| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 50.0, | |
| "global_step": 9650, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.52, | |
| "learning_rate": 3.7125e-06, | |
| "loss": 19.477, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.04, | |
| "learning_rate": 7.4625e-06, | |
| "loss": 11.5043, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.55, | |
| "learning_rate": 1.1212499999999998e-05, | |
| "loss": 7.4168, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 2.07, | |
| "learning_rate": 1.49625e-05, | |
| "loss": 6.086, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 2.59, | |
| "learning_rate": 1.8712499999999997e-05, | |
| "loss": 5.0417, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 2.59, | |
| "eval_loss": 5.148427963256836, | |
| "eval_runtime": 121.8982, | |
| "eval_samples_per_second": 22.092, | |
| "eval_steps_per_second": 2.765, | |
| "eval_wer": 1.0, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 3.11, | |
| "learning_rate": 2.2462499999999997e-05, | |
| "loss": 4.2126, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 3.63, | |
| "learning_rate": 2.6212499999999997e-05, | |
| "loss": 3.6981, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 4.15, | |
| "learning_rate": 2.99625e-05, | |
| "loss": 3.496, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 4.66, | |
| "learning_rate": 3.37125e-05, | |
| "loss": 3.4391, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 5.18, | |
| "learning_rate": 3.7462499999999996e-05, | |
| "loss": 3.3722, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 5.18, | |
| "eval_loss": 3.3380324840545654, | |
| "eval_runtime": 122.9217, | |
| "eval_samples_per_second": 21.908, | |
| "eval_steps_per_second": 2.742, | |
| "eval_wer": 1.0001231729348004, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 5.7, | |
| "learning_rate": 4.12125e-05, | |
| "loss": 3.2941, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 6.22, | |
| "learning_rate": 4.4962499999999995e-05, | |
| "loss": 3.2158, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 6.74, | |
| "learning_rate": 4.871249999999999e-05, | |
| "loss": 2.7897, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 7.25, | |
| "learning_rate": 5.2462499999999994e-05, | |
| "loss": 2.2331, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 7.77, | |
| "learning_rate": 5.62125e-05, | |
| "loss": 1.9752, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 7.77, | |
| "eval_loss": 1.3909666538238525, | |
| "eval_runtime": 123.3452, | |
| "eval_samples_per_second": 21.833, | |
| "eval_steps_per_second": 2.732, | |
| "eval_wer": 1.007431433732961, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 8.29, | |
| "learning_rate": 5.9962499999999994e-05, | |
| "loss": 1.8341, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 8.81, | |
| "learning_rate": 6.37125e-05, | |
| "loss": 1.7481, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 9.33, | |
| "learning_rate": 6.746249999999999e-05, | |
| "loss": 1.6736, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 9.84, | |
| "learning_rate": 7.121249999999999e-05, | |
| "loss": 1.6481, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 10.36, | |
| "learning_rate": 7.492499999999999e-05, | |
| "loss": 1.5868, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 10.36, | |
| "eval_loss": 1.029773235321045, | |
| "eval_runtime": 125.3959, | |
| "eval_samples_per_second": 21.476, | |
| "eval_steps_per_second": 2.687, | |
| "eval_wer": 1.0084168172113648, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 10.88, | |
| "learning_rate": 7.404901960784314e-05, | |
| "loss": 1.5598, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 11.4, | |
| "learning_rate": 7.306862745098039e-05, | |
| "loss": 1.5205, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 11.92, | |
| "learning_rate": 7.208823529411764e-05, | |
| "loss": 1.4995, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 12.44, | |
| "learning_rate": 7.11078431372549e-05, | |
| "loss": 1.451, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 12.95, | |
| "learning_rate": 7.012745098039215e-05, | |
| "loss": 1.4413, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 12.95, | |
| "eval_loss": 0.9313108921051025, | |
| "eval_runtime": 124.7489, | |
| "eval_samples_per_second": 21.587, | |
| "eval_steps_per_second": 2.701, | |
| "eval_wer": 1.0174905567416652, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 13.47, | |
| "learning_rate": 6.91470588235294e-05, | |
| "loss": 1.3987, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 13.99, | |
| "learning_rate": 6.816666666666665e-05, | |
| "loss": 1.3931, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 14.51, | |
| "learning_rate": 6.718627450980392e-05, | |
| "loss": 1.3632, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 15.03, | |
| "learning_rate": 6.620588235294117e-05, | |
| "loss": 1.3609, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 15.54, | |
| "learning_rate": 6.522549019607843e-05, | |
| "loss": 1.3296, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 15.54, | |
| "eval_loss": 0.8966061472892761, | |
| "eval_runtime": 122.7764, | |
| "eval_samples_per_second": 21.934, | |
| "eval_steps_per_second": 2.745, | |
| "eval_wer": 1.0193792084086057, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 16.06, | |
| "learning_rate": 6.424509803921568e-05, | |
| "loss": 1.3424, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 16.58, | |
| "learning_rate": 6.326470588235293e-05, | |
| "loss": 1.3207, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 17.1, | |
| "learning_rate": 6.229411764705881e-05, | |
| "loss": 1.2947, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 17.62, | |
| "learning_rate": 6.131372549019608e-05, | |
| "loss": 1.2949, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 18.13, | |
| "learning_rate": 6.033333333333333e-05, | |
| "loss": 1.2746, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 18.13, | |
| "eval_loss": 0.8875203728675842, | |
| "eval_runtime": 121.1222, | |
| "eval_samples_per_second": 22.234, | |
| "eval_steps_per_second": 2.782, | |
| "eval_wer": 1.0097306618492363, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 18.65, | |
| "learning_rate": 5.9352941176470584e-05, | |
| "loss": 1.2649, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 19.17, | |
| "learning_rate": 5.837254901960784e-05, | |
| "loss": 1.2446, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 19.69, | |
| "learning_rate": 5.739215686274509e-05, | |
| "loss": 1.2393, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 20.21, | |
| "learning_rate": 5.641176470588235e-05, | |
| "loss": 1.2187, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 20.73, | |
| "learning_rate": 5.54313725490196e-05, | |
| "loss": 1.2147, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 20.73, | |
| "eval_loss": 0.8745562434196472, | |
| "eval_runtime": 124.1654, | |
| "eval_samples_per_second": 21.689, | |
| "eval_steps_per_second": 2.714, | |
| "eval_wer": 1.0089095089505666, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 21.24, | |
| "learning_rate": 5.445098039215686e-05, | |
| "loss": 1.2066, | |
| "step": 4100 | |
| }, | |
| { | |
| "epoch": 21.76, | |
| "learning_rate": 5.347058823529411e-05, | |
| "loss": 1.1995, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 22.28, | |
| "learning_rate": 5.2490196078431365e-05, | |
| "loss": 1.1985, | |
| "step": 4300 | |
| }, | |
| { | |
| "epoch": 22.8, | |
| "learning_rate": 5.150980392156863e-05, | |
| "loss": 1.1856, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 23.32, | |
| "learning_rate": 5.052941176470588e-05, | |
| "loss": 1.1774, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 23.32, | |
| "eval_loss": 0.8383361101150513, | |
| "eval_runtime": 126.9391, | |
| "eval_samples_per_second": 21.215, | |
| "eval_steps_per_second": 2.655, | |
| "eval_wer": 1.019830842502874, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 23.83, | |
| "learning_rate": 4.9549019607843137e-05, | |
| "loss": 1.1655, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 24.35, | |
| "learning_rate": 4.856862745098039e-05, | |
| "loss": 1.1548, | |
| "step": 4700 | |
| }, | |
| { | |
| "epoch": 24.87, | |
| "learning_rate": 4.7588235294117644e-05, | |
| "loss": 1.1539, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 25.39, | |
| "learning_rate": 4.6607843137254895e-05, | |
| "loss": 1.1403, | |
| "step": 4900 | |
| }, | |
| { | |
| "epoch": 25.91, | |
| "learning_rate": 4.562745098039215e-05, | |
| "loss": 1.129, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 25.91, | |
| "eval_loss": 0.7847502827644348, | |
| "eval_runtime": 122.1221, | |
| "eval_samples_per_second": 22.052, | |
| "eval_steps_per_second": 2.76, | |
| "eval_wer": 1.016710461487929, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 26.42, | |
| "learning_rate": 4.464705882352941e-05, | |
| "loss": 1.1236, | |
| "step": 5100 | |
| }, | |
| { | |
| "epoch": 26.94, | |
| "learning_rate": 4.366666666666666e-05, | |
| "loss": 1.1171, | |
| "step": 5200 | |
| }, | |
| { | |
| "epoch": 27.46, | |
| "learning_rate": 4.268627450980392e-05, | |
| "loss": 1.1144, | |
| "step": 5300 | |
| }, | |
| { | |
| "epoch": 27.98, | |
| "learning_rate": 4.170588235294117e-05, | |
| "loss": 1.0954, | |
| "step": 5400 | |
| }, | |
| { | |
| "epoch": 28.5, | |
| "learning_rate": 4.0725490196078425e-05, | |
| "loss": 1.0995, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 28.5, | |
| "eval_loss": 0.7992498874664307, | |
| "eval_runtime": 122.009, | |
| "eval_samples_per_second": 22.072, | |
| "eval_steps_per_second": 2.762, | |
| "eval_wer": 1.0209804565610117, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 29.02, | |
| "learning_rate": 3.974509803921569e-05, | |
| "loss": 1.1048, | |
| "step": 5600 | |
| }, | |
| { | |
| "epoch": 29.53, | |
| "learning_rate": 3.876470588235294e-05, | |
| "loss": 1.0818, | |
| "step": 5700 | |
| }, | |
| { | |
| "epoch": 30.05, | |
| "learning_rate": 3.7784313725490196e-05, | |
| "loss": 1.0747, | |
| "step": 5800 | |
| }, | |
| { | |
| "epoch": 30.57, | |
| "learning_rate": 3.680392156862745e-05, | |
| "loss": 1.0711, | |
| "step": 5900 | |
| }, | |
| { | |
| "epoch": 31.09, | |
| "learning_rate": 3.5823529411764704e-05, | |
| "loss": 1.0665, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 31.09, | |
| "eval_loss": 0.7878175973892212, | |
| "eval_runtime": 122.0835, | |
| "eval_samples_per_second": 22.059, | |
| "eval_steps_per_second": 2.76, | |
| "eval_wer": 1.01071604532764, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 31.61, | |
| "learning_rate": 3.484313725490196e-05, | |
| "loss": 1.0736, | |
| "step": 6100 | |
| }, | |
| { | |
| "epoch": 32.12, | |
| "learning_rate": 3.386274509803921e-05, | |
| "loss": 1.0588, | |
| "step": 6200 | |
| }, | |
| { | |
| "epoch": 32.64, | |
| "learning_rate": 3.288235294117647e-05, | |
| "loss": 1.0569, | |
| "step": 6300 | |
| }, | |
| { | |
| "epoch": 33.16, | |
| "learning_rate": 3.190196078431372e-05, | |
| "loss": 1.0361, | |
| "step": 6400 | |
| }, | |
| { | |
| "epoch": 33.68, | |
| "learning_rate": 3.092156862745098e-05, | |
| "loss": 1.0321, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 33.68, | |
| "eval_loss": 0.7653242945671082, | |
| "eval_runtime": 125.1204, | |
| "eval_samples_per_second": 21.523, | |
| "eval_steps_per_second": 2.693, | |
| "eval_wer": 1.0081704713417639, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 34.2, | |
| "learning_rate": 2.9941176470588234e-05, | |
| "loss": 1.0294, | |
| "step": 6600 | |
| }, | |
| { | |
| "epoch": 34.72, | |
| "learning_rate": 2.8960784313725488e-05, | |
| "loss": 1.0174, | |
| "step": 6700 | |
| }, | |
| { | |
| "epoch": 35.23, | |
| "learning_rate": 2.7980392156862742e-05, | |
| "loss": 1.0086, | |
| "step": 6800 | |
| }, | |
| { | |
| "epoch": 35.75, | |
| "learning_rate": 2.6999999999999996e-05, | |
| "loss": 1.0138, | |
| "step": 6900 | |
| }, | |
| { | |
| "epoch": 36.27, | |
| "learning_rate": 2.601960784313725e-05, | |
| "loss": 1.0068, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 36.27, | |
| "eval_loss": 0.7634544968605042, | |
| "eval_runtime": 131.459, | |
| "eval_samples_per_second": 20.485, | |
| "eval_steps_per_second": 2.564, | |
| "eval_wer": 1.0064871078994908, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 36.79, | |
| "learning_rate": 2.503921568627451e-05, | |
| "loss": 1.0158, | |
| "step": 7100 | |
| }, | |
| { | |
| "epoch": 37.31, | |
| "learning_rate": 2.406862745098039e-05, | |
| "loss": 1.0167, | |
| "step": 7200 | |
| }, | |
| { | |
| "epoch": 37.82, | |
| "learning_rate": 2.3088235294117645e-05, | |
| "loss": 1.0112, | |
| "step": 7300 | |
| }, | |
| { | |
| "epoch": 38.34, | |
| "learning_rate": 2.21078431372549e-05, | |
| "loss": 0.999, | |
| "step": 7400 | |
| }, | |
| { | |
| "epoch": 38.86, | |
| "learning_rate": 2.1127450980392156e-05, | |
| "loss": 0.9916, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 38.86, | |
| "eval_loss": 0.7728472352027893, | |
| "eval_runtime": 123.666, | |
| "eval_samples_per_second": 21.776, | |
| "eval_steps_per_second": 2.725, | |
| "eval_wer": 1.009032681885367, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 39.38, | |
| "learning_rate": 2.014705882352941e-05, | |
| "loss": 0.9859, | |
| "step": 7600 | |
| }, | |
| { | |
| "epoch": 39.9, | |
| "learning_rate": 1.9166666666666664e-05, | |
| "loss": 0.9878, | |
| "step": 7700 | |
| }, | |
| { | |
| "epoch": 40.41, | |
| "learning_rate": 1.818627450980392e-05, | |
| "loss": 0.9844, | |
| "step": 7800 | |
| }, | |
| { | |
| "epoch": 40.93, | |
| "learning_rate": 1.7205882352941175e-05, | |
| "loss": 0.9794, | |
| "step": 7900 | |
| }, | |
| { | |
| "epoch": 41.45, | |
| "learning_rate": 1.6225490196078432e-05, | |
| "loss": 0.9735, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 41.45, | |
| "eval_loss": 0.7687555551528931, | |
| "eval_runtime": 126.9882, | |
| "eval_samples_per_second": 21.207, | |
| "eval_steps_per_second": 2.654, | |
| "eval_wer": 1.0070208572836261, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 41.97, | |
| "learning_rate": 1.5245098039215684e-05, | |
| "loss": 0.9671, | |
| "step": 8100 | |
| }, | |
| { | |
| "epoch": 42.49, | |
| "learning_rate": 1.426470588235294e-05, | |
| "loss": 0.9732, | |
| "step": 8200 | |
| }, | |
| { | |
| "epoch": 43.01, | |
| "learning_rate": 1.3284313725490195e-05, | |
| "loss": 0.9689, | |
| "step": 8300 | |
| }, | |
| { | |
| "epoch": 43.52, | |
| "learning_rate": 1.230392156862745e-05, | |
| "loss": 0.9695, | |
| "step": 8400 | |
| }, | |
| { | |
| "epoch": 44.04, | |
| "learning_rate": 1.1323529411764705e-05, | |
| "loss": 0.9745, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 44.04, | |
| "eval_loss": 0.7455455660820007, | |
| "eval_runtime": 121.9802, | |
| "eval_samples_per_second": 22.077, | |
| "eval_steps_per_second": 2.763, | |
| "eval_wer": 1.0097306618492363, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 44.56, | |
| "learning_rate": 1.034313725490196e-05, | |
| "loss": 0.9536, | |
| "step": 8600 | |
| }, | |
| { | |
| "epoch": 45.08, | |
| "learning_rate": 9.362745098039214e-06, | |
| "loss": 0.9507, | |
| "step": 8700 | |
| }, | |
| { | |
| "epoch": 45.6, | |
| "learning_rate": 8.38235294117647e-06, | |
| "loss": 0.9537, | |
| "step": 8800 | |
| }, | |
| { | |
| "epoch": 46.11, | |
| "learning_rate": 7.4019607843137246e-06, | |
| "loss": 0.943, | |
| "step": 8900 | |
| }, | |
| { | |
| "epoch": 46.63, | |
| "learning_rate": 6.42156862745098e-06, | |
| "loss": 0.9677, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 46.63, | |
| "eval_loss": 0.7604856491088867, | |
| "eval_runtime": 125.3047, | |
| "eval_samples_per_second": 21.492, | |
| "eval_steps_per_second": 2.689, | |
| "eval_wer": 1.0098948924289703, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 47.15, | |
| "learning_rate": 5.441176470588236e-06, | |
| "loss": 0.9385, | |
| "step": 9100 | |
| }, | |
| { | |
| "epoch": 47.67, | |
| "learning_rate": 4.4607843137254895e-06, | |
| "loss": 0.9402, | |
| "step": 9200 | |
| }, | |
| { | |
| "epoch": 48.19, | |
| "learning_rate": 3.480392156862745e-06, | |
| "loss": 0.9361, | |
| "step": 9300 | |
| }, | |
| { | |
| "epoch": 48.7, | |
| "learning_rate": 2.4999999999999998e-06, | |
| "loss": 0.9271, | |
| "step": 9400 | |
| }, | |
| { | |
| "epoch": 49.22, | |
| "learning_rate": 1.519607843137255e-06, | |
| "loss": 0.9313, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 49.22, | |
| "eval_loss": 0.7526801824569702, | |
| "eval_runtime": 121.7181, | |
| "eval_samples_per_second": 22.125, | |
| "eval_steps_per_second": 2.769, | |
| "eval_wer": 1.0096896042043029, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 49.74, | |
| "learning_rate": 5.392156862745097e-07, | |
| "loss": 0.9178, | |
| "step": 9600 | |
| }, | |
| { | |
| "epoch": 50.0, | |
| "step": 9650, | |
| "total_flos": 4.012526296607357e+19, | |
| "train_loss": 1.8175229964972777, | |
| "train_runtime": 20359.4006, | |
| "train_samples_per_second": 15.167, | |
| "train_steps_per_second": 0.474 | |
| } | |
| ], | |
| "max_steps": 9650, | |
| "num_train_epochs": 50, | |
| "total_flos": 4.012526296607357e+19, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |