{ "best_metric": 20.044683596085466, "best_model_checkpoint": "all_lang_models/marathi_models/whisper-medium-mr_alldata_multigpu/checkpoint-8200", "epoch": 2.8472222222222223, "global_step": 8200, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.0, "learning_rate": 2.445014948199563e-06, "loss": 2.2897, "step": 10 }, { "epoch": 0.01, "learning_rate": 4.654519848766012e-06, "loss": 1.1234, "step": 20 }, { "epoch": 0.01, "learning_rate": 5.605150925598839e-06, "loss": 0.7141, "step": 30 }, { "epoch": 0.01, "learning_rate": 6.219461347633451e-06, "loss": 0.5167, "step": 40 }, { "epoch": 0.02, "learning_rate": 6.6741959581207635e-06, "loss": 0.4023, "step": 50 }, { "epoch": 0.02, "learning_rate": 7.035392983998264e-06, "loss": 0.3122, "step": 60 }, { "epoch": 0.02, "learning_rate": 7.3350448445986885e-06, "loss": 0.2807, "step": 70 }, { "epoch": 0.03, "learning_rate": 7.5911031495554354e-06, "loss": 0.2622, "step": 80 }, { "epoch": 0.03, "learning_rate": 7.814655826165288e-06, "loss": 0.2502, "step": 90 }, { "epoch": 0.03, "learning_rate": 8.013033860827016e-06, "loss": 0.2301, "step": 100 }, { "epoch": 0.03, "eval_loss": 0.304443359375, "eval_runtime": 284.7229, "eval_samples_per_second": 10.006, "eval_steps_per_second": 0.081, "eval_wer": 45.1272853142012, "step": 100 }, { "epoch": 0.04, "learning_rate": 8.19133763518252e-06, "loss": 0.2152, "step": 110 }, { "epoch": 0.04, "learning_rate": 8.353259112018159e-06, "loss": 0.209, "step": 120 }, { "epoch": 0.05, "learning_rate": 8.50155696580508e-06, "loss": 0.2003, "step": 130 }, { "epoch": 0.05, "learning_rate": 8.6383468479903e-06, "loss": 0.1952, "step": 140 }, { "epoch": 0.05, "learning_rate": 8.765286902998114e-06, "loss": 0.1897, "step": 150 }, { "epoch": 0.06, "learning_rate": 8.883700858687214e-06, "loss": 0.1869, "step": 160 }, { "epoch": 0.06, "learning_rate": 8.994662316723439e-06, "loss": 0.1825, "step": 170 }, { "epoch": 0.06, "learning_rate": 9.099054052977428e-06, "loss": 0.1779, "step": 180 }, { "epoch": 0.07, "learning_rate": 9.19761072313247e-06, "loss": 0.1767, "step": 190 }, { "epoch": 0.07, "learning_rate": 9.290950252080677e-06, "loss": 0.1658, "step": 200 }, { "epoch": 0.07, "eval_loss": 0.2354736328125, "eval_runtime": 1180.4962, "eval_samples_per_second": 2.413, "eval_steps_per_second": 0.019, "eval_wer": 37.77966581704899, "step": 200 }, { "epoch": 0.07, "learning_rate": 9.379597325032726e-06, "loss": 0.1638, "step": 210 }, { "epoch": 0.08, "learning_rate": 9.464001251372113e-06, "loss": 0.1654, "step": 220 }, { "epoch": 0.08, "learning_rate": 9.544549745165139e-06, "loss": 0.1603, "step": 230 }, { "epoch": 0.08, "learning_rate": 9.621579693581943e-06, "loss": 0.1589, "step": 240 }, { "epoch": 0.09, "learning_rate": 9.695385670292832e-06, "loss": 0.1557, "step": 250 }, { "epoch": 0.09, "learning_rate": 9.766226737805446e-06, "loss": 0.1523, "step": 260 }, { "epoch": 0.09, "learning_rate": 9.83433193552004e-06, "loss": 0.1576, "step": 270 }, { "epoch": 0.1, "learning_rate": 9.899904746919066e-06, "loss": 0.1504, "step": 280 }, { "epoch": 0.1, "learning_rate": 9.96312676561764e-06, "loss": 0.1468, "step": 290 }, { "epoch": 0.1, "learning_rate": 9.998955067920586e-06, "loss": 0.1492, "step": 300 }, { "epoch": 0.1, "eval_loss": 0.2061767578125, "eval_runtime": 514.1736, "eval_samples_per_second": 5.541, "eval_steps_per_second": 0.045, "eval_wer": 33.78646275842537, "step": 300 }, { "epoch": 0.11, "learning_rate": 9.995471960989204e-06, "loss": 0.1447, "step": 310 }, { "epoch": 0.11, "learning_rate": 9.99198885405782e-06, "loss": 0.1476, "step": 320 }, { "epoch": 0.11, "learning_rate": 9.988505747126437e-06, "loss": 0.142, "step": 330 }, { "epoch": 0.12, "learning_rate": 9.985022640195054e-06, "loss": 0.1433, "step": 340 }, { "epoch": 0.12, "learning_rate": 9.981539533263672e-06, "loss": 0.1379, "step": 350 }, { "epoch": 0.12, "learning_rate": 9.978056426332289e-06, "loss": 0.1418, "step": 360 }, { "epoch": 0.13, "learning_rate": 9.974573319400907e-06, "loss": 0.141, "step": 370 }, { "epoch": 0.13, "learning_rate": 9.971090212469524e-06, "loss": 0.1389, "step": 380 }, { "epoch": 0.14, "learning_rate": 9.96760710553814e-06, "loss": 0.1352, "step": 390 }, { "epoch": 0.14, "learning_rate": 9.964123998606759e-06, "loss": 0.139, "step": 400 }, { "epoch": 0.14, "eval_loss": 0.1884765625, "eval_runtime": 739.6853, "eval_samples_per_second": 3.852, "eval_steps_per_second": 0.031, "eval_wer": 32.59070455332137, "step": 400 }, { "epoch": 0.14, "learning_rate": 9.960640891675375e-06, "loss": 0.132, "step": 410 }, { "epoch": 0.15, "learning_rate": 9.957157784743993e-06, "loss": 0.1298, "step": 420 }, { "epoch": 0.15, "learning_rate": 9.95367467781261e-06, "loss": 0.1269, "step": 430 }, { "epoch": 0.15, "learning_rate": 9.950191570881227e-06, "loss": 0.1242, "step": 440 }, { "epoch": 0.16, "learning_rate": 9.946708463949845e-06, "loss": 0.1283, "step": 450 }, { "epoch": 0.16, "learning_rate": 9.943225357018461e-06, "loss": 0.131, "step": 460 }, { "epoch": 0.16, "learning_rate": 9.939742250087078e-06, "loss": 0.1232, "step": 470 }, { "epoch": 0.17, "learning_rate": 9.936259143155695e-06, "loss": 0.1228, "step": 480 }, { "epoch": 0.17, "learning_rate": 9.932776036224313e-06, "loss": 0.1249, "step": 490 }, { "epoch": 0.17, "learning_rate": 9.92929292929293e-06, "loss": 0.1238, "step": 500 }, { "epoch": 0.17, "eval_loss": 0.177001953125, "eval_runtime": 759.5608, "eval_samples_per_second": 3.751, "eval_steps_per_second": 0.03, "eval_wer": 30.28729664243683, "step": 500 }, { "epoch": 0.18, "learning_rate": 9.925809822361548e-06, "loss": 0.1245, "step": 510 }, { "epoch": 0.18, "learning_rate": 9.922326715430164e-06, "loss": 0.1251, "step": 520 }, { "epoch": 0.18, "learning_rate": 9.918843608498781e-06, "loss": 0.1222, "step": 530 }, { "epoch": 0.19, "learning_rate": 9.915360501567399e-06, "loss": 0.1218, "step": 540 }, { "epoch": 0.19, "learning_rate": 9.911877394636016e-06, "loss": 0.1235, "step": 550 }, { "epoch": 0.19, "learning_rate": 9.908394287704634e-06, "loss": 0.1222, "step": 560 }, { "epoch": 0.2, "learning_rate": 9.90491118077325e-06, "loss": 0.1217, "step": 570 }, { "epoch": 0.2, "learning_rate": 9.901428073841867e-06, "loss": 0.121, "step": 580 }, { "epoch": 0.2, "learning_rate": 9.897944966910485e-06, "loss": 0.1178, "step": 590 }, { "epoch": 0.21, "learning_rate": 9.894461859979102e-06, "loss": 0.1214, "step": 600 }, { "epoch": 0.21, "eval_loss": 0.1671142578125, "eval_runtime": 424.7805, "eval_samples_per_second": 6.707, "eval_steps_per_second": 0.054, "eval_wer": 29.554108058780955, "step": 600 }, { "epoch": 0.21, "learning_rate": 9.89097875304772e-06, "loss": 0.1203, "step": 610 }, { "epoch": 0.22, "learning_rate": 9.887495646116337e-06, "loss": 0.1154, "step": 620 }, { "epoch": 0.22, "learning_rate": 9.884012539184953e-06, "loss": 0.1181, "step": 630 }, { "epoch": 0.22, "learning_rate": 9.88052943225357e-06, "loss": 0.1124, "step": 640 }, { "epoch": 0.23, "learning_rate": 9.877046325322188e-06, "loss": 0.1183, "step": 650 }, { "epoch": 0.23, "learning_rate": 9.873563218390807e-06, "loss": 0.1114, "step": 660 }, { "epoch": 0.23, "learning_rate": 9.870080111459423e-06, "loss": 0.1125, "step": 670 }, { "epoch": 0.24, "learning_rate": 9.86659700452804e-06, "loss": 0.1138, "step": 680 }, { "epoch": 0.24, "learning_rate": 9.863113897596656e-06, "loss": 0.1127, "step": 690 }, { "epoch": 0.24, "learning_rate": 9.859630790665275e-06, "loss": 0.11, "step": 700 }, { "epoch": 0.24, "eval_loss": 0.159423828125, "eval_runtime": 741.6949, "eval_samples_per_second": 3.841, "eval_steps_per_second": 0.031, "eval_wer": 28.60064822681645, "step": 700 }, { "epoch": 0.25, "learning_rate": 9.856147683733891e-06, "loss": 0.1093, "step": 710 }, { "epoch": 0.25, "learning_rate": 9.85266457680251e-06, "loss": 0.1184, "step": 720 }, { "epoch": 0.25, "learning_rate": 9.849181469871126e-06, "loss": 0.1098, "step": 730 }, { "epoch": 0.26, "learning_rate": 9.845698362939743e-06, "loss": 0.1111, "step": 740 }, { "epoch": 0.26, "learning_rate": 9.842215256008361e-06, "loss": 0.1048, "step": 750 }, { "epoch": 0.26, "learning_rate": 9.838732149076977e-06, "loss": 0.1092, "step": 760 }, { "epoch": 0.27, "learning_rate": 9.835249042145594e-06, "loss": 0.107, "step": 770 }, { "epoch": 0.27, "learning_rate": 9.83176593521421e-06, "loss": 0.1136, "step": 780 }, { "epoch": 0.27, "learning_rate": 9.828282828282829e-06, "loss": 0.1092, "step": 790 }, { "epoch": 0.28, "learning_rate": 9.824799721351447e-06, "loss": 0.1116, "step": 800 }, { "epoch": 0.28, "eval_loss": 0.1539306640625, "eval_runtime": 290.1579, "eval_samples_per_second": 9.819, "eval_steps_per_second": 0.079, "eval_wer": 28.02794298121401, "step": 800 }, { "epoch": 0.28, "learning_rate": 9.821316614420064e-06, "loss": 0.1067, "step": 810 }, { "epoch": 0.28, "learning_rate": 9.81783350748868e-06, "loss": 0.1072, "step": 820 }, { "epoch": 0.29, "learning_rate": 9.814350400557297e-06, "loss": 0.1105, "step": 830 }, { "epoch": 0.29, "learning_rate": 9.810867293625915e-06, "loss": 0.108, "step": 840 }, { "epoch": 0.3, "learning_rate": 9.807384186694532e-06, "loss": 0.1042, "step": 850 }, { "epoch": 0.3, "learning_rate": 9.80390107976315e-06, "loss": 0.1092, "step": 860 }, { "epoch": 0.3, "learning_rate": 9.800417972831767e-06, "loss": 0.105, "step": 870 }, { "epoch": 0.31, "learning_rate": 9.796934865900383e-06, "loss": 0.1067, "step": 880 }, { "epoch": 0.31, "learning_rate": 9.793451758969001e-06, "loss": 0.1066, "step": 890 }, { "epoch": 0.31, "learning_rate": 9.789968652037618e-06, "loss": 0.104, "step": 900 }, { "epoch": 0.31, "eval_loss": 0.150634765625, "eval_runtime": 1070.3526, "eval_samples_per_second": 2.662, "eval_steps_per_second": 0.021, "eval_wer": 27.301047861795524, "step": 900 }, { "epoch": 0.32, "learning_rate": 9.786485545106236e-06, "loss": 0.1086, "step": 910 }, { "epoch": 0.32, "learning_rate": 9.783002438174853e-06, "loss": 0.1033, "step": 920 }, { "epoch": 0.32, "learning_rate": 9.77951933124347e-06, "loss": 0.1054, "step": 930 }, { "epoch": 0.33, "learning_rate": 9.776036224312088e-06, "loss": 0.1089, "step": 940 }, { "epoch": 0.33, "learning_rate": 9.772553117380704e-06, "loss": 0.1043, "step": 950 }, { "epoch": 0.33, "learning_rate": 9.769070010449323e-06, "loss": 0.1046, "step": 960 }, { "epoch": 0.34, "learning_rate": 9.76558690351794e-06, "loss": 0.1, "step": 970 }, { "epoch": 0.34, "learning_rate": 9.762103796586556e-06, "loss": 0.102, "step": 980 }, { "epoch": 0.34, "learning_rate": 9.758620689655172e-06, "loss": 0.1049, "step": 990 }, { "epoch": 0.35, "learning_rate": 9.75513758272379e-06, "loss": 0.1023, "step": 1000 }, { "epoch": 0.35, "eval_loss": 0.146728515625, "eval_runtime": 692.4983, "eval_samples_per_second": 4.114, "eval_steps_per_second": 0.033, "eval_wer": 26.86679882941565, "step": 1000 }, { "epoch": 0.35, "learning_rate": 9.751654475792409e-06, "loss": 0.1022, "step": 1010 }, { "epoch": 0.35, "learning_rate": 9.748171368861024e-06, "loss": 0.0972, "step": 1020 }, { "epoch": 