| { | |
| "best_global_step": 36872, | |
| "best_metric": 0.995850622406639, | |
| "best_model_checkpoint": "./rubert-good-bad/checkpoint-36872", | |
| "epoch": 12.0, | |
| "eval_steps": 500, | |
| "global_step": 40224, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.14916467780429593, | |
| "grad_norm": 7.852961540222168, | |
| "learning_rate": 1.975188941925219e-05, | |
| "loss": 0.4052, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.29832935560859186, | |
| "grad_norm": 3.1220028400421143, | |
| "learning_rate": 1.9503281622911697e-05, | |
| "loss": 0.2525, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.44749403341288785, | |
| "grad_norm": 4.411465644836426, | |
| "learning_rate": 1.9254673826571203e-05, | |
| "loss": 0.2181, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.5966587112171837, | |
| "grad_norm": 4.210575580596924, | |
| "learning_rate": 1.900606603023071e-05, | |
| "loss": 0.1906, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.7458233890214797, | |
| "grad_norm": 7.871537685394287, | |
| "learning_rate": 1.8757458233890215e-05, | |
| "loss": 0.1713, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.8949880668257757, | |
| "grad_norm": 4.302709102630615, | |
| "learning_rate": 1.8508850437549724e-05, | |
| "loss": 0.1618, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.9545433353536296, | |
| "eval_loss": 0.11475327610969543, | |
| "eval_runtime": 5.422, | |
| "eval_samples_per_second": 3955.922, | |
| "eval_steps_per_second": 123.755, | |
| "step": 3352 | |
| }, | |
| { | |
| "epoch": 1.0441527446300716, | |
| "grad_norm": 6.954808235168457, | |
| "learning_rate": 1.826024264120923e-05, | |
| "loss": 0.1452, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 1.1933174224343674, | |
| "grad_norm": 6.441762924194336, | |
| "learning_rate": 1.8011634844868736e-05, | |
| "loss": 0.1349, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 1.3424821002386635, | |
| "grad_norm": 2.3034775257110596, | |
| "learning_rate": 1.7763027048528242e-05, | |
| "loss": 0.1261, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 1.4916467780429594, | |
| "grad_norm": 1.0179331302642822, | |
| "learning_rate": 1.751441925218775e-05, | |
| "loss": 0.116, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 1.6408114558472553, | |
| "grad_norm": 8.965306282043457, | |
| "learning_rate": 1.7265811455847257e-05, | |
| "loss": 0.1095, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 1.7899761336515514, | |
| "grad_norm": 2.703307867050171, | |
| "learning_rate": 1.7017203659506763e-05, | |
| "loss": 0.114, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 1.9391408114558473, | |
| "grad_norm": 8.306221961975098, | |
| "learning_rate": 1.676859586316627e-05, | |
| "loss": 0.1088, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.9744510233577323, | |
| "eval_loss": 0.06880993396043777, | |
| "eval_runtime": 5.2887, | |
| "eval_samples_per_second": 4055.651, | |
| "eval_steps_per_second": 126.875, | |
| "step": 6704 | |
| }, | |
| { | |
| "epoch": 2.088305489260143, | |
| "grad_norm": 5.413713455200195, | |
| "learning_rate": 1.6519988066825775e-05, | |
| "loss": 0.096, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 2.2374701670644392, | |
| "grad_norm": 9.367850303649902, | |
| "learning_rate": 1.6271380270485284e-05, | |
| "loss": 0.0894, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 2.386634844868735, | |
| "grad_norm": 0.4695867896080017, | |
| "learning_rate": 1.602277247414479e-05, | |
| "loss": 0.0882, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 2.535799522673031, | |
| "grad_norm": 8.3032808303833, | |
| "learning_rate": 1.57741646778043e-05, | |
| "loss": 0.0885, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 2.684964200477327, | |
| "grad_norm": 3.0729820728302, | |
| "learning_rate": 1.5525556881463802e-05, | |
| "loss": 0.0844, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 2.834128878281623, | |
| "grad_norm": 6.244034767150879, | |
| "learning_rate": 1.527694908512331e-05, | |
| "loss": 0.0854, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 2.983293556085919, | |
| "grad_norm": 14.806967735290527, | |
| "learning_rate": 1.5028341288782817e-05, | |
| "loss": 0.0871, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.9813044897198004, | |
| "eval_loss": 0.04988732188940048, | |
| "eval_runtime": 5.3443, | |
| "eval_samples_per_second": 4013.427, | |
| "eval_steps_per_second": 125.554, | |
| "step": 10056 | |
| }, | |
| { | |
| "epoch": 3.132458233890215, | |
| "grad_norm": 3.9732472896575928, | |
| "learning_rate": 1.4779733492442325e-05, | |
| "loss": 0.0745, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 3.2816229116945106, | |
| "grad_norm": 3.8117780685424805, | |
| "learning_rate": 1.4531125696101831e-05, | |
| "loss": 0.0697, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 3.4307875894988067, | |
| "grad_norm": 1.4716224670410156, | |
| "learning_rate": 1.4282517899761337e-05, | |
| "loss": 0.0707, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 3.579952267303103, | |
| "grad_norm": 1.516440749168396, | |
| "learning_rate": 1.4033910103420844e-05, | |
| "loss": 0.0687, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 3.7291169451073984, | |
| "grad_norm": 9.91383171081543, | |
| "learning_rate": 1.378530230708035e-05, | |
| "loss": 0.074, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 3.8782816229116945, | |
| "grad_norm": 0.14348161220550537, | |
| "learning_rate": 1.3536694510739858e-05, | |
| "loss": 0.0743, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.986386311716164, | |
| "eval_loss": 0.037734854966402054, | |
| "eval_runtime": 5.3579, | |
| "eval_samples_per_second": 4003.226, | |
| "eval_steps_per_second": 125.235, | |
| "step": 13408 | |
| }, | |
| { | |
| "epoch": 4.02744630071599, | |
| "grad_norm": 3.358393430709839, | |
| "learning_rate": 1.3288086714399364e-05, | |
| "loss": 0.0687, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 4.176610978520286, | |
| "grad_norm": 11.118879318237305, | |
| "learning_rate": 1.3039478918058872e-05, | |
| "loss": 0.0572, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 4.325775656324582, | |
| "grad_norm": 3.1460227966308594, | |
| "learning_rate": 1.2790871121718378e-05, | |
| "loss": 0.0604, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 4.4749403341288785, | |
| "grad_norm": 5.454690933227539, | |
| "learning_rate": 1.2542263325377885e-05, | |
| "loss": 0.0576, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 4.624105011933175, | |
| "grad_norm": 5.29376220703125, | |
| "learning_rate": 1.2293655529037391e-05, | |
| "loss": 0.0619, | |
| "step": 15500 | |
| }, | |
| { | |
| "epoch": 4.77326968973747, | |
| "grad_norm": 0.4051016867160797, | |
| "learning_rate": 1.2045047732696897e-05, | |
| "loss": 0.0591, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 4.922434367541766, | |
| "grad_norm": 3.9757938385009766, | |
| "learning_rate": 1.1796439936356405e-05, | |
| "loss": 0.0566, | |
| "step": 16500 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.9901627115483239, | |
| "eval_loss": 0.02832028456032276, | |
| "eval_runtime": 5.3872, | |
| "eval_samples_per_second": 3981.457, | |
| "eval_steps_per_second": 124.554, | |
| "step": 16760 | |
| }, | |
| { | |
| "epoch": 5.071599045346062, | |
| "grad_norm": 5.767542839050293, | |
| "learning_rate": 1.154783214001591e-05, | |
| "loss": 0.0554, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 5.220763723150358, | |
| "grad_norm": 0.15083061158657074, | |
| "learning_rate": 1.1299224343675418e-05, | |
| "loss": 0.0507, | |
| "step": 17500 | |
| }, | |
| { | |
| "epoch": 5.369928400954654, | |
| "grad_norm": 3.9027342796325684, | |
| "learning_rate": 1.1050616547334924e-05, | |
| "loss": 0.0526, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 5.519093078758949, | |
| "grad_norm": 0.34978094696998596, | |
| "learning_rate": 1.0802008750994433e-05, | |
| "loss": 0.0489, | |
| "step": 18500 | |
| }, | |
| { | |
| "epoch": 5.6682577565632455, | |
| "grad_norm": 24.885211944580078, | |
| "learning_rate": 1.0553400954653938e-05, | |
| "loss": 0.0496, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 5.817422434367542, | |
| "grad_norm": 3.8451194763183594, | |
| "learning_rate": 1.0304793158313447e-05, | |
| "loss": 0.0462, | |
| "step": 19500 | |
| }, | |
| { | |
| "epoch": 5.966587112171838, | |
| "grad_norm": 2.7605342864990234, | |
| "learning_rate": 1.0056185361972953e-05, | |
| "loss": 0.0524, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.9920276003543289, | |
| "eval_loss": 0.023812290281057358, | |
| "eval_runtime": 5.7441, | |
| "eval_samples_per_second": 3734.067, | |
| "eval_steps_per_second": 116.815, | |
| "step": 20112 | |
| }, | |
| { | |
| "epoch": 6.115751789976134, | |
| "grad_norm": 12.299605369567871, | |
| "learning_rate": 9.807577565632459e-06, | |
| "loss": 0.0451, | |
| "step": 20500 | |
| }, | |
| { | |
| "epoch": 6.26491646778043, | |
| "grad_norm": 0.014037308283150196, | |
| "learning_rate": 9.558969769291967e-06, | |
| "loss": 0.0439, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 6.414081145584726, | |
| "grad_norm": 4.366552829742432, | |
| "learning_rate": 9.310361972951472e-06, | |
| "loss": 0.0447, | |
| "step": 21500 | |
| }, | |
| { | |
| "epoch": 6.563245823389021, | |
| "grad_norm": 0.3234240412712097, | |
| "learning_rate": 9.06175417661098e-06, | |
| "loss": 0.0474, | |
| "step": 22000 | |
| }, | |
| { | |
| "epoch": 6.712410501193317, | |
| "grad_norm": 5.625737190246582, | |
| "learning_rate": 8.813146380270486e-06, | |
| "loss": 0.0401, | |
| "step": 22500 | |
| }, | |
| { | |
| "epoch": 6.861575178997613, | |
| "grad_norm": 1.9961190223693848, | |
| "learning_rate": 8.564538583929994e-06, | |
| "loss": 0.0442, | |
| "step": 23000 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.9929134225371812, | |
| "eval_loss": 0.02055698074400425, | |
| "eval_runtime": 5.2686, | |
| "eval_samples_per_second": 4071.116, | |
| "eval_steps_per_second": 127.359, | |
| "step": 23464 | |
| }, | |
| { | |
| "epoch": 7.0107398568019095, | |
| "grad_norm": 0.7006446123123169, | |
| "learning_rate": 8.3159307875895e-06, | |
| "loss": 0.0446, | |
| "step": 23500 | |
| }, | |
| { | |
| "epoch": 7.159904534606206, | |
| "grad_norm": 0.009974448010325432, | |
| "learning_rate": 8.067322991249006e-06, | |
| "loss": 0.0353, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 7.309069212410501, | |
| "grad_norm": 13.073637962341309, | |
| "learning_rate": 7.818715194908513e-06, | |
| "loss": 0.0327, | |
| "step": 24500 | |
| }, | |
| { | |
| "epoch": 7.458233890214797, | |
| "grad_norm": 0.47934669256210327, | |
| "learning_rate": 7.57010739856802e-06, | |
| "loss": 0.0415, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 7.607398568019093, | |
| "grad_norm": 9.82204532623291, | |
| "learning_rate": 7.321499602227527e-06, | |
| "loss": 0.0396, | |
| "step": 25500 | |
| }, | |
| { | |
| "epoch": 7.756563245823389, | |
| "grad_norm": 3.7507293224334717, | |
| "learning_rate": 7.0728918058870335e-06, | |
| "loss": 0.0419, | |
| "step": 26000 | |
| }, | |
| { | |
| "epoch": 7.905727923627685, | |
| "grad_norm": 12.792013168334961, | |
| "learning_rate": 6.82428400954654e-06, | |
| "loss": 0.0437, | |
| "step": 26500 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.9944053335819851, | |
| "eval_loss": 0.017598649486899376, | |
| "eval_runtime": 5.3397, | |
| "eval_samples_per_second": 4016.869, | |
| "eval_steps_per_second": 125.662, | |
| "step": 26816 | |
| }, | |
| { | |
| "epoch": 8.05489260143198, | |
| "grad_norm": 2.062351703643799, | |
| "learning_rate": 6.575676213206047e-06, | |
| "loss": 0.037, | |
| "step": 27000 | |
| }, | |
| { | |
| "epoch": 8.204057279236277, | |
| "grad_norm": 7.230820178985596, | |
| "learning_rate": 6.327068416865554e-06, | |
| "loss": 0.0363, | |
| "step": 27500 | |
| }, | |
| { | |
| "epoch": 8.353221957040573, | |
| "grad_norm": 9.108207702636719, | |
| "learning_rate": 6.078460620525061e-06, | |
| "loss": 0.0343, | |
| "step": 28000 | |
| }, | |
| { | |
| "epoch": 8.50238663484487, | |
| "grad_norm": 3.1893274784088135, | |
| "learning_rate": 5.829852824184567e-06, | |
| "loss": 0.0374, | |
| "step": 28500 | |
| }, | |
| { | |
| "epoch": 8.651551312649165, | |
| "grad_norm": 7.297391414642334, | |
| "learning_rate": 5.581245027844073e-06, | |
| "loss": 0.0371, | |
| "step": 29000 | |
| }, | |
| { | |
| "epoch": 8.80071599045346, | |
| "grad_norm": 0.8091713190078735, | |
| "learning_rate": 5.33263723150358e-06, | |
| "loss": 0.0322, | |
| "step": 29500 | |
| }, | |
| { | |
| "epoch": 8.949880668257757, | |
| "grad_norm": 9.286097526550293, | |
| "learning_rate": 5.084029435163087e-06, | |
| "loss": 0.0366, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_accuracy": 0.995104666884237, | |
| "eval_loss": 0.015091103501617908, | |
| "eval_runtime": 5.5227, | |
| "eval_samples_per_second": 3883.811, | |
| "eval_steps_per_second": 121.499, | |
| "step": 30168 | |
| }, | |
| { | |
| "epoch": 9.099045346062052, | |
| "grad_norm": 2.9286885261535645, | |
| "learning_rate": 4.835421638822594e-06, | |
| "loss": 0.0295, | |
| "step": 30500 | |
| }, | |
| { | |
| "epoch": 9.24821002386635, | |
| "grad_norm": 0.06805476546287537, | |
| "learning_rate": 4.5868138424821005e-06, | |
| "loss": 0.0337, | |
| "step": 31000 | |
| }, | |
| { | |
| "epoch": 9.397374701670644, | |
| "grad_norm": 1.571494698524475, | |
| "learning_rate": 4.338206046141607e-06, | |
| "loss": 0.0331, | |
| "step": 31500 | |
| }, | |
| { | |
| "epoch": 9.54653937947494, | |
| "grad_norm": 0.042112987488508224, | |
| "learning_rate": 4.089598249801114e-06, | |
| "loss": 0.0298, | |
| "step": 32000 | |
| }, | |
| { | |
| "epoch": 9.695704057279237, | |
| "grad_norm": 0.060358405113220215, | |
| "learning_rate": 3.840990453460621e-06, | |
| "loss": 0.0354, | |
| "step": 32500 | |
| }, | |
| { | |
| "epoch": 9.844868735083532, | |
| "grad_norm": 17.142175674438477, | |
| "learning_rate": 3.5923826571201277e-06, | |
| "loss": 0.0315, | |
| "step": 33000 | |
| }, | |
| { | |
| "epoch": 9.994033412887829, | |
| "grad_norm": 0.03519833832979202, | |
| "learning_rate": 3.3437748607796345e-06, | |
| "loss": 0.0341, | |
| "step": 33500 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_accuracy": 0.9953844002051377, | |
| "eval_loss": 0.014557760208845139, | |
| "eval_runtime": 5.4843, | |
| "eval_samples_per_second": 3911.0, | |
| "eval_steps_per_second": 122.35, | |
| "step": 33520 | |
| }, | |
| { | |
| "epoch": 10.143198090692124, | |
| "grad_norm": 8.01278018951416, | |
| "learning_rate": 3.0951670644391413e-06, | |
| "loss": 0.0291, | |
| "step": 34000 | |
| }, | |
| { | |
| "epoch": 10.292362768496421, | |
| "grad_norm": 1.2233223915100098, | |
| "learning_rate": 2.8465592680986476e-06, | |
| "loss": 0.0295, | |
| "step": 34500 | |
| }, | |
| { | |
| "epoch": 10.441527446300716, | |
| "grad_norm": 0.13490265607833862, | |
| "learning_rate": 2.5979514717581544e-06, | |
| "loss": 0.0314, | |
| "step": 35000 | |
| }, | |
| { | |
| "epoch": 10.590692124105011, | |
| "grad_norm": 22.723793029785156, | |
| "learning_rate": 2.349343675417661e-06, | |
| "loss": 0.0323, | |
| "step": 35500 | |
| }, | |
| { | |
| "epoch": 10.739856801909308, | |
| "grad_norm": 0.0038608636241406202, | |
| "learning_rate": 2.100735879077168e-06, | |
| "loss": 0.0304, | |
| "step": 36000 | |
| }, | |
| { | |
| "epoch": 10.889021479713604, | |
| "grad_norm": 0.08787856996059418, | |
| "learning_rate": 1.8521280827366748e-06, | |
| "loss": 0.031, | |
| "step": 36500 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "eval_accuracy": 0.995850622406639, | |
| "eval_loss": 0.013228594325482845, | |
| "eval_runtime": 5.4141, | |
| "eval_samples_per_second": 3961.682, | |
| "eval_steps_per_second": 123.935, | |
| "step": 36872 | |
| }, | |
| { | |
| "epoch": 11.0381861575179, | |
| "grad_norm": 0.11552688479423523, | |
| "learning_rate": 1.6035202863961815e-06, | |
| "loss": 0.0318, | |
| "step": 37000 | |
| }, | |
| { | |
| "epoch": 11.187350835322196, | |
| "grad_norm": 0.19442684948444366, | |
| "learning_rate": 1.3549124900556881e-06, | |
| "loss": 0.027, | |
| "step": 37500 | |
| }, | |
| { | |
| "epoch": 11.336515513126491, | |
| "grad_norm": 10.4265718460083, | |
| "learning_rate": 1.106304693715195e-06, | |
| "loss": 0.027, | |
| "step": 38000 | |
| }, | |
| { | |
| "epoch": 11.485680190930788, | |
| "grad_norm": 0.03579527512192726, | |
| "learning_rate": 8.576968973747017e-07, | |
| "loss": 0.0325, | |
| "step": 38500 | |
| }, | |
| { | |
| "epoch": 11.634844868735083, | |
| "grad_norm": 3.7483468055725098, | |
| "learning_rate": 6.090891010342085e-07, | |
| "loss": 0.0267, | |
| "step": 39000 | |
| }, | |
| { | |
| "epoch": 11.78400954653938, | |
| "grad_norm": 0.031889185309410095, | |
| "learning_rate": 3.604813046937152e-07, | |
| "loss": 0.0268, | |
| "step": 39500 | |
| }, | |
| { | |
| "epoch": 11.933174224343675, | |
| "grad_norm": 0.04797011986374855, | |
| "learning_rate": 1.1187350835322197e-07, | |
| "loss": 0.0273, | |
| "step": 40000 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "eval_accuracy": 0.9957573779663388, | |
| "eval_loss": 0.013085125014185905, | |
| "eval_runtime": 5.2528, | |
| "eval_samples_per_second": 4083.366, | |
| "eval_steps_per_second": 127.742, | |
| "step": 40224 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 40224, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 12, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1159020529550640.0, | |
| "train_batch_size": 32, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |