| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 20.0, | |
| "global_step": 2300, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.87, | |
| "learning_rate": 1.3439999999999998e-05, | |
| "loss": 12.0308, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.74, | |
| "learning_rate": 2.7439999999999998e-05, | |
| "loss": 4.6075, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 2.61, | |
| "learning_rate": 4.1439999999999996e-05, | |
| "loss": 3.4071, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 3.48, | |
| "learning_rate": 5.544e-05, | |
| "loss": 3.0602, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 4.35, | |
| "learning_rate": 6.944e-05, | |
| "loss": 3.0067, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 4.35, | |
| "eval_loss": 2.9632411003112793, | |
| "eval_runtime": 75.847, | |
| "eval_samples_per_second": 22.624, | |
| "eval_steps_per_second": 0.712, | |
| "eval_wer": 1.0, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 5.22, | |
| "learning_rate": 6.626666666666666e-05, | |
| "loss": 2.9665, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 6.09, | |
| "learning_rate": 6.237777777777777e-05, | |
| "loss": 2.8916, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 6.96, | |
| "learning_rate": 5.8488888888888886e-05, | |
| "loss": 2.7907, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 7.83, | |
| "learning_rate": 5.46e-05, | |
| "loss": 2.0776, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 8.7, | |
| "learning_rate": 5.0711111111111105e-05, | |
| "loss": 1.4939, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 8.7, | |
| "eval_loss": 0.5004750490188599, | |
| "eval_runtime": 75.9315, | |
| "eval_samples_per_second": 22.599, | |
| "eval_steps_per_second": 0.711, | |
| "eval_wer": 0.41569049578927375, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 9.57, | |
| "learning_rate": 4.682222222222222e-05, | |
| "loss": 1.2841, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 10.43, | |
| "learning_rate": 4.2933333333333324e-05, | |
| "loss": 1.1469, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 11.3, | |
| "learning_rate": 3.9044444444444444e-05, | |
| "loss": 1.0776, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 12.17, | |
| "learning_rate": 3.515555555555555e-05, | |
| "loss": 1.0147, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 13.04, | |
| "learning_rate": 3.126666666666666e-05, | |
| "loss": 0.9982, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 13.04, | |
| "eval_loss": 0.19668780267238617, | |
| "eval_runtime": 75.4326, | |
| "eval_samples_per_second": 22.749, | |
| "eval_steps_per_second": 0.716, | |
| "eval_wer": 0.185651871082125, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 13.91, | |
| "learning_rate": 2.7377777777777776e-05, | |
| "loss": 0.9517, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 14.78, | |
| "learning_rate": 2.3488888888888886e-05, | |
| "loss": 0.9291, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 15.65, | |
| "learning_rate": 1.96e-05, | |
| "loss": 0.8937, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 16.52, | |
| "learning_rate": 1.5711111111111108e-05, | |
| "loss": 0.8918, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 17.39, | |
| "learning_rate": 1.1822222222222221e-05, | |
| "loss": 0.8726, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 17.39, | |
| "eval_loss": 0.1586812287569046, | |
| "eval_runtime": 75.8623, | |
| "eval_samples_per_second": 22.62, | |
| "eval_steps_per_second": 0.712, | |
| "eval_wer": 0.1563984043563604, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 18.26, | |
| "learning_rate": 7.933333333333332e-06, | |
| "loss": 0.8785, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 19.13, | |
| "learning_rate": 4.044444444444444e-06, | |
| "loss": 0.8559, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "learning_rate": 1.5555555555555554e-07, | |
| "loss": 0.8653, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "step": 2300, | |
| "total_flos": 1.0844366081729495e+19, | |
| "train_loss": 2.2170696955141813, | |
| "train_runtime": 4486.5584, | |
| "train_samples_per_second": 16.36, | |
| "train_steps_per_second": 0.513 | |
| } | |
| ], | |
| "max_steps": 2300, | |
| "num_train_epochs": 20, | |
| "total_flos": 1.0844366081729495e+19, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |