| { | |
| "best_metric": 54.06850588964662, | |
| "best_model_checkpoint": "./checkpoint-400", | |
| "epoch": 116.74074074074075, | |
| "global_step": 701, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.59, | |
| "learning_rate": 4.2e-08, | |
| "loss": 2.5631, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 3.3, | |
| "learning_rate": 1.02e-07, | |
| "loss": 2.7338, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 4.89, | |
| "learning_rate": 1.56e-07, | |
| "loss": 2.5033, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 6.59, | |
| "learning_rate": 2.1599999999999998e-07, | |
| "loss": 2.5045, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 8.3, | |
| "learning_rate": 2.76e-07, | |
| "loss": 2.1675, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 9.89, | |
| "learning_rate": 2.9810526315789476e-07, | |
| "loss": 1.7642, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 11.59, | |
| "learning_rate": 2.949473684210526e-07, | |
| "loss": 1.6382, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 13.3, | |
| "learning_rate": 2.917894736842105e-07, | |
| "loss": 1.469, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 14.89, | |
| "learning_rate": 2.886315789473684e-07, | |
| "loss": 1.2432, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 16.59, | |
| "learning_rate": 2.8547368421052633e-07, | |
| "loss": 1.2281, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 16.59, | |
| "eval_loss": 1.0950794219970703, | |
| "eval_runtime": 2208.5997, | |
| "eval_samples_per_second": 0.232, | |
| "eval_steps_per_second": 0.058, | |
| "eval_wer": 69.31184128952262, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 18.3, | |
| "learning_rate": 2.823157894736842e-07, | |
| "loss": 1.1316, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 19.89, | |
| "learning_rate": 2.7915789473684206e-07, | |
| "loss": 0.9889, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 21.59, | |
| "learning_rate": 2.76e-07, | |
| "loss": 0.9964, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 23.3, | |
| "learning_rate": 2.728421052631579e-07, | |
| "loss": 0.9559, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 24.89, | |
| "learning_rate": 2.6968421052631577e-07, | |
| "loss": 0.8493, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 26.59, | |
| "learning_rate": 2.665263157894737e-07, | |
| "loss": 0.8809, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 28.3, | |
| "learning_rate": 2.6336842105263155e-07, | |
| "loss": 0.8447, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 29.89, | |
| "learning_rate": 2.602105263157895e-07, | |
| "loss": 0.7582, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 31.59, | |
| "learning_rate": 2.5705263157894734e-07, | |
| "loss": 0.789, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 33.3, | |
| "learning_rate": 2.5389473684210526e-07, | |
| "loss": 0.7529, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 33.3, | |
| "eval_loss": 0.8693189024925232, | |
| "eval_runtime": 2268.4256, | |
| "eval_samples_per_second": 0.226, | |
| "eval_steps_per_second": 0.056, | |
| "eval_wer": 57.563546187228766, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 34.89, | |
| "learning_rate": 2.5073684210526313e-07, | |
| "loss": 0.6902, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 36.59, | |
| "learning_rate": 2.4757894736842105e-07, | |
| "loss": 0.7184, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 38.3, | |
| "learning_rate": 2.4442105263157897e-07, | |
| "loss": 0.6878, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 39.89, | |
| "learning_rate": 2.4126315789473683e-07, | |
| "loss": 0.6384, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 41.59, | |
| "learning_rate": 2.3810526315789472e-07, | |
| "loss": 0.6526, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 43.3, | |
| "learning_rate": 2.3494736842105262e-07, | |
| "loss": 0.6423, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 44.89, | |
| "learning_rate": 2.317894736842105e-07, | |
| "loss": 0.5752, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 46.59, | |
| "learning_rate": 2.286315789473684e-07, | |
| "loss": 0.6023, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 48.3, | |
| "learning_rate": 2.254736842105263e-07, | |
| "loss": 0.5882, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 49.89, | |
| "learning_rate": 2.2231578947368422e-07, | |
| "loss": 0.5372, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 49.89, | |
| "eval_loss": 0.8399219512939453, | |
| "eval_runtime": 2313.3276, | |
| "eval_samples_per_second": 0.221, | |
| "eval_steps_per_second": 0.055, | |
| "eval_wer": 54.73496590204587, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 51.59, | |
| "learning_rate": 2.191578947368421e-07, | |
| "loss": 0.5521, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 53.3, | |
| "learning_rate": 2.1599999999999998e-07, | |
| "loss": 0.5448, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 54.89, | |
| "learning_rate": 2.128421052631579e-07, | |
| "loss": 0.4926, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 56.59, | |
| "learning_rate": 2.096842105263158e-07, | |
| "loss": 0.5193, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 58.3, | |
| "learning_rate": 2.0652631578947368e-07, | |
| "loss": 0.4955, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 59.89, | |
| "learning_rate": 2.0336842105263155e-07, | |
| "loss": 0.4537, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 61.59, | |
| "learning_rate": 2.0021052631578947e-07, | |
| "loss": 0.4724, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 63.3, | |
| "learning_rate": 1.9705263157894736e-07, | |
| "loss": 0.461, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 64.89, | |
| "learning_rate": 1.9389473684210525e-07, | |
| "loss": 0.4134, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 66.59, | |
| "learning_rate": 1.9073684210526317e-07, | |
| "loss": 0.4398, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 66.59, | |
| "eval_loss": 0.8623208999633789, | |
| "eval_runtime": 2300.4889, | |
| "eval_samples_per_second": 0.223, | |
| "eval_steps_per_second": 0.056, | |
| "eval_wer": 54.06850588964662, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 68.3, | |
| "learning_rate": 1.8757894736842104e-07, | |
| "loss": 0.4255, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 69.89, | |
| "learning_rate": 1.8442105263157893e-07, | |
| "loss": 0.3835, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 71.59, | |
| "learning_rate": 1.8126315789473682e-07, | |
| "loss": 0.4041, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 73.3, | |
| "learning_rate": 1.7810526315789474e-07, | |
| "loss": 0.3895, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 74.89, | |
| "learning_rate": 1.749473684210526e-07, | |
| "loss": 0.35, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 76.59, | |
| "learning_rate": 1.717894736842105e-07, | |
| "loss": 0.3722, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 78.3, | |
| "learning_rate": 1.6863157894736842e-07, | |
| "loss": 0.3569, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 79.89, | |
| "learning_rate": 1.6547368421052632e-07, | |
| "loss": 0.3314, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 81.59, | |
| "learning_rate": 1.6231578947368418e-07, | |
| "loss": 0.341, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 83.3, | |
| "learning_rate": 1.591578947368421e-07, | |
| "loss": 0.3244, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 83.3, | |
| "eval_loss": 0.9098332524299622, | |
| "eval_runtime": 2322.6222, | |
| "eval_samples_per_second": 0.22, | |
| "eval_steps_per_second": 0.055, | |
| "eval_wer": 54.750464972101675, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 84.89, | |
| "learning_rate": 1.56e-07, | |
| "loss": 0.3066, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 86.59, | |
| "learning_rate": 1.528421052631579e-07, | |
| "loss": 0.3158, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 88.3, | |
| "learning_rate": 1.4968421052631578e-07, | |
| "loss": 0.304, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 89.89, | |
| "learning_rate": 1.4652631578947367e-07, | |
| "loss": 0.2782, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 91.59, | |
| "learning_rate": 1.4336842105263157e-07, | |
| "loss": 0.293, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 93.3, | |
| "learning_rate": 1.4021052631578949e-07, | |
| "loss": 0.2797, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 94.89, | |
| "learning_rate": 1.3705263157894735e-07, | |
| "loss": 0.2595, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 96.59, | |
| "learning_rate": 1.3389473684210527e-07, | |
| "loss": 0.2648, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 98.3, | |
| "learning_rate": 1.3073684210526314e-07, | |
| "loss": 0.2648, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 99.89, | |
| "learning_rate": 1.2757894736842106e-07, | |
| "loss": 0.238, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 99.89, | |
| "eval_loss": 0.9607409238815308, | |
| "eval_runtime": 2334.4802, | |
| "eval_samples_per_second": 0.219, | |
| "eval_steps_per_second": 0.055, | |
| "eval_wer": 55.37817730936144, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 101.59, | |
| "learning_rate": 1.2442105263157895e-07, | |
| "loss": 0.2493, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 103.3, | |
| "learning_rate": 1.2126315789473684e-07, | |
| "loss": 0.2368, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 104.89, | |
| "learning_rate": 1.1810526315789474e-07, | |
| "loss": 0.2222, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 106.59, | |
| "learning_rate": 1.1494736842105262e-07, | |
| "loss": 0.2325, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 108.3, | |
| "learning_rate": 1.1178947368421052e-07, | |
| "loss": 0.2216, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 109.89, | |
| "learning_rate": 1.0863157894736843e-07, | |
| "loss": 0.2045, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 111.59, | |
| "learning_rate": 1.0547368421052631e-07, | |
| "loss": 0.2168, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 113.3, | |
| "learning_rate": 1.0231578947368421e-07, | |
| "loss": 0.2104, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 114.89, | |
| "learning_rate": 9.915789473684209e-08, | |
| "loss": 0.1888, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 116.59, | |
| "learning_rate": 9.6e-08, | |
| "loss": 0.2014, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 116.59, | |
| "eval_loss": 1.007683515548706, | |
| "eval_runtime": 2341.2281, | |
| "eval_samples_per_second": 0.219, | |
| "eval_steps_per_second": 0.055, | |
| "eval_wer": 55.92064476131432, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 116.74, | |
| "step": 701, | |
| "total_flos": 1.068294730825728e+20, | |
| "train_loss": 0.0002270207000356938, | |
| "train_runtime": 19.3242, | |
| "train_samples_per_second": 2318.342, | |
| "train_steps_per_second": 36.224 | |
| } | |
| ], | |
| "max_steps": 700, | |
| "num_train_epochs": 117, | |
| "total_flos": 1.068294730825728e+20, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |