| { | |
| "best_metric": 70.1955074875208, | |
| "best_model_checkpoint": "./whisper-it/checkpoint-300", | |
| "epoch": 2.099236641221374, | |
| "eval_steps": 100, | |
| "global_step": 1100, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 4.4e-07, | |
| "loss": 3.4673, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "learning_rate": 9.400000000000001e-07, | |
| "loss": 2.7417, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "learning_rate": 1.44e-06, | |
| "loss": 2.0307, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "learning_rate": 1.94e-06, | |
| "loss": 1.2496, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "eval_loss": 1.233077049255371, | |
| "eval_runtime": 1969.1907, | |
| "eval_samples_per_second": 0.762, | |
| "eval_steps_per_second": 0.095, | |
| "eval_wer": 176.42124237382143, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "learning_rate": 2.4000000000000003e-06, | |
| "loss": 0.9794, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "learning_rate": 2.9e-06, | |
| "loss": 0.8642, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "learning_rate": 3.4000000000000005e-06, | |
| "loss": 0.8044, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "learning_rate": 3.88e-06, | |
| "loss": 0.7389, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "eval_loss": 0.8330782055854797, | |
| "eval_runtime": 1755.11, | |
| "eval_samples_per_second": 0.855, | |
| "eval_steps_per_second": 0.107, | |
| "eval_wer": 80.49084858569051, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "learning_rate": 4.38e-06, | |
| "loss": 0.6293, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "learning_rate": 4.880000000000001e-06, | |
| "loss": 0.5066, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "learning_rate": 5.380000000000001e-06, | |
| "loss": 0.3526, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "learning_rate": 5.8800000000000005e-06, | |
| "loss": 0.2951, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "eval_loss": 0.4260523319244385, | |
| "eval_runtime": 1708.1953, | |
| "eval_samples_per_second": 0.878, | |
| "eval_steps_per_second": 0.11, | |
| "eval_wer": 70.1955074875208, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "learning_rate": 6.380000000000001e-06, | |
| "loss": 0.2553, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "learning_rate": 6.88e-06, | |
| "loss": 0.2744, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "learning_rate": 7.3800000000000005e-06, | |
| "loss": 0.279, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "learning_rate": 7.88e-06, | |
| "loss": 0.2703, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "eval_loss": 0.40512609481811523, | |
| "eval_runtime": 1849.1681, | |
| "eval_samples_per_second": 0.811, | |
| "eval_steps_per_second": 0.102, | |
| "eval_wer": 101.62922906267333, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "learning_rate": 8.380000000000001e-06, | |
| "loss": 0.2514, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "learning_rate": 8.880000000000001e-06, | |
| "loss": 0.2826, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "learning_rate": 9.38e-06, | |
| "loss": 0.2381, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "learning_rate": 9.88e-06, | |
| "loss": 0.2491, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "eval_loss": 0.3922654092311859, | |
| "eval_runtime": 1735.7941, | |
| "eval_samples_per_second": 0.864, | |
| "eval_steps_per_second": 0.108, | |
| "eval_wer": 112.23655019412091, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "learning_rate": 9.945714285714286e-06, | |
| "loss": 0.2621, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 1.05, | |
| "learning_rate": 9.874285714285715e-06, | |
| "loss": 0.1729, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 1.1, | |
| "learning_rate": 9.802857142857144e-06, | |
| "loss": 0.1725, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 1.15, | |
| "learning_rate": 9.731428571428573e-06, | |
| "loss": 0.17, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.15, | |
| "eval_loss": 0.3860141336917877, | |
| "eval_runtime": 1739.7902, | |
| "eval_samples_per_second": 0.862, | |
| "eval_steps_per_second": 0.108, | |
| "eval_wer": 107.05074875207987, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.19, | |
| "learning_rate": 9.66e-06, | |
| "loss": 0.1744, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 1.24, | |
| "learning_rate": 9.58857142857143e-06, | |
| "loss": 0.1843, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 1.29, | |
| "learning_rate": 9.517142857142859e-06, | |
| "loss": 0.1708, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 1.34, | |
| "learning_rate": 9.445714285714288e-06, | |
| "loss": 0.1603, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 1.34, | |
| "eval_loss": 0.38356906175613403, | |
| "eval_runtime": 1678.1272, | |
| "eval_samples_per_second": 0.894, | |
| "eval_steps_per_second": 0.112, | |
| "eval_wer": 90.3563505268996, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 1.38, | |
| "learning_rate": 9.374285714285715e-06, | |
| "loss": 0.1626, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 1.43, | |
| "learning_rate": 9.302857142857144e-06, | |
| "loss": 0.1674, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 1.48, | |
| "learning_rate": 9.231428571428573e-06, | |
| "loss": 0.1717, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 1.53, | |
| "learning_rate": 9.16e-06, | |
| "loss": 0.1607, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.53, | |
| "eval_loss": 0.3789268732070923, | |
| "eval_runtime": 1834.2673, | |
| "eval_samples_per_second": 0.818, | |
| "eval_steps_per_second": 0.102, | |
| "eval_wer": 134.98336106489182, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.57, | |
| "learning_rate": 9.08857142857143e-06, | |
| "loss": 0.1636, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 1.62, | |
| "learning_rate": 9.017142857142858e-06, | |
| "loss": 0.1625, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 1.67, | |
| "learning_rate": 8.945714285714286e-06, | |
| "loss": 0.1762, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 1.72, | |
| "learning_rate": 8.874285714285715e-06, | |
| "loss": 0.154, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 1.72, | |
| "eval_loss": 0.3783314824104309, | |
| "eval_runtime": 1787.7259, | |
| "eval_samples_per_second": 0.839, | |
| "eval_steps_per_second": 0.105, | |
| "eval_wer": 99.0501941209096, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 1.77, | |
| "learning_rate": 8.802857142857144e-06, | |
| "loss": 0.1703, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 1.81, | |
| "learning_rate": 8.731428571428571e-06, | |
| "loss": 0.1593, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 1.86, | |
| "learning_rate": 8.66e-06, | |
| "loss": 0.1648, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 1.91, | |
| "learning_rate": 8.588571428571429e-06, | |
| "loss": 0.1562, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.91, | |
| "eval_loss": 0.36686423420906067, | |
| "eval_runtime": 1815.5283, | |
| "eval_samples_per_second": 0.826, | |
| "eval_steps_per_second": 0.104, | |
| "eval_wer": 98.31530782029951, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.96, | |
| "learning_rate": 8.517142857142858e-06, | |
| "loss": 0.1751, | |
| "step": 1025 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "learning_rate": 8.445714285714285e-06, | |
| "loss": 0.1593, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 2.05, | |
| "learning_rate": 8.374285714285714e-06, | |
| "loss": 0.0693, | |
| "step": 1075 | |
| }, | |
| { | |
| "epoch": 2.1, | |
| "learning_rate": 8.302857142857143e-06, | |
| "loss": 0.0723, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 2.1, | |
| "eval_loss": 0.3756558299064636, | |
| "eval_runtime": 1775.3191, | |
| "eval_samples_per_second": 0.845, | |
| "eval_steps_per_second": 0.106, | |
| "eval_wer": 158.88103161397672, | |
| "step": 1100 | |
| } | |
| ], | |
| "logging_steps": 25, | |
| "max_steps": 4000, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 8, | |
| "save_steps": 100, | |
| "total_flos": 1.015416589565952e+19, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |