| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 100.0, | |
| "global_step": 4400, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 2.27, | |
| "learning_rate": 1.3579999999999999e-05, | |
| "loss": 11.686, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 4.55, | |
| "learning_rate": 2.758e-05, | |
| "loss": 4.1183, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 6.82, | |
| "learning_rate": 4.157999999999999e-05, | |
| "loss": 3.0783, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 9.09, | |
| "learning_rate": 5.558e-05, | |
| "loss": 2.7958, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 11.36, | |
| "learning_rate": 6.958e-05, | |
| "loss": 2.7064, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 11.36, | |
| "eval_loss": 2.711225986480713, | |
| "eval_runtime": 6.87, | |
| "eval_samples_per_second": 21.689, | |
| "eval_steps_per_second": 0.728, | |
| "eval_wer": 1.0, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 13.64, | |
| "learning_rate": 6.825897435897435e-05, | |
| "loss": 2.6552, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 15.91, | |
| "learning_rate": 6.646410256410256e-05, | |
| "loss": 2.5236, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 18.18, | |
| "learning_rate": 6.466923076923076e-05, | |
| "loss": 1.8116, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 20.45, | |
| "learning_rate": 6.287435897435897e-05, | |
| "loss": 1.4245, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 22.73, | |
| "learning_rate": 6.107948717948717e-05, | |
| "loss": 1.3079, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 22.73, | |
| "eval_loss": 0.7336559891700745, | |
| "eval_runtime": 6.7043, | |
| "eval_samples_per_second": 22.225, | |
| "eval_steps_per_second": 0.746, | |
| "eval_wer": 0.7775876817792986, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 25.0, | |
| "learning_rate": 5.928461538461538e-05, | |
| "loss": 1.2335, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 27.27, | |
| "learning_rate": 5.748974358974358e-05, | |
| "loss": 1.1818, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 29.55, | |
| "learning_rate": 5.569487179487179e-05, | |
| "loss": 1.1441, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 31.82, | |
| "learning_rate": 5.3899999999999996e-05, | |
| "loss": 1.1166, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 34.09, | |
| "learning_rate": 5.21051282051282e-05, | |
| "loss": 1.0919, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 34.09, | |
| "eval_loss": 0.5937860608100891, | |
| "eval_runtime": 6.763, | |
| "eval_samples_per_second": 22.032, | |
| "eval_steps_per_second": 0.739, | |
| "eval_wer": 0.7023096663815227, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 36.36, | |
| "learning_rate": 5.03102564102564e-05, | |
| "loss": 1.0684, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 38.64, | |
| "learning_rate": 4.853333333333333e-05, | |
| "loss": 1.0238, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 40.91, | |
| "learning_rate": 4.6738461538461535e-05, | |
| "loss": 1.0182, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 43.18, | |
| "learning_rate": 4.494358974358974e-05, | |
| "loss": 0.9795, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 45.45, | |
| "learning_rate": 4.314871794871794e-05, | |
| "loss": 0.9546, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 45.45, | |
| "eval_loss": 0.5698409080505371, | |
| "eval_runtime": 6.6992, | |
| "eval_samples_per_second": 22.241, | |
| "eval_steps_per_second": 0.746, | |
| "eval_wer": 0.613344739093242, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 47.73, | |
| "learning_rate": 4.135384615384616e-05, | |
| "loss": 0.9478, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 50.0, | |
| "learning_rate": 3.955897435897436e-05, | |
| "loss": 0.9329, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 52.27, | |
| "learning_rate": 3.776410256410256e-05, | |
| "loss": 0.9001, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 54.55, | |
| "learning_rate": 3.5969230769230765e-05, | |
| "loss": 0.8957, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 56.82, | |
| "learning_rate": 3.417435897435897e-05, | |
| "loss": 0.8895, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 56.82, | |
| "eval_loss": 0.5738843083381653, | |
| "eval_runtime": 6.762, | |
| "eval_samples_per_second": 22.035, | |
| "eval_steps_per_second": 0.739, | |
| "eval_wer": 0.6142001710863987, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 59.09, | |
| "learning_rate": 3.237948717948718e-05, | |
| "loss": 0.8718, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 61.36, | |
| "learning_rate": 3.058461538461538e-05, | |
| "loss": 0.8642, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 63.64, | |
| "learning_rate": 2.878974358974359e-05, | |
| "loss": 0.8443, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 65.91, | |
| "learning_rate": 2.6994871794871793e-05, | |
| "loss": 0.8272, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 68.18, | |
| "learning_rate": 2.5199999999999996e-05, | |
| "loss": 0.8152, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 68.18, | |
| "eval_loss": 0.5578901171684265, | |
| "eval_runtime": 6.7178, | |
| "eval_samples_per_second": 22.18, | |
| "eval_steps_per_second": 0.744, | |
| "eval_wer": 0.6090675791274593, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 70.45, | |
| "learning_rate": 2.3405128205128206e-05, | |
| "loss": 0.8141, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 72.73, | |
| "learning_rate": 2.161025641025641e-05, | |
| "loss": 0.8087, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 75.0, | |
| "learning_rate": 1.9833333333333332e-05, | |
| "loss": 0.7897, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 77.27, | |
| "learning_rate": 1.8038461538461535e-05, | |
| "loss": 0.7837, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 79.55, | |
| "learning_rate": 1.624358974358974e-05, | |
| "loss": 0.7703, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 79.55, | |
| "eval_loss": 0.5813038349151611, | |
| "eval_runtime": 6.7449, | |
| "eval_samples_per_second": 22.091, | |
| "eval_steps_per_second": 0.741, | |
| "eval_wer": 0.621043627031651, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 81.82, | |
| "learning_rate": 1.4448717948717947e-05, | |
| "loss": 0.7612, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 84.09, | |
| "learning_rate": 1.2653846153846152e-05, | |
| "loss": 0.7575, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 86.36, | |
| "learning_rate": 1.0858974358974358e-05, | |
| "loss": 0.7483, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 88.64, | |
| "learning_rate": 9.064102564102564e-06, | |
| "loss": 0.7496, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 90.91, | |
| "learning_rate": 7.269230769230769e-06, | |
| "loss": 0.732, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 90.91, | |
| "eval_loss": 0.5756056308746338, | |
| "eval_runtime": 6.9674, | |
| "eval_samples_per_second": 21.385, | |
| "eval_steps_per_second": 0.718, | |
| "eval_wer": 0.5859709153122327, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 93.18, | |
| "learning_rate": 5.4743589743589735e-06, | |
| "loss": 0.7262, | |
| "step": 4100 | |
| }, | |
| { | |
| "epoch": 95.45, | |
| "learning_rate": 3.6794871794871793e-06, | |
| "loss": 0.7307, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 97.73, | |
| "learning_rate": 1.8846153846153845e-06, | |
| "loss": 0.7265, | |
| "step": 4300 | |
| }, | |
| { | |
| "epoch": 100.0, | |
| "learning_rate": 1.0769230769230769e-07, | |
| "loss": 0.7226, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 100.0, | |
| "step": 4400, | |
| "total_flos": 1.9662171121467834e+19, | |
| "train_loss": 1.4620501448891379, | |
| "train_runtime": 7639.7148, | |
| "train_samples_per_second": 18.142, | |
| "train_steps_per_second": 0.576 | |
| } | |
| ], | |
| "max_steps": 4400, | |
| "num_train_epochs": 100, | |
| "total_flos": 1.9662171121467834e+19, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |