| { | |
| "best_metric": 0.2917510463075469, | |
| "best_model_checkpoint": "./whisper-base-ckb/checkpoint-1000", | |
| "epoch": 10.869565217391305, | |
| "eval_steps": 100, | |
| "global_step": 1000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.11, | |
| "learning_rate": 2.0735116692035353e-06, | |
| "loss": 3.7642, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "learning_rate": 4.841064020920768e-06, | |
| "loss": 2.8375, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "learning_rate": 5.9179056274086315e-06, | |
| "loss": 1.8942, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "learning_rate": 6.599279204266558e-06, | |
| "loss": 1.2854, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "learning_rate": 7.098857723096461e-06, | |
| "loss": 0.9078, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "learning_rate": 7.493495838168052e-06, | |
| "loss": 0.6876, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "learning_rate": 7.819717921297098e-06, | |
| "loss": 0.5329, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "learning_rate": 8.097777358972581e-06, | |
| "loss": 0.4405, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "learning_rate": 8.340082901971508e-06, | |
| "loss": 0.3864, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 1.09, | |
| "learning_rate": 8.55479047323724e-06, | |
| "loss": 0.3434, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.09, | |
| "eval_loss": 0.384033203125, | |
| "eval_runtime": 114.2378, | |
| "eval_samples_per_second": 43.243, | |
| "eval_steps_per_second": 0.061, | |
| "eval_wer": 0.6053732955312542, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "learning_rate": 8.747548830298285e-06, | |
| "loss": 0.3216, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.3, | |
| "learning_rate": 8.92243233400783e-06, | |
| "loss": 0.3038, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.41, | |
| "learning_rate": 9.082476609942404e-06, | |
| "loss": 0.2841, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 1.52, | |
| "learning_rate": 9.230004150181585e-06, | |
| "loss": 0.2701, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 1.63, | |
| "learning_rate": 9.36683155598379e-06, | |
| "loss": 0.255, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.74, | |
| "learning_rate": 9.494406569816089e-06, | |
| "loss": 0.2549, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 1.85, | |
| "learning_rate": 9.613901637842896e-06, | |
| "loss": 0.2433, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 1.96, | |
| "learning_rate": 9.726279567611587e-06, | |
| "loss": 0.2373, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 2.07, | |
| "learning_rate": 9.832340708264348e-06, | |
| "loss": 0.2232, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 2.17, | |
| "learning_rate": 9.93275756211191e-06, | |
| "loss": 0.2089, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 2.17, | |
| "eval_loss": 0.265380859375, | |
| "eval_runtime": 112.176, | |
| "eval_samples_per_second": 44.038, | |
| "eval_steps_per_second": 0.062, | |
| "eval_wer": 0.4739773187525314, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 2.28, | |
| "learning_rate": 9.975000000000002e-06, | |
| "loss": 0.2122, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 2.39, | |
| "learning_rate": 9.85e-06, | |
| "loss": 0.2123, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 2.5, | |
| "learning_rate": 9.725000000000001e-06, | |
| "loss": 0.2044, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 2.61, | |
| "learning_rate": 9.600000000000001e-06, | |
| "loss": 0.1994, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 2.72, | |
| "learning_rate": 9.475000000000002e-06, | |
| "loss": 0.1911, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 2.83, | |
| "learning_rate": 9.350000000000002e-06, | |
| "loss": 0.1927, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 2.93, | |
| "learning_rate": 9.225e-06, | |
| "loss": 0.1913, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 3.04, | |
| "learning_rate": 9.100000000000001e-06, | |
| "loss": 0.1812, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 3.15, | |
| "learning_rate": 8.975e-06, | |
| "loss": 0.1715, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 3.26, | |
| "learning_rate": 8.85e-06, | |
| "loss": 0.167, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 3.26, | |
| "eval_loss": 0.224609375, | |
| "eval_runtime": 112.6918, | |
| "eval_samples_per_second": 43.836, | |
| "eval_steps_per_second": 0.062, | |
| "eval_wer": 0.41899554475496154, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 3.37, | |
| "learning_rate": 8.725000000000002e-06, | |
| "loss": 0.1683, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 3.48, | |
| "learning_rate": 8.6e-06, | |
| "loss": 0.1728, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 3.59, | |
| "learning_rate": 8.475000000000001e-06, | |
| "loss": 0.1692, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 3.7, | |
| "learning_rate": 8.35e-06, | |
| "loss": 0.1654, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 3.8, | |
| "learning_rate": 8.225e-06, | |
| "loss": 0.1622, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 3.91, | |
| "learning_rate": 8.1e-06, | |
| "loss": 0.1597, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 4.02, | |
| "learning_rate": 7.975e-06, | |
| "loss": 0.1626, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 4.13, | |
| "learning_rate": 7.850000000000001e-06, | |
| "loss": 0.151, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 4.24, | |
| "learning_rate": 7.725e-06, | |
| "loss": 0.1481, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 4.35, | |
| "learning_rate": 7.600000000000001e-06, | |
| "loss": 0.1452, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 4.35, | |
| "eval_loss": 0.1964111328125, | |
| "eval_runtime": 112.4796, | |
| "eval_samples_per_second": 43.919, | |
| "eval_steps_per_second": 0.062, | |
| "eval_wer": 0.38034966923180774, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 4.46, | |
| "learning_rate": 7.475000000000001e-06, | |
| "loss": 0.1437, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 4.57, | |
| "learning_rate": 7.350000000000001e-06, | |
| "loss": 0.1477, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 4.67, | |
| "learning_rate": 7.225000000000001e-06, | |
| "loss": 0.146, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 4.78, | |
| "learning_rate": 7.100000000000001e-06, | |
| "loss": 0.1432, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 4.89, | |
| "learning_rate": 6.975000000000001e-06, | |
| "loss": 0.1383, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "learning_rate": 6.850000000000001e-06, | |
| "loss": 0.1389, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 5.11, | |
| "learning_rate": 6.725000000000001e-06, | |
| "loss": 0.1416, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 5.22, | |
| "learning_rate": 6.600000000000001e-06, | |
| "loss": 0.1356, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 5.33, | |
| "learning_rate": 6.475e-06, | |
| "loss": 0.1302, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 5.43, | |
| "learning_rate": 6.35e-06, | |
| "loss": 0.1287, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 5.43, | |
| "eval_loss": 0.1788330078125, | |
| "eval_runtime": 133.6018, | |
| "eval_samples_per_second": 36.976, | |
| "eval_steps_per_second": 0.052, | |
| "eval_wer": 0.35419198055893075, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 5.54, | |
| "learning_rate": 6.225000000000001e-06, | |
| "loss": 0.1217, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 5.65, | |
| "learning_rate": 6.1e-06, | |
| "loss": 0.1332, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 5.76, | |
| "learning_rate": 5.975e-06, | |
| "loss": 0.1307, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 5.87, | |
| "learning_rate": 5.85e-06, | |
| "loss": 0.1281, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 5.98, | |
| "learning_rate": 5.725e-06, | |
| "loss": 0.1257, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 6.09, | |
| "learning_rate": 5.600000000000001e-06, | |
| "loss": 0.1176, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 6.2, | |
| "learning_rate": 5.475e-06, | |
| "loss": 0.1249, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 6.3, | |
| "learning_rate": 5.3500000000000004e-06, | |
| "loss": 0.1214, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 6.41, | |
| "learning_rate": 5.225e-06, | |
| "loss": 0.1166, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 6.52, | |
| "learning_rate": 5.1e-06, | |
| "loss": 0.1163, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 6.52, | |
| "eval_loss": 0.1650390625, | |
| "eval_runtime": 120.7458, | |
| "eval_samples_per_second": 40.912, | |
| "eval_steps_per_second": 0.058, | |
| "eval_wer": 0.33255704063723507, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 6.63, | |
| "learning_rate": 4.975000000000001e-06, | |
| "loss": 0.1139, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 6.74, | |
| "learning_rate": 4.85e-06, | |
| "loss": 0.1216, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 6.85, | |
| "learning_rate": 4.7250000000000005e-06, | |
| "loss": 0.1189, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 6.96, | |
| "learning_rate": 4.600000000000001e-06, | |
| "loss": 0.1167, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 7.07, | |
| "learning_rate": 4.475e-06, | |
| "loss": 0.1132, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 7.17, | |
| "learning_rate": 4.350000000000001e-06, | |
| "loss": 0.1046, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 7.28, | |
| "learning_rate": 4.225e-06, | |
| "loss": 0.111, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 7.39, | |
| "learning_rate": 4.1e-06, | |
| "loss": 0.115, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 7.5, | |
| "learning_rate": 3.975000000000001e-06, | |
| "loss": 0.1097, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 7.61, | |
| "learning_rate": 3.85e-06, | |
| "loss": 0.1068, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 7.61, | |
| "eval_loss": 0.156005859375, | |
| "eval_runtime": 128.5945, | |
| "eval_samples_per_second": 38.415, | |
| "eval_steps_per_second": 0.054, | |
| "eval_wer": 0.31554610503577696, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 7.72, | |
| "learning_rate": 3.7250000000000003e-06, | |
| "loss": 0.1061, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 7.83, | |
| "learning_rate": 3.6000000000000003e-06, | |
| "loss": 0.1112, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 7.93, | |
| "learning_rate": 3.475e-06, | |
| "loss": 0.1128, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 8.04, | |
| "learning_rate": 3.3500000000000005e-06, | |
| "loss": 0.107, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 8.15, | |
| "learning_rate": 3.2250000000000005e-06, | |
| "loss": 0.1022, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 8.26, | |
| "learning_rate": 3.1000000000000004e-06, | |
| "loss": 0.0977, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 8.37, | |
| "learning_rate": 2.9750000000000003e-06, | |
| "loss": 0.1016, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 8.48, | |
| "learning_rate": 2.85e-06, | |
| "loss": 0.1089, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 8.59, | |
| "learning_rate": 2.7250000000000006e-06, | |
| "loss": 0.105, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 8.7, | |
| "learning_rate": 2.6e-06, | |
| "loss": 0.1015, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 8.7, | |
| "eval_loss": 0.14892578125, | |
| "eval_runtime": 117.3243, | |
| "eval_samples_per_second": 42.106, | |
| "eval_steps_per_second": 0.06, | |
| "eval_wer": 0.30592682597542864, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 8.8, | |
| "learning_rate": 2.475e-06, | |
| "loss": 0.0977, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 8.91, | |
| "learning_rate": 2.35e-06, | |
| "loss": 0.1023, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 9.02, | |
| "learning_rate": 2.2250000000000003e-06, | |
| "loss": 0.1064, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 9.13, | |
| "learning_rate": 2.1000000000000002e-06, | |
| "loss": 0.0964, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 9.24, | |
| "learning_rate": 1.975e-06, | |
| "loss": 0.0973, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 9.35, | |
| "learning_rate": 1.85e-06, | |
| "loss": 0.0946, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 9.46, | |
| "learning_rate": 1.725e-06, | |
| "loss": 0.0965, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 9.57, | |
| "learning_rate": 1.6000000000000001e-06, | |
| "loss": 0.1048, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 9.67, | |
| "learning_rate": 1.475e-06, | |
| "loss": 0.0982, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 9.78, | |
| "learning_rate": 1.3500000000000002e-06, | |
| "loss": 0.0968, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 9.78, | |
| "eval_loss": 0.14404296875, | |
| "eval_runtime": 113.4503, | |
| "eval_samples_per_second": 43.543, | |
| "eval_steps_per_second": 0.062, | |
| "eval_wer": 0.2953962467935736, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 9.89, | |
| "learning_rate": 1.2250000000000001e-06, | |
| "loss": 0.0951, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "learning_rate": 1.1e-06, | |
| "loss": 0.0937, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 10.11, | |
| "learning_rate": 9.750000000000002e-07, | |
| "loss": 0.1012, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 10.22, | |
| "learning_rate": 8.500000000000001e-07, | |
| "loss": 0.0963, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 10.33, | |
| "learning_rate": 7.25e-07, | |
| "loss": 0.0941, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 10.43, | |
| "learning_rate": 6.000000000000001e-07, | |
| "loss": 0.0922, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 10.54, | |
| "learning_rate": 4.7500000000000006e-07, | |
| "loss": 0.0897, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 10.65, | |
| "learning_rate": 3.5000000000000004e-07, | |
| "loss": 0.1039, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 10.76, | |
| "learning_rate": 2.2500000000000002e-07, | |
| "loss": 0.0978, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 10.87, | |
| "learning_rate": 1.0000000000000001e-07, | |
| "loss": 0.0939, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 10.87, | |
| "eval_loss": 0.1419677734375, | |
| "eval_runtime": 116.6648, | |
| "eval_samples_per_second": 42.344, | |
| "eval_steps_per_second": 0.06, | |
| "eval_wer": 0.2917510463075469, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 10.87, | |
| "step": 1000, | |
| "total_flos": 7.47187145428435e+19, | |
| "train_loss": 0.25973586964607237, | |
| "train_runtime": 25779.7961, | |
| "train_samples_per_second": 44.686, | |
| "train_steps_per_second": 0.039 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 1000, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 11, | |
| "save_steps": 100, | |
| "total_flos": 7.47187145428435e+19, | |
| "train_batch_size": 192, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |