| { | |
| "best_metric": 32.68135868933731, | |
| "best_model_checkpoint": "./checkpoint-800", | |
| "epoch": 19.23076923076923, | |
| "global_step": 1000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.48, | |
| "learning_rate": 2.3000000000000004e-06, | |
| "loss": 1.1175, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "learning_rate": 4.800000000000001e-06, | |
| "loss": 0.7683, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 1.44, | |
| "learning_rate": 7.3e-06, | |
| "loss": 0.5699, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 1.92, | |
| "learning_rate": 9.800000000000001e-06, | |
| "loss": 0.5088, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "learning_rate": 9.744444444444445e-06, | |
| "loss": 0.3963, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 2.88, | |
| "learning_rate": 9.466666666666667e-06, | |
| "loss": 0.3721, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 3.37, | |
| "learning_rate": 9.188888888888889e-06, | |
| "loss": 0.2968, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 3.85, | |
| "learning_rate": 8.91111111111111e-06, | |
| "loss": 0.2634, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 3.85, | |
| "eval_loss": 0.5562102198600769, | |
| "eval_runtime": 1383.7564, | |
| "eval_samples_per_second": 2.386, | |
| "eval_steps_per_second": 0.075, | |
| "eval_wer": 43.35179518958906, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 4.33, | |
| "learning_rate": 8.633333333333334e-06, | |
| "loss": 0.2223, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 4.81, | |
| "learning_rate": 8.355555555555556e-06, | |
| "loss": 0.1871, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 5.29, | |
| "learning_rate": 8.077777777777778e-06, | |
| "loss": 0.1595, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 5.77, | |
| "learning_rate": 7.800000000000002e-06, | |
| "loss": 0.1393, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 6.25, | |
| "learning_rate": 7.5222222222222226e-06, | |
| "loss": 0.1161, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 6.73, | |
| "learning_rate": 7.244444444444445e-06, | |
| "loss": 0.0923, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 7.21, | |
| "learning_rate": 6.966666666666667e-06, | |
| "loss": 0.0851, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 7.69, | |
| "learning_rate": 6.688888888888889e-06, | |
| "loss": 0.0592, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 7.69, | |
| "eval_loss": 0.6270533800125122, | |
| "eval_runtime": 1352.8123, | |
| "eval_samples_per_second": 2.441, | |
| "eval_steps_per_second": 0.077, | |
| "eval_wer": 40.880746736899184, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 8.17, | |
| "learning_rate": 6.411111111111111e-06, | |
| "loss": 0.0564, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 8.65, | |
| "learning_rate": 6.133333333333334e-06, | |
| "loss": 0.0423, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 9.13, | |
| "learning_rate": 5.855555555555556e-06, | |
| "loss": 0.0354, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 9.62, | |
| "learning_rate": 5.577777777777778e-06, | |
| "loss": 0.0261, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 10.1, | |
| "learning_rate": 5.300000000000001e-06, | |
| "loss": 0.026, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 10.58, | |
| "learning_rate": 5.022222222222223e-06, | |
| "loss": 0.0177, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 11.06, | |
| "learning_rate": 4.744444444444445e-06, | |
| "loss": 0.0168, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 11.54, | |
| "learning_rate": 4.4666666666666665e-06, | |
| "loss": 0.0121, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 11.54, | |
| "eval_loss": 0.7297975420951843, | |
| "eval_runtime": 1209.6571, | |
| "eval_samples_per_second": 2.73, | |
| "eval_steps_per_second": 0.086, | |
| "eval_wer": 35.45063712769665, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 12.02, | |
| "learning_rate": 4.188888888888889e-06, | |
| "loss": 0.0116, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 12.5, | |
| "learning_rate": 3.911111111111112e-06, | |
| "loss": 0.0087, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 12.98, | |
| "learning_rate": 3.633333333333334e-06, | |
| "loss": 0.0085, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 13.46, | |
| "learning_rate": 3.3555555555555557e-06, | |
| "loss": 0.0064, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 13.94, | |
| "learning_rate": 3.077777777777778e-06, | |
| "loss": 0.0069, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 14.42, | |
| "learning_rate": 2.8000000000000003e-06, | |
| "loss": 0.0055, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 14.9, | |
| "learning_rate": 2.5222222222222226e-06, | |
| "loss": 0.0055, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 15.38, | |
| "learning_rate": 2.2444444444444445e-06, | |
| "loss": 0.0048, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 15.38, | |
| "eval_loss": 0.780348539352417, | |
| "eval_runtime": 995.2632, | |
| "eval_samples_per_second": 3.318, | |
| "eval_steps_per_second": 0.104, | |
| "eval_wer": 32.68135868933731, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 15.87, | |
| "learning_rate": 1.9666666666666668e-06, | |
| "loss": 0.0048, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 16.35, | |
| "learning_rate": 1.688888888888889e-06, | |
| "loss": 0.0043, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 16.83, | |
| "learning_rate": 1.4111111111111111e-06, | |
| "loss": 0.0042, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 17.31, | |
| "learning_rate": 1.1333333333333334e-06, | |
| "loss": 0.0041, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 17.79, | |
| "learning_rate": 8.555555555555556e-07, | |
| "loss": 0.0041, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 18.27, | |
| "learning_rate": 5.777777777777778e-07, | |
| "loss": 0.0039, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 18.75, | |
| "learning_rate": 3.0000000000000004e-07, | |
| "loss": 0.0039, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 19.23, | |
| "learning_rate": 2.2222222222222224e-08, | |
| "loss": 0.0039, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 19.23, | |
| "eval_loss": 0.7939738631248474, | |
| "eval_runtime": 1004.5956, | |
| "eval_samples_per_second": 3.287, | |
| "eval_steps_per_second": 0.104, | |
| "eval_wer": 33.32429606104032, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 19.23, | |
| "step": 1000, | |
| "total_flos": 3.6664775110656e+19, | |
| "train_loss": 0.14194686130434275, | |
| "train_runtime": 13358.5887, | |
| "train_samples_per_second": 9.582, | |
| "train_steps_per_second": 0.075 | |
| } | |
| ], | |
| "max_steps": 1000, | |
| "num_train_epochs": 20, | |
| "total_flos": 3.6664775110656e+19, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |