| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 4.9978245105148655, | |
| "global_step": 1720, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.29, | |
| "learning_rate": 7.151470588235293e-05, | |
| "loss": 5.1135, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "learning_rate": 6.710294117647058e-05, | |
| "loss": 3.0957, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "learning_rate": 6.269117647058824e-05, | |
| "loss": 3.0078, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.16, | |
| "learning_rate": 5.827941176470588e-05, | |
| "loss": 2.9785, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.45, | |
| "learning_rate": 5.3867647058823525e-05, | |
| "loss": 2.9069, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.45, | |
| "eval_loss": 2.9046826362609863, | |
| "eval_runtime": 128.7688, | |
| "eval_samples_per_second": 35.878, | |
| "eval_steps_per_second": 4.489, | |
| "eval_wer": 1.0, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.74, | |
| "learning_rate": 4.945588235294117e-05, | |
| "loss": 2.8537, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 2.03, | |
| "learning_rate": 4.504411764705882e-05, | |
| "loss": 2.7928, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 2.32, | |
| "learning_rate": 4.063235294117647e-05, | |
| "loss": 2.6988, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 2.61, | |
| "learning_rate": 3.6220588235294115e-05, | |
| "loss": 2.6342, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 2.91, | |
| "learning_rate": 3.180882352941176e-05, | |
| "loss": 2.5875, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 2.91, | |
| "eval_loss": 2.415876865386963, | |
| "eval_runtime": 164.5912, | |
| "eval_samples_per_second": 28.07, | |
| "eval_steps_per_second": 3.512, | |
| "eval_wer": 1.199687939790736, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 3.2, | |
| "learning_rate": 2.739705882352941e-05, | |
| "loss": 2.5214, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 3.49, | |
| "learning_rate": 2.2985294117647057e-05, | |
| "loss": 2.4174, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 3.78, | |
| "learning_rate": 1.8573529411764705e-05, | |
| "loss": 2.3271, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 4.07, | |
| "learning_rate": 1.416176470588235e-05, | |
| "loss": 2.2784, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 4.36, | |
| "learning_rate": 9.75e-06, | |
| "loss": 2.2043, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 4.36, | |
| "eval_loss": 1.6191935539245605, | |
| "eval_runtime": 127.3244, | |
| "eval_samples_per_second": 36.285, | |
| "eval_steps_per_second": 4.54, | |
| "eval_wer": 0.9597993024536499, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 4.65, | |
| "learning_rate": 5.338235294117647e-06, | |
| "loss": 2.1805, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 4.94, | |
| "learning_rate": 9.26470588235294e-07, | |
| "loss": 2.164, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "step": 1720, | |
| "total_flos": 5.074147598591555e+18, | |
| "train_loss": 2.7439405862675157, | |
| "train_runtime": 3003.9637, | |
| "train_samples_per_second": 18.359, | |
| "train_steps_per_second": 0.573 | |
| } | |
| ], | |
| "max_steps": 1720, | |
| "num_train_epochs": 5, | |
| "total_flos": 5.074147598591555e+18, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |