| { |
| "best_global_step": null, |
| "best_metric": null, |
| "best_model_checkpoint": null, |
| "epoch": 2.0, |
| "eval_steps": 88.0, |
| "global_step": 176, |
| "is_hyper_param_search": false, |
| "is_local_process_zero": true, |
| "is_world_process_zero": true, |
| "log_history": [ |
| { |
| "epoch": 0.011363636363636364, |
| "grad_norm": 195.9052499197566, |
| "learning_rate": 1.111111111111111e-06, |
| "loss": 1.6436803340911865, |
| "step": 1, |
| "token_acc": 0.7368421052631579 |
| }, |
| { |
| "epoch": 0.022727272727272728, |
| "grad_norm": 197.8529757812739, |
| "learning_rate": 2.222222222222222e-06, |
| "loss": 2.2848076820373535, |
| "step": 2, |
| "token_acc": 0.6666666666666666 |
| }, |
| { |
| "epoch": 0.03409090909090909, |
| "grad_norm": 200.12115706482385, |
| "learning_rate": 3.3333333333333333e-06, |
| "loss": 1.9579904079437256, |
| "step": 3, |
| "token_acc": 0.7027027027027027 |
| }, |
| { |
| "epoch": 0.045454545454545456, |
| "grad_norm": 156.1490927498441, |
| "learning_rate": 4.444444444444444e-06, |
| "loss": 1.67714524269104, |
| "step": 4, |
| "token_acc": 0.7894736842105263 |
| }, |
| { |
| "epoch": 0.056818181818181816, |
| "grad_norm": 128.56892038276865, |
| "learning_rate": 5.555555555555557e-06, |
| "loss": 0.5448644757270813, |
| "step": 5, |
| "token_acc": 0.8055555555555556 |
| }, |
| { |
| "epoch": 0.06818181818181818, |
| "grad_norm": 53.96240527920626, |
| "learning_rate": 6.666666666666667e-06, |
| "loss": 0.5201314687728882, |
| "step": 6, |
| "token_acc": 0.9210526315789473 |
| }, |
| { |
| "epoch": 0.07954545454545454, |
| "grad_norm": 58.79727047532702, |
| "learning_rate": 7.77777777777778e-06, |
| "loss": 0.21907192468643188, |
| "step": 7, |
| "token_acc": 0.9523809523809523 |
| }, |
| { |
| "epoch": 0.09090909090909091, |
| "grad_norm": 21.610610913212, |
| "learning_rate": 8.888888888888888e-06, |
| "loss": 0.23302510380744934, |
| "step": 8, |
| "token_acc": 0.9210526315789473 |
| }, |
| { |
| "epoch": 0.10227272727272728, |
| "grad_norm": 48.00136565721641, |
| "learning_rate": 1e-05, |
| "loss": 0.35925620794296265, |
| "step": 9, |
| "token_acc": 0.8947368421052632 |
| }, |
| { |
| "epoch": 0.11363636363636363, |
| "grad_norm": 11.165394060812915, |
| "learning_rate": 9.999115304121459e-06, |
| "loss": 0.16875913739204407, |
| "step": 10, |
| "token_acc": 0.918918918918919 |
| }, |
| { |
| "epoch": 0.125, |
| "grad_norm": 20.67801478310866, |
| "learning_rate": 9.996461529560553e-06, |
| "loss": 0.4627538323402405, |
| "step": 11, |
| "token_acc": 0.8947368421052632 |
| }, |
| { |
| "epoch": 0.13636363636363635, |
| "grad_norm": 14.547090071423504, |
| "learning_rate": 9.992039615430648e-06, |
| "loss": 0.2966996729373932, |
| "step": 12, |
| "token_acc": 0.8378378378378378 |
| }, |
| { |
| "epoch": 0.14772727272727273, |
| "grad_norm": 8.68950614952946, |
| "learning_rate": 9.985851126551428e-06, |
| "loss": 0.17632704973220825, |
| "step": 13, |
| "token_acc": 0.8974358974358975 |
| }, |
| { |
| "epoch": 0.1590909090909091, |
| "grad_norm": 4.809531150727445, |
| "learning_rate": 9.977898252895133e-06, |
| "loss": 0.09936656802892685, |
| "step": 14, |
| "token_acc": 0.95 |
| }, |
| { |
| "epoch": 0.17045454545454544, |
| "grad_norm": 14.570648679905538, |
| "learning_rate": 9.968183808811586e-06, |
| "loss": 0.3718123435974121, |
| "step": 15, |
| "token_acc": 0.9210526315789473 |
| }, |
| { |
| "epoch": 0.18181818181818182, |
| "grad_norm": 13.903576246289267, |
| "learning_rate": 9.95671123203224e-06, |
| "loss": 0.46726879477500916, |
| "step": 16, |
| "token_acc": 0.8888888888888888 |
| }, |
| { |
| "epoch": 0.19318181818181818, |
| "grad_norm": 10.425063032585191, |
| "learning_rate": 9.943484582453653e-06, |
| "loss": 0.3023732304573059, |
| "step": 17, |
| "token_acc": 0.8947368421052632 |
| }, |
| { |
| "epoch": 0.20454545454545456, |
| "grad_norm": 13.968698952157565, |
| "learning_rate": 9.928508540700775e-06, |
| "loss": 0.4589124619960785, |
| "step": 18, |
| "token_acc": 0.8863636363636364 |
| }, |
| { |
| "epoch": 0.2159090909090909, |
| "grad_norm": 9.835347726071497, |
| "learning_rate": 9.91178840647057e-06, |
| "loss": 0.22297608852386475, |
| "step": 19, |
| "token_acc": 0.918918918918919 |
| }, |
| { |
| "epoch": 0.22727272727272727, |
| "grad_norm": 9.579424829686822, |
| "learning_rate": 9.893330096656576e-06, |
| "loss": 0.28242772817611694, |
| "step": 20, |
| "token_acc": 0.9 |
| }, |
| { |
| "epoch": 0.23863636363636365, |
| "grad_norm": 7.881240704755622, |
| "learning_rate": 9.873140143255035e-06, |
| "loss": 0.23008175194263458, |
| "step": 21, |
| "token_acc": 0.918918918918919 |
| }, |
| { |
| "epoch": 0.25, |
| "grad_norm": 10.872886421097451, |
| "learning_rate": 9.851225691053382e-06, |
| "loss": 0.33348631858825684, |
| "step": 22, |
| "token_acc": 0.9024390243902439 |
| }, |
| { |
| "epoch": 0.26136363636363635, |
| "grad_norm": 14.264270357002637, |
| "learning_rate": 9.827594495101824e-06, |
| "loss": 0.5246642827987671, |
| "step": 23, |
| "token_acc": 0.868421052631579 |
| }, |
| { |
| "epoch": 0.2727272727272727, |
| "grad_norm": 9.618701211682444, |
| "learning_rate": 9.802254917969033e-06, |
| "loss": 0.21203476190567017, |
| "step": 24, |
| "token_acc": 0.9210526315789473 |
| }, |
| { |
| "epoch": 0.2840909090909091, |
| "grad_norm": 12.947332773482199, |
| "learning_rate": 9.775215926782788e-06, |
| "loss": 0.4625626802444458, |
| "step": 25, |
| "token_acc": 0.9069767441860465 |
| }, |
| { |
| "epoch": 0.29545454545454547, |
| "grad_norm": 12.458809644831751, |
| "learning_rate": 9.746487090056712e-06, |
| "loss": 0.3603699803352356, |
| "step": 26, |
| "token_acc": 0.875 |
| }, |
| { |
| "epoch": 0.3068181818181818, |
| "grad_norm": 8.931946298485734, |
| "learning_rate": 9.71607857430419e-06, |
| "loss": 0.2773553729057312, |
| "step": 27, |
| "token_acc": 0.9512195121951219 |
| }, |
| { |
| "epoch": 0.3181818181818182, |
| "grad_norm": 11.945759677616323, |
| "learning_rate": 9.68400114044064e-06, |
| "loss": 0.2876451909542084, |
| "step": 28, |
| "token_acc": 0.8918918918918919 |
| }, |
| { |
| "epoch": 0.32954545454545453, |
| "grad_norm": 8.563275162731655, |
| "learning_rate": 9.650266139975474e-06, |
| "loss": 0.23299528658390045, |
| "step": 29, |
| "token_acc": 0.9 |
| }, |
| { |
| "epoch": 0.3409090909090909, |
| "grad_norm": 7.639203814648641, |
| "learning_rate": 9.614885510995047e-06, |
| "loss": 0.2409423291683197, |
| "step": 30, |
| "token_acc": 0.9285714285714286 |
| }, |
| { |
| "epoch": 0.3522727272727273, |
| "grad_norm": 7.594291994556256, |
| "learning_rate": 9.577871773938013e-06, |
| "loss": 0.19314445555210114, |
| "step": 31, |
| "token_acc": 0.9210526315789473 |
| }, |
| { |
| "epoch": 0.36363636363636365, |
| "grad_norm": 10.22139388599444, |
| "learning_rate": 9.539238027164618e-06, |
| "loss": 0.2277001440525055, |
| "step": 32, |
| "token_acc": 0.925 |
| }, |
| { |
| "epoch": 0.375, |
| "grad_norm": 9.825388651759896, |
| "learning_rate": 9.498997942321484e-06, |
| "loss": 0.21605941653251648, |
| "step": 33, |
| "token_acc": 0.8918918918918919 |
| }, |
| { |
| "epoch": 0.38636363636363635, |
| "grad_norm": 8.071349735769445, |
| "learning_rate": 9.457165759503492e-06, |
| "loss": 0.20501622557640076, |
| "step": 34, |
| "token_acc": 0.9 |
| }, |
| { |
| "epoch": 0.3977272727272727, |
| "grad_norm": 4.498644705505135, |
| "learning_rate": 9.413756282214538e-06, |
| "loss": 0.11810928583145142, |
| "step": 35, |
| "token_acc": 0.9545454545454546 |
| }, |
| { |
| "epoch": 0.4090909090909091, |
| "grad_norm": 8.914258150643857, |
| "learning_rate": 9.368784872128877e-06, |
| "loss": 0.18691639602184296, |
| "step": 36, |
| "token_acc": 0.8974358974358975 |
| }, |
| { |
| "epoch": 0.42045454545454547, |
| "grad_norm": 9.393728864095808, |
| "learning_rate": 9.322267443654974e-06, |
| "loss": 0.2205682396888733, |
| "step": 37, |
| "token_acc": 0.918918918918919 |
| }, |
| { |
| "epoch": 0.4318181818181818, |
| "grad_norm": 10.426207266539826, |
| "learning_rate": 9.274220458303727e-06, |
| "loss": 0.3519130349159241, |
| "step": 38, |
| "token_acc": 0.8717948717948718 |
| }, |
| { |
| "epoch": 0.4431818181818182, |
| "grad_norm": 14.34736799438511, |
| "learning_rate": 9.224660918863104e-06, |
| "loss": 0.47112351655960083, |
| "step": 39, |
| "token_acc": 0.8648648648648649 |
| }, |
| { |
| "epoch": 0.45454545454545453, |
| "grad_norm": 15.21652321877103, |
| "learning_rate": 9.173606363381218e-06, |
| "loss": 0.4981301426887512, |
| "step": 40, |
| "token_acc": 0.85 |
| }, |
| { |
| "epoch": 0.4659090909090909, |
| "grad_norm": 8.101928185799775, |
| "learning_rate": 9.121074858959997e-06, |
| "loss": 0.2851291298866272, |
| "step": 41, |
| "token_acc": 0.918918918918919 |
| }, |
| { |
| "epoch": 0.4772727272727273, |
| "grad_norm": 7.246952073824331, |
| "learning_rate": 9.067084995361623e-06, |
| "loss": 0.23519396781921387, |
| "step": 42, |
| "token_acc": 0.9736842105263158 |
| }, |
| { |
| "epoch": 0.48863636363636365, |
| "grad_norm": 6.858716885736816, |
| "learning_rate": 9.011655878430018e-06, |
| "loss": 0.2327650487422943, |
| "step": 43, |
| "token_acc": 0.9210526315789473 |
| }, |
| { |
| "epoch": 0.5, |
| "grad_norm": 5.963237806929174, |
| "learning_rate": 8.954807123329703e-06, |
| "loss": 0.17163364589214325, |
| "step": 44, |
| "token_acc": 0.972972972972973 |
| }, |
| { |
| "epoch": 0.5113636363636364, |
| "grad_norm": 8.927579293527703, |
| "learning_rate": 8.896558847604414e-06, |
| "loss": 0.29938608407974243, |
| "step": 45, |
| "token_acc": 0.918918918918919 |
| }, |
| { |
| "epoch": 0.5227272727272727, |
| "grad_norm": 9.90580188343264, |
| "learning_rate": 8.836931664057935e-06, |
| "loss": 0.26813453435897827, |
| "step": 46, |
| "token_acc": 0.918918918918919 |
| }, |
| { |
| "epoch": 0.5340909090909091, |
| "grad_norm": 9.069170357073718, |
| "learning_rate": 8.775946673459682e-06, |
| "loss": 0.3316657841205597, |
| "step": 47, |
| "token_acc": 0.926829268292683 |
| }, |
| { |
| "epoch": 0.5454545454545454, |
| "grad_norm": 9.467364120777885, |
| "learning_rate": 8.713625457077585e-06, |
| "loss": 0.3275446593761444, |
| "step": 48, |
| "token_acc": 0.8604651162790697 |
| }, |
| { |
| "epoch": 0.5568181818181818, |
| "grad_norm": 6.527486895359135, |
| "learning_rate": 8.64999006904096e-06, |
| "loss": 0.10486947745084763, |
| "step": 49, |
| "token_acc": 0.9473684210526315 |
| }, |
| { |
| "epoch": 0.5681818181818182, |
| "grad_norm": 8.03171901804604, |
| "learning_rate": 8.585063028536015e-06, |
| "loss": 0.1567389965057373, |
| "step": 50, |
| "token_acc": 0.9736842105263158 |
| }, |
| { |
| "epoch": 0.5795454545454546, |
| "grad_norm": 9.54100355711753, |
| "learning_rate": 8.518867311836808e-06, |
| "loss": 0.18998514115810394, |
| "step": 51, |
| "token_acc": 0.9743589743589743 |
| }, |
| { |
| "epoch": 0.5909090909090909, |
| "grad_norm": 14.315562799137096, |
| "learning_rate": 8.451426344174433e-06, |
| "loss": 0.41271671652793884, |
| "step": 52, |
| "token_acc": 0.8888888888888888 |
| }, |
| { |
| "epoch": 0.6022727272727273, |
| "grad_norm": 7.6591623398110364, |
| "learning_rate": 8.382763991447344e-06, |
| "loss": 0.1697944849729538, |
| "step": 53, |
| "token_acc": 0.972972972972973 |
| }, |
| { |
| "epoch": 0.6136363636363636, |
| "grad_norm": 7.728658786307115, |
| "learning_rate": 8.312904551775731e-06, |
| "loss": 0.12445592880249023, |
| "step": 54, |
| "token_acc": 0.9722222222222222 |
| }, |
| { |
| "epoch": 0.625, |
| "grad_norm": 6.928973081487085, |
| "learning_rate": 8.241872746902934e-06, |
| "loss": 0.18071499466896057, |
| "step": 55, |
| "token_acc": 0.95 |
| }, |
| { |
| "epoch": 0.6363636363636364, |
| "grad_norm": 10.93589823580153, |
| "learning_rate": 8.16969371344696e-06, |
| "loss": 0.27051395177841187, |
| "step": 56, |
| "token_acc": 0.9047619047619048 |
| }, |
| { |
| "epoch": 0.6477272727272727, |
| "grad_norm": 6.347167386694028, |
| "learning_rate": 8.096392994005177e-06, |
| "loss": 0.1373193860054016, |
| "step": 57, |
| "token_acc": 0.9512195121951219 |
| }, |
| { |
| "epoch": 0.6590909090909091, |
| "grad_norm": 9.945237728537498, |
| "learning_rate": 8.021996528115335e-06, |
| "loss": 0.5271890163421631, |
| "step": 58, |
| "token_acc": 0.9210526315789473 |
| }, |
| { |
| "epoch": 0.6704545454545454, |
| "grad_norm": 14.74078839957054, |
| "learning_rate": 7.946530643076138e-06, |
| "loss": 0.493736207485199, |
| "step": 59, |
| "token_acc": 0.8947368421052632 |
| }, |
| { |
| "epoch": 0.6818181818181818, |
| "grad_norm": 11.065972717032892, |
| "learning_rate": 7.870022044630569e-06, |
| "loss": 0.32494378089904785, |
| "step": 60, |
| "token_acc": 0.925 |
| }, |
| { |
| "epoch": 0.6931818181818182, |
| "grad_norm": 11.44327848116244, |
| "learning_rate": 7.792497807515317e-06, |
| "loss": 0.31102028489112854, |
| "step": 61, |
| "token_acc": 0.9210526315789473 |
| }, |
| { |
| "epoch": 0.7045454545454546, |
| "grad_norm": 8.66805616454484, |
| "learning_rate": 7.713985365879607e-06, |
| "loss": 0.16446304321289062, |
| "step": 62, |
| "token_acc": 0.9 |
| }, |
| { |
| "epoch": 0.7159090909090909, |
| "grad_norm": 9.211547349947377, |
| "learning_rate": 7.63451250357685e-06, |
| "loss": 0.19309315085411072, |
| "step": 63, |
| "token_acc": 0.9 |
| }, |
| { |
| "epoch": 0.7272727272727273, |
| "grad_norm": 6.570773561644991, |
| "learning_rate": 7.55410734433254e-06, |
| "loss": 0.15972477197647095, |
| "step": 64, |
| "token_acc": 0.9473684210526315 |
| }, |
| { |
| "epoch": 0.7386363636363636, |
| "grad_norm": 11.004459112406472, |
| "learning_rate": 7.472798341791877e-06, |
| "loss": 0.33975672721862793, |
| "step": 65, |
| "token_acc": 0.8947368421052632 |
| }, |
| { |
| "epoch": 0.75, |
| "grad_norm": 7.056423289090194, |
| "learning_rate": 7.390614269450633e-06, |
| "loss": 0.19755582511425018, |
| "step": 66, |
| "token_acc": 0.9459459459459459 |
| }, |
| { |
| "epoch": 0.7613636363636364, |
| "grad_norm": 8.873371196309954, |
| "learning_rate": 7.3075842104728445e-06, |
| "loss": 0.29561012983322144, |
| "step": 67, |
| "token_acc": 0.9230769230769231 |
| }, |
| { |
| "epoch": 0.7727272727272727, |
| "grad_norm": 7.651685537791527, |
| "learning_rate": 7.223737547398898e-06, |
| "loss": 0.239907369017601, |
| "step": 68, |
| "token_acc": 0.925 |
| }, |
| { |
| "epoch": 0.7840909090909091, |
| "grad_norm": 7.502542695671493, |
| "learning_rate": 7.139103951747694e-06, |
| "loss": 0.17635251581668854, |
| "step": 69, |
| "token_acc": 0.9459459459459459 |
| }, |
| { |
| "epoch": 0.7954545454545454, |
| "grad_norm": 8.741695509084057, |
| "learning_rate": 7.053713373516538e-06, |
| "loss": 0.21274474263191223, |
| "step": 70, |
| "token_acc": 0.95 |
| }, |
| { |
| "epoch": 0.8068181818181818, |
| "grad_norm": 6.790076837073194, |
| "learning_rate": 6.9675960305824785e-06, |
| "loss": 0.22346122562885284, |
| "step": 71, |
| "token_acc": 0.8947368421052632 |
| }, |
| { |
| "epoch": 0.8181818181818182, |
| "grad_norm": 7.431039250490218, |
| "learning_rate": 6.880782398008862e-06, |
| "loss": 0.1527853161096573, |
| "step": 72, |
| "token_acc": 0.9487179487179487 |
| }, |
| { |
| "epoch": 0.8295454545454546, |
| "grad_norm": 8.164971968902563, |
| "learning_rate": 6.7933031972608644e-06, |
| "loss": 0.1806386411190033, |
| "step": 73, |
| "token_acc": 0.9512195121951219 |
| }, |
| { |
| "epoch": 0.8409090909090909, |
| "grad_norm": 10.66558090913365, |
| "learning_rate": 6.70518938533383e-06, |
| "loss": 0.13778291642665863, |
| "step": 74, |
| "token_acc": 0.9473684210526315 |
| }, |
| { |
| "epoch": 0.8522727272727273, |
| "grad_norm": 9.529132284960879, |
| "learning_rate": 6.61647214379826e-06, |
| "loss": 0.2421799749135971, |
| "step": 75, |
| "token_acc": 0.925 |
| }, |
| { |
| "epoch": 0.8636363636363636, |
| "grad_norm": 9.227761599729329, |
| "learning_rate": 6.527182867765333e-06, |
| "loss": 0.14697305858135223, |
| "step": 76, |
| "token_acc": 0.9459459459459459 |
| }, |
| { |
| "epoch": 0.875, |
| "grad_norm": 9.81449017329525, |
| "learning_rate": 6.437353154776848e-06, |
| "loss": 0.29360121488571167, |
| "step": 77, |
| "token_acc": 0.8974358974358975 |
| }, |
| { |
| "epoch": 0.8863636363636364, |
| "grad_norm": 9.389854827380821, |
| "learning_rate": 6.3470147936235485e-06, |
| "loss": 0.31358498334884644, |
| "step": 78, |
| "token_acc": 0.926829268292683 |
| }, |
| { |
| "epoch": 0.8977272727272727, |
| "grad_norm": 6.1997657994467446, |
| "learning_rate": 6.256199753095745e-06, |
| "loss": 0.11065918952226639, |
| "step": 79, |
| "token_acc": 0.9736842105263158 |
| }, |
| { |
| "epoch": 0.9090909090909091, |
| "grad_norm": 7.145552043992015, |
| "learning_rate": 6.164940170670266e-06, |
| "loss": 0.16367042064666748, |
| "step": 80, |
| "token_acc": 0.9285714285714286 |
| }, |
| { |
| "epoch": 0.9204545454545454, |
| "grad_norm": 7.93747744022566, |
| "learning_rate": 6.073268341137694e-06, |
| "loss": 0.22145897150039673, |
| "step": 81, |
| "token_acc": 0.9459459459459459 |
| }, |
| { |
| "epoch": 0.9318181818181818, |
| "grad_norm": 6.472591335113881, |
| "learning_rate": 5.98121670517393e-06, |
| "loss": 0.15611425042152405, |
| "step": 82, |
| "token_acc": 0.9534883720930233 |
| }, |
| { |
| "epoch": 0.9431818181818182, |
| "grad_norm": 9.728846060957679, |
| "learning_rate": 5.8888178378601565e-06, |
| "loss": 0.33416748046875, |
| "step": 83, |
| "token_acc": 0.8974358974358975 |
| }, |
| { |
| "epoch": 0.9545454545454546, |
| "grad_norm": 5.949524265644311, |
| "learning_rate": 5.796104437155213e-06, |
| "loss": 0.23228313028812408, |
| "step": 84, |
| "token_acc": 0.925 |
| }, |
| { |
| "epoch": 0.9659090909090909, |
| "grad_norm": 8.506133041858282, |
| "learning_rate": 5.703109312324493e-06, |
| "loss": 0.20393462479114532, |
| "step": 85, |
| "token_acc": 0.95 |
| }, |
| { |
| "epoch": 0.9772727272727273, |
| "grad_norm": 7.684069112631382, |
| "learning_rate": 5.609865372329461e-06, |
| "loss": 0.23599931597709656, |
| "step": 86, |
| "token_acc": 0.8918918918918919 |
| }, |
| { |
| "epoch": 0.9886363636363636, |
| "grad_norm": 5.561375344471036, |
| "learning_rate": 5.516405614181883e-06, |
| "loss": 0.10231108218431473, |
| "step": 87, |
| "token_acc": 0.975 |
| }, |
| { |
| "epoch": 1.0, |
| "grad_norm": 6.402313033196499, |
| "learning_rate": 5.4227631112668955e-06, |
| "loss": 0.1993713080883026, |
| "step": 88, |
| "token_acc": 0.918918918918919 |
| }, |
| { |
| "epoch": 1.0113636363636365, |
| "grad_norm": 7.199893764507934, |
| "learning_rate": 5.328971001639054e-06, |
| "loss": 0.18095773458480835, |
| "step": 89, |
| "token_acc": 0.95 |
| }, |
| { |
| "epoch": 1.0227272727272727, |
| "grad_norm": 2.47366012333541, |
| "learning_rate": 5.235062476295488e-06, |
| "loss": 0.0230905469506979, |
| "step": 90, |
| "token_acc": 1.0 |
| }, |
| { |
| "epoch": 1.0340909090909092, |
| "grad_norm": 7.279321402429015, |
| "learning_rate": 5.141070767430331e-06, |
| "loss": 0.1365898996591568, |
| "step": 91, |
| "token_acc": 0.9459459459459459 |
| }, |
| { |
| "epoch": 1.0454545454545454, |
| "grad_norm": 5.350657202009064, |
| "learning_rate": 5.047029136674563e-06, |
| "loss": 0.11321850121021271, |
| "step": 92, |
| "token_acc": 0.9545454545454546 |
| }, |
| { |
| "epoch": 1.0568181818181819, |
| "grad_norm": 13.953533805140324, |
| "learning_rate": 4.95297086332544e-06, |
| "loss": 0.27081143856048584, |
| "step": 93, |
| "token_acc": 0.9210526315789473 |
| }, |
| { |
| "epoch": 1.0681818181818181, |
| "grad_norm": 9.411721345338792, |
| "learning_rate": 4.858929232569671e-06, |
| "loss": 0.13699771463871002, |
| "step": 94, |
| "token_acc": 0.9473684210526315 |
| }, |
| { |
| "epoch": 1.0795454545454546, |
| "grad_norm": 6.675002235100726, |
| "learning_rate": 4.7649375237045135e-06, |
| "loss": 0.10652704536914825, |
| "step": 95, |
| "token_acc": 0.9487179487179487 |
| }, |
| { |
| "epoch": 1.0909090909090908, |
| "grad_norm": 8.181756584986672, |
| "learning_rate": 4.671028998360947e-06, |
| "loss": 0.1227627545595169, |
| "step": 96, |
| "token_acc": 0.9487179487179487 |
| }, |
| { |
| "epoch": 1.1022727272727273, |
| "grad_norm": 13.821525224054058, |
| "learning_rate": 4.5772368887331044e-06, |
| "loss": 0.3747377395629883, |
| "step": 97, |
| "token_acc": 0.925 |
| }, |
| { |
| "epoch": 1.1136363636363635, |
| "grad_norm": 9.008201950815346, |
| "learning_rate": 4.483594385818119e-06, |
| "loss": 0.09657552093267441, |
| "step": 98, |
| "token_acc": 0.9444444444444444 |
| }, |
| { |
| "epoch": 1.125, |
| "grad_norm": 0.3399503581431662, |
| "learning_rate": 4.39013462767054e-06, |
| "loss": 0.0029641843866556883, |
| "step": 99, |
| "token_acc": 1.0 |
| }, |
| { |
| "epoch": 1.1363636363636362, |
| "grad_norm": 8.404769260512008, |
| "learning_rate": 4.29689068767551e-06, |
| "loss": 0.1632961481809616, |
| "step": 100, |
| "token_acc": 0.9230769230769231 |
| }, |
| { |
| "epoch": 1.1477272727272727, |
| "grad_norm": 6.156169046008191, |
| "learning_rate": 4.203895562844789e-06, |
| "loss": 0.11006340384483337, |
| "step": 101, |
| "token_acc": 0.9444444444444444 |
| }, |
| { |
| "epoch": 1.1590909090909092, |
| "grad_norm": 9.581907369768807, |
| "learning_rate": 4.111182162139844e-06, |
| "loss": 0.15489846467971802, |
| "step": 102, |
| "token_acc": 0.9545454545454546 |
| }, |
| { |
| "epoch": 1.1704545454545454, |
| "grad_norm": 2.1476395609705206, |
| "learning_rate": 4.018783294826071e-06, |
| "loss": 0.025181613862514496, |
| "step": 103, |
| "token_acc": 1.0 |
| }, |
| { |
| "epoch": 1.1818181818181819, |
| "grad_norm": 8.653026525644153, |
| "learning_rate": 3.926731658862307e-06, |
| "loss": 0.1841011792421341, |
| "step": 104, |
| "token_acc": 0.9285714285714286 |
| }, |
| { |
| "epoch": 1.1931818181818181, |
| "grad_norm": 7.504232851195428, |
| "learning_rate": 3.8350598293297345e-06, |
| "loss": 0.12672147154808044, |
| "step": 105, |
| "token_acc": 0.926829268292683 |
| }, |
| { |
| "epoch": 1.2045454545454546, |
| "grad_norm": 9.885762505275693, |
| "learning_rate": 3.7438002469042567e-06, |
| "loss": 0.276999294757843, |
| "step": 106, |
| "token_acc": 0.875 |
| }, |
| { |
| "epoch": 1.2159090909090908, |
| "grad_norm": 12.477323997898832, |
| "learning_rate": 3.652985206376455e-06, |
| "loss": 0.11598174273967743, |
| "step": 107, |
| "token_acc": 0.9166666666666666 |
| }, |
| { |
| "epoch": 1.2272727272727273, |
| "grad_norm": 6.978502854316774, |
| "learning_rate": 3.5626468452231534e-06, |
| "loss": 0.14181096851825714, |
| "step": 108, |
| "token_acc": 0.9512195121951219 |
| }, |
| { |
| "epoch": 1.2386363636363638, |
| "grad_norm": 10.418438484418086, |
| "learning_rate": 3.472817132234669e-06, |
| "loss": 0.24340537190437317, |
| "step": 109, |
| "token_acc": 0.95 |
| }, |
| { |
| "epoch": 1.25, |
| "grad_norm": 9.635231513627847, |
| "learning_rate": 3.3835278562017405e-06, |
| "loss": 0.1451936662197113, |
| "step": 110, |
| "token_acc": 0.95 |
| }, |
| { |
| "epoch": 1.2613636363636362, |
| "grad_norm": 6.495559629271991, |
| "learning_rate": 3.29481061466617e-06, |
| "loss": 0.13670389354228973, |
| "step": 111, |
| "token_acc": 0.9285714285714286 |
| }, |
| { |
| "epoch": 1.2727272727272727, |
| "grad_norm": 7.599787807019207, |
| "learning_rate": 3.2066968027391377e-06, |
| "loss": 0.1283184140920639, |
| "step": 112, |
| "token_acc": 0.975 |
| }, |
| { |
| "epoch": 1.2840909090909092, |
| "grad_norm": 7.8214885017197995, |
| "learning_rate": 3.119217601991139e-06, |
| "loss": 0.12655146420001984, |
| "step": 113, |
| "token_acc": 0.9459459459459459 |
| }, |
| { |
| "epoch": 1.2954545454545454, |
| "grad_norm": 1.33376059820044, |
| "learning_rate": 3.032403969417523e-06, |
| "loss": 0.012109911069273949, |
| "step": 114, |
| "token_acc": 1.0 |
| }, |
| { |
| "epoch": 1.3068181818181819, |
| "grad_norm": 1.855469756126943, |
| "learning_rate": 2.946286626483463e-06, |
| "loss": 0.016417205333709717, |
| "step": 115, |
| "token_acc": 1.0 |
| }, |
| { |
| "epoch": 1.3181818181818181, |
| "grad_norm": 14.519179269847893, |
| "learning_rate": 2.8608960482523058e-06, |
| "loss": 0.4724964499473572, |
| "step": 116, |
| "token_acc": 0.8648648648648649 |
| }, |
| { |
| "epoch": 1.3295454545454546, |
| "grad_norm": 8.04566880478037, |
| "learning_rate": 2.776262452601104e-06, |
| "loss": 0.17560063302516937, |
| "step": 117, |
| "token_acc": 0.9459459459459459 |
| }, |
| { |
| "epoch": 1.3409090909090908, |
| "grad_norm": 12.657358483238154, |
| "learning_rate": 2.6924157895271563e-06, |
| "loss": 0.3591260313987732, |
| "step": 118, |
| "token_acc": 0.8780487804878049 |
| }, |
| { |
| "epoch": 1.3522727272727273, |
| "grad_norm": 4.27840695352133, |
| "learning_rate": 2.6093857305493666e-06, |
| "loss": 0.06306027621030807, |
| "step": 119, |
| "token_acc": 0.9736842105263158 |
| }, |
| { |
| "epoch": 1.3636363636363638, |
| "grad_norm": 9.075217088447092, |
| "learning_rate": 2.5272016582081236e-06, |
| "loss": 0.2458903044462204, |
| "step": 120, |
| "token_acc": 0.9473684210526315 |
| }, |
| { |
| "epoch": 1.375, |
| "grad_norm": 1.7685033476767302, |
| "learning_rate": 2.445892655667462e-06, |
| "loss": 0.01817174255847931, |
| "step": 121, |
| "token_acc": 1.0 |
| }, |
| { |
| "epoch": 1.3863636363636362, |
| "grad_norm": 7.4739153510813185, |
| "learning_rate": 2.365487496423152e-06, |
| "loss": 0.17428866028785706, |
| "step": 122, |
| "token_acc": 0.95 |
| }, |
| { |
| "epoch": 1.3977272727272727, |
| "grad_norm": 7.659624146496388, |
| "learning_rate": 2.2860146341203936e-06, |
| "loss": 0.12651793658733368, |
| "step": 123, |
| "token_acc": 0.918918918918919 |
| }, |
| { |
| "epoch": 1.4090909090909092, |
| "grad_norm": 7.85241582131157, |
| "learning_rate": 2.207502192484685e-06, |
| "loss": 0.16867241263389587, |
| "step": 124, |
| "token_acc": 0.9459459459459459 |
| }, |
| { |
| "epoch": 1.4204545454545454, |
| "grad_norm": 5.969979957166302, |
| "learning_rate": 2.1299779553694323e-06, |
| "loss": 0.07473576813936234, |
| "step": 125, |
| "token_acc": 0.975 |
| }, |
| { |
| "epoch": 1.4318181818181819, |
| "grad_norm": 1.8255317910915432, |
| "learning_rate": 2.053469356923865e-06, |
| "loss": 0.016128040850162506, |
| "step": 126, |
| "token_acc": 1.0 |
| }, |
| { |
| "epoch": 1.4431818181818181, |
| "grad_norm": 5.5394567300337005, |
| "learning_rate": 1.9780034718846653e-06, |
| "loss": 0.055757757276296616, |
| "step": 127, |
| "token_acc": 0.9736842105263158 |
| }, |
| { |
| "epoch": 1.4545454545454546, |
| "grad_norm": 6.583725345555674, |
| "learning_rate": 1.9036070059948253e-06, |
| "loss": 0.1242956891655922, |
| "step": 128, |
| "token_acc": 0.9210526315789473 |
| }, |
| { |
| "epoch": 1.4659090909090908, |
| "grad_norm": 5.9848420129991835, |
| "learning_rate": 1.8303062865530407e-06, |
| "loss": 0.15906336903572083, |
| "step": 129, |
| "token_acc": 0.975 |
| }, |
| { |
| "epoch": 1.4772727272727273, |
| "grad_norm": 6.110563378659318, |
| "learning_rate": 1.7581272530970666e-06, |
| "loss": 0.13482968509197235, |
| "step": 130, |
| "token_acc": 0.9722222222222222 |
| }, |
| { |
| "epoch": 1.4886363636363638, |
| "grad_norm": 2.433994954138156, |
| "learning_rate": 1.6870954482242707e-06, |
| "loss": 0.01966552808880806, |
| "step": 131, |
| "token_acc": 1.0 |
| }, |
| { |
| "epoch": 1.5, |
| "grad_norm": 7.238276976507611, |
| "learning_rate": 1.6172360085526567e-06, |
| "loss": 0.15612998604774475, |
| "step": 132, |
| "token_acc": 0.9487179487179487 |
| }, |
| { |
| "epoch": 1.5113636363636362, |
| "grad_norm": 3.269861711718459, |
| "learning_rate": 1.54857365582557e-06, |
| "loss": 0.0246112123131752, |
| "step": 133, |
| "token_acc": 1.0 |
| }, |
| { |
| "epoch": 1.5227272727272727, |
| "grad_norm": 9.569596217425156, |
| "learning_rate": 1.4811326881631937e-06, |
| "loss": 0.29205116629600525, |
| "step": 134, |
| "token_acc": 0.8974358974358975 |
| }, |
| { |
| "epoch": 1.5340909090909092, |
| "grad_norm": 5.934444868229808, |
| "learning_rate": 1.4149369714639856e-06, |
| "loss": 0.1366189867258072, |
| "step": 135, |
| "token_acc": 0.9285714285714286 |
| }, |
| { |
| "epoch": 1.5454545454545454, |
| "grad_norm": 4.946059819569902, |
| "learning_rate": 1.3500099309590397e-06, |
| "loss": 0.05375118553638458, |
| "step": 136, |
| "token_acc": 0.9743589743589743 |
| }, |
| { |
| "epoch": 1.5568181818181817, |
| "grad_norm": 5.420651409348639, |
| "learning_rate": 1.2863745429224145e-06, |
| "loss": 0.07174789160490036, |
| "step": 137, |
| "token_acc": 0.9487179487179487 |
| }, |
| { |
| "epoch": 1.5681818181818183, |
| "grad_norm": 4.022143313508885, |
| "learning_rate": 1.22405332654032e-06, |
| "loss": 0.028939109295606613, |
| "step": 138, |
| "token_acc": 1.0 |
| }, |
| { |
| "epoch": 1.5795454545454546, |
| "grad_norm": 9.697441443288636, |
| "learning_rate": 1.1630683359420653e-06, |
| "loss": 0.16048294305801392, |
| "step": 139, |
| "token_acc": 0.918918918918919 |
| }, |
| { |
| "epoch": 1.5909090909090908, |
| "grad_norm": 8.229479639223838, |
| "learning_rate": 1.103441152395588e-06, |
| "loss": 0.1701526790857315, |
| "step": 140, |
| "token_acc": 0.9459459459459459 |
| }, |
| { |
| "epoch": 1.6022727272727273, |
| "grad_norm": 6.954228918415347, |
| "learning_rate": 1.045192876670298e-06, |
| "loss": 0.1258842945098877, |
| "step": 141, |
| "token_acc": 0.9743589743589743 |
| }, |
| { |
| "epoch": 1.6136363636363638, |
| "grad_norm": 2.4947934548232165, |
| "learning_rate": 9.883441215699824e-07, |
| "loss": 0.018796103075146675, |
| "step": 142, |
| "token_acc": 1.0 |
| }, |
| { |
| "epoch": 1.625, |
| "grad_norm": 7.6826017807589695, |
| "learning_rate": 9.329150046383773e-07, |
| "loss": 0.14171677827835083, |
| "step": 143, |
| "token_acc": 0.9487179487179487 |
| }, |
| { |
| "epoch": 1.6363636363636362, |
| "grad_norm": 1.981506306155957, |
| "learning_rate": 8.789251410400024e-07, |
| "loss": 0.021310359239578247, |
| "step": 144, |
| "token_acc": 1.0 |
| }, |
| { |
| "epoch": 1.6477272727272727, |
| "grad_norm": 9.316889291597604, |
| "learning_rate": 8.263936366187825e-07, |
| "loss": 0.21884095668792725, |
| "step": 145, |
| "token_acc": 0.95 |
| }, |
| { |
| "epoch": 1.6590909090909092, |
| "grad_norm": 7.099583247790577, |
| "learning_rate": 7.753390811368972e-07, |
| "loss": 0.14693105220794678, |
| "step": 146, |
| "token_acc": 0.926829268292683 |
| }, |
| { |
| "epoch": 1.6704545454545454, |
| "grad_norm": 7.061561118746622, |
| "learning_rate": 7.257795416962754e-07, |
| "loss": 0.09787449985742569, |
| "step": 147, |
| "token_acc": 0.95 |
| }, |
| { |
| "epoch": 1.6818181818181817, |
| "grad_norm": 8.540582560750158, |
| "learning_rate": 6.777325563450282e-07, |
| "loss": 0.1575159877538681, |
| "step": 148, |
| "token_acc": 0.9230769230769231 |
| }, |
| { |
| "epoch": 1.6931818181818183, |
| "grad_norm": 2.181121101987422, |
| "learning_rate": 6.312151278711237e-07, |
| "loss": 0.02192770503461361, |
| "step": 149, |
| "token_acc": 1.0 |
| }, |
| { |
| "epoch": 1.7045454545454546, |
| "grad_norm": 7.3305693576878195, |
| "learning_rate": 5.862437177854629e-07, |
| "loss": 0.10576251894235611, |
| "step": 150, |
| "token_acc": 0.95 |
| }, |
| { |
| "epoch": 1.7159090909090908, |
| "grad_norm": 7.275408392352814, |
| "learning_rate": 5.428342404965076e-07, |
| "loss": 0.09697937965393066, |
| "step": 151, |
| "token_acc": 0.9523809523809523 |
| }, |
| { |
| "epoch": 1.7272727272727273, |
| "grad_norm": 7.7167780665745545, |
| "learning_rate": 5.010020576785174e-07, |
| "loss": 0.11863157898187637, |
| "step": 152, |
| "token_acc": 0.972972972972973 |
| }, |
| { |
| "epoch": 1.7386363636363638, |
| "grad_norm": 6.854566961260753, |
| "learning_rate": 4.607619728353818e-07, |
| "loss": 0.08705897629261017, |
| "step": 153, |
| "token_acc": 0.9459459459459459 |
| }, |
| { |
| "epoch": 1.75, |
| "grad_norm": 1.4950103182848966, |
| "learning_rate": 4.221282260619891e-07, |
| "loss": 0.016769861802458763, |
| "step": 154, |
| "token_acc": 1.0 |
| }, |
| { |
| "epoch": 1.7613636363636362, |
| "grad_norm": 5.335402155261971, |
| "learning_rate": 3.851144890049535e-07, |
| "loss": 0.07611395418643951, |
| "step": 155, |
| "token_acc": 0.9444444444444444 |
| }, |
| { |
| "epoch": 1.7727272727272727, |
| "grad_norm": 3.997548118257428, |
| "learning_rate": 3.497338600245254e-07, |
| "loss": 0.03954974189400673, |
| "step": 156, |
| "token_acc": 1.0 |
| }, |
| { |
| "epoch": 1.7840909090909092, |
| "grad_norm": 6.029329008445983, |
| "learning_rate": 3.159988595593616e-07, |
| "loss": 0.10782109200954437, |
| "step": 157, |
| "token_acc": 0.975 |
| }, |
| { |
| "epoch": 1.7954545454545454, |
| "grad_norm": 2.113198810909374, |
| "learning_rate": 2.839214256958106e-07, |
| "loss": 0.01702558435499668, |
| "step": 158, |
| "token_acc": 1.0 |
| }, |
| { |
| "epoch": 1.8068181818181817, |
| "grad_norm": 6.085772397449453, |
| "learning_rate": 2.5351290994328703e-07, |
| "loss": 0.10801274329423904, |
| "step": 159, |
| "token_acc": 0.9736842105263158 |
| }, |
| { |
| "epoch": 1.8181818181818183, |
| "grad_norm": 4.822542333419725, |
| "learning_rate": 2.2478407321721295e-07, |
| "loss": 0.07818000018596649, |
| "step": 160, |
| "token_acc": 0.9473684210526315 |
| }, |
| { |
| "epoch": 1.8295454545454546, |
| "grad_norm": 6.90692198121901, |
| "learning_rate": 1.9774508203096843e-07, |
| "loss": 0.10510721802711487, |
| "step": 161, |
| "token_acc": 0.8974358974358975 |
| }, |
| { |
| "epoch": 1.8409090909090908, |
| "grad_norm": 9.424366100232511, |
| "learning_rate": 1.7240550489817652e-07, |
| "loss": 0.31693142652511597, |
| "step": 162, |
| "token_acc": 0.9210526315789473 |
| }, |
| { |
| "epoch": 1.8522727272727273, |
| "grad_norm": 5.9906515122333275, |
| "learning_rate": 1.4877430894662037e-07, |
| "loss": 0.0935223400592804, |
| "step": 163, |
| "token_acc": 0.9512195121951219 |
| }, |
| { |
| "epoch": 1.8636363636363638, |
| "grad_norm": 0.7001975036371686, |
| "learning_rate": 1.268598567449647e-07, |
| "loss": 0.005389641970396042, |
| "step": 164, |
| "token_acc": 1.0 |
| }, |
| { |
| "epoch": 1.875, |
| "grad_norm": 5.440457762197436, |
| "learning_rate": 1.0666990334342708e-07, |
| "loss": 0.09289371967315674, |
| "step": 165, |
| "token_acc": 0.9459459459459459 |
| }, |
| { |
| "epoch": 1.8863636363636362, |
| "grad_norm": 4.629011735151434, |
| "learning_rate": 8.821159352943142e-08, |
| "loss": 0.05701345205307007, |
| "step": 166, |
| "token_acc": 0.95 |
| }, |
| { |
| "epoch": 1.8977272727272727, |
| "grad_norm": 9.266170148953467, |
| "learning_rate": 7.149145929922607e-08, |
| "loss": 0.2318773865699768, |
| "step": 167, |
| "token_acc": 0.926829268292683 |
| }, |
| { |
| "epoch": 1.9090909090909092, |
| "grad_norm": 7.325295151739083, |
| "learning_rate": 5.651541754634726e-08, |
| "loss": 0.11617127805948257, |
| "step": 168, |
| "token_acc": 0.9459459459459459 |
| }, |
| { |
| "epoch": 1.9204545454545454, |
| "grad_norm": 4.679409964801964, |
| "learning_rate": 4.328876796776071e-08, |
| "loss": 0.048571497201919556, |
| "step": 169, |
| "token_acc": 0.972972972972973 |
| }, |
| { |
| "epoch": 1.9318181818181817, |
| "grad_norm": 5.203562284759042, |
| "learning_rate": 3.181619118841517e-08, |
| "loss": 0.10277199745178223, |
| "step": 170, |
| "token_acc": 0.9761904761904762 |
| }, |
| { |
| "epoch": 1.9431818181818183, |
| "grad_norm": 6.026296193559956, |
| "learning_rate": 2.210174710486679e-08, |
| "loss": 0.09030674397945404, |
| "step": 171, |
| "token_acc": 0.9512195121951219 |
| }, |
| { |
| "epoch": 1.9545454545454546, |
| "grad_norm": 8.383579125272592, |
| "learning_rate": 1.4148873448573408e-08, |
| "loss": 0.1585637778043747, |
| "step": 172, |
| "token_acc": 0.925 |
| }, |
| { |
| "epoch": 1.9659090909090908, |
| "grad_norm": 5.941814331196659, |
| "learning_rate": 7.96038456935322e-09, |
| "loss": 0.09052297472953796, |
| "step": 173, |
| "token_acc": 0.9473684210526315 |
| }, |
| { |
| "epoch": 1.9772727272727273, |
| "grad_norm": 5.31219068311242, |
| "learning_rate": 3.538470439448105e-09, |
| "loss": 0.08900852501392365, |
| "step": 174, |
| "token_acc": 0.9534883720930233 |
| }, |
| { |
| "epoch": 1.9886363636363638, |
| "grad_norm": 6.436701396284728, |
| "learning_rate": 8.846958785418969e-10, |
| "loss": 0.09907500445842743, |
| "step": 175, |
| "token_acc": 0.9459459459459459 |
| }, |
| { |
| "epoch": 2.0, |
| "grad_norm": 6.805455894721991, |
| "learning_rate": 0.0, |
| "loss": 0.10433489084243774, |
| "step": 176, |
| "token_acc": 0.972972972972973 |
| } |
| ], |
| "logging_steps": 1, |
| "max_steps": 176, |
| "num_input_tokens_seen": 0, |
| "num_train_epochs": 2, |
| "save_steps": 88, |
| "stateful_callbacks": { |
| "TrainerControl": { |
| "args": { |
| "should_epoch_stop": false, |
| "should_evaluate": false, |
| "should_log": false, |
| "should_save": true, |
| "should_training_stop": true |
| }, |
| "attributes": {} |
| } |
| }, |
| "total_flos": 1241146570240.0, |
| "train_batch_size": 2, |
| "trial_name": null, |
| "trial_params": null |
| } |
|
|
