| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 49.98039215686274, | |
| "global_step": 1250, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 3.98, | |
| "learning_rate": 3.675e-06, | |
| "loss": 20.0794, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 7.98, | |
| "learning_rate": 7.425e-06, | |
| "loss": 10.5776, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 11.98, | |
| "learning_rate": 1.1174999999999999e-05, | |
| "loss": 6.6033, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 15.98, | |
| "learning_rate": 1.4925e-05, | |
| "loss": 5.3857, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 19.98, | |
| "learning_rate": 1.8675e-05, | |
| "loss": 4.4431, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 19.98, | |
| "eval_loss": 4.144110202789307, | |
| "eval_runtime": 12.9526, | |
| "eval_samples_per_second": 26.327, | |
| "eval_steps_per_second": 3.32, | |
| "eval_wer": 1.0, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 23.98, | |
| "learning_rate": 2.2424999999999996e-05, | |
| "loss": 3.7383, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 27.98, | |
| "learning_rate": 2.6174999999999996e-05, | |
| "loss": 3.361, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 31.98, | |
| "learning_rate": 2.9925e-05, | |
| "loss": 3.2219, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 35.98, | |
| "learning_rate": 3.3675e-05, | |
| "loss": 3.104, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 39.98, | |
| "learning_rate": 3.7424999999999995e-05, | |
| "loss": 3.0399, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 39.98, | |
| "eval_loss": 2.9768528938293457, | |
| "eval_runtime": 12.8843, | |
| "eval_samples_per_second": 26.466, | |
| "eval_steps_per_second": 3.337, | |
| "eval_wer": 0.9988929889298893, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 43.98, | |
| "learning_rate": 4.1175e-05, | |
| "loss": 2.9893, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 47.98, | |
| "learning_rate": 4.4924999999999994e-05, | |
| "loss": 2.953, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 49.98, | |
| "step": 1250, | |
| "total_flos": 4.919672600088422e+18, | |
| "train_loss": 5.6772947265625, | |
| "train_runtime": 2084.0288, | |
| "train_samples_per_second": 19.434, | |
| "train_steps_per_second": 0.6 | |
| } | |
| ], | |
| "max_steps": 1250, | |
| "num_train_epochs": 50, | |
| "total_flos": 4.919672600088422e+18, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |