| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 10.0, | |
| "eval_steps": 500, | |
| "global_step": 15950, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.31347962382445144, | |
| "grad_norm": 0.9161905646324158, | |
| "learning_rate": 1.9840000000000003e-05, | |
| "loss": 5.5619, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.6269592476489029, | |
| "grad_norm": 0.6389903426170349, | |
| "learning_rate": 1.9357928802589e-05, | |
| "loss": 1.7174, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.9404388714733543, | |
| "grad_norm": 0.5035005211830139, | |
| "learning_rate": 1.8710679611650487e-05, | |
| "loss": 1.6907, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_bleu": 12.658054871280777, | |
| "eval_loss": 1.6638020277023315, | |
| "eval_runtime": 246.1231, | |
| "eval_samples_per_second": 11.441, | |
| "eval_steps_per_second": 0.715, | |
| "step": 1595 | |
| }, | |
| { | |
| "epoch": 1.2539184952978055, | |
| "grad_norm": 0.5225879549980164, | |
| "learning_rate": 1.8063430420711977e-05, | |
| "loss": 1.6638, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 1.567398119122257, | |
| "grad_norm": 0.497714102268219, | |
| "learning_rate": 1.7416181229773465e-05, | |
| "loss": 1.6555, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 1.8808777429467085, | |
| "grad_norm": 0.5155704021453857, | |
| "learning_rate": 1.6768932038834952e-05, | |
| "loss": 1.6542, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_bleu": 13.064274040118429, | |
| "eval_loss": 1.6535193920135498, | |
| "eval_runtime": 263.0779, | |
| "eval_samples_per_second": 10.704, | |
| "eval_steps_per_second": 0.669, | |
| "step": 3190 | |
| }, | |
| { | |
| "epoch": 2.19435736677116, | |
| "grad_norm": 0.48234814405441284, | |
| "learning_rate": 1.612168284789644e-05, | |
| "loss": 1.6385, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 2.507836990595611, | |
| "grad_norm": 0.5366239547729492, | |
| "learning_rate": 1.547443365695793e-05, | |
| "loss": 1.6303, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 2.8213166144200628, | |
| "grad_norm": 0.4303179383277893, | |
| "learning_rate": 1.482718446601942e-05, | |
| "loss": 1.6291, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_bleu": 14.326650897191437, | |
| "eval_loss": 1.648271083831787, | |
| "eval_runtime": 238.3692, | |
| "eval_samples_per_second": 11.814, | |
| "eval_steps_per_second": 0.738, | |
| "step": 4785 | |
| }, | |
| { | |
| "epoch": 3.134796238244514, | |
| "grad_norm": 0.40259647369384766, | |
| "learning_rate": 1.4179935275080907e-05, | |
| "loss": 1.6216, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 3.4482758620689653, | |
| "grad_norm": 0.49842897057533264, | |
| "learning_rate": 1.3532686084142396e-05, | |
| "loss": 1.6111, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 3.761755485893417, | |
| "grad_norm": 0.4395376741886139, | |
| "learning_rate": 1.2885436893203884e-05, | |
| "loss": 1.6123, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_bleu": 14.710438534445657, | |
| "eval_loss": 1.6465048789978027, | |
| "eval_runtime": 229.3865, | |
| "eval_samples_per_second": 12.276, | |
| "eval_steps_per_second": 0.767, | |
| "step": 6380 | |
| }, | |
| { | |
| "epoch": 4.075235109717869, | |
| "grad_norm": 0.6033243536949158, | |
| "learning_rate": 1.2238187702265374e-05, | |
| "loss": 1.6092, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 4.38871473354232, | |
| "grad_norm": 0.541384220123291, | |
| "learning_rate": 1.1590938511326861e-05, | |
| "loss": 1.5984, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 4.702194357366771, | |
| "grad_norm": 0.42259281873703003, | |
| "learning_rate": 1.094368932038835e-05, | |
| "loss": 1.5995, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_bleu": 14.942770930991193, | |
| "eval_loss": 1.646419644355774, | |
| "eval_runtime": 231.9426, | |
| "eval_samples_per_second": 12.141, | |
| "eval_steps_per_second": 0.759, | |
| "step": 7975 | |
| }, | |
| { | |
| "epoch": 5.015673981191223, | |
| "grad_norm": 0.45884329080581665, | |
| "learning_rate": 1.0296440129449838e-05, | |
| "loss": 1.5988, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 5.329153605015674, | |
| "grad_norm": 0.3725912272930145, | |
| "learning_rate": 9.649190938511328e-06, | |
| "loss": 1.5874, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 5.6426332288401255, | |
| "grad_norm": 0.3701029419898987, | |
| "learning_rate": 9.001941747572817e-06, | |
| "loss": 1.5886, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 5.956112852664576, | |
| "grad_norm": 0.46482861042022705, | |
| "learning_rate": 8.354692556634304e-06, | |
| "loss": 1.5893, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_bleu": 14.678090282520607, | |
| "eval_loss": 1.6463446617126465, | |
| "eval_runtime": 231.5813, | |
| "eval_samples_per_second": 12.16, | |
| "eval_steps_per_second": 0.76, | |
| "step": 9570 | |
| }, | |
| { | |
| "epoch": 6.269592476489028, | |
| "grad_norm": 0.4049539864063263, | |
| "learning_rate": 7.707443365695793e-06, | |
| "loss": 1.5804, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 6.58307210031348, | |
| "grad_norm": 0.42662039399147034, | |
| "learning_rate": 7.060194174757282e-06, | |
| "loss": 1.5804, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 6.896551724137931, | |
| "grad_norm": 0.5612569451332092, | |
| "learning_rate": 6.412944983818771e-06, | |
| "loss": 1.5807, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_bleu": 15.24653511299691, | |
| "eval_loss": 1.6464687585830688, | |
| "eval_runtime": 228.543, | |
| "eval_samples_per_second": 12.322, | |
| "eval_steps_per_second": 0.77, | |
| "step": 11165 | |
| }, | |
| { | |
| "epoch": 7.210031347962382, | |
| "grad_norm": 0.4263598620891571, | |
| "learning_rate": 5.765695792880259e-06, | |
| "loss": 1.5754, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 7.523510971786834, | |
| "grad_norm": 0.39763760566711426, | |
| "learning_rate": 5.118446601941748e-06, | |
| "loss": 1.574, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 7.836990595611285, | |
| "grad_norm": 0.4740435779094696, | |
| "learning_rate": 4.471197411003236e-06, | |
| "loss": 1.5739, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_bleu": 15.692886975359345, | |
| "eval_loss": 1.6478267908096313, | |
| "eval_runtime": 228.9879, | |
| "eval_samples_per_second": 12.298, | |
| "eval_steps_per_second": 0.769, | |
| "step": 12760 | |
| }, | |
| { | |
| "epoch": 8.150470219435737, | |
| "grad_norm": 0.35859590768814087, | |
| "learning_rate": 3.823948220064725e-06, | |
| "loss": 1.571, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 8.463949843260188, | |
| "grad_norm": 0.32463109493255615, | |
| "learning_rate": 3.176699029126214e-06, | |
| "loss": 1.569, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 8.77742946708464, | |
| "grad_norm": 0.4066263437271118, | |
| "learning_rate": 2.5294498381877025e-06, | |
| "loss": 1.5694, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_bleu": 15.603090223915896, | |
| "eval_loss": 1.647833228111267, | |
| "eval_runtime": 226.5331, | |
| "eval_samples_per_second": 12.431, | |
| "eval_steps_per_second": 0.777, | |
| "step": 14355 | |
| }, | |
| { | |
| "epoch": 9.090909090909092, | |
| "grad_norm": 0.4182904362678528, | |
| "learning_rate": 1.882200647249191e-06, | |
| "loss": 1.568, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 9.404388714733543, | |
| "grad_norm": 0.45490360260009766, | |
| "learning_rate": 1.2349514563106797e-06, | |
| "loss": 1.5655, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 9.717868338557993, | |
| "grad_norm": 0.42171570658683777, | |
| "learning_rate": 5.877022653721683e-07, | |
| "loss": 1.5658, | |
| "step": 15500 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_bleu": 15.780359854009903, | |
| "eval_loss": 1.6481457948684692, | |
| "eval_runtime": 230.3301, | |
| "eval_samples_per_second": 12.226, | |
| "eval_steps_per_second": 0.764, | |
| "step": 15950 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 15950, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 10, | |
| "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": 1.3821793152270336e+17, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |