| { | |
| "best_metric": 0.6574257425742575, | |
| "best_model_checkpoint": "tiny-bert-sst2-distilled/run-1/checkpoint-2079", | |
| "epoch": 9.0, | |
| "eval_steps": 500, | |
| "global_step": 2673, | |
| "is_hyper_param_search": true, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.8816747665405273, | |
| "learning_rate": 0.0001319007893143593, | |
| "loss": 0.4812, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.5188118811881188, | |
| "eval_loss": 0.4719693958759308, | |
| "eval_runtime": 14.5929, | |
| "eval_samples_per_second": 34.606, | |
| "eval_steps_per_second": 1.096, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 0.699450671672821, | |
| "learning_rate": 0.00011724514605720826, | |
| "loss": 0.4696, | |
| "step": 594 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.5801980198019802, | |
| "eval_loss": 0.46854168176651, | |
| "eval_runtime": 15.1826, | |
| "eval_samples_per_second": 33.262, | |
| "eval_steps_per_second": 1.054, | |
| "step": 594 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 1.2772634029388428, | |
| "learning_rate": 0.00010258950280005722, | |
| "loss": 0.4576, | |
| "step": 891 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.6336633663366337, | |
| "eval_loss": 0.46670353412628174, | |
| "eval_runtime": 14.6186, | |
| "eval_samples_per_second": 34.545, | |
| "eval_steps_per_second": 1.094, | |
| "step": 891 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 1.0283604860305786, | |
| "learning_rate": 8.793385954290618e-05, | |
| "loss": 0.4458, | |
| "step": 1188 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.6356435643564357, | |
| "eval_loss": 0.4543105363845825, | |
| "eval_runtime": 15.6547, | |
| "eval_samples_per_second": 32.259, | |
| "eval_steps_per_second": 1.022, | |
| "step": 1188 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 1.320448398590088, | |
| "learning_rate": 7.327821628575516e-05, | |
| "loss": 0.4413, | |
| "step": 1485 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.6415841584158416, | |
| "eval_loss": 0.4486277401447296, | |
| "eval_runtime": 14.9223, | |
| "eval_samples_per_second": 33.842, | |
| "eval_steps_per_second": 1.072, | |
| "step": 1485 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "grad_norm": 1.926311731338501, | |
| "learning_rate": 5.862257302860413e-05, | |
| "loss": 0.4367, | |
| "step": 1782 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.6336633663366337, | |
| "eval_loss": 0.44990047812461853, | |
| "eval_runtime": 15.0105, | |
| "eval_samples_per_second": 33.643, | |
| "eval_steps_per_second": 1.066, | |
| "step": 1782 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "grad_norm": 1.3797311782836914, | |
| "learning_rate": 4.396692977145309e-05, | |
| "loss": 0.4336, | |
| "step": 2079 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.6574257425742575, | |
| "eval_loss": 0.4465705454349518, | |
| "eval_runtime": 14.9751, | |
| "eval_samples_per_second": 33.723, | |
| "eval_steps_per_second": 1.068, | |
| "step": 2079 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 1.7872868776321411, | |
| "learning_rate": 2.9311286514302065e-05, | |
| "loss": 0.4305, | |
| "step": 2376 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.6534653465346535, | |
| "eval_loss": 0.44232580065727234, | |
| "eval_runtime": 15.2623, | |
| "eval_samples_per_second": 33.088, | |
| "eval_steps_per_second": 1.048, | |
| "step": 2376 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "grad_norm": 1.2580420970916748, | |
| "learning_rate": 1.4655643257151032e-05, | |
| "loss": 0.4287, | |
| "step": 2673 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_accuracy": 0.6574257425742575, | |
| "eval_loss": 0.4441685378551483, | |
| "eval_runtime": 17.0285, | |
| "eval_samples_per_second": 29.656, | |
| "eval_steps_per_second": 0.94, | |
| "step": 2673 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 2970, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 10, | |
| "save_steps": 500, | |
| "total_flos": 7000589548260.0, | |
| "train_batch_size": 32, | |
| "trial_name": null, | |
| "trial_params": { | |
| "alpha": 0.6312656697409034, | |
| "learning_rate": 0.00014655643257151032, | |
| "num_train_epochs": 10, | |
| "temperature": 21 | |
| } | |
| } | |