| { | |
| "best_metric": 7.581081081081082, | |
| "best_model_checkpoint": "./checkpoint-9000", | |
| "epoch": 5.48847420417124, | |
| "global_step": 10000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.27, | |
| "learning_rate": 9.9e-06, | |
| "loss": 0.365, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "learning_rate": 9.478947368421053e-06, | |
| "loss": 0.1912, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "eval_cer": 7.035654676637185, | |
| "eval_loss": 0.18279600143432617, | |
| "eval_runtime": 7736.1561, | |
| "eval_samples_per_second": 0.595, | |
| "eval_steps_per_second": 0.074, | |
| "eval_wer": 11.23141891891892, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "learning_rate": 8.95263157894737e-06, | |
| "loss": 0.1649, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 1.1, | |
| "learning_rate": 8.426315789473684e-06, | |
| "loss": 0.1329, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 1.1, | |
| "eval_cer": 5.9028005940255746, | |
| "eval_loss": 0.16182316839694977, | |
| "eval_runtime": 7650.4506, | |
| "eval_samples_per_second": 0.602, | |
| "eval_steps_per_second": 0.075, | |
| "eval_wer": 9.41722972972973, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 1.37, | |
| "learning_rate": 7.9e-06, | |
| "loss": 0.0954, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 1.65, | |
| "learning_rate": 7.373684210526316e-06, | |
| "loss": 0.0912, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 1.65, | |
| "eval_cer": 5.47108067971245, | |
| "eval_loss": 0.16157789528369904, | |
| "eval_runtime": 7517.9849, | |
| "eval_samples_per_second": 0.612, | |
| "eval_steps_per_second": 0.077, | |
| "eval_wer": 8.925675675675675, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 1.92, | |
| "learning_rate": 6.8473684210526325e-06, | |
| "loss": 0.0887, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 2.2, | |
| "learning_rate": 6.321052631578948e-06, | |
| "loss": 0.0576, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 2.2, | |
| "eval_cer": 5.305489479701936, | |
| "eval_loss": 0.1663784682750702, | |
| "eval_runtime": 7600.817, | |
| "eval_samples_per_second": 0.606, | |
| "eval_steps_per_second": 0.076, | |
| "eval_wer": 8.58614864864865, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 2.47, | |
| "learning_rate": 5.794736842105264e-06, | |
| "loss": 0.047, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 2.74, | |
| "learning_rate": 5.269473684210527e-06, | |
| "loss": 0.0449, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 2.74, | |
| "eval_cer": 5.2930044289074925, | |
| "eval_loss": 0.16421984136104584, | |
| "eval_runtime": 7476.5946, | |
| "eval_samples_per_second": 0.616, | |
| "eval_steps_per_second": 0.077, | |
| "eval_wer": 8.451013513513514, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 3.02, | |
| "learning_rate": 4.743157894736842e-06, | |
| "loss": 0.0431, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 3.29, | |
| "learning_rate": 4.217894736842106e-06, | |
| "loss": 0.02, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 3.29, | |
| "eval_cer": 5.035418117780027, | |
| "eval_loss": 0.17986369132995605, | |
| "eval_runtime": 7470.6199, | |
| "eval_samples_per_second": 0.616, | |
| "eval_steps_per_second": 0.077, | |
| "eval_wer": 8.153716216216218, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 3.57, | |
| "learning_rate": 3.6915789473684216e-06, | |
| "loss": 0.0207, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 3.84, | |
| "learning_rate": 3.1652631578947375e-06, | |
| "loss": 0.019, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 3.84, | |
| "eval_cer": 5.082729889211602, | |
| "eval_loss": 0.180050328373909, | |
| "eval_runtime": 7452.1074, | |
| "eval_samples_per_second": 0.618, | |
| "eval_steps_per_second": 0.077, | |
| "eval_wer": 8.125, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 4.12, | |
| "learning_rate": 2.6389473684210526e-06, | |
| "loss": 0.014, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 4.39, | |
| "learning_rate": 2.1126315789473685e-06, | |
| "loss": 0.0067, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 4.39, | |
| "eval_cer": 4.8133156352262425, | |
| "eval_loss": 0.20030897855758667, | |
| "eval_runtime": 7521.4475, | |
| "eval_samples_per_second": 0.612, | |
| "eval_steps_per_second": 0.077, | |
| "eval_wer": 7.841216216216217, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 4.67, | |
| "learning_rate": 1.5863157894736845e-06, | |
| "loss": 0.0062, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 4.94, | |
| "learning_rate": 1.06e-06, | |
| "loss": 0.006, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 4.94, | |
| "eval_cer": 4.70226439394935, | |
| "eval_loss": 0.2070685774087906, | |
| "eval_runtime": 7515.5986, | |
| "eval_samples_per_second": 0.613, | |
| "eval_steps_per_second": 0.077, | |
| "eval_wer": 7.581081081081082, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 5.21, | |
| "learning_rate": 5.336842105263159e-07, | |
| "loss": 0.0029, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 5.49, | |
| "learning_rate": 7.3684210526315796e-09, | |
| "loss": 0.0022, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 5.49, | |
| "eval_cer": 4.718692092363091, | |
| "eval_loss": 0.2283746749162674, | |
| "eval_runtime": 7466.269, | |
| "eval_samples_per_second": 0.617, | |
| "eval_steps_per_second": 0.077, | |
| "eval_wer": 7.64527027027027, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 5.49, | |
| "step": 10000, | |
| "total_flos": 3.39643668197376e+20, | |
| "train_loss": 0.07097679934501648, | |
| "train_runtime": 114325.2073, | |
| "train_samples_per_second": 1.4, | |
| "train_steps_per_second": 0.087 | |
| } | |
| ], | |
| "max_steps": 10000, | |
| "num_train_epochs": 6, | |
| "total_flos": 3.39643668197376e+20, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |