| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 18.46153846153846, | |
| "global_step": 1200, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 3.08, | |
| "learning_rate": 0.00011999999999999999, | |
| "loss": 5.7855, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 3.08, | |
| "eval_loss": 3.094594717025757, | |
| "eval_runtime": 32.9209, | |
| "eval_samples_per_second": 7.017, | |
| "eval_wer": 1.0, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 6.15, | |
| "learning_rate": 0.00023999999999999998, | |
| "loss": 3.0293, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 6.15, | |
| "eval_loss": 3.005324125289917, | |
| "eval_runtime": 32.0726, | |
| "eval_samples_per_second": 7.202, | |
| "eval_wer": 1.0, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 9.23, | |
| "learning_rate": 0.0002625, | |
| "loss": 1.9629, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 9.23, | |
| "eval_loss": 1.0649160146713257, | |
| "eval_runtime": 32.4082, | |
| "eval_samples_per_second": 7.128, | |
| "eval_wer": 0.7503426222019187, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 12.31, | |
| "learning_rate": 0.00018749999999999998, | |
| "loss": 0.9013, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 12.31, | |
| "eval_loss": 0.8863689303398132, | |
| "eval_runtime": 32.197, | |
| "eval_samples_per_second": 7.175, | |
| "eval_wer": 0.6196893558702604, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 15.38, | |
| "learning_rate": 0.0001125, | |
| "loss": 0.6283, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 15.38, | |
| "eval_loss": 0.8295900821685791, | |
| "eval_runtime": 32.1739, | |
| "eval_samples_per_second": 7.18, | |
| "eval_wer": 0.5616719963453631, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 18.46, | |
| "learning_rate": 3.75e-05, | |
| "loss": 0.4995, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 18.46, | |
| "eval_loss": 0.8448638319969177, | |
| "eval_runtime": 32.5186, | |
| "eval_samples_per_second": 7.104, | |
| "eval_wer": 0.5415714938328003, | |
| "step": 1200 | |
| } | |
| ], | |
| "max_steps": 1300, | |
| "num_train_epochs": 20, | |
| "total_flos": 7.999595747693107e+18, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |