| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 25.0, | |
| "eval_steps": 500, | |
| "global_step": 10100, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.12, | |
| "learning_rate": 1.0000000000000002e-06, | |
| "loss": 10.8557, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "learning_rate": 2.0000000000000003e-06, | |
| "loss": 10.5052, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "learning_rate": 3e-06, | |
| "loss": 9.5, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "learning_rate": 4.000000000000001e-06, | |
| "loss": 7.2099, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "learning_rate": 5e-06, | |
| "loss": 5.6515, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "learning_rate": 6e-06, | |
| "loss": 5.1371, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "learning_rate": 7e-06, | |
| "loss": 4.6222, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "learning_rate": 8.000000000000001e-06, | |
| "loss": 4.4197, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_cer": 0.9960395255351591, | |
| "eval_loss": 4.07672643661499, | |
| "eval_runtime": 48.0448, | |
| "eval_samples_per_second": 134.541, | |
| "eval_steps_per_second": 8.409, | |
| "eval_wer": 0.9928168659986352, | |
| "step": 404 | |
| }, | |
| { | |
| "epoch": 1.11, | |
| "learning_rate": 9e-06, | |
| "loss": 4.178, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 1.24, | |
| "learning_rate": 1e-05, | |
| "loss": 3.9224, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.36, | |
| "learning_rate": 1.1000000000000001e-05, | |
| "loss": 3.7071, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 1.49, | |
| "learning_rate": 1.2e-05, | |
| "loss": 3.5329, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.61, | |
| "learning_rate": 1.3000000000000001e-05, | |
| "loss": 3.3935, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 1.73, | |
| "learning_rate": 1.4e-05, | |
| "loss": 3.18, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 1.86, | |
| "learning_rate": 1.5000000000000002e-05, | |
| "loss": 3.0269, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 1.98, | |
| "learning_rate": 1.6000000000000003e-05, | |
| "loss": 2.8984, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_cer": 0.9960395255351591, | |
| "eval_loss": 2.7950191497802734, | |
| "eval_runtime": 47.7431, | |
| "eval_samples_per_second": 135.391, | |
| "eval_steps_per_second": 8.462, | |
| "eval_wer": 0.9928168659986352, | |
| "step": 808 | |
| }, | |
| { | |
| "epoch": 2.1, | |
| "learning_rate": 1.7e-05, | |
| "loss": 2.7876, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 2.23, | |
| "learning_rate": 1.8e-05, | |
| "loss": 2.7058, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 2.35, | |
| "learning_rate": 1.9e-05, | |
| "loss": 2.6375, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 2.48, | |
| "learning_rate": 2e-05, | |
| "loss": 2.5938, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 2.6, | |
| "learning_rate": 1.989010989010989e-05, | |
| "loss": 2.5391, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 2.72, | |
| "learning_rate": 1.9780219780219784e-05, | |
| "loss": 2.4169, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 2.85, | |
| "learning_rate": 1.967032967032967e-05, | |
| "loss": 2.264, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 2.97, | |
| "learning_rate": 1.9560439560439563e-05, | |
| "loss": 2.1179, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_cer": 0.9960395255351591, | |
| "eval_loss": 1.9177764654159546, | |
| "eval_runtime": 47.4792, | |
| "eval_samples_per_second": 136.144, | |
| "eval_steps_per_second": 8.509, | |
| "eval_wer": 0.9928168659986352, | |
| "step": 1212 | |
| }, | |
| { | |
| "epoch": 3.09, | |
| "learning_rate": 1.9450549450549452e-05, | |
| "loss": 2.0151, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 3.22, | |
| "learning_rate": 1.9340659340659342e-05, | |
| "loss": 1.9512, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 3.34, | |
| "learning_rate": 1.923076923076923e-05, | |
| "loss": 1.8304, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 3.47, | |
| "learning_rate": 1.9120879120879124e-05, | |
| "loss": 1.7452, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 3.59, | |
| "learning_rate": 1.901098901098901e-05, | |
| "loss": 1.6516, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 3.71, | |
| "learning_rate": 1.8901098901098903e-05, | |
| "loss": 1.5666, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 3.84, | |
| "learning_rate": 1.8791208791208793e-05, | |
| "loss": 1.5196, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 3.96, | |
| "learning_rate": 1.8681318681318682e-05, | |
| "loss": 1.4282, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_cer": 0.4791183983841264, | |
| "eval_loss": 1.0854228734970093, | |
| "eval_runtime": 47.4019, | |
| "eval_samples_per_second": 136.366, | |
| "eval_steps_per_second": 8.523, | |
| "eval_wer": 0.6262256222389828, | |
| "step": 1616 | |
| }, | |
| { | |
| "epoch": 4.08, | |
| "learning_rate": 1.8571428571428575e-05, | |
| "loss": 1.372, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 4.21, | |
| "learning_rate": 1.8461538461538465e-05, | |
| "loss": 1.3009, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 4.33, | |
| "learning_rate": 1.835164835164835e-05, | |
| "loss": 1.2673, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 4.46, | |
| "learning_rate": 1.8241758241758244e-05, | |
| "loss": 1.1968, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 4.58, | |
| "learning_rate": 1.8131868131868133e-05, | |
| "loss": 1.1639, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 4.7, | |
| "learning_rate": 1.8021978021978023e-05, | |
| "loss": 1.1464, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 4.83, | |
| "learning_rate": 1.7912087912087915e-05, | |
| "loss": 1.0984, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 4.95, | |
| "learning_rate": 1.78021978021978e-05, | |
| "loss": 1.0793, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_cer": 0.294362264599299, | |
| "eval_loss": 0.7671939134597778, | |
| "eval_runtime": 47.1803, | |
| "eval_samples_per_second": 137.006, | |
| "eval_steps_per_second": 8.563, | |
| "eval_wer": 0.49962288546492833, | |
| "step": 2020 | |
| }, | |
| { | |
| "epoch": 5.07, | |
| "learning_rate": 1.7692307692307694e-05, | |
| "loss": 1.0631, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 5.2, | |
| "learning_rate": 1.7582417582417584e-05, | |
| "loss": 1.0267, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 5.32, | |
| "learning_rate": 1.7472527472527473e-05, | |
| "loss": 1.0496, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 5.45, | |
| "learning_rate": 1.7362637362637363e-05, | |
| "loss": 0.9783, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 5.57, | |
| "learning_rate": 1.7252747252747256e-05, | |
| "loss": 0.9673, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 5.69, | |
| "learning_rate": 1.7142857142857142e-05, | |
| "loss": 0.9404, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 5.82, | |
| "learning_rate": 1.7032967032967035e-05, | |
| "loss": 0.9489, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 5.94, | |
| "learning_rate": 1.6923076923076924e-05, | |
| "loss": 0.9064, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_cer": 0.27366878552050533, | |
| "eval_loss": 0.6211672425270081, | |
| "eval_runtime": 47.5588, | |
| "eval_samples_per_second": 135.916, | |
| "eval_steps_per_second": 8.495, | |
| "eval_wer": 0.4573142261968897, | |
| "step": 2424 | |
| }, | |
| { | |
| "epoch": 6.06, | |
| "learning_rate": 1.6813186813186814e-05, | |
| "loss": 0.9205, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 6.19, | |
| "learning_rate": 1.6703296703296707e-05, | |
| "loss": 0.8951, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 6.31, | |
| "learning_rate": 1.6593406593406596e-05, | |
| "loss": 0.8795, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 6.44, | |
| "learning_rate": 1.6483516483516486e-05, | |
| "loss": 0.8895, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 6.56, | |
| "learning_rate": 1.6373626373626375e-05, | |
| "loss": 0.8843, | |
| "step": 2650 | |
| }, | |
| { | |
| "epoch": 6.68, | |
| "learning_rate": 1.6263736263736265e-05, | |
| "loss": 0.8283, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 6.81, | |
| "learning_rate": 1.6153846153846154e-05, | |
| "loss": 0.8294, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 6.93, | |
| "learning_rate": 1.6043956043956047e-05, | |
| "loss": 0.8366, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_cer": 0.24499495039505734, | |
| "eval_loss": 0.5246793627738953, | |
| "eval_runtime": 47.4984, | |
| "eval_samples_per_second": 136.089, | |
| "eval_steps_per_second": 8.506, | |
| "eval_wer": 0.413173867758503, | |
| "step": 2828 | |
| }, | |
| { | |
| "epoch": 7.05, | |
| "learning_rate": 1.5934065934065933e-05, | |
| "loss": 0.8077, | |
| "step": 2850 | |
| }, | |
| { | |
| "epoch": 7.18, | |
| "learning_rate": 1.5824175824175826e-05, | |
| "loss": 0.8373, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 7.3, | |
| "learning_rate": 1.5714285714285715e-05, | |
| "loss": 0.7896, | |
| "step": 2950 | |
| }, | |
| { | |
| "epoch": 7.43, | |
| "learning_rate": 1.5604395604395605e-05, | |
| "loss": 0.7666, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 7.55, | |
| "learning_rate": 1.5494505494505498e-05, | |
| "loss": 0.7846, | |
| "step": 3050 | |
| }, | |
| { | |
| "epoch": 7.67, | |
| "learning_rate": 1.5384615384615387e-05, | |
| "loss": 0.7748, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 7.8, | |
| "learning_rate": 1.5274725274725277e-05, | |
| "loss": 0.7596, | |
| "step": 3150 | |
| }, | |
| { | |
| "epoch": 7.92, | |
| "learning_rate": 1.5164835164835166e-05, | |
| "loss": 0.7425, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_cer": 0.2257272421236064, | |
| "eval_loss": 0.45023345947265625, | |
| "eval_runtime": 47.1685, | |
| "eval_samples_per_second": 137.041, | |
| "eval_steps_per_second": 8.565, | |
| "eval_wer": 0.3785511618719247, | |
| "step": 3232 | |
| }, | |
| { | |
| "epoch": 8.04, | |
| "learning_rate": 1.5054945054945056e-05, | |
| "loss": 0.7275, | |
| "step": 3250 | |
| }, | |
| { | |
| "epoch": 8.17, | |
| "learning_rate": 1.4945054945054947e-05, | |
| "loss": 0.7249, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 8.29, | |
| "learning_rate": 1.4835164835164836e-05, | |
| "loss": 0.7246, | |
| "step": 3350 | |
| }, | |
| { | |
| "epoch": 8.42, | |
| "learning_rate": 1.4725274725274727e-05, | |
| "loss": 0.7058, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 8.54, | |
| "learning_rate": 1.4615384615384615e-05, | |
| "loss": 0.7114, | |
| "step": 3450 | |
| }, | |
| { | |
| "epoch": 8.66, | |
| "learning_rate": 1.4505494505494506e-05, | |
| "loss": 0.7049, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 8.79, | |
| "learning_rate": 1.4395604395604396e-05, | |
| "loss": 0.674, | |
| "step": 3550 | |
| }, | |
| { | |
| "epoch": 8.91, | |
| "learning_rate": 1.4285714285714287e-05, | |
| "loss": 0.7017, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_cer": 0.2082021426166855, | |
| "eval_loss": 0.39116355776786804, | |
| "eval_runtime": 47.8558, | |
| "eval_samples_per_second": 135.072, | |
| "eval_steps_per_second": 8.442, | |
| "eval_wer": 0.3509320116366771, | |
| "step": 3636 | |
| }, | |
| { | |
| "epoch": 9.03, | |
| "learning_rate": 1.4175824175824178e-05, | |
| "loss": 0.6818, | |
| "step": 3650 | |
| }, | |
| { | |
| "epoch": 9.16, | |
| "learning_rate": 1.4065934065934068e-05, | |
| "loss": 0.6718, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 9.28, | |
| "learning_rate": 1.3956043956043957e-05, | |
| "loss": 0.6451, | |
| "step": 3750 | |
| }, | |
| { | |
| "epoch": 9.41, | |
| "learning_rate": 1.3846153846153847e-05, | |
| "loss": 0.6496, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 9.53, | |
| "learning_rate": 1.3736263736263738e-05, | |
| "loss": 0.6437, | |
| "step": 3850 | |
| }, | |
| { | |
| "epoch": 9.65, | |
| "learning_rate": 1.3626373626373627e-05, | |
| "loss": 0.6289, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 9.78, | |
| "learning_rate": 1.3516483516483519e-05, | |
| "loss": 0.6422, | |
| "step": 3950 | |
| }, | |
| { | |
| "epoch": 9.9, | |
| "learning_rate": 1.3406593406593406e-05, | |
| "loss": 0.6275, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_cer": 0.19792471138042336, | |
| "eval_loss": 0.3406831622123718, | |
| "eval_runtime": 47.4155, | |
| "eval_samples_per_second": 136.327, | |
| "eval_steps_per_second": 8.52, | |
| "eval_wer": 0.332794598283231, | |
| "step": 4040 | |
| }, | |
| { | |
| "epoch": 10.02, | |
| "learning_rate": 1.3296703296703298e-05, | |
| "loss": 0.6054, | |
| "step": 4050 | |
| }, | |
| { | |
| "epoch": 10.15, | |
| "learning_rate": 1.3186813186813187e-05, | |
| "loss": 0.6368, | |
| "step": 4100 | |
| }, | |
| { | |
| "epoch": 10.27, | |
| "learning_rate": 1.3076923076923078e-05, | |
| "loss": 0.6517, | |
| "step": 4150 | |
| }, | |
| { | |
| "epoch": 10.4, | |
| "learning_rate": 1.296703296703297e-05, | |
| "loss": 0.5973, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 10.52, | |
| "learning_rate": 1.2857142857142859e-05, | |
| "loss": 0.6143, | |
| "step": 4250 | |
| }, | |
| { | |
| "epoch": 10.64, | |
| "learning_rate": 1.2747252747252747e-05, | |
| "loss": 0.6007, | |
| "step": 4300 | |
| }, | |
| { | |
| "epoch": 10.77, | |
| "learning_rate": 1.2637362637362638e-05, | |
| "loss": 0.6099, | |
| "step": 4350 | |
| }, | |
| { | |
| "epoch": 10.89, | |
| "learning_rate": 1.2527472527472529e-05, | |
| "loss": 0.5853, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "eval_cer": 0.19198399968316204, | |
| "eval_loss": 0.3044804036617279, | |
| "eval_runtime": 47.1639, | |
| "eval_samples_per_second": 137.054, | |
| "eval_steps_per_second": 8.566, | |
| "eval_wer": 0.32263046367129977, | |
| "step": 4444 | |
| }, | |
| { | |
| "epoch": 11.01, | |
| "learning_rate": 1.2417582417582419e-05, | |
| "loss": 0.5824, | |
| "step": 4450 | |
| }, | |
| { | |
| "epoch": 11.14, | |
| "learning_rate": 1.230769230769231e-05, | |
| "loss": 0.5675, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 11.26, | |
| "learning_rate": 1.21978021978022e-05, | |
| "loss": 0.5906, | |
| "step": 4550 | |
| }, | |
| { | |
| "epoch": 11.39, | |
| "learning_rate": 1.2087912087912089e-05, | |
| "loss": 0.6308, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 11.51, | |
| "learning_rate": 1.1978021978021978e-05, | |
| "loss": 0.5545, | |
| "step": 4650 | |
| }, | |
| { | |
| "epoch": 11.63, | |
| "learning_rate": 1.186813186813187e-05, | |
| "loss": 0.5603, | |
| "step": 4700 | |
| }, | |
| { | |
| "epoch": 11.76, | |
| "learning_rate": 1.1758241758241759e-05, | |
| "loss": 0.54, | |
| "step": 4750 | |
| }, | |
| { | |
| "epoch": 11.88, | |
| "learning_rate": 1.164835164835165e-05, | |
| "loss": 0.5551, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "eval_cer": 0.1864789401770332, | |
| "eval_loss": 0.2657150328159332, | |
| "eval_runtime": 47.4966, | |
| "eval_samples_per_second": 136.094, | |
| "eval_steps_per_second": 8.506, | |
| "eval_wer": 0.31386704018963474, | |
| "step": 4848 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "learning_rate": 1.1538461538461538e-05, | |
| "loss": 0.5421, | |
| "step": 4850 | |
| }, | |
| { | |
| "epoch": 12.13, | |
| "learning_rate": 1.1428571428571429e-05, | |
| "loss": 0.5352, | |
| "step": 4900 | |
| }, | |
| { | |
| "epoch": 12.25, | |
| "learning_rate": 1.131868131868132e-05, | |
| "loss": 0.5306, | |
| "step": 4950 | |
| }, | |
| { | |
| "epoch": 12.38, | |
| "learning_rate": 1.120879120879121e-05, | |
| "loss": 0.5574, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 12.5, | |
| "learning_rate": 1.1098901098901101e-05, | |
| "loss": 0.555, | |
| "step": 5050 | |
| }, | |
| { | |
| "epoch": 12.62, | |
| "learning_rate": 1.098901098901099e-05, | |
| "loss": 0.5331, | |
| "step": 5100 | |
| }, | |
| { | |
| "epoch": 12.75, | |
| "learning_rate": 1.087912087912088e-05, | |
| "loss": 0.5215, | |
| "step": 5150 | |
| }, | |
| { | |
| "epoch": 12.87, | |
| "learning_rate": 1.076923076923077e-05, | |
| "loss": 0.5225, | |
| "step": 5200 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "learning_rate": 1.065934065934066e-05, | |
| "loss": 0.5105, | |
| "step": 5250 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "eval_cer": 0.18267688469078597, | |
| "eval_loss": 0.24552097916603088, | |
| "eval_runtime": 47.4722, | |
| "eval_samples_per_second": 136.164, | |
| "eval_steps_per_second": 8.51, | |
| "eval_wer": 0.30855152102862476, | |
| "step": 5252 | |
| }, | |
| { | |
| "epoch": 13.12, | |
| "learning_rate": 1.054945054945055e-05, | |
| "loss": 0.5578, | |
| "step": 5300 | |
| }, | |
| { | |
| "epoch": 13.24, | |
| "learning_rate": 1.0439560439560441e-05, | |
| "loss": 0.5168, | |
| "step": 5350 | |
| }, | |
| { | |
| "epoch": 13.37, | |
| "learning_rate": 1.0329670329670332e-05, | |
| "loss": 0.5363, | |
| "step": 5400 | |
| }, | |
| { | |
| "epoch": 13.49, | |
| "learning_rate": 1.021978021978022e-05, | |
| "loss": 0.512, | |
| "step": 5450 | |
| }, | |
| { | |
| "epoch": 13.61, | |
| "learning_rate": 1.010989010989011e-05, | |
| "loss": 0.5085, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 13.74, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4896, | |
| "step": 5550 | |
| }, | |
| { | |
| "epoch": 13.86, | |
| "learning_rate": 9.890109890109892e-06, | |
| "loss": 0.5197, | |
| "step": 5600 | |
| }, | |
| { | |
| "epoch": 13.99, | |
| "learning_rate": 9.780219780219781e-06, | |
| "loss": 0.5073, | |
| "step": 5650 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "eval_cer": 0.18317194399889106, | |
| "eval_loss": 0.23889391124248505, | |
| "eval_runtime": 47.6163, | |
| "eval_samples_per_second": 135.752, | |
| "eval_steps_per_second": 8.484, | |
| "eval_wer": 0.3091620874187408, | |
| "step": 5656 | |
| }, | |
| { | |
| "epoch": 14.11, | |
| "learning_rate": 9.670329670329671e-06, | |
| "loss": 0.4856, | |
| "step": 5700 | |
| }, | |
| { | |
| "epoch": 14.23, | |
| "learning_rate": 9.560439560439562e-06, | |
| "loss": 0.5116, | |
| "step": 5750 | |
| }, | |
| { | |
| "epoch": 14.36, | |
| "learning_rate": 9.450549450549452e-06, | |
| "loss": 0.4966, | |
| "step": 5800 | |
| }, | |
| { | |
| "epoch": 14.48, | |
| "learning_rate": 9.340659340659341e-06, | |
| "loss": 0.4882, | |
| "step": 5850 | |
| }, | |
| { | |
| "epoch": 14.6, | |
| "learning_rate": 9.230769230769232e-06, | |
| "loss": 0.4852, | |
| "step": 5900 | |
| }, | |
| { | |
| "epoch": 14.73, | |
| "learning_rate": 9.120879120879122e-06, | |
| "loss": 0.4962, | |
| "step": 5950 | |
| }, | |
| { | |
| "epoch": 14.85, | |
| "learning_rate": 9.010989010989011e-06, | |
| "loss": 0.4839, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 14.98, | |
| "learning_rate": 8.9010989010989e-06, | |
| "loss": 0.4722, | |
| "step": 6050 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "eval_cer": 0.17808273431157054, | |
| "eval_loss": 0.2169932723045349, | |
| "eval_runtime": 46.7989, | |
| "eval_samples_per_second": 138.123, | |
| "eval_steps_per_second": 8.633, | |
| "eval_wer": 0.30298459217756707, | |
| "step": 6060 | |
| }, | |
| { | |
| "epoch": 15.1, | |
| "learning_rate": 8.791208791208792e-06, | |
| "loss": 0.4853, | |
| "step": 6100 | |
| }, | |
| { | |
| "epoch": 15.22, | |
| "learning_rate": 8.681318681318681e-06, | |
| "loss": 0.507, | |
| "step": 6150 | |
| }, | |
| { | |
| "epoch": 15.35, | |
| "learning_rate": 8.571428571428571e-06, | |
| "loss": 0.4966, | |
| "step": 6200 | |
| }, | |
| { | |
| "epoch": 15.47, | |
| "learning_rate": 8.461538461538462e-06, | |
| "loss": 0.5092, | |
| "step": 6250 | |
| }, | |
| { | |
| "epoch": 15.59, | |
| "learning_rate": 8.351648351648353e-06, | |
| "loss": 0.4625, | |
| "step": 6300 | |
| }, | |
| { | |
| "epoch": 15.72, | |
| "learning_rate": 8.241758241758243e-06, | |
| "loss": 0.4779, | |
| "step": 6350 | |
| }, | |
| { | |
| "epoch": 15.84, | |
| "learning_rate": 8.131868131868132e-06, | |
| "loss": 0.4631, | |
| "step": 6400 | |
| }, | |
| { | |
| "epoch": 15.97, | |
| "learning_rate": 8.021978021978023e-06, | |
| "loss": 0.481, | |
| "step": 6450 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "eval_cer": 0.17830056040713677, | |
| "eval_loss": 0.20887333154678345, | |
| "eval_runtime": 47.3142, | |
| "eval_samples_per_second": 136.619, | |
| "eval_steps_per_second": 8.539, | |
| "eval_wer": 0.3022662787774306, | |
| "step": 6464 | |
| }, | |
| { | |
| "epoch": 16.09, | |
| "learning_rate": 7.912087912087913e-06, | |
| "loss": 0.4574, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 16.21, | |
| "learning_rate": 7.802197802197802e-06, | |
| "loss": 0.4749, | |
| "step": 6550 | |
| }, | |
| { | |
| "epoch": 16.34, | |
| "learning_rate": 7.692307692307694e-06, | |
| "loss": 0.466, | |
| "step": 6600 | |
| }, | |
| { | |
| "epoch": 16.46, | |
| "learning_rate": 7.582417582417583e-06, | |
| "loss": 0.4489, | |
| "step": 6650 | |
| }, | |
| { | |
| "epoch": 16.58, | |
| "learning_rate": 7.472527472527473e-06, | |
| "loss": 0.4841, | |
| "step": 6700 | |
| }, | |
| { | |
| "epoch": 16.71, | |
| "learning_rate": 7.362637362637364e-06, | |
| "loss": 0.4736, | |
| "step": 6750 | |
| }, | |
| { | |
| "epoch": 16.83, | |
| "learning_rate": 7.252747252747253e-06, | |
| "loss": 0.4453, | |
| "step": 6800 | |
| }, | |
| { | |
| "epoch": 16.96, | |
| "learning_rate": 7.1428571428571436e-06, | |
| "loss": 0.4738, | |
| "step": 6850 | |
| }, | |
| { | |
| "epoch": 17.0, | |
| "eval_cer": 0.17634012554704054, | |
| "eval_loss": 0.20021960139274597, | |
| "eval_runtime": 46.7477, | |
| "eval_samples_per_second": 138.274, | |
| "eval_steps_per_second": 8.642, | |
| "eval_wer": 0.3004345796070826, | |
| "step": 6868 | |
| }, | |
| { | |
| "epoch": 17.08, | |
| "learning_rate": 7.032967032967034e-06, | |
| "loss": 0.4469, | |
| "step": 6900 | |
| }, | |
| { | |
| "epoch": 17.2, | |
| "learning_rate": 6.923076923076923e-06, | |
| "loss": 0.4408, | |
| "step": 6950 | |
| }, | |
| { | |
| "epoch": 17.33, | |
| "learning_rate": 6.813186813186814e-06, | |
| "loss": 0.4471, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 17.45, | |
| "learning_rate": 6.703296703296703e-06, | |
| "loss": 0.4468, | |
| "step": 7050 | |
| }, | |
| { | |
| "epoch": 17.57, | |
| "learning_rate": 6.5934065934065935e-06, | |
| "loss": 0.4348, | |
| "step": 7100 | |
| }, | |
| { | |
| "epoch": 17.7, | |
| "learning_rate": 6.483516483516485e-06, | |
| "loss": 0.433, | |
| "step": 7150 | |
| }, | |
| { | |
| "epoch": 17.82, | |
| "learning_rate": 6.373626373626373e-06, | |
| "loss": 0.4479, | |
| "step": 7200 | |
| }, | |
| { | |
| "epoch": 17.95, | |
| "learning_rate": 6.2637362637362645e-06, | |
| "loss": 0.4518, | |
| "step": 7250 | |
| }, | |
| { | |
| "epoch": 18.0, | |
| "eval_cer": 0.17651834689795837, | |
| "eval_loss": 0.19896449148654938, | |
| "eval_runtime": 46.712, | |
| "eval_samples_per_second": 138.38, | |
| "eval_steps_per_second": 8.649, | |
| "eval_wer": 0.3006141579571167, | |
| "step": 7272 | |
| }, | |
| { | |
| "epoch": 18.07, | |
| "learning_rate": 6.153846153846155e-06, | |
| "loss": 0.4314, | |
| "step": 7300 | |
| }, | |
| { | |
| "epoch": 18.19, | |
| "learning_rate": 6.043956043956044e-06, | |
| "loss": 0.4413, | |
| "step": 7350 | |
| }, | |
| { | |
| "epoch": 18.32, | |
| "learning_rate": 5.934065934065935e-06, | |
| "loss": 0.4276, | |
| "step": 7400 | |
| }, | |
| { | |
| "epoch": 18.44, | |
| "learning_rate": 5.824175824175825e-06, | |
| "loss": 0.4326, | |
| "step": 7450 | |
| }, | |
| { | |
| "epoch": 18.56, | |
| "learning_rate": 5.7142857142857145e-06, | |
| "loss": 0.4662, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 18.69, | |
| "learning_rate": 5.604395604395605e-06, | |
| "loss": 0.4266, | |
| "step": 7550 | |
| }, | |
| { | |
| "epoch": 18.81, | |
| "learning_rate": 5.494505494505495e-06, | |
| "loss": 0.4281, | |
| "step": 7600 | |
| }, | |
| { | |
| "epoch": 18.94, | |
| "learning_rate": 5.384615384615385e-06, | |
| "loss": 0.4402, | |
| "step": 7650 | |
| }, | |
| { | |
| "epoch": 19.0, | |
| "eval_cer": 0.17635992791936475, | |
| "eval_loss": 0.19003288447856903, | |
| "eval_runtime": 47.6013, | |
| "eval_samples_per_second": 135.795, | |
| "eval_steps_per_second": 8.487, | |
| "eval_wer": 0.2998599288869734, | |
| "step": 7676 | |
| }, | |
| { | |
| "epoch": 19.06, | |
| "learning_rate": 5.274725274725275e-06, | |
| "loss": 0.427, | |
| "step": 7700 | |
| }, | |
| { | |
| "epoch": 19.18, | |
| "learning_rate": 5.164835164835166e-06, | |
| "loss": 0.4166, | |
| "step": 7750 | |
| }, | |
| { | |
| "epoch": 19.31, | |
| "learning_rate": 5.054945054945055e-06, | |
| "loss": 0.4612, | |
| "step": 7800 | |
| }, | |
| { | |
| "epoch": 19.43, | |
| "learning_rate": 4.945054945054946e-06, | |
| "loss": 0.4286, | |
| "step": 7850 | |
| }, | |
| { | |
| "epoch": 19.55, | |
| "learning_rate": 4.8351648351648355e-06, | |
| "loss": 0.4271, | |
| "step": 7900 | |
| }, | |
| { | |
| "epoch": 19.68, | |
| "learning_rate": 4.725274725274726e-06, | |
| "loss": 0.4215, | |
| "step": 7950 | |
| }, | |
| { | |
| "epoch": 19.8, | |
| "learning_rate": 4.615384615384616e-06, | |
| "loss": 0.4486, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 19.93, | |
| "learning_rate": 4.505494505494506e-06, | |
| "loss": 0.4387, | |
| "step": 8050 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "eval_cer": 0.17396384086813602, | |
| "eval_loss": 0.18259279429912567, | |
| "eval_runtime": 47.0879, | |
| "eval_samples_per_second": 137.275, | |
| "eval_steps_per_second": 8.58, | |
| "eval_wer": 0.29695075961642065, | |
| "step": 8080 | |
| }, | |
| { | |
| "epoch": 20.05, | |
| "learning_rate": 4.395604395604396e-06, | |
| "loss": 0.4323, | |
| "step": 8100 | |
| }, | |
| { | |
| "epoch": 20.17, | |
| "learning_rate": 4.2857142857142855e-06, | |
| "loss": 0.4124, | |
| "step": 8150 | |
| }, | |
| { | |
| "epoch": 20.3, | |
| "learning_rate": 4.175824175824177e-06, | |
| "loss": 0.4112, | |
| "step": 8200 | |
| }, | |
| { | |
| "epoch": 20.42, | |
| "learning_rate": 4.065934065934066e-06, | |
| "loss": 0.4347, | |
| "step": 8250 | |
| }, | |
| { | |
| "epoch": 20.54, | |
| "learning_rate": 3.9560439560439565e-06, | |
| "loss": 0.4031, | |
| "step": 8300 | |
| }, | |
| { | |
| "epoch": 20.67, | |
| "learning_rate": 3.846153846153847e-06, | |
| "loss": 0.4122, | |
| "step": 8350 | |
| }, | |
| { | |
| "epoch": 20.79, | |
| "learning_rate": 3.7362637362637367e-06, | |
| "loss": 0.4125, | |
| "step": 8400 | |
| }, | |
| { | |
| "epoch": 20.92, | |
| "learning_rate": 3.6263736263736266e-06, | |
| "loss": 0.4212, | |
| "step": 8450 | |
| }, | |
| { | |
| "epoch": 21.0, | |
| "eval_cer": 0.17331036258143726, | |
| "eval_loss": 0.17667880654335022, | |
| "eval_runtime": 47.2671, | |
| "eval_samples_per_second": 136.755, | |
| "eval_steps_per_second": 8.547, | |
| "eval_wer": 0.29547821714614086, | |
| "step": 8484 | |
| }, | |
| { | |
| "epoch": 21.04, | |
| "learning_rate": 3.516483516483517e-06, | |
| "loss": 0.4213, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 21.16, | |
| "learning_rate": 3.406593406593407e-06, | |
| "loss": 0.4322, | |
| "step": 8550 | |
| }, | |
| { | |
| "epoch": 21.29, | |
| "learning_rate": 3.2967032967032968e-06, | |
| "loss": 0.4116, | |
| "step": 8600 | |
| }, | |
| { | |
| "epoch": 21.41, | |
| "learning_rate": 3.1868131868131867e-06, | |
| "loss": 0.404, | |
| "step": 8650 | |
| }, | |
| { | |
| "epoch": 21.53, | |
| "learning_rate": 3.0769230769230774e-06, | |
| "loss": 0.4016, | |
| "step": 8700 | |
| }, | |
| { | |
| "epoch": 21.66, | |
| "learning_rate": 2.9670329670329673e-06, | |
| "loss": 0.4031, | |
| "step": 8750 | |
| }, | |
| { | |
| "epoch": 21.78, | |
| "learning_rate": 2.8571428571428573e-06, | |
| "loss": 0.4072, | |
| "step": 8800 | |
| }, | |
| { | |
| "epoch": 21.91, | |
| "learning_rate": 2.7472527472527476e-06, | |
| "loss": 0.3893, | |
| "step": 8850 | |
| }, | |
| { | |
| "epoch": 22.0, | |
| "eval_cer": 0.17188459177409454, | |
| "eval_loss": 0.17065875232219696, | |
| "eval_runtime": 47.0103, | |
| "eval_samples_per_second": 137.502, | |
| "eval_steps_per_second": 8.594, | |
| "eval_wer": 0.2937183493158065, | |
| "step": 8888 | |
| }, | |
| { | |
| "epoch": 22.03, | |
| "learning_rate": 2.6373626373626375e-06, | |
| "loss": 0.4191, | |
| "step": 8900 | |
| }, | |
| { | |
| "epoch": 22.15, | |
| "learning_rate": 2.5274725274725274e-06, | |
| "loss": 0.4094, | |
| "step": 8950 | |
| }, | |
| { | |
| "epoch": 22.28, | |
| "learning_rate": 2.4175824175824177e-06, | |
| "loss": 0.4275, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 22.4, | |
| "learning_rate": 2.307692307692308e-06, | |
| "loss": 0.4044, | |
| "step": 9050 | |
| }, | |
| { | |
| "epoch": 22.52, | |
| "learning_rate": 2.197802197802198e-06, | |
| "loss": 0.4023, | |
| "step": 9100 | |
| }, | |
| { | |
| "epoch": 22.65, | |
| "learning_rate": 2.0879120879120883e-06, | |
| "loss": 0.3937, | |
| "step": 9150 | |
| }, | |
| { | |
| "epoch": 22.77, | |
| "learning_rate": 1.9780219780219782e-06, | |
| "loss": 0.4118, | |
| "step": 9200 | |
| }, | |
| { | |
| "epoch": 22.9, | |
| "learning_rate": 1.8681318681318684e-06, | |
| "loss": 0.4055, | |
| "step": 9250 | |
| }, | |
| { | |
| "epoch": 23.0, | |
| "eval_cer": 0.17228063922057862, | |
| "eval_loss": 0.17037759721279144, | |
| "eval_runtime": 47.8098, | |
| "eval_samples_per_second": 135.202, | |
| "eval_steps_per_second": 8.45, | |
| "eval_wer": 0.2942930000359157, | |
| "step": 9292 | |
| }, | |
| { | |
| "epoch": 23.02, | |
| "learning_rate": 1.7582417582417585e-06, | |
| "loss": 0.3926, | |
| "step": 9300 | |
| }, | |
| { | |
| "epoch": 23.14, | |
| "learning_rate": 1.6483516483516484e-06, | |
| "loss": 0.4367, | |
| "step": 9350 | |
| }, | |
| { | |
| "epoch": 23.27, | |
| "learning_rate": 1.5384615384615387e-06, | |
| "loss": 0.4156, | |
| "step": 9400 | |
| }, | |
| { | |
| "epoch": 23.39, | |
| "learning_rate": 1.4285714285714286e-06, | |
| "loss": 0.3998, | |
| "step": 9450 | |
| }, | |
| { | |
| "epoch": 23.51, | |
| "learning_rate": 1.3186813186813187e-06, | |
| "loss": 0.4496, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 23.64, | |
| "learning_rate": 1.2087912087912089e-06, | |
| "loss": 0.4061, | |
| "step": 9550 | |
| }, | |
| { | |
| "epoch": 23.76, | |
| "learning_rate": 1.098901098901099e-06, | |
| "loss": 0.3822, | |
| "step": 9600 | |
| }, | |
| { | |
| "epoch": 23.89, | |
| "learning_rate": 9.890109890109891e-07, | |
| "loss": 0.394, | |
| "step": 9650 | |
| }, | |
| { | |
| "epoch": 24.0, | |
| "eval_cer": 0.17158755618923147, | |
| "eval_loss": 0.16843372583389282, | |
| "eval_runtime": 47.3805, | |
| "eval_samples_per_second": 136.427, | |
| "eval_steps_per_second": 8.527, | |
| "eval_wer": 0.29339510828574505, | |
| "step": 9696 | |
| }, | |
| { | |
| "epoch": 24.01, | |
| "learning_rate": 8.791208791208792e-07, | |
| "loss": 0.4072, | |
| "step": 9700 | |
| }, | |
| { | |
| "epoch": 24.13, | |
| "learning_rate": 7.692307692307694e-07, | |
| "loss": 0.3867, | |
| "step": 9750 | |
| }, | |
| { | |
| "epoch": 24.26, | |
| "learning_rate": 6.593406593406594e-07, | |
| "loss": 0.4118, | |
| "step": 9800 | |
| }, | |
| { | |
| "epoch": 24.38, | |
| "learning_rate": 5.494505494505495e-07, | |
| "loss": 0.3978, | |
| "step": 9850 | |
| }, | |
| { | |
| "epoch": 24.5, | |
| "learning_rate": 4.395604395604396e-07, | |
| "loss": 0.4148, | |
| "step": 9900 | |
| }, | |
| { | |
| "epoch": 24.63, | |
| "learning_rate": 3.296703296703297e-07, | |
| "loss": 0.4074, | |
| "step": 9950 | |
| }, | |
| { | |
| "epoch": 24.75, | |
| "learning_rate": 2.197802197802198e-07, | |
| "loss": 0.4055, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 24.88, | |
| "learning_rate": 1.098901098901099e-07, | |
| "loss": 0.4309, | |
| "step": 10050 | |
| }, | |
| { | |
| "epoch": 25.0, | |
| "learning_rate": 0.0, | |
| "loss": 0.3997, | |
| "step": 10100 | |
| }, | |
| { | |
| "epoch": 25.0, | |
| "eval_cer": 0.17101328739182955, | |
| "eval_loss": 0.16850610077381134, | |
| "eval_runtime": 47.0414, | |
| "eval_samples_per_second": 137.411, | |
| "eval_steps_per_second": 8.588, | |
| "eval_wer": 0.29274862622562225, | |
| "step": 10100 | |
| }, | |
| { | |
| "epoch": 25.0, | |
| "step": 10100, | |
| "total_flos": 5.390210696735239e+18, | |
| "train_loss": 1.0812275544723662, | |
| "train_runtime": 21945.8534, | |
| "train_samples_per_second": 29.454, | |
| "train_steps_per_second": 0.46 | |
| } | |
| ], | |
| "logging_steps": 50, | |
| "max_steps": 10100, | |
| "num_train_epochs": 25, | |
| "save_steps": 500, | |
| "total_flos": 5.390210696735239e+18, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |