| { | |
| "best_metric": 10.587474512857398, | |
| "best_model_checkpoint": "./whisper-small-mix-it/checkpoint-5000", | |
| "epoch": 1.0, | |
| "eval_steps": 1000, | |
| "global_step": 5000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.005, | |
| "grad_norm": 7.571747303009033, | |
| "learning_rate": 4.800000000000001e-07, | |
| "loss": 0.8106, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 4.4452080726623535, | |
| "learning_rate": 9.800000000000001e-07, | |
| "loss": 0.6948, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.015, | |
| "grad_norm": 3.8859105110168457, | |
| "learning_rate": 1.48e-06, | |
| "loss": 0.5709, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 2.8216567039489746, | |
| "learning_rate": 1.98e-06, | |
| "loss": 0.4134, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.025, | |
| "grad_norm": 2.741655111312866, | |
| "learning_rate": 2.4800000000000004e-06, | |
| "loss": 0.3693, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 2.7941854000091553, | |
| "learning_rate": 2.9800000000000003e-06, | |
| "loss": 0.3578, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.035, | |
| "grad_norm": 2.9997148513793945, | |
| "learning_rate": 3.48e-06, | |
| "loss": 0.3242, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 3.136533498764038, | |
| "learning_rate": 3.980000000000001e-06, | |
| "loss": 0.3224, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.045, | |
| "grad_norm": 2.4372665882110596, | |
| "learning_rate": 4.48e-06, | |
| "loss": 0.3051, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 2.798402786254883, | |
| "learning_rate": 4.980000000000001e-06, | |
| "loss": 0.3132, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.055, | |
| "grad_norm": 2.634129524230957, | |
| "learning_rate": 5.480000000000001e-06, | |
| "loss": 0.3025, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 2.955035924911499, | |
| "learning_rate": 5.98e-06, | |
| "loss": 0.3381, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.065, | |
| "grad_norm": 2.6961045265197754, | |
| "learning_rate": 6.480000000000001e-06, | |
| "loss": 0.3108, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 2.5583386421203613, | |
| "learning_rate": 6.98e-06, | |
| "loss": 0.3108, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.075, | |
| "grad_norm": 2.4596619606018066, | |
| "learning_rate": 7.48e-06, | |
| "loss": 0.3041, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 2.315897226333618, | |
| "learning_rate": 7.980000000000002e-06, | |
| "loss": 0.2813, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.085, | |
| "grad_norm": 2.350675582885742, | |
| "learning_rate": 8.48e-06, | |
| "loss": 0.2673, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 2.274202585220337, | |
| "learning_rate": 8.98e-06, | |
| "loss": 0.2589, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.095, | |
| "grad_norm": 2.398346424102783, | |
| "learning_rate": 9.48e-06, | |
| "loss": 0.2554, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 2.4121956825256348, | |
| "learning_rate": 9.980000000000001e-06, | |
| "loss": 0.2365, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.105, | |
| "grad_norm": 2.320568323135376, | |
| "learning_rate": 9.946666666666667e-06, | |
| "loss": 0.2527, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 2.2950375080108643, | |
| "learning_rate": 9.891111111111113e-06, | |
| "loss": 0.2731, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.115, | |
| "grad_norm": 2.712926149368286, | |
| "learning_rate": 9.835555555555556e-06, | |
| "loss": 0.3282, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 2.2590723037719727, | |
| "learning_rate": 9.780000000000001e-06, | |
| "loss": 0.3311, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.125, | |
| "grad_norm": 2.4605965614318848, | |
| "learning_rate": 9.724444444444445e-06, | |
| "loss": 0.2867, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 2.426866292953491, | |
| "learning_rate": 9.66888888888889e-06, | |
| "loss": 0.2805, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.135, | |
| "grad_norm": 2.631782293319702, | |
| "learning_rate": 9.613333333333335e-06, | |
| "loss": 0.2511, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 2.4071285724639893, | |
| "learning_rate": 9.557777777777777e-06, | |
| "loss": 0.2404, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.145, | |
| "grad_norm": 2.018366575241089, | |
| "learning_rate": 9.502222222222223e-06, | |
| "loss": 0.2236, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 2.0572237968444824, | |
| "learning_rate": 9.446666666666667e-06, | |
| "loss": 0.2244, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.155, | |
| "grad_norm": 2.3840184211730957, | |
| "learning_rate": 9.391111111111111e-06, | |
| "loss": 0.2346, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 2.5831480026245117, | |
| "learning_rate": 9.335555555555557e-06, | |
| "loss": 0.325, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.165, | |
| "grad_norm": 2.831907272338867, | |
| "learning_rate": 9.280000000000001e-06, | |
| "loss": 0.332, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 2.4622819423675537, | |
| "learning_rate": 9.224444444444445e-06, | |
| "loss": 0.3338, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.175, | |
| "grad_norm": 2.7406504154205322, | |
| "learning_rate": 9.168888888888889e-06, | |
| "loss": 0.2918, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 2.1826331615448, | |
| "learning_rate": 9.113333333333335e-06, | |
| "loss": 0.2737, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.185, | |
| "grad_norm": 2.2491261959075928, | |
| "learning_rate": 9.057777777777779e-06, | |
| "loss": 0.2247, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "grad_norm": 1.9993724822998047, | |
| "learning_rate": 9.002222222222223e-06, | |
| "loss": 0.2251, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.195, | |
| "grad_norm": 2.3380165100097656, | |
| "learning_rate": 8.946666666666669e-06, | |
| "loss": 0.2129, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 2.147649049758911, | |
| "learning_rate": 8.891111111111111e-06, | |
| "loss": 0.2213, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "eval_loss": 0.24065828323364258, | |
| "eval_runtime": 704.3288, | |
| "eval_samples_per_second": 21.517, | |
| "eval_steps_per_second": 2.691, | |
| "eval_wer": 13.460510600671737, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.205, | |
| "grad_norm": 2.1159000396728516, | |
| "learning_rate": 8.835555555555557e-06, | |
| "loss": 0.2184, | |
| "step": 1025 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "grad_norm": 1.7620288133621216, | |
| "learning_rate": 8.78e-06, | |
| "loss": 0.2027, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 0.215, | |
| "grad_norm": 1.9889628887176514, | |
| "learning_rate": 8.724444444444445e-06, | |
| "loss": 0.2087, | |
| "step": 1075 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "grad_norm": 1.8565176725387573, | |
| "learning_rate": 8.66888888888889e-06, | |
| "loss": 0.213, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.225, | |
| "grad_norm": 1.9121789932250977, | |
| "learning_rate": 8.613333333333333e-06, | |
| "loss": 0.1944, | |
| "step": 1125 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "grad_norm": 2.0282061100006104, | |
| "learning_rate": 8.557777777777778e-06, | |
| "loss": 0.2029, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 0.235, | |
| "grad_norm": 2.0103440284729004, | |
| "learning_rate": 8.502222222222223e-06, | |
| "loss": 0.193, | |
| "step": 1175 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 1.8842862844467163, | |
| "learning_rate": 8.446666666666668e-06, | |
| "loss": 0.1881, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.245, | |
| "grad_norm": 2.2236127853393555, | |
| "learning_rate": 8.391111111111112e-06, | |
| "loss": 0.1776, | |
| "step": 1225 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 1.9313597679138184, | |
| "learning_rate": 8.335555555555556e-06, | |
| "loss": 0.1725, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 0.255, | |
| "grad_norm": 1.8937866687774658, | |
| "learning_rate": 8.28e-06, | |
| "loss": 0.1967, | |
| "step": 1275 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "grad_norm": 2.6142194271087646, | |
| "learning_rate": 8.224444444444444e-06, | |
| "loss": 0.2598, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.265, | |
| "grad_norm": 2.1339993476867676, | |
| "learning_rate": 8.16888888888889e-06, | |
| "loss": 0.2464, | |
| "step": 1325 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "grad_norm": 1.7635416984558105, | |
| "learning_rate": 8.113333333333334e-06, | |
| "loss": 0.2103, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 0.275, | |
| "grad_norm": 1.9434435367584229, | |
| "learning_rate": 8.057777777777778e-06, | |
| "loss": 0.2077, | |
| "step": 1375 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "grad_norm": 1.7413808107376099, | |
| "learning_rate": 8.002222222222222e-06, | |
| "loss": 0.2123, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.285, | |
| "grad_norm": 1.676006555557251, | |
| "learning_rate": 7.946666666666666e-06, | |
| "loss": 0.1803, | |
| "step": 1425 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "grad_norm": 1.9897364377975464, | |
| "learning_rate": 7.891111111111112e-06, | |
| "loss": 0.1647, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 0.295, | |
| "grad_norm": 2.0311105251312256, | |
| "learning_rate": 7.835555555555556e-06, | |
| "loss": 0.1721, | |
| "step": 1475 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "grad_norm": 2.149789333343506, | |
| "learning_rate": 7.78e-06, | |
| "loss": 0.1799, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.305, | |
| "grad_norm": 2.448418617248535, | |
| "learning_rate": 7.724444444444446e-06, | |
| "loss": 0.1805, | |
| "step": 1525 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "grad_norm": 2.129755735397339, | |
| "learning_rate": 7.66888888888889e-06, | |
| "loss": 0.2502, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 0.315, | |
| "grad_norm": 1.7567689418792725, | |
| "learning_rate": 7.613333333333334e-06, | |
| "loss": 0.2249, | |
| "step": 1575 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 1.5997744798660278, | |
| "learning_rate": 7.557777777777779e-06, | |
| "loss": 0.1647, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 0.325, | |
| "grad_norm": 1.6361225843429565, | |
| "learning_rate": 7.502222222222223e-06, | |
| "loss": 0.1646, | |
| "step": 1625 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "grad_norm": 1.6844338178634644, | |
| "learning_rate": 7.446666666666668e-06, | |
| "loss": 0.1532, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 0.335, | |
| "grad_norm": 1.4503517150878906, | |
| "learning_rate": 7.3911111111111125e-06, | |
| "loss": 0.1494, | |
| "step": 1675 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "grad_norm": 1.5753148794174194, | |
| "learning_rate": 7.335555555555556e-06, | |
| "loss": 0.1488, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 0.345, | |
| "grad_norm": 1.8972795009613037, | |
| "learning_rate": 7.280000000000001e-06, | |
| "loss": 0.1671, | |
| "step": 1725 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "grad_norm": 2.3437507152557373, | |
| "learning_rate": 7.224444444444445e-06, | |
| "loss": 0.1862, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 0.355, | |
| "grad_norm": 1.9804730415344238, | |
| "learning_rate": 7.1688888888888895e-06, | |
| "loss": 0.1904, | |
| "step": 1775 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "grad_norm": 2.2449533939361572, | |
| "learning_rate": 7.113333333333334e-06, | |
| "loss": 0.1938, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 0.365, | |
| "grad_norm": 2.0063376426696777, | |
| "learning_rate": 7.057777777777778e-06, | |
| "loss": 0.1933, | |
| "step": 1825 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "grad_norm": 1.804547905921936, | |
| "learning_rate": 7.0022222222222225e-06, | |
| "loss": 0.1592, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 0.375, | |
| "grad_norm": 1.7678463459014893, | |
| "learning_rate": 6.946666666666667e-06, | |
| "loss": 0.1508, | |
| "step": 1875 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "grad_norm": 1.8199127912521362, | |
| "learning_rate": 6.891111111111111e-06, | |
| "loss": 0.1678, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 0.385, | |
| "grad_norm": 2.3224880695343018, | |
| "learning_rate": 6.835555555555556e-06, | |
| "loss": 0.1972, | |
| "step": 1925 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "grad_norm": 2.035727024078369, | |
| "learning_rate": 6.780000000000001e-06, | |
| "loss": 0.2074, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 0.395, | |
| "grad_norm": 2.022366523742676, | |
| "learning_rate": 6.724444444444444e-06, | |
| "loss": 0.1959, | |
| "step": 1975 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 2.046525716781616, | |
| "learning_rate": 6.668888888888889e-06, | |
| "loss": 0.1582, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "eval_loss": 0.21434400975704193, | |
| "eval_runtime": 707.9201, | |
| "eval_samples_per_second": 21.408, | |
| "eval_steps_per_second": 2.677, | |
| "eval_wer": 12.264182068929111, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.405, | |
| "grad_norm": 1.9399248361587524, | |
| "learning_rate": 6.613333333333334e-06, | |
| "loss": 0.1593, | |
| "step": 2025 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "grad_norm": 1.952101469039917, | |
| "learning_rate": 6.557777777777778e-06, | |
| "loss": 0.1676, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 0.415, | |
| "grad_norm": 2.004068613052368, | |
| "learning_rate": 6.502222222222223e-06, | |
| "loss": 0.2002, | |
| "step": 2075 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "grad_norm": 2.298147439956665, | |
| "learning_rate": 6.446666666666668e-06, | |
| "loss": 0.1977, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 0.425, | |
| "grad_norm": 2.5287961959838867, | |
| "learning_rate": 6.391111111111111e-06, | |
| "loss": 0.3416, | |
| "step": 2125 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "grad_norm": 1.9228953123092651, | |
| "learning_rate": 6.335555555555556e-06, | |
| "loss": 0.3219, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 0.435, | |
| "grad_norm": 1.8267163038253784, | |
| "learning_rate": 6.280000000000001e-06, | |
| "loss": 0.23, | |
| "step": 2175 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "grad_norm": 1.5915461778640747, | |
| "learning_rate": 6.224444444444445e-06, | |
| "loss": 0.1586, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 0.445, | |
| "grad_norm": 1.4649773836135864, | |
| "learning_rate": 6.16888888888889e-06, | |
| "loss": 0.1314, | |
| "step": 2225 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "grad_norm": 1.800898790359497, | |
| "learning_rate": 6.113333333333333e-06, | |
| "loss": 0.1303, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 0.455, | |
| "grad_norm": 1.6255425214767456, | |
| "learning_rate": 6.057777777777778e-06, | |
| "loss": 0.1482, | |
| "step": 2275 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "grad_norm": 1.6611112356185913, | |
| "learning_rate": 6.002222222222223e-06, | |
| "loss": 0.1262, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 0.465, | |
| "grad_norm": 1.6362543106079102, | |
| "learning_rate": 5.946666666666668e-06, | |
| "loss": 0.1413, | |
| "step": 2325 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "grad_norm": 1.7811627388000488, | |
| "learning_rate": 5.891111111111112e-06, | |
| "loss": 0.1587, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 0.475, | |
| "grad_norm": 1.9331281185150146, | |
| "learning_rate": 5.8355555555555565e-06, | |
| "loss": 0.1623, | |
| "step": 2375 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 1.6756068468093872, | |
| "learning_rate": 5.78e-06, | |
| "loss": 0.1495, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 0.485, | |
| "grad_norm": 1.714150309562683, | |
| "learning_rate": 5.724444444444445e-06, | |
| "loss": 0.1542, | |
| "step": 2425 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "grad_norm": 1.4862266778945923, | |
| "learning_rate": 5.6688888888888895e-06, | |
| "loss": 0.1569, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 0.495, | |
| "grad_norm": 1.72484290599823, | |
| "learning_rate": 5.613333333333334e-06, | |
| "loss": 0.1449, | |
| "step": 2475 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 1.7387865781784058, | |
| "learning_rate": 5.557777777777778e-06, | |
| "loss": 0.137, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.505, | |
| "grad_norm": 2.0343728065490723, | |
| "learning_rate": 5.5022222222222224e-06, | |
| "loss": 0.1448, | |
| "step": 2525 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "grad_norm": 1.5081489086151123, | |
| "learning_rate": 5.4466666666666665e-06, | |
| "loss": 0.1304, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 0.515, | |
| "grad_norm": 1.6394362449645996, | |
| "learning_rate": 5.391111111111111e-06, | |
| "loss": 0.1362, | |
| "step": 2575 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "grad_norm": 1.968766689300537, | |
| "learning_rate": 5.335555555555556e-06, | |
| "loss": 0.157, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 0.525, | |
| "grad_norm": 2.227083206176758, | |
| "learning_rate": 5.28e-06, | |
| "loss": 0.1589, | |
| "step": 2625 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "grad_norm": 1.7511582374572754, | |
| "learning_rate": 5.224444444444445e-06, | |
| "loss": 0.1544, | |
| "step": 2650 | |
| }, | |
| { | |
| "epoch": 0.535, | |
| "grad_norm": 1.838564157485962, | |
| "learning_rate": 5.168888888888889e-06, | |
| "loss": 0.1408, | |
| "step": 2675 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "grad_norm": 1.5015802383422852, | |
| "learning_rate": 5.113333333333333e-06, | |
| "loss": 0.1452, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 0.545, | |
| "grad_norm": 2.10288667678833, | |
| "learning_rate": 5.057777777777778e-06, | |
| "loss": 0.1664, | |
| "step": 2725 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "grad_norm": 2.287795305252075, | |
| "learning_rate": 5.002222222222223e-06, | |
| "loss": 0.1751, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 0.555, | |
| "grad_norm": 2.386364698410034, | |
| "learning_rate": 4.946666666666667e-06, | |
| "loss": 0.1635, | |
| "step": 2775 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "grad_norm": 1.9961427450180054, | |
| "learning_rate": 4.891111111111111e-06, | |
| "loss": 0.1544, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 0.565, | |
| "grad_norm": 1.952929973602295, | |
| "learning_rate": 4.835555555555556e-06, | |
| "loss": 0.1682, | |
| "step": 2825 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "grad_norm": 1.7093037366867065, | |
| "learning_rate": 4.78e-06, | |
| "loss": 0.1738, | |
| "step": 2850 | |
| }, | |
| { | |
| "epoch": 0.575, | |
| "grad_norm": 1.9177967309951782, | |
| "learning_rate": 4.724444444444445e-06, | |
| "loss": 0.1369, | |
| "step": 2875 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "grad_norm": 1.7161290645599365, | |
| "learning_rate": 4.66888888888889e-06, | |
| "loss": 0.1386, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 0.585, | |
| "grad_norm": 2.1072142124176025, | |
| "learning_rate": 4.613333333333334e-06, | |
| "loss": 0.222, | |
| "step": 2925 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "grad_norm": 1.521346092224121, | |
| "learning_rate": 4.557777777777778e-06, | |
| "loss": 0.1801, | |
| "step": 2950 | |
| }, | |
| { | |
| "epoch": 0.595, | |
| "grad_norm": 1.8204540014266968, | |
| "learning_rate": 4.502222222222223e-06, | |
| "loss": 0.155, | |
| "step": 2975 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 1.8535057306289673, | |
| "learning_rate": 4.446666666666667e-06, | |
| "loss": 0.1913, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "eval_loss": 0.20219513773918152, | |
| "eval_runtime": 700.7499, | |
| "eval_samples_per_second": 21.627, | |
| "eval_steps_per_second": 2.704, | |
| "eval_wer": 11.232818861475318, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.605, | |
| "grad_norm": 1.7448265552520752, | |
| "learning_rate": 4.391111111111112e-06, | |
| "loss": 0.1911, | |
| "step": 3025 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "grad_norm": 1.9879142045974731, | |
| "learning_rate": 4.3355555555555565e-06, | |
| "loss": 0.1887, | |
| "step": 3050 | |
| }, | |
| { | |
| "epoch": 0.615, | |
| "grad_norm": 1.6797877550125122, | |
| "learning_rate": 4.2800000000000005e-06, | |
| "loss": 0.1838, | |
| "step": 3075 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "grad_norm": 2.2724123001098633, | |
| "learning_rate": 4.2244444444444446e-06, | |
| "loss": 0.1914, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 0.625, | |
| "grad_norm": 2.0982179641723633, | |
| "learning_rate": 4.168888888888889e-06, | |
| "loss": 0.1798, | |
| "step": 3125 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "grad_norm": 2.1598305702209473, | |
| "learning_rate": 4.1133333333333335e-06, | |
| "loss": 0.1717, | |
| "step": 3150 | |
| }, | |
| { | |
| "epoch": 0.635, | |
| "grad_norm": 2.225451707839966, | |
| "learning_rate": 4.057777777777778e-06, | |
| "loss": 0.1788, | |
| "step": 3175 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 1.8868625164031982, | |
| "learning_rate": 4.002222222222222e-06, | |
| "loss": 0.1768, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 0.645, | |
| "grad_norm": 1.864670991897583, | |
| "learning_rate": 3.946666666666667e-06, | |
| "loss": 0.1661, | |
| "step": 3225 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "grad_norm": 2.103564500808716, | |
| "learning_rate": 3.891111111111111e-06, | |
| "loss": 0.1931, | |
| "step": 3250 | |
| }, | |
| { | |
| "epoch": 0.655, | |
| "grad_norm": 1.8054132461547852, | |
| "learning_rate": 3.835555555555555e-06, | |
| "loss": 0.1883, | |
| "step": 3275 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "grad_norm": 1.7523831129074097, | |
| "learning_rate": 3.7800000000000002e-06, | |
| "loss": 0.1575, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 0.665, | |
| "grad_norm": 1.9287195205688477, | |
| "learning_rate": 3.724444444444445e-06, | |
| "loss": 0.1597, | |
| "step": 3325 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "grad_norm": 1.4184482097625732, | |
| "learning_rate": 3.668888888888889e-06, | |
| "loss": 0.1468, | |
| "step": 3350 | |
| }, | |
| { | |
| "epoch": 0.675, | |
| "grad_norm": 1.6231290102005005, | |
| "learning_rate": 3.6133333333333336e-06, | |
| "loss": 0.139, | |
| "step": 3375 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "grad_norm": 1.2594499588012695, | |
| "learning_rate": 3.5577777777777785e-06, | |
| "loss": 0.1551, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 0.685, | |
| "grad_norm": 1.8359473943710327, | |
| "learning_rate": 3.5022222222222225e-06, | |
| "loss": 0.1454, | |
| "step": 3425 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "grad_norm": 1.5489341020584106, | |
| "learning_rate": 3.446666666666667e-06, | |
| "loss": 0.1627, | |
| "step": 3450 | |
| }, | |
| { | |
| "epoch": 0.695, | |
| "grad_norm": 1.7028127908706665, | |
| "learning_rate": 3.391111111111111e-06, | |
| "loss": 0.1429, | |
| "step": 3475 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "grad_norm": 1.9561749696731567, | |
| "learning_rate": 3.335555555555556e-06, | |
| "loss": 0.1368, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.705, | |
| "grad_norm": 2.1024222373962402, | |
| "learning_rate": 3.2800000000000004e-06, | |
| "loss": 0.1661, | |
| "step": 3525 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "grad_norm": 2.004310131072998, | |
| "learning_rate": 3.2244444444444444e-06, | |
| "loss": 0.1432, | |
| "step": 3550 | |
| }, | |
| { | |
| "epoch": 0.715, | |
| "grad_norm": 1.3818813562393188, | |
| "learning_rate": 3.1688888888888893e-06, | |
| "loss": 0.1251, | |
| "step": 3575 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "grad_norm": 1.6478841304779053, | |
| "learning_rate": 3.1133333333333337e-06, | |
| "loss": 0.1261, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 0.725, | |
| "grad_norm": 1.5140513181686401, | |
| "learning_rate": 3.0577777777777778e-06, | |
| "loss": 0.1178, | |
| "step": 3625 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "grad_norm": 1.7134217023849487, | |
| "learning_rate": 3.0022222222222227e-06, | |
| "loss": 0.1448, | |
| "step": 3650 | |
| }, | |
| { | |
| "epoch": 0.735, | |
| "grad_norm": 2.0633442401885986, | |
| "learning_rate": 2.946666666666667e-06, | |
| "loss": 0.1637, | |
| "step": 3675 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "grad_norm": 1.6094017028808594, | |
| "learning_rate": 2.891111111111111e-06, | |
| "loss": 0.1342, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 0.745, | |
| "grad_norm": 1.776197910308838, | |
| "learning_rate": 2.835555555555556e-06, | |
| "loss": 0.1567, | |
| "step": 3725 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "grad_norm": 2.1370580196380615, | |
| "learning_rate": 2.7800000000000005e-06, | |
| "loss": 0.1423, | |
| "step": 3750 | |
| }, | |
| { | |
| "epoch": 0.755, | |
| "grad_norm": 2.156872510910034, | |
| "learning_rate": 2.7244444444444445e-06, | |
| "loss": 0.1311, | |
| "step": 3775 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "grad_norm": 1.609500765800476, | |
| "learning_rate": 2.6688888888888894e-06, | |
| "loss": 0.1392, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 0.765, | |
| "grad_norm": 1.677157998085022, | |
| "learning_rate": 2.6133333333333334e-06, | |
| "loss": 0.1383, | |
| "step": 3825 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "grad_norm": 1.6813416481018066, | |
| "learning_rate": 2.557777777777778e-06, | |
| "loss": 0.1359, | |
| "step": 3850 | |
| }, | |
| { | |
| "epoch": 0.775, | |
| "grad_norm": 2.159526824951172, | |
| "learning_rate": 2.5022222222222224e-06, | |
| "loss": 0.1295, | |
| "step": 3875 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "grad_norm": 1.3956880569458008, | |
| "learning_rate": 2.446666666666667e-06, | |
| "loss": 0.127, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 0.785, | |
| "grad_norm": 1.8989673852920532, | |
| "learning_rate": 2.3911111111111113e-06, | |
| "loss": 0.1273, | |
| "step": 3925 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "grad_norm": 1.9193179607391357, | |
| "learning_rate": 2.3355555555555557e-06, | |
| "loss": 0.1647, | |
| "step": 3950 | |
| }, | |
| { | |
| "epoch": 0.795, | |
| "grad_norm": 1.7393921613693237, | |
| "learning_rate": 2.28e-06, | |
| "loss": 0.1726, | |
| "step": 3975 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 1.9714525938034058, | |
| "learning_rate": 2.2244444444444447e-06, | |
| "loss": 0.1538, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "eval_loss": 0.19509215652942657, | |
| "eval_runtime": 768.8189, | |
| "eval_samples_per_second": 19.712, | |
| "eval_steps_per_second": 2.465, | |
| "eval_wer": 11.118662857067445, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.805, | |
| "grad_norm": 1.8712801933288574, | |
| "learning_rate": 2.168888888888889e-06, | |
| "loss": 0.1365, | |
| "step": 4025 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "grad_norm": 1.3689473867416382, | |
| "learning_rate": 2.1133333333333336e-06, | |
| "loss": 0.121, | |
| "step": 4050 | |
| }, | |
| { | |
| "epoch": 0.815, | |
| "grad_norm": 2.0017051696777344, | |
| "learning_rate": 2.057777777777778e-06, | |
| "loss": 0.1405, | |
| "step": 4075 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "grad_norm": 2.032273292541504, | |
| "learning_rate": 2.0022222222222225e-06, | |
| "loss": 0.1652, | |
| "step": 4100 | |
| }, | |
| { | |
| "epoch": 0.825, | |
| "grad_norm": 2.2192623615264893, | |
| "learning_rate": 1.9466666666666665e-06, | |
| "loss": 0.1721, | |
| "step": 4125 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "grad_norm": 2.077394723892212, | |
| "learning_rate": 1.8911111111111114e-06, | |
| "loss": 0.1687, | |
| "step": 4150 | |
| }, | |
| { | |
| "epoch": 0.835, | |
| "grad_norm": 1.6695953607559204, | |
| "learning_rate": 1.8355555555555557e-06, | |
| "loss": 0.1539, | |
| "step": 4175 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "grad_norm": 1.6516363620758057, | |
| "learning_rate": 1.7800000000000001e-06, | |
| "loss": 0.1506, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 0.845, | |
| "grad_norm": 1.743139386177063, | |
| "learning_rate": 1.7244444444444448e-06, | |
| "loss": 0.1628, | |
| "step": 4225 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "grad_norm": 1.993188500404358, | |
| "learning_rate": 1.668888888888889e-06, | |
| "loss": 0.1706, | |
| "step": 4250 | |
| }, | |
| { | |
| "epoch": 0.855, | |
| "grad_norm": 1.659005045890808, | |
| "learning_rate": 1.6133333333333335e-06, | |
| "loss": 0.1684, | |
| "step": 4275 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "grad_norm": 2.1015443801879883, | |
| "learning_rate": 1.5577777777777777e-06, | |
| "loss": 0.165, | |
| "step": 4300 | |
| }, | |
| { | |
| "epoch": 0.865, | |
| "grad_norm": 1.6033889055252075, | |
| "learning_rate": 1.5022222222222224e-06, | |
| "loss": 0.1458, | |
| "step": 4325 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "grad_norm": 1.5493004322052002, | |
| "learning_rate": 1.4466666666666669e-06, | |
| "loss": 0.1456, | |
| "step": 4350 | |
| }, | |
| { | |
| "epoch": 0.875, | |
| "grad_norm": 1.7360889911651611, | |
| "learning_rate": 1.3911111111111111e-06, | |
| "loss": 0.1351, | |
| "step": 4375 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "grad_norm": 1.69246244430542, | |
| "learning_rate": 1.3355555555555558e-06, | |
| "loss": 0.1433, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 0.885, | |
| "grad_norm": 2.408151149749756, | |
| "learning_rate": 1.28e-06, | |
| "loss": 0.1997, | |
| "step": 4425 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "grad_norm": 2.8492488861083984, | |
| "learning_rate": 1.2244444444444445e-06, | |
| "loss": 0.387, | |
| "step": 4450 | |
| }, | |
| { | |
| "epoch": 0.895, | |
| "grad_norm": 2.1923134326934814, | |
| "learning_rate": 1.168888888888889e-06, | |
| "loss": 0.3364, | |
| "step": 4475 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "grad_norm": 2.078341484069824, | |
| "learning_rate": 1.1133333333333334e-06, | |
| "loss": 0.2295, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 0.905, | |
| "grad_norm": 1.9317213296890259, | |
| "learning_rate": 1.0577777777777779e-06, | |
| "loss": 0.1771, | |
| "step": 4525 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "grad_norm": 1.9359657764434814, | |
| "learning_rate": 1.0022222222222223e-06, | |
| "loss": 0.1849, | |
| "step": 4550 | |
| }, | |
| { | |
| "epoch": 0.915, | |
| "grad_norm": 1.6038821935653687, | |
| "learning_rate": 9.466666666666667e-07, | |
| "loss": 0.1669, | |
| "step": 4575 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "grad_norm": 1.6796211004257202, | |
| "learning_rate": 8.911111111111112e-07, | |
| "loss": 0.1604, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 0.925, | |
| "grad_norm": 2.241725206375122, | |
| "learning_rate": 8.355555555555556e-07, | |
| "loss": 0.1614, | |
| "step": 4625 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "grad_norm": 1.6054015159606934, | |
| "learning_rate": 7.8e-07, | |
| "loss": 0.1528, | |
| "step": 4650 | |
| }, | |
| { | |
| "epoch": 0.935, | |
| "grad_norm": 1.6806738376617432, | |
| "learning_rate": 7.244444444444446e-07, | |
| "loss": 0.1255, | |
| "step": 4675 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "grad_norm": 1.7540109157562256, | |
| "learning_rate": 6.68888888888889e-07, | |
| "loss": 0.1163, | |
| "step": 4700 | |
| }, | |
| { | |
| "epoch": 0.945, | |
| "grad_norm": 1.2936944961547852, | |
| "learning_rate": 6.133333333333333e-07, | |
| "loss": 0.1199, | |
| "step": 4725 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "grad_norm": 1.535158634185791, | |
| "learning_rate": 5.577777777777779e-07, | |
| "loss": 0.1166, | |
| "step": 4750 | |
| }, | |
| { | |
| "epoch": 0.955, | |
| "grad_norm": 1.7007349729537964, | |
| "learning_rate": 5.022222222222222e-07, | |
| "loss": 0.1122, | |
| "step": 4775 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 1.4036020040512085, | |
| "learning_rate": 4.466666666666667e-07, | |
| "loss": 0.1155, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 0.965, | |
| "grad_norm": 1.7395439147949219, | |
| "learning_rate": 3.9111111111111115e-07, | |
| "loss": 0.1244, | |
| "step": 4825 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "grad_norm": 1.4144448041915894, | |
| "learning_rate": 3.3555555555555556e-07, | |
| "loss": 0.122, | |
| "step": 4850 | |
| }, | |
| { | |
| "epoch": 0.975, | |
| "grad_norm": 1.5096015930175781, | |
| "learning_rate": 2.8e-07, | |
| "loss": 0.1314, | |
| "step": 4875 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "grad_norm": 1.5810081958770752, | |
| "learning_rate": 2.2444444444444445e-07, | |
| "loss": 0.1249, | |
| "step": 4900 | |
| }, | |
| { | |
| "epoch": 0.985, | |
| "grad_norm": 1.706178903579712, | |
| "learning_rate": 1.6888888888888888e-07, | |
| "loss": 0.1295, | |
| "step": 4925 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "grad_norm": 1.7708213329315186, | |
| "learning_rate": 1.1333333333333336e-07, | |
| "loss": 0.1282, | |
| "step": 4950 | |
| }, | |
| { | |
| "epoch": 0.995, | |
| "grad_norm": 1.7594830989837646, | |
| "learning_rate": 5.777777777777778e-08, | |
| "loss": 0.1366, | |
| "step": 4975 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 2.096646308898926, | |
| "learning_rate": 2.2222222222222225e-09, | |
| "loss": 0.1286, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_loss": 0.19087658822536469, | |
| "eval_runtime": 819.4158, | |
| "eval_samples_per_second": 18.495, | |
| "eval_steps_per_second": 2.313, | |
| "eval_wer": 10.587474512857398, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "step": 5000, | |
| "total_flos": 9.23473281024e+19, | |
| "train_loss": 0.1977850432395935, | |
| "train_runtime": 19215.1319, | |
| "train_samples_per_second": 16.654, | |
| "train_steps_per_second": 0.26 | |
| } | |
| ], | |
| "logging_steps": 25, | |
| "max_steps": 5000, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 9223372036854775807, | |
| "save_steps": 1000, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 9.23473281024e+19, | |
| "train_batch_size": 64, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |