| { | |
| "best_metric": 56.651029055690074, | |
| "best_model_checkpoint": "./checkpoint-600", | |
| "epoch": 100.0, | |
| "global_step": 700, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 3.7, | |
| "learning_rate": 1.9200000000000003e-06, | |
| "loss": 2.1183, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 3.7, | |
| "eval_loss": 1.3170489072799683, | |
| "eval_runtime": 435.5063, | |
| "eval_samples_per_second": 1.176, | |
| "eval_steps_per_second": 0.147, | |
| "eval_wer": 76.95217917675545, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 7.41, | |
| "learning_rate": 3.920000000000001e-06, | |
| "loss": 0.8565, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 7.41, | |
| "eval_loss": 0.9366902709007263, | |
| "eval_runtime": 473.9252, | |
| "eval_samples_per_second": 1.08, | |
| "eval_steps_per_second": 0.135, | |
| "eval_wer": 61.99303874092009, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 7.78, | |
| "learning_rate": 4.12e-06, | |
| "loss": 0.5824, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 8.15, | |
| "learning_rate": 4.32e-06, | |
| "loss": 0.5195, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 8.52, | |
| "learning_rate": 4.520000000000001e-06, | |
| "loss": 0.4472, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 8.89, | |
| "learning_rate": 4.7200000000000005e-06, | |
| "loss": 0.4615, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 9.26, | |
| "learning_rate": 4.92e-06, | |
| "loss": 0.4068, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 9.63, | |
| "learning_rate": 5.12e-06, | |
| "loss": 0.3408, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "learning_rate": 5.320000000000001e-06, | |
| "loss": 0.3718, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 10.37, | |
| "learning_rate": 5.5200000000000005e-06, | |
| "loss": 0.2749, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 10.74, | |
| "learning_rate": 5.72e-06, | |
| "loss": 0.2625, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 11.11, | |
| "learning_rate": 5.92e-06, | |
| "loss": 0.2246, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 11.11, | |
| "eval_loss": 0.9642460942268372, | |
| "eval_runtime": 439.7231, | |
| "eval_samples_per_second": 1.164, | |
| "eval_steps_per_second": 0.146, | |
| "eval_wer": 58.830205811138015, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 11.48, | |
| "learning_rate": 6.120000000000001e-06, | |
| "loss": 0.1991, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 11.85, | |
| "learning_rate": 6.3200000000000005e-06, | |
| "loss": 0.1798, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 12.22, | |
| "learning_rate": 6.520000000000001e-06, | |
| "loss": 0.1426, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 12.59, | |
| "learning_rate": 6.720000000000001e-06, | |
| "loss": 0.125, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 12.96, | |
| "learning_rate": 6.92e-06, | |
| "loss": 0.1177, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 13.33, | |
| "learning_rate": 7.1200000000000004e-06, | |
| "loss": 0.0874, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 13.7, | |
| "learning_rate": 7.32e-06, | |
| "loss": 0.0852, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 14.07, | |
| "learning_rate": 7.520000000000001e-06, | |
| "loss": 0.0666, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 14.44, | |
| "learning_rate": 7.72e-06, | |
| "loss": 0.0531, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 14.81, | |
| "learning_rate": 7.92e-06, | |
| "loss": 0.054, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 14.81, | |
| "eval_loss": 1.0876343250274658, | |
| "eval_runtime": 428.9567, | |
| "eval_samples_per_second": 1.194, | |
| "eval_steps_per_second": 0.149, | |
| "eval_wer": 57.99031476997578, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 15.19, | |
| "learning_rate": 8.120000000000002e-06, | |
| "loss": 0.0465, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 15.56, | |
| "learning_rate": 8.32e-06, | |
| "loss": 0.0392, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 15.93, | |
| "learning_rate": 8.52e-06, | |
| "loss": 0.0354, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 16.3, | |
| "learning_rate": 8.720000000000001e-06, | |
| "loss": 0.0296, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 16.67, | |
| "learning_rate": 8.920000000000001e-06, | |
| "loss": 0.0255, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 17.04, | |
| "learning_rate": 9.12e-06, | |
| "loss": 0.0274, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 17.41, | |
| "learning_rate": 9.32e-06, | |
| "loss": 0.0216, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 17.78, | |
| "learning_rate": 9.52e-06, | |
| "loss": 0.022, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 18.15, | |
| "learning_rate": 9.72e-06, | |
| "loss": 0.0219, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 18.52, | |
| "learning_rate": 9.920000000000002e-06, | |
| "loss": 0.0159, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 18.52, | |
| "eval_loss": 1.1797882318496704, | |
| "eval_runtime": 450.9533, | |
| "eval_samples_per_second": 1.135, | |
| "eval_steps_per_second": 0.142, | |
| "eval_wer": 57.87681598062954, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 18.89, | |
| "learning_rate": 9.4e-06, | |
| "loss": 0.019, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 19.26, | |
| "learning_rate": 8.400000000000001e-06, | |
| "loss": 0.021, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 19.63, | |
| "learning_rate": 7.4e-06, | |
| "loss": 0.0139, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "learning_rate": 6.4000000000000006e-06, | |
| "loss": 0.013, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 20.37, | |
| "learning_rate": 5.400000000000001e-06, | |
| "loss": 0.007, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 20.74, | |
| "learning_rate": 4.4e-06, | |
| "loss": 0.0082, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 21.11, | |
| "learning_rate": 3.4000000000000005e-06, | |
| "loss": 0.009, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 21.48, | |
| "learning_rate": 2.4000000000000003e-06, | |
| "loss": 0.0045, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 21.85, | |
| "learning_rate": 1.4000000000000001e-06, | |
| "loss": 0.0049, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 22.22, | |
| "learning_rate": 4.0000000000000003e-07, | |
| "loss": 0.0045, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 22.22, | |
| "eval_loss": 1.2309296131134033, | |
| "eval_runtime": 449.2432, | |
| "eval_samples_per_second": 1.14, | |
| "eval_steps_per_second": 0.142, | |
| "eval_wer": 56.651029055690074, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 87.14, | |
| "learning_rate": 4.7e-06, | |
| "loss": 0.0038, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 88.57, | |
| "learning_rate": 4.2000000000000004e-06, | |
| "loss": 0.0033, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 90.0, | |
| "learning_rate": 3.7e-06, | |
| "loss": 0.0033, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 91.43, | |
| "learning_rate": 3.2000000000000003e-06, | |
| "loss": 0.0031, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 92.86, | |
| "learning_rate": 2.7000000000000004e-06, | |
| "loss": 0.0029, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 94.29, | |
| "learning_rate": 2.2e-06, | |
| "loss": 0.0028, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 95.71, | |
| "learning_rate": 1.7000000000000002e-06, | |
| "loss": 0.0027, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 97.14, | |
| "learning_rate": 1.2000000000000002e-06, | |
| "loss": 0.0026, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 98.57, | |
| "learning_rate": 7.000000000000001e-07, | |
| "loss": 0.0026, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 100.0, | |
| "learning_rate": 2.0000000000000002e-07, | |
| "loss": 0.0026, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 100.0, | |
| "eval_loss": 1.2580721378326416, | |
| "eval_runtime": 471.9052, | |
| "eval_samples_per_second": 1.085, | |
| "eval_steps_per_second": 0.068, | |
| "eval_wer": 56.847760290556906, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 100.0, | |
| "step": 700, | |
| "total_flos": 4.53742824923136e+18, | |
| "train_loss": 0.0004244298700775419, | |
| "train_runtime": 836.7871, | |
| "train_samples_per_second": 53.538, | |
| "train_steps_per_second": 0.837 | |
| } | |
| ], | |
| "max_steps": 700, | |
| "num_train_epochs": 100, | |
| "total_flos": 4.53742824923136e+18, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |