": 0.5099, "mean_token_accuracy": 0.8685393929481506, "num_tokens": 562621753.0, "step": 14822 }, { "epoch": 1.8856379595471315, "ewc_loss": 0.10693359375, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.696152210235596, "learning_rate": 1e-05, "loss": 0.4247, "mean_token_accuracy": 0.8890259861946106, "num_tokens": 562654064.0, "step": 14823 }, { "epoch": 1.885765169825722, "ewc_loss": 0.140625, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 0.00011730194091796875, "grad_norm": 7.417820930480957, "learning_rate": 1e-05, "loss": 0.5487, "mean_token_accuracy": 0.8589484095573425, "num_tokens": 562692575.0, "step": 14824 }, { "epoch": 1.8858923801043126, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 5.9351701736450195, "learning_rate": 1e-05, "loss": 0.4677, "mean_token_accuracy": 0.8800770044326782, "num_tokens": 562732655.0, "step": 14825 }, { "epoch": 1.886019590382903, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.562880516052246, "learning_rate": 1e-05, "loss": 0.5402, "mean_token_accuracy": 0.8663519024848938, "num_tokens": 562768618.0, "step": 14826 }, { "epoch": 1.8861468006614934, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.32218074798584, "learning_rate": 1e-05, "loss": 0.5527, "mean_token_accuracy": 0.8541889786720276, "num_tokens": 562805357.0, "step": 14827 }, { "epoch": 1.886274010940084, "ewc_loss": 0.140625, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 0.00011682510375976562, "grad_norm": 7.3362531661987305, "learning_rate": 1e-05, "loss": 0.548, "mean_token_accuracy": 0.8639795780181885, "num_tokens": 562848798.0, "step": 14828 }, { "epoch": 1.8864012212186745, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.487484931945801, "learning_rate": 1e-05, "loss": 0.4828, "mean_token_accuracy": 0.8777520060539246, "num_tokens": 562890108.0, "step": 14829 }, { "epoch": 1.886528431497265, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.9779791831970215, "learning_rate": 1e-05, "loss": 0.4818, "mean_token_accuracy": 0.8752198219299316, "num_tokens": 562921089.0, "step": 14830 }, { "epoch": 1.8866556417758555, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.4498162269592285, "learning_rate": 1e-05, "loss": 0.6058, "mean_token_accuracy": 0.8345316052436829, "num_tokens": 562961705.0, "step": 14831 }, { "epoch": 1.8867828520544458, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.653526306152344, "learning_rate": 1e-05, "loss": 0.4538, "mean_token_accuracy": 0.884475588798523, "num_tokens": 563000011.0, "step": 14832 }, { "epoch": 1.8869100623330364, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.33030891418457, "learning_rate": 1e-05, "loss": 0.4851, "mean_token_accuracy": 0.8738027811050415, "num_tokens": 563042568.0, "step": 14833 }, { "epoch": 1.887037272611627, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.483212471008301, "learning_rate": 1e-05, "loss": 0.4794, "mean_token_accuracy": 0.876503586769104, "num_tokens": 563075892.0, "step": 14834 }, { "epoch": 1.8871644828902174, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.189615726470947, "learning_rate": 1e-05, "loss": 0.5228, "mean_token_accuracy": 0.867719292640686, "num_tokens": 563117276.0, "step": 14835 }, { "epoch": 1.887291693168808, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.443776607513428, "learning_rate": 1e-05, "loss": 0.4962, "mean_token_accuracy": 0.8698368072509766, "num_tokens": 563152557.0, "step": 14836 }, { "epoch": 1.8874189034473985, "ewc_loss": 0.11181640625, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 6.033321857452393, "learning_rate": 1e-05, "loss": 0.4807, "mean_token_accuracy": 0.8727968335151672, "num_tokens": 563188620.0, "step": 14837 }, { "epoch": 1.887546113725989, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.52005672454834, "learning_rate": 1e-05, "loss": 0.5161, "mean_token_accuracy": 0.8642324805259705, "num_tokens": 563227582.0, "step": 14838 }, { "epoch": 1.8876733240045795, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.869177341461182, "learning_rate": 1e-05, "loss": 0.4419, "mean_token_accuracy": 0.8815492391586304, "num_tokens": 563260057.0, "step": 14839 }, { "epoch": 1.88780053428317, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.7044243812561035, "learning_rate": 1e-05, "loss": 0.503, "mean_token_accuracy": 0.8672736883163452, "num_tokens": 563299612.0, "step": 14840 }, { "epoch": 1.8879277445617606, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.729668140411377, "learning_rate": 1e-05, "loss": 0.4759, "mean_token_accuracy": 0.8704395294189453, "num_tokens": 563331659.0, "step": 14841 }, { "epoch": 1.8880549548403511, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 7.005216121673584, "learning_rate": 1e-05, "loss": 0.5295, "mean_token_accuracy": 0.8612377643585205, "num_tokens": 563371411.0, "step": 14842 }, { "epoch": 1.8881821651189417, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.714823246002197, "learning_rate": 1e-05, "loss": 0.4885, "mean_token_accuracy": 0.8696497082710266, "num_tokens": 563401498.0, "step": 14843 }, { "epoch": 1.8883093753975322, "ewc_loss": 0.140625, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 0.00011730194091796875, "grad_norm": 7.327457427978516, "learning_rate": 1e-05, "loss": 0.4663, "mean_token_accuracy": 0.8841745853424072, "num_tokens": 563441383.0, "step": 14844 }, { "epoch": 1.8884365856761227, "ewc_loss": 0.11474609375, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 5.9152445793151855, "learning_rate": 1e-05, "loss": 0.5197, "mean_token_accuracy": 0.8590823411941528, "num_tokens": 563478547.0, "step": 14845 }, { "epoch": 1.8885637959547132, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 0.0001201629638671875, "grad_norm": 7.442398548126221, "learning_rate": 1e-05, "loss": 0.528, "mean_token_accuracy": 0.8681610226631165, "num_tokens": 563521366.0, "step": 14846 }, { "epoch": 1.8886910062333038, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.284812927246094, "learning_rate": 1e-05, "loss": 0.5525, "mean_token_accuracy": 0.8517787456512451, "num_tokens": 563560062.0, "step": 14847 }, { "epoch": 1.8888182165118943, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 0.00011205673217773438, "grad_norm": 7.1569061279296875, "learning_rate": 1e-05, "loss": 0.5388, "mean_token_accuracy": 0.8618487119674683, "num_tokens": 563598448.0, "step": 14848 }, { "epoch": 1.8889454267904848, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.484512805938721, "learning_rate": 1e-05, "loss": 0.4835, "mean_token_accuracy": 0.8769570589065552, "num_tokens": 563639274.0, "step": 14849 }, { "epoch": 1.8890726370690751, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.671498775482178, "learning_rate": 1e-05, "loss": 0.5259, "mean_token_accuracy": 0.8672735095024109, "num_tokens": 563679309.0, "step": 14850 }, { "epoch": 1.8891998473476657, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.555526256561279, "learning_rate": 1e-05, "loss": 0.5099, "mean_token_accuracy": 0.8685726523399353, "num_tokens": 563717594.0, "step": 14851 }, { "epoch": 1.8893270576262562, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.188103675842285, "learning_rate": 1e-05, "loss": 0.4836, "mean_token_accuracy": 0.8787415027618408, "num_tokens": 563749442.0, "step": 14852 }, { "epoch": 1.8894542679048467, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.68980073928833, "learning_rate": 1e-05, "loss": 0.5231, "mean_token_accuracy": 0.8614481687545776, "num_tokens": 563788519.0, "step": 14853 }, { "epoch": 1.8895814781834372, "ewc_loss": 0.10791015625, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.827265739440918, "learning_rate": 1e-05, "loss": 0.5208, "mean_token_accuracy": 0.8593301773071289, "num_tokens": 563827233.0, "step": 14854 }, { "epoch": 1.8897086884620278, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.9638352394104, "learning_rate": 1e-05, "loss": 0.5132, "mean_token_accuracy": 0.8689132928848267, "num_tokens": 563863534.0, "step": 14855 }, { "epoch": 1.889835898740618, "ewc_loss": 0.10595703125, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.677830696105957, "learning_rate": 1e-05, "loss": 0.5408, "mean_token_accuracy": 0.8518205881118774, "num_tokens": 563903237.0, "step": 14856 }, { "epoch": 1.8899631090192086, "ewc_loss": 0.140625, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 0.00011730194091796875, "grad_norm": 7.32790994644165, "learning_rate": 1e-05, "loss": 0.4445, "mean_token_accuracy": 0.8913589119911194, "num_tokens": 563944334.0, "step": 14857 }, { "epoch": 1.8900903192977991, "ewc_loss": 0.11376953125, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 5.821455955505371, "learning_rate": 1e-05, "loss": 0.5049, "mean_token_accuracy": 0.8638367056846619, "num_tokens": 563979801.0, "step": 14858 }, { "epoch": 1.8902175295763897, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.566582202911377, "learning_rate": 1e-05, "loss": 0.5409, "mean_token_accuracy": 0.8624343872070312, "num_tokens": 564022348.0, "step": 14859 }, { "epoch": 1.8903447398549802, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.221986293792725, "learning_rate": 1e-05, "loss": 0.4617, "mean_token_accuracy": 0.8819407820701599, "num_tokens": 564063285.0, "step": 14860 }, { "epoch": 1.8904719501335707, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 0.00011873245239257812, "grad_norm": 7.42350959777832, "learning_rate": 1e-05, "loss": 0.5315, "mean_token_accuracy": 0.8691390156745911, "num_tokens": 564105323.0, "step": 14861 }, { "epoch": 1.8905991604121613, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.477586269378662, "learning_rate": 1e-05, "loss": 0.5387, "mean_token_accuracy": 0.8588382005691528, "num_tokens": 564140238.0, "step": 14862 }, { "epoch": 1.8907263706907518, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 0.00011014938354492188, "grad_norm": 7.080653190612793, "learning_rate": 1e-05, "loss": 0.5995, "mean_token_accuracy": 0.8403761386871338, "num_tokens": 564178441.0, "step": 14863 }, { "epoch": 1.8908535809693423, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.472299098968506, "learning_rate": 1e-05, "loss": 0.5081, "mean_token_accuracy": 0.8689113855361938, "num_tokens": 564217586.0, "step": 14864 }, { "epoch": 1.8909807912479328, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.738605976104736, "learning_rate": 1e-05, "loss": 0.5352, "mean_token_accuracy": 0.8585728406906128, "num_tokens": 564259198.0, "step": 14865 }, { "epoch": 1.8911080015265234, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.359194755554199, "learning_rate": 1e-05, "loss": 0.4603, "mean_token_accuracy": 0.8827956914901733, "num_tokens": 564298622.0, "step": 14866 }, { "epoch": 1.891235211805114, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.530806541442871, "learning_rate": 1e-05, "loss": 0.4618, "mean_token_accuracy": 0.8752782344818115, "num_tokens": 564331573.0, "step": 14867 }, { "epoch": 1.8913624220837044, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.2279953956604, "learning_rate": 1e-05, "loss": 0.5098, "mean_token_accuracy": 0.86249840259552, "num_tokens": 564371461.0, "step": 14868 }, { "epoch": 1.891489632362295, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.455595970153809, "learning_rate": 1e-05, "loss": 0.4363, "mean_token_accuracy": 0.88722163438797, "num_tokens": 564410160.0, "step": 14869 }, { "epoch": 1.8916168426408855, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 6.069427013397217, "learning_rate": 1e-05, "loss": 0.5264, "mean_token_accuracy": 0.8572857975959778, "num_tokens": 564455954.0, "step": 14870 }, { "epoch": 1.891744052919476, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.510997295379639, "learning_rate": 1e-05, "loss": 0.504, "mean_token_accuracy": 0.869423508644104, "num_tokens": 564490196.0, "step": 14871 }, { "epoch": 1.8918712631980665, "ewc_loss": 0.10888671875, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.910231113433838, "learning_rate": 1e-05, "loss": 0.4635, "mean_token_accuracy": 0.8749998807907104, "num_tokens": 564530362.0, "step": 14872 }, { "epoch": 1.891998473476657, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.672745227813721, "learning_rate": 1e-05, "loss": 0.5117, "mean_token_accuracy": 0.8657230138778687, "num_tokens": 564563528.0, "step": 14873 }, { "epoch": 1.8921256837552476, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.773414134979248, "learning_rate": 1e-05, "loss": 0.4845, "mean_token_accuracy": 0.871024489402771, "num_tokens": 564596865.0, "step": 14874 }, { "epoch": 1.892252894033838, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.94399356842041, "learning_rate": 1e-05, "loss": 0.5368, "mean_token_accuracy": 0.8614874482154846, "num_tokens": 564635740.0, "step": 14875 }, { "epoch": 1.8923801043124284, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.710963726043701, "learning_rate": 1e-05, "loss": 0.5126, "mean_token_accuracy": 0.8576615452766418, "num_tokens": 564679007.0, "step": 14876 }, { "epoch": 1.892507314591019, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 0.00011539459228515625, "grad_norm": 7.275452613830566, "learning_rate": 1e-05, "loss": 0.5093, "mean_token_accuracy": 0.8726585507392883, "num_tokens": 564719211.0, "step": 14877 }, { "epoch": 1.8926345248696095, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 5.871275901794434, "learning_rate": 1e-05, "loss": 0.4441, "mean_token_accuracy": 0.883094310760498, "num_tokens": 564756482.0, "step": 14878 }, { "epoch": 1.8927617351482, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 0.00011920928955078125, "grad_norm": 7.438432216644287, "learning_rate": 1e-05, "loss": 0.5336, "mean_token_accuracy": 0.869477391242981, "num_tokens": 564796304.0, "step": 14879 }, { "epoch": 1.8928889454267905, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.248723030090332, "learning_rate": 1e-05, "loss": 0.5086, "mean_token_accuracy": 0.8672741055488586, "num_tokens": 564835444.0, "step": 14880 }, { "epoch": 1.8930161557053808, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 0.00011301040649414062, "grad_norm": 7.198442459106445, "learning_rate": 1e-05, "loss": 0.5587, "mean_token_accuracy": 0.8549633026123047, "num_tokens": 564873637.0, "step": 14881 }, { "epoch": 1.8931433659839714, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.4857659339904785, "learning_rate": 1e-05, "loss": 0.4968, "mean_token_accuracy": 0.8704946041107178, "num_tokens": 564914742.0, "step": 14882 }, { "epoch": 1.893270576262562, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.700911045074463, "learning_rate": 1e-05, "loss": 0.4996, "mean_token_accuracy": 0.8704540133476257, "num_tokens": 564953519.0, "step": 14883 }, { "epoch": 1.8933977865411524, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.579972743988037, "learning_rate": 1e-05, "loss": 0.526, "mean_token_accuracy": 0.8625825643539429, "num_tokens": 564984994.0, "step": 14884 }, { "epoch": 1.893524996819743, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.180827617645264, "learning_rate": 1e-05, "loss": 0.5255, "mean_token_accuracy": 0.8583323359489441, "num_tokens": 565028584.0, "step": 14885 }, { "epoch": 1.8936522070983335, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.713797569274902, "learning_rate": 1e-05, "loss": 0.5216, "mean_token_accuracy": 0.86710524559021, "num_tokens": 565068726.0, "step": 14886 }, { "epoch": 1.893779417376924, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.810826301574707, "learning_rate": 1e-05, "loss": 0.5271, "mean_token_accuracy": 0.8602776527404785, "num_tokens": 565112156.0, "step": 14887 }, { "epoch": 1.8939066276555145, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 7.009053707122803, "learning_rate": 1e-05, "loss": 0.4953, "mean_token_accuracy": 0.87162846326828, "num_tokens": 565146158.0, "step": 14888 }, { "epoch": 1.894033837934105, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.674769401550293, "learning_rate": 1e-05, "loss": 0.4438, "mean_token_accuracy": 0.8818446397781372, "num_tokens": 565176571.0, "step": 14889 }, { "epoch": 1.8941610482126956, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 0.00011873245239257812, "grad_norm": 7.3893327713012695, "learning_rate": 1e-05, "loss": 0.5273, "mean_token_accuracy": 0.8662213087081909, "num_tokens": 565216162.0, "step": 14890 }, { "epoch": 1.8942882584912861, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 5.869449138641357, "learning_rate": 1e-05, "loss": 0.4444, "mean_token_accuracy": 0.8821113109588623, "num_tokens": 565252010.0, "step": 14891 }, { "epoch": 1.8944154687698767, "ewc_loss": 0.1474609375, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 0.0001239776611328125, "grad_norm": 7.584556579589844, "learning_rate": 1e-05, "loss": 0.5719, "mean_token_accuracy": 0.851858377456665, "num_tokens": 565297159.0, "step": 14892 }, { "epoch": 1.8945426790484672, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.276182174682617, "learning_rate": 1e-05, "loss": 0.4887, "mean_token_accuracy": 0.8721378445625305, "num_tokens": 565337074.0, "step": 14893 }, { "epoch": 1.8946698893270577, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 0.00011920928955078125, "grad_norm": 7.386911869049072, "learning_rate": 1e-05, "loss": 0.5505, "mean_token_accuracy": 0.8618016242980957, "num_tokens": 565379551.0, "step": 14894 }, { "epoch": 1.8947970996056482, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.495285987854004, "learning_rate": 1e-05, "loss": 0.4978, "mean_token_accuracy": 0.8726847767829895, "num_tokens": 565418611.0, "step": 14895 }, { "epoch": 1.8949243098842388, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 7.012963771820068, "learning_rate": 1e-05, "loss": 0.5103, "mean_token_accuracy": 0.870983362197876, "num_tokens": 565452804.0, "step": 14896 }, { "epoch": 1.8950515201628293, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.479633808135986, "learning_rate": 1e-05, "loss": 0.5234, "mean_token_accuracy": 0.8638916015625, "num_tokens": 565490628.0, "step": 14897 }, { "epoch": 1.8951787304414198, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.682003021240234, "learning_rate": 1e-05, "loss": 0.524, "mean_token_accuracy": 0.8605411052703857, "num_tokens": 565523990.0, "step": 14898 }, { "epoch": 1.8953059407200101, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.370983600616455, "learning_rate": 1e-05, "loss": 0.4884, "mean_token_accuracy": 0.8755972385406494, "num_tokens": 565562684.0, "step": 14899 }, { "epoch": 1.8954331509986007, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.4780402183532715, "learning_rate": 1e-05, "loss": 0.4307, "mean_token_accuracy": 0.890315055847168, "num_tokens": 565596479.0, "step": 14900 }, { "epoch": 1.8955603612771912, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.2213311195373535, "learning_rate": 1e-05, "loss": 0.4975, "mean_token_accuracy": 0.8664399981498718, "num_tokens": 565639112.0, "step": 14901 }, { "epoch": 1.8956875715557817, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.425344467163086, "learning_rate": 1e-05, "loss": 0.4692, "mean_token_accuracy": 0.8763779997825623, "num_tokens": 565671379.0, "step": 14902 }, { "epoch": 1.8958147818343722, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 6.061630725860596, "learning_rate": 1e-05, "loss": 0.4785, "mean_token_accuracy": 0.8766332864761353, "num_tokens": 565708115.0, "step": 14903 }, { "epoch": 1.8959419921129628, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.488891124725342, "learning_rate": 1e-05, "loss": 0.5181, "mean_token_accuracy": 0.8630322217941284, "num_tokens": 565744767.0, "step": 14904 }, { "epoch": 1.896069202391553, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.889542579650879, "learning_rate": 1e-05, "loss": 0.4948, "mean_token_accuracy": 0.8685469627380371, "num_tokens": 565783709.0, "step": 14905 }, { "epoch": 1.8961964126701436, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.658792495727539, "learning_rate": 1e-05, "loss": 0.4927, "mean_token_accuracy": 0.8732839822769165, "num_tokens": 565818503.0, "step": 14906 }, { "epoch": 1.8963236229487341, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.737647533416748, "learning_rate": 1e-05, "loss": 0.4263, "mean_token_accuracy": 0.8886970281600952, "num_tokens": 565850188.0, "step": 14907 }, { "epoch": 1.8964508332273247, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.953652381896973, "learning_rate": 1e-05, "loss": 0.5283, "mean_token_accuracy": 0.8624076247215271, "num_tokens": 565889177.0, "step": 14908 }, { "epoch": 1.8965780435059152, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.672244071960449, "learning_rate": 1e-05, "loss": 0.5557, "mean_token_accuracy": 0.8485836982727051, "num_tokens": 565929316.0, "step": 14909 }, { "epoch": 1.8967052537845057, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 0.00011777877807617188, "grad_norm": 7.32059383392334, "learning_rate": 1e-05, "loss": 0.4722, "mean_token_accuracy": 0.8867493867874146, "num_tokens": 565961874.0, "step": 14910 }, { "epoch": 1.8968324640630962, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 5.866913318634033, "learning_rate": 1e-05, "loss": 0.5355, "mean_token_accuracy": 0.8532041907310486, "num_tokens": 565995143.0, "step": 14911 }, { "epoch": 1.8969596743416868, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 0.00012111663818359375, "grad_norm": 7.508044719696045, "learning_rate": 1e-05, "loss": 0.5309, "mean_token_accuracy": 0.8686614632606506, "num_tokens": 566032297.0, "step": 14912 }, { "epoch": 1.8970868846202773, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.2569475173950195, "learning_rate": 1e-05, "loss": 0.4855, "mean_token_accuracy": 0.8732988834381104, "num_tokens": 566073138.0, "step": 14913 }, { "epoch": 1.8972140948988678, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 0.00011587142944335938, "grad_norm": 7.285159587860107, "learning_rate": 1e-05, "loss": 0.5197, "mean_token_accuracy": 0.8732439279556274, "num_tokens": 566115471.0, "step": 14914 }, { "epoch": 1.8973413051774584, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.509161472320557, "learning_rate": 1e-05, "loss": 0.5191, "mean_token_accuracy": 0.8656491637229919, "num_tokens": 566152644.0, "step": 14915 }, { "epoch": 1.8974685154560489, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.791599750518799, "learning_rate": 1e-05, "loss": 0.5104, "mean_token_accuracy": 0.8662874698638916, "num_tokens": 566197411.0, "step": 14916 }, { "epoch": 1.8975957257346394, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.567660808563232, "learning_rate": 1e-05, "loss": 0.4296, "mean_token_accuracy": 0.892568051815033, "num_tokens": 566234738.0, "step": 14917 }, { "epoch": 1.89772293601323, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.295968532562256, "learning_rate": 1e-05, "loss": 0.5518, "mean_token_accuracy": 0.851955771446228, "num_tokens": 566274445.0, "step": 14918 }, { "epoch": 1.8978501462918205, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.664116382598877, "learning_rate": 1e-05, "loss": 0.4513, "mean_token_accuracy": 0.8879886269569397, "num_tokens": 566312710.0, "step": 14919 }, { "epoch": 1.897977356570411, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.911548137664795, "learning_rate": 1e-05, "loss": 0.5076, "mean_token_accuracy": 0.866196870803833, "num_tokens": 566349229.0, "step": 14920 }, { "epoch": 1.8981045668490015, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.899307727813721, "learning_rate": 1e-05, "loss": 0.54, "mean_token_accuracy": 0.8619763851165771, "num_tokens": 566387208.0, "step": 14921 }, { "epoch": 1.898231777127592, "ewc_loss": 0.10498046875, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.706384658813477, "learning_rate": 1e-05, "loss": 0.4826, "mean_token_accuracy": 0.8684505224227905, "num_tokens": 566423616.0, "step": 14922 }, { "epoch": 1.8983589874061826, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 0.00011396408081054688, "grad_norm": 7.240169525146484, "learning_rate": 1e-05, "loss": 0.5022, "mean_token_accuracy": 0.87315833568573, "num_tokens": 566457984.0, "step": 14923 }, { "epoch": 1.898486197684773, "ewc_loss": 0.109375, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.786275386810303, "learning_rate": 1e-05, "loss": 0.5102, "mean_token_accuracy": 0.8642737865447998, "num_tokens": 566495664.0, "step": 14924 }, { "epoch": 1.8986134079633634, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 0.0001201629638671875, "grad_norm": 7.504148960113525, "learning_rate": 1e-05, "loss": 0.5451, "mean_token_accuracy": 0.8615961670875549, "num_tokens": 566533750.0, "step": 14925 }, { "epoch": 1.898740618241954, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.119394779205322, "learning_rate": 1e-05, "loss": 0.523, "mean_token_accuracy": 0.859843373298645, "num_tokens": 566576247.0, "step": 14926 }, { "epoch": 1.8988678285205445, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 0.00012111663818359375, "grad_norm": 7.458662986755371, "learning_rate": 1e-05, "loss": 0.5783, "mean_token_accuracy": 0.8524460196495056, "num_tokens": 566618185.0, "step": 14927 }, { "epoch": 1.898995038799135, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.425439834594727, "learning_rate": 1e-05, "loss": 0.5378, "mean_token_accuracy": 0.8559308052062988, "num_tokens": 566659219.0, "step": 14928 }, { "epoch": 1.8991222490777255, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 0.00011110305786132812, "grad_norm": 7.1311469078063965, "learning_rate": 1e-05, "loss": 0.4911, "mean_token_accuracy": 0.8756098747253418, "num_tokens": 566695101.0, "step": 14929 }, { "epoch": 1.8992494593563158, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.4587721824646, "learning_rate": 1e-05, "loss": 0.5207, "mean_token_accuracy": 0.8675507307052612, "num_tokens": 566735286.0, "step": 14930 }, { "epoch": 1.8993766696349064, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.793536186218262, "learning_rate": 1e-05, "loss": 0.5002, "mean_token_accuracy": 0.8702659010887146, "num_tokens": 566773534.0, "step": 14931 }, { "epoch": 1.899503879913497, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.385920524597168, "learning_rate": 1e-05, "loss": 0.4875, "mean_token_accuracy": 0.8748795986175537, "num_tokens": 566819141.0, "step": 14932 }, { "epoch": 1.8996310901920874, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.53569221496582, "learning_rate": 1e-05, "loss": 0.4606, "mean_token_accuracy": 0.8846702575683594, "num_tokens": 566855095.0, "step": 14933 }, { "epoch": 1.899758300470678, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.274683475494385, "learning_rate": 1e-05, "loss": 0.5187, "mean_token_accuracy": 0.8614835739135742, "num_tokens": 566888989.0, "step": 14934 }, { "epoch": 1.8998855107492685, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.431802749633789, "learning_rate": 1e-05, "loss": 0.4615, "mean_token_accuracy": 0.8807462453842163, "num_tokens": 566923049.0, "step": 14935 }, { "epoch": 1.900012721027859, "ewc_loss": 0.11474609375, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.132330894470215, "learning_rate": 1e-05, "loss": 0.5588, "mean_token_accuracy": 0.8489636778831482, "num_tokens": 566962049.0, "step": 14936 }, { "epoch": 1.9001399313064495, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.425838470458984, "learning_rate": 1e-05, "loss": 0.4521, "mean_token_accuracy": 0.8840346336364746, "num_tokens": 567003538.0, "step": 14937 }, { "epoch": 1.90026714158504, "ewc_loss": 0.10986328125, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.968064308166504, "learning_rate": 1e-05, "loss": 0.4622, "mean_token_accuracy": 0.8796243071556091, "num_tokens": 567035821.0, "step": 14938 }, { "epoch": 1.9003943518636306, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.5470685958862305, "learning_rate": 1e-05, "loss": 0.4815, "mean_token_accuracy": 0.8777862787246704, "num_tokens": 567068095.0, "step": 14939 }, { "epoch": 1.9005215621422211, "ewc_loss": 0.10693359375, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.810737133026123, "learning_rate": 1e-05, "loss": 0.4679, "mean_token_accuracy": 0.8735697269439697, "num_tokens": 567104741.0, "step": 14940 }, { "epoch": 1.9006487724208116, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.772103309631348, "learning_rate": 1e-05, "loss": 0.5077, "mean_token_accuracy": 0.8677850961685181, "num_tokens": 567142500.0, "step": 14941 }, { "epoch": 1.9007759826994022, "ewc_loss": 0.10595703125, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.677044868469238, "learning_rate": 1e-05, "loss": 0.4963, "mean_token_accuracy": 0.8674114942550659, "num_tokens": 567182001.0, "step": 14942 }, { "epoch": 1.9009031929779927, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 0.00011014938354492188, "grad_norm": 7.1380085945129395, "learning_rate": 1e-05, "loss": 0.5315, "mean_token_accuracy": 0.865899384021759, "num_tokens": 567219809.0, "step": 14943 }, { "epoch": 1.9010304032565832, "ewc_loss": 0.10791015625, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.719945907592773, "learning_rate": 1e-05, "loss": 0.5748, "mean_token_accuracy": 0.8408703804016113, "num_tokens": 567260309.0, "step": 14944 }, { "epoch": 1.9011576135351738, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 0.00011968612670898438, "grad_norm": 7.4714436531066895, "learning_rate": 1e-05, "loss": 0.5581, "mean_token_accuracy": 0.8561750054359436, "num_tokens": 567295840.0, "step": 14945 }, { "epoch": 1.9012848238137643, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.052767276763916, "learning_rate": 1e-05, "loss": 0.5642, "mean_token_accuracy": 0.847809374332428, "num_tokens": 567335546.0, "step": 14946 }, { "epoch": 1.9014120340923548, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 0.00011968612670898438, "grad_norm": 7.451577663421631, "learning_rate": 1e-05, "loss": 0.4778, "mean_token_accuracy": 0.8861263990402222, "num_tokens": 567372996.0, "step": 14947 }, { "epoch": 1.9015392443709451, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.424853801727295, "learning_rate": 1e-05, "loss": 0.4901, "mean_token_accuracy": 0.8774121403694153, "num_tokens": 567411758.0, "step": 14948 }, { "epoch": 1.9016664546495357, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 0.00010967254638671875, "grad_norm": 7.063912391662598, "learning_rate": 1e-05, "loss": 0.5275, "mean_token_accuracy": 0.865050733089447, "num_tokens": 567446859.0, "step": 14949 }, { "epoch": 1.9017936649281262, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.556533336639404, "learning_rate": 1e-05, "loss": 0.4922, "mean_token_accuracy": 0.8790882229804993, "num_tokens": 567480890.0, "step": 14950 }, { "epoch": 1.9019208752067167, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.526377201080322, "learning_rate": 1e-05, "loss": 0.4763, "mean_token_accuracy": 0.8753626346588135, "num_tokens": 567523381.0, "step": 14951 }, { "epoch": 1.9020480854853072, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.62384557723999, "learning_rate": 1e-05, "loss": 0.523, "mean_token_accuracy": 0.8677405118942261, "num_tokens": 567564926.0, "step": 14952 }, { "epoch": 1.9021752957638978, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.073254585266113, "learning_rate": 1e-05, "loss": 0.5007, "mean_token_accuracy": 0.8669576048851013, "num_tokens": 567603351.0, "step": 14953 }, { "epoch": 1.902302506042488, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.775855541229248, "learning_rate": 1e-05, "loss": 0.4631, "mean_token_accuracy": 0.8830200433731079, "num_tokens": 567636710.0, "step": 14954 }, { "epoch": 1.9024297163210786, "ewc_loss": 0.10693359375, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.780738353729248, "learning_rate": 1e-05, "loss": 0.4464, "mean_token_accuracy": 0.8819308280944824, "num_tokens": 567676325.0, "step": 14955 }, { "epoch": 1.9025569265996691, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 7.065457344055176, "learning_rate": 1e-05, "loss": 0.4542, "mean_token_accuracy": 0.8893531560897827, "num_tokens": 567720663.0, "step": 14956 }, { "epoch": 1.9026841368782597, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.718384265899658, "learning_rate": 1e-05, "loss": 0.5245, "mean_token_accuracy": 0.8606173396110535, "num_tokens": 567761589.0, "step": 14957 }, { "epoch": 1.9028113471568502, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 0.00011873245239257812, "grad_norm": 7.385768890380859, "learning_rate": 1e-05, "loss": 0.4864, "mean_token_accuracy": 0.8802186250686646, "num_tokens": 567801649.0, "step": 14958 }, { "epoch": 1.9029385574354407, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 5.9409379959106445, "learning_rate": 1e-05, "loss": 0.4728, "mean_token_accuracy": 0.8781350255012512, "num_tokens": 567839953.0, "step": 14959 }, { "epoch": 1.9030657677140312, "ewc_loss": 0.1455078125, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 0.00012159347534179688, "grad_norm": 7.498129367828369, "learning_rate": 1e-05, "loss": 0.5041, "mean_token_accuracy": 0.8765062689781189, "num_tokens": 567877634.0, "step": 14960 }, { "epoch": 1.9031929779926218, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.291154384613037, "learning_rate": 1e-05, "loss": 0.5473, "mean_token_accuracy": 0.8558281064033508, "num_tokens": 567911504.0, "step": 14961 }, { "epoch": 1.9033201882712123, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 0.00011587142944335938, "grad_norm": 7.282384395599365, "learning_rate": 1e-05, "loss": 0.5125, "mean_token_accuracy": 0.874217689037323, "num_tokens": 567954904.0, "step": 14962 }, { "epoch": 1.9034473985498028, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.431703567504883, "learning_rate": 1e-05, "loss": 0.5357, "mean_token_accuracy": 0.862305760383606, "num_tokens": 567992384.0, "step": 14963 }, { "epoch": 1.9035746088283934, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.9567108154296875, "learning_rate": 1e-05, "loss": 0.5305, "mean_token_accuracy": 0.8662407398223877, "num_tokens": 568027692.0, "step": 14964 }, { "epoch": 1.9037018191069839, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.413227558135986, "learning_rate": 1e-05, "loss": 0.5179, "mean_token_accuracy": 0.862822413444519, "num_tokens": 568063385.0, "step": 14965 }, { "epoch": 1.9038290293855744, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.657705783843994, "learning_rate": 1e-05, "loss": 0.4756, "mean_token_accuracy": 0.8787755966186523, "num_tokens": 568098315.0, "step": 14966 }, { "epoch": 1.903956239664165, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.3070387840271, "learning_rate": 1e-05, "loss": 0.5399, "mean_token_accuracy": 0.8601443767547607, "num_tokens": 568144032.0, "step": 14967 }, { "epoch": 1.9040834499427555, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.494986534118652, "learning_rate": 1e-05, "loss": 0.5088, "mean_token_accuracy": 0.8669681549072266, "num_tokens": 568174988.0, "step": 14968 }, { "epoch": 1.904210660221346, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.188258171081543, "learning_rate": 1e-05, "loss": 0.5361, "mean_token_accuracy": 0.8573421239852905, "num_tokens": 568219794.0, "step": 14969 }, { "epoch": 1.9043378704999365, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.438572883605957, "learning_rate": 1e-05, "loss": 0.4817, "mean_token_accuracy": 0.8766816854476929, "num_tokens": 568264995.0, "step": 14970 }, { "epoch": 1.904465080778527, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 6.049986362457275, "learning_rate": 1e-05, "loss": 0.476, "mean_token_accuracy": 0.8729023933410645, "num_tokens": 568300876.0, "step": 14971 }, { "epoch": 1.9045922910571176, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.490556716918945, "learning_rate": 1e-05, "loss": 0.5179, "mean_token_accuracy": 0.8657512664794922, "num_tokens": 568337341.0, "step": 14972 }, { "epoch": 1.9047195013357079, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.894113063812256, "learning_rate": 1e-05, "loss": 0.4974, "mean_token_accuracy": 0.8649234771728516, "num_tokens": 568373524.0, "step": 14973 }, { "epoch": 1.9048467116142984, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.655104160308838, "learning_rate": 1e-05, "loss": 0.4534, "mean_token_accuracy": 0.8831766843795776, "num_tokens": 568412575.0, "step": 14974 }, { "epoch": 1.904973921892889, "ewc_loss": 0.10693359375, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.74976921081543, "learning_rate": 1e-05, "loss": 0.4786, "mean_token_accuracy": 0.8723762035369873, "num_tokens": 568446794.0, "step": 14975 }, { "epoch": 1.9051011321714795, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.934783935546875, "learning_rate": 1e-05, "loss": 0.5295, "mean_token_accuracy": 0.8651360273361206, "num_tokens": 568490645.0, "step": 14976 }, { "epoch": 1.90522834245007, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.701654434204102, "learning_rate": 1e-05, "loss": 0.4912, "mean_token_accuracy": 0.8700714111328125, "num_tokens": 568529281.0, "step": 14977 }, { "epoch": 1.9053555527286605, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 0.00011491775512695312, "grad_norm": 7.272058010101318, "learning_rate": 1e-05, "loss": 0.4999, "mean_token_accuracy": 0.8795315027236938, "num_tokens": 568568722.0, "step": 14978 }, { "epoch": 1.9054827630072508, "ewc_loss": 0.11474609375, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 5.864114284515381, "learning_rate": 1e-05, "loss": 0.4818, "mean_token_accuracy": 0.8734379410743713, "num_tokens": 568604736.0, "step": 14979 }, { "epoch": 1.9056099732858414, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 0.000118255615234375, "grad_norm": 7.437570571899414, "learning_rate": 1e-05, "loss": 0.4809, "mean_token_accuracy": 0.8831936120986938, "num_tokens": 568643720.0, "step": 14980 }, { "epoch": 1.905737183564432, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.218959331512451, "learning_rate": 1e-05, "loss": 0.4695, "mean_token_accuracy": 0.8809552788734436, "num_tokens": 568681805.0, "step": 14981 }, { "epoch": 1.9058643938430224, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 0.00011396408081054688, "grad_norm": 7.216315269470215, "learning_rate": 1e-05, "loss": 0.5299, "mean_token_accuracy": 0.8641579151153564, "num_tokens": 568723107.0, "step": 14982 }, { "epoch": 1.905991604121613, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.472332000732422, "learning_rate": 1e-05, "loss": 0.5697, "mean_token_accuracy": 0.8510477542877197, "num_tokens": 568760229.0, "step": 14983 }, { "epoch": 1.9061188144002035, "ewc_loss": 0.125, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.730645179748535, "learning_rate": 1e-05, "loss": 0.5512, "mean_token_accuracy": 0.856521725654602, "num_tokens": 568794539.0, "step": 14984 }, { "epoch": 1.906246024678794, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.5460381507873535, "learning_rate": 1e-05, "loss": 0.5399, "mean_token_accuracy": 0.8629326224327087, "num_tokens": 568834808.0, "step": 14985 }, { "epoch": 1.9063732349573845, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.229219436645508, "learning_rate": 1e-05, "loss": 0.4995, "mean_token_accuracy": 0.8697049617767334, "num_tokens": 568875004.0, "step": 14986 }, { "epoch": 1.906500445235975, "ewc_loss": 0.125, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.690485000610352, "learning_rate": 1e-05, "loss": 0.494, "mean_token_accuracy": 0.8723794221878052, "num_tokens": 568917372.0, "step": 14987 }, { "epoch": 1.9066276555145656, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.833868503570557, "learning_rate": 1e-05, "loss": 0.5029, "mean_token_accuracy": 0.8627218008041382, "num_tokens": 568953567.0, "step": 14988 }, { "epoch": 1.9067548657931561, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 7.0035400390625, "learning_rate": 1e-05, "loss": 0.5418, "mean_token_accuracy": 0.8592720627784729, "num_tokens": 568988634.0, "step": 14989 }, { "epoch": 1.9068820760717466, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.652571678161621, "learning_rate": 1e-05, "loss": 0.4706, "mean_token_accuracy": 0.8743883371353149, "num_tokens": 569023929.0, "step": 14990 }, { "epoch": 1.9070092863503372, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 0.00011920928955078125, "grad_norm": 7.420860290527344, "learning_rate": 1e-05, "loss": 0.4932, "mean_token_accuracy": 0.8777888417243958, "num_tokens": 569054483.0, "step": 14991 }, { "epoch": 1.9071364966289277, "ewc_loss": 0.11474609375, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 5.85281229019165, "learning_rate": 1e-05, "loss": 0.528, "mean_token_accuracy": 0.8584703207015991, "num_tokens": 569096919.0, "step": 14992 }, { "epoch": 1.9072637069075182, "ewc_loss": 0.1484375, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 0.0001239776611328125, "grad_norm": 7.645036697387695, "learning_rate": 1e-05, "loss": 0.5492, "mean_token_accuracy": 0.8648649454116821, "num_tokens": 569134161.0, "step": 14993 }, { "epoch": 1.9073909171861088, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.317195892333984, "learning_rate": 1e-05, "loss": 0.4968, "mean_token_accuracy": 0.8697264194488525, "num_tokens": 569173391.0, "step": 14994 }, { "epoch": 1.9075181274646993, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 0.0001201629638671875, "grad_norm": 7.437545299530029, "learning_rate": 1e-05, "loss": 0.5905, "mean_token_accuracy": 0.8509677052497864, "num_tokens": 569206173.0, "step": 14995 }, { "epoch": 1.9076453377432898, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.556168556213379, "learning_rate": 1e-05, "loss": 0.5204, "mean_token_accuracy": 0.8629088401794434, "num_tokens": 569241692.0, "step": 14996 }, { "epoch": 1.9077725480218801, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 7.0406694412231445, "learning_rate": 1e-05, "loss": 0.4874, "mean_token_accuracy": 0.8791370391845703, "num_tokens": 569276361.0, "step": 14997 }, { "epoch": 1.9078997583004706, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.5045881271362305, "learning_rate": 1e-05, "loss": 0.5829, "mean_token_accuracy": 0.8431587219238281, "num_tokens": 569314478.0, "step": 14998 }, { "epoch": 1.9080269685790612, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.697307586669922, "learning_rate": 1e-05, "loss": 0.4836, "mean_token_accuracy": 0.8729168176651001, "num_tokens": 569354867.0, "step": 14999 }, { "epoch": 1.9081541788576517, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.374258995056152, "learning_rate": 1e-05, "loss": 0.4883, "mean_token_accuracy": 0.8762152791023254, "num_tokens": 569394031.0, "step": 15000 }, { "epoch": 1.9082813891362422, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.502649784088135, "learning_rate": 1e-05, "loss": 0.4224, "mean_token_accuracy": 0.8947440385818481, "num_tokens": 569430899.0, "step": 15001 }, { "epoch": 1.9084085994148328, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.224355697631836, "learning_rate": 1e-05, "loss": 0.5445, "mean_token_accuracy": 0.8498167991638184, "num_tokens": 569469871.0, "step": 15002 }, { "epoch": 1.908535809693423, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.454914569854736, "learning_rate": 1e-05, "loss": 0.5031, "mean_token_accuracy": 0.8694825768470764, "num_tokens": 569508385.0, "step": 15003 }, { "epoch": 1.9086630199720136, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 6.0580363273620605, "learning_rate": 1e-05, "loss": 0.4933, "mean_token_accuracy": 0.8698012232780457, "num_tokens": 569549705.0, "step": 15004 }, { "epoch": 1.9087902302506041, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.539563179016113, "learning_rate": 1e-05, "loss": 0.5376, "mean_token_accuracy": 0.8579371571540833, "num_tokens": 569584092.0, "step": 15005 }, { "epoch": 1.9089174405291947, "ewc_loss": 0.10888671875, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.887911319732666, "learning_rate": 1e-05, "loss": 0.4765, "mean_token_accuracy": 0.8687987327575684, "num_tokens": 569619683.0, "step": 15006 }, { "epoch": 1.9090446508077852, "ewc_loss": 0.125, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.7142438888549805, "learning_rate": 1e-05, "loss": 0.4837, "mean_token_accuracy": 0.8774048089981079, "num_tokens": 569656347.0, "step": 15007 }, { "epoch": 1.9091718610863757, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.747137546539307, "learning_rate": 1e-05, "loss": 0.4637, "mean_token_accuracy": 0.8760979175567627, "num_tokens": 569689633.0, "step": 15008 }, { "epoch": 1.9092990713649662, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 7.023926258087158, "learning_rate": 1e-05, "loss": 0.5518, "mean_token_accuracy": 0.8560796976089478, "num_tokens": 569730685.0, "step": 15009 }, { "epoch": 1.9094262816435568, "ewc_loss": 0.10693359375, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.714078903198242, "learning_rate": 1e-05, "loss": 0.5303, "mean_token_accuracy": 0.8573818206787109, "num_tokens": 569773343.0, "step": 15010 }, { "epoch": 1.9095534919221473, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 0.000118255615234375, "grad_norm": 7.371496200561523, "learning_rate": 1e-05, "loss": 0.5264, "mean_token_accuracy": 0.8690563440322876, "num_tokens": 569814417.0, "step": 15011 }, { "epoch": 1.9096807022007378, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 5.951683044433594, "learning_rate": 1e-05, "loss": 0.4833, "mean_token_accuracy": 0.8760464787483215, "num_tokens": 569854436.0, "step": 15012 }, { "epoch": 1.9098079124793284, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 0.0001201629638671875, "grad_norm": 7.475693702697754, "learning_rate": 1e-05, "loss": 0.5756, "mean_token_accuracy": 0.852613627910614, "num_tokens": 569897530.0, "step": 15013 }, { "epoch": 1.9099351227579189, "ewc_loss": 0.125, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.3425421714782715, "learning_rate": 1e-05, "loss": 0.4938, "mean_token_accuracy": 0.8742388486862183, "num_tokens": 569931927.0, "step": 15014 }, { "epoch": 1.9100623330365094, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 0.00011301040649414062, "grad_norm": 7.171577453613281, "learning_rate": 1e-05, "loss": 0.5005, "mean_token_accuracy": 0.872283935546875, "num_tokens": 569968283.0, "step": 15015 }, { "epoch": 1.9101895433151, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.533231735229492, "learning_rate": 1e-05, "loss": 0.5167, "mean_token_accuracy": 0.8649662733078003, "num_tokens": 570009073.0, "step": 15016 }, { "epoch": 1.9103167535936905, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.664727687835693, "learning_rate": 1e-05, "loss": 0.5068, "mean_token_accuracy": 0.8682976365089417, "num_tokens": 570046421.0, "step": 15017 }, { "epoch": 1.910443963872281, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.595209121704102, "learning_rate": 1e-05, "loss": 0.5214, "mean_token_accuracy": 0.8646003007888794, "num_tokens": 570093679.0, "step": 15018 }, { "epoch": 1.9105711741508715, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 6.158247947692871, "learning_rate": 1e-05, "loss": 0.4655, "mean_token_accuracy": 0.8772341012954712, "num_tokens": 570132575.0, "step": 15019 }, { "epoch": 1.910698384429462, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.752173900604248, "learning_rate": 1e-05, "loss": 0.5061, "mean_token_accuracy": 0.8669575452804565, "num_tokens": 570169782.0, "step": 15020 }, { "epoch": 1.9108255947080524, "ewc_loss": 0.10693359375, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.819941997528076, "learning_rate": 1e-05, "loss": 0.4769, "mean_token_accuracy": 0.8746103644371033, "num_tokens": 570205034.0, "step": 15021 }, { "epoch": 1.9109528049866429, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 7.024928092956543, "learning_rate": 1e-05, "loss": 0.496, "mean_token_accuracy": 0.8708533048629761, "num_tokens": 570236804.0, "step": 15022 }, { "epoch": 1.9110800152652334, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.709405899047852, "learning_rate": 1e-05, "loss": 0.4707, "mean_token_accuracy": 0.8710877299308777, "num_tokens": 570266893.0, "step": 15023 }, { "epoch": 1.911207225543824, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 0.000118255615234375, "grad_norm": 7.377089023590088, "learning_rate": 1e-05, "loss": 0.4926, "mean_token_accuracy": 0.8800239562988281, "num_tokens": 570297779.0, "step": 15024 }, { "epoch": 1.9113344358224145, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 5.885669708251953, "learning_rate": 1e-05, "loss": 0.4883, "mean_token_accuracy": 0.8739253282546997, "num_tokens": 570334774.0, "step": 15025 }, { "epoch": 1.911461646101005, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 0.0001220703125, "grad_norm": 7.5593414306640625, "learning_rate": 1e-05, "loss": 0.5366, "mean_token_accuracy": 0.863411545753479, "num_tokens": 570373788.0, "step": 15026 }, { "epoch": 1.9115888563795955, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.276573181152344, "learning_rate": 1e-05, "loss": 0.4873, "mean_token_accuracy": 0.8746145963668823, "num_tokens": 570412156.0, "step": 15027 }, { "epoch": 1.9117160666581858, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 0.00011777877807617188, "grad_norm": 7.384041786193848, "learning_rate": 1e-05, "loss": 0.5081, "mean_token_accuracy": 0.8755816221237183, "num_tokens": 570442702.0, "step": 15028 }, { "epoch": 1.9118432769367764, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.488768577575684, "learning_rate": 1e-05, "loss": 0.5561, "mean_token_accuracy": 0.8564987182617188, "num_tokens": 570485302.0, "step": 15029 }, { "epoch": 1.9119704872153669, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 7.012178897857666, "learning_rate": 1e-05, "loss": 0.497, "mean_token_accuracy": 0.8729899525642395, "num_tokens": 570524820.0, "step": 15030 }, { "epoch": 1.9120976974939574, "ewc_loss": 0.125, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.470775127410889, "learning_rate": 1e-05, "loss": 0.5484, "mean_token_accuracy": 0.8571687936782837, "num_tokens": 570567973.0, "step": 15031 }, { "epoch": 1.912224907772548, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.68773889541626, "learning_rate": 1e-05, "loss": 0.4828, "mean_token_accuracy": 0.8746789693832397, "num_tokens": 570610369.0, "step": 15032 }, { "epoch": 1.9123521180511385, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.360196590423584, "learning_rate": 1e-05, "loss": 0.5223, "mean_token_accuracy": 0.8634526133537292, "num_tokens": 570653226.0, "step": 15033 }, { "epoch": 1.912479328329729, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.507138252258301, "learning_rate": 1e-05, "loss": 0.5213, "mean_token_accuracy": 0.8648964166641235, "num_tokens": 570691505.0, "step": 15034 }, { "epoch": 1.9126065386083195, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.21822452545166, "learning_rate": 1e-05, "loss": 0.5235, "mean_token_accuracy": 0.8609068393707275, "num_tokens": 570733781.0, "step": 15035 }, { "epoch": 1.91273374888691, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.44268798828125, "learning_rate": 1e-05, "loss": 0.4861, "mean_token_accuracy": 0.8744754195213318, "num_tokens": 570772675.0, "step": 15036 }, { "epoch": 1.9128609591655006, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 6.051896095275879, "learning_rate": 1e-05, "loss": 0.4432, "mean_token_accuracy": 0.8834216594696045, "num_tokens": 570813967.0, "step": 15037 }, { "epoch": 1.9129881694440911, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.509941101074219, "learning_rate": 1e-05, "loss": 0.4661, "mean_token_accuracy": 0.8786746263504028, "num_tokens": 570850315.0, "step": 15038 }, { "epoch": 1.9131153797226816, "ewc_loss": 0.109375, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.901757717132568, "learning_rate": 1e-05, "loss": 0.509, "mean_token_accuracy": 0.8619078397750854, "num_tokens": 570890191.0, "step": 15039 }, { "epoch": 1.9132425900012722, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.672900676727295, "learning_rate": 1e-05, "loss": 0.4918, "mean_token_accuracy": 0.8728458285331726, "num_tokens": 570926564.0, "step": 15040 }, { "epoch": 1.9133698002798627, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.752179145812988, "learning_rate": 1e-05, "loss": 0.4746, "mean_token_accuracy": 0.8719894886016846, "num_tokens": 570966334.0, "step": 15041 }, { "epoch": 1.9134970105584532, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.974728107452393, "learning_rate": 1e-05, "loss": 0.5341, "mean_token_accuracy": 0.8629781603813171, "num_tokens": 571008008.0, "step": 15042 }, { "epoch": 1.9136242208370438, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.695614814758301, "learning_rate": 1e-05, "loss": 0.5298, "mean_token_accuracy": 0.8585774898529053, "num_tokens": 571046915.0, "step": 15043 }, { "epoch": 1.9137514311156343, "ewc_loss": 0.140625, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 0.0001163482666015625, "grad_norm": 7.336305141448975, "learning_rate": 1e-05, "loss": 0.5222, "mean_token_accuracy": 0.8692482709884644, "num_tokens": 571085945.0, "step": 15044 }, { "epoch": 1.9138786413942248, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 5.888383388519287, "learning_rate": 1e-05, "loss": 0.4522, "mean_token_accuracy": 0.884158194065094, "num_tokens": 571119839.0, "step": 15045 }, { "epoch": 1.9140058516728151, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 0.00012063980102539062, "grad_norm": 7.506472110748291, "learning_rate": 1e-05, "loss": 0.4948, "mean_token_accuracy": 0.8805934190750122, "num_tokens": 571156888.0, "step": 15046 }, { "epoch": 1.9141330619514056, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.286416530609131, "learning_rate": 1e-05, "loss": 0.4432, "mean_token_accuracy": 0.8892803192138672, "num_tokens": 571197022.0, "step": 15047 }, { "epoch": 1.9142602722299962, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 0.00011491775512695312, "grad_norm": 7.252315998077393, "learning_rate": 1e-05, "loss": 0.5364, "mean_token_accuracy": 0.8621550798416138, "num_tokens": 571234780.0, "step": 15048 }, { "epoch": 1.9143874825085867, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.535181522369385, "learning_rate": 1e-05, "loss": 0.4957, "mean_token_accuracy": 0.8730341196060181, "num_tokens": 571272804.0, "step": 15049 }, { "epoch": 1.9145146927871772, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.723517894744873, "learning_rate": 1e-05, "loss": 0.5061, "mean_token_accuracy": 0.8673973679542542, "num_tokens": 571315455.0, "step": 15050 }, { "epoch": 1.9146419030657678, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.602146625518799, "learning_rate": 1e-05, "loss": 0.4889, "mean_token_accuracy": 0.8753941059112549, "num_tokens": 571350522.0, "step": 15051 }, { "epoch": 1.914769113344358, "ewc_loss": 0.11474609375, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.20412540435791, "learning_rate": 1e-05, "loss": 0.4668, "mean_token_accuracy": 0.8760820627212524, "num_tokens": 571384452.0, "step": 15052 }, { "epoch": 1.9148963236229486, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.727588176727295, "learning_rate": 1e-05, "loss": 0.4577, "mean_token_accuracy": 0.886538028717041, "num_tokens": 571423041.0, "step": 15053 }, { "epoch": 1.9150235339015391, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.831444263458252, "learning_rate": 1e-05, "loss": 0.4409, "mean_token_accuracy": 0.8844308257102966, "num_tokens": 571454744.0, "step": 15054 }, { "epoch": 1.9151507441801296, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 7.020986080169678, "learning_rate": 1e-05, "loss": 0.4746, "mean_token_accuracy": 0.879826545715332, "num_tokens": 571492937.0, "step": 15055 }, { "epoch": 1.9152779544587202, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.6896185874938965, "learning_rate": 1e-05, "loss": 0.443, "mean_token_accuracy": 0.8835792541503906, "num_tokens": 571531756.0, "step": 15056 }, { "epoch": 1.9154051647373107, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 0.00011873245239257812, "grad_norm": 7.381338596343994, "learning_rate": 1e-05, "loss": 0.5394, "mean_token_accuracy": 0.859767735004425, "num_tokens": 571566095.0, "step": 15057 }, { "epoch": 1.9155323750159012, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 5.88607120513916, "learning_rate": 1e-05, "loss": 0.4777, "mean_token_accuracy": 0.874571681022644, "num_tokens": 571599169.0, "step": 15058 }, { "epoch": 1.9156595852944918, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 0.00012111663818359375, "grad_norm": 7.5757856369018555, "learning_rate": 1e-05, "loss": 0.4948, "mean_token_accuracy": 0.8787992000579834, "num_tokens": 571632063.0, "step": 15059 }, { "epoch": 1.9157867955730823, "ewc_loss": 0.125, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.279909610748291, "learning_rate": 1e-05, "loss": 0.5061, "mean_token_accuracy": 0.8720690011978149, "num_tokens": 571670867.0, "step": 15060 }, { "epoch": 1.9159140058516728, "ewc_loss": 0.140625, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 0.00011730194091796875, "grad_norm": 7.3765668869018555, "learning_rate": 1e-05, "loss": 0.5207, "mean_token_accuracy": 0.8695660829544067, "num_tokens": 571710445.0, "step": 15061 }, { "epoch": 1.9160412161302633, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.482757568359375, "learning_rate": 1e-05, "loss": 0.4472, "mean_token_accuracy": 0.888480544090271, "num_tokens": 571748410.0, "step": 15062 }, { "epoch": 1.9161684264088539, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 7.0205607414245605, "learning_rate": 1e-05, "loss": 0.5549, "mean_token_accuracy": 0.8568699359893799, "num_tokens": 571786386.0, "step": 15063 }, { "epoch": 1.9162956366874444, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.454872131347656, "learning_rate": 1e-05, "loss": 0.5275, "mean_token_accuracy": 0.8643359541893005, "num_tokens": 571824414.0, "step": 15064 }, { "epoch": 1.916422846966035, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.707793712615967, "learning_rate": 1e-05, "loss": 0.5648, "mean_token_accuracy": 0.8492575287818909, "num_tokens": 571863085.0, "step": 15065 }, { "epoch": 1.9165500572446255, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.354499816894531, "learning_rate": 1e-05, "loss": 0.5062, "mean_token_accuracy": 0.8677918314933777, "num_tokens": 571894749.0, "step": 15066 }, { "epoch": 1.916677267523216, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.00024127960205078125, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.526434421539307, "learning_rate": 1e-05, "loss": 0.5179, "mean_token_accuracy": 0.8594803810119629, "num_tokens": 571925962.0, "step": 15067 }, { "epoch": 1.9168044778018065, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.19141960144043, "learning_rate": 1e-05, "loss": 0.519, "mean_token_accuracy": 0.8598257899284363, "num_tokens": 571966021.0, "step": 15068 }, { "epoch": 1.916931688080397, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.478837490081787, "learning_rate": 1e-05, "loss": 0.4727, "mean_token_accuracy": 0.880222499370575, "num_tokens": 572002684.0, "step": 15069 }, { "epoch": 1.9170588983589874, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 6.026151180267334, "learning_rate": 1e-05, "loss": 0.4884, "mean_token_accuracy": 0.8713535070419312, "num_tokens": 572038054.0, "step": 15070 }, { "epoch": 1.9171861086375779, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.00024127960205078125, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.5641255378723145, "learning_rate": 1e-05, "loss": 0.507, "mean_token_accuracy": 0.8671583533287048, "num_tokens": 572076851.0, "step": 15071 }, { "epoch": 1.9173133189161684, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.00024127960205078125, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.863877773284912, "learning_rate": 1e-05, "loss": 0.4664, "mean_token_accuracy": 0.8742673397064209, "num_tokens": 572115642.0, "step": 15072 }, { "epoch": 1.917440529194759, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.00024127960205078125, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.765067100524902, "learning_rate": 1e-05, "loss": 0.4791, "mean_token_accuracy": 0.8784036636352539, "num_tokens": 572154803.0, "step": 15073 }, { "epoch": 1.9175677394733495, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.00024127960205078125, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.746485233306885, "learning_rate": 1e-05, "loss": 0.4953, "mean_token_accuracy": 0.8655408620834351, "num_tokens": 572189241.0, "step": 15074 }, { "epoch": 1.91769494975194, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.00024127960205078125, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 7.058028697967529, "learning_rate": 1e-05, "loss": 0.475, "mean_token_accuracy": 0.8840367794036865, "num_tokens": 572230464.0, "step": 15075 }, { "epoch": 1.9178221600305305, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.00024127960205078125, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.770204544067383, "learning_rate": 1e-05, "loss": 0.4658, "mean_token_accuracy": 0.8810800909996033, "num_tokens": 572268863.0, "step": 15076 }, { "epoch": 1.9179493703091208, "ewc_loss": 0.140625, "ewc_loss_diag": 0.00024127960205078125, "ewc_loss_parallel": 0.0001163482666015625, "grad_norm": 7.333531379699707, "learning_rate": 1e-05, "loss": 0.5283, "mean_token_accuracy": 0.8659133911132812, "num_tokens": 572307484.0, "step": 15077 }, { "epoch": 1.9180765805877114, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.00024127960205078125, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.003976821899414, "learning_rate": 1e-05, "loss": 0.4961, "mean_token_accuracy": 0.8683410882949829, "num_tokens": 572347653.0, "step": 15078 }, { "epoch": 1.9182037908663019, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.00024127960205078125, "ewc_loss_parallel": 0.00011730194091796875, "grad_norm": 7.362937927246094, "learning_rate": 1e-05, "loss": 0.5687, "mean_token_accuracy": 0.8569738864898682, "num_tokens": 572386850.0, "step": 15079 }, { "epoch": 1.9183310011448924, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.00024127960205078125, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.332106113433838, "learning_rate": 1e-05, "loss": 0.5471, "mean_token_accuracy": 0.859460175037384, "num_tokens": 572422416.0, "step": 15080 }, { "epoch": 1.918458211423483, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.00024127960205078125, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 7.017706394195557, "learning_rate": 1e-05, "loss": 0.5378, "mean_token_accuracy": 0.8586472272872925, "num_tokens": 572458845.0, "step": 15081 }, { "epoch": 1.9185854217020735, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.00024127960205078125, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.495029926300049, "learning_rate": 1e-05, "loss": 0.5031, "mean_token_accuracy": 0.8712800145149231, "num_tokens": 572500553.0, "step": 15082 }, { "epoch": 1.918712631980664, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.00024127960205078125, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.506008625030518, "learning_rate": 1e-05, "loss": 0.5153, "mean_token_accuracy": 0.8616712093353271, "num_tokens": 572541315.0, "step": 15083 }, { "epoch": 1.9188398422592545, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.00024127960205078125, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.594939231872559, "learning_rate": 1e-05, "loss": 0.4751, "mean_token_accuracy": 0.878189742565155, "num_tokens": 572574461.0, "step": 15084 }, { "epoch": 1.918967052537845, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.00024127960205078125, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 6.0200324058532715, "learning_rate": 1e-05, "loss": 0.5174, "mean_token_accuracy": 0.8609490394592285, "num_tokens": 572616860.0, "step": 15085 }, { "epoch": 1.9190942628164356, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.00024127960205078125, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.818886756896973, "learning_rate": 1e-05, "loss": 0.4864, "mean_token_accuracy": 0.8758016228675842, "num_tokens": 572654228.0, "step": 15086 }, { "epoch": 1.919221473095026, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.00024127960205078125, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.691065311431885, "learning_rate": 1e-05, "loss": 0.4859, "mean_token_accuracy": 0.8683511018753052, "num_tokens": 572693603.0, "step": 15087 }, { "epoch": 1.9193486833736166, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.00024127960205078125, "ewc_loss_parallel": 0.00011444091796875, "grad_norm": 7.208881855010986, "learning_rate": 1e-05, "loss": 0.5429, "mean_token_accuracy": 0.8627195954322815, "num_tokens": 572730292.0, "step": 15088 }, { "epoch": 1.9194758936522072, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.00024127960205078125, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.683752059936523, "learning_rate": 1e-05, "loss": 0.509, "mean_token_accuracy": 0.8645029067993164, "num_tokens": 572764348.0, "step": 15089 }, { "epoch": 1.9196031039307977, "ewc_loss": 0.1474609375, "ewc_loss_diag": 0.00024127960205078125, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.615474224090576, "learning_rate": 1e-05, "loss": 0.5058, "mean_token_accuracy": 0.8757736086845398, "num_tokens": 572798410.0, "step": 15090 }, { "epoch": 1.9197303142093882, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.00024127960205078125, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.067777633666992, "learning_rate": 1e-05, "loss": 0.5561, "mean_token_accuracy": 0.8561550378799438, "num_tokens": 572832067.0, "step": 15091 }, { "epoch": 1.9198575244879788, "ewc_loss": 0.1494140625, "ewc_loss_diag": 0.00024127960205078125, "ewc_loss_parallel": 0.00012493133544921875, "grad_norm": 7.670467853546143, "learning_rate": 1e-05, "loss": 0.5084, "mean_token_accuracy": 0.8749129772186279, "num_tokens": 572866676.0, "step": 15092 }, { "epoch": 1.9199847347665693, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.00024127960205078125, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.511720657348633, "learning_rate": 1e-05, "loss": 0.542, "mean_token_accuracy": 0.8591976761817932, "num_tokens": 572913631.0, "step": 15093 }, { "epoch": 1.9201119450451598, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.00024127960205078125, "ewc_loss_parallel": 0.00011444091796875, "grad_norm": 7.3050103187561035, "learning_rate": 1e-05, "loss": 0.5032, "mean_token_accuracy": 0.8752322196960449, "num_tokens": 572953154.0, "step": 15094 }, { "epoch": 1.9202391553237501, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.00024127960205078125, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.61378812789917, "learning_rate": 1e-05, "loss": 0.527, "mean_token_accuracy": 0.8634514808654785, "num_tokens": 572992259.0, "step": 15095 }, { "epoch": 1.9203663656023406, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.00024127960205078125, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.8759918212890625, "learning_rate": 1e-05, "loss": 0.4888, "mean_token_accuracy": 0.8708905577659607, "num_tokens": 573024464.0, "step": 15096 }, { "epoch": 1.9204935758809312, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.00024127960205078125, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.506296157836914, "learning_rate": 1e-05, "loss": 0.4853, "mean_token_accuracy": 0.8755241632461548, "num_tokens": 573060620.0, "step": 15097 }, { "epoch": 1.9206207861595217, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.00024127960205078125, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.577478408813477, "learning_rate": 1e-05, "loss": 0.511, "mean_token_accuracy": 0.8643470406532288, "num_tokens": 573098284.0, "step": 15098 }, { "epoch": 1.9207479964381122, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.00024127960205078125, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.3668107986450195, "learning_rate": 1e-05, "loss": 0.4926, "mean_token_accuracy": 0.8709735870361328, "num_tokens": 573136253.0, "step": 15099 }, { "epoch": 1.9208752067167028, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.00024127960205078125, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.429891109466553, "learning_rate": 1e-05, "loss": 0.5176, "mean_token_accuracy": 0.8703827857971191, "num_tokens": 573174843.0, "step": 15100 }, { "epoch": 1.921002416995293, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.00024127960205078125, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.205097675323486, "learning_rate": 1e-05, "loss": 0.5312, "mean_token_accuracy": 0.8571922779083252, "num_tokens": 573215759.0, "step": 15101 }, { "epoch": 1.9211296272738836, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.00024127960205078125, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.408470630645752, "learning_rate": 1e-05, "loss": 0.4914, "mean_token_accuracy": 0.8715340495109558, "num_tokens": 573255297.0, "step": 15102 }, { "epoch": 1.9212568375524741, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.00024127960205078125, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 6.025502681732178, "learning_rate": 1e-05, "loss": 0.4045, "mean_token_accuracy": 0.8980021476745605, "num_tokens": 573293152.0, "step": 15103 }, { "epoch": 1.9213840478310646, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.00024127960205078125, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.506998062133789, "learning_rate": 1e-05, "loss": 0.5048, "mean_token_accuracy": 0.8687319755554199, "num_tokens": 573330456.0, "step": 15104 }, { "epoch": 1.9215112581096552, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.00024127960205078125, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.855495929718018, "learning_rate": 1e-05, "loss": 0.5085, "mean_token_accuracy": 0.861332893371582, "num_tokens": 573371595.0, "step": 15105 }, { "epoch": 1.9216384683882457, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.00024127960205078125, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.715873718261719, "learning_rate": 1e-05, "loss": 0.5268, "mean_token_accuracy": 0.8645871877670288, "num_tokens": 573417414.0, "step": 15106 }, { "epoch": 1.9217656786668362, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.00024127960205078125, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.718284606933594, "learning_rate": 1e-05, "loss": 0.5098, "mean_token_accuracy": 0.8645930290222168, "num_tokens": 573452146.0, "step": 15107 }, { "epoch": 1.9218928889454268, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.00024127960205078125, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 7.063310623168945, "learning_rate": 1e-05, "loss": 0.5352, "mean_token_accuracy": 0.8620493412017822, "num_tokens": 573485469.0, "step": 15108 }, { "epoch": 1.9220200992240173, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.00024127960205078125, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.669713497161865, "learning_rate": 1e-05, "loss": 0.4817, "mean_token_accuracy": 0.8738605380058289, "num_tokens": 573524763.0, "step": 15109 }, { "epoch": 1.9221473095026078, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.00024127960205078125, "ewc_loss_parallel": 0.00011968612670898438, "grad_norm": 7.478217601776123, "learning_rate": 1e-05, "loss": 0.5378, "mean_token_accuracy": 0.8653931617736816, "num_tokens": 573559326.0, "step": 15110 }, { "epoch": 1.9222745197811983, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.00024127960205078125, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 5.944761276245117, "learning_rate": 1e-05, "loss": 0.5108, "mean_token_accuracy": 0.8653469085693359, "num_tokens": 573604700.0, "step": 15111 }, { "epoch": 1.9224017300597889, "ewc_loss": 0.1484375, "ewc_loss_diag": 0.00024127960205078125, "ewc_loss_parallel": 0.0001239776611328125, "grad_norm": 7.647747993469238, "learning_rate": 1e-05, "loss": 0.485, "mean_token_accuracy": 0.8822996616363525, "num_tokens": 573639570.0, "step": 15112 }, { "epoch": 1.9225289403383794, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.00024127960205078125, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.435214519500732, "learning_rate": 1e-05, "loss": 0.5267, "mean_token_accuracy": 0.8652291297912598, "num_tokens": 573669849.0, "step": 15113 }, { "epoch": 1.92265615061697, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.00024127960205078125, "ewc_loss_parallel": 0.00011444091796875, "grad_norm": 7.3004045486450195, "learning_rate": 1e-05, "loss": 0.5203, "mean_token_accuracy": 0.866520881652832, "num_tokens": 573702402.0, "step": 15114 }, { "epoch": 1.9227833608955605, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.00024127960205078125, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.627435684204102, "learning_rate": 1e-05, "loss": 0.5202, "mean_token_accuracy": 0.8654849529266357, "num_tokens": 573737428.0, "step": 15115 }, { "epoch": 1.922910571174151, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.00024127960205078125, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.774723529815674, "learning_rate": 1e-05, "loss": 0.4836, "mean_token_accuracy": 0.8760625720024109, "num_tokens": 573775111.0, "step": 15116 }, { "epoch": 1.9230377814527415, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.00024127960205078125, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.626154899597168, "learning_rate": 1e-05, "loss": 0.531, "mean_token_accuracy": 0.8607487082481384, "num_tokens": 573814550.0, "step": 15117 }, { "epoch": 1.923164991731332, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.00024127960205078125, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.3014607429504395, "learning_rate": 1e-05, "loss": 0.529, "mean_token_accuracy": 0.858446478843689, "num_tokens": 573852932.0, "step": 15118 }, { "epoch": 1.9232922020099223, "ewc_loss": 0.125, "ewc_loss_diag": 0.00024127960205078125, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.690842628479004, "learning_rate": 1e-05, "loss": 0.4801, "mean_token_accuracy": 0.8783798813819885, "num_tokens": 573887769.0, "step": 15119 }, { "epoch": 1.9234194122885129, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.00024127960205078125, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.957684516906738, "learning_rate": 1e-05, "loss": 0.4999, "mean_token_accuracy": 0.8638681173324585, "num_tokens": 573925258.0, "step": 15120 }, { "epoch": 1.9235466225671034, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.00024127960205078125, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.8683247566223145, "learning_rate": 1e-05, "loss": 0.4745, "mean_token_accuracy": 0.8803265690803528, "num_tokens": 573956690.0, "step": 15121 }, { "epoch": 1.923673832845694, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.00024127960205078125, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.797179222106934, "learning_rate": 1e-05, "loss": 0.5221, "mean_token_accuracy": 0.8565975427627563, "num_tokens": 573993822.0, "step": 15122 }, { "epoch": 1.9238010431242845, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.00024127960205078125, "ewc_loss_parallel": 0.00011014938354492188, "grad_norm": 7.119019031524658, "learning_rate": 1e-05, "loss": 0.5244, "mean_token_accuracy": 0.8682791590690613, "num_tokens": 574030087.0, "step": 15123 }, { "epoch": 1.923928253402875, "ewc_loss": 0.109375, "ewc_loss_diag": 0.00024127960205078125, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.799312591552734, "learning_rate": 1e-05, "loss": 0.4657, "mean_token_accuracy": 0.8759398460388184, "num_tokens": 574071964.0, "step": 15124 }, { "epoch": 1.9240554636814655, "ewc_loss": 0.140625, "ewc_loss_diag": 0.00024127960205078125, "ewc_loss_parallel": 0.00011682510375976562, "grad_norm": 7.37445068359375, "learning_rate": 1e-05, "loss": 0.5401, "mean_token_accuracy": 0.8638596534729004, "num_tokens": 574105706.0, "step": 15125 }, { "epoch": 1.9241826739600558, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.00024127960205078125, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.015519142150879, "learning_rate": 1e-05, "loss": 0.549, "mean_token_accuracy": 0.8522538542747498, "num_tokens": 574145798.0, "step": 15126 }, { "epoch": 1.9243098842386464, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.00024127960205078125, "ewc_loss_parallel": 0.00011920928955078125, "grad_norm": 7.418333530426025, "learning_rate": 1e-05, "loss": 0.4737, "mean_token_accuracy": 0.8851273655891418, "num_tokens": 574192631.0, "step": 15127 }, { "epoch": 1.9244370945172369, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.00024127960205078125, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.282094955444336, "learning_rate": 1e-05, "loss": 0.5589, "mean_token_accuracy": 0.8500547409057617, "num_tokens": 574228601.0, "step": 15128 }, { "epoch": 1.9245643047958274, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.00024127960205078125, "ewc_loss_parallel": 0.0001125335693359375, "grad_norm": 7.22175407409668, "learning_rate": 1e-05, "loss": 0.5055, "mean_token_accuracy": 0.8760947585105896, "num_tokens": 574266416.0, "step": 15129 }, { "epoch": 1.924691515074418, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.00024127960205078125, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.398447513580322, "learning_rate": 1e-05, "loss": 0.5447, "mean_token_accuracy": 0.8582625389099121, "num_tokens": 574297414.0, "step": 15130 }, { "epoch": 1.9248187253530085, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.00024127960205078125, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.914725303649902, "learning_rate": 1e-05, "loss": 0.4549, "mean_token_accuracy": 0.8869953155517578, "num_tokens": 574333749.0, "step": 15131 }, { "epoch": 1.924945935631599, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.00024127960205078125, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.368107795715332, "learning_rate": 1e-05, "loss": 0.5018, "mean_token_accuracy": 0.8684287071228027, "num_tokens": 574377186.0, "step": 15132 }, { "epoch": 1.9250731459101895, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.00024127960205078125, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.637448787689209, "learning_rate": 1e-05, "loss": 0.4814, "mean_token_accuracy": 0.874231219291687, "num_tokens": 574414814.0, "step": 15133 }, { "epoch": 1.92520035618878, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.00024127960205078125, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.2973198890686035, "learning_rate": 1e-05, "loss": 0.5171, "mean_token_accuracy": 0.8617486357688904, "num_tokens": 574453070.0, "step": 15134 }, { "epoch": 1.9253275664673706, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.00024127960205078125, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.477396488189697, "learning_rate": 1e-05, "loss": 0.4629, "mean_token_accuracy": 0.8763574957847595, "num_tokens": 574487564.0, "step": 15135 }, { "epoch": 1.925454776745961, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.00024127960205078125, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.1881866455078125, "learning_rate": 1e-05, "loss": 0.5181, "mean_token_accuracy": 0.8664776682853699, "num_tokens": 574528456.0, "step": 15136 }, { "epoch": 1.9255819870245516, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.418120384216309, "learning_rate": 1e-05, "loss": 0.497, "mean_token_accuracy": 0.8711041212081909, "num_tokens": 574567724.0, "step": 15137 }, { "epoch": 1.9257091973031422, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 6.069943428039551, "learning_rate": 1e-05, "loss": 0.4806, "mean_token_accuracy": 0.870421290397644, "num_tokens": 574599702.0, "step": 15138 }, { "epoch": 1.9258364075817327, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.453515529632568, "learning_rate": 1e-05, "loss": 0.4824, "mean_token_accuracy": 0.8752844333648682, "num_tokens": 574633709.0, "step": 15139 }, { "epoch": 1.9259636178603232, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.931070327758789, "learning_rate": 1e-05, "loss": 0.5028, "mean_token_accuracy": 0.8680474758148193, "num_tokens": 574670490.0, "step": 15140 }, { "epoch": 1.9260908281389137, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.577119827270508, "learning_rate": 1e-05, "loss": 0.5407, "mean_token_accuracy": 0.8547832369804382, "num_tokens": 574706563.0, "step": 15141 }, { "epoch": 1.9262180384175043, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.789515018463135, "learning_rate": 1e-05, "loss": 0.4837, "mean_token_accuracy": 0.8729874491691589, "num_tokens": 574744743.0, "step": 15142 }, { "epoch": 1.9263452486960948, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.8131537437438965, "learning_rate": 1e-05, "loss": 0.509, "mean_token_accuracy": 0.8678853511810303, "num_tokens": 574782236.0, "step": 15143 }, { "epoch": 1.926472458974685, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.678612232208252, "learning_rate": 1e-05, "loss": 0.4492, "mean_token_accuracy": 0.8819398880004883, "num_tokens": 574826457.0, "step": 15144 }, { "epoch": 1.9265996692532756, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 0.00011205673217773438, "grad_norm": 7.159214019775391, "learning_rate": 1e-05, "loss": 0.5376, "mean_token_accuracy": 0.8617525696754456, "num_tokens": 574870362.0, "step": 15145 }, { "epoch": 1.9267268795318662, "ewc_loss": 0.109375, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.74131441116333, "learning_rate": 1e-05, "loss": 0.4677, "mean_token_accuracy": 0.8761551380157471, "num_tokens": 574907984.0, "step": 15146 }, { "epoch": 1.9268540898104567, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 0.0001201629638671875, "grad_norm": 7.462851524353027, "learning_rate": 1e-05, "loss": 0.5498, "mean_token_accuracy": 0.8627254366874695, "num_tokens": 574950813.0, "step": 15147 }, { "epoch": 1.9269813000890472, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.086265563964844, "learning_rate": 1e-05, "loss": 0.4883, "mean_token_accuracy": 0.8732808828353882, "num_tokens": 574985210.0, "step": 15148 }, { "epoch": 1.9271085103676378, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 0.00011920928955078125, "grad_norm": 7.421270370483398, "learning_rate": 1e-05, "loss": 0.4838, "mean_token_accuracy": 0.87693190574646, "num_tokens": 575020263.0, "step": 15149 }, { "epoch": 1.927235720646228, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.42182731628418, "learning_rate": 1e-05, "loss": 0.5304, "mean_token_accuracy": 0.8666689395904541, "num_tokens": 575060632.0, "step": 15150 }, { "epoch": 1.9273629309248186, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 7.004724502563477, "learning_rate": 1e-05, "loss": 0.523, "mean_token_accuracy": 0.8659585118293762, "num_tokens": 575102725.0, "step": 15151 }, { "epoch": 1.9274901412034091, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.541136264801025, "learning_rate": 1e-05, "loss": 0.4568, "mean_token_accuracy": 0.8842712640762329, "num_tokens": 575143069.0, "step": 15152 }, { "epoch": 1.9276173514819996, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.498202800750732, "learning_rate": 1e-05, "loss": 0.4873, "mean_token_accuracy": 0.8722397089004517, "num_tokens": 575182618.0, "step": 15153 }, { "epoch": 1.9277445617605902, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.61681604385376, "learning_rate": 1e-05, "loss": 0.5081, "mean_token_accuracy": 0.8685696125030518, "num_tokens": 575224576.0, "step": 15154 }, { "epoch": 1.9278717720391807, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 6.050195693969727, "learning_rate": 1e-05, "loss": 0.4532, "mean_token_accuracy": 0.8804906606674194, "num_tokens": 575260152.0, "step": 15155 }, { "epoch": 1.9279989823177712, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.812722682952881, "learning_rate": 1e-05, "loss": 0.5061, "mean_token_accuracy": 0.8669842481613159, "num_tokens": 575292308.0, "step": 15156 }, { "epoch": 1.9281261925963618, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.777479648590088, "learning_rate": 1e-05, "loss": 0.4958, "mean_token_accuracy": 0.8637691736221313, "num_tokens": 575326305.0, "step": 15157 }, { "epoch": 1.9282534028749523, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 0.00011014938354492188, "grad_norm": 7.113025188446045, "learning_rate": 1e-05, "loss": 0.5277, "mean_token_accuracy": 0.8658798933029175, "num_tokens": 575361775.0, "step": 15158 }, { "epoch": 1.9283806131535428, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.732793807983398, "learning_rate": 1e-05, "loss": 0.4544, "mean_token_accuracy": 0.8785012364387512, "num_tokens": 575397508.0, "step": 15159 }, { "epoch": 1.9285078234321333, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 0.00011920928955078125, "grad_norm": 7.426102161407471, "learning_rate": 1e-05, "loss": 0.5164, "mean_token_accuracy": 0.8671259880065918, "num_tokens": 575437770.0, "step": 15160 }, { "epoch": 1.9286350337107239, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 5.982144355773926, "learning_rate": 1e-05, "loss": 0.492, "mean_token_accuracy": 0.8739747405052185, "num_tokens": 575478322.0, "step": 15161 }, { "epoch": 1.9287622439893144, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 0.00012063980102539062, "grad_norm": 7.51938009262085, "learning_rate": 1e-05, "loss": 0.5686, "mean_token_accuracy": 0.8577398657798767, "num_tokens": 575518816.0, "step": 15162 }, { "epoch": 1.928889454267905, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.335238933563232, "learning_rate": 1e-05, "loss": 0.5248, "mean_token_accuracy": 0.8616970181465149, "num_tokens": 575554276.0, "step": 15163 }, { "epoch": 1.9290166645464955, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 0.00011587142944335938, "grad_norm": 7.280400276184082, "learning_rate": 1e-05, "loss": 0.5383, "mean_token_accuracy": 0.8635307550430298, "num_tokens": 575591331.0, "step": 15164 }, { "epoch": 1.929143874825086, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.468404293060303, "learning_rate": 1e-05, "loss": 0.5252, "mean_token_accuracy": 0.8675401210784912, "num_tokens": 575629449.0, "step": 15165 }, { "epoch": 1.9292710851036765, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.9243645668029785, "learning_rate": 1e-05, "loss": 0.5509, "mean_token_accuracy": 0.8557577729225159, "num_tokens": 575668169.0, "step": 15166 }, { "epoch": 1.929398295382267, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.417372703552246, "learning_rate": 1e-05, "loss": 0.5373, "mean_token_accuracy": 0.8602137565612793, "num_tokens": 575708481.0, "step": 15167 }, { "epoch": 1.9295255056608573, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.645769119262695, "learning_rate": 1e-05, "loss": 0.5668, "mean_token_accuracy": 0.8483917117118835, "num_tokens": 575744879.0, "step": 15168 }, { "epoch": 1.9296527159394479, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.31392765045166, "learning_rate": 1e-05, "loss": 0.4336, "mean_token_accuracy": 0.8909590244293213, "num_tokens": 575781587.0, "step": 15169 }, { "epoch": 1.9297799262180384, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.4821553230285645, "learning_rate": 1e-05, "loss": 0.5104, "mean_token_accuracy": 0.867051362991333, "num_tokens": 575816662.0, "step": 15170 }, { "epoch": 1.929907136496629, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.178231239318848, "learning_rate": 1e-05, "loss": 0.5206, "mean_token_accuracy": 0.8641158938407898, "num_tokens": 575855169.0, "step": 15171 }, { "epoch": 1.9300343467752195, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.446347713470459, "learning_rate": 1e-05, "loss": 0.4714, "mean_token_accuracy": 0.8815029859542847, "num_tokens": 575896939.0, "step": 15172 }, { "epoch": 1.93016155705381, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 6.035090923309326, "learning_rate": 1e-05, "loss": 0.4855, "mean_token_accuracy": 0.873604416847229, "num_tokens": 575930587.0, "step": 15173 }, { "epoch": 1.9302887673324005, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.528256893157959, "learning_rate": 1e-05, "loss": 0.5467, "mean_token_accuracy": 0.8535959124565125, "num_tokens": 575964432.0, "step": 15174 }, { "epoch": 1.9304159776109908, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.888705730438232, "learning_rate": 1e-05, "loss": 0.4795, "mean_token_accuracy": 0.8724395036697388, "num_tokens": 576004916.0, "step": 15175 }, { "epoch": 1.9305431878895813, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.689804553985596, "learning_rate": 1e-05, "loss": 0.5688, "mean_token_accuracy": 0.8497483730316162, "num_tokens": 576047857.0, "step": 15176 }, { "epoch": 1.9306703981681719, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.758580207824707, "learning_rate": 1e-05, "loss": 0.5148, "mean_token_accuracy": 0.8596109747886658, "num_tokens": 576089087.0, "step": 15177 }, { "epoch": 1.9307976084467624, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.947505950927734, "learning_rate": 1e-05, "loss": 0.5088, "mean_token_accuracy": 0.8723717331886292, "num_tokens": 576134106.0, "step": 15178 }, { "epoch": 1.930924818725353, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.70660400390625, "learning_rate": 1e-05, "loss": 0.4916, "mean_token_accuracy": 0.8669195771217346, "num_tokens": 576172566.0, "step": 15179 }, { "epoch": 1.9310520290039435, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 0.00011587142944335938, "grad_norm": 7.29097318649292, "learning_rate": 1e-05, "loss": 0.6147, "mean_token_accuracy": 0.844178318977356, "num_tokens": 576214469.0, "step": 15180 }, { "epoch": 1.931179239282534, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 5.882778167724609, "learning_rate": 1e-05, "loss": 0.4243, "mean_token_accuracy": 0.8915736079216003, "num_tokens": 576249316.0, "step": 15181 }, { "epoch": 1.9313064495611245, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 0.00012063980102539062, "grad_norm": 7.4445085525512695, "learning_rate": 1e-05, "loss": 0.5824, "mean_token_accuracy": 0.8493562340736389, "num_tokens": 576290748.0, "step": 15182 }, { "epoch": 1.931433659839715, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.246715545654297, "learning_rate": 1e-05, "loss": 0.4997, "mean_token_accuracy": 0.871417224407196, "num_tokens": 576329137.0, "step": 15183 }, { "epoch": 1.9315608701183056, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 0.00011396408081054688, "grad_norm": 7.201745986938477, "learning_rate": 1e-05, "loss": 0.4803, "mean_token_accuracy": 0.8822874426841736, "num_tokens": 576367147.0, "step": 15184 }, { "epoch": 1.931688080396896, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.4817938804626465, "learning_rate": 1e-05, "loss": 0.5579, "mean_token_accuracy": 0.8611865043640137, "num_tokens": 576406068.0, "step": 15185 }, { "epoch": 1.9318152906754866, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.724227428436279, "learning_rate": 1e-05, "loss": 0.5268, "mean_token_accuracy": 0.8646930456161499, "num_tokens": 576438150.0, "step": 15186 }, { "epoch": 1.9319425009540772, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.568602085113525, "learning_rate": 1e-05, "loss": 0.495, "mean_token_accuracy": 0.8738484382629395, "num_tokens": 576470630.0, "step": 15187 }, { "epoch": 1.9320697112326677, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.196660995483398, "learning_rate": 1e-05, "loss": 0.4728, "mean_token_accuracy": 0.8753544092178345, "num_tokens": 576505896.0, "step": 15188 }, { "epoch": 1.9321969215112582, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.721033096313477, "learning_rate": 1e-05, "loss": 0.4751, "mean_token_accuracy": 0.8786602020263672, "num_tokens": 576545204.0, "step": 15189 }, { "epoch": 1.9323241317898487, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.821675777435303, "learning_rate": 1e-05, "loss": 0.5495, "mean_token_accuracy": 0.8578252792358398, "num_tokens": 576590746.0, "step": 15190 }, { "epoch": 1.9324513420684393, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 7.024888515472412, "learning_rate": 1e-05, "loss": 0.5198, "mean_token_accuracy": 0.8674145936965942, "num_tokens": 576631185.0, "step": 15191 }, { "epoch": 1.9325785523470298, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.680111408233643, "learning_rate": 1e-05, "loss": 0.5053, "mean_token_accuracy": 0.8670640587806702, "num_tokens": 576668945.0, "step": 15192 }, { "epoch": 1.93270576262562, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 0.00011873245239257812, "grad_norm": 7.423941612243652, "learning_rate": 1e-05, "loss": 0.5237, "mean_token_accuracy": 0.869523286819458, "num_tokens": 576706052.0, "step": 15193 }, { "epoch": 1.9328329729042106, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 5.88455057144165, "learning_rate": 1e-05, "loss": 0.4957, "mean_token_accuracy": 0.869123637676239, "num_tokens": 576740668.0, "step": 15194 }, { "epoch": 1.9329601831828012, "ewc_loss": 0.1484375, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 0.0001239776611328125, "grad_norm": 7.61953067779541, "learning_rate": 1e-05, "loss": 0.5591, "mean_token_accuracy": 0.8589795827865601, "num_tokens": 576780312.0, "step": 15195 }, { "epoch": 1.9330873934613917, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.303314685821533, "learning_rate": 1e-05, "loss": 0.4666, "mean_token_accuracy": 0.8817205429077148, "num_tokens": 576817130.0, "step": 15196 }, { "epoch": 1.9332146037399822, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 0.0001201629638671875, "grad_norm": 7.429209232330322, "learning_rate": 1e-05, "loss": 0.4702, "mean_token_accuracy": 0.8841737508773804, "num_tokens": 576849823.0, "step": 15197 }, { "epoch": 1.9333418140185727, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.516687870025635, "learning_rate": 1e-05, "loss": 0.5267, "mean_token_accuracy": 0.861185610294342, "num_tokens": 576890890.0, "step": 15198 }, { "epoch": 1.933469024297163, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 0.00010967254638671875, "grad_norm": 7.079867362976074, "learning_rate": 1e-05, "loss": 0.5521, "mean_token_accuracy": 0.8543370366096497, "num_tokens": 576927926.0, "step": 15199 }, { "epoch": 1.9335962345757536, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.492447376251221, "learning_rate": 1e-05, "loss": 0.5458, "mean_token_accuracy": 0.8588547706604004, "num_tokens": 576966194.0, "step": 15200 }, { "epoch": 1.933723444854344, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.7605671882629395, "learning_rate": 1e-05, "loss": 0.5466, "mean_token_accuracy": 0.8579690456390381, "num_tokens": 577007218.0, "step": 15201 }, { "epoch": 1.9338506551329346, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.373019218444824, "learning_rate": 1e-05, "loss": 0.447, "mean_token_accuracy": 0.887181282043457, "num_tokens": 577040551.0, "step": 15202 }, { "epoch": 1.9339778654115252, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.559943199157715, "learning_rate": 1e-05, "loss": 0.509, "mean_token_accuracy": 0.8711602091789246, "num_tokens": 577082676.0, "step": 15203 }, { "epoch": 1.9341050756901157, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.22038459777832, "learning_rate": 1e-05, "loss": 0.48, "mean_token_accuracy": 0.8752373456954956, "num_tokens": 577124572.0, "step": 15204 }, { "epoch": 1.9342322859687062, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.503324031829834, "learning_rate": 1e-05, "loss": 0.4682, "mean_token_accuracy": 0.8814404010772705, "num_tokens": 577167571.0, "step": 15205 }, { "epoch": 1.9343594962472968, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 6.059370517730713, "learning_rate": 1e-05, "loss": 0.5447, "mean_token_accuracy": 0.8505681753158569, "num_tokens": 577211976.0, "step": 15206 }, { "epoch": 1.9344867065258873, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.568155765533447, "learning_rate": 1e-05, "loss": 0.4779, "mean_token_accuracy": 0.8766608238220215, "num_tokens": 577252572.0, "step": 15207 }, { "epoch": 1.9346139168044778, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.908365726470947, "learning_rate": 1e-05, "loss": 0.5256, "mean_token_accuracy": 0.8551658391952515, "num_tokens": 577286923.0, "step": 15208 }, { "epoch": 1.9347411270830683, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.7287726402282715, "learning_rate": 1e-05, "loss": 0.511, "mean_token_accuracy": 0.8668696880340576, "num_tokens": 577324520.0, "step": 15209 }, { "epoch": 1.9348683373616589, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.78200101852417, "learning_rate": 1e-05, "loss": 0.488, "mean_token_accuracy": 0.864737868309021, "num_tokens": 577361100.0, "step": 15210 }, { "epoch": 1.9349955476402494, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.978333950042725, "learning_rate": 1e-05, "loss": 0.5126, "mean_token_accuracy": 0.8674837350845337, "num_tokens": 577402714.0, "step": 15211 }, { "epoch": 1.93512275791884, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.7668609619140625, "learning_rate": 1e-05, "loss": 0.5386, "mean_token_accuracy": 0.858414888381958, "num_tokens": 577437226.0, "step": 15212 }, { "epoch": 1.9352499681974304, "ewc_loss": 0.140625, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 0.0001163482666015625, "grad_norm": 7.266358375549316, "learning_rate": 1e-05, "loss": 0.5258, "mean_token_accuracy": 0.8726929426193237, "num_tokens": 577475564.0, "step": 15213 }, { "epoch": 1.935377178476021, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 5.9393229484558105, "learning_rate": 1e-05, "loss": 0.5707, "mean_token_accuracy": 0.8508168458938599, "num_tokens": 577512623.0, "step": 15214 }, { "epoch": 1.9355043887546115, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 0.00011777877807617188, "grad_norm": 7.361897945404053, "learning_rate": 1e-05, "loss": 0.5217, "mean_token_accuracy": 0.8699979782104492, "num_tokens": 577545515.0, "step": 15215 }, { "epoch": 1.935631599033202, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.256990909576416, "learning_rate": 1e-05, "loss": 0.5002, "mean_token_accuracy": 0.8696044683456421, "num_tokens": 577577043.0, "step": 15216 }, { "epoch": 1.9357588093117923, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 0.00011110305786132812, "grad_norm": 7.091170787811279, "learning_rate": 1e-05, "loss": 0.4939, "mean_token_accuracy": 0.8767035007476807, "num_tokens": 577616311.0, "step": 15217 }, { "epoch": 1.9358860195903829, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.445046901702881, "learning_rate": 1e-05, "loss": 0.4851, "mean_token_accuracy": 0.8729056119918823, "num_tokens": 577654963.0, "step": 15218 }, { "epoch": 1.9360132298689734, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.613031387329102, "learning_rate": 1e-05, "loss": 0.4661, "mean_token_accuracy": 0.8797482252120972, "num_tokens": 577688283.0, "step": 15219 }, { "epoch": 1.936140440147564, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.555249214172363, "learning_rate": 1e-05, "loss": 0.5307, "mean_token_accuracy": 0.8606807589530945, "num_tokens": 577726855.0, "step": 15220 }, { "epoch": 1.9362676504261545, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.123292446136475, "learning_rate": 1e-05, "loss": 0.5471, "mean_token_accuracy": 0.857282817363739, "num_tokens": 577765699.0, "step": 15221 }, { "epoch": 1.936394860704745, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.752148151397705, "learning_rate": 1e-05, "loss": 0.4379, "mean_token_accuracy": 0.8927266597747803, "num_tokens": 577797393.0, "step": 15222 }, { "epoch": 1.9365220709833355, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.730747222900391, "learning_rate": 1e-05, "loss": 0.4687, "mean_token_accuracy": 0.8777971267700195, "num_tokens": 577838768.0, "step": 15223 }, { "epoch": 1.9366492812619258, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 0.00011110305786132812, "grad_norm": 7.130562782287598, "learning_rate": 1e-05, "loss": 0.5109, "mean_token_accuracy": 0.8722292184829712, "num_tokens": 577882455.0, "step": 15224 }, { "epoch": 1.9367764915405163, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.624732971191406, "learning_rate": 1e-05, "loss": 0.4437, "mean_token_accuracy": 0.8839643001556396, "num_tokens": 577922649.0, "step": 15225 }, { "epoch": 1.9369037018191069, "ewc_loss": 0.1474609375, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.600760459899902, "learning_rate": 1e-05, "loss": 0.5239, "mean_token_accuracy": 0.871088445186615, "num_tokens": 577961679.0, "step": 15226 }, { "epoch": 1.9370309120976974, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 5.975175857543945, "learning_rate": 1e-05, "loss": 0.4772, "mean_token_accuracy": 0.8749195337295532, "num_tokens": 578001848.0, "step": 15227 }, { "epoch": 1.937158122376288, "ewc_loss": 0.1533203125, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 0.00012874603271484375, "grad_norm": 7.766101360321045, "learning_rate": 1e-05, "loss": 0.5168, "mean_token_accuracy": 0.8735274076461792, "num_tokens": 578039278.0, "step": 15228 }, { "epoch": 1.9372853326548785, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 6.5182390213012695, "learning_rate": 1e-05, "loss": 0.5182, "mean_token_accuracy": 0.8680189847946167, "num_tokens": 578078152.0, "step": 15229 }, { "epoch": 1.937412542933469, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 0.00011873245239257812, "grad_norm": 7.406322002410889, "learning_rate": 1e-05, "loss": 0.5448, "mean_token_accuracy": 0.8634312152862549, "num_tokens": 578115927.0, "step": 15230 }, { "epoch": 1.9375397532120595, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 6.654682159423828, "learning_rate": 1e-05, "loss": 0.4996, "mean_token_accuracy": 0.8733338117599487, "num_tokens": 578157425.0, "step": 15231 }, { "epoch": 1.93766696349065, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.976800441741943, "learning_rate": 1e-05, "loss": 0.4831, "mean_token_accuracy": 0.8759318590164185, "num_tokens": 578201985.0, "step": 15232 }, { "epoch": 1.9377941737692406, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.551624298095703, "learning_rate": 1e-05, "loss": 0.4985, "mean_token_accuracy": 0.8692145347595215, "num_tokens": 578239471.0, "step": 15233 }, { "epoch": 1.937921384047831, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.651712417602539, "learning_rate": 1e-05, "loss": 0.4769, "mean_token_accuracy": 0.8777834177017212, "num_tokens": 578272302.0, "step": 15234 }, { "epoch": 1.9380485943264216, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.405375957489014, "learning_rate": 1e-05, "loss": 0.5004, "mean_token_accuracy": 0.8677226901054382, "num_tokens": 578314585.0, "step": 15235 }, { "epoch": 1.9381758046050122, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.481449127197266, "learning_rate": 1e-05, "loss": 0.5581, "mean_token_accuracy": 0.8492375612258911, "num_tokens": 578355561.0, "step": 15236 }, { "epoch": 1.9383030148836027, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.223104476928711, "learning_rate": 1e-05, "loss": 0.5052, "mean_token_accuracy": 0.8663976788520813, "num_tokens": 578398903.0, "step": 15237 }, { "epoch": 1.9384302251621932, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.463772773742676, "learning_rate": 1e-05, "loss": 0.5914, "mean_token_accuracy": 0.8430964946746826, "num_tokens": 578439293.0, "step": 15238 }, { "epoch": 1.9385574354407837, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 6.035586357116699, "learning_rate": 1e-05, "loss": 0.5369, "mean_token_accuracy": 0.8524947166442871, "num_tokens": 578477965.0, "step": 15239 }, { "epoch": 1.9386846457193743, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.555843830108643, "learning_rate": 1e-05, "loss": 0.4845, "mean_token_accuracy": 0.8746196031570435, "num_tokens": 578515577.0, "step": 15240 }, { "epoch": 1.9388118559979648, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.857377052307129, "learning_rate": 1e-05, "loss": 0.4873, "mean_token_accuracy": 0.8696354627609253, "num_tokens": 578557415.0, "step": 15241 }, { "epoch": 1.938939066276555, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.775285720825195, "learning_rate": 1e-05, "loss": 0.4908, "mean_token_accuracy": 0.8737651109695435, "num_tokens": 578594556.0, "step": 15242 }, { "epoch": 1.9390662765551456, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.709808826446533, "learning_rate": 1e-05, "loss": 0.4549, "mean_token_accuracy": 0.8792818188667297, "num_tokens": 578627219.0, "step": 15243 }, { "epoch": 1.9391934868337362, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 0.00011157989501953125, "grad_norm": 7.140530109405518, "learning_rate": 1e-05, "loss": 0.5613, "mean_token_accuracy": 0.8550418615341187, "num_tokens": 578669135.0, "step": 15244 }, { "epoch": 1.9393206971123267, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.704949855804443, "learning_rate": 1e-05, "loss": 0.4634, "mean_token_accuracy": 0.8750990629196167, "num_tokens": 578706174.0, "step": 15245 }, { "epoch": 1.9394479073909172, "ewc_loss": 0.1455078125, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 0.00012111663818359375, "grad_norm": 7.536957263946533, "learning_rate": 1e-05, "loss": 0.5179, "mean_token_accuracy": 0.8670206665992737, "num_tokens": 578738942.0, "step": 15246 }, { "epoch": 1.9395751176695077, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.051253318786621, "learning_rate": 1e-05, "loss": 0.4302, "mean_token_accuracy": 0.8906644582748413, "num_tokens": 578771553.0, "step": 15247 }, { "epoch": 1.939702327948098, "ewc_loss": 0.1474609375, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.579998016357422, "learning_rate": 1e-05, "loss": 0.5414, "mean_token_accuracy": 0.865226149559021, "num_tokens": 578811532.0, "step": 15248 }, { "epoch": 1.9398295382266886, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.495638370513916, "learning_rate": 1e-05, "loss": 0.5171, "mean_token_accuracy": 0.8671547174453735, "num_tokens": 578852858.0, "step": 15249 }, { "epoch": 1.939956748505279, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 0.00011110305786132812, "grad_norm": 7.220062732696533, "learning_rate": 1e-05, "loss": 0.5366, "mean_token_accuracy": 0.8651818037033081, "num_tokens": 578890340.0, "step": 15250 }, { "epoch": 1.9400839587838696, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.654004096984863, "learning_rate": 1e-05, "loss": 0.4825, "mean_token_accuracy": 0.877246081829071, "num_tokens": 578928953.0, "step": 15251 }, { "epoch": 1.9402111690624602, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.637995719909668, "learning_rate": 1e-05, "loss": 0.4631, "mean_token_accuracy": 0.8789066672325134, "num_tokens": 578967978.0, "step": 15252 }, { "epoch": 1.9403383793410507, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.662615776062012, "learning_rate": 1e-05, "loss": 0.5, "mean_token_accuracy": 0.8763020038604736, "num_tokens": 579005484.0, "step": 15253 }, { "epoch": 1.9404655896196412, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.167177677154541, "learning_rate": 1e-05, "loss": 0.5503, "mean_token_accuracy": 0.851942241191864, "num_tokens": 579045172.0, "step": 15254 }, { "epoch": 1.9405927998982317, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.779123306274414, "learning_rate": 1e-05, "loss": 0.4715, "mean_token_accuracy": 0.8791925311088562, "num_tokens": 579075001.0, "step": 15255 }, { "epoch": 1.9407200101768223, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.85719633102417, "learning_rate": 1e-05, "loss": 0.4872, "mean_token_accuracy": 0.8688317537307739, "num_tokens": 579116029.0, "step": 15256 }, { "epoch": 1.9408472204554128, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 7.022892475128174, "learning_rate": 1e-05, "loss": 0.5177, "mean_token_accuracy": 0.8673690557479858, "num_tokens": 579157741.0, "step": 15257 }, { "epoch": 1.9409744307340033, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.7616682052612305, "learning_rate": 1e-05, "loss": 0.4866, "mean_token_accuracy": 0.8699057102203369, "num_tokens": 579196033.0, "step": 15258 }, { "epoch": 1.9411016410125939, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 0.00011539459228515625, "grad_norm": 7.314162731170654, "learning_rate": 1e-05, "loss": 0.4869, "mean_token_accuracy": 0.8798303008079529, "num_tokens": 579235060.0, "step": 15259 }, { "epoch": 1.9412288512911844, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 5.920564651489258, "learning_rate": 1e-05, "loss": 0.5072, "mean_token_accuracy": 0.8711773157119751, "num_tokens": 579272184.0, "step": 15260 }, { "epoch": 1.941356061569775, "ewc_loss": 0.1455078125, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 0.00012111663818359375, "grad_norm": 7.465335845947266, "learning_rate": 1e-05, "loss": 0.5262, "mean_token_accuracy": 0.8710930347442627, "num_tokens": 579309461.0, "step": 15261 }, { "epoch": 1.9414832718483654, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.225028038024902, "learning_rate": 1e-05, "loss": 0.5499, "mean_token_accuracy": 0.8521596193313599, "num_tokens": 579351871.0, "step": 15262 }, { "epoch": 1.941610482126956, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 0.00011587142944335938, "grad_norm": 7.317094802856445, "learning_rate": 1e-05, "loss": 0.5724, "mean_token_accuracy": 0.8547850847244263, "num_tokens": 579390123.0, "step": 15263 }, { "epoch": 1.9417376924055465, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.414079189300537, "learning_rate": 1e-05, "loss": 0.5415, "mean_token_accuracy": 0.8596673011779785, "num_tokens": 579426015.0, "step": 15264 }, { "epoch": 1.941864902684137, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.999924659729004, "learning_rate": 1e-05, "loss": 0.5449, "mean_token_accuracy": 0.8565483093261719, "num_tokens": 579465853.0, "step": 15265 }, { "epoch": 1.9419921129627273, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.421384811401367, "learning_rate": 1e-05, "loss": 0.5233, "mean_token_accuracy": 0.8649345636367798, "num_tokens": 579499740.0, "step": 15266 }, { "epoch": 1.9421193232413179, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.682283401489258, "learning_rate": 1e-05, "loss": 0.4599, "mean_token_accuracy": 0.8850505948066711, "num_tokens": 579538908.0, "step": 15267 }, { "epoch": 1.9422465335199084, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.339439868927002, "learning_rate": 1e-05, "loss": 0.4868, "mean_token_accuracy": 0.871129035949707, "num_tokens": 579576026.0, "step": 15268 }, { "epoch": 1.942373743798499, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.483165740966797, "learning_rate": 1e-05, "loss": 0.4841, "mean_token_accuracy": 0.8738663792610168, "num_tokens": 579618980.0, "step": 15269 }, { "epoch": 1.9425009540770894, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.231245517730713, "learning_rate": 1e-05, "loss": 0.5407, "mean_token_accuracy": 0.8557978272438049, "num_tokens": 579659205.0, "step": 15270 }, { "epoch": 1.94262816435568, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.408730983734131, "learning_rate": 1e-05, "loss": 0.4532, "mean_token_accuracy": 0.8809831142425537, "num_tokens": 579695598.0, "step": 15271 }, { "epoch": 1.9427553746342705, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 6.092316150665283, "learning_rate": 1e-05, "loss": 0.4845, "mean_token_accuracy": 0.8713605999946594, "num_tokens": 579730871.0, "step": 15272 }, { "epoch": 1.9428825849128608, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.436849117279053, "learning_rate": 1e-05, "loss": 0.5113, "mean_token_accuracy": 0.8658526539802551, "num_tokens": 579772879.0, "step": 15273 }, { "epoch": 1.9430097951914513, "ewc_loss": 0.109375, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.940151214599609, "learning_rate": 1e-05, "loss": 0.4942, "mean_token_accuracy": 0.8653965592384338, "num_tokens": 579812854.0, "step": 15274 }, { "epoch": 1.9431370054700419, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.5808186531066895, "learning_rate": 1e-05, "loss": 0.5228, "mean_token_accuracy": 0.8623573780059814, "num_tokens": 579857324.0, "step": 15275 }, { "epoch": 1.9432642157486324, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.7726240158081055, "learning_rate": 1e-05, "loss": 0.4845, "mean_token_accuracy": 0.8697883486747742, "num_tokens": 579897103.0, "step": 15276 }, { "epoch": 1.943391426027223, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.859308242797852, "learning_rate": 1e-05, "loss": 0.5222, "mean_token_accuracy": 0.8625394105911255, "num_tokens": 579937382.0, "step": 15277 }, { "epoch": 1.9435186363058135, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.661189079284668, "learning_rate": 1e-05, "loss": 0.4758, "mean_token_accuracy": 0.8760973215103149, "num_tokens": 579978591.0, "step": 15278 }, { "epoch": 1.943645846584404, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 0.00011444091796875, "grad_norm": 7.2326555252075195, "learning_rate": 1e-05, "loss": 0.5117, "mean_token_accuracy": 0.8752636909484863, "num_tokens": 580014354.0, "step": 15279 }, { "epoch": 1.9437730568629945, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.754376411437988, "learning_rate": 1e-05, "loss": 0.4665, "mean_token_accuracy": 0.8764359354972839, "num_tokens": 580055575.0, "step": 15280 }, { "epoch": 1.943900267141585, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 0.0001220703125, "grad_norm": 7.549648284912109, "learning_rate": 1e-05, "loss": 0.5511, "mean_token_accuracy": 0.8629087209701538, "num_tokens": 580105007.0, "step": 15281 }, { "epoch": 1.9440274774201756, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.1487956047058105, "learning_rate": 1e-05, "loss": 0.5322, "mean_token_accuracy": 0.861718475818634, "num_tokens": 580150820.0, "step": 15282 }, { "epoch": 1.944154687698766, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 0.00011968612670898438, "grad_norm": 7.461763858795166, "learning_rate": 1e-05, "loss": 0.4932, "mean_token_accuracy": 0.8796721696853638, "num_tokens": 580188303.0, "step": 15283 }, { "epoch": 1.9442818979773566, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.512507438659668, "learning_rate": 1e-05, "loss": 0.5134, "mean_token_accuracy": 0.8682624101638794, "num_tokens": 580221824.0, "step": 15284 }, { "epoch": 1.9444091082559471, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.994629383087158, "learning_rate": 1e-05, "loss": 0.5107, "mean_token_accuracy": 0.8671356439590454, "num_tokens": 580262615.0, "step": 15285 }, { "epoch": 1.9445363185345377, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.601889610290527, "learning_rate": 1e-05, "loss": 0.4871, "mean_token_accuracy": 0.8747580051422119, "num_tokens": 580293535.0, "step": 15286 }, { "epoch": 1.9446635288131282, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.4722490310668945, "learning_rate": 1e-05, "loss": 0.5119, "mean_token_accuracy": 0.8687096834182739, "num_tokens": 580328645.0, "step": 15287 }, { "epoch": 1.9447907390917187, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.650976181030273, "learning_rate": 1e-05, "loss": 0.4988, "mean_token_accuracy": 0.8690372705459595, "num_tokens": 580365157.0, "step": 15288 }, { "epoch": 1.9449179493703093, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 6.038695812225342, "learning_rate": 1e-05, "loss": 0.4576, "mean_token_accuracy": 0.8774305582046509, "num_tokens": 580398195.0, "step": 15289 }, { "epoch": 1.9450451596488998, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.814275741577148, "learning_rate": 1e-05, "loss": 0.5213, "mean_token_accuracy": 0.8663411140441895, "num_tokens": 580440238.0, "step": 15290 }, { "epoch": 1.94517236992749, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.800746917724609, "learning_rate": 1e-05, "loss": 0.5078, "mean_token_accuracy": 0.8628149032592773, "num_tokens": 580477997.0, "step": 15291 }, { "epoch": 1.9452995802060806, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 0.00010967254638671875, "grad_norm": 7.0857439041137695, "learning_rate": 1e-05, "loss": 0.5082, "mean_token_accuracy": 0.8713040351867676, "num_tokens": 580516335.0, "step": 15292 }, { "epoch": 1.9454267904846712, "ewc_loss": 0.109375, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.774778366088867, "learning_rate": 1e-05, "loss": 0.4573, "mean_token_accuracy": 0.879498302936554, "num_tokens": 580546492.0, "step": 15293 }, { "epoch": 1.9455540007632617, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 0.00011777877807617188, "grad_norm": 7.345443248748779, "learning_rate": 1e-05, "loss": 0.4808, "mean_token_accuracy": 0.882056713104248, "num_tokens": 580583804.0, "step": 15294 }, { "epoch": 1.9456812110418522, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 5.979188919067383, "learning_rate": 1e-05, "loss": 0.5313, "mean_token_accuracy": 0.8587883114814758, "num_tokens": 580618111.0, "step": 15295 }, { "epoch": 1.9458084213204427, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 0.00011920928955078125, "grad_norm": 7.408923149108887, "learning_rate": 1e-05, "loss": 0.5337, "mean_token_accuracy": 0.867458701133728, "num_tokens": 580657058.0, "step": 15296 }, { "epoch": 1.945935631599033, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.255779266357422, "learning_rate": 1e-05, "loss": 0.5286, "mean_token_accuracy": 0.8603381514549255, "num_tokens": 580701628.0, "step": 15297 }, { "epoch": 1.9460628418776236, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 0.00011396408081054688, "grad_norm": 7.229090690612793, "learning_rate": 1e-05, "loss": 0.5021, "mean_token_accuracy": 0.8734601140022278, "num_tokens": 580738766.0, "step": 15298 }, { "epoch": 1.946190052156214, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.379806995391846, "learning_rate": 1e-05, "loss": 0.5033, "mean_token_accuracy": 0.8712379932403564, "num_tokens": 580780110.0, "step": 15299 }, { "epoch": 1.9463172624348046, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.940888404846191, "learning_rate": 1e-05, "loss": 0.4862, "mean_token_accuracy": 0.877630352973938, "num_tokens": 580825208.0, "step": 15300 }, { "epoch": 1.9464444727133952, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.3607354164123535, "learning_rate": 1e-05, "loss": 0.5623, "mean_token_accuracy": 0.8511718511581421, "num_tokens": 580870339.0, "step": 15301 }, { "epoch": 1.9465716829919857, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.678527355194092, "learning_rate": 1e-05, "loss": 0.4698, "mean_token_accuracy": 0.8796854019165039, "num_tokens": 580899509.0, "step": 15302 }, { "epoch": 1.9466988932705762, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.284492015838623, "learning_rate": 1e-05, "loss": 0.499, "mean_token_accuracy": 0.8721328377723694, "num_tokens": 580941345.0, "step": 15303 }, { "epoch": 1.9468261035491667, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.518449783325195, "learning_rate": 1e-05, "loss": 0.4822, "mean_token_accuracy": 0.8759567737579346, "num_tokens": 580976819.0, "step": 15304 }, { "epoch": 1.9469533138277573, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.164371013641357, "learning_rate": 1e-05, "loss": 0.4763, "mean_token_accuracy": 0.8753809928894043, "num_tokens": 581011971.0, "step": 15305 }, { "epoch": 1.9470805241063478, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.488419532775879, "learning_rate": 1e-05, "loss": 0.5487, "mean_token_accuracy": 0.8539161682128906, "num_tokens": 581045871.0, "step": 15306 }, { "epoch": 1.9472077343849383, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 6.036355018615723, "learning_rate": 1e-05, "loss": 0.4763, "mean_token_accuracy": 0.8787959218025208, "num_tokens": 581077986.0, "step": 15307 }, { "epoch": 1.9473349446635289, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.534865379333496, "learning_rate": 1e-05, "loss": 0.4955, "mean_token_accuracy": 0.8718776702880859, "num_tokens": 581112988.0, "step": 15308 }, { "epoch": 1.9474621549421194, "ewc_loss": 0.109375, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.885396480560303, "learning_rate": 1e-05, "loss": 0.5184, "mean_token_accuracy": 0.863084077835083, "num_tokens": 581155314.0, "step": 15309 }, { "epoch": 1.94758936522071, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.698570251464844, "learning_rate": 1e-05, "loss": 0.4827, "mean_token_accuracy": 0.8758215308189392, "num_tokens": 581203622.0, "step": 15310 }, { "epoch": 1.9477165754993004, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.769553184509277, "learning_rate": 1e-05, "loss": 0.4768, "mean_token_accuracy": 0.8735126256942749, "num_tokens": 581234306.0, "step": 15311 }, { "epoch": 1.947843785777891, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.961780548095703, "learning_rate": 1e-05, "loss": 0.5658, "mean_token_accuracy": 0.8520115613937378, "num_tokens": 581275884.0, "step": 15312 }, { "epoch": 1.9479709960564815, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.739725589752197, "learning_rate": 1e-05, "loss": 0.4486, "mean_token_accuracy": 0.8846375942230225, "num_tokens": 581312674.0, "step": 15313 }, { "epoch": 1.948098206335072, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 0.00011587142944335938, "grad_norm": 7.248292922973633, "learning_rate": 1e-05, "loss": 0.5341, "mean_token_accuracy": 0.8617899417877197, "num_tokens": 581352400.0, "step": 15314 }, { "epoch": 1.9482254166136623, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 5.915254592895508, "learning_rate": 1e-05, "loss": 0.5137, "mean_token_accuracy": 0.8644629716873169, "num_tokens": 581390136.0, "step": 15315 }, { "epoch": 1.9483526268922529, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 0.00011730194091796875, "grad_norm": 7.355703353881836, "learning_rate": 1e-05, "loss": 0.482, "mean_token_accuracy": 0.8814250826835632, "num_tokens": 581424864.0, "step": 15316 }, { "epoch": 1.9484798371708434, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.23197603225708, "learning_rate": 1e-05, "loss": 0.5633, "mean_token_accuracy": 0.8539345860481262, "num_tokens": 581467791.0, "step": 15317 }, { "epoch": 1.948607047449434, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 7.100914478302002, "learning_rate": 1e-05, "loss": 0.5151, "mean_token_accuracy": 0.8695928454399109, "num_tokens": 581505121.0, "step": 15318 }, { "epoch": 1.9487342577280244, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.424071788787842, "learning_rate": 1e-05, "loss": 0.488, "mean_token_accuracy": 0.8740532398223877, "num_tokens": 581543306.0, "step": 15319 }, { "epoch": 1.948861468006615, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.644678115844727, "learning_rate": 1e-05, "loss": 0.5012, "mean_token_accuracy": 0.8662214279174805, "num_tokens": 581573476.0, "step": 15320 }, { "epoch": 1.9489886782852053, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.533975601196289, "learning_rate": 1e-05, "loss": 0.4718, "mean_token_accuracy": 0.8819345235824585, "num_tokens": 581606693.0, "step": 15321 }, { "epoch": 1.9491158885637958, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.144289493560791, "learning_rate": 1e-05, "loss": 0.4901, "mean_token_accuracy": 0.8694985508918762, "num_tokens": 581637826.0, "step": 15322 }, { "epoch": 1.9492430988423863, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.7136759757995605, "learning_rate": 1e-05, "loss": 0.461, "mean_token_accuracy": 0.8823569416999817, "num_tokens": 581670590.0, "step": 15323 }, { "epoch": 1.9493703091209769, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.772014617919922, "learning_rate": 1e-05, "loss": 0.4924, "mean_token_accuracy": 0.8702112436294556, "num_tokens": 581704091.0, "step": 15324 }, { "epoch": 1.9494975193995674, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 7.044976234436035, "learning_rate": 1e-05, "loss": 0.4896, "mean_token_accuracy": 0.8771365880966187, "num_tokens": 581747362.0, "step": 15325 }, { "epoch": 1.949624729678158, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.634137153625488, "learning_rate": 1e-05, "loss": 0.4819, "mean_token_accuracy": 0.8726465106010437, "num_tokens": 581781105.0, "step": 15326 }, { "epoch": 1.9497519399567484, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 0.00012063980102539062, "grad_norm": 7.494790554046631, "learning_rate": 1e-05, "loss": 0.5115, "mean_token_accuracy": 0.8748681545257568, "num_tokens": 581824800.0, "step": 15327 }, { "epoch": 1.949879150235339, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 5.8902997970581055, "learning_rate": 1e-05, "loss": 0.5294, "mean_token_accuracy": 0.8586183786392212, "num_tokens": 581865984.0, "step": 15328 }, { "epoch": 1.9500063605139295, "ewc_loss": 0.1513671875, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 0.00012683868408203125, "grad_norm": 7.70469331741333, "learning_rate": 1e-05, "loss": 0.5316, "mean_token_accuracy": 0.870673656463623, "num_tokens": 581903098.0, "step": 15329 }, { "epoch": 1.95013357079252, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.384397983551025, "learning_rate": 1e-05, "loss": 0.4627, "mean_token_accuracy": 0.883294939994812, "num_tokens": 581940704.0, "step": 15330 }, { "epoch": 1.9502607810711106, "ewc_loss": 0.1455078125, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 0.00012159347534179688, "grad_norm": 7.454527378082275, "learning_rate": 1e-05, "loss": 0.562, "mean_token_accuracy": 0.8588102459907532, "num_tokens": 581983598.0, "step": 15331 }, { "epoch": 1.950387991349701, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 6.593008041381836, "learning_rate": 1e-05, "loss": 0.4979, "mean_token_accuracy": 0.8728625774383545, "num_tokens": 582020073.0, "step": 15332 }, { "epoch": 1.9505152016282916, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 0.00010967254638671875, "grad_norm": 7.051314830780029, "learning_rate": 1e-05, "loss": 0.4658, "mean_token_accuracy": 0.8843241333961487, "num_tokens": 582056597.0, "step": 15333 }, { "epoch": 1.9506424119068821, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.529603481292725, "learning_rate": 1e-05, "loss": 0.4654, "mean_token_accuracy": 0.8844969272613525, "num_tokens": 582090164.0, "step": 15334 }, { "epoch": 1.9507696221854727, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.713571071624756, "learning_rate": 1e-05, "loss": 0.4844, "mean_token_accuracy": 0.8781867027282715, "num_tokens": 582130553.0, "step": 15335 }, { "epoch": 1.9508968324640632, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.3966827392578125, "learning_rate": 1e-05, "loss": 0.4854, "mean_token_accuracy": 0.8766087293624878, "num_tokens": 582161175.0, "step": 15336 }, { "epoch": 1.9510240427426537, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.527604579925537, "learning_rate": 1e-05, "loss": 0.4906, "mean_token_accuracy": 0.871698260307312, "num_tokens": 582194424.0, "step": 15337 }, { "epoch": 1.9511512530212443, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.234803676605225, "learning_rate": 1e-05, "loss": 0.5101, "mean_token_accuracy": 0.8662895560264587, "num_tokens": 582231446.0, "step": 15338 }, { "epoch": 1.9512784632998348, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.462137222290039, "learning_rate": 1e-05, "loss": 0.5101, "mean_token_accuracy": 0.8671201467514038, "num_tokens": 582270486.0, "step": 15339 }, { "epoch": 1.951405673578425, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 6.070143699645996, "learning_rate": 1e-05, "loss": 0.4867, "mean_token_accuracy": 0.8728482723236084, "num_tokens": 582303583.0, "step": 15340 }, { "epoch": 1.9515328838570156, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.519040107727051, "learning_rate": 1e-05, "loss": 0.5068, "mean_token_accuracy": 0.869204044342041, "num_tokens": 582344982.0, "step": 15341 }, { "epoch": 1.9516600941356061, "ewc_loss": 0.10888671875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.896320819854736, "learning_rate": 1e-05, "loss": 0.4648, "mean_token_accuracy": 0.8785128593444824, "num_tokens": 582383373.0, "step": 15342 }, { "epoch": 1.9517873044141967, "ewc_loss": 0.125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.692350387573242, "learning_rate": 1e-05, "loss": 0.4948, "mean_token_accuracy": 0.8701189756393433, "num_tokens": 582426671.0, "step": 15343 }, { "epoch": 1.9519145146927872, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.752978324890137, "learning_rate": 1e-05, "loss": 0.51, "mean_token_accuracy": 0.8630276918411255, "num_tokens": 582464801.0, "step": 15344 }, { "epoch": 1.9520417249713777, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.99688720703125, "learning_rate": 1e-05, "loss": 0.5114, "mean_token_accuracy": 0.8681306838989258, "num_tokens": 582507627.0, "step": 15345 }, { "epoch": 1.952168935249968, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.697486400604248, "learning_rate": 1e-05, "loss": 0.4542, "mean_token_accuracy": 0.8809468746185303, "num_tokens": 582544072.0, "step": 15346 }, { "epoch": 1.9522961455285586, "ewc_loss": 0.140625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.0001163482666015625, "grad_norm": 7.359989166259766, "learning_rate": 1e-05, "loss": 0.5244, "mean_token_accuracy": 0.8662683367729187, "num_tokens": 582587928.0, "step": 15347 }, { "epoch": 1.952423355807149, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 5.903575420379639, "learning_rate": 1e-05, "loss": 0.4934, "mean_token_accuracy": 0.8698435425758362, "num_tokens": 582622182.0, "step": 15348 }, { "epoch": 1.9525505660857396, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 0.0001220703125, "grad_norm": 7.54486608505249, "learning_rate": 1e-05, "loss": 0.471, "mean_token_accuracy": 0.8888813257217407, "num_tokens": 582654890.0, "step": 15349 }, { "epoch": 1.9526777763643302, "ewc_loss": 0.125, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.315916538238525, "learning_rate": 1e-05, "loss": 0.4955, "mean_token_accuracy": 0.8742642998695374, "num_tokens": 582690874.0, "step": 15350 }, { "epoch": 1.9528049866429207, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 0.00011491775512695312, "grad_norm": 7.256706237792969, "learning_rate": 1e-05, "loss": 0.4894, "mean_token_accuracy": 0.8743563890457153, "num_tokens": 582726543.0, "step": 15351 }, { "epoch": 1.9529321969215112, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.54891300201416, "learning_rate": 1e-05, "loss": 0.5427, "mean_token_accuracy": 0.8601647615432739, "num_tokens": 582764552.0, "step": 15352 }, { "epoch": 1.9530594072001017, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.743683338165283, "learning_rate": 1e-05, "loss": 0.5327, "mean_token_accuracy": 0.8578861951828003, "num_tokens": 582804613.0, "step": 15353 }, { "epoch": 1.9531866174786923, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.599827289581299, "learning_rate": 1e-05, "loss": 0.491, "mean_token_accuracy": 0.8748584985733032, "num_tokens": 582841510.0, "step": 15354 }, { "epoch": 1.9533138277572828, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.2438249588012695, "learning_rate": 1e-05, "loss": 0.5768, "mean_token_accuracy": 0.8409504890441895, "num_tokens": 582879371.0, "step": 15355 }, { "epoch": 1.9534410380358733, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.713101863861084, "learning_rate": 1e-05, "loss": 0.5322, "mean_token_accuracy": 0.8626265525817871, "num_tokens": 582915354.0, "step": 15356 }, { "epoch": 1.9535682483144639, "ewc_loss": 0.10791015625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.87166166305542, "learning_rate": 1e-05, "loss": 0.4562, "mean_token_accuracy": 0.8776608109474182, "num_tokens": 582949897.0, "step": 15357 }, { "epoch": 1.9536954585930544, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.968934059143066, "learning_rate": 1e-05, "loss": 0.5525, "mean_token_accuracy": 0.8573394417762756, "num_tokens": 582984383.0, "step": 15358 }, { "epoch": 1.953822668871645, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.715339183807373, "learning_rate": 1e-05, "loss": 0.5116, "mean_token_accuracy": 0.864497184753418, "num_tokens": 583023146.0, "step": 15359 }, { "epoch": 1.9539498791502354, "ewc_loss": 0.140625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00011587142944335938, "grad_norm": 7.308605670928955, "learning_rate": 1e-05, "loss": 0.4815, "mean_token_accuracy": 0.8821649551391602, "num_tokens": 583061485.0, "step": 15360 }, { "epoch": 1.954077089428826, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 5.841857433319092, "learning_rate": 1e-05, "loss": 0.5008, "mean_token_accuracy": 0.8661574721336365, "num_tokens": 583097278.0, "step": 15361 }, { "epoch": 1.9542042997074165, "ewc_loss": 0.1474609375, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.52246618270874, "learning_rate": 1e-05, "loss": 0.5116, "mean_token_accuracy": 0.8756201267242432, "num_tokens": 583140458.0, "step": 15362 }, { "epoch": 1.954331509986007, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.184370040893555, "learning_rate": 1e-05, "loss": 0.4737, "mean_token_accuracy": 0.877220869064331, "num_tokens": 583184798.0, "step": 15363 }, { "epoch": 1.9544587202645973, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.0001201629638671875, "grad_norm": 7.407928466796875, "learning_rate": 1e-05, "loss": 0.4938, "mean_token_accuracy": 0.8786641955375671, "num_tokens": 583220517.0, "step": 15364 }, { "epoch": 1.9545859305431879, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.447192668914795, "learning_rate": 1e-05, "loss": 0.5438, "mean_token_accuracy": 0.8591697812080383, "num_tokens": 583256615.0, "step": 15365 }, { "epoch": 1.9547131408217784, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 7.111566066741943, "learning_rate": 1e-05, "loss": 0.5144, "mean_token_accuracy": 0.8653696775436401, "num_tokens": 583294533.0, "step": 15366 }, { "epoch": 1.954840351100369, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.466987133026123, "learning_rate": 1e-05, "loss": 0.4673, "mean_token_accuracy": 0.8803802132606506, "num_tokens": 583333610.0, "step": 15367 }, { "epoch": 1.9549675613789594, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.766416072845459, "learning_rate": 1e-05, "loss": 0.488, "mean_token_accuracy": 0.8789225816726685, "num_tokens": 583366424.0, "step": 15368 }, { "epoch": 1.95509477165755, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.372261047363281, "learning_rate": 1e-05, "loss": 0.5102, "mean_token_accuracy": 0.8699414730072021, "num_tokens": 583403939.0, "step": 15369 }, { "epoch": 1.9552219819361403, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.543224334716797, "learning_rate": 1e-05, "loss": 0.4928, "mean_token_accuracy": 0.8688259124755859, "num_tokens": 583441053.0, "step": 15370 }, { "epoch": 1.9553491922147308, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.249112606048584, "learning_rate": 1e-05, "loss": 0.4182, "mean_token_accuracy": 0.8918001651763916, "num_tokens": 583471168.0, "step": 15371 }, { "epoch": 1.9554764024933213, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.463046073913574, "learning_rate": 1e-05, "loss": 0.5045, "mean_token_accuracy": 0.8653042316436768, "num_tokens": 583510604.0, "step": 15372 }, { "epoch": 1.9556036127719119, "ewc_loss": 0.11474609375, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.099770545959473, "learning_rate": 1e-05, "loss": 0.4862, "mean_token_accuracy": 0.8722177147865295, "num_tokens": 583546186.0, "step": 15373 }, { "epoch": 1.9557308230505024, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.498312473297119, "learning_rate": 1e-05, "loss": 0.5358, "mean_token_accuracy": 0.8556999564170837, "num_tokens": 583581301.0, "step": 15374 }, { "epoch": 1.955858033329093, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.946748733520508, "learning_rate": 1e-05, "loss": 0.5514, "mean_token_accuracy": 0.8535556793212891, "num_tokens": 583624005.0, "step": 15375 }, { "epoch": 1.9559852436076834, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.622494697570801, "learning_rate": 1e-05, "loss": 0.4935, "mean_token_accuracy": 0.8699409365653992, "num_tokens": 583662738.0, "step": 15376 }, { "epoch": 1.956112453886274, "ewc_loss": 0.10791015625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.805432319641113, "learning_rate": 1e-05, "loss": 0.4606, "mean_token_accuracy": 0.8794769644737244, "num_tokens": 583700742.0, "step": 15377 }, { "epoch": 1.9562396641648645, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.862060070037842, "learning_rate": 1e-05, "loss": 0.4814, "mean_token_accuracy": 0.8760759830474854, "num_tokens": 583740529.0, "step": 15378 }, { "epoch": 1.956366874443455, "ewc_loss": 0.10791015625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.7104878425598145, "learning_rate": 1e-05, "loss": 0.4431, "mean_token_accuracy": 0.8830453753471375, "num_tokens": 583776866.0, "step": 15379 }, { "epoch": 1.9564940847220456, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00011444091796875, "grad_norm": 7.200779914855957, "learning_rate": 1e-05, "loss": 0.4987, "mean_token_accuracy": 0.8755704164505005, "num_tokens": 583815561.0, "step": 15380 }, { "epoch": 1.956621295000636, "ewc_loss": 0.11083984375, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.813620567321777, "learning_rate": 1e-05, "loss": 0.4556, "mean_token_accuracy": 0.8768329620361328, "num_tokens": 583846701.0, "step": 15381 }, { "epoch": 1.9567485052792266, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00011968612670898438, "grad_norm": 7.426146030426025, "learning_rate": 1e-05, "loss": 0.4842, "mean_token_accuracy": 0.8836368322372437, "num_tokens": 583885561.0, "step": 15382 }, { "epoch": 1.9568757155578171, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.134298801422119, "learning_rate": 1e-05, "loss": 0.5147, "mean_token_accuracy": 0.863767147064209, "num_tokens": 583928662.0, "step": 15383 }, { "epoch": 1.9570029258364077, "ewc_loss": 0.140625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00011587142944335938, "grad_norm": 7.305147171020508, "learning_rate": 1e-05, "loss": 0.4957, "mean_token_accuracy": 0.8759794235229492, "num_tokens": 583969097.0, "step": 15384 }, { "epoch": 1.9571301361149982, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.429076671600342, "learning_rate": 1e-05, "loss": 0.5135, "mean_token_accuracy": 0.8663873672485352, "num_tokens": 584008547.0, "step": 15385 }, { "epoch": 1.9572573463935887, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.890978813171387, "learning_rate": 1e-05, "loss": 0.5815, "mean_token_accuracy": 0.8522853255271912, "num_tokens": 584046982.0, "step": 15386 }, { "epoch": 1.9573845566721793, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.529178142547607, "learning_rate": 1e-05, "loss": 0.4661, "mean_token_accuracy": 0.8833231925964355, "num_tokens": 584081011.0, "step": 15387 }, { "epoch": 1.9575117669507698, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.37308406829834, "learning_rate": 1e-05, "loss": 0.5333, "mean_token_accuracy": 0.8619064092636108, "num_tokens": 584115550.0, "step": 15388 }, { "epoch": 1.95763897722936, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.646581172943115, "learning_rate": 1e-05, "loss": 0.4842, "mean_token_accuracy": 0.8784521818161011, "num_tokens": 584153733.0, "step": 15389 }, { "epoch": 1.9577661875079506, "ewc_loss": 0.109375, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.938056945800781, "learning_rate": 1e-05, "loss": 0.5026, "mean_token_accuracy": 0.865605354309082, "num_tokens": 584182097.0, "step": 15390 }, { "epoch": 1.9578933977865411, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.8784260749816895, "learning_rate": 1e-05, "loss": 0.4399, "mean_token_accuracy": 0.8909106254577637, "num_tokens": 584215089.0, "step": 15391 }, { "epoch": 1.9580206080651317, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.6924238204956055, "learning_rate": 1e-05, "loss": 0.4343, "mean_token_accuracy": 0.8882816433906555, "num_tokens": 584244635.0, "step": 15392 }, { "epoch": 1.9581478183437222, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00011396408081054688, "grad_norm": 7.239943981170654, "learning_rate": 1e-05, "loss": 0.4889, "mean_token_accuracy": 0.876728892326355, "num_tokens": 584280267.0, "step": 15393 }, { "epoch": 1.9582750286223127, "ewc_loss": 0.10986328125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.745778560638428, "learning_rate": 1e-05, "loss": 0.4354, "mean_token_accuracy": 0.8852238655090332, "num_tokens": 584311998.0, "step": 15394 }, { "epoch": 1.958402238900903, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.0001220703125, "grad_norm": 7.57053804397583, "learning_rate": 1e-05, "loss": 0.5334, "mean_token_accuracy": 0.8701615333557129, "num_tokens": 584342975.0, "step": 15395 }, { "epoch": 1.9585294491794936, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.122432708740234, "learning_rate": 1e-05, "loss": 0.494, "mean_token_accuracy": 0.869828999042511, "num_tokens": 584379087.0, "step": 15396 }, { "epoch": 1.958656659458084, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.0001220703125, "grad_norm": 7.5230607986450195, "learning_rate": 1e-05, "loss": 0.545, "mean_token_accuracy": 0.8626044988632202, "num_tokens": 584420126.0, "step": 15397 }, { "epoch": 1.9587838697366746, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.459002494812012, "learning_rate": 1e-05, "loss": 0.499, "mean_token_accuracy": 0.8720969557762146, "num_tokens": 584457742.0, "step": 15398 }, { "epoch": 1.9589110800152651, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00011396408081054688, "grad_norm": 7.189312934875488, "learning_rate": 1e-05, "loss": 0.4902, "mean_token_accuracy": 0.8756698966026306, "num_tokens": 584497439.0, "step": 15399 }, { "epoch": 1.9590382902938557, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.519103050231934, "learning_rate": 1e-05, "loss": 0.4691, "mean_token_accuracy": 0.8812933564186096, "num_tokens": 584532922.0, "step": 15400 }, { "epoch": 1.9591655005724462, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.811757564544678, "learning_rate": 1e-05, "loss": 0.4768, "mean_token_accuracy": 0.8819134831428528, "num_tokens": 584574834.0, "step": 15401 }, { "epoch": 1.9592927108510367, "ewc_loss": 0.125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.444743633270264, "learning_rate": 1e-05, "loss": 0.5047, "mean_token_accuracy": 0.8709261417388916, "num_tokens": 584615494.0, "step": 15402 }, { "epoch": 1.9594199211296273, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.551356792449951, "learning_rate": 1e-05, "loss": 0.5131, "mean_token_accuracy": 0.8613947629928589, "num_tokens": 584651494.0, "step": 15403 }, { "epoch": 1.9595471314082178, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.317724704742432, "learning_rate": 1e-05, "loss": 0.5018, "mean_token_accuracy": 0.8616360425949097, "num_tokens": 584691556.0, "step": 15404 }, { "epoch": 1.9596743416868083, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.4253926277160645, "learning_rate": 1e-05, "loss": 0.5569, "mean_token_accuracy": 0.8504559993743896, "num_tokens": 584735960.0, "step": 15405 }, { "epoch": 1.9598015519653988, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.171845436096191, "learning_rate": 1e-05, "loss": 0.5078, "mean_token_accuracy": 0.863810122013092, "num_tokens": 584770808.0, "step": 15406 }, { "epoch": 1.9599287622439894, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.4138689041137695, "learning_rate": 1e-05, "loss": 0.4824, "mean_token_accuracy": 0.8742597699165344, "num_tokens": 584811451.0, "step": 15407 }, { "epoch": 1.96005597252258, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 6.019031524658203, "learning_rate": 1e-05, "loss": 0.4915, "mean_token_accuracy": 0.8700435161590576, "num_tokens": 584845050.0, "step": 15408 }, { "epoch": 1.9601831828011704, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.497920036315918, "learning_rate": 1e-05, "loss": 0.485, "mean_token_accuracy": 0.8716089725494385, "num_tokens": 584883857.0, "step": 15409 }, { "epoch": 1.960310393079761, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.863908290863037, "learning_rate": 1e-05, "loss": 0.501, "mean_token_accuracy": 0.8655698299407959, "num_tokens": 584918118.0, "step": 15410 }, { "epoch": 1.9604376033583515, "ewc_loss": 0.125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.692103385925293, "learning_rate": 1e-05, "loss": 0.504, "mean_token_accuracy": 0.8687044382095337, "num_tokens": 584960789.0, "step": 15411 }, { "epoch": 1.960564813636942, "ewc_loss": 0.10693359375, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.697521686553955, "learning_rate": 1e-05, "loss": 0.44, "mean_token_accuracy": 0.8851888179779053, "num_tokens": 585000631.0, "step": 15412 }, { "epoch": 1.9606920239155323, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 7.053894519805908, "learning_rate": 1e-05, "loss": 0.5072, "mean_token_accuracy": 0.8721937537193298, "num_tokens": 585036615.0, "step": 15413 }, { "epoch": 1.9608192341941229, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.648702621459961, "learning_rate": 1e-05, "loss": 0.4465, "mean_token_accuracy": 0.8830774426460266, "num_tokens": 585075076.0, "step": 15414 }, { "epoch": 1.9609464444727134, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.0001220703125, "grad_norm": 7.491990089416504, "learning_rate": 1e-05, "loss": 0.4999, "mean_token_accuracy": 0.8768377304077148, "num_tokens": 585112215.0, "step": 15415 }, { "epoch": 1.961073654751304, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 5.928257465362549, "learning_rate": 1e-05, "loss": 0.4779, "mean_token_accuracy": 0.8767675757408142, "num_tokens": 585148920.0, "step": 15416 }, { "epoch": 1.9612008650298944, "ewc_loss": 0.1513671875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00012683868408203125, "grad_norm": 7.672422885894775, "learning_rate": 1e-05, "loss": 0.5527, "mean_token_accuracy": 0.8633915185928345, "num_tokens": 585191128.0, "step": 15417 }, { "epoch": 1.961328075308485, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.422492980957031, "learning_rate": 1e-05, "loss": 0.5135, "mean_token_accuracy": 0.8665294647216797, "num_tokens": 585234130.0, "step": 15418 }, { "epoch": 1.9614552855870753, "ewc_loss": 0.140625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00011682510375976562, "grad_norm": 7.349733829498291, "learning_rate": 1e-05, "loss": 0.5116, "mean_token_accuracy": 0.8675327301025391, "num_tokens": 585267089.0, "step": 15419 }, { "epoch": 1.9615824958656658, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 6.639211654663086, "learning_rate": 1e-05, "loss": 0.4977, "mean_token_accuracy": 0.8745381236076355, "num_tokens": 585310508.0, "step": 15420 }, { "epoch": 1.9617097061442563, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.827558994293213, "learning_rate": 1e-05, "loss": 0.4768, "mean_token_accuracy": 0.8766999244689941, "num_tokens": 585346579.0, "step": 15421 }, { "epoch": 1.9618369164228469, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.63927698135376, "learning_rate": 1e-05, "loss": 0.5547, "mean_token_accuracy": 0.8650273680686951, "num_tokens": 585383914.0, "step": 15422 }, { "epoch": 1.9619641267014374, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.354531288146973, "learning_rate": 1e-05, "loss": 0.5173, "mean_token_accuracy": 0.8626583814620972, "num_tokens": 585419322.0, "step": 15423 }, { "epoch": 1.962091336980028, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.674615383148193, "learning_rate": 1e-05, "loss": 0.4883, "mean_token_accuracy": 0.8760082125663757, "num_tokens": 585459081.0, "step": 15424 }, { "epoch": 1.9622185472586184, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 6.013516426086426, "learning_rate": 1e-05, "loss": 0.5043, "mean_token_accuracy": 0.8641718626022339, "num_tokens": 585497116.0, "step": 15425 }, { "epoch": 1.962345757537209, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.803283214569092, "learning_rate": 1e-05, "loss": 0.5256, "mean_token_accuracy": 0.8618727326393127, "num_tokens": 585534704.0, "step": 15426 }, { "epoch": 1.9624729678157995, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.829682350158691, "learning_rate": 1e-05, "loss": 0.463, "mean_token_accuracy": 0.8780574202537537, "num_tokens": 585575806.0, "step": 15427 }, { "epoch": 1.96260017809439, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 7.018782138824463, "learning_rate": 1e-05, "loss": 0.5002, "mean_token_accuracy": 0.8774980902671814, "num_tokens": 585612585.0, "step": 15428 }, { "epoch": 1.9627273883729806, "ewc_loss": 0.10888671875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.7967448234558105, "learning_rate": 1e-05, "loss": 0.4614, "mean_token_accuracy": 0.8756905198097229, "num_tokens": 585647414.0, "step": 15429 }, { "epoch": 1.962854598651571, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00011396408081054688, "grad_norm": 7.2334794998168945, "learning_rate": 1e-05, "loss": 0.5264, "mean_token_accuracy": 0.8675277829170227, "num_tokens": 585693802.0, "step": 15430 }, { "epoch": 1.9629818089301616, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 5.9196553230285645, "learning_rate": 1e-05, "loss": 0.4774, "mean_token_accuracy": 0.8729099035263062, "num_tokens": 585727955.0, "step": 15431 }, { "epoch": 1.9631090192087521, "ewc_loss": 0.140625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00011682510375976562, "grad_norm": 7.35790491104126, "learning_rate": 1e-05, "loss": 0.5512, "mean_token_accuracy": 0.8594023585319519, "num_tokens": 585769784.0, "step": 15432 }, { "epoch": 1.9632362294873427, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.14077615737915, "learning_rate": 1e-05, "loss": 0.48, "mean_token_accuracy": 0.8762847185134888, "num_tokens": 585809989.0, "step": 15433 }, { "epoch": 1.9633634397659332, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00011491775512695312, "grad_norm": 7.281612396240234, "learning_rate": 1e-05, "loss": 0.5148, "mean_token_accuracy": 0.8724097013473511, "num_tokens": 585851054.0, "step": 15434 }, { "epoch": 1.9634906500445237, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.311294078826904, "learning_rate": 1e-05, "loss": 0.5385, "mean_token_accuracy": 0.8598805069923401, "num_tokens": 585886638.0, "step": 15435 }, { "epoch": 1.9636178603231143, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 7.031698226928711, "learning_rate": 1e-05, "loss": 0.5321, "mean_token_accuracy": 0.8617191314697266, "num_tokens": 585925572.0, "step": 15436 }, { "epoch": 1.9637450706017048, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.352087020874023, "learning_rate": 1e-05, "loss": 0.4856, "mean_token_accuracy": 0.869745135307312, "num_tokens": 585961939.0, "step": 15437 }, { "epoch": 1.963872280880295, "ewc_loss": 0.125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.7406134605407715, "learning_rate": 1e-05, "loss": 0.5132, "mean_token_accuracy": 0.8671549558639526, "num_tokens": 585996804.0, "step": 15438 }, { "epoch": 1.9639994911588856, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.310936450958252, "learning_rate": 1e-05, "loss": 0.4701, "mean_token_accuracy": 0.8800169825553894, "num_tokens": 586040347.0, "step": 15439 }, { "epoch": 1.9641267014374761, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.518197536468506, "learning_rate": 1e-05, "loss": 0.4822, "mean_token_accuracy": 0.8776955604553223, "num_tokens": 586078495.0, "step": 15440 }, { "epoch": 1.9642539117160667, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.247483253479004, "learning_rate": 1e-05, "loss": 0.475, "mean_token_accuracy": 0.8784593343734741, "num_tokens": 586114743.0, "step": 15441 }, { "epoch": 1.9643811219946572, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.390402793884277, "learning_rate": 1e-05, "loss": 0.4763, "mean_token_accuracy": 0.8788222074508667, "num_tokens": 586150462.0, "step": 15442 }, { "epoch": 1.9645083322732477, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.179852485656738, "learning_rate": 1e-05, "loss": 0.4957, "mean_token_accuracy": 0.8695240020751953, "num_tokens": 586182860.0, "step": 15443 }, { "epoch": 1.964635542551838, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.338443279266357, "learning_rate": 1e-05, "loss": 0.4924, "mean_token_accuracy": 0.8672665953636169, "num_tokens": 586226438.0, "step": 15444 }, { "epoch": 1.9647627528304286, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.061197757720947, "learning_rate": 1e-05, "loss": 0.4734, "mean_token_accuracy": 0.8721325397491455, "num_tokens": 586267022.0, "step": 15445 }, { "epoch": 1.964889963109019, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.364949703216553, "learning_rate": 1e-05, "loss": 0.4671, "mean_token_accuracy": 0.8777023553848267, "num_tokens": 586299415.0, "step": 15446 }, { "epoch": 1.9650171733876096, "ewc_loss": 0.11083984375, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.9372758865356445, "learning_rate": 1e-05, "loss": 0.4913, "mean_token_accuracy": 0.872251033782959, "num_tokens": 586337351.0, "step": 15447 }, { "epoch": 1.9651443836662001, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.487560272216797, "learning_rate": 1e-05, "loss": 0.5272, "mean_token_accuracy": 0.8620811700820923, "num_tokens": 586378311.0, "step": 15448 }, { "epoch": 1.9652715939447907, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.806903839111328, "learning_rate": 1e-05, "loss": 0.4866, "mean_token_accuracy": 0.8717201352119446, "num_tokens": 586414137.0, "step": 15449 }, { "epoch": 1.9653988042233812, "ewc_loss": 0.125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.71278715133667, "learning_rate": 1e-05, "loss": 0.5264, "mean_token_accuracy": 0.8657160997390747, "num_tokens": 586451536.0, "step": 15450 }, { "epoch": 1.9655260145019717, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.6735100746154785, "learning_rate": 1e-05, "loss": 0.4873, "mean_token_accuracy": 0.8673993349075317, "num_tokens": 586484643.0, "step": 15451 }, { "epoch": 1.9656532247805623, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00011014938354492188, "grad_norm": 7.069182395935059, "learning_rate": 1e-05, "loss": 0.5376, "mean_token_accuracy": 0.8602142333984375, "num_tokens": 586523926.0, "step": 15452 }, { "epoch": 1.9657804350591528, "ewc_loss": 0.10693359375, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.626528739929199, "learning_rate": 1e-05, "loss": 0.4721, "mean_token_accuracy": 0.8735382556915283, "num_tokens": 586564999.0, "step": 15453 }, { "epoch": 1.9659076453377433, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.0001220703125, "grad_norm": 7.517071723937988, "learning_rate": 1e-05, "loss": 0.5614, "mean_token_accuracy": 0.8630439639091492, "num_tokens": 586603455.0, "step": 15454 }, { "epoch": 1.9660348556163338, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 5.921984672546387, "learning_rate": 1e-05, "loss": 0.5007, "mean_token_accuracy": 0.8674170970916748, "num_tokens": 586646345.0, "step": 15455 }, { "epoch": 1.9661620658949244, "ewc_loss": 0.1513671875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00012683868408203125, "grad_norm": 7.702734470367432, "learning_rate": 1e-05, "loss": 0.5451, "mean_token_accuracy": 0.8671573400497437, "num_tokens": 586688065.0, "step": 15456 }, { "epoch": 1.966289276173515, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.432367324829102, "learning_rate": 1e-05, "loss": 0.4761, "mean_token_accuracy": 0.8759231567382812, "num_tokens": 586722631.0, "step": 15457 }, { "epoch": 1.9664164864521054, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.000118255615234375, "grad_norm": 7.372185230255127, "learning_rate": 1e-05, "loss": 0.464, "mean_token_accuracy": 0.8885775804519653, "num_tokens": 586762630.0, "step": 15458 }, { "epoch": 1.966543696730696, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 6.637518405914307, "learning_rate": 1e-05, "loss": 0.512, "mean_token_accuracy": 0.8705897331237793, "num_tokens": 586799514.0, "step": 15459 }, { "epoch": 1.9666709070092865, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.864849090576172, "learning_rate": 1e-05, "loss": 0.4738, "mean_token_accuracy": 0.8810737133026123, "num_tokens": 586839507.0, "step": 15460 }, { "epoch": 1.966798117287877, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.613813400268555, "learning_rate": 1e-05, "loss": 0.5426, "mean_token_accuracy": 0.8602831363677979, "num_tokens": 586879263.0, "step": 15461 }, { "epoch": 1.9669253275664673, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.414938926696777, "learning_rate": 1e-05, "loss": 0.4845, "mean_token_accuracy": 0.8715490698814392, "num_tokens": 586920297.0, "step": 15462 }, { "epoch": 1.9670525378450578, "ewc_loss": 0.125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.61596155166626, "learning_rate": 1e-05, "loss": 0.5299, "mean_token_accuracy": 0.860003650188446, "num_tokens": 586961171.0, "step": 15463 }, { "epoch": 1.9671797481236484, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 6.094382286071777, "learning_rate": 1e-05, "loss": 0.4795, "mean_token_accuracy": 0.8712490797042847, "num_tokens": 587003808.0, "step": 15464 }, { "epoch": 1.967306958402239, "ewc_loss": 0.125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.68596076965332, "learning_rate": 1e-05, "loss": 0.5345, "mean_token_accuracy": 0.8602014780044556, "num_tokens": 587042056.0, "step": 15465 }, { "epoch": 1.9674341686808294, "ewc_loss": 0.10986328125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.9093146324157715, "learning_rate": 1e-05, "loss": 0.5346, "mean_token_accuracy": 0.8568673133850098, "num_tokens": 587083009.0, "step": 15466 }, { "epoch": 1.96756137895942, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.816233158111572, "learning_rate": 1e-05, "loss": 0.5681, "mean_token_accuracy": 0.8502132892608643, "num_tokens": 587125129.0, "step": 15467 }, { "epoch": 1.9676885892380103, "ewc_loss": 0.109375, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.815005302429199, "learning_rate": 1e-05, "loss": 0.5332, "mean_token_accuracy": 0.8548620939254761, "num_tokens": 587163314.0, "step": 15468 }, { "epoch": 1.9678157995166008, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 7.005875587463379, "learning_rate": 1e-05, "loss": 0.5074, "mean_token_accuracy": 0.8674213886260986, "num_tokens": 587208275.0, "step": 15469 }, { "epoch": 1.9679430097951913, "ewc_loss": 0.10986328125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.7993855476379395, "learning_rate": 1e-05, "loss": 0.4829, "mean_token_accuracy": 0.8683084845542908, "num_tokens": 587242885.0, "step": 15470 }, { "epoch": 1.9680702200737819, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00011396408081054688, "grad_norm": 7.209041118621826, "learning_rate": 1e-05, "loss": 0.537, "mean_token_accuracy": 0.8623554110527039, "num_tokens": 587287787.0, "step": 15471 }, { "epoch": 1.9681974303523724, "ewc_loss": 0.11474609375, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 5.907591819763184, "learning_rate": 1e-05, "loss": 0.4496, "mean_token_accuracy": 0.8837010860443115, "num_tokens": 587325445.0, "step": 15472 }, { "epoch": 1.968324640630963, "ewc_loss": 0.140625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00011682510375976562, "grad_norm": 7.325799465179443, "learning_rate": 1e-05, "loss": 0.5176, "mean_token_accuracy": 0.8716781139373779, "num_tokens": 587366691.0, "step": 15473 }, { "epoch": 1.9684518509095534, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.113734722137451, "learning_rate": 1e-05, "loss": 0.5268, "mean_token_accuracy": 0.8612551689147949, "num_tokens": 587406310.0, "step": 15474 }, { "epoch": 1.968579061188144, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00011396408081054688, "grad_norm": 7.244231224060059, "learning_rate": 1e-05, "loss": 0.5123, "mean_token_accuracy": 0.8682683110237122, "num_tokens": 587446654.0, "step": 15475 }, { "epoch": 1.9687062714667345, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.290099143981934, "learning_rate": 1e-05, "loss": 0.5098, "mean_token_accuracy": 0.8665604591369629, "num_tokens": 587481089.0, "step": 15476 }, { "epoch": 1.968833481745325, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 6.979311943054199, "learning_rate": 1e-05, "loss": 0.494, "mean_token_accuracy": 0.8699756860733032, "num_tokens": 587521293.0, "step": 15477 }, { "epoch": 1.9689606920239155, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.354405403137207, "learning_rate": 1e-05, "loss": 0.5264, "mean_token_accuracy": 0.8628991842269897, "num_tokens": 587564193.0, "step": 15478 }, { "epoch": 1.969087902302506, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.667764663696289, "learning_rate": 1e-05, "loss": 0.5179, "mean_token_accuracy": 0.8625274896621704, "num_tokens": 587603487.0, "step": 15479 }, { "epoch": 1.9692151125810966, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.34440803527832, "learning_rate": 1e-05, "loss": 0.5489, "mean_token_accuracy": 0.8599294424057007, "num_tokens": 587641131.0, "step": 15480 }, { "epoch": 1.9693423228596871, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.3943400382995605, "learning_rate": 1e-05, "loss": 0.5588, "mean_token_accuracy": 0.8515152931213379, "num_tokens": 587680012.0, "step": 15481 }, { "epoch": 1.9694695331382777, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.315123558044434, "learning_rate": 1e-05, "loss": 0.4796, "mean_token_accuracy": 0.8724943399429321, "num_tokens": 587721904.0, "step": 15482 }, { "epoch": 1.9695967434168682, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.218447208404541, "learning_rate": 1e-05, "loss": 0.4769, "mean_token_accuracy": 0.8751212358474731, "num_tokens": 587752080.0, "step": 15483 }, { "epoch": 1.9697239536954587, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.276184558868408, "learning_rate": 1e-05, "loss": 0.4586, "mean_token_accuracy": 0.8809155225753784, "num_tokens": 587787838.0, "step": 15484 }, { "epoch": 1.9698511639740492, "ewc_loss": 0.11474609375, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.116875171661377, "learning_rate": 1e-05, "loss": 0.4907, "mean_token_accuracy": 0.8703557252883911, "num_tokens": 587826370.0, "step": 15485 }, { "epoch": 1.9699783742526398, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.231177806854248, "learning_rate": 1e-05, "loss": 0.4488, "mean_token_accuracy": 0.8860984444618225, "num_tokens": 587866716.0, "step": 15486 }, { "epoch": 1.97010558453123, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.0747599601745605, "learning_rate": 1e-05, "loss": 0.4799, "mean_token_accuracy": 0.8731568455696106, "num_tokens": 587902766.0, "step": 15487 }, { "epoch": 1.9702327948098206, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.204562187194824, "learning_rate": 1e-05, "loss": 0.5222, "mean_token_accuracy": 0.8582268953323364, "num_tokens": 587944584.0, "step": 15488 }, { "epoch": 1.9703600050884111, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 6.0505828857421875, "learning_rate": 1e-05, "loss": 0.5281, "mean_token_accuracy": 0.8585994839668274, "num_tokens": 587983482.0, "step": 15489 }, { "epoch": 1.9704872153670017, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.203347682952881, "learning_rate": 1e-05, "loss": 0.4726, "mean_token_accuracy": 0.8785548806190491, "num_tokens": 588019621.0, "step": 15490 }, { "epoch": 1.9706144256455922, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 5.99806022644043, "learning_rate": 1e-05, "loss": 0.463, "mean_token_accuracy": 0.8760454654693604, "num_tokens": 588056845.0, "step": 15491 }, { "epoch": 1.9707416359241827, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.222153186798096, "learning_rate": 1e-05, "loss": 0.4787, "mean_token_accuracy": 0.8762921094894409, "num_tokens": 588093731.0, "step": 15492 }, { "epoch": 1.970868846202773, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.941590309143066, "learning_rate": 1e-05, "loss": 0.4378, "mean_token_accuracy": 0.8841719031333923, "num_tokens": 588128908.0, "step": 15493 }, { "epoch": 1.9709960564813636, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.298998832702637, "learning_rate": 1e-05, "loss": 0.528, "mean_token_accuracy": 0.856472373008728, "num_tokens": 588167729.0, "step": 15494 }, { "epoch": 1.971123266759954, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.867081165313721, "learning_rate": 1e-05, "loss": 0.5392, "mean_token_accuracy": 0.8538761138916016, "num_tokens": 588200337.0, "step": 15495 }, { "epoch": 1.9712504770385446, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.425312042236328, "learning_rate": 1e-05, "loss": 0.4956, "mean_token_accuracy": 0.8697121143341064, "num_tokens": 588240736.0, "step": 15496 }, { "epoch": 1.9713776873171351, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.702963352203369, "learning_rate": 1e-05, "loss": 0.5036, "mean_token_accuracy": 0.860346257686615, "num_tokens": 588275187.0, "step": 15497 }, { "epoch": 1.9715048975957257, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.716711044311523, "learning_rate": 1e-05, "loss": 0.5218, "mean_token_accuracy": 0.8632816672325134, "num_tokens": 588308800.0, "step": 15498 }, { "epoch": 1.9716321078743162, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.488488674163818, "learning_rate": 1e-05, "loss": 0.4402, "mean_token_accuracy": 0.8841551542282104, "num_tokens": 588347688.0, "step": 15499 }, { "epoch": 1.9717593181529067, "ewc_loss": 0.140625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.0001163482666015625, "grad_norm": 7.2577595710754395, "learning_rate": 1e-05, "loss": 0.5123, "mean_token_accuracy": 0.8713306784629822, "num_tokens": 588380791.0, "step": 15500 }, { "epoch": 1.9718865284314973, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.406620025634766, "learning_rate": 1e-05, "loss": 0.518, "mean_token_accuracy": 0.8576626777648926, "num_tokens": 588413667.0, "step": 15501 }, { "epoch": 1.9720137387100878, "ewc_loss": 0.162109375, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.0001373291015625, "grad_norm": 8.061001777648926, "learning_rate": 1e-05, "loss": 0.5272, "mean_token_accuracy": 0.8714266419410706, "num_tokens": 588453884.0, "step": 15502 }, { "epoch": 1.9721409489886783, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.033944606781006, "learning_rate": 1e-05, "loss": 0.5115, "mean_token_accuracy": 0.8722776770591736, "num_tokens": 588488895.0, "step": 15503 }, { "epoch": 1.9722681592672688, "ewc_loss": 0.1689453125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00014495849609375, "grad_norm": 8.402729034423828, "learning_rate": 1e-05, "loss": 0.5359, "mean_token_accuracy": 0.874832272529602, "num_tokens": 588526807.0, "step": 15504 }, { "epoch": 1.9723953695458594, "ewc_loss": 0.1513671875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00012683868408203125, "grad_norm": 7.058258533477783, "learning_rate": 1e-05, "loss": 0.5186, "mean_token_accuracy": 0.874101996421814, "num_tokens": 588562272.0, "step": 15505 }, { "epoch": 1.97252257982445, "ewc_loss": 0.15234375, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.0001277923583984375, "grad_norm": 7.779861927032471, "learning_rate": 1e-05, "loss": 0.5402, "mean_token_accuracy": 0.8671581745147705, "num_tokens": 588598298.0, "step": 15506 }, { "epoch": 1.9726497901030404, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.0001220703125, "grad_norm": 7.107568264007568, "learning_rate": 1e-05, "loss": 0.5194, "mean_token_accuracy": 0.870924711227417, "num_tokens": 588637833.0, "step": 15507 }, { "epoch": 1.972777000381631, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00011730194091796875, "grad_norm": 7.2938947677612305, "learning_rate": 1e-05, "loss": 0.5126, "mean_token_accuracy": 0.8679459095001221, "num_tokens": 588674933.0, "step": 15508 }, { "epoch": 1.9729042106602215, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00011157989501953125, "grad_norm": 6.776486873626709, "learning_rate": 1e-05, "loss": 0.5225, "mean_token_accuracy": 0.872113049030304, "num_tokens": 588711899.0, "step": 15509 }, { "epoch": 1.973031420938812, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 6.991280555725098, "learning_rate": 1e-05, "loss": 0.5141, "mean_token_accuracy": 0.8690711259841919, "num_tokens": 588749702.0, "step": 15510 }, { "epoch": 1.9731586312174023, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.501709461212158, "learning_rate": 1e-05, "loss": 0.4836, "mean_token_accuracy": 0.8776876926422119, "num_tokens": 588787066.0, "step": 15511 }, { "epoch": 1.9732858414959928, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.848675727844238, "learning_rate": 1e-05, "loss": 0.5036, "mean_token_accuracy": 0.8716045618057251, "num_tokens": 588824062.0, "step": 15512 }, { "epoch": 1.9734130517745834, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.254035949707031, "learning_rate": 1e-05, "loss": 0.5036, "mean_token_accuracy": 0.8689988255500793, "num_tokens": 588864858.0, "step": 15513 }, { "epoch": 1.973540262053174, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.833655834197998, "learning_rate": 1e-05, "loss": 0.515, "mean_token_accuracy": 0.866777241230011, "num_tokens": 588900219.0, "step": 15514 }, { "epoch": 1.9736674723317644, "ewc_loss": 0.11474609375, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.058255672454834, "learning_rate": 1e-05, "loss": 0.5096, "mean_token_accuracy": 0.8655266165733337, "num_tokens": 588937256.0, "step": 15515 }, { "epoch": 1.973794682610355, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.964141845703125, "learning_rate": 1e-05, "loss": 0.48, "mean_token_accuracy": 0.87806236743927, "num_tokens": 588971263.0, "step": 15516 }, { "epoch": 1.9739218928889453, "ewc_loss": 0.10986328125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.904252529144287, "learning_rate": 1e-05, "loss": 0.4958, "mean_token_accuracy": 0.8649072647094727, "num_tokens": 589010710.0, "step": 15517 }, { "epoch": 1.9740491031675358, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00011301040649414062, "grad_norm": 7.254992485046387, "learning_rate": 1e-05, "loss": 0.5011, "mean_token_accuracy": 0.8746373653411865, "num_tokens": 589043940.0, "step": 15518 }, { "epoch": 1.9741763134461263, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 5.940906524658203, "learning_rate": 1e-05, "loss": 0.5068, "mean_token_accuracy": 0.8654494881629944, "num_tokens": 589082187.0, "step": 15519 }, { "epoch": 1.9743035237247168, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.0001201629638671875, "grad_norm": 7.513321876525879, "learning_rate": 1e-05, "loss": 0.4959, "mean_token_accuracy": 0.8773004412651062, "num_tokens": 589126870.0, "step": 15520 }, { "epoch": 1.9744307340033074, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.243342876434326, "learning_rate": 1e-05, "loss": 0.5057, "mean_token_accuracy": 0.865365743637085, "num_tokens": 589168378.0, "step": 15521 }, { "epoch": 1.974557944281898, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00011777877807617188, "grad_norm": 7.452714920043945, "learning_rate": 1e-05, "loss": 0.5754, "mean_token_accuracy": 0.8544833660125732, "num_tokens": 589213070.0, "step": 15522 }, { "epoch": 1.9746851545604884, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.581328392028809, "learning_rate": 1e-05, "loss": 0.5268, "mean_token_accuracy": 0.8683422207832336, "num_tokens": 589253008.0, "step": 15523 }, { "epoch": 1.974812364839079, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 7.00618314743042, "learning_rate": 1e-05, "loss": 0.5223, "mean_token_accuracy": 0.8628085851669312, "num_tokens": 589291786.0, "step": 15524 }, { "epoch": 1.9749395751176695, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.671388626098633, "learning_rate": 1e-05, "loss": 0.5105, "mean_token_accuracy": 0.8715977072715759, "num_tokens": 589332045.0, "step": 15525 }, { "epoch": 1.97506678539626, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.445547580718994, "learning_rate": 1e-05, "loss": 0.5649, "mean_token_accuracy": 0.8487319946289062, "num_tokens": 589374983.0, "step": 15526 }, { "epoch": 1.9751939956748505, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.722949981689453, "learning_rate": 1e-05, "loss": 0.4382, "mean_token_accuracy": 0.8942465782165527, "num_tokens": 589419786.0, "step": 15527 }, { "epoch": 1.975321205953441, "ewc_loss": 0.10791015625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.96055269241333, "learning_rate": 1e-05, "loss": 0.4295, "mean_token_accuracy": 0.8891811370849609, "num_tokens": 589454365.0, "step": 15528 }, { "epoch": 1.9754484162320316, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.922738552093506, "learning_rate": 1e-05, "loss": 0.4973, "mean_token_accuracy": 0.8751131296157837, "num_tokens": 589486829.0, "step": 15529 }, { "epoch": 1.9755756265106221, "ewc_loss": 0.10546875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.671694755554199, "learning_rate": 1e-05, "loss": 0.4782, "mean_token_accuracy": 0.8721626996994019, "num_tokens": 589527500.0, "step": 15530 }, { "epoch": 1.9757028367892127, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00011539459228515625, "grad_norm": 7.329343318939209, "learning_rate": 1e-05, "loss": 0.4794, "mean_token_accuracy": 0.8823797702789307, "num_tokens": 589570064.0, "step": 15531 }, { "epoch": 1.9758300470678032, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.709999084472656, "learning_rate": 1e-05, "loss": 0.5346, "mean_token_accuracy": 0.8541832566261292, "num_tokens": 589605327.0, "step": 15532 }, { "epoch": 1.9759572573463937, "ewc_loss": 0.1513671875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00012683868408203125, "grad_norm": 7.7601847648620605, "learning_rate": 1e-05, "loss": 0.4935, "mean_token_accuracy": 0.8798022270202637, "num_tokens": 589640886.0, "step": 15533 }, { "epoch": 1.9760844676249842, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.189975261688232, "learning_rate": 1e-05, "loss": 0.5515, "mean_token_accuracy": 0.8562922477722168, "num_tokens": 589681178.0, "step": 15534 }, { "epoch": 1.9762116779035748, "ewc_loss": 0.150390625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.000125885009765625, "grad_norm": 7.750193119049072, "learning_rate": 1e-05, "loss": 0.4814, "mean_token_accuracy": 0.8817137479782104, "num_tokens": 589719073.0, "step": 15535 }, { "epoch": 1.976338888182165, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 6.682587146759033, "learning_rate": 1e-05, "loss": 0.5367, "mean_token_accuracy": 0.8611707091331482, "num_tokens": 589760528.0, "step": 15536 }, { "epoch": 1.9764660984607556, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00011396408081054688, "grad_norm": 7.296747207641602, "learning_rate": 1e-05, "loss": 0.5171, "mean_token_accuracy": 0.870230495929718, "num_tokens": 589795946.0, "step": 15537 }, { "epoch": 1.9765933087393461, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 6.728890895843506, "learning_rate": 1e-05, "loss": 0.5122, "mean_token_accuracy": 0.8701348900794983, "num_tokens": 589829139.0, "step": 15538 }, { "epoch": 1.9767205190179367, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.860806465148926, "learning_rate": 1e-05, "loss": 0.4619, "mean_token_accuracy": 0.8836813569068909, "num_tokens": 589862916.0, "step": 15539 }, { "epoch": 1.9768477292965272, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.57715368270874, "learning_rate": 1e-05, "loss": 0.4909, "mean_token_accuracy": 0.876801609992981, "num_tokens": 589900230.0, "step": 15540 }, { "epoch": 1.9769749395751177, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.551448345184326, "learning_rate": 1e-05, "loss": 0.4905, "mean_token_accuracy": 0.869881808757782, "num_tokens": 589938737.0, "step": 15541 }, { "epoch": 1.977102149853708, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.410114765167236, "learning_rate": 1e-05, "loss": 0.4713, "mean_token_accuracy": 0.8765137791633606, "num_tokens": 589977989.0, "step": 15542 }, { "epoch": 1.9772293601322986, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.41377592086792, "learning_rate": 1e-05, "loss": 0.4293, "mean_token_accuracy": 0.8872547745704651, "num_tokens": 590015547.0, "step": 15543 }, { "epoch": 1.977356570410889, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.239370822906494, "learning_rate": 1e-05, "loss": 0.4875, "mean_token_accuracy": 0.8721352219581604, "num_tokens": 590055257.0, "step": 15544 }, { "epoch": 1.9774837806894796, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.41572380065918, "learning_rate": 1e-05, "loss": 0.4915, "mean_token_accuracy": 0.8734896779060364, "num_tokens": 590087911.0, "step": 15545 }, { "epoch": 1.9776109909680701, "ewc_loss": 0.11181640625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 6.057930946350098, "learning_rate": 1e-05, "loss": 0.4965, "mean_token_accuracy": 0.8674393892288208, "num_tokens": 590128889.0, "step": 15546 }, { "epoch": 1.9777382012466607, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.500514507293701, "learning_rate": 1e-05, "loss": 0.534, "mean_token_accuracy": 0.8589717149734497, "num_tokens": 590168078.0, "step": 15547 }, { "epoch": 1.9778654115252512, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.873900413513184, "learning_rate": 1e-05, "loss": 0.543, "mean_token_accuracy": 0.8529636859893799, "num_tokens": 590211361.0, "step": 15548 }, { "epoch": 1.9779926218038417, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.72238302230835, "learning_rate": 1e-05, "loss": 0.5274, "mean_token_accuracy": 0.8607499599456787, "num_tokens": 590249922.0, "step": 15549 }, { "epoch": 1.9781198320824323, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.687511920928955, "learning_rate": 1e-05, "loss": 0.4822, "mean_token_accuracy": 0.871351957321167, "num_tokens": 590288284.0, "step": 15550 }, { "epoch": 1.9782470423610228, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00010967254638671875, "grad_norm": 7.1050286293029785, "learning_rate": 1e-05, "loss": 0.547, "mean_token_accuracy": 0.8567347526550293, "num_tokens": 590324373.0, "step": 15551 }, { "epoch": 1.9783742526396133, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.6320695877075195, "learning_rate": 1e-05, "loss": 0.4841, "mean_token_accuracy": 0.8713592886924744, "num_tokens": 590365374.0, "step": 15552 }, { "epoch": 1.9785014629182038, "ewc_loss": 0.1474609375, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.6006999015808105, "learning_rate": 1e-05, "loss": 0.5212, "mean_token_accuracy": 0.8688271045684814, "num_tokens": 590402712.0, "step": 15553 }, { "epoch": 1.9786286731967944, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 5.98356294631958, "learning_rate": 1e-05, "loss": 0.5848, "mean_token_accuracy": 0.852632462978363, "num_tokens": 590438434.0, "step": 15554 }, { "epoch": 1.978755883475385, "ewc_loss": 0.1533203125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00012874603271484375, "grad_norm": 7.794557571411133, "learning_rate": 1e-05, "loss": 0.5701, "mean_token_accuracy": 0.8609822988510132, "num_tokens": 590482619.0, "step": 15555 }, { "epoch": 1.9788830937539754, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 6.540487289428711, "learning_rate": 1e-05, "loss": 0.5637, "mean_token_accuracy": 0.8540139198303223, "num_tokens": 590527828.0, "step": 15556 }, { "epoch": 1.979010304032566, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00011730194091796875, "grad_norm": 7.417803764343262, "learning_rate": 1e-05, "loss": 0.5083, "mean_token_accuracy": 0.8729447722434998, "num_tokens": 590561758.0, "step": 15557 }, { "epoch": 1.9791375143111565, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00011110305786132812, "grad_norm": 6.749699592590332, "learning_rate": 1e-05, "loss": 0.4676, "mean_token_accuracy": 0.8823826313018799, "num_tokens": 590598323.0, "step": 15558 }, { "epoch": 1.979264724589747, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.870865821838379, "learning_rate": 1e-05, "loss": 0.5272, "mean_token_accuracy": 0.8649860620498657, "num_tokens": 590642469.0, "step": 15559 }, { "epoch": 1.9793919348683373, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.689640522003174, "learning_rate": 1e-05, "loss": 0.5225, "mean_token_accuracy": 0.8656898140907288, "num_tokens": 590683603.0, "step": 15560 }, { "epoch": 1.9795191451469278, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.419703006744385, "learning_rate": 1e-05, "loss": 0.4533, "mean_token_accuracy": 0.8786729574203491, "num_tokens": 590710136.0, "step": 15561 }, { "epoch": 1.9796463554255184, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.676649570465088, "learning_rate": 1e-05, "loss": 0.5124, "mean_token_accuracy": 0.8675099611282349, "num_tokens": 590743104.0, "step": 15562 }, { "epoch": 1.979773565704109, "ewc_loss": 0.11181640625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 6.097738742828369, "learning_rate": 1e-05, "loss": 0.502, "mean_token_accuracy": 0.8645543456077576, "num_tokens": 590781851.0, "step": 15563 }, { "epoch": 1.9799007759826994, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.751007556915283, "learning_rate": 1e-05, "loss": 0.5372, "mean_token_accuracy": 0.8592469692230225, "num_tokens": 590819865.0, "step": 15564 }, { "epoch": 1.98002798626129, "ewc_loss": 0.10888671875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.891776084899902, "learning_rate": 1e-05, "loss": 0.484, "mean_token_accuracy": 0.8788741827011108, "num_tokens": 590856478.0, "step": 15565 }, { "epoch": 1.9801551965398803, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.930270671844482, "learning_rate": 1e-05, "loss": 0.5408, "mean_token_accuracy": 0.8585217595100403, "num_tokens": 590895765.0, "step": 15566 }, { "epoch": 1.9802824068184708, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.79374885559082, "learning_rate": 1e-05, "loss": 0.4855, "mean_token_accuracy": 0.8734619617462158, "num_tokens": 590931238.0, "step": 15567 }, { "epoch": 1.9804096170970613, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00011157989501953125, "grad_norm": 7.182074546813965, "learning_rate": 1e-05, "loss": 0.5006, "mean_token_accuracy": 0.8748418092727661, "num_tokens": 590969875.0, "step": 15568 }, { "epoch": 1.9805368273756518, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 5.852678298950195, "learning_rate": 1e-05, "loss": 0.5285, "mean_token_accuracy": 0.8567392826080322, "num_tokens": 591007964.0, "step": 15569 }, { "epoch": 1.9806640376542424, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00011730194091796875, "grad_norm": 7.387284278869629, "learning_rate": 1e-05, "loss": 0.511, "mean_token_accuracy": 0.8750792741775513, "num_tokens": 591049339.0, "step": 15570 }, { "epoch": 1.980791247932833, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.068702697753906, "learning_rate": 1e-05, "loss": 0.4663, "mean_token_accuracy": 0.8791367411613464, "num_tokens": 591086429.0, "step": 15571 }, { "epoch": 1.9809184582114234, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00011730194091796875, "grad_norm": 7.398857116699219, "learning_rate": 1e-05, "loss": 0.5076, "mean_token_accuracy": 0.8740044236183167, "num_tokens": 591128252.0, "step": 15572 }, { "epoch": 1.981045668490014, "ewc_loss": 0.125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.323487758636475, "learning_rate": 1e-05, "loss": 0.4552, "mean_token_accuracy": 0.8821802139282227, "num_tokens": 591165049.0, "step": 15573 }, { "epoch": 1.9811728787686045, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.0001125335693359375, "grad_norm": 7.163716793060303, "learning_rate": 1e-05, "loss": 0.4856, "mean_token_accuracy": 0.8810384273529053, "num_tokens": 591196805.0, "step": 15574 }, { "epoch": 1.981300089047195, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.414083003997803, "learning_rate": 1e-05, "loss": 0.4987, "mean_token_accuracy": 0.87204909324646, "num_tokens": 591242897.0, "step": 15575 }, { "epoch": 1.9814272993257855, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.83410120010376, "learning_rate": 1e-05, "loss": 0.4972, "mean_token_accuracy": 0.8689025044441223, "num_tokens": 591280684.0, "step": 15576 }, { "epoch": 1.981554509604376, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.408135414123535, "learning_rate": 1e-05, "loss": 0.4834, "mean_token_accuracy": 0.8732248544692993, "num_tokens": 591315264.0, "step": 15577 }, { "epoch": 1.9816817198829666, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.525763034820557, "learning_rate": 1e-05, "loss": 0.509, "mean_token_accuracy": 0.867756187915802, "num_tokens": 591358028.0, "step": 15578 }, { "epoch": 1.9818089301615571, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.350972652435303, "learning_rate": 1e-05, "loss": 0.5658, "mean_token_accuracy": 0.8525671362876892, "num_tokens": 591404272.0, "step": 15579 }, { "epoch": 1.9819361404401477, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.329620838165283, "learning_rate": 1e-05, "loss": 0.448, "mean_token_accuracy": 0.8832447528839111, "num_tokens": 591439896.0, "step": 15580 }, { "epoch": 1.9820633507187382, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.270273685455322, "learning_rate": 1e-05, "loss": 0.4774, "mean_token_accuracy": 0.8767179250717163, "num_tokens": 591477318.0, "step": 15581 }, { "epoch": 1.9821905609973287, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.25120210647583, "learning_rate": 1e-05, "loss": 0.511, "mean_token_accuracy": 0.8661028146743774, "num_tokens": 591518600.0, "step": 15582 }, { "epoch": 1.9823177712759192, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.19843053817749, "learning_rate": 1e-05, "loss": 0.4909, "mean_token_accuracy": 0.8671166300773621, "num_tokens": 591547414.0, "step": 15583 }, { "epoch": 1.9824449815545098, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.222321510314941, "learning_rate": 1e-05, "loss": 0.5341, "mean_token_accuracy": 0.8583918809890747, "num_tokens": 591594270.0, "step": 15584 }, { "epoch": 1.9825721918331, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.092813968658447, "learning_rate": 1e-05, "loss": 0.4748, "mean_token_accuracy": 0.873837947845459, "num_tokens": 591638252.0, "step": 15585 }, { "epoch": 1.9826994021116906, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.263765335083008, "learning_rate": 1e-05, "loss": 0.5658, "mean_token_accuracy": 0.8493880033493042, "num_tokens": 591680268.0, "step": 15586 }, { "epoch": 1.9828266123902811, "ewc_loss": 0.11083984375, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.978171348571777, "learning_rate": 1e-05, "loss": 0.483, "mean_token_accuracy": 0.8743104934692383, "num_tokens": 591717724.0, "step": 15587 }, { "epoch": 1.9829538226688717, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.370238304138184, "learning_rate": 1e-05, "loss": 0.4503, "mean_token_accuracy": 0.8850383758544922, "num_tokens": 591754562.0, "step": 15588 }, { "epoch": 1.9830810329474622, "ewc_loss": 0.10791015625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.819980621337891, "learning_rate": 1e-05, "loss": 0.4671, "mean_token_accuracy": 0.8743306994438171, "num_tokens": 591794838.0, "step": 15589 }, { "epoch": 1.9832082432260527, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.607379913330078, "learning_rate": 1e-05, "loss": 0.5596, "mean_token_accuracy": 0.8514400720596313, "num_tokens": 591834771.0, "step": 15590 }, { "epoch": 1.983335453504643, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.626969337463379, "learning_rate": 1e-05, "loss": 0.5038, "mean_token_accuracy": 0.8577353954315186, "num_tokens": 591872764.0, "step": 15591 }, { "epoch": 1.9834626637832335, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 7.014647006988525, "learning_rate": 1e-05, "loss": 0.5172, "mean_token_accuracy": 0.8653693199157715, "num_tokens": 591907035.0, "step": 15592 }, { "epoch": 1.983589874061824, "ewc_loss": 0.10498046875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 8.058547973632812e-05, "grad_norm": 5.498849868774414, "learning_rate": 1e-05, "loss": 0.4592, "mean_token_accuracy": 0.8751024007797241, "num_tokens": 591940797.0, "step": 15593 }, { "epoch": 1.9837170843404146, "ewc_loss": 0.1494140625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00012493133544921875, "grad_norm": 7.618733882904053, "learning_rate": 1e-05, "loss": 0.4739, "mean_token_accuracy": 0.8837399482727051, "num_tokens": 591976872.0, "step": 15594 }, { "epoch": 1.9838442946190051, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 5.772818565368652, "learning_rate": 1e-05, "loss": 0.5237, "mean_token_accuracy": 0.8549346923828125, "num_tokens": 592015933.0, "step": 15595 }, { "epoch": 1.9839715048975957, "ewc_loss": 0.1611328125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00013637542724609375, "grad_norm": 8.075242042541504, "learning_rate": 1e-05, "loss": 0.5076, "mean_token_accuracy": 0.877160906791687, "num_tokens": 592046796.0, "step": 15596 }, { "epoch": 1.9840987151761862, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00011301040649414062, "grad_norm": 6.5556488037109375, "learning_rate": 1e-05, "loss": 0.5161, "mean_token_accuracy": 0.8695533275604248, "num_tokens": 592092841.0, "step": 15597 }, { "epoch": 1.9842259254547767, "ewc_loss": 0.1513671875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00012683868408203125, "grad_norm": 7.782837390899658, "learning_rate": 1e-05, "loss": 0.584, "mean_token_accuracy": 0.8538720607757568, "num_tokens": 592134118.0, "step": 15598 }, { "epoch": 1.9843531357333672, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00011777877807617188, "grad_norm": 6.911629676818848, "learning_rate": 1e-05, "loss": 0.5302, "mean_token_accuracy": 0.8641916513442993, "num_tokens": 592167292.0, "step": 15599 }, { "epoch": 1.9844803460119578, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00011301040649414062, "grad_norm": 7.243807315826416, "learning_rate": 1e-05, "loss": 0.4821, "mean_token_accuracy": 0.8812088370323181, "num_tokens": 592205616.0, "step": 15600 }, { "epoch": 1.9846075562905483, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00011014938354492188, "grad_norm": 6.746554374694824, "learning_rate": 1e-05, "loss": 0.4905, "mean_token_accuracy": 0.8772611618041992, "num_tokens": 592246650.0, "step": 15601 }, { "epoch": 1.9847347665691388, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.868346691131592, "learning_rate": 1e-05, "loss": 0.4866, "mean_token_accuracy": 0.8745408654212952, "num_tokens": 592284729.0, "step": 15602 }, { "epoch": 1.9848619768477294, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.54294490814209, "learning_rate": 1e-05, "loss": 0.488, "mean_token_accuracy": 0.8736915588378906, "num_tokens": 592318423.0, "step": 15603 }, { "epoch": 1.9849891871263199, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.60591983795166, "learning_rate": 1e-05, "loss": 0.4913, "mean_token_accuracy": 0.8684178590774536, "num_tokens": 592351304.0, "step": 15604 }, { "epoch": 1.9851163974049104, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.369034290313721, "learning_rate": 1e-05, "loss": 0.4893, "mean_token_accuracy": 0.8726888298988342, "num_tokens": 592388353.0, "step": 15605 }, { "epoch": 1.985243607683501, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.475427627563477, "learning_rate": 1e-05, "loss": 0.4822, "mean_token_accuracy": 0.8769645690917969, "num_tokens": 592427651.0, "step": 15606 }, { "epoch": 1.9853708179620915, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.203994274139404, "learning_rate": 1e-05, "loss": 0.4918, "mean_token_accuracy": 0.8707048892974854, "num_tokens": 592463638.0, "step": 15607 }, { "epoch": 1.985498028240682, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.444784164428711, "learning_rate": 1e-05, "loss": 0.5313, "mean_token_accuracy": 0.8608363270759583, "num_tokens": 592501508.0, "step": 15608 }, { "epoch": 1.9856252385192723, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 6.036776542663574, "learning_rate": 1e-05, "loss": 0.4682, "mean_token_accuracy": 0.8786359429359436, "num_tokens": 592538418.0, "step": 15609 }, { "epoch": 1.9857524487978628, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.525779724121094, "learning_rate": 1e-05, "loss": 0.4498, "mean_token_accuracy": 0.8850162625312805, "num_tokens": 592573552.0, "step": 15610 }, { "epoch": 1.9858796590764534, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.848664283752441, "learning_rate": 1e-05, "loss": 0.4709, "mean_token_accuracy": 0.8727052807807922, "num_tokens": 592607212.0, "step": 15611 }, { "epoch": 1.986006869355044, "ewc_loss": 0.125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.765836238861084, "learning_rate": 1e-05, "loss": 0.5271, "mean_token_accuracy": 0.8645486831665039, "num_tokens": 592649910.0, "step": 15612 }, { "epoch": 1.9861340796336344, "ewc_loss": 0.10595703125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.658645153045654, "learning_rate": 1e-05, "loss": 0.4782, "mean_token_accuracy": 0.872040867805481, "num_tokens": 592689555.0, "step": 15613 }, { "epoch": 1.986261289912225, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00011301040649414062, "grad_norm": 7.200463771820068, "learning_rate": 1e-05, "loss": 0.4977, "mean_token_accuracy": 0.8756904006004333, "num_tokens": 592731921.0, "step": 15614 }, { "epoch": 1.9863885001908153, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.655999183654785, "learning_rate": 1e-05, "loss": 0.4867, "mean_token_accuracy": 0.8740379810333252, "num_tokens": 592763631.0, "step": 15615 }, { "epoch": 1.9865157104694058, "ewc_loss": 0.1494140625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00012493133544921875, "grad_norm": 7.706393241882324, "learning_rate": 1e-05, "loss": 0.4662, "mean_token_accuracy": 0.8894091844558716, "num_tokens": 592796961.0, "step": 15616 }, { "epoch": 1.9866429207479963, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.076089859008789, "learning_rate": 1e-05, "loss": 0.4571, "mean_token_accuracy": 0.8774031400680542, "num_tokens": 592831380.0, "step": 15617 }, { "epoch": 1.9867701310265868, "ewc_loss": 0.15234375, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.0001277923583984375, "grad_norm": 7.819963455200195, "learning_rate": 1e-05, "loss": 0.5552, "mean_token_accuracy": 0.8640413880348206, "num_tokens": 592873606.0, "step": 15618 }, { "epoch": 1.9868973413051774, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 6.650089740753174, "learning_rate": 1e-05, "loss": 0.5421, "mean_token_accuracy": 0.8631862998008728, "num_tokens": 592919311.0, "step": 15619 }, { "epoch": 1.987024551583768, "ewc_loss": 0.140625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00011587142944335938, "grad_norm": 7.368946552276611, "learning_rate": 1e-05, "loss": 0.5864, "mean_token_accuracy": 0.8478604555130005, "num_tokens": 592961818.0, "step": 15620 }, { "epoch": 1.9871517618623584, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00011157989501953125, "grad_norm": 6.781128406524658, "learning_rate": 1e-05, "loss": 0.5146, "mean_token_accuracy": 0.8697071075439453, "num_tokens": 593003320.0, "step": 15621 }, { "epoch": 1.987278972140949, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.8324198722839355, "learning_rate": 1e-05, "loss": 0.537, "mean_token_accuracy": 0.8593511581420898, "num_tokens": 593050887.0, "step": 15622 }, { "epoch": 1.9874061824195395, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.698426723480225, "learning_rate": 1e-05, "loss": 0.5386, "mean_token_accuracy": 0.8611142635345459, "num_tokens": 593086296.0, "step": 15623 }, { "epoch": 1.98753339269813, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.416049957275391, "learning_rate": 1e-05, "loss": 0.5514, "mean_token_accuracy": 0.8525480628013611, "num_tokens": 593122135.0, "step": 15624 }, { "epoch": 1.9876606029767205, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.658008575439453, "learning_rate": 1e-05, "loss": 0.5254, "mean_token_accuracy": 0.8622608780860901, "num_tokens": 593158845.0, "step": 15625 }, { "epoch": 1.987787813255311, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 6.110142230987549, "learning_rate": 1e-05, "loss": 0.4568, "mean_token_accuracy": 0.8771746158599854, "num_tokens": 593199386.0, "step": 15626 }, { "epoch": 1.9879150235339016, "ewc_loss": 0.125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.699852466583252, "learning_rate": 1e-05, "loss": 0.4652, "mean_token_accuracy": 0.8792897462844849, "num_tokens": 593237061.0, "step": 15627 }, { "epoch": 1.9880422338124921, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.927728652954102, "learning_rate": 1e-05, "loss": 0.49, "mean_token_accuracy": 0.8713775277137756, "num_tokens": 593275220.0, "step": 15628 }, { "epoch": 1.9881694440910826, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.843487739562988, "learning_rate": 1e-05, "loss": 0.5642, "mean_token_accuracy": 0.8497321605682373, "num_tokens": 593314576.0, "step": 15629 }, { "epoch": 1.9882966543696732, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.799868106842041, "learning_rate": 1e-05, "loss": 0.4572, "mean_token_accuracy": 0.8780971765518188, "num_tokens": 593362905.0, "step": 15630 }, { "epoch": 1.9884238646482637, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 7.097267150878906, "learning_rate": 1e-05, "loss": 0.4789, "mean_token_accuracy": 0.878294825553894, "num_tokens": 593402486.0, "step": 15631 }, { "epoch": 1.9885510749268542, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.788308143615723, "learning_rate": 1e-05, "loss": 0.4472, "mean_token_accuracy": 0.8810113668441772, "num_tokens": 593440555.0, "step": 15632 }, { "epoch": 1.9886782852054448, "ewc_loss": 0.140625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00011682510375976562, "grad_norm": 7.36440372467041, "learning_rate": 1e-05, "loss": 0.5211, "mean_token_accuracy": 0.8695611953735352, "num_tokens": 593480488.0, "step": 15633 }, { "epoch": 1.988805495484035, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 5.981023788452148, "learning_rate": 1e-05, "loss": 0.464, "mean_token_accuracy": 0.876333475112915, "num_tokens": 593518615.0, "step": 15634 }, { "epoch": 1.9889327057626256, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00011968612670898438, "grad_norm": 7.467934608459473, "learning_rate": 1e-05, "loss": 0.4996, "mean_token_accuracy": 0.8768993616104126, "num_tokens": 593556990.0, "step": 15635 }, { "epoch": 1.9890599160412161, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.265710353851318, "learning_rate": 1e-05, "loss": 0.5374, "mean_token_accuracy": 0.8613163828849792, "num_tokens": 593599438.0, "step": 15636 }, { "epoch": 1.9891871263198067, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00011539459228515625, "grad_norm": 7.289977073669434, "learning_rate": 1e-05, "loss": 0.5568, "mean_token_accuracy": 0.8572871685028076, "num_tokens": 593635991.0, "step": 15637 }, { "epoch": 1.9893143365983972, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.43044376373291, "learning_rate": 1e-05, "loss": 0.5055, "mean_token_accuracy": 0.8716980218887329, "num_tokens": 593676600.0, "step": 15638 }, { "epoch": 1.9894415468769877, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.967576026916504, "learning_rate": 1e-05, "loss": 0.4807, "mean_token_accuracy": 0.8800437450408936, "num_tokens": 593709069.0, "step": 15639 }, { "epoch": 1.989568757155578, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.446597099304199, "learning_rate": 1e-05, "loss": 0.4341, "mean_token_accuracy": 0.8951816558837891, "num_tokens": 593740361.0, "step": 15640 }, { "epoch": 1.9896959674341685, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.611318111419678, "learning_rate": 1e-05, "loss": 0.4374, "mean_token_accuracy": 0.885672926902771, "num_tokens": 593775424.0, "step": 15641 }, { "epoch": 1.989823177712759, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.408048152923584, "learning_rate": 1e-05, "loss": 0.5288, "mean_token_accuracy": 0.863211989402771, "num_tokens": 593816562.0, "step": 15642 }, { "epoch": 1.9899503879913496, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.357883930206299, "learning_rate": 1e-05, "loss": 0.5058, "mean_token_accuracy": 0.8714839816093445, "num_tokens": 593853272.0, "step": 15643 }, { "epoch": 1.9900775982699401, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.336974620819092, "learning_rate": 1e-05, "loss": 0.466, "mean_token_accuracy": 0.8759914636611938, "num_tokens": 593885308.0, "step": 15644 }, { "epoch": 1.9902048085485307, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.214892864227295, "learning_rate": 1e-05, "loss": 0.4507, "mean_token_accuracy": 0.8829840421676636, "num_tokens": 593918150.0, "step": 15645 }, { "epoch": 1.9903320188271212, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.266015529632568, "learning_rate": 1e-05, "loss": 0.5084, "mean_token_accuracy": 0.8652762174606323, "num_tokens": 593961947.0, "step": 15646 }, { "epoch": 1.9904592291057117, "ewc_loss": 0.11474609375, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.156097888946533, "learning_rate": 1e-05, "loss": 0.536, "mean_token_accuracy": 0.8548640012741089, "num_tokens": 594000659.0, "step": 15647 }, { "epoch": 1.9905864393843022, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.191864967346191, "learning_rate": 1e-05, "loss": 0.4897, "mean_token_accuracy": 0.8712806701660156, "num_tokens": 594040033.0, "step": 15648 }, { "epoch": 1.9907136496628928, "ewc_loss": 0.11474609375, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.134514331817627, "learning_rate": 1e-05, "loss": 0.4735, "mean_token_accuracy": 0.8766056299209595, "num_tokens": 594080534.0, "step": 15649 }, { "epoch": 1.9908408599414833, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.119054794311523, "learning_rate": 1e-05, "loss": 0.5074, "mean_token_accuracy": 0.8672155737876892, "num_tokens": 594118015.0, "step": 15650 }, { "epoch": 1.9909680702200738, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.1512131690979, "learning_rate": 1e-05, "loss": 0.4856, "mean_token_accuracy": 0.8742339611053467, "num_tokens": 594153839.0, "step": 15651 }, { "epoch": 1.9910952804986644, "ewc_loss": 0.11376953125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 6.047482967376709, "learning_rate": 1e-05, "loss": 0.46, "mean_token_accuracy": 0.8809001445770264, "num_tokens": 594195672.0, "step": 15652 }, { "epoch": 1.9912224907772549, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.185502052307129, "learning_rate": 1e-05, "loss": 0.5064, "mean_token_accuracy": 0.8648209571838379, "num_tokens": 594239178.0, "step": 15653 }, { "epoch": 1.9913497010558454, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 5.975614070892334, "learning_rate": 1e-05, "loss": 0.4758, "mean_token_accuracy": 0.8728154897689819, "num_tokens": 594274938.0, "step": 15654 }, { "epoch": 1.991476911334436, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.271324157714844, "learning_rate": 1e-05, "loss": 0.4643, "mean_token_accuracy": 0.8800896406173706, "num_tokens": 594310632.0, "step": 15655 }, { "epoch": 1.9916041216130265, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.862676620483398, "learning_rate": 1e-05, "loss": 0.5027, "mean_token_accuracy": 0.8644244074821472, "num_tokens": 594344308.0, "step": 15656 }, { "epoch": 1.991731331891617, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.431649208068848, "learning_rate": 1e-05, "loss": 0.5033, "mean_token_accuracy": 0.8686151504516602, "num_tokens": 594385287.0, "step": 15657 }, { "epoch": 1.9918585421702073, "ewc_loss": 0.10693359375, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.681868076324463, "learning_rate": 1e-05, "loss": 0.4496, "mean_token_accuracy": 0.881784975528717, "num_tokens": 594422521.0, "step": 15658 }, { "epoch": 1.9919857524487978, "ewc_loss": 0.125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.766921043395996, "learning_rate": 1e-05, "loss": 0.5542, "mean_token_accuracy": 0.8555269241333008, "num_tokens": 594466837.0, "step": 15659 }, { "epoch": 1.9921129627273884, "ewc_loss": 0.1044921875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 8.0108642578125e-05, "grad_norm": 5.471110820770264, "learning_rate": 1e-05, "loss": 0.4937, "mean_token_accuracy": 0.8699109554290771, "num_tokens": 594501283.0, "step": 15660 }, { "epoch": 1.9922401730059789, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00011777877807617188, "grad_norm": 7.3563456535339355, "learning_rate": 1e-05, "loss": 0.5634, "mean_token_accuracy": 0.8564720153808594, "num_tokens": 594544627.0, "step": 15661 }, { "epoch": 1.9923673832845694, "ewc_loss": 0.10595703125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.442130088806152, "learning_rate": 1e-05, "loss": 0.5495, "mean_token_accuracy": 0.8491957783699036, "num_tokens": 594586363.0, "step": 15662 }, { "epoch": 1.99249459356316, "ewc_loss": 0.162109375, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.0001373291015625, "grad_norm": 8.168298721313477, "learning_rate": 1e-05, "loss": 0.5275, "mean_token_accuracy": 0.867708683013916, "num_tokens": 594624950.0, "step": 15663 }, { "epoch": 1.9926218038417502, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.222685813903809, "learning_rate": 1e-05, "loss": 0.5586, "mean_token_accuracy": 0.8556276559829712, "num_tokens": 594665103.0, "step": 15664 }, { "epoch": 1.9927490141203408, "ewc_loss": 0.1669921875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00014209747314453125, "grad_norm": 8.333393096923828, "learning_rate": 1e-05, "loss": 0.5034, "mean_token_accuracy": 0.8812084197998047, "num_tokens": 594699506.0, "step": 15665 }, { "epoch": 1.9928762243989313, "ewc_loss": 0.1513671875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00012683868408203125, "grad_norm": 7.131301403045654, "learning_rate": 1e-05, "loss": 0.551, "mean_token_accuracy": 0.8658308982849121, "num_tokens": 594741771.0, "step": 15666 }, { "epoch": 1.9930034346775218, "ewc_loss": 0.1484375, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.0001239776611328125, "grad_norm": 7.653353214263916, "learning_rate": 1e-05, "loss": 0.5025, "mean_token_accuracy": 0.8788880705833435, "num_tokens": 594784705.0, "step": 15667 }, { "epoch": 1.9931306449561124, "ewc_loss": 0.1455078125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00012111663818359375, "grad_norm": 7.064284801483154, "learning_rate": 1e-05, "loss": 0.4978, "mean_token_accuracy": 0.8758472204208374, "num_tokens": 594827800.0, "step": 15668 }, { "epoch": 1.993257855234703, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00011348724365234375, "grad_norm": 7.206897735595703, "learning_rate": 1e-05, "loss": 0.5228, "mean_token_accuracy": 0.86735600233078, "num_tokens": 594865150.0, "step": 15669 }, { "epoch": 1.9933850655132934, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 6.741137504577637, "learning_rate": 1e-05, "loss": 0.5613, "mean_token_accuracy": 0.8562427759170532, "num_tokens": 594903440.0, "step": 15670 }, { "epoch": 1.993512275791884, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.9287004470825195, "learning_rate": 1e-05, "loss": 0.4892, "mean_token_accuracy": 0.8742918968200684, "num_tokens": 594936761.0, "step": 15671 }, { "epoch": 1.9936394860704745, "ewc_loss": 0.125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.4755473136901855, "learning_rate": 1e-05, "loss": 0.5236, "mean_token_accuracy": 0.8569133281707764, "num_tokens": 594972298.0, "step": 15672 }, { "epoch": 1.993766696349065, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.783443450927734, "learning_rate": 1e-05, "loss": 0.5056, "mean_token_accuracy": 0.8706638216972351, "num_tokens": 595009530.0, "step": 15673 }, { "epoch": 1.9938939066276555, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.2265167236328125, "learning_rate": 1e-05, "loss": 0.4914, "mean_token_accuracy": 0.8737108707427979, "num_tokens": 595048147.0, "step": 15674 }, { "epoch": 1.994021116906246, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.799592018127441, "learning_rate": 1e-05, "loss": 0.4708, "mean_token_accuracy": 0.882203221321106, "num_tokens": 595083580.0, "step": 15675 }, { "epoch": 1.9941483271848366, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 6.012897491455078, "learning_rate": 1e-05, "loss": 0.5544, "mean_token_accuracy": 0.8507410287857056, "num_tokens": 595122989.0, "step": 15676 }, { "epoch": 1.9942755374634271, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.984210014343262, "learning_rate": 1e-05, "loss": 0.5187, "mean_token_accuracy": 0.8661861419677734, "num_tokens": 595165378.0, "step": 15677 }, { "epoch": 1.9944027477420176, "ewc_loss": 0.10693359375, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.882570743560791, "learning_rate": 1e-05, "loss": 0.5624, "mean_token_accuracy": 0.848565936088562, "num_tokens": 595203802.0, "step": 15678 }, { "epoch": 1.9945299580206082, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00011348724365234375, "grad_norm": 7.308678150177002, "learning_rate": 1e-05, "loss": 0.5293, "mean_token_accuracy": 0.8645416498184204, "num_tokens": 595242056.0, "step": 15679 }, { "epoch": 1.9946571682991987, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 5.929780960083008, "learning_rate": 1e-05, "loss": 0.5343, "mean_token_accuracy": 0.855820894241333, "num_tokens": 595276381.0, "step": 15680 }, { "epoch": 1.9947843785777892, "ewc_loss": 0.1474609375, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.592559814453125, "learning_rate": 1e-05, "loss": 0.5224, "mean_token_accuracy": 0.8710347414016724, "num_tokens": 595318551.0, "step": 15681 }, { "epoch": 1.9949115888563798, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.296414375305176, "learning_rate": 1e-05, "loss": 0.5285, "mean_token_accuracy": 0.8603099584579468, "num_tokens": 595361802.0, "step": 15682 }, { "epoch": 1.99503879913497, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00011777877807617188, "grad_norm": 7.478453159332275, "learning_rate": 1e-05, "loss": 0.5213, "mean_token_accuracy": 0.8713208436965942, "num_tokens": 595402853.0, "step": 15683 }, { "epoch": 1.9951660094135606, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.6384100914001465, "learning_rate": 1e-05, "loss": 0.4973, "mean_token_accuracy": 0.8753278851509094, "num_tokens": 595443774.0, "step": 15684 }, { "epoch": 1.9952932196921511, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.972956657409668, "learning_rate": 1e-05, "loss": 0.5237, "mean_token_accuracy": 0.8636540174484253, "num_tokens": 595485395.0, "step": 15685 }, { "epoch": 1.9954204299707416, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.717085838317871, "learning_rate": 1e-05, "loss": 0.5254, "mean_token_accuracy": 0.8650698065757751, "num_tokens": 595527724.0, "step": 15686 }, { "epoch": 1.9955476402493322, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.411115646362305, "learning_rate": 1e-05, "loss": 0.537, "mean_token_accuracy": 0.8572306632995605, "num_tokens": 595569081.0, "step": 15687 }, { "epoch": 1.9956748505279227, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.762743949890137, "learning_rate": 1e-05, "loss": 0.5325, "mean_token_accuracy": 0.8610101938247681, "num_tokens": 595611400.0, "step": 15688 }, { "epoch": 1.995802060806513, "ewc_loss": 0.10791015625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.955042839050293, "learning_rate": 1e-05, "loss": 0.5067, "mean_token_accuracy": 0.867615818977356, "num_tokens": 595648063.0, "step": 15689 }, { "epoch": 1.9959292710851035, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.968750953674316, "learning_rate": 1e-05, "loss": 0.555, "mean_token_accuracy": 0.85642409324646, "num_tokens": 595688679.0, "step": 15690 }, { "epoch": 1.996056481363694, "ewc_loss": 0.10595703125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.699552536010742, "learning_rate": 1e-05, "loss": 0.5128, "mean_token_accuracy": 0.8594892024993896, "num_tokens": 595726120.0, "step": 15691 }, { "epoch": 1.9961836916422846, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00011491775512695312, "grad_norm": 7.341031551361084, "learning_rate": 1e-05, "loss": 0.5032, "mean_token_accuracy": 0.8710870742797852, "num_tokens": 595762020.0, "step": 15692 }, { "epoch": 1.9963109019208751, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.753775596618652, "learning_rate": 1e-05, "loss": 0.5475, "mean_token_accuracy": 0.8481658697128296, "num_tokens": 595797909.0, "step": 15693 }, { "epoch": 1.9964381121994657, "ewc_loss": 0.1513671875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00012683868408203125, "grad_norm": 7.7355756759643555, "learning_rate": 1e-05, "loss": 0.5376, "mean_token_accuracy": 0.8660141229629517, "num_tokens": 595837542.0, "step": 15694 }, { "epoch": 1.9965653224780562, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.232332706451416, "learning_rate": 1e-05, "loss": 0.4862, "mean_token_accuracy": 0.873487114906311, "num_tokens": 595874006.0, "step": 15695 }, { "epoch": 1.9966925327566467, "ewc_loss": 0.1474609375, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.6880269050598145, "learning_rate": 1e-05, "loss": 0.5436, "mean_token_accuracy": 0.863431453704834, "num_tokens": 595904098.0, "step": 15696 }, { "epoch": 1.9968197430352372, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 6.654874324798584, "learning_rate": 1e-05, "loss": 0.4966, "mean_token_accuracy": 0.8741025924682617, "num_tokens": 595945725.0, "step": 15697 }, { "epoch": 1.9969469533138278, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00011301040649414062, "grad_norm": 7.252157211303711, "learning_rate": 1e-05, "loss": 0.516, "mean_token_accuracy": 0.8694678544998169, "num_tokens": 595977826.0, "step": 15698 }, { "epoch": 1.9970741635924183, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.690512180328369, "learning_rate": 1e-05, "loss": 0.5365, "mean_token_accuracy": 0.8615756034851074, "num_tokens": 596011350.0, "step": 15699 }, { "epoch": 1.9972013738710088, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.823324203491211, "learning_rate": 1e-05, "loss": 0.4574, "mean_token_accuracy": 0.884189248085022, "num_tokens": 596050688.0, "step": 15700 }, { "epoch": 1.9973285841495994, "ewc_loss": 0.125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.545240879058838, "learning_rate": 1e-05, "loss": 0.4754, "mean_token_accuracy": 0.8803881406784058, "num_tokens": 596090711.0, "step": 15701 }, { "epoch": 1.9974557944281899, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.5478997230529785, "learning_rate": 1e-05, "loss": 0.5085, "mean_token_accuracy": 0.8684918284416199, "num_tokens": 596130346.0, "step": 15702 }, { "epoch": 1.9975830047067804, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.374507427215576, "learning_rate": 1e-05, "loss": 0.5436, "mean_token_accuracy": 0.8608444929122925, "num_tokens": 596173035.0, "step": 15703 }, { "epoch": 1.997710214985371, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.456033706665039, "learning_rate": 1e-05, "loss": 0.5465, "mean_token_accuracy": 0.8519048690795898, "num_tokens": 596208105.0, "step": 15704 }, { "epoch": 1.9978374252639615, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.170616149902344, "learning_rate": 1e-05, "loss": 0.4515, "mean_token_accuracy": 0.8840685486793518, "num_tokens": 596247367.0, "step": 15705 }, { "epoch": 1.997964635542552, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.490431308746338, "learning_rate": 1e-05, "loss": 0.5142, "mean_token_accuracy": 0.8682152032852173, "num_tokens": 596288226.0, "step": 15706 }, { "epoch": 1.9980918458211423, "ewc_loss": 0.10888671875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.981945991516113, "learning_rate": 1e-05, "loss": 0.5206, "mean_token_accuracy": 0.8580148220062256, "num_tokens": 596329422.0, "step": 15707 }, { "epoch": 1.9982190560997328, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.646749973297119, "learning_rate": 1e-05, "loss": 0.4918, "mean_token_accuracy": 0.8734951019287109, "num_tokens": 596367447.0, "step": 15708 }, { "epoch": 1.9983462663783234, "ewc_loss": 0.10693359375, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.7961015701293945, "learning_rate": 1e-05, "loss": 0.4946, "mean_token_accuracy": 0.8659558296203613, "num_tokens": 596401564.0, "step": 15709 }, { "epoch": 1.9984734766569139, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.952146530151367, "learning_rate": 1e-05, "loss": 0.5298, "mean_token_accuracy": 0.8643592596054077, "num_tokens": 596441076.0, "step": 15710 }, { "epoch": 1.9986006869355044, "ewc_loss": 0.10595703125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.671106338500977, "learning_rate": 1e-05, "loss": 0.4589, "mean_token_accuracy": 0.8785421848297119, "num_tokens": 596480360.0, "step": 15711 }, { "epoch": 1.998727897214095, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00011873245239257812, "grad_norm": 7.4084978103637695, "learning_rate": 1e-05, "loss": 0.5464, "mean_token_accuracy": 0.8597294092178345, "num_tokens": 596512836.0, "step": 15712 }, { "epoch": 1.9988551074926852, "ewc_loss": 0.11376953125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 5.847135543823242, "learning_rate": 1e-05, "loss": 0.4526, "mean_token_accuracy": 0.8840709328651428, "num_tokens": 596549154.0, "step": 15713 }, { "epoch": 1.9989823177712758, "ewc_loss": 0.1513671875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00012683868408203125, "grad_norm": 7.720056533813477, "learning_rate": 1e-05, "loss": 0.5128, "mean_token_accuracy": 0.8775889873504639, "num_tokens": 596590528.0, "step": 15714 }, { "epoch": 1.9991095280498663, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.379136085510254, "learning_rate": 1e-05, "loss": 0.5055, "mean_token_accuracy": 0.8703203797340393, "num_tokens": 596623926.0, "step": 15715 }, { "epoch": 1.9992367383284568, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.0001201629638671875, "grad_norm": 7.485362529754639, "learning_rate": 1e-05, "loss": 0.519, "mean_token_accuracy": 0.8716487884521484, "num_tokens": 596661102.0, "step": 15716 }, { "epoch": 1.9993639486070474, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 6.695847988128662, "learning_rate": 1e-05, "loss": 0.4759, "mean_token_accuracy": 0.8835022449493408, "num_tokens": 596701113.0, "step": 15717 }, { "epoch": 1.9994911588856379, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.941699981689453, "learning_rate": 1e-05, "loss": 0.482, "mean_token_accuracy": 0.8799293041229248, "num_tokens": 596735130.0, "step": 15718 }, { "epoch": 1.9996183691642284, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.709041118621826, "learning_rate": 1e-05, "loss": 0.5532, "mean_token_accuracy": 0.8562644720077515, "num_tokens": 596774612.0, "step": 15719 }, { "epoch": 1.999745579442819, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.423510551452637, "learning_rate": 1e-05, "loss": 0.4955, "mean_token_accuracy": 0.8688672780990601, "num_tokens": 596811033.0, "step": 15720 }, { "epoch": 1.9998727897214095, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.729362964630127, "learning_rate": 1e-05, "loss": 0.5336, "mean_token_accuracy": 0.8581240177154541, "num_tokens": 596849169.0, "step": 15721 }, { "epoch": 2.0, "ewc_loss": 0.11083984375, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 6.035297393798828, "learning_rate": 1e-05, "loss": 0.464, "mean_token_accuracy": 0.8764654994010925, "num_tokens": 596886661.0, "step": 15722 }, { "epoch": 2.0001272102785905, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.860963344573975, "learning_rate": 1e-05, "loss": 0.4656, "mean_token_accuracy": 0.8837143778800964, "num_tokens": 596927769.0, "step": 15723 }, { "epoch": 2.000254420557181, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.817877769470215, "learning_rate": 1e-05, "loss": 0.4628, "mean_token_accuracy": 0.8786648511886597, "num_tokens": 596966558.0, "step": 15724 }, { "epoch": 2.0003816308357716, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00011014938354492188, "grad_norm": 7.1239399909973145, "learning_rate": 1e-05, "loss": 0.4617, "mean_token_accuracy": 0.884048342704773, "num_tokens": 597007424.0, "step": 15725 }, { "epoch": 2.000508841114362, "ewc_loss": 0.10986328125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.805589199066162, "learning_rate": 1e-05, "loss": 0.4519, "mean_token_accuracy": 0.8788670301437378, "num_tokens": 597043544.0, "step": 15726 }, { "epoch": 2.0006360513929526, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00011730194091796875, "grad_norm": 7.397198677062988, "learning_rate": 1e-05, "loss": 0.4891, "mean_token_accuracy": 0.877028226852417, "num_tokens": 597081079.0, "step": 15727 }, { "epoch": 2.000763261671543, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.013129711151123, "learning_rate": 1e-05, "loss": 0.4268, "mean_token_accuracy": 0.8867338299751282, "num_tokens": 597122188.0, "step": 15728 }, { "epoch": 2.0008904719501337, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.0001201629638671875, "grad_norm": 7.484764099121094, "learning_rate": 1e-05, "loss": 0.5094, "mean_token_accuracy": 0.8749066591262817, "num_tokens": 597164835.0, "step": 15729 }, { "epoch": 2.0010176822287242, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.320323944091797, "learning_rate": 1e-05, "loss": 0.4558, "mean_token_accuracy": 0.8832019567489624, "num_tokens": 597201266.0, "step": 15730 }, { "epoch": 2.0011448925073148, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00011491775512695312, "grad_norm": 7.282039642333984, "learning_rate": 1e-05, "loss": 0.4752, "mean_token_accuracy": 0.8820981979370117, "num_tokens": 597244095.0, "step": 15731 }, { "epoch": 2.0012721027859053, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.460658073425293, "learning_rate": 1e-05, "loss": 0.4748, "mean_token_accuracy": 0.8771325945854187, "num_tokens": 597285462.0, "step": 15732 }, { "epoch": 2.001399313064496, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.9505934715271, "learning_rate": 1e-05, "loss": 0.4646, "mean_token_accuracy": 0.8836065530776978, "num_tokens": 597323436.0, "step": 15733 }, { "epoch": 2.0015265233430863, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.439534664154053, "learning_rate": 1e-05, "loss": 0.4915, "mean_token_accuracy": 0.8721516132354736, "num_tokens": 597361775.0, "step": 15734 }, { "epoch": 2.0016537336216764, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.644208908081055, "learning_rate": 1e-05, "loss": 0.5068, "mean_token_accuracy": 0.8681479692459106, "num_tokens": 597402772.0, "step": 15735 }, { "epoch": 2.001780943900267, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.337716102600098, "learning_rate": 1e-05, "loss": 0.4555, "mean_token_accuracy": 0.8840843439102173, "num_tokens": 597444245.0, "step": 15736 }, { "epoch": 2.0019081541788575, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.475996971130371, "learning_rate": 1e-05, "loss": 0.4657, "mean_token_accuracy": 0.8793704509735107, "num_tokens": 597481742.0, "step": 15737 }, { "epoch": 2.002035364457448, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.213262557983398, "learning_rate": 1e-05, "loss": 0.5028, "mean_token_accuracy": 0.8684579133987427, "num_tokens": 597520339.0, "step": 15738 }, { "epoch": 2.0021625747360385, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.4336042404174805, "learning_rate": 1e-05, "loss": 0.4442, "mean_token_accuracy": 0.885709285736084, "num_tokens": 597552756.0, "step": 15739 }, { "epoch": 2.002289785014629, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 6.062907695770264, "learning_rate": 1e-05, "loss": 0.4311, "mean_token_accuracy": 0.8872283697128296, "num_tokens": 597582224.0, "step": 15740 }, { "epoch": 2.0024169952932196, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.490992069244385, "learning_rate": 1e-05, "loss": 0.4686, "mean_token_accuracy": 0.878810703754425, "num_tokens": 597619121.0, "step": 15741 }, { "epoch": 2.00254420557181, "ewc_loss": 0.10986328125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.914342880249023, "learning_rate": 1e-05, "loss": 0.4961, "mean_token_accuracy": 0.8687319159507751, "num_tokens": 597654074.0, "step": 15742 }, { "epoch": 2.0026714158504006, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.6355462074279785, "learning_rate": 1e-05, "loss": 0.4657, "mean_token_accuracy": 0.8812166452407837, "num_tokens": 597695873.0, "step": 15743 }, { "epoch": 2.002798626128991, "ewc_loss": 0.10693359375, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.7621684074401855, "learning_rate": 1e-05, "loss": 0.4392, "mean_token_accuracy": 0.8819870948791504, "num_tokens": 597733072.0, "step": 15744 }, { "epoch": 2.0029258364075817, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.9210591316223145, "learning_rate": 1e-05, "loss": 0.4141, "mean_token_accuracy": 0.8999003171920776, "num_tokens": 597772777.0, "step": 15745 }, { "epoch": 2.0030530466861722, "ewc_loss": 0.10595703125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.693543910980225, "learning_rate": 1e-05, "loss": 0.4803, "mean_token_accuracy": 0.8686158657073975, "num_tokens": 597809276.0, "step": 15746 }, { "epoch": 2.0031802569647628, "ewc_loss": 0.140625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00011682510375976562, "grad_norm": 7.286266326904297, "learning_rate": 1e-05, "loss": 0.5031, "mean_token_accuracy": 0.8716626167297363, "num_tokens": 597848956.0, "step": 15747 }, { "epoch": 2.0033074672433533, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 5.831340312957764, "learning_rate": 1e-05, "loss": 0.4514, "mean_token_accuracy": 0.8822349309921265, "num_tokens": 597890664.0, "step": 15748 }, { "epoch": 2.003434677521944, "ewc_loss": 0.1474609375, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.550561904907227, "learning_rate": 1e-05, "loss": 0.4985, "mean_token_accuracy": 0.8772990703582764, "num_tokens": 597931017.0, "step": 15749 }, { "epoch": 2.0035618878005343, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.236540794372559, "learning_rate": 1e-05, "loss": 0.4316, "mean_token_accuracy": 0.8910433053970337, "num_tokens": 597971552.0, "step": 15750 }, { "epoch": 2.003689098079125, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00011873245239257812, "grad_norm": 7.375426769256592, "learning_rate": 1e-05, "loss": 0.5117, "mean_token_accuracy": 0.869156539440155, "num_tokens": 598011554.0, "step": 15751 }, { "epoch": 2.0038163083577154, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.534646034240723, "learning_rate": 1e-05, "loss": 0.4761, "mean_token_accuracy": 0.877720832824707, "num_tokens": 598050633.0, "step": 15752 }, { "epoch": 2.003943518636306, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.8907270431518555, "learning_rate": 1e-05, "loss": 0.4522, "mean_token_accuracy": 0.8854020833969116, "num_tokens": 598088022.0, "step": 15753 }, { "epoch": 2.0040707289148965, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.6014885902404785, "learning_rate": 1e-05, "loss": 0.4835, "mean_token_accuracy": 0.8764801621437073, "num_tokens": 598126333.0, "step": 15754 }, { "epoch": 2.004197939193487, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.380075931549072, "learning_rate": 1e-05, "loss": 0.4706, "mean_token_accuracy": 0.8751841187477112, "num_tokens": 598161725.0, "step": 15755 }, { "epoch": 2.0043251494720775, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.669130325317383, "learning_rate": 1e-05, "loss": 0.4682, "mean_token_accuracy": 0.8825607895851135, "num_tokens": 598203158.0, "step": 15756 }, { "epoch": 2.004452359750668, "ewc_loss": 0.11083984375, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.9834113121032715, "learning_rate": 1e-05, "loss": 0.4481, "mean_token_accuracy": 0.8821545839309692, "num_tokens": 598241093.0, "step": 15757 }, { "epoch": 2.0045795700292586, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.870247840881348, "learning_rate": 1e-05, "loss": 0.4859, "mean_token_accuracy": 0.8782654404640198, "num_tokens": 598276389.0, "step": 15758 }, { "epoch": 2.0047067803078487, "ewc_loss": 0.10693359375, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.7649335861206055, "learning_rate": 1e-05, "loss": 0.4467, "mean_token_accuracy": 0.8803131580352783, "num_tokens": 598316645.0, "step": 15759 }, { "epoch": 2.004833990586439, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.0001125335693359375, "grad_norm": 7.180842399597168, "learning_rate": 1e-05, "loss": 0.4992, "mean_token_accuracy": 0.8753461837768555, "num_tokens": 598361443.0, "step": 15760 }, { "epoch": 2.0049612008650297, "ewc_loss": 0.109375, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.780531883239746, "learning_rate": 1e-05, "loss": 0.4565, "mean_token_accuracy": 0.8795939683914185, "num_tokens": 598402934.0, "step": 15761 }, { "epoch": 2.0050884111436202, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00012063980102539062, "grad_norm": 7.469526290893555, "learning_rate": 1e-05, "loss": 0.5123, "mean_token_accuracy": 0.8726032376289368, "num_tokens": 598436315.0, "step": 15762 }, { "epoch": 2.0052156214222108, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.071250915527344, "learning_rate": 1e-05, "loss": 0.4634, "mean_token_accuracy": 0.8856271505355835, "num_tokens": 598476562.0, "step": 15763 }, { "epoch": 2.0053428317008013, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00012159347534179688, "grad_norm": 7.475547790527344, "learning_rate": 1e-05, "loss": 0.4542, "mean_token_accuracy": 0.8916362524032593, "num_tokens": 598515180.0, "step": 15764 }, { "epoch": 2.005470041979392, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.371273994445801, "learning_rate": 1e-05, "loss": 0.4413, "mean_token_accuracy": 0.8888589143753052, "num_tokens": 598554177.0, "step": 15765 }, { "epoch": 2.0055972522579824, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00011539459228515625, "grad_norm": 7.21598482131958, "learning_rate": 1e-05, "loss": 0.5065, "mean_token_accuracy": 0.8754084706306458, "num_tokens": 598589091.0, "step": 15766 }, { "epoch": 2.005724462536573, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.4664788246154785, "learning_rate": 1e-05, "loss": 0.4571, "mean_token_accuracy": 0.8839641809463501, "num_tokens": 598624570.0, "step": 15767 }, { "epoch": 2.0058516728151634, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.87913703918457, "learning_rate": 1e-05, "loss": 0.41, "mean_token_accuracy": 0.8987226486206055, "num_tokens": 598660350.0, "step": 15768 }, { "epoch": 2.005978883093754, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.391577243804932, "learning_rate": 1e-05, "loss": 0.5129, "mean_token_accuracy": 0.8666146993637085, "num_tokens": 598701823.0, "step": 15769 }, { "epoch": 2.0061060933723445, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.630649089813232, "learning_rate": 1e-05, "loss": 0.4489, "mean_token_accuracy": 0.8857694864273071, "num_tokens": 598738637.0, "step": 15770 }, { "epoch": 2.006233303650935, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.289217472076416, "learning_rate": 1e-05, "loss": 0.4719, "mean_token_accuracy": 0.8776534795761108, "num_tokens": 598773459.0, "step": 15771 }, { "epoch": 2.0063605139295255, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.509596347808838, "learning_rate": 1e-05, "loss": 0.4491, "mean_token_accuracy": 0.8859090805053711, "num_tokens": 598810098.0, "step": 15772 }, { "epoch": 2.006487724208116, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.140303134918213, "learning_rate": 1e-05, "loss": 0.4797, "mean_token_accuracy": 0.8705434799194336, "num_tokens": 598850701.0, "step": 15773 }, { "epoch": 2.0066149344867066, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.509432315826416, "learning_rate": 1e-05, "loss": 0.4207, "mean_token_accuracy": 0.8947674632072449, "num_tokens": 598887745.0, "step": 15774 }, { "epoch": 2.006742144765297, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.981897830963135, "learning_rate": 1e-05, "loss": 0.4485, "mean_token_accuracy": 0.8820363879203796, "num_tokens": 598928360.0, "step": 15775 }, { "epoch": 2.0068693550438876, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.645659446716309, "learning_rate": 1e-05, "loss": 0.47, "mean_token_accuracy": 0.8794454336166382, "num_tokens": 598965628.0, "step": 15776 }, { "epoch": 2.006996565322478, "ewc_loss": 0.109375, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.832772254943848, "learning_rate": 1e-05, "loss": 0.4666, "mean_token_accuracy": 0.8762553930282593, "num_tokens": 599008445.0, "step": 15777 }, { "epoch": 2.0071237756010687, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.874806880950928, "learning_rate": 1e-05, "loss": 0.4319, "mean_token_accuracy": 0.8958910703659058, "num_tokens": 599041730.0, "step": 15778 }, { "epoch": 2.007250985879659, "ewc_loss": 0.10888671875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.736478805541992, "learning_rate": 1e-05, "loss": 0.4504, "mean_token_accuracy": 0.8815947771072388, "num_tokens": 599082481.0, "step": 15779 }, { "epoch": 2.0073781961582498, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00011444091796875, "grad_norm": 7.211907386779785, "learning_rate": 1e-05, "loss": 0.4821, "mean_token_accuracy": 0.8781262636184692, "num_tokens": 599122503.0, "step": 15780 }, { "epoch": 2.0075054064368403, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 5.8344502449035645, "learning_rate": 1e-05, "loss": 0.488, "mean_token_accuracy": 0.866798996925354, "num_tokens": 599158558.0, "step": 15781 }, { "epoch": 2.007632616715431, "ewc_loss": 0.1455078125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00012111663818359375, "grad_norm": 7.443897247314453, "learning_rate": 1e-05, "loss": 0.4803, "mean_token_accuracy": 0.8840488195419312, "num_tokens": 599196769.0, "step": 15782 }, { "epoch": 2.0077598269940213, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.163970470428467, "learning_rate": 1e-05, "loss": 0.4572, "mean_token_accuracy": 0.881456732749939, "num_tokens": 599231050.0, "step": 15783 }, { "epoch": 2.0078870372726114, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00011730194091796875, "grad_norm": 7.315747261047363, "learning_rate": 1e-05, "loss": 0.4632, "mean_token_accuracy": 0.8867465257644653, "num_tokens": 599271826.0, "step": 15784 }, { "epoch": 2.008014247551202, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.461875915527344, "learning_rate": 1e-05, "loss": 0.5065, "mean_token_accuracy": 0.8728695511817932, "num_tokens": 599311134.0, "step": 15785 }, { "epoch": 2.0081414578297925, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.871126174926758, "learning_rate": 1e-05, "loss": 0.4782, "mean_token_accuracy": 0.8816300630569458, "num_tokens": 599349240.0, "step": 15786 }, { "epoch": 2.008268668108383, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.565788269042969, "learning_rate": 1e-05, "loss": 0.471, "mean_token_accuracy": 0.8809987306594849, "num_tokens": 599381624.0, "step": 15787 }, { "epoch": 2.0083958783869735, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.348926067352295, "learning_rate": 1e-05, "loss": 0.4526, "mean_token_accuracy": 0.8809834718704224, "num_tokens": 599421524.0, "step": 15788 }, { "epoch": 2.008523088665564, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.680211544036865, "learning_rate": 1e-05, "loss": 0.4902, "mean_token_accuracy": 0.871613621711731, "num_tokens": 599459285.0, "step": 15789 }, { "epoch": 2.0086502989441546, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.9220709800720215, "learning_rate": 1e-05, "loss": 0.4101, "mean_token_accuracy": 0.8930139541625977, "num_tokens": 599501682.0, "step": 15790 }, { "epoch": 2.008777509222745, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.935352325439453, "learning_rate": 1e-05, "loss": 0.4484, "mean_token_accuracy": 0.8892560005187988, "num_tokens": 599538518.0, "step": 15791 }, { "epoch": 2.0089047195013356, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.70738410949707, "learning_rate": 1e-05, "loss": 0.4431, "mean_token_accuracy": 0.8876164555549622, "num_tokens": 599572802.0, "step": 15792 }, { "epoch": 2.009031929779926, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 0.00011730194091796875, "grad_norm": 7.315832138061523, "learning_rate": 1e-05, "loss": 0.4687, "mean_token_accuracy": 0.8849731683731079, "num_tokens": 599614366.0, "step": 15793 }, { "epoch": 2.0091591400585167, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 5.790760517120361, "learning_rate": 1e-05, "loss": 0.4622, "mean_token_accuracy": 0.8774400949478149, "num_tokens": 599654814.0, "step": 15794 }, { "epoch": 2.0092863503371072, "ewc_loss": 0.1494140625, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 0.00012493133544921875, "grad_norm": 7.632327079772949, "learning_rate": 1e-05, "loss": 0.4914, "mean_token_accuracy": 0.8802895545959473, "num_tokens": 599689678.0, "step": 15795 }, { "epoch": 2.0094135606156978, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.204586982727051, "learning_rate": 1e-05, "loss": 0.5009, "mean_token_accuracy": 0.8740834593772888, "num_tokens": 599725932.0, "step": 15796 }, { "epoch": 2.0095407708942883, "ewc_loss": 0.1474609375, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.550569534301758, "learning_rate": 1e-05, "loss": 0.4624, "mean_token_accuracy": 0.890927791595459, "num_tokens": 599762041.0, "step": 15797 }, { "epoch": 2.009667981172879, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.525299549102783, "learning_rate": 1e-05, "loss": 0.4615, "mean_token_accuracy": 0.883613109588623, "num_tokens": 599800706.0, "step": 15798 }, { "epoch": 2.0097951914514693, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 0.00011396408081054688, "grad_norm": 7.210263729095459, "learning_rate": 1e-05, "loss": 0.5235, "mean_token_accuracy": 0.8644180297851562, "num_tokens": 599833483.0, "step": 15799 }, { "epoch": 2.00992240173006, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.550917625427246, "learning_rate": 1e-05, "loss": 0.4645, "mean_token_accuracy": 0.8800902366638184, "num_tokens": 599872172.0, "step": 15800 }, { "epoch": 2.0100496120086504, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.840816020965576, "learning_rate": 1e-05, "loss": 0.4977, "mean_token_accuracy": 0.8739519119262695, "num_tokens": 599914149.0, "step": 15801 }, { "epoch": 2.010176822287241, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.446742057800293, "learning_rate": 1e-05, "loss": 0.4614, "mean_token_accuracy": 0.8817223906517029, "num_tokens": 599950603.0, "step": 15802 }, { "epoch": 2.0103040325658315, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.597235679626465, "learning_rate": 1e-05, "loss": 0.491, "mean_token_accuracy": 0.8720101118087769, "num_tokens": 599991844.0, "step": 15803 }, { "epoch": 2.010431242844422, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.3010478019714355, "learning_rate": 1e-05, "loss": 0.4524, "mean_token_accuracy": 0.8844021558761597, "num_tokens": 600033303.0, "step": 15804 }, { "epoch": 2.0105584531230125, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.494022846221924, "learning_rate": 1e-05, "loss": 0.422, "mean_token_accuracy": 0.8953118324279785, "num_tokens": 600065633.0, "step": 15805 }, { "epoch": 2.010685663401603, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.138177871704102, "learning_rate": 1e-05, "loss": 0.457, "mean_token_accuracy": 0.8793110847473145, "num_tokens": 600104333.0, "step": 15806 }, { "epoch": 2.0108128736801936, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.517524719238281, "learning_rate": 1e-05, "loss": 0.4876, "mean_token_accuracy": 0.8723341226577759, "num_tokens": 600141329.0, "step": 15807 }, { "epoch": 2.0109400839587837, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.974304676055908, "learning_rate": 1e-05, "loss": 0.4838, "mean_token_accuracy": 0.8695055842399597, "num_tokens": 600184171.0, "step": 15808 }, { "epoch": 2.011067294237374, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.6525750160217285, "learning_rate": 1e-05, "loss": 0.5209, "mean_token_accuracy": 0.8640871644020081, "num_tokens": 600220249.0, "step": 15809 }, { "epoch": 2.0111945045159647, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.816893577575684, "learning_rate": 1e-05, "loss": 0.4841, "mean_token_accuracy": 0.8747695684432983, "num_tokens": 600256015.0, "step": 15810 }, { "epoch": 2.0113217147945552, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.9167656898498535, "learning_rate": 1e-05, "loss": 0.4668, "mean_token_accuracy": 0.879708468914032, "num_tokens": 600282893.0, "step": 15811 }, { "epoch": 2.0114489250731458, "ewc_loss": 0.109375, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.730493068695068, "learning_rate": 1e-05, "loss": 0.5224, "mean_token_accuracy": 0.8566133975982666, "num_tokens": 600320940.0, "step": 15812 }, { "epoch": 2.0115761353517363, "ewc_loss": 0.140625, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 0.0001163482666015625, "grad_norm": 7.246758937835693, "learning_rate": 1e-05, "loss": 0.5468, "mean_token_accuracy": 0.8615573644638062, "num_tokens": 600362054.0, "step": 15813 }, { "epoch": 2.011703345630327, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 5.865556716918945, "learning_rate": 1e-05, "loss": 0.5088, "mean_token_accuracy": 0.8639698624610901, "num_tokens": 600395080.0, "step": 15814 }, { "epoch": 2.0118305559089174, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 0.0001220703125, "grad_norm": 7.463733673095703, "learning_rate": 1e-05, "loss": 0.468, "mean_token_accuracy": 0.8868660926818848, "num_tokens": 600428060.0, "step": 15815 }, { "epoch": 2.011957766187508, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.197021961212158, "learning_rate": 1e-05, "loss": 0.4534, "mean_token_accuracy": 0.8843097686767578, "num_tokens": 600469833.0, "step": 15816 }, { "epoch": 2.0120849764660984, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 0.00011777877807617188, "grad_norm": 7.318617820739746, "learning_rate": 1e-05, "loss": 0.4701, "mean_token_accuracy": 0.8825660943984985, "num_tokens": 600498006.0, "step": 15817 }, { "epoch": 2.012212186744689, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.489189624786377, "learning_rate": 1e-05, "loss": 0.4648, "mean_token_accuracy": 0.8813716173171997, "num_tokens": 600534744.0, "step": 15818 }, { "epoch": 2.0123393970232795, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.852856636047363, "learning_rate": 1e-05, "loss": 0.4897, "mean_token_accuracy": 0.8782386779785156, "num_tokens": 600570282.0, "step": 15819 }, { "epoch": 2.01246660730187, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.585718631744385, "learning_rate": 1e-05, "loss": 0.448, "mean_token_accuracy": 0.8880264759063721, "num_tokens": 600603858.0, "step": 15820 }, { "epoch": 2.0125938175804605, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.331099987030029, "learning_rate": 1e-05, "loss": 0.5122, "mean_token_accuracy": 0.8636311888694763, "num_tokens": 600642078.0, "step": 15821 }, { "epoch": 2.012721027859051, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.70525598526001, "learning_rate": 1e-05, "loss": 0.4953, "mean_token_accuracy": 0.8729434609413147, "num_tokens": 600679424.0, "step": 15822 }, { "epoch": 2.0128482381376416, "ewc_loss": 0.109375, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.8890886306762695, "learning_rate": 1e-05, "loss": 0.43, "mean_token_accuracy": 0.8865052461624146, "num_tokens": 600716310.0, "step": 15823 }, { "epoch": 2.012975448416232, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 6.9820475578308105, "learning_rate": 1e-05, "loss": 0.5237, "mean_token_accuracy": 0.8663774728775024, "num_tokens": 600757031.0, "step": 15824 }, { "epoch": 2.0131026586948226, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.6859130859375, "learning_rate": 1e-05, "loss": 0.446, "mean_token_accuracy": 0.8787232637405396, "num_tokens": 600795414.0, "step": 15825 }, { "epoch": 2.013229868973413, "ewc_loss": 0.1455078125, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 0.00012063980102539062, "grad_norm": 7.394535064697266, "learning_rate": 1e-05, "loss": 0.5347, "mean_token_accuracy": 0.864657998085022, "num_tokens": 600829239.0, "step": 15826 }, { "epoch": 2.0133570792520037, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 5.827759742736816, "learning_rate": 1e-05, "loss": 0.4274, "mean_token_accuracy": 0.8887379765510559, "num_tokens": 600868155.0, "step": 15827 }, { "epoch": 2.013484289530594, "ewc_loss": 0.1513671875, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 0.00012683868408203125, "grad_norm": 7.675924301147461, "learning_rate": 1e-05, "loss": 0.5099, "mean_token_accuracy": 0.8724258542060852, "num_tokens": 600906797.0, "step": 15828 }, { "epoch": 2.0136114998091847, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.289597988128662, "learning_rate": 1e-05, "loss": 0.4952, "mean_token_accuracy": 0.869636058807373, "num_tokens": 600948399.0, "step": 15829 }, { "epoch": 2.0137387100877753, "ewc_loss": 0.1474609375, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.517770290374756, "learning_rate": 1e-05, "loss": 0.5041, "mean_token_accuracy": 0.8755826354026794, "num_tokens": 600989682.0, "step": 15830 }, { "epoch": 2.013865920366366, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 6.581741809844971, "learning_rate": 1e-05, "loss": 0.4867, "mean_token_accuracy": 0.8751633167266846, "num_tokens": 601025387.0, "step": 15831 }, { "epoch": 2.0139931306449563, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 0.00011157989501953125, "grad_norm": 7.128656387329102, "learning_rate": 1e-05, "loss": 0.4472, "mean_token_accuracy": 0.890229344367981, "num_tokens": 601061124.0, "step": 15832 }, { "epoch": 2.0141203409235464, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.551787376403809, "learning_rate": 1e-05, "loss": 0.4954, "mean_token_accuracy": 0.8739280700683594, "num_tokens": 601105789.0, "step": 15833 }, { "epoch": 2.014247551202137, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.790365695953369, "learning_rate": 1e-05, "loss": 0.5167, "mean_token_accuracy": 0.8633474111557007, "num_tokens": 601138697.0, "step": 15834 }, { "epoch": 2.0143747614807275, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.427127361297607, "learning_rate": 1e-05, "loss": 0.4365, "mean_token_accuracy": 0.8900368809700012, "num_tokens": 601172875.0, "step": 15835 }, { "epoch": 2.014501971759318, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.583442211151123, "learning_rate": 1e-05, "loss": 0.4327, "mean_token_accuracy": 0.8919083476066589, "num_tokens": 601206303.0, "step": 15836 }, { "epoch": 2.0146291820379085, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.260989189147949, "learning_rate": 1e-05, "loss": 0.4168, "mean_token_accuracy": 0.8958324193954468, "num_tokens": 601243879.0, "step": 15837 }, { "epoch": 2.014756392316499, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.518793106079102, "learning_rate": 1e-05, "loss": 0.4673, "mean_token_accuracy": 0.8804474472999573, "num_tokens": 601282639.0, "step": 15838 }, { "epoch": 2.0148836025950896, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.084538459777832, "learning_rate": 1e-05, "loss": 0.4744, "mean_token_accuracy": 0.8754804730415344, "num_tokens": 601326204.0, "step": 15839 }, { "epoch": 2.01501081287368, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.583148956298828, "learning_rate": 1e-05, "loss": 0.4484, "mean_token_accuracy": 0.8847035765647888, "num_tokens": 601358936.0, "step": 15840 }, { "epoch": 2.0151380231522706, "ewc_loss": 0.109375, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.917293548583984, "learning_rate": 1e-05, "loss": 0.4152, "mean_token_accuracy": 0.8910094499588013, "num_tokens": 601398253.0, "step": 15841 }, { "epoch": 2.015265233430861, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.759899139404297, "learning_rate": 1e-05, "loss": 0.392, "mean_token_accuracy": 0.9046772718429565, "num_tokens": 601434806.0, "step": 15842 }, { "epoch": 2.0153924437094517, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.788146495819092, "learning_rate": 1e-05, "loss": 0.4284, "mean_token_accuracy": 0.8836237788200378, "num_tokens": 601471834.0, "step": 15843 }, { "epoch": 2.0155196539880422, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 7.064493656158447, "learning_rate": 1e-05, "loss": 0.4525, "mean_token_accuracy": 0.8890471458435059, "num_tokens": 601511256.0, "step": 15844 }, { "epoch": 2.0156468642666328, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.762843608856201, "learning_rate": 1e-05, "loss": 0.4158, "mean_token_accuracy": 0.8904374837875366, "num_tokens": 601547434.0, "step": 15845 }, { "epoch": 2.0157740745452233, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 0.000118255615234375, "grad_norm": 7.397704601287842, "learning_rate": 1e-05, "loss": 0.4612, "mean_token_accuracy": 0.8858516216278076, "num_tokens": 601581375.0, "step": 15846 }, { "epoch": 2.015901284823814, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.004831790924072, "learning_rate": 1e-05, "loss": 0.48, "mean_token_accuracy": 0.8745303153991699, "num_tokens": 601615067.0, "step": 15847 }, { "epoch": 2.0160284951024043, "ewc_loss": 0.1455078125, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 0.00012111663818359375, "grad_norm": 7.479458332061768, "learning_rate": 1e-05, "loss": 0.4265, "mean_token_accuracy": 0.8974286317825317, "num_tokens": 601654168.0, "step": 15848 }, { "epoch": 2.016155705380995, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.371970176696777, "learning_rate": 1e-05, "loss": 0.437, "mean_token_accuracy": 0.8921063542366028, "num_tokens": 601687113.0, "step": 15849 }, { "epoch": 2.0162829156595854, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 0.00011348724365234375, "grad_norm": 7.165534973144531, "learning_rate": 1e-05, "loss": 0.499, "mean_token_accuracy": 0.871995210647583, "num_tokens": 601723993.0, "step": 15850 }, { "epoch": 2.016410125938176, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.568350315093994, "learning_rate": 1e-05, "loss": 0.5321, "mean_token_accuracy": 0.8612029552459717, "num_tokens": 601758522.0, "step": 15851 }, { "epoch": 2.0165373362167665, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.642462730407715, "learning_rate": 1e-05, "loss": 0.4059, "mean_token_accuracy": 0.9021328091621399, "num_tokens": 601794221.0, "step": 15852 }, { "epoch": 2.016664546495357, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.621222019195557, "learning_rate": 1e-05, "loss": 0.451, "mean_token_accuracy": 0.8879404067993164, "num_tokens": 601830430.0, "step": 15853 }, { "epoch": 2.0167917567739475, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.1523118019104, "learning_rate": 1e-05, "loss": 0.452, "mean_token_accuracy": 0.8816066384315491, "num_tokens": 601871834.0, "step": 15854 }, { "epoch": 2.016918967052538, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.770854473114014, "learning_rate": 1e-05, "loss": 0.4852, "mean_token_accuracy": 0.8755049705505371, "num_tokens": 601911432.0, "step": 15855 }, { "epoch": 2.0170461773311286, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.804931163787842, "learning_rate": 1e-05, "loss": 0.4224, "mean_token_accuracy": 0.8892391920089722, "num_tokens": 601955592.0, "step": 15856 }, { "epoch": 2.0171733876097186, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 0.00011014938354492188, "grad_norm": 7.091108798980713, "learning_rate": 1e-05, "loss": 0.4989, "mean_token_accuracy": 0.8729106783866882, "num_tokens": 601996843.0, "step": 15857 }, { "epoch": 2.017300597888309, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.707128047943115, "learning_rate": 1e-05, "loss": 0.4292, "mean_token_accuracy": 0.8872679471969604, "num_tokens": 602034963.0, "step": 15858 }, { "epoch": 2.0174278081668997, "ewc_loss": 0.1455078125, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 0.00012063980102539062, "grad_norm": 7.4874653816223145, "learning_rate": 1e-05, "loss": 0.5154, "mean_token_accuracy": 0.878154993057251, "num_tokens": 602073460.0, "step": 15859 }, { "epoch": 2.0175550184454902, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 5.953211307525635, "learning_rate": 1e-05, "loss": 0.5088, "mean_token_accuracy": 0.8652902841567993, "num_tokens": 602117442.0, "step": 15860 }, { "epoch": 2.0176822287240808, "ewc_loss": 0.1494140625, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 0.00012493133544921875, "grad_norm": 7.662481307983398, "learning_rate": 1e-05, "loss": 0.4731, "mean_token_accuracy": 0.8845590353012085, "num_tokens": 602161253.0, "step": 15861 }, { "epoch": 2.0178094390026713, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.385832786560059, "learning_rate": 1e-05, "loss": 0.4936, "mean_token_accuracy": 0.8705469965934753, "num_tokens": 602204324.0, "step": 15862 }, { "epoch": 2.017936649281262, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 0.00011920928955078125, "grad_norm": 7.420237064361572, "learning_rate": 1e-05, "loss": 0.4675, "mean_token_accuracy": 0.882948637008667, "num_tokens": 602236335.0, "step": 15863 }, { "epoch": 2.0180638595598523, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 6.60100793838501, "learning_rate": 1e-05, "loss": 0.5036, "mean_token_accuracy": 0.8732931017875671, "num_tokens": 602271302.0, "step": 15864 }, { "epoch": 2.018191069838443, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 7.020205974578857, "learning_rate": 1e-05, "loss": 0.4522, "mean_token_accuracy": 0.8862346410751343, "num_tokens": 602304103.0, "step": 15865 }, { "epoch": 2.0183182801170334, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.532421112060547, "learning_rate": 1e-05, "loss": 0.4676, "mean_token_accuracy": 0.8809282183647156, "num_tokens": 602343293.0, "step": 15866 }, { "epoch": 2.018445490395624, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.694919586181641, "learning_rate": 1e-05, "loss": 0.4848, "mean_token_accuracy": 0.872741162776947, "num_tokens": 602383650.0, "step": 15867 }, { "epoch": 2.0185727006742145, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.410610675811768, "learning_rate": 1e-05, "loss": 0.4446, "mean_token_accuracy": 0.8889850974082947, "num_tokens": 602423436.0, "step": 15868 }, { "epoch": 2.018699910952805, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.497876167297363, "learning_rate": 1e-05, "loss": 0.4628, "mean_token_accuracy": 0.880582869052887, "num_tokens": 602460909.0, "step": 15869 }, { "epoch": 2.0188271212313955, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.261441230773926, "learning_rate": 1e-05, "loss": 0.472, "mean_token_accuracy": 0.8756889700889587, "num_tokens": 602499610.0, "step": 15870 }, { "epoch": 2.018954331509986, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.435188293457031, "learning_rate": 1e-05, "loss": 0.4822, "mean_token_accuracy": 0.8750799298286438, "num_tokens": 602540783.0, "step": 15871 }, { "epoch": 2.0190815417885766, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 6.091630458831787, "learning_rate": 1e-05, "loss": 0.47, "mean_token_accuracy": 0.8728322982788086, "num_tokens": 602579205.0, "step": 15872 }, { "epoch": 2.019208752067167, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.512685775756836, "learning_rate": 1e-05, "loss": 0.4696, "mean_token_accuracy": 0.8787279725074768, "num_tokens": 602613720.0, "step": 15873 }, { "epoch": 2.0193359623457576, "ewc_loss": 0.109375, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.923067569732666, "learning_rate": 1e-05, "loss": 0.4429, "mean_token_accuracy": 0.8844881653785706, "num_tokens": 602651954.0, "step": 15874 }, { "epoch": 2.019463172624348, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.682943344116211, "learning_rate": 1e-05, "loss": 0.4392, "mean_token_accuracy": 0.8911224603652954, "num_tokens": 602685553.0, "step": 15875 }, { "epoch": 2.0195903829029387, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.764501571655273, "learning_rate": 1e-05, "loss": 0.4801, "mean_token_accuracy": 0.8735084533691406, "num_tokens": 602727854.0, "step": 15876 }, { "epoch": 2.019717593181529, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.987969875335693, "learning_rate": 1e-05, "loss": 0.4918, "mean_token_accuracy": 0.8809689879417419, "num_tokens": 602765993.0, "step": 15877 }, { "epoch": 2.0198448034601197, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.692113876342773, "learning_rate": 1e-05, "loss": 0.4174, "mean_token_accuracy": 0.8901512622833252, "num_tokens": 602807450.0, "step": 15878 }, { "epoch": 2.0199720137387103, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 0.00011730194091796875, "grad_norm": 7.3920769691467285, "learning_rate": 1e-05, "loss": 0.5334, "mean_token_accuracy": 0.861549973487854, "num_tokens": 602848689.0, "step": 15879 }, { "epoch": 2.020099224017301, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 5.884509086608887, "learning_rate": 1e-05, "loss": 0.4349, "mean_token_accuracy": 0.8921322226524353, "num_tokens": 602888010.0, "step": 15880 }, { "epoch": 2.0202264342958913, "ewc_loss": 0.1494140625, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 0.00012493133544921875, "grad_norm": 7.630034923553467, "learning_rate": 1e-05, "loss": 0.5526, "mean_token_accuracy": 0.8631013631820679, "num_tokens": 602923345.0, "step": 15881 }, { "epoch": 2.0203536445744814, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.339937686920166, "learning_rate": 1e-05, "loss": 0.464, "mean_token_accuracy": 0.8821781873703003, "num_tokens": 602967320.0, "step": 15882 }, { "epoch": 2.020480854853072, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 0.00011777877807617188, "grad_norm": 7.366803169250488, "learning_rate": 1e-05, "loss": 0.4439, "mean_token_accuracy": 0.8903362154960632, "num_tokens": 603005177.0, "step": 15883 }, { "epoch": 2.0206080651316625, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 6.612900733947754, "learning_rate": 1e-05, "loss": 0.4412, "mean_token_accuracy": 0.8921282291412354, "num_tokens": 603039562.0, "step": 15884 }, { "epoch": 2.020735275410253, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.85844612121582, "learning_rate": 1e-05, "loss": 0.463, "mean_token_accuracy": 0.8861485719680786, "num_tokens": 603079075.0, "step": 15885 }, { "epoch": 2.0208624856888435, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.6383514404296875, "learning_rate": 1e-05, "loss": 0.4738, "mean_token_accuracy": 0.8798708915710449, "num_tokens": 603118077.0, "step": 15886 }, { "epoch": 2.020989695967434, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.362686634063721, "learning_rate": 1e-05, "loss": 0.4867, "mean_token_accuracy": 0.871150553226471, "num_tokens": 603156921.0, "step": 15887 }, { "epoch": 2.0211169062460246, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.686862468719482, "learning_rate": 1e-05, "loss": 0.4661, "mean_token_accuracy": 0.8806897401809692, "num_tokens": 603202062.0, "step": 15888 }, { "epoch": 2.021244116524615, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 5.997070789337158, "learning_rate": 1e-05, "loss": 0.4409, "mean_token_accuracy": 0.8856585025787354, "num_tokens": 603236240.0, "step": 15889 }, { "epoch": 2.0213713268032056, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.853207588195801, "learning_rate": 1e-05, "loss": 0.4903, "mean_token_accuracy": 0.8732197880744934, "num_tokens": 603277376.0, "step": 15890 }, { "epoch": 2.021498537081796, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.803369522094727, "learning_rate": 1e-05, "loss": 0.4554, "mean_token_accuracy": 0.8770197629928589, "num_tokens": 603314740.0, "step": 15891 }, { "epoch": 2.0216257473603867, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 7.120286464691162, "learning_rate": 1e-05, "loss": 0.4773, "mean_token_accuracy": 0.8814513683319092, "num_tokens": 603354992.0, "step": 15892 }, { "epoch": 2.021752957638977, "ewc_loss": 0.109375, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.795783042907715, "learning_rate": 1e-05, "loss": 0.4635, "mean_token_accuracy": 0.875978946685791, "num_tokens": 603387271.0, "step": 15893 }, { "epoch": 2.0218801679175677, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 0.00011730194091796875, "grad_norm": 7.3873114585876465, "learning_rate": 1e-05, "loss": 0.463, "mean_token_accuracy": 0.8876603841781616, "num_tokens": 603427092.0, "step": 15894 }, { "epoch": 2.0220073781961583, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.004927158355713, "learning_rate": 1e-05, "loss": 0.4871, "mean_token_accuracy": 0.8696309328079224, "num_tokens": 603465188.0, "step": 15895 }, { "epoch": 2.022134588474749, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 0.0001201629638671875, "grad_norm": 7.4695940017700195, "learning_rate": 1e-05, "loss": 0.491, "mean_token_accuracy": 0.8767383694648743, "num_tokens": 603499290.0, "step": 15896 }, { "epoch": 2.0222617987533393, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.297633647918701, "learning_rate": 1e-05, "loss": 0.4698, "mean_token_accuracy": 0.8839055895805359, "num_tokens": 603530594.0, "step": 15897 }, { "epoch": 2.02238900903193, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 0.00011539459228515625, "grad_norm": 7.274710178375244, "learning_rate": 1e-05, "loss": 0.4735, "mean_token_accuracy": 0.8820273280143738, "num_tokens": 603569337.0, "step": 15898 }, { "epoch": 2.0225162193105204, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.428312301635742, "learning_rate": 1e-05, "loss": 0.46, "mean_token_accuracy": 0.8825557231903076, "num_tokens": 603609704.0, "step": 15899 }, { "epoch": 2.022643429589111, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 6.954684257507324, "learning_rate": 1e-05, "loss": 0.479, "mean_token_accuracy": 0.8788149356842041, "num_tokens": 603647623.0, "step": 15900 }, { "epoch": 2.0227706398677014, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.412346363067627, "learning_rate": 1e-05, "loss": 0.4596, "mean_token_accuracy": 0.8834834694862366, "num_tokens": 603682237.0, "step": 15901 }, { "epoch": 2.022897850146292, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.668924331665039, "learning_rate": 1e-05, "loss": 0.4823, "mean_token_accuracy": 0.8734700679779053, "num_tokens": 603720348.0, "step": 15902 }, { "epoch": 2.0230250604248825, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.333212852478027, "learning_rate": 1e-05, "loss": 0.4596, "mean_token_accuracy": 0.8815460205078125, "num_tokens": 603759857.0, "step": 15903 }, { "epoch": 2.023152270703473, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.489733695983887, "learning_rate": 1e-05, "loss": 0.4782, "mean_token_accuracy": 0.8748390674591064, "num_tokens": 603796970.0, "step": 15904 }, { "epoch": 2.0232794809820636, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.219400405883789, "learning_rate": 1e-05, "loss": 0.4794, "mean_token_accuracy": 0.8770793676376343, "num_tokens": 603838385.0, "step": 15905 }, { "epoch": 2.0234066912606536, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.4287919998168945, "learning_rate": 1e-05, "loss": 0.4687, "mean_token_accuracy": 0.8787882328033447, "num_tokens": 603884262.0, "step": 15906 }, { "epoch": 2.023533901539244, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 6.11004638671875, "learning_rate": 1e-05, "loss": 0.4775, "mean_token_accuracy": 0.8740430474281311, "num_tokens": 603925013.0, "step": 15907 }, { "epoch": 2.0236611118178347, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.457934856414795, "learning_rate": 1e-05, "loss": 0.4906, "mean_token_accuracy": 0.871044397354126, "num_tokens": 603971173.0, "step": 15908 }, { "epoch": 2.0237883220964252, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.949073314666748, "learning_rate": 1e-05, "loss": 0.4776, "mean_token_accuracy": 0.8726019859313965, "num_tokens": 604009091.0, "step": 15909 }, { "epoch": 2.0239155323750158, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.608395576477051, "learning_rate": 1e-05, "loss": 0.4674, "mean_token_accuracy": 0.8780549764633179, "num_tokens": 604045163.0, "step": 15910 }, { "epoch": 2.0240427426536063, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.7747802734375, "learning_rate": 1e-05, "loss": 0.4334, "mean_token_accuracy": 0.8827821016311646, "num_tokens": 604091379.0, "step": 15911 }, { "epoch": 2.024169952932197, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.898233413696289, "learning_rate": 1e-05, "loss": 0.4842, "mean_token_accuracy": 0.8770731687545776, "num_tokens": 604132159.0, "step": 15912 }, { "epoch": 2.0242971632107873, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.674034118652344, "learning_rate": 1e-05, "loss": 0.4859, "mean_token_accuracy": 0.8701480031013489, "num_tokens": 604169689.0, "step": 15913 }, { "epoch": 2.024424373489378, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 0.00011730194091796875, "grad_norm": 7.291426658630371, "learning_rate": 1e-05, "loss": 0.4274, "mean_token_accuracy": 0.8979237079620361, "num_tokens": 604201913.0, "step": 15914 }, { "epoch": 2.0245515837679684, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 5.793325424194336, "learning_rate": 1e-05, "loss": 0.4948, "mean_token_accuracy": 0.8708301782608032, "num_tokens": 604238275.0, "step": 15915 }, { "epoch": 2.024678794046559, "ewc_loss": 0.1474609375, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.578061580657959, "learning_rate": 1e-05, "loss": 0.4638, "mean_token_accuracy": 0.8869450092315674, "num_tokens": 604279862.0, "step": 15916 }, { "epoch": 2.0248060043251495, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.195802688598633, "learning_rate": 1e-05, "loss": 0.4772, "mean_token_accuracy": 0.8749457597732544, "num_tokens": 604319091.0, "step": 15917 }, { "epoch": 2.02493321460374, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 0.00011968612670898438, "grad_norm": 7.5057806968688965, "learning_rate": 1e-05, "loss": 0.509, "mean_token_accuracy": 0.8724690675735474, "num_tokens": 604363614.0, "step": 15918 }, { "epoch": 2.0250604248823305, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 6.568594932556152, "learning_rate": 1e-05, "loss": 0.4781, "mean_token_accuracy": 0.8777209520339966, "num_tokens": 604394422.0, "step": 15919 }, { "epoch": 2.025187635160921, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 6.987752914428711, "learning_rate": 1e-05, "loss": 0.4964, "mean_token_accuracy": 0.8759406208992004, "num_tokens": 604432844.0, "step": 15920 }, { "epoch": 2.0253148454395116, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.6170573234558105, "learning_rate": 1e-05, "loss": 0.4752, "mean_token_accuracy": 0.8796284794807434, "num_tokens": 604472074.0, "step": 15921 }, { "epoch": 2.025442055718102, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.469506740570068, "learning_rate": 1e-05, "loss": 0.4101, "mean_token_accuracy": 0.8959048986434937, "num_tokens": 604507827.0, "step": 15922 }, { "epoch": 2.0255692659966926, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.677770137786865, "learning_rate": 1e-05, "loss": 0.4897, "mean_token_accuracy": 0.8742270469665527, "num_tokens": 604548600.0, "step": 15923 }, { "epoch": 2.025696476275283, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.067314147949219, "learning_rate": 1e-05, "loss": 0.5113, "mean_token_accuracy": 0.864579439163208, "num_tokens": 604582701.0, "step": 15924 }, { "epoch": 2.0258236865538737, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.837931156158447, "learning_rate": 1e-05, "loss": 0.4938, "mean_token_accuracy": 0.8702184557914734, "num_tokens": 604624089.0, "step": 15925 }, { "epoch": 2.025950896832464, "ewc_loss": 0.109375, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.823129653930664, "learning_rate": 1e-05, "loss": 0.4293, "mean_token_accuracy": 0.8884139060974121, "num_tokens": 604655764.0, "step": 15926 }, { "epoch": 2.0260781071110547, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 0.00011157989501953125, "grad_norm": 7.10690450668335, "learning_rate": 1e-05, "loss": 0.4507, "mean_token_accuracy": 0.8898887038230896, "num_tokens": 604688000.0, "step": 15927 }, { "epoch": 2.0262053173896453, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.8015007972717285, "learning_rate": 1e-05, "loss": 0.4821, "mean_token_accuracy": 0.8735657334327698, "num_tokens": 604724602.0, "step": 15928 }, { "epoch": 2.026332527668236, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 0.00011777877807617188, "grad_norm": 7.357414245605469, "learning_rate": 1e-05, "loss": 0.4764, "mean_token_accuracy": 0.8827592134475708, "num_tokens": 604761794.0, "step": 15929 }, { "epoch": 2.0264597379468263, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 5.992866516113281, "learning_rate": 1e-05, "loss": 0.4171, "mean_token_accuracy": 0.8937174081802368, "num_tokens": 604800239.0, "step": 15930 }, { "epoch": 2.0265869482254164, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 0.00011968612670898438, "grad_norm": 7.438476085662842, "learning_rate": 1e-05, "loss": 0.4989, "mean_token_accuracy": 0.8791744709014893, "num_tokens": 604837143.0, "step": 15931 }, { "epoch": 2.026714158504007, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.270053863525391, "learning_rate": 1e-05, "loss": 0.4785, "mean_token_accuracy": 0.8798651099205017, "num_tokens": 604874074.0, "step": 15932 }, { "epoch": 2.0268413687825975, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 0.00011539459228515625, "grad_norm": 7.266737937927246, "learning_rate": 1e-05, "loss": 0.5046, "mean_token_accuracy": 0.8732972145080566, "num_tokens": 604911639.0, "step": 15933 }, { "epoch": 2.026968579061188, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.408693790435791, "learning_rate": 1e-05, "loss": 0.442, "mean_token_accuracy": 0.8904401659965515, "num_tokens": 604944963.0, "step": 15934 }, { "epoch": 2.0270957893397785, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 6.974083423614502, "learning_rate": 1e-05, "loss": 0.4821, "mean_token_accuracy": 0.8794656991958618, "num_tokens": 604986048.0, "step": 15935 }, { "epoch": 2.027222999618369, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.393346309661865, "learning_rate": 1e-05, "loss": 0.4492, "mean_token_accuracy": 0.8851654529571533, "num_tokens": 605023995.0, "step": 15936 }, { "epoch": 2.0273502098969596, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.6953043937683105, "learning_rate": 1e-05, "loss": 0.4533, "mean_token_accuracy": 0.8861740231513977, "num_tokens": 605064612.0, "step": 15937 }, { "epoch": 2.02747742017555, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.310462951660156, "learning_rate": 1e-05, "loss": 0.4994, "mean_token_accuracy": 0.874255895614624, "num_tokens": 605099332.0, "step": 15938 }, { "epoch": 2.0276046304541406, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.535070896148682, "learning_rate": 1e-05, "loss": 0.4737, "mean_token_accuracy": 0.8795844912528992, "num_tokens": 605137616.0, "step": 15939 }, { "epoch": 2.027731840732731, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.195897579193115, "learning_rate": 1e-05, "loss": 0.4778, "mean_token_accuracy": 0.8762978911399841, "num_tokens": 605177974.0, "step": 15940 }, { "epoch": 2.0278590510113217, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.470615863800049, "learning_rate": 1e-05, "loss": 0.4813, "mean_token_accuracy": 0.8777463436126709, "num_tokens": 605219156.0, "step": 15941 }, { "epoch": 2.027986261289912, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.068737983703613, "learning_rate": 1e-05, "loss": 0.4453, "mean_token_accuracy": 0.8811897039413452, "num_tokens": 605261131.0, "step": 15942 }, { "epoch": 2.0281134715685027, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.520329475402832, "learning_rate": 1e-05, "loss": 0.4736, "mean_token_accuracy": 0.8747371435165405, "num_tokens": 605303640.0, "step": 15943 }, { "epoch": 2.0282406818470933, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 5.9337544441223145, "learning_rate": 1e-05, "loss": 0.4227, "mean_token_accuracy": 0.8878815174102783, "num_tokens": 605341823.0, "step": 15944 }, { "epoch": 2.028367892125684, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.673776626586914, "learning_rate": 1e-05, "loss": 0.4421, "mean_token_accuracy": 0.8897215127944946, "num_tokens": 605380282.0, "step": 15945 }, { "epoch": 2.0284951024042743, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.796179294586182, "learning_rate": 1e-05, "loss": 0.43, "mean_token_accuracy": 0.8850795030593872, "num_tokens": 605415077.0, "step": 15946 }, { "epoch": 2.028622312682865, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.9372429847717285, "learning_rate": 1e-05, "loss": 0.4702, "mean_token_accuracy": 0.8799602389335632, "num_tokens": 605453805.0, "step": 15947 }, { "epoch": 2.0287495229614554, "ewc_loss": 0.109375, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.742712497711182, "learning_rate": 1e-05, "loss": 0.4614, "mean_token_accuracy": 0.8795791268348694, "num_tokens": 605488740.0, "step": 15948 }, { "epoch": 2.028876733240046, "ewc_loss": 0.140625, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 0.00011587142944335938, "grad_norm": 7.229646682739258, "learning_rate": 1e-05, "loss": 0.5108, "mean_token_accuracy": 0.8710874319076538, "num_tokens": 605529155.0, "step": 15949 }, { "epoch": 2.0290039435186364, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 5.86408805847168, "learning_rate": 1e-05, "loss": 0.4672, "mean_token_accuracy": 0.8776723742485046, "num_tokens": 605567610.0, "step": 15950 }, { "epoch": 2.029131153797227, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 0.00011920928955078125, "grad_norm": 7.408290386199951, "learning_rate": 1e-05, "loss": 0.5092, "mean_token_accuracy": 0.8740627765655518, "num_tokens": 605609485.0, "step": 15951 }, { "epoch": 2.0292583640758175, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.174452781677246, "learning_rate": 1e-05, "loss": 0.4354, "mean_token_accuracy": 0.8899089097976685, "num_tokens": 605644718.0, "step": 15952 }, { "epoch": 2.029385574354408, "ewc_loss": 0.140625, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 0.0001163482666015625, "grad_norm": 7.248349189758301, "learning_rate": 1e-05, "loss": 0.483, "mean_token_accuracy": 0.8802245855331421, "num_tokens": 605686893.0, "step": 15953 }, { "epoch": 2.0295127846329986, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.422390460968018, "learning_rate": 1e-05, "loss": 0.4717, "mean_token_accuracy": 0.8806866407394409, "num_tokens": 605724007.0, "step": 15954 }, { "epoch": 2.0296399949115886, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.842234134674072, "learning_rate": 1e-05, "loss": 0.5016, "mean_token_accuracy": 0.872899055480957, "num_tokens": 605761213.0, "step": 15955 }, { "epoch": 2.029767205190179, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.518547534942627, "learning_rate": 1e-05, "loss": 0.4725, "mean_token_accuracy": 0.8813730478286743, "num_tokens": 605803274.0, "step": 15956 }, { "epoch": 2.0298944154687697, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.349971294403076, "learning_rate": 1e-05, "loss": 0.4766, "mean_token_accuracy": 0.8771336078643799, "num_tokens": 605838326.0, "step": 15957 }, { "epoch": 2.0300216257473602, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.64025354385376, "learning_rate": 1e-05, "loss": 0.4652, "mean_token_accuracy": 0.8829402923583984, "num_tokens": 605878882.0, "step": 15958 }, { "epoch": 2.0301488360259508, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.935194492340088, "learning_rate": 1e-05, "loss": 0.4317, "mean_token_accuracy": 0.8851529955863953, "num_tokens": 605913574.0, "step": 15959 }, { "epoch": 2.0302760463045413, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.887723922729492, "learning_rate": 1e-05, "loss": 0.4459, "mean_token_accuracy": 0.8887453079223633, "num_tokens": 605944845.0, "step": 15960 }, { "epoch": 2.030403256583132, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.70772647857666, "learning_rate": 1e-05, "loss": 0.519, "mean_token_accuracy": 0.857966423034668, "num_tokens": 605982604.0, "step": 15961 }, { "epoch": 2.0305304668617223, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 0.00011491775512695312, "grad_norm": 7.246232032775879, "learning_rate": 1e-05, "loss": 0.5068, "mean_token_accuracy": 0.8740646243095398, "num_tokens": 606023506.0, "step": 15962 }, { "epoch": 2.030657677140313, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.750110626220703, "learning_rate": 1e-05, "loss": 0.4921, "mean_token_accuracy": 0.86911541223526, "num_tokens": 606063319.0, "step": 15963 }, { "epoch": 2.0307848874189034, "ewc_loss": 0.1474609375, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.6025919914245605, "learning_rate": 1e-05, "loss": 0.4486, "mean_token_accuracy": 0.8944650888442993, "num_tokens": 606103393.0, "step": 15964 }, { "epoch": 2.030912097697494, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.1194915771484375, "learning_rate": 1e-05, "loss": 0.4385, "mean_token_accuracy": 0.889439582824707, "num_tokens": 606144191.0, "step": 15965 }, { "epoch": 2.0310393079760845, "ewc_loss": 0.1474609375, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.5918073654174805, "learning_rate": 1e-05, "loss": 0.4889, "mean_token_accuracy": 0.8829090595245361, "num_tokens": 606183219.0, "step": 15966 }, { "epoch": 2.031166518254675, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.482275009155273, "learning_rate": 1e-05, "loss": 0.4819, "mean_token_accuracy": 0.8756022453308105, "num_tokens": 606221429.0, "step": 15967 }, { "epoch": 2.0312937285332655, "ewc_loss": 0.140625, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 0.0001163482666015625, "grad_norm": 7.256818771362305, "learning_rate": 1e-05, "loss": 0.4635, "mean_token_accuracy": 0.8846181631088257, "num_tokens": 606260810.0, "step": 15968 }, { "epoch": 2.031420938811856, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.551318168640137, "learning_rate": 1e-05, "loss": 0.458, "mean_token_accuracy": 0.8850629329681396, "num_tokens": 606299602.0, "step": 15969 }, { "epoch": 2.0315481490904466, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.882568359375, "learning_rate": 1e-05, "loss": 0.4648, "mean_token_accuracy": 0.8828479647636414, "num_tokens": 606341130.0, "step": 15970 }, { "epoch": 2.031675359369037, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.466101169586182, "learning_rate": 1e-05, "loss": 0.4606, "mean_token_accuracy": 0.8834459781646729, "num_tokens": 606378341.0, "step": 15971 }, { "epoch": 2.0318025696476276, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.613459587097168, "learning_rate": 1e-05, "loss": 0.5181, "mean_token_accuracy": 0.862143874168396, "num_tokens": 606413799.0, "step": 15972 }, { "epoch": 2.031929779926218, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.346522808074951, "learning_rate": 1e-05, "loss": 0.4391, "mean_token_accuracy": 0.8863345980644226, "num_tokens": 606447428.0, "step": 15973 }, { "epoch": 2.0320569902048087, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.475516319274902, "learning_rate": 1e-05, "loss": 0.4701, "mean_token_accuracy": 0.8787468671798706, "num_tokens": 606481652.0, "step": 15974 }, { "epoch": 2.032184200483399, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.193657875061035, "learning_rate": 1e-05, "loss": 0.4087, "mean_token_accuracy": 0.8971750140190125, "num_tokens": 606519279.0, "step": 15975 }, { "epoch": 2.0323114107619897, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.4471235275268555, "learning_rate": 1e-05, "loss": 0.4585, "mean_token_accuracy": 0.8800654411315918, "num_tokens": 606556399.0, "step": 15976 }, { "epoch": 2.0324386210405803, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.048646926879883, "learning_rate": 1e-05, "loss": 0.5067, "mean_token_accuracy": 0.8645331859588623, "num_tokens": 606596807.0, "step": 15977 }, { "epoch": 2.032565831319171, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.522256851196289, "learning_rate": 1e-05, "loss": 0.4261, "mean_token_accuracy": 0.8923200368881226, "num_tokens": 606629231.0, "step": 15978 }, { "epoch": 2.032693041597761, "ewc_loss": 0.109375, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.891059398651123, "learning_rate": 1e-05, "loss": 0.4639, "mean_token_accuracy": 0.8748786449432373, "num_tokens": 606669042.0, "step": 15979 }, { "epoch": 2.0328202518763514, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.707526683807373, "learning_rate": 1e-05, "loss": 0.5073, "mean_token_accuracy": 0.8698644638061523, "num_tokens": 606705550.0, "step": 15980 }, { "epoch": 2.032947462154942, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.752326011657715, "learning_rate": 1e-05, "loss": 0.4589, "mean_token_accuracy": 0.8739485740661621, "num_tokens": 606739866.0, "step": 15981 }, { "epoch": 2.0330746724335325, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 7.015803813934326, "learning_rate": 1e-05, "loss": 0.4857, "mean_token_accuracy": 0.8790242671966553, "num_tokens": 606782923.0, "step": 15982 }, { "epoch": 2.033201882712123, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.7031660079956055, "learning_rate": 1e-05, "loss": 0.439, "mean_token_accuracy": 0.8840070962905884, "num_tokens": 606822831.0, "step": 15983 }, { "epoch": 2.0333290929907135, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 0.00011968612670898438, "grad_norm": 7.409618377685547, "learning_rate": 1e-05, "loss": 0.4791, "mean_token_accuracy": 0.8830756545066833, "num_tokens": 606861129.0, "step": 15984 }, { "epoch": 2.033456303269304, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 5.918025493621826, "learning_rate": 1e-05, "loss": 0.5182, "mean_token_accuracy": 0.8623453974723816, "num_tokens": 606902744.0, "step": 15985 }, { "epoch": 2.0335835135478946, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 0.0001220703125, "grad_norm": 7.5911545753479, "learning_rate": 1e-05, "loss": 0.5105, "mean_token_accuracy": 0.8725831508636475, "num_tokens": 606935137.0, "step": 15986 }, { "epoch": 2.033710723826485, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.338732719421387, "learning_rate": 1e-05, "loss": 0.4525, "mean_token_accuracy": 0.8832424879074097, "num_tokens": 606972215.0, "step": 15987 }, { "epoch": 2.0338379341050756, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 0.00011730194091796875, "grad_norm": 7.3329572677612305, "learning_rate": 1e-05, "loss": 0.4905, "mean_token_accuracy": 0.8817726373672485, "num_tokens": 607004489.0, "step": 15988 }, { "epoch": 2.033965144383666, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 6.595047950744629, "learning_rate": 1e-05, "loss": 0.4965, "mean_token_accuracy": 0.8744994401931763, "num_tokens": 607039457.0, "step": 15989 }, { "epoch": 2.0340923546622567, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.806874752044678, "learning_rate": 1e-05, "loss": 0.4935, "mean_token_accuracy": 0.8722468614578247, "num_tokens": 607084863.0, "step": 15990 }, { "epoch": 2.034219564940847, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.634158611297607, "learning_rate": 1e-05, "loss": 0.5155, "mean_token_accuracy": 0.8682563900947571, "num_tokens": 607125013.0, "step": 15991 }, { "epoch": 2.0343467752194377, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.3215813636779785, "learning_rate": 1e-05, "loss": 0.4547, "mean_token_accuracy": 0.8823803663253784, "num_tokens": 607160501.0, "step": 15992 }, { "epoch": 2.0344739854980283, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.698729038238525, "learning_rate": 1e-05, "loss": 0.4986, "mean_token_accuracy": 0.8718712329864502, "num_tokens": 607204621.0, "step": 15993 }, { "epoch": 2.034601195776619, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 5.962034702301025, "learning_rate": 1e-05, "loss": 0.4403, "mean_token_accuracy": 0.8875781297683716, "num_tokens": 607245904.0, "step": 15994 }, { "epoch": 2.0347284060552093, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.889209747314453, "learning_rate": 1e-05, "loss": 0.4814, "mean_token_accuracy": 0.8801496028900146, "num_tokens": 607286907.0, "step": 15995 }, { "epoch": 2.0348556163338, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.791409969329834, "learning_rate": 1e-05, "loss": 0.4812, "mean_token_accuracy": 0.8663182258605957, "num_tokens": 607325990.0, "step": 15996 }, { "epoch": 2.0349828266123904, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 0.00011301040649414062, "grad_norm": 7.1739630699157715, "learning_rate": 1e-05, "loss": 0.5038, "mean_token_accuracy": 0.8718125820159912, "num_tokens": 607363000.0, "step": 15997 }, { "epoch": 2.035110036890981, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 5.802603721618652, "learning_rate": 1e-05, "loss": 0.4683, "mean_token_accuracy": 0.8776235580444336, "num_tokens": 607405913.0, "step": 15998 }, { "epoch": 2.0352372471695714, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 0.00011968612670898438, "grad_norm": 7.43812370300293, "learning_rate": 1e-05, "loss": 0.4888, "mean_token_accuracy": 0.8800433278083801, "num_tokens": 607440242.0, "step": 15999 }, { "epoch": 2.035364457448162, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.053342342376709, "learning_rate": 1e-05, "loss": 0.4658, "mean_token_accuracy": 0.8791306018829346, "num_tokens": 607485705.0, "step": 16000 }, { "epoch": 2.0354916677267525, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 0.0001201629638671875, "grad_norm": 7.470634937286377, "learning_rate": 1e-05, "loss": 0.5164, "mean_token_accuracy": 0.870729923248291, "num_tokens": 607524026.0, "step": 16001 }, { "epoch": 2.035618878005343, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.338149547576904, "learning_rate": 1e-05, "loss": 0.4948, "mean_token_accuracy": 0.8722362518310547, "num_tokens": 607562925.0, "step": 16002 }, { "epoch": 2.0357460882839336, "ewc_loss": 0.140625, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 0.00011587142944335938, "grad_norm": 7.2514166831970215, "learning_rate": 1e-05, "loss": 0.5277, "mean_token_accuracy": 0.8663884401321411, "num_tokens": 607598126.0, "step": 16003 }, { "epoch": 2.0358732985625236, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.45700740814209, "learning_rate": 1e-05, "loss": 0.4468, "mean_token_accuracy": 0.8909097909927368, "num_tokens": 607631911.0, "step": 16004 }, { "epoch": 2.036000508841114, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.923931121826172, "learning_rate": 1e-05, "loss": 0.5063, "mean_token_accuracy": 0.8678762912750244, "num_tokens": 607669817.0, "step": 16005 }, { "epoch": 2.0361277191197047, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.413005352020264, "learning_rate": 1e-05, "loss": 0.4598, "mean_token_accuracy": 0.8842057585716248, "num_tokens": 607710713.0, "step": 16006 }, { "epoch": 2.036254929398295, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.638413429260254, "learning_rate": 1e-05, "loss": 0.4193, "mean_token_accuracy": 0.8962665796279907, "num_tokens": 607751540.0, "step": 16007 }, { "epoch": 2.0363821396768857, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.320337772369385, "learning_rate": 1e-05, "loss": 0.4645, "mean_token_accuracy": 0.8789395689964294, "num_tokens": 607792831.0, "step": 16008 }, { "epoch": 2.0365093499554763, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.494929790496826, "learning_rate": 1e-05, "loss": 0.4771, "mean_token_accuracy": 0.8769960403442383, "num_tokens": 607824175.0, "step": 16009 }, { "epoch": 2.036636560234067, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.196805953979492, "learning_rate": 1e-05, "loss": 0.4926, "mean_token_accuracy": 0.8696376085281372, "num_tokens": 607864302.0, "step": 16010 }, { "epoch": 2.0367637705126573, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.462604522705078, "learning_rate": 1e-05, "loss": 0.4924, "mean_token_accuracy": 0.871290922164917, "num_tokens": 607903231.0, "step": 16011 }, { "epoch": 2.036890980791248, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.0447468757629395, "learning_rate": 1e-05, "loss": 0.4471, "mean_token_accuracy": 0.8846752643585205, "num_tokens": 607942795.0, "step": 16012 }, { "epoch": 2.0370181910698384, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.553651809692383, "learning_rate": 1e-05, "loss": 0.4642, "mean_token_accuracy": 0.8767179250717163, "num_tokens": 607978034.0, "step": 16013 }, { "epoch": 2.037145401348429, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.894720554351807, "learning_rate": 1e-05, "loss": 0.4332, "mean_token_accuracy": 0.8891931772232056, "num_tokens": 608014260.0, "step": 16014 }, { "epoch": 2.0372726116270194, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.730231285095215, "learning_rate": 1e-05, "loss": 0.4909, "mean_token_accuracy": 0.8689143061637878, "num_tokens": 608052637.0, "step": 16015 }, { "epoch": 2.03739982190561, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.772171974182129, "learning_rate": 1e-05, "loss": 0.462, "mean_token_accuracy": 0.8736773729324341, "num_tokens": 608085877.0, "step": 16016 }, { "epoch": 2.0375270321842005, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 7.009862422943115, "learning_rate": 1e-05, "loss": 0.5057, "mean_token_accuracy": 0.8771307468414307, "num_tokens": 608125059.0, "step": 16017 }, { "epoch": 2.037654242462791, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.7416672706604, "learning_rate": 1e-05, "loss": 0.4372, "mean_token_accuracy": 0.8848050236701965, "num_tokens": 608161357.0, "step": 16018 }, { "epoch": 2.0377814527413816, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 0.00011777877807617188, "grad_norm": 7.352234840393066, "learning_rate": 1e-05, "loss": 0.4964, "mean_token_accuracy": 0.875877320766449, "num_tokens": 608196534.0, "step": 16019 }, { "epoch": 2.037908663019972, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 5.968654632568359, "learning_rate": 1e-05, "loss": 0.4573, "mean_token_accuracy": 0.8784184455871582, "num_tokens": 608237130.0, "step": 16020 }, { "epoch": 2.0380358732985626, "ewc_loss": 0.1455078125, "ewc_loss_diag": 0.000247955322265625, "ewc_loss_parallel": 0.00012111663818359375, "grad_norm": 7.420365333557129, "learning_rate": 1e-05, "loss": 0.4464, "mean_token_accuracy": 0.8961230516433716, "num_tokens": 608279064.0, "step": 16021 }, { "epoch": 2.038163083577153, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.000247955322265625, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.271383285522461, "learning_rate": 1e-05, "loss": 0.4406, "mean_token_accuracy": 0.8885154724121094, "num_tokens": 608316019.0, "step": 16022 }, { "epoch": 2.0382902938557437, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.000247955322265625, "ewc_loss_parallel": 0.00011444091796875, "grad_norm": 7.178407669067383, "learning_rate": 1e-05, "loss": 0.45, "mean_token_accuracy": 0.8904412388801575, "num_tokens": 608357199.0, "step": 16023 }, { "epoch": 2.038417504134334, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.000247955322265625, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.475154399871826, "learning_rate": 1e-05, "loss": 0.4676, "mean_token_accuracy": 0.8777682781219482, "num_tokens": 608392149.0, "step": 16024 }, { "epoch": 2.0385447144129247, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.000247955322265625, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.721940517425537, "learning_rate": 1e-05, "loss": 0.4529, "mean_token_accuracy": 0.8872537016868591, "num_tokens": 608431200.0, "step": 16025 }, { "epoch": 2.0386719246915153, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.000247955322265625, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.557072639465332, "learning_rate": 1e-05, "loss": 0.4549, "mean_token_accuracy": 0.8849191069602966, "num_tokens": 608471832.0, "step": 16026 }, { "epoch": 2.038799134970106, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.000247955322265625, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.250192165374756, "learning_rate": 1e-05, "loss": 0.4872, "mean_token_accuracy": 0.8697499632835388, "num_tokens": 608508362.0, "step": 16027 }, { "epoch": 2.0389263452486963, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.000247955322265625, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.688461780548096, "learning_rate": 1e-05, "loss": 0.4952, "mean_token_accuracy": 0.8742243051528931, "num_tokens": 608547988.0, "step": 16028 }, { "epoch": 2.0390535555272864, "ewc_loss": 0.11083984375, "ewc_loss_diag": 0.000247955322265625, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.8761467933654785, "learning_rate": 1e-05, "loss": 0.4314, "mean_token_accuracy": 0.8875135183334351, "num_tokens": 608580448.0, "step": 16029 }, { "epoch": 2.039180765805877, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.000247955322265625, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 6.9674553871154785, "learning_rate": 1e-05, "loss": 0.5146, "mean_token_accuracy": 0.8690581321716309, "num_tokens": 608618573.0, "step": 16030 }, { "epoch": 2.0393079760844675, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.000247955322265625, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.690464019775391, "learning_rate": 1e-05, "loss": 0.4595, "mean_token_accuracy": 0.875150740146637, "num_tokens": 608658161.0, "step": 16031 }, { "epoch": 2.039435186363058, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.000247955322265625, "ewc_loss_parallel": 0.00011873245239257812, "grad_norm": 7.370145797729492, "learning_rate": 1e-05, "loss": 0.5088, "mean_token_accuracy": 0.8720606565475464, "num_tokens": 608696599.0, "step": 16032 }, { "epoch": 2.0395623966416485, "ewc_loss": 0.11474609375, "ewc_loss_diag": 0.000247955322265625, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 5.805506229400635, "learning_rate": 1e-05, "loss": 0.4393, "mean_token_accuracy": 0.8871228098869324, "num_tokens": 608738347.0, "step": 16033 }, { "epoch": 2.039689606920239, "ewc_loss": 0.150390625, "ewc_loss_diag": 0.000247955322265625, "ewc_loss_parallel": 0.000125885009765625, "grad_norm": 7.664313793182373, "learning_rate": 1e-05, "loss": 0.4832, "mean_token_accuracy": 0.8828824758529663, "num_tokens": 608779483.0, "step": 16034 }, { "epoch": 2.0398168171988296, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.000247955322265625, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.239696979522705, "learning_rate": 1e-05, "loss": 0.4812, "mean_token_accuracy": 0.8773033618927002, "num_tokens": 608821369.0, "step": 16035 }, { "epoch": 2.03994402747742, "ewc_loss": 0.1474609375, "ewc_loss_diag": 0.000247955322265625, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.568648815155029, "learning_rate": 1e-05, "loss": 0.5148, "mean_token_accuracy": 0.8784456849098206, "num_tokens": 608854749.0, "step": 16036 }, { "epoch": 2.0400712377560106, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.000247955322265625, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 6.552000999450684, "learning_rate": 1e-05, "loss": 0.45, "mean_token_accuracy": 0.8875383138656616, "num_tokens": 608893692.0, "step": 16037 }, { "epoch": 2.040198448034601, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.000247955322265625, "ewc_loss_parallel": 0.00011444091796875, "grad_norm": 7.1864447593688965, "learning_rate": 1e-05, "loss": 0.4956, "mean_token_accuracy": 0.8732849955558777, "num_tokens": 608930721.0, "step": 16038 }, { "epoch": 2.0403256583131917, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.000247955322265625, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.572307586669922, "learning_rate": 1e-05, "loss": 0.5012, "mean_token_accuracy": 0.8717021942138672, "num_tokens": 608967022.0, "step": 16039 }, { "epoch": 2.040452868591782, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.000247955322265625, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.816640377044678, "learning_rate": 1e-05, "loss": 0.5311, "mean_token_accuracy": 0.8603544235229492, "num_tokens": 609009995.0, "step": 16040 }, { "epoch": 2.0405800788703727, "ewc_loss": 0.125, "ewc_loss_diag": 0.000247955322265625, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.458219528198242, "learning_rate": 1e-05, "loss": 0.4828, "mean_token_accuracy": 0.8754737377166748, "num_tokens": 609044636.0, "step": 16041 }, { "epoch": 2.0407072891489633, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.000247955322265625, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.5751214027404785, "learning_rate": 1e-05, "loss": 0.4488, "mean_token_accuracy": 0.8860616087913513, "num_tokens": 609076328.0, "step": 16042 }, { "epoch": 2.040834499427554, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.000247955322265625, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.311905384063721, "learning_rate": 1e-05, "loss": 0.5099, "mean_token_accuracy": 0.8632593750953674, "num_tokens": 609118955.0, "step": 16043 }, { "epoch": 2.0409617097061443, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.000247955322265625, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.462985038757324, "learning_rate": 1e-05, "loss": 0.4959, "mean_token_accuracy": 0.8685104250907898, "num_tokens": 609157958.0, "step": 16044 }, { "epoch": 2.041088919984735, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.000247955322265625, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.153474807739258, "learning_rate": 1e-05, "loss": 0.4439, "mean_token_accuracy": 0.8835241794586182, "num_tokens": 609197636.0, "step": 16045 }, { "epoch": 2.0412161302633254, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.000247955322265625, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.4850616455078125, "learning_rate": 1e-05, "loss": 0.5316, "mean_token_accuracy": 0.8595823049545288, "num_tokens": 609239648.0, "step": 16046 }, { "epoch": 2.041343340541916, "ewc_loss": 0.11181640625, "ewc_loss_diag": 0.000247955322265625, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.976499080657959, "learning_rate": 1e-05, "loss": 0.418, "mean_token_accuracy": 0.8906642198562622, "num_tokens": 609281012.0, "step": 16047 }, { "epoch": 2.0414705508205064, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.000247955322265625, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.619514465332031, "learning_rate": 1e-05, "loss": 0.4786, "mean_token_accuracy": 0.8770670890808105, "num_tokens": 609323775.0, "step": 16048 }, { "epoch": 2.041597761099097, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.000247955322265625, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.812353134155273, "learning_rate": 1e-05, "loss": 0.4139, "mean_token_accuracy": 0.892013669013977, "num_tokens": 609358960.0, "step": 16049 }, { "epoch": 2.0417249713776875, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.000247955322265625, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.890984535217285, "learning_rate": 1e-05, "loss": 0.5098, "mean_token_accuracy": 0.8673617243766785, "num_tokens": 609395447.0, "step": 16050 }, { "epoch": 2.041852181656278, "ewc_loss": 0.10888671875, "ewc_loss_diag": 0.000247955322265625, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.701295852661133, "learning_rate": 1e-05, "loss": 0.4149, "mean_token_accuracy": 0.8914428353309631, "num_tokens": 609429811.0, "step": 16051 }, { "epoch": 2.0419793919348685, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.000247955322265625, "ewc_loss_parallel": 0.00011539459228515625, "grad_norm": 7.267895698547363, "learning_rate": 1e-05, "loss": 0.4715, "mean_token_accuracy": 0.8830602765083313, "num_tokens": 609466104.0, "step": 16052 }, { "epoch": 2.0421066022134586, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.000247955322265625, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 5.786256313323975, "learning_rate": 1e-05, "loss": 0.4236, "mean_token_accuracy": 0.8866744637489319, "num_tokens": 609503513.0, "step": 16053 }, { "epoch": 2.042233812492049, "ewc_loss": 0.1484375, "ewc_loss_diag": 0.000247955322265625, "ewc_loss_parallel": 0.0001239776611328125, "grad_norm": 7.57754373550415, "learning_rate": 1e-05, "loss": 0.4597, "mean_token_accuracy": 0.8895124197006226, "num_tokens": 609546899.0, "step": 16054 }, { "epoch": 2.0423610227706397, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.000247955322265625, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.1918625831604, "learning_rate": 1e-05, "loss": 0.5007, "mean_token_accuracy": 0.8723081350326538, "num_tokens": 609589297.0, "step": 16055 }, { "epoch": 2.04248823304923, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.000247955322265625, "ewc_loss_parallel": 0.00012159347534179688, "grad_norm": 7.482308864593506, "learning_rate": 1e-05, "loss": 0.4669, "mean_token_accuracy": 0.8869516849517822, "num_tokens": 609624839.0, "step": 16056 }, { "epoch": 2.0426154433278207, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.000247955322265625, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 6.544817924499512, "learning_rate": 1e-05, "loss": 0.4456, "mean_token_accuracy": 0.8874891996383667, "num_tokens": 609661708.0, "step": 16057 }, { "epoch": 2.0427426536064113, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.000247955322265625, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 7.018482685089111, "learning_rate": 1e-05, "loss": 0.5141, "mean_token_accuracy": 0.8669064044952393, "num_tokens": 609704133.0, "step": 16058 }, { "epoch": 2.042869863885002, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.000247955322265625, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.624073028564453, "learning_rate": 1e-05, "loss": 0.4782, "mean_token_accuracy": 0.8781816363334656, "num_tokens": 609740743.0, "step": 16059 }, { "epoch": 2.0429970741635923, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.000247955322265625, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.505586624145508, "learning_rate": 1e-05, "loss": 0.5004, "mean_token_accuracy": 0.8692457675933838, "num_tokens": 609777520.0, "step": 16060 }, { "epoch": 2.043124284442183, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.000247955322265625, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.670060157775879, "learning_rate": 1e-05, "loss": 0.4795, "mean_token_accuracy": 0.8784395456314087, "num_tokens": 609812974.0, "step": 16061 }, { "epoch": 2.0432514947207734, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.000247955322265625, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 6.077408790588379, "learning_rate": 1e-05, "loss": 0.4458, "mean_token_accuracy": 0.8854191303253174, "num_tokens": 609853287.0, "step": 16062 }, { "epoch": 2.043378704999364, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.000247955322265625, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.829634666442871, "learning_rate": 1e-05, "loss": 0.4371, "mean_token_accuracy": 0.8900234699249268, "num_tokens": 609888768.0, "step": 16063 }, { "epoch": 2.0435059152779544, "ewc_loss": 0.10888671875, "ewc_loss_diag": 0.000247955322265625, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.816044807434082, "learning_rate": 1e-05, "loss": 0.4346, "mean_token_accuracy": 0.8856712579727173, "num_tokens": 609927976.0, "step": 16064 }, { "epoch": 2.043633125556545, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.000247955322265625, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 7.110607147216797, "learning_rate": 1e-05, "loss": 0.4561, "mean_token_accuracy": 0.8855233788490295, "num_tokens": 609964404.0, "step": 16065 }, { "epoch": 2.0437603358351355, "ewc_loss": 0.109375, "ewc_loss_diag": 0.000247955322265625, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.770578384399414, "learning_rate": 1e-05, "loss": 0.4614, "mean_token_accuracy": 0.87790846824646, "num_tokens": 610005188.0, "step": 16066 }, { "epoch": 2.043887546113726, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.000247955322265625, "ewc_loss_parallel": 0.00011920928955078125, "grad_norm": 7.423642158508301, "learning_rate": 1e-05, "loss": 0.4756, "mean_token_accuracy": 0.8849285840988159, "num_tokens": 610039118.0, "step": 16067 }, { "epoch": 2.0440147563923166, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.000247955322265625, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 5.990851879119873, "learning_rate": 1e-05, "loss": 0.4333, "mean_token_accuracy": 0.888970136642456, "num_tokens": 610074867.0, "step": 16068 }, { "epoch": 2.044141966670907, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.000247955322265625, "ewc_loss_parallel": 0.0001220703125, "grad_norm": 7.526439189910889, "learning_rate": 1e-05, "loss": 0.4805, "mean_token_accuracy": 0.8817290663719177, "num_tokens": 610111382.0, "step": 16069 }, { "epoch": 2.0442691769494976, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.000247955322265625, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.314842700958252, "learning_rate": 1e-05, "loss": 0.4691, "mean_token_accuracy": 0.8772994875907898, "num_tokens": 610148026.0, "step": 16070 }, { "epoch": 2.044396387228088, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.000247955322265625, "ewc_loss_parallel": 0.00011730194091796875, "grad_norm": 7.335744857788086, "learning_rate": 1e-05, "loss": 0.459, "mean_token_accuracy": 0.8865716457366943, "num_tokens": 610179841.0, "step": 16071 }, { "epoch": 2.0445235975066787, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.000247955322265625, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.472531795501709, "learning_rate": 1e-05, "loss": 0.4797, "mean_token_accuracy": 0.8752598762512207, "num_tokens": 610214224.0, "step": 16072 }, { "epoch": 2.044650807785269, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.000247955322265625, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 6.994268894195557, "learning_rate": 1e-05, "loss": 0.4787, "mean_token_accuracy": 0.8810656070709229, "num_tokens": 610257640.0, "step": 16073 }, { "epoch": 2.0447780180638597, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.000247955322265625, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.452654838562012, "learning_rate": 1e-05, "loss": 0.4356, "mean_token_accuracy": 0.8903363943099976, "num_tokens": 610295438.0, "step": 16074 }, { "epoch": 2.0449052283424503, "ewc_loss": 0.125, "ewc_loss_diag": 0.000247955322265625, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.695855617523193, "learning_rate": 1e-05, "loss": 0.4449, "mean_token_accuracy": 0.8875570297241211, "num_tokens": 610336390.0, "step": 16075 }, { "epoch": 2.045032438621041, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.000247955322265625, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.3632354736328125, "learning_rate": 1e-05, "loss": 0.451, "mean_token_accuracy": 0.8832786083221436, "num_tokens": 610369860.0, "step": 16076 }, { "epoch": 2.045159648899631, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.000247955322265625, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.517485618591309, "learning_rate": 1e-05, "loss": 0.457, "mean_token_accuracy": 0.8827369213104248, "num_tokens": 610408116.0, "step": 16077 }, { "epoch": 2.0452868591782214, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.000247955322265625, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.2398176193237305, "learning_rate": 1e-05, "loss": 0.493, "mean_token_accuracy": 0.8699395656585693, "num_tokens": 610444317.0, "step": 16078 }, { "epoch": 2.045414069456812, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.000247955322265625, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.469152450561523, "learning_rate": 1e-05, "loss": 0.4904, "mean_token_accuracy": 0.8720779418945312, "num_tokens": 610485675.0, "step": 16079 }, { "epoch": 2.0455412797354025, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.000247955322265625, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.071675777435303, "learning_rate": 1e-05, "loss": 0.4437, "mean_token_accuracy": 0.8847758173942566, "num_tokens": 610524697.0, "step": 16080 }, { "epoch": 2.045668490013993, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.000247955322265625, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.537808418273926, "learning_rate": 1e-05, "loss": 0.466, "mean_token_accuracy": 0.8829938769340515, "num_tokens": 610567470.0, "step": 16081 }, { "epoch": 2.0457957002925835, "ewc_loss": 0.11083984375, "ewc_loss_diag": 0.000247955322265625, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.912889003753662, "learning_rate": 1e-05, "loss": 0.4424, "mean_token_accuracy": 0.8812284469604492, "num_tokens": 610603168.0, "step": 16082 }, { "epoch": 2.045922910571174, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.000247955322265625, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.722011566162109, "learning_rate": 1e-05, "loss": 0.4615, "mean_token_accuracy": 0.8791235685348511, "num_tokens": 610643416.0, "step": 16083 }, { "epoch": 2.0460501208497646, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.000247955322265625, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.7678117752075195, "learning_rate": 1e-05, "loss": 0.4396, "mean_token_accuracy": 0.8819706439971924, "num_tokens": 610676513.0, "step": 16084 }, { "epoch": 2.046177331128355, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.000247955322265625, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 7.026015758514404, "learning_rate": 1e-05, "loss": 0.4642, "mean_token_accuracy": 0.8819384574890137, "num_tokens": 610710842.0, "step": 16085 }, { "epoch": 2.0463045414069456, "ewc_loss": 0.109375, "ewc_loss_diag": 0.000247955322265625, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.720970630645752, "learning_rate": 1e-05, "loss": 0.4873, "mean_token_accuracy": 0.8667573928833008, "num_tokens": 610747770.0, "step": 16086 }, { "epoch": 2.046431751685536, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.000247955322265625, "ewc_loss_parallel": 0.00011873245239257812, "grad_norm": 7.366127014160156, "learning_rate": 1e-05, "loss": 0.4697, "mean_token_accuracy": 0.8856320977210999, "num_tokens": 610788780.0, "step": 16087 }, { "epoch": 2.0465589619641267, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.000247955322265625, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 5.920650482177734, "learning_rate": 1e-05, "loss": 0.4452, "mean_token_accuracy": 0.8838902115821838, "num_tokens": 610832930.0, "step": 16088 }, { "epoch": 2.046686172242717, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.000247955322265625, "ewc_loss_parallel": 0.0001220703125, "grad_norm": 7.5114240646362305, "learning_rate": 1e-05, "loss": 0.4662, "mean_token_accuracy": 0.8836173415184021, "num_tokens": 610875828.0, "step": 16089 }, { "epoch": 2.0468133825213077, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.000247955322265625, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.314820289611816, "learning_rate": 1e-05, "loss": 0.4487, "mean_token_accuracy": 0.8876827955245972, "num_tokens": 610913920.0, "step": 16090 }, { "epoch": 2.0469405927998983, "ewc_loss": 0.140625, "ewc_loss_diag": 0.000247955322265625, "ewc_loss_parallel": 0.0001163482666015625, "grad_norm": 7.237416744232178, "learning_rate": 1e-05, "loss": 0.4782, "mean_token_accuracy": 0.8801286220550537, "num_tokens": 610956887.0, "step": 16091 }, { "epoch": 2.047067803078489, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.000247955322265625, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.5384650230407715, "learning_rate": 1e-05, "loss": 0.5118, "mean_token_accuracy": 0.8707594871520996, "num_tokens": 611000889.0, "step": 16092 }, { "epoch": 2.0471950133570793, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.000247955322265625, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.75055456161499, "learning_rate": 1e-05, "loss": 0.4776, "mean_token_accuracy": 0.8780761957168579, "num_tokens": 611040573.0, "step": 16093 }, { "epoch": 2.04732222363567, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.000247955322265625, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.609371662139893, "learning_rate": 1e-05, "loss": 0.4467, "mean_token_accuracy": 0.8876618146896362, "num_tokens": 611075925.0, "step": 16094 }, { "epoch": 2.0474494339142604, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.000247955322265625, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.239840984344482, "learning_rate": 1e-05, "loss": 0.4821, "mean_token_accuracy": 0.8713911175727844, "num_tokens": 611116492.0, "step": 16095 }, { "epoch": 2.047576644192851, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.000247955322265625, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.751652240753174, "learning_rate": 1e-05, "loss": 0.4559, "mean_token_accuracy": 0.8838236927986145, "num_tokens": 611155683.0, "step": 16096 }, { "epoch": 2.0477038544714414, "ewc_loss": 0.109375, "ewc_loss_diag": 0.000247955322265625, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.858055114746094, "learning_rate": 1e-05, "loss": 0.3686, "mean_token_accuracy": 0.9059421420097351, "num_tokens": 611194352.0, "step": 16097 }, { "epoch": 2.047831064750032, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.000247955322265625, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 7.027385711669922, "learning_rate": 1e-05, "loss": 0.46, "mean_token_accuracy": 0.8840161561965942, "num_tokens": 611229919.0, "step": 16098 }, { "epoch": 2.0479582750286225, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.000247955322265625, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.726870059967041, "learning_rate": 1e-05, "loss": 0.4633, "mean_token_accuracy": 0.8773011565208435, "num_tokens": 611267394.0, "step": 16099 }, { "epoch": 2.048085485307213, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.000247955322265625, "ewc_loss_parallel": 0.00011968612670898438, "grad_norm": 7.3769450187683105, "learning_rate": 1e-05, "loss": 0.5769, "mean_token_accuracy": 0.8512859344482422, "num_tokens": 611312009.0, "step": 16100 }, { "epoch": 2.0482126955858035, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.000247955322265625, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 5.901896953582764, "learning_rate": 1e-05, "loss": 0.5153, "mean_token_accuracy": 0.8654805421829224, "num_tokens": 611346723.0, "step": 16101 }, { "epoch": 2.0483399058643936, "ewc_loss": 0.1474609375, "ewc_loss_diag": 0.000247955322265625, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.575843811035156, "learning_rate": 1e-05, "loss": 0.4792, "mean_token_accuracy": 0.8837226629257202, "num_tokens": 611381803.0, "step": 16102 }, { "epoch": 2.048467116142984, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.000247955322265625, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.2728753089904785, "learning_rate": 1e-05, "loss": 0.4791, "mean_token_accuracy": 0.8783516883850098, "num_tokens": 611419579.0, "step": 16103 }, { "epoch": 2.0485943264215747, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.000247955322265625, "ewc_loss_parallel": 0.00011873245239257812, "grad_norm": 7.419486999511719, "learning_rate": 1e-05, "loss": 0.4874, "mean_token_accuracy": 0.8786582350730896, "num_tokens": 611456904.0, "step": 16104 }, { "epoch": 2.048721536700165, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.000247955322265625, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.482999324798584, "learning_rate": 1e-05, "loss": 0.4396, "mean_token_accuracy": 0.8893443942070007, "num_tokens": 611495597.0, "step": 16105 }, { "epoch": 2.0488487469787557, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.000247955322265625, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 7.098679542541504, "learning_rate": 1e-05, "loss": 0.4546, "mean_token_accuracy": 0.8844884634017944, "num_tokens": 611531307.0, "step": 16106 }, { "epoch": 2.0489759572573463, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.000247955322265625, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.483907222747803, "learning_rate": 1e-05, "loss": 0.5212, "mean_token_accuracy": 0.8642839789390564, "num_tokens": 611567657.0, "step": 16107 }, { "epoch": 2.049103167535937, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.000247955322265625, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.777315139770508, "learning_rate": 1e-05, "loss": 0.484, "mean_token_accuracy": 0.8769789934158325, "num_tokens": 611608107.0, "step": 16108 }, { "epoch": 2.0492303778145273, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.000247955322265625, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.397952079772949, "learning_rate": 1e-05, "loss": 0.4789, "mean_token_accuracy": 0.8783538341522217, "num_tokens": 611644695.0, "step": 16109 }, { "epoch": 2.049357588093118, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.000247955322265625, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.5614447593688965, "learning_rate": 1e-05, "loss": 0.4542, "mean_token_accuracy": 0.8838849663734436, "num_tokens": 611679330.0, "step": 16110 }, { "epoch": 2.0494847983717084, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.000247955322265625, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.260222434997559, "learning_rate": 1e-05, "loss": 0.5068, "mean_token_accuracy": 0.8652565479278564, "num_tokens": 611718792.0, "step": 16111 }, { "epoch": 2.049612008650299, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.000247955322265625, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.494617462158203, "learning_rate": 1e-05, "loss": 0.4902, "mean_token_accuracy": 0.8702294826507568, "num_tokens": 611752559.0, "step": 16112 }, { "epoch": 2.0497392189288894, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.000247955322265625, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.10245943069458, "learning_rate": 1e-05, "loss": 0.4459, "mean_token_accuracy": 0.8839452862739563, "num_tokens": 611791616.0, "step": 16113 }, { "epoch": 2.04986642920748, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.000247955322265625, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.560793399810791, "learning_rate": 1e-05, "loss": 0.4921, "mean_token_accuracy": 0.869645893573761, "num_tokens": 611826577.0, "step": 16114 }, { "epoch": 2.0499936394860705, "ewc_loss": 0.11083984375, "ewc_loss_diag": 0.000247955322265625, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.9396443367004395, "learning_rate": 1e-05, "loss": 0.4632, "mean_token_accuracy": 0.8763834238052368, "num_tokens": 611861777.0, "step": 16115 }, { "epoch": 2.050120849764661, "ewc_loss": 0.125, "ewc_loss_diag": 0.000247955322265625, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.717743873596191, "learning_rate": 1e-05, "loss": 0.4938, "mean_token_accuracy": 0.8714834451675415, "num_tokens": 611897933.0, "step": 16116 }, { "epoch": 2.0502480600432516, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.000247955322265625, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.793161392211914, "learning_rate": 1e-05, "loss": 0.4515, "mean_token_accuracy": 0.8794467449188232, "num_tokens": 611935285.0, "step": 16117 }, { "epoch": 2.050375270321842, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.000247955322265625, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 7.0091729164123535, "learning_rate": 1e-05, "loss": 0.4496, "mean_token_accuracy": 0.8877831697463989, "num_tokens": 611969879.0, "step": 16118 }, { "epoch": 2.0505024806004326, "ewc_loss": 0.10888671875, "ewc_loss_diag": 0.000247955322265625, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.750799655914307, "learning_rate": 1e-05, "loss": 0.4337, "mean_token_accuracy": 0.885676383972168, "num_tokens": 612003608.0, "step": 16119 }, { "epoch": 2.050629690879023, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.000247955322265625, "ewc_loss_parallel": 0.00011777877807617188, "grad_norm": 7.336887359619141, "learning_rate": 1e-05, "loss": 0.515, "mean_token_accuracy": 0.8746511936187744, "num_tokens": 612044247.0, "step": 16120 }, { "epoch": 2.0507569011576137, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.000247955322265625, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 5.93628454208374, "learning_rate": 1e-05, "loss": 0.5012, "mean_token_accuracy": 0.8715686798095703, "num_tokens": 612080776.0, "step": 16121 }, { "epoch": 2.050884111436204, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.000247955322265625, "ewc_loss_parallel": 0.00012159347534179688, "grad_norm": 7.479263782501221, "learning_rate": 1e-05, "loss": 0.5229, "mean_token_accuracy": 0.8687925338745117, "num_tokens": 612119162.0, "step": 16122 }, { "epoch": 2.0510113217147947, "ewc_loss": 0.125, "ewc_loss_diag": 0.000247955322265625, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.281851291656494, "learning_rate": 1e-05, "loss": 0.4351, "mean_token_accuracy": 0.8905083537101746, "num_tokens": 612157774.0, "step": 16123 }, { "epoch": 2.0511385319933853, "ewc_loss": 0.140625, "ewc_loss_diag": 0.000247955322265625, "ewc_loss_parallel": 0.00011587142944335938, "grad_norm": 7.250297546386719, "learning_rate": 1e-05, "loss": 0.4882, "mean_token_accuracy": 0.8799974322319031, "num_tokens": 612192169.0, "step": 16124 }, { "epoch": 2.051265742271976, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.000247955322265625, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.5120320320129395, "learning_rate": 1e-05, "loss": 0.454, "mean_token_accuracy": 0.8847346901893616, "num_tokens": 612225936.0, "step": 16125 }, { "epoch": 2.0513929525505663, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.000247955322265625, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.778297424316406, "learning_rate": 1e-05, "loss": 0.4603, "mean_token_accuracy": 0.8841613531112671, "num_tokens": 612262509.0, "step": 16126 }, { "epoch": 2.0515201628291564, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.000247955322265625, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.590490818023682, "learning_rate": 1e-05, "loss": 0.4879, "mean_token_accuracy": 0.8752244710922241, "num_tokens": 612297952.0, "step": 16127 }, { "epoch": 2.051647373107747, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.000247955322265625, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.2723469734191895, "learning_rate": 1e-05, "loss": 0.4509, "mean_token_accuracy": 0.881786584854126, "num_tokens": 612332512.0, "step": 16128 }, { "epoch": 2.0517745833863374, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.000247955322265625, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.72307825088501, "learning_rate": 1e-05, "loss": 0.453, "mean_token_accuracy": 0.8862794041633606, "num_tokens": 612370350.0, "step": 16129 }, { "epoch": 2.051901793664928, "ewc_loss": 0.109375, "ewc_loss_diag": 0.000247955322265625, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.882308483123779, "learning_rate": 1e-05, "loss": 0.482, "mean_token_accuracy": 0.8689666986465454, "num_tokens": 612404775.0, "step": 16130 }, { "epoch": 2.0520290039435185, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.000247955322265625, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 6.987757682800293, "learning_rate": 1e-05, "loss": 0.4953, "mean_token_accuracy": 0.8788936138153076, "num_tokens": 612447226.0, "step": 16131 }, { "epoch": 2.052156214222109, "ewc_loss": 0.10888671875, "ewc_loss_diag": 0.000247955322265625, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.715229034423828, "learning_rate": 1e-05, "loss": 0.489, "mean_token_accuracy": 0.8673075437545776, "num_tokens": 612481358.0, "step": 16132 }, { "epoch": 2.0522834245006996, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.000247955322265625, "ewc_loss_parallel": 0.00011920928955078125, "grad_norm": 7.348209381103516, "learning_rate": 1e-05, "loss": 0.4573, "mean_token_accuracy": 0.8895574808120728, "num_tokens": 612519132.0, "step": 16133 }, { "epoch": 2.05241063477929, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.000247955322265625, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 5.8326416015625, "learning_rate": 1e-05, "loss": 0.497, "mean_token_accuracy": 0.8697957396507263, "num_tokens": 612559187.0, "step": 16134 }, { "epoch": 2.0525378450578806, "ewc_loss": 0.150390625, "ewc_loss_diag": 0.000247955322265625, "ewc_loss_parallel": 0.000125885009765625, "grad_norm": 8.037280082702637, "learning_rate": 1e-05, "loss": 0.4766, "mean_token_accuracy": 0.8835433721542358, "num_tokens": 612598208.0, "step": 16135 }, { "epoch": 2.052665055336471, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002498626708984375, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 6.444532871246338, "learning_rate": 1e-05, "loss": 0.4785, "mean_token_accuracy": 0.8801698088645935, "num_tokens": 612633951.0, "step": 16136 }, { "epoch": 2.0527922656150617, "ewc_loss": 0.1474609375, "ewc_loss_diag": 0.0002498626708984375, "ewc_loss_parallel": 0.0001220703125, "grad_norm": 7.512548923492432, "learning_rate": 1e-05, "loss": 0.5046, "mean_token_accuracy": 0.873802661895752, "num_tokens": 612673978.0, "step": 16137 }, { "epoch": 2.052919475893652, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.0002498626708984375, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 6.633289337158203, "learning_rate": 1e-05, "loss": 0.4692, "mean_token_accuracy": 0.8809574842453003, "num_tokens": 612720488.0, "step": 16138 }, { "epoch": 2.0530466861722427, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002498626708984375, "ewc_loss_parallel": 0.00011348724365234375, "grad_norm": 7.152493476867676, "learning_rate": 1e-05, "loss": 0.5076, "mean_token_accuracy": 0.8750956058502197, "num_tokens": 612755009.0, "step": 16139 }, { "epoch": 2.0531738964508333, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002498626708984375, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 6.635979652404785, "learning_rate": 1e-05, "loss": 0.5008, "mean_token_accuracy": 0.8748515248298645, "num_tokens": 612790653.0, "step": 16140 }, { "epoch": 2.053301106729424, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.00025177001953125, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.776533126831055, "learning_rate": 1e-05, "loss": 0.4634, "mean_token_accuracy": 0.8825429677963257, "num_tokens": 612829406.0, "step": 16141 }, { "epoch": 2.0534283170080143, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002498626708984375, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.458988666534424, "learning_rate": 1e-05, "loss": 0.5163, "mean_token_accuracy": 0.8686004877090454, "num_tokens": 612872304.0, "step": 16142 }, { "epoch": 2.053555527286605, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002498626708984375, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.571174144744873, "learning_rate": 1e-05, "loss": 0.4824, "mean_token_accuracy": 0.8757981061935425, "num_tokens": 612907909.0, "step": 16143 }, { "epoch": 2.0536827375651954, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.0002498626708984375, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.291168689727783, "learning_rate": 1e-05, "loss": 0.4341, "mean_token_accuracy": 0.8902501463890076, "num_tokens": 612943794.0, "step": 16144 }, { "epoch": 2.053809947843786, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002498626708984375, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.5082268714904785, "learning_rate": 1e-05, "loss": 0.4492, "mean_token_accuracy": 0.884286642074585, "num_tokens": 612987125.0, "step": 16145 }, { "epoch": 2.0539371581223764, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.0002498626708984375, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.097517967224121, "learning_rate": 1e-05, "loss": 0.4917, "mean_token_accuracy": 0.8705257177352905, "num_tokens": 613028232.0, "step": 16146 }, { "epoch": 2.054064368400967, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002498626708984375, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.591619968414307, "learning_rate": 1e-05, "loss": 0.4613, "mean_token_accuracy": 0.8815878629684448, "num_tokens": 613066651.0, "step": 16147 }, { "epoch": 2.0541915786795575, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.0002498626708984375, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 5.961235523223877, "learning_rate": 1e-05, "loss": 0.4392, "mean_token_accuracy": 0.8864110708236694, "num_tokens": 613105153.0, "step": 16148 }, { "epoch": 2.054318788958148, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002498626708984375, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.725380897521973, "learning_rate": 1e-05, "loss": 0.4449, "mean_token_accuracy": 0.8873139023780823, "num_tokens": 613146817.0, "step": 16149 }, { "epoch": 2.0544459992367385, "ewc_loss": 0.11083984375, "ewc_loss_diag": 0.0002498626708984375, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.783132076263428, "learning_rate": 1e-05, "loss": 0.4803, "mean_token_accuracy": 0.8715405464172363, "num_tokens": 613192644.0, "step": 16150 }, { "epoch": 2.0545732095153286, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.0002498626708984375, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 7.024443626403809, "learning_rate": 1e-05, "loss": 0.5049, "mean_token_accuracy": 0.8765971660614014, "num_tokens": 613232283.0, "step": 16151 }, { "epoch": 2.054700419793919, "ewc_loss": 0.11083984375, "ewc_loss_diag": 0.0002498626708984375, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.730379104614258, "learning_rate": 1e-05, "loss": 0.4382, "mean_token_accuracy": 0.8848813772201538, "num_tokens": 613270916.0, "step": 16152 }, { "epoch": 2.0548276300725097, "ewc_loss": 0.1455078125, "ewc_loss_diag": 0.0002498626708984375, "ewc_loss_parallel": 0.0001201629638671875, "grad_norm": 7.412464141845703, "learning_rate": 1e-05, "loss": 0.4656, "mean_token_accuracy": 0.8883121609687805, "num_tokens": 613303095.0, "step": 16153 }, { "epoch": 2.0549548403511, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0002498626708984375, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 5.9426350593566895, "learning_rate": 1e-05, "loss": 0.4771, "mean_token_accuracy": 0.873275876045227, "num_tokens": 613341551.0, "step": 16154 }, { "epoch": 2.0550820506296907, "ewc_loss": 0.150390625, "ewc_loss_diag": 0.0002498626708984375, "ewc_loss_parallel": 0.00012493133544921875, "grad_norm": 7.578981399536133, "learning_rate": 1e-05, "loss": 0.4661, "mean_token_accuracy": 0.8873430490493774, "num_tokens": 613377511.0, "step": 16155 }, { "epoch": 2.0552092609082813, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002498626708984375, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.363924980163574, "learning_rate": 1e-05, "loss": 0.4543, "mean_token_accuracy": 0.8864299654960632, "num_tokens": 613408534.0, "step": 16156 }, { "epoch": 2.055336471186872, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.0002498626708984375, "ewc_loss_parallel": 0.00011730194091796875, "grad_norm": 7.309925079345703, "learning_rate": 1e-05, "loss": 0.4473, "mean_token_accuracy": 0.8928803205490112, "num_tokens": 613443461.0, "step": 16157 }, { "epoch": 2.0554636814654623, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002498626708984375, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 6.579750061035156, "learning_rate": 1e-05, "loss": 0.5108, "mean_token_accuracy": 0.8670664429664612, "num_tokens": 613482438.0, "step": 16158 }, { "epoch": 2.055590891744053, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002498626708984375, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.815550327301025, "learning_rate": 1e-05, "loss": 0.5064, "mean_token_accuracy": 0.8701255321502686, "num_tokens": 613521937.0, "step": 16159 }, { "epoch": 2.0557181020226434, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.0002498626708984375, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.638551235198975, "learning_rate": 1e-05, "loss": 0.5066, "mean_token_accuracy": 0.8667871952056885, "num_tokens": 613563471.0, "step": 16160 }, { "epoch": 2.055845312301234, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.0002498626708984375, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.306582927703857, "learning_rate": 1e-05, "loss": 0.4661, "mean_token_accuracy": 0.8785053491592407, "num_tokens": 613603753.0, "step": 16161 }, { "epoch": 2.0559725225798244, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002498626708984375, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.738963603973389, "learning_rate": 1e-05, "loss": 0.4547, "mean_token_accuracy": 0.8867473006248474, "num_tokens": 613637602.0, "step": 16162 }, { "epoch": 2.056099732858415, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.00025177001953125, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.930005073547363, "learning_rate": 1e-05, "loss": 0.4535, "mean_token_accuracy": 0.8784340620040894, "num_tokens": 613673807.0, "step": 16163 }, { "epoch": 2.0562269431370055, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.00025177001953125, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 6.96558141708374, "learning_rate": 1e-05, "loss": 0.4894, "mean_token_accuracy": 0.8762946128845215, "num_tokens": 613710702.0, "step": 16164 }, { "epoch": 2.056354153415596, "ewc_loss": 0.109375, "ewc_loss_diag": 0.00025177001953125, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.752068042755127, "learning_rate": 1e-05, "loss": 0.4552, "mean_token_accuracy": 0.8776658177375793, "num_tokens": 613748890.0, "step": 16165 }, { "epoch": 2.0564813636941865, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.00025177001953125, "ewc_loss_parallel": 0.00011682510375976562, "grad_norm": 7.305778503417969, "learning_rate": 1e-05, "loss": 0.497, "mean_token_accuracy": 0.8759040236473083, "num_tokens": 613791039.0, "step": 16166 }, { "epoch": 2.056608573972777, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.00025177001953125, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 5.840210437774658, "learning_rate": 1e-05, "loss": 0.4467, "mean_token_accuracy": 0.8822647333145142, "num_tokens": 613821345.0, "step": 16167 }, { "epoch": 2.0567357842513676, "ewc_loss": 0.1484375, "ewc_loss_diag": 0.00025177001953125, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.585519313812256, "learning_rate": 1e-05, "loss": 0.5709, "mean_token_accuracy": 0.851909339427948, "num_tokens": 613857770.0, "step": 16168 }, { "epoch": 2.056862994529958, "ewc_loss": 0.125, "ewc_loss_diag": 0.00025177001953125, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.191593170166016, "learning_rate": 1e-05, "loss": 0.4427, "mean_token_accuracy": 0.8841047286987305, "num_tokens": 613898844.0, "step": 16169 }, { "epoch": 2.0569902048085487, "ewc_loss": 0.1474609375, "ewc_loss_diag": 0.00025177001953125, "ewc_loss_parallel": 0.0001220703125, "grad_norm": 7.519516468048096, "learning_rate": 1e-05, "loss": 0.5165, "mean_token_accuracy": 0.868278443813324, "num_tokens": 613940884.0, "step": 16170 }, { "epoch": 2.057117415087139, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.00025177001953125, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.483479022979736, "learning_rate": 1e-05, "loss": 0.4779, "mean_token_accuracy": 0.8796168565750122, "num_tokens": 613981664.0, "step": 16171 }, { "epoch": 2.0572446253657297, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.00025177001953125, "ewc_loss_parallel": 0.00011301040649414062, "grad_norm": 7.199107646942139, "learning_rate": 1e-05, "loss": 0.4987, "mean_token_accuracy": 0.8729514479637146, "num_tokens": 614013965.0, "step": 16172 }, { "epoch": 2.0573718356443202, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.00025177001953125, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.545299053192139, "learning_rate": 1e-05, "loss": 0.4754, "mean_token_accuracy": 0.8800914287567139, "num_tokens": 614048841.0, "step": 16173 }, { "epoch": 2.0574990459229108, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.00025177001953125, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.839564323425293, "learning_rate": 1e-05, "loss": 0.4641, "mean_token_accuracy": 0.8810527324676514, "num_tokens": 614087677.0, "step": 16174 }, { "epoch": 2.057626256201501, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.00025177001953125, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.441222190856934, "learning_rate": 1e-05, "loss": 0.4716, "mean_token_accuracy": 0.8782714605331421, "num_tokens": 614128546.0, "step": 16175 }, { "epoch": 2.0577534664800914, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.00025177001953125, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.591273784637451, "learning_rate": 1e-05, "loss": 0.4486, "mean_token_accuracy": 0.8864244222640991, "num_tokens": 614163947.0, "step": 16176 }, { "epoch": 2.057880676758682, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.00025177001953125, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.321325302124023, "learning_rate": 1e-05, "loss": 0.4742, "mean_token_accuracy": 0.8766491413116455, "num_tokens": 614202353.0, "step": 16177 }, { "epoch": 2.0580078870372724, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.00025177001953125, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.4863176345825195, "learning_rate": 1e-05, "loss": 0.4833, "mean_token_accuracy": 0.8745033144950867, "num_tokens": 614244781.0, "step": 16178 }, { "epoch": 2.058135097315863, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.00025177001953125, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.165171146392822, "learning_rate": 1e-05, "loss": 0.4583, "mean_token_accuracy": 0.8808249235153198, "num_tokens": 614278335.0, "step": 16179 }, { "epoch": 2.0582623075944535, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.00025177001953125, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.498457431793213, "learning_rate": 1e-05, "loss": 0.4892, "mean_token_accuracy": 0.8741616010665894, "num_tokens": 614317726.0, "step": 16180 }, { "epoch": 2.058389517873044, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.00025177001953125, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 6.006292343139648, "learning_rate": 1e-05, "loss": 0.4398, "mean_token_accuracy": 0.8855303525924683, "num_tokens": 614355403.0, "step": 16181 }, { "epoch": 2.0585167281516346, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.00025177001953125, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.604576110839844, "learning_rate": 1e-05, "loss": 0.5126, "mean_token_accuracy": 0.8653424382209778, "num_tokens": 614404370.0, "step": 16182 }, { "epoch": 2.058643938430225, "ewc_loss": 0.109375, "ewc_loss_diag": 0.00025177001953125, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.854411602020264, "learning_rate": 1e-05, "loss": 0.4254, "mean_token_accuracy": 0.8864150047302246, "num_tokens": 614439613.0, "step": 16183 }, { "epoch": 2.0587711487088156, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.00025177001953125, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.834900379180908, "learning_rate": 1e-05, "loss": 0.4958, "mean_token_accuracy": 0.871662974357605, "num_tokens": 614481634.0, "step": 16184 }, { "epoch": 2.058898358987406, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.00025177001953125, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.742563724517822, "learning_rate": 1e-05, "loss": 0.4334, "mean_token_accuracy": 0.8842841386795044, "num_tokens": 614517178.0, "step": 16185 }, { "epoch": 2.0590255692659967, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.00025177001953125, "ewc_loss_parallel": 0.00011205673217773438, "grad_norm": 7.180351734161377, "learning_rate": 1e-05, "loss": 0.509, "mean_token_accuracy": 0.8654223680496216, "num_tokens": 614551531.0, "step": 16186 }, { "epoch": 2.059152779544587, "ewc_loss": 0.11083984375, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.778779029846191, "learning_rate": 1e-05, "loss": 0.4799, "mean_token_accuracy": 0.8722788095474243, "num_tokens": 614589173.0, "step": 16187 }, { "epoch": 2.0592799898231777, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 0.00012063980102539062, "grad_norm": 7.483856678009033, "learning_rate": 1e-05, "loss": 0.4865, "mean_token_accuracy": 0.8771200180053711, "num_tokens": 614633657.0, "step": 16188 }, { "epoch": 2.0594072001017683, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.0982537269592285, "learning_rate": 1e-05, "loss": 0.4762, "mean_token_accuracy": 0.8757270574569702, "num_tokens": 614670013.0, "step": 16189 }, { "epoch": 2.059534410380359, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 0.00011968612670898438, "grad_norm": 7.461272716522217, "learning_rate": 1e-05, "loss": 0.4991, "mean_token_accuracy": 0.8727585077285767, "num_tokens": 614706414.0, "step": 16190 }, { "epoch": 2.0596616206589493, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.455687522888184, "learning_rate": 1e-05, "loss": 0.4875, "mean_token_accuracy": 0.8728867173194885, "num_tokens": 614746397.0, "step": 16191 }, { "epoch": 2.05978883093754, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 0.00010967254638671875, "grad_norm": 7.052464008331299, "learning_rate": 1e-05, "loss": 0.4569, "mean_token_accuracy": 0.8866621255874634, "num_tokens": 614785587.0, "step": 16192 }, { "epoch": 2.0599160412161304, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.582594394683838, "learning_rate": 1e-05, "loss": 0.4762, "mean_token_accuracy": 0.8817143440246582, "num_tokens": 614823304.0, "step": 16193 }, { "epoch": 2.060043251494721, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.539653301239014, "learning_rate": 1e-05, "loss": 0.4769, "mean_token_accuracy": 0.8798809051513672, "num_tokens": 614857544.0, "step": 16194 }, { "epoch": 2.0601704617733114, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.6536078453063965, "learning_rate": 1e-05, "loss": 0.5042, "mean_token_accuracy": 0.874587893486023, "num_tokens": 614893095.0, "step": 16195 }, { "epoch": 2.060297672051902, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.07169246673584, "learning_rate": 1e-05, "loss": 0.4787, "mean_token_accuracy": 0.8738012909889221, "num_tokens": 614932634.0, "step": 16196 }, { "epoch": 2.0604248823304925, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.839139461517334, "learning_rate": 1e-05, "loss": 0.4706, "mean_token_accuracy": 0.8766160011291504, "num_tokens": 614975039.0, "step": 16197 }, { "epoch": 2.060552092609083, "ewc_loss": 0.109375, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.797362804412842, "learning_rate": 1e-05, "loss": 0.4536, "mean_token_accuracy": 0.8796192407608032, "num_tokens": 615019092.0, "step": 16198 }, { "epoch": 2.0606793028876735, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 0.00011301040649414062, "grad_norm": 7.14479923248291, "learning_rate": 1e-05, "loss": 0.4544, "mean_token_accuracy": 0.8884541392326355, "num_tokens": 615060183.0, "step": 16199 }, { "epoch": 2.0608065131662636, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.763127326965332, "learning_rate": 1e-05, "loss": 0.4673, "mean_token_accuracy": 0.8767406940460205, "num_tokens": 615098376.0, "step": 16200 }, { "epoch": 2.060933723444854, "ewc_loss": 0.1455078125, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 0.0001201629638671875, "grad_norm": 7.48288106918335, "learning_rate": 1e-05, "loss": 0.4814, "mean_token_accuracy": 0.8844563364982605, "num_tokens": 615134401.0, "step": 16201 }, { "epoch": 2.0610609337234447, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.017152786254883, "learning_rate": 1e-05, "loss": 0.4637, "mean_token_accuracy": 0.8830089569091797, "num_tokens": 615176145.0, "step": 16202 }, { "epoch": 2.061188144002035, "ewc_loss": 0.1474609375, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 0.0001220703125, "grad_norm": 7.570955753326416, "learning_rate": 1e-05, "loss": 0.5139, "mean_token_accuracy": 0.8716447949409485, "num_tokens": 615210281.0, "step": 16203 }, { "epoch": 2.0613153542806257, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.3708906173706055, "learning_rate": 1e-05, "loss": 0.4275, "mean_token_accuracy": 0.8952395915985107, "num_tokens": 615246952.0, "step": 16204 }, { "epoch": 2.0614425645592163, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 0.0001163482666015625, "grad_norm": 7.310088157653809, "learning_rate": 1e-05, "loss": 0.4557, "mean_token_accuracy": 0.8903316855430603, "num_tokens": 615285807.0, "step": 16205 }, { "epoch": 2.061569774837807, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.505646228790283, "learning_rate": 1e-05, "loss": 0.4333, "mean_token_accuracy": 0.8914910554885864, "num_tokens": 615319641.0, "step": 16206 }, { "epoch": 2.0616969851163973, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.96433687210083, "learning_rate": 1e-05, "loss": 0.5245, "mean_token_accuracy": 0.8660281896591187, "num_tokens": 615358072.0, "step": 16207 }, { "epoch": 2.061824195394988, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.461079120635986, "learning_rate": 1e-05, "loss": 0.4588, "mean_token_accuracy": 0.880682110786438, "num_tokens": 615393924.0, "step": 16208 }, { "epoch": 2.0619514056735784, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.6579766273498535, "learning_rate": 1e-05, "loss": 0.4716, "mean_token_accuracy": 0.8787294626235962, "num_tokens": 615436771.0, "step": 16209 }, { "epoch": 2.062078615952169, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.356671333312988, "learning_rate": 1e-05, "loss": 0.4817, "mean_token_accuracy": 0.8737503290176392, "num_tokens": 615473577.0, "step": 16210 }, { "epoch": 2.0622058262307594, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.510303974151611, "learning_rate": 1e-05, "loss": 0.514, "mean_token_accuracy": 0.8622148633003235, "num_tokens": 615507585.0, "step": 16211 }, { "epoch": 2.06233303650935, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.218991279602051, "learning_rate": 1e-05, "loss": 0.4411, "mean_token_accuracy": 0.8859796524047852, "num_tokens": 615542638.0, "step": 16212 }, { "epoch": 2.0624602467879405, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.474392890930176, "learning_rate": 1e-05, "loss": 0.4855, "mean_token_accuracy": 0.874207615852356, "num_tokens": 615579567.0, "step": 16213 }, { "epoch": 2.062587457066531, "ewc_loss": 0.11376953125, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 6.055459022521973, "learning_rate": 1e-05, "loss": 0.4314, "mean_token_accuracy": 0.8871802091598511, "num_tokens": 615616563.0, "step": 16214 }, { "epoch": 2.0627146673451215, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.566854953765869, "learning_rate": 1e-05, "loss": 0.501, "mean_token_accuracy": 0.8681008815765381, "num_tokens": 615653243.0, "step": 16215 }, { "epoch": 2.062841877623712, "ewc_loss": 0.11083984375, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.887015342712402, "learning_rate": 1e-05, "loss": 0.425, "mean_token_accuracy": 0.8866175413131714, "num_tokens": 615689552.0, "step": 16216 }, { "epoch": 2.0629690879023026, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.782059192657471, "learning_rate": 1e-05, "loss": 0.5062, "mean_token_accuracy": 0.8685544729232788, "num_tokens": 615723803.0, "step": 16217 }, { "epoch": 2.063096298180893, "ewc_loss": 0.10888671875, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.753400802612305, "learning_rate": 1e-05, "loss": 0.4449, "mean_token_accuracy": 0.8811241388320923, "num_tokens": 615761181.0, "step": 16218 }, { "epoch": 2.0632235084594837, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 0.00011157989501953125, "grad_norm": 7.102110385894775, "learning_rate": 1e-05, "loss": 0.4573, "mean_token_accuracy": 0.8869572877883911, "num_tokens": 615802200.0, "step": 16219 }, { "epoch": 2.063350718738074, "ewc_loss": 0.109375, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.753888130187988, "learning_rate": 1e-05, "loss": 0.4386, "mean_token_accuracy": 0.8835428357124329, "num_tokens": 615843299.0, "step": 16220 }, { "epoch": 2.0634779290166647, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 0.00011968612670898438, "grad_norm": 7.442535400390625, "learning_rate": 1e-05, "loss": 0.4959, "mean_token_accuracy": 0.8772697448730469, "num_tokens": 615886148.0, "step": 16221 }, { "epoch": 2.0636051392952552, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.007913589477539, "learning_rate": 1e-05, "loss": 0.4751, "mean_token_accuracy": 0.8719545602798462, "num_tokens": 615930595.0, "step": 16222 }, { "epoch": 2.0637323495738458, "ewc_loss": 0.1474609375, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 0.0001220703125, "grad_norm": 7.538352012634277, "learning_rate": 1e-05, "loss": 0.4906, "mean_token_accuracy": 0.8791791796684265, "num_tokens": 615966873.0, "step": 16223 }, { "epoch": 2.0638595598524363, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.407742977142334, "learning_rate": 1e-05, "loss": 0.4253, "mean_token_accuracy": 0.8928773403167725, "num_tokens": 616002021.0, "step": 16224 }, { "epoch": 2.0639867701310264, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 0.00011396408081054688, "grad_norm": 7.191913604736328, "learning_rate": 1e-05, "loss": 0.486, "mean_token_accuracy": 0.8790661096572876, "num_tokens": 616043692.0, "step": 16225 }, { "epoch": 2.064113980409617, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.595815181732178, "learning_rate": 1e-05, "loss": 0.4508, "mean_token_accuracy": 0.8900443315505981, "num_tokens": 616083811.0, "step": 16226 }, { "epoch": 2.0642411906882074, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.67147970199585, "learning_rate": 1e-05, "loss": 0.4317, "mean_token_accuracy": 0.8915825486183167, "num_tokens": 616121714.0, "step": 16227 }, { "epoch": 2.064368400966798, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.65057897567749, "learning_rate": 1e-05, "loss": 0.4861, "mean_token_accuracy": 0.876403272151947, "num_tokens": 616156571.0, "step": 16228 }, { "epoch": 2.0644956112453885, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.1901984214782715, "learning_rate": 1e-05, "loss": 0.4924, "mean_token_accuracy": 0.8733433485031128, "num_tokens": 616197198.0, "step": 16229 }, { "epoch": 2.064622821523979, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.783015251159668, "learning_rate": 1e-05, "loss": 0.5124, "mean_token_accuracy": 0.8677573204040527, "num_tokens": 616233495.0, "step": 16230 }, { "epoch": 2.0647500318025696, "ewc_loss": 0.11083984375, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.861987590789795, "learning_rate": 1e-05, "loss": 0.4203, "mean_token_accuracy": 0.8911160230636597, "num_tokens": 616268894.0, "step": 16231 }, { "epoch": 2.06487724208116, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 7.057352066040039, "learning_rate": 1e-05, "loss": 0.4861, "mean_token_accuracy": 0.8788864612579346, "num_tokens": 616302899.0, "step": 16232 }, { "epoch": 2.0650044523597506, "ewc_loss": 0.109375, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.756564617156982, "learning_rate": 1e-05, "loss": 0.5151, "mean_token_accuracy": 0.859249472618103, "num_tokens": 616342937.0, "step": 16233 }, { "epoch": 2.065131662638341, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 0.00011873245239257812, "grad_norm": 7.378527641296387, "learning_rate": 1e-05, "loss": 0.4888, "mean_token_accuracy": 0.8799125552177429, "num_tokens": 616391581.0, "step": 16234 }, { "epoch": 2.0652588729169317, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 5.92288064956665, "learning_rate": 1e-05, "loss": 0.4648, "mean_token_accuracy": 0.8812757730484009, "num_tokens": 616430303.0, "step": 16235 }, { "epoch": 2.065386083195522, "ewc_loss": 0.1474609375, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 0.0001220703125, "grad_norm": 7.554423809051514, "learning_rate": 1e-05, "loss": 0.4589, "mean_token_accuracy": 0.8878766298294067, "num_tokens": 616468294.0, "step": 16236 }, { "epoch": 2.0655132934741127, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.280735015869141, "learning_rate": 1e-05, "loss": 0.4518, "mean_token_accuracy": 0.8815699815750122, "num_tokens": 616503551.0, "step": 16237 }, { "epoch": 2.0656405037527032, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 0.000118255615234375, "grad_norm": 7.400639533996582, "learning_rate": 1e-05, "loss": 0.466, "mean_token_accuracy": 0.8893132209777832, "num_tokens": 616541374.0, "step": 16238 }, { "epoch": 2.065767714031294, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.475857257843018, "learning_rate": 1e-05, "loss": 0.467, "mean_token_accuracy": 0.8820438385009766, "num_tokens": 616575480.0, "step": 16239 }, { "epoch": 2.0658949243098843, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 0.00011110305786132812, "grad_norm": 7.069950580596924, "learning_rate": 1e-05, "loss": 0.4927, "mean_token_accuracy": 0.8773788809776306, "num_tokens": 616620323.0, "step": 16240 }, { "epoch": 2.066022134588475, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.464606285095215, "learning_rate": 1e-05, "loss": 0.439, "mean_token_accuracy": 0.8922522068023682, "num_tokens": 616659843.0, "step": 16241 }, { "epoch": 2.0661493448670654, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.769069194793701, "learning_rate": 1e-05, "loss": 0.4722, "mean_token_accuracy": 0.8821791410446167, "num_tokens": 616696065.0, "step": 16242 }, { "epoch": 2.066276555145656, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.372311592102051, "learning_rate": 1e-05, "loss": 0.4664, "mean_token_accuracy": 0.8838253617286682, "num_tokens": 616729444.0, "step": 16243 }, { "epoch": 2.0664037654242464, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.579995155334473, "learning_rate": 1e-05, "loss": 0.4733, "mean_token_accuracy": 0.8787656426429749, "num_tokens": 616773351.0, "step": 16244 }, { "epoch": 2.066530975702837, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.244744300842285, "learning_rate": 1e-05, "loss": 0.4686, "mean_token_accuracy": 0.8788081407546997, "num_tokens": 616812161.0, "step": 16245 }, { "epoch": 2.0666581859814275, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.5251336097717285, "learning_rate": 1e-05, "loss": 0.5402, "mean_token_accuracy": 0.8636040687561035, "num_tokens": 616847623.0, "step": 16246 }, { "epoch": 2.066785396260018, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.095621585845947, "learning_rate": 1e-05, "loss": 0.4638, "mean_token_accuracy": 0.8773741722106934, "num_tokens": 616884078.0, "step": 16247 }, { "epoch": 2.0669126065386085, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.570163249969482, "learning_rate": 1e-05, "loss": 0.4699, "mean_token_accuracy": 0.8785583972930908, "num_tokens": 616924291.0, "step": 16248 }, { "epoch": 2.0670398168171986, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.934218406677246, "learning_rate": 1e-05, "loss": 0.4372, "mean_token_accuracy": 0.8854429721832275, "num_tokens": 616964580.0, "step": 16249 }, { "epoch": 2.067167027095789, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.743465900421143, "learning_rate": 1e-05, "loss": 0.441, "mean_token_accuracy": 0.8892645835876465, "num_tokens": 617000574.0, "step": 16250 }, { "epoch": 2.0672942373743797, "ewc_loss": 0.109375, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.793632507324219, "learning_rate": 1e-05, "loss": 0.4528, "mean_token_accuracy": 0.8784207105636597, "num_tokens": 617043323.0, "step": 16251 }, { "epoch": 2.06742144765297, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 7.031824111938477, "learning_rate": 1e-05, "loss": 0.4644, "mean_token_accuracy": 0.8839161396026611, "num_tokens": 617082585.0, "step": 16252 }, { "epoch": 2.0675486579315607, "ewc_loss": 0.109375, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.773913383483887, "learning_rate": 1e-05, "loss": 0.4397, "mean_token_accuracy": 0.8821336627006531, "num_tokens": 617118384.0, "step": 16253 }, { "epoch": 2.0676758682101513, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 0.00011730194091796875, "grad_norm": 7.344899654388428, "learning_rate": 1e-05, "loss": 0.4406, "mean_token_accuracy": 0.8951777219772339, "num_tokens": 617153354.0, "step": 16254 }, { "epoch": 2.067803078488742, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 5.976354598999023, "learning_rate": 1e-05, "loss": 0.4636, "mean_token_accuracy": 0.8811362385749817, "num_tokens": 617192829.0, "step": 16255 }, { "epoch": 2.0679302887673323, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 0.00011873245239257812, "grad_norm": 7.445672035217285, "learning_rate": 1e-05, "loss": 0.4848, "mean_token_accuracy": 0.8808358311653137, "num_tokens": 617230824.0, "step": 16256 }, { "epoch": 2.068057499045923, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.3293962478637695, "learning_rate": 1e-05, "loss": 0.5001, "mean_token_accuracy": 0.8695536851882935, "num_tokens": 617268791.0, "step": 16257 }, { "epoch": 2.0681847093245134, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 0.0001125335693359375, "grad_norm": 7.144367218017578, "learning_rate": 1e-05, "loss": 0.4832, "mean_token_accuracy": 0.8777901530265808, "num_tokens": 617305366.0, "step": 16258 }, { "epoch": 2.068311919603104, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.534299373626709, "learning_rate": 1e-05, "loss": 0.4864, "mean_token_accuracy": 0.8778793811798096, "num_tokens": 617342294.0, "step": 16259 }, { "epoch": 2.0684391298816944, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.632645130157471, "learning_rate": 1e-05, "loss": 0.4733, "mean_token_accuracy": 0.8807092308998108, "num_tokens": 617380456.0, "step": 16260 }, { "epoch": 2.068566340160285, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.625213146209717, "learning_rate": 1e-05, "loss": 0.4491, "mean_token_accuracy": 0.8857638239860535, "num_tokens": 617420705.0, "step": 16261 }, { "epoch": 2.0686935504388755, "ewc_loss": 0.11376953125, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 6.11706018447876, "learning_rate": 1e-05, "loss": 0.4401, "mean_token_accuracy": 0.8840386867523193, "num_tokens": 617462740.0, "step": 16262 }, { "epoch": 2.068820760717466, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.833159446716309, "learning_rate": 1e-05, "loss": 0.4448, "mean_token_accuracy": 0.8874218463897705, "num_tokens": 617497496.0, "step": 16263 }, { "epoch": 2.0689479709960565, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.769538402557373, "learning_rate": 1e-05, "loss": 0.4167, "mean_token_accuracy": 0.8893513679504395, "num_tokens": 617535297.0, "step": 16264 }, { "epoch": 2.069075181274647, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 0.00011301040649414062, "grad_norm": 7.197634696960449, "learning_rate": 1e-05, "loss": 0.4916, "mean_token_accuracy": 0.8780182003974915, "num_tokens": 617568332.0, "step": 16265 }, { "epoch": 2.0692023915532376, "ewc_loss": 0.109375, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.6997880935668945, "learning_rate": 1e-05, "loss": 0.4608, "mean_token_accuracy": 0.8760879635810852, "num_tokens": 617608934.0, "step": 16266 }, { "epoch": 2.069329601831828, "ewc_loss": 0.1474609375, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 0.0001220703125, "grad_norm": 7.647200107574463, "learning_rate": 1e-05, "loss": 0.5122, "mean_token_accuracy": 0.8699455261230469, "num_tokens": 617643794.0, "step": 16267 }, { "epoch": 2.0694568121104187, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.076478481292725, "learning_rate": 1e-05, "loss": 0.4539, "mean_token_accuracy": 0.8838231563568115, "num_tokens": 617676448.0, "step": 16268 }, { "epoch": 2.069584022389009, "ewc_loss": 0.15234375, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 0.00012683868408203125, "grad_norm": 7.728994846343994, "learning_rate": 1e-05, "loss": 0.5724, "mean_token_accuracy": 0.8550540208816528, "num_tokens": 617713460.0, "step": 16269 }, { "epoch": 2.0697112326675997, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.53399133682251, "learning_rate": 1e-05, "loss": 0.4984, "mean_token_accuracy": 0.8714913129806519, "num_tokens": 617749767.0, "step": 16270 }, { "epoch": 2.0698384429461902, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 0.000118255615234375, "grad_norm": 7.388571262359619, "learning_rate": 1e-05, "loss": 0.5364, "mean_token_accuracy": 0.8634290099143982, "num_tokens": 617784991.0, "step": 16271 }, { "epoch": 2.0699656532247808, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 6.640201568603516, "learning_rate": 1e-05, "loss": 0.445, "mean_token_accuracy": 0.8881075978279114, "num_tokens": 617820241.0, "step": 16272 }, { "epoch": 2.070092863503371, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.980154037475586, "learning_rate": 1e-05, "loss": 0.508, "mean_token_accuracy": 0.8676697015762329, "num_tokens": 617861838.0, "step": 16273 }, { "epoch": 2.0702200737819614, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.535903453826904, "learning_rate": 1e-05, "loss": 0.4646, "mean_token_accuracy": 0.8786565661430359, "num_tokens": 617893001.0, "step": 16274 }, { "epoch": 2.070347284060552, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.691610813140869, "learning_rate": 1e-05, "loss": 0.4833, "mean_token_accuracy": 0.8741910457611084, "num_tokens": 617936361.0, "step": 16275 }, { "epoch": 2.0704744943391424, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.377196788787842, "learning_rate": 1e-05, "loss": 0.4791, "mean_token_accuracy": 0.8736923933029175, "num_tokens": 617968267.0, "step": 16276 }, { "epoch": 2.070601704617733, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.571030139923096, "learning_rate": 1e-05, "loss": 0.4762, "mean_token_accuracy": 0.8774563670158386, "num_tokens": 618007710.0, "step": 16277 }, { "epoch": 2.0707289148963235, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.194224834442139, "learning_rate": 1e-05, "loss": 0.456, "mean_token_accuracy": 0.8798385858535767, "num_tokens": 618043185.0, "step": 16278 }, { "epoch": 2.070856125174914, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.584079265594482, "learning_rate": 1e-05, "loss": 0.4357, "mean_token_accuracy": 0.8889052867889404, "num_tokens": 618085460.0, "step": 16279 }, { "epoch": 2.0709833354535045, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 6.005395889282227, "learning_rate": 1e-05, "loss": 0.4488, "mean_token_accuracy": 0.8871728777885437, "num_tokens": 618117155.0, "step": 16280 }, { "epoch": 2.071110545732095, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.722378253936768, "learning_rate": 1e-05, "loss": 0.4577, "mean_token_accuracy": 0.8863006830215454, "num_tokens": 618153380.0, "step": 16281 }, { "epoch": 2.0712377560106856, "ewc_loss": 0.109375, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.8472161293029785, "learning_rate": 1e-05, "loss": 0.4566, "mean_token_accuracy": 0.8797958493232727, "num_tokens": 618190125.0, "step": 16282 }, { "epoch": 2.071364966289276, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.984800338745117, "learning_rate": 1e-05, "loss": 0.4702, "mean_token_accuracy": 0.8805631399154663, "num_tokens": 618225281.0, "step": 16283 }, { "epoch": 2.0714921765678667, "ewc_loss": 0.109375, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.777032852172852, "learning_rate": 1e-05, "loss": 0.446, "mean_token_accuracy": 0.8840982913970947, "num_tokens": 618262213.0, "step": 16284 }, { "epoch": 2.071619386846457, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 0.00011682510375976562, "grad_norm": 7.326803684234619, "learning_rate": 1e-05, "loss": 0.5409, "mean_token_accuracy": 0.8587501049041748, "num_tokens": 618292780.0, "step": 16285 }, { "epoch": 2.0717465971250477, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 5.930517196655273, "learning_rate": 1e-05, "loss": 0.5083, "mean_token_accuracy": 0.8633884191513062, "num_tokens": 618332968.0, "step": 16286 }, { "epoch": 2.0718738074036382, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 0.00012063980102539062, "grad_norm": 7.493201732635498, "learning_rate": 1e-05, "loss": 0.4537, "mean_token_accuracy": 0.8886021375656128, "num_tokens": 618369585.0, "step": 16287 }, { "epoch": 2.0720010176822288, "ewc_loss": 0.125, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.281896591186523, "learning_rate": 1e-05, "loss": 0.4238, "mean_token_accuracy": 0.8941276669502258, "num_tokens": 618410388.0, "step": 16288 }, { "epoch": 2.0721282279608193, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 0.0001163482666015625, "grad_norm": 7.279904842376709, "learning_rate": 1e-05, "loss": 0.4395, "mean_token_accuracy": 0.8948078155517578, "num_tokens": 618453591.0, "step": 16289 }, { "epoch": 2.07225543823941, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.5416259765625, "learning_rate": 1e-05, "loss": 0.4624, "mean_token_accuracy": 0.8836002349853516, "num_tokens": 618489757.0, "step": 16290 }, { "epoch": 2.0723826485180004, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.794299125671387, "learning_rate": 1e-05, "loss": 0.451, "mean_token_accuracy": 0.8849362730979919, "num_tokens": 618524733.0, "step": 16291 }, { "epoch": 2.072509858796591, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.6098504066467285, "learning_rate": 1e-05, "loss": 0.4948, "mean_token_accuracy": 0.8740957975387573, "num_tokens": 618564734.0, "step": 16292 }, { "epoch": 2.0726370690751814, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.291534423828125, "learning_rate": 1e-05, "loss": 0.4542, "mean_token_accuracy": 0.8823628425598145, "num_tokens": 618602503.0, "step": 16293 }, { "epoch": 2.072764279353772, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.728790283203125, "learning_rate": 1e-05, "loss": 0.4347, "mean_token_accuracy": 0.8909298777580261, "num_tokens": 618639907.0, "step": 16294 }, { "epoch": 2.0728914896323625, "ewc_loss": 0.11083984375, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.902184963226318, "learning_rate": 1e-05, "loss": 0.4442, "mean_token_accuracy": 0.8840519189834595, "num_tokens": 618681366.0, "step": 16295 }, { "epoch": 2.073018699910953, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.988025188446045, "learning_rate": 1e-05, "loss": 0.4225, "mean_token_accuracy": 0.8952608108520508, "num_tokens": 618718610.0, "step": 16296 }, { "epoch": 2.0731459101895435, "ewc_loss": 0.10791015625, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.708651065826416, "learning_rate": 1e-05, "loss": 0.4298, "mean_token_accuracy": 0.8850314617156982, "num_tokens": 618757741.0, "step": 16297 }, { "epoch": 2.0732731204681336, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 0.000118255615234375, "grad_norm": 7.385797023773193, "learning_rate": 1e-05, "loss": 0.4612, "mean_token_accuracy": 0.8867311477661133, "num_tokens": 618798189.0, "step": 16298 }, { "epoch": 2.073400330746724, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 5.840394020080566, "learning_rate": 1e-05, "loss": 0.4664, "mean_token_accuracy": 0.8786801099777222, "num_tokens": 618837308.0, "step": 16299 }, { "epoch": 2.0735275410253147, "ewc_loss": 0.1513671875, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 0.000125885009765625, "grad_norm": 7.674299240112305, "learning_rate": 1e-05, "loss": 0.5421, "mean_token_accuracy": 0.8649086952209473, "num_tokens": 618876058.0, "step": 16300 }, { "epoch": 2.073654751303905, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.2638750076293945, "learning_rate": 1e-05, "loss": 0.5044, "mean_token_accuracy": 0.869988739490509, "num_tokens": 618916246.0, "step": 16301 }, { "epoch": 2.0737819615824957, "ewc_loss": 0.1474609375, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 0.0001220703125, "grad_norm": 7.562684059143066, "learning_rate": 1e-05, "loss": 0.5113, "mean_token_accuracy": 0.8720483183860779, "num_tokens": 618957639.0, "step": 16302 }, { "epoch": 2.0739091718610863, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.57322883605957, "learning_rate": 1e-05, "loss": 0.4555, "mean_token_accuracy": 0.8887128829956055, "num_tokens": 618999505.0, "step": 16303 }, { "epoch": 2.074036382139677, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 0.00011396408081054688, "grad_norm": 7.182115077972412, "learning_rate": 1e-05, "loss": 0.4608, "mean_token_accuracy": 0.8874694108963013, "num_tokens": 619034807.0, "step": 16304 }, { "epoch": 2.0741635924182673, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.587163925170898, "learning_rate": 1e-05, "loss": 0.448, "mean_token_accuracy": 0.8889108896255493, "num_tokens": 619073262.0, "step": 16305 }, { "epoch": 2.074290802696858, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.814155101776123, "learning_rate": 1e-05, "loss": 0.4956, "mean_token_accuracy": 0.8705956339836121, "num_tokens": 619112706.0, "step": 16306 }, { "epoch": 2.0744180129754484, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.501524925231934, "learning_rate": 1e-05, "loss": 0.4932, "mean_token_accuracy": 0.8717424273490906, "num_tokens": 619144444.0, "step": 16307 }, { "epoch": 2.074545223254039, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.529260635375977, "learning_rate": 1e-05, "loss": 0.4391, "mean_token_accuracy": 0.8886786699295044, "num_tokens": 619183291.0, "step": 16308 }, { "epoch": 2.0746724335326294, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.371856689453125, "learning_rate": 1e-05, "loss": 0.4441, "mean_token_accuracy": 0.8870277404785156, "num_tokens": 619223865.0, "step": 16309 }, { "epoch": 2.07479964381122, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.416300296783447, "learning_rate": 1e-05, "loss": 0.4332, "mean_token_accuracy": 0.8876008987426758, "num_tokens": 619257418.0, "step": 16310 }, { "epoch": 2.0749268540898105, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.227221488952637, "learning_rate": 1e-05, "loss": 0.4927, "mean_token_accuracy": 0.87348473072052, "num_tokens": 619298724.0, "step": 16311 }, { "epoch": 2.075054064368401, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.419168949127197, "learning_rate": 1e-05, "loss": 0.4727, "mean_token_accuracy": 0.8814902901649475, "num_tokens": 619332497.0, "step": 16312 }, { "epoch": 2.0751812746469915, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.061299800872803, "learning_rate": 1e-05, "loss": 0.4472, "mean_token_accuracy": 0.8812532424926758, "num_tokens": 619363371.0, "step": 16313 }, { "epoch": 2.075308484925582, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.515982627868652, "learning_rate": 1e-05, "loss": 0.4492, "mean_token_accuracy": 0.8840146064758301, "num_tokens": 619396068.0, "step": 16314 }, { "epoch": 2.0754356952041726, "ewc_loss": 0.10986328125, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.891201496124268, "learning_rate": 1e-05, "loss": 0.464, "mean_token_accuracy": 0.877951979637146, "num_tokens": 619433728.0, "step": 16315 }, { "epoch": 2.075562905482763, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.722044944763184, "learning_rate": 1e-05, "loss": 0.4929, "mean_token_accuracy": 0.8699464797973633, "num_tokens": 619476185.0, "step": 16316 }, { "epoch": 2.0756901157613536, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.707123279571533, "learning_rate": 1e-05, "loss": 0.3913, "mean_token_accuracy": 0.8980976343154907, "num_tokens": 619513184.0, "step": 16317 }, { "epoch": 2.075817326039944, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 7.100666522979736, "learning_rate": 1e-05, "loss": 0.4974, "mean_token_accuracy": 0.8736136555671692, "num_tokens": 619550168.0, "step": 16318 }, { "epoch": 2.0759445363185347, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.656787872314453, "learning_rate": 1e-05, "loss": 0.4749, "mean_token_accuracy": 0.8751254677772522, "num_tokens": 619588960.0, "step": 16319 }, { "epoch": 2.0760717465971252, "ewc_loss": 0.150390625, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 0.00012493133544921875, "grad_norm": 7.57472562789917, "learning_rate": 1e-05, "loss": 0.4866, "mean_token_accuracy": 0.8797308206558228, "num_tokens": 619624907.0, "step": 16320 }, { "epoch": 2.0761989568757158, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 5.954729080200195, "learning_rate": 1e-05, "loss": 0.4499, "mean_token_accuracy": 0.884781002998352, "num_tokens": 619670576.0, "step": 16321 }, { "epoch": 2.0763261671543063, "ewc_loss": 0.1533203125, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 0.0001277923583984375, "grad_norm": 7.790954113006592, "learning_rate": 1e-05, "loss": 0.4721, "mean_token_accuracy": 0.8870567083358765, "num_tokens": 619710452.0, "step": 16322 }, { "epoch": 2.0764533774328964, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 6.5258684158325195, "learning_rate": 1e-05, "loss": 0.4924, "mean_token_accuracy": 0.8806172609329224, "num_tokens": 619750104.0, "step": 16323 }, { "epoch": 2.076580587711487, "ewc_loss": 0.1455078125, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 0.00011968612670898438, "grad_norm": 7.426900386810303, "learning_rate": 1e-05, "loss": 0.4774, "mean_token_accuracy": 0.8840367197990417, "num_tokens": 619791364.0, "step": 16324 }, { "epoch": 2.0767077979900774, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 0.00011157989501953125, "grad_norm": 6.743004322052002, "learning_rate": 1e-05, "loss": 0.5024, "mean_token_accuracy": 0.8754739761352539, "num_tokens": 619831138.0, "step": 16325 }, { "epoch": 2.076835008268668, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.897140979766846, "learning_rate": 1e-05, "loss": 0.465, "mean_token_accuracy": 0.8835701942443848, "num_tokens": 619862897.0, "step": 16326 }, { "epoch": 2.0769622185472585, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.697587013244629, "learning_rate": 1e-05, "loss": 0.4478, "mean_token_accuracy": 0.8896432518959045, "num_tokens": 619897096.0, "step": 16327 }, { "epoch": 2.077089428825849, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.411205768585205, "learning_rate": 1e-05, "loss": 0.4289, "mean_token_accuracy": 0.892120361328125, "num_tokens": 619937458.0, "step": 16328 }, { "epoch": 2.0772166391044395, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.7104034423828125, "learning_rate": 1e-05, "loss": 0.5157, "mean_token_accuracy": 0.8669930100440979, "num_tokens": 619974735.0, "step": 16329 }, { "epoch": 2.07734384938303, "ewc_loss": 0.11474609375, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 6.089283466339111, "learning_rate": 1e-05, "loss": 0.4734, "mean_token_accuracy": 0.874807596206665, "num_tokens": 620012497.0, "step": 16330 }, { "epoch": 2.0774710596616206, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.796351432800293, "learning_rate": 1e-05, "loss": 0.483, "mean_token_accuracy": 0.8741547465324402, "num_tokens": 620051120.0, "step": 16331 }, { "epoch": 2.077598269940211, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.90089225769043, "learning_rate": 1e-05, "loss": 0.4509, "mean_token_accuracy": 0.8805631399154663, "num_tokens": 620091479.0, "step": 16332 }, { "epoch": 2.0777254802188017, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 6.970669269561768, "learning_rate": 1e-05, "loss": 0.4774, "mean_token_accuracy": 0.8807889223098755, "num_tokens": 620126509.0, "step": 16333 }, { "epoch": 2.077852690497392, "ewc_loss": 0.11083984375, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.829928874969482, "learning_rate": 1e-05, "loss": 0.4483, "mean_token_accuracy": 0.882278323173523, "num_tokens": 620164278.0, "step": 16334 }, { "epoch": 2.0779799007759827, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 0.00011348724365234375, "grad_norm": 7.209789752960205, "learning_rate": 1e-05, "loss": 0.5075, "mean_token_accuracy": 0.8717914819717407, "num_tokens": 620202388.0, "step": 16335 }, { "epoch": 2.0781071110545732, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 5.898508548736572, "learning_rate": 1e-05, "loss": 0.4472, "mean_token_accuracy": 0.8831386566162109, "num_tokens": 620236981.0, "step": 16336 }, { "epoch": 2.0782343213331638, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 0.000118255615234375, "grad_norm": 7.378229141235352, "learning_rate": 1e-05, "loss": 0.4853, "mean_token_accuracy": 0.8832242488861084, "num_tokens": 620275946.0, "step": 16337 }, { "epoch": 2.0783615316117543, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.10503625869751, "learning_rate": 1e-05, "loss": 0.5089, "mean_token_accuracy": 0.8647541999816895, "num_tokens": 620317385.0, "step": 16338 }, { "epoch": 2.078488741890345, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 0.00011730194091796875, "grad_norm": 7.353318691253662, "learning_rate": 1e-05, "loss": 0.4627, "mean_token_accuracy": 0.8868427872657776, "num_tokens": 620353625.0, "step": 16339 }, { "epoch": 2.0786159521689354, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.317491054534912, "learning_rate": 1e-05, "loss": 0.5392, "mean_token_accuracy": 0.8624979853630066, "num_tokens": 620389421.0, "step": 16340 }, { "epoch": 2.078743162447526, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 0.00011205673217773438, "grad_norm": 7.129523277282715, "learning_rate": 1e-05, "loss": 0.4705, "mean_token_accuracy": 0.88434237241745, "num_tokens": 620428233.0, "step": 16341 }, { "epoch": 2.0788703727261164, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.391295909881592, "learning_rate": 1e-05, "loss": 0.5471, "mean_token_accuracy": 0.8606959581375122, "num_tokens": 620465156.0, "step": 16342 }, { "epoch": 2.078997583004707, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.821169376373291, "learning_rate": 1e-05, "loss": 0.4886, "mean_token_accuracy": 0.8773987293243408, "num_tokens": 620501731.0, "step": 16343 }, { "epoch": 2.0791247932832975, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.375799655914307, "learning_rate": 1e-05, "loss": 0.4811, "mean_token_accuracy": 0.8735476136207581, "num_tokens": 620538507.0, "step": 16344 }, { "epoch": 2.079252003561888, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.54841947555542, "learning_rate": 1e-05, "loss": 0.4178, "mean_token_accuracy": 0.8938814401626587, "num_tokens": 620577355.0, "step": 16345 }, { "epoch": 2.0793792138404785, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.320882320404053, "learning_rate": 1e-05, "loss": 0.4809, "mean_token_accuracy": 0.8754671812057495, "num_tokens": 620613320.0, "step": 16346 }, { "epoch": 2.0795064241190686, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.377473831176758, "learning_rate": 1e-05, "loss": 0.4207, "mean_token_accuracy": 0.8911476135253906, "num_tokens": 620655167.0, "step": 16347 }, { "epoch": 2.079633634397659, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.2555460929870605, "learning_rate": 1e-05, "loss": 0.5055, "mean_token_accuracy": 0.8632081747055054, "num_tokens": 620697898.0, "step": 16348 }, { "epoch": 2.0797608446762497, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.301782608032227, "learning_rate": 1e-05, "loss": 0.4514, "mean_token_accuracy": 0.8848105669021606, "num_tokens": 620736548.0, "step": 16349 }, { "epoch": 2.07988805495484, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.163138389587402, "learning_rate": 1e-05, "loss": 0.4997, "mean_token_accuracy": 0.8688369989395142, "num_tokens": 620781368.0, "step": 16350 }, { "epoch": 2.0800152652334307, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.293863773345947, "learning_rate": 1e-05, "loss": 0.459, "mean_token_accuracy": 0.8798180818557739, "num_tokens": 620821749.0, "step": 16351 }, { "epoch": 2.0801424755120212, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.065323829650879, "learning_rate": 1e-05, "loss": 0.4754, "mean_token_accuracy": 0.8755413293838501, "num_tokens": 620857906.0, "step": 16352 }, { "epoch": 2.0802696857906118, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.3609232902526855, "learning_rate": 1e-05, "loss": 0.4668, "mean_token_accuracy": 0.8803749084472656, "num_tokens": 620897124.0, "step": 16353 }, { "epoch": 2.0803968960692023, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 5.941574573516846, "learning_rate": 1e-05, "loss": 0.5048, "mean_token_accuracy": 0.867851197719574, "num_tokens": 620932123.0, "step": 16354 }, { "epoch": 2.080524106347793, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.50946569442749, "learning_rate": 1e-05, "loss": 0.4717, "mean_token_accuracy": 0.8797338008880615, "num_tokens": 620970906.0, "step": 16355 }, { "epoch": 2.0806513166263834, "ewc_loss": 0.10986328125, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.794571399688721, "learning_rate": 1e-05, "loss": 0.4845, "mean_token_accuracy": 0.8702144622802734, "num_tokens": 621007387.0, "step": 16356 }, { "epoch": 2.080778526904974, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.752748489379883, "learning_rate": 1e-05, "loss": 0.4699, "mean_token_accuracy": 0.8805743455886841, "num_tokens": 621043723.0, "step": 16357 }, { "epoch": 2.0809057371835644, "ewc_loss": 0.10791015625, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.621192932128906, "learning_rate": 1e-05, "loss": 0.4233, "mean_token_accuracy": 0.8880504369735718, "num_tokens": 621083686.0, "step": 16358 }, { "epoch": 2.081032947462155, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 0.00011396408081054688, "grad_norm": 7.191226005554199, "learning_rate": 1e-05, "loss": 0.4804, "mean_token_accuracy": 0.879734992980957, "num_tokens": 621125758.0, "step": 16359 }, { "epoch": 2.0811601577407455, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.588469982147217, "learning_rate": 1e-05, "loss": 0.458, "mean_token_accuracy": 0.880562424659729, "num_tokens": 621163407.0, "step": 16360 }, { "epoch": 2.081287368019336, "ewc_loss": 0.1552734375, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 0.00012969970703125, "grad_norm": 7.760567665100098, "learning_rate": 1e-05, "loss": 0.5227, "mean_token_accuracy": 0.8712026476860046, "num_tokens": 621203321.0, "step": 16361 }, { "epoch": 2.0814145782979265, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.022458553314209, "learning_rate": 1e-05, "loss": 0.4699, "mean_token_accuracy": 0.874817430973053, "num_tokens": 621239777.0, "step": 16362 }, { "epoch": 2.081541788576517, "ewc_loss": 0.1591796875, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 0.000133514404296875, "grad_norm": 7.9399943351745605, "learning_rate": 1e-05, "loss": 0.4845, "mean_token_accuracy": 0.8845024108886719, "num_tokens": 621271841.0, "step": 16363 }, { "epoch": 2.0816689988551076, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 0.00011396408081054688, "grad_norm": 6.65513277053833, "learning_rate": 1e-05, "loss": 0.4836, "mean_token_accuracy": 0.8812496662139893, "num_tokens": 621309393.0, "step": 16364 }, { "epoch": 2.081796209133698, "ewc_loss": 0.1484375, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.535886764526367, "learning_rate": 1e-05, "loss": 0.5359, "mean_token_accuracy": 0.8661578893661499, "num_tokens": 621344362.0, "step": 16365 }, { "epoch": 2.0819234194122886, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 0.00011301040649414062, "grad_norm": 6.79585599899292, "learning_rate": 1e-05, "loss": 0.5319, "mean_token_accuracy": 0.8611265420913696, "num_tokens": 621383999.0, "step": 16366 }, { "epoch": 2.082050629690879, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 0.00011014938354492188, "grad_norm": 7.045596599578857, "learning_rate": 1e-05, "loss": 0.4568, "mean_token_accuracy": 0.8856353759765625, "num_tokens": 621424475.0, "step": 16367 }, { "epoch": 2.0821778399694697, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 6.667938709259033, "learning_rate": 1e-05, "loss": 0.5121, "mean_token_accuracy": 0.8664004802703857, "num_tokens": 621458241.0, "step": 16368 }, { "epoch": 2.0823050502480602, "ewc_loss": 0.125, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.669050693511963, "learning_rate": 1e-05, "loss": 0.4424, "mean_token_accuracy": 0.8871523141860962, "num_tokens": 621495424.0, "step": 16369 }, { "epoch": 2.0824322605266508, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.5358357429504395, "learning_rate": 1e-05, "loss": 0.4706, "mean_token_accuracy": 0.8805003762245178, "num_tokens": 621531256.0, "step": 16370 }, { "epoch": 2.082559470805241, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.416789531707764, "learning_rate": 1e-05, "loss": 0.4596, "mean_token_accuracy": 0.8814375996589661, "num_tokens": 621571665.0, "step": 16371 }, { "epoch": 2.0826866810838314, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.4273457527160645, "learning_rate": 1e-05, "loss": 0.4602, "mean_token_accuracy": 0.8842836618423462, "num_tokens": 621604552.0, "step": 16372 }, { "epoch": 2.082813891362422, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.2692766189575195, "learning_rate": 1e-05, "loss": 0.5193, "mean_token_accuracy": 0.8600870370864868, "num_tokens": 621651677.0, "step": 16373 }, { "epoch": 2.0829411016410124, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.335212707519531, "learning_rate": 1e-05, "loss": 0.4408, "mean_token_accuracy": 0.8865261673927307, "num_tokens": 621695108.0, "step": 16374 }, { "epoch": 2.083068311919603, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.200912952423096, "learning_rate": 1e-05, "loss": 0.4328, "mean_token_accuracy": 0.8885849118232727, "num_tokens": 621725768.0, "step": 16375 }, { "epoch": 2.0831955221981935, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.25168514251709, "learning_rate": 1e-05, "loss": 0.4902, "mean_token_accuracy": 0.8693625330924988, "num_tokens": 621768559.0, "step": 16376 }, { "epoch": 2.083322732476784, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.167195796966553, "learning_rate": 1e-05, "loss": 0.4697, "mean_token_accuracy": 0.8782449960708618, "num_tokens": 621805464.0, "step": 16377 }, { "epoch": 2.0834499427553745, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.2024993896484375, "learning_rate": 1e-05, "loss": 0.496, "mean_token_accuracy": 0.8692519664764404, "num_tokens": 621843129.0, "step": 16378 }, { "epoch": 2.083577153033965, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.132334232330322, "learning_rate": 1e-05, "loss": 0.4515, "mean_token_accuracy": 0.8798031210899353, "num_tokens": 621882450.0, "step": 16379 }, { "epoch": 2.0837043633125556, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.16424036026001, "learning_rate": 1e-05, "loss": 0.4585, "mean_token_accuracy": 0.8766952753067017, "num_tokens": 621915797.0, "step": 16380 }, { "epoch": 2.083831573591146, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.125066757202148, "learning_rate": 1e-05, "loss": 0.4804, "mean_token_accuracy": 0.8700313568115234, "num_tokens": 621953563.0, "step": 16381 }, { "epoch": 2.0839587838697367, "ewc_loss": 0.11474609375, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 6.128106594085693, "learning_rate": 1e-05, "loss": 0.4248, "mean_token_accuracy": 0.891309380531311, "num_tokens": 621989579.0, "step": 16382 }, { "epoch": 2.084085994148327, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.116952419281006, "learning_rate": 1e-05, "loss": 0.4153, "mean_token_accuracy": 0.8926708698272705, "num_tokens": 622027446.0, "step": 16383 }, { "epoch": 2.0842132044269177, "ewc_loss": 0.11474609375, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 6.122377395629883, "learning_rate": 1e-05, "loss": 0.492, "mean_token_accuracy": 0.8674749135971069, "num_tokens": 622064972.0, "step": 16384 }, { "epoch": 2.0843404147055082, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.107053279876709, "learning_rate": 1e-05, "loss": 0.4261, "mean_token_accuracy": 0.8916044235229492, "num_tokens": 622103260.0, "step": 16385 }, { "epoch": 2.0844676249840988, "ewc_loss": 0.11474609375, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 6.1209540367126465, "learning_rate": 1e-05, "loss": 0.4475, "mean_token_accuracy": 0.883021354675293, "num_tokens": 622140894.0, "step": 16386 }, { "epoch": 2.0845948352626893, "ewc_loss": 0.11474609375, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 6.092442989349365, "learning_rate": 1e-05, "loss": 0.4411, "mean_token_accuracy": 0.8860260248184204, "num_tokens": 622177537.0, "step": 16387 }, { "epoch": 2.08472204554128, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.132417678833008, "learning_rate": 1e-05, "loss": 0.4583, "mean_token_accuracy": 0.8805490732192993, "num_tokens": 622211172.0, "step": 16388 }, { "epoch": 2.0848492558198704, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.076306343078613, "learning_rate": 1e-05, "loss": 0.4706, "mean_token_accuracy": 0.876104474067688, "num_tokens": 622244805.0, "step": 16389 }, { "epoch": 2.084976466098461, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.1502485275268555, "learning_rate": 1e-05, "loss": 0.4275, "mean_token_accuracy": 0.8902366161346436, "num_tokens": 622279496.0, "step": 16390 }, { "epoch": 2.0851036763770514, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 6.009485721588135, "learning_rate": 1e-05, "loss": 0.4352, "mean_token_accuracy": 0.8870911598205566, "num_tokens": 622318818.0, "step": 16391 }, { "epoch": 2.085230886655642, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.210945129394531, "learning_rate": 1e-05, "loss": 0.4445, "mean_token_accuracy": 0.8833186030387878, "num_tokens": 622359313.0, "step": 16392 }, { "epoch": 2.0853580969342325, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 5.933475017547607, "learning_rate": 1e-05, "loss": 0.4691, "mean_token_accuracy": 0.8745549917221069, "num_tokens": 622392840.0, "step": 16393 }, { "epoch": 2.085485307212823, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.324033737182617, "learning_rate": 1e-05, "loss": 0.5066, "mean_token_accuracy": 0.8648008108139038, "num_tokens": 622433522.0, "step": 16394 }, { "epoch": 2.0856125174914135, "ewc_loss": 0.10986328125, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.804839134216309, "learning_rate": 1e-05, "loss": 0.4583, "mean_token_accuracy": 0.8800691366195679, "num_tokens": 622468805.0, "step": 16395 }, { "epoch": 2.0857397277700036, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.5481953620910645, "learning_rate": 1e-05, "loss": 0.4357, "mean_token_accuracy": 0.8899490833282471, "num_tokens": 622511462.0, "step": 16396 }, { "epoch": 2.085866938048594, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.581360340118408, "learning_rate": 1e-05, "loss": 0.4516, "mean_token_accuracy": 0.8820242881774902, "num_tokens": 622546246.0, "step": 16397 }, { "epoch": 2.0859941483271847, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 7.0203094482421875, "learning_rate": 1e-05, "loss": 0.4522, "mean_token_accuracy": 0.8881722688674927, "num_tokens": 622583113.0, "step": 16398 }, { "epoch": 2.086121358605775, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 8.106231689453125e-05, "grad_norm": 5.324741363525391, "learning_rate": 1e-05, "loss": 0.4487, "mean_token_accuracy": 0.8765162229537964, "num_tokens": 622621414.0, "step": 16399 }, { "epoch": 2.0862485688843657, "ewc_loss": 0.158203125, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 0.00013256072998046875, "grad_norm": 7.902528762817383, "learning_rate": 1e-05, "loss": 0.5534, "mean_token_accuracy": 0.8619486093521118, "num_tokens": 622660079.0, "step": 16400 }, { "epoch": 2.0863757791629562, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 5.63564395904541, "learning_rate": 1e-05, "loss": 0.4701, "mean_token_accuracy": 0.8773985505104065, "num_tokens": 622700413.0, "step": 16401 }, { "epoch": 2.0865029894415468, "ewc_loss": 0.1787109375, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 0.00015354156494140625, "grad_norm": 8.790514945983887, "learning_rate": 1e-05, "loss": 0.5163, "mean_token_accuracy": 0.8839428424835205, "num_tokens": 622736448.0, "step": 16402 }, { "epoch": 2.0866301997201373, "ewc_loss": 0.1572265625, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 0.0001316070556640625, "grad_norm": 7.095867156982422, "learning_rate": 1e-05, "loss": 0.5004, "mean_token_accuracy": 0.8801361322402954, "num_tokens": 622777014.0, "step": 16403 }, { "epoch": 2.086757409998728, "ewc_loss": 0.1630859375, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 0.0001373291015625, "grad_norm": 8.191373825073242, "learning_rate": 1e-05, "loss": 0.5027, "mean_token_accuracy": 0.8823763132095337, "num_tokens": 622817608.0, "step": 16404 }, { "epoch": 2.0868846202773184, "ewc_loss": 0.158203125, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 0.00013256072998046875, "grad_norm": 7.487984657287598, "learning_rate": 1e-05, "loss": 0.4949, "mean_token_accuracy": 0.8788020610809326, "num_tokens": 622853090.0, "step": 16405 }, { "epoch": 2.087011830555909, "ewc_loss": 0.1474609375, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 0.0001220703125, "grad_norm": 7.477599143981934, "learning_rate": 1e-05, "loss": 0.4464, "mean_token_accuracy": 0.8941528797149658, "num_tokens": 622891843.0, "step": 16406 }, { "epoch": 2.0871390408344994, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 0.00011873245239257812, "grad_norm": 7.066232204437256, "learning_rate": 1e-05, "loss": 0.5521, "mean_token_accuracy": 0.8569455146789551, "num_tokens": 622930364.0, "step": 16407 }, { "epoch": 2.08726625111309, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 0.0001125335693359375, "grad_norm": 7.156251430511475, "learning_rate": 1e-05, "loss": 0.5505, "mean_token_accuracy": 0.8597396016120911, "num_tokens": 622970199.0, "step": 16408 }, { "epoch": 2.0873934613916805, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.6856465339660645, "learning_rate": 1e-05, "loss": 0.4971, "mean_token_accuracy": 0.8714021444320679, "num_tokens": 623003669.0, "step": 16409 }, { "epoch": 2.087520671670271, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 6.972239971160889, "learning_rate": 1e-05, "loss": 0.5196, "mean_token_accuracy": 0.8644088506698608, "num_tokens": 623043110.0, "step": 16410 }, { "epoch": 2.0876478819488615, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.399337291717529, "learning_rate": 1e-05, "loss": 0.5383, "mean_token_accuracy": 0.8634811043739319, "num_tokens": 623082474.0, "step": 16411 }, { "epoch": 2.087775092227452, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.942138671875, "learning_rate": 1e-05, "loss": 0.5271, "mean_token_accuracy": 0.8631688356399536, "num_tokens": 623119650.0, "step": 16412 }, { "epoch": 2.0879023025060426, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.156768321990967, "learning_rate": 1e-05, "loss": 0.4975, "mean_token_accuracy": 0.867794394493103, "num_tokens": 623150359.0, "step": 16413 }, { "epoch": 2.088029512784633, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 7.081057071685791, "learning_rate": 1e-05, "loss": 0.4767, "mean_token_accuracy": 0.8786877393722534, "num_tokens": 623189251.0, "step": 16414 }, { "epoch": 2.0881567230632236, "ewc_loss": 0.11376953125, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 5.976985454559326, "learning_rate": 1e-05, "loss": 0.4212, "mean_token_accuracy": 0.8905999660491943, "num_tokens": 623231763.0, "step": 16415 }, { "epoch": 2.088283933341814, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 0.0001163482666015625, "grad_norm": 7.393458366394043, "learning_rate": 1e-05, "loss": 0.4648, "mean_token_accuracy": 0.8861885070800781, "num_tokens": 623269322.0, "step": 16416 }, { "epoch": 2.0884111436204047, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.019433498382568, "learning_rate": 1e-05, "loss": 0.4389, "mean_token_accuracy": 0.8856953382492065, "num_tokens": 623303765.0, "step": 16417 }, { "epoch": 2.0885383538989952, "ewc_loss": 0.150390625, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 0.00012493133544921875, "grad_norm": 7.7173004150390625, "learning_rate": 1e-05, "loss": 0.5245, "mean_token_accuracy": 0.8716380596160889, "num_tokens": 623355484.0, "step": 16418 }, { "epoch": 2.0886655641775858, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.405251502990723, "learning_rate": 1e-05, "loss": 0.5047, "mean_token_accuracy": 0.8741462230682373, "num_tokens": 623399441.0, "step": 16419 }, { "epoch": 2.0887927744561763, "ewc_loss": 0.1455078125, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 0.0001201629638671875, "grad_norm": 7.607020378112793, "learning_rate": 1e-05, "loss": 0.5695, "mean_token_accuracy": 0.8554648160934448, "num_tokens": 623438690.0, "step": 16420 }, { "epoch": 2.0889199847347664, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.77431058883667, "learning_rate": 1e-05, "loss": 0.5033, "mean_token_accuracy": 0.870773434638977, "num_tokens": 623476291.0, "step": 16421 }, { "epoch": 2.089047195013357, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 7.071032524108887, "learning_rate": 1e-05, "loss": 0.4577, "mean_token_accuracy": 0.8896181583404541, "num_tokens": 623513830.0, "step": 16422 }, { "epoch": 2.0891744052919474, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.8060688972473145, "learning_rate": 1e-05, "loss": 0.4972, "mean_token_accuracy": 0.8727300763130188, "num_tokens": 623554789.0, "step": 16423 }, { "epoch": 2.089301615570538, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.512356281280518, "learning_rate": 1e-05, "loss": 0.4771, "mean_token_accuracy": 0.880628228187561, "num_tokens": 623593646.0, "step": 16424 }, { "epoch": 2.0894288258491285, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.804128170013428, "learning_rate": 1e-05, "loss": 0.4533, "mean_token_accuracy": 0.887073814868927, "num_tokens": 623630646.0, "step": 16425 }, { "epoch": 2.089556036127719, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 6.035847187042236, "learning_rate": 1e-05, "loss": 0.5262, "mean_token_accuracy": 0.8580690622329712, "num_tokens": 623675488.0, "step": 16426 }, { "epoch": 2.0896832464063095, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.961656093597412, "learning_rate": 1e-05, "loss": 0.5132, "mean_token_accuracy": 0.8672994375228882, "num_tokens": 623719717.0, "step": 16427 }, { "epoch": 2.0898104566849, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.7540507316589355, "learning_rate": 1e-05, "loss": 0.4319, "mean_token_accuracy": 0.8846983313560486, "num_tokens": 623753248.0, "step": 16428 }, { "epoch": 2.0899376669634906, "ewc_loss": 0.140625, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 0.00011491775512695312, "grad_norm": 7.3184895515441895, "learning_rate": 1e-05, "loss": 0.5306, "mean_token_accuracy": 0.8649609088897705, "num_tokens": 623791562.0, "step": 16429 }, { "epoch": 2.090064877242081, "ewc_loss": 0.10888671875, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.73935604095459, "learning_rate": 1e-05, "loss": 0.4992, "mean_token_accuracy": 0.8640637397766113, "num_tokens": 623834187.0, "step": 16430 }, { "epoch": 2.0901920875206716, "ewc_loss": 0.15234375, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 0.00012683868408203125, "grad_norm": 7.767780303955078, "learning_rate": 1e-05, "loss": 0.5559, "mean_token_accuracy": 0.865481972694397, "num_tokens": 623869664.0, "step": 16431 }, { "epoch": 2.090319297799262, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.182846546173096, "learning_rate": 1e-05, "loss": 0.4755, "mean_token_accuracy": 0.8768002390861511, "num_tokens": 623907461.0, "step": 16432 }, { "epoch": 2.0904465080778527, "ewc_loss": 0.1533203125, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 0.0001277923583984375, "grad_norm": 7.818575859069824, "learning_rate": 1e-05, "loss": 0.5088, "mean_token_accuracy": 0.8759124279022217, "num_tokens": 623947679.0, "step": 16433 }, { "epoch": 2.0905737183564432, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 0.00011110305786132812, "grad_norm": 6.697244644165039, "learning_rate": 1e-05, "loss": 0.4656, "mean_token_accuracy": 0.8819626569747925, "num_tokens": 623992659.0, "step": 16434 }, { "epoch": 2.0907009286350338, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 0.00011730194091796875, "grad_norm": 7.38515567779541, "learning_rate": 1e-05, "loss": 0.4557, "mean_token_accuracy": 0.8891031742095947, "num_tokens": 624028305.0, "step": 16435 }, { "epoch": 2.0908281389136243, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 6.790303707122803, "learning_rate": 1e-05, "loss": 0.4719, "mean_token_accuracy": 0.8793339133262634, "num_tokens": 624062585.0, "step": 16436 }, { "epoch": 2.090955349192215, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.92871618270874, "learning_rate": 1e-05, "loss": 0.4654, "mean_token_accuracy": 0.8819477558135986, "num_tokens": 624100525.0, "step": 16437 }, { "epoch": 2.0910825594708053, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.6412248611450195, "learning_rate": 1e-05, "loss": 0.4467, "mean_token_accuracy": 0.887561559677124, "num_tokens": 624135982.0, "step": 16438 }, { "epoch": 2.091209769749396, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.60174036026001, "learning_rate": 1e-05, "loss": 0.4382, "mean_token_accuracy": 0.8872359395027161, "num_tokens": 624172867.0, "step": 16439 }, { "epoch": 2.0913369800279864, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.472172260284424, "learning_rate": 1e-05, "loss": 0.4229, "mean_token_accuracy": 0.8937598466873169, "num_tokens": 624208588.0, "step": 16440 }, { "epoch": 2.091464190306577, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.45603084564209, "learning_rate": 1e-05, "loss": 0.4744, "mean_token_accuracy": 0.8734914064407349, "num_tokens": 624241060.0, "step": 16441 }, { "epoch": 2.0915914005851675, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.278556823730469, "learning_rate": 1e-05, "loss": 0.4853, "mean_token_accuracy": 0.8704030513763428, "num_tokens": 624284508.0, "step": 16442 }, { "epoch": 2.091718610863758, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.446554183959961, "learning_rate": 1e-05, "loss": 0.4584, "mean_token_accuracy": 0.8812466859817505, "num_tokens": 624326787.0, "step": 16443 }, { "epoch": 2.0918458211423485, "ewc_loss": 0.11376953125, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 6.080169200897217, "learning_rate": 1e-05, "loss": 0.4324, "mean_token_accuracy": 0.8870056867599487, "num_tokens": 624363375.0, "step": 16444 }, { "epoch": 2.0919730314209386, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.553254127502441, "learning_rate": 1e-05, "loss": 0.4933, "mean_token_accuracy": 0.8703645467758179, "num_tokens": 624405508.0, "step": 16445 }, { "epoch": 2.092100241699529, "ewc_loss": 0.10986328125, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.908267021179199, "learning_rate": 1e-05, "loss": 0.4237, "mean_token_accuracy": 0.8872357606887817, "num_tokens": 624438505.0, "step": 16446 }, { "epoch": 2.0922274519781197, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.747650146484375, "learning_rate": 1e-05, "loss": 0.4694, "mean_token_accuracy": 0.8799658417701721, "num_tokens": 624475463.0, "step": 16447 }, { "epoch": 2.09235466225671, "ewc_loss": 0.10693359375, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.733612537384033, "learning_rate": 1e-05, "loss": 0.4359, "mean_token_accuracy": 0.8840289115905762, "num_tokens": 624511577.0, "step": 16448 }, { "epoch": 2.0924818725353007, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 7.133632659912109, "learning_rate": 1e-05, "loss": 0.4848, "mean_token_accuracy": 0.8805913329124451, "num_tokens": 624552213.0, "step": 16449 }, { "epoch": 2.0926090828138912, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.67255163192749, "learning_rate": 1e-05, "loss": 0.4318, "mean_token_accuracy": 0.8853006362915039, "num_tokens": 624589026.0, "step": 16450 }, { "epoch": 2.0927362930924818, "ewc_loss": 0.1484375, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.625935077667236, "learning_rate": 1e-05, "loss": 0.5031, "mean_token_accuracy": 0.8762583136558533, "num_tokens": 624623245.0, "step": 16451 }, { "epoch": 2.0928635033710723, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.016528129577637, "learning_rate": 1e-05, "loss": 0.4811, "mean_token_accuracy": 0.8739839792251587, "num_tokens": 624659713.0, "step": 16452 }, { "epoch": 2.092990713649663, "ewc_loss": 0.154296875, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 0.00012874603271484375, "grad_norm": 7.819526672363281, "learning_rate": 1e-05, "loss": 0.4876, "mean_token_accuracy": 0.8816105127334595, "num_tokens": 624695539.0, "step": 16453 }, { "epoch": 2.0931179239282534, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 6.599290370941162, "learning_rate": 1e-05, "loss": 0.4971, "mean_token_accuracy": 0.8716269731521606, "num_tokens": 624735311.0, "step": 16454 }, { "epoch": 2.093245134206844, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 0.00011873245239257812, "grad_norm": 7.421694278717041, "learning_rate": 1e-05, "loss": 0.4851, "mean_token_accuracy": 0.8828436732292175, "num_tokens": 624771941.0, "step": 16455 }, { "epoch": 2.0933723444854344, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 0.00011157989501953125, "grad_norm": 6.7726874351501465, "learning_rate": 1e-05, "loss": 0.4492, "mean_token_accuracy": 0.8899668455123901, "num_tokens": 624809963.0, "step": 16456 }, { "epoch": 2.093499554764025, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.881689548492432, "learning_rate": 1e-05, "loss": 0.4824, "mean_token_accuracy": 0.8776651620864868, "num_tokens": 624849839.0, "step": 16457 }, { "epoch": 2.0936267650426155, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.730321884155273, "learning_rate": 1e-05, "loss": 0.4844, "mean_token_accuracy": 0.8774375915527344, "num_tokens": 624883920.0, "step": 16458 }, { "epoch": 2.093753975321206, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.425641059875488, "learning_rate": 1e-05, "loss": 0.4465, "mean_token_accuracy": 0.8860732913017273, "num_tokens": 624916024.0, "step": 16459 }, { "epoch": 2.0938811855997965, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.712484836578369, "learning_rate": 1e-05, "loss": 0.4729, "mean_token_accuracy": 0.8816986680030823, "num_tokens": 624950978.0, "step": 16460 }, { "epoch": 2.094008395878387, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.085881233215332, "learning_rate": 1e-05, "loss": 0.4259, "mean_token_accuracy": 0.8878500461578369, "num_tokens": 624992177.0, "step": 16461 }, { "epoch": 2.0941356061569776, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.8114728927612305, "learning_rate": 1e-05, "loss": 0.475, "mean_token_accuracy": 0.878734290599823, "num_tokens": 625028211.0, "step": 16462 }, { "epoch": 2.094262816435568, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.894872188568115, "learning_rate": 1e-05, "loss": 0.4093, "mean_token_accuracy": 0.8953230381011963, "num_tokens": 625065895.0, "step": 16463 }, { "epoch": 2.0943900267141586, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 7.010532855987549, "learning_rate": 1e-05, "loss": 0.4629, "mean_token_accuracy": 0.8858702778816223, "num_tokens": 625100521.0, "step": 16464 }, { "epoch": 2.094517236992749, "ewc_loss": 0.11181640625, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.814746856689453, "learning_rate": 1e-05, "loss": 0.4614, "mean_token_accuracy": 0.8800119161605835, "num_tokens": 625144810.0, "step": 16465 }, { "epoch": 2.0946444472713397, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 0.00011396408081054688, "grad_norm": 7.2825517654418945, "learning_rate": 1e-05, "loss": 0.4988, "mean_token_accuracy": 0.8750113248825073, "num_tokens": 625178001.0, "step": 16466 }, { "epoch": 2.09477165754993, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 5.940433502197266, "learning_rate": 1e-05, "loss": 0.4486, "mean_token_accuracy": 0.8814051151275635, "num_tokens": 625214857.0, "step": 16467 }, { "epoch": 2.0948988678285207, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 0.00011920928955078125, "grad_norm": 7.43159818649292, "learning_rate": 1e-05, "loss": 0.4995, "mean_token_accuracy": 0.8813098669052124, "num_tokens": 625253396.0, "step": 16468 }, { "epoch": 2.095026078107111, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.161067008972168, "learning_rate": 1e-05, "loss": 0.4967, "mean_token_accuracy": 0.865360677242279, "num_tokens": 625291922.0, "step": 16469 }, { "epoch": 2.0951532883857014, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 0.00011730194091796875, "grad_norm": 7.389472007751465, "learning_rate": 1e-05, "loss": 0.5057, "mean_token_accuracy": 0.8764406442642212, "num_tokens": 625326548.0, "step": 16470 }, { "epoch": 2.095280498664292, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.372129917144775, "learning_rate": 1e-05, "loss": 0.4584, "mean_token_accuracy": 0.8836734294891357, "num_tokens": 625361354.0, "step": 16471 }, { "epoch": 2.0954077089428824, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 0.00011205673217773438, "grad_norm": 7.128705978393555, "learning_rate": 1e-05, "loss": 0.5136, "mean_token_accuracy": 0.8684554100036621, "num_tokens": 625400885.0, "step": 16472 }, { "epoch": 2.095534919221473, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.419943809509277, "learning_rate": 1e-05, "loss": 0.4648, "mean_token_accuracy": 0.8800967931747437, "num_tokens": 625438975.0, "step": 16473 }, { "epoch": 2.0956621295000635, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.830434799194336, "learning_rate": 1e-05, "loss": 0.5347, "mean_token_accuracy": 0.8599194884300232, "num_tokens": 625475384.0, "step": 16474 }, { "epoch": 2.095789339778654, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.391338348388672, "learning_rate": 1e-05, "loss": 0.4275, "mean_token_accuracy": 0.8934718370437622, "num_tokens": 625508422.0, "step": 16475 }, { "epoch": 2.0959165500572445, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.57727575302124, "learning_rate": 1e-05, "loss": 0.4402, "mean_token_accuracy": 0.8875960111618042, "num_tokens": 625546434.0, "step": 16476 }, { "epoch": 2.096043760335835, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.312508583068848, "learning_rate": 1e-05, "loss": 0.458, "mean_token_accuracy": 0.8828346729278564, "num_tokens": 625584005.0, "step": 16477 }, { "epoch": 2.0961709706144256, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.436155796051025, "learning_rate": 1e-05, "loss": 0.4201, "mean_token_accuracy": 0.8953913450241089, "num_tokens": 625616713.0, "step": 16478 }, { "epoch": 2.096298180893016, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.213990211486816, "learning_rate": 1e-05, "loss": 0.4652, "mean_token_accuracy": 0.8799888491630554, "num_tokens": 625657738.0, "step": 16479 }, { "epoch": 2.0964253911716066, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.385281085968018, "learning_rate": 1e-05, "loss": 0.4648, "mean_token_accuracy": 0.8748666048049927, "num_tokens": 625693485.0, "step": 16480 }, { "epoch": 2.096552601450197, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.102222442626953, "learning_rate": 1e-05, "loss": 0.4316, "mean_token_accuracy": 0.8857910633087158, "num_tokens": 625727219.0, "step": 16481 }, { "epoch": 2.0966798117287877, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.430087089538574, "learning_rate": 1e-05, "loss": 0.4261, "mean_token_accuracy": 0.892120361328125, "num_tokens": 625762858.0, "step": 16482 }, { "epoch": 2.0968070220073782, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.9644880294799805, "learning_rate": 1e-05, "loss": 0.4695, "mean_token_accuracy": 0.8759973645210266, "num_tokens": 625799752.0, "step": 16483 }, { "epoch": 2.0969342322859688, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.565748691558838, "learning_rate": 1e-05, "loss": 0.4283, "mean_token_accuracy": 0.8906694650650024, "num_tokens": 625835453.0, "step": 16484 }, { "epoch": 2.0970614425645593, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.8140764236450195, "learning_rate": 1e-05, "loss": 0.5034, "mean_token_accuracy": 0.8656477928161621, "num_tokens": 625875181.0, "step": 16485 }, { "epoch": 2.09718865284315, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.810823917388916, "learning_rate": 1e-05, "loss": 0.4535, "mean_token_accuracy": 0.8908497095108032, "num_tokens": 625918851.0, "step": 16486 }, { "epoch": 2.0973158631217403, "ewc_loss": 0.109375, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.685050010681152, "learning_rate": 1e-05, "loss": 0.4417, "mean_token_accuracy": 0.8809808492660522, "num_tokens": 625954859.0, "step": 16487 }, { "epoch": 2.097443073400331, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 0.00011301040649414062, "grad_norm": 7.205723762512207, "learning_rate": 1e-05, "loss": 0.4641, "mean_token_accuracy": 0.8844602108001709, "num_tokens": 625986042.0, "step": 16488 }, { "epoch": 2.0975702836789214, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.724833011627197, "learning_rate": 1e-05, "loss": 0.5298, "mean_token_accuracy": 0.8575201034545898, "num_tokens": 626023604.0, "step": 16489 }, { "epoch": 2.097697493957512, "ewc_loss": 0.1494140625, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 0.0001239776611328125, "grad_norm": 7.59084415435791, "learning_rate": 1e-05, "loss": 0.4428, "mean_token_accuracy": 0.8973712921142578, "num_tokens": 626059193.0, "step": 16490 }, { "epoch": 2.0978247042361025, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.060786247253418, "learning_rate": 1e-05, "loss": 0.4891, "mean_token_accuracy": 0.8722370862960815, "num_tokens": 626099994.0, "step": 16491 }, { "epoch": 2.097951914514693, "ewc_loss": 0.15234375, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 0.00012683868408203125, "grad_norm": 7.65907621383667, "learning_rate": 1e-05, "loss": 0.5055, "mean_token_accuracy": 0.8780959844589233, "num_tokens": 626133105.0, "step": 16492 }, { "epoch": 2.0980791247932835, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 0.00010967254638671875, "grad_norm": 6.517179012298584, "learning_rate": 1e-05, "loss": 0.4708, "mean_token_accuracy": 0.8861150145530701, "num_tokens": 626171837.0, "step": 16493 }, { "epoch": 2.0982063350718736, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 0.00011587142944335938, "grad_norm": 7.261244297027588, "learning_rate": 1e-05, "loss": 0.5064, "mean_token_accuracy": 0.8732777833938599, "num_tokens": 626214057.0, "step": 16494 }, { "epoch": 2.098333545350464, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 6.652853012084961, "learning_rate": 1e-05, "loss": 0.4653, "mean_token_accuracy": 0.8842767477035522, "num_tokens": 626249046.0, "step": 16495 }, { "epoch": 2.0984607556290547, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.771005153656006, "learning_rate": 1e-05, "loss": 0.5145, "mean_token_accuracy": 0.8673673868179321, "num_tokens": 626283991.0, "step": 16496 }, { "epoch": 2.098587965907645, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.657063961029053, "learning_rate": 1e-05, "loss": 0.4949, "mean_token_accuracy": 0.8757066130638123, "num_tokens": 626321919.0, "step": 16497 }, { "epoch": 2.0987151761862357, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.316741466522217, "learning_rate": 1e-05, "loss": 0.4309, "mean_token_accuracy": 0.8926124572753906, "num_tokens": 626360281.0, "step": 16498 }, { "epoch": 2.0988423864648262, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.685574531555176, "learning_rate": 1e-05, "loss": 0.4815, "mean_token_accuracy": 0.8800089359283447, "num_tokens": 626404314.0, "step": 16499 }, { "epoch": 2.0989695967434168, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 6.010650634765625, "learning_rate": 1e-05, "loss": 0.51, "mean_token_accuracy": 0.8642970323562622, "num_tokens": 626452743.0, "step": 16500 }, { "epoch": 2.0990968070220073, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.8248066902160645, "learning_rate": 1e-05, "loss": 0.4331, "mean_token_accuracy": 0.8950424194335938, "num_tokens": 626494310.0, "step": 16501 }, { "epoch": 2.099224017300598, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.842695236206055, "learning_rate": 1e-05, "loss": 0.4583, "mean_token_accuracy": 0.8809249401092529, "num_tokens": 626534995.0, "step": 16502 }, { "epoch": 2.0993512275791884, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 7.0399627685546875, "learning_rate": 1e-05, "loss": 0.4532, "mean_token_accuracy": 0.8872362375259399, "num_tokens": 626575084.0, "step": 16503 }, { "epoch": 2.099478437857779, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.825887203216553, "learning_rate": 1e-05, "loss": 0.4485, "mean_token_accuracy": 0.8834162950515747, "num_tokens": 626608922.0, "step": 16504 }, { "epoch": 2.0996056481363694, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 0.0001163482666015625, "grad_norm": 7.265622138977051, "learning_rate": 1e-05, "loss": 0.4631, "mean_token_accuracy": 0.8884857892990112, "num_tokens": 626645480.0, "step": 16505 }, { "epoch": 2.09973285841496, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 5.9401774406433105, "learning_rate": 1e-05, "loss": 0.495, "mean_token_accuracy": 0.870921790599823, "num_tokens": 626687641.0, "step": 16506 }, { "epoch": 2.0998600686935505, "ewc_loss": 0.1455078125, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 0.0001201629638671875, "grad_norm": 7.406853675842285, "learning_rate": 1e-05, "loss": 0.4577, "mean_token_accuracy": 0.8892412185668945, "num_tokens": 626720624.0, "step": 16507 }, { "epoch": 2.099987278972141, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.169207572937012, "learning_rate": 1e-05, "loss": 0.5271, "mean_token_accuracy": 0.8572497367858887, "num_tokens": 626760174.0, "step": 16508 }, { "epoch": 2.1001144892507315, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 0.00011777877807617188, "grad_norm": 7.31231164932251, "learning_rate": 1e-05, "loss": 0.4316, "mean_token_accuracy": 0.8936277627944946, "num_tokens": 626792970.0, "step": 16509 }, { "epoch": 2.100241699529322, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.343961238861084, "learning_rate": 1e-05, "loss": 0.4869, "mean_token_accuracy": 0.8765830993652344, "num_tokens": 626826793.0, "step": 16510 }, { "epoch": 2.1003689098079126, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 0.00011205673217773438, "grad_norm": 7.070130825042725, "learning_rate": 1e-05, "loss": 0.4774, "mean_token_accuracy": 0.8803511261940002, "num_tokens": 626862938.0, "step": 16511 }, { "epoch": 2.100496120086503, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.376588344573975, "learning_rate": 1e-05, "loss": 0.4709, "mean_token_accuracy": 0.8794481754302979, "num_tokens": 626900738.0, "step": 16512 }, { "epoch": 2.1006233303650936, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.849385738372803, "learning_rate": 1e-05, "loss": 0.4479, "mean_token_accuracy": 0.8881965279579163, "num_tokens": 626932670.0, "step": 16513 }, { "epoch": 2.100750540643684, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.330907821655273, "learning_rate": 1e-05, "loss": 0.4838, "mean_token_accuracy": 0.8753976225852966, "num_tokens": 626971683.0, "step": 16514 }, { "epoch": 2.1008777509222747, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.605566024780273, "learning_rate": 1e-05, "loss": 0.4892, "mean_token_accuracy": 0.871886134147644, "num_tokens": 627009955.0, "step": 16515 }, { "epoch": 2.101004961200865, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.2598557472229, "learning_rate": 1e-05, "loss": 0.5093, "mean_token_accuracy": 0.866051197052002, "num_tokens": 627052469.0, "step": 16516 }, { "epoch": 2.1011321714794557, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.472597122192383, "learning_rate": 1e-05, "loss": 0.4774, "mean_token_accuracy": 0.8757541179656982, "num_tokens": 627088314.0, "step": 16517 }, { "epoch": 2.1012593817580463, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.163199424743652, "learning_rate": 1e-05, "loss": 0.4744, "mean_token_accuracy": 0.8788888454437256, "num_tokens": 627132030.0, "step": 16518 }, { "epoch": 2.1013865920366364, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.440653324127197, "learning_rate": 1e-05, "loss": 0.4878, "mean_token_accuracy": 0.8744757771492004, "num_tokens": 627171135.0, "step": 16519 }, { "epoch": 2.101513802315227, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 6.053001403808594, "learning_rate": 1e-05, "loss": 0.4531, "mean_token_accuracy": 0.8809739947319031, "num_tokens": 627206303.0, "step": 16520 }, { "epoch": 2.1016410125938174, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.500147342681885, "learning_rate": 1e-05, "loss": 0.4742, "mean_token_accuracy": 0.8772980570793152, "num_tokens": 627242859.0, "step": 16521 }, { "epoch": 2.101768222872408, "ewc_loss": 0.11083984375, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.920244216918945, "learning_rate": 1e-05, "loss": 0.4676, "mean_token_accuracy": 0.8739626407623291, "num_tokens": 627284659.0, "step": 16522 }, { "epoch": 2.1018954331509985, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.653096675872803, "learning_rate": 1e-05, "loss": 0.46, "mean_token_accuracy": 0.8840371370315552, "num_tokens": 627320012.0, "step": 16523 }, { "epoch": 2.102022643429589, "ewc_loss": 0.109375, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.781964302062988, "learning_rate": 1e-05, "loss": 0.4401, "mean_token_accuracy": 0.8826132416725159, "num_tokens": 627358117.0, "step": 16524 }, { "epoch": 2.1021498537081795, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.930319786071777, "learning_rate": 1e-05, "loss": 0.5168, "mean_token_accuracy": 0.8669081926345825, "num_tokens": 627393331.0, "step": 16525 }, { "epoch": 2.10227706398677, "ewc_loss": 0.10986328125, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.719110012054443, "learning_rate": 1e-05, "loss": 0.4868, "mean_token_accuracy": 0.8687968254089355, "num_tokens": 627433905.0, "step": 16526 }, { "epoch": 2.1024042742653606, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 0.00011730194091796875, "grad_norm": 7.260286331176758, "learning_rate": 1e-05, "loss": 0.46, "mean_token_accuracy": 0.8910757303237915, "num_tokens": 627469673.0, "step": 16527 }, { "epoch": 2.102531484543951, "ewc_loss": 0.11474609375, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 5.847017288208008, "learning_rate": 1e-05, "loss": 0.4777, "mean_token_accuracy": 0.8753776550292969, "num_tokens": 627502917.0, "step": 16528 }, { "epoch": 2.1026586948225416, "ewc_loss": 0.1474609375, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 0.00012159347534179688, "grad_norm": 7.491647720336914, "learning_rate": 1e-05, "loss": 0.4596, "mean_token_accuracy": 0.891493558883667, "num_tokens": 627536632.0, "step": 16529 }, { "epoch": 2.102785905101132, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.187209129333496, "learning_rate": 1e-05, "loss": 0.4495, "mean_token_accuracy": 0.8856071829795837, "num_tokens": 627580820.0, "step": 16530 }, { "epoch": 2.1029131153797227, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 0.00011777877807617188, "grad_norm": 7.36006498336792, "learning_rate": 1e-05, "loss": 0.4354, "mean_token_accuracy": 0.8982582092285156, "num_tokens": 627616895.0, "step": 16531 }, { "epoch": 2.1030403256583132, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.470578670501709, "learning_rate": 1e-05, "loss": 0.4875, "mean_token_accuracy": 0.871978759765625, "num_tokens": 627655648.0, "step": 16532 }, { "epoch": 2.1031675359369038, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 6.931888580322266, "learning_rate": 1e-05, "loss": 0.4954, "mean_token_accuracy": 0.8780566453933716, "num_tokens": 627692259.0, "step": 16533 }, { "epoch": 2.1032947462154943, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.562095642089844, "learning_rate": 1e-05, "loss": 0.4639, "mean_token_accuracy": 0.8835426568984985, "num_tokens": 627731601.0, "step": 16534 }, { "epoch": 2.103421956494085, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.447865962982178, "learning_rate": 1e-05, "loss": 0.4717, "mean_token_accuracy": 0.8775028586387634, "num_tokens": 627768381.0, "step": 16535 }, { "epoch": 2.1035491667726753, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.638206481933594, "learning_rate": 1e-05, "loss": 0.4569, "mean_token_accuracy": 0.8847448825836182, "num_tokens": 627808327.0, "step": 16536 }, { "epoch": 2.103676377051266, "ewc_loss": 0.11474609375, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.060727119445801, "learning_rate": 1e-05, "loss": 0.445, "mean_token_accuracy": 0.8811267614364624, "num_tokens": 627838779.0, "step": 16537 }, { "epoch": 2.1038035873298564, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.804683685302734, "learning_rate": 1e-05, "loss": 0.4755, "mean_token_accuracy": 0.8775102496147156, "num_tokens": 627875634.0, "step": 16538 }, { "epoch": 2.103930797608447, "ewc_loss": 0.11083984375, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.79900598526001, "learning_rate": 1e-05, "loss": 0.4579, "mean_token_accuracy": 0.8794490098953247, "num_tokens": 627918350.0, "step": 16539 }, { "epoch": 2.1040580078870375, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 7.100759983062744, "learning_rate": 1e-05, "loss": 0.4883, "mean_token_accuracy": 0.8755018711090088, "num_tokens": 627955649.0, "step": 16540 }, { "epoch": 2.104185218165628, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.760378360748291, "learning_rate": 1e-05, "loss": 0.4611, "mean_token_accuracy": 0.8791794776916504, "num_tokens": 627989374.0, "step": 16541 }, { "epoch": 2.1043124284442185, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 0.00011920928955078125, "grad_norm": 7.437012672424316, "learning_rate": 1e-05, "loss": 0.5327, "mean_token_accuracy": 0.8689563274383545, "num_tokens": 628026338.0, "step": 16542 }, { "epoch": 2.1044396387228086, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 5.97072172164917, "learning_rate": 1e-05, "loss": 0.4546, "mean_token_accuracy": 0.884627640247345, "num_tokens": 628062550.0, "step": 16543 }, { "epoch": 2.104566849001399, "ewc_loss": 0.1474609375, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 0.0001220703125, "grad_norm": 7.570889949798584, "learning_rate": 1e-05, "loss": 0.5135, "mean_token_accuracy": 0.8737637996673584, "num_tokens": 628104868.0, "step": 16544 }, { "epoch": 2.1046940592799896, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.306721210479736, "learning_rate": 1e-05, "loss": 0.5032, "mean_token_accuracy": 0.8714998960494995, "num_tokens": 628142009.0, "step": 16545 }, { "epoch": 2.10482126955858, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 0.00011920928955078125, "grad_norm": 7.3994598388671875, "learning_rate": 1e-05, "loss": 0.4911, "mean_token_accuracy": 0.8762285709381104, "num_tokens": 628180992.0, "step": 16546 }, { "epoch": 2.1049484798371707, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.49481725692749, "learning_rate": 1e-05, "loss": 0.4827, "mean_token_accuracy": 0.8787908554077148, "num_tokens": 628215636.0, "step": 16547 }, { "epoch": 2.1050756901157612, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 0.00011157989501953125, "grad_norm": 7.051285743713379, "learning_rate": 1e-05, "loss": 0.4564, "mean_token_accuracy": 0.8854929804801941, "num_tokens": 628256993.0, "step": 16548 }, { "epoch": 2.1052029003943518, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.474691867828369, "learning_rate": 1e-05, "loss": 0.5001, "mean_token_accuracy": 0.8758176565170288, "num_tokens": 628293408.0, "step": 16549 }, { "epoch": 2.1053301106729423, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.741573810577393, "learning_rate": 1e-05, "loss": 0.4728, "mean_token_accuracy": 0.8809846043586731, "num_tokens": 628329353.0, "step": 16550 }, { "epoch": 2.105457320951533, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.384424686431885, "learning_rate": 1e-05, "loss": 0.5074, "mean_token_accuracy": 0.8673821687698364, "num_tokens": 628364312.0, "step": 16551 }, { "epoch": 2.1055845312301233, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.539791584014893, "learning_rate": 1e-05, "loss": 0.4554, "mean_token_accuracy": 0.8812747001647949, "num_tokens": 628401354.0, "step": 16552 }, { "epoch": 2.105711741508714, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.264134407043457, "learning_rate": 1e-05, "loss": 0.4464, "mean_token_accuracy": 0.8857883214950562, "num_tokens": 628432840.0, "step": 16553 }, { "epoch": 2.1058389517873044, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.4570417404174805, "learning_rate": 1e-05, "loss": 0.4812, "mean_token_accuracy": 0.8756983876228333, "num_tokens": 628476300.0, "step": 16554 }, { "epoch": 2.105966162065895, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.11989164352417, "learning_rate": 1e-05, "loss": 0.4725, "mean_token_accuracy": 0.8770163059234619, "num_tokens": 628514248.0, "step": 16555 }, { "epoch": 2.1060933723444855, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.485373020172119, "learning_rate": 1e-05, "loss": 0.4561, "mean_token_accuracy": 0.8825151920318604, "num_tokens": 628557161.0, "step": 16556 }, { "epoch": 2.106220582623076, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.973474979400635, "learning_rate": 1e-05, "loss": 0.4393, "mean_token_accuracy": 0.8844038248062134, "num_tokens": 628591073.0, "step": 16557 }, { "epoch": 2.1063477929016665, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.621593475341797, "learning_rate": 1e-05, "loss": 0.4553, "mean_token_accuracy": 0.8831184506416321, "num_tokens": 628627215.0, "step": 16558 }, { "epoch": 2.106475003180257, "ewc_loss": 0.109375, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.829945087432861, "learning_rate": 1e-05, "loss": 0.4637, "mean_token_accuracy": 0.8759523630142212, "num_tokens": 628669858.0, "step": 16559 }, { "epoch": 2.1066022134588476, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.861490249633789, "learning_rate": 1e-05, "loss": 0.4281, "mean_token_accuracy": 0.8943654298782349, "num_tokens": 628704504.0, "step": 16560 }, { "epoch": 2.106729423737438, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.724273204803467, "learning_rate": 1e-05, "loss": 0.4892, "mean_token_accuracy": 0.8637614846229553, "num_tokens": 628743444.0, "step": 16561 }, { "epoch": 2.1068566340160286, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 0.00011301040649414062, "grad_norm": 7.215959072113037, "learning_rate": 1e-05, "loss": 0.5245, "mean_token_accuracy": 0.8655709028244019, "num_tokens": 628782256.0, "step": 16562 }, { "epoch": 2.106983844294619, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 5.78395414352417, "learning_rate": 1e-05, "loss": 0.4515, "mean_token_accuracy": 0.8837283849716187, "num_tokens": 628822150.0, "step": 16563 }, { "epoch": 2.1071110545732097, "ewc_loss": 0.1484375, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.534091949462891, "learning_rate": 1e-05, "loss": 0.4638, "mean_token_accuracy": 0.8901593685150146, "num_tokens": 628862445.0, "step": 16564 }, { "epoch": 2.1072382648518, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.146678447723389, "learning_rate": 1e-05, "loss": 0.4586, "mean_token_accuracy": 0.8797736763954163, "num_tokens": 628897166.0, "step": 16565 }, { "epoch": 2.1073654751303907, "ewc_loss": 0.1474609375, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 0.0001220703125, "grad_norm": 7.4722113609313965, "learning_rate": 1e-05, "loss": 0.4784, "mean_token_accuracy": 0.8821940422058105, "num_tokens": 628933443.0, "step": 16566 }, { "epoch": 2.107492685408981, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.4872331619262695, "learning_rate": 1e-05, "loss": 0.4552, "mean_token_accuracy": 0.8851166367530823, "num_tokens": 628969614.0, "step": 16567 }, { "epoch": 2.1076198956875714, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 0.00011110305786132812, "grad_norm": 7.044607162475586, "learning_rate": 1e-05, "loss": 0.4683, "mean_token_accuracy": 0.8833674192428589, "num_tokens": 629010386.0, "step": 16568 }, { "epoch": 2.107747105966162, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.61106538772583, "learning_rate": 1e-05, "loss": 0.5045, "mean_token_accuracy": 0.8714713454246521, "num_tokens": 629041388.0, "step": 16569 }, { "epoch": 2.1078743162447524, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.527899742126465, "learning_rate": 1e-05, "loss": 0.438, "mean_token_accuracy": 0.8895326852798462, "num_tokens": 629077306.0, "step": 16570 }, { "epoch": 2.108001526523343, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.662353515625, "learning_rate": 1e-05, "loss": 0.4537, "mean_token_accuracy": 0.8865945339202881, "num_tokens": 629110636.0, "step": 16571 }, { "epoch": 2.1081287368019335, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 6.091070652008057, "learning_rate": 1e-05, "loss": 0.4078, "mean_token_accuracy": 0.8979636430740356, "num_tokens": 629145834.0, "step": 16572 }, { "epoch": 2.108255947080524, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.823750019073486, "learning_rate": 1e-05, "loss": 0.4923, "mean_token_accuracy": 0.8751561641693115, "num_tokens": 629187148.0, "step": 16573 }, { "epoch": 2.1083831573591145, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.829647064208984, "learning_rate": 1e-05, "loss": 0.4268, "mean_token_accuracy": 0.8888379335403442, "num_tokens": 629224391.0, "step": 16574 }, { "epoch": 2.108510367637705, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 0.00011014938354492188, "grad_norm": 7.110408782958984, "learning_rate": 1e-05, "loss": 0.4827, "mean_token_accuracy": 0.8811922073364258, "num_tokens": 629261074.0, "step": 16575 }, { "epoch": 2.1086375779162956, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.776369571685791, "learning_rate": 1e-05, "loss": 0.4557, "mean_token_accuracy": 0.8796907663345337, "num_tokens": 629296568.0, "step": 16576 }, { "epoch": 2.108764788194886, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 0.00011920928955078125, "grad_norm": 7.411038398742676, "learning_rate": 1e-05, "loss": 0.4826, "mean_token_accuracy": 0.8819757699966431, "num_tokens": 629342784.0, "step": 16577 }, { "epoch": 2.1088919984734766, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 5.976963520050049, "learning_rate": 1e-05, "loss": 0.4449, "mean_token_accuracy": 0.8847448825836182, "num_tokens": 629377705.0, "step": 16578 }, { "epoch": 2.109019208752067, "ewc_loss": 0.1484375, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.531322956085205, "learning_rate": 1e-05, "loss": 0.4788, "mean_token_accuracy": 0.8833109140396118, "num_tokens": 629414823.0, "step": 16579 }, { "epoch": 2.1091464190306577, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.3065338134765625, "learning_rate": 1e-05, "loss": 0.5037, "mean_token_accuracy": 0.8680700063705444, "num_tokens": 629452404.0, "step": 16580 }, { "epoch": 2.109273629309248, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 0.00011873245239257812, "grad_norm": 7.341364860534668, "learning_rate": 1e-05, "loss": 0.4823, "mean_token_accuracy": 0.8835744857788086, "num_tokens": 629486147.0, "step": 16581 }, { "epoch": 2.1094008395878387, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.476243495941162, "learning_rate": 1e-05, "loss": 0.4642, "mean_token_accuracy": 0.8844146728515625, "num_tokens": 629523581.0, "step": 16582 }, { "epoch": 2.1095280498664293, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 7.002092361450195, "learning_rate": 1e-05, "loss": 0.4753, "mean_token_accuracy": 0.8789222240447998, "num_tokens": 629562125.0, "step": 16583 }, { "epoch": 2.10965526014502, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.452036380767822, "learning_rate": 1e-05, "loss": 0.4502, "mean_token_accuracy": 0.8876309394836426, "num_tokens": 629601786.0, "step": 16584 }, { "epoch": 2.1097824704236103, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.697611331939697, "learning_rate": 1e-05, "loss": 0.452, "mean_token_accuracy": 0.8870208859443665, "num_tokens": 629645022.0, "step": 16585 }, { "epoch": 2.109909680702201, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.3835296630859375, "learning_rate": 1e-05, "loss": 0.4354, "mean_token_accuracy": 0.8912540674209595, "num_tokens": 629675604.0, "step": 16586 }, { "epoch": 2.1100368909807914, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.497349262237549, "learning_rate": 1e-05, "loss": 0.4838, "mean_token_accuracy": 0.871779203414917, "num_tokens": 629718508.0, "step": 16587 }, { "epoch": 2.110164101259382, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.26782751083374, "learning_rate": 1e-05, "loss": 0.5248, "mean_token_accuracy": 0.8605113625526428, "num_tokens": 629758476.0, "step": 16588 }, { "epoch": 2.1102913115379724, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.433564186096191, "learning_rate": 1e-05, "loss": 0.4852, "mean_token_accuracy": 0.8723894357681274, "num_tokens": 629793692.0, "step": 16589 }, { "epoch": 2.110418521816563, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.14059591293335, "learning_rate": 1e-05, "loss": 0.4845, "mean_token_accuracy": 0.8702582120895386, "num_tokens": 629829750.0, "step": 16590 }, { "epoch": 2.1105457320951535, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.459149360656738, "learning_rate": 1e-05, "loss": 0.5224, "mean_token_accuracy": 0.8605557680130005, "num_tokens": 629865832.0, "step": 16591 }, { "epoch": 2.1106729423737436, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 5.990907192230225, "learning_rate": 1e-05, "loss": 0.4923, "mean_token_accuracy": 0.868351936340332, "num_tokens": 629906450.0, "step": 16592 }, { "epoch": 2.110800152652334, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.584904193878174, "learning_rate": 1e-05, "loss": 0.5216, "mean_token_accuracy": 0.8650363087654114, "num_tokens": 629939650.0, "step": 16593 }, { "epoch": 2.1109273629309246, "ewc_loss": 0.109375, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.830700874328613, "learning_rate": 1e-05, "loss": 0.4582, "mean_token_accuracy": 0.8791011571884155, "num_tokens": 629977692.0, "step": 16594 }, { "epoch": 2.111054573209515, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.815630912780762, "learning_rate": 1e-05, "loss": 0.4699, "mean_token_accuracy": 0.8808115720748901, "num_tokens": 630019111.0, "step": 16595 }, { "epoch": 2.1111817834881057, "ewc_loss": 0.109375, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.702148914337158, "learning_rate": 1e-05, "loss": 0.4197, "mean_token_accuracy": 0.8869648575782776, "num_tokens": 630054924.0, "step": 16596 }, { "epoch": 2.1113089937666962, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 0.00011444091796875, "grad_norm": 7.192323684692383, "learning_rate": 1e-05, "loss": 0.513, "mean_token_accuracy": 0.8729223012924194, "num_tokens": 630100344.0, "step": 16597 }, { "epoch": 2.1114362040452868, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.726718902587891, "learning_rate": 1e-05, "loss": 0.4494, "mean_token_accuracy": 0.8803285956382751, "num_tokens": 630137296.0, "step": 16598 }, { "epoch": 2.1115634143238773, "ewc_loss": 0.1484375, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.58041524887085, "learning_rate": 1e-05, "loss": 0.494, "mean_token_accuracy": 0.8816562294960022, "num_tokens": 630177477.0, "step": 16599 }, { "epoch": 2.111690624602468, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.0918073654174805, "learning_rate": 1e-05, "loss": 0.4143, "mean_token_accuracy": 0.8960003852844238, "num_tokens": 630205830.0, "step": 16600 }, { "epoch": 2.1118178348810583, "ewc_loss": 0.1484375, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.619255542755127, "learning_rate": 1e-05, "loss": 0.5042, "mean_token_accuracy": 0.8778504133224487, "num_tokens": 630247114.0, "step": 16601 }, { "epoch": 2.111945045159649, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 6.510498046875, "learning_rate": 1e-05, "loss": 0.4467, "mean_token_accuracy": 0.8929804563522339, "num_tokens": 630279848.0, "step": 16602 }, { "epoch": 2.1120722554382394, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 0.00011587142944335938, "grad_norm": 7.235050201416016, "learning_rate": 1e-05, "loss": 0.4709, "mean_token_accuracy": 0.8889143466949463, "num_tokens": 630320993.0, "step": 16603 }, { "epoch": 2.11219946571683, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 6.641358375549316, "learning_rate": 1e-05, "loss": 0.5484, "mean_token_accuracy": 0.856645405292511, "num_tokens": 630354703.0, "step": 16604 }, { "epoch": 2.1123266759954205, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.753749847412109, "learning_rate": 1e-05, "loss": 0.4886, "mean_token_accuracy": 0.8741096258163452, "num_tokens": 630388902.0, "step": 16605 }, { "epoch": 2.112453886274011, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.633333206176758, "learning_rate": 1e-05, "loss": 0.4771, "mean_token_accuracy": 0.8803783655166626, "num_tokens": 630431446.0, "step": 16606 }, { "epoch": 2.1125810965526015, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.538433074951172, "learning_rate": 1e-05, "loss": 0.4737, "mean_token_accuracy": 0.8744419813156128, "num_tokens": 630468334.0, "step": 16607 }, { "epoch": 2.112708306831192, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.564032554626465, "learning_rate": 1e-05, "loss": 0.4572, "mean_token_accuracy": 0.8835920095443726, "num_tokens": 630498607.0, "step": 16608 }, { "epoch": 2.1128355171097826, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.109686851501465, "learning_rate": 1e-05, "loss": 0.4416, "mean_token_accuracy": 0.8824293613433838, "num_tokens": 630534900.0, "step": 16609 }, { "epoch": 2.112962727388373, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.604732036590576, "learning_rate": 1e-05, "loss": 0.5012, "mean_token_accuracy": 0.8690416812896729, "num_tokens": 630577045.0, "step": 16610 }, { "epoch": 2.1130899376669636, "ewc_loss": 0.11376953125, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 5.923316478729248, "learning_rate": 1e-05, "loss": 0.4487, "mean_token_accuracy": 0.8819092512130737, "num_tokens": 630613477.0, "step": 16611 }, { "epoch": 2.113217147945554, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.762842655181885, "learning_rate": 1e-05, "loss": 0.4742, "mean_token_accuracy": 0.8783988356590271, "num_tokens": 630651868.0, "step": 16612 }, { "epoch": 2.1133443582241447, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.801084041595459, "learning_rate": 1e-05, "loss": 0.4532, "mean_token_accuracy": 0.881247341632843, "num_tokens": 630691483.0, "step": 16613 }, { "epoch": 2.113471568502735, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 6.987893581390381, "learning_rate": 1e-05, "loss": 0.4768, "mean_token_accuracy": 0.8829708099365234, "num_tokens": 630730392.0, "step": 16614 }, { "epoch": 2.1135987787813257, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.75546407699585, "learning_rate": 1e-05, "loss": 0.442, "mean_token_accuracy": 0.8862528800964355, "num_tokens": 630764336.0, "step": 16615 }, { "epoch": 2.1137259890599163, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 0.000118255615234375, "grad_norm": 7.285419464111328, "learning_rate": 1e-05, "loss": 0.5222, "mean_token_accuracy": 0.8711630702018738, "num_tokens": 630805912.0, "step": 16616 }, { "epoch": 2.1138531993385064, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 5.861658573150635, "learning_rate": 1e-05, "loss": 0.4903, "mean_token_accuracy": 0.8715385794639587, "num_tokens": 630844667.0, "step": 16617 }, { "epoch": 2.113980409617097, "ewc_loss": 0.1484375, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.480857849121094, "learning_rate": 1e-05, "loss": 0.4907, "mean_token_accuracy": 0.8793172836303711, "num_tokens": 630888217.0, "step": 16618 }, { "epoch": 2.1141076198956874, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.159040451049805, "learning_rate": 1e-05, "loss": 0.4347, "mean_token_accuracy": 0.8878377676010132, "num_tokens": 630922415.0, "step": 16619 }, { "epoch": 2.114234830174278, "ewc_loss": 0.1484375, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 0.0001220703125, "grad_norm": 7.396175861358643, "learning_rate": 1e-05, "loss": 0.495, "mean_token_accuracy": 0.8819600343704224, "num_tokens": 630955469.0, "step": 16620 }, { "epoch": 2.1143620404528685, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.3857879638671875, "learning_rate": 1e-05, "loss": 0.4705, "mean_token_accuracy": 0.8834853768348694, "num_tokens": 630994641.0, "step": 16621 }, { "epoch": 2.114489250731459, "ewc_loss": 0.140625, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 0.00011444091796875, "grad_norm": 7.09453010559082, "learning_rate": 1e-05, "loss": 0.5101, "mean_token_accuracy": 0.8695017695426941, "num_tokens": 631034671.0, "step": 16622 }, { "epoch": 2.1146164610100495, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.463204383850098, "learning_rate": 1e-05, "loss": 0.5006, "mean_token_accuracy": 0.872862696647644, "num_tokens": 631073518.0, "step": 16623 }, { "epoch": 2.11474367128864, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.736205101013184, "learning_rate": 1e-05, "loss": 0.5335, "mean_token_accuracy": 0.8627913594245911, "num_tokens": 631111631.0, "step": 16624 }, { "epoch": 2.1148708815672306, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.4368438720703125, "learning_rate": 1e-05, "loss": 0.4699, "mean_token_accuracy": 0.8768868446350098, "num_tokens": 631148081.0, "step": 16625 }, { "epoch": 2.114998091845821, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.440035343170166, "learning_rate": 1e-05, "loss": 0.5017, "mean_token_accuracy": 0.8717484474182129, "num_tokens": 631183176.0, "step": 16626 }, { "epoch": 2.1151253021244116, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.404767036437988, "learning_rate": 1e-05, "loss": 0.4574, "mean_token_accuracy": 0.8812402486801147, "num_tokens": 631213841.0, "step": 16627 }, { "epoch": 2.115252512403002, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.238936901092529, "learning_rate": 1e-05, "loss": 0.5009, "mean_token_accuracy": 0.8649440407752991, "num_tokens": 631249410.0, "step": 16628 }, { "epoch": 2.1153797226815927, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.370695114135742, "learning_rate": 1e-05, "loss": 0.468, "mean_token_accuracy": 0.8769758939743042, "num_tokens": 631292153.0, "step": 16629 }, { "epoch": 2.115506932960183, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.115372657775879, "learning_rate": 1e-05, "loss": 0.4727, "mean_token_accuracy": 0.8768684267997742, "num_tokens": 631330461.0, "step": 16630 }, { "epoch": 2.1156341432387737, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.3634161949157715, "learning_rate": 1e-05, "loss": 0.4991, "mean_token_accuracy": 0.8730673789978027, "num_tokens": 631367384.0, "step": 16631 }, { "epoch": 2.1157613535173643, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 6.040904521942139, "learning_rate": 1e-05, "loss": 0.4742, "mean_token_accuracy": 0.8722820281982422, "num_tokens": 631402558.0, "step": 16632 }, { "epoch": 2.115888563795955, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.362372875213623, "learning_rate": 1e-05, "loss": 0.5216, "mean_token_accuracy": 0.8582488298416138, "num_tokens": 631444040.0, "step": 16633 }, { "epoch": 2.1160157740745453, "ewc_loss": 0.11376953125, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 5.949533462524414, "learning_rate": 1e-05, "loss": 0.4482, "mean_token_accuracy": 0.8860787153244019, "num_tokens": 631486350.0, "step": 16634 }, { "epoch": 2.116142984353136, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.439654350280762, "learning_rate": 1e-05, "loss": 0.4388, "mean_token_accuracy": 0.8869598507881165, "num_tokens": 631519145.0, "step": 16635 }, { "epoch": 2.1162701946317264, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.867733478546143, "learning_rate": 1e-05, "loss": 0.4742, "mean_token_accuracy": 0.8730934262275696, "num_tokens": 631554637.0, "step": 16636 }, { "epoch": 2.116397404910317, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.578344345092773, "learning_rate": 1e-05, "loss": 0.4451, "mean_token_accuracy": 0.8876013159751892, "num_tokens": 631593580.0, "step": 16637 }, { "epoch": 2.1165246151889074, "ewc_loss": 0.10986328125, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.697984218597412, "learning_rate": 1e-05, "loss": 0.4632, "mean_token_accuracy": 0.8767749071121216, "num_tokens": 631630410.0, "step": 16638 }, { "epoch": 2.116651825467498, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.898750305175781, "learning_rate": 1e-05, "loss": 0.4846, "mean_token_accuracy": 0.8797447681427002, "num_tokens": 631661894.0, "step": 16639 }, { "epoch": 2.116779035746088, "ewc_loss": 0.10888671875, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.5493388175964355, "learning_rate": 1e-05, "loss": 0.4424, "mean_token_accuracy": 0.8770933747291565, "num_tokens": 631694514.0, "step": 16640 }, { "epoch": 2.1169062460246786, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 0.00012063980102539062, "grad_norm": 7.401546001434326, "learning_rate": 1e-05, "loss": 0.5068, "mean_token_accuracy": 0.8706361651420593, "num_tokens": 631737715.0, "step": 16641 }, { "epoch": 2.117033456303269, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.620635032653809, "learning_rate": 1e-05, "loss": 0.4983, "mean_token_accuracy": 0.8682447671890259, "num_tokens": 631773972.0, "step": 16642 }, { "epoch": 2.1171606665818596, "ewc_loss": 0.162109375, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 0.0001354217529296875, "grad_norm": 7.977424144744873, "learning_rate": 1e-05, "loss": 0.5155, "mean_token_accuracy": 0.8742888569831848, "num_tokens": 631812292.0, "step": 16643 }, { "epoch": 2.11728787686045, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.231170177459717, "learning_rate": 1e-05, "loss": 0.4487, "mean_token_accuracy": 0.8887600898742676, "num_tokens": 631849518.0, "step": 16644 }, { "epoch": 2.1174150871390407, "ewc_loss": 0.162109375, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 0.0001354217529296875, "grad_norm": 8.023755073547363, "learning_rate": 1e-05, "loss": 0.5043, "mean_token_accuracy": 0.8792837262153625, "num_tokens": 631889821.0, "step": 16645 }, { "epoch": 2.1175422974176312, "ewc_loss": 0.1484375, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 0.0001220703125, "grad_norm": 6.857207298278809, "learning_rate": 1e-05, "loss": 0.4547, "mean_token_accuracy": 0.8905045986175537, "num_tokens": 631925192.0, "step": 16646 }, { "epoch": 2.1176695076962218, "ewc_loss": 0.1484375, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.517007350921631, "learning_rate": 1e-05, "loss": 0.4951, "mean_token_accuracy": 0.878162145614624, "num_tokens": 631967876.0, "step": 16647 }, { "epoch": 2.1177967179748123, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 0.00011682510375976562, "grad_norm": 6.863386631011963, "learning_rate": 1e-05, "loss": 0.4766, "mean_token_accuracy": 0.8834590315818787, "num_tokens": 632006026.0, "step": 16648 }, { "epoch": 2.117923928253403, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 0.00011014938354492188, "grad_norm": 7.064713001251221, "learning_rate": 1e-05, "loss": 0.4471, "mean_token_accuracy": 0.8909094333648682, "num_tokens": 632045886.0, "step": 16649 }, { "epoch": 2.1180511385319933, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 6.680709362030029, "learning_rate": 1e-05, "loss": 0.4635, "mean_token_accuracy": 0.8851525783538818, "num_tokens": 632083583.0, "step": 16650 }, { "epoch": 2.118178348810584, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.731451988220215, "learning_rate": 1e-05, "loss": 0.4673, "mean_token_accuracy": 0.8815441131591797, "num_tokens": 632124427.0, "step": 16651 }, { "epoch": 2.1183055590891744, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.5165114402771, "learning_rate": 1e-05, "loss": 0.5007, "mean_token_accuracy": 0.8742465972900391, "num_tokens": 632159787.0, "step": 16652 }, { "epoch": 2.118432769367765, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.506252288818359, "learning_rate": 1e-05, "loss": 0.4431, "mean_token_accuracy": 0.8867371082305908, "num_tokens": 632194737.0, "step": 16653 }, { "epoch": 2.1185599796463555, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.363349437713623, "learning_rate": 1e-05, "loss": 0.4586, "mean_token_accuracy": 0.8827440738677979, "num_tokens": 632230737.0, "step": 16654 }, { "epoch": 2.118687189924946, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.39207649230957, "learning_rate": 1e-05, "loss": 0.4728, "mean_token_accuracy": 0.8766235113143921, "num_tokens": 632271374.0, "step": 16655 }, { "epoch": 2.1188144002035365, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.244739055633545, "learning_rate": 1e-05, "loss": 0.4369, "mean_token_accuracy": 0.8864091038703918, "num_tokens": 632305326.0, "step": 16656 }, { "epoch": 2.118941610482127, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.349564552307129, "learning_rate": 1e-05, "loss": 0.4693, "mean_token_accuracy": 0.878236711025238, "num_tokens": 632338807.0, "step": 16657 }, { "epoch": 2.1190688207607176, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.136985778808594, "learning_rate": 1e-05, "loss": 0.5063, "mean_token_accuracy": 0.8671852946281433, "num_tokens": 632373933.0, "step": 16658 }, { "epoch": 2.119196031039308, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.3691253662109375, "learning_rate": 1e-05, "loss": 0.461, "mean_token_accuracy": 0.878395140171051, "num_tokens": 632414149.0, "step": 16659 }, { "epoch": 2.1193232413178986, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 5.999366760253906, "learning_rate": 1e-05, "loss": 0.4363, "mean_token_accuracy": 0.8851138353347778, "num_tokens": 632453622.0, "step": 16660 }, { "epoch": 2.119450451596489, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.484969139099121, "learning_rate": 1e-05, "loss": 0.5083, "mean_token_accuracy": 0.8652650713920593, "num_tokens": 632488398.0, "step": 16661 }, { "epoch": 2.1195776618750797, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.826395511627197, "learning_rate": 1e-05, "loss": 0.4702, "mean_token_accuracy": 0.8735114932060242, "num_tokens": 632524013.0, "step": 16662 }, { "epoch": 2.11970487215367, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.7310357093811035, "learning_rate": 1e-05, "loss": 0.5042, "mean_token_accuracy": 0.8720768690109253, "num_tokens": 632562247.0, "step": 16663 }, { "epoch": 2.1198320824322607, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.641026973724365, "learning_rate": 1e-05, "loss": 0.4951, "mean_token_accuracy": 0.8687806129455566, "num_tokens": 632600066.0, "step": 16664 }, { "epoch": 2.119959292710851, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 0.00011205673217773438, "grad_norm": 7.169334888458252, "learning_rate": 1e-05, "loss": 0.4992, "mean_token_accuracy": 0.872857928276062, "num_tokens": 632638157.0, "step": 16665 }, { "epoch": 2.1200865029894413, "ewc_loss": 0.1083984375, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.539974689483643, "learning_rate": 1e-05, "loss": 0.5286, "mean_token_accuracy": 0.854604959487915, "num_tokens": 632682534.0, "step": 16666 }, { "epoch": 2.120213713268032, "ewc_loss": 0.15625, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 0.00013065338134765625, "grad_norm": 7.822537422180176, "learning_rate": 1e-05, "loss": 0.4784, "mean_token_accuracy": 0.8866613507270813, "num_tokens": 632717574.0, "step": 16667 }, { "epoch": 2.1203409235466224, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 5.967110633850098, "learning_rate": 1e-05, "loss": 0.4479, "mean_token_accuracy": 0.8813717365264893, "num_tokens": 632754113.0, "step": 16668 }, { "epoch": 2.120468133825213, "ewc_loss": 0.162109375, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 0.00013637542724609375, "grad_norm": 8.165472984313965, "learning_rate": 1e-05, "loss": 0.5487, "mean_token_accuracy": 0.8647032380104065, "num_tokens": 632797776.0, "step": 16669 }, { "epoch": 2.1205953441038035, "ewc_loss": 0.1455078125, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 0.00011968612670898438, "grad_norm": 6.806826591491699, "learning_rate": 1e-05, "loss": 0.4931, "mean_token_accuracy": 0.8778826594352722, "num_tokens": 632834841.0, "step": 16670 }, { "epoch": 2.120722554382394, "ewc_loss": 0.15234375, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 0.00012683868408203125, "grad_norm": 7.6955389976501465, "learning_rate": 1e-05, "loss": 0.4838, "mean_token_accuracy": 0.8854861259460449, "num_tokens": 632872698.0, "step": 16671 }, { "epoch": 2.1208497646609845, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 0.00011920928955078125, "grad_norm": 6.999523162841797, "learning_rate": 1e-05, "loss": 0.5051, "mean_token_accuracy": 0.8757196664810181, "num_tokens": 632911067.0, "step": 16672 }, { "epoch": 2.120976974939575, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 0.00011110305786132812, "grad_norm": 7.14990758895874, "learning_rate": 1e-05, "loss": 0.4771, "mean_token_accuracy": 0.8829354643821716, "num_tokens": 632952037.0, "step": 16673 }, { "epoch": 2.1211041852181656, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 0.00011014938354492188, "grad_norm": 6.812643527984619, "learning_rate": 1e-05, "loss": 0.5004, "mean_token_accuracy": 0.8726469874382019, "num_tokens": 632984667.0, "step": 16674 }, { "epoch": 2.121231395496756, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 8.78373908996582, "learning_rate": 1e-05, "loss": 0.5096, "mean_token_accuracy": 0.8685946464538574, "num_tokens": 633027165.0, "step": 16675 }, { "epoch": 2.1213586057753466, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 0.0001125335693359375, "grad_norm": 6.845117568969727, "learning_rate": 1e-05, "loss": 0.4578, "mean_token_accuracy": 0.8903223872184753, "num_tokens": 633068636.0, "step": 16676 }, { "epoch": 2.121485816053937, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.870504856109619, "learning_rate": 1e-05, "loss": 0.4565, "mean_token_accuracy": 0.8865954875946045, "num_tokens": 633103546.0, "step": 16677 }, { "epoch": 2.1216130263325277, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.468984127044678, "learning_rate": 1e-05, "loss": 0.5072, "mean_token_accuracy": 0.8673741817474365, "num_tokens": 633141529.0, "step": 16678 }, { "epoch": 2.121740236611118, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.758011341094971, "learning_rate": 1e-05, "loss": 0.4495, "mean_token_accuracy": 0.8883143067359924, "num_tokens": 633177324.0, "step": 16679 }, { "epoch": 2.1218674468897087, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.340707302093506, "learning_rate": 1e-05, "loss": 0.47, "mean_token_accuracy": 0.8793964982032776, "num_tokens": 633216726.0, "step": 16680 }, { "epoch": 2.1219946571682993, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.618139266967773, "learning_rate": 1e-05, "loss": 0.4924, "mean_token_accuracy": 0.8737491965293884, "num_tokens": 633255583.0, "step": 16681 }, { "epoch": 2.12212186744689, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.207571983337402, "learning_rate": 1e-05, "loss": 0.4624, "mean_token_accuracy": 0.8824781179428101, "num_tokens": 633295167.0, "step": 16682 }, { "epoch": 2.1222490777254803, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.6628007888793945, "learning_rate": 1e-05, "loss": 0.5214, "mean_token_accuracy": 0.8630690574645996, "num_tokens": 633336662.0, "step": 16683 }, { "epoch": 2.122376288004071, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.0044169425964355, "learning_rate": 1e-05, "loss": 0.4488, "mean_token_accuracy": 0.8829156756401062, "num_tokens": 633379826.0, "step": 16684 }, { "epoch": 2.1225034982826614, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.850152969360352, "learning_rate": 1e-05, "loss": 0.5136, "mean_token_accuracy": 0.8675228953361511, "num_tokens": 633421447.0, "step": 16685 }, { "epoch": 2.122630708561252, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 5.854936122894287, "learning_rate": 1e-05, "loss": 0.4466, "mean_token_accuracy": 0.887317419052124, "num_tokens": 633462704.0, "step": 16686 }, { "epoch": 2.1227579188398424, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 0.00011396408081054688, "grad_norm": 7.170196056365967, "learning_rate": 1e-05, "loss": 0.4833, "mean_token_accuracy": 0.8774694204330444, "num_tokens": 633496452.0, "step": 16687 }, { "epoch": 2.122885129118433, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 5.828418731689453, "learning_rate": 1e-05, "loss": 0.4133, "mean_token_accuracy": 0.8961286544799805, "num_tokens": 633534560.0, "step": 16688 }, { "epoch": 2.1230123393970235, "ewc_loss": 0.1484375, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 0.0001220703125, "grad_norm": 7.538784503936768, "learning_rate": 1e-05, "loss": 0.4837, "mean_token_accuracy": 0.8832694888114929, "num_tokens": 633572429.0, "step": 16689 }, { "epoch": 2.1231395496756136, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.14858865737915, "learning_rate": 1e-05, "loss": 0.497, "mean_token_accuracy": 0.8701916933059692, "num_tokens": 633603588.0, "step": 16690 }, { "epoch": 2.123266759954204, "ewc_loss": 0.1484375, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.625096797943115, "learning_rate": 1e-05, "loss": 0.489, "mean_token_accuracy": 0.8811039924621582, "num_tokens": 633640975.0, "step": 16691 }, { "epoch": 2.1233939702327946, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 6.572293758392334, "learning_rate": 1e-05, "loss": 0.4751, "mean_token_accuracy": 0.882428765296936, "num_tokens": 633684340.0, "step": 16692 }, { "epoch": 2.123521180511385, "ewc_loss": 0.140625, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 0.00011444091796875, "grad_norm": 7.160515785217285, "learning_rate": 1e-05, "loss": 0.4797, "mean_token_accuracy": 0.8805388808250427, "num_tokens": 633720969.0, "step": 16693 }, { "epoch": 2.1236483907899757, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 6.719242095947266, "learning_rate": 1e-05, "loss": 0.4362, "mean_token_accuracy": 0.8898468613624573, "num_tokens": 633751726.0, "step": 16694 }, { "epoch": 2.123775601068566, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.5794291496276855, "learning_rate": 1e-05, "loss": 0.4593, "mean_token_accuracy": 0.884026050567627, "num_tokens": 633784645.0, "step": 16695 }, { "epoch": 2.1239028113471567, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.777482032775879, "learning_rate": 1e-05, "loss": 0.434, "mean_token_accuracy": 0.8908366560935974, "num_tokens": 633816028.0, "step": 16696 }, { "epoch": 2.1240300216257473, "ewc_loss": 0.11474609375, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.055780410766602, "learning_rate": 1e-05, "loss": 0.485, "mean_token_accuracy": 0.8699233531951904, "num_tokens": 633859018.0, "step": 16697 }, { "epoch": 2.124157231904338, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 6.987497329711914, "learning_rate": 1e-05, "loss": 0.4938, "mean_token_accuracy": 0.8728675842285156, "num_tokens": 633897549.0, "step": 16698 }, { "epoch": 2.1242844421829283, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.755115509033203, "learning_rate": 1e-05, "loss": 0.4233, "mean_token_accuracy": 0.8892934322357178, "num_tokens": 633933111.0, "step": 16699 }, { "epoch": 2.124411652461519, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 0.00011777877807617188, "grad_norm": 7.370305061340332, "learning_rate": 1e-05, "loss": 0.5142, "mean_token_accuracy": 0.8668489456176758, "num_tokens": 633972409.0, "step": 16700 }, { "epoch": 2.1245388627401094, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.809304714202881, "learning_rate": 1e-05, "loss": 0.517, "mean_token_accuracy": 0.8602371215820312, "num_tokens": 634007112.0, "step": 16701 }, { "epoch": 2.1246660730187, "ewc_loss": 0.154296875, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 0.0001277923583984375, "grad_norm": 7.741833686828613, "learning_rate": 1e-05, "loss": 0.4611, "mean_token_accuracy": 0.8918119668960571, "num_tokens": 634043509.0, "step": 16702 }, { "epoch": 2.1247932832972904, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.276216506958008, "learning_rate": 1e-05, "loss": 0.4781, "mean_token_accuracy": 0.8782728314399719, "num_tokens": 634081883.0, "step": 16703 }, { "epoch": 2.124920493575881, "ewc_loss": 0.150390625, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 0.00012493133544921875, "grad_norm": 7.664271831512451, "learning_rate": 1e-05, "loss": 0.4784, "mean_token_accuracy": 0.8883236050605774, "num_tokens": 634116203.0, "step": 16704 }, { "epoch": 2.1250477038544715, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 0.00011014938354492188, "grad_norm": 6.634263515472412, "learning_rate": 1e-05, "loss": 0.4833, "mean_token_accuracy": 0.8800696134567261, "num_tokens": 634152538.0, "step": 16705 }, { "epoch": 2.125174914133062, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 0.0001163482666015625, "grad_norm": 7.264903545379639, "learning_rate": 1e-05, "loss": 0.4832, "mean_token_accuracy": 0.8800902962684631, "num_tokens": 634188563.0, "step": 16706 }, { "epoch": 2.1253021244116526, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.6468729972839355, "learning_rate": 1e-05, "loss": 0.514, "mean_token_accuracy": 0.8648555278778076, "num_tokens": 634235164.0, "step": 16707 }, { "epoch": 2.125429334690243, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.900282382965088, "learning_rate": 1e-05, "loss": 0.4622, "mean_token_accuracy": 0.8852683305740356, "num_tokens": 634275136.0, "step": 16708 }, { "epoch": 2.1255565449688336, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.503523349761963, "learning_rate": 1e-05, "loss": 0.4608, "mean_token_accuracy": 0.8834251165390015, "num_tokens": 634312056.0, "step": 16709 }, { "epoch": 2.125683755247424, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.672342777252197, "learning_rate": 1e-05, "loss": 0.5276, "mean_token_accuracy": 0.8620911836624146, "num_tokens": 634345737.0, "step": 16710 }, { "epoch": 2.1258109655260147, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.334367752075195, "learning_rate": 1e-05, "loss": 0.4741, "mean_token_accuracy": 0.876892626285553, "num_tokens": 634381122.0, "step": 16711 }, { "epoch": 2.125938175804605, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.585030555725098, "learning_rate": 1e-05, "loss": 0.4813, "mean_token_accuracy": 0.8770169019699097, "num_tokens": 634420529.0, "step": 16712 }, { "epoch": 2.1260653860831957, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.140240669250488, "learning_rate": 1e-05, "loss": 0.4504, "mean_token_accuracy": 0.8815177083015442, "num_tokens": 634460540.0, "step": 16713 }, { "epoch": 2.1261925963617863, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.650018692016602, "learning_rate": 1e-05, "loss": 0.5113, "mean_token_accuracy": 0.8686598539352417, "num_tokens": 634497818.0, "step": 16714 }, { "epoch": 2.1263198066403763, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.995297908782959, "learning_rate": 1e-05, "loss": 0.4838, "mean_token_accuracy": 0.8691331148147583, "num_tokens": 634532262.0, "step": 16715 }, { "epoch": 2.126447016918967, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.796444416046143, "learning_rate": 1e-05, "loss": 0.4867, "mean_token_accuracy": 0.8770561218261719, "num_tokens": 634567316.0, "step": 16716 }, { "epoch": 2.1265742271975574, "ewc_loss": 0.10986328125, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.820908069610596, "learning_rate": 1e-05, "loss": 0.4766, "mean_token_accuracy": 0.8743036985397339, "num_tokens": 634608603.0, "step": 16717 }, { "epoch": 2.126701437476148, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 0.00011014938354492188, "grad_norm": 7.085587501525879, "learning_rate": 1e-05, "loss": 0.4955, "mean_token_accuracy": 0.8759596347808838, "num_tokens": 634646539.0, "step": 16718 }, { "epoch": 2.1268286477547385, "ewc_loss": 0.11083984375, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.8217244148254395, "learning_rate": 1e-05, "loss": 0.434, "mean_token_accuracy": 0.8831813335418701, "num_tokens": 634673998.0, "step": 16719 }, { "epoch": 2.126955858033329, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 0.00011777877807617188, "grad_norm": 7.375357151031494, "learning_rate": 1e-05, "loss": 0.4832, "mean_token_accuracy": 0.8793585300445557, "num_tokens": 634707609.0, "step": 16720 }, { "epoch": 2.1270830683119195, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.030944347381592, "learning_rate": 1e-05, "loss": 0.458, "mean_token_accuracy": 0.8810746669769287, "num_tokens": 634742893.0, "step": 16721 }, { "epoch": 2.12721027859051, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 0.00012063980102539062, "grad_norm": 7.439730167388916, "learning_rate": 1e-05, "loss": 0.4898, "mean_token_accuracy": 0.8831338882446289, "num_tokens": 634780887.0, "step": 16722 }, { "epoch": 2.1273374888691006, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.379334449768066, "learning_rate": 1e-05, "loss": 0.4713, "mean_token_accuracy": 0.8790180087089539, "num_tokens": 634819512.0, "step": 16723 }, { "epoch": 2.127464699147691, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 0.0001125335693359375, "grad_norm": 7.122115612030029, "learning_rate": 1e-05, "loss": 0.5273, "mean_token_accuracy": 0.8627681136131287, "num_tokens": 634857213.0, "step": 16724 }, { "epoch": 2.1275919094262816, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.55471134185791, "learning_rate": 1e-05, "loss": 0.5097, "mean_token_accuracy": 0.8704237937927246, "num_tokens": 634899221.0, "step": 16725 }, { "epoch": 2.127719119704872, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.6223554611206055, "learning_rate": 1e-05, "loss": 0.4255, "mean_token_accuracy": 0.8938505053520203, "num_tokens": 634943486.0, "step": 16726 }, { "epoch": 2.1278463299834627, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.65133810043335, "learning_rate": 1e-05, "loss": 0.5061, "mean_token_accuracy": 0.8704257011413574, "num_tokens": 634981303.0, "step": 16727 }, { "epoch": 2.127973540262053, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 6.127265453338623, "learning_rate": 1e-05, "loss": 0.4748, "mean_token_accuracy": 0.8761964440345764, "num_tokens": 635021411.0, "step": 16728 }, { "epoch": 2.1281007505406437, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.827546119689941, "learning_rate": 1e-05, "loss": 0.4599, "mean_token_accuracy": 0.886676549911499, "num_tokens": 635063449.0, "step": 16729 }, { "epoch": 2.1282279608192343, "ewc_loss": 0.10888671875, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.777341365814209, "learning_rate": 1e-05, "loss": 0.4982, "mean_token_accuracy": 0.8651995658874512, "num_tokens": 635107052.0, "step": 16730 }, { "epoch": 2.128355171097825, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 0.00011301040649414062, "grad_norm": 7.200134754180908, "learning_rate": 1e-05, "loss": 0.4575, "mean_token_accuracy": 0.8876571655273438, "num_tokens": 635141036.0, "step": 16731 }, { "epoch": 2.1284823813764153, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.691193103790283, "learning_rate": 1e-05, "loss": 0.4818, "mean_token_accuracy": 0.8687504529953003, "num_tokens": 635181060.0, "step": 16732 }, { "epoch": 2.128609591655006, "ewc_loss": 0.150390625, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 0.0001239776611328125, "grad_norm": 7.632204055786133, "learning_rate": 1e-05, "loss": 0.5013, "mean_token_accuracy": 0.8768434524536133, "num_tokens": 635222679.0, "step": 16733 }, { "epoch": 2.1287368019335964, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.036867618560791, "learning_rate": 1e-05, "loss": 0.4544, "mean_token_accuracy": 0.8822184801101685, "num_tokens": 635258351.0, "step": 16734 }, { "epoch": 2.128864012212187, "ewc_loss": 0.154296875, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 0.00012874603271484375, "grad_norm": 7.775467395782471, "learning_rate": 1e-05, "loss": 0.5757, "mean_token_accuracy": 0.8600264191627502, "num_tokens": 635300647.0, "step": 16735 }, { "epoch": 2.1289912224907774, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 6.5398712158203125, "learning_rate": 1e-05, "loss": 0.4903, "mean_token_accuracy": 0.8774878978729248, "num_tokens": 635335239.0, "step": 16736 }, { "epoch": 2.129118432769368, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 0.00011920928955078125, "grad_norm": 7.429093360900879, "learning_rate": 1e-05, "loss": 0.4859, "mean_token_accuracy": 0.882632851600647, "num_tokens": 635370786.0, "step": 16737 }, { "epoch": 2.129245643047958, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 6.68614387512207, "learning_rate": 1e-05, "loss": 0.4935, "mean_token_accuracy": 0.8782638907432556, "num_tokens": 635410033.0, "step": 16738 }, { "epoch": 2.129372853326549, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 7.028045177459717, "learning_rate": 1e-05, "loss": 0.4512, "mean_token_accuracy": 0.8899981379508972, "num_tokens": 635447157.0, "step": 16739 }, { "epoch": 2.129500063605139, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.591415882110596, "learning_rate": 1e-05, "loss": 0.4623, "mean_token_accuracy": 0.8856995105743408, "num_tokens": 635485580.0, "step": 16740 }, { "epoch": 2.1296272738837296, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.686845302581787, "learning_rate": 1e-05, "loss": 0.4272, "mean_token_accuracy": 0.8914546966552734, "num_tokens": 635520289.0, "step": 16741 }, { "epoch": 2.12975448416232, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.464155197143555, "learning_rate": 1e-05, "loss": 0.4493, "mean_token_accuracy": 0.8836672306060791, "num_tokens": 635552390.0, "step": 16742 }, { "epoch": 2.1298816944409107, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.505953788757324, "learning_rate": 1e-05, "loss": 0.4728, "mean_token_accuracy": 0.8777646422386169, "num_tokens": 635588281.0, "step": 16743 }, { "epoch": 2.130008904719501, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.286498546600342, "learning_rate": 1e-05, "loss": 0.5116, "mean_token_accuracy": 0.8646391034126282, "num_tokens": 635632229.0, "step": 16744 }, { "epoch": 2.1301361149980917, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.464157581329346, "learning_rate": 1e-05, "loss": 0.4576, "mean_token_accuracy": 0.8826490044593811, "num_tokens": 635671315.0, "step": 16745 }, { "epoch": 2.1302633252766823, "ewc_loss": 0.11474609375, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.107271671295166, "learning_rate": 1e-05, "loss": 0.4998, "mean_token_accuracy": 0.8734608888626099, "num_tokens": 635711765.0, "step": 16746 }, { "epoch": 2.130390535555273, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.546016693115234, "learning_rate": 1e-05, "loss": 0.5178, "mean_token_accuracy": 0.8604962825775146, "num_tokens": 635747710.0, "step": 16747 }, { "epoch": 2.1305177458338633, "ewc_loss": 0.11181640625, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.940272808074951, "learning_rate": 1e-05, "loss": 0.4985, "mean_token_accuracy": 0.8668118715286255, "num_tokens": 635781355.0, "step": 16748 }, { "epoch": 2.130644956112454, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.729249000549316, "learning_rate": 1e-05, "loss": 0.472, "mean_token_accuracy": 0.8784590363502502, "num_tokens": 635821340.0, "step": 16749 }, { "epoch": 2.1307721663910444, "ewc_loss": 0.109375, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.785181522369385, "learning_rate": 1e-05, "loss": 0.4865, "mean_token_accuracy": 0.8681408166885376, "num_tokens": 635857943.0, "step": 16750 }, { "epoch": 2.130899376669635, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 7.047616004943848, "learning_rate": 1e-05, "loss": 0.5075, "mean_token_accuracy": 0.8715879917144775, "num_tokens": 635896808.0, "step": 16751 }, { "epoch": 2.1310265869482254, "ewc_loss": 0.10986328125, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.70838737487793, "learning_rate": 1e-05, "loss": 0.4414, "mean_token_accuracy": 0.8836441040039062, "num_tokens": 635934952.0, "step": 16752 }, { "epoch": 2.131153797226816, "ewc_loss": 0.1455078125, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 0.00011968612670898438, "grad_norm": 7.474477291107178, "learning_rate": 1e-05, "loss": 0.5012, "mean_token_accuracy": 0.8789838552474976, "num_tokens": 635972968.0, "step": 16753 }, { "epoch": 2.1312810075054065, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 5.928907871246338, "learning_rate": 1e-05, "loss": 0.4161, "mean_token_accuracy": 0.8904026746749878, "num_tokens": 636004761.0, "step": 16754 }, { "epoch": 2.131408217783997, "ewc_loss": 0.15234375, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 0.00012683868408203125, "grad_norm": 7.694783687591553, "learning_rate": 1e-05, "loss": 0.5176, "mean_token_accuracy": 0.8714268207550049, "num_tokens": 636043003.0, "step": 16755 }, { "epoch": 2.1315354280625876, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.407683372497559, "learning_rate": 1e-05, "loss": 0.4745, "mean_token_accuracy": 0.8773065805435181, "num_tokens": 636083580.0, "step": 16756 }, { "epoch": 2.131662638341178, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 0.0001201629638671875, "grad_norm": 7.432556629180908, "learning_rate": 1e-05, "loss": 0.4912, "mean_token_accuracy": 0.8820573687553406, "num_tokens": 636124248.0, "step": 16757 }, { "epoch": 2.1317898486197686, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 6.666274070739746, "learning_rate": 1e-05, "loss": 0.5113, "mean_token_accuracy": 0.8690075874328613, "num_tokens": 636164665.0, "step": 16758 }, { "epoch": 2.131917058898359, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.91970682144165, "learning_rate": 1e-05, "loss": 0.4856, "mean_token_accuracy": 0.8746165037155151, "num_tokens": 636201526.0, "step": 16759 }, { "epoch": 2.1320442691769497, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.674590110778809, "learning_rate": 1e-05, "loss": 0.4865, "mean_token_accuracy": 0.8766530752182007, "num_tokens": 636247793.0, "step": 16760 }, { "epoch": 2.13217147945554, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.427767753601074, "learning_rate": 1e-05, "loss": 0.4588, "mean_token_accuracy": 0.8852517604827881, "num_tokens": 636289989.0, "step": 16761 }, { "epoch": 2.1322986897341307, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.707112789154053, "learning_rate": 1e-05, "loss": 0.5278, "mean_token_accuracy": 0.8647065758705139, "num_tokens": 636328729.0, "step": 16762 }, { "epoch": 2.132425900012721, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 6.081618309020996, "learning_rate": 1e-05, "loss": 0.4697, "mean_token_accuracy": 0.8753716945648193, "num_tokens": 636362593.0, "step": 16763 }, { "epoch": 2.1325531102913113, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.841329574584961, "learning_rate": 1e-05, "loss": 0.4635, "mean_token_accuracy": 0.8867288827896118, "num_tokens": 636402790.0, "step": 16764 }, { "epoch": 2.132680320569902, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.8558349609375, "learning_rate": 1e-05, "loss": 0.4707, "mean_token_accuracy": 0.8746678829193115, "num_tokens": 636443052.0, "step": 16765 }, { "epoch": 2.1328075308484924, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 0.00010967254638671875, "grad_norm": 7.083082675933838, "learning_rate": 1e-05, "loss": 0.4822, "mean_token_accuracy": 0.878399133682251, "num_tokens": 636481345.0, "step": 16766 }, { "epoch": 2.132934741127083, "ewc_loss": 0.11083984375, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.805725574493408, "learning_rate": 1e-05, "loss": 0.4599, "mean_token_accuracy": 0.876514732837677, "num_tokens": 636516638.0, "step": 16767 }, { "epoch": 2.1330619514056735, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 0.00011682510375976562, "grad_norm": 7.349662780761719, "learning_rate": 1e-05, "loss": 0.4435, "mean_token_accuracy": 0.8915275931358337, "num_tokens": 636551606.0, "step": 16768 }, { "epoch": 2.133189161684264, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 5.9580397605896, "learning_rate": 1e-05, "loss": 0.49, "mean_token_accuracy": 0.8707269430160522, "num_tokens": 636589298.0, "step": 16769 }, { "epoch": 2.1333163719628545, "ewc_loss": 0.1484375, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 0.0001220703125, "grad_norm": 7.51248836517334, "learning_rate": 1e-05, "loss": 0.4881, "mean_token_accuracy": 0.8773311376571655, "num_tokens": 636622010.0, "step": 16770 }, { "epoch": 2.133443582241445, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.244068622589111, "learning_rate": 1e-05, "loss": 0.4647, "mean_token_accuracy": 0.8813893795013428, "num_tokens": 636665355.0, "step": 16771 }, { "epoch": 2.1335707925200356, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 0.000118255615234375, "grad_norm": 7.387705326080322, "learning_rate": 1e-05, "loss": 0.4839, "mean_token_accuracy": 0.8848174810409546, "num_tokens": 636706135.0, "step": 16772 }, { "epoch": 2.133698002798626, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.443572044372559, "learning_rate": 1e-05, "loss": 0.4336, "mean_token_accuracy": 0.8910976648330688, "num_tokens": 636744587.0, "step": 16773 }, { "epoch": 2.1338252130772166, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 0.00011157989501953125, "grad_norm": 7.0822625160217285, "learning_rate": 1e-05, "loss": 0.4413, "mean_token_accuracy": 0.888590931892395, "num_tokens": 636784175.0, "step": 16774 }, { "epoch": 2.133952423355807, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.471123218536377, "learning_rate": 1e-05, "loss": 0.5005, "mean_token_accuracy": 0.8667406439781189, "num_tokens": 636821752.0, "step": 16775 }, { "epoch": 2.1340796336343977, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.785059928894043, "learning_rate": 1e-05, "loss": 0.5438, "mean_token_accuracy": 0.8628610372543335, "num_tokens": 636855507.0, "step": 16776 }, { "epoch": 2.134206843912988, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.409012794494629, "learning_rate": 1e-05, "loss": 0.5061, "mean_token_accuracy": 0.8706507086753845, "num_tokens": 636895401.0, "step": 16777 }, { "epoch": 2.1343340541915787, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.53752326965332, "learning_rate": 1e-05, "loss": 0.5292, "mean_token_accuracy": 0.8589048385620117, "num_tokens": 636940225.0, "step": 16778 }, { "epoch": 2.1344612644701693, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.318601608276367, "learning_rate": 1e-05, "loss": 0.4826, "mean_token_accuracy": 0.8733880519866943, "num_tokens": 636976685.0, "step": 16779 }, { "epoch": 2.13458847474876, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.408310890197754, "learning_rate": 1e-05, "loss": 0.4766, "mean_token_accuracy": 0.8770087957382202, "num_tokens": 637018170.0, "step": 16780 }, { "epoch": 2.1347156850273503, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.2161149978637695, "learning_rate": 1e-05, "loss": 0.4559, "mean_token_accuracy": 0.8800489902496338, "num_tokens": 637055115.0, "step": 16781 }, { "epoch": 2.134842895305941, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.397568225860596, "learning_rate": 1e-05, "loss": 0.5039, "mean_token_accuracy": 0.8651666045188904, "num_tokens": 637093964.0, "step": 16782 }, { "epoch": 2.1349701055845314, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 6.062639236450195, "learning_rate": 1e-05, "loss": 0.4055, "mean_token_accuracy": 0.8980982899665833, "num_tokens": 637135242.0, "step": 16783 }, { "epoch": 2.135097315863122, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.480425834655762, "learning_rate": 1e-05, "loss": 0.4771, "mean_token_accuracy": 0.8762226104736328, "num_tokens": 637173407.0, "step": 16784 }, { "epoch": 2.1352245261417124, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.9253106117248535, "learning_rate": 1e-05, "loss": 0.4677, "mean_token_accuracy": 0.8760430812835693, "num_tokens": 637207613.0, "step": 16785 }, { "epoch": 2.135351736420303, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.6483154296875, "learning_rate": 1e-05, "loss": 0.426, "mean_token_accuracy": 0.8916188478469849, "num_tokens": 637242010.0, "step": 16786 }, { "epoch": 2.1354789466988935, "ewc_loss": 0.109375, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.758832931518555, "learning_rate": 1e-05, "loss": 0.4375, "mean_token_accuracy": 0.8823148012161255, "num_tokens": 637284034.0, "step": 16787 }, { "epoch": 2.1356061569774836, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.980535984039307, "learning_rate": 1e-05, "loss": 0.4613, "mean_token_accuracy": 0.884392261505127, "num_tokens": 637325657.0, "step": 16788 }, { "epoch": 2.135733367256074, "ewc_loss": 0.10888671875, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.64807653427124, "learning_rate": 1e-05, "loss": 0.4242, "mean_token_accuracy": 0.8886127471923828, "num_tokens": 637360946.0, "step": 16789 }, { "epoch": 2.1358605775346646, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 0.00011873245239257812, "grad_norm": 7.447232723236084, "learning_rate": 1e-05, "loss": 0.5098, "mean_token_accuracy": 0.8742395043373108, "num_tokens": 637397540.0, "step": 16790 }, { "epoch": 2.135987787813255, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 5.817989826202393, "learning_rate": 1e-05, "loss": 0.4522, "mean_token_accuracy": 0.8830727338790894, "num_tokens": 637443148.0, "step": 16791 }, { "epoch": 2.1361149980918457, "ewc_loss": 0.154296875, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 0.00012874603271484375, "grad_norm": 7.781094074249268, "learning_rate": 1e-05, "loss": 0.4749, "mean_token_accuracy": 0.8864524364471436, "num_tokens": 637480796.0, "step": 16792 }, { "epoch": 2.136242208370436, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.365461349487305, "learning_rate": 1e-05, "loss": 0.503, "mean_token_accuracy": 0.8725777864456177, "num_tokens": 637525465.0, "step": 16793 }, { "epoch": 2.1363694186490267, "ewc_loss": 0.1484375, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.573670387268066, "learning_rate": 1e-05, "loss": 0.4823, "mean_token_accuracy": 0.8802733421325684, "num_tokens": 637563235.0, "step": 16794 }, { "epoch": 2.1364966289276173, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 0.00011110305786132812, "grad_norm": 6.6970624923706055, "learning_rate": 1e-05, "loss": 0.4654, "mean_token_accuracy": 0.8848421573638916, "num_tokens": 637600061.0, "step": 16795 }, { "epoch": 2.136623839206208, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 7.0607991218566895, "learning_rate": 1e-05, "loss": 0.4407, "mean_token_accuracy": 0.892634928226471, "num_tokens": 637640877.0, "step": 16796 }, { "epoch": 2.1367510494847983, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 6.701332092285156, "learning_rate": 1e-05, "loss": 0.4879, "mean_token_accuracy": 0.87410569190979, "num_tokens": 637679248.0, "step": 16797 }, { "epoch": 2.136878259763389, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.582701683044434, "learning_rate": 1e-05, "loss": 0.478, "mean_token_accuracy": 0.87326580286026, "num_tokens": 637715973.0, "step": 16798 }, { "epoch": 2.1370054700419794, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.675292491912842, "learning_rate": 1e-05, "loss": 0.4645, "mean_token_accuracy": 0.883819043636322, "num_tokens": 637753358.0, "step": 16799 }, { "epoch": 2.13713268032057, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.217021942138672, "learning_rate": 1e-05, "loss": 0.473, "mean_token_accuracy": 0.8780456781387329, "num_tokens": 637798608.0, "step": 16800 }, { "epoch": 2.1372598905991604, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.732213020324707, "learning_rate": 1e-05, "loss": 0.4746, "mean_token_accuracy": 0.8799106478691101, "num_tokens": 637837173.0, "step": 16801 }, { "epoch": 2.137387100877751, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.99224328994751, "learning_rate": 1e-05, "loss": 0.4501, "mean_token_accuracy": 0.8849796056747437, "num_tokens": 637874072.0, "step": 16802 }, { "epoch": 2.1375143111563415, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.847090244293213, "learning_rate": 1e-05, "loss": 0.5185, "mean_token_accuracy": 0.8690627813339233, "num_tokens": 637909369.0, "step": 16803 }, { "epoch": 2.137641521434932, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.863919258117676, "learning_rate": 1e-05, "loss": 0.4288, "mean_token_accuracy": 0.8887420892715454, "num_tokens": 637946526.0, "step": 16804 }, { "epoch": 2.1377687317135226, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 7.045315742492676, "learning_rate": 1e-05, "loss": 0.4962, "mean_token_accuracy": 0.8734337091445923, "num_tokens": 637984476.0, "step": 16805 }, { "epoch": 2.137895941992113, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.84268045425415, "learning_rate": 1e-05, "loss": 0.4723, "mean_token_accuracy": 0.8738024234771729, "num_tokens": 638019098.0, "step": 16806 }, { "epoch": 2.1380231522707036, "ewc_loss": 0.140625, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 0.00011491775512695312, "grad_norm": 7.263693809509277, "learning_rate": 1e-05, "loss": 0.4868, "mean_token_accuracy": 0.8783336281776428, "num_tokens": 638059049.0, "step": 16807 }, { "epoch": 2.138150362549294, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 5.946060657501221, "learning_rate": 1e-05, "loss": 0.4635, "mean_token_accuracy": 0.8785274028778076, "num_tokens": 638095887.0, "step": 16808 }, { "epoch": 2.1382775728278847, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 0.00011730194091796875, "grad_norm": 7.400614261627197, "learning_rate": 1e-05, "loss": 0.5624, "mean_token_accuracy": 0.8592368364334106, "num_tokens": 638141855.0, "step": 16809 }, { "epoch": 2.138404783106475, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.160699844360352, "learning_rate": 1e-05, "loss": 0.4381, "mean_token_accuracy": 0.8901050686836243, "num_tokens": 638184491.0, "step": 16810 }, { "epoch": 2.1385319933850653, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 0.00011682510375976562, "grad_norm": 7.3469390869140625, "learning_rate": 1e-05, "loss": 0.4502, "mean_token_accuracy": 0.8897078633308411, "num_tokens": 638218341.0, "step": 16811 }, { "epoch": 2.1386592036636562, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.352668285369873, "learning_rate": 1e-05, "loss": 0.497, "mean_token_accuracy": 0.8721685409545898, "num_tokens": 638257422.0, "step": 16812 }, { "epoch": 2.1387864139422463, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 0.00011157989501953125, "grad_norm": 7.09376859664917, "learning_rate": 1e-05, "loss": 0.4722, "mean_token_accuracy": 0.8816927075386047, "num_tokens": 638298053.0, "step": 16813 }, { "epoch": 2.138913624220837, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.415774822235107, "learning_rate": 1e-05, "loss": 0.535, "mean_token_accuracy": 0.8604482412338257, "num_tokens": 638337870.0, "step": 16814 }, { "epoch": 2.1390408344994274, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.7847676277160645, "learning_rate": 1e-05, "loss": 0.4794, "mean_token_accuracy": 0.8768930435180664, "num_tokens": 638371587.0, "step": 16815 }, { "epoch": 2.139168044778018, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.393862247467041, "learning_rate": 1e-05, "loss": 0.4764, "mean_token_accuracy": 0.8771753311157227, "num_tokens": 638416879.0, "step": 16816 }, { "epoch": 2.1392952550566084, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.5371413230896, "learning_rate": 1e-05, "loss": 0.4371, "mean_token_accuracy": 0.8875637054443359, "num_tokens": 638452809.0, "step": 16817 }, { "epoch": 2.139422465335199, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.347324848175049, "learning_rate": 1e-05, "loss": 0.482, "mean_token_accuracy": 0.8724944591522217, "num_tokens": 638487027.0, "step": 16818 }, { "epoch": 2.1395496756137895, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.362496376037598, "learning_rate": 1e-05, "loss": 0.4569, "mean_token_accuracy": 0.8817063570022583, "num_tokens": 638517922.0, "step": 16819 }, { "epoch": 2.13967688589238, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.274310111999512, "learning_rate": 1e-05, "loss": 0.5023, "mean_token_accuracy": 0.867607057094574, "num_tokens": 638562360.0, "step": 16820 }, { "epoch": 2.1398040961709706, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.284049987792969, "learning_rate": 1e-05, "loss": 0.4641, "mean_token_accuracy": 0.8750165700912476, "num_tokens": 638595649.0, "step": 16821 }, { "epoch": 2.139931306449561, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.206695556640625, "learning_rate": 1e-05, "loss": 0.4679, "mean_token_accuracy": 0.8805931210517883, "num_tokens": 638632588.0, "step": 16822 }, { "epoch": 2.1400585167281516, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.25678014755249, "learning_rate": 1e-05, "loss": 0.457, "mean_token_accuracy": 0.8814469575881958, "num_tokens": 638663000.0, "step": 16823 }, { "epoch": 2.140185727006742, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.12491512298584, "learning_rate": 1e-05, "loss": 0.4704, "mean_token_accuracy": 0.8757089376449585, "num_tokens": 638703737.0, "step": 16824 }, { "epoch": 2.1403129372853327, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.282402038574219, "learning_rate": 1e-05, "loss": 0.5134, "mean_token_accuracy": 0.8643242120742798, "num_tokens": 638745526.0, "step": 16825 }, { "epoch": 2.140440147563923, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 6.0284833908081055, "learning_rate": 1e-05, "loss": 0.4446, "mean_token_accuracy": 0.884503185749054, "num_tokens": 638780865.0, "step": 16826 }, { "epoch": 2.1405673578425137, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.366616249084473, "learning_rate": 1e-05, "loss": 0.4806, "mean_token_accuracy": 0.8755528926849365, "num_tokens": 638822274.0, "step": 16827 }, { "epoch": 2.1406945681211043, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.908474445343018, "learning_rate": 1e-05, "loss": 0.4957, "mean_token_accuracy": 0.8670517206192017, "num_tokens": 638859077.0, "step": 16828 }, { "epoch": 2.140821778399695, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.532266139984131, "learning_rate": 1e-05, "loss": 0.4694, "mean_token_accuracy": 0.878970742225647, "num_tokens": 638898055.0, "step": 16829 }, { "epoch": 2.1409489886782853, "ewc_loss": 0.109375, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.737611293792725, "learning_rate": 1e-05, "loss": 0.5078, "mean_token_accuracy": 0.8647077083587646, "num_tokens": 638939597.0, "step": 16830 }, { "epoch": 2.141076198956876, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.848877906799316, "learning_rate": 1e-05, "loss": 0.4934, "mean_token_accuracy": 0.8723299503326416, "num_tokens": 638978268.0, "step": 16831 }, { "epoch": 2.1412034092354664, "ewc_loss": 0.107421875, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 8.153915405273438e-05, "grad_norm": 5.553760528564453, "learning_rate": 1e-05, "loss": 0.4174, "mean_token_accuracy": 0.8907262682914734, "num_tokens": 639010554.0, "step": 16832 }, { "epoch": 2.141330619514057, "ewc_loss": 0.1455078125, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 0.00011920928955078125, "grad_norm": 7.384944438934326, "learning_rate": 1e-05, "loss": 0.4273, "mean_token_accuracy": 0.9011355638504028, "num_tokens": 639046721.0, "step": 16833 }, { "epoch": 2.1414578297926474, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.611906051635742, "learning_rate": 1e-05, "loss": 0.4749, "mean_token_accuracy": 0.8721438646316528, "num_tokens": 639080078.0, "step": 16834 }, { "epoch": 2.141585040071238, "ewc_loss": 0.1611328125, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 0.00013446807861328125, "grad_norm": 7.986404895782471, "learning_rate": 1e-05, "loss": 0.5025, "mean_token_accuracy": 0.8795053362846375, "num_tokens": 639119469.0, "step": 16835 }, { "epoch": 2.141712250349828, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.223326683044434, "learning_rate": 1e-05, "loss": 0.486, "mean_token_accuracy": 0.8741501569747925, "num_tokens": 639159954.0, "step": 16836 }, { "epoch": 2.141839460628419, "ewc_loss": 0.1611328125, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 0.00013446807861328125, "grad_norm": 8.070301055908203, "learning_rate": 1e-05, "loss": 0.51, "mean_token_accuracy": 0.8810597658157349, "num_tokens": 639202455.0, "step": 16837 }, { "epoch": 2.141966670907009, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 0.0001201629638671875, "grad_norm": 6.889774322509766, "learning_rate": 1e-05, "loss": 0.4592, "mean_token_accuracy": 0.8875330686569214, "num_tokens": 639240889.0, "step": 16838 }, { "epoch": 2.1420938811855996, "ewc_loss": 0.1494140625, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.5566725730896, "learning_rate": 1e-05, "loss": 0.4942, "mean_token_accuracy": 0.8787217140197754, "num_tokens": 639275079.0, "step": 16839 }, { "epoch": 2.14222109146419, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 0.0001163482666015625, "grad_norm": 6.912247657775879, "learning_rate": 1e-05, "loss": 0.5125, "mean_token_accuracy": 0.8683172464370728, "num_tokens": 639308166.0, "step": 16840 }, { "epoch": 2.1423483017427807, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 7.104252815246582, "learning_rate": 1e-05, "loss": 0.4732, "mean_token_accuracy": 0.887123167514801, "num_tokens": 639341862.0, "step": 16841 }, { "epoch": 2.142475512021371, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.675378799438477, "learning_rate": 1e-05, "loss": 0.4634, "mean_token_accuracy": 0.8879278898239136, "num_tokens": 639381065.0, "step": 16842 }, { "epoch": 2.1426027222999617, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.7863616943359375, "learning_rate": 1e-05, "loss": 0.4527, "mean_token_accuracy": 0.8865987062454224, "num_tokens": 639421232.0, "step": 16843 }, { "epoch": 2.1427299325785523, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.483071327209473, "learning_rate": 1e-05, "loss": 0.4958, "mean_token_accuracy": 0.8728600740432739, "num_tokens": 639459462.0, "step": 16844 }, { "epoch": 2.142857142857143, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.611865520477295, "learning_rate": 1e-05, "loss": 0.4785, "mean_token_accuracy": 0.8786911964416504, "num_tokens": 639499759.0, "step": 16845 }, { "epoch": 2.1429843531357333, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.294187545776367, "learning_rate": 1e-05, "loss": 0.4607, "mean_token_accuracy": 0.8804646730422974, "num_tokens": 639534285.0, "step": 16846 }, { "epoch": 2.143111563414324, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.563149452209473, "learning_rate": 1e-05, "loss": 0.5005, "mean_token_accuracy": 0.872745156288147, "num_tokens": 639576058.0, "step": 16847 }, { "epoch": 2.1432387736929144, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.1161088943481445, "learning_rate": 1e-05, "loss": 0.4628, "mean_token_accuracy": 0.8810950517654419, "num_tokens": 639610464.0, "step": 16848 }, { "epoch": 2.143365983971505, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.630425453186035, "learning_rate": 1e-05, "loss": 0.4291, "mean_token_accuracy": 0.8911415338516235, "num_tokens": 639645327.0, "step": 16849 }, { "epoch": 2.1434931942500954, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.9338836669921875, "learning_rate": 1e-05, "loss": 0.4913, "mean_token_accuracy": 0.8676163554191589, "num_tokens": 639684501.0, "step": 16850 }, { "epoch": 2.143620404528686, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.835076808929443, "learning_rate": 1e-05, "loss": 0.5039, "mean_token_accuracy": 0.8708810806274414, "num_tokens": 639721687.0, "step": 16851 }, { "epoch": 2.1437476148072765, "ewc_loss": 0.109375, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.760377883911133, "learning_rate": 1e-05, "loss": 0.5041, "mean_token_accuracy": 0.8603235483169556, "num_tokens": 639761503.0, "step": 16852 }, { "epoch": 2.143874825085867, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 0.00011301040649414062, "grad_norm": 7.214033126831055, "learning_rate": 1e-05, "loss": 0.4798, "mean_token_accuracy": 0.8812017440795898, "num_tokens": 639800938.0, "step": 16853 }, { "epoch": 2.1440020353644575, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.750140190124512, "learning_rate": 1e-05, "loss": 0.4093, "mean_token_accuracy": 0.8922255635261536, "num_tokens": 639835318.0, "step": 16854 }, { "epoch": 2.144129245643048, "ewc_loss": 0.150390625, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 0.0001239776611328125, "grad_norm": 7.645940780639648, "learning_rate": 1e-05, "loss": 0.4816, "mean_token_accuracy": 0.8834602236747742, "num_tokens": 639878155.0, "step": 16855 }, { "epoch": 2.1442564559216386, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.113749980926514, "learning_rate": 1e-05, "loss": 0.4617, "mean_token_accuracy": 0.8815034031867981, "num_tokens": 639920369.0, "step": 16856 }, { "epoch": 2.144383666200229, "ewc_loss": 0.15234375, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 0.000125885009765625, "grad_norm": 7.7294206619262695, "learning_rate": 1e-05, "loss": 0.4923, "mean_token_accuracy": 0.8805067539215088, "num_tokens": 639964690.0, "step": 16857 }, { "epoch": 2.1445108764788197, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 0.00011014938354492188, "grad_norm": 6.628653526306152, "learning_rate": 1e-05, "loss": 0.5263, "mean_token_accuracy": 0.8656631708145142, "num_tokens": 640002969.0, "step": 16858 }, { "epoch": 2.14463808675741, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 0.00011539459228515625, "grad_norm": 7.281600475311279, "learning_rate": 1e-05, "loss": 0.4541, "mean_token_accuracy": 0.8895968198776245, "num_tokens": 640044061.0, "step": 16859 }, { "epoch": 2.1447652970360007, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 6.763860702514648, "learning_rate": 1e-05, "loss": 0.4678, "mean_token_accuracy": 0.8836298584938049, "num_tokens": 640082803.0, "step": 16860 }, { "epoch": 2.144892507314591, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.720622539520264, "learning_rate": 1e-05, "loss": 0.4298, "mean_token_accuracy": 0.8961154222488403, "num_tokens": 640123223.0, "step": 16861 }, { "epoch": 2.1450197175931813, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.746922016143799, "learning_rate": 1e-05, "loss": 0.4302, "mean_token_accuracy": 0.8947353363037109, "num_tokens": 640160674.0, "step": 16862 }, { "epoch": 2.145146927871772, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.248952388763428, "learning_rate": 1e-05, "loss": 0.449, "mean_token_accuracy": 0.8863925933837891, "num_tokens": 640197347.0, "step": 16863 }, { "epoch": 2.1452741381503624, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.832115650177002, "learning_rate": 1e-05, "loss": 0.4842, "mean_token_accuracy": 0.8777926564216614, "num_tokens": 640235599.0, "step": 16864 }, { "epoch": 2.145401348428953, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.935699939727783, "learning_rate": 1e-05, "loss": 0.4719, "mean_token_accuracy": 0.8736327886581421, "num_tokens": 640277753.0, "step": 16865 }, { "epoch": 2.1455285587075434, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 7.051594257354736, "learning_rate": 1e-05, "loss": 0.4676, "mean_token_accuracy": 0.8829408288002014, "num_tokens": 640317486.0, "step": 16866 }, { "epoch": 2.145655768986134, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.831235408782959, "learning_rate": 1e-05, "loss": 0.4986, "mean_token_accuracy": 0.8630117177963257, "num_tokens": 640357240.0, "step": 16867 }, { "epoch": 2.1457829792647245, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 0.00011587142944335938, "grad_norm": 7.336738109588623, "learning_rate": 1e-05, "loss": 0.474, "mean_token_accuracy": 0.8836210370063782, "num_tokens": 640392429.0, "step": 16868 }, { "epoch": 2.145910189543315, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 5.938197135925293, "learning_rate": 1e-05, "loss": 0.5012, "mean_token_accuracy": 0.8655527830123901, "num_tokens": 640432529.0, "step": 16869 }, { "epoch": 2.1460373998219056, "ewc_loss": 0.1455078125, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 0.00011968612670898438, "grad_norm": 7.5280680656433105, "learning_rate": 1e-05, "loss": 0.4894, "mean_token_accuracy": 0.8789337873458862, "num_tokens": 640468443.0, "step": 16870 }, { "epoch": 2.146164610100496, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.238460540771484, "learning_rate": 1e-05, "loss": 0.5031, "mean_token_accuracy": 0.8687611222267151, "num_tokens": 640505647.0, "step": 16871 }, { "epoch": 2.1462918203790866, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 0.000118255615234375, "grad_norm": 7.435357093811035, "learning_rate": 1e-05, "loss": 0.5313, "mean_token_accuracy": 0.8644070625305176, "num_tokens": 640551180.0, "step": 16872 }, { "epoch": 2.146419030657677, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.445512294769287, "learning_rate": 1e-05, "loss": 0.4592, "mean_token_accuracy": 0.8849494457244873, "num_tokens": 640590939.0, "step": 16873 }, { "epoch": 2.1465462409362677, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 0.00011205673217773438, "grad_norm": 7.161508560180664, "learning_rate": 1e-05, "loss": 0.5269, "mean_token_accuracy": 0.8637586832046509, "num_tokens": 640625688.0, "step": 16874 }, { "epoch": 2.146673451214858, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.475164413452148, "learning_rate": 1e-05, "loss": 0.5536, "mean_token_accuracy": 0.8564245700836182, "num_tokens": 640669142.0, "step": 16875 }, { "epoch": 2.1468006614934487, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.8560380935668945, "learning_rate": 1e-05, "loss": 0.5064, "mean_token_accuracy": 0.8689626455307007, "num_tokens": 640713345.0, "step": 16876 }, { "epoch": 2.1469278717720393, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.4181623458862305, "learning_rate": 1e-05, "loss": 0.5377, "mean_token_accuracy": 0.8548756837844849, "num_tokens": 640750930.0, "step": 16877 }, { "epoch": 2.14705508205063, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.614280700683594, "learning_rate": 1e-05, "loss": 0.454, "mean_token_accuracy": 0.8838169574737549, "num_tokens": 640792029.0, "step": 16878 }, { "epoch": 2.1471822923292203, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.304157257080078, "learning_rate": 1e-05, "loss": 0.4557, "mean_token_accuracy": 0.8828038573265076, "num_tokens": 640833952.0, "step": 16879 }, { "epoch": 2.147309502607811, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.512112140655518, "learning_rate": 1e-05, "loss": 0.463, "mean_token_accuracy": 0.8768670558929443, "num_tokens": 640867980.0, "step": 16880 }, { "epoch": 2.1474367128864014, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.176054954528809, "learning_rate": 1e-05, "loss": 0.5142, "mean_token_accuracy": 0.8634642362594604, "num_tokens": 640913241.0, "step": 16881 }, { "epoch": 2.147563923164992, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.504077911376953, "learning_rate": 1e-05, "loss": 0.4263, "mean_token_accuracy": 0.8936718702316284, "num_tokens": 640946705.0, "step": 16882 }, { "epoch": 2.1476911334435824, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 6.037217617034912, "learning_rate": 1e-05, "loss": 0.4381, "mean_token_accuracy": 0.8858174681663513, "num_tokens": 640985658.0, "step": 16883 }, { "epoch": 2.147818343722173, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.595691204071045, "learning_rate": 1e-05, "loss": 0.496, "mean_token_accuracy": 0.8703067898750305, "num_tokens": 641025963.0, "step": 16884 }, { "epoch": 2.1479455540007635, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.889886856079102, "learning_rate": 1e-05, "loss": 0.4419, "mean_token_accuracy": 0.8855326771736145, "num_tokens": 641070853.0, "step": 16885 }, { "epoch": 2.1480727642793536, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.800812244415283, "learning_rate": 1e-05, "loss": 0.4645, "mean_token_accuracy": 0.8815763592720032, "num_tokens": 641106336.0, "step": 16886 }, { "epoch": 2.148199974557944, "ewc_loss": 0.109375, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.765950679779053, "learning_rate": 1e-05, "loss": 0.4092, "mean_token_accuracy": 0.8943912982940674, "num_tokens": 641147842.0, "step": 16887 }, { "epoch": 2.1483271848365346, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 0.00010967254638671875, "grad_norm": 7.117717266082764, "learning_rate": 1e-05, "loss": 0.4485, "mean_token_accuracy": 0.888450562953949, "num_tokens": 641186329.0, "step": 16888 }, { "epoch": 2.148454395115125, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.7806267738342285, "learning_rate": 1e-05, "loss": 0.4805, "mean_token_accuracy": 0.8721954822540283, "num_tokens": 641227892.0, "step": 16889 }, { "epoch": 2.1485816053937157, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 0.0001201629638671875, "grad_norm": 7.441976070404053, "learning_rate": 1e-05, "loss": 0.4711, "mean_token_accuracy": 0.8836398124694824, "num_tokens": 641265611.0, "step": 16890 }, { "epoch": 2.148708815672306, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.0510783195495605, "learning_rate": 1e-05, "loss": 0.5207, "mean_token_accuracy": 0.8598501086235046, "num_tokens": 641299564.0, "step": 16891 }, { "epoch": 2.1488360259508967, "ewc_loss": 0.1474609375, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 0.00012111663818359375, "grad_norm": 7.483849048614502, "learning_rate": 1e-05, "loss": 0.4905, "mean_token_accuracy": 0.8798626661300659, "num_tokens": 641335892.0, "step": 16892 }, { "epoch": 2.1489632362294873, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.396109580993652, "learning_rate": 1e-05, "loss": 0.4883, "mean_token_accuracy": 0.8766111135482788, "num_tokens": 641375470.0, "step": 16893 }, { "epoch": 2.149090446508078, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 0.00011205673217773438, "grad_norm": 7.155739784240723, "learning_rate": 1e-05, "loss": 0.4886, "mean_token_accuracy": 0.8785643577575684, "num_tokens": 641408793.0, "step": 16894 }, { "epoch": 2.1492176567866683, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.5683441162109375, "learning_rate": 1e-05, "loss": 0.4879, "mean_token_accuracy": 0.8741496801376343, "num_tokens": 641444949.0, "step": 16895 }, { "epoch": 2.149344867065259, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.6493754386901855, "learning_rate": 1e-05, "loss": 0.4759, "mean_token_accuracy": 0.8761011362075806, "num_tokens": 641482775.0, "step": 16896 }, { "epoch": 2.1494720773438494, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.628701210021973, "learning_rate": 1e-05, "loss": 0.4563, "mean_token_accuracy": 0.8839324712753296, "num_tokens": 641520385.0, "step": 16897 }, { "epoch": 2.14959928762244, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.176794052124023, "learning_rate": 1e-05, "loss": 0.4964, "mean_token_accuracy": 0.8648050427436829, "num_tokens": 641563885.0, "step": 16898 }, { "epoch": 2.1497264979010304, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.776656627655029, "learning_rate": 1e-05, "loss": 0.4707, "mean_token_accuracy": 0.8785083889961243, "num_tokens": 641600868.0, "step": 16899 }, { "epoch": 2.149853708179621, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.850555896759033, "learning_rate": 1e-05, "loss": 0.458, "mean_token_accuracy": 0.8766733407974243, "num_tokens": 641643936.0, "step": 16900 }, { "epoch": 2.1499809184582115, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 7.081847190856934, "learning_rate": 1e-05, "loss": 0.5018, "mean_token_accuracy": 0.8710721731185913, "num_tokens": 641680248.0, "step": 16901 }, { "epoch": 2.150108128736802, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.7421393394470215, "learning_rate": 1e-05, "loss": 0.4392, "mean_token_accuracy": 0.8853496313095093, "num_tokens": 641721280.0, "step": 16902 }, { "epoch": 2.1502353390153925, "ewc_loss": 0.1455078125, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 0.00011920928955078125, "grad_norm": 7.455604553222656, "learning_rate": 1e-05, "loss": 0.5008, "mean_token_accuracy": 0.8748347163200378, "num_tokens": 641760168.0, "step": 16903 }, { "epoch": 2.150362549293983, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 5.946571350097656, "learning_rate": 1e-05, "loss": 0.5139, "mean_token_accuracy": 0.8653217554092407, "num_tokens": 641798517.0, "step": 16904 }, { "epoch": 2.1504897595725736, "ewc_loss": 0.1513671875, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 0.00012493133544921875, "grad_norm": 7.634908676147461, "learning_rate": 1e-05, "loss": 0.5113, "mean_token_accuracy": 0.8724915385246277, "num_tokens": 641841486.0, "step": 16905 }, { "epoch": 2.150616969851164, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.347720623016357, "learning_rate": 1e-05, "loss": 0.5136, "mean_token_accuracy": 0.868445873260498, "num_tokens": 641880042.0, "step": 16906 }, { "epoch": 2.1507441801297547, "ewc_loss": 0.1455078125, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 0.00011968612670898438, "grad_norm": 7.432663917541504, "learning_rate": 1e-05, "loss": 0.4859, "mean_token_accuracy": 0.8803935647010803, "num_tokens": 641916069.0, "step": 16907 }, { "epoch": 2.150871390408345, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.5507121086120605, "learning_rate": 1e-05, "loss": 0.4945, "mean_token_accuracy": 0.876175045967102, "num_tokens": 641951298.0, "step": 16908 }, { "epoch": 2.1509986006869353, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 0.00011014938354492188, "grad_norm": 7.07504415512085, "learning_rate": 1e-05, "loss": 0.4949, "mean_token_accuracy": 0.8810312151908875, "num_tokens": 641987577.0, "step": 16909 }, { "epoch": 2.1511258109655262, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.499495983123779, "learning_rate": 1e-05, "loss": 0.5293, "mean_token_accuracy": 0.8619804382324219, "num_tokens": 642031914.0, "step": 16910 }, { "epoch": 2.1512530212441163, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.776968479156494, "learning_rate": 1e-05, "loss": 0.4865, "mean_token_accuracy": 0.8727222084999084, "num_tokens": 642069729.0, "step": 16911 }, { "epoch": 2.151380231522707, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.400291919708252, "learning_rate": 1e-05, "loss": 0.471, "mean_token_accuracy": 0.8753843307495117, "num_tokens": 642106831.0, "step": 16912 }, { "epoch": 2.1515074418012974, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.598611831665039, "learning_rate": 1e-05, "loss": 0.449, "mean_token_accuracy": 0.8859814405441284, "num_tokens": 642142408.0, "step": 16913 }, { "epoch": 2.151634652079888, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.2592315673828125, "learning_rate": 1e-05, "loss": 0.4898, "mean_token_accuracy": 0.8732120394706726, "num_tokens": 642181860.0, "step": 16914 }, { "epoch": 2.1517618623584784, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.5481438636779785, "learning_rate": 1e-05, "loss": 0.436, "mean_token_accuracy": 0.8928943872451782, "num_tokens": 642219073.0, "step": 16915 }, { "epoch": 2.151889072637069, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.107966899871826, "learning_rate": 1e-05, "loss": 0.4704, "mean_token_accuracy": 0.8755403161048889, "num_tokens": 642254872.0, "step": 16916 }, { "epoch": 2.1520162829156595, "ewc_loss": 0.125, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.602330684661865, "learning_rate": 1e-05, "loss": 0.4414, "mean_token_accuracy": 0.8893406391143799, "num_tokens": 642295379.0, "step": 16917 }, { "epoch": 2.15214349319425, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.960690021514893, "learning_rate": 1e-05, "loss": 0.5166, "mean_token_accuracy": 0.8595597743988037, "num_tokens": 642333316.0, "step": 16918 }, { "epoch": 2.1522707034728406, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.758405685424805, "learning_rate": 1e-05, "loss": 0.505, "mean_token_accuracy": 0.8694583773612976, "num_tokens": 642373677.0, "step": 16919 }, { "epoch": 2.152397913751431, "ewc_loss": 0.11083984375, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.8308820724487305, "learning_rate": 1e-05, "loss": 0.5007, "mean_token_accuracy": 0.8659974336624146, "num_tokens": 642416066.0, "step": 16920 }, { "epoch": 2.1525251240300216, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 7.022136211395264, "learning_rate": 1e-05, "loss": 0.528, "mean_token_accuracy": 0.8642628788948059, "num_tokens": 642456818.0, "step": 16921 }, { "epoch": 2.152652334308612, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.78614616394043, "learning_rate": 1e-05, "loss": 0.4679, "mean_token_accuracy": 0.8747395277023315, "num_tokens": 642498021.0, "step": 16922 }, { "epoch": 2.1527795445872027, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 0.0001163482666015625, "grad_norm": 7.341027736663818, "learning_rate": 1e-05, "loss": 0.5185, "mean_token_accuracy": 0.8677936792373657, "num_tokens": 642533999.0, "step": 16923 }, { "epoch": 2.152906754865793, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 5.97105598449707, "learning_rate": 1e-05, "loss": 0.4715, "mean_token_accuracy": 0.8780407905578613, "num_tokens": 642571943.0, "step": 16924 }, { "epoch": 2.1530339651443837, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 0.00011968612670898438, "grad_norm": 7.44942569732666, "learning_rate": 1e-05, "loss": 0.4691, "mean_token_accuracy": 0.8837804794311523, "num_tokens": 642605145.0, "step": 16925 }, { "epoch": 2.1531611754229742, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.316281318664551, "learning_rate": 1e-05, "loss": 0.4866, "mean_token_accuracy": 0.8736054301261902, "num_tokens": 642644686.0, "step": 16926 }, { "epoch": 2.1532883857015648, "ewc_loss": 0.140625, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 0.00011396408081054688, "grad_norm": 7.189108371734619, "learning_rate": 1e-05, "loss": 0.4626, "mean_token_accuracy": 0.8851065039634705, "num_tokens": 642677815.0, "step": 16927 }, { "epoch": 2.1534155959801553, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.5139970779418945, "learning_rate": 1e-05, "loss": 0.4428, "mean_token_accuracy": 0.8926358222961426, "num_tokens": 642716245.0, "step": 16928 }, { "epoch": 2.153542806258746, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.722530364990234, "learning_rate": 1e-05, "loss": 0.4941, "mean_token_accuracy": 0.8751475811004639, "num_tokens": 642750428.0, "step": 16929 }, { "epoch": 2.1536700165373364, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.6069207191467285, "learning_rate": 1e-05, "loss": 0.5057, "mean_token_accuracy": 0.8702654838562012, "num_tokens": 642790349.0, "step": 16930 }, { "epoch": 2.153797226815927, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.201059818267822, "learning_rate": 1e-05, "loss": 0.4516, "mean_token_accuracy": 0.8825380206108093, "num_tokens": 642835063.0, "step": 16931 }, { "epoch": 2.1539244370945174, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.7901692390441895, "learning_rate": 1e-05, "loss": 0.4861, "mean_token_accuracy": 0.8750719428062439, "num_tokens": 642875013.0, "step": 16932 }, { "epoch": 2.154051647373108, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.818367004394531, "learning_rate": 1e-05, "loss": 0.4861, "mean_token_accuracy": 0.8688733577728271, "num_tokens": 642914934.0, "step": 16933 }, { "epoch": 2.154178857651698, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 0.00011157989501953125, "grad_norm": 7.133839130401611, "learning_rate": 1e-05, "loss": 0.4839, "mean_token_accuracy": 0.8776563405990601, "num_tokens": 642949368.0, "step": 16934 }, { "epoch": 2.1543060679302886, "ewc_loss": 0.109375, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.699719429016113, "learning_rate": 1e-05, "loss": 0.4828, "mean_token_accuracy": 0.8694334030151367, "num_tokens": 642988083.0, "step": 16935 }, { "epoch": 2.154433278208879, "ewc_loss": 0.1484375, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 0.0001220703125, "grad_norm": 7.568135738372803, "learning_rate": 1e-05, "loss": 0.5557, "mean_token_accuracy": 0.859879732131958, "num_tokens": 643028868.0, "step": 16936 }, { "epoch": 2.1545604884874696, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 5.973492622375488, "learning_rate": 1e-05, "loss": 0.4871, "mean_token_accuracy": 0.8730853796005249, "num_tokens": 643072251.0, "step": 16937 }, { "epoch": 2.15468769876606, "ewc_loss": 0.154296875, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 0.0001277923583984375, "grad_norm": 7.764749050140381, "learning_rate": 1e-05, "loss": 0.4953, "mean_token_accuracy": 0.8792734146118164, "num_tokens": 643109864.0, "step": 16938 }, { "epoch": 2.1548149090446507, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.471244812011719, "learning_rate": 1e-05, "loss": 0.4507, "mean_token_accuracy": 0.8902373909950256, "num_tokens": 643147868.0, "step": 16939 }, { "epoch": 2.154942119323241, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 0.00012063980102539062, "grad_norm": 7.476235389709473, "learning_rate": 1e-05, "loss": 0.4212, "mean_token_accuracy": 0.9004397988319397, "num_tokens": 643182173.0, "step": 16940 }, { "epoch": 2.1550693296018317, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 6.660247802734375, "learning_rate": 1e-05, "loss": 0.5083, "mean_token_accuracy": 0.8717300891876221, "num_tokens": 643217006.0, "step": 16941 }, { "epoch": 2.1551965398804223, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 0.00011110305786132812, "grad_norm": 7.0680365562438965, "learning_rate": 1e-05, "loss": 0.5416, "mean_token_accuracy": 0.8599333763122559, "num_tokens": 643253724.0, "step": 16942 }, { "epoch": 2.155323750159013, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.582602024078369, "learning_rate": 1e-05, "loss": 0.5306, "mean_token_accuracy": 0.8634248971939087, "num_tokens": 643294633.0, "step": 16943 }, { "epoch": 2.1554509604376033, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.746060371398926, "learning_rate": 1e-05, "loss": 0.4983, "mean_token_accuracy": 0.8696428537368774, "num_tokens": 643333269.0, "step": 16944 }, { "epoch": 2.155578170716194, "ewc_loss": 0.125, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.425621032714844, "learning_rate": 1e-05, "loss": 0.4564, "mean_token_accuracy": 0.8840451240539551, "num_tokens": 643374892.0, "step": 16945 }, { "epoch": 2.1557053809947844, "ewc_loss": 0.125, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.576221942901611, "learning_rate": 1e-05, "loss": 0.4429, "mean_token_accuracy": 0.8890630006790161, "num_tokens": 643411774.0, "step": 16946 }, { "epoch": 2.155832591273375, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.251748085021973, "learning_rate": 1e-05, "loss": 0.4444, "mean_token_accuracy": 0.8861886262893677, "num_tokens": 643450034.0, "step": 16947 }, { "epoch": 2.1559598015519654, "ewc_loss": 0.125, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.568068027496338, "learning_rate": 1e-05, "loss": 0.476, "mean_token_accuracy": 0.8789548873901367, "num_tokens": 643482901.0, "step": 16948 }, { "epoch": 2.156087011830556, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 6.072109222412109, "learning_rate": 1e-05, "loss": 0.4837, "mean_token_accuracy": 0.8723095655441284, "num_tokens": 643524918.0, "step": 16949 }, { "epoch": 2.1562142221091465, "ewc_loss": 0.125, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.666714191436768, "learning_rate": 1e-05, "loss": 0.502, "mean_token_accuracy": 0.8677386045455933, "num_tokens": 643563931.0, "step": 16950 }, { "epoch": 2.156341432387737, "ewc_loss": 0.11083984375, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.904830455780029, "learning_rate": 1e-05, "loss": 0.462, "mean_token_accuracy": 0.8768448829650879, "num_tokens": 643602838.0, "step": 16951 }, { "epoch": 2.1564686426663275, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.900792598724365, "learning_rate": 1e-05, "loss": 0.4841, "mean_token_accuracy": 0.8770644068717957, "num_tokens": 643641128.0, "step": 16952 }, { "epoch": 2.156595852944918, "ewc_loss": 0.11083984375, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.791051387786865, "learning_rate": 1e-05, "loss": 0.4486, "mean_token_accuracy": 0.8814230561256409, "num_tokens": 643674853.0, "step": 16953 }, { "epoch": 2.1567230632235086, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 0.00011348724365234375, "grad_norm": 7.241044998168945, "learning_rate": 1e-05, "loss": 0.5065, "mean_token_accuracy": 0.8723517656326294, "num_tokens": 643709924.0, "step": 16954 }, { "epoch": 2.156850273502099, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 5.853096961975098, "learning_rate": 1e-05, "loss": 0.4407, "mean_token_accuracy": 0.8821758031845093, "num_tokens": 643744899.0, "step": 16955 }, { "epoch": 2.1569774837806897, "ewc_loss": 0.1484375, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 0.00012159347534179688, "grad_norm": 7.520749568939209, "learning_rate": 1e-05, "loss": 0.467, "mean_token_accuracy": 0.8889101147651672, "num_tokens": 643781316.0, "step": 16956 }, { "epoch": 2.15710469405928, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.167726516723633, "learning_rate": 1e-05, "loss": 0.4614, "mean_token_accuracy": 0.8801931142807007, "num_tokens": 643824267.0, "step": 16957 }, { "epoch": 2.1572319043378707, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 0.00011968612670898438, "grad_norm": 7.481652736663818, "learning_rate": 1e-05, "loss": 0.5064, "mean_token_accuracy": 0.8756685853004456, "num_tokens": 643867155.0, "step": 16958 }, { "epoch": 2.157359114616461, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.517168045043945, "learning_rate": 1e-05, "loss": 0.463, "mean_token_accuracy": 0.8853293061256409, "num_tokens": 643907813.0, "step": 16959 }, { "epoch": 2.1574863248950513, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 0.00011110305786132812, "grad_norm": 7.082692623138428, "learning_rate": 1e-05, "loss": 0.4668, "mean_token_accuracy": 0.8836514949798584, "num_tokens": 643940387.0, "step": 16960 }, { "epoch": 2.157613535173642, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.6389384269714355, "learning_rate": 1e-05, "loss": 0.5037, "mean_token_accuracy": 0.8694567680358887, "num_tokens": 643980387.0, "step": 16961 }, { "epoch": 2.1577407454522324, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.550680637359619, "learning_rate": 1e-05, "loss": 0.4773, "mean_token_accuracy": 0.8788512945175171, "num_tokens": 644020883.0, "step": 16962 }, { "epoch": 2.157867955730823, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.71984338760376, "learning_rate": 1e-05, "loss": 0.5354, "mean_token_accuracy": 0.860819935798645, "num_tokens": 644054492.0, "step": 16963 }, { "epoch": 2.1579951660094134, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.089691162109375, "learning_rate": 1e-05, "loss": 0.4977, "mean_token_accuracy": 0.8656871318817139, "num_tokens": 644095009.0, "step": 16964 }, { "epoch": 2.158122376288004, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.893502235412598, "learning_rate": 1e-05, "loss": 0.4464, "mean_token_accuracy": 0.8882409334182739, "num_tokens": 644128469.0, "step": 16965 }, { "epoch": 2.1582495865665945, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.800797462463379, "learning_rate": 1e-05, "loss": 0.5081, "mean_token_accuracy": 0.8594131469726562, "num_tokens": 644167102.0, "step": 16966 }, { "epoch": 2.158376796845185, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 0.0001125335693359375, "grad_norm": 7.230462074279785, "learning_rate": 1e-05, "loss": 0.5118, "mean_token_accuracy": 0.8739099502563477, "num_tokens": 644204319.0, "step": 16967 }, { "epoch": 2.1585040071237755, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.765739917755127, "learning_rate": 1e-05, "loss": 0.5102, "mean_token_accuracy": 0.8668358325958252, "num_tokens": 644246950.0, "step": 16968 }, { "epoch": 2.158631217402366, "ewc_loss": 0.1494140625, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.6003193855285645, "learning_rate": 1e-05, "loss": 0.4947, "mean_token_accuracy": 0.8792496919631958, "num_tokens": 644288744.0, "step": 16969 }, { "epoch": 2.1587584276809566, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.092385292053223, "learning_rate": 1e-05, "loss": 0.4623, "mean_token_accuracy": 0.8784040212631226, "num_tokens": 644326719.0, "step": 16970 }, { "epoch": 2.158885637959547, "ewc_loss": 0.1494140625, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.665793418884277, "learning_rate": 1e-05, "loss": 0.5151, "mean_token_accuracy": 0.8726949691772461, "num_tokens": 644362027.0, "step": 16971 }, { "epoch": 2.1590128482381377, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.505283355712891, "learning_rate": 1e-05, "loss": 0.491, "mean_token_accuracy": 0.8750789165496826, "num_tokens": 644401917.0, "step": 16972 }, { "epoch": 2.159140058516728, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 0.00011873245239257812, "grad_norm": 7.348504066467285, "learning_rate": 1e-05, "loss": 0.5109, "mean_token_accuracy": 0.8710777759552002, "num_tokens": 644441735.0, "step": 16973 }, { "epoch": 2.1592672687953187, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 6.613315105438232, "learning_rate": 1e-05, "loss": 0.4365, "mean_token_accuracy": 0.8908367156982422, "num_tokens": 644475833.0, "step": 16974 }, { "epoch": 2.1593944790739092, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 6.968235015869141, "learning_rate": 1e-05, "loss": 0.4798, "mean_token_accuracy": 0.8785222172737122, "num_tokens": 644516506.0, "step": 16975 }, { "epoch": 2.1595216893524998, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.53104305267334, "learning_rate": 1e-05, "loss": 0.4568, "mean_token_accuracy": 0.8832716345787048, "num_tokens": 644554081.0, "step": 16976 }, { "epoch": 2.1596488996310903, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.664370059967041, "learning_rate": 1e-05, "loss": 0.5195, "mean_token_accuracy": 0.8645422458648682, "num_tokens": 644599514.0, "step": 16977 }, { "epoch": 2.159776109909681, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.400462627410889, "learning_rate": 1e-05, "loss": 0.4954, "mean_token_accuracy": 0.8670089244842529, "num_tokens": 644639569.0, "step": 16978 }, { "epoch": 2.1599033201882714, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.529246807098389, "learning_rate": 1e-05, "loss": 0.4377, "mean_token_accuracy": 0.8916118741035461, "num_tokens": 644685721.0, "step": 16979 }, { "epoch": 2.160030530466862, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.231449604034424, "learning_rate": 1e-05, "loss": 0.5188, "mean_token_accuracy": 0.862844705581665, "num_tokens": 644729939.0, "step": 16980 }, { "epoch": 2.1601577407454524, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.540033340454102, "learning_rate": 1e-05, "loss": 0.5002, "mean_token_accuracy": 0.8696112632751465, "num_tokens": 644768181.0, "step": 16981 }, { "epoch": 2.160284951024043, "ewc_loss": 0.11376953125, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 6.057021141052246, "learning_rate": 1e-05, "loss": 0.4995, "mean_token_accuracy": 0.8687605857849121, "num_tokens": 644806288.0, "step": 16982 }, { "epoch": 2.1604121613026335, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.647396087646484, "learning_rate": 1e-05, "loss": 0.4537, "mean_token_accuracy": 0.8844478130340576, "num_tokens": 644846116.0, "step": 16983 }, { "epoch": 2.1605393715812236, "ewc_loss": 0.11083984375, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.883114814758301, "learning_rate": 1e-05, "loss": 0.4803, "mean_token_accuracy": 0.8708330392837524, "num_tokens": 644885330.0, "step": 16984 }, { "epoch": 2.160666581859814, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.889464378356934, "learning_rate": 1e-05, "loss": 0.5189, "mean_token_accuracy": 0.8671592473983765, "num_tokens": 644923341.0, "step": 16985 }, { "epoch": 2.1607937921384046, "ewc_loss": 0.10986328125, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.768036365509033, "learning_rate": 1e-05, "loss": 0.4913, "mean_token_accuracy": 0.8697660565376282, "num_tokens": 644964165.0, "step": 16986 }, { "epoch": 2.160921002416995, "ewc_loss": 0.140625, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 0.00011444091796875, "grad_norm": 7.235698699951172, "learning_rate": 1e-05, "loss": 0.4411, "mean_token_accuracy": 0.8913555145263672, "num_tokens": 644999319.0, "step": 16987 }, { "epoch": 2.1610482126955857, "ewc_loss": 0.11376953125, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 5.818449020385742, "learning_rate": 1e-05, "loss": 0.4369, "mean_token_accuracy": 0.8846088647842407, "num_tokens": 645040687.0, "step": 16988 }, { "epoch": 2.161175422974176, "ewc_loss": 0.1484375, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 0.00012159347534179688, "grad_norm": 7.568133354187012, "learning_rate": 1e-05, "loss": 0.5106, "mean_token_accuracy": 0.8733885288238525, "num_tokens": 645075772.0, "step": 16989 }, { "epoch": 2.1613026332527667, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.180052280426025, "learning_rate": 1e-05, "loss": 0.4825, "mean_token_accuracy": 0.8748220801353455, "num_tokens": 645111856.0, "step": 16990 }, { "epoch": 2.1614298435313573, "ewc_loss": 0.1484375, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 0.00012159347534179688, "grad_norm": 7.498411178588867, "learning_rate": 1e-05, "loss": 0.5106, "mean_token_accuracy": 0.8700278401374817, "num_tokens": 645155995.0, "step": 16991 }, { "epoch": 2.161557053809948, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.54044771194458, "learning_rate": 1e-05, "loss": 0.5303, "mean_token_accuracy": 0.8612505197525024, "num_tokens": 645200101.0, "step": 16992 }, { "epoch": 2.1616842640885383, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 0.00011014938354492188, "grad_norm": 7.0781378746032715, "learning_rate": 1e-05, "loss": 0.4591, "mean_token_accuracy": 0.8891023397445679, "num_tokens": 645235243.0, "step": 16993 }, { "epoch": 2.161811474367129, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 6.644394397735596, "learning_rate": 1e-05, "loss": 0.4577, "mean_token_accuracy": 0.8889424800872803, "num_tokens": 645270369.0, "step": 16994 }, { "epoch": 2.1619386846457194, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.548813343048096, "learning_rate": 1e-05, "loss": 0.4316, "mean_token_accuracy": 0.8897168636322021, "num_tokens": 645304111.0, "step": 16995 }, { "epoch": 2.16206589492431, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.7016448974609375, "learning_rate": 1e-05, "loss": 0.508, "mean_token_accuracy": 0.8637967109680176, "num_tokens": 645339884.0, "step": 16996 }, { "epoch": 2.1621931052029004, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.098816394805908, "learning_rate": 1e-05, "loss": 0.4062, "mean_token_accuracy": 0.8933998942375183, "num_tokens": 645372465.0, "step": 16997 }, { "epoch": 2.162320315481491, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.875125885009766, "learning_rate": 1e-05, "loss": 0.4739, "mean_token_accuracy": 0.8808383345603943, "num_tokens": 645413022.0, "step": 16998 }, { "epoch": 2.1624475257600815, "ewc_loss": 0.10986328125, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.823755264282227, "learning_rate": 1e-05, "loss": 0.424, "mean_token_accuracy": 0.8899399042129517, "num_tokens": 645451344.0, "step": 16999 }, { "epoch": 2.162574736038672, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 0.0001125335693359375, "grad_norm": 7.177645683288574, "learning_rate": 1e-05, "loss": 0.4313, "mean_token_accuracy": 0.8949490785598755, "num_tokens": 645493463.0, "step": 17000 }, { "epoch": 2.1627019463172625, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.78593111038208, "learning_rate": 1e-05, "loss": 0.4434, "mean_token_accuracy": 0.8849378228187561, "num_tokens": 645533508.0, "step": 17001 }, { "epoch": 2.162829156595853, "ewc_loss": 0.1484375, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 0.00012159347534179688, "grad_norm": 7.5337347984313965, "learning_rate": 1e-05, "loss": 0.502, "mean_token_accuracy": 0.8797370195388794, "num_tokens": 645573866.0, "step": 17002 }, { "epoch": 2.1629563668744436, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.067261695861816, "learning_rate": 1e-05, "loss": 0.4456, "mean_token_accuracy": 0.8840365409851074, "num_tokens": 645609835.0, "step": 17003 }, { "epoch": 2.163083577153034, "ewc_loss": 0.1494140625, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.617644786834717, "learning_rate": 1e-05, "loss": 0.4849, "mean_token_accuracy": 0.8832995891571045, "num_tokens": 645644228.0, "step": 17004 }, { "epoch": 2.1632107874316246, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.42352819442749, "learning_rate": 1e-05, "loss": 0.4777, "mean_token_accuracy": 0.8790180683135986, "num_tokens": 645682619.0, "step": 17005 }, { "epoch": 2.163337997710215, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 0.000118255615234375, "grad_norm": 7.361408233642578, "learning_rate": 1e-05, "loss": 0.4773, "mean_token_accuracy": 0.881282389163971, "num_tokens": 645719506.0, "step": 17006 }, { "epoch": 2.1634652079888053, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.550151348114014, "learning_rate": 1e-05, "loss": 0.4919, "mean_token_accuracy": 0.872817873954773, "num_tokens": 645761930.0, "step": 17007 }, { "epoch": 2.1635924182673962, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 7.0160651206970215, "learning_rate": 1e-05, "loss": 0.5252, "mean_token_accuracy": 0.8659005165100098, "num_tokens": 645800322.0, "step": 17008 }, { "epoch": 2.1637196285459863, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.496480464935303, "learning_rate": 1e-05, "loss": 0.487, "mean_token_accuracy": 0.8763955235481262, "num_tokens": 645842914.0, "step": 17009 }, { "epoch": 2.163846838824577, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.718392372131348, "learning_rate": 1e-05, "loss": 0.5082, "mean_token_accuracy": 0.8695691823959351, "num_tokens": 645885798.0, "step": 17010 }, { "epoch": 2.1639740491031674, "ewc_loss": 0.125, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.387562274932861, "learning_rate": 1e-05, "loss": 0.4821, "mean_token_accuracy": 0.8749188184738159, "num_tokens": 645922187.0, "step": 17011 }, { "epoch": 2.164101259381758, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.563843727111816, "learning_rate": 1e-05, "loss": 0.4303, "mean_token_accuracy": 0.8906970620155334, "num_tokens": 645955594.0, "step": 17012 }, { "epoch": 2.1642284696603484, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.254751682281494, "learning_rate": 1e-05, "loss": 0.4509, "mean_token_accuracy": 0.8865447640419006, "num_tokens": 645992579.0, "step": 17013 }, { "epoch": 2.164355679938939, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.528228282928467, "learning_rate": 1e-05, "loss": 0.4809, "mean_token_accuracy": 0.8740759491920471, "num_tokens": 646033243.0, "step": 17014 }, { "epoch": 2.1644828902175295, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 6.206816673278809, "learning_rate": 1e-05, "loss": 0.439, "mean_token_accuracy": 0.8873717784881592, "num_tokens": 646069309.0, "step": 17015 }, { "epoch": 2.16461010049612, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.507525444030762, "learning_rate": 1e-05, "loss": 0.474, "mean_token_accuracy": 0.8772647380828857, "num_tokens": 646100720.0, "step": 17016 }, { "epoch": 2.1647373107747105, "ewc_loss": 0.11376953125, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 5.997843265533447, "learning_rate": 1e-05, "loss": 0.401, "mean_token_accuracy": 0.8973860144615173, "num_tokens": 646135917.0, "step": 17017 }, { "epoch": 2.164864521053301, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.610477924346924, "learning_rate": 1e-05, "loss": 0.4275, "mean_token_accuracy": 0.8918010592460632, "num_tokens": 646173227.0, "step": 17018 }, { "epoch": 2.1649917313318916, "ewc_loss": 0.11083984375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.862878799438477, "learning_rate": 1e-05, "loss": 0.4976, "mean_token_accuracy": 0.8693807721138, "num_tokens": 646210023.0, "step": 17019 }, { "epoch": 2.165118941610482, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.857268810272217, "learning_rate": 1e-05, "loss": 0.4706, "mean_token_accuracy": 0.8804378509521484, "num_tokens": 646241124.0, "step": 17020 }, { "epoch": 2.1652461518890727, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.747299671173096, "learning_rate": 1e-05, "loss": 0.5214, "mean_token_accuracy": 0.8592973947525024, "num_tokens": 646276892.0, "step": 17021 }, { "epoch": 2.165373362167663, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 0.00011539459228515625, "grad_norm": 7.221304893493652, "learning_rate": 1e-05, "loss": 0.4496, "mean_token_accuracy": 0.8894796371459961, "num_tokens": 646308959.0, "step": 17022 }, { "epoch": 2.1655005724462537, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.781719207763672, "learning_rate": 1e-05, "loss": 0.4414, "mean_token_accuracy": 0.883162796497345, "num_tokens": 646342614.0, "step": 17023 }, { "epoch": 2.1656277827248442, "ewc_loss": 0.1484375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 0.0001220703125, "grad_norm": 7.551498889923096, "learning_rate": 1e-05, "loss": 0.4943, "mean_token_accuracy": 0.876479983329773, "num_tokens": 646380395.0, "step": 17024 }, { "epoch": 2.1657549930034348, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.119316101074219, "learning_rate": 1e-05, "loss": 0.5176, "mean_token_accuracy": 0.8674712181091309, "num_tokens": 646417068.0, "step": 17025 }, { "epoch": 2.1658822032820253, "ewc_loss": 0.1494140625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.55733060836792, "learning_rate": 1e-05, "loss": 0.4998, "mean_token_accuracy": 0.8762649893760681, "num_tokens": 646456655.0, "step": 17026 }, { "epoch": 2.166009413560616, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 6.494076728820801, "learning_rate": 1e-05, "loss": 0.4281, "mean_token_accuracy": 0.8989318609237671, "num_tokens": 646495035.0, "step": 17027 }, { "epoch": 2.1661366238392064, "ewc_loss": 0.140625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 0.00011396408081054688, "grad_norm": 7.169840335845947, "learning_rate": 1e-05, "loss": 0.4417, "mean_token_accuracy": 0.8904259204864502, "num_tokens": 646529695.0, "step": 17028 }, { "epoch": 2.166263834117797, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 6.634529113769531, "learning_rate": 1e-05, "loss": 0.5392, "mean_token_accuracy": 0.8593469858169556, "num_tokens": 646566563.0, "step": 17029 }, { "epoch": 2.1663910443963874, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.652956962585449, "learning_rate": 1e-05, "loss": 0.5008, "mean_token_accuracy": 0.8750982284545898, "num_tokens": 646609515.0, "step": 17030 }, { "epoch": 2.166518254674978, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.6748175621032715, "learning_rate": 1e-05, "loss": 0.4661, "mean_token_accuracy": 0.8841941356658936, "num_tokens": 646650024.0, "step": 17031 }, { "epoch": 2.166645464953568, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.193465232849121, "learning_rate": 1e-05, "loss": 0.4416, "mean_token_accuracy": 0.8877987265586853, "num_tokens": 646690547.0, "step": 17032 }, { "epoch": 2.1667726752321586, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.819986820220947, "learning_rate": 1e-05, "loss": 0.4534, "mean_token_accuracy": 0.8848373293876648, "num_tokens": 646722060.0, "step": 17033 }, { "epoch": 2.166899885510749, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.893176078796387, "learning_rate": 1e-05, "loss": 0.4706, "mean_token_accuracy": 0.8750476241111755, "num_tokens": 646761972.0, "step": 17034 }, { "epoch": 2.1670270957893396, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 7.082431316375732, "learning_rate": 1e-05, "loss": 0.5297, "mean_token_accuracy": 0.8655091524124146, "num_tokens": 646794847.0, "step": 17035 }, { "epoch": 2.16715430606793, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.790444374084473, "learning_rate": 1e-05, "loss": 0.4352, "mean_token_accuracy": 0.8870621919631958, "num_tokens": 646833744.0, "step": 17036 }, { "epoch": 2.1672815163465207, "ewc_loss": 0.1455078125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 0.00011873245239257812, "grad_norm": 7.394134998321533, "learning_rate": 1e-05, "loss": 0.4729, "mean_token_accuracy": 0.88297438621521, "num_tokens": 646874257.0, "step": 17037 }, { "epoch": 2.167408726625111, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 5.943411350250244, "learning_rate": 1e-05, "loss": 0.4343, "mean_token_accuracy": 0.889897882938385, "num_tokens": 646912439.0, "step": 17038 }, { "epoch": 2.1675359369037017, "ewc_loss": 0.1484375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 0.0001220703125, "grad_norm": 7.580383777618408, "learning_rate": 1e-05, "loss": 0.4621, "mean_token_accuracy": 0.888164758682251, "num_tokens": 646946311.0, "step": 17039 }, { "epoch": 2.1676631471822922, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.280036449432373, "learning_rate": 1e-05, "loss": 0.5111, "mean_token_accuracy": 0.8659881949424744, "num_tokens": 646982426.0, "step": 17040 }, { "epoch": 2.1677903574608828, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 0.00011968612670898438, "grad_norm": 7.4325480461120605, "learning_rate": 1e-05, "loss": 0.4586, "mean_token_accuracy": 0.8912568092346191, "num_tokens": 647023244.0, "step": 17041 }, { "epoch": 2.1679175677394733, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.489355087280273, "learning_rate": 1e-05, "loss": 0.5345, "mean_token_accuracy": 0.859775185585022, "num_tokens": 647062434.0, "step": 17042 }, { "epoch": 2.168044778018064, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 0.0001125335693359375, "grad_norm": 7.1278462409973145, "learning_rate": 1e-05, "loss": 0.494, "mean_token_accuracy": 0.8737403154373169, "num_tokens": 647098021.0, "step": 17043 }, { "epoch": 2.1681719882966544, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.500609874725342, "learning_rate": 1e-05, "loss": 0.4974, "mean_token_accuracy": 0.8743549585342407, "num_tokens": 647134709.0, "step": 17044 }, { "epoch": 2.168299198575245, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.812417030334473, "learning_rate": 1e-05, "loss": 0.504, "mean_token_accuracy": 0.8629018664360046, "num_tokens": 647169267.0, "step": 17045 }, { "epoch": 2.1684264088538354, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.4100213050842285, "learning_rate": 1e-05, "loss": 0.4833, "mean_token_accuracy": 0.8739900588989258, "num_tokens": 647207399.0, "step": 17046 }, { "epoch": 2.168553619132426, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.605438232421875, "learning_rate": 1e-05, "loss": 0.4865, "mean_token_accuracy": 0.8736732006072998, "num_tokens": 647244722.0, "step": 17047 }, { "epoch": 2.1686808294110165, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.298058986663818, "learning_rate": 1e-05, "loss": 0.4538, "mean_token_accuracy": 0.8841570615768433, "num_tokens": 647279051.0, "step": 17048 }, { "epoch": 2.168808039689607, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.523491382598877, "learning_rate": 1e-05, "loss": 0.4792, "mean_token_accuracy": 0.8751195669174194, "num_tokens": 647319201.0, "step": 17049 }, { "epoch": 2.1689352499681975, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.1717915534973145, "learning_rate": 1e-05, "loss": 0.488, "mean_token_accuracy": 0.873485267162323, "num_tokens": 647358306.0, "step": 17050 }, { "epoch": 2.169062460246788, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.5282979011535645, "learning_rate": 1e-05, "loss": 0.5006, "mean_token_accuracy": 0.8665146827697754, "num_tokens": 647393879.0, "step": 17051 }, { "epoch": 2.1691896705253786, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 6.030201435089111, "learning_rate": 1e-05, "loss": 0.4811, "mean_token_accuracy": 0.8699156641960144, "num_tokens": 647434848.0, "step": 17052 }, { "epoch": 2.169316880803969, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.642786979675293, "learning_rate": 1e-05, "loss": 0.4512, "mean_token_accuracy": 0.8845027089118958, "num_tokens": 647469711.0, "step": 17053 }, { "epoch": 2.1694440910825596, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.8873491287231445, "learning_rate": 1e-05, "loss": 0.5148, "mean_token_accuracy": 0.861443281173706, "num_tokens": 647507788.0, "step": 17054 }, { "epoch": 2.16957130136115, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.858700752258301, "learning_rate": 1e-05, "loss": 0.488, "mean_token_accuracy": 0.8797520399093628, "num_tokens": 647542193.0, "step": 17055 }, { "epoch": 2.1696985116397407, "ewc_loss": 0.11083984375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.762826442718506, "learning_rate": 1e-05, "loss": 0.405, "mean_token_accuracy": 0.8960249423980713, "num_tokens": 647578692.0, "step": 17056 }, { "epoch": 2.169825721918331, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 0.00011348724365234375, "grad_norm": 7.1854987144470215, "learning_rate": 1e-05, "loss": 0.4767, "mean_token_accuracy": 0.8821924924850464, "num_tokens": 647618158.0, "step": 17057 }, { "epoch": 2.1699529321969213, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 5.801630973815918, "learning_rate": 1e-05, "loss": 0.4585, "mean_token_accuracy": 0.8806931972503662, "num_tokens": 647660788.0, "step": 17058 }, { "epoch": 2.170080142475512, "ewc_loss": 0.1484375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 0.00012159347534179688, "grad_norm": 7.497666358947754, "learning_rate": 1e-05, "loss": 0.5213, "mean_token_accuracy": 0.8671165704727173, "num_tokens": 647700418.0, "step": 17059 }, { "epoch": 2.1702073527541024, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.101285934448242, "learning_rate": 1e-05, "loss": 0.4482, "mean_token_accuracy": 0.886299729347229, "num_tokens": 647738248.0, "step": 17060 }, { "epoch": 2.170334563032693, "ewc_loss": 0.1474609375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 0.00012111663818359375, "grad_norm": 7.5090227127075195, "learning_rate": 1e-05, "loss": 0.476, "mean_token_accuracy": 0.8855577707290649, "num_tokens": 647777287.0, "step": 17061 }, { "epoch": 2.1704617733112834, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.457736492156982, "learning_rate": 1e-05, "loss": 0.4906, "mean_token_accuracy": 0.8741545677185059, "num_tokens": 647823213.0, "step": 17062 }, { "epoch": 2.170588983589874, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 0.00011301040649414062, "grad_norm": 7.13059663772583, "learning_rate": 1e-05, "loss": 0.4561, "mean_token_accuracy": 0.8913092613220215, "num_tokens": 647861506.0, "step": 17063 }, { "epoch": 2.1707161938684645, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.592321872711182, "learning_rate": 1e-05, "loss": 0.4428, "mean_token_accuracy": 0.8898813724517822, "num_tokens": 647900039.0, "step": 17064 }, { "epoch": 2.170843404147055, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.630392074584961, "learning_rate": 1e-05, "loss": 0.4872, "mean_token_accuracy": 0.8748384714126587, "num_tokens": 647938368.0, "step": 17065 }, { "epoch": 2.1709706144256455, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.66702127456665, "learning_rate": 1e-05, "loss": 0.517, "mean_token_accuracy": 0.8659798502922058, "num_tokens": 647972549.0, "step": 17066 }, { "epoch": 2.171097824704236, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.1537299156188965, "learning_rate": 1e-05, "loss": 0.4749, "mean_token_accuracy": 0.8715460300445557, "num_tokens": 648016923.0, "step": 17067 }, { "epoch": 2.1712250349828266, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.838132858276367, "learning_rate": 1e-05, "loss": 0.468, "mean_token_accuracy": 0.8833572864532471, "num_tokens": 648059880.0, "step": 17068 }, { "epoch": 2.171352245261417, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.8516435623168945, "learning_rate": 1e-05, "loss": 0.4838, "mean_token_accuracy": 0.8716113567352295, "num_tokens": 648093657.0, "step": 17069 }, { "epoch": 2.1714794555400077, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 0.00011110305786132812, "grad_norm": 7.140868663787842, "learning_rate": 1e-05, "loss": 0.4731, "mean_token_accuracy": 0.8791505694389343, "num_tokens": 648130723.0, "step": 17070 }, { "epoch": 2.171606665818598, "ewc_loss": 0.11083984375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.775114059448242, "learning_rate": 1e-05, "loss": 0.4882, "mean_token_accuracy": 0.8697564601898193, "num_tokens": 648176657.0, "step": 17071 }, { "epoch": 2.1717338760971887, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 0.00011968612670898438, "grad_norm": 7.475667476654053, "learning_rate": 1e-05, "loss": 0.4974, "mean_token_accuracy": 0.8817928433418274, "num_tokens": 648214505.0, "step": 17072 }, { "epoch": 2.1718610863757792, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.007890701293945, "learning_rate": 1e-05, "loss": 0.4522, "mean_token_accuracy": 0.885746419429779, "num_tokens": 648247816.0, "step": 17073 }, { "epoch": 2.1719882966543698, "ewc_loss": 0.1494140625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.6224751472473145, "learning_rate": 1e-05, "loss": 0.4894, "mean_token_accuracy": 0.8794816732406616, "num_tokens": 648276418.0, "step": 17074 }, { "epoch": 2.1721155069329603, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.3746018409729, "learning_rate": 1e-05, "loss": 0.5155, "mean_token_accuracy": 0.8655779361724854, "num_tokens": 648314722.0, "step": 17075 }, { "epoch": 2.172242717211551, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 0.000118255615234375, "grad_norm": 7.384188175201416, "learning_rate": 1e-05, "loss": 0.5079, "mean_token_accuracy": 0.8706701993942261, "num_tokens": 648354733.0, "step": 17076 }, { "epoch": 2.1723699274901414, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.5425567626953125, "learning_rate": 1e-05, "loss": 0.4904, "mean_token_accuracy": 0.8753630518913269, "num_tokens": 648392164.0, "step": 17077 }, { "epoch": 2.172497137768732, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 7.043066501617432, "learning_rate": 1e-05, "loss": 0.4787, "mean_token_accuracy": 0.8820987939834595, "num_tokens": 648431728.0, "step": 17078 }, { "epoch": 2.1726243480473224, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.493403434753418, "learning_rate": 1e-05, "loss": 0.4956, "mean_token_accuracy": 0.8716090321540833, "num_tokens": 648471225.0, "step": 17079 }, { "epoch": 2.172751558325913, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.757235527038574, "learning_rate": 1e-05, "loss": 0.4766, "mean_token_accuracy": 0.8773760795593262, "num_tokens": 648507742.0, "step": 17080 }, { "epoch": 2.1728787686045035, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.382668972015381, "learning_rate": 1e-05, "loss": 0.4951, "mean_token_accuracy": 0.8741520643234253, "num_tokens": 648548641.0, "step": 17081 }, { "epoch": 2.1730059788830935, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.5818190574646, "learning_rate": 1e-05, "loss": 0.4462, "mean_token_accuracy": 0.8865469694137573, "num_tokens": 648583540.0, "step": 17082 }, { "epoch": 2.173133189161684, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.243634223937988, "learning_rate": 1e-05, "loss": 0.4309, "mean_token_accuracy": 0.8911526203155518, "num_tokens": 648627225.0, "step": 17083 }, { "epoch": 2.1732603994402746, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.542278289794922, "learning_rate": 1e-05, "loss": 0.4338, "mean_token_accuracy": 0.8916600942611694, "num_tokens": 648662760.0, "step": 17084 }, { "epoch": 2.173387609718865, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.092794895172119, "learning_rate": 1e-05, "loss": 0.4714, "mean_token_accuracy": 0.8715893030166626, "num_tokens": 648697748.0, "step": 17085 }, { "epoch": 2.1735148199974557, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.6209330558776855, "learning_rate": 1e-05, "loss": 0.4925, "mean_token_accuracy": 0.8718301653862, "num_tokens": 648732981.0, "step": 17086 }, { "epoch": 2.173642030276046, "ewc_loss": 0.11181640625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.941123962402344, "learning_rate": 1e-05, "loss": 0.4581, "mean_token_accuracy": 0.8781896233558655, "num_tokens": 648764356.0, "step": 17087 }, { "epoch": 2.1737692405546367, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.805606842041016, "learning_rate": 1e-05, "loss": 0.4888, "mean_token_accuracy": 0.8748258352279663, "num_tokens": 648809064.0, "step": 17088 }, { "epoch": 2.1738964508332272, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.811051845550537, "learning_rate": 1e-05, "loss": 0.4306, "mean_token_accuracy": 0.886283814907074, "num_tokens": 648843299.0, "step": 17089 }, { "epoch": 2.1740236611118178, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 0.00011110305786132812, "grad_norm": 7.101650238037109, "learning_rate": 1e-05, "loss": 0.5053, "mean_token_accuracy": 0.8697028160095215, "num_tokens": 648881251.0, "step": 17090 }, { "epoch": 2.1741508713904083, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.78910493850708, "learning_rate": 1e-05, "loss": 0.4746, "mean_token_accuracy": 0.8742021322250366, "num_tokens": 648915837.0, "step": 17091 }, { "epoch": 2.174278081668999, "ewc_loss": 0.1455078125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 0.00011920928955078125, "grad_norm": 7.427077293395996, "learning_rate": 1e-05, "loss": 0.5304, "mean_token_accuracy": 0.8668490648269653, "num_tokens": 648953354.0, "step": 17092 }, { "epoch": 2.1744052919475894, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.019100666046143, "learning_rate": 1e-05, "loss": 0.4508, "mean_token_accuracy": 0.8813945651054382, "num_tokens": 648993224.0, "step": 17093 }, { "epoch": 2.17453250222618, "ewc_loss": 0.1474609375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 0.00012111663818359375, "grad_norm": 7.528092384338379, "learning_rate": 1e-05, "loss": 0.5023, "mean_token_accuracy": 0.8738154172897339, "num_tokens": 649026429.0, "step": 17094 }, { "epoch": 2.1746597125047704, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.396048545837402, "learning_rate": 1e-05, "loss": 0.4626, "mean_token_accuracy": 0.8819071650505066, "num_tokens": 649064672.0, "step": 17095 }, { "epoch": 2.174786922783361, "ewc_loss": 0.140625, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 0.00011444091796875, "grad_norm": 7.203361511230469, "learning_rate": 1e-05, "loss": 0.4886, "mean_token_accuracy": 0.8795367479324341, "num_tokens": 649097926.0, "step": 17096 }, { "epoch": 2.1749141330619515, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.589614391326904, "learning_rate": 1e-05, "loss": 0.4967, "mean_token_accuracy": 0.8753859400749207, "num_tokens": 649138502.0, "step": 17097 }, { "epoch": 2.175041343340542, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.699456691741943, "learning_rate": 1e-05, "loss": 0.5062, "mean_token_accuracy": 0.8690491914749146, "num_tokens": 649175867.0, "step": 17098 }, { "epoch": 2.1751685536191325, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.653618335723877, "learning_rate": 1e-05, "loss": 0.4435, "mean_token_accuracy": 0.8899289965629578, "num_tokens": 649214132.0, "step": 17099 }, { "epoch": 2.175295763897723, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.204284191131592, "learning_rate": 1e-05, "loss": 0.4849, "mean_token_accuracy": 0.8739504814147949, "num_tokens": 649262397.0, "step": 17100 }, { "epoch": 2.1754229741763136, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.838817596435547, "learning_rate": 1e-05, "loss": 0.5388, "mean_token_accuracy": 0.8620762228965759, "num_tokens": 649298165.0, "step": 17101 }, { "epoch": 2.175550184454904, "ewc_loss": 0.11181640625, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.850362300872803, "learning_rate": 1e-05, "loss": 0.4679, "mean_token_accuracy": 0.8738352060317993, "num_tokens": 649337802.0, "step": 17102 }, { "epoch": 2.1756773947334946, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 0.00011157989501953125, "grad_norm": 7.132753849029541, "learning_rate": 1e-05, "loss": 0.4581, "mean_token_accuracy": 0.8863771557807922, "num_tokens": 649378922.0, "step": 17103 }, { "epoch": 2.175804605012085, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.740360260009766, "learning_rate": 1e-05, "loss": 0.427, "mean_token_accuracy": 0.8876560926437378, "num_tokens": 649414332.0, "step": 17104 }, { "epoch": 2.1759318152906753, "ewc_loss": 0.1474609375, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 0.00012111663818359375, "grad_norm": 7.538715362548828, "learning_rate": 1e-05, "loss": 0.4675, "mean_token_accuracy": 0.8857723474502563, "num_tokens": 649448965.0, "step": 17105 }, { "epoch": 2.1760590255692662, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 5.999363422393799, "learning_rate": 1e-05, "loss": 0.4671, "mean_token_accuracy": 0.8755496740341187, "num_tokens": 649487322.0, "step": 17106 }, { "epoch": 2.1761862358478563, "ewc_loss": 0.1533203125, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 0.00012683868408203125, "grad_norm": 7.699112415313721, "learning_rate": 1e-05, "loss": 0.5935, "mean_token_accuracy": 0.8501962423324585, "num_tokens": 649526524.0, "step": 17107 }, { "epoch": 2.176313446126447, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.437588691711426, "learning_rate": 1e-05, "loss": 0.5114, "mean_token_accuracy": 0.8696564435958862, "num_tokens": 649566169.0, "step": 17108 }, { "epoch": 2.1764406564050374, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 0.0001201629638671875, "grad_norm": 7.442203044891357, "learning_rate": 1e-05, "loss": 0.4614, "mean_token_accuracy": 0.8878207802772522, "num_tokens": 649603725.0, "step": 17109 }, { "epoch": 2.176567866683628, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 6.611259460449219, "learning_rate": 1e-05, "loss": 0.5015, "mean_token_accuracy": 0.8718392848968506, "num_tokens": 649644789.0, "step": 17110 }, { "epoch": 2.1766950769622184, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 0.00011110305786132812, "grad_norm": 7.062897682189941, "learning_rate": 1e-05, "loss": 0.5082, "mean_token_accuracy": 0.870527446269989, "num_tokens": 649682051.0, "step": 17111 }, { "epoch": 2.176822287240809, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.551201820373535, "learning_rate": 1e-05, "loss": 0.448, "mean_token_accuracy": 0.8894045948982239, "num_tokens": 649715730.0, "step": 17112 }, { "epoch": 2.1769494975193995, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.7523698806762695, "learning_rate": 1e-05, "loss": 0.4951, "mean_token_accuracy": 0.8741909265518188, "num_tokens": 649750237.0, "step": 17113 }, { "epoch": 2.17707670779799, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.420970439910889, "learning_rate": 1e-05, "loss": 0.4447, "mean_token_accuracy": 0.8870927691459656, "num_tokens": 649791250.0, "step": 17114 }, { "epoch": 2.1772039180765805, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.573312282562256, "learning_rate": 1e-05, "loss": 0.4801, "mean_token_accuracy": 0.8797833919525146, "num_tokens": 649831750.0, "step": 17115 }, { "epoch": 2.177331128355171, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.278820037841797, "learning_rate": 1e-05, "loss": 0.487, "mean_token_accuracy": 0.873320996761322, "num_tokens": 649869245.0, "step": 17116 }, { "epoch": 2.1774583386337616, "ewc_loss": 0.125, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.5410566329956055, "learning_rate": 1e-05, "loss": 0.4874, "mean_token_accuracy": 0.8737772703170776, "num_tokens": 649903763.0, "step": 17117 }, { "epoch": 2.177585548912352, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 6.1000657081604, "learning_rate": 1e-05, "loss": 0.4292, "mean_token_accuracy": 0.889297604560852, "num_tokens": 649946601.0, "step": 17118 }, { "epoch": 2.1777127591909426, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.622279167175293, "learning_rate": 1e-05, "loss": 0.4386, "mean_token_accuracy": 0.8887130618095398, "num_tokens": 649979308.0, "step": 17119 }, { "epoch": 2.177839969469533, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.9326300621032715, "learning_rate": 1e-05, "loss": 0.4361, "mean_token_accuracy": 0.8881072998046875, "num_tokens": 650018068.0, "step": 17120 }, { "epoch": 2.1779671797481237, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.817852020263672, "learning_rate": 1e-05, "loss": 0.5066, "mean_token_accuracy": 0.870159387588501, "num_tokens": 650059559.0, "step": 17121 }, { "epoch": 2.1780943900267142, "ewc_loss": 0.11181640625, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.809085845947266, "learning_rate": 1e-05, "loss": 0.4543, "mean_token_accuracy": 0.8791160583496094, "num_tokens": 650103266.0, "step": 17122 }, { "epoch": 2.1782216003053048, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 0.00011157989501953125, "grad_norm": 7.117343425750732, "learning_rate": 1e-05, "loss": 0.5134, "mean_token_accuracy": 0.8710824251174927, "num_tokens": 650139525.0, "step": 17123 }, { "epoch": 2.1783488105838953, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.798712253570557, "learning_rate": 1e-05, "loss": 0.4213, "mean_token_accuracy": 0.8915191292762756, "num_tokens": 650179249.0, "step": 17124 }, { "epoch": 2.178476020862486, "ewc_loss": 0.1455078125, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 0.00011873245239257812, "grad_norm": 7.458399772644043, "learning_rate": 1e-05, "loss": 0.5298, "mean_token_accuracy": 0.8650954961776733, "num_tokens": 650217258.0, "step": 17125 }, { "epoch": 2.1786032311410763, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.035697937011719, "learning_rate": 1e-05, "loss": 0.4089, "mean_token_accuracy": 0.8969889879226685, "num_tokens": 650261923.0, "step": 17126 }, { "epoch": 2.178730441419667, "ewc_loss": 0.1484375, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 0.0001220703125, "grad_norm": 7.550081729888916, "learning_rate": 1e-05, "loss": 0.4728, "mean_token_accuracy": 0.8861361742019653, "num_tokens": 650295562.0, "step": 17127 }, { "epoch": 2.1788576516982574, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.426785945892334, "learning_rate": 1e-05, "loss": 0.4837, "mean_token_accuracy": 0.8744767904281616, "num_tokens": 650331988.0, "step": 17128 }, { "epoch": 2.178984861976848, "ewc_loss": 0.140625, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 0.00011444091796875, "grad_norm": 7.211688041687012, "learning_rate": 1e-05, "loss": 0.4683, "mean_token_accuracy": 0.8848698735237122, "num_tokens": 650370572.0, "step": 17129 }, { "epoch": 2.179112072255438, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.608331680297852, "learning_rate": 1e-05, "loss": 0.4536, "mean_token_accuracy": 0.8879369497299194, "num_tokens": 650412838.0, "step": 17130 }, { "epoch": 2.1792392825340285, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.704911708831787, "learning_rate": 1e-05, "loss": 0.4728, "mean_token_accuracy": 0.8780419826507568, "num_tokens": 650455588.0, "step": 17131 }, { "epoch": 2.179366492812619, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.668602466583252, "learning_rate": 1e-05, "loss": 0.4477, "mean_token_accuracy": 0.8877716660499573, "num_tokens": 650496369.0, "step": 17132 }, { "epoch": 2.1794937030912096, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.233956813812256, "learning_rate": 1e-05, "loss": 0.4863, "mean_token_accuracy": 0.8678324818611145, "num_tokens": 650535366.0, "step": 17133 }, { "epoch": 2.1796209133698, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.795870304107666, "learning_rate": 1e-05, "loss": 0.4525, "mean_token_accuracy": 0.8867625594139099, "num_tokens": 650574171.0, "step": 17134 }, { "epoch": 2.1797481236483907, "ewc_loss": 0.11181640625, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.8908257484436035, "learning_rate": 1e-05, "loss": 0.4176, "mean_token_accuracy": 0.8916610479354858, "num_tokens": 650612132.0, "step": 17135 }, { "epoch": 2.179875333926981, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 0.00010967254638671875, "grad_norm": 7.073544025421143, "learning_rate": 1e-05, "loss": 0.4536, "mean_token_accuracy": 0.8875671029090881, "num_tokens": 650648813.0, "step": 17136 }, { "epoch": 2.1800025442055717, "ewc_loss": 0.10986328125, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.776027679443359, "learning_rate": 1e-05, "loss": 0.4449, "mean_token_accuracy": 0.8832864761352539, "num_tokens": 650678807.0, "step": 17137 }, { "epoch": 2.1801297544841622, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 0.00011777877807617188, "grad_norm": 7.416164398193359, "learning_rate": 1e-05, "loss": 0.5043, "mean_token_accuracy": 0.870072066783905, "num_tokens": 650716011.0, "step": 17138 }, { "epoch": 2.1802569647627528, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 5.920801639556885, "learning_rate": 1e-05, "loss": 0.4143, "mean_token_accuracy": 0.8941457867622375, "num_tokens": 650749763.0, "step": 17139 }, { "epoch": 2.1803841750413433, "ewc_loss": 0.1494140625, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.63297700881958, "learning_rate": 1e-05, "loss": 0.504, "mean_token_accuracy": 0.8769494295120239, "num_tokens": 650786977.0, "step": 17140 }, { "epoch": 2.180511385319934, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.298187255859375, "learning_rate": 1e-05, "loss": 0.5033, "mean_token_accuracy": 0.8668859004974365, "num_tokens": 650823293.0, "step": 17141 }, { "epoch": 2.1806385955985244, "ewc_loss": 0.1474609375, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 0.00012063980102539062, "grad_norm": 7.495323181152344, "learning_rate": 1e-05, "loss": 0.509, "mean_token_accuracy": 0.8746724128723145, "num_tokens": 650865037.0, "step": 17142 }, { "epoch": 2.180765805877115, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.536391735076904, "learning_rate": 1e-05, "loss": 0.4661, "mean_token_accuracy": 0.8834633827209473, "num_tokens": 650901803.0, "step": 17143 }, { "epoch": 2.1808930161557054, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 0.00011205673217773438, "grad_norm": 7.15143346786499, "learning_rate": 1e-05, "loss": 0.4667, "mean_token_accuracy": 0.8837355375289917, "num_tokens": 650942163.0, "step": 17144 }, { "epoch": 2.181020226434296, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.538835048675537, "learning_rate": 1e-05, "loss": 0.4787, "mean_token_accuracy": 0.875867486000061, "num_tokens": 650977989.0, "step": 17145 }, { "epoch": 2.1811474367128865, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.827225685119629, "learning_rate": 1e-05, "loss": 0.4672, "mean_token_accuracy": 0.8815061450004578, "num_tokens": 651016831.0, "step": 17146 }, { "epoch": 2.181274646991477, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.454890727996826, "learning_rate": 1e-05, "loss": 0.5039, "mean_token_accuracy": 0.8724471926689148, "num_tokens": 651052493.0, "step": 17147 }, { "epoch": 2.1814018572700675, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.575647830963135, "learning_rate": 1e-05, "loss": 0.5143, "mean_token_accuracy": 0.8653385639190674, "num_tokens": 651093140.0, "step": 17148 }, { "epoch": 2.181529067548658, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.33866548538208, "learning_rate": 1e-05, "loss": 0.4772, "mean_token_accuracy": 0.8702399134635925, "num_tokens": 651129367.0, "step": 17149 }, { "epoch": 2.1816562778272486, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.476013660430908, "learning_rate": 1e-05, "loss": 0.4622, "mean_token_accuracy": 0.8824608325958252, "num_tokens": 651169573.0, "step": 17150 }, { "epoch": 2.181783488105839, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.198102951049805, "learning_rate": 1e-05, "loss": 0.4369, "mean_token_accuracy": 0.887761116027832, "num_tokens": 651206771.0, "step": 17151 }, { "epoch": 2.1819106983844296, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.488155841827393, "learning_rate": 1e-05, "loss": 0.4447, "mean_token_accuracy": 0.8883568644523621, "num_tokens": 651251322.0, "step": 17152 }, { "epoch": 2.18203790866302, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 6.052593231201172, "learning_rate": 1e-05, "loss": 0.4408, "mean_token_accuracy": 0.8849999904632568, "num_tokens": 651287495.0, "step": 17153 }, { "epoch": 2.1821651189416107, "ewc_loss": 0.125, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.602102279663086, "learning_rate": 1e-05, "loss": 0.5068, "mean_token_accuracy": 0.8655384182929993, "num_tokens": 651326005.0, "step": 17154 }, { "epoch": 2.1822923292202008, "ewc_loss": 0.11181640625, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.891174793243408, "learning_rate": 1e-05, "loss": 0.5057, "mean_token_accuracy": 0.8648973703384399, "num_tokens": 651364843.0, "step": 17155 }, { "epoch": 2.1824195394987913, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.828402519226074, "learning_rate": 1e-05, "loss": 0.45, "mean_token_accuracy": 0.8854933381080627, "num_tokens": 651403981.0, "step": 17156 }, { "epoch": 2.182546749777382, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.746739387512207, "learning_rate": 1e-05, "loss": 0.4301, "mean_token_accuracy": 0.8868452310562134, "num_tokens": 651443751.0, "step": 17157 }, { "epoch": 2.1826739600559724, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 0.0001125335693359375, "grad_norm": 7.186069965362549, "learning_rate": 1e-05, "loss": 0.5152, "mean_token_accuracy": 0.8726402521133423, "num_tokens": 651487248.0, "step": 17158 }, { "epoch": 2.182801170334563, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.751192569732666, "learning_rate": 1e-05, "loss": 0.5141, "mean_token_accuracy": 0.8675994277000427, "num_tokens": 651523055.0, "step": 17159 }, { "epoch": 2.1829283806131534, "ewc_loss": 0.1494140625, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.579124450683594, "learning_rate": 1e-05, "loss": 0.4953, "mean_token_accuracy": 0.87636399269104, "num_tokens": 651554749.0, "step": 17160 }, { "epoch": 2.183055590891744, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.063847541809082, "learning_rate": 1e-05, "loss": 0.5371, "mean_token_accuracy": 0.8577823042869568, "num_tokens": 651597342.0, "step": 17161 }, { "epoch": 2.1831828011703345, "ewc_loss": 0.1533203125, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 0.00012683868408203125, "grad_norm": 7.660789966583252, "learning_rate": 1e-05, "loss": 0.5025, "mean_token_accuracy": 0.8785489797592163, "num_tokens": 651634578.0, "step": 17162 }, { "epoch": 2.183310011448925, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.507643699645996, "learning_rate": 1e-05, "loss": 0.4562, "mean_token_accuracy": 0.8855550289154053, "num_tokens": 651674561.0, "step": 17163 }, { "epoch": 2.1834372217275155, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 0.0001163482666015625, "grad_norm": 7.279262065887451, "learning_rate": 1e-05, "loss": 0.4636, "mean_token_accuracy": 0.8891616463661194, "num_tokens": 651713426.0, "step": 17164 }, { "epoch": 2.183564432006106, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 6.685177803039551, "learning_rate": 1e-05, "loss": 0.4773, "mean_token_accuracy": 0.881078839302063, "num_tokens": 651753267.0, "step": 17165 }, { "epoch": 2.1836916422846966, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.749227046966553, "learning_rate": 1e-05, "loss": 0.458, "mean_token_accuracy": 0.8835930824279785, "num_tokens": 651797580.0, "step": 17166 }, { "epoch": 2.183818852563287, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.70978307723999, "learning_rate": 1e-05, "loss": 0.4959, "mean_token_accuracy": 0.8702973127365112, "num_tokens": 651828294.0, "step": 17167 }, { "epoch": 2.1839460628418776, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.293771743774414, "learning_rate": 1e-05, "loss": 0.5301, "mean_token_accuracy": 0.8621118068695068, "num_tokens": 651865572.0, "step": 17168 }, { "epoch": 2.184073273120468, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.779928684234619, "learning_rate": 1e-05, "loss": 0.4698, "mean_token_accuracy": 0.8803514242172241, "num_tokens": 651897453.0, "step": 17169 }, { "epoch": 2.1842004833990587, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 5.970977306365967, "learning_rate": 1e-05, "loss": 0.4688, "mean_token_accuracy": 0.8731236457824707, "num_tokens": 651933096.0, "step": 17170 }, { "epoch": 2.1843276936776492, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.977944850921631, "learning_rate": 1e-05, "loss": 0.4485, "mean_token_accuracy": 0.8859094381332397, "num_tokens": 651970042.0, "step": 17171 }, { "epoch": 2.1844549039562398, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.835546493530273, "learning_rate": 1e-05, "loss": 0.4786, "mean_token_accuracy": 0.8736697435379028, "num_tokens": 652006513.0, "step": 17172 }, { "epoch": 2.1845821142348303, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 0.00011348724365234375, "grad_norm": 7.24649715423584, "learning_rate": 1e-05, "loss": 0.4578, "mean_token_accuracy": 0.8877779245376587, "num_tokens": 652042706.0, "step": 17173 }, { "epoch": 2.184709324513421, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 5.890420436859131, "learning_rate": 1e-05, "loss": 0.4388, "mean_token_accuracy": 0.8873612880706787, "num_tokens": 652084219.0, "step": 17174 }, { "epoch": 2.1848365347920113, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 0.00011968612670898438, "grad_norm": 7.459619045257568, "learning_rate": 1e-05, "loss": 0.4825, "mean_token_accuracy": 0.8858963251113892, "num_tokens": 652130640.0, "step": 17175 }, { "epoch": 2.184963745070602, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.1334710121154785, "learning_rate": 1e-05, "loss": 0.4832, "mean_token_accuracy": 0.873167872428894, "num_tokens": 652169812.0, "step": 17176 }, { "epoch": 2.1850909553491924, "ewc_loss": 0.1455078125, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 0.00011873245239257812, "grad_norm": 7.458367347717285, "learning_rate": 1e-05, "loss": 0.4538, "mean_token_accuracy": 0.8945564031600952, "num_tokens": 652206317.0, "step": 17177 }, { "epoch": 2.185218165627783, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.380099296569824, "learning_rate": 1e-05, "loss": 0.4861, "mean_token_accuracy": 0.8739858865737915, "num_tokens": 652241420.0, "step": 17178 }, { "epoch": 2.1853453759063735, "ewc_loss": 0.140625, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 0.00011396408081054688, "grad_norm": 7.225215911865234, "learning_rate": 1e-05, "loss": 0.4752, "mean_token_accuracy": 0.881747305393219, "num_tokens": 652283951.0, "step": 17179 }, { "epoch": 2.1854725861849635, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.4526238441467285, "learning_rate": 1e-05, "loss": 0.4707, "mean_token_accuracy": 0.8806757926940918, "num_tokens": 652320729.0, "step": 17180 }, { "epoch": 2.185599796463554, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.930130958557129, "learning_rate": 1e-05, "loss": 0.5007, "mean_token_accuracy": 0.8727178573608398, "num_tokens": 652362969.0, "step": 17181 }, { "epoch": 2.1857270067421446, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.418064594268799, "learning_rate": 1e-05, "loss": 0.4879, "mean_token_accuracy": 0.873866081237793, "num_tokens": 652403120.0, "step": 17182 }, { "epoch": 2.185854217020735, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.69712495803833, "learning_rate": 1e-05, "loss": 0.5315, "mean_token_accuracy": 0.8632798194885254, "num_tokens": 652436904.0, "step": 17183 }, { "epoch": 2.1859814272993257, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.324314117431641, "learning_rate": 1e-05, "loss": 0.4378, "mean_token_accuracy": 0.8893201947212219, "num_tokens": 652474938.0, "step": 17184 }, { "epoch": 2.186108637577916, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.564995765686035, "learning_rate": 1e-05, "loss": 0.4629, "mean_token_accuracy": 0.881691575050354, "num_tokens": 652503718.0, "step": 17185 }, { "epoch": 2.1862358478565067, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.208747386932373, "learning_rate": 1e-05, "loss": 0.4594, "mean_token_accuracy": 0.8804723024368286, "num_tokens": 652542235.0, "step": 17186 }, { "epoch": 2.1863630581350972, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.540774345397949, "learning_rate": 1e-05, "loss": 0.5217, "mean_token_accuracy": 0.8613905906677246, "num_tokens": 652584779.0, "step": 17187 }, { "epoch": 2.1864902684136878, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 6.071305751800537, "learning_rate": 1e-05, "loss": 0.4866, "mean_token_accuracy": 0.8697656393051147, "num_tokens": 652627132.0, "step": 17188 }, { "epoch": 2.1866174786922783, "ewc_loss": 0.125, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.606112480163574, "learning_rate": 1e-05, "loss": 0.5009, "mean_token_accuracy": 0.871552050113678, "num_tokens": 652670156.0, "step": 17189 }, { "epoch": 2.186744688970869, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.940110206604004, "learning_rate": 1e-05, "loss": 0.4294, "mean_token_accuracy": 0.8887256383895874, "num_tokens": 652705246.0, "step": 17190 }, { "epoch": 2.1868718992494594, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.764549255371094, "learning_rate": 1e-05, "loss": 0.49, "mean_token_accuracy": 0.8774396777153015, "num_tokens": 652742234.0, "step": 17191 }, { "epoch": 2.18699910952805, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.827400207519531, "learning_rate": 1e-05, "loss": 0.4433, "mean_token_accuracy": 0.8845266699790955, "num_tokens": 652779100.0, "step": 17192 }, { "epoch": 2.1871263198066404, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 7.039217472076416, "learning_rate": 1e-05, "loss": 0.4795, "mean_token_accuracy": 0.8814043402671814, "num_tokens": 652821174.0, "step": 17193 }, { "epoch": 2.187253530085231, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.811161994934082, "learning_rate": 1e-05, "loss": 0.4253, "mean_token_accuracy": 0.8877079486846924, "num_tokens": 652856271.0, "step": 17194 }, { "epoch": 2.1873807403638215, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 0.00011730194091796875, "grad_norm": 7.3119401931762695, "learning_rate": 1e-05, "loss": 0.5011, "mean_token_accuracy": 0.875922441482544, "num_tokens": 652898737.0, "step": 17195 }, { "epoch": 2.187507950642412, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 5.981547832489014, "learning_rate": 1e-05, "loss": 0.458, "mean_token_accuracy": 0.8813730478286743, "num_tokens": 652938778.0, "step": 17196 }, { "epoch": 2.1876351609210025, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 0.00011777877807617188, "grad_norm": 7.402365207672119, "learning_rate": 1e-05, "loss": 0.504, "mean_token_accuracy": 0.8767090439796448, "num_tokens": 652982202.0, "step": 17197 }, { "epoch": 2.187762371199593, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.284271240234375, "learning_rate": 1e-05, "loss": 0.4967, "mean_token_accuracy": 0.8683856725692749, "num_tokens": 653024990.0, "step": 17198 }, { "epoch": 2.1878895814781836, "ewc_loss": 0.140625, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 0.00011396408081054688, "grad_norm": 7.157424449920654, "learning_rate": 1e-05, "loss": 0.5107, "mean_token_accuracy": 0.8711187839508057, "num_tokens": 653061913.0, "step": 17199 }, { "epoch": 2.188016791756774, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.490546703338623, "learning_rate": 1e-05, "loss": 0.4833, "mean_token_accuracy": 0.8767082691192627, "num_tokens": 653105944.0, "step": 17200 }, { "epoch": 2.1881440020353646, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.680748462677002, "learning_rate": 1e-05, "loss": 0.4405, "mean_token_accuracy": 0.8898391723632812, "num_tokens": 653139819.0, "step": 17201 }, { "epoch": 2.188271212313955, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.59296178817749, "learning_rate": 1e-05, "loss": 0.4502, "mean_token_accuracy": 0.8892842531204224, "num_tokens": 653184569.0, "step": 17202 }, { "epoch": 2.1883984225925452, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.185133457183838, "learning_rate": 1e-05, "loss": 0.4004, "mean_token_accuracy": 0.9007676839828491, "num_tokens": 653220346.0, "step": 17203 }, { "epoch": 2.188525632871136, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.80618953704834, "learning_rate": 1e-05, "loss": 0.4804, "mean_token_accuracy": 0.8771101236343384, "num_tokens": 653257288.0, "step": 17204 }, { "epoch": 2.1886528431497263, "ewc_loss": 0.10986328125, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.811954498291016, "learning_rate": 1e-05, "loss": 0.4809, "mean_token_accuracy": 0.8715430498123169, "num_tokens": 653293055.0, "step": 17205 }, { "epoch": 2.188780053428317, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 0.0001125335693359375, "grad_norm": 7.177781581878662, "learning_rate": 1e-05, "loss": 0.5111, "mean_token_accuracy": 0.8690919876098633, "num_tokens": 653329988.0, "step": 17206 }, { "epoch": 2.1889072637069074, "ewc_loss": 0.11083984375, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.69766092300415, "learning_rate": 1e-05, "loss": 0.4232, "mean_token_accuracy": 0.8920613527297974, "num_tokens": 653367704.0, "step": 17207 }, { "epoch": 2.189034473985498, "ewc_loss": 0.150390625, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 0.0001239776611328125, "grad_norm": 7.631128787994385, "learning_rate": 1e-05, "loss": 0.4905, "mean_token_accuracy": 0.8831921815872192, "num_tokens": 653403915.0, "step": 17208 }, { "epoch": 2.1891616842640884, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.0209736824035645, "learning_rate": 1e-05, "loss": 0.4479, "mean_token_accuracy": 0.8848460912704468, "num_tokens": 653438045.0, "step": 17209 }, { "epoch": 2.189288894542679, "ewc_loss": 0.154296875, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 0.0001277923583984375, "grad_norm": 7.787870407104492, "learning_rate": 1e-05, "loss": 0.4973, "mean_token_accuracy": 0.877184271812439, "num_tokens": 653469552.0, "step": 17210 }, { "epoch": 2.1894161048212695, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 6.513350009918213, "learning_rate": 1e-05, "loss": 0.4674, "mean_token_accuracy": 0.8836706876754761, "num_tokens": 653511809.0, "step": 17211 }, { "epoch": 2.18954331509986, "ewc_loss": 0.1455078125, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 0.00011920928955078125, "grad_norm": 7.490612506866455, "learning_rate": 1e-05, "loss": 0.4554, "mean_token_accuracy": 0.8905489444732666, "num_tokens": 653543812.0, "step": 17212 }, { "epoch": 2.1896705253784505, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 0.00011014938354492188, "grad_norm": 6.671997547149658, "learning_rate": 1e-05, "loss": 0.5141, "mean_token_accuracy": 0.8687331080436707, "num_tokens": 653586232.0, "step": 17213 }, { "epoch": 2.189797735657041, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 0.0001125335693359375, "grad_norm": 7.088866710662842, "learning_rate": 1e-05, "loss": 0.4942, "mean_token_accuracy": 0.8755140900611877, "num_tokens": 653626700.0, "step": 17214 }, { "epoch": 2.1899249459356316, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.584470272064209, "learning_rate": 1e-05, "loss": 0.4852, "mean_token_accuracy": 0.8747802972793579, "num_tokens": 653666648.0, "step": 17215 }, { "epoch": 2.190052156214222, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.781208038330078, "learning_rate": 1e-05, "loss": 0.4539, "mean_token_accuracy": 0.8859657645225525, "num_tokens": 653698991.0, "step": 17216 }, { "epoch": 2.1901793664928126, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.437208652496338, "learning_rate": 1e-05, "loss": 0.4405, "mean_token_accuracy": 0.8872983455657959, "num_tokens": 653737696.0, "step": 17217 }, { "epoch": 2.190306576771403, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.603090763092041, "learning_rate": 1e-05, "loss": 0.4477, "mean_token_accuracy": 0.8854833841323853, "num_tokens": 653776386.0, "step": 17218 }, { "epoch": 2.1904337870499937, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.281010627746582, "learning_rate": 1e-05, "loss": 0.435, "mean_token_accuracy": 0.8888986110687256, "num_tokens": 653811346.0, "step": 17219 }, { "epoch": 2.1905609973285842, "ewc_loss": 0.125, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.56867790222168, "learning_rate": 1e-05, "loss": 0.4874, "mean_token_accuracy": 0.872350811958313, "num_tokens": 653847966.0, "step": 17220 }, { "epoch": 2.1906882076071748, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.102860927581787, "learning_rate": 1e-05, "loss": 0.4078, "mean_token_accuracy": 0.892970860004425, "num_tokens": 653882381.0, "step": 17221 }, { "epoch": 2.1908154178857653, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.6541547775268555, "learning_rate": 1e-05, "loss": 0.4704, "mean_token_accuracy": 0.8841096758842468, "num_tokens": 653925026.0, "step": 17222 }, { "epoch": 2.190942628164356, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.940973281860352, "learning_rate": 1e-05, "loss": 0.4555, "mean_token_accuracy": 0.8808654546737671, "num_tokens": 653962210.0, "step": 17223 }, { "epoch": 2.1910698384429463, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.8646721839904785, "learning_rate": 1e-05, "loss": 0.4674, "mean_token_accuracy": 0.8791526556015015, "num_tokens": 653998713.0, "step": 17224 }, { "epoch": 2.191197048721537, "ewc_loss": 0.11083984375, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.821814060211182, "learning_rate": 1e-05, "loss": 0.4755, "mean_token_accuracy": 0.8729639053344727, "num_tokens": 654033383.0, "step": 17225 }, { "epoch": 2.1913242590001274, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 0.0001125335693359375, "grad_norm": 7.15548038482666, "learning_rate": 1e-05, "loss": 0.5012, "mean_token_accuracy": 0.8760353326797485, "num_tokens": 654074712.0, "step": 17226 }, { "epoch": 2.191451469278718, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.839730739593506, "learning_rate": 1e-05, "loss": 0.4947, "mean_token_accuracy": 0.873680830001831, "num_tokens": 654110650.0, "step": 17227 }, { "epoch": 2.191578679557308, "ewc_loss": 0.1455078125, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 0.00011873245239257812, "grad_norm": 7.453127384185791, "learning_rate": 1e-05, "loss": 0.4506, "mean_token_accuracy": 0.8915149569511414, "num_tokens": 654146628.0, "step": 17228 }, { "epoch": 2.1917058898358985, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.081169128417969, "learning_rate": 1e-05, "loss": 0.4985, "mean_token_accuracy": 0.869756281375885, "num_tokens": 654192239.0, "step": 17229 }, { "epoch": 2.191833100114489, "ewc_loss": 0.1474609375, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 0.00012111663818359375, "grad_norm": 7.495402812957764, "learning_rate": 1e-05, "loss": 0.5615, "mean_token_accuracy": 0.8574779033660889, "num_tokens": 654232458.0, "step": 17230 }, { "epoch": 2.1919603103930796, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.433345794677734, "learning_rate": 1e-05, "loss": 0.5389, "mean_token_accuracy": 0.8587861061096191, "num_tokens": 654271266.0, "step": 17231 }, { "epoch": 2.19208752067167, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 0.0001125335693359375, "grad_norm": 7.156398773193359, "learning_rate": 1e-05, "loss": 0.4865, "mean_token_accuracy": 0.8778551816940308, "num_tokens": 654307801.0, "step": 17232 }, { "epoch": 2.1922147309502606, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.600543022155762, "learning_rate": 1e-05, "loss": 0.4442, "mean_token_accuracy": 0.8928041458129883, "num_tokens": 654348903.0, "step": 17233 }, { "epoch": 2.192341941228851, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.644664287567139, "learning_rate": 1e-05, "loss": 0.4593, "mean_token_accuracy": 0.8835862874984741, "num_tokens": 654388659.0, "step": 17234 }, { "epoch": 2.1924691515074417, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.681962966918945, "learning_rate": 1e-05, "loss": 0.4645, "mean_token_accuracy": 0.8809579014778137, "num_tokens": 654425231.0, "step": 17235 }, { "epoch": 2.1925963617860322, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.159586429595947, "learning_rate": 1e-05, "loss": 0.4507, "mean_token_accuracy": 0.8845804929733276, "num_tokens": 654457560.0, "step": 17236 }, { "epoch": 2.1927235720646228, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.8509202003479, "learning_rate": 1e-05, "loss": 0.487, "mean_token_accuracy": 0.8760859966278076, "num_tokens": 654491000.0, "step": 17237 }, { "epoch": 2.1928507823432133, "ewc_loss": 0.11083984375, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.824519157409668, "learning_rate": 1e-05, "loss": 0.4711, "mean_token_accuracy": 0.8752595782279968, "num_tokens": 654531142.0, "step": 17238 }, { "epoch": 2.192977992621804, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 0.00011205673217773438, "grad_norm": 7.176197052001953, "learning_rate": 1e-05, "loss": 0.4902, "mean_token_accuracy": 0.8763118982315063, "num_tokens": 654574133.0, "step": 17239 }, { "epoch": 2.1931052029003943, "ewc_loss": 0.11083984375, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.759540557861328, "learning_rate": 1e-05, "loss": 0.4803, "mean_token_accuracy": 0.87204909324646, "num_tokens": 654612745.0, "step": 17240 }, { "epoch": 2.193232413178985, "ewc_loss": 0.1474609375, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 0.00012111663818359375, "grad_norm": 7.54569673538208, "learning_rate": 1e-05, "loss": 0.4941, "mean_token_accuracy": 0.8794780969619751, "num_tokens": 654646097.0, "step": 17241 }, { "epoch": 2.1933596234575754, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.024020671844482, "learning_rate": 1e-05, "loss": 0.4778, "mean_token_accuracy": 0.8747884631156921, "num_tokens": 654688216.0, "step": 17242 }, { "epoch": 2.193486833736166, "ewc_loss": 0.1513671875, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 0.00012493133544921875, "grad_norm": 7.692227363586426, "learning_rate": 1e-05, "loss": 0.4417, "mean_token_accuracy": 0.8997729420661926, "num_tokens": 654725697.0, "step": 17243 }, { "epoch": 2.1936140440147565, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.45066499710083, "learning_rate": 1e-05, "loss": 0.494, "mean_token_accuracy": 0.8778132796287537, "num_tokens": 654762440.0, "step": 17244 }, { "epoch": 2.193741254293347, "ewc_loss": 0.1455078125, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 0.00011920928955078125, "grad_norm": 7.435698509216309, "learning_rate": 1e-05, "loss": 0.5144, "mean_token_accuracy": 0.8703339099884033, "num_tokens": 654795843.0, "step": 17245 }, { "epoch": 2.1938684645719375, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.608387470245361, "learning_rate": 1e-05, "loss": 0.4662, "mean_token_accuracy": 0.8809905648231506, "num_tokens": 654835212.0, "step": 17246 }, { "epoch": 2.193995674850528, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 7.056917190551758, "learning_rate": 1e-05, "loss": 0.512, "mean_token_accuracy": 0.8704664707183838, "num_tokens": 654869092.0, "step": 17247 }, { "epoch": 2.1941228851291186, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.555232524871826, "learning_rate": 1e-05, "loss": 0.4601, "mean_token_accuracy": 0.8839694261550903, "num_tokens": 654907202.0, "step": 17248 }, { "epoch": 2.194250095407709, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.747198581695557, "learning_rate": 1e-05, "loss": 0.4673, "mean_token_accuracy": 0.8793971538543701, "num_tokens": 654938867.0, "step": 17249 }, { "epoch": 2.1943773056862996, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.431878089904785, "learning_rate": 1e-05, "loss": 0.4897, "mean_token_accuracy": 0.8731094002723694, "num_tokens": 654978838.0, "step": 17250 }, { "epoch": 2.19450451596489, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.573019981384277, "learning_rate": 1e-05, "loss": 0.4503, "mean_token_accuracy": 0.8854995965957642, "num_tokens": 655016987.0, "step": 17251 }, { "epoch": 2.1946317262434807, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.292973518371582, "learning_rate": 1e-05, "loss": 0.4642, "mean_token_accuracy": 0.8793566226959229, "num_tokens": 655056797.0, "step": 17252 }, { "epoch": 2.1947589365220708, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.519787788391113, "learning_rate": 1e-05, "loss": 0.4652, "mean_token_accuracy": 0.8817481994628906, "num_tokens": 655094959.0, "step": 17253 }, { "epoch": 2.1948861468006613, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.132428169250488, "learning_rate": 1e-05, "loss": 0.4902, "mean_token_accuracy": 0.8688212633132935, "num_tokens": 655136490.0, "step": 17254 }, { "epoch": 2.195013357079252, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.582602500915527, "learning_rate": 1e-05, "loss": 0.47, "mean_token_accuracy": 0.8810197114944458, "num_tokens": 655175174.0, "step": 17255 }, { "epoch": 2.1951405673578424, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 5.9698567390441895, "learning_rate": 1e-05, "loss": 0.4771, "mean_token_accuracy": 0.8719760179519653, "num_tokens": 655214753.0, "step": 17256 }, { "epoch": 2.195267777636433, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.758060455322266, "learning_rate": 1e-05, "loss": 0.4346, "mean_token_accuracy": 0.8924965858459473, "num_tokens": 655249737.0, "step": 17257 }, { "epoch": 2.1953949879150234, "ewc_loss": 0.11083984375, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.825444221496582, "learning_rate": 1e-05, "loss": 0.4382, "mean_token_accuracy": 0.8859254121780396, "num_tokens": 655286553.0, "step": 17258 }, { "epoch": 2.195522198193614, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 0.00010967254638671875, "grad_norm": 7.0575785636901855, "learning_rate": 1e-05, "loss": 0.477, "mean_token_accuracy": 0.8807377815246582, "num_tokens": 655320409.0, "step": 17259 }, { "epoch": 2.1956494084722045, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.773471355438232, "learning_rate": 1e-05, "loss": 0.459, "mean_token_accuracy": 0.8777405619621277, "num_tokens": 655362436.0, "step": 17260 }, { "epoch": 2.195776618750795, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 0.000118255615234375, "grad_norm": 7.416191101074219, "learning_rate": 1e-05, "loss": 0.501, "mean_token_accuracy": 0.8737953305244446, "num_tokens": 655405059.0, "step": 17261 }, { "epoch": 2.1959038290293855, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 5.950830936431885, "learning_rate": 1e-05, "loss": 0.475, "mean_token_accuracy": 0.8741341233253479, "num_tokens": 655448688.0, "step": 17262 }, { "epoch": 2.196031039307976, "ewc_loss": 0.1484375, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 0.0001220703125, "grad_norm": 7.617528915405273, "learning_rate": 1e-05, "loss": 0.4782, "mean_token_accuracy": 0.8836533427238464, "num_tokens": 655486361.0, "step": 17263 }, { "epoch": 2.1961582495865666, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.369326591491699, "learning_rate": 1e-05, "loss": 0.4582, "mean_token_accuracy": 0.879385232925415, "num_tokens": 655518610.0, "step": 17264 }, { "epoch": 2.196285459865157, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 0.00011777877807617188, "grad_norm": 7.391223430633545, "learning_rate": 1e-05, "loss": 0.5085, "mean_token_accuracy": 0.8746224045753479, "num_tokens": 655557860.0, "step": 17265 }, { "epoch": 2.1964126701437476, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 6.6098246574401855, "learning_rate": 1e-05, "loss": 0.4861, "mean_token_accuracy": 0.8779283165931702, "num_tokens": 655597885.0, "step": 17266 }, { "epoch": 2.196539880422338, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.890573024749756, "learning_rate": 1e-05, "loss": 0.4458, "mean_token_accuracy": 0.8877400159835815, "num_tokens": 655632847.0, "step": 17267 }, { "epoch": 2.1966670907009287, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.6684746742248535, "learning_rate": 1e-05, "loss": 0.4983, "mean_token_accuracy": 0.8748270273208618, "num_tokens": 655665374.0, "step": 17268 }, { "epoch": 2.196794300979519, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.378279685974121, "learning_rate": 1e-05, "loss": 0.4514, "mean_token_accuracy": 0.8840559124946594, "num_tokens": 655705926.0, "step": 17269 }, { "epoch": 2.1969215112581097, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.746911525726318, "learning_rate": 1e-05, "loss": 0.4359, "mean_token_accuracy": 0.8897155523300171, "num_tokens": 655743500.0, "step": 17270 }, { "epoch": 2.1970487215367003, "ewc_loss": 0.11376953125, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 6.00020694732666, "learning_rate": 1e-05, "loss": 0.4843, "mean_token_accuracy": 0.8689232468605042, "num_tokens": 655788063.0, "step": 17271 }, { "epoch": 2.197175931815291, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.958474159240723, "learning_rate": 1e-05, "loss": 0.4471, "mean_token_accuracy": 0.8887736797332764, "num_tokens": 655828393.0, "step": 17272 }, { "epoch": 2.1973031420938813, "ewc_loss": 0.11181640625, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.8004021644592285, "learning_rate": 1e-05, "loss": 0.4395, "mean_token_accuracy": 0.8840359449386597, "num_tokens": 655871666.0, "step": 17273 }, { "epoch": 2.197430352372472, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 0.00011491775512695312, "grad_norm": 7.288973331451416, "learning_rate": 1e-05, "loss": 0.4455, "mean_token_accuracy": 0.8940337300300598, "num_tokens": 655905396.0, "step": 17274 }, { "epoch": 2.1975575626510624, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 5.849776268005371, "learning_rate": 1e-05, "loss": 0.4686, "mean_token_accuracy": 0.8757790327072144, "num_tokens": 655944515.0, "step": 17275 }, { "epoch": 2.197684772929653, "ewc_loss": 0.1494140625, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.586883544921875, "learning_rate": 1e-05, "loss": 0.5366, "mean_token_accuracy": 0.8677452206611633, "num_tokens": 655979773.0, "step": 17276 }, { "epoch": 2.1978119832082434, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.161677360534668, "learning_rate": 1e-05, "loss": 0.4717, "mean_token_accuracy": 0.8801376819610596, "num_tokens": 656017786.0, "step": 17277 }, { "epoch": 2.1979391934868335, "ewc_loss": 0.1484375, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 0.0001220703125, "grad_norm": 7.586533546447754, "learning_rate": 1e-05, "loss": 0.5735, "mean_token_accuracy": 0.8541266918182373, "num_tokens": 656058776.0, "step": 17278 }, { "epoch": 2.198066403765424, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.475358009338379, "learning_rate": 1e-05, "loss": 0.473, "mean_token_accuracy": 0.8829336166381836, "num_tokens": 656100231.0, "step": 17279 }, { "epoch": 2.1981936140440146, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 0.00011587142944335938, "grad_norm": 7.405341148376465, "learning_rate": 1e-05, "loss": 0.4881, "mean_token_accuracy": 0.879381537437439, "num_tokens": 656131786.0, "step": 17280 }, { "epoch": 2.198320824322605, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.579853057861328, "learning_rate": 1e-05, "loss": 0.4619, "mean_token_accuracy": 0.8831365704536438, "num_tokens": 656172041.0, "step": 17281 }, { "epoch": 2.1984480346011956, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 6.98103141784668, "learning_rate": 1e-05, "loss": 0.4538, "mean_token_accuracy": 0.8867071866989136, "num_tokens": 656203152.0, "step": 17282 }, { "epoch": 2.198575244879786, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.521438121795654, "learning_rate": 1e-05, "loss": 0.508, "mean_token_accuracy": 0.8720262050628662, "num_tokens": 656241568.0, "step": 17283 }, { "epoch": 2.1987024551583767, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.682474613189697, "learning_rate": 1e-05, "loss": 0.5128, "mean_token_accuracy": 0.8644036054611206, "num_tokens": 656275221.0, "step": 17284 }, { "epoch": 2.1988296654369672, "ewc_loss": 0.125, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.418368816375732, "learning_rate": 1e-05, "loss": 0.498, "mean_token_accuracy": 0.8706799745559692, "num_tokens": 656317535.0, "step": 17285 }, { "epoch": 2.1989568757155578, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.506902694702148, "learning_rate": 1e-05, "loss": 0.4633, "mean_token_accuracy": 0.8834155797958374, "num_tokens": 656351440.0, "step": 17286 }, { "epoch": 2.1990840859941483, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.2896013259887695, "learning_rate": 1e-05, "loss": 0.4848, "mean_token_accuracy": 0.8728330731391907, "num_tokens": 656390942.0, "step": 17287 }, { "epoch": 2.199211296272739, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.451948642730713, "learning_rate": 1e-05, "loss": 0.4564, "mean_token_accuracy": 0.8845183849334717, "num_tokens": 656427314.0, "step": 17288 }, { "epoch": 2.1993385065513293, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.157242298126221, "learning_rate": 1e-05, "loss": 0.4825, "mean_token_accuracy": 0.8682904243469238, "num_tokens": 656464014.0, "step": 17289 }, { "epoch": 2.19946571682992, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.487077236175537, "learning_rate": 1e-05, "loss": 0.4625, "mean_token_accuracy": 0.8788812160491943, "num_tokens": 656497783.0, "step": 17290 }, { "epoch": 2.1995929271085104, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 6.010963439941406, "learning_rate": 1e-05, "loss": 0.4825, "mean_token_accuracy": 0.8755146265029907, "num_tokens": 656538160.0, "step": 17291 }, { "epoch": 2.199720137387101, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.6029767990112305, "learning_rate": 1e-05, "loss": 0.4522, "mean_token_accuracy": 0.8854813575744629, "num_tokens": 656578296.0, "step": 17292 }, { "epoch": 2.1998473476656915, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.865143775939941, "learning_rate": 1e-05, "loss": 0.4643, "mean_token_accuracy": 0.8793286085128784, "num_tokens": 656620662.0, "step": 17293 }, { "epoch": 2.199974557944282, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.838479518890381, "learning_rate": 1e-05, "loss": 0.4748, "mean_token_accuracy": 0.8805674314498901, "num_tokens": 656665511.0, "step": 17294 }, { "epoch": 2.2001017682228725, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.737857341766357, "learning_rate": 1e-05, "loss": 0.4727, "mean_token_accuracy": 0.8772823810577393, "num_tokens": 656703079.0, "step": 17295 }, { "epoch": 2.200228978501463, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.0001125335693359375, "grad_norm": 7.207200050354004, "learning_rate": 1e-05, "loss": 0.5105, "mean_token_accuracy": 0.8705689907073975, "num_tokens": 656741430.0, "step": 17296 }, { "epoch": 2.2003561887800536, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.748138427734375, "learning_rate": 1e-05, "loss": 0.4309, "mean_token_accuracy": 0.8875119686126709, "num_tokens": 656778836.0, "step": 17297 }, { "epoch": 2.200483399058644, "ewc_loss": 0.150390625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.611919403076172, "learning_rate": 1e-05, "loss": 0.4948, "mean_token_accuracy": 0.8794147968292236, "num_tokens": 656811331.0, "step": 17298 }, { "epoch": 2.2006106093372346, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.085052013397217, "learning_rate": 1e-05, "loss": 0.4456, "mean_token_accuracy": 0.8875102996826172, "num_tokens": 656845579.0, "step": 17299 }, { "epoch": 2.200737819615825, "ewc_loss": 0.154296875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00012683868408203125, "grad_norm": 7.687242031097412, "learning_rate": 1e-05, "loss": 0.5067, "mean_token_accuracy": 0.8746887445449829, "num_tokens": 656880718.0, "step": 17300 }, { "epoch": 2.2008650298944152, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 6.527050018310547, "learning_rate": 1e-05, "loss": 0.4733, "mean_token_accuracy": 0.8792077302932739, "num_tokens": 656918350.0, "step": 17301 }, { "epoch": 2.200992240173006, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00011682510375976562, "grad_norm": 7.29979133605957, "learning_rate": 1e-05, "loss": 0.5001, "mean_token_accuracy": 0.8749852180480957, "num_tokens": 656954661.0, "step": 17302 }, { "epoch": 2.2011194504515963, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 6.683282375335693, "learning_rate": 1e-05, "loss": 0.4723, "mean_token_accuracy": 0.881156325340271, "num_tokens": 656993233.0, "step": 17303 }, { "epoch": 2.201246660730187, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.788632869720459, "learning_rate": 1e-05, "loss": 0.4913, "mean_token_accuracy": 0.8730005025863647, "num_tokens": 657032843.0, "step": 17304 }, { "epoch": 2.2013738710087773, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.678851127624512, "learning_rate": 1e-05, "loss": 0.4943, "mean_token_accuracy": 0.8769379258155823, "num_tokens": 657078625.0, "step": 17305 }, { "epoch": 2.201501081287368, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.3469157218933105, "learning_rate": 1e-05, "loss": 0.4744, "mean_token_accuracy": 0.8808749318122864, "num_tokens": 657113696.0, "step": 17306 }, { "epoch": 2.2016282915659584, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.746246337890625, "learning_rate": 1e-05, "loss": 0.4923, "mean_token_accuracy": 0.875291645526886, "num_tokens": 657149224.0, "step": 17307 }, { "epoch": 2.201755501844549, "ewc_loss": 0.11474609375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 6.038085460662842, "learning_rate": 1e-05, "loss": 0.475, "mean_token_accuracy": 0.8721674680709839, "num_tokens": 657181946.0, "step": 17308 }, { "epoch": 2.2018827121231395, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.895193576812744, "learning_rate": 1e-05, "loss": 0.5527, "mean_token_accuracy": 0.8544632196426392, "num_tokens": 657214372.0, "step": 17309 }, { "epoch": 2.20200992240173, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.869426250457764, "learning_rate": 1e-05, "loss": 0.4575, "mean_token_accuracy": 0.8784220218658447, "num_tokens": 657255572.0, "step": 17310 }, { "epoch": 2.2021371326803205, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00011110305786132812, "grad_norm": 7.128870010375977, "learning_rate": 1e-05, "loss": 0.4683, "mean_token_accuracy": 0.8795136213302612, "num_tokens": 657288336.0, "step": 17311 }, { "epoch": 2.202264342958911, "ewc_loss": 0.11474609375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 5.861631393432617, "learning_rate": 1e-05, "loss": 0.4792, "mean_token_accuracy": 0.8741558790206909, "num_tokens": 657327954.0, "step": 17312 }, { "epoch": 2.2023915532375016, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.000118255615234375, "grad_norm": 7.358513355255127, "learning_rate": 1e-05, "loss": 0.5218, "mean_token_accuracy": 0.8694635629653931, "num_tokens": 657366691.0, "step": 17313 }, { "epoch": 2.202518763516092, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.014293193817139, "learning_rate": 1e-05, "loss": 0.4733, "mean_token_accuracy": 0.8765155076980591, "num_tokens": 657405694.0, "step": 17314 }, { "epoch": 2.2026459737946826, "ewc_loss": 0.1484375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00012111663818359375, "grad_norm": 7.440847396850586, "learning_rate": 1e-05, "loss": 0.4452, "mean_token_accuracy": 0.8948138952255249, "num_tokens": 657442461.0, "step": 17315 }, { "epoch": 2.202773184073273, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.246980667114258, "learning_rate": 1e-05, "loss": 0.4565, "mean_token_accuracy": 0.8842791318893433, "num_tokens": 657476384.0, "step": 17316 }, { "epoch": 2.2029003943518637, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00011777877807617188, "grad_norm": 7.298473358154297, "learning_rate": 1e-05, "loss": 0.5232, "mean_token_accuracy": 0.8687307834625244, "num_tokens": 657511291.0, "step": 17317 }, { "epoch": 2.203027604630454, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.388749599456787, "learning_rate": 1e-05, "loss": 0.4894, "mean_token_accuracy": 0.8701659440994263, "num_tokens": 657546820.0, "step": 17318 }, { "epoch": 2.2031548149090447, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00011110305786132812, "grad_norm": 7.026859283447266, "learning_rate": 1e-05, "loss": 0.4707, "mean_token_accuracy": 0.8813563585281372, "num_tokens": 657582513.0, "step": 17319 }, { "epoch": 2.2032820251876353, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.405110836029053, "learning_rate": 1e-05, "loss": 0.4699, "mean_token_accuracy": 0.8802274465560913, "num_tokens": 657617108.0, "step": 17320 }, { "epoch": 2.203409235466226, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.739882469177246, "learning_rate": 1e-05, "loss": 0.4477, "mean_token_accuracy": 0.8865568041801453, "num_tokens": 657656220.0, "step": 17321 }, { "epoch": 2.2035364457448163, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.368871212005615, "learning_rate": 1e-05, "loss": 0.4457, "mean_token_accuracy": 0.8875476121902466, "num_tokens": 657691386.0, "step": 17322 }, { "epoch": 2.203663656023407, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.537262439727783, "learning_rate": 1e-05, "loss": 0.5111, "mean_token_accuracy": 0.8647903203964233, "num_tokens": 657728118.0, "step": 17323 }, { "epoch": 2.2037908663019974, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.278261184692383, "learning_rate": 1e-05, "loss": 0.431, "mean_token_accuracy": 0.8916889429092407, "num_tokens": 657768832.0, "step": 17324 }, { "epoch": 2.203918076580588, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.450870990753174, "learning_rate": 1e-05, "loss": 0.4562, "mean_token_accuracy": 0.8837548494338989, "num_tokens": 657805511.0, "step": 17325 }, { "epoch": 2.204045286859178, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.1767144203186035, "learning_rate": 1e-05, "loss": 0.4807, "mean_token_accuracy": 0.876288652420044, "num_tokens": 657843250.0, "step": 17326 }, { "epoch": 2.2041724971377685, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.445723533630371, "learning_rate": 1e-05, "loss": 0.4345, "mean_token_accuracy": 0.8883837461471558, "num_tokens": 657878882.0, "step": 17327 }, { "epoch": 2.204299707416359, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.0650224685668945, "learning_rate": 1e-05, "loss": 0.4063, "mean_token_accuracy": 0.8991884589195251, "num_tokens": 657917007.0, "step": 17328 }, { "epoch": 2.2044269176949496, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.5249738693237305, "learning_rate": 1e-05, "loss": 0.508, "mean_token_accuracy": 0.868545651435852, "num_tokens": 657956741.0, "step": 17329 }, { "epoch": 2.20455412797354, "ewc_loss": 0.11376953125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.925496578216553, "learning_rate": 1e-05, "loss": 0.4417, "mean_token_accuracy": 0.8817743062973022, "num_tokens": 657995050.0, "step": 17330 }, { "epoch": 2.2046813382521306, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.713552951812744, "learning_rate": 1e-05, "loss": 0.4891, "mean_token_accuracy": 0.8752975463867188, "num_tokens": 658032708.0, "step": 17331 }, { "epoch": 2.204808548530721, "ewc_loss": 0.11083984375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.781199932098389, "learning_rate": 1e-05, "loss": 0.4143, "mean_token_accuracy": 0.8940142393112183, "num_tokens": 658065725.0, "step": 17332 }, { "epoch": 2.2049357588093117, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 7.015376091003418, "learning_rate": 1e-05, "loss": 0.4559, "mean_token_accuracy": 0.8866667151451111, "num_tokens": 658107290.0, "step": 17333 }, { "epoch": 2.2050629690879022, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.700272083282471, "learning_rate": 1e-05, "loss": 0.4432, "mean_token_accuracy": 0.8822222948074341, "num_tokens": 658147516.0, "step": 17334 }, { "epoch": 2.2051901793664928, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00011920928955078125, "grad_norm": 7.423254013061523, "learning_rate": 1e-05, "loss": 0.4723, "mean_token_accuracy": 0.885521411895752, "num_tokens": 658188477.0, "step": 17335 }, { "epoch": 2.2053173896450833, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 5.866072654724121, "learning_rate": 1e-05, "loss": 0.4576, "mean_token_accuracy": 0.8815914988517761, "num_tokens": 658223517.0, "step": 17336 }, { "epoch": 2.205444599923674, "ewc_loss": 0.154296875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.0001277923583984375, "grad_norm": 7.71827507019043, "learning_rate": 1e-05, "loss": 0.4858, "mean_token_accuracy": 0.8840112686157227, "num_tokens": 658259167.0, "step": 17337 }, { "epoch": 2.2055718102022643, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.326904773712158, "learning_rate": 1e-05, "loss": 0.4674, "mean_token_accuracy": 0.8809425830841064, "num_tokens": 658294825.0, "step": 17338 }, { "epoch": 2.205699020480855, "ewc_loss": 0.150390625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.526472568511963, "learning_rate": 1e-05, "loss": 0.5187, "mean_token_accuracy": 0.8709725141525269, "num_tokens": 658336898.0, "step": 17339 }, { "epoch": 2.2058262307594454, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 6.626784801483154, "learning_rate": 1e-05, "loss": 0.4377, "mean_token_accuracy": 0.8931519985198975, "num_tokens": 658373201.0, "step": 17340 }, { "epoch": 2.205953441038036, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00010967254638671875, "grad_norm": 7.069793701171875, "learning_rate": 1e-05, "loss": 0.5192, "mean_token_accuracy": 0.8710647821426392, "num_tokens": 658413147.0, "step": 17341 }, { "epoch": 2.2060806513166265, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00010967254638671875, "grad_norm": 6.662213325500488, "learning_rate": 1e-05, "loss": 0.5202, "mean_token_accuracy": 0.8673141002655029, "num_tokens": 658453832.0, "step": 17342 }, { "epoch": 2.206207861595217, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.585160255432129, "learning_rate": 1e-05, "loss": 0.5011, "mean_token_accuracy": 0.870801568031311, "num_tokens": 658498302.0, "step": 17343 }, { "epoch": 2.2063350718738075, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.686270236968994, "learning_rate": 1e-05, "loss": 0.4757, "mean_token_accuracy": 0.8823479413986206, "num_tokens": 658535017.0, "step": 17344 }, { "epoch": 2.206462282152398, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.195028305053711, "learning_rate": 1e-05, "loss": 0.5167, "mean_token_accuracy": 0.860824465751648, "num_tokens": 658575495.0, "step": 17345 }, { "epoch": 2.2065894924309886, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.772330284118652, "learning_rate": 1e-05, "loss": 0.455, "mean_token_accuracy": 0.8872390985488892, "num_tokens": 658619493.0, "step": 17346 }, { "epoch": 2.206716702709579, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.954566955566406, "learning_rate": 1e-05, "loss": 0.4606, "mean_token_accuracy": 0.8777865171432495, "num_tokens": 658654371.0, "step": 17347 }, { "epoch": 2.2068439129881696, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 7.037905216217041, "learning_rate": 1e-05, "loss": 0.5056, "mean_token_accuracy": 0.8705054521560669, "num_tokens": 658694543.0, "step": 17348 }, { "epoch": 2.20697112326676, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.8499908447265625, "learning_rate": 1e-05, "loss": 0.4682, "mean_token_accuracy": 0.8750847578048706, "num_tokens": 658724859.0, "step": 17349 }, { "epoch": 2.2070983335453507, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00011491775512695312, "grad_norm": 7.241228103637695, "learning_rate": 1e-05, "loss": 0.4535, "mean_token_accuracy": 0.8915262222290039, "num_tokens": 658759941.0, "step": 17350 }, { "epoch": 2.2072255438239408, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 5.897928714752197, "learning_rate": 1e-05, "loss": 0.4367, "mean_token_accuracy": 0.8860844969749451, "num_tokens": 658801254.0, "step": 17351 }, { "epoch": 2.2073527541025313, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.0001201629638671875, "grad_norm": 7.4509782791137695, "learning_rate": 1e-05, "loss": 0.5041, "mean_token_accuracy": 0.8754570484161377, "num_tokens": 658835177.0, "step": 17352 }, { "epoch": 2.207479964381122, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.1296539306640625, "learning_rate": 1e-05, "loss": 0.4589, "mean_token_accuracy": 0.8842130899429321, "num_tokens": 658874636.0, "step": 17353 }, { "epoch": 2.2076071746597123, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.0001201629638671875, "grad_norm": 7.440708637237549, "learning_rate": 1e-05, "loss": 0.5099, "mean_token_accuracy": 0.8717172145843506, "num_tokens": 658911388.0, "step": 17354 }, { "epoch": 2.207734384938303, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.354809761047363, "learning_rate": 1e-05, "loss": 0.459, "mean_token_accuracy": 0.8838688731193542, "num_tokens": 658954685.0, "step": 17355 }, { "epoch": 2.2078615952168934, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00011491775512695312, "grad_norm": 7.227182388305664, "learning_rate": 1e-05, "loss": 0.4756, "mean_token_accuracy": 0.8811652064323425, "num_tokens": 658992573.0, "step": 17356 }, { "epoch": 2.207988805495484, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.44667387008667, "learning_rate": 1e-05, "loss": 0.4987, "mean_token_accuracy": 0.8708169460296631, "num_tokens": 659024460.0, "step": 17357 }, { "epoch": 2.2081160157740745, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00010967254638671875, "grad_norm": 6.935807704925537, "learning_rate": 1e-05, "loss": 0.547, "mean_token_accuracy": 0.8597184419631958, "num_tokens": 659066103.0, "step": 17358 }, { "epoch": 2.208243226052665, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.429037570953369, "learning_rate": 1e-05, "loss": 0.4498, "mean_token_accuracy": 0.8870928287506104, "num_tokens": 659097293.0, "step": 17359 }, { "epoch": 2.2083704363312555, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.6656060218811035, "learning_rate": 1e-05, "loss": 0.4913, "mean_token_accuracy": 0.8727997541427612, "num_tokens": 659141221.0, "step": 17360 }, { "epoch": 2.208497646609846, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.362049579620361, "learning_rate": 1e-05, "loss": 0.4545, "mean_token_accuracy": 0.8809159398078918, "num_tokens": 659181026.0, "step": 17361 }, { "epoch": 2.2086248568884366, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.493480682373047, "learning_rate": 1e-05, "loss": 0.4769, "mean_token_accuracy": 0.8742138147354126, "num_tokens": 659218503.0, "step": 17362 }, { "epoch": 2.208752067167027, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.273953914642334, "learning_rate": 1e-05, "loss": 0.4538, "mean_token_accuracy": 0.8840542435646057, "num_tokens": 659255940.0, "step": 17363 }, { "epoch": 2.2088792774456176, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.4192681312561035, "learning_rate": 1e-05, "loss": 0.5402, "mean_token_accuracy": 0.8565579652786255, "num_tokens": 659296994.0, "step": 17364 }, { "epoch": 2.209006487724208, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.163737773895264, "learning_rate": 1e-05, "loss": 0.4906, "mean_token_accuracy": 0.8686182498931885, "num_tokens": 659335372.0, "step": 17365 }, { "epoch": 2.2091336980027987, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.43710470199585, "learning_rate": 1e-05, "loss": 0.4764, "mean_token_accuracy": 0.8747696280479431, "num_tokens": 659375954.0, "step": 17366 }, { "epoch": 2.209260908281389, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.047248363494873, "learning_rate": 1e-05, "loss": 0.4672, "mean_token_accuracy": 0.8773971796035767, "num_tokens": 659409995.0, "step": 17367 }, { "epoch": 2.2093881185599797, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.532804489135742, "learning_rate": 1e-05, "loss": 0.4496, "mean_token_accuracy": 0.8856685161590576, "num_tokens": 659445640.0, "step": 17368 }, { "epoch": 2.2095153288385703, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.898487567901611, "learning_rate": 1e-05, "loss": 0.4698, "mean_token_accuracy": 0.8765542507171631, "num_tokens": 659488609.0, "step": 17369 }, { "epoch": 2.209642539117161, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.733789920806885, "learning_rate": 1e-05, "loss": 0.4372, "mean_token_accuracy": 0.8896869421005249, "num_tokens": 659533077.0, "step": 17370 }, { "epoch": 2.2097697493957513, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.741962909698486, "learning_rate": 1e-05, "loss": 0.4599, "mean_token_accuracy": 0.8788232803344727, "num_tokens": 659574080.0, "step": 17371 }, { "epoch": 2.209896959674342, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00011157989501953125, "grad_norm": 7.096583366394043, "learning_rate": 1e-05, "loss": 0.4919, "mean_token_accuracy": 0.8776848912239075, "num_tokens": 659610508.0, "step": 17372 }, { "epoch": 2.2100241699529324, "ewc_loss": 0.11181640625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.682413101196289, "learning_rate": 1e-05, "loss": 0.4636, "mean_token_accuracy": 0.8759424686431885, "num_tokens": 659648378.0, "step": 17373 }, { "epoch": 2.210151380231523, "ewc_loss": 0.1484375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.0001220703125, "grad_norm": 7.54857063293457, "learning_rate": 1e-05, "loss": 0.5169, "mean_token_accuracy": 0.8703598976135254, "num_tokens": 659681639.0, "step": 17374 }, { "epoch": 2.2102785905101134, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 5.935299873352051, "learning_rate": 1e-05, "loss": 0.4927, "mean_token_accuracy": 0.8690397143363953, "num_tokens": 659721579.0, "step": 17375 }, { "epoch": 2.2104058007887035, "ewc_loss": 0.15625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00012874603271484375, "grad_norm": 7.78864860534668, "learning_rate": 1e-05, "loss": 0.5171, "mean_token_accuracy": 0.8750409483909607, "num_tokens": 659761577.0, "step": 17376 }, { "epoch": 2.210533011067294, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 6.448643684387207, "learning_rate": 1e-05, "loss": 0.486, "mean_token_accuracy": 0.8778573274612427, "num_tokens": 659799393.0, "step": 17377 }, { "epoch": 2.2106602213458846, "ewc_loss": 0.1484375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00012111663818359375, "grad_norm": 7.507984161376953, "learning_rate": 1e-05, "loss": 0.5562, "mean_token_accuracy": 0.8591363430023193, "num_tokens": 659839118.0, "step": 17378 }, { "epoch": 2.210787431624475, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.0001125335693359375, "grad_norm": 6.704622745513916, "learning_rate": 1e-05, "loss": 0.4539, "mean_token_accuracy": 0.889509916305542, "num_tokens": 659874955.0, "step": 17379 }, { "epoch": 2.2109146419030656, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 7.003413677215576, "learning_rate": 1e-05, "loss": 0.5257, "mean_token_accuracy": 0.8691954612731934, "num_tokens": 659908590.0, "step": 17380 }, { "epoch": 2.211041852181656, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 6.6953511238098145, "learning_rate": 1e-05, "loss": 0.4939, "mean_token_accuracy": 0.8751042485237122, "num_tokens": 659945753.0, "step": 17381 }, { "epoch": 2.2111690624602467, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.53851842880249, "learning_rate": 1e-05, "loss": 0.445, "mean_token_accuracy": 0.8871402740478516, "num_tokens": 659981284.0, "step": 17382 }, { "epoch": 2.211296272738837, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.679556369781494, "learning_rate": 1e-05, "loss": 0.4855, "mean_token_accuracy": 0.8751517534255981, "num_tokens": 660016160.0, "step": 17383 }, { "epoch": 2.2114234830174277, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.207062721252441, "learning_rate": 1e-05, "loss": 0.4535, "mean_token_accuracy": 0.8824611902236938, "num_tokens": 660053755.0, "step": 17384 }, { "epoch": 2.2115506932960183, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.718008041381836, "learning_rate": 1e-05, "loss": 0.4119, "mean_token_accuracy": 0.8995876312255859, "num_tokens": 660092808.0, "step": 17385 }, { "epoch": 2.211677903574609, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 6.0177483558654785, "learning_rate": 1e-05, "loss": 0.4819, "mean_token_accuracy": 0.8695936799049377, "num_tokens": 660128996.0, "step": 17386 }, { "epoch": 2.2118051138531993, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.8299479484558105, "learning_rate": 1e-05, "loss": 0.4267, "mean_token_accuracy": 0.8956990242004395, "num_tokens": 660163345.0, "step": 17387 }, { "epoch": 2.21193232413179, "ewc_loss": 0.11376953125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.898890495300293, "learning_rate": 1e-05, "loss": 0.4536, "mean_token_accuracy": 0.8819190263748169, "num_tokens": 660196483.0, "step": 17388 }, { "epoch": 2.2120595344103804, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 7.019879341125488, "learning_rate": 1e-05, "loss": 0.496, "mean_token_accuracy": 0.8737226724624634, "num_tokens": 660226039.0, "step": 17389 }, { "epoch": 2.212186744688971, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.841558456420898, "learning_rate": 1e-05, "loss": 0.4626, "mean_token_accuracy": 0.8765429854393005, "num_tokens": 660274301.0, "step": 17390 }, { "epoch": 2.2123139549675614, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00011587142944335938, "grad_norm": 7.24364709854126, "learning_rate": 1e-05, "loss": 0.5252, "mean_token_accuracy": 0.8693221807479858, "num_tokens": 660312893.0, "step": 17391 }, { "epoch": 2.212441165246152, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 5.912787914276123, "learning_rate": 1e-05, "loss": 0.4444, "mean_token_accuracy": 0.8852258920669556, "num_tokens": 660349396.0, "step": 17392 }, { "epoch": 2.2125683755247425, "ewc_loss": 0.1455078125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00011873245239257812, "grad_norm": 7.416679859161377, "learning_rate": 1e-05, "loss": 0.4754, "mean_token_accuracy": 0.8825975656509399, "num_tokens": 660385902.0, "step": 17393 }, { "epoch": 2.212695585803333, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.119295120239258, "learning_rate": 1e-05, "loss": 0.4871, "mean_token_accuracy": 0.8734424710273743, "num_tokens": 660421527.0, "step": 17394 }, { "epoch": 2.2128227960819236, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00011920928955078125, "grad_norm": 7.411969184875488, "learning_rate": 1e-05, "loss": 0.4787, "mean_token_accuracy": 0.881367564201355, "num_tokens": 660459687.0, "step": 17395 }, { "epoch": 2.212950006360514, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.322732925415039, "learning_rate": 1e-05, "loss": 0.5313, "mean_token_accuracy": 0.8617191314697266, "num_tokens": 660502415.0, "step": 17396 }, { "epoch": 2.2130772166391046, "ewc_loss": 0.140625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00011348724365234375, "grad_norm": 7.195805072784424, "learning_rate": 1e-05, "loss": 0.4902, "mean_token_accuracy": 0.8799158930778503, "num_tokens": 660542503.0, "step": 17397 }, { "epoch": 2.213204426917695, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.443988800048828, "learning_rate": 1e-05, "loss": 0.4994, "mean_token_accuracy": 0.8700770139694214, "num_tokens": 660576482.0, "step": 17398 }, { "epoch": 2.2133316371962852, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.85775089263916, "learning_rate": 1e-05, "loss": 0.4268, "mean_token_accuracy": 0.8961315155029297, "num_tokens": 660615151.0, "step": 17399 }, { "epoch": 2.213458847474876, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.457965850830078, "learning_rate": 1e-05, "loss": 0.4899, "mean_token_accuracy": 0.8746871948242188, "num_tokens": 660656555.0, "step": 17400 }, { "epoch": 2.2135860577534663, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.556569576263428, "learning_rate": 1e-05, "loss": 0.5117, "mean_token_accuracy": 0.8653953075408936, "num_tokens": 660695991.0, "step": 17401 }, { "epoch": 2.213713268032057, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.435568332672119, "learning_rate": 1e-05, "loss": 0.4729, "mean_token_accuracy": 0.8781336545944214, "num_tokens": 660727840.0, "step": 17402 }, { "epoch": 2.2138404783106473, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.325640678405762, "learning_rate": 1e-05, "loss": 0.4567, "mean_token_accuracy": 0.8825478553771973, "num_tokens": 660770107.0, "step": 17403 }, { "epoch": 2.213967688589238, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.415402412414551, "learning_rate": 1e-05, "loss": 0.5296, "mean_token_accuracy": 0.8612873554229736, "num_tokens": 660809813.0, "step": 17404 }, { "epoch": 2.2140948988678284, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.175564289093018, "learning_rate": 1e-05, "loss": 0.5069, "mean_token_accuracy": 0.8661794662475586, "num_tokens": 660849427.0, "step": 17405 }, { "epoch": 2.214222109146419, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.386734962463379, "learning_rate": 1e-05, "loss": 0.4399, "mean_token_accuracy": 0.8861266374588013, "num_tokens": 660888456.0, "step": 17406 }, { "epoch": 2.2143493194250095, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.084907531738281, "learning_rate": 1e-05, "loss": 0.4344, "mean_token_accuracy": 0.8874337673187256, "num_tokens": 660931889.0, "step": 17407 }, { "epoch": 2.2144765297036, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.411159038543701, "learning_rate": 1e-05, "loss": 0.4574, "mean_token_accuracy": 0.8820106387138367, "num_tokens": 660972965.0, "step": 17408 }, { "epoch": 2.2146037399821905, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.016383171081543, "learning_rate": 1e-05, "loss": 0.4764, "mean_token_accuracy": 0.8760946393013, "num_tokens": 661007352.0, "step": 17409 }, { "epoch": 2.214730950260781, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.46475887298584, "learning_rate": 1e-05, "loss": 0.4594, "mean_token_accuracy": 0.8821107149124146, "num_tokens": 661044202.0, "step": 17410 }, { "epoch": 2.2148581605393716, "ewc_loss": 0.11474609375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 5.911132335662842, "learning_rate": 1e-05, "loss": 0.4692, "mean_token_accuracy": 0.876714289188385, "num_tokens": 661088542.0, "step": 17411 }, { "epoch": 2.214985370817962, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.608969688415527, "learning_rate": 1e-05, "loss": 0.5071, "mean_token_accuracy": 0.8698834180831909, "num_tokens": 661130060.0, "step": 17412 }, { "epoch": 2.2151125810965526, "ewc_loss": 0.11083984375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.762534141540527, "learning_rate": 1e-05, "loss": 0.4882, "mean_token_accuracy": 0.8696394562721252, "num_tokens": 661170325.0, "step": 17413 }, { "epoch": 2.215239791375143, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.889741897583008, "learning_rate": 1e-05, "loss": 0.4994, "mean_token_accuracy": 0.8722569942474365, "num_tokens": 661216151.0, "step": 17414 }, { "epoch": 2.2153670016537337, "ewc_loss": 0.109375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.595670223236084, "learning_rate": 1e-05, "loss": 0.4488, "mean_token_accuracy": 0.8825686573982239, "num_tokens": 661257857.0, "step": 17415 }, { "epoch": 2.215494211932324, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00011920928955078125, "grad_norm": 7.404813289642334, "learning_rate": 1e-05, "loss": 0.4938, "mean_token_accuracy": 0.8789210319519043, "num_tokens": 661294113.0, "step": 17416 }, { "epoch": 2.2156214222109147, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.6262030601501465, "learning_rate": 1e-05, "loss": 0.4714, "mean_token_accuracy": 0.873859167098999, "num_tokens": 661332681.0, "step": 17417 }, { "epoch": 2.2157486324895053, "ewc_loss": 0.162109375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00013446807861328125, "grad_norm": 7.9998602867126465, "learning_rate": 1e-05, "loss": 0.5323, "mean_token_accuracy": 0.8720759749412537, "num_tokens": 661373567.0, "step": 17418 }, { "epoch": 2.215875842768096, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.204580783843994, "learning_rate": 1e-05, "loss": 0.5219, "mean_token_accuracy": 0.8698516488075256, "num_tokens": 661415727.0, "step": 17419 }, { "epoch": 2.2160030530466863, "ewc_loss": 0.162109375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00013446807861328125, "grad_norm": 8.122445106506348, "learning_rate": 1e-05, "loss": 0.4938, "mean_token_accuracy": 0.8827788829803467, "num_tokens": 661454934.0, "step": 17420 }, { "epoch": 2.216130263325277, "ewc_loss": 0.1484375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00012159347534179688, "grad_norm": 6.889753341674805, "learning_rate": 1e-05, "loss": 0.5594, "mean_token_accuracy": 0.8607617020606995, "num_tokens": 661499070.0, "step": 17421 }, { "epoch": 2.2162574736038674, "ewc_loss": 0.150390625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.624373435974121, "learning_rate": 1e-05, "loss": 0.4943, "mean_token_accuracy": 0.8812088370323181, "num_tokens": 661538595.0, "step": 17422 }, { "epoch": 2.216384683882458, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.000118255615234375, "grad_norm": 6.933110237121582, "learning_rate": 1e-05, "loss": 0.5175, "mean_token_accuracy": 0.8687443733215332, "num_tokens": 661578472.0, "step": 17423 }, { "epoch": 2.216511894161048, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00011157989501953125, "grad_norm": 7.160774230957031, "learning_rate": 1e-05, "loss": 0.5023, "mean_token_accuracy": 0.8718761205673218, "num_tokens": 661615127.0, "step": 17424 }, { "epoch": 2.2166391044396385, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 6.726082801818848, "learning_rate": 1e-05, "loss": 0.5362, "mean_token_accuracy": 0.8612961173057556, "num_tokens": 661655514.0, "step": 17425 }, { "epoch": 2.216766314718229, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.839170455932617, "learning_rate": 1e-05, "loss": 0.5151, "mean_token_accuracy": 0.8664022088050842, "num_tokens": 661690838.0, "step": 17426 }, { "epoch": 2.2168935249968196, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.533219814300537, "learning_rate": 1e-05, "loss": 0.4773, "mean_token_accuracy": 0.8771861791610718, "num_tokens": 661735697.0, "step": 17427 }, { "epoch": 2.21702073527541, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.630215644836426, "learning_rate": 1e-05, "loss": 0.5436, "mean_token_accuracy": 0.8613631725311279, "num_tokens": 661766875.0, "step": 17428 }, { "epoch": 2.2171479455540006, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.3675713539123535, "learning_rate": 1e-05, "loss": 0.4733, "mean_token_accuracy": 0.8772237300872803, "num_tokens": 661798059.0, "step": 17429 }, { "epoch": 2.217275155832591, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.526393890380859, "learning_rate": 1e-05, "loss": 0.4453, "mean_token_accuracy": 0.8855249881744385, "num_tokens": 661833128.0, "step": 17430 }, { "epoch": 2.2174023661111817, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.198011875152588, "learning_rate": 1e-05, "loss": 0.4963, "mean_token_accuracy": 0.8671646118164062, "num_tokens": 661872672.0, "step": 17431 }, { "epoch": 2.217529576389772, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.525366306304932, "learning_rate": 1e-05, "loss": 0.5026, "mean_token_accuracy": 0.8681254386901855, "num_tokens": 661919611.0, "step": 17432 }, { "epoch": 2.2176567866683627, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.046211242675781, "learning_rate": 1e-05, "loss": 0.5329, "mean_token_accuracy": 0.8590952157974243, "num_tokens": 661955349.0, "step": 17433 }, { "epoch": 2.2177839969469533, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.661730766296387, "learning_rate": 1e-05, "loss": 0.4535, "mean_token_accuracy": 0.8839434385299683, "num_tokens": 661990178.0, "step": 17434 }, { "epoch": 2.217911207225544, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.861817359924316, "learning_rate": 1e-05, "loss": 0.4514, "mean_token_accuracy": 0.881425142288208, "num_tokens": 662035855.0, "step": 17435 }, { "epoch": 2.2180384175041343, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.923536777496338, "learning_rate": 1e-05, "loss": 0.5011, "mean_token_accuracy": 0.8675436973571777, "num_tokens": 662073959.0, "step": 17436 }, { "epoch": 2.218165627782725, "ewc_loss": 0.10986328125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.7022013664245605, "learning_rate": 1e-05, "loss": 0.4023, "mean_token_accuracy": 0.8953839540481567, "num_tokens": 662108598.0, "step": 17437 }, { "epoch": 2.2182928380613154, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00011777877807617188, "grad_norm": 7.391895771026611, "learning_rate": 1e-05, "loss": 0.4987, "mean_token_accuracy": 0.8775327205657959, "num_tokens": 662136316.0, "step": 17438 }, { "epoch": 2.218420048339906, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 5.778205871582031, "learning_rate": 1e-05, "loss": 0.4614, "mean_token_accuracy": 0.8787549734115601, "num_tokens": 662175771.0, "step": 17439 }, { "epoch": 2.2185472586184964, "ewc_loss": 0.158203125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00013065338134765625, "grad_norm": 7.8473615646362305, "learning_rate": 1e-05, "loss": 0.5062, "mean_token_accuracy": 0.8743720054626465, "num_tokens": 662219031.0, "step": 17440 }, { "epoch": 2.218674468897087, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.289019584655762, "learning_rate": 1e-05, "loss": 0.4571, "mean_token_accuracy": 0.8838644027709961, "num_tokens": 662257734.0, "step": 17441 }, { "epoch": 2.2188016791756775, "ewc_loss": 0.15625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00012874603271484375, "grad_norm": 7.793851375579834, "learning_rate": 1e-05, "loss": 0.5058, "mean_token_accuracy": 0.8778284788131714, "num_tokens": 662297329.0, "step": 17442 }, { "epoch": 2.218928889454268, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00011491775512695312, "grad_norm": 6.788131237030029, "learning_rate": 1e-05, "loss": 0.4969, "mean_token_accuracy": 0.8730581998825073, "num_tokens": 662334597.0, "step": 17443 }, { "epoch": 2.2190560997328586, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00011491775512695312, "grad_norm": 7.264615058898926, "learning_rate": 1e-05, "loss": 0.5009, "mean_token_accuracy": 0.8716902732849121, "num_tokens": 662370369.0, "step": 17444 }, { "epoch": 2.219183310011449, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00011205673217773438, "grad_norm": 6.827253818511963, "learning_rate": 1e-05, "loss": 0.5046, "mean_token_accuracy": 0.8705542087554932, "num_tokens": 662406193.0, "step": 17445 }, { "epoch": 2.2193105202900396, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.740198135375977, "learning_rate": 1e-05, "loss": 0.4393, "mean_token_accuracy": 0.8909215927124023, "num_tokens": 662439692.0, "step": 17446 }, { "epoch": 2.21943773056863, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.7573347091674805, "learning_rate": 1e-05, "loss": 0.4665, "mean_token_accuracy": 0.881100058555603, "num_tokens": 662479768.0, "step": 17447 }, { "epoch": 2.2195649408472207, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.320242404937744, "learning_rate": 1e-05, "loss": 0.4547, "mean_token_accuracy": 0.8810033202171326, "num_tokens": 662523591.0, "step": 17448 }, { "epoch": 2.2196921511258108, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.767327785491943, "learning_rate": 1e-05, "loss": 0.4794, "mean_token_accuracy": 0.8832120299339294, "num_tokens": 662557693.0, "step": 17449 }, { "epoch": 2.2198193614044013, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.062899112701416, "learning_rate": 1e-05, "loss": 0.4686, "mean_token_accuracy": 0.8760484457015991, "num_tokens": 662595404.0, "step": 17450 }, { "epoch": 2.219946571682992, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.877367973327637, "learning_rate": 1e-05, "loss": 0.4814, "mean_token_accuracy": 0.8771843314170837, "num_tokens": 662637033.0, "step": 17451 }, { "epoch": 2.2200737819615823, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 5.916572570800781, "learning_rate": 1e-05, "loss": 0.4757, "mean_token_accuracy": 0.8754101991653442, "num_tokens": 662675199.0, "step": 17452 }, { "epoch": 2.220200992240173, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 7.063933849334717, "learning_rate": 1e-05, "loss": 0.4939, "mean_token_accuracy": 0.8773077130317688, "num_tokens": 662720809.0, "step": 17453 }, { "epoch": 2.2203282025187634, "ewc_loss": 0.11474609375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 5.870244979858398, "learning_rate": 1e-05, "loss": 0.4252, "mean_token_accuracy": 0.8904236555099487, "num_tokens": 662759894.0, "step": 17454 }, { "epoch": 2.220455412797354, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00011587142944335938, "grad_norm": 7.310620307922363, "learning_rate": 1e-05, "loss": 0.4626, "mean_token_accuracy": 0.8854087591171265, "num_tokens": 662795373.0, "step": 17455 }, { "epoch": 2.2205826230759445, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 5.965994358062744, "learning_rate": 1e-05, "loss": 0.4444, "mean_token_accuracy": 0.8852189183235168, "num_tokens": 662832764.0, "step": 17456 }, { "epoch": 2.220709833354535, "ewc_loss": 0.1455078125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00011873245239257812, "grad_norm": 7.4695353507995605, "learning_rate": 1e-05, "loss": 0.5563, "mean_token_accuracy": 0.8591134548187256, "num_tokens": 662872796.0, "step": 17457 }, { "epoch": 2.2208370436331255, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.202212810516357, "learning_rate": 1e-05, "loss": 0.5244, "mean_token_accuracy": 0.8675929307937622, "num_tokens": 662914027.0, "step": 17458 }, { "epoch": 2.220964253911716, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.000118255615234375, "grad_norm": 7.409992694854736, "learning_rate": 1e-05, "loss": 0.5255, "mean_token_accuracy": 0.8666998744010925, "num_tokens": 662953924.0, "step": 17459 }, { "epoch": 2.2210914641903066, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.508150577545166, "learning_rate": 1e-05, "loss": 0.5059, "mean_token_accuracy": 0.8737436532974243, "num_tokens": 662982792.0, "step": 17460 }, { "epoch": 2.221218674468897, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00010967254638671875, "grad_norm": 7.052079200744629, "learning_rate": 1e-05, "loss": 0.4786, "mean_token_accuracy": 0.8813207149505615, "num_tokens": 663023674.0, "step": 17461 }, { "epoch": 2.2213458847474876, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.460147380828857, "learning_rate": 1e-05, "loss": 0.5102, "mean_token_accuracy": 0.8688771724700928, "num_tokens": 663064187.0, "step": 17462 }, { "epoch": 2.221473095026078, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.72274923324585, "learning_rate": 1e-05, "loss": 0.4355, "mean_token_accuracy": 0.8919491767883301, "num_tokens": 663109146.0, "step": 17463 }, { "epoch": 2.2216003053046687, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.403931617736816, "learning_rate": 1e-05, "loss": 0.4595, "mean_token_accuracy": 0.8815391659736633, "num_tokens": 663147988.0, "step": 17464 }, { "epoch": 2.221727515583259, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.495026588439941, "learning_rate": 1e-05, "loss": 0.4609, "mean_token_accuracy": 0.881038248538971, "num_tokens": 663190072.0, "step": 17465 }, { "epoch": 2.2218547258618497, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.349339485168457, "learning_rate": 1e-05, "loss": 0.4682, "mean_token_accuracy": 0.8778258562088013, "num_tokens": 663222896.0, "step": 17466 }, { "epoch": 2.2219819361404403, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.371399402618408, "learning_rate": 1e-05, "loss": 0.4799, "mean_token_accuracy": 0.8761149048805237, "num_tokens": 663261846.0, "step": 17467 }, { "epoch": 2.222109146419031, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.261768341064453, "learning_rate": 1e-05, "loss": 0.4145, "mean_token_accuracy": 0.8947572112083435, "num_tokens": 663295766.0, "step": 17468 }, { "epoch": 2.2222363566976213, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.328944683074951, "learning_rate": 1e-05, "loss": 0.4256, "mean_token_accuracy": 0.8894863724708557, "num_tokens": 663333218.0, "step": 17469 }, { "epoch": 2.222363566976212, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.167415142059326, "learning_rate": 1e-05, "loss": 0.4729, "mean_token_accuracy": 0.878504753112793, "num_tokens": 663371491.0, "step": 17470 }, { "epoch": 2.2224907772548024, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.358354091644287, "learning_rate": 1e-05, "loss": 0.436, "mean_token_accuracy": 0.8902484774589539, "num_tokens": 663410824.0, "step": 17471 }, { "epoch": 2.222617987533393, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.041295051574707, "learning_rate": 1e-05, "loss": 0.4571, "mean_token_accuracy": 0.8832871913909912, "num_tokens": 663445624.0, "step": 17472 }, { "epoch": 2.2227451978119834, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.475532531738281, "learning_rate": 1e-05, "loss": 0.5161, "mean_token_accuracy": 0.8678144216537476, "num_tokens": 663483223.0, "step": 17473 }, { "epoch": 2.2228724080905735, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.890085697174072, "learning_rate": 1e-05, "loss": 0.4702, "mean_token_accuracy": 0.8773046731948853, "num_tokens": 663520577.0, "step": 17474 }, { "epoch": 2.222999618369164, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.671740531921387, "learning_rate": 1e-05, "loss": 0.4747, "mean_token_accuracy": 0.8774576187133789, "num_tokens": 663554667.0, "step": 17475 }, { "epoch": 2.2231268286477546, "ewc_loss": 0.10986328125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.718441963195801, "learning_rate": 1e-05, "loss": 0.4455, "mean_token_accuracy": 0.8830984830856323, "num_tokens": 663593475.0, "step": 17476 }, { "epoch": 2.223254038926345, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 7.034054756164551, "learning_rate": 1e-05, "loss": 0.4671, "mean_token_accuracy": 0.8846783638000488, "num_tokens": 663635957.0, "step": 17477 }, { "epoch": 2.2233812492049356, "ewc_loss": 0.10986328125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.598546504974365, "learning_rate": 1e-05, "loss": 0.4233, "mean_token_accuracy": 0.8902470469474792, "num_tokens": 663678610.0, "step": 17478 }, { "epoch": 2.223508459483526, "ewc_loss": 0.150390625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.594864845275879, "learning_rate": 1e-05, "loss": 0.493, "mean_token_accuracy": 0.8786395788192749, "num_tokens": 663718204.0, "step": 17479 }, { "epoch": 2.2236356697621167, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 5.8117804527282715, "learning_rate": 1e-05, "loss": 0.4657, "mean_token_accuracy": 0.8806060552597046, "num_tokens": 663760501.0, "step": 17480 }, { "epoch": 2.223762880040707, "ewc_loss": 0.162109375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.0001354217529296875, "grad_norm": 8.031702995300293, "learning_rate": 1e-05, "loss": 0.485, "mean_token_accuracy": 0.8867563009262085, "num_tokens": 663802515.0, "step": 17481 }, { "epoch": 2.2238900903192977, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 6.509406089782715, "learning_rate": 1e-05, "loss": 0.5029, "mean_token_accuracy": 0.8692164421081543, "num_tokens": 663839095.0, "step": 17482 }, { "epoch": 2.2240173005978883, "ewc_loss": 0.154296875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.0001277923583984375, "grad_norm": 7.817488193511963, "learning_rate": 1e-05, "loss": 0.5492, "mean_token_accuracy": 0.8649131655693054, "num_tokens": 663875734.0, "step": 17483 }, { "epoch": 2.224144510876479, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00011730194091796875, "grad_norm": 6.889926910400391, "learning_rate": 1e-05, "loss": 0.4703, "mean_token_accuracy": 0.8820990920066833, "num_tokens": 663911321.0, "step": 17484 }, { "epoch": 2.2242717211550693, "ewc_loss": 0.140625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00011444091796875, "grad_norm": 7.287410736083984, "learning_rate": 1e-05, "loss": 0.5025, "mean_token_accuracy": 0.8746000528335571, "num_tokens": 663942334.0, "step": 17485 }, { "epoch": 2.22439893143366, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.0001125335693359375, "grad_norm": 6.8031768798828125, "learning_rate": 1e-05, "loss": 0.4841, "mean_token_accuracy": 0.8753402233123779, "num_tokens": 663976306.0, "step": 17486 }, { "epoch": 2.2245261417122504, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.86185884475708, "learning_rate": 1e-05, "loss": 0.4761, "mean_token_accuracy": 0.880805492401123, "num_tokens": 664016234.0, "step": 17487 }, { "epoch": 2.224653351990841, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.626250267028809, "learning_rate": 1e-05, "loss": 0.4969, "mean_token_accuracy": 0.8726638555526733, "num_tokens": 664057012.0, "step": 17488 }, { "epoch": 2.2247805622694314, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.584305763244629, "learning_rate": 1e-05, "loss": 0.4779, "mean_token_accuracy": 0.8771633505821228, "num_tokens": 664092911.0, "step": 17489 }, { "epoch": 2.224907772548022, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.47395133972168, "learning_rate": 1e-05, "loss": 0.4998, "mean_token_accuracy": 0.8703140616416931, "num_tokens": 664133069.0, "step": 17490 }, { "epoch": 2.2250349828266125, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.425734519958496, "learning_rate": 1e-05, "loss": 0.464, "mean_token_accuracy": 0.8808699250221252, "num_tokens": 664166059.0, "step": 17491 }, { "epoch": 2.225162193105203, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.343489170074463, "learning_rate": 1e-05, "loss": 0.5067, "mean_token_accuracy": 0.870164155960083, "num_tokens": 664204918.0, "step": 17492 }, { "epoch": 2.2252894033837936, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.34696102142334, "learning_rate": 1e-05, "loss": 0.4759, "mean_token_accuracy": 0.8782908916473389, "num_tokens": 664240561.0, "step": 17493 }, { "epoch": 2.225416613662384, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.2352190017700195, "learning_rate": 1e-05, "loss": 0.4808, "mean_token_accuracy": 0.870806097984314, "num_tokens": 664274905.0, "step": 17494 }, { "epoch": 2.2255438239409746, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.344747066497803, "learning_rate": 1e-05, "loss": 0.4829, "mean_token_accuracy": 0.871925950050354, "num_tokens": 664316666.0, "step": 17495 }, { "epoch": 2.225671034219565, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 6.108712673187256, "learning_rate": 1e-05, "loss": 0.4444, "mean_token_accuracy": 0.8857377767562866, "num_tokens": 664358590.0, "step": 17496 }, { "epoch": 2.225798244498155, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.406705856323242, "learning_rate": 1e-05, "loss": 0.4744, "mean_token_accuracy": 0.8773214817047119, "num_tokens": 664394764.0, "step": 17497 }, { "epoch": 2.225925454776746, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 5.988909721374512, "learning_rate": 1e-05, "loss": 0.451, "mean_token_accuracy": 0.8818989396095276, "num_tokens": 664433730.0, "step": 17498 }, { "epoch": 2.2260526650553363, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.541388511657715, "learning_rate": 1e-05, "loss": 0.478, "mean_token_accuracy": 0.8767268061637878, "num_tokens": 664469926.0, "step": 17499 }, { "epoch": 2.226179875333927, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.797978401184082, "learning_rate": 1e-05, "loss": 0.4867, "mean_token_accuracy": 0.867586076259613, "num_tokens": 664511204.0, "step": 17500 }, { "epoch": 2.2263070856125173, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.863815784454346, "learning_rate": 1e-05, "loss": 0.5477, "mean_token_accuracy": 0.8584948778152466, "num_tokens": 664547474.0, "step": 17501 }, { "epoch": 2.226434295891108, "ewc_loss": 0.10888671875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.605642795562744, "learning_rate": 1e-05, "loss": 0.4631, "mean_token_accuracy": 0.8776683211326599, "num_tokens": 664579252.0, "step": 17502 }, { "epoch": 2.2265615061696984, "ewc_loss": 0.1455078125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00011873245239257812, "grad_norm": 7.397376537322998, "learning_rate": 1e-05, "loss": 0.521, "mean_token_accuracy": 0.8686275482177734, "num_tokens": 664612129.0, "step": 17503 }, { "epoch": 2.226688716448289, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.5962653160095215, "learning_rate": 1e-05, "loss": 0.4195, "mean_token_accuracy": 0.8920354843139648, "num_tokens": 664650847.0, "step": 17504 }, { "epoch": 2.2268159267268794, "ewc_loss": 0.1640625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00013637542724609375, "grad_norm": 8.060961723327637, "learning_rate": 1e-05, "loss": 0.5392, "mean_token_accuracy": 0.8684526085853577, "num_tokens": 664694056.0, "step": 17505 }, { "epoch": 2.22694313700547, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.247147560119629, "learning_rate": 1e-05, "loss": 0.5035, "mean_token_accuracy": 0.8700553178787231, "num_tokens": 664732868.0, "step": 17506 }, { "epoch": 2.2270703472840605, "ewc_loss": 0.1640625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00013637542724609375, "grad_norm": 8.162014961242676, "learning_rate": 1e-05, "loss": 0.488, "mean_token_accuracy": 0.8851386308670044, "num_tokens": 664772345.0, "step": 17507 }, { "epoch": 2.227197557562651, "ewc_loss": 0.150390625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 6.980184555053711, "learning_rate": 1e-05, "loss": 0.5221, "mean_token_accuracy": 0.8711010813713074, "num_tokens": 664811398.0, "step": 17508 }, { "epoch": 2.2273247678412416, "ewc_loss": 0.150390625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.615565776824951, "learning_rate": 1e-05, "loss": 0.5158, "mean_token_accuracy": 0.8725454211235046, "num_tokens": 664846773.0, "step": 17509 }, { "epoch": 2.227451978119832, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00011777877807617188, "grad_norm": 6.9812469482421875, "learning_rate": 1e-05, "loss": 0.4835, "mean_token_accuracy": 0.8816524744033813, "num_tokens": 664887027.0, "step": 17510 }, { "epoch": 2.2275791883984226, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 7.153722286224365, "learning_rate": 1e-05, "loss": 0.4454, "mean_token_accuracy": 0.892009973526001, "num_tokens": 664922798.0, "step": 17511 }, { "epoch": 2.227706398677013, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 6.740048408508301, "learning_rate": 1e-05, "loss": 0.4834, "mean_token_accuracy": 0.879863440990448, "num_tokens": 664965862.0, "step": 17512 }, { "epoch": 2.2278336089556037, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.834662437438965, "learning_rate": 1e-05, "loss": 0.5407, "mean_token_accuracy": 0.8591129779815674, "num_tokens": 665006053.0, "step": 17513 }, { "epoch": 2.227960819234194, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.543912887573242, "learning_rate": 1e-05, "loss": 0.4799, "mean_token_accuracy": 0.8764185309410095, "num_tokens": 665040003.0, "step": 17514 }, { "epoch": 2.2280880295127847, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.638794898986816, "learning_rate": 1e-05, "loss": 0.494, "mean_token_accuracy": 0.871951162815094, "num_tokens": 665079989.0, "step": 17515 }, { "epoch": 2.2282152397913753, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.345077037811279, "learning_rate": 1e-05, "loss": 0.461, "mean_token_accuracy": 0.8832471370697021, "num_tokens": 665114009.0, "step": 17516 }, { "epoch": 2.228342450069966, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.571530818939209, "learning_rate": 1e-05, "loss": 0.4735, "mean_token_accuracy": 0.8788942694664001, "num_tokens": 665153259.0, "step": 17517 }, { "epoch": 2.2284696603485563, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.14398193359375, "learning_rate": 1e-05, "loss": 0.4419, "mean_token_accuracy": 0.8874497413635254, "num_tokens": 665195024.0, "step": 17518 }, { "epoch": 2.228596870627147, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.629274845123291, "learning_rate": 1e-05, "loss": 0.4768, "mean_token_accuracy": 0.8786892294883728, "num_tokens": 665236000.0, "step": 17519 }, { "epoch": 2.2287240809057374, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.962850093841553, "learning_rate": 1e-05, "loss": 0.5221, "mean_token_accuracy": 0.8573886156082153, "num_tokens": 665277837.0, "step": 17520 }, { "epoch": 2.228851291184328, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.836243152618408, "learning_rate": 1e-05, "loss": 0.4691, "mean_token_accuracy": 0.8824267387390137, "num_tokens": 665312927.0, "step": 17521 }, { "epoch": 2.228978501462918, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.793409824371338, "learning_rate": 1e-05, "loss": 0.45, "mean_token_accuracy": 0.8810710310935974, "num_tokens": 665346962.0, "step": 17522 }, { "epoch": 2.2291057117415085, "ewc_loss": 0.140625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00011348724365234375, "grad_norm": 7.215586185455322, "learning_rate": 1e-05, "loss": 0.4861, "mean_token_accuracy": 0.8811023235321045, "num_tokens": 665385684.0, "step": 17523 }, { "epoch": 2.229232922020099, "ewc_loss": 0.11083984375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.746488094329834, "learning_rate": 1e-05, "loss": 0.3963, "mean_token_accuracy": 0.898524820804596, "num_tokens": 665421194.0, "step": 17524 }, { "epoch": 2.2293601322986896, "ewc_loss": 0.15234375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00012493133544921875, "grad_norm": 7.689213275909424, "learning_rate": 1e-05, "loss": 0.5064, "mean_token_accuracy": 0.8747138381004333, "num_tokens": 665459731.0, "step": 17525 }, { "epoch": 2.22948734257728, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.1116790771484375, "learning_rate": 1e-05, "loss": 0.4663, "mean_token_accuracy": 0.8799156546592712, "num_tokens": 665490310.0, "step": 17526 }, { "epoch": 2.2296145528558706, "ewc_loss": 0.154296875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.0001277923583984375, "grad_norm": 7.820155620574951, "learning_rate": 1e-05, "loss": 0.5391, "mean_token_accuracy": 0.8711638450622559, "num_tokens": 665527543.0, "step": 17527 }, { "epoch": 2.229741763134461, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00010967254638671875, "grad_norm": 6.649914264678955, "learning_rate": 1e-05, "loss": 0.4419, "mean_token_accuracy": 0.8866902589797974, "num_tokens": 665562555.0, "step": 17528 }, { "epoch": 2.2298689734130517, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00011730194091796875, "grad_norm": 7.392453670501709, "learning_rate": 1e-05, "loss": 0.4952, "mean_token_accuracy": 0.8791835308074951, "num_tokens": 665598165.0, "step": 17529 }, { "epoch": 2.229996183691642, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00011205673217773438, "grad_norm": 6.817700386047363, "learning_rate": 1e-05, "loss": 0.5242, "mean_token_accuracy": 0.862754762172699, "num_tokens": 665637225.0, "step": 17530 }, { "epoch": 2.2301233939702327, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.831481456756592, "learning_rate": 1e-05, "loss": 0.5169, "mean_token_accuracy": 0.8703832030296326, "num_tokens": 665676126.0, "step": 17531 }, { "epoch": 2.2302506042488233, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 6.780159950256348, "learning_rate": 1e-05, "loss": 0.4778, "mean_token_accuracy": 0.8819807767868042, "num_tokens": 665713570.0, "step": 17532 }, { "epoch": 2.230377814527414, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.368829250335693, "learning_rate": 1e-05, "loss": 0.4446, "mean_token_accuracy": 0.8858712911605835, "num_tokens": 665743648.0, "step": 17533 }, { "epoch": 2.2305050248060043, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.816529273986816, "learning_rate": 1e-05, "loss": 0.4671, "mean_token_accuracy": 0.8790761828422546, "num_tokens": 665780249.0, "step": 17534 }, { "epoch": 2.230632235084595, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 6.032495498657227, "learning_rate": 1e-05, "loss": 0.4253, "mean_token_accuracy": 0.8922098875045776, "num_tokens": 665821175.0, "step": 17535 }, { "epoch": 2.2307594453631854, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.981403827667236, "learning_rate": 1e-05, "loss": 0.4404, "mean_token_accuracy": 0.8892802596092224, "num_tokens": 665854598.0, "step": 17536 }, { "epoch": 2.230886655641776, "ewc_loss": 0.11181640625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.8553667068481445, "learning_rate": 1e-05, "loss": 0.434, "mean_token_accuracy": 0.8883837461471558, "num_tokens": 665892294.0, "step": 17537 }, { "epoch": 2.2310138659203664, "ewc_loss": 0.140625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00011396408081054688, "grad_norm": 7.2708516120910645, "learning_rate": 1e-05, "loss": 0.5034, "mean_token_accuracy": 0.8696060180664062, "num_tokens": 665928575.0, "step": 17538 }, { "epoch": 2.231141076198957, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 5.881997585296631, "learning_rate": 1e-05, "loss": 0.4295, "mean_token_accuracy": 0.891103982925415, "num_tokens": 665963136.0, "step": 17539 }, { "epoch": 2.2312682864775475, "ewc_loss": 0.1484375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.0001220703125, "grad_norm": 7.534127235412598, "learning_rate": 1e-05, "loss": 0.4955, "mean_token_accuracy": 0.8789218664169312, "num_tokens": 666008556.0, "step": 17540 }, { "epoch": 2.231395496756138, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.148936748504639, "learning_rate": 1e-05, "loss": 0.4724, "mean_token_accuracy": 0.8782632350921631, "num_tokens": 666047066.0, "step": 17541 }, { "epoch": 2.2315227070347285, "ewc_loss": 0.1484375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00012159347534179688, "grad_norm": 7.564803600311279, "learning_rate": 1e-05, "loss": 0.4812, "mean_token_accuracy": 0.8813540935516357, "num_tokens": 666084998.0, "step": 17542 }, { "epoch": 2.231649917313319, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.441338539123535, "learning_rate": 1e-05, "loss": 0.4903, "mean_token_accuracy": 0.8737543225288391, "num_tokens": 666124935.0, "step": 17543 }, { "epoch": 2.2317771275919096, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.0001163482666015625, "grad_norm": 7.3168559074401855, "learning_rate": 1e-05, "loss": 0.5388, "mean_token_accuracy": 0.8621633052825928, "num_tokens": 666165580.0, "step": 17544 }, { "epoch": 2.2319043378705, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.54349422454834, "learning_rate": 1e-05, "loss": 0.5043, "mean_token_accuracy": 0.8692464828491211, "num_tokens": 666206020.0, "step": 17545 }, { "epoch": 2.2320315481490907, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.9823317527771, "learning_rate": 1e-05, "loss": 0.523, "mean_token_accuracy": 0.8665727376937866, "num_tokens": 666239757.0, "step": 17546 }, { "epoch": 2.2321587584276807, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.477971076965332, "learning_rate": 1e-05, "loss": 0.4327, "mean_token_accuracy": 0.8932526707649231, "num_tokens": 666278426.0, "step": 17547 }, { "epoch": 2.2322859687062713, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.717780113220215, "learning_rate": 1e-05, "loss": 0.4603, "mean_token_accuracy": 0.8829088807106018, "num_tokens": 666315784.0, "step": 17548 }, { "epoch": 2.232413178984862, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.397007942199707, "learning_rate": 1e-05, "loss": 0.4984, "mean_token_accuracy": 0.8686530590057373, "num_tokens": 666353851.0, "step": 17549 }, { "epoch": 2.2325403892634523, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.541001796722412, "learning_rate": 1e-05, "loss": 0.4948, "mean_token_accuracy": 0.871788740158081, "num_tokens": 666391208.0, "step": 17550 }, { "epoch": 2.232667599542043, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.272435188293457, "learning_rate": 1e-05, "loss": 0.4729, "mean_token_accuracy": 0.8758676052093506, "num_tokens": 666434477.0, "step": 17551 }, { "epoch": 2.2327948098206334, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.4941229820251465, "learning_rate": 1e-05, "loss": 0.4906, "mean_token_accuracy": 0.8746892213821411, "num_tokens": 666472481.0, "step": 17552 }, { "epoch": 2.232922020099224, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.145367622375488, "learning_rate": 1e-05, "loss": 0.4684, "mean_token_accuracy": 0.876736044883728, "num_tokens": 666502858.0, "step": 17553 }, { "epoch": 2.2330492303778144, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.547273635864258, "learning_rate": 1e-05, "loss": 0.5312, "mean_token_accuracy": 0.8610789775848389, "num_tokens": 666547311.0, "step": 17554 }, { "epoch": 2.233176440656405, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 5.982821941375732, "learning_rate": 1e-05, "loss": 0.4411, "mean_token_accuracy": 0.8873322606086731, "num_tokens": 666587818.0, "step": 17555 }, { "epoch": 2.2333036509349955, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.700064182281494, "learning_rate": 1e-05, "loss": 0.4675, "mean_token_accuracy": 0.8785949945449829, "num_tokens": 666626071.0, "step": 17556 }, { "epoch": 2.233430861213586, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.83919620513916, "learning_rate": 1e-05, "loss": 0.4118, "mean_token_accuracy": 0.8926236629486084, "num_tokens": 666659221.0, "step": 17557 }, { "epoch": 2.2335580714921766, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.979183197021484, "learning_rate": 1e-05, "loss": 0.4974, "mean_token_accuracy": 0.8719379901885986, "num_tokens": 666696189.0, "step": 17558 }, { "epoch": 2.233685281770767, "ewc_loss": 0.11083984375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.751065731048584, "learning_rate": 1e-05, "loss": 0.5021, "mean_token_accuracy": 0.8675699234008789, "num_tokens": 666738495.0, "step": 17559 }, { "epoch": 2.2338124920493576, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.0001163482666015625, "grad_norm": 7.349654197692871, "learning_rate": 1e-05, "loss": 0.4727, "mean_token_accuracy": 0.8833174705505371, "num_tokens": 666772848.0, "step": 17560 }, { "epoch": 2.233939702327948, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 5.867401599884033, "learning_rate": 1e-05, "loss": 0.4512, "mean_token_accuracy": 0.8803019523620605, "num_tokens": 666813849.0, "step": 17561 }, { "epoch": 2.2340669126065387, "ewc_loss": 0.150390625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.62937593460083, "learning_rate": 1e-05, "loss": 0.5106, "mean_token_accuracy": 0.8779703378677368, "num_tokens": 666848564.0, "step": 17562 }, { "epoch": 2.234194122885129, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.30039644241333, "learning_rate": 1e-05, "loss": 0.4795, "mean_token_accuracy": 0.8785718083381653, "num_tokens": 666877381.0, "step": 17563 }, { "epoch": 2.2343213331637197, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00011968612670898438, "grad_norm": 7.493033409118652, "learning_rate": 1e-05, "loss": 0.5323, "mean_token_accuracy": 0.8669782876968384, "num_tokens": 666920452.0, "step": 17564 }, { "epoch": 2.2344485434423103, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 6.605721473693848, "learning_rate": 1e-05, "loss": 0.4626, "mean_token_accuracy": 0.8835302591323853, "num_tokens": 666958799.0, "step": 17565 }, { "epoch": 2.234575753720901, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 6.98558235168457, "learning_rate": 1e-05, "loss": 0.4762, "mean_token_accuracy": 0.8811204433441162, "num_tokens": 666993899.0, "step": 17566 }, { "epoch": 2.2347029639994913, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.667471408843994, "learning_rate": 1e-05, "loss": 0.4974, "mean_token_accuracy": 0.8738141059875488, "num_tokens": 667035154.0, "step": 17567 }, { "epoch": 2.234830174278082, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.467735290527344, "learning_rate": 1e-05, "loss": 0.4967, "mean_token_accuracy": 0.8711873888969421, "num_tokens": 667077047.0, "step": 17568 }, { "epoch": 2.2349573845566724, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.732007026672363, "learning_rate": 1e-05, "loss": 0.4634, "mean_token_accuracy": 0.8795482516288757, "num_tokens": 667114298.0, "step": 17569 }, { "epoch": 2.235084594835263, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 6.0527729988098145, "learning_rate": 1e-05, "loss": 0.4572, "mean_token_accuracy": 0.8807997703552246, "num_tokens": 667157942.0, "step": 17570 }, { "epoch": 2.2352118051138534, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 6.959650039672852, "learning_rate": 1e-05, "loss": 0.5233, "mean_token_accuracy": 0.8680065870285034, "num_tokens": 667195547.0, "step": 17571 }, { "epoch": 2.2353390153924435, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.833981037139893, "learning_rate": 1e-05, "loss": 0.4208, "mean_token_accuracy": 0.890938401222229, "num_tokens": 667232484.0, "step": 17572 }, { "epoch": 2.235466225671034, "ewc_loss": 0.140625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00011348724365234375, "grad_norm": 7.277538776397705, "learning_rate": 1e-05, "loss": 0.516, "mean_token_accuracy": 0.8677997589111328, "num_tokens": 667267661.0, "step": 17573 }, { "epoch": 2.2355934359496246, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 5.869631767272949, "learning_rate": 1e-05, "loss": 0.4246, "mean_token_accuracy": 0.8912457227706909, "num_tokens": 667303963.0, "step": 17574 }, { "epoch": 2.235720646228215, "ewc_loss": 0.1484375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00012111663818359375, "grad_norm": 7.552509784698486, "learning_rate": 1e-05, "loss": 0.5013, "mean_token_accuracy": 0.8755441904067993, "num_tokens": 667344103.0, "step": 17575 }, { "epoch": 2.2358478565068056, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.1573967933654785, "learning_rate": 1e-05, "loss": 0.4737, "mean_token_accuracy": 0.8787389993667603, "num_tokens": 667377782.0, "step": 17576 }, { "epoch": 2.235975066785396, "ewc_loss": 0.1484375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.0001220703125, "grad_norm": 7.544110298156738, "learning_rate": 1e-05, "loss": 0.5015, "mean_token_accuracy": 0.8790510892868042, "num_tokens": 667419015.0, "step": 17577 }, { "epoch": 2.2361022770639867, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.440296649932861, "learning_rate": 1e-05, "loss": 0.4751, "mean_token_accuracy": 0.8794545531272888, "num_tokens": 667457465.0, "step": 17578 }, { "epoch": 2.236229487342577, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00011491775512695312, "grad_norm": 7.27384614944458, "learning_rate": 1e-05, "loss": 0.5198, "mean_token_accuracy": 0.868969738483429, "num_tokens": 667497671.0, "step": 17579 }, { "epoch": 2.2363566976211677, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.5261945724487305, "learning_rate": 1e-05, "loss": 0.4448, "mean_token_accuracy": 0.8863259553909302, "num_tokens": 667534258.0, "step": 17580 }, { "epoch": 2.2364839078997583, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 6.951694965362549, "learning_rate": 1e-05, "loss": 0.4746, "mean_token_accuracy": 0.879758894443512, "num_tokens": 667572689.0, "step": 17581 }, { "epoch": 2.236611118178349, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.464691638946533, "learning_rate": 1e-05, "loss": 0.4951, "mean_token_accuracy": 0.8726712465286255, "num_tokens": 667612425.0, "step": 17582 }, { "epoch": 2.2367383284569393, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.708045482635498, "learning_rate": 1e-05, "loss": 0.4979, "mean_token_accuracy": 0.8708123564720154, "num_tokens": 667653309.0, "step": 17583 }, { "epoch": 2.23686553873553, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.332348823547363, "learning_rate": 1e-05, "loss": 0.4767, "mean_token_accuracy": 0.8797831535339355, "num_tokens": 667699174.0, "step": 17584 }, { "epoch": 2.2369927490141204, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.60441780090332, "learning_rate": 1e-05, "loss": 0.4704, "mean_token_accuracy": 0.8797248601913452, "num_tokens": 667738987.0, "step": 17585 }, { "epoch": 2.237119959292711, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.1913018226623535, "learning_rate": 1e-05, "loss": 0.4643, "mean_token_accuracy": 0.8812713623046875, "num_tokens": 667780678.0, "step": 17586 }, { "epoch": 2.2372471695713014, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.615077018737793, "learning_rate": 1e-05, "loss": 0.4612, "mean_token_accuracy": 0.881712794303894, "num_tokens": 667816298.0, "step": 17587 }, { "epoch": 2.237374379849892, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 6.055398464202881, "learning_rate": 1e-05, "loss": 0.5078, "mean_token_accuracy": 0.8669357299804688, "num_tokens": 667855723.0, "step": 17588 }, { "epoch": 2.2375015901284825, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.737957000732422, "learning_rate": 1e-05, "loss": 0.517, "mean_token_accuracy": 0.867852509021759, "num_tokens": 667893895.0, "step": 17589 }, { "epoch": 2.237628800407073, "ewc_loss": 0.11181640625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.89224100112915, "learning_rate": 1e-05, "loss": 0.4162, "mean_token_accuracy": 0.8942741751670837, "num_tokens": 667932061.0, "step": 17590 }, { "epoch": 2.2377560106856635, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.998124599456787, "learning_rate": 1e-05, "loss": 0.4568, "mean_token_accuracy": 0.8868334293365479, "num_tokens": 667965723.0, "step": 17591 }, { "epoch": 2.237883220964254, "ewc_loss": 0.11181640625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.824474811553955, "learning_rate": 1e-05, "loss": 0.4732, "mean_token_accuracy": 0.8724339008331299, "num_tokens": 668002310.0, "step": 17592 }, { "epoch": 2.2380104312428446, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00011491775512695312, "grad_norm": 7.295909404754639, "learning_rate": 1e-05, "loss": 0.5317, "mean_token_accuracy": 0.8628321290016174, "num_tokens": 668044970.0, "step": 17593 }, { "epoch": 2.238137641521435, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 5.9466447830200195, "learning_rate": 1e-05, "loss": 0.5035, "mean_token_accuracy": 0.8680381774902344, "num_tokens": 668081898.0, "step": 17594 }, { "epoch": 2.238264851800025, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00011920928955078125, "grad_norm": 7.478684425354004, "learning_rate": 1e-05, "loss": 0.4983, "mean_token_accuracy": 0.8756349682807922, "num_tokens": 668119072.0, "step": 17595 }, { "epoch": 2.238392062078616, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.256577491760254, "learning_rate": 1e-05, "loss": 0.4551, "mean_token_accuracy": 0.8820189237594604, "num_tokens": 668157063.0, "step": 17596 }, { "epoch": 2.2385192723572063, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00011587142944335938, "grad_norm": 7.307373046875, "learning_rate": 1e-05, "loss": 0.5588, "mean_token_accuracy": 0.8553148508071899, "num_tokens": 668196985.0, "step": 17597 }, { "epoch": 2.238646482635797, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.513290882110596, "learning_rate": 1e-05, "loss": 0.4613, "mean_token_accuracy": 0.8845573663711548, "num_tokens": 668235954.0, "step": 17598 }, { "epoch": 2.2387736929143873, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.839017391204834, "learning_rate": 1e-05, "loss": 0.4826, "mean_token_accuracy": 0.875969648361206, "num_tokens": 668275442.0, "step": 17599 }, { "epoch": 2.238900903192978, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.640606880187988, "learning_rate": 1e-05, "loss": 0.5024, "mean_token_accuracy": 0.8781574368476868, "num_tokens": 668311413.0, "step": 17600 }, { "epoch": 2.2390281134715684, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.3084917068481445, "learning_rate": 1e-05, "loss": 0.4255, "mean_token_accuracy": 0.8886113166809082, "num_tokens": 668346928.0, "step": 17601 }, { "epoch": 2.239155323750159, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.795777797698975, "learning_rate": 1e-05, "loss": 0.4596, "mean_token_accuracy": 0.886793851852417, "num_tokens": 668388566.0, "step": 17602 }, { "epoch": 2.2392825340287494, "ewc_loss": 0.11181640625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.880829334259033, "learning_rate": 1e-05, "loss": 0.4243, "mean_token_accuracy": 0.8895279169082642, "num_tokens": 668430253.0, "step": 17603 }, { "epoch": 2.23940974430734, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00011157989501953125, "grad_norm": 7.134007453918457, "learning_rate": 1e-05, "loss": 0.5154, "mean_token_accuracy": 0.8676601648330688, "num_tokens": 668470561.0, "step": 17604 }, { "epoch": 2.2395369545859305, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.7138214111328125, "learning_rate": 1e-05, "loss": 0.4512, "mean_token_accuracy": 0.87763512134552, "num_tokens": 668509718.0, "step": 17605 }, { "epoch": 2.239664164864521, "ewc_loss": 0.1484375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.0001220703125, "grad_norm": 7.568504810333252, "learning_rate": 1e-05, "loss": 0.5129, "mean_token_accuracy": 0.8714321255683899, "num_tokens": 668548367.0, "step": 17606 }, { "epoch": 2.2397913751431116, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 5.95683479309082, "learning_rate": 1e-05, "loss": 0.5309, "mean_token_accuracy": 0.8580654263496399, "num_tokens": 668586977.0, "step": 17607 }, { "epoch": 2.239918585421702, "ewc_loss": 0.15625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00012969970703125, "grad_norm": 7.807094097137451, "learning_rate": 1e-05, "loss": 0.5093, "mean_token_accuracy": 0.8779274821281433, "num_tokens": 668617801.0, "step": 17608 }, { "epoch": 2.2400457957002926, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.465240001678467, "learning_rate": 1e-05, "loss": 0.501, "mean_token_accuracy": 0.8700541853904724, "num_tokens": 668656074.0, "step": 17609 }, { "epoch": 2.240173005978883, "ewc_loss": 0.1484375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00012111663818359375, "grad_norm": 7.55206823348999, "learning_rate": 1e-05, "loss": 0.4728, "mean_token_accuracy": 0.8860195279121399, "num_tokens": 668698354.0, "step": 17610 }, { "epoch": 2.2403002162574737, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 6.685688018798828, "learning_rate": 1e-05, "loss": 0.4726, "mean_token_accuracy": 0.8845144510269165, "num_tokens": 668734822.0, "step": 17611 }, { "epoch": 2.240427426536064, "ewc_loss": 0.140625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00011348724365234375, "grad_norm": 7.146883010864258, "learning_rate": 1e-05, "loss": 0.4203, "mean_token_accuracy": 0.8992966413497925, "num_tokens": 668773136.0, "step": 17612 }, { "epoch": 2.2405546368146547, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.614511966705322, "learning_rate": 1e-05, "loss": 0.4909, "mean_token_accuracy": 0.8777607083320618, "num_tokens": 668811580.0, "step": 17613 }, { "epoch": 2.2406818470932452, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.818647861480713, "learning_rate": 1e-05, "loss": 0.4662, "mean_token_accuracy": 0.8788172602653503, "num_tokens": 668846743.0, "step": 17614 }, { "epoch": 2.2408090573718358, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.471288681030273, "learning_rate": 1e-05, "loss": 0.4544, "mean_token_accuracy": 0.8855018615722656, "num_tokens": 668885634.0, "step": 17615 }, { "epoch": 2.2409362676504263, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.636112213134766, "learning_rate": 1e-05, "loss": 0.4655, "mean_token_accuracy": 0.8785900473594666, "num_tokens": 668920319.0, "step": 17616 }, { "epoch": 2.241063477929017, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.299416542053223, "learning_rate": 1e-05, "loss": 0.4692, "mean_token_accuracy": 0.8797590136528015, "num_tokens": 668959741.0, "step": 17617 }, { "epoch": 2.2411906882076074, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.606988430023193, "learning_rate": 1e-05, "loss": 0.553, "mean_token_accuracy": 0.8543190956115723, "num_tokens": 668998469.0, "step": 17618 }, { "epoch": 2.241317898486198, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.132419109344482, "learning_rate": 1e-05, "loss": 0.5054, "mean_token_accuracy": 0.8656173944473267, "num_tokens": 669030018.0, "step": 17619 }, { "epoch": 2.241445108764788, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.677525997161865, "learning_rate": 1e-05, "loss": 0.4449, "mean_token_accuracy": 0.8868029713630676, "num_tokens": 669066393.0, "step": 17620 }, { "epoch": 2.2415723190433785, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.949442386627197, "learning_rate": 1e-05, "loss": 0.4403, "mean_token_accuracy": 0.887503981590271, "num_tokens": 669108321.0, "step": 17621 }, { "epoch": 2.241699529321969, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.8880391120910645, "learning_rate": 1e-05, "loss": 0.474, "mean_token_accuracy": 0.8781112432479858, "num_tokens": 669147878.0, "step": 17622 }, { "epoch": 2.2418267396005596, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.822530746459961, "learning_rate": 1e-05, "loss": 0.4966, "mean_token_accuracy": 0.8650840520858765, "num_tokens": 669192348.0, "step": 17623 }, { "epoch": 2.24195394987915, "ewc_loss": 0.140625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00011444091796875, "grad_norm": 7.224465370178223, "learning_rate": 1e-05, "loss": 0.5233, "mean_token_accuracy": 0.8688490390777588, "num_tokens": 669233763.0, "step": 17624 }, { "epoch": 2.2420811601577406, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 5.860280990600586, "learning_rate": 1e-05, "loss": 0.4407, "mean_token_accuracy": 0.8836355805397034, "num_tokens": 669269609.0, "step": 17625 }, { "epoch": 2.242208370436331, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.0001201629638671875, "grad_norm": 7.513507843017578, "learning_rate": 1e-05, "loss": 0.4992, "mean_token_accuracy": 0.8753726482391357, "num_tokens": 669308649.0, "step": 17626 }, { "epoch": 2.2423355807149217, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.15216064453125, "learning_rate": 1e-05, "loss": 0.4754, "mean_token_accuracy": 0.8776973485946655, "num_tokens": 669350550.0, "step": 17627 }, { "epoch": 2.242462790993512, "ewc_loss": 0.1484375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00012159347534179688, "grad_norm": 7.503249168395996, "learning_rate": 1e-05, "loss": 0.5059, "mean_token_accuracy": 0.8750772476196289, "num_tokens": 669392795.0, "step": 17628 }, { "epoch": 2.2425900012721027, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.511282444000244, "learning_rate": 1e-05, "loss": 0.4967, "mean_token_accuracy": 0.8710146546363831, "num_tokens": 669426928.0, "step": 17629 }, { "epoch": 2.2427172115506933, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00011110305786132812, "grad_norm": 7.092067241668701, "learning_rate": 1e-05, "loss": 0.4677, "mean_token_accuracy": 0.8837246894836426, "num_tokens": 669467064.0, "step": 17630 }, { "epoch": 2.242844421829284, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.645583629608154, "learning_rate": 1e-05, "loss": 0.5085, "mean_token_accuracy": 0.8683291673660278, "num_tokens": 669506656.0, "step": 17631 }, { "epoch": 2.2429716321078743, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.5716094970703125, "learning_rate": 1e-05, "loss": 0.5697, "mean_token_accuracy": 0.847712516784668, "num_tokens": 669545787.0, "step": 17632 }, { "epoch": 2.243098842386465, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.720183849334717, "learning_rate": 1e-05, "loss": 0.489, "mean_token_accuracy": 0.8760215044021606, "num_tokens": 669581242.0, "step": 17633 }, { "epoch": 2.2432260526650554, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 6.418150901794434, "learning_rate": 1e-05, "loss": 0.4241, "mean_token_accuracy": 0.8904215097427368, "num_tokens": 669621586.0, "step": 17634 }, { "epoch": 2.243353262943646, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.684051990509033, "learning_rate": 1e-05, "loss": 0.4784, "mean_token_accuracy": 0.8776629567146301, "num_tokens": 669662903.0, "step": 17635 }, { "epoch": 2.2434804732222364, "ewc_loss": 0.11376953125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.931124687194824, "learning_rate": 1e-05, "loss": 0.4564, "mean_token_accuracy": 0.8783537149429321, "num_tokens": 669698409.0, "step": 17636 }, { "epoch": 2.243607683500827, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.8704447746276855, "learning_rate": 1e-05, "loss": 0.5375, "mean_token_accuracy": 0.8614879846572876, "num_tokens": 669741344.0, "step": 17637 }, { "epoch": 2.2437348937794175, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.74686336517334, "learning_rate": 1e-05, "loss": 0.4351, "mean_token_accuracy": 0.8848349452018738, "num_tokens": 669780042.0, "step": 17638 }, { "epoch": 2.243862104058008, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00011491775512695312, "grad_norm": 7.241022109985352, "learning_rate": 1e-05, "loss": 0.5449, "mean_token_accuracy": 0.8590928316116333, "num_tokens": 669819340.0, "step": 17639 }, { "epoch": 2.2439893143365985, "ewc_loss": 0.11376953125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.761649131774902, "learning_rate": 1e-05, "loss": 0.4283, "mean_token_accuracy": 0.8888630867004395, "num_tokens": 669859269.0, "step": 17640 }, { "epoch": 2.244116524615189, "ewc_loss": 0.15234375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00012493133544921875, "grad_norm": 7.594585418701172, "learning_rate": 1e-05, "loss": 0.4985, "mean_token_accuracy": 0.8800216913223267, "num_tokens": 669903085.0, "step": 17641 }, { "epoch": 2.2442437348937796, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.090954780578613, "learning_rate": 1e-05, "loss": 0.4533, "mean_token_accuracy": 0.8856329917907715, "num_tokens": 669945329.0, "step": 17642 }, { "epoch": 2.24437094517237, "ewc_loss": 0.15234375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.000125885009765625, "grad_norm": 7.6600728034973145, "learning_rate": 1e-05, "loss": 0.4848, "mean_token_accuracy": 0.885187029838562, "num_tokens": 669985141.0, "step": 17643 }, { "epoch": 2.2444981554509607, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 6.480633735656738, "learning_rate": 1e-05, "loss": 0.5292, "mean_token_accuracy": 0.8663877248764038, "num_tokens": 670024034.0, "step": 17644 }, { "epoch": 2.2446253657295507, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.000118255615234375, "grad_norm": 7.3697357177734375, "learning_rate": 1e-05, "loss": 0.4869, "mean_token_accuracy": 0.8821898102760315, "num_tokens": 670062098.0, "step": 17645 }, { "epoch": 2.2447525760081413, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 6.615706443786621, "learning_rate": 1e-05, "loss": 0.4989, "mean_token_accuracy": 0.8744691610336304, "num_tokens": 670100801.0, "step": 17646 }, { "epoch": 2.244879786286732, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 6.95758581161499, "learning_rate": 1e-05, "loss": 0.5017, "mean_token_accuracy": 0.872291088104248, "num_tokens": 670142221.0, "step": 17647 }, { "epoch": 2.2450069965653223, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.560649871826172, "learning_rate": 1e-05, "loss": 0.4633, "mean_token_accuracy": 0.8833761811256409, "num_tokens": 670183131.0, "step": 17648 }, { "epoch": 2.245134206843913, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.617089748382568, "learning_rate": 1e-05, "loss": 0.418, "mean_token_accuracy": 0.8935278654098511, "num_tokens": 670219207.0, "step": 17649 }, { "epoch": 2.2452614171225034, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.481732368469238, "learning_rate": 1e-05, "loss": 0.4847, "mean_token_accuracy": 0.8770643472671509, "num_tokens": 670254085.0, "step": 17650 }, { "epoch": 2.245388627401094, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.416208744049072, "learning_rate": 1e-05, "loss": 0.519, "mean_token_accuracy": 0.8656002283096313, "num_tokens": 670299669.0, "step": 17651 }, { "epoch": 2.2455158376796844, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.382336616516113, "learning_rate": 1e-05, "loss": 0.4257, "mean_token_accuracy": 0.8918500542640686, "num_tokens": 670332898.0, "step": 17652 }, { "epoch": 2.245643047958275, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.327661037445068, "learning_rate": 1e-05, "loss": 0.4768, "mean_token_accuracy": 0.8766512870788574, "num_tokens": 670371635.0, "step": 17653 }, { "epoch": 2.2457702582368655, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.2682013511657715, "learning_rate": 1e-05, "loss": 0.49, "mean_token_accuracy": 0.8748866319656372, "num_tokens": 670412568.0, "step": 17654 }, { "epoch": 2.245897468515456, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.309437274932861, "learning_rate": 1e-05, "loss": 0.4391, "mean_token_accuracy": 0.8884032964706421, "num_tokens": 670454338.0, "step": 17655 }, { "epoch": 2.2460246787940465, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.164449691772461, "learning_rate": 1e-05, "loss": 0.4411, "mean_token_accuracy": 0.8833566308021545, "num_tokens": 670490826.0, "step": 17656 }, { "epoch": 2.246151889072637, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.352118968963623, "learning_rate": 1e-05, "loss": 0.4993, "mean_token_accuracy": 0.8669955134391785, "num_tokens": 670528806.0, "step": 17657 }, { "epoch": 2.2462790993512276, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.0381855964660645, "learning_rate": 1e-05, "loss": 0.4687, "mean_token_accuracy": 0.8741147518157959, "num_tokens": 670563349.0, "step": 17658 }, { "epoch": 2.246406309629818, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.4803853034973145, "learning_rate": 1e-05, "loss": 0.4853, "mean_token_accuracy": 0.8729528188705444, "num_tokens": 670604682.0, "step": 17659 }, { "epoch": 2.2465335199084087, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.872867107391357, "learning_rate": 1e-05, "loss": 0.5016, "mean_token_accuracy": 0.8657426834106445, "num_tokens": 670644807.0, "step": 17660 }, { "epoch": 2.246660730186999, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.715137958526611, "learning_rate": 1e-05, "loss": 0.4681, "mean_token_accuracy": 0.8818262815475464, "num_tokens": 670683788.0, "step": 17661 }, { "epoch": 2.2467879404655897, "ewc_loss": 0.10986328125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.684070110321045, "learning_rate": 1e-05, "loss": 0.4627, "mean_token_accuracy": 0.8829824924468994, "num_tokens": 670728270.0, "step": 17662 }, { "epoch": 2.2469151507441802, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00011205673217773438, "grad_norm": 7.15005350112915, "learning_rate": 1e-05, "loss": 0.4782, "mean_token_accuracy": 0.8792750239372253, "num_tokens": 670763203.0, "step": 17663 }, { "epoch": 2.2470423610227708, "ewc_loss": 0.11083984375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.563520431518555, "learning_rate": 1e-05, "loss": 0.4834, "mean_token_accuracy": 0.8709635734558105, "num_tokens": 670804548.0, "step": 17664 }, { "epoch": 2.2471695713013613, "ewc_loss": 0.15625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00012874603271484375, "grad_norm": 7.776607990264893, "learning_rate": 1e-05, "loss": 0.5221, "mean_token_accuracy": 0.8743300437927246, "num_tokens": 670844922.0, "step": 17665 }, { "epoch": 2.247296781579952, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 5.89931583404541, "learning_rate": 1e-05, "loss": 0.4353, "mean_token_accuracy": 0.8871043920516968, "num_tokens": 670882369.0, "step": 17666 }, { "epoch": 2.2474239918585424, "ewc_loss": 0.1640625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.0001373291015625, "grad_norm": 8.194713592529297, "learning_rate": 1e-05, "loss": 0.4798, "mean_token_accuracy": 0.8896292448043823, "num_tokens": 670922716.0, "step": 17667 }, { "epoch": 2.247551202137133, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00011682510375976562, "grad_norm": 6.704811096191406, "learning_rate": 1e-05, "loss": 0.516, "mean_token_accuracy": 0.8720163106918335, "num_tokens": 670964618.0, "step": 17668 }, { "epoch": 2.2476784124157234, "ewc_loss": 0.15625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00012874603271484375, "grad_norm": 7.834259510040283, "learning_rate": 1e-05, "loss": 0.4538, "mean_token_accuracy": 0.8958631157875061, "num_tokens": 670997915.0, "step": 17669 }, { "epoch": 2.2478056226943135, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00011968612670898438, "grad_norm": 7.012889862060547, "learning_rate": 1e-05, "loss": 0.5144, "mean_token_accuracy": 0.869957447052002, "num_tokens": 671031675.0, "step": 17670 }, { "epoch": 2.247932832972904, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00011539459228515625, "grad_norm": 7.303515434265137, "learning_rate": 1e-05, "loss": 0.5266, "mean_token_accuracy": 0.8685011863708496, "num_tokens": 671068060.0, "step": 17671 }, { "epoch": 2.2480600432514946, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00011205673217773438, "grad_norm": 6.8232035636901855, "learning_rate": 1e-05, "loss": 0.4875, "mean_token_accuracy": 0.8799776434898376, "num_tokens": 671109014.0, "step": 17672 }, { "epoch": 2.248187253530085, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.928404808044434, "learning_rate": 1e-05, "loss": 0.4691, "mean_token_accuracy": 0.8803602457046509, "num_tokens": 671140173.0, "step": 17673 }, { "epoch": 2.2483144638086756, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.62495756149292, "learning_rate": 1e-05, "loss": 0.4974, "mean_token_accuracy": 0.8728009462356567, "num_tokens": 671177644.0, "step": 17674 }, { "epoch": 2.248441674087266, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.670523166656494, "learning_rate": 1e-05, "loss": 0.4785, "mean_token_accuracy": 0.8753200173377991, "num_tokens": 671221860.0, "step": 17675 }, { "epoch": 2.2485688843658567, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.431891441345215, "learning_rate": 1e-05, "loss": 0.4777, "mean_token_accuracy": 0.878456711769104, "num_tokens": 671262567.0, "step": 17676 }, { "epoch": 2.248696094644447, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.542422771453857, "learning_rate": 1e-05, "loss": 0.4925, "mean_token_accuracy": 0.8716998100280762, "num_tokens": 671301195.0, "step": 17677 }, { "epoch": 2.2488233049230377, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.25326681137085, "learning_rate": 1e-05, "loss": 0.4873, "mean_token_accuracy": 0.8740646839141846, "num_tokens": 671346028.0, "step": 17678 }, { "epoch": 2.2489505152016283, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.531435966491699, "learning_rate": 1e-05, "loss": 0.45, "mean_token_accuracy": 0.8828428387641907, "num_tokens": 671386927.0, "step": 17679 }, { "epoch": 2.249077725480219, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.09213399887085, "learning_rate": 1e-05, "loss": 0.4844, "mean_token_accuracy": 0.8729279041290283, "num_tokens": 671422447.0, "step": 17680 }, { "epoch": 2.2492049357588093, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.630019187927246, "learning_rate": 1e-05, "loss": 0.4803, "mean_token_accuracy": 0.873336911201477, "num_tokens": 671461390.0, "step": 17681 }, { "epoch": 2.2493321460374, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.907935619354248, "learning_rate": 1e-05, "loss": 0.4631, "mean_token_accuracy": 0.8780854940414429, "num_tokens": 671499150.0, "step": 17682 }, { "epoch": 2.2494593563159904, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.875301837921143, "learning_rate": 1e-05, "loss": 0.4878, "mean_token_accuracy": 0.8765950202941895, "num_tokens": 671537529.0, "step": 17683 }, { "epoch": 2.249586566594581, "ewc_loss": 0.109375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.735763072967529, "learning_rate": 1e-05, "loss": 0.4814, "mean_token_accuracy": 0.8708208799362183, "num_tokens": 671577661.0, "step": 17684 }, { "epoch": 2.2497137768731714, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00011539459228515625, "grad_norm": 7.304812431335449, "learning_rate": 1e-05, "loss": 0.4669, "mean_token_accuracy": 0.883759081363678, "num_tokens": 671611375.0, "step": 17685 }, { "epoch": 2.249840987151762, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.729030132293701, "learning_rate": 1e-05, "loss": 0.4905, "mean_token_accuracy": 0.868023693561554, "num_tokens": 671655430.0, "step": 17686 }, { "epoch": 2.2499681974303525, "ewc_loss": 0.15625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00012874603271484375, "grad_norm": 7.79453706741333, "learning_rate": 1e-05, "loss": 0.4893, "mean_token_accuracy": 0.8806034326553345, "num_tokens": 671696451.0, "step": 17687 }, { "epoch": 2.250095407708943, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.176627159118652, "learning_rate": 1e-05, "loss": 0.4983, "mean_token_accuracy": 0.8704727292060852, "num_tokens": 671736576.0, "step": 17688 }, { "epoch": 2.2502226179875335, "ewc_loss": 0.158203125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00013065338134765625, "grad_norm": 7.897451877593994, "learning_rate": 1e-05, "loss": 0.5095, "mean_token_accuracy": 0.8782972097396851, "num_tokens": 671774638.0, "step": 17689 }, { "epoch": 2.250349828266124, "ewc_loss": 0.140625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00011348724365234375, "grad_norm": 6.752941131591797, "learning_rate": 1e-05, "loss": 0.4937, "mean_token_accuracy": 0.8737561702728271, "num_tokens": 671812626.0, "step": 17690 }, { "epoch": 2.2504770385447146, "ewc_loss": 0.1455078125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00011873245239257812, "grad_norm": 7.414791107177734, "learning_rate": 1e-05, "loss": 0.4946, "mean_token_accuracy": 0.882006049156189, "num_tokens": 671844752.0, "step": 17691 }, { "epoch": 2.250604248823305, "ewc_loss": 0.140625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00011348724365234375, "grad_norm": 6.85542106628418, "learning_rate": 1e-05, "loss": 0.5128, "mean_token_accuracy": 0.8727884888648987, "num_tokens": 671883161.0, "step": 17692 }, { "epoch": 2.250731459101895, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.883069038391113, "learning_rate": 1e-05, "loss": 0.4648, "mean_token_accuracy": 0.8822999000549316, "num_tokens": 671924386.0, "step": 17693 }, { "epoch": 2.250858669380486, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 6.774168491363525, "learning_rate": 1e-05, "loss": 0.4915, "mean_token_accuracy": 0.8750274777412415, "num_tokens": 671958452.0, "step": 17694 }, { "epoch": 2.2509858796590763, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.451268196105957, "learning_rate": 1e-05, "loss": 0.4855, "mean_token_accuracy": 0.871707022190094, "num_tokens": 671995335.0, "step": 17695 }, { "epoch": 2.251113089937667, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.753566265106201, "learning_rate": 1e-05, "loss": 0.4855, "mean_token_accuracy": 0.8723821043968201, "num_tokens": 672025640.0, "step": 17696 }, { "epoch": 2.2512403002162573, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.16248893737793, "learning_rate": 1e-05, "loss": 0.4955, "mean_token_accuracy": 0.8693605661392212, "num_tokens": 672066013.0, "step": 17697 }, { "epoch": 2.251367510494848, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.801940441131592, "learning_rate": 1e-05, "loss": 0.4783, "mean_token_accuracy": 0.8780108690261841, "num_tokens": 672107243.0, "step": 17698 }, { "epoch": 2.2514947207734384, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 5.975515842437744, "learning_rate": 1e-05, "loss": 0.4616, "mean_token_accuracy": 0.8777342438697815, "num_tokens": 672145753.0, "step": 17699 }, { "epoch": 2.251621931052029, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.963921546936035, "learning_rate": 1e-05, "loss": 0.4713, "mean_token_accuracy": 0.8789949417114258, "num_tokens": 672185489.0, "step": 17700 }, { "epoch": 2.2517491413306194, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.863925457000732, "learning_rate": 1e-05, "loss": 0.4439, "mean_token_accuracy": 0.8806979656219482, "num_tokens": 672225702.0, "step": 17701 }, { "epoch": 2.25187635160921, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00011301040649414062, "grad_norm": 7.219664573669434, "learning_rate": 1e-05, "loss": 0.5078, "mean_token_accuracy": 0.8711820840835571, "num_tokens": 672265901.0, "step": 17702 }, { "epoch": 2.2520035618878005, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 5.8979668617248535, "learning_rate": 1e-05, "loss": 0.5009, "mean_token_accuracy": 0.8676309585571289, "num_tokens": 672304081.0, "step": 17703 }, { "epoch": 2.252130772166391, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00011920928955078125, "grad_norm": 7.446977138519287, "learning_rate": 1e-05, "loss": 0.4918, "mean_token_accuracy": 0.8799989223480225, "num_tokens": 672343425.0, "step": 17704 }, { "epoch": 2.2522579824449815, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.089698791503906, "learning_rate": 1e-05, "loss": 0.4543, "mean_token_accuracy": 0.8829768896102905, "num_tokens": 672377745.0, "step": 17705 }, { "epoch": 2.252385192723572, "ewc_loss": 0.1484375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00012111663818359375, "grad_norm": 7.504645347595215, "learning_rate": 1e-05, "loss": 0.472, "mean_token_accuracy": 0.8858283758163452, "num_tokens": 672413811.0, "step": 17706 }, { "epoch": 2.2525124030021626, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.35089111328125, "learning_rate": 1e-05, "loss": 0.4942, "mean_token_accuracy": 0.8726959228515625, "num_tokens": 672455083.0, "step": 17707 }, { "epoch": 2.252639613280753, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00011491775512695312, "grad_norm": 7.301214694976807, "learning_rate": 1e-05, "loss": 0.4673, "mean_token_accuracy": 0.8858671188354492, "num_tokens": 672502401.0, "step": 17708 }, { "epoch": 2.2527668235593437, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.487508773803711, "learning_rate": 1e-05, "loss": 0.4718, "mean_token_accuracy": 0.8806172609329224, "num_tokens": 672540504.0, "step": 17709 }, { "epoch": 2.252894033837934, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 6.974991321563721, "learning_rate": 1e-05, "loss": 0.4773, "mean_token_accuracy": 0.8796011209487915, "num_tokens": 672578157.0, "step": 17710 }, { "epoch": 2.2530212441165247, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.489842891693115, "learning_rate": 1e-05, "loss": 0.4808, "mean_token_accuracy": 0.8778467178344727, "num_tokens": 672611859.0, "step": 17711 }, { "epoch": 2.2531484543951152, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.653044700622559, "learning_rate": 1e-05, "loss": 0.4858, "mean_token_accuracy": 0.8730506896972656, "num_tokens": 672651779.0, "step": 17712 }, { "epoch": 2.2532756646737058, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.431786060333252, "learning_rate": 1e-05, "loss": 0.4945, "mean_token_accuracy": 0.8729935884475708, "num_tokens": 672691559.0, "step": 17713 }, { "epoch": 2.2534028749522963, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.4442572593688965, "learning_rate": 1e-05, "loss": 0.4672, "mean_token_accuracy": 0.8786843419075012, "num_tokens": 672731353.0, "step": 17714 }, { "epoch": 2.253530085230887, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.353926181793213, "learning_rate": 1e-05, "loss": 0.4532, "mean_token_accuracy": 0.8818938732147217, "num_tokens": 672770427.0, "step": 17715 }, { "epoch": 2.2536572955094774, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.324528217315674, "learning_rate": 1e-05, "loss": 0.4868, "mean_token_accuracy": 0.874387800693512, "num_tokens": 672808817.0, "step": 17716 }, { "epoch": 2.253784505788068, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.271744251251221, "learning_rate": 1e-05, "loss": 0.4692, "mean_token_accuracy": 0.8762823343276978, "num_tokens": 672848368.0, "step": 17717 }, { "epoch": 2.253911716066658, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.299036502838135, "learning_rate": 1e-05, "loss": 0.4327, "mean_token_accuracy": 0.8901995420455933, "num_tokens": 672885074.0, "step": 17718 }, { "epoch": 2.254038926345249, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.179422855377197, "learning_rate": 1e-05, "loss": 0.4566, "mean_token_accuracy": 0.8792287707328796, "num_tokens": 672924866.0, "step": 17719 }, { "epoch": 2.254166136623839, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.331057071685791, "learning_rate": 1e-05, "loss": 0.4516, "mean_token_accuracy": 0.8829391002655029, "num_tokens": 672962051.0, "step": 17720 }, { "epoch": 2.2542933469024296, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.071915626525879, "learning_rate": 1e-05, "loss": 0.455, "mean_token_accuracy": 0.8810132741928101, "num_tokens": 672999954.0, "step": 17721 }, { "epoch": 2.25442055718102, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.425879001617432, "learning_rate": 1e-05, "loss": 0.5232, "mean_token_accuracy": 0.8629988431930542, "num_tokens": 673037825.0, "step": 17722 }, { "epoch": 2.2545477674596106, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 5.92698860168457, "learning_rate": 1e-05, "loss": 0.4229, "mean_token_accuracy": 0.891531765460968, "num_tokens": 673072036.0, "step": 17723 }, { "epoch": 2.254674977738201, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.613948822021484, "learning_rate": 1e-05, "loss": 0.5039, "mean_token_accuracy": 0.8657340407371521, "num_tokens": 673110733.0, "step": 17724 }, { "epoch": 2.2548021880167917, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.7516889572143555, "learning_rate": 1e-05, "loss": 0.4313, "mean_token_accuracy": 0.8889457583427429, "num_tokens": 673146611.0, "step": 17725 }, { "epoch": 2.254929398295382, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 6.961482048034668, "learning_rate": 1e-05, "loss": 0.4971, "mean_token_accuracy": 0.8748056888580322, "num_tokens": 673189965.0, "step": 17726 }, { "epoch": 2.2550566085739727, "ewc_loss": 0.109375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.5897135734558105, "learning_rate": 1e-05, "loss": 0.4998, "mean_token_accuracy": 0.8653284907341003, "num_tokens": 673226990.0, "step": 17727 }, { "epoch": 2.2551838188525632, "ewc_loss": 0.1484375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00012111663818359375, "grad_norm": 7.536214351654053, "learning_rate": 1e-05, "loss": 0.5075, "mean_token_accuracy": 0.8731899261474609, "num_tokens": 673263109.0, "step": 17728 }, { "epoch": 2.2553110291311538, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 5.700993061065674, "learning_rate": 1e-05, "loss": 0.4421, "mean_token_accuracy": 0.8833236694335938, "num_tokens": 673304081.0, "step": 17729 }, { "epoch": 2.2554382394097443, "ewc_loss": 0.1640625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00013637542724609375, "grad_norm": 8.11418628692627, "learning_rate": 1e-05, "loss": 0.4908, "mean_token_accuracy": 0.8864421248435974, "num_tokens": 673345510.0, "step": 17730 }, { "epoch": 2.255565449688335, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00011110305786132812, "grad_norm": 6.441101551055908, "learning_rate": 1e-05, "loss": 0.5102, "mean_token_accuracy": 0.872529149055481, "num_tokens": 673383607.0, "step": 17731 }, { "epoch": 2.2556926599669254, "ewc_loss": 0.16015625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00013256072998046875, "grad_norm": 7.999457836151123, "learning_rate": 1e-05, "loss": 0.5116, "mean_token_accuracy": 0.8742560148239136, "num_tokens": 673422676.0, "step": 17732 }, { "epoch": 2.255819870245516, "ewc_loss": 0.1484375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.0001220703125, "grad_norm": 6.966167449951172, "learning_rate": 1e-05, "loss": 0.4995, "mean_token_accuracy": 0.8784087300300598, "num_tokens": 673466135.0, "step": 17733 }, { "epoch": 2.2559470805241064, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00011968612670898438, "grad_norm": 7.464719295501709, "learning_rate": 1e-05, "loss": 0.5285, "mean_token_accuracy": 0.8669499754905701, "num_tokens": 673509003.0, "step": 17734 }, { "epoch": 2.256074290802697, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00011491775512695312, "grad_norm": 6.896303653717041, "learning_rate": 1e-05, "loss": 0.5156, "mean_token_accuracy": 0.8766355514526367, "num_tokens": 673546231.0, "step": 17735 }, { "epoch": 2.2562015010812875, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 7.0349297523498535, "learning_rate": 1e-05, "loss": 0.4936, "mean_token_accuracy": 0.8735086917877197, "num_tokens": 673585812.0, "step": 17736 }, { "epoch": 2.256328711359878, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.684556007385254, "learning_rate": 1e-05, "loss": 0.4526, "mean_token_accuracy": 0.8878135681152344, "num_tokens": 673621566.0, "step": 17737 }, { "epoch": 2.2564559216384685, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.719589710235596, "learning_rate": 1e-05, "loss": 0.4215, "mean_token_accuracy": 0.8955621719360352, "num_tokens": 673665659.0, "step": 17738 }, { "epoch": 2.256583131917059, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.49656343460083, "learning_rate": 1e-05, "loss": 0.451, "mean_token_accuracy": 0.8866064548492432, "num_tokens": 673705936.0, "step": 17739 }, { "epoch": 2.2567103421956496, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.567162036895752, "learning_rate": 1e-05, "loss": 0.4919, "mean_token_accuracy": 0.8701401948928833, "num_tokens": 673740741.0, "step": 17740 }, { "epoch": 2.2568375524742397, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.332028388977051, "learning_rate": 1e-05, "loss": 0.4613, "mean_token_accuracy": 0.8803049325942993, "num_tokens": 673778304.0, "step": 17741 }, { "epoch": 2.2569647627528306, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.497618198394775, "learning_rate": 1e-05, "loss": 0.4656, "mean_token_accuracy": 0.8794053792953491, "num_tokens": 673815136.0, "step": 17742 }, { "epoch": 2.2570919730314207, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.184475421905518, "learning_rate": 1e-05, "loss": 0.4634, "mean_token_accuracy": 0.8813091516494751, "num_tokens": 673847227.0, "step": 17743 }, { "epoch": 2.2572191833100113, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.532834529876709, "learning_rate": 1e-05, "loss": 0.4426, "mean_token_accuracy": 0.8882821798324585, "num_tokens": 673885110.0, "step": 17744 }, { "epoch": 2.257346393588602, "ewc_loss": 0.11474609375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 5.997978687286377, "learning_rate": 1e-05, "loss": 0.4933, "mean_token_accuracy": 0.8683404326438904, "num_tokens": 673929534.0, "step": 17745 }, { "epoch": 2.2574736038671923, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.720959186553955, "learning_rate": 1e-05, "loss": 0.4959, "mean_token_accuracy": 0.8686691522598267, "num_tokens": 673972353.0, "step": 17746 }, { "epoch": 2.257600814145783, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.800960063934326, "learning_rate": 1e-05, "loss": 0.4173, "mean_token_accuracy": 0.8876843452453613, "num_tokens": 674009441.0, "step": 17747 }, { "epoch": 2.2577280244243734, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 7.045459270477295, "learning_rate": 1e-05, "loss": 0.4518, "mean_token_accuracy": 0.8875328302383423, "num_tokens": 674045397.0, "step": 17748 }, { "epoch": 2.257855234702964, "ewc_loss": 0.109375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.676997184753418, "learning_rate": 1e-05, "loss": 0.4756, "mean_token_accuracy": 0.8697346448898315, "num_tokens": 674084896.0, "step": 17749 }, { "epoch": 2.2579824449815544, "ewc_loss": 0.1484375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00012111663818359375, "grad_norm": 7.586453914642334, "learning_rate": 1e-05, "loss": 0.535, "mean_token_accuracy": 0.8664419651031494, "num_tokens": 674121223.0, "step": 17750 }, { "epoch": 2.258109655260145, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 5.870654106140137, "learning_rate": 1e-05, "loss": 0.4826, "mean_token_accuracy": 0.8748014569282532, "num_tokens": 674157395.0, "step": 17751 }, { "epoch": 2.2582368655387355, "ewc_loss": 0.16015625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00013256072998046875, "grad_norm": 8.009254455566406, "learning_rate": 1e-05, "loss": 0.5816, "mean_token_accuracy": 0.8563377857208252, "num_tokens": 674200157.0, "step": 17752 }, { "epoch": 2.258364075817326, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00010967254638671875, "grad_norm": 6.530981540679932, "learning_rate": 1e-05, "loss": 0.5259, "mean_token_accuracy": 0.8652048707008362, "num_tokens": 674234696.0, "step": 17753 }, { "epoch": 2.2584912860959165, "ewc_loss": 0.154296875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00012683868408203125, "grad_norm": 7.751464366912842, "learning_rate": 1e-05, "loss": 0.5168, "mean_token_accuracy": 0.8756095170974731, "num_tokens": 674277456.0, "step": 17754 }, { "epoch": 2.258618496374507, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00011587142944335938, "grad_norm": 6.894204139709473, "learning_rate": 1e-05, "loss": 0.4587, "mean_token_accuracy": 0.8911222219467163, "num_tokens": 674316913.0, "step": 17755 }, { "epoch": 2.2587457066530976, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00011157989501953125, "grad_norm": 7.20375394821167, "learning_rate": 1e-05, "loss": 0.4869, "mean_token_accuracy": 0.8800595998764038, "num_tokens": 674355539.0, "step": 17756 }, { "epoch": 2.258872916931688, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00011110305786132812, "grad_norm": 6.842601299285889, "learning_rate": 1e-05, "loss": 0.512, "mean_token_accuracy": 0.8685952425003052, "num_tokens": 674386759.0, "step": 17757 }, { "epoch": 2.2590001272102787, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.719658851623535, "learning_rate": 1e-05, "loss": 0.4506, "mean_token_accuracy": 0.8837814331054688, "num_tokens": 674427098.0, "step": 17758 }, { "epoch": 2.259127337488869, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.733277320861816, "learning_rate": 1e-05, "loss": 0.4705, "mean_token_accuracy": 0.8828685283660889, "num_tokens": 674464766.0, "step": 17759 }, { "epoch": 2.2592545477674597, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.390578746795654, "learning_rate": 1e-05, "loss": 0.4829, "mean_token_accuracy": 0.8736952543258667, "num_tokens": 674506260.0, "step": 17760 }, { "epoch": 2.2593817580460502, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.681750774383545, "learning_rate": 1e-05, "loss": 0.4818, "mean_token_accuracy": 0.8801002502441406, "num_tokens": 674544255.0, "step": 17761 }, { "epoch": 2.2595089683246408, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.166756629943848, "learning_rate": 1e-05, "loss": 0.4252, "mean_token_accuracy": 0.8870224952697754, "num_tokens": 674579806.0, "step": 17762 }, { "epoch": 2.2596361786032313, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.704132556915283, "learning_rate": 1e-05, "loss": 0.5253, "mean_token_accuracy": 0.864722490310669, "num_tokens": 674620734.0, "step": 17763 }, { "epoch": 2.259763388881822, "ewc_loss": 0.11474609375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 6.014261245727539, "learning_rate": 1e-05, "loss": 0.4635, "mean_token_accuracy": 0.8825234174728394, "num_tokens": 674661213.0, "step": 17764 }, { "epoch": 2.2598905991604123, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.804689884185791, "learning_rate": 1e-05, "loss": 0.4563, "mean_token_accuracy": 0.8858262300491333, "num_tokens": 674696323.0, "step": 17765 }, { "epoch": 2.2600178094390024, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.8897809982299805, "learning_rate": 1e-05, "loss": 0.4621, "mean_token_accuracy": 0.8750240802764893, "num_tokens": 674728041.0, "step": 17766 }, { "epoch": 2.2601450197175934, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 7.003260135650635, "learning_rate": 1e-05, "loss": 0.4692, "mean_token_accuracy": 0.8838406801223755, "num_tokens": 674767518.0, "step": 17767 }, { "epoch": 2.2602722299961835, "ewc_loss": 0.11181640625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.800354957580566, "learning_rate": 1e-05, "loss": 0.4497, "mean_token_accuracy": 0.8794758319854736, "num_tokens": 674807510.0, "step": 17768 }, { "epoch": 2.260399440274774, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.0001163482666015625, "grad_norm": 7.311821460723877, "learning_rate": 1e-05, "loss": 0.478, "mean_token_accuracy": 0.8772549629211426, "num_tokens": 674844851.0, "step": 17769 }, { "epoch": 2.2605266505533645, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 5.86503267288208, "learning_rate": 1e-05, "loss": 0.4265, "mean_token_accuracy": 0.8912274837493896, "num_tokens": 674881839.0, "step": 17770 }, { "epoch": 2.260653860831955, "ewc_loss": 0.150390625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.6011061668396, "learning_rate": 1e-05, "loss": 0.5126, "mean_token_accuracy": 0.8755300045013428, "num_tokens": 674920450.0, "step": 17771 }, { "epoch": 2.2607810711105456, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.176994800567627, "learning_rate": 1e-05, "loss": 0.462, "mean_token_accuracy": 0.8796600103378296, "num_tokens": 674960031.0, "step": 17772 }, { "epoch": 2.260908281389136, "ewc_loss": 0.150390625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.6080851554870605, "learning_rate": 1e-05, "loss": 0.4679, "mean_token_accuracy": 0.885041356086731, "num_tokens": 674993354.0, "step": 17773 }, { "epoch": 2.2610354916677267, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.502682685852051, "learning_rate": 1e-05, "loss": 0.4784, "mean_token_accuracy": 0.8814899921417236, "num_tokens": 675029186.0, "step": 17774 }, { "epoch": 2.261162701946317, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00011539459228515625, "grad_norm": 7.280253887176514, "learning_rate": 1e-05, "loss": 0.4547, "mean_token_accuracy": 0.8883504867553711, "num_tokens": 675065170.0, "step": 17775 }, { "epoch": 2.2612899122249077, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 6.606324195861816, "learning_rate": 1e-05, "loss": 0.4482, "mean_token_accuracy": 0.8908185958862305, "num_tokens": 675102774.0, "step": 17776 }, { "epoch": 2.2614171225034982, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.871983528137207, "learning_rate": 1e-05, "loss": 0.4938, "mean_token_accuracy": 0.8713805675506592, "num_tokens": 675136890.0, "step": 17777 }, { "epoch": 2.2615443327820888, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.5843586921691895, "learning_rate": 1e-05, "loss": 0.4553, "mean_token_accuracy": 0.8853036761283875, "num_tokens": 675171959.0, "step": 17778 }, { "epoch": 2.2616715430606793, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.518514633178711, "learning_rate": 1e-05, "loss": 0.4554, "mean_token_accuracy": 0.8822596073150635, "num_tokens": 675203931.0, "step": 17779 }, { "epoch": 2.26179875333927, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.53721284866333, "learning_rate": 1e-05, "loss": 0.4304, "mean_token_accuracy": 0.8906147480010986, "num_tokens": 675243645.0, "step": 17780 }, { "epoch": 2.2619259636178604, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.264091491699219, "learning_rate": 1e-05, "loss": 0.4535, "mean_token_accuracy": 0.8800349831581116, "num_tokens": 675280318.0, "step": 17781 }, { "epoch": 2.262053173896451, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.51689338684082, "learning_rate": 1e-05, "loss": 0.4849, "mean_token_accuracy": 0.8739714026451111, "num_tokens": 675321442.0, "step": 17782 }, { "epoch": 2.2621803841750414, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.124603748321533, "learning_rate": 1e-05, "loss": 0.4176, "mean_token_accuracy": 0.893451452255249, "num_tokens": 675351039.0, "step": 17783 }, { "epoch": 2.262307594453632, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.5229058265686035, "learning_rate": 1e-05, "loss": 0.46, "mean_token_accuracy": 0.8802427053451538, "num_tokens": 675385756.0, "step": 17784 }, { "epoch": 2.2624348047322225, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 6.006361961364746, "learning_rate": 1e-05, "loss": 0.4145, "mean_token_accuracy": 0.8939048051834106, "num_tokens": 675429282.0, "step": 17785 }, { "epoch": 2.262562015010813, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.587038516998291, "learning_rate": 1e-05, "loss": 0.4388, "mean_token_accuracy": 0.8877581357955933, "num_tokens": 675466661.0, "step": 17786 }, { "epoch": 2.2626892252894035, "ewc_loss": 0.11376953125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.9009270668029785, "learning_rate": 1e-05, "loss": 0.4322, "mean_token_accuracy": 0.8859797716140747, "num_tokens": 675501885.0, "step": 17787 }, { "epoch": 2.262816435567994, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.746048450469971, "learning_rate": 1e-05, "loss": 0.5172, "mean_token_accuracy": 0.8669893145561218, "num_tokens": 675544423.0, "step": 17788 }, { "epoch": 2.2629436458465846, "ewc_loss": 0.11083984375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.771631240844727, "learning_rate": 1e-05, "loss": 0.464, "mean_token_accuracy": 0.8754830360412598, "num_tokens": 675585301.0, "step": 17789 }, { "epoch": 2.263070856125175, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 7.040092945098877, "learning_rate": 1e-05, "loss": 0.4864, "mean_token_accuracy": 0.8764162063598633, "num_tokens": 675625244.0, "step": 17790 }, { "epoch": 2.263198066403765, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.681442737579346, "learning_rate": 1e-05, "loss": 0.4219, "mean_token_accuracy": 0.8903970718383789, "num_tokens": 675657310.0, "step": 17791 }, { "epoch": 2.263325276682356, "ewc_loss": 0.1455078125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00011873245239257812, "grad_norm": 7.478256702423096, "learning_rate": 1e-05, "loss": 0.4968, "mean_token_accuracy": 0.878599226474762, "num_tokens": 675690702.0, "step": 17792 }, { "epoch": 2.2634524869609463, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 5.820135116577148, "learning_rate": 1e-05, "loss": 0.4371, "mean_token_accuracy": 0.8870372772216797, "num_tokens": 675725211.0, "step": 17793 }, { "epoch": 2.263579697239537, "ewc_loss": 0.158203125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.0001316070556640625, "grad_norm": 7.856362819671631, "learning_rate": 1e-05, "loss": 0.4718, "mean_token_accuracy": 0.890221118927002, "num_tokens": 675762418.0, "step": 17794 }, { "epoch": 2.2637069075181273, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.321935653686523, "learning_rate": 1e-05, "loss": 0.5287, "mean_token_accuracy": 0.8644533753395081, "num_tokens": 675807640.0, "step": 17795 }, { "epoch": 2.263834117796718, "ewc_loss": 0.15625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00012874603271484375, "grad_norm": 7.79374885559082, "learning_rate": 1e-05, "loss": 0.4945, "mean_token_accuracy": 0.879173994064331, "num_tokens": 675843032.0, "step": 17796 }, { "epoch": 2.2639613280753084, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00011491775512695312, "grad_norm": 6.732719898223877, "learning_rate": 1e-05, "loss": 0.4813, "mean_token_accuracy": 0.8803435564041138, "num_tokens": 675875150.0, "step": 17797 }, { "epoch": 2.264088538353899, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00011777877807617188, "grad_norm": 7.35103178024292, "learning_rate": 1e-05, "loss": 0.5464, "mean_token_accuracy": 0.8628361821174622, "num_tokens": 675917635.0, "step": 17798 }, { "epoch": 2.2642157486324894, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.0001125335693359375, "grad_norm": 6.7434163093566895, "learning_rate": 1e-05, "loss": 0.4944, "mean_token_accuracy": 0.8741883635520935, "num_tokens": 675955577.0, "step": 17799 }, { "epoch": 2.26434295891108, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.924844264984131, "learning_rate": 1e-05, "loss": 0.5257, "mean_token_accuracy": 0.862560510635376, "num_tokens": 675996326.0, "step": 17800 }, { "epoch": 2.2644701691896705, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.65445613861084, "learning_rate": 1e-05, "loss": 0.5289, "mean_token_accuracy": 0.8628295660018921, "num_tokens": 676040857.0, "step": 17801 }, { "epoch": 2.264597379468261, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.55286169052124, "learning_rate": 1e-05, "loss": 0.431, "mean_token_accuracy": 0.8896653056144714, "num_tokens": 676077601.0, "step": 17802 }, { "epoch": 2.2647245897468515, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.58149528503418, "learning_rate": 1e-05, "loss": 0.4854, "mean_token_accuracy": 0.87333083152771, "num_tokens": 676115558.0, "step": 17803 }, { "epoch": 2.264851800025442, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.317183017730713, "learning_rate": 1e-05, "loss": 0.4806, "mean_token_accuracy": 0.8759570121765137, "num_tokens": 676157828.0, "step": 17804 }, { "epoch": 2.2649790103040326, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.5299248695373535, "learning_rate": 1e-05, "loss": 0.5105, "mean_token_accuracy": 0.8663546442985535, "num_tokens": 676199147.0, "step": 17805 }, { "epoch": 2.265106220582623, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.14521598815918, "learning_rate": 1e-05, "loss": 0.4273, "mean_token_accuracy": 0.8894715309143066, "num_tokens": 676241271.0, "step": 17806 }, { "epoch": 2.2652334308612136, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.542065143585205, "learning_rate": 1e-05, "loss": 0.4757, "mean_token_accuracy": 0.8785926103591919, "num_tokens": 676276990.0, "step": 17807 }, { "epoch": 2.265360641139804, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.018847465515137, "learning_rate": 1e-05, "loss": 0.4674, "mean_token_accuracy": 0.8775649070739746, "num_tokens": 676313441.0, "step": 17808 }, { "epoch": 2.2654878514183947, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.615058422088623, "learning_rate": 1e-05, "loss": 0.4899, "mean_token_accuracy": 0.875938892364502, "num_tokens": 676350322.0, "step": 17809 }, { "epoch": 2.2656150616969852, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.889621734619141, "learning_rate": 1e-05, "loss": 0.4491, "mean_token_accuracy": 0.8828900456428528, "num_tokens": 676386113.0, "step": 17810 }, { "epoch": 2.2657422719755758, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.794952392578125, "learning_rate": 1e-05, "loss": 0.4261, "mean_token_accuracy": 0.8911404609680176, "num_tokens": 676421995.0, "step": 17811 }, { "epoch": 2.2658694822541663, "ewc_loss": 0.11083984375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.733977317810059, "learning_rate": 1e-05, "loss": 0.4575, "mean_token_accuracy": 0.876569926738739, "num_tokens": 676459761.0, "step": 17812 }, { "epoch": 2.265996692532757, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.00011205673217773438, "grad_norm": 7.150691509246826, "learning_rate": 1e-05, "loss": 0.5, "mean_token_accuracy": 0.8752728700637817, "num_tokens": 676503870.0, "step": 17813 }, { "epoch": 2.2661239028113473, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.663817882537842, "learning_rate": 1e-05, "loss": 0.4241, "mean_token_accuracy": 0.8892026543617249, "num_tokens": 676544173.0, "step": 17814 }, { "epoch": 2.266251113089938, "ewc_loss": 0.150390625, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.63800573348999, "learning_rate": 1e-05, "loss": 0.498, "mean_token_accuracy": 0.87847900390625, "num_tokens": 676580896.0, "step": 17815 }, { "epoch": 2.266378323368528, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 5.931649684906006, "learning_rate": 1e-05, "loss": 0.4497, "mean_token_accuracy": 0.882477879524231, "num_tokens": 676620127.0, "step": 17816 }, { "epoch": 2.266505533647119, "ewc_loss": 0.16015625, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.00013256072998046875, "grad_norm": 7.942659378051758, "learning_rate": 1e-05, "loss": 0.4862, "mean_token_accuracy": 0.8868018388748169, "num_tokens": 676655988.0, "step": 17817 }, { "epoch": 2.266632743925709, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.00011157989501953125, "grad_norm": 6.52916955947876, "learning_rate": 1e-05, "loss": 0.481, "mean_token_accuracy": 0.8767052292823792, "num_tokens": 676689125.0, "step": 17818 }, { "epoch": 2.2667599542042995, "ewc_loss": 0.154296875, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.00012683868408203125, "grad_norm": 7.676119327545166, "learning_rate": 1e-05, "loss": 0.5124, "mean_token_accuracy": 0.8747684955596924, "num_tokens": 676729231.0, "step": 17819 }, { "epoch": 2.26688716448289, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.00011444091796875, "grad_norm": 6.817399501800537, "learning_rate": 1e-05, "loss": 0.5127, "mean_token_accuracy": 0.8721659183502197, "num_tokens": 676770130.0, "step": 17820 }, { "epoch": 2.2670143747614806, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.00011205673217773438, "grad_norm": 7.174643039703369, "learning_rate": 1e-05, "loss": 0.4714, "mean_token_accuracy": 0.882291316986084, "num_tokens": 676807847.0, "step": 17821 }, { "epoch": 2.267141585040071, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 6.732584476470947, "learning_rate": 1e-05, "loss": 0.4621, "mean_token_accuracy": 0.8875025510787964, "num_tokens": 676841050.0, "step": 17822 }, { "epoch": 2.2672687953186617, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.767822265625, "learning_rate": 1e-05, "loss": 0.4814, "mean_token_accuracy": 0.8744840025901794, "num_tokens": 676877114.0, "step": 17823 }, { "epoch": 2.267396005597252, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.618694305419922, "learning_rate": 1e-05, "loss": 0.4251, "mean_token_accuracy": 0.8951181769371033, "num_tokens": 676917913.0, "step": 17824 }, { "epoch": 2.2675232158758427, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.45718240737915, "learning_rate": 1e-05, "loss": 0.426, "mean_token_accuracy": 0.8902791142463684, "num_tokens": 676955974.0, "step": 17825 }, { "epoch": 2.2676504261544332, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.525445461273193, "learning_rate": 1e-05, "loss": 0.4405, "mean_token_accuracy": 0.8888235688209534, "num_tokens": 676991285.0, "step": 17826 }, { "epoch": 2.2677776364330238, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.276567459106445, "learning_rate": 1e-05, "loss": 0.4963, "mean_token_accuracy": 0.8717472553253174, "num_tokens": 677025472.0, "step": 17827 }, { "epoch": 2.2679048467116143, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.457176208496094, "learning_rate": 1e-05, "loss": 0.4579, "mean_token_accuracy": 0.8827728629112244, "num_tokens": 677065047.0, "step": 17828 }, { "epoch": 2.268032056990205, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.17115592956543, "learning_rate": 1e-05, "loss": 0.4897, "mean_token_accuracy": 0.8729634881019592, "num_tokens": 677103027.0, "step": 17829 }, { "epoch": 2.2681592672687954, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.429022312164307, "learning_rate": 1e-05, "loss": 0.4327, "mean_token_accuracy": 0.89058518409729, "num_tokens": 677142302.0, "step": 17830 }, { "epoch": 2.268286477547386, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.063849449157715, "learning_rate": 1e-05, "loss": 0.4806, "mean_token_accuracy": 0.8734480142593384, "num_tokens": 677186963.0, "step": 17831 }, { "epoch": 2.2684136878259764, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.477640151977539, "learning_rate": 1e-05, "loss": 0.5375, "mean_token_accuracy": 0.8570985794067383, "num_tokens": 677232227.0, "step": 17832 }, { "epoch": 2.268540898104567, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 5.954174518585205, "learning_rate": 1e-05, "loss": 0.4187, "mean_token_accuracy": 0.895095705986023, "num_tokens": 677274213.0, "step": 17833 }, { "epoch": 2.2686681083831575, "ewc_loss": 0.125, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.613222599029541, "learning_rate": 1e-05, "loss": 0.4573, "mean_token_accuracy": 0.8835437297821045, "num_tokens": 677313870.0, "step": 17834 }, { "epoch": 2.268795318661748, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.7997589111328125, "learning_rate": 1e-05, "loss": 0.4412, "mean_token_accuracy": 0.8818075656890869, "num_tokens": 677350478.0, "step": 17835 }, { "epoch": 2.2689225289403385, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.890378952026367, "learning_rate": 1e-05, "loss": 0.4908, "mean_token_accuracy": 0.8727712631225586, "num_tokens": 677392253.0, "step": 17836 }, { "epoch": 2.269049739218929, "ewc_loss": 0.109375, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.625425815582275, "learning_rate": 1e-05, "loss": 0.44, "mean_token_accuracy": 0.886157751083374, "num_tokens": 677429024.0, "step": 17837 }, { "epoch": 2.2691769494975196, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.00011777877807617188, "grad_norm": 7.395211696624756, "learning_rate": 1e-05, "loss": 0.4521, "mean_token_accuracy": 0.8929510116577148, "num_tokens": 677461705.0, "step": 17838 }, { "epoch": 2.2693041597761097, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.633225917816162, "learning_rate": 1e-05, "loss": 0.4638, "mean_token_accuracy": 0.8782652020454407, "num_tokens": 677505586.0, "step": 17839 }, { "epoch": 2.2694313700547006, "ewc_loss": 0.162109375, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.00013446807861328125, "grad_norm": 8.00103759765625, "learning_rate": 1e-05, "loss": 0.5073, "mean_token_accuracy": 0.878243088722229, "num_tokens": 677544892.0, "step": 17840 }, { "epoch": 2.2695585803332907, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.224231243133545, "learning_rate": 1e-05, "loss": 0.4858, "mean_token_accuracy": 0.876174807548523, "num_tokens": 677586721.0, "step": 17841 }, { "epoch": 2.2696857906118812, "ewc_loss": 0.1630859375, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.0001354217529296875, "grad_norm": 8.098876953125, "learning_rate": 1e-05, "loss": 0.5429, "mean_token_accuracy": 0.8706939220428467, "num_tokens": 677627657.0, "step": 17842 }, { "epoch": 2.2698130008904718, "ewc_loss": 0.1474609375, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.00012063980102539062, "grad_norm": 6.890420913696289, "learning_rate": 1e-05, "loss": 0.4362, "mean_token_accuracy": 0.8968422412872314, "num_tokens": 677667513.0, "step": 17843 }, { "epoch": 2.2699402111690623, "ewc_loss": 0.150390625, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.599939823150635, "learning_rate": 1e-05, "loss": 0.4869, "mean_token_accuracy": 0.8805112838745117, "num_tokens": 677702120.0, "step": 17844 }, { "epoch": 2.270067421447653, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.0001163482666015625, "grad_norm": 6.944044589996338, "learning_rate": 1e-05, "loss": 0.5323, "mean_token_accuracy": 0.8663048148155212, "num_tokens": 677739353.0, "step": 17845 }, { "epoch": 2.2701946317262434, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.00011014938354492188, "grad_norm": 7.142796993255615, "learning_rate": 1e-05, "loss": 0.5327, "mean_token_accuracy": 0.8652997016906738, "num_tokens": 677778464.0, "step": 17846 }, { "epoch": 2.270321842004834, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 6.745674133300781, "learning_rate": 1e-05, "loss": 0.4885, "mean_token_accuracy": 0.8769218921661377, "num_tokens": 677815169.0, "step": 17847 }, { "epoch": 2.2704490522834244, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.7922773361206055, "learning_rate": 1e-05, "loss": 0.4404, "mean_token_accuracy": 0.8900958299636841, "num_tokens": 677852824.0, "step": 17848 }, { "epoch": 2.270576262562015, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.554562091827393, "learning_rate": 1e-05, "loss": 0.5022, "mean_token_accuracy": 0.8689302206039429, "num_tokens": 677895274.0, "step": 17849 }, { "epoch": 2.2707034728406055, "ewc_loss": 0.125, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.585397720336914, "learning_rate": 1e-05, "loss": 0.4424, "mean_token_accuracy": 0.8864569664001465, "num_tokens": 677929956.0, "step": 17850 }, { "epoch": 2.270830683119196, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.3925299644470215, "learning_rate": 1e-05, "loss": 0.4977, "mean_token_accuracy": 0.8703118562698364, "num_tokens": 677965921.0, "step": 17851 }, { "epoch": 2.2709578933977865, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.473968505859375, "learning_rate": 1e-05, "loss": 0.4376, "mean_token_accuracy": 0.8889697790145874, "num_tokens": 678005701.0, "step": 17852 }, { "epoch": 2.271085103676377, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.2520928382873535, "learning_rate": 1e-05, "loss": 0.4708, "mean_token_accuracy": 0.8761645555496216, "num_tokens": 678040275.0, "step": 17853 }, { "epoch": 2.2712123139549676, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.454553127288818, "learning_rate": 1e-05, "loss": 0.4602, "mean_token_accuracy": 0.8801102638244629, "num_tokens": 678071963.0, "step": 17854 }, { "epoch": 2.271339524233558, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.11210823059082, "learning_rate": 1e-05, "loss": 0.4517, "mean_token_accuracy": 0.876460075378418, "num_tokens": 678103422.0, "step": 17855 }, { "epoch": 2.2714667345121486, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.527181625366211, "learning_rate": 1e-05, "loss": 0.4799, "mean_token_accuracy": 0.8796284198760986, "num_tokens": 678144288.0, "step": 17856 }, { "epoch": 2.271593944790739, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.944394588470459, "learning_rate": 1e-05, "loss": 0.4125, "mean_token_accuracy": 0.8905260562896729, "num_tokens": 678175575.0, "step": 17857 }, { "epoch": 2.2717211550693297, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.722181797027588, "learning_rate": 1e-05, "loss": 0.511, "mean_token_accuracy": 0.865496814250946, "num_tokens": 678218089.0, "step": 17858 }, { "epoch": 2.2718483653479202, "ewc_loss": 0.109375, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.7631049156188965, "learning_rate": 1e-05, "loss": 0.4797, "mean_token_accuracy": 0.8708920478820801, "num_tokens": 678258777.0, "step": 17859 }, { "epoch": 2.2719755756265108, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.00010967254638671875, "grad_norm": 7.103309154510498, "learning_rate": 1e-05, "loss": 0.5058, "mean_token_accuracy": 0.8718855977058411, "num_tokens": 678300354.0, "step": 17860 }, { "epoch": 2.2721027859051013, "ewc_loss": 0.109375, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.637320041656494, "learning_rate": 1e-05, "loss": 0.4657, "mean_token_accuracy": 0.8743830919265747, "num_tokens": 678343579.0, "step": 17861 }, { "epoch": 2.272229996183692, "ewc_loss": 0.15234375, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.00012493133544921875, "grad_norm": 7.67678689956665, "learning_rate": 1e-05, "loss": 0.5095, "mean_token_accuracy": 0.8790327310562134, "num_tokens": 678377059.0, "step": 17862 }, { "epoch": 2.2723572064622823, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 5.920127868652344, "learning_rate": 1e-05, "loss": 0.4955, "mean_token_accuracy": 0.8711589574813843, "num_tokens": 678417608.0, "step": 17863 }, { "epoch": 2.2724844167408724, "ewc_loss": 0.1611328125, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.000133514404296875, "grad_norm": 8.067097663879395, "learning_rate": 1e-05, "loss": 0.5216, "mean_token_accuracy": 0.8764995336532593, "num_tokens": 678455473.0, "step": 17864 }, { "epoch": 2.2726116270194634, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.00011301040649414062, "grad_norm": 6.632664203643799, "learning_rate": 1e-05, "loss": 0.5442, "mean_token_accuracy": 0.8662224411964417, "num_tokens": 678500258.0, "step": 17865 }, { "epoch": 2.2727388372980535, "ewc_loss": 0.1533203125, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.000125885009765625, "grad_norm": 7.721038341522217, "learning_rate": 1e-05, "loss": 0.515, "mean_token_accuracy": 0.8738741874694824, "num_tokens": 678542219.0, "step": 17866 }, { "epoch": 2.272866047576644, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.00011587142944335938, "grad_norm": 6.946470260620117, "learning_rate": 1e-05, "loss": 0.493, "mean_token_accuracy": 0.8759307265281677, "num_tokens": 678579364.0, "step": 17867 }, { "epoch": 2.2729932578552345, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 7.166625022888184, "learning_rate": 1e-05, "loss": 0.5254, "mean_token_accuracy": 0.8647085428237915, "num_tokens": 678616549.0, "step": 17868 }, { "epoch": 2.273120468133825, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.00011157989501953125, "grad_norm": 6.830453872680664, "learning_rate": 1e-05, "loss": 0.4404, "mean_token_accuracy": 0.8924390077590942, "num_tokens": 678652324.0, "step": 17869 }, { "epoch": 2.2732476784124156, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.7042012214660645, "learning_rate": 1e-05, "loss": 0.4879, "mean_token_accuracy": 0.8741274476051331, "num_tokens": 678693574.0, "step": 17870 }, { "epoch": 2.273374888691006, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.715909957885742, "learning_rate": 1e-05, "loss": 0.5492, "mean_token_accuracy": 0.862140953540802, "num_tokens": 678730097.0, "step": 17871 }, { "epoch": 2.2735020989695967, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.395863056182861, "learning_rate": 1e-05, "loss": 0.4819, "mean_token_accuracy": 0.8764487504959106, "num_tokens": 678768392.0, "step": 17872 }, { "epoch": 2.273629309248187, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.64579439163208, "learning_rate": 1e-05, "loss": 0.51, "mean_token_accuracy": 0.8722609281539917, "num_tokens": 678804290.0, "step": 17873 }, { "epoch": 2.2737565195267777, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.182511329650879, "learning_rate": 1e-05, "loss": 0.4105, "mean_token_accuracy": 0.8965885639190674, "num_tokens": 678839535.0, "step": 17874 }, { "epoch": 2.2738837298053682, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.635574817657471, "learning_rate": 1e-05, "loss": 0.4643, "mean_token_accuracy": 0.8795392513275146, "num_tokens": 678879179.0, "step": 17875 }, { "epoch": 2.2740109400839588, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.039113521575928, "learning_rate": 1e-05, "loss": 0.4343, "mean_token_accuracy": 0.8872182369232178, "num_tokens": 678916804.0, "step": 17876 }, { "epoch": 2.2741381503625493, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.730472564697266, "learning_rate": 1e-05, "loss": 0.4809, "mean_token_accuracy": 0.8784680962562561, "num_tokens": 678950386.0, "step": 17877 }, { "epoch": 2.27426536064114, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.894443988800049, "learning_rate": 1e-05, "loss": 0.4551, "mean_token_accuracy": 0.8782764673233032, "num_tokens": 678988367.0, "step": 17878 }, { "epoch": 2.2743925709197303, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.921304225921631, "learning_rate": 1e-05, "loss": 0.4451, "mean_token_accuracy": 0.8888916373252869, "num_tokens": 679022577.0, "step": 17879 }, { "epoch": 2.274519781198321, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.7760725021362305, "learning_rate": 1e-05, "loss": 0.4607, "mean_token_accuracy": 0.8783487677574158, "num_tokens": 679059613.0, "step": 17880 }, { "epoch": 2.2746469914769114, "ewc_loss": 0.140625, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.00011396408081054688, "grad_norm": 7.247404098510742, "learning_rate": 1e-05, "loss": 0.4455, "mean_token_accuracy": 0.8908257484436035, "num_tokens": 679092956.0, "step": 17881 }, { "epoch": 2.274774201755502, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 5.7896199226379395, "learning_rate": 1e-05, "loss": 0.4652, "mean_token_accuracy": 0.8803070783615112, "num_tokens": 679131124.0, "step": 17882 }, { "epoch": 2.2749014120340925, "ewc_loss": 0.150390625, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.615207672119141, "learning_rate": 1e-05, "loss": 0.4517, "mean_token_accuracy": 0.8930791616439819, "num_tokens": 679166077.0, "step": 17883 }, { "epoch": 2.275028622312683, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.079246520996094, "learning_rate": 1e-05, "loss": 0.482, "mean_token_accuracy": 0.8776959776878357, "num_tokens": 679204427.0, "step": 17884 }, { "epoch": 2.2751558325912735, "ewc_loss": 0.154296875, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.00012683868408203125, "grad_norm": 7.729602336883545, "learning_rate": 1e-05, "loss": 0.4997, "mean_token_accuracy": 0.8804847002029419, "num_tokens": 679236003.0, "step": 17885 }, { "epoch": 2.275283042869864, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 6.491727352142334, "learning_rate": 1e-05, "loss": 0.4819, "mean_token_accuracy": 0.8793164491653442, "num_tokens": 679276347.0, "step": 17886 }, { "epoch": 2.2754102531484546, "ewc_loss": 0.1455078125, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.00011873245239257812, "grad_norm": 7.4420576095581055, "learning_rate": 1e-05, "loss": 0.4824, "mean_token_accuracy": 0.8822308778762817, "num_tokens": 679318335.0, "step": 17887 }, { "epoch": 2.275537463427045, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.00010967254638671875, "grad_norm": 6.652533054351807, "learning_rate": 1e-05, "loss": 0.5245, "mean_token_accuracy": 0.8633553981781006, "num_tokens": 679361253.0, "step": 17888 }, { "epoch": 2.275664673705635, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.00010967254638671875, "grad_norm": 7.023388385772705, "learning_rate": 1e-05, "loss": 0.5052, "mean_token_accuracy": 0.8719592094421387, "num_tokens": 679400234.0, "step": 17889 }, { "epoch": 2.275791883984226, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.620983123779297, "learning_rate": 1e-05, "loss": 0.4926, "mean_token_accuracy": 0.8764487504959106, "num_tokens": 679438538.0, "step": 17890 }, { "epoch": 2.2759190942628162, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.639983177185059, "learning_rate": 1e-05, "loss": 0.4838, "mean_token_accuracy": 0.8740243911743164, "num_tokens": 679475425.0, "step": 17891 }, { "epoch": 2.2760463045414068, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.551828861236572, "learning_rate": 1e-05, "loss": 0.4671, "mean_token_accuracy": 0.8826478719711304, "num_tokens": 679517447.0, "step": 17892 }, { "epoch": 2.2761735148199973, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.375311374664307, "learning_rate": 1e-05, "loss": 0.4578, "mean_token_accuracy": 0.879637598991394, "num_tokens": 679549933.0, "step": 17893 }, { "epoch": 2.276300725098588, "ewc_loss": 0.125, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.499937534332275, "learning_rate": 1e-05, "loss": 0.4569, "mean_token_accuracy": 0.8827623724937439, "num_tokens": 679590498.0, "step": 17894 }, { "epoch": 2.2764279353771784, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.200198173522949, "learning_rate": 1e-05, "loss": 0.4808, "mean_token_accuracy": 0.8752012252807617, "num_tokens": 679634582.0, "step": 17895 }, { "epoch": 2.276555145655769, "ewc_loss": 0.125, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.475790977478027, "learning_rate": 1e-05, "loss": 0.4546, "mean_token_accuracy": 0.882990300655365, "num_tokens": 679672324.0, "step": 17896 }, { "epoch": 2.2766823559343594, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.084085941314697, "learning_rate": 1e-05, "loss": 0.4512, "mean_token_accuracy": 0.8811980485916138, "num_tokens": 679706426.0, "step": 17897 }, { "epoch": 2.27680956621295, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.500558376312256, "learning_rate": 1e-05, "loss": 0.4267, "mean_token_accuracy": 0.8922551870346069, "num_tokens": 679744706.0, "step": 17898 }, { "epoch": 2.2769367764915405, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 5.9844255447387695, "learning_rate": 1e-05, "loss": 0.4619, "mean_token_accuracy": 0.8795639872550964, "num_tokens": 679779939.0, "step": 17899 }, { "epoch": 2.277063986770131, "ewc_loss": 0.125, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.5939459800720215, "learning_rate": 1e-05, "loss": 0.4679, "mean_token_accuracy": 0.8804643154144287, "num_tokens": 679820286.0, "step": 17900 }, { "epoch": 2.2771911970487215, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.848544597625732, "learning_rate": 1e-05, "loss": 0.4736, "mean_token_accuracy": 0.8717135190963745, "num_tokens": 679857985.0, "step": 17901 }, { "epoch": 2.277318407327312, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.811710357666016, "learning_rate": 1e-05, "loss": 0.4478, "mean_token_accuracy": 0.888908863067627, "num_tokens": 679896379.0, "step": 17902 }, { "epoch": 2.2774456176059026, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.693872928619385, "learning_rate": 1e-05, "loss": 0.4834, "mean_token_accuracy": 0.8696845769882202, "num_tokens": 679933162.0, "step": 17903 }, { "epoch": 2.277572827884493, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.00011444091796875, "grad_norm": 7.198693752288818, "learning_rate": 1e-05, "loss": 0.5338, "mean_token_accuracy": 0.8627906441688538, "num_tokens": 679975828.0, "step": 17904 }, { "epoch": 2.2777000381630836, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.650381565093994, "learning_rate": 1e-05, "loss": 0.5284, "mean_token_accuracy": 0.8553653955459595, "num_tokens": 680014678.0, "step": 17905 }, { "epoch": 2.277827248441674, "ewc_loss": 0.1552734375, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.0001277923583984375, "grad_norm": 7.707263469696045, "learning_rate": 1e-05, "loss": 0.4957, "mean_token_accuracy": 0.8784863352775574, "num_tokens": 680050912.0, "step": 17906 }, { "epoch": 2.2779544587202647, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 5.977570056915283, "learning_rate": 1e-05, "loss": 0.452, "mean_token_accuracy": 0.8839602470397949, "num_tokens": 680082591.0, "step": 17907 }, { "epoch": 2.2780816689988552, "ewc_loss": 0.162109375, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.00013446807861328125, "grad_norm": 7.973948955535889, "learning_rate": 1e-05, "loss": 0.5111, "mean_token_accuracy": 0.8758500814437866, "num_tokens": 680122100.0, "step": 17908 }, { "epoch": 2.2782088792774458, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.00011444091796875, "grad_norm": 6.577851295471191, "learning_rate": 1e-05, "loss": 0.5199, "mean_token_accuracy": 0.8682461977005005, "num_tokens": 680160926.0, "step": 17909 }, { "epoch": 2.2783360895560363, "ewc_loss": 0.1552734375, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.0001277923583984375, "grad_norm": 7.734319686889648, "learning_rate": 1e-05, "loss": 0.5098, "mean_token_accuracy": 0.8752579689025879, "num_tokens": 680202621.0, "step": 17910 }, { "epoch": 2.278463299834627, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.00011777877807617188, "grad_norm": 6.8620452880859375, "learning_rate": 1e-05, "loss": 0.4928, "mean_token_accuracy": 0.8773689270019531, "num_tokens": 680241129.0, "step": 17911 }, { "epoch": 2.2785905101132173, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.00011157989501953125, "grad_norm": 7.181647300720215, "learning_rate": 1e-05, "loss": 0.4816, "mean_token_accuracy": 0.8795064091682434, "num_tokens": 680279832.0, "step": 17912 }, { "epoch": 2.278717720391808, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.00011110305786132812, "grad_norm": 6.766323566436768, "learning_rate": 1e-05, "loss": 0.5231, "mean_token_accuracy": 0.8677018880844116, "num_tokens": 680313541.0, "step": 17913 }, { "epoch": 2.278844930670398, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.772590637207031, "learning_rate": 1e-05, "loss": 0.504, "mean_token_accuracy": 0.8697562217712402, "num_tokens": 680355756.0, "step": 17914 }, { "epoch": 2.278972140948989, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.627569675445557, "learning_rate": 1e-05, "loss": 0.4933, "mean_token_accuracy": 0.8728549480438232, "num_tokens": 680396276.0, "step": 17915 }, { "epoch": 2.279099351227579, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.4760212898254395, "learning_rate": 1e-05, "loss": 0.4995, "mean_token_accuracy": 0.8703868985176086, "num_tokens": 680435196.0, "step": 17916 }, { "epoch": 2.2792265615061695, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.546567440032959, "learning_rate": 1e-05, "loss": 0.4423, "mean_token_accuracy": 0.8900726437568665, "num_tokens": 680473088.0, "step": 17917 }, { "epoch": 2.27935377178476, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.294902324676514, "learning_rate": 1e-05, "loss": 0.4977, "mean_token_accuracy": 0.8706015348434448, "num_tokens": 680506269.0, "step": 17918 }, { "epoch": 2.2794809820633506, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.470499038696289, "learning_rate": 1e-05, "loss": 0.5276, "mean_token_accuracy": 0.8625040054321289, "num_tokens": 680543464.0, "step": 17919 }, { "epoch": 2.279608192341941, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.171835899353027, "learning_rate": 1e-05, "loss": 0.4791, "mean_token_accuracy": 0.8723781108856201, "num_tokens": 680582730.0, "step": 17920 }, { "epoch": 2.2797354026205316, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.448646068572998, "learning_rate": 1e-05, "loss": 0.4669, "mean_token_accuracy": 0.8800166845321655, "num_tokens": 680625773.0, "step": 17921 }, { "epoch": 2.279862612899122, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.069432258605957, "learning_rate": 1e-05, "loss": 0.469, "mean_token_accuracy": 0.8741140961647034, "num_tokens": 680655890.0, "step": 17922 }, { "epoch": 2.2799898231777127, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.496082782745361, "learning_rate": 1e-05, "loss": 0.467, "mean_token_accuracy": 0.8786990642547607, "num_tokens": 680689921.0, "step": 17923 }, { "epoch": 2.2801170334563032, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 5.928406238555908, "learning_rate": 1e-05, "loss": 0.4969, "mean_token_accuracy": 0.866324245929718, "num_tokens": 680735744.0, "step": 17924 }, { "epoch": 2.2802442437348938, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.640081405639648, "learning_rate": 1e-05, "loss": 0.4945, "mean_token_accuracy": 0.8746093511581421, "num_tokens": 680766925.0, "step": 17925 }, { "epoch": 2.2803714540134843, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.771377086639404, "learning_rate": 1e-05, "loss": 0.4476, "mean_token_accuracy": 0.8812760710716248, "num_tokens": 680804324.0, "step": 17926 }, { "epoch": 2.280498664292075, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.9575605392456055, "learning_rate": 1e-05, "loss": 0.4593, "mean_token_accuracy": 0.8851709365844727, "num_tokens": 680842005.0, "step": 17927 }, { "epoch": 2.2806258745706653, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.619015693664551, "learning_rate": 1e-05, "loss": 0.4349, "mean_token_accuracy": 0.8853706121444702, "num_tokens": 680873475.0, "step": 17928 }, { "epoch": 2.280753084849256, "ewc_loss": 0.1484375, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.00012159347534179688, "grad_norm": 7.475630760192871, "learning_rate": 1e-05, "loss": 0.4784, "mean_token_accuracy": 0.8824251294136047, "num_tokens": 680911322.0, "step": 17929 }, { "epoch": 2.2808802951278464, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 5.69758939743042, "learning_rate": 1e-05, "loss": 0.5272, "mean_token_accuracy": 0.8529185056686401, "num_tokens": 680949778.0, "step": 17930 }, { "epoch": 2.281007505406437, "ewc_loss": 0.1640625, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.00013637542724609375, "grad_norm": 8.02206039428711, "learning_rate": 1e-05, "loss": 0.5291, "mean_token_accuracy": 0.8731521368026733, "num_tokens": 680992795.0, "step": 17931 }, { "epoch": 2.2811347156850275, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 6.330929756164551, "learning_rate": 1e-05, "loss": 0.5215, "mean_token_accuracy": 0.8636171817779541, "num_tokens": 681026430.0, "step": 17932 }, { "epoch": 2.281261925963618, "ewc_loss": 0.1630859375, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.0001354217529296875, "grad_norm": 8.002528190612793, "learning_rate": 1e-05, "loss": 0.4039, "mean_token_accuracy": 0.9134242534637451, "num_tokens": 681062998.0, "step": 17933 }, { "epoch": 2.2813891362422085, "ewc_loss": 0.1474609375, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.00012063980102539062, "grad_norm": 6.8864874839782715, "learning_rate": 1e-05, "loss": 0.4967, "mean_token_accuracy": 0.8753976225852966, "num_tokens": 681100956.0, "step": 17934 }, { "epoch": 2.281516346520799, "ewc_loss": 0.1494140625, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.0001220703125, "grad_norm": 7.4971113204956055, "learning_rate": 1e-05, "loss": 0.4808, "mean_token_accuracy": 0.8830501437187195, "num_tokens": 681136406.0, "step": 17935 }, { "epoch": 2.2816435567993896, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.00011682510375976562, "grad_norm": 6.894660472869873, "learning_rate": 1e-05, "loss": 0.5324, "mean_token_accuracy": 0.8638655543327332, "num_tokens": 681174899.0, "step": 17936 }, { "epoch": 2.2817707670779797, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 7.042766571044922, "learning_rate": 1e-05, "loss": 0.5131, "mean_token_accuracy": 0.8706842660903931, "num_tokens": 681207662.0, "step": 17937 }, { "epoch": 2.2818979773565706, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 6.702565670013428, "learning_rate": 1e-05, "loss": 0.5161, "mean_token_accuracy": 0.870658814907074, "num_tokens": 681248819.0, "step": 17938 }, { "epoch": 2.2820251876351607, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.700678825378418, "learning_rate": 1e-05, "loss": 0.4758, "mean_token_accuracy": 0.8771077990531921, "num_tokens": 681286416.0, "step": 17939 }, { "epoch": 2.2821523979137512, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.55008602142334, "learning_rate": 1e-05, "loss": 0.5236, "mean_token_accuracy": 0.8601018190383911, "num_tokens": 681325629.0, "step": 17940 }, { "epoch": 2.2822796081923418, "ewc_loss": 0.125, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.501034259796143, "learning_rate": 1e-05, "loss": 0.5176, "mean_token_accuracy": 0.8624680638313293, "num_tokens": 681362310.0, "step": 17941 }, { "epoch": 2.2824068184709323, "ewc_loss": 0.125, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.413464069366455, "learning_rate": 1e-05, "loss": 0.475, "mean_token_accuracy": 0.8785665035247803, "num_tokens": 681401123.0, "step": 17942 }, { "epoch": 2.282534028749523, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.391780376434326, "learning_rate": 1e-05, "loss": 0.4695, "mean_token_accuracy": 0.8773350715637207, "num_tokens": 681439801.0, "step": 17943 }, { "epoch": 2.2826612390281134, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.305615425109863, "learning_rate": 1e-05, "loss": 0.4249, "mean_token_accuracy": 0.8902870416641235, "num_tokens": 681470032.0, "step": 17944 }, { "epoch": 2.282788449306704, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.335936069488525, "learning_rate": 1e-05, "loss": 0.4698, "mean_token_accuracy": 0.8834263682365417, "num_tokens": 681512421.0, "step": 17945 }, { "epoch": 2.2829156595852944, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.197566032409668, "learning_rate": 1e-05, "loss": 0.4824, "mean_token_accuracy": 0.876198947429657, "num_tokens": 681550226.0, "step": 17946 }, { "epoch": 2.283042869863885, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.336968898773193, "learning_rate": 1e-05, "loss": 0.4529, "mean_token_accuracy": 0.8825122714042664, "num_tokens": 681587885.0, "step": 17947 }, { "epoch": 2.2831700801424755, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.087741374969482, "learning_rate": 1e-05, "loss": 0.4149, "mean_token_accuracy": 0.892791211605072, "num_tokens": 681623230.0, "step": 17948 }, { "epoch": 2.283297290421066, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.4109392166137695, "learning_rate": 1e-05, "loss": 0.4511, "mean_token_accuracy": 0.8857496976852417, "num_tokens": 681659419.0, "step": 17949 }, { "epoch": 2.2834245006996565, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 5.959147930145264, "learning_rate": 1e-05, "loss": 0.4635, "mean_token_accuracy": 0.8756201267242432, "num_tokens": 681699845.0, "step": 17950 }, { "epoch": 2.283551710978247, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.5802202224731445, "learning_rate": 1e-05, "loss": 0.5117, "mean_token_accuracy": 0.8666720390319824, "num_tokens": 681737579.0, "step": 17951 }, { "epoch": 2.2836789212568376, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.7733635902404785, "learning_rate": 1e-05, "loss": 0.4816, "mean_token_accuracy": 0.8699285387992859, "num_tokens": 681771174.0, "step": 17952 }, { "epoch": 2.283806131535428, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.913547039031982, "learning_rate": 1e-05, "loss": 0.4949, "mean_token_accuracy": 0.8714659214019775, "num_tokens": 681810641.0, "step": 17953 }, { "epoch": 2.2839333418140186, "ewc_loss": 0.109375, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 8.20159912109375e-05, "grad_norm": 5.572704315185547, "learning_rate": 1e-05, "loss": 0.5019, "mean_token_accuracy": 0.8682926893234253, "num_tokens": 681849273.0, "step": 17954 }, { "epoch": 2.284060552092609, "ewc_loss": 0.1474609375, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.00012063980102539062, "grad_norm": 7.4974517822265625, "learning_rate": 1e-05, "loss": 0.4814, "mean_token_accuracy": 0.8842741250991821, "num_tokens": 681893105.0, "step": 17955 }, { "epoch": 2.2841877623711997, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.619771957397461, "learning_rate": 1e-05, "loss": 0.4446, "mean_token_accuracy": 0.8827359676361084, "num_tokens": 681931364.0, "step": 17956 }, { "epoch": 2.28431497264979, "ewc_loss": 0.166015625, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.00013828277587890625, "grad_norm": 8.183469772338867, "learning_rate": 1e-05, "loss": 0.5115, "mean_token_accuracy": 0.881630539894104, "num_tokens": 681970809.0, "step": 17957 }, { "epoch": 2.2844421829283807, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.00010967254638671875, "grad_norm": 6.398590564727783, "learning_rate": 1e-05, "loss": 0.4853, "mean_token_accuracy": 0.8779299855232239, "num_tokens": 682004581.0, "step": 17958 }, { "epoch": 2.2845693932069713, "ewc_loss": 0.1640625, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.00013637542724609375, "grad_norm": 8.17220401763916, "learning_rate": 1e-05, "loss": 0.5004, "mean_token_accuracy": 0.8827468752861023, "num_tokens": 682041809.0, "step": 17959 }, { "epoch": 2.284696603485562, "ewc_loss": 0.15234375, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.00012493133544921875, "grad_norm": 7.092587471008301, "learning_rate": 1e-05, "loss": 0.5284, "mean_token_accuracy": 0.8686147928237915, "num_tokens": 682078058.0, "step": 17960 }, { "epoch": 2.2848238137641523, "ewc_loss": 0.150390625, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.5523481369018555, "learning_rate": 1e-05, "loss": 0.4743, "mean_token_accuracy": 0.88567054271698, "num_tokens": 682111743.0, "step": 17961 }, { "epoch": 2.2849510240427424, "ewc_loss": 0.1455078125, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.000118255615234375, "grad_norm": 7.015214920043945, "learning_rate": 1e-05, "loss": 0.5032, "mean_token_accuracy": 0.8727110624313354, "num_tokens": 682147710.0, "step": 17962 }, { "epoch": 2.2850782343213334, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 7.091024875640869, "learning_rate": 1e-05, "loss": 0.5053, "mean_token_accuracy": 0.8715083599090576, "num_tokens": 682187139.0, "step": 17963 }, { "epoch": 2.2852054445999235, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 6.758403301239014, "learning_rate": 1e-05, "loss": 0.4687, "mean_token_accuracy": 0.8851868510246277, "num_tokens": 682226345.0, "step": 17964 }, { "epoch": 2.285332654878514, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.781941890716553, "learning_rate": 1e-05, "loss": 0.4688, "mean_token_accuracy": 0.8826979398727417, "num_tokens": 682262479.0, "step": 17965 }, { "epoch": 2.2854598651571045, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.552638053894043, "learning_rate": 1e-05, "loss": 0.4882, "mean_token_accuracy": 0.8733254671096802, "num_tokens": 682300062.0, "step": 17966 }, { "epoch": 2.285587075435695, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.6112060546875, "learning_rate": 1e-05, "loss": 0.4667, "mean_token_accuracy": 0.8816674947738647, "num_tokens": 682338075.0, "step": 17967 }, { "epoch": 2.2857142857142856, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.364894390106201, "learning_rate": 1e-05, "loss": 0.4752, "mean_token_accuracy": 0.8765144348144531, "num_tokens": 682373036.0, "step": 17968 }, { "epoch": 2.285841495992876, "ewc_loss": 0.125, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.546901702880859, "learning_rate": 1e-05, "loss": 0.5077, "mean_token_accuracy": 0.8659518957138062, "num_tokens": 682411131.0, "step": 17969 }, { "epoch": 2.2859687062714666, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.313393592834473, "learning_rate": 1e-05, "loss": 0.5354, "mean_token_accuracy": 0.8565230369567871, "num_tokens": 682449142.0, "step": 17970 }, { "epoch": 2.286095916550057, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.478123664855957, "learning_rate": 1e-05, "loss": 0.4178, "mean_token_accuracy": 0.8964713215827942, "num_tokens": 682486369.0, "step": 17971 }, { "epoch": 2.2862231268286477, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.095666885375977, "learning_rate": 1e-05, "loss": 0.4506, "mean_token_accuracy": 0.8803634643554688, "num_tokens": 682519756.0, "step": 17972 }, { "epoch": 2.2863503371072382, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.55388879776001, "learning_rate": 1e-05, "loss": 0.4798, "mean_token_accuracy": 0.8725565075874329, "num_tokens": 682559930.0, "step": 17973 }, { "epoch": 2.2864775473858288, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.91686487197876, "learning_rate": 1e-05, "loss": 0.4925, "mean_token_accuracy": 0.8690537214279175, "num_tokens": 682601379.0, "step": 17974 }, { "epoch": 2.2866047576644193, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.785899639129639, "learning_rate": 1e-05, "loss": 0.4769, "mean_token_accuracy": 0.8767194747924805, "num_tokens": 682639070.0, "step": 17975 }, { "epoch": 2.28673196794301, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.739259243011475, "learning_rate": 1e-05, "loss": 0.4398, "mean_token_accuracy": 0.8843774795532227, "num_tokens": 682675684.0, "step": 17976 }, { "epoch": 2.2868591782216003, "ewc_loss": 0.140625, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.00011348724365234375, "grad_norm": 7.202337265014648, "learning_rate": 1e-05, "loss": 0.4883, "mean_token_accuracy": 0.8813552856445312, "num_tokens": 682711026.0, "step": 17977 }, { "epoch": 2.286986388500191, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.665740013122559, "learning_rate": 1e-05, "loss": 0.4044, "mean_token_accuracy": 0.89157634973526, "num_tokens": 682743074.0, "step": 17978 }, { "epoch": 2.2871135987787814, "ewc_loss": 0.15625, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.00012874603271484375, "grad_norm": 7.768846035003662, "learning_rate": 1e-05, "loss": 0.4895, "mean_token_accuracy": 0.8802770376205444, "num_tokens": 682778875.0, "step": 17979 }, { "epoch": 2.287240809057372, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.037072658538818, "learning_rate": 1e-05, "loss": 0.4653, "mean_token_accuracy": 0.880555272102356, "num_tokens": 682817699.0, "step": 17980 }, { "epoch": 2.2873680193359625, "ewc_loss": 0.1611328125, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.000133514404296875, "grad_norm": 8.058368682861328, "learning_rate": 1e-05, "loss": 0.541, "mean_token_accuracy": 0.8677318692207336, "num_tokens": 682852742.0, "step": 17981 }, { "epoch": 2.287495229614553, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.00011539459228515625, "grad_norm": 6.7366719245910645, "learning_rate": 1e-05, "loss": 0.5067, "mean_token_accuracy": 0.8728536367416382, "num_tokens": 682894911.0, "step": 17982 }, { "epoch": 2.2876224398931435, "ewc_loss": 0.150390625, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.634649753570557, "learning_rate": 1e-05, "loss": 0.5139, "mean_token_accuracy": 0.8731347918510437, "num_tokens": 682932387.0, "step": 17983 }, { "epoch": 2.287749650171734, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.00011587142944335938, "grad_norm": 6.948363304138184, "learning_rate": 1e-05, "loss": 0.4762, "mean_token_accuracy": 0.8838323354721069, "num_tokens": 682965072.0, "step": 17984 }, { "epoch": 2.2878768604503246, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.00010967254638671875, "grad_norm": 7.090473651885986, "learning_rate": 1e-05, "loss": 0.4683, "mean_token_accuracy": 0.8822515606880188, "num_tokens": 683006008.0, "step": 17985 }, { "epoch": 2.288004070728915, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 6.816986083984375, "learning_rate": 1e-05, "loss": 0.4834, "mean_token_accuracy": 0.880003035068512, "num_tokens": 683042043.0, "step": 17986 }, { "epoch": 2.288131281007505, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.671173572540283, "learning_rate": 1e-05, "loss": 0.4847, "mean_token_accuracy": 0.874537467956543, "num_tokens": 683079429.0, "step": 17987 }, { "epoch": 2.288258491286096, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.7028889656066895, "learning_rate": 1e-05, "loss": 0.5078, "mean_token_accuracy": 0.869626522064209, "num_tokens": 683115686.0, "step": 17988 }, { "epoch": 2.2883857015646862, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.3601603507995605, "learning_rate": 1e-05, "loss": 0.443, "mean_token_accuracy": 0.887829601764679, "num_tokens": 683156715.0, "step": 17989 }, { "epoch": 2.2885129118432768, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.6392502784729, "learning_rate": 1e-05, "loss": 0.4717, "mean_token_accuracy": 0.8778171539306641, "num_tokens": 683191758.0, "step": 17990 }, { "epoch": 2.2886401221218673, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.177196025848389, "learning_rate": 1e-05, "loss": 0.4598, "mean_token_accuracy": 0.8778979778289795, "num_tokens": 683225146.0, "step": 17991 }, { "epoch": 2.288767332400458, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.653602600097656, "learning_rate": 1e-05, "loss": 0.4476, "mean_token_accuracy": 0.8877229690551758, "num_tokens": 683259552.0, "step": 17992 }, { "epoch": 2.2888945426790483, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 6.0261406898498535, "learning_rate": 1e-05, "loss": 0.4401, "mean_token_accuracy": 0.888835072517395, "num_tokens": 683296925.0, "step": 17993 }, { "epoch": 2.289021752957639, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.739604949951172, "learning_rate": 1e-05, "loss": 0.4711, "mean_token_accuracy": 0.8799522519111633, "num_tokens": 683328836.0, "step": 17994 }, { "epoch": 2.2891489632362294, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.884557723999023, "learning_rate": 1e-05, "loss": 0.4437, "mean_token_accuracy": 0.8844926953315735, "num_tokens": 683359306.0, "step": 17995 }, { "epoch": 2.28927617351482, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.947924613952637, "learning_rate": 1e-05, "loss": 0.4274, "mean_token_accuracy": 0.8950843811035156, "num_tokens": 683395743.0, "step": 17996 }, { "epoch": 2.2894033837934105, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.764122486114502, "learning_rate": 1e-05, "loss": 0.4467, "mean_token_accuracy": 0.881557047367096, "num_tokens": 683434528.0, "step": 17997 }, { "epoch": 2.289530594072001, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.00011587142944335938, "grad_norm": 7.307887554168701, "learning_rate": 1e-05, "loss": 0.5077, "mean_token_accuracy": 0.8715522289276123, "num_tokens": 683470540.0, "step": 17998 }, { "epoch": 2.2896578043505915, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 5.800290584564209, "learning_rate": 1e-05, "loss": 0.4749, "mean_token_accuracy": 0.8726742267608643, "num_tokens": 683502805.0, "step": 17999 }, { "epoch": 2.289785014629182, "ewc_loss": 0.15234375, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.00012493133544921875, "grad_norm": 7.689968585968018, "learning_rate": 1e-05, "loss": 0.5107, "mean_token_accuracy": 0.8781145215034485, "num_tokens": 683540306.0, "step": 18000 }, { "epoch": 2.2899122249077726, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.1481122970581055, "learning_rate": 1e-05, "loss": 0.4735, "mean_token_accuracy": 0.8762054443359375, "num_tokens": 683573089.0, "step": 18001 }, { "epoch": 2.290039435186363, "ewc_loss": 0.1552734375, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.0001277923583984375, "grad_norm": 7.754977226257324, "learning_rate": 1e-05, "loss": 0.5215, "mean_token_accuracy": 0.8722279667854309, "num_tokens": 683613540.0, "step": 18002 }, { "epoch": 2.2901666454649536, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 6.564650058746338, "learning_rate": 1e-05, "loss": 0.503, "mean_token_accuracy": 0.8733575344085693, "num_tokens": 683655016.0, "step": 18003 }, { "epoch": 2.290293855743544, "ewc_loss": 0.1455078125, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.000118255615234375, "grad_norm": 7.427896499633789, "learning_rate": 1e-05, "loss": 0.4542, "mean_token_accuracy": 0.889562726020813, "num_tokens": 683692607.0, "step": 18004 }, { "epoch": 2.2904210660221347, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.00011110305786132812, "grad_norm": 6.698232173919678, "learning_rate": 1e-05, "loss": 0.4704, "mean_token_accuracy": 0.8813412189483643, "num_tokens": 683726825.0, "step": 18005 }, { "epoch": 2.290548276300725, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 6.997585773468018, "learning_rate": 1e-05, "loss": 0.5293, "mean_token_accuracy": 0.8631039261817932, "num_tokens": 683763905.0, "step": 18006 }, { "epoch": 2.2906754865793157, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.643256664276123, "learning_rate": 1e-05, "loss": 0.518, "mean_token_accuracy": 0.8663673996925354, "num_tokens": 683806926.0, "step": 18007 }, { "epoch": 2.2908026968579063, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.61947774887085, "learning_rate": 1e-05, "loss": 0.4953, "mean_token_accuracy": 0.8730469346046448, "num_tokens": 683846069.0, "step": 18008 }, { "epoch": 2.290929907136497, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.575156211853027, "learning_rate": 1e-05, "loss": 0.4846, "mean_token_accuracy": 0.876189649105072, "num_tokens": 683887530.0, "step": 18009 }, { "epoch": 2.2910571174150873, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.35200834274292, "learning_rate": 1e-05, "loss": 0.4884, "mean_token_accuracy": 0.8731034994125366, "num_tokens": 683925721.0, "step": 18010 }, { "epoch": 2.291184327693678, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.514028549194336, "learning_rate": 1e-05, "loss": 0.4619, "mean_token_accuracy": 0.8812257647514343, "num_tokens": 683967867.0, "step": 18011 }, { "epoch": 2.291311537972268, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.189113140106201, "learning_rate": 1e-05, "loss": 0.4767, "mean_token_accuracy": 0.8754428029060364, "num_tokens": 684006360.0, "step": 18012 }, { "epoch": 2.291438748250859, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.494719505310059, "learning_rate": 1e-05, "loss": 0.4575, "mean_token_accuracy": 0.8840694427490234, "num_tokens": 684047559.0, "step": 18013 }, { "epoch": 2.291565958529449, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 6.074293613433838, "learning_rate": 1e-05, "loss": 0.4862, "mean_token_accuracy": 0.872088611125946, "num_tokens": 684092028.0, "step": 18014 }, { "epoch": 2.2916931688080395, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.5282883644104, "learning_rate": 1e-05, "loss": 0.497, "mean_token_accuracy": 0.868442952632904, "num_tokens": 684131441.0, "step": 18015 }, { "epoch": 2.29182037908663, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.9689741134643555, "learning_rate": 1e-05, "loss": 0.4499, "mean_token_accuracy": 0.8818219900131226, "num_tokens": 684163864.0, "step": 18016 }, { "epoch": 2.2919475893652206, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.624239444732666, "learning_rate": 1e-05, "loss": 0.4624, "mean_token_accuracy": 0.8792716264724731, "num_tokens": 684200962.0, "step": 18017 }, { "epoch": 2.292074799643811, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.825800895690918, "learning_rate": 1e-05, "loss": 0.4285, "mean_token_accuracy": 0.8877622485160828, "num_tokens": 684237060.0, "step": 18018 }, { "epoch": 2.2922020099224016, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.871273040771484, "learning_rate": 1e-05, "loss": 0.4689, "mean_token_accuracy": 0.8816595077514648, "num_tokens": 684271299.0, "step": 18019 }, { "epoch": 2.292329220200992, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.672711372375488, "learning_rate": 1e-05, "loss": 0.4862, "mean_token_accuracy": 0.8686139583587646, "num_tokens": 684310008.0, "step": 18020 }, { "epoch": 2.2924564304795827, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.00011587142944335938, "grad_norm": 7.2878007888793945, "learning_rate": 1e-05, "loss": 0.4885, "mean_token_accuracy": 0.8775906562805176, "num_tokens": 684348477.0, "step": 18021 }, { "epoch": 2.2925836407581732, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.679821014404297, "learning_rate": 1e-05, "loss": 0.4886, "mean_token_accuracy": 0.8678516149520874, "num_tokens": 684383090.0, "step": 18022 }, { "epoch": 2.2927108510367638, "ewc_loss": 0.1591796875, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.0001316070556640625, "grad_norm": 7.79143762588501, "learning_rate": 1e-05, "loss": 0.5649, "mean_token_accuracy": 0.8592806458473206, "num_tokens": 684415142.0, "step": 18023 }, { "epoch": 2.2928380613153543, "ewc_loss": 0.125, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.076435089111328, "learning_rate": 1e-05, "loss": 0.4743, "mean_token_accuracy": 0.8798607587814331, "num_tokens": 684455878.0, "step": 18024 }, { "epoch": 2.292965271593945, "ewc_loss": 0.1611328125, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.000133514404296875, "grad_norm": 7.968676567077637, "learning_rate": 1e-05, "loss": 0.542, "mean_token_accuracy": 0.8718398213386536, "num_tokens": 684501140.0, "step": 18025 }, { "epoch": 2.2930924818725353, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.00011587142944335938, "grad_norm": 6.680963516235352, "learning_rate": 1e-05, "loss": 0.4989, "mean_token_accuracy": 0.8729044795036316, "num_tokens": 684542682.0, "step": 18026 }, { "epoch": 2.293219692151126, "ewc_loss": 0.1513671875, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.0001239776611328125, "grad_norm": 7.593043804168701, "learning_rate": 1e-05, "loss": 0.523, "mean_token_accuracy": 0.8686699271202087, "num_tokens": 684581301.0, "step": 18027 }, { "epoch": 2.2933469024297164, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.00011491775512695312, "grad_norm": 6.8407440185546875, "learning_rate": 1e-05, "loss": 0.4622, "mean_token_accuracy": 0.8859031200408936, "num_tokens": 684620124.0, "step": 18028 }, { "epoch": 2.293474112708307, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.0001125335693359375, "grad_norm": 7.1033711433410645, "learning_rate": 1e-05, "loss": 0.5114, "mean_token_accuracy": 0.8697148561477661, "num_tokens": 684657237.0, "step": 18029 }, { "epoch": 2.2936013229868975, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.00010967254638671875, "grad_norm": 6.7263970375061035, "learning_rate": 1e-05, "loss": 0.4629, "mean_token_accuracy": 0.8840063214302063, "num_tokens": 684696717.0, "step": 18030 }, { "epoch": 2.293728533265488, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.716163158416748, "learning_rate": 1e-05, "loss": 0.4808, "mean_token_accuracy": 0.8763307332992554, "num_tokens": 684734823.0, "step": 18031 }, { "epoch": 2.2938557435440785, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.605656147003174, "learning_rate": 1e-05, "loss": 0.4858, "mean_token_accuracy": 0.8743396997451782, "num_tokens": 684773036.0, "step": 18032 }, { "epoch": 2.293982953822669, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.4322896003723145, "learning_rate": 1e-05, "loss": 0.4576, "mean_token_accuracy": 0.8821827173233032, "num_tokens": 684811210.0, "step": 18033 }, { "epoch": 2.2941101641012596, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.5217719078063965, "learning_rate": 1e-05, "loss": 0.4791, "mean_token_accuracy": 0.8779537677764893, "num_tokens": 684850656.0, "step": 18034 }, { "epoch": 2.2942373743798496, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.2689528465271, "learning_rate": 1e-05, "loss": 0.4832, "mean_token_accuracy": 0.8754961490631104, "num_tokens": 684890113.0, "step": 18035 }, { "epoch": 2.2943645846584406, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.457032680511475, "learning_rate": 1e-05, "loss": 0.4529, "mean_token_accuracy": 0.8826051354408264, "num_tokens": 684930418.0, "step": 18036 }, { "epoch": 2.2944917949370307, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.154279708862305, "learning_rate": 1e-05, "loss": 0.4801, "mean_token_accuracy": 0.8776166439056396, "num_tokens": 684967068.0, "step": 18037 }, { "epoch": 2.2946190052156212, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.454157829284668, "learning_rate": 1e-05, "loss": 0.5481, "mean_token_accuracy": 0.8542628288269043, "num_tokens": 685007602.0, "step": 18038 }, { "epoch": 2.2947462154942118, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.048630237579346, "learning_rate": 1e-05, "loss": 0.4628, "mean_token_accuracy": 0.8802977800369263, "num_tokens": 685043332.0, "step": 18039 }, { "epoch": 2.2948734257728023, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.519787788391113, "learning_rate": 1e-05, "loss": 0.4593, "mean_token_accuracy": 0.881401777267456, "num_tokens": 685078327.0, "step": 18040 }, { "epoch": 2.295000636051393, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.913323879241943, "learning_rate": 1e-05, "loss": 0.4632, "mean_token_accuracy": 0.8783211708068848, "num_tokens": 685119065.0, "step": 18041 }, { "epoch": 2.2951278463299833, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.699978351593018, "learning_rate": 1e-05, "loss": 0.5728, "mean_token_accuracy": 0.8519327044487, "num_tokens": 685154988.0, "step": 18042 }, { "epoch": 2.295255056608574, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.747034549713135, "learning_rate": 1e-05, "loss": 0.4639, "mean_token_accuracy": 0.8749622702598572, "num_tokens": 685194739.0, "step": 18043 }, { "epoch": 2.2953822668871644, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 7.02907133102417, "learning_rate": 1e-05, "loss": 0.5429, "mean_token_accuracy": 0.8636248111724854, "num_tokens": 685235619.0, "step": 18044 }, { "epoch": 2.295509477165755, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.604086875915527, "learning_rate": 1e-05, "loss": 0.4766, "mean_token_accuracy": 0.8725683689117432, "num_tokens": 685275189.0, "step": 18045 }, { "epoch": 2.2956366874443455, "ewc_loss": 0.15234375, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.00012493133544921875, "grad_norm": 7.578465461730957, "learning_rate": 1e-05, "loss": 0.5333, "mean_token_accuracy": 0.8695934414863586, "num_tokens": 685319928.0, "step": 18046 }, { "epoch": 2.295763897722936, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 5.760468482971191, "learning_rate": 1e-05, "loss": 0.448, "mean_token_accuracy": 0.8854706287384033, "num_tokens": 685355698.0, "step": 18047 }, { "epoch": 2.2958911080015265, "ewc_loss": 0.1640625, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.00013637542724609375, "grad_norm": 8.103225708007812, "learning_rate": 1e-05, "loss": 0.4958, "mean_token_accuracy": 0.87921541929245, "num_tokens": 685394873.0, "step": 18048 }, { "epoch": 2.296018318280117, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.0001125335693359375, "grad_norm": 6.48049783706665, "learning_rate": 1e-05, "loss": 0.4723, "mean_token_accuracy": 0.8821589350700378, "num_tokens": 685433896.0, "step": 18049 }, { "epoch": 2.2961455285587076, "ewc_loss": 0.16015625, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.00013256072998046875, "grad_norm": 7.948418617248535, "learning_rate": 1e-05, "loss": 0.5199, "mean_token_accuracy": 0.8738948106765747, "num_tokens": 685476036.0, "step": 18050 }, { "epoch": 2.296272738837298, "ewc_loss": 0.1494140625, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.0001220703125, "grad_norm": 6.9555864334106445, "learning_rate": 1e-05, "loss": 0.4936, "mean_token_accuracy": 0.8797354698181152, "num_tokens": 685514020.0, "step": 18051 }, { "epoch": 2.2963999491158886, "ewc_loss": 0.1455078125, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.000118255615234375, "grad_norm": 7.403736591339111, "learning_rate": 1e-05, "loss": 0.4727, "mean_token_accuracy": 0.8854383230209351, "num_tokens": 685547644.0, "step": 18052 }, { "epoch": 2.296527159394479, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.00011539459228515625, "grad_norm": 6.8777079582214355, "learning_rate": 1e-05, "loss": 0.5137, "mean_token_accuracy": 0.8705346584320068, "num_tokens": 685584147.0, "step": 18053 }, { "epoch": 2.2966543696730697, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 6.9618682861328125, "learning_rate": 1e-05, "loss": 0.4916, "mean_token_accuracy": 0.8777748942375183, "num_tokens": 685623093.0, "step": 18054 }, { "epoch": 2.29678157995166, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.6857075691223145, "learning_rate": 1e-05, "loss": 0.4774, "mean_token_accuracy": 0.8831479549407959, "num_tokens": 685657837.0, "step": 18055 }, { "epoch": 2.2969087902302507, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.6613287925720215, "learning_rate": 1e-05, "loss": 0.481, "mean_token_accuracy": 0.8779240846633911, "num_tokens": 685691077.0, "step": 18056 }, { "epoch": 2.2970360005088413, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.53813362121582, "learning_rate": 1e-05, "loss": 0.4878, "mean_token_accuracy": 0.8720642924308777, "num_tokens": 685725011.0, "step": 18057 }, { "epoch": 2.297163210787432, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.470339775085449, "learning_rate": 1e-05, "loss": 0.5284, "mean_token_accuracy": 0.8617785573005676, "num_tokens": 685764212.0, "step": 18058 }, { "epoch": 2.2972904210660223, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.395086765289307, "learning_rate": 1e-05, "loss": 0.5416, "mean_token_accuracy": 0.8561525344848633, "num_tokens": 685800780.0, "step": 18059 }, { "epoch": 2.2974176313446124, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.3718132972717285, "learning_rate": 1e-05, "loss": 0.4827, "mean_token_accuracy": 0.8754938840866089, "num_tokens": 685837728.0, "step": 18060 }, { "epoch": 2.2975448416232034, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.28924036026001, "learning_rate": 1e-05, "loss": 0.5184, "mean_token_accuracy": 0.8598983287811279, "num_tokens": 685869349.0, "step": 18061 }, { "epoch": 2.2976720519017935, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.324249744415283, "learning_rate": 1e-05, "loss": 0.449, "mean_token_accuracy": 0.8846081495285034, "num_tokens": 685908114.0, "step": 18062 }, { "epoch": 2.297799262180384, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.1748576164245605, "learning_rate": 1e-05, "loss": 0.5088, "mean_token_accuracy": 0.8669360876083374, "num_tokens": 685944980.0, "step": 18063 }, { "epoch": 2.2979264724589745, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.348467826843262, "learning_rate": 1e-05, "loss": 0.4549, "mean_token_accuracy": 0.883213460445404, "num_tokens": 685982729.0, "step": 18064 }, { "epoch": 2.298053682737565, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.06099271774292, "learning_rate": 1e-05, "loss": 0.4196, "mean_token_accuracy": 0.8938924074172974, "num_tokens": 686014090.0, "step": 18065 }, { "epoch": 2.2981808930161556, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.4386796951293945, "learning_rate": 1e-05, "loss": 0.4571, "mean_token_accuracy": 0.8829115629196167, "num_tokens": 686047345.0, "step": 18066 }, { "epoch": 2.298308103294746, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 5.915958404541016, "learning_rate": 1e-05, "loss": 0.5091, "mean_token_accuracy": 0.8691555261611938, "num_tokens": 686088662.0, "step": 18067 }, { "epoch": 2.2984353135733366, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.645944118499756, "learning_rate": 1e-05, "loss": 0.5129, "mean_token_accuracy": 0.8664804697036743, "num_tokens": 686124210.0, "step": 18068 }, { "epoch": 2.298562523851927, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.717956066131592, "learning_rate": 1e-05, "loss": 0.4596, "mean_token_accuracy": 0.8781307935714722, "num_tokens": 686157759.0, "step": 18069 }, { "epoch": 2.2986897341305177, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.00010967254638671875, "grad_norm": 7.047563076019287, "learning_rate": 1e-05, "loss": 0.5048, "mean_token_accuracy": 0.8709893822669983, "num_tokens": 686198486.0, "step": 18070 }, { "epoch": 2.298816944409108, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.536987781524658, "learning_rate": 1e-05, "loss": 0.5185, "mean_token_accuracy": 0.855707049369812, "num_tokens": 686235859.0, "step": 18071 }, { "epoch": 2.2989441546876987, "ewc_loss": 0.1533203125, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.000125885009765625, "grad_norm": 7.706625938415527, "learning_rate": 1e-05, "loss": 0.5006, "mean_token_accuracy": 0.8836067914962769, "num_tokens": 686274185.0, "step": 18072 }, { "epoch": 2.2990713649662893, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 5.752540588378906, "learning_rate": 1e-05, "loss": 0.426, "mean_token_accuracy": 0.8906469345092773, "num_tokens": 686308706.0, "step": 18073 }, { "epoch": 2.29919857524488, "ewc_loss": 0.169921875, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.00014209747314453125, "grad_norm": 8.322382926940918, "learning_rate": 1e-05, "loss": 0.4935, "mean_token_accuracy": 0.8875260353088379, "num_tokens": 686347871.0, "step": 18074 }, { "epoch": 2.2993257855234703, "ewc_loss": 0.1455078125, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.00011873245239257812, "grad_norm": 6.718687057495117, "learning_rate": 1e-05, "loss": 0.5276, "mean_token_accuracy": 0.8665000200271606, "num_tokens": 686395284.0, "step": 18075 }, { "epoch": 2.299452995802061, "ewc_loss": 0.162109375, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.00013446807861328125, "grad_norm": 8.006711959838867, "learning_rate": 1e-05, "loss": 0.551, "mean_token_accuracy": 0.8678920269012451, "num_tokens": 686436566.0, "step": 18076 }, { "epoch": 2.2995802060806514, "ewc_loss": 0.150390625, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.126464366912842, "learning_rate": 1e-05, "loss": 0.4816, "mean_token_accuracy": 0.8830390572547913, "num_tokens": 686477332.0, "step": 18077 }, { "epoch": 2.299707416359242, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.00011730194091796875, "grad_norm": 7.417236328125, "learning_rate": 1e-05, "loss": 0.4659, "mean_token_accuracy": 0.8895654678344727, "num_tokens": 686512000.0, "step": 18078 }, { "epoch": 2.2998346266378324, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.00011444091796875, "grad_norm": 6.949233531951904, "learning_rate": 1e-05, "loss": 0.4892, "mean_token_accuracy": 0.8787875175476074, "num_tokens": 686544732.0, "step": 18079 }, { "epoch": 2.299961836916423, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 7.020140647888184, "learning_rate": 1e-05, "loss": 0.4979, "mean_token_accuracy": 0.8771618008613586, "num_tokens": 686583100.0, "step": 18080 }, { "epoch": 2.3000890471950135, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.685775279998779, "learning_rate": 1e-05, "loss": 0.4852, "mean_token_accuracy": 0.8753794431686401, "num_tokens": 686620824.0, "step": 18081 }, { "epoch": 2.300216257473604, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.755695343017578, "learning_rate": 1e-05, "loss": 0.465, "mean_token_accuracy": 0.883305549621582, "num_tokens": 686660410.0, "step": 18082 }, { "epoch": 2.3003434677521946, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.47831392288208, "learning_rate": 1e-05, "loss": 0.4399, "mean_token_accuracy": 0.8884000778198242, "num_tokens": 686694385.0, "step": 18083 }, { "epoch": 2.300470678030785, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.617790699005127, "learning_rate": 1e-05, "loss": 0.4622, "mean_token_accuracy": 0.8811739683151245, "num_tokens": 686736844.0, "step": 18084 }, { "epoch": 2.300597888309375, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.277514934539795, "learning_rate": 1e-05, "loss": 0.4983, "mean_token_accuracy": 0.8702309727668762, "num_tokens": 686781104.0, "step": 18085 }, { "epoch": 2.300725098587966, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.613672256469727, "learning_rate": 1e-05, "loss": 0.4392, "mean_token_accuracy": 0.8907299041748047, "num_tokens": 686815244.0, "step": 18086 }, { "epoch": 2.3008523088665562, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 6.0916643142700195, "learning_rate": 1e-05, "loss": 0.4711, "mean_token_accuracy": 0.8760721683502197, "num_tokens": 686858133.0, "step": 18087 }, { "epoch": 2.3009795191451468, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.724482536315918, "learning_rate": 1e-05, "loss": 0.5027, "mean_token_accuracy": 0.8711904287338257, "num_tokens": 686901250.0, "step": 18088 }, { "epoch": 2.3011067294237373, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.898487567901611, "learning_rate": 1e-05, "loss": 0.448, "mean_token_accuracy": 0.8839830160140991, "num_tokens": 686936408.0, "step": 18089 }, { "epoch": 2.301233939702328, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 7.014460563659668, "learning_rate": 1e-05, "loss": 0.4875, "mean_token_accuracy": 0.8759762644767761, "num_tokens": 686981924.0, "step": 18090 }, { "epoch": 2.3013611499809183, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.760835647583008, "learning_rate": 1e-05, "loss": 0.471, "mean_token_accuracy": 0.8729588985443115, "num_tokens": 687016052.0, "step": 18091 }, { "epoch": 2.301488360259509, "ewc_loss": 0.1455078125, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.00011873245239257812, "grad_norm": 7.455819606781006, "learning_rate": 1e-05, "loss": 0.4832, "mean_token_accuracy": 0.8846174478530884, "num_tokens": 687059107.0, "step": 18092 }, { "epoch": 2.3016155705380994, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 5.86138391494751, "learning_rate": 1e-05, "loss": 0.4665, "mean_token_accuracy": 0.8819870352745056, "num_tokens": 687104015.0, "step": 18093 }, { "epoch": 2.30174278081669, "ewc_loss": 0.15625, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.00012874603271484375, "grad_norm": 7.8722076416015625, "learning_rate": 1e-05, "loss": 0.5253, "mean_token_accuracy": 0.870642900466919, "num_tokens": 687143719.0, "step": 18094 }, { "epoch": 2.3018699910952805, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.613810062408447, "learning_rate": 1e-05, "loss": 0.4819, "mean_token_accuracy": 0.8786355257034302, "num_tokens": 687179732.0, "step": 18095 }, { "epoch": 2.301997201373871, "ewc_loss": 0.1474609375, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.0001201629638671875, "grad_norm": 7.50960636138916, "learning_rate": 1e-05, "loss": 0.5247, "mean_token_accuracy": 0.8723300695419312, "num_tokens": 687222309.0, "step": 18096 }, { "epoch": 2.3021244116524615, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 6.66127347946167, "learning_rate": 1e-05, "loss": 0.4604, "mean_token_accuracy": 0.8855423927307129, "num_tokens": 687259114.0, "step": 18097 }, { "epoch": 2.302251621931052, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 0.00010967254638671875, "grad_norm": 7.073703765869141, "learning_rate": 1e-05, "loss": 0.4718, "mean_token_accuracy": 0.8817815780639648, "num_tokens": 687293052.0, "step": 18098 }, { "epoch": 2.3023788322096426, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 6.6786932945251465, "learning_rate": 1e-05, "loss": 0.4607, "mean_token_accuracy": 0.8844183683395386, "num_tokens": 687333245.0, "step": 18099 }, { "epoch": 2.302506042488233, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.629342555999756, "learning_rate": 1e-05, "loss": 0.4867, "mean_token_accuracy": 0.8745830059051514, "num_tokens": 687371740.0, "step": 18100 }, { "epoch": 2.3026332527668236, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.681112289428711, "learning_rate": 1e-05, "loss": 0.5444, "mean_token_accuracy": 0.8540414571762085, "num_tokens": 687407298.0, "step": 18101 }, { "epoch": 2.302760463045414, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.249912738800049, "learning_rate": 1e-05, "loss": 0.455, "mean_token_accuracy": 0.883718729019165, "num_tokens": 687447425.0, "step": 18102 }, { "epoch": 2.3028876733240047, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.736278057098389, "learning_rate": 1e-05, "loss": 0.484, "mean_token_accuracy": 0.8782525062561035, "num_tokens": 687488528.0, "step": 18103 }, { "epoch": 2.303014883602595, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 5.991340637207031, "learning_rate": 1e-05, "loss": 0.4627, "mean_token_accuracy": 0.8746204376220703, "num_tokens": 687521086.0, "step": 18104 }, { "epoch": 2.3031420938811857, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.912220001220703, "learning_rate": 1e-05, "loss": 0.4978, "mean_token_accuracy": 0.8702147006988525, "num_tokens": 687558882.0, "step": 18105 }, { "epoch": 2.3032693041597763, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.815184593200684, "learning_rate": 1e-05, "loss": 0.427, "mean_token_accuracy": 0.8900489211082458, "num_tokens": 687592058.0, "step": 18106 }, { "epoch": 2.303396514438367, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.00011444091796875, "grad_norm": 7.23069953918457, "learning_rate": 1e-05, "loss": 0.4971, "mean_token_accuracy": 0.8750891089439392, "num_tokens": 687629715.0, "step": 18107 }, { "epoch": 2.3035237247169573, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.785624027252197, "learning_rate": 1e-05, "loss": 0.4711, "mean_token_accuracy": 0.8739011883735657, "num_tokens": 687666201.0, "step": 18108 }, { "epoch": 2.303650934995548, "ewc_loss": 0.1513671875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.0001239776611328125, "grad_norm": 7.627511978149414, "learning_rate": 1e-05, "loss": 0.4922, "mean_token_accuracy": 0.8825681209564209, "num_tokens": 687706897.0, "step": 18109 }, { "epoch": 2.303778145274138, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.082138538360596, "learning_rate": 1e-05, "loss": 0.4689, "mean_token_accuracy": 0.879220187664032, "num_tokens": 687741449.0, "step": 18110 }, { "epoch": 2.303905355552729, "ewc_loss": 0.1552734375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.0001277923583984375, "grad_norm": 7.758908748626709, "learning_rate": 1e-05, "loss": 0.4773, "mean_token_accuracy": 0.8899316787719727, "num_tokens": 687776758.0, "step": 18111 }, { "epoch": 2.304032565831319, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 6.519354820251465, "learning_rate": 1e-05, "loss": 0.5264, "mean_token_accuracy": 0.8625997304916382, "num_tokens": 687814636.0, "step": 18112 }, { "epoch": 2.3041597761099095, "ewc_loss": 0.1484375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.00012159347534179688, "grad_norm": 7.507968902587891, "learning_rate": 1e-05, "loss": 0.536, "mean_token_accuracy": 0.8599022626876831, "num_tokens": 687843685.0, "step": 18113 }, { "epoch": 2.3042869863885, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.00011157989501953125, "grad_norm": 6.697809219360352, "learning_rate": 1e-05, "loss": 0.4683, "mean_token_accuracy": 0.8843507766723633, "num_tokens": 687875040.0, "step": 18114 }, { "epoch": 2.3044141966670906, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.00011205673217773438, "grad_norm": 7.078482151031494, "learning_rate": 1e-05, "loss": 0.5143, "mean_token_accuracy": 0.8729265928268433, "num_tokens": 687912952.0, "step": 18115 }, { "epoch": 2.304541406945681, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.63373327255249, "learning_rate": 1e-05, "loss": 0.4825, "mean_token_accuracy": 0.8780826926231384, "num_tokens": 687951366.0, "step": 18116 }, { "epoch": 2.3046686172242716, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.722132205963135, "learning_rate": 1e-05, "loss": 0.4749, "mean_token_accuracy": 0.8775656819343567, "num_tokens": 687993593.0, "step": 18117 }, { "epoch": 2.304795827502862, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.534904956817627, "learning_rate": 1e-05, "loss": 0.4903, "mean_token_accuracy": 0.874359130859375, "num_tokens": 688033449.0, "step": 18118 }, { "epoch": 2.3049230377814527, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.47713041305542, "learning_rate": 1e-05, "loss": 0.4816, "mean_token_accuracy": 0.8748786449432373, "num_tokens": 688073239.0, "step": 18119 }, { "epoch": 2.305050248060043, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.424948215484619, "learning_rate": 1e-05, "loss": 0.4936, "mean_token_accuracy": 0.8718700408935547, "num_tokens": 688112760.0, "step": 18120 }, { "epoch": 2.3051774583386337, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.369602203369141, "learning_rate": 1e-05, "loss": 0.5141, "mean_token_accuracy": 0.8675075769424438, "num_tokens": 688150553.0, "step": 18121 }, { "epoch": 2.3053046686172243, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.304596900939941, "learning_rate": 1e-05, "loss": 0.4775, "mean_token_accuracy": 0.8730658888816833, "num_tokens": 688190582.0, "step": 18122 }, { "epoch": 2.305431878895815, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.337341785430908, "learning_rate": 1e-05, "loss": 0.486, "mean_token_accuracy": 0.876031219959259, "num_tokens": 688226512.0, "step": 18123 }, { "epoch": 2.3055590891744053, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.182524681091309, "learning_rate": 1e-05, "loss": 0.4754, "mean_token_accuracy": 0.8763394355773926, "num_tokens": 688264891.0, "step": 18124 }, { "epoch": 2.305686299452996, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.382136344909668, "learning_rate": 1e-05, "loss": 0.4729, "mean_token_accuracy": 0.8751199841499329, "num_tokens": 688301280.0, "step": 18125 }, { "epoch": 2.3058135097315864, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 6.058159351348877, "learning_rate": 1e-05, "loss": 0.4648, "mean_token_accuracy": 0.8793255090713501, "num_tokens": 688337178.0, "step": 18126 }, { "epoch": 2.305940720010177, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.48691987991333, "learning_rate": 1e-05, "loss": 0.4511, "mean_token_accuracy": 0.884804368019104, "num_tokens": 688375839.0, "step": 18127 }, { "epoch": 2.3060679302887674, "ewc_loss": 0.11376953125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.887611389160156, "learning_rate": 1e-05, "loss": 0.4702, "mean_token_accuracy": 0.8725318312644958, "num_tokens": 688416395.0, "step": 18128 }, { "epoch": 2.306195140567358, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.739547252655029, "learning_rate": 1e-05, "loss": 0.4583, "mean_token_accuracy": 0.8872643709182739, "num_tokens": 688454801.0, "step": 18129 }, { "epoch": 2.3063223508459485, "ewc_loss": 0.10986328125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.691154956817627, "learning_rate": 1e-05, "loss": 0.4701, "mean_token_accuracy": 0.876448392868042, "num_tokens": 688489265.0, "step": 18130 }, { "epoch": 2.306449561124539, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.00011205673217773438, "grad_norm": 7.188845634460449, "learning_rate": 1e-05, "loss": 0.5126, "mean_token_accuracy": 0.8701695799827576, "num_tokens": 688525392.0, "step": 18131 }, { "epoch": 2.3065767714031296, "ewc_loss": 0.10986328125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.593214511871338, "learning_rate": 1e-05, "loss": 0.508, "mean_token_accuracy": 0.8596147298812866, "num_tokens": 688562389.0, "step": 18132 }, { "epoch": 2.3067039816817196, "ewc_loss": 0.15625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.00012874603271484375, "grad_norm": 7.813868522644043, "learning_rate": 1e-05, "loss": 0.5014, "mean_token_accuracy": 0.8784752488136292, "num_tokens": 688597728.0, "step": 18133 }, { "epoch": 2.3068311919603106, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 5.966848373413086, "learning_rate": 1e-05, "loss": 0.4872, "mean_token_accuracy": 0.8739498853683472, "num_tokens": 688633065.0, "step": 18134 }, { "epoch": 2.3069584022389007, "ewc_loss": 0.1640625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.00013637542724609375, "grad_norm": 8.174369812011719, "learning_rate": 1e-05, "loss": 0.5166, "mean_token_accuracy": 0.8785288333892822, "num_tokens": 688677095.0, "step": 18135 }, { "epoch": 2.3070856125174912, "ewc_loss": 0.1455078125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.00011873245239257812, "grad_norm": 6.766793251037598, "learning_rate": 1e-05, "loss": 0.4757, "mean_token_accuracy": 0.8810722827911377, "num_tokens": 688710421.0, "step": 18136 }, { "epoch": 2.3072128227960818, "ewc_loss": 0.154296875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.00012683868408203125, "grad_norm": 7.76724910736084, "learning_rate": 1e-05, "loss": 0.5026, "mean_token_accuracy": 0.8784432411193848, "num_tokens": 688747300.0, "step": 18137 }, { "epoch": 2.3073400330746723, "ewc_loss": 0.1474609375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.0001201629638671875, "grad_norm": 7.016255855560303, "learning_rate": 1e-05, "loss": 0.4788, "mean_token_accuracy": 0.8854057788848877, "num_tokens": 688781867.0, "step": 18138 }, { "epoch": 2.307467243353263, "ewc_loss": 0.140625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.00011348724365234375, "grad_norm": 7.214416980743408, "learning_rate": 1e-05, "loss": 0.4767, "mean_token_accuracy": 0.8794368505477905, "num_tokens": 688822777.0, "step": 18139 }, { "epoch": 2.3075944536318533, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.0001125335693359375, "grad_norm": 6.835343837738037, "learning_rate": 1e-05, "loss": 0.5157, "mean_token_accuracy": 0.8696805238723755, "num_tokens": 688862295.0, "step": 18140 }, { "epoch": 2.307721663910444, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.812378406524658, "learning_rate": 1e-05, "loss": 0.4415, "mean_token_accuracy": 0.8907389044761658, "num_tokens": 688904814.0, "step": 18141 }, { "epoch": 2.3078488741890344, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.657020092010498, "learning_rate": 1e-05, "loss": 0.5154, "mean_token_accuracy": 0.8669540882110596, "num_tokens": 688940403.0, "step": 18142 }, { "epoch": 2.307976084467625, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.564524173736572, "learning_rate": 1e-05, "loss": 0.5375, "mean_token_accuracy": 0.858374834060669, "num_tokens": 688980256.0, "step": 18143 }, { "epoch": 2.3081032947462155, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.502994060516357, "learning_rate": 1e-05, "loss": 0.5376, "mean_token_accuracy": 0.8592469692230225, "num_tokens": 689020629.0, "step": 18144 }, { "epoch": 2.308230505024806, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.428468227386475, "learning_rate": 1e-05, "loss": 0.5246, "mean_token_accuracy": 0.8614777326583862, "num_tokens": 689051257.0, "step": 18145 }, { "epoch": 2.3083577153033965, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.361164569854736, "learning_rate": 1e-05, "loss": 0.4644, "mean_token_accuracy": 0.8803375363349915, "num_tokens": 689083723.0, "step": 18146 }, { "epoch": 2.308484925581987, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.358282089233398, "learning_rate": 1e-05, "loss": 0.5346, "mean_token_accuracy": 0.8575628995895386, "num_tokens": 689123785.0, "step": 18147 }, { "epoch": 2.3086121358605776, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.243377685546875, "learning_rate": 1e-05, "loss": 0.4732, "mean_token_accuracy": 0.8731445074081421, "num_tokens": 689164053.0, "step": 18148 }, { "epoch": 2.308739346139168, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.3446173667907715, "learning_rate": 1e-05, "loss": 0.4639, "mean_token_accuracy": 0.8800818920135498, "num_tokens": 689198221.0, "step": 18149 }, { "epoch": 2.3088665564177586, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.127525806427002, "learning_rate": 1e-05, "loss": 0.4866, "mean_token_accuracy": 0.8742743134498596, "num_tokens": 689232330.0, "step": 18150 }, { "epoch": 2.308993766696349, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.3930816650390625, "learning_rate": 1e-05, "loss": 0.5089, "mean_token_accuracy": 0.8675622940063477, "num_tokens": 689277663.0, "step": 18151 }, { "epoch": 2.3091209769749397, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.010330677032471, "learning_rate": 1e-05, "loss": 0.4428, "mean_token_accuracy": 0.8845596313476562, "num_tokens": 689308563.0, "step": 18152 }, { "epoch": 2.30924818725353, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.519599914550781, "learning_rate": 1e-05, "loss": 0.4849, "mean_token_accuracy": 0.8742885589599609, "num_tokens": 689346897.0, "step": 18153 }, { "epoch": 2.3093753975321207, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.839334011077881, "learning_rate": 1e-05, "loss": 0.5024, "mean_token_accuracy": 0.866619884967804, "num_tokens": 689380689.0, "step": 18154 }, { "epoch": 2.3095026078107113, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.783292293548584, "learning_rate": 1e-05, "loss": 0.4713, "mean_token_accuracy": 0.8762506246566772, "num_tokens": 689418587.0, "step": 18155 }, { "epoch": 2.309629818089302, "ewc_loss": 0.10986328125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.6324076652526855, "learning_rate": 1e-05, "loss": 0.4573, "mean_token_accuracy": 0.8812267184257507, "num_tokens": 689455742.0, "step": 18156 }, { "epoch": 2.3097570283678923, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.00011587142944335938, "grad_norm": 7.292110919952393, "learning_rate": 1e-05, "loss": 0.4967, "mean_token_accuracy": 0.8758792877197266, "num_tokens": 689495618.0, "step": 18157 }, { "epoch": 2.3098842386464824, "ewc_loss": 0.1103515625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.558361053466797, "learning_rate": 1e-05, "loss": 0.5207, "mean_token_accuracy": 0.8566849231719971, "num_tokens": 689529856.0, "step": 18158 }, { "epoch": 2.3100114489250734, "ewc_loss": 0.162109375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.00013446807861328125, "grad_norm": 7.9881815910339355, "learning_rate": 1e-05, "loss": 0.5345, "mean_token_accuracy": 0.8728185296058655, "num_tokens": 689566785.0, "step": 18159 }, { "epoch": 2.3101386592036635, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.082629203796387, "learning_rate": 1e-05, "loss": 0.5027, "mean_token_accuracy": 0.8728982210159302, "num_tokens": 689604332.0, "step": 18160 }, { "epoch": 2.310265869482254, "ewc_loss": 0.16796875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.00014019012451171875, "grad_norm": 8.316859245300293, "learning_rate": 1e-05, "loss": 0.499, "mean_token_accuracy": 0.8834205269813538, "num_tokens": 689639159.0, "step": 18161 }, { "epoch": 2.3103930797608445, "ewc_loss": 0.1513671875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.0001239776611328125, "grad_norm": 6.972319602966309, "learning_rate": 1e-05, "loss": 0.5396, "mean_token_accuracy": 0.8681893348693848, "num_tokens": 689681660.0, "step": 18162 }, { "epoch": 2.310520290039435, "ewc_loss": 0.1552734375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.0001277923583984375, "grad_norm": 7.773593902587891, "learning_rate": 1e-05, "loss": 0.4862, "mean_token_accuracy": 0.8867865800857544, "num_tokens": 689717070.0, "step": 18163 }, { "epoch": 2.3106475003180256, "ewc_loss": 0.1484375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.00012159347534179688, "grad_norm": 7.099498748779297, "learning_rate": 1e-05, "loss": 0.4991, "mean_token_accuracy": 0.8789932727813721, "num_tokens": 689758885.0, "step": 18164 }, { "epoch": 2.310774710596616, "ewc_loss": 0.140625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.00011348724365234375, "grad_norm": 7.2613701820373535, "learning_rate": 1e-05, "loss": 0.4439, "mean_token_accuracy": 0.8909363746643066, "num_tokens": 689794943.0, "step": 18165 }, { "epoch": 2.3109019208752066, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.00011301040649414062, "grad_norm": 6.830122947692871, "learning_rate": 1e-05, "loss": 0.4817, "mean_token_accuracy": 0.8789147734642029, "num_tokens": 689836943.0, "step": 18166 }, { "epoch": 2.311029131153797, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 6.934138774871826, "learning_rate": 1e-05, "loss": 0.4894, "mean_token_accuracy": 0.8789392709732056, "num_tokens": 689870787.0, "step": 18167 }, { "epoch": 2.3111563414323877, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.585394859313965, "learning_rate": 1e-05, "loss": 0.4902, "mean_token_accuracy": 0.8737759590148926, "num_tokens": 689908302.0, "step": 18168 }, { "epoch": 2.311283551710978, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.7345476150512695, "learning_rate": 1e-05, "loss": 0.513, "mean_token_accuracy": 0.8698717355728149, "num_tokens": 689950031.0, "step": 18169 }, { "epoch": 2.3114107619895687, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.375728130340576, "learning_rate": 1e-05, "loss": 0.4729, "mean_token_accuracy": 0.8803775906562805, "num_tokens": 689994674.0, "step": 18170 }, { "epoch": 2.3115379722681593, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.655623435974121, "learning_rate": 1e-05, "loss": 0.4723, "mean_token_accuracy": 0.8812684416770935, "num_tokens": 690038697.0, "step": 18171 }, { "epoch": 2.31166518254675, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.195348739624023, "learning_rate": 1e-05, "loss": 0.5153, "mean_token_accuracy": 0.8715241551399231, "num_tokens": 690077449.0, "step": 18172 }, { "epoch": 2.3117923928253403, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.711480617523193, "learning_rate": 1e-05, "loss": 0.4877, "mean_token_accuracy": 0.8719050884246826, "num_tokens": 690113659.0, "step": 18173 }, { "epoch": 2.311919603103931, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 6.008113384246826, "learning_rate": 1e-05, "loss": 0.5055, "mean_token_accuracy": 0.8698657155036926, "num_tokens": 690153928.0, "step": 18174 }, { "epoch": 2.3120468133825214, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.884862422943115, "learning_rate": 1e-05, "loss": 0.4786, "mean_token_accuracy": 0.8790092468261719, "num_tokens": 690189526.0, "step": 18175 }, { "epoch": 2.312174023661112, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.848040580749512, "learning_rate": 1e-05, "loss": 0.4948, "mean_token_accuracy": 0.8724276423454285, "num_tokens": 690230220.0, "step": 18176 }, { "epoch": 2.3123012339397024, "ewc_loss": 0.140625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.00011396408081054688, "grad_norm": 7.241353511810303, "learning_rate": 1e-05, "loss": 0.5129, "mean_token_accuracy": 0.8741856813430786, "num_tokens": 690264998.0, "step": 18177 }, { "epoch": 2.312428444218293, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.817323207855225, "learning_rate": 1e-05, "loss": 0.3964, "mean_token_accuracy": 0.9019404649734497, "num_tokens": 690304323.0, "step": 18178 }, { "epoch": 2.3125556544968835, "ewc_loss": 0.15234375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.00012493133544921875, "grad_norm": 7.6626715660095215, "learning_rate": 1e-05, "loss": 0.5055, "mean_token_accuracy": 0.8763153553009033, "num_tokens": 690346832.0, "step": 18179 }, { "epoch": 2.312682864775474, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.149473667144775, "learning_rate": 1e-05, "loss": 0.4605, "mean_token_accuracy": 0.8800387382507324, "num_tokens": 690385829.0, "step": 18180 }, { "epoch": 2.3128100750540646, "ewc_loss": 0.154296875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.00012683868408203125, "grad_norm": 7.773869037628174, "learning_rate": 1e-05, "loss": 0.4994, "mean_token_accuracy": 0.8793250322341919, "num_tokens": 690429949.0, "step": 18181 }, { "epoch": 2.312937285332655, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 6.6366472244262695, "learning_rate": 1e-05, "loss": 0.4819, "mean_token_accuracy": 0.8798470497131348, "num_tokens": 690473252.0, "step": 18182 }, { "epoch": 2.313064495611245, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.00011682510375976562, "grad_norm": 7.376021385192871, "learning_rate": 1e-05, "loss": 0.4797, "mean_token_accuracy": 0.8826392889022827, "num_tokens": 690508765.0, "step": 18183 }, { "epoch": 2.313191705889836, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.00011014938354492188, "grad_norm": 6.812100410461426, "learning_rate": 1e-05, "loss": 0.4842, "mean_token_accuracy": 0.8768889307975769, "num_tokens": 690546749.0, "step": 18184 }, { "epoch": 2.313318916168426, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.816418647766113, "learning_rate": 1e-05, "loss": 0.4717, "mean_token_accuracy": 0.876578688621521, "num_tokens": 690579808.0, "step": 18185 }, { "epoch": 2.3134461264470167, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.810647487640381, "learning_rate": 1e-05, "loss": 0.4845, "mean_token_accuracy": 0.8752518892288208, "num_tokens": 690611667.0, "step": 18186 }, { "epoch": 2.3135733367256073, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.3125104904174805, "learning_rate": 1e-05, "loss": 0.4979, "mean_token_accuracy": 0.8680540323257446, "num_tokens": 690654991.0, "step": 18187 }, { "epoch": 2.313700547004198, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.895413398742676, "learning_rate": 1e-05, "loss": 0.5372, "mean_token_accuracy": 0.8631628751754761, "num_tokens": 690692344.0, "step": 18188 }, { "epoch": 2.3138277572827883, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.974649429321289, "learning_rate": 1e-05, "loss": 0.433, "mean_token_accuracy": 0.8845599293708801, "num_tokens": 690726812.0, "step": 18189 }, { "epoch": 2.313954967561379, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.00010967254638671875, "grad_norm": 7.1229939460754395, "learning_rate": 1e-05, "loss": 0.5062, "mean_token_accuracy": 0.872828483581543, "num_tokens": 690768229.0, "step": 18190 }, { "epoch": 2.3140821778399694, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.841327667236328, "learning_rate": 1e-05, "loss": 0.4574, "mean_token_accuracy": 0.8788528442382812, "num_tokens": 690810614.0, "step": 18191 }, { "epoch": 2.31420938811856, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.00011920928955078125, "grad_norm": 7.464780330657959, "learning_rate": 1e-05, "loss": 0.5534, "mean_token_accuracy": 0.863347053527832, "num_tokens": 690855422.0, "step": 18192 }, { "epoch": 2.3143365983971504, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 5.99169921875, "learning_rate": 1e-05, "loss": 0.4409, "mean_token_accuracy": 0.8850375413894653, "num_tokens": 690893751.0, "step": 18193 }, { "epoch": 2.314463808675741, "ewc_loss": 0.15234375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.00012493133544921875, "grad_norm": 7.694243431091309, "learning_rate": 1e-05, "loss": 0.5016, "mean_token_accuracy": 0.8785325288772583, "num_tokens": 690936734.0, "step": 18194 }, { "epoch": 2.3145910189543315, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.379908561706543, "learning_rate": 1e-05, "loss": 0.4731, "mean_token_accuracy": 0.8820866346359253, "num_tokens": 690975005.0, "step": 18195 }, { "epoch": 2.314718229232922, "ewc_loss": 0.1494140625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.0001220703125, "grad_norm": 7.554160118103027, "learning_rate": 1e-05, "loss": 0.4947, "mean_token_accuracy": 0.8803362846374512, "num_tokens": 691011607.0, "step": 18196 }, { "epoch": 2.3148454395115126, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 6.610551357269287, "learning_rate": 1e-05, "loss": 0.4767, "mean_token_accuracy": 0.8805165886878967, "num_tokens": 691047635.0, "step": 18197 }, { "epoch": 2.314972649790103, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.00011444091796875, "grad_norm": 7.203104496002197, "learning_rate": 1e-05, "loss": 0.5117, "mean_token_accuracy": 0.8712148666381836, "num_tokens": 691095290.0, "step": 18198 }, { "epoch": 2.3150998600686936, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.584805011749268, "learning_rate": 1e-05, "loss": 0.4811, "mean_token_accuracy": 0.8780735731124878, "num_tokens": 691140641.0, "step": 18199 }, { "epoch": 2.315227070347284, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.893678188323975, "learning_rate": 1e-05, "loss": 0.486, "mean_token_accuracy": 0.8762972354888916, "num_tokens": 691181744.0, "step": 18200 }, { "epoch": 2.3153542806258747, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.4736409187316895, "learning_rate": 1e-05, "loss": 0.4876, "mean_token_accuracy": 0.8769515752792358, "num_tokens": 691221279.0, "step": 18201 }, { "epoch": 2.315481490904465, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.705801010131836, "learning_rate": 1e-05, "loss": 0.4744, "mean_token_accuracy": 0.8822647333145142, "num_tokens": 691258519.0, "step": 18202 }, { "epoch": 2.3156087011830557, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.319234848022461, "learning_rate": 1e-05, "loss": 0.4917, "mean_token_accuracy": 0.872480571269989, "num_tokens": 691297063.0, "step": 18203 }, { "epoch": 2.3157359114616463, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.642232894897461, "learning_rate": 1e-05, "loss": 0.477, "mean_token_accuracy": 0.8778554797172546, "num_tokens": 691337146.0, "step": 18204 }, { "epoch": 2.315863121740237, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.1572089195251465, "learning_rate": 1e-05, "loss": 0.4489, "mean_token_accuracy": 0.8829047679901123, "num_tokens": 691373475.0, "step": 18205 }, { "epoch": 2.3159903320188273, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.707091808319092, "learning_rate": 1e-05, "loss": 0.4931, "mean_token_accuracy": 0.8737263679504395, "num_tokens": 691413286.0, "step": 18206 }, { "epoch": 2.316117542297418, "ewc_loss": 0.11376953125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.997626304626465, "learning_rate": 1e-05, "loss": 0.426, "mean_token_accuracy": 0.8900338411331177, "num_tokens": 691444950.0, "step": 18207 }, { "epoch": 2.316244752576008, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.890351295471191, "learning_rate": 1e-05, "loss": 0.4983, "mean_token_accuracy": 0.8735539317131042, "num_tokens": 691486767.0, "step": 18208 }, { "epoch": 2.316371962854599, "ewc_loss": 0.11181640625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.878464221954346, "learning_rate": 1e-05, "loss": 0.4746, "mean_token_accuracy": 0.8734137415885925, "num_tokens": 691527149.0, "step": 18209 }, { "epoch": 2.316499173133189, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.00011205673217773438, "grad_norm": 7.183407306671143, "learning_rate": 1e-05, "loss": 0.5117, "mean_token_accuracy": 0.8689000606536865, "num_tokens": 691563963.0, "step": 18210 }, { "epoch": 2.3166263834117795, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 5.902941703796387, "learning_rate": 1e-05, "loss": 0.4879, "mean_token_accuracy": 0.870351254940033, "num_tokens": 691596423.0, "step": 18211 }, { "epoch": 2.31675359369037, "ewc_loss": 0.1455078125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.000118255615234375, "grad_norm": 7.442215919494629, "learning_rate": 1e-05, "loss": 0.5164, "mean_token_accuracy": 0.874198317527771, "num_tokens": 691636928.0, "step": 18212 }, { "epoch": 2.3168808039689606, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.148698806762695, "learning_rate": 1e-05, "loss": 0.4766, "mean_token_accuracy": 0.8761948347091675, "num_tokens": 691673427.0, "step": 18213 }, { "epoch": 2.317008014247551, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.00011777877807617188, "grad_norm": 7.425604820251465, "learning_rate": 1e-05, "loss": 0.5122, "mean_token_accuracy": 0.8719271421432495, "num_tokens": 691710427.0, "step": 18214 }, { "epoch": 2.3171352245261416, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.461730003356934, "learning_rate": 1e-05, "loss": 0.5254, "mean_token_accuracy": 0.8677823543548584, "num_tokens": 691746809.0, "step": 18215 }, { "epoch": 2.317262434804732, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.00011014938354492188, "grad_norm": 7.055600643157959, "learning_rate": 1e-05, "loss": 0.4801, "mean_token_accuracy": 0.8812324404716492, "num_tokens": 691786037.0, "step": 18216 }, { "epoch": 2.3173896450833227, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.616955757141113, "learning_rate": 1e-05, "loss": 0.496, "mean_token_accuracy": 0.8723344802856445, "num_tokens": 691822270.0, "step": 18217 }, { "epoch": 2.317516855361913, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.524683475494385, "learning_rate": 1e-05, "loss": 0.4188, "mean_token_accuracy": 0.8964946269989014, "num_tokens": 691858538.0, "step": 18218 }, { "epoch": 2.3176440656405037, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.723271369934082, "learning_rate": 1e-05, "loss": 0.4692, "mean_token_accuracy": 0.8819577097892761, "num_tokens": 691898810.0, "step": 18219 }, { "epoch": 2.3177712759190943, "ewc_loss": 0.11376953125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 6.049877166748047, "learning_rate": 1e-05, "loss": 0.4668, "mean_token_accuracy": 0.878170371055603, "num_tokens": 691933708.0, "step": 18220 }, { "epoch": 2.317898486197685, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.9694294929504395, "learning_rate": 1e-05, "loss": 0.4545, "mean_token_accuracy": 0.8897369503974915, "num_tokens": 691968293.0, "step": 18221 }, { "epoch": 2.3180256964762753, "ewc_loss": 0.10986328125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.745848655700684, "learning_rate": 1e-05, "loss": 0.4973, "mean_token_accuracy": 0.8671735525131226, "num_tokens": 692004532.0, "step": 18222 }, { "epoch": 2.318152906754866, "ewc_loss": 0.1455078125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.000118255615234375, "grad_norm": 7.4098615646362305, "learning_rate": 1e-05, "loss": 0.4893, "mean_token_accuracy": 0.8784624934196472, "num_tokens": 692042323.0, "step": 18223 }, { "epoch": 2.3182801170334564, "ewc_loss": 0.11376953125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.783540725708008, "learning_rate": 1e-05, "loss": 0.4725, "mean_token_accuracy": 0.8747798204421997, "num_tokens": 692080970.0, "step": 18224 }, { "epoch": 2.318407327312047, "ewc_loss": 0.15625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.00012874603271484375, "grad_norm": 7.836611270904541, "learning_rate": 1e-05, "loss": 0.4983, "mean_token_accuracy": 0.8808664083480835, "num_tokens": 692115027.0, "step": 18225 }, { "epoch": 2.3185345375906374, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.298816204071045, "learning_rate": 1e-05, "loss": 0.49, "mean_token_accuracy": 0.8741859793663025, "num_tokens": 692146602.0, "step": 18226 }, { "epoch": 2.318661747869228, "ewc_loss": 0.1552734375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.0001277923583984375, "grad_norm": 7.761873245239258, "learning_rate": 1e-05, "loss": 0.5596, "mean_token_accuracy": 0.8611019849777222, "num_tokens": 692189721.0, "step": 18227 }, { "epoch": 2.3187889581478185, "ewc_loss": 0.140625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.00011396408081054688, "grad_norm": 6.7169928550720215, "learning_rate": 1e-05, "loss": 0.5269, "mean_token_accuracy": 0.8677877187728882, "num_tokens": 692233032.0, "step": 18228 }, { "epoch": 2.318916168426409, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.00011730194091796875, "grad_norm": 7.320147514343262, "learning_rate": 1e-05, "loss": 0.5313, "mean_token_accuracy": 0.8673050403594971, "num_tokens": 692269033.0, "step": 18229 }, { "epoch": 2.3190433787049995, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 6.741649150848389, "learning_rate": 1e-05, "loss": 0.5008, "mean_token_accuracy": 0.874546468257904, "num_tokens": 692307415.0, "step": 18230 }, { "epoch": 2.3191705889835896, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.898601055145264, "learning_rate": 1e-05, "loss": 0.4667, "mean_token_accuracy": 0.8829805254936218, "num_tokens": 692347003.0, "step": 18231 }, { "epoch": 2.3192977992621806, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.6060333251953125, "learning_rate": 1e-05, "loss": 0.4908, "mean_token_accuracy": 0.8759061098098755, "num_tokens": 692385141.0, "step": 18232 }, { "epoch": 2.3194250095407707, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.622363090515137, "learning_rate": 1e-05, "loss": 0.434, "mean_token_accuracy": 0.8902420997619629, "num_tokens": 692421731.0, "step": 18233 }, { "epoch": 2.319552219819361, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.436984539031982, "learning_rate": 1e-05, "loss": 0.4946, "mean_token_accuracy": 0.8757903575897217, "num_tokens": 692466400.0, "step": 18234 }, { "epoch": 2.3196794300979517, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.535928726196289, "learning_rate": 1e-05, "loss": 0.4263, "mean_token_accuracy": 0.8921729326248169, "num_tokens": 692501201.0, "step": 18235 }, { "epoch": 2.3198066403765423, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.000274658203125, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.242982387542725, "learning_rate": 1e-05, "loss": 0.4378, "mean_token_accuracy": 0.8901106119155884, "num_tokens": 692541549.0, "step": 18236 }, { "epoch": 2.319933850655133, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.57076358795166, "learning_rate": 1e-05, "loss": 0.4769, "mean_token_accuracy": 0.8827627897262573, "num_tokens": 692582156.0, "step": 18237 }, { "epoch": 2.3200610609337233, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 6.048752307891846, "learning_rate": 1e-05, "loss": 0.5065, "mean_token_accuracy": 0.8639609813690186, "num_tokens": 692620522.0, "step": 18238 }, { "epoch": 2.320188271212314, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.729312896728516, "learning_rate": 1e-05, "loss": 0.5026, "mean_token_accuracy": 0.8720160722732544, "num_tokens": 692658372.0, "step": 18239 }, { "epoch": 2.3203154814909044, "ewc_loss": 0.11181640625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.881004333496094, "learning_rate": 1e-05, "loss": 0.4418, "mean_token_accuracy": 0.8801953196525574, "num_tokens": 692687338.0, "step": 18240 }, { "epoch": 2.320442691769495, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 7.015804290771484, "learning_rate": 1e-05, "loss": 0.4955, "mean_token_accuracy": 0.8740643262863159, "num_tokens": 692724417.0, "step": 18241 }, { "epoch": 2.3205699020480854, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.75760555267334, "learning_rate": 1e-05, "loss": 0.4683, "mean_token_accuracy": 0.8762975335121155, "num_tokens": 692767881.0, "step": 18242 }, { "epoch": 2.320697112326676, "ewc_loss": 0.1455078125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.000118255615234375, "grad_norm": 7.451538562774658, "learning_rate": 1e-05, "loss": 0.5095, "mean_token_accuracy": 0.8722327351570129, "num_tokens": 692803154.0, "step": 18243 }, { "epoch": 2.3208243226052665, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 5.889344215393066, "learning_rate": 1e-05, "loss": 0.4215, "mean_token_accuracy": 0.8938204050064087, "num_tokens": 692842435.0, "step": 18244 }, { "epoch": 2.320951532883857, "ewc_loss": 0.15625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.00012874603271484375, "grad_norm": 7.7888569831848145, "learning_rate": 1e-05, "loss": 0.5401, "mean_token_accuracy": 0.8680185675621033, "num_tokens": 692884491.0, "step": 18245 }, { "epoch": 2.3210787431624476, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.382346153259277, "learning_rate": 1e-05, "loss": 0.4832, "mean_token_accuracy": 0.8717333674430847, "num_tokens": 692920825.0, "step": 18246 }, { "epoch": 2.321205953441038, "ewc_loss": 0.1513671875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.0001239776611328125, "grad_norm": 7.6211934089660645, "learning_rate": 1e-05, "loss": 0.4847, "mean_token_accuracy": 0.8831156492233276, "num_tokens": 692956413.0, "step": 18247 }, { "epoch": 2.3213331637196286, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.00011205673217773438, "grad_norm": 6.733603477478027, "learning_rate": 1e-05, "loss": 0.5287, "mean_token_accuracy": 0.8680109977722168, "num_tokens": 692995860.0, "step": 18248 }, { "epoch": 2.321460373998219, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.0001125335693359375, "grad_norm": 7.116372585296631, "learning_rate": 1e-05, "loss": 0.4613, "mean_token_accuracy": 0.8952167630195618, "num_tokens": 693031077.0, "step": 18249 }, { "epoch": 2.3215875842768097, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 6.772107124328613, "learning_rate": 1e-05, "loss": 0.5339, "mean_token_accuracy": 0.8640288710594177, "num_tokens": 693067181.0, "step": 18250 }, { "epoch": 2.3217147945554, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.600127696990967, "learning_rate": 1e-05, "loss": 0.4756, "mean_token_accuracy": 0.88059401512146, "num_tokens": 693103919.0, "step": 18251 }, { "epoch": 2.3218420048339907, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.7701029777526855, "learning_rate": 1e-05, "loss": 0.5293, "mean_token_accuracy": 0.8604221343994141, "num_tokens": 693138471.0, "step": 18252 }, { "epoch": 2.3219692151125813, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.2063117027282715, "learning_rate": 1e-05, "loss": 0.5057, "mean_token_accuracy": 0.8674753904342651, "num_tokens": 693174220.0, "step": 18253 }, { "epoch": 2.322096425391172, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.857546329498291, "learning_rate": 1e-05, "loss": 0.4698, "mean_token_accuracy": 0.8858889937400818, "num_tokens": 693217222.0, "step": 18254 }, { "epoch": 2.3222236356697623, "ewc_loss": 0.11376953125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.9474616050720215, "learning_rate": 1e-05, "loss": 0.4308, "mean_token_accuracy": 0.8898251056671143, "num_tokens": 693260052.0, "step": 18255 }, { "epoch": 2.3223508459483524, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.00011014938354492188, "grad_norm": 7.079255104064941, "learning_rate": 1e-05, "loss": 0.5161, "mean_token_accuracy": 0.865068793296814, "num_tokens": 693293775.0, "step": 18256 }, { "epoch": 2.3224780562269434, "ewc_loss": 0.11376953125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.861770153045654, "learning_rate": 1e-05, "loss": 0.4483, "mean_token_accuracy": 0.8808414936065674, "num_tokens": 693335876.0, "step": 18257 }, { "epoch": 2.3226052665055335, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.00011682510375976562, "grad_norm": 7.358721733093262, "learning_rate": 1e-05, "loss": 0.5239, "mean_token_accuracy": 0.8678336143493652, "num_tokens": 693376886.0, "step": 18258 }, { "epoch": 2.322732476784124, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 5.983543872833252, "learning_rate": 1e-05, "loss": 0.5126, "mean_token_accuracy": 0.8661177158355713, "num_tokens": 693416386.0, "step": 18259 }, { "epoch": 2.3228596870627145, "ewc_loss": 0.1474609375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.0001201629638671875, "grad_norm": 7.537670612335205, "learning_rate": 1e-05, "loss": 0.4639, "mean_token_accuracy": 0.8884047269821167, "num_tokens": 693454926.0, "step": 18260 }, { "epoch": 2.322986897341305, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.237316608428955, "learning_rate": 1e-05, "loss": 0.4892, "mean_token_accuracy": 0.874991238117218, "num_tokens": 693492926.0, "step": 18261 }, { "epoch": 2.3231141076198956, "ewc_loss": 0.1474609375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.0001201629638671875, "grad_norm": 7.516471862792969, "learning_rate": 1e-05, "loss": 0.5034, "mean_token_accuracy": 0.8789845705032349, "num_tokens": 693534684.0, "step": 18262 }, { "epoch": 2.323241317898486, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.467276096343994, "learning_rate": 1e-05, "loss": 0.5124, "mean_token_accuracy": 0.8646255731582642, "num_tokens": 693570856.0, "step": 18263 }, { "epoch": 2.3233685281770766, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.00011539459228515625, "grad_norm": 7.2327704429626465, "learning_rate": 1e-05, "loss": 0.5285, "mean_token_accuracy": 0.8652122020721436, "num_tokens": 693613286.0, "step": 18264 }, { "epoch": 2.323495738455667, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.517562389373779, "learning_rate": 1e-05, "loss": 0.4555, "mean_token_accuracy": 0.8841533660888672, "num_tokens": 693648748.0, "step": 18265 }, { "epoch": 2.3236229487342577, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.920694828033447, "learning_rate": 1e-05, "loss": 0.4436, "mean_token_accuracy": 0.8919095993041992, "num_tokens": 693688997.0, "step": 18266 }, { "epoch": 2.323750159012848, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.469825267791748, "learning_rate": 1e-05, "loss": 0.4952, "mean_token_accuracy": 0.8725613355636597, "num_tokens": 693727653.0, "step": 18267 }, { "epoch": 2.3238773692914387, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.664359092712402, "learning_rate": 1e-05, "loss": 0.4888, "mean_token_accuracy": 0.8733055591583252, "num_tokens": 693771146.0, "step": 18268 }, { "epoch": 2.3240045795700293, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.398342132568359, "learning_rate": 1e-05, "loss": 0.5633, "mean_token_accuracy": 0.8494271039962769, "num_tokens": 693809135.0, "step": 18269 }, { "epoch": 2.32413178984862, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.5122880935668945, "learning_rate": 1e-05, "loss": 0.4533, "mean_token_accuracy": 0.8849465250968933, "num_tokens": 693847127.0, "step": 18270 }, { "epoch": 2.3242590001272103, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.281589984893799, "learning_rate": 1e-05, "loss": 0.4673, "mean_token_accuracy": 0.8805687427520752, "num_tokens": 693887186.0, "step": 18271 }, { "epoch": 2.324386210405801, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.466210842132568, "learning_rate": 1e-05, "loss": 0.455, "mean_token_accuracy": 0.8853024244308472, "num_tokens": 693925943.0, "step": 18272 }, { "epoch": 2.3245134206843914, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.159454822540283, "learning_rate": 1e-05, "loss": 0.4179, "mean_token_accuracy": 0.893222987651825, "num_tokens": 693958825.0, "step": 18273 }, { "epoch": 2.324640630962982, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.511181354522705, "learning_rate": 1e-05, "loss": 0.5166, "mean_token_accuracy": 0.8640251159667969, "num_tokens": 694003787.0, "step": 18274 }, { "epoch": 2.3247678412415724, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 6.026319980621338, "learning_rate": 1e-05, "loss": 0.4775, "mean_token_accuracy": 0.8770260810852051, "num_tokens": 694047253.0, "step": 18275 }, { "epoch": 2.324895051520163, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.644961833953857, "learning_rate": 1e-05, "loss": 0.5036, "mean_token_accuracy": 0.8715088367462158, "num_tokens": 694085913.0, "step": 18276 }, { "epoch": 2.3250222617987535, "ewc_loss": 0.11181640625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.877008438110352, "learning_rate": 1e-05, "loss": 0.5106, "mean_token_accuracy": 0.8623509407043457, "num_tokens": 694122848.0, "step": 18277 }, { "epoch": 2.325149472077344, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.951592445373535, "learning_rate": 1e-05, "loss": 0.5, "mean_token_accuracy": 0.8699953556060791, "num_tokens": 694160545.0, "step": 18278 }, { "epoch": 2.3252766823559345, "ewc_loss": 0.11181640625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.745084285736084, "learning_rate": 1e-05, "loss": 0.4536, "mean_token_accuracy": 0.8776943683624268, "num_tokens": 694194272.0, "step": 18279 }, { "epoch": 2.325403892634525, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.00011491775512695312, "grad_norm": 7.329779148101807, "learning_rate": 1e-05, "loss": 0.4844, "mean_token_accuracy": 0.8803020715713501, "num_tokens": 694227164.0, "step": 18280 }, { "epoch": 2.325531102913115, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 5.806641101837158, "learning_rate": 1e-05, "loss": 0.4311, "mean_token_accuracy": 0.8891528844833374, "num_tokens": 694266589.0, "step": 18281 }, { "epoch": 2.325658313191706, "ewc_loss": 0.1533203125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.000125885009765625, "grad_norm": 7.698731899261475, "learning_rate": 1e-05, "loss": 0.5272, "mean_token_accuracy": 0.8680053353309631, "num_tokens": 694301558.0, "step": 18282 }, { "epoch": 2.325785523470296, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.223998546600342, "learning_rate": 1e-05, "loss": 0.5157, "mean_token_accuracy": 0.8640150427818298, "num_tokens": 694343332.0, "step": 18283 }, { "epoch": 2.3259127337488867, "ewc_loss": 0.15234375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.00012493133544921875, "grad_norm": 7.666523456573486, "learning_rate": 1e-05, "loss": 0.4679, "mean_token_accuracy": 0.8880212306976318, "num_tokens": 694383602.0, "step": 18284 }, { "epoch": 2.3260399440274773, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 6.6423234939575195, "learning_rate": 1e-05, "loss": 0.5255, "mean_token_accuracy": 0.8660755157470703, "num_tokens": 694417664.0, "step": 18285 }, { "epoch": 2.326167154306068, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.00011587142944335938, "grad_norm": 7.226931095123291, "learning_rate": 1e-05, "loss": 0.5432, "mean_token_accuracy": 0.8608826398849487, "num_tokens": 694456176.0, "step": 18286 }, { "epoch": 2.3262943645846583, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.00011014938354492188, "grad_norm": 6.714197635650635, "learning_rate": 1e-05, "loss": 0.5264, "mean_token_accuracy": 0.8707215785980225, "num_tokens": 694497114.0, "step": 18287 }, { "epoch": 2.326421574863249, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.719143867492676, "learning_rate": 1e-05, "loss": 0.4644, "mean_token_accuracy": 0.8826476335525513, "num_tokens": 694540832.0, "step": 18288 }, { "epoch": 2.3265487851418394, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.704935550689697, "learning_rate": 1e-05, "loss": 0.4776, "mean_token_accuracy": 0.8805938959121704, "num_tokens": 694578673.0, "step": 18289 }, { "epoch": 2.32667599542043, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.321288108825684, "learning_rate": 1e-05, "loss": 0.5592, "mean_token_accuracy": 0.8582069277763367, "num_tokens": 694624736.0, "step": 18290 }, { "epoch": 2.3268032056990204, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.766146659851074, "learning_rate": 1e-05, "loss": 0.4322, "mean_token_accuracy": 0.8927035927772522, "num_tokens": 694659123.0, "step": 18291 }, { "epoch": 2.326930415977611, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.04537296295166, "learning_rate": 1e-05, "loss": 0.4417, "mean_token_accuracy": 0.8878956437110901, "num_tokens": 694695310.0, "step": 18292 }, { "epoch": 2.3270576262562015, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.000274658203125, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.910130977630615, "learning_rate": 1e-05, "loss": 0.5022, "mean_token_accuracy": 0.8707101345062256, "num_tokens": 694732470.0, "step": 18293 }, { "epoch": 2.327184836534792, "ewc_loss": 0.11474609375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 5.900684356689453, "learning_rate": 1e-05, "loss": 0.4782, "mean_token_accuracy": 0.8725603818893433, "num_tokens": 694763693.0, "step": 18294 }, { "epoch": 2.3273120468133826, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.00011157989501953125, "grad_norm": 7.1477789878845215, "learning_rate": 1e-05, "loss": 0.4632, "mean_token_accuracy": 0.8871962428092957, "num_tokens": 694798127.0, "step": 18295 }, { "epoch": 2.327439257091973, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 5.882318496704102, "learning_rate": 1e-05, "loss": 0.4605, "mean_token_accuracy": 0.8789997696876526, "num_tokens": 694833462.0, "step": 18296 }, { "epoch": 2.3275664673705636, "ewc_loss": 0.1455078125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.000118255615234375, "grad_norm": 7.380760192871094, "learning_rate": 1e-05, "loss": 0.5188, "mean_token_accuracy": 0.8722715377807617, "num_tokens": 694870257.0, "step": 18297 }, { "epoch": 2.327693677649154, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.022653579711914, "learning_rate": 1e-05, "loss": 0.4876, "mean_token_accuracy": 0.8681284189224243, "num_tokens": 694910455.0, "step": 18298 }, { "epoch": 2.3278208879277447, "ewc_loss": 0.1484375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.00012159347534179688, "grad_norm": 7.500231742858887, "learning_rate": 1e-05, "loss": 0.4882, "mean_token_accuracy": 0.8781923055648804, "num_tokens": 694947040.0, "step": 18299 }, { "epoch": 2.327948098206335, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.261041164398193, "learning_rate": 1e-05, "loss": 0.4959, "mean_token_accuracy": 0.8725100755691528, "num_tokens": 694989043.0, "step": 18300 }, { "epoch": 2.3280753084849257, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.00011968612670898438, "grad_norm": 7.405082702636719, "learning_rate": 1e-05, "loss": 0.4629, "mean_token_accuracy": 0.890899658203125, "num_tokens": 695027094.0, "step": 18301 }, { "epoch": 2.3282025187635162, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.447721004486084, "learning_rate": 1e-05, "loss": 0.4884, "mean_token_accuracy": 0.8771237730979919, "num_tokens": 695056419.0, "step": 18302 }, { "epoch": 2.3283297290421068, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.00011205673217773438, "grad_norm": 7.125024318695068, "learning_rate": 1e-05, "loss": 0.5077, "mean_token_accuracy": 0.8735037446022034, "num_tokens": 695101430.0, "step": 18303 }, { "epoch": 2.3284569393206973, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.476268768310547, "learning_rate": 1e-05, "loss": 0.5221, "mean_token_accuracy": 0.8636676669120789, "num_tokens": 695141446.0, "step": 18304 }, { "epoch": 2.328584149599288, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.825878143310547, "learning_rate": 1e-05, "loss": 0.4865, "mean_token_accuracy": 0.8769014477729797, "num_tokens": 695179761.0, "step": 18305 }, { "epoch": 2.328711359877878, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.429060459136963, "learning_rate": 1e-05, "loss": 0.428, "mean_token_accuracy": 0.8966871500015259, "num_tokens": 695217224.0, "step": 18306 }, { "epoch": 2.328838570156469, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.586519241333008, "learning_rate": 1e-05, "loss": 0.4814, "mean_token_accuracy": 0.8747885823249817, "num_tokens": 695253481.0, "step": 18307 }, { "epoch": 2.328965780435059, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.364294528961182, "learning_rate": 1e-05, "loss": 0.4832, "mean_token_accuracy": 0.8775092363357544, "num_tokens": 695287002.0, "step": 18308 }, { "epoch": 2.3290929907136495, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.4520978927612305, "learning_rate": 1e-05, "loss": 0.455, "mean_token_accuracy": 0.8824693560600281, "num_tokens": 695328296.0, "step": 18309 }, { "epoch": 2.32922020099224, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.268248558044434, "learning_rate": 1e-05, "loss": 0.4688, "mean_token_accuracy": 0.8782474994659424, "num_tokens": 695364291.0, "step": 18310 }, { "epoch": 2.3293474112708306, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.40683126449585, "learning_rate": 1e-05, "loss": 0.5312, "mean_token_accuracy": 0.8583534955978394, "num_tokens": 695407263.0, "step": 18311 }, { "epoch": 2.329474621549421, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.163679599761963, "learning_rate": 1e-05, "loss": 0.5265, "mean_token_accuracy": 0.8589491248130798, "num_tokens": 695444243.0, "step": 18312 }, { "epoch": 2.3296018318280116, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.445985794067383, "learning_rate": 1e-05, "loss": 0.4431, "mean_token_accuracy": 0.8848643898963928, "num_tokens": 695475207.0, "step": 18313 }, { "epoch": 2.329729042106602, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 6.031978130340576, "learning_rate": 1e-05, "loss": 0.4491, "mean_token_accuracy": 0.8830113410949707, "num_tokens": 695515915.0, "step": 18314 }, { "epoch": 2.3298562523851927, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.561893463134766, "learning_rate": 1e-05, "loss": 0.4737, "mean_token_accuracy": 0.877913236618042, "num_tokens": 695555168.0, "step": 18315 }, { "epoch": 2.329983462663783, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.890017986297607, "learning_rate": 1e-05, "loss": 0.479, "mean_token_accuracy": 0.8749210834503174, "num_tokens": 695591829.0, "step": 18316 }, { "epoch": 2.3301106729423737, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.789002895355225, "learning_rate": 1e-05, "loss": 0.4936, "mean_token_accuracy": 0.8725051283836365, "num_tokens": 695628518.0, "step": 18317 }, { "epoch": 2.3302378832209643, "ewc_loss": 0.11083984375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.7424421310424805, "learning_rate": 1e-05, "loss": 0.4597, "mean_token_accuracy": 0.879705548286438, "num_tokens": 695661769.0, "step": 18318 }, { "epoch": 2.330365093499555, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.0001125335693359375, "grad_norm": 7.172839641571045, "learning_rate": 1e-05, "loss": 0.5522, "mean_token_accuracy": 0.8573144674301147, "num_tokens": 695699830.0, "step": 18319 }, { "epoch": 2.3304923037781453, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.680614948272705, "learning_rate": 1e-05, "loss": 0.4461, "mean_token_accuracy": 0.881135106086731, "num_tokens": 695738871.0, "step": 18320 }, { "epoch": 2.330619514056736, "ewc_loss": 0.1533203125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.000125885009765625, "grad_norm": 7.6770100593566895, "learning_rate": 1e-05, "loss": 0.5254, "mean_token_accuracy": 0.8684808611869812, "num_tokens": 695771868.0, "step": 18321 }, { "epoch": 2.3307467243353264, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.006115436553955, "learning_rate": 1e-05, "loss": 0.5074, "mean_token_accuracy": 0.8678792119026184, "num_tokens": 695808145.0, "step": 18322 }, { "epoch": 2.330873934613917, "ewc_loss": 0.1591796875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.0001316070556640625, "grad_norm": 7.90086030960083, "learning_rate": 1e-05, "loss": 0.4702, "mean_token_accuracy": 0.894713819026947, "num_tokens": 695837611.0, "step": 18323 }, { "epoch": 2.3310011448925074, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.0001125335693359375, "grad_norm": 6.571487903594971, "learning_rate": 1e-05, "loss": 0.4587, "mean_token_accuracy": 0.8898402452468872, "num_tokens": 695874640.0, "step": 18324 }, { "epoch": 2.331128355171098, "ewc_loss": 0.1513671875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.0001239776611328125, "grad_norm": 7.577757358551025, "learning_rate": 1e-05, "loss": 0.55, "mean_token_accuracy": 0.8647475838661194, "num_tokens": 695912580.0, "step": 18325 }, { "epoch": 2.3312555654496885, "ewc_loss": 0.140625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.00011396408081054688, "grad_norm": 6.7886271476745605, "learning_rate": 1e-05, "loss": 0.4682, "mean_token_accuracy": 0.8848123550415039, "num_tokens": 695952061.0, "step": 18326 }, { "epoch": 2.331382775728279, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.00011157989501953125, "grad_norm": 7.103363037109375, "learning_rate": 1e-05, "loss": 0.4551, "mean_token_accuracy": 0.8885459899902344, "num_tokens": 695986389.0, "step": 18327 }, { "epoch": 2.3315099860068695, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 6.704713344573975, "learning_rate": 1e-05, "loss": 0.474, "mean_token_accuracy": 0.8833192586898804, "num_tokens": 696026153.0, "step": 18328 }, { "epoch": 2.3316371962854596, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.694253444671631, "learning_rate": 1e-05, "loss": 0.5234, "mean_token_accuracy": 0.8613710403442383, "num_tokens": 696062693.0, "step": 18329 }, { "epoch": 2.3317644065640506, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.6206512451171875, "learning_rate": 1e-05, "loss": 0.4639, "mean_token_accuracy": 0.8813453912734985, "num_tokens": 696101933.0, "step": 18330 }, { "epoch": 2.3318916168426407, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.394253730773926, "learning_rate": 1e-05, "loss": 0.4957, "mean_token_accuracy": 0.8703348636627197, "num_tokens": 696137607.0, "step": 18331 }, { "epoch": 2.332018827121231, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.569096565246582, "learning_rate": 1e-05, "loss": 0.488, "mean_token_accuracy": 0.8771417737007141, "num_tokens": 696181277.0, "step": 18332 }, { "epoch": 2.3321460373998217, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.210745334625244, "learning_rate": 1e-05, "loss": 0.4888, "mean_token_accuracy": 0.8763535618782043, "num_tokens": 696216592.0, "step": 18333 }, { "epoch": 2.3322732476784123, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.56040096282959, "learning_rate": 1e-05, "loss": 0.4297, "mean_token_accuracy": 0.8927911520004272, "num_tokens": 696250762.0, "step": 18334 }, { "epoch": 2.332400457957003, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.0866899490356445, "learning_rate": 1e-05, "loss": 0.4646, "mean_token_accuracy": 0.875441312789917, "num_tokens": 696289819.0, "step": 18335 }, { "epoch": 2.3325276682355933, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.600943088531494, "learning_rate": 1e-05, "loss": 0.5587, "mean_token_accuracy": 0.850314736366272, "num_tokens": 696335551.0, "step": 18336 }, { "epoch": 2.332654878514184, "ewc_loss": 0.11474609375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 5.973991394042969, "learning_rate": 1e-05, "loss": 0.458, "mean_token_accuracy": 0.8809459209442139, "num_tokens": 696369973.0, "step": 18337 }, { "epoch": 2.3327820887927744, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.73253870010376, "learning_rate": 1e-05, "loss": 0.4928, "mean_token_accuracy": 0.8701446652412415, "num_tokens": 696403718.0, "step": 18338 }, { "epoch": 2.332909299071365, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.837143898010254, "learning_rate": 1e-05, "loss": 0.4665, "mean_token_accuracy": 0.8724774122238159, "num_tokens": 696440769.0, "step": 18339 }, { "epoch": 2.3330365093499554, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.96425724029541, "learning_rate": 1e-05, "loss": 0.4841, "mean_token_accuracy": 0.8753024339675903, "num_tokens": 696482570.0, "step": 18340 }, { "epoch": 2.333163719628546, "ewc_loss": 0.11181640625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.731049060821533, "learning_rate": 1e-05, "loss": 0.4253, "mean_token_accuracy": 0.8893387317657471, "num_tokens": 696519199.0, "step": 18341 }, { "epoch": 2.3332909299071365, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.00011682510375976562, "grad_norm": 7.340640544891357, "learning_rate": 1e-05, "loss": 0.5383, "mean_token_accuracy": 0.8613290786743164, "num_tokens": 696558460.0, "step": 18342 }, { "epoch": 2.333418140185727, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 5.7918477058410645, "learning_rate": 1e-05, "loss": 0.5583, "mean_token_accuracy": 0.8491268157958984, "num_tokens": 696600395.0, "step": 18343 }, { "epoch": 2.3335453504643175, "ewc_loss": 0.154296875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.00012683868408203125, "grad_norm": 7.724238395690918, "learning_rate": 1e-05, "loss": 0.5199, "mean_token_accuracy": 0.874189019203186, "num_tokens": 696631478.0, "step": 18344 }, { "epoch": 2.333672560742908, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.151978969573975, "learning_rate": 1e-05, "loss": 0.4703, "mean_token_accuracy": 0.8816210031509399, "num_tokens": 696669634.0, "step": 18345 }, { "epoch": 2.3337997710214986, "ewc_loss": 0.15625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.00012874603271484375, "grad_norm": 7.784774303436279, "learning_rate": 1e-05, "loss": 0.5268, "mean_token_accuracy": 0.8692387342453003, "num_tokens": 696707642.0, "step": 18346 }, { "epoch": 2.333926981300089, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.00011205673217773438, "grad_norm": 6.56604528427124, "learning_rate": 1e-05, "loss": 0.4732, "mean_token_accuracy": 0.8825860023498535, "num_tokens": 696747426.0, "step": 18347 }, { "epoch": 2.3340541915786797, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.00011968612670898438, "grad_norm": 7.448044776916504, "learning_rate": 1e-05, "loss": 0.5422, "mean_token_accuracy": 0.8635004758834839, "num_tokens": 696784114.0, "step": 18348 }, { "epoch": 2.33418140185727, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.00011205673217773438, "grad_norm": 6.809324741363525, "learning_rate": 1e-05, "loss": 0.466, "mean_token_accuracy": 0.8838623762130737, "num_tokens": 696822303.0, "step": 18349 }, { "epoch": 2.3343086121358607, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 6.942300796508789, "learning_rate": 1e-05, "loss": 0.5073, "mean_token_accuracy": 0.8746768236160278, "num_tokens": 696855460.0, "step": 18350 }, { "epoch": 2.3344358224144512, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 6.656878471374512, "learning_rate": 1e-05, "loss": 0.4181, "mean_token_accuracy": 0.8997926115989685, "num_tokens": 696891070.0, "step": 18351 }, { "epoch": 2.3345630326930418, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.558813571929932, "learning_rate": 1e-05, "loss": 0.4676, "mean_token_accuracy": 0.8775120973587036, "num_tokens": 696924679.0, "step": 18352 }, { "epoch": 2.3346902429716323, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.597340106964111, "learning_rate": 1e-05, "loss": 0.4637, "mean_token_accuracy": 0.8810463547706604, "num_tokens": 696958894.0, "step": 18353 }, { "epoch": 2.3348174532502224, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.30393123626709, "learning_rate": 1e-05, "loss": 0.4899, "mean_token_accuracy": 0.8705677390098572, "num_tokens": 696995371.0, "step": 18354 }, { "epoch": 2.3349446635288134, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.568606376647949, "learning_rate": 1e-05, "loss": 0.4422, "mean_token_accuracy": 0.8888710141181946, "num_tokens": 697034284.0, "step": 18355 }, { "epoch": 2.3350718738074034, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.145336627960205, "learning_rate": 1e-05, "loss": 0.436, "mean_token_accuracy": 0.8900991678237915, "num_tokens": 697069805.0, "step": 18356 }, { "epoch": 2.335199084085994, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.582573413848877, "learning_rate": 1e-05, "loss": 0.4723, "mean_token_accuracy": 0.878130316734314, "num_tokens": 697105326.0, "step": 18357 }, { "epoch": 2.3353262943645845, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.02783203125, "learning_rate": 1e-05, "loss": 0.4612, "mean_token_accuracy": 0.8802089095115662, "num_tokens": 697144157.0, "step": 18358 }, { "epoch": 2.335453504643175, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.654872894287109, "learning_rate": 1e-05, "loss": 0.4552, "mean_token_accuracy": 0.885857880115509, "num_tokens": 697178261.0, "step": 18359 }, { "epoch": 2.3355807149217656, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 5.896520137786865, "learning_rate": 1e-05, "loss": 0.4524, "mean_token_accuracy": 0.8809635043144226, "num_tokens": 697224044.0, "step": 18360 }, { "epoch": 2.335707925200356, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.859429359436035, "learning_rate": 1e-05, "loss": 0.4749, "mean_token_accuracy": 0.8814138174057007, "num_tokens": 697262007.0, "step": 18361 }, { "epoch": 2.3358351354789466, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.758001804351807, "learning_rate": 1e-05, "loss": 0.4436, "mean_token_accuracy": 0.8843714594841003, "num_tokens": 697301689.0, "step": 18362 }, { "epoch": 2.335962345757537, "ewc_loss": 0.140625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.00011396408081054688, "grad_norm": 7.193150997161865, "learning_rate": 1e-05, "loss": 0.4807, "mean_token_accuracy": 0.8803710341453552, "num_tokens": 697343509.0, "step": 18363 }, { "epoch": 2.3360895560361277, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.719370365142822, "learning_rate": 1e-05, "loss": 0.4624, "mean_token_accuracy": 0.8785507678985596, "num_tokens": 697380620.0, "step": 18364 }, { "epoch": 2.336216766314718, "ewc_loss": 0.15234375, "ewc_loss_diag": 0.000274658203125, "ewc_loss_parallel": 0.00012493133544921875, "grad_norm": 7.637364864349365, "learning_rate": 1e-05, "loss": 0.5193, "mean_token_accuracy": 0.8715022802352905, "num_tokens": 697422811.0, "step": 18365 }, { "epoch": 2.3363439765933087, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 5.991260051727295, "learning_rate": 1e-05, "loss": 0.4879, "mean_token_accuracy": 0.8730076551437378, "num_tokens": 697460527.0, "step": 18366 }, { "epoch": 2.3364711868718993, "ewc_loss": 0.1591796875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.0001316070556640625, "grad_norm": 7.91971492767334, "learning_rate": 1e-05, "loss": 0.4999, "mean_token_accuracy": 0.8818902969360352, "num_tokens": 697495417.0, "step": 18367 }, { "epoch": 2.33659839715049, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 0.00011157989501953125, "grad_norm": 6.5144195556640625, "learning_rate": 1e-05, "loss": 0.4711, "mean_token_accuracy": 0.8824474811553955, "num_tokens": 697534745.0, "step": 18368 }, { "epoch": 2.3367256074290803, "ewc_loss": 0.154296875, "ewc_loss_diag": 0.000274658203125, "ewc_loss_parallel": 0.00012683868408203125, "grad_norm": 7.631469249725342, "learning_rate": 1e-05, "loss": 0.5096, "mean_token_accuracy": 0.8772009611129761, "num_tokens": 697572074.0, "step": 18369 }, { "epoch": 2.336852817707671, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.000274658203125, "ewc_loss_parallel": 0.00011539459228515625, "grad_norm": 6.771509647369385, "learning_rate": 1e-05, "loss": 0.5069, "mean_token_accuracy": 0.872173011302948, "num_tokens": 697606943.0, "step": 18370 }, { "epoch": 2.3369800279862614, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.000274658203125, "ewc_loss_parallel": 0.00011396408081054688, "grad_norm": 7.156714916229248, "learning_rate": 1e-05, "loss": 0.4919, "mean_token_accuracy": 0.8754835724830627, "num_tokens": 697643634.0, "step": 18371 }, { "epoch": 2.337107238264852, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.000274658203125, "ewc_loss_parallel": 0.00011110305786132812, "grad_norm": 6.722723960876465, "learning_rate": 1e-05, "loss": 0.4944, "mean_token_accuracy": 0.8755197525024414, "num_tokens": 697682429.0, "step": 18372 }, { "epoch": 2.3372344485434424, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.000274658203125, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.726744174957275, "learning_rate": 1e-05, "loss": 0.4754, "mean_token_accuracy": 0.8747358322143555, "num_tokens": 697718039.0, "step": 18373 }, { "epoch": 2.337361658822033, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.000274658203125, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.632784843444824, "learning_rate": 1e-05, "loss": 0.4736, "mean_token_accuracy": 0.8820110559463501, "num_tokens": 697756179.0, "step": 18374 }, { "epoch": 2.3374888691006235, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.000274658203125, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.419035911560059, "learning_rate": 1e-05, "loss": 0.4661, "mean_token_accuracy": 0.8806078433990479, "num_tokens": 697795327.0, "step": 18375 }, { "epoch": 2.337616079379214, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.000274658203125, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.57303524017334, "learning_rate": 1e-05, "loss": 0.4867, "mean_token_accuracy": 0.8736761212348938, "num_tokens": 697830423.0, "step": 18376 }, { "epoch": 2.3377432896578045, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.000274658203125, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.22675895690918, "learning_rate": 1e-05, "loss": 0.5345, "mean_token_accuracy": 0.8552975654602051, "num_tokens": 697870890.0, "step": 18377 }, { "epoch": 2.337870499936395, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.000274658203125, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.54353141784668, "learning_rate": 1e-05, "loss": 0.5039, "mean_token_accuracy": 0.8673957586288452, "num_tokens": 697910880.0, "step": 18378 }, { "epoch": 2.337997710214985, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.000274658203125, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.0943498611450195, "learning_rate": 1e-05, "loss": 0.443, "mean_token_accuracy": 0.8853983283042908, "num_tokens": 697945509.0, "step": 18379 }, { "epoch": 2.338124920493576, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.000274658203125, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.593453407287598, "learning_rate": 1e-05, "loss": 0.5471, "mean_token_accuracy": 0.8568406105041504, "num_tokens": 697987374.0, "step": 18380 }, { "epoch": 2.338252130772166, "ewc_loss": 0.11474609375, "ewc_loss_diag": 0.000274658203125, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 5.955355644226074, "learning_rate": 1e-05, "loss": 0.469, "mean_token_accuracy": 0.8748034834861755, "num_tokens": 698025130.0, "step": 18381 }, { "epoch": 2.3383793410507567, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.000274658203125, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.740084171295166, "learning_rate": 1e-05, "loss": 0.4808, "mean_token_accuracy": 0.8755010962486267, "num_tokens": 698057785.0, "step": 18382 }, { "epoch": 2.3385065513293473, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.000274658203125, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.810351848602295, "learning_rate": 1e-05, "loss": 0.4461, "mean_token_accuracy": 0.8854879140853882, "num_tokens": 698098952.0, "step": 18383 }, { "epoch": 2.338633761607938, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.000274658203125, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 7.00883150100708, "learning_rate": 1e-05, "loss": 0.4623, "mean_token_accuracy": 0.8855375051498413, "num_tokens": 698142280.0, "step": 18384 }, { "epoch": 2.3387609718865283, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.000274658203125, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.704749584197998, "learning_rate": 1e-05, "loss": 0.4872, "mean_token_accuracy": 0.8702956438064575, "num_tokens": 698183083.0, "step": 18385 }, { "epoch": 2.338888182165119, "ewc_loss": 0.1474609375, "ewc_loss_diag": 0.000274658203125, "ewc_loss_parallel": 0.00011968612670898438, "grad_norm": 7.445408344268799, "learning_rate": 1e-05, "loss": 0.5208, "mean_token_accuracy": 0.8739742040634155, "num_tokens": 698215446.0, "step": 18386 }, { "epoch": 2.3390153924437094, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.000274658203125, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 5.796729564666748, "learning_rate": 1e-05, "loss": 0.4899, "mean_token_accuracy": 0.8717876672744751, "num_tokens": 698257353.0, "step": 18387 }, { "epoch": 2.3391426027223, "ewc_loss": 0.1611328125, "ewc_loss_diag": 0.000274658203125, "ewc_loss_parallel": 0.000133514404296875, "grad_norm": 7.872539520263672, "learning_rate": 1e-05, "loss": 0.5183, "mean_token_accuracy": 0.8739719986915588, "num_tokens": 698292753.0, "step": 18388 }, { "epoch": 2.3392698130008904, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.000274658203125, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.2898993492126465, "learning_rate": 1e-05, "loss": 0.4755, "mean_token_accuracy": 0.8784539103507996, "num_tokens": 698327282.0, "step": 18389 }, { "epoch": 2.339397023279481, "ewc_loss": 0.158203125, "ewc_loss_diag": 0.000274658203125, "ewc_loss_parallel": 0.00013065338134765625, "grad_norm": 7.838622570037842, "learning_rate": 1e-05, "loss": 0.5366, "mean_token_accuracy": 0.8744667768478394, "num_tokens": 698368006.0, "step": 18390 }, { "epoch": 2.3395242335580715, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.000274658203125, "ewc_loss_parallel": 0.00011587142944335938, "grad_norm": 6.739304065704346, "learning_rate": 1e-05, "loss": 0.5775, "mean_token_accuracy": 0.8467044234275818, "num_tokens": 698406898.0, "step": 18391 }, { "epoch": 2.339651443836662, "ewc_loss": 0.1455078125, "ewc_loss_diag": 0.000274658203125, "ewc_loss_parallel": 0.00011777877807617188, "grad_norm": 7.405041217803955, "learning_rate": 1e-05, "loss": 0.5165, "mean_token_accuracy": 0.8709496259689331, "num_tokens": 698448159.0, "step": 18392 }, { "epoch": 2.3397786541152525, "ewc_loss": 0.140625, "ewc_loss_diag": 0.000274658203125, "ewc_loss_parallel": 0.00011348724365234375, "grad_norm": 6.773925304412842, "learning_rate": 1e-05, "loss": 0.4767, "mean_token_accuracy": 0.8840165138244629, "num_tokens": 698489075.0, "step": 18393 }, { "epoch": 2.339905864393843, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.000274658203125, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 6.967366695404053, "learning_rate": 1e-05, "loss": 0.4968, "mean_token_accuracy": 0.873689591884613, "num_tokens": 698527619.0, "step": 18394 }, { "epoch": 2.3400330746724336, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.000274658203125, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 6.681632995605469, "learning_rate": 1e-05, "loss": 0.5177, "mean_token_accuracy": 0.8690232038497925, "num_tokens": 698565250.0, "step": 18395 }, { "epoch": 2.340160284951024, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.000274658203125, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.61616325378418, "learning_rate": 1e-05, "loss": 0.4725, "mean_token_accuracy": 0.8778549432754517, "num_tokens": 698603207.0, "step": 18396 }, { "epoch": 2.3402874952296147, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.000274658203125, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.565160274505615, "learning_rate": 1e-05, "loss": 0.4457, "mean_token_accuracy": 0.8883877992630005, "num_tokens": 698636747.0, "step": 18397 }, { "epoch": 2.340414705508205, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.000274658203125, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.381580829620361, "learning_rate": 1e-05, "loss": 0.4493, "mean_token_accuracy": 0.8860073089599609, "num_tokens": 698677885.0, "step": 18398 }, { "epoch": 2.3405419157867957, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.000274658203125, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.479389667510986, "learning_rate": 1e-05, "loss": 0.4834, "mean_token_accuracy": 0.8734910488128662, "num_tokens": 698721924.0, "step": 18399 }, { "epoch": 2.3406691260653862, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.000274658203125, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.273013591766357, "learning_rate": 1e-05, "loss": 0.51, "mean_token_accuracy": 0.8648487329483032, "num_tokens": 698754847.0, "step": 18400 }, { "epoch": 2.3407963363439768, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.000274658203125, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.410484790802002, "learning_rate": 1e-05, "loss": 0.444, "mean_token_accuracy": 0.8872858285903931, "num_tokens": 698798567.0, "step": 18401 }, { "epoch": 2.3409235466225673, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.000274658203125, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.177371978759766, "learning_rate": 1e-05, "loss": 0.4459, "mean_token_accuracy": 0.8821973204612732, "num_tokens": 698837574.0, "step": 18402 }, { "epoch": 2.341050756901158, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.000274658203125, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.401772499084473, "learning_rate": 1e-05, "loss": 0.5207, "mean_token_accuracy": 0.872759222984314, "num_tokens": 698867550.0, "step": 18403 }, { "epoch": 2.341177967179748, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.000274658203125, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.085402488708496, "learning_rate": 1e-05, "loss": 0.4392, "mean_token_accuracy": 0.8871054649353027, "num_tokens": 698898859.0, "step": 18404 }, { "epoch": 2.341305177458339, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.000274658203125, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.442899227142334, "learning_rate": 1e-05, "loss": 0.4711, "mean_token_accuracy": 0.880720853805542, "num_tokens": 698933449.0, "step": 18405 }, { "epoch": 2.341432387736929, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.000274658203125, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 5.971596717834473, "learning_rate": 1e-05, "loss": 0.4327, "mean_token_accuracy": 0.8882358074188232, "num_tokens": 698969045.0, "step": 18406 }, { "epoch": 2.3415595980155195, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.000274658203125, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.5813307762146, "learning_rate": 1e-05, "loss": 0.5107, "mean_token_accuracy": 0.8651177287101746, "num_tokens": 699006201.0, "step": 18407 }, { "epoch": 2.34168680829411, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.000274658203125, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.803630828857422, "learning_rate": 1e-05, "loss": 0.4417, "mean_token_accuracy": 0.88396155834198, "num_tokens": 699044091.0, "step": 18408 }, { "epoch": 2.3418140185727006, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.000274658203125, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.868102550506592, "learning_rate": 1e-05, "loss": 0.4847, "mean_token_accuracy": 0.8748378753662109, "num_tokens": 699075519.0, "step": 18409 }, { "epoch": 2.341941228851291, "ewc_loss": 0.10986328125, "ewc_loss_diag": 0.000274658203125, "ewc_loss_parallel": 8.249282836914062e-05, "grad_norm": 5.61090612411499, "learning_rate": 1e-05, "loss": 0.4438, "mean_token_accuracy": 0.8853917121887207, "num_tokens": 699113129.0, "step": 18410 }, { "epoch": 2.3420684391298816, "ewc_loss": 0.1474609375, "ewc_loss_diag": 0.000274658203125, "ewc_loss_parallel": 0.0001201629638671875, "grad_norm": 7.3961615562438965, "learning_rate": 1e-05, "loss": 0.4897, "mean_token_accuracy": 0.8775500655174255, "num_tokens": 699147515.0, "step": 18411 }, { "epoch": 2.342195649408472, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.000274658203125, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.594042778015137, "learning_rate": 1e-05, "loss": 0.4831, "mean_token_accuracy": 0.8712190389633179, "num_tokens": 699190578.0, "step": 18412 }, { "epoch": 2.3423228596870627, "ewc_loss": 0.1640625, "ewc_loss_diag": 0.000274658203125, "ewc_loss_parallel": 0.00013637542724609375, "grad_norm": 8.078169822692871, "learning_rate": 1e-05, "loss": 0.5343, "mean_token_accuracy": 0.8696566224098206, "num_tokens": 699230422.0, "step": 18413 }, { "epoch": 2.342450069965653, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.000274658203125, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.2171630859375, "learning_rate": 1e-05, "loss": 0.4823, "mean_token_accuracy": 0.8782330751419067, "num_tokens": 699268662.0, "step": 18414 }, { "epoch": 2.3425772802442437, "ewc_loss": 0.166015625, "ewc_loss_diag": 0.000274658203125, "ewc_loss_parallel": 0.00013828277587890625, "grad_norm": 8.256939888000488, "learning_rate": 1e-05, "loss": 0.5327, "mean_token_accuracy": 0.8697353005409241, "num_tokens": 699312498.0, "step": 18415 }, { "epoch": 2.3427044905228342, "ewc_loss": 0.15234375, "ewc_loss_diag": 0.000274658203125, "ewc_loss_parallel": 0.00012493133544921875, "grad_norm": 7.019546985626221, "learning_rate": 1e-05, "loss": 0.471, "mean_token_accuracy": 0.8863658905029297, "num_tokens": 699354143.0, "step": 18416 }, { "epoch": 2.3428317008014248, "ewc_loss": 0.15234375, "ewc_loss_diag": 0.000274658203125, "ewc_loss_parallel": 0.00012493133544921875, "grad_norm": 7.6940789222717285, "learning_rate": 1e-05, "loss": 0.481, "mean_token_accuracy": 0.8830284476280212, "num_tokens": 699392099.0, "step": 18417 }, { "epoch": 2.3429589110800153, "ewc_loss": 0.1484375, "ewc_loss_diag": 0.000274658203125, "ewc_loss_parallel": 0.00012063980102539062, "grad_norm": 7.043713092803955, "learning_rate": 1e-05, "loss": 0.4821, "mean_token_accuracy": 0.8829779624938965, "num_tokens": 699428449.0, "step": 18418 }, { "epoch": 2.343086121358606, "ewc_loss": 0.140625, "ewc_loss_diag": 0.000274658203125, "ewc_loss_parallel": 0.00011301040649414062, "grad_norm": 7.217791557312012, "learning_rate": 1e-05, "loss": 0.5011, "mean_token_accuracy": 0.8734487295150757, "num_tokens": 699469359.0, "step": 18419 }, { "epoch": 2.3432133316371964, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.000274658203125, "ewc_loss_parallel": 0.00011110305786132812, "grad_norm": 6.790958881378174, "learning_rate": 1e-05, "loss": 0.4697, "mean_token_accuracy": 0.8833876252174377, "num_tokens": 699507481.0, "step": 18420 }, { "epoch": 2.343340541915787, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.000274658203125, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.896246910095215, "learning_rate": 1e-05, "loss": 0.5128, "mean_token_accuracy": 0.8686834573745728, "num_tokens": 699548335.0, "step": 18421 }, { "epoch": 2.3434677521943774, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.000274658203125, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.574125289916992, "learning_rate": 1e-05, "loss": 0.5117, "mean_token_accuracy": 0.8679244518280029, "num_tokens": 699587460.0, "step": 18422 }, { "epoch": 2.343594962472968, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.721154689788818, "learning_rate": 1e-05, "loss": 0.5082, "mean_token_accuracy": 0.8687428832054138, "num_tokens": 699631128.0, "step": 18423 }, { "epoch": 2.3437221727515585, "ewc_loss": 0.125, "ewc_loss_diag": 0.000274658203125, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.365276336669922, "learning_rate": 1e-05, "loss": 0.5135, "mean_token_accuracy": 0.8663371801376343, "num_tokens": 699672133.0, "step": 18424 }, { "epoch": 2.343849383030149, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.000274658203125, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.660004138946533, "learning_rate": 1e-05, "loss": 0.4921, "mean_token_accuracy": 0.8721822500228882, "num_tokens": 699705280.0, "step": 18425 }, { "epoch": 2.3439765933087395, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.000274658203125, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.17942476272583, "learning_rate": 1e-05, "loss": 0.486, "mean_token_accuracy": 0.8713632822036743, "num_tokens": 699740959.0, "step": 18426 }, { "epoch": 2.3441038035873296, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.000274658203125, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.715123176574707, "learning_rate": 1e-05, "loss": 0.5304, "mean_token_accuracy": 0.8590957522392273, "num_tokens": 699775546.0, "step": 18427 }, { "epoch": 2.3442310138659206, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.000274658203125, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.9950666427612305, "learning_rate": 1e-05, "loss": 0.4665, "mean_token_accuracy": 0.8775566220283508, "num_tokens": 699809777.0, "step": 18428 }, { "epoch": 2.3443582241445107, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.000274658203125, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.911857604980469, "learning_rate": 1e-05, "loss": 0.4464, "mean_token_accuracy": 0.8885849714279175, "num_tokens": 699846140.0, "step": 18429 }, { "epoch": 2.344485434423101, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.000274658203125, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.824181079864502, "learning_rate": 1e-05, "loss": 0.4528, "mean_token_accuracy": 0.8772176504135132, "num_tokens": 699882769.0, "step": 18430 }, { "epoch": 2.3446126447016917, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.000274658203125, "ewc_loss_parallel": 0.00011539459228515625, "grad_norm": 7.285235404968262, "learning_rate": 1e-05, "loss": 0.4956, "mean_token_accuracy": 0.8743965029716492, "num_tokens": 699923658.0, "step": 18431 }, { "epoch": 2.3447398549802823, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.000274658203125, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.819955825805664, "learning_rate": 1e-05, "loss": 0.4156, "mean_token_accuracy": 0.8894727826118469, "num_tokens": 699957457.0, "step": 18432 }, { "epoch": 2.344867065258873, "ewc_loss": 0.154296875, "ewc_loss_diag": 0.000274658203125, "ewc_loss_parallel": 0.00012683868408203125, "grad_norm": 7.715224742889404, "learning_rate": 1e-05, "loss": 0.4741, "mean_token_accuracy": 0.8866432309150696, "num_tokens": 699994852.0, "step": 18433 }, { "epoch": 2.3449942755374633, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.000274658203125, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.205431938171387, "learning_rate": 1e-05, "loss": 0.4524, "mean_token_accuracy": 0.8830423355102539, "num_tokens": 700030983.0, "step": 18434 }, { "epoch": 2.345121485816054, "ewc_loss": 0.154296875, "ewc_loss_diag": 0.000274658203125, "ewc_loss_parallel": 0.00012683868408203125, "grad_norm": 7.772372245788574, "learning_rate": 1e-05, "loss": 0.5005, "mean_token_accuracy": 0.8816806674003601, "num_tokens": 700067907.0, "step": 18435 }, { "epoch": 2.3452486960946444, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.000274658203125, "ewc_loss_parallel": 0.00011110305786132812, "grad_norm": 6.683804512023926, "learning_rate": 1e-05, "loss": 0.4892, "mean_token_accuracy": 0.877341628074646, "num_tokens": 700107989.0, "step": 18436 }, { "epoch": 2.345375906373235, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.000274658203125, "ewc_loss_parallel": 0.00011539459228515625, "grad_norm": 7.315141677856445, "learning_rate": 1e-05, "loss": 0.4604, "mean_token_accuracy": 0.8898951411247253, "num_tokens": 700145870.0, "step": 18437 }, { "epoch": 2.3455031166518254, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.000274658203125, "ewc_loss_parallel": 0.00011157989501953125, "grad_norm": 6.818445205688477, "learning_rate": 1e-05, "loss": 0.4707, "mean_token_accuracy": 0.8851486444473267, "num_tokens": 700186762.0, "step": 18438 }, { "epoch": 2.345630326930416, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.000274658203125, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.752639293670654, "learning_rate": 1e-05, "loss": 0.4612, "mean_token_accuracy": 0.8820598125457764, "num_tokens": 700227307.0, "step": 18439 }, { "epoch": 2.3457575372090065, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.000274658203125, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 6.814811706542969, "learning_rate": 1e-05, "loss": 0.5112, "mean_token_accuracy": 0.8719039559364319, "num_tokens": 700263660.0, "step": 18440 }, { "epoch": 2.345884747487597, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.000274658203125, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.290464878082275, "learning_rate": 1e-05, "loss": 0.4293, "mean_token_accuracy": 0.8925602436065674, "num_tokens": 700294835.0, "step": 18441 }, { "epoch": 2.3460119577661875, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.893189430236816, "learning_rate": 1e-05, "loss": 0.4614, "mean_token_accuracy": 0.8836392164230347, "num_tokens": 700335977.0, "step": 18442 }, { "epoch": 2.346139168044778, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.000274658203125, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.984762191772461, "learning_rate": 1e-05, "loss": 0.4619, "mean_token_accuracy": 0.8752409219741821, "num_tokens": 700369388.0, "step": 18443 }, { "epoch": 2.3462663783233686, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.000274658203125, "ewc_loss_parallel": 0.00011110305786132812, "grad_norm": 7.115757465362549, "learning_rate": 1e-05, "loss": 0.4903, "mean_token_accuracy": 0.8753722906112671, "num_tokens": 700409575.0, "step": 18444 }, { "epoch": 2.346393588601959, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.854034423828125, "learning_rate": 1e-05, "loss": 0.4351, "mean_token_accuracy": 0.8874837160110474, "num_tokens": 700449463.0, "step": 18445 }, { "epoch": 2.3465207988805497, "ewc_loss": 0.1455078125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00011777877807617188, "grad_norm": 7.439173221588135, "learning_rate": 1e-05, "loss": 0.4871, "mean_token_accuracy": 0.8801628947257996, "num_tokens": 700483765.0, "step": 18446 }, { "epoch": 2.34664800915914, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 5.982761383056641, "learning_rate": 1e-05, "loss": 0.4601, "mean_token_accuracy": 0.8827041387557983, "num_tokens": 700521953.0, "step": 18447 }, { "epoch": 2.3467752194377307, "ewc_loss": 0.1513671875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.0001239776611328125, "grad_norm": 7.669391632080078, "learning_rate": 1e-05, "loss": 0.5152, "mean_token_accuracy": 0.8704656362533569, "num_tokens": 700559662.0, "step": 18448 }, { "epoch": 2.3469024297163212, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.336196422576904, "learning_rate": 1e-05, "loss": 0.4818, "mean_token_accuracy": 0.8777549266815186, "num_tokens": 700598880.0, "step": 18449 }, { "epoch": 2.3470296399949118, "ewc_loss": 0.1494140625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.0001220703125, "grad_norm": 7.569791793823242, "learning_rate": 1e-05, "loss": 0.4439, "mean_token_accuracy": 0.8940080404281616, "num_tokens": 700631712.0, "step": 18450 }, { "epoch": 2.3471568502735023, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 6.602814197540283, "learning_rate": 1e-05, "loss": 0.5219, "mean_token_accuracy": 0.8652143478393555, "num_tokens": 700665420.0, "step": 18451 }, { "epoch": 2.3472840605520924, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00011396408081054688, "grad_norm": 7.218929767608643, "learning_rate": 1e-05, "loss": 0.4818, "mean_token_accuracy": 0.8805648684501648, "num_tokens": 700704603.0, "step": 18452 }, { "epoch": 2.3474112708306833, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.598691463470459, "learning_rate": 1e-05, "loss": 0.4561, "mean_token_accuracy": 0.8822745084762573, "num_tokens": 700738730.0, "step": 18453 }, { "epoch": 2.3475384811092734, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.895170211791992, "learning_rate": 1e-05, "loss": 0.4385, "mean_token_accuracy": 0.8950279951095581, "num_tokens": 700773848.0, "step": 18454 }, { "epoch": 2.347665691387864, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.493928909301758, "learning_rate": 1e-05, "loss": 0.4856, "mean_token_accuracy": 0.8772140741348267, "num_tokens": 700810352.0, "step": 18455 }, { "epoch": 2.3477929016664545, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.682268142700195, "learning_rate": 1e-05, "loss": 0.4942, "mean_token_accuracy": 0.8727564811706543, "num_tokens": 700846432.0, "step": 18456 }, { "epoch": 2.347920111945045, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.3556060791015625, "learning_rate": 1e-05, "loss": 0.4551, "mean_token_accuracy": 0.8827656507492065, "num_tokens": 700883674.0, "step": 18457 }, { "epoch": 2.3480473222236355, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.6000871658325195, "learning_rate": 1e-05, "loss": 0.5441, "mean_token_accuracy": 0.858579695224762, "num_tokens": 700925096.0, "step": 18458 }, { "epoch": 2.348174532502226, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.202150821685791, "learning_rate": 1e-05, "loss": 0.4275, "mean_token_accuracy": 0.8925845623016357, "num_tokens": 700957301.0, "step": 18459 }, { "epoch": 2.3483017427808166, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.628311634063721, "learning_rate": 1e-05, "loss": 0.4512, "mean_token_accuracy": 0.886184811592102, "num_tokens": 700990636.0, "step": 18460 }, { "epoch": 2.348428953059407, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 6.048367500305176, "learning_rate": 1e-05, "loss": 0.4258, "mean_token_accuracy": 0.8889986872673035, "num_tokens": 701026379.0, "step": 18461 }, { "epoch": 2.3485561633379977, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.760934829711914, "learning_rate": 1e-05, "loss": 0.4763, "mean_token_accuracy": 0.876983642578125, "num_tokens": 701070377.0, "step": 18462 }, { "epoch": 2.348683373616588, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.897388935089111, "learning_rate": 1e-05, "loss": 0.4379, "mean_token_accuracy": 0.884120523929596, "num_tokens": 701107691.0, "step": 18463 }, { "epoch": 2.3488105838951787, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 7.019232273101807, "learning_rate": 1e-05, "loss": 0.4793, "mean_token_accuracy": 0.8795926570892334, "num_tokens": 701147775.0, "step": 18464 }, { "epoch": 2.3489377941737692, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.822375297546387, "learning_rate": 1e-05, "loss": 0.4337, "mean_token_accuracy": 0.8866650462150574, "num_tokens": 701187732.0, "step": 18465 }, { "epoch": 2.3490650044523598, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00011730194091796875, "grad_norm": 7.361578941345215, "learning_rate": 1e-05, "loss": 0.4561, "mean_token_accuracy": 0.8899307250976562, "num_tokens": 701220916.0, "step": 18466 }, { "epoch": 2.3491922147309503, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 5.9549360275268555, "learning_rate": 1e-05, "loss": 0.4603, "mean_token_accuracy": 0.8806647658348083, "num_tokens": 701258831.0, "step": 18467 }, { "epoch": 2.349319425009541, "ewc_loss": 0.1494140625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.0001220703125, "grad_norm": 7.581012725830078, "learning_rate": 1e-05, "loss": 0.4778, "mean_token_accuracy": 0.8846075534820557, "num_tokens": 701299310.0, "step": 18468 }, { "epoch": 2.3494466352881314, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.323225021362305, "learning_rate": 1e-05, "loss": 0.4895, "mean_token_accuracy": 0.869236946105957, "num_tokens": 701331471.0, "step": 18469 }, { "epoch": 2.349573845566722, "ewc_loss": 0.1455078125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00011777877807617188, "grad_norm": 7.4008402824401855, "learning_rate": 1e-05, "loss": 0.4978, "mean_token_accuracy": 0.8766216039657593, "num_tokens": 701371126.0, "step": 18470 }, { "epoch": 2.3497010558453124, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.573347091674805, "learning_rate": 1e-05, "loss": 0.4145, "mean_token_accuracy": 0.8999453783035278, "num_tokens": 701413612.0, "step": 18471 }, { "epoch": 2.349828266123903, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.937177658081055, "learning_rate": 1e-05, "loss": 0.5522, "mean_token_accuracy": 0.8560559153556824, "num_tokens": 701456085.0, "step": 18472 }, { "epoch": 2.3499554764024935, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.670487403869629, "learning_rate": 1e-05, "loss": 0.4827, "mean_token_accuracy": 0.8757340908050537, "num_tokens": 701494409.0, "step": 18473 }, { "epoch": 2.350082686681084, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.421631813049316, "learning_rate": 1e-05, "loss": 0.4737, "mean_token_accuracy": 0.8755521178245544, "num_tokens": 701533610.0, "step": 18474 }, { "epoch": 2.3502098969596745, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.7828216552734375, "learning_rate": 1e-05, "loss": 0.5172, "mean_token_accuracy": 0.8636715412139893, "num_tokens": 701571324.0, "step": 18475 }, { "epoch": 2.350337107238265, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 6.009176254272461, "learning_rate": 1e-05, "loss": 0.5204, "mean_token_accuracy": 0.8641460537910461, "num_tokens": 701609077.0, "step": 18476 }, { "epoch": 2.350464317516855, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 7.029462814331055, "learning_rate": 1e-05, "loss": 0.4235, "mean_token_accuracy": 0.898026704788208, "num_tokens": 701647281.0, "step": 18477 }, { "epoch": 2.350591527795446, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.784119606018066, "learning_rate": 1e-05, "loss": 0.4357, "mean_token_accuracy": 0.8871418237686157, "num_tokens": 701680039.0, "step": 18478 }, { "epoch": 2.350718738074036, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00011920928955078125, "grad_norm": 7.420403480529785, "learning_rate": 1e-05, "loss": 0.511, "mean_token_accuracy": 0.8737562298774719, "num_tokens": 701721679.0, "step": 18479 }, { "epoch": 2.3508459483526267, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 5.870663166046143, "learning_rate": 1e-05, "loss": 0.4151, "mean_token_accuracy": 0.8955587148666382, "num_tokens": 701758692.0, "step": 18480 }, { "epoch": 2.3509731586312173, "ewc_loss": 0.154296875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00012683868408203125, "grad_norm": 7.756315231323242, "learning_rate": 1e-05, "loss": 0.5638, "mean_token_accuracy": 0.8628970980644226, "num_tokens": 701793038.0, "step": 18481 }, { "epoch": 2.351100368909808, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.305318355560303, "learning_rate": 1e-05, "loss": 0.4815, "mean_token_accuracy": 0.8770779371261597, "num_tokens": 701831172.0, "step": 18482 }, { "epoch": 2.3512275791883983, "ewc_loss": 0.150390625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.665528297424316, "learning_rate": 1e-05, "loss": 0.4628, "mean_token_accuracy": 0.8890056610107422, "num_tokens": 701865644.0, "step": 18483 }, { "epoch": 2.351354789466989, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 6.636660099029541, "learning_rate": 1e-05, "loss": 0.4854, "mean_token_accuracy": 0.8781596422195435, "num_tokens": 701903928.0, "step": 18484 }, { "epoch": 2.3514819997455794, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.0001163482666015625, "grad_norm": 7.290556907653809, "learning_rate": 1e-05, "loss": 0.5324, "mean_token_accuracy": 0.8621938228607178, "num_tokens": 701937440.0, "step": 18485 }, { "epoch": 2.35160921002417, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.6515021324157715, "learning_rate": 1e-05, "loss": 0.495, "mean_token_accuracy": 0.8792198896408081, "num_tokens": 701975936.0, "step": 18486 }, { "epoch": 2.3517364203027604, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.931027412414551, "learning_rate": 1e-05, "loss": 0.4801, "mean_token_accuracy": 0.8741995096206665, "num_tokens": 702009504.0, "step": 18487 }, { "epoch": 2.351863630581351, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.542913436889648, "learning_rate": 1e-05, "loss": 0.4808, "mean_token_accuracy": 0.8725451231002808, "num_tokens": 702046631.0, "step": 18488 }, { "epoch": 2.3519908408599415, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.680334568023682, "learning_rate": 1e-05, "loss": 0.4808, "mean_token_accuracy": 0.8780678510665894, "num_tokens": 702086915.0, "step": 18489 }, { "epoch": 2.352118051138532, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.3954176902771, "learning_rate": 1e-05, "loss": 0.4813, "mean_token_accuracy": 0.8731436729431152, "num_tokens": 702129214.0, "step": 18490 }, { "epoch": 2.3522452614171225, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.581829071044922, "learning_rate": 1e-05, "loss": 0.4889, "mean_token_accuracy": 0.8727583289146423, "num_tokens": 702167450.0, "step": 18491 }, { "epoch": 2.352372471695713, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.244067192077637, "learning_rate": 1e-05, "loss": 0.4689, "mean_token_accuracy": 0.8788795471191406, "num_tokens": 702205967.0, "step": 18492 }, { "epoch": 2.3524996819743036, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.592586040496826, "learning_rate": 1e-05, "loss": 0.4783, "mean_token_accuracy": 0.8736064434051514, "num_tokens": 702239570.0, "step": 18493 }, { "epoch": 2.352626892252894, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.080078601837158, "learning_rate": 1e-05, "loss": 0.4844, "mean_token_accuracy": 0.8724867105484009, "num_tokens": 702279276.0, "step": 18494 }, { "epoch": 2.3527541025314846, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.716238021850586, "learning_rate": 1e-05, "loss": 0.4622, "mean_token_accuracy": 0.8837437629699707, "num_tokens": 702319695.0, "step": 18495 }, { "epoch": 2.352881312810075, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.901055812835693, "learning_rate": 1e-05, "loss": 0.4465, "mean_token_accuracy": 0.8828877806663513, "num_tokens": 702361933.0, "step": 18496 }, { "epoch": 2.3530085230886657, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 6.981378078460693, "learning_rate": 1e-05, "loss": 0.4841, "mean_token_accuracy": 0.8783961534500122, "num_tokens": 702398532.0, "step": 18497 }, { "epoch": 2.3531357333672562, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.80389928817749, "learning_rate": 1e-05, "loss": 0.4556, "mean_token_accuracy": 0.8784884214401245, "num_tokens": 702437880.0, "step": 18498 }, { "epoch": 2.3532629436458468, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.0001163482666015625, "grad_norm": 7.333606719970703, "learning_rate": 1e-05, "loss": 0.5155, "mean_token_accuracy": 0.8723391890525818, "num_tokens": 702478908.0, "step": 18499 }, { "epoch": 2.353390153924437, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 5.890605449676514, "learning_rate": 1e-05, "loss": 0.4917, "mean_token_accuracy": 0.8698230981826782, "num_tokens": 702523817.0, "step": 18500 }, { "epoch": 2.353517364203028, "ewc_loss": 0.150390625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.628983497619629, "learning_rate": 1e-05, "loss": 0.4913, "mean_token_accuracy": 0.8788619637489319, "num_tokens": 702569682.0, "step": 18501 }, { "epoch": 2.353644574481618, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.266960144042969, "learning_rate": 1e-05, "loss": 0.4966, "mean_token_accuracy": 0.8705083131790161, "num_tokens": 702604746.0, "step": 18502 }, { "epoch": 2.353771784760209, "ewc_loss": 0.1494140625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00012159347534179688, "grad_norm": 7.5453901290893555, "learning_rate": 1e-05, "loss": 0.5262, "mean_token_accuracy": 0.8682465553283691, "num_tokens": 702637660.0, "step": 18503 }, { "epoch": 2.353898995038799, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 6.597047328948975, "learning_rate": 1e-05, "loss": 0.4318, "mean_token_accuracy": 0.8953272104263306, "num_tokens": 702676942.0, "step": 18504 }, { "epoch": 2.3540262053173895, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00011110305786132812, "grad_norm": 7.106899261474609, "learning_rate": 1e-05, "loss": 0.5304, "mean_token_accuracy": 0.8638336658477783, "num_tokens": 702718629.0, "step": 18505 }, { "epoch": 2.35415341559598, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.690891265869141, "learning_rate": 1e-05, "loss": 0.4841, "mean_token_accuracy": 0.8790573477745056, "num_tokens": 702757698.0, "step": 18506 }, { "epoch": 2.3542806258745705, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.5894455909729, "learning_rate": 1e-05, "loss": 0.4551, "mean_token_accuracy": 0.8847588896751404, "num_tokens": 702797337.0, "step": 18507 }, { "epoch": 2.354407836153161, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.751438140869141, "learning_rate": 1e-05, "loss": 0.5122, "mean_token_accuracy": 0.8680188655853271, "num_tokens": 702832002.0, "step": 18508 }, { "epoch": 2.3545350464317516, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.151010513305664, "learning_rate": 1e-05, "loss": 0.4586, "mean_token_accuracy": 0.8830606341362, "num_tokens": 702874472.0, "step": 18509 }, { "epoch": 2.354662256710342, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.934754371643066, "learning_rate": 1e-05, "loss": 0.5203, "mean_token_accuracy": 0.8740924596786499, "num_tokens": 702909867.0, "step": 18510 }, { "epoch": 2.3547894669889327, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.8701581954956055, "learning_rate": 1e-05, "loss": 0.4317, "mean_token_accuracy": 0.8850990533828735, "num_tokens": 702945817.0, "step": 18511 }, { "epoch": 2.354916677267523, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.0001125335693359375, "grad_norm": 7.231104373931885, "learning_rate": 1e-05, "loss": 0.4478, "mean_token_accuracy": 0.8948937058448792, "num_tokens": 702986373.0, "step": 18512 }, { "epoch": 2.3550438875461137, "ewc_loss": 0.11376953125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.836180686950684, "learning_rate": 1e-05, "loss": 0.4698, "mean_token_accuracy": 0.8755666017532349, "num_tokens": 703026077.0, "step": 18513 }, { "epoch": 2.3551710978247042, "ewc_loss": 0.1494140625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.0001220703125, "grad_norm": 7.5785322189331055, "learning_rate": 1e-05, "loss": 0.5015, "mean_token_accuracy": 0.8762581944465637, "num_tokens": 703067042.0, "step": 18514 }, { "epoch": 2.3552983081032948, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.100292682647705, "learning_rate": 1e-05, "loss": 0.4507, "mean_token_accuracy": 0.8842167258262634, "num_tokens": 703105290.0, "step": 18515 }, { "epoch": 2.3554255183818853, "ewc_loss": 0.15234375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00012493133544921875, "grad_norm": 7.68021297454834, "learning_rate": 1e-05, "loss": 0.52, "mean_token_accuracy": 0.8735880851745605, "num_tokens": 703144446.0, "step": 18516 }, { "epoch": 2.355552728660476, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.4800705909729, "learning_rate": 1e-05, "loss": 0.5045, "mean_token_accuracy": 0.8709943294525146, "num_tokens": 703179490.0, "step": 18517 }, { "epoch": 2.3556799389390664, "ewc_loss": 0.1484375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00012063980102539062, "grad_norm": 7.422226905822754, "learning_rate": 1e-05, "loss": 0.5116, "mean_token_accuracy": 0.8714682459831238, "num_tokens": 703214218.0, "step": 18518 }, { "epoch": 2.355807149217657, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.603185653686523, "learning_rate": 1e-05, "loss": 0.4702, "mean_token_accuracy": 0.8827618956565857, "num_tokens": 703252576.0, "step": 18519 }, { "epoch": 2.3559343594962474, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00011110305786132812, "grad_norm": 7.057704925537109, "learning_rate": 1e-05, "loss": 0.4219, "mean_token_accuracy": 0.897747278213501, "num_tokens": 703291129.0, "step": 18520 }, { "epoch": 2.356061569774838, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.557776927947998, "learning_rate": 1e-05, "loss": 0.5222, "mean_token_accuracy": 0.8681141138076782, "num_tokens": 703326864.0, "step": 18521 }, { "epoch": 2.3561887800534285, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.762094974517822, "learning_rate": 1e-05, "loss": 0.4807, "mean_token_accuracy": 0.8779329061508179, "num_tokens": 703369719.0, "step": 18522 }, { "epoch": 2.356315990332019, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.436756610870361, "learning_rate": 1e-05, "loss": 0.477, "mean_token_accuracy": 0.8775306940078735, "num_tokens": 703415304.0, "step": 18523 }, { "epoch": 2.3564432006106095, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.615877628326416, "learning_rate": 1e-05, "loss": 0.4977, "mean_token_accuracy": 0.871361255645752, "num_tokens": 703456273.0, "step": 18524 }, { "epoch": 2.3565704108891996, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.297712326049805, "learning_rate": 1e-05, "loss": 0.4931, "mean_token_accuracy": 0.870803952217102, "num_tokens": 703491355.0, "step": 18525 }, { "epoch": 2.3566976211677906, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.591802597045898, "learning_rate": 1e-05, "loss": 0.476, "mean_token_accuracy": 0.8798624277114868, "num_tokens": 703532951.0, "step": 18526 }, { "epoch": 2.3568248314463807, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.1382551193237305, "learning_rate": 1e-05, "loss": 0.4754, "mean_token_accuracy": 0.8770033717155457, "num_tokens": 703575306.0, "step": 18527 }, { "epoch": 2.356952041724971, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.668533802032471, "learning_rate": 1e-05, "loss": 0.4496, "mean_token_accuracy": 0.8851386308670044, "num_tokens": 703615198.0, "step": 18528 }, { "epoch": 2.3570792520035617, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.978487491607666, "learning_rate": 1e-05, "loss": 0.4335, "mean_token_accuracy": 0.8871794939041138, "num_tokens": 703651780.0, "step": 18529 }, { "epoch": 2.3572064622821522, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.8888630867004395, "learning_rate": 1e-05, "loss": 0.4931, "mean_token_accuracy": 0.8724766969680786, "num_tokens": 703685273.0, "step": 18530 }, { "epoch": 2.3573336725607428, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.8394455909729, "learning_rate": 1e-05, "loss": 0.4284, "mean_token_accuracy": 0.8885971903800964, "num_tokens": 703722185.0, "step": 18531 }, { "epoch": 2.3574608828393333, "ewc_loss": 0.140625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00011301040649414062, "grad_norm": 7.211265563964844, "learning_rate": 1e-05, "loss": 0.4747, "mean_token_accuracy": 0.8835645914077759, "num_tokens": 703760229.0, "step": 18532 }, { "epoch": 2.357588093117924, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.853000164031982, "learning_rate": 1e-05, "loss": 0.4353, "mean_token_accuracy": 0.8855858445167542, "num_tokens": 703794286.0, "step": 18533 }, { "epoch": 2.3577153033965144, "ewc_loss": 0.1474609375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.0001201629638671875, "grad_norm": 7.532251358032227, "learning_rate": 1e-05, "loss": 0.4822, "mean_token_accuracy": 0.8827655911445618, "num_tokens": 703827486.0, "step": 18534 }, { "epoch": 2.357842513675105, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.1269659996032715, "learning_rate": 1e-05, "loss": 0.5142, "mean_token_accuracy": 0.8632501363754272, "num_tokens": 703861398.0, "step": 18535 }, { "epoch": 2.3579697239536954, "ewc_loss": 0.1494140625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.0001220703125, "grad_norm": 7.563070297241211, "learning_rate": 1e-05, "loss": 0.5151, "mean_token_accuracy": 0.8719000816345215, "num_tokens": 703901690.0, "step": 18536 }, { "epoch": 2.358096934232286, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.488028049468994, "learning_rate": 1e-05, "loss": 0.4744, "mean_token_accuracy": 0.878900408744812, "num_tokens": 703943828.0, "step": 18537 }, { "epoch": 2.3582241445108765, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00011396408081054688, "grad_norm": 7.206662178039551, "learning_rate": 1e-05, "loss": 0.5108, "mean_token_accuracy": 0.870679497718811, "num_tokens": 703984893.0, "step": 18538 }, { "epoch": 2.358351354789467, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 6.670099258422852, "learning_rate": 1e-05, "loss": 0.5179, "mean_token_accuracy": 0.8678687214851379, "num_tokens": 704023633.0, "step": 18539 }, { "epoch": 2.3584785650680575, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.682237148284912, "learning_rate": 1e-05, "loss": 0.4831, "mean_token_accuracy": 0.8756299614906311, "num_tokens": 704058762.0, "step": 18540 }, { "epoch": 2.358605775346648, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.730946063995361, "learning_rate": 1e-05, "loss": 0.4899, "mean_token_accuracy": 0.874016523361206, "num_tokens": 704099100.0, "step": 18541 }, { "epoch": 2.3587329856252386, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.2119903564453125, "learning_rate": 1e-05, "loss": 0.5261, "mean_token_accuracy": 0.8589857816696167, "num_tokens": 704131900.0, "step": 18542 }, { "epoch": 2.358860195903829, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.904172420501709, "learning_rate": 1e-05, "loss": 0.4728, "mean_token_accuracy": 0.8820719718933105, "num_tokens": 704161940.0, "step": 18543 }, { "epoch": 2.3589874061824196, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.877838611602783, "learning_rate": 1e-05, "loss": 0.4295, "mean_token_accuracy": 0.8853979706764221, "num_tokens": 704200698.0, "step": 18544 }, { "epoch": 2.35911461646101, "ewc_loss": 0.140625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00011301040649414062, "grad_norm": 7.222123622894287, "learning_rate": 1e-05, "loss": 0.4962, "mean_token_accuracy": 0.8753838539123535, "num_tokens": 704243317.0, "step": 18545 }, { "epoch": 2.3592418267396007, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.802398204803467, "learning_rate": 1e-05, "loss": 0.4291, "mean_token_accuracy": 0.8884034156799316, "num_tokens": 704277258.0, "step": 18546 }, { "epoch": 2.3593690370181912, "ewc_loss": 0.1494140625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.0001220703125, "grad_norm": 7.606026649475098, "learning_rate": 1e-05, "loss": 0.4763, "mean_token_accuracy": 0.887366771697998, "num_tokens": 704315932.0, "step": 18547 }, { "epoch": 2.3594962472967818, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.077494144439697, "learning_rate": 1e-05, "loss": 0.5184, "mean_token_accuracy": 0.8629289865493774, "num_tokens": 704357355.0, "step": 18548 }, { "epoch": 2.3596234575753723, "ewc_loss": 0.1552734375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.0001277923583984375, "grad_norm": 7.736205577850342, "learning_rate": 1e-05, "loss": 0.526, "mean_token_accuracy": 0.8700631856918335, "num_tokens": 704392981.0, "step": 18549 }, { "epoch": 2.3597506678539624, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 6.489847183227539, "learning_rate": 1e-05, "loss": 0.4874, "mean_token_accuracy": 0.8792399764060974, "num_tokens": 704430530.0, "step": 18550 }, { "epoch": 2.3598778781325533, "ewc_loss": 0.1474609375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.0001201629638671875, "grad_norm": 7.469552993774414, "learning_rate": 1e-05, "loss": 0.462, "mean_token_accuracy": 0.8874914646148682, "num_tokens": 704467444.0, "step": 18551 }, { "epoch": 2.3600050884111434, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010967254638671875, "grad_norm": 6.643754005432129, "learning_rate": 1e-05, "loss": 0.4519, "mean_token_accuracy": 0.8894031643867493, "num_tokens": 704504179.0, "step": 18552 }, { "epoch": 2.360132298689734, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00011157989501953125, "grad_norm": 7.097484588623047, "learning_rate": 1e-05, "loss": 0.4729, "mean_token_accuracy": 0.8807665109634399, "num_tokens": 704545346.0, "step": 18553 }, { "epoch": 2.3602595089683245, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.584511756896973, "learning_rate": 1e-05, "loss": 0.5259, "mean_token_accuracy": 0.8661805391311646, "num_tokens": 704583453.0, "step": 18554 }, { "epoch": 2.360386719246915, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.798664093017578, "learning_rate": 1e-05, "loss": 0.484, "mean_token_accuracy": 0.8757210969924927, "num_tokens": 704617210.0, "step": 18555 }, { "epoch": 2.3605139295255055, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.4725751876831055, "learning_rate": 1e-05, "loss": 0.4504, "mean_token_accuracy": 0.8832775950431824, "num_tokens": 704651002.0, "step": 18556 }, { "epoch": 2.360641139804096, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.6180524826049805, "learning_rate": 1e-05, "loss": 0.4848, "mean_token_accuracy": 0.8749202489852905, "num_tokens": 704694718.0, "step": 18557 }, { "epoch": 2.3607683500826866, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.322971820831299, "learning_rate": 1e-05, "loss": 0.4432, "mean_token_accuracy": 0.8867874145507812, "num_tokens": 704730266.0, "step": 18558 }, { "epoch": 2.360895560361277, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.577898025512695, "learning_rate": 1e-05, "loss": 0.47, "mean_token_accuracy": 0.8813873529434204, "num_tokens": 704766422.0, "step": 18559 }, { "epoch": 2.3610227706398677, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.1607136726379395, "learning_rate": 1e-05, "loss": 0.4715, "mean_token_accuracy": 0.8788530826568604, "num_tokens": 704802133.0, "step": 18560 }, { "epoch": 2.361149980918458, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.662900924682617, "learning_rate": 1e-05, "loss": 0.4515, "mean_token_accuracy": 0.8857004046440125, "num_tokens": 704832108.0, "step": 18561 }, { "epoch": 2.3612771911970487, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 6.004124641418457, "learning_rate": 1e-05, "loss": 0.4842, "mean_token_accuracy": 0.8698222637176514, "num_tokens": 704867056.0, "step": 18562 }, { "epoch": 2.3614044014756392, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.823986530303955, "learning_rate": 1e-05, "loss": 0.5238, "mean_token_accuracy": 0.8641144037246704, "num_tokens": 704905565.0, "step": 18563 }, { "epoch": 2.3615316117542298, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.856655120849609, "learning_rate": 1e-05, "loss": 0.4625, "mean_token_accuracy": 0.8781650066375732, "num_tokens": 704949690.0, "step": 18564 }, { "epoch": 2.3616588220328203, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00011205673217773438, "grad_norm": 7.132354259490967, "learning_rate": 1e-05, "loss": 0.5623, "mean_token_accuracy": 0.8592650294303894, "num_tokens": 704985123.0, "step": 18565 }, { "epoch": 2.361786032311411, "ewc_loss": 0.11474609375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.835343837738037, "learning_rate": 1e-05, "loss": 0.4516, "mean_token_accuracy": 0.8816030621528625, "num_tokens": 705020317.0, "step": 18566 }, { "epoch": 2.3619132425900013, "ewc_loss": 0.1474609375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.0001201629638671875, "grad_norm": 7.450920581817627, "learning_rate": 1e-05, "loss": 0.4797, "mean_token_accuracy": 0.8829449415206909, "num_tokens": 705059025.0, "step": 18567 }, { "epoch": 2.362040452868592, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.030107498168945, "learning_rate": 1e-05, "loss": 0.4187, "mean_token_accuracy": 0.8925433158874512, "num_tokens": 705094157.0, "step": 18568 }, { "epoch": 2.3621676631471824, "ewc_loss": 0.150390625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.596965312957764, "learning_rate": 1e-05, "loss": 0.4947, "mean_token_accuracy": 0.8780359029769897, "num_tokens": 705133642.0, "step": 18569 }, { "epoch": 2.362294873425773, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.431219100952148, "learning_rate": 1e-05, "loss": 0.5138, "mean_token_accuracy": 0.8678774833679199, "num_tokens": 705165253.0, "step": 18570 }, { "epoch": 2.3624220837043635, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00011730194091796875, "grad_norm": 7.3042426109313965, "learning_rate": 1e-05, "loss": 0.4594, "mean_token_accuracy": 0.8871515989303589, "num_tokens": 705203847.0, "step": 18571 }, { "epoch": 2.362549293982954, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 6.624064922332764, "learning_rate": 1e-05, "loss": 0.5025, "mean_token_accuracy": 0.87221759557724, "num_tokens": 705241984.0, "step": 18572 }, { "epoch": 2.3626765042615445, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.81845235824585, "learning_rate": 1e-05, "loss": 0.519, "mean_token_accuracy": 0.8659573197364807, "num_tokens": 705278462.0, "step": 18573 }, { "epoch": 2.362803714540135, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.691288948059082, "learning_rate": 1e-05, "loss": 0.4622, "mean_token_accuracy": 0.8865319490432739, "num_tokens": 705313004.0, "step": 18574 }, { "epoch": 2.362930924818725, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.318082332611084, "learning_rate": 1e-05, "loss": 0.4356, "mean_token_accuracy": 0.8897996544837952, "num_tokens": 705350225.0, "step": 18575 }, { "epoch": 2.363058135097316, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.833334922790527, "learning_rate": 1e-05, "loss": 0.4647, "mean_token_accuracy": 0.8823881149291992, "num_tokens": 705380213.0, "step": 18576 }, { "epoch": 2.363185345375906, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.943893909454346, "learning_rate": 1e-05, "loss": 0.497, "mean_token_accuracy": 0.8686833381652832, "num_tokens": 705419059.0, "step": 18577 }, { "epoch": 2.3633125556544967, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00011157989501953125, "grad_norm": 7.106603145599365, "learning_rate": 1e-05, "loss": 0.4936, "mean_token_accuracy": 0.8738258481025696, "num_tokens": 705459506.0, "step": 18578 }, { "epoch": 2.3634397659330872, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.792915344238281, "learning_rate": 1e-05, "loss": 0.473, "mean_token_accuracy": 0.8772287368774414, "num_tokens": 705495719.0, "step": 18579 }, { "epoch": 2.3635669762116778, "ewc_loss": 0.1484375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00012063980102539062, "grad_norm": 7.488434791564941, "learning_rate": 1e-05, "loss": 0.5334, "mean_token_accuracy": 0.8623472452163696, "num_tokens": 705536040.0, "step": 18580 }, { "epoch": 2.3636941864902683, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 5.952898979187012, "learning_rate": 1e-05, "loss": 0.464, "mean_token_accuracy": 0.8769932985305786, "num_tokens": 705571541.0, "step": 18581 }, { "epoch": 2.363821396768859, "ewc_loss": 0.15625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00012874603271484375, "grad_norm": 7.733227729797363, "learning_rate": 1e-05, "loss": 0.4976, "mean_token_accuracy": 0.8805781602859497, "num_tokens": 705610048.0, "step": 18582 }, { "epoch": 2.3639486070474494, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.362927436828613, "learning_rate": 1e-05, "loss": 0.4544, "mean_token_accuracy": 0.8840522766113281, "num_tokens": 705645098.0, "step": 18583 }, { "epoch": 2.36407581732604, "ewc_loss": 0.150390625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.58235502243042, "learning_rate": 1e-05, "loss": 0.5008, "mean_token_accuracy": 0.8759839534759521, "num_tokens": 705680962.0, "step": 18584 }, { "epoch": 2.3642030276046304, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 6.618016242980957, "learning_rate": 1e-05, "loss": 0.493, "mean_token_accuracy": 0.8742332458496094, "num_tokens": 705716983.0, "step": 18585 }, { "epoch": 2.364330237883221, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00011444091796875, "grad_norm": 7.22641134262085, "learning_rate": 1e-05, "loss": 0.4942, "mean_token_accuracy": 0.8780679702758789, "num_tokens": 705749735.0, "step": 18586 }, { "epoch": 2.3644574481618115, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.609339714050293, "learning_rate": 1e-05, "loss": 0.5105, "mean_token_accuracy": 0.8691362738609314, "num_tokens": 705788747.0, "step": 18587 }, { "epoch": 2.364584658440402, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.892622947692871, "learning_rate": 1e-05, "loss": 0.5004, "mean_token_accuracy": 0.8721932172775269, "num_tokens": 705827475.0, "step": 18588 }, { "epoch": 2.3647118687189925, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.506687641143799, "learning_rate": 1e-05, "loss": 0.4437, "mean_token_accuracy": 0.8908641338348389, "num_tokens": 705864940.0, "step": 18589 }, { "epoch": 2.364839078997583, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.6536545753479, "learning_rate": 1e-05, "loss": 0.4185, "mean_token_accuracy": 0.8941214084625244, "num_tokens": 705898274.0, "step": 18590 }, { "epoch": 2.3649662892761736, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.378999710083008, "learning_rate": 1e-05, "loss": 0.4879, "mean_token_accuracy": 0.8740854263305664, "num_tokens": 705936619.0, "step": 18591 }, { "epoch": 2.365093499554764, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.557537078857422, "learning_rate": 1e-05, "loss": 0.4057, "mean_token_accuracy": 0.9011354446411133, "num_tokens": 705975208.0, "step": 18592 }, { "epoch": 2.3652207098333546, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.238304138183594, "learning_rate": 1e-05, "loss": 0.5069, "mean_token_accuracy": 0.8700331449508667, "num_tokens": 706019948.0, "step": 18593 }, { "epoch": 2.365347920111945, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.563117027282715, "learning_rate": 1e-05, "loss": 0.4594, "mean_token_accuracy": 0.8801511526107788, "num_tokens": 706056862.0, "step": 18594 }, { "epoch": 2.3654751303905357, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.098593711853027, "learning_rate": 1e-05, "loss": 0.4889, "mean_token_accuracy": 0.8723922967910767, "num_tokens": 706092591.0, "step": 18595 }, { "epoch": 2.3656023406691262, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.661848068237305, "learning_rate": 1e-05, "loss": 0.4983, "mean_token_accuracy": 0.8742591738700867, "num_tokens": 706132572.0, "step": 18596 }, { "epoch": 2.3657295509477168, "ewc_loss": 0.11376953125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.94172477722168, "learning_rate": 1e-05, "loss": 0.4504, "mean_token_accuracy": 0.8799737691879272, "num_tokens": 706168264.0, "step": 18597 }, { "epoch": 2.365856761226307, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.873162746429443, "learning_rate": 1e-05, "loss": 0.4975, "mean_token_accuracy": 0.8757733702659607, "num_tokens": 706206684.0, "step": 18598 }, { "epoch": 2.365983971504898, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.818000316619873, "learning_rate": 1e-05, "loss": 0.4155, "mean_token_accuracy": 0.8925032615661621, "num_tokens": 706247547.0, "step": 18599 }, { "epoch": 2.366111181783488, "ewc_loss": 0.140625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00011348724365234375, "grad_norm": 7.195784568786621, "learning_rate": 1e-05, "loss": 0.4485, "mean_token_accuracy": 0.8916712403297424, "num_tokens": 706286286.0, "step": 18600 }, { "epoch": 2.3662383920620784, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 5.819920063018799, "learning_rate": 1e-05, "loss": 0.474, "mean_token_accuracy": 0.8718700408935547, "num_tokens": 706329990.0, "step": 18601 }, { "epoch": 2.366365602340669, "ewc_loss": 0.1494140625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.0001220703125, "grad_norm": 7.5572614669799805, "learning_rate": 1e-05, "loss": 0.5088, "mean_token_accuracy": 0.8726774454116821, "num_tokens": 706370423.0, "step": 18602 }, { "epoch": 2.3664928126192595, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.099012851715088, "learning_rate": 1e-05, "loss": 0.4635, "mean_token_accuracy": 0.8813034892082214, "num_tokens": 706414380.0, "step": 18603 }, { "epoch": 2.36662002289785, "ewc_loss": 0.150390625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.629097938537598, "learning_rate": 1e-05, "loss": 0.5095, "mean_token_accuracy": 0.8770069479942322, "num_tokens": 706455403.0, "step": 18604 }, { "epoch": 2.3667472331764405, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.498788356781006, "learning_rate": 1e-05, "loss": 0.4214, "mean_token_accuracy": 0.8970977067947388, "num_tokens": 706490114.0, "step": 18605 }, { "epoch": 2.366874443455031, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00011587142944335938, "grad_norm": 7.287454128265381, "learning_rate": 1e-05, "loss": 0.4491, "mean_token_accuracy": 0.8900710344314575, "num_tokens": 706523926.0, "step": 18606 }, { "epoch": 2.3670016537336216, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 6.66917085647583, "learning_rate": 1e-05, "loss": 0.4417, "mean_token_accuracy": 0.8917738199234009, "num_tokens": 706562626.0, "step": 18607 }, { "epoch": 2.367128864012212, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.781253814697266, "learning_rate": 1e-05, "loss": 0.4415, "mean_token_accuracy": 0.8902148008346558, "num_tokens": 706598735.0, "step": 18608 }, { "epoch": 2.3672560742908026, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.723687171936035, "learning_rate": 1e-05, "loss": 0.523, "mean_token_accuracy": 0.8659056425094604, "num_tokens": 706638399.0, "step": 18609 }, { "epoch": 2.367383284569393, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.292115688323975, "learning_rate": 1e-05, "loss": 0.4693, "mean_token_accuracy": 0.877867579460144, "num_tokens": 706680274.0, "step": 18610 }, { "epoch": 2.3675104948479837, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.865301609039307, "learning_rate": 1e-05, "loss": 0.5275, "mean_token_accuracy": 0.8612107038497925, "num_tokens": 706713918.0, "step": 18611 }, { "epoch": 2.3676377051265742, "ewc_loss": 0.11376953125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.9603047370910645, "learning_rate": 1e-05, "loss": 0.4715, "mean_token_accuracy": 0.8754816651344299, "num_tokens": 706751758.0, "step": 18612 }, { "epoch": 2.3677649154051648, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00011110305786132812, "grad_norm": 7.117476463317871, "learning_rate": 1e-05, "loss": 0.4672, "mean_token_accuracy": 0.8874198198318481, "num_tokens": 706788243.0, "step": 18613 }, { "epoch": 2.3678921256837553, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.842076778411865, "learning_rate": 1e-05, "loss": 0.4466, "mean_token_accuracy": 0.885857105255127, "num_tokens": 706821960.0, "step": 18614 }, { "epoch": 2.368019335962346, "ewc_loss": 0.1455078125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00011777877807617188, "grad_norm": 7.4290900230407715, "learning_rate": 1e-05, "loss": 0.4793, "mean_token_accuracy": 0.8813170194625854, "num_tokens": 706858768.0, "step": 18615 }, { "epoch": 2.3681465462409363, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 5.979108810424805, "learning_rate": 1e-05, "loss": 0.4242, "mean_token_accuracy": 0.8919527530670166, "num_tokens": 706898889.0, "step": 18616 }, { "epoch": 2.368273756519527, "ewc_loss": 0.150390625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.644901752471924, "learning_rate": 1e-05, "loss": 0.4991, "mean_token_accuracy": 0.877210795879364, "num_tokens": 706934245.0, "step": 18617 }, { "epoch": 2.3684009667981174, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.323218822479248, "learning_rate": 1e-05, "loss": 0.4938, "mean_token_accuracy": 0.8738665580749512, "num_tokens": 706982092.0, "step": 18618 }, { "epoch": 2.368528177076708, "ewc_loss": 0.1494140625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00012159347534179688, "grad_norm": 7.529256343841553, "learning_rate": 1e-05, "loss": 0.4736, "mean_token_accuracy": 0.8843141794204712, "num_tokens": 707014748.0, "step": 18619 }, { "epoch": 2.3686553873552985, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.54683780670166, "learning_rate": 1e-05, "loss": 0.4654, "mean_token_accuracy": 0.8825454711914062, "num_tokens": 707048363.0, "step": 18620 }, { "epoch": 2.368782597633889, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00011491775512695312, "grad_norm": 7.227502346038818, "learning_rate": 1e-05, "loss": 0.5204, "mean_token_accuracy": 0.8700388669967651, "num_tokens": 707084089.0, "step": 18621 }, { "epoch": 2.3689098079124795, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.552341938018799, "learning_rate": 1e-05, "loss": 0.5072, "mean_token_accuracy": 0.8682394623756409, "num_tokens": 707127334.0, "step": 18622 }, { "epoch": 2.3690370181910696, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.921677589416504, "learning_rate": 1e-05, "loss": 0.4815, "mean_token_accuracy": 0.880711555480957, "num_tokens": 707166485.0, "step": 18623 }, { "epoch": 2.3691642284696606, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.472647190093994, "learning_rate": 1e-05, "loss": 0.4697, "mean_token_accuracy": 0.8783775568008423, "num_tokens": 707202031.0, "step": 18624 }, { "epoch": 2.3692914387482507, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.712994575500488, "learning_rate": 1e-05, "loss": 0.4691, "mean_token_accuracy": 0.8806205987930298, "num_tokens": 707240649.0, "step": 18625 }, { "epoch": 2.369418649026841, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.3521246910095215, "learning_rate": 1e-05, "loss": 0.4747, "mean_token_accuracy": 0.8771963119506836, "num_tokens": 707275996.0, "step": 18626 }, { "epoch": 2.3695458593054317, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.620732307434082, "learning_rate": 1e-05, "loss": 0.5351, "mean_token_accuracy": 0.8624736070632935, "num_tokens": 707312776.0, "step": 18627 }, { "epoch": 2.3696730695840222, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.2024335861206055, "learning_rate": 1e-05, "loss": 0.4955, "mean_token_accuracy": 0.8714807033538818, "num_tokens": 707349519.0, "step": 18628 }, { "epoch": 2.3698002798626128, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.647646427154541, "learning_rate": 1e-05, "loss": 0.4786, "mean_token_accuracy": 0.8752393126487732, "num_tokens": 707389180.0, "step": 18629 }, { "epoch": 2.3699274901412033, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.036892414093018, "learning_rate": 1e-05, "loss": 0.4617, "mean_token_accuracy": 0.8793349266052246, "num_tokens": 707428603.0, "step": 18630 }, { "epoch": 2.370054700419794, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.8053131103515625, "learning_rate": 1e-05, "loss": 0.5154, "mean_token_accuracy": 0.8648190498352051, "num_tokens": 707470297.0, "step": 18631 }, { "epoch": 2.3701819106983844, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.899392127990723, "learning_rate": 1e-05, "loss": 0.4647, "mean_token_accuracy": 0.8787884712219238, "num_tokens": 707507353.0, "step": 18632 }, { "epoch": 2.370309120976975, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00011014938354492188, "grad_norm": 7.072976589202881, "learning_rate": 1e-05, "loss": 0.4692, "mean_token_accuracy": 0.8858198523521423, "num_tokens": 707541795.0, "step": 18633 }, { "epoch": 2.3704363312555654, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.8572187423706055, "learning_rate": 1e-05, "loss": 0.4613, "mean_token_accuracy": 0.8750894069671631, "num_tokens": 707577333.0, "step": 18634 }, { "epoch": 2.370563541534156, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.0001163482666015625, "grad_norm": 7.38355827331543, "learning_rate": 1e-05, "loss": 0.5354, "mean_token_accuracy": 0.8652863502502441, "num_tokens": 707611306.0, "step": 18635 }, { "epoch": 2.3706907518127465, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.017708778381348, "learning_rate": 1e-05, "loss": 0.4883, "mean_token_accuracy": 0.8719372749328613, "num_tokens": 707647698.0, "step": 18636 }, { "epoch": 2.370817962091337, "ewc_loss": 0.1484375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00012111663818359375, "grad_norm": 7.517316818237305, "learning_rate": 1e-05, "loss": 0.488, "mean_token_accuracy": 0.8819853067398071, "num_tokens": 707683934.0, "step": 18637 }, { "epoch": 2.3709451723699275, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.34365701675415, "learning_rate": 1e-05, "loss": 0.4963, "mean_token_accuracy": 0.8719999194145203, "num_tokens": 707723216.0, "step": 18638 }, { "epoch": 2.371072382648518, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.0001163482666015625, "grad_norm": 7.284623146057129, "learning_rate": 1e-05, "loss": 0.4938, "mean_token_accuracy": 0.8812058568000793, "num_tokens": 707761194.0, "step": 18639 }, { "epoch": 2.3711995929271086, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.574385166168213, "learning_rate": 1e-05, "loss": 0.5289, "mean_token_accuracy": 0.8626279234886169, "num_tokens": 707807327.0, "step": 18640 }, { "epoch": 2.371326803205699, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.793440818786621, "learning_rate": 1e-05, "loss": 0.4823, "mean_token_accuracy": 0.8781951665878296, "num_tokens": 707846394.0, "step": 18641 }, { "epoch": 2.3714540134842896, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.674263954162598, "learning_rate": 1e-05, "loss": 0.4978, "mean_token_accuracy": 0.8730989694595337, "num_tokens": 707889099.0, "step": 18642 }, { "epoch": 2.37158122376288, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.2860188484191895, "learning_rate": 1e-05, "loss": 0.4367, "mean_token_accuracy": 0.8844250440597534, "num_tokens": 707924729.0, "step": 18643 }, { "epoch": 2.3717084340414707, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.854846000671387, "learning_rate": 1e-05, "loss": 0.5586, "mean_token_accuracy": 0.8548282980918884, "num_tokens": 707965096.0, "step": 18644 }, { "epoch": 2.371835644320061, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.894824504852295, "learning_rate": 1e-05, "loss": 0.457, "mean_token_accuracy": 0.8790835738182068, "num_tokens": 708002277.0, "step": 18645 }, { "epoch": 2.3719628545986517, "ewc_loss": 0.140625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00011348724365234375, "grad_norm": 7.203027725219727, "learning_rate": 1e-05, "loss": 0.5155, "mean_token_accuracy": 0.8714536428451538, "num_tokens": 708038978.0, "step": 18646 }, { "epoch": 2.3720900648772423, "ewc_loss": 0.11181640625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.739464282989502, "learning_rate": 1e-05, "loss": 0.4213, "mean_token_accuracy": 0.8915319442749023, "num_tokens": 708076022.0, "step": 18647 }, { "epoch": 2.3722172751558324, "ewc_loss": 0.15234375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00012493133544921875, "grad_norm": 7.656714916229248, "learning_rate": 1e-05, "loss": 0.5352, "mean_token_accuracy": 0.8639799952507019, "num_tokens": 708119606.0, "step": 18648 }, { "epoch": 2.3723444854344233, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.022481441497803, "learning_rate": 1e-05, "loss": 0.4356, "mean_token_accuracy": 0.8880217671394348, "num_tokens": 708159570.0, "step": 18649 }, { "epoch": 2.3724716957130134, "ewc_loss": 0.1572265625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00012969970703125, "grad_norm": 7.876043319702148, "learning_rate": 1e-05, "loss": 0.5542, "mean_token_accuracy": 0.862319827079773, "num_tokens": 708201092.0, "step": 18650 }, { "epoch": 2.372598905991604, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 6.558512210845947, "learning_rate": 1e-05, "loss": 0.4729, "mean_token_accuracy": 0.8804702758789062, "num_tokens": 708235808.0, "step": 18651 }, { "epoch": 2.3727261162701945, "ewc_loss": 0.150390625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.575254440307617, "learning_rate": 1e-05, "loss": 0.4518, "mean_token_accuracy": 0.8961420059204102, "num_tokens": 708270863.0, "step": 18652 }, { "epoch": 2.372853326548785, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.0001125335693359375, "grad_norm": 6.743760108947754, "learning_rate": 1e-05, "loss": 0.5369, "mean_token_accuracy": 0.8643076419830322, "num_tokens": 708309653.0, "step": 18653 }, { "epoch": 2.3729805368273755, "ewc_loss": 0.140625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00011348724365234375, "grad_norm": 7.152912616729736, "learning_rate": 1e-05, "loss": 0.4874, "mean_token_accuracy": 0.881850004196167, "num_tokens": 708346242.0, "step": 18654 }, { "epoch": 2.373107747105966, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 6.66921854019165, "learning_rate": 1e-05, "loss": 0.5193, "mean_token_accuracy": 0.8685208559036255, "num_tokens": 708383391.0, "step": 18655 }, { "epoch": 2.3732349573845566, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.815972328186035, "learning_rate": 1e-05, "loss": 0.4664, "mean_token_accuracy": 0.884697437286377, "num_tokens": 708417094.0, "step": 18656 }, { "epoch": 2.373362167663147, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.519725322723389, "learning_rate": 1e-05, "loss": 0.4912, "mean_token_accuracy": 0.8744723200798035, "num_tokens": 708456648.0, "step": 18657 }, { "epoch": 2.3734893779417376, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.627812385559082, "learning_rate": 1e-05, "loss": 0.5046, "mean_token_accuracy": 0.8687063455581665, "num_tokens": 708499125.0, "step": 18658 }, { "epoch": 2.373616588220328, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.340635776519775, "learning_rate": 1e-05, "loss": 0.4654, "mean_token_accuracy": 0.8792438507080078, "num_tokens": 708544844.0, "step": 18659 }, { "epoch": 2.3737437984989187, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.586978435516357, "learning_rate": 1e-05, "loss": 0.489, "mean_token_accuracy": 0.8730568885803223, "num_tokens": 708586852.0, "step": 18660 }, { "epoch": 2.3738710087775092, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.1716508865356445, "learning_rate": 1e-05, "loss": 0.4649, "mean_token_accuracy": 0.878935694694519, "num_tokens": 708624108.0, "step": 18661 }, { "epoch": 2.3739982190560998, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.658645153045654, "learning_rate": 1e-05, "loss": 0.4207, "mean_token_accuracy": 0.896020233631134, "num_tokens": 708658214.0, "step": 18662 }, { "epoch": 2.3741254293346903, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 5.995000839233398, "learning_rate": 1e-05, "loss": 0.4264, "mean_token_accuracy": 0.8916199207305908, "num_tokens": 708695315.0, "step": 18663 }, { "epoch": 2.374252639613281, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.857073783874512, "learning_rate": 1e-05, "loss": 0.502, "mean_token_accuracy": 0.8676784634590149, "num_tokens": 708734034.0, "step": 18664 }, { "epoch": 2.3743798498918713, "ewc_loss": 0.11376953125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.863458156585693, "learning_rate": 1e-05, "loss": 0.4658, "mean_token_accuracy": 0.8769534826278687, "num_tokens": 708769758.0, "step": 18665 }, { "epoch": 2.374507060170462, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.0001125335693359375, "grad_norm": 7.164361953735352, "learning_rate": 1e-05, "loss": 0.5171, "mean_token_accuracy": 0.8671029806137085, "num_tokens": 708803639.0, "step": 18666 }, { "epoch": 2.3746342704490524, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 5.852386474609375, "learning_rate": 1e-05, "loss": 0.4886, "mean_token_accuracy": 0.8730083107948303, "num_tokens": 708837085.0, "step": 18667 }, { "epoch": 2.374761480727643, "ewc_loss": 0.1474609375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.0001201629638671875, "grad_norm": 7.494751453399658, "learning_rate": 1e-05, "loss": 0.4908, "mean_token_accuracy": 0.8798857927322388, "num_tokens": 708876102.0, "step": 18668 }, { "epoch": 2.3748886910062335, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.053019046783447, "learning_rate": 1e-05, "loss": 0.4493, "mean_token_accuracy": 0.8881108164787292, "num_tokens": 708914260.0, "step": 18669 }, { "epoch": 2.375015901284824, "ewc_loss": 0.150390625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.639172554016113, "learning_rate": 1e-05, "loss": 0.5341, "mean_token_accuracy": 0.8652024269104004, "num_tokens": 708952071.0, "step": 18670 }, { "epoch": 2.3751431115634145, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.458831310272217, "learning_rate": 1e-05, "loss": 0.4983, "mean_token_accuracy": 0.87065589427948, "num_tokens": 708989394.0, "step": 18671 }, { "epoch": 2.375270321842005, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.0001163482666015625, "grad_norm": 7.339108467102051, "learning_rate": 1e-05, "loss": 0.5836, "mean_token_accuracy": 0.8530080318450928, "num_tokens": 709033558.0, "step": 18672 }, { "epoch": 2.375397532120595, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010967254638671875, "grad_norm": 6.666059970855713, "learning_rate": 1e-05, "loss": 0.4942, "mean_token_accuracy": 0.875032901763916, "num_tokens": 709071420.0, "step": 18673 }, { "epoch": 2.375524742399186, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.846979141235352, "learning_rate": 1e-05, "loss": 0.4508, "mean_token_accuracy": 0.8863139152526855, "num_tokens": 709104810.0, "step": 18674 }, { "epoch": 2.375651952677776, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.710511684417725, "learning_rate": 1e-05, "loss": 0.5019, "mean_token_accuracy": 0.8703154921531677, "num_tokens": 709142789.0, "step": 18675 }, { "epoch": 2.3757791629563667, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.3500189781188965, "learning_rate": 1e-05, "loss": 0.4474, "mean_token_accuracy": 0.8857560753822327, "num_tokens": 709177590.0, "step": 18676 }, { "epoch": 2.3759063732349572, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.827950477600098, "learning_rate": 1e-05, "loss": 0.4666, "mean_token_accuracy": 0.885022759437561, "num_tokens": 709217177.0, "step": 18677 }, { "epoch": 2.3760335835135478, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.990527153015137, "learning_rate": 1e-05, "loss": 0.5649, "mean_token_accuracy": 0.8465167284011841, "num_tokens": 709252566.0, "step": 18678 }, { "epoch": 2.3761607937921383, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 7.070631980895996, "learning_rate": 1e-05, "loss": 0.4416, "mean_token_accuracy": 0.8904468417167664, "num_tokens": 709290386.0, "step": 18679 }, { "epoch": 2.376288004070729, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.808088302612305, "learning_rate": 1e-05, "loss": 0.4552, "mean_token_accuracy": 0.8778055906295776, "num_tokens": 709331110.0, "step": 18680 }, { "epoch": 2.3764152143493193, "ewc_loss": 0.1455078125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00011777877807617188, "grad_norm": 7.425673007965088, "learning_rate": 1e-05, "loss": 0.4601, "mean_token_accuracy": 0.8892218470573425, "num_tokens": 709371390.0, "step": 18681 }, { "epoch": 2.37654242462791, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 5.926340103149414, "learning_rate": 1e-05, "loss": 0.4868, "mean_token_accuracy": 0.8754967451095581, "num_tokens": 709413428.0, "step": 18682 }, { "epoch": 2.3766696349065004, "ewc_loss": 0.1533203125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.000125885009765625, "grad_norm": 7.6996026039123535, "learning_rate": 1e-05, "loss": 0.4947, "mean_token_accuracy": 0.8777282238006592, "num_tokens": 709450332.0, "step": 18683 }, { "epoch": 2.376796845185091, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.312833309173584, "learning_rate": 1e-05, "loss": 0.545, "mean_token_accuracy": 0.8543975949287415, "num_tokens": 709490039.0, "step": 18684 }, { "epoch": 2.3769240554636815, "ewc_loss": 0.150390625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.598413467407227, "learning_rate": 1e-05, "loss": 0.4796, "mean_token_accuracy": 0.885665774345398, "num_tokens": 709531839.0, "step": 18685 }, { "epoch": 2.377051265742272, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 6.587127685546875, "learning_rate": 1e-05, "loss": 0.4726, "mean_token_accuracy": 0.8808483481407166, "num_tokens": 709568126.0, "step": 18686 }, { "epoch": 2.3771784760208625, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00011539459228515625, "grad_norm": 7.256138324737549, "learning_rate": 1e-05, "loss": 0.4837, "mean_token_accuracy": 0.8801883459091187, "num_tokens": 709610546.0, "step": 18687 }, { "epoch": 2.377305686299453, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.604687690734863, "learning_rate": 1e-05, "loss": 0.5122, "mean_token_accuracy": 0.8688046932220459, "num_tokens": 709656869.0, "step": 18688 }, { "epoch": 2.3774328965780436, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.920036315917969, "learning_rate": 1e-05, "loss": 0.4557, "mean_token_accuracy": 0.8859857320785522, "num_tokens": 709695691.0, "step": 18689 }, { "epoch": 2.377560106856634, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.505819320678711, "learning_rate": 1e-05, "loss": 0.4757, "mean_token_accuracy": 0.8762980699539185, "num_tokens": 709736691.0, "step": 18690 }, { "epoch": 2.3776873171352246, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.692082405090332, "learning_rate": 1e-05, "loss": 0.4962, "mean_token_accuracy": 0.8715562224388123, "num_tokens": 709776277.0, "step": 18691 }, { "epoch": 2.377814527413815, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.380808353424072, "learning_rate": 1e-05, "loss": 0.4971, "mean_token_accuracy": 0.8683465719223022, "num_tokens": 709817222.0, "step": 18692 }, { "epoch": 2.3779417376924057, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.596104145050049, "learning_rate": 1e-05, "loss": 0.4879, "mean_token_accuracy": 0.8734142780303955, "num_tokens": 709859060.0, "step": 18693 }, { "epoch": 2.378068947970996, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.227246284484863, "learning_rate": 1e-05, "loss": 0.4713, "mean_token_accuracy": 0.8760818243026733, "num_tokens": 709897396.0, "step": 18694 }, { "epoch": 2.3781961582495867, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.613104343414307, "learning_rate": 1e-05, "loss": 0.47, "mean_token_accuracy": 0.8799895644187927, "num_tokens": 709939797.0, "step": 18695 }, { "epoch": 2.378323368528177, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.067864418029785, "learning_rate": 1e-05, "loss": 0.4729, "mean_token_accuracy": 0.873863935470581, "num_tokens": 709977102.0, "step": 18696 }, { "epoch": 2.378450578806768, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.750871181488037, "learning_rate": 1e-05, "loss": 0.4879, "mean_token_accuracy": 0.875272274017334, "num_tokens": 710020824.0, "step": 18697 }, { "epoch": 2.378577789085358, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.91892671585083, "learning_rate": 1e-05, "loss": 0.4554, "mean_token_accuracy": 0.8781910538673401, "num_tokens": 710052263.0, "step": 18698 }, { "epoch": 2.3787049993639484, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.985648155212402, "learning_rate": 1e-05, "loss": 0.4715, "mean_token_accuracy": 0.878948450088501, "num_tokens": 710095196.0, "step": 18699 }, { "epoch": 2.378832209642539, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.843052387237549, "learning_rate": 1e-05, "loss": 0.4915, "mean_token_accuracy": 0.870578944683075, "num_tokens": 710127255.0, "step": 18700 }, { "epoch": 2.3789594199211295, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00011539459228515625, "grad_norm": 7.321290969848633, "learning_rate": 1e-05, "loss": 0.5008, "mean_token_accuracy": 0.8751153945922852, "num_tokens": 710159129.0, "step": 18701 }, { "epoch": 2.37908663019972, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 5.916446685791016, "learning_rate": 1e-05, "loss": 0.4818, "mean_token_accuracy": 0.8741598129272461, "num_tokens": 710197244.0, "step": 18702 }, { "epoch": 2.3792138404783105, "ewc_loss": 0.1484375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00012063980102539062, "grad_norm": 7.57697868347168, "learning_rate": 1e-05, "loss": 0.5639, "mean_token_accuracy": 0.8599543571472168, "num_tokens": 710241133.0, "step": 18703 }, { "epoch": 2.379341050756901, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.243202209472656, "learning_rate": 1e-05, "loss": 0.4815, "mean_token_accuracy": 0.8794736266136169, "num_tokens": 710283264.0, "step": 18704 }, { "epoch": 2.3794682610354916, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00011920928955078125, "grad_norm": 7.495625972747803, "learning_rate": 1e-05, "loss": 0.5184, "mean_token_accuracy": 0.8703610897064209, "num_tokens": 710329203.0, "step": 18705 }, { "epoch": 2.379595471314082, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.568922996520996, "learning_rate": 1e-05, "loss": 0.4624, "mean_token_accuracy": 0.8824614882469177, "num_tokens": 710366772.0, "step": 18706 }, { "epoch": 2.3797226815926726, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 7.0689520835876465, "learning_rate": 1e-05, "loss": 0.4977, "mean_token_accuracy": 0.8737049698829651, "num_tokens": 710407458.0, "step": 18707 }, { "epoch": 2.379849891871263, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.673616409301758, "learning_rate": 1e-05, "loss": 0.4741, "mean_token_accuracy": 0.878300666809082, "num_tokens": 710449200.0, "step": 18708 }, { "epoch": 2.3799771021498537, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.535482406616211, "learning_rate": 1e-05, "loss": 0.422, "mean_token_accuracy": 0.8976612687110901, "num_tokens": 710490764.0, "step": 18709 }, { "epoch": 2.3801043124284442, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.779275417327881, "learning_rate": 1e-05, "loss": 0.4908, "mean_token_accuracy": 0.8764101266860962, "num_tokens": 710526282.0, "step": 18710 }, { "epoch": 2.3802315227070348, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 6.084754467010498, "learning_rate": 1e-05, "loss": 0.449, "mean_token_accuracy": 0.883761465549469, "num_tokens": 710566523.0, "step": 18711 }, { "epoch": 2.3803587329856253, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 6.99835205078125, "learning_rate": 1e-05, "loss": 0.4724, "mean_token_accuracy": 0.8803917169570923, "num_tokens": 710607640.0, "step": 18712 }, { "epoch": 2.380485943264216, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.8352861404418945, "learning_rate": 1e-05, "loss": 0.4668, "mean_token_accuracy": 0.87501460313797, "num_tokens": 710644785.0, "step": 18713 }, { "epoch": 2.3806131535428063, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00011539459228515625, "grad_norm": 7.358537197113037, "learning_rate": 1e-05, "loss": 0.4829, "mean_token_accuracy": 0.8797686696052551, "num_tokens": 710681857.0, "step": 18714 }, { "epoch": 2.380740363821397, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 5.86856746673584, "learning_rate": 1e-05, "loss": 0.4422, "mean_token_accuracy": 0.8851281404495239, "num_tokens": 710718119.0, "step": 18715 }, { "epoch": 2.3808675740999874, "ewc_loss": 0.15234375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00012493133544921875, "grad_norm": 7.676350116729736, "learning_rate": 1e-05, "loss": 0.5575, "mean_token_accuracy": 0.86448734998703, "num_tokens": 710758573.0, "step": 18716 }, { "epoch": 2.380994784378578, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.222832679748535, "learning_rate": 1e-05, "loss": 0.4404, "mean_token_accuracy": 0.8908440470695496, "num_tokens": 710791567.0, "step": 18717 }, { "epoch": 2.3811219946571685, "ewc_loss": 0.15234375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00012493133544921875, "grad_norm": 7.66785192489624, "learning_rate": 1e-05, "loss": 0.4864, "mean_token_accuracy": 0.8834452629089355, "num_tokens": 710833060.0, "step": 18718 }, { "epoch": 2.381249204935759, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 6.56310510635376, "learning_rate": 1e-05, "loss": 0.507, "mean_token_accuracy": 0.871737003326416, "num_tokens": 710872175.0, "step": 18719 }, { "epoch": 2.3813764152143495, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.0001163482666015625, "grad_norm": 7.344468593597412, "learning_rate": 1e-05, "loss": 0.4774, "mean_token_accuracy": 0.8806958198547363, "num_tokens": 710910442.0, "step": 18720 }, { "epoch": 2.3815036254929396, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 6.626455307006836, "learning_rate": 1e-05, "loss": 0.4844, "mean_token_accuracy": 0.877336859703064, "num_tokens": 710948656.0, "step": 18721 }, { "epoch": 2.3816308357715306, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 6.987557411193848, "learning_rate": 1e-05, "loss": 0.5094, "mean_token_accuracy": 0.873234748840332, "num_tokens": 710989300.0, "step": 18722 }, { "epoch": 2.3817580460501206, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.552539825439453, "learning_rate": 1e-05, "loss": 0.456, "mean_token_accuracy": 0.8894622921943665, "num_tokens": 711027661.0, "step": 18723 }, { "epoch": 2.381885256328711, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.724766731262207, "learning_rate": 1e-05, "loss": 0.5108, "mean_token_accuracy": 0.8673699498176575, "num_tokens": 711067895.0, "step": 18724 }, { "epoch": 2.3820124666073017, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.419963359832764, "learning_rate": 1e-05, "loss": 0.4948, "mean_token_accuracy": 0.8707510232925415, "num_tokens": 711105567.0, "step": 18725 }, { "epoch": 2.3821396768858922, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.604728698730469, "learning_rate": 1e-05, "loss": 0.4313, "mean_token_accuracy": 0.8911811709403992, "num_tokens": 711145263.0, "step": 18726 }, { "epoch": 2.3822668871644828, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.264247894287109, "learning_rate": 1e-05, "loss": 0.4624, "mean_token_accuracy": 0.8816981315612793, "num_tokens": 711187024.0, "step": 18727 }, { "epoch": 2.3823940974430733, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.60872745513916, "learning_rate": 1e-05, "loss": 0.4778, "mean_token_accuracy": 0.8751285076141357, "num_tokens": 711226261.0, "step": 18728 }, { "epoch": 2.382521307721664, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.105345249176025, "learning_rate": 1e-05, "loss": 0.4834, "mean_token_accuracy": 0.8699514865875244, "num_tokens": 711267005.0, "step": 18729 }, { "epoch": 2.3826485180002543, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.740105152130127, "learning_rate": 1e-05, "loss": 0.5119, "mean_token_accuracy": 0.8699248433113098, "num_tokens": 711307145.0, "step": 18730 }, { "epoch": 2.382775728278845, "ewc_loss": 0.11376953125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.930386543273926, "learning_rate": 1e-05, "loss": 0.4575, "mean_token_accuracy": 0.8808309435844421, "num_tokens": 711351128.0, "step": 18731 }, { "epoch": 2.3829029385574354, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 6.993070602416992, "learning_rate": 1e-05, "loss": 0.4681, "mean_token_accuracy": 0.8842750787734985, "num_tokens": 711385344.0, "step": 18732 }, { "epoch": 2.383030148836026, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.832493305206299, "learning_rate": 1e-05, "loss": 0.4779, "mean_token_accuracy": 0.8721283674240112, "num_tokens": 711421548.0, "step": 18733 }, { "epoch": 2.3831573591146165, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00011587142944335938, "grad_norm": 7.3251261711120605, "learning_rate": 1e-05, "loss": 0.4791, "mean_token_accuracy": 0.8802805542945862, "num_tokens": 711456952.0, "step": 18734 }, { "epoch": 2.383284569393207, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 5.925030708312988, "learning_rate": 1e-05, "loss": 0.4508, "mean_token_accuracy": 0.8812054395675659, "num_tokens": 711486985.0, "step": 18735 }, { "epoch": 2.3834117796717975, "ewc_loss": 0.1494140625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.0001220703125, "grad_norm": 7.588573455810547, "learning_rate": 1e-05, "loss": 0.4639, "mean_token_accuracy": 0.893460750579834, "num_tokens": 711521047.0, "step": 18736 }, { "epoch": 2.383538989950388, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.242018222808838, "learning_rate": 1e-05, "loss": 0.4623, "mean_token_accuracy": 0.8852779865264893, "num_tokens": 711561151.0, "step": 18737 }, { "epoch": 2.3836662002289786, "ewc_loss": 0.1484375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00012111663818359375, "grad_norm": 7.518139362335205, "learning_rate": 1e-05, "loss": 0.5563, "mean_token_accuracy": 0.8589183688163757, "num_tokens": 711597975.0, "step": 18738 }, { "epoch": 2.383793410507569, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.575149059295654, "learning_rate": 1e-05, "loss": 0.4836, "mean_token_accuracy": 0.8762200474739075, "num_tokens": 711637923.0, "step": 18739 }, { "epoch": 2.3839206207861596, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 7.081024646759033, "learning_rate": 1e-05, "loss": 0.5175, "mean_token_accuracy": 0.8735792636871338, "num_tokens": 711675408.0, "step": 18740 }, { "epoch": 2.38404783106475, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 6.688399314880371, "learning_rate": 1e-05, "loss": 0.4627, "mean_token_accuracy": 0.8850606083869934, "num_tokens": 711712781.0, "step": 18741 }, { "epoch": 2.3841750413433407, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.545285701751709, "learning_rate": 1e-05, "loss": 0.4785, "mean_token_accuracy": 0.8789397478103638, "num_tokens": 711752974.0, "step": 18742 }, { "epoch": 2.384302251621931, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.794845104217529, "learning_rate": 1e-05, "loss": 0.4803, "mean_token_accuracy": 0.8782665729522705, "num_tokens": 711793526.0, "step": 18743 }, { "epoch": 2.3844294619005217, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 6.082113742828369, "learning_rate": 1e-05, "loss": 0.4167, "mean_token_accuracy": 0.8918830156326294, "num_tokens": 711830409.0, "step": 18744 }, { "epoch": 2.3845566721791123, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 7.025228023529053, "learning_rate": 1e-05, "loss": 0.4588, "mean_token_accuracy": 0.8869950175285339, "num_tokens": 711873476.0, "step": 18745 }, { "epoch": 2.3846838824577024, "ewc_loss": 0.11083984375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.828988552093506, "learning_rate": 1e-05, "loss": 0.448, "mean_token_accuracy": 0.8862631916999817, "num_tokens": 711912507.0, "step": 18746 }, { "epoch": 2.3848110927362933, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00011730194091796875, "grad_norm": 7.414608478546143, "learning_rate": 1e-05, "loss": 0.4655, "mean_token_accuracy": 0.8871436715126038, "num_tokens": 711946884.0, "step": 18747 }, { "epoch": 2.3849383030148834, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 5.890753746032715, "learning_rate": 1e-05, "loss": 0.496, "mean_token_accuracy": 0.8665356636047363, "num_tokens": 711987060.0, "step": 18748 }, { "epoch": 2.385065513293474, "ewc_loss": 0.1533203125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.000125885009765625, "grad_norm": 7.719967365264893, "learning_rate": 1e-05, "loss": 0.4556, "mean_token_accuracy": 0.8934241533279419, "num_tokens": 712023197.0, "step": 18749 }, { "epoch": 2.3851927235720645, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.266943454742432, "learning_rate": 1e-05, "loss": 0.4409, "mean_token_accuracy": 0.8883889317512512, "num_tokens": 712060337.0, "step": 18750 }, { "epoch": 2.385319933850655, "ewc_loss": 0.1513671875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.0001239776611328125, "grad_norm": 7.679935932159424, "learning_rate": 1e-05, "loss": 0.4758, "mean_token_accuracy": 0.8880602121353149, "num_tokens": 712101601.0, "step": 18751 }, { "epoch": 2.3854471441292455, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 6.611349105834961, "learning_rate": 1e-05, "loss": 0.4528, "mean_token_accuracy": 0.8865104913711548, "num_tokens": 712136311.0, "step": 18752 }, { "epoch": 2.385574354407836, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00011730194091796875, "grad_norm": 7.3319172859191895, "learning_rate": 1e-05, "loss": 0.5182, "mean_token_accuracy": 0.8694097995758057, "num_tokens": 712174386.0, "step": 18753 }, { "epoch": 2.3857015646864266, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 6.652518272399902, "learning_rate": 1e-05, "loss": 0.4891, "mean_token_accuracy": 0.8817631006240845, "num_tokens": 712209540.0, "step": 18754 }, { "epoch": 2.385828774965017, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 6.986209869384766, "learning_rate": 1e-05, "loss": 0.5057, "mean_token_accuracy": 0.8708364963531494, "num_tokens": 712245405.0, "step": 18755 }, { "epoch": 2.3859559852436076, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.558285713195801, "learning_rate": 1e-05, "loss": 0.5389, "mean_token_accuracy": 0.856464147567749, "num_tokens": 712285013.0, "step": 18756 }, { "epoch": 2.386083195522198, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.720693111419678, "learning_rate": 1e-05, "loss": 0.5195, "mean_token_accuracy": 0.8606811165809631, "num_tokens": 712324850.0, "step": 18757 }, { "epoch": 2.3862104058007887, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.397803783416748, "learning_rate": 1e-05, "loss": 0.4939, "mean_token_accuracy": 0.8731095790863037, "num_tokens": 712364336.0, "step": 18758 }, { "epoch": 2.386337616079379, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.636383533477783, "learning_rate": 1e-05, "loss": 0.533, "mean_token_accuracy": 0.861892580986023, "num_tokens": 712403773.0, "step": 18759 }, { "epoch": 2.3864648263579697, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.237191677093506, "learning_rate": 1e-05, "loss": 0.5145, "mean_token_accuracy": 0.8639493584632874, "num_tokens": 712440307.0, "step": 18760 }, { "epoch": 2.3865920366365603, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.657006740570068, "learning_rate": 1e-05, "loss": 0.467, "mean_token_accuracy": 0.8812536001205444, "num_tokens": 712472008.0, "step": 18761 }, { "epoch": 2.386719246915151, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 6.067803859710693, "learning_rate": 1e-05, "loss": 0.4424, "mean_token_accuracy": 0.8831249475479126, "num_tokens": 712512935.0, "step": 18762 }, { "epoch": 2.3868464571937413, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.793586254119873, "learning_rate": 1e-05, "loss": 0.4581, "mean_token_accuracy": 0.8851759433746338, "num_tokens": 712552783.0, "step": 18763 }, { "epoch": 2.386973667472332, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.92490816116333, "learning_rate": 1e-05, "loss": 0.3977, "mean_token_accuracy": 0.8977527022361755, "num_tokens": 712583752.0, "step": 18764 }, { "epoch": 2.3871008777509224, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 7.052119731903076, "learning_rate": 1e-05, "loss": 0.5378, "mean_token_accuracy": 0.8654090166091919, "num_tokens": 712623125.0, "step": 18765 }, { "epoch": 2.387228088029513, "ewc_loss": 0.11376953125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.849584102630615, "learning_rate": 1e-05, "loss": 0.4861, "mean_token_accuracy": 0.8698420524597168, "num_tokens": 712664965.0, "step": 18766 }, { "epoch": 2.3873552983081034, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00011682510375976562, "grad_norm": 7.381678104400635, "learning_rate": 1e-05, "loss": 0.5084, "mean_token_accuracy": 0.8738014101982117, "num_tokens": 712703571.0, "step": 18767 }, { "epoch": 2.387482508586694, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 5.988626480102539, "learning_rate": 1e-05, "loss": 0.4621, "mean_token_accuracy": 0.8802810907363892, "num_tokens": 712739830.0, "step": 18768 }, { "epoch": 2.3876097188652845, "ewc_loss": 0.1494140625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00012159347534179688, "grad_norm": 7.575904369354248, "learning_rate": 1e-05, "loss": 0.5209, "mean_token_accuracy": 0.8714529275894165, "num_tokens": 712782367.0, "step": 18769 }, { "epoch": 2.387736929143875, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.331637859344482, "learning_rate": 1e-05, "loss": 0.4705, "mean_token_accuracy": 0.8799446821212769, "num_tokens": 712816937.0, "step": 18770 }, { "epoch": 2.387864139422465, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00011873245239257812, "grad_norm": 7.405287742614746, "learning_rate": 1e-05, "loss": 0.4641, "mean_token_accuracy": 0.8880034685134888, "num_tokens": 712850198.0, "step": 18771 }, { "epoch": 2.387991349701056, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.589761734008789, "learning_rate": 1e-05, "loss": 0.4734, "mean_token_accuracy": 0.8806083798408508, "num_tokens": 712886201.0, "step": 18772 }, { "epoch": 2.388118559979646, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.947797775268555, "learning_rate": 1e-05, "loss": 0.5206, "mean_token_accuracy": 0.8702531456947327, "num_tokens": 712924309.0, "step": 18773 }, { "epoch": 2.3882457702582367, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.672812461853027, "learning_rate": 1e-05, "loss": 0.4417, "mean_token_accuracy": 0.8930943608283997, "num_tokens": 712964508.0, "step": 18774 }, { "epoch": 2.3883729805368272, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.432864189147949, "learning_rate": 1e-05, "loss": 0.4953, "mean_token_accuracy": 0.8701643943786621, "num_tokens": 713002737.0, "step": 18775 }, { "epoch": 2.3885001908154178, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.79245662689209, "learning_rate": 1e-05, "loss": 0.4853, "mean_token_accuracy": 0.8774250149726868, "num_tokens": 713040859.0, "step": 18776 }, { "epoch": 2.3886274010940083, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 6.018359184265137, "learning_rate": 1e-05, "loss": 0.4487, "mean_token_accuracy": 0.8806477785110474, "num_tokens": 713073504.0, "step": 18777 }, { "epoch": 2.388754611372599, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 7.047400951385498, "learning_rate": 1e-05, "loss": 0.4758, "mean_token_accuracy": 0.8803724050521851, "num_tokens": 713110285.0, "step": 18778 }, { "epoch": 2.3888818216511893, "ewc_loss": 0.11083984375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.788370132446289, "learning_rate": 1e-05, "loss": 0.4788, "mean_token_accuracy": 0.8714348673820496, "num_tokens": 713146227.0, "step": 18779 }, { "epoch": 2.38900903192978, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00011873245239257812, "grad_norm": 7.451048374176025, "learning_rate": 1e-05, "loss": 0.5123, "mean_token_accuracy": 0.87443608045578, "num_tokens": 713187632.0, "step": 18780 }, { "epoch": 2.3891362422083704, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 5.8906660079956055, "learning_rate": 1e-05, "loss": 0.4785, "mean_token_accuracy": 0.8740876317024231, "num_tokens": 713223023.0, "step": 18781 }, { "epoch": 2.389263452486961, "ewc_loss": 0.15625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00012874603271484375, "grad_norm": 7.774294853210449, "learning_rate": 1e-05, "loss": 0.5276, "mean_token_accuracy": 0.8684265613555908, "num_tokens": 713264390.0, "step": 18782 }, { "epoch": 2.3893906627655515, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.320376396179199, "learning_rate": 1e-05, "loss": 0.454, "mean_token_accuracy": 0.8846548795700073, "num_tokens": 713306504.0, "step": 18783 }, { "epoch": 2.389517873044142, "ewc_loss": 0.15234375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00012493133544921875, "grad_norm": 7.696119785308838, "learning_rate": 1e-05, "loss": 0.5006, "mean_token_accuracy": 0.8783268928527832, "num_tokens": 713347556.0, "step": 18784 }, { "epoch": 2.3896450833227325, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010967254638671875, "grad_norm": 6.664906024932861, "learning_rate": 1e-05, "loss": 0.5615, "mean_token_accuracy": 0.8507698178291321, "num_tokens": 713386645.0, "step": 18785 }, { "epoch": 2.389772293601323, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00011730194091796875, "grad_norm": 7.313115119934082, "learning_rate": 1e-05, "loss": 0.4798, "mean_token_accuracy": 0.881108283996582, "num_tokens": 713419754.0, "step": 18786 }, { "epoch": 2.3898995038799136, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 6.681045055389404, "learning_rate": 1e-05, "loss": 0.5001, "mean_token_accuracy": 0.8765860795974731, "num_tokens": 713456374.0, "step": 18787 }, { "epoch": 2.390026714158504, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 6.944581985473633, "learning_rate": 1e-05, "loss": 0.4784, "mean_token_accuracy": 0.8790546655654907, "num_tokens": 713487961.0, "step": 18788 }, { "epoch": 2.3901539244370946, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.563711643218994, "learning_rate": 1e-05, "loss": 0.4861, "mean_token_accuracy": 0.8766187429428101, "num_tokens": 713525697.0, "step": 18789 }, { "epoch": 2.390281134715685, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.697263717651367, "learning_rate": 1e-05, "loss": 0.4897, "mean_token_accuracy": 0.8753784894943237, "num_tokens": 713563902.0, "step": 18790 }, { "epoch": 2.3904083449942757, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.413727760314941, "learning_rate": 1e-05, "loss": 0.5162, "mean_token_accuracy": 0.8684511184692383, "num_tokens": 713602996.0, "step": 18791 }, { "epoch": 2.390535555272866, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.606668949127197, "learning_rate": 1e-05, "loss": 0.4871, "mean_token_accuracy": 0.8774522542953491, "num_tokens": 713641725.0, "step": 18792 }, { "epoch": 2.3906627655514567, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.239367961883545, "learning_rate": 1e-05, "loss": 0.5264, "mean_token_accuracy": 0.8560722470283508, "num_tokens": 713682973.0, "step": 18793 }, { "epoch": 2.390789975830047, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.654794692993164, "learning_rate": 1e-05, "loss": 0.4723, "mean_token_accuracy": 0.8821092844009399, "num_tokens": 713718919.0, "step": 18794 }, { "epoch": 2.390917186108638, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 6.06328010559082, "learning_rate": 1e-05, "loss": 0.4813, "mean_token_accuracy": 0.8734103441238403, "num_tokens": 713757416.0, "step": 18795 }, { "epoch": 2.391044396387228, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.805642127990723, "learning_rate": 1e-05, "loss": 0.5375, "mean_token_accuracy": 0.8596519231796265, "num_tokens": 713795648.0, "step": 18796 }, { "epoch": 2.3911716066658184, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.891615867614746, "learning_rate": 1e-05, "loss": 0.4119, "mean_token_accuracy": 0.8958667516708374, "num_tokens": 713830095.0, "step": 18797 }, { "epoch": 2.391298816944409, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00011014938354492188, "grad_norm": 7.099374771118164, "learning_rate": 1e-05, "loss": 0.5143, "mean_token_accuracy": 0.8697813153266907, "num_tokens": 713868958.0, "step": 18798 }, { "epoch": 2.3914260272229995, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.836729526519775, "learning_rate": 1e-05, "loss": 0.508, "mean_token_accuracy": 0.8665062189102173, "num_tokens": 713911533.0, "step": 18799 }, { "epoch": 2.39155323750159, "ewc_loss": 0.1474609375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00011968612670898438, "grad_norm": 7.459787368774414, "learning_rate": 1e-05, "loss": 0.5292, "mean_token_accuracy": 0.868356466293335, "num_tokens": 713952585.0, "step": 18800 }, { "epoch": 2.3916804477801805, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.018659591674805, "learning_rate": 1e-05, "loss": 0.4235, "mean_token_accuracy": 0.8941611051559448, "num_tokens": 713990929.0, "step": 18801 }, { "epoch": 2.391807658058771, "ewc_loss": 0.1513671875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.0001239776611328125, "grad_norm": 7.643935203552246, "learning_rate": 1e-05, "loss": 0.5254, "mean_token_accuracy": 0.8673720359802246, "num_tokens": 714027098.0, "step": 18802 }, { "epoch": 2.3919348683373616, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.406840801239014, "learning_rate": 1e-05, "loss": 0.504, "mean_token_accuracy": 0.8679221868515015, "num_tokens": 714062444.0, "step": 18803 }, { "epoch": 2.392062078615952, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00011873245239257812, "grad_norm": 7.404685974121094, "learning_rate": 1e-05, "loss": 0.4429, "mean_token_accuracy": 0.8950462341308594, "num_tokens": 714099091.0, "step": 18804 }, { "epoch": 2.3921892888945426, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 6.647085189819336, "learning_rate": 1e-05, "loss": 0.5045, "mean_token_accuracy": 0.8716987371444702, "num_tokens": 714135694.0, "step": 18805 }, { "epoch": 2.392316499173133, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.93029260635376, "learning_rate": 1e-05, "loss": 0.4532, "mean_token_accuracy": 0.8904656171798706, "num_tokens": 714170324.0, "step": 18806 }, { "epoch": 2.3924437094517237, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 6.707603931427002, "learning_rate": 1e-05, "loss": 0.4815, "mean_token_accuracy": 0.8779489994049072, "num_tokens": 714205080.0, "step": 18807 }, { "epoch": 2.392570919730314, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.424582004547119, "learning_rate": 1e-05, "loss": 0.4764, "mean_token_accuracy": 0.8745130896568298, "num_tokens": 714245297.0, "step": 18808 }, { "epoch": 2.3926981300089047, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.807363033294678, "learning_rate": 1e-05, "loss": 0.45, "mean_token_accuracy": 0.8896405100822449, "num_tokens": 714284155.0, "step": 18809 }, { "epoch": 2.3928253402874953, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 6.022912979125977, "learning_rate": 1e-05, "loss": 0.4495, "mean_token_accuracy": 0.8812360763549805, "num_tokens": 714318688.0, "step": 18810 }, { "epoch": 2.392952550566086, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 7.053158760070801, "learning_rate": 1e-05, "loss": 0.4346, "mean_token_accuracy": 0.8930186033248901, "num_tokens": 714353730.0, "step": 18811 }, { "epoch": 2.3930797608446763, "ewc_loss": 0.11376953125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.8281779289245605, "learning_rate": 1e-05, "loss": 0.468, "mean_token_accuracy": 0.8758664131164551, "num_tokens": 714390156.0, "step": 18812 }, { "epoch": 2.393206971123267, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00011873245239257812, "grad_norm": 7.401054859161377, "learning_rate": 1e-05, "loss": 0.5356, "mean_token_accuracy": 0.8624894618988037, "num_tokens": 714429435.0, "step": 18813 }, { "epoch": 2.3933341814018574, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 5.890071868896484, "learning_rate": 1e-05, "loss": 0.4148, "mean_token_accuracy": 0.8965165019035339, "num_tokens": 714470034.0, "step": 18814 }, { "epoch": 2.393461391680448, "ewc_loss": 0.154296875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00012683868408203125, "grad_norm": 7.717559337615967, "learning_rate": 1e-05, "loss": 0.5033, "mean_token_accuracy": 0.8748353719711304, "num_tokens": 714505989.0, "step": 18815 }, { "epoch": 2.3935886019590384, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.2733154296875, "learning_rate": 1e-05, "loss": 0.5046, "mean_token_accuracy": 0.8715823888778687, "num_tokens": 714541727.0, "step": 18816 }, { "epoch": 2.393715812237629, "ewc_loss": 0.1533203125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.000125885009765625, "grad_norm": 7.676419258117676, "learning_rate": 1e-05, "loss": 0.504, "mean_token_accuracy": 0.8787823915481567, "num_tokens": 714581429.0, "step": 18817 }, { "epoch": 2.3938430225162195, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 6.592806816101074, "learning_rate": 1e-05, "loss": 0.4865, "mean_token_accuracy": 0.8770973682403564, "num_tokens": 714621839.0, "step": 18818 }, { "epoch": 2.3939702327948096, "ewc_loss": 0.1455078125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00011777877807617188, "grad_norm": 7.338217735290527, "learning_rate": 1e-05, "loss": 0.4996, "mean_token_accuracy": 0.8758090734481812, "num_tokens": 714665459.0, "step": 18819 }, { "epoch": 2.3940974430734006, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 6.662814617156982, "learning_rate": 1e-05, "loss": 0.4753, "mean_token_accuracy": 0.8828465938568115, "num_tokens": 714699259.0, "step": 18820 }, { "epoch": 2.3942246533519906, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 6.967464447021484, "learning_rate": 1e-05, "loss": 0.5091, "mean_token_accuracy": 0.8686954975128174, "num_tokens": 714740330.0, "step": 18821 }, { "epoch": 2.394351863630581, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.582751750946045, "learning_rate": 1e-05, "loss": 0.4696, "mean_token_accuracy": 0.8805707097053528, "num_tokens": 714773330.0, "step": 18822 }, { "epoch": 2.3944790739091717, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.688541889190674, "learning_rate": 1e-05, "loss": 0.4959, "mean_token_accuracy": 0.8678145408630371, "num_tokens": 714812707.0, "step": 18823 }, { "epoch": 2.3946062841877622, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.447418212890625, "learning_rate": 1e-05, "loss": 0.4624, "mean_token_accuracy": 0.8808209300041199, "num_tokens": 714853496.0, "step": 18824 }, { "epoch": 2.3947334944663528, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.55251932144165, "learning_rate": 1e-05, "loss": 0.4614, "mean_token_accuracy": 0.8823262453079224, "num_tokens": 714895766.0, "step": 18825 }, { "epoch": 2.3948607047449433, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.30890417098999, "learning_rate": 1e-05, "loss": 0.4629, "mean_token_accuracy": 0.8794206380844116, "num_tokens": 714936089.0, "step": 18826 }, { "epoch": 2.394987915023534, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.533968925476074, "learning_rate": 1e-05, "loss": 0.4296, "mean_token_accuracy": 0.8917008638381958, "num_tokens": 714980759.0, "step": 18827 }, { "epoch": 2.3951151253021243, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.151430130004883, "learning_rate": 1e-05, "loss": 0.4415, "mean_token_accuracy": 0.88267982006073, "num_tokens": 715019671.0, "step": 18828 }, { "epoch": 2.395242335580715, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.6419677734375, "learning_rate": 1e-05, "loss": 0.5378, "mean_token_accuracy": 0.8585816025733948, "num_tokens": 715059639.0, "step": 18829 }, { "epoch": 2.3953695458593054, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.967962741851807, "learning_rate": 1e-05, "loss": 0.4315, "mean_token_accuracy": 0.8894565105438232, "num_tokens": 715096315.0, "step": 18830 }, { "epoch": 2.395496756137896, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.864330768585205, "learning_rate": 1e-05, "loss": 0.4766, "mean_token_accuracy": 0.8773500323295593, "num_tokens": 715130800.0, "step": 18831 }, { "epoch": 2.3956239664164865, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.8187479972839355, "learning_rate": 1e-05, "loss": 0.4747, "mean_token_accuracy": 0.8721849322319031, "num_tokens": 715175369.0, "step": 18832 }, { "epoch": 2.395751176695077, "ewc_loss": 0.140625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00011301040649414062, "grad_norm": 7.205207347869873, "learning_rate": 1e-05, "loss": 0.5088, "mean_token_accuracy": 0.872582733631134, "num_tokens": 715216480.0, "step": 18833 }, { "epoch": 2.3958783869736675, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.796498775482178, "learning_rate": 1e-05, "loss": 0.4394, "mean_token_accuracy": 0.8845802545547485, "num_tokens": 715257182.0, "step": 18834 }, { "epoch": 2.396005597252258, "ewc_loss": 0.1494140625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.0001220703125, "grad_norm": 7.6265177726745605, "learning_rate": 1e-05, "loss": 0.5156, "mean_token_accuracy": 0.872133731842041, "num_tokens": 715288793.0, "step": 18835 }, { "epoch": 2.3961328075308486, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.093600273132324, "learning_rate": 1e-05, "loss": 0.4683, "mean_token_accuracy": 0.8768293857574463, "num_tokens": 715323720.0, "step": 18836 }, { "epoch": 2.396260017809439, "ewc_loss": 0.1533203125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.000125885009765625, "grad_norm": 7.759546279907227, "learning_rate": 1e-05, "loss": 0.5399, "mean_token_accuracy": 0.8713160157203674, "num_tokens": 715356104.0, "step": 18837 }, { "epoch": 2.3963872280880296, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 6.550908088684082, "learning_rate": 1e-05, "loss": 0.4669, "mean_token_accuracy": 0.8857111930847168, "num_tokens": 715390727.0, "step": 18838 }, { "epoch": 2.39651443836662, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00011873245239257812, "grad_norm": 7.413904666900635, "learning_rate": 1e-05, "loss": 0.4896, "mean_token_accuracy": 0.8817611336708069, "num_tokens": 715426340.0, "step": 18839 }, { "epoch": 2.3966416486452107, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00011014938354492188, "grad_norm": 6.744193077087402, "learning_rate": 1e-05, "loss": 0.4032, "mean_token_accuracy": 0.903535008430481, "num_tokens": 715458195.0, "step": 18840 }, { "epoch": 2.396768858923801, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.878488540649414, "learning_rate": 1e-05, "loss": 0.4535, "mean_token_accuracy": 0.8886719346046448, "num_tokens": 715496896.0, "step": 18841 }, { "epoch": 2.3968960692023917, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.7668914794921875, "learning_rate": 1e-05, "loss": 0.5164, "mean_token_accuracy": 0.864523708820343, "num_tokens": 715537985.0, "step": 18842 }, { "epoch": 2.3970232794809823, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.387492656707764, "learning_rate": 1e-05, "loss": 0.4145, "mean_token_accuracy": 0.8921498656272888, "num_tokens": 715574767.0, "step": 18843 }, { "epoch": 2.3971504897595723, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.858747959136963, "learning_rate": 1e-05, "loss": 0.4772, "mean_token_accuracy": 0.8761854767799377, "num_tokens": 715608649.0, "step": 18844 }, { "epoch": 2.3972777000381633, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 6.045083045959473, "learning_rate": 1e-05, "loss": 0.4635, "mean_token_accuracy": 0.8773446679115295, "num_tokens": 715650278.0, "step": 18845 }, { "epoch": 2.3974049103167534, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 7.052146911621094, "learning_rate": 1e-05, "loss": 0.495, "mean_token_accuracy": 0.8726837635040283, "num_tokens": 715692678.0, "step": 18846 }, { "epoch": 2.397532120595344, "ewc_loss": 0.11474609375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.891721725463867, "learning_rate": 1e-05, "loss": 0.4942, "mean_token_accuracy": 0.868045449256897, "num_tokens": 715733128.0, "step": 18847 }, { "epoch": 2.3976593308739345, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00011539459228515625, "grad_norm": 7.338456630706787, "learning_rate": 1e-05, "loss": 0.5144, "mean_token_accuracy": 0.8745764493942261, "num_tokens": 715767272.0, "step": 18848 }, { "epoch": 2.397786541152525, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 5.954759120941162, "learning_rate": 1e-05, "loss": 0.4033, "mean_token_accuracy": 0.8964138031005859, "num_tokens": 715803497.0, "step": 18849 }, { "epoch": 2.3979137514311155, "ewc_loss": 0.1494140625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00012159347534179688, "grad_norm": 7.541524887084961, "learning_rate": 1e-05, "loss": 0.4914, "mean_token_accuracy": 0.8809809684753418, "num_tokens": 715847069.0, "step": 18850 }, { "epoch": 2.398040961709706, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.204080104827881, "learning_rate": 1e-05, "loss": 0.4627, "mean_token_accuracy": 0.883047342300415, "num_tokens": 715887639.0, "step": 18851 }, { "epoch": 2.3981681719882966, "ewc_loss": 0.1494140625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00012159347534179688, "grad_norm": 7.560194969177246, "learning_rate": 1e-05, "loss": 0.4338, "mean_token_accuracy": 0.8975309133529663, "num_tokens": 715919427.0, "step": 18852 }, { "epoch": 2.398295382266887, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.499821186065674, "learning_rate": 1e-05, "loss": 0.5362, "mean_token_accuracy": 0.8625832796096802, "num_tokens": 715961287.0, "step": 18853 }, { "epoch": 2.3984225925454776, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00011539459228515625, "grad_norm": 7.263375282287598, "learning_rate": 1e-05, "loss": 0.5055, "mean_token_accuracy": 0.8747245073318481, "num_tokens": 715995021.0, "step": 18854 }, { "epoch": 2.398549802824068, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.507837772369385, "learning_rate": 1e-05, "loss": 0.4416, "mean_token_accuracy": 0.8923280239105225, "num_tokens": 716039517.0, "step": 18855 }, { "epoch": 2.3986770131026587, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.967403888702393, "learning_rate": 1e-05, "loss": 0.5249, "mean_token_accuracy": 0.8623629808425903, "num_tokens": 716077116.0, "step": 18856 }, { "epoch": 2.398804223381249, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.479026794433594, "learning_rate": 1e-05, "loss": 0.5013, "mean_token_accuracy": 0.8683856725692749, "num_tokens": 716112800.0, "step": 18857 }, { "epoch": 2.3989314336598397, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.724246501922607, "learning_rate": 1e-05, "loss": 0.4492, "mean_token_accuracy": 0.8886063694953918, "num_tokens": 716148165.0, "step": 18858 }, { "epoch": 2.3990586439384303, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.370058536529541, "learning_rate": 1e-05, "loss": 0.4936, "mean_token_accuracy": 0.8733733892440796, "num_tokens": 716185745.0, "step": 18859 }, { "epoch": 2.399185854217021, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.604273796081543, "learning_rate": 1e-05, "loss": 0.4324, "mean_token_accuracy": 0.8910883665084839, "num_tokens": 716225177.0, "step": 18860 }, { "epoch": 2.3993130644956113, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.24690055847168, "learning_rate": 1e-05, "loss": 0.421, "mean_token_accuracy": 0.8970956802368164, "num_tokens": 716265969.0, "step": 18861 }, { "epoch": 2.399440274774202, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.61074686050415, "learning_rate": 1e-05, "loss": 0.4743, "mean_token_accuracy": 0.87819504737854, "num_tokens": 716302056.0, "step": 18862 }, { "epoch": 2.3995674850527924, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.101013660430908, "learning_rate": 1e-05, "loss": 0.4944, "mean_token_accuracy": 0.8675256967544556, "num_tokens": 716339452.0, "step": 18863 }, { "epoch": 2.399694695331383, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.731093406677246, "learning_rate": 1e-05, "loss": 0.4805, "mean_token_accuracy": 0.8784527778625488, "num_tokens": 716379930.0, "step": 18864 }, { "epoch": 2.3998219056099734, "ewc_loss": 0.11474609375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.95259428024292, "learning_rate": 1e-05, "loss": 0.4673, "mean_token_accuracy": 0.874387264251709, "num_tokens": 716416367.0, "step": 18865 }, { "epoch": 2.399949115888564, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.948906421661377, "learning_rate": 1e-05, "loss": 0.5392, "mean_token_accuracy": 0.8603296279907227, "num_tokens": 716460112.0, "step": 18866 }, { "epoch": 2.4000763261671545, "ewc_loss": 0.11474609375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.8592448234558105, "learning_rate": 1e-05, "loss": 0.4682, "mean_token_accuracy": 0.8729274272918701, "num_tokens": 716497085.0, "step": 18867 }, { "epoch": 2.400203536445745, "ewc_loss": 0.140625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00011348724365234375, "grad_norm": 7.234233379364014, "learning_rate": 1e-05, "loss": 0.4878, "mean_token_accuracy": 0.8788662552833557, "num_tokens": 716542892.0, "step": 18868 }, { "epoch": 2.400330746724335, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 5.9072980880737305, "learning_rate": 1e-05, "loss": 0.4997, "mean_token_accuracy": 0.8646754026412964, "num_tokens": 716578765.0, "step": 18869 }, { "epoch": 2.400457957002926, "ewc_loss": 0.1474609375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.0001201629638671875, "grad_norm": 7.483093738555908, "learning_rate": 1e-05, "loss": 0.495, "mean_token_accuracy": 0.8790967464447021, "num_tokens": 716618881.0, "step": 18870 }, { "epoch": 2.400585167281516, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.16357421875, "learning_rate": 1e-05, "loss": 0.4697, "mean_token_accuracy": 0.8763583898544312, "num_tokens": 716651583.0, "step": 18871 }, { "epoch": 2.4007123775601067, "ewc_loss": 0.1474609375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00011968612670898438, "grad_norm": 7.47056770324707, "learning_rate": 1e-05, "loss": 0.5259, "mean_token_accuracy": 0.8662116527557373, "num_tokens": 716687118.0, "step": 18872 }, { "epoch": 2.400839587838697, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.458878993988037, "learning_rate": 1e-05, "loss": 0.5095, "mean_token_accuracy": 0.8678537011146545, "num_tokens": 716722543.0, "step": 18873 }, { "epoch": 2.4009667981172877, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.0001125335693359375, "grad_norm": 7.1144561767578125, "learning_rate": 1e-05, "loss": 0.479, "mean_token_accuracy": 0.8799657821655273, "num_tokens": 716757447.0, "step": 18874 }, { "epoch": 2.4010940083958783, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.594632625579834, "learning_rate": 1e-05, "loss": 0.4835, "mean_token_accuracy": 0.8761450052261353, "num_tokens": 716799736.0, "step": 18875 }, { "epoch": 2.401221218674469, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.620265960693359, "learning_rate": 1e-05, "loss": 0.5032, "mean_token_accuracy": 0.8729604482650757, "num_tokens": 716844850.0, "step": 18876 }, { "epoch": 2.4013484289530593, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.704469680786133, "learning_rate": 1e-05, "loss": 0.4944, "mean_token_accuracy": 0.8752647042274475, "num_tokens": 716878801.0, "step": 18877 }, { "epoch": 2.40147563923165, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.1468048095703125, "learning_rate": 1e-05, "loss": 0.4663, "mean_token_accuracy": 0.876698911190033, "num_tokens": 716914271.0, "step": 18878 }, { "epoch": 2.4016028495102404, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.92144250869751, "learning_rate": 1e-05, "loss": 0.4935, "mean_token_accuracy": 0.8750784397125244, "num_tokens": 716954094.0, "step": 18879 }, { "epoch": 2.401730059788831, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.826824188232422, "learning_rate": 1e-05, "loss": 0.4557, "mean_token_accuracy": 0.8801227807998657, "num_tokens": 716995061.0, "step": 18880 }, { "epoch": 2.4018572700674214, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00011491775512695312, "grad_norm": 7.301053524017334, "learning_rate": 1e-05, "loss": 0.4971, "mean_token_accuracy": 0.8755372762680054, "num_tokens": 717032271.0, "step": 18881 }, { "epoch": 2.401984480346012, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.787109851837158, "learning_rate": 1e-05, "loss": 0.4544, "mean_token_accuracy": 0.8808305263519287, "num_tokens": 717068551.0, "step": 18882 }, { "epoch": 2.4021116906246025, "ewc_loss": 0.1552734375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.0001277923583984375, "grad_norm": 7.723216533660889, "learning_rate": 1e-05, "loss": 0.4986, "mean_token_accuracy": 0.881044864654541, "num_tokens": 717107460.0, "step": 18883 }, { "epoch": 2.402238900903193, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.146553993225098, "learning_rate": 1e-05, "loss": 0.4876, "mean_token_accuracy": 0.8735440969467163, "num_tokens": 717144774.0, "step": 18884 }, { "epoch": 2.4023661111817836, "ewc_loss": 0.15625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00012874603271484375, "grad_norm": 7.820432186126709, "learning_rate": 1e-05, "loss": 0.5046, "mean_token_accuracy": 0.8745596408843994, "num_tokens": 717180808.0, "step": 18885 }, { "epoch": 2.402493321460374, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010967254638671875, "grad_norm": 6.601217746734619, "learning_rate": 1e-05, "loss": 0.4639, "mean_token_accuracy": 0.8841280937194824, "num_tokens": 717220788.0, "step": 18886 }, { "epoch": 2.4026205317389646, "ewc_loss": 0.1484375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00012063980102539062, "grad_norm": 7.503393650054932, "learning_rate": 1e-05, "loss": 0.4913, "mean_token_accuracy": 0.8827467560768127, "num_tokens": 717260257.0, "step": 18887 }, { "epoch": 2.402747742017555, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00011014938354492188, "grad_norm": 6.730428695678711, "learning_rate": 1e-05, "loss": 0.4631, "mean_token_accuracy": 0.8834978342056274, "num_tokens": 717291911.0, "step": 18888 }, { "epoch": 2.4028749522961457, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00011157989501953125, "grad_norm": 7.113419532775879, "learning_rate": 1e-05, "loss": 0.5043, "mean_token_accuracy": 0.8741512298583984, "num_tokens": 717322356.0, "step": 18889 }, { "epoch": 2.403002162574736, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.629374027252197, "learning_rate": 1e-05, "loss": 0.4981, "mean_token_accuracy": 0.8762410879135132, "num_tokens": 717363439.0, "step": 18890 }, { "epoch": 2.4031293728533267, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.792750358581543, "learning_rate": 1e-05, "loss": 0.4862, "mean_token_accuracy": 0.8740037083625793, "num_tokens": 717401050.0, "step": 18891 }, { "epoch": 2.403256583131917, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.496726036071777, "learning_rate": 1e-05, "loss": 0.4905, "mean_token_accuracy": 0.8746458888053894, "num_tokens": 717442346.0, "step": 18892 }, { "epoch": 2.403383793410508, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.621187210083008, "learning_rate": 1e-05, "loss": 0.4665, "mean_token_accuracy": 0.8803364038467407, "num_tokens": 717483387.0, "step": 18893 }, { "epoch": 2.403511003689098, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.335417747497559, "learning_rate": 1e-05, "loss": 0.4757, "mean_token_accuracy": 0.878467321395874, "num_tokens": 717522208.0, "step": 18894 }, { "epoch": 2.4036382139676884, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.5825276374816895, "learning_rate": 1e-05, "loss": 0.4561, "mean_token_accuracy": 0.8879566788673401, "num_tokens": 717563338.0, "step": 18895 }, { "epoch": 2.403765424246279, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.178916931152344, "learning_rate": 1e-05, "loss": 0.4787, "mean_token_accuracy": 0.8760837912559509, "num_tokens": 717603351.0, "step": 18896 }, { "epoch": 2.4038926345248695, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.659299373626709, "learning_rate": 1e-05, "loss": 0.4368, "mean_token_accuracy": 0.8893133401870728, "num_tokens": 717639886.0, "step": 18897 }, { "epoch": 2.40401984480346, "ewc_loss": 0.11474609375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 6.033206462860107, "learning_rate": 1e-05, "loss": 0.5439, "mean_token_accuracy": 0.8547669649124146, "num_tokens": 717672228.0, "step": 18898 }, { "epoch": 2.4041470550820505, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.83569860458374, "learning_rate": 1e-05, "loss": 0.5118, "mean_token_accuracy": 0.8691068291664124, "num_tokens": 717709703.0, "step": 18899 }, { "epoch": 2.404274265360641, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.869811058044434, "learning_rate": 1e-05, "loss": 0.4734, "mean_token_accuracy": 0.8736011981964111, "num_tokens": 717746785.0, "step": 18900 }, { "epoch": 2.4044014756392316, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00011157989501953125, "grad_norm": 7.154306888580322, "learning_rate": 1e-05, "loss": 0.5079, "mean_token_accuracy": 0.8712683916091919, "num_tokens": 717785786.0, "step": 18901 }, { "epoch": 2.404528685917822, "ewc_loss": 0.11376953125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.8350443840026855, "learning_rate": 1e-05, "loss": 0.472, "mean_token_accuracy": 0.8736516237258911, "num_tokens": 717821833.0, "step": 18902 }, { "epoch": 2.4046558961964126, "ewc_loss": 0.1484375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00012063980102539062, "grad_norm": 7.509984970092773, "learning_rate": 1e-05, "loss": 0.4856, "mean_token_accuracy": 0.88367760181427, "num_tokens": 717858107.0, "step": 18903 }, { "epoch": 2.404783106475003, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.04594612121582, "learning_rate": 1e-05, "loss": 0.4245, "mean_token_accuracy": 0.8926030397415161, "num_tokens": 717893090.0, "step": 18904 }, { "epoch": 2.4049103167535937, "ewc_loss": 0.15234375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00012493133544921875, "grad_norm": 7.6805548667907715, "learning_rate": 1e-05, "loss": 0.4743, "mean_token_accuracy": 0.8879170417785645, "num_tokens": 717919703.0, "step": 18905 }, { "epoch": 2.405037527032184, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.45534086227417, "learning_rate": 1e-05, "loss": 0.453, "mean_token_accuracy": 0.8886466026306152, "num_tokens": 717956877.0, "step": 18906 }, { "epoch": 2.4051647373107747, "ewc_loss": 0.1455078125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00011777877807617188, "grad_norm": 7.406579971313477, "learning_rate": 1e-05, "loss": 0.5358, "mean_token_accuracy": 0.8669384121894836, "num_tokens": 718002812.0, "step": 18907 }, { "epoch": 2.4052919475893653, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 6.688199520111084, "learning_rate": 1e-05, "loss": 0.4928, "mean_token_accuracy": 0.8786637783050537, "num_tokens": 718041170.0, "step": 18908 }, { "epoch": 2.405419157867956, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.895800590515137, "learning_rate": 1e-05, "loss": 0.4958, "mean_token_accuracy": 0.8734339475631714, "num_tokens": 718084168.0, "step": 18909 }, { "epoch": 2.4055463681465463, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.747486114501953, "learning_rate": 1e-05, "loss": 0.4526, "mean_token_accuracy": 0.8876070976257324, "num_tokens": 718118663.0, "step": 18910 }, { "epoch": 2.405673578425137, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.390557765960693, "learning_rate": 1e-05, "loss": 0.4601, "mean_token_accuracy": 0.8775216341018677, "num_tokens": 718155490.0, "step": 18911 }, { "epoch": 2.4058007887037274, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.859393119812012, "learning_rate": 1e-05, "loss": 0.4974, "mean_token_accuracy": 0.8733981847763062, "num_tokens": 718192301.0, "step": 18912 }, { "epoch": 2.405927998982318, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 6.007817268371582, "learning_rate": 1e-05, "loss": 0.437, "mean_token_accuracy": 0.8873059749603271, "num_tokens": 718231390.0, "step": 18913 }, { "epoch": 2.4060552092609084, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 7.1128458976745605, "learning_rate": 1e-05, "loss": 0.4536, "mean_token_accuracy": 0.8888596296310425, "num_tokens": 718269248.0, "step": 18914 }, { "epoch": 2.406182419539499, "ewc_loss": 0.11376953125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.844103813171387, "learning_rate": 1e-05, "loss": 0.4593, "mean_token_accuracy": 0.8809435367584229, "num_tokens": 718305961.0, "step": 18915 }, { "epoch": 2.4063096298180895, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00011873245239257812, "grad_norm": 7.462557315826416, "learning_rate": 1e-05, "loss": 0.5228, "mean_token_accuracy": 0.8717186450958252, "num_tokens": 718344659.0, "step": 18916 }, { "epoch": 2.4064368400966796, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 5.968233585357666, "learning_rate": 1e-05, "loss": 0.516, "mean_token_accuracy": 0.864151120185852, "num_tokens": 718386624.0, "step": 18917 }, { "epoch": 2.4065640503752705, "ewc_loss": 0.15234375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00012493133544921875, "grad_norm": 7.700601577758789, "learning_rate": 1e-05, "loss": 0.4577, "mean_token_accuracy": 0.8903274536132812, "num_tokens": 718422154.0, "step": 18918 }, { "epoch": 2.4066912606538606, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.3379807472229, "learning_rate": 1e-05, "loss": 0.5191, "mean_token_accuracy": 0.8621383905410767, "num_tokens": 718460253.0, "step": 18919 }, { "epoch": 2.406818470932451, "ewc_loss": 0.150390625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.586819648742676, "learning_rate": 1e-05, "loss": 0.5056, "mean_token_accuracy": 0.879968523979187, "num_tokens": 718495328.0, "step": 18920 }, { "epoch": 2.4069456812110417, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 6.595244884490967, "learning_rate": 1e-05, "loss": 0.4813, "mean_token_accuracy": 0.8798204660415649, "num_tokens": 718531244.0, "step": 18921 }, { "epoch": 2.407072891489632, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00011491775512695312, "grad_norm": 7.257966041564941, "learning_rate": 1e-05, "loss": 0.4914, "mean_token_accuracy": 0.8792614936828613, "num_tokens": 718569766.0, "step": 18922 }, { "epoch": 2.4072001017682227, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.591762542724609, "learning_rate": 1e-05, "loss": 0.4973, "mean_token_accuracy": 0.8768712282180786, "num_tokens": 718612496.0, "step": 18923 }, { "epoch": 2.4073273120468133, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 6.9419708251953125, "learning_rate": 1e-05, "loss": 0.4748, "mean_token_accuracy": 0.8779571652412415, "num_tokens": 718651251.0, "step": 18924 }, { "epoch": 2.407454522325404, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.519460201263428, "learning_rate": 1e-05, "loss": 0.4885, "mean_token_accuracy": 0.8738739490509033, "num_tokens": 718683204.0, "step": 18925 }, { "epoch": 2.4075817326039943, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.7118024826049805, "learning_rate": 1e-05, "loss": 0.4417, "mean_token_accuracy": 0.8875728845596313, "num_tokens": 718717114.0, "step": 18926 }, { "epoch": 2.407708942882585, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.385132789611816, "learning_rate": 1e-05, "loss": 0.5048, "mean_token_accuracy": 0.8684449791908264, "num_tokens": 718754756.0, "step": 18927 }, { "epoch": 2.4078361531611754, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.61814022064209, "learning_rate": 1e-05, "loss": 0.5146, "mean_token_accuracy": 0.8688728213310242, "num_tokens": 718792312.0, "step": 18928 }, { "epoch": 2.407963363439766, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.231118202209473, "learning_rate": 1e-05, "loss": 0.4826, "mean_token_accuracy": 0.8740483522415161, "num_tokens": 718834771.0, "step": 18929 }, { "epoch": 2.4080905737183564, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.671199798583984, "learning_rate": 1e-05, "loss": 0.4977, "mean_token_accuracy": 0.8730257153511047, "num_tokens": 718871124.0, "step": 18930 }, { "epoch": 2.408217783996947, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.074664115905762, "learning_rate": 1e-05, "loss": 0.4674, "mean_token_accuracy": 0.8754992485046387, "num_tokens": 718912037.0, "step": 18931 }, { "epoch": 2.4083449942755375, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.795781135559082, "learning_rate": 1e-05, "loss": 0.5188, "mean_token_accuracy": 0.8686795830726624, "num_tokens": 718953048.0, "step": 18932 }, { "epoch": 2.408472204554128, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.933791160583496, "learning_rate": 1e-05, "loss": 0.4881, "mean_token_accuracy": 0.8700299859046936, "num_tokens": 718991825.0, "step": 18933 }, { "epoch": 2.4085994148327186, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 7.052064418792725, "learning_rate": 1e-05, "loss": 0.4942, "mean_token_accuracy": 0.8757708072662354, "num_tokens": 719026847.0, "step": 18934 }, { "epoch": 2.408726625111309, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.864601135253906, "learning_rate": 1e-05, "loss": 0.4468, "mean_token_accuracy": 0.881507158279419, "num_tokens": 719061415.0, "step": 18935 }, { "epoch": 2.4088538353898996, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00011730194091796875, "grad_norm": 7.367892265319824, "learning_rate": 1e-05, "loss": 0.4876, "mean_token_accuracy": 0.8823332786560059, "num_tokens": 719097432.0, "step": 18936 }, { "epoch": 2.40898104566849, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 5.998744964599609, "learning_rate": 1e-05, "loss": 0.4619, "mean_token_accuracy": 0.8801178932189941, "num_tokens": 719129301.0, "step": 18937 }, { "epoch": 2.4091082559470807, "ewc_loss": 0.150390625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.546683311462402, "learning_rate": 1e-05, "loss": 0.5048, "mean_token_accuracy": 0.8758458495140076, "num_tokens": 719171772.0, "step": 18938 }, { "epoch": 2.409235466225671, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.318460941314697, "learning_rate": 1e-05, "loss": 0.4999, "mean_token_accuracy": 0.8707805871963501, "num_tokens": 719211751.0, "step": 18939 }, { "epoch": 2.4093626765042617, "ewc_loss": 0.1455078125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00011777877807617188, "grad_norm": 7.369589805603027, "learning_rate": 1e-05, "loss": 0.4465, "mean_token_accuracy": 0.894798994064331, "num_tokens": 719246812.0, "step": 18940 }, { "epoch": 2.4094898867828523, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.563559532165527, "learning_rate": 1e-05, "loss": 0.5393, "mean_token_accuracy": 0.8586959838867188, "num_tokens": 719290962.0, "step": 18941 }, { "epoch": 2.4096170970614423, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.915859699249268, "learning_rate": 1e-05, "loss": 0.4922, "mean_token_accuracy": 0.8782035708427429, "num_tokens": 719323506.0, "step": 18942 }, { "epoch": 2.4097443073400333, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.672497272491455, "learning_rate": 1e-05, "loss": 0.5179, "mean_token_accuracy": 0.8676852583885193, "num_tokens": 719361435.0, "step": 18943 }, { "epoch": 2.4098715176186234, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.394961357116699, "learning_rate": 1e-05, "loss": 0.4408, "mean_token_accuracy": 0.8881004452705383, "num_tokens": 719405617.0, "step": 18944 }, { "epoch": 2.409998727897214, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.80357027053833, "learning_rate": 1e-05, "loss": 0.4866, "mean_token_accuracy": 0.8757278919219971, "num_tokens": 719450588.0, "step": 18945 }, { "epoch": 2.4101259381758044, "ewc_loss": 0.11474609375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.972701072692871, "learning_rate": 1e-05, "loss": 0.4302, "mean_token_accuracy": 0.8861753940582275, "num_tokens": 719489028.0, "step": 18946 }, { "epoch": 2.410253148454395, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00011157989501953125, "grad_norm": 7.119937419891357, "learning_rate": 1e-05, "loss": 0.5191, "mean_token_accuracy": 0.8720318675041199, "num_tokens": 719527921.0, "step": 18947 }, { "epoch": 2.4103803587329855, "ewc_loss": 0.11181640625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.749392032623291, "learning_rate": 1e-05, "loss": 0.4553, "mean_token_accuracy": 0.8789204359054565, "num_tokens": 719572461.0, "step": 18948 }, { "epoch": 2.410507569011576, "ewc_loss": 0.150390625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.56856632232666, "learning_rate": 1e-05, "loss": 0.5022, "mean_token_accuracy": 0.8741917014122009, "num_tokens": 719613478.0, "step": 18949 }, { "epoch": 2.4106347792901666, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 5.9359283447265625, "learning_rate": 1e-05, "loss": 0.4316, "mean_token_accuracy": 0.8897702097892761, "num_tokens": 719643665.0, "step": 18950 }, { "epoch": 2.410761989568757, "ewc_loss": 0.15625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00012874603271484375, "grad_norm": 7.864120006561279, "learning_rate": 1e-05, "loss": 0.5265, "mean_token_accuracy": 0.8672308921813965, "num_tokens": 719684168.0, "step": 18951 }, { "epoch": 2.4108891998473476, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.445831298828125, "learning_rate": 1e-05, "loss": 0.481, "mean_token_accuracy": 0.8765190243721008, "num_tokens": 719722484.0, "step": 18952 }, { "epoch": 2.411016410125938, "ewc_loss": 0.15234375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00012493133544921875, "grad_norm": 7.669708728790283, "learning_rate": 1e-05, "loss": 0.4881, "mean_token_accuracy": 0.8793555498123169, "num_tokens": 719756659.0, "step": 18953 }, { "epoch": 2.4111436204045287, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.0001125335693359375, "grad_norm": 6.750988006591797, "learning_rate": 1e-05, "loss": 0.4956, "mean_token_accuracy": 0.8775413036346436, "num_tokens": 719787302.0, "step": 18954 }, { "epoch": 2.411270830683119, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00011539459228515625, "grad_norm": 7.227377414703369, "learning_rate": 1e-05, "loss": 0.4898, "mean_token_accuracy": 0.8799879550933838, "num_tokens": 719823359.0, "step": 18955 }, { "epoch": 2.4113980409617097, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 6.702589511871338, "learning_rate": 1e-05, "loss": 0.4881, "mean_token_accuracy": 0.8781888484954834, "num_tokens": 719861263.0, "step": 18956 }, { "epoch": 2.4115252512403003, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.848864555358887, "learning_rate": 1e-05, "loss": 0.4639, "mean_token_accuracy": 0.8805826902389526, "num_tokens": 719897987.0, "step": 18957 }, { "epoch": 2.411652461518891, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.580169677734375, "learning_rate": 1e-05, "loss": 0.4601, "mean_token_accuracy": 0.8837171792984009, "num_tokens": 719931067.0, "step": 18958 }, { "epoch": 2.4117796717974813, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.625577449798584, "learning_rate": 1e-05, "loss": 0.4417, "mean_token_accuracy": 0.886992871761322, "num_tokens": 719971174.0, "step": 18959 }, { "epoch": 2.411906882076072, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.418663501739502, "learning_rate": 1e-05, "loss": 0.4666, "mean_token_accuracy": 0.8835489749908447, "num_tokens": 720005781.0, "step": 18960 }, { "epoch": 2.4120340923546624, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.536661148071289, "learning_rate": 1e-05, "loss": 0.4398, "mean_token_accuracy": 0.8857381343841553, "num_tokens": 720040917.0, "step": 18961 }, { "epoch": 2.412161302633253, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.262328147888184, "learning_rate": 1e-05, "loss": 0.5066, "mean_token_accuracy": 0.8647085428237915, "num_tokens": 720080490.0, "step": 18962 }, { "epoch": 2.4122885129118434, "ewc_loss": 0.125, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.553186893463135, "learning_rate": 1e-05, "loss": 0.4789, "mean_token_accuracy": 0.8769544959068298, "num_tokens": 720116461.0, "step": 18963 }, { "epoch": 2.412415723190434, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.091937065124512, "learning_rate": 1e-05, "loss": 0.3982, "mean_token_accuracy": 0.8972804546356201, "num_tokens": 720156756.0, "step": 18964 }, { "epoch": 2.4125429334690245, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.677281379699707, "learning_rate": 1e-05, "loss": 0.4868, "mean_token_accuracy": 0.8770069479942322, "num_tokens": 720195271.0, "step": 18965 }, { "epoch": 2.412670143747615, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.91201639175415, "learning_rate": 1e-05, "loss": 0.432, "mean_token_accuracy": 0.8871968984603882, "num_tokens": 720232159.0, "step": 18966 }, { "epoch": 2.412797354026205, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.9689249992370605, "learning_rate": 1e-05, "loss": 0.4717, "mean_token_accuracy": 0.8832982778549194, "num_tokens": 720264591.0, "step": 18967 }, { "epoch": 2.412924564304796, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.767475605010986, "learning_rate": 1e-05, "loss": 0.4534, "mean_token_accuracy": 0.8797073364257812, "num_tokens": 720302957.0, "step": 18968 }, { "epoch": 2.413051774583386, "ewc_loss": 0.1455078125, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 0.00011777877807617188, "grad_norm": 7.403797149658203, "learning_rate": 1e-05, "loss": 0.5246, "mean_token_accuracy": 0.8701943755149841, "num_tokens": 720335679.0, "step": 18969 }, { "epoch": 2.4131789848619767, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 5.834291458129883, "learning_rate": 1e-05, "loss": 0.4877, "mean_token_accuracy": 0.8743121027946472, "num_tokens": 720377261.0, "step": 18970 }, { "epoch": 2.413306195140567, "ewc_loss": 0.15625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 0.00012874603271484375, "grad_norm": 7.805346965789795, "learning_rate": 1e-05, "loss": 0.5017, "mean_token_accuracy": 0.8768228888511658, "num_tokens": 720415344.0, "step": 18971 }, { "epoch": 2.4134334054191577, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.28727388381958, "learning_rate": 1e-05, "loss": 0.4861, "mean_token_accuracy": 0.8741284608840942, "num_tokens": 720449133.0, "step": 18972 }, { "epoch": 2.4135606156977483, "ewc_loss": 0.15625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 0.0001277923583984375, "grad_norm": 7.778183937072754, "learning_rate": 1e-05, "loss": 0.5367, "mean_token_accuracy": 0.8675724267959595, "num_tokens": 720484250.0, "step": 18973 }, { "epoch": 2.413687825976339, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 0.00011205673217773438, "grad_norm": 6.755425930023193, "learning_rate": 1e-05, "loss": 0.5005, "mean_token_accuracy": 0.8719600439071655, "num_tokens": 720520466.0, "step": 18974 }, { "epoch": 2.4138150362549293, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 0.00011587142944335938, "grad_norm": 7.288021087646484, "learning_rate": 1e-05, "loss": 0.489, "mean_token_accuracy": 0.8793073296546936, "num_tokens": 720557013.0, "step": 18975 }, { "epoch": 2.41394224653352, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 6.786047458648682, "learning_rate": 1e-05, "loss": 0.4852, "mean_token_accuracy": 0.8806081414222717, "num_tokens": 720594957.0, "step": 18976 }, { "epoch": 2.4140694568121104, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.7748284339904785, "learning_rate": 1e-05, "loss": 0.4875, "mean_token_accuracy": 0.8751437664031982, "num_tokens": 720630531.0, "step": 18977 }, { "epoch": 2.414196667090701, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.758886814117432, "learning_rate": 1e-05, "loss": 0.5103, "mean_token_accuracy": 0.872414231300354, "num_tokens": 720673502.0, "step": 18978 }, { "epoch": 2.4143238773692914, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.352552890777588, "learning_rate": 1e-05, "loss": 0.4692, "mean_token_accuracy": 0.8775124549865723, "num_tokens": 720716765.0, "step": 18979 }, { "epoch": 2.414451087647882, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.809760093688965, "learning_rate": 1e-05, "loss": 0.4429, "mean_token_accuracy": 0.8894820213317871, "num_tokens": 720754711.0, "step": 18980 }, { "epoch": 2.4145782979264725, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 6.076602458953857, "learning_rate": 1e-05, "loss": 0.4838, "mean_token_accuracy": 0.8713294267654419, "num_tokens": 720789067.0, "step": 18981 }, { "epoch": 2.414705508205063, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.957431316375732, "learning_rate": 1e-05, "loss": 0.5051, "mean_token_accuracy": 0.8694897890090942, "num_tokens": 720821725.0, "step": 18982 }, { "epoch": 2.4148327184836536, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 5.926035404205322, "learning_rate": 1e-05, "loss": 0.4571, "mean_token_accuracy": 0.88080894947052, "num_tokens": 720857587.0, "step": 18983 }, { "epoch": 2.414959928762244, "ewc_loss": 0.140625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 0.00011348724365234375, "grad_norm": 7.187573432922363, "learning_rate": 1e-05, "loss": 0.4861, "mean_token_accuracy": 0.8810519576072693, "num_tokens": 720900481.0, "step": 18984 }, { "epoch": 2.4150871390408346, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 5.915920734405518, "learning_rate": 1e-05, "loss": 0.4711, "mean_token_accuracy": 0.8773254156112671, "num_tokens": 720933131.0, "step": 18985 }, { "epoch": 2.415214349319425, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 0.00011873245239257812, "grad_norm": 7.424198150634766, "learning_rate": 1e-05, "loss": 0.4993, "mean_token_accuracy": 0.8760488033294678, "num_tokens": 720972142.0, "step": 18986 }, { "epoch": 2.4153415595980157, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.043766975402832, "learning_rate": 1e-05, "loss": 0.4401, "mean_token_accuracy": 0.8871454000473022, "num_tokens": 721014251.0, "step": 18987 }, { "epoch": 2.415468769876606, "ewc_loss": 0.150390625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 0.0001220703125, "grad_norm": 7.552551746368408, "learning_rate": 1e-05, "loss": 0.5404, "mean_token_accuracy": 0.8664781451225281, "num_tokens": 721053327.0, "step": 18988 }, { "epoch": 2.4155959801551967, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.302935600280762, "learning_rate": 1e-05, "loss": 0.4701, "mean_token_accuracy": 0.8810456395149231, "num_tokens": 721092047.0, "step": 18989 }, { "epoch": 2.415723190433787, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 0.00011920928955078125, "grad_norm": 7.44620943069458, "learning_rate": 1e-05, "loss": 0.4367, "mean_token_accuracy": 0.8971722722053528, "num_tokens": 721125318.0, "step": 18990 }, { "epoch": 2.4158504007123778, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.487353324890137, "learning_rate": 1e-05, "loss": 0.4477, "mean_token_accuracy": 0.8917076587677002, "num_tokens": 721162058.0, "step": 18991 }, { "epoch": 2.415977610990968, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 0.00011396408081054688, "grad_norm": 7.157955646514893, "learning_rate": 1e-05, "loss": 0.4898, "mean_token_accuracy": 0.8805936574935913, "num_tokens": 721199219.0, "step": 18992 }, { "epoch": 2.4161048212695584, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.51124382019043, "learning_rate": 1e-05, "loss": 0.4888, "mean_token_accuracy": 0.8735007643699646, "num_tokens": 721239696.0, "step": 18993 }, { "epoch": 2.416232031548149, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.863839149475098, "learning_rate": 1e-05, "loss": 0.5061, "mean_token_accuracy": 0.8703648447990417, "num_tokens": 721277660.0, "step": 18994 }, { "epoch": 2.4163592418267394, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.472846984863281, "learning_rate": 1e-05, "loss": 0.4777, "mean_token_accuracy": 0.8782907724380493, "num_tokens": 721311805.0, "step": 18995 }, { "epoch": 2.41648645210533, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.6255693435668945, "learning_rate": 1e-05, "loss": 0.534, "mean_token_accuracy": 0.8581675887107849, "num_tokens": 721355968.0, "step": 18996 }, { "epoch": 2.4166136623839205, "ewc_loss": 0.125, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.386349678039551, "learning_rate": 1e-05, "loss": 0.4903, "mean_token_accuracy": 0.8711177110671997, "num_tokens": 721397930.0, "step": 18997 }, { "epoch": 2.416740872662511, "ewc_loss": 0.125, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.494597434997559, "learning_rate": 1e-05, "loss": 0.4394, "mean_token_accuracy": 0.8883156776428223, "num_tokens": 721432408.0, "step": 18998 }, { "epoch": 2.4168680829411016, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.290013790130615, "learning_rate": 1e-05, "loss": 0.4159, "mean_token_accuracy": 0.8972266912460327, "num_tokens": 721468903.0, "step": 18999 }, { "epoch": 2.416995293219692, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.472742080688477, "learning_rate": 1e-05, "loss": 0.4307, "mean_token_accuracy": 0.8930538892745972, "num_tokens": 721509687.0, "step": 19000 }, { "epoch": 2.4171225034982826, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.176493167877197, "learning_rate": 1e-05, "loss": 0.497, "mean_token_accuracy": 0.8696425557136536, "num_tokens": 721548130.0, "step": 19001 }, { "epoch": 2.417249713776873, "ewc_loss": 0.125, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.529159069061279, "learning_rate": 1e-05, "loss": 0.4734, "mean_token_accuracy": 0.8834896683692932, "num_tokens": 721588504.0, "step": 19002 }, { "epoch": 2.4173769240554637, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 6.031993865966797, "learning_rate": 1e-05, "loss": 0.4795, "mean_token_accuracy": 0.8756939172744751, "num_tokens": 721632621.0, "step": 19003 }, { "epoch": 2.417504134334054, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.672085762023926, "learning_rate": 1e-05, "loss": 0.4878, "mean_token_accuracy": 0.8761544823646545, "num_tokens": 721670898.0, "step": 19004 }, { "epoch": 2.4176313446126447, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.870259761810303, "learning_rate": 1e-05, "loss": 0.4367, "mean_token_accuracy": 0.886158287525177, "num_tokens": 721709012.0, "step": 19005 }, { "epoch": 2.4177585548912353, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.965410232543945, "learning_rate": 1e-05, "loss": 0.4791, "mean_token_accuracy": 0.8779999017715454, "num_tokens": 721747135.0, "step": 19006 }, { "epoch": 2.417885765169826, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.752377510070801, "learning_rate": 1e-05, "loss": 0.4606, "mean_token_accuracy": 0.8773118257522583, "num_tokens": 721784447.0, "step": 19007 }, { "epoch": 2.4180129754484163, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 0.000118255615234375, "grad_norm": 7.396828651428223, "learning_rate": 1e-05, "loss": 0.5224, "mean_token_accuracy": 0.8676395416259766, "num_tokens": 721817853.0, "step": 19008 }, { "epoch": 2.418140185727007, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 5.819647312164307, "learning_rate": 1e-05, "loss": 0.4695, "mean_token_accuracy": 0.873357892036438, "num_tokens": 721852537.0, "step": 19009 }, { "epoch": 2.4182673960055974, "ewc_loss": 0.15625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 0.00012874603271484375, "grad_norm": 7.795103073120117, "learning_rate": 1e-05, "loss": 0.4998, "mean_token_accuracy": 0.879449188709259, "num_tokens": 721892156.0, "step": 19010 }, { "epoch": 2.418394606284188, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.261228084564209, "learning_rate": 1e-05, "loss": 0.478, "mean_token_accuracy": 0.8806285858154297, "num_tokens": 721929425.0, "step": 19011 }, { "epoch": 2.4185218165627784, "ewc_loss": 0.15625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 0.0001277923583984375, "grad_norm": 7.770714282989502, "learning_rate": 1e-05, "loss": 0.5299, "mean_token_accuracy": 0.86866295337677, "num_tokens": 721967492.0, "step": 19012 }, { "epoch": 2.418649026841369, "ewc_loss": 0.140625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 0.00011348724365234375, "grad_norm": 6.700591564178467, "learning_rate": 1e-05, "loss": 0.5055, "mean_token_accuracy": 0.8741247653961182, "num_tokens": 722007754.0, "step": 19013 }, { "epoch": 2.4187762371199595, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 0.0001163482666015625, "grad_norm": 7.312704086303711, "learning_rate": 1e-05, "loss": 0.4952, "mean_token_accuracy": 0.8776434063911438, "num_tokens": 722043380.0, "step": 19014 }, { "epoch": 2.4189034473985496, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 0.00011157989501953125, "grad_norm": 6.778987884521484, "learning_rate": 1e-05, "loss": 0.5049, "mean_token_accuracy": 0.8727368116378784, "num_tokens": 722086200.0, "step": 19015 }, { "epoch": 2.4190306576771405, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.797274112701416, "learning_rate": 1e-05, "loss": 0.4682, "mean_token_accuracy": 0.8809171915054321, "num_tokens": 722122907.0, "step": 19016 }, { "epoch": 2.4191578679557306, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.761425495147705, "learning_rate": 1e-05, "loss": 0.4729, "mean_token_accuracy": 0.8803935050964355, "num_tokens": 722155951.0, "step": 19017 }, { "epoch": 2.419285078234321, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.371945858001709, "learning_rate": 1e-05, "loss": 0.4791, "mean_token_accuracy": 0.8729486465454102, "num_tokens": 722194479.0, "step": 19018 }, { "epoch": 2.4194122885129117, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.807143211364746, "learning_rate": 1e-05, "loss": 0.4754, "mean_token_accuracy": 0.8801885843276978, "num_tokens": 722228470.0, "step": 19019 }, { "epoch": 2.419539498791502, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 6.104189395904541, "learning_rate": 1e-05, "loss": 0.4988, "mean_token_accuracy": 0.8687822818756104, "num_tokens": 722265960.0, "step": 19020 }, { "epoch": 2.4196667090700927, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.936098098754883, "learning_rate": 1e-05, "loss": 0.5065, "mean_token_accuracy": 0.8731420040130615, "num_tokens": 722301377.0, "step": 19021 }, { "epoch": 2.4197939193486833, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 5.943939208984375, "learning_rate": 1e-05, "loss": 0.448, "mean_token_accuracy": 0.8808767795562744, "num_tokens": 722338534.0, "step": 19022 }, { "epoch": 2.419921129627274, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 0.00011157989501953125, "grad_norm": 7.12971830368042, "learning_rate": 1e-05, "loss": 0.496, "mean_token_accuracy": 0.8748655319213867, "num_tokens": 722377460.0, "step": 19023 }, { "epoch": 2.4200483399058643, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 5.922756671905518, "learning_rate": 1e-05, "loss": 0.5043, "mean_token_accuracy": 0.8669410347938538, "num_tokens": 722415167.0, "step": 19024 }, { "epoch": 2.420175550184455, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 0.00011730194091796875, "grad_norm": 7.352923393249512, "learning_rate": 1e-05, "loss": 0.538, "mean_token_accuracy": 0.8644858598709106, "num_tokens": 722456114.0, "step": 19025 }, { "epoch": 2.4203027604630454, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.0290093421936035, "learning_rate": 1e-05, "loss": 0.4572, "mean_token_accuracy": 0.8832922577857971, "num_tokens": 722496378.0, "step": 19026 }, { "epoch": 2.420429970741636, "ewc_loss": 0.1484375, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 0.0001201629638671875, "grad_norm": 7.466709613800049, "learning_rate": 1e-05, "loss": 0.5261, "mean_token_accuracy": 0.8688374161720276, "num_tokens": 722534903.0, "step": 19027 }, { "epoch": 2.4205571810202264, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.233034133911133, "learning_rate": 1e-05, "loss": 0.4936, "mean_token_accuracy": 0.8727145791053772, "num_tokens": 722576649.0, "step": 19028 }, { "epoch": 2.420684391298817, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 0.00011920928955078125, "grad_norm": 7.400055885314941, "learning_rate": 1e-05, "loss": 0.5191, "mean_token_accuracy": 0.8680026531219482, "num_tokens": 722618555.0, "step": 19029 }, { "epoch": 2.4208116015774075, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.40598726272583, "learning_rate": 1e-05, "loss": 0.4253, "mean_token_accuracy": 0.8949444890022278, "num_tokens": 722652045.0, "step": 19030 }, { "epoch": 2.420938811855998, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 0.00011396408081054688, "grad_norm": 7.1573967933654785, "learning_rate": 1e-05, "loss": 0.4986, "mean_token_accuracy": 0.875291645526886, "num_tokens": 722686987.0, "step": 19031 }, { "epoch": 2.4210660221345885, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.463823318481445, "learning_rate": 1e-05, "loss": 0.4964, "mean_token_accuracy": 0.8708939552307129, "num_tokens": 722724161.0, "step": 19032 }, { "epoch": 2.421193232413179, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.8661603927612305, "learning_rate": 1e-05, "loss": 0.471, "mean_token_accuracy": 0.8842355012893677, "num_tokens": 722766473.0, "step": 19033 }, { "epoch": 2.4213204426917696, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.42893123626709, "learning_rate": 1e-05, "loss": 0.4714, "mean_token_accuracy": 0.8782850503921509, "num_tokens": 722809499.0, "step": 19034 }, { "epoch": 2.42144765297036, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.631324768066406, "learning_rate": 1e-05, "loss": 0.4606, "mean_token_accuracy": 0.8821537494659424, "num_tokens": 722847295.0, "step": 19035 }, { "epoch": 2.4215748632489507, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.384031295776367, "learning_rate": 1e-05, "loss": 0.472, "mean_token_accuracy": 0.878915548324585, "num_tokens": 722884313.0, "step": 19036 }, { "epoch": 2.421702073527541, "ewc_loss": 0.125, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.4867448806762695, "learning_rate": 1e-05, "loss": 0.4682, "mean_token_accuracy": 0.8818731307983398, "num_tokens": 722924530.0, "step": 19037 }, { "epoch": 2.4218292838061317, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.305578708648682, "learning_rate": 1e-05, "loss": 0.5027, "mean_token_accuracy": 0.8666319847106934, "num_tokens": 722962968.0, "step": 19038 }, { "epoch": 2.4219564940847222, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.439483165740967, "learning_rate": 1e-05, "loss": 0.4449, "mean_token_accuracy": 0.8844298720359802, "num_tokens": 722997801.0, "step": 19039 }, { "epoch": 2.4220837043633123, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.2071614265441895, "learning_rate": 1e-05, "loss": 0.502, "mean_token_accuracy": 0.8655136227607727, "num_tokens": 723037724.0, "step": 19040 }, { "epoch": 2.4222109146419033, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.450632095336914, "learning_rate": 1e-05, "loss": 0.5185, "mean_token_accuracy": 0.8686934113502502, "num_tokens": 723078199.0, "step": 19041 }, { "epoch": 2.4223381249204934, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.10204553604126, "learning_rate": 1e-05, "loss": 0.4958, "mean_token_accuracy": 0.8700820207595825, "num_tokens": 723110894.0, "step": 19042 }, { "epoch": 2.422465335199084, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.539371967315674, "learning_rate": 1e-05, "loss": 0.5088, "mean_token_accuracy": 0.8677213191986084, "num_tokens": 723150688.0, "step": 19043 }, { "epoch": 2.4225925454776744, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 5.965085983276367, "learning_rate": 1e-05, "loss": 0.4705, "mean_token_accuracy": 0.8753980994224548, "num_tokens": 723189863.0, "step": 19044 }, { "epoch": 2.422719755756265, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.737014293670654, "learning_rate": 1e-05, "loss": 0.5065, "mean_token_accuracy": 0.8665802478790283, "num_tokens": 723232333.0, "step": 19045 }, { "epoch": 2.4228469660348555, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.806363105773926, "learning_rate": 1e-05, "loss": 0.5242, "mean_token_accuracy": 0.8635953664779663, "num_tokens": 723267469.0, "step": 19046 }, { "epoch": 2.422974176313446, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 0.00010967254638671875, "grad_norm": 7.070368766784668, "learning_rate": 1e-05, "loss": 0.4489, "mean_token_accuracy": 0.888759970664978, "num_tokens": 723304367.0, "step": 19047 }, { "epoch": 2.4231013865920366, "ewc_loss": 0.11181640625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.699248313903809, "learning_rate": 1e-05, "loss": 0.4651, "mean_token_accuracy": 0.873235285282135, "num_tokens": 723343247.0, "step": 19048 }, { "epoch": 2.423228596870627, "ewc_loss": 0.1494140625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 0.00012159347534179688, "grad_norm": 7.547122478485107, "learning_rate": 1e-05, "loss": 0.469, "mean_token_accuracy": 0.8879773616790771, "num_tokens": 723378022.0, "step": 19049 }, { "epoch": 2.4233558071492176, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 5.885634422302246, "learning_rate": 1e-05, "loss": 0.4649, "mean_token_accuracy": 0.8797447681427002, "num_tokens": 723409827.0, "step": 19050 }, { "epoch": 2.423483017427808, "ewc_loss": 0.16015625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 0.0001316070556640625, "grad_norm": 7.909718990325928, "learning_rate": 1e-05, "loss": 0.4835, "mean_token_accuracy": 0.8856772184371948, "num_tokens": 723447192.0, "step": 19051 }, { "epoch": 2.4236102277063987, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.411802291870117, "learning_rate": 1e-05, "loss": 0.4734, "mean_token_accuracy": 0.8808190822601318, "num_tokens": 723492123.0, "step": 19052 }, { "epoch": 2.423737437984989, "ewc_loss": 0.15625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 0.0001277923583984375, "grad_norm": 7.759962558746338, "learning_rate": 1e-05, "loss": 0.4581, "mean_token_accuracy": 0.8931683897972107, "num_tokens": 723525680.0, "step": 19053 }, { "epoch": 2.4238646482635797, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 0.00011587142944335938, "grad_norm": 6.785494804382324, "learning_rate": 1e-05, "loss": 0.5057, "mean_token_accuracy": 0.8754032850265503, "num_tokens": 723565667.0, "step": 19054 }, { "epoch": 2.4239918585421703, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 0.00011730194091796875, "grad_norm": 7.287125110626221, "learning_rate": 1e-05, "loss": 0.5126, "mean_token_accuracy": 0.8716592788696289, "num_tokens": 723599486.0, "step": 19055 }, { "epoch": 2.424119068820761, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 0.00011110305786132812, "grad_norm": 6.77647590637207, "learning_rate": 1e-05, "loss": 0.4681, "mean_token_accuracy": 0.8834509253501892, "num_tokens": 723631992.0, "step": 19056 }, { "epoch": 2.4242462790993513, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.82836389541626, "learning_rate": 1e-05, "loss": 0.4448, "mean_token_accuracy": 0.8859912157058716, "num_tokens": 723667654.0, "step": 19057 }, { "epoch": 2.424373489377942, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.687493801116943, "learning_rate": 1e-05, "loss": 0.4529, "mean_token_accuracy": 0.8835210800170898, "num_tokens": 723705368.0, "step": 19058 }, { "epoch": 2.4245006996565324, "ewc_loss": 0.125, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.486324787139893, "learning_rate": 1e-05, "loss": 0.4533, "mean_token_accuracy": 0.8792468905448914, "num_tokens": 723744973.0, "step": 19059 }, { "epoch": 2.424627909935123, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.655842304229736, "learning_rate": 1e-05, "loss": 0.515, "mean_token_accuracy": 0.8671855926513672, "num_tokens": 723778868.0, "step": 19060 }, { "epoch": 2.4247551202137134, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.249790191650391, "learning_rate": 1e-05, "loss": 0.4692, "mean_token_accuracy": 0.8807199001312256, "num_tokens": 723822413.0, "step": 19061 }, { "epoch": 2.424882330492304, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.653782844543457, "learning_rate": 1e-05, "loss": 0.4979, "mean_token_accuracy": 0.8700929880142212, "num_tokens": 723865826.0, "step": 19062 }, { "epoch": 2.4250095407708945, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.106622219085693, "learning_rate": 1e-05, "loss": 0.4508, "mean_token_accuracy": 0.8822786808013916, "num_tokens": 723901839.0, "step": 19063 }, { "epoch": 2.425136751049485, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.694302082061768, "learning_rate": 1e-05, "loss": 0.4686, "mean_token_accuracy": 0.8841283321380615, "num_tokens": 723944883.0, "step": 19064 }, { "epoch": 2.425263961328075, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 5.974472522735596, "learning_rate": 1e-05, "loss": 0.4452, "mean_token_accuracy": 0.8843498229980469, "num_tokens": 723985084.0, "step": 19065 }, { "epoch": 2.425391171606666, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.831653594970703, "learning_rate": 1e-05, "loss": 0.4938, "mean_token_accuracy": 0.8758764863014221, "num_tokens": 724027919.0, "step": 19066 }, { "epoch": 2.425518381885256, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.866342067718506, "learning_rate": 1e-05, "loss": 0.4533, "mean_token_accuracy": 0.8825570344924927, "num_tokens": 724064593.0, "step": 19067 }, { "epoch": 2.4256455921638467, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 0.00011014938354492188, "grad_norm": 7.08007287979126, "learning_rate": 1e-05, "loss": 0.4579, "mean_token_accuracy": 0.887722373008728, "num_tokens": 724095967.0, "step": 19068 }, { "epoch": 2.425772802442437, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.797554969787598, "learning_rate": 1e-05, "loss": 0.4241, "mean_token_accuracy": 0.8905010223388672, "num_tokens": 724128420.0, "step": 19069 }, { "epoch": 2.4259000127210277, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 0.00011873245239257812, "grad_norm": 7.424499034881592, "learning_rate": 1e-05, "loss": 0.5197, "mean_token_accuracy": 0.8688908815383911, "num_tokens": 724165073.0, "step": 19070 }, { "epoch": 2.4260272229996183, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 5.898819923400879, "learning_rate": 1e-05, "loss": 0.4808, "mean_token_accuracy": 0.8768824338912964, "num_tokens": 724201377.0, "step": 19071 }, { "epoch": 2.426154433278209, "ewc_loss": 0.154296875, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 0.00012683868408203125, "grad_norm": 7.701984882354736, "learning_rate": 1e-05, "loss": 0.4615, "mean_token_accuracy": 0.8936018347740173, "num_tokens": 724241783.0, "step": 19072 }, { "epoch": 2.4262816435567993, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.261313438415527, "learning_rate": 1e-05, "loss": 0.4725, "mean_token_accuracy": 0.8767902851104736, "num_tokens": 724277903.0, "step": 19073 }, { "epoch": 2.42640885383539, "ewc_loss": 0.15234375, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 0.00012493133544921875, "grad_norm": 7.648146152496338, "learning_rate": 1e-05, "loss": 0.4723, "mean_token_accuracy": 0.8850818872451782, "num_tokens": 724318834.0, "step": 19074 }, { "epoch": 2.4265360641139804, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 6.583596229553223, "learning_rate": 1e-05, "loss": 0.4754, "mean_token_accuracy": 0.882072925567627, "num_tokens": 724356406.0, "step": 19075 }, { "epoch": 2.426663274392571, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 0.00011539459228515625, "grad_norm": 7.279404640197754, "learning_rate": 1e-05, "loss": 0.462, "mean_token_accuracy": 0.8876767158508301, "num_tokens": 724394774.0, "step": 19076 }, { "epoch": 2.4267904846711614, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 6.680146217346191, "learning_rate": 1e-05, "loss": 0.5089, "mean_token_accuracy": 0.8712226152420044, "num_tokens": 724431825.0, "step": 19077 }, { "epoch": 2.426917694949752, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.858004093170166, "learning_rate": 1e-05, "loss": 0.485, "mean_token_accuracy": 0.8774295449256897, "num_tokens": 724466402.0, "step": 19078 }, { "epoch": 2.4270449052283425, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.630977630615234, "learning_rate": 1e-05, "loss": 0.5084, "mean_token_accuracy": 0.8695461750030518, "num_tokens": 724503353.0, "step": 19079 }, { "epoch": 2.427172115506933, "ewc_loss": 0.125, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.493968963623047, "learning_rate": 1e-05, "loss": 0.4534, "mean_token_accuracy": 0.8845100402832031, "num_tokens": 724540348.0, "step": 19080 }, { "epoch": 2.4272993257855235, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.605423927307129, "learning_rate": 1e-05, "loss": 0.4955, "mean_token_accuracy": 0.8745923638343811, "num_tokens": 724577999.0, "step": 19081 }, { "epoch": 2.427426536064114, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.244178771972656, "learning_rate": 1e-05, "loss": 0.4154, "mean_token_accuracy": 0.8921334147453308, "num_tokens": 724614754.0, "step": 19082 }, { "epoch": 2.4275537463427046, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.617005825042725, "learning_rate": 1e-05, "loss": 0.5045, "mean_token_accuracy": 0.8690433502197266, "num_tokens": 724653232.0, "step": 19083 }, { "epoch": 2.427680956621295, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.089869022369385, "learning_rate": 1e-05, "loss": 0.5029, "mean_token_accuracy": 0.8688725829124451, "num_tokens": 724686461.0, "step": 19084 }, { "epoch": 2.4278081668998857, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.6734113693237305, "learning_rate": 1e-05, "loss": 0.4497, "mean_token_accuracy": 0.8883714079856873, "num_tokens": 724727330.0, "step": 19085 }, { "epoch": 2.427935377178476, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 5.957180976867676, "learning_rate": 1e-05, "loss": 0.5008, "mean_token_accuracy": 0.8681067228317261, "num_tokens": 724766509.0, "step": 19086 }, { "epoch": 2.4280625874570667, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.821989059448242, "learning_rate": 1e-05, "loss": 0.436, "mean_token_accuracy": 0.8923002481460571, "num_tokens": 724795128.0, "step": 19087 }, { "epoch": 2.428189797735657, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.8426127433776855, "learning_rate": 1e-05, "loss": 0.4816, "mean_token_accuracy": 0.8720361590385437, "num_tokens": 724833941.0, "step": 19088 }, { "epoch": 2.4283170080142478, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 0.00011014938354492188, "grad_norm": 7.080956935882568, "learning_rate": 1e-05, "loss": 0.4751, "mean_token_accuracy": 0.8827232122421265, "num_tokens": 724870429.0, "step": 19089 }, { "epoch": 2.428444218292838, "ewc_loss": 0.11474609375, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.770382404327393, "learning_rate": 1e-05, "loss": 0.4497, "mean_token_accuracy": 0.8810399174690247, "num_tokens": 724905574.0, "step": 19090 }, { "epoch": 2.4285714285714284, "ewc_loss": 0.1484375, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 0.00012063980102539062, "grad_norm": 7.438397407531738, "learning_rate": 1e-05, "loss": 0.4792, "mean_token_accuracy": 0.8847079277038574, "num_tokens": 724949845.0, "step": 19091 }, { "epoch": 2.428698638850019, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 5.887325763702393, "learning_rate": 1e-05, "loss": 0.5027, "mean_token_accuracy": 0.8708665370941162, "num_tokens": 724987586.0, "step": 19092 }, { "epoch": 2.4288258491286094, "ewc_loss": 0.15625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 0.0001277923583984375, "grad_norm": 7.751804351806641, "learning_rate": 1e-05, "loss": 0.4766, "mean_token_accuracy": 0.8840301036834717, "num_tokens": 725026432.0, "step": 19093 }, { "epoch": 2.4289530594072, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.268514633178711, "learning_rate": 1e-05, "loss": 0.5475, "mean_token_accuracy": 0.8579767942428589, "num_tokens": 725070627.0, "step": 19094 }, { "epoch": 2.4290802696857905, "ewc_loss": 0.154296875, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 0.00012683868408203125, "grad_norm": 7.71429443359375, "learning_rate": 1e-05, "loss": 0.5104, "mean_token_accuracy": 0.8775716423988342, "num_tokens": 725106005.0, "step": 19095 }, { "epoch": 2.429207479964381, "ewc_loss": 0.140625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 0.0001125335693359375, "grad_norm": 6.628096580505371, "learning_rate": 1e-05, "loss": 0.4704, "mean_token_accuracy": 0.882116436958313, "num_tokens": 725149667.0, "step": 19096 }, { "epoch": 2.4293346902429716, "ewc_loss": 0.1455078125, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 0.00011777877807617188, "grad_norm": 7.335876941680908, "learning_rate": 1e-05, "loss": 0.5107, "mean_token_accuracy": 0.8810974955558777, "num_tokens": 725185526.0, "step": 19097 }, { "epoch": 2.429461900521562, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 0.00011205673217773438, "grad_norm": 6.713700294494629, "learning_rate": 1e-05, "loss": 0.4791, "mean_token_accuracy": 0.8819968700408936, "num_tokens": 725221789.0, "step": 19098 }, { "epoch": 2.4295891108001526, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.883557319641113, "learning_rate": 1e-05, "loss": 0.5126, "mean_token_accuracy": 0.8706971406936646, "num_tokens": 725258275.0, "step": 19099 }, { "epoch": 2.429716321078743, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.673010349273682, "learning_rate": 1e-05, "loss": 0.523, "mean_token_accuracy": 0.864875853061676, "num_tokens": 725297753.0, "step": 19100 }, { "epoch": 2.4298435313573337, "ewc_loss": 0.125, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.502274036407471, "learning_rate": 1e-05, "loss": 0.4473, "mean_token_accuracy": 0.8851741552352905, "num_tokens": 725331978.0, "step": 19101 }, { "epoch": 2.429970741635924, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.638222694396973, "learning_rate": 1e-05, "loss": 0.4659, "mean_token_accuracy": 0.8823370933532715, "num_tokens": 725373876.0, "step": 19102 }, { "epoch": 2.4300979519145147, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.255228519439697, "learning_rate": 1e-05, "loss": 0.4384, "mean_token_accuracy": 0.8848898410797119, "num_tokens": 725411280.0, "step": 19103 }, { "epoch": 2.4302251621931052, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.647614479064941, "learning_rate": 1e-05, "loss": 0.5251, "mean_token_accuracy": 0.8640996217727661, "num_tokens": 725452050.0, "step": 19104 }, { "epoch": 2.4303523724716958, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.08741569519043, "learning_rate": 1e-05, "loss": 0.4531, "mean_token_accuracy": 0.8851152062416077, "num_tokens": 725489922.0, "step": 19105 }, { "epoch": 2.4304795827502863, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.711653232574463, "learning_rate": 1e-05, "loss": 0.4908, "mean_token_accuracy": 0.8732531666755676, "num_tokens": 725527871.0, "step": 19106 }, { "epoch": 2.430606793028877, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 5.952160358428955, "learning_rate": 1e-05, "loss": 0.4479, "mean_token_accuracy": 0.888835608959198, "num_tokens": 725565045.0, "step": 19107 }, { "epoch": 2.4307340033074674, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.8603644371032715, "learning_rate": 1e-05, "loss": 0.4529, "mean_token_accuracy": 0.8876160979270935, "num_tokens": 725596604.0, "step": 19108 }, { "epoch": 2.430861213586058, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.832555294036865, "learning_rate": 1e-05, "loss": 0.4531, "mean_token_accuracy": 0.8764650225639343, "num_tokens": 725632299.0, "step": 19109 }, { "epoch": 2.4309884238646484, "ewc_loss": 0.140625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 0.0001125335693359375, "grad_norm": 7.135745525360107, "learning_rate": 1e-05, "loss": 0.495, "mean_token_accuracy": 0.8755457401275635, "num_tokens": 725673778.0, "step": 19110 }, { "epoch": 2.431115634143239, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 5.780043601989746, "learning_rate": 1e-05, "loss": 0.5092, "mean_token_accuracy": 0.8661019802093506, "num_tokens": 725715087.0, "step": 19111 }, { "epoch": 2.4312428444218295, "ewc_loss": 0.1494140625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 0.00012159347534179688, "grad_norm": 7.506859302520752, "learning_rate": 1e-05, "loss": 0.4942, "mean_token_accuracy": 0.8801548480987549, "num_tokens": 725754047.0, "step": 19112 }, { "epoch": 2.4313700547004196, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 5.94316291809082, "learning_rate": 1e-05, "loss": 0.5052, "mean_token_accuracy": 0.8665024042129517, "num_tokens": 725793168.0, "step": 19113 }, { "epoch": 2.4314972649790105, "ewc_loss": 0.15625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 0.00012874603271484375, "grad_norm": 7.759061813354492, "learning_rate": 1e-05, "loss": 0.5057, "mean_token_accuracy": 0.8813116550445557, "num_tokens": 725834713.0, "step": 19114 }, { "epoch": 2.4316244752576006, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.3445820808410645, "learning_rate": 1e-05, "loss": 0.4776, "mean_token_accuracy": 0.8788155317306519, "num_tokens": 725872759.0, "step": 19115 }, { "epoch": 2.431751685536191, "ewc_loss": 0.15234375, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 0.0001239776611328125, "grad_norm": 7.658596992492676, "learning_rate": 1e-05, "loss": 0.4926, "mean_token_accuracy": 0.8808850646018982, "num_tokens": 725916192.0, "step": 19116 }, { "epoch": 2.4318788958147817, "ewc_loss": 0.140625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 0.0001125335693359375, "grad_norm": 6.654252529144287, "learning_rate": 1e-05, "loss": 0.433, "mean_token_accuracy": 0.893976092338562, "num_tokens": 725949838.0, "step": 19117 }, { "epoch": 2.432006106093372, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 0.0001163482666015625, "grad_norm": 7.268603801727295, "learning_rate": 1e-05, "loss": 0.5103, "mean_token_accuracy": 0.8728641271591187, "num_tokens": 725986538.0, "step": 19118 }, { "epoch": 2.4321333163719627, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 0.00011157989501953125, "grad_norm": 6.6841301918029785, "learning_rate": 1e-05, "loss": 0.4677, "mean_token_accuracy": 0.8822273015975952, "num_tokens": 726029007.0, "step": 19119 }, { "epoch": 2.4322605266505533, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.838924884796143, "learning_rate": 1e-05, "loss": 0.4782, "mean_token_accuracy": 0.8792651891708374, "num_tokens": 726066745.0, "step": 19120 }, { "epoch": 2.432387736929144, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.641615390777588, "learning_rate": 1e-05, "loss": 0.5041, "mean_token_accuracy": 0.8723266124725342, "num_tokens": 726106768.0, "step": 19121 }, { "epoch": 2.4325149472077343, "ewc_loss": 0.125, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.481357097625732, "learning_rate": 1e-05, "loss": 0.471, "mean_token_accuracy": 0.8780220746994019, "num_tokens": 726144879.0, "step": 19122 }, { "epoch": 2.432642157486325, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.611562728881836, "learning_rate": 1e-05, "loss": 0.4742, "mean_token_accuracy": 0.8780719041824341, "num_tokens": 726181278.0, "step": 19123 }, { "epoch": 2.4327693677649154, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.245595455169678, "learning_rate": 1e-05, "loss": 0.4778, "mean_token_accuracy": 0.8803361654281616, "num_tokens": 726226090.0, "step": 19124 }, { "epoch": 2.432896578043506, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.61268949508667, "learning_rate": 1e-05, "loss": 0.4896, "mean_token_accuracy": 0.8749507665634155, "num_tokens": 726261567.0, "step": 19125 }, { "epoch": 2.4330237883220964, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.09429407119751, "learning_rate": 1e-05, "loss": 0.4669, "mean_token_accuracy": 0.8778822422027588, "num_tokens": 726300527.0, "step": 19126 }, { "epoch": 2.433150998600687, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.660571098327637, "learning_rate": 1e-05, "loss": 0.4576, "mean_token_accuracy": 0.8838489055633545, "num_tokens": 726341240.0, "step": 19127 }, { "epoch": 2.4332782088792775, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 5.968818187713623, "learning_rate": 1e-05, "loss": 0.4888, "mean_token_accuracy": 0.8692617416381836, "num_tokens": 726375286.0, "step": 19128 }, { "epoch": 2.433405419157868, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.78852653503418, "learning_rate": 1e-05, "loss": 0.441, "mean_token_accuracy": 0.8920316696166992, "num_tokens": 726414810.0, "step": 19129 }, { "epoch": 2.4335326294364585, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.8403191566467285, "learning_rate": 1e-05, "loss": 0.4151, "mean_token_accuracy": 0.8942630290985107, "num_tokens": 726451204.0, "step": 19130 }, { "epoch": 2.433659839715049, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 0.00011014938354492188, "grad_norm": 7.051459312438965, "learning_rate": 1e-05, "loss": 0.485, "mean_token_accuracy": 0.8771381378173828, "num_tokens": 726491107.0, "step": 19131 }, { "epoch": 2.4337870499936396, "ewc_loss": 0.11376953125, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.750001430511475, "learning_rate": 1e-05, "loss": 0.4759, "mean_token_accuracy": 0.8761461973190308, "num_tokens": 726535297.0, "step": 19132 }, { "epoch": 2.43391426027223, "ewc_loss": 0.1484375, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 0.00012063980102539062, "grad_norm": 7.450375080108643, "learning_rate": 1e-05, "loss": 0.5205, "mean_token_accuracy": 0.8689744472503662, "num_tokens": 726574841.0, "step": 19133 }, { "epoch": 2.4340414705508207, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 5.863460063934326, "learning_rate": 1e-05, "loss": 0.4838, "mean_token_accuracy": 0.8720860481262207, "num_tokens": 726611101.0, "step": 19134 }, { "epoch": 2.434168680829411, "ewc_loss": 0.158203125, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 0.00012969970703125, "grad_norm": 7.803310394287109, "learning_rate": 1e-05, "loss": 0.5071, "mean_token_accuracy": 0.872816264629364, "num_tokens": 726654323.0, "step": 19135 }, { "epoch": 2.4342958911080017, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.283816814422607, "learning_rate": 1e-05, "loss": 0.4817, "mean_token_accuracy": 0.8800082802772522, "num_tokens": 726691504.0, "step": 19136 }, { "epoch": 2.4344231013865922, "ewc_loss": 0.15625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 0.0001277923583984375, "grad_norm": 7.775946617126465, "learning_rate": 1e-05, "loss": 0.526, "mean_token_accuracy": 0.8720636367797852, "num_tokens": 726737257.0, "step": 19137 }, { "epoch": 2.4345503116651823, "ewc_loss": 0.140625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 0.00011348724365234375, "grad_norm": 6.67397403717041, "learning_rate": 1e-05, "loss": 0.4612, "mean_token_accuracy": 0.887475848197937, "num_tokens": 726779917.0, "step": 19138 }, { "epoch": 2.4346775219437733, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 0.00011682510375976562, "grad_norm": 7.372040748596191, "learning_rate": 1e-05, "loss": 0.4819, "mean_token_accuracy": 0.879889965057373, "num_tokens": 726823130.0, "step": 19139 }, { "epoch": 2.4348047322223634, "ewc_loss": 0.140625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 0.0001125335693359375, "grad_norm": 6.747743129730225, "learning_rate": 1e-05, "loss": 0.4809, "mean_token_accuracy": 0.8809664249420166, "num_tokens": 726861787.0, "step": 19140 }, { "epoch": 2.434931942500954, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.918824195861816, "learning_rate": 1e-05, "loss": 0.5319, "mean_token_accuracy": 0.8613241910934448, "num_tokens": 726904413.0, "step": 19141 }, { "epoch": 2.4350591527795444, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 6.689527988433838, "learning_rate": 1e-05, "loss": 0.4581, "mean_token_accuracy": 0.8898482322692871, "num_tokens": 726946984.0, "step": 19142 }, { "epoch": 2.435186363058135, "ewc_loss": 0.125, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.544712543487549, "learning_rate": 1e-05, "loss": 0.4885, "mean_token_accuracy": 0.8770971298217773, "num_tokens": 726982587.0, "step": 19143 }, { "epoch": 2.4353135733367255, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.651610374450684, "learning_rate": 1e-05, "loss": 0.4574, "mean_token_accuracy": 0.8814947605133057, "num_tokens": 727017255.0, "step": 19144 }, { "epoch": 2.435440783615316, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.282320976257324, "learning_rate": 1e-05, "loss": 0.4438, "mean_token_accuracy": 0.8848317861557007, "num_tokens": 727046147.0, "step": 19145 }, { "epoch": 2.4355679938939065, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.63629674911499, "learning_rate": 1e-05, "loss": 0.474, "mean_token_accuracy": 0.8771774768829346, "num_tokens": 727085982.0, "step": 19146 }, { "epoch": 2.435695204172497, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.114189624786377, "learning_rate": 1e-05, "loss": 0.4319, "mean_token_accuracy": 0.8907607197761536, "num_tokens": 727123991.0, "step": 19147 }, { "epoch": 2.4358224144510876, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.673520565032959, "learning_rate": 1e-05, "loss": 0.5284, "mean_token_accuracy": 0.8624929189682007, "num_tokens": 727162155.0, "step": 19148 }, { "epoch": 2.435949624729678, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 5.984111785888672, "learning_rate": 1e-05, "loss": 0.4937, "mean_token_accuracy": 0.8683756589889526, "num_tokens": 727201234.0, "step": 19149 }, { "epoch": 2.4360768350082687, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.805209636688232, "learning_rate": 1e-05, "loss": 0.4903, "mean_token_accuracy": 0.8763168454170227, "num_tokens": 727243429.0, "step": 19150 }, { "epoch": 2.436204045286859, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.858474254608154, "learning_rate": 1e-05, "loss": 0.4704, "mean_token_accuracy": 0.8762986063957214, "num_tokens": 727279705.0, "step": 19151 }, { "epoch": 2.4363312555654497, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 7.059505462646484, "learning_rate": 1e-05, "loss": 0.4667, "mean_token_accuracy": 0.883859395980835, "num_tokens": 727315503.0, "step": 19152 }, { "epoch": 2.4364584658440402, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.755150318145752, "learning_rate": 1e-05, "loss": 0.4749, "mean_token_accuracy": 0.8751773238182068, "num_tokens": 727355272.0, "step": 19153 }, { "epoch": 2.4365856761226308, "ewc_loss": 0.1484375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 0.00012063980102539062, "grad_norm": 7.463431358337402, "learning_rate": 1e-05, "loss": 0.5353, "mean_token_accuracy": 0.8655892610549927, "num_tokens": 727391058.0, "step": 19154 }, { "epoch": 2.4367128864012213, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 5.868384838104248, "learning_rate": 1e-05, "loss": 0.4835, "mean_token_accuracy": 0.8707563877105713, "num_tokens": 727433011.0, "step": 19155 }, { "epoch": 2.436840096679812, "ewc_loss": 0.1572265625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 0.00012874603271484375, "grad_norm": 7.818215370178223, "learning_rate": 1e-05, "loss": 0.4864, "mean_token_accuracy": 0.884007453918457, "num_tokens": 727462282.0, "step": 19156 }, { "epoch": 2.4369673069584024, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.304876804351807, "learning_rate": 1e-05, "loss": 0.5024, "mean_token_accuracy": 0.8754254579544067, "num_tokens": 727501636.0, "step": 19157 }, { "epoch": 2.437094517236993, "ewc_loss": 0.15625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 0.0001277923583984375, "grad_norm": 7.758749008178711, "learning_rate": 1e-05, "loss": 0.4444, "mean_token_accuracy": 0.8954598903656006, "num_tokens": 727536828.0, "step": 19158 }, { "epoch": 2.4372217275155834, "ewc_loss": 0.140625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 0.00011301040649414062, "grad_norm": 6.681914806365967, "learning_rate": 1e-05, "loss": 0.5647, "mean_token_accuracy": 0.850477933883667, "num_tokens": 727574858.0, "step": 19159 }, { "epoch": 2.437348937794174, "ewc_loss": 0.1455078125, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 0.00011730194091796875, "grad_norm": 7.360165596008301, "learning_rate": 1e-05, "loss": 0.4508, "mean_token_accuracy": 0.8928426504135132, "num_tokens": 727618168.0, "step": 19160 }, { "epoch": 2.4374761480727645, "ewc_loss": 0.140625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 0.0001125335693359375, "grad_norm": 6.740999221801758, "learning_rate": 1e-05, "loss": 0.5308, "mean_token_accuracy": 0.8667514324188232, "num_tokens": 727660205.0, "step": 19161 }, { "epoch": 2.437603358351355, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.918703079223633, "learning_rate": 1e-05, "loss": 0.465, "mean_token_accuracy": 0.8834057450294495, "num_tokens": 727701980.0, "step": 19162 }, { "epoch": 2.437730568629945, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 6.668478012084961, "learning_rate": 1e-05, "loss": 0.5222, "mean_token_accuracy": 0.865381121635437, "num_tokens": 727743766.0, "step": 19163 }, { "epoch": 2.437857778908536, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.5590434074401855, "learning_rate": 1e-05, "loss": 0.4701, "mean_token_accuracy": 0.8793468475341797, "num_tokens": 727785504.0, "step": 19164 }, { "epoch": 2.437984989187126, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.604358196258545, "learning_rate": 1e-05, "loss": 0.4615, "mean_token_accuracy": 0.8837222456932068, "num_tokens": 727831296.0, "step": 19165 }, { "epoch": 2.4381121994657167, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.323784351348877, "learning_rate": 1e-05, "loss": 0.4419, "mean_token_accuracy": 0.8835651874542236, "num_tokens": 727867343.0, "step": 19166 }, { "epoch": 2.438239409744307, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.576140880584717, "learning_rate": 1e-05, "loss": 0.4738, "mean_token_accuracy": 0.8769688010215759, "num_tokens": 727903856.0, "step": 19167 }, { "epoch": 2.4383666200228977, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.173457622528076, "learning_rate": 1e-05, "loss": 0.4818, "mean_token_accuracy": 0.8752634525299072, "num_tokens": 727938385.0, "step": 19168 }, { "epoch": 2.4384938303014883, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.5995707511901855, "learning_rate": 1e-05, "loss": 0.4443, "mean_token_accuracy": 0.8868973255157471, "num_tokens": 727965566.0, "step": 19169 }, { "epoch": 2.438621040580079, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 6.053109169006348, "learning_rate": 1e-05, "loss": 0.4346, "mean_token_accuracy": 0.8917409777641296, "num_tokens": 728002203.0, "step": 19170 }, { "epoch": 2.4387482508586693, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.657647132873535, "learning_rate": 1e-05, "loss": 0.4561, "mean_token_accuracy": 0.883945107460022, "num_tokens": 728037055.0, "step": 19171 }, { "epoch": 2.43887546113726, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 5.935849666595459, "learning_rate": 1e-05, "loss": 0.4942, "mean_token_accuracy": 0.869152843952179, "num_tokens": 728077226.0, "step": 19172 }, { "epoch": 2.4390026714158504, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.826093673706055, "learning_rate": 1e-05, "loss": 0.4679, "mean_token_accuracy": 0.8834075927734375, "num_tokens": 728111396.0, "step": 19173 }, { "epoch": 2.439129881694441, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.789398670196533, "learning_rate": 1e-05, "loss": 0.4728, "mean_token_accuracy": 0.8737186789512634, "num_tokens": 728149609.0, "step": 19174 }, { "epoch": 2.4392570919730314, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 0.00011348724365234375, "grad_norm": 7.164877891540527, "learning_rate": 1e-05, "loss": 0.4949, "mean_token_accuracy": 0.8771322965621948, "num_tokens": 728186668.0, "step": 19175 }, { "epoch": 2.439384302251622, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.720198154449463, "learning_rate": 1e-05, "loss": 0.5125, "mean_token_accuracy": 0.8608204126358032, "num_tokens": 728226504.0, "step": 19176 }, { "epoch": 2.4395115125302125, "ewc_loss": 0.1513671875, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.616331100463867, "learning_rate": 1e-05, "loss": 0.5002, "mean_token_accuracy": 0.8760883808135986, "num_tokens": 728264648.0, "step": 19177 }, { "epoch": 2.439638722808803, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 5.937356472015381, "learning_rate": 1e-05, "loss": 0.4601, "mean_token_accuracy": 0.8788397908210754, "num_tokens": 728300292.0, "step": 19178 }, { "epoch": 2.4397659330873935, "ewc_loss": 0.16015625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 0.0001316070556640625, "grad_norm": 7.940252304077148, "learning_rate": 1e-05, "loss": 0.5202, "mean_token_accuracy": 0.8754183053970337, "num_tokens": 728336169.0, "step": 19179 }, { "epoch": 2.439893143365984, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 6.4670305252075195, "learning_rate": 1e-05, "loss": 0.4697, "mean_token_accuracy": 0.8831473588943481, "num_tokens": 728380160.0, "step": 19180 }, { "epoch": 2.4400203536445746, "ewc_loss": 0.15625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 0.0001277923583984375, "grad_norm": 7.775745868682861, "learning_rate": 1e-05, "loss": 0.4984, "mean_token_accuracy": 0.8789530992507935, "num_tokens": 728417978.0, "step": 19181 }, { "epoch": 2.440147563923165, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 0.00011539459228515625, "grad_norm": 6.811546802520752, "learning_rate": 1e-05, "loss": 0.4788, "mean_token_accuracy": 0.8809572458267212, "num_tokens": 728455264.0, "step": 19182 }, { "epoch": 2.4402747742017556, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 0.00011682510375976562, "grad_norm": 7.300665378570557, "learning_rate": 1e-05, "loss": 0.5285, "mean_token_accuracy": 0.8662734627723694, "num_tokens": 728498880.0, "step": 19183 }, { "epoch": 2.440401984480346, "ewc_loss": 0.140625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 0.0001125335693359375, "grad_norm": 6.803488254547119, "learning_rate": 1e-05, "loss": 0.4994, "mean_token_accuracy": 0.8741412162780762, "num_tokens": 728531925.0, "step": 19184 }, { "epoch": 2.4405291947589367, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.851740837097168, "learning_rate": 1e-05, "loss": 0.499, "mean_token_accuracy": 0.8723358511924744, "num_tokens": 728570573.0, "step": 19185 }, { "epoch": 2.440656405037527, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.705413341522217, "learning_rate": 1e-05, "loss": 0.4799, "mean_token_accuracy": 0.8792824149131775, "num_tokens": 728610877.0, "step": 19186 }, { "epoch": 2.4407836153161178, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.512166500091553, "learning_rate": 1e-05, "loss": 0.4814, "mean_token_accuracy": 0.8770637512207031, "num_tokens": 728650963.0, "step": 19187 }, { "epoch": 2.440910825594708, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.626004219055176, "learning_rate": 1e-05, "loss": 0.4479, "mean_token_accuracy": 0.8880964517593384, "num_tokens": 728688699.0, "step": 19188 }, { "epoch": 2.4410380358732984, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.29516077041626, "learning_rate": 1e-05, "loss": 0.5122, "mean_token_accuracy": 0.8657792806625366, "num_tokens": 728730142.0, "step": 19189 }, { "epoch": 2.441165246151889, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.593008518218994, "learning_rate": 1e-05, "loss": 0.5002, "mean_token_accuracy": 0.8696008920669556, "num_tokens": 728767808.0, "step": 19190 }, { "epoch": 2.4412924564304794, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.161875247955322, "learning_rate": 1e-05, "loss": 0.5078, "mean_token_accuracy": 0.8638622760772705, "num_tokens": 728816171.0, "step": 19191 }, { "epoch": 2.44141966670907, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.610320091247559, "learning_rate": 1e-05, "loss": 0.5364, "mean_token_accuracy": 0.8605400919914246, "num_tokens": 728858350.0, "step": 19192 }, { "epoch": 2.4415468769876605, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 6.035665035247803, "learning_rate": 1e-05, "loss": 0.4265, "mean_token_accuracy": 0.8893080949783325, "num_tokens": 728894408.0, "step": 19193 }, { "epoch": 2.441674087266251, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.700225830078125, "learning_rate": 1e-05, "loss": 0.5379, "mean_token_accuracy": 0.8565672636032104, "num_tokens": 728936563.0, "step": 19194 }, { "epoch": 2.4418012975448415, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.891201019287109, "learning_rate": 1e-05, "loss": 0.5048, "mean_token_accuracy": 0.8694109916687012, "num_tokens": 728978796.0, "step": 19195 }, { "epoch": 2.441928507823432, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.9212965965271, "learning_rate": 1e-05, "loss": 0.4684, "mean_token_accuracy": 0.8857914209365845, "num_tokens": 729018524.0, "step": 19196 }, { "epoch": 2.4420557181020226, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.760018825531006, "learning_rate": 1e-05, "loss": 0.4758, "mean_token_accuracy": 0.8740667700767517, "num_tokens": 729055336.0, "step": 19197 }, { "epoch": 2.442182928380613, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 0.00011587142944335938, "grad_norm": 7.29854679107666, "learning_rate": 1e-05, "loss": 0.5197, "mean_token_accuracy": 0.8680695295333862, "num_tokens": 729095505.0, "step": 19198 }, { "epoch": 2.4423101386592037, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.7327704429626465, "learning_rate": 1e-05, "loss": 0.4817, "mean_token_accuracy": 0.8754311800003052, "num_tokens": 729138384.0, "step": 19199 }, { "epoch": 2.442437348937794, "ewc_loss": 0.1572265625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 0.00012874603271484375, "grad_norm": 7.777760982513428, "learning_rate": 1e-05, "loss": 0.5703, "mean_token_accuracy": 0.8555964827537537, "num_tokens": 729179942.0, "step": 19200 }, { "epoch": 2.4425645592163847, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.098179340362549, "learning_rate": 1e-05, "loss": 0.5355, "mean_token_accuracy": 0.8585315942764282, "num_tokens": 729223051.0, "step": 19201 }, { "epoch": 2.4426917694949752, "ewc_loss": 0.162109375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 0.000133514404296875, "grad_norm": 7.958003520965576, "learning_rate": 1e-05, "loss": 0.5, "mean_token_accuracy": 0.8790069222450256, "num_tokens": 729259368.0, "step": 19202 }, { "epoch": 2.4428189797735658, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 0.00011491775512695312, "grad_norm": 6.653655529022217, "learning_rate": 1e-05, "loss": 0.5026, "mean_token_accuracy": 0.8727116584777832, "num_tokens": 729295148.0, "step": 19203 }, { "epoch": 2.4429461900521563, "ewc_loss": 0.15234375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 0.0001239776611328125, "grad_norm": 7.606420993804932, "learning_rate": 1e-05, "loss": 0.5285, "mean_token_accuracy": 0.8693327903747559, "num_tokens": 729330649.0, "step": 19204 }, { "epoch": 2.443073400330747, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 0.00011539459228515625, "grad_norm": 6.8285722732543945, "learning_rate": 1e-05, "loss": 0.4748, "mean_token_accuracy": 0.8853399157524109, "num_tokens": 729369260.0, "step": 19205 }, { "epoch": 2.4432006106093374, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 0.00011205673217773438, "grad_norm": 7.1265339851379395, "learning_rate": 1e-05, "loss": 0.4877, "mean_token_accuracy": 0.8784225583076477, "num_tokens": 729406029.0, "step": 19206 }, { "epoch": 2.443327820887928, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 0.00011014938354492188, "grad_norm": 6.749138832092285, "learning_rate": 1e-05, "loss": 0.4994, "mean_token_accuracy": 0.8725106120109558, "num_tokens": 729444102.0, "step": 19207 }, { "epoch": 2.4434550311665184, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.720345973968506, "learning_rate": 1e-05, "loss": 0.4062, "mean_token_accuracy": 0.8992923498153687, "num_tokens": 729477121.0, "step": 19208 }, { "epoch": 2.443582241445109, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.670426368713379, "learning_rate": 1e-05, "loss": 0.5135, "mean_token_accuracy": 