| { | |
| "best_metric": 9.06746912032304, | |
| "best_model_checkpoint": "./fine-tuned-whisper/openai/whisper-small/lr_1e-05_warmup_500_epochs_0_batch_32_grad_acc_1_max_steps_10000/dataset_myst/2023-09-09_11-20-19/checkpoint-1500", | |
| "epoch": 4.0, | |
| "eval_steps": 500, | |
| "global_step": 2500, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.16, | |
| "learning_rate": 1.94e-06, | |
| "loss": 1.7092, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "learning_rate": 3.94e-06, | |
| "loss": 0.6661, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "learning_rate": 5.94e-06, | |
| "loss": 0.4875, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "learning_rate": 7.94e-06, | |
| "loss": 0.4176, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "learning_rate": 9.940000000000001e-06, | |
| "loss": 0.3943, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "eval_loss": 0.23756489157676697, | |
| "eval_runtime": 611.1384, | |
| "eval_samples_per_second": 5.171, | |
| "eval_steps_per_second": 0.324, | |
| "eval_wer": 10.145123229361573, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "learning_rate": 9.897894736842107e-06, | |
| "loss": 0.3814, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.12, | |
| "learning_rate": 9.79263157894737e-06, | |
| "loss": 0.3181, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 1.28, | |
| "learning_rate": 9.687368421052632e-06, | |
| "loss": 0.3291, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.44, | |
| "learning_rate": 9.582105263157897e-06, | |
| "loss": 0.324, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "learning_rate": 9.476842105263158e-06, | |
| "loss": 0.3302, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "eval_loss": 0.22787314653396606, | |
| "eval_runtime": 601.1188, | |
| "eval_samples_per_second": 5.257, | |
| "eval_steps_per_second": 0.329, | |
| "eval_wer": 9.153630009709175, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.76, | |
| "learning_rate": 9.371578947368421e-06, | |
| "loss": 0.3232, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 1.92, | |
| "learning_rate": 9.266315789473685e-06, | |
| "loss": 0.3179, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 2.08, | |
| "learning_rate": 9.161052631578948e-06, | |
| "loss": 0.2809, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 2.24, | |
| "learning_rate": 9.055789473684211e-06, | |
| "loss": 0.2445, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "learning_rate": 8.950526315789474e-06, | |
| "loss": 0.2523, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "eval_loss": 0.23648202419281006, | |
| "eval_runtime": 599.2062, | |
| "eval_samples_per_second": 5.274, | |
| "eval_steps_per_second": 0.33, | |
| "eval_wer": 9.06746912032304, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 2.56, | |
| "learning_rate": 8.845263157894738e-06, | |
| "loss": 0.2481, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 2.72, | |
| "learning_rate": 8.740000000000001e-06, | |
| "loss": 0.2596, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 2.88, | |
| "learning_rate": 8.634736842105264e-06, | |
| "loss": 0.256, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 3.04, | |
| "learning_rate": 8.529473684210528e-06, | |
| "loss": 0.2419, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 3.2, | |
| "learning_rate": 8.42421052631579e-06, | |
| "loss": 0.1874, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 3.2, | |
| "eval_loss": 0.24421058595180511, | |
| "eval_runtime": 604.8436, | |
| "eval_samples_per_second": 5.224, | |
| "eval_steps_per_second": 0.327, | |
| "eval_wer": 9.441690296612077, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 3.36, | |
| "learning_rate": 8.318947368421052e-06, | |
| "loss": 0.1942, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 3.52, | |
| "learning_rate": 8.213684210526316e-06, | |
| "loss": 0.2072, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 3.68, | |
| "learning_rate": 8.10842105263158e-06, | |
| "loss": 0.1809, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 3.84, | |
| "learning_rate": 8.003157894736842e-06, | |
| "loss": 0.1791, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "learning_rate": 7.897894736842106e-06, | |
| "loss": 0.2087, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_loss": 0.2534370422363281, | |
| "eval_runtime": 605.6058, | |
| "eval_samples_per_second": 5.218, | |
| "eval_steps_per_second": 0.327, | |
| "eval_wer": 9.089973830237328, | |
| "step": 2500 | |
| } | |
| ], | |
| "logging_steps": 100, | |
| "max_steps": 10000, | |
| "num_train_epochs": 16, | |
| "save_steps": 500, | |
| "total_flos": 2.307182558478336e+19, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |