| { | |
| "best_metric": 0.0379740335047245, | |
| "best_model_checkpoint": "doc-topic-model_eval-03_train-01/checkpoint-16000", | |
| "epoch": 10.355029585798816, | |
| "eval_steps": 1000, | |
| "global_step": 21000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.2465483234714004, | |
| "grad_norm": 0.32817578315734863, | |
| "learning_rate": 1.9950690335305722e-05, | |
| "loss": 0.1659, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.4930966469428008, | |
| "grad_norm": 0.34106990694999695, | |
| "learning_rate": 1.9901380670611442e-05, | |
| "loss": 0.0941, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.4930966469428008, | |
| "eval_accuracy": 0.981439120143997, | |
| "eval_f1": 0.0, | |
| "eval_loss": 0.09035909175872803, | |
| "eval_precision": 0.0, | |
| "eval_recall": 0.0, | |
| "eval_runtime": 12.1739, | |
| "eval_samples_per_second": 666.182, | |
| "eval_steps_per_second": 2.629, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.7396449704142012, | |
| "grad_norm": 0.4004117548465729, | |
| "learning_rate": 1.9852071005917162e-05, | |
| "loss": 0.088, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.9861932938856016, | |
| "grad_norm": 0.37588435411453247, | |
| "learning_rate": 1.980276134122288e-05, | |
| "loss": 0.0787, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.9861932938856016, | |
| "eval_accuracy": 0.981439120143997, | |
| "eval_f1": 0.0, | |
| "eval_loss": 0.07049160450696945, | |
| "eval_precision": 0.0, | |
| "eval_recall": 0.0, | |
| "eval_runtime": 12.1677, | |
| "eval_samples_per_second": 666.518, | |
| "eval_steps_per_second": 2.63, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 1.232741617357002, | |
| "grad_norm": 0.43946701288223267, | |
| "learning_rate": 1.9753451676528602e-05, | |
| "loss": 0.0696, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 1.4792899408284024, | |
| "grad_norm": 0.4252438545227051, | |
| "learning_rate": 1.9704142011834322e-05, | |
| "loss": 0.0628, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 1.4792899408284024, | |
| "eval_accuracy": 0.9822044588063551, | |
| "eval_f1": 0.1170182841068917, | |
| "eval_loss": 0.05713700130581856, | |
| "eval_precision": 0.7402135231316725, | |
| "eval_recall": 0.06353084911423336, | |
| "eval_runtime": 14.6983, | |
| "eval_samples_per_second": 551.763, | |
| "eval_steps_per_second": 2.177, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 1.725838264299803, | |
| "grad_norm": 0.3535268306732178, | |
| "learning_rate": 1.965483234714004e-05, | |
| "loss": 0.0584, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 1.972386587771203, | |
| "grad_norm": 0.3976692855358124, | |
| "learning_rate": 1.9605522682445763e-05, | |
| "loss": 0.0537, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 1.972386587771203, | |
| "eval_accuracy": 0.9841787490964752, | |
| "eval_f1": 0.3162859067801801, | |
| "eval_loss": 0.05010608583688736, | |
| "eval_precision": 0.7991333952336738, | |
| "eval_recall": 0.19715943799633476, | |
| "eval_runtime": 14.5719, | |
| "eval_samples_per_second": 556.549, | |
| "eval_steps_per_second": 2.196, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 2.2189349112426036, | |
| "grad_norm": 0.24940212070941925, | |
| "learning_rate": 1.9556213017751483e-05, | |
| "loss": 0.0499, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 2.465483234714004, | |
| "grad_norm": 0.40094295144081116, | |
| "learning_rate": 1.95069033530572e-05, | |
| "loss": 0.0478, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 2.465483234714004, | |
| "eval_accuracy": 0.9851084938418584, | |
| "eval_f1": 0.42113382182799847, | |
| "eval_loss": 0.046862900257110596, | |
| "eval_precision": 0.7560830860534125, | |
| "eval_recall": 0.2918448381185095, | |
| "eval_runtime": 14.4838, | |
| "eval_samples_per_second": 559.937, | |
| "eval_steps_per_second": 2.209, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 2.712031558185404, | |
| "grad_norm": 0.42049625515937805, | |
| "learning_rate": 1.9457593688362923e-05, | |
| "loss": 0.0463, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 2.9585798816568047, | |
| "grad_norm": 0.41673073172569275, | |
| "learning_rate": 1.940828402366864e-05, | |
| "loss": 0.0453, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 2.9585798816568047, | |
| "eval_accuracy": 0.985744858766671, | |
| "eval_f1": 0.4941152801529021, | |
| "eval_loss": 0.044261422008275986, | |
| "eval_precision": 0.7238432066018273, | |
| "eval_recall": 0.3750763591936469, | |
| "eval_runtime": 14.4397, | |
| "eval_samples_per_second": 561.645, | |
| "eval_steps_per_second": 2.216, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 3.2051282051282053, | |
| "grad_norm": 0.3240315914154053, | |
| "learning_rate": 1.935897435897436e-05, | |
| "loss": 0.0424, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 3.4516765285996054, | |
| "grad_norm": 0.39667147397994995, | |
| "learning_rate": 1.930966469428008e-05, | |
| "loss": 0.0389, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 3.4516765285996054, | |
| "eval_accuracy": 0.9863174454696203, | |
| "eval_f1": 0.5358653846153846, | |
| "eval_loss": 0.04172312840819359, | |
| "eval_precision": 0.7233904465212876, | |
| "eval_recall": 0.42554978619425776, | |
| "eval_runtime": 14.5013, | |
| "eval_samples_per_second": 559.261, | |
| "eval_steps_per_second": 2.207, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 3.698224852071006, | |
| "grad_norm": 0.39966270327568054, | |
| "learning_rate": 1.92603550295858e-05, | |
| "loss": 0.0403, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 3.9447731755424065, | |
| "grad_norm": 0.42601555585861206, | |
| "learning_rate": 1.921104536489152e-05, | |
| "loss": 0.0393, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 3.9447731755424065, | |
| "eval_accuracy": 0.9861601258556911, | |
| "eval_f1": 0.5412262156448203, | |
| "eval_loss": 0.04098295420408249, | |
| "eval_precision": 0.7033825863963854, | |
| "eval_recall": 0.4398289554062309, | |
| "eval_runtime": 14.4597, | |
| "eval_samples_per_second": 560.87, | |
| "eval_steps_per_second": 2.213, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 4.191321499013807, | |
| "grad_norm": 0.41485270857810974, | |
| "learning_rate": 1.916173570019724e-05, | |
| "loss": 0.0361, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 4.437869822485207, | |
| "grad_norm": 0.4282112717628479, | |
| "learning_rate": 1.911242603550296e-05, | |
| "loss": 0.0349, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 4.437869822485207, | |
| "eval_accuracy": 0.9867851524299502, | |
| "eval_f1": 0.5692506698697218, | |
| "eval_loss": 0.03965826332569122, | |
| "eval_precision": 0.7205847953216374, | |
| "eval_recall": 0.4704489920586439, | |
| "eval_runtime": 14.5109, | |
| "eval_samples_per_second": 558.891, | |
| "eval_steps_per_second": 2.205, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 4.684418145956608, | |
| "grad_norm": 0.5737874507904053, | |
| "learning_rate": 1.906311637080868e-05, | |
| "loss": 0.0358, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 4.930966469428008, | |
| "grad_norm": 0.42196542024612427, | |
| "learning_rate": 1.90138067061144e-05, | |
| "loss": 0.0344, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 4.930966469428008, | |
| "eval_accuracy": 0.9869849908584548, | |
| "eval_f1": 0.5743684820393975, | |
| "eval_loss": 0.03886643052101135, | |
| "eval_precision": 0.730746550300743, | |
| "eval_recall": 0.4731215638362859, | |
| "eval_runtime": 14.4599, | |
| "eval_samples_per_second": 560.863, | |
| "eval_steps_per_second": 2.213, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 5.177514792899408, | |
| "grad_norm": 0.2073846012353897, | |
| "learning_rate": 1.896459566074951e-05, | |
| "loss": 0.0315, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 5.424063116370808, | |
| "grad_norm": 0.3816153109073639, | |
| "learning_rate": 1.891528599605523e-05, | |
| "loss": 0.0302, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 5.424063116370808, | |
| "eval_accuracy": 0.9871692390549485, | |
| "eval_f1": 0.5890790250102129, | |
| "eval_loss": 0.03835758566856384, | |
| "eval_precision": 0.7262451035254617, | |
| "eval_recall": 0.495494807574832, | |
| "eval_runtime": 14.4472, | |
| "eval_samples_per_second": 561.355, | |
| "eval_steps_per_second": 2.215, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 5.670611439842209, | |
| "grad_norm": 0.4880998730659485, | |
| "learning_rate": 1.886597633136095e-05, | |
| "loss": 0.0321, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 5.9171597633136095, | |
| "grad_norm": 0.31077542901039124, | |
| "learning_rate": 1.881666666666667e-05, | |
| "loss": 0.0305, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 5.9171597633136095, | |
| "eval_accuracy": 0.9869538103944329, | |
| "eval_f1": 0.5894107676524377, | |
| "eval_loss": 0.03864369913935661, | |
| "eval_precision": 0.7086774643355144, | |
| "eval_recall": 0.504505192425168, | |
| "eval_runtime": 14.4003, | |
| "eval_samples_per_second": 563.181, | |
| "eval_steps_per_second": 2.222, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 6.16370808678501, | |
| "grad_norm": 0.4404219686985016, | |
| "learning_rate": 1.876735700197239e-05, | |
| "loss": 0.0281, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 6.410256410256411, | |
| "grad_norm": 0.39860355854034424, | |
| "learning_rate": 1.8718145956607497e-05, | |
| "loss": 0.027, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 6.410256410256411, | |
| "eval_accuracy": 0.9872528593902802, | |
| "eval_f1": 0.5966454390528299, | |
| "eval_loss": 0.03840399533510208, | |
| "eval_precision": 0.7228863290589003, | |
| "eval_recall": 0.5079413561392792, | |
| "eval_runtime": 14.5776, | |
| "eval_samples_per_second": 556.332, | |
| "eval_steps_per_second": 2.195, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 6.65680473372781, | |
| "grad_norm": 0.3162294924259186, | |
| "learning_rate": 1.8668836291913217e-05, | |
| "loss": 0.0267, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 6.903353057199211, | |
| "grad_norm": 0.3698256313800812, | |
| "learning_rate": 1.8619526627218937e-05, | |
| "loss": 0.0282, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 6.903353057199211, | |
| "eval_accuracy": 0.9873704947772722, | |
| "eval_f1": 0.6017786119676454, | |
| "eval_loss": 0.038046352565288544, | |
| "eval_precision": 0.7254606184678375, | |
| "eval_recall": 0.5141264508246793, | |
| "eval_runtime": 14.743, | |
| "eval_samples_per_second": 550.092, | |
| "eval_steps_per_second": 2.171, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 7.149901380670611, | |
| "grad_norm": 0.42440325021743774, | |
| "learning_rate": 1.8570216962524657e-05, | |
| "loss": 0.0252, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 7.396449704142012, | |
| "grad_norm": 0.5403586626052856, | |
| "learning_rate": 1.8521005917159765e-05, | |
| "loss": 0.0235, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 7.396449704142012, | |
| "eval_accuracy": 0.987438524880593, | |
| "eval_f1": 0.6184510740884239, | |
| "eval_loss": 0.03824068605899811, | |
| "eval_precision": 0.708872002368499, | |
| "eval_recall": 0.5484880879657911, | |
| "eval_runtime": 14.4874, | |
| "eval_samples_per_second": 559.797, | |
| "eval_steps_per_second": 2.209, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 7.642998027613412, | |
| "grad_norm": 0.3601667582988739, | |
| "learning_rate": 1.8471696252465485e-05, | |
| "loss": 0.0243, | |
| "step": 15500 | |
| }, | |
| { | |
| "epoch": 7.889546351084813, | |
| "grad_norm": 0.4192611575126648, | |
| "learning_rate": 1.8422386587771205e-05, | |
| "loss": 0.0255, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 7.889546351084813, | |
| "eval_accuracy": 0.9874456113496889, | |
| "eval_f1": 0.6197630494505495, | |
| "eval_loss": 0.0379740335047245, | |
| "eval_precision": 0.7077450980392157, | |
| "eval_recall": 0.55123701893708, | |
| "eval_runtime": 14.5961, | |
| "eval_samples_per_second": 555.63, | |
| "eval_steps_per_second": 2.192, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 8.136094674556213, | |
| "grad_norm": 0.13463503122329712, | |
| "learning_rate": 1.8373076923076926e-05, | |
| "loss": 0.0224, | |
| "step": 16500 | |
| }, | |
| { | |
| "epoch": 8.382642998027613, | |
| "grad_norm": 0.25073572993278503, | |
| "learning_rate": 1.8323865877712033e-05, | |
| "loss": 0.0214, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 8.382642998027613, | |
| "eval_accuracy": 0.9875703332057769, | |
| "eval_f1": 0.6291754756871036, | |
| "eval_loss": 0.03818249702453613, | |
| "eval_precision": 0.7049459920409323, | |
| "eval_recall": 0.5681124007330483, | |
| "eval_runtime": 14.6338, | |
| "eval_samples_per_second": 554.198, | |
| "eval_steps_per_second": 2.187, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 8.629191321499015, | |
| "grad_norm": 0.39663228392601013, | |
| "learning_rate": 1.8274556213017754e-05, | |
| "loss": 0.0215, | |
| "step": 17500 | |
| }, | |
| { | |
| "epoch": 8.875739644970414, | |
| "grad_norm": 0.7587671279907227, | |
| "learning_rate": 1.822524654832347e-05, | |
| "loss": 0.0222, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 8.875739644970414, | |
| "eval_accuracy": 0.9875816715563304, | |
| "eval_f1": 0.6271172014639543, | |
| "eval_loss": 0.03859669715166092, | |
| "eval_precision": 0.7083253220534512, | |
| "eval_recall": 0.5626145387904704, | |
| "eval_runtime": 14.5227, | |
| "eval_samples_per_second": 558.436, | |
| "eval_steps_per_second": 2.203, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 9.122287968441814, | |
| "grad_norm": 0.14514634013175964, | |
| "learning_rate": 1.8175936883629194e-05, | |
| "loss": 0.0198, | |
| "step": 18500 | |
| }, | |
| { | |
| "epoch": 9.368836291913215, | |
| "grad_norm": 0.44190242886543274, | |
| "learning_rate": 1.8126725838264302e-05, | |
| "loss": 0.0192, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 9.368836291913215, | |
| "eval_accuracy": 0.9874087617103902, | |
| "eval_f1": 0.6294318845415867, | |
| "eval_loss": 0.03965187072753906, | |
| "eval_precision": 0.6936017650303364, | |
| "eval_recall": 0.5761301160659743, | |
| "eval_runtime": 14.4812, | |
| "eval_samples_per_second": 560.038, | |
| "eval_steps_per_second": 2.21, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 9.615384615384615, | |
| "grad_norm": 0.3838333785533905, | |
| "learning_rate": 1.8077416173570022e-05, | |
| "loss": 0.0193, | |
| "step": 19500 | |
| }, | |
| { | |
| "epoch": 9.861932938856016, | |
| "grad_norm": 0.36484745144844055, | |
| "learning_rate": 1.802810650887574e-05, | |
| "loss": 0.0189, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 9.861932938856016, | |
| "eval_accuracy": 0.9875037203962753, | |
| "eval_f1": 0.6300197222105661, | |
| "eval_loss": 0.03964517265558243, | |
| "eval_precision": 0.6993013507219376, | |
| "eval_recall": 0.5732284667073916, | |
| "eval_runtime": 14.5377, | |
| "eval_samples_per_second": 557.86, | |
| "eval_steps_per_second": 2.201, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 10.108481262327416, | |
| "grad_norm": 0.18375088274478912, | |
| "learning_rate": 1.7978796844181462e-05, | |
| "loss": 0.0175, | |
| "step": 20500 | |
| }, | |
| { | |
| "epoch": 10.355029585798816, | |
| "grad_norm": 0.13978295028209686, | |
| "learning_rate": 1.7929487179487182e-05, | |
| "loss": 0.0159, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 10.355029585798816, | |
| "eval_accuracy": 0.9876893858865882, | |
| "eval_f1": 0.6361732428583396, | |
| "eval_loss": 0.040131811052560806, | |
| "eval_precision": 0.7045834106513268, | |
| "eval_recall": 0.5798717165546732, | |
| "eval_runtime": 14.7405, | |
| "eval_samples_per_second": 550.183, | |
| "eval_steps_per_second": 2.171, | |
| "step": 21000 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 202800, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 100, | |
| "save_steps": 1000, | |
| "stateful_callbacks": { | |
| "EarlyStoppingCallback": { | |
| "args": { | |
| "early_stopping_patience": 5, | |
| "early_stopping_threshold": 0.0 | |
| }, | |
| "attributes": { | |
| "early_stopping_patience_counter": 0 | |
| } | |
| }, | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 474932252325492.0, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |