| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 99.98461538461538, | |
| "global_step": 3200, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 3.12, | |
| "learning_rate": 1.9200000000000003e-06, | |
| "loss": 10.7305, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 6.25, | |
| "learning_rate": 3.920000000000001e-06, | |
| "loss": 3.0098, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 9.37, | |
| "learning_rate": 5.92e-06, | |
| "loss": 2.9327, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 12.49, | |
| "learning_rate": 7.92e-06, | |
| "loss": 2.8216, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 15.62, | |
| "learning_rate": 9.920000000000002e-06, | |
| "loss": 2.3731, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 15.62, | |
| "eval_loss": 1.5517226457595825, | |
| "eval_runtime": 24.8327, | |
| "eval_samples_per_second": 20.497, | |
| "eval_steps_per_second": 1.289, | |
| "eval_wer": 0.9499121265377856, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 18.74, | |
| "learning_rate": 1.1920000000000001e-05, | |
| "loss": 1.9105, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 21.86, | |
| "learning_rate": 1.392e-05, | |
| "loss": 1.714, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 24.98, | |
| "learning_rate": 1.5920000000000003e-05, | |
| "loss": 1.5476, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 28.12, | |
| "learning_rate": 1.792e-05, | |
| "loss": 1.4238, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 31.25, | |
| "learning_rate": 1.9920000000000002e-05, | |
| "loss": 1.3312, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 31.25, | |
| "eval_loss": 0.8717297911643982, | |
| "eval_runtime": 24.7966, | |
| "eval_samples_per_second": 20.527, | |
| "eval_steps_per_second": 1.29, | |
| "eval_wer": 0.6189220855301699, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 34.37, | |
| "learning_rate": 1.912727272727273e-05, | |
| "loss": 1.2049, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 37.49, | |
| "learning_rate": 1.821818181818182e-05, | |
| "loss": 1.1346, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 40.62, | |
| "learning_rate": 1.730909090909091e-05, | |
| "loss": 1.0533, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 43.74, | |
| "learning_rate": 1.64e-05, | |
| "loss": 0.9638, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 46.86, | |
| "learning_rate": 1.549090909090909e-05, | |
| "loss": 0.9135, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 46.86, | |
| "eval_loss": 0.8298946619033813, | |
| "eval_runtime": 24.721, | |
| "eval_samples_per_second": 20.59, | |
| "eval_steps_per_second": 1.294, | |
| "eval_wer": 0.5310486233157586, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 49.98, | |
| "learning_rate": 1.4581818181818184e-05, | |
| "loss": 0.8568, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 53.12, | |
| "learning_rate": 1.3672727272727273e-05, | |
| "loss": 0.8141, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 56.25, | |
| "learning_rate": 1.2763636363636365e-05, | |
| "loss": 0.7526, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 59.37, | |
| "learning_rate": 1.1854545454545457e-05, | |
| "loss": 0.7177, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 62.49, | |
| "learning_rate": 1.0945454545454545e-05, | |
| "loss": 0.6719, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 62.49, | |
| "eval_loss": 0.8842366933822632, | |
| "eval_runtime": 25.0435, | |
| "eval_samples_per_second": 20.325, | |
| "eval_steps_per_second": 1.278, | |
| "eval_wer": 0.5043936731107206, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 65.62, | |
| "learning_rate": 1.0036363636363637e-05, | |
| "loss": 0.6552, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 68.74, | |
| "learning_rate": 9.127272727272727e-06, | |
| "loss": 0.6145, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 71.86, | |
| "learning_rate": 8.21818181818182e-06, | |
| "loss": 0.596, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 74.98, | |
| "learning_rate": 7.30909090909091e-06, | |
| "loss": 0.5719, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 78.12, | |
| "learning_rate": 6.4000000000000006e-06, | |
| "loss": 0.5583, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 78.12, | |
| "eval_loss": 0.9093144536018372, | |
| "eval_runtime": 24.6074, | |
| "eval_samples_per_second": 20.685, | |
| "eval_steps_per_second": 1.3, | |
| "eval_wer": 0.4800820152314001, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 81.25, | |
| "learning_rate": 5.490909090909091e-06, | |
| "loss": 0.5417, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 84.37, | |
| "learning_rate": 4.581818181818183e-06, | |
| "loss": 0.5241, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 87.49, | |
| "learning_rate": 3.672727272727273e-06, | |
| "loss": 0.4901, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 90.62, | |
| "learning_rate": 2.763636363636364e-06, | |
| "loss": 0.4882, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 93.74, | |
| "learning_rate": 1.8545454545454546e-06, | |
| "loss": 0.4728, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 93.74, | |
| "eval_loss": 0.9488239884376526, | |
| "eval_runtime": 24.6884, | |
| "eval_samples_per_second": 20.617, | |
| "eval_steps_per_second": 1.296, | |
| "eval_wer": 0.48125366139425896, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 96.86, | |
| "learning_rate": 9.454545454545455e-07, | |
| "loss": 0.4682, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 99.98, | |
| "learning_rate": 3.636363636363637e-08, | |
| "loss": 0.4634, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 99.98, | |
| "step": 3200, | |
| "total_flos": 3.840629068156852e+19, | |
| "train_loss": 1.4163260304927825, | |
| "train_runtime": 8387.3816, | |
| "train_samples_per_second": 12.34, | |
| "train_steps_per_second": 0.382 | |
| } | |
| ], | |
| "max_steps": 3200, | |
| "num_train_epochs": 100, | |
| "total_flos": 3.840629068156852e+19, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |