| { | |
| "best_metric": 0.8307240704500979, | |
| "best_model_checkpoint": "tiny-bert-sst2-distilled/run-1/checkpoint-728", | |
| "epoch": 8.0, | |
| "eval_steps": 500, | |
| "global_step": 728, | |
| "is_hyper_param_search": true, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 5.826261043548584, | |
| "learning_rate": 0.0002239970993716639, | |
| "loss": 0.481, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.7749510763209393, | |
| "eval_f1": 0.7993019197207678, | |
| "eval_loss": 0.4282819330692291, | |
| "eval_precision": 0.721259842519685, | |
| "eval_recall": 0.8962818003913894, | |
| "eval_runtime": 28.5148, | |
| "eval_samples_per_second": 35.841, | |
| "eval_steps_per_second": 1.122, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 8.367091178894043, | |
| "learning_rate": 0.00019599746195020593, | |
| "loss": 0.4267, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.7896281800391389, | |
| "eval_f1": 0.8043676069153776, | |
| "eval_loss": 0.4049767255783081, | |
| "eval_precision": 0.7517006802721088, | |
| "eval_recall": 0.8649706457925636, | |
| "eval_runtime": 28.5272, | |
| "eval_samples_per_second": 35.825, | |
| "eval_steps_per_second": 1.122, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 3.821913480758667, | |
| "learning_rate": 0.00016799782452874793, | |
| "loss": 0.4057, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.8101761252446184, | |
| "eval_f1": 0.8116504854368931, | |
| "eval_loss": 0.38922348618507385, | |
| "eval_precision": 0.8053949903660886, | |
| "eval_recall": 0.8180039138943248, | |
| "eval_runtime": 28.5222, | |
| "eval_samples_per_second": 35.832, | |
| "eval_steps_per_second": 1.122, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 4.150608539581299, | |
| "learning_rate": 0.00013999818710728996, | |
| "loss": 0.3899, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.8209393346379648, | |
| "eval_f1": 0.8221574344023325, | |
| "eval_loss": 0.38097265362739563, | |
| "eval_precision": 0.8166023166023166, | |
| "eval_recall": 0.8277886497064579, | |
| "eval_runtime": 27.9162, | |
| "eval_samples_per_second": 36.61, | |
| "eval_steps_per_second": 1.146, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 15.217473030090332, | |
| "learning_rate": 0.00011199854968583195, | |
| "loss": 0.3811, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.8209393346379648, | |
| "eval_f1": 0.8288119738072964, | |
| "eval_loss": 0.38663867115974426, | |
| "eval_precision": 0.7939068100358423, | |
| "eval_recall": 0.8669275929549902, | |
| "eval_runtime": 28.6175, | |
| "eval_samples_per_second": 35.712, | |
| "eval_steps_per_second": 1.118, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "grad_norm": 2.0855538845062256, | |
| "learning_rate": 8.399891226437396e-05, | |
| "loss": 0.3782, | |
| "step": 546 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.8111545988258317, | |
| "eval_f1": 0.8305531167690957, | |
| "eval_loss": 0.39196252822875977, | |
| "eval_precision": 0.7531847133757962, | |
| "eval_recall": 0.9256360078277887, | |
| "eval_runtime": 28.4313, | |
| "eval_samples_per_second": 35.946, | |
| "eval_steps_per_second": 1.126, | |
| "step": 546 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "grad_norm": 4.913670063018799, | |
| "learning_rate": 5.5999274842915974e-05, | |
| "loss": 0.3707, | |
| "step": 637 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.8170254403131115, | |
| "eval_f1": 0.8310749774164408, | |
| "eval_loss": 0.38759666681289673, | |
| "eval_precision": 0.7718120805369127, | |
| "eval_recall": 0.9001956947162426, | |
| "eval_runtime": 28.3099, | |
| "eval_samples_per_second": 36.1, | |
| "eval_steps_per_second": 1.13, | |
| "step": 637 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 8.997469902038574, | |
| "learning_rate": 2.7999637421457987e-05, | |
| "loss": 0.3696, | |
| "step": 728 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.8307240704500979, | |
| "eval_f1": 0.8422971741112123, | |
| "eval_loss": 0.3802284896373749, | |
| "eval_precision": 0.78839590443686, | |
| "eval_recall": 0.9041095890410958, | |
| "eval_runtime": 28.5564, | |
| "eval_samples_per_second": 35.789, | |
| "eval_steps_per_second": 1.121, | |
| "step": 728 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 819, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 9, | |
| "save_steps": 500, | |
| "total_flos": 1885561578240.0, | |
| "train_batch_size": 34, | |
| "trial_name": null, | |
| "trial_params": { | |
| "alpha": 0.7088780913019314, | |
| "learning_rate": 0.0002519967367931219, | |
| "num_train_epochs": 9, | |
| "per_device_train_batch_size": 34, | |
| "temperature": 19 | |
| } | |
| } | |