| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 15.0, | |
| "global_step": 1635, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.92, | |
| "eval_loss": 3.5559017658233643, | |
| "eval_runtime": 96.7485, | |
| "eval_samples_per_second": 17.024, | |
| "eval_steps_per_second": 2.129, | |
| "eval_wer": 1.0, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.83, | |
| "eval_loss": 3.0160555839538574, | |
| "eval_runtime": 94.1056, | |
| "eval_samples_per_second": 17.502, | |
| "eval_steps_per_second": 2.189, | |
| "eval_wer": 0.9998978653865795, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 2.75, | |
| "eval_loss": 0.858742356300354, | |
| "eval_runtime": 93.3532, | |
| "eval_samples_per_second": 17.643, | |
| "eval_steps_per_second": 2.207, | |
| "eval_wer": 0.7442549279950975, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 3.67, | |
| "eval_loss": 0.5855284333229065, | |
| "eval_runtime": 95.5853, | |
| "eval_samples_per_second": 17.231, | |
| "eval_steps_per_second": 2.155, | |
| "eval_wer": 0.6120927382289858, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 4.59, | |
| "learning_rate": 0.0003, | |
| "loss": 3.1095, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 4.59, | |
| "eval_loss": 0.4840559959411621, | |
| "eval_runtime": 95.1562, | |
| "eval_samples_per_second": 17.308, | |
| "eval_steps_per_second": 2.165, | |
| "eval_wer": 0.5203758553773874, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 5.5, | |
| "eval_loss": 0.453325480222702, | |
| "eval_runtime": 94.8597, | |
| "eval_samples_per_second": 17.362, | |
| "eval_steps_per_second": 2.172, | |
| "eval_wer": 0.49228883668675316, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 6.42, | |
| "eval_loss": 0.41565799713134766, | |
| "eval_runtime": 94.1725, | |
| "eval_samples_per_second": 17.489, | |
| "eval_steps_per_second": 2.187, | |
| "eval_wer": 0.43417424165049534, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 7.34, | |
| "eval_loss": 0.4304071068763733, | |
| "eval_runtime": 94.2253, | |
| "eval_samples_per_second": 17.479, | |
| "eval_steps_per_second": 2.186, | |
| "eval_wer": 0.43335716474313146, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 8.26, | |
| "eval_loss": 0.4097221791744232, | |
| "eval_runtime": 93.7406, | |
| "eval_samples_per_second": 17.57, | |
| "eval_steps_per_second": 2.198, | |
| "eval_wer": 0.4068021652538045, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 9.17, | |
| "learning_rate": 0.00016784140969162994, | |
| "loss": 0.2249, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 9.17, | |
| "eval_loss": 0.4049021899700165, | |
| "eval_runtime": 92.8281, | |
| "eval_samples_per_second": 17.742, | |
| "eval_steps_per_second": 2.219, | |
| "eval_wer": 0.38811153099785517, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 10.09, | |
| "eval_loss": 0.39933764934539795, | |
| "eval_runtime": 94.1592, | |
| "eval_samples_per_second": 17.492, | |
| "eval_steps_per_second": 2.188, | |
| "eval_wer": 0.3808599734450005, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 11.01, | |
| "eval_loss": 0.38551005721092224, | |
| "eval_runtime": 93.9927, | |
| "eval_samples_per_second": 17.523, | |
| "eval_steps_per_second": 2.192, | |
| "eval_wer": 0.37820447349606784, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 11.93, | |
| "eval_loss": 0.3923419117927551, | |
| "eval_runtime": 94.4081, | |
| "eval_samples_per_second": 17.446, | |
| "eval_steps_per_second": 2.182, | |
| "eval_wer": 0.3712593197834746, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 12.84, | |
| "eval_loss": 0.3832719624042511, | |
| "eval_runtime": 93.677, | |
| "eval_samples_per_second": 17.582, | |
| "eval_steps_per_second": 2.199, | |
| "eval_wer": 0.35910530078643654, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 13.76, | |
| "learning_rate": 3.568281938325991e-05, | |
| "loss": 0.1029, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 13.76, | |
| "eval_loss": 0.3811333179473877, | |
| "eval_runtime": 93.1148, | |
| "eval_samples_per_second": 17.688, | |
| "eval_steps_per_second": 2.212, | |
| "eval_wer": 0.3569604739046063, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 14.68, | |
| "eval_loss": 0.38338083028793335, | |
| "eval_runtime": 94.0888, | |
| "eval_samples_per_second": 17.505, | |
| "eval_steps_per_second": 2.189, | |
| "eval_wer": 0.3499131855785926, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "step": 1635, | |
| "total_flos": 6.432633269893007e+18, | |
| "train_loss": 1.057601281845606, | |
| "train_runtime": 4550.4131, | |
| "train_samples_per_second": 11.465, | |
| "train_steps_per_second": 0.359 | |
| } | |
| ], | |
| "max_steps": 1635, | |
| "num_train_epochs": 15, | |
| "total_flos": 6.432633269893007e+18, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |