| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.4, | |
| "eval_steps": 500, | |
| "global_step": 2000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 1.9600000000000002e-05, | |
| "loss": 1.5719, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 3.960000000000001e-05, | |
| "loss": 0.3718, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "learning_rate": 5.96e-05, | |
| "loss": 0.3261, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "learning_rate": 7.960000000000001e-05, | |
| "loss": 0.3219, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "learning_rate": 9.960000000000001e-05, | |
| "loss": 0.3252, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "eval_combined_wer": 0.19175330812854444, | |
| "eval_f1_score": 0.6148314606741573, | |
| "eval_label_f1": 0.7577528089887641, | |
| "eval_loss": 0.3395695686340332, | |
| "eval_runtime": 98.1651, | |
| "eval_samples_per_second": 10.187, | |
| "eval_steps_per_second": 0.163, | |
| "eval_wer": 0.11928060336491975, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "learning_rate": 9.98830238119205e-05, | |
| "loss": 0.321, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "learning_rate": 9.952307128483256e-05, | |
| "loss": 0.3095, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "learning_rate": 9.892184733248666e-05, | |
| "loss": 0.2932, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "learning_rate": 9.808228105754376e-05, | |
| "loss": 0.2845, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "learning_rate": 9.700846274250251e-05, | |
| "loss": 0.2729, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "eval_combined_wer": 0.17300724637681159, | |
| "eval_f1_score": 0.6449152542372881, | |
| "eval_label_f1": 0.7906779661016949, | |
| "eval_loss": 0.31582510471343994, | |
| "eval_runtime": 99.4294, | |
| "eval_samples_per_second": 10.057, | |
| "eval_steps_per_second": 0.161, | |
| "eval_wer": 0.10578224714755366, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "learning_rate": 9.570562392225396e-05, | |
| "loss": 0.2719, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "learning_rate": 9.418011189656941e-05, | |
| "loss": 0.2593, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "learning_rate": 9.24393588066941e-05, | |
| "loss": 0.2545, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "learning_rate": 9.049184542670199e-05, | |
| "loss": 0.2467, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "learning_rate": 8.834705984601708e-05, | |
| "loss": 0.2369, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "eval_combined_wer": 0.1735586011342155, | |
| "eval_f1_score": 0.6916524701873935, | |
| "eval_label_f1": 0.8083475298126065, | |
| "eval_loss": 0.29707884788513184, | |
| "eval_runtime": 99.4491, | |
| "eval_samples_per_second": 10.055, | |
| "eval_steps_per_second": 0.161, | |
| "eval_wer": 0.10667182363179269, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "learning_rate": 8.601545124439535e-05, | |
| "loss": 0.2262, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "learning_rate": 8.350837898457143e-05, | |
| "loss": 0.2196, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "learning_rate": 8.083805727058513e-05, | |
| "loss": 0.2168, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "learning_rate": 7.801749564140724e-05, | |
| "loss": 0.2063, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "learning_rate": 7.506043558977321e-05, | |
| "loss": 0.1967, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "eval_combined_wer": 0.1633585381222432, | |
| "eval_f1_score": 0.6915401301518438, | |
| "eval_label_f1": 0.8095444685466376, | |
| "eval_loss": 0.282340943813324, | |
| "eval_runtime": 99.8374, | |
| "eval_samples_per_second": 10.016, | |
| "eval_steps_per_second": 0.16, | |
| "eval_wer": 0.09990330690388706, | |
| "step": 2000 | |
| } | |
| ], | |
| "logging_steps": 100, | |
| "max_steps": 5000, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 9223372036854775807, | |
| "save_steps": 500, | |
| "total_flos": 7.387786248192e+19, | |
| "train_batch_size": 64, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |