| { | |
| "best_metric": 0.6604696673189824, | |
| "best_model_checkpoint": "tiny-bert-sst2-distilled/run-0/checkpoint-486", | |
| "epoch": 9.0, | |
| "eval_steps": 500, | |
| "global_step": 729, | |
| "is_hyper_param_search": true, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.29800447821617126, | |
| "learning_rate": 2.6647930950340397e-05, | |
| "loss": 0.3887, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.5, | |
| "eval_f1": 0.0, | |
| "eval_loss": 0.3728242516517639, | |
| "eval_precision": 0.0, | |
| "eval_recall": 0.0, | |
| "eval_runtime": 30.0921, | |
| "eval_samples_per_second": 33.962, | |
| "eval_steps_per_second": 1.063, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 1.0294488668441772, | |
| "learning_rate": 2.331693958154785e-05, | |
| "loss": 0.3659, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.5371819960861057, | |
| "eval_f1": 0.19145299145299144, | |
| "eval_loss": 0.35036981105804443, | |
| "eval_precision": 0.7567567567567568, | |
| "eval_recall": 0.1095890410958904, | |
| "eval_runtime": 30.0254, | |
| "eval_samples_per_second": 34.038, | |
| "eval_steps_per_second": 1.066, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 1.3886942863464355, | |
| "learning_rate": 1.99859482127553e-05, | |
| "loss": 0.3406, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.5802348336594912, | |
| "eval_f1": 0.32863849765258213, | |
| "eval_loss": 0.32384729385375977, | |
| "eval_precision": 0.8203125, | |
| "eval_recall": 0.2054794520547945, | |
| "eval_runtime": 29.2219, | |
| "eval_samples_per_second": 34.974, | |
| "eval_steps_per_second": 1.095, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 0.869124174118042, | |
| "learning_rate": 1.665495684396275e-05, | |
| "loss": 0.3253, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.6360078277886497, | |
| "eval_f1": 0.48901098901098894, | |
| "eval_loss": 0.31458520889282227, | |
| "eval_precision": 0.8202764976958525, | |
| "eval_recall": 0.34833659491193736, | |
| "eval_runtime": 29.7469, | |
| "eval_samples_per_second": 34.356, | |
| "eval_steps_per_second": 1.076, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 1.3957839012145996, | |
| "learning_rate": 1.3323965475170199e-05, | |
| "loss": 0.3227, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.6487279843444227, | |
| "eval_f1": 0.521970705725699, | |
| "eval_loss": 0.3122331500053406, | |
| "eval_precision": 0.8166666666666667, | |
| "eval_recall": 0.3835616438356164, | |
| "eval_runtime": 29.6103, | |
| "eval_samples_per_second": 34.515, | |
| "eval_steps_per_second": 1.081, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "grad_norm": 1.8226208686828613, | |
| "learning_rate": 9.99297410637765e-06, | |
| "loss": 0.3203, | |
| "step": 486 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.6604696673189824, | |
| "eval_f1": 0.5522580645161289, | |
| "eval_loss": 0.31177300214767456, | |
| "eval_precision": 0.8106060606060606, | |
| "eval_recall": 0.4187866927592955, | |
| "eval_runtime": 29.3492, | |
| "eval_samples_per_second": 34.822, | |
| "eval_steps_per_second": 1.09, | |
| "step": 486 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "grad_norm": 1.4845212697982788, | |
| "learning_rate": 6.661982737585099e-06, | |
| "loss": 0.3188, | |
| "step": 567 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.6340508806262231, | |
| "eval_f1": 0.48484848484848486, | |
| "eval_loss": 0.3102571964263916, | |
| "eval_precision": 0.8186046511627907, | |
| "eval_recall": 0.34442270058708413, | |
| "eval_runtime": 29.5252, | |
| "eval_samples_per_second": 34.614, | |
| "eval_steps_per_second": 1.084, | |
| "step": 567 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 1.0866426229476929, | |
| "learning_rate": 3.3309913687925497e-06, | |
| "loss": 0.3164, | |
| "step": 648 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.6545988258317026, | |
| "eval_f1": 0.5397653194263363, | |
| "eval_loss": 0.3101910948753357, | |
| "eval_precision": 0.80859375, | |
| "eval_recall": 0.4050880626223092, | |
| "eval_runtime": 30.0772, | |
| "eval_samples_per_second": 33.979, | |
| "eval_steps_per_second": 1.064, | |
| "step": 648 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "grad_norm": 0.8858753442764282, | |
| "learning_rate": 0.0, | |
| "loss": 0.316, | |
| "step": 729 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_accuracy": 0.649706457925636, | |
| "eval_f1": 0.5289473684210527, | |
| "eval_loss": 0.30977755784988403, | |
| "eval_precision": 0.8072289156626506, | |
| "eval_recall": 0.3933463796477495, | |
| "eval_runtime": 30.1686, | |
| "eval_samples_per_second": 33.876, | |
| "eval_steps_per_second": 1.061, | |
| "step": 729 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 729, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 9, | |
| "save_steps": 500, | |
| "total_flos": 2121256775520.0, | |
| "train_batch_size": 38, | |
| "trial_name": null, | |
| "trial_params": { | |
| "alpha": 0.462182377373202, | |
| "learning_rate": 2.997892231913295e-05, | |
| "num_train_epochs": 9, | |
| "per_device_train_batch_size": 38, | |
| "temperature": 10 | |
| } | |
| } | |