0.8672439455986023, "num_tokens": 729511682.0, "step": 19209 }, { "epoch": 2.4437094517236995, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.433318138122559, "learning_rate": 1e-05, "loss": 0.4724, "mean_token_accuracy": 0.8826838135719299, "num_tokens": 729545675.0, "step": 19210 }, { "epoch": 2.4438366620022896, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.616209506988525, "learning_rate": 1e-05, "loss": 0.4934, "mean_token_accuracy": 0.8765004873275757, "num_tokens": 729585323.0, "step": 19211 }, { "epoch": 2.4439638722808805, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.237256050109863, "learning_rate": 1e-05, "loss": 0.4878, "mean_token_accuracy": 0.869794487953186, "num_tokens": 729623881.0, "step": 19212 }, { "epoch": 2.4440910825594706, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.595734119415283, "learning_rate": 1e-05, "loss": 0.4387, "mean_token_accuracy": 0.8923026323318481, "num_tokens": 729655486.0, "step": 19213 }, { "epoch": 2.444218292838061, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.107807159423828, "learning_rate": 1e-05, "loss": 0.4712, "mean_token_accuracy": 0.8776724338531494, "num_tokens": 729693606.0, "step": 19214 }, { "epoch": 2.4443455031166517, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.647139549255371, "learning_rate": 1e-05, "loss": 0.5063, "mean_token_accuracy": 0.8678007125854492, "num_tokens": 729729467.0, "step": 19215 }, { "epoch": 2.444472713395242, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 5.977268218994141, "learning_rate": 1e-05, "loss": 0.4372, "mean_token_accuracy": 0.8862724304199219, "num_tokens": 729767264.0, "step": 19216 }, { "epoch": 2.4445999236738327, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.793041706085205, "learning_rate": 1e-05, "loss": 0.4835, "mean_token_accuracy": 0.8801680207252502, "num_tokens": 729807352.0, "step": 19217 }, { "epoch": 2.4447271339524232, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.8356804847717285, "learning_rate": 1e-05, "loss": 0.5253, "mean_token_accuracy": 0.8585975170135498, "num_tokens": 729847489.0, "step": 19218 }, { "epoch": 2.4448543442310138, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 7.075894832611084, "learning_rate": 1e-05, "loss": 0.4697, "mean_token_accuracy": 0.8836328983306885, "num_tokens": 729889675.0, "step": 19219 }, { "epoch": 2.4449815545096043, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.720088958740234, "learning_rate": 1e-05, "loss": 0.4715, "mean_token_accuracy": 0.8748760223388672, "num_tokens": 729926562.0, "step": 19220 }, { "epoch": 2.445108764788195, "ewc_loss": 0.150390625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 0.0001220703125, "grad_norm": 7.526459693908691, "learning_rate": 1e-05, "loss": 0.5634, "mean_token_accuracy": 0.8612358570098877, "num_tokens": 729964533.0, "step": 19221 }, { "epoch": 2.4452359750667854, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 5.849420547485352, "learning_rate": 1e-05, "loss": 0.4399, "mean_token_accuracy": 0.8888535499572754, "num_tokens": 730004206.0, "step": 19222 }, { "epoch": 2.445363185345376, "ewc_loss": 0.162109375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 0.000133514404296875, "grad_norm": 7.934709072113037, "learning_rate": 1e-05, "loss": 0.5345, "mean_token_accuracy": 0.8704543709754944, "num_tokens": 730046779.0, "step": 19223 }, { "epoch": 2.4454903956239664, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.37739896774292, "learning_rate": 1e-05, "loss": 0.5176, "mean_token_accuracy": 0.8722223043441772, "num_tokens": 730085882.0, "step": 19224 }, { "epoch": 2.445617605902557, "ewc_loss": 0.158203125, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 0.00012969970703125, "grad_norm": 7.854683876037598, "learning_rate": 1e-05, "loss": 0.526, "mean_token_accuracy": 0.8745657801628113, "num_tokens": 730126072.0, "step": 19225 }, { "epoch": 2.4457448161811475, "ewc_loss": 0.1455078125, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 0.00011730194091796875, "grad_norm": 6.7929768562316895, "learning_rate": 1e-05, "loss": 0.4509, "mean_token_accuracy": 0.892790675163269, "num_tokens": 730160038.0, "step": 19226 }, { "epoch": 2.445872026459738, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.00011873245239257812, "grad_norm": 7.406540870666504, "learning_rate": 1e-05, "loss": 0.5043, "mean_token_accuracy": 0.8789240121841431, "num_tokens": 730195701.0, "step": 19227 }, { "epoch": 2.4459992367383285, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.00011348724365234375, "grad_norm": 6.815498352050781, "learning_rate": 1e-05, "loss": 0.4712, "mean_token_accuracy": 0.8813663721084595, "num_tokens": 730233938.0, "step": 19228 }, { "epoch": 2.446126447016919, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.943719863891602, "learning_rate": 1e-05, "loss": 0.4365, "mean_token_accuracy": 0.8926354646682739, "num_tokens": 730272333.0, "step": 19229 }, { "epoch": 2.4462536572955096, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 6.730586528778076, "learning_rate": 1e-05, "loss": 0.5147, "mean_token_accuracy": 0.8663676381111145, "num_tokens": 730312967.0, "step": 19230 }, { "epoch": 2.4463808675741, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.578707695007324, "learning_rate": 1e-05, "loss": 0.4694, "mean_token_accuracy": 0.8799170851707458, "num_tokens": 730345525.0, "step": 19231 }, { "epoch": 2.4465080778526906, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.636856555938721, "learning_rate": 1e-05, "loss": 0.4763, "mean_token_accuracy": 0.8815171122550964, "num_tokens": 730384329.0, "step": 19232 }, { "epoch": 2.446635288131281, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.335651397705078, "learning_rate": 1e-05, "loss": 0.4242, "mean_token_accuracy": 0.8928993940353394, "num_tokens": 730423139.0, "step": 19233 }, { "epoch": 2.4467624984098717, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.590210437774658, "learning_rate": 1e-05, "loss": 0.4794, "mean_token_accuracy": 0.8782872557640076, "num_tokens": 730463989.0, "step": 19234 }, { "epoch": 2.4468897086884622, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.201496124267578, "learning_rate": 1e-05, "loss": 0.4875, "mean_token_accuracy": 0.8696568012237549, "num_tokens": 730500854.0, "step": 19235 }, { "epoch": 2.4470169189670523, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.579980850219727, "learning_rate": 1e-05, "loss": 0.4274, "mean_token_accuracy": 0.8943028450012207, "num_tokens": 730537742.0, "step": 19236 }, { "epoch": 2.4471441292456433, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.088385581970215, "learning_rate": 1e-05, "loss": 0.4106, "mean_token_accuracy": 0.8957062363624573, "num_tokens": 730573090.0, "step": 19237 }, { "epoch": 2.4472713395242334, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.633870601654053, "learning_rate": 1e-05, "loss": 0.4293, "mean_token_accuracy": 0.894263744354248, "num_tokens": 730611865.0, "step": 19238 }, { "epoch": 2.447398549802824, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 5.970929145812988, "learning_rate": 1e-05, "loss": 0.5324, "mean_token_accuracy": 0.8592746257781982, "num_tokens": 730646831.0, "step": 19239 }, { "epoch": 2.4475257600814144, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.782176494598389, "learning_rate": 1e-05, "loss": 0.4806, "mean_token_accuracy": 0.8812615275382996, "num_tokens": 730689789.0, "step": 19240 }, { "epoch": 2.447652970360005, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.830390453338623, "learning_rate": 1e-05, "loss": 0.4733, "mean_token_accuracy": 0.8795483112335205, "num_tokens": 730720289.0, "step": 19241 }, { "epoch": 2.4477801806385955, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 0.00010967254638671875, "grad_norm": 7.057657241821289, "learning_rate": 1e-05, "loss": 0.4636, "mean_token_accuracy": 0.8856678009033203, "num_tokens": 730757499.0, "step": 19242 }, { "epoch": 2.447907390917186, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.7060546875, "learning_rate": 1e-05, "loss": 0.452, "mean_token_accuracy": 0.8821439146995544, "num_tokens": 730799897.0, "step": 19243 }, { "epoch": 2.4480346011957765, "ewc_loss": 0.1494140625, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.00012159347534179688, "grad_norm": 7.511404514312744, "learning_rate": 1e-05, "loss": 0.4807, "mean_token_accuracy": 0.88419508934021, "num_tokens": 730837709.0, "step": 19244 }, { "epoch": 2.448161811474367, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 5.824874401092529, "learning_rate": 1e-05, "loss": 0.4131, "mean_token_accuracy": 0.8962886333465576, "num_tokens": 730874128.0, "step": 19245 }, { "epoch": 2.4482890217529576, "ewc_loss": 0.1611328125, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.00013256072998046875, "grad_norm": 7.952836990356445, "learning_rate": 1e-05, "loss": 0.5243, "mean_token_accuracy": 0.8725670576095581, "num_tokens": 730917027.0, "step": 19246 }, { "epoch": 2.448416232031548, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.369950294494629, "learning_rate": 1e-05, "loss": 0.5226, "mean_token_accuracy": 0.8663594722747803, "num_tokens": 730955107.0, "step": 19247 }, { "epoch": 2.4485434423101387, "ewc_loss": 0.1591796875, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.00013065338134765625, "grad_norm": 7.8921427726745605, "learning_rate": 1e-05, "loss": 0.5442, "mean_token_accuracy": 0.8654758930206299, "num_tokens": 730995807.0, "step": 19248 }, { "epoch": 2.448670652588729, "ewc_loss": 0.1455078125, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.00011777877807617188, "grad_norm": 6.805948257446289, "learning_rate": 1e-05, "loss": 0.4564, "mean_token_accuracy": 0.891351044178009, "num_tokens": 731034935.0, "step": 19249 }, { "epoch": 2.4487978628673197, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.000118255615234375, "grad_norm": 7.426120758056641, "learning_rate": 1e-05, "loss": 0.5011, "mean_token_accuracy": 0.8752409815788269, "num_tokens": 731070934.0, "step": 19250 }, { "epoch": 2.4489250731459102, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.00011348724365234375, "grad_norm": 6.825589656829834, "learning_rate": 1e-05, "loss": 0.4733, "mean_token_accuracy": 0.8850393891334534, "num_tokens": 731112625.0, "step": 19251 }, { "epoch": 2.4490522834245008, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 6.987971782684326, "learning_rate": 1e-05, "loss": 0.493, "mean_token_accuracy": 0.8732374906539917, "num_tokens": 731151011.0, "step": 19252 }, { "epoch": 2.4491794937030913, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 6.7222747802734375, "learning_rate": 1e-05, "loss": 0.4807, "mean_token_accuracy": 0.8790367841720581, "num_tokens": 731190885.0, "step": 19253 }, { "epoch": 2.449306703981682, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.623330593109131, "learning_rate": 1e-05, "loss": 0.4657, "mean_token_accuracy": 0.8802827596664429, "num_tokens": 731229056.0, "step": 19254 }, { "epoch": 2.4494339142602723, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.631325721740723, "learning_rate": 1e-05, "loss": 0.4933, "mean_token_accuracy": 0.877083957195282, "num_tokens": 731268098.0, "step": 19255 }, { "epoch": 2.449561124538863, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.387782573699951, "learning_rate": 1e-05, "loss": 0.5175, "mean_token_accuracy": 0.8626053333282471, "num_tokens": 731307809.0, "step": 19256 }, { "epoch": 2.4496883348174534, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.573750019073486, "learning_rate": 1e-05, "loss": 0.5045, "mean_token_accuracy": 0.8714143633842468, "num_tokens": 731343904.0, "step": 19257 }, { "epoch": 2.449815545096044, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.228550910949707, "learning_rate": 1e-05, "loss": 0.4266, "mean_token_accuracy": 0.891721785068512, "num_tokens": 731383120.0, "step": 19258 }, { "epoch": 2.4499427553746345, "ewc_loss": 0.125, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.56235408782959, "learning_rate": 1e-05, "loss": 0.4759, "mean_token_accuracy": 0.8794335722923279, "num_tokens": 731423020.0, "step": 19259 }, { "epoch": 2.450069965653225, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.098949909210205, "learning_rate": 1e-05, "loss": 0.4788, "mean_token_accuracy": 0.8730899691581726, "num_tokens": 731463522.0, "step": 19260 }, { "epoch": 2.450197175931815, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.613780975341797, "learning_rate": 1e-05, "loss": 0.5052, "mean_token_accuracy": 0.8673805594444275, "num_tokens": 731502327.0, "step": 19261 }, { "epoch": 2.450324386210406, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 5.970159530639648, "learning_rate": 1e-05, "loss": 0.4304, "mean_token_accuracy": 0.8891853094100952, "num_tokens": 731545253.0, "step": 19262 }, { "epoch": 2.450451596488996, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.790321350097656, "learning_rate": 1e-05, "loss": 0.4484, "mean_token_accuracy": 0.8855236768722534, "num_tokens": 731575388.0, "step": 19263 }, { "epoch": 2.4505788067675867, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.811544418334961, "learning_rate": 1e-05, "loss": 0.4706, "mean_token_accuracy": 0.8754600286483765, "num_tokens": 731618195.0, "step": 19264 }, { "epoch": 2.450706017046177, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 7.114051342010498, "learning_rate": 1e-05, "loss": 0.5158, "mean_token_accuracy": 0.8688533306121826, "num_tokens": 731658551.0, "step": 19265 }, { "epoch": 2.4508332273247677, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.685179233551025, "learning_rate": 1e-05, "loss": 0.4219, "mean_token_accuracy": 0.8897017240524292, "num_tokens": 731700477.0, "step": 19266 }, { "epoch": 2.4509604376033582, "ewc_loss": 0.15234375, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.0001239776611328125, "grad_norm": 7.615102291107178, "learning_rate": 1e-05, "loss": 0.5003, "mean_token_accuracy": 0.8794282674789429, "num_tokens": 731735841.0, "step": 19267 }, { "epoch": 2.4510876478819488, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 5.890073776245117, "learning_rate": 1e-05, "loss": 0.4894, "mean_token_accuracy": 0.8715041875839233, "num_tokens": 731774974.0, "step": 19268 }, { "epoch": 2.4512148581605393, "ewc_loss": 0.1630859375, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.00013446807861328125, "grad_norm": 8.006290435791016, "learning_rate": 1e-05, "loss": 0.5007, "mean_token_accuracy": 0.8820674419403076, "num_tokens": 731806010.0, "step": 19269 }, { "epoch": 2.45134206843913, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 6.468728542327881, "learning_rate": 1e-05, "loss": 0.4368, "mean_token_accuracy": 0.8945513963699341, "num_tokens": 731841773.0, "step": 19270 }, { "epoch": 2.4514692787177204, "ewc_loss": 0.1572265625, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.00012874603271484375, "grad_norm": 7.855062007904053, "learning_rate": 1e-05, "loss": 0.4823, "mean_token_accuracy": 0.885819137096405, "num_tokens": 731877974.0, "step": 19271 }, { "epoch": 2.451596488996311, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.00011873245239257812, "grad_norm": 6.853541374206543, "learning_rate": 1e-05, "loss": 0.4917, "mean_token_accuracy": 0.8793724775314331, "num_tokens": 731919375.0, "step": 19272 }, { "epoch": 2.4517236992749014, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.00011682510375976562, "grad_norm": 7.383680820465088, "learning_rate": 1e-05, "loss": 0.4848, "mean_token_accuracy": 0.8795473575592041, "num_tokens": 731957518.0, "step": 19273 }, { "epoch": 2.451850909553492, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.00011348724365234375, "grad_norm": 6.82058572769165, "learning_rate": 1e-05, "loss": 0.4591, "mean_token_accuracy": 0.8874819278717041, "num_tokens": 731996597.0, "step": 19274 }, { "epoch": 2.4519781198320825, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.950207710266113, "learning_rate": 1e-05, "loss": 0.488, "mean_token_accuracy": 0.878027081489563, "num_tokens": 732035888.0, "step": 19275 }, { "epoch": 2.452105330110673, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.698281764984131, "learning_rate": 1e-05, "loss": 0.4934, "mean_token_accuracy": 0.8752095699310303, "num_tokens": 732073026.0, "step": 19276 }, { "epoch": 2.4522325403892635, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.646101474761963, "learning_rate": 1e-05, "loss": 0.5547, "mean_token_accuracy": 0.8539575338363647, "num_tokens": 732110794.0, "step": 19277 }, { "epoch": 2.452359750667854, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.587277889251709, "learning_rate": 1e-05, "loss": 0.4502, "mean_token_accuracy": 0.8857187032699585, "num_tokens": 732141051.0, "step": 19278 }, { "epoch": 2.4524869609464446, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.422129154205322, "learning_rate": 1e-05, "loss": 0.4653, "mean_token_accuracy": 0.882370114326477, "num_tokens": 732187023.0, "step": 19279 }, { "epoch": 2.452614171225035, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.506870746612549, "learning_rate": 1e-05, "loss": 0.5101, "mean_token_accuracy": 0.864392876625061, "num_tokens": 732220960.0, "step": 19280 }, { "epoch": 2.4527413815036256, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.305877685546875, "learning_rate": 1e-05, "loss": 0.4757, "mean_token_accuracy": 0.8786399364471436, "num_tokens": 732260656.0, "step": 19281 }, { "epoch": 2.452868591782216, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.450133800506592, "learning_rate": 1e-05, "loss": 0.5553, "mean_token_accuracy": 0.8503333926200867, "num_tokens": 732301771.0, "step": 19282 }, { "epoch": 2.4529958020608067, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.2155561447143555, "learning_rate": 1e-05, "loss": 0.4985, "mean_token_accuracy": 0.8712433576583862, "num_tokens": 732341618.0, "step": 19283 }, { "epoch": 2.453123012339397, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.419239521026611, "learning_rate": 1e-05, "loss": 0.3852, "mean_token_accuracy": 0.9082093238830566, "num_tokens": 732375797.0, "step": 19284 }, { "epoch": 2.4532502226179878, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.134777545928955, "learning_rate": 1e-05, "loss": 0.4425, "mean_token_accuracy": 0.8855723738670349, "num_tokens": 732414234.0, "step": 19285 }, { "epoch": 2.453377432896578, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.448692798614502, "learning_rate": 1e-05, "loss": 0.4864, "mean_token_accuracy": 0.8731988668441772, "num_tokens": 732461063.0, "step": 19286 }, { "epoch": 2.4535046431751684, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 6.022335529327393, "learning_rate": 1e-05, "loss": 0.4565, "mean_token_accuracy": 0.8796173334121704, "num_tokens": 732501622.0, "step": 19287 }, { "epoch": 2.453631853453759, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.565789699554443, "learning_rate": 1e-05, "loss": 0.4938, "mean_token_accuracy": 0.8714483380317688, "num_tokens": 732546938.0, "step": 19288 }, { "epoch": 2.4537590637323494, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 5.882734298706055, "learning_rate": 1e-05, "loss": 0.4178, "mean_token_accuracy": 0.893036961555481, "num_tokens": 732581250.0, "step": 19289 }, { "epoch": 2.45388627401094, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.798398971557617, "learning_rate": 1e-05, "loss": 0.4534, "mean_token_accuracy": 0.8885160684585571, "num_tokens": 732621081.0, "step": 19290 }, { "epoch": 2.4540134842895305, "ewc_loss": 0.111328125, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.688653945922852, "learning_rate": 1e-05, "loss": 0.4371, "mean_token_accuracy": 0.8849934339523315, "num_tokens": 732660625.0, "step": 19291 }, { "epoch": 2.454140694568121, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 0.00011539459228515625, "grad_norm": 7.2611589431762695, "learning_rate": 1e-05, "loss": 0.4819, "mean_token_accuracy": 0.8810908794403076, "num_tokens": 732700100.0, "step": 19292 }, { "epoch": 2.4542679048467115, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.582493305206299, "learning_rate": 1e-05, "loss": 0.4628, "mean_token_accuracy": 0.879380464553833, "num_tokens": 732735252.0, "step": 19293 }, { "epoch": 2.454395115125302, "ewc_loss": 0.1611328125, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.00013256072998046875, "grad_norm": 7.953397274017334, "learning_rate": 1e-05, "loss": 0.5376, "mean_token_accuracy": 0.8726412653923035, "num_tokens": 732767912.0, "step": 19294 }, { "epoch": 2.4545223254038926, "ewc_loss": 0.125, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 5.995206356048584, "learning_rate": 1e-05, "loss": 0.4972, "mean_token_accuracy": 0.8715583086013794, "num_tokens": 732808499.0, "step": 19295 }, { "epoch": 2.454649535682483, "ewc_loss": 0.1708984375, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.00014209747314453125, "grad_norm": 8.379159927368164, "learning_rate": 1e-05, "loss": 0.5207, "mean_token_accuracy": 0.8768444061279297, "num_tokens": 732840754.0, "step": 19296 }, { "epoch": 2.4547767459610736, "ewc_loss": 0.150390625, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.0001220703125, "grad_norm": 6.902014255523682, "learning_rate": 1e-05, "loss": 0.5674, "mean_token_accuracy": 0.8562866449356079, "num_tokens": 732882432.0, "step": 19297 }, { "epoch": 2.454903956239664, "ewc_loss": 0.1572265625, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.00012874603271484375, "grad_norm": 7.938904285430908, "learning_rate": 1e-05, "loss": 0.4881, "mean_token_accuracy": 0.8868966102600098, "num_tokens": 732912418.0, "step": 19298 }, { "epoch": 2.4550311665182547, "ewc_loss": 0.15234375, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.0001239776611328125, "grad_norm": 7.149867534637451, "learning_rate": 1e-05, "loss": 0.4927, "mean_token_accuracy": 0.8805195093154907, "num_tokens": 732946050.0, "step": 19299 }, { "epoch": 2.4551583767968452, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.00011682510375976562, "grad_norm": 7.41217565536499, "learning_rate": 1e-05, "loss": 0.4712, "mean_token_accuracy": 0.8855543732643127, "num_tokens": 732980667.0, "step": 19300 }, { "epoch": 2.4552855870754358, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.00011444091796875, "grad_norm": 6.906306743621826, "learning_rate": 1e-05, "loss": 0.4922, "mean_token_accuracy": 0.8800815343856812, "num_tokens": 733016169.0, "step": 19301 }, { "epoch": 2.4554127973540263, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.00011014938354492188, "grad_norm": 7.063092231750488, "learning_rate": 1e-05, "loss": 0.5039, "mean_token_accuracy": 0.8725453019142151, "num_tokens": 733055120.0, "step": 19302 }, { "epoch": 2.455540007632617, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.659729957580566, "learning_rate": 1e-05, "loss": 0.5196, "mean_token_accuracy": 0.8661394715309143, "num_tokens": 733094550.0, "step": 19303 }, { "epoch": 2.4556672179112073, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.851884365081787, "learning_rate": 1e-05, "loss": 0.4326, "mean_token_accuracy": 0.8917103409767151, "num_tokens": 733132338.0, "step": 19304 }, { "epoch": 2.455794428189798, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.448126316070557, "learning_rate": 1e-05, "loss": 0.489, "mean_token_accuracy": 0.8744111657142639, "num_tokens": 733167053.0, "step": 19305 }, { "epoch": 2.4559216384683884, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.764893054962158, "learning_rate": 1e-05, "loss": 0.5131, "mean_token_accuracy": 0.8684518337249756, "num_tokens": 733208175.0, "step": 19306 }, { "epoch": 2.456048848746979, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.250180244445801, "learning_rate": 1e-05, "loss": 0.5037, "mean_token_accuracy": 0.8662327527999878, "num_tokens": 733248336.0, "step": 19307 }, { "epoch": 2.4561760590255695, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.823845386505127, "learning_rate": 1e-05, "loss": 0.4875, "mean_token_accuracy": 0.8763677477836609, "num_tokens": 733283314.0, "step": 19308 }, { "epoch": 2.4563032693041595, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 6.0584492683410645, "learning_rate": 1e-05, "loss": 0.462, "mean_token_accuracy": 0.8767557144165039, "num_tokens": 733317541.0, "step": 19309 }, { "epoch": 2.4564304795827505, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 7.016757488250732, "learning_rate": 1e-05, "loss": 0.423, "mean_token_accuracy": 0.8974305391311646, "num_tokens": 733349975.0, "step": 19310 }, { "epoch": 2.4565576898613406, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.901548862457275, "learning_rate": 1e-05, "loss": 0.4377, "mean_token_accuracy": 0.8880913257598877, "num_tokens": 733388085.0, "step": 19311 }, { "epoch": 2.456684900139931, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.00011587142944335938, "grad_norm": 7.3755784034729, "learning_rate": 1e-05, "loss": 0.4608, "mean_token_accuracy": 0.8864949941635132, "num_tokens": 733428420.0, "step": 19312 }, { "epoch": 2.4568121104185217, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 5.9431843757629395, "learning_rate": 1e-05, "loss": 0.4453, "mean_token_accuracy": 0.8821220397949219, "num_tokens": 733460271.0, "step": 19313 }, { "epoch": 2.456939320697112, "ewc_loss": 0.1533203125, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.00012493133544921875, "grad_norm": 7.734300136566162, "learning_rate": 1e-05, "loss": 0.5088, "mean_token_accuracy": 0.8760839700698853, "num_tokens": 733498610.0, "step": 19314 }, { "epoch": 2.4570665309757027, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.33189582824707, "learning_rate": 1e-05, "loss": 0.4998, "mean_token_accuracy": 0.8701204061508179, "num_tokens": 733530501.0, "step": 19315 }, { "epoch": 2.4571937412542932, "ewc_loss": 0.15234375, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.0001239776611328125, "grad_norm": 7.687395095825195, "learning_rate": 1e-05, "loss": 0.4808, "mean_token_accuracy": 0.8827639818191528, "num_tokens": 733563208.0, "step": 19316 }, { "epoch": 2.4573209515328838, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.00010967254638671875, "grad_norm": 6.724748134613037, "learning_rate": 1e-05, "loss": 0.4742, "mean_token_accuracy": 0.8850505352020264, "num_tokens": 733598729.0, "step": 19317 }, { "epoch": 2.4574481618114743, "ewc_loss": 0.140625, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.00011301040649414062, "grad_norm": 7.2054219245910645, "learning_rate": 1e-05, "loss": 0.4971, "mean_token_accuracy": 0.8742544054985046, "num_tokens": 733632033.0, "step": 19318 }, { "epoch": 2.457575372090065, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.00011014938354492188, "grad_norm": 6.812807559967041, "learning_rate": 1e-05, "loss": 0.5196, "mean_token_accuracy": 0.8704348802566528, "num_tokens": 733673088.0, "step": 19319 }, { "epoch": 2.4577025823686554, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.64218807220459, "learning_rate": 1e-05, "loss": 0.5139, "mean_token_accuracy": 0.867120623588562, "num_tokens": 733714096.0, "step": 19320 }, { "epoch": 2.457829792647246, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 6.856210231781006, "learning_rate": 1e-05, "loss": 0.5121, "mean_token_accuracy": 0.8698141574859619, "num_tokens": 733754495.0, "step": 19321 }, { "epoch": 2.4579570029258364, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.1612548828125, "learning_rate": 1e-05, "loss": 0.4636, "mean_token_accuracy": 0.8773722052574158, "num_tokens": 733788904.0, "step": 19322 }, { "epoch": 2.458084213204427, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 7.036191463470459, "learning_rate": 1e-05, "loss": 0.4773, "mean_token_accuracy": 0.8818441033363342, "num_tokens": 733831041.0, "step": 19323 }, { "epoch": 2.4582114234830175, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.879586219787598, "learning_rate": 1e-05, "loss": 0.4932, "mean_token_accuracy": 0.8671430349349976, "num_tokens": 733872209.0, "step": 19324 }, { "epoch": 2.458338633761608, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.00011539459228515625, "grad_norm": 7.378467082977295, "learning_rate": 1e-05, "loss": 0.5156, "mean_token_accuracy": 0.8702248334884644, "num_tokens": 733906884.0, "step": 19325 }, { "epoch": 2.4584658440401985, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 5.875524520874023, "learning_rate": 1e-05, "loss": 0.4224, "mean_token_accuracy": 0.8932563066482544, "num_tokens": 733949262.0, "step": 19326 }, { "epoch": 2.458593054318789, "ewc_loss": 0.1552734375, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.00012683868408203125, "grad_norm": 7.753484725952148, "learning_rate": 1e-05, "loss": 0.4842, "mean_token_accuracy": 0.884084165096283, "num_tokens": 733991111.0, "step": 19327 }, { "epoch": 2.4587202645973796, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.244124889373779, "learning_rate": 1e-05, "loss": 0.5002, "mean_token_accuracy": 0.8740983605384827, "num_tokens": 734030968.0, "step": 19328 }, { "epoch": 2.45884747487597, "ewc_loss": 0.1552734375, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.00012683868408203125, "grad_norm": 7.7906293869018555, "learning_rate": 1e-05, "loss": 0.5291, "mean_token_accuracy": 0.8692604303359985, "num_tokens": 734070612.0, "step": 19329 }, { "epoch": 2.4589746851545606, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.00010967254638671875, "grad_norm": 6.643971920013428, "learning_rate": 1e-05, "loss": 0.4783, "mean_token_accuracy": 0.8813787698745728, "num_tokens": 734108682.0, "step": 19330 }, { "epoch": 2.459101895433151, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.000118255615234375, "grad_norm": 7.456847667694092, "learning_rate": 1e-05, "loss": 0.4822, "mean_token_accuracy": 0.8829720616340637, "num_tokens": 734141565.0, "step": 19331 }, { "epoch": 2.4592291057117417, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 6.715560436248779, "learning_rate": 1e-05, "loss": 0.4752, "mean_token_accuracy": 0.8816831707954407, "num_tokens": 734183048.0, "step": 19332 }, { "epoch": 2.459356315990332, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 7.092818737030029, "learning_rate": 1e-05, "loss": 0.5062, "mean_token_accuracy": 0.8708720207214355, "num_tokens": 734216141.0, "step": 19333 }, { "epoch": 2.4594835262689223, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.608752727508545, "learning_rate": 1e-05, "loss": 0.481, "mean_token_accuracy": 0.8776645660400391, "num_tokens": 734256365.0, "step": 19334 }, { "epoch": 2.4596107365475133, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.822309970855713, "learning_rate": 1e-05, "loss": 0.4463, "mean_token_accuracy": 0.8884848356246948, "num_tokens": 734297679.0, "step": 19335 }, { "epoch": 2.4597379468261034, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.468940734863281, "learning_rate": 1e-05, "loss": 0.5207, "mean_token_accuracy": 0.8633868098258972, "num_tokens": 734333245.0, "step": 19336 }, { "epoch": 2.459865157104694, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.70864200592041, "learning_rate": 1e-05, "loss": 0.5103, "mean_token_accuracy": 0.8680135011672974, "num_tokens": 734374408.0, "step": 19337 }, { "epoch": 2.4599923673832844, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.286815643310547, "learning_rate": 1e-05, "loss": 0.4583, "mean_token_accuracy": 0.8821759223937988, "num_tokens": 734408105.0, "step": 19338 }, { "epoch": 2.460119577661875, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.726068019866943, "learning_rate": 1e-05, "loss": 0.4689, "mean_token_accuracy": 0.8803609013557434, "num_tokens": 734451700.0, "step": 19339 }, { "epoch": 2.4602467879404655, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.109648704528809, "learning_rate": 1e-05, "loss": 0.5093, "mean_token_accuracy": 0.869120717048645, "num_tokens": 734492343.0, "step": 19340 }, { "epoch": 2.460373998219056, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.867790699005127, "learning_rate": 1e-05, "loss": 0.4799, "mean_token_accuracy": 0.8783016204833984, "num_tokens": 734531720.0, "step": 19341 }, { "epoch": 2.4605012084976465, "ewc_loss": 0.11376953125, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.959293365478516, "learning_rate": 1e-05, "loss": 0.4473, "mean_token_accuracy": 0.8816163539886475, "num_tokens": 734567809.0, "step": 19342 }, { "epoch": 2.460628418776237, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 7.116533279418945, "learning_rate": 1e-05, "loss": 0.4198, "mean_token_accuracy": 0.8962423801422119, "num_tokens": 734605491.0, "step": 19343 }, { "epoch": 2.4607556290548276, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.92068338394165, "learning_rate": 1e-05, "loss": 0.4436, "mean_token_accuracy": 0.8857868313789368, "num_tokens": 734637169.0, "step": 19344 }, { "epoch": 2.460882839333418, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.00011682510375976562, "grad_norm": 7.419780254364014, "learning_rate": 1e-05, "loss": 0.4686, "mean_token_accuracy": 0.8829573392868042, "num_tokens": 734668167.0, "step": 19345 }, { "epoch": 2.4610100496120086, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.075917720794678, "learning_rate": 1e-05, "loss": 0.4267, "mean_token_accuracy": 0.8923643827438354, "num_tokens": 734705641.0, "step": 19346 }, { "epoch": 2.461137259890599, "ewc_loss": 0.1474609375, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.00011968612670898438, "grad_norm": 7.541713714599609, "learning_rate": 1e-05, "loss": 0.4991, "mean_token_accuracy": 0.8734827637672424, "num_tokens": 734742412.0, "step": 19347 }, { "epoch": 2.4612644701691897, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.4078240394592285, "learning_rate": 1e-05, "loss": 0.4401, "mean_token_accuracy": 0.8903565406799316, "num_tokens": 734782930.0, "step": 19348 }, { "epoch": 2.4613916804477802, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.00011491775512695312, "grad_norm": 7.294830322265625, "learning_rate": 1e-05, "loss": 0.4662, "mean_token_accuracy": 0.8874574303627014, "num_tokens": 734825051.0, "step": 19349 }, { "epoch": 2.4615188907263708, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.630209445953369, "learning_rate": 1e-05, "loss": 0.5067, "mean_token_accuracy": 0.867932915687561, "num_tokens": 734863524.0, "step": 19350 }, { "epoch": 2.4616461010049613, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.803201675415039, "learning_rate": 1e-05, "loss": 0.4943, "mean_token_accuracy": 0.8713353276252747, "num_tokens": 734899692.0, "step": 19351 }, { "epoch": 2.461773311283552, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.730787754058838, "learning_rate": 1e-05, "loss": 0.5459, "mean_token_accuracy": 0.8608059883117676, "num_tokens": 734933114.0, "step": 19352 }, { "epoch": 2.4619005215621423, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.273180961608887, "learning_rate": 1e-05, "loss": 0.5344, "mean_token_accuracy": 0.8587839603424072, "num_tokens": 734975896.0, "step": 19353 }, { "epoch": 2.462027731840733, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.8930511474609375, "learning_rate": 1e-05, "loss": 0.5212, "mean_token_accuracy": 0.8674612641334534, "num_tokens": 735017969.0, "step": 19354 }, { "epoch": 2.4621549421193234, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.890254020690918, "learning_rate": 1e-05, "loss": 0.4781, "mean_token_accuracy": 0.8741754293441772, "num_tokens": 735057052.0, "step": 19355 }, { "epoch": 2.462282152397914, "ewc_loss": 0.140625, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.00011301040649414062, "grad_norm": 7.249853610992432, "learning_rate": 1e-05, "loss": 0.4446, "mean_token_accuracy": 0.8929724097251892, "num_tokens": 735094211.0, "step": 19356 }, { "epoch": 2.4624093626765045, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.76397705078125, "learning_rate": 1e-05, "loss": 0.5282, "mean_token_accuracy": 0.8586294651031494, "num_tokens": 735134870.0, "step": 19357 }, { "epoch": 2.462536572955095, "ewc_loss": 0.154296875, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.000125885009765625, "grad_norm": 7.713281154632568, "learning_rate": 1e-05, "loss": 0.5415, "mean_token_accuracy": 0.8649020195007324, "num_tokens": 735177218.0, "step": 19358 }, { "epoch": 2.462663783233685, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.079834938049316, "learning_rate": 1e-05, "loss": 0.5621, "mean_token_accuracy": 0.8509626984596252, "num_tokens": 735208310.0, "step": 19359 }, { "epoch": 2.462790993512276, "ewc_loss": 0.158203125, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.00012969970703125, "grad_norm": 7.9104204177856445, "learning_rate": 1e-05, "loss": 0.4753, "mean_token_accuracy": 0.8861072063446045, "num_tokens": 735243514.0, "step": 19360 }, { "epoch": 2.462918203790866, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 6.604703903198242, "learning_rate": 1e-05, "loss": 0.5091, "mean_token_accuracy": 0.8710293769836426, "num_tokens": 735278765.0, "step": 19361 }, { "epoch": 2.4630454140694567, "ewc_loss": 0.1494140625, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.00012159347534179688, "grad_norm": 7.59782075881958, "learning_rate": 1e-05, "loss": 0.5343, "mean_token_accuracy": 0.8614858388900757, "num_tokens": 735317966.0, "step": 19362 }, { "epoch": 2.463172624348047, "ewc_loss": 0.140625, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.00011301040649414062, "grad_norm": 6.792098522186279, "learning_rate": 1e-05, "loss": 0.4683, "mean_token_accuracy": 0.8819571733474731, "num_tokens": 735355552.0, "step": 19363 }, { "epoch": 2.4632998346266377, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.00011348724365234375, "grad_norm": 7.166067600250244, "learning_rate": 1e-05, "loss": 0.5114, "mean_token_accuracy": 0.8706569075584412, "num_tokens": 735398011.0, "step": 19364 }, { "epoch": 2.4634270449052282, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 6.70090389251709, "learning_rate": 1e-05, "loss": 0.452, "mean_token_accuracy": 0.890203595161438, "num_tokens": 735436743.0, "step": 19365 }, { "epoch": 2.4635542551838188, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.835935115814209, "learning_rate": 1e-05, "loss": 0.4415, "mean_token_accuracy": 0.8912132978439331, "num_tokens": 735467932.0, "step": 19366 }, { "epoch": 2.4636814654624093, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.5516037940979, "learning_rate": 1e-05, "loss": 0.4516, "mean_token_accuracy": 0.8870550394058228, "num_tokens": 735510057.0, "step": 19367 }, { "epoch": 2.463808675741, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.639240264892578, "learning_rate": 1e-05, "loss": 0.4872, "mean_token_accuracy": 0.8735846281051636, "num_tokens": 735553700.0, "step": 19368 }, { "epoch": 2.4639358860195903, "ewc_loss": 0.125, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.392186641693115, "learning_rate": 1e-05, "loss": 0.5411, "mean_token_accuracy": 0.8589909076690674, "num_tokens": 735598490.0, "step": 19369 }, { "epoch": 2.464063096298181, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.597588062286377, "learning_rate": 1e-05, "loss": 0.4795, "mean_token_accuracy": 0.8719218969345093, "num_tokens": 735637987.0, "step": 19370 }, { "epoch": 2.4641903065767714, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.208273887634277, "learning_rate": 1e-05, "loss": 0.4089, "mean_token_accuracy": 0.8986879587173462, "num_tokens": 735676327.0, "step": 19371 }, { "epoch": 2.464317516855362, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.662691116333008, "learning_rate": 1e-05, "loss": 0.4963, "mean_token_accuracy": 0.8745590448379517, "num_tokens": 735713015.0, "step": 19372 }, { "epoch": 2.4644447271339525, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 6.035450458526611, "learning_rate": 1e-05, "loss": 0.4552, "mean_token_accuracy": 0.8793678283691406, "num_tokens": 735750561.0, "step": 19373 }, { "epoch": 2.464571937412543, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.863585948944092, "learning_rate": 1e-05, "loss": 0.5085, "mean_token_accuracy": 0.8688666224479675, "num_tokens": 735788264.0, "step": 19374 }, { "epoch": 2.4646991476911335, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.87378454208374, "learning_rate": 1e-05, "loss": 0.4812, "mean_token_accuracy": 0.8770772218704224, "num_tokens": 735829384.0, "step": 19375 }, { "epoch": 2.464826357969724, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.00011157989501953125, "grad_norm": 7.181967258453369, "learning_rate": 1e-05, "loss": 0.4423, "mean_token_accuracy": 0.8932808637619019, "num_tokens": 735868323.0, "step": 19376 }, { "epoch": 2.4649535682483146, "ewc_loss": 0.11474609375, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.827284336090088, "learning_rate": 1e-05, "loss": 0.4295, "mean_token_accuracy": 0.8873497247695923, "num_tokens": 735902107.0, "step": 19377 }, { "epoch": 2.465080778526905, "ewc_loss": 0.1513671875, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.577896595001221, "learning_rate": 1e-05, "loss": 0.5016, "mean_token_accuracy": 0.8798433542251587, "num_tokens": 735941501.0, "step": 19378 }, { "epoch": 2.4652079888054956, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.063670635223389, "learning_rate": 1e-05, "loss": 0.4962, "mean_token_accuracy": 0.8716331720352173, "num_tokens": 735987721.0, "step": 19379 }, { "epoch": 2.465335199084086, "ewc_loss": 0.1552734375, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.00012683868408203125, "grad_norm": 7.748384475708008, "learning_rate": 1e-05, "loss": 0.4689, "mean_token_accuracy": 0.8916223645210266, "num_tokens": 736022126.0, "step": 19380 }, { "epoch": 2.4654624093626767, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.5138983726501465, "learning_rate": 1e-05, "loss": 0.4729, "mean_token_accuracy": 0.8843711614608765, "num_tokens": 736059782.0, "step": 19381 }, { "epoch": 2.4655896196412668, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.000118255615234375, "grad_norm": 7.451646327972412, "learning_rate": 1e-05, "loss": 0.5253, "mean_token_accuracy": 0.8655343651771545, "num_tokens": 736101514.0, "step": 19382 }, { "epoch": 2.4657168299198577, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.00011205673217773438, "grad_norm": 6.744332313537598, "learning_rate": 1e-05, "loss": 0.4584, "mean_token_accuracy": 0.8853142857551575, "num_tokens": 736137861.0, "step": 19383 }, { "epoch": 2.465844040198448, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.939979076385498, "learning_rate": 1e-05, "loss": 0.4922, "mean_token_accuracy": 0.8731669783592224, "num_tokens": 736177567.0, "step": 19384 }, { "epoch": 2.4659712504770384, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 6.772735595703125, "learning_rate": 1e-05, "loss": 0.4835, "mean_token_accuracy": 0.8785707950592041, "num_tokens": 736214893.0, "step": 19385 }, { "epoch": 2.466098460755629, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.452746391296387, "learning_rate": 1e-05, "loss": 0.5202, "mean_token_accuracy": 0.864945650100708, "num_tokens": 736250539.0, "step": 19386 }, { "epoch": 2.4662256710342194, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.83196496963501, "learning_rate": 1e-05, "loss": 0.4448, "mean_token_accuracy": 0.8912700414657593, "num_tokens": 736280924.0, "step": 19387 }, { "epoch": 2.46635288131281, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 6.076200008392334, "learning_rate": 1e-05, "loss": 0.4714, "mean_token_accuracy": 0.8779462575912476, "num_tokens": 736322446.0, "step": 19388 }, { "epoch": 2.4664800915914005, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 7.030792713165283, "learning_rate": 1e-05, "loss": 0.4768, "mean_token_accuracy": 0.8812659978866577, "num_tokens": 736361370.0, "step": 19389 }, { "epoch": 2.466607301869991, "ewc_loss": 0.11474609375, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.880948066711426, "learning_rate": 1e-05, "loss": 0.4642, "mean_token_accuracy": 0.8795732259750366, "num_tokens": 736401931.0, "step": 19390 }, { "epoch": 2.4667345121485815, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.00011491775512695312, "grad_norm": 7.3423171043396, "learning_rate": 1e-05, "loss": 0.4926, "mean_token_accuracy": 0.8804479241371155, "num_tokens": 736445125.0, "step": 19391 }, { "epoch": 2.466861722427172, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 5.913853645324707, "learning_rate": 1e-05, "loss": 0.4171, "mean_token_accuracy": 0.8973392844200134, "num_tokens": 736488409.0, "step": 19392 }, { "epoch": 2.4669889327057626, "ewc_loss": 0.1513671875, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.6363372802734375, "learning_rate": 1e-05, "loss": 0.5127, "mean_token_accuracy": 0.8745930790901184, "num_tokens": 736526979.0, "step": 19393 }, { "epoch": 2.467116142984353, "ewc_loss": 0.125, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.205509185791016, "learning_rate": 1e-05, "loss": 0.4448, "mean_token_accuracy": 0.8862501382827759, "num_tokens": 736571933.0, "step": 19394 }, { "epoch": 2.4672433532629436, "ewc_loss": 0.15234375, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.0001239776611328125, "grad_norm": 7.673588275909424, "learning_rate": 1e-05, "loss": 0.5565, "mean_token_accuracy": 0.8608361482620239, "num_tokens": 736610423.0, "step": 19395 }, { "epoch": 2.467370563541534, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.544448375701904, "learning_rate": 1e-05, "loss": 0.5141, "mean_token_accuracy": 0.8689873814582825, "num_tokens": 736648432.0, "step": 19396 }, { "epoch": 2.4674977738201247, "ewc_loss": 0.1455078125, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.00011730194091796875, "grad_norm": 7.392965793609619, "learning_rate": 1e-05, "loss": 0.5504, "mean_token_accuracy": 0.8600631356239319, "num_tokens": 736688032.0, "step": 19397 }, { "epoch": 2.4676249840987152, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 6.646913528442383, "learning_rate": 1e-05, "loss": 0.4853, "mean_token_accuracy": 0.8760359287261963, "num_tokens": 736725785.0, "step": 19398 }, { "epoch": 2.4677521943773058, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.00010967254638671875, "grad_norm": 7.026521682739258, "learning_rate": 1e-05, "loss": 0.5523, "mean_token_accuracy": 0.8570824265480042, "num_tokens": 736760229.0, "step": 19399 }, { "epoch": 2.4678794046558963, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.594804763793945, "learning_rate": 1e-05, "loss": 0.4744, "mean_token_accuracy": 0.8782727718353271, "num_tokens": 736794981.0, "step": 19400 }, { "epoch": 2.468006614934487, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.731553554534912, "learning_rate": 1e-05, "loss": 0.4818, "mean_token_accuracy": 0.8765468597412109, "num_tokens": 736835387.0, "step": 19401 }, { "epoch": 2.4681338252130773, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.471263408660889, "learning_rate": 1e-05, "loss": 0.4217, "mean_token_accuracy": 0.8965082764625549, "num_tokens": 736877047.0, "step": 19402 }, { "epoch": 2.468261035491668, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.565802574157715, "learning_rate": 1e-05, "loss": 0.4959, "mean_token_accuracy": 0.8702690601348877, "num_tokens": 736916637.0, "step": 19403 }, { "epoch": 2.4683882457702584, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.344816207885742, "learning_rate": 1e-05, "loss": 0.4349, "mean_token_accuracy": 0.8855776786804199, "num_tokens": 736954905.0, "step": 19404 }, { "epoch": 2.468515456048849, "ewc_loss": 0.125, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.536144733428955, "learning_rate": 1e-05, "loss": 0.452, "mean_token_accuracy": 0.8877172470092773, "num_tokens": 736992264.0, "step": 19405 }, { "epoch": 2.4686426663274394, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.202778339385986, "learning_rate": 1e-05, "loss": 0.5196, "mean_token_accuracy": 0.8617780208587646, "num_tokens": 737032896.0, "step": 19406 }, { "epoch": 2.4687698766060295, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.589064598083496, "learning_rate": 1e-05, "loss": 0.4499, "mean_token_accuracy": 0.8843474388122559, "num_tokens": 737073734.0, "step": 19407 }, { "epoch": 2.4688970868846205, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 6.042330741882324, "learning_rate": 1e-05, "loss": 0.4522, "mean_token_accuracy": 0.8817681074142456, "num_tokens": 737112246.0, "step": 19408 }, { "epoch": 2.4690242971632106, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.775763511657715, "learning_rate": 1e-05, "loss": 0.5004, "mean_token_accuracy": 0.8708928227424622, "num_tokens": 737148378.0, "step": 19409 }, { "epoch": 2.469151507441801, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.884927272796631, "learning_rate": 1e-05, "loss": 0.5012, "mean_token_accuracy": 0.8647425770759583, "num_tokens": 737182815.0, "step": 19410 }, { "epoch": 2.4692787177203916, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 7.061402320861816, "learning_rate": 1e-05, "loss": 0.4822, "mean_token_accuracy": 0.882017195224762, "num_tokens": 737223301.0, "step": 19411 }, { "epoch": 2.469405927998982, "ewc_loss": 0.11279296875, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.795170307159424, "learning_rate": 1e-05, "loss": 0.4832, "mean_token_accuracy": 0.8701207637786865, "num_tokens": 737254184.0, "step": 19412 }, { "epoch": 2.4695331382775727, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.000118255615234375, "grad_norm": 7.4483962059021, "learning_rate": 1e-05, "loss": 0.4358, "mean_token_accuracy": 0.8958603143692017, "num_tokens": 737292540.0, "step": 19413 }, { "epoch": 2.4696603485561632, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 5.913699626922607, "learning_rate": 1e-05, "loss": 0.5218, "mean_token_accuracy": 0.86171555519104, "num_tokens": 737329822.0, "step": 19414 }, { "epoch": 2.4697875588347538, "ewc_loss": 0.1572265625, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.00012874603271484375, "grad_norm": 7.758651256561279, "learning_rate": 1e-05, "loss": 0.4888, "mean_token_accuracy": 0.8846787214279175, "num_tokens": 737365416.0, "step": 19415 }, { "epoch": 2.4699147691133443, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.34426212310791, "learning_rate": 1e-05, "loss": 0.4805, "mean_token_accuracy": 0.8797130584716797, "num_tokens": 737411498.0, "step": 19416 }, { "epoch": 2.470041979391935, "ewc_loss": 0.15234375, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.0001239776611328125, "grad_norm": 7.635234355926514, "learning_rate": 1e-05, "loss": 0.4861, "mean_token_accuracy": 0.8843191862106323, "num_tokens": 737453667.0, "step": 19417 }, { "epoch": 2.4701691896705253, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.00011110305786132812, "grad_norm": 6.706919193267822, "learning_rate": 1e-05, "loss": 0.4625, "mean_token_accuracy": 0.8834720849990845, "num_tokens": 737487095.0, "step": 19418 }, { "epoch": 2.470296399949116, "ewc_loss": 0.140625, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.0001125335693359375, "grad_norm": 7.162802696228027, "learning_rate": 1e-05, "loss": 0.4595, "mean_token_accuracy": 0.8866028785705566, "num_tokens": 737521513.0, "step": 19419 }, { "epoch": 2.4704236102277064, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 6.755667209625244, "learning_rate": 1e-05, "loss": 0.4783, "mean_token_accuracy": 0.8856872320175171, "num_tokens": 737559218.0, "step": 19420 }, { "epoch": 2.470550820506297, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.651663780212402, "learning_rate": 1e-05, "loss": 0.5227, "mean_token_accuracy": 0.8660946488380432, "num_tokens": 737600486.0, "step": 19421 }, { "epoch": 2.4706780307848875, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.790826320648193, "learning_rate": 1e-05, "loss": 0.5212, "mean_token_accuracy": 0.8655343055725098, "num_tokens": 737639035.0, "step": 19422 }, { "epoch": 2.470805241063478, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.232065200805664, "learning_rate": 1e-05, "loss": 0.4085, "mean_token_accuracy": 0.8960394859313965, "num_tokens": 737676582.0, "step": 19423 }, { "epoch": 2.4709324513420685, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.927707195281982, "learning_rate": 1e-05, "loss": 0.4858, "mean_token_accuracy": 0.8795298933982849, "num_tokens": 737714809.0, "step": 19424 }, { "epoch": 2.471059661620659, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 5.975481033325195, "learning_rate": 1e-05, "loss": 0.4713, "mean_token_accuracy": 0.8735344409942627, "num_tokens": 737757803.0, "step": 19425 }, { "epoch": 2.4711868718992496, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.00011157989501953125, "grad_norm": 7.144253730773926, "learning_rate": 1e-05, "loss": 0.4884, "mean_token_accuracy": 0.875654935836792, "num_tokens": 737796563.0, "step": 19426 }, { "epoch": 2.47131408217784, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 5.915302276611328, "learning_rate": 1e-05, "loss": 0.4977, "mean_token_accuracy": 0.8716439008712769, "num_tokens": 737829711.0, "step": 19427 }, { "epoch": 2.4714412924564306, "ewc_loss": 0.1455078125, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.00011777877807617188, "grad_norm": 7.409470081329346, "learning_rate": 1e-05, "loss": 0.4872, "mean_token_accuracy": 0.8803907632827759, "num_tokens": 737862632.0, "step": 19428 }, { "epoch": 2.471568502735021, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.042911529541016, "learning_rate": 1e-05, "loss": 0.4801, "mean_token_accuracy": 0.8777315616607666, "num_tokens": 737900188.0, "step": 19429 }, { "epoch": 2.4716957130136117, "ewc_loss": 0.1494140625, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.00012111663818359375, "grad_norm": 7.546370029449463, "learning_rate": 1e-05, "loss": 0.4777, "mean_token_accuracy": 0.8845713138580322, "num_tokens": 737934667.0, "step": 19430 }, { "epoch": 2.471822923292202, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.285318374633789, "learning_rate": 1e-05, "loss": 0.4445, "mean_token_accuracy": 0.8847191333770752, "num_tokens": 737968386.0, "step": 19431 }, { "epoch": 2.4719501335707923, "ewc_loss": 0.1484375, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.0001201629638671875, "grad_norm": 7.455112457275391, "learning_rate": 1e-05, "loss": 0.5049, "mean_token_accuracy": 0.8758547306060791, "num_tokens": 738011761.0, "step": 19432 }, { "epoch": 2.4720773438493833, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.458724021911621, "learning_rate": 1e-05, "loss": 0.4726, "mean_token_accuracy": 0.8829125761985779, "num_tokens": 738049319.0, "step": 19433 }, { "epoch": 2.4722045541279734, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.00011491775512695312, "grad_norm": 7.216806888580322, "learning_rate": 1e-05, "loss": 0.4968, "mean_token_accuracy": 0.8728660345077515, "num_tokens": 738083725.0, "step": 19434 }, { "epoch": 2.472331764406564, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.492111682891846, "learning_rate": 1e-05, "loss": 0.5067, "mean_token_accuracy": 0.8712937235832214, "num_tokens": 738126330.0, "step": 19435 }, { "epoch": 2.4724589746851544, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.934789657592773, "learning_rate": 1e-05, "loss": 0.501, "mean_token_accuracy": 0.874535083770752, "num_tokens": 738163564.0, "step": 19436 }, { "epoch": 2.472586184963745, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.456237316131592, "learning_rate": 1e-05, "loss": 0.5026, "mean_token_accuracy": 0.8708159923553467, "num_tokens": 738197014.0, "step": 19437 }, { "epoch": 2.4727133952423355, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.721712112426758, "learning_rate": 1e-05, "loss": 0.4787, "mean_token_accuracy": 0.8786695003509521, "num_tokens": 738232263.0, "step": 19438 }, { "epoch": 2.472840605520926, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.367257118225098, "learning_rate": 1e-05, "loss": 0.4477, "mean_token_accuracy": 0.8877437114715576, "num_tokens": 738273839.0, "step": 19439 }, { "epoch": 2.4729678157995165, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.607118606567383, "learning_rate": 1e-05, "loss": 0.432, "mean_token_accuracy": 0.8912911415100098, "num_tokens": 738314062.0, "step": 19440 }, { "epoch": 2.473095026078107, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.270310401916504, "learning_rate": 1e-05, "loss": 0.497, "mean_token_accuracy": 0.8719638586044312, "num_tokens": 738352330.0, "step": 19441 }, { "epoch": 2.4732222363566976, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.602107048034668, "learning_rate": 1e-05, "loss": 0.5341, "mean_token_accuracy": 0.8569254875183105, "num_tokens": 738388527.0, "step": 19442 }, { "epoch": 2.473349446635288, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.147724151611328, "learning_rate": 1e-05, "loss": 0.4889, "mean_token_accuracy": 0.8717302680015564, "num_tokens": 738422913.0, "step": 19443 }, { "epoch": 2.4734766569138786, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.680220127105713, "learning_rate": 1e-05, "loss": 0.4968, "mean_token_accuracy": 0.8715248107910156, "num_tokens": 738463739.0, "step": 19444 }, { "epoch": 2.473603867192469, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 6.005979537963867, "learning_rate": 1e-05, "loss": 0.4871, "mean_token_accuracy": 0.8668457269668579, "num_tokens": 738499160.0, "step": 19445 }, { "epoch": 2.4737310774710597, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.845348358154297, "learning_rate": 1e-05, "loss": 0.4555, "mean_token_accuracy": 0.8869026899337769, "num_tokens": 738546634.0, "step": 19446 }, { "epoch": 2.47385828774965, "ewc_loss": 0.11474609375, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.8887553215026855, "learning_rate": 1e-05, "loss": 0.4426, "mean_token_accuracy": 0.8838694095611572, "num_tokens": 738581021.0, "step": 19447 }, { "epoch": 2.4739854980282407, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.00011157989501953125, "grad_norm": 7.136135101318359, "learning_rate": 1e-05, "loss": 0.4818, "mean_token_accuracy": 0.8784164190292358, "num_tokens": 738622606.0, "step": 19448 }, { "epoch": 2.4741127083068313, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 5.856137275695801, "learning_rate": 1e-05, "loss": 0.4628, "mean_token_accuracy": 0.8759013414382935, "num_tokens": 738656244.0, "step": 19449 }, { "epoch": 2.474239918585422, "ewc_loss": 0.1484375, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.0001201629638671875, "grad_norm": 7.456369876861572, "learning_rate": 1e-05, "loss": 0.5273, "mean_token_accuracy": 0.8674293756484985, "num_tokens": 738694743.0, "step": 19450 }, { "epoch": 2.4743671288640123, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.04729700088501, "learning_rate": 1e-05, "loss": 0.4938, "mean_token_accuracy": 0.8715072870254517, "num_tokens": 738731447.0, "step": 19451 }, { "epoch": 2.474494339142603, "ewc_loss": 0.1513671875, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.58408260345459, "learning_rate": 1e-05, "loss": 0.4749, "mean_token_accuracy": 0.8885363936424255, "num_tokens": 738764631.0, "step": 19452 }, { "epoch": 2.4746215494211934, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.390804767608643, "learning_rate": 1e-05, "loss": 0.4617, "mean_token_accuracy": 0.885908842086792, "num_tokens": 738807685.0, "step": 19453 }, { "epoch": 2.474748759699784, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.00011682510375976562, "grad_norm": 7.33528470993042, "learning_rate": 1e-05, "loss": 0.5089, "mean_token_accuracy": 0.8743994235992432, "num_tokens": 738851717.0, "step": 19454 }, { "epoch": 2.4748759699783744, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.610507965087891, "learning_rate": 1e-05, "loss": 0.5405, "mean_token_accuracy": 0.8618050217628479, "num_tokens": 738889432.0, "step": 19455 }, { "epoch": 2.475003180256965, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.866469383239746, "learning_rate": 1e-05, "loss": 0.5004, "mean_token_accuracy": 0.8698240518569946, "num_tokens": 738923197.0, "step": 19456 }, { "epoch": 2.475130390535555, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.7002716064453125, "learning_rate": 1e-05, "loss": 0.4738, "mean_token_accuracy": 0.8813211917877197, "num_tokens": 738959699.0, "step": 19457 }, { "epoch": 2.475257600814146, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.364616870880127, "learning_rate": 1e-05, "loss": 0.434, "mean_token_accuracy": 0.8932331800460815, "num_tokens": 738990444.0, "step": 19458 }, { "epoch": 2.475384811092736, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.842197418212891, "learning_rate": 1e-05, "loss": 0.4694, "mean_token_accuracy": 0.8875949382781982, "num_tokens": 739030864.0, "step": 19459 }, { "epoch": 2.4755120213713266, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 5.978792667388916, "learning_rate": 1e-05, "loss": 0.4892, "mean_token_accuracy": 0.8709956407546997, "num_tokens": 739068407.0, "step": 19460 }, { "epoch": 2.475639231649917, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.00011205673217773438, "grad_norm": 7.135073184967041, "learning_rate": 1e-05, "loss": 0.4921, "mean_token_accuracy": 0.8803337812423706, "num_tokens": 739105560.0, "step": 19461 }, { "epoch": 2.4757664419285077, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.83177375793457, "learning_rate": 1e-05, "loss": 0.4949, "mean_token_accuracy": 0.8668491244316101, "num_tokens": 739142056.0, "step": 19462 }, { "epoch": 2.4758936522070982, "ewc_loss": 0.1484375, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.0001201629638671875, "grad_norm": 7.497206211090088, "learning_rate": 1e-05, "loss": 0.5629, "mean_token_accuracy": 0.8581103086471558, "num_tokens": 739185407.0, "step": 19463 }, { "epoch": 2.4760208624856888, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 5.97308874130249, "learning_rate": 1e-05, "loss": 0.4841, "mean_token_accuracy": 0.872896134853363, "num_tokens": 739222594.0, "step": 19464 }, { "epoch": 2.4761480727642793, "ewc_loss": 0.154296875, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.000125885009765625, "grad_norm": 7.737569332122803, "learning_rate": 1e-05, "loss": 0.4999, "mean_token_accuracy": 0.8808367252349854, "num_tokens": 739261021.0, "step": 19465 }, { "epoch": 2.47627528304287, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.371125221252441, "learning_rate": 1e-05, "loss": 0.5076, "mean_token_accuracy": 0.8715188503265381, "num_tokens": 739300828.0, "step": 19466 }, { "epoch": 2.4764024933214603, "ewc_loss": 0.150390625, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.0001220703125, "grad_norm": 7.598071098327637, "learning_rate": 1e-05, "loss": 0.4733, "mean_token_accuracy": 0.8860708475112915, "num_tokens": 739338724.0, "step": 19467 }, { "epoch": 2.476529703600051, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 6.611199378967285, "learning_rate": 1e-05, "loss": 0.4603, "mean_token_accuracy": 0.8853347897529602, "num_tokens": 739377821.0, "step": 19468 }, { "epoch": 2.4766569138786414, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.00011682510375976562, "grad_norm": 7.265680313110352, "learning_rate": 1e-05, "loss": 0.4554, "mean_token_accuracy": 0.8887214660644531, "num_tokens": 739411438.0, "step": 19469 }, { "epoch": 2.476784124157232, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.615450382232666, "learning_rate": 1e-05, "loss": 0.5149, "mean_token_accuracy": 0.866856038570404, "num_tokens": 739457836.0, "step": 19470 }, { "epoch": 2.4769113344358225, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.941335201263428, "learning_rate": 1e-05, "loss": 0.5122, "mean_token_accuracy": 0.8674865961074829, "num_tokens": 739496336.0, "step": 19471 }, { "epoch": 2.477038544714413, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.524864673614502, "learning_rate": 1e-05, "loss": 0.4816, "mean_token_accuracy": 0.8782496452331543, "num_tokens": 739539777.0, "step": 19472 }, { "epoch": 2.4771657549930035, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.722137928009033, "learning_rate": 1e-05, "loss": 0.4378, "mean_token_accuracy": 0.8931275010108948, "num_tokens": 739572560.0, "step": 19473 }, { "epoch": 2.477292965271594, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.391003131866455, "learning_rate": 1e-05, "loss": 0.4531, "mean_token_accuracy": 0.8826141953468323, "num_tokens": 739604260.0, "step": 19474 }, { "epoch": 2.4774201755501846, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.63289213180542, "learning_rate": 1e-05, "loss": 0.4827, "mean_token_accuracy": 0.8780888915061951, "num_tokens": 739645241.0, "step": 19475 }, { "epoch": 2.477547385828775, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.245315074920654, "learning_rate": 1e-05, "loss": 0.4676, "mean_token_accuracy": 0.8804608583450317, "num_tokens": 739683516.0, "step": 19476 }, { "epoch": 2.4776745961073656, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.664741516113281, "learning_rate": 1e-05, "loss": 0.4912, "mean_token_accuracy": 0.8755565285682678, "num_tokens": 739720422.0, "step": 19477 }, { "epoch": 2.477801806385956, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.086514949798584, "learning_rate": 1e-05, "loss": 0.4233, "mean_token_accuracy": 0.892151951789856, "num_tokens": 739758071.0, "step": 19478 }, { "epoch": 2.4779290166645467, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.807974815368652, "learning_rate": 1e-05, "loss": 0.4739, "mean_token_accuracy": 0.8817373514175415, "num_tokens": 739796143.0, "step": 19479 }, { "epoch": 2.4780562269431368, "ewc_loss": 0.11474609375, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.941144943237305, "learning_rate": 1e-05, "loss": 0.4934, "mean_token_accuracy": 0.8722677230834961, "num_tokens": 739832038.0, "step": 19480 }, { "epoch": 2.4781834372217277, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 7.071102619171143, "learning_rate": 1e-05, "loss": 0.4795, "mean_token_accuracy": 0.8801871538162231, "num_tokens": 739873232.0, "step": 19481 }, { "epoch": 2.478310647500318, "ewc_loss": 0.11474609375, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.864414215087891, "learning_rate": 1e-05, "loss": 0.4394, "mean_token_accuracy": 0.8850011825561523, "num_tokens": 739912213.0, "step": 19482 }, { "epoch": 2.4784378577789083, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.00011682510375976562, "grad_norm": 7.402914524078369, "learning_rate": 1e-05, "loss": 0.5025, "mean_token_accuracy": 0.8735322952270508, "num_tokens": 739949363.0, "step": 19483 }, { "epoch": 2.478565068057499, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.005871772766113, "learning_rate": 1e-05, "loss": 0.4608, "mean_token_accuracy": 0.8818296194076538, "num_tokens": 739979360.0, "step": 19484 }, { "epoch": 2.4786922783360894, "ewc_loss": 0.1513671875, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.601327896118164, "learning_rate": 1e-05, "loss": 0.5526, "mean_token_accuracy": 0.8605774641036987, "num_tokens": 740019691.0, "step": 19485 }, { "epoch": 2.47881948861468, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.34856653213501, "learning_rate": 1e-05, "loss": 0.4609, "mean_token_accuracy": 0.8836472034454346, "num_tokens": 740053604.0, "step": 19486 }, { "epoch": 2.4789466988932705, "ewc_loss": 0.1455078125, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.00011777877807617188, "grad_norm": 7.421419143676758, "learning_rate": 1e-05, "loss": 0.5263, "mean_token_accuracy": 0.8724654316902161, "num_tokens": 740090220.0, "step": 19487 }, { "epoch": 2.479073909171861, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.598100185394287, "learning_rate": 1e-05, "loss": 0.4494, "mean_token_accuracy": 0.8894413113594055, "num_tokens": 740126361.0, "step": 19488 }, { "epoch": 2.4792011194504515, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.964502334594727, "learning_rate": 1e-05, "loss": 0.4871, "mean_token_accuracy": 0.8768056631088257, "num_tokens": 740164752.0, "step": 19489 }, { "epoch": 2.479328329729042, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.692376136779785, "learning_rate": 1e-05, "loss": 0.467, "mean_token_accuracy": 0.8803755044937134, "num_tokens": 740199622.0, "step": 19490 }, { "epoch": 2.4794555400076326, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.45334529876709, "learning_rate": 1e-05, "loss": 0.4766, "mean_token_accuracy": 0.8775384426116943, "num_tokens": 740236457.0, "step": 19491 }, { "epoch": 2.479582750286223, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.8087687492370605, "learning_rate": 1e-05, "loss": 0.5086, "mean_token_accuracy": 0.8697959184646606, "num_tokens": 740275094.0, "step": 19492 }, { "epoch": 2.4797099605648136, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 6.036707878112793, "learning_rate": 1e-05, "loss": 0.4756, "mean_token_accuracy": 0.8748393654823303, "num_tokens": 740308564.0, "step": 19493 }, { "epoch": 2.479837170843404, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 7.074983596801758, "learning_rate": 1e-05, "loss": 0.4841, "mean_token_accuracy": 0.8777995109558105, "num_tokens": 740354994.0, "step": 19494 }, { "epoch": 2.4799643811219947, "ewc_loss": 0.11181640625, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.808915138244629, "learning_rate": 1e-05, "loss": 0.445, "mean_token_accuracy": 0.8878524303436279, "num_tokens": 740394924.0, "step": 19495 }, { "epoch": 2.480091591400585, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.00011873245239257812, "grad_norm": 7.482873439788818, "learning_rate": 1e-05, "loss": 0.4647, "mean_token_accuracy": 0.8892140984535217, "num_tokens": 740428362.0, "step": 19496 }, { "epoch": 2.4802188016791757, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 5.90863561630249, "learning_rate": 1e-05, "loss": 0.5008, "mean_token_accuracy": 0.8687275052070618, "num_tokens": 740473987.0, "step": 19497 }, { "epoch": 2.4803460119577663, "ewc_loss": 0.1572265625, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.00012874603271484375, "grad_norm": 7.823142051696777, "learning_rate": 1e-05, "loss": 0.5532, "mean_token_accuracy": 0.8619459271430969, "num_tokens": 740512458.0, "step": 19498 }, { "epoch": 2.480473222236357, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.355737209320068, "learning_rate": 1e-05, "loss": 0.5459, "mean_token_accuracy": 0.8595215082168579, "num_tokens": 740558744.0, "step": 19499 }, { "epoch": 2.4806004325149473, "ewc_loss": 0.15625, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.0001277923583984375, "grad_norm": 7.731675148010254, "learning_rate": 1e-05, "loss": 0.5412, "mean_token_accuracy": 0.8658515810966492, "num_tokens": 740599889.0, "step": 19500 }, { "epoch": 2.480727642793538, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 6.691092014312744, "learning_rate": 1e-05, "loss": 0.4844, "mean_token_accuracy": 0.8788883090019226, "num_tokens": 740636162.0, "step": 19501 }, { "epoch": 2.4808548530721284, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.00011539459228515625, "grad_norm": 7.338304042816162, "learning_rate": 1e-05, "loss": 0.5482, "mean_token_accuracy": 0.8601793050765991, "num_tokens": 740679595.0, "step": 19502 }, { "epoch": 2.480982063350719, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 6.710235595703125, "learning_rate": 1e-05, "loss": 0.4622, "mean_token_accuracy": 0.8845826983451843, "num_tokens": 740714830.0, "step": 19503 }, { "epoch": 2.4811092736293094, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 6.969653129577637, "learning_rate": 1e-05, "loss": 0.467, "mean_token_accuracy": 0.8813209533691406, "num_tokens": 740749748.0, "step": 19504 }, { "epoch": 2.4812364839078995, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.59785270690918, "learning_rate": 1e-05, "loss": 0.5193, "mean_token_accuracy": 0.8642436265945435, "num_tokens": 740785110.0, "step": 19505 }, { "epoch": 2.4813636941864905, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 7.2818379402160645, "learning_rate": 1e-05, "loss": 0.4755, "mean_token_accuracy": 0.8805819749832153, "num_tokens": 740824432.0, "step": 19506 }, { "epoch": 2.4814909044650806, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.364333152770996, "learning_rate": 1e-05, "loss": 0.4887, "mean_token_accuracy": 0.8763116002082825, "num_tokens": 740861403.0, "step": 19507 }, { "epoch": 2.481618114743671, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.883513927459717, "learning_rate": 1e-05, "loss": 0.4907, "mean_token_accuracy": 0.8743624091148376, "num_tokens": 740899299.0, "step": 19508 }, { "epoch": 2.4817453250222616, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.188966751098633, "learning_rate": 1e-05, "loss": 0.456, "mean_token_accuracy": 0.8830331563949585, "num_tokens": 740934178.0, "step": 19509 }, { "epoch": 2.481872535300852, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 6.958658695220947, "learning_rate": 1e-05, "loss": 0.5003, "mean_token_accuracy": 0.8730531930923462, "num_tokens": 740967010.0, "step": 19510 }, { "epoch": 2.4819997455794427, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.129575729370117, "learning_rate": 1e-05, "loss": 0.4781, "mean_token_accuracy": 0.8737293481826782, "num_tokens": 741004655.0, "step": 19511 }, { "epoch": 2.4821269558580332, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 0.00011157989501953125, "grad_norm": 7.038822650909424, "learning_rate": 1e-05, "loss": 0.4583, "mean_token_accuracy": 0.8881707191467285, "num_tokens": 741037365.0, "step": 19512 }, { "epoch": 2.4822541661366238, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.122012138366699, "learning_rate": 1e-05, "loss": 0.4846, "mean_token_accuracy": 0.873056948184967, "num_tokens": 741072304.0, "step": 19513 }, { "epoch": 2.4823813764152143, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 0.00011348724365234375, "grad_norm": 7.123803615570068, "learning_rate": 1e-05, "loss": 0.4762, "mean_token_accuracy": 0.8807388544082642, "num_tokens": 741107187.0, "step": 19514 }, { "epoch": 2.482508586693805, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.1676530838012695, "learning_rate": 1e-05, "loss": 0.4804, "mean_token_accuracy": 0.8745310306549072, "num_tokens": 741142152.0, "step": 19515 }, { "epoch": 2.4826357969723953, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 0.00011444091796875, "grad_norm": 7.126852035522461, "learning_rate": 1e-05, "loss": 0.4834, "mean_token_accuracy": 0.8836818933486938, "num_tokens": 741175704.0, "step": 19516 }, { "epoch": 2.482763007250986, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.266933441162109, "learning_rate": 1e-05, "loss": 0.4479, "mean_token_accuracy": 0.8873468637466431, "num_tokens": 741212187.0, "step": 19517 }, { "epoch": 2.4828902175295764, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 0.00011110305786132812, "grad_norm": 7.00216817855835, "learning_rate": 1e-05, "loss": 0.4877, "mean_token_accuracy": 0.878750205039978, "num_tokens": 741246830.0, "step": 19518 }, { "epoch": 2.483017427808167, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.360560894012451, "learning_rate": 1e-05, "loss": 0.487, "mean_token_accuracy": 0.8752341866493225, "num_tokens": 741279672.0, "step": 19519 }, { "epoch": 2.4831446380867574, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.751524925231934, "learning_rate": 1e-05, "loss": 0.4546, "mean_token_accuracy": 0.8875126838684082, "num_tokens": 741318205.0, "step": 19520 }, { "epoch": 2.483271848365348, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.445353984832764, "learning_rate": 1e-05, "loss": 0.4587, "mean_token_accuracy": 0.8835816383361816, "num_tokens": 741356312.0, "step": 19521 }, { "epoch": 2.4833990586439385, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.433235168457031, "learning_rate": 1e-05, "loss": 0.5276, "mean_token_accuracy": 0.8676000833511353, "num_tokens": 741393370.0, "step": 19522 }, { "epoch": 2.483526268922529, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.555634498596191, "learning_rate": 1e-05, "loss": 0.4974, "mean_token_accuracy": 0.866529643535614, "num_tokens": 741428533.0, "step": 19523 }, { "epoch": 2.4836534792011196, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.107203483581543, "learning_rate": 1e-05, "loss": 0.4975, "mean_token_accuracy": 0.869128942489624, "num_tokens": 741464614.0, "step": 19524 }, { "epoch": 2.48378068947971, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.706062316894531, "learning_rate": 1e-05, "loss": 0.4869, "mean_token_accuracy": 0.874211847782135, "num_tokens": 741499126.0, "step": 19525 }, { "epoch": 2.4839078997583006, "ewc_loss": 0.11474609375, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.838627338409424, "learning_rate": 1e-05, "loss": 0.4501, "mean_token_accuracy": 0.8812264204025269, "num_tokens": 741544381.0, "step": 19526 }, { "epoch": 2.484035110036891, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 0.00011014938354492188, "grad_norm": 6.996315002441406, "learning_rate": 1e-05, "loss": 0.448, "mean_token_accuracy": 0.8929076194763184, "num_tokens": 741588627.0, "step": 19527 }, { "epoch": 2.4841623203154817, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.65571928024292, "learning_rate": 1e-05, "loss": 0.4526, "mean_token_accuracy": 0.8823350667953491, "num_tokens": 741631165.0, "step": 19528 }, { "epoch": 2.484289530594072, "ewc_loss": 0.150390625, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 0.00012159347534179688, "grad_norm": 7.498099327087402, "learning_rate": 1e-05, "loss": 0.502, "mean_token_accuracy": 0.8778085708618164, "num_tokens": 741668667.0, "step": 19529 }, { "epoch": 2.4844167408726623, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 5.729074001312256, "learning_rate": 1e-05, "loss": 0.467, "mean_token_accuracy": 0.8785427212715149, "num_tokens": 741703625.0, "step": 19530 }, { "epoch": 2.4845439511512533, "ewc_loss": 0.1650390625, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 0.00013637542724609375, "grad_norm": 8.043920516967773, "learning_rate": 1e-05, "loss": 0.525, "mean_token_accuracy": 0.8740466237068176, "num_tokens": 741744371.0, "step": 19531 }, { "epoch": 2.4846711614298433, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 6.314792633056641, "learning_rate": 1e-05, "loss": 0.5159, "mean_token_accuracy": 0.8690651655197144, "num_tokens": 741786733.0, "step": 19532 }, { "epoch": 2.484798371708434, "ewc_loss": 0.1640625, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 0.0001354217529296875, "grad_norm": 8.06716537475586, "learning_rate": 1e-05, "loss": 0.5224, "mean_token_accuracy": 0.8775471448898315, "num_tokens": 741825693.0, "step": 19533 }, { "epoch": 2.4849255819870244, "ewc_loss": 0.1513671875, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 6.973992824554443, "learning_rate": 1e-05, "loss": 0.5648, "mean_token_accuracy": 0.8587085008621216, "num_tokens": 741866176.0, "step": 19534 }, { "epoch": 2.485052792265615, "ewc_loss": 0.150390625, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 0.0001220703125, "grad_norm": 7.503279685974121, "learning_rate": 1e-05, "loss": 0.5443, "mean_token_accuracy": 0.8672886490821838, "num_tokens": 741902523.0, "step": 19535 }, { "epoch": 2.4851800025442055, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.000118255615234375, "grad_norm": 6.935705184936523, "learning_rate": 1e-05, "loss": 0.5293, "mean_token_accuracy": 0.8671426773071289, "num_tokens": 741939005.0, "step": 19536 }, { "epoch": 2.485307212822796, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 7.0398430824279785, "learning_rate": 1e-05, "loss": 0.543, "mean_token_accuracy": 0.8612119555473328, "num_tokens": 741975012.0, "step": 19537 }, { "epoch": 2.4854344231013865, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00011014938354492188, "grad_norm": 6.777874946594238, "learning_rate": 1e-05, "loss": 0.5014, "mean_token_accuracy": 0.8778262734413147, "num_tokens": 742008351.0, "step": 19538 }, { "epoch": 2.485561633379977, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.668915748596191, "learning_rate": 1e-05, "loss": 0.4774, "mean_token_accuracy": 0.8791154623031616, "num_tokens": 742045477.0, "step": 19539 }, { "epoch": 2.4856888436585676, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.661843299865723, "learning_rate": 1e-05, "loss": 0.5182, "mean_token_accuracy": 0.8649261593818665, "num_tokens": 742080070.0, "step": 19540 }, { "epoch": 2.485816053937158, "ewc_loss": 0.125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.435894012451172, "learning_rate": 1e-05, "loss": 0.4671, "mean_token_accuracy": 0.8801437616348267, "num_tokens": 742116735.0, "step": 19541 }, { "epoch": 2.4859432642157486, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.5914130210876465, "learning_rate": 1e-05, "loss": 0.4899, "mean_token_accuracy": 0.8749442100524902, "num_tokens": 742151492.0, "step": 19542 }, { "epoch": 2.486070474494339, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.272568702697754, "learning_rate": 1e-05, "loss": 0.4409, "mean_token_accuracy": 0.8891732096672058, "num_tokens": 742196913.0, "step": 19543 }, { "epoch": 2.4861976847729297, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.53134822845459, "learning_rate": 1e-05, "loss": 0.4335, "mean_token_accuracy": 0.8915584087371826, "num_tokens": 742241894.0, "step": 19544 }, { "epoch": 2.48632489505152, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.1679158210754395, "learning_rate": 1e-05, "loss": 0.4551, "mean_token_accuracy": 0.8838476538658142, "num_tokens": 742276256.0, "step": 19545 }, { "epoch": 2.4864521053301107, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.52682638168335, "learning_rate": 1e-05, "loss": 0.4344, "mean_token_accuracy": 0.8926968574523926, "num_tokens": 742316834.0, "step": 19546 }, { "epoch": 2.4865793156087013, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.066240310668945, "learning_rate": 1e-05, "loss": 0.4622, "mean_token_accuracy": 0.8790875673294067, "num_tokens": 742359132.0, "step": 19547 }, { "epoch": 2.486706525887292, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.62869119644165, "learning_rate": 1e-05, "loss": 0.523, "mean_token_accuracy": 0.860929548740387, "num_tokens": 742399746.0, "step": 19548 }, { "epoch": 2.4868337361658823, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 5.932348728179932, "learning_rate": 1e-05, "loss": 0.5054, "mean_token_accuracy": 0.8656247854232788, "num_tokens": 742435400.0, "step": 19549 }, { "epoch": 2.486960946444473, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.85848331451416, "learning_rate": 1e-05, "loss": 0.4578, "mean_token_accuracy": 0.8847390413284302, "num_tokens": 742472935.0, "step": 19550 }, { "epoch": 2.4870881567230634, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.7277045249938965, "learning_rate": 1e-05, "loss": 0.4607, "mean_token_accuracy": 0.8806571364402771, "num_tokens": 742516841.0, "step": 19551 }, { "epoch": 2.487215367001654, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.0001163482666015625, "grad_norm": 7.291077136993408, "learning_rate": 1e-05, "loss": 0.5289, "mean_token_accuracy": 0.8710572719573975, "num_tokens": 742551057.0, "step": 19552 }, { "epoch": 2.4873425772802444, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.662589073181152, "learning_rate": 1e-05, "loss": 0.4732, "mean_token_accuracy": 0.8745609521865845, "num_tokens": 742584891.0, "step": 19553 }, { "epoch": 2.487469787558835, "ewc_loss": 0.16015625, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.0001316070556640625, "grad_norm": 7.888826847076416, "learning_rate": 1e-05, "loss": 0.4916, "mean_token_accuracy": 0.8845811486244202, "num_tokens": 742622574.0, "step": 19554 }, { "epoch": 2.487596997837425, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.031374931335449, "learning_rate": 1e-05, "loss": 0.4753, "mean_token_accuracy": 0.8765279054641724, "num_tokens": 742664458.0, "step": 19555 }, { "epoch": 2.487724208116016, "ewc_loss": 0.166015625, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.0001373291015625, "grad_norm": 8.236205101013184, "learning_rate": 1e-05, "loss": 0.5015, "mean_token_accuracy": 0.8801395297050476, "num_tokens": 742704006.0, "step": 19556 }, { "epoch": 2.487851418394606, "ewc_loss": 0.1494140625, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00012063980102539062, "grad_norm": 6.817807197570801, "learning_rate": 1e-05, "loss": 0.5338, "mean_token_accuracy": 0.8669741153717041, "num_tokens": 742736885.0, "step": 19557 }, { "epoch": 2.4879786286731966, "ewc_loss": 0.15625, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.0001277923583984375, "grad_norm": 7.830353736877441, "learning_rate": 1e-05, "loss": 0.5198, "mean_token_accuracy": 0.8733893036842346, "num_tokens": 742773496.0, "step": 19558 }, { "epoch": 2.488105838951787, "ewc_loss": 0.1494140625, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00012063980102539062, "grad_norm": 7.031777381896973, "learning_rate": 1e-05, "loss": 0.4879, "mean_token_accuracy": 0.8828146457672119, "num_tokens": 742812444.0, "step": 19559 }, { "epoch": 2.4882330492303777, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00011587142944335938, "grad_norm": 7.31739616394043, "learning_rate": 1e-05, "loss": 0.5019, "mean_token_accuracy": 0.8751330375671387, "num_tokens": 742849910.0, "step": 19560 }, { "epoch": 2.488360259508968, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00011348724365234375, "grad_norm": 6.861301898956299, "learning_rate": 1e-05, "loss": 0.4623, "mean_token_accuracy": 0.8880422115325928, "num_tokens": 742891412.0, "step": 19561 }, { "epoch": 2.4884874697875587, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.926984786987305, "learning_rate": 1e-05, "loss": 0.4986, "mean_token_accuracy": 0.8729292750358582, "num_tokens": 742929510.0, "step": 19562 }, { "epoch": 2.4886146800661493, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.686814785003662, "learning_rate": 1e-05, "loss": 0.5207, "mean_token_accuracy": 0.8667575120925903, "num_tokens": 742969767.0, "step": 19563 }, { "epoch": 2.48874189034474, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.675541877746582, "learning_rate": 1e-05, "loss": 0.4551, "mean_token_accuracy": 0.8839063048362732, "num_tokens": 743001416.0, "step": 19564 }, { "epoch": 2.4888691006233303, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.52531623840332, "learning_rate": 1e-05, "loss": 0.4894, "mean_token_accuracy": 0.8736828565597534, "num_tokens": 743037273.0, "step": 19565 }, { "epoch": 2.488996310901921, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.533781051635742, "learning_rate": 1e-05, "loss": 0.4926, "mean_token_accuracy": 0.8732792735099792, "num_tokens": 743072774.0, "step": 19566 }, { "epoch": 2.4891235211805114, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.418829441070557, "learning_rate": 1e-05, "loss": 0.5056, "mean_token_accuracy": 0.8658447265625, "num_tokens": 743115931.0, "step": 19567 }, { "epoch": 2.489250731459102, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.438902378082275, "learning_rate": 1e-05, "loss": 0.4509, "mean_token_accuracy": 0.8882675766944885, "num_tokens": 743158500.0, "step": 19568 }, { "epoch": 2.4893779417376924, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.288569450378418, "learning_rate": 1e-05, "loss": 0.4697, "mean_token_accuracy": 0.8791733980178833, "num_tokens": 743195311.0, "step": 19569 }, { "epoch": 2.489505152016283, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.432725429534912, "learning_rate": 1e-05, "loss": 0.4585, "mean_token_accuracy": 0.8814293146133423, "num_tokens": 743227978.0, "step": 19570 }, { "epoch": 2.4896323622948735, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.162929058074951, "learning_rate": 1e-05, "loss": 0.4534, "mean_token_accuracy": 0.8824668526649475, "num_tokens": 743264297.0, "step": 19571 }, { "epoch": 2.489759572573464, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.49821662902832, "learning_rate": 1e-05, "loss": 0.4829, "mean_token_accuracy": 0.8731070756912231, "num_tokens": 743305351.0, "step": 19572 }, { "epoch": 2.4898867828520546, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 6.02818489074707, "learning_rate": 1e-05, "loss": 0.4947, "mean_token_accuracy": 0.8711032271385193, "num_tokens": 743345105.0, "step": 19573 }, { "epoch": 2.490013993130645, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.669614791870117, "learning_rate": 1e-05, "loss": 0.4503, "mean_token_accuracy": 0.8866928815841675, "num_tokens": 743383889.0, "step": 19574 }, { "epoch": 2.4901412034092356, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.8492512702941895, "learning_rate": 1e-05, "loss": 0.4611, "mean_token_accuracy": 0.8772746324539185, "num_tokens": 743420045.0, "step": 19575 }, { "epoch": 2.490268413687826, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.989173889160156, "learning_rate": 1e-05, "loss": 0.4881, "mean_token_accuracy": 0.8761393427848816, "num_tokens": 743455358.0, "step": 19576 }, { "epoch": 2.4903956239664167, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 8.344650268554688e-05, "grad_norm": 5.678096771240234, "learning_rate": 1e-05, "loss": 0.5154, "mean_token_accuracy": 0.8586612939834595, "num_tokens": 743492337.0, "step": 19577 }, { "epoch": 2.4905228342450068, "ewc_loss": 0.1494140625, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00012063980102539062, "grad_norm": 7.531900882720947, "learning_rate": 1e-05, "loss": 0.4744, "mean_token_accuracy": 0.885994017124176, "num_tokens": 743526023.0, "step": 19578 }, { "epoch": 2.4906500445235977, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 5.726779937744141, "learning_rate": 1e-05, "loss": 0.4545, "mean_token_accuracy": 0.8789033889770508, "num_tokens": 743560224.0, "step": 19579 }, { "epoch": 2.490777254802188, "ewc_loss": 0.1650390625, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00013637542724609375, "grad_norm": 8.156417846679688, "learning_rate": 1e-05, "loss": 0.5952, "mean_token_accuracy": 0.8500717878341675, "num_tokens": 743601403.0, "step": 19580 }, { "epoch": 2.4909044650807783, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 6.398530960083008, "learning_rate": 1e-05, "loss": 0.4986, "mean_token_accuracy": 0.8790178298950195, "num_tokens": 743641677.0, "step": 19581 }, { "epoch": 2.491031675359369, "ewc_loss": 0.1650390625, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00013637542724609375, "grad_norm": 8.160367012023926, "learning_rate": 1e-05, "loss": 0.5223, "mean_token_accuracy": 0.8760608434677124, "num_tokens": 743684615.0, "step": 19582 }, { "epoch": 2.4911588856379594, "ewc_loss": 0.15234375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.0001239776611328125, "grad_norm": 7.052896499633789, "learning_rate": 1e-05, "loss": 0.4549, "mean_token_accuracy": 0.8915736079216003, "num_tokens": 743718960.0, "step": 19583 }, { "epoch": 2.49128609591655, "ewc_loss": 0.15234375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.0001239776611328125, "grad_norm": 7.594607830047607, "learning_rate": 1e-05, "loss": 0.4778, "mean_token_accuracy": 0.8870145082473755, "num_tokens": 743758336.0, "step": 19584 }, { "epoch": 2.4914133061951405, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.000118255615234375, "grad_norm": 7.025515556335449, "learning_rate": 1e-05, "loss": 0.524, "mean_token_accuracy": 0.8672347068786621, "num_tokens": 743803243.0, "step": 19585 }, { "epoch": 2.491540516473731, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00011110305786132812, "grad_norm": 7.128419399261475, "learning_rate": 1e-05, "loss": 0.4671, "mean_token_accuracy": 0.8828980922698975, "num_tokens": 743840945.0, "step": 19586 }, { "epoch": 2.4916677267523215, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00010967254638671875, "grad_norm": 6.797668933868408, "learning_rate": 1e-05, "loss": 0.483, "mean_token_accuracy": 0.8803693056106567, "num_tokens": 743884338.0, "step": 19587 }, { "epoch": 2.491794937030912, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.802566051483154, "learning_rate": 1e-05, "loss": 0.4921, "mean_token_accuracy": 0.8764259219169617, "num_tokens": 743925499.0, "step": 19588 }, { "epoch": 2.4919221473095026, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.6244893074035645, "learning_rate": 1e-05, "loss": 0.528, "mean_token_accuracy": 0.863113284111023, "num_tokens": 743965016.0, "step": 19589 }, { "epoch": 2.492049357588093, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.606239318847656, "learning_rate": 1e-05, "loss": 0.487, "mean_token_accuracy": 0.870939314365387, "num_tokens": 744000523.0, "step": 19590 }, { "epoch": 2.4921765678666836, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.457815647125244, "learning_rate": 1e-05, "loss": 0.4947, "mean_token_accuracy": 0.873142421245575, "num_tokens": 744043862.0, "step": 19591 }, { "epoch": 2.492303778145274, "ewc_loss": 0.125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.5121588706970215, "learning_rate": 1e-05, "loss": 0.5331, "mean_token_accuracy": 0.8589766025543213, "num_tokens": 744085161.0, "step": 19592 }, { "epoch": 2.4924309884238647, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.3280181884765625, "learning_rate": 1e-05, "loss": 0.4605, "mean_token_accuracy": 0.8805437684059143, "num_tokens": 744117575.0, "step": 19593 }, { "epoch": 2.492558198702455, "ewc_loss": 0.125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.497357368469238, "learning_rate": 1e-05, "loss": 0.5636, "mean_token_accuracy": 0.8507939577102661, "num_tokens": 744156312.0, "step": 19594 }, { "epoch": 2.4926854089810457, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.187660217285156, "learning_rate": 1e-05, "loss": 0.4608, "mean_token_accuracy": 0.8790494203567505, "num_tokens": 744191177.0, "step": 19595 }, { "epoch": 2.4928126192596363, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.557349681854248, "learning_rate": 1e-05, "loss": 0.4838, "mean_token_accuracy": 0.872296929359436, "num_tokens": 744230407.0, "step": 19596 }, { "epoch": 2.492939829538227, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.016254901885986, "learning_rate": 1e-05, "loss": 0.4825, "mean_token_accuracy": 0.8730790615081787, "num_tokens": 744270265.0, "step": 19597 }, { "epoch": 2.4930670398168173, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.740962028503418, "learning_rate": 1e-05, "loss": 0.4766, "mean_token_accuracy": 0.8778930306434631, "num_tokens": 744310913.0, "step": 19598 }, { "epoch": 2.493194250095408, "ewc_loss": 0.11376953125, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.837765693664551, "learning_rate": 1e-05, "loss": 0.4865, "mean_token_accuracy": 0.871878445148468, "num_tokens": 744342918.0, "step": 19599 }, { "epoch": 2.4933214603739984, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 0.00011014938354492188, "grad_norm": 7.096347808837891, "learning_rate": 1e-05, "loss": 0.4978, "mean_token_accuracy": 0.874602198600769, "num_tokens": 744379000.0, "step": 19600 }, { "epoch": 2.493448670652589, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.680232524871826, "learning_rate": 1e-05, "loss": 0.4493, "mean_token_accuracy": 0.8761879801750183, "num_tokens": 744414641.0, "step": 19601 }, { "epoch": 2.4935758809311794, "ewc_loss": 0.1533203125, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 0.00012493133544921875, "grad_norm": 7.674449443817139, "learning_rate": 1e-05, "loss": 0.5224, "mean_token_accuracy": 0.870663046836853, "num_tokens": 744444860.0, "step": 19602 }, { "epoch": 2.4937030912097695, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 5.860755920410156, "learning_rate": 1e-05, "loss": 0.4514, "mean_token_accuracy": 0.8866515755653381, "num_tokens": 744479657.0, "step": 19603 }, { "epoch": 2.4938303014883605, "ewc_loss": 0.1650390625, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 0.00013637542724609375, "grad_norm": 8.168656349182129, "learning_rate": 1e-05, "loss": 0.5125, "mean_token_accuracy": 0.8772034645080566, "num_tokens": 744520048.0, "step": 19604 }, { "epoch": 2.4939575117669506, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 0.00011301040649414062, "grad_norm": 6.60235595703125, "learning_rate": 1e-05, "loss": 0.5174, "mean_token_accuracy": 0.8707358837127686, "num_tokens": 744555265.0, "step": 19605 }, { "epoch": 2.494084722045541, "ewc_loss": 0.1591796875, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 0.00013065338134765625, "grad_norm": 7.966915607452393, "learning_rate": 1e-05, "loss": 0.512, "mean_token_accuracy": 0.8728331923484802, "num_tokens": 744591011.0, "step": 19606 }, { "epoch": 2.4942119323241316, "ewc_loss": 0.1484375, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 0.0001201629638671875, "grad_norm": 7.0310797691345215, "learning_rate": 1e-05, "loss": 0.5254, "mean_token_accuracy": 0.8712843060493469, "num_tokens": 744632302.0, "step": 19607 }, { "epoch": 2.494339142602722, "ewc_loss": 0.1455078125, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 0.00011682510375976562, "grad_norm": 7.406635284423828, "learning_rate": 1e-05, "loss": 0.4604, "mean_token_accuracy": 0.8862476944923401, "num_tokens": 744669827.0, "step": 19608 }, { "epoch": 2.4944663528813127, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 0.00011444091796875, "grad_norm": 6.953587055206299, "learning_rate": 1e-05, "loss": 0.4667, "mean_token_accuracy": 0.8827923536300659, "num_tokens": 744708438.0, "step": 19609 }, { "epoch": 2.494593563159903, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.948317527770996, "learning_rate": 1e-05, "loss": 0.4874, "mean_token_accuracy": 0.8780617713928223, "num_tokens": 744744218.0, "step": 19610 }, { "epoch": 2.4947207734384937, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.768126487731934, "learning_rate": 1e-05, "loss": 0.4708, "mean_token_accuracy": 0.8823271989822388, "num_tokens": 744782805.0, "step": 19611 }, { "epoch": 2.4948479837170843, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.630964756011963, "learning_rate": 1e-05, "loss": 0.4147, "mean_token_accuracy": 0.8956645727157593, "num_tokens": 744814011.0, "step": 19612 }, { "epoch": 2.494975193995675, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.638160228729248, "learning_rate": 1e-05, "loss": 0.4814, "mean_token_accuracy": 0.8788675665855408, "num_tokens": 744857385.0, "step": 19613 }, { "epoch": 2.4951024042742653, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.4349846839904785, "learning_rate": 1e-05, "loss": 0.5407, "mean_token_accuracy": 0.8585883378982544, "num_tokens": 744898212.0, "step": 19614 }, { "epoch": 2.495229614552856, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.542076587677002, "learning_rate": 1e-05, "loss": 0.4889, "mean_token_accuracy": 0.8741220831871033, "num_tokens": 744930665.0, "step": 19615 }, { "epoch": 2.4953568248314464, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.308823585510254, "learning_rate": 1e-05, "loss": 0.5143, "mean_token_accuracy": 0.8651621341705322, "num_tokens": 744968725.0, "step": 19616 }, { "epoch": 2.495484035110037, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.474045753479004, "learning_rate": 1e-05, "loss": 0.5068, "mean_token_accuracy": 0.865654706954956, "num_tokens": 745008386.0, "step": 19617 }, { "epoch": 2.4956112453886274, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.222896099090576, "learning_rate": 1e-05, "loss": 0.4815, "mean_token_accuracy": 0.8740569353103638, "num_tokens": 745047059.0, "step": 19618 }, { "epoch": 2.495738455667218, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.445962429046631, "learning_rate": 1e-05, "loss": 0.508, "mean_token_accuracy": 0.8672618269920349, "num_tokens": 745089768.0, "step": 19619 }, { "epoch": 2.4958656659458085, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.159363746643066, "learning_rate": 1e-05, "loss": 0.4838, "mean_token_accuracy": 0.8732860684394836, "num_tokens": 745130848.0, "step": 19620 }, { "epoch": 2.495992876224399, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.471538066864014, "learning_rate": 1e-05, "loss": 0.4825, "mean_token_accuracy": 0.8735262155532837, "num_tokens": 745166715.0, "step": 19621 }, { "epoch": 2.4961200865029896, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.049467086791992, "learning_rate": 1e-05, "loss": 0.4922, "mean_token_accuracy": 0.8696761727333069, "num_tokens": 745200718.0, "step": 19622 }, { "epoch": 2.49624729678158, "ewc_loss": 0.125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.5862040519714355, "learning_rate": 1e-05, "loss": 0.4949, "mean_token_accuracy": 0.8680691123008728, "num_tokens": 745240736.0, "step": 19623 }, { "epoch": 2.4963745070601706, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.895169734954834, "learning_rate": 1e-05, "loss": 0.478, "mean_token_accuracy": 0.8752686381340027, "num_tokens": 745279040.0, "step": 19624 }, { "epoch": 2.496501717338761, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.830163955688477, "learning_rate": 1e-05, "loss": 0.5116, "mean_token_accuracy": 0.8703334331512451, "num_tokens": 745316936.0, "step": 19625 }, { "epoch": 2.4966289276173517, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.699856281280518, "learning_rate": 1e-05, "loss": 0.5371, "mean_token_accuracy": 0.8549298048019409, "num_tokens": 745355438.0, "step": 19626 }, { "epoch": 2.496756137895942, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.0001163482666015625, "grad_norm": 7.304130554199219, "learning_rate": 1e-05, "loss": 0.4854, "mean_token_accuracy": 0.8812911510467529, "num_tokens": 745394509.0, "step": 19627 }, { "epoch": 2.4968833481745323, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.59061861038208, "learning_rate": 1e-05, "loss": 0.4793, "mean_token_accuracy": 0.873589813709259, "num_tokens": 745426347.0, "step": 19628 }, { "epoch": 2.4970105584531233, "ewc_loss": 0.1630859375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00013446807861328125, "grad_norm": 8.010383605957031, "learning_rate": 1e-05, "loss": 0.493, "mean_token_accuracy": 0.8848536014556885, "num_tokens": 745457523.0, "step": 19629 }, { "epoch": 2.4971377687317133, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.042665481567383, "learning_rate": 1e-05, "loss": 0.5322, "mean_token_accuracy": 0.8603905439376831, "num_tokens": 745495155.0, "step": 19630 }, { "epoch": 2.497264979010304, "ewc_loss": 0.1708984375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.0001430511474609375, "grad_norm": 8.407033920288086, "learning_rate": 1e-05, "loss": 0.5368, "mean_token_accuracy": 0.8712366819381714, "num_tokens": 745535466.0, "step": 19631 }, { "epoch": 2.4973921892888944, "ewc_loss": 0.1533203125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00012493133544921875, "grad_norm": 6.9658074378967285, "learning_rate": 1e-05, "loss": 0.5079, "mean_token_accuracy": 0.8730883598327637, "num_tokens": 745569184.0, "step": 19632 }, { "epoch": 2.497519399567485, "ewc_loss": 0.16015625, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.0001316070556640625, "grad_norm": 7.932692527770996, "learning_rate": 1e-05, "loss": 0.5355, "mean_token_accuracy": 0.8699513673782349, "num_tokens": 745610735.0, "step": 19633 }, { "epoch": 2.4976466098460754, "ewc_loss": 0.15234375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.0001239776611328125, "grad_norm": 7.174626350402832, "learning_rate": 1e-05, "loss": 0.5411, "mean_token_accuracy": 0.865833580493927, "num_tokens": 745653513.0, "step": 19634 }, { "epoch": 2.497773820124666, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00011587142944335938, "grad_norm": 7.393383979797363, "learning_rate": 1e-05, "loss": 0.5088, "mean_token_accuracy": 0.8727431297302246, "num_tokens": 745690007.0, "step": 19635 }, { "epoch": 2.4979010304032565, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00011539459228515625, "grad_norm": 6.933608055114746, "learning_rate": 1e-05, "loss": 0.5403, "mean_token_accuracy": 0.8668684959411621, "num_tokens": 745724886.0, "step": 19636 }, { "epoch": 2.498028240681847, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00010967254638671875, "grad_norm": 7.040950298309326, "learning_rate": 1e-05, "loss": 0.4747, "mean_token_accuracy": 0.882463812828064, "num_tokens": 745760782.0, "step": 19637 }, { "epoch": 2.4981554509604376, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.682680606842041, "learning_rate": 1e-05, "loss": 0.4404, "mean_token_accuracy": 0.8910900354385376, "num_tokens": 745794535.0, "step": 19638 }, { "epoch": 2.498282661239028, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.829832553863525, "learning_rate": 1e-05, "loss": 0.4986, "mean_token_accuracy": 0.8766306638717651, "num_tokens": 745831010.0, "step": 19639 }, { "epoch": 2.4984098715176186, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.463249683380127, "learning_rate": 1e-05, "loss": 0.5037, "mean_token_accuracy": 0.8691445589065552, "num_tokens": 745866761.0, "step": 19640 }, { "epoch": 2.498537081796209, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.748198509216309, "learning_rate": 1e-05, "loss": 0.4751, "mean_token_accuracy": 0.8787638545036316, "num_tokens": 745909081.0, "step": 19641 }, { "epoch": 2.4986642920747997, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.264342784881592, "learning_rate": 1e-05, "loss": 0.4284, "mean_token_accuracy": 0.892014741897583, "num_tokens": 745942907.0, "step": 19642 }, { "epoch": 2.49879150235339, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.7813310623168945, "learning_rate": 1e-05, "loss": 0.5011, "mean_token_accuracy": 0.8732264637947083, "num_tokens": 745981558.0, "step": 19643 }, { "epoch": 2.4989187126319807, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.070871353149414, "learning_rate": 1e-05, "loss": 0.4707, "mean_token_accuracy": 0.8776293992996216, "num_tokens": 746015568.0, "step": 19644 }, { "epoch": 2.4990459229105713, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.987925052642822, "learning_rate": 1e-05, "loss": 0.4733, "mean_token_accuracy": 0.8832553625106812, "num_tokens": 746049503.0, "step": 19645 }, { "epoch": 2.499173133189162, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.899930477142334, "learning_rate": 1e-05, "loss": 0.467, "mean_token_accuracy": 0.8774398565292358, "num_tokens": 746086979.0, "step": 19646 }, { "epoch": 2.4993003434677523, "ewc_loss": 0.1455078125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00011682510375976562, "grad_norm": 7.354592323303223, "learning_rate": 1e-05, "loss": 0.4972, "mean_token_accuracy": 0.8781595230102539, "num_tokens": 746125073.0, "step": 19647 }, { "epoch": 2.499427553746343, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 5.898611068725586, "learning_rate": 1e-05, "loss": 0.4615, "mean_token_accuracy": 0.8791232109069824, "num_tokens": 746161621.0, "step": 19648 }, { "epoch": 2.4995547640249334, "ewc_loss": 0.154296875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.000125885009765625, "grad_norm": 7.758258819580078, "learning_rate": 1e-05, "loss": 0.5295, "mean_token_accuracy": 0.871060848236084, "num_tokens": 746203032.0, "step": 19649 }, { "epoch": 2.499681974303524, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.27880859375, "learning_rate": 1e-05, "loss": 0.4844, "mean_token_accuracy": 0.8749713897705078, "num_tokens": 746233939.0, "step": 19650 }, { "epoch": 2.499809184582114, "ewc_loss": 0.1552734375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00012683868408203125, "grad_norm": 7.804944038391113, "learning_rate": 1e-05, "loss": 0.4843, "mean_token_accuracy": 0.886574387550354, "num_tokens": 746269175.0, "step": 19651 }, { "epoch": 2.499936394860705, "ewc_loss": 0.140625, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00011157989501953125, "grad_norm": 6.7376532554626465, "learning_rate": 1e-05, "loss": 0.5059, "mean_token_accuracy": 0.871681809425354, "num_tokens": 746310634.0, "step": 19652 }, { "epoch": 2.500063605139295, "ewc_loss": 0.1455078125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00011682510375976562, "grad_norm": 7.342250823974609, "learning_rate": 1e-05, "loss": 0.4838, "mean_token_accuracy": 0.8817311525344849, "num_tokens": 746353683.0, "step": 19653 }, { "epoch": 2.500190815417886, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00011110305786132812, "grad_norm": 6.860782623291016, "learning_rate": 1e-05, "loss": 0.5049, "mean_token_accuracy": 0.8739640712738037, "num_tokens": 746392299.0, "step": 19654 }, { "epoch": 2.500318025696476, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.782372951507568, "learning_rate": 1e-05, "loss": 0.4693, "mean_token_accuracy": 0.8805872201919556, "num_tokens": 746435276.0, "step": 19655 }, { "epoch": 2.5004452359750666, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 6.860255241394043, "learning_rate": 1e-05, "loss": 0.5274, "mean_token_accuracy": 0.8690302968025208, "num_tokens": 746476175.0, "step": 19656 }, { "epoch": 2.500572446253657, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.293632984161377, "learning_rate": 1e-05, "loss": 0.4353, "mean_token_accuracy": 0.8866194486618042, "num_tokens": 746517084.0, "step": 19657 }, { "epoch": 2.5006996565322477, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.981708526611328, "learning_rate": 1e-05, "loss": 0.526, "mean_token_accuracy": 0.861560046672821, "num_tokens": 746556073.0, "step": 19658 }, { "epoch": 2.500826866810838, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.978757381439209, "learning_rate": 1e-05, "loss": 0.4279, "mean_token_accuracy": 0.8896409273147583, "num_tokens": 746590188.0, "step": 19659 }, { "epoch": 2.5009540770894287, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00011301040649414062, "grad_norm": 7.265255451202393, "learning_rate": 1e-05, "loss": 0.5944, "mean_token_accuracy": 0.8490157127380371, "num_tokens": 746634667.0, "step": 19660 }, { "epoch": 2.5010812873680193, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.87261438369751, "learning_rate": 1e-05, "loss": 0.453, "mean_token_accuracy": 0.8823909163475037, "num_tokens": 746677398.0, "step": 19661 }, { "epoch": 2.50120849764661, "ewc_loss": 0.150390625, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.0001220703125, "grad_norm": 7.6392645835876465, "learning_rate": 1e-05, "loss": 0.4969, "mean_token_accuracy": 0.8831638693809509, "num_tokens": 746711607.0, "step": 19662 }, { "epoch": 2.5013357079252003, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.117660045623779, "learning_rate": 1e-05, "loss": 0.4625, "mean_token_accuracy": 0.8841329216957092, "num_tokens": 746751853.0, "step": 19663 }, { "epoch": 2.501462918203791, "ewc_loss": 0.154296875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.000125885009765625, "grad_norm": 7.798803329467773, "learning_rate": 1e-05, "loss": 0.5444, "mean_token_accuracy": 0.8641053438186646, "num_tokens": 746792134.0, "step": 19664 }, { "epoch": 2.5015901284823814, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.546479225158691, "learning_rate": 1e-05, "loss": 0.5152, "mean_token_accuracy": 0.867900013923645, "num_tokens": 746830087.0, "step": 19665 }, { "epoch": 2.501717338760972, "ewc_loss": 0.1484375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.0001201629638671875, "grad_norm": 7.5477070808410645, "learning_rate": 1e-05, "loss": 0.4889, "mean_token_accuracy": 0.8818928599357605, "num_tokens": 746864738.0, "step": 19666 }, { "epoch": 2.5018445490395624, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00011014938354492188, "grad_norm": 6.724787712097168, "learning_rate": 1e-05, "loss": 0.5371, "mean_token_accuracy": 0.8600051403045654, "num_tokens": 746904021.0, "step": 19667 }, { "epoch": 2.501971759318153, "ewc_loss": 0.140625, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.0001125335693359375, "grad_norm": 7.160886287689209, "learning_rate": 1e-05, "loss": 0.4699, "mean_token_accuracy": 0.8821858167648315, "num_tokens": 746942710.0, "step": 19668 }, { "epoch": 2.5020989695967435, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.644963264465332, "learning_rate": 1e-05, "loss": 0.475, "mean_token_accuracy": 0.8773393630981445, "num_tokens": 746986694.0, "step": 19669 }, { "epoch": 2.502226179875334, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.844194412231445, "learning_rate": 1e-05, "loss": 0.4509, "mean_token_accuracy": 0.8890877962112427, "num_tokens": 747028440.0, "step": 19670 }, { "epoch": 2.5023533901539246, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.517961502075195, "learning_rate": 1e-05, "loss": 0.447, "mean_token_accuracy": 0.8865625858306885, "num_tokens": 747063862.0, "step": 19671 }, { "epoch": 2.502480600432515, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.679561614990234, "learning_rate": 1e-05, "loss": 0.4919, "mean_token_accuracy": 0.872727632522583, "num_tokens": 747100469.0, "step": 19672 }, { "epoch": 2.5026078107111056, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.355177402496338, "learning_rate": 1e-05, "loss": 0.4313, "mean_token_accuracy": 0.8892069458961487, "num_tokens": 747133074.0, "step": 19673 }, { "epoch": 2.5027350209896957, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.658344745635986, "learning_rate": 1e-05, "loss": 0.515, "mean_token_accuracy": 0.8638269901275635, "num_tokens": 747167010.0, "step": 19674 }, { "epoch": 2.5028622312682867, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.17117977142334, "learning_rate": 1e-05, "loss": 0.4801, "mean_token_accuracy": 0.8714634776115417, "num_tokens": 747201397.0, "step": 19675 }, { "epoch": 2.5029894415468767, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.7390875816345215, "learning_rate": 1e-05, "loss": 0.4582, "mean_token_accuracy": 0.885029673576355, "num_tokens": 747238430.0, "step": 19676 }, { "epoch": 2.5031166518254677, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 6.009609699249268, "learning_rate": 1e-05, "loss": 0.4842, "mean_token_accuracy": 0.8773738145828247, "num_tokens": 747277760.0, "step": 19677 }, { "epoch": 2.503243862104058, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.940551280975342, "learning_rate": 1e-05, "loss": 0.4872, "mean_token_accuracy": 0.8764652609825134, "num_tokens": 747317763.0, "step": 19678 }, { "epoch": 2.5033710723826488, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.879603385925293, "learning_rate": 1e-05, "loss": 0.4959, "mean_token_accuracy": 0.8672431111335754, "num_tokens": 747362733.0, "step": 19679 }, { "epoch": 2.503498282661239, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00011396408081054688, "grad_norm": 7.266170501708984, "learning_rate": 1e-05, "loss": 0.4635, "mean_token_accuracy": 0.8857597708702087, "num_tokens": 747400735.0, "step": 19680 }, { "epoch": 2.5036254929398294, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 5.90175199508667, "learning_rate": 1e-05, "loss": 0.4824, "mean_token_accuracy": 0.8733431100845337, "num_tokens": 747440263.0, "step": 19681 }, { "epoch": 2.50375270321842, "ewc_loss": 0.150390625, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00012159347534179688, "grad_norm": 7.596750259399414, "learning_rate": 1e-05, "loss": 0.4941, "mean_token_accuracy": 0.8787946701049805, "num_tokens": 747477918.0, "step": 19682 }, { "epoch": 2.5038799134970104, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.193426609039307, "learning_rate": 1e-05, "loss": 0.4753, "mean_token_accuracy": 0.8759763240814209, "num_tokens": 747513235.0, "step": 19683 }, { "epoch": 2.504007123775601, "ewc_loss": 0.1513671875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.601468086242676, "learning_rate": 1e-05, "loss": 0.4657, "mean_token_accuracy": 0.8940017223358154, "num_tokens": 747553280.0, "step": 19684 }, { "epoch": 2.5041343340541915, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.57413911819458, "learning_rate": 1e-05, "loss": 0.5076, "mean_token_accuracy": 0.8709885478019714, "num_tokens": 747594389.0, "step": 19685 }, { "epoch": 2.504261544332782, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00011396408081054688, "grad_norm": 7.193595886230469, "learning_rate": 1e-05, "loss": 0.4745, "mean_token_accuracy": 0.8807310461997986, "num_tokens": 747634805.0, "step": 19686 }, { "epoch": 2.5043887546113726, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 6.731726169586182, "learning_rate": 1e-05, "loss": 0.5318, "mean_token_accuracy": 0.8667943477630615, "num_tokens": 747673963.0, "step": 19687 }, { "epoch": 2.504515964889963, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.655621528625488, "learning_rate": 1e-05, "loss": 0.4979, "mean_token_accuracy": 0.8680514097213745, "num_tokens": 747715067.0, "step": 19688 }, { "epoch": 2.5046431751685536, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.8224687576293945, "learning_rate": 1e-05, "loss": 0.462, "mean_token_accuracy": 0.8824295401573181, "num_tokens": 747752799.0, "step": 19689 }, { "epoch": 2.504770385447144, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 6.171174049377441, "learning_rate": 1e-05, "loss": 0.5085, "mean_token_accuracy": 0.8669306635856628, "num_tokens": 747789325.0, "step": 19690 }, { "epoch": 2.5048975957257347, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 7.0316972732543945, "learning_rate": 1e-05, "loss": 0.4905, "mean_token_accuracy": 0.8758339285850525, "num_tokens": 747823438.0, "step": 19691 }, { "epoch": 2.505024806004325, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.852357864379883, "learning_rate": 1e-05, "loss": 0.4389, "mean_token_accuracy": 0.8843737840652466, "num_tokens": 747862310.0, "step": 19692 }, { "epoch": 2.5051520162829157, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.000118255615234375, "grad_norm": 7.405235767364502, "learning_rate": 1e-05, "loss": 0.4938, "mean_token_accuracy": 0.8780920505523682, "num_tokens": 747903154.0, "step": 19693 }, { "epoch": 2.5052792265615063, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 5.879957675933838, "learning_rate": 1e-05, "loss": 0.4808, "mean_token_accuracy": 0.8731688261032104, "num_tokens": 747937849.0, "step": 19694 }, { "epoch": 2.505406436840097, "ewc_loss": 0.15625, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.0001277923583984375, "grad_norm": 7.786612510681152, "learning_rate": 1e-05, "loss": 0.4955, "mean_token_accuracy": 0.8762384653091431, "num_tokens": 747972387.0, "step": 19695 }, { "epoch": 2.5055336471186873, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.283288955688477, "learning_rate": 1e-05, "loss": 0.5387, "mean_token_accuracy": 0.8612774610519409, "num_tokens": 748007181.0, "step": 19696 }, { "epoch": 2.505660857397278, "ewc_loss": 0.15625, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.0001277923583984375, "grad_norm": 7.761529445648193, "learning_rate": 1e-05, "loss": 0.498, "mean_token_accuracy": 0.8820474147796631, "num_tokens": 748044342.0, "step": 19697 }, { "epoch": 2.5057880676758684, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00011014938354492188, "grad_norm": 6.663281440734863, "learning_rate": 1e-05, "loss": 0.5076, "mean_token_accuracy": 0.8696227073669434, "num_tokens": 748074012.0, "step": 19698 }, { "epoch": 2.5059152779544585, "ewc_loss": 0.1484375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00011968612670898438, "grad_norm": 7.411080837249756, "learning_rate": 1e-05, "loss": 0.4695, "mean_token_accuracy": 0.8859771490097046, "num_tokens": 748111055.0, "step": 19699 }, { "epoch": 2.5060424882330494, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 6.700917720794678, "learning_rate": 1e-05, "loss": 0.5467, "mean_token_accuracy": 0.857471227645874, "num_tokens": 748151922.0, "step": 19700 }, { "epoch": 2.5061696985116395, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00011110305786132812, "grad_norm": 7.049775123596191, "learning_rate": 1e-05, "loss": 0.4626, "mean_token_accuracy": 0.886640191078186, "num_tokens": 748186090.0, "step": 19701 }, { "epoch": 2.5062969087902305, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.613524436950684, "learning_rate": 1e-05, "loss": 0.4621, "mean_token_accuracy": 0.8832682967185974, "num_tokens": 748223281.0, "step": 19702 }, { "epoch": 2.5064241190688206, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.789044380187988, "learning_rate": 1e-05, "loss": 0.4708, "mean_token_accuracy": 0.8793856501579285, "num_tokens": 748255657.0, "step": 19703 }, { "epoch": 2.5065513293474115, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.4546308517456055, "learning_rate": 1e-05, "loss": 0.5196, "mean_token_accuracy": 0.8662225008010864, "num_tokens": 748296954.0, "step": 19704 }, { "epoch": 2.5066785396260016, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.671322822570801, "learning_rate": 1e-05, "loss": 0.4879, "mean_token_accuracy": 0.8703147172927856, "num_tokens": 748334753.0, "step": 19705 }, { "epoch": 2.506805749904592, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.281599044799805, "learning_rate": 1e-05, "loss": 0.4776, "mean_token_accuracy": 0.8795929551124573, "num_tokens": 748377880.0, "step": 19706 }, { "epoch": 2.5069329601831827, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.7027788162231445, "learning_rate": 1e-05, "loss": 0.4999, "mean_token_accuracy": 0.8727710843086243, "num_tokens": 748418230.0, "step": 19707 }, { "epoch": 2.507060170461773, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.11225700378418, "learning_rate": 1e-05, "loss": 0.4587, "mean_token_accuracy": 0.8820127844810486, "num_tokens": 748459320.0, "step": 19708 }, { "epoch": 2.5071873807403637, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.845070838928223, "learning_rate": 1e-05, "loss": 0.5335, "mean_token_accuracy": 0.8599446415901184, "num_tokens": 748503951.0, "step": 19709 }, { "epoch": 2.5073145910189543, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 5.954470634460449, "learning_rate": 1e-05, "loss": 0.468, "mean_token_accuracy": 0.8778495192527771, "num_tokens": 748544736.0, "step": 19710 }, { "epoch": 2.507441801297545, "ewc_loss": 0.140625, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00011205673217773438, "grad_norm": 7.117031097412109, "learning_rate": 1e-05, "loss": 0.4957, "mean_token_accuracy": 0.8781668543815613, "num_tokens": 748583174.0, "step": 19711 }, { "epoch": 2.5075690115761353, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 5.894885063171387, "learning_rate": 1e-05, "loss": 0.4648, "mean_token_accuracy": 0.8797667622566223, "num_tokens": 748624580.0, "step": 19712 }, { "epoch": 2.507696221854726, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00011777877807617188, "grad_norm": 7.446267604827881, "learning_rate": 1e-05, "loss": 0.5444, "mean_token_accuracy": 0.8614481091499329, "num_tokens": 748660343.0, "step": 19713 }, { "epoch": 2.5078234321333164, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.055643081665039, "learning_rate": 1e-05, "loss": 0.5167, "mean_token_accuracy": 0.8650104999542236, "num_tokens": 748695006.0, "step": 19714 }, { "epoch": 2.507950642411907, "ewc_loss": 0.150390625, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.0001220703125, "grad_norm": 7.611059188842773, "learning_rate": 1e-05, "loss": 0.4933, "mean_token_accuracy": 0.8819002509117126, "num_tokens": 748732142.0, "step": 19715 }, { "epoch": 2.5080778526904974, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.401488780975342, "learning_rate": 1e-05, "loss": 0.5164, "mean_token_accuracy": 0.8674137592315674, "num_tokens": 748771263.0, "step": 19716 }, { "epoch": 2.508205062969088, "ewc_loss": 0.1455078125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00011730194091796875, "grad_norm": 7.399810791015625, "learning_rate": 1e-05, "loss": 0.4979, "mean_token_accuracy": 0.8725297451019287, "num_tokens": 748812796.0, "step": 19717 }, { "epoch": 2.5083322732476785, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 6.639164447784424, "learning_rate": 1e-05, "loss": 0.4843, "mean_token_accuracy": 0.8759899139404297, "num_tokens": 748852701.0, "step": 19718 }, { "epoch": 2.508459483526269, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.924968719482422, "learning_rate": 1e-05, "loss": 0.5326, "mean_token_accuracy": 0.8680059909820557, "num_tokens": 748888799.0, "step": 19719 }, { "epoch": 2.5085866938048595, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.720006942749023, "learning_rate": 1e-05, "loss": 0.4829, "mean_token_accuracy": 0.878137469291687, "num_tokens": 748927056.0, "step": 19720 }, { "epoch": 2.50871390408345, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.409337997436523, "learning_rate": 1e-05, "loss": 0.5227, "mean_token_accuracy": 0.8617575168609619, "num_tokens": 748961829.0, "step": 19721 }, { "epoch": 2.5088411143620406, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.850690841674805, "learning_rate": 1e-05, "loss": 0.4614, "mean_token_accuracy": 0.8857723474502563, "num_tokens": 748995281.0, "step": 19722 }, { "epoch": 2.508968324640631, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 6.011469841003418, "learning_rate": 1e-05, "loss": 0.4486, "mean_token_accuracy": 0.8826515078544617, "num_tokens": 749027574.0, "step": 19723 }, { "epoch": 2.509095534919221, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 7.120649337768555, "learning_rate": 1e-05, "loss": 0.4761, "mean_token_accuracy": 0.8816417455673218, "num_tokens": 749065092.0, "step": 19724 }, { "epoch": 2.509222745197812, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.804940700531006, "learning_rate": 1e-05, "loss": 0.4937, "mean_token_accuracy": 0.8680694103240967, "num_tokens": 749104389.0, "step": 19725 }, { "epoch": 2.5093499554764023, "ewc_loss": 0.1484375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00011968612670898438, "grad_norm": 7.529212951660156, "learning_rate": 1e-05, "loss": 0.4528, "mean_token_accuracy": 0.8897278308868408, "num_tokens": 749138433.0, "step": 19726 }, { "epoch": 2.5094771657549932, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 5.937819957733154, "learning_rate": 1e-05, "loss": 0.4558, "mean_token_accuracy": 0.8837476372718811, "num_tokens": 749178095.0, "step": 19727 }, { "epoch": 2.5096043760335833, "ewc_loss": 0.1572265625, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00012874603271484375, "grad_norm": 7.838779926300049, "learning_rate": 1e-05, "loss": 0.4684, "mean_token_accuracy": 0.8920353651046753, "num_tokens": 749215547.0, "step": 19728 }, { "epoch": 2.5097315863121743, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.410158157348633, "learning_rate": 1e-05, "loss": 0.5193, "mean_token_accuracy": 0.8637251257896423, "num_tokens": 749254935.0, "step": 19729 }, { "epoch": 2.5098587965907644, "ewc_loss": 0.1533203125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00012493133544921875, "grad_norm": 7.70172643661499, "learning_rate": 1e-05, "loss": 0.4909, "mean_token_accuracy": 0.8880037069320679, "num_tokens": 749290425.0, "step": 19730 }, { "epoch": 2.509986006869355, "ewc_loss": 0.140625, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00011205673217773438, "grad_norm": 6.724830627441406, "learning_rate": 1e-05, "loss": 0.4684, "mean_token_accuracy": 0.8814760446548462, "num_tokens": 749319390.0, "step": 19731 }, { "epoch": 2.5101132171479454, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.0001163482666015625, "grad_norm": 7.2982916831970215, "learning_rate": 1e-05, "loss": 0.5316, "mean_token_accuracy": 0.8642150163650513, "num_tokens": 749358517.0, "step": 19732 }, { "epoch": 2.510240427426536, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 6.708810329437256, "learning_rate": 1e-05, "loss": 0.4918, "mean_token_accuracy": 0.8759434819221497, "num_tokens": 749402219.0, "step": 19733 }, { "epoch": 2.5103676377051265, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.953586578369141, "learning_rate": 1e-05, "loss": 0.5695, "mean_token_accuracy": 0.8485193848609924, "num_tokens": 749437514.0, "step": 19734 }, { "epoch": 2.510494847983717, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.5761284828186035, "learning_rate": 1e-05, "loss": 0.4793, "mean_token_accuracy": 0.879047155380249, "num_tokens": 749477148.0, "step": 19735 }, { "epoch": 2.5106220582623076, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.7230987548828125, "learning_rate": 1e-05, "loss": 0.4746, "mean_token_accuracy": 0.8798242807388306, "num_tokens": 749511174.0, "step": 19736 }, { "epoch": 2.510749268540898, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.430408000946045, "learning_rate": 1e-05, "loss": 0.468, "mean_token_accuracy": 0.8837373852729797, "num_tokens": 749547117.0, "step": 19737 }, { "epoch": 2.5108764788194886, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.6319193840026855, "learning_rate": 1e-05, "loss": 0.4572, "mean_token_accuracy": 0.8852421045303345, "num_tokens": 749583755.0, "step": 19738 }, { "epoch": 2.511003689098079, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.262609481811523, "learning_rate": 1e-05, "loss": 0.4634, "mean_token_accuracy": 0.8812135457992554, "num_tokens": 749621742.0, "step": 19739 }, { "epoch": 2.5111308993766697, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.663801193237305, "learning_rate": 1e-05, "loss": 0.4466, "mean_token_accuracy": 0.8895885348320007, "num_tokens": 749659178.0, "step": 19740 }, { "epoch": 2.51125810965526, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 6.096385955810547, "learning_rate": 1e-05, "loss": 0.4956, "mean_token_accuracy": 0.8677024841308594, "num_tokens": 749698828.0, "step": 19741 }, { "epoch": 2.5113853199338507, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.823126316070557, "learning_rate": 1e-05, "loss": 0.5288, "mean_token_accuracy": 0.8620678186416626, "num_tokens": 749732609.0, "step": 19742 }, { "epoch": 2.5115125302124413, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.926734447479248, "learning_rate": 1e-05, "loss": 0.4133, "mean_token_accuracy": 0.895078182220459, "num_tokens": 749768240.0, "step": 19743 }, { "epoch": 2.511639740491032, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 7.11346960067749, "learning_rate": 1e-05, "loss": 0.4637, "mean_token_accuracy": 0.8825347423553467, "num_tokens": 749809435.0, "step": 19744 }, { "epoch": 2.5117669507696223, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 5.84914493560791, "learning_rate": 1e-05, "loss": 0.4992, "mean_token_accuracy": 0.8678461313247681, "num_tokens": 749852790.0, "step": 19745 }, { "epoch": 2.511894161048213, "ewc_loss": 0.1484375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00011968612670898438, "grad_norm": 7.487359523773193, "learning_rate": 1e-05, "loss": 0.5088, "mean_token_accuracy": 0.8724343180656433, "num_tokens": 749895111.0, "step": 19746 }, { "epoch": 2.5120213713268034, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.021646976470947, "learning_rate": 1e-05, "loss": 0.4435, "mean_token_accuracy": 0.8858067989349365, "num_tokens": 749936142.0, "step": 19747 }, { "epoch": 2.512148581605394, "ewc_loss": 0.1533203125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00012493133544921875, "grad_norm": 7.698011875152588, "learning_rate": 1e-05, "loss": 0.4822, "mean_token_accuracy": 0.8842821717262268, "num_tokens": 749972106.0, "step": 19748 }, { "epoch": 2.512275791883984, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.434401988983154, "learning_rate": 1e-05, "loss": 0.5207, "mean_token_accuracy": 0.8642023205757141, "num_tokens": 750014090.0, "step": 19749 }, { "epoch": 2.512403002162575, "ewc_loss": 0.1474609375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00011873245239257812, "grad_norm": 7.477826118469238, "learning_rate": 1e-05, "loss": 0.4833, "mean_token_accuracy": 0.8845906257629395, "num_tokens": 750052542.0, "step": 19750 }, { "epoch": 2.512530212441165, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 6.699464321136475, "learning_rate": 1e-05, "loss": 0.5426, "mean_token_accuracy": 0.8634141087532043, "num_tokens": 750091846.0, "step": 19751 }, { "epoch": 2.512657422719756, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.974234580993652, "learning_rate": 1e-05, "loss": 0.5358, "mean_token_accuracy": 0.8607074022293091, "num_tokens": 750135721.0, "step": 19752 }, { "epoch": 2.512784632998346, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 6.761404514312744, "learning_rate": 1e-05, "loss": 0.5233, "mean_token_accuracy": 0.8658585548400879, "num_tokens": 750177786.0, "step": 19753 }, { "epoch": 2.5129118432769366, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.455512046813965, "learning_rate": 1e-05, "loss": 0.4851, "mean_token_accuracy": 0.8738933801651001, "num_tokens": 750209577.0, "step": 19754 }, { "epoch": 2.513039053555527, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.8534417152404785, "learning_rate": 1e-05, "loss": 0.4401, "mean_token_accuracy": 0.8921353816986084, "num_tokens": 750243168.0, "step": 19755 }, { "epoch": 2.5131662638341177, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 6.040586948394775, "learning_rate": 1e-05, "loss": 0.4755, "mean_token_accuracy": 0.8727262020111084, "num_tokens": 750288304.0, "step": 19756 }, { "epoch": 2.513293474112708, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00011110305786132812, "grad_norm": 7.112207889556885, "learning_rate": 1e-05, "loss": 0.5089, "mean_token_accuracy": 0.8728563785552979, "num_tokens": 750329662.0, "step": 19757 }, { "epoch": 2.5134206843912987, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.856956958770752, "learning_rate": 1e-05, "loss": 0.4829, "mean_token_accuracy": 0.8695927858352661, "num_tokens": 750364581.0, "step": 19758 }, { "epoch": 2.5135478946698893, "ewc_loss": 0.1474609375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00011873245239257812, "grad_norm": 7.479231834411621, "learning_rate": 1e-05, "loss": 0.5039, "mean_token_accuracy": 0.8808587789535522, "num_tokens": 750403129.0, "step": 19759 }, { "epoch": 2.51367510494848, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 5.962873458862305, "learning_rate": 1e-05, "loss": 0.4604, "mean_token_accuracy": 0.8777366876602173, "num_tokens": 750439738.0, "step": 19760 }, { "epoch": 2.5138023152270703, "ewc_loss": 0.15625, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.0001277923583984375, "grad_norm": 7.754073143005371, "learning_rate": 1e-05, "loss": 0.552, "mean_token_accuracy": 0.8608371019363403, "num_tokens": 750479302.0, "step": 19761 }, { "epoch": 2.513929525505661, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.34298849105835, "learning_rate": 1e-05, "loss": 0.4755, "mean_token_accuracy": 0.8783650994300842, "num_tokens": 750517658.0, "step": 19762 }, { "epoch": 2.5140567357842514, "ewc_loss": 0.1513671875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.656557083129883, "learning_rate": 1e-05, "loss": 0.4601, "mean_token_accuracy": 0.8908301591873169, "num_tokens": 750558192.0, "step": 19763 }, { "epoch": 2.514183946062842, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 6.63551139831543, "learning_rate": 1e-05, "loss": 0.4757, "mean_token_accuracy": 0.8790553212165833, "num_tokens": 750595209.0, "step": 19764 }, { "epoch": 2.5143111563414324, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00011587142944335938, "grad_norm": 7.325252056121826, "learning_rate": 1e-05, "loss": 0.5122, "mean_token_accuracy": 0.8711208701133728, "num_tokens": 750634560.0, "step": 19765 }, { "epoch": 2.514438366620023, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 6.6568379402160645, "learning_rate": 1e-05, "loss": 0.4727, "mean_token_accuracy": 0.879692792892456, "num_tokens": 750675223.0, "step": 19766 }, { "epoch": 2.5145655768986135, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 6.972987174987793, "learning_rate": 1e-05, "loss": 0.4899, "mean_token_accuracy": 0.8757866621017456, "num_tokens": 750717302.0, "step": 19767 }, { "epoch": 2.514692787177204, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.571045398712158, "learning_rate": 1e-05, "loss": 0.503, "mean_token_accuracy": 0.8744233846664429, "num_tokens": 750754615.0, "step": 19768 }, { "epoch": 2.5148199974557945, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.736501216888428, "learning_rate": 1e-05, "loss": 0.4428, "mean_token_accuracy": 0.890493631362915, "num_tokens": 750788745.0, "step": 19769 }, { "epoch": 2.514947207734385, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.438710689544678, "learning_rate": 1e-05, "loss": 0.5277, "mean_token_accuracy": 0.8608911037445068, "num_tokens": 750829727.0, "step": 19770 }, { "epoch": 2.5150744180129756, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.618594646453857, "learning_rate": 1e-05, "loss": 0.5308, "mean_token_accuracy": 0.8628939986228943, "num_tokens": 750873095.0, "step": 19771 }, { "epoch": 2.5152016282915657, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.30021858215332, "learning_rate": 1e-05, "loss": 0.5064, "mean_token_accuracy": 0.8647782206535339, "num_tokens": 750909513.0, "step": 19772 }, { "epoch": 2.5153288385701567, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.63035249710083, "learning_rate": 1e-05, "loss": 0.494, "mean_token_accuracy": 0.875314474105835, "num_tokens": 750947234.0, "step": 19773 }, { "epoch": 2.5154560488487467, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.131206035614014, "learning_rate": 1e-05, "loss": 0.427, "mean_token_accuracy": 0.8871382474899292, "num_tokens": 750984926.0, "step": 19774 }, { "epoch": 2.5155832591273377, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.735219478607178, "learning_rate": 1e-05, "loss": 0.4949, "mean_token_accuracy": 0.8764991760253906, "num_tokens": 751023700.0, "step": 19775 }, { "epoch": 2.515710469405928, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 5.980837345123291, "learning_rate": 1e-05, "loss": 0.4656, "mean_token_accuracy": 0.8783468008041382, "num_tokens": 751065444.0, "step": 19776 }, { "epoch": 2.5158376796845188, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.9743781089782715, "learning_rate": 1e-05, "loss": 0.5339, "mean_token_accuracy": 0.8597984910011292, "num_tokens": 751101627.0, "step": 19777 }, { "epoch": 2.515964889963109, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.864195823669434, "learning_rate": 1e-05, "loss": 0.4422, "mean_token_accuracy": 0.8830159306526184, "num_tokens": 751138006.0, "step": 19778 }, { "epoch": 2.5160921002416994, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00011539459228515625, "grad_norm": 7.307465553283691, "learning_rate": 1e-05, "loss": 0.4547, "mean_token_accuracy": 0.8881183862686157, "num_tokens": 751179187.0, "step": 19779 }, { "epoch": 2.51621931052029, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 5.9119062423706055, "learning_rate": 1e-05, "loss": 0.4566, "mean_token_accuracy": 0.8801642060279846, "num_tokens": 751217086.0, "step": 19780 }, { "epoch": 2.5163465207988804, "ewc_loss": 0.1513671875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.610199928283691, "learning_rate": 1e-05, "loss": 0.5155, "mean_token_accuracy": 0.8731685876846313, "num_tokens": 751259330.0, "step": 19781 }, { "epoch": 2.516473731077471, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.21547794342041, "learning_rate": 1e-05, "loss": 0.4771, "mean_token_accuracy": 0.8778669834136963, "num_tokens": 751300873.0, "step": 19782 }, { "epoch": 2.5166009413560615, "ewc_loss": 0.150390625, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.0001220703125, "grad_norm": 7.608402729034424, "learning_rate": 1e-05, "loss": 0.4557, "mean_token_accuracy": 0.8847925066947937, "num_tokens": 751330783.0, "step": 19783 }, { "epoch": 2.516728151634652, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.575098514556885, "learning_rate": 1e-05, "loss": 0.4806, "mean_token_accuracy": 0.8814328908920288, "num_tokens": 751370032.0, "step": 19784 }, { "epoch": 2.5168553619132426, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00011444091796875, "grad_norm": 7.204675674438477, "learning_rate": 1e-05, "loss": 0.4849, "mean_token_accuracy": 0.879956841468811, "num_tokens": 751412561.0, "step": 19785 }, { "epoch": 2.516982572191833, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 6.719740390777588, "learning_rate": 1e-05, "loss": 0.4867, "mean_token_accuracy": 0.8768196105957031, "num_tokens": 751442168.0, "step": 19786 }, { "epoch": 2.5171097824704236, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.681091785430908, "learning_rate": 1e-05, "loss": 0.5483, "mean_token_accuracy": 0.8563956022262573, "num_tokens": 751488656.0, "step": 19787 }, { "epoch": 2.517236992749014, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.782854080200195, "learning_rate": 1e-05, "loss": 0.4944, "mean_token_accuracy": 0.8736900687217712, "num_tokens": 751533305.0, "step": 19788 }, { "epoch": 2.5173642030276047, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.207982063293457, "learning_rate": 1e-05, "loss": 0.4744, "mean_token_accuracy": 0.8748352527618408, "num_tokens": 751570693.0, "step": 19789 }, { "epoch": 2.517491413306195, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 6.966325283050537, "learning_rate": 1e-05, "loss": 0.4517, "mean_token_accuracy": 0.8925970196723938, "num_tokens": 751604994.0, "step": 19790 }, { "epoch": 2.5176186235847857, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.911621570587158, "learning_rate": 1e-05, "loss": 0.4565, "mean_token_accuracy": 0.8776607513427734, "num_tokens": 751642552.0, "step": 19791 }, { "epoch": 2.5177458338633762, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00011444091796875, "grad_norm": 7.291960716247559, "learning_rate": 1e-05, "loss": 0.471, "mean_token_accuracy": 0.884670615196228, "num_tokens": 751677253.0, "step": 19792 }, { "epoch": 2.5178730441419668, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 5.869851112365723, "learning_rate": 1e-05, "loss": 0.4536, "mean_token_accuracy": 0.8815293908119202, "num_tokens": 751717737.0, "step": 19793 }, { "epoch": 2.5180002544205573, "ewc_loss": 0.15234375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.0001239776611328125, "grad_norm": 7.631867408752441, "learning_rate": 1e-05, "loss": 0.4477, "mean_token_accuracy": 0.8976052403450012, "num_tokens": 751750453.0, "step": 19794 }, { "epoch": 2.518127464699148, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.138234615325928, "learning_rate": 1e-05, "loss": 0.4781, "mean_token_accuracy": 0.8759608864784241, "num_tokens": 751792179.0, "step": 19795 }, { "epoch": 2.5182546749777384, "ewc_loss": 0.1552734375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00012683868408203125, "grad_norm": 7.731836318969727, "learning_rate": 1e-05, "loss": 0.5046, "mean_token_accuracy": 0.8780288100242615, "num_tokens": 751833367.0, "step": 19796 }, { "epoch": 2.5183818852563284, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 6.51966667175293, "learning_rate": 1e-05, "loss": 0.4811, "mean_token_accuracy": 0.8801683187484741, "num_tokens": 751874042.0, "step": 19797 }, { "epoch": 2.5185090955349194, "ewc_loss": 0.1484375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00011968612670898438, "grad_norm": 7.471302509307861, "learning_rate": 1e-05, "loss": 0.575, "mean_token_accuracy": 0.8533135652542114, "num_tokens": 751914990.0, "step": 19798 }, { "epoch": 2.5186363058135095, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 6.678185939788818, "learning_rate": 1e-05, "loss": 0.5337, "mean_token_accuracy": 0.8614998459815979, "num_tokens": 751946266.0, "step": 19799 }, { "epoch": 2.5187635160921005, "ewc_loss": 0.140625, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.0001125335693359375, "grad_norm": 7.090656280517578, "learning_rate": 1e-05, "loss": 0.4671, "mean_token_accuracy": 0.8869626522064209, "num_tokens": 751990359.0, "step": 19800 }, { "epoch": 2.5188907263706906, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.608221054077148, "learning_rate": 1e-05, "loss": 0.4911, "mean_token_accuracy": 0.8750278353691101, "num_tokens": 752026955.0, "step": 19801 }, { "epoch": 2.5190179366492815, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.800858497619629, "learning_rate": 1e-05, "loss": 0.5096, "mean_token_accuracy": 0.8663432598114014, "num_tokens": 752065187.0, "step": 19802 }, { "epoch": 2.5191451469278716, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.491227626800537, "learning_rate": 1e-05, "loss": 0.432, "mean_token_accuracy": 0.8905634880065918, "num_tokens": 752100855.0, "step": 19803 }, { "epoch": 2.519272357206462, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.638016700744629, "learning_rate": 1e-05, "loss": 0.4708, "mean_token_accuracy": 0.8809201717376709, "num_tokens": 752136672.0, "step": 19804 }, { "epoch": 2.5193995674850527, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.3492608070373535, "learning_rate": 1e-05, "loss": 0.4552, "mean_token_accuracy": 0.8839629888534546, "num_tokens": 752167529.0, "step": 19805 }, { "epoch": 2.519526777763643, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.612165927886963, "learning_rate": 1e-05, "loss": 0.516, "mean_token_accuracy": 0.8636414408683777, "num_tokens": 752201865.0, "step": 19806 }, { "epoch": 2.5196539880422337, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.196195602416992, "learning_rate": 1e-05, "loss": 0.4657, "mean_token_accuracy": 0.8783153295516968, "num_tokens": 752239315.0, "step": 19807 }, { "epoch": 2.5197811983208243, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.671283721923828, "learning_rate": 1e-05, "loss": 0.4574, "mean_token_accuracy": 0.8846239447593689, "num_tokens": 752278505.0, "step": 19808 }, { "epoch": 2.519908408599415, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 6.058741569519043, "learning_rate": 1e-05, "loss": 0.4481, "mean_token_accuracy": 0.8844102621078491, "num_tokens": 752319346.0, "step": 19809 }, { "epoch": 2.5200356188780053, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.83274507522583, "learning_rate": 1e-05, "loss": 0.5235, "mean_token_accuracy": 0.8664897084236145, "num_tokens": 752364210.0, "step": 19810 }, { "epoch": 2.520162829156596, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.906400680541992, "learning_rate": 1e-05, "loss": 0.4272, "mean_token_accuracy": 0.8901850581169128, "num_tokens": 752400679.0, "step": 19811 }, { "epoch": 2.5202900394351864, "ewc_loss": 0.140625, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00011205673217773438, "grad_norm": 7.121032238006592, "learning_rate": 1e-05, "loss": 0.5004, "mean_token_accuracy": 0.8779246211051941, "num_tokens": 752436003.0, "step": 19812 }, { "epoch": 2.520417249713777, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.837557792663574, "learning_rate": 1e-05, "loss": 0.5006, "mean_token_accuracy": 0.8681889176368713, "num_tokens": 752478690.0, "step": 19813 }, { "epoch": 2.5205444599923674, "ewc_loss": 0.1494140625, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00012111663818359375, "grad_norm": 7.491945266723633, "learning_rate": 1e-05, "loss": 0.5531, "mean_token_accuracy": 0.8616228103637695, "num_tokens": 752513326.0, "step": 19814 }, { "epoch": 2.520671670270958, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.019554615020752, "learning_rate": 1e-05, "loss": 0.5089, "mean_token_accuracy": 0.8669573664665222, "num_tokens": 752556553.0, "step": 19815 }, { "epoch": 2.5207988805495485, "ewc_loss": 0.154296875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.000125885009765625, "grad_norm": 7.677005290985107, "learning_rate": 1e-05, "loss": 0.4727, "mean_token_accuracy": 0.8855244517326355, "num_tokens": 752599219.0, "step": 19816 }, { "epoch": 2.520926090828139, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.405566215515137, "learning_rate": 1e-05, "loss": 0.4745, "mean_token_accuracy": 0.8816201090812683, "num_tokens": 752639308.0, "step": 19817 }, { "epoch": 2.5210533011067295, "ewc_loss": 0.1484375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00011968612670898438, "grad_norm": 7.455180644989014, "learning_rate": 1e-05, "loss": 0.5539, "mean_token_accuracy": 0.8605791330337524, "num_tokens": 752679384.0, "step": 19818 }, { "epoch": 2.52118051138532, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 6.661672592163086, "learning_rate": 1e-05, "loss": 0.5061, "mean_token_accuracy": 0.8725705146789551, "num_tokens": 752722395.0, "step": 19819 }, { "epoch": 2.5213077216639106, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 6.988449573516846, "learning_rate": 1e-05, "loss": 0.5186, "mean_token_accuracy": 0.875487208366394, "num_tokens": 752754491.0, "step": 19820 }, { "epoch": 2.521434931942501, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 6.731976509094238, "learning_rate": 1e-05, "loss": 0.4623, "mean_token_accuracy": 0.8845103979110718, "num_tokens": 752792595.0, "step": 19821 }, { "epoch": 2.521562142221091, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.472940921783447, "learning_rate": 1e-05, "loss": 0.4925, "mean_token_accuracy": 0.8739312887191772, "num_tokens": 752830689.0, "step": 19822 }, { "epoch": 2.521689352499682, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.840222358703613, "learning_rate": 1e-05, "loss": 0.4363, "mean_token_accuracy": 0.8918768167495728, "num_tokens": 752862848.0, "step": 19823 }, { "epoch": 2.5218165627782723, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 6.081787109375, "learning_rate": 1e-05, "loss": 0.461, "mean_token_accuracy": 0.8791937828063965, "num_tokens": 752900566.0, "step": 19824 }, { "epoch": 2.5219437730568632, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 7.052521705627441, "learning_rate": 1e-05, "loss": 0.4647, "mean_token_accuracy": 0.8856785297393799, "num_tokens": 752939201.0, "step": 19825 }, { "epoch": 2.5220709833354533, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 5.885869979858398, "learning_rate": 1e-05, "loss": 0.4457, "mean_token_accuracy": 0.8845124244689941, "num_tokens": 752973820.0, "step": 19826 }, { "epoch": 2.522198193614044, "ewc_loss": 0.1455078125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00011730194091796875, "grad_norm": 7.366706371307373, "learning_rate": 1e-05, "loss": 0.5681, "mean_token_accuracy": 0.8555556535720825, "num_tokens": 753018563.0, "step": 19827 }, { "epoch": 2.5223254038926344, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 5.9272141456604, "learning_rate": 1e-05, "loss": 0.4923, "mean_token_accuracy": 0.8687984943389893, "num_tokens": 753060097.0, "step": 19828 }, { "epoch": 2.522452614171225, "ewc_loss": 0.1513671875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.649765968322754, "learning_rate": 1e-05, "loss": 0.5286, "mean_token_accuracy": 0.8699530363082886, "num_tokens": 753099200.0, "step": 19829 }, { "epoch": 2.5225798244498154, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.241481781005859, "learning_rate": 1e-05, "loss": 0.4593, "mean_token_accuracy": 0.8796782493591309, "num_tokens": 753132421.0, "step": 19830 }, { "epoch": 2.522707034728406, "ewc_loss": 0.15234375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.0001239776611328125, "grad_norm": 7.645046234130859, "learning_rate": 1e-05, "loss": 0.5267, "mean_token_accuracy": 0.8668233156204224, "num_tokens": 753165380.0, "step": 19831 }, { "epoch": 2.5228342450069965, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 6.525913715362549, "learning_rate": 1e-05, "loss": 0.4747, "mean_token_accuracy": 0.878110408782959, "num_tokens": 753203501.0, "step": 19832 }, { "epoch": 2.522961455285587, "ewc_loss": 0.1455078125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00011730194091796875, "grad_norm": 7.360597133636475, "learning_rate": 1e-05, "loss": 0.5351, "mean_token_accuracy": 0.8658562302589417, "num_tokens": 753247674.0, "step": 19833 }, { "epoch": 2.5230886655641775, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 6.630474090576172, "learning_rate": 1e-05, "loss": 0.4771, "mean_token_accuracy": 0.879011332988739, "num_tokens": 753285967.0, "step": 19834 }, { "epoch": 2.523215875842768, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 7.008732318878174, "learning_rate": 1e-05, "loss": 0.5023, "mean_token_accuracy": 0.8715827465057373, "num_tokens": 753319080.0, "step": 19835 }, { "epoch": 2.5233430861213586, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.545906066894531, "learning_rate": 1e-05, "loss": 0.5128, "mean_token_accuracy": 0.867893636226654, "num_tokens": 753357924.0, "step": 19836 }, { "epoch": 2.523470296399949, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.744819641113281, "learning_rate": 1e-05, "loss": 0.4724, "mean_token_accuracy": 0.8836350440979004, "num_tokens": 753397922.0, "step": 19837 }, { "epoch": 2.5235975066785397, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.451580047607422, "learning_rate": 1e-05, "loss": 0.5552, "mean_token_accuracy": 0.8537193536758423, "num_tokens": 753431268.0, "step": 19838 }, { "epoch": 2.52372471695713, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.616799831390381, "learning_rate": 1e-05, "loss": 0.5032, "mean_token_accuracy": 0.8710249662399292, "num_tokens": 753465169.0, "step": 19839 }, { "epoch": 2.5238519272357207, "ewc_loss": 0.125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.309374809265137, "learning_rate": 1e-05, "loss": 0.4731, "mean_token_accuracy": 0.8773220777511597, "num_tokens": 753501017.0, "step": 19840 }, { "epoch": 2.5239791375143112, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.594350337982178, "learning_rate": 1e-05, "loss": 0.4623, "mean_token_accuracy": 0.8795645236968994, "num_tokens": 753541039.0, "step": 19841 }, { "epoch": 2.5241063477929018, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.153896331787109, "learning_rate": 1e-05, "loss": 0.56, "mean_token_accuracy": 0.8491178154945374, "num_tokens": 753585793.0, "step": 19842 }, { "epoch": 2.5242335580714923, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.680713176727295, "learning_rate": 1e-05, "loss": 0.4378, "mean_token_accuracy": 0.8887888789176941, "num_tokens": 753623752.0, "step": 19843 }, { "epoch": 2.524360768350083, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 6.006420612335205, "learning_rate": 1e-05, "loss": 0.4877, "mean_token_accuracy": 0.8697682619094849, "num_tokens": 753662939.0, "step": 19844 }, { "epoch": 2.5244879786286734, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.8845038414001465, "learning_rate": 1e-05, "loss": 0.4776, "mean_token_accuracy": 0.8798156976699829, "num_tokens": 753699200.0, "step": 19845 }, { "epoch": 2.524615188907264, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.8798065185546875, "learning_rate": 1e-05, "loss": 0.4934, "mean_token_accuracy": 0.8667130470275879, "num_tokens": 753743867.0, "step": 19846 }, { "epoch": 2.524742399185854, "ewc_loss": 0.140625, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.0001125335693359375, "grad_norm": 7.19439172744751, "learning_rate": 1e-05, "loss": 0.5332, "mean_token_accuracy": 0.8644534349441528, "num_tokens": 753781494.0, "step": 19847 }, { "epoch": 2.524869609464445, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 5.862172603607178, "learning_rate": 1e-05, "loss": 0.4753, "mean_token_accuracy": 0.878426730632782, "num_tokens": 753823518.0, "step": 19848 }, { "epoch": 2.524996819743035, "ewc_loss": 0.150390625, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.0001220703125, "grad_norm": 7.573287487030029, "learning_rate": 1e-05, "loss": 0.5217, "mean_token_accuracy": 0.8723348379135132, "num_tokens": 753855058.0, "step": 19849 }, { "epoch": 2.525124030021626, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.083070755004883, "learning_rate": 1e-05, "loss": 0.4634, "mean_token_accuracy": 0.8806953430175781, "num_tokens": 753896721.0, "step": 19850 }, { "epoch": 2.525251240300216, "ewc_loss": 0.154296875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.000125885009765625, "grad_norm": 7.695201873779297, "learning_rate": 1e-05, "loss": 0.4676, "mean_token_accuracy": 0.8902255296707153, "num_tokens": 753933225.0, "step": 19851 }, { "epoch": 2.5253784505788066, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.491108417510986, "learning_rate": 1e-05, "loss": 0.4978, "mean_token_accuracy": 0.8714888095855713, "num_tokens": 753970373.0, "step": 19852 }, { "epoch": 2.525505660857397, "ewc_loss": 0.1455078125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00011730194091796875, "grad_norm": 7.41408634185791, "learning_rate": 1e-05, "loss": 0.4623, "mean_token_accuracy": 0.8851908445358276, "num_tokens": 754007544.0, "step": 19853 }, { "epoch": 2.5256328711359877, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00011110305786132812, "grad_norm": 6.706645965576172, "learning_rate": 1e-05, "loss": 0.4848, "mean_token_accuracy": 0.8810076117515564, "num_tokens": 754040858.0, "step": 19854 }, { "epoch": 2.525760081414578, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.9202752113342285, "learning_rate": 1e-05, "loss": 0.5558, "mean_token_accuracy": 0.8591172695159912, "num_tokens": 754083877.0, "step": 19855 }, { "epoch": 2.5258872916931687, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 6.740586280822754, "learning_rate": 1e-05, "loss": 0.4556, "mean_token_accuracy": 0.8883765339851379, "num_tokens": 754122792.0, "step": 19856 }, { "epoch": 2.5260145019717593, "ewc_loss": 0.125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.435364246368408, "learning_rate": 1e-05, "loss": 0.473, "mean_token_accuracy": 0.877039909362793, "num_tokens": 754157986.0, "step": 19857 }, { "epoch": 2.52614171225035, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.835000038146973, "learning_rate": 1e-05, "loss": 0.4948, "mean_token_accuracy": 0.8765456676483154, "num_tokens": 754198321.0, "step": 19858 }, { "epoch": 2.5262689225289403, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 6.073030471801758, "learning_rate": 1e-05, "loss": 0.5258, "mean_token_accuracy": 0.8611023426055908, "num_tokens": 754235463.0, "step": 19859 }, { "epoch": 2.526396132807531, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00010967254638671875, "grad_norm": 7.046299934387207, "learning_rate": 1e-05, "loss": 0.4928, "mean_token_accuracy": 0.8772001266479492, "num_tokens": 754274319.0, "step": 19860 }, { "epoch": 2.5265233430861214, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.882138252258301, "learning_rate": 1e-05, "loss": 0.4614, "mean_token_accuracy": 0.8800814747810364, "num_tokens": 754315206.0, "step": 19861 }, { "epoch": 2.526650553364712, "ewc_loss": 0.1455078125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00011682510375976562, "grad_norm": 7.353434085845947, "learning_rate": 1e-05, "loss": 0.5269, "mean_token_accuracy": 0.8714617490768433, "num_tokens": 754353300.0, "step": 19862 }, { "epoch": 2.5267777636433024, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 5.939781665802002, "learning_rate": 1e-05, "loss": 0.5112, "mean_token_accuracy": 0.8681244850158691, "num_tokens": 754392692.0, "step": 19863 }, { "epoch": 2.526904973921893, "ewc_loss": 0.1513671875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.625836372375488, "learning_rate": 1e-05, "loss": 0.5068, "mean_token_accuracy": 0.8771160840988159, "num_tokens": 754428688.0, "step": 19864 }, { "epoch": 2.5270321842004835, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.217202186584473, "learning_rate": 1e-05, "loss": 0.4624, "mean_token_accuracy": 0.8772059679031372, "num_tokens": 754467647.0, "step": 19865 }, { "epoch": 2.527159394479074, "ewc_loss": 0.1513671875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.62268590927124, "learning_rate": 1e-05, "loss": 0.5291, "mean_token_accuracy": 0.8624394536018372, "num_tokens": 754501511.0, "step": 19866 }, { "epoch": 2.5272866047576645, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 6.503762722015381, "learning_rate": 1e-05, "loss": 0.4475, "mean_token_accuracy": 0.8894963264465332, "num_tokens": 754543000.0, "step": 19867 }, { "epoch": 2.527413815036255, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00011730194091796875, "grad_norm": 7.362095832824707, "learning_rate": 1e-05, "loss": 0.4962, "mean_token_accuracy": 0.8782508373260498, "num_tokens": 754583485.0, "step": 19868 }, { "epoch": 2.5275410253148456, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 6.598552227020264, "learning_rate": 1e-05, "loss": 0.4772, "mean_token_accuracy": 0.8796589970588684, "num_tokens": 754620472.0, "step": 19869 }, { "epoch": 2.5276682355934357, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00011014938354492188, "grad_norm": 7.034351348876953, "learning_rate": 1e-05, "loss": 0.479, "mean_token_accuracy": 0.8813822269439697, "num_tokens": 754659906.0, "step": 19870 }, { "epoch": 2.5277954458720266, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.5516533851623535, "learning_rate": 1e-05, "loss": 0.4774, "mean_token_accuracy": 0.8768064379692078, "num_tokens": 754696128.0, "step": 19871 }, { "epoch": 2.5279226561506167, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.771032333374023, "learning_rate": 1e-05, "loss": 0.481, "mean_token_accuracy": 0.8768541812896729, "num_tokens": 754729553.0, "step": 19872 }, { "epoch": 2.5280498664292077, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.447945594787598, "learning_rate": 1e-05, "loss": 0.5286, "mean_token_accuracy": 0.8616526126861572, "num_tokens": 754773689.0, "step": 19873 }, { "epoch": 2.528177076707798, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.6294145584106445, "learning_rate": 1e-05, "loss": 0.501, "mean_token_accuracy": 0.8715024590492249, "num_tokens": 754809326.0, "step": 19874 }, { "epoch": 2.5283042869863888, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.320404529571533, "learning_rate": 1e-05, "loss": 0.4755, "mean_token_accuracy": 0.8773391246795654, "num_tokens": 754850627.0, "step": 19875 }, { "epoch": 2.528431497264979, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.609186172485352, "learning_rate": 1e-05, "loss": 0.4535, "mean_token_accuracy": 0.882962167263031, "num_tokens": 754887810.0, "step": 19876 }, { "epoch": 2.5285587075435694, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.1808552742004395, "learning_rate": 1e-05, "loss": 0.4816, "mean_token_accuracy": 0.8738014698028564, "num_tokens": 754923778.0, "step": 19877 }, { "epoch": 2.52868591782216, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.682997703552246, "learning_rate": 1e-05, "loss": 0.4816, "mean_token_accuracy": 0.8779787421226501, "num_tokens": 754963486.0, "step": 19878 }, { "epoch": 2.5288131281007504, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.0220794677734375, "learning_rate": 1e-05, "loss": 0.4327, "mean_token_accuracy": 0.8884673118591309, "num_tokens": 755000302.0, "step": 19879 }, { "epoch": 2.528940338379341, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.8580002784729, "learning_rate": 1e-05, "loss": 0.4403, "mean_token_accuracy": 0.8901932835578918, "num_tokens": 755040319.0, "step": 19880 }, { "epoch": 2.5290675486579315, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.9084367752075195, "learning_rate": 1e-05, "loss": 0.4585, "mean_token_accuracy": 0.8824034929275513, "num_tokens": 755073410.0, "step": 19881 }, { "epoch": 2.529194758936522, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 7.144213676452637, "learning_rate": 1e-05, "loss": 0.4853, "mean_token_accuracy": 0.8820164799690247, "num_tokens": 755112279.0, "step": 19882 }, { "epoch": 2.5293219692151125, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 5.873100280761719, "learning_rate": 1e-05, "loss": 0.527, "mean_token_accuracy": 0.8583488464355469, "num_tokens": 755146904.0, "step": 19883 }, { "epoch": 2.529449179493703, "ewc_loss": 0.1484375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00011968612670898438, "grad_norm": 7.485280990600586, "learning_rate": 1e-05, "loss": 0.5186, "mean_token_accuracy": 0.8695047497749329, "num_tokens": 755179881.0, "step": 19884 }, { "epoch": 2.5295763897722936, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.051388740539551, "learning_rate": 1e-05, "loss": 0.4417, "mean_token_accuracy": 0.8887385129928589, "num_tokens": 755214504.0, "step": 19885 }, { "epoch": 2.529703600050884, "ewc_loss": 0.15234375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.0001239776611328125, "grad_norm": 7.628333568572998, "learning_rate": 1e-05, "loss": 0.5141, "mean_token_accuracy": 0.8759732246398926, "num_tokens": 755254229.0, "step": 19886 }, { "epoch": 2.5298308103294747, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.416844367980957, "learning_rate": 1e-05, "loss": 0.5036, "mean_token_accuracy": 0.8707597255706787, "num_tokens": 755296150.0, "step": 19887 }, { "epoch": 2.529958020608065, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.000118255615234375, "grad_norm": 7.373228073120117, "learning_rate": 1e-05, "loss": 0.4841, "mean_token_accuracy": 0.883065938949585, "num_tokens": 755336508.0, "step": 19888 }, { "epoch": 2.5300852308866557, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 6.634293079376221, "learning_rate": 1e-05, "loss": 0.5111, "mean_token_accuracy": 0.8642566204071045, "num_tokens": 755376863.0, "step": 19889 }, { "epoch": 2.5302124411652462, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.903332233428955, "learning_rate": 1e-05, "loss": 0.4713, "mean_token_accuracy": 0.8841009140014648, "num_tokens": 755414209.0, "step": 19890 }, { "epoch": 2.5303396514438368, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 6.720480918884277, "learning_rate": 1e-05, "loss": 0.5386, "mean_token_accuracy": 0.8599668741226196, "num_tokens": 755454281.0, "step": 19891 }, { "epoch": 2.5304668617224273, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.405783653259277, "learning_rate": 1e-05, "loss": 0.4225, "mean_token_accuracy": 0.8939933776855469, "num_tokens": 755492441.0, "step": 19892 }, { "epoch": 2.530594072001018, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.851467132568359, "learning_rate": 1e-05, "loss": 0.4758, "mean_token_accuracy": 0.8758969306945801, "num_tokens": 755526264.0, "step": 19893 }, { "epoch": 2.5307212822796084, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.038476467132568, "learning_rate": 1e-05, "loss": 0.4629, "mean_token_accuracy": 0.8804881572723389, "num_tokens": 755560819.0, "step": 19894 }, { "epoch": 2.5308484925581984, "ewc_loss": 0.140625, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00011157989501953125, "grad_norm": 7.106583595275879, "learning_rate": 1e-05, "loss": 0.4902, "mean_token_accuracy": 0.8747962713241577, "num_tokens": 755592103.0, "step": 19895 }, { "epoch": 2.5309757028367894, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.846123218536377, "learning_rate": 1e-05, "loss": 0.4486, "mean_token_accuracy": 0.887753963470459, "num_tokens": 755631864.0, "step": 19896 }, { "epoch": 2.5311029131153795, "ewc_loss": 0.1494140625, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00012111663818359375, "grad_norm": 7.485475063323975, "learning_rate": 1e-05, "loss": 0.516, "mean_token_accuracy": 0.8701052665710449, "num_tokens": 755672555.0, "step": 19897 }, { "epoch": 2.5312301233939705, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 5.951836585998535, "learning_rate": 1e-05, "loss": 0.4358, "mean_token_accuracy": 0.8893428444862366, "num_tokens": 755705606.0, "step": 19898 }, { "epoch": 2.5313573336725606, "ewc_loss": 0.15625, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.0001277923583984375, "grad_norm": 7.777247905731201, "learning_rate": 1e-05, "loss": 0.533, "mean_token_accuracy": 0.8717705011367798, "num_tokens": 755738844.0, "step": 19899 }, { "epoch": 2.5314845439511515, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.340447902679443, "learning_rate": 1e-05, "loss": 0.4555, "mean_token_accuracy": 0.8858514428138733, "num_tokens": 755774164.0, "step": 19900 }, { "epoch": 2.5316117542297416, "ewc_loss": 0.154296875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.000125885009765625, "grad_norm": 7.703092575073242, "learning_rate": 1e-05, "loss": 0.5443, "mean_token_accuracy": 0.8685125112533569, "num_tokens": 755816848.0, "step": 19901 }, { "epoch": 2.531738964508332, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 6.6418585777282715, "learning_rate": 1e-05, "loss": 0.4981, "mean_token_accuracy": 0.8729212880134583, "num_tokens": 755854281.0, "step": 19902 }, { "epoch": 2.5318661747869227, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00011777877807617188, "grad_norm": 7.370319843292236, "learning_rate": 1e-05, "loss": 0.558, "mean_token_accuracy": 0.8606933355331421, "num_tokens": 755893769.0, "step": 19903 }, { "epoch": 2.531993385065513, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 6.681059837341309, "learning_rate": 1e-05, "loss": 0.499, "mean_token_accuracy": 0.8727573156356812, "num_tokens": 755931690.0, "step": 19904 }, { "epoch": 2.5321205953441037, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00010967254638671875, "grad_norm": 7.0172529220581055, "learning_rate": 1e-05, "loss": 0.4856, "mean_token_accuracy": 0.8766841292381287, "num_tokens": 755967177.0, "step": 19905 }, { "epoch": 2.5322478056226942, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.590048789978027, "learning_rate": 1e-05, "loss": 0.4504, "mean_token_accuracy": 0.885566234588623, "num_tokens": 755998440.0, "step": 19906 }, { "epoch": 2.5323750159012848, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.753125190734863, "learning_rate": 1e-05, "loss": 0.4922, "mean_token_accuracy": 0.8727548122406006, "num_tokens": 756038049.0, "step": 19907 }, { "epoch": 2.5325022261798753, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.45029878616333, "learning_rate": 1e-05, "loss": 0.4451, "mean_token_accuracy": 0.8886001110076904, "num_tokens": 756073185.0, "step": 19908 }, { "epoch": 2.532629436458466, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.639861106872559, "learning_rate": 1e-05, "loss": 0.506, "mean_token_accuracy": 0.8677517175674438, "num_tokens": 756113439.0, "step": 19909 }, { "epoch": 2.5327566467370564, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.294338226318359, "learning_rate": 1e-05, "loss": 0.4857, "mean_token_accuracy": 0.879242479801178, "num_tokens": 756155516.0, "step": 19910 }, { "epoch": 2.532883857015647, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.652799129486084, "learning_rate": 1e-05, "loss": 0.4799, "mean_token_accuracy": 0.8770419955253601, "num_tokens": 756196685.0, "step": 19911 }, { "epoch": 2.5330110672942374, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.14204740524292, "learning_rate": 1e-05, "loss": 0.4744, "mean_token_accuracy": 0.8733574151992798, "num_tokens": 756230150.0, "step": 19912 }, { "epoch": 2.533138277572828, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.7584309577941895, "learning_rate": 1e-05, "loss": 0.4527, "mean_token_accuracy": 0.8878148794174194, "num_tokens": 756270806.0, "step": 19913 }, { "epoch": 2.5332654878514185, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 5.994145393371582, "learning_rate": 1e-05, "loss": 0.4251, "mean_token_accuracy": 0.8878487944602966, "num_tokens": 756300387.0, "step": 19914 }, { "epoch": 2.533392698130009, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.992182731628418, "learning_rate": 1e-05, "loss": 0.4819, "mean_token_accuracy": 0.8807666897773743, "num_tokens": 756335436.0, "step": 19915 }, { "epoch": 2.5335199084085995, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.87404203414917, "learning_rate": 1e-05, "loss": 0.5107, "mean_token_accuracy": 0.8654203414916992, "num_tokens": 756371942.0, "step": 19916 }, { "epoch": 2.53364711868719, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.0001163482666015625, "grad_norm": 7.328220367431641, "learning_rate": 1e-05, "loss": 0.4993, "mean_token_accuracy": 0.8790307641029358, "num_tokens": 756416102.0, "step": 19917 }, { "epoch": 2.5337743289657806, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 5.922015190124512, "learning_rate": 1e-05, "loss": 0.5094, "mean_token_accuracy": 0.8637298345565796, "num_tokens": 756457066.0, "step": 19918 }, { "epoch": 2.533901539244371, "ewc_loss": 0.1513671875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.634615898132324, "learning_rate": 1e-05, "loss": 0.5015, "mean_token_accuracy": 0.877443253993988, "num_tokens": 756493977.0, "step": 19919 }, { "epoch": 2.534028749522961, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.250941276550293, "learning_rate": 1e-05, "loss": 0.5351, "mean_token_accuracy": 0.8557965159416199, "num_tokens": 756535926.0, "step": 19920 }, { "epoch": 2.534155959801552, "ewc_loss": 0.150390625, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.0001220703125, "grad_norm": 7.592438697814941, "learning_rate": 1e-05, "loss": 0.4674, "mean_token_accuracy": 0.890891432762146, "num_tokens": 756578406.0, "step": 19921 }, { "epoch": 2.5342831700801423, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.608634948730469, "learning_rate": 1e-05, "loss": 0.5354, "mean_token_accuracy": 0.8591241240501404, "num_tokens": 756615104.0, "step": 19922 }, { "epoch": 2.5344103803587332, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00011348724365234375, "grad_norm": 7.164482593536377, "learning_rate": 1e-05, "loss": 0.5213, "mean_token_accuracy": 0.8694121241569519, "num_tokens": 756652059.0, "step": 19923 }, { "epoch": 2.5345375906373233, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 6.709242343902588, "learning_rate": 1e-05, "loss": 0.4369, "mean_token_accuracy": 0.8951732516288757, "num_tokens": 756688640.0, "step": 19924 }, { "epoch": 2.534664800915914, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.657664775848389, "learning_rate": 1e-05, "loss": 0.4501, "mean_token_accuracy": 0.8886431455612183, "num_tokens": 756721596.0, "step": 19925 }, { "epoch": 2.5347920111945044, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.783039093017578, "learning_rate": 1e-05, "loss": 0.5167, "mean_token_accuracy": 0.8660945892333984, "num_tokens": 756757663.0, "step": 19926 }, { "epoch": 2.534919221473095, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.1923089027404785, "learning_rate": 1e-05, "loss": 0.4657, "mean_token_accuracy": 0.8781274557113647, "num_tokens": 756802021.0, "step": 19927 }, { "epoch": 2.5350464317516854, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 6.980639457702637, "learning_rate": 1e-05, "loss": 0.5156, "mean_token_accuracy": 0.8673957586288452, "num_tokens": 756842039.0, "step": 19928 }, { "epoch": 2.535173642030276, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.904799938201904, "learning_rate": 1e-05, "loss": 0.4738, "mean_token_accuracy": 0.8781191110610962, "num_tokens": 756875788.0, "step": 19929 }, { "epoch": 2.5353008523088665, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00011444091796875, "grad_norm": 7.322474956512451, "learning_rate": 1e-05, "loss": 0.4747, "mean_token_accuracy": 0.8851202726364136, "num_tokens": 756913292.0, "step": 19930 }, { "epoch": 2.535428062587457, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.859206199645996, "learning_rate": 1e-05, "loss": 0.4295, "mean_token_accuracy": 0.8944705724716187, "num_tokens": 756953005.0, "step": 19931 }, { "epoch": 2.5355552728660475, "ewc_loss": 0.1533203125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00012493133544921875, "grad_norm": 7.693010330200195, "learning_rate": 1e-05, "loss": 0.5045, "mean_token_accuracy": 0.8779948949813843, "num_tokens": 756987656.0, "step": 19932 }, { "epoch": 2.535682483144638, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.171144962310791, "learning_rate": 1e-05, "loss": 0.4561, "mean_token_accuracy": 0.8836261034011841, "num_tokens": 757021040.0, "step": 19933 }, { "epoch": 2.5358096934232286, "ewc_loss": 0.1572265625, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00012874603271484375, "grad_norm": 7.768802642822266, "learning_rate": 1e-05, "loss": 0.4677, "mean_token_accuracy": 0.8906630277633667, "num_tokens": 757053039.0, "step": 19934 }, { "epoch": 2.535936903701819, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 6.563688278198242, "learning_rate": 1e-05, "loss": 0.51, "mean_token_accuracy": 0.8703550100326538, "num_tokens": 757090057.0, "step": 19935 }, { "epoch": 2.5360641139804097, "ewc_loss": 0.1484375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.0001201629638671875, "grad_norm": 7.488863468170166, "learning_rate": 1e-05, "loss": 0.4937, "mean_token_accuracy": 0.8811390399932861, "num_tokens": 757129537.0, "step": 19936 }, { "epoch": 2.536191324259, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00011014938354492188, "grad_norm": 6.691540718078613, "learning_rate": 1e-05, "loss": 0.4763, "mean_token_accuracy": 0.88669753074646, "num_tokens": 757164892.0, "step": 19937 }, { "epoch": 2.5363185345375907, "ewc_loss": 0.140625, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00011157989501953125, "grad_norm": 7.11580753326416, "learning_rate": 1e-05, "loss": 0.5138, "mean_token_accuracy": 0.8694013357162476, "num_tokens": 757204717.0, "step": 19938 }, { "epoch": 2.5364457448161812, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.628092288970947, "learning_rate": 1e-05, "loss": 0.4933, "mean_token_accuracy": 0.8740439414978027, "num_tokens": 757239725.0, "step": 19939 }, { "epoch": 2.5365729550947718, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.822342395782471, "learning_rate": 1e-05, "loss": 0.5386, "mean_token_accuracy": 0.8582267761230469, "num_tokens": 757280356.0, "step": 19940 }, { "epoch": 2.5367001653733623, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.495243549346924, "learning_rate": 1e-05, "loss": 0.4965, "mean_token_accuracy": 0.8714865446090698, "num_tokens": 757316498.0, "step": 19941 }, { "epoch": 2.536827375651953, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.668112754821777, "learning_rate": 1e-05, "loss": 0.4972, "mean_token_accuracy": 0.8712402582168579, "num_tokens": 757351675.0, "step": 19942 }, { "epoch": 2.5369545859305433, "ewc_loss": 0.125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.345130920410156, "learning_rate": 1e-05, "loss": 0.487, "mean_token_accuracy": 0.8723084330558777, "num_tokens": 757393310.0, "step": 19943 }, { "epoch": 2.537081796209134, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.645486831665039, "learning_rate": 1e-05, "loss": 0.5034, "mean_token_accuracy": 0.8677005767822266, "num_tokens": 757429800.0, "step": 19944 }, { "epoch": 2.537209006487724, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.186215400695801, "learning_rate": 1e-05, "loss": 0.4325, "mean_token_accuracy": 0.8904343843460083, "num_tokens": 757463456.0, "step": 19945 }, { "epoch": 2.537336216766315, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.744381427764893, "learning_rate": 1e-05, "loss": 0.514, "mean_token_accuracy": 0.8644251823425293, "num_tokens": 757496977.0, "step": 19946 }, { "epoch": 2.537463427044905, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 6.008118629455566, "learning_rate": 1e-05, "loss": 0.4432, "mean_token_accuracy": 0.8839927911758423, "num_tokens": 757536003.0, "step": 19947 }, { "epoch": 2.537590637323496, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.966763019561768, "learning_rate": 1e-05, "loss": 0.5113, "mean_token_accuracy": 0.8688616752624512, "num_tokens": 757575017.0, "step": 19948 }, { "epoch": 2.537717847602086, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.8751912117004395, "learning_rate": 1e-05, "loss": 0.4643, "mean_token_accuracy": 0.8764575719833374, "num_tokens": 757617098.0, "step": 19949 }, { "epoch": 2.5378450578806766, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00011539459228515625, "grad_norm": 7.310093402862549, "learning_rate": 1e-05, "loss": 0.5255, "mean_token_accuracy": 0.8670367002487183, "num_tokens": 757652343.0, "step": 19950 }, { "epoch": 2.537972268159267, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 5.9130682945251465, "learning_rate": 1e-05, "loss": 0.4716, "mean_token_accuracy": 0.8768064975738525, "num_tokens": 757691021.0, "step": 19951 }, { "epoch": 2.5380994784378577, "ewc_loss": 0.1513671875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.632299423217773, "learning_rate": 1e-05, "loss": 0.4927, "mean_token_accuracy": 0.8809807300567627, "num_tokens": 757727805.0, "step": 19952 }, { "epoch": 2.538226688716448, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.228915691375732, "learning_rate": 1e-05, "loss": 0.5084, "mean_token_accuracy": 0.8684377670288086, "num_tokens": 757767979.0, "step": 19953 }, { "epoch": 2.5383538989950387, "ewc_loss": 0.15234375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.0001239776611328125, "grad_norm": 7.619754314422607, "learning_rate": 1e-05, "loss": 0.4933, "mean_token_accuracy": 0.8792949914932251, "num_tokens": 757807488.0, "step": 19954 }, { "epoch": 2.5384811092736292, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 6.612541198730469, "learning_rate": 1e-05, "loss": 0.5161, "mean_token_accuracy": 0.8661398887634277, "num_tokens": 757838298.0, "step": 19955 }, { "epoch": 2.5386083195522198, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00011301040649414062, "grad_norm": 7.199639797210693, "learning_rate": 1e-05, "loss": 0.5097, "mean_token_accuracy": 0.8715252876281738, "num_tokens": 757875016.0, "step": 19956 }, { "epoch": 2.5387355298308103, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00011014938354492188, "grad_norm": 6.802920341491699, "learning_rate": 1e-05, "loss": 0.4854, "mean_token_accuracy": 0.8796013593673706, "num_tokens": 757921306.0, "step": 19957 }, { "epoch": 2.538862740109401, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.617591381072998, "learning_rate": 1e-05, "loss": 0.4836, "mean_token_accuracy": 0.8801436424255371, "num_tokens": 757956555.0, "step": 19958 }, { "epoch": 2.5389899503879914, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 6.840517520904541, "learning_rate": 1e-05, "loss": 0.5265, "mean_token_accuracy": 0.8696688413619995, "num_tokens": 757990433.0, "step": 19959 }, { "epoch": 2.539117160666582, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.1436448097229, "learning_rate": 1e-05, "loss": 0.444, "mean_token_accuracy": 0.8873313069343567, "num_tokens": 758026606.0, "step": 19960 }, { "epoch": 2.5392443709451724, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00011110305786132812, "grad_norm": 7.065379619598389, "learning_rate": 1e-05, "loss": 0.4467, "mean_token_accuracy": 0.8912711143493652, "num_tokens": 758060995.0, "step": 19961 }, { "epoch": 2.539371581223763, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.887943267822266, "learning_rate": 1e-05, "loss": 0.4536, "mean_token_accuracy": 0.8815172910690308, "num_tokens": 758097479.0, "step": 19962 }, { "epoch": 2.5394987915023535, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.000118255615234375, "grad_norm": 7.415032386779785, "learning_rate": 1e-05, "loss": 0.5199, "mean_token_accuracy": 0.8759676814079285, "num_tokens": 758134541.0, "step": 19963 }, { "epoch": 2.539626001780944, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 5.917266845703125, "learning_rate": 1e-05, "loss": 0.4387, "mean_token_accuracy": 0.8865797519683838, "num_tokens": 758173885.0, "step": 19964 }, { "epoch": 2.5397532120595345, "ewc_loss": 0.15625, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.0001277923583984375, "grad_norm": 7.746370315551758, "learning_rate": 1e-05, "loss": 0.5417, "mean_token_accuracy": 0.8674114942550659, "num_tokens": 758219394.0, "step": 19965 }, { "epoch": 2.539880422338125, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.276968955993652, "learning_rate": 1e-05, "loss": 0.4904, "mean_token_accuracy": 0.8734482526779175, "num_tokens": 758261146.0, "step": 19966 }, { "epoch": 2.5400076326167156, "ewc_loss": 0.15625, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.0001277923583984375, "grad_norm": 7.723638534545898, "learning_rate": 1e-05, "loss": 0.4528, "mean_token_accuracy": 0.8945018649101257, "num_tokens": 758297935.0, "step": 19967 }, { "epoch": 2.5401348428953057, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00010967254638671875, "grad_norm": 6.636423587799072, "learning_rate": 1e-05, "loss": 0.5136, "mean_token_accuracy": 0.8698601722717285, "num_tokens": 758328579.0, "step": 19968 }, { "epoch": 2.5402620531738966, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.000118255615234375, "grad_norm": 7.393235206604004, "learning_rate": 1e-05, "loss": 0.5128, "mean_token_accuracy": 0.8741755485534668, "num_tokens": 758361731.0, "step": 19969 }, { "epoch": 2.5403892634524867, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 6.691863536834717, "learning_rate": 1e-05, "loss": 0.4833, "mean_token_accuracy": 0.8819337487220764, "num_tokens": 758394651.0, "step": 19970 }, { "epoch": 2.5405164737310777, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 7.0237932205200195, "learning_rate": 1e-05, "loss": 0.5004, "mean_token_accuracy": 0.8761947154998779, "num_tokens": 758436455.0, "step": 19971 }, { "epoch": 2.540643684009668, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.585757255554199, "learning_rate": 1e-05, "loss": 0.4794, "mean_token_accuracy": 0.8813445568084717, "num_tokens": 758471880.0, "step": 19972 }, { "epoch": 2.5407708942882588, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.772952556610107, "learning_rate": 1e-05, "loss": 0.4808, "mean_token_accuracy": 0.8772950172424316, "num_tokens": 758514354.0, "step": 19973 }, { "epoch": 2.540898104566849, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.449191093444824, "learning_rate": 1e-05, "loss": 0.4541, "mean_token_accuracy": 0.8831499814987183, "num_tokens": 758554263.0, "step": 19974 }, { "epoch": 2.5410253148454394, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.672547817230225, "learning_rate": 1e-05, "loss": 0.4887, "mean_token_accuracy": 0.8755178451538086, "num_tokens": 758588907.0, "step": 19975 }, { "epoch": 2.54115252512403, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.301151275634766, "learning_rate": 1e-05, "loss": 0.4942, "mean_token_accuracy": 0.8705103397369385, "num_tokens": 758618573.0, "step": 19976 }, { "epoch": 2.5412797354026204, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.678525924682617, "learning_rate": 1e-05, "loss": 0.4746, "mean_token_accuracy": 0.8783175349235535, "num_tokens": 758654315.0, "step": 19977 }, { "epoch": 2.541406945681211, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.1322197914123535, "learning_rate": 1e-05, "loss": 0.4906, "mean_token_accuracy": 0.8712266683578491, "num_tokens": 758690670.0, "step": 19978 }, { "epoch": 2.5415341559598015, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.804433345794678, "learning_rate": 1e-05, "loss": 0.4687, "mean_token_accuracy": 0.8781964778900146, "num_tokens": 758725686.0, "step": 19979 }, { "epoch": 2.541661366238392, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.963535308837891, "learning_rate": 1e-05, "loss": 0.4518, "mean_token_accuracy": 0.8816706538200378, "num_tokens": 758757002.0, "step": 19980 }, { "epoch": 2.5417885765169825, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 7.063426494598389, "learning_rate": 1e-05, "loss": 0.4706, "mean_token_accuracy": 0.8822453022003174, "num_tokens": 758793695.0, "step": 19981 }, { "epoch": 2.541915786795573, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.856861591339111, "learning_rate": 1e-05, "loss": 0.4728, "mean_token_accuracy": 0.8737701177597046, "num_tokens": 758837278.0, "step": 19982 }, { "epoch": 2.5420429970741636, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00011777877807617188, "grad_norm": 7.421741008758545, "learning_rate": 1e-05, "loss": 0.4986, "mean_token_accuracy": 0.8734206557273865, "num_tokens": 758871252.0, "step": 19983 }, { "epoch": 2.542170207352754, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 5.971912860870361, "learning_rate": 1e-05, "loss": 0.4744, "mean_token_accuracy": 0.8772079944610596, "num_tokens": 758912250.0, "step": 19984 }, { "epoch": 2.5422974176313446, "ewc_loss": 0.154296875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.000125885009765625, "grad_norm": 7.668797016143799, "learning_rate": 1e-05, "loss": 0.5093, "mean_token_accuracy": 0.8763426542282104, "num_tokens": 758951773.0, "step": 19985 }, { "epoch": 2.542424627909935, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.351351261138916, "learning_rate": 1e-05, "loss": 0.4632, "mean_token_accuracy": 0.8858672380447388, "num_tokens": 758989054.0, "step": 19986 }, { "epoch": 2.5425518381885257, "ewc_loss": 0.150390625, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.0001220703125, "grad_norm": 7.5200653076171875, "learning_rate": 1e-05, "loss": 0.478, "mean_token_accuracy": 0.8840089440345764, "num_tokens": 759026917.0, "step": 19987 }, { "epoch": 2.5426790484671162, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 6.659859657287598, "learning_rate": 1e-05, "loss": 0.4952, "mean_token_accuracy": 0.8770034313201904, "num_tokens": 759065752.0, "step": 19988 }, { "epoch": 2.5428062587457068, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00011110305786132812, "grad_norm": 7.041500568389893, "learning_rate": 1e-05, "loss": 0.4642, "mean_token_accuracy": 0.8862565755844116, "num_tokens": 759100650.0, "step": 19989 }, { "epoch": 2.5429334690242973, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 6.766191482543945, "learning_rate": 1e-05, "loss": 0.5218, "mean_token_accuracy": 0.866710901260376, "num_tokens": 759138274.0, "step": 19990 }, { "epoch": 2.543060679302888, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.515462398529053, "learning_rate": 1e-05, "loss": 0.5121, "mean_token_accuracy": 0.8660145998001099, "num_tokens": 759176060.0, "step": 19991 }, { "epoch": 2.5431878895814783, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.851068019866943, "learning_rate": 1e-05, "loss": 0.4667, "mean_token_accuracy": 0.8848700523376465, "num_tokens": 759208323.0, "step": 19992 }, { "epoch": 2.5433150998600684, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.08886194229126, "learning_rate": 1e-05, "loss": 0.443, "mean_token_accuracy": 0.890316367149353, "num_tokens": 759239180.0, "step": 19993 }, { "epoch": 2.5434423101386594, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00010967254638671875, "grad_norm": 7.071926116943359, "learning_rate": 1e-05, "loss": 0.4748, "mean_token_accuracy": 0.8821209073066711, "num_tokens": 759278489.0, "step": 19994 }, { "epoch": 2.5435695204172495, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.859004497528076, "learning_rate": 1e-05, "loss": 0.466, "mean_token_accuracy": 0.8753364682197571, "num_tokens": 759315254.0, "step": 19995 }, { "epoch": 2.5436967306958405, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00011777877807617188, "grad_norm": 7.448462009429932, "learning_rate": 1e-05, "loss": 0.4972, "mean_token_accuracy": 0.8799176216125488, "num_tokens": 759345969.0, "step": 19996 }, { "epoch": 2.5438239409744305, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 5.927450656890869, "learning_rate": 1e-05, "loss": 0.483, "mean_token_accuracy": 0.8734418749809265, "num_tokens": 759386530.0, "step": 19997 }, { "epoch": 2.5439511512530215, "ewc_loss": 0.1572265625, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00012874603271484375, "grad_norm": 7.7542195320129395, "learning_rate": 1e-05, "loss": 0.5274, "mean_token_accuracy": 0.8735260963439941, "num_tokens": 759428755.0, "step": 19998 }, { "epoch": 2.5440783615316116, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.322352409362793, "learning_rate": 1e-05, "loss": 0.4833, "mean_token_accuracy": 0.8751975297927856, "num_tokens": 759471335.0, "step": 19999 }, { "epoch": 2.544205571810202, "ewc_loss": 0.1533203125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00012493133544921875, "grad_norm": 7.691834926605225, "learning_rate": 1e-05, "loss": 0.5031, "mean_token_accuracy": 0.8775315284729004, "num_tokens": 759512541.0, "step": 20000 }, { "epoch": 2.5443327820887927, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 6.638903617858887, "learning_rate": 1e-05, "loss": 0.5087, "mean_token_accuracy": 0.8731995820999146, "num_tokens": 759551015.0, "step": 20001 }, { "epoch": 2.544459992367383, "ewc_loss": 0.1455078125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00011730194091796875, "grad_norm": 7.350640296936035, "learning_rate": 1e-05, "loss": 0.5276, "mean_token_accuracy": 0.8662067651748657, "num_tokens": 759588698.0, "step": 20002 }, { "epoch": 2.5445872026459737, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 6.680181980133057, "learning_rate": 1e-05, "loss": 0.5375, "mean_token_accuracy": 0.8607367277145386, "num_tokens": 759630180.0, "step": 20003 }, { "epoch": 2.5447144129245642, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 7.007904052734375, "learning_rate": 1e-05, "loss": 0.5531, "mean_token_accuracy": 0.8570242524147034, "num_tokens": 759664416.0, "step": 20004 }, { "epoch": 2.5448416232031548, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.586519718170166, "learning_rate": 1e-05, "loss": 0.5112, "mean_token_accuracy": 0.867398738861084, "num_tokens": 759708868.0, "step": 20005 }, { "epoch": 2.5449688334817453, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.753047466278076, "learning_rate": 1e-05, "loss": 0.4722, "mean_token_accuracy": 0.8796312808990479, "num_tokens": 759742431.0, "step": 20006 }, { "epoch": 2.545096043760336, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.446810722351074, "learning_rate": 1e-05, "loss": 0.4714, "mean_token_accuracy": 0.8807646036148071, "num_tokens": 759776504.0, "step": 20007 }, { "epoch": 2.5452232540389264, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.63962984085083, "learning_rate": 1e-05, "loss": 0.4415, "mean_token_accuracy": 0.8899332880973816, "num_tokens": 759808246.0, "step": 20008 }, { "epoch": 2.545350464317517, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.291415214538574, "learning_rate": 1e-05, "loss": 0.4331, "mean_token_accuracy": 0.8915213346481323, "num_tokens": 759846654.0, "step": 20009 }, { "epoch": 2.5454776745961074, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.65904426574707, "learning_rate": 1e-05, "loss": 0.4947, "mean_token_accuracy": 0.8733479976654053, "num_tokens": 759882594.0, "step": 20010 }, { "epoch": 2.545604884874698, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.133763790130615, "learning_rate": 1e-05, "loss": 0.4562, "mean_token_accuracy": 0.8828041553497314, "num_tokens": 759923557.0, "step": 20011 }, { "epoch": 2.5457320951532885, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.798455715179443, "learning_rate": 1e-05, "loss": 0.5046, "mean_token_accuracy": 0.8692406415939331, "num_tokens": 759961152.0, "step": 20012 }, { "epoch": 2.545859305431879, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.979406356811523, "learning_rate": 1e-05, "loss": 0.4663, "mean_token_accuracy": 0.8797611594200134, "num_tokens": 759995505.0, "step": 20013 }, { "epoch": 2.5459865157104695, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 7.04555606842041, "learning_rate": 1e-05, "loss": 0.4929, "mean_token_accuracy": 0.875952959060669, "num_tokens": 760033910.0, "step": 20014 }, { "epoch": 2.54611372598906, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.873818397521973, "learning_rate": 1e-05, "loss": 0.4248, "mean_token_accuracy": 0.8905670642852783, "num_tokens": 760067605.0, "step": 20015 }, { "epoch": 2.5462409362676506, "ewc_loss": 0.1455078125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00011682510375976562, "grad_norm": 7.392214775085449, "learning_rate": 1e-05, "loss": 0.5395, "mean_token_accuracy": 0.8615747690200806, "num_tokens": 760101151.0, "step": 20016 }, { "epoch": 2.546368146546241, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 5.965728282928467, "learning_rate": 1e-05, "loss": 0.511, "mean_token_accuracy": 0.8665310144424438, "num_tokens": 760140253.0, "step": 20017 }, { "epoch": 2.546495356824831, "ewc_loss": 0.1533203125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00012493133544921875, "grad_norm": 7.654801368713379, "learning_rate": 1e-05, "loss": 0.5076, "mean_token_accuracy": 0.8770686388015747, "num_tokens": 760175988.0, "step": 20018 }, { "epoch": 2.546622567103422, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.3116044998168945, "learning_rate": 1e-05, "loss": 0.4936, "mean_token_accuracy": 0.8726440668106079, "num_tokens": 760211734.0, "step": 20019 }, { "epoch": 2.5467497773820122, "ewc_loss": 0.150390625, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00012159347534179688, "grad_norm": 7.574073791503906, "learning_rate": 1e-05, "loss": 0.5046, "mean_token_accuracy": 0.8721646070480347, "num_tokens": 760248887.0, "step": 20020 }, { "epoch": 2.546876987660603, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 6.633967399597168, "learning_rate": 1e-05, "loss": 0.4965, "mean_token_accuracy": 0.8751891255378723, "num_tokens": 760284836.0, "step": 20021 }, { "epoch": 2.5470041979391933, "ewc_loss": 0.140625, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00011205673217773438, "grad_norm": 7.124981880187988, "learning_rate": 1e-05, "loss": 0.5053, "mean_token_accuracy": 0.8756521940231323, "num_tokens": 760327328.0, "step": 20022 }, { "epoch": 2.547131408217784, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 6.734625816345215, "learning_rate": 1e-05, "loss": 0.4803, "mean_token_accuracy": 0.8781037330627441, "num_tokens": 760362876.0, "step": 20023 }, { "epoch": 2.5472586184963744, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.604568958282471, "learning_rate": 1e-05, "loss": 0.4526, "mean_token_accuracy": 0.8827415704727173, "num_tokens": 760399652.0, "step": 20024 }, { "epoch": 2.547385828774965, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 6.821217060089111, "learning_rate": 1e-05, "loss": 0.4682, "mean_token_accuracy": 0.8841378092765808, "num_tokens": 760436346.0, "step": 20025 }, { "epoch": 2.5475130390535554, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.161588668823242, "learning_rate": 1e-05, "loss": 0.4808, "mean_token_accuracy": 0.8743999004364014, "num_tokens": 760468540.0, "step": 20026 }, { "epoch": 2.547640249332146, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 7.032942295074463, "learning_rate": 1e-05, "loss": 0.4875, "mean_token_accuracy": 0.8788673877716064, "num_tokens": 760504088.0, "step": 20027 }, { "epoch": 2.5477674596107365, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.894222736358643, "learning_rate": 1e-05, "loss": 0.4812, "mean_token_accuracy": 0.8742386102676392, "num_tokens": 760542543.0, "step": 20028 }, { "epoch": 2.547894669889327, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00011539459228515625, "grad_norm": 7.363208770751953, "learning_rate": 1e-05, "loss": 0.5231, "mean_token_accuracy": 0.8659726977348328, "num_tokens": 760582472.0, "step": 20029 }, { "epoch": 2.5480218801679175, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 5.913527011871338, "learning_rate": 1e-05, "loss": 0.4922, "mean_token_accuracy": 0.8705112934112549, "num_tokens": 760618410.0, "step": 20030 }, { "epoch": 2.548149090446508, "ewc_loss": 0.15234375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.0001239776611328125, "grad_norm": 7.681807041168213, "learning_rate": 1e-05, "loss": 0.4848, "mean_token_accuracy": 0.8816313147544861, "num_tokens": 760656223.0, "step": 20031 }, { "epoch": 2.5482763007250986, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.202251434326172, "learning_rate": 1e-05, "loss": 0.4348, "mean_token_accuracy": 0.8922432065010071, "num_tokens": 760700898.0, "step": 20032 }, { "epoch": 2.548403511003689, "ewc_loss": 0.154296875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.000125885009765625, "grad_norm": 7.745178699493408, "learning_rate": 1e-05, "loss": 0.5249, "mean_token_accuracy": 0.8758532404899597, "num_tokens": 760739608.0, "step": 20033 }, { "epoch": 2.5485307212822796, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 6.581088066101074, "learning_rate": 1e-05, "loss": 0.5259, "mean_token_accuracy": 0.86607825756073, "num_tokens": 760777808.0, "step": 20034 }, { "epoch": 2.54865793156087, "ewc_loss": 0.1484375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00011968612670898438, "grad_norm": 7.4348955154418945, "learning_rate": 1e-05, "loss": 0.4864, "mean_token_accuracy": 0.8800209760665894, "num_tokens": 760817497.0, "step": 20035 }, { "epoch": 2.5487851418394607, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 6.69002628326416, "learning_rate": 1e-05, "loss": 0.5046, "mean_token_accuracy": 0.871982991695404, "num_tokens": 760855699.0, "step": 20036 }, { "epoch": 2.5489123521180512, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00011014938354492188, "grad_norm": 7.06538200378418, "learning_rate": 1e-05, "loss": 0.538, "mean_token_accuracy": 0.8597366809844971, "num_tokens": 760890796.0, "step": 20037 }, { "epoch": 2.5490395623966418, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.6101531982421875, "learning_rate": 1e-05, "loss": 0.4896, "mean_token_accuracy": 0.8772037029266357, "num_tokens": 760929056.0, "step": 20038 }, { "epoch": 2.5491667726752323, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.76209020614624, "learning_rate": 1e-05, "loss": 0.4422, "mean_token_accuracy": 0.8884939551353455, "num_tokens": 760963941.0, "step": 20039 }, { "epoch": 2.549293982953823, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.482554912567139, "learning_rate": 1e-05, "loss": 0.4806, "mean_token_accuracy": 0.8762873411178589, "num_tokens": 761009703.0, "step": 20040 }, { "epoch": 2.5494211932324133, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.6138129234313965, "learning_rate": 1e-05, "loss": 0.4711, "mean_token_accuracy": 0.8793599009513855, "num_tokens": 761051408.0, "step": 20041 }, { "epoch": 2.549548403511004, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.355078220367432, "learning_rate": 1e-05, "loss": 0.5042, "mean_token_accuracy": 0.8688492774963379, "num_tokens": 761093764.0, "step": 20042 }, { "epoch": 2.549675613789594, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.591158390045166, "learning_rate": 1e-05, "loss": 0.4796, "mean_token_accuracy": 0.878635823726654, "num_tokens": 761129975.0, "step": 20043 }, { "epoch": 2.549802824068185, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.200141429901123, "learning_rate": 1e-05, "loss": 0.4964, "mean_token_accuracy": 0.8691892623901367, "num_tokens": 761169845.0, "step": 20044 }, { "epoch": 2.549930034346775, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.678086757659912, "learning_rate": 1e-05, "loss": 0.5098, "mean_token_accuracy": 0.874022364616394, "num_tokens": 761207667.0, "step": 20045 }, { "epoch": 2.550057244625366, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 6.023375034332275, "learning_rate": 1e-05, "loss": 0.471, "mean_token_accuracy": 0.8776150941848755, "num_tokens": 761249649.0, "step": 20046 }, { "epoch": 2.550184454903956, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.893368244171143, "learning_rate": 1e-05, "loss": 0.5921, "mean_token_accuracy": 0.8455716967582703, "num_tokens": 761289879.0, "step": 20047 }, { "epoch": 2.5503116651825466, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.880233287811279, "learning_rate": 1e-05, "loss": 0.4886, "mean_token_accuracy": 0.8694960474967957, "num_tokens": 761331398.0, "step": 20048 }, { "epoch": 2.550438875461137, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00011301040649414062, "grad_norm": 7.230588436126709, "learning_rate": 1e-05, "loss": 0.4861, "mean_token_accuracy": 0.8786181211471558, "num_tokens": 761363928.0, "step": 20049 }, { "epoch": 2.5505660857397277, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 5.838636875152588, "learning_rate": 1e-05, "loss": 0.5313, "mean_token_accuracy": 0.8595796823501587, "num_tokens": 761405737.0, "step": 20050 }, { "epoch": 2.550693296018318, "ewc_loss": 0.150390625, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.0001220703125, "grad_norm": 7.644257545471191, "learning_rate": 1e-05, "loss": 0.4784, "mean_token_accuracy": 0.8842592239379883, "num_tokens": 761435934.0, "step": 20051 }, { "epoch": 2.5508205062969087, "ewc_loss": 0.125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.113667964935303, "learning_rate": 1e-05, "loss": 0.464, "mean_token_accuracy": 0.8808915615081787, "num_tokens": 761466700.0, "step": 20052 }, { "epoch": 2.5509477165754992, "ewc_loss": 0.15625, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.0001277923583984375, "grad_norm": 7.783134460449219, "learning_rate": 1e-05, "loss": 0.5048, "mean_token_accuracy": 0.8781119585037231, "num_tokens": 761505682.0, "step": 20053 }, { "epoch": 2.5510749268540898, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 6.557011127471924, "learning_rate": 1e-05, "loss": 0.491, "mean_token_accuracy": 0.8764358758926392, "num_tokens": 761544071.0, "step": 20054 }, { "epoch": 2.5512021371326803, "ewc_loss": 0.1494140625, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00012063980102539062, "grad_norm": 7.461209774017334, "learning_rate": 1e-05, "loss": 0.4805, "mean_token_accuracy": 0.884975254535675, "num_tokens": 761586595.0, "step": 20055 }, { "epoch": 2.551329347411271, "ewc_loss": 0.140625, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.0001125335693359375, "grad_norm": 6.770524978637695, "learning_rate": 1e-05, "loss": 0.4947, "mean_token_accuracy": 0.8766616582870483, "num_tokens": 761623924.0, "step": 20056 }, { "epoch": 2.5514565576898613, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.936095714569092, "learning_rate": 1e-05, "loss": 0.4974, "mean_token_accuracy": 0.872907280921936, "num_tokens": 761660952.0, "step": 20057 }, { "epoch": 2.551583767968452, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 6.785078525543213, "learning_rate": 1e-05, "loss": 0.4577, "mean_token_accuracy": 0.8890775442123413, "num_tokens": 761701451.0, "step": 20058 }, { "epoch": 2.5517109782470424, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.466087818145752, "learning_rate": 1e-05, "loss": 0.4781, "mean_token_accuracy": 0.8766980171203613, "num_tokens": 761733638.0, "step": 20059 }, { "epoch": 2.551838188525633, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.865230560302734, "learning_rate": 1e-05, "loss": 0.4835, "mean_token_accuracy": 0.8796017169952393, "num_tokens": 761763978.0, "step": 20060 }, { "epoch": 2.5519653988042235, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.102829933166504, "learning_rate": 1e-05, "loss": 0.4553, "mean_token_accuracy": 0.8843986988067627, "num_tokens": 761802653.0, "step": 20061 }, { "epoch": 2.552092609082814, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 7.04171085357666, "learning_rate": 1e-05, "loss": 0.4849, "mean_token_accuracy": 0.8791295886039734, "num_tokens": 761840671.0, "step": 20062 }, { "epoch": 2.5522198193614045, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 5.936031341552734, "learning_rate": 1e-05, "loss": 0.4911, "mean_token_accuracy": 0.8709883689880371, "num_tokens": 761883823.0, "step": 20063 }, { "epoch": 2.552347029639995, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00011491775512695312, "grad_norm": 7.315093040466309, "learning_rate": 1e-05, "loss": 0.4991, "mean_token_accuracy": 0.8776342868804932, "num_tokens": 761921597.0, "step": 20064 }, { "epoch": 2.5524742399185856, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 5.937082290649414, "learning_rate": 1e-05, "loss": 0.4439, "mean_token_accuracy": 0.8838729858398438, "num_tokens": 761963400.0, "step": 20065 }, { "epoch": 2.5526014501971757, "ewc_loss": 0.150390625, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.0001220703125, "grad_norm": 7.591197967529297, "learning_rate": 1e-05, "loss": 0.5125, "mean_token_accuracy": 0.8753771781921387, "num_tokens": 762004442.0, "step": 20066 }, { "epoch": 2.5527286604757666, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.196496963500977, "learning_rate": 1e-05, "loss": 0.5222, "mean_token_accuracy": 0.8621490597724915, "num_tokens": 762040679.0, "step": 20067 }, { "epoch": 2.5528558707543567, "ewc_loss": 0.1513671875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.621397972106934, "learning_rate": 1e-05, "loss": 0.513, "mean_token_accuracy": 0.8737430572509766, "num_tokens": 762085249.0, "step": 20068 }, { "epoch": 2.5529830810329477, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.4607672691345215, "learning_rate": 1e-05, "loss": 0.4428, "mean_token_accuracy": 0.8913571834564209, "num_tokens": 762125551.0, "step": 20069 }, { "epoch": 2.5531102913115378, "ewc_loss": 0.1474609375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00011873245239257812, "grad_norm": 7.414653778076172, "learning_rate": 1e-05, "loss": 0.4966, "mean_token_accuracy": 0.881657600402832, "num_tokens": 762162503.0, "step": 20070 }, { "epoch": 2.5532375015901287, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.589287757873535, "learning_rate": 1e-05, "loss": 0.472, "mean_token_accuracy": 0.8820586204528809, "num_tokens": 762197302.0, "step": 20071 }, { "epoch": 2.553364711868719, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00011014938354492188, "grad_norm": 7.077907085418701, "learning_rate": 1e-05, "loss": 0.4818, "mean_token_accuracy": 0.8825143575668335, "num_tokens": 762235069.0, "step": 20072 }, { "epoch": 2.5534919221473094, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.553905487060547, "learning_rate": 1e-05, "loss": 0.4394, "mean_token_accuracy": 0.8890659809112549, "num_tokens": 762272014.0, "step": 20073 }, { "epoch": 2.5536191324259, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.799466133117676, "learning_rate": 1e-05, "loss": 0.4422, "mean_token_accuracy": 0.8886653184890747, "num_tokens": 762312840.0, "step": 20074 }, { "epoch": 2.5537463427044904, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.47735071182251, "learning_rate": 1e-05, "loss": 0.4846, "mean_token_accuracy": 0.8740308880805969, "num_tokens": 762347997.0, "step": 20075 }, { "epoch": 2.553873552983081, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.638427257537842, "learning_rate": 1e-05, "loss": 0.508, "mean_token_accuracy": 0.871507465839386, "num_tokens": 762390706.0, "step": 20076 }, { "epoch": 2.5540007632616715, "ewc_loss": 0.125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.360755443572998, "learning_rate": 1e-05, "loss": 0.4861, "mean_token_accuracy": 0.8753396272659302, "num_tokens": 762425894.0, "step": 20077 }, { "epoch": 2.554127973540262, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.584178447723389, "learning_rate": 1e-05, "loss": 0.516, "mean_token_accuracy": 0.8616828322410583, "num_tokens": 762461962.0, "step": 20078 }, { "epoch": 2.5542551838188525, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.2230753898620605, "learning_rate": 1e-05, "loss": 0.4757, "mean_token_accuracy": 0.8744131922721863, "num_tokens": 762500128.0, "step": 20079 }, { "epoch": 2.554382394097443, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.632710933685303, "learning_rate": 1e-05, "loss": 0.4293, "mean_token_accuracy": 0.8946155309677124, "num_tokens": 762536256.0, "step": 20080 }, { "epoch": 2.5545096043760336, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.093852519989014, "learning_rate": 1e-05, "loss": 0.5433, "mean_token_accuracy": 0.8566763997077942, "num_tokens": 762569797.0, "step": 20081 }, { "epoch": 2.554636814654624, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.764062404632568, "learning_rate": 1e-05, "loss": 0.4885, "mean_token_accuracy": 0.8745900392532349, "num_tokens": 762606849.0, "step": 20082 }, { "epoch": 2.5547640249332146, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.936706066131592, "learning_rate": 1e-05, "loss": 0.5034, "mean_token_accuracy": 0.8672693967819214, "num_tokens": 762649526.0, "step": 20083 }, { "epoch": 2.554891235211805, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 7.031276702880859, "learning_rate": 1e-05, "loss": 0.4729, "mean_token_accuracy": 0.8789374828338623, "num_tokens": 762686240.0, "step": 20084 }, { "epoch": 2.5550184454903957, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.838221073150635, "learning_rate": 1e-05, "loss": 0.4931, "mean_token_accuracy": 0.8671082258224487, "num_tokens": 762727908.0, "step": 20085 }, { "epoch": 2.5551456557689862, "ewc_loss": 0.1474609375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00011873245239257812, "grad_norm": 7.402528285980225, "learning_rate": 1e-05, "loss": 0.457, "mean_token_accuracy": 0.8906520009040833, "num_tokens": 762765966.0, "step": 20086 }, { "epoch": 2.5552728660475768, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 5.940232753753662, "learning_rate": 1e-05, "loss": 0.4739, "mean_token_accuracy": 0.8762814998626709, "num_tokens": 762806031.0, "step": 20087 }, { "epoch": 2.5554000763261673, "ewc_loss": 0.154296875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.000125885009765625, "grad_norm": 7.699701309204102, "learning_rate": 1e-05, "loss": 0.4762, "mean_token_accuracy": 0.885884165763855, "num_tokens": 762847531.0, "step": 20088 }, { "epoch": 2.555527286604758, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.3079328536987305, "learning_rate": 1e-05, "loss": 0.4743, "mean_token_accuracy": 0.8787810802459717, "num_tokens": 762878409.0, "step": 20089 }, { "epoch": 2.5556544968833483, "ewc_loss": 0.15234375, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.0001239776611328125, "grad_norm": 7.618247032165527, "learning_rate": 1e-05, "loss": 0.5039, "mean_token_accuracy": 0.8766847848892212, "num_tokens": 762918964.0, "step": 20090 }, { "epoch": 2.5557817071619384, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 6.630515098571777, "learning_rate": 1e-05, "loss": 0.461, "mean_token_accuracy": 0.8854738473892212, "num_tokens": 762958962.0, "step": 20091 }, { "epoch": 2.5559089174405294, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00011301040649414062, "grad_norm": 7.197669506072998, "learning_rate": 1e-05, "loss": 0.5407, "mean_token_accuracy": 0.8630847334861755, "num_tokens": 762996062.0, "step": 20092 }, { "epoch": 2.5560361277191195, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00010967254638671875, "grad_norm": 6.734129905700684, "learning_rate": 1e-05, "loss": 0.4919, "mean_token_accuracy": 0.878318190574646, "num_tokens": 763033387.0, "step": 20093 }, { "epoch": 2.5561633379977104, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.702788829803467, "learning_rate": 1e-05, "loss": 0.4958, "mean_token_accuracy": 0.8722738027572632, "num_tokens": 763067826.0, "step": 20094 }, { "epoch": 2.5562905482763005, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.784940242767334, "learning_rate": 1e-05, "loss": 0.4568, "mean_token_accuracy": 0.884463369846344, "num_tokens": 763100442.0, "step": 20095 }, { "epoch": 2.5564177585548915, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.258498191833496, "learning_rate": 1e-05, "loss": 0.4295, "mean_token_accuracy": 0.8896543383598328, "num_tokens": 763137659.0, "step": 20096 }, { "epoch": 2.5565449688334816, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 6.938415050506592, "learning_rate": 1e-05, "loss": 0.4734, "mean_token_accuracy": 0.886147141456604, "num_tokens": 763180000.0, "step": 20097 }, { "epoch": 2.556672179112072, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 5.983007431030273, "learning_rate": 1e-05, "loss": 0.5162, "mean_token_accuracy": 0.8648068904876709, "num_tokens": 763218863.0, "step": 20098 }, { "epoch": 2.5567993893906626, "ewc_loss": 0.140625, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00011205673217773438, "grad_norm": 7.201225757598877, "learning_rate": 1e-05, "loss": 0.4553, "mean_token_accuracy": 0.8877041339874268, "num_tokens": 763252973.0, "step": 20099 }, { "epoch": 2.556926599669253, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 5.9094696044921875, "learning_rate": 1e-05, "loss": 0.5374, "mean_token_accuracy": 0.8587212562561035, "num_tokens": 763288781.0, "step": 20100 }, { "epoch": 2.5570538099478437, "ewc_loss": 0.1484375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00011968612670898438, "grad_norm": 7.500627040863037, "learning_rate": 1e-05, "loss": 0.5106, "mean_token_accuracy": 0.8704121112823486, "num_tokens": 763321839.0, "step": 20101 }, { "epoch": 2.5571810202264342, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.058785915374756, "learning_rate": 1e-05, "loss": 0.4834, "mean_token_accuracy": 0.8734772205352783, "num_tokens": 763357607.0, "step": 20102 }, { "epoch": 2.5573082305050248, "ewc_loss": 0.1533203125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00012493133544921875, "grad_norm": 7.650961399078369, "learning_rate": 1e-05, "loss": 0.5268, "mean_token_accuracy": 0.8724299669265747, "num_tokens": 763399489.0, "step": 20103 }, { "epoch": 2.5574354407836153, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.363067626953125, "learning_rate": 1e-05, "loss": 0.5487, "mean_token_accuracy": 0.8596765995025635, "num_tokens": 763441256.0, "step": 20104 }, { "epoch": 2.557562651062206, "ewc_loss": 0.150390625, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00012159347534179688, "grad_norm": 7.563274383544922, "learning_rate": 1e-05, "loss": 0.5256, "mean_token_accuracy": 0.8673601150512695, "num_tokens": 763475019.0, "step": 20105 }, { "epoch": 2.5576898613407963, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 6.570430278778076, "learning_rate": 1e-05, "loss": 0.4881, "mean_token_accuracy": 0.8767006993293762, "num_tokens": 763510456.0, "step": 20106 }, { "epoch": 2.557817071619387, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00011491775512695312, "grad_norm": 7.2445878982543945, "learning_rate": 1e-05, "loss": 0.5149, "mean_token_accuracy": 0.8723684549331665, "num_tokens": 763546181.0, "step": 20107 }, { "epoch": 2.5579442818979774, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.588150978088379, "learning_rate": 1e-05, "loss": 0.4761, "mean_token_accuracy": 0.8797073364257812, "num_tokens": 763578763.0, "step": 20108 }, { "epoch": 2.558071492176568, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.931448459625244, "learning_rate": 1e-05, "loss": 0.5232, "mean_token_accuracy": 0.8654905557632446, "num_tokens": 763615615.0, "step": 20109 }, { "epoch": 2.5581987024551585, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.525792121887207, "learning_rate": 1e-05, "loss": 0.4578, "mean_token_accuracy": 0.888149619102478, "num_tokens": 763655546.0, "step": 20110 }, { "epoch": 2.558325912733749, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.701841354370117, "learning_rate": 1e-05, "loss": 0.503, "mean_token_accuracy": 0.8703508973121643, "num_tokens": 763695400.0, "step": 20111 }, { "epoch": 2.5584531230123395, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.433786869049072, "learning_rate": 1e-05, "loss": 0.5368, "mean_token_accuracy": 0.8577686548233032, "num_tokens": 763725968.0, "step": 20112 }, { "epoch": 2.55858033329093, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.583828449249268, "learning_rate": 1e-05, "loss": 0.4727, "mean_token_accuracy": 0.8790401220321655, "num_tokens": 763763708.0, "step": 20113 }, { "epoch": 2.5587075435695206, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.297924995422363, "learning_rate": 1e-05, "loss": 0.5102, "mean_token_accuracy": 0.8637802004814148, "num_tokens": 763800836.0, "step": 20114 }, { "epoch": 2.558834753848111, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.5811686515808105, "learning_rate": 1e-05, "loss": 0.5052, "mean_token_accuracy": 0.8696009516716003, "num_tokens": 763842119.0, "step": 20115 }, { "epoch": 2.558961964126701, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.1636738777160645, "learning_rate": 1e-05, "loss": 0.4752, "mean_token_accuracy": 0.875266432762146, "num_tokens": 763876797.0, "step": 20116 }, { "epoch": 2.559089174405292, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.678674697875977, "learning_rate": 1e-05, "loss": 0.4967, "mean_token_accuracy": 0.873318076133728, "num_tokens": 763913131.0, "step": 20117 }, { "epoch": 2.5592163846838822, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 6.012412071228027, "learning_rate": 1e-05, "loss": 0.4635, "mean_token_accuracy": 0.8774653673171997, "num_tokens": 763952430.0, "step": 20118 }, { "epoch": 2.559343594962473, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.8730692863464355, "learning_rate": 1e-05, "loss": 0.4896, "mean_token_accuracy": 0.8781861662864685, "num_tokens": 763988978.0, "step": 20119 }, { "epoch": 2.5594708052410633, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.873326301574707, "learning_rate": 1e-05, "loss": 0.4963, "mean_token_accuracy": 0.8739252090454102, "num_tokens": 764027473.0, "step": 20120 }, { "epoch": 2.559598015519654, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00011301040649414062, "grad_norm": 7.209643840789795, "learning_rate": 1e-05, "loss": 0.4818, "mean_token_accuracy": 0.8824599981307983, "num_tokens": 764063681.0, "step": 20121 }, { "epoch": 2.5597252257982444, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 5.836422443389893, "learning_rate": 1e-05, "loss": 0.4724, "mean_token_accuracy": 0.8792533874511719, "num_tokens": 764102463.0, "step": 20122 }, { "epoch": 2.559852436076835, "ewc_loss": 0.15234375, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.6082377433776855, "learning_rate": 1e-05, "loss": 0.485, "mean_token_accuracy": 0.8857762813568115, "num_tokens": 764141912.0, "step": 20123 }, { "epoch": 2.5599796463554254, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.072902202606201, "learning_rate": 1e-05, "loss": 0.4418, "mean_token_accuracy": 0.8883748054504395, "num_tokens": 764176448.0, "step": 20124 }, { "epoch": 2.560106856634016, "ewc_loss": 0.1552734375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00012683868408203125, "grad_norm": 7.770924091339111, "learning_rate": 1e-05, "loss": 0.5584, "mean_token_accuracy": 0.8607592582702637, "num_tokens": 764217287.0, "step": 20125 }, { "epoch": 2.5602340669126065, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 6.526263236999512, "learning_rate": 1e-05, "loss": 0.5004, "mean_token_accuracy": 0.8747446537017822, "num_tokens": 764255593.0, "step": 20126 }, { "epoch": 2.560361277191197, "ewc_loss": 0.1494140625, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00012063980102539062, "grad_norm": 7.475515842437744, "learning_rate": 1e-05, "loss": 0.5165, "mean_token_accuracy": 0.8731139898300171, "num_tokens": 764294624.0, "step": 20127 }, { "epoch": 2.5604884874697875, "ewc_loss": 0.140625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.0001125335693359375, "grad_norm": 6.745428562164307, "learning_rate": 1e-05, "loss": 0.4589, "mean_token_accuracy": 0.8868807554244995, "num_tokens": 764328495.0, "step": 20128 }, { "epoch": 2.560615697748378, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 6.962871551513672, "learning_rate": 1e-05, "loss": 0.4598, "mean_token_accuracy": 0.8835440874099731, "num_tokens": 764359986.0, "step": 20129 }, { "epoch": 2.5607429080269686, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 6.779574871063232, "learning_rate": 1e-05, "loss": 0.5179, "mean_token_accuracy": 0.869070291519165, "num_tokens": 764394039.0, "step": 20130 }, { "epoch": 2.560870118305559, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.472752094268799, "learning_rate": 1e-05, "loss": 0.5172, "mean_token_accuracy": 0.868325412273407, "num_tokens": 764436106.0, "step": 20131 }, { "epoch": 2.5609973285841496, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.845255374908447, "learning_rate": 1e-05, "loss": 0.4711, "mean_token_accuracy": 0.8794938325881958, "num_tokens": 764470004.0, "step": 20132 }, { "epoch": 2.56112453886274, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.1172003746032715, "learning_rate": 1e-05, "loss": 0.4482, "mean_token_accuracy": 0.8844000697135925, "num_tokens": 764507679.0, "step": 20133 }, { "epoch": 2.5612517491413307, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 7.032780170440674, "learning_rate": 1e-05, "loss": 0.459, "mean_token_accuracy": 0.8902066349983215, "num_tokens": 764548818.0, "step": 20134 }, { "epoch": 2.561378959419921, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 5.943356037139893, "learning_rate": 1e-05, "loss": 0.5012, "mean_token_accuracy": 0.8724688291549683, "num_tokens": 764590289.0, "step": 20135 }, { "epoch": 2.5615061696985117, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00011396408081054688, "grad_norm": 7.31061315536499, "learning_rate": 1e-05, "loss": 0.4675, "mean_token_accuracy": 0.8830201625823975, "num_tokens": 764624210.0, "step": 20136 }, { "epoch": 2.5616333799771023, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 5.968698024749756, "learning_rate": 1e-05, "loss": 0.4886, "mean_token_accuracy": 0.8746069669723511, "num_tokens": 764657071.0, "step": 20137 }, { "epoch": 2.561760590255693, "ewc_loss": 0.1494140625, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00012063980102539062, "grad_norm": 7.5335564613342285, "learning_rate": 1e-05, "loss": 0.4939, "mean_token_accuracy": 0.8775038123130798, "num_tokens": 764691439.0, "step": 20138 }, { "epoch": 2.5618878005342833, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.173688888549805, "learning_rate": 1e-05, "loss": 0.4275, "mean_token_accuracy": 0.8922145366668701, "num_tokens": 764731551.0, "step": 20139 }, { "epoch": 2.562015010812874, "ewc_loss": 0.150390625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.0001220703125, "grad_norm": 7.573554515838623, "learning_rate": 1e-05, "loss": 0.4643, "mean_token_accuracy": 0.8927861452102661, "num_tokens": 764763233.0, "step": 20140 }, { "epoch": 2.562142221091464, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.415966033935547, "learning_rate": 1e-05, "loss": 0.4197, "mean_token_accuracy": 0.8968651294708252, "num_tokens": 764796937.0, "step": 20141 }, { "epoch": 2.562269431370055, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.000118255615234375, "grad_norm": 7.383368015289307, "learning_rate": 1e-05, "loss": 0.482, "mean_token_accuracy": 0.8790397644042969, "num_tokens": 764831808.0, "step": 20142 }, { "epoch": 2.562396641648645, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.534074306488037, "learning_rate": 1e-05, "loss": 0.4862, "mean_token_accuracy": 0.8768115639686584, "num_tokens": 764863288.0, "step": 20143 }, { "epoch": 2.562523851927236, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00011110305786132812, "grad_norm": 7.090076923370361, "learning_rate": 1e-05, "loss": 0.4523, "mean_token_accuracy": 0.888141930103302, "num_tokens": 764905417.0, "step": 20144 }, { "epoch": 2.562651062205826, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.511269569396973, "learning_rate": 1e-05, "loss": 0.4743, "mean_token_accuracy": 0.8797807693481445, "num_tokens": 764942058.0, "step": 20145 }, { "epoch": 2.5627782724844166, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.8295488357543945, "learning_rate": 1e-05, "loss": 0.4978, "mean_token_accuracy": 0.8732199668884277, "num_tokens": 764983652.0, "step": 20146 }, { "epoch": 2.562905482763007, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.452536106109619, "learning_rate": 1e-05, "loss": 0.4739, "mean_token_accuracy": 0.8802993297576904, "num_tokens": 765023966.0, "step": 20147 }, { "epoch": 2.5630326930415976, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.650832653045654, "learning_rate": 1e-05, "loss": 0.4835, "mean_token_accuracy": 0.872455358505249, "num_tokens": 765055847.0, "step": 20148 }, { "epoch": 2.563159903320188, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.685728549957275, "learning_rate": 1e-05, "loss": 0.5127, "mean_token_accuracy": 0.8673405647277832, "num_tokens": 765094387.0, "step": 20149 }, { "epoch": 2.5632871135987787, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.362894535064697, "learning_rate": 1e-05, "loss": 0.5048, "mean_token_accuracy": 0.8686668872833252, "num_tokens": 765128193.0, "step": 20150 }, { "epoch": 2.5634143238773692, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.451385498046875, "learning_rate": 1e-05, "loss": 0.4831, "mean_token_accuracy": 0.8750157356262207, "num_tokens": 765166550.0, "step": 20151 }, { "epoch": 2.5635415341559598, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.24135684967041, "learning_rate": 1e-05, "loss": 0.4829, "mean_token_accuracy": 0.8746541738510132, "num_tokens": 765204228.0, "step": 20152 }, { "epoch": 2.5636687444345503, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.404451847076416, "learning_rate": 1e-05, "loss": 0.4604, "mean_token_accuracy": 0.8807506561279297, "num_tokens": 765247148.0, "step": 20153 }, { "epoch": 2.563795954713141, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.217280387878418, "learning_rate": 1e-05, "loss": 0.4194, "mean_token_accuracy": 0.8895807862281799, "num_tokens": 765287902.0, "step": 20154 }, { "epoch": 2.5639231649917313, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.363605499267578, "learning_rate": 1e-05, "loss": 0.4677, "mean_token_accuracy": 0.8831510543823242, "num_tokens": 765331118.0, "step": 20155 }, { "epoch": 2.564050375270322, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.211723804473877, "learning_rate": 1e-05, "loss": 0.4521, "mean_token_accuracy": 0.8849847912788391, "num_tokens": 765369942.0, "step": 20156 }, { "epoch": 2.5641775855489124, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.342383861541748, "learning_rate": 1e-05, "loss": 0.4953, "mean_token_accuracy": 0.8775888085365295, "num_tokens": 765407664.0, "step": 20157 }, { "epoch": 2.564304795827503, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.188786506652832, "learning_rate": 1e-05, "loss": 0.4965, "mean_token_accuracy": 0.8762763738632202, "num_tokens": 765453303.0, "step": 20158 }, { "epoch": 2.5644320061060935, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.445490837097168, "learning_rate": 1e-05, "loss": 0.4795, "mean_token_accuracy": 0.8777066469192505, "num_tokens": 765492558.0, "step": 20159 }, { "epoch": 2.564559216384684, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.081699848175049, "learning_rate": 1e-05, "loss": 0.5402, "mean_token_accuracy": 0.8537117838859558, "num_tokens": 765529964.0, "step": 20160 }, { "epoch": 2.5646864266632745, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.496978759765625, "learning_rate": 1e-05, "loss": 0.4411, "mean_token_accuracy": 0.8897688984870911, "num_tokens": 765570158.0, "step": 20161 }, { "epoch": 2.564813636941865, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 5.946337699890137, "learning_rate": 1e-05, "loss": 0.5309, "mean_token_accuracy": 0.8555667996406555, "num_tokens": 765609304.0, "step": 20162 }, { "epoch": 2.5649408472204556, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.711526870727539, "learning_rate": 1e-05, "loss": 0.4853, "mean_token_accuracy": 0.8751745820045471, "num_tokens": 765651124.0, "step": 20163 }, { "epoch": 2.5650680574990457, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.736894607543945, "learning_rate": 1e-05, "loss": 0.4265, "mean_token_accuracy": 0.8903692960739136, "num_tokens": 765685544.0, "step": 20164 }, { "epoch": 2.5651952677776366, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00011396408081054688, "grad_norm": 7.129739284515381, "learning_rate": 1e-05, "loss": 0.4657, "mean_token_accuracy": 0.887974739074707, "num_tokens": 765726120.0, "step": 20165 }, { "epoch": 2.5653224780562267, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.551828384399414, "learning_rate": 1e-05, "loss": 0.5007, "mean_token_accuracy": 0.8637923002243042, "num_tokens": 765764145.0, "step": 20166 }, { "epoch": 2.5654496883348177, "ewc_loss": 0.16015625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.0001316070556640625, "grad_norm": 7.855151176452637, "learning_rate": 1e-05, "loss": 0.4679, "mean_token_accuracy": 0.8886832594871521, "num_tokens": 765795530.0, "step": 20167 }, { "epoch": 2.5655768986134078, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 5.754027366638184, "learning_rate": 1e-05, "loss": 0.4866, "mean_token_accuracy": 0.8728457689285278, "num_tokens": 765831793.0, "step": 20168 }, { "epoch": 2.5657041088919987, "ewc_loss": 0.177734375, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.000148773193359375, "grad_norm": 8.572762489318848, "learning_rate": 1e-05, "loss": 0.5455, "mean_token_accuracy": 0.8694374561309814, "num_tokens": 765876724.0, "step": 20169 }, { "epoch": 2.565831319170589, "ewc_loss": 0.154296875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00012493133544921875, "grad_norm": 6.81116247177124, "learning_rate": 1e-05, "loss": 0.49, "mean_token_accuracy": 0.8902183771133423, "num_tokens": 765917604.0, "step": 20170 }, { "epoch": 2.5659585294491793, "ewc_loss": 0.169921875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00014019012451171875, "grad_norm": 8.316350936889648, "learning_rate": 1e-05, "loss": 0.5154, "mean_token_accuracy": 0.878082275390625, "num_tokens": 765953717.0, "step": 20171 }, { "epoch": 2.56608573972777, "ewc_loss": 0.16015625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.0001316070556640625, "grad_norm": 7.377532005310059, "learning_rate": 1e-05, "loss": 0.5193, "mean_token_accuracy": 0.8743127584457397, "num_tokens": 765989967.0, "step": 20172 }, { "epoch": 2.5662129500063604, "ewc_loss": 0.154296875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00012493133544921875, "grad_norm": 7.661014080047607, "learning_rate": 1e-05, "loss": 0.5027, "mean_token_accuracy": 0.8811874985694885, "num_tokens": 766029276.0, "step": 20173 }, { "epoch": 2.566340160284951, "ewc_loss": 0.150390625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.0001220703125, "grad_norm": 7.130598545074463, "learning_rate": 1e-05, "loss": 0.4811, "mean_token_accuracy": 0.8839402794837952, "num_tokens": 766064819.0, "step": 20174 }, { "epoch": 2.5664673705635415, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00011587142944335938, "grad_norm": 7.274723529815674, "learning_rate": 1e-05, "loss": 0.5069, "mean_token_accuracy": 0.8710330724716187, "num_tokens": 766103621.0, "step": 20175 }, { "epoch": 2.566594580842132, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00011014938354492188, "grad_norm": 6.8072357177734375, "learning_rate": 1e-05, "loss": 0.4584, "mean_token_accuracy": 0.8849585652351379, "num_tokens": 766138945.0, "step": 20176 }, { "epoch": 2.5667217911207225, "ewc_loss": 0.140625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00011110305786132812, "grad_norm": 7.043607711791992, "learning_rate": 1e-05, "loss": 0.5051, "mean_token_accuracy": 0.8718670010566711, "num_tokens": 766180771.0, "step": 20177 }, { "epoch": 2.566849001399313, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.55059289932251, "learning_rate": 1e-05, "loss": 0.5151, "mean_token_accuracy": 0.868224024772644, "num_tokens": 766216907.0, "step": 20178 }, { "epoch": 2.5669762116779036, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.94383430480957, "learning_rate": 1e-05, "loss": 0.5002, "mean_token_accuracy": 0.8728863000869751, "num_tokens": 766256302.0, "step": 20179 }, { "epoch": 2.567103421956494, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.316267490386963, "learning_rate": 1e-05, "loss": 0.4763, "mean_token_accuracy": 0.8786501288414001, "num_tokens": 766291255.0, "step": 20180 }, { "epoch": 2.5672306322350846, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 7.010037422180176, "learning_rate": 1e-05, "loss": 0.4714, "mean_token_accuracy": 0.882413387298584, "num_tokens": 766323895.0, "step": 20181 }, { "epoch": 2.567357842513675, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.103469371795654, "learning_rate": 1e-05, "loss": 0.4476, "mean_token_accuracy": 0.8830320835113525, "num_tokens": 766364518.0, "step": 20182 }, { "epoch": 2.5674850527922657, "ewc_loss": 0.140625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.0001125335693359375, "grad_norm": 7.24717378616333, "learning_rate": 1e-05, "loss": 0.5277, "mean_token_accuracy": 0.865339994430542, "num_tokens": 766402256.0, "step": 20183 }, { "epoch": 2.567612263070856, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.008703708648682, "learning_rate": 1e-05, "loss": 0.4843, "mean_token_accuracy": 0.8771764039993286, "num_tokens": 766435186.0, "step": 20184 }, { "epoch": 2.5677394733494467, "ewc_loss": 0.150390625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.0001220703125, "grad_norm": 7.614212989807129, "learning_rate": 1e-05, "loss": 0.5414, "mean_token_accuracy": 0.8705387115478516, "num_tokens": 766476950.0, "step": 20185 }, { "epoch": 2.5678666836280373, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.166962623596191, "learning_rate": 1e-05, "loss": 0.4399, "mean_token_accuracy": 0.8911833763122559, "num_tokens": 766517908.0, "step": 20186 }, { "epoch": 2.567993893906628, "ewc_loss": 0.15625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.0001277923583984375, "grad_norm": 7.846046447753906, "learning_rate": 1e-05, "loss": 0.5479, "mean_token_accuracy": 0.8648491501808167, "num_tokens": 766556251.0, "step": 20187 }, { "epoch": 2.5681211041852183, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.629880905151367, "learning_rate": 1e-05, "loss": 0.5208, "mean_token_accuracy": 0.8650556802749634, "num_tokens": 766593865.0, "step": 20188 }, { "epoch": 2.5682483144638084, "ewc_loss": 0.1484375, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00011968612670898438, "grad_norm": 7.539241313934326, "learning_rate": 1e-05, "loss": 0.506, "mean_token_accuracy": 0.873258113861084, "num_tokens": 766627594.0, "step": 20189 }, { "epoch": 2.5683755247423994, "ewc_loss": 0.140625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00011205673217773438, "grad_norm": 6.874873161315918, "learning_rate": 1e-05, "loss": 0.4712, "mean_token_accuracy": 0.8819213509559631, "num_tokens": 766669620.0, "step": 20190 }, { "epoch": 2.5685027350209895, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.967640399932861, "learning_rate": 1e-05, "loss": 0.4631, "mean_token_accuracy": 0.8843621015548706, "num_tokens": 766711553.0, "step": 20191 }, { "epoch": 2.5686299452995804, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 6.890974521636963, "learning_rate": 1e-05, "loss": 0.488, "mean_token_accuracy": 0.8757272362709045, "num_tokens": 766751717.0, "step": 20192 }, { "epoch": 2.5687571555781705, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.420266628265381, "learning_rate": 1e-05, "loss": 0.4638, "mean_token_accuracy": 0.8790230751037598, "num_tokens": 766782207.0, "step": 20193 }, { "epoch": 2.5688843658567615, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 6.987209796905518, "learning_rate": 1e-05, "loss": 0.5, "mean_token_accuracy": 0.8734190464019775, "num_tokens": 766822050.0, "step": 20194 }, { "epoch": 2.5690115761353516, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.991785049438477, "learning_rate": 1e-05, "loss": 0.4989, "mean_token_accuracy": 0.866301417350769, "num_tokens": 766867713.0, "step": 20195 }, { "epoch": 2.569138786413942, "ewc_loss": 0.140625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00011205673217773438, "grad_norm": 7.263975620269775, "learning_rate": 1e-05, "loss": 0.4792, "mean_token_accuracy": 0.8811655044555664, "num_tokens": 766908191.0, "step": 20196 }, { "epoch": 2.5692659966925326, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.824678897857666, "learning_rate": 1e-05, "loss": 0.5026, "mean_token_accuracy": 0.8657796382904053, "num_tokens": 766945774.0, "step": 20197 }, { "epoch": 2.569393206971123, "ewc_loss": 0.154296875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00012493133544921875, "grad_norm": 7.698378086090088, "learning_rate": 1e-05, "loss": 0.4643, "mean_token_accuracy": 0.8911257982254028, "num_tokens": 766986255.0, "step": 20198 }, { "epoch": 2.5695204172497137, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.051990509033203, "learning_rate": 1e-05, "loss": 0.478, "mean_token_accuracy": 0.8757560849189758, "num_tokens": 767020930.0, "step": 20199 }, { "epoch": 2.5696476275283042, "ewc_loss": 0.16015625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.0001316070556640625, "grad_norm": 7.982645511627197, "learning_rate": 1e-05, "loss": 0.5309, "mean_token_accuracy": 0.8723022937774658, "num_tokens": 767059863.0, "step": 20200 }, { "epoch": 2.5697748378068948, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 6.5989766120910645, "learning_rate": 1e-05, "loss": 0.5757, "mean_token_accuracy": 0.8504140377044678, "num_tokens": 767100566.0, "step": 20201 }, { "epoch": 2.5699020480854853, "ewc_loss": 0.154296875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00012493133544921875, "grad_norm": 7.72230339050293, "learning_rate": 1e-05, "loss": 0.5229, "mean_token_accuracy": 0.8695334196090698, "num_tokens": 767137892.0, "step": 20202 }, { "epoch": 2.570029258364076, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00011396408081054688, "grad_norm": 6.877274513244629, "learning_rate": 1e-05, "loss": 0.5675, "mean_token_accuracy": 0.8566639423370361, "num_tokens": 767175340.0, "step": 20203 }, { "epoch": 2.5701564686426663, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00011539459228515625, "grad_norm": 7.272302150726318, "learning_rate": 1e-05, "loss": 0.5444, "mean_token_accuracy": 0.8605964183807373, "num_tokens": 767209601.0, "step": 20204 }, { "epoch": 2.570283678921257, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010967254638671875, "grad_norm": 6.789968013763428, "learning_rate": 1e-05, "loss": 0.4327, "mean_token_accuracy": 0.8957319259643555, "num_tokens": 767247500.0, "step": 20205 }, { "epoch": 2.5704108891998474, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.894992351531982, "learning_rate": 1e-05, "loss": 0.5342, "mean_token_accuracy": 0.8624467849731445, "num_tokens": 767285251.0, "step": 20206 }, { "epoch": 2.570538099478438, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.653063774108887, "learning_rate": 1e-05, "loss": 0.4613, "mean_token_accuracy": 0.8852091431617737, "num_tokens": 767321493.0, "step": 20207 }, { "epoch": 2.5706653097570284, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.685523986816406, "learning_rate": 1e-05, "loss": 0.4594, "mean_token_accuracy": 0.8848174214363098, "num_tokens": 767351375.0, "step": 20208 }, { "epoch": 2.570792520035619, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.4507975578308105, "learning_rate": 1e-05, "loss": 0.4825, "mean_token_accuracy": 0.8744402527809143, "num_tokens": 767392990.0, "step": 20209 }, { "epoch": 2.5709197303142095, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.628952980041504, "learning_rate": 1e-05, "loss": 0.4685, "mean_token_accuracy": 0.8806995153427124, "num_tokens": 767431584.0, "step": 20210 }, { "epoch": 2.5710469405928, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.261753082275391, "learning_rate": 1e-05, "loss": 0.4832, "mean_token_accuracy": 0.8732929229736328, "num_tokens": 767468451.0, "step": 20211 }, { "epoch": 2.5711741508713906, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.694532871246338, "learning_rate": 1e-05, "loss": 0.5048, "mean_token_accuracy": 0.8695907592773438, "num_tokens": 767508133.0, "step": 20212 }, { "epoch": 2.571301361149981, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 6.071666717529297, "learning_rate": 1e-05, "loss": 0.4514, "mean_token_accuracy": 0.882373571395874, "num_tokens": 767539427.0, "step": 20213 }, { "epoch": 2.571428571428571, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.892678737640381, "learning_rate": 1e-05, "loss": 0.4899, "mean_token_accuracy": 0.875465989112854, "num_tokens": 767574842.0, "step": 20214 }, { "epoch": 2.571555781707162, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.895716667175293, "learning_rate": 1e-05, "loss": 0.4771, "mean_token_accuracy": 0.8732892274856567, "num_tokens": 767609378.0, "step": 20215 }, { "epoch": 2.5716829919857522, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00011301040649414062, "grad_norm": 7.246631622314453, "learning_rate": 1e-05, "loss": 0.5072, "mean_token_accuracy": 0.869776725769043, "num_tokens": 767642283.0, "step": 20216 }, { "epoch": 2.571810202264343, "ewc_loss": 0.11474609375, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.842985153198242, "learning_rate": 1e-05, "loss": 0.4519, "mean_token_accuracy": 0.8814917206764221, "num_tokens": 767680019.0, "step": 20217 }, { "epoch": 2.5719374125429333, "ewc_loss": 0.15234375, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.0001239776611328125, "grad_norm": 7.681270599365234, "learning_rate": 1e-05, "loss": 0.5315, "mean_token_accuracy": 0.8693097829818726, "num_tokens": 767715570.0, "step": 20218 }, { "epoch": 2.572064622821524, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.123993873596191, "learning_rate": 1e-05, "loss": 0.4721, "mean_token_accuracy": 0.8767738342285156, "num_tokens": 767757257.0, "step": 20219 }, { "epoch": 2.5721918331001143, "ewc_loss": 0.158203125, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00012874603271484375, "grad_norm": 7.859521389007568, "learning_rate": 1e-05, "loss": 0.4477, "mean_token_accuracy": 0.8979828357696533, "num_tokens": 767791684.0, "step": 20220 }, { "epoch": 2.572319043378705, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00011014938354492188, "grad_norm": 6.635734558105469, "learning_rate": 1e-05, "loss": 0.4905, "mean_token_accuracy": 0.8771979808807373, "num_tokens": 767831687.0, "step": 20221 }, { "epoch": 2.5724462536572954, "ewc_loss": 0.1484375, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00011920928955078125, "grad_norm": 7.504899501800537, "learning_rate": 1e-05, "loss": 0.482, "mean_token_accuracy": 0.8831061124801636, "num_tokens": 767871349.0, "step": 20222 }, { "epoch": 2.572573463935886, "ewc_loss": 0.140625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.0001125335693359375, "grad_norm": 6.842121601104736, "learning_rate": 1e-05, "loss": 0.4716, "mean_token_accuracy": 0.8841874599456787, "num_tokens": 767908301.0, "step": 20223 }, { "epoch": 2.5727006742144765, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.966851711273193, "learning_rate": 1e-05, "loss": 0.4796, "mean_token_accuracy": 0.8796161413192749, "num_tokens": 767944809.0, "step": 20224 }, { "epoch": 2.572827884493067, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 6.84603214263916, "learning_rate": 1e-05, "loss": 0.5133, "mean_token_accuracy": 0.870751142501831, "num_tokens": 767984629.0, "step": 20225 }, { "epoch": 2.5729550947716575, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.484015941619873, "learning_rate": 1e-05, "loss": 0.5075, "mean_token_accuracy": 0.8639885187149048, "num_tokens": 768017697.0, "step": 20226 }, { "epoch": 2.573082305050248, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.88723087310791, "learning_rate": 1e-05, "loss": 0.5077, "mean_token_accuracy": 0.8724363446235657, "num_tokens": 768060259.0, "step": 20227 }, { "epoch": 2.5732095153288386, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.126040458679199, "learning_rate": 1e-05, "loss": 0.4879, "mean_token_accuracy": 0.8682843446731567, "num_tokens": 768096248.0, "step": 20228 }, { "epoch": 2.573336725607429, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010967254638671875, "grad_norm": 7.072671413421631, "learning_rate": 1e-05, "loss": 0.4874, "mean_token_accuracy": 0.8799861669540405, "num_tokens": 768131730.0, "step": 20229 }, { "epoch": 2.5734639358860196, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 5.9455342292785645, "learning_rate": 1e-05, "loss": 0.4455, "mean_token_accuracy": 0.8866169452667236, "num_tokens": 768167551.0, "step": 20230 }, { "epoch": 2.57359114616461, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00011587142944335938, "grad_norm": 7.366870880126953, "learning_rate": 1e-05, "loss": 0.5003, "mean_token_accuracy": 0.8764641284942627, "num_tokens": 768198212.0, "step": 20231 }, { "epoch": 2.5737183564432007, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 5.972988128662109, "learning_rate": 1e-05, "loss": 0.4838, "mean_token_accuracy": 0.8740191459655762, "num_tokens": 768233052.0, "step": 20232 }, { "epoch": 2.573845566721791, "ewc_loss": 0.15234375, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.63527250289917, "learning_rate": 1e-05, "loss": 0.4632, "mean_token_accuracy": 0.8891181945800781, "num_tokens": 768272961.0, "step": 20233 }, { "epoch": 2.5739727770003817, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.224318504333496, "learning_rate": 1e-05, "loss": 0.4511, "mean_token_accuracy": 0.8853914737701416, "num_tokens": 768308748.0, "step": 20234 }, { "epoch": 2.5740999872789723, "ewc_loss": 0.15234375, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.0001239776611328125, "grad_norm": 7.666093826293945, "learning_rate": 1e-05, "loss": 0.5389, "mean_token_accuracy": 0.8673310279846191, "num_tokens": 768346592.0, "step": 20235 }, { "epoch": 2.574227197557563, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.525099277496338, "learning_rate": 1e-05, "loss": 0.4852, "mean_token_accuracy": 0.8798947930335999, "num_tokens": 768394367.0, "step": 20236 }, { "epoch": 2.5743544078361533, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00011777877807617188, "grad_norm": 7.418298721313477, "learning_rate": 1e-05, "loss": 0.5313, "mean_token_accuracy": 0.8702576160430908, "num_tokens": 768437060.0, "step": 20237 }, { "epoch": 2.574481618114744, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 6.640493869781494, "learning_rate": 1e-05, "loss": 0.4852, "mean_token_accuracy": 0.8765993118286133, "num_tokens": 768470373.0, "step": 20238 }, { "epoch": 2.574608828393334, "ewc_loss": 0.140625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00011157989501953125, "grad_norm": 7.06915283203125, "learning_rate": 1e-05, "loss": 0.4707, "mean_token_accuracy": 0.8838185667991638, "num_tokens": 768507567.0, "step": 20239 }, { "epoch": 2.574736038671925, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.598047256469727, "learning_rate": 1e-05, "loss": 0.4628, "mean_token_accuracy": 0.8857616186141968, "num_tokens": 768547217.0, "step": 20240 }, { "epoch": 2.574863248950515, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.767046928405762, "learning_rate": 1e-05, "loss": 0.4667, "mean_token_accuracy": 0.8828469514846802, "num_tokens": 768586720.0, "step": 20241 }, { "epoch": 2.574990459229106, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.4985551834106445, "learning_rate": 1e-05, "loss": 0.4644, "mean_token_accuracy": 0.8791012763977051, "num_tokens": 768627890.0, "step": 20242 }, { "epoch": 2.575117669507696, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.6228251457214355, "learning_rate": 1e-05, "loss": 0.5141, "mean_token_accuracy": 0.8696678876876831, "num_tokens": 768663393.0, "step": 20243 }, { "epoch": 2.5752448797862866, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.361232280731201, "learning_rate": 1e-05, "loss": 0.4503, "mean_token_accuracy": 0.8874691128730774, "num_tokens": 768706950.0, "step": 20244 }, { "epoch": 2.575372090064877, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.584121227264404, "learning_rate": 1e-05, "loss": 0.5179, "mean_token_accuracy": 0.8664877414703369, "num_tokens": 768747391.0, "step": 20245 }, { "epoch": 2.5754993003434676, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.212285041809082, "learning_rate": 1e-05, "loss": 0.4069, "mean_token_accuracy": 0.8960883617401123, "num_tokens": 768781170.0, "step": 20246 }, { "epoch": 2.575626510622058, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.6525068283081055, "learning_rate": 1e-05, "loss": 0.4923, "mean_token_accuracy": 0.8727733492851257, "num_tokens": 768820505.0, "step": 20247 }, { "epoch": 2.5757537209006487, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 6.060735702514648, "learning_rate": 1e-05, "loss": 0.4786, "mean_token_accuracy": 0.8749666810035706, "num_tokens": 768859647.0, "step": 20248 }, { "epoch": 2.575880931179239, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.836755752563477, "learning_rate": 1e-05, "loss": 0.4369, "mean_token_accuracy": 0.8912231922149658, "num_tokens": 768894883.0, "step": 20249 }, { "epoch": 2.5760081414578297, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.90287971496582, "learning_rate": 1e-05, "loss": 0.4346, "mean_token_accuracy": 0.8899851441383362, "num_tokens": 768934648.0, "step": 20250 }, { "epoch": 2.5761353517364203, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 7.134200572967529, "learning_rate": 1e-05, "loss": 0.4687, "mean_token_accuracy": 0.8818756341934204, "num_tokens": 768969568.0, "step": 20251 }, { "epoch": 2.576262562015011, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.839128494262695, "learning_rate": 1e-05, "loss": 0.5021, "mean_token_accuracy": 0.8665996789932251, "num_tokens": 769009423.0, "step": 20252 }, { "epoch": 2.5763897722936013, "ewc_loss": 0.150390625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00012111663818359375, "grad_norm": 7.539093971252441, "learning_rate": 1e-05, "loss": 0.4965, "mean_token_accuracy": 0.8800380229949951, "num_tokens": 769043812.0, "step": 20253 }, { "epoch": 2.576516982572192, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.012306213378906, "learning_rate": 1e-05, "loss": 0.457, "mean_token_accuracy": 0.8809106945991516, "num_tokens": 769079959.0, "step": 20254 }, { "epoch": 2.5766441928507824, "ewc_loss": 0.15625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.0001277923583984375, "grad_norm": 7.793022632598877, "learning_rate": 1e-05, "loss": 0.4723, "mean_token_accuracy": 0.8885700702667236, "num_tokens": 769119375.0, "step": 20255 }, { "epoch": 2.576771403129373, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.457371234893799, "learning_rate": 1e-05, "loss": 0.4655, "mean_token_accuracy": 0.8854813575744629, "num_tokens": 769157085.0, "step": 20256 }, { "epoch": 2.5768986134079634, "ewc_loss": 0.15234375, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.578670978546143, "learning_rate": 1e-05, "loss": 0.5107, "mean_token_accuracy": 0.8722900152206421, "num_tokens": 769190883.0, "step": 20257 }, { "epoch": 2.577025823686554, "ewc_loss": 0.140625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00011157989501953125, "grad_norm": 6.747417449951172, "learning_rate": 1e-05, "loss": 0.4797, "mean_token_accuracy": 0.8792823553085327, "num_tokens": 769228370.0, "step": 20258 }, { "epoch": 2.5771530339651445, "ewc_loss": 0.140625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00011110305786132812, "grad_norm": 7.063770771026611, "learning_rate": 1e-05, "loss": 0.4157, "mean_token_accuracy": 0.9029949903488159, "num_tokens": 769265551.0, "step": 20259 }, { "epoch": 2.577280244243735, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 6.8043694496154785, "learning_rate": 1e-05, "loss": 0.4764, "mean_token_accuracy": 0.8798452615737915, "num_tokens": 769299482.0, "step": 20260 }, { "epoch": 2.5774074545223256, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.547760963439941, "learning_rate": 1e-05, "loss": 0.4754, "mean_token_accuracy": 0.8791645765304565, "num_tokens": 769339165.0, "step": 20261 }, { "epoch": 2.5775346648009156, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.864699363708496, "learning_rate": 1e-05, "loss": 0.4729, "mean_token_accuracy": 0.8820289373397827, "num_tokens": 769374066.0, "step": 20262 }, { "epoch": 2.5776618750795066, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.157027721405029, "learning_rate": 1e-05, "loss": 0.4777, "mean_token_accuracy": 0.8760954141616821, "num_tokens": 769415536.0, "step": 20263 }, { "epoch": 2.5777890853580967, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010967254638671875, "grad_norm": 7.054952621459961, "learning_rate": 1e-05, "loss": 0.496, "mean_token_accuracy": 0.8748500347137451, "num_tokens": 769454016.0, "step": 20264 }, { "epoch": 2.5779162956366877, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 5.947094917297363, "learning_rate": 1e-05, "loss": 0.4634, "mean_token_accuracy": 0.8797701001167297, "num_tokens": 769493150.0, "step": 20265 }, { "epoch": 2.5780435059152778, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00011491775512695312, "grad_norm": 7.358011245727539, "learning_rate": 1e-05, "loss": 0.4966, "mean_token_accuracy": 0.8733422160148621, "num_tokens": 769524247.0, "step": 20266 }, { "epoch": 2.5781707161938687, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 5.952563762664795, "learning_rate": 1e-05, "loss": 0.4299, "mean_token_accuracy": 0.8895060420036316, "num_tokens": 769566476.0, "step": 20267 }, { "epoch": 2.578297926472459, "ewc_loss": 0.154296875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00012493133544921875, "grad_norm": 7.644360542297363, "learning_rate": 1e-05, "loss": 0.5134, "mean_token_accuracy": 0.8743672966957092, "num_tokens": 769603474.0, "step": 20268 }, { "epoch": 2.5784251367510493, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.218770980834961, "learning_rate": 1e-05, "loss": 0.4838, "mean_token_accuracy": 0.8737360239028931, "num_tokens": 769640446.0, "step": 20269 }, { "epoch": 2.57855234702964, "ewc_loss": 0.154296875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00012493133544921875, "grad_norm": 7.674943923950195, "learning_rate": 1e-05, "loss": 0.4953, "mean_token_accuracy": 0.8830164074897766, "num_tokens": 769679809.0, "step": 20270 }, { "epoch": 2.5786795573082304, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 6.515440464019775, "learning_rate": 1e-05, "loss": 0.4389, "mean_token_accuracy": 0.8952292203903198, "num_tokens": 769717221.0, "step": 20271 }, { "epoch": 2.578806767586821, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.000118255615234375, "grad_norm": 7.428516864776611, "learning_rate": 1e-05, "loss": 0.4461, "mean_token_accuracy": 0.8904857635498047, "num_tokens": 769749350.0, "step": 20272 }, { "epoch": 2.5789339778654115, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.628582000732422, "learning_rate": 1e-05, "loss": 0.5113, "mean_token_accuracy": 0.8667809963226318, "num_tokens": 769786604.0, "step": 20273 }, { "epoch": 2.579061188144002, "ewc_loss": 0.140625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00011110305786132812, "grad_norm": 7.08643102645874, "learning_rate": 1e-05, "loss": 0.443, "mean_token_accuracy": 0.8919574022293091, "num_tokens": 769827709.0, "step": 20274 }, { "epoch": 2.5791883984225925, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.590307712554932, "learning_rate": 1e-05, "loss": 0.4976, "mean_token_accuracy": 0.8743757605552673, "num_tokens": 769866713.0, "step": 20275 }, { "epoch": 2.579315608701183, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.797619342803955, "learning_rate": 1e-05, "loss": 0.4988, "mean_token_accuracy": 0.8777451515197754, "num_tokens": 769907365.0, "step": 20276 }, { "epoch": 2.5794428189797736, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.483833312988281, "learning_rate": 1e-05, "loss": 0.4831, "mean_token_accuracy": 0.8775544166564941, "num_tokens": 769945573.0, "step": 20277 }, { "epoch": 2.579570029258364, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.643593788146973, "learning_rate": 1e-05, "loss": 0.4856, "mean_token_accuracy": 0.8742663860321045, "num_tokens": 769980936.0, "step": 20278 }, { "epoch": 2.5796972395369546, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.359560966491699, "learning_rate": 1e-05, "loss": 0.5251, "mean_token_accuracy": 0.8631410598754883, "num_tokens": 770022244.0, "step": 20279 }, { "epoch": 2.579824449815545, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.604793071746826, "learning_rate": 1e-05, "loss": 0.4516, "mean_token_accuracy": 0.8846937417984009, "num_tokens": 770059078.0, "step": 20280 }, { "epoch": 2.5799516600941357, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.229830741882324, "learning_rate": 1e-05, "loss": 0.513, "mean_token_accuracy": 0.8652193546295166, "num_tokens": 770100816.0, "step": 20281 }, { "epoch": 2.580078870372726, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.650206565856934, "learning_rate": 1e-05, "loss": 0.4619, "mean_token_accuracy": 0.8828881978988647, "num_tokens": 770138912.0, "step": 20282 }, { "epoch": 2.5802060806513167, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.07589864730835, "learning_rate": 1e-05, "loss": 0.449, "mean_token_accuracy": 0.8846385478973389, "num_tokens": 770178915.0, "step": 20283 }, { "epoch": 2.5803332909299073, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.819686412811279, "learning_rate": 1e-05, "loss": 0.4582, "mean_token_accuracy": 0.8854488134384155, "num_tokens": 770217707.0, "step": 20284 }, { "epoch": 2.580460501208498, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.92409086227417, "learning_rate": 1e-05, "loss": 0.4651, "mean_token_accuracy": 0.8772287368774414, "num_tokens": 770251659.0, "step": 20285 }, { "epoch": 2.5805877114870883, "ewc_loss": 0.140625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00011157989501953125, "grad_norm": 7.112341403961182, "learning_rate": 1e-05, "loss": 0.5112, "mean_token_accuracy": 0.8721590638160706, "num_tokens": 770291612.0, "step": 20286 }, { "epoch": 2.5807149217656784, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.838507652282715, "learning_rate": 1e-05, "loss": 0.4865, "mean_token_accuracy": 0.8701874017715454, "num_tokens": 770333822.0, "step": 20287 }, { "epoch": 2.5808421320442694, "ewc_loss": 0.1494140625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00012063980102539062, "grad_norm": 7.507319927215576, "learning_rate": 1e-05, "loss": 0.5094, "mean_token_accuracy": 0.8727960586547852, "num_tokens": 770369850.0, "step": 20288 }, { "epoch": 2.5809693423228595, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 5.995246410369873, "learning_rate": 1e-05, "loss": 0.5112, "mean_token_accuracy": 0.8651199340820312, "num_tokens": 770408825.0, "step": 20289 }, { "epoch": 2.5810965526014504, "ewc_loss": 0.15625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00012683868408203125, "grad_norm": 7.751849174499512, "learning_rate": 1e-05, "loss": 0.4702, "mean_token_accuracy": 0.8903239965438843, "num_tokens": 770451320.0, "step": 20290 }, { "epoch": 2.5812237628800405, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.420640468597412, "learning_rate": 1e-05, "loss": 0.5038, "mean_token_accuracy": 0.8713699579238892, "num_tokens": 770488099.0, "step": 20291 }, { "epoch": 2.5813509731586315, "ewc_loss": 0.150390625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00012159347534179688, "grad_norm": 7.554666519165039, "learning_rate": 1e-05, "loss": 0.5256, "mean_token_accuracy": 0.8689607381820679, "num_tokens": 770533689.0, "step": 20292 }, { "epoch": 2.5814781834372216, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010967254638671875, "grad_norm": 6.708773612976074, "learning_rate": 1e-05, "loss": 0.4943, "mean_token_accuracy": 0.8792487382888794, "num_tokens": 770570998.0, "step": 20293 }, { "epoch": 2.581605393715812, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00011014938354492188, "grad_norm": 7.058374881744385, "learning_rate": 1e-05, "loss": 0.4506, "mean_token_accuracy": 0.8903664350509644, "num_tokens": 770610248.0, "step": 20294 }, { "epoch": 2.5817326039944026, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 6.77571964263916, "learning_rate": 1e-05, "loss": 0.5305, "mean_token_accuracy": 0.8615756630897522, "num_tokens": 770650230.0, "step": 20295 }, { "epoch": 2.581859814272993, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.53521203994751, "learning_rate": 1e-05, "loss": 0.454, "mean_token_accuracy": 0.8848609924316406, "num_tokens": 770694740.0, "step": 20296 }, { "epoch": 2.5819870245515837, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.853054046630859, "learning_rate": 1e-05, "loss": 0.4647, "mean_token_accuracy": 0.8837951421737671, "num_tokens": 770734004.0, "step": 20297 }, { "epoch": 2.582114234830174, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.139455318450928, "learning_rate": 1e-05, "loss": 0.474, "mean_token_accuracy": 0.8739280700683594, "num_tokens": 770770421.0, "step": 20298 }, { "epoch": 2.5822414451087647, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00011014938354492188, "grad_norm": 7.0566887855529785, "learning_rate": 1e-05, "loss": 0.5048, "mean_token_accuracy": 0.871857225894928, "num_tokens": 770810790.0, "step": 20299 }, { "epoch": 2.5823686553873553, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.927346706390381, "learning_rate": 1e-05, "loss": 0.4844, "mean_token_accuracy": 0.8730642795562744, "num_tokens": 770848614.0, "step": 20300 }, { "epoch": 2.582495865665946, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00011539459228515625, "grad_norm": 7.356391906738281, "learning_rate": 1e-05, "loss": 0.4395, "mean_token_accuracy": 0.8969599008560181, "num_tokens": 770889888.0, "step": 20301 }, { "epoch": 2.5826230759445363, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 5.968997478485107, "learning_rate": 1e-05, "loss": 0.5501, "mean_token_accuracy": 0.8523669242858887, "num_tokens": 770925918.0, "step": 20302 }, { "epoch": 2.582750286223127, "ewc_loss": 0.15234375, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.613381385803223, "learning_rate": 1e-05, "loss": 0.4392, "mean_token_accuracy": 0.895062267780304, "num_tokens": 770963130.0, "step": 20303 }, { "epoch": 2.5828774965017174, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.227928161621094, "learning_rate": 1e-05, "loss": 0.5024, "mean_token_accuracy": 0.8673713207244873, "num_tokens": 770998308.0, "step": 20304 }, { "epoch": 2.583004706780308, "ewc_loss": 0.15234375, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.614301681518555, "learning_rate": 1e-05, "loss": 0.5165, "mean_token_accuracy": 0.8701585531234741, "num_tokens": 771032532.0, "step": 20305 }, { "epoch": 2.5831319170588984, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.50519323348999, "learning_rate": 1e-05, "loss": 0.465, "mean_token_accuracy": 0.8836987018585205, "num_tokens": 771070859.0, "step": 20306 }, { "epoch": 2.583259127337489, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.0001163482666015625, "grad_norm": 7.359319686889648, "learning_rate": 1e-05, "loss": 0.5159, "mean_token_accuracy": 0.8705717325210571, "num_tokens": 771111216.0, "step": 20307 }, { "epoch": 2.5833863376160795, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.5988545417785645, "learning_rate": 1e-05, "loss": 0.4921, "mean_token_accuracy": 0.8780065774917603, "num_tokens": 771146703.0, "step": 20308 }, { "epoch": 2.58351354789467, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 7.040495872497559, "learning_rate": 1e-05, "loss": 0.4812, "mean_token_accuracy": 0.88133704662323, "num_tokens": 771178778.0, "step": 20309 }, { "epoch": 2.5836407581732606, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.541264057159424, "learning_rate": 1e-05, "loss": 0.5139, "mean_token_accuracy": 0.866431474685669, "num_tokens": 771217272.0, "step": 20310 }, { "epoch": 2.583767968451851, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.782999038696289, "learning_rate": 1e-05, "loss": 0.4963, "mean_token_accuracy": 0.8721659183502197, "num_tokens": 771252085.0, "step": 20311 }, { "epoch": 2.583895178730441, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.445796489715576, "learning_rate": 1e-05, "loss": 0.4594, "mean_token_accuracy": 0.8810583353042603, "num_tokens": 771285367.0, "step": 20312 }, { "epoch": 2.584022389009032, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.646697521209717, "learning_rate": 1e-05, "loss": 0.4549, "mean_token_accuracy": 0.8831182718276978, "num_tokens": 771324388.0, "step": 20313 }, { "epoch": 2.5841495992876222, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.317084312438965, "learning_rate": 1e-05, "loss": 0.4511, "mean_token_accuracy": 0.8837980628013611, "num_tokens": 771361307.0, "step": 20314 }, { "epoch": 2.584276809566213, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.630507469177246, "learning_rate": 1e-05, "loss": 0.5114, "mean_token_accuracy": 0.8664288520812988, "num_tokens": 771400383.0, "step": 20315 }, { "epoch": 2.5844040198448033, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.173428535461426, "learning_rate": 1e-05, "loss": 0.4407, "mean_token_accuracy": 0.8870930671691895, "num_tokens": 771435065.0, "step": 20316 }, { "epoch": 2.584531230123394, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.720473766326904, "learning_rate": 1e-05, "loss": 0.5257, "mean_token_accuracy": 0.8618659377098083, "num_tokens": 771480573.0, "step": 20317 }, { "epoch": 2.5846584404019843, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 6.022321701049805, "learning_rate": 1e-05, "loss": 0.4514, "mean_token_accuracy": 0.8838580846786499, "num_tokens": 771517172.0, "step": 20318 }, { "epoch": 2.584785650680575, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.929628372192383, "learning_rate": 1e-05, "loss": 0.5214, "mean_token_accuracy": 0.8658876419067383, "num_tokens": 771554347.0, "step": 20319 }, { "epoch": 2.5849128609591654, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 5.894273281097412, "learning_rate": 1e-05, "loss": 0.4373, "mean_token_accuracy": 0.8841859698295593, "num_tokens": 771591209.0, "step": 20320 }, { "epoch": 2.585040071237756, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00011396408081054688, "grad_norm": 7.23767614364624, "learning_rate": 1e-05, "loss": 0.4629, "mean_token_accuracy": 0.887773334980011, "num_tokens": 771626604.0, "step": 20321 }, { "epoch": 2.5851672815163464, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 5.8937201499938965, "learning_rate": 1e-05, "loss": 0.4756, "mean_token_accuracy": 0.8748877048492432, "num_tokens": 771662525.0, "step": 20322 }, { "epoch": 2.585294491794937, "ewc_loss": 0.1494140625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00012063980102539062, "grad_norm": 7.547743320465088, "learning_rate": 1e-05, "loss": 0.4737, "mean_token_accuracy": 0.887226939201355, "num_tokens": 771701616.0, "step": 20323 }, { "epoch": 2.5854217020735275, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.12756872177124, "learning_rate": 1e-05, "loss": 0.5143, "mean_token_accuracy": 0.8692585229873657, "num_tokens": 771736864.0, "step": 20324 }, { "epoch": 2.585548912352118, "ewc_loss": 0.154296875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00012493133544921875, "grad_norm": 7.630680084228516, "learning_rate": 1e-05, "loss": 0.4783, "mean_token_accuracy": 0.8857145309448242, "num_tokens": 771774900.0, "step": 20325 }, { "epoch": 2.5856761226307086, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.486355304718018, "learning_rate": 1e-05, "loss": 0.5022, "mean_token_accuracy": 0.8720028400421143, "num_tokens": 771812078.0, "step": 20326 }, { "epoch": 2.585803332909299, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.0001163482666015625, "grad_norm": 7.320652961730957, "learning_rate": 1e-05, "loss": 0.4977, "mean_token_accuracy": 0.8778582811355591, "num_tokens": 771852247.0, "step": 20327 }, { "epoch": 2.5859305431878896, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010967254638671875, "grad_norm": 6.669670104980469, "learning_rate": 1e-05, "loss": 0.5102, "mean_token_accuracy": 0.8725699186325073, "num_tokens": 771891264.0, "step": 20328 }, { "epoch": 2.58605775346648, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.83165168762207, "learning_rate": 1e-05, "loss": 0.5246, "mean_token_accuracy": 0.8649030327796936, "num_tokens": 771935356.0, "step": 20329 }, { "epoch": 2.5861849637450707, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.757480144500732, "learning_rate": 1e-05, "loss": 0.5527, "mean_token_accuracy": 0.8540101051330566, "num_tokens": 771967284.0, "step": 20330 }, { "epoch": 2.586312174023661, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.3275227546691895, "learning_rate": 1e-05, "loss": 0.4564, "mean_token_accuracy": 0.8832986354827881, "num_tokens": 772013465.0, "step": 20331 }, { "epoch": 2.5864393843022517, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.908665180206299, "learning_rate": 1e-05, "loss": 0.496, "mean_token_accuracy": 0.8720758557319641, "num_tokens": 772046978.0, "step": 20332 }, { "epoch": 2.5865665945808423, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 5.9823431968688965, "learning_rate": 1e-05, "loss": 0.4534, "mean_token_accuracy": 0.8857582211494446, "num_tokens": 772081630.0, "step": 20333 }, { "epoch": 2.586693804859433, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00011301040649414062, "grad_norm": 7.199380874633789, "learning_rate": 1e-05, "loss": 0.5312, "mean_token_accuracy": 0.8664776086807251, "num_tokens": 772122197.0, "step": 20334 }, { "epoch": 2.5868210151380233, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.846545219421387, "learning_rate": 1e-05, "loss": 0.4562, "mean_token_accuracy": 0.8793623447418213, "num_tokens": 772162744.0, "step": 20335 }, { "epoch": 2.586948225416614, "ewc_loss": 0.150390625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.0001220703125, "grad_norm": 7.586175918579102, "learning_rate": 1e-05, "loss": 0.5152, "mean_token_accuracy": 0.8710776567459106, "num_tokens": 772207520.0, "step": 20336 }, { "epoch": 2.587075435695204, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.04377555847168, "learning_rate": 1e-05, "loss": 0.4805, "mean_token_accuracy": 0.8754396438598633, "num_tokens": 772244869.0, "step": 20337 }, { "epoch": 2.587202645973795, "ewc_loss": 0.158203125, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00012874603271484375, "grad_norm": 7.8073201179504395, "learning_rate": 1e-05, "loss": 0.5173, "mean_token_accuracy": 0.8775396943092346, "num_tokens": 772279132.0, "step": 20338 }, { "epoch": 2.587329856252385, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.4509124755859375, "learning_rate": 1e-05, "loss": 0.4541, "mean_token_accuracy": 0.8884004950523376, "num_tokens": 772323668.0, "step": 20339 }, { "epoch": 2.587457066530976, "ewc_loss": 0.15234375, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.0001239776611328125, "grad_norm": 7.62678861618042, "learning_rate": 1e-05, "loss": 0.5065, "mean_token_accuracy": 0.8762261867523193, "num_tokens": 772367834.0, "step": 20340 }, { "epoch": 2.587584276809566, "ewc_loss": 0.140625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00011157989501953125, "grad_norm": 6.69457483291626, "learning_rate": 1e-05, "loss": 0.4387, "mean_token_accuracy": 0.8933350443840027, "num_tokens": 772398777.0, "step": 20341 }, { "epoch": 2.5877114870881566, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00011587142944335938, "grad_norm": 7.248548984527588, "learning_rate": 1e-05, "loss": 0.4939, "mean_token_accuracy": 0.8772032260894775, "num_tokens": 772432330.0, "step": 20342 }, { "epoch": 2.587838697366747, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.670568466186523, "learning_rate": 1e-05, "loss": 0.5258, "mean_token_accuracy": 0.8697017431259155, "num_tokens": 772472412.0, "step": 20343 }, { "epoch": 2.5879659076453376, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.899751663208008, "learning_rate": 1e-05, "loss": 0.4795, "mean_token_accuracy": 0.8777502775192261, "num_tokens": 772509896.0, "step": 20344 }, { "epoch": 2.588093117923928, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.564359664916992, "learning_rate": 1e-05, "loss": 0.459, "mean_token_accuracy": 0.885071873664856, "num_tokens": 772549443.0, "step": 20345 }, { "epoch": 2.5882203282025187, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.679502964019775, "learning_rate": 1e-05, "loss": 0.5269, "mean_token_accuracy": 0.8662973642349243, "num_tokens": 772592086.0, "step": 20346 }, { "epoch": 2.588347538481109, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.424867153167725, "learning_rate": 1e-05, "loss": 0.5238, "mean_token_accuracy": 0.8637093901634216, "num_tokens": 772632805.0, "step": 20347 }, { "epoch": 2.5884747487596997, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.624866008758545, "learning_rate": 1e-05, "loss": 0.5192, "mean_token_accuracy": 0.8651201725006104, "num_tokens": 772673806.0, "step": 20348 }, { "epoch": 2.5886019590382903, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.268099308013916, "learning_rate": 1e-05, "loss": 0.4165, "mean_token_accuracy": 0.9002628922462463, "num_tokens": 772709434.0, "step": 20349 }, { "epoch": 2.588729169316881, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.669421195983887, "learning_rate": 1e-05, "loss": 0.4716, "mean_token_accuracy": 0.8801345825195312, "num_tokens": 772748702.0, "step": 20350 }, { "epoch": 2.5888563795954713, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.1086955070495605, "learning_rate": 1e-05, "loss": 0.521, "mean_token_accuracy": 0.8590472936630249, "num_tokens": 772792235.0, "step": 20351 }, { "epoch": 2.588983589874062, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.826048851013184, "learning_rate": 1e-05, "loss": 0.4866, "mean_token_accuracy": 0.8710128664970398, "num_tokens": 772827253.0, "step": 20352 }, { "epoch": 2.5891108001526524, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.935482501983643, "learning_rate": 1e-05, "loss": 0.4284, "mean_token_accuracy": 0.8897876143455505, "num_tokens": 772871951.0, "step": 20353 }, { "epoch": 2.589238010431243, "ewc_loss": 0.140625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00011110305786132812, "grad_norm": 7.1193060874938965, "learning_rate": 1e-05, "loss": 0.474, "mean_token_accuracy": 0.8851180076599121, "num_tokens": 772908585.0, "step": 20354 }, { "epoch": 2.5893652207098334, "ewc_loss": 0.11474609375, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.859128952026367, "learning_rate": 1e-05, "loss": 0.4927, "mean_token_accuracy": 0.8697525262832642, "num_tokens": 772946883.0, "step": 20355 }, { "epoch": 2.589492430988424, "ewc_loss": 0.1484375, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00011968612670898438, "grad_norm": 7.508497714996338, "learning_rate": 1e-05, "loss": 0.5007, "mean_token_accuracy": 0.8811835646629333, "num_tokens": 772985143.0, "step": 20356 }, { "epoch": 2.5896196412670145, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.005273818969727, "learning_rate": 1e-05, "loss": 0.4629, "mean_token_accuracy": 0.8813575506210327, "num_tokens": 773027004.0, "step": 20357 }, { "epoch": 2.589746851545605, "ewc_loss": 0.15625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00012683868408203125, "grad_norm": 7.756617546081543, "learning_rate": 1e-05, "loss": 0.5395, "mean_token_accuracy": 0.8656278848648071, "num_tokens": 773067807.0, "step": 20358 }, { "epoch": 2.5898740618241956, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.429337501525879, "learning_rate": 1e-05, "loss": 0.4774, "mean_token_accuracy": 0.8824509382247925, "num_tokens": 773103160.0, "step": 20359 }, { "epoch": 2.5900012721027856, "ewc_loss": 0.15234375, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.560240268707275, "learning_rate": 1e-05, "loss": 0.4902, "mean_token_accuracy": 0.8840129375457764, "num_tokens": 773143036.0, "step": 20360 }, { "epoch": 2.5901284823813766, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 6.7115583419799805, "learning_rate": 1e-05, "loss": 0.4707, "mean_token_accuracy": 0.8828264474868774, "num_tokens": 773180930.0, "step": 20361 }, { "epoch": 2.5902556926599667, "ewc_loss": 0.140625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00011110305786132812, "grad_norm": 7.084825038909912, "learning_rate": 1e-05, "loss": 0.5551, "mean_token_accuracy": 0.8581262826919556, "num_tokens": 773220056.0, "step": 20362 }, { "epoch": 2.5903829029385577, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010967254638671875, "grad_norm": 6.786834716796875, "learning_rate": 1e-05, "loss": 0.4342, "mean_token_accuracy": 0.8947234749794006, "num_tokens": 773255890.0, "step": 20363 }, { "epoch": 2.5905101132171477, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.556142330169678, "learning_rate": 1e-05, "loss": 0.4722, "mean_token_accuracy": 0.8812808990478516, "num_tokens": 773298727.0, "step": 20364 }, { "epoch": 2.5906373234957387, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 6.854733943939209, "learning_rate": 1e-05, "loss": 0.5517, "mean_token_accuracy": 0.8585572838783264, "num_tokens": 773338137.0, "step": 20365 }, { "epoch": 2.590764533774329, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.148623466491699, "learning_rate": 1e-05, "loss": 0.4539, "mean_token_accuracy": 0.8836118578910828, "num_tokens": 773381532.0, "step": 20366 }, { "epoch": 2.5908917440529193, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00011014938354492188, "grad_norm": 7.053439140319824, "learning_rate": 1e-05, "loss": 0.5185, "mean_token_accuracy": 0.8658804893493652, "num_tokens": 773416354.0, "step": 20367 }, { "epoch": 2.59101895433151, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.920327663421631, "learning_rate": 1e-05, "loss": 0.4766, "mean_token_accuracy": 0.8760740756988525, "num_tokens": 773456130.0, "step": 20368 }, { "epoch": 2.5911461646101004, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00011587142944335938, "grad_norm": 7.366335868835449, "learning_rate": 1e-05, "loss": 0.4789, "mean_token_accuracy": 0.8845561742782593, "num_tokens": 773493812.0, "step": 20369 }, { "epoch": 2.591273374888691, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 5.926722526550293, "learning_rate": 1e-05, "loss": 0.4651, "mean_token_accuracy": 0.878672182559967, "num_tokens": 773533000.0, "step": 20370 }, { "epoch": 2.5914005851672814, "ewc_loss": 0.154296875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00012493133544921875, "grad_norm": 7.674809455871582, "learning_rate": 1e-05, "loss": 0.5113, "mean_token_accuracy": 0.8729501962661743, "num_tokens": 773572355.0, "step": 20371 }, { "epoch": 2.591527795445872, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.228010654449463, "learning_rate": 1e-05, "loss": 0.4443, "mean_token_accuracy": 0.8893404006958008, "num_tokens": 773609154.0, "step": 20372 }, { "epoch": 2.5916550057244625, "ewc_loss": 0.154296875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.000125885009765625, "grad_norm": 7.701077938079834, "learning_rate": 1e-05, "loss": 0.4757, "mean_token_accuracy": 0.8855999708175659, "num_tokens": 773654430.0, "step": 20373 }, { "epoch": 2.591782216003053, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.553168773651123, "learning_rate": 1e-05, "loss": 0.4768, "mean_token_accuracy": 0.881608784198761, "num_tokens": 773693710.0, "step": 20374 }, { "epoch": 2.5919094262816436, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.000118255615234375, "grad_norm": 7.444418907165527, "learning_rate": 1e-05, "loss": 0.5611, "mean_token_accuracy": 0.8522293567657471, "num_tokens": 773728858.0, "step": 20375 }, { "epoch": 2.592036636560234, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 6.668736934661865, "learning_rate": 1e-05, "loss": 0.5088, "mean_token_accuracy": 0.8704754114151001, "num_tokens": 773761080.0, "step": 20376 }, { "epoch": 2.5921638468388246, "ewc_loss": 0.140625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00011110305786132812, "grad_norm": 7.081974506378174, "learning_rate": 1e-05, "loss": 0.4596, "mean_token_accuracy": 0.888563334941864, "num_tokens": 773799084.0, "step": 20377 }, { "epoch": 2.592291057117415, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.598390579223633, "learning_rate": 1e-05, "loss": 0.4591, "mean_token_accuracy": 0.8822349309921265, "num_tokens": 773840378.0, "step": 20378 }, { "epoch": 2.5924182673960057, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.808831214904785, "learning_rate": 1e-05, "loss": 0.4772, "mean_token_accuracy": 0.8797270059585571, "num_tokens": 773878627.0, "step": 20379 }, { "epoch": 2.592545477674596, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.481418609619141, "learning_rate": 1e-05, "loss": 0.505, "mean_token_accuracy": 0.8688945770263672, "num_tokens": 773921492.0, "step": 20380 }, { "epoch": 2.5926726879531867, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.674261569976807, "learning_rate": 1e-05, "loss": 0.5245, "mean_token_accuracy": 0.861393392086029, "num_tokens": 773959603.0, "step": 20381 }, { "epoch": 2.5927998982317773, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.34149694442749, "learning_rate": 1e-05, "loss": 0.5158, "mean_token_accuracy": 0.8653069138526917, "num_tokens": 773996550.0, "step": 20382 }, { "epoch": 2.592927108510368, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.663306713104248, "learning_rate": 1e-05, "loss": 0.4718, "mean_token_accuracy": 0.8776320219039917, "num_tokens": 774033815.0, "step": 20383 }, { "epoch": 2.5930543187889583, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.176558017730713, "learning_rate": 1e-05, "loss": 0.4708, "mean_token_accuracy": 0.8764305114746094, "num_tokens": 774068930.0, "step": 20384 }, { "epoch": 2.5931815290675484, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.766543388366699, "learning_rate": 1e-05, "loss": 0.4827, "mean_token_accuracy": 0.8755464553833008, "num_tokens": 774112014.0, "step": 20385 }, { "epoch": 2.5933087393461394, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 6.008920192718506, "learning_rate": 1e-05, "loss": 0.4776, "mean_token_accuracy": 0.8758859634399414, "num_tokens": 774149676.0, "step": 20386 }, { "epoch": 2.5934359496247295, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.991711616516113, "learning_rate": 1e-05, "loss": 0.4337, "mean_token_accuracy": 0.8955899477005005, "num_tokens": 774190160.0, "step": 20387 }, { "epoch": 2.5935631599033204, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.888028621673584, "learning_rate": 1e-05, "loss": 0.4784, "mean_token_accuracy": 0.8746217489242554, "num_tokens": 774235133.0, "step": 20388 }, { "epoch": 2.5936903701819105, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00011491775512695312, "grad_norm": 7.32258415222168, "learning_rate": 1e-05, "loss": 0.4775, "mean_token_accuracy": 0.8829635381698608, "num_tokens": 774277962.0, "step": 20389 }, { "epoch": 2.5938175804605015, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 5.940078258514404, "learning_rate": 1e-05, "loss": 0.4613, "mean_token_accuracy": 0.8791056275367737, "num_tokens": 774316946.0, "step": 20390 }, { "epoch": 2.5939447907390916, "ewc_loss": 0.150390625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.0001220703125, "grad_norm": 7.619153022766113, "learning_rate": 1e-05, "loss": 0.489, "mean_token_accuracy": 0.8799306750297546, "num_tokens": 774356376.0, "step": 20391 }, { "epoch": 2.594072001017682, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.247298717498779, "learning_rate": 1e-05, "loss": 0.4447, "mean_token_accuracy": 0.8881406784057617, "num_tokens": 774392543.0, "step": 20392 }, { "epoch": 2.5941992112962726, "ewc_loss": 0.150390625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00012111663818359375, "grad_norm": 7.580626964569092, "learning_rate": 1e-05, "loss": 0.52, "mean_token_accuracy": 0.8722730875015259, "num_tokens": 774432114.0, "step": 20393 }, { "epoch": 2.594326421574863, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 6.5924506187438965, "learning_rate": 1e-05, "loss": 0.5127, "mean_token_accuracy": 0.8683258891105652, "num_tokens": 774468758.0, "step": 20394 }, { "epoch": 2.5944536318534537, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00011348724365234375, "grad_norm": 7.176347255706787, "learning_rate": 1e-05, "loss": 0.5363, "mean_token_accuracy": 0.8648297786712646, "num_tokens": 774503136.0, "step": 20395 }, { "epoch": 2.594580842132044, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.718059062957764, "learning_rate": 1e-05, "loss": 0.4584, "mean_token_accuracy": 0.886523425579071, "num_tokens": 774545162.0, "step": 20396 }, { "epoch": 2.5947080524106347, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.650405406951904, "learning_rate": 1e-05, "loss": 0.4823, "mean_token_accuracy": 0.8774312138557434, "num_tokens": 774581831.0, "step": 20397 }, { "epoch": 2.5948352626892253, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 6.816021919250488, "learning_rate": 1e-05, "loss": 0.4692, "mean_token_accuracy": 0.883170485496521, "num_tokens": 774619258.0, "step": 20398 }, { "epoch": 2.594962472967816, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.176075458526611, "learning_rate": 1e-05, "loss": 0.5085, "mean_token_accuracy": 0.8636302947998047, "num_tokens": 774650244.0, "step": 20399 }, { "epoch": 2.5950896832464063, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010967254638671875, "grad_norm": 7.032303810119629, "learning_rate": 1e-05, "loss": 0.5363, "mean_token_accuracy": 0.8636336326599121, "num_tokens": 774690060.0, "step": 20400 }, { "epoch": 2.595216893524997, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.871169090270996, "learning_rate": 1e-05, "loss": 0.4904, "mean_token_accuracy": 0.8725144863128662, "num_tokens": 774733925.0, "step": 20401 }, { "epoch": 2.5953441038035874, "ewc_loss": 0.1455078125, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00011682510375976562, "grad_norm": 7.412424564361572, "learning_rate": 1e-05, "loss": 0.4722, "mean_token_accuracy": 0.8821244239807129, "num_tokens": 774769344.0, "step": 20402 }, { "epoch": 2.595471314082178, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 5.868265151977539, "learning_rate": 1e-05, "loss": 0.4218, "mean_token_accuracy": 0.8937597274780273, "num_tokens": 774809548.0, "step": 20403 }, { "epoch": 2.5955985243607684, "ewc_loss": 0.158203125, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00012874603271484375, "grad_norm": 7.810033321380615, "learning_rate": 1e-05, "loss": 0.5225, "mean_token_accuracy": 0.8741397261619568, "num_tokens": 774848042.0, "step": 20404 }, { "epoch": 2.595725734639359, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.281808853149414, "learning_rate": 1e-05, "loss": 0.5023, "mean_token_accuracy": 0.8708086013793945, "num_tokens": 774884226.0, "step": 20405 }, { "epoch": 2.5958529449179495, "ewc_loss": 0.158203125, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00012874603271484375, "grad_norm": 7.811709403991699, "learning_rate": 1e-05, "loss": 0.5049, "mean_token_accuracy": 0.8737412691116333, "num_tokens": 774913697.0, "step": 20406 }, { "epoch": 2.59598015519654, "ewc_loss": 0.140625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00011157989501953125, "grad_norm": 6.68549108505249, "learning_rate": 1e-05, "loss": 0.4451, "mean_token_accuracy": 0.8926421999931335, "num_tokens": 774948369.0, "step": 20407 }, { "epoch": 2.5961073654751305, "ewc_loss": 0.1484375, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00011920928955078125, "grad_norm": 7.439816474914551, "learning_rate": 1e-05, "loss": 0.4938, "mean_token_accuracy": 0.8803437352180481, "num_tokens": 774992260.0, "step": 20408 }, { "epoch": 2.596234575753721, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 6.762983798980713, "learning_rate": 1e-05, "loss": 0.607, "mean_token_accuracy": 0.8358235359191895, "num_tokens": 775023613.0, "step": 20409 }, { "epoch": 2.596361786032311, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00011014938354492188, "grad_norm": 7.044455528259277, "learning_rate": 1e-05, "loss": 0.5147, "mean_token_accuracy": 0.8739532828330994, "num_tokens": 775063874.0, "step": 20410 }, { "epoch": 2.596488996310902, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.64237117767334, "learning_rate": 1e-05, "loss": 0.505, "mean_token_accuracy": 0.8715813159942627, "num_tokens": 775104447.0, "step": 20411 }, { "epoch": 2.596616206589492, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.7660136222839355, "learning_rate": 1e-05, "loss": 0.4614, "mean_token_accuracy": 0.8848116993904114, "num_tokens": 775143472.0, "step": 20412 }, { "epoch": 2.596743416868083, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.506555080413818, "learning_rate": 1e-05, "loss": 0.4775, "mean_token_accuracy": 0.8820877075195312, "num_tokens": 775176274.0, "step": 20413 }, { "epoch": 2.5968706271466733, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.636592388153076, "learning_rate": 1e-05, "loss": 0.5052, "mean_token_accuracy": 0.8709686398506165, "num_tokens": 775216485.0, "step": 20414 }, { "epoch": 2.596997837425264, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.324458122253418, "learning_rate": 1e-05, "loss": 0.445, "mean_token_accuracy": 0.8861554265022278, "num_tokens": 775256828.0, "step": 20415 }, { "epoch": 2.5971250477038543, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.663589000701904, "learning_rate": 1e-05, "loss": 0.5109, "mean_token_accuracy": 0.8670711517333984, "num_tokens": 775296577.0, "step": 20416 }, { "epoch": 2.597252257982445, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.152939796447754, "learning_rate": 1e-05, "loss": 0.4803, "mean_token_accuracy": 0.8746157884597778, "num_tokens": 775329567.0, "step": 20417 }, { "epoch": 2.5973794682610354, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.799413204193115, "learning_rate": 1e-05, "loss": 0.4707, "mean_token_accuracy": 0.8803329467773438, "num_tokens": 775363729.0, "step": 20418 }, { "epoch": 2.597506678539626, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 5.982218265533447, "learning_rate": 1e-05, "loss": 0.4706, "mean_token_accuracy": 0.8763293027877808, "num_tokens": 775397547.0, "step": 20419 }, { "epoch": 2.5976338888182164, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010967254638671875, "grad_norm": 7.0528435707092285, "learning_rate": 1e-05, "loss": 0.515, "mean_token_accuracy": 0.8676036596298218, "num_tokens": 775441033.0, "step": 20420 }, { "epoch": 2.597761099096807, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.867027282714844, "learning_rate": 1e-05, "loss": 0.448, "mean_token_accuracy": 0.8828491568565369, "num_tokens": 775477516.0, "step": 20421 }, { "epoch": 2.5978883093753975, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00011730194091796875, "grad_norm": 7.4383769035339355, "learning_rate": 1e-05, "loss": 0.4825, "mean_token_accuracy": 0.8820945024490356, "num_tokens": 775515573.0, "step": 20422 }, { "epoch": 2.598015519653988, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 5.957430839538574, "learning_rate": 1e-05, "loss": 0.4686, "mean_token_accuracy": 0.8819555640220642, "num_tokens": 775552970.0, "step": 20423 }, { "epoch": 2.5981427299325786, "ewc_loss": 0.15625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00012683868408203125, "grad_norm": 7.734962463378906, "learning_rate": 1e-05, "loss": 0.5193, "mean_token_accuracy": 0.8737971782684326, "num_tokens": 775590180.0, "step": 20424 }, { "epoch": 2.598269940211169, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.341185569763184, "learning_rate": 1e-05, "loss": 0.4761, "mean_token_accuracy": 0.8819235563278198, "num_tokens": 775627346.0, "step": 20425 }, { "epoch": 2.5983971504897596, "ewc_loss": 0.15234375, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.0001239776611328125, "grad_norm": 7.632511138916016, "learning_rate": 1e-05, "loss": 0.5101, "mean_token_accuracy": 0.8762596845626831, "num_tokens": 775667173.0, "step": 20426 }, { "epoch": 2.59852436076835, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010967254638671875, "grad_norm": 6.69697380065918, "learning_rate": 1e-05, "loss": 0.5254, "mean_token_accuracy": 0.864150881767273, "num_tokens": 775702710.0, "step": 20427 }, { "epoch": 2.5986515710469407, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00011301040649414062, "grad_norm": 7.170219898223877, "learning_rate": 1e-05, "loss": 0.5002, "mean_token_accuracy": 0.8753335475921631, "num_tokens": 775743934.0, "step": 20428 }, { "epoch": 2.598778781325531, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 6.785821437835693, "learning_rate": 1e-05, "loss": 0.4794, "mean_token_accuracy": 0.8802198171615601, "num_tokens": 775783184.0, "step": 20429 }, { "epoch": 2.5989059916041217, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.649608612060547, "learning_rate": 1e-05, "loss": 0.4912, "mean_token_accuracy": 0.8700273036956787, "num_tokens": 775820019.0, "step": 20430 }, { "epoch": 2.5990332018827123, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.842824459075928, "learning_rate": 1e-05, "loss": 0.4886, "mean_token_accuracy": 0.8764088153839111, "num_tokens": 775858740.0, "step": 20431 }, { "epoch": 2.599160412161303, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.202451229095459, "learning_rate": 1e-05, "loss": 0.4757, "mean_token_accuracy": 0.8768603205680847, "num_tokens": 775900615.0, "step": 20432 }, { "epoch": 2.5992876224398933, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 7.0347580909729, "learning_rate": 1e-05, "loss": 0.4574, "mean_token_accuracy": 0.8871117830276489, "num_tokens": 775938953.0, "step": 20433 }, { "epoch": 2.599414832718484, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.949347496032715, "learning_rate": 1e-05, "loss": 0.4776, "mean_token_accuracy": 0.8724066615104675, "num_tokens": 775973203.0, "step": 20434 }, { "epoch": 2.599542042997074, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00011491775512695312, "grad_norm": 7.320331573486328, "learning_rate": 1e-05, "loss": 0.535, "mean_token_accuracy": 0.8653786182403564, "num_tokens": 776014984.0, "step": 20435 }, { "epoch": 2.599669253275665, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 5.909852504730225, "learning_rate": 1e-05, "loss": 0.5211, "mean_token_accuracy": 0.8607866764068604, "num_tokens": 776059883.0, "step": 20436 }, { "epoch": 2.599796463554255, "ewc_loss": 0.150390625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.0001220703125, "grad_norm": 7.646273136138916, "learning_rate": 1e-05, "loss": 0.5425, "mean_token_accuracy": 0.8628802299499512, "num_tokens": 776101025.0, "step": 20437 }, { "epoch": 2.599923673832846, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.175896644592285, "learning_rate": 1e-05, "loss": 0.4511, "mean_token_accuracy": 0.8824474811553955, "num_tokens": 776134356.0, "step": 20438 }, { "epoch": 2.600050884111436, "ewc_loss": 0.154296875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.000125885009765625, "grad_norm": 7.7137556076049805, "learning_rate": 1e-05, "loss": 0.4823, "mean_token_accuracy": 0.8813415169715881, "num_tokens": 776176158.0, "step": 20439 }, { "epoch": 2.6001780943900266, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.513509273529053, "learning_rate": 1e-05, "loss": 0.4752, "mean_token_accuracy": 0.8778377771377563, "num_tokens": 776209586.0, "step": 20440 }, { "epoch": 2.600305304668617, "ewc_loss": 0.1494140625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00012063980102539062, "grad_norm": 7.493595123291016, "learning_rate": 1e-05, "loss": 0.5194, "mean_token_accuracy": 0.8684896230697632, "num_tokens": 776240707.0, "step": 20441 }, { "epoch": 2.6004325149472076, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 6.649982929229736, "learning_rate": 1e-05, "loss": 0.4946, "mean_token_accuracy": 0.8747856020927429, "num_tokens": 776280448.0, "step": 20442 }, { "epoch": 2.600559725225798, "ewc_loss": 0.140625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.0001125335693359375, "grad_norm": 7.147248268127441, "learning_rate": 1e-05, "loss": 0.465, "mean_token_accuracy": 0.889677882194519, "num_tokens": 776320112.0, "step": 20443 }, { "epoch": 2.6006869355043887, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.607359409332275, "learning_rate": 1e-05, "loss": 0.4991, "mean_token_accuracy": 0.8743836283683777, "num_tokens": 776362720.0, "step": 20444 }, { "epoch": 2.600814145782979, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.858704090118408, "learning_rate": 1e-05, "loss": 0.4937, "mean_token_accuracy": 0.8736410140991211, "num_tokens": 776403006.0, "step": 20445 }, { "epoch": 2.6009413560615697, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.518078327178955, "learning_rate": 1e-05, "loss": 0.4833, "mean_token_accuracy": 0.8736119866371155, "num_tokens": 776436039.0, "step": 20446 }, { "epoch": 2.6010685663401603, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.676867961883545, "learning_rate": 1e-05, "loss": 0.4836, "mean_token_accuracy": 0.8762876391410828, "num_tokens": 776476337.0, "step": 20447 }, { "epoch": 2.601195776618751, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.381234645843506, "learning_rate": 1e-05, "loss": 0.4384, "mean_token_accuracy": 0.887284517288208, "num_tokens": 776512552.0, "step": 20448 }, { "epoch": 2.6013229868973413, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.643558979034424, "learning_rate": 1e-05, "loss": 0.4916, "mean_token_accuracy": 0.8736304044723511, "num_tokens": 776549727.0, "step": 20449 }, { "epoch": 2.601450197175932, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.232572078704834, "learning_rate": 1e-05, "loss": 0.4951, "mean_token_accuracy": 0.8740371465682983, "num_tokens": 776586701.0, "step": 20450 }, { "epoch": 2.6015774074545224, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.712507724761963, "learning_rate": 1e-05, "loss": 0.5088, "mean_token_accuracy": 0.8738238215446472, "num_tokens": 776627386.0, "step": 20451 }, { "epoch": 2.601704617733113, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.068594455718994, "learning_rate": 1e-05, "loss": 0.4596, "mean_token_accuracy": 0.8807225227355957, "num_tokens": 776668389.0, "step": 20452 }, { "epoch": 2.6018318280117034, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.896631717681885, "learning_rate": 1e-05, "loss": 0.5041, "mean_token_accuracy": 0.8726451396942139, "num_tokens": 776708106.0, "step": 20453 }, { "epoch": 2.601959038290294, "ewc_loss": 0.11474609375, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.927919387817383, "learning_rate": 1e-05, "loss": 0.4668, "mean_token_accuracy": 0.8765964508056641, "num_tokens": 776744577.0, "step": 20454 }, { "epoch": 2.6020862485688845, "ewc_loss": 0.140625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.0001125335693359375, "grad_norm": 7.212552070617676, "learning_rate": 1e-05, "loss": 0.5261, "mean_token_accuracy": 0.8640741109848022, "num_tokens": 776784350.0, "step": 20455 }, { "epoch": 2.602213458847475, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 5.885844707489014, "learning_rate": 1e-05, "loss": 0.454, "mean_token_accuracy": 0.8848254680633545, "num_tokens": 776823728.0, "step": 20456 }, { "epoch": 2.6023406691260655, "ewc_loss": 0.150390625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00012159347534179688, "grad_norm": 7.577188968658447, "learning_rate": 1e-05, "loss": 0.5081, "mean_token_accuracy": 0.87718266248703, "num_tokens": 776860097.0, "step": 20457 }, { "epoch": 2.6024678794046556, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.119831085205078, "learning_rate": 1e-05, "loss": 0.4631, "mean_token_accuracy": 0.878974199295044, "num_tokens": 776895876.0, "step": 20458 }, { "epoch": 2.6025950896832466, "ewc_loss": 0.154296875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.000125885009765625, "grad_norm": 7.7087321281433105, "learning_rate": 1e-05, "loss": 0.5211, "mean_token_accuracy": 0.8716281056404114, "num_tokens": 776930003.0, "step": 20459 }, { "epoch": 2.6027222999618367, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.504026889801025, "learning_rate": 1e-05, "loss": 0.4236, "mean_token_accuracy": 0.895843505859375, "num_tokens": 776969699.0, "step": 20460 }, { "epoch": 2.6028495102404277, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00011777877807617188, "grad_norm": 7.421888828277588, "learning_rate": 1e-05, "loss": 0.4913, "mean_token_accuracy": 0.8784351944923401, "num_tokens": 777006517.0, "step": 20461 }, { "epoch": 2.6029767205190177, "ewc_loss": 0.140625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00011110305786132812, "grad_norm": 6.729990005493164, "learning_rate": 1e-05, "loss": 0.5546, "mean_token_accuracy": 0.858208179473877, "num_tokens": 777044160.0, "step": 20462 }, { "epoch": 2.6031039307976087, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.92003870010376, "learning_rate": 1e-05, "loss": 0.4862, "mean_token_accuracy": 0.8799434304237366, "num_tokens": 777079890.0, "step": 20463 }, { "epoch": 2.603231141076199, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.771974563598633, "learning_rate": 1e-05, "loss": 0.4504, "mean_token_accuracy": 0.8892972469329834, "num_tokens": 777120928.0, "step": 20464 }, { "epoch": 2.6033583513547893, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.408811092376709, "learning_rate": 1e-05, "loss": 0.5174, "mean_token_accuracy": 0.8649674654006958, "num_tokens": 777166363.0, "step": 20465 }, { "epoch": 2.60348556163338, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.9037861824035645, "learning_rate": 1e-05, "loss": 0.5005, "mean_token_accuracy": 0.876369297504425, "num_tokens": 777201146.0, "step": 20466 }, { "epoch": 2.6036127719119704, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 6.054256439208984, "learning_rate": 1e-05, "loss": 0.5374, "mean_token_accuracy": 0.8566656708717346, "num_tokens": 777239884.0, "step": 20467 }, { "epoch": 2.603739982190561, "ewc_loss": 0.140625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00011110305786132812, "grad_norm": 7.147643089294434, "learning_rate": 1e-05, "loss": 0.5494, "mean_token_accuracy": 0.8612804412841797, "num_tokens": 777279898.0, "step": 20468 }, { "epoch": 2.6038671924691514, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.887886047363281, "learning_rate": 1e-05, "loss": 0.5005, "mean_token_accuracy": 0.8709973096847534, "num_tokens": 777320614.0, "step": 20469 }, { "epoch": 2.603994402747742, "ewc_loss": 0.1484375, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00011968612670898438, "grad_norm": 7.49876070022583, "learning_rate": 1e-05, "loss": 0.5134, "mean_token_accuracy": 0.8720607757568359, "num_tokens": 777364782.0, "step": 20470 }, { "epoch": 2.6041216130263325, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.014697074890137, "learning_rate": 1e-05, "loss": 0.4887, "mean_token_accuracy": 0.8687517046928406, "num_tokens": 777398276.0, "step": 20471 }, { "epoch": 2.604248823304923, "ewc_loss": 0.154296875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00012493133544921875, "grad_norm": 7.731415748596191, "learning_rate": 1e-05, "loss": 0.4936, "mean_token_accuracy": 0.8832061886787415, "num_tokens": 777434510.0, "step": 20472 }, { "epoch": 2.6043760335835135, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.346382141113281, "learning_rate": 1e-05, "loss": 0.479, "mean_token_accuracy": 0.8773802518844604, "num_tokens": 777469266.0, "step": 20473 }, { "epoch": 2.604503243862104, "ewc_loss": 0.15234375, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.656373023986816, "learning_rate": 1e-05, "loss": 0.459, "mean_token_accuracy": 0.8888005614280701, "num_tokens": 777500770.0, "step": 20474 }, { "epoch": 2.6046304541406946, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 6.613707065582275, "learning_rate": 1e-05, "loss": 0.4922, "mean_token_accuracy": 0.8777986168861389, "num_tokens": 777537977.0, "step": 20475 }, { "epoch": 2.604757664419285, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.0001163482666015625, "grad_norm": 7.346168518066406, "learning_rate": 1e-05, "loss": 0.4839, "mean_token_accuracy": 0.8794841766357422, "num_tokens": 777574065.0, "step": 20476 }, { "epoch": 2.6048848746978757, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.645153045654297, "learning_rate": 1e-05, "loss": 0.466, "mean_token_accuracy": 0.88517165184021, "num_tokens": 777613366.0, "step": 20477 }, { "epoch": 2.605012084976466, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00011014938354492188, "grad_norm": 7.011290550231934, "learning_rate": 1e-05, "loss": 0.5075, "mean_token_accuracy": 0.8728134632110596, "num_tokens": 777650954.0, "step": 20478 }, { "epoch": 2.6051392952550567, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.571355819702148, "learning_rate": 1e-05, "loss": 0.5209, "mean_token_accuracy": 0.8638817667961121, "num_tokens": 777694370.0, "step": 20479 }, { "epoch": 2.6052665055336472, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.755894660949707, "learning_rate": 1e-05, "loss": 0.4969, "mean_token_accuracy": 0.8746618032455444, "num_tokens": 777734039.0, "step": 20480 }, { "epoch": 2.6053937158122378, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.452398300170898, "learning_rate": 1e-05, "loss": 0.4831, "mean_token_accuracy": 0.8752380013465881, "num_tokens": 777776237.0, "step": 20481 }, { "epoch": 2.6055209260908283, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.647892475128174, "learning_rate": 1e-05, "loss": 0.4703, "mean_token_accuracy": 0.8813093900680542, "num_tokens": 777813767.0, "step": 20482 }, { "epoch": 2.6056481363694184, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.312257766723633, "learning_rate": 1e-05, "loss": 0.4559, "mean_token_accuracy": 0.8790817260742188, "num_tokens": 777848642.0, "step": 20483 }, { "epoch": 2.6057753466480094, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.652723789215088, "learning_rate": 1e-05, "loss": 0.4555, "mean_token_accuracy": 0.8823736906051636, "num_tokens": 777885748.0, "step": 20484 }, { "epoch": 2.6059025569265994, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.154915809631348, "learning_rate": 1e-05, "loss": 0.5102, "mean_token_accuracy": 0.8643103837966919, "num_tokens": 777922366.0, "step": 20485 }, { "epoch": 2.6060297672051904, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.775213241577148, "learning_rate": 1e-05, "loss": 0.4962, "mean_token_accuracy": 0.8751169443130493, "num_tokens": 777965130.0, "step": 20486 }, { "epoch": 2.6061569774837805, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 5.9878034591674805, "learning_rate": 1e-05, "loss": 0.422, "mean_token_accuracy": 0.8907040953636169, "num_tokens": 777999425.0, "step": 20487 }, { "epoch": 2.6062841877623715, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 7.023210525512695, "learning_rate": 1e-05, "loss": 0.5246, "mean_token_accuracy": 0.8644008636474609, "num_tokens": 778042378.0, "step": 20488 }, { "epoch": 2.6064113980409616, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.872708320617676, "learning_rate": 1e-05, "loss": 0.4646, "mean_token_accuracy": 0.8783622980117798, "num_tokens": 778083858.0, "step": 20489 }, { "epoch": 2.606538608319552, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00011730194091796875, "grad_norm": 7.37775993347168, "learning_rate": 1e-05, "loss": 0.4853, "mean_token_accuracy": 0.8840377926826477, "num_tokens": 778122103.0, "step": 20490 }, { "epoch": 2.6066658185981426, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 5.9392547607421875, "learning_rate": 1e-05, "loss": 0.4738, "mean_token_accuracy": 0.8770254850387573, "num_tokens": 778158207.0, "step": 20491 }, { "epoch": 2.606793028876733, "ewc_loss": 0.154296875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.000125885009765625, "grad_norm": 7.69618558883667, "learning_rate": 1e-05, "loss": 0.4935, "mean_token_accuracy": 0.8825111985206604, "num_tokens": 778191542.0, "step": 20492 }, { "epoch": 2.6069202391553237, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.284298896789551, "learning_rate": 1e-05, "loss": 0.483, "mean_token_accuracy": 0.8785910606384277, "num_tokens": 778234929.0, "step": 20493 }, { "epoch": 2.607047449433914, "ewc_loss": 0.15234375, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.0001239776611328125, "grad_norm": 7.638322830200195, "learning_rate": 1e-05, "loss": 0.4437, "mean_token_accuracy": 0.8936905264854431, "num_tokens": 778273225.0, "step": 20494 }, { "epoch": 2.6071746597125047, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010967254638671875, "grad_norm": 6.650843620300293, "learning_rate": 1e-05, "loss": 0.4996, "mean_token_accuracy": 0.8716377019882202, "num_tokens": 778310998.0, "step": 20495 }, { "epoch": 2.6073018699910953, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00011396408081054688, "grad_norm": 7.207168102264404, "learning_rate": 1e-05, "loss": 0.5177, "mean_token_accuracy": 0.8687253594398499, "num_tokens": 778348379.0, "step": 20496 }, { "epoch": 2.607429080269686, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010967254638671875, "grad_norm": 6.774749755859375, "learning_rate": 1e-05, "loss": 0.5083, "mean_token_accuracy": 0.8717910051345825, "num_tokens": 778390742.0, "step": 20497 }, { "epoch": 2.6075562905482763, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.6671342849731445, "learning_rate": 1e-05, "loss": 0.4645, "mean_token_accuracy": 0.8825070858001709, "num_tokens": 778429239.0, "step": 20498 }, { "epoch": 2.607683500826867, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.837592124938965, "learning_rate": 1e-05, "loss": 0.4557, "mean_token_accuracy": 0.886465311050415, "num_tokens": 778467606.0, "step": 20499 }, { "epoch": 2.6078107111054574, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.218087196350098, "learning_rate": 1e-05, "loss": 0.4955, "mean_token_accuracy": 0.8680334091186523, "num_tokens": 778500790.0, "step": 20500 }, { "epoch": 2.607937921384048, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 7.036171913146973, "learning_rate": 1e-05, "loss": 0.4825, "mean_token_accuracy": 0.8792523145675659, "num_tokens": 778536412.0, "step": 20501 }, { "epoch": 2.6080651316626384, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.929396152496338, "learning_rate": 1e-05, "loss": 0.4865, "mean_token_accuracy": 0.8709904551506042, "num_tokens": 778579812.0, "step": 20502 }, { "epoch": 2.608192341941229, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00011539459228515625, "grad_norm": 7.344367980957031, "learning_rate": 1e-05, "loss": 0.4686, "mean_token_accuracy": 0.8839744329452515, "num_tokens": 778626135.0, "step": 20503 }, { "epoch": 2.6083195522198195, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 5.912729263305664, "learning_rate": 1e-05, "loss": 0.4566, "mean_token_accuracy": 0.8836144208908081, "num_tokens": 778666163.0, "step": 20504 }, { "epoch": 2.60844676249841, "ewc_loss": 0.154296875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.000125885009765625, "grad_norm": 7.685992240905762, "learning_rate": 1e-05, "loss": 0.5076, "mean_token_accuracy": 0.8790585994720459, "num_tokens": 778703168.0, "step": 20505 }, { "epoch": 2.6085739727770005, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.218827724456787, "learning_rate": 1e-05, "loss": 0.5092, "mean_token_accuracy": 0.8679524660110474, "num_tokens": 778741696.0, "step": 20506 }, { "epoch": 2.608701183055591, "ewc_loss": 0.154296875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.000125885009765625, "grad_norm": 7.718979358673096, "learning_rate": 1e-05, "loss": 0.5092, "mean_token_accuracy": 0.875914990901947, "num_tokens": 778777450.0, "step": 20507 }, { "epoch": 2.608828393334181, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 6.555571556091309, "learning_rate": 1e-05, "loss": 0.4837, "mean_token_accuracy": 0.878633975982666, "num_tokens": 778811448.0, "step": 20508 }, { "epoch": 2.608955603612772, "ewc_loss": 0.1484375, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00011968612670898438, "grad_norm": 7.451701641082764, "learning_rate": 1e-05, "loss": 0.5027, "mean_token_accuracy": 0.8769278526306152, "num_tokens": 778849333.0, "step": 20509 }, { "epoch": 2.609082813891362, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 6.6550679206848145, "learning_rate": 1e-05, "loss": 0.4756, "mean_token_accuracy": 0.8836669325828552, "num_tokens": 778885264.0, "step": 20510 }, { "epoch": 2.609210024169953, "ewc_loss": 0.140625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00011205673217773438, "grad_norm": 7.121099948883057, "learning_rate": 1e-05, "loss": 0.4944, "mean_token_accuracy": 0.8764289617538452, "num_tokens": 778918160.0, "step": 20511 }, { "epoch": 2.6093372344485433, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.5822038650512695, "learning_rate": 1e-05, "loss": 0.4731, "mean_token_accuracy": 0.8835045099258423, "num_tokens": 778955938.0, "step": 20512 }, { "epoch": 2.609464444727134, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.857805252075195, "learning_rate": 1e-05, "loss": 0.5051, "mean_token_accuracy": 0.8714817762374878, "num_tokens": 778994412.0, "step": 20513 }, { "epoch": 2.6095916550057243, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.4735188484191895, "learning_rate": 1e-05, "loss": 0.5223, "mean_token_accuracy": 0.8650099039077759, "num_tokens": 779030818.0, "step": 20514 }, { "epoch": 2.609718865284315, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.71274471282959, "learning_rate": 1e-05, "loss": 0.4624, "mean_token_accuracy": 0.8870556950569153, "num_tokens": 779070217.0, "step": 20515 }, { "epoch": 2.6098460755629054, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.335798263549805, "learning_rate": 1e-05, "loss": 0.4362, "mean_token_accuracy": 0.8896158933639526, "num_tokens": 779105668.0, "step": 20516 }, { "epoch": 2.609973285841496, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.703863620758057, "learning_rate": 1e-05, "loss": 0.4242, "mean_token_accuracy": 0.8974129557609558, "num_tokens": 779140005.0, "step": 20517 }, { "epoch": 2.6101004961200864, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.1832990646362305, "learning_rate": 1e-05, "loss": 0.4657, "mean_token_accuracy": 0.8799480199813843, "num_tokens": 779176269.0, "step": 20518 }, { "epoch": 2.610227706398677, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.802566051483154, "learning_rate": 1e-05, "loss": 0.4521, "mean_token_accuracy": 0.8899643421173096, "num_tokens": 779209914.0, "step": 20519 }, { "epoch": 2.6103549166772675, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.039849281311035, "learning_rate": 1e-05, "loss": 0.4443, "mean_token_accuracy": 0.8866720795631409, "num_tokens": 779244994.0, "step": 20520 }, { "epoch": 2.610482126955858, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.994073390960693, "learning_rate": 1e-05, "loss": 0.4628, "mean_token_accuracy": 0.8845680952072144, "num_tokens": 779283700.0, "step": 20521 }, { "epoch": 2.6106093372344485, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 5.931829929351807, "learning_rate": 1e-05, "loss": 0.4435, "mean_token_accuracy": 0.8879674673080444, "num_tokens": 779324827.0, "step": 20522 }, { "epoch": 2.610736547513039, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00011444091796875, "grad_norm": 7.293255805969238, "learning_rate": 1e-05, "loss": 0.5317, "mean_token_accuracy": 0.8683547973632812, "num_tokens": 779363593.0, "step": 20523 }, { "epoch": 2.6108637577916296, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 5.9766693115234375, "learning_rate": 1e-05, "loss": 0.4644, "mean_token_accuracy": 0.8810564875602722, "num_tokens": 779401295.0, "step": 20524 }, { "epoch": 2.61099096807022, "ewc_loss": 0.150390625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00012159347534179688, "grad_norm": 7.553197383880615, "learning_rate": 1e-05, "loss": 0.5499, "mean_token_accuracy": 0.8614791035652161, "num_tokens": 779440375.0, "step": 20525 }, { "epoch": 2.6111181783488107, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.235873699188232, "learning_rate": 1e-05, "loss": 0.4936, "mean_token_accuracy": 0.8741355538368225, "num_tokens": 779475881.0, "step": 20526 }, { "epoch": 2.611245388627401, "ewc_loss": 0.1484375, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00011920928955078125, "grad_norm": 7.529317378997803, "learning_rate": 1e-05, "loss": 0.534, "mean_token_accuracy": 0.8664132356643677, "num_tokens": 779520155.0, "step": 20527 }, { "epoch": 2.6113725989059917, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.533358573913574, "learning_rate": 1e-05, "loss": 0.495, "mean_token_accuracy": 0.8747061491012573, "num_tokens": 779563952.0, "step": 20528 }, { "epoch": 2.6114998091845822, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00011301040649414062, "grad_norm": 7.163806438446045, "learning_rate": 1e-05, "loss": 0.4685, "mean_token_accuracy": 0.8873105049133301, "num_tokens": 779601618.0, "step": 20529 }, { "epoch": 2.6116270194631728, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 6.6811137199401855, "learning_rate": 1e-05, "loss": 0.4964, "mean_token_accuracy": 0.871645987033844, "num_tokens": 779645228.0, "step": 20530 }, { "epoch": 2.6117542297417633, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.656850337982178, "learning_rate": 1e-05, "loss": 0.48, "mean_token_accuracy": 0.8756858706474304, "num_tokens": 779681780.0, "step": 20531 }, { "epoch": 2.611881440020354, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.79799747467041, "learning_rate": 1e-05, "loss": 0.4796, "mean_token_accuracy": 0.8789957761764526, "num_tokens": 779717783.0, "step": 20532 }, { "epoch": 2.612008650298944, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 6.173918724060059, "learning_rate": 1e-05, "loss": 0.4457, "mean_token_accuracy": 0.8835622072219849, "num_tokens": 779752542.0, "step": 20533 }, { "epoch": 2.612135860577535, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 7.022793292999268, "learning_rate": 1e-05, "loss": 0.4933, "mean_token_accuracy": 0.8756278157234192, "num_tokens": 779795134.0, "step": 20534 }, { "epoch": 2.612263070856125, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.872716426849365, "learning_rate": 1e-05, "loss": 0.445, "mean_token_accuracy": 0.8823399543762207, "num_tokens": 779828589.0, "step": 20535 }, { "epoch": 2.612390281134716, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00011777877807617188, "grad_norm": 7.405846118927002, "learning_rate": 1e-05, "loss": 0.5336, "mean_token_accuracy": 0.864616334438324, "num_tokens": 779865904.0, "step": 20536 }, { "epoch": 2.612517491413306, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 5.845146179199219, "learning_rate": 1e-05, "loss": 0.4585, "mean_token_accuracy": 0.8777152299880981, "num_tokens": 779910204.0, "step": 20537 }, { "epoch": 2.6126447016918966, "ewc_loss": 0.158203125, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00012874603271484375, "grad_norm": 7.822613716125488, "learning_rate": 1e-05, "loss": 0.4861, "mean_token_accuracy": 0.8838652968406677, "num_tokens": 779951032.0, "step": 20538 }, { "epoch": 2.612771911970487, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.2634758949279785, "learning_rate": 1e-05, "loss": 0.4468, "mean_token_accuracy": 0.8885578513145447, "num_tokens": 779983316.0, "step": 20539 }, { "epoch": 2.6128991222490776, "ewc_loss": 0.158203125, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00012874603271484375, "grad_norm": 7.854745864868164, "learning_rate": 1e-05, "loss": 0.5016, "mean_token_accuracy": 0.8764452934265137, "num_tokens": 780021167.0, "step": 20540 }, { "epoch": 2.613026332527668, "ewc_loss": 0.140625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.0001125335693359375, "grad_norm": 6.7014288902282715, "learning_rate": 1e-05, "loss": 0.5005, "mean_token_accuracy": 0.876339852809906, "num_tokens": 780061276.0, "step": 20541 }, { "epoch": 2.6131535428062587, "ewc_loss": 0.150390625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00012111663818359375, "grad_norm": 7.489809513092041, "learning_rate": 1e-05, "loss": 0.4816, "mean_token_accuracy": 0.8807471990585327, "num_tokens": 780097204.0, "step": 20542 }, { "epoch": 2.613280753084849, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 6.789611339569092, "learning_rate": 1e-05, "loss": 0.5044, "mean_token_accuracy": 0.8678973913192749, "num_tokens": 780137208.0, "step": 20543 }, { "epoch": 2.6134079633634397, "ewc_loss": 0.140625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00011157989501953125, "grad_norm": 7.088131427764893, "learning_rate": 1e-05, "loss": 0.5127, "mean_token_accuracy": 0.8702371716499329, "num_tokens": 780177815.0, "step": 20544 }, { "epoch": 2.6135351736420303, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.687244415283203, "learning_rate": 1e-05, "loss": 0.4407, "mean_token_accuracy": 0.8913582563400269, "num_tokens": 780215798.0, "step": 20545 }, { "epoch": 2.613662383920621, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.787201881408691, "learning_rate": 1e-05, "loss": 0.5208, "mean_token_accuracy": 0.8608163595199585, "num_tokens": 780252252.0, "step": 20546 }, { "epoch": 2.6137895941992113, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.550757884979248, "learning_rate": 1e-05, "loss": 0.4311, "mean_token_accuracy": 0.8939379453659058, "num_tokens": 780288100.0, "step": 20547 }, { "epoch": 2.613916804477802, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.628829002380371, "learning_rate": 1e-05, "loss": 0.4836, "mean_token_accuracy": 0.8770554065704346, "num_tokens": 780331922.0, "step": 20548 }, { "epoch": 2.6140440147563924, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.386745452880859, "learning_rate": 1e-05, "loss": 0.4893, "mean_token_accuracy": 0.8702470660209656, "num_tokens": 780369930.0, "step": 20549 }, { "epoch": 2.614171225034983, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.6153340339660645, "learning_rate": 1e-05, "loss": 0.4811, "mean_token_accuracy": 0.877973198890686, "num_tokens": 780403893.0, "step": 20550 }, { "epoch": 2.6142984353135734, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.218809127807617, "learning_rate": 1e-05, "loss": 0.4356, "mean_token_accuracy": 0.8901581764221191, "num_tokens": 780438426.0, "step": 20551 }, { "epoch": 2.614425645592164, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.719045162200928, "learning_rate": 1e-05, "loss": 0.4756, "mean_token_accuracy": 0.8802824020385742, "num_tokens": 780483267.0, "step": 20552 }, { "epoch": 2.6145528558707545, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 6.027639389038086, "learning_rate": 1e-05, "loss": 0.4624, "mean_token_accuracy": 0.8795914649963379, "num_tokens": 780520355.0, "step": 20553 }, { "epoch": 2.614680066149345, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.9384565353393555, "learning_rate": 1e-05, "loss": 0.4773, "mean_token_accuracy": 0.8775184154510498, "num_tokens": 780563449.0, "step": 20554 }, { "epoch": 2.6148072764279355, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.875181674957275, "learning_rate": 1e-05, "loss": 0.4223, "mean_token_accuracy": 0.8893764019012451, "num_tokens": 780598607.0, "step": 20555 }, { "epoch": 2.6149344867065256, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00011491775512695312, "grad_norm": 7.297067165374756, "learning_rate": 1e-05, "loss": 0.4961, "mean_token_accuracy": 0.8791201114654541, "num_tokens": 780638425.0, "step": 20556 }, { "epoch": 2.6150616969851166, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 5.8619160652160645, "learning_rate": 1e-05, "loss": 0.4976, "mean_token_accuracy": 0.8665298223495483, "num_tokens": 780683441.0, "step": 20557 }, { "epoch": 2.6151889072637067, "ewc_loss": 0.154296875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00012493133544921875, "grad_norm": 7.710831642150879, "learning_rate": 1e-05, "loss": 0.4622, "mean_token_accuracy": 0.8919870257377625, "num_tokens": 780720281.0, "step": 20558 }, { "epoch": 2.6153161175422976, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.170928001403809, "learning_rate": 1e-05, "loss": 0.4465, "mean_token_accuracy": 0.8888964056968689, "num_tokens": 780761038.0, "step": 20559 }, { "epoch": 2.6154433278208877, "ewc_loss": 0.158203125, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00012874603271484375, "grad_norm": 7.821550369262695, "learning_rate": 1e-05, "loss": 0.5805, "mean_token_accuracy": 0.8542472124099731, "num_tokens": 780802358.0, "step": 20560 }, { "epoch": 2.6155705380994787, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00011014938354492188, "grad_norm": 6.642623424530029, "learning_rate": 1e-05, "loss": 0.4702, "mean_token_accuracy": 0.8816633820533752, "num_tokens": 780839820.0, "step": 20561 }, { "epoch": 2.615697748378069, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00011777877807617188, "grad_norm": 7.436335563659668, "learning_rate": 1e-05, "loss": 0.4984, "mean_token_accuracy": 0.8771370649337769, "num_tokens": 780879969.0, "step": 20562 }, { "epoch": 2.6158249586566593, "ewc_loss": 0.140625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00011205673217773438, "grad_norm": 6.821291923522949, "learning_rate": 1e-05, "loss": 0.4851, "mean_token_accuracy": 0.8778440356254578, "num_tokens": 780916119.0, "step": 20563 }, { "epoch": 2.61595216893525, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.901618480682373, "learning_rate": 1e-05, "loss": 0.4838, "mean_token_accuracy": 0.8761452436447144, "num_tokens": 780957311.0, "step": 20564 }, { "epoch": 2.6160793792138404, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 6.83976411819458, "learning_rate": 1e-05, "loss": 0.5055, "mean_token_accuracy": 0.871013879776001, "num_tokens": 780998322.0, "step": 20565 }, { "epoch": 2.616206589492431, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.4129719734191895, "learning_rate": 1e-05, "loss": 0.5195, "mean_token_accuracy": 0.8623414635658264, "num_tokens": 781036249.0, "step": 20566 }, { "epoch": 2.6163337997710214, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 6.934104919433594, "learning_rate": 1e-05, "loss": 0.4798, "mean_token_accuracy": 0.8793798089027405, "num_tokens": 781075228.0, "step": 20567 }, { "epoch": 2.616461010049612, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.080536365509033, "learning_rate": 1e-05, "loss": 0.4365, "mean_token_accuracy": 0.8847161531448364, "num_tokens": 781110907.0, "step": 20568 }, { "epoch": 2.6165882203282025, "ewc_loss": 0.140625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00011157989501953125, "grad_norm": 7.1527509689331055, "learning_rate": 1e-05, "loss": 0.5272, "mean_token_accuracy": 0.8681873083114624, "num_tokens": 781148615.0, "step": 20569 }, { "epoch": 2.616715430606793, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 5.9492692947387695, "learning_rate": 1e-05, "loss": 0.4543, "mean_token_accuracy": 0.88340163230896, "num_tokens": 781181159.0, "step": 20570 }, { "epoch": 2.6168426408853835, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00011730194091796875, "grad_norm": 7.439577579498291, "learning_rate": 1e-05, "loss": 0.5775, "mean_token_accuracy": 0.8529170155525208, "num_tokens": 781221563.0, "step": 20571 }, { "epoch": 2.616969851163974, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.02785062789917, "learning_rate": 1e-05, "loss": 0.5091, "mean_token_accuracy": 0.864676296710968, "num_tokens": 781257606.0, "step": 20572 }, { "epoch": 2.6170970614425646, "ewc_loss": 0.154296875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00012493133544921875, "grad_norm": 7.646353244781494, "learning_rate": 1e-05, "loss": 0.5088, "mean_token_accuracy": 0.874323844909668, "num_tokens": 781296653.0, "step": 20573 }, { "epoch": 2.617224271721155, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.3002543449401855, "learning_rate": 1e-05, "loss": 0.4304, "mean_token_accuracy": 0.8927999138832092, "num_tokens": 781331596.0, "step": 20574 }, { "epoch": 2.6173514819997457, "ewc_loss": 0.15234375, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.617684841156006, "learning_rate": 1e-05, "loss": 0.4352, "mean_token_accuracy": 0.9009206891059875, "num_tokens": 781365599.0, "step": 20575 }, { "epoch": 2.617478692278336, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.541561126708984, "learning_rate": 1e-05, "loss": 0.4856, "mean_token_accuracy": 0.875878095626831, "num_tokens": 781406169.0, "step": 20576 }, { "epoch": 2.6176059025569267, "ewc_loss": 0.1455078125, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00011682510375976562, "grad_norm": 7.364830017089844, "learning_rate": 1e-05, "loss": 0.5086, "mean_token_accuracy": 0.8679544925689697, "num_tokens": 781444570.0, "step": 20577 }, { "epoch": 2.6177331128355172, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.6058831214904785, "learning_rate": 1e-05, "loss": 0.508, "mean_token_accuracy": 0.865996241569519, "num_tokens": 781482752.0, "step": 20578 }, { "epoch": 2.6178603231141078, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00011014938354492188, "grad_norm": 7.04372501373291, "learning_rate": 1e-05, "loss": 0.53, "mean_token_accuracy": 0.8651304841041565, "num_tokens": 781524549.0, "step": 20579 }, { "epoch": 2.6179875333926983, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.568624973297119, "learning_rate": 1e-05, "loss": 0.5264, "mean_token_accuracy": 0.8645952343940735, "num_tokens": 781560152.0, "step": 20580 }, { "epoch": 2.6181147436712884, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.775564670562744, "learning_rate": 1e-05, "loss": 0.4935, "mean_token_accuracy": 0.8752316832542419, "num_tokens": 781606399.0, "step": 20581 }, { "epoch": 2.6182419539498794, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.471046447753906, "learning_rate": 1e-05, "loss": 0.4894, "mean_token_accuracy": 0.8779551982879639, "num_tokens": 781646386.0, "step": 20582 }, { "epoch": 2.6183691642284694, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.637832164764404, "learning_rate": 1e-05, "loss": 0.5005, "mean_token_accuracy": 0.8717772960662842, "num_tokens": 781685629.0, "step": 20583 }, { "epoch": 2.6184963745070604, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.352068901062012, "learning_rate": 1e-05, "loss": 0.4158, "mean_token_accuracy": 0.8948974013328552, "num_tokens": 781717198.0, "step": 20584 }, { "epoch": 2.6186235847856505, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.621471881866455, "learning_rate": 1e-05, "loss": 0.5234, "mean_token_accuracy": 0.8633599281311035, "num_tokens": 781749281.0, "step": 20585 }, { "epoch": 2.6187507950642415, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.193875312805176, "learning_rate": 1e-05, "loss": 0.4577, "mean_token_accuracy": 0.8807927370071411, "num_tokens": 781791544.0, "step": 20586 }, { "epoch": 2.6188780053428315, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.694899559020996, "learning_rate": 1e-05, "loss": 0.4973, "mean_token_accuracy": 0.8707602024078369, "num_tokens": 781832864.0, "step": 20587 }, { "epoch": 2.619005215621422, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.038585186004639, "learning_rate": 1e-05, "loss": 0.4827, "mean_token_accuracy": 0.8781017065048218, "num_tokens": 781875757.0, "step": 20588 }, { "epoch": 2.6191324259000126, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.886629104614258, "learning_rate": 1e-05, "loss": 0.4762, "mean_token_accuracy": 0.8805355429649353, "num_tokens": 781911200.0, "step": 20589 }, { "epoch": 2.619259636178603, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.914943695068359, "learning_rate": 1e-05, "loss": 0.4334, "mean_token_accuracy": 0.8909620046615601, "num_tokens": 781948372.0, "step": 20590 }, { "epoch": 2.6193868464571937, "ewc_loss": 0.140625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.0001125335693359375, "grad_norm": 7.201736927032471, "learning_rate": 1e-05, "loss": 0.4998, "mean_token_accuracy": 0.8748988509178162, "num_tokens": 781985360.0, "step": 20591 }, { "epoch": 2.619514056735784, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 5.870145797729492, "learning_rate": 1e-05, "loss": 0.4398, "mean_token_accuracy": 0.8820033073425293, "num_tokens": 782019439.0, "step": 20592 }, { "epoch": 2.6196412670143747, "ewc_loss": 0.150390625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00012159347534179688, "grad_norm": 7.568124294281006, "learning_rate": 1e-05, "loss": 0.4194, "mean_token_accuracy": 0.9036364555358887, "num_tokens": 782055805.0, "step": 20593 }, { "epoch": 2.6197684772929652, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.085716724395752, "learning_rate": 1e-05, "loss": 0.491, "mean_token_accuracy": 0.8731576204299927, "num_tokens": 782092879.0, "step": 20594 }, { "epoch": 2.6198956875715558, "ewc_loss": 0.15625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00012683868408203125, "grad_norm": 7.735342979431152, "learning_rate": 1e-05, "loss": 0.5168, "mean_token_accuracy": 0.8734147548675537, "num_tokens": 782133123.0, "step": 20595 }, { "epoch": 2.6200228978501463, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 6.5035719871521, "learning_rate": 1e-05, "loss": 0.5301, "mean_token_accuracy": 0.8646153211593628, "num_tokens": 782169999.0, "step": 20596 }, { "epoch": 2.620150108128737, "ewc_loss": 0.1474609375, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00011873245239257812, "grad_norm": 7.448838710784912, "learning_rate": 1e-05, "loss": 0.4919, "mean_token_accuracy": 0.8809453248977661, "num_tokens": 782208269.0, "step": 20597 }, { "epoch": 2.6202773184073274, "ewc_loss": 0.140625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00011110305786132812, "grad_norm": 6.732702255249023, "learning_rate": 1e-05, "loss": 0.4993, "mean_token_accuracy": 0.8745055794715881, "num_tokens": 782242927.0, "step": 20598 }, { "epoch": 2.620404528685918, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.932317733764648, "learning_rate": 1e-05, "loss": 0.5028, "mean_token_accuracy": 0.8731399178504944, "num_tokens": 782281018.0, "step": 20599 }, { "epoch": 2.6205317389645084, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 6.805370330810547, "learning_rate": 1e-05, "loss": 0.5868, "mean_token_accuracy": 0.8492515087127686, "num_tokens": 782315336.0, "step": 20600 }, { "epoch": 2.620658949243099, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.416821479797363, "learning_rate": 1e-05, "loss": 0.4443, "mean_token_accuracy": 0.8901054859161377, "num_tokens": 782351163.0, "step": 20601 }, { "epoch": 2.6207861595216895, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 6.9673261642456055, "learning_rate": 1e-05, "loss": 0.5464, "mean_token_accuracy": 0.8599303364753723, "num_tokens": 782396955.0, "step": 20602 }, { "epoch": 2.62091336980028, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 6.027803421020508, "learning_rate": 1e-05, "loss": 0.4321, "mean_token_accuracy": 0.8895978927612305, "num_tokens": 782432005.0, "step": 20603 }, { "epoch": 2.6210405800788705, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00011348724365234375, "grad_norm": 7.2326436042785645, "learning_rate": 1e-05, "loss": 0.4917, "mean_token_accuracy": 0.8808956146240234, "num_tokens": 782467659.0, "step": 20604 }, { "epoch": 2.621167790357461, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 5.900332927703857, "learning_rate": 1e-05, "loss": 0.4733, "mean_token_accuracy": 0.875662088394165, "num_tokens": 782502621.0, "step": 20605 }, { "epoch": 2.621295000636051, "ewc_loss": 0.150390625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.0001220703125, "grad_norm": 7.587371826171875, "learning_rate": 1e-05, "loss": 0.4834, "mean_token_accuracy": 0.8856341242790222, "num_tokens": 782535803.0, "step": 20606 }, { "epoch": 2.621422210914642, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.089606285095215, "learning_rate": 1e-05, "loss": 0.4719, "mean_token_accuracy": 0.8765283823013306, "num_tokens": 782574720.0, "step": 20607 }, { "epoch": 2.621549421193232, "ewc_loss": 0.154296875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.000125885009765625, "grad_norm": 7.739317417144775, "learning_rate": 1e-05, "loss": 0.4857, "mean_token_accuracy": 0.8808624744415283, "num_tokens": 782615477.0, "step": 20608 }, { "epoch": 2.621676631471823, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.464025974273682, "learning_rate": 1e-05, "loss": 0.5138, "mean_token_accuracy": 0.8691357374191284, "num_tokens": 782653153.0, "step": 20609 }, { "epoch": 2.6218038417504133, "ewc_loss": 0.150390625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.0001220703125, "grad_norm": 7.5579328536987305, "learning_rate": 1e-05, "loss": 0.5376, "mean_token_accuracy": 0.8667073249816895, "num_tokens": 782691024.0, "step": 20610 }, { "epoch": 2.621931052029004, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 6.639538288116455, "learning_rate": 1e-05, "loss": 0.4887, "mean_token_accuracy": 0.8786782622337341, "num_tokens": 782732031.0, "step": 20611 }, { "epoch": 2.6220582623075943, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00011444091796875, "grad_norm": 7.228701591491699, "learning_rate": 1e-05, "loss": 0.5034, "mean_token_accuracy": 0.8743094801902771, "num_tokens": 782773430.0, "step": 20612 }, { "epoch": 2.622185472586185, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.61112117767334, "learning_rate": 1e-05, "loss": 0.4986, "mean_token_accuracy": 0.8718923330307007, "num_tokens": 782815852.0, "step": 20613 }, { "epoch": 2.6223126828647754, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 6.942600727081299, "learning_rate": 1e-05, "loss": 0.4693, "mean_token_accuracy": 0.88529372215271, "num_tokens": 782851327.0, "step": 20614 }, { "epoch": 2.622439893143366, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.515884876251221, "learning_rate": 1e-05, "loss": 0.5119, "mean_token_accuracy": 0.8660309314727783, "num_tokens": 782887382.0, "step": 20615 }, { "epoch": 2.6225671034219564, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.766537666320801, "learning_rate": 1e-05, "loss": 0.4501, "mean_token_accuracy": 0.8871743083000183, "num_tokens": 782923667.0, "step": 20616 }, { "epoch": 2.622694313700547, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.37406587600708, "learning_rate": 1e-05, "loss": 0.4717, "mean_token_accuracy": 0.8837308883666992, "num_tokens": 782961200.0, "step": 20617 }, { "epoch": 2.6228215239791375, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.733587741851807, "learning_rate": 1e-05, "loss": 0.4905, "mean_token_accuracy": 0.8750721216201782, "num_tokens": 782995832.0, "step": 20618 }, { "epoch": 2.622948734257728, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.208331108093262, "learning_rate": 1e-05, "loss": 0.4738, "mean_token_accuracy": 0.8787345290184021, "num_tokens": 783040880.0, "step": 20619 }, { "epoch": 2.6230759445363185, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.827369689941406, "learning_rate": 1e-05, "loss": 0.4543, "mean_token_accuracy": 0.8838315010070801, "num_tokens": 783075068.0, "step": 20620 }, { "epoch": 2.623203154814909, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 6.059389114379883, "learning_rate": 1e-05, "loss": 0.4618, "mean_token_accuracy": 0.880113422870636, "num_tokens": 783109321.0, "step": 20621 }, { "epoch": 2.6233303650934996, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 7.002264976501465, "learning_rate": 1e-05, "loss": 0.4892, "mean_token_accuracy": 0.8812371492385864, "num_tokens": 783145222.0, "step": 20622 }, { "epoch": 2.62345757537209, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 5.963155269622803, "learning_rate": 1e-05, "loss": 0.4938, "mean_token_accuracy": 0.8696106672286987, "num_tokens": 783185017.0, "step": 20623 }, { "epoch": 2.6235847856506807, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00011396408081054688, "grad_norm": 7.280717849731445, "learning_rate": 1e-05, "loss": 0.5179, "mean_token_accuracy": 0.8716555237770081, "num_tokens": 783228137.0, "step": 20624 }, { "epoch": 2.623711995929271, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 5.995389938354492, "learning_rate": 1e-05, "loss": 0.4813, "mean_token_accuracy": 0.8757525682449341, "num_tokens": 783266525.0, "step": 20625 }, { "epoch": 2.6238392062078617, "ewc_loss": 0.1494140625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00012063980102539062, "grad_norm": 7.5164337158203125, "learning_rate": 1e-05, "loss": 0.5169, "mean_token_accuracy": 0.8733447194099426, "num_tokens": 783301090.0, "step": 20626 }, { "epoch": 2.6239664164864522, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.240238666534424, "learning_rate": 1e-05, "loss": 0.4742, "mean_token_accuracy": 0.8791365623474121, "num_tokens": 783336816.0, "step": 20627 }, { "epoch": 2.6240936267650428, "ewc_loss": 0.1474609375, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00011873245239257812, "grad_norm": 7.4604034423828125, "learning_rate": 1e-05, "loss": 0.4686, "mean_token_accuracy": 0.888805091381073, "num_tokens": 783376766.0, "step": 20628 }, { "epoch": 2.6242208370436333, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.514980792999268, "learning_rate": 1e-05, "loss": 0.4487, "mean_token_accuracy": 0.8885064125061035, "num_tokens": 783416780.0, "step": 20629 }, { "epoch": 2.624348047322224, "ewc_loss": 0.140625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.0001125335693359375, "grad_norm": 7.112026691436768, "learning_rate": 1e-05, "loss": 0.458, "mean_token_accuracy": 0.8893669247627258, "num_tokens": 783449957.0, "step": 20630 }, { "epoch": 2.624475257600814, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 6.6761627197265625, "learning_rate": 1e-05, "loss": 0.4293, "mean_token_accuracy": 0.8948870897293091, "num_tokens": 783480488.0, "step": 20631 }, { "epoch": 2.624602467879405, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.590503215789795, "learning_rate": 1e-05, "loss": 0.4887, "mean_token_accuracy": 0.8746352791786194, "num_tokens": 783521296.0, "step": 20632 }, { "epoch": 2.624729678157995, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.800650119781494, "learning_rate": 1e-05, "loss": 0.5152, "mean_token_accuracy": 0.8675405383110046, "num_tokens": 783561937.0, "step": 20633 }, { "epoch": 2.624856888436586, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 6.115564346313477, "learning_rate": 1e-05, "loss": 0.4484, "mean_token_accuracy": 0.8836672902107239, "num_tokens": 783598928.0, "step": 20634 }, { "epoch": 2.624984098715176, "ewc_loss": 0.140625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00011110305786132812, "grad_norm": 7.076214790344238, "learning_rate": 1e-05, "loss": 0.4709, "mean_token_accuracy": 0.8830506801605225, "num_tokens": 783635538.0, "step": 20635 }, { "epoch": 2.6251113089937665, "ewc_loss": 0.11376953125, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.798494815826416, "learning_rate": 1e-05, "loss": 0.4267, "mean_token_accuracy": 0.8889461755752563, "num_tokens": 783675712.0, "step": 20636 }, { "epoch": 2.625238519272357, "ewc_loss": 0.1494140625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00012063980102539062, "grad_norm": 7.537280082702637, "learning_rate": 1e-05, "loss": 0.5026, "mean_token_accuracy": 0.8781792521476746, "num_tokens": 783715091.0, "step": 20637 }, { "epoch": 2.6253657295509476, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 5.859673023223877, "learning_rate": 1e-05, "loss": 0.44, "mean_token_accuracy": 0.8875988125801086, "num_tokens": 783753263.0, "step": 20638 }, { "epoch": 2.625492939829538, "ewc_loss": 0.162109375, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.000133514404296875, "grad_norm": 7.97366189956665, "learning_rate": 1e-05, "loss": 0.529, "mean_token_accuracy": 0.8706639409065247, "num_tokens": 783786251.0, "step": 20639 }, { "epoch": 2.6256201501081287, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.389847278594971, "learning_rate": 1e-05, "loss": 0.4835, "mean_token_accuracy": 0.8780335783958435, "num_tokens": 783829751.0, "step": 20640 }, { "epoch": 2.625747360386719, "ewc_loss": 0.16015625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00013065338134765625, "grad_norm": 7.909875392913818, "learning_rate": 1e-05, "loss": 0.5304, "mean_token_accuracy": 0.8687169551849365, "num_tokens": 783867786.0, "step": 20641 }, { "epoch": 2.6258745706653097, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00011444091796875, "grad_norm": 6.843844413757324, "learning_rate": 1e-05, "loss": 0.5253, "mean_token_accuracy": 0.8692303895950317, "num_tokens": 783904245.0, "step": 20642 }, { "epoch": 2.6260017809439002, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.000118255615234375, "grad_norm": 7.442582130432129, "learning_rate": 1e-05, "loss": 0.4573, "mean_token_accuracy": 0.8895683288574219, "num_tokens": 783946892.0, "step": 20643 }, { "epoch": 2.6261289912224908, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00011301040649414062, "grad_norm": 6.850193500518799, "learning_rate": 1e-05, "loss": 0.5406, "mean_token_accuracy": 0.866237998008728, "num_tokens": 783988132.0, "step": 20644 }, { "epoch": 2.6262562015010813, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00011014938354492188, "grad_norm": 7.023298740386963, "learning_rate": 1e-05, "loss": 0.5416, "mean_token_accuracy": 0.8688549995422363, "num_tokens": 784024321.0, "step": 20645 }, { "epoch": 2.626383411779672, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.713768482208252, "learning_rate": 1e-05, "loss": 0.5085, "mean_token_accuracy": 0.8699336051940918, "num_tokens": 784061266.0, "step": 20646 }, { "epoch": 2.6265106220582624, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.739387512207031, "learning_rate": 1e-05, "loss": 0.4954, "mean_token_accuracy": 0.8726953268051147, "num_tokens": 784101991.0, "step": 20647 }, { "epoch": 2.626637832336853, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.543738842010498, "learning_rate": 1e-05, "loss": 0.486, "mean_token_accuracy": 0.876265823841095, "num_tokens": 784137726.0, "step": 20648 }, { "epoch": 2.6267650426154434, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.618003845214844, "learning_rate": 1e-05, "loss": 0.4966, "mean_token_accuracy": 0.8727803826332092, "num_tokens": 784177773.0, "step": 20649 }, { "epoch": 2.626892252894034, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.369946002960205, "learning_rate": 1e-05, "loss": 0.4843, "mean_token_accuracy": 0.8762052655220032, "num_tokens": 784217429.0, "step": 20650 }, { "epoch": 2.6270194631726245, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.624650001525879, "learning_rate": 1e-05, "loss": 0.4444, "mean_token_accuracy": 0.8884384632110596, "num_tokens": 784252289.0, "step": 20651 }, { "epoch": 2.627146673451215, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.17942476272583, "learning_rate": 1e-05, "loss": 0.552, "mean_token_accuracy": 0.8532103300094604, "num_tokens": 784293472.0, "step": 20652 }, { "epoch": 2.6272738837298055, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.755721569061279, "learning_rate": 1e-05, "loss": 0.4904, "mean_token_accuracy": 0.8736099004745483, "num_tokens": 784339423.0, "step": 20653 }, { "epoch": 2.6274010940083956, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.998842239379883, "learning_rate": 1e-05, "loss": 0.4194, "mean_token_accuracy": 0.8886492848396301, "num_tokens": 784367026.0, "step": 20654 }, { "epoch": 2.6275283042869866, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 7.016955852508545, "learning_rate": 1e-05, "loss": 0.4902, "mean_token_accuracy": 0.8769433498382568, "num_tokens": 784404004.0, "step": 20655 }, { "epoch": 2.6276555145655767, "ewc_loss": 0.11474609375, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.845541954040527, "learning_rate": 1e-05, "loss": 0.426, "mean_token_accuracy": 0.8901877999305725, "num_tokens": 784439335.0, "step": 20656 }, { "epoch": 2.6277827248441676, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00011777877807617188, "grad_norm": 7.422545433044434, "learning_rate": 1e-05, "loss": 0.5163, "mean_token_accuracy": 0.8708117008209229, "num_tokens": 784480141.0, "step": 20657 }, { "epoch": 2.6279099351227577, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 5.905827522277832, "learning_rate": 1e-05, "loss": 0.4636, "mean_token_accuracy": 0.8812059164047241, "num_tokens": 784514785.0, "step": 20658 }, { "epoch": 2.6280371454013487, "ewc_loss": 0.158203125, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00012874603271484375, "grad_norm": 7.829785346984863, "learning_rate": 1e-05, "loss": 0.5289, "mean_token_accuracy": 0.8673248291015625, "num_tokens": 784549776.0, "step": 20659 }, { "epoch": 2.628164355679939, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.3252644538879395, "learning_rate": 1e-05, "loss": 0.5494, "mean_token_accuracy": 0.8604956269264221, "num_tokens": 784578558.0, "step": 20660 }, { "epoch": 2.6282915659585293, "ewc_loss": 0.15625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.0001277923583984375, "grad_norm": 7.808223724365234, "learning_rate": 1e-05, "loss": 0.4944, "mean_token_accuracy": 0.88161301612854, "num_tokens": 784614056.0, "step": 20661 }, { "epoch": 2.62841877623712, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00011348724365234375, "grad_norm": 6.74137544631958, "learning_rate": 1e-05, "loss": 0.4851, "mean_token_accuracy": 0.8790918588638306, "num_tokens": 784651663.0, "step": 20662 }, { "epoch": 2.6285459865157104, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00011777877807617188, "grad_norm": 7.365557670593262, "learning_rate": 1e-05, "loss": 0.5187, "mean_token_accuracy": 0.8688303828239441, "num_tokens": 784690343.0, "step": 20663 }, { "epoch": 2.628673196794301, "ewc_loss": 0.140625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00011205673217773438, "grad_norm": 6.833220958709717, "learning_rate": 1e-05, "loss": 0.4919, "mean_token_accuracy": 0.8790163993835449, "num_tokens": 784728569.0, "step": 20664 }, { "epoch": 2.6288004070728914, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.8613409996032715, "learning_rate": 1e-05, "loss": 0.4549, "mean_token_accuracy": 0.8841273784637451, "num_tokens": 784762376.0, "step": 20665 }, { "epoch": 2.628927617351482, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 6.83389139175415, "learning_rate": 1e-05, "loss": 0.4947, "mean_token_accuracy": 0.8777463436126709, "num_tokens": 784793154.0, "step": 20666 }, { "epoch": 2.6290548276300725, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.419082164764404, "learning_rate": 1e-05, "loss": 0.4731, "mean_token_accuracy": 0.8794354796409607, "num_tokens": 784829640.0, "step": 20667 }, { "epoch": 2.629182037908663, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.888728141784668, "learning_rate": 1e-05, "loss": 0.4576, "mean_token_accuracy": 0.8870586156845093, "num_tokens": 784867045.0, "step": 20668 }, { "epoch": 2.6293092481872535, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.122284412384033, "learning_rate": 1e-05, "loss": 0.4662, "mean_token_accuracy": 0.8783246874809265, "num_tokens": 784907157.0, "step": 20669 }, { "epoch": 2.629436458465844, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 7.0499653816223145, "learning_rate": 1e-05, "loss": 0.4882, "mean_token_accuracy": 0.8797585964202881, "num_tokens": 784942247.0, "step": 20670 }, { "epoch": 2.6295636687444346, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 5.979587078094482, "learning_rate": 1e-05, "loss": 0.4769, "mean_token_accuracy": 0.8748016357421875, "num_tokens": 784976910.0, "step": 20671 }, { "epoch": 2.629690879023025, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00011444091796875, "grad_norm": 7.302656650543213, "learning_rate": 1e-05, "loss": 0.5437, "mean_token_accuracy": 0.8596962690353394, "num_tokens": 785017969.0, "step": 20672 }, { "epoch": 2.6298180893016156, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 5.976611614227295, "learning_rate": 1e-05, "loss": 0.4635, "mean_token_accuracy": 0.8813422918319702, "num_tokens": 785056182.0, "step": 20673 }, { "epoch": 2.629945299580206, "ewc_loss": 0.150390625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00012159347534179688, "grad_norm": 7.5447001457214355, "learning_rate": 1e-05, "loss": 0.4782, "mean_token_accuracy": 0.8855658769607544, "num_tokens": 785091996.0, "step": 20674 }, { "epoch": 2.6300725098587967, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.163510799407959, "learning_rate": 1e-05, "loss": 0.4601, "mean_token_accuracy": 0.8829880952835083, "num_tokens": 785129268.0, "step": 20675 }, { "epoch": 2.6301997201373872, "ewc_loss": 0.15234375, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.623574733734131, "learning_rate": 1e-05, "loss": 0.4672, "mean_token_accuracy": 0.8879126906394958, "num_tokens": 785168418.0, "step": 20676 }, { "epoch": 2.6303269304159778, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.4250407218933105, "learning_rate": 1e-05, "loss": 0.4708, "mean_token_accuracy": 0.8832881450653076, "num_tokens": 785209981.0, "step": 20677 }, { "epoch": 2.6304541406945683, "ewc_loss": 0.1484375, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.0001201629638671875, "grad_norm": 7.455009937286377, "learning_rate": 1e-05, "loss": 0.5019, "mean_token_accuracy": 0.8799474239349365, "num_tokens": 785250354.0, "step": 20678 }, { "epoch": 2.6305813509731584, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.582978248596191, "learning_rate": 1e-05, "loss": 0.4904, "mean_token_accuracy": 0.8762308955192566, "num_tokens": 785288594.0, "step": 20679 }, { "epoch": 2.6307085612517493, "ewc_loss": 0.140625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.0001125335693359375, "grad_norm": 7.1385297775268555, "learning_rate": 1e-05, "loss": 0.4482, "mean_token_accuracy": 0.8911965489387512, "num_tokens": 785323885.0, "step": 20680 }, { "epoch": 2.6308357715303394, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.579863548278809, "learning_rate": 1e-05, "loss": 0.4451, "mean_token_accuracy": 0.8903282880783081, "num_tokens": 785361950.0, "step": 20681 }, { "epoch": 2.6309629818089304, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.929446697235107, "learning_rate": 1e-05, "loss": 0.5058, "mean_token_accuracy": 0.8694440722465515, "num_tokens": 785398650.0, "step": 20682 }, { "epoch": 2.6310901920875205, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.492417335510254, "learning_rate": 1e-05, "loss": 0.5037, "mean_token_accuracy": 0.8704695701599121, "num_tokens": 785451273.0, "step": 20683 }, { "epoch": 2.6312174023661115, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.681591033935547, "learning_rate": 1e-05, "loss": 0.4901, "mean_token_accuracy": 0.8732436895370483, "num_tokens": 785491244.0, "step": 20684 }, { "epoch": 2.6313446126447015, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.38768196105957, "learning_rate": 1e-05, "loss": 0.4702, "mean_token_accuracy": 0.8795362710952759, "num_tokens": 785536558.0, "step": 20685 }, { "epoch": 2.631471822923292, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.61701774597168, "learning_rate": 1e-05, "loss": 0.4953, "mean_token_accuracy": 0.873910665512085, "num_tokens": 785572635.0, "step": 20686 }, { "epoch": 2.6315990332018826, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.282789707183838, "learning_rate": 1e-05, "loss": 0.4914, "mean_token_accuracy": 0.8736924529075623, "num_tokens": 785615765.0, "step": 20687 }, { "epoch": 2.631726243480473, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.614659309387207, "learning_rate": 1e-05, "loss": 0.4785, "mean_token_accuracy": 0.8787168860435486, "num_tokens": 785652934.0, "step": 20688 }, { "epoch": 2.6318534537590637, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.163272380828857, "learning_rate": 1e-05, "loss": 0.4815, "mean_token_accuracy": 0.8739157915115356, "num_tokens": 785689018.0, "step": 20689 }, { "epoch": 2.631980664037654, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.71441125869751, "learning_rate": 1e-05, "loss": 0.5018, "mean_token_accuracy": 0.8700577020645142, "num_tokens": 785728257.0, "step": 20690 }, { "epoch": 2.6321078743162447, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.015237808227539, "learning_rate": 1e-05, "loss": 0.4925, "mean_token_accuracy": 0.8685957789421082, "num_tokens": 785768381.0, "step": 20691 }, { "epoch": 2.6322350845948352, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.922685146331787, "learning_rate": 1e-05, "loss": 0.4818, "mean_token_accuracy": 0.875185489654541, "num_tokens": 785806700.0, "step": 20692 }, { "epoch": 2.6323622948734258, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.889930725097656, "learning_rate": 1e-05, "loss": 0.518, "mean_token_accuracy": 0.8609260320663452, "num_tokens": 785848602.0, "step": 20693 }, { "epoch": 2.6324895051520163, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00011396408081054688, "grad_norm": 7.243216037750244, "learning_rate": 1e-05, "loss": 0.4901, "mean_token_accuracy": 0.8793523907661438, "num_tokens": 785888714.0, "step": 20694 }, { "epoch": 2.632616715430607, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 5.8727569580078125, "learning_rate": 1e-05, "loss": 0.4338, "mean_token_accuracy": 0.8882230520248413, "num_tokens": 785925163.0, "step": 20695 }, { "epoch": 2.6327439257091974, "ewc_loss": 0.1513671875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.0001220703125, "grad_norm": 7.610221862792969, "learning_rate": 1e-05, "loss": 0.5467, "mean_token_accuracy": 0.8601855039596558, "num_tokens": 785963566.0, "step": 20696 }, { "epoch": 2.632871135987788, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.1381988525390625, "learning_rate": 1e-05, "loss": 0.4998, "mean_token_accuracy": 0.867001473903656, "num_tokens": 785997654.0, "step": 20697 }, { "epoch": 2.6329983462663784, "ewc_loss": 0.154296875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00012493133544921875, "grad_norm": 7.692305088043213, "learning_rate": 1e-05, "loss": 0.5185, "mean_token_accuracy": 0.872295081615448, "num_tokens": 786038761.0, "step": 20698 }, { "epoch": 2.633125556544969, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 6.530652046203613, "learning_rate": 1e-05, "loss": 0.4808, "mean_token_accuracy": 0.8807818293571472, "num_tokens": 786072380.0, "step": 20699 }, { "epoch": 2.6332527668235595, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00011730194091796875, "grad_norm": 7.367276191711426, "learning_rate": 1e-05, "loss": 0.5043, "mean_token_accuracy": 0.8747873902320862, "num_tokens": 786112751.0, "step": 20700 }, { "epoch": 2.63337997710215, "ewc_loss": 0.140625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00011157989501953125, "grad_norm": 6.705317974090576, "learning_rate": 1e-05, "loss": 0.506, "mean_token_accuracy": 0.8763518333435059, "num_tokens": 786156841.0, "step": 20701 }, { "epoch": 2.6335071873807405, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.864519119262695, "learning_rate": 1e-05, "loss": 0.5145, "mean_token_accuracy": 0.8701580762863159, "num_tokens": 786191015.0, "step": 20702 }, { "epoch": 2.633634397659331, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 6.774462699890137, "learning_rate": 1e-05, "loss": 0.5426, "mean_token_accuracy": 0.8594595193862915, "num_tokens": 786231420.0, "step": 20703 }, { "epoch": 2.633761607937921, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.367365837097168, "learning_rate": 1e-05, "loss": 0.4582, "mean_token_accuracy": 0.8834110498428345, "num_tokens": 786266586.0, "step": 20704 }, { "epoch": 2.633888818216512, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.913252830505371, "learning_rate": 1e-05, "loss": 0.4937, "mean_token_accuracy": 0.8755625486373901, "num_tokens": 786307489.0, "step": 20705 }, { "epoch": 2.634016028495102, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 6.0219902992248535, "learning_rate": 1e-05, "loss": 0.4541, "mean_token_accuracy": 0.8828179836273193, "num_tokens": 786340960.0, "step": 20706 }, { "epoch": 2.634143238773693, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.0001125335693359375, "grad_norm": 7.180078506469727, "learning_rate": 1e-05, "loss": 0.5003, "mean_token_accuracy": 0.8738633394241333, "num_tokens": 786383086.0, "step": 20707 }, { "epoch": 2.6342704490522832, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 5.884786605834961, "learning_rate": 1e-05, "loss": 0.443, "mean_token_accuracy": 0.8841491937637329, "num_tokens": 786421093.0, "step": 20708 }, { "epoch": 2.6343976593308738, "ewc_loss": 0.150390625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00012111663818359375, "grad_norm": 7.537109375, "learning_rate": 1e-05, "loss": 0.5056, "mean_token_accuracy": 0.8746795654296875, "num_tokens": 786462712.0, "step": 20709 }, { "epoch": 2.6345248696094643, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.035605430603027, "learning_rate": 1e-05, "loss": 0.4833, "mean_token_accuracy": 0.8740029335021973, "num_tokens": 786501607.0, "step": 20710 }, { "epoch": 2.634652079888055, "ewc_loss": 0.15625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00012683868408203125, "grad_norm": 7.762220859527588, "learning_rate": 1e-05, "loss": 0.5007, "mean_token_accuracy": 0.8784447908401489, "num_tokens": 786537163.0, "step": 20711 }, { "epoch": 2.6347792901666454, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.405370712280273, "learning_rate": 1e-05, "loss": 0.4516, "mean_token_accuracy": 0.8869351744651794, "num_tokens": 786572880.0, "step": 20712 }, { "epoch": 2.634906500445236, "ewc_loss": 0.15234375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.617059230804443, "learning_rate": 1e-05, "loss": 0.5015, "mean_token_accuracy": 0.8787885308265686, "num_tokens": 786610082.0, "step": 20713 }, { "epoch": 2.6350337107238264, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010967254638671875, "grad_norm": 6.641839504241943, "learning_rate": 1e-05, "loss": 0.4983, "mean_token_accuracy": 0.874825119972229, "num_tokens": 786648875.0, "step": 20714 }, { "epoch": 2.635160921002417, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00011587142944335938, "grad_norm": 7.283222675323486, "learning_rate": 1e-05, "loss": 0.5139, "mean_token_accuracy": 0.8728075623512268, "num_tokens": 786688076.0, "step": 20715 }, { "epoch": 2.6352881312810075, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.6561784744262695, "learning_rate": 1e-05, "loss": 0.4979, "mean_token_accuracy": 0.8731257915496826, "num_tokens": 786723969.0, "step": 20716 }, { "epoch": 2.635415341559598, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.940438747406006, "learning_rate": 1e-05, "loss": 0.5044, "mean_token_accuracy": 0.8706095218658447, "num_tokens": 786764018.0, "step": 20717 }, { "epoch": 2.6355425518381885, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.563102722167969, "learning_rate": 1e-05, "loss": 0.454, "mean_token_accuracy": 0.8837257623672485, "num_tokens": 786796586.0, "step": 20718 }, { "epoch": 2.635669762116779, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.767919540405273, "learning_rate": 1e-05, "loss": 0.525, "mean_token_accuracy": 0.8642657995223999, "num_tokens": 786833797.0, "step": 20719 }, { "epoch": 2.6357969723953696, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.4236297607421875, "learning_rate": 1e-05, "loss": 0.5189, "mean_token_accuracy": 0.8659169673919678, "num_tokens": 786869971.0, "step": 20720 }, { "epoch": 2.63592418267396, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.654989719390869, "learning_rate": 1e-05, "loss": 0.4804, "mean_token_accuracy": 0.8793593645095825, "num_tokens": 786905854.0, "step": 20721 }, { "epoch": 2.6360513929525506, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.270859718322754, "learning_rate": 1e-05, "loss": 0.4748, "mean_token_accuracy": 0.8772327899932861, "num_tokens": 786940535.0, "step": 20722 }, { "epoch": 2.636178603231141, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.692080497741699, "learning_rate": 1e-05, "loss": 0.5053, "mean_token_accuracy": 0.8662129640579224, "num_tokens": 786974976.0, "step": 20723 }, { "epoch": 2.6363058135097317, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.1146674156188965, "learning_rate": 1e-05, "loss": 0.4801, "mean_token_accuracy": 0.8743098378181458, "num_tokens": 787013389.0, "step": 20724 }, { "epoch": 2.6364330237883222, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.847352981567383, "learning_rate": 1e-05, "loss": 0.4402, "mean_token_accuracy": 0.8912005424499512, "num_tokens": 787044463.0, "step": 20725 }, { "epoch": 2.6365602340669128, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 5.977613925933838, "learning_rate": 1e-05, "loss": 0.4616, "mean_token_accuracy": 0.878541886806488, "num_tokens": 787075123.0, "step": 20726 }, { "epoch": 2.636687444345503, "ewc_loss": 0.140625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00011110305786132812, "grad_norm": 7.096512317657471, "learning_rate": 1e-05, "loss": 0.5088, "mean_token_accuracy": 0.8723248839378357, "num_tokens": 787114790.0, "step": 20727 }, { "epoch": 2.636814654624094, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 5.900813102722168, "learning_rate": 1e-05, "loss": 0.4368, "mean_token_accuracy": 0.8879253268241882, "num_tokens": 787153263.0, "step": 20728 }, { "epoch": 2.636941864902684, "ewc_loss": 0.1474609375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00011873245239257812, "grad_norm": 7.401214599609375, "learning_rate": 1e-05, "loss": 0.4712, "mean_token_accuracy": 0.8864998817443848, "num_tokens": 787191955.0, "step": 20729 }, { "epoch": 2.637069075181275, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.018636703491211, "learning_rate": 1e-05, "loss": 0.4684, "mean_token_accuracy": 0.8792334794998169, "num_tokens": 787228330.0, "step": 20730 }, { "epoch": 2.637196285459865, "ewc_loss": 0.1533203125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.0001239776611328125, "grad_norm": 7.6288676261901855, "learning_rate": 1e-05, "loss": 0.5711, "mean_token_accuracy": 0.855633020401001, "num_tokens": 787271927.0, "step": 20731 }, { "epoch": 2.637323495738456, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.328566074371338, "learning_rate": 1e-05, "loss": 0.454, "mean_token_accuracy": 0.8890159130096436, "num_tokens": 787310515.0, "step": 20732 }, { "epoch": 2.637450706017046, "ewc_loss": 0.1494140625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00012063980102539062, "grad_norm": 7.503958702087402, "learning_rate": 1e-05, "loss": 0.5033, "mean_token_accuracy": 0.8765735030174255, "num_tokens": 787346720.0, "step": 20733 }, { "epoch": 2.6375779162956365, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 6.612804412841797, "learning_rate": 1e-05, "loss": 0.495, "mean_token_accuracy": 0.8742272257804871, "num_tokens": 787387591.0, "step": 20734 }, { "epoch": 2.637705126574227, "ewc_loss": 0.140625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00011205673217773438, "grad_norm": 7.082586288452148, "learning_rate": 1e-05, "loss": 0.5071, "mean_token_accuracy": 0.8741800785064697, "num_tokens": 787427762.0, "step": 20735 }, { "epoch": 2.6378323368528176, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 6.721498489379883, "learning_rate": 1e-05, "loss": 0.5194, "mean_token_accuracy": 0.8658920526504517, "num_tokens": 787463047.0, "step": 20736 }, { "epoch": 2.637959547131408, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.553668022155762, "learning_rate": 1e-05, "loss": 0.4404, "mean_token_accuracy": 0.889406144618988, "num_tokens": 787500862.0, "step": 20737 }, { "epoch": 2.6380867574099987, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.837808132171631, "learning_rate": 1e-05, "loss": 0.4772, "mean_token_accuracy": 0.8861677646636963, "num_tokens": 787529625.0, "step": 20738 }, { "epoch": 2.638213967688589, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.145462989807129, "learning_rate": 1e-05, "loss": 0.4623, "mean_token_accuracy": 0.8814778327941895, "num_tokens": 787567248.0, "step": 20739 }, { "epoch": 2.6383411779671797, "ewc_loss": 0.140625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00011157989501953125, "grad_norm": 7.049041748046875, "learning_rate": 1e-05, "loss": 0.5086, "mean_token_accuracy": 0.8738797903060913, "num_tokens": 787606763.0, "step": 20740 }, { "epoch": 2.6384683882457702, "ewc_loss": 0.11376953125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.866701126098633, "learning_rate": 1e-05, "loss": 0.4693, "mean_token_accuracy": 0.8742834329605103, "num_tokens": 787638795.0, "step": 20741 }, { "epoch": 2.6385955985243608, "ewc_loss": 0.1484375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00011920928955078125, "grad_norm": 7.421860694885254, "learning_rate": 1e-05, "loss": 0.5202, "mean_token_accuracy": 0.8704402446746826, "num_tokens": 787679495.0, "step": 20742 }, { "epoch": 2.6387228088029513, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 5.889934539794922, "learning_rate": 1e-05, "loss": 0.4624, "mean_token_accuracy": 0.8779584765434265, "num_tokens": 787718583.0, "step": 20743 }, { "epoch": 2.638850019081542, "ewc_loss": 0.1591796875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00012969970703125, "grad_norm": 7.779592514038086, "learning_rate": 1e-05, "loss": 0.4568, "mean_token_accuracy": 0.8935739994049072, "num_tokens": 787756558.0, "step": 20744 }, { "epoch": 2.6389772293601323, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.270313262939453, "learning_rate": 1e-05, "loss": 0.4758, "mean_token_accuracy": 0.8795157074928284, "num_tokens": 787789796.0, "step": 20745 }, { "epoch": 2.639104439638723, "ewc_loss": 0.158203125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00012874603271484375, "grad_norm": 7.7753376960754395, "learning_rate": 1e-05, "loss": 0.5433, "mean_token_accuracy": 0.8657571077346802, "num_tokens": 787832054.0, "step": 20746 }, { "epoch": 2.6392316499173134, "ewc_loss": 0.140625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00011205673217773438, "grad_norm": 6.645514488220215, "learning_rate": 1e-05, "loss": 0.5, "mean_token_accuracy": 0.8764052987098694, "num_tokens": 787876723.0, "step": 20747 }, { "epoch": 2.639358860195904, "ewc_loss": 0.1494140625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.0001201629638671875, "grad_norm": 7.44582462310791, "learning_rate": 1e-05, "loss": 0.5305, "mean_token_accuracy": 0.864211916923523, "num_tokens": 787912183.0, "step": 20748 }, { "epoch": 2.6394860704744945, "ewc_loss": 0.140625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00011110305786132812, "grad_norm": 6.739496231079102, "learning_rate": 1e-05, "loss": 0.5383, "mean_token_accuracy": 0.8646726608276367, "num_tokens": 787946865.0, "step": 20749 }, { "epoch": 2.639613280753085, "ewc_loss": 0.140625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00011110305786132812, "grad_norm": 7.047108173370361, "learning_rate": 1e-05, "loss": 0.4709, "mean_token_accuracy": 0.8855639696121216, "num_tokens": 787984286.0, "step": 20750 }, { "epoch": 2.6397404910316755, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.639216423034668, "learning_rate": 1e-05, "loss": 0.4634, "mean_token_accuracy": 0.8885200619697571, "num_tokens": 788020565.0, "step": 20751 }, { "epoch": 2.6398677013102656, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.771540641784668, "learning_rate": 1e-05, "loss": 0.4644, "mean_token_accuracy": 0.88675856590271, "num_tokens": 788051833.0, "step": 20752 }, { "epoch": 2.6399949115888566, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.505562782287598, "learning_rate": 1e-05, "loss": 0.4401, "mean_token_accuracy": 0.8891445994377136, "num_tokens": 788092780.0, "step": 20753 }, { "epoch": 2.6401221218674467, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.632573127746582, "learning_rate": 1e-05, "loss": 0.5019, "mean_token_accuracy": 0.869774580001831, "num_tokens": 788140088.0, "step": 20754 }, { "epoch": 2.6402493321460376, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.349660873413086, "learning_rate": 1e-05, "loss": 0.4603, "mean_token_accuracy": 0.8817300796508789, "num_tokens": 788177951.0, "step": 20755 }, { "epoch": 2.6403765424246277, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.6375203132629395, "learning_rate": 1e-05, "loss": 0.4666, "mean_token_accuracy": 0.8830875158309937, "num_tokens": 788212124.0, "step": 20756 }, { "epoch": 2.6405037527032187, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.172458171844482, "learning_rate": 1e-05, "loss": 0.4684, "mean_token_accuracy": 0.8753606081008911, "num_tokens": 788255860.0, "step": 20757 }, { "epoch": 2.6406309629818088, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.754638671875, "learning_rate": 1e-05, "loss": 0.448, "mean_token_accuracy": 0.8865947723388672, "num_tokens": 788296481.0, "step": 20758 }, { "epoch": 2.6407581732603993, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.010831356048584, "learning_rate": 1e-05, "loss": 0.4348, "mean_token_accuracy": 0.8875510692596436, "num_tokens": 788328818.0, "step": 20759 }, { "epoch": 2.64088538353899, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.9895782470703125, "learning_rate": 1e-05, "loss": 0.4333, "mean_token_accuracy": 0.8958450555801392, "num_tokens": 788370359.0, "step": 20760 }, { "epoch": 2.6410125938175804, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.882136344909668, "learning_rate": 1e-05, "loss": 0.4612, "mean_token_accuracy": 0.8780627250671387, "num_tokens": 788412425.0, "step": 20761 }, { "epoch": 2.641139804096171, "ewc_loss": 0.1455078125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00011682510375976562, "grad_norm": 7.349058151245117, "learning_rate": 1e-05, "loss": 0.5445, "mean_token_accuracy": 0.8617351055145264, "num_tokens": 788445137.0, "step": 20762 }, { "epoch": 2.6412670143747614, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 5.911435127258301, "learning_rate": 1e-05, "loss": 0.442, "mean_token_accuracy": 0.8845807313919067, "num_tokens": 788478185.0, "step": 20763 }, { "epoch": 2.641394224653352, "ewc_loss": 0.154296875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00012493133544921875, "grad_norm": 7.6781744956970215, "learning_rate": 1e-05, "loss": 0.5104, "mean_token_accuracy": 0.8751304149627686, "num_tokens": 788514329.0, "step": 20764 }, { "epoch": 2.6415214349319425, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.224337100982666, "learning_rate": 1e-05, "loss": 0.4728, "mean_token_accuracy": 0.880294144153595, "num_tokens": 788554234.0, "step": 20765 }, { "epoch": 2.641648645210533, "ewc_loss": 0.1552734375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.000125885009765625, "grad_norm": 7.693909645080566, "learning_rate": 1e-05, "loss": 0.4934, "mean_token_accuracy": 0.8810623288154602, "num_tokens": 788591684.0, "step": 20766 }, { "epoch": 2.6417758554891235, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010967254638671875, "grad_norm": 6.614225387573242, "learning_rate": 1e-05, "loss": 0.4739, "mean_token_accuracy": 0.879546582698822, "num_tokens": 788626802.0, "step": 20767 }, { "epoch": 2.641903065767714, "ewc_loss": 0.1455078125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.0001163482666015625, "grad_norm": 7.298629283905029, "learning_rate": 1e-05, "loss": 0.4893, "mean_token_accuracy": 0.8810536861419678, "num_tokens": 788673826.0, "step": 20768 }, { "epoch": 2.6420302760463046, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00011014938354492188, "grad_norm": 6.7558465003967285, "learning_rate": 1e-05, "loss": 0.4581, "mean_token_accuracy": 0.8864375948905945, "num_tokens": 788708393.0, "step": 20769 }, { "epoch": 2.642157486324895, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.788010597229004, "learning_rate": 1e-05, "loss": 0.463, "mean_token_accuracy": 0.8854461908340454, "num_tokens": 788746480.0, "step": 20770 }, { "epoch": 2.6422846966034856, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 6.79427433013916, "learning_rate": 1e-05, "loss": 0.4951, "mean_token_accuracy": 0.8776485323905945, "num_tokens": 788785295.0, "step": 20771 }, { "epoch": 2.642411906882076, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.322980880737305, "learning_rate": 1e-05, "loss": 0.4575, "mean_token_accuracy": 0.8823880553245544, "num_tokens": 788818803.0, "step": 20772 }, { "epoch": 2.6425391171606667, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 6.939906597137451, "learning_rate": 1e-05, "loss": 0.507, "mean_token_accuracy": 0.8741328716278076, "num_tokens": 788857266.0, "step": 20773 }, { "epoch": 2.6426663274392572, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 6.013442039489746, "learning_rate": 1e-05, "loss": 0.514, "mean_token_accuracy": 0.8586220741271973, "num_tokens": 788898433.0, "step": 20774 }, { "epoch": 2.6427935377178478, "ewc_loss": 0.140625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00011205673217773438, "grad_norm": 7.192661762237549, "learning_rate": 1e-05, "loss": 0.5274, "mean_token_accuracy": 0.8710902333259583, "num_tokens": 788942259.0, "step": 20775 }, { "epoch": 2.6429207479964383, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 5.894423007965088, "learning_rate": 1e-05, "loss": 0.5026, "mean_token_accuracy": 0.8679019212722778, "num_tokens": 788981484.0, "step": 20776 }, { "epoch": 2.6430479582750284, "ewc_loss": 0.1484375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00011968612670898438, "grad_norm": 7.5363593101501465, "learning_rate": 1e-05, "loss": 0.4971, "mean_token_accuracy": 0.8790867328643799, "num_tokens": 789019863.0, "step": 20777 }, { "epoch": 2.6431751685536193, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.041168689727783, "learning_rate": 1e-05, "loss": 0.4752, "mean_token_accuracy": 0.8773082494735718, "num_tokens": 789059139.0, "step": 20778 }, { "epoch": 2.6433023788322094, "ewc_loss": 0.1572265625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.0001277923583984375, "grad_norm": 7.741063117980957, "learning_rate": 1e-05, "loss": 0.4946, "mean_token_accuracy": 0.8811742663383484, "num_tokens": 789100429.0, "step": 20779 }, { "epoch": 2.6434295891108004, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.404179096221924, "learning_rate": 1e-05, "loss": 0.4749, "mean_token_accuracy": 0.8811407089233398, "num_tokens": 789136219.0, "step": 20780 }, { "epoch": 2.6435567993893905, "ewc_loss": 0.15234375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.618517875671387, "learning_rate": 1e-05, "loss": 0.4984, "mean_token_accuracy": 0.880302906036377, "num_tokens": 789177541.0, "step": 20781 }, { "epoch": 2.6436840096679814, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 6.645720481872559, "learning_rate": 1e-05, "loss": 0.4389, "mean_token_accuracy": 0.894010066986084, "num_tokens": 789217165.0, "step": 20782 }, { "epoch": 2.6438112199465715, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00011539459228515625, "grad_norm": 7.279195308685303, "learning_rate": 1e-05, "loss": 0.477, "mean_token_accuracy": 0.8807587623596191, "num_tokens": 789255384.0, "step": 20783 }, { "epoch": 2.643938430225162, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.6534199714660645, "learning_rate": 1e-05, "loss": 0.4701, "mean_token_accuracy": 0.8802149891853333, "num_tokens": 789294153.0, "step": 20784 }, { "epoch": 2.6440656405037526, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.955690860748291, "learning_rate": 1e-05, "loss": 0.5484, "mean_token_accuracy": 0.8598443865776062, "num_tokens": 789334933.0, "step": 20785 }, { "epoch": 2.644192850782343, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.573277473449707, "learning_rate": 1e-05, "loss": 0.5417, "mean_token_accuracy": 0.8560189604759216, "num_tokens": 789370068.0, "step": 20786 }, { "epoch": 2.6443200610609336, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.710809707641602, "learning_rate": 1e-05, "loss": 0.4643, "mean_token_accuracy": 0.8828779458999634, "num_tokens": 789413887.0, "step": 20787 }, { "epoch": 2.644447271339524, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.46038818359375, "learning_rate": 1e-05, "loss": 0.4518, "mean_token_accuracy": 0.8878720998764038, "num_tokens": 789452319.0, "step": 20788 }, { "epoch": 2.6445744816181147, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.6085591316223145, "learning_rate": 1e-05, "loss": 0.4544, "mean_token_accuracy": 0.8868747353553772, "num_tokens": 789488988.0, "step": 20789 }, { "epoch": 2.6447016918967052, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.326038837432861, "learning_rate": 1e-05, "loss": 0.5053, "mean_token_accuracy": 0.8704882860183716, "num_tokens": 789529849.0, "step": 20790 }, { "epoch": 2.6448289021752958, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.61827278137207, "learning_rate": 1e-05, "loss": 0.4807, "mean_token_accuracy": 0.8755123019218445, "num_tokens": 789572877.0, "step": 20791 }, { "epoch": 2.6449561124538863, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.1975555419921875, "learning_rate": 1e-05, "loss": 0.5061, "mean_token_accuracy": 0.8667236566543579, "num_tokens": 789605189.0, "step": 20792 }, { "epoch": 2.645083322732477, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.706363201141357, "learning_rate": 1e-05, "loss": 0.4592, "mean_token_accuracy": 0.8843902349472046, "num_tokens": 789641589.0, "step": 20793 }, { "epoch": 2.6452105330110673, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.050658702850342, "learning_rate": 1e-05, "loss": 0.4905, "mean_token_accuracy": 0.8669651746749878, "num_tokens": 789673869.0, "step": 20794 }, { "epoch": 2.645337743289658, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.910233020782471, "learning_rate": 1e-05, "loss": 0.4817, "mean_token_accuracy": 0.8781086206436157, "num_tokens": 789710605.0, "step": 20795 }, { "epoch": 2.6454649535682484, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.89040470123291, "learning_rate": 1e-05, "loss": 0.4328, "mean_token_accuracy": 0.8860837817192078, "num_tokens": 789747641.0, "step": 20796 }, { "epoch": 2.645592163846839, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00011396408081054688, "grad_norm": 7.244400978088379, "learning_rate": 1e-05, "loss": 0.4452, "mean_token_accuracy": 0.8915346264839172, "num_tokens": 789781345.0, "step": 20797 }, { "epoch": 2.6457193741254295, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 5.864659309387207, "learning_rate": 1e-05, "loss": 0.4821, "mean_token_accuracy": 0.8751645088195801, "num_tokens": 789819582.0, "step": 20798 }, { "epoch": 2.64584658440402, "ewc_loss": 0.15234375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.623448848724365, "learning_rate": 1e-05, "loss": 0.4621, "mean_token_accuracy": 0.8884075880050659, "num_tokens": 789856281.0, "step": 20799 }, { "epoch": 2.6459737946826105, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.104226589202881, "learning_rate": 1e-05, "loss": 0.4468, "mean_token_accuracy": 0.8870445489883423, "num_tokens": 789895711.0, "step": 20800 }, { "epoch": 2.646101004961201, "ewc_loss": 0.158203125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00012874603271484375, "grad_norm": 7.781596660614014, "learning_rate": 1e-05, "loss": 0.4766, "mean_token_accuracy": 0.8912544250488281, "num_tokens": 789931751.0, "step": 20801 }, { "epoch": 2.646228215239791, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010967254638671875, "grad_norm": 6.546076774597168, "learning_rate": 1e-05, "loss": 0.5105, "mean_token_accuracy": 0.8723483085632324, "num_tokens": 789969207.0, "step": 20802 }, { "epoch": 2.646355425518382, "ewc_loss": 0.1484375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00011968612670898438, "grad_norm": 7.480409622192383, "learning_rate": 1e-05, "loss": 0.5044, "mean_token_accuracy": 0.8739309906959534, "num_tokens": 790005706.0, "step": 20803 }, { "epoch": 2.646482635796972, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.0001125335693359375, "grad_norm": 6.760804653167725, "learning_rate": 1e-05, "loss": 0.4961, "mean_token_accuracy": 0.8747544288635254, "num_tokens": 790047389.0, "step": 20804 }, { "epoch": 2.646609846075563, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 6.975292682647705, "learning_rate": 1e-05, "loss": 0.4908, "mean_token_accuracy": 0.880915641784668, "num_tokens": 790085430.0, "step": 20805 }, { "epoch": 2.6467370563541532, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 6.800542831420898, "learning_rate": 1e-05, "loss": 0.5036, "mean_token_accuracy": 0.8739327192306519, "num_tokens": 790126620.0, "step": 20806 }, { "epoch": 2.6468642666327438, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.4594268798828125, "learning_rate": 1e-05, "loss": 0.4566, "mean_token_accuracy": 0.8815171122550964, "num_tokens": 790166033.0, "step": 20807 }, { "epoch": 2.6469914769113343, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 6.92179536819458, "learning_rate": 1e-05, "loss": 0.4547, "mean_token_accuracy": 0.8864872455596924, "num_tokens": 790202463.0, "step": 20808 }, { "epoch": 2.647118687189925, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.084694862365723, "learning_rate": 1e-05, "loss": 0.4558, "mean_token_accuracy": 0.8815033435821533, "num_tokens": 790241600.0, "step": 20809 }, { "epoch": 2.6472458974685154, "ewc_loss": 0.140625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00011110305786132812, "grad_norm": 7.167119026184082, "learning_rate": 1e-05, "loss": 0.5626, "mean_token_accuracy": 0.8550605177879333, "num_tokens": 790277669.0, "step": 20810 }, { "epoch": 2.647373107747106, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 5.9376726150512695, "learning_rate": 1e-05, "loss": 0.4884, "mean_token_accuracy": 0.8695585131645203, "num_tokens": 790317453.0, "step": 20811 }, { "epoch": 2.6475003180256964, "ewc_loss": 0.1474609375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.000118255615234375, "grad_norm": 7.475872993469238, "learning_rate": 1e-05, "loss": 0.4976, "mean_token_accuracy": 0.8768582344055176, "num_tokens": 790354460.0, "step": 20812 }, { "epoch": 2.647627528304287, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.039027690887451, "learning_rate": 1e-05, "loss": 0.4655, "mean_token_accuracy": 0.877995491027832, "num_tokens": 790393933.0, "step": 20813 }, { "epoch": 2.6477547385828775, "ewc_loss": 0.154296875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00012493133544921875, "grad_norm": 7.700959205627441, "learning_rate": 1e-05, "loss": 0.4972, "mean_token_accuracy": 0.8790925741195679, "num_tokens": 790425477.0, "step": 20814 }, { "epoch": 2.647881948861468, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.341212749481201, "learning_rate": 1e-05, "loss": 0.539, "mean_token_accuracy": 0.8567050099372864, "num_tokens": 790470501.0, "step": 20815 }, { "epoch": 2.6480091591400585, "ewc_loss": 0.1513671875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.0001220703125, "grad_norm": 7.625479221343994, "learning_rate": 1e-05, "loss": 0.5227, "mean_token_accuracy": 0.8733600378036499, "num_tokens": 790508869.0, "step": 20816 }, { "epoch": 2.648136369418649, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 6.581485748291016, "learning_rate": 1e-05, "loss": 0.4391, "mean_token_accuracy": 0.8925325274467468, "num_tokens": 790545675.0, "step": 20817 }, { "epoch": 2.6482635796972396, "ewc_loss": 0.1455078125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00011682510375976562, "grad_norm": 7.334123611450195, "learning_rate": 1e-05, "loss": 0.4909, "mean_token_accuracy": 0.8816854953765869, "num_tokens": 790586832.0, "step": 20818 }, { "epoch": 2.64839078997583, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.615716934204102, "learning_rate": 1e-05, "loss": 0.5014, "mean_token_accuracy": 0.8735958337783813, "num_tokens": 790633587.0, "step": 20819 }, { "epoch": 2.6485180002544206, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010967254638671875, "grad_norm": 7.036197185516357, "learning_rate": 1e-05, "loss": 0.5209, "mean_token_accuracy": 0.8679492473602295, "num_tokens": 790668768.0, "step": 20820 }, { "epoch": 2.648645210533011, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.544778347015381, "learning_rate": 1e-05, "loss": 0.4553, "mean_token_accuracy": 0.88717120885849, "num_tokens": 790708393.0, "step": 20821 }, { "epoch": 2.6487724208116017, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.820215225219727, "learning_rate": 1e-05, "loss": 0.5132, "mean_token_accuracy": 0.8664702773094177, "num_tokens": 790742632.0, "step": 20822 }, { "epoch": 2.648899631090192, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.439385890960693, "learning_rate": 1e-05, "loss": 0.5153, "mean_token_accuracy": 0.8649806976318359, "num_tokens": 790779740.0, "step": 20823 }, { "epoch": 2.6490268413687827, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.708308219909668, "learning_rate": 1e-05, "loss": 0.4515, "mean_token_accuracy": 0.8851581811904907, "num_tokens": 790819040.0, "step": 20824 }, { "epoch": 2.649154051647373, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.292593002319336, "learning_rate": 1e-05, "loss": 0.4354, "mean_token_accuracy": 0.89227294921875, "num_tokens": 790853382.0, "step": 20825 }, { "epoch": 2.649281261925964, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.738569736480713, "learning_rate": 1e-05, "loss": 0.5044, "mean_token_accuracy": 0.8694932460784912, "num_tokens": 790890606.0, "step": 20826 }, { "epoch": 2.649408472204554, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.1382832527160645, "learning_rate": 1e-05, "loss": 0.4693, "mean_token_accuracy": 0.8841745853424072, "num_tokens": 790929469.0, "step": 20827 }, { "epoch": 2.649535682483145, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.864862442016602, "learning_rate": 1e-05, "loss": 0.4914, "mean_token_accuracy": 0.8762780427932739, "num_tokens": 790970745.0, "step": 20828 }, { "epoch": 2.649662892761735, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.011430263519287, "learning_rate": 1e-05, "loss": 0.5038, "mean_token_accuracy": 0.8677475452423096, "num_tokens": 791010365.0, "step": 20829 }, { "epoch": 2.649790103040326, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00011014938354492188, "grad_norm": 7.091676712036133, "learning_rate": 1e-05, "loss": 0.491, "mean_token_accuracy": 0.8793801069259644, "num_tokens": 791047012.0, "step": 20830 }, { "epoch": 2.649917313318916, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 5.9394073486328125, "learning_rate": 1e-05, "loss": 0.4511, "mean_token_accuracy": 0.880824089050293, "num_tokens": 791092836.0, "step": 20831 }, { "epoch": 2.6500445235975065, "ewc_loss": 0.1455078125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.0001163482666015625, "grad_norm": 7.373227119445801, "learning_rate": 1e-05, "loss": 0.4681, "mean_token_accuracy": 0.8890485763549805, "num_tokens": 791127725.0, "step": 20832 }, { "epoch": 2.650171733876097, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.060314178466797, "learning_rate": 1e-05, "loss": 0.4471, "mean_token_accuracy": 0.8842406272888184, "num_tokens": 791164530.0, "step": 20833 }, { "epoch": 2.6502989441546876, "ewc_loss": 0.1494140625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00012063980102539062, "grad_norm": 7.527754306793213, "learning_rate": 1e-05, "loss": 0.4496, "mean_token_accuracy": 0.8930482864379883, "num_tokens": 791202384.0, "step": 20834 }, { "epoch": 2.650426154433278, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.335507392883301, "learning_rate": 1e-05, "loss": 0.4327, "mean_token_accuracy": 0.8914622068405151, "num_tokens": 791237998.0, "step": 20835 }, { "epoch": 2.6505533647118686, "ewc_loss": 0.1455078125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.0001163482666015625, "grad_norm": 7.374403953552246, "learning_rate": 1e-05, "loss": 0.5464, "mean_token_accuracy": 0.8568719625473022, "num_tokens": 791277755.0, "step": 20836 }, { "epoch": 2.650680574990459, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.5674967765808105, "learning_rate": 1e-05, "loss": 0.51, "mean_token_accuracy": 0.8674086928367615, "num_tokens": 791315779.0, "step": 20837 }, { "epoch": 2.6508077852690497, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.947864055633545, "learning_rate": 1e-05, "loss": 0.5098, "mean_token_accuracy": 0.8739218711853027, "num_tokens": 791359212.0, "step": 20838 }, { "epoch": 2.6509349955476402, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.699450969696045, "learning_rate": 1e-05, "loss": 0.4281, "mean_token_accuracy": 0.8956155776977539, "num_tokens": 791388308.0, "step": 20839 }, { "epoch": 2.6510622058262308, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.435091972351074, "learning_rate": 1e-05, "loss": 0.4939, "mean_token_accuracy": 0.8744910359382629, "num_tokens": 791425645.0, "step": 20840 }, { "epoch": 2.6511894161048213, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.871209621429443, "learning_rate": 1e-05, "loss": 0.4809, "mean_token_accuracy": 0.8760219812393188, "num_tokens": 791462632.0, "step": 20841 }, { "epoch": 2.651316626383412, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.979093551635742, "learning_rate": 1e-05, "loss": 0.4098, "mean_token_accuracy": 0.8947452902793884, "num_tokens": 791502834.0, "step": 20842 }, { "epoch": 2.6514438366620023, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00011301040649414062, "grad_norm": 7.231142044067383, "learning_rate": 1e-05, "loss": 0.4952, "mean_token_accuracy": 0.8747809529304504, "num_tokens": 791538973.0, "step": 20843 }, { "epoch": 2.651571046940593, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.774216651916504, "learning_rate": 1e-05, "loss": 0.4596, "mean_token_accuracy": 0.8793836236000061, "num_tokens": 791577804.0, "step": 20844 }, { "epoch": 2.6516982572191834, "ewc_loss": 0.1552734375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.000125885009765625, "grad_norm": 7.714718818664551, "learning_rate": 1e-05, "loss": 0.4597, "mean_token_accuracy": 0.8895440101623535, "num_tokens": 791618100.0, "step": 20845 }, { "epoch": 2.651825467497774, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.014642238616943, "learning_rate": 1e-05, "loss": 0.4872, "mean_token_accuracy": 0.873537003993988, "num_tokens": 791660112.0, "step": 20846 }, { "epoch": 2.6519526777763645, "ewc_loss": 0.1630859375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.000133514404296875, "grad_norm": 8.016631126403809, "learning_rate": 1e-05, "loss": 0.5129, "mean_token_accuracy": 0.876591682434082, "num_tokens": 791698133.0, "step": 20847 }, { "epoch": 2.652079888054955, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 6.594536304473877, "learning_rate": 1e-05, "loss": 0.4769, "mean_token_accuracy": 0.8830952048301697, "num_tokens": 791737810.0, "step": 20848 }, { "epoch": 2.6522070983335455, "ewc_loss": 0.1552734375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.000125885009765625, "grad_norm": 7.7584228515625, "learning_rate": 1e-05, "loss": 0.5161, "mean_token_accuracy": 0.8715212345123291, "num_tokens": 791775187.0, "step": 20849 }, { "epoch": 2.6523343086121356, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00011491775512695312, "grad_norm": 6.865755081176758, "learning_rate": 1e-05, "loss": 0.4689, "mean_token_accuracy": 0.8855844140052795, "num_tokens": 791810673.0, "step": 20850 }, { "epoch": 2.6524615188907266, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00011587142944335938, "grad_norm": 7.30742073059082, "learning_rate": 1e-05, "loss": 0.427, "mean_token_accuracy": 0.9000297784805298, "num_tokens": 791848144.0, "step": 20851 }, { "epoch": 2.6525887291693167, "ewc_loss": 0.140625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00011110305786132812, "grad_norm": 6.7912116050720215, "learning_rate": 1e-05, "loss": 0.4894, "mean_token_accuracy": 0.8767894506454468, "num_tokens": 791885025.0, "step": 20852 }, { "epoch": 2.6527159394479076, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.950407981872559, "learning_rate": 1e-05, "loss": 0.495, "mean_token_accuracy": 0.875586986541748, "num_tokens": 791922885.0, "step": 20853 }, { "epoch": 2.6528431497264977, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.635410785675049, "learning_rate": 1e-05, "loss": 0.543, "mean_token_accuracy": 0.8585504293441772, "num_tokens": 791959791.0, "step": 20854 }, { "epoch": 2.6529703600050887, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.734330177307129, "learning_rate": 1e-05, "loss": 0.4726, "mean_token_accuracy": 0.8789235353469849, "num_tokens": 791998559.0, "step": 20855 }, { "epoch": 2.6530975702836788, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.465475559234619, "learning_rate": 1e-05, "loss": 0.4773, "mean_token_accuracy": 0.8778294324874878, "num_tokens": 792038428.0, "step": 20856 }, { "epoch": 2.6532247805622693, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.679691314697266, "learning_rate": 1e-05, "loss": 0.5291, "mean_token_accuracy": 0.8659945726394653, "num_tokens": 792074405.0, "step": 20857 }, { "epoch": 2.65335199084086, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.288910388946533, "learning_rate": 1e-05, "loss": 0.5067, "mean_token_accuracy": 0.8666701316833496, "num_tokens": 792112831.0, "step": 20858 }, { "epoch": 2.6534792011194503, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.722236156463623, "learning_rate": 1e-05, "loss": 0.5028, "mean_token_accuracy": 0.8733136653900146, "num_tokens": 792154122.0, "step": 20859 }, { "epoch": 2.653606411398041, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.119058132171631, "learning_rate": 1e-05, "loss": 0.5228, "mean_token_accuracy": 0.8592637181282043, "num_tokens": 792191818.0, "step": 20860 }, { "epoch": 2.6537336216766314, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.885214805603027, "learning_rate": 1e-05, "loss": 0.4829, "mean_token_accuracy": 0.878532886505127, "num_tokens": 792226753.0, "step": 20861 }, { "epoch": 2.653860831955222, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 5.959100723266602, "learning_rate": 1e-05, "loss": 0.4623, "mean_token_accuracy": 0.875540018081665, "num_tokens": 792263941.0, "step": 20862 }, { "epoch": 2.6539880422338125, "ewc_loss": 0.140625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00011205673217773438, "grad_norm": 7.1726603507995605, "learning_rate": 1e-05, "loss": 0.4856, "mean_token_accuracy": 0.8807039856910706, "num_tokens": 792301433.0, "step": 20863 }, { "epoch": 2.654115252512403, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 5.890684127807617, "learning_rate": 1e-05, "loss": 0.4706, "mean_token_accuracy": 0.8774397969245911, "num_tokens": 792340188.0, "step": 20864 }, { "epoch": 2.6542424627909935, "ewc_loss": 0.150390625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00012159347534179688, "grad_norm": 7.561984062194824, "learning_rate": 1e-05, "loss": 0.526, "mean_token_accuracy": 0.8705574870109558, "num_tokens": 792383498.0, "step": 20865 }, { "epoch": 2.654369673069584, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.0658955574035645, "learning_rate": 1e-05, "loss": 0.4403, "mean_token_accuracy": 0.8869715332984924, "num_tokens": 792417160.0, "step": 20866 }, { "epoch": 2.6544968833481746, "ewc_loss": 0.1572265625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.0001277923583984375, "grad_norm": 7.788206577301025, "learning_rate": 1e-05, "loss": 0.5239, "mean_token_accuracy": 0.8736149668693542, "num_tokens": 792451609.0, "step": 20867 }, { "epoch": 2.654624093626765, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.484951496124268, "learning_rate": 1e-05, "loss": 0.531, "mean_token_accuracy": 0.8650161027908325, "num_tokens": 792498309.0, "step": 20868 }, { "epoch": 2.6547513039053556, "ewc_loss": 0.15234375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.563534736633301, "learning_rate": 1e-05, "loss": 0.5508, "mean_token_accuracy": 0.8680198192596436, "num_tokens": 792538954.0, "step": 20869 }, { "epoch": 2.654878514183946, "ewc_loss": 0.140625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00011157989501953125, "grad_norm": 6.763424396514893, "learning_rate": 1e-05, "loss": 0.5362, "mean_token_accuracy": 0.8626366853713989, "num_tokens": 792577037.0, "step": 20870 }, { "epoch": 2.6550057244625367, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00011014938354492188, "grad_norm": 7.067280292510986, "learning_rate": 1e-05, "loss": 0.5215, "mean_token_accuracy": 0.8651490807533264, "num_tokens": 792609538.0, "step": 20871 }, { "epoch": 2.655132934741127, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010967254638671875, "grad_norm": 6.821218967437744, "learning_rate": 1e-05, "loss": 0.504, "mean_token_accuracy": 0.877016544342041, "num_tokens": 792647793.0, "step": 20872 }, { "epoch": 2.6552601450197177, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.536527156829834, "learning_rate": 1e-05, "loss": 0.4607, "mean_token_accuracy": 0.8812355995178223, "num_tokens": 792686746.0, "step": 20873 }, { "epoch": 2.6553873552983083, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 6.900640487670898, "learning_rate": 1e-05, "loss": 0.466, "mean_token_accuracy": 0.8799670338630676, "num_tokens": 792716324.0, "step": 20874 }, { "epoch": 2.6555145655768984, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.130923748016357, "learning_rate": 1e-05, "loss": 0.4537, "mean_token_accuracy": 0.8823566436767578, "num_tokens": 792757242.0, "step": 20875 }, { "epoch": 2.6556417758554893, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 7.110080718994141, "learning_rate": 1e-05, "loss": 0.5017, "mean_token_accuracy": 0.8718856573104858, "num_tokens": 792795545.0, "step": 20876 }, { "epoch": 2.6557689861340794, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 5.926026344299316, "learning_rate": 1e-05, "loss": 0.4713, "mean_token_accuracy": 0.8753429055213928, "num_tokens": 792835772.0, "step": 20877 }, { "epoch": 2.6558961964126704, "ewc_loss": 0.1474609375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.000118255615234375, "grad_norm": 7.442470550537109, "learning_rate": 1e-05, "loss": 0.5353, "mean_token_accuracy": 0.8660646677017212, "num_tokens": 792877901.0, "step": 20878 }, { "epoch": 2.6560234066912605, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 5.965921401977539, "learning_rate": 1e-05, "loss": 0.4779, "mean_token_accuracy": 0.8768498301506042, "num_tokens": 792918528.0, "step": 20879 }, { "epoch": 2.6561506169698514, "ewc_loss": 0.15625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00012683868408203125, "grad_norm": 7.736640453338623, "learning_rate": 1e-05, "loss": 0.4595, "mean_token_accuracy": 0.8905420303344727, "num_tokens": 792954251.0, "step": 20880 }, { "epoch": 2.6562778272484415, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.304759979248047, "learning_rate": 1e-05, "loss": 0.4881, "mean_token_accuracy": 0.8788487315177917, "num_tokens": 792991452.0, "step": 20881 }, { "epoch": 2.656405037527032, "ewc_loss": 0.154296875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00012493133544921875, "grad_norm": 7.724874019622803, "learning_rate": 1e-05, "loss": 0.521, "mean_token_accuracy": 0.8705639839172363, "num_tokens": 793024488.0, "step": 20882 }, { "epoch": 2.6565322478056226, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00011014938354492188, "grad_norm": 6.620713710784912, "learning_rate": 1e-05, "loss": 0.4749, "mean_token_accuracy": 0.8850961327552795, "num_tokens": 793064070.0, "step": 20883 }, { "epoch": 2.656659458084213, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00011777877807617188, "grad_norm": 7.402867794036865, "learning_rate": 1e-05, "loss": 0.5235, "mean_token_accuracy": 0.8667556643486023, "num_tokens": 793103088.0, "step": 20884 }, { "epoch": 2.6567866683628036, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00011014938354492188, "grad_norm": 6.687004566192627, "learning_rate": 1e-05, "loss": 0.5415, "mean_token_accuracy": 0.8629394769668579, "num_tokens": 793146415.0, "step": 20885 }, { "epoch": 2.656913878641394, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010967254638671875, "grad_norm": 7.054886341094971, "learning_rate": 1e-05, "loss": 0.4951, "mean_token_accuracy": 0.8795135617256165, "num_tokens": 793181588.0, "step": 20886 }, { "epoch": 2.6570410889199847, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.614415168762207, "learning_rate": 1e-05, "loss": 0.5249, "mean_token_accuracy": 0.8596465587615967, "num_tokens": 793217764.0, "step": 20887 }, { "epoch": 2.6571682991985752, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.7934441566467285, "learning_rate": 1e-05, "loss": 0.4644, "mean_token_accuracy": 0.8824869394302368, "num_tokens": 793255945.0, "step": 20888 }, { "epoch": 2.6572955094771658, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.4999165534973145, "learning_rate": 1e-05, "loss": 0.4951, "mean_token_accuracy": 0.8746048212051392, "num_tokens": 793288981.0, "step": 20889 }, { "epoch": 2.6574227197557563, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.648491859436035, "learning_rate": 1e-05, "loss": 0.4707, "mean_token_accuracy": 0.8797051906585693, "num_tokens": 793327635.0, "step": 20890 }, { "epoch": 2.657549930034347, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.365128040313721, "learning_rate": 1e-05, "loss": 0.471, "mean_token_accuracy": 0.8768029808998108, "num_tokens": 793357077.0, "step": 20891 }, { "epoch": 2.6576771403129373, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.617197513580322, "learning_rate": 1e-05, "loss": 0.4317, "mean_token_accuracy": 0.8929599523544312, "num_tokens": 793389664.0, "step": 20892 }, { "epoch": 2.657804350591528, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.214977741241455, "learning_rate": 1e-05, "loss": 0.4356, "mean_token_accuracy": 0.8907995820045471, "num_tokens": 793429951.0, "step": 20893 }, { "epoch": 2.6579315608701184, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.692975997924805, "learning_rate": 1e-05, "loss": 0.5176, "mean_token_accuracy": 0.8663519024848938, "num_tokens": 793471404.0, "step": 20894 }, { "epoch": 2.658058771148709, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 6.056218147277832, "learning_rate": 1e-05, "loss": 0.4814, "mean_token_accuracy": 0.8766759634017944, "num_tokens": 793510140.0, "step": 20895 }, { "epoch": 2.6581859814272994, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.891528606414795, "learning_rate": 1e-05, "loss": 0.4884, "mean_token_accuracy": 0.8785281777381897, "num_tokens": 793553927.0, "step": 20896 }, { "epoch": 2.65831319170589, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.906794548034668, "learning_rate": 1e-05, "loss": 0.4594, "mean_token_accuracy": 0.8829212784767151, "num_tokens": 793590413.0, "step": 20897 }, { "epoch": 2.6584404019844805, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00011348724365234375, "grad_norm": 7.227565765380859, "learning_rate": 1e-05, "loss": 0.553, "mean_token_accuracy": 0.8580600023269653, "num_tokens": 793626320.0, "step": 20898 }, { "epoch": 2.658567612263071, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 5.860885143280029, "learning_rate": 1e-05, "loss": 0.4529, "mean_token_accuracy": 0.877997875213623, "num_tokens": 793665733.0, "step": 20899 }, { "epoch": 2.658694822541661, "ewc_loss": 0.1533203125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.0001239776611328125, "grad_norm": 7.624965667724609, "learning_rate": 1e-05, "loss": 0.5044, "mean_token_accuracy": 0.8774905204772949, "num_tokens": 793705593.0, "step": 20900 }, { "epoch": 2.658822032820252, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.097782611846924, "learning_rate": 1e-05, "loss": 0.4569, "mean_token_accuracy": 0.882031261920929, "num_tokens": 793744874.0, "step": 20901 }, { "epoch": 2.658949243098842, "ewc_loss": 0.158203125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00012874603271484375, "grad_norm": 7.800352573394775, "learning_rate": 1e-05, "loss": 0.5496, "mean_token_accuracy": 0.8632553219795227, "num_tokens": 793786598.0, "step": 20902 }, { "epoch": 2.659076453377433, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 6.545100688934326, "learning_rate": 1e-05, "loss": 0.4829, "mean_token_accuracy": 0.8758010268211365, "num_tokens": 793826548.0, "step": 20903 }, { "epoch": 2.6592036636560232, "ewc_loss": 0.1494140625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.0001201629638671875, "grad_norm": 7.512495040893555, "learning_rate": 1e-05, "loss": 0.5531, "mean_token_accuracy": 0.8626919984817505, "num_tokens": 793868411.0, "step": 20904 }, { "epoch": 2.6593308739346138, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.0001125335693359375, "grad_norm": 6.775219917297363, "learning_rate": 1e-05, "loss": 0.5173, "mean_token_accuracy": 0.8711665868759155, "num_tokens": 793905931.0, "step": 20905 }, { "epoch": 2.6594580842132043, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 7.001857757568359, "learning_rate": 1e-05, "loss": 0.495, "mean_token_accuracy": 0.8741801977157593, "num_tokens": 793939711.0, "step": 20906 }, { "epoch": 2.659585294491795, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010967254638671875, "grad_norm": 6.803546905517578, "learning_rate": 1e-05, "loss": 0.489, "mean_token_accuracy": 0.8790566921234131, "num_tokens": 793976746.0, "step": 20907 }, { "epoch": 2.6597125047703853, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.506357669830322, "learning_rate": 1e-05, "loss": 0.5314, "mean_token_accuracy": 0.8586336374282837, "num_tokens": 794014132.0, "step": 20908 }, { "epoch": 2.659839715048976, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 6.88460636138916, "learning_rate": 1e-05, "loss": 0.5353, "mean_token_accuracy": 0.8634471297264099, "num_tokens": 794051958.0, "step": 20909 }, { "epoch": 2.6599669253275664, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 6.12789249420166, "learning_rate": 1e-05, "loss": 0.4556, "mean_token_accuracy": 0.8803337812423706, "num_tokens": 794087494.0, "step": 20910 }, { "epoch": 2.660094135606157, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00011014938354492188, "grad_norm": 7.088073253631592, "learning_rate": 1e-05, "loss": 0.5369, "mean_token_accuracy": 0.8638170957565308, "num_tokens": 794123156.0, "step": 20911 }, { "epoch": 2.6602213458847475, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 5.931123733520508, "learning_rate": 1e-05, "loss": 0.4999, "mean_token_accuracy": 0.8643103241920471, "num_tokens": 794169023.0, "step": 20912 }, { "epoch": 2.660348556163338, "ewc_loss": 0.1455078125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.0001163482666015625, "grad_norm": 7.400208950042725, "learning_rate": 1e-05, "loss": 0.4559, "mean_token_accuracy": 0.8891240358352661, "num_tokens": 794202145.0, "step": 20913 }, { "epoch": 2.6604757664419285, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 5.980818748474121, "learning_rate": 1e-05, "loss": 0.4424, "mean_token_accuracy": 0.8917769193649292, "num_tokens": 794239791.0, "step": 20914 }, { "epoch": 2.660602976720519, "ewc_loss": 0.1533203125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.0001239776611328125, "grad_norm": 7.652184009552002, "learning_rate": 1e-05, "loss": 0.485, "mean_token_accuracy": 0.8844490647315979, "num_tokens": 794285184.0, "step": 20915 }, { "epoch": 2.6607301869991096, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.260210990905762, "learning_rate": 1e-05, "loss": 0.4384, "mean_token_accuracy": 0.8925639390945435, "num_tokens": 794324705.0, "step": 20916 }, { "epoch": 2.6608573972777, "ewc_loss": 0.154296875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00012493133544921875, "grad_norm": 7.660856246948242, "learning_rate": 1e-05, "loss": 0.5183, "mean_token_accuracy": 0.8707515001296997, "num_tokens": 794360712.0, "step": 20917 }, { "epoch": 2.6609846075562906, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.550417423248291, "learning_rate": 1e-05, "loss": 0.4704, "mean_token_accuracy": 0.8825699090957642, "num_tokens": 794401634.0, "step": 20918 }, { "epoch": 2.661111817834881, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00011777877807617188, "grad_norm": 7.398808002471924, "learning_rate": 1e-05, "loss": 0.4841, "mean_token_accuracy": 0.8800767660140991, "num_tokens": 794433460.0, "step": 20919 }, { "epoch": 2.6612390281134717, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 6.6445417404174805, "learning_rate": 1e-05, "loss": 0.4338, "mean_token_accuracy": 0.8939730525016785, "num_tokens": 794468258.0, "step": 20920 }, { "epoch": 2.661366238392062, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00011014938354492188, "grad_norm": 7.056203365325928, "learning_rate": 1e-05, "loss": 0.5077, "mean_token_accuracy": 0.8740242719650269, "num_tokens": 794505493.0, "step": 20921 }, { "epoch": 2.6614934486706527, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.5888142585754395, "learning_rate": 1e-05, "loss": 0.4132, "mean_token_accuracy": 0.900382936000824, "num_tokens": 794539370.0, "step": 20922 }, { "epoch": 2.661620658949243, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.783262252807617, "learning_rate": 1e-05, "loss": 0.4905, "mean_token_accuracy": 0.873881459236145, "num_tokens": 794580820.0, "step": 20923 }, { "epoch": 2.661747869227834, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.489286422729492, "learning_rate": 1e-05, "loss": 0.4655, "mean_token_accuracy": 0.8838720917701721, "num_tokens": 794617503.0, "step": 20924 }, { "epoch": 2.661875079506424, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.640670299530029, "learning_rate": 1e-05, "loss": 0.4771, "mean_token_accuracy": 0.8787020444869995, "num_tokens": 794653701.0, "step": 20925 }, { "epoch": 2.662002289785015, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.374189853668213, "learning_rate": 1e-05, "loss": 0.4781, "mean_token_accuracy": 0.873699426651001, "num_tokens": 794689938.0, "step": 20926 }, { "epoch": 2.662129500063605, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.6081695556640625, "learning_rate": 1e-05, "loss": 0.5013, "mean_token_accuracy": 0.8673909902572632, "num_tokens": 794730460.0, "step": 20927 }, { "epoch": 2.662256710342196, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.22089958190918, "learning_rate": 1e-05, "loss": 0.4465, "mean_token_accuracy": 0.8847182989120483, "num_tokens": 794770686.0, "step": 20928 }, { "epoch": 2.662383920620786, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.6739888191223145, "learning_rate": 1e-05, "loss": 0.5298, "mean_token_accuracy": 0.8644163608551025, "num_tokens": 794814396.0, "step": 20929 }, { "epoch": 2.6625111308993765, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 6.065199851989746, "learning_rate": 1e-05, "loss": 0.4674, "mean_token_accuracy": 0.8815544843673706, "num_tokens": 794857425.0, "step": 20930 }, { "epoch": 2.662638341177967, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.866983413696289, "learning_rate": 1e-05, "loss": 0.4597, "mean_token_accuracy": 0.8826819658279419, "num_tokens": 794894115.0, "step": 20931 }, { "epoch": 2.6627655514565576, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.984803199768066, "learning_rate": 1e-05, "loss": 0.4791, "mean_token_accuracy": 0.8737217783927917, "num_tokens": 794929423.0, "step": 20932 }, { "epoch": 2.662892761735148, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00011014938354492188, "grad_norm": 7.1073994636535645, "learning_rate": 1e-05, "loss": 0.4937, "mean_token_accuracy": 0.8757354617118835, "num_tokens": 794966019.0, "step": 20933 }, { "epoch": 2.6630199720137386, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.863588809967041, "learning_rate": 1e-05, "loss": 0.4768, "mean_token_accuracy": 0.8739420771598816, "num_tokens": 795006501.0, "step": 20934 }, { "epoch": 2.663147182292329, "ewc_loss": 0.1484375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00011968612670898438, "grad_norm": 7.464919567108154, "learning_rate": 1e-05, "loss": 0.5065, "mean_token_accuracy": 0.8773613572120667, "num_tokens": 795048068.0, "step": 20935 }, { "epoch": 2.6632743925709197, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.000020503997803, "learning_rate": 1e-05, "loss": 0.5047, "mean_token_accuracy": 0.8658718466758728, "num_tokens": 795084918.0, "step": 20936 }, { "epoch": 2.66340160284951, "ewc_loss": 0.1552734375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.000125885009765625, "grad_norm": 7.707596778869629, "learning_rate": 1e-05, "loss": 0.4979, "mean_token_accuracy": 0.8830404281616211, "num_tokens": 795119261.0, "step": 20937 }, { "epoch": 2.6635288131281007, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.3659539222717285, "learning_rate": 1e-05, "loss": 0.4538, "mean_token_accuracy": 0.8867557048797607, "num_tokens": 795162559.0, "step": 20938 }, { "epoch": 2.6636560234066913, "ewc_loss": 0.15234375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.55353307723999, "learning_rate": 1e-05, "loss": 0.49, "mean_token_accuracy": 0.8818863034248352, "num_tokens": 795200853.0, "step": 20939 }, { "epoch": 2.663783233685282, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00011014938354492188, "grad_norm": 6.675041675567627, "learning_rate": 1e-05, "loss": 0.4683, "mean_token_accuracy": 0.8854516744613647, "num_tokens": 795238405.0, "step": 20940 }, { "epoch": 2.6639104439638723, "ewc_loss": 0.140625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00011157989501953125, "grad_norm": 7.087430953979492, "learning_rate": 1e-05, "loss": 0.5059, "mean_token_accuracy": 0.8730008006095886, "num_tokens": 795278341.0, "step": 20941 }, { "epoch": 2.664037654242463, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 6.766659736633301, "learning_rate": 1e-05, "loss": 0.5146, "mean_token_accuracy": 0.8707935214042664, "num_tokens": 795319108.0, "step": 20942 }, { "epoch": 2.6641648645210534, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.581387042999268, "learning_rate": 1e-05, "loss": 0.4796, "mean_token_accuracy": 0.8760213851928711, "num_tokens": 795354641.0, "step": 20943 }, { "epoch": 2.664292074799644, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.843020439147949, "learning_rate": 1e-05, "loss": 0.539, "mean_token_accuracy": 0.8628510236740112, "num_tokens": 795399098.0, "step": 20944 }, { "epoch": 2.6644192850782344, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.146585941314697, "learning_rate": 1e-05, "loss": 0.4556, "mean_token_accuracy": 0.8819193243980408, "num_tokens": 795444191.0, "step": 20945 }, { "epoch": 2.664546495356825, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00011014938354492188, "grad_norm": 7.076348304748535, "learning_rate": 1e-05, "loss": 0.4883, "mean_token_accuracy": 0.8755279779434204, "num_tokens": 795476450.0, "step": 20946 }, { "epoch": 2.6646737056354155, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 5.913759231567383, "learning_rate": 1e-05, "loss": 0.4743, "mean_token_accuracy": 0.8750357627868652, "num_tokens": 795518740.0, "step": 20947 }, { "epoch": 2.6648009159140056, "ewc_loss": 0.1455078125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00011682510375976562, "grad_norm": 7.399699687957764, "learning_rate": 1e-05, "loss": 0.5238, "mean_token_accuracy": 0.8717978000640869, "num_tokens": 795558944.0, "step": 20948 }, { "epoch": 2.6649281261925966, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 5.961846351623535, "learning_rate": 1e-05, "loss": 0.4579, "mean_token_accuracy": 0.8790353536605835, "num_tokens": 795592685.0, "step": 20949 }, { "epoch": 2.6650553364711866, "ewc_loss": 0.154296875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00012493133544921875, "grad_norm": 7.686119079589844, "learning_rate": 1e-05, "loss": 0.5312, "mean_token_accuracy": 0.8705503344535828, "num_tokens": 795627878.0, "step": 20950 }, { "epoch": 2.6651825467497776, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.2548322677612305, "learning_rate": 1e-05, "loss": 0.5211, "mean_token_accuracy": 0.8667703866958618, "num_tokens": 795666784.0, "step": 20951 }, { "epoch": 2.6653097570283677, "ewc_loss": 0.154296875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00012493133544921875, "grad_norm": 7.684600353240967, "learning_rate": 1e-05, "loss": 0.512, "mean_token_accuracy": 0.8775127530097961, "num_tokens": 795701921.0, "step": 20952 }, { "epoch": 2.6654369673069587, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.54144287109375, "learning_rate": 1e-05, "loss": 0.51, "mean_token_accuracy": 0.87083899974823, "num_tokens": 795744497.0, "step": 20953 }, { "epoch": 2.6655641775855488, "ewc_loss": 0.1484375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00011968612670898438, "grad_norm": 7.437565326690674, "learning_rate": 1e-05, "loss": 0.5413, "mean_token_accuracy": 0.8635179400444031, "num_tokens": 795786871.0, "step": 20954 }, { "epoch": 2.6656913878641393, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 6.639771938323975, "learning_rate": 1e-05, "loss": 0.4627, "mean_token_accuracy": 0.8865407705307007, "num_tokens": 795829403.0, "step": 20955 }, { "epoch": 2.66581859814273, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.0001125335693359375, "grad_norm": 7.100732803344727, "learning_rate": 1e-05, "loss": 0.5488, "mean_token_accuracy": 0.8605715036392212, "num_tokens": 795867352.0, "step": 20956 }, { "epoch": 2.6659458084213203, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.598262310028076, "learning_rate": 1e-05, "loss": 0.5225, "mean_token_accuracy": 0.868750810623169, "num_tokens": 795907209.0, "step": 20957 }, { "epoch": 2.666073018699911, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.8347086906433105, "learning_rate": 1e-05, "loss": 0.5058, "mean_token_accuracy": 0.873436689376831, "num_tokens": 795946515.0, "step": 20958 }, { "epoch": 2.6662002289785014, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.501695156097412, "learning_rate": 1e-05, "loss": 0.4644, "mean_token_accuracy": 0.8854063749313354, "num_tokens": 795983328.0, "step": 20959 }, { "epoch": 2.666327439257092, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.68150520324707, "learning_rate": 1e-05, "loss": 0.5212, "mean_token_accuracy": 0.8589428663253784, "num_tokens": 796018609.0, "step": 20960 }, { "epoch": 2.6664546495356825, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.37047004699707, "learning_rate": 1e-05, "loss": 0.4979, "mean_token_accuracy": 0.8703674674034119, "num_tokens": 796059676.0, "step": 20961 }, { "epoch": 2.666581859814273, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.6553449630737305, "learning_rate": 1e-05, "loss": 0.4885, "mean_token_accuracy": 0.8739666938781738, "num_tokens": 796100896.0, "step": 20962 }, { "epoch": 2.6667090700928635, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.217586994171143, "learning_rate": 1e-05, "loss": 0.4343, "mean_token_accuracy": 0.8908042311668396, "num_tokens": 796140513.0, "step": 20963 }, { "epoch": 2.666836280371454, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.733185768127441, "learning_rate": 1e-05, "loss": 0.4514, "mean_token_accuracy": 0.8878182172775269, "num_tokens": 796174339.0, "step": 20964 }, { "epoch": 2.6669634906500446, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.052537441253662, "learning_rate": 1e-05, "loss": 0.4724, "mean_token_accuracy": 0.8795570731163025, "num_tokens": 796216624.0, "step": 20965 }, { "epoch": 2.667090700928635, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.936331272125244, "learning_rate": 1e-05, "loss": 0.5233, "mean_token_accuracy": 0.8622987270355225, "num_tokens": 796250681.0, "step": 20966 }, { "epoch": 2.6672179112072256, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.920417785644531, "learning_rate": 1e-05, "loss": 0.4554, "mean_token_accuracy": 0.8799081444740295, "num_tokens": 796284860.0, "step": 20967 }, { "epoch": 2.667345121485816, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00011444091796875, "grad_norm": 7.249430179595947, "learning_rate": 1e-05, "loss": 0.4557, "mean_token_accuracy": 0.8898786306381226, "num_tokens": 796325833.0, "step": 20968 }, { "epoch": 2.6674723317644067, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 5.905074119567871, "learning_rate": 1e-05, "loss": 0.4419, "mean_token_accuracy": 0.8859912157058716, "num_tokens": 796354955.0, "step": 20969 }, { "epoch": 2.667599542042997, "ewc_loss": 0.15234375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.594826698303223, "learning_rate": 1e-05, "loss": 0.5096, "mean_token_accuracy": 0.8755475282669067, "num_tokens": 796390505.0, "step": 20970 }, { "epoch": 2.6677267523215877, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.140364170074463, "learning_rate": 1e-05, "loss": 0.4558, "mean_token_accuracy": 0.8853387832641602, "num_tokens": 796428003.0, "step": 20971 }, { "epoch": 2.6678539626001783, "ewc_loss": 0.15625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00012683868408203125, "grad_norm": 7.695431709289551, "learning_rate": 1e-05, "loss": 0.4802, "mean_token_accuracy": 0.8851083517074585, "num_tokens": 796467396.0, "step": 20972 }, { "epoch": 2.6679811728787683, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 6.524107933044434, "learning_rate": 1e-05, "loss": 0.5009, "mean_token_accuracy": 0.8743689656257629, "num_tokens": 796505320.0, "step": 20973 }, { "epoch": 2.6681083831573593, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00011777877807617188, "grad_norm": 7.3867950439453125, "learning_rate": 1e-05, "loss": 0.4822, "mean_token_accuracy": 0.8852876424789429, "num_tokens": 796542109.0, "step": 20974 }, { "epoch": 2.6682355934359494, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 6.721557140350342, "learning_rate": 1e-05, "loss": 0.4985, "mean_token_accuracy": 0.8739714622497559, "num_tokens": 796588523.0, "step": 20975 }, { "epoch": 2.6683628037145404, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.881140232086182, "learning_rate": 1e-05, "loss": 0.5471, "mean_token_accuracy": 0.8593064546585083, "num_tokens": 796624493.0, "step": 20976 }, { "epoch": 2.6684900139931305, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 6.778509140014648, "learning_rate": 1e-05, "loss": 0.4548, "mean_token_accuracy": 0.8892850875854492, "num_tokens": 796662147.0, "step": 20977 }, { "epoch": 2.668617224271721, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.388829231262207, "learning_rate": 1e-05, "loss": 0.4877, "mean_token_accuracy": 0.8746863603591919, "num_tokens": 796700024.0, "step": 20978 }, { "epoch": 2.6687444345503115, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 6.908505439758301, "learning_rate": 1e-05, "loss": 0.4651, "mean_token_accuracy": 0.8821221590042114, "num_tokens": 796735058.0, "step": 20979 }, { "epoch": 2.668871644828902, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 6.031033992767334, "learning_rate": 1e-05, "loss": 0.4859, "mean_token_accuracy": 0.8695008754730225, "num_tokens": 796778419.0, "step": 20980 }, { "epoch": 2.6689988551074926, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.0001125335693359375, "grad_norm": 7.184506893157959, "learning_rate": 1e-05, "loss": 0.4717, "mean_token_accuracy": 0.8826438188552856, "num_tokens": 796817495.0, "step": 20981 }, { "epoch": 2.669126065386083, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 5.896661758422852, "learning_rate": 1e-05, "loss": 0.483, "mean_token_accuracy": 0.8730975389480591, "num_tokens": 796850432.0, "step": 20982 }, { "epoch": 2.6692532756646736, "ewc_loss": 0.150390625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00012111663818359375, "grad_norm": 7.5417680740356445, "learning_rate": 1e-05, "loss": 0.5272, "mean_token_accuracy": 0.8705212473869324, "num_tokens": 796889584.0, "step": 20983 }, { "epoch": 2.669380485943264, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.039985179901123, "learning_rate": 1e-05, "loss": 0.4694, "mean_token_accuracy": 0.8763443827629089, "num_tokens": 796924052.0, "step": 20984 }, { "epoch": 2.6695076962218547, "ewc_loss": 0.1552734375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.000125885009765625, "grad_norm": 7.770593166351318, "learning_rate": 1e-05, "loss": 0.5285, "mean_token_accuracy": 0.8681403398513794, "num_tokens": 796963795.0, "step": 20985 }, { "epoch": 2.669634906500445, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.4070892333984375, "learning_rate": 1e-05, "loss": 0.4535, "mean_token_accuracy": 0.8870717883110046, "num_tokens": 797002621.0, "step": 20986 }, { "epoch": 2.6697621167790357, "ewc_loss": 0.1533203125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.0001239776611328125, "grad_norm": 7.624938488006592, "learning_rate": 1e-05, "loss": 0.5715, "mean_token_accuracy": 0.8539121150970459, "num_tokens": 797047173.0, "step": 20987 }, { "epoch": 2.6698893270576263, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 6.648516654968262, "learning_rate": 1e-05, "loss": 0.5039, "mean_token_accuracy": 0.8745914697647095, "num_tokens": 797088394.0, "step": 20988 }, { "epoch": 2.670016537336217, "ewc_loss": 0.1455078125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.0001163482666015625, "grad_norm": 7.2864580154418945, "learning_rate": 1e-05, "loss": 0.482, "mean_token_accuracy": 0.8848756551742554, "num_tokens": 797126108.0, "step": 20989 }, { "epoch": 2.6701437476148073, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 6.6683807373046875, "learning_rate": 1e-05, "loss": 0.4807, "mean_token_accuracy": 0.8830618262290955, "num_tokens": 797162718.0, "step": 20990 }, { "epoch": 2.670270957893398, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.954492568969727, "learning_rate": 1e-05, "loss": 0.5199, "mean_token_accuracy": 0.8667072057723999, "num_tokens": 797197552.0, "step": 20991 }, { "epoch": 2.6703981681719884, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.587381839752197, "learning_rate": 1e-05, "loss": 0.5455, "mean_token_accuracy": 0.8556311726570129, "num_tokens": 797232523.0, "step": 20992 }, { "epoch": 2.670525378450579, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.716342926025391, "learning_rate": 1e-05, "loss": 0.4698, "mean_token_accuracy": 0.879538357257843, "num_tokens": 797266620.0, "step": 20993 }, { "epoch": 2.6706525887291694, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.460195541381836, "learning_rate": 1e-05, "loss": 0.4643, "mean_token_accuracy": 0.8838238716125488, "num_tokens": 797305378.0, "step": 20994 }, { "epoch": 2.67077979900776, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.623281478881836, "learning_rate": 1e-05, "loss": 0.4598, "mean_token_accuracy": 0.8828821182250977, "num_tokens": 797343269.0, "step": 20995 }, { "epoch": 2.6709070092863505, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.302792549133301, "learning_rate": 1e-05, "loss": 0.3993, "mean_token_accuracy": 0.9010373950004578, "num_tokens": 797375034.0, "step": 20996 }, { "epoch": 2.671034219564941, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.664947032928467, "learning_rate": 1e-05, "loss": 0.4891, "mean_token_accuracy": 0.8727458119392395, "num_tokens": 797410869.0, "step": 20997 }, { "epoch": 2.671161429843531, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.139883518218994, "learning_rate": 1e-05, "loss": 0.4474, "mean_token_accuracy": 0.8869522213935852, "num_tokens": 797449189.0, "step": 20998 }, { "epoch": 2.671288640122122, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.800176620483398, "learning_rate": 1e-05, "loss": 0.4692, "mean_token_accuracy": 0.8814737200737, "num_tokens": 797492182.0, "step": 20999 }, { "epoch": 2.671415850400712, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 5.964768409729004, "learning_rate": 1e-05, "loss": 0.4101, "mean_token_accuracy": 0.8941760063171387, "num_tokens": 797529043.0, "step": 21000 }, { "epoch": 2.671543060679303, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00011014938354492188, "grad_norm": 7.083416938781738, "learning_rate": 1e-05, "loss": 0.513, "mean_token_accuracy": 0.8704541921615601, "num_tokens": 797560630.0, "step": 21001 }, { "epoch": 2.6716702709578932, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.862268447875977, "learning_rate": 1e-05, "loss": 0.4931, "mean_token_accuracy": 0.8714083433151245, "num_tokens": 797598667.0, "step": 21002 }, { "epoch": 2.6717974812364838, "ewc_loss": 0.1494140625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.0001201629638671875, "grad_norm": 7.4684157371521, "learning_rate": 1e-05, "loss": 0.4732, "mean_token_accuracy": 0.8867782354354858, "num_tokens": 797637420.0, "step": 21003 }, { "epoch": 2.6719246915150743, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 5.966523170471191, "learning_rate": 1e-05, "loss": 0.4626, "mean_token_accuracy": 0.8826454281806946, "num_tokens": 797670969.0, "step": 21004 }, { "epoch": 2.672051901793665, "ewc_loss": 0.158203125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00012874603271484375, "grad_norm": 7.770328998565674, "learning_rate": 1e-05, "loss": 0.5001, "mean_token_accuracy": 0.8785349726676941, "num_tokens": 797711915.0, "step": 21005 }, { "epoch": 2.6721791120722553, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.365726470947266, "learning_rate": 1e-05, "loss": 0.4831, "mean_token_accuracy": 0.8752175569534302, "num_tokens": 797751036.0, "step": 21006 }, { "epoch": 2.672306322350846, "ewc_loss": 0.154296875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00012493133544921875, "grad_norm": 7.661888122558594, "learning_rate": 1e-05, "loss": 0.5264, "mean_token_accuracy": 0.8698413372039795, "num_tokens": 797793999.0, "step": 21007 }, { "epoch": 2.6724335326294364, "ewc_loss": 0.140625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00011157989501953125, "grad_norm": 6.704616069793701, "learning_rate": 1e-05, "loss": 0.5031, "mean_token_accuracy": 0.8752468824386597, "num_tokens": 797834836.0, "step": 21008 }, { "epoch": 2.672560742908027, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00011301040649414062, "grad_norm": 7.204422473907471, "learning_rate": 1e-05, "loss": 0.53, "mean_token_accuracy": 0.8662738800048828, "num_tokens": 797872855.0, "step": 21009 }, { "epoch": 2.6726879531866174, "ewc_loss": 0.140625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00011110305786132812, "grad_norm": 6.801244735717773, "learning_rate": 1e-05, "loss": 0.5129, "mean_token_accuracy": 0.8741808533668518, "num_tokens": 797910925.0, "step": 21010 }, { "epoch": 2.672815163465208, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.685955047607422, "learning_rate": 1e-05, "loss": 0.493, "mean_token_accuracy": 0.8700481653213501, "num_tokens": 797952201.0, "step": 21011 }, { "epoch": 2.6729423737437985, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 6.842397689819336, "learning_rate": 1e-05, "loss": 0.5498, "mean_token_accuracy": 0.8576335906982422, "num_tokens": 797988706.0, "step": 21012 }, { "epoch": 2.673069584022389, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.25856351852417, "learning_rate": 1e-05, "loss": 0.4982, "mean_token_accuracy": 0.8703915476799011, "num_tokens": 798028438.0, "step": 21013 }, { "epoch": 2.6731967943009796, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 6.9882402420043945, "learning_rate": 1e-05, "loss": 0.4566, "mean_token_accuracy": 0.8874621987342834, "num_tokens": 798062864.0, "step": 21014 }, { "epoch": 2.67332400457957, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 5.996420383453369, "learning_rate": 1e-05, "loss": 0.458, "mean_token_accuracy": 0.8809462785720825, "num_tokens": 798098271.0, "step": 21015 }, { "epoch": 2.6734512148581606, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00011348724365234375, "grad_norm": 7.256039619445801, "learning_rate": 1e-05, "loss": 0.4766, "mean_token_accuracy": 0.8830022811889648, "num_tokens": 798131286.0, "step": 21016 }, { "epoch": 2.673578425136751, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 5.931898593902588, "learning_rate": 1e-05, "loss": 0.4679, "mean_token_accuracy": 0.8778334856033325, "num_tokens": 798164711.0, "step": 21017 }, { "epoch": 2.6737056354153417, "ewc_loss": 0.1513671875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.0001220703125, "grad_norm": 7.5497918128967285, "learning_rate": 1e-05, "loss": 0.5061, "mean_token_accuracy": 0.8768202066421509, "num_tokens": 798205280.0, "step": 21018 }, { "epoch": 2.673832845693932, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.103862285614014, "learning_rate": 1e-05, "loss": 0.4678, "mean_token_accuracy": 0.8775770664215088, "num_tokens": 798246561.0, "step": 21019 }, { "epoch": 2.6739600559725227, "ewc_loss": 0.15625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00012683868408203125, "grad_norm": 7.6890339851379395, "learning_rate": 1e-05, "loss": 0.496, "mean_token_accuracy": 0.8789661526679993, "num_tokens": 798287494.0, "step": 21020 }, { "epoch": 2.674087266251113, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.421586990356445, "learning_rate": 1e-05, "loss": 0.4762, "mean_token_accuracy": 0.8784738779067993, "num_tokens": 798325305.0, "step": 21021 }, { "epoch": 2.674214476529704, "ewc_loss": 0.150390625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00012111663818359375, "grad_norm": 7.52739953994751, "learning_rate": 1e-05, "loss": 0.4678, "mean_token_accuracy": 0.8872382640838623, "num_tokens": 798363083.0, "step": 21022 }, { "epoch": 2.674341686808294, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 6.604341506958008, "learning_rate": 1e-05, "loss": 0.5556, "mean_token_accuracy": 0.8580405116081238, "num_tokens": 798403324.0, "step": 21023 }, { "epoch": 2.674468897086885, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00011491775512695312, "grad_norm": 7.211480617523193, "learning_rate": 1e-05, "loss": 0.4721, "mean_token_accuracy": 0.8834998607635498, "num_tokens": 798441548.0, "step": 21024 }, { "epoch": 2.674596107365475, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.59283447265625, "learning_rate": 1e-05, "loss": 0.508, "mean_token_accuracy": 0.8716340661048889, "num_tokens": 798486038.0, "step": 21025 }, { "epoch": 2.674723317644066, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.916153430938721, "learning_rate": 1e-05, "loss": 0.4608, "mean_token_accuracy": 0.8858940005302429, "num_tokens": 798530657.0, "step": 21026 }, { "epoch": 2.674850527922656, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.520358562469482, "learning_rate": 1e-05, "loss": 0.4748, "mean_token_accuracy": 0.8783705234527588, "num_tokens": 798571027.0, "step": 21027 }, { "epoch": 2.6749777382012465, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.724308013916016, "learning_rate": 1e-05, "loss": 0.4623, "mean_token_accuracy": 0.8853330612182617, "num_tokens": 798609874.0, "step": 21028 }, { "epoch": 2.675104948479837, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.423453330993652, "learning_rate": 1e-05, "loss": 0.5023, "mean_token_accuracy": 0.8697535991668701, "num_tokens": 798645191.0, "step": 21029 }, { "epoch": 2.6752321587584276, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.6340837478637695, "learning_rate": 1e-05, "loss": 0.4582, "mean_token_accuracy": 0.8846815824508667, "num_tokens": 798688445.0, "step": 21030 }, { "epoch": 2.675359369037018, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.283995151519775, "learning_rate": 1e-05, "loss": 0.478, "mean_token_accuracy": 0.8789024949073792, "num_tokens": 798729582.0, "step": 21031 }, { "epoch": 2.6754865793156086, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.663444995880127, "learning_rate": 1e-05, "loss": 0.4545, "mean_token_accuracy": 0.8836848139762878, "num_tokens": 798765178.0, "step": 21032 }, { "epoch": 2.675613789594199, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.1367292404174805, "learning_rate": 1e-05, "loss": 0.4819, "mean_token_accuracy": 0.8769123554229736, "num_tokens": 798806048.0, "step": 21033 }, { "epoch": 2.6757409998727897, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.796191692352295, "learning_rate": 1e-05, "loss": 0.4702, "mean_token_accuracy": 0.8836457133293152, "num_tokens": 798845554.0, "step": 21034 }, { "epoch": 2.67586821015138, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 5.994481563568115, "learning_rate": 1e-05, "loss": 0.4211, "mean_token_accuracy": 0.8939328193664551, "num_tokens": 798883124.0, "step": 21035 }, { "epoch": 2.6759954204299707, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 7.036796569824219, "learning_rate": 1e-05, "loss": 0.4552, "mean_token_accuracy": 0.8911611437797546, "num_tokens": 798916040.0, "step": 21036 }, { "epoch": 2.6761226307085613, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 5.893643379211426, "learning_rate": 1e-05, "loss": 0.4709, "mean_token_accuracy": 0.8766287565231323, "num_tokens": 798952562.0, "step": 21037 }, { "epoch": 2.676249840987152, "ewc_loss": 0.1455078125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.0001163482666015625, "grad_norm": 7.381308078765869, "learning_rate": 1e-05, "loss": 0.5285, "mean_token_accuracy": 0.8713253140449524, "num_tokens": 798990689.0, "step": 21038 }, { "epoch": 2.6763770512657423, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 5.952369689941406, "learning_rate": 1e-05, "loss": 0.4557, "mean_token_accuracy": 0.8825122714042664, "num_tokens": 799029398.0, "step": 21039 }, { "epoch": 2.676504261544333, "ewc_loss": 0.1533203125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.0001239776611328125, "grad_norm": 7.6658759117126465, "learning_rate": 1e-05, "loss": 0.4668, "mean_token_accuracy": 0.8904995918273926, "num_tokens": 799073200.0, "step": 21040 }, { "epoch": 2.6766314718229234, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.279216289520264, "learning_rate": 1e-05, "loss": 0.456, "mean_token_accuracy": 0.884364664554596, "num_tokens": 799115020.0, "step": 21041 }, { "epoch": 2.676758682101514, "ewc_loss": 0.15234375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.613007068634033, "learning_rate": 1e-05, "loss": 0.4673, "mean_token_accuracy": 0.8881606459617615, "num_tokens": 799152348.0, "step": 21042 }, { "epoch": 2.6768858923801044, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 6.614870548248291, "learning_rate": 1e-05, "loss": 0.4885, "mean_token_accuracy": 0.875005841255188, "num_tokens": 799191120.0, "step": 21043 }, { "epoch": 2.677013102658695, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00011396408081054688, "grad_norm": 7.197092533111572, "learning_rate": 1e-05, "loss": 0.4995, "mean_token_accuracy": 0.8780791759490967, "num_tokens": 799229044.0, "step": 21044 }, { "epoch": 2.6771403129372855, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 6.73251485824585, "learning_rate": 1e-05, "loss": 0.4751, "mean_token_accuracy": 0.8845562934875488, "num_tokens": 799269148.0, "step": 21045 }, { "epoch": 2.6772675232158756, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.686837196350098, "learning_rate": 1e-05, "loss": 0.4425, "mean_token_accuracy": 0.886319637298584, "num_tokens": 799305188.0, "step": 21046 }, { "epoch": 2.6773947334944665, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.837126731872559, "learning_rate": 1e-05, "loss": 0.5132, "mean_token_accuracy": 0.8697623014450073, "num_tokens": 799344156.0, "step": 21047 }, { "epoch": 2.6775219437730566, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.225502967834473, "learning_rate": 1e-05, "loss": 0.4971, "mean_token_accuracy": 0.8654165267944336, "num_tokens": 799374792.0, "step": 21048 }, { "epoch": 2.6776491540516476, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 7.028986930847168, "learning_rate": 1e-05, "loss": 0.4812, "mean_token_accuracy": 0.8786671161651611, "num_tokens": 799408425.0, "step": 21049 }, { "epoch": 2.6777763643302377, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.93410062789917, "learning_rate": 1e-05, "loss": 0.4502, "mean_token_accuracy": 0.8824958801269531, "num_tokens": 799448362.0, "step": 21050 }, { "epoch": 2.6779035746088287, "ewc_loss": 0.1455078125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.0001163482666015625, "grad_norm": 7.3623857498168945, "learning_rate": 1e-05, "loss": 0.4717, "mean_token_accuracy": 0.8831392526626587, "num_tokens": 799482225.0, "step": 21051 }, { "epoch": 2.6780307848874187, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 5.905880451202393, "learning_rate": 1e-05, "loss": 0.5063, "mean_token_accuracy": 0.8636597394943237, "num_tokens": 799523053.0, "step": 21052 }, { "epoch": 2.6781579951660093, "ewc_loss": 0.1552734375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.000125885009765625, "grad_norm": 7.7194061279296875, "learning_rate": 1e-05, "loss": 0.5045, "mean_token_accuracy": 0.8784910440444946, "num_tokens": 799563070.0, "step": 21053 }, { "epoch": 2.6782852054446, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.216050148010254, "learning_rate": 1e-05, "loss": 0.4498, "mean_token_accuracy": 0.8858762979507446, "num_tokens": 799595465.0, "step": 21054 }, { "epoch": 2.6784124157231903, "ewc_loss": 0.15625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00012683868408203125, "grad_norm": 7.776448726654053, "learning_rate": 1e-05, "loss": 0.4804, "mean_token_accuracy": 0.8867922425270081, "num_tokens": 799632308.0, "step": 21055 }, { "epoch": 2.678539626001781, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 6.575790882110596, "learning_rate": 1e-05, "loss": 0.4493, "mean_token_accuracy": 0.8914339542388916, "num_tokens": 799669563.0, "step": 21056 }, { "epoch": 2.6786668362803714, "ewc_loss": 0.150390625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00012159347534179688, "grad_norm": 7.513616561889648, "learning_rate": 1e-05, "loss": 0.5083, "mean_token_accuracy": 0.8747923374176025, "num_tokens": 799714575.0, "step": 21057 }, { "epoch": 2.678794046558962, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00011014938354492188, "grad_norm": 6.700881481170654, "learning_rate": 1e-05, "loss": 0.489, "mean_token_accuracy": 0.8806540966033936, "num_tokens": 799752730.0, "step": 21058 }, { "epoch": 2.6789212568375524, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00011444091796875, "grad_norm": 7.167788505554199, "learning_rate": 1e-05, "loss": 0.529, "mean_token_accuracy": 0.8669846653938293, "num_tokens": 799788110.0, "step": 21059 }, { "epoch": 2.679048467116143, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.637397289276123, "learning_rate": 1e-05, "loss": 0.4579, "mean_token_accuracy": 0.8855277299880981, "num_tokens": 799825561.0, "step": 21060 }, { "epoch": 2.6791756773947335, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.875808238983154, "learning_rate": 1e-05, "loss": 0.4575, "mean_token_accuracy": 0.884902834892273, "num_tokens": 799865412.0, "step": 21061 }, { "epoch": 2.679302887673324, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.529342174530029, "learning_rate": 1e-05, "loss": 0.5038, "mean_token_accuracy": 0.866429328918457, "num_tokens": 799900238.0, "step": 21062 }, { "epoch": 2.6794300979519146, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.711720943450928, "learning_rate": 1e-05, "loss": 0.4957, "mean_token_accuracy": 0.8718641400337219, "num_tokens": 799936724.0, "step": 21063 }, { "epoch": 2.679557308230505, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.38424825668335, "learning_rate": 1e-05, "loss": 0.4924, "mean_token_accuracy": 0.8731409311294556, "num_tokens": 799978347.0, "step": 21064 }, { "epoch": 2.6796845185090956, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.674042701721191, "learning_rate": 1e-05, "loss": 0.5064, "mean_token_accuracy": 0.8723032474517822, "num_tokens": 800015419.0, "step": 21065 }, { "epoch": 2.679811728787686, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.218753337860107, "learning_rate": 1e-05, "loss": 0.415, "mean_token_accuracy": 0.8981322050094604, "num_tokens": 800048435.0, "step": 21066 }, { "epoch": 2.6799389390662767, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.781095504760742, "learning_rate": 1e-05, "loss": 0.4756, "mean_token_accuracy": 0.8781888484954834, "num_tokens": 800083317.0, "step": 21067 }, { "epoch": 2.680066149344867, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.052975654602051, "learning_rate": 1e-05, "loss": 0.4784, "mean_token_accuracy": 0.8747416138648987, "num_tokens": 800123303.0, "step": 21068 }, { "epoch": 2.6801933596234577, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 6.989391326904297, "learning_rate": 1e-05, "loss": 0.4829, "mean_token_accuracy": 0.8802818059921265, "num_tokens": 800165157.0, "step": 21069 }, { "epoch": 2.6803205699020483, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.93340539932251, "learning_rate": 1e-05, "loss": 0.4908, "mean_token_accuracy": 0.8710787892341614, "num_tokens": 800203657.0, "step": 21070 }, { "epoch": 2.6804477801806383, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00011539459228515625, "grad_norm": 7.303635597229004, "learning_rate": 1e-05, "loss": 0.4753, "mean_token_accuracy": 0.8840388059616089, "num_tokens": 800245008.0, "step": 21071 }, { "epoch": 2.6805749904592293, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 5.9437174797058105, "learning_rate": 1e-05, "loss": 0.4659, "mean_token_accuracy": 0.8785648345947266, "num_tokens": 800285438.0, "step": 21072 }, { "epoch": 2.6807022007378194, "ewc_loss": 0.1513671875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.0001220703125, "grad_norm": 7.615082263946533, "learning_rate": 1e-05, "loss": 0.5143, "mean_token_accuracy": 0.8723558187484741, "num_tokens": 800321345.0, "step": 21073 }, { "epoch": 2.6808294110164104, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.22613525390625, "learning_rate": 1e-05, "loss": 0.4979, "mean_token_accuracy": 0.871289849281311, "num_tokens": 800361657.0, "step": 21074 }, { "epoch": 2.6809566212950005, "ewc_loss": 0.15234375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.634009838104248, "learning_rate": 1e-05, "loss": 0.5594, "mean_token_accuracy": 0.8569943904876709, "num_tokens": 800393809.0, "step": 21075 }, { "epoch": 2.681083831573591, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.578334808349609, "learning_rate": 1e-05, "loss": 0.4798, "mean_token_accuracy": 0.8791345357894897, "num_tokens": 800432541.0, "step": 21076 }, { "epoch": 2.6812110418521815, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00011491775512695312, "grad_norm": 7.274026393890381, "learning_rate": 1e-05, "loss": 0.498, "mean_token_accuracy": 0.873881459236145, "num_tokens": 800468552.0, "step": 21077 }, { "epoch": 2.681338252130772, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 6.728128433227539, "learning_rate": 1e-05, "loss": 0.4797, "mean_token_accuracy": 0.8837023377418518, "num_tokens": 800503911.0, "step": 21078 }, { "epoch": 2.6814654624093626, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.741888523101807, "learning_rate": 1e-05, "loss": 0.4846, "mean_token_accuracy": 0.8763105273246765, "num_tokens": 800549755.0, "step": 21079 }, { "epoch": 2.681592672687953, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 6.821972370147705, "learning_rate": 1e-05, "loss": 0.46, "mean_token_accuracy": 0.8855205774307251, "num_tokens": 800583240.0, "step": 21080 }, { "epoch": 2.6817198829665436, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.2427520751953125, "learning_rate": 1e-05, "loss": 0.5306, "mean_token_accuracy": 0.8608143329620361, "num_tokens": 800624016.0, "step": 21081 }, { "epoch": 2.681847093245134, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010967254638671875, "grad_norm": 7.037843227386475, "learning_rate": 1e-05, "loss": 0.5394, "mean_token_accuracy": 0.8595951795578003, "num_tokens": 800661052.0, "step": 21082 }, { "epoch": 2.6819743035237247, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.923417091369629, "learning_rate": 1e-05, "loss": 0.4622, "mean_token_accuracy": 0.8789435029029846, "num_tokens": 800701549.0, "step": 21083 }, { "epoch": 2.682101513802315, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00011777877807617188, "grad_norm": 7.401001453399658, "learning_rate": 1e-05, "loss": 0.4955, "mean_token_accuracy": 0.8755616545677185, "num_tokens": 800738650.0, "step": 21084 }, { "epoch": 2.6822287240809057, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 5.885895729064941, "learning_rate": 1e-05, "loss": 0.4555, "mean_token_accuracy": 0.880335807800293, "num_tokens": 800773687.0, "step": 21085 }, { "epoch": 2.6823559343594963, "ewc_loss": 0.158203125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00012874603271484375, "grad_norm": 7.782317161560059, "learning_rate": 1e-05, "loss": 0.4914, "mean_token_accuracy": 0.8862637281417847, "num_tokens": 800812523.0, "step": 21086 }, { "epoch": 2.682483144638087, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.2391180992126465, "learning_rate": 1e-05, "loss": 0.4846, "mean_token_accuracy": 0.8793017268180847, "num_tokens": 800850838.0, "step": 21087 }, { "epoch": 2.6826103549166773, "ewc_loss": 0.158203125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00012874603271484375, "grad_norm": 7.848324298858643, "learning_rate": 1e-05, "loss": 0.4731, "mean_token_accuracy": 0.8872859477996826, "num_tokens": 800886734.0, "step": 21088 }, { "epoch": 2.682737565195268, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 6.648016452789307, "learning_rate": 1e-05, "loss": 0.5018, "mean_token_accuracy": 0.8731740117073059, "num_tokens": 800930816.0, "step": 21089 }, { "epoch": 2.6828647754738584, "ewc_loss": 0.150390625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00012111663818359375, "grad_norm": 7.544007778167725, "learning_rate": 1e-05, "loss": 0.5105, "mean_token_accuracy": 0.8756194114685059, "num_tokens": 800962982.0, "step": 21090 }, { "epoch": 2.682991985752449, "ewc_loss": 0.140625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00011110305786132812, "grad_norm": 6.7667694091796875, "learning_rate": 1e-05, "loss": 0.5287, "mean_token_accuracy": 0.864865779876709, "num_tokens": 801003608.0, "step": 21091 }, { "epoch": 2.6831191960310394, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00011348724365234375, "grad_norm": 7.160373210906982, "learning_rate": 1e-05, "loss": 0.5107, "mean_token_accuracy": 0.8724452257156372, "num_tokens": 801032889.0, "step": 21092 }, { "epoch": 2.68324640630963, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.681965351104736, "learning_rate": 1e-05, "loss": 0.5147, "mean_token_accuracy": 0.8666815161705017, "num_tokens": 801068717.0, "step": 21093 }, { "epoch": 2.6833736165882205, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.865091800689697, "learning_rate": 1e-05, "loss": 0.4919, "mean_token_accuracy": 0.8773237466812134, "num_tokens": 801105120.0, "step": 21094 }, { "epoch": 2.683500826866811, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.542718410491943, "learning_rate": 1e-05, "loss": 0.4975, "mean_token_accuracy": 0.8711303472518921, "num_tokens": 801140301.0, "step": 21095 }, { "epoch": 2.683628037145401, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.716867923736572, "learning_rate": 1e-05, "loss": 0.4213, "mean_token_accuracy": 0.8978010416030884, "num_tokens": 801176314.0, "step": 21096 }, { "epoch": 2.683755247423992, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.3662943840026855, "learning_rate": 1e-05, "loss": 0.4743, "mean_token_accuracy": 0.8821197748184204, "num_tokens": 801212906.0, "step": 21097 }, { "epoch": 2.683882457702582, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.741425037384033, "learning_rate": 1e-05, "loss": 0.5172, "mean_token_accuracy": 0.8673330545425415, "num_tokens": 801243084.0, "step": 21098 }, { "epoch": 2.684009667981173, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.182048320770264, "learning_rate": 1e-05, "loss": 0.4848, "mean_token_accuracy": 0.8745871782302856, "num_tokens": 801282701.0, "step": 21099 }, { "epoch": 2.684136878259763, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.874095439910889, "learning_rate": 1e-05, "loss": 0.4675, "mean_token_accuracy": 0.8853954672813416, "num_tokens": 801315363.0, "step": 21100 }, { "epoch": 2.6842640885383537, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 6.011218070983887, "learning_rate": 1e-05, "loss": 0.4895, "mean_token_accuracy": 0.8720887303352356, "num_tokens": 801354103.0, "step": 21101 }, { "epoch": 2.6843912988169443, "ewc_loss": 0.140625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00011205673217773438, "grad_norm": 7.131938457489014, "learning_rate": 1e-05, "loss": 0.496, "mean_token_accuracy": 0.8763045072555542, "num_tokens": 801393351.0, "step": 21102 }, { "epoch": 2.684518509095535, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 5.920529365539551, "learning_rate": 1e-05, "loss": 0.4696, "mean_token_accuracy": 0.8723815679550171, "num_tokens": 801432829.0, "step": 21103 }, { "epoch": 2.6846457193741253, "ewc_loss": 0.1484375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00011920928955078125, "grad_norm": 7.485720157623291, "learning_rate": 1e-05, "loss": 0.5147, "mean_token_accuracy": 0.8734945058822632, "num_tokens": 801468735.0, "step": 21104 }, { "epoch": 2.684772929652716, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.056394577026367, "learning_rate": 1e-05, "loss": 0.4405, "mean_token_accuracy": 0.8882709741592407, "num_tokens": 801507238.0, "step": 21105 }, { "epoch": 2.6849001399313064, "ewc_loss": 0.1552734375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.000125885009765625, "grad_norm": 7.688491344451904, "learning_rate": 1e-05, "loss": 0.4919, "mean_token_accuracy": 0.8839854001998901, "num_tokens": 801543040.0, "step": 21106 }, { "epoch": 2.685027350209897, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.421354293823242, "learning_rate": 1e-05, "loss": 0.4785, "mean_token_accuracy": 0.877228319644928, "num_tokens": 801584466.0, "step": 21107 }, { "epoch": 2.6851545604884874, "ewc_loss": 0.1494140625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.0001201629638671875, "grad_norm": 7.507523536682129, "learning_rate": 1e-05, "loss": 0.4774, "mean_token_accuracy": 0.8848979473114014, "num_tokens": 801617466.0, "step": 21108 }, { "epoch": 2.685281770767078, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010967254638671875, "grad_norm": 6.69993257522583, "learning_rate": 1e-05, "loss": 0.5252, "mean_token_accuracy": 0.8671223521232605, "num_tokens": 801653408.0, "step": 21109 }, { "epoch": 2.6854089810456685, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010967254638671875, "grad_norm": 7.017438888549805, "learning_rate": 1e-05, "loss": 0.4942, "mean_token_accuracy": 0.8765625953674316, "num_tokens": 801687200.0, "step": 21110 }, { "epoch": 2.685536191324259, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 6.781966209411621, "learning_rate": 1e-05, "loss": 0.5, "mean_token_accuracy": 0.8743486404418945, "num_tokens": 801728547.0, "step": 21111 }, { "epoch": 2.6856634016028496, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.494801044464111, "learning_rate": 1e-05, "loss": 0.505, "mean_token_accuracy": 0.8699108362197876, "num_tokens": 801765425.0, "step": 21112 }, { "epoch": 2.68579061188144, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 6.895756244659424, "learning_rate": 1e-05, "loss": 0.4716, "mean_token_accuracy": 0.8798296451568604, "num_tokens": 801800694.0, "step": 21113 }, { "epoch": 2.6859178221600306, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 6.061683177947998, "learning_rate": 1e-05, "loss": 0.4293, "mean_token_accuracy": 0.8890456557273865, "num_tokens": 801835748.0, "step": 21114 }, { "epoch": 2.686045032438621, "ewc_loss": 0.140625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00011205673217773438, "grad_norm": 7.176555156707764, "learning_rate": 1e-05, "loss": 0.5049, "mean_token_accuracy": 0.8717144727706909, "num_tokens": 801873527.0, "step": 21115 }, { "epoch": 2.6861722427172117, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.8535027503967285, "learning_rate": 1e-05, "loss": 0.4474, "mean_token_accuracy": 0.885294497013092, "num_tokens": 801910630.0, "step": 21116 }, { "epoch": 2.686299452995802, "ewc_loss": 0.15234375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.5860209465026855, "learning_rate": 1e-05, "loss": 0.4918, "mean_token_accuracy": 0.8813257217407227, "num_tokens": 801946136.0, "step": 21117 }, { "epoch": 2.6864266632743927, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 5.995795249938965, "learning_rate": 1e-05, "loss": 0.4788, "mean_token_accuracy": 0.8732013702392578, "num_tokens": 801982283.0, "step": 21118 }, { "epoch": 2.686553873552983, "ewc_loss": 0.158203125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00012874603271484375, "grad_norm": 7.902734756469727, "learning_rate": 1e-05, "loss": 0.5217, "mean_token_accuracy": 0.8732243776321411, "num_tokens": 802022263.0, "step": 21119 }, { "epoch": 2.686681083831574, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.473158359527588, "learning_rate": 1e-05, "loss": 0.5983, "mean_token_accuracy": 0.8377017378807068, "num_tokens": 802061837.0, "step": 21120 }, { "epoch": 2.686808294110164, "ewc_loss": 0.15625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00012683868408203125, "grad_norm": 7.744287490844727, "learning_rate": 1e-05, "loss": 0.4812, "mean_token_accuracy": 0.8850836753845215, "num_tokens": 802094528.0, "step": 21121 }, { "epoch": 2.686935504388755, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00011396408081054688, "grad_norm": 6.770744323730469, "learning_rate": 1e-05, "loss": 0.5067, "mean_token_accuracy": 0.8720080256462097, "num_tokens": 802135157.0, "step": 21122 }, { "epoch": 2.687062714667345, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00011730194091796875, "grad_norm": 7.337031841278076, "learning_rate": 1e-05, "loss": 0.4564, "mean_token_accuracy": 0.8875226378440857, "num_tokens": 802168818.0, "step": 21123 }, { "epoch": 2.687189924945936, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010967254638671875, "grad_norm": 6.756018161773682, "learning_rate": 1e-05, "loss": 0.4975, "mean_token_accuracy": 0.8753762245178223, "num_tokens": 802209486.0, "step": 21124 }, { "epoch": 2.687317135224526, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 6.968626499176025, "learning_rate": 1e-05, "loss": 0.5217, "mean_token_accuracy": 0.8648092746734619, "num_tokens": 802248534.0, "step": 21125 }, { "epoch": 2.6874443455031165, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.633683681488037, "learning_rate": 1e-05, "loss": 0.4835, "mean_token_accuracy": 0.8770078420639038, "num_tokens": 802289330.0, "step": 21126 }, { "epoch": 2.687571555781707, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.734038352966309, "learning_rate": 1e-05, "loss": 0.484, "mean_token_accuracy": 0.8773151636123657, "num_tokens": 802322065.0, "step": 21127 }, { "epoch": 2.6876987660602976, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.4904561042785645, "learning_rate": 1e-05, "loss": 0.4863, "mean_token_accuracy": 0.8741597533226013, "num_tokens": 802361820.0, "step": 21128 }, { "epoch": 2.687825976338888, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.6401214599609375, "learning_rate": 1e-05, "loss": 0.4704, "mean_token_accuracy": 0.878551721572876, "num_tokens": 802397748.0, "step": 21129 }, { "epoch": 2.6879531866174786, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.328882217407227, "learning_rate": 1e-05, "loss": 0.5109, "mean_token_accuracy": 0.863105833530426, "num_tokens": 802432152.0, "step": 21130 }, { "epoch": 2.688080396896069, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.6660475730896, "learning_rate": 1e-05, "loss": 0.4598, "mean_token_accuracy": 0.8817987442016602, "num_tokens": 802471464.0, "step": 21131 }, { "epoch": 2.6882076071746597, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.162103176116943, "learning_rate": 1e-05, "loss": 0.4677, "mean_token_accuracy": 0.8770339488983154, "num_tokens": 802503096.0, "step": 21132 }, { "epoch": 2.68833481745325, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.8206377029418945, "learning_rate": 1e-05, "loss": 0.5282, "mean_token_accuracy": 0.8672688603401184, "num_tokens": 802540036.0, "step": 21133 }, { "epoch": 2.6884620277318407, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 5.980861663818359, "learning_rate": 1e-05, "loss": 0.4753, "mean_token_accuracy": 0.8754953145980835, "num_tokens": 802580798.0, "step": 21134 }, { "epoch": 2.6885892380104313, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010967254638671875, "grad_norm": 7.08802604675293, "learning_rate": 1e-05, "loss": 0.5002, "mean_token_accuracy": 0.8715792894363403, "num_tokens": 802619472.0, "step": 21135 }, { "epoch": 2.688716448289022, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.8621721267700195, "learning_rate": 1e-05, "loss": 0.4637, "mean_token_accuracy": 0.878643274307251, "num_tokens": 802656834.0, "step": 21136 }, { "epoch": 2.6888436585676123, "ewc_loss": 0.1494140625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.0001201629638671875, "grad_norm": 7.491395950317383, "learning_rate": 1e-05, "loss": 0.4336, "mean_token_accuracy": 0.8977357745170593, "num_tokens": 802690269.0, "step": 21137 }, { "epoch": 2.688970868846203, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 5.972157001495361, "learning_rate": 1e-05, "loss": 0.441, "mean_token_accuracy": 0.8865888118743896, "num_tokens": 802726822.0, "step": 21138 }, { "epoch": 2.6890980791247934, "ewc_loss": 0.158203125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00012874603271484375, "grad_norm": 7.823156833648682, "learning_rate": 1e-05, "loss": 0.4837, "mean_token_accuracy": 0.8864372968673706, "num_tokens": 802764479.0, "step": 21139 }, { "epoch": 2.689225289403384, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.392185688018799, "learning_rate": 1e-05, "loss": 0.4684, "mean_token_accuracy": 0.8807123899459839, "num_tokens": 802800938.0, "step": 21140 }, { "epoch": 2.6893524996819744, "ewc_loss": 0.1552734375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.000125885009765625, "grad_norm": 7.71109676361084, "learning_rate": 1e-05, "loss": 0.491, "mean_token_accuracy": 0.8804546594619751, "num_tokens": 802841593.0, "step": 21141 }, { "epoch": 2.689479709960565, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00011539459228515625, "grad_norm": 6.764153957366943, "learning_rate": 1e-05, "loss": 0.5041, "mean_token_accuracy": 0.8742291331291199, "num_tokens": 802875847.0, "step": 21142 }, { "epoch": 2.6896069202391555, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00011396408081054688, "grad_norm": 7.232922554016113, "learning_rate": 1e-05, "loss": 0.4775, "mean_token_accuracy": 0.8823215961456299, "num_tokens": 802915786.0, "step": 21143 }, { "epoch": 2.6897341305177456, "ewc_loss": 0.140625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00011205673217773438, "grad_norm": 6.824310302734375, "learning_rate": 1e-05, "loss": 0.4995, "mean_token_accuracy": 0.8733074069023132, "num_tokens": 802958774.0, "step": 21144 }, { "epoch": 2.6898613407963365, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.7155585289001465, "learning_rate": 1e-05, "loss": 0.4438, "mean_token_accuracy": 0.8910977840423584, "num_tokens": 802991360.0, "step": 21145 }, { "epoch": 2.6899885510749266, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 6.8502020835876465, "learning_rate": 1e-05, "loss": 0.4691, "mean_token_accuracy": 0.8841302394866943, "num_tokens": 803030852.0, "step": 21146 }, { "epoch": 2.6901157613535176, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.284399032592773, "learning_rate": 1e-05, "loss": 0.4597, "mean_token_accuracy": 0.8814849853515625, "num_tokens": 803068965.0, "step": 21147 }, { "epoch": 2.6902429716321077, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 6.987644672393799, "learning_rate": 1e-05, "loss": 0.5068, "mean_token_accuracy": 0.8708821535110474, "num_tokens": 803103331.0, "step": 21148 }, { "epoch": 2.6903701819106987, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.023179531097412, "learning_rate": 1e-05, "loss": 0.4603, "mean_token_accuracy": 0.8810069561004639, "num_tokens": 803144242.0, "step": 21149 }, { "epoch": 2.6904973921892887, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00011301040649414062, "grad_norm": 7.2399749755859375, "learning_rate": 1e-05, "loss": 0.5406, "mean_token_accuracy": 0.8612322807312012, "num_tokens": 803178355.0, "step": 21150 }, { "epoch": 2.6906246024678793, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 5.943352222442627, "learning_rate": 1e-05, "loss": 0.4857, "mean_token_accuracy": 0.8739618062973022, "num_tokens": 803218783.0, "step": 21151 }, { "epoch": 2.69075181274647, "ewc_loss": 0.150390625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00012159347534179688, "grad_norm": 7.5507659912109375, "learning_rate": 1e-05, "loss": 0.5749, "mean_token_accuracy": 0.8573114275932312, "num_tokens": 803256391.0, "step": 21152 }, { "epoch": 2.6908790230250603, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.11387825012207, "learning_rate": 1e-05, "loss": 0.4787, "mean_token_accuracy": 0.8761540651321411, "num_tokens": 803285708.0, "step": 21153 }, { "epoch": 2.691006233303651, "ewc_loss": 0.154296875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00012493133544921875, "grad_norm": 7.706535339355469, "learning_rate": 1e-05, "loss": 0.511, "mean_token_accuracy": 0.878830075263977, "num_tokens": 803319711.0, "step": 21154 }, { "epoch": 2.6911334435822414, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.4196882247924805, "learning_rate": 1e-05, "loss": 0.5239, "mean_token_accuracy": 0.8627023696899414, "num_tokens": 803353916.0, "step": 21155 }, { "epoch": 2.691260653860832, "ewc_loss": 0.150390625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00012159347534179688, "grad_norm": 7.572943210601807, "learning_rate": 1e-05, "loss": 0.466, "mean_token_accuracy": 0.8879417181015015, "num_tokens": 803388579.0, "step": 21156 }, { "epoch": 2.6913878641394224, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 6.597047328948975, "learning_rate": 1e-05, "loss": 0.4525, "mean_token_accuracy": 0.8875671625137329, "num_tokens": 803427715.0, "step": 21157 }, { "epoch": 2.691515074418013, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00011587142944335938, "grad_norm": 7.263857841491699, "learning_rate": 1e-05, "loss": 0.5295, "mean_token_accuracy": 0.8661741018295288, "num_tokens": 803470363.0, "step": 21158 }, { "epoch": 2.6916422846966035, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.633334159851074, "learning_rate": 1e-05, "loss": 0.4977, "mean_token_accuracy": 0.8731670379638672, "num_tokens": 803515009.0, "step": 21159 }, { "epoch": 2.691769494975194, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.946300506591797, "learning_rate": 1e-05, "loss": 0.4831, "mean_token_accuracy": 0.879538357257843, "num_tokens": 803554773.0, "step": 21160 }, { "epoch": 2.6918967052537845, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.568778038024902, "learning_rate": 1e-05, "loss": 0.455, "mean_token_accuracy": 0.8830526471138, "num_tokens": 803590404.0, "step": 21161 }, { "epoch": 2.692023915532375, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.726616859436035, "learning_rate": 1e-05, "loss": 0.4633, "mean_token_accuracy": 0.8870348930358887, "num_tokens": 803634388.0, "step": 21162 }, { "epoch": 2.6921511258109656, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.46900749206543, "learning_rate": 1e-05, "loss": 0.4678, "mean_token_accuracy": 0.8793689608573914, "num_tokens": 803673164.0, "step": 21163 }, { "epoch": 2.692278336089556, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.60575532913208, "learning_rate": 1e-05, "loss": 0.485, "mean_token_accuracy": 0.877183198928833, "num_tokens": 803715271.0, "step": 21164 }, { "epoch": 2.6924055463681467, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.359521389007568, "learning_rate": 1e-05, "loss": 0.4524, "mean_token_accuracy": 0.8848661184310913, "num_tokens": 803748295.0, "step": 21165 }, { "epoch": 2.692532756646737, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.580938816070557, "learning_rate": 1e-05, "loss": 0.5043, "mean_token_accuracy": 0.868276834487915, "num_tokens": 803787591.0, "step": 21166 }, { "epoch": 2.6926599669253277, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.226073265075684, "learning_rate": 1e-05, "loss": 0.4621, "mean_token_accuracy": 0.8792429566383362, "num_tokens": 803817640.0, "step": 21167 }, { "epoch": 2.6927871772039182, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.662771701812744, "learning_rate": 1e-05, "loss": 0.5074, "mean_token_accuracy": 0.8684358596801758, "num_tokens": 803847932.0, "step": 21168 }, { "epoch": 2.6929143874825083, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.063257694244385, "learning_rate": 1e-05, "loss": 0.486, "mean_token_accuracy": 0.8706790208816528, "num_tokens": 803884206.0, "step": 21169 }, { "epoch": 2.6930415977610993, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.8509955406188965, "learning_rate": 1e-05, "loss": 0.4755, "mean_token_accuracy": 0.8805389404296875, "num_tokens": 803923225.0, "step": 21170 }, { "epoch": 2.6931688080396894, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.915886878967285, "learning_rate": 1e-05, "loss": 0.5146, "mean_token_accuracy": 0.8599741458892822, "num_tokens": 803961880.0, "step": 21171 }, { "epoch": 2.6932960183182804, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.0001125335693359375, "grad_norm": 7.174376010894775, "learning_rate": 1e-05, "loss": 0.49, "mean_token_accuracy": 0.8803520798683167, "num_tokens": 804004676.0, "step": 21172 }, { "epoch": 2.6934232285968704, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.821799278259277, "learning_rate": 1e-05, "loss": 0.4523, "mean_token_accuracy": 0.8792990446090698, "num_tokens": 804047875.0, "step": 21173 }, { "epoch": 2.693550438875461, "ewc_loss": 0.15234375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.631086349487305, "learning_rate": 1e-05, "loss": 0.4403, "mean_token_accuracy": 0.895412802696228, "num_tokens": 804083467.0, "step": 21174 }, { "epoch": 2.6936776491540515, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.02217435836792, "learning_rate": 1e-05, "loss": 0.4483, "mean_token_accuracy": 0.8862996101379395, "num_tokens": 804124653.0, "step": 21175 }, { "epoch": 2.693804859432642, "ewc_loss": 0.16015625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00013065338134765625, "grad_norm": 7.909850120544434, "learning_rate": 1e-05, "loss": 0.4858, "mean_token_accuracy": 0.8842841982841492, "num_tokens": 804159081.0, "step": 21176 }, { "epoch": 2.6939320697112326, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 6.526806831359863, "learning_rate": 1e-05, "loss": 0.5044, "mean_token_accuracy": 0.8679322600364685, "num_tokens": 804200687.0, "step": 21177 }, { "epoch": 2.694059279989823, "ewc_loss": 0.1552734375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.000125885009765625, "grad_norm": 7.672520160675049, "learning_rate": 1e-05, "loss": 0.5203, "mean_token_accuracy": 0.872999370098114, "num_tokens": 804237192.0, "step": 21178 }, { "epoch": 2.6941864902684136, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00011301040649414062, "grad_norm": 6.820003986358643, "learning_rate": 1e-05, "loss": 0.4735, "mean_token_accuracy": 0.881523847579956, "num_tokens": 804275078.0, "step": 21179 }, { "epoch": 2.694313700547004, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00011301040649414062, "grad_norm": 7.1689558029174805, "learning_rate": 1e-05, "loss": 0.5203, "mean_token_accuracy": 0.8708428740501404, "num_tokens": 804308551.0, "step": 21180 }, { "epoch": 2.6944409108255947, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00011014938354492188, "grad_norm": 6.83421516418457, "learning_rate": 1e-05, "loss": 0.4895, "mean_token_accuracy": 0.8766774535179138, "num_tokens": 804347530.0, "step": 21181 }, { "epoch": 2.694568121104185, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.6790080070495605, "learning_rate": 1e-05, "loss": 0.5107, "mean_token_accuracy": 0.871300458908081, "num_tokens": 804388663.0, "step": 21182 }, { "epoch": 2.6946953313827757, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.85134744644165, "learning_rate": 1e-05, "loss": 0.4727, "mean_token_accuracy": 0.8833698034286499, "num_tokens": 804423956.0, "step": 21183 }, { "epoch": 2.6948225416613663, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.296882152557373, "learning_rate": 1e-05, "loss": 0.4908, "mean_token_accuracy": 0.8735166192054749, "num_tokens": 804466736.0, "step": 21184 }, { "epoch": 2.694949751939957, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 6.9668145179748535, "learning_rate": 1e-05, "loss": 0.5801, "mean_token_accuracy": 0.8504208326339722, "num_tokens": 804505101.0, "step": 21185 }, { "epoch": 2.6950769622185473, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.062598705291748, "learning_rate": 1e-05, "loss": 0.4465, "mean_token_accuracy": 0.8843994140625, "num_tokens": 804542520.0, "step": 21186 }, { "epoch": 2.695204172497138, "ewc_loss": 0.140625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00011157989501953125, "grad_norm": 7.1510725021362305, "learning_rate": 1e-05, "loss": 0.4694, "mean_token_accuracy": 0.8851250410079956, "num_tokens": 804577749.0, "step": 21187 }, { "epoch": 2.6953313827757284, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 5.977971076965332, "learning_rate": 1e-05, "loss": 0.4172, "mean_token_accuracy": 0.8934453725814819, "num_tokens": 804611370.0, "step": 21188 }, { "epoch": 2.695458593054319, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00011539459228515625, "grad_norm": 7.4079766273498535, "learning_rate": 1e-05, "loss": 0.4553, "mean_token_accuracy": 0.8882936835289001, "num_tokens": 804643696.0, "step": 21189 }, { "epoch": 2.6955858033329094, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.049623966217041, "learning_rate": 1e-05, "loss": 0.4247, "mean_token_accuracy": 0.8950368165969849, "num_tokens": 804678343.0, "step": 21190 }, { "epoch": 2.6957130136115, "ewc_loss": 0.1513671875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.0001220703125, "grad_norm": 7.592240810394287, "learning_rate": 1e-05, "loss": 0.4561, "mean_token_accuracy": 0.8905407190322876, "num_tokens": 804711303.0, "step": 21191 }, { "epoch": 2.6958402238900905, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.269556999206543, "learning_rate": 1e-05, "loss": 0.5464, "mean_token_accuracy": 0.8539174795150757, "num_tokens": 804749016.0, "step": 21192 }, { "epoch": 2.695967434168681, "ewc_loss": 0.1513671875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.0001220703125, "grad_norm": 7.5819268226623535, "learning_rate": 1e-05, "loss": 0.5186, "mean_token_accuracy": 0.8693900108337402, "num_tokens": 804789600.0, "step": 21193 }, { "epoch": 2.696094644447271, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.479647159576416, "learning_rate": 1e-05, "loss": 0.4717, "mean_token_accuracy": 0.8827533721923828, "num_tokens": 804829033.0, "step": 21194 }, { "epoch": 2.696221854725862, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00011730194091796875, "grad_norm": 7.3694281578063965, "learning_rate": 1e-05, "loss": 0.5369, "mean_token_accuracy": 0.8670045733451843, "num_tokens": 804865114.0, "step": 21195 }, { "epoch": 2.696349065004452, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.572465419769287, "learning_rate": 1e-05, "loss": 0.5387, "mean_token_accuracy": 0.8618028163909912, "num_tokens": 804907604.0, "step": 21196 }, { "epoch": 2.696476275283043, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00011014938354492188, "grad_norm": 7.072290897369385, "learning_rate": 1e-05, "loss": 0.5255, "mean_token_accuracy": 0.8697187900543213, "num_tokens": 804942361.0, "step": 21197 }, { "epoch": 2.696603485561633, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.557947635650635, "learning_rate": 1e-05, "loss": 0.4869, "mean_token_accuracy": 0.8777487277984619, "num_tokens": 804980452.0, "step": 21198 }, { "epoch": 2.6967306958402237, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.791895389556885, "learning_rate": 1e-05, "loss": 0.4586, "mean_token_accuracy": 0.8832894563674927, "num_tokens": 805017969.0, "step": 21199 }, { "epoch": 2.6968579061188143, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.49974250793457, "learning_rate": 1e-05, "loss": 0.4616, "mean_token_accuracy": 0.8841139078140259, "num_tokens": 805057053.0, "step": 21200 }, { "epoch": 2.696985116397405, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.609704971313477, "learning_rate": 1e-05, "loss": 0.4775, "mean_token_accuracy": 0.8804014921188354, "num_tokens": 805098928.0, "step": 21201 }, { "epoch": 2.6971123266759953, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.406571388244629, "learning_rate": 1e-05, "loss": 0.4711, "mean_token_accuracy": 0.8790658712387085, "num_tokens": 805140729.0, "step": 21202 }, { "epoch": 2.697239536954586, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.5513834953308105, "learning_rate": 1e-05, "loss": 0.4699, "mean_token_accuracy": 0.8817052841186523, "num_tokens": 805177150.0, "step": 21203 }, { "epoch": 2.6973667472331764, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.297982215881348, "learning_rate": 1e-05, "loss": 0.4693, "mean_token_accuracy": 0.8773361444473267, "num_tokens": 805221233.0, "step": 21204 }, { "epoch": 2.697493957511767, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.560614585876465, "learning_rate": 1e-05, "loss": 0.4452, "mean_token_accuracy": 0.8896948099136353, "num_tokens": 805255075.0, "step": 21205 }, { "epoch": 2.6976211677903574, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.175529956817627, "learning_rate": 1e-05, "loss": 0.4927, "mean_token_accuracy": 0.8717020153999329, "num_tokens": 805292468.0, "step": 21206 }, { "epoch": 2.697748378068948, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.6526007652282715, "learning_rate": 1e-05, "loss": 0.4659, "mean_token_accuracy": 0.879296064376831, "num_tokens": 805330648.0, "step": 21207 }, { "epoch": 2.6978755883475385, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.038503646850586, "learning_rate": 1e-05, "loss": 0.4623, "mean_token_accuracy": 0.8791583776473999, "num_tokens": 805368137.0, "step": 21208 }, { "epoch": 2.698002798626129, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.850221633911133, "learning_rate": 1e-05, "loss": 0.4771, "mean_token_accuracy": 0.8789055943489075, "num_tokens": 805399898.0, "step": 21209 }, { "epoch": 2.6981300089047195, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.885156154632568, "learning_rate": 1e-05, "loss": 0.456, "mean_token_accuracy": 0.8843767642974854, "num_tokens": 805439292.0, "step": 21210 }, { "epoch": 2.69825721918331, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00011396408081054688, "grad_norm": 7.189724445343018, "learning_rate": 1e-05, "loss": 0.5389, "mean_token_accuracy": 0.8628265261650085, "num_tokens": 805479464.0, "step": 21211 }, { "epoch": 2.6983844294619006, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.801559925079346, "learning_rate": 1e-05, "loss": 0.3919, "mean_token_accuracy": 0.9012121558189392, "num_tokens": 805518844.0, "step": 21212 }, { "epoch": 2.698511639740491, "ewc_loss": 0.15234375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.647181510925293, "learning_rate": 1e-05, "loss": 0.468, "mean_token_accuracy": 0.8873054385185242, "num_tokens": 805552192.0, "step": 21213 }, { "epoch": 2.6986388500190817, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.004240989685059, "learning_rate": 1e-05, "loss": 0.4628, "mean_token_accuracy": 0.8818799257278442, "num_tokens": 805587343.0, "step": 21214 }, { "epoch": 2.698766060297672, "ewc_loss": 0.1611328125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.0001316070556640625, "grad_norm": 7.9497857093811035, "learning_rate": 1e-05, "loss": 0.5005, "mean_token_accuracy": 0.8794273138046265, "num_tokens": 805623256.0, "step": 21215 }, { "epoch": 2.6988932705762627, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 6.52601957321167, "learning_rate": 1e-05, "loss": 0.498, "mean_token_accuracy": 0.8765588998794556, "num_tokens": 805664627.0, "step": 21216 }, { "epoch": 2.699020480854853, "ewc_loss": 0.15625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00012683868408203125, "grad_norm": 7.729219436645508, "learning_rate": 1e-05, "loss": 0.5422, "mean_token_accuracy": 0.8679478764533997, "num_tokens": 805698808.0, "step": 21217 }, { "epoch": 2.6991476911334438, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00011539459228515625, "grad_norm": 6.831521987915039, "learning_rate": 1e-05, "loss": 0.5062, "mean_token_accuracy": 0.879305899143219, "num_tokens": 805738468.0, "step": 21218 }, { "epoch": 2.699274901412034, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00011539459228515625, "grad_norm": 7.235841751098633, "learning_rate": 1e-05, "loss": 0.5345, "mean_token_accuracy": 0.8637773990631104, "num_tokens": 805776233.0, "step": 21219 }, { "epoch": 2.699402111690625, "ewc_loss": 0.140625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00011205673217773438, "grad_norm": 6.82131290435791, "learning_rate": 1e-05, "loss": 0.4486, "mean_token_accuracy": 0.8941277265548706, "num_tokens": 805818459.0, "step": 21220 }, { "epoch": 2.699529321969215, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.7499918937683105, "learning_rate": 1e-05, "loss": 0.4573, "mean_token_accuracy": 0.8869747519493103, "num_tokens": 805858340.0, "step": 21221 }, { "epoch": 2.699656532247806, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 6.8136887550354, "learning_rate": 1e-05, "loss": 0.4998, "mean_token_accuracy": 0.8738787174224854, "num_tokens": 805895920.0, "step": 21222 }, { "epoch": 2.699783742526396, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.375144958496094, "learning_rate": 1e-05, "loss": 0.4655, "mean_token_accuracy": 0.8822811245918274, "num_tokens": 805937692.0, "step": 21223 }, { "epoch": 2.6999109528049865, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.8846306800842285, "learning_rate": 1e-05, "loss": 0.5064, "mean_token_accuracy": 0.8716002106666565, "num_tokens": 805974573.0, "step": 21224 }, { "epoch": 2.700038163083577, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.139162063598633, "learning_rate": 1e-05, "loss": 0.4634, "mean_token_accuracy": 0.8802609443664551, "num_tokens": 806013643.0, "step": 21225 }, { "epoch": 2.7001653733621676, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 6.999776363372803, "learning_rate": 1e-05, "loss": 0.484, "mean_token_accuracy": 0.8806562423706055, "num_tokens": 806048027.0, "step": 21226 }, { "epoch": 2.700292583640758, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 6.023736953735352, "learning_rate": 1e-05, "loss": 0.4952, "mean_token_accuracy": 0.8686254620552063, "num_tokens": 806087269.0, "step": 21227 }, { "epoch": 2.7004197939193486, "ewc_loss": 0.140625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00011205673217773438, "grad_norm": 7.171036243438721, "learning_rate": 1e-05, "loss": 0.5152, "mean_token_accuracy": 0.870448887348175, "num_tokens": 806126924.0, "step": 21228 }, { "epoch": 2.700547004197939, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 5.99875545501709, "learning_rate": 1e-05, "loss": 0.4755, "mean_token_accuracy": 0.8773527145385742, "num_tokens": 806162510.0, "step": 21229 }, { "epoch": 2.7006742144765297, "ewc_loss": 0.1455078125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.0001163482666015625, "grad_norm": 7.3743486404418945, "learning_rate": 1e-05, "loss": 0.4665, "mean_token_accuracy": 0.8856006860733032, "num_tokens": 806196519.0, "step": 21230 }, { "epoch": 2.70080142475512, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.079678058624268, "learning_rate": 1e-05, "loss": 0.5163, "mean_token_accuracy": 0.8641762137413025, "num_tokens": 806233995.0, "step": 21231 }, { "epoch": 2.7009286350337107, "ewc_loss": 0.1494140625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00012063980102539062, "grad_norm": 7.5187602043151855, "learning_rate": 1e-05, "loss": 0.4935, "mean_token_accuracy": 0.8796133399009705, "num_tokens": 806268660.0, "step": 21232 }, { "epoch": 2.7010558453123013, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.2469482421875, "learning_rate": 1e-05, "loss": 0.5115, "mean_token_accuracy": 0.8688102960586548, "num_tokens": 806304187.0, "step": 21233 }, { "epoch": 2.701183055590892, "ewc_loss": 0.1484375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00011968612670898438, "grad_norm": 7.495269298553467, "learning_rate": 1e-05, "loss": 0.5095, "mean_token_accuracy": 0.8751189112663269, "num_tokens": 806341524.0, "step": 21234 }, { "epoch": 2.7013102658694823, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.424021244049072, "learning_rate": 1e-05, "loss": 0.5184, "mean_token_accuracy": 0.866389811038971, "num_tokens": 806384983.0, "step": 21235 }, { "epoch": 2.701437476148073, "ewc_loss": 0.1455078125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.0001163482666015625, "grad_norm": 7.285749912261963, "learning_rate": 1e-05, "loss": 0.4919, "mean_token_accuracy": 0.8811179995536804, "num_tokens": 806429923.0, "step": 21236 }, { "epoch": 2.7015646864266634, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.516412734985352, "learning_rate": 1e-05, "loss": 0.4674, "mean_token_accuracy": 0.8853263854980469, "num_tokens": 806467562.0, "step": 21237 }, { "epoch": 2.701691896705254, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 7.003873348236084, "learning_rate": 1e-05, "loss": 0.501, "mean_token_accuracy": 0.8759927153587341, "num_tokens": 806508634.0, "step": 21238 }, { "epoch": 2.7018191069838444, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.518650531768799, "learning_rate": 1e-05, "loss": 0.5062, "mean_token_accuracy": 0.8687185049057007, "num_tokens": 806547972.0, "step": 21239 }, { "epoch": 2.701946317262435, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.7283501625061035, "learning_rate": 1e-05, "loss": 0.5059, "mean_token_accuracy": 0.8690147399902344, "num_tokens": 806591461.0, "step": 21240 }, { "epoch": 2.7020735275410255, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.4986572265625, "learning_rate": 1e-05, "loss": 0.4406, "mean_token_accuracy": 0.8916985392570496, "num_tokens": 806623276.0, "step": 21241 }, { "epoch": 2.7022007378196156, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.521782398223877, "learning_rate": 1e-05, "loss": 0.4616, "mean_token_accuracy": 0.8805015087127686, "num_tokens": 806657732.0, "step": 21242 }, { "epoch": 2.7023279480982065, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.466561317443848, "learning_rate": 1e-05, "loss": 0.4429, "mean_token_accuracy": 0.8857220411300659, "num_tokens": 806695018.0, "step": 21243 }, { "epoch": 2.7024551583767966, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.393570423126221, "learning_rate": 1e-05, "loss": 0.4204, "mean_token_accuracy": 0.8961853981018066, "num_tokens": 806735344.0, "step": 21244 }, { "epoch": 2.7025823686553876, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.417900085449219, "learning_rate": 1e-05, "loss": 0.4791, "mean_token_accuracy": 0.8759665489196777, "num_tokens": 806776854.0, "step": 21245 }, { "epoch": 2.7027095789339777, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.343661785125732, "learning_rate": 1e-05, "loss": 0.4565, "mean_token_accuracy": 0.8844670057296753, "num_tokens": 806811594.0, "step": 21246 }, { "epoch": 2.7028367892125686, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.379508972167969, "learning_rate": 1e-05, "loss": 0.5074, "mean_token_accuracy": 0.8706990480422974, "num_tokens": 806845249.0, "step": 21247 }, { "epoch": 2.7029639994911587, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.319401741027832, "learning_rate": 1e-05, "loss": 0.5032, "mean_token_accuracy": 0.8709632754325867, "num_tokens": 806885110.0, "step": 21248 }, { "epoch": 2.7030912097697493, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.3219895362854, "learning_rate": 1e-05, "loss": 0.4623, "mean_token_accuracy": 0.8774502277374268, "num_tokens": 806924809.0, "step": 21249 }, { "epoch": 2.70321842004834, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.311773777008057, "learning_rate": 1e-05, "loss": 0.4224, "mean_token_accuracy": 0.8938820362091064, "num_tokens": 806962386.0, "step": 21250 }, { "epoch": 2.7033456303269303, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.278800010681152, "learning_rate": 1e-05, "loss": 0.4749, "mean_token_accuracy": 0.8726814389228821, "num_tokens": 806998754.0, "step": 21251 }, { "epoch": 2.703472840605521, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.333797454833984, "learning_rate": 1e-05, "loss": 0.4684, "mean_token_accuracy": 0.8788644075393677, "num_tokens": 807036140.0, "step": 21252 }, { "epoch": 2.7036000508841114, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.234971523284912, "learning_rate": 1e-05, "loss": 0.4811, "mean_token_accuracy": 0.873767614364624, "num_tokens": 807077144.0, "step": 21253 }, { "epoch": 2.703727261162702, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.369609832763672, "learning_rate": 1e-05, "loss": 0.4777, "mean_token_accuracy": 0.875091552734375, "num_tokens": 807114928.0, "step": 21254 }, { "epoch": 2.7038544714412924, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.1742424964904785, "learning_rate": 1e-05, "loss": 0.4716, "mean_token_accuracy": 0.8815235495567322, "num_tokens": 807151067.0, "step": 21255 }, { "epoch": 2.703981681719883, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.435549736022949, "learning_rate": 1e-05, "loss": 0.5042, "mean_token_accuracy": 0.8680826425552368, "num_tokens": 807185992.0, "step": 21256 }, { "epoch": 2.7041088919984735, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.059863090515137, "learning_rate": 1e-05, "loss": 0.436, "mean_token_accuracy": 0.8871858716011047, "num_tokens": 807229122.0, "step": 21257 }, { "epoch": 2.704236102277064, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.605766773223877, "learning_rate": 1e-05, "loss": 0.4913, "mean_token_accuracy": 0.8717242479324341, "num_tokens": 807264118.0, "step": 21258 }, { "epoch": 2.7043633125556545, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 5.891218185424805, "learning_rate": 1e-05, "loss": 0.4585, "mean_token_accuracy": 0.8810052275657654, "num_tokens": 807298087.0, "step": 21259 }, { "epoch": 2.704490522834245, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.919190883636475, "learning_rate": 1e-05, "loss": 0.4886, "mean_token_accuracy": 0.8787066340446472, "num_tokens": 807338367.0, "step": 21260 }, { "epoch": 2.7046177331128356, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.658109664916992, "learning_rate": 1e-05, "loss": 0.4324, "mean_token_accuracy": 0.8873491883277893, "num_tokens": 807376858.0, "step": 21261 }, { "epoch": 2.704744943391426, "ewc_loss": 0.1494140625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.0001201629638671875, "grad_norm": 7.502668380737305, "learning_rate": 1e-05, "loss": 0.5303, "mean_token_accuracy": 0.8668358325958252, "num_tokens": 807410840.0, "step": 21262 }, { "epoch": 2.7048721536700167, "ewc_loss": 0.11376953125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.576615810394287, "learning_rate": 1e-05, "loss": 0.4133, "mean_token_accuracy": 0.8919952511787415, "num_tokens": 807445128.0, "step": 21263 }, { "epoch": 2.704999363948607, "ewc_loss": 0.173828125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00014495849609375, "grad_norm": 8.386035919189453, "learning_rate": 1e-05, "loss": 0.5429, "mean_token_accuracy": 0.8704617023468018, "num_tokens": 807478884.0, "step": 21264 }, { "epoch": 2.7051265742271977, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 6.294447898864746, "learning_rate": 1e-05, "loss": 0.4601, "mean_token_accuracy": 0.8865789771080017, "num_tokens": 807517020.0, "step": 21265 }, { "epoch": 2.7052537845057882, "ewc_loss": 0.181640625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.000152587890625, "grad_norm": 8.715973854064941, "learning_rate": 1e-05, "loss": 0.5777, "mean_token_accuracy": 0.864412784576416, "num_tokens": 807557037.0, "step": 21266 }, { "epoch": 2.7053809947843783, "ewc_loss": 0.1611328125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.0001316070556640625, "grad_norm": 7.388373374938965, "learning_rate": 1e-05, "loss": 0.4725, "mean_token_accuracy": 0.8904096484184265, "num_tokens": 807596876.0, "step": 21267 }, { "epoch": 2.7055082050629693, "ewc_loss": 0.162109375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00013256072998046875, "grad_norm": 8.013312339782715, "learning_rate": 1e-05, "loss": 0.5383, "mean_token_accuracy": 0.8690075874328613, "num_tokens": 807638107.0, "step": 21268 }, { "epoch": 2.7056354153415594, "ewc_loss": 0.1552734375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.000125885009765625, "grad_norm": 7.387149333953857, "learning_rate": 1e-05, "loss": 0.5414, "mean_token_accuracy": 0.8631919622421265, "num_tokens": 807678751.0, "step": 21269 }, { "epoch": 2.7057626256201504, "ewc_loss": 0.1513671875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.0001220703125, "grad_norm": 7.534140586853027, "learning_rate": 1e-05, "loss": 0.5706, "mean_token_accuracy": 0.8547119498252869, "num_tokens": 807711447.0, "step": 21270 }, { "epoch": 2.7058898358987404, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00011444091796875, "grad_norm": 7.0116400718688965, "learning_rate": 1e-05, "loss": 0.4727, "mean_token_accuracy": 0.8824423551559448, "num_tokens": 807747157.0, "step": 21271 }, { "epoch": 2.706017046177331, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00011396408081054688, "grad_norm": 7.254012584686279, "learning_rate": 1e-05, "loss": 0.5134, "mean_token_accuracy": 0.8697475790977478, "num_tokens": 807782695.0, "step": 21272 }, { "epoch": 2.7061442564559215, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.70393705368042, "learning_rate": 1e-05, "loss": 0.4919, "mean_token_accuracy": 0.8792929649353027, "num_tokens": 807816850.0, "step": 21273 }, { "epoch": 2.706271466734512, "ewc_loss": 0.140625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00011157989501953125, "grad_norm": 7.1322922706604, "learning_rate": 1e-05, "loss": 0.5268, "mean_token_accuracy": 0.8670542240142822, "num_tokens": 807856246.0, "step": 21274 }, { "epoch": 2.7063986770131025, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.454331874847412, "learning_rate": 1e-05, "loss": 0.4504, "mean_token_accuracy": 0.884270191192627, "num_tokens": 807891417.0, "step": 21275 }, { "epoch": 2.706525887291693, "ewc_loss": 0.140625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00011110305786132812, "grad_norm": 7.150195598602295, "learning_rate": 1e-05, "loss": 0.4633, "mean_token_accuracy": 0.8855389356613159, "num_tokens": 807924831.0, "step": 21276 }, { "epoch": 2.7066530975702836, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.249668121337891, "learning_rate": 1e-05, "loss": 0.4282, "mean_token_accuracy": 0.8936009407043457, "num_tokens": 807955427.0, "step": 21277 }, { "epoch": 2.706780307848874, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00011396408081054688, "grad_norm": 7.327043056488037, "learning_rate": 1e-05, "loss": 0.4951, "mean_token_accuracy": 0.8768707513809204, "num_tokens": 807990799.0, "step": 21278 }, { "epoch": 2.7069075181274647, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.14168119430542, "learning_rate": 1e-05, "loss": 0.5169, "mean_token_accuracy": 0.8647904992103577, "num_tokens": 808027028.0, "step": 21279 }, { "epoch": 2.707034728406055, "ewc_loss": 0.15234375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.63237190246582, "learning_rate": 1e-05, "loss": 0.4961, "mean_token_accuracy": 0.8769932389259338, "num_tokens": 808056750.0, "step": 21280 }, { "epoch": 2.7071619386846457, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.288573741912842, "learning_rate": 1e-05, "loss": 0.4875, "mean_token_accuracy": 0.8727105259895325, "num_tokens": 808096372.0, "step": 21281 }, { "epoch": 2.7072891489632362, "ewc_loss": 0.154296875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00012493133544921875, "grad_norm": 7.779372215270996, "learning_rate": 1e-05, "loss": 0.5015, "mean_token_accuracy": 0.8759620189666748, "num_tokens": 808136783.0, "step": 21282 }, { "epoch": 2.7074163592418268, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.652778148651123, "learning_rate": 1e-05, "loss": 0.4736, "mean_token_accuracy": 0.8765815496444702, "num_tokens": 808172310.0, "step": 21283 }, { "epoch": 2.7075435695204173, "ewc_loss": 0.1474609375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00011873245239257812, "grad_norm": 7.495110034942627, "learning_rate": 1e-05, "loss": 0.4809, "mean_token_accuracy": 0.883723258972168, "num_tokens": 808207745.0, "step": 21284 }, { "epoch": 2.707670779799008, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 6.8621368408203125, "learning_rate": 1e-05, "loss": 0.5056, "mean_token_accuracy": 0.8760839700698853, "num_tokens": 808243135.0, "step": 21285 }, { "epoch": 2.7077979900775984, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.963754653930664, "learning_rate": 1e-05, "loss": 0.4853, "mean_token_accuracy": 0.8762091398239136, "num_tokens": 808280601.0, "step": 21286 }, { "epoch": 2.707925200356189, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.895143985748291, "learning_rate": 1e-05, "loss": 0.4581, "mean_token_accuracy": 0.8834915161132812, "num_tokens": 808315541.0, "step": 21287 }, { "epoch": 2.7080524106347794, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.4237589836120605, "learning_rate": 1e-05, "loss": 0.4917, "mean_token_accuracy": 0.8713071942329407, "num_tokens": 808355670.0, "step": 21288 }, { "epoch": 2.70817962091337, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.973222732543945, "learning_rate": 1e-05, "loss": 0.4776, "mean_token_accuracy": 0.8832815289497375, "num_tokens": 808394189.0, "step": 21289 }, { "epoch": 2.7083068311919605, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 6.0125532150268555, "learning_rate": 1e-05, "loss": 0.433, "mean_token_accuracy": 0.8886457085609436, "num_tokens": 808432861.0, "step": 21290 }, { "epoch": 2.708434041470551, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00011301040649414062, "grad_norm": 7.250056743621826, "learning_rate": 1e-05, "loss": 0.5068, "mean_token_accuracy": 0.8735870122909546, "num_tokens": 808462423.0, "step": 21291 }, { "epoch": 2.708561251749141, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.81109094619751, "learning_rate": 1e-05, "loss": 0.4184, "mean_token_accuracy": 0.8929422497749329, "num_tokens": 808494112.0, "step": 21292 }, { "epoch": 2.708688462027732, "ewc_loss": 0.154296875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00012493133544921875, "grad_norm": 7.705102443695068, "learning_rate": 1e-05, "loss": 0.5011, "mean_token_accuracy": 0.8814396858215332, "num_tokens": 808538399.0, "step": 21293 }, { "epoch": 2.708815672306322, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 5.976522445678711, "learning_rate": 1e-05, "loss": 0.4682, "mean_token_accuracy": 0.8790252804756165, "num_tokens": 808575857.0, "step": 21294 }, { "epoch": 2.708942882584913, "ewc_loss": 0.1640625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00013446807861328125, "grad_norm": 8.104074478149414, "learning_rate": 1e-05, "loss": 0.5295, "mean_token_accuracy": 0.8733155727386475, "num_tokens": 808615529.0, "step": 21295 }, { "epoch": 2.709070092863503, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00011014938354492188, "grad_norm": 6.607314109802246, "learning_rate": 1e-05, "loss": 0.4553, "mean_token_accuracy": 0.8906961679458618, "num_tokens": 808653899.0, "step": 21296 }, { "epoch": 2.7091973031420937, "ewc_loss": 0.158203125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00012874603271484375, "grad_norm": 7.883411884307861, "learning_rate": 1e-05, "loss": 0.5343, "mean_token_accuracy": 0.8678156733512878, "num_tokens": 808697372.0, "step": 21297 }, { "epoch": 2.7093245134206843, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00011730194091796875, "grad_norm": 7.003042697906494, "learning_rate": 1e-05, "loss": 0.512, "mean_token_accuracy": 0.8744791150093079, "num_tokens": 808733854.0, "step": 21298 }, { "epoch": 2.709451723699275, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00011730194091796875, "grad_norm": 7.341315269470215, "learning_rate": 1e-05, "loss": 0.5048, "mean_token_accuracy": 0.8784292936325073, "num_tokens": 808777026.0, "step": 21299 }, { "epoch": 2.7095789339778653, "ewc_loss": 0.140625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00011157989501953125, "grad_norm": 6.93055534362793, "learning_rate": 1e-05, "loss": 0.4654, "mean_token_accuracy": 0.8897662162780762, "num_tokens": 808814142.0, "step": 21300 }, { "epoch": 2.709706144256456, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.925662994384766, "learning_rate": 1e-05, "loss": 0.5148, "mean_token_accuracy": 0.8705564141273499, "num_tokens": 808853355.0, "step": 21301 }, { "epoch": 2.7098333545350464, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.76242733001709, "learning_rate": 1e-05, "loss": 0.479, "mean_token_accuracy": 0.8777757883071899, "num_tokens": 808892782.0, "step": 21302 }, { "epoch": 2.709960564813637, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.6589674949646, "learning_rate": 1e-05, "loss": 0.4974, "mean_token_accuracy": 0.8704742193222046, "num_tokens": 808931935.0, "step": 21303 }, { "epoch": 2.7100877750922274, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.59923791885376, "learning_rate": 1e-05, "loss": 0.4757, "mean_token_accuracy": 0.8760886192321777, "num_tokens": 808973525.0, "step": 21304 }, { "epoch": 2.710214985370818, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.551423072814941, "learning_rate": 1e-05, "loss": 0.5073, "mean_token_accuracy": 0.8669402599334717, "num_tokens": 809008198.0, "step": 21305 }, { "epoch": 2.7103421956494085, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.433864593505859, "learning_rate": 1e-05, "loss": 0.4119, "mean_token_accuracy": 0.8969221711158752, "num_tokens": 809038267.0, "step": 21306 }, { "epoch": 2.710469405927999, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.531282901763916, "learning_rate": 1e-05, "loss": 0.4883, "mean_token_accuracy": 0.8693294525146484, "num_tokens": 809073380.0, "step": 21307 }, { "epoch": 2.7105966162065895, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.264717102050781, "learning_rate": 1e-05, "loss": 0.4423, "mean_token_accuracy": 0.8862406611442566, "num_tokens": 809114635.0, "step": 21308 }, { "epoch": 2.71072382648518, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.587148666381836, "learning_rate": 1e-05, "loss": 0.413, "mean_token_accuracy": 0.8991459012031555, "num_tokens": 809149441.0, "step": 21309 }, { "epoch": 2.7108510367637706, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.103537082672119, "learning_rate": 1e-05, "loss": 0.4507, "mean_token_accuracy": 0.8824394941329956, "num_tokens": 809190902.0, "step": 21310 }, { "epoch": 2.710978247042361, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.745205879211426, "learning_rate": 1e-05, "loss": 0.471, "mean_token_accuracy": 0.883021354675293, "num_tokens": 809239297.0, "step": 21311 }, { "epoch": 2.7111054573209517, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.906854152679443, "learning_rate": 1e-05, "loss": 0.4581, "mean_token_accuracy": 0.8817853927612305, "num_tokens": 809277304.0, "step": 21312 }, { "epoch": 2.711232667599542, "ewc_loss": 0.140625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00011157989501953125, "grad_norm": 7.1083831787109375, "learning_rate": 1e-05, "loss": 0.533, "mean_token_accuracy": 0.8639134168624878, "num_tokens": 809313975.0, "step": 21313 }, { "epoch": 2.7113598778781327, "ewc_loss": 0.11328125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 8.392333984375e-05, "grad_norm": 5.7349700927734375, "learning_rate": 1e-05, "loss": 0.4112, "mean_token_accuracy": 0.8931686282157898, "num_tokens": 809354214.0, "step": 21314 }, { "epoch": 2.711487088156723, "ewc_loss": 0.154296875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00012493133544921875, "grad_norm": 7.664330005645752, "learning_rate": 1e-05, "loss": 0.4925, "mean_token_accuracy": 0.8832895159721375, "num_tokens": 809388248.0, "step": 21315 }, { "epoch": 2.7116142984353138, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 5.881227016448975, "learning_rate": 1e-05, "loss": 0.4685, "mean_token_accuracy": 0.8760492205619812, "num_tokens": 809428256.0, "step": 21316 }, { "epoch": 2.711741508713904, "ewc_loss": 0.166015625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00013637542724609375, "grad_norm": 8.189353942871094, "learning_rate": 1e-05, "loss": 0.4936, "mean_token_accuracy": 0.8828638195991516, "num_tokens": 809463706.0, "step": 21317 }, { "epoch": 2.711868718992495, "ewc_loss": 0.140625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00011205673217773438, "grad_norm": 6.587663173675537, "learning_rate": 1e-05, "loss": 0.51, "mean_token_accuracy": 0.8763413429260254, "num_tokens": 809503172.0, "step": 21318 }, { "epoch": 2.711995929271085, "ewc_loss": 0.162109375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00013256072998046875, "grad_norm": 8.028908729553223, "learning_rate": 1e-05, "loss": 0.5292, "mean_token_accuracy": 0.8742284774780273, "num_tokens": 809539894.0, "step": 21319 }, { "epoch": 2.712123139549676, "ewc_loss": 0.1494140625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00012063980102539062, "grad_norm": 7.067810535430908, "learning_rate": 1e-05, "loss": 0.5554, "mean_token_accuracy": 0.861864447593689, "num_tokens": 809582781.0, "step": 21320 }, { "epoch": 2.712250349828266, "ewc_loss": 0.1484375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00011968612670898438, "grad_norm": 7.451197624206543, "learning_rate": 1e-05, "loss": 0.4985, "mean_token_accuracy": 0.8803175687789917, "num_tokens": 809627077.0, "step": 21321 }, { "epoch": 2.7123775601068565, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00011491775512695312, "grad_norm": 7.024538040161133, "learning_rate": 1e-05, "loss": 0.5144, "mean_token_accuracy": 0.8705978393554688, "num_tokens": 809663063.0, "step": 21322 }, { "epoch": 2.712504770385447, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.963298320770264, "learning_rate": 1e-05, "loss": 0.5319, "mean_token_accuracy": 0.8637791275978088, "num_tokens": 809702386.0, "step": 21323 }, { "epoch": 2.7126319806640375, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 6.86612606048584, "learning_rate": 1e-05, "loss": 0.4951, "mean_token_accuracy": 0.8716193437576294, "num_tokens": 809745927.0, "step": 21324 }, { "epoch": 2.712759190942628, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.607010364532471, "learning_rate": 1e-05, "loss": 0.4502, "mean_token_accuracy": 0.8839054107666016, "num_tokens": 809777796.0, "step": 21325 }, { "epoch": 2.7128864012212186, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.772782325744629, "learning_rate": 1e-05, "loss": 0.4677, "mean_token_accuracy": 0.8839988708496094, "num_tokens": 809813450.0, "step": 21326 }, { "epoch": 2.713013611499809, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.392083168029785, "learning_rate": 1e-05, "loss": 0.5209, "mean_token_accuracy": 0.8632173538208008, "num_tokens": 809850740.0, "step": 21327 }, { "epoch": 2.7131408217783997, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.7232842445373535, "learning_rate": 1e-05, "loss": 0.4659, "mean_token_accuracy": 0.8830193877220154, "num_tokens": 809886170.0, "step": 21328 }, { "epoch": 2.71326803205699, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.231374740600586, "learning_rate": 1e-05, "loss": 0.4574, "mean_token_accuracy": 0.8824257254600525, "num_tokens": 809922078.0, "step": 21329 }, { "epoch": 2.7133952423355807, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.752004146575928, "learning_rate": 1e-05, "loss": 0.4869, "mean_token_accuracy": 0.8743070363998413, "num_tokens": 809962495.0, "step": 21330 }, { "epoch": 2.7135224526141712, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.096767902374268, "learning_rate": 1e-05, "loss": 0.475, "mean_token_accuracy": 0.8734099864959717, "num_tokens": 810002278.0, "step": 21331 }, { "epoch": 2.7136496628927618, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.862678050994873, "learning_rate": 1e-05, "loss": 0.4848, "mean_token_accuracy": 0.8795864582061768, "num_tokens": 810041236.0, "step": 21332 }, { "epoch": 2.7137768731713523, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 5.963984966278076, "learning_rate": 1e-05, "loss": 0.4269, "mean_token_accuracy": 0.8926762342453003, "num_tokens": 810078649.0, "step": 21333 }, { "epoch": 2.713904083449943, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00011014938354492188, "grad_norm": 7.102558135986328, "learning_rate": 1e-05, "loss": 0.5284, "mean_token_accuracy": 0.8682379722595215, "num_tokens": 810118246.0, "step": 21334 }, { "epoch": 2.7140312937285334, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.8547844886779785, "learning_rate": 1e-05, "loss": 0.4314, "mean_token_accuracy": 0.8904855847358704, "num_tokens": 810153407.0, "step": 21335 }, { "epoch": 2.714158504007124, "ewc_loss": 0.1484375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00011968612670898438, "grad_norm": 7.4801411628723145, "learning_rate": 1e-05, "loss": 0.5, "mean_token_accuracy": 0.8772986531257629, "num_tokens": 810188110.0, "step": 21336 }, { "epoch": 2.7142857142857144, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 5.915611743927002, "learning_rate": 1e-05, "loss": 0.4888, "mean_token_accuracy": 0.869646430015564, "num_tokens": 810225290.0, "step": 21337 }, { "epoch": 2.714412924564305, "ewc_loss": 0.158203125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00012874603271484375, "grad_norm": 7.851027965545654, "learning_rate": 1e-05, "loss": 0.4894, "mean_token_accuracy": 0.8843452334403992, "num_tokens": 810259801.0, "step": 21338 }, { "epoch": 2.7145401348428955, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.2954511642456055, "learning_rate": 1e-05, "loss": 0.4544, "mean_token_accuracy": 0.8857298493385315, "num_tokens": 810296556.0, "step": 21339 }, { "epoch": 2.7146673451214856, "ewc_loss": 0.158203125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00012874603271484375, "grad_norm": 7.885013580322266, "learning_rate": 1e-05, "loss": 0.4979, "mean_token_accuracy": 0.879418134689331, "num_tokens": 810332282.0, "step": 21340 }, { "epoch": 2.7147945554000765, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00011444091796875, "grad_norm": 6.73151159286499, "learning_rate": 1e-05, "loss": 0.4917, "mean_token_accuracy": 0.8792492151260376, "num_tokens": 810368017.0, "step": 21341 }, { "epoch": 2.7149217656786666, "ewc_loss": 0.1494140625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.0001201629638671875, "grad_norm": 7.502228736877441, "learning_rate": 1e-05, "loss": 0.5559, "mean_token_accuracy": 0.8588685393333435, "num_tokens": 810412239.0, "step": 21342 }, { "epoch": 2.7150489759572576, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00011348724365234375, "grad_norm": 6.837107181549072, "learning_rate": 1e-05, "loss": 0.5113, "mean_token_accuracy": 0.8706815242767334, "num_tokens": 810452319.0, "step": 21343 }, { "epoch": 2.7151761862358477, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 7.032257556915283, "learning_rate": 1e-05, "loss": 0.5026, "mean_token_accuracy": 0.8715707659721375, "num_tokens": 810490936.0, "step": 21344 }, { "epoch": 2.7153033965144386, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 6.772647380828857, "learning_rate": 1e-05, "loss": 0.4941, "mean_token_accuracy": 0.876872181892395, "num_tokens": 810533565.0, "step": 21345 }, { "epoch": 2.7154306067930287, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.640031337738037, "learning_rate": 1e-05, "loss": 0.4975, "mean_token_accuracy": 0.8750452995300293, "num_tokens": 810574623.0, "step": 21346 }, { "epoch": 2.7155578170716193, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.723533630371094, "learning_rate": 1e-05, "loss": 0.4592, "mean_token_accuracy": 0.8842017650604248, "num_tokens": 810609423.0, "step": 21347 }, { "epoch": 2.71568502735021, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.380697727203369, "learning_rate": 1e-05, "loss": 0.47, "mean_token_accuracy": 0.883293867111206, "num_tokens": 810643814.0, "step": 21348 }, { "epoch": 2.7158122376288003, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.699835300445557, "learning_rate": 1e-05, "loss": 0.4776, "mean_token_accuracy": 0.8790696263313293, "num_tokens": 810682393.0, "step": 21349 }, { "epoch": 2.715939447907391, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.215730667114258, "learning_rate": 1e-05, "loss": 0.4342, "mean_token_accuracy": 0.8898550271987915, "num_tokens": 810718604.0, "step": 21350 }, { "epoch": 2.7160666581859814, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.735528945922852, "learning_rate": 1e-05, "loss": 0.4718, "mean_token_accuracy": 0.8817373514175415, "num_tokens": 810751259.0, "step": 21351 }, { "epoch": 2.716193868464572, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.085270404815674, "learning_rate": 1e-05, "loss": 0.4766, "mean_token_accuracy": 0.877877414226532, "num_tokens": 810781741.0, "step": 21352 }, { "epoch": 2.7163210787431624, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.88185977935791, "learning_rate": 1e-05, "loss": 0.5661, "mean_token_accuracy": 0.8503362536430359, "num_tokens": 810820258.0, "step": 21353 }, { "epoch": 2.716448289021753, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 5.936906814575195, "learning_rate": 1e-05, "loss": 0.4531, "mean_token_accuracy": 0.884334921836853, "num_tokens": 810861341.0, "step": 21354 }, { "epoch": 2.7165754993003435, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00011014938354492188, "grad_norm": 7.100704193115234, "learning_rate": 1e-05, "loss": 0.5028, "mean_token_accuracy": 0.8720425367355347, "num_tokens": 810897963.0, "step": 21355 }, { "epoch": 2.716702709578934, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 5.821954727172852, "learning_rate": 1e-05, "loss": 0.4693, "mean_token_accuracy": 0.8789887428283691, "num_tokens": 810938042.0, "step": 21356 }, { "epoch": 2.7168299198575245, "ewc_loss": 0.150390625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00012111663818359375, "grad_norm": 7.495527744293213, "learning_rate": 1e-05, "loss": 0.4944, "mean_token_accuracy": 0.8782762289047241, "num_tokens": 810969905.0, "step": 21357 }, { "epoch": 2.716957130136115, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 5.9039106369018555, "learning_rate": 1e-05, "loss": 0.4372, "mean_token_accuracy": 0.8851800560951233, "num_tokens": 811006818.0, "step": 21358 }, { "epoch": 2.7170843404147056, "ewc_loss": 0.16015625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00013065338134765625, "grad_norm": 7.870204448699951, "learning_rate": 1e-05, "loss": 0.5104, "mean_token_accuracy": 0.8779919147491455, "num_tokens": 811045867.0, "step": 21359 }, { "epoch": 2.717211550693296, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.322455883026123, "learning_rate": 1e-05, "loss": 0.4776, "mean_token_accuracy": 0.8818182349205017, "num_tokens": 811084207.0, "step": 21360 }, { "epoch": 2.7173387609718866, "ewc_loss": 0.158203125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00012874603271484375, "grad_norm": 7.875574588775635, "learning_rate": 1e-05, "loss": 0.5377, "mean_token_accuracy": 0.8661165833473206, "num_tokens": 811123093.0, "step": 21361 }, { "epoch": 2.717465971250477, "ewc_loss": 0.1455078125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.0001163482666015625, "grad_norm": 6.752901077270508, "learning_rate": 1e-05, "loss": 0.4862, "mean_token_accuracy": 0.8787835836410522, "num_tokens": 811157951.0, "step": 21362 }, { "epoch": 2.7175931815290677, "ewc_loss": 0.1494140625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.0001201629638671875, "grad_norm": 7.457650661468506, "learning_rate": 1e-05, "loss": 0.4386, "mean_token_accuracy": 0.8985758423805237, "num_tokens": 811192743.0, "step": 21363 }, { "epoch": 2.7177203918076582, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00011444091796875, "grad_norm": 6.825662136077881, "learning_rate": 1e-05, "loss": 0.4883, "mean_token_accuracy": 0.8782474994659424, "num_tokens": 811230042.0, "step": 21364 }, { "epoch": 2.7178476020862483, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00011014938354492188, "grad_norm": 6.985297203063965, "learning_rate": 1e-05, "loss": 0.4907, "mean_token_accuracy": 0.8771809339523315, "num_tokens": 811270175.0, "step": 21365 }, { "epoch": 2.7179748123648393, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010967254638671875, "grad_norm": 6.787348747253418, "learning_rate": 1e-05, "loss": 0.477, "mean_token_accuracy": 0.8820679187774658, "num_tokens": 811302926.0, "step": 21366 }, { "epoch": 2.7181020226434294, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.582726955413818, "learning_rate": 1e-05, "loss": 0.4144, "mean_token_accuracy": 0.8991541862487793, "num_tokens": 811334503.0, "step": 21367 }, { "epoch": 2.7182292329220203, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.738997936248779, "learning_rate": 1e-05, "loss": 0.4791, "mean_token_accuracy": 0.8798699378967285, "num_tokens": 811371480.0, "step": 21368 }, { "epoch": 2.7183564432006104, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.3278489112854, "learning_rate": 1e-05, "loss": 0.4795, "mean_token_accuracy": 0.8774047493934631, "num_tokens": 811410044.0, "step": 21369 }, { "epoch": 2.718483653479201, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.74399471282959, "learning_rate": 1e-05, "loss": 0.5609, "mean_token_accuracy": 0.8509405255317688, "num_tokens": 811447815.0, "step": 21370 }, { "epoch": 2.7186108637577915, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.170392036437988, "learning_rate": 1e-05, "loss": 0.4974, "mean_token_accuracy": 0.8683911561965942, "num_tokens": 811484530.0, "step": 21371 }, { "epoch": 2.718738074036382, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.7868781089782715, "learning_rate": 1e-05, "loss": 0.4706, "mean_token_accuracy": 0.8816699981689453, "num_tokens": 811523010.0, "step": 21372 }, { "epoch": 2.7188652843149725, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.03848934173584, "learning_rate": 1e-05, "loss": 0.4662, "mean_token_accuracy": 0.8785152435302734, "num_tokens": 811559031.0, "step": 21373 }, { "epoch": 2.718992494593563, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.920712947845459, "learning_rate": 1e-05, "loss": 0.5401, "mean_token_accuracy": 0.8592904806137085, "num_tokens": 811596501.0, "step": 21374 }, { "epoch": 2.7191197048721536, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 5.923759460449219, "learning_rate": 1e-05, "loss": 0.524, "mean_token_accuracy": 0.8574902415275574, "num_tokens": 811634619.0, "step": 21375 }, { "epoch": 2.719246915150744, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.0001125335693359375, "grad_norm": 7.159246921539307, "learning_rate": 1e-05, "loss": 0.5359, "mean_token_accuracy": 0.8647712469100952, "num_tokens": 811678142.0, "step": 21376 }, { "epoch": 2.7193741254293347, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 5.868567943572998, "learning_rate": 1e-05, "loss": 0.4895, "mean_token_accuracy": 0.8716138601303101, "num_tokens": 811711072.0, "step": 21377 }, { "epoch": 2.719501335707925, "ewc_loss": 0.150390625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00012159347534179688, "grad_norm": 7.497981071472168, "learning_rate": 1e-05, "loss": 0.4535, "mean_token_accuracy": 0.8929095268249512, "num_tokens": 811746372.0, "step": 21378 }, { "epoch": 2.7196285459865157, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 5.963779926300049, "learning_rate": 1e-05, "loss": 0.5186, "mean_token_accuracy": 0.864677906036377, "num_tokens": 811788763.0, "step": 21379 }, { "epoch": 2.7197557562651062, "ewc_loss": 0.158203125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00012874603271484375, "grad_norm": 7.781739234924316, "learning_rate": 1e-05, "loss": 0.4795, "mean_token_accuracy": 0.883558988571167, "num_tokens": 811820866.0, "step": 21380 }, { "epoch": 2.7198829665436968, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.3277459144592285, "learning_rate": 1e-05, "loss": 0.4824, "mean_token_accuracy": 0.8802205324172974, "num_tokens": 811866684.0, "step": 21381 }, { "epoch": 2.7200101768222873, "ewc_loss": 0.1572265625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.0001277923583984375, "grad_norm": 7.739330768585205, "learning_rate": 1e-05, "loss": 0.4956, "mean_token_accuracy": 0.8800389766693115, "num_tokens": 811902728.0, "step": 21382 }, { "epoch": 2.720137387100878, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00011348724365234375, "grad_norm": 6.667672634124756, "learning_rate": 1e-05, "loss": 0.4787, "mean_token_accuracy": 0.8804792761802673, "num_tokens": 811946051.0, "step": 21383 }, { "epoch": 2.7202645973794684, "ewc_loss": 0.1474609375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.000118255615234375, "grad_norm": 7.384196758270264, "learning_rate": 1e-05, "loss": 0.5053, "mean_token_accuracy": 0.8755941390991211, "num_tokens": 811984148.0, "step": 21384 }, { "epoch": 2.720391807658059, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00011301040649414062, "grad_norm": 6.754170894622803, "learning_rate": 1e-05, "loss": 0.5394, "mean_token_accuracy": 0.8633038997650146, "num_tokens": 812020173.0, "step": 21385 }, { "epoch": 2.7205190179366494, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 6.937745094299316, "learning_rate": 1e-05, "loss": 0.451, "mean_token_accuracy": 0.8871861100196838, "num_tokens": 812054509.0, "step": 21386 }, { "epoch": 2.72064622821524, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 6.7228827476501465, "learning_rate": 1e-05, "loss": 0.4692, "mean_token_accuracy": 0.8826307654380798, "num_tokens": 812088717.0, "step": 21387 }, { "epoch": 2.7207734384938305, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.549437046051025, "learning_rate": 1e-05, "loss": 0.4882, "mean_token_accuracy": 0.873232901096344, "num_tokens": 812126230.0, "step": 21388 }, { "epoch": 2.720900648772421, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.713316917419434, "learning_rate": 1e-05, "loss": 0.4603, "mean_token_accuracy": 0.8867648839950562, "num_tokens": 812157385.0, "step": 21389 }, { "epoch": 2.721027859051011, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.313363552093506, "learning_rate": 1e-05, "loss": 0.5421, "mean_token_accuracy": 0.8592265844345093, "num_tokens": 812184653.0, "step": 21390 }, { "epoch": 2.721155069329602, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.706151962280273, "learning_rate": 1e-05, "loss": 0.4507, "mean_token_accuracy": 0.88834547996521, "num_tokens": 812226784.0, "step": 21391 }, { "epoch": 2.721282279608192, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.150227069854736, "learning_rate": 1e-05, "loss": 0.4883, "mean_token_accuracy": 0.8740370273590088, "num_tokens": 812261088.0, "step": 21392 }, { "epoch": 2.721409489886783, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.7621307373046875, "learning_rate": 1e-05, "loss": 0.4445, "mean_token_accuracy": 0.8905735611915588, "num_tokens": 812296859.0, "step": 21393 }, { "epoch": 2.721536700165373, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.023025989532471, "learning_rate": 1e-05, "loss": 0.4559, "mean_token_accuracy": 0.882205605506897, "num_tokens": 812334027.0, "step": 21394 }, { "epoch": 2.7216639104439637, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.904872417449951, "learning_rate": 1e-05, "loss": 0.4891, "mean_token_accuracy": 0.8760563135147095, "num_tokens": 812370425.0, "step": 21395 }, { "epoch": 2.7217911207225542, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 5.916242599487305, "learning_rate": 1e-05, "loss": 0.4656, "mean_token_accuracy": 0.8789437413215637, "num_tokens": 812402067.0, "step": 21396 }, { "epoch": 2.7219183310011448, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.0001125335693359375, "grad_norm": 7.145224094390869, "learning_rate": 1e-05, "loss": 0.4409, "mean_token_accuracy": 0.892856240272522, "num_tokens": 812436482.0, "step": 21397 }, { "epoch": 2.7220455412797353, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 5.851144790649414, "learning_rate": 1e-05, "loss": 0.51, "mean_token_accuracy": 0.8644427061080933, "num_tokens": 812476059.0, "step": 21398 }, { "epoch": 2.722172751558326, "ewc_loss": 0.150390625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00012159347534179688, "grad_norm": 7.50820255279541, "learning_rate": 1e-05, "loss": 0.4586, "mean_token_accuracy": 0.8905757665634155, "num_tokens": 812513846.0, "step": 21399 }, { "epoch": 2.7222999618369164, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 5.962225914001465, "learning_rate": 1e-05, "loss": 0.463, "mean_token_accuracy": 0.8850011825561523, "num_tokens": 812550000.0, "step": 21400 }, { "epoch": 2.722427172115507, "ewc_loss": 0.16015625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00013065338134765625, "grad_norm": 7.814468860626221, "learning_rate": 1e-05, "loss": 0.5023, "mean_token_accuracy": 0.8781591057777405, "num_tokens": 812589539.0, "step": 21401 }, { "epoch": 2.7225543823940974, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.3416547775268555, "learning_rate": 1e-05, "loss": 0.4483, "mean_token_accuracy": 0.8898573517799377, "num_tokens": 812622699.0, "step": 21402 }, { "epoch": 2.722681592672688, "ewc_loss": 0.158203125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00012874603271484375, "grad_norm": 7.776276111602783, "learning_rate": 1e-05, "loss": 0.4762, "mean_token_accuracy": 0.8860294222831726, "num_tokens": 812662945.0, "step": 21403 }, { "epoch": 2.7228088029512785, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00011396408081054688, "grad_norm": 6.679352760314941, "learning_rate": 1e-05, "loss": 0.4862, "mean_token_accuracy": 0.8814278841018677, "num_tokens": 812700650.0, "step": 21404 }, { "epoch": 2.722936013229869, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00011777877807617188, "grad_norm": 7.397727012634277, "learning_rate": 1e-05, "loss": 0.4574, "mean_token_accuracy": 0.8925248384475708, "num_tokens": 812742651.0, "step": 21405 }, { "epoch": 2.7230632235084595, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00011301040649414062, "grad_norm": 6.7670979499816895, "learning_rate": 1e-05, "loss": 0.5124, "mean_token_accuracy": 0.8691586256027222, "num_tokens": 812783792.0, "step": 21406 }, { "epoch": 2.72319043378705, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 6.949296474456787, "learning_rate": 1e-05, "loss": 0.5046, "mean_token_accuracy": 0.8713279962539673, "num_tokens": 812829059.0, "step": 21407 }, { "epoch": 2.7233176440656406, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 6.75618314743042, "learning_rate": 1e-05, "loss": 0.4624, "mean_token_accuracy": 0.8828758001327515, "num_tokens": 812859088.0, "step": 21408 }, { "epoch": 2.723444854344231, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.55568790435791, "learning_rate": 1e-05, "loss": 0.5122, "mean_token_accuracy": 0.8687921166419983, "num_tokens": 812899057.0, "step": 21409 }, { "epoch": 2.7235720646228216, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.741319179534912, "learning_rate": 1e-05, "loss": 0.5156, "mean_token_accuracy": 0.8689699172973633, "num_tokens": 812940711.0, "step": 21410 }, { "epoch": 2.723699274901412, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.274534225463867, "learning_rate": 1e-05, "loss": 0.5266, "mean_token_accuracy": 0.858260989189148, "num_tokens": 812986967.0, "step": 21411 }, { "epoch": 2.7238264851800027, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.776123523712158, "learning_rate": 1e-05, "loss": 0.4781, "mean_token_accuracy": 0.879638671875, "num_tokens": 813024396.0, "step": 21412 }, { "epoch": 2.723953695458593, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.101479530334473, "learning_rate": 1e-05, "loss": 0.4839, "mean_token_accuracy": 0.8709180355072021, "num_tokens": 813058204.0, "step": 21413 }, { "epoch": 2.7240809057371838, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.880715847015381, "learning_rate": 1e-05, "loss": 0.5115, "mean_token_accuracy": 0.8689504861831665, "num_tokens": 813098477.0, "step": 21414 }, { "epoch": 2.724208116015774, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 5.958158016204834, "learning_rate": 1e-05, "loss": 0.4939, "mean_token_accuracy": 0.8756489753723145, "num_tokens": 813138091.0, "step": 21415 }, { "epoch": 2.724335326294365, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 7.099365234375, "learning_rate": 1e-05, "loss": 0.5075, "mean_token_accuracy": 0.8716022968292236, "num_tokens": 813170280.0, "step": 21416 }, { "epoch": 2.724462536572955, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 5.875221252441406, "learning_rate": 1e-05, "loss": 0.5045, "mean_token_accuracy": 0.8666871786117554, "num_tokens": 813209198.0, "step": 21417 }, { "epoch": 2.724589746851546, "ewc_loss": 0.1474609375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00011873245239257812, "grad_norm": 7.429195880889893, "learning_rate": 1e-05, "loss": 0.4282, "mean_token_accuracy": 0.8996431827545166, "num_tokens": 813248836.0, "step": 21418 }, { "epoch": 2.724716957130136, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 5.941037654876709, "learning_rate": 1e-05, "loss": 0.4911, "mean_token_accuracy": 0.8716726303100586, "num_tokens": 813279421.0, "step": 21419 }, { "epoch": 2.7248441674087265, "ewc_loss": 0.158203125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00012874603271484375, "grad_norm": 7.737051963806152, "learning_rate": 1e-05, "loss": 0.5385, "mean_token_accuracy": 0.8677890300750732, "num_tokens": 813321615.0, "step": 21420 }, { "epoch": 2.724971377687317, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.244607448577881, "learning_rate": 1e-05, "loss": 0.4789, "mean_token_accuracy": 0.8788751363754272, "num_tokens": 813357338.0, "step": 21421 }, { "epoch": 2.7250985879659075, "ewc_loss": 0.158203125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00012874603271484375, "grad_norm": 7.7786946296691895, "learning_rate": 1e-05, "loss": 0.5347, "mean_token_accuracy": 0.8755003213882446, "num_tokens": 813397774.0, "step": 21422 }, { "epoch": 2.725225798244498, "ewc_loss": 0.140625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00011110305786132812, "grad_norm": 6.613635540008545, "learning_rate": 1e-05, "loss": 0.4808, "mean_token_accuracy": 0.8847662210464478, "num_tokens": 813432373.0, "step": 21423 }, { "epoch": 2.7253530085230886, "ewc_loss": 0.1494140625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.0001201629638671875, "grad_norm": 7.467195510864258, "learning_rate": 1e-05, "loss": 0.4478, "mean_token_accuracy": 0.8958843946456909, "num_tokens": 813472354.0, "step": 21424 }, { "epoch": 2.725480218801679, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.0001125335693359375, "grad_norm": 6.743055820465088, "learning_rate": 1e-05, "loss": 0.459, "mean_token_accuracy": 0.8858746290206909, "num_tokens": 813505498.0, "step": 21425 }, { "epoch": 2.7256074290802697, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00011014938354492188, "grad_norm": 7.044565677642822, "learning_rate": 1e-05, "loss": 0.4741, "mean_token_accuracy": 0.8813233375549316, "num_tokens": 813546886.0, "step": 21426 }, { "epoch": 2.72573463935886, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00011014938354492188, "grad_norm": 6.720724582672119, "learning_rate": 1e-05, "loss": 0.4803, "mean_token_accuracy": 0.881519079208374, "num_tokens": 813583753.0, "step": 21427 }, { "epoch": 2.7258618496374507, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.663207054138184, "learning_rate": 1e-05, "loss": 0.4885, "mean_token_accuracy": 0.8765271902084351, "num_tokens": 813617561.0, "step": 21428 }, { "epoch": 2.7259890599160412, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.689282417297363, "learning_rate": 1e-05, "loss": 0.5346, "mean_token_accuracy": 0.8615595102310181, "num_tokens": 813652440.0, "step": 21429 }, { "epoch": 2.7261162701946318, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.380788326263428, "learning_rate": 1e-05, "loss": 0.4551, "mean_token_accuracy": 0.8835691213607788, "num_tokens": 813691118.0, "step": 21430 }, { "epoch": 2.7262434804732223, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.675510883331299, "learning_rate": 1e-05, "loss": 0.5124, "mean_token_accuracy": 0.869597315788269, "num_tokens": 813733238.0, "step": 21431 }, { "epoch": 2.726370690751813, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.205780029296875, "learning_rate": 1e-05, "loss": 0.4582, "mean_token_accuracy": 0.883466362953186, "num_tokens": 813770002.0, "step": 21432 }, { "epoch": 2.7264979010304033, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.697887420654297, "learning_rate": 1e-05, "loss": 0.4929, "mean_token_accuracy": 0.8740624785423279, "num_tokens": 813810778.0, "step": 21433 }, { "epoch": 2.726625111308994, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.09052038192749, "learning_rate": 1e-05, "loss": 0.4649, "mean_token_accuracy": 0.8758300542831421, "num_tokens": 813844004.0, "step": 21434 }, { "epoch": 2.7267523215875844, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.774744987487793, "learning_rate": 1e-05, "loss": 0.4684, "mean_token_accuracy": 0.8818225860595703, "num_tokens": 813884399.0, "step": 21435 }, { "epoch": 2.726879531866175, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 5.9726386070251465, "learning_rate": 1e-05, "loss": 0.4417, "mean_token_accuracy": 0.8891609907150269, "num_tokens": 813920454.0, "step": 21436 }, { "epoch": 2.7270067421447655, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.975616931915283, "learning_rate": 1e-05, "loss": 0.4562, "mean_token_accuracy": 0.8853955864906311, "num_tokens": 813954629.0, "step": 21437 }, { "epoch": 2.7271339524233555, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 5.84193754196167, "learning_rate": 1e-05, "loss": 0.456, "mean_token_accuracy": 0.8812727928161621, "num_tokens": 813992661.0, "step": 21438 }, { "epoch": 2.7272611627019465, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00011682510375976562, "grad_norm": 7.323453903198242, "learning_rate": 1e-05, "loss": 0.4802, "mean_token_accuracy": 0.8804997801780701, "num_tokens": 814024401.0, "step": 21439 }, { "epoch": 2.7273883729805366, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 5.819922924041748, "learning_rate": 1e-05, "loss": 0.47, "mean_token_accuracy": 0.8792133331298828, "num_tokens": 814064950.0, "step": 21440 }, { "epoch": 2.7275155832591276, "ewc_loss": 0.1572265625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.0001277923583984375, "grad_norm": 7.727653503417969, "learning_rate": 1e-05, "loss": 0.4999, "mean_token_accuracy": 0.8801831007003784, "num_tokens": 814109503.0, "step": 21441 }, { "epoch": 2.7276427935377177, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.102385520935059, "learning_rate": 1e-05, "loss": 0.5012, "mean_token_accuracy": 0.8725113868713379, "num_tokens": 814143981.0, "step": 21442 }, { "epoch": 2.7277700038163086, "ewc_loss": 0.1630859375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.000133514404296875, "grad_norm": 7.947154998779297, "learning_rate": 1e-05, "loss": 0.4794, "mean_token_accuracy": 0.8870601654052734, "num_tokens": 814180781.0, "step": 21443 }, { "epoch": 2.7278972140948987, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00011301040649414062, "grad_norm": 6.581156253814697, "learning_rate": 1e-05, "loss": 0.4957, "mean_token_accuracy": 0.8748704791069031, "num_tokens": 814224841.0, "step": 21444 }, { "epoch": 2.7280244243734892, "ewc_loss": 0.1552734375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.000125885009765625, "grad_norm": 7.701397895812988, "learning_rate": 1e-05, "loss": 0.5026, "mean_token_accuracy": 0.8791770339012146, "num_tokens": 814258618.0, "step": 21445 }, { "epoch": 2.7281516346520798, "ewc_loss": 0.1455078125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00011682510375976562, "grad_norm": 6.821935653686523, "learning_rate": 1e-05, "loss": 0.4932, "mean_token_accuracy": 0.8810304999351501, "num_tokens": 814297286.0, "step": 21446 }, { "epoch": 2.7282788449306703, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00011539459228515625, "grad_norm": 7.254031181335449, "learning_rate": 1e-05, "loss": 0.4932, "mean_token_accuracy": 0.8776425719261169, "num_tokens": 814331540.0, "step": 21447 }, { "epoch": 2.728406055209261, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00011348724365234375, "grad_norm": 6.801467418670654, "learning_rate": 1e-05, "loss": 0.5347, "mean_token_accuracy": 0.8631876707077026, "num_tokens": 814371870.0, "step": 21448 }, { "epoch": 2.7285332654878514, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.8167877197265625, "learning_rate": 1e-05, "loss": 0.4591, "mean_token_accuracy": 0.8835591077804565, "num_tokens": 814406561.0, "step": 21449 }, { "epoch": 2.728660475766442, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.73487663269043, "learning_rate": 1e-05, "loss": 0.4729, "mean_token_accuracy": 0.8809610605239868, "num_tokens": 814444176.0, "step": 21450 }, { "epoch": 2.7287876860450324, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.505868911743164, "learning_rate": 1e-05, "loss": 0.4897, "mean_token_accuracy": 0.87452232837677, "num_tokens": 814485224.0, "step": 21451 }, { "epoch": 2.728914896323623, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.69044303894043, "learning_rate": 1e-05, "loss": 0.4618, "mean_token_accuracy": 0.8833644390106201, "num_tokens": 814518006.0, "step": 21452 }, { "epoch": 2.7290421066022135, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.3090362548828125, "learning_rate": 1e-05, "loss": 0.4849, "mean_token_accuracy": 0.8758348226547241, "num_tokens": 814562883.0, "step": 21453 }, { "epoch": 2.729169316880804, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.675744533538818, "learning_rate": 1e-05, "loss": 0.4653, "mean_token_accuracy": 0.8832930326461792, "num_tokens": 814604619.0, "step": 21454 }, { "epoch": 2.7292965271593945, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.182422637939453, "learning_rate": 1e-05, "loss": 0.4764, "mean_token_accuracy": 0.8730454444885254, "num_tokens": 814641112.0, "step": 21455 }, { "epoch": 2.729423737437985, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.721740245819092, "learning_rate": 1e-05, "loss": 0.4758, "mean_token_accuracy": 0.8764878511428833, "num_tokens": 814674300.0, "step": 21456 }, { "epoch": 2.7295509477165756, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.049199104309082, "learning_rate": 1e-05, "loss": 0.4544, "mean_token_accuracy": 0.8828679323196411, "num_tokens": 814711013.0, "step": 21457 }, { "epoch": 2.729678157995166, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.851008415222168, "learning_rate": 1e-05, "loss": 0.4601, "mean_token_accuracy": 0.8841497898101807, "num_tokens": 814750789.0, "step": 21458 }, { "epoch": 2.7298053682737566, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 5.913865566253662, "learning_rate": 1e-05, "loss": 0.4673, "mean_token_accuracy": 0.8779098987579346, "num_tokens": 814791498.0, "step": 21459 }, { "epoch": 2.729932578552347, "ewc_loss": 0.140625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00011205673217773438, "grad_norm": 7.120453357696533, "learning_rate": 1e-05, "loss": 0.4845, "mean_token_accuracy": 0.8799920082092285, "num_tokens": 814829811.0, "step": 21460 }, { "epoch": 2.7300597888309377, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 5.79600715637207, "learning_rate": 1e-05, "loss": 0.4699, "mean_token_accuracy": 0.8768285512924194, "num_tokens": 814864878.0, "step": 21461 }, { "epoch": 2.7301869991095282, "ewc_loss": 0.1513671875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.0001220703125, "grad_norm": 7.557328224182129, "learning_rate": 1e-05, "loss": 0.4839, "mean_token_accuracy": 0.8822153806686401, "num_tokens": 814903843.0, "step": 21462 }, { "epoch": 2.7303142093881183, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 5.900907039642334, "learning_rate": 1e-05, "loss": 0.4504, "mean_token_accuracy": 0.8820033073425293, "num_tokens": 814943076.0, "step": 21463 }, { "epoch": 2.7304414196667093, "ewc_loss": 0.1630859375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.000133514404296875, "grad_norm": 7.945437908172607, "learning_rate": 1e-05, "loss": 0.4907, "mean_token_accuracy": 0.8863067626953125, "num_tokens": 814981257.0, "step": 21464 }, { "epoch": 2.7305686299452994, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 6.376132965087891, "learning_rate": 1e-05, "loss": 0.4603, "mean_token_accuracy": 0.8852682113647461, "num_tokens": 815019763.0, "step": 21465 }, { "epoch": 2.7306958402238903, "ewc_loss": 0.1611328125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.0001316070556640625, "grad_norm": 7.902297019958496, "learning_rate": 1e-05, "loss": 0.5073, "mean_token_accuracy": 0.8792212009429932, "num_tokens": 815061444.0, "step": 21466 }, { "epoch": 2.7308230505024804, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00011730194091796875, "grad_norm": 6.805889129638672, "learning_rate": 1e-05, "loss": 0.5078, "mean_token_accuracy": 0.8733880519866943, "num_tokens": 815105141.0, "step": 21467 }, { "epoch": 2.730950260781071, "ewc_loss": 0.1494140625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00012063980102539062, "grad_norm": 7.473235607147217, "learning_rate": 1e-05, "loss": 0.5082, "mean_token_accuracy": 0.8759484887123108, "num_tokens": 815142839.0, "step": 21468 }, { "epoch": 2.7310774710596615, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00011444091796875, "grad_norm": 6.8596320152282715, "learning_rate": 1e-05, "loss": 0.4635, "mean_token_accuracy": 0.8869181871414185, "num_tokens": 815183077.0, "step": 21469 }, { "epoch": 2.731204681338252, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 7.014675617218018, "learning_rate": 1e-05, "loss": 0.4874, "mean_token_accuracy": 0.8779391646385193, "num_tokens": 815222601.0, "step": 21470 }, { "epoch": 2.7313318916168425, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010967254638671875, "grad_norm": 6.777885913848877, "learning_rate": 1e-05, "loss": 0.5081, "mean_token_accuracy": 0.8737657070159912, "num_tokens": 815264587.0, "step": 21471 }, { "epoch": 2.731459101895433, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.639313220977783, "learning_rate": 1e-05, "loss": 0.4899, "mean_token_accuracy": 0.8758840560913086, "num_tokens": 815302575.0, "step": 21472 }, { "epoch": 2.7315863121740236, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.705115795135498, "learning_rate": 1e-05, "loss": 0.4827, "mean_token_accuracy": 0.8787959814071655, "num_tokens": 815342230.0, "step": 21473 }, { "epoch": 2.731713522452614, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.403861045837402, "learning_rate": 1e-05, "loss": 0.4534, "mean_token_accuracy": 0.8847736716270447, "num_tokens": 815375906.0, "step": 21474 }, { "epoch": 2.7318407327312046, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.675902843475342, "learning_rate": 1e-05, "loss": 0.4774, "mean_token_accuracy": 0.8738554120063782, "num_tokens": 815415766.0, "step": 21475 }, { "epoch": 2.731967943009795, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.243984699249268, "learning_rate": 1e-05, "loss": 0.4628, "mean_token_accuracy": 0.8847119212150574, "num_tokens": 815449979.0, "step": 21476 }, { "epoch": 2.7320951532883857, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.6961517333984375, "learning_rate": 1e-05, "loss": 0.4991, "mean_token_accuracy": 0.8723914623260498, "num_tokens": 815489736.0, "step": 21477 }, { "epoch": 2.7322223635669762, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.101491928100586, "learning_rate": 1e-05, "loss": 0.4396, "mean_token_accuracy": 0.8868647813796997, "num_tokens": 815525713.0, "step": 21478 }, { "epoch": 2.7323495738455668, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.790560245513916, "learning_rate": 1e-05, "loss": 0.5019, "mean_token_accuracy": 0.8724278211593628, "num_tokens": 815569231.0, "step": 21479 }, { "epoch": 2.7324767841241573, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 5.969925403594971, "learning_rate": 1e-05, "loss": 0.497, "mean_token_accuracy": 0.870090901851654, "num_tokens": 815607958.0, "step": 21480 }, { "epoch": 2.732603994402748, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 7.01098108291626, "learning_rate": 1e-05, "loss": 0.5062, "mean_token_accuracy": 0.8689652681350708, "num_tokens": 815645919.0, "step": 21481 }, { "epoch": 2.7327312046813383, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.83522367477417, "learning_rate": 1e-05, "loss": 0.4889, "mean_token_accuracy": 0.8722837567329407, "num_tokens": 815688568.0, "step": 21482 }, { "epoch": 2.732858414959929, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00011730194091796875, "grad_norm": 7.384089469909668, "learning_rate": 1e-05, "loss": 0.5185, "mean_token_accuracy": 0.8712534308433533, "num_tokens": 815723130.0, "step": 21483 }, { "epoch": 2.7329856252385194, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 5.81838321685791, "learning_rate": 1e-05, "loss": 0.4741, "mean_token_accuracy": 0.8759634494781494, "num_tokens": 815764806.0, "step": 21484 }, { "epoch": 2.73311283551711, "ewc_loss": 0.16015625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00013065338134765625, "grad_norm": 7.855214595794678, "learning_rate": 1e-05, "loss": 0.534, "mean_token_accuracy": 0.8694342374801636, "num_tokens": 815800226.0, "step": 21485 }, { "epoch": 2.7332400457957005, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.177822113037109, "learning_rate": 1e-05, "loss": 0.444, "mean_token_accuracy": 0.8875043392181396, "num_tokens": 815833625.0, "step": 21486 }, { "epoch": 2.733367256074291, "ewc_loss": 0.1640625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00013446807861328125, "grad_norm": 8.022411346435547, "learning_rate": 1e-05, "loss": 0.4845, "mean_token_accuracy": 0.8859140276908875, "num_tokens": 815870617.0, "step": 21487 }, { "epoch": 2.733494466352881, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00011682510375976562, "grad_norm": 6.727370738983154, "learning_rate": 1e-05, "loss": 0.5128, "mean_token_accuracy": 0.8726369142532349, "num_tokens": 815910854.0, "step": 21488 }, { "epoch": 2.733621676631472, "ewc_loss": 0.1533203125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.0001239776611328125, "grad_norm": 7.67563533782959, "learning_rate": 1e-05, "loss": 0.486, "mean_token_accuracy": 0.882393479347229, "num_tokens": 815946467.0, "step": 21489 }, { "epoch": 2.733748886910062, "ewc_loss": 0.1455078125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.0001163482666015625, "grad_norm": 6.905763149261475, "learning_rate": 1e-05, "loss": 0.5525, "mean_token_accuracy": 0.8621688485145569, "num_tokens": 815986924.0, "step": 21490 }, { "epoch": 2.733876097188653, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00011348724365234375, "grad_norm": 7.181199073791504, "learning_rate": 1e-05, "loss": 0.4645, "mean_token_accuracy": 0.884341835975647, "num_tokens": 816023960.0, "step": 21491 }, { "epoch": 2.734003307467243, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.0001125335693359375, "grad_norm": 6.818600654602051, "learning_rate": 1e-05, "loss": 0.5057, "mean_token_accuracy": 0.8737219572067261, "num_tokens": 816063911.0, "step": 21492 }, { "epoch": 2.7341305177458337, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.779317855834961, "learning_rate": 1e-05, "loss": 0.5531, "mean_token_accuracy": 0.8556470274925232, "num_tokens": 816099979.0, "step": 21493 }, { "epoch": 2.7342577280244242, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.73689079284668, "learning_rate": 1e-05, "loss": 0.5158, "mean_token_accuracy": 0.8664487600326538, "num_tokens": 816135226.0, "step": 21494 }, { "epoch": 2.7343849383030148, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.470554828643799, "learning_rate": 1e-05, "loss": 0.5101, "mean_token_accuracy": 0.8699450492858887, "num_tokens": 816174031.0, "step": 21495 }, { "epoch": 2.7345121485816053, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.679644584655762, "learning_rate": 1e-05, "loss": 0.4653, "mean_token_accuracy": 0.8824259638786316, "num_tokens": 816211194.0, "step": 21496 }, { "epoch": 2.734639358860196, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.281583786010742, "learning_rate": 1e-05, "loss": 0.4679, "mean_token_accuracy": 0.8791919946670532, "num_tokens": 816253849.0, "step": 21497 }, { "epoch": 2.7347665691387864, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.690088748931885, "learning_rate": 1e-05, "loss": 0.4806, "mean_token_accuracy": 0.8769931793212891, "num_tokens": 816290081.0, "step": 21498 }, { "epoch": 2.734893779417377, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.139306545257568, "learning_rate": 1e-05, "loss": 0.4035, "mean_token_accuracy": 0.8972029685974121, "num_tokens": 816324818.0, "step": 21499 }, { "epoch": 2.7350209896959674, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.762570381164551, "learning_rate": 1e-05, "loss": 0.503, "mean_token_accuracy": 0.8758103847503662, "num_tokens": 816369946.0, "step": 21500 }, { "epoch": 2.735148199974558, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 6.008598327636719, "learning_rate": 1e-05, "loss": 0.5153, "mean_token_accuracy": 0.8635437488555908, "num_tokens": 816405541.0, "step": 21501 }, { "epoch": 2.7352754102531485, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.9444260597229, "learning_rate": 1e-05, "loss": 0.4933, "mean_token_accuracy": 0.8724528551101685, "num_tokens": 816441232.0, "step": 21502 }, { "epoch": 2.735402620531739, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.868178844451904, "learning_rate": 1e-05, "loss": 0.4747, "mean_token_accuracy": 0.8741657733917236, "num_tokens": 816474538.0, "step": 21503 }, { "epoch": 2.7355298308103295, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00011444091796875, "grad_norm": 7.2711687088012695, "learning_rate": 1e-05, "loss": 0.4986, "mean_token_accuracy": 0.8741387724876404, "num_tokens": 816512706.0, "step": 21504 }, { "epoch": 2.73565704108892, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 5.798493385314941, "learning_rate": 1e-05, "loss": 0.4434, "mean_token_accuracy": 0.8863165974617004, "num_tokens": 816548638.0, "step": 21505 }, { "epoch": 2.7357842513675106, "ewc_loss": 0.1572265625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.0001277923583984375, "grad_norm": 7.724350929260254, "learning_rate": 1e-05, "loss": 0.5292, "mean_token_accuracy": 0.8720101714134216, "num_tokens": 816583721.0, "step": 21506 }, { "epoch": 2.735911461646101, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.042736530303955, "learning_rate": 1e-05, "loss": 0.4706, "mean_token_accuracy": 0.879027247428894, "num_tokens": 816621873.0, "step": 21507 }, { "epoch": 2.7360386719246916, "ewc_loss": 0.1640625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00013446807861328125, "grad_norm": 8.012499809265137, "learning_rate": 1e-05, "loss": 0.5481, "mean_token_accuracy": 0.8658218383789062, "num_tokens": 816659372.0, "step": 21508 }, { "epoch": 2.736165882203282, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00011396408081054688, "grad_norm": 6.586818695068359, "learning_rate": 1e-05, "loss": 0.4512, "mean_token_accuracy": 0.8921018242835999, "num_tokens": 816696426.0, "step": 21509 }, { "epoch": 2.7362930924818727, "ewc_loss": 0.1572265625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.0001277923583984375, "grad_norm": 7.773557186126709, "learning_rate": 1e-05, "loss": 0.51, "mean_token_accuracy": 0.8817535638809204, "num_tokens": 816735344.0, "step": 21510 }, { "epoch": 2.7364203027604628, "ewc_loss": 0.1474609375, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.000118255615234375, "grad_norm": 6.875442981719971, "learning_rate": 1e-05, "loss": 0.4643, "mean_token_accuracy": 0.888073205947876, "num_tokens": 816773026.0, "step": 21511 }, { "epoch": 2.7365475130390537, "ewc_loss": 0.1455078125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.0001163482666015625, "grad_norm": 7.304779529571533, "learning_rate": 1e-05, "loss": 0.4807, "mean_token_accuracy": 0.8812465667724609, "num_tokens": 816809777.0, "step": 21512 }, { "epoch": 2.736674723317644, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00011444091796875, "grad_norm": 6.845347881317139, "learning_rate": 1e-05, "loss": 0.5239, "mean_token_accuracy": 0.8638736605644226, "num_tokens": 816849018.0, "step": 21513 }, { "epoch": 2.736801933596235, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.860606670379639, "learning_rate": 1e-05, "loss": 0.4994, "mean_token_accuracy": 0.8722479939460754, "num_tokens": 816888867.0, "step": 21514 }, { "epoch": 2.736929143874825, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.756097793579102, "learning_rate": 1e-05, "loss": 0.4571, "mean_token_accuracy": 0.8844351768493652, "num_tokens": 816929221.0, "step": 21515 }, { "epoch": 2.737056354153416, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.533975601196289, "learning_rate": 1e-05, "loss": 0.4644, "mean_token_accuracy": 0.882762610912323, "num_tokens": 816965628.0, "step": 21516 }, { "epoch": 2.737183564432006, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.703973293304443, "learning_rate": 1e-05, "loss": 0.4935, "mean_token_accuracy": 0.8736224174499512, "num_tokens": 816997908.0, "step": 21517 }, { "epoch": 2.7373107747105965, "ewc_loss": 0.125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.327249526977539, "learning_rate": 1e-05, "loss": 0.4694, "mean_token_accuracy": 0.8801064491271973, "num_tokens": 817034339.0, "step": 21518 }, { "epoch": 2.737437984989187, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.690889358520508, "learning_rate": 1e-05, "loss": 0.5085, "mean_token_accuracy": 0.8698912858963013, "num_tokens": 817073006.0, "step": 21519 }, { "epoch": 2.7375651952677775, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.178830146789551, "learning_rate": 1e-05, "loss": 0.4861, "mean_token_accuracy": 0.8723343014717102, "num_tokens": 817114517.0, "step": 21520 }, { "epoch": 2.737692405546368, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.732118606567383, "learning_rate": 1e-05, "loss": 0.4724, "mean_token_accuracy": 0.8820111751556396, "num_tokens": 817158249.0, "step": 21521 }, { "epoch": 2.7378196158249586, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.052923679351807, "learning_rate": 1e-05, "loss": 0.4644, "mean_token_accuracy": 0.8821989297866821, "num_tokens": 817197178.0, "step": 21522 }, { "epoch": 2.737946826103549, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.867640495300293, "learning_rate": 1e-05, "loss": 0.4589, "mean_token_accuracy": 0.8842781186103821, "num_tokens": 817231406.0, "step": 21523 }, { "epoch": 2.7380740363821396, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 5.918719291687012, "learning_rate": 1e-05, "loss": 0.4724, "mean_token_accuracy": 0.8794514536857605, "num_tokens": 817269984.0, "step": 21524 }, { "epoch": 2.73820124666073, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.0001125335693359375, "grad_norm": 7.144028663635254, "learning_rate": 1e-05, "loss": 0.4927, "mean_token_accuracy": 0.882194459438324, "num_tokens": 817308997.0, "step": 21525 }, { "epoch": 2.7383284569393207, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.814126491546631, "learning_rate": 1e-05, "loss": 0.5039, "mean_token_accuracy": 0.8678398132324219, "num_tokens": 817349329.0, "step": 21526 }, { "epoch": 2.7384556672179112, "ewc_loss": 0.15234375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.575489044189453, "learning_rate": 1e-05, "loss": 0.5041, "mean_token_accuracy": 0.8766836524009705, "num_tokens": 817384683.0, "step": 21527 }, { "epoch": 2.7385828774965018, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 5.928832530975342, "learning_rate": 1e-05, "loss": 0.458, "mean_token_accuracy": 0.8779380321502686, "num_tokens": 817423202.0, "step": 21528 }, { "epoch": 2.7387100877750923, "ewc_loss": 0.162109375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00013256072998046875, "grad_norm": 7.956672668457031, "learning_rate": 1e-05, "loss": 0.4822, "mean_token_accuracy": 0.8882651329040527, "num_tokens": 817462048.0, "step": 21529 }, { "epoch": 2.738837298053683, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 6.399485111236572, "learning_rate": 1e-05, "loss": 0.5007, "mean_token_accuracy": 0.8728054761886597, "num_tokens": 817508820.0, "step": 21530 }, { "epoch": 2.7389645083322733, "ewc_loss": 0.16015625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00013065338134765625, "grad_norm": 7.898308277130127, "learning_rate": 1e-05, "loss": 0.519, "mean_token_accuracy": 0.8729262351989746, "num_tokens": 817554140.0, "step": 21531 }, { "epoch": 2.739091718610864, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00011777877807617188, "grad_norm": 6.816921710968018, "learning_rate": 1e-05, "loss": 0.5417, "mean_token_accuracy": 0.8640611171722412, "num_tokens": 817601629.0, "step": 21532 }, { "epoch": 2.7392189288894544, "ewc_loss": 0.1494140625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.0001201629638671875, "grad_norm": 7.471084117889404, "learning_rate": 1e-05, "loss": 0.5268, "mean_token_accuracy": 0.8716514110565186, "num_tokens": 817641900.0, "step": 21533 }, { "epoch": 2.739346139168045, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00011539459228515625, "grad_norm": 6.86074686050415, "learning_rate": 1e-05, "loss": 0.4593, "mean_token_accuracy": 0.8896292448043823, "num_tokens": 817681843.0, "step": 21534 }, { "epoch": 2.7394733494466355, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 6.99663782119751, "learning_rate": 1e-05, "loss": 0.541, "mean_token_accuracy": 0.860549807548523, "num_tokens": 817728548.0, "step": 21535 }, { "epoch": 2.7396005597252255, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00011014938354492188, "grad_norm": 6.781143665313721, "learning_rate": 1e-05, "loss": 0.4718, "mean_token_accuracy": 0.882757306098938, "num_tokens": 817759087.0, "step": 21536 }, { "epoch": 2.7397277700038165, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.624539852142334, "learning_rate": 1e-05, "loss": 0.5098, "mean_token_accuracy": 0.871120810508728, "num_tokens": 817799056.0, "step": 21537 }, { "epoch": 2.7398549802824066, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.732158660888672, "learning_rate": 1e-05, "loss": 0.4819, "mean_token_accuracy": 0.8772956132888794, "num_tokens": 817831859.0, "step": 21538 }, { "epoch": 2.7399821905609976, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.3718647956848145, "learning_rate": 1e-05, "loss": 0.4709, "mean_token_accuracy": 0.8784257173538208, "num_tokens": 817869699.0, "step": 21539 }, { "epoch": 2.7401094008395877, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.6978912353515625, "learning_rate": 1e-05, "loss": 0.5129, "mean_token_accuracy": 0.8683897256851196, "num_tokens": 817909218.0, "step": 21540 }, { "epoch": 2.7402366111181786, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.204611778259277, "learning_rate": 1e-05, "loss": 0.4588, "mean_token_accuracy": 0.8823204636573792, "num_tokens": 817948398.0, "step": 21541 }, { "epoch": 2.7403638213967687, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.731851577758789, "learning_rate": 1e-05, "loss": 0.513, "mean_token_accuracy": 0.868635892868042, "num_tokens": 817987351.0, "step": 21542 }, { "epoch": 2.7404910316753592, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.068216800689697, "learning_rate": 1e-05, "loss": 0.4499, "mean_token_accuracy": 0.8831246495246887, "num_tokens": 818031306.0, "step": 21543 }, { "epoch": 2.7406182419539498, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.851964950561523, "learning_rate": 1e-05, "loss": 0.4748, "mean_token_accuracy": 0.8796295523643494, "num_tokens": 818067658.0, "step": 21544 }, { "epoch": 2.7407454522325403, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 5.939421653747559, "learning_rate": 1e-05, "loss": 0.5128, "mean_token_accuracy": 0.8650991320610046, "num_tokens": 818109766.0, "step": 21545 }, { "epoch": 2.740872662511131, "ewc_loss": 0.140625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00011157989501953125, "grad_norm": 7.118675708770752, "learning_rate": 1e-05, "loss": 0.5358, "mean_token_accuracy": 0.8668349981307983, "num_tokens": 818144793.0, "step": 21546 }, { "epoch": 2.7409998727897213, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.824045181274414, "learning_rate": 1e-05, "loss": 0.4985, "mean_token_accuracy": 0.8703207969665527, "num_tokens": 818184140.0, "step": 21547 }, { "epoch": 2.741127083068312, "ewc_loss": 0.150390625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00012159347534179688, "grad_norm": 7.524709701538086, "learning_rate": 1e-05, "loss": 0.5173, "mean_token_accuracy": 0.872735857963562, "num_tokens": 818222216.0, "step": 21548 }, { "epoch": 2.7412542933469024, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 5.908741474151611, "learning_rate": 1e-05, "loss": 0.4487, "mean_token_accuracy": 0.8852680921554565, "num_tokens": 818256356.0, "step": 21549 }, { "epoch": 2.741381503625493, "ewc_loss": 0.162109375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00013256072998046875, "grad_norm": 7.9204182624816895, "learning_rate": 1e-05, "loss": 0.5472, "mean_token_accuracy": 0.868899941444397, "num_tokens": 818300012.0, "step": 21550 }, { "epoch": 2.7415087139040835, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.35313081741333, "learning_rate": 1e-05, "loss": 0.4915, "mean_token_accuracy": 0.8739326000213623, "num_tokens": 818336090.0, "step": 21551 }, { "epoch": 2.741635924182674, "ewc_loss": 0.1611328125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.0001316070556640625, "grad_norm": 7.9082770347595215, "learning_rate": 1e-05, "loss": 0.5065, "mean_token_accuracy": 0.8783653974533081, "num_tokens": 818376475.0, "step": 21552 }, { "epoch": 2.7417631344612645, "ewc_loss": 0.1455078125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00011682510375976562, "grad_norm": 6.786603927612305, "learning_rate": 1e-05, "loss": 0.4744, "mean_token_accuracy": 0.8861504197120667, "num_tokens": 818417483.0, "step": 21553 }, { "epoch": 2.741890344739855, "ewc_loss": 0.1494140625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00012063980102539062, "grad_norm": 7.49453067779541, "learning_rate": 1e-05, "loss": 0.5038, "mean_token_accuracy": 0.8749969005584717, "num_tokens": 818460323.0, "step": 21554 }, { "epoch": 2.7420175550184456, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00011491775512695312, "grad_norm": 6.8686065673828125, "learning_rate": 1e-05, "loss": 0.5015, "mean_token_accuracy": 0.8759922385215759, "num_tokens": 818496989.0, "step": 21555 }, { "epoch": 2.742144765297036, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 7.026236534118652, "learning_rate": 1e-05, "loss": 0.4479, "mean_token_accuracy": 0.8896759748458862, "num_tokens": 818532755.0, "step": 21556 }, { "epoch": 2.7422719755756266, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010967254638671875, "grad_norm": 6.787003517150879, "learning_rate": 1e-05, "loss": 0.5425, "mean_token_accuracy": 0.8597648739814758, "num_tokens": 818575237.0, "step": 21557 }, { "epoch": 2.742399185854217, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.64151668548584, "learning_rate": 1e-05, "loss": 0.4584, "mean_token_accuracy": 0.882804811000824, "num_tokens": 818607865.0, "step": 21558 }, { "epoch": 2.7425263961328077, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.7165727615356445, "learning_rate": 1e-05, "loss": 0.4638, "mean_token_accuracy": 0.8859567642211914, "num_tokens": 818647807.0, "step": 21559 }, { "epoch": 2.742653606411398, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.391537666320801, "learning_rate": 1e-05, "loss": 0.4828, "mean_token_accuracy": 0.8765356540679932, "num_tokens": 818686951.0, "step": 21560 }, { "epoch": 2.7427808166899883, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.694223880767822, "learning_rate": 1e-05, "loss": 0.484, "mean_token_accuracy": 0.873996376991272, "num_tokens": 818724813.0, "step": 21561 }, { "epoch": 2.7429080269685793, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.233173847198486, "learning_rate": 1e-05, "loss": 0.46, "mean_token_accuracy": 0.8810684680938721, "num_tokens": 818757865.0, "step": 21562 }, { "epoch": 2.7430352372471694, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.699994087219238, "learning_rate": 1e-05, "loss": 0.5062, "mean_token_accuracy": 0.8661547899246216, "num_tokens": 818796994.0, "step": 21563 }, { "epoch": 2.7431624475257603, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.106686592102051, "learning_rate": 1e-05, "loss": 0.4471, "mean_token_accuracy": 0.8846762180328369, "num_tokens": 818836408.0, "step": 21564 }, { "epoch": 2.7432896578043504, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.775055408477783, "learning_rate": 1e-05, "loss": 0.4822, "mean_token_accuracy": 0.8790668845176697, "num_tokens": 818876134.0, "step": 21565 }, { "epoch": 2.743416868082941, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 5.990145206451416, "learning_rate": 1e-05, "loss": 0.4506, "mean_token_accuracy": 0.8777340650558472, "num_tokens": 818911357.0, "step": 21566 }, { "epoch": 2.7435440783615315, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 6.96769380569458, "learning_rate": 1e-05, "loss": 0.484, "mean_token_accuracy": 0.8776441812515259, "num_tokens": 818942743.0, "step": 21567 }, { "epoch": 2.743671288640122, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.85361385345459, "learning_rate": 1e-05, "loss": 0.4702, "mean_token_accuracy": 0.8753774166107178, "num_tokens": 818979217.0, "step": 21568 }, { "epoch": 2.7437984989187125, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00011539459228515625, "grad_norm": 7.300719738006592, "learning_rate": 1e-05, "loss": 0.5274, "mean_token_accuracy": 0.8677058219909668, "num_tokens": 819019991.0, "step": 21569 }, { "epoch": 2.743925709197303, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 5.814599990844727, "learning_rate": 1e-05, "loss": 0.4876, "mean_token_accuracy": 0.8717725276947021, "num_tokens": 819059264.0, "step": 21570 }, { "epoch": 2.7440529194758936, "ewc_loss": 0.1572265625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.0001277923583984375, "grad_norm": 7.744159698486328, "learning_rate": 1e-05, "loss": 0.5188, "mean_token_accuracy": 0.8776217699050903, "num_tokens": 819099649.0, "step": 21571 }, { "epoch": 2.744180129754484, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.075806140899658, "learning_rate": 1e-05, "loss": 0.4912, "mean_token_accuracy": 0.8727965354919434, "num_tokens": 819142434.0, "step": 21572 }, { "epoch": 2.7443073400330746, "ewc_loss": 0.1630859375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.000133514404296875, "grad_norm": 7.996445655822754, "learning_rate": 1e-05, "loss": 0.5766, "mean_token_accuracy": 0.8573509454727173, "num_tokens": 819183908.0, "step": 21573 }, { "epoch": 2.744434550311665, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00011348724365234375, "grad_norm": 6.61680793762207, "learning_rate": 1e-05, "loss": 0.4479, "mean_token_accuracy": 0.8920525908470154, "num_tokens": 819218975.0, "step": 21574 }, { "epoch": 2.7445617605902557, "ewc_loss": 0.1572265625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.0001277923583984375, "grad_norm": 7.716653823852539, "learning_rate": 1e-05, "loss": 0.4991, "mean_token_accuracy": 0.8789536952972412, "num_tokens": 819264427.0, "step": 21575 }, { "epoch": 2.7446889708688462, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00011777877807617188, "grad_norm": 6.863575458526611, "learning_rate": 1e-05, "loss": 0.5299, "mean_token_accuracy": 0.8687769770622253, "num_tokens": 819304947.0, "step": 21576 }, { "epoch": 2.7448161811474368, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00011444091796875, "grad_norm": 7.236356735229492, "learning_rate": 1e-05, "loss": 0.472, "mean_token_accuracy": 0.885824978351593, "num_tokens": 819345540.0, "step": 21577 }, { "epoch": 2.7449433914260273, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.0001125335693359375, "grad_norm": 6.825419902801514, "learning_rate": 1e-05, "loss": 0.4918, "mean_token_accuracy": 0.8768601417541504, "num_tokens": 819383007.0, "step": 21578 }, { "epoch": 2.745070601704618, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.809924602508545, "learning_rate": 1e-05, "loss": 0.5263, "mean_token_accuracy": 0.8694922924041748, "num_tokens": 819415640.0, "step": 21579 }, { "epoch": 2.7451978119832083, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.738572597503662, "learning_rate": 1e-05, "loss": 0.5186, "mean_token_accuracy": 0.8667276501655579, "num_tokens": 819459497.0, "step": 21580 }, { "epoch": 2.745325022261799, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.497724533081055, "learning_rate": 1e-05, "loss": 0.5235, "mean_token_accuracy": 0.8651898503303528, "num_tokens": 819497010.0, "step": 21581 }, { "epoch": 2.7454522325403894, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.692325115203857, "learning_rate": 1e-05, "loss": 0.4906, "mean_token_accuracy": 0.8779657483100891, "num_tokens": 819535659.0, "step": 21582 }, { "epoch": 2.74557944281898, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.282680034637451, "learning_rate": 1e-05, "loss": 0.4843, "mean_token_accuracy": 0.8768348693847656, "num_tokens": 819572774.0, "step": 21583 }, { "epoch": 2.7457066530975704, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.702500343322754, "learning_rate": 1e-05, "loss": 0.4378, "mean_token_accuracy": 0.8904964327812195, "num_tokens": 819607461.0, "step": 21584 }, { "epoch": 2.745833863376161, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.135631561279297, "learning_rate": 1e-05, "loss": 0.4537, "mean_token_accuracy": 0.8832947015762329, "num_tokens": 819645213.0, "step": 21585 }, { "epoch": 2.745961073654751, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.784821510314941, "learning_rate": 1e-05, "loss": 0.4742, "mean_token_accuracy": 0.8806982040405273, "num_tokens": 819679537.0, "step": 21586 }, { "epoch": 2.746088283933342, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 5.985195159912109, "learning_rate": 1e-05, "loss": 0.4582, "mean_token_accuracy": 0.8821849226951599, "num_tokens": 819717414.0, "step": 21587 }, { "epoch": 2.746215494211932, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.97538423538208, "learning_rate": 1e-05, "loss": 0.472, "mean_token_accuracy": 0.8848068714141846, "num_tokens": 819757875.0, "step": 21588 }, { "epoch": 2.746342704490523, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.858805179595947, "learning_rate": 1e-05, "loss": 0.4814, "mean_token_accuracy": 0.8725412487983704, "num_tokens": 819796002.0, "step": 21589 }, { "epoch": 2.746469914769113, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00011587142944335938, "grad_norm": 7.315469741821289, "learning_rate": 1e-05, "loss": 0.551, "mean_token_accuracy": 0.8626998662948608, "num_tokens": 819834736.0, "step": 21590 }, { "epoch": 2.7465971250477037, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 5.826777935028076, "learning_rate": 1e-05, "loss": 0.4306, "mean_token_accuracy": 0.8880141973495483, "num_tokens": 819877566.0, "step": 21591 }, { "epoch": 2.7467243353262942, "ewc_loss": 0.158203125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 0.00012874603271484375, "grad_norm": 7.757747173309326, "learning_rate": 1e-05, "loss": 0.512, "mean_token_accuracy": 0.8743351697921753, "num_tokens": 819914843.0, "step": 21592 }, { "epoch": 2.7468515456048848, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.103629112243652, "learning_rate": 1e-05, "loss": 0.4957, "mean_token_accuracy": 0.8751674294471741, "num_tokens": 819957181.0, "step": 21593 }, { "epoch": 2.7469787558834753, "ewc_loss": 0.1630859375, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.000133514404296875, "grad_norm": 7.9857048988342285, "learning_rate": 1e-05, "loss": 0.5062, "mean_token_accuracy": 0.8798609972000122, "num_tokens": 819991291.0, "step": 21594 }, { "epoch": 2.747105966162066, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00011301040649414062, "grad_norm": 6.630286693572998, "learning_rate": 1e-05, "loss": 0.5015, "mean_token_accuracy": 0.873282253742218, "num_tokens": 820033096.0, "step": 21595 }, { "epoch": 2.7472331764406563, "ewc_loss": 0.1552734375, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.000125885009765625, "grad_norm": 7.708685874938965, "learning_rate": 1e-05, "loss": 0.4756, "mean_token_accuracy": 0.8856123685836792, "num_tokens": 820070200.0, "step": 21596 }, { "epoch": 2.747360386719247, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00011730194091796875, "grad_norm": 6.868635177612305, "learning_rate": 1e-05, "loss": 0.4616, "mean_token_accuracy": 0.8895531296730042, "num_tokens": 820100989.0, "step": 21597 }, { "epoch": 2.7474875969978374, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00011444091796875, "grad_norm": 7.241541862487793, "learning_rate": 1e-05, "loss": 0.5078, "mean_token_accuracy": 0.8743541240692139, "num_tokens": 820145052.0, "step": 21598 }, { "epoch": 2.747614807276428, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00011157989501953125, "grad_norm": 6.820859432220459, "learning_rate": 1e-05, "loss": 0.4545, "mean_token_accuracy": 0.8878346085548401, "num_tokens": 820181463.0, "step": 21599 }, { "epoch": 2.7477420175550185, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.8344526290893555, "learning_rate": 1e-05, "loss": 0.4965, "mean_token_accuracy": 0.8735678195953369, "num_tokens": 820219226.0, "step": 21600 }, { "epoch": 2.747869227833609, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.740645408630371, "learning_rate": 1e-05, "loss": 0.5025, "mean_token_accuracy": 0.8737264275550842, "num_tokens": 820255953.0, "step": 21601 }, { "epoch": 2.7479964381121995, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.510883808135986, "learning_rate": 1e-05, "loss": 0.4925, "mean_token_accuracy": 0.8744975328445435, "num_tokens": 820296587.0, "step": 21602 }, { "epoch": 2.74812364839079, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.686663627624512, "learning_rate": 1e-05, "loss": 0.4805, "mean_token_accuracy": 0.8801572322845459, "num_tokens": 820333815.0, "step": 21603 }, { "epoch": 2.7482508586693806, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.31051778793335, "learning_rate": 1e-05, "loss": 0.473, "mean_token_accuracy": 0.8777455687522888, "num_tokens": 820376401.0, "step": 21604 }, { "epoch": 2.748378068947971, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.661181926727295, "learning_rate": 1e-05, "loss": 0.4664, "mean_token_accuracy": 0.8809885382652283, "num_tokens": 820417355.0, "step": 21605 }, { "epoch": 2.7485052792265616, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.181386470794678, "learning_rate": 1e-05, "loss": 0.3889, "mean_token_accuracy": 0.9042038917541504, "num_tokens": 820456788.0, "step": 21606 }, { "epoch": 2.748632489505152, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.690086364746094, "learning_rate": 1e-05, "loss": 0.4364, "mean_token_accuracy": 0.8923143148422241, "num_tokens": 820499767.0, "step": 21607 }, { "epoch": 2.7487596997837427, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.088406085968018, "learning_rate": 1e-05, "loss": 0.4538, "mean_token_accuracy": 0.8837073445320129, "num_tokens": 820543052.0, "step": 21608 }, { "epoch": 2.7488869100623328, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.797156810760498, "learning_rate": 1e-05, "loss": 0.4483, "mean_token_accuracy": 0.8902978897094727, "num_tokens": 820586162.0, "step": 21609 }, { "epoch": 2.7490141203409237, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 5.966959476470947, "learning_rate": 1e-05, "loss": 0.5049, "mean_token_accuracy": 0.8698829412460327, "num_tokens": 820626703.0, "step": 21610 }, { "epoch": 2.749141330619514, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 7.022058010101318, "learning_rate": 1e-05, "loss": 0.5215, "mean_token_accuracy": 0.8656143546104431, "num_tokens": 820663936.0, "step": 21611 }, { "epoch": 2.749268540898105, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.81818962097168, "learning_rate": 1e-05, "loss": 0.4596, "mean_token_accuracy": 0.8812172412872314, "num_tokens": 820703761.0, "step": 21612 }, { "epoch": 2.749395751176695, "ewc_loss": 0.1474609375, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.000118255615234375, "grad_norm": 7.41451358795166, "learning_rate": 1e-05, "loss": 0.5298, "mean_token_accuracy": 0.8693451881408691, "num_tokens": 820742505.0, "step": 21613 }, { "epoch": 2.749522961455286, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 5.820699691772461, "learning_rate": 1e-05, "loss": 0.4372, "mean_token_accuracy": 0.8887773752212524, "num_tokens": 820784185.0, "step": 21614 }, { "epoch": 2.749650171733876, "ewc_loss": 0.162109375, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00013256072998046875, "grad_norm": 7.88817024230957, "learning_rate": 1e-05, "loss": 0.5603, "mean_token_accuracy": 0.859688401222229, "num_tokens": 820826163.0, "step": 21615 }, { "epoch": 2.7497773820124665, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.205934047698975, "learning_rate": 1e-05, "loss": 0.4916, "mean_token_accuracy": 0.872252881526947, "num_tokens": 820861598.0, "step": 21616 }, { "epoch": 2.749904592291057, "ewc_loss": 0.1650390625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.0001354217529296875, "grad_norm": 8.046966552734375, "learning_rate": 1e-05, "loss": 0.5347, "mean_token_accuracy": 0.8734984397888184, "num_tokens": 820901201.0, "step": 21617 }, { "epoch": 2.7500318025696475, "ewc_loss": 0.1474609375, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00011777877807617188, "grad_norm": 6.751939296722412, "learning_rate": 1e-05, "loss": 0.4767, "mean_token_accuracy": 0.886406660079956, "num_tokens": 820941773.0, "step": 21618 }, { "epoch": 2.750159012848238, "ewc_loss": 0.1533203125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.0001239776611328125, "grad_norm": 7.6799421310424805, "learning_rate": 1e-05, "loss": 0.5781, "mean_token_accuracy": 0.8545869588851929, "num_tokens": 820984610.0, "step": 21619 }, { "epoch": 2.7502862231268286, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00011730194091796875, "grad_norm": 6.91379976272583, "learning_rate": 1e-05, "loss": 0.4613, "mean_token_accuracy": 0.8886710405349731, "num_tokens": 821020814.0, "step": 21620 }, { "epoch": 2.750413433405419, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00011301040649414062, "grad_norm": 7.202554702758789, "learning_rate": 1e-05, "loss": 0.4599, "mean_token_accuracy": 0.8862472772598267, "num_tokens": 821053073.0, "step": 21621 }, { "epoch": 2.7505406436840096, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.0001125335693359375, "grad_norm": 6.828177452087402, "learning_rate": 1e-05, "loss": 0.5, "mean_token_accuracy": 0.8724032044410706, "num_tokens": 821088872.0, "step": 21622 }, { "epoch": 2.7506678539626, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.797102928161621, "learning_rate": 1e-05, "loss": 0.5115, "mean_token_accuracy": 0.8683935403823853, "num_tokens": 821124314.0, "step": 21623 }, { "epoch": 2.7507950642411907, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.745931625366211, "learning_rate": 1e-05, "loss": 0.4837, "mean_token_accuracy": 0.8759822249412537, "num_tokens": 821152025.0, "step": 21624 }, { "epoch": 2.750922274519781, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.503786087036133, "learning_rate": 1e-05, "loss": 0.5109, "mean_token_accuracy": 0.865842342376709, "num_tokens": 821198138.0, "step": 21625 }, { "epoch": 2.7510494847983717, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.669503688812256, "learning_rate": 1e-05, "loss": 0.4854, "mean_token_accuracy": 0.8770416975021362, "num_tokens": 821241256.0, "step": 21626 }, { "epoch": 2.7511766950769623, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.331864833831787, "learning_rate": 1e-05, "loss": 0.4641, "mean_token_accuracy": 0.8847258687019348, "num_tokens": 821274932.0, "step": 21627 }, { "epoch": 2.751303905355553, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.63665246963501, "learning_rate": 1e-05, "loss": 0.4782, "mean_token_accuracy": 0.8779746294021606, "num_tokens": 821310769.0, "step": 21628 }, { "epoch": 2.7514311156341433, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.206271648406982, "learning_rate": 1e-05, "loss": 0.467, "mean_token_accuracy": 0.8773562908172607, "num_tokens": 821346474.0, "step": 21629 }, { "epoch": 2.751558325912734, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.667686939239502, "learning_rate": 1e-05, "loss": 0.4848, "mean_token_accuracy": 0.8778519630432129, "num_tokens": 821385293.0, "step": 21630 }, { "epoch": 2.7516855361913244, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.089608192443848, "learning_rate": 1e-05, "loss": 0.4369, "mean_token_accuracy": 0.8884186148643494, "num_tokens": 821421161.0, "step": 21631 }, { "epoch": 2.751812746469915, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.772274494171143, "learning_rate": 1e-05, "loss": 0.4634, "mean_token_accuracy": 0.8873298764228821, "num_tokens": 821459937.0, "step": 21632 }, { "epoch": 2.7519399567485054, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 5.948282718658447, "learning_rate": 1e-05, "loss": 0.4639, "mean_token_accuracy": 0.879399299621582, "num_tokens": 821503892.0, "step": 21633 }, { "epoch": 2.7520671670270955, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 7.010739803314209, "learning_rate": 1e-05, "loss": 0.4754, "mean_token_accuracy": 0.8833486437797546, "num_tokens": 821537781.0, "step": 21634 }, { "epoch": 2.7521943773056865, "ewc_loss": 0.11474609375, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.812600612640381, "learning_rate": 1e-05, "loss": 0.5014, "mean_token_accuracy": 0.8638790845870972, "num_tokens": 821576114.0, "step": 21635 }, { "epoch": 2.7523215875842766, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00011730194091796875, "grad_norm": 7.431458473205566, "learning_rate": 1e-05, "loss": 0.4933, "mean_token_accuracy": 0.882868766784668, "num_tokens": 821613587.0, "step": 21636 }, { "epoch": 2.7524487978628676, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 5.800901412963867, "learning_rate": 1e-05, "loss": 0.392, "mean_token_accuracy": 0.9006881713867188, "num_tokens": 821650956.0, "step": 21637 }, { "epoch": 2.7525760081414576, "ewc_loss": 0.162109375, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00013256072998046875, "grad_norm": 7.925752639770508, "learning_rate": 1e-05, "loss": 0.5429, "mean_token_accuracy": 0.8689683079719543, "num_tokens": 821692943.0, "step": 21638 }, { "epoch": 2.7527032184200486, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.219714641571045, "learning_rate": 1e-05, "loss": 0.532, "mean_token_accuracy": 0.8615701198577881, "num_tokens": 821735344.0, "step": 21639 }, { "epoch": 2.7528304286986387, "ewc_loss": 0.1650390625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.0001354217529296875, "grad_norm": 8.09066104888916, "learning_rate": 1e-05, "loss": 0.5164, "mean_token_accuracy": 0.8800634741783142, "num_tokens": 821772910.0, "step": 21640 }, { "epoch": 2.7529576389772292, "ewc_loss": 0.1484375, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00011920928955078125, "grad_norm": 6.794497966766357, "learning_rate": 1e-05, "loss": 0.4788, "mean_token_accuracy": 0.8844008445739746, "num_tokens": 821810211.0, "step": 21641 }, { "epoch": 2.7530848492558198, "ewc_loss": 0.154296875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00012493133544921875, "grad_norm": 7.707740783691406, "learning_rate": 1e-05, "loss": 0.4991, "mean_token_accuracy": 0.8764925599098206, "num_tokens": 821840782.0, "step": 21642 }, { "epoch": 2.7532120595344103, "ewc_loss": 0.1474609375, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00011777877807617188, "grad_norm": 6.947999000549316, "learning_rate": 1e-05, "loss": 0.5059, "mean_token_accuracy": 0.8752217292785645, "num_tokens": 821879160.0, "step": 21643 }, { "epoch": 2.753339269813001, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00011444091796875, "grad_norm": 7.2278947830200195, "learning_rate": 1e-05, "loss": 0.4738, "mean_token_accuracy": 0.8875821828842163, "num_tokens": 821915344.0, "step": 21644 }, { "epoch": 2.7534664800915913, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00011301040649414062, "grad_norm": 6.837711811065674, "learning_rate": 1e-05, "loss": 0.5484, "mean_token_accuracy": 0.8560227155685425, "num_tokens": 821955628.0, "step": 21645 }, { "epoch": 2.753593690370182, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.840270519256592, "learning_rate": 1e-05, "loss": 0.5013, "mean_token_accuracy": 0.8725046515464783, "num_tokens": 821996917.0, "step": 21646 }, { "epoch": 2.7537209006487724, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.7120442390441895, "learning_rate": 1e-05, "loss": 0.4752, "mean_token_accuracy": 0.8795503973960876, "num_tokens": 822034222.0, "step": 21647 }, { "epoch": 2.753848110927363, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.568238735198975, "learning_rate": 1e-05, "loss": 0.4851, "mean_token_accuracy": 0.8748370409011841, "num_tokens": 822072350.0, "step": 21648 }, { "epoch": 2.7539753212059535, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.62708044052124, "learning_rate": 1e-05, "loss": 0.4938, "mean_token_accuracy": 0.8746094703674316, "num_tokens": 822111495.0, "step": 21649 }, { "epoch": 2.754102531484544, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.4027934074401855, "learning_rate": 1e-05, "loss": 0.4817, "mean_token_accuracy": 0.8754268288612366, "num_tokens": 822148132.0, "step": 21650 }, { "epoch": 2.7542297417631345, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.556211948394775, "learning_rate": 1e-05, "loss": 0.482, "mean_token_accuracy": 0.878771185874939, "num_tokens": 822187440.0, "step": 21651 }, { "epoch": 2.754356952041725, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.297844409942627, "learning_rate": 1e-05, "loss": 0.469, "mean_token_accuracy": 0.8786486387252808, "num_tokens": 822225936.0, "step": 21652 }, { "epoch": 2.7544841623203156, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.5327887535095215, "learning_rate": 1e-05, "loss": 0.4912, "mean_token_accuracy": 0.8744414448738098, "num_tokens": 822266187.0, "step": 21653 }, { "epoch": 2.754611372598906, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.201920509338379, "learning_rate": 1e-05, "loss": 0.4464, "mean_token_accuracy": 0.8853176832199097, "num_tokens": 822302963.0, "step": 21654 }, { "epoch": 2.7547385828774966, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.559933185577393, "learning_rate": 1e-05, "loss": 0.4476, "mean_token_accuracy": 0.887398362159729, "num_tokens": 822336969.0, "step": 21655 }, { "epoch": 2.754865793156087, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.10610294342041, "learning_rate": 1e-05, "loss": 0.4711, "mean_token_accuracy": 0.8763163089752197, "num_tokens": 822376847.0, "step": 21656 }, { "epoch": 2.7549930034346777, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.651241302490234, "learning_rate": 1e-05, "loss": 0.5001, "mean_token_accuracy": 0.8717349767684937, "num_tokens": 822417858.0, "step": 21657 }, { "epoch": 2.755120213713268, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 5.957603931427002, "learning_rate": 1e-05, "loss": 0.5101, "mean_token_accuracy": 0.8671112060546875, "num_tokens": 822459946.0, "step": 21658 }, { "epoch": 2.7552474239918583, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.892677307128906, "learning_rate": 1e-05, "loss": 0.5108, "mean_token_accuracy": 0.8711413145065308, "num_tokens": 822498093.0, "step": 21659 }, { "epoch": 2.7553746342704493, "ewc_loss": 0.11376953125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.767318248748779, "learning_rate": 1e-05, "loss": 0.4437, "mean_token_accuracy": 0.8837674260139465, "num_tokens": 822535918.0, "step": 21660 }, { "epoch": 2.7555018445490393, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00011682510375976562, "grad_norm": 7.334993839263916, "learning_rate": 1e-05, "loss": 0.4987, "mean_token_accuracy": 0.8748825192451477, "num_tokens": 822574589.0, "step": 21661 }, { "epoch": 2.7556290548276303, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.664982795715332, "learning_rate": 1e-05, "loss": 0.4374, "mean_token_accuracy": 0.882919430732727, "num_tokens": 822612437.0, "step": 21662 }, { "epoch": 2.7557562651062204, "ewc_loss": 0.1630859375, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.000133514404296875, "grad_norm": 7.984928131103516, "learning_rate": 1e-05, "loss": 0.5202, "mean_token_accuracy": 0.8791898488998413, "num_tokens": 822654834.0, "step": 21663 }, { "epoch": 2.755883475384811, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.071039199829102, "learning_rate": 1e-05, "loss": 0.4552, "mean_token_accuracy": 0.8862996697425842, "num_tokens": 822689555.0, "step": 21664 }, { "epoch": 2.7560106856634015, "ewc_loss": 0.1708984375, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.000141143798828125, "grad_norm": 8.378226280212402, "learning_rate": 1e-05, "loss": 0.5734, "mean_token_accuracy": 0.8624261021614075, "num_tokens": 822731609.0, "step": 21665 }, { "epoch": 2.756137895941992, "ewc_loss": 0.1513671875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.0001220703125, "grad_norm": 6.915805816650391, "learning_rate": 1e-05, "loss": 0.5223, "mean_token_accuracy": 0.8683434724807739, "num_tokens": 822769881.0, "step": 21666 }, { "epoch": 2.7562651062205825, "ewc_loss": 0.16015625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00013065338134765625, "grad_norm": 7.960363864898682, "learning_rate": 1e-05, "loss": 0.5582, "mean_token_accuracy": 0.8648451566696167, "num_tokens": 822810369.0, "step": 21667 }, { "epoch": 2.756392316499173, "ewc_loss": 0.15234375, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.163457870483398, "learning_rate": 1e-05, "loss": 0.5162, "mean_token_accuracy": 0.8744550943374634, "num_tokens": 822852476.0, "step": 21668 }, { "epoch": 2.7565195267777636, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00011730194091796875, "grad_norm": 7.4214324951171875, "learning_rate": 1e-05, "loss": 0.5516, "mean_token_accuracy": 0.8618513345718384, "num_tokens": 822890745.0, "step": 21669 }, { "epoch": 2.756646737056354, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00011396408081054688, "grad_norm": 6.955687999725342, "learning_rate": 1e-05, "loss": 0.4914, "mean_token_accuracy": 0.876402735710144, "num_tokens": 822931243.0, "step": 21670 }, { "epoch": 2.7567739473349446, "ewc_loss": 0.140625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 7.041772842407227, "learning_rate": 1e-05, "loss": 0.4908, "mean_token_accuracy": 0.8762038946151733, "num_tokens": 822967560.0, "step": 21671 }, { "epoch": 2.756901157613535, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.748459815979004, "learning_rate": 1e-05, "loss": 0.4827, "mean_token_accuracy": 0.8801186084747314, "num_tokens": 823007129.0, "step": 21672 }, { "epoch": 2.7570283678921257, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.7906270027160645, "learning_rate": 1e-05, "loss": 0.4997, "mean_token_accuracy": 0.8727681636810303, "num_tokens": 823043722.0, "step": 21673 }, { "epoch": 2.757155578170716, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.574880123138428, "learning_rate": 1e-05, "loss": 0.5417, "mean_token_accuracy": 0.8602581024169922, "num_tokens": 823079937.0, "step": 21674 }, { "epoch": 2.7572827884493067, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.648929119110107, "learning_rate": 1e-05, "loss": 0.4948, "mean_token_accuracy": 0.8737415671348572, "num_tokens": 823120455.0, "step": 21675 }, { "epoch": 2.7574099987278973, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.437308311462402, "learning_rate": 1e-05, "loss": 0.4981, "mean_token_accuracy": 0.8706936836242676, "num_tokens": 823151415.0, "step": 21676 }, { "epoch": 2.757537209006488, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.601933002471924, "learning_rate": 1e-05, "loss": 0.4849, "mean_token_accuracy": 0.8782640695571899, "num_tokens": 823188281.0, "step": 21677 }, { "epoch": 2.7576644192850783, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.275121688842773, "learning_rate": 1e-05, "loss": 0.4475, "mean_token_accuracy": 0.8808803558349609, "num_tokens": 823226040.0, "step": 21678 }, { "epoch": 2.757791629563669, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.6541032791137695, "learning_rate": 1e-05, "loss": 0.4877, "mean_token_accuracy": 0.8735549449920654, "num_tokens": 823261246.0, "step": 21679 }, { "epoch": 2.7579188398422594, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.098480224609375, "learning_rate": 1e-05, "loss": 0.4721, "mean_token_accuracy": 0.8790997266769409, "num_tokens": 823300034.0, "step": 21680 }, { "epoch": 2.75804605012085, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.834568977355957, "learning_rate": 1e-05, "loss": 0.4317, "mean_token_accuracy": 0.8923230171203613, "num_tokens": 823332152.0, "step": 21681 }, { "epoch": 2.7581732603994404, "ewc_loss": 0.11474609375, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.900146484375, "learning_rate": 1e-05, "loss": 0.4674, "mean_token_accuracy": 0.8768875002861023, "num_tokens": 823373319.0, "step": 21682 }, { "epoch": 2.758300470678031, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.0001125335693359375, "grad_norm": 7.188703536987305, "learning_rate": 1e-05, "loss": 0.4491, "mean_token_accuracy": 0.8915867209434509, "num_tokens": 823409875.0, "step": 21683 }, { "epoch": 2.758427680956621, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.75649881362915, "learning_rate": 1e-05, "loss": 0.4241, "mean_token_accuracy": 0.8920677304267883, "num_tokens": 823450941.0, "step": 21684 }, { "epoch": 2.758554891235212, "ewc_loss": 0.15625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00012683868408203125, "grad_norm": 7.754942417144775, "learning_rate": 1e-05, "loss": 0.4808, "mean_token_accuracy": 0.8843859434127808, "num_tokens": 823493483.0, "step": 21685 }, { "epoch": 2.758682101513802, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 5.9833526611328125, "learning_rate": 1e-05, "loss": 0.4615, "mean_token_accuracy": 0.8813633918762207, "num_tokens": 823531682.0, "step": 21686 }, { "epoch": 2.758809311792393, "ewc_loss": 0.166015625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00013637542724609375, "grad_norm": 8.216658592224121, "learning_rate": 1e-05, "loss": 0.494, "mean_token_accuracy": 0.8863672018051147, "num_tokens": 823564089.0, "step": 21687 }, { "epoch": 2.758936522070983, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00011539459228515625, "grad_norm": 6.7105488777160645, "learning_rate": 1e-05, "loss": 0.437, "mean_token_accuracy": 0.8955889344215393, "num_tokens": 823598700.0, "step": 21688 }, { "epoch": 2.7590637323495737, "ewc_loss": 0.158203125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00012874603271484375, "grad_norm": 7.918009281158447, "learning_rate": 1e-05, "loss": 0.5188, "mean_token_accuracy": 0.8720677495002747, "num_tokens": 823640248.0, "step": 21689 }, { "epoch": 2.7591909426281642, "ewc_loss": 0.150390625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00012063980102539062, "grad_norm": 7.091898441314697, "learning_rate": 1e-05, "loss": 0.5063, "mean_token_accuracy": 0.872362494468689, "num_tokens": 823676303.0, "step": 21690 }, { "epoch": 2.7593181529067548, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00011396408081054688, "grad_norm": 7.332923412322998, "learning_rate": 1e-05, "loss": 0.5192, "mean_token_accuracy": 0.8702603578567505, "num_tokens": 823717736.0, "step": 21691 }, { "epoch": 2.7594453631853453, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00011444091796875, "grad_norm": 7.002619743347168, "learning_rate": 1e-05, "loss": 0.5033, "mean_token_accuracy": 0.8739570379257202, "num_tokens": 823755602.0, "step": 21692 }, { "epoch": 2.759572573463936, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.847168922424316, "learning_rate": 1e-05, "loss": 0.5313, "mean_token_accuracy": 0.8634772300720215, "num_tokens": 823792257.0, "step": 21693 }, { "epoch": 2.7596997837425263, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 6.875779628753662, "learning_rate": 1e-05, "loss": 0.487, "mean_token_accuracy": 0.8779590129852295, "num_tokens": 823835655.0, "step": 21694 }, { "epoch": 2.759826994021117, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.518784046173096, "learning_rate": 1e-05, "loss": 0.4879, "mean_token_accuracy": 0.8711550235748291, "num_tokens": 823874840.0, "step": 21695 }, { "epoch": 2.7599542042997074, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.824044704437256, "learning_rate": 1e-05, "loss": 0.5189, "mean_token_accuracy": 0.8680050373077393, "num_tokens": 823910844.0, "step": 21696 }, { "epoch": 2.760081414578298, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.288457870483398, "learning_rate": 1e-05, "loss": 0.4363, "mean_token_accuracy": 0.8889659643173218, "num_tokens": 823948472.0, "step": 21697 }, { "epoch": 2.7602086248568884, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.839211463928223, "learning_rate": 1e-05, "loss": 0.4558, "mean_token_accuracy": 0.8877862691879272, "num_tokens": 823997532.0, "step": 21698 }, { "epoch": 2.760335835135479, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.1375908851623535, "learning_rate": 1e-05, "loss": 0.4938, "mean_token_accuracy": 0.8685389161109924, "num_tokens": 824037426.0, "step": 21699 }, { "epoch": 2.7604630454140695, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.942342758178711, "learning_rate": 1e-05, "loss": 0.5132, "mean_token_accuracy": 0.8715047836303711, "num_tokens": 824076731.0, "step": 21700 }, { "epoch": 2.76059025569266, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 6.011597633361816, "learning_rate": 1e-05, "loss": 0.5201, "mean_token_accuracy": 0.8612581491470337, "num_tokens": 824119445.0, "step": 21701 }, { "epoch": 2.7607174659712506, "ewc_loss": 0.140625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00011110305786132812, "grad_norm": 7.146979808807373, "learning_rate": 1e-05, "loss": 0.5275, "mean_token_accuracy": 0.8664514422416687, "num_tokens": 824157372.0, "step": 21702 }, { "epoch": 2.760844676249841, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 5.917576789855957, "learning_rate": 1e-05, "loss": 0.4712, "mean_token_accuracy": 0.8766369819641113, "num_tokens": 824204312.0, "step": 21703 }, { "epoch": 2.7609718865284316, "ewc_loss": 0.1474609375, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00011777877807617188, "grad_norm": 7.464763164520264, "learning_rate": 1e-05, "loss": 0.4779, "mean_token_accuracy": 0.8833874464035034, "num_tokens": 824245951.0, "step": 21704 }, { "epoch": 2.761099096807022, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 5.988246440887451, "learning_rate": 1e-05, "loss": 0.4768, "mean_token_accuracy": 0.8764389753341675, "num_tokens": 824288742.0, "step": 21705 }, { "epoch": 2.7612263070856127, "ewc_loss": 0.15625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00012683868408203125, "grad_norm": 7.7711896896362305, "learning_rate": 1e-05, "loss": 0.5181, "mean_token_accuracy": 0.8712178468704224, "num_tokens": 824326435.0, "step": 21706 }, { "epoch": 2.7613535173642028, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.322171688079834, "learning_rate": 1e-05, "loss": 0.473, "mean_token_accuracy": 0.8771286606788635, "num_tokens": 824359442.0, "step": 21707 }, { "epoch": 2.7614807276427937, "ewc_loss": 0.154296875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00012493133544921875, "grad_norm": 7.766307353973389, "learning_rate": 1e-05, "loss": 0.5396, "mean_token_accuracy": 0.8664498329162598, "num_tokens": 824392387.0, "step": 21708 }, { "epoch": 2.761607937921384, "ewc_loss": 0.140625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 6.6524658203125, "learning_rate": 1e-05, "loss": 0.5091, "mean_token_accuracy": 0.8721981644630432, "num_tokens": 824427860.0, "step": 21709 }, { "epoch": 2.761735148199975, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00011682510375976562, "grad_norm": 7.425562381744385, "learning_rate": 1e-05, "loss": 0.4986, "mean_token_accuracy": 0.8750206232070923, "num_tokens": 824466011.0, "step": 21710 }, { "epoch": 2.761862358478565, "ewc_loss": 0.140625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 6.742647647857666, "learning_rate": 1e-05, "loss": 0.4509, "mean_token_accuracy": 0.8887711763381958, "num_tokens": 824507235.0, "step": 21711 }, { "epoch": 2.761989568757156, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 7.003512859344482, "learning_rate": 1e-05, "loss": 0.4788, "mean_token_accuracy": 0.8819336891174316, "num_tokens": 824538807.0, "step": 21712 }, { "epoch": 2.762116779035746, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.714534759521484, "learning_rate": 1e-05, "loss": 0.4937, "mean_token_accuracy": 0.8767633438110352, "num_tokens": 824575047.0, "step": 21713 }, { "epoch": 2.7622439893143365, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.639886856079102, "learning_rate": 1e-05, "loss": 0.4222, "mean_token_accuracy": 0.8955792188644409, "num_tokens": 824604113.0, "step": 21714 }, { "epoch": 2.762371199592927, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.667415618896484, "learning_rate": 1e-05, "loss": 0.5183, "mean_token_accuracy": 0.864783763885498, "num_tokens": 824640912.0, "step": 21715 }, { "epoch": 2.7624984098715175, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.388919830322266, "learning_rate": 1e-05, "loss": 0.4617, "mean_token_accuracy": 0.8819398283958435, "num_tokens": 824682739.0, "step": 21716 }, { "epoch": 2.762625620150108, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.65177583694458, "learning_rate": 1e-05, "loss": 0.5402, "mean_token_accuracy": 0.8583505153656006, "num_tokens": 824719973.0, "step": 21717 }, { "epoch": 2.7627528304286986, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.245432376861572, "learning_rate": 1e-05, "loss": 0.4616, "mean_token_accuracy": 0.8808465600013733, "num_tokens": 824761214.0, "step": 21718 }, { "epoch": 2.762880040707289, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.637918472290039, "learning_rate": 1e-05, "loss": 0.5054, "mean_token_accuracy": 0.8670628070831299, "num_tokens": 824801114.0, "step": 21719 }, { "epoch": 2.7630072509858796, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.146787166595459, "learning_rate": 1e-05, "loss": 0.489, "mean_token_accuracy": 0.8727293014526367, "num_tokens": 824832734.0, "step": 21720 }, { "epoch": 2.76313446126447, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.680975914001465, "learning_rate": 1e-05, "loss": 0.4932, "mean_token_accuracy": 0.8701691627502441, "num_tokens": 824867089.0, "step": 21721 }, { "epoch": 2.7632616715430607, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.03251838684082, "learning_rate": 1e-05, "loss": 0.4306, "mean_token_accuracy": 0.8892508149147034, "num_tokens": 824909173.0, "step": 21722 }, { "epoch": 2.763388881821651, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.818124294281006, "learning_rate": 1e-05, "loss": 0.4569, "mean_token_accuracy": 0.8847293853759766, "num_tokens": 824943815.0, "step": 21723 }, { "epoch": 2.7635160921002417, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.889119625091553, "learning_rate": 1e-05, "loss": 0.4214, "mean_token_accuracy": 0.892054557800293, "num_tokens": 824978708.0, "step": 21724 }, { "epoch": 2.7636433023788323, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00011157989501953125, "grad_norm": 7.126570224761963, "learning_rate": 1e-05, "loss": 0.5591, "mean_token_accuracy": 0.8614628314971924, "num_tokens": 825015226.0, "step": 21725 }, { "epoch": 2.763770512657423, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.785068035125732, "learning_rate": 1e-05, "loss": 0.4918, "mean_token_accuracy": 0.8678199648857117, "num_tokens": 825054417.0, "step": 21726 }, { "epoch": 2.7638977229360133, "ewc_loss": 0.15234375, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.559570789337158, "learning_rate": 1e-05, "loss": 0.4789, "mean_token_accuracy": 0.8908405303955078, "num_tokens": 825093757.0, "step": 21727 }, { "epoch": 2.764024933214604, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 5.895693302154541, "learning_rate": 1e-05, "loss": 0.5321, "mean_token_accuracy": 0.8606619834899902, "num_tokens": 825129955.0, "step": 21728 }, { "epoch": 2.7641521434931944, "ewc_loss": 0.162109375, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00013256072998046875, "grad_norm": 7.987295150756836, "learning_rate": 1e-05, "loss": 0.4818, "mean_token_accuracy": 0.8868871927261353, "num_tokens": 825168426.0, "step": 21729 }, { "epoch": 2.764279353771785, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 6.373169422149658, "learning_rate": 1e-05, "loss": 0.5341, "mean_token_accuracy": 0.8632339835166931, "num_tokens": 825213475.0, "step": 21730 }, { "epoch": 2.7644065640503754, "ewc_loss": 0.1611328125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.0001316070556640625, "grad_norm": 7.987961769104004, "learning_rate": 1e-05, "loss": 0.4708, "mean_token_accuracy": 0.887231171131134, "num_tokens": 825249146.0, "step": 21731 }, { "epoch": 2.7645337743289655, "ewc_loss": 0.1484375, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00011873245239257812, "grad_norm": 6.850461959838867, "learning_rate": 1e-05, "loss": 0.4487, "mean_token_accuracy": 0.8952105641365051, "num_tokens": 825286504.0, "step": 21732 }, { "epoch": 2.7646609846075565, "ewc_loss": 0.15234375, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.555355072021484, "learning_rate": 1e-05, "loss": 0.5133, "mean_token_accuracy": 0.8724969029426575, "num_tokens": 825327820.0, "step": 21733 }, { "epoch": 2.7647881948861466, "ewc_loss": 0.1455078125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00011587142944335938, "grad_norm": 6.910808086395264, "learning_rate": 1e-05, "loss": 0.4651, "mean_token_accuracy": 0.8877911567687988, "num_tokens": 825365967.0, "step": 21734 }, { "epoch": 2.7649154051647375, "ewc_loss": 0.140625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 7.0990495681762695, "learning_rate": 1e-05, "loss": 0.4982, "mean_token_accuracy": 0.8760131597518921, "num_tokens": 825406959.0, "step": 21735 }, { "epoch": 2.7650426154433276, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00011014938354492188, "grad_norm": 6.802719593048096, "learning_rate": 1e-05, "loss": 0.5276, "mean_token_accuracy": 0.8650453090667725, "num_tokens": 825447355.0, "step": 21736 }, { "epoch": 2.7651698257219186, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.730042457580566, "learning_rate": 1e-05, "loss": 0.5221, "mean_token_accuracy": 0.8646984696388245, "num_tokens": 825485618.0, "step": 21737 }, { "epoch": 2.7652970360005087, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.699143409729004, "learning_rate": 1e-05, "loss": 0.4756, "mean_token_accuracy": 0.8831452131271362, "num_tokens": 825522287.0, "step": 21738 }, { "epoch": 2.765424246279099, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.479241847991943, "learning_rate": 1e-05, "loss": 0.4774, "mean_token_accuracy": 0.8776500225067139, "num_tokens": 825566214.0, "step": 21739 }, { "epoch": 2.7655514565576897, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.629655361175537, "learning_rate": 1e-05, "loss": 0.4782, "mean_token_accuracy": 0.8773735165596008, "num_tokens": 825608169.0, "step": 21740 }, { "epoch": 2.7656786668362803, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.334335803985596, "learning_rate": 1e-05, "loss": 0.4624, "mean_token_accuracy": 0.882763147354126, "num_tokens": 825643076.0, "step": 21741 }, { "epoch": 2.765805877114871, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.612077236175537, "learning_rate": 1e-05, "loss": 0.4794, "mean_token_accuracy": 0.8762882947921753, "num_tokens": 825679868.0, "step": 21742 }, { "epoch": 2.7659330873934613, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.2111077308654785, "learning_rate": 1e-05, "loss": 0.4683, "mean_token_accuracy": 0.8755395412445068, "num_tokens": 825719391.0, "step": 21743 }, { "epoch": 2.766060297672052, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.632105350494385, "learning_rate": 1e-05, "loss": 0.4803, "mean_token_accuracy": 0.876147985458374, "num_tokens": 825761972.0, "step": 21744 }, { "epoch": 2.7661875079506424, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.098076820373535, "learning_rate": 1e-05, "loss": 0.48, "mean_token_accuracy": 0.8742189407348633, "num_tokens": 825802672.0, "step": 21745 }, { "epoch": 2.766314718229233, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.726897239685059, "learning_rate": 1e-05, "loss": 0.486, "mean_token_accuracy": 0.8728823065757751, "num_tokens": 825840219.0, "step": 21746 }, { "epoch": 2.7664419285078234, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 5.962104797363281, "learning_rate": 1e-05, "loss": 0.4678, "mean_token_accuracy": 0.8765488266944885, "num_tokens": 825876793.0, "step": 21747 }, { "epoch": 2.766569138786414, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 6.967311382293701, "learning_rate": 1e-05, "loss": 0.5121, "mean_token_accuracy": 0.8706996440887451, "num_tokens": 825913231.0, "step": 21748 }, { "epoch": 2.7666963490650045, "ewc_loss": 0.1142578125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.8020734786987305, "learning_rate": 1e-05, "loss": 0.502, "mean_token_accuracy": 0.8649888038635254, "num_tokens": 825952600.0, "step": 21749 }, { "epoch": 2.766823559343595, "ewc_loss": 0.1455078125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.0001163482666015625, "grad_norm": 7.379761219024658, "learning_rate": 1e-05, "loss": 0.5167, "mean_token_accuracy": 0.8753246068954468, "num_tokens": 825997239.0, "step": 21750 }, { "epoch": 2.7669507696221856, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 5.751978874206543, "learning_rate": 1e-05, "loss": 0.4511, "mean_token_accuracy": 0.8823685646057129, "num_tokens": 826037492.0, "step": 21751 }, { "epoch": 2.767077979900776, "ewc_loss": 0.162109375, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00013256072998046875, "grad_norm": 7.941464424133301, "learning_rate": 1e-05, "loss": 0.5093, "mean_token_accuracy": 0.879063606262207, "num_tokens": 826077607.0, "step": 21752 }, { "epoch": 2.7672051901793666, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.147683143615723, "learning_rate": 1e-05, "loss": 0.4992, "mean_token_accuracy": 0.8714260458946228, "num_tokens": 826115503.0, "step": 21753 }, { "epoch": 2.767332400457957, "ewc_loss": 0.166015625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00013637542724609375, "grad_norm": 8.188006401062012, "learning_rate": 1e-05, "loss": 0.53, "mean_token_accuracy": 0.8720829486846924, "num_tokens": 826154526.0, "step": 21754 }, { "epoch": 2.7674596107365477, "ewc_loss": 0.1484375, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00011920928955078125, "grad_norm": 6.835651397705078, "learning_rate": 1e-05, "loss": 0.5417, "mean_token_accuracy": 0.8670971989631653, "num_tokens": 826194945.0, "step": 21755 }, { "epoch": 2.767586821015138, "ewc_loss": 0.15625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00012683868408203125, "grad_norm": 7.792614459991455, "learning_rate": 1e-05, "loss": 0.498, "mean_token_accuracy": 0.8797807097434998, "num_tokens": 826231000.0, "step": 21756 }, { "epoch": 2.7677140312937283, "ewc_loss": 0.1494140625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.0001201629638671875, "grad_norm": 7.012023448944092, "learning_rate": 1e-05, "loss": 0.4708, "mean_token_accuracy": 0.8879128694534302, "num_tokens": 826266453.0, "step": 21757 }, { "epoch": 2.7678412415723193, "ewc_loss": 0.1455078125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00011587142944335938, "grad_norm": 7.299530029296875, "learning_rate": 1e-05, "loss": 0.4777, "mean_token_accuracy": 0.8815118074417114, "num_tokens": 826310442.0, "step": 21758 }, { "epoch": 2.7679684518509093, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00011348724365234375, "grad_norm": 6.875072956085205, "learning_rate": 1e-05, "loss": 0.511, "mean_token_accuracy": 0.876673698425293, "num_tokens": 826344595.0, "step": 21759 }, { "epoch": 2.7680956621295003, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.915618419647217, "learning_rate": 1e-05, "loss": 0.4972, "mean_token_accuracy": 0.8778820633888245, "num_tokens": 826384556.0, "step": 21760 }, { "epoch": 2.7682228724080904, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.7366623878479, "learning_rate": 1e-05, "loss": 0.5409, "mean_token_accuracy": 0.8546801209449768, "num_tokens": 826418344.0, "step": 21761 }, { "epoch": 2.768350082686681, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.637640476226807, "learning_rate": 1e-05, "loss": 0.4838, "mean_token_accuracy": 0.8760509490966797, "num_tokens": 826460027.0, "step": 21762 }, { "epoch": 2.7684772929652715, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.605371952056885, "learning_rate": 1e-05, "loss": 0.4875, "mean_token_accuracy": 0.8761094212532043, "num_tokens": 826497422.0, "step": 21763 }, { "epoch": 2.768604503243862, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.477292537689209, "learning_rate": 1e-05, "loss": 0.519, "mean_token_accuracy": 0.8700345754623413, "num_tokens": 826536624.0, "step": 21764 }, { "epoch": 2.7687317135224525, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.515364646911621, "learning_rate": 1e-05, "loss": 0.4801, "mean_token_accuracy": 0.8724501132965088, "num_tokens": 826572303.0, "step": 21765 }, { "epoch": 2.768858923801043, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.4030537605285645, "learning_rate": 1e-05, "loss": 0.5345, "mean_token_accuracy": 0.856762707233429, "num_tokens": 826613326.0, "step": 21766 }, { "epoch": 2.7689861340796336, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.434935092926025, "learning_rate": 1e-05, "loss": 0.4806, "mean_token_accuracy": 0.8751320838928223, "num_tokens": 826649059.0, "step": 21767 }, { "epoch": 2.769113344358224, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.354670524597168, "learning_rate": 1e-05, "loss": 0.4672, "mean_token_accuracy": 0.8804807662963867, "num_tokens": 826682983.0, "step": 21768 }, { "epoch": 2.7692405546368146, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.37205696105957, "learning_rate": 1e-05, "loss": 0.4819, "mean_token_accuracy": 0.8742835521697998, "num_tokens": 826715179.0, "step": 21769 }, { "epoch": 2.769367764915405, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.328474521636963, "learning_rate": 1e-05, "loss": 0.5021, "mean_token_accuracy": 0.86552894115448, "num_tokens": 826748442.0, "step": 21770 }, { "epoch": 2.7694949751939957, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.318173885345459, "learning_rate": 1e-05, "loss": 0.4493, "mean_token_accuracy": 0.885393500328064, "num_tokens": 826791014.0, "step": 21771 }, { "epoch": 2.769622185472586, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.30877685546875, "learning_rate": 1e-05, "loss": 0.4845, "mean_token_accuracy": 0.8709840774536133, "num_tokens": 826827987.0, "step": 21772 }, { "epoch": 2.7697493957511767, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.274003505706787, "learning_rate": 1e-05, "loss": 0.4605, "mean_token_accuracy": 0.8809018731117249, "num_tokens": 826869315.0, "step": 21773 }, { "epoch": 2.7698766060297673, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.318596839904785, "learning_rate": 1e-05, "loss": 0.4833, "mean_token_accuracy": 0.8745156526565552, "num_tokens": 826912679.0, "step": 21774 }, { "epoch": 2.770003816308358, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.242636203765869, "learning_rate": 1e-05, "loss": 0.4624, "mean_token_accuracy": 0.8801066875457764, "num_tokens": 826949276.0, "step": 21775 }, { "epoch": 2.7701310265869483, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.341018199920654, "learning_rate": 1e-05, "loss": 0.4566, "mean_token_accuracy": 0.8845803737640381, "num_tokens": 826989441.0, "step": 21776 }, { "epoch": 2.770258236865539, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.193613052368164, "learning_rate": 1e-05, "loss": 0.4968, "mean_token_accuracy": 0.8701403737068176, "num_tokens": 827025030.0, "step": 21777 }, { "epoch": 2.7703854471441294, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.402306079864502, "learning_rate": 1e-05, "loss": 0.4785, "mean_token_accuracy": 0.875569224357605, "num_tokens": 827060252.0, "step": 21778 }, { "epoch": 2.77051265742272, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.12198543548584, "learning_rate": 1e-05, "loss": 0.4736, "mean_token_accuracy": 0.8773250579833984, "num_tokens": 827100063.0, "step": 21779 }, { "epoch": 2.77063986770131, "ewc_loss": 0.125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.501334190368652, "learning_rate": 1e-05, "loss": 0.4393, "mean_token_accuracy": 0.8892917633056641, "num_tokens": 827131772.0, "step": 21780 }, { "epoch": 2.770767077979901, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 5.998473167419434, "learning_rate": 1e-05, "loss": 0.5111, "mean_token_accuracy": 0.8640526533126831, "num_tokens": 827167659.0, "step": 21781 }, { "epoch": 2.770894288258491, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.706476211547852, "learning_rate": 1e-05, "loss": 0.5035, "mean_token_accuracy": 0.868780791759491, "num_tokens": 827211215.0, "step": 21782 }, { "epoch": 2.771021498537082, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.776203632354736, "learning_rate": 1e-05, "loss": 0.4981, "mean_token_accuracy": 0.8676353693008423, "num_tokens": 827248075.0, "step": 21783 }, { "epoch": 2.771148708815672, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00011301040649414062, "grad_norm": 7.152350902557373, "learning_rate": 1e-05, "loss": 0.4757, "mean_token_accuracy": 0.881981611251831, "num_tokens": 827283270.0, "step": 21784 }, { "epoch": 2.771275919094263, "ewc_loss": 0.1123046875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 8.296966552734375e-05, "grad_norm": 5.54894495010376, "learning_rate": 1e-05, "loss": 0.4484, "mean_token_accuracy": 0.8822634816169739, "num_tokens": 827317520.0, "step": 21785 }, { "epoch": 2.771403129372853, "ewc_loss": 0.162109375, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00013256072998046875, "grad_norm": 7.929802417755127, "learning_rate": 1e-05, "loss": 0.5353, "mean_token_accuracy": 0.8689677715301514, "num_tokens": 827355899.0, "step": 21786 }, { "epoch": 2.7715303396514437, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 5.7400221824646, "learning_rate": 1e-05, "loss": 0.5407, "mean_token_accuracy": 0.8570981025695801, "num_tokens": 827390038.0, "step": 21787 }, { "epoch": 2.771657549930034, "ewc_loss": 0.1826171875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00015354156494140625, "grad_norm": 8.767988204956055, "learning_rate": 1e-05, "loss": 0.5548, "mean_token_accuracy": 0.8722667694091797, "num_tokens": 827428563.0, "step": 21788 }, { "epoch": 2.7717847602086247, "ewc_loss": 0.1572265625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.0001277923583984375, "grad_norm": 6.943639278411865, "learning_rate": 1e-05, "loss": 0.4847, "mean_token_accuracy": 0.8840943574905396, "num_tokens": 827469658.0, "step": 21789 }, { "epoch": 2.7719119704872153, "ewc_loss": 0.173828125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00014495849609375, "grad_norm": 8.47192668914795, "learning_rate": 1e-05, "loss": 0.5054, "mean_token_accuracy": 0.8802698254585266, "num_tokens": 827504277.0, "step": 21790 }, { "epoch": 2.772039180765806, "ewc_loss": 0.1640625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00013446807861328125, "grad_norm": 7.563868999481201, "learning_rate": 1e-05, "loss": 0.5354, "mean_token_accuracy": 0.8712825179100037, "num_tokens": 827544840.0, "step": 21791 }, { "epoch": 2.7721663910443963, "ewc_loss": 0.1572265625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.0001277923583984375, "grad_norm": 7.73655891418457, "learning_rate": 1e-05, "loss": 0.5559, "mean_token_accuracy": 0.8632040619850159, "num_tokens": 827590607.0, "step": 21792 }, { "epoch": 2.772293601322987, "ewc_loss": 0.1533203125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.0001239776611328125, "grad_norm": 7.253536701202393, "learning_rate": 1e-05, "loss": 0.447, "mean_token_accuracy": 0.8923721313476562, "num_tokens": 827622039.0, "step": 21793 }, { "epoch": 2.7724208116015774, "ewc_loss": 0.1484375, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00011920928955078125, "grad_norm": 7.370981693267822, "learning_rate": 1e-05, "loss": 0.4785, "mean_token_accuracy": 0.8845342397689819, "num_tokens": 827659234.0, "step": 21794 }, { "epoch": 2.772548021880168, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00011157989501953125, "grad_norm": 6.875394344329834, "learning_rate": 1e-05, "loss": 0.4878, "mean_token_accuracy": 0.8786371350288391, "num_tokens": 827702044.0, "step": 21795 }, { "epoch": 2.7726752321587584, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00011157989501953125, "grad_norm": 7.170134544372559, "learning_rate": 1e-05, "loss": 0.4759, "mean_token_accuracy": 0.8816438913345337, "num_tokens": 827738189.0, "step": 21796 }, { "epoch": 2.772802442437349, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.592569828033447, "learning_rate": 1e-05, "loss": 0.5194, "mean_token_accuracy": 0.8670570850372314, "num_tokens": 827780483.0, "step": 21797 }, { "epoch": 2.7729296527159395, "ewc_loss": 0.140625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 7.124988079071045, "learning_rate": 1e-05, "loss": 0.4823, "mean_token_accuracy": 0.8796546459197998, "num_tokens": 827817156.0, "step": 21798 }, { "epoch": 2.77305686299453, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.354467868804932, "learning_rate": 1e-05, "loss": 0.4736, "mean_token_accuracy": 0.878905713558197, "num_tokens": 827851591.0, "step": 21799 }, { "epoch": 2.7731840732731206, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00011205673217773438, "grad_norm": 7.216898441314697, "learning_rate": 1e-05, "loss": 0.4784, "mean_token_accuracy": 0.8837875723838806, "num_tokens": 827891289.0, "step": 21800 }, { "epoch": 2.773311283551711, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.1793413162231445, "learning_rate": 1e-05, "loss": 0.4073, "mean_token_accuracy": 0.8973914980888367, "num_tokens": 827926466.0, "step": 21801 }, { "epoch": 2.7734384938303016, "ewc_loss": 0.1455078125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.0001163482666015625, "grad_norm": 7.491501808166504, "learning_rate": 1e-05, "loss": 0.5632, "mean_token_accuracy": 0.8497211933135986, "num_tokens": 827964885.0, "step": 21802 }, { "epoch": 2.773565704108892, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.161108493804932, "learning_rate": 1e-05, "loss": 0.4995, "mean_token_accuracy": 0.8680906891822815, "num_tokens": 828006063.0, "step": 21803 }, { "epoch": 2.7736929143874827, "ewc_loss": 0.154296875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00012493133544921875, "grad_norm": 7.8131422996521, "learning_rate": 1e-05, "loss": 0.4819, "mean_token_accuracy": 0.8833845853805542, "num_tokens": 828037533.0, "step": 21804 }, { "epoch": 2.7738201246660728, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.471582412719727, "learning_rate": 1e-05, "loss": 0.4615, "mean_token_accuracy": 0.885284423828125, "num_tokens": 828072298.0, "step": 21805 }, { "epoch": 2.7739473349446637, "ewc_loss": 0.154296875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00012493133544921875, "grad_norm": 7.787410259246826, "learning_rate": 1e-05, "loss": 0.4787, "mean_token_accuracy": 0.8835117816925049, "num_tokens": 828107900.0, "step": 21806 }, { "epoch": 2.774074545223254, "ewc_loss": 0.140625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 6.845188140869141, "learning_rate": 1e-05, "loss": 0.4762, "mean_token_accuracy": 0.881150484085083, "num_tokens": 828143323.0, "step": 21807 }, { "epoch": 2.774201755501845, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00011444091796875, "grad_norm": 7.318326950073242, "learning_rate": 1e-05, "loss": 0.455, "mean_token_accuracy": 0.8918472528457642, "num_tokens": 828183439.0, "step": 21808 }, { "epoch": 2.774328965780435, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 6.939477920532227, "learning_rate": 1e-05, "loss": 0.48, "mean_token_accuracy": 0.8817870020866394, "num_tokens": 828218360.0, "step": 21809 }, { "epoch": 2.774456176059026, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.763901233673096, "learning_rate": 1e-05, "loss": 0.5025, "mean_token_accuracy": 0.8700623512268066, "num_tokens": 828254042.0, "step": 21810 }, { "epoch": 2.774583386337616, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 6.939586162567139, "learning_rate": 1e-05, "loss": 0.4948, "mean_token_accuracy": 0.8760449290275574, "num_tokens": 828290992.0, "step": 21811 }, { "epoch": 2.7747105966162064, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.262009143829346, "learning_rate": 1e-05, "loss": 0.4515, "mean_token_accuracy": 0.8821252584457397, "num_tokens": 828323351.0, "step": 21812 }, { "epoch": 2.774837806894797, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 7.073418140411377, "learning_rate": 1e-05, "loss": 0.4664, "mean_token_accuracy": 0.8830176591873169, "num_tokens": 828354555.0, "step": 21813 }, { "epoch": 2.7749650171733875, "ewc_loss": 0.11376953125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.92400598526001, "learning_rate": 1e-05, "loss": 0.4549, "mean_token_accuracy": 0.8807581663131714, "num_tokens": 828395641.0, "step": 21814 }, { "epoch": 2.775092227451978, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00011539459228515625, "grad_norm": 7.4111480712890625, "learning_rate": 1e-05, "loss": 0.4978, "mean_token_accuracy": 0.8788572549819946, "num_tokens": 828437871.0, "step": 21815 }, { "epoch": 2.7752194377305686, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.835419654846191, "learning_rate": 1e-05, "loss": 0.5054, "mean_token_accuracy": 0.8637222051620483, "num_tokens": 828480545.0, "step": 21816 }, { "epoch": 2.775346648009159, "ewc_loss": 0.1591796875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00012969970703125, "grad_norm": 7.904987335205078, "learning_rate": 1e-05, "loss": 0.5009, "mean_token_accuracy": 0.8796036243438721, "num_tokens": 828513519.0, "step": 21817 }, { "epoch": 2.7754738582877496, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.209895610809326, "learning_rate": 1e-05, "loss": 0.4893, "mean_token_accuracy": 0.8721197247505188, "num_tokens": 828549329.0, "step": 21818 }, { "epoch": 2.77560106856634, "ewc_loss": 0.1630859375, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.000133514404296875, "grad_norm": 8.097599983215332, "learning_rate": 1e-05, "loss": 0.5595, "mean_token_accuracy": 0.8633809089660645, "num_tokens": 828592357.0, "step": 21819 }, { "epoch": 2.7757282788449307, "ewc_loss": 0.1455078125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.0001163482666015625, "grad_norm": 6.848602294921875, "learning_rate": 1e-05, "loss": 0.5083, "mean_token_accuracy": 0.8755826950073242, "num_tokens": 828627622.0, "step": 21820 }, { "epoch": 2.775855489123521, "ewc_loss": 0.15234375, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.658839225769043, "learning_rate": 1e-05, "loss": 0.4733, "mean_token_accuracy": 0.8873177766799927, "num_tokens": 828668680.0, "step": 21821 }, { "epoch": 2.7759826994021117, "ewc_loss": 0.1455078125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.0001163482666015625, "grad_norm": 6.990700721740723, "learning_rate": 1e-05, "loss": 0.4958, "mean_token_accuracy": 0.875248908996582, "num_tokens": 828705233.0, "step": 21822 }, { "epoch": 2.7761099096807023, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00011301040649414062, "grad_norm": 7.166281223297119, "learning_rate": 1e-05, "loss": 0.5213, "mean_token_accuracy": 0.8663938045501709, "num_tokens": 828745183.0, "step": 21823 }, { "epoch": 2.776237119959293, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 6.8575592041015625, "learning_rate": 1e-05, "loss": 0.5111, "mean_token_accuracy": 0.8717978596687317, "num_tokens": 828787769.0, "step": 21824 }, { "epoch": 2.7763643302378833, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.813989639282227, "learning_rate": 1e-05, "loss": 0.4886, "mean_token_accuracy": 0.8727282285690308, "num_tokens": 828826511.0, "step": 21825 }, { "epoch": 2.776491540516474, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.695131778717041, "learning_rate": 1e-05, "loss": 0.4794, "mean_token_accuracy": 0.8767972588539124, "num_tokens": 828868632.0, "step": 21826 }, { "epoch": 2.7766187507950644, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.6219611167907715, "learning_rate": 1e-05, "loss": 0.5051, "mean_token_accuracy": 0.8689954876899719, "num_tokens": 828904342.0, "step": 21827 }, { "epoch": 2.776745961073655, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.529311656951904, "learning_rate": 1e-05, "loss": 0.5216, "mean_token_accuracy": 0.8735291957855225, "num_tokens": 828947074.0, "step": 21828 }, { "epoch": 2.7768731713522454, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.556400775909424, "learning_rate": 1e-05, "loss": 0.4645, "mean_token_accuracy": 0.8799557089805603, "num_tokens": 828989583.0, "step": 21829 }, { "epoch": 2.7770003816308355, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.362452507019043, "learning_rate": 1e-05, "loss": 0.553, "mean_token_accuracy": 0.849055290222168, "num_tokens": 829028848.0, "step": 21830 }, { "epoch": 2.7771275919094265, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.582205295562744, "learning_rate": 1e-05, "loss": 0.4771, "mean_token_accuracy": 0.8770152926445007, "num_tokens": 829063904.0, "step": 21831 }, { "epoch": 2.7772548021880166, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.1964030265808105, "learning_rate": 1e-05, "loss": 0.5141, "mean_token_accuracy": 0.8628594875335693, "num_tokens": 829102416.0, "step": 21832 }, { "epoch": 2.7773820124666075, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.708761692047119, "learning_rate": 1e-05, "loss": 0.5157, "mean_token_accuracy": 0.8728272914886475, "num_tokens": 829138870.0, "step": 21833 }, { "epoch": 2.7775092227451976, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 6.015300750732422, "learning_rate": 1e-05, "loss": 0.5085, "mean_token_accuracy": 0.8691051602363586, "num_tokens": 829176704.0, "step": 21834 }, { "epoch": 2.7776364330237886, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.95286750793457, "learning_rate": 1e-05, "loss": 0.467, "mean_token_accuracy": 0.883417546749115, "num_tokens": 829210170.0, "step": 21835 }, { "epoch": 2.7777636433023787, "ewc_loss": 0.11376953125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.81491756439209, "learning_rate": 1e-05, "loss": 0.4307, "mean_token_accuracy": 0.8902690410614014, "num_tokens": 829249692.0, "step": 21836 }, { "epoch": 2.777890853580969, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00011682510375976562, "grad_norm": 7.420324802398682, "learning_rate": 1e-05, "loss": 0.4763, "mean_token_accuracy": 0.8844618201255798, "num_tokens": 829285342.0, "step": 21837 }, { "epoch": 2.7780180638595597, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 5.796501636505127, "learning_rate": 1e-05, "loss": 0.4797, "mean_token_accuracy": 0.8769286274909973, "num_tokens": 829320281.0, "step": 21838 }, { "epoch": 2.7781452741381503, "ewc_loss": 0.162109375, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00013256072998046875, "grad_norm": 7.992462158203125, "learning_rate": 1e-05, "loss": 0.5295, "mean_token_accuracy": 0.8717268705368042, "num_tokens": 829359990.0, "step": 21839 }, { "epoch": 2.778272484416741, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.25827169418335, "learning_rate": 1e-05, "loss": 0.4526, "mean_token_accuracy": 0.8866022825241089, "num_tokens": 829395330.0, "step": 21840 }, { "epoch": 2.7783996946953313, "ewc_loss": 0.1640625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00013446807861328125, "grad_norm": 8.149683952331543, "learning_rate": 1e-05, "loss": 0.5391, "mean_token_accuracy": 0.8747982978820801, "num_tokens": 829432127.0, "step": 21841 }, { "epoch": 2.778526904973922, "ewc_loss": 0.1484375, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00011873245239257812, "grad_norm": 6.930610656738281, "learning_rate": 1e-05, "loss": 0.5132, "mean_token_accuracy": 0.8730206489562988, "num_tokens": 829469735.0, "step": 21842 }, { "epoch": 2.7786541152525124, "ewc_loss": 0.1533203125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.0001239776611328125, "grad_norm": 7.650012969970703, "learning_rate": 1e-05, "loss": 0.5123, "mean_token_accuracy": 0.8755354881286621, "num_tokens": 829502893.0, "step": 21843 }, { "epoch": 2.778781325531103, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00011682510375976562, "grad_norm": 7.057038307189941, "learning_rate": 1e-05, "loss": 0.4859, "mean_token_accuracy": 0.8829845190048218, "num_tokens": 829538307.0, "step": 21844 }, { "epoch": 2.7789085358096934, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00010967254638671875, "grad_norm": 7.101036071777344, "learning_rate": 1e-05, "loss": 0.4598, "mean_token_accuracy": 0.88558429479599, "num_tokens": 829575992.0, "step": 21845 }, { "epoch": 2.779035746088284, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00010967254638671875, "grad_norm": 6.917725563049316, "learning_rate": 1e-05, "loss": 0.499, "mean_token_accuracy": 0.8694201111793518, "num_tokens": 829610167.0, "step": 21846 }, { "epoch": 2.7791629563668745, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.692392826080322, "learning_rate": 1e-05, "loss": 0.5158, "mean_token_accuracy": 0.8667894601821899, "num_tokens": 829646828.0, "step": 21847 }, { "epoch": 2.779290166645465, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.836410045623779, "learning_rate": 1e-05, "loss": 0.5217, "mean_token_accuracy": 0.8675469160079956, "num_tokens": 829683593.0, "step": 21848 }, { "epoch": 2.7794173769240555, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.401614665985107, "learning_rate": 1e-05, "loss": 0.4752, "mean_token_accuracy": 0.8742009401321411, "num_tokens": 829721080.0, "step": 21849 }, { "epoch": 2.779544587202646, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.810340881347656, "learning_rate": 1e-05, "loss": 0.4367, "mean_token_accuracy": 0.8921514749526978, "num_tokens": 829755871.0, "step": 21850 }, { "epoch": 2.7796717974812366, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.219515800476074, "learning_rate": 1e-05, "loss": 0.4715, "mean_token_accuracy": 0.8779957294464111, "num_tokens": 829793377.0, "step": 21851 }, { "epoch": 2.779799007759827, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.8655195236206055, "learning_rate": 1e-05, "loss": 0.4833, "mean_token_accuracy": 0.8740404844284058, "num_tokens": 829833165.0, "step": 21852 }, { "epoch": 2.7799262180384177, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.067774295806885, "learning_rate": 1e-05, "loss": 0.4613, "mean_token_accuracy": 0.8797355890274048, "num_tokens": 829869558.0, "step": 21853 }, { "epoch": 2.780053428317008, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 7.026487827301025, "learning_rate": 1e-05, "loss": 0.5014, "mean_token_accuracy": 0.8779342770576477, "num_tokens": 829906385.0, "step": 21854 }, { "epoch": 2.7801806385955983, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 5.942112922668457, "learning_rate": 1e-05, "loss": 0.5036, "mean_token_accuracy": 0.8673763871192932, "num_tokens": 829944216.0, "step": 21855 }, { "epoch": 2.7803078488741892, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00011491775512695312, "grad_norm": 7.3071746826171875, "learning_rate": 1e-05, "loss": 0.57, "mean_token_accuracy": 0.8535617589950562, "num_tokens": 829976817.0, "step": 21856 }, { "epoch": 2.7804350591527793, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 5.918779373168945, "learning_rate": 1e-05, "loss": 0.5035, "mean_token_accuracy": 0.8684729337692261, "num_tokens": 830020434.0, "step": 21857 }, { "epoch": 2.7805622694313703, "ewc_loss": 0.1533203125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.0001239776611328125, "grad_norm": 7.65531587600708, "learning_rate": 1e-05, "loss": 0.5032, "mean_token_accuracy": 0.877650797367096, "num_tokens": 830054469.0, "step": 21858 }, { "epoch": 2.7806894797099604, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.134164810180664, "learning_rate": 1e-05, "loss": 0.4797, "mean_token_accuracy": 0.878124475479126, "num_tokens": 830091536.0, "step": 21859 }, { "epoch": 2.780816689988551, "ewc_loss": 0.158203125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00012874603271484375, "grad_norm": 7.812626361846924, "learning_rate": 1e-05, "loss": 0.5315, "mean_token_accuracy": 0.8687450289726257, "num_tokens": 830129425.0, "step": 21860 }, { "epoch": 2.7809439002671414, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 6.509724140167236, "learning_rate": 1e-05, "loss": 0.503, "mean_token_accuracy": 0.8708142042160034, "num_tokens": 830171137.0, "step": 21861 }, { "epoch": 2.781071110545732, "ewc_loss": 0.1533203125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.0001239776611328125, "grad_norm": 7.618557453155518, "learning_rate": 1e-05, "loss": 0.473, "mean_token_accuracy": 0.8890931010246277, "num_tokens": 830211434.0, "step": 21862 }, { "epoch": 2.7811983208243225, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00011157989501953125, "grad_norm": 6.732468128204346, "learning_rate": 1e-05, "loss": 0.4495, "mean_token_accuracy": 0.8902180194854736, "num_tokens": 830250843.0, "step": 21863 }, { "epoch": 2.781325531102913, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00011396408081054688, "grad_norm": 7.211550235748291, "learning_rate": 1e-05, "loss": 0.5293, "mean_token_accuracy": 0.8670507669448853, "num_tokens": 830292131.0, "step": 21864 }, { "epoch": 2.7814527413815036, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 6.739956855773926, "learning_rate": 1e-05, "loss": 0.4548, "mean_token_accuracy": 0.8853702545166016, "num_tokens": 830329678.0, "step": 21865 }, { "epoch": 2.781579951660094, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.817440509796143, "learning_rate": 1e-05, "loss": 0.4653, "mean_token_accuracy": 0.884874165058136, "num_tokens": 830367822.0, "step": 21866 }, { "epoch": 2.7817071619386846, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.695402145385742, "learning_rate": 1e-05, "loss": 0.5031, "mean_token_accuracy": 0.874556839466095, "num_tokens": 830408170.0, "step": 21867 }, { "epoch": 2.781834372217275, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.502185821533203, "learning_rate": 1e-05, "loss": 0.4638, "mean_token_accuracy": 0.8806257247924805, "num_tokens": 830447402.0, "step": 21868 }, { "epoch": 2.7819615824958657, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.673187732696533, "learning_rate": 1e-05, "loss": 0.5567, "mean_token_accuracy": 0.8549848794937134, "num_tokens": 830485411.0, "step": 21869 }, { "epoch": 2.782088792774456, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.305373191833496, "learning_rate": 1e-05, "loss": 0.4755, "mean_token_accuracy": 0.8768545389175415, "num_tokens": 830522562.0, "step": 21870 }, { "epoch": 2.7822160030530467, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.658355236053467, "learning_rate": 1e-05, "loss": 0.5186, "mean_token_accuracy": 0.8639094829559326, "num_tokens": 830565979.0, "step": 21871 }, { "epoch": 2.7823432133316373, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.177123546600342, "learning_rate": 1e-05, "loss": 0.4756, "mean_token_accuracy": 0.8735194206237793, "num_tokens": 830602771.0, "step": 21872 }, { "epoch": 2.782470423610228, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.681265354156494, "learning_rate": 1e-05, "loss": 0.5061, "mean_token_accuracy": 0.8660861253738403, "num_tokens": 830639891.0, "step": 21873 }, { "epoch": 2.7825976338888183, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.0700178146362305, "learning_rate": 1e-05, "loss": 0.4199, "mean_token_accuracy": 0.8870574235916138, "num_tokens": 830676044.0, "step": 21874 }, { "epoch": 2.782724844167409, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.783441066741943, "learning_rate": 1e-05, "loss": 0.4918, "mean_token_accuracy": 0.8749873042106628, "num_tokens": 830718252.0, "step": 21875 }, { "epoch": 2.7828520544459994, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 5.947928428649902, "learning_rate": 1e-05, "loss": 0.434, "mean_token_accuracy": 0.8882650136947632, "num_tokens": 830757563.0, "step": 21876 }, { "epoch": 2.78297926472459, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 7.020521640777588, "learning_rate": 1e-05, "loss": 0.5015, "mean_token_accuracy": 0.8707641363143921, "num_tokens": 830797808.0, "step": 21877 }, { "epoch": 2.78310647500318, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.817071437835693, "learning_rate": 1e-05, "loss": 0.4823, "mean_token_accuracy": 0.8730213046073914, "num_tokens": 830838691.0, "step": 21878 }, { "epoch": 2.783233685281771, "ewc_loss": 0.1484375, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00011873245239257812, "grad_norm": 7.398828029632568, "learning_rate": 1e-05, "loss": 0.4669, "mean_token_accuracy": 0.8893755674362183, "num_tokens": 830877016.0, "step": 21879 }, { "epoch": 2.783360895560361, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 5.839415550231934, "learning_rate": 1e-05, "loss": 0.4748, "mean_token_accuracy": 0.8805207014083862, "num_tokens": 830908841.0, "step": 21880 }, { "epoch": 2.783488105838952, "ewc_loss": 0.16015625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00013065338134765625, "grad_norm": 7.832283973693848, "learning_rate": 1e-05, "loss": 0.5354, "mean_token_accuracy": 0.8722273111343384, "num_tokens": 830947580.0, "step": 21881 }, { "epoch": 2.783615316117542, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.190395832061768, "learning_rate": 1e-05, "loss": 0.4871, "mean_token_accuracy": 0.8765736222267151, "num_tokens": 830984385.0, "step": 21882 }, { "epoch": 2.783742526396133, "ewc_loss": 0.1640625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00013446807861328125, "grad_norm": 7.99338960647583, "learning_rate": 1e-05, "loss": 0.4927, "mean_token_accuracy": 0.8873257040977478, "num_tokens": 831022280.0, "step": 21883 }, { "epoch": 2.783869736674723, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00011539459228515625, "grad_norm": 6.694568157196045, "learning_rate": 1e-05, "loss": 0.4951, "mean_token_accuracy": 0.8798024654388428, "num_tokens": 831055764.0, "step": 21884 }, { "epoch": 2.7839969469533137, "ewc_loss": 0.154296875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00012493133544921875, "grad_norm": 7.670560836791992, "learning_rate": 1e-05, "loss": 0.5715, "mean_token_accuracy": 0.8652600049972534, "num_tokens": 831093397.0, "step": 21885 }, { "epoch": 2.784124157231904, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00011730194091796875, "grad_norm": 6.872089862823486, "learning_rate": 1e-05, "loss": 0.5004, "mean_token_accuracy": 0.8770765066146851, "num_tokens": 831130720.0, "step": 21886 }, { "epoch": 2.7842513675104947, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00011348724365234375, "grad_norm": 7.208400726318359, "learning_rate": 1e-05, "loss": 0.5019, "mean_token_accuracy": 0.8769851922988892, "num_tokens": 831164092.0, "step": 21887 }, { "epoch": 2.7843785777890853, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00011205673217773438, "grad_norm": 6.808396339416504, "learning_rate": 1e-05, "loss": 0.504, "mean_token_accuracy": 0.8717774748802185, "num_tokens": 831202570.0, "step": 21888 }, { "epoch": 2.784505788067676, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.799617767333984, "learning_rate": 1e-05, "loss": 0.5271, "mean_token_accuracy": 0.8628170490264893, "num_tokens": 831242596.0, "step": 21889 }, { "epoch": 2.7846329983462663, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.720423221588135, "learning_rate": 1e-05, "loss": 0.4481, "mean_token_accuracy": 0.8890195488929749, "num_tokens": 831283305.0, "step": 21890 }, { "epoch": 2.784760208624857, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.498682022094727, "learning_rate": 1e-05, "loss": 0.444, "mean_token_accuracy": 0.8899548649787903, "num_tokens": 831322355.0, "step": 21891 }, { "epoch": 2.7848874189034474, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.682948589324951, "learning_rate": 1e-05, "loss": 0.4987, "mean_token_accuracy": 0.8740512132644653, "num_tokens": 831361410.0, "step": 21892 }, { "epoch": 2.785014629182038, "ewc_loss": 0.125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.311457633972168, "learning_rate": 1e-05, "loss": 0.4998, "mean_token_accuracy": 0.8692681789398193, "num_tokens": 831402493.0, "step": 21893 }, { "epoch": 2.7851418394606284, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.675354480743408, "learning_rate": 1e-05, "loss": 0.5553, "mean_token_accuracy": 0.8514800667762756, "num_tokens": 831443231.0, "step": 21894 }, { "epoch": 2.785269049739219, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.173933982849121, "learning_rate": 1e-05, "loss": 0.4083, "mean_token_accuracy": 0.9009230732917786, "num_tokens": 831474716.0, "step": 21895 }, { "epoch": 2.7853962600178095, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.727783679962158, "learning_rate": 1e-05, "loss": 0.5329, "mean_token_accuracy": 0.8600701093673706, "num_tokens": 831519410.0, "step": 21896 }, { "epoch": 2.7855234702964, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.031994342803955, "learning_rate": 1e-05, "loss": 0.4651, "mean_token_accuracy": 0.8817279934883118, "num_tokens": 831562040.0, "step": 21897 }, { "epoch": 2.7856506805749905, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.88883638381958, "learning_rate": 1e-05, "loss": 0.5212, "mean_token_accuracy": 0.867591142654419, "num_tokens": 831604956.0, "step": 21898 }, { "epoch": 2.785777890853581, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.896815299987793, "learning_rate": 1e-05, "loss": 0.4539, "mean_token_accuracy": 0.879474937915802, "num_tokens": 831637270.0, "step": 21899 }, { "epoch": 2.7859051011321716, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00011348724365234375, "grad_norm": 7.1875762939453125, "learning_rate": 1e-05, "loss": 0.4687, "mean_token_accuracy": 0.8832073211669922, "num_tokens": 831671342.0, "step": 21900 }, { "epoch": 2.786032311410762, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 5.785180568695068, "learning_rate": 1e-05, "loss": 0.4798, "mean_token_accuracy": 0.8733463287353516, "num_tokens": 831709222.0, "step": 21901 }, { "epoch": 2.7861595216893527, "ewc_loss": 0.15234375, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.646465301513672, "learning_rate": 1e-05, "loss": 0.4864, "mean_token_accuracy": 0.885267436504364, "num_tokens": 831749690.0, "step": 21902 }, { "epoch": 2.7862867319679427, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 5.955482006072998, "learning_rate": 1e-05, "loss": 0.5341, "mean_token_accuracy": 0.8641828298568726, "num_tokens": 831787963.0, "step": 21903 }, { "epoch": 2.7864139422465337, "ewc_loss": 0.1640625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00013446807861328125, "grad_norm": 8.016928672790527, "learning_rate": 1e-05, "loss": 0.4936, "mean_token_accuracy": 0.8834095001220703, "num_tokens": 831820925.0, "step": 21904 }, { "epoch": 2.786541152525124, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00011205673217773438, "grad_norm": 6.502748489379883, "learning_rate": 1e-05, "loss": 0.5158, "mean_token_accuracy": 0.8721104264259338, "num_tokens": 831856190.0, "step": 21905 }, { "epoch": 2.7866683628037148, "ewc_loss": 0.1611328125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.0001316070556640625, "grad_norm": 7.872514724731445, "learning_rate": 1e-05, "loss": 0.5535, "mean_token_accuracy": 0.8638842701911926, "num_tokens": 831900339.0, "step": 21906 }, { "epoch": 2.786795573082305, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00011730194091796875, "grad_norm": 6.86937141418457, "learning_rate": 1e-05, "loss": 0.5151, "mean_token_accuracy": 0.8693951368331909, "num_tokens": 831932166.0, "step": 21907 }, { "epoch": 2.786922783360896, "ewc_loss": 0.1474609375, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.000118255615234375, "grad_norm": 7.417809009552002, "learning_rate": 1e-05, "loss": 0.4851, "mean_token_accuracy": 0.8795473575592041, "num_tokens": 831964796.0, "step": 21908 }, { "epoch": 2.787049993639486, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00011444091796875, "grad_norm": 6.865764141082764, "learning_rate": 1e-05, "loss": 0.5066, "mean_token_accuracy": 0.87372225522995, "num_tokens": 832004603.0, "step": 21909 }, { "epoch": 2.7871772039180764, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 6.965732574462891, "learning_rate": 1e-05, "loss": 0.459, "mean_token_accuracy": 0.8883898258209229, "num_tokens": 832042847.0, "step": 21910 }, { "epoch": 2.787304414196667, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 6.76997184753418, "learning_rate": 1e-05, "loss": 0.4813, "mean_token_accuracy": 0.8767485022544861, "num_tokens": 832079746.0, "step": 21911 }, { "epoch": 2.7874316244752575, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.616392612457275, "learning_rate": 1e-05, "loss": 0.5105, "mean_token_accuracy": 0.8691771030426025, "num_tokens": 832116300.0, "step": 21912 }, { "epoch": 2.787558834753848, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.7130913734436035, "learning_rate": 1e-05, "loss": 0.4578, "mean_token_accuracy": 0.8876940011978149, "num_tokens": 832151985.0, "step": 21913 }, { "epoch": 2.7876860450324386, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.3876166343688965, "learning_rate": 1e-05, "loss": 0.4945, "mean_token_accuracy": 0.8763183355331421, "num_tokens": 832181327.0, "step": 21914 }, { "epoch": 2.787813255311029, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.674016952514648, "learning_rate": 1e-05, "loss": 0.4797, "mean_token_accuracy": 0.8810676336288452, "num_tokens": 832217710.0, "step": 21915 }, { "epoch": 2.7879404655896196, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.2313151359558105, "learning_rate": 1e-05, "loss": 0.4971, "mean_token_accuracy": 0.873220682144165, "num_tokens": 832255628.0, "step": 21916 }, { "epoch": 2.78806767586821, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.701889514923096, "learning_rate": 1e-05, "loss": 0.4815, "mean_token_accuracy": 0.8770015239715576, "num_tokens": 832294239.0, "step": 21917 }, { "epoch": 2.7881948861468007, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.0887956619262695, "learning_rate": 1e-05, "loss": 0.4538, "mean_token_accuracy": 0.8812524080276489, "num_tokens": 832331270.0, "step": 21918 }, { "epoch": 2.788322096425391, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.808906555175781, "learning_rate": 1e-05, "loss": 0.5179, "mean_token_accuracy": 0.868106484413147, "num_tokens": 832372276.0, "step": 21919 }, { "epoch": 2.7884493067039817, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 5.950859546661377, "learning_rate": 1e-05, "loss": 0.4586, "mean_token_accuracy": 0.8854286670684814, "num_tokens": 832410352.0, "step": 21920 }, { "epoch": 2.7885765169825723, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 7.053749084472656, "learning_rate": 1e-05, "loss": 0.5281, "mean_token_accuracy": 0.8693528175354004, "num_tokens": 832452633.0, "step": 21921 }, { "epoch": 2.788703727261163, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.809082508087158, "learning_rate": 1e-05, "loss": 0.4546, "mean_token_accuracy": 0.8826483488082886, "num_tokens": 832492100.0, "step": 21922 }, { "epoch": 2.7888309375397533, "ewc_loss": 0.1494140625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00011968612670898438, "grad_norm": 7.4756903648376465, "learning_rate": 1e-05, "loss": 0.5066, "mean_token_accuracy": 0.8730658292770386, "num_tokens": 832529972.0, "step": 21923 }, { "epoch": 2.788958147818344, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 5.851528167724609, "learning_rate": 1e-05, "loss": 0.4462, "mean_token_accuracy": 0.8873207569122314, "num_tokens": 832571042.0, "step": 21924 }, { "epoch": 2.7890853580969344, "ewc_loss": 0.1630859375, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.000133514404296875, "grad_norm": 7.927757263183594, "learning_rate": 1e-05, "loss": 0.5385, "mean_token_accuracy": 0.867821216583252, "num_tokens": 832610575.0, "step": 21925 }, { "epoch": 2.789212568375525, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.289691925048828, "learning_rate": 1e-05, "loss": 0.4431, "mean_token_accuracy": 0.8889670372009277, "num_tokens": 832649477.0, "step": 21926 }, { "epoch": 2.7893397786541154, "ewc_loss": 0.1640625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00013446807861328125, "grad_norm": 8.01344108581543, "learning_rate": 1e-05, "loss": 0.4965, "mean_token_accuracy": 0.8830127716064453, "num_tokens": 832688958.0, "step": 21927 }, { "epoch": 2.7894669889327055, "ewc_loss": 0.1484375, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00011873245239257812, "grad_norm": 6.803857803344727, "learning_rate": 1e-05, "loss": 0.4646, "mean_token_accuracy": 0.8890906572341919, "num_tokens": 832724368.0, "step": 21928 }, { "epoch": 2.7895941992112965, "ewc_loss": 0.1533203125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.0001239776611328125, "grad_norm": 7.600837230682373, "learning_rate": 1e-05, "loss": 0.5272, "mean_token_accuracy": 0.8725475072860718, "num_tokens": 832765950.0, "step": 21929 }, { "epoch": 2.7897214094898866, "ewc_loss": 0.1474609375, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.000118255615234375, "grad_norm": 6.919313907623291, "learning_rate": 1e-05, "loss": 0.4425, "mean_token_accuracy": 0.892378032207489, "num_tokens": 832800666.0, "step": 21930 }, { "epoch": 2.7898486197684775, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00011301040649414062, "grad_norm": 7.131228923797607, "learning_rate": 1e-05, "loss": 0.5332, "mean_token_accuracy": 0.8655860424041748, "num_tokens": 832835830.0, "step": 21931 }, { "epoch": 2.7899758300470676, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.0001125335693359375, "grad_norm": 6.826370716094971, "learning_rate": 1e-05, "loss": 0.4633, "mean_token_accuracy": 0.8869580626487732, "num_tokens": 832866370.0, "step": 21932 }, { "epoch": 2.7901030403256586, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.7281880378723145, "learning_rate": 1e-05, "loss": 0.4983, "mean_token_accuracy": 0.8755788803100586, "num_tokens": 832914174.0, "step": 21933 }, { "epoch": 2.7902302506042487, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.733471393585205, "learning_rate": 1e-05, "loss": 0.4477, "mean_token_accuracy": 0.8889498710632324, "num_tokens": 832951397.0, "step": 21934 }, { "epoch": 2.790357460882839, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.47127103805542, "learning_rate": 1e-05, "loss": 0.4688, "mean_token_accuracy": 0.8821935653686523, "num_tokens": 832986282.0, "step": 21935 }, { "epoch": 2.7904846711614297, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.670222759246826, "learning_rate": 1e-05, "loss": 0.4608, "mean_token_accuracy": 0.8848139047622681, "num_tokens": 833023288.0, "step": 21936 }, { "epoch": 2.7906118814400203, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.313921928405762, "learning_rate": 1e-05, "loss": 0.4945, "mean_token_accuracy": 0.8735054731369019, "num_tokens": 833053744.0, "step": 21937 }, { "epoch": 2.790739091718611, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.633853435516357, "learning_rate": 1e-05, "loss": 0.4556, "mean_token_accuracy": 0.8863852620124817, "num_tokens": 833096499.0, "step": 21938 }, { "epoch": 2.7908663019972013, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.195506572723389, "learning_rate": 1e-05, "loss": 0.4431, "mean_token_accuracy": 0.8867172598838806, "num_tokens": 833132405.0, "step": 21939 }, { "epoch": 2.790993512275792, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.66619873046875, "learning_rate": 1e-05, "loss": 0.4858, "mean_token_accuracy": 0.8772289752960205, "num_tokens": 833175581.0, "step": 21940 }, { "epoch": 2.7911207225543824, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.079831123352051, "learning_rate": 1e-05, "loss": 0.5055, "mean_token_accuracy": 0.8675308227539062, "num_tokens": 833213466.0, "step": 21941 }, { "epoch": 2.791247932832973, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.7843451499938965, "learning_rate": 1e-05, "loss": 0.4912, "mean_token_accuracy": 0.8710470199584961, "num_tokens": 833250344.0, "step": 21942 }, { "epoch": 2.7913751431115634, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 5.943910598754883, "learning_rate": 1e-05, "loss": 0.4676, "mean_token_accuracy": 0.8761758804321289, "num_tokens": 833291032.0, "step": 21943 }, { "epoch": 2.791502353390154, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 7.0335307121276855, "learning_rate": 1e-05, "loss": 0.4747, "mean_token_accuracy": 0.8812750577926636, "num_tokens": 833331213.0, "step": 21944 }, { "epoch": 2.7916295636687445, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.785069942474365, "learning_rate": 1e-05, "loss": 0.4407, "mean_token_accuracy": 0.8846242427825928, "num_tokens": 833373906.0, "step": 21945 }, { "epoch": 2.791756773947335, "ewc_loss": 0.1484375, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00011920928955078125, "grad_norm": 7.4853434562683105, "learning_rate": 1e-05, "loss": 0.5187, "mean_token_accuracy": 0.8697472214698792, "num_tokens": 833411724.0, "step": 21946 }, { "epoch": 2.7918839842259255, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 5.821452617645264, "learning_rate": 1e-05, "loss": 0.4505, "mean_token_accuracy": 0.8859726786613464, "num_tokens": 833446337.0, "step": 21947 }, { "epoch": 2.792011194504516, "ewc_loss": 0.1640625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00013446807861328125, "grad_norm": 7.995014190673828, "learning_rate": 1e-05, "loss": 0.552, "mean_token_accuracy": 0.8643057346343994, "num_tokens": 833484444.0, "step": 21948 }, { "epoch": 2.7921384047831066, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.2866411209106445, "learning_rate": 1e-05, "loss": 0.4857, "mean_token_accuracy": 0.8791053295135498, "num_tokens": 833521375.0, "step": 21949 }, { "epoch": 2.792265615061697, "ewc_loss": 0.1650390625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.0001354217529296875, "grad_norm": 8.10059642791748, "learning_rate": 1e-05, "loss": 0.5026, "mean_token_accuracy": 0.8854501247406006, "num_tokens": 833557784.0, "step": 21950 }, { "epoch": 2.7923928253402877, "ewc_loss": 0.1494140625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00011968612670898438, "grad_norm": 6.867243766784668, "learning_rate": 1e-05, "loss": 0.4546, "mean_token_accuracy": 0.8951736688613892, "num_tokens": 833598872.0, "step": 21951 }, { "epoch": 2.792520035618878, "ewc_loss": 0.1533203125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.0001239776611328125, "grad_norm": 7.668532848358154, "learning_rate": 1e-05, "loss": 0.4814, "mean_token_accuracy": 0.885459303855896, "num_tokens": 833635799.0, "step": 21952 }, { "epoch": 2.7926472458974683, "ewc_loss": 0.1474609375, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.000118255615234375, "grad_norm": 6.984714031219482, "learning_rate": 1e-05, "loss": 0.4749, "mean_token_accuracy": 0.8836355805397034, "num_tokens": 833671814.0, "step": 21953 }, { "epoch": 2.7927744561760592, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00011396408081054688, "grad_norm": 7.167048454284668, "learning_rate": 1e-05, "loss": 0.5403, "mean_token_accuracy": 0.8659052848815918, "num_tokens": 833718081.0, "step": 21954 }, { "epoch": 2.7929016664546493, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.0001125335693359375, "grad_norm": 6.851810932159424, "learning_rate": 1e-05, "loss": 0.4961, "mean_token_accuracy": 0.874427080154419, "num_tokens": 833759197.0, "step": 21955 }, { "epoch": 2.7930288767332403, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.805426120758057, "learning_rate": 1e-05, "loss": 0.4531, "mean_token_accuracy": 0.884273111820221, "num_tokens": 833785582.0, "step": 21956 }, { "epoch": 2.7931560870118304, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.722634792327881, "learning_rate": 1e-05, "loss": 0.4723, "mean_token_accuracy": 0.883770227432251, "num_tokens": 833826611.0, "step": 21957 }, { "epoch": 2.793283297290421, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.53403377532959, "learning_rate": 1e-05, "loss": 0.4965, "mean_token_accuracy": 0.8723037838935852, "num_tokens": 833870657.0, "step": 21958 }, { "epoch": 2.7934105075690114, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.631746292114258, "learning_rate": 1e-05, "loss": 0.483, "mean_token_accuracy": 0.8767643570899963, "num_tokens": 833907621.0, "step": 21959 }, { "epoch": 2.793537717847602, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.385354042053223, "learning_rate": 1e-05, "loss": 0.4627, "mean_token_accuracy": 0.8822960257530212, "num_tokens": 833944364.0, "step": 21960 }, { "epoch": 2.7936649281261925, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.577641487121582, "learning_rate": 1e-05, "loss": 0.4472, "mean_token_accuracy": 0.8874738216400146, "num_tokens": 833983840.0, "step": 21961 }, { "epoch": 2.793792138404783, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.2785162925720215, "learning_rate": 1e-05, "loss": 0.4357, "mean_token_accuracy": 0.8903183341026306, "num_tokens": 834016367.0, "step": 21962 }, { "epoch": 2.7939193486833735, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.560207366943359, "learning_rate": 1e-05, "loss": 0.4475, "mean_token_accuracy": 0.8860107660293579, "num_tokens": 834050190.0, "step": 21963 }, { "epoch": 2.794046558961964, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.186978340148926, "learning_rate": 1e-05, "loss": 0.5033, "mean_token_accuracy": 0.8703423738479614, "num_tokens": 834087013.0, "step": 21964 }, { "epoch": 2.7941737692405546, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.579431533813477, "learning_rate": 1e-05, "loss": 0.4815, "mean_token_accuracy": 0.8754148483276367, "num_tokens": 834131771.0, "step": 21965 }, { "epoch": 2.794300979519145, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.092278003692627, "learning_rate": 1e-05, "loss": 0.4542, "mean_token_accuracy": 0.8834330439567566, "num_tokens": 834169055.0, "step": 21966 }, { "epoch": 2.7944281897977357, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.692968368530273, "learning_rate": 1e-05, "loss": 0.4637, "mean_token_accuracy": 0.8818121552467346, "num_tokens": 834206081.0, "step": 21967 }, { "epoch": 2.794555400076326, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 5.940064430236816, "learning_rate": 1e-05, "loss": 0.4422, "mean_token_accuracy": 0.885996401309967, "num_tokens": 834240846.0, "step": 21968 }, { "epoch": 2.7946826103549167, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.962987422943115, "learning_rate": 1e-05, "loss": 0.4843, "mean_token_accuracy": 0.8766877055168152, "num_tokens": 834275037.0, "step": 21969 }, { "epoch": 2.7948098206335072, "ewc_loss": 0.11376953125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 8.440017700195312e-05, "grad_norm": 5.746000289916992, "learning_rate": 1e-05, "loss": 0.4659, "mean_token_accuracy": 0.8786735534667969, "num_tokens": 834316408.0, "step": 21970 }, { "epoch": 2.7949370309120978, "ewc_loss": 0.1484375, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00011873245239257812, "grad_norm": 7.431403160095215, "learning_rate": 1e-05, "loss": 0.4667, "mean_token_accuracy": 0.887452244758606, "num_tokens": 834358473.0, "step": 21971 }, { "epoch": 2.7950642411906883, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.700570106506348, "learning_rate": 1e-05, "loss": 0.4734, "mean_token_accuracy": 0.8765588998794556, "num_tokens": 834399982.0, "step": 21972 }, { "epoch": 2.795191451469279, "ewc_loss": 0.166015625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00013637542724609375, "grad_norm": 8.102448463439941, "learning_rate": 1e-05, "loss": 0.4844, "mean_token_accuracy": 0.8859621286392212, "num_tokens": 834429731.0, "step": 21973 }, { "epoch": 2.7953186617478694, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.20623254776001, "learning_rate": 1e-05, "loss": 0.4958, "mean_token_accuracy": 0.8745933175086975, "num_tokens": 834476958.0, "step": 21974 }, { "epoch": 2.79544587202646, "ewc_loss": 0.171875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00014209747314453125, "grad_norm": 8.366281509399414, "learning_rate": 1e-05, "loss": 0.5115, "mean_token_accuracy": 0.8824810981750488, "num_tokens": 834516820.0, "step": 21975 }, { "epoch": 2.79557308230505, "ewc_loss": 0.154296875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00012493133544921875, "grad_norm": 7.014009952545166, "learning_rate": 1e-05, "loss": 0.509, "mean_token_accuracy": 0.877504825592041, "num_tokens": 834552455.0, "step": 21976 }, { "epoch": 2.795700292583641, "ewc_loss": 0.1591796875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00012969970703125, "grad_norm": 7.875816822052002, "learning_rate": 1e-05, "loss": 0.5223, "mean_token_accuracy": 0.8748584985733032, "num_tokens": 834593059.0, "step": 21977 }, { "epoch": 2.795827502862231, "ewc_loss": 0.15234375, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.135299205780029, "learning_rate": 1e-05, "loss": 0.5381, "mean_token_accuracy": 0.8674555420875549, "num_tokens": 834640134.0, "step": 21978 }, { "epoch": 2.795954713140822, "ewc_loss": 0.1455078125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.0001163482666015625, "grad_norm": 7.389063358306885, "learning_rate": 1e-05, "loss": 0.5085, "mean_token_accuracy": 0.874809741973877, "num_tokens": 834679172.0, "step": 21979 }, { "epoch": 2.796081923419412, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00011539459228515625, "grad_norm": 6.924404144287109, "learning_rate": 1e-05, "loss": 0.5332, "mean_token_accuracy": 0.8657031655311584, "num_tokens": 834713463.0, "step": 21980 }, { "epoch": 2.796209133698003, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00011014938354492188, "grad_norm": 7.034637451171875, "learning_rate": 1e-05, "loss": 0.5511, "mean_token_accuracy": 0.8549833297729492, "num_tokens": 834757782.0, "step": 21981 }, { "epoch": 2.796336343976593, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.710472106933594, "learning_rate": 1e-05, "loss": 0.5265, "mean_token_accuracy": 0.8691619634628296, "num_tokens": 834791598.0, "step": 21982 }, { "epoch": 2.7964635542551837, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.810275077819824, "learning_rate": 1e-05, "loss": 0.5196, "mean_token_accuracy": 0.8671591877937317, "num_tokens": 834835205.0, "step": 21983 }, { "epoch": 2.796590764533774, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.514819622039795, "learning_rate": 1e-05, "loss": 0.4449, "mean_token_accuracy": 0.8885149955749512, "num_tokens": 834867450.0, "step": 21984 }, { "epoch": 2.7967179748123647, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.710377216339111, "learning_rate": 1e-05, "loss": 0.4945, "mean_token_accuracy": 0.8765814304351807, "num_tokens": 834911067.0, "step": 21985 }, { "epoch": 2.7968451850909553, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.331024169921875, "learning_rate": 1e-05, "loss": 0.4659, "mean_token_accuracy": 0.877781331539154, "num_tokens": 834949117.0, "step": 21986 }, { "epoch": 2.796972395369546, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.736973285675049, "learning_rate": 1e-05, "loss": 0.499, "mean_token_accuracy": 0.8740836381912231, "num_tokens": 834988909.0, "step": 21987 }, { "epoch": 2.7970996056481363, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.1533098220825195, "learning_rate": 1e-05, "loss": 0.4565, "mean_token_accuracy": 0.8823508024215698, "num_tokens": 835031168.0, "step": 21988 }, { "epoch": 2.797226815926727, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.888343334197998, "learning_rate": 1e-05, "loss": 0.4599, "mean_token_accuracy": 0.8845280408859253, "num_tokens": 835072110.0, "step": 21989 }, { "epoch": 2.7973540262053174, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 5.974747180938721, "learning_rate": 1e-05, "loss": 0.4353, "mean_token_accuracy": 0.8905212879180908, "num_tokens": 835104605.0, "step": 21990 }, { "epoch": 2.797481236483908, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00011301040649414062, "grad_norm": 7.212351322174072, "learning_rate": 1e-05, "loss": 0.5495, "mean_token_accuracy": 0.8614219427108765, "num_tokens": 835141970.0, "step": 21991 }, { "epoch": 2.7976084467624984, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.879941940307617, "learning_rate": 1e-05, "loss": 0.5239, "mean_token_accuracy": 0.859602153301239, "num_tokens": 835177469.0, "step": 21992 }, { "epoch": 2.797735657041089, "ewc_loss": 0.15234375, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.651811599731445, "learning_rate": 1e-05, "loss": 0.5188, "mean_token_accuracy": 0.8736718893051147, "num_tokens": 835220286.0, "step": 21993 }, { "epoch": 2.7978628673196795, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.050374507904053, "learning_rate": 1e-05, "loss": 0.4561, "mean_token_accuracy": 0.880050778388977, "num_tokens": 835254290.0, "step": 21994 }, { "epoch": 2.79799007759827, "ewc_loss": 0.1611328125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.0001316070556640625, "grad_norm": 7.987701892852783, "learning_rate": 1e-05, "loss": 0.513, "mean_token_accuracy": 0.8732874393463135, "num_tokens": 835286375.0, "step": 21995 }, { "epoch": 2.7981172878768605, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 6.605919361114502, "learning_rate": 1e-05, "loss": 0.4853, "mean_token_accuracy": 0.8779974579811096, "num_tokens": 835315472.0, "step": 21996 }, { "epoch": 2.798244498155451, "ewc_loss": 0.1552734375, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.000125885009765625, "grad_norm": 7.725809574127197, "learning_rate": 1e-05, "loss": 0.4808, "mean_token_accuracy": 0.8849948048591614, "num_tokens": 835355865.0, "step": 21997 }, { "epoch": 2.7983717084340416, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00011491775512695312, "grad_norm": 6.936537265777588, "learning_rate": 1e-05, "loss": 0.5034, "mean_token_accuracy": 0.8772895336151123, "num_tokens": 835391251.0, "step": 21998 }, { "epoch": 2.798498918712632, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00011301040649414062, "grad_norm": 7.176848411560059, "learning_rate": 1e-05, "loss": 0.4539, "mean_token_accuracy": 0.8886123299598694, "num_tokens": 835424869.0, "step": 21999 }, { "epoch": 2.7986261289912227, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00011396408081054688, "grad_norm": 6.9346160888671875, "learning_rate": 1e-05, "loss": 0.4958, "mean_token_accuracy": 0.878315806388855, "num_tokens": 835464916.0, "step": 22000 }, { "epoch": 2.7987533392698127, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.653925895690918, "learning_rate": 1e-05, "loss": 0.5124, "mean_token_accuracy": 0.8687300086021423, "num_tokens": 835503114.0, "step": 22001 }, { "epoch": 2.7988805495484037, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 6.942263126373291, "learning_rate": 1e-05, "loss": 0.5339, "mean_token_accuracy": 0.8668146133422852, "num_tokens": 835541446.0, "step": 22002 }, { "epoch": 2.799007759826994, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.256176948547363, "learning_rate": 1e-05, "loss": 0.4305, "mean_token_accuracy": 0.892875075340271, "num_tokens": 835579261.0, "step": 22003 }, { "epoch": 2.7991349701055848, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00011014938354492188, "grad_norm": 7.080399036407471, "learning_rate": 1e-05, "loss": 0.5016, "mean_token_accuracy": 0.8739114999771118, "num_tokens": 835615847.0, "step": 22004 }, { "epoch": 2.799262180384175, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.020056247711182, "learning_rate": 1e-05, "loss": 0.5326, "mean_token_accuracy": 0.8575717806816101, "num_tokens": 835658461.0, "step": 22005 }, { "epoch": 2.799389390662766, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00011539459228515625, "grad_norm": 7.338898658752441, "learning_rate": 1e-05, "loss": 0.4602, "mean_token_accuracy": 0.8845233917236328, "num_tokens": 835693489.0, "step": 22006 }, { "epoch": 2.799516600941356, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 5.976139068603516, "learning_rate": 1e-05, "loss": 0.4265, "mean_token_accuracy": 0.8913968801498413, "num_tokens": 835733925.0, "step": 22007 }, { "epoch": 2.7996438112199464, "ewc_loss": 0.15234375, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.652846336364746, "learning_rate": 1e-05, "loss": 0.5314, "mean_token_accuracy": 0.8703495264053345, "num_tokens": 835775071.0, "step": 22008 }, { "epoch": 2.799771021498537, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.209137916564941, "learning_rate": 1e-05, "loss": 0.4917, "mean_token_accuracy": 0.8742400407791138, "num_tokens": 835813940.0, "step": 22009 }, { "epoch": 2.7998982317771275, "ewc_loss": 0.154296875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00012493133544921875, "grad_norm": 7.774383544921875, "learning_rate": 1e-05, "loss": 0.555, "mean_token_accuracy": 0.8623079657554626, "num_tokens": 835849667.0, "step": 22010 }, { "epoch": 2.800025442055718, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.553412914276123, "learning_rate": 1e-05, "loss": 0.4702, "mean_token_accuracy": 0.8814082741737366, "num_tokens": 835880928.0, "step": 22011 }, { "epoch": 2.8001526523343085, "ewc_loss": 0.15234375, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.557417869567871, "learning_rate": 1e-05, "loss": 0.4478, "mean_token_accuracy": 0.8947564363479614, "num_tokens": 835917502.0, "step": 22012 }, { "epoch": 2.800279862612899, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00011014938354492188, "grad_norm": 6.694137096405029, "learning_rate": 1e-05, "loss": 0.5222, "mean_token_accuracy": 0.8666608333587646, "num_tokens": 835958547.0, "step": 22013 }, { "epoch": 2.8004070728914896, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00011444091796875, "grad_norm": 7.214412212371826, "learning_rate": 1e-05, "loss": 0.5081, "mean_token_accuracy": 0.8754910230636597, "num_tokens": 835997427.0, "step": 22014 }, { "epoch": 2.80053428317008, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.648787498474121, "learning_rate": 1e-05, "loss": 0.4632, "mean_token_accuracy": 0.8831018209457397, "num_tokens": 836039096.0, "step": 22015 }, { "epoch": 2.8006614934486707, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.927547454833984, "learning_rate": 1e-05, "loss": 0.4936, "mean_token_accuracy": 0.8760491609573364, "num_tokens": 836079613.0, "step": 22016 }, { "epoch": 2.800788703727261, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.553168773651123, "learning_rate": 1e-05, "loss": 0.4661, "mean_token_accuracy": 0.8818770051002502, "num_tokens": 836117500.0, "step": 22017 }, { "epoch": 2.8009159140058517, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.755949974060059, "learning_rate": 1e-05, "loss": 0.4756, "mean_token_accuracy": 0.8778780698776245, "num_tokens": 836149680.0, "step": 22018 }, { "epoch": 2.8010431242844422, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.414767742156982, "learning_rate": 1e-05, "loss": 0.4183, "mean_token_accuracy": 0.8952935338020325, "num_tokens": 836179869.0, "step": 22019 }, { "epoch": 2.8011703345630328, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.707897663116455, "learning_rate": 1e-05, "loss": 0.491, "mean_token_accuracy": 0.8725247383117676, "num_tokens": 836222305.0, "step": 22020 }, { "epoch": 2.8012975448416233, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.263605117797852, "learning_rate": 1e-05, "loss": 0.4922, "mean_token_accuracy": 0.8728158473968506, "num_tokens": 836257687.0, "step": 22021 }, { "epoch": 2.801424755120214, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.7735819816589355, "learning_rate": 1e-05, "loss": 0.489, "mean_token_accuracy": 0.8751621246337891, "num_tokens": 836299105.0, "step": 22022 }, { "epoch": 2.8015519653988044, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.102643013000488, "learning_rate": 1e-05, "loss": 0.4595, "mean_token_accuracy": 0.8795875906944275, "num_tokens": 836337661.0, "step": 22023 }, { "epoch": 2.801679175677395, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.938394546508789, "learning_rate": 1e-05, "loss": 0.5215, "mean_token_accuracy": 0.8679947257041931, "num_tokens": 836378792.0, "step": 22024 }, { "epoch": 2.8018063859559854, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 5.984444618225098, "learning_rate": 1e-05, "loss": 0.4826, "mean_token_accuracy": 0.8729751110076904, "num_tokens": 836412754.0, "step": 22025 }, { "epoch": 2.8019335962345755, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00011301040649414062, "grad_norm": 7.21511173248291, "learning_rate": 1e-05, "loss": 0.4726, "mean_token_accuracy": 0.8826790452003479, "num_tokens": 836452173.0, "step": 22026 }, { "epoch": 2.8020608065131665, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 5.949811935424805, "learning_rate": 1e-05, "loss": 0.444, "mean_token_accuracy": 0.8859155178070068, "num_tokens": 836491398.0, "step": 22027 }, { "epoch": 2.8021880167917566, "ewc_loss": 0.1494140625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.0001201629638671875, "grad_norm": 7.534552574157715, "learning_rate": 1e-05, "loss": 0.5365, "mean_token_accuracy": 0.868032693862915, "num_tokens": 836525130.0, "step": 22028 }, { "epoch": 2.8023152270703475, "ewc_loss": 0.125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.149692058563232, "learning_rate": 1e-05, "loss": 0.489, "mean_token_accuracy": 0.8743699193000793, "num_tokens": 836561288.0, "step": 22029 }, { "epoch": 2.8024424373489376, "ewc_loss": 0.1513671875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.0001220703125, "grad_norm": 7.644100189208984, "learning_rate": 1e-05, "loss": 0.5569, "mean_token_accuracy": 0.8614175319671631, "num_tokens": 836603595.0, "step": 22030 }, { "epoch": 2.8025696476275286, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.500268459320068, "learning_rate": 1e-05, "loss": 0.5224, "mean_token_accuracy": 0.8699421286582947, "num_tokens": 836640460.0, "step": 22031 }, { "epoch": 2.8026968579061187, "ewc_loss": 0.1455078125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.0001163482666015625, "grad_norm": 7.367671012878418, "learning_rate": 1e-05, "loss": 0.532, "mean_token_accuracy": 0.8678305149078369, "num_tokens": 836678024.0, "step": 22032 }, { "epoch": 2.802824068184709, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 6.695784091949463, "learning_rate": 1e-05, "loss": 0.4472, "mean_token_accuracy": 0.892900824546814, "num_tokens": 836718383.0, "step": 22033 }, { "epoch": 2.8029512784632997, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.881775379180908, "learning_rate": 1e-05, "loss": 0.523, "mean_token_accuracy": 0.8638865947723389, "num_tokens": 836755564.0, "step": 22034 }, { "epoch": 2.8030784887418903, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 6.783295154571533, "learning_rate": 1e-05, "loss": 0.4972, "mean_token_accuracy": 0.8742839097976685, "num_tokens": 836792891.0, "step": 22035 }, { "epoch": 2.803205699020481, "ewc_loss": 0.125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.379225254058838, "learning_rate": 1e-05, "loss": 0.4894, "mean_token_accuracy": 0.8726301193237305, "num_tokens": 836830024.0, "step": 22036 }, { "epoch": 2.8033329092990713, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 6.937722682952881, "learning_rate": 1e-05, "loss": 0.5194, "mean_token_accuracy": 0.8683497905731201, "num_tokens": 836871600.0, "step": 22037 }, { "epoch": 2.803460119577662, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 5.990676403045654, "learning_rate": 1e-05, "loss": 0.4371, "mean_token_accuracy": 0.8841661810874939, "num_tokens": 836911495.0, "step": 22038 }, { "epoch": 2.8035873298562524, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00011396408081054688, "grad_norm": 7.281315803527832, "learning_rate": 1e-05, "loss": 0.4996, "mean_token_accuracy": 0.8783754706382751, "num_tokens": 836952840.0, "step": 22039 }, { "epoch": 2.803714540134843, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.85404109954834, "learning_rate": 1e-05, "loss": 0.4802, "mean_token_accuracy": 0.873064398765564, "num_tokens": 836988355.0, "step": 22040 }, { "epoch": 2.8038417504134334, "ewc_loss": 0.1552734375, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.000125885009765625, "grad_norm": 7.7154316902160645, "learning_rate": 1e-05, "loss": 0.502, "mean_token_accuracy": 0.8778579235076904, "num_tokens": 837021559.0, "step": 22041 }, { "epoch": 2.803968960692024, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.10205602645874, "learning_rate": 1e-05, "loss": 0.4869, "mean_token_accuracy": 0.8723751902580261, "num_tokens": 837060401.0, "step": 22042 }, { "epoch": 2.8040961709706145, "ewc_loss": 0.162109375, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00013256072998046875, "grad_norm": 7.926985263824463, "learning_rate": 1e-05, "loss": 0.4816, "mean_token_accuracy": 0.8864753246307373, "num_tokens": 837094769.0, "step": 22043 }, { "epoch": 2.804223381249205, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 6.597733974456787, "learning_rate": 1e-05, "loss": 0.4455, "mean_token_accuracy": 0.8913515210151672, "num_tokens": 837125820.0, "step": 22044 }, { "epoch": 2.8043505915277955, "ewc_loss": 0.154296875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00012493133544921875, "grad_norm": 7.680946350097656, "learning_rate": 1e-05, "loss": 0.4855, "mean_token_accuracy": 0.885094165802002, "num_tokens": 837167392.0, "step": 22045 }, { "epoch": 2.804477801806386, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00011348724365234375, "grad_norm": 6.808934211730957, "learning_rate": 1e-05, "loss": 0.4891, "mean_token_accuracy": 0.8766103386878967, "num_tokens": 837207439.0, "step": 22046 }, { "epoch": 2.8046050120849766, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00011539459228515625, "grad_norm": 7.273232460021973, "learning_rate": 1e-05, "loss": 0.4931, "mean_token_accuracy": 0.8820850849151611, "num_tokens": 837240038.0, "step": 22047 }, { "epoch": 2.804732222363567, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 6.745309829711914, "learning_rate": 1e-05, "loss": 0.5178, "mean_token_accuracy": 0.8712875247001648, "num_tokens": 837278842.0, "step": 22048 }, { "epoch": 2.8048594326421576, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.924975872039795, "learning_rate": 1e-05, "loss": 0.437, "mean_token_accuracy": 0.8940989375114441, "num_tokens": 837317519.0, "step": 22049 }, { "epoch": 2.804986642920748, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.615024089813232, "learning_rate": 1e-05, "loss": 0.4522, "mean_token_accuracy": 0.8860559463500977, "num_tokens": 837349053.0, "step": 22050 }, { "epoch": 2.8051138531993383, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.728769302368164, "learning_rate": 1e-05, "loss": 0.4871, "mean_token_accuracy": 0.8742890954017639, "num_tokens": 837386018.0, "step": 22051 }, { "epoch": 2.8052410634779292, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.461709499359131, "learning_rate": 1e-05, "loss": 0.5, "mean_token_accuracy": 0.8708096742630005, "num_tokens": 837423640.0, "step": 22052 }, { "epoch": 2.8053682737565193, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.664649963378906, "learning_rate": 1e-05, "loss": 0.4755, "mean_token_accuracy": 0.877799928188324, "num_tokens": 837462198.0, "step": 22053 }, { "epoch": 2.8054954840351103, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.317789077758789, "learning_rate": 1e-05, "loss": 0.4827, "mean_token_accuracy": 0.8741120100021362, "num_tokens": 837492722.0, "step": 22054 }, { "epoch": 2.8056226943137004, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.701374053955078, "learning_rate": 1e-05, "loss": 0.4707, "mean_token_accuracy": 0.8822820782661438, "num_tokens": 837531598.0, "step": 22055 }, { "epoch": 2.805749904592291, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.13165283203125, "learning_rate": 1e-05, "loss": 0.4936, "mean_token_accuracy": 0.8701567649841309, "num_tokens": 837572580.0, "step": 22056 }, { "epoch": 2.8058771148708814, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.875533103942871, "learning_rate": 1e-05, "loss": 0.4629, "mean_token_accuracy": 0.8850783109664917, "num_tokens": 837610398.0, "step": 22057 }, { "epoch": 2.806004325149472, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.968575477600098, "learning_rate": 1e-05, "loss": 0.4722, "mean_token_accuracy": 0.874201238155365, "num_tokens": 837653518.0, "step": 22058 }, { "epoch": 2.8061315354280625, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00011301040649414062, "grad_norm": 7.171384811401367, "learning_rate": 1e-05, "loss": 0.5073, "mean_token_accuracy": 0.8744126558303833, "num_tokens": 837697397.0, "step": 22059 }, { "epoch": 2.806258745706653, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 5.881038188934326, "learning_rate": 1e-05, "loss": 0.4181, "mean_token_accuracy": 0.8932408094406128, "num_tokens": 837736565.0, "step": 22060 }, { "epoch": 2.8063859559852435, "ewc_loss": 0.150390625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00012063980102539062, "grad_norm": 7.583161354064941, "learning_rate": 1e-05, "loss": 0.5018, "mean_token_accuracy": 0.8779780268669128, "num_tokens": 837772514.0, "step": 22061 }, { "epoch": 2.806513166263834, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.0527567863464355, "learning_rate": 1e-05, "loss": 0.4428, "mean_token_accuracy": 0.8859160542488098, "num_tokens": 837805147.0, "step": 22062 }, { "epoch": 2.8066403765424246, "ewc_loss": 0.1591796875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00012969970703125, "grad_norm": 7.843029499053955, "learning_rate": 1e-05, "loss": 0.557, "mean_token_accuracy": 0.864378809928894, "num_tokens": 837846795.0, "step": 22063 }, { "epoch": 2.806767586821015, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.498024940490723, "learning_rate": 1e-05, "loss": 0.4818, "mean_token_accuracy": 0.8798314332962036, "num_tokens": 837882703.0, "step": 22064 }, { "epoch": 2.8068947970996057, "ewc_loss": 0.1533203125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 0.0001239776611328125, "grad_norm": 7.6399126052856445, "learning_rate": 1e-05, "loss": 0.5542, "mean_token_accuracy": 0.8614280819892883, "num_tokens": 837924698.0, "step": 22065 }, { "epoch": 2.807022007378196, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00011348724365234375, "grad_norm": 6.798399925231934, "learning_rate": 1e-05, "loss": 0.4843, "mean_token_accuracy": 0.879766583442688, "num_tokens": 837958553.0, "step": 22066 }, { "epoch": 2.8071492176567867, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00011301040649414062, "grad_norm": 7.1407551765441895, "learning_rate": 1e-05, "loss": 0.4228, "mean_token_accuracy": 0.9009919166564941, "num_tokens": 837998674.0, "step": 22067 }, { "epoch": 2.8072764279353772, "ewc_loss": 0.140625, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 6.829630374908447, "learning_rate": 1e-05, "loss": 0.4761, "mean_token_accuracy": 0.877659797668457, "num_tokens": 838035111.0, "step": 22068 }, { "epoch": 2.8074036382139678, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.644710540771484, "learning_rate": 1e-05, "loss": 0.5003, "mean_token_accuracy": 0.8736408948898315, "num_tokens": 838075444.0, "step": 22069 }, { "epoch": 2.8075308484925583, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 6.866456985473633, "learning_rate": 1e-05, "loss": 0.5659, "mean_token_accuracy": 0.8569477796554565, "num_tokens": 838117972.0, "step": 22070 }, { "epoch": 2.807658058771149, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.252156734466553, "learning_rate": 1e-05, "loss": 0.5062, "mean_token_accuracy": 0.8659668564796448, "num_tokens": 838158797.0, "step": 22071 }, { "epoch": 2.8077852690497394, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 7.026554107666016, "learning_rate": 1e-05, "loss": 0.5148, "mean_token_accuracy": 0.8701122999191284, "num_tokens": 838198958.0, "step": 22072 }, { "epoch": 2.80791247932833, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 5.995102882385254, "learning_rate": 1e-05, "loss": 0.476, "mean_token_accuracy": 0.873626172542572, "num_tokens": 838233936.0, "step": 22073 }, { "epoch": 2.80803968960692, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00011396408081054688, "grad_norm": 7.30249547958374, "learning_rate": 1e-05, "loss": 0.5109, "mean_token_accuracy": 0.8715337514877319, "num_tokens": 838274333.0, "step": 22074 }, { "epoch": 2.808166899885511, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 5.9530439376831055, "learning_rate": 1e-05, "loss": 0.4585, "mean_token_accuracy": 0.8799258470535278, "num_tokens": 838304932.0, "step": 22075 }, { "epoch": 2.808294110164101, "ewc_loss": 0.1513671875, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.0001220703125, "grad_norm": 7.62188196182251, "learning_rate": 1e-05, "loss": 0.5492, "mean_token_accuracy": 0.8665632009506226, "num_tokens": 838345377.0, "step": 22076 }, { "epoch": 2.808421320442692, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.14927339553833, "learning_rate": 1e-05, "loss": 0.4977, "mean_token_accuracy": 0.8757295608520508, "num_tokens": 838383105.0, "step": 22077 }, { "epoch": 2.808548530721282, "ewc_loss": 0.1572265625, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.0001277923583984375, "grad_norm": 7.756035804748535, "learning_rate": 1e-05, "loss": 0.5502, "mean_token_accuracy": 0.8634848594665527, "num_tokens": 838421857.0, "step": 22078 }, { "epoch": 2.808675740999873, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.490825176239014, "learning_rate": 1e-05, "loss": 0.5289, "mean_token_accuracy": 0.8650770783424377, "num_tokens": 838459439.0, "step": 22079 }, { "epoch": 2.808802951278463, "ewc_loss": 0.150390625, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00012111663818359375, "grad_norm": 7.5670013427734375, "learning_rate": 1e-05, "loss": 0.5253, "mean_token_accuracy": 0.8737468719482422, "num_tokens": 838501176.0, "step": 22080 }, { "epoch": 2.8089301615570537, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00010967254638671875, "grad_norm": 6.678652763366699, "learning_rate": 1e-05, "loss": 0.5291, "mean_token_accuracy": 0.862861156463623, "num_tokens": 838538725.0, "step": 22081 }, { "epoch": 2.809057371835644, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00011491775512695312, "grad_norm": 7.225277423858643, "learning_rate": 1e-05, "loss": 0.5077, "mean_token_accuracy": 0.8707799911499023, "num_tokens": 838577365.0, "step": 22082 }, { "epoch": 2.8091845821142347, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 6.658432483673096, "learning_rate": 1e-05, "loss": 0.474, "mean_token_accuracy": 0.8826254606246948, "num_tokens": 838617872.0, "step": 22083 }, { "epoch": 2.8093117923928252, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.9140143394470215, "learning_rate": 1e-05, "loss": 0.5027, "mean_token_accuracy": 0.8711949586868286, "num_tokens": 838656599.0, "step": 22084 }, { "epoch": 2.8094390026714158, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.574695587158203, "learning_rate": 1e-05, "loss": 0.4947, "mean_token_accuracy": 0.875740647315979, "num_tokens": 838693270.0, "step": 22085 }, { "epoch": 2.8095662129500063, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.707168102264404, "learning_rate": 1e-05, "loss": 0.5084, "mean_token_accuracy": 0.8706223964691162, "num_tokens": 838728493.0, "step": 22086 }, { "epoch": 2.809693423228597, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.442393779754639, "learning_rate": 1e-05, "loss": 0.4746, "mean_token_accuracy": 0.8773390650749207, "num_tokens": 838767136.0, "step": 22087 }, { "epoch": 2.8098206335071874, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.645091533660889, "learning_rate": 1e-05, "loss": 0.4868, "mean_token_accuracy": 0.8748468160629272, "num_tokens": 838801894.0, "step": 22088 }, { "epoch": 2.809947843785778, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.286499977111816, "learning_rate": 1e-05, "loss": 0.4169, "mean_token_accuracy": 0.8969501256942749, "num_tokens": 838838732.0, "step": 22089 }, { "epoch": 2.8100750540643684, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.703954219818115, "learning_rate": 1e-05, "loss": 0.4136, "mean_token_accuracy": 0.9001982808113098, "num_tokens": 838876331.0, "step": 22090 }, { "epoch": 2.810202264342959, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.119691848754883, "learning_rate": 1e-05, "loss": 0.457, "mean_token_accuracy": 0.8798649907112122, "num_tokens": 838909613.0, "step": 22091 }, { "epoch": 2.8103294746215495, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.86441707611084, "learning_rate": 1e-05, "loss": 0.5047, "mean_token_accuracy": 0.8669604063034058, "num_tokens": 838944694.0, "step": 22092 }, { "epoch": 2.81045668490014, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 5.9685444831848145, "learning_rate": 1e-05, "loss": 0.4567, "mean_token_accuracy": 0.8825621604919434, "num_tokens": 838984595.0, "step": 22093 }, { "epoch": 2.8105838951787305, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00011205673217773438, "grad_norm": 7.149906158447266, "learning_rate": 1e-05, "loss": 0.4923, "mean_token_accuracy": 0.873518705368042, "num_tokens": 839018528.0, "step": 22094 }, { "epoch": 2.810711105457321, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 5.887699127197266, "learning_rate": 1e-05, "loss": 0.4767, "mean_token_accuracy": 0.8744577765464783, "num_tokens": 839056447.0, "step": 22095 }, { "epoch": 2.8108383157359116, "ewc_loss": 0.150390625, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00012063980102539062, "grad_norm": 7.529163360595703, "learning_rate": 1e-05, "loss": 0.4802, "mean_token_accuracy": 0.8835119009017944, "num_tokens": 839093612.0, "step": 22096 }, { "epoch": 2.810965526014502, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.043135166168213, "learning_rate": 1e-05, "loss": 0.4763, "mean_token_accuracy": 0.8757805228233337, "num_tokens": 839130192.0, "step": 22097 }, { "epoch": 2.8110927362930926, "ewc_loss": 0.1572265625, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.0001277923583984375, "grad_norm": 7.768594264984131, "learning_rate": 1e-05, "loss": 0.5205, "mean_token_accuracy": 0.8719673752784729, "num_tokens": 839167821.0, "step": 22098 }, { "epoch": 2.8112199465716827, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.4490227699279785, "learning_rate": 1e-05, "loss": 0.5166, "mean_token_accuracy": 0.8666307926177979, "num_tokens": 839200124.0, "step": 22099 }, { "epoch": 2.8113471568502737, "ewc_loss": 0.15234375, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.580865859985352, "learning_rate": 1e-05, "loss": 0.5162, "mean_token_accuracy": 0.8726643919944763, "num_tokens": 839242787.0, "step": 22100 }, { "epoch": 2.811474367128864, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00011157989501953125, "grad_norm": 6.725535869598389, "learning_rate": 1e-05, "loss": 0.4509, "mean_token_accuracy": 0.8913450837135315, "num_tokens": 839277517.0, "step": 22101 }, { "epoch": 2.8116015774074548, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00011301040649414062, "grad_norm": 7.115062713623047, "learning_rate": 1e-05, "loss": 0.4554, "mean_token_accuracy": 0.8904637694358826, "num_tokens": 839318609.0, "step": 22102 }, { "epoch": 2.811728787686045, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00011014938354492188, "grad_norm": 6.800244331359863, "learning_rate": 1e-05, "loss": 0.5154, "mean_token_accuracy": 0.8713502287864685, "num_tokens": 839353964.0, "step": 22103 }, { "epoch": 2.811855997964636, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.593451023101807, "learning_rate": 1e-05, "loss": 0.4361, "mean_token_accuracy": 0.8925583362579346, "num_tokens": 839393926.0, "step": 22104 }, { "epoch": 2.811983208243226, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 6.870194435119629, "learning_rate": 1e-05, "loss": 0.4914, "mean_token_accuracy": 0.8756679892539978, "num_tokens": 839437857.0, "step": 22105 }, { "epoch": 2.8121104185218164, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.19980525970459, "learning_rate": 1e-05, "loss": 0.5092, "mean_token_accuracy": 0.8665663003921509, "num_tokens": 839478444.0, "step": 22106 }, { "epoch": 2.812237628800407, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00010967254638671875, "grad_norm": 7.0672783851623535, "learning_rate": 1e-05, "loss": 0.4686, "mean_token_accuracy": 0.8802724480628967, "num_tokens": 839512752.0, "step": 22107 }, { "epoch": 2.8123648390789975, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 5.969904899597168, "learning_rate": 1e-05, "loss": 0.5094, "mean_token_accuracy": 0.8631536364555359, "num_tokens": 839550024.0, "step": 22108 }, { "epoch": 2.812492049357588, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00011539459228515625, "grad_norm": 7.362431049346924, "learning_rate": 1e-05, "loss": 0.4712, "mean_token_accuracy": 0.884314775466919, "num_tokens": 839582612.0, "step": 22109 }, { "epoch": 2.8126192596361785, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 5.966428756713867, "learning_rate": 1e-05, "loss": 0.4588, "mean_token_accuracy": 0.8801103830337524, "num_tokens": 839621791.0, "step": 22110 }, { "epoch": 2.812746469914769, "ewc_loss": 0.154296875, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00012493133544921875, "grad_norm": 7.672264575958252, "learning_rate": 1e-05, "loss": 0.5528, "mean_token_accuracy": 0.8629511594772339, "num_tokens": 839660829.0, "step": 22111 }, { "epoch": 2.8128736801933596, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.207979202270508, "learning_rate": 1e-05, "loss": 0.4811, "mean_token_accuracy": 0.8783012628555298, "num_tokens": 839701675.0, "step": 22112 }, { "epoch": 2.81300089047195, "ewc_loss": 0.15625, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00012683868408203125, "grad_norm": 7.75618314743042, "learning_rate": 1e-05, "loss": 0.5407, "mean_token_accuracy": 0.867850661277771, "num_tokens": 839742958.0, "step": 22113 }, { "epoch": 2.8131281007505406, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.535109996795654, "learning_rate": 1e-05, "loss": 0.4904, "mean_token_accuracy": 0.877060055732727, "num_tokens": 839787552.0, "step": 22114 }, { "epoch": 2.813255311029131, "ewc_loss": 0.1513671875, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00012159347534179688, "grad_norm": 7.5384297370910645, "learning_rate": 1e-05, "loss": 0.5014, "mean_token_accuracy": 0.8754552602767944, "num_tokens": 839824630.0, "step": 22115 }, { "epoch": 2.8133825213077217, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 6.691601276397705, "learning_rate": 1e-05, "loss": 0.459, "mean_token_accuracy": 0.887711226940155, "num_tokens": 839857671.0, "step": 22116 }, { "epoch": 2.8135097315863122, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00011491775512695312, "grad_norm": 7.197230339050293, "learning_rate": 1e-05, "loss": 0.5012, "mean_token_accuracy": 0.8752086162567139, "num_tokens": 839896395.0, "step": 22117 }, { "epoch": 2.8136369418649028, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.640022277832031, "learning_rate": 1e-05, "loss": 0.4907, "mean_token_accuracy": 0.8768869638442993, "num_tokens": 839942188.0, "step": 22118 }, { "epoch": 2.8137641521434933, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.913860321044922, "learning_rate": 1e-05, "loss": 0.4281, "mean_token_accuracy": 0.8919529318809509, "num_tokens": 839970510.0, "step": 22119 }, { "epoch": 2.813891362422084, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.555217266082764, "learning_rate": 1e-05, "loss": 0.4752, "mean_token_accuracy": 0.8775290250778198, "num_tokens": 840007458.0, "step": 22120 }, { "epoch": 2.8140185727006743, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.7236199378967285, "learning_rate": 1e-05, "loss": 0.4869, "mean_token_accuracy": 0.8751512765884399, "num_tokens": 840049729.0, "step": 22121 }, { "epoch": 2.814145782979265, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.4268798828125, "learning_rate": 1e-05, "loss": 0.4816, "mean_token_accuracy": 0.8786683082580566, "num_tokens": 840089883.0, "step": 22122 }, { "epoch": 2.8142729932578554, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.660601615905762, "learning_rate": 1e-05, "loss": 0.4338, "mean_token_accuracy": 0.8938642144203186, "num_tokens": 840121237.0, "step": 22123 }, { "epoch": 2.8144002035364455, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.30225944519043, "learning_rate": 1e-05, "loss": 0.5043, "mean_token_accuracy": 0.8671183586120605, "num_tokens": 840157213.0, "step": 22124 }, { "epoch": 2.8145274138150365, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.695245265960693, "learning_rate": 1e-05, "loss": 0.4659, "mean_token_accuracy": 0.8822720646858215, "num_tokens": 840191404.0, "step": 22125 }, { "epoch": 2.8146546240936265, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.150747299194336, "learning_rate": 1e-05, "loss": 0.5072, "mean_token_accuracy": 0.8699374198913574, "num_tokens": 840227803.0, "step": 22126 }, { "epoch": 2.8147818343722175, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.837676525115967, "learning_rate": 1e-05, "loss": 0.5577, "mean_token_accuracy": 0.8552316427230835, "num_tokens": 840266686.0, "step": 22127 }, { "epoch": 2.8149090446508076, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 5.990301132202148, "learning_rate": 1e-05, "loss": 0.4492, "mean_token_accuracy": 0.8842054605484009, "num_tokens": 840310437.0, "step": 22128 }, { "epoch": 2.815036254929398, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00011157989501953125, "grad_norm": 7.110852241516113, "learning_rate": 1e-05, "loss": 0.4985, "mean_token_accuracy": 0.8764806389808655, "num_tokens": 840349599.0, "step": 22129 }, { "epoch": 2.8151634652079887, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 5.89939022064209, "learning_rate": 1e-05, "loss": 0.4594, "mean_token_accuracy": 0.8815139532089233, "num_tokens": 840389143.0, "step": 22130 }, { "epoch": 2.815290675486579, "ewc_loss": 0.1474609375, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00011777877807617188, "grad_norm": 7.457857131958008, "learning_rate": 1e-05, "loss": 0.4826, "mean_token_accuracy": 0.8834008574485779, "num_tokens": 840429440.0, "step": 22131 }, { "epoch": 2.8154178857651697, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.014801025390625, "learning_rate": 1e-05, "loss": 0.4808, "mean_token_accuracy": 0.8767880797386169, "num_tokens": 840466056.0, "step": 22132 }, { "epoch": 2.8155450960437602, "ewc_loss": 0.154296875, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00012493133544921875, "grad_norm": 7.73131799697876, "learning_rate": 1e-05, "loss": 0.5085, "mean_token_accuracy": 0.8759715557098389, "num_tokens": 840503039.0, "step": 22133 }, { "epoch": 2.8156723063223508, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.368542194366455, "learning_rate": 1e-05, "loss": 0.4763, "mean_token_accuracy": 0.8806612491607666, "num_tokens": 840546809.0, "step": 22134 }, { "epoch": 2.8157995166009413, "ewc_loss": 0.1533203125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.0001239776611328125, "grad_norm": 7.61838436126709, "learning_rate": 1e-05, "loss": 0.5161, "mean_token_accuracy": 0.8740012645721436, "num_tokens": 840582617.0, "step": 22135 }, { "epoch": 2.815926726879532, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00010967254638671875, "grad_norm": 6.686959266662598, "learning_rate": 1e-05, "loss": 0.4874, "mean_token_accuracy": 0.8792346119880676, "num_tokens": 840620279.0, "step": 22136 }, { "epoch": 2.8160539371581224, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00011444091796875, "grad_norm": 7.188676834106445, "learning_rate": 1e-05, "loss": 0.4899, "mean_token_accuracy": 0.8798345327377319, "num_tokens": 840655925.0, "step": 22137 }, { "epoch": 2.816181147436713, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00011014938354492188, "grad_norm": 6.7857160568237305, "learning_rate": 1e-05, "loss": 0.4866, "mean_token_accuracy": 0.8780933618545532, "num_tokens": 840692328.0, "step": 22138 }, { "epoch": 2.8163083577153034, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.672555446624756, "learning_rate": 1e-05, "loss": 0.4741, "mean_token_accuracy": 0.8786205053329468, "num_tokens": 840730674.0, "step": 22139 }, { "epoch": 2.816435567993894, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.854317665100098, "learning_rate": 1e-05, "loss": 0.504, "mean_token_accuracy": 0.8706185221672058, "num_tokens": 840764791.0, "step": 22140 }, { "epoch": 2.8165627782724845, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.229630470275879, "learning_rate": 1e-05, "loss": 0.4649, "mean_token_accuracy": 0.878352701663971, "num_tokens": 840803527.0, "step": 22141 }, { "epoch": 2.816689988551075, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 7.056044101715088, "learning_rate": 1e-05, "loss": 0.5113, "mean_token_accuracy": 0.8691196441650391, "num_tokens": 840843823.0, "step": 22142 }, { "epoch": 2.8168171988296655, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 5.959192752838135, "learning_rate": 1e-05, "loss": 0.5009, "mean_token_accuracy": 0.8671090602874756, "num_tokens": 840879648.0, "step": 22143 }, { "epoch": 2.816944409108256, "ewc_loss": 0.1455078125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00011587142944335938, "grad_norm": 7.37850284576416, "learning_rate": 1e-05, "loss": 0.4493, "mean_token_accuracy": 0.8909080624580383, "num_tokens": 840914979.0, "step": 22144 }, { "epoch": 2.8170716193868466, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 5.953039646148682, "learning_rate": 1e-05, "loss": 0.4697, "mean_token_accuracy": 0.8804227113723755, "num_tokens": 840951534.0, "step": 22145 }, { "epoch": 2.817198829665437, "ewc_loss": 0.154296875, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00012493133544921875, "grad_norm": 7.7033891677856445, "learning_rate": 1e-05, "loss": 0.5262, "mean_token_accuracy": 0.8698751926422119, "num_tokens": 840987416.0, "step": 22146 }, { "epoch": 2.8173260399440276, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.22713565826416, "learning_rate": 1e-05, "loss": 0.5349, "mean_token_accuracy": 0.8593138456344604, "num_tokens": 841026734.0, "step": 22147 }, { "epoch": 2.817453250222618, "ewc_loss": 0.1572265625, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.0001277923583984375, "grad_norm": 7.779092311859131, "learning_rate": 1e-05, "loss": 0.4636, "mean_token_accuracy": 0.8901124000549316, "num_tokens": 841065333.0, "step": 22148 }, { "epoch": 2.8175804605012083, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 6.570072650909424, "learning_rate": 1e-05, "loss": 0.4582, "mean_token_accuracy": 0.8851805925369263, "num_tokens": 841102257.0, "step": 22149 }, { "epoch": 2.8177076707797992, "ewc_loss": 0.150390625, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00012111663818359375, "grad_norm": 7.540235996246338, "learning_rate": 1e-05, "loss": 0.5079, "mean_token_accuracy": 0.8776471018791199, "num_tokens": 841140253.0, "step": 22150 }, { "epoch": 2.8178348810583893, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 6.691150188446045, "learning_rate": 1e-05, "loss": 0.4902, "mean_token_accuracy": 0.8807774782180786, "num_tokens": 841181675.0, "step": 22151 }, { "epoch": 2.8179620913369803, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00011396408081054688, "grad_norm": 7.1957316398620605, "learning_rate": 1e-05, "loss": 0.4833, "mean_token_accuracy": 0.8824523091316223, "num_tokens": 841219922.0, "step": 22152 }, { "epoch": 2.8180893016155704, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.654698848724365, "learning_rate": 1e-05, "loss": 0.4734, "mean_token_accuracy": 0.8833242654800415, "num_tokens": 841256296.0, "step": 22153 }, { "epoch": 2.818216511894161, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.912930488586426, "learning_rate": 1e-05, "loss": 0.4683, "mean_token_accuracy": 0.8840720653533936, "num_tokens": 841296737.0, "step": 22154 }, { "epoch": 2.8183437221727514, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.545831203460693, "learning_rate": 1e-05, "loss": 0.5353, "mean_token_accuracy": 0.8640904426574707, "num_tokens": 841329863.0, "step": 22155 }, { "epoch": 2.818470932451342, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.751660346984863, "learning_rate": 1e-05, "loss": 0.4946, "mean_token_accuracy": 0.8718883991241455, "num_tokens": 841365057.0, "step": 22156 }, { "epoch": 2.8185981427299325, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.408485412597656, "learning_rate": 1e-05, "loss": 0.4652, "mean_token_accuracy": 0.8806020021438599, "num_tokens": 841404885.0, "step": 22157 }, { "epoch": 2.818725353008523, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.705862998962402, "learning_rate": 1e-05, "loss": 0.5018, "mean_token_accuracy": 0.8733533024787903, "num_tokens": 841445067.0, "step": 22158 }, { "epoch": 2.8188525632871135, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.25607442855835, "learning_rate": 1e-05, "loss": 0.4173, "mean_token_accuracy": 0.8934556841850281, "num_tokens": 841480569.0, "step": 22159 }, { "epoch": 2.818979773565704, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.7747673988342285, "learning_rate": 1e-05, "loss": 0.4579, "mean_token_accuracy": 0.8846650123596191, "num_tokens": 841520540.0, "step": 22160 }, { "epoch": 2.8191069838442946, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 6.109022617340088, "learning_rate": 1e-05, "loss": 0.4494, "mean_token_accuracy": 0.884848952293396, "num_tokens": 841557765.0, "step": 22161 }, { "epoch": 2.819234194122885, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.962116241455078, "learning_rate": 1e-05, "loss": 0.5013, "mean_token_accuracy": 0.8736071586608887, "num_tokens": 841593345.0, "step": 22162 }, { "epoch": 2.8193614044014756, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 5.982037544250488, "learning_rate": 1e-05, "loss": 0.5076, "mean_token_accuracy": 0.867943525314331, "num_tokens": 841629977.0, "step": 22163 }, { "epoch": 2.819488614680066, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00011348724365234375, "grad_norm": 7.234004497528076, "learning_rate": 1e-05, "loss": 0.5006, "mean_token_accuracy": 0.878476619720459, "num_tokens": 841671901.0, "step": 22164 }, { "epoch": 2.8196158249586567, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 5.963375091552734, "learning_rate": 1e-05, "loss": 0.4754, "mean_token_accuracy": 0.8779557943344116, "num_tokens": 841706072.0, "step": 22165 }, { "epoch": 2.8197430352372472, "ewc_loss": 0.1494140625, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.0001201629638671875, "grad_norm": 7.53532600402832, "learning_rate": 1e-05, "loss": 0.5226, "mean_token_accuracy": 0.8672569990158081, "num_tokens": 841745707.0, "step": 22166 }, { "epoch": 2.8198702455158378, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.15756368637085, "learning_rate": 1e-05, "loss": 0.4657, "mean_token_accuracy": 0.8786078691482544, "num_tokens": 841781468.0, "step": 22167 }, { "epoch": 2.8199974557944283, "ewc_loss": 0.15234375, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.62147855758667, "learning_rate": 1e-05, "loss": 0.4773, "mean_token_accuracy": 0.8855116367340088, "num_tokens": 841821890.0, "step": 22168 }, { "epoch": 2.820124666073019, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.498188018798828, "learning_rate": 1e-05, "loss": 0.5221, "mean_token_accuracy": 0.8667953610420227, "num_tokens": 841857668.0, "step": 22169 }, { "epoch": 2.8202518763516093, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00011730194091796875, "grad_norm": 7.342467784881592, "learning_rate": 1e-05, "loss": 0.4951, "mean_token_accuracy": 0.8762683272361755, "num_tokens": 841892595.0, "step": 22170 }, { "epoch": 2.8203790866302, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00010967254638671875, "grad_norm": 6.6856889724731445, "learning_rate": 1e-05, "loss": 0.4526, "mean_token_accuracy": 0.8908856511116028, "num_tokens": 841928077.0, "step": 22171 }, { "epoch": 2.82050629690879, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.856688022613525, "learning_rate": 1e-05, "loss": 0.4713, "mean_token_accuracy": 0.8802515864372253, "num_tokens": 841967847.0, "step": 22172 }, { "epoch": 2.820633507187381, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 6.76774263381958, "learning_rate": 1e-05, "loss": 0.5491, "mean_token_accuracy": 0.8646421432495117, "num_tokens": 842008931.0, "step": 22173 }, { "epoch": 2.820760717465971, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.36807107925415, "learning_rate": 1e-05, "loss": 0.4763, "mean_token_accuracy": 0.8743069171905518, "num_tokens": 842042870.0, "step": 22174 }, { "epoch": 2.820887927744562, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 6.94677209854126, "learning_rate": 1e-05, "loss": 0.4864, "mean_token_accuracy": 0.8792577981948853, "num_tokens": 842080193.0, "step": 22175 }, { "epoch": 2.821015138023152, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 5.977686405181885, "learning_rate": 1e-05, "loss": 0.4106, "mean_token_accuracy": 0.8947986364364624, "num_tokens": 842120953.0, "step": 22176 }, { "epoch": 2.821142348301743, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00011348724365234375, "grad_norm": 7.292147636413574, "learning_rate": 1e-05, "loss": 0.4589, "mean_token_accuracy": 0.8868451118469238, "num_tokens": 842158466.0, "step": 22177 }, { "epoch": 2.821269558580333, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.84837532043457, "learning_rate": 1e-05, "loss": 0.4713, "mean_token_accuracy": 0.8758717775344849, "num_tokens": 842194164.0, "step": 22178 }, { "epoch": 2.8213967688589237, "ewc_loss": 0.15625, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00012683868408203125, "grad_norm": 7.7357988357543945, "learning_rate": 1e-05, "loss": 0.5143, "mean_token_accuracy": 0.8756892085075378, "num_tokens": 842232137.0, "step": 22179 }, { "epoch": 2.821523979137514, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.111639499664307, "learning_rate": 1e-05, "loss": 0.5095, "mean_token_accuracy": 0.8677875995635986, "num_tokens": 842263770.0, "step": 22180 }, { "epoch": 2.8216511894161047, "ewc_loss": 0.1630859375, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.000133514404296875, "grad_norm": 7.949918270111084, "learning_rate": 1e-05, "loss": 0.5379, "mean_token_accuracy": 0.8684569597244263, "num_tokens": 842309420.0, "step": 22181 }, { "epoch": 2.8217783996946952, "ewc_loss": 0.140625, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 6.59528112411499, "learning_rate": 1e-05, "loss": 0.5397, "mean_token_accuracy": 0.8618612289428711, "num_tokens": 842350533.0, "step": 22182 }, { "epoch": 2.8219056099732858, "ewc_loss": 0.1533203125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.0001239776611328125, "grad_norm": 7.714842319488525, "learning_rate": 1e-05, "loss": 0.5392, "mean_token_accuracy": 0.8653812408447266, "num_tokens": 842389734.0, "step": 22183 }, { "epoch": 2.8220328202518763, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00011396408081054688, "grad_norm": 6.816819667816162, "learning_rate": 1e-05, "loss": 0.5238, "mean_token_accuracy": 0.8674200773239136, "num_tokens": 842424425.0, "step": 22184 }, { "epoch": 2.822160030530467, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00011682510375976562, "grad_norm": 7.314548015594482, "learning_rate": 1e-05, "loss": 0.4899, "mean_token_accuracy": 0.8765779733657837, "num_tokens": 842460269.0, "step": 22185 }, { "epoch": 2.8222872408090574, "ewc_loss": 0.140625, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 6.771306037902832, "learning_rate": 1e-05, "loss": 0.5044, "mean_token_accuracy": 0.8737843036651611, "num_tokens": 842497018.0, "step": 22186 }, { "epoch": 2.822414451087648, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 6.973718643188477, "learning_rate": 1e-05, "loss": 0.4791, "mean_token_accuracy": 0.8795222043991089, "num_tokens": 842537563.0, "step": 22187 }, { "epoch": 2.8225416613662384, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.627281188964844, "learning_rate": 1e-05, "loss": 0.4309, "mean_token_accuracy": 0.895612359046936, "num_tokens": 842573842.0, "step": 22188 }, { "epoch": 2.822668871644829, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.780313491821289, "learning_rate": 1e-05, "loss": 0.5156, "mean_token_accuracy": 0.8646993637084961, "num_tokens": 842614216.0, "step": 22189 }, { "epoch": 2.8227960819234195, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.466413497924805, "learning_rate": 1e-05, "loss": 0.5503, "mean_token_accuracy": 0.8583756685256958, "num_tokens": 842651823.0, "step": 22190 }, { "epoch": 2.82292329220201, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.738742351531982, "learning_rate": 1e-05, "loss": 0.5358, "mean_token_accuracy": 0.8585695028305054, "num_tokens": 842695113.0, "step": 22191 }, { "epoch": 2.8230505024806005, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.283458709716797, "learning_rate": 1e-05, "loss": 0.4231, "mean_token_accuracy": 0.8928571343421936, "num_tokens": 842731341.0, "step": 22192 }, { "epoch": 2.823177712759191, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.807655334472656, "learning_rate": 1e-05, "loss": 0.5085, "mean_token_accuracy": 0.8668654561042786, "num_tokens": 842769611.0, "step": 22193 }, { "epoch": 2.8233049230377816, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.11110258102417, "learning_rate": 1e-05, "loss": 0.5035, "mean_token_accuracy": 0.8675974607467651, "num_tokens": 842807106.0, "step": 22194 }, { "epoch": 2.823432133316372, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.997744560241699, "learning_rate": 1e-05, "loss": 0.4853, "mean_token_accuracy": 0.8807541131973267, "num_tokens": 842847957.0, "step": 22195 }, { "epoch": 2.8235593435949626, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 5.972255229949951, "learning_rate": 1e-05, "loss": 0.5121, "mean_token_accuracy": 0.8647207021713257, "num_tokens": 842889222.0, "step": 22196 }, { "epoch": 2.8236865538735527, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00011491775512695312, "grad_norm": 7.318663597106934, "learning_rate": 1e-05, "loss": 0.4706, "mean_token_accuracy": 0.8842440843582153, "num_tokens": 842925164.0, "step": 22197 }, { "epoch": 2.8238137641521437, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 5.962287425994873, "learning_rate": 1e-05, "loss": 0.5096, "mean_token_accuracy": 0.862185001373291, "num_tokens": 842963832.0, "step": 22198 }, { "epoch": 2.8239409744307338, "ewc_loss": 0.15234375, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.652472972869873, "learning_rate": 1e-05, "loss": 0.5122, "mean_token_accuracy": 0.8707840442657471, "num_tokens": 843004357.0, "step": 22199 }, { "epoch": 2.8240681847093247, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.230461597442627, "learning_rate": 1e-05, "loss": 0.5415, "mean_token_accuracy": 0.8562275171279907, "num_tokens": 843045241.0, "step": 22200 }, { "epoch": 2.824195394987915, "ewc_loss": 0.154296875, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00012493133544921875, "grad_norm": 7.709972858428955, "learning_rate": 1e-05, "loss": 0.5252, "mean_token_accuracy": 0.8699705600738525, "num_tokens": 843084473.0, "step": 22201 }, { "epoch": 2.824322605266506, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.611496925354004, "learning_rate": 1e-05, "loss": 0.5002, "mean_token_accuracy": 0.8747094869613647, "num_tokens": 843122118.0, "step": 22202 }, { "epoch": 2.824449815545096, "ewc_loss": 0.1455078125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00011587142944335938, "grad_norm": 7.358672618865967, "learning_rate": 1e-05, "loss": 0.5414, "mean_token_accuracy": 0.8641649484634399, "num_tokens": 843159936.0, "step": 22203 }, { "epoch": 2.8245770258236864, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00011014938354492188, "grad_norm": 6.780749797821045, "learning_rate": 1e-05, "loss": 0.5102, "mean_token_accuracy": 0.8729608058929443, "num_tokens": 843196212.0, "step": 22204 }, { "epoch": 2.824704236102277, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.8328962326049805, "learning_rate": 1e-05, "loss": 0.4953, "mean_token_accuracy": 0.879684329032898, "num_tokens": 843236870.0, "step": 22205 }, { "epoch": 2.8248314463808675, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 6.862331390380859, "learning_rate": 1e-05, "loss": 0.4761, "mean_token_accuracy": 0.8814911842346191, "num_tokens": 843270701.0, "step": 22206 }, { "epoch": 2.824958656659458, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.3219733238220215, "learning_rate": 1e-05, "loss": 0.4547, "mean_token_accuracy": 0.8807464241981506, "num_tokens": 843308547.0, "step": 22207 }, { "epoch": 2.8250858669380485, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00010967254638671875, "grad_norm": 7.032749176025391, "learning_rate": 1e-05, "loss": 0.4555, "mean_token_accuracy": 0.8883732557296753, "num_tokens": 843354496.0, "step": 22208 }, { "epoch": 2.825213077216639, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.985781192779541, "learning_rate": 1e-05, "loss": 0.4881, "mean_token_accuracy": 0.8676633834838867, "num_tokens": 843389995.0, "step": 22209 }, { "epoch": 2.8253402874952296, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00011491775512695312, "grad_norm": 7.363205909729004, "learning_rate": 1e-05, "loss": 0.5226, "mean_token_accuracy": 0.8731149435043335, "num_tokens": 843428444.0, "step": 22210 }, { "epoch": 2.82546749777382, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 5.902367115020752, "learning_rate": 1e-05, "loss": 0.4608, "mean_token_accuracy": 0.8800716400146484, "num_tokens": 843466390.0, "step": 22211 }, { "epoch": 2.8255947080524106, "ewc_loss": 0.1552734375, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.000125885009765625, "grad_norm": 7.779854774475098, "learning_rate": 1e-05, "loss": 0.5833, "mean_token_accuracy": 0.8597870469093323, "num_tokens": 843503424.0, "step": 22212 }, { "epoch": 2.825721918331001, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.196965217590332, "learning_rate": 1e-05, "loss": 0.5089, "mean_token_accuracy": 0.8661343455314636, "num_tokens": 843544669.0, "step": 22213 }, { "epoch": 2.8258491286095917, "ewc_loss": 0.1591796875, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00012969970703125, "grad_norm": 7.935139179229736, "learning_rate": 1e-05, "loss": 0.4628, "mean_token_accuracy": 0.8920215368270874, "num_tokens": 843579777.0, "step": 22214 }, { "epoch": 2.8259763388881822, "ewc_loss": 0.140625, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 6.655592918395996, "learning_rate": 1e-05, "loss": 0.4842, "mean_token_accuracy": 0.8770425319671631, "num_tokens": 843613481.0, "step": 22215 }, { "epoch": 2.8261035491667728, "ewc_loss": 0.1533203125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.0001239776611328125, "grad_norm": 7.637559413909912, "learning_rate": 1e-05, "loss": 0.4595, "mean_token_accuracy": 0.8915014266967773, "num_tokens": 843656870.0, "step": 22216 }, { "epoch": 2.8262307594453633, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.0001125335693359375, "grad_norm": 6.8197021484375, "learning_rate": 1e-05, "loss": 0.4776, "mean_token_accuracy": 0.8778928518295288, "num_tokens": 843693859.0, "step": 22217 }, { "epoch": 2.826357969723954, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00011491775512695312, "grad_norm": 7.24526309967041, "learning_rate": 1e-05, "loss": 0.4523, "mean_token_accuracy": 0.8917359709739685, "num_tokens": 843726999.0, "step": 22218 }, { "epoch": 2.8264851800025443, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.739959239959717, "learning_rate": 1e-05, "loss": 0.5104, "mean_token_accuracy": 0.8711849451065063, "num_tokens": 843767244.0, "step": 22219 }, { "epoch": 2.826612390281135, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.935623645782471, "learning_rate": 1e-05, "loss": 0.4878, "mean_token_accuracy": 0.8740502595901489, "num_tokens": 843799828.0, "step": 22220 }, { "epoch": 2.8267396005597254, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.60405969619751, "learning_rate": 1e-05, "loss": 0.4748, "mean_token_accuracy": 0.8768738508224487, "num_tokens": 843837485.0, "step": 22221 }, { "epoch": 2.8268668108383155, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.753366947174072, "learning_rate": 1e-05, "loss": 0.46, "mean_token_accuracy": 0.8841148614883423, "num_tokens": 843878796.0, "step": 22222 }, { "epoch": 2.8269940211169065, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.448146820068359, "learning_rate": 1e-05, "loss": 0.5015, "mean_token_accuracy": 0.869821310043335, "num_tokens": 843920200.0, "step": 22223 }, { "epoch": 2.8271212313954965, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.710849285125732, "learning_rate": 1e-05, "loss": 0.4697, "mean_token_accuracy": 0.8850661516189575, "num_tokens": 843961719.0, "step": 22224 }, { "epoch": 2.8272484416740875, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.265824317932129, "learning_rate": 1e-05, "loss": 0.5055, "mean_token_accuracy": 0.8688451051712036, "num_tokens": 844005085.0, "step": 22225 }, { "epoch": 2.8273756519526776, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.803905487060547, "learning_rate": 1e-05, "loss": 0.507, "mean_token_accuracy": 0.8679348230361938, "num_tokens": 844036964.0, "step": 22226 }, { "epoch": 2.827502862231268, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.09360933303833, "learning_rate": 1e-05, "loss": 0.5096, "mean_token_accuracy": 0.8705859184265137, "num_tokens": 844078297.0, "step": 22227 }, { "epoch": 2.8276300725098586, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 7.009968280792236, "learning_rate": 1e-05, "loss": 0.4488, "mean_token_accuracy": 0.8895900845527649, "num_tokens": 844113749.0, "step": 22228 }, { "epoch": 2.827757282788449, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 5.957710266113281, "learning_rate": 1e-05, "loss": 0.4972, "mean_token_accuracy": 0.8683453798294067, "num_tokens": 844155993.0, "step": 22229 }, { "epoch": 2.8278844930670397, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00011491775512695312, "grad_norm": 7.332233428955078, "learning_rate": 1e-05, "loss": 0.4957, "mean_token_accuracy": 0.8758326768875122, "num_tokens": 844204302.0, "step": 22230 }, { "epoch": 2.8280117033456302, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 5.974068641662598, "learning_rate": 1e-05, "loss": 0.4602, "mean_token_accuracy": 0.8809918761253357, "num_tokens": 844241490.0, "step": 22231 }, { "epoch": 2.8281389136242208, "ewc_loss": 0.15234375, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.671695232391357, "learning_rate": 1e-05, "loss": 0.4928, "mean_token_accuracy": 0.8791884183883667, "num_tokens": 844282891.0, "step": 22232 }, { "epoch": 2.8282661239028113, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.230097770690918, "learning_rate": 1e-05, "loss": 0.4878, "mean_token_accuracy": 0.8735613822937012, "num_tokens": 844324407.0, "step": 22233 }, { "epoch": 2.828393334181402, "ewc_loss": 0.1552734375, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.000125885009765625, "grad_norm": 7.75473165512085, "learning_rate": 1e-05, "loss": 0.5036, "mean_token_accuracy": 0.8789536952972412, "num_tokens": 844356656.0, "step": 22234 }, { "epoch": 2.8285205444599923, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 6.635626792907715, "learning_rate": 1e-05, "loss": 0.5108, "mean_token_accuracy": 0.8701488375663757, "num_tokens": 844389789.0, "step": 22235 }, { "epoch": 2.828647754738583, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00011682510375976562, "grad_norm": 7.393331050872803, "learning_rate": 1e-05, "loss": 0.4933, "mean_token_accuracy": 0.8761428594589233, "num_tokens": 844423599.0, "step": 22236 }, { "epoch": 2.8287749650171734, "ewc_loss": 0.140625, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00011110305786132812, "grad_norm": 6.799888610839844, "learning_rate": 1e-05, "loss": 0.4861, "mean_token_accuracy": 0.879667341709137, "num_tokens": 844460117.0, "step": 22237 }, { "epoch": 2.828902175295764, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.85378885269165, "learning_rate": 1e-05, "loss": 0.5099, "mean_token_accuracy": 0.866263210773468, "num_tokens": 844501263.0, "step": 22238 }, { "epoch": 2.8290293855743545, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 6.859898567199707, "learning_rate": 1e-05, "loss": 0.4569, "mean_token_accuracy": 0.87957763671875, "num_tokens": 844541573.0, "step": 22239 }, { "epoch": 2.829156595852945, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.3476457595825195, "learning_rate": 1e-05, "loss": 0.474, "mean_token_accuracy": 0.876039445400238, "num_tokens": 844583728.0, "step": 22240 }, { "epoch": 2.8292838061315355, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 7.031525611877441, "learning_rate": 1e-05, "loss": 0.4821, "mean_token_accuracy": 0.8798620104789734, "num_tokens": 844616611.0, "step": 22241 }, { "epoch": 2.829411016410126, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 6.016070365905762, "learning_rate": 1e-05, "loss": 0.5244, "mean_token_accuracy": 0.8602334856987, "num_tokens": 844652067.0, "step": 22242 }, { "epoch": 2.8295382266887166, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00011444091796875, "grad_norm": 7.332387924194336, "learning_rate": 1e-05, "loss": 0.5137, "mean_token_accuracy": 0.8696258068084717, "num_tokens": 844688582.0, "step": 22243 }, { "epoch": 2.829665436967307, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 5.918953895568848, "learning_rate": 1e-05, "loss": 0.4706, "mean_token_accuracy": 0.8776083588600159, "num_tokens": 844727120.0, "step": 22244 }, { "epoch": 2.8297926472458976, "ewc_loss": 0.154296875, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00012493133544921875, "grad_norm": 7.7043914794921875, "learning_rate": 1e-05, "loss": 0.5013, "mean_token_accuracy": 0.8775873184204102, "num_tokens": 844772328.0, "step": 22245 }, { "epoch": 2.829919857524488, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.167941570281982, "learning_rate": 1e-05, "loss": 0.4268, "mean_token_accuracy": 0.8895337581634521, "num_tokens": 844812528.0, "step": 22246 }, { "epoch": 2.8300470678030782, "ewc_loss": 0.158203125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00012874603271484375, "grad_norm": 7.870614528656006, "learning_rate": 1e-05, "loss": 0.5327, "mean_token_accuracy": 0.8707581758499146, "num_tokens": 844857208.0, "step": 22247 }, { "epoch": 2.830174278081669, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 6.59304666519165, "learning_rate": 1e-05, "loss": 0.5115, "mean_token_accuracy": 0.868546724319458, "num_tokens": 844890362.0, "step": 22248 }, { "epoch": 2.8303014883602593, "ewc_loss": 0.15234375, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.625793933868408, "learning_rate": 1e-05, "loss": 0.5449, "mean_token_accuracy": 0.8634026050567627, "num_tokens": 844928664.0, "step": 22249 }, { "epoch": 2.8304286986388503, "ewc_loss": 0.140625, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 6.777206897735596, "learning_rate": 1e-05, "loss": 0.4404, "mean_token_accuracy": 0.8958940505981445, "num_tokens": 844957043.0, "step": 22250 }, { "epoch": 2.8305559089174404, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00011539459228515625, "grad_norm": 7.245047569274902, "learning_rate": 1e-05, "loss": 0.5081, "mean_token_accuracy": 0.8729349970817566, "num_tokens": 844991808.0, "step": 22251 }, { "epoch": 2.830683119196031, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 6.729772567749023, "learning_rate": 1e-05, "loss": 0.4914, "mean_token_accuracy": 0.8774863481521606, "num_tokens": 845025571.0, "step": 22252 }, { "epoch": 2.8308103294746214, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.912049770355225, "learning_rate": 1e-05, "loss": 0.5229, "mean_token_accuracy": 0.8676679134368896, "num_tokens": 845063154.0, "step": 22253 }, { "epoch": 2.830937539753212, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.606471538543701, "learning_rate": 1e-05, "loss": 0.4686, "mean_token_accuracy": 0.8847140073776245, "num_tokens": 845100256.0, "step": 22254 }, { "epoch": 2.8310647500318025, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.732898712158203, "learning_rate": 1e-05, "loss": 0.4872, "mean_token_accuracy": 0.877212405204773, "num_tokens": 845138111.0, "step": 22255 }, { "epoch": 2.831191960310393, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.4489240646362305, "learning_rate": 1e-05, "loss": 0.4587, "mean_token_accuracy": 0.881335973739624, "num_tokens": 845175759.0, "step": 22256 }, { "epoch": 2.8313191705889835, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.7047529220581055, "learning_rate": 1e-05, "loss": 0.4779, "mean_token_accuracy": 0.8753693103790283, "num_tokens": 845212232.0, "step": 22257 }, { "epoch": 2.831446380867574, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.273073673248291, "learning_rate": 1e-05, "loss": 0.4201, "mean_token_accuracy": 0.8933245539665222, "num_tokens": 845247141.0, "step": 22258 }, { "epoch": 2.8315735911461646, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.796099662780762, "learning_rate": 1e-05, "loss": 0.47, "mean_token_accuracy": 0.8819255828857422, "num_tokens": 845287908.0, "step": 22259 }, { "epoch": 2.831700801424755, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 6.081850528717041, "learning_rate": 1e-05, "loss": 0.4558, "mean_token_accuracy": 0.8801038265228271, "num_tokens": 845328134.0, "step": 22260 }, { "epoch": 2.8318280117033456, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 7.028669834136963, "learning_rate": 1e-05, "loss": 0.4801, "mean_token_accuracy": 0.8812494874000549, "num_tokens": 845366096.0, "step": 22261 }, { "epoch": 2.831955221981936, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 5.95103645324707, "learning_rate": 1e-05, "loss": 0.4585, "mean_token_accuracy": 0.877899706363678, "num_tokens": 845397737.0, "step": 22262 }, { "epoch": 2.8320824322605267, "ewc_loss": 0.1455078125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.0001163482666015625, "grad_norm": 7.372363567352295, "learning_rate": 1e-05, "loss": 0.488, "mean_token_accuracy": 0.8804118037223816, "num_tokens": 845437015.0, "step": 22263 }, { "epoch": 2.8322096425391172, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 5.957716464996338, "learning_rate": 1e-05, "loss": 0.4387, "mean_token_accuracy": 0.8878803849220276, "num_tokens": 845473778.0, "step": 22264 }, { "epoch": 2.8323368528177078, "ewc_loss": 0.1552734375, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.000125885009765625, "grad_norm": 7.724002361297607, "learning_rate": 1e-05, "loss": 0.4654, "mean_token_accuracy": 0.889541506767273, "num_tokens": 845509254.0, "step": 22265 }, { "epoch": 2.8324640630962983, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.280209064483643, "learning_rate": 1e-05, "loss": 0.5084, "mean_token_accuracy": 0.8690100908279419, "num_tokens": 845547682.0, "step": 22266 }, { "epoch": 2.832591273374889, "ewc_loss": 0.154296875, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00012493133544921875, "grad_norm": 7.737151145935059, "learning_rate": 1e-05, "loss": 0.4641, "mean_token_accuracy": 0.8878152370452881, "num_tokens": 845585315.0, "step": 22267 }, { "epoch": 2.8327184836534793, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 6.665360450744629, "learning_rate": 1e-05, "loss": 0.4809, "mean_token_accuracy": 0.8833768963813782, "num_tokens": 845625065.0, "step": 22268 }, { "epoch": 2.83284569393207, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00011730194091796875, "grad_norm": 7.331779479980469, "learning_rate": 1e-05, "loss": 0.4665, "mean_token_accuracy": 0.8877543807029724, "num_tokens": 845664148.0, "step": 22269 }, { "epoch": 2.83297290421066, "ewc_loss": 0.140625, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00011110305786132812, "grad_norm": 6.826513290405273, "learning_rate": 1e-05, "loss": 0.4622, "mean_token_accuracy": 0.8812311291694641, "num_tokens": 845695986.0, "step": 22270 }, { "epoch": 2.833100114489251, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.794026851654053, "learning_rate": 1e-05, "loss": 0.491, "mean_token_accuracy": 0.8773595094680786, "num_tokens": 845732807.0, "step": 22271 }, { "epoch": 2.833227324767841, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 6.886767864227295, "learning_rate": 1e-05, "loss": 0.4675, "mean_token_accuracy": 0.8831861019134521, "num_tokens": 845763238.0, "step": 22272 }, { "epoch": 2.833354535046432, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.291440010070801, "learning_rate": 1e-05, "loss": 0.4341, "mean_token_accuracy": 0.8905152082443237, "num_tokens": 845802218.0, "step": 22273 }, { "epoch": 2.833481745325022, "ewc_loss": 0.140625, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 7.068849563598633, "learning_rate": 1e-05, "loss": 0.5564, "mean_token_accuracy": 0.85593581199646, "num_tokens": 845845056.0, "step": 22274 }, { "epoch": 2.833608955603613, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.983373641967773, "learning_rate": 1e-05, "loss": 0.4761, "mean_token_accuracy": 0.8746815919876099, "num_tokens": 845884909.0, "step": 22275 }, { "epoch": 2.833736165882203, "ewc_loss": 0.1455078125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00011587142944335938, "grad_norm": 7.4112091064453125, "learning_rate": 1e-05, "loss": 0.5035, "mean_token_accuracy": 0.8754026889801025, "num_tokens": 845923894.0, "step": 22276 }, { "epoch": 2.8338633761607936, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 5.947520732879639, "learning_rate": 1e-05, "loss": 0.5318, "mean_token_accuracy": 0.857882559299469, "num_tokens": 845962452.0, "step": 22277 }, { "epoch": 2.833990586439384, "ewc_loss": 0.1552734375, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.000125885009765625, "grad_norm": 7.793332099914551, "learning_rate": 1e-05, "loss": 0.503, "mean_token_accuracy": 0.8764432668685913, "num_tokens": 845998031.0, "step": 22278 }, { "epoch": 2.8341177967179747, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.266338348388672, "learning_rate": 1e-05, "loss": 0.488, "mean_token_accuracy": 0.8734695315361023, "num_tokens": 846036920.0, "step": 22279 }, { "epoch": 2.8342450069965652, "ewc_loss": 0.158203125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00012874603271484375, "grad_norm": 7.8624677658081055, "learning_rate": 1e-05, "loss": 0.5505, "mean_token_accuracy": 0.8660550713539124, "num_tokens": 846075978.0, "step": 22280 }, { "epoch": 2.8343722172751558, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00010967254638671875, "grad_norm": 6.667185306549072, "learning_rate": 1e-05, "loss": 0.4919, "mean_token_accuracy": 0.8759416341781616, "num_tokens": 846111893.0, "step": 22281 }, { "epoch": 2.8344994275537463, "ewc_loss": 0.1513671875, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.0001220703125, "grad_norm": 7.5592780113220215, "learning_rate": 1e-05, "loss": 0.5168, "mean_token_accuracy": 0.8731009364128113, "num_tokens": 846153417.0, "step": 22282 }, { "epoch": 2.834626637832337, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00011157989501953125, "grad_norm": 6.774240493774414, "learning_rate": 1e-05, "loss": 0.5158, "mean_token_accuracy": 0.8713851571083069, "num_tokens": 846194375.0, "step": 22283 }, { "epoch": 2.8347538481109273, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00011301040649414062, "grad_norm": 7.187510967254639, "learning_rate": 1e-05, "loss": 0.4703, "mean_token_accuracy": 0.8820053935050964, "num_tokens": 846231040.0, "step": 22284 }, { "epoch": 2.834881058389518, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.684845447540283, "learning_rate": 1e-05, "loss": 0.4565, "mean_token_accuracy": 0.8874506950378418, "num_tokens": 846270645.0, "step": 22285 }, { "epoch": 2.8350082686681084, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.925459384918213, "learning_rate": 1e-05, "loss": 0.4636, "mean_token_accuracy": 0.8835909366607666, "num_tokens": 846304038.0, "step": 22286 }, { "epoch": 2.835135478946699, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.546792984008789, "learning_rate": 1e-05, "loss": 0.5185, "mean_token_accuracy": 0.8666809797286987, "num_tokens": 846338136.0, "step": 22287 }, { "epoch": 2.8352626892252895, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.788804054260254, "learning_rate": 1e-05, "loss": 0.484, "mean_token_accuracy": 0.8754582405090332, "num_tokens": 846372654.0, "step": 22288 }, { "epoch": 2.83538989950388, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.378589153289795, "learning_rate": 1e-05, "loss": 0.4854, "mean_token_accuracy": 0.8740060329437256, "num_tokens": 846409180.0, "step": 22289 }, { "epoch": 2.8355171097824705, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.79450798034668, "learning_rate": 1e-05, "loss": 0.5253, "mean_token_accuracy": 0.8649756908416748, "num_tokens": 846453410.0, "step": 22290 }, { "epoch": 2.835644320061061, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.197620868682861, "learning_rate": 1e-05, "loss": 0.451, "mean_token_accuracy": 0.8853416442871094, "num_tokens": 846488271.0, "step": 22291 }, { "epoch": 2.8357715303396516, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.928197860717773, "learning_rate": 1e-05, "loss": 0.4563, "mean_token_accuracy": 0.8831275701522827, "num_tokens": 846528730.0, "step": 22292 }, { "epoch": 2.835898740618242, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 6.039392948150635, "learning_rate": 1e-05, "loss": 0.4213, "mean_token_accuracy": 0.8898928761482239, "num_tokens": 846567276.0, "step": 22293 }, { "epoch": 2.8360259508968326, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.0001125335693359375, "grad_norm": 7.189310073852539, "learning_rate": 1e-05, "loss": 0.4749, "mean_token_accuracy": 0.8845688104629517, "num_tokens": 846608482.0, "step": 22294 }, { "epoch": 2.8361531611754227, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 5.983030319213867, "learning_rate": 1e-05, "loss": 0.4918, "mean_token_accuracy": 0.8722347617149353, "num_tokens": 846648582.0, "step": 22295 }, { "epoch": 2.8362803714540137, "ewc_loss": 0.1494140625, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00011968612670898438, "grad_norm": 7.500320911407471, "learning_rate": 1e-05, "loss": 0.4796, "mean_token_accuracy": 0.884813666343689, "num_tokens": 846687392.0, "step": 22296 }, { "epoch": 2.8364075817326038, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.135525703430176, "learning_rate": 1e-05, "loss": 0.4606, "mean_token_accuracy": 0.882768988609314, "num_tokens": 846721542.0, "step": 22297 }, { "epoch": 2.8365347920111947, "ewc_loss": 0.15234375, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.65677547454834, "learning_rate": 1e-05, "loss": 0.5373, "mean_token_accuracy": 0.8678462505340576, "num_tokens": 846759262.0, "step": 22298 }, { "epoch": 2.836662002289785, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.465022563934326, "learning_rate": 1e-05, "loss": 0.5058, "mean_token_accuracy": 0.8719733953475952, "num_tokens": 846796979.0, "step": 22299 }, { "epoch": 2.836789212568376, "ewc_loss": 0.1474609375, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.000118255615234375, "grad_norm": 7.447885990142822, "learning_rate": 1e-05, "loss": 0.5099, "mean_token_accuracy": 0.877066969871521, "num_tokens": 846832417.0, "step": 22300 }, { "epoch": 2.836916422846966, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.698066711425781, "learning_rate": 1e-05, "loss": 0.5142, "mean_token_accuracy": 0.8718799352645874, "num_tokens": 846870289.0, "step": 22301 }, { "epoch": 2.8370436331255564, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 6.965830326080322, "learning_rate": 1e-05, "loss": 0.4914, "mean_token_accuracy": 0.8783752918243408, "num_tokens": 846909821.0, "step": 22302 }, { "epoch": 2.837170843404147, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 6.791715621948242, "learning_rate": 1e-05, "loss": 0.4902, "mean_token_accuracy": 0.8787946701049805, "num_tokens": 846945512.0, "step": 22303 }, { "epoch": 2.8372980536827375, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.4413886070251465, "learning_rate": 1e-05, "loss": 0.4629, "mean_token_accuracy": 0.8833397030830383, "num_tokens": 846980834.0, "step": 22304 }, { "epoch": 2.837425263961328, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 6.952510356903076, "learning_rate": 1e-05, "loss": 0.4827, "mean_token_accuracy": 0.878902792930603, "num_tokens": 847025320.0, "step": 22305 }, { "epoch": 2.8375524742399185, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 6.018564224243164, "learning_rate": 1e-05, "loss": 0.433, "mean_token_accuracy": 0.8876131772994995, "num_tokens": 847063637.0, "step": 22306 }, { "epoch": 2.837679684518509, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00011348724365234375, "grad_norm": 7.286067485809326, "learning_rate": 1e-05, "loss": 0.4647, "mean_token_accuracy": 0.8844884634017944, "num_tokens": 847102314.0, "step": 22307 }, { "epoch": 2.8378068947970996, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.848657131195068, "learning_rate": 1e-05, "loss": 0.4542, "mean_token_accuracy": 0.8802202343940735, "num_tokens": 847141956.0, "step": 22308 }, { "epoch": 2.83793410507569, "ewc_loss": 0.15625, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00012683868408203125, "grad_norm": 7.755885601043701, "learning_rate": 1e-05, "loss": 0.527, "mean_token_accuracy": 0.869792103767395, "num_tokens": 847178855.0, "step": 22309 }, { "epoch": 2.8380613153542806, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.090463638305664, "learning_rate": 1e-05, "loss": 0.4845, "mean_token_accuracy": 0.878104567527771, "num_tokens": 847214228.0, "step": 22310 }, { "epoch": 2.838188525632871, "ewc_loss": 0.162109375, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00013256072998046875, "grad_norm": 8.019740104675293, "learning_rate": 1e-05, "loss": 0.52, "mean_token_accuracy": 0.8749859929084778, "num_tokens": 847255942.0, "step": 22311 }, { "epoch": 2.8383157359114617, "ewc_loss": 0.140625, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00011110305786132812, "grad_norm": 6.640124320983887, "learning_rate": 1e-05, "loss": 0.5371, "mean_token_accuracy": 0.8632627725601196, "num_tokens": 847294742.0, "step": 22312 }, { "epoch": 2.838442946190052, "ewc_loss": 0.1552734375, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.000125885009765625, "grad_norm": 7.754886627197266, "learning_rate": 1e-05, "loss": 0.5186, "mean_token_accuracy": 0.8750806450843811, "num_tokens": 847330105.0, "step": 22313 }, { "epoch": 2.8385701564686427, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00011491775512695312, "grad_norm": 6.885271072387695, "learning_rate": 1e-05, "loss": 0.5065, "mean_token_accuracy": 0.8710986971855164, "num_tokens": 847373291.0, "step": 22314 }, { "epoch": 2.8386973667472333, "ewc_loss": 0.1455078125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.0001163482666015625, "grad_norm": 7.311954498291016, "learning_rate": 1e-05, "loss": 0.459, "mean_token_accuracy": 0.8890936374664307, "num_tokens": 847409059.0, "step": 22315 }, { "epoch": 2.838824577025824, "ewc_loss": 0.140625, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 6.8176727294921875, "learning_rate": 1e-05, "loss": 0.488, "mean_token_accuracy": 0.8808070421218872, "num_tokens": 847445869.0, "step": 22316 }, { "epoch": 2.8389517873044143, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.946157932281494, "learning_rate": 1e-05, "loss": 0.4779, "mean_token_accuracy": 0.8810936212539673, "num_tokens": 847486816.0, "step": 22317 }, { "epoch": 2.839078997583005, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.673281192779541, "learning_rate": 1e-05, "loss": 0.4686, "mean_token_accuracy": 0.8846824765205383, "num_tokens": 847527482.0, "step": 22318 }, { "epoch": 2.8392062078615954, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.734806060791016, "learning_rate": 1e-05, "loss": 0.5263, "mean_token_accuracy": 0.8647209405899048, "num_tokens": 847570850.0, "step": 22319 }, { "epoch": 2.8393334181401855, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.5194292068481445, "learning_rate": 1e-05, "loss": 0.4989, "mean_token_accuracy": 0.8725177645683289, "num_tokens": 847608872.0, "step": 22320 }, { "epoch": 2.8394606284187764, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.668398857116699, "learning_rate": 1e-05, "loss": 0.5016, "mean_token_accuracy": 0.8711414337158203, "num_tokens": 847647813.0, "step": 22321 }, { "epoch": 2.8395878386973665, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.335723400115967, "learning_rate": 1e-05, "loss": 0.4634, "mean_token_accuracy": 0.8792604207992554, "num_tokens": 847685726.0, "step": 22322 }, { "epoch": 2.8397150489759575, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.733015060424805, "learning_rate": 1e-05, "loss": 0.4989, "mean_token_accuracy": 0.8701679706573486, "num_tokens": 847723029.0, "step": 22323 }, { "epoch": 2.8398422592545476, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.140498161315918, "learning_rate": 1e-05, "loss": 0.4524, "mean_token_accuracy": 0.8835110664367676, "num_tokens": 847761922.0, "step": 22324 }, { "epoch": 2.839969469533138, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.906854152679443, "learning_rate": 1e-05, "loss": 0.5001, "mean_token_accuracy": 0.873313307762146, "num_tokens": 847806229.0, "step": 22325 }, { "epoch": 2.8400966798117286, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.966070652008057, "learning_rate": 1e-05, "loss": 0.4775, "mean_token_accuracy": 0.8762189745903015, "num_tokens": 847844448.0, "step": 22326 }, { "epoch": 2.840223890090319, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00011348724365234375, "grad_norm": 7.231875896453857, "learning_rate": 1e-05, "loss": 0.5223, "mean_token_accuracy": 0.8674747943878174, "num_tokens": 847889387.0, "step": 22327 }, { "epoch": 2.8403511003689097, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.885382175445557, "learning_rate": 1e-05, "loss": 0.4707, "mean_token_accuracy": 0.8740266561508179, "num_tokens": 847926993.0, "step": 22328 }, { "epoch": 2.8404783106475002, "ewc_loss": 0.15234375, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.649282932281494, "learning_rate": 1e-05, "loss": 0.4744, "mean_token_accuracy": 0.887550950050354, "num_tokens": 847958884.0, "step": 22329 }, { "epoch": 2.8406055209260908, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.095577239990234, "learning_rate": 1e-05, "loss": 0.4957, "mean_token_accuracy": 0.874187707901001, "num_tokens": 847997894.0, "step": 22330 }, { "epoch": 2.8407327312046813, "ewc_loss": 0.16015625, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00013065338134765625, "grad_norm": 7.899016857147217, "learning_rate": 1e-05, "loss": 0.5187, "mean_token_accuracy": 0.8766285181045532, "num_tokens": 848037392.0, "step": 22331 }, { "epoch": 2.840859941483272, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00011014938354492188, "grad_norm": 6.567215442657471, "learning_rate": 1e-05, "loss": 0.4889, "mean_token_accuracy": 0.8774797320365906, "num_tokens": 848072844.0, "step": 22332 }, { "epoch": 2.8409871517618623, "ewc_loss": 0.1533203125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.0001239776611328125, "grad_norm": 7.656774997711182, "learning_rate": 1e-05, "loss": 0.5425, "mean_token_accuracy": 0.8650230765342712, "num_tokens": 848107477.0, "step": 22333 }, { "epoch": 2.841114362040453, "ewc_loss": 0.1455078125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00011587142944335938, "grad_norm": 6.853048324584961, "learning_rate": 1e-05, "loss": 0.5001, "mean_token_accuracy": 0.8771857023239136, "num_tokens": 848143124.0, "step": 22334 }, { "epoch": 2.8412415723190434, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.0001125335693359375, "grad_norm": 7.129284381866455, "learning_rate": 1e-05, "loss": 0.4899, "mean_token_accuracy": 0.8797447681427002, "num_tokens": 848181454.0, "step": 22335 }, { "epoch": 2.841368782597634, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00011205673217773438, "grad_norm": 6.877139568328857, "learning_rate": 1e-05, "loss": 0.4688, "mean_token_accuracy": 0.8842531442642212, "num_tokens": 848219300.0, "step": 22336 }, { "epoch": 2.8414959928762245, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.610479354858398, "learning_rate": 1e-05, "loss": 0.4735, "mean_token_accuracy": 0.8805687427520752, "num_tokens": 848256650.0, "step": 22337 }, { "epoch": 2.841623203154815, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 6.9341206550598145, "learning_rate": 1e-05, "loss": 0.5071, "mean_token_accuracy": 0.8682076930999756, "num_tokens": 848294836.0, "step": 22338 }, { "epoch": 2.8417504134334055, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.2122297286987305, "learning_rate": 1e-05, "loss": 0.4697, "mean_token_accuracy": 0.8786509037017822, "num_tokens": 848326306.0, "step": 22339 }, { "epoch": 2.841877623711996, "ewc_loss": 0.140625, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 7.108040809631348, "learning_rate": 1e-05, "loss": 0.467, "mean_token_accuracy": 0.8863337635993958, "num_tokens": 848360886.0, "step": 22340 }, { "epoch": 2.8420048339905866, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.003441333770752, "learning_rate": 1e-05, "loss": 0.4756, "mean_token_accuracy": 0.8758426308631897, "num_tokens": 848399364.0, "step": 22341 }, { "epoch": 2.842132044269177, "ewc_loss": 0.1455078125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00011587142944335938, "grad_norm": 7.395545482635498, "learning_rate": 1e-05, "loss": 0.4797, "mean_token_accuracy": 0.8833041787147522, "num_tokens": 848436978.0, "step": 22342 }, { "epoch": 2.8422592545477676, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.001001358032227, "learning_rate": 1e-05, "loss": 0.516, "mean_token_accuracy": 0.8649510741233826, "num_tokens": 848474661.0, "step": 22343 }, { "epoch": 2.842386464826358, "ewc_loss": 0.1533203125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.0001239776611328125, "grad_norm": 7.681443691253662, "learning_rate": 1e-05, "loss": 0.4719, "mean_token_accuracy": 0.8876453042030334, "num_tokens": 848514469.0, "step": 22344 }, { "epoch": 2.8425136751049482, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.25096321105957, "learning_rate": 1e-05, "loss": 0.4623, "mean_token_accuracy": 0.8819600343704224, "num_tokens": 848548606.0, "step": 22345 }, { "epoch": 2.842640885383539, "ewc_loss": 0.1552734375, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.000125885009765625, "grad_norm": 7.7424211502075195, "learning_rate": 1e-05, "loss": 0.5264, "mean_token_accuracy": 0.8697065711021423, "num_tokens": 848585309.0, "step": 22346 }, { "epoch": 2.8427680956621293, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.554003715515137, "learning_rate": 1e-05, "loss": 0.4874, "mean_token_accuracy": 0.8781309127807617, "num_tokens": 848625062.0, "step": 22347 }, { "epoch": 2.8428953059407203, "ewc_loss": 0.1494140625, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.0001201629638671875, "grad_norm": 7.517111301422119, "learning_rate": 1e-05, "loss": 0.5084, "mean_token_accuracy": 0.8747889399528503, "num_tokens": 848661914.0, "step": 22348 }, { "epoch": 2.8430225162193103, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 6.685604572296143, "learning_rate": 1e-05, "loss": 0.5866, "mean_token_accuracy": 0.8480648994445801, "num_tokens": 848704343.0, "step": 22349 }, { "epoch": 2.843149726497901, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00011348724365234375, "grad_norm": 7.1881513595581055, "learning_rate": 1e-05, "loss": 0.4525, "mean_token_accuracy": 0.8921927213668823, "num_tokens": 848737415.0, "step": 22350 }, { "epoch": 2.8432769367764914, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.643916606903076, "learning_rate": 1e-05, "loss": 0.4895, "mean_token_accuracy": 0.8748670220375061, "num_tokens": 848770795.0, "step": 22351 }, { "epoch": 2.843404147055082, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.9004998207092285, "learning_rate": 1e-05, "loss": 0.4598, "mean_token_accuracy": 0.8830369114875793, "num_tokens": 848815086.0, "step": 22352 }, { "epoch": 2.8435313573336725, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.537867546081543, "learning_rate": 1e-05, "loss": 0.4238, "mean_token_accuracy": 0.8946206569671631, "num_tokens": 848853545.0, "step": 22353 }, { "epoch": 2.843658567612263, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.7526445388793945, "learning_rate": 1e-05, "loss": 0.4607, "mean_token_accuracy": 0.8825393319129944, "num_tokens": 848886047.0, "step": 22354 }, { "epoch": 2.8437857778908535, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.399670600891113, "learning_rate": 1e-05, "loss": 0.4641, "mean_token_accuracy": 0.8811226487159729, "num_tokens": 848929147.0, "step": 22355 }, { "epoch": 2.843912988169444, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.726828098297119, "learning_rate": 1e-05, "loss": 0.4811, "mean_token_accuracy": 0.8754615783691406, "num_tokens": 848962912.0, "step": 22356 }, { "epoch": 2.8440401984480346, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.263022422790527, "learning_rate": 1e-05, "loss": 0.4778, "mean_token_accuracy": 0.8767275214195251, "num_tokens": 848997954.0, "step": 22357 }, { "epoch": 2.844167408726625, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.790132999420166, "learning_rate": 1e-05, "loss": 0.4944, "mean_token_accuracy": 0.8732423782348633, "num_tokens": 849044418.0, "step": 22358 }, { "epoch": 2.8442946190052156, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.09540319442749, "learning_rate": 1e-05, "loss": 0.4726, "mean_token_accuracy": 0.8758414387702942, "num_tokens": 849085591.0, "step": 22359 }, { "epoch": 2.844421829283806, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 7.0004401206970215, "learning_rate": 1e-05, "loss": 0.5009, "mean_token_accuracy": 0.8740432262420654, "num_tokens": 849128046.0, "step": 22360 }, { "epoch": 2.8445490395623967, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 5.970957279205322, "learning_rate": 1e-05, "loss": 0.4893, "mean_token_accuracy": 0.8702985644340515, "num_tokens": 849165231.0, "step": 22361 }, { "epoch": 2.844676249840987, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00011444091796875, "grad_norm": 7.316944122314453, "learning_rate": 1e-05, "loss": 0.4885, "mean_token_accuracy": 0.879055917263031, "num_tokens": 849203414.0, "step": 22362 }, { "epoch": 2.8448034601195777, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 5.965182781219482, "learning_rate": 1e-05, "loss": 0.4512, "mean_token_accuracy": 0.881580650806427, "num_tokens": 849242463.0, "step": 22363 }, { "epoch": 2.8449306703981683, "ewc_loss": 0.1513671875, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.0001220703125, "grad_norm": 7.644932746887207, "learning_rate": 1e-05, "loss": 0.492, "mean_token_accuracy": 0.8816683292388916, "num_tokens": 849280098.0, "step": 22364 }, { "epoch": 2.845057880676759, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.233649730682373, "learning_rate": 1e-05, "loss": 0.4597, "mean_token_accuracy": 0.8814381957054138, "num_tokens": 849317460.0, "step": 22365 }, { "epoch": 2.8451850909553493, "ewc_loss": 0.1533203125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.0001239776611328125, "grad_norm": 7.702116966247559, "learning_rate": 1e-05, "loss": 0.4911, "mean_token_accuracy": 0.8807948231697083, "num_tokens": 849354852.0, "step": 22366 }, { "epoch": 2.84531230123394, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.587213516235352, "learning_rate": 1e-05, "loss": 0.492, "mean_token_accuracy": 0.8767682909965515, "num_tokens": 849397132.0, "step": 22367 }, { "epoch": 2.84543951151253, "ewc_loss": 0.1455078125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00011587142944335938, "grad_norm": 7.351216793060303, "learning_rate": 1e-05, "loss": 0.4897, "mean_token_accuracy": 0.879063069820404, "num_tokens": 849441751.0, "step": 22368 }, { "epoch": 2.845566721791121, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00010967254638671875, "grad_norm": 6.763024806976318, "learning_rate": 1e-05, "loss": 0.5125, "mean_token_accuracy": 0.8701684474945068, "num_tokens": 849480177.0, "step": 22369 }, { "epoch": 2.845693932069711, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.826126575469971, "learning_rate": 1e-05, "loss": 0.5131, "mean_token_accuracy": 0.8706649541854858, "num_tokens": 849520211.0, "step": 22370 }, { "epoch": 2.845821142348302, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 6.845150947570801, "learning_rate": 1e-05, "loss": 0.4806, "mean_token_accuracy": 0.877524733543396, "num_tokens": 849551591.0, "step": 22371 }, { "epoch": 2.845948352626892, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.326690673828125, "learning_rate": 1e-05, "loss": 0.4967, "mean_token_accuracy": 0.8695204257965088, "num_tokens": 849592700.0, "step": 22372 }, { "epoch": 2.846075562905483, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 7.03501558303833, "learning_rate": 1e-05, "loss": 0.5305, "mean_token_accuracy": 0.868245005607605, "num_tokens": 849631220.0, "step": 22373 }, { "epoch": 2.846202773184073, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 5.983277320861816, "learning_rate": 1e-05, "loss": 0.4515, "mean_token_accuracy": 0.8844057321548462, "num_tokens": 849670349.0, "step": 22374 }, { "epoch": 2.8463299834626636, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00011539459228515625, "grad_norm": 7.392156600952148, "learning_rate": 1e-05, "loss": 0.5796, "mean_token_accuracy": 0.850266695022583, "num_tokens": 849711000.0, "step": 22375 }, { "epoch": 2.846457193741254, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 5.914174556732178, "learning_rate": 1e-05, "loss": 0.5023, "mean_token_accuracy": 0.8642612099647522, "num_tokens": 849751199.0, "step": 22376 }, { "epoch": 2.8465844040198447, "ewc_loss": 0.154296875, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00012493133544921875, "grad_norm": 7.796444416046143, "learning_rate": 1e-05, "loss": 0.4378, "mean_token_accuracy": 0.9007697105407715, "num_tokens": 849785416.0, "step": 22377 }, { "epoch": 2.8467116142984352, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.222148418426514, "learning_rate": 1e-05, "loss": 0.5136, "mean_token_accuracy": 0.8679608702659607, "num_tokens": 849818553.0, "step": 22378 }, { "epoch": 2.8468388245770258, "ewc_loss": 0.1591796875, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00012969970703125, "grad_norm": 7.914546012878418, "learning_rate": 1e-05, "loss": 0.4878, "mean_token_accuracy": 0.8829960823059082, "num_tokens": 849854994.0, "step": 22379 }, { "epoch": 2.8469660348556163, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00011157989501953125, "grad_norm": 6.690123558044434, "learning_rate": 1e-05, "loss": 0.538, "mean_token_accuracy": 0.8592067360877991, "num_tokens": 849896603.0, "step": 22380 }, { "epoch": 2.847093245134207, "ewc_loss": 0.15234375, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.608669281005859, "learning_rate": 1e-05, "loss": 0.5457, "mean_token_accuracy": 0.8649364709854126, "num_tokens": 849935999.0, "step": 22381 }, { "epoch": 2.8472204554127973, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00011301040649414062, "grad_norm": 6.795016288757324, "learning_rate": 1e-05, "loss": 0.4716, "mean_token_accuracy": 0.8817022442817688, "num_tokens": 849976858.0, "step": 22382 }, { "epoch": 2.847347665691388, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00011491775512695312, "grad_norm": 7.230106353759766, "learning_rate": 1e-05, "loss": 0.4595, "mean_token_accuracy": 0.8903212547302246, "num_tokens": 850012802.0, "step": 22383 }, { "epoch": 2.8474748759699784, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 6.712836742401123, "learning_rate": 1e-05, "loss": 0.5117, "mean_token_accuracy": 0.8705745935440063, "num_tokens": 850047330.0, "step": 22384 }, { "epoch": 2.847602086248569, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 6.944190979003906, "learning_rate": 1e-05, "loss": 0.5184, "mean_token_accuracy": 0.8702205419540405, "num_tokens": 850087543.0, "step": 22385 }, { "epoch": 2.8477292965271594, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.573736667633057, "learning_rate": 1e-05, "loss": 0.4599, "mean_token_accuracy": 0.8825283050537109, "num_tokens": 850126864.0, "step": 22386 }, { "epoch": 2.84785650680575, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.78213357925415, "learning_rate": 1e-05, "loss": 0.4473, "mean_token_accuracy": 0.8891253471374512, "num_tokens": 850162476.0, "step": 22387 }, { "epoch": 2.8479837170843405, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.417963027954102, "learning_rate": 1e-05, "loss": 0.4841, "mean_token_accuracy": 0.8750413656234741, "num_tokens": 850197606.0, "step": 22388 }, { "epoch": 2.848110927362931, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.771910667419434, "learning_rate": 1e-05, "loss": 0.4884, "mean_token_accuracy": 0.8757334351539612, "num_tokens": 850232223.0, "step": 22389 }, { "epoch": 2.8482381376415216, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.240261554718018, "learning_rate": 1e-05, "loss": 0.4519, "mean_token_accuracy": 0.884652316570282, "num_tokens": 850268936.0, "step": 22390 }, { "epoch": 2.848365347920112, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.882131576538086, "learning_rate": 1e-05, "loss": 0.514, "mean_token_accuracy": 0.8701846599578857, "num_tokens": 850301652.0, "step": 22391 }, { "epoch": 2.8484925581987026, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.085926532745361, "learning_rate": 1e-05, "loss": 0.4524, "mean_token_accuracy": 0.8795245885848999, "num_tokens": 850335033.0, "step": 22392 }, { "epoch": 2.8486197684772927, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00011014938354492188, "grad_norm": 7.096614360809326, "learning_rate": 1e-05, "loss": 0.49, "mean_token_accuracy": 0.876104474067688, "num_tokens": 850369677.0, "step": 22393 }, { "epoch": 2.8487469787558837, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 5.987024307250977, "learning_rate": 1e-05, "loss": 0.5155, "mean_token_accuracy": 0.8639692068099976, "num_tokens": 850411464.0, "step": 22394 }, { "epoch": 2.8488741890344738, "ewc_loss": 0.1474609375, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00011777877807617188, "grad_norm": 7.414667129516602, "learning_rate": 1e-05, "loss": 0.4838, "mean_token_accuracy": 0.8797488212585449, "num_tokens": 850447578.0, "step": 22395 }, { "epoch": 2.8490013993130647, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.0428290367126465, "learning_rate": 1e-05, "loss": 0.4514, "mean_token_accuracy": 0.881250262260437, "num_tokens": 850489689.0, "step": 22396 }, { "epoch": 2.849128609591655, "ewc_loss": 0.15234375, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.669912338256836, "learning_rate": 1e-05, "loss": 0.4821, "mean_token_accuracy": 0.8837944269180298, "num_tokens": 850528598.0, "step": 22397 }, { "epoch": 2.849255819870246, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.374013423919678, "learning_rate": 1e-05, "loss": 0.4827, "mean_token_accuracy": 0.8786621689796448, "num_tokens": 850566953.0, "step": 22398 }, { "epoch": 2.849383030148836, "ewc_loss": 0.150390625, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00012063980102539062, "grad_norm": 7.5705108642578125, "learning_rate": 1e-05, "loss": 0.4793, "mean_token_accuracy": 0.8841476440429688, "num_tokens": 850604286.0, "step": 22399 }, { "epoch": 2.8495102404274264, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 6.681486129760742, "learning_rate": 1e-05, "loss": 0.5141, "mean_token_accuracy": 0.8725026845932007, "num_tokens": 850640724.0, "step": 22400 }, { "epoch": 2.849637450706017, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00011301040649414062, "grad_norm": 7.126420021057129, "learning_rate": 1e-05, "loss": 0.5043, "mean_token_accuracy": 0.8759770393371582, "num_tokens": 850682377.0, "step": 22401 }, { "epoch": 2.8497646609846075, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00010967254638671875, "grad_norm": 6.803211212158203, "learning_rate": 1e-05, "loss": 0.4927, "mean_token_accuracy": 0.8748466968536377, "num_tokens": 850721637.0, "step": 22402 }, { "epoch": 2.849891871263198, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.582175254821777, "learning_rate": 1e-05, "loss": 0.4966, "mean_token_accuracy": 0.8713868856430054, "num_tokens": 850754884.0, "step": 22403 }, { "epoch": 2.8500190815417885, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 6.926393032073975, "learning_rate": 1e-05, "loss": 0.4981, "mean_token_accuracy": 0.8717919588088989, "num_tokens": 850793706.0, "step": 22404 }, { "epoch": 2.850146291820379, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.127050876617432, "learning_rate": 1e-05, "loss": 0.5, "mean_token_accuracy": 0.8665921688079834, "num_tokens": 850827708.0, "step": 22405 }, { "epoch": 2.8502735020989696, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00011157989501953125, "grad_norm": 7.199325084686279, "learning_rate": 1e-05, "loss": 0.5107, "mean_token_accuracy": 0.8677274584770203, "num_tokens": 850870808.0, "step": 22406 }, { "epoch": 2.85040071237756, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.880510330200195, "learning_rate": 1e-05, "loss": 0.4588, "mean_token_accuracy": 0.8797777891159058, "num_tokens": 850906129.0, "step": 22407 }, { "epoch": 2.8505279226561506, "ewc_loss": 0.1513671875, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.0001220703125, "grad_norm": 7.637156009674072, "learning_rate": 1e-05, "loss": 0.5107, "mean_token_accuracy": 0.8777320384979248, "num_tokens": 850944021.0, "step": 22408 }, { "epoch": 2.850655132934741, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 5.994439601898193, "learning_rate": 1e-05, "loss": 0.4652, "mean_token_accuracy": 0.8825514316558838, "num_tokens": 850983040.0, "step": 22409 }, { "epoch": 2.8507823432133317, "ewc_loss": 0.1611328125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.0001316070556640625, "grad_norm": 7.999476432800293, "learning_rate": 1e-05, "loss": 0.5094, "mean_token_accuracy": 0.8774057626724243, "num_tokens": 851017400.0, "step": 22410 }, { "epoch": 2.850909553491922, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 6.513234615325928, "learning_rate": 1e-05, "loss": 0.4811, "mean_token_accuracy": 0.8775233030319214, "num_tokens": 851055559.0, "step": 22411 }, { "epoch": 2.8510367637705127, "ewc_loss": 0.1572265625, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.0001277923583984375, "grad_norm": 7.853233337402344, "learning_rate": 1e-05, "loss": 0.4786, "mean_token_accuracy": 0.8867337703704834, "num_tokens": 851088810.0, "step": 22412 }, { "epoch": 2.8511639740491033, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00011539459228515625, "grad_norm": 6.8701372146606445, "learning_rate": 1e-05, "loss": 0.5051, "mean_token_accuracy": 0.8760582208633423, "num_tokens": 851121531.0, "step": 22413 }, { "epoch": 2.851291184327694, "ewc_loss": 0.1474609375, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.000118255615234375, "grad_norm": 7.413511276245117, "learning_rate": 1e-05, "loss": 0.4916, "mean_token_accuracy": 0.8769440054893494, "num_tokens": 851160601.0, "step": 22414 }, { "epoch": 2.8514183946062843, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.0001125335693359375, "grad_norm": 6.851335525512695, "learning_rate": 1e-05, "loss": 0.5224, "mean_token_accuracy": 0.865217924118042, "num_tokens": 851195887.0, "step": 22415 }, { "epoch": 2.851545604884875, "ewc_loss": 0.140625, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 7.018313884735107, "learning_rate": 1e-05, "loss": 0.4633, "mean_token_accuracy": 0.8831992149353027, "num_tokens": 851231761.0, "step": 22416 }, { "epoch": 2.8516728151634654, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.706099033355713, "learning_rate": 1e-05, "loss": 0.5353, "mean_token_accuracy": 0.8596836924552917, "num_tokens": 851270765.0, "step": 22417 }, { "epoch": 2.8518000254420555, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.781209945678711, "learning_rate": 1e-05, "loss": 0.5483, "mean_token_accuracy": 0.8633498549461365, "num_tokens": 851312200.0, "step": 22418 }, { "epoch": 2.8519272357206464, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.532190322875977, "learning_rate": 1e-05, "loss": 0.498, "mean_token_accuracy": 0.8720687031745911, "num_tokens": 851350854.0, "step": 22419 }, { "epoch": 2.8520544459992365, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.699877738952637, "learning_rate": 1e-05, "loss": 0.5159, "mean_token_accuracy": 0.8635205030441284, "num_tokens": 851389242.0, "step": 22420 }, { "epoch": 2.8521816562778275, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.3497395515441895, "learning_rate": 1e-05, "loss": 0.472, "mean_token_accuracy": 0.8760556578636169, "num_tokens": 851425450.0, "step": 22421 }, { "epoch": 2.8523088665564176, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.750643253326416, "learning_rate": 1e-05, "loss": 0.4836, "mean_token_accuracy": 0.8758615255355835, "num_tokens": 851459470.0, "step": 22422 }, { "epoch": 2.852436076835008, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 6.164677143096924, "learning_rate": 1e-05, "loss": 0.5004, "mean_token_accuracy": 0.8666476011276245, "num_tokens": 851500849.0, "step": 22423 }, { "epoch": 2.8525632871135986, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.914918899536133, "learning_rate": 1e-05, "loss": 0.489, "mean_token_accuracy": 0.8776288628578186, "num_tokens": 851540048.0, "step": 22424 }, { "epoch": 2.852690497392189, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 5.995704650878906, "learning_rate": 1e-05, "loss": 0.4842, "mean_token_accuracy": 0.8735962510108948, "num_tokens": 851573343.0, "step": 22425 }, { "epoch": 2.8528177076707797, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00011348724365234375, "grad_norm": 7.229895114898682, "learning_rate": 1e-05, "loss": 0.5379, "mean_token_accuracy": 0.8611243963241577, "num_tokens": 851613658.0, "step": 22426 }, { "epoch": 2.85294491794937, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 5.910007953643799, "learning_rate": 1e-05, "loss": 0.4363, "mean_token_accuracy": 0.8872957229614258, "num_tokens": 851649961.0, "step": 22427 }, { "epoch": 2.8530721282279607, "ewc_loss": 0.15234375, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.633213996887207, "learning_rate": 1e-05, "loss": 0.5072, "mean_token_accuracy": 0.8757646083831787, "num_tokens": 851685465.0, "step": 22428 }, { "epoch": 2.8531993385065513, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.103187084197998, "learning_rate": 1e-05, "loss": 0.4262, "mean_token_accuracy": 0.8916364312171936, "num_tokens": 851718434.0, "step": 22429 }, { "epoch": 2.853326548785142, "ewc_loss": 0.158203125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00012874603271484375, "grad_norm": 7.865730285644531, "learning_rate": 1e-05, "loss": 0.4948, "mean_token_accuracy": 0.8839403390884399, "num_tokens": 851756207.0, "step": 22430 }, { "epoch": 2.8534537590637323, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.571622371673584, "learning_rate": 1e-05, "loss": 0.5133, "mean_token_accuracy": 0.8719110488891602, "num_tokens": 851795398.0, "step": 22431 }, { "epoch": 2.853580969342323, "ewc_loss": 0.1513671875, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.0001220703125, "grad_norm": 7.602957248687744, "learning_rate": 1e-05, "loss": 0.5382, "mean_token_accuracy": 0.8650873303413391, "num_tokens": 851832569.0, "step": 22432 }, { "epoch": 2.8537081796209134, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00011348724365234375, "grad_norm": 6.832424640655518, "learning_rate": 1e-05, "loss": 0.491, "mean_token_accuracy": 0.8800735473632812, "num_tokens": 851870749.0, "step": 22433 }, { "epoch": 2.853835389899504, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00011205673217773438, "grad_norm": 7.07396125793457, "learning_rate": 1e-05, "loss": 0.4872, "mean_token_accuracy": 0.8797734975814819, "num_tokens": 851912591.0, "step": 22434 }, { "epoch": 2.8539626001780944, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00011014938354492188, "grad_norm": 6.866570949554443, "learning_rate": 1e-05, "loss": 0.5, "mean_token_accuracy": 0.8761602640151978, "num_tokens": 851948792.0, "step": 22435 }, { "epoch": 2.854089810456685, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.5699615478515625, "learning_rate": 1e-05, "loss": 0.5035, "mean_token_accuracy": 0.8679081201553345, "num_tokens": 851988811.0, "step": 22436 }, { "epoch": 2.8542170207352755, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 6.939849853515625, "learning_rate": 1e-05, "loss": 0.4746, "mean_token_accuracy": 0.8840895891189575, "num_tokens": 852024781.0, "step": 22437 }, { "epoch": 2.854344231013866, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.179938793182373, "learning_rate": 1e-05, "loss": 0.4525, "mean_token_accuracy": 0.883489727973938, "num_tokens": 852056676.0, "step": 22438 }, { "epoch": 2.8544714412924566, "ewc_loss": 0.140625, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 7.153042793273926, "learning_rate": 1e-05, "loss": 0.533, "mean_token_accuracy": 0.8646227121353149, "num_tokens": 852095500.0, "step": 22439 }, { "epoch": 2.854598651571047, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 5.960038185119629, "learning_rate": 1e-05, "loss": 0.4422, "mean_token_accuracy": 0.8871533870697021, "num_tokens": 852133658.0, "step": 22440 }, { "epoch": 2.8547258618496376, "ewc_loss": 0.1484375, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00011920928955078125, "grad_norm": 7.488113880157471, "learning_rate": 1e-05, "loss": 0.5442, "mean_token_accuracy": 0.8593218326568604, "num_tokens": 852173963.0, "step": 22441 }, { "epoch": 2.854853072128228, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.018782615661621, "learning_rate": 1e-05, "loss": 0.4799, "mean_token_accuracy": 0.8741180896759033, "num_tokens": 852209957.0, "step": 22442 }, { "epoch": 2.8549802824068182, "ewc_loss": 0.1572265625, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.0001277923583984375, "grad_norm": 7.767979621887207, "learning_rate": 1e-05, "loss": 0.5003, "mean_token_accuracy": 0.8781072497367859, "num_tokens": 852252281.0, "step": 22443 }, { "epoch": 2.855107492685409, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.336187839508057, "learning_rate": 1e-05, "loss": 0.4285, "mean_token_accuracy": 0.8937947750091553, "num_tokens": 852285287.0, "step": 22444 }, { "epoch": 2.8552347029639993, "ewc_loss": 0.15625, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00012683868408203125, "grad_norm": 7.767980098724365, "learning_rate": 1e-05, "loss": 0.5397, "mean_token_accuracy": 0.8673085570335388, "num_tokens": 852326596.0, "step": 22445 }, { "epoch": 2.8553619132425903, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00010967254638671875, "grad_norm": 6.657088756561279, "learning_rate": 1e-05, "loss": 0.466, "mean_token_accuracy": 0.8832404613494873, "num_tokens": 852358062.0, "step": 22446 }, { "epoch": 2.8554891235211803, "ewc_loss": 0.1484375, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00011873245239257812, "grad_norm": 7.483260631561279, "learning_rate": 1e-05, "loss": 0.5387, "mean_token_accuracy": 0.8670099377632141, "num_tokens": 852395828.0, "step": 22447 }, { "epoch": 2.855616333799771, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00011014938354492188, "grad_norm": 6.735268592834473, "learning_rate": 1e-05, "loss": 0.5176, "mean_token_accuracy": 0.8709139227867126, "num_tokens": 852439466.0, "step": 22448 }, { "epoch": 2.8557435440783614, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.0001125335693359375, "grad_norm": 7.1097636222839355, "learning_rate": 1e-05, "loss": 0.4876, "mean_token_accuracy": 0.8788043260574341, "num_tokens": 852477992.0, "step": 22449 }, { "epoch": 2.855870754356952, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.660019874572754, "learning_rate": 1e-05, "loss": 0.4729, "mean_token_accuracy": 0.883386492729187, "num_tokens": 852514814.0, "step": 22450 }, { "epoch": 2.8559979646355425, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.841729640960693, "learning_rate": 1e-05, "loss": 0.475, "mean_token_accuracy": 0.8793489933013916, "num_tokens": 852553866.0, "step": 22451 }, { "epoch": 2.856125174914133, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.548952102661133, "learning_rate": 1e-05, "loss": 0.5153, "mean_token_accuracy": 0.8660506010055542, "num_tokens": 852594291.0, "step": 22452 }, { "epoch": 2.8562523851927235, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.702662944793701, "learning_rate": 1e-05, "loss": 0.4686, "mean_token_accuracy": 0.8826268911361694, "num_tokens": 852630586.0, "step": 22453 }, { "epoch": 2.856379595471314, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.403117656707764, "learning_rate": 1e-05, "loss": 0.4885, "mean_token_accuracy": 0.871745765209198, "num_tokens": 852670013.0, "step": 22454 }, { "epoch": 2.8565068057499046, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.687368869781494, "learning_rate": 1e-05, "loss": 0.485, "mean_token_accuracy": 0.8753300905227661, "num_tokens": 852704097.0, "step": 22455 }, { "epoch": 2.856634016028495, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.2452216148376465, "learning_rate": 1e-05, "loss": 0.4884, "mean_token_accuracy": 0.8706855773925781, "num_tokens": 852739771.0, "step": 22456 }, { "epoch": 2.8567612263070856, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.774083614349365, "learning_rate": 1e-05, "loss": 0.5253, "mean_token_accuracy": 0.865932047367096, "num_tokens": 852778604.0, "step": 22457 }, { "epoch": 2.856888436585676, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.083159923553467, "learning_rate": 1e-05, "loss": 0.542, "mean_token_accuracy": 0.8554561138153076, "num_tokens": 852814966.0, "step": 22458 }, { "epoch": 2.8570156468642667, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.979892730712891, "learning_rate": 1e-05, "loss": 0.5125, "mean_token_accuracy": 0.8663887977600098, "num_tokens": 852849562.0, "step": 22459 }, { "epoch": 2.857142857142857, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 5.950778484344482, "learning_rate": 1e-05, "loss": 0.452, "mean_token_accuracy": 0.8817510008811951, "num_tokens": 852884183.0, "step": 22460 }, { "epoch": 2.8572700674214477, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00011491775512695312, "grad_norm": 7.295764923095703, "learning_rate": 1e-05, "loss": 0.5436, "mean_token_accuracy": 0.8678839802742004, "num_tokens": 852925121.0, "step": 22461 }, { "epoch": 2.8573972777000383, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 5.9311628341674805, "learning_rate": 1e-05, "loss": 0.4585, "mean_token_accuracy": 0.8828084468841553, "num_tokens": 852965962.0, "step": 22462 }, { "epoch": 2.857524487978629, "ewc_loss": 0.15234375, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.652397155761719, "learning_rate": 1e-05, "loss": 0.5154, "mean_token_accuracy": 0.8759475946426392, "num_tokens": 853004886.0, "step": 22463 }, { "epoch": 2.8576516982572193, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.188054084777832, "learning_rate": 1e-05, "loss": 0.4884, "mean_token_accuracy": 0.8708251714706421, "num_tokens": 853038898.0, "step": 22464 }, { "epoch": 2.85777890853581, "ewc_loss": 0.15625, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00012683868408203125, "grad_norm": 7.75435209274292, "learning_rate": 1e-05, "loss": 0.5593, "mean_token_accuracy": 0.8606605529785156, "num_tokens": 853079064.0, "step": 22465 }, { "epoch": 2.8579061188144, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 6.588648319244385, "learning_rate": 1e-05, "loss": 0.5509, "mean_token_accuracy": 0.8588888645172119, "num_tokens": 853125704.0, "step": 22466 }, { "epoch": 2.858033329092991, "ewc_loss": 0.1484375, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00011873245239257812, "grad_norm": 7.421610355377197, "learning_rate": 1e-05, "loss": 0.5202, "mean_token_accuracy": 0.8714636564254761, "num_tokens": 853168760.0, "step": 22467 }, { "epoch": 2.858160539371581, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00011157989501953125, "grad_norm": 6.786296844482422, "learning_rate": 1e-05, "loss": 0.4952, "mean_token_accuracy": 0.8752851486206055, "num_tokens": 853207146.0, "step": 22468 }, { "epoch": 2.858287749650172, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.900955677032471, "learning_rate": 1e-05, "loss": 0.4716, "mean_token_accuracy": 0.8820763230323792, "num_tokens": 853248764.0, "step": 22469 }, { "epoch": 2.858414959928762, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 6.835715293884277, "learning_rate": 1e-05, "loss": 0.4672, "mean_token_accuracy": 0.8822863698005676, "num_tokens": 853283325.0, "step": 22470 }, { "epoch": 2.858542170207353, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.418769836425781, "learning_rate": 1e-05, "loss": 0.4915, "mean_token_accuracy": 0.874940037727356, "num_tokens": 853316579.0, "step": 22471 }, { "epoch": 2.858669380485943, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 6.97166633605957, "learning_rate": 1e-05, "loss": 0.4929, "mean_token_accuracy": 0.8784092664718628, "num_tokens": 853353109.0, "step": 22472 }, { "epoch": 2.8587965907645336, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.055984973907471, "learning_rate": 1e-05, "loss": 0.4718, "mean_token_accuracy": 0.8795832991600037, "num_tokens": 853394974.0, "step": 22473 }, { "epoch": 2.858923801043124, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.0001125335693359375, "grad_norm": 7.25020694732666, "learning_rate": 1e-05, "loss": 0.4491, "mean_token_accuracy": 0.8913086652755737, "num_tokens": 853433819.0, "step": 22474 }, { "epoch": 2.8590510113217147, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 5.916042804718018, "learning_rate": 1e-05, "loss": 0.4709, "mean_token_accuracy": 0.8753202557563782, "num_tokens": 853471947.0, "step": 22475 }, { "epoch": 2.859178221600305, "ewc_loss": 0.15234375, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.626522064208984, "learning_rate": 1e-05, "loss": 0.503, "mean_token_accuracy": 0.8762618899345398, "num_tokens": 853505936.0, "step": 22476 }, { "epoch": 2.8593054318788957, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.099531650543213, "learning_rate": 1e-05, "loss": 0.4256, "mean_token_accuracy": 0.895448625087738, "num_tokens": 853540957.0, "step": 22477 }, { "epoch": 2.8594326421574863, "ewc_loss": 0.158203125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00012874603271484375, "grad_norm": 7.8304362297058105, "learning_rate": 1e-05, "loss": 0.4721, "mean_token_accuracy": 0.8899428844451904, "num_tokens": 853574117.0, "step": 22478 }, { "epoch": 2.859559852436077, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.483274936676025, "learning_rate": 1e-05, "loss": 0.5085, "mean_token_accuracy": 0.8698018789291382, "num_tokens": 853612424.0, "step": 22479 }, { "epoch": 2.8596870627146673, "ewc_loss": 0.154296875, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00012493133544921875, "grad_norm": 7.677709102630615, "learning_rate": 1e-05, "loss": 0.5105, "mean_token_accuracy": 0.8743566274642944, "num_tokens": 853652659.0, "step": 22480 }, { "epoch": 2.859814272993258, "ewc_loss": 0.140625, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00011110305786132812, "grad_norm": 6.717098712921143, "learning_rate": 1e-05, "loss": 0.473, "mean_token_accuracy": 0.8831202387809753, "num_tokens": 853690763.0, "step": 22481 }, { "epoch": 2.8599414832718484, "ewc_loss": 0.1455078125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00011587142944335938, "grad_norm": 7.321239948272705, "learning_rate": 1e-05, "loss": 0.4599, "mean_token_accuracy": 0.889275848865509, "num_tokens": 853727437.0, "step": 22482 }, { "epoch": 2.860068693550439, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 6.710918426513672, "learning_rate": 1e-05, "loss": 0.5257, "mean_token_accuracy": 0.8677527904510498, "num_tokens": 853769179.0, "step": 22483 }, { "epoch": 2.8601959038290294, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 6.994851112365723, "learning_rate": 1e-05, "loss": 0.4468, "mean_token_accuracy": 0.8905256390571594, "num_tokens": 853803374.0, "step": 22484 }, { "epoch": 2.86032311410762, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.607651710510254, "learning_rate": 1e-05, "loss": 0.5069, "mean_token_accuracy": 0.8697980642318726, "num_tokens": 853834286.0, "step": 22485 }, { "epoch": 2.8604503243862105, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.794487953186035, "learning_rate": 1e-05, "loss": 0.4692, "mean_token_accuracy": 0.8833974599838257, "num_tokens": 853872369.0, "step": 22486 }, { "epoch": 2.860577534664801, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.466651916503906, "learning_rate": 1e-05, "loss": 0.5245, "mean_token_accuracy": 0.8657665848731995, "num_tokens": 853916470.0, "step": 22487 }, { "epoch": 2.8607047449433916, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.715588092803955, "learning_rate": 1e-05, "loss": 0.4285, "mean_token_accuracy": 0.8962719440460205, "num_tokens": 853956802.0, "step": 22488 }, { "epoch": 2.860831955221982, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.294459819793701, "learning_rate": 1e-05, "loss": 0.4708, "mean_token_accuracy": 0.8817098140716553, "num_tokens": 853990914.0, "step": 22489 }, { "epoch": 2.8609591655005726, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.787773609161377, "learning_rate": 1e-05, "loss": 0.4713, "mean_token_accuracy": 0.8827602863311768, "num_tokens": 854029707.0, "step": 22490 }, { "epoch": 2.8610863757791627, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.127544403076172, "learning_rate": 1e-05, "loss": 0.409, "mean_token_accuracy": 0.898370623588562, "num_tokens": 854065584.0, "step": 22491 }, { "epoch": 2.8612135860577537, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 6.978962421417236, "learning_rate": 1e-05, "loss": 0.4938, "mean_token_accuracy": 0.8763625621795654, "num_tokens": 854105937.0, "step": 22492 }, { "epoch": 2.8613407963363438, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 5.99036979675293, "learning_rate": 1e-05, "loss": 0.4534, "mean_token_accuracy": 0.8822596073150635, "num_tokens": 854146746.0, "step": 22493 }, { "epoch": 2.8614680066149347, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00011444091796875, "grad_norm": 7.288315773010254, "learning_rate": 1e-05, "loss": 0.4643, "mean_token_accuracy": 0.8882269859313965, "num_tokens": 854179953.0, "step": 22494 }, { "epoch": 2.861595216893525, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 5.970870494842529, "learning_rate": 1e-05, "loss": 0.4695, "mean_token_accuracy": 0.8777320384979248, "num_tokens": 854220765.0, "step": 22495 }, { "epoch": 2.861722427172116, "ewc_loss": 0.15234375, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.60731315612793, "learning_rate": 1e-05, "loss": 0.4976, "mean_token_accuracy": 0.8794690370559692, "num_tokens": 854258544.0, "step": 22496 }, { "epoch": 2.861849637450706, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.205554962158203, "learning_rate": 1e-05, "loss": 0.4914, "mean_token_accuracy": 0.8773139119148254, "num_tokens": 854297250.0, "step": 22497 }, { "epoch": 2.8619768477292964, "ewc_loss": 0.1533203125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.0001239776611328125, "grad_norm": 7.680142879486084, "learning_rate": 1e-05, "loss": 0.517, "mean_token_accuracy": 0.8737902045249939, "num_tokens": 854340283.0, "step": 22498 }, { "epoch": 2.862104058007887, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.558377265930176, "learning_rate": 1e-05, "loss": 0.446, "mean_token_accuracy": 0.8914765119552612, "num_tokens": 854380145.0, "step": 22499 }, { "epoch": 2.8622312682864774, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00011730194091796875, "grad_norm": 7.375606060028076, "learning_rate": 1e-05, "loss": 0.4991, "mean_token_accuracy": 0.877727210521698, "num_tokens": 854422809.0, "step": 22500 }, { "epoch": 2.862358478565068, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00010967254638671875, "grad_norm": 6.740021228790283, "learning_rate": 1e-05, "loss": 0.47, "mean_token_accuracy": 0.8838757872581482, "num_tokens": 854461926.0, "step": 22501 }, { "epoch": 2.8624856888436585, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.872628211975098, "learning_rate": 1e-05, "loss": 0.5209, "mean_token_accuracy": 0.8666445016860962, "num_tokens": 854502431.0, "step": 22502 }, { "epoch": 2.862612899122249, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 6.820487976074219, "learning_rate": 1e-05, "loss": 0.4371, "mean_token_accuracy": 0.8923265933990479, "num_tokens": 854537440.0, "step": 22503 }, { "epoch": 2.8627401094008396, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.38115930557251, "learning_rate": 1e-05, "loss": 0.4662, "mean_token_accuracy": 0.8822140693664551, "num_tokens": 854575121.0, "step": 22504 }, { "epoch": 2.86286731967943, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 6.985203266143799, "learning_rate": 1e-05, "loss": 0.4458, "mean_token_accuracy": 0.8915300369262695, "num_tokens": 854603624.0, "step": 22505 }, { "epoch": 2.8629945299580206, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.0221028327941895, "learning_rate": 1e-05, "loss": 0.513, "mean_token_accuracy": 0.8603286743164062, "num_tokens": 854644616.0, "step": 22506 }, { "epoch": 2.863121740236611, "ewc_loss": 0.1455078125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00011587142944335938, "grad_norm": 7.303271293640137, "learning_rate": 1e-05, "loss": 0.514, "mean_token_accuracy": 0.8750367164611816, "num_tokens": 854689772.0, "step": 22507 }, { "epoch": 2.8632489505152017, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 5.8944549560546875, "learning_rate": 1e-05, "loss": 0.4387, "mean_token_accuracy": 0.886959433555603, "num_tokens": 854725549.0, "step": 22508 }, { "epoch": 2.863376160793792, "ewc_loss": 0.1533203125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.0001239776611328125, "grad_norm": 7.700485706329346, "learning_rate": 1e-05, "loss": 0.4896, "mean_token_accuracy": 0.8836613893508911, "num_tokens": 854768509.0, "step": 22509 }, { "epoch": 2.8635033710723827, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.140068531036377, "learning_rate": 1e-05, "loss": 0.4904, "mean_token_accuracy": 0.8722935914993286, "num_tokens": 854804822.0, "step": 22510 }, { "epoch": 2.8636305813509733, "ewc_loss": 0.158203125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00012874603271484375, "grad_norm": 7.892425537109375, "learning_rate": 1e-05, "loss": 0.5001, "mean_token_accuracy": 0.8819655776023865, "num_tokens": 854837458.0, "step": 22511 }, { "epoch": 2.863757791629564, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.562491416931152, "learning_rate": 1e-05, "loss": 0.4895, "mean_token_accuracy": 0.8779314756393433, "num_tokens": 854877996.0, "step": 22512 }, { "epoch": 2.8638850019081543, "ewc_loss": 0.154296875, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00012493133544921875, "grad_norm": 7.685255527496338, "learning_rate": 1e-05, "loss": 0.5294, "mean_token_accuracy": 0.8692808151245117, "num_tokens": 854913078.0, "step": 22513 }, { "epoch": 2.864012212186745, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.0001125335693359375, "grad_norm": 6.782446384429932, "learning_rate": 1e-05, "loss": 0.47, "mean_token_accuracy": 0.8845325708389282, "num_tokens": 854952300.0, "step": 22514 }, { "epoch": 2.8641394224653354, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00011491775512695312, "grad_norm": 7.2996625900268555, "learning_rate": 1e-05, "loss": 0.4697, "mean_token_accuracy": 0.883621096611023, "num_tokens": 854992657.0, "step": 22515 }, { "epoch": 2.8642666327439255, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 6.750919342041016, "learning_rate": 1e-05, "loss": 0.4847, "mean_token_accuracy": 0.8789392113685608, "num_tokens": 855028147.0, "step": 22516 }, { "epoch": 2.8643938430225164, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.968153953552246, "learning_rate": 1e-05, "loss": 0.4979, "mean_token_accuracy": 0.875264048576355, "num_tokens": 855069930.0, "step": 22517 }, { "epoch": 2.8645210533011065, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.622549057006836, "learning_rate": 1e-05, "loss": 0.4717, "mean_token_accuracy": 0.8805989027023315, "num_tokens": 855105224.0, "step": 22518 }, { "epoch": 2.8646482635796975, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.777558326721191, "learning_rate": 1e-05, "loss": 0.5218, "mean_token_accuracy": 0.864230215549469, "num_tokens": 855139533.0, "step": 22519 }, { "epoch": 2.8647754738582876, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.482516765594482, "learning_rate": 1e-05, "loss": 0.5268, "mean_token_accuracy": 0.8641825914382935, "num_tokens": 855178380.0, "step": 22520 }, { "epoch": 2.864902684136878, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.724344253540039, "learning_rate": 1e-05, "loss": 0.4928, "mean_token_accuracy": 0.8743784427642822, "num_tokens": 855212700.0, "step": 22521 }, { "epoch": 2.8650298944154686, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.308975696563721, "learning_rate": 1e-05, "loss": 0.4545, "mean_token_accuracy": 0.8837831020355225, "num_tokens": 855250340.0, "step": 22522 }, { "epoch": 2.865157104694059, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.779322624206543, "learning_rate": 1e-05, "loss": 0.4749, "mean_token_accuracy": 0.8784914016723633, "num_tokens": 855285341.0, "step": 22523 }, { "epoch": 2.8652843149726497, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.1284613609313965, "learning_rate": 1e-05, "loss": 0.4786, "mean_token_accuracy": 0.8759797811508179, "num_tokens": 855331038.0, "step": 22524 }, { "epoch": 2.86541152525124, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.963168621063232, "learning_rate": 1e-05, "loss": 0.4616, "mean_token_accuracy": 0.8828704357147217, "num_tokens": 855363643.0, "step": 22525 }, { "epoch": 2.8655387355298307, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 5.997103214263916, "learning_rate": 1e-05, "loss": 0.4474, "mean_token_accuracy": 0.8852983117103577, "num_tokens": 855397342.0, "step": 22526 }, { "epoch": 2.8656659458084213, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00011348724365234375, "grad_norm": 7.262448787689209, "learning_rate": 1e-05, "loss": 0.4578, "mean_token_accuracy": 0.8887003660202026, "num_tokens": 855432867.0, "step": 22527 }, { "epoch": 2.865793156087012, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 5.966316223144531, "learning_rate": 1e-05, "loss": 0.4892, "mean_token_accuracy": 0.8719083070755005, "num_tokens": 855466091.0, "step": 22528 }, { "epoch": 2.8659203663656023, "ewc_loss": 0.1513671875, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00012159347534179688, "grad_norm": 7.60560941696167, "learning_rate": 1e-05, "loss": 0.5397, "mean_token_accuracy": 0.8653533458709717, "num_tokens": 855501200.0, "step": 22529 }, { "epoch": 2.866047576644193, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.150405406951904, "learning_rate": 1e-05, "loss": 0.4494, "mean_token_accuracy": 0.8875421285629272, "num_tokens": 855543625.0, "step": 22530 }, { "epoch": 2.8661747869227834, "ewc_loss": 0.15625, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00012683868408203125, "grad_norm": 7.748215198516846, "learning_rate": 1e-05, "loss": 0.5339, "mean_token_accuracy": 0.8725913763046265, "num_tokens": 855579653.0, "step": 22531 }, { "epoch": 2.866301997201374, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.5481719970703125, "learning_rate": 1e-05, "loss": 0.4825, "mean_token_accuracy": 0.8775012493133545, "num_tokens": 855611081.0, "step": 22532 }, { "epoch": 2.8664292074799644, "ewc_loss": 0.1474609375, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.000118255615234375, "grad_norm": 7.466353893280029, "learning_rate": 1e-05, "loss": 0.4599, "mean_token_accuracy": 0.8900442123413086, "num_tokens": 855642389.0, "step": 22533 }, { "epoch": 2.866556417758555, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 6.760704517364502, "learning_rate": 1e-05, "loss": 0.5065, "mean_token_accuracy": 0.8722066879272461, "num_tokens": 855686240.0, "step": 22534 }, { "epoch": 2.8666836280371455, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 6.964651584625244, "learning_rate": 1e-05, "loss": 0.4999, "mean_token_accuracy": 0.8722054958343506, "num_tokens": 855727979.0, "step": 22535 }, { "epoch": 2.866810838315736, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 6.841718673706055, "learning_rate": 1e-05, "loss": 0.5228, "mean_token_accuracy": 0.866696298122406, "num_tokens": 855769119.0, "step": 22536 }, { "epoch": 2.8669380485943265, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.43166446685791, "learning_rate": 1e-05, "loss": 0.4297, "mean_token_accuracy": 0.8889771103858948, "num_tokens": 855811472.0, "step": 22537 }, { "epoch": 2.867065258872917, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 6.990579128265381, "learning_rate": 1e-05, "loss": 0.4759, "mean_token_accuracy": 0.879914402961731, "num_tokens": 855842739.0, "step": 22538 }, { "epoch": 2.8671924691515076, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.052818298339844, "learning_rate": 1e-05, "loss": 0.4783, "mean_token_accuracy": 0.8760901689529419, "num_tokens": 855874520.0, "step": 22539 }, { "epoch": 2.867319679430098, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00011348724365234375, "grad_norm": 7.276067733764648, "learning_rate": 1e-05, "loss": 0.4637, "mean_token_accuracy": 0.8861422538757324, "num_tokens": 855915151.0, "step": 22540 }, { "epoch": 2.867446889708688, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 5.914806842803955, "learning_rate": 1e-05, "loss": 0.5137, "mean_token_accuracy": 0.8621081113815308, "num_tokens": 855948944.0, "step": 22541 }, { "epoch": 2.867574099987279, "ewc_loss": 0.15234375, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.658506393432617, "learning_rate": 1e-05, "loss": 0.4987, "mean_token_accuracy": 0.8751037120819092, "num_tokens": 855981604.0, "step": 22542 }, { "epoch": 2.8677013102658693, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.097258567810059, "learning_rate": 1e-05, "loss": 0.4443, "mean_token_accuracy": 0.8876663446426392, "num_tokens": 856020407.0, "step": 22543 }, { "epoch": 2.8678285205444602, "ewc_loss": 0.158203125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00012874603271484375, "grad_norm": 7.882008075714111, "learning_rate": 1e-05, "loss": 0.5434, "mean_token_accuracy": 0.8650642037391663, "num_tokens": 856056514.0, "step": 22544 }, { "epoch": 2.8679557308230503, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.5319390296936035, "learning_rate": 1e-05, "loss": 0.4965, "mean_token_accuracy": 0.8751364350318909, "num_tokens": 856091385.0, "step": 22545 }, { "epoch": 2.868082941101641, "ewc_loss": 0.154296875, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00012493133544921875, "grad_norm": 7.680782318115234, "learning_rate": 1e-05, "loss": 0.4531, "mean_token_accuracy": 0.8956687450408936, "num_tokens": 856133077.0, "step": 22546 }, { "epoch": 2.8682101513802314, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00011205673217773438, "grad_norm": 6.7544684410095215, "learning_rate": 1e-05, "loss": 0.4747, "mean_token_accuracy": 0.8827565908432007, "num_tokens": 856174682.0, "step": 22547 }, { "epoch": 2.868337361658822, "ewc_loss": 0.1455078125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.0001163482666015625, "grad_norm": 7.325748443603516, "learning_rate": 1e-05, "loss": 0.5373, "mean_token_accuracy": 0.861110270023346, "num_tokens": 856214486.0, "step": 22548 }, { "epoch": 2.8684645719374124, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 6.748148441314697, "learning_rate": 1e-05, "loss": 0.501, "mean_token_accuracy": 0.8689590692520142, "num_tokens": 856246816.0, "step": 22549 }, { "epoch": 2.868591782216003, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 6.981908798217773, "learning_rate": 1e-05, "loss": 0.5211, "mean_token_accuracy": 0.8708041310310364, "num_tokens": 856286232.0, "step": 22550 }, { "epoch": 2.8687189924945935, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.627849578857422, "learning_rate": 1e-05, "loss": 0.4555, "mean_token_accuracy": 0.8840912580490112, "num_tokens": 856326451.0, "step": 22551 }, { "epoch": 2.868846202773184, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.763465881347656, "learning_rate": 1e-05, "loss": 0.5044, "mean_token_accuracy": 0.8726794719696045, "num_tokens": 856367766.0, "step": 22552 }, { "epoch": 2.8689734130517746, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.493450164794922, "learning_rate": 1e-05, "loss": 0.4906, "mean_token_accuracy": 0.8720660209655762, "num_tokens": 856396085.0, "step": 22553 }, { "epoch": 2.869100623330365, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.680703163146973, "learning_rate": 1e-05, "loss": 0.4394, "mean_token_accuracy": 0.8910453915596008, "num_tokens": 856439835.0, "step": 22554 }, { "epoch": 2.8692278336089556, "ewc_loss": 0.125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.334955215454102, "learning_rate": 1e-05, "loss": 0.4518, "mean_token_accuracy": 0.8846898078918457, "num_tokens": 856470890.0, "step": 22555 }, { "epoch": 2.869355043887546, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.7256245613098145, "learning_rate": 1e-05, "loss": 0.5216, "mean_token_accuracy": 0.8680258393287659, "num_tokens": 856510281.0, "step": 22556 }, { "epoch": 2.8694822541661367, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.171791076660156, "learning_rate": 1e-05, "loss": 0.5093, "mean_token_accuracy": 0.8648711442947388, "num_tokens": 856558194.0, "step": 22557 }, { "epoch": 2.869609464444727, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.878376483917236, "learning_rate": 1e-05, "loss": 0.4814, "mean_token_accuracy": 0.877860963344574, "num_tokens": 856594001.0, "step": 22558 }, { "epoch": 2.8697366747233177, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 6.014815330505371, "learning_rate": 1e-05, "loss": 0.4544, "mean_token_accuracy": 0.8803407549858093, "num_tokens": 856636424.0, "step": 22559 }, { "epoch": 2.8698638850019083, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00011205673217773438, "grad_norm": 7.1715874671936035, "learning_rate": 1e-05, "loss": 0.52, "mean_token_accuracy": 0.8666640520095825, "num_tokens": 856670561.0, "step": 22560 }, { "epoch": 2.869991095280499, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 5.912042140960693, "learning_rate": 1e-05, "loss": 0.4944, "mean_token_accuracy": 0.8679195046424866, "num_tokens": 856704852.0, "step": 22561 }, { "epoch": 2.8701183055590893, "ewc_loss": 0.1513671875, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.00012159347534179688, "grad_norm": 7.55941915512085, "learning_rate": 1e-05, "loss": 0.4839, "mean_token_accuracy": 0.8843340873718262, "num_tokens": 856743943.0, "step": 22562 }, { "epoch": 2.87024551583768, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.042141437530518, "learning_rate": 1e-05, "loss": 0.4943, "mean_token_accuracy": 0.8686904907226562, "num_tokens": 856782944.0, "step": 22563 }, { "epoch": 2.87037272611627, "ewc_loss": 0.1572265625, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.0001277923583984375, "grad_norm": 7.861242294311523, "learning_rate": 1e-05, "loss": 0.5067, "mean_token_accuracy": 0.8765485286712646, "num_tokens": 856821240.0, "step": 22564 }, { "epoch": 2.870499936394861, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.460010051727295, "learning_rate": 1e-05, "loss": 0.4942, "mean_token_accuracy": 0.8737781643867493, "num_tokens": 856860113.0, "step": 22565 }, { "epoch": 2.870627146673451, "ewc_loss": 0.1533203125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 0.0001239776611328125, "grad_norm": 7.697882175445557, "learning_rate": 1e-05, "loss": 0.5002, "mean_token_accuracy": 0.8777148127555847, "num_tokens": 856902501.0, "step": 22566 }, { "epoch": 2.870754356952042, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 0.0001125335693359375, "grad_norm": 6.791234970092773, "learning_rate": 1e-05, "loss": 0.4704, "mean_token_accuracy": 0.88520348072052, "num_tokens": 856940023.0, "step": 22567 }, { "epoch": 2.870881567230632, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 0.00011444091796875, "grad_norm": 7.220800399780273, "learning_rate": 1e-05, "loss": 0.5628, "mean_token_accuracy": 0.855900764465332, "num_tokens": 856980653.0, "step": 22568 }, { "epoch": 2.871008777509223, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 0.00011157989501953125, "grad_norm": 6.867069244384766, "learning_rate": 1e-05, "loss": 0.4656, "mean_token_accuracy": 0.8839755058288574, "num_tokens": 857008081.0, "step": 22569 }, { "epoch": 2.871135987787813, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.69021463394165, "learning_rate": 1e-05, "loss": 0.4398, "mean_token_accuracy": 0.8899781703948975, "num_tokens": 857040884.0, "step": 22570 }, { "epoch": 2.8712631980664036, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 6.917646408081055, "learning_rate": 1e-05, "loss": 0.5094, "mean_token_accuracy": 0.8730719685554504, "num_tokens": 857078519.0, "step": 22571 }, { "epoch": 2.871390408344994, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.260531425476074, "learning_rate": 1e-05, "loss": 0.4756, "mean_token_accuracy": 0.8774704337120056, "num_tokens": 857119250.0, "step": 22572 }, { "epoch": 2.8715176186235847, "ewc_loss": 0.140625, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 0.00011014938354492188, "grad_norm": 7.077378749847412, "learning_rate": 1e-05, "loss": 0.4838, "mean_token_accuracy": 0.8785707950592041, "num_tokens": 857160003.0, "step": 22573 }, { "epoch": 2.871644828902175, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 6.008932590484619, "learning_rate": 1e-05, "loss": 0.4594, "mean_token_accuracy": 0.8796971440315247, "num_tokens": 857200918.0, "step": 22574 }, { "epoch": 2.8717720391807657, "ewc_loss": 0.1474609375, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 0.00011777877807617188, "grad_norm": 7.36284875869751, "learning_rate": 1e-05, "loss": 0.471, "mean_token_accuracy": 0.8859471678733826, "num_tokens": 857241157.0, "step": 22575 }, { "epoch": 2.8718992494593563, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 5.993175029754639, "learning_rate": 1e-05, "loss": 0.4625, "mean_token_accuracy": 0.8763514161109924, "num_tokens": 857276237.0, "step": 22576 }, { "epoch": 2.872026459737947, "ewc_loss": 0.154296875, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 0.0001239776611328125, "grad_norm": 7.667786598205566, "learning_rate": 1e-05, "loss": 0.46, "mean_token_accuracy": 0.8939027190208435, "num_tokens": 857311059.0, "step": 22577 }, { "epoch": 2.8721536700165373, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.224707126617432, "learning_rate": 1e-05, "loss": 0.528, "mean_token_accuracy": 0.8623968362808228, "num_tokens": 857350326.0, "step": 22578 }, { "epoch": 2.872280880295128, "ewc_loss": 0.15625, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 0.00012683868408203125, "grad_norm": 7.757937908172607, "learning_rate": 1e-05, "loss": 0.5331, "mean_token_accuracy": 0.8664997816085815, "num_tokens": 857387708.0, "step": 22579 }, { "epoch": 2.8724080905737184, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.543280601501465, "learning_rate": 1e-05, "loss": 0.4753, "mean_token_accuracy": 0.882230818271637, "num_tokens": 857427658.0, "step": 22580 }, { "epoch": 2.872535300852309, "ewc_loss": 0.150390625, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 0.00012063980102539062, "grad_norm": 7.536394119262695, "learning_rate": 1e-05, "loss": 0.5204, "mean_token_accuracy": 0.8740915060043335, "num_tokens": 857468529.0, "step": 22581 }, { "epoch": 2.8726625111308994, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 6.676795959472656, "learning_rate": 1e-05, "loss": 0.4873, "mean_token_accuracy": 0.8804277181625366, "num_tokens": 857509406.0, "step": 22582 }, { "epoch": 2.87278972140949, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 0.00011396408081054688, "grad_norm": 7.203215599060059, "learning_rate": 1e-05, "loss": 0.4999, "mean_token_accuracy": 0.87943434715271, "num_tokens": 857548302.0, "step": 22583 }, { "epoch": 2.8729169316880805, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.6492919921875, "learning_rate": 1e-05, "loss": 0.4331, "mean_token_accuracy": 0.8936487436294556, "num_tokens": 857584022.0, "step": 22584 }, { "epoch": 2.873044141966671, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.918216705322266, "learning_rate": 1e-05, "loss": 0.5104, "mean_token_accuracy": 0.8694816827774048, "num_tokens": 857627334.0, "step": 22585 }, { "epoch": 2.8731713522452615, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.555159091949463, "learning_rate": 1e-05, "loss": 0.4869, "mean_token_accuracy": 0.8777132630348206, "num_tokens": 857663317.0, "step": 22586 }, { "epoch": 2.873298562523852, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.766848564147949, "learning_rate": 1e-05, "loss": 0.524, "mean_token_accuracy": 0.8619492053985596, "num_tokens": 857697358.0, "step": 22587 }, { "epoch": 2.8734257728024426, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.413036823272705, "learning_rate": 1e-05, "loss": 0.4945, "mean_token_accuracy": 0.8738112449645996, "num_tokens": 857737686.0, "step": 22588 }, { "epoch": 2.8735529830810327, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.73748779296875, "learning_rate": 1e-05, "loss": 0.5173, "mean_token_accuracy": 0.8673628568649292, "num_tokens": 857774730.0, "step": 22589 }, { "epoch": 2.8736801933596237, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.259546279907227, "learning_rate": 1e-05, "loss": 0.5017, "mean_token_accuracy": 0.8718913793563843, "num_tokens": 857811503.0, "step": 22590 }, { "epoch": 2.8738074036382137, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.825290679931641, "learning_rate": 1e-05, "loss": 0.5075, "mean_token_accuracy": 0.8696690797805786, "num_tokens": 857850696.0, "step": 22591 }, { "epoch": 2.8739346139168047, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.102573394775391, "learning_rate": 1e-05, "loss": 0.455, "mean_token_accuracy": 0.8831042051315308, "num_tokens": 857894849.0, "step": 22592 }, { "epoch": 2.874061824195395, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 7.021454334259033, "learning_rate": 1e-05, "loss": 0.4671, "mean_token_accuracy": 0.882440447807312, "num_tokens": 857938876.0, "step": 22593 }, { "epoch": 2.8741890344739858, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.001617431640625, "learning_rate": 1e-05, "loss": 0.4549, "mean_token_accuracy": 0.8832608461380005, "num_tokens": 857973424.0, "step": 22594 }, { "epoch": 2.874316244752576, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 0.00011396408081054688, "grad_norm": 7.320387840270996, "learning_rate": 1e-05, "loss": 0.5112, "mean_token_accuracy": 0.8704804182052612, "num_tokens": 858008105.0, "step": 22595 }, { "epoch": 2.8744434550311664, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 5.999521255493164, "learning_rate": 1e-05, "loss": 0.4649, "mean_token_accuracy": 0.8813425302505493, "num_tokens": 858043898.0, "step": 22596 }, { "epoch": 2.874570665309757, "ewc_loss": 0.15234375, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 0.0001220703125, "grad_norm": 7.6005048751831055, "learning_rate": 1e-05, "loss": 0.5509, "mean_token_accuracy": 0.8624175786972046, "num_tokens": 858080671.0, "step": 22597 }, { "epoch": 2.8746978755883474, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.233631134033203, "learning_rate": 1e-05, "loss": 0.4309, "mean_token_accuracy": 0.8918881416320801, "num_tokens": 858118619.0, "step": 22598 }, { "epoch": 2.874825085866938, "ewc_loss": 0.15234375, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 0.0001220703125, "grad_norm": 7.6382622718811035, "learning_rate": 1e-05, "loss": 0.5144, "mean_token_accuracy": 0.8751505017280579, "num_tokens": 858160207.0, "step": 22599 }, { "epoch": 2.8749522961455285, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.5630316734313965, "learning_rate": 1e-05, "loss": 0.5136, "mean_token_accuracy": 0.8694256544113159, "num_tokens": 858201634.0, "step": 22600 }, { "epoch": 2.875079506424119, "ewc_loss": 0.1455078125, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 0.00011587142944335938, "grad_norm": 7.312812328338623, "learning_rate": 1e-05, "loss": 0.5266, "mean_token_accuracy": 0.8686761260032654, "num_tokens": 858241618.0, "step": 22601 }, { "epoch": 2.8752067167027096, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 6.7439398765563965, "learning_rate": 1e-05, "loss": 0.4801, "mean_token_accuracy": 0.8753384351730347, "num_tokens": 858275393.0, "step": 22602 }, { "epoch": 2.8753339269813, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.794125556945801, "learning_rate": 1e-05, "loss": 0.4941, "mean_token_accuracy": 0.8727696537971497, "num_tokens": 858316470.0, "step": 22603 }, { "epoch": 2.8754611372598906, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 6.850359916687012, "learning_rate": 1e-05, "loss": 0.4659, "mean_token_accuracy": 0.8834511041641235, "num_tokens": 858350567.0, "step": 22604 }, { "epoch": 2.875588347538481, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.276060104370117, "learning_rate": 1e-05, "loss": 0.4889, "mean_token_accuracy": 0.8766021132469177, "num_tokens": 858393998.0, "step": 22605 }, { "epoch": 2.8757155578170717, "ewc_loss": 0.140625, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 7.075109481811523, "learning_rate": 1e-05, "loss": 0.472, "mean_token_accuracy": 0.8887860178947449, "num_tokens": 858430987.0, "step": 22606 }, { "epoch": 2.875842768095662, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.944504737854004, "learning_rate": 1e-05, "loss": 0.4756, "mean_token_accuracy": 0.8728162050247192, "num_tokens": 858464230.0, "step": 22607 }, { "epoch": 2.8759699783742527, "ewc_loss": 0.1484375, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 0.00011920928955078125, "grad_norm": 7.462905406951904, "learning_rate": 1e-05, "loss": 0.5186, "mean_token_accuracy": 0.871795654296875, "num_tokens": 858505965.0, "step": 22608 }, { "epoch": 2.8760971886528433, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 5.908048629760742, "learning_rate": 1e-05, "loss": 0.4533, "mean_token_accuracy": 0.8803685903549194, "num_tokens": 858542629.0, "step": 22609 }, { "epoch": 2.876224398931434, "ewc_loss": 0.16015625, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 0.00013065338134765625, "grad_norm": 7.881731986999512, "learning_rate": 1e-05, "loss": 0.4744, "mean_token_accuracy": 0.8879271745681763, "num_tokens": 858583377.0, "step": 22610 }, { "epoch": 2.8763516092100243, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.303542137145996, "learning_rate": 1e-05, "loss": 0.4487, "mean_token_accuracy": 0.8910747170448303, "num_tokens": 858620546.0, "step": 22611 }, { "epoch": 2.876478819488615, "ewc_loss": 0.16015625, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 0.00013065338134765625, "grad_norm": 7.934838771820068, "learning_rate": 1e-05, "loss": 0.5356, "mean_token_accuracy": 0.8704317808151245, "num_tokens": 858659469.0, "step": 22612 }, { "epoch": 2.8766060297672054, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 0.00011396408081054688, "grad_norm": 6.7599382400512695, "learning_rate": 1e-05, "loss": 0.4932, "mean_token_accuracy": 0.8772444725036621, "num_tokens": 858694519.0, "step": 22613 }, { "epoch": 2.8767332400457954, "ewc_loss": 0.150390625, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 0.00012111663818359375, "grad_norm": 7.5567755699157715, "learning_rate": 1e-05, "loss": 0.4751, "mean_token_accuracy": 0.8864888548851013, "num_tokens": 858728371.0, "step": 22614 }, { "epoch": 2.8768604503243864, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 0.00011348724365234375, "grad_norm": 6.850064277648926, "learning_rate": 1e-05, "loss": 0.4697, "mean_token_accuracy": 0.8845118880271912, "num_tokens": 858765195.0, "step": 22615 }, { "epoch": 2.8769876606029765, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 0.00011301040649414062, "grad_norm": 7.142232418060303, "learning_rate": 1e-05, "loss": 0.4647, "mean_token_accuracy": 0.8859102129936218, "num_tokens": 858804129.0, "step": 22616 }, { "epoch": 2.8771148708815675, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.749965190887451, "learning_rate": 1e-05, "loss": 0.4707, "mean_token_accuracy": 0.8851157426834106, "num_tokens": 858841267.0, "step": 22617 }, { "epoch": 2.8772420811601576, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.836461544036865, "learning_rate": 1e-05, "loss": 0.4849, "mean_token_accuracy": 0.8761997222900391, "num_tokens": 858882405.0, "step": 22618 }, { "epoch": 2.877369291438748, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.603903293609619, "learning_rate": 1e-05, "loss": 0.4696, "mean_token_accuracy": 0.8809760808944702, "num_tokens": 858920176.0, "step": 22619 }, { "epoch": 2.8774965017173386, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.696654796600342, "learning_rate": 1e-05, "loss": 0.4988, "mean_token_accuracy": 0.8719812631607056, "num_tokens": 858959065.0, "step": 22620 }, { "epoch": 2.877623711995929, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.437465190887451, "learning_rate": 1e-05, "loss": 0.4491, "mean_token_accuracy": 0.8874553442001343, "num_tokens": 858997497.0, "step": 22621 }, { "epoch": 2.8777509222745197, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.700303554534912, "learning_rate": 1e-05, "loss": 0.459, "mean_token_accuracy": 0.8842998147010803, "num_tokens": 859036425.0, "step": 22622 }, { "epoch": 2.87787813255311, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.2594428062438965, "learning_rate": 1e-05, "loss": 0.4797, "mean_token_accuracy": 0.876275897026062, "num_tokens": 859073602.0, "step": 22623 }, { "epoch": 2.8780053428317007, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.8173112869262695, "learning_rate": 1e-05, "loss": 0.503, "mean_token_accuracy": 0.8741950988769531, "num_tokens": 859107576.0, "step": 22624 }, { "epoch": 2.8781325531102913, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.074558258056641, "learning_rate": 1e-05, "loss": 0.4955, "mean_token_accuracy": 0.8695889711380005, "num_tokens": 859141866.0, "step": 22625 }, { "epoch": 2.878259763388882, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 7.0641069412231445, "learning_rate": 1e-05, "loss": 0.4584, "mean_token_accuracy": 0.8870928287506104, "num_tokens": 859180592.0, "step": 22626 }, { "epoch": 2.8783869736674723, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 5.922033786773682, "learning_rate": 1e-05, "loss": 0.5032, "mean_token_accuracy": 0.8726382851600647, "num_tokens": 859216365.0, "step": 22627 }, { "epoch": 2.878514183946063, "ewc_loss": 0.1484375, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 0.00011873245239257812, "grad_norm": 7.450335502624512, "learning_rate": 1e-05, "loss": 0.509, "mean_token_accuracy": 0.874458372592926, "num_tokens": 859254167.0, "step": 22628 }, { "epoch": 2.8786413942246534, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 5.953122138977051, "learning_rate": 1e-05, "loss": 0.4506, "mean_token_accuracy": 0.884463906288147, "num_tokens": 859288980.0, "step": 22629 }, { "epoch": 2.878768604503244, "ewc_loss": 0.158203125, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 0.0001277923583984375, "grad_norm": 7.838282108306885, "learning_rate": 1e-05, "loss": 0.4641, "mean_token_accuracy": 0.8934922218322754, "num_tokens": 859326741.0, "step": 22630 }, { "epoch": 2.8788958147818344, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.338874340057373, "learning_rate": 1e-05, "loss": 0.493, "mean_token_accuracy": 0.87173992395401, "num_tokens": 859369530.0, "step": 22631 }, { "epoch": 2.879023025060425, "ewc_loss": 0.158203125, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 0.00012874603271484375, "grad_norm": 7.825583457946777, "learning_rate": 1e-05, "loss": 0.5589, "mean_token_accuracy": 0.8599322438240051, "num_tokens": 859409870.0, "step": 22632 }, { "epoch": 2.8791502353390155, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 0.00011348724365234375, "grad_norm": 6.7760772705078125, "learning_rate": 1e-05, "loss": 0.5347, "mean_token_accuracy": 0.8626041412353516, "num_tokens": 859443419.0, "step": 22633 }, { "epoch": 2.879277445617606, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 0.00011730194091796875, "grad_norm": 7.385927200317383, "learning_rate": 1e-05, "loss": 0.4784, "mean_token_accuracy": 0.8829148411750793, "num_tokens": 859474832.0, "step": 22634 }, { "epoch": 2.8794046558961965, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 0.00011205673217773438, "grad_norm": 6.866570949554443, "learning_rate": 1e-05, "loss": 0.5094, "mean_token_accuracy": 0.8751270771026611, "num_tokens": 859517668.0, "step": 22635 }, { "epoch": 2.879531866174787, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.86207389831543, "learning_rate": 1e-05, "loss": 0.4806, "mean_token_accuracy": 0.8809232711791992, "num_tokens": 859555954.0, "step": 22636 }, { "epoch": 2.8796590764533776, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 0.00010967254638671875, "grad_norm": 6.880490303039551, "learning_rate": 1e-05, "loss": 0.5142, "mean_token_accuracy": 0.8719930648803711, "num_tokens": 859591221.0, "step": 22637 }, { "epoch": 2.879786286731968, "ewc_loss": 0.125, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.3975749015808105, "learning_rate": 1e-05, "loss": 0.4237, "mean_token_accuracy": 0.8925564289093018, "num_tokens": 859631879.0, "step": 22638 }, { "epoch": 2.879913497010558, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 6.992495059967041, "learning_rate": 1e-05, "loss": 0.5255, "mean_token_accuracy": 0.8707126379013062, "num_tokens": 859669259.0, "step": 22639 }, { "epoch": 2.880040707289149, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.110645294189453, "learning_rate": 1e-05, "loss": 0.4449, "mean_token_accuracy": 0.8842616081237793, "num_tokens": 859706168.0, "step": 22640 }, { "epoch": 2.8801679175677393, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 0.00011205673217773438, "grad_norm": 7.21290397644043, "learning_rate": 1e-05, "loss": 0.481, "mean_token_accuracy": 0.8789572715759277, "num_tokens": 859743182.0, "step": 22641 }, { "epoch": 2.8802951278463302, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 5.980557441711426, "learning_rate": 1e-05, "loss": 0.4649, "mean_token_accuracy": 0.8798736929893494, "num_tokens": 859786120.0, "step": 22642 }, { "epoch": 2.8804223381249203, "ewc_loss": 0.1494140625, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 0.00011968612670898438, "grad_norm": 7.532014846801758, "learning_rate": 1e-05, "loss": 0.4851, "mean_token_accuracy": 0.8832972049713135, "num_tokens": 859824097.0, "step": 22643 }, { "epoch": 2.880549548403511, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.077580451965332, "learning_rate": 1e-05, "loss": 0.5069, "mean_token_accuracy": 0.8661216497421265, "num_tokens": 859860081.0, "step": 22644 }, { "epoch": 2.8806767586821014, "ewc_loss": 0.15625, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 0.000125885009765625, "grad_norm": 7.769558429718018, "learning_rate": 1e-05, "loss": 0.5248, "mean_token_accuracy": 0.8708321452140808, "num_tokens": 859898731.0, "step": 22645 }, { "epoch": 2.880803968960692, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.379728317260742, "learning_rate": 1e-05, "loss": 0.4818, "mean_token_accuracy": 0.878913164138794, "num_tokens": 859938406.0, "step": 22646 }, { "epoch": 2.8809311792392824, "ewc_loss": 0.154296875, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 0.00012493133544921875, "grad_norm": 7.720193862915039, "learning_rate": 1e-05, "loss": 0.4633, "mean_token_accuracy": 0.8914496898651123, "num_tokens": 859974270.0, "step": 22647 }, { "epoch": 2.881058389517873, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 6.65877628326416, "learning_rate": 1e-05, "loss": 0.475, "mean_token_accuracy": 0.8814336061477661, "num_tokens": 860007261.0, "step": 22648 }, { "epoch": 2.8811855997964635, "ewc_loss": 0.1484375, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 0.000118255615234375, "grad_norm": 7.422528266906738, "learning_rate": 1e-05, "loss": 0.4988, "mean_token_accuracy": 0.8767103552818298, "num_tokens": 860051170.0, "step": 22649 }, { "epoch": 2.881312810075054, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 6.71100378036499, "learning_rate": 1e-05, "loss": 0.4883, "mean_token_accuracy": 0.8810441493988037, "num_tokens": 860082771.0, "step": 22650 }, { "epoch": 2.8814400203536445, "ewc_loss": 0.140625, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 7.073516845703125, "learning_rate": 1e-05, "loss": 0.5403, "mean_token_accuracy": 0.8631730675697327, "num_tokens": 860124184.0, "step": 22651 }, { "epoch": 2.881567230632235, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.654452800750732, "learning_rate": 1e-05, "loss": 0.5212, "mean_token_accuracy": 0.8643361330032349, "num_tokens": 860158453.0, "step": 22652 }, { "epoch": 2.8816944409108256, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.806652545928955, "learning_rate": 1e-05, "loss": 0.4648, "mean_token_accuracy": 0.8851633667945862, "num_tokens": 860195403.0, "step": 22653 }, { "epoch": 2.881821651189416, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.547226428985596, "learning_rate": 1e-05, "loss": 0.4995, "mean_token_accuracy": 0.8689025044441223, "num_tokens": 860232813.0, "step": 22654 }, { "epoch": 2.8819488614680067, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.663760185241699, "learning_rate": 1e-05, "loss": 0.4953, "mean_token_accuracy": 0.8749043345451355, "num_tokens": 860269542.0, "step": 22655 }, { "epoch": 2.882076071746597, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.433988094329834, "learning_rate": 1e-05, "loss": 0.4994, "mean_token_accuracy": 0.8709995746612549, "num_tokens": 860305386.0, "step": 22656 }, { "epoch": 2.8822032820251877, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.6290154457092285, "learning_rate": 1e-05, "loss": 0.4849, "mean_token_accuracy": 0.8768764734268188, "num_tokens": 860343137.0, "step": 22657 }, { "epoch": 2.8823304923037782, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.311441898345947, "learning_rate": 1e-05, "loss": 0.4908, "mean_token_accuracy": 0.871735155582428, "num_tokens": 860382968.0, "step": 22658 }, { "epoch": 2.8824577025823688, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.677659034729004, "learning_rate": 1e-05, "loss": 0.5053, "mean_token_accuracy": 0.8711373805999756, "num_tokens": 860426985.0, "step": 22659 }, { "epoch": 2.8825849128609593, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.147495746612549, "learning_rate": 1e-05, "loss": 0.5065, "mean_token_accuracy": 0.8663483262062073, "num_tokens": 860460950.0, "step": 22660 }, { "epoch": 2.88271212313955, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.839935302734375, "learning_rate": 1e-05, "loss": 0.4642, "mean_token_accuracy": 0.8840344548225403, "num_tokens": 860499328.0, "step": 22661 }, { "epoch": 2.88283933341814, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 6.004754066467285, "learning_rate": 1e-05, "loss": 0.478, "mean_token_accuracy": 0.8775891065597534, "num_tokens": 860532411.0, "step": 22662 }, { "epoch": 2.882966543696731, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 0.00010967254638671875, "grad_norm": 7.107266426086426, "learning_rate": 1e-05, "loss": 0.4678, "mean_token_accuracy": 0.8876489400863647, "num_tokens": 860567565.0, "step": 22663 }, { "epoch": 2.883093753975321, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.879766464233398, "learning_rate": 1e-05, "loss": 0.4885, "mean_token_accuracy": 0.8708549737930298, "num_tokens": 860611687.0, "step": 22664 }, { "epoch": 2.883220964253912, "ewc_loss": 0.150390625, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 0.00012063980102539062, "grad_norm": 7.515442848205566, "learning_rate": 1e-05, "loss": 0.5549, "mean_token_accuracy": 0.861664891242981, "num_tokens": 860652220.0, "step": 22665 }, { "epoch": 2.883348174532502, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 5.965388298034668, "learning_rate": 1e-05, "loss": 0.4545, "mean_token_accuracy": 0.8842041492462158, "num_tokens": 860689641.0, "step": 22666 }, { "epoch": 2.883475384811093, "ewc_loss": 0.16015625, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 0.00013065338134765625, "grad_norm": 7.894656658172607, "learning_rate": 1e-05, "loss": 0.5104, "mean_token_accuracy": 0.8808910846710205, "num_tokens": 860727490.0, "step": 22667 }, { "epoch": 2.883602595089683, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.405210494995117, "learning_rate": 1e-05, "loss": 0.4866, "mean_token_accuracy": 0.8732291460037231, "num_tokens": 860768625.0, "step": 22668 }, { "epoch": 2.8837298053682736, "ewc_loss": 0.158203125, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 0.0001277923583984375, "grad_norm": 7.80047082901001, "learning_rate": 1e-05, "loss": 0.4907, "mean_token_accuracy": 0.8841755390167236, "num_tokens": 860806294.0, "step": 22669 }, { "epoch": 2.883857015646864, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 0.00011491775512695312, "grad_norm": 6.795628070831299, "learning_rate": 1e-05, "loss": 0.5434, "mean_token_accuracy": 0.8617558479309082, "num_tokens": 860842672.0, "step": 22670 }, { "epoch": 2.8839842259254547, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 0.00011682510375976562, "grad_norm": 7.34438943862915, "learning_rate": 1e-05, "loss": 0.5152, "mean_token_accuracy": 0.8692301511764526, "num_tokens": 860879882.0, "step": 22671 }, { "epoch": 2.884111436204045, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 0.00011301040649414062, "grad_norm": 6.866870880126953, "learning_rate": 1e-05, "loss": 0.48, "mean_token_accuracy": 0.8781089782714844, "num_tokens": 860916099.0, "step": 22672 }, { "epoch": 2.8842386464826357, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.8138041496276855, "learning_rate": 1e-05, "loss": 0.4851, "mean_token_accuracy": 0.8765873908996582, "num_tokens": 860957982.0, "step": 22673 }, { "epoch": 2.8843658567612263, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 6.877169609069824, "learning_rate": 1e-05, "loss": 0.4664, "mean_token_accuracy": 0.8859579563140869, "num_tokens": 860995528.0, "step": 22674 }, { "epoch": 2.884493067039817, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.384558200836182, "learning_rate": 1e-05, "loss": 0.4873, "mean_token_accuracy": 0.8748085498809814, "num_tokens": 861034676.0, "step": 22675 }, { "epoch": 2.8846202773184073, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 6.980161666870117, "learning_rate": 1e-05, "loss": 0.5198, "mean_token_accuracy": 0.8684899806976318, "num_tokens": 861069767.0, "step": 22676 }, { "epoch": 2.884747487596998, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.117151260375977, "learning_rate": 1e-05, "loss": 0.4611, "mean_token_accuracy": 0.8791568875312805, "num_tokens": 861109787.0, "step": 22677 }, { "epoch": 2.8848746978755884, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 0.00011205673217773438, "grad_norm": 7.180293083190918, "learning_rate": 1e-05, "loss": 0.48, "mean_token_accuracy": 0.8805925846099854, "num_tokens": 861144603.0, "step": 22678 }, { "epoch": 2.885001908154179, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 5.996214389801025, "learning_rate": 1e-05, "loss": 0.4505, "mean_token_accuracy": 0.8838025331497192, "num_tokens": 861184750.0, "step": 22679 }, { "epoch": 2.8851291184327694, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 0.00011730194091796875, "grad_norm": 7.449202537536621, "learning_rate": 1e-05, "loss": 0.504, "mean_token_accuracy": 0.8752731680870056, "num_tokens": 861223591.0, "step": 22680 }, { "epoch": 2.88525632871136, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.075920104980469, "learning_rate": 1e-05, "loss": 0.4805, "mean_token_accuracy": 0.8729861974716187, "num_tokens": 861260456.0, "step": 22681 }, { "epoch": 2.8853835389899505, "ewc_loss": 0.154296875, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 0.0001239776611328125, "grad_norm": 7.647481441497803, "learning_rate": 1e-05, "loss": 0.5102, "mean_token_accuracy": 0.8759787082672119, "num_tokens": 861298266.0, "step": 22682 }, { "epoch": 2.885510749268541, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.311493396759033, "learning_rate": 1e-05, "loss": 0.4676, "mean_token_accuracy": 0.881909966468811, "num_tokens": 861334646.0, "step": 22683 }, { "epoch": 2.8856379595471315, "ewc_loss": 0.15234375, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.632162570953369, "learning_rate": 1e-05, "loss": 0.5249, "mean_token_accuracy": 0.8689972162246704, "num_tokens": 861370734.0, "step": 22684 }, { "epoch": 2.885765169825722, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.539800643920898, "learning_rate": 1e-05, "loss": 0.4386, "mean_token_accuracy": 0.8939655423164368, "num_tokens": 861403111.0, "step": 22685 }, { "epoch": 2.8858923801043126, "ewc_loss": 0.1474609375, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 0.00011777877807617188, "grad_norm": 7.3951616287231445, "learning_rate": 1e-05, "loss": 0.4836, "mean_token_accuracy": 0.8792095184326172, "num_tokens": 861440668.0, "step": 22686 }, { "epoch": 2.8860195903829027, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.610212326049805, "learning_rate": 1e-05, "loss": 0.5252, "mean_token_accuracy": 0.8675268292427063, "num_tokens": 861478586.0, "step": 22687 }, { "epoch": 2.8861468006614936, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 0.00011205673217773438, "grad_norm": 7.09132194519043, "learning_rate": 1e-05, "loss": 0.4837, "mean_token_accuracy": 0.87569260597229, "num_tokens": 861511637.0, "step": 22688 }, { "epoch": 2.8862740109400837, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.595261096954346, "learning_rate": 1e-05, "loss": 0.4861, "mean_token_accuracy": 0.8760300874710083, "num_tokens": 861545421.0, "step": 22689 }, { "epoch": 2.8864012212186747, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.8158650398254395, "learning_rate": 1e-05, "loss": 0.5166, "mean_token_accuracy": 0.872011125087738, "num_tokens": 861586230.0, "step": 22690 }, { "epoch": 2.886528431497265, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.515719890594482, "learning_rate": 1e-05, "loss": 0.4591, "mean_token_accuracy": 0.8856594562530518, "num_tokens": 861620556.0, "step": 22691 }, { "epoch": 2.8866556417758558, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.669270992279053, "learning_rate": 1e-05, "loss": 0.4831, "mean_token_accuracy": 0.8743312954902649, "num_tokens": 861655826.0, "step": 22692 }, { "epoch": 2.886782852054446, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.405075550079346, "learning_rate": 1e-05, "loss": 0.4613, "mean_token_accuracy": 0.8812676072120667, "num_tokens": 861699160.0, "step": 22693 }, { "epoch": 2.8869100623330364, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.628624439239502, "learning_rate": 1e-05, "loss": 0.5256, "mean_token_accuracy": 0.8692901134490967, "num_tokens": 861736784.0, "step": 22694 }, { "epoch": 2.887037272611627, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.296962738037109, "learning_rate": 1e-05, "loss": 0.4435, "mean_token_accuracy": 0.8878073692321777, "num_tokens": 861775359.0, "step": 22695 }, { "epoch": 2.8871644828902174, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.664266109466553, "learning_rate": 1e-05, "loss": 0.5315, "mean_token_accuracy": 0.8643923997879028, "num_tokens": 861813852.0, "step": 22696 }, { "epoch": 2.887291693168808, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.159387588500977, "learning_rate": 1e-05, "loss": 0.4827, "mean_token_accuracy": 0.8756355047225952, "num_tokens": 861853730.0, "step": 22697 }, { "epoch": 2.8874189034473985, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.797389030456543, "learning_rate": 1e-05, "loss": 0.4851, "mean_token_accuracy": 0.8744258880615234, "num_tokens": 861891405.0, "step": 22698 }, { "epoch": 2.887546113725989, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.002959251403809, "learning_rate": 1e-05, "loss": 0.437, "mean_token_accuracy": 0.8870620131492615, "num_tokens": 861929595.0, "step": 22699 }, { "epoch": 2.8876733240045795, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 7.060781002044678, "learning_rate": 1e-05, "loss": 0.5211, "mean_token_accuracy": 0.8648871183395386, "num_tokens": 861963263.0, "step": 22700 }, { "epoch": 2.88780053428317, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.893918037414551, "learning_rate": 1e-05, "loss": 0.5043, "mean_token_accuracy": 0.868927001953125, "num_tokens": 862002259.0, "step": 22701 }, { "epoch": 2.8879277445617606, "ewc_loss": 0.1484375, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 0.00011920928955078125, "grad_norm": 7.443858623504639, "learning_rate": 1e-05, "loss": 0.561, "mean_token_accuracy": 0.8604360222816467, "num_tokens": 862039869.0, "step": 22702 }, { "epoch": 2.888054954840351, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 5.937839031219482, "learning_rate": 1e-05, "loss": 0.4695, "mean_token_accuracy": 0.8775032758712769, "num_tokens": 862068979.0, "step": 22703 }, { "epoch": 2.8881821651189417, "ewc_loss": 0.158203125, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 0.00012874603271484375, "grad_norm": 7.7974114418029785, "learning_rate": 1e-05, "loss": 0.5532, "mean_token_accuracy": 0.8628631830215454, "num_tokens": 862117384.0, "step": 22704 }, { "epoch": 2.888309375397532, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.292078495025635, "learning_rate": 1e-05, "loss": 0.5187, "mean_token_accuracy": 0.8667464256286621, "num_tokens": 862157961.0, "step": 22705 }, { "epoch": 2.8884365856761227, "ewc_loss": 0.158203125, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 0.00012874603271484375, "grad_norm": 7.807251930236816, "learning_rate": 1e-05, "loss": 0.5082, "mean_token_accuracy": 0.8790880441665649, "num_tokens": 862199023.0, "step": 22706 }, { "epoch": 2.8885637959547132, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 0.0001125335693359375, "grad_norm": 6.702648162841797, "learning_rate": 1e-05, "loss": 0.4928, "mean_token_accuracy": 0.8778713941574097, "num_tokens": 862237839.0, "step": 22707 }, { "epoch": 2.8886910062333038, "ewc_loss": 0.1484375, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 0.00011873245239257812, "grad_norm": 7.413371562957764, "learning_rate": 1e-05, "loss": 0.5043, "mean_token_accuracy": 0.8726281523704529, "num_tokens": 862273060.0, "step": 22708 }, { "epoch": 2.8888182165118943, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 0.00011348724365234375, "grad_norm": 6.832047939300537, "learning_rate": 1e-05, "loss": 0.5213, "mean_token_accuracy": 0.8673285245895386, "num_tokens": 862316469.0, "step": 22709 }, { "epoch": 2.888945426790485, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.905568599700928, "learning_rate": 1e-05, "loss": 0.5153, "mean_token_accuracy": 0.8696900606155396, "num_tokens": 862359367.0, "step": 22710 }, { "epoch": 2.8890726370690754, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 6.845631122589111, "learning_rate": 1e-05, "loss": 0.4702, "mean_token_accuracy": 0.8829094171524048, "num_tokens": 862397113.0, "step": 22711 }, { "epoch": 2.8891998473476654, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.462353229522705, "learning_rate": 1e-05, "loss": 0.5014, "mean_token_accuracy": 0.8715877532958984, "num_tokens": 862435627.0, "step": 22712 }, { "epoch": 2.8893270576262564, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 6.9302849769592285, "learning_rate": 1e-05, "loss": 0.4861, "mean_token_accuracy": 0.877467691898346, "num_tokens": 862476948.0, "step": 22713 }, { "epoch": 2.8894542679048465, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.155752658843994, "learning_rate": 1e-05, "loss": 0.4848, "mean_token_accuracy": 0.8754738569259644, "num_tokens": 862513429.0, "step": 22714 }, { "epoch": 2.8895814781834375, "ewc_loss": 0.140625, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 7.126341342926025, "learning_rate": 1e-05, "loss": 0.5216, "mean_token_accuracy": 0.866224467754364, "num_tokens": 862548570.0, "step": 22715 }, { "epoch": 2.8897086884620276, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.008962154388428, "learning_rate": 1e-05, "loss": 0.48, "mean_token_accuracy": 0.8759029507637024, "num_tokens": 862586265.0, "step": 22716 }, { "epoch": 2.889835898740618, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 0.00011730194091796875, "grad_norm": 7.395412921905518, "learning_rate": 1e-05, "loss": 0.5405, "mean_token_accuracy": 0.8625645637512207, "num_tokens": 862620552.0, "step": 22717 }, { "epoch": 2.8899631090192086, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.025402069091797, "learning_rate": 1e-05, "loss": 0.4364, "mean_token_accuracy": 0.8893176913261414, "num_tokens": 862656449.0, "step": 22718 }, { "epoch": 2.890090319297799, "ewc_loss": 0.15234375, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.623203277587891, "learning_rate": 1e-05, "loss": 0.5292, "mean_token_accuracy": 0.8692408800125122, "num_tokens": 862697612.0, "step": 22719 }, { "epoch": 2.8902175295763897, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.245553493499756, "learning_rate": 1e-05, "loss": 0.5176, "mean_token_accuracy": 0.8662295341491699, "num_tokens": 862742304.0, "step": 22720 }, { "epoch": 2.89034473985498, "ewc_loss": 0.154296875, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 0.0001239776611328125, "grad_norm": 7.659346103668213, "learning_rate": 1e-05, "loss": 0.5058, "mean_token_accuracy": 0.8793261647224426, "num_tokens": 862781738.0, "step": 22721 }, { "epoch": 2.8904719501335707, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.506182670593262, "learning_rate": 1e-05, "loss": 0.5632, "mean_token_accuracy": 0.8486833572387695, "num_tokens": 862823124.0, "step": 22722 }, { "epoch": 2.8905991604121613, "ewc_loss": 0.150390625, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 0.00012063980102539062, "grad_norm": 7.449806213378906, "learning_rate": 1e-05, "loss": 0.5442, "mean_token_accuracy": 0.8644521236419678, "num_tokens": 862857706.0, "step": 22723 }, { "epoch": 2.890726370690752, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.621514797210693, "learning_rate": 1e-05, "loss": 0.5429, "mean_token_accuracy": 0.8590983152389526, "num_tokens": 862899966.0, "step": 22724 }, { "epoch": 2.8908535809693423, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 0.00011205673217773438, "grad_norm": 7.128452777862549, "learning_rate": 1e-05, "loss": 0.4657, "mean_token_accuracy": 0.8775202631950378, "num_tokens": 862931090.0, "step": 22725 }, { "epoch": 2.890980791247933, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.605045318603516, "learning_rate": 1e-05, "loss": 0.4516, "mean_token_accuracy": 0.882712721824646, "num_tokens": 862962833.0, "step": 22726 }, { "epoch": 2.8911080015265234, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.855309009552002, "learning_rate": 1e-05, "loss": 0.4998, "mean_token_accuracy": 0.8715585470199585, "num_tokens": 863003769.0, "step": 22727 }, { "epoch": 2.891235211805114, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.527952671051025, "learning_rate": 1e-05, "loss": 0.4839, "mean_token_accuracy": 0.8770596981048584, "num_tokens": 863038130.0, "step": 22728 }, { "epoch": 2.8913624220837044, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.676912784576416, "learning_rate": 1e-05, "loss": 0.4427, "mean_token_accuracy": 0.890680193901062, "num_tokens": 863077943.0, "step": 22729 }, { "epoch": 2.891489632362295, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.4322099685668945, "learning_rate": 1e-05, "loss": 0.4904, "mean_token_accuracy": 0.8730112314224243, "num_tokens": 863113853.0, "step": 22730 }, { "epoch": 2.8916168426408855, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.613317966461182, "learning_rate": 1e-05, "loss": 0.4699, "mean_token_accuracy": 0.8827992081642151, "num_tokens": 863149407.0, "step": 22731 }, { "epoch": 2.891744052919476, "ewc_loss": 0.125, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.314995765686035, "learning_rate": 1e-05, "loss": 0.4982, "mean_token_accuracy": 0.8721017837524414, "num_tokens": 863183324.0, "step": 22732 }, { "epoch": 2.8918712631980665, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.63955020904541, "learning_rate": 1e-05, "loss": 0.4385, "mean_token_accuracy": 0.8914060592651367, "num_tokens": 863220400.0, "step": 22733 }, { "epoch": 2.891998473476657, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.180325031280518, "learning_rate": 1e-05, "loss": 0.4548, "mean_token_accuracy": 0.88453209400177, "num_tokens": 863259565.0, "step": 22734 }, { "epoch": 2.8921256837552476, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.7756123542785645, "learning_rate": 1e-05, "loss": 0.5139, "mean_token_accuracy": 0.8696534037590027, "num_tokens": 863293480.0, "step": 22735 }, { "epoch": 2.892252894033838, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.033728122711182, "learning_rate": 1e-05, "loss": 0.4869, "mean_token_accuracy": 0.8760243058204651, "num_tokens": 863331702.0, "step": 22736 }, { "epoch": 2.892380104312428, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 7.007277488708496, "learning_rate": 1e-05, "loss": 0.5165, "mean_token_accuracy": 0.8711228966712952, "num_tokens": 863366195.0, "step": 22737 }, { "epoch": 2.892507314591019, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 5.8940253257751465, "learning_rate": 1e-05, "loss": 0.4873, "mean_token_accuracy": 0.8698898553848267, "num_tokens": 863406674.0, "step": 22738 }, { "epoch": 2.8926345248696093, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 0.00011682510375976562, "grad_norm": 7.376552581787109, "learning_rate": 1e-05, "loss": 0.51, "mean_token_accuracy": 0.8738963007926941, "num_tokens": 863441837.0, "step": 22739 }, { "epoch": 2.8927617351482002, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 5.899126052856445, "learning_rate": 1e-05, "loss": 0.4435, "mean_token_accuracy": 0.8837220668792725, "num_tokens": 863476461.0, "step": 22740 }, { "epoch": 2.8928889454267903, "ewc_loss": 0.158203125, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 0.00012874603271484375, "grad_norm": 7.770660400390625, "learning_rate": 1e-05, "loss": 0.5328, "mean_token_accuracy": 0.8676038384437561, "num_tokens": 863513715.0, "step": 22741 }, { "epoch": 2.893016155705381, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.21735143661499, "learning_rate": 1e-05, "loss": 0.4701, "mean_token_accuracy": 0.8810637593269348, "num_tokens": 863553172.0, "step": 22742 }, { "epoch": 2.8931433659839714, "ewc_loss": 0.1611328125, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 0.00013065338134765625, "grad_norm": 7.876142501831055, "learning_rate": 1e-05, "loss": 0.5529, "mean_token_accuracy": 0.8635634183883667, "num_tokens": 863585353.0, "step": 22743 }, { "epoch": 2.893270576262562, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 0.00011348724365234375, "grad_norm": 6.674988269805908, "learning_rate": 1e-05, "loss": 0.5257, "mean_token_accuracy": 0.8698980808258057, "num_tokens": 863627510.0, "step": 22744 }, { "epoch": 2.8933977865411524, "ewc_loss": 0.1513671875, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 0.00012159347534179688, "grad_norm": 7.503833770751953, "learning_rate": 1e-05, "loss": 0.5324, "mean_token_accuracy": 0.8689346313476562, "num_tokens": 863670683.0, "step": 22745 }, { "epoch": 2.893524996819743, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 0.00011491775512695312, "grad_norm": 6.8341875076293945, "learning_rate": 1e-05, "loss": 0.5136, "mean_token_accuracy": 0.8713728189468384, "num_tokens": 863710793.0, "step": 22746 }, { "epoch": 2.8936522070983335, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 0.00010967254638671875, "grad_norm": 7.003565788269043, "learning_rate": 1e-05, "loss": 0.4713, "mean_token_accuracy": 0.8830130696296692, "num_tokens": 863745963.0, "step": 22747 }, { "epoch": 2.893779417376924, "ewc_loss": 0.140625, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 6.840730667114258, "learning_rate": 1e-05, "loss": 0.4565, "mean_token_accuracy": 0.8887045979499817, "num_tokens": 863783729.0, "step": 22748 }, { "epoch": 2.8939066276555145, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.544084072113037, "learning_rate": 1e-05, "loss": 0.4745, "mean_token_accuracy": 0.876971960067749, "num_tokens": 863820070.0, "step": 22749 }, { "epoch": 2.894033837934105, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 6.901500225067139, "learning_rate": 1e-05, "loss": 0.4674, "mean_token_accuracy": 0.8804086446762085, "num_tokens": 863853438.0, "step": 22750 }, { "epoch": 2.8941610482126956, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.229340076446533, "learning_rate": 1e-05, "loss": 0.4724, "mean_token_accuracy": 0.8796368837356567, "num_tokens": 863894151.0, "step": 22751 }, { "epoch": 2.894288258491286, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 7.037652969360352, "learning_rate": 1e-05, "loss": 0.4742, "mean_token_accuracy": 0.8834383487701416, "num_tokens": 863929712.0, "step": 22752 }, { "epoch": 2.8944154687698767, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.034232139587402, "learning_rate": 1e-05, "loss": 0.5293, "mean_token_accuracy": 0.8585292100906372, "num_tokens": 863970236.0, "step": 22753 }, { "epoch": 2.894542679048467, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 0.00011396408081054688, "grad_norm": 7.270848751068115, "learning_rate": 1e-05, "loss": 0.5251, "mean_token_accuracy": 0.8659114837646484, "num_tokens": 864009164.0, "step": 22754 }, { "epoch": 2.8946698893270577, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.000120162963867, "learning_rate": 1e-05, "loss": 0.4826, "mean_token_accuracy": 0.8724241852760315, "num_tokens": 864048941.0, "step": 22755 }, { "epoch": 2.8947970996056482, "ewc_loss": 0.1513671875, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 0.00012111663818359375, "grad_norm": 7.537181854248047, "learning_rate": 1e-05, "loss": 0.5354, "mean_token_accuracy": 0.8709367513656616, "num_tokens": 864084041.0, "step": 22756 }, { "epoch": 2.8949243098842388, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.143430233001709, "learning_rate": 1e-05, "loss": 0.4582, "mean_token_accuracy": 0.8804436922073364, "num_tokens": 864118370.0, "step": 22757 }, { "epoch": 2.8950515201628293, "ewc_loss": 0.154296875, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 0.0001239776611328125, "grad_norm": 7.677606582641602, "learning_rate": 1e-05, "loss": 0.4651, "mean_token_accuracy": 0.889614462852478, "num_tokens": 864148888.0, "step": 22758 }, { "epoch": 2.89517873044142, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.389144420623779, "learning_rate": 1e-05, "loss": 0.4667, "mean_token_accuracy": 0.8814167380332947, "num_tokens": 864185850.0, "step": 22759 }, { "epoch": 2.89530594072001, "ewc_loss": 0.1533203125, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.593114376068115, "learning_rate": 1e-05, "loss": 0.557, "mean_token_accuracy": 0.8593788146972656, "num_tokens": 864224914.0, "step": 22760 }, { "epoch": 2.895433150998601, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 6.572070598602295, "learning_rate": 1e-05, "loss": 0.4975, "mean_token_accuracy": 0.8716925382614136, "num_tokens": 864266430.0, "step": 22761 }, { "epoch": 2.895560361277191, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 0.0001163482666015625, "grad_norm": 7.3238959312438965, "learning_rate": 1e-05, "loss": 0.4938, "mean_token_accuracy": 0.8733530640602112, "num_tokens": 864299769.0, "step": 22762 }, { "epoch": 2.895687571555782, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.629237651824951, "learning_rate": 1e-05, "loss": 0.5027, "mean_token_accuracy": 0.8705941438674927, "num_tokens": 864340713.0, "step": 22763 }, { "epoch": 2.895814781834372, "ewc_loss": 0.140625, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 7.01066255569458, "learning_rate": 1e-05, "loss": 0.5264, "mean_token_accuracy": 0.8665993213653564, "num_tokens": 864382930.0, "step": 22764 }, { "epoch": 2.895941992112963, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.590351581573486, "learning_rate": 1e-05, "loss": 0.46, "mean_token_accuracy": 0.8865261077880859, "num_tokens": 864424632.0, "step": 22765 }, { "epoch": 2.896069202391553, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.749776840209961, "learning_rate": 1e-05, "loss": 0.459, "mean_token_accuracy": 0.8877681493759155, "num_tokens": 864463444.0, "step": 22766 }, { "epoch": 2.8961964126701436, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.523107051849365, "learning_rate": 1e-05, "loss": 0.5466, "mean_token_accuracy": 0.8567968010902405, "num_tokens": 864506651.0, "step": 22767 }, { "epoch": 2.896323622948734, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.599431991577148, "learning_rate": 1e-05, "loss": 0.4656, "mean_token_accuracy": 0.8810185194015503, "num_tokens": 864544075.0, "step": 22768 }, { "epoch": 2.8964508332273247, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.435620307922363, "learning_rate": 1e-05, "loss": 0.4327, "mean_token_accuracy": 0.893699586391449, "num_tokens": 864579814.0, "step": 22769 }, { "epoch": 2.896578043505915, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.550135612487793, "learning_rate": 1e-05, "loss": 0.5404, "mean_token_accuracy": 0.8582490086555481, "num_tokens": 864616690.0, "step": 22770 }, { "epoch": 2.8967052537845057, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.3209333419799805, "learning_rate": 1e-05, "loss": 0.4629, "mean_token_accuracy": 0.8813388347625732, "num_tokens": 864662035.0, "step": 22771 }, { "epoch": 2.8968324640630962, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.590301513671875, "learning_rate": 1e-05, "loss": 0.5218, "mean_token_accuracy": 0.864189088344574, "num_tokens": 864695219.0, "step": 22772 }, { "epoch": 2.8969596743416868, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.19408655166626, "learning_rate": 1e-05, "loss": 0.4748, "mean_token_accuracy": 0.8790276050567627, "num_tokens": 864731990.0, "step": 22773 }, { "epoch": 2.8970868846202773, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.703455448150635, "learning_rate": 1e-05, "loss": 0.4696, "mean_token_accuracy": 0.8795537948608398, "num_tokens": 864769642.0, "step": 22774 }, { "epoch": 2.897214094898868, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.038786888122559, "learning_rate": 1e-05, "loss": 0.4916, "mean_token_accuracy": 0.8712209463119507, "num_tokens": 864805329.0, "step": 22775 }, { "epoch": 2.8973413051774584, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.9246625900268555, "learning_rate": 1e-05, "loss": 0.4649, "mean_token_accuracy": 0.8859595060348511, "num_tokens": 864847748.0, "step": 22776 }, { "epoch": 2.897468515456049, "ewc_loss": 0.11669921875, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.8758721351623535, "learning_rate": 1e-05, "loss": 0.507, "mean_token_accuracy": 0.8640753626823425, "num_tokens": 864884621.0, "step": 22777 }, { "epoch": 2.8975957257346394, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 0.00011491775512695312, "grad_norm": 7.310213565826416, "learning_rate": 1e-05, "loss": 0.4796, "mean_token_accuracy": 0.8847044706344604, "num_tokens": 864927398.0, "step": 22778 }, { "epoch": 2.89772293601323, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 5.817760944366455, "learning_rate": 1e-05, "loss": 0.4465, "mean_token_accuracy": 0.8871064186096191, "num_tokens": 864961038.0, "step": 22779 }, { "epoch": 2.8978501462918205, "ewc_loss": 0.1591796875, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 0.00012874603271484375, "grad_norm": 7.8126068115234375, "learning_rate": 1e-05, "loss": 0.5278, "mean_token_accuracy": 0.8716322183609009, "num_tokens": 865000307.0, "step": 22780 }, { "epoch": 2.897977356570411, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.1125006675720215, "learning_rate": 1e-05, "loss": 0.4442, "mean_token_accuracy": 0.8873485326766968, "num_tokens": 865035898.0, "step": 22781 }, { "epoch": 2.8981045668490015, "ewc_loss": 0.166015625, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 0.0001354217529296875, "grad_norm": 8.064446449279785, "learning_rate": 1e-05, "loss": 0.4616, "mean_token_accuracy": 0.8958826065063477, "num_tokens": 865076325.0, "step": 22782 }, { "epoch": 2.898231777127592, "ewc_loss": 0.1455078125, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 0.00011587142944335938, "grad_norm": 6.689629077911377, "learning_rate": 1e-05, "loss": 0.5143, "mean_token_accuracy": 0.8711760640144348, "num_tokens": 865117290.0, "step": 22783 }, { "epoch": 2.8983589874061826, "ewc_loss": 0.158203125, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 0.0001277923583984375, "grad_norm": 7.748895645141602, "learning_rate": 1e-05, "loss": 0.494, "mean_token_accuracy": 0.8828158974647522, "num_tokens": 865153003.0, "step": 22784 }, { "epoch": 2.8984861976847727, "ewc_loss": 0.1474609375, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 0.00011730194091796875, "grad_norm": 6.937935829162598, "learning_rate": 1e-05, "loss": 0.5133, "mean_token_accuracy": 0.8717592358589172, "num_tokens": 865195113.0, "step": 22785 }, { "epoch": 2.8986134079633636, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 0.00011491775512695312, "grad_norm": 7.2417755126953125, "learning_rate": 1e-05, "loss": 0.4759, "mean_token_accuracy": 0.883704423904419, "num_tokens": 865231319.0, "step": 22786 }, { "epoch": 2.8987406182419537, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 0.00011396408081054688, "grad_norm": 6.889825820922852, "learning_rate": 1e-05, "loss": 0.4858, "mean_token_accuracy": 0.8768705129623413, "num_tokens": 865265862.0, "step": 22787 }, { "epoch": 2.8988678285205447, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.788206577301025, "learning_rate": 1e-05, "loss": 0.5032, "mean_token_accuracy": 0.8730983734130859, "num_tokens": 865303873.0, "step": 22788 }, { "epoch": 2.898995038799135, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 6.842657089233398, "learning_rate": 1e-05, "loss": 0.4631, "mean_token_accuracy": 0.885568380355835, "num_tokens": 865335163.0, "step": 22789 }, { "epoch": 2.8991222490777258, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.46510648727417, "learning_rate": 1e-05, "loss": 0.4856, "mean_token_accuracy": 0.8761390447616577, "num_tokens": 865373498.0, "step": 22790 }, { "epoch": 2.899249459356316, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.8321452140808105, "learning_rate": 1e-05, "loss": 0.4839, "mean_token_accuracy": 0.8761633038520813, "num_tokens": 865411042.0, "step": 22791 }, { "epoch": 2.8993766696349064, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.261277675628662, "learning_rate": 1e-05, "loss": 0.4688, "mean_token_accuracy": 0.8789891600608826, "num_tokens": 865447676.0, "step": 22792 }, { "epoch": 2.899503879913497, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.873170375823975, "learning_rate": 1e-05, "loss": 0.5036, "mean_token_accuracy": 0.8716030120849609, "num_tokens": 865490905.0, "step": 22793 }, { "epoch": 2.8996310901920874, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.128318786621094, "learning_rate": 1e-05, "loss": 0.485, "mean_token_accuracy": 0.8790366649627686, "num_tokens": 865530572.0, "step": 22794 }, { "epoch": 2.899758300470678, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.991736888885498, "learning_rate": 1e-05, "loss": 0.4776, "mean_token_accuracy": 0.8798609972000122, "num_tokens": 865565999.0, "step": 22795 }, { "epoch": 2.8998855107492685, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 6.020657539367676, "learning_rate": 1e-05, "loss": 0.4851, "mean_token_accuracy": 0.8691183924674988, "num_tokens": 865602455.0, "step": 22796 }, { "epoch": 2.900012721027859, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 0.0001125335693359375, "grad_norm": 7.201615333557129, "learning_rate": 1e-05, "loss": 0.4348, "mean_token_accuracy": 0.8985233902931213, "num_tokens": 865637723.0, "step": 22797 }, { "epoch": 2.9001399313064495, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 5.9767608642578125, "learning_rate": 1e-05, "loss": 0.4927, "mean_token_accuracy": 0.8680166006088257, "num_tokens": 865669668.0, "step": 22798 }, { "epoch": 2.90026714158504, "ewc_loss": 0.1484375, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 0.000118255615234375, "grad_norm": 7.44121789932251, "learning_rate": 1e-05, "loss": 0.4999, "mean_token_accuracy": 0.8780813813209534, "num_tokens": 865708255.0, "step": 22799 }, { "epoch": 2.9003943518636306, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.033333778381348, "learning_rate": 1e-05, "loss": 0.5036, "mean_token_accuracy": 0.8662108182907104, "num_tokens": 865749607.0, "step": 22800 }, { "epoch": 2.900521562142221, "ewc_loss": 0.1552734375, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 0.00012493133544921875, "grad_norm": 7.6630401611328125, "learning_rate": 1e-05, "loss": 0.5405, "mean_token_accuracy": 0.8693684935569763, "num_tokens": 865786372.0, "step": 22801 }, { "epoch": 2.9006487724208116, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.280750751495361, "learning_rate": 1e-05, "loss": 0.4679, "mean_token_accuracy": 0.8808856010437012, "num_tokens": 865822939.0, "step": 22802 }, { "epoch": 2.900775982699402, "ewc_loss": 0.154296875, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 0.0001239776611328125, "grad_norm": 7.655101776123047, "learning_rate": 1e-05, "loss": 0.4691, "mean_token_accuracy": 0.8867985606193542, "num_tokens": 865859012.0, "step": 22803 }, { "epoch": 2.9009031929779927, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 0.00011110305786132812, "grad_norm": 6.543365478515625, "learning_rate": 1e-05, "loss": 0.475, "mean_token_accuracy": 0.8832289576530457, "num_tokens": 865897641.0, "step": 22804 }, { "epoch": 2.9010304032565832, "ewc_loss": 0.1484375, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 0.00011873245239257812, "grad_norm": 7.381684303283691, "learning_rate": 1e-05, "loss": 0.51, "mean_token_accuracy": 0.8739824891090393, "num_tokens": 865938270.0, "step": 22805 }, { "epoch": 2.9011576135351738, "ewc_loss": 0.140625, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 0.00011014938354492188, "grad_norm": 6.660346508026123, "learning_rate": 1e-05, "loss": 0.4763, "mean_token_accuracy": 0.8814092874526978, "num_tokens": 865971066.0, "step": 22806 }, { "epoch": 2.9012848238137643, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 0.00010967254638671875, "grad_norm": 7.010268211364746, "learning_rate": 1e-05, "loss": 0.4648, "mean_token_accuracy": 0.8867887258529663, "num_tokens": 866008595.0, "step": 22807 }, { "epoch": 2.901412034092355, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.662019729614258, "learning_rate": 1e-05, "loss": 0.4872, "mean_token_accuracy": 0.8805844783782959, "num_tokens": 866046650.0, "step": 22808 }, { "epoch": 2.9015392443709453, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.657905578613281, "learning_rate": 1e-05, "loss": 0.5085, "mean_token_accuracy": 0.8685159087181091, "num_tokens": 866086305.0, "step": 22809 }, { "epoch": 2.9016664546495354, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.645808219909668, "learning_rate": 1e-05, "loss": 0.4842, "mean_token_accuracy": 0.8756405711174011, "num_tokens": 866125770.0, "step": 22810 }, { "epoch": 2.9017936649281264, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.408955097198486, "learning_rate": 1e-05, "loss": 0.4542, "mean_token_accuracy": 0.8860288858413696, "num_tokens": 866160777.0, "step": 22811 }, { "epoch": 2.9019208752067165, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.655943870544434, "learning_rate": 1e-05, "loss": 0.4541, "mean_token_accuracy": 0.8861349821090698, "num_tokens": 866195212.0, "step": 22812 }, { "epoch": 2.9020480854853075, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.245067596435547, "learning_rate": 1e-05, "loss": 0.4552, "mean_token_accuracy": 0.8832641839981079, "num_tokens": 866236252.0, "step": 22813 }, { "epoch": 2.9021752957638975, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.689124584197998, "learning_rate": 1e-05, "loss": 0.4496, "mean_token_accuracy": 0.8875402212142944, "num_tokens": 866275705.0, "step": 22814 }, { "epoch": 2.902302506042488, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.125138282775879, "learning_rate": 1e-05, "loss": 0.4866, "mean_token_accuracy": 0.8751869201660156, "num_tokens": 866316458.0, "step": 22815 }, { "epoch": 2.9024297163210786, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.789351463317871, "learning_rate": 1e-05, "loss": 0.5037, "mean_token_accuracy": 0.869707465171814, "num_tokens": 866351107.0, "step": 22816 }, { "epoch": 2.902556926599669, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 5.998468399047852, "learning_rate": 1e-05, "loss": 0.4259, "mean_token_accuracy": 0.8938220739364624, "num_tokens": 866385381.0, "step": 22817 }, { "epoch": 2.9026841368782597, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 6.993197441101074, "learning_rate": 1e-05, "loss": 0.4656, "mean_token_accuracy": 0.8848381042480469, "num_tokens": 866427018.0, "step": 22818 }, { "epoch": 2.90281134715685, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 5.859979629516602, "learning_rate": 1e-05, "loss": 0.4208, "mean_token_accuracy": 0.891942024230957, "num_tokens": 866463973.0, "step": 22819 }, { "epoch": 2.9029385574354407, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 0.00011682510375976562, "grad_norm": 7.346765995025635, "learning_rate": 1e-05, "loss": 0.5124, "mean_token_accuracy": 0.8738726377487183, "num_tokens": 866500900.0, "step": 22820 }, { "epoch": 2.9030657677140312, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 5.8321638107299805, "learning_rate": 1e-05, "loss": 0.4257, "mean_token_accuracy": 0.8929414749145508, "num_tokens": 866539757.0, "step": 22821 }, { "epoch": 2.9031929779926218, "ewc_loss": 0.1611328125, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 0.00013065338134765625, "grad_norm": 7.798634052276611, "learning_rate": 1e-05, "loss": 0.512, "mean_token_accuracy": 0.8764734268188477, "num_tokens": 866577454.0, "step": 22822 }, { "epoch": 2.9033201882712123, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.0997161865234375, "learning_rate": 1e-05, "loss": 0.4449, "mean_token_accuracy": 0.8887077569961548, "num_tokens": 866615032.0, "step": 22823 }, { "epoch": 2.903447398549803, "ewc_loss": 0.166015625, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 0.0001354217529296875, "grad_norm": 8.055646896362305, "learning_rate": 1e-05, "loss": 0.5309, "mean_token_accuracy": 0.8707435131072998, "num_tokens": 866649700.0, "step": 22824 }, { "epoch": 2.9035746088283934, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 0.00011491775512695312, "grad_norm": 6.642449855804443, "learning_rate": 1e-05, "loss": 0.5056, "mean_token_accuracy": 0.8740154504776001, "num_tokens": 866684720.0, "step": 22825 }, { "epoch": 2.903701819106984, "ewc_loss": 0.1591796875, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 0.00012874603271484375, "grad_norm": 7.796180248260498, "learning_rate": 1e-05, "loss": 0.5206, "mean_token_accuracy": 0.8766605854034424, "num_tokens": 866722127.0, "step": 22826 }, { "epoch": 2.9038290293855744, "ewc_loss": 0.1484375, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 0.000118255615234375, "grad_norm": 6.89872407913208, "learning_rate": 1e-05, "loss": 0.4944, "mean_token_accuracy": 0.8796820640563965, "num_tokens": 866757093.0, "step": 22827 }, { "epoch": 2.903956239664165, "ewc_loss": 0.1474609375, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 0.00011730194091796875, "grad_norm": 7.328436851501465, "learning_rate": 1e-05, "loss": 0.5308, "mean_token_accuracy": 0.8696650266647339, "num_tokens": 866796229.0, "step": 22828 }, { "epoch": 2.9040834499427555, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 0.00011444091796875, "grad_norm": 6.8552141189575195, "learning_rate": 1e-05, "loss": 0.4844, "mean_token_accuracy": 0.8800999522209167, "num_tokens": 866831867.0, "step": 22829 }, { "epoch": 2.904210660221346, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.904850006103516, "learning_rate": 1e-05, "loss": 0.5123, "mean_token_accuracy": 0.8675233125686646, "num_tokens": 866876246.0, "step": 22830 }, { "epoch": 2.9043378704999365, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.790426254272461, "learning_rate": 1e-05, "loss": 0.4958, "mean_token_accuracy": 0.8768424391746521, "num_tokens": 866909555.0, "step": 22831 }, { "epoch": 2.904465080778527, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.573779582977295, "learning_rate": 1e-05, "loss": 0.4833, "mean_token_accuracy": 0.8740720748901367, "num_tokens": 866945208.0, "step": 22832 }, { "epoch": 2.9045922910571176, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.761203289031982, "learning_rate": 1e-05, "loss": 0.4769, "mean_token_accuracy": 0.8812308311462402, "num_tokens": 866976832.0, "step": 22833 }, { "epoch": 2.904719501335708, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.352902412414551, "learning_rate": 1e-05, "loss": 0.4897, "mean_token_accuracy": 0.8731470108032227, "num_tokens": 867013964.0, "step": 22834 }, { "epoch": 2.904846711614298, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.748579978942871, "learning_rate": 1e-05, "loss": 0.4932, "mean_token_accuracy": 0.8741230964660645, "num_tokens": 867054361.0, "step": 22835 }, { "epoch": 2.904973921892889, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.194289684295654, "learning_rate": 1e-05, "loss": 0.421, "mean_token_accuracy": 0.8936355113983154, "num_tokens": 867098732.0, "step": 22836 }, { "epoch": 2.9051011321714793, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.807183265686035, "learning_rate": 1e-05, "loss": 0.5319, "mean_token_accuracy": 0.8615185022354126, "num_tokens": 867133839.0, "step": 22837 }, { "epoch": 2.9052283424500702, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.062455177307129, "learning_rate": 1e-05, "loss": 0.5392, "mean_token_accuracy": 0.8541982173919678, "num_tokens": 867175419.0, "step": 22838 }, { "epoch": 2.9053555527286603, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 6.981895446777344, "learning_rate": 1e-05, "loss": 0.5092, "mean_token_accuracy": 0.8714600801467896, "num_tokens": 867216585.0, "step": 22839 }, { "epoch": 2.905482763007251, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 5.911555290222168, "learning_rate": 1e-05, "loss": 0.4529, "mean_token_accuracy": 0.881999135017395, "num_tokens": 867257228.0, "step": 22840 }, { "epoch": 2.9056099732858414, "ewc_loss": 0.1455078125, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 0.00011587142944335938, "grad_norm": 7.308465480804443, "learning_rate": 1e-05, "loss": 0.5094, "mean_token_accuracy": 0.8756470084190369, "num_tokens": 867297504.0, "step": 22841 }, { "epoch": 2.905737183564432, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 5.857616424560547, "learning_rate": 1e-05, "loss": 0.4735, "mean_token_accuracy": 0.8793895244598389, "num_tokens": 867336799.0, "step": 22842 }, { "epoch": 2.9058643938430224, "ewc_loss": 0.158203125, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 0.0001277923583984375, "grad_norm": 7.745248794555664, "learning_rate": 1e-05, "loss": 0.5107, "mean_token_accuracy": 0.8773000240325928, "num_tokens": 867371337.0, "step": 22843 }, { "epoch": 2.905991604121613, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.07885217666626, "learning_rate": 1e-05, "loss": 0.4734, "mean_token_accuracy": 0.8797001838684082, "num_tokens": 867411870.0, "step": 22844 }, { "epoch": 2.9061188144002035, "ewc_loss": 0.166015625, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 0.0001354217529296875, "grad_norm": 8.01414966583252, "learning_rate": 1e-05, "loss": 0.5502, "mean_token_accuracy": 0.8652503490447998, "num_tokens": 867450278.0, "step": 22845 }, { "epoch": 2.906246024678794, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 0.00011348724365234375, "grad_norm": 6.581247329711914, "learning_rate": 1e-05, "loss": 0.5081, "mean_token_accuracy": 0.8703291416168213, "num_tokens": 867490316.0, "step": 22846 }, { "epoch": 2.9063732349573845, "ewc_loss": 0.1591796875, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 0.00012874603271484375, "grad_norm": 7.798818588256836, "learning_rate": 1e-05, "loss": 0.4851, "mean_token_accuracy": 0.8847204446792603, "num_tokens": 867526243.0, "step": 22847 }, { "epoch": 2.906500445235975, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 0.00011682510375976562, "grad_norm": 6.878334045410156, "learning_rate": 1e-05, "loss": 0.5132, "mean_token_accuracy": 0.8747760057449341, "num_tokens": 867568757.0, "step": 22848 }, { "epoch": 2.9066276555145656, "ewc_loss": 0.1474609375, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 0.00011730194091796875, "grad_norm": 7.334237575531006, "learning_rate": 1e-05, "loss": 0.481, "mean_token_accuracy": 0.8817325234413147, "num_tokens": 867602164.0, "step": 22849 }, { "epoch": 2.906754865793156, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 0.00011348724365234375, "grad_norm": 6.860663890838623, "learning_rate": 1e-05, "loss": 0.5125, "mean_token_accuracy": 0.8706651926040649, "num_tokens": 867640722.0, "step": 22850 }, { "epoch": 2.9068820760717466, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.884993553161621, "learning_rate": 1e-05, "loss": 0.4855, "mean_token_accuracy": 0.8794348239898682, "num_tokens": 867679568.0, "step": 22851 }, { "epoch": 2.907009286350337, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.804471492767334, "learning_rate": 1e-05, "loss": 0.5052, "mean_token_accuracy": 0.8707035183906555, "num_tokens": 867714991.0, "step": 22852 }, { "epoch": 2.9071364966289277, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.539929389953613, "learning_rate": 1e-05, "loss": 0.4943, "mean_token_accuracy": 0.8717960119247437, "num_tokens": 867753180.0, "step": 22853 }, { "epoch": 2.9072637069075182, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.7695088386535645, "learning_rate": 1e-05, "loss": 0.4766, "mean_token_accuracy": 0.8774369359016418, "num_tokens": 867789535.0, "step": 22854 }, { "epoch": 2.9073909171861088, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.318511009216309, "learning_rate": 1e-05, "loss": 0.4579, "mean_token_accuracy": 0.8804203867912292, "num_tokens": 867821830.0, "step": 22855 }, { "epoch": 2.9075181274646993, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.771188735961914, "learning_rate": 1e-05, "loss": 0.4734, "mean_token_accuracy": 0.8824996948242188, "num_tokens": 867859878.0, "step": 22856 }, { "epoch": 2.90764533774329, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.178429126739502, "learning_rate": 1e-05, "loss": 0.485, "mean_token_accuracy": 0.8716506361961365, "num_tokens": 867897212.0, "step": 22857 }, { "epoch": 2.90777254802188, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.843950271606445, "learning_rate": 1e-05, "loss": 0.5213, "mean_token_accuracy": 0.8653538227081299, "num_tokens": 867929899.0, "step": 22858 }, { "epoch": 2.907899758300471, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.045773029327393, "learning_rate": 1e-05, "loss": 0.443, "mean_token_accuracy": 0.8850358128547668, "num_tokens": 867962313.0, "step": 22859 }, { "epoch": 2.908026968579061, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 7.026238918304443, "learning_rate": 1e-05, "loss": 0.5681, "mean_token_accuracy": 0.854730486869812, "num_tokens": 867998886.0, "step": 22860 }, { "epoch": 2.908154178857652, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 5.911440372467041, "learning_rate": 1e-05, "loss": 0.4694, "mean_token_accuracy": 0.8769630193710327, "num_tokens": 868039151.0, "step": 22861 }, { "epoch": 2.908281389136242, "ewc_loss": 0.1455078125, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 0.00011587142944335938, "grad_norm": 7.351701736450195, "learning_rate": 1e-05, "loss": 0.4869, "mean_token_accuracy": 0.8803870677947998, "num_tokens": 868071021.0, "step": 22862 }, { "epoch": 2.908408599414833, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 5.876281261444092, "learning_rate": 1e-05, "loss": 0.4655, "mean_token_accuracy": 0.8807957768440247, "num_tokens": 868109872.0, "step": 22863 }, { "epoch": 2.908535809693423, "ewc_loss": 0.158203125, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 0.0001277923583984375, "grad_norm": 7.7592339515686035, "learning_rate": 1e-05, "loss": 0.5104, "mean_token_accuracy": 0.8781862258911133, "num_tokens": 868150258.0, "step": 22864 }, { "epoch": 2.9086630199720136, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.12798547744751, "learning_rate": 1e-05, "loss": 0.5078, "mean_token_accuracy": 0.8658537268638611, "num_tokens": 868195202.0, "step": 22865 }, { "epoch": 2.908790230250604, "ewc_loss": 0.1640625, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 0.000133514404296875, "grad_norm": 7.976306915283203, "learning_rate": 1e-05, "loss": 0.4897, "mean_token_accuracy": 0.8877710103988647, "num_tokens": 868226458.0, "step": 22866 }, { "epoch": 2.9089174405291947, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 0.00011348724365234375, "grad_norm": 6.611538410186768, "learning_rate": 1e-05, "loss": 0.496, "mean_token_accuracy": 0.8752690553665161, "num_tokens": 868262908.0, "step": 22867 }, { "epoch": 2.909044650807785, "ewc_loss": 0.158203125, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 0.0001277923583984375, "grad_norm": 7.734258651733398, "learning_rate": 1e-05, "loss": 0.5369, "mean_token_accuracy": 0.8675105571746826, "num_tokens": 868300013.0, "step": 22868 }, { "epoch": 2.9091718610863757, "ewc_loss": 0.1455078125, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 0.00011587142944335938, "grad_norm": 6.855053424835205, "learning_rate": 1e-05, "loss": 0.4881, "mean_token_accuracy": 0.8777026534080505, "num_tokens": 868342794.0, "step": 22869 }, { "epoch": 2.9092990713649662, "ewc_loss": 0.1455078125, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 0.00011539459228515625, "grad_norm": 7.281864643096924, "learning_rate": 1e-05, "loss": 0.5135, "mean_token_accuracy": 0.8724238872528076, "num_tokens": 868382009.0, "step": 22870 }, { "epoch": 2.9094262816435568, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 0.00011301040649414062, "grad_norm": 6.843305587768555, "learning_rate": 1e-05, "loss": 0.5316, "mean_token_accuracy": 0.86440110206604, "num_tokens": 868417770.0, "step": 22871 }, { "epoch": 2.9095534919221473, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.841363430023193, "learning_rate": 1e-05, "loss": 0.4685, "mean_token_accuracy": 0.8826967477798462, "num_tokens": 868455447.0, "step": 22872 }, { "epoch": 2.909680702200738, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.76820707321167, "learning_rate": 1e-05, "loss": 0.5102, "mean_token_accuracy": 0.8725336790084839, "num_tokens": 868500388.0, "step": 22873 }, { "epoch": 2.9098079124793284, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.531239032745361, "learning_rate": 1e-05, "loss": 0.4857, "mean_token_accuracy": 0.876802921295166, "num_tokens": 868535707.0, "step": 22874 }, { "epoch": 2.909935122757919, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.736526966094971, "learning_rate": 1e-05, "loss": 0.5483, "mean_token_accuracy": 0.8603734970092773, "num_tokens": 868576634.0, "step": 22875 }, { "epoch": 2.9100623330365094, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.331312656402588, "learning_rate": 1e-05, "loss": 0.4557, "mean_token_accuracy": 0.8840306997299194, "num_tokens": 868619914.0, "step": 22876 }, { "epoch": 2.9101895433151, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.728583812713623, "learning_rate": 1e-05, "loss": 0.4764, "mean_token_accuracy": 0.8799775838851929, "num_tokens": 868652279.0, "step": 22877 }, { "epoch": 2.9103167535936905, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.199720859527588, "learning_rate": 1e-05, "loss": 0.4953, "mean_token_accuracy": 0.8691772222518921, "num_tokens": 868689418.0, "step": 22878 }, { "epoch": 2.910443963872281, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.769979953765869, "learning_rate": 1e-05, "loss": 0.4361, "mean_token_accuracy": 0.8929193019866943, "num_tokens": 868726695.0, "step": 22879 }, { "epoch": 2.9105711741508715, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.0795722007751465, "learning_rate": 1e-05, "loss": 0.4821, "mean_token_accuracy": 0.8711535930633545, "num_tokens": 868766120.0, "step": 22880 }, { "epoch": 2.910698384429462, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.8912272453308105, "learning_rate": 1e-05, "loss": 0.4507, "mean_token_accuracy": 0.8886992931365967, "num_tokens": 868806041.0, "step": 22881 }, { "epoch": 2.9108255947080526, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 5.947080135345459, "learning_rate": 1e-05, "loss": 0.4713, "mean_token_accuracy": 0.8755968809127808, "num_tokens": 868852980.0, "step": 22882 }, { "epoch": 2.9109528049866427, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 0.00011301040649414062, "grad_norm": 7.162516117095947, "learning_rate": 1e-05, "loss": 0.5309, "mean_token_accuracy": 0.8656236529350281, "num_tokens": 868890951.0, "step": 22883 }, { "epoch": 2.9110800152652336, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 5.852859020233154, "learning_rate": 1e-05, "loss": 0.505, "mean_token_accuracy": 0.869305431842804, "num_tokens": 868926412.0, "step": 22884 }, { "epoch": 2.9112072255438237, "ewc_loss": 0.15234375, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 0.0001220703125, "grad_norm": 7.57288122177124, "learning_rate": 1e-05, "loss": 0.5356, "mean_token_accuracy": 0.8694103956222534, "num_tokens": 868965181.0, "step": 22885 }, { "epoch": 2.9113344358224147, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 5.940325736999512, "learning_rate": 1e-05, "loss": 0.4665, "mean_token_accuracy": 0.8800976872444153, "num_tokens": 869004635.0, "step": 22886 }, { "epoch": 2.9114616461010048, "ewc_loss": 0.1630859375, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 0.00013256072998046875, "grad_norm": 7.952887058258057, "learning_rate": 1e-05, "loss": 0.5148, "mean_token_accuracy": 0.8790773153305054, "num_tokens": 869043833.0, "step": 22887 }, { "epoch": 2.9115888563795957, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 6.394335746765137, "learning_rate": 1e-05, "loss": 0.5206, "mean_token_accuracy": 0.8674601316452026, "num_tokens": 869078868.0, "step": 22888 }, { "epoch": 2.911716066658186, "ewc_loss": 0.162109375, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 0.0001316070556640625, "grad_norm": 7.922093868255615, "learning_rate": 1e-05, "loss": 0.5079, "mean_token_accuracy": 0.8810793161392212, "num_tokens": 869113444.0, "step": 22889 }, { "epoch": 2.9118432769367764, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 0.00011682510375976562, "grad_norm": 6.807744026184082, "learning_rate": 1e-05, "loss": 0.4767, "mean_token_accuracy": 0.881873369216919, "num_tokens": 869147741.0, "step": 22890 }, { "epoch": 2.911970487215367, "ewc_loss": 0.150390625, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 0.00012063980102539062, "grad_norm": 7.5094380378723145, "learning_rate": 1e-05, "loss": 0.5117, "mean_token_accuracy": 0.8739984631538391, "num_tokens": 869180247.0, "step": 22891 }, { "epoch": 2.9120976974939574, "ewc_loss": 0.1455078125, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 0.00011539459228515625, "grad_norm": 6.866531848907471, "learning_rate": 1e-05, "loss": 0.4519, "mean_token_accuracy": 0.8928915858268738, "num_tokens": 869221602.0, "step": 22892 }, { "epoch": 2.912224907772548, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 0.00011157989501953125, "grad_norm": 7.0480780601501465, "learning_rate": 1e-05, "loss": 0.4826, "mean_token_accuracy": 0.8773012757301331, "num_tokens": 869259601.0, "step": 22893 }, { "epoch": 2.9123521180511385, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 0.00011157989501953125, "grad_norm": 6.800012111663818, "learning_rate": 1e-05, "loss": 0.4507, "mean_token_accuracy": 0.8896477222442627, "num_tokens": 869303366.0, "step": 22894 }, { "epoch": 2.912479328329729, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.667171955108643, "learning_rate": 1e-05, "loss": 0.5324, "mean_token_accuracy": 0.8610242605209351, "num_tokens": 869338579.0, "step": 22895 }, { "epoch": 2.9126065386083195, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.762582778930664, "learning_rate": 1e-05, "loss": 0.5333, "mean_token_accuracy": 0.86380934715271, "num_tokens": 869384129.0, "step": 22896 }, { "epoch": 2.91273374888691, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.410316467285156, "learning_rate": 1e-05, "loss": 0.4821, "mean_token_accuracy": 0.8769198656082153, "num_tokens": 869421752.0, "step": 22897 }, { "epoch": 2.9128609591655006, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.74418830871582, "learning_rate": 1e-05, "loss": 0.5007, "mean_token_accuracy": 0.872295618057251, "num_tokens": 869459309.0, "step": 22898 }, { "epoch": 2.912988169444091, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.240029811859131, "learning_rate": 1e-05, "loss": 0.4462, "mean_token_accuracy": 0.8878562450408936, "num_tokens": 869497718.0, "step": 22899 }, { "epoch": 2.9131153797226816, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.780738353729248, "learning_rate": 1e-05, "loss": 0.5139, "mean_token_accuracy": 0.869780421257019, "num_tokens": 869540522.0, "step": 22900 }, { "epoch": 2.913242590001272, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.105968475341797, "learning_rate": 1e-05, "loss": 0.5021, "mean_token_accuracy": 0.8685697913169861, "num_tokens": 869579610.0, "step": 22901 }, { "epoch": 2.9133698002798627, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.9148993492126465, "learning_rate": 1e-05, "loss": 0.4606, "mean_token_accuracy": 0.8869326114654541, "num_tokens": 869619186.0, "step": 22902 }, { "epoch": 2.9134970105584532, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 5.977704048156738, "learning_rate": 1e-05, "loss": 0.4899, "mean_token_accuracy": 0.8690226078033447, "num_tokens": 869657415.0, "step": 22903 }, { "epoch": 2.9136242208370438, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 0.0001125335693359375, "grad_norm": 7.15610933303833, "learning_rate": 1e-05, "loss": 0.5086, "mean_token_accuracy": 0.8746161460876465, "num_tokens": 869698319.0, "step": 22904 }, { "epoch": 2.9137514311156343, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 5.867319107055664, "learning_rate": 1e-05, "loss": 0.4443, "mean_token_accuracy": 0.8847202062606812, "num_tokens": 869738630.0, "step": 22905 }, { "epoch": 2.913878641394225, "ewc_loss": 0.1513671875, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 0.00012111663818359375, "grad_norm": 7.531989097595215, "learning_rate": 1e-05, "loss": 0.5382, "mean_token_accuracy": 0.8640041947364807, "num_tokens": 869780713.0, "step": 22906 }, { "epoch": 2.9140058516728153, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 5.951321601867676, "learning_rate": 1e-05, "loss": 0.4371, "mean_token_accuracy": 0.8896211385726929, "num_tokens": 869817046.0, "step": 22907 }, { "epoch": 2.9141330619514054, "ewc_loss": 0.1630859375, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 0.00013256072998046875, "grad_norm": 7.894062519073486, "learning_rate": 1e-05, "loss": 0.503, "mean_token_accuracy": 0.8852227330207825, "num_tokens": 869855099.0, "step": 22908 }, { "epoch": 2.9142602722299964, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.360565185546875, "learning_rate": 1e-05, "loss": 0.5043, "mean_token_accuracy": 0.8750119805335999, "num_tokens": 869893207.0, "step": 22909 }, { "epoch": 2.9143874825085865, "ewc_loss": 0.1611328125, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 0.00013065338134765625, "grad_norm": 7.90689754486084, "learning_rate": 1e-05, "loss": 0.5115, "mean_token_accuracy": 0.877763032913208, "num_tokens": 869929925.0, "step": 22910 }, { "epoch": 2.9145146927871775, "ewc_loss": 0.1474609375, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 0.00011730194091796875, "grad_norm": 6.747638702392578, "learning_rate": 1e-05, "loss": 0.49, "mean_token_accuracy": 0.8809279203414917, "num_tokens": 869974607.0, "step": 22911 }, { "epoch": 2.9146419030657675, "ewc_loss": 0.1513671875, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 0.00012111663818359375, "grad_norm": 7.5313286781311035, "learning_rate": 1e-05, "loss": 0.498, "mean_token_accuracy": 0.8742592334747314, "num_tokens": 870013056.0, "step": 22912 }, { "epoch": 2.914769113344358, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 0.00011444091796875, "grad_norm": 6.848850727081299, "learning_rate": 1e-05, "loss": 0.5236, "mean_token_accuracy": 0.8683239817619324, "num_tokens": 870053735.0, "step": 22913 }, { "epoch": 2.9148963236229486, "ewc_loss": 0.140625, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 7.08604621887207, "learning_rate": 1e-05, "loss": 0.4759, "mean_token_accuracy": 0.8849291205406189, "num_tokens": 870093607.0, "step": 22914 }, { "epoch": 2.915023533901539, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 0.00011110305786132812, "grad_norm": 6.795428276062012, "learning_rate": 1e-05, "loss": 0.4881, "mean_token_accuracy": 0.8765789866447449, "num_tokens": 870133118.0, "step": 22915 }, { "epoch": 2.9151507441801296, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.69700288772583, "learning_rate": 1e-05, "loss": 0.4753, "mean_token_accuracy": 0.8794640302658081, "num_tokens": 870169782.0, "step": 22916 }, { "epoch": 2.91527795445872, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.755019664764404, "learning_rate": 1e-05, "loss": 0.4871, "mean_token_accuracy": 0.877345860004425, "num_tokens": 870206891.0, "step": 22917 }, { "epoch": 2.9154051647373107, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.418659687042236, "learning_rate": 1e-05, "loss": 0.5092, "mean_token_accuracy": 0.869233250617981, "num_tokens": 870253035.0, "step": 22918 }, { "epoch": 2.9155323750159012, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.742130279541016, "learning_rate": 1e-05, "loss": 0.4938, "mean_token_accuracy": 0.8764875531196594, "num_tokens": 870292286.0, "step": 22919 }, { "epoch": 2.9156595852944918, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.241296768188477, "learning_rate": 1e-05, "loss": 0.5031, "mean_token_accuracy": 0.8669514656066895, "num_tokens": 870335594.0, "step": 22920 }, { "epoch": 2.9157867955730823, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.780974864959717, "learning_rate": 1e-05, "loss": 0.4699, "mean_token_accuracy": 0.8771997690200806, "num_tokens": 870379465.0, "step": 22921 }, { "epoch": 2.915914005851673, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.104359149932861, "learning_rate": 1e-05, "loss": 0.4443, "mean_token_accuracy": 0.8868122100830078, "num_tokens": 870418060.0, "step": 22922 }, { "epoch": 2.9160412161302633, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.912613868713379, "learning_rate": 1e-05, "loss": 0.4558, "mean_token_accuracy": 0.8832027316093445, "num_tokens": 870455565.0, "step": 22923 }, { "epoch": 2.916168426408854, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 5.971487045288086, "learning_rate": 1e-05, "loss": 0.4783, "mean_token_accuracy": 0.8750423192977905, "num_tokens": 870492609.0, "step": 22924 }, { "epoch": 2.9162956366874444, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 0.00011157989501953125, "grad_norm": 7.158176422119141, "learning_rate": 1e-05, "loss": 0.4967, "mean_token_accuracy": 0.8759990930557251, "num_tokens": 870528150.0, "step": 22925 }, { "epoch": 2.916422846966035, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 5.8730363845825195, "learning_rate": 1e-05, "loss": 0.4525, "mean_token_accuracy": 0.8815051913261414, "num_tokens": 870568188.0, "step": 22926 }, { "epoch": 2.9165500572446255, "ewc_loss": 0.150390625, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 0.00012063980102539062, "grad_norm": 7.536812782287598, "learning_rate": 1e-05, "loss": 0.4878, "mean_token_accuracy": 0.8805338144302368, "num_tokens": 870601459.0, "step": 22927 }, { "epoch": 2.916677267523216, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 5.948119640350342, "learning_rate": 1e-05, "loss": 0.4747, "mean_token_accuracy": 0.8792029619216919, "num_tokens": 870638976.0, "step": 22928 }, { "epoch": 2.9168044778018065, "ewc_loss": 0.162109375, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 0.0001316070556640625, "grad_norm": 7.915802955627441, "learning_rate": 1e-05, "loss": 0.5433, "mean_token_accuracy": 0.862606406211853, "num_tokens": 870671163.0, "step": 22929 }, { "epoch": 2.916931688080397, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.354036808013916, "learning_rate": 1e-05, "loss": 0.4986, "mean_token_accuracy": 0.8783510327339172, "num_tokens": 870705806.0, "step": 22930 }, { "epoch": 2.917058898358987, "ewc_loss": 0.162109375, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 0.0001316070556640625, "grad_norm": 7.936566352844238, "learning_rate": 1e-05, "loss": 0.5245, "mean_token_accuracy": 0.872435450553894, "num_tokens": 870743045.0, "step": 22931 }, { "epoch": 2.917186108637578, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 0.0001163482666015625, "grad_norm": 6.758684158325195, "learning_rate": 1e-05, "loss": 0.4661, "mean_token_accuracy": 0.8863592147827148, "num_tokens": 870780965.0, "step": 22932 }, { "epoch": 2.917313318916168, "ewc_loss": 0.15234375, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 0.0001220703125, "grad_norm": 7.551970481872559, "learning_rate": 1e-05, "loss": 0.4659, "mean_token_accuracy": 0.8896106481552124, "num_tokens": 870814694.0, "step": 22933 }, { "epoch": 2.917440529194759, "ewc_loss": 0.1455078125, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00011539459228515625, "grad_norm": 6.872267246246338, "learning_rate": 1e-05, "loss": 0.5147, "mean_token_accuracy": 0.8718689680099487, "num_tokens": 870845654.0, "step": 22934 }, { "epoch": 2.9175677394733492, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.0001125335693359375, "grad_norm": 7.089407444000244, "learning_rate": 1e-05, "loss": 0.4914, "mean_token_accuracy": 0.8766289949417114, "num_tokens": 870881720.0, "step": 22935 }, { "epoch": 2.91769494975194, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00011157989501953125, "grad_norm": 6.8039960861206055, "learning_rate": 1e-05, "loss": 0.4755, "mean_token_accuracy": 0.8820251226425171, "num_tokens": 870927480.0, "step": 22936 }, { "epoch": 2.9178221600305303, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.707625865936279, "learning_rate": 1e-05, "loss": 0.4878, "mean_token_accuracy": 0.8811652660369873, "num_tokens": 870962407.0, "step": 22937 }, { "epoch": 2.917949370309121, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.749218463897705, "learning_rate": 1e-05, "loss": 0.4644, "mean_token_accuracy": 0.8854818344116211, "num_tokens": 870994638.0, "step": 22938 }, { "epoch": 2.9180765805877114, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.434800148010254, "learning_rate": 1e-05, "loss": 0.4684, "mean_token_accuracy": 0.8821428418159485, "num_tokens": 871032638.0, "step": 22939 }, { "epoch": 2.918203790866302, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.717855930328369, "learning_rate": 1e-05, "loss": 0.486, "mean_token_accuracy": 0.875468373298645, "num_tokens": 871070955.0, "step": 22940 }, { "epoch": 2.9183310011448924, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.263665676116943, "learning_rate": 1e-05, "loss": 0.4882, "mean_token_accuracy": 0.8761289119720459, "num_tokens": 871109282.0, "step": 22941 }, { "epoch": 2.918458211423483, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.73904275894165, "learning_rate": 1e-05, "loss": 0.5008, "mean_token_accuracy": 0.867634654045105, "num_tokens": 871151214.0, "step": 22942 }, { "epoch": 2.9185854217020735, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.149996280670166, "learning_rate": 1e-05, "loss": 0.5086, "mean_token_accuracy": 0.8687412738800049, "num_tokens": 871192909.0, "step": 22943 }, { "epoch": 2.918712631980664, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.805448055267334, "learning_rate": 1e-05, "loss": 0.5068, "mean_token_accuracy": 0.8728780746459961, "num_tokens": 871234952.0, "step": 22944 }, { "epoch": 2.9188398422592545, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.029973030090332, "learning_rate": 1e-05, "loss": 0.4734, "mean_token_accuracy": 0.8765188455581665, "num_tokens": 871270789.0, "step": 22945 }, { "epoch": 2.918967052537845, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 6.989273548126221, "learning_rate": 1e-05, "loss": 0.5226, "mean_token_accuracy": 0.8697035312652588, "num_tokens": 871312425.0, "step": 22946 }, { "epoch": 2.9190942628164356, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 5.895163059234619, "learning_rate": 1e-05, "loss": 0.4357, "mean_token_accuracy": 0.8908605575561523, "num_tokens": 871347685.0, "step": 22947 }, { "epoch": 2.919221473095026, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.0001163482666015625, "grad_norm": 7.332974910736084, "learning_rate": 1e-05, "loss": 0.4934, "mean_token_accuracy": 0.8770661950111389, "num_tokens": 871381789.0, "step": 22948 }, { "epoch": 2.9193486833736166, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 5.845025539398193, "learning_rate": 1e-05, "loss": 0.4359, "mean_token_accuracy": 0.8882039785385132, "num_tokens": 871417578.0, "step": 22949 }, { "epoch": 2.919475893652207, "ewc_loss": 0.16015625, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00012969970703125, "grad_norm": 7.785664081573486, "learning_rate": 1e-05, "loss": 0.5113, "mean_token_accuracy": 0.8761041760444641, "num_tokens": 871454816.0, "step": 22950 }, { "epoch": 2.9196031039307977, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.099125385284424, "learning_rate": 1e-05, "loss": 0.4713, "mean_token_accuracy": 0.87773597240448, "num_tokens": 871488898.0, "step": 22951 }, { "epoch": 2.9197303142093882, "ewc_loss": 0.166015625, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.0001354217529296875, "grad_norm": 8.049488067626953, "learning_rate": 1e-05, "loss": 0.5033, "mean_token_accuracy": 0.8840670585632324, "num_tokens": 871531895.0, "step": 22952 }, { "epoch": 2.9198575244879788, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00011348724365234375, "grad_norm": 6.630873680114746, "learning_rate": 1e-05, "loss": 0.4766, "mean_token_accuracy": 0.8852688670158386, "num_tokens": 871566151.0, "step": 22953 }, { "epoch": 2.9199847347665693, "ewc_loss": 0.158203125, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.0001277923583984375, "grad_norm": 7.821500778198242, "learning_rate": 1e-05, "loss": 0.5137, "mean_token_accuracy": 0.874666690826416, "num_tokens": 871603777.0, "step": 22954 }, { "epoch": 2.92011194504516, "ewc_loss": 0.1484375, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00011873245239257812, "grad_norm": 6.917923927307129, "learning_rate": 1e-05, "loss": 0.4798, "mean_token_accuracy": 0.8820388913154602, "num_tokens": 871636118.0, "step": 22955 }, { "epoch": 2.92023915532375, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00011682510375976562, "grad_norm": 7.344141006469727, "learning_rate": 1e-05, "loss": 0.4976, "mean_token_accuracy": 0.87520432472229, "num_tokens": 871670948.0, "step": 22956 }, { "epoch": 2.920366365602341, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00011491775512695312, "grad_norm": 6.88493013381958, "learning_rate": 1e-05, "loss": 0.5263, "mean_token_accuracy": 0.871147632598877, "num_tokens": 871707987.0, "step": 22957 }, { "epoch": 2.920493575880931, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.882923126220703, "learning_rate": 1e-05, "loss": 0.4728, "mean_token_accuracy": 0.8852237462997437, "num_tokens": 871753861.0, "step": 22958 }, { "epoch": 2.920620786159522, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 6.8090033531188965, "learning_rate": 1e-05, "loss": 0.5105, "mean_token_accuracy": 0.8729069232940674, "num_tokens": 871792667.0, "step": 22959 }, { "epoch": 2.920747996438112, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.555269241333008, "learning_rate": 1e-05, "loss": 0.5385, "mean_token_accuracy": 0.8607253432273865, "num_tokens": 871830994.0, "step": 22960 }, { "epoch": 2.920875206716703, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.780600547790527, "learning_rate": 1e-05, "loss": 0.5337, "mean_token_accuracy": 0.8629617094993591, "num_tokens": 871869984.0, "step": 22961 }, { "epoch": 2.921002416995293, "ewc_loss": 0.125, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.335015773773193, "learning_rate": 1e-05, "loss": 0.513, "mean_token_accuracy": 0.8663309216499329, "num_tokens": 871906757.0, "step": 22962 }, { "epoch": 2.9211296272738836, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.774190902709961, "learning_rate": 1e-05, "loss": 0.4777, "mean_token_accuracy": 0.8842706084251404, "num_tokens": 871945387.0, "step": 22963 }, { "epoch": 2.921256837552474, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.188697338104248, "learning_rate": 1e-05, "loss": 0.4716, "mean_token_accuracy": 0.8781190514564514, "num_tokens": 871981456.0, "step": 22964 }, { "epoch": 2.9213840478310646, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.834722995758057, "learning_rate": 1e-05, "loss": 0.4665, "mean_token_accuracy": 0.8842231035232544, "num_tokens": 872022953.0, "step": 22965 }, { "epoch": 2.921511258109655, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.0571746826171875, "learning_rate": 1e-05, "loss": 0.4662, "mean_token_accuracy": 0.8776917457580566, "num_tokens": 872059103.0, "step": 22966 }, { "epoch": 2.9216384683882457, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 7.155742645263672, "learning_rate": 1e-05, "loss": 0.4756, "mean_token_accuracy": 0.8777136206626892, "num_tokens": 872091910.0, "step": 22967 }, { "epoch": 2.9217656786668362, "ewc_loss": 0.11865234375, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 5.90402889251709, "learning_rate": 1e-05, "loss": 0.4667, "mean_token_accuracy": 0.8769023418426514, "num_tokens": 872130863.0, "step": 22968 }, { "epoch": 2.9218928889454268, "ewc_loss": 0.1494140625, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00011920928955078125, "grad_norm": 7.452911376953125, "learning_rate": 1e-05, "loss": 0.5407, "mean_token_accuracy": 0.8676155805587769, "num_tokens": 872168104.0, "step": 22969 }, { "epoch": 2.9220200992240173, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 5.921815395355225, "learning_rate": 1e-05, "loss": 0.4921, "mean_token_accuracy": 0.8718538284301758, "num_tokens": 872197554.0, "step": 22970 }, { "epoch": 2.922147309502608, "ewc_loss": 0.16015625, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00012969970703125, "grad_norm": 7.842855930328369, "learning_rate": 1e-05, "loss": 0.5135, "mean_token_accuracy": 0.8756992816925049, "num_tokens": 872232283.0, "step": 22971 }, { "epoch": 2.9222745197811983, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.244025707244873, "learning_rate": 1e-05, "loss": 0.4679, "mean_token_accuracy": 0.8794531226158142, "num_tokens": 872263459.0, "step": 22972 }, { "epoch": 2.922401730059789, "ewc_loss": 0.1611328125, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00013065338134765625, "grad_norm": 7.944445610046387, "learning_rate": 1e-05, "loss": 0.4777, "mean_token_accuracy": 0.889237642288208, "num_tokens": 872303056.0, "step": 22973 }, { "epoch": 2.9225289403383794, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00011587142944335938, "grad_norm": 6.6943888664245605, "learning_rate": 1e-05, "loss": 0.5171, "mean_token_accuracy": 0.8752145767211914, "num_tokens": 872341030.0, "step": 22974 }, { "epoch": 2.92265615061697, "ewc_loss": 0.1552734375, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00012493133544921875, "grad_norm": 7.638168811798096, "learning_rate": 1e-05, "loss": 0.534, "mean_token_accuracy": 0.8714999556541443, "num_tokens": 872384006.0, "step": 22975 }, { "epoch": 2.9227833608955605, "ewc_loss": 0.1484375, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00011777877807617188, "grad_norm": 6.859158515930176, "learning_rate": 1e-05, "loss": 0.5317, "mean_token_accuracy": 0.8674522638320923, "num_tokens": 872425575.0, "step": 22976 }, { "epoch": 2.922910571174151, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00011444091796875, "grad_norm": 7.173676013946533, "learning_rate": 1e-05, "loss": 0.4825, "mean_token_accuracy": 0.8818308711051941, "num_tokens": 872463483.0, "step": 22977 }, { "epoch": 2.9230377814527415, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00011301040649414062, "grad_norm": 6.8212995529174805, "learning_rate": 1e-05, "loss": 0.5072, "mean_token_accuracy": 0.875103235244751, "num_tokens": 872496066.0, "step": 22978 }, { "epoch": 2.923164991731332, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.789177894592285, "learning_rate": 1e-05, "loss": 0.4794, "mean_token_accuracy": 0.8793960213661194, "num_tokens": 872529330.0, "step": 22979 }, { "epoch": 2.9232922020099226, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.754307746887207, "learning_rate": 1e-05, "loss": 0.4752, "mean_token_accuracy": 0.8808742761611938, "num_tokens": 872563989.0, "step": 22980 }, { "epoch": 2.9234194122885127, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.485649585723877, "learning_rate": 1e-05, "loss": 0.4556, "mean_token_accuracy": 0.8838123679161072, "num_tokens": 872601506.0, "step": 22981 }, { "epoch": 2.9235466225671036, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.710090637207031, "learning_rate": 1e-05, "loss": 0.4734, "mean_token_accuracy": 0.8809059858322144, "num_tokens": 872633258.0, "step": 22982 }, { "epoch": 2.9236738328456937, "ewc_loss": 0.125, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.330873489379883, "learning_rate": 1e-05, "loss": 0.513, "mean_token_accuracy": 0.8650280833244324, "num_tokens": 872665803.0, "step": 22983 }, { "epoch": 2.9238010431242847, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.696708679199219, "learning_rate": 1e-05, "loss": 0.487, "mean_token_accuracy": 0.8755462169647217, "num_tokens": 872703607.0, "step": 22984 }, { "epoch": 2.9239282534028748, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.210934638977051, "learning_rate": 1e-05, "loss": 0.4875, "mean_token_accuracy": 0.8745166063308716, "num_tokens": 872741966.0, "step": 22985 }, { "epoch": 2.9240554636814657, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.741657257080078, "learning_rate": 1e-05, "loss": 0.4741, "mean_token_accuracy": 0.8794803619384766, "num_tokens": 872773052.0, "step": 22986 }, { "epoch": 2.924182673960056, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.089301109313965, "learning_rate": 1e-05, "loss": 0.4944, "mean_token_accuracy": 0.8674591779708862, "num_tokens": 872813667.0, "step": 22987 }, { "epoch": 2.9243098842386464, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.867063522338867, "learning_rate": 1e-05, "loss": 0.5356, "mean_token_accuracy": 0.8652514815330505, "num_tokens": 872854320.0, "step": 22988 }, { "epoch": 2.924437094517237, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 5.9502177238464355, "learning_rate": 1e-05, "loss": 0.4662, "mean_token_accuracy": 0.877665102481842, "num_tokens": 872895190.0, "step": 22989 }, { "epoch": 2.9245643047958274, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00011396408081054688, "grad_norm": 7.143764019012451, "learning_rate": 1e-05, "loss": 0.494, "mean_token_accuracy": 0.8791772127151489, "num_tokens": 872932288.0, "step": 22990 }, { "epoch": 2.924691515074418, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 5.838503360748291, "learning_rate": 1e-05, "loss": 0.46, "mean_token_accuracy": 0.8818811774253845, "num_tokens": 872963617.0, "step": 22991 }, { "epoch": 2.9248187253530085, "ewc_loss": 0.1533203125, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.58082389831543, "learning_rate": 1e-05, "loss": 0.5289, "mean_token_accuracy": 0.8693603277206421, "num_tokens": 872997635.0, "step": 22992 }, { "epoch": 2.924945935631599, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 5.9094648361206055, "learning_rate": 1e-05, "loss": 0.5256, "mean_token_accuracy": 0.860704779624939, "num_tokens": 873037275.0, "step": 22993 }, { "epoch": 2.9250731459101895, "ewc_loss": 0.1650390625, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00013446807861328125, "grad_norm": 8.010344505310059, "learning_rate": 1e-05, "loss": 0.54, "mean_token_accuracy": 0.8690540790557861, "num_tokens": 873074791.0, "step": 22994 }, { "epoch": 2.92520035618878, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 6.376658916473389, "learning_rate": 1e-05, "loss": 0.4668, "mean_token_accuracy": 0.8840317726135254, "num_tokens": 873111521.0, "step": 22995 }, { "epoch": 2.9253275664673706, "ewc_loss": 0.1650390625, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00013446807861328125, "grad_norm": 8.03970718383789, "learning_rate": 1e-05, "loss": 0.5757, "mean_token_accuracy": 0.857284665107727, "num_tokens": 873150007.0, "step": 22996 }, { "epoch": 2.925454776745961, "ewc_loss": 0.150390625, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.0001201629638671875, "grad_norm": 6.862757682800293, "learning_rate": 1e-05, "loss": 0.5045, "mean_token_accuracy": 0.8777682781219482, "num_tokens": 873188360.0, "step": 22997 }, { "epoch": 2.9255819870245516, "ewc_loss": 0.1533203125, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.610118389129639, "learning_rate": 1e-05, "loss": 0.4679, "mean_token_accuracy": 0.884223461151123, "num_tokens": 873222962.0, "step": 22998 }, { "epoch": 2.925709197303142, "ewc_loss": 0.1494140625, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00011920928955078125, "grad_norm": 6.939637184143066, "learning_rate": 1e-05, "loss": 0.4757, "mean_token_accuracy": 0.8847111463546753, "num_tokens": 873260579.0, "step": 22999 }, { "epoch": 2.9258364075817327, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00011301040649414062, "grad_norm": 7.138247489929199, "learning_rate": 1e-05, "loss": 0.4634, "mean_token_accuracy": 0.8868836164474487, "num_tokens": 873292862.0, "step": 23000 }, { "epoch": 2.925963617860323, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00011157989501953125, "grad_norm": 6.850466251373291, "learning_rate": 1e-05, "loss": 0.5701, "mean_token_accuracy": 0.8521167039871216, "num_tokens": 873331824.0, "step": 23001 }, { "epoch": 2.9260908281389137, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.750791072845459, "learning_rate": 1e-05, "loss": 0.4814, "mean_token_accuracy": 0.8767375946044922, "num_tokens": 873364902.0, "step": 23002 }, { "epoch": 2.9262180384175043, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.755362033843994, "learning_rate": 1e-05, "loss": 0.4428, "mean_token_accuracy": 0.8913662433624268, "num_tokens": 873398523.0, "step": 23003 }, { "epoch": 2.926345248696095, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.489495277404785, "learning_rate": 1e-05, "loss": 0.479, "mean_token_accuracy": 0.8793415427207947, "num_tokens": 873435779.0, "step": 23004 }, { "epoch": 2.9264724589746853, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.700366020202637, "learning_rate": 1e-05, "loss": 0.4608, "mean_token_accuracy": 0.8876468539237976, "num_tokens": 873476706.0, "step": 23005 }, { "epoch": 2.9265996692532754, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.326503276824951, "learning_rate": 1e-05, "loss": 0.4767, "mean_token_accuracy": 0.8787823915481567, "num_tokens": 873507762.0, "step": 23006 }, { "epoch": 2.9267268795318664, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.6961283683776855, "learning_rate": 1e-05, "loss": 0.4509, "mean_token_accuracy": 0.884527862071991, "num_tokens": 873545692.0, "step": 23007 }, { "epoch": 2.9268540898104565, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.182915210723877, "learning_rate": 1e-05, "loss": 0.4653, "mean_token_accuracy": 0.8805989027023315, "num_tokens": 873589131.0, "step": 23008 }, { "epoch": 2.9269813000890474, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.768435001373291, "learning_rate": 1e-05, "loss": 0.5381, "mean_token_accuracy": 0.860654890537262, "num_tokens": 873628593.0, "step": 23009 }, { "epoch": 2.9271085103676375, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.057252407073975, "learning_rate": 1e-05, "loss": 0.4937, "mean_token_accuracy": 0.8708485960960388, "num_tokens": 873671525.0, "step": 23010 }, { "epoch": 2.927235720646228, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.942112445831299, "learning_rate": 1e-05, "loss": 0.4995, "mean_token_accuracy": 0.8776369094848633, "num_tokens": 873710618.0, "step": 23011 }, { "epoch": 2.9273629309248186, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 5.902427673339844, "learning_rate": 1e-05, "loss": 0.5278, "mean_token_accuracy": 0.8594964742660522, "num_tokens": 873751731.0, "step": 23012 }, { "epoch": 2.927490141203409, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00011587142944335938, "grad_norm": 7.288712978363037, "learning_rate": 1e-05, "loss": 0.4943, "mean_token_accuracy": 0.8823955059051514, "num_tokens": 873791948.0, "step": 23013 }, { "epoch": 2.9276173514819996, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.805424213409424, "learning_rate": 1e-05, "loss": 0.4885, "mean_token_accuracy": 0.8702934980392456, "num_tokens": 873834193.0, "step": 23014 }, { "epoch": 2.92774456176059, "ewc_loss": 0.1591796875, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00012874603271484375, "grad_norm": 7.804190635681152, "learning_rate": 1e-05, "loss": 0.5352, "mean_token_accuracy": 0.8705365657806396, "num_tokens": 873870035.0, "step": 23015 }, { "epoch": 2.9278717720391807, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.043818950653076, "learning_rate": 1e-05, "loss": 0.4665, "mean_token_accuracy": 0.8824359774589539, "num_tokens": 873902007.0, "step": 23016 }, { "epoch": 2.9279989823177712, "ewc_loss": 0.1669921875, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00013637542724609375, "grad_norm": 8.14988899230957, "learning_rate": 1e-05, "loss": 0.5781, "mean_token_accuracy": 0.8584730625152588, "num_tokens": 873947063.0, "step": 23017 }, { "epoch": 2.9281261925963618, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00011587142944335938, "grad_norm": 6.650661468505859, "learning_rate": 1e-05, "loss": 0.5363, "mean_token_accuracy": 0.8665295243263245, "num_tokens": 873987542.0, "step": 23018 }, { "epoch": 2.9282534028749523, "ewc_loss": 0.162109375, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.0001316070556640625, "grad_norm": 7.934311389923096, "learning_rate": 1e-05, "loss": 0.5173, "mean_token_accuracy": 0.8773221969604492, "num_tokens": 874020492.0, "step": 23019 }, { "epoch": 2.928380613153543, "ewc_loss": 0.1513671875, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00012111663818359375, "grad_norm": 6.988714218139648, "learning_rate": 1e-05, "loss": 0.4677, "mean_token_accuracy": 0.8871089220046997, "num_tokens": 874061902.0, "step": 23020 }, { "epoch": 2.9285078234321333, "ewc_loss": 0.1484375, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.000118255615234375, "grad_norm": 7.429302215576172, "learning_rate": 1e-05, "loss": 0.4551, "mean_token_accuracy": 0.8918986320495605, "num_tokens": 874093277.0, "step": 23021 }, { "epoch": 2.928635033710724, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00011491775512695312, "grad_norm": 6.927345275878906, "learning_rate": 1e-05, "loss": 0.4662, "mean_token_accuracy": 0.8853948712348938, "num_tokens": 874136543.0, "step": 23022 }, { "epoch": 2.9287622439893144, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 6.995070457458496, "learning_rate": 1e-05, "loss": 0.4584, "mean_token_accuracy": 0.8868468999862671, "num_tokens": 874180933.0, "step": 23023 }, { "epoch": 2.928889454267905, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 6.804502010345459, "learning_rate": 1e-05, "loss": 0.4776, "mean_token_accuracy": 0.881089448928833, "num_tokens": 874221980.0, "step": 23024 }, { "epoch": 2.9290166645464955, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.675471305847168, "learning_rate": 1e-05, "loss": 0.4808, "mean_token_accuracy": 0.8793626427650452, "num_tokens": 874255814.0, "step": 23025 }, { "epoch": 2.929143874825086, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.718952655792236, "learning_rate": 1e-05, "loss": 0.5176, "mean_token_accuracy": 0.8682795763015747, "num_tokens": 874296230.0, "step": 23026 }, { "epoch": 2.9292710851036765, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.463135719299316, "learning_rate": 1e-05, "loss": 0.5162, "mean_token_accuracy": 0.8642567992210388, "num_tokens": 874331854.0, "step": 23027 }, { "epoch": 2.929398295382267, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.645086765289307, "learning_rate": 1e-05, "loss": 0.4719, "mean_token_accuracy": 0.8814756870269775, "num_tokens": 874373348.0, "step": 23028 }, { "epoch": 2.929525505660857, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.518220901489258, "learning_rate": 1e-05, "loss": 0.4572, "mean_token_accuracy": 0.8802196979522705, "num_tokens": 874415226.0, "step": 23029 }, { "epoch": 2.929652715939448, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.48741340637207, "learning_rate": 1e-05, "loss": 0.5423, "mean_token_accuracy": 0.8599758148193359, "num_tokens": 874451894.0, "step": 23030 }, { "epoch": 2.929779926218038, "ewc_loss": 0.125, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.38883113861084, "learning_rate": 1e-05, "loss": 0.4655, "mean_token_accuracy": 0.881790280342102, "num_tokens": 874490248.0, "step": 23031 }, { "epoch": 2.929907136496629, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.392072677612305, "learning_rate": 1e-05, "loss": 0.4723, "mean_token_accuracy": 0.8803821206092834, "num_tokens": 874527421.0, "step": 23032 }, { "epoch": 2.9300343467752192, "ewc_loss": 0.125, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.37183952331543, "learning_rate": 1e-05, "loss": 0.5009, "mean_token_accuracy": 0.8707930445671082, "num_tokens": 874570834.0, "step": 23033 }, { "epoch": 2.93016155705381, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.35297155380249, "learning_rate": 1e-05, "loss": 0.5082, "mean_token_accuracy": 0.867543637752533, "num_tokens": 874607956.0, "step": 23034 }, { "epoch": 2.9302887673324003, "ewc_loss": 0.125, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.3632025718688965, "learning_rate": 1e-05, "loss": 0.4928, "mean_token_accuracy": 0.8717643618583679, "num_tokens": 874644599.0, "step": 23035 }, { "epoch": 2.930415977610991, "ewc_loss": 0.125, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.313858509063721, "learning_rate": 1e-05, "loss": 0.505, "mean_token_accuracy": 0.8708794116973877, "num_tokens": 874682214.0, "step": 23036 }, { "epoch": 2.9305431878895813, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.383556365966797, "learning_rate": 1e-05, "loss": 0.5121, "mean_token_accuracy": 0.8655083179473877, "num_tokens": 874719298.0, "step": 23037 }, { "epoch": 2.930670398168172, "ewc_loss": 0.12451171875, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.2452545166015625, "learning_rate": 1e-05, "loss": 0.4524, "mean_token_accuracy": 0.887511134147644, "num_tokens": 874757391.0, "step": 23038 }, { "epoch": 2.9307976084467624, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.4341349601745605, "learning_rate": 1e-05, "loss": 0.4632, "mean_token_accuracy": 0.8821032047271729, "num_tokens": 874788307.0, "step": 23039 }, { "epoch": 2.930924818725353, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.173126697540283, "learning_rate": 1e-05, "loss": 0.4753, "mean_token_accuracy": 0.874418318271637, "num_tokens": 874825648.0, "step": 23040 }, { "epoch": 2.9310520290039435, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.534920692443848, "learning_rate": 1e-05, "loss": 0.4899, "mean_token_accuracy": 0.8729093074798584, "num_tokens": 874862275.0, "step": 23041 }, { "epoch": 2.931179239282534, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.044210433959961, "learning_rate": 1e-05, "loss": 0.5006, "mean_token_accuracy": 0.8709691166877747, "num_tokens": 874898411.0, "step": 23042 }, { "epoch": 2.9313064495611245, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.728717803955078, "learning_rate": 1e-05, "loss": 0.471, "mean_token_accuracy": 0.8778136968612671, "num_tokens": 874935599.0, "step": 23043 }, { "epoch": 2.931433659839715, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 5.849931716918945, "learning_rate": 1e-05, "loss": 0.4814, "mean_token_accuracy": 0.8743748664855957, "num_tokens": 874966302.0, "step": 23044 }, { "epoch": 2.9315608701183056, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00011205673217773438, "grad_norm": 7.112727165222168, "learning_rate": 1e-05, "loss": 0.4536, "mean_token_accuracy": 0.8909615278244019, "num_tokens": 875003158.0, "step": 23045 }, { "epoch": 2.931688080396896, "ewc_loss": 0.11572265625, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.606679916381836, "learning_rate": 1e-05, "loss": 0.4657, "mean_token_accuracy": 0.8774475455284119, "num_tokens": 875042163.0, "step": 23046 }, { "epoch": 2.9318152906754866, "ewc_loss": 0.16015625, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00012969970703125, "grad_norm": 7.8492560386657715, "learning_rate": 1e-05, "loss": 0.5268, "mean_token_accuracy": 0.8724967837333679, "num_tokens": 875081957.0, "step": 23047 }, { "epoch": 2.931942500954077, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 5.714865207672119, "learning_rate": 1e-05, "loss": 0.4662, "mean_token_accuracy": 0.8775671124458313, "num_tokens": 875122286.0, "step": 23048 }, { "epoch": 2.9320697112326677, "ewc_loss": 0.18359375, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00015354156494140625, "grad_norm": 8.726151466369629, "learning_rate": 1e-05, "loss": 0.586, "mean_token_accuracy": 0.8610247373580933, "num_tokens": 875163320.0, "step": 23049 }, { "epoch": 2.932196921511258, "ewc_loss": 0.154296875, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.0001239776611328125, "grad_norm": 6.8204665184021, "learning_rate": 1e-05, "loss": 0.484, "mean_token_accuracy": 0.8854566216468811, "num_tokens": 875203819.0, "step": 23050 }, { "epoch": 2.9323241317898487, "ewc_loss": 0.177734375, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00014781951904296875, "grad_norm": 8.573600769042969, "learning_rate": 1e-05, "loss": 0.5192, "mean_token_accuracy": 0.8776763677597046, "num_tokens": 875239028.0, "step": 23051 }, { "epoch": 2.9324513420684393, "ewc_loss": 0.1650390625, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00013446807861328125, "grad_norm": 7.578275203704834, "learning_rate": 1e-05, "loss": 0.5565, "mean_token_accuracy": 0.8667915463447571, "num_tokens": 875276324.0, "step": 23052 }, { "epoch": 2.93257855234703, "ewc_loss": 0.1591796875, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00012874603271484375, "grad_norm": 7.82294225692749, "learning_rate": 1e-05, "loss": 0.4878, "mean_token_accuracy": 0.8834041953086853, "num_tokens": 875316918.0, "step": 23053 }, { "epoch": 2.93270576262562, "ewc_loss": 0.1552734375, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00012493133544921875, "grad_norm": 7.319071292877197, "learning_rate": 1e-05, "loss": 0.5296, "mean_token_accuracy": 0.8694204688072205, "num_tokens": 875351009.0, "step": 23054 }, { "epoch": 2.932832972904211, "ewc_loss": 0.1494140625, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00011920928955078125, "grad_norm": 7.415302276611328, "learning_rate": 1e-05, "loss": 0.4696, "mean_token_accuracy": 0.8863466382026672, "num_tokens": 875386988.0, "step": 23055 }, { "epoch": 2.932960183182801, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00011205673217773438, "grad_norm": 6.935062885284424, "learning_rate": 1e-05, "loss": 0.4839, "mean_token_accuracy": 0.8803187608718872, "num_tokens": 875429607.0, "step": 23056 }, { "epoch": 2.933087393461392, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.0001125335693359375, "grad_norm": 7.1895222663879395, "learning_rate": 1e-05, "loss": 0.484, "mean_token_accuracy": 0.8819194436073303, "num_tokens": 875461071.0, "step": 23057 }, { "epoch": 2.933214603739982, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.656745910644531, "learning_rate": 1e-05, "loss": 0.4706, "mean_token_accuracy": 0.8823552131652832, "num_tokens": 875498146.0, "step": 23058 }, { "epoch": 2.933341814018573, "ewc_loss": 0.140625, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 7.102114677429199, "learning_rate": 1e-05, "loss": 0.4969, "mean_token_accuracy": 0.8760792016983032, "num_tokens": 875536155.0, "step": 23059 }, { "epoch": 2.933469024297163, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.412895679473877, "learning_rate": 1e-05, "loss": 0.4912, "mean_token_accuracy": 0.8726145029067993, "num_tokens": 875569868.0, "step": 23060 }, { "epoch": 2.9335962345757536, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00011157989501953125, "grad_norm": 7.157180309295654, "learning_rate": 1e-05, "loss": 0.4978, "mean_token_accuracy": 0.8746803402900696, "num_tokens": 875609298.0, "step": 23061 }, { "epoch": 2.933723444854344, "ewc_loss": 0.125, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.226378917694092, "learning_rate": 1e-05, "loss": 0.4313, "mean_token_accuracy": 0.8895841836929321, "num_tokens": 875642837.0, "step": 23062 }, { "epoch": 2.9338506551329346, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00011491775512695312, "grad_norm": 7.371894359588623, "learning_rate": 1e-05, "loss": 0.5171, "mean_token_accuracy": 0.8714255690574646, "num_tokens": 875683273.0, "step": 23063 }, { "epoch": 2.933977865411525, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.149599552154541, "learning_rate": 1e-05, "loss": 0.5122, "mean_token_accuracy": 0.8609603643417358, "num_tokens": 875717461.0, "step": 23064 }, { "epoch": 2.9341050756901157, "ewc_loss": 0.15234375, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.0001220703125, "grad_norm": 7.707956314086914, "learning_rate": 1e-05, "loss": 0.5323, "mean_token_accuracy": 0.8665589094161987, "num_tokens": 875756975.0, "step": 23065 }, { "epoch": 2.934232285968706, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.316056728363037, "learning_rate": 1e-05, "loss": 0.4598, "mean_token_accuracy": 0.880896806716919, "num_tokens": 875793414.0, "step": 23066 }, { "epoch": 2.9343594962472968, "ewc_loss": 0.1572265625, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00012683868408203125, "grad_norm": 7.876692295074463, "learning_rate": 1e-05, "loss": 0.5652, "mean_token_accuracy": 0.8569028973579407, "num_tokens": 875822477.0, "step": 23067 }, { "epoch": 2.9344867065258873, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.725964069366455, "learning_rate": 1e-05, "loss": 0.4831, "mean_token_accuracy": 0.8777564167976379, "num_tokens": 875856220.0, "step": 23068 }, { "epoch": 2.934613916804478, "ewc_loss": 0.1484375, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00011873245239257812, "grad_norm": 7.564346790313721, "learning_rate": 1e-05, "loss": 0.501, "mean_token_accuracy": 0.8774385452270508, "num_tokens": 875890572.0, "step": 23069 }, { "epoch": 2.9347411270830683, "ewc_loss": 0.140625, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00011014938354492188, "grad_norm": 6.940837860107422, "learning_rate": 1e-05, "loss": 0.4815, "mean_token_accuracy": 0.885261058807373, "num_tokens": 875928231.0, "step": 23070 }, { "epoch": 2.934868337361659, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.9971208572387695, "learning_rate": 1e-05, "loss": 0.5173, "mean_token_accuracy": 0.8694612979888916, "num_tokens": 875966211.0, "step": 23071 }, { "epoch": 2.9349955476402494, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 6.959271430969238, "learning_rate": 1e-05, "loss": 0.4845, "mean_token_accuracy": 0.877804160118103, "num_tokens": 876004857.0, "step": 23072 }, { "epoch": 2.93512275791884, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.445425033569336, "learning_rate": 1e-05, "loss": 0.5196, "mean_token_accuracy": 0.8618320822715759, "num_tokens": 876041966.0, "step": 23073 }, { "epoch": 2.9352499681974304, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 7.071005344390869, "learning_rate": 1e-05, "loss": 0.4711, "mean_token_accuracy": 0.881626307964325, "num_tokens": 876075732.0, "step": 23074 }, { "epoch": 2.935377178476021, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 6.025238513946533, "learning_rate": 1e-05, "loss": 0.4881, "mean_token_accuracy": 0.8728620409965515, "num_tokens": 876116582.0, "step": 23075 }, { "epoch": 2.9355043887546115, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00011491775512695312, "grad_norm": 7.393088340759277, "learning_rate": 1e-05, "loss": 0.4498, "mean_token_accuracy": 0.8918067216873169, "num_tokens": 876150122.0, "step": 23076 }, { "epoch": 2.935631599033202, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.863699436187744, "learning_rate": 1e-05, "loss": 0.4417, "mean_token_accuracy": 0.8872364163398743, "num_tokens": 876187902.0, "step": 23077 }, { "epoch": 2.9357588093117926, "ewc_loss": 0.1591796875, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00012874603271484375, "grad_norm": 7.884295463562012, "learning_rate": 1e-05, "loss": 0.5225, "mean_token_accuracy": 0.8734725117683411, "num_tokens": 876224304.0, "step": 23078 }, { "epoch": 2.9358860195903826, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.167916297912598, "learning_rate": 1e-05, "loss": 0.5331, "mean_token_accuracy": 0.8595201969146729, "num_tokens": 876261436.0, "step": 23079 }, { "epoch": 2.9360132298689736, "ewc_loss": 0.1650390625, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00013446807861328125, "grad_norm": 8.155220985412598, "learning_rate": 1e-05, "loss": 0.5369, "mean_token_accuracy": 0.8684919476509094, "num_tokens": 876299996.0, "step": 23080 }, { "epoch": 2.9361404401475637, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00011444091796875, "grad_norm": 6.79218864440918, "learning_rate": 1e-05, "loss": 0.5217, "mean_token_accuracy": 0.867091178894043, "num_tokens": 876337793.0, "step": 23081 }, { "epoch": 2.9362676504261547, "ewc_loss": 0.158203125, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.0001277923583984375, "grad_norm": 7.81065559387207, "learning_rate": 1e-05, "loss": 0.5053, "mean_token_accuracy": 0.8797498345375061, "num_tokens": 876371837.0, "step": 23082 }, { "epoch": 2.9363948607047448, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00011682510375976562, "grad_norm": 7.0262885093688965, "learning_rate": 1e-05, "loss": 0.479, "mean_token_accuracy": 0.8853586912155151, "num_tokens": 876414402.0, "step": 23083 }, { "epoch": 2.9365220709833357, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.0001163482666015625, "grad_norm": 7.306557655334473, "learning_rate": 1e-05, "loss": 0.4473, "mean_token_accuracy": 0.8926195502281189, "num_tokens": 876451790.0, "step": 23084 }, { "epoch": 2.936649281261926, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 6.9223408699035645, "learning_rate": 1e-05, "loss": 0.5291, "mean_token_accuracy": 0.8608729839324951, "num_tokens": 876486891.0, "step": 23085 }, { "epoch": 2.9367764915405163, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.9187750816345215, "learning_rate": 1e-05, "loss": 0.4579, "mean_token_accuracy": 0.8861398100852966, "num_tokens": 876525560.0, "step": 23086 }, { "epoch": 2.936903701819107, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.733283996582031, "learning_rate": 1e-05, "loss": 0.516, "mean_token_accuracy": 0.8700917959213257, "num_tokens": 876569037.0, "step": 23087 }, { "epoch": 2.9370309120976974, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.721049785614014, "learning_rate": 1e-05, "loss": 0.4955, "mean_token_accuracy": 0.8717531561851501, "num_tokens": 876605354.0, "step": 23088 }, { "epoch": 2.937158122376288, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.539482593536377, "learning_rate": 1e-05, "loss": 0.4541, "mean_token_accuracy": 0.8833937644958496, "num_tokens": 876644251.0, "step": 23089 }, { "epoch": 2.9372853326548785, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.66952657699585, "learning_rate": 1e-05, "loss": 0.4599, "mean_token_accuracy": 0.886060893535614, "num_tokens": 876686558.0, "step": 23090 }, { "epoch": 2.937412542933469, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.351232051849365, "learning_rate": 1e-05, "loss": 0.4694, "mean_token_accuracy": 0.8774933815002441, "num_tokens": 876718629.0, "step": 23091 }, { "epoch": 2.9375397532120595, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.740312576293945, "learning_rate": 1e-05, "loss": 0.5163, "mean_token_accuracy": 0.864876389503479, "num_tokens": 876758037.0, "step": 23092 }, { "epoch": 2.93766696349065, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.150254249572754, "learning_rate": 1e-05, "loss": 0.4716, "mean_token_accuracy": 0.8782463073730469, "num_tokens": 876795201.0, "step": 23093 }, { "epoch": 2.9377941737692406, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.930904865264893, "learning_rate": 1e-05, "loss": 0.4511, "mean_token_accuracy": 0.8894131183624268, "num_tokens": 876833465.0, "step": 23094 }, { "epoch": 2.937921384047831, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 5.960836887359619, "learning_rate": 1e-05, "loss": 0.4775, "mean_token_accuracy": 0.8751999139785767, "num_tokens": 876869915.0, "step": 23095 }, { "epoch": 2.9380485943264216, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00011348724365234375, "grad_norm": 7.290175437927246, "learning_rate": 1e-05, "loss": 0.4758, "mean_token_accuracy": 0.8834078311920166, "num_tokens": 876906811.0, "step": 23096 }, { "epoch": 2.938175804605012, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 5.870364665985107, "learning_rate": 1e-05, "loss": 0.4938, "mean_token_accuracy": 0.8680094480514526, "num_tokens": 876944247.0, "step": 23097 }, { "epoch": 2.9383030148836027, "ewc_loss": 0.1572265625, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00012683868408203125, "grad_norm": 7.777108192443848, "learning_rate": 1e-05, "loss": 0.502, "mean_token_accuracy": 0.8786638379096985, "num_tokens": 876978557.0, "step": 23098 }, { "epoch": 2.938430225162193, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.135475158691406, "learning_rate": 1e-05, "loss": 0.5291, "mean_token_accuracy": 0.8598587512969971, "num_tokens": 877021241.0, "step": 23099 }, { "epoch": 2.9385574354407837, "ewc_loss": 0.1630859375, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00013256072998046875, "grad_norm": 8.036901473999023, "learning_rate": 1e-05, "loss": 0.5375, "mean_token_accuracy": 0.8709932565689087, "num_tokens": 877060777.0, "step": 23100 }, { "epoch": 2.9386846457193743, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00011157989501953125, "grad_norm": 6.7175822257995605, "learning_rate": 1e-05, "loss": 0.5304, "mean_token_accuracy": 0.8674092292785645, "num_tokens": 877096955.0, "step": 23101 }, { "epoch": 2.938811855997965, "ewc_loss": 0.154296875, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.0001239776611328125, "grad_norm": 7.717160224914551, "learning_rate": 1e-05, "loss": 0.5266, "mean_token_accuracy": 0.874393880367279, "num_tokens": 877130202.0, "step": 23102 }, { "epoch": 2.9389390662765553, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00011587142944335938, "grad_norm": 6.9810895919799805, "learning_rate": 1e-05, "loss": 0.5171, "mean_token_accuracy": 0.8695064187049866, "num_tokens": 877168777.0, "step": 23103 }, { "epoch": 2.9390662765551454, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00011205673217773438, "grad_norm": 7.14374303817749, "learning_rate": 1e-05, "loss": 0.5076, "mean_token_accuracy": 0.8738298416137695, "num_tokens": 877209737.0, "step": 23104 }, { "epoch": 2.9391934868337364, "ewc_loss": 0.140625, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 6.949000835418701, "learning_rate": 1e-05, "loss": 0.4904, "mean_token_accuracy": 0.8773741722106934, "num_tokens": 877245826.0, "step": 23105 }, { "epoch": 2.9393206971123265, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.638706207275391, "learning_rate": 1e-05, "loss": 0.4821, "mean_token_accuracy": 0.8823074102401733, "num_tokens": 877285249.0, "step": 23106 }, { "epoch": 2.9394479073909174, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 6.960925102233887, "learning_rate": 1e-05, "loss": 0.4399, "mean_token_accuracy": 0.8925549983978271, "num_tokens": 877321971.0, "step": 23107 }, { "epoch": 2.9395751176695075, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.278736591339111, "learning_rate": 1e-05, "loss": 0.5152, "mean_token_accuracy": 0.8720784187316895, "num_tokens": 877361479.0, "step": 23108 }, { "epoch": 2.939702327948098, "ewc_loss": 0.140625, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00010967254638671875, "grad_norm": 7.087940692901611, "learning_rate": 1e-05, "loss": 0.4908, "mean_token_accuracy": 0.8751933574676514, "num_tokens": 877400166.0, "step": 23109 }, { "epoch": 2.9398295382266886, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.077548980712891, "learning_rate": 1e-05, "loss": 0.4679, "mean_token_accuracy": 0.8786516785621643, "num_tokens": 877437755.0, "step": 23110 }, { "epoch": 2.939956748505279, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00011348724365234375, "grad_norm": 7.307655334472656, "learning_rate": 1e-05, "loss": 0.5444, "mean_token_accuracy": 0.8609973192214966, "num_tokens": 877471496.0, "step": 23111 }, { "epoch": 2.9400839587838696, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.038265705108643, "learning_rate": 1e-05, "loss": 0.4347, "mean_token_accuracy": 0.8897719383239746, "num_tokens": 877504576.0, "step": 23112 }, { "epoch": 2.94021116906246, "ewc_loss": 0.150390625, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00012063980102539062, "grad_norm": 7.565051555633545, "learning_rate": 1e-05, "loss": 0.5237, "mean_token_accuracy": 0.8720887899398804, "num_tokens": 877542795.0, "step": 23113 }, { "epoch": 2.9403383793410507, "ewc_loss": 0.125, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.178362846374512, "learning_rate": 1e-05, "loss": 0.4943, "mean_token_accuracy": 0.8709659576416016, "num_tokens": 877581731.0, "step": 23114 }, { "epoch": 2.940465589619641, "ewc_loss": 0.1552734375, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00012493133544921875, "grad_norm": 7.70058012008667, "learning_rate": 1e-05, "loss": 0.5309, "mean_token_accuracy": 0.8715503811836243, "num_tokens": 877618405.0, "step": 23115 }, { "epoch": 2.9405927998982317, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.438406467437744, "learning_rate": 1e-05, "loss": 0.4996, "mean_token_accuracy": 0.8781405091285706, "num_tokens": 877657792.0, "step": 23116 }, { "epoch": 2.9407200101768223, "ewc_loss": 0.15234375, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.0001220703125, "grad_norm": 7.611917972564697, "learning_rate": 1e-05, "loss": 0.4723, "mean_token_accuracy": 0.8880689144134521, "num_tokens": 877688965.0, "step": 23117 }, { "epoch": 2.940847220455413, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.617969989776611, "learning_rate": 1e-05, "loss": 0.4592, "mean_token_accuracy": 0.8863170146942139, "num_tokens": 877726595.0, "step": 23118 }, { "epoch": 2.9409744307340033, "ewc_loss": 0.1455078125, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00011539459228515625, "grad_norm": 7.344814777374268, "learning_rate": 1e-05, "loss": 0.4987, "mean_token_accuracy": 0.8787216544151306, "num_tokens": 877771713.0, "step": 23119 }, { "epoch": 2.941101641012594, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.648576259613037, "learning_rate": 1e-05, "loss": 0.4764, "mean_token_accuracy": 0.8814457654953003, "num_tokens": 877805048.0, "step": 23120 }, { "epoch": 2.9412288512911844, "ewc_loss": 0.140625, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00011014938354492188, "grad_norm": 7.049831867218018, "learning_rate": 1e-05, "loss": 0.4708, "mean_token_accuracy": 0.8817540407180786, "num_tokens": 877840469.0, "step": 23121 }, { "epoch": 2.941356061569775, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.596289157867432, "learning_rate": 1e-05, "loss": 0.5012, "mean_token_accuracy": 0.8696247935295105, "num_tokens": 877876747.0, "step": 23122 }, { "epoch": 2.9414832718483654, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.833197593688965, "learning_rate": 1e-05, "loss": 0.474, "mean_token_accuracy": 0.8805183172225952, "num_tokens": 877913871.0, "step": 23123 }, { "epoch": 2.941610482126956, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.496356010437012, "learning_rate": 1e-05, "loss": 0.5695, "mean_token_accuracy": 0.862094521522522, "num_tokens": 877954195.0, "step": 23124 }, { "epoch": 2.9417376924055465, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.723211765289307, "learning_rate": 1e-05, "loss": 0.5191, "mean_token_accuracy": 0.8623108267784119, "num_tokens": 877989124.0, "step": 23125 }, { "epoch": 2.941864902684137, "ewc_loss": 0.125, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.361568450927734, "learning_rate": 1e-05, "loss": 0.486, "mean_token_accuracy": 0.8744574785232544, "num_tokens": 878030188.0, "step": 23126 }, { "epoch": 2.941992112962727, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.714416027069092, "learning_rate": 1e-05, "loss": 0.4242, "mean_token_accuracy": 0.8965569138526917, "num_tokens": 878073839.0, "step": 23127 }, { "epoch": 2.942119323241318, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.228256702423096, "learning_rate": 1e-05, "loss": 0.4654, "mean_token_accuracy": 0.8793292045593262, "num_tokens": 878111446.0, "step": 23128 }, { "epoch": 2.942246533519908, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.826049327850342, "learning_rate": 1e-05, "loss": 0.543, "mean_token_accuracy": 0.857681155204773, "num_tokens": 878154543.0, "step": 23129 }, { "epoch": 2.942373743798499, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.0839715003967285, "learning_rate": 1e-05, "loss": 0.4589, "mean_token_accuracy": 0.8841980695724487, "num_tokens": 878194265.0, "step": 23130 }, { "epoch": 2.9425009540770892, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 7.052277088165283, "learning_rate": 1e-05, "loss": 0.5463, "mean_token_accuracy": 0.8618744015693665, "num_tokens": 878231475.0, "step": 23131 }, { "epoch": 2.94262816435568, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 5.967444896697998, "learning_rate": 1e-05, "loss": 0.4681, "mean_token_accuracy": 0.8784868717193604, "num_tokens": 878269492.0, "step": 23132 }, { "epoch": 2.9427553746342703, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.0001163482666015625, "grad_norm": 7.362009048461914, "learning_rate": 1e-05, "loss": 0.542, "mean_token_accuracy": 0.8638852834701538, "num_tokens": 878307047.0, "step": 23133 }, { "epoch": 2.942882584912861, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.009196758270264, "learning_rate": 1e-05, "loss": 0.4891, "mean_token_accuracy": 0.8729920387268066, "num_tokens": 878343971.0, "step": 23134 }, { "epoch": 2.9430097951914513, "ewc_loss": 0.15234375, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.0001220703125, "grad_norm": 7.648416996002197, "learning_rate": 1e-05, "loss": 0.4902, "mean_token_accuracy": 0.8829259276390076, "num_tokens": 878380858.0, "step": 23135 }, { "epoch": 2.943137005470042, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.272313117980957, "learning_rate": 1e-05, "loss": 0.4736, "mean_token_accuracy": 0.8840739727020264, "num_tokens": 878420854.0, "step": 23136 }, { "epoch": 2.9432642157486324, "ewc_loss": 0.154296875, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.0001239776611328125, "grad_norm": 7.655055999755859, "learning_rate": 1e-05, "loss": 0.5401, "mean_token_accuracy": 0.8664395213127136, "num_tokens": 878455465.0, "step": 23137 }, { "epoch": 2.943391426027223, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.615278244018555, "learning_rate": 1e-05, "loss": 0.4597, "mean_token_accuracy": 0.8885675668716431, "num_tokens": 878487122.0, "step": 23138 }, { "epoch": 2.9435186363058135, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00011491775512695312, "grad_norm": 7.289852619171143, "learning_rate": 1e-05, "loss": 0.5453, "mean_token_accuracy": 0.8622272610664368, "num_tokens": 878523881.0, "step": 23139 }, { "epoch": 2.943645846584404, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 7.14243221282959, "learning_rate": 1e-05, "loss": 0.5178, "mean_token_accuracy": 0.8667061924934387, "num_tokens": 878564351.0, "step": 23140 }, { "epoch": 2.9437730568629945, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.5431694984436035, "learning_rate": 1e-05, "loss": 0.4438, "mean_token_accuracy": 0.8879361152648926, "num_tokens": 878605759.0, "step": 23141 }, { "epoch": 2.943900267141585, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 7.032135486602783, "learning_rate": 1e-05, "loss": 0.4673, "mean_token_accuracy": 0.8867713809013367, "num_tokens": 878644975.0, "step": 23142 }, { "epoch": 2.9440274774201756, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.061076641082764, "learning_rate": 1e-05, "loss": 0.4572, "mean_token_accuracy": 0.8834286332130432, "num_tokens": 878679882.0, "step": 23143 }, { "epoch": 2.944154687698766, "ewc_loss": 0.1484375, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00011777877807617188, "grad_norm": 7.404842853546143, "learning_rate": 1e-05, "loss": 0.4543, "mean_token_accuracy": 0.8908630609512329, "num_tokens": 878723539.0, "step": 23144 }, { "epoch": 2.9442818979773566, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 5.924433708190918, "learning_rate": 1e-05, "loss": 0.4453, "mean_token_accuracy": 0.8862115740776062, "num_tokens": 878761993.0, "step": 23145 }, { "epoch": 2.944409108255947, "ewc_loss": 0.16015625, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00012969970703125, "grad_norm": 7.849203586578369, "learning_rate": 1e-05, "loss": 0.5051, "mean_token_accuracy": 0.8779427409172058, "num_tokens": 878794355.0, "step": 23146 }, { "epoch": 2.9445363185345377, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.273801803588867, "learning_rate": 1e-05, "loss": 0.5445, "mean_token_accuracy": 0.8600925207138062, "num_tokens": 878828477.0, "step": 23147 }, { "epoch": 2.944663528813128, "ewc_loss": 0.162109375, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.0001316070556640625, "grad_norm": 7.966310501098633, "learning_rate": 1e-05, "loss": 0.5451, "mean_token_accuracy": 0.864771842956543, "num_tokens": 878868134.0, "step": 23148 }, { "epoch": 2.9447907390917187, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00011301040649414062, "grad_norm": 6.718108177185059, "learning_rate": 1e-05, "loss": 0.488, "mean_token_accuracy": 0.8806964159011841, "num_tokens": 878908912.0, "step": 23149 }, { "epoch": 2.9449179493703093, "ewc_loss": 0.1533203125, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.6402974128723145, "learning_rate": 1e-05, "loss": 0.5069, "mean_token_accuracy": 0.8774298429489136, "num_tokens": 878947354.0, "step": 23150 }, { "epoch": 2.9450451596489, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.0001125335693359375, "grad_norm": 6.836087226867676, "learning_rate": 1e-05, "loss": 0.4817, "mean_token_accuracy": 0.8811219930648804, "num_tokens": 878985538.0, "step": 23151 }, { "epoch": 2.94517236992749, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.00011587142944335938, "grad_norm": 7.27402400970459, "learning_rate": 1e-05, "loss": 0.5237, "mean_token_accuracy": 0.8684072494506836, "num_tokens": 879028884.0, "step": 23152 }, { "epoch": 2.945299580206081, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.71131706237793, "learning_rate": 1e-05, "loss": 0.4704, "mean_token_accuracy": 0.8808358907699585, "num_tokens": 879064388.0, "step": 23153 }, { "epoch": 2.945426790484671, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 7.024791717529297, "learning_rate": 1e-05, "loss": 0.4649, "mean_token_accuracy": 0.883194088935852, "num_tokens": 879100353.0, "step": 23154 }, { "epoch": 2.945554000763262, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.533987998962402, "learning_rate": 1e-05, "loss": 0.4665, "mean_token_accuracy": 0.8830077648162842, "num_tokens": 879138447.0, "step": 23155 }, { "epoch": 2.945681211041852, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.943830490112305, "learning_rate": 1e-05, "loss": 0.538, "mean_token_accuracy": 0.8624024391174316, "num_tokens": 879172825.0, "step": 23156 }, { "epoch": 2.945808421320443, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.342997074127197, "learning_rate": 1e-05, "loss": 0.4922, "mean_token_accuracy": 0.870446503162384, "num_tokens": 879212086.0, "step": 23157 }, { "epoch": 2.945935631599033, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 6.9943976402282715, "learning_rate": 1e-05, "loss": 0.4833, "mean_token_accuracy": 0.8801301717758179, "num_tokens": 879247689.0, "step": 23158 }, { "epoch": 2.9460628418776236, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.178684234619141, "learning_rate": 1e-05, "loss": 0.4532, "mean_token_accuracy": 0.8857886791229248, "num_tokens": 879283510.0, "step": 23159 }, { "epoch": 2.946190052156214, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.0001125335693359375, "grad_norm": 7.162068843841553, "learning_rate": 1e-05, "loss": 0.4986, "mean_token_accuracy": 0.8763571977615356, "num_tokens": 879322037.0, "step": 23160 }, { "epoch": 2.9463172624348046, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.088905334472656, "learning_rate": 1e-05, "loss": 0.4987, "mean_token_accuracy": 0.8670364022254944, "num_tokens": 879366400.0, "step": 23161 }, { "epoch": 2.946444472713395, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.0001163482666015625, "grad_norm": 7.396218776702881, "learning_rate": 1e-05, "loss": 0.5702, "mean_token_accuracy": 0.855867326259613, "num_tokens": 879397768.0, "step": 23162 }, { "epoch": 2.9465716829919857, "ewc_loss": 0.125, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.166154384613037, "learning_rate": 1e-05, "loss": 0.479, "mean_token_accuracy": 0.876237154006958, "num_tokens": 879436751.0, "step": 23163 }, { "epoch": 2.946698893270576, "ewc_loss": 0.150390625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.00011968612670898438, "grad_norm": 7.5308518409729, "learning_rate": 1e-05, "loss": 0.557, "mean_token_accuracy": 0.8602166175842285, "num_tokens": 879477750.0, "step": 23164 }, { "epoch": 2.9468261035491667, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.392074108123779, "learning_rate": 1e-05, "loss": 0.4858, "mean_token_accuracy": 0.8741886019706726, "num_tokens": 879516049.0, "step": 23165 }, { "epoch": 2.9469533138277573, "ewc_loss": 0.1474609375, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.00011730194091796875, "grad_norm": 7.4013895988464355, "learning_rate": 1e-05, "loss": 0.5481, "mean_token_accuracy": 0.8615425825119019, "num_tokens": 879553176.0, "step": 23166 }, { "epoch": 2.947080524106348, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.609371185302734, "learning_rate": 1e-05, "loss": 0.4449, "mean_token_accuracy": 0.8899661302566528, "num_tokens": 879588435.0, "step": 23167 }, { "epoch": 2.9472077343849383, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 7.016493797302246, "learning_rate": 1e-05, "loss": 0.4539, "mean_token_accuracy": 0.8900740146636963, "num_tokens": 879623232.0, "step": 23168 }, { "epoch": 2.947334944663529, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.720624923706055, "learning_rate": 1e-05, "loss": 0.5276, "mean_token_accuracy": 0.8666099905967712, "num_tokens": 879669470.0, "step": 23169 }, { "epoch": 2.9474621549421194, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.538341045379639, "learning_rate": 1e-05, "loss": 0.4581, "mean_token_accuracy": 0.8875970840454102, "num_tokens": 879708173.0, "step": 23170 }, { "epoch": 2.94758936522071, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 6.844397068023682, "learning_rate": 1e-05, "loss": 0.5249, "mean_token_accuracy": 0.8693811297416687, "num_tokens": 879749086.0, "step": 23171 }, { "epoch": 2.9477165754993004, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 6.117704391479492, "learning_rate": 1e-05, "loss": 0.4914, "mean_token_accuracy": 0.8771535754203796, "num_tokens": 879786521.0, "step": 23172 }, { "epoch": 2.947843785777891, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.00011157989501953125, "grad_norm": 7.1121602058410645, "learning_rate": 1e-05, "loss": 0.4755, "mean_token_accuracy": 0.8850827813148499, "num_tokens": 879820332.0, "step": 23173 }, { "epoch": 2.9479709960564815, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 8.535385131835938e-05, "grad_norm": 5.826208114624023, "learning_rate": 1e-05, "loss": 0.4736, "mean_token_accuracy": 0.8758935332298279, "num_tokens": 879859224.0, "step": 23174 }, { "epoch": 2.948098206335072, "ewc_loss": 0.15234375, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.00012159347534179688, "grad_norm": 7.58828067779541, "learning_rate": 1e-05, "loss": 0.4827, "mean_token_accuracy": 0.8812068700790405, "num_tokens": 879892177.0, "step": 23175 }, { "epoch": 2.9482254166136626, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 5.8703436851501465, "learning_rate": 1e-05, "loss": 0.4233, "mean_token_accuracy": 0.8958694934844971, "num_tokens": 879924195.0, "step": 23176 }, { "epoch": 2.9483526268922526, "ewc_loss": 0.1669921875, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.00013637542724609375, "grad_norm": 8.063389778137207, "learning_rate": 1e-05, "loss": 0.5299, "mean_token_accuracy": 0.87346351146698, "num_tokens": 879965679.0, "step": 23177 }, { "epoch": 2.9484798371708436, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.410519123077393, "learning_rate": 1e-05, "loss": 0.4694, "mean_token_accuracy": 0.8816359043121338, "num_tokens": 880006441.0, "step": 23178 }, { "epoch": 2.9486070474494337, "ewc_loss": 0.1650390625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.00013446807861328125, "grad_norm": 8.051186561584473, "learning_rate": 1e-05, "loss": 0.5575, "mean_token_accuracy": 0.8643908500671387, "num_tokens": 880042823.0, "step": 23179 }, { "epoch": 2.9487342577280247, "ewc_loss": 0.1494140625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.00011920928955078125, "grad_norm": 6.940876483917236, "learning_rate": 1e-05, "loss": 0.4969, "mean_token_accuracy": 0.8767621517181396, "num_tokens": 880073949.0, "step": 23180 }, { "epoch": 2.9488614680066147, "ewc_loss": 0.15234375, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.00012159347534179688, "grad_norm": 7.558603286743164, "learning_rate": 1e-05, "loss": 0.4502, "mean_token_accuracy": 0.8957464694976807, "num_tokens": 880112921.0, "step": 23181 }, { "epoch": 2.9489886782852053, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.00011587142944335938, "grad_norm": 6.968855381011963, "learning_rate": 1e-05, "loss": 0.5292, "mean_token_accuracy": 0.8681188225746155, "num_tokens": 880152384.0, "step": 23182 }, { "epoch": 2.949115888563796, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.00011301040649414062, "grad_norm": 7.108996391296387, "learning_rate": 1e-05, "loss": 0.509, "mean_token_accuracy": 0.8768081068992615, "num_tokens": 880187107.0, "step": 23183 }, { "epoch": 2.9492430988423863, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 6.826849937438965, "learning_rate": 1e-05, "loss": 0.4824, "mean_token_accuracy": 0.8792302012443542, "num_tokens": 880226072.0, "step": 23184 }, { "epoch": 2.949370309120977, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.788008689880371, "learning_rate": 1e-05, "loss": 0.4802, "mean_token_accuracy": 0.8783973455429077, "num_tokens": 880259607.0, "step": 23185 }, { "epoch": 2.9494975193995674, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.673371315002441, "learning_rate": 1e-05, "loss": 0.5119, "mean_token_accuracy": 0.8688770532608032, "num_tokens": 880301238.0, "step": 23186 }, { "epoch": 2.949624729678158, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 9.965896606445312e-05, "grad_norm": 6.622263431549072, "learning_rate": 1e-05, "loss": 0.4989, "mean_token_accuracy": 0.8690980672836304, "num_tokens": 880336449.0, "step": 23187 }, { "epoch": 2.9497519399567484, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.519344329833984, "learning_rate": 1e-05, "loss": 0.496, "mean_token_accuracy": 0.8713973760604858, "num_tokens": 880373527.0, "step": 23188 }, { "epoch": 2.949879150235339, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.564781665802002, "learning_rate": 1e-05, "loss": 0.478, "mean_token_accuracy": 0.8775131702423096, "num_tokens": 880412676.0, "step": 23189 }, { "epoch": 2.9500063605139295, "ewc_loss": 0.125, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.37637996673584, "learning_rate": 1e-05, "loss": 0.4675, "mean_token_accuracy": 0.8801811933517456, "num_tokens": 880448171.0, "step": 23190 }, { "epoch": 2.95013357079252, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.596639633178711, "learning_rate": 1e-05, "loss": 0.4901, "mean_token_accuracy": 0.87446528673172, "num_tokens": 880487056.0, "step": 23191 }, { "epoch": 2.9502607810711106, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 6.21997594833374, "learning_rate": 1e-05, "loss": 0.4193, "mean_token_accuracy": 0.8961834907531738, "num_tokens": 880520990.0, "step": 23192 }, { "epoch": 2.950387991349701, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.722303867340088, "learning_rate": 1e-05, "loss": 0.5203, "mean_token_accuracy": 0.864676833152771, "num_tokens": 880557919.0, "step": 23193 }, { "epoch": 2.9505152016282916, "ewc_loss": 0.11962890625, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 6.038650989532471, "learning_rate": 1e-05, "loss": 0.4406, "mean_token_accuracy": 0.8861802816390991, "num_tokens": 880590578.0, "step": 23194 }, { "epoch": 2.950642411906882, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 7.013775825500488, "learning_rate": 1e-05, "loss": 0.4823, "mean_token_accuracy": 0.8777377009391785, "num_tokens": 880631607.0, "step": 23195 }, { "epoch": 2.9507696221854727, "ewc_loss": 0.1162109375, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 8.58306884765625e-05, "grad_norm": 5.840485572814941, "learning_rate": 1e-05, "loss": 0.4568, "mean_token_accuracy": 0.8808480501174927, "num_tokens": 880674301.0, "step": 23196 }, { "epoch": 2.950896832464063, "ewc_loss": 0.1484375, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.000118255615234375, "grad_norm": 7.467979907989502, "learning_rate": 1e-05, "loss": 0.5088, "mean_token_accuracy": 0.8749755024909973, "num_tokens": 880714254.0, "step": 23197 }, { "epoch": 2.9510240427426537, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 5.812922954559326, "learning_rate": 1e-05, "loss": 0.4097, "mean_token_accuracy": 0.8958415985107422, "num_tokens": 880751562.0, "step": 23198 }, { "epoch": 2.9511512530212443, "ewc_loss": 0.166015625, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.0001354217529296875, "grad_norm": 8.072429656982422, "learning_rate": 1e-05, "loss": 0.5529, "mean_token_accuracy": 0.8654083609580994, "num_tokens": 880785062.0, "step": 23199 }, { "epoch": 2.951278463299835, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.31666374206543, "learning_rate": 1e-05, "loss": 0.4992, "mean_token_accuracy": 0.8723391890525818, "num_tokens": 880820673.0, "step": 23200 }, { "epoch": 2.9514056735784253, "ewc_loss": 0.1669921875, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00013637542724609375, "grad_norm": 8.213278770446777, "learning_rate": 1e-05, "loss": 0.5319, "mean_token_accuracy": 0.872458815574646, "num_tokens": 880860707.0, "step": 23201 }, { "epoch": 2.9515328838570154, "ewc_loss": 0.150390625, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00012063980102539062, "grad_norm": 6.970057487487793, "learning_rate": 1e-05, "loss": 0.4904, "mean_token_accuracy": 0.8798549175262451, "num_tokens": 880903504.0, "step": 23202 }, { "epoch": 2.9516600941356064, "ewc_loss": 0.1552734375, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00012493133544921875, "grad_norm": 7.717681407928467, "learning_rate": 1e-05, "loss": 0.469, "mean_token_accuracy": 0.8884433507919312, "num_tokens": 880943414.0, "step": 23203 }, { "epoch": 2.9517873044141965, "ewc_loss": 0.1484375, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00011873245239257812, "grad_norm": 7.092060089111328, "learning_rate": 1e-05, "loss": 0.488, "mean_token_accuracy": 0.8841490745544434, "num_tokens": 880977736.0, "step": 23204 }, { "epoch": 2.9519145146927874, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.0001125335693359375, "grad_norm": 7.178709030151367, "learning_rate": 1e-05, "loss": 0.4477, "mean_token_accuracy": 0.8920901417732239, "num_tokens": 881019793.0, "step": 23205 }, { "epoch": 2.9520417249713775, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.0001125335693359375, "grad_norm": 6.952390193939209, "learning_rate": 1e-05, "loss": 0.4625, "mean_token_accuracy": 0.8893402814865112, "num_tokens": 881057703.0, "step": 23206 }, { "epoch": 2.952168935249968, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.772867679595947, "learning_rate": 1e-05, "loss": 0.468, "mean_token_accuracy": 0.882995069026947, "num_tokens": 881093649.0, "step": 23207 }, { "epoch": 2.9522961455285586, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.85704231262207, "learning_rate": 1e-05, "loss": 0.5443, "mean_token_accuracy": 0.8579867482185364, "num_tokens": 881125620.0, "step": 23208 }, { "epoch": 2.952423355807149, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.489162921905518, "learning_rate": 1e-05, "loss": 0.4409, "mean_token_accuracy": 0.8897026181221008, "num_tokens": 881165313.0, "step": 23209 }, { "epoch": 2.9525505660857396, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.79770565032959, "learning_rate": 1e-05, "loss": 0.4733, "mean_token_accuracy": 0.8791341781616211, "num_tokens": 881195390.0, "step": 23210 }, { "epoch": 2.95267777636433, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.307967662811279, "learning_rate": 1e-05, "loss": 0.5021, "mean_token_accuracy": 0.8693192005157471, "num_tokens": 881242005.0, "step": 23211 }, { "epoch": 2.9528049866429207, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.7945027351379395, "learning_rate": 1e-05, "loss": 0.4792, "mean_token_accuracy": 0.8813778758049011, "num_tokens": 881284587.0, "step": 23212 }, { "epoch": 2.952932196921511, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.177586078643799, "learning_rate": 1e-05, "loss": 0.4643, "mean_token_accuracy": 0.8819173574447632, "num_tokens": 881324174.0, "step": 23213 }, { "epoch": 2.9530594072001017, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00010538101196289062, "grad_norm": 6.887454032897949, "learning_rate": 1e-05, "loss": 0.5413, "mean_token_accuracy": 0.8587713241577148, "num_tokens": 881360727.0, "step": 23214 }, { "epoch": 2.9531866174786923, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.052680015563965, "learning_rate": 1e-05, "loss": 0.4718, "mean_token_accuracy": 0.8783200979232788, "num_tokens": 881400482.0, "step": 23215 }, { "epoch": 2.953313827757283, "ewc_loss": 0.140625, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 7.081926345825195, "learning_rate": 1e-05, "loss": 0.495, "mean_token_accuracy": 0.8764874339103699, "num_tokens": 881440157.0, "step": 23216 }, { "epoch": 2.9534410380358733, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 5.922612190246582, "learning_rate": 1e-05, "loss": 0.4485, "mean_token_accuracy": 0.8854767680168152, "num_tokens": 881486814.0, "step": 23217 }, { "epoch": 2.953568248314464, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00011682510375976562, "grad_norm": 7.421621322631836, "learning_rate": 1e-05, "loss": 0.5173, "mean_token_accuracy": 0.8719732761383057, "num_tokens": 881526997.0, "step": 23218 }, { "epoch": 2.9536954585930544, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 5.931190013885498, "learning_rate": 1e-05, "loss": 0.4676, "mean_token_accuracy": 0.878085732460022, "num_tokens": 881569103.0, "step": 23219 }, { "epoch": 2.953822668871645, "ewc_loss": 0.16015625, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00012969970703125, "grad_norm": 7.802165508270264, "learning_rate": 1e-05, "loss": 0.4911, "mean_token_accuracy": 0.8820714354515076, "num_tokens": 881604734.0, "step": 23220 }, { "epoch": 2.9539498791502354, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.212839126586914, "learning_rate": 1e-05, "loss": 0.4779, "mean_token_accuracy": 0.8788482546806335, "num_tokens": 881645788.0, "step": 23221 }, { "epoch": 2.954077089428826, "ewc_loss": 0.162109375, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.0001316070556640625, "grad_norm": 7.973597526550293, "learning_rate": 1e-05, "loss": 0.4715, "mean_token_accuracy": 0.8889989852905273, "num_tokens": 881674738.0, "step": 23222 }, { "epoch": 2.9542042997074165, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.00011348724365234375, "grad_norm": 6.675556659698486, "learning_rate": 1e-05, "loss": 0.5093, "mean_token_accuracy": 0.8763266801834106, "num_tokens": 881711685.0, "step": 23223 }, { "epoch": 2.954331509986007, "ewc_loss": 0.154296875, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.0001239776611328125, "grad_norm": 7.670473575592041, "learning_rate": 1e-05, "loss": 0.4609, "mean_token_accuracy": 0.8918318748474121, "num_tokens": 881748165.0, "step": 23224 }, { "epoch": 2.954458720264597, "ewc_loss": 0.1455078125, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00011539459228515625, "grad_norm": 6.860087871551514, "learning_rate": 1e-05, "loss": 0.4807, "mean_token_accuracy": 0.8837612867355347, "num_tokens": 881781332.0, "step": 23225 }, { "epoch": 2.954585930543188, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00011444091796875, "grad_norm": 7.222543239593506, "learning_rate": 1e-05, "loss": 0.4905, "mean_token_accuracy": 0.8783827424049377, "num_tokens": 881814526.0, "step": 23226 }, { "epoch": 2.954713140821778, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00011205673217773438, "grad_norm": 6.822154998779297, "learning_rate": 1e-05, "loss": 0.5374, "mean_token_accuracy": 0.8657785058021545, "num_tokens": 881856625.0, "step": 23227 }, { "epoch": 2.954840351100369, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.792420387268066, "learning_rate": 1e-05, "loss": 0.474, "mean_token_accuracy": 0.8838481903076172, "num_tokens": 881895092.0, "step": 23228 }, { "epoch": 2.954967561378959, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.778346061706543, "learning_rate": 1e-05, "loss": 0.4859, "mean_token_accuracy": 0.8787959814071655, "num_tokens": 881936789.0, "step": 23229 }, { "epoch": 2.95509477165755, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.470478057861328, "learning_rate": 1e-05, "loss": 0.4594, "mean_token_accuracy": 0.8868235349655151, "num_tokens": 881973471.0, "step": 23230 }, { "epoch": 2.9552219819361403, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.771973133087158, "learning_rate": 1e-05, "loss": 0.497, "mean_token_accuracy": 0.8725458383560181, "num_tokens": 882011039.0, "step": 23231 }, { "epoch": 2.955349192214731, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.271561145782471, "learning_rate": 1e-05, "loss": 0.468, "mean_token_accuracy": 0.8780102729797363, "num_tokens": 882049221.0, "step": 23232 }, { "epoch": 2.9554764024933213, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.803779602050781, "learning_rate": 1e-05, "loss": 0.4642, "mean_token_accuracy": 0.8831825256347656, "num_tokens": 882086527.0, "step": 23233 }, { "epoch": 2.955603612771912, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.1334333419799805, "learning_rate": 1e-05, "loss": 0.4053, "mean_token_accuracy": 0.8981623649597168, "num_tokens": 882118375.0, "step": 23234 }, { "epoch": 2.9557308230505024, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.919877052307129, "learning_rate": 1e-05, "loss": 0.4927, "mean_token_accuracy": 0.8745799660682678, "num_tokens": 882150604.0, "step": 23235 }, { "epoch": 2.955858033329093, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 5.997185230255127, "learning_rate": 1e-05, "loss": 0.4697, "mean_token_accuracy": 0.8763709664344788, "num_tokens": 882185822.0, "step": 23236 }, { "epoch": 2.9559852436076834, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00011301040649414062, "grad_norm": 7.144455909729004, "learning_rate": 1e-05, "loss": 0.4919, "mean_token_accuracy": 0.8799865245819092, "num_tokens": 882224829.0, "step": 23237 }, { "epoch": 2.956112453886274, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 5.902858257293701, "learning_rate": 1e-05, "loss": 0.5155, "mean_token_accuracy": 0.8595479726791382, "num_tokens": 882263107.0, "step": 23238 }, { "epoch": 2.9562396641648645, "ewc_loss": 0.150390625, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00012063980102539062, "grad_norm": 7.499660968780518, "learning_rate": 1e-05, "loss": 0.4994, "mean_token_accuracy": 0.8784114122390747, "num_tokens": 882294302.0, "step": 23239 }, { "epoch": 2.956366874443455, "ewc_loss": 0.12255859375, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 5.974915027618408, "learning_rate": 1e-05, "loss": 0.4563, "mean_token_accuracy": 0.8810432553291321, "num_tokens": 882327081.0, "step": 23240 }, { "epoch": 2.9564940847220456, "ewc_loss": 0.16015625, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00012969970703125, "grad_norm": 7.833684921264648, "learning_rate": 1e-05, "loss": 0.4975, "mean_token_accuracy": 0.8820905685424805, "num_tokens": 882361793.0, "step": 23241 }, { "epoch": 2.956621295000636, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.307243824005127, "learning_rate": 1e-05, "loss": 0.428, "mean_token_accuracy": 0.8941763639450073, "num_tokens": 882395255.0, "step": 23242 }, { "epoch": 2.9567485052792266, "ewc_loss": 0.1591796875, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00012874603271484375, "grad_norm": 7.870380401611328, "learning_rate": 1e-05, "loss": 0.5084, "mean_token_accuracy": 0.8786569237709045, "num_tokens": 882431915.0, "step": 23243 }, { "epoch": 2.956875715557817, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00011396408081054688, "grad_norm": 6.691880702972412, "learning_rate": 1e-05, "loss": 0.4972, "mean_token_accuracy": 0.8746832609176636, "num_tokens": 882472688.0, "step": 23244 }, { "epoch": 2.9570029258364077, "ewc_loss": 0.1513671875, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00012111663818359375, "grad_norm": 7.542867183685303, "learning_rate": 1e-05, "loss": 0.4949, "mean_token_accuracy": 0.8788127899169922, "num_tokens": 882506851.0, "step": 23245 }, { "epoch": 2.957130136114998, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00011444091796875, "grad_norm": 6.831487655639648, "learning_rate": 1e-05, "loss": 0.535, "mean_token_accuracy": 0.8698058128356934, "num_tokens": 882541935.0, "step": 23246 }, { "epoch": 2.9572573463935887, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00011205673217773438, "grad_norm": 7.093879699707031, "learning_rate": 1e-05, "loss": 0.4729, "mean_token_accuracy": 0.8826186656951904, "num_tokens": 882577860.0, "step": 23247 }, { "epoch": 2.9573845566721793, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 6.786214351654053, "learning_rate": 1e-05, "loss": 0.5406, "mean_token_accuracy": 0.8621968030929565, "num_tokens": 882616182.0, "step": 23248 }, { "epoch": 2.95751176695077, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.7027201652526855, "learning_rate": 1e-05, "loss": 0.5437, "mean_token_accuracy": 0.8589514493942261, "num_tokens": 882658847.0, "step": 23249 }, { "epoch": 2.95763897722936, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.739681720733643, "learning_rate": 1e-05, "loss": 0.4945, "mean_token_accuracy": 0.874367356300354, "num_tokens": 882699842.0, "step": 23250 }, { "epoch": 2.957766187507951, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.431647300720215, "learning_rate": 1e-05, "loss": 0.4501, "mean_token_accuracy": 0.8846722841262817, "num_tokens": 882733143.0, "step": 23251 }, { "epoch": 2.957893397786541, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.740788459777832, "learning_rate": 1e-05, "loss": 0.4896, "mean_token_accuracy": 0.8728628754615784, "num_tokens": 882763233.0, "step": 23252 }, { "epoch": 2.958020608065132, "ewc_loss": 0.125, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.255770206451416, "learning_rate": 1e-05, "loss": 0.4748, "mean_token_accuracy": 0.877539873123169, "num_tokens": 882807612.0, "step": 23253 }, { "epoch": 2.958147818343722, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.751195907592773, "learning_rate": 1e-05, "loss": 0.4172, "mean_token_accuracy": 0.8975268006324768, "num_tokens": 882842135.0, "step": 23254 }, { "epoch": 2.958275028622313, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.137972354888916, "learning_rate": 1e-05, "loss": 0.4562, "mean_token_accuracy": 0.882943868637085, "num_tokens": 882876687.0, "step": 23255 }, { "epoch": 2.958402238900903, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.847565650939941, "learning_rate": 1e-05, "loss": 0.4955, "mean_token_accuracy": 0.8732342720031738, "num_tokens": 882919142.0, "step": 23256 }, { "epoch": 2.9585294491794936, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 6.0045623779296875, "learning_rate": 1e-05, "loss": 0.4523, "mean_token_accuracy": 0.8858940005302429, "num_tokens": 882955301.0, "step": 23257 }, { "epoch": 2.958656659458084, "ewc_loss": 0.140625, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00010967254638671875, "grad_norm": 7.0555925369262695, "learning_rate": 1e-05, "loss": 0.4492, "mean_token_accuracy": 0.8913769721984863, "num_tokens": 882985436.0, "step": 23258 }, { "epoch": 2.9587838697366746, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 5.886252403259277, "learning_rate": 1e-05, "loss": 0.4924, "mean_token_accuracy": 0.8739705085754395, "num_tokens": 883028082.0, "step": 23259 }, { "epoch": 2.958911080015265, "ewc_loss": 0.1484375, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.000118255615234375, "grad_norm": 7.419999599456787, "learning_rate": 1e-05, "loss": 0.5069, "mean_token_accuracy": 0.875684916973114, "num_tokens": 883063771.0, "step": 23260 }, { "epoch": 2.9590382902938557, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 5.890707969665527, "learning_rate": 1e-05, "loss": 0.4496, "mean_token_accuracy": 0.8806492686271667, "num_tokens": 883096742.0, "step": 23261 }, { "epoch": 2.959165500572446, "ewc_loss": 0.162109375, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.0001316070556640625, "grad_norm": 7.8428802490234375, "learning_rate": 1e-05, "loss": 0.5276, "mean_token_accuracy": 0.8746006488800049, "num_tokens": 883137916.0, "step": 23262 }, { "epoch": 2.9592927108510367, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00010013580322265625, "grad_norm": 6.203141212463379, "learning_rate": 1e-05, "loss": 0.4972, "mean_token_accuracy": 0.8728135824203491, "num_tokens": 883172730.0, "step": 23263 }, { "epoch": 2.9594199211296273, "ewc_loss": 0.1650390625, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00013446807861328125, "grad_norm": 8.016437530517578, "learning_rate": 1e-05, "loss": 0.5234, "mean_token_accuracy": 0.8759133219718933, "num_tokens": 883212139.0, "step": 23264 }, { "epoch": 2.959547131408218, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.00011396408081054688, "grad_norm": 6.7001214027404785, "learning_rate": 1e-05, "loss": 0.4724, "mean_token_accuracy": 0.8856500387191772, "num_tokens": 883244866.0, "step": 23265 }, { "epoch": 2.9596743416868083, "ewc_loss": 0.1572265625, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00012683868408203125, "grad_norm": 7.7203288078308105, "learning_rate": 1e-05, "loss": 0.5093, "mean_token_accuracy": 0.8818617463111877, "num_tokens": 883277595.0, "step": 23266 }, { "epoch": 2.959801551965399, "ewc_loss": 0.1484375, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.000118255615234375, "grad_norm": 6.899264812469482, "learning_rate": 1e-05, "loss": 0.5047, "mean_token_accuracy": 0.8760095834732056, "num_tokens": 883311246.0, "step": 23267 }, { "epoch": 2.9599287622439894, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00011491775512695312, "grad_norm": 7.254181861877441, "learning_rate": 1e-05, "loss": 0.5021, "mean_token_accuracy": 0.8766919374465942, "num_tokens": 883353351.0, "step": 23268 }, { "epoch": 2.96005597252258, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.0001125335693359375, "grad_norm": 6.8376383781433105, "learning_rate": 1e-05, "loss": 0.4789, "mean_token_accuracy": 0.8819979429244995, "num_tokens": 883392245.0, "step": 23269 }, { "epoch": 2.9601831828011704, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.831003189086914, "learning_rate": 1e-05, "loss": 0.4675, "mean_token_accuracy": 0.8819438815116882, "num_tokens": 883430404.0, "step": 23270 }, { "epoch": 2.960310393079761, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.77548885345459, "learning_rate": 1e-05, "loss": 0.5594, "mean_token_accuracy": 0.8570350408554077, "num_tokens": 883477894.0, "step": 23271 }, { "epoch": 2.9604376033583515, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.528379917144775, "learning_rate": 1e-05, "loss": 0.5247, "mean_token_accuracy": 0.8615099191665649, "num_tokens": 883520396.0, "step": 23272 }, { "epoch": 2.960564813636942, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.743098735809326, "learning_rate": 1e-05, "loss": 0.499, "mean_token_accuracy": 0.8754037618637085, "num_tokens": 883559120.0, "step": 23273 }, { "epoch": 2.9606920239155325, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.331637382507324, "learning_rate": 1e-05, "loss": 0.4895, "mean_token_accuracy": 0.8729371428489685, "num_tokens": 883603061.0, "step": 23274 }, { "epoch": 2.9608192341941226, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.739152908325195, "learning_rate": 1e-05, "loss": 0.4877, "mean_token_accuracy": 0.8752212524414062, "num_tokens": 883637674.0, "step": 23275 }, { "epoch": 2.9609464444727136, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.204528331756592, "learning_rate": 1e-05, "loss": 0.4556, "mean_token_accuracy": 0.8827348947525024, "num_tokens": 883674631.0, "step": 23276 }, { "epoch": 2.9610736547513037, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.792568206787109, "learning_rate": 1e-05, "loss": 0.4898, "mean_token_accuracy": 0.8746322393417358, "num_tokens": 883716823.0, "step": 23277 }, { "epoch": 2.9612008650298947, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.08743953704834, "learning_rate": 1e-05, "loss": 0.4917, "mean_token_accuracy": 0.8751277923583984, "num_tokens": 883751849.0, "step": 23278 }, { "epoch": 2.9613280753084847, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.935268402099609, "learning_rate": 1e-05, "loss": 0.4594, "mean_token_accuracy": 0.886759877204895, "num_tokens": 883790698.0, "step": 23279 }, { "epoch": 2.9614552855870753, "ewc_loss": 0.11767578125, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 5.934664249420166, "learning_rate": 1e-05, "loss": 0.4239, "mean_token_accuracy": 0.8912408351898193, "num_tokens": 883823238.0, "step": 23280 }, { "epoch": 2.961582495865666, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00011396408081054688, "grad_norm": 7.230149745941162, "learning_rate": 1e-05, "loss": 0.4881, "mean_token_accuracy": 0.8789545297622681, "num_tokens": 883863036.0, "step": 23281 }, { "epoch": 2.9617097061442563, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 5.841690540313721, "learning_rate": 1e-05, "loss": 0.4919, "mean_token_accuracy": 0.866051435470581, "num_tokens": 883902556.0, "step": 23282 }, { "epoch": 2.961836916422847, "ewc_loss": 0.15625, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.000125885009765625, "grad_norm": 7.6958770751953125, "learning_rate": 1e-05, "loss": 0.5436, "mean_token_accuracy": 0.8625028729438782, "num_tokens": 883937039.0, "step": 23283 }, { "epoch": 2.9619641267014374, "ewc_loss": 0.125, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.011924743652344, "learning_rate": 1e-05, "loss": 0.4766, "mean_token_accuracy": 0.8760662078857422, "num_tokens": 883969047.0, "step": 23284 }, { "epoch": 2.962091336980028, "ewc_loss": 0.166015625, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.0001354217529296875, "grad_norm": 8.039586067199707, "learning_rate": 1e-05, "loss": 0.5376, "mean_token_accuracy": 0.8720850944519043, "num_tokens": 884013431.0, "step": 23285 }, { "epoch": 2.9622185472586184, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.0001125335693359375, "grad_norm": 6.529091835021973, "learning_rate": 1e-05, "loss": 0.5003, "mean_token_accuracy": 0.8733325004577637, "num_tokens": 884053714.0, "step": 23286 }, { "epoch": 2.962345757537209, "ewc_loss": 0.1611328125, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00013065338134765625, "grad_norm": 7.90460205078125, "learning_rate": 1e-05, "loss": 0.5331, "mean_token_accuracy": 0.8702015280723572, "num_tokens": 884094624.0, "step": 23287 }, { "epoch": 2.9624729678157995, "ewc_loss": 0.1484375, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.000118255615234375, "grad_norm": 6.899743556976318, "learning_rate": 1e-05, "loss": 0.5055, "mean_token_accuracy": 0.8752419948577881, "num_tokens": 884131759.0, "step": 23288 }, { "epoch": 2.96260017809439, "ewc_loss": 0.1494140625, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00011920928955078125, "grad_norm": 7.434065818786621, "learning_rate": 1e-05, "loss": 0.4872, "mean_token_accuracy": 0.8820717334747314, "num_tokens": 884175333.0, "step": 23289 }, { "epoch": 2.9627273883729806, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00011444091796875, "grad_norm": 6.898850917816162, "learning_rate": 1e-05, "loss": 0.453, "mean_token_accuracy": 0.8909279704093933, "num_tokens": 884209778.0, "step": 23290 }, { "epoch": 2.962854598651571, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 6.991859436035156, "learning_rate": 1e-05, "loss": 0.486, "mean_token_accuracy": 0.8775177001953125, "num_tokens": 884245581.0, "step": 23291 }, { "epoch": 2.9629818089301616, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 6.806573867797852, "learning_rate": 1e-05, "loss": 0.4604, "mean_token_accuracy": 0.8863062262535095, "num_tokens": 884284925.0, "step": 23292 }, { "epoch": 2.963109019208752, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.6432671546936035, "learning_rate": 1e-05, "loss": 0.4668, "mean_token_accuracy": 0.8837979435920715, "num_tokens": 884326056.0, "step": 23293 }, { "epoch": 2.9632362294873427, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.73536491394043, "learning_rate": 1e-05, "loss": 0.4773, "mean_token_accuracy": 0.8781806826591492, "num_tokens": 884368169.0, "step": 23294 }, { "epoch": 2.963363439765933, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.4257283210754395, "learning_rate": 1e-05, "loss": 0.4903, "mean_token_accuracy": 0.8776030540466309, "num_tokens": 884402785.0, "step": 23295 }, { "epoch": 2.9634906500445237, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.698006629943848, "learning_rate": 1e-05, "loss": 0.4677, "mean_token_accuracy": 0.8792113065719604, "num_tokens": 884444755.0, "step": 23296 }, { "epoch": 2.9636178603231143, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.268378734588623, "learning_rate": 1e-05, "loss": 0.4647, "mean_token_accuracy": 0.8816818594932556, "num_tokens": 884486299.0, "step": 23297 }, { "epoch": 2.963745070601705, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.793905735015869, "learning_rate": 1e-05, "loss": 0.5457, "mean_token_accuracy": 0.8567174673080444, "num_tokens": 884532076.0, "step": 23298 }, { "epoch": 2.9638722808802953, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.114454746246338, "learning_rate": 1e-05, "loss": 0.4796, "mean_token_accuracy": 0.8782317042350769, "num_tokens": 884569518.0, "step": 23299 }, { "epoch": 2.9639994911588854, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.9083147048950195, "learning_rate": 1e-05, "loss": 0.4762, "mean_token_accuracy": 0.879445493221283, "num_tokens": 884601173.0, "step": 23300 }, { "epoch": 2.9641267014374764, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 5.949677467346191, "learning_rate": 1e-05, "loss": 0.4607, "mean_token_accuracy": 0.8806004524230957, "num_tokens": 884637558.0, "step": 23301 }, { "epoch": 2.9642539117160664, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00011301040649414062, "grad_norm": 7.2016119956970215, "learning_rate": 1e-05, "loss": 0.4912, "mean_token_accuracy": 0.8790216445922852, "num_tokens": 884673377.0, "step": 23302 }, { "epoch": 2.9643811219946574, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 5.8393025398254395, "learning_rate": 1e-05, "loss": 0.4289, "mean_token_accuracy": 0.8924881815910339, "num_tokens": 884710334.0, "step": 23303 }, { "epoch": 2.9645083322732475, "ewc_loss": 0.154296875, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.0001239776611328125, "grad_norm": 7.647052764892578, "learning_rate": 1e-05, "loss": 0.4699, "mean_token_accuracy": 0.8901834487915039, "num_tokens": 884741515.0, "step": 23304 }, { "epoch": 2.964635542551838, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 5.963658332824707, "learning_rate": 1e-05, "loss": 0.4974, "mean_token_accuracy": 0.8697085380554199, "num_tokens": 884782204.0, "step": 23305 }, { "epoch": 2.9647627528304286, "ewc_loss": 0.166015625, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.0001354217529296875, "grad_norm": 8.060540199279785, "learning_rate": 1e-05, "loss": 0.4882, "mean_token_accuracy": 0.8870750665664673, "num_tokens": 884819613.0, "step": 23306 }, { "epoch": 2.964889963109019, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00011157989501953125, "grad_norm": 6.482271194458008, "learning_rate": 1e-05, "loss": 0.5033, "mean_token_accuracy": 0.8758860230445862, "num_tokens": 884857381.0, "step": 23307 }, { "epoch": 2.9650171733876096, "ewc_loss": 0.1630859375, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00013256072998046875, "grad_norm": 7.993622303009033, "learning_rate": 1e-05, "loss": 0.5789, "mean_token_accuracy": 0.8550107479095459, "num_tokens": 884893562.0, "step": 23308 }, { "epoch": 2.9651443836662, "ewc_loss": 0.1494140625, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00011920928955078125, "grad_norm": 6.908022403717041, "learning_rate": 1e-05, "loss": 0.5079, "mean_token_accuracy": 0.8756126165390015, "num_tokens": 884934359.0, "step": 23309 }, { "epoch": 2.9652715939447907, "ewc_loss": 0.1513671875, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00012111663818359375, "grad_norm": 7.529688358306885, "learning_rate": 1e-05, "loss": 0.4684, "mean_token_accuracy": 0.8900367021560669, "num_tokens": 884974446.0, "step": 23310 }, { "epoch": 2.965398804223381, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.0001163482666015625, "grad_norm": 6.933394432067871, "learning_rate": 1e-05, "loss": 0.5216, "mean_token_accuracy": 0.8701035976409912, "num_tokens": 885016447.0, "step": 23311 }, { "epoch": 2.9655260145019717, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00011205673217773438, "grad_norm": 7.077712059020996, "learning_rate": 1e-05, "loss": 0.5119, "mean_token_accuracy": 0.870813250541687, "num_tokens": 885055064.0, "step": 23312 }, { "epoch": 2.9656532247805623, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 6.832984924316406, "learning_rate": 1e-05, "loss": 0.4879, "mean_token_accuracy": 0.8800840973854065, "num_tokens": 885092380.0, "step": 23313 }, { "epoch": 2.965780435059153, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.715344429016113, "learning_rate": 1e-05, "loss": 0.4595, "mean_token_accuracy": 0.8839014768600464, "num_tokens": 885130717.0, "step": 23314 }, { "epoch": 2.9659076453377433, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.74540901184082, "learning_rate": 1e-05, "loss": 0.4838, "mean_token_accuracy": 0.8804134130477905, "num_tokens": 885168291.0, "step": 23315 }, { "epoch": 2.966034855616334, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.4916205406188965, "learning_rate": 1e-05, "loss": 0.5381, "mean_token_accuracy": 0.8599735498428345, "num_tokens": 885203839.0, "step": 23316 }, { "epoch": 2.9661620658949244, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.6799211502075195, "learning_rate": 1e-05, "loss": 0.5044, "mean_token_accuracy": 0.8683939576148987, "num_tokens": 885239336.0, "step": 23317 }, { "epoch": 2.966289276173515, "ewc_loss": 0.125, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.349534511566162, "learning_rate": 1e-05, "loss": 0.5199, "mean_token_accuracy": 0.8585494756698608, "num_tokens": 885275964.0, "step": 23318 }, { "epoch": 2.9664164864521054, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.647250652313232, "learning_rate": 1e-05, "loss": 0.4791, "mean_token_accuracy": 0.876937210559845, "num_tokens": 885307444.0, "step": 23319 }, { "epoch": 2.966543696730696, "ewc_loss": 0.12353515625, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.2301025390625, "learning_rate": 1e-05, "loss": 0.4428, "mean_token_accuracy": 0.8885167837142944, "num_tokens": 885345379.0, "step": 23320 }, { "epoch": 2.9666709070092865, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.675834655761719, "learning_rate": 1e-05, "loss": 0.4614, "mean_token_accuracy": 0.88343346118927, "num_tokens": 885384105.0, "step": 23321 }, { "epoch": 2.966798117287877, "ewc_loss": 0.12158203125, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.111327171325684, "learning_rate": 1e-05, "loss": 0.475, "mean_token_accuracy": 0.875913143157959, "num_tokens": 885420374.0, "step": 23322 }, { "epoch": 2.966925327566467, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.800572872161865, "learning_rate": 1e-05, "loss": 0.4844, "mean_token_accuracy": 0.8804166316986084, "num_tokens": 885453450.0, "step": 23323 }, { "epoch": 2.967052537845058, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 5.985278606414795, "learning_rate": 1e-05, "loss": 0.4721, "mean_token_accuracy": 0.8812050819396973, "num_tokens": 885486331.0, "step": 23324 }, { "epoch": 2.967179748123648, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 7.039758205413818, "learning_rate": 1e-05, "loss": 0.4617, "mean_token_accuracy": 0.8903822302818298, "num_tokens": 885520474.0, "step": 23325 }, { "epoch": 2.967306958402239, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.8246259689331055, "learning_rate": 1e-05, "loss": 0.4865, "mean_token_accuracy": 0.874194324016571, "num_tokens": 885563128.0, "step": 23326 }, { "epoch": 2.967434168680829, "ewc_loss": 0.150390625, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00011968612670898438, "grad_norm": 7.47959041595459, "learning_rate": 1e-05, "loss": 0.4452, "mean_token_accuracy": 0.8985212445259094, "num_tokens": 885604219.0, "step": 23327 }, { "epoch": 2.96756137895942, "ewc_loss": 0.12060546875, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 5.813689231872559, "learning_rate": 1e-05, "loss": 0.4624, "mean_token_accuracy": 0.8832101821899414, "num_tokens": 885639932.0, "step": 23328 }, { "epoch": 2.9676885892380103, "ewc_loss": 0.1650390625, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00013446807861328125, "grad_norm": 8.028966903686523, "learning_rate": 1e-05, "loss": 0.551, "mean_token_accuracy": 0.8636974692344666, "num_tokens": 885675914.0, "step": 23329 }, { "epoch": 2.967815799516601, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.261906147003174, "learning_rate": 1e-05, "loss": 0.4685, "mean_token_accuracy": 0.8842897415161133, "num_tokens": 885711910.0, "step": 23330 }, { "epoch": 2.9679430097951913, "ewc_loss": 0.1689453125, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00013828277587890625, "grad_norm": 8.211336135864258, "learning_rate": 1e-05, "loss": 0.4884, "mean_token_accuracy": 0.8863064050674438, "num_tokens": 885751615.0, "step": 23331 }, { "epoch": 2.968070220073782, "ewc_loss": 0.15234375, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00012159347534179688, "grad_norm": 6.891918182373047, "learning_rate": 1e-05, "loss": 0.4868, "mean_token_accuracy": 0.8831215500831604, "num_tokens": 885788394.0, "step": 23332 }, { "epoch": 2.9681974303523724, "ewc_loss": 0.158203125, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.0001277923583984375, "grad_norm": 7.799700736999512, "learning_rate": 1e-05, "loss": 0.4845, "mean_token_accuracy": 0.8886333703994751, "num_tokens": 885820462.0, "step": 23333 }, { "epoch": 2.968324640630963, "ewc_loss": 0.1513671875, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00012111663818359375, "grad_norm": 7.0524702072143555, "learning_rate": 1e-05, "loss": 0.4826, "mean_token_accuracy": 0.8860963582992554, "num_tokens": 885853714.0, "step": 23334 }, { "epoch": 2.9684518509095534, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00011491775512695312, "grad_norm": 7.313805103302002, "learning_rate": 1e-05, "loss": 0.493, "mean_token_accuracy": 0.878892719745636, "num_tokens": 885887508.0, "step": 23335 }, { "epoch": 2.968579061188144, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00011348724365234375, "grad_norm": 6.911249160766602, "learning_rate": 1e-05, "loss": 0.5031, "mean_token_accuracy": 0.8725615739822388, "num_tokens": 885923314.0, "step": 23336 }, { "epoch": 2.9687062714667345, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.909336090087891, "learning_rate": 1e-05, "loss": 0.4528, "mean_token_accuracy": 0.8895763754844666, "num_tokens": 885964466.0, "step": 23337 }, { "epoch": 2.968833481745325, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.768948078155518, "learning_rate": 1e-05, "loss": 0.4909, "mean_token_accuracy": 0.8753950595855713, "num_tokens": 886001483.0, "step": 23338 }, { "epoch": 2.9689606920239155, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.664731979370117, "learning_rate": 1e-05, "loss": 0.4791, "mean_token_accuracy": 0.880391001701355, "num_tokens": 886040130.0, "step": 23339 }, { "epoch": 2.969087902302506, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.64874792098999, "learning_rate": 1e-05, "loss": 0.4833, "mean_token_accuracy": 0.878097653388977, "num_tokens": 886076291.0, "step": 23340 }, { "epoch": 2.9692151125810966, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.546380996704102, "learning_rate": 1e-05, "loss": 0.5293, "mean_token_accuracy": 0.8638640642166138, "num_tokens": 886109430.0, "step": 23341 }, { "epoch": 2.969342322859687, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.522421836853027, "learning_rate": 1e-05, "loss": 0.4695, "mean_token_accuracy": 0.8818434476852417, "num_tokens": 886147158.0, "step": 23342 }, { "epoch": 2.9694695331382777, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.46427059173584, "learning_rate": 1e-05, "loss": 0.4828, "mean_token_accuracy": 0.8789352178573608, "num_tokens": 886184810.0, "step": 23343 }, { "epoch": 2.969596743416868, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.432398319244385, "learning_rate": 1e-05, "loss": 0.4616, "mean_token_accuracy": 0.8811612725257874, "num_tokens": 886225511.0, "step": 23344 }, { "epoch": 2.9697239536954587, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.4380035400390625, "learning_rate": 1e-05, "loss": 0.459, "mean_token_accuracy": 0.8803202509880066, "num_tokens": 886265185.0, "step": 23345 }, { "epoch": 2.9698511639740492, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.339285850524902, "learning_rate": 1e-05, "loss": 0.4811, "mean_token_accuracy": 0.8777724504470825, "num_tokens": 886304897.0, "step": 23346 }, { "epoch": 2.9699783742526398, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 9.632110595703125e-05, "grad_norm": 6.476848602294922, "learning_rate": 1e-05, "loss": 0.5042, "mean_token_accuracy": 0.8687622547149658, "num_tokens": 886343097.0, "step": 23347 }, { "epoch": 2.97010558453123, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.240640163421631, "learning_rate": 1e-05, "loss": 0.4815, "mean_token_accuracy": 0.875131368637085, "num_tokens": 886377428.0, "step": 23348 }, { "epoch": 2.970232794809821, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.559606075286865, "learning_rate": 1e-05, "loss": 0.4906, "mean_token_accuracy": 0.8705538511276245, "num_tokens": 886418344.0, "step": 23349 }, { "epoch": 2.970360005088411, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 6.109150409698486, "learning_rate": 1e-05, "loss": 0.5479, "mean_token_accuracy": 0.8585613369941711, "num_tokens": 886458983.0, "step": 23350 }, { "epoch": 2.970487215367002, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.71769380569458, "learning_rate": 1e-05, "loss": 0.4357, "mean_token_accuracy": 0.8914753198623657, "num_tokens": 886492174.0, "step": 23351 }, { "epoch": 2.970614425645592, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 5.922857284545898, "learning_rate": 1e-05, "loss": 0.5081, "mean_token_accuracy": 0.8658705949783325, "num_tokens": 886537145.0, "step": 23352 }, { "epoch": 2.970741635924183, "ewc_loss": 0.140625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.00010967254638671875, "grad_norm": 7.055920124053955, "learning_rate": 1e-05, "loss": 0.5403, "mean_token_accuracy": 0.8664889931678772, "num_tokens": 886573420.0, "step": 23353 }, { "epoch": 2.970868846202773, "ewc_loss": 0.115234375, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 8.487701416015625e-05, "grad_norm": 5.7067179679870605, "learning_rate": 1e-05, "loss": 0.4602, "mean_token_accuracy": 0.8790979385375977, "num_tokens": 886610489.0, "step": 23354 }, { "epoch": 2.9709960564813636, "ewc_loss": 0.154296875, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.0001239776611328125, "grad_norm": 7.669959545135498, "learning_rate": 1e-05, "loss": 0.4699, "mean_token_accuracy": 0.8862864971160889, "num_tokens": 886646954.0, "step": 23355 }, { "epoch": 2.971123266759954, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 5.720896244049072, "learning_rate": 1e-05, "loss": 0.4341, "mean_token_accuracy": 0.8878477811813354, "num_tokens": 886686979.0, "step": 23356 }, { "epoch": 2.9712504770385446, "ewc_loss": 0.1748046875, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.00014495849609375, "grad_norm": 8.478561401367188, "learning_rate": 1e-05, "loss": 0.6419, "mean_token_accuracy": 0.8492878079414368, "num_tokens": 886719734.0, "step": 23357 }, { "epoch": 2.971377687317135, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.0001163482666015625, "grad_norm": 6.557271480560303, "learning_rate": 1e-05, "loss": 0.4824, "mean_token_accuracy": 0.881616473197937, "num_tokens": 886754762.0, "step": 23358 }, { "epoch": 2.9715048975957257, "ewc_loss": 0.17578125, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.00014591217041015625, "grad_norm": 8.510286331176758, "learning_rate": 1e-05, "loss": 0.4985, "mean_token_accuracy": 0.8836233615875244, "num_tokens": 886794830.0, "step": 23359 }, { "epoch": 2.971632107874316, "ewc_loss": 0.162109375, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.0001316070556640625, "grad_norm": 7.360695838928223, "learning_rate": 1e-05, "loss": 0.527, "mean_token_accuracy": 0.872774600982666, "num_tokens": 886834952.0, "step": 23360 }, { "epoch": 2.9717593181529067, "ewc_loss": 0.1591796875, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.00012874603271484375, "grad_norm": 7.854081630706787, "learning_rate": 1e-05, "loss": 0.5466, "mean_token_accuracy": 0.8628032207489014, "num_tokens": 886874236.0, "step": 23361 }, { "epoch": 2.9718865284314973, "ewc_loss": 0.1552734375, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.00012493133544921875, "grad_norm": 7.280294895172119, "learning_rate": 1e-05, "loss": 0.5301, "mean_token_accuracy": 0.866828203201294, "num_tokens": 886907723.0, "step": 23362 }, { "epoch": 2.972013738710088, "ewc_loss": 0.1474609375, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.00011730194091796875, "grad_norm": 7.3998918533325195, "learning_rate": 1e-05, "loss": 0.5172, "mean_token_accuracy": 0.8742902278900146, "num_tokens": 886939914.0, "step": 23363 }, { "epoch": 2.9721409489886783, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.00011301040649414062, "grad_norm": 6.966467380523682, "learning_rate": 1e-05, "loss": 0.5302, "mean_token_accuracy": 0.8664906620979309, "num_tokens": 886970533.0, "step": 23364 }, { "epoch": 2.972268159267269, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.00011110305786132812, "grad_norm": 7.098201274871826, "learning_rate": 1e-05, "loss": 0.5329, "mean_token_accuracy": 0.8630144000053406, "num_tokens": 887012103.0, "step": 23365 }, { "epoch": 2.9723953695458594, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.691082000732422, "learning_rate": 1e-05, "loss": 0.5219, "mean_token_accuracy": 0.8647444844245911, "num_tokens": 887055789.0, "step": 23366 }, { "epoch": 2.97252257982445, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.936835289001465, "learning_rate": 1e-05, "loss": 0.4547, "mean_token_accuracy": 0.8860302567481995, "num_tokens": 887094505.0, "step": 23367 }, { "epoch": 2.9726497901030404, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.465178966522217, "learning_rate": 1e-05, "loss": 0.4504, "mean_token_accuracy": 0.8887923359870911, "num_tokens": 887131009.0, "step": 23368 }, { "epoch": 2.972777000381631, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.913451671600342, "learning_rate": 1e-05, "loss": 0.438, "mean_token_accuracy": 0.8906450271606445, "num_tokens": 887170969.0, "step": 23369 }, { "epoch": 2.9729042106602215, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.270391941070557, "learning_rate": 1e-05, "loss": 0.4415, "mean_token_accuracy": 0.8865326642990112, "num_tokens": 887205123.0, "step": 23370 }, { "epoch": 2.973031420938812, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 7.053483963012695, "learning_rate": 1e-05, "loss": 0.4548, "mean_token_accuracy": 0.8877248764038086, "num_tokens": 887235908.0, "step": 23371 }, { "epoch": 2.9731586312174025, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.069125652313232, "learning_rate": 1e-05, "loss": 0.4427, "mean_token_accuracy": 0.8848844170570374, "num_tokens": 887273929.0, "step": 23372 }, { "epoch": 2.9732858414959926, "ewc_loss": 0.1455078125, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.00011539459228515625, "grad_norm": 7.375954627990723, "learning_rate": 1e-05, "loss": 0.4819, "mean_token_accuracy": 0.8816008567810059, "num_tokens": 887309379.0, "step": 23373 }, { "epoch": 2.9734130517745836, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 8.916854858398438e-05, "grad_norm": 6.009237766265869, "learning_rate": 1e-05, "loss": 0.5143, "mean_token_accuracy": 0.8693197965621948, "num_tokens": 887346108.0, "step": 23374 }, { "epoch": 2.9735402620531737, "ewc_loss": 0.15625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.000125885009765625, "grad_norm": 7.793212890625, "learning_rate": 1e-05, "loss": 0.4754, "mean_token_accuracy": 0.8905117511749268, "num_tokens": 887380066.0, "step": 23375 }, { "epoch": 2.9736674723317646, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.2818732261657715, "learning_rate": 1e-05, "loss": 0.4868, "mean_token_accuracy": 0.8745458722114563, "num_tokens": 887414043.0, "step": 23376 }, { "epoch": 2.9737946826103547, "ewc_loss": 0.1591796875, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.00012874603271484375, "grad_norm": 7.949860095977783, "learning_rate": 1e-05, "loss": 0.5203, "mean_token_accuracy": 0.8734220266342163, "num_tokens": 887451921.0, "step": 23377 }, { "epoch": 2.9739218928889453, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.0001125335693359375, "grad_norm": 6.779961585998535, "learning_rate": 1e-05, "loss": 0.5043, "mean_token_accuracy": 0.879761815071106, "num_tokens": 887483977.0, "step": 23378 }, { "epoch": 2.974049103167536, "ewc_loss": 0.1494140625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.00011873245239257812, "grad_norm": 7.562041282653809, "learning_rate": 1e-05, "loss": 0.4851, "mean_token_accuracy": 0.8813813924789429, "num_tokens": 887516795.0, "step": 23379 }, { "epoch": 2.9741763134461263, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.00011444091796875, "grad_norm": 6.9728193283081055, "learning_rate": 1e-05, "loss": 0.5046, "mean_token_accuracy": 0.8721579313278198, "num_tokens": 887554791.0, "step": 23380 }, { "epoch": 2.974303523724717, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.97266149520874, "learning_rate": 1e-05, "loss": 0.4795, "mean_token_accuracy": 0.8830313682556152, "num_tokens": 887598451.0, "step": 23381 }, { "epoch": 2.9744307340033074, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 6.9939069747924805, "learning_rate": 1e-05, "loss": 0.452, "mean_token_accuracy": 0.8929804563522339, "num_tokens": 887633242.0, "step": 23382 }, { "epoch": 2.974557944281898, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 9.393692016601562e-05, "grad_norm": 6.442996025085449, "learning_rate": 1e-05, "loss": 0.4494, "mean_token_accuracy": 0.8838770389556885, "num_tokens": 887665084.0, "step": 23383 }, { "epoch": 2.9746851545604884, "ewc_loss": 0.140625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.00011014938354492188, "grad_norm": 7.098809719085693, "learning_rate": 1e-05, "loss": 0.5789, "mean_token_accuracy": 0.8512853384017944, "num_tokens": 887706086.0, "step": 23384 }, { "epoch": 2.974812364839079, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 6.063418865203857, "learning_rate": 1e-05, "loss": 0.4943, "mean_token_accuracy": 0.8691854476928711, "num_tokens": 887747395.0, "step": 23385 }, { "epoch": 2.9749395751176695, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.0001163482666015625, "grad_norm": 7.406811714172363, "learning_rate": 1e-05, "loss": 0.5283, "mean_token_accuracy": 0.8690671920776367, "num_tokens": 887789834.0, "step": 23386 }, { "epoch": 2.97506678539626, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 5.943539619445801, "learning_rate": 1e-05, "loss": 0.5222, "mean_token_accuracy": 0.8602110147476196, "num_tokens": 887831416.0, "step": 23387 }, { "epoch": 2.9751939956748505, "ewc_loss": 0.158203125, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.0001277923583984375, "grad_norm": 7.829329490661621, "learning_rate": 1e-05, "loss": 0.5229, "mean_token_accuracy": 0.8669325113296509, "num_tokens": 887869781.0, "step": 23388 }, { "epoch": 2.975321205953441, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.225947856903076, "learning_rate": 1e-05, "loss": 0.4797, "mean_token_accuracy": 0.8795748353004456, "num_tokens": 887906844.0, "step": 23389 }, { "epoch": 2.9754484162320316, "ewc_loss": 0.162109375, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.0001316070556640625, "grad_norm": 8.027027130126953, "learning_rate": 1e-05, "loss": 0.543, "mean_token_accuracy": 0.8686505556106567, "num_tokens": 887939514.0, "step": 23390 }, { "epoch": 2.975575626510622, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.0001125335693359375, "grad_norm": 6.726556777954102, "learning_rate": 1e-05, "loss": 0.5389, "mean_token_accuracy": 0.8629922270774841, "num_tokens": 887976041.0, "step": 23391 }, { "epoch": 2.9757028367892127, "ewc_loss": 0.15625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.000125885009765625, "grad_norm": 7.745354652404785, "learning_rate": 1e-05, "loss": 0.5569, "mean_token_accuracy": 0.8645759224891663, "num_tokens": 888010007.0, "step": 23392 }, { "epoch": 2.975830047067803, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.00011444091796875, "grad_norm": 6.908368110656738, "learning_rate": 1e-05, "loss": 0.481, "mean_token_accuracy": 0.8789365291595459, "num_tokens": 888045654.0, "step": 23393 }, { "epoch": 2.9759572573463937, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.00011587142944335938, "grad_norm": 7.321326732635498, "learning_rate": 1e-05, "loss": 0.4861, "mean_token_accuracy": 0.8813971281051636, "num_tokens": 888084075.0, "step": 23394 }, { "epoch": 2.9760844676249842, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 6.829322338104248, "learning_rate": 1e-05, "loss": 0.5052, "mean_token_accuracy": 0.8747915029525757, "num_tokens": 888124891.0, "step": 23395 }, { "epoch": 2.9762116779035748, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.9848103523254395, "learning_rate": 1e-05, "loss": 0.4724, "mean_token_accuracy": 0.8843272924423218, "num_tokens": 888161078.0, "step": 23396 }, { "epoch": 2.9763388881821653, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.6648640632629395, "learning_rate": 1e-05, "loss": 0.4641, "mean_token_accuracy": 0.8818304538726807, "num_tokens": 888197013.0, "step": 23397 }, { "epoch": 2.9764660984607554, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.823840618133545, "learning_rate": 1e-05, "loss": 0.4984, "mean_token_accuracy": 0.8712563514709473, "num_tokens": 888231676.0, "step": 23398 }, { "epoch": 2.9765933087393464, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.473623752593994, "learning_rate": 1e-05, "loss": 0.4595, "mean_token_accuracy": 0.8824107050895691, "num_tokens": 888272676.0, "step": 23399 }, { "epoch": 2.9767205190179364, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.815954685211182, "learning_rate": 1e-05, "loss": 0.4642, "mean_token_accuracy": 0.8822195529937744, "num_tokens": 888310064.0, "step": 23400 }, { "epoch": 2.9768477292965274, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.277896404266357, "learning_rate": 1e-05, "loss": 0.4701, "mean_token_accuracy": 0.879146933555603, "num_tokens": 888344103.0, "step": 23401 }, { "epoch": 2.9769749395751175, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.922147274017334, "learning_rate": 1e-05, "loss": 0.5112, "mean_token_accuracy": 0.8712029457092285, "num_tokens": 888379284.0, "step": 23402 }, { "epoch": 2.977102149853708, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.095442295074463, "learning_rate": 1e-05, "loss": 0.4472, "mean_token_accuracy": 0.8866889476776123, "num_tokens": 888414058.0, "step": 23403 }, { "epoch": 2.9772293601322986, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.00011110305786132812, "grad_norm": 7.1894354820251465, "learning_rate": 1e-05, "loss": 0.5188, "mean_token_accuracy": 0.8662725687026978, "num_tokens": 888454756.0, "step": 23404 }, { "epoch": 2.977356570410889, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 8.821487426757812e-05, "grad_norm": 5.976347923278809, "learning_rate": 1e-05, "loss": 0.4676, "mean_token_accuracy": 0.8791590332984924, "num_tokens": 888491995.0, "step": 23405 }, { "epoch": 2.9774837806894796, "ewc_loss": 0.150390625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.0001201629638671875, "grad_norm": 7.579324722290039, "learning_rate": 1e-05, "loss": 0.5696, "mean_token_accuracy": 0.8587247133255005, "num_tokens": 888533073.0, "step": 23406 }, { "epoch": 2.97761099096807, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.1072821617126465, "learning_rate": 1e-05, "loss": 0.4084, "mean_token_accuracy": 0.8981143236160278, "num_tokens": 888570392.0, "step": 23407 }, { "epoch": 2.9777382012466607, "ewc_loss": 0.158203125, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.0001277923583984375, "grad_norm": 7.837193489074707, "learning_rate": 1e-05, "loss": 0.5096, "mean_token_accuracy": 0.87779700756073, "num_tokens": 888612078.0, "step": 23408 }, { "epoch": 2.977865411525251, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.53185510635376, "learning_rate": 1e-05, "loss": 0.4702, "mean_token_accuracy": 0.8847547173500061, "num_tokens": 888653603.0, "step": 23409 }, { "epoch": 2.9779926218038417, "ewc_loss": 0.15234375, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.00012159347534179688, "grad_norm": 7.639659404754639, "learning_rate": 1e-05, "loss": 0.5334, "mean_token_accuracy": 0.8691282272338867, "num_tokens": 888687089.0, "step": 23410 }, { "epoch": 2.9781198320824323, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.00011110305786132812, "grad_norm": 6.8402533531188965, "learning_rate": 1e-05, "loss": 0.5189, "mean_token_accuracy": 0.8635329008102417, "num_tokens": 888732221.0, "step": 23411 }, { "epoch": 2.978247042361023, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.00011110305786132812, "grad_norm": 7.102849006652832, "learning_rate": 1e-05, "loss": 0.448, "mean_token_accuracy": 0.8915956020355225, "num_tokens": 888768992.0, "step": 23412 }, { "epoch": 2.9783742526396133, "ewc_loss": 0.140625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.00011014938354492188, "grad_norm": 6.914456367492676, "learning_rate": 1e-05, "loss": 0.4639, "mean_token_accuracy": 0.8869454860687256, "num_tokens": 888809834.0, "step": 23413 }, { "epoch": 2.978501462918204, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.536234378814697, "learning_rate": 1e-05, "loss": 0.4615, "mean_token_accuracy": 0.8810727596282959, "num_tokens": 888851943.0, "step": 23414 }, { "epoch": 2.9786286731967944, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 7.028965473175049, "learning_rate": 1e-05, "loss": 0.4769, "mean_token_accuracy": 0.8802760243415833, "num_tokens": 888890465.0, "step": 23415 }, { "epoch": 2.978755883475385, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 6.095376014709473, "learning_rate": 1e-05, "loss": 0.4844, "mean_token_accuracy": 0.8756043314933777, "num_tokens": 888932954.0, "step": 23416 }, { "epoch": 2.9788830937539754, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.00011444091796875, "grad_norm": 7.336884498596191, "learning_rate": 1e-05, "loss": 0.4929, "mean_token_accuracy": 0.8789098262786865, "num_tokens": 888973916.0, "step": 23417 }, { "epoch": 2.979010304032566, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 8.630752563476562e-05, "grad_norm": 5.9255547523498535, "learning_rate": 1e-05, "loss": 0.5254, "mean_token_accuracy": 0.8580145835876465, "num_tokens": 889017202.0, "step": 23418 }, { "epoch": 2.9791375143111565, "ewc_loss": 0.1552734375, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.00012493133544921875, "grad_norm": 7.754199981689453, "learning_rate": 1e-05, "loss": 0.5596, "mean_token_accuracy": 0.8622891902923584, "num_tokens": 889053604.0, "step": 23419 }, { "epoch": 2.979264724589747, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.157872676849365, "learning_rate": 1e-05, "loss": 0.4782, "mean_token_accuracy": 0.8770170211791992, "num_tokens": 889087328.0, "step": 23420 }, { "epoch": 2.979391934868337, "ewc_loss": 0.1611328125, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.00013065338134765625, "grad_norm": 7.9737162590026855, "learning_rate": 1e-05, "loss": 0.5358, "mean_token_accuracy": 0.8712356686592102, "num_tokens": 889125209.0, "step": 23421 }, { "epoch": 2.979519145146928, "ewc_loss": 0.140625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.00010967254638671875, "grad_norm": 6.711911678314209, "learning_rate": 1e-05, "loss": 0.5139, "mean_token_accuracy": 0.8723636269569397, "num_tokens": 889160968.0, "step": 23422 }, { "epoch": 2.979646355425518, "ewc_loss": 0.1533203125, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.645920276641846, "learning_rate": 1e-05, "loss": 0.5398, "mean_token_accuracy": 0.8695859313011169, "num_tokens": 889201710.0, "step": 23423 }, { "epoch": 2.979773565704109, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.00011110305786132812, "grad_norm": 6.813396453857422, "learning_rate": 1e-05, "loss": 0.557, "mean_token_accuracy": 0.8616462349891663, "num_tokens": 889237003.0, "step": 23424 }, { "epoch": 2.979900775982699, "ewc_loss": 0.1455078125, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.00011491775512695312, "grad_norm": 7.276190757751465, "learning_rate": 1e-05, "loss": 0.5311, "mean_token_accuracy": 0.8673571348190308, "num_tokens": 889278775.0, "step": 23425 }, { "epoch": 2.98002798626129, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.7307305335998535, "learning_rate": 1e-05, "loss": 0.5236, "mean_token_accuracy": 0.8698608875274658, "num_tokens": 889320888.0, "step": 23426 }, { "epoch": 2.9801551965398803, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.986057758331299, "learning_rate": 1e-05, "loss": 0.4938, "mean_token_accuracy": 0.8786134719848633, "num_tokens": 889356746.0, "step": 23427 }, { "epoch": 2.980282406818471, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.588546276092529, "learning_rate": 1e-05, "loss": 0.4659, "mean_token_accuracy": 0.8845310211181641, "num_tokens": 889400771.0, "step": 23428 }, { "epoch": 2.9804096170970613, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.846028804779053, "learning_rate": 1e-05, "loss": 0.5116, "mean_token_accuracy": 0.8701021075248718, "num_tokens": 889440459.0, "step": 23429 }, { "epoch": 2.980536827375652, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.402233123779297, "learning_rate": 1e-05, "loss": 0.4679, "mean_token_accuracy": 0.8836535215377808, "num_tokens": 889479782.0, "step": 23430 }, { "epoch": 2.9806640376542424, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.859187126159668, "learning_rate": 1e-05, "loss": 0.4314, "mean_token_accuracy": 0.8941787481307983, "num_tokens": 889520167.0, "step": 23431 }, { "epoch": 2.980791247932833, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.220550537109375, "learning_rate": 1e-05, "loss": 0.4749, "mean_token_accuracy": 0.8795653581619263, "num_tokens": 889567589.0, "step": 23432 }, { "epoch": 2.9809184582114234, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 7.0071892738342285, "learning_rate": 1e-05, "loss": 0.5, "mean_token_accuracy": 0.8768427968025208, "num_tokens": 889607703.0, "step": 23433 }, { "epoch": 2.981045668490014, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.073546886444092, "learning_rate": 1e-05, "loss": 0.4842, "mean_token_accuracy": 0.8718605637550354, "num_tokens": 889645735.0, "step": 23434 }, { "epoch": 2.9811728787686045, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.00011348724365234375, "grad_norm": 7.262495040893555, "learning_rate": 1e-05, "loss": 0.4772, "mean_token_accuracy": 0.8813989162445068, "num_tokens": 889680035.0, "step": 23435 }, { "epoch": 2.981300089047195, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.023748397827148, "learning_rate": 1e-05, "loss": 0.4754, "mean_token_accuracy": 0.8756081461906433, "num_tokens": 889720234.0, "step": 23436 }, { "epoch": 2.9814272993257855, "ewc_loss": 0.1513671875, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.00012111663818359375, "grad_norm": 7.5709147453308105, "learning_rate": 1e-05, "loss": 0.4736, "mean_token_accuracy": 0.8859117031097412, "num_tokens": 889756070.0, "step": 23437 }, { "epoch": 2.981554509604376, "ewc_loss": 0.126953125, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 9.679794311523438e-05, "grad_norm": 6.206273555755615, "learning_rate": 1e-05, "loss": 0.4772, "mean_token_accuracy": 0.8781039714813232, "num_tokens": 889796499.0, "step": 23438 }, { "epoch": 2.9816817198829666, "ewc_loss": 0.1533203125, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.697905540466309, "learning_rate": 1e-05, "loss": 0.5831, "mean_token_accuracy": 0.8576185703277588, "num_tokens": 889830804.0, "step": 23439 }, { "epoch": 2.981808930161557, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.555765628814697, "learning_rate": 1e-05, "loss": 0.5279, "mean_token_accuracy": 0.8640245199203491, "num_tokens": 889865228.0, "step": 23440 }, { "epoch": 2.9819361404401477, "ewc_loss": 0.1474609375, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.00011682510375976562, "grad_norm": 7.401314735412598, "learning_rate": 1e-05, "loss": 0.5044, "mean_token_accuracy": 0.8736066818237305, "num_tokens": 889900780.0, "step": 23441 }, { "epoch": 2.982063350718738, "ewc_loss": 0.140625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.00011014938354492188, "grad_norm": 6.754140853881836, "learning_rate": 1e-05, "loss": 0.5023, "mean_token_accuracy": 0.8773201704025269, "num_tokens": 889944366.0, "step": 23442 }, { "epoch": 2.9821905609973287, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.895083904266357, "learning_rate": 1e-05, "loss": 0.4317, "mean_token_accuracy": 0.8950393199920654, "num_tokens": 889979739.0, "step": 23443 }, { "epoch": 2.9823177712759192, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 6.8432230949401855, "learning_rate": 1e-05, "loss": 0.4758, "mean_token_accuracy": 0.8788912892341614, "num_tokens": 890017288.0, "step": 23444 }, { "epoch": 2.9824449815545098, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.393959999084473, "learning_rate": 1e-05, "loss": 0.4518, "mean_token_accuracy": 0.8817031979560852, "num_tokens": 890050848.0, "step": 23445 }, { "epoch": 2.9825721918331, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 7.009486675262451, "learning_rate": 1e-05, "loss": 0.4839, "mean_token_accuracy": 0.8785958290100098, "num_tokens": 890087985.0, "step": 23446 }, { "epoch": 2.982699402111691, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 6.000131607055664, "learning_rate": 1e-05, "loss": 0.4989, "mean_token_accuracy": 0.8681145310401917, "num_tokens": 890128292.0, "step": 23447 }, { "epoch": 2.982826612390281, "ewc_loss": 0.1455078125, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.00011539459228515625, "grad_norm": 7.348647117614746, "learning_rate": 1e-05, "loss": 0.4905, "mean_token_accuracy": 0.8797950744628906, "num_tokens": 890176130.0, "step": 23448 }, { "epoch": 2.982953822668872, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 5.865268230438232, "learning_rate": 1e-05, "loss": 0.4714, "mean_token_accuracy": 0.8780460357666016, "num_tokens": 890212547.0, "step": 23449 }, { "epoch": 2.983081032947462, "ewc_loss": 0.16015625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.00012969970703125, "grad_norm": 7.81267786026001, "learning_rate": 1e-05, "loss": 0.5391, "mean_token_accuracy": 0.8707941770553589, "num_tokens": 890257423.0, "step": 23450 }, { "epoch": 2.983208243226053, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.154141426086426, "learning_rate": 1e-05, "loss": 0.4857, "mean_token_accuracy": 0.87525475025177, "num_tokens": 890299465.0, "step": 23451 }, { "epoch": 2.983335453504643, "ewc_loss": 0.1640625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.000133514404296875, "grad_norm": 8.016979217529297, "learning_rate": 1e-05, "loss": 0.4923, "mean_token_accuracy": 0.8835522532463074, "num_tokens": 890340010.0, "step": 23452 }, { "epoch": 2.9834626637832335, "ewc_loss": 0.1435546875, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.00011348724365234375, "grad_norm": 6.696226596832275, "learning_rate": 1e-05, "loss": 0.5067, "mean_token_accuracy": 0.8748024702072144, "num_tokens": 890380865.0, "step": 23453 }, { "epoch": 2.983589874061824, "ewc_loss": 0.15625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.000125885009765625, "grad_norm": 7.712287902832031, "learning_rate": 1e-05, "loss": 0.4757, "mean_token_accuracy": 0.8887282609939575, "num_tokens": 890416710.0, "step": 23454 }, { "epoch": 2.9837170843404146, "ewc_loss": 0.1455078125, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.00011491775512695312, "grad_norm": 6.896939277648926, "learning_rate": 1e-05, "loss": 0.4673, "mean_token_accuracy": 0.8881828188896179, "num_tokens": 890452963.0, "step": 23455 }, { "epoch": 2.983844294619005, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.0001163482666015625, "grad_norm": 7.274981498718262, "learning_rate": 1e-05, "loss": 0.4827, "mean_token_accuracy": 0.8818212151527405, "num_tokens": 890493308.0, "step": 23456 }, { "epoch": 2.9839715048975957, "ewc_loss": 0.140625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.00011014938354492188, "grad_norm": 6.829599380493164, "learning_rate": 1e-05, "loss": 0.5147, "mean_token_accuracy": 0.8715066909790039, "num_tokens": 890529279.0, "step": 23457 }, { "epoch": 2.984098715176186, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.913461208343506, "learning_rate": 1e-05, "loss": 0.547, "mean_token_accuracy": 0.8568502068519592, "num_tokens": 890568866.0, "step": 23458 }, { "epoch": 2.9842259254547767, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.6826934814453125, "learning_rate": 1e-05, "loss": 0.4796, "mean_token_accuracy": 0.8802752494812012, "num_tokens": 890613077.0, "step": 23459 }, { "epoch": 2.9843531357333672, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.718178749084473, "learning_rate": 1e-05, "loss": 0.4836, "mean_token_accuracy": 0.8782765865325928, "num_tokens": 890653599.0, "step": 23460 }, { "epoch": 2.9844803460119578, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.507230758666992, "learning_rate": 1e-05, "loss": 0.4793, "mean_token_accuracy": 0.87860107421875, "num_tokens": 890694581.0, "step": 23461 }, { "epoch": 2.9846075562905483, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.688826560974121, "learning_rate": 1e-05, "loss": 0.4546, "mean_token_accuracy": 0.8827646970748901, "num_tokens": 890725912.0, "step": 23462 }, { "epoch": 2.984734766569139, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 9.5367431640625e-05, "grad_norm": 6.3322434425354, "learning_rate": 1e-05, "loss": 0.4893, "mean_token_accuracy": 0.8757486939430237, "num_tokens": 890763028.0, "step": 23463 }, { "epoch": 2.9848619768477294, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.754481315612793, "learning_rate": 1e-05, "loss": 0.4885, "mean_token_accuracy": 0.8772984743118286, "num_tokens": 890803318.0, "step": 23464 }, { "epoch": 2.98498918712632, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.159355640411377, "learning_rate": 1e-05, "loss": 0.4896, "mean_token_accuracy": 0.873705267906189, "num_tokens": 890841405.0, "step": 23465 }, { "epoch": 2.9851163974049104, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.938995838165283, "learning_rate": 1e-05, "loss": 0.4696, "mean_token_accuracy": 0.8810649514198303, "num_tokens": 890884271.0, "step": 23466 }, { "epoch": 2.985243607683501, "ewc_loss": 0.1171875, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 5.977959156036377, "learning_rate": 1e-05, "loss": 0.4442, "mean_token_accuracy": 0.885222315788269, "num_tokens": 890917747.0, "step": 23467 }, { "epoch": 2.9853708179620915, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.00011444091796875, "grad_norm": 7.29109525680542, "learning_rate": 1e-05, "loss": 0.4833, "mean_token_accuracy": 0.8826318979263306, "num_tokens": 890958659.0, "step": 23468 }, { "epoch": 2.985498028240682, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 5.898664951324463, "learning_rate": 1e-05, "loss": 0.4293, "mean_token_accuracy": 0.8915505409240723, "num_tokens": 890993792.0, "step": 23469 }, { "epoch": 2.9856252385192725, "ewc_loss": 0.154296875, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.0001239776611328125, "grad_norm": 7.745382308959961, "learning_rate": 1e-05, "loss": 0.4888, "mean_token_accuracy": 0.8796515464782715, "num_tokens": 891030873.0, "step": 23470 }, { "epoch": 2.9857524487978626, "ewc_loss": 0.1259765625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 9.584426879882812e-05, "grad_norm": 6.140603065490723, "learning_rate": 1e-05, "loss": 0.436, "mean_token_accuracy": 0.8930694460868835, "num_tokens": 891068062.0, "step": 23471 }, { "epoch": 2.9858796590764536, "ewc_loss": 0.1630859375, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.00013256072998046875, "grad_norm": 7.99616813659668, "learning_rate": 1e-05, "loss": 0.5476, "mean_token_accuracy": 0.8678687810897827, "num_tokens": 891108749.0, "step": 23472 }, { "epoch": 2.9860068693550437, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.00011157989501953125, "grad_norm": 6.662350654602051, "learning_rate": 1e-05, "loss": 0.4978, "mean_token_accuracy": 0.8749456405639648, "num_tokens": 891144749.0, "step": 23473 }, { "epoch": 2.9861340796336346, "ewc_loss": 0.1552734375, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.00012493133544921875, "grad_norm": 7.706178665161133, "learning_rate": 1e-05, "loss": 0.5576, "mean_token_accuracy": 0.8642771244049072, "num_tokens": 891180795.0, "step": 23474 }, { "epoch": 2.9862612899122247, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.00011587142944335938, "grad_norm": 6.94183349609375, "learning_rate": 1e-05, "loss": 0.5125, "mean_token_accuracy": 0.8720411062240601, "num_tokens": 891221333.0, "step": 23475 }, { "epoch": 2.9863885001908153, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.0001125335693359375, "grad_norm": 7.152667045593262, "learning_rate": 1e-05, "loss": 0.4942, "mean_token_accuracy": 0.8782040476799011, "num_tokens": 891258414.0, "step": 23476 }, { "epoch": 2.986515710469406, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.0001125335693359375, "grad_norm": 6.941944599151611, "learning_rate": 1e-05, "loss": 0.5258, "mean_token_accuracy": 0.8692532181739807, "num_tokens": 891296864.0, "step": 23477 }, { "epoch": 2.9866429207479963, "ewc_loss": 0.1298828125, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 9.918212890625e-05, "grad_norm": 6.625197410583496, "learning_rate": 1e-05, "loss": 0.4332, "mean_token_accuracy": 0.8933557868003845, "num_tokens": 891338191.0, "step": 23478 }, { "epoch": 2.986770131026587, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 6.971130847930908, "learning_rate": 1e-05, "loss": 0.5042, "mean_token_accuracy": 0.873504638671875, "num_tokens": 891377354.0, "step": 23479 }, { "epoch": 2.9868973413051774, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.253167629241943, "learning_rate": 1e-05, "loss": 0.4635, "mean_token_accuracy": 0.8795591592788696, "num_tokens": 891415868.0, "step": 23480 }, { "epoch": 2.987024551583768, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 7.13662052154541, "learning_rate": 1e-05, "loss": 0.4744, "mean_token_accuracy": 0.8801044225692749, "num_tokens": 891449548.0, "step": 23481 }, { "epoch": 2.9871517618623584, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 6.053833961486816, "learning_rate": 1e-05, "loss": 0.5269, "mean_token_accuracy": 0.8580702543258667, "num_tokens": 891485937.0, "step": 23482 }, { "epoch": 2.987278972140949, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.00011587142944335938, "grad_norm": 7.395498275756836, "learning_rate": 1e-05, "loss": 0.4787, "mean_token_accuracy": 0.8811244368553162, "num_tokens": 891522458.0, "step": 23483 }, { "epoch": 2.9874061824195395, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.03950834274292, "learning_rate": 1e-05, "loss": 0.5119, "mean_token_accuracy": 0.8665363192558289, "num_tokens": 891560804.0, "step": 23484 }, { "epoch": 2.98753339269813, "ewc_loss": 0.1533203125, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.966067314147949, "learning_rate": 1e-05, "loss": 0.473, "mean_token_accuracy": 0.8891150951385498, "num_tokens": 891595692.0, "step": 23485 }, { "epoch": 2.9876606029767205, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.3640851974487305, "learning_rate": 1e-05, "loss": 0.4624, "mean_token_accuracy": 0.886133074760437, "num_tokens": 891632007.0, "step": 23486 }, { "epoch": 2.987787813255311, "ewc_loss": 0.15625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.000125885009765625, "grad_norm": 7.792169094085693, "learning_rate": 1e-05, "loss": 0.4908, "mean_token_accuracy": 0.8830987811088562, "num_tokens": 891671605.0, "step": 23487 }, { "epoch": 2.9879150235339016, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 6.653108596801758, "learning_rate": 1e-05, "loss": 0.5304, "mean_token_accuracy": 0.8660416603088379, "num_tokens": 891708369.0, "step": 23488 }, { "epoch": 2.988042233812492, "ewc_loss": 0.15234375, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.0001220703125, "grad_norm": 7.549666404724121, "learning_rate": 1e-05, "loss": 0.476, "mean_token_accuracy": 0.8845220804214478, "num_tokens": 891744673.0, "step": 23489 }, { "epoch": 2.9881694440910826, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.00011157989501953125, "grad_norm": 6.748345375061035, "learning_rate": 1e-05, "loss": 0.5025, "mean_token_accuracy": 0.874062180519104, "num_tokens": 891781971.0, "step": 23490 }, { "epoch": 2.988296654369673, "ewc_loss": 0.1455078125, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.00011539459228515625, "grad_norm": 7.217126369476318, "learning_rate": 1e-05, "loss": 0.4729, "mean_token_accuracy": 0.886987566947937, "num_tokens": 891814079.0, "step": 23491 }, { "epoch": 2.9884238646482637, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.00010776519775390625, "grad_norm": 6.69984769821167, "learning_rate": 1e-05, "loss": 0.4753, "mean_token_accuracy": 0.883985161781311, "num_tokens": 891852188.0, "step": 23492 }, { "epoch": 2.9885510749268542, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 6.924520015716553, "learning_rate": 1e-05, "loss": 0.499, "mean_token_accuracy": 0.874727725982666, "num_tokens": 891894048.0, "step": 23493 }, { "epoch": 2.9886782852054448, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.584684371948242, "learning_rate": 1e-05, "loss": 0.4895, "mean_token_accuracy": 0.8760331869125366, "num_tokens": 891935199.0, "step": 23494 }, { "epoch": 2.9888054954840353, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.770040035247803, "learning_rate": 1e-05, "loss": 0.4851, "mean_token_accuracy": 0.8783085942268372, "num_tokens": 891977686.0, "step": 23495 }, { "epoch": 2.9889327057626254, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 9.870529174804688e-05, "grad_norm": 6.427201747894287, "learning_rate": 1e-05, "loss": 0.5087, "mean_token_accuracy": 0.868537187576294, "num_tokens": 892017501.0, "step": 23496 }, { "epoch": 2.9890599160412163, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.762718200683594, "learning_rate": 1e-05, "loss": 0.5148, "mean_token_accuracy": 0.8659903407096863, "num_tokens": 892057369.0, "step": 23497 }, { "epoch": 2.9891871263198064, "ewc_loss": 0.125, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 6.266531944274902, "learning_rate": 1e-05, "loss": 0.4785, "mean_token_accuracy": 0.8762477040290833, "num_tokens": 892093998.0, "step": 23498 }, { "epoch": 2.9893143365983974, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.00010728836059570312, "grad_norm": 6.8528571128845215, "learning_rate": 1e-05, "loss": 0.4579, "mean_token_accuracy": 0.8898051977157593, "num_tokens": 892125835.0, "step": 23499 }, { "epoch": 2.9894415468769875, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.120622158050537, "learning_rate": 1e-05, "loss": 0.4822, "mean_token_accuracy": 0.8777961730957031, "num_tokens": 892160340.0, "step": 23500 }, { "epoch": 2.989568757155578, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 7.041350364685059, "learning_rate": 1e-05, "loss": 0.4951, "mean_token_accuracy": 0.8742733001708984, "num_tokens": 892202096.0, "step": 23501 }, { "epoch": 2.9896959674341685, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 5.991684913635254, "learning_rate": 1e-05, "loss": 0.4478, "mean_token_accuracy": 0.8863733410835266, "num_tokens": 892241942.0, "step": 23502 }, { "epoch": 2.989823177712759, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.0001163482666015625, "grad_norm": 7.3555192947387695, "learning_rate": 1e-05, "loss": 0.5295, "mean_token_accuracy": 0.8672621250152588, "num_tokens": 892284724.0, "step": 23503 }, { "epoch": 2.9899503879913496, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.005382061004639, "learning_rate": 1e-05, "loss": 0.4435, "mean_token_accuracy": 0.8891528248786926, "num_tokens": 892324761.0, "step": 23504 }, { "epoch": 2.99007759826994, "ewc_loss": 0.1552734375, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.00012493133544921875, "grad_norm": 7.678013324737549, "learning_rate": 1e-05, "loss": 0.468, "mean_token_accuracy": 0.8878943920135498, "num_tokens": 892358128.0, "step": 23505 }, { "epoch": 2.9902048085485307, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.284394264221191, "learning_rate": 1e-05, "loss": 0.5058, "mean_token_accuracy": 0.8721393346786499, "num_tokens": 892400380.0, "step": 23506 }, { "epoch": 2.990332018827121, "ewc_loss": 0.15625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.000125885009765625, "grad_norm": 7.709624767303467, "learning_rate": 1e-05, "loss": 0.5392, "mean_token_accuracy": 0.8708041906356812, "num_tokens": 892439141.0, "step": 23507 }, { "epoch": 2.9904592291057117, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 6.65793514251709, "learning_rate": 1e-05, "loss": 0.4918, "mean_token_accuracy": 0.8743818998336792, "num_tokens": 892480983.0, "step": 23508 }, { "epoch": 2.9905864393843022, "ewc_loss": 0.146484375, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.00011587142944335938, "grad_norm": 7.318568706512451, "learning_rate": 1e-05, "loss": 0.5108, "mean_token_accuracy": 0.8763891458511353, "num_tokens": 892518220.0, "step": 23509 }, { "epoch": 2.9907136496628928, "ewc_loss": 0.140625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.00010967254638671875, "grad_norm": 6.809515953063965, "learning_rate": 1e-05, "loss": 0.4855, "mean_token_accuracy": 0.8778596520423889, "num_tokens": 892549308.0, "step": 23510 }, { "epoch": 2.9908408599414833, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.00010395050048828125, "grad_norm": 6.7916412353515625, "learning_rate": 1e-05, "loss": 0.4985, "mean_token_accuracy": 0.8756796717643738, "num_tokens": 892590154.0, "step": 23511 }, { "epoch": 2.990968070220074, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 6.886486053466797, "learning_rate": 1e-05, "loss": 0.4383, "mean_token_accuracy": 0.8942533731460571, "num_tokens": 892630951.0, "step": 23512 }, { "epoch": 2.9910952804986644, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 9.250640869140625e-05, "grad_norm": 6.297884941101074, "learning_rate": 1e-05, "loss": 0.4963, "mean_token_accuracy": 0.8714638948440552, "num_tokens": 892675345.0, "step": 23513 }, { "epoch": 2.991222490777255, "ewc_loss": 0.140625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.00011014938354492188, "grad_norm": 7.096504211425781, "learning_rate": 1e-05, "loss": 0.5167, "mean_token_accuracy": 0.8702897429466248, "num_tokens": 892718129.0, "step": 23514 }, { "epoch": 2.9913497010558454, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 5.98118782043457, "learning_rate": 1e-05, "loss": 0.476, "mean_token_accuracy": 0.8767857551574707, "num_tokens": 892757159.0, "step": 23515 }, { "epoch": 2.991476911334436, "ewc_loss": 0.1484375, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.000118255615234375, "grad_norm": 7.447488784790039, "learning_rate": 1e-05, "loss": 0.498, "mean_token_accuracy": 0.8789028525352478, "num_tokens": 892797872.0, "step": 23516 }, { "epoch": 2.9916041216130265, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 9.059906005859375e-05, "grad_norm": 5.955480575561523, "learning_rate": 1e-05, "loss": 0.4953, "mean_token_accuracy": 0.8696799278259277, "num_tokens": 892837213.0, "step": 23517 }, { "epoch": 2.991731331891617, "ewc_loss": 0.16015625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.00012969970703125, "grad_norm": 7.823639392852783, "learning_rate": 1e-05, "loss": 0.5216, "mean_token_accuracy": 0.8718290328979492, "num_tokens": 892878651.0, "step": 23518 }, { "epoch": 2.991858542170207, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.0001010894775390625, "grad_norm": 6.284294128417969, "learning_rate": 1e-05, "loss": 0.4454, "mean_token_accuracy": 0.8852198123931885, "num_tokens": 892913739.0, "step": 23519 }, { "epoch": 2.991985752448798, "ewc_loss": 0.16015625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.00012969970703125, "grad_norm": 7.900963306427002, "learning_rate": 1e-05, "loss": 0.5182, "mean_token_accuracy": 0.8741090893745422, "num_tokens": 892949649.0, "step": 23520 }, { "epoch": 2.992112962727388, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.00011157989501953125, "grad_norm": 6.689999580383301, "learning_rate": 1e-05, "loss": 0.4878, "mean_token_accuracy": 0.8798250555992126, "num_tokens": 892988831.0, "step": 23521 }, { "epoch": 2.992240173005979, "ewc_loss": 0.15234375, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.0001220703125, "grad_norm": 7.594807147979736, "learning_rate": 1e-05, "loss": 0.5029, "mean_token_accuracy": 0.8785877227783203, "num_tokens": 893025311.0, "step": 23522 }, { "epoch": 2.992367383284569, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.00011157989501953125, "grad_norm": 6.808629989624023, "learning_rate": 1e-05, "loss": 0.5153, "mean_token_accuracy": 0.8709297180175781, "num_tokens": 893072461.0, "step": 23523 }, { "epoch": 2.99249459356316, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.00011444091796875, "grad_norm": 7.227005958557129, "learning_rate": 1e-05, "loss": 0.5211, "mean_token_accuracy": 0.8701073527336121, "num_tokens": 893109268.0, "step": 23524 }, { "epoch": 2.9926218038417502, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.737699031829834, "learning_rate": 1e-05, "loss": 0.5232, "mean_token_accuracy": 0.8698724508285522, "num_tokens": 893139033.0, "step": 23525 }, { "epoch": 2.992749014120341, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.00010585784912109375, "grad_norm": 6.918167591094971, "learning_rate": 1e-05, "loss": 0.4634, "mean_token_accuracy": 0.8837538957595825, "num_tokens": 893180056.0, "step": 23526 }, { "epoch": 2.9928762243989313, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.597210884094238, "learning_rate": 1e-05, "loss": 0.4985, "mean_token_accuracy": 0.8737114667892456, "num_tokens": 893214164.0, "step": 23527 }, { "epoch": 2.993003434677522, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.00010347366333007812, "grad_norm": 6.772409439086914, "learning_rate": 1e-05, "loss": 0.4806, "mean_token_accuracy": 0.8780994415283203, "num_tokens": 893255030.0, "step": 23528 }, { "epoch": 2.9931306449561124, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 9.72747802734375e-05, "grad_norm": 6.436008930206299, "learning_rate": 1e-05, "loss": 0.4933, "mean_token_accuracy": 0.8702113628387451, "num_tokens": 893287661.0, "step": 23529 }, { "epoch": 2.993257855234703, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.772076606750488, "learning_rate": 1e-05, "loss": 0.5537, "mean_token_accuracy": 0.854320764541626, "num_tokens": 893328461.0, "step": 23530 }, { "epoch": 2.9933850655132934, "ewc_loss": 0.1240234375, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 9.34600830078125e-05, "grad_norm": 6.246729850769043, "learning_rate": 1e-05, "loss": 0.4604, "mean_token_accuracy": 0.8810974359512329, "num_tokens": 893367543.0, "step": 23531 }, { "epoch": 2.993512275791884, "ewc_loss": 0.134765625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.00010442733764648438, "grad_norm": 6.889366626739502, "learning_rate": 1e-05, "loss": 0.4566, "mean_token_accuracy": 0.8877580165863037, "num_tokens": 893401180.0, "step": 23532 }, { "epoch": 2.9936394860704745, "ewc_loss": 0.12109375, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 9.012222290039062e-05, "grad_norm": 6.08091926574707, "learning_rate": 1e-05, "loss": 0.413, "mean_token_accuracy": 0.8957371115684509, "num_tokens": 893429823.0, "step": 23533 }, { "epoch": 2.993766696349065, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 7.134507179260254, "learning_rate": 1e-05, "loss": 0.4877, "mean_token_accuracy": 0.8755931854248047, "num_tokens": 893472730.0, "step": 23534 }, { "epoch": 2.9938939066276555, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 5.967755317687988, "learning_rate": 1e-05, "loss": 0.4569, "mean_token_accuracy": 0.8802030682563782, "num_tokens": 893516506.0, "step": 23535 }, { "epoch": 2.994021116906246, "ewc_loss": 0.1494140625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.00011873245239257812, "grad_norm": 7.485907554626465, "learning_rate": 1e-05, "loss": 0.5047, "mean_token_accuracy": 0.8761180639266968, "num_tokens": 893560539.0, "step": 23536 }, { "epoch": 2.9941483271848366, "ewc_loss": 0.123046875, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 9.298324584960938e-05, "grad_norm": 6.065730094909668, "learning_rate": 1e-05, "loss": 0.4739, "mean_token_accuracy": 0.8765928745269775, "num_tokens": 893595986.0, "step": 23537 }, { "epoch": 2.994275537463427, "ewc_loss": 0.1572265625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.00012683868408203125, "grad_norm": 7.744157791137695, "learning_rate": 1e-05, "loss": 0.4849, "mean_token_accuracy": 0.8844754695892334, "num_tokens": 893632997.0, "step": 23538 }, { "epoch": 2.9944027477420176, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.406753063201904, "learning_rate": 1e-05, "loss": 0.5029, "mean_token_accuracy": 0.8746496438980103, "num_tokens": 893671889.0, "step": 23539 }, { "epoch": 2.994529958020608, "ewc_loss": 0.154296875, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.0001239776611328125, "grad_norm": 7.635275363922119, "learning_rate": 1e-05, "loss": 0.476, "mean_token_accuracy": 0.886360764503479, "num_tokens": 893708781.0, "step": 23540 }, { "epoch": 2.9946571682991987, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 6.7137017250061035, "learning_rate": 1e-05, "loss": 0.4951, "mean_token_accuracy": 0.8721731901168823, "num_tokens": 893745590.0, "step": 23541 }, { "epoch": 2.9947843785777892, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.00011444091796875, "grad_norm": 7.205050468444824, "learning_rate": 1e-05, "loss": 0.4806, "mean_token_accuracy": 0.8821901082992554, "num_tokens": 893783599.0, "step": 23542 }, { "epoch": 2.9949115888563798, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.00010919570922851562, "grad_norm": 6.82442045211792, "learning_rate": 1e-05, "loss": 0.4352, "mean_token_accuracy": 0.893925130367279, "num_tokens": 893822603.0, "step": 23543 }, { "epoch": 2.99503879913497, "ewc_loss": 0.130859375, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.00010061264038085938, "grad_norm": 6.671535491943359, "learning_rate": 1e-05, "loss": 0.4757, "mean_token_accuracy": 0.8814840316772461, "num_tokens": 893857318.0, "step": 23544 }, { "epoch": 2.995166009413561, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 6.909336566925049, "learning_rate": 1e-05, "loss": 0.4858, "mean_token_accuracy": 0.8747467994689941, "num_tokens": 893896269.0, "step": 23545 }, { "epoch": 2.995293219692151, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 9.202957153320312e-05, "grad_norm": 6.223996639251709, "learning_rate": 1e-05, "loss": 0.5122, "mean_token_accuracy": 0.8636997938156128, "num_tokens": 893936877.0, "step": 23546 }, { "epoch": 2.995420429970742, "ewc_loss": 0.1416015625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.000110626220703125, "grad_norm": 7.137539863586426, "learning_rate": 1e-05, "loss": 0.5216, "mean_token_accuracy": 0.8757803440093994, "num_tokens": 893970014.0, "step": 23547 }, { "epoch": 2.995547640249332, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 8.7738037109375e-05, "grad_norm": 5.96136999130249, "learning_rate": 1e-05, "loss": 0.4692, "mean_token_accuracy": 0.8779021501541138, "num_tokens": 894005568.0, "step": 23548 }, { "epoch": 2.995674850527923, "ewc_loss": 0.1494140625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.00011920928955078125, "grad_norm": 7.498823165893555, "learning_rate": 1e-05, "loss": 0.5051, "mean_token_accuracy": 0.8770570158958435, "num_tokens": 894053880.0, "step": 23549 }, { "epoch": 2.995802060806513, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 9.1552734375e-05, "grad_norm": 5.972341537475586, "learning_rate": 1e-05, "loss": 0.429, "mean_token_accuracy": 0.8907656073570251, "num_tokens": 894083981.0, "step": 23550 }, { "epoch": 2.9959292710851035, "ewc_loss": 0.16015625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.00012969970703125, "grad_norm": 7.860446453094482, "learning_rate": 1e-05, "loss": 0.5069, "mean_token_accuracy": 0.880885124206543, "num_tokens": 894125315.0, "step": 23551 }, { "epoch": 2.996056481363694, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.3339056968688965, "learning_rate": 1e-05, "loss": 0.5035, "mean_token_accuracy": 0.8726838231086731, "num_tokens": 894165195.0, "step": 23552 }, { "epoch": 2.9961836916422846, "ewc_loss": 0.16015625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.00012969970703125, "grad_norm": 7.884614944458008, "learning_rate": 1e-05, "loss": 0.5328, "mean_token_accuracy": 0.8697516322135925, "num_tokens": 894206443.0, "step": 23553 }, { "epoch": 2.996310901920875, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.00011157989501953125, "grad_norm": 6.720956325531006, "learning_rate": 1e-05, "loss": 0.4844, "mean_token_accuracy": 0.8809723854064941, "num_tokens": 894240897.0, "step": 23554 }, { "epoch": 2.9964381121994657, "ewc_loss": 0.1533203125, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.00012302398681640625, "grad_norm": 7.558457374572754, "learning_rate": 1e-05, "loss": 0.4734, "mean_token_accuracy": 0.8877670764923096, "num_tokens": 894277475.0, "step": 23555 }, { "epoch": 2.996565322478056, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.00011205673217773438, "grad_norm": 6.812950611114502, "learning_rate": 1e-05, "loss": 0.5253, "mean_token_accuracy": 0.8668510913848877, "num_tokens": 894320231.0, "step": 23556 }, { "epoch": 2.9966925327566467, "ewc_loss": 0.14453125, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.00011444091796875, "grad_norm": 7.185187339782715, "learning_rate": 1e-05, "loss": 0.4951, "mean_token_accuracy": 0.8776806592941284, "num_tokens": 894359196.0, "step": 23557 }, { "epoch": 2.9968197430352372, "ewc_loss": 0.13671875, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.00010633468627929688, "grad_norm": 6.727563858032227, "learning_rate": 1e-05, "loss": 0.5081, "mean_token_accuracy": 0.8702442646026611, "num_tokens": 894401362.0, "step": 23558 }, { "epoch": 2.9969469533138278, "ewc_loss": 0.1357421875, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.0001049041748046875, "grad_norm": 6.883364677429199, "learning_rate": 1e-05, "loss": 0.4583, "mean_token_accuracy": 0.8849571347236633, "num_tokens": 894435445.0, "step": 23559 }, { "epoch": 2.9970741635924183, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.00010204315185546875, "grad_norm": 6.607568264007568, "learning_rate": 1e-05, "loss": 0.4957, "mean_token_accuracy": 0.8720683455467224, "num_tokens": 894475968.0, "step": 23560 }, { "epoch": 2.997201373871009, "ewc_loss": 0.1328125, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.00010251998901367188, "grad_norm": 6.731816291809082, "learning_rate": 1e-05, "loss": 0.4632, "mean_token_accuracy": 0.8823314309120178, "num_tokens": 894511692.0, "step": 23561 }, { "epoch": 2.9973285841495994, "ewc_loss": 0.12890625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 9.822845458984375e-05, "grad_norm": 6.45832633972168, "learning_rate": 1e-05, "loss": 0.5102, "mean_token_accuracy": 0.8714941740036011, "num_tokens": 894550886.0, "step": 23562 }, { "epoch": 2.99745579442819, "ewc_loss": 0.1318359375, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.00010156631469726562, "grad_norm": 6.700594425201416, "learning_rate": 1e-05, "loss": 0.4492, "mean_token_accuracy": 0.8883222341537476, "num_tokens": 894584811.0, "step": 23563 }, { "epoch": 2.9975830047067804, "ewc_loss": 0.125, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 9.441375732421875e-05, "grad_norm": 6.291189193725586, "learning_rate": 1e-05, "loss": 0.493, "mean_token_accuracy": 0.8731387853622437, "num_tokens": 894622866.0, "step": 23564 }, { "epoch": 2.997710214985371, "ewc_loss": 0.1337890625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.000102996826171875, "grad_norm": 6.785587787628174, "learning_rate": 1e-05, "loss": 0.4503, "mean_token_accuracy": 0.8924210667610168, "num_tokens": 894663513.0, "step": 23565 }, { "epoch": 2.9978374252639615, "ewc_loss": 0.1220703125, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 9.107589721679688e-05, "grad_norm": 6.130539417266846, "learning_rate": 1e-05, "loss": 0.4894, "mean_token_accuracy": 0.8720368146896362, "num_tokens": 894702873.0, "step": 23566 }, { "epoch": 2.997964635542552, "ewc_loss": 0.1376953125, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.0001068115234375, "grad_norm": 6.994378089904785, "learning_rate": 1e-05, "loss": 0.4761, "mean_token_accuracy": 0.8800245523452759, "num_tokens": 894737990.0, "step": 23567 }, { "epoch": 2.9980918458211425, "ewc_loss": 0.1181640625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 8.726119995117188e-05, "grad_norm": 5.9834113121032715, "learning_rate": 1e-05, "loss": 0.4632, "mean_token_accuracy": 0.8818668127059937, "num_tokens": 894775408.0, "step": 23568 }, { "epoch": 2.9982190560997326, "ewc_loss": 0.1455078125, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.00011539459228515625, "grad_norm": 7.322310447692871, "learning_rate": 1e-05, "loss": 0.5095, "mean_token_accuracy": 0.8719329833984375, "num_tokens": 894815702.0, "step": 23569 }, { "epoch": 2.9983462663783236, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 5.959479331970215, "learning_rate": 1e-05, "loss": 0.4685, "mean_token_accuracy": 0.8785632848739624, "num_tokens": 894849116.0, "step": 23570 }, { "epoch": 2.9984734766569137, "ewc_loss": 0.15625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.000125885009765625, "grad_norm": 7.68562650680542, "learning_rate": 1e-05, "loss": 0.4718, "mean_token_accuracy": 0.8865255117416382, "num_tokens": 894882020.0, "step": 23571 }, { "epoch": 2.9986006869355046, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.210912704467773, "learning_rate": 1e-05, "loss": 0.4761, "mean_token_accuracy": 0.8792531490325928, "num_tokens": 894916994.0, "step": 23572 }, { "epoch": 2.9987278972140947, "ewc_loss": 0.158203125, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.0001277923583984375, "grad_norm": 7.813921928405762, "learning_rate": 1e-05, "loss": 0.5343, "mean_token_accuracy": 0.8705546855926514, "num_tokens": 894952763.0, "step": 23573 }, { "epoch": 2.9988551074926852, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 6.62925386428833, "learning_rate": 1e-05, "loss": 0.4848, "mean_token_accuracy": 0.8793808221817017, "num_tokens": 894983427.0, "step": 23574 }, { "epoch": 2.9989823177712758, "ewc_loss": 0.150390625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.0001201629638671875, "grad_norm": 7.487193584442139, "learning_rate": 1e-05, "loss": 0.5537, "mean_token_accuracy": 0.8644642233848572, "num_tokens": 895019082.0, "step": 23575 }, { "epoch": 2.9991095280498663, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.0001125335693359375, "grad_norm": 6.8170905113220215, "learning_rate": 1e-05, "loss": 0.4589, "mean_token_accuracy": 0.889344334602356, "num_tokens": 895054051.0, "step": 23576 }, { "epoch": 2.999236738328457, "ewc_loss": 0.138671875, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.00010824203491210938, "grad_norm": 6.982884407043457, "learning_rate": 1e-05, "loss": 0.4831, "mean_token_accuracy": 0.8780465126037598, "num_tokens": 895090185.0, "step": 23577 }, { "epoch": 2.9993639486070474, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 6.868707656860352, "learning_rate": 1e-05, "loss": 0.5036, "mean_token_accuracy": 0.8759957551956177, "num_tokens": 895130159.0, "step": 23578 }, { "epoch": 2.999491158885638, "ewc_loss": 0.1279296875, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 9.775161743164062e-05, "grad_norm": 6.48787260055542, "learning_rate": 1e-05, "loss": 0.4814, "mean_token_accuracy": 0.8761958479881287, "num_tokens": 895169639.0, "step": 23579 }, { "epoch": 2.9996183691642284, "ewc_loss": 0.1396484375, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.0001087188720703125, "grad_norm": 6.98527717590332, "learning_rate": 1e-05, "loss": 0.4409, "mean_token_accuracy": 0.8929726481437683, "num_tokens": 895209939.0, "step": 23580 }, { "epoch": 2.999745579442819, "ewc_loss": 0.1201171875, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 8.96453857421875e-05, "grad_norm": 6.1177754402160645, "learning_rate": 1e-05, "loss": 0.4911, "mean_token_accuracy": 0.8705339431762695, "num_tokens": 895253073.0, "step": 23581 }, { "epoch": 2.9998727897214095, "ewc_loss": 0.142578125, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 0.0001125335693359375, "grad_norm": 7.241332054138184, "learning_rate": 1e-05, "loss": 0.5054, "mean_token_accuracy": 0.8742351531982422, "num_tokens": 895295058.0, "step": 23582 }, { "epoch": 3.0, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 8.869171142578125e-05, "grad_norm": 5.950682163238525, "learning_rate": 1e-05, "loss": 0.4557, "mean_token_accuracy": 0.8818272948265076, "num_tokens": 895335272.0, "step": 23583 }, { "epoch": 3.0, "ewc_loss": 0.119140625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 8.869171142578125e-05, "step": 23583, "total_flos": 5.156325133756924e+19, "train_loss": 0.5050858836974098, "train_runtime": 50355.4873, "train_samples_per_second": 7.493, "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 }