0.36, "learning_rate": 9.744688261929642e-06, "loss": 0.1008, "step": 1030 }, { "epoch": 0.36, "learning_rate": 9.741205154998259e-06, "loss": 0.1059, "step": 1040 }, { "epoch": 0.36, "learning_rate": 9.737722048066877e-06, "loss": 0.1021, "step": 1050 }, { "epoch": 0.37, "learning_rate": 9.734238941135494e-06, "loss": 0.1034, "step": 1060 }, { "epoch": 0.37, "learning_rate": 9.73075583420411e-06, "loss": 0.1015, "step": 1070 }, { "epoch": 0.38, "learning_rate": 9.727272727272728e-06, "loss": 0.1003, "step": 1080 }, { "epoch": 0.38, "learning_rate": 9.723789620341345e-06, "loss": 0.1025, "step": 1090 }, { "epoch": 0.38, "learning_rate": 9.720306513409963e-06, "loss": 0.0972, "step": 1100 }, { "epoch": 0.38, "eval_loss": 0.1441650390625, "eval_runtime": 527.1217, "eval_samples_per_second": 5.405, "eval_steps_per_second": 0.044, "eval_wer": 25.756002391516407, "step": 1100 }, { "epoch": 0.39, "learning_rate": 9.71682340647858e-06, "loss": 0.1017, "step": 1110 }, { "epoch": 0.39, "learning_rate": 9.713340299547196e-06, "loss": 0.1021, "step": 1120 }, { "epoch": 0.39, "learning_rate": 9.709857192615813e-06, "loss": 0.1011, "step": 1130 }, { "epoch": 0.4, "learning_rate": 9.706374085684431e-06, "loss": 0.0979, "step": 1140 }, { "epoch": 0.4, "learning_rate": 9.70289097875305e-06, "loss": 0.0998, "step": 1150 }, { "epoch": 0.4, "learning_rate": 9.699407871821666e-06, "loss": 0.0966, "step": 1160 }, { "epoch": 0.41, "learning_rate": 9.695924764890283e-06, "loss": 0.099, "step": 1170 }, { "epoch": 0.41, "learning_rate": 9.6924416579589e-06, "loss": 0.0984, "step": 1180 }, { "epoch": 0.41, "learning_rate": 9.688958551027518e-06, "loss": 0.0986, "step": 1190 }, { "epoch": 0.42, "learning_rate": 9.685475444096134e-06, "loss": 0.0972, "step": 1200 }, { "epoch": 0.42, "eval_loss": 0.1383056640625, "eval_runtime": 777.6072, "eval_samples_per_second": 3.664, "eval_steps_per_second": 0.03, "eval_wer": 25.350073948204788, "step": 1200 }, { "epoch": 0.42, "learning_rate": 9.681992337164752e-06, "loss": 0.0975, "step": 1210 }, { "epoch": 0.42, "learning_rate": 9.678509230233369e-06, "loss": 0.0978, "step": 1220 }, { "epoch": 0.43, "learning_rate": 9.675026123301986e-06, "loss": 0.0996, "step": 1230 }, { "epoch": 0.43, "learning_rate": 9.671543016370604e-06, "loss": 0.0978, "step": 1240 }, { "epoch": 0.43, "learning_rate": 9.66805990943922e-06, "loss": 0.1, "step": 1250 }, { "epoch": 0.44, "learning_rate": 9.664576802507839e-06, "loss": 0.0924, "step": 1260 }, { "epoch": 0.44, "learning_rate": 9.661093695576455e-06, "loss": 0.0998, "step": 1270 }, { "epoch": 0.44, "learning_rate": 9.657610588645072e-06, "loss": 0.0961, "step": 1280 }, { "epoch": 0.45, "learning_rate": 9.65412748171369e-06, "loss": 0.0979, "step": 1290 }, { "epoch": 0.45, "learning_rate": 9.650644374782307e-06, "loss": 0.0977, "step": 1300 }, { "epoch": 0.45, "eval_loss": 0.13623046875, "eval_runtime": 782.2381, "eval_samples_per_second": 3.642, "eval_steps_per_second": 0.029, "eval_wer": 25.43188898329085, "step": 1300 }, { "epoch": 0.45, "learning_rate": 9.647161267850923e-06, "loss": 0.0979, "step": 1310 }, { "epoch": 0.46, "learning_rate": 9.64367816091954e-06, "loss": 0.0976, "step": 1320 }, { "epoch": 0.46, "learning_rate": 9.640195053988158e-06, "loss": 0.0981, "step": 1330 }, { "epoch": 0.47, "learning_rate": 9.636711947056775e-06, "loss": 0.0955, "step": 1340 }, { "epoch": 0.47, "learning_rate": 9.633228840125393e-06, "loss": 0.096, "step": 1350 }, { "epoch": 0.47, "learning_rate": 9.62974573319401e-06, "loss": 0.0944, "step": 1360 }, { "epoch": 0.48, "learning_rate": 9.626262626262626e-06, "loss": 0.0968, "step": 1370 }, { "epoch": 0.48, "learning_rate": 9.622779519331244e-06, "loss": 0.0917, "step": 1380 }, { "epoch": 0.48, "learning_rate": 9.619296412399861e-06, "loss": 0.0977, "step": 1390 }, { "epoch": 0.49, "learning_rate": 9.61581330546848e-06, "loss": 0.0934, "step": 1400 }, { "epoch": 0.49, "eval_loss": 0.1353759765625, "eval_runtime": 476.9686, "eval_samples_per_second": 5.973, "eval_steps_per_second": 0.048, "eval_wer": 25.249378520406555, "step": 1400 }, { "epoch": 0.49, "learning_rate": 9.612330198537096e-06, "loss": 0.0918, "step": 1410 }, { "epoch": 0.49, "learning_rate": 9.608847091605712e-06, "loss": 0.097, "step": 1420 }, { "epoch": 0.5, "learning_rate": 9.60536398467433e-06, "loss": 0.0933, "step": 1430 }, { "epoch": 0.5, "learning_rate": 9.601880877742947e-06, "loss": 0.0944, "step": 1440 }, { "epoch": 0.5, "learning_rate": 9.598397770811566e-06, "loss": 0.0957, "step": 1450 }, { "epoch": 0.51, "learning_rate": 9.594914663880182e-06, "loss": 0.0948, "step": 1460 }, { "epoch": 0.51, "learning_rate": 9.591431556948799e-06, "loss": 0.0955, "step": 1470 }, { "epoch": 0.51, "learning_rate": 9.587948450017415e-06, "loss": 0.0967, "step": 1480 }, { "epoch": 0.52, "learning_rate": 9.584465343086034e-06, "loss": 0.0935, "step": 1490 }, { "epoch": 0.52, "learning_rate": 9.580982236154652e-06, "loss": 0.0965, "step": 1500 }, { "epoch": 0.52, "eval_loss": 0.1326904296875, "eval_runtime": 295.4901, "eval_samples_per_second": 9.642, "eval_steps_per_second": 0.078, "eval_wer": 25.243085056169168, "step": 1500 }, { "epoch": 0.52, "learning_rate": 9.577499129223268e-06, "loss": 0.097, "step": 1510 }, { "epoch": 0.53, "learning_rate": 9.574016022291885e-06, "loss": 0.0891, "step": 1520 }, { "epoch": 0.53, "learning_rate": 9.570532915360502e-06, "loss": 0.0953, "step": 1530 }, { "epoch": 0.53, "learning_rate": 9.56704980842912e-06, "loss": 0.0948, "step": 1540 }, { "epoch": 0.54, "learning_rate": 9.563566701497736e-06, "loss": 0.0949, "step": 1550 }, { "epoch": 0.54, "learning_rate": 9.560083594566355e-06, "loss": 0.0911, "step": 1560 }, { "epoch": 0.55, "learning_rate": 9.556600487634971e-06, "loss": 0.0901, "step": 1570 }, { "epoch": 0.55, "learning_rate": 9.553117380703588e-06, "loss": 0.0953, "step": 1580 }, { "epoch": 0.55, "learning_rate": 9.549634273772206e-06, "loss": 0.0932, "step": 1590 }, { "epoch": 0.56, "learning_rate": 9.546151166840823e-06, "loss": 0.0937, "step": 1600 }, { "epoch": 0.56, "eval_loss": 0.130126953125, "eval_runtime": 803.927, "eval_samples_per_second": 3.544, "eval_steps_per_second": 0.029, "eval_wer": 24.390320652002895, "step": 1600 }, { "epoch": 0.56, "learning_rate": 9.54266805990944e-06, "loss": 0.0912, "step": 1610 }, { "epoch": 0.56, "learning_rate": 9.539184952978056e-06, "loss": 0.092, "step": 1620 }, { "epoch": 0.57, "learning_rate": 9.535701846046674e-06, "loss": 0.0923, "step": 1630 }, { "epoch": 0.57, "learning_rate": 9.532218739115292e-06, "loss": 0.0886, "step": 1640 }, { "epoch": 0.57, "learning_rate": 9.528735632183909e-06, "loss": 0.0891, "step": 1650 }, { "epoch": 0.58, "learning_rate": 9.525252525252526e-06, "loss": 0.0864, "step": 1660 }, { "epoch": 0.58, "learning_rate": 9.521769418321142e-06, "loss": 0.0898, "step": 1670 }, { "epoch": 0.58, "learning_rate": 9.51828631138976e-06, "loss": 0.0913, "step": 1680 }, { "epoch": 0.59, "learning_rate": 9.514803204458377e-06, "loss": 0.089, "step": 1690 }, { "epoch": 0.59, "learning_rate": 9.511320097526995e-06, "loss": 0.0926, "step": 1700 }, { "epoch": 0.59, "eval_loss": 0.12744140625, "eval_runtime": 779.0613, "eval_samples_per_second": 3.657, "eval_steps_per_second": 0.03, "eval_wer": 24.30221215267944, "step": 1700 }, { "epoch": 0.59, "learning_rate": 9.507836990595612e-06, "loss": 0.0934, "step": 1710 }, { "epoch": 0.6, "learning_rate": 9.504353883664229e-06, "loss": 0.0866, "step": 1720 }, { "epoch": 0.6, "learning_rate": 9.500870776732847e-06, "loss": 0.0901, "step": 1730 }, { "epoch": 0.6, "learning_rate": 9.497387669801463e-06, "loss": 0.0879, "step": 1740 }, { "epoch": 0.61, "learning_rate": 9.493904562870082e-06, "loss": 0.0925, "step": 1750 }, { "epoch": 0.61, "learning_rate": 9.490421455938698e-06, "loss": 0.0883, "step": 1760 }, { "epoch": 0.61, "learning_rate": 9.486938349007315e-06, "loss": 0.0904, "step": 1770 }, { "epoch": 0.62, "learning_rate": 9.483455242075933e-06, "loss": 0.0895, "step": 1780 }, { "epoch": 0.62, "learning_rate": 9.47997213514455e-06, "loss": 0.0944, "step": 1790 }, { "epoch": 0.62, "learning_rate": 9.476489028213168e-06, "loss": 0.0951, "step": 1800 }, { "epoch": 0.62, "eval_loss": 0.1268310546875, "eval_runtime": 625.0326, "eval_samples_per_second": 4.558, "eval_steps_per_second": 0.037, "eval_wer": 24.51933666886938, "step": 1800 }, { "epoch": 0.63, "learning_rate": 9.473005921281785e-06, "loss": 0.0887, "step": 1810 }, { "epoch": 0.63, "learning_rate": 9.469522814350401e-06, "loss": 0.0914, "step": 1820 }, { "epoch": 0.64, "learning_rate": 9.466039707419018e-06, "loss": 0.0906, "step": 1830 }, { "epoch": 0.64, "learning_rate": 9.462556600487636e-06, "loss": 0.0908, "step": 1840 }, { "epoch": 0.64, "learning_rate": 9.459073493556253e-06, "loss": 0.0901, "step": 1850 }, { "epoch": 0.65, "learning_rate": 9.455590386624869e-06, "loss": 0.088, "step": 1860 }, { "epoch": 0.65, "learning_rate": 9.452107279693487e-06, "loss": 0.0825, "step": 1870 }, { "epoch": 0.65, "learning_rate": 9.448624172762104e-06, "loss": 0.0909, "step": 1880 }, { "epoch": 0.66, "learning_rate": 9.445141065830722e-06, "loss": 0.0896, "step": 1890 }, { "epoch": 0.66, "learning_rate": 9.441657958899339e-06, "loss": 0.0922, "step": 1900 }, { "epoch": 0.66, "eval_loss": 0.1253662109375, "eval_runtime": 548.6186, "eval_samples_per_second": 5.193, "eval_steps_per_second": 0.042, "eval_wer": 23.770414424620032, "step": 1900 }, { "epoch": 0.66, "learning_rate": 9.438174851967955e-06, "loss": 0.0845, "step": 1910 }, { "epoch": 0.67, "learning_rate": 9.434691745036574e-06, "loss": 0.0843, "step": 1920 }, { "epoch": 0.67, "learning_rate": 9.43120863810519e-06, "loss": 0.0826, "step": 1930 }, { "epoch": 0.67, "learning_rate": 9.427725531173809e-06, "loss": 0.0879, "step": 1940 }, { "epoch": 0.68, "learning_rate": 9.424242424242425e-06, "loss": 0.0879, "step": 1950 }, { "epoch": 0.68, "learning_rate": 9.420759317311042e-06, "loss": 0.0885, "step": 1960 }, { "epoch": 0.68, "learning_rate": 9.417276210379658e-06, "loss": 0.09, "step": 1970 }, { "epoch": 0.69, "learning_rate": 9.413793103448277e-06, "loss": 0.0898, "step": 1980 }, { "epoch": 0.69, "learning_rate": 9.410309996516895e-06, "loss": 0.0884, "step": 1990 }, { "epoch": 0.69, "learning_rate": 9.406826889585511e-06, "loss": 0.0881, "step": 2000 }, { "epoch": 0.69, "eval_loss": 0.1258544921875, "eval_runtime": 767.5109, "eval_samples_per_second": 3.712, "eval_steps_per_second": 0.03, "eval_wer": 23.386513106139276, "step": 2000 }, { "epoch": 0.7, "learning_rate": 9.403343782654128e-06, "loss": 0.0888, "step": 2010 }, { "epoch": 0.7, "learning_rate": 9.399860675722745e-06, "loss": 0.0862, "step": 2020 }, { "epoch": 0.7, "learning_rate": 9.396377568791363e-06, "loss": 0.0918, "step": 2030 }, { "epoch": 0.71, "learning_rate": 9.39289446185998e-06, "loss": 0.0899, "step": 2040 }, { "epoch": 0.71, "learning_rate": 9.389411354928598e-06, "loss": 0.0854, "step": 2050 }, { "epoch": 0.72, "learning_rate": 9.385928247997214e-06, "loss": 0.0858, "step": 2060 }, { "epoch": 0.72, "learning_rate": 9.382445141065831e-06, "loss": 0.0898, "step": 2070 }, { "epoch": 0.72, "learning_rate": 9.378962034134449e-06, "loss": 0.0873, "step": 2080 }, { "epoch": 0.73, "learning_rate": 9.375478927203066e-06, "loss": 0.087, "step": 2090 }, { "epoch": 0.73, "learning_rate": 9.371995820271684e-06, "loss": 0.0867, "step": 2100 }, { "epoch": 0.73, "eval_loss": 0.12060546875, "eval_runtime": 595.8484, "eval_samples_per_second": 4.781, "eval_steps_per_second": 0.039, "eval_wer": 23.399100034614055, "step": 2100 }, { "epoch": 0.73, "learning_rate": 9.3685127133403e-06, "loss": 0.0865, "step": 2110 }, { "epoch": 0.74, "learning_rate": 9.365029606408917e-06, "loss": 0.0868, "step": 2120 }, { "epoch": 0.74, "learning_rate": 9.361546499477535e-06, "loss": 0.0851, "step": 2130 }, { "epoch": 0.74, "learning_rate": 9.358063392546152e-06, "loss": 0.0825, "step": 2140 }, { "epoch": 0.75, "learning_rate": 9.354580285614769e-06, "loss": 0.0871, "step": 2150 }, { "epoch": 0.75, "learning_rate": 9.351097178683385e-06, "loss": 0.0876, "step": 2160 }, { "epoch": 0.75, "learning_rate": 9.347614071752003e-06, "loss": 0.0842, "step": 2170 }, { "epoch": 0.76, "learning_rate": 9.34413096482062e-06, "loss": 0.0854, "step": 2180 }, { "epoch": 0.76, "learning_rate": 9.340647857889238e-06, "loss": 0.0883, "step": 2190 }, { "epoch": 0.76, "learning_rate": 9.337164750957855e-06, "loss": 0.0818, "step": 2200 }, { "epoch": 0.76, "eval_loss": 0.1219482421875, "eval_runtime": 360.525, "eval_samples_per_second": 7.902, "eval_steps_per_second": 0.064, "eval_wer": 23.2512036250354, "step": 2200 }, { "epoch": 0.77, "learning_rate": 9.333681644026471e-06, "loss": 0.0826, "step": 2210 }, { "epoch": 0.77, "learning_rate": 9.33019853709509e-06, "loss": 0.0823, "step": 2220 }, { "epoch": 0.77, "learning_rate": 9.326715430163706e-06, "loss": 0.0845, "step": 2230 }, { "epoch": 0.78, "learning_rate": 9.323232323232325e-06, "loss": 0.0814, "step": 2240 }, { "epoch": 0.78, "learning_rate": 9.319749216300941e-06, "loss": 0.088, "step": 2250 }, { "epoch": 0.78, "learning_rate": 9.316266109369558e-06, "loss": 0.0844, "step": 2260 }, { "epoch": 0.79, "learning_rate": 9.312783002438176e-06, "loss": 0.0867, "step": 2270 }, { "epoch": 0.79, "learning_rate": 9.309299895506793e-06, "loss": 0.083, "step": 2280 }, { "epoch": 0.8, "learning_rate": 9.305816788575411e-06, "loss": 0.0814, "step": 2290 }, { "epoch": 0.8, "learning_rate": 9.302333681644027e-06, "loss": 0.0857, "step": 2300 }, { "epoch": 0.8, "eval_loss": 0.12091064453125, "eval_runtime": 790.1779, "eval_samples_per_second": 3.606, "eval_steps_per_second": 0.029, "eval_wer": 23.18512225054281, "step": 2300 }, { "epoch": 0.8, "learning_rate": 9.298850574712644e-06, "loss": 0.0849, "step": 2310 }, { "epoch": 0.81, "learning_rate": 9.295367467781262e-06, "loss": 0.0821, "step": 2320 }, { "epoch": 0.81, "learning_rate": 9.291884360849879e-06, "loss": 0.0845, "step": 2330 }, { "epoch": 0.81, "learning_rate": 9.288401253918497e-06, "loss": 0.0839, "step": 2340 }, { "epoch": 0.82, "learning_rate": 9.284918146987114e-06, "loss": 0.0839, "step": 2350 }, { "epoch": 0.82, "learning_rate": 9.28143504005573e-06, "loss": 0.0827, "step": 2360 }, { "epoch": 0.82, "learning_rate": 9.277951933124347e-06, "loss": 0.0835, "step": 2370 }, { "epoch": 0.83, "learning_rate": 9.274468826192965e-06, "loss": 0.0851, "step": 2380 }, { "epoch": 0.83, "learning_rate": 9.270985719261583e-06, "loss": 0.0833, "step": 2390 }, { "epoch": 0.83, "learning_rate": 9.267502612330198e-06, "loss": 0.0856, "step": 2400 }, { "epoch": 0.83, "eval_loss": 0.1209716796875, "eval_runtime": 462.8014, "eval_samples_per_second": 6.156, "eval_steps_per_second": 0.05, "eval_wer": 23.074986626388498, "step": 2400 }, { "epoch": 0.84, "learning_rate": 9.264019505398817e-06, "loss": 0.0829, "step": 2410 }, { "epoch": 0.84, "learning_rate": 9.260536398467433e-06, "loss": 0.0809, "step": 2420 }, { "epoch": 0.84, "learning_rate": 9.257053291536052e-06, "loss": 0.0838, "step": 2430 }, { "epoch": 0.85, "learning_rate": 9.253570184604668e-06, "loss": 0.0845, "step": 2440 }, { "epoch": 0.85, "learning_rate": 9.250087077673285e-06, "loss": 0.0813, "step": 2450 }, { "epoch": 0.85, "learning_rate": 9.246603970741903e-06, "loss": 0.0866, "step": 2460 }, { "epoch": 0.86, "learning_rate": 9.24312086381052e-06, "loss": 0.0851, "step": 2470 }, { "epoch": 0.86, "learning_rate": 9.239637756879138e-06, "loss": 0.0822, "step": 2480 }, { "epoch": 0.86, "learning_rate": 9.236154649947754e-06, "loss": 0.0789, "step": 2490 }, { "epoch": 0.87, "learning_rate": 9.232671543016371e-06, "loss": 0.0812, "step": 2500 }, { "epoch": 0.87, "eval_loss": 0.1201171875, "eval_runtime": 682.3065, "eval_samples_per_second": 4.176, "eval_steps_per_second": 0.034, "eval_wer": 23.134774536643697, "step": 2500 }, { "epoch": 0.87, "learning_rate": 9.229188436084988e-06, "loss": 0.0859, "step": 2510 }, { "epoch": 0.88, "learning_rate": 9.225705329153606e-06, "loss": 0.0801, "step": 2520 }, { "epoch": 0.88, "learning_rate": 9.222222222222224e-06, "loss": 0.0819, "step": 2530 }, { "epoch": 0.88, "learning_rate": 9.21873911529084e-06, "loss": 0.0807, "step": 2540 }, { "epoch": 0.89, "learning_rate": 9.215256008359457e-06, "loss": 0.085, "step": 2550 }, { "epoch": 0.89, "learning_rate": 9.211772901428074e-06, "loss": 0.0822, "step": 2560 }, { "epoch": 0.89, "learning_rate": 9.208289794496692e-06, "loss": 0.0837, "step": 2570 }, { "epoch": 0.9, "learning_rate": 9.204806687565309e-06, "loss": 0.0865, "step": 2580 }, { "epoch": 0.9, "learning_rate": 9.201323580633927e-06, "loss": 0.0851, "step": 2590 }, { "epoch": 0.9, "learning_rate": 9.197840473702544e-06, "loss": 0.0825, "step": 2600 }, { "epoch": 0.9, "eval_loss": 0.11773681640625, "eval_runtime": 780.5221, "eval_samples_per_second": 3.65, "eval_steps_per_second": 0.029, "eval_wer": 22.73828628968816, "step": 2600 }, { "epoch": 0.91, "learning_rate": 9.19435736677116e-06, "loss": 0.0857, "step": 2610 }, { "epoch": 0.91, "learning_rate": 9.190874259839778e-06, "loss": 0.0839, "step": 2620 }, { "epoch": 0.91, "learning_rate": 9.187391152908395e-06, "loss": 0.0814, "step": 2630 }, { "epoch": 0.92, "learning_rate": 9.183908045977013e-06, "loss": 0.0841, "step": 2640 }, { "epoch": 0.92, "learning_rate": 9.18042493904563e-06, "loss": 0.08, "step": 2650 }, { "epoch": 0.92, "learning_rate": 9.176941832114246e-06, "loss": 0.0842, "step": 2660 }, { "epoch": 0.93, "learning_rate": 9.173458725182865e-06, "loss": 0.0826, "step": 2670 }, { "epoch": 0.93, "learning_rate": 9.169975618251481e-06, "loss": 0.082, "step": 2680 }, { "epoch": 0.93, "learning_rate": 9.166492511320098e-06, "loss": 0.0838, "step": 2690 }, { "epoch": 0.94, "learning_rate": 9.163009404388714e-06, "loss": 0.0824, "step": 2700 }, { "epoch": 0.94, "eval_loss": 0.11480712890625, "eval_runtime": 515.5719, "eval_samples_per_second": 5.526, "eval_steps_per_second": 0.045, "eval_wer": 22.555775826803863, "step": 2700 }, { "epoch": 0.94, "learning_rate": 9.159526297457333e-06, "loss": 0.0819, "step": 2710 }, { "epoch": 0.94, "learning_rate": 9.15604319052595e-06, "loss": 0.08, "step": 2720 }, { "epoch": 0.95, "learning_rate": 9.152560083594568e-06, "loss": 0.0785, "step": 2730 }, { "epoch": 0.95, "learning_rate": 9.149076976663184e-06, "loss": 0.081, "step": 2740 }, { "epoch": 0.95, "learning_rate": 9.1455938697318e-06, "loss": 0.0728, "step": 2750 }, { "epoch": 0.96, "learning_rate": 9.142110762800419e-06, "loss": 0.0836, "step": 2760 }, { "epoch": 0.96, "learning_rate": 9.138627655869036e-06, "loss": 0.0779, "step": 2770 }, { "epoch": 0.97, "learning_rate": 9.135144548937654e-06, "loss": 0.0773, "step": 2780 }, { "epoch": 0.97, "learning_rate": 9.13166144200627e-06, "loss": 0.0816, "step": 2790 }, { "epoch": 0.97, "learning_rate": 9.128178335074887e-06, "loss": 0.0801, "step": 2800 }, { "epoch": 0.97, "eval_loss": 0.1163330078125, "eval_runtime": 770.6819, "eval_samples_per_second": 3.697, "eval_steps_per_second": 0.03, "eval_wer": 22.228515686459613, "step": 2800 }, { "epoch": 0.98, "learning_rate": 9.124695228143505e-06, "loss": 0.0821, "step": 2810 }, { "epoch": 0.98, "learning_rate": 9.121212121212122e-06, "loss": 0.0825, "step": 2820 }, { "epoch": 0.98, "learning_rate": 9.11772901428074e-06, "loss": 0.0816, "step": 2830 }, { "epoch": 0.99, "learning_rate": 9.114245907349357e-06, "loss": 0.0789, "step": 2840 }, { "epoch": 0.99, "learning_rate": 9.110762800417973e-06, "loss": 0.0835, "step": 2850 }, { "epoch": 0.99, "learning_rate": 9.10727969348659e-06, "loss": 0.0807, "step": 2860 }, { "epoch": 1.0, "learning_rate": 9.103796586555208e-06, "loss": 0.083, "step": 2870 }, { "epoch": 1.0, "learning_rate": 9.100313479623826e-06, "loss": 0.0773, "step": 2880 }, { "epoch": 1.0, "learning_rate": 9.096830372692443e-06, "loss": 0.0709, "step": 2890 }, { "epoch": 1.01, "learning_rate": 9.09334726576106e-06, "loss": 0.0753, "step": 2900 }, { "epoch": 1.01, "eval_loss": 0.1170654296875, "eval_runtime": 291.6827, "eval_samples_per_second": 9.767, "eval_steps_per_second": 0.079, "eval_wer": 22.228515686459613, "step": 2900 }, { "epoch": 1.01, "learning_rate": 9.089864158829676e-06, "loss": 0.0767, "step": 2910 }, { "epoch": 1.01, "learning_rate": 9.086381051898294e-06, "loss": 0.0717, "step": 2920 }, { "epoch": 1.02, "learning_rate": 9.082897944966911e-06, "loss": 0.0736, "step": 2930 }, { "epoch": 1.02, "learning_rate": 9.07941483803553e-06, "loss": 0.0724, "step": 2940 }, { "epoch": 1.02, "learning_rate": 9.075931731104146e-06, "loss": 0.0726, "step": 2950 }, { "epoch": 1.03, "learning_rate": 9.072448624172762e-06, "loss": 0.0723, "step": 2960 }, { "epoch": 1.03, "learning_rate": 9.06896551724138e-06, "loss": 0.0735, "step": 2970 }, { "epoch": 1.03, "learning_rate": 9.065482410309997e-06, "loss": 0.0724, "step": 2980 }, { "epoch": 1.04, "learning_rate": 9.061999303378614e-06, "loss": 0.07, "step": 2990 }, { "epoch": 1.04, "learning_rate": 9.05851619644723e-06, "loss": 0.0698, "step": 3000 }, { "epoch": 1.04, "eval_loss": 0.1173095703125, "eval_runtime": 667.2017, "eval_samples_per_second": 4.27, "eval_steps_per_second": 0.034, "eval_wer": 22.59983007646559, "step": 3000 }, { "epoch": 1.05, "learning_rate": 9.055033089515849e-06, "loss": 0.0736, "step": 3010 }, { "epoch": 1.05, "learning_rate": 9.051549982584467e-06, "loss": 0.0697, "step": 3020 }, { "epoch": 1.05, "learning_rate": 9.048066875653084e-06, "loss": 0.0716, "step": 3030 }, { "epoch": 1.06, "learning_rate": 9.0445837687217e-06, "loss": 0.0685, "step": 3040 }, { "epoch": 1.06, "learning_rate": 9.041100661790317e-06, "loss": 0.0744, "step": 3050 }, { "epoch": 1.06, "learning_rate": 9.037617554858935e-06, "loss": 0.0712, "step": 3060 }, { "epoch": 1.07, "learning_rate": 9.034134447927552e-06, "loss": 0.0708, "step": 3070 }, { "epoch": 1.07, "learning_rate": 9.03065134099617e-06, "loss": 0.0739, "step": 3080 }, { "epoch": 1.07, "learning_rate": 9.027168234064786e-06, "loss": 0.0712, "step": 3090 }, { "epoch": 1.08, "learning_rate": 9.023685127133403e-06, "loss": 0.0738, "step": 3100 }, { "epoch": 1.08, "eval_loss": 0.11474609375, "eval_runtime": 772.0192, "eval_samples_per_second": 3.69, "eval_steps_per_second": 0.03, "eval_wer": 22.162434311967022, "step": 3100 }, { "epoch": 1.08, "learning_rate": 9.020202020202021e-06, "loss": 0.0725, "step": 3110 }, { "epoch": 1.08, "learning_rate": 9.016718913270638e-06, "loss": 0.0753, "step": 3120 }, { "epoch": 1.09, "learning_rate": 9.013235806339256e-06, "loss": 0.0705, "step": 3130 }, { "epoch": 1.09, "learning_rate": 9.009752699407873e-06, "loss": 0.0715, "step": 3140 }, { "epoch": 1.09, "learning_rate": 9.00626959247649e-06, "loss": 0.0774, "step": 3150 }, { "epoch": 1.1, "learning_rate": 9.002786485545108e-06, "loss": 0.0714, "step": 3160 }, { "epoch": 1.1, "learning_rate": 8.999303378613724e-06, "loss": 0.0699, "step": 3170 }, { "epoch": 1.1, "learning_rate": 8.995820271682342e-06, "loss": 0.0756, "step": 3180 }, { "epoch": 1.11, "learning_rate": 8.992337164750959e-06, "loss": 0.0715, "step": 3190 }, { "epoch": 1.11, "learning_rate": 8.988854057819576e-06, "loss": 0.0704, "step": 3200 }, { "epoch": 1.11, "eval_loss": 0.11749267578125, "eval_runtime": 486.4812, "eval_samples_per_second": 5.856, "eval_steps_per_second": 0.047, "eval_wer": 22.017684634507066, "step": 3200 }, { "epoch": 1.11, "learning_rate": 8.985370950888192e-06, "loss": 0.0723, "step": 3210 }, { "epoch": 1.12, "learning_rate": 8.98188784395681e-06, "loss": 0.0726, "step": 3220 }, { "epoch": 1.12, "learning_rate": 8.978404737025427e-06, "loss": 0.0711, "step": 3230 }, { "epoch": 1.12, "learning_rate": 8.974921630094044e-06, "loss": 0.0704, "step": 3240 }, { "epoch": 1.13, "learning_rate": 8.971438523162662e-06, "loss": 0.0716, "step": 3250 }, { "epoch": 1.13, "learning_rate": 8.967955416231279e-06, "loss": 0.0723, "step": 3260 }, { "epoch": 1.14, "learning_rate": 8.964472309299897e-06, "loss": 0.0751, "step": 3270 }, { "epoch": 1.14, "learning_rate": 8.960989202368513e-06, "loss": 0.072, "step": 3280 }, { "epoch": 1.14, "learning_rate": 8.95750609543713e-06, "loss": 0.0707, "step": 3290 }, { "epoch": 1.15, "learning_rate": 8.954022988505748e-06, "loss": 0.0671, "step": 3300 }, { "epoch": 1.15, "eval_loss": 0.11810302734375, "eval_runtime": 700.0763, "eval_samples_per_second": 4.07, "eval_steps_per_second": 0.033, "eval_wer": 22.03341829510054, "step": 3300 }, { "epoch": 1.15, "learning_rate": 8.950539881574365e-06, "loss": 0.0725, "step": 3310 }, { "epoch": 1.15, "learning_rate": 8.947056774642983e-06, "loss": 0.0762, "step": 3320 }, { "epoch": 1.16, "learning_rate": 8.9435736677116e-06, "loss": 0.0706, "step": 3330 }, { "epoch": 1.16, "learning_rate": 8.940090560780216e-06, "loss": 0.0692, "step": 3340 }, { "epoch": 1.16, "learning_rate": 8.936607453848833e-06, "loss": 0.0739, "step": 3350 }, { "epoch": 1.17, "learning_rate": 8.933124346917451e-06, "loss": 0.0751, "step": 3360 }, { "epoch": 1.17, "learning_rate": 8.92964123998607e-06, "loss": 0.0733, "step": 3370 }, { "epoch": 1.17, "learning_rate": 8.926158133054686e-06, "loss": 0.0692, "step": 3380 }, { "epoch": 1.18, "learning_rate": 8.922675026123303e-06, "loss": 0.0713, "step": 3390 }, { "epoch": 1.18, "learning_rate": 8.919191919191919e-06, "loss": 0.0699, "step": 3400 }, { "epoch": 1.18, "eval_loss": 0.11517333984375, "eval_runtime": 774.6303, "eval_samples_per_second": 3.678, "eval_steps_per_second": 0.03, "eval_wer": 22.04285849145662, "step": 3400 }, { "epoch": 1.18, "learning_rate": 8.915708812260537e-06, "loss": 0.0737, "step": 3410 }, { "epoch": 1.19, "learning_rate": 8.912225705329154e-06, "loss": 0.0709, "step": 3420 }, { "epoch": 1.19, "learning_rate": 8.908742598397772e-06, "loss": 0.0702, "step": 3430 }, { "epoch": 1.19, "learning_rate": 8.905259491466389e-06, "loss": 0.0685, "step": 3440 }, { "epoch": 1.2, "learning_rate": 8.901776384535005e-06, "loss": 0.0739, "step": 3450 }, { "epoch": 1.2, "learning_rate": 8.898293277603624e-06, "loss": 0.0718, "step": 3460 }, { "epoch": 1.2, "learning_rate": 8.89481017067224e-06, "loss": 0.0717, "step": 3470 }, { "epoch": 1.21, "learning_rate": 8.891327063740859e-06, "loss": 0.0744, "step": 3480 }, { "epoch": 1.21, "learning_rate": 8.887843956809475e-06, "loss": 0.0717, "step": 3490 }, { "epoch": 1.22, "learning_rate": 8.884360849878092e-06, "loss": 0.0686, "step": 3500 }, { "epoch": 1.22, "eval_loss": 0.11749267578125, "eval_runtime": 492.6786, "eval_samples_per_second": 5.783, "eval_steps_per_second": 0.047, "eval_wer": 22.231662418578306, "step": 3500 }, { "epoch": 1.22, "learning_rate": 8.88087774294671e-06, "loss": 0.0727, "step": 3510 }, { "epoch": 1.22, "learning_rate": 8.877394636015327e-06, "loss": 0.0704, "step": 3520 }, { "epoch": 1.23, "learning_rate": 8.873911529083943e-06, "loss": 0.0737, "step": 3530 }, { "epoch": 1.23, "learning_rate": 8.87042842215256e-06, "loss": 0.0728, "step": 3540 }, { "epoch": 1.23, "learning_rate": 8.866945315221178e-06, "loss": 0.0764, "step": 3550 }, { "epoch": 1.24, "learning_rate": 8.863462208289795e-06, "loss": 0.0745, "step": 3560 }, { "epoch": 1.24, "learning_rate": 8.859979101358413e-06, "loss": 0.0688, "step": 3570 }, { "epoch": 1.24, "learning_rate": 8.85649599442703e-06, "loss": 0.0715, "step": 3580 }, { "epoch": 1.25, "learning_rate": 8.853012887495646e-06, "loss": 0.0704, "step": 3590 }, { "epoch": 1.25, "learning_rate": 8.849529780564264e-06, "loss": 0.0742, "step": 3600 }, { "epoch": 1.25, "eval_loss": 0.113525390625, "eval_runtime": 293.9147, "eval_samples_per_second": 9.693, "eval_steps_per_second": 0.078, "eval_wer": 22.181314704679192, "step": 3600 }, { "epoch": 1.25, "learning_rate": 8.846046673632881e-06, "loss": 0.0726, "step": 3610 }, { "epoch": 1.26, "learning_rate": 8.8425635667015e-06, "loss": 0.0731, "step": 3620 }, { "epoch": 1.26, "learning_rate": 8.839080459770116e-06, "loss": 0.0712, "step": 3630 }, { "epoch": 1.26, "learning_rate": 8.835597352838732e-06, "loss": 0.0672, "step": 3640 }, { "epoch": 1.27, "learning_rate": 8.83211424590735e-06, "loss": 0.0704, "step": 3650 }, { "epoch": 1.27, "learning_rate": 8.828631138975967e-06, "loss": 0.0721, "step": 3660 }, { "epoch": 1.27, "learning_rate": 8.825148032044585e-06, "loss": 0.0705, "step": 3670 }, { "epoch": 1.28, "learning_rate": 8.821664925113202e-06, "loss": 0.0722, "step": 3680 }, { "epoch": 1.28, "learning_rate": 8.818181818181819e-06, "loss": 0.0749, "step": 3690 }, { "epoch": 1.28, "learning_rate": 8.814698711250435e-06, "loss": 0.0698, "step": 3700 }, { "epoch": 1.28, "eval_loss": 0.113037109375, "eval_runtime": 720.2659, "eval_samples_per_second": 3.955, "eval_steps_per_second": 0.032, "eval_wer": 22.03027156298184, "step": 3700 }, { "epoch": 1.29, "learning_rate": 8.811215604319053e-06, "loss": 0.0668, "step": 3710 }, { "epoch": 1.29, "learning_rate": 8.807732497387672e-06, "loss": 0.0712, "step": 3720 }, { "epoch": 1.3, "learning_rate": 8.804249390456288e-06, "loss": 0.068, "step": 3730 }, { "epoch": 1.3, "learning_rate": 8.800766283524905e-06, "loss": 0.07, "step": 3740 }, { "epoch": 1.3, "learning_rate": 8.797283176593521e-06, "loss": 0.0704, "step": 3750 }, { "epoch": 1.31, "learning_rate": 8.79380006966214e-06, "loss": 0.0703, "step": 3760 }, { "epoch": 1.31, "learning_rate": 8.790316962730756e-06, "loss": 0.0739, "step": 3770 }, { "epoch": 1.31, "learning_rate": 8.786833855799373e-06, "loss": 0.0736, "step": 3780 }, { "epoch": 1.32, "learning_rate": 8.783350748867991e-06, "loss": 0.0684, "step": 3790 }, { "epoch": 1.32, "learning_rate": 8.779867641936608e-06, "loss": 0.0712, "step": 3800 }, { "epoch": 1.32, "eval_loss": 0.11163330078125, "eval_runtime": 782.2761, "eval_samples_per_second": 3.642, "eval_steps_per_second": 0.029, "eval_wer": 21.920135938827528, "step": 3800 }, { "epoch": 1.32, "learning_rate": 8.776384535005226e-06, "loss": 0.0717, "step": 3810 }, { "epoch": 1.33, "learning_rate": 8.772901428073843e-06, "loss": 0.0703, "step": 3820 }, { "epoch": 1.33, "learning_rate": 8.76941832114246e-06, "loss": 0.073, "step": 3830 }, { "epoch": 1.33, "learning_rate": 8.765935214211076e-06, "loss": 0.0728, "step": 3840 }, { "epoch": 1.34, "learning_rate": 8.762452107279694e-06, "loss": 0.0713, "step": 3850 }, { "epoch": 1.34, "learning_rate": 8.758969000348312e-06, "loss": 0.0701, "step": 3860 }, { "epoch": 1.34, "learning_rate": 8.755485893416929e-06, "loss": 0.0732, "step": 3870 }, { "epoch": 1.35, "learning_rate": 8.752002786485546e-06, "loss": 0.0734, "step": 3880 }, { "epoch": 1.35, "learning_rate": 8.748519679554162e-06, "loss": 0.07, "step": 3890 }, { "epoch": 1.35, "learning_rate": 8.74503657262278e-06, "loss": 0.0705, "step": 3900 }, { "epoch": 1.35, "eval_loss": 0.114990234375, "eval_runtime": 518.7116, "eval_samples_per_second": 5.492, "eval_steps_per_second": 0.044, "eval_wer": 21.680984297806727, "step": 3900 }, { "epoch": 1.36, "learning_rate": 8.741553465691397e-06, "loss": 0.0702, "step": 3910 }, { "epoch": 1.36, "learning_rate": 8.738070358760015e-06, "loss": 0.0704, "step": 3920 }, { "epoch": 1.36, "learning_rate": 8.734587251828632e-06, "loss": 0.0689, "step": 3930 }, { "epoch": 1.37, "learning_rate": 8.731104144897248e-06, "loss": 0.0709, "step": 3940 }, { "epoch": 1.37, "learning_rate": 8.727621037965867e-06, "loss": 0.07, "step": 3950 }, { "epoch": 1.38, "learning_rate": 8.724137931034483e-06, "loss": 0.0722, "step": 3960 }, { "epoch": 1.38, "learning_rate": 8.720654824103102e-06, "loss": 0.0717, "step": 3970 }, { "epoch": 1.38, "learning_rate": 8.717171717171718e-06, "loss": 0.0716, "step": 3980 }, { "epoch": 1.39, "learning_rate": 8.713688610240335e-06, "loss": 0.068, "step": 3990 }, { "epoch": 1.39, "learning_rate": 8.710205503308953e-06, "loss": 0.0701, "step": 4000 }, { "epoch": 1.39, "eval_loss": 0.11181640625, "eval_runtime": 775.9038, "eval_samples_per_second": 3.672, "eval_steps_per_second": 0.03, "eval_wer": 21.636930048145, "step": 4000 }, { "epoch": 1.39, "learning_rate": 8.70672239637757e-06, "loss": 0.0732, "step": 4010 }, { "epoch": 1.4, "learning_rate": 8.703239289446188e-06, "loss": 0.0666, "step": 4020 }, { "epoch": 1.4, "learning_rate": 8.699756182514804e-06, "loss": 0.0701, "step": 4030 }, { "epoch": 1.4, "learning_rate": 8.696273075583421e-06, "loss": 0.0726, "step": 4040 }, { "epoch": 1.41, "learning_rate": 8.692789968652038e-06, "loss": 0.0724, "step": 4050 }, { "epoch": 1.41, "learning_rate": 8.689306861720656e-06, "loss": 0.0701, "step": 4060 }, { "epoch": 1.41, "learning_rate": 8.685823754789272e-06, "loss": 0.0713, "step": 4070 }, { "epoch": 1.42, "learning_rate": 8.682340647857889e-06, "loss": 0.0689, "step": 4080 }, { "epoch": 1.42, "learning_rate": 8.678857540926507e-06, "loss": 0.0707, "step": 4090 }, { "epoch": 1.42, "learning_rate": 8.675374433995124e-06, "loss": 0.0675, "step": 4100 }, { "epoch": 1.42, "eval_loss": 0.1124267578125, "eval_runtime": 663.0481, "eval_samples_per_second": 4.297, "eval_steps_per_second": 0.035, "eval_wer": 21.68413102992542, "step": 4100 }, { "epoch": 1.43, "learning_rate": 8.671891327063742e-06, "loss": 0.0713, "step": 4110 }, { "epoch": 1.43, "learning_rate": 8.668408220132359e-06, "loss": 0.0678, "step": 4120 }, { "epoch": 1.43, "learning_rate": 8.664925113200975e-06, "loss": 0.0729, "step": 4130 }, { "epoch": 1.44, "learning_rate": 8.661442006269594e-06, "loss": 0.0703, "step": 4140 }, { "epoch": 1.44, "learning_rate": 8.65795889933821e-06, "loss": 0.0749, "step": 4150 }, { "epoch": 1.44, "learning_rate": 8.654475792406828e-06, "loss": 0.0687, "step": 4160 }, { "epoch": 1.45, "learning_rate": 8.650992685475445e-06, "loss": 0.0697, "step": 4170 }, { "epoch": 1.45, "learning_rate": 8.647509578544062e-06, "loss": 0.0698, "step": 4180 }, { "epoch": 1.45, "learning_rate": 8.644026471612678e-06, "loss": 0.0722, "step": 4190 }, { "epoch": 1.46, "learning_rate": 8.640543364681296e-06, "loss": 0.0671, "step": 4200 }, { "epoch": 1.46, "eval_loss": 0.109130859375, "eval_runtime": 486.9376, "eval_samples_per_second": 5.851, "eval_steps_per_second": 0.047, "eval_wer": 21.253028729664244, "step": 4200 }, { "epoch": 1.46, "learning_rate": 8.637060257749915e-06, "loss": 0.0704, "step": 4210 }, { "epoch": 1.47, "learning_rate": 8.633577150818531e-06, "loss": 0.0704, "step": 4220 }, { "epoch": 1.47, "learning_rate": 8.630094043887148e-06, "loss": 0.073, "step": 4230 }, { "epoch": 1.47, "learning_rate": 8.626610936955764e-06, "loss": 0.0737, "step": 4240 }, { "epoch": 1.48, "learning_rate": 8.623127830024383e-06, "loss": 0.0699, "step": 4250 }, { "epoch": 1.48, "learning_rate": 8.619644723093e-06, "loss": 0.0694, "step": 4260 }, { "epoch": 1.48, "learning_rate": 8.616161616161618e-06, "loss": 0.069, "step": 4270 }, { "epoch": 1.49, "learning_rate": 8.612678509230234e-06, "loss": 0.0711, "step": 4280 }, { "epoch": 1.49, "learning_rate": 8.60919540229885e-06, "loss": 0.0695, "step": 4290 }, { "epoch": 1.49, "learning_rate": 8.605712295367469e-06, "loss": 0.0673, "step": 4300 }, { "epoch": 1.49, "eval_loss": 0.1119384765625, "eval_runtime": 294.9453, "eval_samples_per_second": 9.659, "eval_steps_per_second": 0.078, "eval_wer": 21.341137228987698, "step": 4300 }, { "epoch": 1.5, "learning_rate": 8.602229188436086e-06, "loss": 0.0731, "step": 4310 }, { "epoch": 1.5, "learning_rate": 8.598746081504704e-06, "loss": 0.066, "step": 4320 }, { "epoch": 1.5, "learning_rate": 8.59526297457332e-06, "loss": 0.0702, "step": 4330 }, { "epoch": 1.51, "learning_rate": 8.591779867641937e-06, "loss": 0.0704, "step": 4340 }, { "epoch": 1.51, "learning_rate": 8.588296760710555e-06, "loss": 0.0701, "step": 4350 }, { "epoch": 1.51, "learning_rate": 8.584813653779172e-06, "loss": 0.0736, "step": 4360 }, { "epoch": 1.52, "learning_rate": 8.581330546847788e-06, "loss": 0.07, "step": 4370 }, { "epoch": 1.52, "learning_rate": 8.577847439916405e-06, "loss": 0.0715, "step": 4380 }, { "epoch": 1.52, "learning_rate": 8.574364332985023e-06, "loss": 0.0687, "step": 4390 }, { "epoch": 1.53, "learning_rate": 8.57088122605364e-06, "loss": 0.0719, "step": 4400 }, { "epoch": 1.53, "eval_loss": 0.111572265625, "eval_runtime": 650.2805, "eval_samples_per_second": 4.381, "eval_steps_per_second": 0.035, "eval_wer": 21.385191478649425, "step": 4400 }, { "epoch": 1.53, "learning_rate": 8.567398119122258e-06, "loss": 0.0691, "step": 4410 }, { "epoch": 1.53, "learning_rate": 8.563915012190875e-06, "loss": 0.0664, "step": 4420 }, { "epoch": 1.54, "learning_rate": 8.560431905259491e-06, "loss": 0.0696, "step": 4430 }, { "epoch": 1.54, "learning_rate": 8.55694879832811e-06, "loss": 0.0721, "step": 4440 }, { "epoch": 1.55, "learning_rate": 8.553465691396726e-06, "loss": 0.0717, "step": 4450 }, { "epoch": 1.55, "learning_rate": 8.549982584465344e-06, "loss": 0.0684, "step": 4460 }, { "epoch": 1.55, "learning_rate": 8.546499477533961e-06, "loss": 0.0702, "step": 4470 }, { "epoch": 1.56, "learning_rate": 8.543016370602578e-06, "loss": 0.0689, "step": 4480 }, { "epoch": 1.56, "learning_rate": 8.539533263671196e-06, "loss": 0.0721, "step": 4490 }, { "epoch": 1.56, "learning_rate": 8.536050156739812e-06, "loss": 0.0711, "step": 4500 }, { "epoch": 1.56, "eval_loss": 0.110595703125, "eval_runtime": 785.4826, "eval_samples_per_second": 3.627, "eval_steps_per_second": 0.029, "eval_wer": 21.407218603480285, "step": 4500 }, { "epoch": 1.57, "learning_rate": 8.53256704980843e-06, "loss": 0.0718, "step": 4510 }, { "epoch": 1.57, "learning_rate": 8.529083942877047e-06, "loss": 0.0684, "step": 4520 }, { "epoch": 1.57, "learning_rate": 8.525600835945664e-06, "loss": 0.0664, "step": 4530 }, { "epoch": 1.58, "learning_rate": 8.52211772901428e-06, "loss": 0.0626, "step": 4540 }, { "epoch": 1.58, "learning_rate": 8.518634622082899e-06, "loss": 0.0687, "step": 4550 }, { "epoch": 1.58, "learning_rate": 8.515151515151517e-06, "loss": 0.066, "step": 4560 }, { "epoch": 1.59, "learning_rate": 8.511668408220134e-06, "loss": 0.0724, "step": 4570 }, { "epoch": 1.59, "learning_rate": 8.50818530128875e-06, "loss": 0.0699, "step": 4580 }, { "epoch": 1.59, "learning_rate": 8.504702194357367e-06, "loss": 0.0677, "step": 4590 }, { "epoch": 1.6, "learning_rate": 8.501219087425985e-06, "loss": 0.071, "step": 4600 }, { "epoch": 1.6, "eval_loss": 0.10986328125, "eval_runtime": 486.6542, "eval_samples_per_second": 5.854, "eval_steps_per_second": 0.047, "eval_wer": 21.347430693225085, "step": 4600 }, { "epoch": 1.6, "learning_rate": 8.497735980494602e-06, "loss": 0.0706, "step": 4610 }, { "epoch": 1.6, "learning_rate": 8.494252873563218e-06, "loss": 0.0663, "step": 4620 }, { "epoch": 1.61, "learning_rate": 8.490769766631836e-06, "loss": 0.0686, "step": 4630 }, { "epoch": 1.61, "learning_rate": 8.487286659700453e-06, "loss": 0.0689, "step": 4640 }, { "epoch": 1.61, "learning_rate": 8.483803552769071e-06, "loss": 0.0702, "step": 4650 }, { "epoch": 1.62, "learning_rate": 8.480320445837688e-06, "loss": 0.0679, "step": 4660 }, { "epoch": 1.62, "learning_rate": 8.476837338906305e-06, "loss": 0.069, "step": 4670 }, { "epoch": 1.62, "learning_rate": 8.473354231974921e-06, "loss": 0.0677, "step": 4680 }, { "epoch": 1.63, "learning_rate": 8.46987112504354e-06, "loss": 0.0749, "step": 4690 }, { "epoch": 1.63, "learning_rate": 8.466388018112158e-06, "loss": 0.0702, "step": 4700 }, { "epoch": 1.63, "eval_loss": 0.1082763671875, "eval_runtime": 726.1025, "eval_samples_per_second": 3.924, "eval_steps_per_second": 0.032, "eval_wer": 21.14603983762862, "step": 4700 }, { "epoch": 1.64, "learning_rate": 8.462904911180774e-06, "loss": 0.069, "step": 4710 }, { "epoch": 1.64, "learning_rate": 8.45942180424939e-06, "loss": 0.0668, "step": 4720 }, { "epoch": 1.64, "learning_rate": 8.455938697318007e-06, "loss": 0.0685, "step": 4730 }, { "epoch": 1.65, "learning_rate": 8.452455590386626e-06, "loss": 0.0706, "step": 4740 }, { "epoch": 1.65, "learning_rate": 8.448972483455242e-06, "loss": 0.0693, "step": 4750 }, { "epoch": 1.65, "learning_rate": 8.44548937652386e-06, "loss": 0.0677, "step": 4760 }, { "epoch": 1.66, "learning_rate": 8.442006269592477e-06, "loss": 0.0719, "step": 4770 }, { "epoch": 1.66, "learning_rate": 8.438523162661094e-06, "loss": 0.0692, "step": 4780 }, { "epoch": 1.66, "learning_rate": 8.435040055729712e-06, "loss": 0.0671, "step": 4790 }, { "epoch": 1.67, "learning_rate": 8.431556948798329e-06, "loss": 0.0686, "step": 4800 }, { "epoch": 1.67, "eval_loss": 0.1087646484375, "eval_runtime": 787.116, "eval_samples_per_second": 3.62, "eval_steps_per_second": 0.029, "eval_wer": 21.725038547468454, "step": 4800 }, { "epoch": 1.67, "learning_rate": 8.428073841866947e-06, "loss": 0.066, "step": 4810 }, { "epoch": 1.67, "learning_rate": 8.424590734935563e-06, "loss": 0.0661, "step": 4820 }, { "epoch": 1.68, "learning_rate": 8.42110762800418e-06, "loss": 0.0661, "step": 4830 }, { "epoch": 1.68, "learning_rate": 8.417624521072798e-06, "loss": 0.0677, "step": 4840 }, { "epoch": 1.68, "learning_rate": 8.414141414141415e-06, "loss": 0.0678, "step": 4850 }, { "epoch": 1.69, "learning_rate": 8.410658307210033e-06, "loss": 0.0699, "step": 4860 }, { "epoch": 1.69, "learning_rate": 8.40717520027865e-06, "loss": 0.0696, "step": 4870 }, { "epoch": 1.69, "learning_rate": 8.403692093347266e-06, "loss": 0.073, "step": 4880 }, { "epoch": 1.7, "learning_rate": 8.400208986415883e-06, "loss": 0.0693, "step": 4890 }, { "epoch": 1.7, "learning_rate": 8.396725879484501e-06, "loss": 0.0643, "step": 4900 }, { "epoch": 1.7, "eval_loss": 0.10882568359375, "eval_runtime": 484.8642, "eval_samples_per_second": 5.876, "eval_steps_per_second": 0.047, "eval_wer": 21.312816639919443, "step": 4900 }, { "epoch": 1.7, "learning_rate": 8.393242772553118e-06, "loss": 0.0684, "step": 4910 }, { "epoch": 1.71, "learning_rate": 8.389759665621734e-06, "loss": 0.0676, "step": 4920 }, { "epoch": 1.71, "learning_rate": 8.386276558690353e-06, "loss": 0.0673, "step": 4930 }, { "epoch": 1.72, "learning_rate": 8.382793451758969e-06, "loss": 0.0712, "step": 4940 }, { "epoch": 1.72, "learning_rate": 8.379310344827587e-06, "loss": 0.0666, "step": 4950 }, { "epoch": 1.72, "learning_rate": 8.375827237896204e-06, "loss": 0.0678, "step": 4960 }, { "epoch": 1.73, "learning_rate": 8.37234413096482e-06, "loss": 0.0686, "step": 4970 }, { "epoch": 1.73, "learning_rate": 8.368861024033439e-06, "loss": 0.0696, "step": 4980 }, { "epoch": 1.73, "learning_rate": 8.365377917102055e-06, "loss": 0.0691, "step": 4990 }, { "epoch": 1.74, "learning_rate": 8.361894810170674e-06, "loss": 0.0669, "step": 5000 }, { "epoch": 1.74, "eval_loss": 0.1083984375, "eval_runtime": 302.723, "eval_samples_per_second": 9.411, "eval_steps_per_second": 0.076, "eval_wer": 21.199534283646436, "step": 5000 }, { "epoch": 1.74, "learning_rate": 8.359108324625566e-06, "loss": 0.0652, "step": 5010 }, { "epoch": 1.74, "learning_rate": 8.355625217694185e-06, "loss": 0.0708, "step": 5020 }, { "epoch": 1.75, "learning_rate": 8.352142110762801e-06, "loss": 0.0678, "step": 5030 }, { "epoch": 1.75, "learning_rate": 8.34865900383142e-06, "loss": 0.0698, "step": 5040 }, { "epoch": 1.75, "learning_rate": 8.345175896900036e-06, "loss": 0.0677, "step": 5050 }, { "epoch": 1.76, "learning_rate": 8.341692789968653e-06, "loss": 0.0715, "step": 5060 }, { "epoch": 1.76, "learning_rate": 8.33820968303727e-06, "loss": 0.0699, "step": 5070 }, { "epoch": 1.76, "learning_rate": 8.334726576105887e-06, "loss": 0.0701, "step": 5080 }, { "epoch": 1.77, "learning_rate": 8.331243469174504e-06, "loss": 0.07, "step": 5090 }, { "epoch": 1.77, "learning_rate": 8.32776036224312e-06, "loss": 0.0712, "step": 5100 }, { "epoch": 1.77, "eval_loss": 0.106201171875, "eval_runtime": 311.8176, "eval_samples_per_second": 9.137, "eval_steps_per_second": 0.074, "eval_wer": 21.13030617703515, "step": 5100 }, { "epoch": 1.77, "learning_rate": 8.324277255311739e-06, "loss": 0.0678, "step": 5110 }, { "epoch": 1.78, "learning_rate": 8.320794148380355e-06, "loss": 0.0679, "step": 5120 }, { "epoch": 1.78, "learning_rate": 8.317311041448974e-06, "loss": 0.0691, "step": 5130 }, { "epoch": 1.78, "learning_rate": 8.31382793451759e-06, "loss": 0.0693, "step": 5140 }, { "epoch": 1.79, "learning_rate": 8.310344827586207e-06, "loss": 0.0668, "step": 5150 }, { "epoch": 1.79, "learning_rate": 8.306861720654825e-06, "loss": 0.0666, "step": 5160 }, { "epoch": 1.8, "learning_rate": 8.303378613723442e-06, "loss": 0.0663, "step": 5170 }, { "epoch": 1.8, "learning_rate": 8.29989550679206e-06, "loss": 0.0683, "step": 5180 }, { "epoch": 1.8, "learning_rate": 8.296412399860677e-06, "loss": 0.0663, "step": 5190 }, { "epoch": 1.81, "learning_rate": 8.292929292929293e-06, "loss": 0.066, "step": 5200 }, { "epoch": 1.81, "eval_loss": 0.10882568359375, "eval_runtime": 328.1314, "eval_samples_per_second": 8.682, "eval_steps_per_second": 0.07, "eval_wer": 21.34428396110639, "step": 5200 }, { "epoch": 1.81, "learning_rate": 8.28944618599791e-06, "loss": 0.0708, "step": 5210 }, { "epoch": 1.81, "learning_rate": 8.285963079066528e-06, "loss": 0.0695, "step": 5220 }, { "epoch": 1.82, "learning_rate": 8.282479972135146e-06, "loss": 0.0647, "step": 5230 }, { "epoch": 1.82, "learning_rate": 8.278996865203763e-06, "loss": 0.0673, "step": 5240 }, { "epoch": 1.82, "learning_rate": 8.27551375827238e-06, "loss": 0.0675, "step": 5250 }, { "epoch": 1.83, "learning_rate": 8.272030651340996e-06, "loss": 0.0672, "step": 5260 }, { "epoch": 1.83, "learning_rate": 8.268547544409614e-06, "loss": 0.0693, "step": 5270 }, { "epoch": 1.83, "learning_rate": 8.265064437478231e-06, "loss": 0.0662, "step": 5280 }, { "epoch": 1.84, "learning_rate": 8.26158133054685e-06, "loss": 0.0683, "step": 5290 }, { "epoch": 1.84, "learning_rate": 8.258098223615466e-06, "loss": 0.0688, "step": 5300 }, { "epoch": 1.84, "eval_loss": 0.10693359375, "eval_runtime": 683.9313, "eval_samples_per_second": 4.166, "eval_steps_per_second": 0.034, "eval_wer": 20.994996695931274, "step": 5300 }, { "epoch": 1.84, "learning_rate": 8.254615116684082e-06, "loss": 0.0683, "step": 5310 }, { "epoch": 1.85, "learning_rate": 8.2511320097527e-06, "loss": 0.066, "step": 5320 }, { "epoch": 1.85, "learning_rate": 8.247648902821317e-06, "loss": 0.066, "step": 5330 }, { "epoch": 1.85, "learning_rate": 8.244165795889934e-06, "loss": 0.073, "step": 5340 }, { "epoch": 1.86, "learning_rate": 8.24068268895855e-06, "loss": 0.0679, "step": 5350 }, { "epoch": 1.86, "learning_rate": 8.237199582027169e-06, "loss": 0.0707, "step": 5360 }, { "epoch": 1.86, "learning_rate": 8.233716475095787e-06, "loss": 0.0693, "step": 5370 }, { "epoch": 1.87, "learning_rate": 8.230233368164403e-06, "loss": 0.065, "step": 5380 }, { "epoch": 1.87, "learning_rate": 8.22675026123302e-06, "loss": 0.0672, "step": 5390 }, { "epoch": 1.88, "learning_rate": 8.223267154301637e-06, "loss": 0.0709, "step": 5400 }, { "epoch": 1.88, "eval_loss": 0.103515625, "eval_runtime": 310.8453, "eval_samples_per_second": 9.165, "eval_steps_per_second": 0.074, "eval_wer": 20.686616948299193, "step": 5400 }, { "epoch": 1.88, "learning_rate": 8.219784047370255e-06, "loss": 0.0686, "step": 5410 }, { "epoch": 1.88, "learning_rate": 8.216300940438872e-06, "loss": 0.0679, "step": 5420 }, { "epoch": 1.89, "learning_rate": 8.21281783350749e-06, "loss": 0.0677, "step": 5430 }, { "epoch": 1.89, "learning_rate": 8.209334726576106e-06, "loss": 0.0658, "step": 5440 }, { "epoch": 1.89, "learning_rate": 8.205851619644723e-06, "loss": 0.0667, "step": 5450 }, { "epoch": 1.9, "learning_rate": 8.202368512713341e-06, "loss": 0.0644, "step": 5460 }, { "epoch": 1.9, "learning_rate": 8.198885405781958e-06, "loss": 0.0699, "step": 5470 }, { "epoch": 1.9, "learning_rate": 8.195402298850576e-06, "loss": 0.0694, "step": 5480 }, { "epoch": 1.91, "learning_rate": 8.191919191919193e-06, "loss": 0.0675, "step": 5490 }, { "epoch": 1.91, "learning_rate": 8.18843608498781e-06, "loss": 0.0688, "step": 5500 }, { "epoch": 1.91, "eval_loss": 0.10687255859375, "eval_runtime": 300.4898, "eval_samples_per_second": 9.481, "eval_steps_per_second": 0.077, "eval_wer": 20.699203876773968, "step": 5500 }, { "epoch": 1.91, "learning_rate": 8.184952978056428e-06, "loss": 0.0667, "step": 5510 }, { "epoch": 1.92, "learning_rate": 8.181469871125044e-06, "loss": 0.0683, "step": 5520 }, { "epoch": 1.92, "learning_rate": 8.177986764193662e-06, "loss": 0.0655, "step": 5530 }, { "epoch": 1.92, "learning_rate": 8.174503657262279e-06, "loss": 0.0682, "step": 5540 }, { "epoch": 1.93, "learning_rate": 8.171020550330896e-06, "loss": 0.0662, "step": 5550 }, { "epoch": 1.93, "learning_rate": 8.167537443399512e-06, "loss": 0.0689, "step": 5560 }, { "epoch": 1.93, "learning_rate": 8.16405433646813e-06, "loss": 0.0665, "step": 5570 }, { "epoch": 1.94, "learning_rate": 8.160571229536749e-06, "loss": 0.0675, "step": 5580 }, { "epoch": 1.94, "learning_rate": 8.157088122605365e-06, "loss": 0.0681, "step": 5590 }, { "epoch": 1.94, "learning_rate": 8.153605015673982e-06, "loss": 0.0711, "step": 5600 }, { "epoch": 1.94, "eval_loss": 0.10504150390625, "eval_runtime": 679.7242, "eval_samples_per_second": 4.191, "eval_steps_per_second": 0.034, "eval_wer": 21.001290160168665, "step": 5600 }, { "epoch": 1.95, "learning_rate": 8.150121908742598e-06, "loss": 0.0661, "step": 5610 }, { "epoch": 1.95, "learning_rate": 8.146638801811217e-06, "loss": 0.0658, "step": 5620 }, { "epoch": 1.95, "learning_rate": 8.143155694879833e-06, "loss": 0.0657, "step": 5630 }, { "epoch": 1.96, "learning_rate": 8.13967258794845e-06, "loss": 0.0681, "step": 5640 }, { "epoch": 1.96, "learning_rate": 8.136189481017068e-06, "loss": 0.0671, "step": 5650 }, { "epoch": 1.97, "learning_rate": 8.132706374085685e-06, "loss": 0.0719, "step": 5660 }, { "epoch": 1.97, "learning_rate": 8.129223267154303e-06, "loss": 0.0695, "step": 5670 }, { "epoch": 1.97, "learning_rate": 8.12574016022292e-06, "loss": 0.0654, "step": 5680 }, { "epoch": 1.98, "learning_rate": 8.122257053291536e-06, "loss": 0.0673, "step": 5690 }, { "epoch": 1.98, "learning_rate": 8.118773946360153e-06, "loss": 0.0676, "step": 5700 }, { "epoch": 1.98, "eval_loss": 0.111572265625, "eval_runtime": 263.5896, "eval_samples_per_second": 10.808, "eval_steps_per_second": 0.087, "eval_wer": 20.837660089996536, "step": 5700 }, { "epoch": 1.98, "learning_rate": 8.115290839428771e-06, "loss": 0.0675, "step": 5710 }, { "epoch": 1.99, "learning_rate": 8.11180773249739e-06, "loss": 0.0694, "step": 5720 }, { "epoch": 1.99, "learning_rate": 8.108324625566006e-06, "loss": 0.0713, "step": 5730 }, { "epoch": 1.99, "learning_rate": 8.104841518634622e-06, "loss": 0.0681, "step": 5740 }, { "epoch": 2.0, "learning_rate": 8.101358411703239e-06, "loss": 0.067, "step": 5750 }, { "epoch": 2.0, "learning_rate": 8.097875304771857e-06, "loss": 0.0719, "step": 5760 }, { "epoch": 2.0, "learning_rate": 8.094392197840476e-06, "loss": 0.0601, "step": 5770 }, { "epoch": 2.01, "learning_rate": 8.090909090909092e-06, "loss": 0.0592, "step": 5780 }, { "epoch": 2.01, "learning_rate": 8.087425983977709e-06, "loss": 0.0606, "step": 5790 }, { "epoch": 2.01, "learning_rate": 8.083942877046325e-06, "loss": 0.0591, "step": 5800 }, { "epoch": 2.01, "eval_loss": 0.11163330078125, "eval_runtime": 266.5978, "eval_samples_per_second": 10.687, "eval_steps_per_second": 0.086, "eval_wer": 20.740111394317, "step": 5800 }, { "epoch": 2.02, "learning_rate": 8.080459770114944e-06, "loss": 0.0613, "step": 5810 }, { "epoch": 2.02, "learning_rate": 8.07697666318356e-06, "loss": 0.0589, "step": 5820 }, { "epoch": 2.02, "learning_rate": 8.073493556252178e-06, "loss": 0.0585, "step": 5830 }, { "epoch": 2.03, "learning_rate": 8.070010449320795e-06, "loss": 0.0579, "step": 5840 }, { "epoch": 2.03, "learning_rate": 8.066527342389412e-06, "loss": 0.0576, "step": 5850 }, { "epoch": 2.03, "learning_rate": 8.06304423545803e-06, "loss": 0.0569, "step": 5860 }, { "epoch": 2.04, "learning_rate": 8.059561128526646e-06, "loss": 0.0608, "step": 5870 }, { "epoch": 2.04, "learning_rate": 8.056078021595263e-06, "loss": 0.0568, "step": 5880 }, { "epoch": 2.05, "learning_rate": 8.05259491466388e-06, "loss": 0.0599, "step": 5890 }, { "epoch": 2.05, "learning_rate": 8.049111807732498e-06, "loss": 0.062, "step": 5900 }, { "epoch": 2.05, "eval_loss": 0.11236572265625, "eval_runtime": 270.2208, "eval_samples_per_second": 10.543, "eval_steps_per_second": 0.085, "eval_wer": 20.607948645331824, "step": 5900 }, { "epoch": 2.05, "learning_rate": 8.045628700801116e-06, "loss": 0.0565, "step": 5910 }, { "epoch": 2.06, "learning_rate": 8.042145593869733e-06, "loss": 0.0577, "step": 5920 }, { "epoch": 2.06, "learning_rate": 8.03866248693835e-06, "loss": 0.0571, "step": 5930 }, { "epoch": 2.06, "learning_rate": 8.035179380006966e-06, "loss": 0.0601, "step": 5940 }, { "epoch": 2.07, "learning_rate": 8.031696273075584e-06, "loss": 0.0578, "step": 5950 }, { "epoch": 2.07, "learning_rate": 8.0282131661442e-06, "loss": 0.0577, "step": 5960 }, { "epoch": 2.07, "learning_rate": 8.024730059212819e-06, "loss": 0.0581, "step": 5970 }, { "epoch": 2.08, "learning_rate": 8.021246952281436e-06, "loss": 0.0589, "step": 5980 }, { "epoch": 2.08, "learning_rate": 8.017763845350052e-06, "loss": 0.0601, "step": 5990 }, { "epoch": 2.08, "learning_rate": 8.01428073841867e-06, "loss": 0.0601, "step": 6000 }, { "epoch": 2.08, "eval_loss": 0.11273193359375, "eval_runtime": 269.3473, "eval_samples_per_second": 10.577, "eval_steps_per_second": 0.085, "eval_wer": 20.504106485414898, "step": 6000 }, { "epoch": 2.09, "learning_rate": 8.011494252873565e-06, "loss": 0.0576, "step": 6010 }, { "epoch": 2.09, "learning_rate": 8.008011145942181e-06, "loss": 0.0588, "step": 6020 }, { "epoch": 2.09, "learning_rate": 8.004528039010798e-06, "loss": 0.0611, "step": 6030 }, { "epoch": 2.1, "learning_rate": 8.001044932079416e-06, "loss": 0.057, "step": 6040 }, { "epoch": 2.1, "learning_rate": 7.997561825148033e-06, "loss": 0.0577, "step": 6050 }, { "epoch": 2.1, "learning_rate": 7.99407871821665e-06, "loss": 0.0618, "step": 6060 }, { "epoch": 2.11, "learning_rate": 7.990595611285266e-06, "loss": 0.0614, "step": 6070 }, { "epoch": 2.11, "learning_rate": 7.987112504353884e-06, "loss": 0.0615, "step": 6080 }, { "epoch": 2.11, "learning_rate": 7.9836293974225e-06, "loss": 0.0569, "step": 6090 }, { "epoch": 2.12, "learning_rate": 7.980146290491119e-06, "loss": 0.059, "step": 6100 }, { "epoch": 2.12, "eval_loss": 0.11077880859375, "eval_runtime": 272.3283, "eval_samples_per_second": 10.462, "eval_steps_per_second": 0.084, "eval_wer": 20.57333459202618, "step": 6100 }, { "epoch": 2.12, "learning_rate": 7.976663183559736e-06, "loss": 0.057, "step": 6110 }, { "epoch": 2.12, "learning_rate": 7.973180076628352e-06, "loss": 0.0565, "step": 6120 }, { "epoch": 2.13, "learning_rate": 7.96969696969697e-06, "loss": 0.0599, "step": 6130 }, { "epoch": 2.13, "learning_rate": 7.966213862765587e-06, "loss": 0.0564, "step": 6140 }, { "epoch": 2.14, "learning_rate": 7.962730755834205e-06, "loss": 0.0618, "step": 6150 }, { "epoch": 2.14, "learning_rate": 7.959247648902822e-06, "loss": 0.0581, "step": 6160 }, { "epoch": 2.14, "learning_rate": 7.955764541971439e-06, "loss": 0.0564, "step": 6170 }, { "epoch": 2.15, "learning_rate": 7.952281435040057e-06, "loss": 0.0576, "step": 6180 }, { "epoch": 2.15, "learning_rate": 7.948798328108673e-06, "loss": 0.0605, "step": 6190 }, { "epoch": 2.15, "learning_rate": 7.945315221177292e-06, "loss": 0.0594, "step": 6200 }, { "epoch": 2.15, "eval_loss": 0.11126708984375, "eval_runtime": 271.5968, "eval_samples_per_second": 10.49, "eval_steps_per_second": 0.085, "eval_wer": 20.595361716857045, "step": 6200 }, { "epoch": 2.16, "learning_rate": 7.941832114245908e-06, "loss": 0.0595, "step": 6210 }, { "epoch": 2.16, "learning_rate": 7.938349007314525e-06, "loss": 0.0579, "step": 6220 }, { "epoch": 2.16, "learning_rate": 7.934865900383141e-06, "loss": 0.0574, "step": 6230 }, { "epoch": 2.17, "learning_rate": 7.93138279345176e-06, "loss": 0.0549, "step": 6240 }, { "epoch": 2.17, "learning_rate": 7.927899686520378e-06, "loss": 0.0579, "step": 6250 }, { "epoch": 2.17, "learning_rate": 7.924416579588995e-06, "loss": 0.0588, "step": 6260 }, { "epoch": 2.18, "learning_rate": 7.920933472657611e-06, "loss": 0.057, "step": 6270 }, { "epoch": 2.18, "learning_rate": 7.917450365726228e-06, "loss": 0.0593, "step": 6280 }, { "epoch": 2.18, "learning_rate": 7.913967258794846e-06, "loss": 0.0602, "step": 6290 }, { "epoch": 2.19, "learning_rate": 7.910484151863463e-06, "loss": 0.0587, "step": 6300 }, { "epoch": 2.19, "eval_loss": 0.11077880859375, "eval_runtime": 295.0822, "eval_samples_per_second": 9.655, "eval_steps_per_second": 0.078, "eval_wer": 20.532427074483152, "step": 6300 }, { "epoch": 2.19, "learning_rate": 7.90700104493208e-06, "loss": 0.0593, "step": 6310 }, { "epoch": 2.19, "learning_rate": 7.903517938000697e-06, "loss": 0.0583, "step": 6320 }, { "epoch": 2.2, "learning_rate": 7.900034831069314e-06, "loss": 0.0585, "step": 6330 }, { "epoch": 2.2, "learning_rate": 7.896551724137932e-06, "loss": 0.0583, "step": 6340 }, { "epoch": 2.2, "learning_rate": 7.893068617206549e-06, "loss": 0.0567, "step": 6350 }, { "epoch": 2.21, "learning_rate": 7.889585510275165e-06, "loss": 0.0584, "step": 6360 }, { "epoch": 2.21, "learning_rate": 7.886102403343784e-06, "loss": 0.0588, "step": 6370 }, { "epoch": 2.22, "learning_rate": 7.8826192964124e-06, "loss": 0.058, "step": 6380 }, { "epoch": 2.22, "learning_rate": 7.879136189481019e-06, "loss": 0.0577, "step": 6390 }, { "epoch": 2.22, "learning_rate": 7.875653082549635e-06, "loss": 0.0581, "step": 6400 }, { "epoch": 2.22, "eval_loss": 0.112060546875, "eval_runtime": 273.5545, "eval_samples_per_second": 10.415, "eval_steps_per_second": 0.084, "eval_wer": 20.58592152050096, "step": 6400 }, { "epoch": 2.23, "learning_rate": 7.872169975618252e-06, "loss": 0.0556, "step": 6410 }, { "epoch": 2.23, "learning_rate": 7.868686868686868e-06, "loss": 0.0597, "step": 6420 }, { "epoch": 2.23, "learning_rate": 7.865203761755487e-06, "loss": 0.0618, "step": 6430 }, { "epoch": 2.24, "learning_rate": 7.861720654824105e-06, "loss": 0.0567, "step": 6440 }, { "epoch": 2.24, "learning_rate": 7.858237547892721e-06, "loss": 0.0611, "step": 6450 }, { "epoch": 2.24, "learning_rate": 7.854754440961338e-06, "loss": 0.062, "step": 6460 }, { "epoch": 2.25, "learning_rate": 7.851271334029955e-06, "loss": 0.0563, "step": 6470 }, { "epoch": 2.25, "learning_rate": 7.847788227098573e-06, "loss": 0.0594, "step": 6480 }, { "epoch": 2.25, "learning_rate": 7.84430512016719e-06, "loss": 0.06, "step": 6490 }, { "epoch": 2.26, "learning_rate": 7.840822013235808e-06, "loss": 0.0596, "step": 6500 }, { "epoch": 2.26, "eval_loss": 0.1087646484375, "eval_runtime": 268.8291, "eval_samples_per_second": 10.598, "eval_steps_per_second": 0.086, "eval_wer": 20.63941596651877, "step": 6500 }, { "epoch": 2.26, "learning_rate": 7.837338906304424e-06, "loss": 0.061, "step": 6510 }, { "epoch": 2.26, "learning_rate": 7.833855799373041e-06, "loss": 0.0571, "step": 6520 }, { "epoch": 2.27, "learning_rate": 7.830372692441659e-06, "loss": 0.0588, "step": 6530 }, { "epoch": 2.27, "learning_rate": 7.826889585510276e-06, "loss": 0.0571, "step": 6540 }, { "epoch": 2.27, "learning_rate": 7.823406478578894e-06, "loss": 0.0591, "step": 6550 }, { "epoch": 2.28, "learning_rate": 7.81992337164751e-06, "loss": 0.06, "step": 6560 }, { "epoch": 2.28, "learning_rate": 7.816440264716127e-06, "loss": 0.0566, "step": 6570 }, { "epoch": 2.28, "learning_rate": 7.812957157784745e-06, "loss": 0.0595, "step": 6580 }, { "epoch": 2.29, "learning_rate": 7.809474050853362e-06, "loss": 0.0616, "step": 6590 }, { "epoch": 2.29, "learning_rate": 7.805990943921979e-06, "loss": 0.0617, "step": 6600 }, { "epoch": 2.29, "eval_loss": 0.1103515625, "eval_runtime": 274.7709, "eval_samples_per_second": 10.369, "eval_steps_per_second": 0.084, "eval_wer": 20.81877969728437, "step": 6600 }, { "epoch": 2.3, "learning_rate": 7.802507836990595e-06, "loss": 0.058, "step": 6610 }, { "epoch": 2.3, "learning_rate": 7.799024730059213e-06, "loss": 0.0592, "step": 6620 }, { "epoch": 2.3, "learning_rate": 7.79554162312783e-06, "loss": 0.0588, "step": 6630 }, { "epoch": 2.31, "learning_rate": 7.792058516196448e-06, "loss": 0.0548, "step": 6640 }, { "epoch": 2.31, "learning_rate": 7.788575409265065e-06, "loss": 0.0567, "step": 6650 }, { "epoch": 2.31, "learning_rate": 7.785092302333681e-06, "loss": 0.0577, "step": 6660 }, { "epoch": 2.32, "learning_rate": 7.7816091954023e-06, "loss": 0.058, "step": 6670 }, { "epoch": 2.32, "learning_rate": 7.778126088470916e-06, "loss": 0.0584, "step": 6680 }, { "epoch": 2.32, "learning_rate": 7.774642981539535e-06, "loss": 0.0559, "step": 6690 }, { "epoch": 2.33, "learning_rate": 7.771159874608151e-06, "loss": 0.0557, "step": 6700 }, { "epoch": 2.33, "eval_loss": 0.11236572265625, "eval_runtime": 271.2745, "eval_samples_per_second": 10.502, "eval_steps_per_second": 0.085, "eval_wer": 20.557600931432706, "step": 6700 }, { "epoch": 2.33, "learning_rate": 7.767676767676768e-06, "loss": 0.0587, "step": 6710 }, { "epoch": 2.33, "learning_rate": 7.764193660745386e-06, "loss": 0.0599, "step": 6720 }, { "epoch": 2.34, "learning_rate": 7.760710553814003e-06, "loss": 0.0596, "step": 6730 }, { "epoch": 2.34, "learning_rate": 7.757227446882621e-06, "loss": 0.0556, "step": 6740 }, { "epoch": 2.34, "learning_rate": 7.753744339951237e-06, "loss": 0.06, "step": 6750 }, { "epoch": 2.35, "learning_rate": 7.750261233019854e-06, "loss": 0.0574, "step": 6760 }, { "epoch": 2.35, "learning_rate": 7.74677812608847e-06, "loss": 0.0567, "step": 6770 }, { "epoch": 2.35, "learning_rate": 7.743295019157089e-06, "loss": 0.0599, "step": 6780 }, { "epoch": 2.36, "learning_rate": 7.739811912225707e-06, "loss": 0.0562, "step": 6790 }, { "epoch": 2.36, "learning_rate": 7.736328805294324e-06, "loss": 0.0594, "step": 6800 }, { "epoch": 2.36, "eval_loss": 0.10858154296875, "eval_runtime": 272.0479, "eval_samples_per_second": 10.472, "eval_steps_per_second": 0.085, "eval_wer": 20.607948645331824, "step": 6800 }, { "epoch": 2.36, "learning_rate": 7.73284569836294e-06, "loss": 0.0571, "step": 6810 }, { "epoch": 2.37, "learning_rate": 7.729362591431557e-06, "loss": 0.059, "step": 6820 }, { "epoch": 2.37, "learning_rate": 7.725879484500175e-06, "loss": 0.0576, "step": 6830 }, { "epoch": 2.38, "learning_rate": 7.722396377568792e-06, "loss": 0.0599, "step": 6840 }, { "epoch": 2.38, "learning_rate": 7.71891327063741e-06, "loss": 0.056, "step": 6850 }, { "epoch": 2.38, "learning_rate": 7.715430163706027e-06, "loss": 0.0595, "step": 6860 }, { "epoch": 2.39, "learning_rate": 7.711947056774643e-06, "loss": 0.0567, "step": 6870 }, { "epoch": 2.39, "learning_rate": 7.708463949843261e-06, "loss": 0.0581, "step": 6880 }, { "epoch": 2.39, "learning_rate": 7.704980842911878e-06, "loss": 0.059, "step": 6890 }, { "epoch": 2.4, "learning_rate": 7.701497735980495e-06, "loss": 0.0611, "step": 6900 }, { "epoch": 2.4, "eval_loss": 0.11187744140625, "eval_runtime": 270.2, "eval_samples_per_second": 10.544, "eval_steps_per_second": 0.085, "eval_wer": 20.488372824821425, "step": 6900 }, { "epoch": 2.4, "learning_rate": 7.698014629049111e-06, "loss": 0.0633, "step": 6910 }, { "epoch": 2.4, "learning_rate": 7.69453152211773e-06, "loss": 0.0577, "step": 6920 }, { "epoch": 2.41, "learning_rate": 7.691048415186348e-06, "loss": 0.0599, "step": 6930 }, { "epoch": 2.41, "learning_rate": 7.687565308254964e-06, "loss": 0.059, "step": 6940 }, { "epoch": 2.41, "learning_rate": 7.684082201323581e-06, "loss": 0.0576, "step": 6950 }, { "epoch": 2.42, "learning_rate": 7.680599094392198e-06, "loss": 0.0571, "step": 6960 }, { "epoch": 2.42, "learning_rate": 7.677115987460816e-06, "loss": 0.0606, "step": 6970 }, { "epoch": 2.42, "learning_rate": 7.673632880529432e-06, "loss": 0.061, "step": 6980 }, { "epoch": 2.43, "learning_rate": 7.67014977359805e-06, "loss": 0.0603, "step": 6990 }, { "epoch": 2.43, "learning_rate": 7.666666666666667e-06, "loss": 0.0574, "step": 7000 }, { "epoch": 2.43, "eval_loss": 0.112060546875, "eval_runtime": 270.472, "eval_samples_per_second": 10.533, "eval_steps_per_second": 0.085, "eval_wer": 20.516693413889676, "step": 7000 }, { "epoch": 2.43, "learning_rate": 7.663183559735284e-06, "loss": 0.0595, "step": 7010 }, { "epoch": 2.44, "learning_rate": 7.660397074190178e-06, "loss": 0.0587, "step": 7020 }, { "epoch": 2.44, "learning_rate": 7.656913967258796e-06, "loss": 0.0575, "step": 7030 }, { "epoch": 2.44, "learning_rate": 7.653430860327413e-06, "loss": 0.0591, "step": 7040 }, { "epoch": 2.45, "learning_rate": 7.64994775339603e-06, "loss": 0.0569, "step": 7050 }, { "epoch": 2.45, "learning_rate": 7.646464646464648e-06, "loss": 0.0564, "step": 7060 }, { "epoch": 2.45, "learning_rate": 7.642981539533264e-06, "loss": 0.0626, "step": 7070 }, { "epoch": 2.46, "learning_rate": 7.639498432601881e-06, "loss": 0.0585, "step": 7080 }, { "epoch": 2.46, "learning_rate": 7.636015325670498e-06, "loss": 0.061, "step": 7090 }, { "epoch": 2.47, "learning_rate": 7.632532218739116e-06, "loss": 0.0593, "step": 7100 }, { "epoch": 2.47, "eval_loss": 0.1129150390625, "eval_runtime": 324.392, "eval_samples_per_second": 8.783, "eval_steps_per_second": 0.071, "eval_wer": 20.45375877151578, "step": 7100 }, { "epoch": 2.47, "learning_rate": 7.629049111807733e-06, "loss": 0.0569, "step": 7110 }, { "epoch": 2.47, "learning_rate": 7.62556600487635e-06, "loss": 0.0579, "step": 7120 }, { "epoch": 2.48, "learning_rate": 7.622082897944968e-06, "loss": 0.0567, "step": 7130 }, { "epoch": 2.48, "learning_rate": 7.618599791013585e-06, "loss": 0.0627, "step": 7140 }, { "epoch": 2.48, "learning_rate": 7.615116684082202e-06, "loss": 0.0583, "step": 7150 }, { "epoch": 2.49, "learning_rate": 7.611633577150819e-06, "loss": 0.0567, "step": 7160 }, { "epoch": 2.49, "learning_rate": 7.608150470219436e-06, "loss": 0.0617, "step": 7170 }, { "epoch": 2.49, "learning_rate": 7.604667363288054e-06, "loss": 0.0585, "step": 7180 }, { "epoch": 2.5, "learning_rate": 7.601184256356671e-06, "loss": 0.0588, "step": 7190 }, { "epoch": 2.5, "learning_rate": 7.597701149425288e-06, "loss": 0.0613, "step": 7200 }, { "epoch": 2.5, "eval_loss": 0.1126708984375, "eval_runtime": 273.1528, "eval_samples_per_second": 10.43, "eval_steps_per_second": 0.084, "eval_wer": 20.582774788382267, "step": 7200 }, { "epoch": 2.5, "learning_rate": 7.594218042493905e-06, "loss": 0.0577, "step": 7210 }, { "epoch": 2.51, "learning_rate": 7.590734935562522e-06, "loss": 0.056, "step": 7220 }, { "epoch": 2.51, "learning_rate": 7.587251828631139e-06, "loss": 0.0582, "step": 7230 }, { "epoch": 2.51, "learning_rate": 7.583768721699757e-06, "loss": 0.0578, "step": 7240 }, { "epoch": 2.52, "learning_rate": 7.580285614768375e-06, "loss": 0.06, "step": 7250 }, { "epoch": 2.52, "learning_rate": 7.576802507836991e-06, "loss": 0.059, "step": 7260 }, { "epoch": 2.52, "learning_rate": 7.573319400905609e-06, "loss": 0.0586, "step": 7270 }, { "epoch": 2.53, "learning_rate": 7.569836293974225e-06, "loss": 0.0555, "step": 7280 }, { "epoch": 2.53, "learning_rate": 7.566353187042843e-06, "loss": 0.0572, "step": 7290 }, { "epoch": 2.53, "learning_rate": 7.562870080111459e-06, "loss": 0.0594, "step": 7300 }, { "epoch": 2.53, "eval_loss": 0.11083984375, "eval_runtime": 271.7249, "eval_samples_per_second": 10.485, "eval_steps_per_second": 0.085, "eval_wer": 20.488372824821425, "step": 7300 }, { "epoch": 2.54, "learning_rate": 7.559386973180078e-06, "loss": 0.059, "step": 7310 }, { "epoch": 2.54, "learning_rate": 7.555903866248695e-06, "loss": 0.0571, "step": 7320 }, { "epoch": 2.55, "learning_rate": 7.552420759317312e-06, "loss": 0.0573, "step": 7330 }, { "epoch": 2.55, "learning_rate": 7.548937652385929e-06, "loss": 0.0599, "step": 7340 }, { "epoch": 2.55, "learning_rate": 7.545454545454546e-06, "loss": 0.0536, "step": 7350 }, { "epoch": 2.56, "learning_rate": 7.541971438523164e-06, "loss": 0.0589, "step": 7360 }, { "epoch": 2.56, "learning_rate": 7.5384883315917805e-06, "loss": 0.0579, "step": 7370 }, { "epoch": 2.56, "learning_rate": 7.535005224660398e-06, "loss": 0.0585, "step": 7380 }, { "epoch": 2.57, "learning_rate": 7.531522117729015e-06, "loss": 0.0595, "step": 7390 }, { "epoch": 2.57, "learning_rate": 7.528039010797632e-06, "loss": 0.0603, "step": 7400 }, { "epoch": 2.57, "eval_loss": 0.11322021484375, "eval_runtime": 268.4354, "eval_samples_per_second": 10.613, "eval_steps_per_second": 0.086, "eval_wer": 20.66458982346833, "step": 7400 }, { "epoch": 2.57, "learning_rate": 7.524555903866249e-06, "loss": 0.0592, "step": 7410 }, { "epoch": 2.58, "learning_rate": 7.521072796934866e-06, "loss": 0.0595, "step": 7420 }, { "epoch": 2.58, "learning_rate": 7.517589690003484e-06, "loss": 0.0585, "step": 7430 }, { "epoch": 2.58, "learning_rate": 7.514106583072101e-06, "loss": 0.0576, "step": 7440 }, { "epoch": 2.59, "learning_rate": 7.510623476140718e-06, "loss": 0.0582, "step": 7450 }, { "epoch": 2.59, "learning_rate": 7.507140369209336e-06, "loss": 0.0614, "step": 7460 }, { "epoch": 2.59, "learning_rate": 7.503657262277952e-06, "loss": 0.0565, "step": 7470 }, { "epoch": 2.6, "learning_rate": 7.5001741553465705e-06, "loss": 0.0626, "step": 7480 }, { "epoch": 2.6, "learning_rate": 7.496691048415187e-06, "loss": 0.0587, "step": 7490 }, { "epoch": 2.6, "learning_rate": 7.4932079414838045e-06, "loss": 0.0602, "step": 7500 }, { "epoch": 2.6, "eval_loss": 0.10906982421875, "eval_runtime": 271.3251, "eval_samples_per_second": 10.5, "eval_steps_per_second": 0.085, "eval_wer": 20.27124830863149, "step": 7500 }, { "epoch": 2.61, "learning_rate": 7.489724834552421e-06, "loss": 0.0595, "step": 7510 }, { "epoch": 2.61, "learning_rate": 7.4862417276210385e-06, "loss": 0.0566, "step": 7520 }, { "epoch": 2.61, "learning_rate": 7.482758620689656e-06, "loss": 0.0584, "step": 7530 }, { "epoch": 2.62, "learning_rate": 7.4792755137582725e-06, "loss": 0.0549, "step": 7540 }, { "epoch": 2.62, "learning_rate": 7.475792406826891e-06, "loss": 0.0598, "step": 7550 }, { "epoch": 2.62, "learning_rate": 7.472309299895507e-06, "loss": 0.0571, "step": 7560 }, { "epoch": 2.63, "learning_rate": 7.468826192964125e-06, "loss": 0.0588, "step": 7570 }, { "epoch": 2.63, "learning_rate": 7.465343086032741e-06, "loss": 0.0558, "step": 7580 }, { "epoch": 2.64, "learning_rate": 7.461859979101359e-06, "loss": 0.059, "step": 7590 }, { "epoch": 2.64, "learning_rate": 7.458376872169977e-06, "loss": 0.0607, "step": 7600 }, { "epoch": 2.64, "eval_loss": 0.10968017578125, "eval_runtime": 271.2796, "eval_samples_per_second": 10.502, "eval_steps_per_second": 0.085, "eval_wer": 20.35306334371755, "step": 7600 }, { "epoch": 2.64, "learning_rate": 7.454893765238594e-06, "loss": 0.0585, "step": 7610 }, { "epoch": 2.65, "learning_rate": 7.451410658307211e-06, "loss": 0.0593, "step": 7620 }, { "epoch": 2.65, "learning_rate": 7.447927551375828e-06, "loss": 0.0596, "step": 7630 }, { "epoch": 2.65, "learning_rate": 7.444444444444445e-06, "loss": 0.0575, "step": 7640 }, { "epoch": 2.66, "learning_rate": 7.440961337513062e-06, "loss": 0.0605, "step": 7650 }, { "epoch": 2.66, "learning_rate": 7.437478230581679e-06, "loss": 0.0611, "step": 7660 }, { "epoch": 2.66, "learning_rate": 7.433995123650297e-06, "loss": 0.0554, "step": 7670 }, { "epoch": 2.67, "learning_rate": 7.430512016718914e-06, "loss": 0.0548, "step": 7680 }, { "epoch": 2.67, "learning_rate": 7.427028909787531e-06, "loss": 0.0589, "step": 7690 }, { "epoch": 2.67, "learning_rate": 7.423545802856148e-06, "loss": 0.0567, "step": 7700 }, { "epoch": 2.67, "eval_loss": 0.10833740234375, "eval_runtime": 276.9395, "eval_samples_per_second": 10.287, "eval_steps_per_second": 0.083, "eval_wer": 20.321596022530603, "step": 7700 }, { "epoch": 2.68, "learning_rate": 7.420062695924765e-06, "loss": 0.0582, "step": 7710 }, { "epoch": 2.68, "learning_rate": 7.416579588993382e-06, "loss": 0.0593, "step": 7720 }, { "epoch": 2.68, "learning_rate": 7.413096482062e-06, "loss": 0.0569, "step": 7730 }, { "epoch": 2.69, "learning_rate": 7.409613375130618e-06, "loss": 0.0613, "step": 7740 }, { "epoch": 2.69, "learning_rate": 7.406130268199234e-06, "loss": 0.0568, "step": 7750 }, { "epoch": 2.69, "learning_rate": 7.402647161267852e-06, "loss": 0.0568, "step": 7760 }, { "epoch": 2.7, "learning_rate": 7.399164054336468e-06, "loss": 0.0591, "step": 7770 }, { "epoch": 2.7, "learning_rate": 7.3956809474050865e-06, "loss": 0.0606, "step": 7780 }, { "epoch": 2.7, "learning_rate": 7.392197840473703e-06, "loss": 0.0609, "step": 7790 }, { "epoch": 2.71, "learning_rate": 7.3887147335423205e-06, "loss": 0.0572, "step": 7800 }, { "epoch": 2.71, "eval_loss": 0.10943603515625, "eval_runtime": 273.1967, "eval_samples_per_second": 10.428, "eval_steps_per_second": 0.084, "eval_wer": 20.548160735076625, "step": 7800 }, { "epoch": 2.71, "learning_rate": 7.385231626610938e-06, "loss": 0.0596, "step": 7810 }, { "epoch": 2.72, "learning_rate": 7.3817485196795545e-06, "loss": 0.0582, "step": 7820 }, { "epoch": 2.72, "learning_rate": 7.378265412748172e-06, "loss": 0.0581, "step": 7830 }, { "epoch": 2.72, "learning_rate": 7.3747823058167885e-06, "loss": 0.059, "step": 7840 }, { "epoch": 2.73, "learning_rate": 7.371299198885407e-06, "loss": 0.0579, "step": 7850 }, { "epoch": 2.73, "learning_rate": 7.367816091954023e-06, "loss": 0.0595, "step": 7860 }, { "epoch": 2.73, "learning_rate": 7.364332985022641e-06, "loss": 0.0604, "step": 7870 }, { "epoch": 2.74, "learning_rate": 7.360849878091258e-06, "loss": 0.0601, "step": 7880 }, { "epoch": 2.74, "learning_rate": 7.357366771159875e-06, "loss": 0.0635, "step": 7890 }, { "epoch": 2.74, "learning_rate": 7.353883664228493e-06, "loss": 0.0563, "step": 7900 }, { "epoch": 2.74, "eval_loss": 0.10894775390625, "eval_runtime": 275.1597, "eval_samples_per_second": 10.354, "eval_steps_per_second": 0.084, "eval_wer": 20.541867270839234, "step": 7900 }, { "epoch": 2.75, "learning_rate": 7.35040055729711e-06, "loss": 0.0593, "step": 7910 }, { "epoch": 2.75, "learning_rate": 7.346917450365727e-06, "loss": 0.0577, "step": 7920 }, { "epoch": 2.75, "learning_rate": 7.343434343434344e-06, "loss": 0.0576, "step": 7930 }, { "epoch": 2.76, "learning_rate": 7.339951236502961e-06, "loss": 0.0615, "step": 7940 }, { "epoch": 2.76, "learning_rate": 7.3364681295715786e-06, "loss": 0.0583, "step": 7950 }, { "epoch": 2.76, "learning_rate": 7.332985022640195e-06, "loss": 0.059, "step": 7960 }, { "epoch": 2.77, "learning_rate": 7.329501915708813e-06, "loss": 0.0591, "step": 7970 }, { "epoch": 2.77, "learning_rate": 7.32601880877743e-06, "loss": 0.0573, "step": 7980 }, { "epoch": 2.77, "learning_rate": 7.322535701846047e-06, "loss": 0.0613, "step": 7990 }, { "epoch": 2.78, "learning_rate": 7.319052594914664e-06, "loss": 0.0581, "step": 8000 }, { "epoch": 2.78, "eval_loss": 0.10784912109375, "eval_runtime": 272.0815, "eval_samples_per_second": 10.471, "eval_steps_per_second": 0.085, "eval_wer": 20.16111268447717, "step": 8000 }, { "epoch": 2.78, "learning_rate": 7.3155694879832814e-06, "loss": 0.0567, "step": 8010 }, { "epoch": 2.78, "learning_rate": 7.312783002438175e-06, "loss": 0.061, "step": 8020 }, { "epoch": 2.79, "learning_rate": 7.309299895506793e-06, "loss": 0.0588, "step": 8030 }, { "epoch": 2.79, "learning_rate": 7.30581678857541e-06, "loss": 0.0577, "step": 8040 }, { "epoch": 2.8, "learning_rate": 7.302333681644027e-06, "loss": 0.0586, "step": 8050 }, { "epoch": 2.8, "learning_rate": 7.298850574712645e-06, "loss": 0.0586, "step": 8060 }, { "epoch": 2.8, "learning_rate": 7.295367467781261e-06, "loss": 0.0598, "step": 8070 }, { "epoch": 2.81, "learning_rate": 7.2918843608498794e-06, "loss": 0.0558, "step": 8080 }, { "epoch": 2.81, "learning_rate": 7.288401253918496e-06, "loss": 0.057, "step": 8090 }, { "epoch": 2.81, "learning_rate": 7.2849181469871134e-06, "loss": 0.0582, "step": 8100 }, { "epoch": 2.81, "eval_loss": 0.11016845703125, "eval_runtime": 270.4011, "eval_samples_per_second": 10.536, "eval_steps_per_second": 0.085, "eval_wer": 20.321596022530603, "step": 8100 }, { "epoch": 2.82, "learning_rate": 7.28143504005573e-06, "loss": 0.057, "step": 8110 }, { "epoch": 2.82, "learning_rate": 7.2779519331243474e-06, "loss": 0.0579, "step": 8120 }, { "epoch": 2.82, "learning_rate": 7.274468826192965e-06, "loss": 0.0567, "step": 8130 }, { "epoch": 2.83, "learning_rate": 7.2709857192615815e-06, "loss": 0.057, "step": 8140 }, { "epoch": 2.83, "learning_rate": 7.2675026123302e-06, "loss": 0.0628, "step": 8150 }, { "epoch": 2.83, "learning_rate": 7.264019505398816e-06, "loss": 0.0574, "step": 8160 }, { "epoch": 2.84, "learning_rate": 7.260536398467434e-06, "loss": 0.0575, "step": 8170 }, { "epoch": 2.84, "learning_rate": 7.25705329153605e-06, "loss": 0.0574, "step": 8180 }, { "epoch": 2.84, "learning_rate": 7.253570184604668e-06, "loss": 0.0585, "step": 8190 }, { "epoch": 2.85, "learning_rate": 7.250087077673286e-06, "loss": 0.0571, "step": 8200 }, { "epoch": 2.85, "eval_loss": 0.11126708984375, "eval_runtime": 271.8988, "eval_samples_per_second": 10.478, "eval_steps_per_second": 0.085, "eval_wer": 20.044683596085466, "step": 8200 } ], "max_steps": 29000, "num_train_epochs": 11, "total_flos": 2.142217063719011e+21, "trial_name": null, "trial_params": null }