| { | |
| "best_metric": 0.7544031311154599, | |
| "best_model_checkpoint": "tiny-bert-sst2-distilled/run-6/checkpoint-744", | |
| "epoch": 8.0, | |
| "eval_steps": 500, | |
| "global_step": 744, | |
| "is_hyper_param_search": true, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.4286664128303528, | |
| "learning_rate": 1.945409831472016e-05, | |
| "loss": 0.4806, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.5, | |
| "eval_f1": 0.0, | |
| "eval_loss": 0.4703535735607147, | |
| "eval_precision": 0.0, | |
| "eval_recall": 0.0, | |
| "eval_runtime": 28.42, | |
| "eval_samples_per_second": 35.961, | |
| "eval_steps_per_second": 1.126, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 0.94158935546875, | |
| "learning_rate": 1.7022336025380143e-05, | |
| "loss": 0.4648, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.5176125244618396, | |
| "eval_f1": 0.10849909584086799, | |
| "eval_loss": 0.4538751244544983, | |
| "eval_precision": 0.7142857142857143, | |
| "eval_recall": 0.05870841487279843, | |
| "eval_runtime": 28.3351, | |
| "eval_samples_per_second": 36.068, | |
| "eval_steps_per_second": 1.129, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 1.0048439502716064, | |
| "learning_rate": 1.459057373604012e-05, | |
| "loss": 0.4406, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.5929549902152642, | |
| "eval_f1": 0.3677811550151976, | |
| "eval_loss": 0.4149659276008606, | |
| "eval_precision": 0.8231292517006803, | |
| "eval_recall": 0.23679060665362034, | |
| "eval_runtime": 28.3925, | |
| "eval_samples_per_second": 35.995, | |
| "eval_steps_per_second": 1.127, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 2.1001150608062744, | |
| "learning_rate": 1.2158811446700102e-05, | |
| "loss": 0.4126, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.7025440313111546, | |
| "eval_f1": 0.6456876456876457, | |
| "eval_loss": 0.39196181297302246, | |
| "eval_precision": 0.7982708933717579, | |
| "eval_recall": 0.5420743639921722, | |
| "eval_runtime": 28.4943, | |
| "eval_samples_per_second": 35.867, | |
| "eval_steps_per_second": 1.123, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 1.2996047735214233, | |
| "learning_rate": 9.72704915736008e-06, | |
| "loss": 0.4021, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.735812133072407, | |
| "eval_f1": 0.7133757961783439, | |
| "eval_loss": 0.3851335346698761, | |
| "eval_precision": 0.7795823665893271, | |
| "eval_recall": 0.6575342465753424, | |
| "eval_runtime": 28.8344, | |
| "eval_samples_per_second": 35.444, | |
| "eval_steps_per_second": 1.11, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "grad_norm": 1.8318911790847778, | |
| "learning_rate": 7.29528686802006e-06, | |
| "loss": 0.3976, | |
| "step": 558 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.7524461839530333, | |
| "eval_f1": 0.7394438722966015, | |
| "eval_loss": 0.3816056251525879, | |
| "eval_precision": 0.7804347826086957, | |
| "eval_recall": 0.7025440313111546, | |
| "eval_runtime": 28.629, | |
| "eval_samples_per_second": 35.698, | |
| "eval_steps_per_second": 1.118, | |
| "step": 558 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "grad_norm": 1.5164391994476318, | |
| "learning_rate": 4.86352457868004e-06, | |
| "loss": 0.3934, | |
| "step": 651 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.7504892367906066, | |
| "eval_f1": 0.7357512953367875, | |
| "eval_loss": 0.37981584668159485, | |
| "eval_precision": 0.7819383259911894, | |
| "eval_recall": 0.6947162426614482, | |
| "eval_runtime": 29.4404, | |
| "eval_samples_per_second": 34.714, | |
| "eval_steps_per_second": 1.087, | |
| "step": 651 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 1.8273214101791382, | |
| "learning_rate": 2.43176228934002e-06, | |
| "loss": 0.3903, | |
| "step": 744 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.7544031311154599, | |
| "eval_f1": 0.7462082912032356, | |
| "eval_loss": 0.3790663480758667, | |
| "eval_precision": 0.7719665271966527, | |
| "eval_recall": 0.7221135029354208, | |
| "eval_runtime": 28.4026, | |
| "eval_samples_per_second": 35.983, | |
| "eval_steps_per_second": 1.127, | |
| "step": 744 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 837, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 9, | |
| "save_steps": 500, | |
| "total_flos": 1885561578240.0, | |
| "train_batch_size": 33, | |
| "trial_name": null, | |
| "trial_params": { | |
| "alpha": 0.6122687021783514, | |
| "learning_rate": 2.188586060406018e-05, | |
| "num_train_epochs": 9, | |
| "per_device_train_batch_size": 33, | |
| "temperature": 14 | |
| } | |
| } | |