| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.243829468960359, | |
| "global_step": 3000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0, | |
| "learning_rate": 2.0000000000000002e-07, | |
| "loss": 18.6192, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "learning_rate": 9.94e-05, | |
| "loss": 7.5397, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "eval_cer": 0.9810978889654326, | |
| "eval_clean_cer": 0.9774387393628106, | |
| "eval_loss": 3.571558952331543, | |
| "eval_runtime": 221.4433, | |
| "eval_samples_per_second": 24.679, | |
| "eval_steps_per_second": 1.544, | |
| "eval_wer": 1.0, | |
| "learning_rate": 9.94e-05, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "learning_rate": 9.613830613830615e-05, | |
| "loss": 1.7478, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "eval_cer": 0.22960449951399964, | |
| "eval_clean_cer": 0.17457099250398456, | |
| "eval_loss": 0.7702281475067139, | |
| "eval_runtime": 224.0929, | |
| "eval_samples_per_second": 24.387, | |
| "eval_steps_per_second": 1.526, | |
| "eval_wer": 0.8097043024248205, | |
| "learning_rate": 9.613830613830615e-05, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.12, | |
| "learning_rate": 9.225330225330226e-05, | |
| "loss": 0.7687, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 1.12, | |
| "eval_cer": 0.14148744420432777, | |
| "eval_clean_cer": 0.11820726398251868, | |
| "eval_loss": 0.4996742308139801, | |
| "eval_runtime": 221.7353, | |
| "eval_samples_per_second": 24.647, | |
| "eval_steps_per_second": 1.542, | |
| "eval_wer": 0.5391836939605402, | |
| "learning_rate": 9.225330225330226e-05, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 1.5, | |
| "learning_rate": 8.836829836829838e-05, | |
| "loss": 0.6064, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 1.5, | |
| "eval_cer": 0.12378853794105713, | |
| "eval_clean_cer": 0.10011770602778763, | |
| "eval_loss": 0.4269550144672394, | |
| "eval_runtime": 224.0891, | |
| "eval_samples_per_second": 24.388, | |
| "eval_steps_per_second": 1.526, | |
| "eval_wer": 0.49563231085897885, | |
| "learning_rate": 8.836829836829838e-05, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 1.87, | |
| "learning_rate": 8.448329448329449e-05, | |
| "loss": 0.5473, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 1.87, | |
| "eval_cer": 0.1105108194060415, | |
| "eval_clean_cer": 0.0898226524968884, | |
| "eval_loss": 0.38087567687034607, | |
| "eval_runtime": 223.664, | |
| "eval_samples_per_second": 24.434, | |
| "eval_steps_per_second": 1.529, | |
| "eval_wer": 0.4489432200411667, | |
| "learning_rate": 8.448329448329449e-05, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 2.24, | |
| "learning_rate": 8.05982905982906e-05, | |
| "loss": 0.454, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 2.24, | |
| "eval_cer": 0.10214972585805014, | |
| "eval_clean_cer": 0.08129037344912453, | |
| "eval_loss": 0.3585045337677002, | |
| "eval_runtime": 225.448, | |
| "eval_samples_per_second": 24.241, | |
| "eval_steps_per_second": 1.517, | |
| "eval_wer": 0.4256237762939907, | |
| "learning_rate": 8.05982905982906e-05, | |
| "step": 3000 | |
| } | |
| ], | |
| "max_steps": 13370, | |
| "num_train_epochs": 10, | |
| "total_flos": 1.2697227005959725e+19, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |