{ "best_metric": 10.134191031082755, "best_model_checkpoint": "whisper-medium-hi_alldata_dropout_0.05_multigpu/checkpoint-13400", "epoch": 10.218978102189782, "global_step": 21000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.01, "learning_rate": 2.385606273598312e-06, "loss": 4.2329, "step": 10 }, { "epoch": 0.01, "learning_rate": 5.5697167615341825e-06, "loss": 1.6441, "step": 20 }, { "epoch": 0.02, "learning_rate": 6.808639180087963e-06, "loss": 0.8689, "step": 30 }, { "epoch": 0.02, "learning_rate": 7.592569699389437e-06, "loss": 0.6238, "step": 40 }, { "epoch": 0.03, "learning_rate": 8.167342277897933e-06, "loss": 0.5, "step": 50 }, { "epoch": 0.03, "learning_rate": 8.621379348003945e-06, "loss": 0.4275, "step": 60 }, { "epoch": 0.04, "learning_rate": 8.996702747267908e-06, "loss": 0.3778, "step": 70 }, { "epoch": 0.04, "learning_rate": 9.316614300602277e-06, "loss": 0.3369, "step": 80 }, { "epoch": 0.05, "learning_rate": 9.59539046188037e-06, "loss": 0.3139, "step": 90 }, { "epoch": 0.05, "learning_rate": 9.842414742769675e-06, "loss": 0.2969, "step": 100 }, { "epoch": 0.05, "eval_loss": 0.24853515625, "eval_runtime": 160.4678, "eval_samples_per_second": 20.122, "eval_steps_per_second": 0.081, "eval_wer": 26.727478843162583, "step": 100 }, { "epoch": 0.06, "learning_rate": 9.998994974874374e-06, "loss": 0.2826, "step": 110 }, { "epoch": 0.06, "learning_rate": 9.993969849246232e-06, "loss": 0.2721, "step": 120 }, { "epoch": 0.07, "learning_rate": 9.988944723618091e-06, "loss": 0.2556, "step": 130 }, { "epoch": 0.07, "learning_rate": 9.983919597989951e-06, "loss": 0.2493, "step": 140 }, { "epoch": 0.08, "learning_rate": 9.97889447236181e-06, "loss": 0.2429, "step": 150 }, { "epoch": 0.08, "learning_rate": 9.973869346733668e-06, "loss": 0.2304, "step": 160 }, { "epoch": 0.09, "learning_rate": 9.968844221105529e-06, "loss": 0.2281, "step": 170 }, { "epoch": 0.09, "learning_rate": 9.963819095477387e-06, "loss": 0.216, "step": 180 }, { "epoch": 0.1, "learning_rate": 9.958793969849248e-06, "loss": 0.2159, "step": 190 }, { "epoch": 0.1, "learning_rate": 9.953768844221106e-06, "loss": 0.2082, "step": 200 }, { "epoch": 0.1, "eval_loss": 0.196533203125, "eval_runtime": 160.8342, "eval_samples_per_second": 20.077, "eval_steps_per_second": 0.081, "eval_wer": 21.684110621918748, "step": 200 }, { "epoch": 0.11, "learning_rate": 9.948743718592965e-06, "loss": 0.2014, "step": 210 }, { "epoch": 0.12, "learning_rate": 9.943718592964825e-06, "loss": 0.1998, "step": 220 }, { "epoch": 0.12, "learning_rate": 9.938693467336684e-06, "loss": 0.1964, "step": 230 }, { "epoch": 0.13, "learning_rate": 9.933668341708543e-06, "loss": 0.1936, "step": 240 }, { "epoch": 0.13, "learning_rate": 9.928643216080403e-06, "loss": 0.1866, "step": 250 }, { "epoch": 0.14, "learning_rate": 9.923618090452262e-06, "loss": 0.1887, "step": 260 }, { "epoch": 0.14, "learning_rate": 9.918592964824122e-06, "loss": 0.1801, "step": 270 }, { "epoch": 0.15, "learning_rate": 9.91356783919598e-06, "loss": 0.1808, "step": 280 }, { "epoch": 0.15, "learning_rate": 9.908542713567841e-06, "loss": 0.1776, "step": 290 }, { "epoch": 0.16, "learning_rate": 9.9035175879397e-06, "loss": 0.1715, "step": 300 }, { "epoch": 0.16, "eval_loss": 0.176513671875, "eval_runtime": 167.3804, "eval_samples_per_second": 19.291, "eval_steps_per_second": 0.078, "eval_wer": 19.3983812712557, "step": 300 }, { "epoch": 0.16, "learning_rate": 9.898492462311558e-06, "loss": 0.1783, "step": 310 }, { "epoch": 0.17, "learning_rate": 9.893467336683417e-06, "loss": 0.1726, "step": 320 }, { "epoch": 0.17, "learning_rate": 9.888442211055277e-06, "loss": 0.1702, "step": 330 }, { "epoch": 0.18, "learning_rate": 9.883417085427136e-06, "loss": 0.1666, "step": 340 }, { "epoch": 0.18, "learning_rate": 9.878391959798995e-06, "loss": 0.17, "step": 350 }, { "epoch": 0.19, "learning_rate": 9.873366834170855e-06, "loss": 0.1653, "step": 360 }, { "epoch": 0.19, "learning_rate": 9.868341708542715e-06, "loss": 0.1611, "step": 370 }, { "epoch": 0.2, "learning_rate": 9.863316582914574e-06, "loss": 0.1612, "step": 380 }, { "epoch": 0.2, "learning_rate": 9.858291457286433e-06, "loss": 0.1604, "step": 390 }, { "epoch": 0.21, "learning_rate": 9.853266331658293e-06, "loss": 0.1605, "step": 400 }, { "epoch": 0.21, "eval_loss": 0.165283203125, "eval_runtime": 170.7476, "eval_samples_per_second": 18.911, "eval_steps_per_second": 0.076, "eval_wer": 18.101289182990165, "step": 400 }, { "epoch": 0.22, "learning_rate": 9.848241206030152e-06, "loss": 0.1535, "step": 410 }, { "epoch": 0.22, "learning_rate": 9.84321608040201e-06, "loss": 0.1579, "step": 420 }, { "epoch": 0.23, "learning_rate": 9.838190954773869e-06, "loss": 0.1509, "step": 430 }, { "epoch": 0.23, "learning_rate": 9.83316582914573e-06, "loss": 0.1477, "step": 440 }, { "epoch": 0.24, "learning_rate": 9.82814070351759e-06, "loss": 0.1558, "step": 450 }, { "epoch": 0.24, "learning_rate": 9.823115577889448e-06, "loss": 0.1528, "step": 460 }, { "epoch": 0.25, "learning_rate": 9.818090452261307e-06, "loss": 0.1476, "step": 470 }, { "epoch": 0.25, "learning_rate": 9.813065326633167e-06, "loss": 0.1492, "step": 480 }, { "epoch": 0.26, "learning_rate": 9.808040201005026e-06, "loss": 0.1414, "step": 490 }, { "epoch": 0.26, "learning_rate": 9.803015075376884e-06, "loss": 0.1445, "step": 500 }, { "epoch": 0.26, "eval_loss": 0.15625, "eval_runtime": 158.1436, "eval_samples_per_second": 20.418, "eval_steps_per_second": 0.082, "eval_wer": 16.943924494476814, "step": 500 }, { "epoch": 0.27, "learning_rate": 9.797989949748745e-06, "loss": 0.1464, "step": 510 }, { "epoch": 0.27, "learning_rate": 9.792964824120605e-06, "loss": 0.1414, "step": 520 }, { "epoch": 0.28, "learning_rate": 9.787939698492464e-06, "loss": 0.1411, "step": 530 }, { "epoch": 0.28, "learning_rate": 9.782914572864322e-06, "loss": 0.1407, "step": 540 }, { "epoch": 0.29, "learning_rate": 9.777889447236181e-06, "loss": 0.1463, "step": 550 }, { "epoch": 0.29, "learning_rate": 9.772864321608041e-06, "loss": 0.1418, "step": 560 }, { "epoch": 0.3, "learning_rate": 9.7678391959799e-06, "loss": 0.1386, "step": 570 }, { "epoch": 0.3, "learning_rate": 9.762814070351759e-06, "loss": 0.1381, "step": 580 }, { "epoch": 0.31, "learning_rate": 9.757788944723619e-06, "loss": 0.1346, "step": 590 }, { "epoch": 0.31, "learning_rate": 9.752763819095478e-06, "loss": 0.1354, "step": 600 }, { "epoch": 0.31, "eval_loss": 0.1500244140625, "eval_runtime": 156.5238, "eval_samples_per_second": 20.629, "eval_steps_per_second": 0.083, "eval_wer": 16.253196593815087, "step": 600 }, { "epoch": 0.32, "learning_rate": 9.747738693467338e-06, "loss": 0.1322, "step": 610 }, { "epoch": 0.33, "learning_rate": 9.742713567839197e-06, "loss": 0.1344, "step": 620 }, { "epoch": 0.33, "learning_rate": 9.737688442211057e-06, "loss": 0.1358, "step": 630 }, { "epoch": 0.34, "learning_rate": 9.732663316582916e-06, "loss": 0.1322, "step": 640 }, { "epoch": 0.34, "learning_rate": 9.727638190954774e-06, "loss": 0.1309, "step": 650 }, { "epoch": 0.35, "learning_rate": 9.722613065326633e-06, "loss": 0.1386, "step": 660 }, { "epoch": 0.35, "learning_rate": 9.717587939698493e-06, "loss": 0.1335, "step": 670 }, { "epoch": 0.36, "learning_rate": 9.712562814070352e-06, "loss": 0.1371, "step": 680 }, { "epoch": 0.36, "learning_rate": 9.707537688442212e-06, "loss": 0.1311, "step": 690 }, { "epoch": 0.37, "learning_rate": 9.702512562814071e-06, "loss": 0.1302, "step": 700 }, { "epoch": 0.37, "eval_loss": 0.1446533203125, "eval_runtime": 158.1947, "eval_samples_per_second": 20.412, "eval_steps_per_second": 0.082, "eval_wer": 15.89201444728586, "step": 700 }, { "epoch": 0.37, "learning_rate": 9.697487437185931e-06, "loss": 0.1279, "step": 710 }, { "epoch": 0.38, "learning_rate": 9.69246231155779e-06, "loss": 0.132, "step": 720 }, { "epoch": 0.38, "learning_rate": 9.687437185929649e-06, "loss": 0.1292, "step": 730 }, { "epoch": 0.39, "learning_rate": 9.682412060301509e-06, "loss": 0.1284, "step": 740 }, { "epoch": 0.39, "learning_rate": 9.677386934673368e-06, "loss": 0.1304, "step": 750 }, { "epoch": 0.4, "learning_rate": 9.672361809045226e-06, "loss": 0.1262, "step": 760 }, { "epoch": 0.4, "learning_rate": 9.667336683417087e-06, "loss": 0.1283, "step": 770 }, { "epoch": 0.41, "learning_rate": 9.662311557788945e-06, "loss": 0.1216, "step": 780 }, { "epoch": 0.41, "learning_rate": 9.657286432160806e-06, "loss": 0.1269, "step": 790 }, { "epoch": 0.42, "learning_rate": 9.652261306532664e-06, "loss": 0.1246, "step": 800 }, { "epoch": 0.42, "eval_loss": 0.143310546875, "eval_runtime": 169.7265, "eval_samples_per_second": 19.025, "eval_steps_per_second": 0.077, "eval_wer": 15.391104901004454, "step": 800 }, { "epoch": 0.42, "learning_rate": 9.647236180904523e-06, "loss": 0.1205, "step": 810 }, { "epoch": 0.43, "learning_rate": 9.642211055276383e-06, "loss": 0.1228, "step": 820 }, { "epoch": 0.44, "learning_rate": 9.637185929648242e-06, "loss": 0.1248, "step": 830 }, { "epoch": 0.44, "learning_rate": 9.6321608040201e-06, "loss": 0.1237, "step": 840 }, { "epoch": 0.45, "learning_rate": 9.62713567839196e-06, "loss": 0.1228, "step": 850 }, { "epoch": 0.45, "learning_rate": 9.62211055276382e-06, "loss": 0.1272, "step": 860 }, { "epoch": 0.46, "learning_rate": 9.61708542713568e-06, "loss": 0.1244, "step": 870 }, { "epoch": 0.46, "learning_rate": 9.612060301507538e-06, "loss": 0.1237, "step": 880 }, { "epoch": 0.47, "learning_rate": 9.607035175879397e-06, "loss": 0.1212, "step": 890 }, { "epoch": 0.47, "learning_rate": 9.602010050251257e-06, "loss": 0.1209, "step": 900 }, { "epoch": 0.47, "eval_loss": 0.1376953125, "eval_runtime": 169.4394, "eval_samples_per_second": 19.057, "eval_steps_per_second": 0.077, "eval_wer": 15.098468271334792, "step": 900 }, { "epoch": 0.48, "learning_rate": 9.596984924623116e-06, "loss": 0.12, "step": 910 }, { "epoch": 0.48, "learning_rate": 9.591959798994975e-06, "loss": 0.1249, "step": 920 }, { "epoch": 0.49, "learning_rate": 9.586934673366835e-06, "loss": 0.1215, "step": 930 }, { "epoch": 0.49, "learning_rate": 9.581909547738695e-06, "loss": 0.1196, "step": 940 }, { "epoch": 0.5, "learning_rate": 9.576884422110554e-06, "loss": 0.12, "step": 950 }, { "epoch": 0.5, "learning_rate": 9.571859296482413e-06, "loss": 0.1194, "step": 960 }, { "epoch": 0.51, "learning_rate": 9.566834170854273e-06, "loss": 0.1181, "step": 970 }, { "epoch": 0.51, "learning_rate": 9.561809045226132e-06, "loss": 0.1182, "step": 980 }, { "epoch": 0.52, "learning_rate": 9.55678391959799e-06, "loss": 0.1158, "step": 990 }, { "epoch": 0.52, "learning_rate": 9.551758793969849e-06, "loss": 0.115, "step": 1000 }, { "epoch": 0.52, "eval_loss": 0.1353759765625, "eval_runtime": 156.6489, "eval_samples_per_second": 20.613, "eval_steps_per_second": 0.083, "eval_wer": 14.97983179984709, "step": 1000 }, { "epoch": 0.53, "learning_rate": 9.54673366834171e-06, "loss": 0.1205, "step": 1010 }, { "epoch": 0.54, "learning_rate": 9.54170854271357e-06, "loss": 0.1198, "step": 1020 }, { "epoch": 0.54, "learning_rate": 9.536683417085428e-06, "loss": 0.1178, "step": 1030 }, { "epoch": 0.55, "learning_rate": 9.531658291457287e-06, "loss": 0.1163, "step": 1040 }, { "epoch": 0.55, "learning_rate": 9.526633165829147e-06, "loss": 0.1169, "step": 1050 }, { "epoch": 0.56, "learning_rate": 9.521608040201006e-06, "loss": 0.1145, "step": 1060 }, { "epoch": 0.56, "learning_rate": 9.516582914572865e-06, "loss": 0.1173, "step": 1070 }, { "epoch": 0.57, "learning_rate": 9.511557788944725e-06, "loss": 0.1139, "step": 1080 }, { "epoch": 0.57, "learning_rate": 9.506532663316584e-06, "loss": 0.114, "step": 1090 }, { "epoch": 0.58, "learning_rate": 9.501507537688442e-06, "loss": 0.1114, "step": 1100 }, { "epoch": 0.58, "eval_loss": 0.130615234375, "eval_runtime": 158.8351, "eval_samples_per_second": 20.329, "eval_steps_per_second": 0.082, "eval_wer": 14.468376789433446, "step": 1100 }, { "epoch": 0.58, "learning_rate": 9.496482412060303e-06, "loss": 0.1138, "step": 1110 }, { "epoch": 0.59, "learning_rate": 9.491457286432161e-06, "loss": 0.112, "step": 1120 }, { "epoch": 0.59, "learning_rate": 9.486432160804022e-06, "loss": 0.1138, "step": 1130 }, { "epoch": 0.6, "learning_rate": 9.48140703517588e-06, "loss": 0.1128, "step": 1140 }, { "epoch": 0.6, "learning_rate": 9.476381909547739e-06, "loss": 0.1118, "step": 1150 }, { "epoch": 0.61, "learning_rate": 9.4713567839196e-06, "loss": 0.1115, "step": 1160 }, { "epoch": 0.61, "learning_rate": 9.466331658291458e-06, "loss": 0.1149, "step": 1170 }, { "epoch": 0.62, "learning_rate": 9.461306532663317e-06, "loss": 0.1114, "step": 1180 }, { "epoch": 0.62, "learning_rate": 9.456281407035177e-06, "loss": 0.1113, "step": 1190 }, { "epoch": 0.63, "learning_rate": 9.451256281407035e-06, "loss": 0.1098, "step": 1200 }, { "epoch": 0.63, "eval_loss": 0.12939453125, "eval_runtime": 152.9256, "eval_samples_per_second": 21.115, "eval_steps_per_second": 0.085, "eval_wer": 14.265376604887821, "step": 1200 }, { "epoch": 0.63, "learning_rate": 9.446231155778896e-06, "loss": 0.1078, "step": 1210 }, { "epoch": 0.64, "learning_rate": 9.441206030150754e-06, "loss": 0.1103, "step": 1220 }, { "epoch": 0.65, "learning_rate": 9.436180904522613e-06, "loss": 0.1112, "step": 1230 }, { "epoch": 0.65, "learning_rate": 9.431155778894473e-06, "loss": 0.1112, "step": 1240 }, { "epoch": 0.66, "learning_rate": 9.426130653266332e-06, "loss": 0.1091, "step": 1250 }, { "epoch": 0.66, "learning_rate": 9.42110552763819e-06, "loss": 0.1119, "step": 1260 }, { "epoch": 0.67, "learning_rate": 9.416080402010051e-06, "loss": 0.1111, "step": 1270 }, { "epoch": 0.67, "learning_rate": 9.411055276381911e-06, "loss": 0.1097, "step": 1280 }, { "epoch": 0.68, "learning_rate": 9.40603015075377e-06, "loss": 0.1127, "step": 1290 }, { "epoch": 0.68, "learning_rate": 9.401005025125629e-06, "loss": 0.109, "step": 1300 }, { "epoch": 0.68, "eval_loss": 0.1268310546875, "eval_runtime": 152.6721, "eval_samples_per_second": 21.15, "eval_steps_per_second": 0.085, "eval_wer": 14.007012733647938, "step": 1300 }, { "epoch": 0.69, "learning_rate": 9.395979899497489e-06, "loss": 0.1062, "step": 1310 }, { "epoch": 0.69, "learning_rate": 9.390954773869348e-06, "loss": 0.1077, "step": 1320 }, { "epoch": 0.7, "learning_rate": 9.385929648241206e-06, "loss": 0.1113, "step": 1330 }, { "epoch": 0.7, "learning_rate": 9.380904522613065e-06, "loss": 0.1057, "step": 1340 }, { "epoch": 0.71, "learning_rate": 9.375879396984925e-06, "loss": 0.1095, "step": 1350 }, { "epoch": 0.71, "learning_rate": 9.370854271356786e-06, "loss": 0.1067, "step": 1360 }, { "epoch": 0.72, "learning_rate": 9.365829145728644e-06, "loss": 0.1047, "step": 1370 }, { "epoch": 0.72, "learning_rate": 9.360804020100503e-06, "loss": 0.1064, "step": 1380 }, { "epoch": 0.73, "learning_rate": 9.355778894472363e-06, "loss": 0.1079, "step": 1390 }, { "epoch": 0.73, "learning_rate": 9.350753768844222e-06, "loss": 0.1074, "step": 1400 }, { "epoch": 0.73, "eval_loss": 0.12481689453125, "eval_runtime": 153.2249, "eval_samples_per_second": 21.074, "eval_steps_per_second": 0.085, "eval_wer": 13.848830771664339, "step": 1400 }, { "epoch": 0.74, "learning_rate": 9.34572864321608e-06, "loss": 0.1076, "step": 1410 }, { "epoch": 0.75, "learning_rate": 9.340703517587941e-06, "loss": 0.1017, "step": 1420 }, { "epoch": 0.75, "learning_rate": 9.3356783919598e-06, "loss": 0.1058, "step": 1430 }, { "epoch": 0.76, "learning_rate": 9.33065326633166e-06, "loss": 0.1058, "step": 1440 }, { "epoch": 0.76, "learning_rate": 9.325628140703519e-06, "loss": 0.1042, "step": 1450 }, { "epoch": 0.77, "learning_rate": 9.320603015075377e-06, "loss": 0.1066, "step": 1460 }, { "epoch": 0.77, "learning_rate": 9.315577889447238e-06, "loss": 0.1095, "step": 1470 }, { "epoch": 0.78, "learning_rate": 9.310552763819096e-06, "loss": 0.1039, "step": 1480 }, { "epoch": 0.78, "learning_rate": 9.305527638190955e-06, "loss": 0.1037, "step": 1490 }, { "epoch": 0.79, "learning_rate": 9.300502512562815e-06, "loss": 0.1009, "step": 1500 }, { "epoch": 0.79, "eval_loss": 0.12353515625, "eval_runtime": 159.3167, "eval_samples_per_second": 20.268, "eval_steps_per_second": 0.082, "eval_wer": 13.645830587118715, "step": 1500 }, { "epoch": 0.79, "learning_rate": 9.295477386934674e-06, "loss": 0.108, "step": 1510 }, { "epoch": 0.8, "learning_rate": 9.290452261306533e-06, "loss": 0.1005, "step": 1520 }, { "epoch": 0.8, "learning_rate": 9.285427135678393e-06, "loss": 0.1061, "step": 1530 }, { "epoch": 0.81, "learning_rate": 9.280402010050251e-06, "loss": 0.1054, "step": 1540 }, { "epoch": 0.81, "learning_rate": 9.275376884422112e-06, "loss": 0.1033, "step": 1550 }, { "epoch": 0.82, "learning_rate": 9.27035175879397e-06, "loss": 0.1063, "step": 1560 }, { "epoch": 0.82, "learning_rate": 9.265326633165829e-06, "loss": 0.0989, "step": 1570 }, { "epoch": 0.83, "learning_rate": 9.26030150753769e-06, "loss": 0.1005, "step": 1580 }, { "epoch": 0.83, "learning_rate": 9.255276381909548e-06, "loss": 0.1028, "step": 1590 }, { "epoch": 0.84, "learning_rate": 9.250251256281407e-06, "loss": 0.1007, "step": 1600 }, { "epoch": 0.84, "eval_loss": 0.1219482421875, "eval_runtime": 169.4322, "eval_samples_per_second": 19.058, "eval_steps_per_second": 0.077, "eval_wer": 13.624739658854235, "step": 1600 }, { "epoch": 0.84, "learning_rate": 9.245226130653267e-06, "loss": 0.1021, "step": 1610 }, { "epoch": 0.85, "learning_rate": 9.240201005025127e-06, "loss": 0.1029, "step": 1620 }, { "epoch": 0.86, "learning_rate": 9.235175879396986e-06, "loss": 0.1011, "step": 1630 }, { "epoch": 0.86, "learning_rate": 9.230150753768845e-06, "loss": 0.1005, "step": 1640 }, { "epoch": 0.87, "learning_rate": 9.225125628140703e-06, "loss": 0.1012, "step": 1650 }, { "epoch": 0.87, "learning_rate": 9.220100502512564e-06, "loss": 0.1017, "step": 1660 }, { "epoch": 0.88, "learning_rate": 9.215075376884422e-06, "loss": 0.1016, "step": 1670 }, { "epoch": 0.88, "learning_rate": 9.210050251256281e-06, "loss": 0.1005, "step": 1680 }, { "epoch": 0.89, "learning_rate": 9.205025125628141e-06, "loss": 0.1004, "step": 1690 }, { "epoch": 0.89, "learning_rate": 9.200000000000002e-06, "loss": 0.0984, "step": 1700 }, { "epoch": 0.89, "eval_loss": 0.11895751953125, "eval_runtime": 151.6841, "eval_samples_per_second": 21.288, "eval_steps_per_second": 0.086, "eval_wer": 13.300466636787853, "step": 1700 }, { "epoch": 0.9, "learning_rate": 9.19497487437186e-06, "loss": 0.1007, "step": 1710 }, { "epoch": 0.9, "learning_rate": 9.189949748743719e-06, "loss": 0.1058, "step": 1720 }, { "epoch": 0.91, "learning_rate": 9.18492462311558e-06, "loss": 0.1014, "step": 1730 }, { "epoch": 0.91, "learning_rate": 9.179899497487438e-06, "loss": 0.1022, "step": 1740 }, { "epoch": 0.92, "learning_rate": 9.174874371859297e-06, "loss": 0.097, "step": 1750 }, { "epoch": 0.92, "learning_rate": 9.169849246231157e-06, "loss": 0.1013, "step": 1760 }, { "epoch": 0.93, "learning_rate": 9.164824120603016e-06, "loss": 0.0986, "step": 1770 }, { "epoch": 0.93, "learning_rate": 9.159798994974876e-06, "loss": 0.0994, "step": 1780 }, { "epoch": 0.94, "learning_rate": 9.154773869346735e-06, "loss": 0.0993, "step": 1790 }, { "epoch": 0.94, "learning_rate": 9.149748743718593e-06, "loss": 0.1032, "step": 1800 }, { "epoch": 0.94, "eval_loss": 0.117431640625, "eval_runtime": 149.7158, "eval_samples_per_second": 21.568, "eval_steps_per_second": 0.087, "eval_wer": 13.121193746539769, "step": 1800 }, { "epoch": 0.95, "learning_rate": 9.144723618090454e-06, "loss": 0.0973, "step": 1810 }, { "epoch": 0.95, "learning_rate": 9.139698492462312e-06, "loss": 0.0981, "step": 1820 }, { "epoch": 0.96, "learning_rate": 9.134673366834171e-06, "loss": 0.0987, "step": 1830 }, { "epoch": 0.97, "learning_rate": 9.129648241206031e-06, "loss": 0.0987, "step": 1840 }, { "epoch": 0.97, "learning_rate": 9.12462311557789e-06, "loss": 0.0985, "step": 1850 }, { "epoch": 0.98, "learning_rate": 9.11959798994975e-06, "loss": 0.1007, "step": 1860 }, { "epoch": 0.98, "learning_rate": 9.114572864321609e-06, "loss": 0.0968, "step": 1870 }, { "epoch": 0.99, "learning_rate": 9.109547738693468e-06, "loss": 0.0989, "step": 1880 }, { "epoch": 0.99, "learning_rate": 9.104522613065328e-06, "loss": 0.0965, "step": 1890 }, { "epoch": 1.0, "learning_rate": 9.099497487437186e-06, "loss": 0.0963, "step": 1900 }, { "epoch": 1.0, "eval_loss": 0.116455078125, "eval_runtime": 159.5626, "eval_samples_per_second": 20.237, "eval_steps_per_second": 0.081, "eval_wer": 13.07110279191163, "step": 1900 }, { "epoch": 1.0, "learning_rate": 9.094472361809045e-06, "loss": 0.0987, "step": 1910 }, { "epoch": 1.01, "learning_rate": 9.089447236180905e-06, "loss": 0.0921, "step": 1920 }, { "epoch": 1.01, "learning_rate": 9.084422110552764e-06, "loss": 0.0951, "step": 1930 }, { "epoch": 1.02, "learning_rate": 9.079396984924624e-06, "loss": 0.0931, "step": 1940 }, { "epoch": 1.02, "learning_rate": 9.074371859296483e-06, "loss": 0.0972, "step": 1950 }, { "epoch": 1.03, "learning_rate": 9.069346733668343e-06, "loss": 0.095, "step": 1960 }, { "epoch": 1.03, "learning_rate": 9.064321608040202e-06, "loss": 0.0929, "step": 1970 }, { "epoch": 1.04, "learning_rate": 9.05929648241206e-06, "loss": 0.0947, "step": 1980 }, { "epoch": 1.04, "learning_rate": 9.05427135678392e-06, "loss": 0.0925, "step": 1990 }, { "epoch": 1.05, "learning_rate": 9.04924623115578e-06, "loss": 0.0929, "step": 2000 }, { "epoch": 1.05, "eval_loss": 0.11529541015625, "eval_runtime": 155.1322, "eval_samples_per_second": 20.815, "eval_steps_per_second": 0.084, "eval_wer": 12.947193588357809, "step": 2000 }, { "epoch": 1.05, "learning_rate": 9.044221105527638e-06, "loss": 0.0925, "step": 2010 }, { "epoch": 1.06, "learning_rate": 9.039195979899497e-06, "loss": 0.0948, "step": 2020 }, { "epoch": 1.07, "learning_rate": 9.034170854271357e-06, "loss": 0.0968, "step": 2030 }, { "epoch": 1.07, "learning_rate": 9.029145728643218e-06, "loss": 0.0932, "step": 2040 }, { "epoch": 1.08, "learning_rate": 9.024120603015076e-06, "loss": 0.0948, "step": 2050 }, { "epoch": 1.08, "learning_rate": 9.019095477386935e-06, "loss": 0.0912, "step": 2060 }, { "epoch": 1.09, "learning_rate": 9.014070351758795e-06, "loss": 0.0935, "step": 2070 }, { "epoch": 1.09, "learning_rate": 9.009045226130654e-06, "loss": 0.0952, "step": 2080 }, { "epoch": 1.1, "learning_rate": 9.004020100502513e-06, "loss": 0.0907, "step": 2090 }, { "epoch": 1.1, "learning_rate": 8.998994974874371e-06, "loss": 0.0951, "step": 2100 }, { "epoch": 1.1, "eval_loss": 0.11480712890625, "eval_runtime": 170.5255, "eval_samples_per_second": 18.936, "eval_steps_per_second": 0.076, "eval_wer": 12.836466214969287, "step": 2100 }, { "epoch": 1.11, "learning_rate": 8.993969849246232e-06, "loss": 0.0922, "step": 2110 }, { "epoch": 1.11, "learning_rate": 8.988944723618092e-06, "loss": 0.0957, "step": 2120 }, { "epoch": 1.12, "learning_rate": 8.98391959798995e-06, "loss": 0.0933, "step": 2130 }, { "epoch": 1.12, "learning_rate": 8.97889447236181e-06, "loss": 0.0934, "step": 2140 }, { "epoch": 1.13, "learning_rate": 8.97386934673367e-06, "loss": 0.0921, "step": 2150 }, { "epoch": 1.13, "learning_rate": 8.968844221105528e-06, "loss": 0.0918, "step": 2160 }, { "epoch": 1.14, "learning_rate": 8.963819095477387e-06, "loss": 0.0938, "step": 2170 }, { "epoch": 1.14, "learning_rate": 8.958793969849247e-06, "loss": 0.0905, "step": 2180 }, { "epoch": 1.15, "learning_rate": 8.953768844221108e-06, "loss": 0.0938, "step": 2190 }, { "epoch": 1.15, "learning_rate": 8.948743718592966e-06, "loss": 0.0943, "step": 2200 }, { "epoch": 1.15, "eval_loss": 0.1126708984375, "eval_runtime": 205.2247, "eval_samples_per_second": 15.734, "eval_steps_per_second": 0.063, "eval_wer": 12.670375154886504, "step": 2200 }, { "epoch": 1.16, "learning_rate": 8.943718592964825e-06, "loss": 0.0962, "step": 2210 }, { "epoch": 1.16, "learning_rate": 8.938693467336684e-06, "loss": 0.0936, "step": 2220 }, { "epoch": 1.17, "learning_rate": 8.933668341708544e-06, "loss": 0.0878, "step": 2230 }, { "epoch": 1.18, "learning_rate": 8.928643216080402e-06, "loss": 0.0865, "step": 2240 }, { "epoch": 1.18, "learning_rate": 8.923618090452261e-06, "loss": 0.0923, "step": 2250 }, { "epoch": 1.19, "learning_rate": 8.918592964824121e-06, "loss": 0.0908, "step": 2260 }, { "epoch": 1.19, "learning_rate": 8.91356783919598e-06, "loss": 0.0913, "step": 2270 }, { "epoch": 1.2, "learning_rate": 8.90854271356784e-06, "loss": 0.0881, "step": 2280 }, { "epoch": 1.2, "learning_rate": 8.903517587939699e-06, "loss": 0.089, "step": 2290 }, { "epoch": 1.21, "learning_rate": 8.89849246231156e-06, "loss": 0.0906, "step": 2300 }, { "epoch": 1.21, "eval_loss": 0.11187744140625, "eval_runtime": 170.8742, "eval_samples_per_second": 18.897, "eval_steps_per_second": 0.076, "eval_wer": 12.720466109514645, "step": 2300 }, { "epoch": 1.21, "learning_rate": 8.893467336683418e-06, "loss": 0.0893, "step": 2310 }, { "epoch": 1.22, "learning_rate": 8.888442211055277e-06, "loss": 0.0911, "step": 2320 }, { "epoch": 1.22, "learning_rate": 8.883417085427135e-06, "loss": 0.0901, "step": 2330 }, { "epoch": 1.23, "learning_rate": 8.878391959798996e-06, "loss": 0.0948, "step": 2340 }, { "epoch": 1.23, "learning_rate": 8.873366834170854e-06, "loss": 0.0872, "step": 2350 }, { "epoch": 1.24, "learning_rate": 8.868341708542715e-06, "loss": 0.087, "step": 2360 }, { "epoch": 1.24, "learning_rate": 8.863316582914573e-06, "loss": 0.0882, "step": 2370 }, { "epoch": 1.25, "learning_rate": 8.858291457286434e-06, "loss": 0.0905, "step": 2380 }, { "epoch": 1.25, "learning_rate": 8.853266331658292e-06, "loss": 0.0884, "step": 2390 }, { "epoch": 1.26, "learning_rate": 8.848241206030151e-06, "loss": 0.09, "step": 2400 }, { "epoch": 1.26, "eval_loss": 0.1112060546875, "eval_runtime": 204.4645, "eval_samples_per_second": 15.792, "eval_steps_per_second": 0.064, "eval_wer": 12.630829664390603, "step": 2400 }, { "epoch": 1.26, "learning_rate": 8.843216080402011e-06, "loss": 0.0899, "step": 2410 }, { "epoch": 1.27, "learning_rate": 8.83819095477387e-06, "loss": 0.0914, "step": 2420 }, { "epoch": 1.27, "learning_rate": 8.833165829145729e-06, "loss": 0.0885, "step": 2430 }, { "epoch": 1.28, "learning_rate": 8.828140703517587e-06, "loss": 0.0884, "step": 2440 }, { "epoch": 1.29, "learning_rate": 8.823115577889448e-06, "loss": 0.0895, "step": 2450 }, { "epoch": 1.29, "learning_rate": 8.818090452261308e-06, "loss": 0.0917, "step": 2460 }, { "epoch": 1.3, "learning_rate": 8.813065326633167e-06, "loss": 0.0907, "step": 2470 }, { "epoch": 1.3, "learning_rate": 8.808040201005025e-06, "loss": 0.093, "step": 2480 }, { "epoch": 1.31, "learning_rate": 8.803015075376886e-06, "loss": 0.0894, "step": 2490 }, { "epoch": 1.31, "learning_rate": 8.797989949748744e-06, "loss": 0.0881, "step": 2500 }, { "epoch": 1.31, "eval_loss": 0.1104736328125, "eval_runtime": 338.1287, "eval_samples_per_second": 9.55, "eval_steps_per_second": 0.038, "eval_wer": 12.593920539927764, "step": 2500 }, { "epoch": 1.32, "learning_rate": 8.792964824120603e-06, "loss": 0.0923, "step": 2510 }, { "epoch": 1.32, "learning_rate": 8.787939698492463e-06, "loss": 0.0894, "step": 2520 }, { "epoch": 1.33, "learning_rate": 8.782914572864324e-06, "loss": 0.0872, "step": 2530 }, { "epoch": 1.33, "learning_rate": 8.777889447236182e-06, "loss": 0.0898, "step": 2540 }, { "epoch": 1.34, "learning_rate": 8.772864321608041e-06, "loss": 0.0864, "step": 2550 }, { "epoch": 1.34, "learning_rate": 8.7678391959799e-06, "loss": 0.0869, "step": 2560 }, { "epoch": 1.35, "learning_rate": 8.76281407035176e-06, "loss": 0.0918, "step": 2570 }, { "epoch": 1.35, "learning_rate": 8.757788944723619e-06, "loss": 0.0849, "step": 2580 }, { "epoch": 1.36, "learning_rate": 8.752763819095477e-06, "loss": 0.0884, "step": 2590 }, { "epoch": 1.36, "learning_rate": 8.747738693467337e-06, "loss": 0.0921, "step": 2600 }, { "epoch": 1.36, "eval_loss": 0.11029052734375, "eval_runtime": 179.0595, "eval_samples_per_second": 18.033, "eval_steps_per_second": 0.073, "eval_wer": 12.33292030265482, "step": 2600 }, { "epoch": 1.37, "learning_rate": 8.742713567839198e-06, "loss": 0.0902, "step": 2610 }, { "epoch": 1.37, "learning_rate": 8.737688442211056e-06, "loss": 0.0889, "step": 2620 }, { "epoch": 1.38, "learning_rate": 8.732663316582915e-06, "loss": 0.0911, "step": 2630 }, { "epoch": 1.39, "learning_rate": 8.727638190954775e-06, "loss": 0.0897, "step": 2640 }, { "epoch": 1.39, "learning_rate": 8.722613065326634e-06, "loss": 0.0873, "step": 2650 }, { "epoch": 1.4, "learning_rate": 8.717587939698493e-06, "loss": 0.0884, "step": 2660 }, { "epoch": 1.4, "learning_rate": 8.712562814070351e-06, "loss": 0.0867, "step": 2670 }, { "epoch": 1.41, "learning_rate": 8.707537688442212e-06, "loss": 0.0873, "step": 2680 }, { "epoch": 1.41, "learning_rate": 8.70251256281407e-06, "loss": 0.0896, "step": 2690 }, { "epoch": 1.42, "learning_rate": 8.69748743718593e-06, "loss": 0.0894, "step": 2700 }, { "epoch": 1.42, "eval_loss": 0.10845947265625, "eval_runtime": 237.2681, "eval_samples_per_second": 13.609, "eval_steps_per_second": 0.055, "eval_wer": 12.37246579315072, "step": 2700 }, { "epoch": 1.42, "learning_rate": 8.69246231155779e-06, "loss": 0.0866, "step": 2710 }, { "epoch": 1.43, "learning_rate": 8.68743718592965e-06, "loss": 0.0849, "step": 2720 }, { "epoch": 1.43, "learning_rate": 8.682412060301508e-06, "loss": 0.0886, "step": 2730 }, { "epoch": 1.44, "learning_rate": 8.677386934673367e-06, "loss": 0.0876, "step": 2740 }, { "epoch": 1.44, "learning_rate": 8.672361809045227e-06, "loss": 0.0865, "step": 2750 }, { "epoch": 1.45, "learning_rate": 8.667336683417086e-06, "loss": 0.0845, "step": 2760 }, { "epoch": 1.45, "learning_rate": 8.662311557788945e-06, "loss": 0.0886, "step": 2770 }, { "epoch": 1.46, "learning_rate": 8.657286432160805e-06, "loss": 0.0856, "step": 2780 }, { "epoch": 1.46, "learning_rate": 8.652261306532664e-06, "loss": 0.0901, "step": 2790 }, { "epoch": 1.47, "learning_rate": 8.647236180904524e-06, "loss": 0.0915, "step": 2800 }, { "epoch": 1.47, "eval_loss": 0.1082763671875, "eval_runtime": 219.7791, "eval_samples_per_second": 14.692, "eval_steps_per_second": 0.059, "eval_wer": 12.44364767604334, "step": 2800 }, { "epoch": 1.47, "learning_rate": 8.642211055276383e-06, "loss": 0.0865, "step": 2810 }, { "epoch": 1.48, "learning_rate": 8.637185929648241e-06, "loss": 0.0885, "step": 2820 }, { "epoch": 1.48, "learning_rate": 8.632160804020102e-06, "loss": 0.0856, "step": 2830 }, { "epoch": 1.49, "learning_rate": 8.62713567839196e-06, "loss": 0.0873, "step": 2840 }, { "epoch": 1.5, "learning_rate": 8.622110552763819e-06, "loss": 0.0856, "step": 2850 }, { "epoch": 1.5, "learning_rate": 8.61708542713568e-06, "loss": 0.0844, "step": 2860 }, { "epoch": 1.51, "learning_rate": 8.61206030150754e-06, "loss": 0.0875, "step": 2870 }, { "epoch": 1.51, "learning_rate": 8.607035175879398e-06, "loss": 0.0872, "step": 2880 }, { "epoch": 1.52, "learning_rate": 8.602010050251257e-06, "loss": 0.0871, "step": 2890 }, { "epoch": 1.52, "learning_rate": 8.596984924623116e-06, "loss": 0.087, "step": 2900 }, { "epoch": 1.52, "eval_loss": 0.1072998046875, "eval_runtime": 277.1368, "eval_samples_per_second": 11.651, "eval_steps_per_second": 0.047, "eval_wer": 12.274920249927499, "step": 2900 }, { "epoch": 1.53, "learning_rate": 8.591959798994976e-06, "loss": 0.0863, "step": 2910 }, { "epoch": 1.53, "learning_rate": 8.586934673366835e-06, "loss": 0.087, "step": 2920 }, { "epoch": 1.54, "learning_rate": 8.581909547738693e-06, "loss": 0.0862, "step": 2930 }, { "epoch": 1.54, "learning_rate": 8.576884422110553e-06, "loss": 0.0867, "step": 2940 }, { "epoch": 1.55, "learning_rate": 8.571859296482414e-06, "loss": 0.0877, "step": 2950 }, { "epoch": 1.55, "learning_rate": 8.566834170854272e-06, "loss": 0.0869, "step": 2960 }, { "epoch": 1.56, "learning_rate": 8.561809045226131e-06, "loss": 0.0859, "step": 2970 }, { "epoch": 1.56, "learning_rate": 8.556783919597991e-06, "loss": 0.0858, "step": 2980 }, { "epoch": 1.57, "learning_rate": 8.55175879396985e-06, "loss": 0.0834, "step": 2990 }, { "epoch": 1.57, "learning_rate": 8.546733668341709e-06, "loss": 0.0852, "step": 3000 }, { "epoch": 1.57, "eval_loss": 0.10693359375, "eval_runtime": 236.0912, "eval_samples_per_second": 13.677, "eval_steps_per_second": 0.055, "eval_wer": 12.032374574885978, "step": 3000 }, { "epoch": 1.58, "learning_rate": 8.541708542713567e-06, "loss": 0.087, "step": 3010 }, { "epoch": 1.58, "learning_rate": 8.536683417085428e-06, "loss": 0.0844, "step": 3020 }, { "epoch": 1.59, "learning_rate": 8.531658291457288e-06, "loss": 0.0867, "step": 3030 }, { "epoch": 1.59, "learning_rate": 8.526633165829147e-06, "loss": 0.0873, "step": 3040 }, { "epoch": 1.6, "learning_rate": 8.521608040201005e-06, "loss": 0.0871, "step": 3050 }, { "epoch": 1.61, "learning_rate": 8.516582914572866e-06, "loss": 0.0846, "step": 3060 }, { "epoch": 1.61, "learning_rate": 8.511557788944724e-06, "loss": 0.0853, "step": 3070 }, { "epoch": 1.62, "learning_rate": 8.506532663316583e-06, "loss": 0.0881, "step": 3080 }, { "epoch": 1.62, "learning_rate": 8.501507537688443e-06, "loss": 0.0861, "step": 3090 }, { "epoch": 1.63, "learning_rate": 8.496482412060302e-06, "loss": 0.0833, "step": 3100 }, { "epoch": 1.63, "eval_loss": 0.10552978515625, "eval_runtime": 363.0182, "eval_samples_per_second": 8.895, "eval_steps_per_second": 0.036, "eval_wer": 12.090374627613299, "step": 3100 }, { "epoch": 1.63, "learning_rate": 8.491457286432162e-06, "loss": 0.0856, "step": 3110 }, { "epoch": 1.64, "learning_rate": 8.486432160804021e-06, "loss": 0.0831, "step": 3120 }, { "epoch": 1.64, "learning_rate": 8.48140703517588e-06, "loss": 0.0847, "step": 3130 }, { "epoch": 1.65, "learning_rate": 8.47638190954774e-06, "loss": 0.0861, "step": 3140 }, { "epoch": 1.65, "learning_rate": 8.471356783919599e-06, "loss": 0.0835, "step": 3150 }, { "epoch": 1.66, "learning_rate": 8.466331658291457e-06, "loss": 0.0839, "step": 3160 }, { "epoch": 1.66, "learning_rate": 8.461306532663318e-06, "loss": 0.0877, "step": 3170 }, { "epoch": 1.67, "learning_rate": 8.456281407035176e-06, "loss": 0.0853, "step": 3180 }, { "epoch": 1.67, "learning_rate": 8.451256281407035e-06, "loss": 0.0844, "step": 3190 }, { "epoch": 1.68, "learning_rate": 8.446231155778895e-06, "loss": 0.0877, "step": 3200 }, { "epoch": 1.68, "eval_loss": 0.10516357421875, "eval_runtime": 336.3241, "eval_samples_per_second": 9.601, "eval_steps_per_second": 0.039, "eval_wer": 12.074556431414937, "step": 3200 }, { "epoch": 1.68, "learning_rate": 8.441206030150756e-06, "loss": 0.0833, "step": 3210 }, { "epoch": 1.69, "learning_rate": 8.436180904522614e-06, "loss": 0.0894, "step": 3220 }, { "epoch": 1.69, "learning_rate": 8.431155778894473e-06, "loss": 0.0831, "step": 3230 }, { "epoch": 1.7, "learning_rate": 8.426130653266332e-06, "loss": 0.0846, "step": 3240 }, { "epoch": 1.71, "learning_rate": 8.421105527638192e-06, "loss": 0.0832, "step": 3250 }, { "epoch": 1.71, "learning_rate": 8.41608040201005e-06, "loss": 0.0869, "step": 3260 }, { "epoch": 1.72, "learning_rate": 8.41105527638191e-06, "loss": 0.0842, "step": 3270 }, { "epoch": 1.72, "learning_rate": 8.40603015075377e-06, "loss": 0.0835, "step": 3280 }, { "epoch": 1.73, "learning_rate": 8.40100502512563e-06, "loss": 0.0836, "step": 3290 }, { "epoch": 1.73, "learning_rate": 8.395979899497488e-06, "loss": 0.0854, "step": 3300 }, { "epoch": 1.73, "eval_loss": 0.10528564453125, "eval_runtime": 190.8873, "eval_samples_per_second": 16.916, "eval_steps_per_second": 0.068, "eval_wer": 11.984919986290897, "step": 3300 }, { "epoch": 1.74, "learning_rate": 8.390954773869347e-06, "loss": 0.0822, "step": 3310 }, { "epoch": 1.74, "learning_rate": 8.385929648241207e-06, "loss": 0.084, "step": 3320 }, { "epoch": 1.75, "learning_rate": 8.380904522613066e-06, "loss": 0.084, "step": 3330 }, { "epoch": 1.75, "learning_rate": 8.375879396984925e-06, "loss": 0.0813, "step": 3340 }, { "epoch": 1.76, "learning_rate": 8.370854271356783e-06, "loss": 0.0829, "step": 3350 }, { "epoch": 1.76, "learning_rate": 8.365829145728644e-06, "loss": 0.0869, "step": 3360 }, { "epoch": 1.77, "learning_rate": 8.360804020100504e-06, "loss": 0.0821, "step": 3370 }, { "epoch": 1.77, "learning_rate": 8.355778894472363e-06, "loss": 0.0846, "step": 3380 }, { "epoch": 1.78, "learning_rate": 8.350753768844221e-06, "loss": 0.0841, "step": 3390 }, { "epoch": 1.78, "learning_rate": 8.345728643216082e-06, "loss": 0.083, "step": 3400 }, { "epoch": 1.78, "eval_loss": 0.10321044921875, "eval_runtime": 267.6827, "eval_samples_per_second": 12.063, "eval_steps_per_second": 0.049, "eval_wer": 11.908465371332156, "step": 3400 }, { "epoch": 1.79, "learning_rate": 8.34070351758794e-06, "loss": 0.0873, "step": 3410 }, { "epoch": 1.79, "learning_rate": 8.335678391959799e-06, "loss": 0.0848, "step": 3420 }, { "epoch": 1.8, "learning_rate": 8.33065326633166e-06, "loss": 0.0827, "step": 3430 }, { "epoch": 1.8, "learning_rate": 8.325628140703518e-06, "loss": 0.0826, "step": 3440 }, { "epoch": 1.81, "learning_rate": 8.320603015075378e-06, "loss": 0.0833, "step": 3450 }, { "epoch": 1.82, "learning_rate": 8.315577889447237e-06, "loss": 0.0869, "step": 3460 }, { "epoch": 1.82, "learning_rate": 8.310552763819096e-06, "loss": 0.0797, "step": 3470 }, { "epoch": 1.83, "learning_rate": 8.305527638190956e-06, "loss": 0.0822, "step": 3480 }, { "epoch": 1.83, "learning_rate": 8.300502512562815e-06, "loss": 0.0808, "step": 3490 }, { "epoch": 1.84, "learning_rate": 8.295477386934673e-06, "loss": 0.0837, "step": 3500 }, { "epoch": 1.84, "eval_loss": 0.10302734375, "eval_runtime": 343.7062, "eval_samples_per_second": 9.395, "eval_steps_per_second": 0.038, "eval_wer": 11.992829084390076, "step": 3500 }, { "epoch": 1.84, "learning_rate": 8.332323232323232e-06, "loss": 0.0844, "step": 3510 }, { "epoch": 1.85, "learning_rate": 8.327272727272728e-06, "loss": 0.0843, "step": 3520 }, { "epoch": 1.85, "learning_rate": 8.322222222222223e-06, "loss": 0.0826, "step": 3530 }, { "epoch": 1.86, "learning_rate": 8.317171717171719e-06, "loss": 0.0819, "step": 3540 }, { "epoch": 1.86, "learning_rate": 8.312121212121213e-06, "loss": 0.0844, "step": 3550 }, { "epoch": 1.87, "learning_rate": 8.307070707070707e-06, "loss": 0.0854, "step": 3560 }, { "epoch": 1.87, "learning_rate": 8.302020202020203e-06, "loss": 0.0813, "step": 3570 }, { "epoch": 1.88, "learning_rate": 8.296969696969697e-06, "loss": 0.0857, "step": 3580 }, { "epoch": 1.88, "learning_rate": 8.291919191919192e-06, "loss": 0.0825, "step": 3590 }, { "epoch": 1.89, "learning_rate": 8.286868686868688e-06, "loss": 0.083, "step": 3600 }, { "epoch": 1.89, "eval_loss": 0.1026611328125, "eval_runtime": 209.4768, "eval_samples_per_second": 15.415, "eval_steps_per_second": 0.062, "eval_wer": 11.797737997943635, "step": 3600 }, { "epoch": 1.89, "learning_rate": 8.281818181818182e-06, "loss": 0.0842, "step": 3610 }, { "epoch": 1.9, "learning_rate": 8.276767676767678e-06, "loss": 0.0847, "step": 3620 }, { "epoch": 1.9, "learning_rate": 8.271717171717172e-06, "loss": 0.0844, "step": 3630 }, { "epoch": 1.91, "learning_rate": 8.266666666666667e-06, "loss": 0.0812, "step": 3640 }, { "epoch": 1.92, "learning_rate": 8.261616161616163e-06, "loss": 0.0838, "step": 3650 }, { "epoch": 1.92, "learning_rate": 8.256565656565657e-06, "loss": 0.0822, "step": 3660 }, { "epoch": 1.93, "learning_rate": 8.251515151515153e-06, "loss": 0.0836, "step": 3670 }, { "epoch": 1.93, "learning_rate": 8.246464646464647e-06, "loss": 0.0852, "step": 3680 }, { "epoch": 1.94, "learning_rate": 8.241414141414143e-06, "loss": 0.0816, "step": 3690 }, { "epoch": 1.94, "learning_rate": 8.236363636363637e-06, "loss": 0.0835, "step": 3700 }, { "epoch": 1.94, "eval_loss": 0.10247802734375, "eval_runtime": 250.486, "eval_samples_per_second": 12.891, "eval_steps_per_second": 0.052, "eval_wer": 11.771374337613034, "step": 3700 }, { "epoch": 1.95, "learning_rate": 8.231313131313132e-06, "loss": 0.0787, "step": 3710 }, { "epoch": 1.95, "learning_rate": 8.226262626262626e-06, "loss": 0.0834, "step": 3720 }, { "epoch": 1.96, "learning_rate": 8.221212121212122e-06, "loss": 0.0814, "step": 3730 }, { "epoch": 1.96, "learning_rate": 8.216161616161616e-06, "loss": 0.0813, "step": 3740 }, { "epoch": 1.97, "learning_rate": 8.211111111111112e-06, "loss": 0.0831, "step": 3750 }, { "epoch": 1.97, "learning_rate": 8.206060606060607e-06, "loss": 0.0842, "step": 3760 }, { "epoch": 1.98, "learning_rate": 8.201010101010103e-06, "loss": 0.0843, "step": 3770 }, { "epoch": 1.98, "learning_rate": 8.195959595959597e-06, "loss": 0.0829, "step": 3780 }, { "epoch": 1.99, "learning_rate": 8.190909090909091e-06, "loss": 0.0858, "step": 3790 }, { "epoch": 1.99, "learning_rate": 8.185858585858587e-06, "loss": 0.0813, "step": 3800 }, { "epoch": 1.99, "eval_loss": 0.10205078125, "eval_runtime": 458.2285, "eval_samples_per_second": 7.047, "eval_steps_per_second": 0.028, "eval_wer": 11.621101473728611, "step": 3800 }, { "epoch": 2.0, "learning_rate": 8.180808080808081e-06, "loss": 0.0803, "step": 3810 }, { "epoch": 2.0, "learning_rate": 8.175757575757577e-06, "loss": 0.0822, "step": 3820 }, { "epoch": 2.01, "learning_rate": 8.170707070707072e-06, "loss": 0.0781, "step": 3830 }, { "epoch": 2.01, "learning_rate": 8.165656565656566e-06, "loss": 0.0801, "step": 3840 }, { "epoch": 2.02, "learning_rate": 8.16060606060606e-06, "loss": 0.0798, "step": 3850 }, { "epoch": 2.03, "learning_rate": 8.155555555555556e-06, "loss": 0.0756, "step": 3860 }, { "epoch": 2.03, "learning_rate": 8.15050505050505e-06, "loss": 0.0798, "step": 3870 }, { "epoch": 2.04, "learning_rate": 8.145454545454547e-06, "loss": 0.0758, "step": 3880 }, { "epoch": 2.04, "learning_rate": 8.140404040404041e-06, "loss": 0.0793, "step": 3890 }, { "epoch": 2.05, "learning_rate": 8.135353535353537e-06, "loss": 0.077, "step": 3900 }, { "epoch": 2.05, "eval_loss": 0.10223388671875, "eval_runtime": 177.8708, "eval_samples_per_second": 18.154, "eval_steps_per_second": 0.073, "eval_wer": 11.634283303893913, "step": 3900 }, { "epoch": 2.05, "learning_rate": 8.130303030303031e-06, "loss": 0.0796, "step": 3910 }, { "epoch": 2.06, "learning_rate": 8.125252525252526e-06, "loss": 0.0777, "step": 3920 }, { "epoch": 2.06, "learning_rate": 8.12020202020202e-06, "loss": 0.0786, "step": 3930 }, { "epoch": 2.07, "learning_rate": 8.115151515151516e-06, "loss": 0.0775, "step": 3940 }, { "epoch": 2.07, "learning_rate": 8.11010101010101e-06, "loss": 0.0828, "step": 3950 }, { "epoch": 2.08, "learning_rate": 8.105050505050506e-06, "loss": 0.0769, "step": 3960 }, { "epoch": 2.08, "learning_rate": 8.1e-06, "loss": 0.0806, "step": 3970 }, { "epoch": 2.09, "learning_rate": 8.094949494949496e-06, "loss": 0.0797, "step": 3980 }, { "epoch": 2.09, "learning_rate": 8.08989898989899e-06, "loss": 0.0779, "step": 3990 }, { "epoch": 2.1, "learning_rate": 8.084848484848485e-06, "loss": 0.0765, "step": 4000 }, { "epoch": 2.1, "eval_loss": 0.10174560546875, "eval_runtime": 176.0647, "eval_samples_per_second": 18.34, "eval_steps_per_second": 0.074, "eval_wer": 11.650101500092273, "step": 4000 }, { "epoch": 2.1, "learning_rate": 8.079797979797981e-06, "loss": 0.0751, "step": 4010 }, { "epoch": 2.11, "learning_rate": 8.074747474747475e-06, "loss": 0.0775, "step": 4020 }, { "epoch": 2.11, "learning_rate": 8.069696969696971e-06, "loss": 0.0795, "step": 4030 }, { "epoch": 2.12, "learning_rate": 8.064646464646466e-06, "loss": 0.0777, "step": 4040 }, { "epoch": 2.12, "learning_rate": 8.05959595959596e-06, "loss": 0.0789, "step": 4050 }, { "epoch": 2.13, "learning_rate": 8.054545454545454e-06, "loss": 0.0794, "step": 4060 }, { "epoch": 2.14, "learning_rate": 8.04949494949495e-06, "loss": 0.0803, "step": 4070 }, { "epoch": 2.14, "learning_rate": 8.044444444444444e-06, "loss": 0.0774, "step": 4080 }, { "epoch": 2.15, "learning_rate": 8.03939393939394e-06, "loss": 0.0764, "step": 4090 }, { "epoch": 2.15, "learning_rate": 8.034343434343435e-06, "loss": 0.079, "step": 4100 }, { "epoch": 2.15, "eval_loss": 0.10064697265625, "eval_runtime": 186.171, "eval_samples_per_second": 17.344, "eval_steps_per_second": 0.07, "eval_wer": 11.520919564472331, "step": 4100 }, { "epoch": 2.16, "learning_rate": 8.02929292929293e-06, "loss": 0.0764, "step": 4110 }, { "epoch": 2.16, "learning_rate": 8.024242424242425e-06, "loss": 0.0764, "step": 4120 }, { "epoch": 2.17, "learning_rate": 8.01919191919192e-06, "loss": 0.0781, "step": 4130 }, { "epoch": 2.17, "learning_rate": 8.014141414141415e-06, "loss": 0.0744, "step": 4140 }, { "epoch": 2.18, "learning_rate": 8.00909090909091e-06, "loss": 0.075, "step": 4150 }, { "epoch": 2.18, "learning_rate": 8.004040404040406e-06, "loss": 0.08, "step": 4160 }, { "epoch": 2.19, "learning_rate": 7.9989898989899e-06, "loss": 0.0771, "step": 4170 }, { "epoch": 2.19, "learning_rate": 7.993939393939396e-06, "loss": 0.0752, "step": 4180 }, { "epoch": 2.2, "learning_rate": 7.98888888888889e-06, "loss": 0.0797, "step": 4190 }, { "epoch": 2.2, "learning_rate": 7.983838383838384e-06, "loss": 0.0776, "step": 4200 }, { "epoch": 2.2, "eval_loss": 0.10064697265625, "eval_runtime": 152.434, "eval_samples_per_second": 21.183, "eval_steps_per_second": 0.085, "eval_wer": 11.44973768157971, "step": 4200 }, { "epoch": 2.21, "learning_rate": 7.978787878787879e-06, "loss": 0.0798, "step": 4210 }, { "epoch": 2.21, "learning_rate": 7.973737373737375e-06, "loss": 0.0805, "step": 4220 }, { "epoch": 2.22, "learning_rate": 7.968686868686869e-06, "loss": 0.0756, "step": 4230 }, { "epoch": 2.22, "learning_rate": 7.963636363636365e-06, "loss": 0.0769, "step": 4240 }, { "epoch": 2.23, "learning_rate": 7.95858585858586e-06, "loss": 0.0754, "step": 4250 }, { "epoch": 2.24, "learning_rate": 7.953535353535354e-06, "loss": 0.0777, "step": 4260 }, { "epoch": 2.24, "learning_rate": 7.948484848484848e-06, "loss": 0.075, "step": 4270 }, { "epoch": 2.25, "learning_rate": 7.943434343434344e-06, "loss": 0.0755, "step": 4280 }, { "epoch": 2.25, "learning_rate": 7.938383838383838e-06, "loss": 0.076, "step": 4290 }, { "epoch": 2.26, "learning_rate": 7.933333333333334e-06, "loss": 0.0778, "step": 4300 }, { "epoch": 2.26, "eval_loss": 0.10003662109375, "eval_runtime": 403.4335, "eval_samples_per_second": 8.004, "eval_steps_per_second": 0.032, "eval_wer": 11.523555930505392, "step": 4300 }, { "epoch": 2.26, "learning_rate": 7.92828282828283e-06, "loss": 0.0763, "step": 4310 }, { "epoch": 2.27, "learning_rate": 7.923232323232324e-06, "loss": 0.0779, "step": 4320 }, { "epoch": 2.27, "learning_rate": 7.918181818181819e-06, "loss": 0.0775, "step": 4330 }, { "epoch": 2.28, "learning_rate": 7.913131313131313e-06, "loss": 0.0771, "step": 4340 }, { "epoch": 2.28, "learning_rate": 7.908080808080809e-06, "loss": 0.0762, "step": 4350 }, { "epoch": 2.29, "learning_rate": 7.903030303030303e-06, "loss": 0.0775, "step": 4360 }, { "epoch": 2.29, "learning_rate": 7.8979797979798e-06, "loss": 0.0743, "step": 4370 }, { "epoch": 2.3, "learning_rate": 7.892929292929294e-06, "loss": 0.0754, "step": 4380 }, { "epoch": 2.3, "learning_rate": 7.88787878787879e-06, "loss": 0.0742, "step": 4390 }, { "epoch": 2.31, "learning_rate": 7.882828282828284e-06, "loss": 0.0753, "step": 4400 }, { "epoch": 2.31, "eval_loss": 0.09893798828125, "eval_runtime": 357.3277, "eval_samples_per_second": 9.037, "eval_steps_per_second": 0.036, "eval_wer": 11.484010440009492, "step": 4400 }, { "epoch": 2.31, "learning_rate": 7.877777777777778e-06, "loss": 0.0768, "step": 4410 }, { "epoch": 2.32, "learning_rate": 7.872727272727273e-06, "loss": 0.0795, "step": 4420 }, { "epoch": 2.32, "learning_rate": 7.867676767676768e-06, "loss": 0.0751, "step": 4430 }, { "epoch": 2.33, "learning_rate": 7.862626262626263e-06, "loss": 0.0774, "step": 4440 }, { "epoch": 2.33, "learning_rate": 7.857575757575759e-06, "loss": 0.0761, "step": 4450 }, { "epoch": 2.34, "learning_rate": 7.852525252525253e-06, "loss": 0.0752, "step": 4460 }, { "epoch": 2.35, "learning_rate": 7.847474747474747e-06, "loss": 0.0727, "step": 4470 }, { "epoch": 2.35, "learning_rate": 7.842424242424243e-06, "loss": 0.0783, "step": 4480 }, { "epoch": 2.36, "learning_rate": 7.837373737373738e-06, "loss": 0.0782, "step": 4490 }, { "epoch": 2.36, "learning_rate": 7.832323232323234e-06, "loss": 0.0774, "step": 4500 }, { "epoch": 2.36, "eval_loss": 0.09832763671875, "eval_runtime": 375.3825, "eval_samples_per_second": 8.602, "eval_steps_per_second": 0.035, "eval_wer": 11.373283066620969, "step": 4500 }, { "epoch": 2.37, "learning_rate": 7.827272727272728e-06, "loss": 0.078, "step": 4510 }, { "epoch": 2.37, "learning_rate": 7.822222222222224e-06, "loss": 0.0752, "step": 4520 }, { "epoch": 2.38, "learning_rate": 7.817171717171718e-06, "loss": 0.0749, "step": 4530 }, { "epoch": 2.38, "learning_rate": 7.812121212121213e-06, "loss": 0.0768, "step": 4540 }, { "epoch": 2.39, "learning_rate": 7.807070707070707e-06, "loss": 0.0749, "step": 4550 }, { "epoch": 2.39, "learning_rate": 7.802020202020203e-06, "loss": 0.0763, "step": 4560 }, { "epoch": 2.4, "learning_rate": 7.796969696969697e-06, "loss": 0.075, "step": 4570 }, { "epoch": 2.4, "learning_rate": 7.791919191919193e-06, "loss": 0.076, "step": 4580 }, { "epoch": 2.41, "learning_rate": 7.786868686868687e-06, "loss": 0.0754, "step": 4590 }, { "epoch": 2.41, "learning_rate": 7.781818181818183e-06, "loss": 0.0748, "step": 4600 }, { "epoch": 2.41, "eval_loss": 0.09857177734375, "eval_runtime": 185.0049, "eval_samples_per_second": 17.454, "eval_steps_per_second": 0.07, "eval_wer": 11.49191953810867, "step": 4600 }, { "epoch": 2.42, "learning_rate": 7.776767676767678e-06, "loss": 0.0765, "step": 4610 }, { "epoch": 2.42, "learning_rate": 7.771717171717172e-06, "loss": 0.0738, "step": 4620 }, { "epoch": 2.43, "learning_rate": 7.766666666666666e-06, "loss": 0.0768, "step": 4630 }, { "epoch": 2.43, "learning_rate": 7.761616161616162e-06, "loss": 0.0746, "step": 4640 }, { "epoch": 2.44, "learning_rate": 7.756565656565658e-06, "loss": 0.073, "step": 4650 }, { "epoch": 2.44, "learning_rate": 7.751515151515153e-06, "loss": 0.0749, "step": 4660 }, { "epoch": 2.45, "learning_rate": 7.746464646464647e-06, "loss": 0.0766, "step": 4670 }, { "epoch": 2.46, "learning_rate": 7.741414141414141e-06, "loss": 0.0766, "step": 4680 }, { "epoch": 2.46, "learning_rate": 7.736363636363637e-06, "loss": 0.075, "step": 4690 }, { "epoch": 2.47, "learning_rate": 7.731313131313131e-06, "loss": 0.0752, "step": 4700 }, { "epoch": 2.47, "eval_loss": 0.09906005859375, "eval_runtime": 286.1402, "eval_samples_per_second": 11.285, "eval_steps_per_second": 0.045, "eval_wer": 11.32846484405895, "step": 4700 }, { "epoch": 2.47, "learning_rate": 7.726262626262627e-06, "loss": 0.0757, "step": 4710 }, { "epoch": 2.48, "learning_rate": 7.721212121212122e-06, "loss": 0.075, "step": 4720 }, { "epoch": 2.48, "learning_rate": 7.716161616161618e-06, "loss": 0.077, "step": 4730 }, { "epoch": 2.49, "learning_rate": 7.711111111111112e-06, "loss": 0.0767, "step": 4740 }, { "epoch": 2.49, "learning_rate": 7.706060606060606e-06, "loss": 0.0762, "step": 4750 }, { "epoch": 2.5, "learning_rate": 7.7010101010101e-06, "loss": 0.0737, "step": 4760 }, { "epoch": 2.5, "learning_rate": 7.695959595959597e-06, "loss": 0.0737, "step": 4770 }, { "epoch": 2.51, "learning_rate": 7.690909090909091e-06, "loss": 0.0732, "step": 4780 }, { "epoch": 2.51, "learning_rate": 7.685858585858587e-06, "loss": 0.0759, "step": 4790 }, { "epoch": 2.52, "learning_rate": 7.680808080808081e-06, "loss": 0.0747, "step": 4800 }, { "epoch": 2.52, "eval_loss": 0.09674072265625, "eval_runtime": 525.1151, "eval_samples_per_second": 6.149, "eval_steps_per_second": 0.025, "eval_wer": 11.225646568769609, "step": 4800 }, { "epoch": 2.52, "learning_rate": 7.675757575757577e-06, "loss": 0.076, "step": 4810 }, { "epoch": 2.53, "learning_rate": 7.670707070707071e-06, "loss": 0.0719, "step": 4820 }, { "epoch": 2.53, "learning_rate": 7.665656565656566e-06, "loss": 0.0791, "step": 4830 }, { "epoch": 2.54, "learning_rate": 7.660606060606062e-06, "loss": 0.0748, "step": 4840 }, { "epoch": 2.54, "learning_rate": 7.655555555555556e-06, "loss": 0.0738, "step": 4850 }, { "epoch": 2.55, "learning_rate": 7.650505050505052e-06, "loss": 0.0759, "step": 4860 }, { "epoch": 2.56, "learning_rate": 7.645454545454546e-06, "loss": 0.0764, "step": 4870 }, { "epoch": 2.56, "learning_rate": 7.64040404040404e-06, "loss": 0.0727, "step": 4880 }, { "epoch": 2.57, "learning_rate": 7.635353535353535e-06, "loss": 0.0745, "step": 4890 }, { "epoch": 2.57, "learning_rate": 7.630303030303031e-06, "loss": 0.0738, "step": 4900 }, { "epoch": 2.57, "eval_loss": 0.0968017578125, "eval_runtime": 331.6628, "eval_samples_per_second": 9.736, "eval_steps_per_second": 0.039, "eval_wer": 11.249373863067149, "step": 4900 }, { "epoch": 2.58, "learning_rate": 7.625252525252525e-06, "loss": 0.0766, "step": 4910 }, { "epoch": 2.58, "learning_rate": 7.620202020202021e-06, "loss": 0.0726, "step": 4920 }, { "epoch": 2.59, "learning_rate": 7.6151515151515155e-06, "loss": 0.0782, "step": 4930 }, { "epoch": 2.59, "learning_rate": 7.610101010101011e-06, "loss": 0.0767, "step": 4940 }, { "epoch": 2.6, "learning_rate": 7.605050505050505e-06, "loss": 0.0737, "step": 4950 }, { "epoch": 2.6, "learning_rate": 7.600000000000001e-06, "loss": 0.0771, "step": 4960 }, { "epoch": 2.61, "learning_rate": 7.594949494949495e-06, "loss": 0.0781, "step": 4970 }, { "epoch": 2.61, "learning_rate": 7.58989898989899e-06, "loss": 0.0748, "step": 4980 }, { "epoch": 2.62, "learning_rate": 7.584848484848486e-06, "loss": 0.0745, "step": 4990 }, { "epoch": 2.62, "learning_rate": 7.579797979797981e-06, "loss": 0.0751, "step": 5000 }, { "epoch": 2.62, "eval_loss": 0.09808349609375, "eval_runtime": 252.8514, "eval_samples_per_second": 12.77, "eval_steps_per_second": 0.051, "eval_wer": 11.204555640505127, "step": 5000 }, { "epoch": 2.63, "learning_rate": 7.574747474747476e-06, "loss": 0.0763, "step": 5010 }, { "epoch": 2.63, "learning_rate": 7.56969696969697e-06, "loss": 0.0706, "step": 5020 }, { "epoch": 2.64, "learning_rate": 7.564646464646466e-06, "loss": 0.0721, "step": 5030 }, { "epoch": 2.64, "learning_rate": 7.55959595959596e-06, "loss": 0.0747, "step": 5040 }, { "epoch": 2.65, "learning_rate": 7.5545454545454555e-06, "loss": 0.0731, "step": 5050 }, { "epoch": 2.65, "learning_rate": 7.54949494949495e-06, "loss": 0.0743, "step": 5060 }, { "epoch": 2.66, "learning_rate": 7.544444444444445e-06, "loss": 0.0743, "step": 5070 }, { "epoch": 2.67, "learning_rate": 7.53939393939394e-06, "loss": 0.0755, "step": 5080 }, { "epoch": 2.67, "learning_rate": 7.534343434343435e-06, "loss": 0.0737, "step": 5090 }, { "epoch": 2.68, "learning_rate": 7.5292929292929295e-06, "loss": 0.0756, "step": 5100 }, { "epoch": 2.68, "eval_loss": 0.096923828125, "eval_runtime": 401.3358, "eval_samples_per_second": 8.046, "eval_steps_per_second": 0.032, "eval_wer": 11.220373836703489, "step": 5100 }, { "epoch": 2.68, "learning_rate": 7.524242424242425e-06, "loss": 0.0736, "step": 5110 }, { "epoch": 2.69, "learning_rate": 7.519191919191919e-06, "loss": 0.0756, "step": 5120 }, { "epoch": 2.69, "learning_rate": 7.514141414141415e-06, "loss": 0.0742, "step": 5130 }, { "epoch": 2.7, "learning_rate": 7.509090909090909e-06, "loss": 0.0747, "step": 5140 }, { "epoch": 2.7, "learning_rate": 7.504040404040404e-06, "loss": 0.0749, "step": 5150 }, { "epoch": 2.71, "learning_rate": 7.4989898989899e-06, "loss": 0.0756, "step": 5160 }, { "epoch": 2.71, "learning_rate": 7.493939393939395e-06, "loss": 0.0759, "step": 5170 }, { "epoch": 2.72, "learning_rate": 7.48888888888889e-06, "loss": 0.0768, "step": 5180 }, { "epoch": 2.72, "learning_rate": 7.483838383838384e-06, "loss": 0.0717, "step": 5190 }, { "epoch": 2.73, "learning_rate": 7.47878787878788e-06, "loss": 0.0742, "step": 5200 }, { "epoch": 2.73, "eval_loss": 0.09613037109375, "eval_runtime": 429.166, "eval_samples_per_second": 7.524, "eval_steps_per_second": 0.03, "eval_wer": 11.228282934802667, "step": 5200 }, { "epoch": 2.73, "learning_rate": 7.4737373737373744e-06, "loss": 0.0737, "step": 5210 }, { "epoch": 2.74, "learning_rate": 7.4686868686868696e-06, "loss": 0.0776, "step": 5220 }, { "epoch": 2.74, "learning_rate": 7.463636363636364e-06, "loss": 0.0757, "step": 5230 }, { "epoch": 2.75, "learning_rate": 7.45858585858586e-06, "loss": 0.0751, "step": 5240 }, { "epoch": 2.75, "learning_rate": 7.453535353535354e-06, "loss": 0.0726, "step": 5250 }, { "epoch": 2.76, "learning_rate": 7.448484848484849e-06, "loss": 0.072, "step": 5260 }, { "epoch": 2.76, "learning_rate": 7.443434343434344e-06, "loss": 0.0722, "step": 5270 }, { "epoch": 2.77, "learning_rate": 7.438383838383839e-06, "loss": 0.0765, "step": 5280 }, { "epoch": 2.78, "learning_rate": 7.433333333333334e-06, "loss": 0.0763, "step": 5290 }, { "epoch": 2.78, "learning_rate": 7.428282828282829e-06, "loss": 0.0758, "step": 5300 }, { "epoch": 2.78, "eval_loss": 0.0966796875, "eval_runtime": 154.4148, "eval_samples_per_second": 20.911, "eval_steps_per_second": 0.084, "eval_wer": 11.141282855711687, "step": 5300 }, { "epoch": 2.79, "learning_rate": 7.423232323232323e-06, "loss": 0.0748, "step": 5310 }, { "epoch": 2.79, "learning_rate": 7.4181818181818185e-06, "loss": 0.0757, "step": 5320 }, { "epoch": 2.8, "learning_rate": 7.4131313131313145e-06, "loss": 0.0752, "step": 5330 }, { "epoch": 2.8, "learning_rate": 7.408080808080809e-06, "loss": 0.0747, "step": 5340 }, { "epoch": 2.81, "learning_rate": 7.403030303030304e-06, "loss": 0.0723, "step": 5350 }, { "epoch": 2.81, "learning_rate": 7.397979797979798e-06, "loss": 0.0772, "step": 5360 }, { "epoch": 2.82, "learning_rate": 7.392929292929294e-06, "loss": 0.075, "step": 5370 }, { "epoch": 2.82, "learning_rate": 7.3878787878787885e-06, "loss": 0.0733, "step": 5380 }, { "epoch": 2.83, "learning_rate": 7.382828282828284e-06, "loss": 0.074, "step": 5390 }, { "epoch": 2.83, "learning_rate": 7.377777777777778e-06, "loss": 0.0724, "step": 5400 }, { "epoch": 2.83, "eval_loss": 0.096435546875, "eval_runtime": 231.1522, "eval_samples_per_second": 13.969, "eval_steps_per_second": 0.056, "eval_wer": 11.236192032901847, "step": 5400 }, { "epoch": 2.84, "learning_rate": 7.372727272727274e-06, "loss": 0.0747, "step": 5410 }, { "epoch": 2.84, "learning_rate": 7.367676767676768e-06, "loss": 0.0716, "step": 5420 }, { "epoch": 2.85, "learning_rate": 7.362626262626263e-06, "loss": 0.0748, "step": 5430 }, { "epoch": 2.85, "learning_rate": 7.357575757575758e-06, "loss": 0.0759, "step": 5440 }, { "epoch": 2.86, "learning_rate": 7.352525252525254e-06, "loss": 0.0755, "step": 5450 }, { "epoch": 2.86, "learning_rate": 7.347474747474748e-06, "loss": 0.0741, "step": 5460 }, { "epoch": 2.87, "learning_rate": 7.342424242424243e-06, "loss": 0.0783, "step": 5470 }, { "epoch": 2.88, "learning_rate": 7.337373737373737e-06, "loss": 0.074, "step": 5480 }, { "epoch": 2.88, "learning_rate": 7.332323232323233e-06, "loss": 0.0752, "step": 5490 }, { "epoch": 2.89, "learning_rate": 7.3272727272727285e-06, "loss": 0.0732, "step": 5500 }, { "epoch": 2.89, "eval_loss": 0.09649658203125, "eval_runtime": 195.7291, "eval_samples_per_second": 16.497, "eval_steps_per_second": 0.066, "eval_wer": 11.017373652157865, "step": 5500 }, { "epoch": 2.89, "learning_rate": 7.322222222222223e-06, "loss": 0.0734, "step": 5510 }, { "epoch": 2.9, "learning_rate": 7.317171717171718e-06, "loss": 0.0733, "step": 5520 }, { "epoch": 2.9, "learning_rate": 7.312121212121212e-06, "loss": 0.0725, "step": 5530 }, { "epoch": 2.91, "learning_rate": 7.307070707070708e-06, "loss": 0.0759, "step": 5540 }, { "epoch": 2.91, "learning_rate": 7.3020202020202025e-06, "loss": 0.0729, "step": 5550 }, { "epoch": 2.92, "learning_rate": 7.296969696969698e-06, "loss": 0.0735, "step": 5560 }, { "epoch": 2.92, "learning_rate": 7.291919191919192e-06, "loss": 0.0708, "step": 5570 }, { "epoch": 2.93, "learning_rate": 7.286868686868688e-06, "loss": 0.0728, "step": 5580 }, { "epoch": 2.93, "learning_rate": 7.281818181818182e-06, "loss": 0.0736, "step": 5590 }, { "epoch": 2.94, "learning_rate": 7.276767676767677e-06, "loss": 0.0734, "step": 5600 }, { "epoch": 2.94, "eval_loss": 0.09649658203125, "eval_runtime": 193.3684, "eval_samples_per_second": 16.699, "eval_steps_per_second": 0.067, "eval_wer": 11.130737391579446, "step": 5600 }, { "epoch": 2.94, "learning_rate": 7.271717171717172e-06, "loss": 0.0753, "step": 5610 }, { "epoch": 2.95, "learning_rate": 7.266666666666668e-06, "loss": 0.0742, "step": 5620 }, { "epoch": 2.95, "learning_rate": 7.261616161616162e-06, "loss": 0.0737, "step": 5630 }, { "epoch": 2.96, "learning_rate": 7.256565656565657e-06, "loss": 0.072, "step": 5640 }, { "epoch": 2.96, "learning_rate": 7.251515151515151e-06, "loss": 0.0746, "step": 5650 }, { "epoch": 2.97, "learning_rate": 7.246464646464647e-06, "loss": 0.0738, "step": 5660 }, { "epoch": 2.97, "learning_rate": 7.241414141414142e-06, "loss": 0.0733, "step": 5670 }, { "epoch": 2.98, "learning_rate": 7.236363636363637e-06, "loss": 0.0766, "step": 5680 }, { "epoch": 2.99, "learning_rate": 7.231313131313132e-06, "loss": 0.0728, "step": 5690 }, { "epoch": 2.99, "learning_rate": 7.226262626262627e-06, "loss": 0.0736, "step": 5700 }, { "epoch": 2.99, "eval_loss": 0.09588623046875, "eval_runtime": 184.4712, "eval_samples_per_second": 17.504, "eval_steps_per_second": 0.07, "eval_wer": 11.117555561414147, "step": 5700 }, { "epoch": 3.0, "learning_rate": 7.221212121212122e-06, "loss": 0.0749, "step": 5710 }, { "epoch": 3.0, "learning_rate": 7.2161616161616166e-06, "loss": 0.073, "step": 5720 }, { "epoch": 3.01, "learning_rate": 7.211111111111112e-06, "loss": 0.0679, "step": 5730 }, { "epoch": 3.01, "learning_rate": 7.206060606060606e-06, "loss": 0.0701, "step": 5740 }, { "epoch": 3.02, "learning_rate": 7.201010101010102e-06, "loss": 0.0699, "step": 5750 }, { "epoch": 3.02, "learning_rate": 7.195959595959596e-06, "loss": 0.0693, "step": 5760 }, { "epoch": 3.03, "learning_rate": 7.1909090909090914e-06, "loss": 0.0702, "step": 5770 }, { "epoch": 3.03, "learning_rate": 7.185858585858586e-06, "loss": 0.0693, "step": 5780 }, { "epoch": 3.04, "learning_rate": 7.180808080808082e-06, "loss": 0.0692, "step": 5790 }, { "epoch": 3.04, "learning_rate": 7.175757575757576e-06, "loss": 0.068, "step": 5800 }, { "epoch": 3.04, "eval_loss": 0.09613037109375, "eval_runtime": 168.0475, "eval_samples_per_second": 19.215, "eval_steps_per_second": 0.077, "eval_wer": 11.025282750257045, "step": 5800 }, { "epoch": 3.05, "learning_rate": 7.170707070707071e-06, "loss": 0.0681, "step": 5810 }, { "epoch": 3.05, "learning_rate": 7.1656565656565655e-06, "loss": 0.0698, "step": 5820 }, { "epoch": 3.06, "learning_rate": 7.1606060606060615e-06, "loss": 0.068, "step": 5830 }, { "epoch": 3.06, "learning_rate": 7.155555555555556e-06, "loss": 0.0688, "step": 5840 }, { "epoch": 3.07, "learning_rate": 7.150505050505051e-06, "loss": 0.0674, "step": 5850 }, { "epoch": 3.07, "learning_rate": 7.145454545454547e-06, "loss": 0.07, "step": 5860 }, { "epoch": 3.08, "learning_rate": 7.140404040404041e-06, "loss": 0.069, "step": 5870 }, { "epoch": 3.08, "learning_rate": 7.135353535353536e-06, "loss": 0.067, "step": 5880 }, { "epoch": 3.09, "learning_rate": 7.130303030303031e-06, "loss": 0.0687, "step": 5890 }, { "epoch": 3.1, "learning_rate": 7.125252525252527e-06, "loss": 0.072, "step": 5900 }, { "epoch": 3.1, "eval_loss": 0.09600830078125, "eval_runtime": 201.6171, "eval_samples_per_second": 16.016, "eval_steps_per_second": 0.064, "eval_wer": 10.906646278769344, "step": 5900 }, { "epoch": 3.1, "learning_rate": 7.120202020202021e-06, "loss": 0.0694, "step": 5910 }, { "epoch": 3.11, "learning_rate": 7.115151515151516e-06, "loss": 0.0693, "step": 5920 }, { "epoch": 3.11, "learning_rate": 7.11010101010101e-06, "loss": 0.0697, "step": 5930 }, { "epoch": 3.12, "learning_rate": 7.1050505050505055e-06, "loss": 0.0716, "step": 5940 }, { "epoch": 3.12, "learning_rate": 7.100000000000001e-06, "loss": 0.0712, "step": 5950 }, { "epoch": 3.13, "learning_rate": 7.094949494949496e-06, "loss": 0.0698, "step": 5960 }, { "epoch": 3.13, "learning_rate": 7.08989898989899e-06, "loss": 0.0694, "step": 5970 }, { "epoch": 3.14, "learning_rate": 7.084848484848485e-06, "loss": 0.069, "step": 5980 }, { "epoch": 3.14, "learning_rate": 7.0797979797979795e-06, "loss": 0.0708, "step": 5990 }, { "epoch": 3.15, "learning_rate": 7.0747474747474755e-06, "loss": 0.0694, "step": 6000 }, { "epoch": 3.15, "eval_loss": 0.09478759765625, "eval_runtime": 191.2322, "eval_samples_per_second": 16.885, "eval_steps_per_second": 0.068, "eval_wer": 10.938282671166064, "step": 6000 }, { "epoch": 3.15, "learning_rate": 7.070707070707071e-06, "loss": 0.0708, "step": 6010 }, { "epoch": 3.16, "learning_rate": 7.065656565656566e-06, "loss": 0.0669, "step": 6020 }, { "epoch": 3.16, "learning_rate": 7.060606060606061e-06, "loss": 0.07, "step": 6030 }, { "epoch": 3.17, "learning_rate": 7.055555555555557e-06, "loss": 0.0678, "step": 6040 }, { "epoch": 3.17, "learning_rate": 7.050505050505051e-06, "loss": 0.0703, "step": 6050 }, { "epoch": 3.18, "learning_rate": 7.045454545454546e-06, "loss": 0.0715, "step": 6060 }, { "epoch": 3.18, "learning_rate": 7.0404040404040404e-06, "loss": 0.0655, "step": 6070 }, { "epoch": 3.19, "learning_rate": 7.0353535353535364e-06, "loss": 0.0702, "step": 6080 }, { "epoch": 3.2, "learning_rate": 7.030303030303031e-06, "loss": 0.0683, "step": 6090 }, { "epoch": 3.2, "learning_rate": 7.025252525252526e-06, "loss": 0.0715, "step": 6100 }, { "epoch": 3.2, "eval_loss": 0.0955810546875, "eval_runtime": 156.6454, "eval_samples_per_second": 20.613, "eval_steps_per_second": 0.083, "eval_wer": 10.848646226042023, "step": 6100 }, { "epoch": 3.21, "learning_rate": 7.02020202020202e-06, "loss": 0.0695, "step": 6110 }, { "epoch": 3.21, "learning_rate": 7.015151515151516e-06, "loss": 0.0706, "step": 6120 }, { "epoch": 3.22, "learning_rate": 7.0101010101010105e-06, "loss": 0.0685, "step": 6130 }, { "epoch": 3.22, "learning_rate": 7.005050505050506e-06, "loss": 0.0689, "step": 6140 }, { "epoch": 3.23, "learning_rate": 7e-06, "loss": 0.0695, "step": 6150 }, { "epoch": 3.23, "learning_rate": 6.994949494949496e-06, "loss": 0.0656, "step": 6160 }, { "epoch": 3.24, "learning_rate": 6.98989898989899e-06, "loss": 0.0716, "step": 6170 }, { "epoch": 3.24, "learning_rate": 6.984848484848485e-06, "loss": 0.0675, "step": 6180 }, { "epoch": 3.25, "learning_rate": 6.979797979797981e-06, "loss": 0.0703, "step": 6190 }, { "epoch": 3.25, "learning_rate": 6.974747474747476e-06, "loss": 0.0688, "step": 6200 }, { "epoch": 3.25, "eval_loss": 0.09490966796875, "eval_runtime": 152.9521, "eval_samples_per_second": 21.111, "eval_steps_per_second": 0.085, "eval_wer": 10.782737075215524, "step": 6200 }, { "epoch": 3.26, "learning_rate": 6.969696969696971e-06, "loss": 0.0674, "step": 6210 }, { "epoch": 3.26, "learning_rate": 6.964646464646465e-06, "loss": 0.0672, "step": 6220 }, { "epoch": 3.27, "learning_rate": 6.95959595959596e-06, "loss": 0.0705, "step": 6230 }, { "epoch": 3.27, "learning_rate": 6.954545454545455e-06, "loss": 0.0715, "step": 6240 }, { "epoch": 3.28, "learning_rate": 6.9494949494949505e-06, "loss": 0.0672, "step": 6250 }, { "epoch": 3.28, "learning_rate": 6.944444444444445e-06, "loss": 0.0713, "step": 6260 }, { "epoch": 3.29, "learning_rate": 6.93939393939394e-06, "loss": 0.0703, "step": 6270 }, { "epoch": 3.29, "learning_rate": 6.934343434343434e-06, "loss": 0.0681, "step": 6280 }, { "epoch": 3.3, "learning_rate": 6.92929292929293e-06, "loss": 0.0687, "step": 6290 }, { "epoch": 3.31, "learning_rate": 6.9242424242424245e-06, "loss": 0.0707, "step": 6300 }, { "epoch": 3.31, "eval_loss": 0.09539794921875, "eval_runtime": 200.4855, "eval_samples_per_second": 16.106, "eval_steps_per_second": 0.065, "eval_wer": 10.772191611083283, "step": 6300 }, { "epoch": 3.31, "learning_rate": 6.91919191919192e-06, "loss": 0.0691, "step": 6310 }, { "epoch": 3.32, "learning_rate": 6.914141414141414e-06, "loss": 0.0679, "step": 6320 }, { "epoch": 3.32, "learning_rate": 6.90909090909091e-06, "loss": 0.0712, "step": 6330 }, { "epoch": 3.33, "learning_rate": 6.904040404040404e-06, "loss": 0.0682, "step": 6340 }, { "epoch": 3.33, "learning_rate": 6.898989898989899e-06, "loss": 0.0719, "step": 6350 }, { "epoch": 3.34, "learning_rate": 6.893939393939395e-06, "loss": 0.0687, "step": 6360 }, { "epoch": 3.34, "learning_rate": 6.88888888888889e-06, "loss": 0.071, "step": 6370 }, { "epoch": 3.35, "learning_rate": 6.883838383838385e-06, "loss": 0.0696, "step": 6380 }, { "epoch": 3.35, "learning_rate": 6.878787878787879e-06, "loss": 0.0712, "step": 6390 }, { "epoch": 3.36, "learning_rate": 6.873737373737375e-06, "loss": 0.0703, "step": 6400 }, { "epoch": 3.36, "eval_loss": 0.0947265625, "eval_runtime": 406.5772, "eval_samples_per_second": 7.942, "eval_steps_per_second": 0.032, "eval_wer": 10.769555245050224, "step": 6400 }, { "epoch": 3.36, "learning_rate": 6.868686868686869e-06, "loss": 0.0693, "step": 6410 }, { "epoch": 3.37, "learning_rate": 6.8636363636363645e-06, "loss": 0.0676, "step": 6420 }, { "epoch": 3.37, "learning_rate": 6.858585858585859e-06, "loss": 0.069, "step": 6430 }, { "epoch": 3.38, "learning_rate": 6.853535353535354e-06, "loss": 0.0664, "step": 6440 }, { "epoch": 3.38, "learning_rate": 6.848484848484849e-06, "loss": 0.0699, "step": 6450 }, { "epoch": 3.39, "learning_rate": 6.843434343434344e-06, "loss": 0.0685, "step": 6460 }, { "epoch": 3.39, "learning_rate": 6.8383838383838386e-06, "loss": 0.0712, "step": 6470 }, { "epoch": 3.4, "learning_rate": 6.833333333333334e-06, "loss": 0.0699, "step": 6480 }, { "epoch": 3.41, "learning_rate": 6.828282828282828e-06, "loss": 0.0688, "step": 6490 }, { "epoch": 3.41, "learning_rate": 6.823232323232324e-06, "loss": 0.0683, "step": 6500 }, { "epoch": 3.41, "eval_loss": 0.09368896484375, "eval_runtime": 343.8939, "eval_samples_per_second": 9.39, "eval_steps_per_second": 0.038, "eval_wer": 10.838100761909782, "step": 6500 }, { "epoch": 3.42, "learning_rate": 6.818181818181818e-06, "loss": 0.0699, "step": 6510 }, { "epoch": 3.42, "learning_rate": 6.813131313131313e-06, "loss": 0.07, "step": 6520 }, { "epoch": 3.43, "learning_rate": 6.808080808080809e-06, "loss": 0.0677, "step": 6530 }, { "epoch": 3.43, "learning_rate": 6.803030303030304e-06, "loss": 0.0693, "step": 6540 }, { "epoch": 3.44, "learning_rate": 6.797979797979799e-06, "loss": 0.0686, "step": 6550 }, { "epoch": 3.44, "learning_rate": 6.792929292929293e-06, "loss": 0.0695, "step": 6560 }, { "epoch": 3.45, "learning_rate": 6.787878787878789e-06, "loss": 0.0667, "step": 6570 }, { "epoch": 3.45, "learning_rate": 6.7828282828282834e-06, "loss": 0.0708, "step": 6580 }, { "epoch": 3.46, "learning_rate": 6.777777777777779e-06, "loss": 0.0688, "step": 6590 }, { "epoch": 3.46, "learning_rate": 6.772727272727273e-06, "loss": 0.0697, "step": 6600 }, { "epoch": 3.46, "eval_loss": 0.093017578125, "eval_runtime": 317.7379, "eval_samples_per_second": 10.162, "eval_steps_per_second": 0.041, "eval_wer": 10.711555192322903, "step": 6600 }, { "epoch": 3.47, "learning_rate": 6.767676767676769e-06, "loss": 0.0691, "step": 6610 }, { "epoch": 3.47, "learning_rate": 6.762626262626263e-06, "loss": 0.0698, "step": 6620 }, { "epoch": 3.48, "learning_rate": 6.757575757575758e-06, "loss": 0.0694, "step": 6630 }, { "epoch": 3.48, "learning_rate": 6.752525252525253e-06, "loss": 0.0665, "step": 6640 }, { "epoch": 3.49, "learning_rate": 6.747474747474749e-06, "loss": 0.0677, "step": 6650 }, { "epoch": 3.49, "learning_rate": 6.742424242424243e-06, "loss": 0.0694, "step": 6660 }, { "epoch": 3.5, "learning_rate": 6.737373737373738e-06, "loss": 0.0705, "step": 6670 }, { "epoch": 3.5, "learning_rate": 6.732323232323232e-06, "loss": 0.0678, "step": 6680 }, { "epoch": 3.51, "learning_rate": 6.7272727272727275e-06, "loss": 0.0696, "step": 6690 }, { "epoch": 3.52, "learning_rate": 6.7222222222222235e-06, "loss": 0.0695, "step": 6700 }, { "epoch": 3.52, "eval_loss": 0.0933837890625, "eval_runtime": 417.2918, "eval_samples_per_second": 7.738, "eval_steps_per_second": 0.031, "eval_wer": 10.885555350504864, "step": 6700 }, { "epoch": 3.52, "learning_rate": 6.717171717171718e-06, "loss": 0.0701, "step": 6710 }, { "epoch": 3.53, "learning_rate": 6.712121212121213e-06, "loss": 0.0674, "step": 6720 }, { "epoch": 3.53, "learning_rate": 6.707070707070707e-06, "loss": 0.0703, "step": 6730 }, { "epoch": 3.54, "learning_rate": 6.702020202020203e-06, "loss": 0.0666, "step": 6740 }, { "epoch": 3.54, "learning_rate": 6.6969696969696975e-06, "loss": 0.0676, "step": 6750 }, { "epoch": 3.55, "learning_rate": 6.691919191919193e-06, "loss": 0.0686, "step": 6760 }, { "epoch": 3.55, "learning_rate": 6.686868686868687e-06, "loss": 0.0672, "step": 6770 }, { "epoch": 3.56, "learning_rate": 6.681818181818183e-06, "loss": 0.0682, "step": 6780 }, { "epoch": 3.56, "learning_rate": 6.676767676767677e-06, "loss": 0.068, "step": 6790 }, { "epoch": 3.57, "learning_rate": 6.671717171717172e-06, "loss": 0.0699, "step": 6800 }, { "epoch": 3.57, "eval_loss": 0.0936279296875, "eval_runtime": 456.4445, "eval_samples_per_second": 7.074, "eval_steps_per_second": 0.028, "eval_wer": 10.740555218686563, "step": 6800 }, { "epoch": 3.57, "learning_rate": 6.666666666666667e-06, "loss": 0.069, "step": 6810 }, { "epoch": 3.58, "learning_rate": 6.661616161616163e-06, "loss": 0.0669, "step": 6820 }, { "epoch": 3.58, "learning_rate": 6.656565656565657e-06, "loss": 0.0661, "step": 6830 }, { "epoch": 3.59, "learning_rate": 6.651515151515152e-06, "loss": 0.0676, "step": 6840 }, { "epoch": 3.59, "learning_rate": 6.646464646464646e-06, "loss": 0.0664, "step": 6850 }, { "epoch": 3.6, "learning_rate": 6.641414141414142e-06, "loss": 0.0687, "step": 6860 }, { "epoch": 3.6, "learning_rate": 6.6363636363636375e-06, "loss": 0.0703, "step": 6870 }, { "epoch": 3.61, "learning_rate": 6.631313131313132e-06, "loss": 0.0702, "step": 6880 }, { "epoch": 3.61, "learning_rate": 6.626262626262627e-06, "loss": 0.0691, "step": 6890 }, { "epoch": 3.62, "learning_rate": 6.621212121212121e-06, "loss": 0.0704, "step": 6900 }, { "epoch": 3.62, "eval_loss": 0.0933837890625, "eval_runtime": 226.6631, "eval_samples_per_second": 14.246, "eval_steps_per_second": 0.057, "eval_wer": 10.751100682818802, "step": 6900 }, { "epoch": 3.63, "learning_rate": 6.616161616161617e-06, "loss": 0.0672, "step": 6910 }, { "epoch": 3.63, "learning_rate": 6.6111111111111115e-06, "loss": 0.0681, "step": 6920 }, { "epoch": 3.64, "learning_rate": 6.606060606060607e-06, "loss": 0.0694, "step": 6930 }, { "epoch": 3.64, "learning_rate": 6.601010101010101e-06, "loss": 0.0681, "step": 6940 }, { "epoch": 3.65, "learning_rate": 6.595959595959597e-06, "loss": 0.0695, "step": 6950 }, { "epoch": 3.65, "learning_rate": 6.590909090909091e-06, "loss": 0.0658, "step": 6960 }, { "epoch": 3.66, "learning_rate": 6.585858585858586e-06, "loss": 0.0668, "step": 6970 }, { "epoch": 3.66, "learning_rate": 6.580808080808081e-06, "loss": 0.07, "step": 6980 }, { "epoch": 3.67, "learning_rate": 6.575757575757577e-06, "loss": 0.0684, "step": 6990 }, { "epoch": 3.67, "learning_rate": 6.570707070707071e-06, "loss": 0.0679, "step": 7000 }, { "epoch": 3.67, "eval_loss": 0.09307861328125, "eval_runtime": 487.6614, "eval_samples_per_second": 6.621, "eval_steps_per_second": 0.027, "eval_wer": 10.62191874719886, "step": 7000 }, { "epoch": 3.68, "learning_rate": 6.566161616161617e-06, "loss": 0.0711, "step": 7010 }, { "epoch": 3.68, "learning_rate": 6.561616161616162e-06, "loss": 0.0688, "step": 7020 }, { "epoch": 3.69, "learning_rate": 6.556565656565656e-06, "loss": 0.0699, "step": 7030 }, { "epoch": 3.69, "learning_rate": 6.551515151515152e-06, "loss": 0.0682, "step": 7040 }, { "epoch": 3.7, "learning_rate": 6.546464646464647e-06, "loss": 0.0682, "step": 7050 }, { "epoch": 3.7, "learning_rate": 6.541414141414142e-06, "loss": 0.0702, "step": 7060 }, { "epoch": 3.71, "learning_rate": 6.536363636363638e-06, "loss": 0.0695, "step": 7070 }, { "epoch": 3.71, "learning_rate": 6.531313131313132e-06, "loss": 0.068, "step": 7080 }, { "epoch": 3.72, "learning_rate": 6.526262626262627e-06, "loss": 0.0671, "step": 7090 }, { "epoch": 3.73, "learning_rate": 6.521212121212121e-06, "loss": 0.0678, "step": 7100 }, { "epoch": 3.73, "eval_loss": 0.09295654296875, "eval_runtime": 268.1778, "eval_samples_per_second": 12.041, "eval_steps_per_second": 0.048, "eval_wer": 10.756373414884923, "step": 7100 }, { "epoch": 3.73, "learning_rate": 6.516161616161617e-06, "loss": 0.0701, "step": 7110 }, { "epoch": 3.74, "learning_rate": 6.511111111111112e-06, "loss": 0.0704, "step": 7120 }, { "epoch": 3.74, "learning_rate": 6.506060606060607e-06, "loss": 0.069, "step": 7130 }, { "epoch": 3.75, "learning_rate": 6.501010101010101e-06, "loss": 0.0668, "step": 7140 }, { "epoch": 3.75, "learning_rate": 6.495959595959597e-06, "loss": 0.0707, "step": 7150 }, { "epoch": 3.76, "learning_rate": 6.490909090909091e-06, "loss": 0.0688, "step": 7160 }, { "epoch": 3.76, "learning_rate": 6.4858585858585865e-06, "loss": 0.0661, "step": 7170 }, { "epoch": 3.77, "learning_rate": 6.480808080808081e-06, "loss": 0.0673, "step": 7180 }, { "epoch": 3.77, "learning_rate": 6.475757575757576e-06, "loss": 0.0695, "step": 7190 }, { "epoch": 3.78, "learning_rate": 6.470707070707071e-06, "loss": 0.0692, "step": 7200 }, { "epoch": 3.78, "eval_loss": 0.0941162109375, "eval_runtime": 284.2293, "eval_samples_per_second": 11.361, "eval_steps_per_second": 0.046, "eval_wer": 10.669373335793942, "step": 7200 }, { "epoch": 3.78, "learning_rate": 6.465656565656566e-06, "loss": 0.0665, "step": 7210 }, { "epoch": 3.79, "learning_rate": 6.460606060606061e-06, "loss": 0.0692, "step": 7220 }, { "epoch": 3.79, "learning_rate": 6.455555555555556e-06, "loss": 0.0669, "step": 7230 }, { "epoch": 3.8, "learning_rate": 6.450505050505052e-06, "loss": 0.0698, "step": 7240 }, { "epoch": 3.8, "learning_rate": 6.445454545454546e-06, "loss": 0.0695, "step": 7250 }, { "epoch": 3.81, "learning_rate": 6.440404040404041e-06, "loss": 0.065, "step": 7260 }, { "epoch": 3.81, "learning_rate": 6.435353535353535e-06, "loss": 0.0697, "step": 7270 }, { "epoch": 3.82, "learning_rate": 6.430303030303031e-06, "loss": 0.0698, "step": 7280 }, { "epoch": 3.82, "learning_rate": 6.425252525252526e-06, "loss": 0.0687, "step": 7290 }, { "epoch": 3.83, "learning_rate": 6.420202020202021e-06, "loss": 0.0691, "step": 7300 }, { "epoch": 3.83, "eval_loss": 0.0926513671875, "eval_runtime": 272.5428, "eval_samples_per_second": 11.848, "eval_steps_per_second": 0.048, "eval_wer": 10.672009701827001, "step": 7300 }, { "epoch": 3.84, "learning_rate": 6.415151515151515e-06, "loss": 0.0669, "step": 7310 }, { "epoch": 3.84, "learning_rate": 6.410101010101011e-06, "loss": 0.0693, "step": 7320 }, { "epoch": 3.85, "learning_rate": 6.405050505050505e-06, "loss": 0.0678, "step": 7330 }, { "epoch": 3.85, "learning_rate": 6.4000000000000006e-06, "loss": 0.0665, "step": 7340 }, { "epoch": 3.86, "learning_rate": 6.394949494949495e-06, "loss": 0.0695, "step": 7350 }, { "epoch": 3.86, "learning_rate": 6.389898989898991e-06, "loss": 0.0685, "step": 7360 }, { "epoch": 3.87, "learning_rate": 6.384848484848485e-06, "loss": 0.0698, "step": 7370 }, { "epoch": 3.87, "learning_rate": 6.37979797979798e-06, "loss": 0.0698, "step": 7380 }, { "epoch": 3.88, "learning_rate": 6.3747474747474754e-06, "loss": 0.0648, "step": 7390 }, { "epoch": 3.88, "learning_rate": 6.3696969696969706e-06, "loss": 0.0692, "step": 7400 }, { "epoch": 3.88, "eval_loss": 0.0924072265625, "eval_runtime": 245.3497, "eval_samples_per_second": 13.161, "eval_steps_per_second": 0.053, "eval_wer": 10.62455511323192, "step": 7400 }, { "epoch": 3.89, "learning_rate": 6.364646464646466e-06, "loss": 0.0701, "step": 7410 }, { "epoch": 3.89, "learning_rate": 6.35959595959596e-06, "loss": 0.0673, "step": 7420 }, { "epoch": 3.9, "learning_rate": 6.354545454545455e-06, "loss": 0.0666, "step": 7430 }, { "epoch": 3.9, "learning_rate": 6.3494949494949494e-06, "loss": 0.0673, "step": 7440 }, { "epoch": 3.91, "learning_rate": 6.3444444444444454e-06, "loss": 0.0687, "step": 7450 }, { "epoch": 3.91, "learning_rate": 6.33939393939394e-06, "loss": 0.0681, "step": 7460 }, { "epoch": 3.92, "learning_rate": 6.334343434343435e-06, "loss": 0.0673, "step": 7470 }, { "epoch": 3.92, "learning_rate": 6.329292929292929e-06, "loss": 0.0689, "step": 7480 }, { "epoch": 3.93, "learning_rate": 6.324242424242425e-06, "loss": 0.0683, "step": 7490 }, { "epoch": 3.93, "learning_rate": 6.3191919191919195e-06, "loss": 0.0673, "step": 7500 }, { "epoch": 3.93, "eval_loss": 0.0908203125, "eval_runtime": 432.3008, "eval_samples_per_second": 7.469, "eval_steps_per_second": 0.03, "eval_wer": 10.53755503414094, "step": 7500 }, { "epoch": 3.94, "learning_rate": 6.314141414141415e-06, "loss": 0.0675, "step": 7510 }, { "epoch": 3.95, "learning_rate": 6.309090909090909e-06, "loss": 0.068, "step": 7520 }, { "epoch": 3.95, "learning_rate": 6.304040404040405e-06, "loss": 0.0641, "step": 7530 }, { "epoch": 3.96, "learning_rate": 6.298989898989899e-06, "loss": 0.0687, "step": 7540 }, { "epoch": 3.96, "learning_rate": 6.293939393939394e-06, "loss": 0.0687, "step": 7550 }, { "epoch": 3.97, "learning_rate": 6.28888888888889e-06, "loss": 0.0649, "step": 7560 }, { "epoch": 3.97, "learning_rate": 6.283838383838385e-06, "loss": 0.07, "step": 7570 }, { "epoch": 3.98, "learning_rate": 6.27878787878788e-06, "loss": 0.0692, "step": 7580 }, { "epoch": 3.98, "learning_rate": 6.273737373737374e-06, "loss": 0.0688, "step": 7590 }, { "epoch": 3.99, "learning_rate": 6.268686868686869e-06, "loss": 0.0686, "step": 7600 }, { "epoch": 3.99, "eval_loss": 0.09173583984375, "eval_runtime": 409.8227, "eval_samples_per_second": 7.879, "eval_steps_per_second": 0.032, "eval_wer": 10.595555086868261, "step": 7600 }, { "epoch": 3.99, "learning_rate": 6.263636363636364e-06, "loss": 0.0671, "step": 7610 }, { "epoch": 4.0, "learning_rate": 6.2585858585858595e-06, "loss": 0.0708, "step": 7620 }, { "epoch": 4.0, "learning_rate": 6.253535353535354e-06, "loss": 0.0646, "step": 7630 }, { "epoch": 4.01, "learning_rate": 6.248484848484849e-06, "loss": 0.0647, "step": 7640 }, { "epoch": 4.01, "learning_rate": 6.243434343434343e-06, "loss": 0.0638, "step": 7650 }, { "epoch": 4.02, "learning_rate": 6.238383838383839e-06, "loss": 0.0623, "step": 7660 }, { "epoch": 4.02, "learning_rate": 6.2333333333333335e-06, "loss": 0.0671, "step": 7670 }, { "epoch": 4.03, "learning_rate": 6.228282828282829e-06, "loss": 0.0642, "step": 7680 }, { "epoch": 4.03, "learning_rate": 6.223232323232323e-06, "loss": 0.0637, "step": 7690 }, { "epoch": 4.04, "learning_rate": 6.218181818181819e-06, "loss": 0.0644, "step": 7700 }, { "epoch": 4.04, "eval_loss": 0.0924072265625, "eval_runtime": 168.9023, "eval_samples_per_second": 19.118, "eval_steps_per_second": 0.077, "eval_wer": 10.57710052463684, "step": 7700 }, { "epoch": 4.05, "learning_rate": 6.213131313131313e-06, "loss": 0.065, "step": 7710 }, { "epoch": 4.05, "learning_rate": 6.208080808080808e-06, "loss": 0.0655, "step": 7720 }, { "epoch": 4.06, "learning_rate": 6.203030303030304e-06, "loss": 0.066, "step": 7730 }, { "epoch": 4.06, "learning_rate": 6.197979797979799e-06, "loss": 0.0629, "step": 7740 }, { "epoch": 4.07, "learning_rate": 6.192929292929294e-06, "loss": 0.0647, "step": 7750 }, { "epoch": 4.07, "learning_rate": 6.187878787878788e-06, "loss": 0.0627, "step": 7760 }, { "epoch": 4.08, "learning_rate": 6.182828282828284e-06, "loss": 0.0632, "step": 7770 }, { "epoch": 4.08, "learning_rate": 6.177777777777778e-06, "loss": 0.0619, "step": 7780 }, { "epoch": 4.09, "learning_rate": 6.1727272727272735e-06, "loss": 0.0665, "step": 7790 }, { "epoch": 4.09, "learning_rate": 6.167676767676768e-06, "loss": 0.063, "step": 7800 }, { "epoch": 4.09, "eval_loss": 0.09228515625, "eval_runtime": 2057.1462, "eval_samples_per_second": 1.57, "eval_steps_per_second": 0.006, "eval_wer": 10.48482771347974, "step": 7800 }, { "epoch": 4.1, "learning_rate": 6.162626262626264e-06, "loss": 0.0627, "step": 7810 }, { "epoch": 4.1, "learning_rate": 6.157575757575758e-06, "loss": 0.0653, "step": 7820 }, { "epoch": 4.11, "learning_rate": 6.152525252525253e-06, "loss": 0.0638, "step": 7830 }, { "epoch": 4.11, "learning_rate": 6.1474747474747476e-06, "loss": 0.0639, "step": 7840 }, { "epoch": 4.12, "learning_rate": 6.142424242424243e-06, "loss": 0.0651, "step": 7850 }, { "epoch": 4.12, "learning_rate": 6.137373737373738e-06, "loss": 0.0641, "step": 7860 }, { "epoch": 4.13, "learning_rate": 6.132323232323233e-06, "loss": 0.0648, "step": 7870 }, { "epoch": 4.13, "learning_rate": 6.127272727272727e-06, "loss": 0.0635, "step": 7880 }, { "epoch": 4.14, "learning_rate": 6.1222222222222224e-06, "loss": 0.0646, "step": 7890 }, { "epoch": 4.14, "learning_rate": 6.117171717171718e-06, "loss": 0.0646, "step": 7900 }, { "epoch": 4.14, "eval_loss": 0.091796875, "eval_runtime": 309.7183, "eval_samples_per_second": 10.426, "eval_steps_per_second": 0.042, "eval_wer": 10.5270095700087, "step": 7900 }, { "epoch": 4.15, "learning_rate": 6.112121212121213e-06, "loss": 0.0628, "step": 7910 }, { "epoch": 4.16, "learning_rate": 6.107070707070708e-06, "loss": 0.0653, "step": 7920 }, { "epoch": 4.16, "learning_rate": 6.102020202020202e-06, "loss": 0.0636, "step": 7930 }, { "epoch": 4.17, "learning_rate": 6.096969696969698e-06, "loss": 0.0615, "step": 7940 }, { "epoch": 4.17, "learning_rate": 6.0919191919191924e-06, "loss": 0.0625, "step": 7950 }, { "epoch": 4.18, "learning_rate": 6.086868686868688e-06, "loss": 0.0632, "step": 7960 }, { "epoch": 4.18, "learning_rate": 6.081818181818182e-06, "loss": 0.0651, "step": 7970 }, { "epoch": 4.19, "learning_rate": 6.076767676767678e-06, "loss": 0.0648, "step": 7980 }, { "epoch": 4.19, "learning_rate": 6.071717171717172e-06, "loss": 0.0631, "step": 7990 }, { "epoch": 4.2, "learning_rate": 6.066666666666667e-06, "loss": 0.0656, "step": 8000 }, { "epoch": 4.2, "eval_loss": 0.09222412109375, "eval_runtime": 209.3674, "eval_samples_per_second": 15.423, "eval_steps_per_second": 0.062, "eval_wer": 10.643009675463341, "step": 8000 }, { "epoch": 4.2, "learning_rate": 6.061616161616162e-06, "loss": 0.0634, "step": 8010 }, { "epoch": 4.21, "learning_rate": 6.057575757575757e-06, "loss": 0.0637, "step": 8020 }, { "epoch": 4.21, "learning_rate": 6.052525252525253e-06, "loss": 0.066, "step": 8030 }, { "epoch": 4.22, "learning_rate": 6.047474747474748e-06, "loss": 0.0643, "step": 8040 }, { "epoch": 4.22, "learning_rate": 6.042424242424243e-06, "loss": 0.0645, "step": 8050 }, { "epoch": 4.23, "learning_rate": 6.037373737373737e-06, "loss": 0.0645, "step": 8060 }, { "epoch": 4.23, "learning_rate": 6.032323232323233e-06, "loss": 0.0644, "step": 8070 }, { "epoch": 4.24, "learning_rate": 6.027272727272728e-06, "loss": 0.0627, "step": 8080 }, { "epoch": 4.24, "learning_rate": 6.0222222222222225e-06, "loss": 0.0644, "step": 8090 }, { "epoch": 4.25, "learning_rate": 6.0171717171717185e-06, "loss": 0.0644, "step": 8100 }, { "epoch": 4.25, "eval_loss": 0.0919189453125, "eval_runtime": 266.5495, "eval_samples_per_second": 12.114, "eval_steps_per_second": 0.049, "eval_wer": 10.669373335793942, "step": 8100 }, { "epoch": 4.25, "learning_rate": 6.012121212121213e-06, "loss": 0.0658, "step": 8110 }, { "epoch": 4.26, "learning_rate": 6.007070707070708e-06, "loss": 0.0631, "step": 8120 }, { "epoch": 4.27, "learning_rate": 6.002020202020202e-06, "loss": 0.0634, "step": 8130 }, { "epoch": 4.27, "learning_rate": 5.996969696969697e-06, "loss": 0.064, "step": 8140 }, { "epoch": 4.28, "learning_rate": 5.9919191919191926e-06, "loss": 0.0641, "step": 8150 }, { "epoch": 4.28, "learning_rate": 5.986868686868688e-06, "loss": 0.0648, "step": 8160 }, { "epoch": 4.29, "learning_rate": 5.981818181818182e-06, "loss": 0.0636, "step": 8170 }, { "epoch": 4.29, "learning_rate": 5.976767676767677e-06, "loss": 0.0643, "step": 8180 }, { "epoch": 4.3, "learning_rate": 5.9717171717171714e-06, "loss": 0.0649, "step": 8190 }, { "epoch": 4.3, "learning_rate": 5.966666666666667e-06, "loss": 0.0655, "step": 8200 }, { "epoch": 4.3, "eval_loss": 0.0919189453125, "eval_runtime": 212.5594, "eval_samples_per_second": 15.191, "eval_steps_per_second": 0.061, "eval_wer": 10.492736811578919, "step": 8200 }, { "epoch": 4.31, "learning_rate": 5.961616161616162e-06, "loss": 0.0638, "step": 8210 }, { "epoch": 4.31, "learning_rate": 5.956565656565657e-06, "loss": 0.0632, "step": 8220 }, { "epoch": 4.32, "learning_rate": 5.951515151515151e-06, "loss": 0.0649, "step": 8230 }, { "epoch": 4.32, "learning_rate": 5.946464646464647e-06, "loss": 0.0618, "step": 8240 }, { "epoch": 4.33, "learning_rate": 5.941414141414142e-06, "loss": 0.064, "step": 8250 }, { "epoch": 4.33, "learning_rate": 5.936363636363637e-06, "loss": 0.0645, "step": 8260 }, { "epoch": 4.34, "learning_rate": 5.9313131313131326e-06, "loss": 0.0623, "step": 8270 }, { "epoch": 4.34, "learning_rate": 5.926262626262627e-06, "loss": 0.0646, "step": 8280 }, { "epoch": 4.35, "learning_rate": 5.921212121212122e-06, "loss": 0.0633, "step": 8290 }, { "epoch": 4.35, "learning_rate": 5.916161616161616e-06, "loss": 0.0644, "step": 8300 }, { "epoch": 4.35, "eval_loss": 0.09149169921875, "eval_runtime": 312.8825, "eval_samples_per_second": 10.32, "eval_steps_per_second": 0.042, "eval_wer": 10.57710052463684, "step": 8300 }, { "epoch": 4.36, "learning_rate": 5.911111111111112e-06, "loss": 0.0622, "step": 8310 }, { "epoch": 4.37, "learning_rate": 5.906060606060607e-06, "loss": 0.0641, "step": 8320 }, { "epoch": 4.37, "learning_rate": 5.901010101010102e-06, "loss": 0.0641, "step": 8330 }, { "epoch": 4.38, "learning_rate": 5.895959595959596e-06, "loss": 0.0644, "step": 8340 }, { "epoch": 4.38, "learning_rate": 5.890909090909091e-06, "loss": 0.064, "step": 8350 }, { "epoch": 4.39, "learning_rate": 5.885858585858586e-06, "loss": 0.0634, "step": 8360 }, { "epoch": 4.39, "learning_rate": 5.8808080808080815e-06, "loss": 0.0645, "step": 8370 }, { "epoch": 4.4, "learning_rate": 5.875757575757576e-06, "loss": 0.0599, "step": 8380 }, { "epoch": 4.4, "learning_rate": 5.870707070707071e-06, "loss": 0.063, "step": 8390 }, { "epoch": 4.41, "learning_rate": 5.865656565656565e-06, "loss": 0.0645, "step": 8400 }, { "epoch": 4.41, "eval_loss": 0.09271240234375, "eval_runtime": 336.1727, "eval_samples_per_second": 9.605, "eval_steps_per_second": 0.039, "eval_wer": 10.635100577364161, "step": 8400 }, { "epoch": 4.41, "learning_rate": 5.860606060606061e-06, "loss": 0.0643, "step": 8410 }, { "epoch": 4.42, "learning_rate": 5.855555555555556e-06, "loss": 0.0642, "step": 8420 }, { "epoch": 4.42, "learning_rate": 5.850505050505051e-06, "loss": 0.0632, "step": 8430 }, { "epoch": 4.43, "learning_rate": 5.845454545454547e-06, "loss": 0.0645, "step": 8440 }, { "epoch": 4.43, "learning_rate": 5.840404040404041e-06, "loss": 0.0653, "step": 8450 }, { "epoch": 4.44, "learning_rate": 5.835353535353536e-06, "loss": 0.0644, "step": 8460 }, { "epoch": 4.44, "learning_rate": 5.83030303030303e-06, "loss": 0.0651, "step": 8470 }, { "epoch": 4.45, "learning_rate": 5.825252525252526e-06, "loss": 0.0632, "step": 8480 }, { "epoch": 4.45, "learning_rate": 5.820202020202021e-06, "loss": 0.0622, "step": 8490 }, { "epoch": 4.46, "learning_rate": 5.815151515151516e-06, "loss": 0.0655, "step": 8500 }, { "epoch": 4.46, "eval_loss": 0.0908203125, "eval_runtime": 179.8176, "eval_samples_per_second": 17.957, "eval_steps_per_second": 0.072, "eval_wer": 10.48219134744668, "step": 8500 }, { "epoch": 4.46, "learning_rate": 5.81010101010101e-06, "loss": 0.0662, "step": 8510 }, { "epoch": 4.47, "learning_rate": 5.805050505050506e-06, "loss": 0.0632, "step": 8520 }, { "epoch": 4.48, "learning_rate": 5.8e-06, "loss": 0.0618, "step": 8530 }, { "epoch": 4.48, "learning_rate": 5.7949494949494955e-06, "loss": 0.0649, "step": 8540 }, { "epoch": 4.49, "learning_rate": 5.78989898989899e-06, "loss": 0.064, "step": 8550 }, { "epoch": 4.49, "learning_rate": 5.784848484848486e-06, "loss": 0.0657, "step": 8560 }, { "epoch": 4.5, "learning_rate": 5.77979797979798e-06, "loss": 0.0645, "step": 8570 }, { "epoch": 4.5, "learning_rate": 5.774747474747475e-06, "loss": 0.0657, "step": 8580 }, { "epoch": 4.51, "learning_rate": 5.76969696969697e-06, "loss": 0.0634, "step": 8590 }, { "epoch": 4.51, "learning_rate": 5.764646464646465e-06, "loss": 0.0636, "step": 8600 }, { "epoch": 4.51, "eval_loss": 0.09185791015625, "eval_runtime": 183.7649, "eval_samples_per_second": 17.571, "eval_steps_per_second": 0.071, "eval_wer": 10.50591864174422, "step": 8600 }, { "epoch": 4.52, "learning_rate": 5.759595959595961e-06, "loss": 0.0642, "step": 8610 }, { "epoch": 4.52, "learning_rate": 5.754545454545455e-06, "loss": 0.0625, "step": 8620 }, { "epoch": 4.53, "learning_rate": 5.74949494949495e-06, "loss": 0.0635, "step": 8630 }, { "epoch": 4.53, "learning_rate": 5.744444444444444e-06, "loss": 0.0627, "step": 8640 }, { "epoch": 4.54, "learning_rate": 5.73939393939394e-06, "loss": 0.0643, "step": 8650 }, { "epoch": 4.54, "learning_rate": 5.734343434343435e-06, "loss": 0.0625, "step": 8660 }, { "epoch": 4.55, "learning_rate": 5.72929292929293e-06, "loss": 0.065, "step": 8670 }, { "epoch": 4.55, "learning_rate": 5.724242424242424e-06, "loss": 0.0653, "step": 8680 }, { "epoch": 4.56, "learning_rate": 5.71919191919192e-06, "loss": 0.062, "step": 8690 }, { "epoch": 4.56, "learning_rate": 5.714141414141414e-06, "loss": 0.0638, "step": 8700 }, { "epoch": 4.56, "eval_loss": 0.09088134765625, "eval_runtime": 169.082, "eval_samples_per_second": 19.097, "eval_steps_per_second": 0.077, "eval_wer": 10.45055495504996, "step": 8700 }, { "epoch": 4.57, "learning_rate": 5.7090909090909096e-06, "loss": 0.0632, "step": 8710 }, { "epoch": 4.58, "learning_rate": 5.704040404040404e-06, "loss": 0.064, "step": 8720 }, { "epoch": 4.58, "learning_rate": 5.6989898989899e-06, "loss": 0.0628, "step": 8730 }, { "epoch": 4.59, "learning_rate": 5.693939393939394e-06, "loss": 0.0663, "step": 8740 }, { "epoch": 4.59, "learning_rate": 5.688888888888889e-06, "loss": 0.0643, "step": 8750 }, { "epoch": 4.6, "learning_rate": 5.6838383838383844e-06, "loss": 0.0634, "step": 8760 }, { "epoch": 4.6, "learning_rate": 5.67878787878788e-06, "loss": 0.0632, "step": 8770 }, { "epoch": 4.61, "learning_rate": 5.673737373737375e-06, "loss": 0.0637, "step": 8780 }, { "epoch": 4.61, "learning_rate": 5.668686868686869e-06, "loss": 0.0659, "step": 8790 }, { "epoch": 4.62, "learning_rate": 5.663636363636364e-06, "loss": 0.0635, "step": 8800 }, { "epoch": 4.62, "eval_loss": 0.09234619140625, "eval_runtime": 175.2089, "eval_samples_per_second": 18.429, "eval_steps_per_second": 0.074, "eval_wer": 10.4874640795128, "step": 8800 }, { "epoch": 4.62, "learning_rate": 5.6585858585858585e-06, "loss": 0.0653, "step": 8810 }, { "epoch": 4.63, "learning_rate": 5.6535353535353544e-06, "loss": 0.064, "step": 8820 }, { "epoch": 4.63, "learning_rate": 5.648484848484849e-06, "loss": 0.0635, "step": 8830 }, { "epoch": 4.64, "learning_rate": 5.643434343434344e-06, "loss": 0.0624, "step": 8840 }, { "epoch": 4.64, "learning_rate": 5.638383838383838e-06, "loss": 0.0655, "step": 8850 }, { "epoch": 4.65, "learning_rate": 5.633333333333334e-06, "loss": 0.0654, "step": 8860 }, { "epoch": 4.65, "learning_rate": 5.6282828282828285e-06, "loss": 0.0647, "step": 8870 }, { "epoch": 4.66, "learning_rate": 5.623232323232324e-06, "loss": 0.0648, "step": 8880 }, { "epoch": 4.66, "learning_rate": 5.618181818181818e-06, "loss": 0.0656, "step": 8890 }, { "epoch": 4.67, "learning_rate": 5.613131313131314e-06, "loss": 0.0642, "step": 8900 }, { "epoch": 4.67, "eval_loss": 0.09088134765625, "eval_runtime": 158.7696, "eval_samples_per_second": 20.338, "eval_steps_per_second": 0.082, "eval_wer": 10.447918589016899, "step": 8900 }, { "epoch": 4.67, "learning_rate": 5.608080808080808e-06, "loss": 0.0633, "step": 8910 }, { "epoch": 4.68, "learning_rate": 5.603030303030303e-06, "loss": 0.0641, "step": 8920 }, { "epoch": 4.69, "learning_rate": 5.597979797979799e-06, "loss": 0.0641, "step": 8930 }, { "epoch": 4.69, "learning_rate": 5.592929292929294e-06, "loss": 0.0644, "step": 8940 }, { "epoch": 4.7, "learning_rate": 5.587878787878789e-06, "loss": 0.0637, "step": 8950 }, { "epoch": 4.7, "learning_rate": 5.582828282828283e-06, "loss": 0.0647, "step": 8960 }, { "epoch": 4.71, "learning_rate": 5.577777777777778e-06, "loss": 0.0631, "step": 8970 }, { "epoch": 4.71, "learning_rate": 5.572727272727273e-06, "loss": 0.0632, "step": 8980 }, { "epoch": 4.72, "learning_rate": 5.5676767676767685e-06, "loss": 0.0616, "step": 8990 }, { "epoch": 4.72, "learning_rate": 5.562626262626263e-06, "loss": 0.0638, "step": 9000 }, { "epoch": 4.72, "eval_loss": 0.0909423828125, "eval_runtime": 160.9683, "eval_samples_per_second": 20.06, "eval_steps_per_second": 0.081, "eval_wer": 10.440009490917719, "step": 9000 }, { "epoch": 4.73, "learning_rate": 5.557575757575758e-06, "loss": 0.0625, "step": 9010 }, { "epoch": 4.73, "learning_rate": 5.5535353535353546e-06, "loss": 0.0645, "step": 9020 }, { "epoch": 4.74, "learning_rate": 5.548484848484849e-06, "loss": 0.0644, "step": 9030 }, { "epoch": 4.74, "learning_rate": 5.543434343434344e-06, "loss": 0.0643, "step": 9040 }, { "epoch": 4.75, "learning_rate": 5.538383838383838e-06, "loss": 0.0624, "step": 9050 }, { "epoch": 4.75, "learning_rate": 5.533333333333334e-06, "loss": 0.0658, "step": 9060 }, { "epoch": 4.76, "learning_rate": 5.528282828282829e-06, "loss": 0.0643, "step": 9070 }, { "epoch": 4.76, "learning_rate": 5.523232323232324e-06, "loss": 0.0624, "step": 9080 }, { "epoch": 4.77, "learning_rate": 5.518181818181818e-06, "loss": 0.0633, "step": 9090 }, { "epoch": 4.77, "learning_rate": 5.513131313131313e-06, "loss": 0.0635, "step": 9100 }, { "epoch": 4.77, "eval_loss": 0.0904541015625, "eval_runtime": 172.0485, "eval_samples_per_second": 18.768, "eval_steps_per_second": 0.076, "eval_wer": 10.57710052463684, "step": 9100 }, { "epoch": 4.78, "learning_rate": 5.508080808080809e-06, "loss": 0.0625, "step": 9110 }, { "epoch": 4.78, "learning_rate": 5.5030303030303034e-06, "loss": 0.0641, "step": 9120 }, { "epoch": 4.79, "learning_rate": 5.497979797979799e-06, "loss": 0.0626, "step": 9130 }, { "epoch": 4.8, "learning_rate": 5.492929292929293e-06, "loss": 0.0639, "step": 9140 }, { "epoch": 4.8, "learning_rate": 5.487878787878789e-06, "loss": 0.0622, "step": 9150 }, { "epoch": 4.81, "learning_rate": 5.482828282828283e-06, "loss": 0.0632, "step": 9160 }, { "epoch": 4.81, "learning_rate": 5.477777777777778e-06, "loss": 0.0646, "step": 9170 }, { "epoch": 4.82, "learning_rate": 5.472727272727273e-06, "loss": 0.0644, "step": 9180 }, { "epoch": 4.82, "learning_rate": 5.467676767676769e-06, "loss": 0.0649, "step": 9190 }, { "epoch": 4.83, "learning_rate": 5.462626262626263e-06, "loss": 0.0634, "step": 9200 }, { "epoch": 4.83, "eval_loss": 0.09051513671875, "eval_runtime": 173.6186, "eval_samples_per_second": 18.598, "eval_steps_per_second": 0.075, "eval_wer": 10.4611004191822, "step": 9200 }, { "epoch": 4.83, "learning_rate": 5.457575757575758e-06, "loss": 0.063, "step": 9210 }, { "epoch": 4.84, "learning_rate": 5.452525252525252e-06, "loss": 0.0622, "step": 9220 }, { "epoch": 4.84, "learning_rate": 5.447474747474748e-06, "loss": 0.0639, "step": 9230 }, { "epoch": 4.85, "learning_rate": 5.442424242424243e-06, "loss": 0.0651, "step": 9240 }, { "epoch": 4.85, "learning_rate": 5.437373737373738e-06, "loss": 0.0629, "step": 9250 }, { "epoch": 4.86, "learning_rate": 5.432323232323232e-06, "loss": 0.0613, "step": 9260 }, { "epoch": 4.86, "learning_rate": 5.427272727272728e-06, "loss": 0.0621, "step": 9270 }, { "epoch": 4.87, "learning_rate": 5.422222222222223e-06, "loss": 0.0643, "step": 9280 }, { "epoch": 4.87, "learning_rate": 5.4171717171717175e-06, "loss": 0.0652, "step": 9290 }, { "epoch": 4.88, "learning_rate": 5.412121212121213e-06, "loss": 0.0636, "step": 9300 }, { "epoch": 4.88, "eval_loss": 0.09100341796875, "eval_runtime": 159.5182, "eval_samples_per_second": 20.242, "eval_steps_per_second": 0.081, "eval_wer": 10.469009517281378, "step": 9300 }, { "epoch": 4.88, "learning_rate": 5.407070707070707e-06, "loss": 0.0635, "step": 9310 }, { "epoch": 4.89, "learning_rate": 5.402020202020203e-06, "loss": 0.0647, "step": 9320 }, { "epoch": 4.9, "learning_rate": 5.396969696969697e-06, "loss": 0.0624, "step": 9330 }, { "epoch": 4.9, "learning_rate": 5.391919191919192e-06, "loss": 0.0661, "step": 9340 }, { "epoch": 4.91, "learning_rate": 5.386868686868687e-06, "loss": 0.0646, "step": 9350 }, { "epoch": 4.91, "learning_rate": 5.381818181818183e-06, "loss": 0.0639, "step": 9360 }, { "epoch": 4.92, "learning_rate": 5.376767676767677e-06, "loss": 0.0634, "step": 9370 }, { "epoch": 4.92, "learning_rate": 5.371717171717172e-06, "loss": 0.0649, "step": 9380 }, { "epoch": 4.93, "learning_rate": 5.366666666666666e-06, "loss": 0.0627, "step": 9390 }, { "epoch": 4.93, "learning_rate": 5.361616161616162e-06, "loss": 0.0624, "step": 9400 }, { "epoch": 4.93, "eval_loss": 0.09051513671875, "eval_runtime": 200.2744, "eval_samples_per_second": 16.123, "eval_steps_per_second": 0.065, "eval_wer": 10.432100392818539, "step": 9400 }, { "epoch": 4.94, "learning_rate": 5.356565656565657e-06, "loss": 0.0653, "step": 9410 }, { "epoch": 4.94, "learning_rate": 5.351515151515152e-06, "loss": 0.0639, "step": 9420 }, { "epoch": 4.95, "learning_rate": 5.346464646464646e-06, "loss": 0.0646, "step": 9430 }, { "epoch": 4.95, "learning_rate": 5.341414141414142e-06, "loss": 0.0622, "step": 9440 }, { "epoch": 4.96, "learning_rate": 5.336363636363637e-06, "loss": 0.0622, "step": 9450 }, { "epoch": 4.96, "learning_rate": 5.3313131313131315e-06, "loss": 0.0643, "step": 9460 }, { "epoch": 4.97, "learning_rate": 5.3262626262626275e-06, "loss": 0.0644, "step": 9470 }, { "epoch": 4.97, "learning_rate": 5.321212121212122e-06, "loss": 0.0647, "step": 9480 }, { "epoch": 4.98, "learning_rate": 5.316161616161617e-06, "loss": 0.0667, "step": 9490 }, { "epoch": 4.98, "learning_rate": 5.311111111111111e-06, "loss": 0.0619, "step": 9500 }, { "epoch": 4.98, "eval_loss": 0.09033203125, "eval_runtime": 161.6443, "eval_samples_per_second": 19.976, "eval_steps_per_second": 0.08, "eval_wer": 10.397827634388758, "step": 9500 }, { "epoch": 4.99, "learning_rate": 5.306060606060606e-06, "loss": 0.0649, "step": 9510 }, { "epoch": 4.99, "learning_rate": 5.3010101010101016e-06, "loss": 0.0629, "step": 9520 }, { "epoch": 5.0, "learning_rate": 5.295959595959597e-06, "loss": 0.0632, "step": 9530 }, { "epoch": 5.01, "learning_rate": 5.290909090909091e-06, "loss": 0.0596, "step": 9540 }, { "epoch": 5.01, "learning_rate": 5.285858585858586e-06, "loss": 0.0615, "step": 9550 }, { "epoch": 5.02, "learning_rate": 5.2808080808080804e-06, "loss": 0.0608, "step": 9560 }, { "epoch": 5.02, "learning_rate": 5.2757575757575764e-06, "loss": 0.0591, "step": 9570 }, { "epoch": 5.03, "learning_rate": 5.270707070707071e-06, "loss": 0.0594, "step": 9580 }, { "epoch": 5.03, "learning_rate": 5.265656565656566e-06, "loss": 0.0604, "step": 9590 }, { "epoch": 5.04, "learning_rate": 5.26060606060606e-06, "loss": 0.0608, "step": 9600 }, { "epoch": 5.04, "eval_loss": 0.09173583984375, "eval_runtime": 176.7516, "eval_samples_per_second": 18.269, "eval_steps_per_second": 0.074, "eval_wer": 10.400464000421819, "step": 9600 }, { "epoch": 5.04, "learning_rate": 5.255555555555556e-06, "loss": 0.0602, "step": 9610 }, { "epoch": 5.05, "learning_rate": 5.250505050505051e-06, "loss": 0.0603, "step": 9620 }, { "epoch": 5.05, "learning_rate": 5.245454545454546e-06, "loss": 0.0623, "step": 9630 }, { "epoch": 5.06, "learning_rate": 5.240404040404042e-06, "loss": 0.059, "step": 9640 }, { "epoch": 5.06, "learning_rate": 5.235353535353536e-06, "loss": 0.0603, "step": 9650 }, { "epoch": 5.07, "learning_rate": 5.230303030303031e-06, "loss": 0.0607, "step": 9660 }, { "epoch": 5.07, "learning_rate": 5.225252525252525e-06, "loss": 0.0608, "step": 9670 }, { "epoch": 5.08, "learning_rate": 5.220202020202021e-06, "loss": 0.0609, "step": 9680 }, { "epoch": 5.08, "learning_rate": 5.215151515151516e-06, "loss": 0.0588, "step": 9690 }, { "epoch": 5.09, "learning_rate": 5.210101010101011e-06, "loss": 0.0582, "step": 9700 }, { "epoch": 5.09, "eval_loss": 0.091064453125, "eval_runtime": 164.2191, "eval_samples_per_second": 19.663, "eval_steps_per_second": 0.079, "eval_wer": 10.45055495504996, "step": 9700 }, { "epoch": 5.09, "learning_rate": 5.205050505050505e-06, "loss": 0.0594, "step": 9710 }, { "epoch": 5.1, "learning_rate": 5.2e-06, "loss": 0.0577, "step": 9720 }, { "epoch": 5.1, "learning_rate": 5.194949494949495e-06, "loss": 0.0611, "step": 9730 }, { "epoch": 5.11, "learning_rate": 5.1898989898989905e-06, "loss": 0.0603, "step": 9740 }, { "epoch": 5.12, "learning_rate": 5.184848484848485e-06, "loss": 0.0598, "step": 9750 }, { "epoch": 5.12, "learning_rate": 5.17979797979798e-06, "loss": 0.0592, "step": 9760 }, { "epoch": 5.13, "learning_rate": 5.174747474747474e-06, "loss": 0.0582, "step": 9770 }, { "epoch": 5.13, "learning_rate": 5.16969696969697e-06, "loss": 0.0606, "step": 9780 }, { "epoch": 5.14, "learning_rate": 5.164646464646465e-06, "loss": 0.0596, "step": 9790 }, { "epoch": 5.14, "learning_rate": 5.15959595959596e-06, "loss": 0.0622, "step": 9800 }, { "epoch": 5.14, "eval_loss": 0.090087890625, "eval_runtime": 165.8673, "eval_samples_per_second": 19.467, "eval_steps_per_second": 0.078, "eval_wer": 10.42682766075242, "step": 9800 }, { "epoch": 5.15, "learning_rate": 5.154545454545456e-06, "loss": 0.0594, "step": 9810 }, { "epoch": 5.15, "learning_rate": 5.14949494949495e-06, "loss": 0.0627, "step": 9820 }, { "epoch": 5.16, "learning_rate": 5.144444444444445e-06, "loss": 0.06, "step": 9830 }, { "epoch": 5.16, "learning_rate": 5.139393939393939e-06, "loss": 0.0593, "step": 9840 }, { "epoch": 5.17, "learning_rate": 5.134343434343435e-06, "loss": 0.0601, "step": 9850 }, { "epoch": 5.17, "learning_rate": 5.12929292929293e-06, "loss": 0.0594, "step": 9860 }, { "epoch": 5.18, "learning_rate": 5.124242424242425e-06, "loss": 0.0614, "step": 9870 }, { "epoch": 5.18, "learning_rate": 5.119191919191919e-06, "loss": 0.0619, "step": 9880 }, { "epoch": 5.19, "learning_rate": 5.114141414141415e-06, "loss": 0.0618, "step": 9890 }, { "epoch": 5.19, "learning_rate": 5.109090909090909e-06, "loss": 0.0581, "step": 9900 }, { "epoch": 5.19, "eval_loss": 0.09136962890625, "eval_runtime": 177.1479, "eval_samples_per_second": 18.228, "eval_steps_per_second": 0.073, "eval_wer": 10.424191294719359, "step": 9900 }, { "epoch": 5.2, "learning_rate": 5.1040404040404045e-06, "loss": 0.0593, "step": 9910 }, { "epoch": 5.2, "learning_rate": 5.098989898989899e-06, "loss": 0.0597, "step": 9920 }, { "epoch": 5.21, "learning_rate": 5.093939393939395e-06, "loss": 0.0597, "step": 9930 }, { "epoch": 5.22, "learning_rate": 5.088888888888889e-06, "loss": 0.0593, "step": 9940 }, { "epoch": 5.22, "learning_rate": 5.083838383838384e-06, "loss": 0.0594, "step": 9950 }, { "epoch": 5.23, "learning_rate": 5.078787878787879e-06, "loss": 0.0607, "step": 9960 }, { "epoch": 5.23, "learning_rate": 5.073737373737374e-06, "loss": 0.0599, "step": 9970 }, { "epoch": 5.24, "learning_rate": 5.06868686868687e-06, "loss": 0.0596, "step": 9980 }, { "epoch": 5.24, "learning_rate": 5.063636363636364e-06, "loss": 0.0596, "step": 9990 }, { "epoch": 5.25, "learning_rate": 5.058585858585859e-06, "loss": 0.0602, "step": 10000 }, { "epoch": 5.25, "eval_loss": 0.0911865234375, "eval_runtime": 182.0565, "eval_samples_per_second": 17.736, "eval_steps_per_second": 0.071, "eval_wer": 10.371463974058157, "step": 10000 }, { "epoch": 5.25, "learning_rate": 5.053535353535353e-06, "loss": 0.0591, "step": 10010 }, { "epoch": 5.26, "learning_rate": 5.04949494949495e-06, "loss": 0.0599, "step": 10020 }, { "epoch": 5.26, "learning_rate": 5.044444444444445e-06, "loss": 0.0616, "step": 10030 }, { "epoch": 5.27, "learning_rate": 5.0393939393939395e-06, "loss": 0.0593, "step": 10040 }, { "epoch": 5.27, "learning_rate": 5.034343434343435e-06, "loss": 0.0608, "step": 10050 }, { "epoch": 5.28, "learning_rate": 5.029292929292929e-06, "loss": 0.0608, "step": 10060 }, { "epoch": 5.28, "learning_rate": 5.024242424242425e-06, "loss": 0.0612, "step": 10070 }, { "epoch": 5.29, "learning_rate": 5.019191919191919e-06, "loss": 0.0632, "step": 10080 }, { "epoch": 5.29, "learning_rate": 5.014141414141414e-06, "loss": 0.0588, "step": 10090 }, { "epoch": 5.3, "learning_rate": 5.009090909090909e-06, "loss": 0.0597, "step": 10100 }, { "epoch": 5.3, "eval_loss": 0.09130859375, "eval_runtime": 202.7835, "eval_samples_per_second": 15.923, "eval_steps_per_second": 0.064, "eval_wer": 10.321373019430018, "step": 10100 }, { "epoch": 5.3, "learning_rate": 5.004040404040405e-06, "loss": 0.0599, "step": 10110 }, { "epoch": 5.31, "learning_rate": 4.9989898989899e-06, "loss": 0.061, "step": 10120 }, { "epoch": 5.31, "learning_rate": 4.993939393939394e-06, "loss": 0.0601, "step": 10130 }, { "epoch": 5.32, "learning_rate": 4.988888888888889e-06, "loss": 0.061, "step": 10140 }, { "epoch": 5.33, "learning_rate": 4.983838383838384e-06, "loss": 0.061, "step": 10150 }, { "epoch": 5.33, "learning_rate": 4.978787878787879e-06, "loss": 0.0591, "step": 10160 }, { "epoch": 5.34, "learning_rate": 4.973737373737374e-06, "loss": 0.0597, "step": 10170 }, { "epoch": 5.34, "learning_rate": 4.968686868686869e-06, "loss": 0.0607, "step": 10180 }, { "epoch": 5.35, "learning_rate": 4.963636363636364e-06, "loss": 0.0602, "step": 10190 }, { "epoch": 5.35, "learning_rate": 4.958585858585858e-06, "loss": 0.0613, "step": 10200 }, { "epoch": 5.35, "eval_loss": 0.09033203125, "eval_runtime": 199.7934, "eval_samples_per_second": 16.162, "eval_steps_per_second": 0.065, "eval_wer": 10.268645698768818, "step": 10200 }, { "epoch": 5.36, "learning_rate": 4.953535353535354e-06, "loss": 0.0594, "step": 10210 }, { "epoch": 5.36, "learning_rate": 4.9484848484848495e-06, "loss": 0.0587, "step": 10220 }, { "epoch": 5.37, "learning_rate": 4.943434343434344e-06, "loss": 0.0589, "step": 10230 }, { "epoch": 5.37, "learning_rate": 4.938383838383839e-06, "loss": 0.0583, "step": 10240 }, { "epoch": 5.38, "learning_rate": 4.933333333333334e-06, "loss": 0.061, "step": 10250 }, { "epoch": 5.38, "learning_rate": 4.928282828282828e-06, "loss": 0.0604, "step": 10260 }, { "epoch": 5.39, "learning_rate": 4.9232323232323235e-06, "loss": 0.0599, "step": 10270 }, { "epoch": 5.39, "learning_rate": 4.918181818181819e-06, "loss": 0.0623, "step": 10280 }, { "epoch": 5.4, "learning_rate": 4.913131313131314e-06, "loss": 0.0631, "step": 10290 }, { "epoch": 5.4, "learning_rate": 4.908080808080808e-06, "loss": 0.0609, "step": 10300 }, { "epoch": 5.4, "eval_loss": 0.0914306640625, "eval_runtime": 176.0119, "eval_samples_per_second": 18.345, "eval_steps_per_second": 0.074, "eval_wer": 10.48219134744668, "step": 10300 }, { "epoch": 5.41, "learning_rate": 4.903030303030303e-06, "loss": 0.0602, "step": 10310 }, { "epoch": 5.41, "learning_rate": 4.897979797979798e-06, "loss": 0.0619, "step": 10320 }, { "epoch": 5.42, "learning_rate": 4.8929292929292936e-06, "loss": 0.0619, "step": 10330 }, { "epoch": 5.42, "learning_rate": 4.887878787878788e-06, "loss": 0.0607, "step": 10340 }, { "epoch": 5.43, "learning_rate": 4.882828282828283e-06, "loss": 0.0603, "step": 10350 }, { "epoch": 5.44, "learning_rate": 4.877777777777778e-06, "loss": 0.0625, "step": 10360 }, { "epoch": 5.44, "learning_rate": 4.872727272727273e-06, "loss": 0.0587, "step": 10370 }, { "epoch": 5.45, "learning_rate": 4.867676767676768e-06, "loss": 0.0618, "step": 10380 }, { "epoch": 5.45, "learning_rate": 4.8626262626262636e-06, "loss": 0.0631, "step": 10390 }, { "epoch": 5.46, "learning_rate": 4.857575757575758e-06, "loss": 0.059, "step": 10400 }, { "epoch": 5.46, "eval_loss": 0.09124755859375, "eval_runtime": 227.6564, "eval_samples_per_second": 14.184, "eval_steps_per_second": 0.057, "eval_wer": 10.408373098520999, "step": 10400 }, { "epoch": 5.46, "learning_rate": 4.852525252525253e-06, "loss": 0.0602, "step": 10410 }, { "epoch": 5.47, "learning_rate": 4.847474747474748e-06, "loss": 0.0572, "step": 10420 }, { "epoch": 5.47, "learning_rate": 4.842424242424243e-06, "loss": 0.0583, "step": 10430 }, { "epoch": 5.48, "learning_rate": 4.837373737373738e-06, "loss": 0.0595, "step": 10440 }, { "epoch": 5.48, "learning_rate": 4.832323232323233e-06, "loss": 0.0599, "step": 10450 }, { "epoch": 5.49, "learning_rate": 4.827272727272728e-06, "loss": 0.0612, "step": 10460 }, { "epoch": 5.49, "learning_rate": 4.822222222222222e-06, "loss": 0.0598, "step": 10470 }, { "epoch": 5.5, "learning_rate": 4.817171717171717e-06, "loss": 0.06, "step": 10480 }, { "epoch": 5.5, "learning_rate": 4.8121212121212125e-06, "loss": 0.0598, "step": 10490 }, { "epoch": 5.51, "learning_rate": 4.807070707070708e-06, "loss": 0.0579, "step": 10500 }, { "epoch": 5.51, "eval_loss": 0.092041015625, "eval_runtime": 223.3021, "eval_samples_per_second": 14.46, "eval_steps_per_second": 0.058, "eval_wer": 10.321373019430018, "step": 10500 }, { "epoch": 5.51, "learning_rate": 4.802020202020202e-06, "loss": 0.057, "step": 10510 }, { "epoch": 5.52, "learning_rate": 4.796969696969697e-06, "loss": 0.0607, "step": 10520 }, { "epoch": 5.52, "learning_rate": 4.791919191919192e-06, "loss": 0.0594, "step": 10530 }, { "epoch": 5.53, "learning_rate": 4.786868686868687e-06, "loss": 0.0615, "step": 10540 }, { "epoch": 5.54, "learning_rate": 4.7818181818181825e-06, "loss": 0.0601, "step": 10550 }, { "epoch": 5.54, "learning_rate": 4.776767676767678e-06, "loss": 0.0595, "step": 10560 }, { "epoch": 5.55, "learning_rate": 4.771717171717172e-06, "loss": 0.0608, "step": 10570 }, { "epoch": 5.55, "learning_rate": 4.766666666666667e-06, "loss": 0.0601, "step": 10580 }, { "epoch": 5.56, "learning_rate": 4.761616161616162e-06, "loss": 0.0632, "step": 10590 }, { "epoch": 5.56, "learning_rate": 4.756565656565657e-06, "loss": 0.0603, "step": 10600 }, { "epoch": 5.56, "eval_loss": 0.0911865234375, "eval_runtime": 172.4041, "eval_samples_per_second": 18.729, "eval_steps_per_second": 0.075, "eval_wer": 10.295009359099417, "step": 10600 }, { "epoch": 5.57, "learning_rate": 4.751515151515152e-06, "loss": 0.0608, "step": 10610 }, { "epoch": 5.57, "learning_rate": 4.746464646464647e-06, "loss": 0.0597, "step": 10620 }, { "epoch": 5.58, "learning_rate": 4.741414141414142e-06, "loss": 0.061, "step": 10630 }, { "epoch": 5.58, "learning_rate": 4.736363636363637e-06, "loss": 0.0595, "step": 10640 }, { "epoch": 5.59, "learning_rate": 4.731313131313131e-06, "loss": 0.0595, "step": 10650 }, { "epoch": 5.59, "learning_rate": 4.7262626262626265e-06, "loss": 0.0612, "step": 10660 }, { "epoch": 5.6, "learning_rate": 4.721212121212122e-06, "loss": 0.0593, "step": 10670 }, { "epoch": 5.6, "learning_rate": 4.716161616161617e-06, "loss": 0.0602, "step": 10680 }, { "epoch": 5.61, "learning_rate": 4.711111111111111e-06, "loss": 0.0595, "step": 10690 }, { "epoch": 5.61, "learning_rate": 4.706060606060606e-06, "loss": 0.059, "step": 10700 }, { "epoch": 5.61, "eval_loss": 0.090576171875, "eval_runtime": 289.2448, "eval_samples_per_second": 11.164, "eval_steps_per_second": 0.045, "eval_wer": 10.287100261000237, "step": 10700 }, { "epoch": 5.62, "learning_rate": 4.701010101010101e-06, "loss": 0.0582, "step": 10710 }, { "epoch": 5.62, "learning_rate": 4.6959595959595965e-06, "loss": 0.0596, "step": 10720 }, { "epoch": 5.63, "learning_rate": 4.690909090909092e-06, "loss": 0.0576, "step": 10730 }, { "epoch": 5.63, "learning_rate": 4.685858585858587e-06, "loss": 0.0607, "step": 10740 }, { "epoch": 5.64, "learning_rate": 4.680808080808081e-06, "loss": 0.0612, "step": 10750 }, { "epoch": 5.65, "learning_rate": 4.675757575757576e-06, "loss": 0.0588, "step": 10760 }, { "epoch": 5.65, "learning_rate": 4.670707070707071e-06, "loss": 0.0604, "step": 10770 }, { "epoch": 5.66, "learning_rate": 4.6656565656565665e-06, "loss": 0.0614, "step": 10780 }, { "epoch": 5.66, "learning_rate": 4.660606060606061e-06, "loss": 0.0612, "step": 10790 }, { "epoch": 5.67, "learning_rate": 4.655555555555556e-06, "loss": 0.0592, "step": 10800 }, { "epoch": 5.67, "eval_loss": 0.09051513671875, "eval_runtime": 177.9005, "eval_samples_per_second": 18.151, "eval_steps_per_second": 0.073, "eval_wer": 10.345100313727558, "step": 10800 }, { "epoch": 5.67, "learning_rate": 4.650505050505051e-06, "loss": 0.0613, "step": 10810 }, { "epoch": 5.68, "learning_rate": 4.645454545454545e-06, "loss": 0.0588, "step": 10820 }, { "epoch": 5.68, "learning_rate": 4.6404040404040406e-06, "loss": 0.0594, "step": 10830 }, { "epoch": 5.69, "learning_rate": 4.635353535353536e-06, "loss": 0.0599, "step": 10840 }, { "epoch": 5.69, "learning_rate": 4.630303030303031e-06, "loss": 0.0622, "step": 10850 }, { "epoch": 5.7, "learning_rate": 4.625252525252525e-06, "loss": 0.0613, "step": 10860 }, { "epoch": 5.7, "learning_rate": 4.62020202020202e-06, "loss": 0.0604, "step": 10870 }, { "epoch": 5.71, "learning_rate": 4.615151515151515e-06, "loss": 0.0599, "step": 10880 }, { "epoch": 5.71, "learning_rate": 4.6101010101010106e-06, "loss": 0.0604, "step": 10890 }, { "epoch": 5.72, "learning_rate": 4.605050505050506e-06, "loss": 0.0585, "step": 10900 }, { "epoch": 5.72, "eval_loss": 0.090576171875, "eval_runtime": 164.5832, "eval_samples_per_second": 19.619, "eval_steps_per_second": 0.079, "eval_wer": 10.313463921330838, "step": 10900 }, { "epoch": 5.72, "learning_rate": 4.600000000000001e-06, "loss": 0.0598, "step": 10910 }, { "epoch": 5.73, "learning_rate": 4.594949494949495e-06, "loss": 0.0624, "step": 10920 }, { "epoch": 5.73, "learning_rate": 4.58989898989899e-06, "loss": 0.0601, "step": 10930 }, { "epoch": 5.74, "learning_rate": 4.5848484848484854e-06, "loss": 0.0601, "step": 10940 }, { "epoch": 5.75, "learning_rate": 4.579797979797981e-06, "loss": 0.0597, "step": 10950 }, { "epoch": 5.75, "learning_rate": 4.574747474747475e-06, "loss": 0.059, "step": 10960 }, { "epoch": 5.76, "learning_rate": 4.56969696969697e-06, "loss": 0.0608, "step": 10970 }, { "epoch": 5.76, "learning_rate": 4.564646464646465e-06, "loss": 0.062, "step": 10980 }, { "epoch": 5.77, "learning_rate": 4.55959595959596e-06, "loss": 0.06, "step": 10990 }, { "epoch": 5.77, "learning_rate": 4.554545454545455e-06, "loss": 0.0608, "step": 11000 }, { "epoch": 5.77, "eval_loss": 0.09063720703125, "eval_runtime": 182.6745, "eval_samples_per_second": 17.676, "eval_steps_per_second": 0.071, "eval_wer": 10.258100234636577, "step": 11000 }, { "epoch": 5.78, "learning_rate": 4.54949494949495e-06, "loss": 0.062, "step": 11010 }, { "epoch": 5.78, "learning_rate": 4.5454545454545455e-06, "loss": 0.0604, "step": 11020 }, { "epoch": 5.79, "learning_rate": 4.540404040404041e-06, "loss": 0.06, "step": 11030 }, { "epoch": 5.79, "learning_rate": 4.535353535353536e-06, "loss": 0.0609, "step": 11040 }, { "epoch": 5.8, "learning_rate": 4.53030303030303e-06, "loss": 0.0611, "step": 11050 }, { "epoch": 5.8, "learning_rate": 4.525252525252526e-06, "loss": 0.0591, "step": 11060 }, { "epoch": 5.81, "learning_rate": 4.520202020202021e-06, "loss": 0.0604, "step": 11070 }, { "epoch": 5.81, "learning_rate": 4.5151515151515155e-06, "loss": 0.0607, "step": 11080 }, { "epoch": 5.82, "learning_rate": 4.510101010101011e-06, "loss": 0.062, "step": 11090 }, { "epoch": 5.82, "learning_rate": 4.505050505050506e-06, "loss": 0.0617, "step": 11100 }, { "epoch": 5.82, "eval_loss": 0.0908203125, "eval_runtime": 188.9731, "eval_samples_per_second": 17.087, "eval_steps_per_second": 0.069, "eval_wer": 10.310827555297777, "step": 11100 }, { "epoch": 5.83, "learning_rate": 4.5e-06, "loss": 0.0614, "step": 11110 }, { "epoch": 5.83, "learning_rate": 4.494949494949495e-06, "loss": 0.0601, "step": 11120 }, { "epoch": 5.84, "learning_rate": 4.48989898989899e-06, "loss": 0.0587, "step": 11130 }, { "epoch": 5.84, "learning_rate": 4.4848484848484855e-06, "loss": 0.0616, "step": 11140 }, { "epoch": 5.85, "learning_rate": 4.47979797979798e-06, "loss": 0.0604, "step": 11150 }, { "epoch": 5.86, "learning_rate": 4.474747474747475e-06, "loss": 0.0606, "step": 11160 }, { "epoch": 5.86, "learning_rate": 4.46969696969697e-06, "loss": 0.0592, "step": 11170 }, { "epoch": 5.87, "learning_rate": 4.464646464646465e-06, "loss": 0.0577, "step": 11180 }, { "epoch": 5.87, "learning_rate": 4.4595959595959596e-06, "loss": 0.0597, "step": 11190 }, { "epoch": 5.88, "learning_rate": 4.454545454545455e-06, "loss": 0.0593, "step": 11200 }, { "epoch": 5.88, "eval_loss": 0.09027099609375, "eval_runtime": 190.6834, "eval_samples_per_second": 16.934, "eval_steps_per_second": 0.068, "eval_wer": 10.179009253644775, "step": 11200 }, { "epoch": 5.88, "learning_rate": 4.44949494949495e-06, "loss": 0.0588, "step": 11210 }, { "epoch": 5.89, "learning_rate": 4.444444444444444e-06, "loss": 0.0608, "step": 11220 }, { "epoch": 5.89, "learning_rate": 4.43939393939394e-06, "loss": 0.0602, "step": 11230 }, { "epoch": 5.9, "learning_rate": 4.434343434343435e-06, "loss": 0.0594, "step": 11240 }, { "epoch": 5.9, "learning_rate": 4.42929292929293e-06, "loss": 0.0627, "step": 11250 }, { "epoch": 5.91, "learning_rate": 4.424242424242425e-06, "loss": 0.06, "step": 11260 }, { "epoch": 5.91, "learning_rate": 4.41919191919192e-06, "loss": 0.0592, "step": 11270 }, { "epoch": 5.92, "learning_rate": 4.414141414141415e-06, "loss": 0.0614, "step": 11280 }, { "epoch": 5.92, "learning_rate": 4.409090909090909e-06, "loss": 0.0594, "step": 11290 }, { "epoch": 5.93, "learning_rate": 4.4040404040404044e-06, "loss": 0.0592, "step": 11300 }, { "epoch": 5.93, "eval_loss": 0.0904541015625, "eval_runtime": 161.3572, "eval_samples_per_second": 20.012, "eval_steps_per_second": 0.081, "eval_wer": 10.179009253644775, "step": 11300 }, { "epoch": 5.93, "learning_rate": 4.3989898989899e-06, "loss": 0.0605, "step": 11310 }, { "epoch": 5.94, "learning_rate": 4.393939393939394e-06, "loss": 0.0591, "step": 11320 }, { "epoch": 5.94, "learning_rate": 4.388888888888889e-06, "loss": 0.0581, "step": 11330 }, { "epoch": 5.95, "learning_rate": 4.383838383838384e-06, "loss": 0.0585, "step": 11340 }, { "epoch": 5.95, "learning_rate": 4.378787878787879e-06, "loss": 0.0573, "step": 11350 }, { "epoch": 5.96, "learning_rate": 4.373737373737374e-06, "loss": 0.0594, "step": 11360 }, { "epoch": 5.97, "learning_rate": 4.368686868686869e-06, "loss": 0.0579, "step": 11370 }, { "epoch": 5.97, "learning_rate": 4.363636363636364e-06, "loss": 0.0595, "step": 11380 }, { "epoch": 5.98, "learning_rate": 4.358585858585859e-06, "loss": 0.0608, "step": 11390 }, { "epoch": 5.98, "learning_rate": 4.353535353535353e-06, "loss": 0.0614, "step": 11400 }, { "epoch": 5.98, "eval_loss": 0.090576171875, "eval_runtime": 159.4162, "eval_samples_per_second": 20.255, "eval_steps_per_second": 0.082, "eval_wer": 10.208009280008437, "step": 11400 }, { "epoch": 5.99, "learning_rate": 4.348484848484849e-06, "loss": 0.0588, "step": 11410 }, { "epoch": 5.99, "learning_rate": 4.343434343434344e-06, "loss": 0.0598, "step": 11420 }, { "epoch": 6.0, "learning_rate": 4.338383838383839e-06, "loss": 0.0591, "step": 11430 }, { "epoch": 6.0, "learning_rate": 4.333333333333334e-06, "loss": 0.0568, "step": 11440 }, { "epoch": 6.01, "learning_rate": 4.328282828282829e-06, "loss": 0.0593, "step": 11450 }, { "epoch": 6.01, "learning_rate": 4.323232323232323e-06, "loss": 0.0567, "step": 11460 }, { "epoch": 6.02, "learning_rate": 4.3181818181818185e-06, "loss": 0.0596, "step": 11470 }, { "epoch": 6.02, "learning_rate": 4.313131313131314e-06, "loss": 0.0547, "step": 11480 }, { "epoch": 6.03, "learning_rate": 4.308080808080809e-06, "loss": 0.057, "step": 11490 }, { "epoch": 6.03, "learning_rate": 4.303030303030303e-06, "loss": 0.055, "step": 11500 }, { "epoch": 6.03, "eval_loss": 0.0919189453125, "eval_runtime": 185.8281, "eval_samples_per_second": 17.376, "eval_steps_per_second": 0.07, "eval_wer": 10.215918378107617, "step": 11500 }, { "epoch": 6.04, "learning_rate": 4.297979797979798e-06, "loss": 0.0552, "step": 11510 }, { "epoch": 6.04, "learning_rate": 4.292929292929293e-06, "loss": 0.0576, "step": 11520 }, { "epoch": 6.05, "learning_rate": 4.287878787878788e-06, "loss": 0.0558, "step": 11530 }, { "epoch": 6.05, "learning_rate": 4.282828282828283e-06, "loss": 0.0561, "step": 11540 }, { "epoch": 6.06, "learning_rate": 4.277777777777778e-06, "loss": 0.0563, "step": 11550 }, { "epoch": 6.07, "learning_rate": 4.272727272727273e-06, "loss": 0.0572, "step": 11560 }, { "epoch": 6.07, "learning_rate": 4.267676767676767e-06, "loss": 0.0556, "step": 11570 }, { "epoch": 6.08, "learning_rate": 4.262626262626263e-06, "loss": 0.0572, "step": 11580 }, { "epoch": 6.08, "learning_rate": 4.2575757575757585e-06, "loss": 0.0601, "step": 11590 }, { "epoch": 6.09, "learning_rate": 4.252525252525253e-06, "loss": 0.058, "step": 11600 }, { "epoch": 6.09, "eval_loss": 0.09112548828125, "eval_runtime": 190.4867, "eval_samples_per_second": 16.951, "eval_steps_per_second": 0.068, "eval_wer": 10.324009385463077, "step": 11600 }, { "epoch": 6.09, "learning_rate": 4.247474747474748e-06, "loss": 0.0585, "step": 11610 }, { "epoch": 6.1, "learning_rate": 4.242424242424243e-06, "loss": 0.0578, "step": 11620 }, { "epoch": 6.1, "learning_rate": 4.237373737373737e-06, "loss": 0.0573, "step": 11630 }, { "epoch": 6.11, "learning_rate": 4.2323232323232325e-06, "loss": 0.0576, "step": 11640 }, { "epoch": 6.11, "learning_rate": 4.227272727272728e-06, "loss": 0.057, "step": 11650 }, { "epoch": 6.12, "learning_rate": 4.222222222222223e-06, "loss": 0.0572, "step": 11660 }, { "epoch": 6.12, "learning_rate": 4.217171717171717e-06, "loss": 0.0572, "step": 11670 }, { "epoch": 6.13, "learning_rate": 4.212121212121212e-06, "loss": 0.057, "step": 11680 }, { "epoch": 6.13, "learning_rate": 4.207070707070707e-06, "loss": 0.0576, "step": 11690 }, { "epoch": 6.14, "learning_rate": 4.2020202020202026e-06, "loss": 0.0582, "step": 11700 }, { "epoch": 6.14, "eval_loss": 0.09130859375, "eval_runtime": 208.0443, "eval_samples_per_second": 15.521, "eval_steps_per_second": 0.062, "eval_wer": 10.197463815876198, "step": 11700 }, { "epoch": 6.14, "learning_rate": 4.196969696969697e-06, "loss": 0.0585, "step": 11710 }, { "epoch": 6.15, "learning_rate": 4.191919191919192e-06, "loss": 0.0573, "step": 11720 }, { "epoch": 6.15, "learning_rate": 4.186868686868687e-06, "loss": 0.0563, "step": 11730 }, { "epoch": 6.16, "learning_rate": 4.181818181818182e-06, "loss": 0.0574, "step": 11740 }, { "epoch": 6.16, "learning_rate": 4.1767676767676774e-06, "loss": 0.0573, "step": 11750 }, { "epoch": 6.17, "learning_rate": 4.1717171717171726e-06, "loss": 0.0566, "step": 11760 }, { "epoch": 6.18, "learning_rate": 4.166666666666667e-06, "loss": 0.0575, "step": 11770 }, { "epoch": 6.18, "learning_rate": 4.161616161616162e-06, "loss": 0.0546, "step": 11780 }, { "epoch": 6.19, "learning_rate": 4.156565656565657e-06, "loss": 0.0573, "step": 11790 }, { "epoch": 6.19, "learning_rate": 4.151515151515152e-06, "loss": 0.0578, "step": 11800 }, { "epoch": 6.19, "eval_loss": 0.091064453125, "eval_runtime": 229.8873, "eval_samples_per_second": 14.046, "eval_steps_per_second": 0.057, "eval_wer": 10.210645646041497, "step": 11800 }, { "epoch": 6.2, "learning_rate": 4.146464646464647e-06, "loss": 0.0561, "step": 11810 }, { "epoch": 6.2, "learning_rate": 4.141414141414142e-06, "loss": 0.0575, "step": 11820 }, { "epoch": 6.21, "learning_rate": 4.136363636363637e-06, "loss": 0.0567, "step": 11830 }, { "epoch": 6.21, "learning_rate": 4.131313131313132e-06, "loss": 0.0584, "step": 11840 }, { "epoch": 6.22, "learning_rate": 4.126262626262626e-06, "loss": 0.0561, "step": 11850 }, { "epoch": 6.22, "learning_rate": 4.1212121212121215e-06, "loss": 0.0578, "step": 11860 }, { "epoch": 6.23, "learning_rate": 4.116161616161617e-06, "loss": 0.058, "step": 11870 }, { "epoch": 6.23, "learning_rate": 4.111111111111111e-06, "loss": 0.0567, "step": 11880 }, { "epoch": 6.24, "learning_rate": 4.106060606060606e-06, "loss": 0.0545, "step": 11890 }, { "epoch": 6.24, "learning_rate": 4.101010101010101e-06, "loss": 0.0552, "step": 11900 }, { "epoch": 6.24, "eval_loss": 0.0906982421875, "eval_runtime": 234.7301, "eval_samples_per_second": 13.756, "eval_steps_per_second": 0.055, "eval_wer": 10.295009359099417, "step": 11900 }, { "epoch": 6.25, "learning_rate": 4.095959595959596e-06, "loss": 0.058, "step": 11910 }, { "epoch": 6.25, "learning_rate": 4.0909090909090915e-06, "loss": 0.0561, "step": 11920 }, { "epoch": 6.26, "learning_rate": 4.085858585858587e-06, "loss": 0.0568, "step": 11930 }, { "epoch": 6.26, "learning_rate": 4.080808080808081e-06, "loss": 0.0558, "step": 11940 }, { "epoch": 6.27, "learning_rate": 4.075757575757576e-06, "loss": 0.0563, "step": 11950 }, { "epoch": 6.27, "learning_rate": 4.070707070707071e-06, "loss": 0.0578, "step": 11960 }, { "epoch": 6.28, "learning_rate": 4.065656565656566e-06, "loss": 0.0586, "step": 11970 }, { "epoch": 6.29, "learning_rate": 4.060606060606061e-06, "loss": 0.0559, "step": 11980 }, { "epoch": 6.29, "learning_rate": 4.055555555555556e-06, "loss": 0.0584, "step": 11990 }, { "epoch": 6.3, "learning_rate": 4.050505050505051e-06, "loss": 0.0568, "step": 12000 }, { "epoch": 6.3, "eval_loss": 0.09124755859375, "eval_runtime": 249.8203, "eval_samples_per_second": 12.925, "eval_steps_per_second": 0.052, "eval_wer": 10.186918351743957, "step": 12000 }, { "epoch": 6.3, "learning_rate": 4.045454545454546e-06, "loss": 0.057, "step": 12010 }, { "epoch": 6.31, "learning_rate": 4.040909090909091e-06, "loss": 0.0585, "step": 12020 }, { "epoch": 6.31, "learning_rate": 4.036363636363637e-06, "loss": 0.0578, "step": 12030 }, { "epoch": 6.32, "learning_rate": 4.031313131313131e-06, "loss": 0.0561, "step": 12040 }, { "epoch": 6.32, "learning_rate": 4.0262626262626264e-06, "loss": 0.0557, "step": 12050 }, { "epoch": 6.33, "learning_rate": 4.0212121212121216e-06, "loss": 0.0583, "step": 12060 }, { "epoch": 6.33, "learning_rate": 4.016161616161616e-06, "loss": 0.058, "step": 12070 }, { "epoch": 6.34, "learning_rate": 4.011111111111111e-06, "loss": 0.0564, "step": 12080 }, { "epoch": 6.34, "learning_rate": 4.006060606060607e-06, "loss": 0.0561, "step": 12090 }, { "epoch": 6.35, "learning_rate": 4.001010101010101e-06, "loss": 0.0583, "step": 12100 }, { "epoch": 6.35, "eval_loss": 0.09112548828125, "eval_runtime": 161.9841, "eval_samples_per_second": 19.934, "eval_steps_per_second": 0.08, "eval_wer": 10.302918457198597, "step": 12100 }, { "epoch": 6.35, "learning_rate": 3.9959595959595964e-06, "loss": 0.0583, "step": 12110 }, { "epoch": 6.36, "learning_rate": 3.990909090909092e-06, "loss": 0.0572, "step": 12120 }, { "epoch": 6.36, "learning_rate": 3.985858585858587e-06, "loss": 0.0583, "step": 12130 }, { "epoch": 6.37, "learning_rate": 3.980808080808081e-06, "loss": 0.0555, "step": 12140 }, { "epoch": 6.37, "learning_rate": 3.975757575757576e-06, "loss": 0.058, "step": 12150 }, { "epoch": 6.38, "learning_rate": 3.970707070707071e-06, "loss": 0.0577, "step": 12160 }, { "epoch": 6.39, "learning_rate": 3.965656565656566e-06, "loss": 0.0576, "step": 12170 }, { "epoch": 6.39, "learning_rate": 3.960606060606061e-06, "loss": 0.0584, "step": 12180 }, { "epoch": 6.4, "learning_rate": 3.955555555555556e-06, "loss": 0.0555, "step": 12190 }, { "epoch": 6.4, "learning_rate": 3.950505050505051e-06, "loss": 0.0562, "step": 12200 }, { "epoch": 6.4, "eval_loss": 0.0914306640625, "eval_runtime": 179.6459, "eval_samples_per_second": 17.974, "eval_steps_per_second": 0.072, "eval_wer": 10.310827555297777, "step": 12200 }, { "epoch": 6.41, "learning_rate": 3.945454545454545e-06, "loss": 0.0566, "step": 12210 }, { "epoch": 6.41, "learning_rate": 3.9404040404040405e-06, "loss": 0.0566, "step": 12220 }, { "epoch": 6.42, "learning_rate": 3.935353535353536e-06, "loss": 0.0573, "step": 12230 }, { "epoch": 6.42, "learning_rate": 3.930303030303031e-06, "loss": 0.0564, "step": 12240 }, { "epoch": 6.43, "learning_rate": 3.925252525252525e-06, "loss": 0.0564, "step": 12250 }, { "epoch": 6.43, "learning_rate": 3.920202020202021e-06, "loss": 0.0556, "step": 12260 }, { "epoch": 6.44, "learning_rate": 3.915151515151515e-06, "loss": 0.0575, "step": 12270 }, { "epoch": 6.44, "learning_rate": 3.9101010101010105e-06, "loss": 0.0567, "step": 12280 }, { "epoch": 6.45, "learning_rate": 3.905050505050506e-06, "loss": 0.058, "step": 12290 }, { "epoch": 6.45, "learning_rate": 3.900000000000001e-06, "loss": 0.0578, "step": 12300 }, { "epoch": 6.45, "eval_loss": 0.091064453125, "eval_runtime": 220.2146, "eval_samples_per_second": 14.663, "eval_steps_per_second": 0.059, "eval_wer": 10.221191110173738, "step": 12300 }, { "epoch": 6.46, "learning_rate": 3.894949494949495e-06, "loss": 0.0576, "step": 12310 }, { "epoch": 6.46, "learning_rate": 3.88989898989899e-06, "loss": 0.0587, "step": 12320 }, { "epoch": 6.47, "learning_rate": 3.884848484848485e-06, "loss": 0.0562, "step": 12330 }, { "epoch": 6.47, "learning_rate": 3.8797979797979805e-06, "loss": 0.0583, "step": 12340 }, { "epoch": 6.48, "learning_rate": 3.874747474747475e-06, "loss": 0.0555, "step": 12350 }, { "epoch": 6.48, "learning_rate": 3.86969696969697e-06, "loss": 0.0566, "step": 12360 }, { "epoch": 6.49, "learning_rate": 3.864646464646465e-06, "loss": 0.0571, "step": 12370 }, { "epoch": 6.5, "learning_rate": 3.859595959595959e-06, "loss": 0.0579, "step": 12380 }, { "epoch": 6.5, "learning_rate": 3.8545454545454545e-06, "loss": 0.0548, "step": 12390 }, { "epoch": 6.51, "learning_rate": 3.84949494949495e-06, "loss": 0.0567, "step": 12400 }, { "epoch": 6.51, "eval_loss": 0.08990478515625, "eval_runtime": 224.1781, "eval_samples_per_second": 14.404, "eval_steps_per_second": 0.058, "eval_wer": 10.186918351743957, "step": 12400 }, { "epoch": 6.51, "learning_rate": 3.844444444444445e-06, "loss": 0.0571, "step": 12410 }, { "epoch": 6.52, "learning_rate": 3.839393939393939e-06, "loss": 0.0544, "step": 12420 }, { "epoch": 6.52, "learning_rate": 3.834343434343435e-06, "loss": 0.0558, "step": 12430 }, { "epoch": 6.53, "learning_rate": 3.82929292929293e-06, "loss": 0.0554, "step": 12440 }, { "epoch": 6.53, "learning_rate": 3.8242424242424245e-06, "loss": 0.0593, "step": 12450 }, { "epoch": 6.54, "learning_rate": 3.81919191919192e-06, "loss": 0.0561, "step": 12460 }, { "epoch": 6.54, "learning_rate": 3.8141414141414144e-06, "loss": 0.0576, "step": 12470 }, { "epoch": 6.55, "learning_rate": 3.8090909090909095e-06, "loss": 0.0561, "step": 12480 }, { "epoch": 6.55, "learning_rate": 3.8040404040404043e-06, "loss": 0.0578, "step": 12490 }, { "epoch": 6.56, "learning_rate": 3.7989898989898994e-06, "loss": 0.0592, "step": 12500 }, { "epoch": 6.56, "eval_loss": 0.09033203125, "eval_runtime": 188.5938, "eval_samples_per_second": 17.121, "eval_steps_per_second": 0.069, "eval_wer": 10.231736574305977, "step": 12500 }, { "epoch": 6.56, "learning_rate": 3.793939393939394e-06, "loss": 0.0553, "step": 12510 }, { "epoch": 6.57, "learning_rate": 3.7888888888888893e-06, "loss": 0.0566, "step": 12520 }, { "epoch": 6.57, "learning_rate": 3.783838383838384e-06, "loss": 0.0582, "step": 12530 }, { "epoch": 6.58, "learning_rate": 3.778787878787879e-06, "loss": 0.0551, "step": 12540 }, { "epoch": 6.58, "learning_rate": 3.773737373737374e-06, "loss": 0.0571, "step": 12550 }, { "epoch": 6.59, "learning_rate": 3.768686868686869e-06, "loss": 0.0582, "step": 12560 }, { "epoch": 6.59, "learning_rate": 3.7636363636363637e-06, "loss": 0.0572, "step": 12570 }, { "epoch": 6.6, "learning_rate": 3.758585858585859e-06, "loss": 0.0551, "step": 12580 }, { "epoch": 6.61, "learning_rate": 3.7535353535353536e-06, "loss": 0.0578, "step": 12590 }, { "epoch": 6.61, "learning_rate": 3.748484848484849e-06, "loss": 0.0567, "step": 12600 }, { "epoch": 6.61, "eval_loss": 0.08966064453125, "eval_runtime": 175.2601, "eval_samples_per_second": 18.424, "eval_steps_per_second": 0.074, "eval_wer": 10.244918404471278, "step": 12600 }, { "epoch": 6.62, "learning_rate": 3.743434343434344e-06, "loss": 0.0576, "step": 12610 }, { "epoch": 6.62, "learning_rate": 3.738383838383839e-06, "loss": 0.057, "step": 12620 }, { "epoch": 6.63, "learning_rate": 3.7333333333333337e-06, "loss": 0.0568, "step": 12630 }, { "epoch": 6.63, "learning_rate": 3.728282828282829e-06, "loss": 0.0581, "step": 12640 }, { "epoch": 6.64, "learning_rate": 3.7232323232323236e-06, "loss": 0.0568, "step": 12650 }, { "epoch": 6.64, "learning_rate": 3.7181818181818187e-06, "loss": 0.0577, "step": 12660 }, { "epoch": 6.65, "learning_rate": 3.7131313131313135e-06, "loss": 0.0564, "step": 12670 }, { "epoch": 6.65, "learning_rate": 3.7080808080808086e-06, "loss": 0.0565, "step": 12680 }, { "epoch": 6.66, "learning_rate": 3.7030303030303033e-06, "loss": 0.058, "step": 12690 }, { "epoch": 6.66, "learning_rate": 3.697979797979798e-06, "loss": 0.0578, "step": 12700 }, { "epoch": 6.66, "eval_loss": 0.0889892578125, "eval_runtime": 195.7645, "eval_samples_per_second": 16.494, "eval_steps_per_second": 0.066, "eval_wer": 10.215918378107617, "step": 12700 }, { "epoch": 6.67, "learning_rate": 3.692929292929293e-06, "loss": 0.0565, "step": 12710 }, { "epoch": 6.67, "learning_rate": 3.687878787878788e-06, "loss": 0.0587, "step": 12720 }, { "epoch": 6.68, "learning_rate": 3.682828282828283e-06, "loss": 0.0566, "step": 12730 }, { "epoch": 6.68, "learning_rate": 3.6777777777777778e-06, "loss": 0.0569, "step": 12740 }, { "epoch": 6.69, "learning_rate": 3.672727272727273e-06, "loss": 0.0577, "step": 12750 }, { "epoch": 6.69, "learning_rate": 3.6676767676767676e-06, "loss": 0.0561, "step": 12760 }, { "epoch": 6.7, "learning_rate": 3.662626262626263e-06, "loss": 0.0582, "step": 12770 }, { "epoch": 6.71, "learning_rate": 3.657575757575758e-06, "loss": 0.059, "step": 12780 }, { "epoch": 6.71, "learning_rate": 3.652525252525253e-06, "loss": 0.058, "step": 12790 }, { "epoch": 6.72, "learning_rate": 3.6474747474747478e-06, "loss": 0.0581, "step": 12800 }, { "epoch": 6.72, "eval_loss": 0.08984375, "eval_runtime": 201.8836, "eval_samples_per_second": 15.994, "eval_steps_per_second": 0.064, "eval_wer": 10.268645698768818, "step": 12800 }, { "epoch": 6.72, "learning_rate": 3.642424242424243e-06, "loss": 0.0586, "step": 12810 }, { "epoch": 6.73, "learning_rate": 3.6373737373737376e-06, "loss": 0.0561, "step": 12820 }, { "epoch": 6.73, "learning_rate": 3.6323232323232328e-06, "loss": 0.0567, "step": 12830 }, { "epoch": 6.74, "learning_rate": 3.6272727272727275e-06, "loss": 0.0573, "step": 12840 }, { "epoch": 6.74, "learning_rate": 3.6222222222222226e-06, "loss": 0.0568, "step": 12850 }, { "epoch": 6.75, "learning_rate": 3.6171717171717174e-06, "loss": 0.0568, "step": 12860 }, { "epoch": 6.75, "learning_rate": 3.6121212121212125e-06, "loss": 0.0572, "step": 12870 }, { "epoch": 6.76, "learning_rate": 3.6070707070707072e-06, "loss": 0.0552, "step": 12880 }, { "epoch": 6.76, "learning_rate": 3.6020202020202024e-06, "loss": 0.0568, "step": 12890 }, { "epoch": 6.77, "learning_rate": 3.596969696969697e-06, "loss": 0.0592, "step": 12900 }, { "epoch": 6.77, "eval_loss": 0.0902099609375, "eval_runtime": 295.3954, "eval_samples_per_second": 10.931, "eval_steps_per_second": 0.044, "eval_wer": 10.197463815876198, "step": 12900 }, { "epoch": 6.77, "learning_rate": 3.5919191919191922e-06, "loss": 0.0585, "step": 12910 }, { "epoch": 6.78, "learning_rate": 3.586868686868687e-06, "loss": 0.0565, "step": 12920 }, { "epoch": 6.78, "learning_rate": 3.5818181818181817e-06, "loss": 0.0585, "step": 12930 }, { "epoch": 6.79, "learning_rate": 3.5767676767676772e-06, "loss": 0.0565, "step": 12940 }, { "epoch": 6.79, "learning_rate": 3.5717171717171724e-06, "loss": 0.0571, "step": 12950 }, { "epoch": 6.8, "learning_rate": 3.566666666666667e-06, "loss": 0.0575, "step": 12960 }, { "epoch": 6.8, "learning_rate": 3.5616161616161622e-06, "loss": 0.0557, "step": 12970 }, { "epoch": 6.81, "learning_rate": 3.556565656565657e-06, "loss": 0.0551, "step": 12980 }, { "epoch": 6.82, "learning_rate": 3.551515151515152e-06, "loss": 0.0571, "step": 12990 }, { "epoch": 6.82, "learning_rate": 3.546464646464647e-06, "loss": 0.0574, "step": 13000 }, { "epoch": 6.82, "eval_loss": 0.09033203125, "eval_runtime": 201.5528, "eval_samples_per_second": 16.021, "eval_steps_per_second": 0.064, "eval_wer": 10.144736495214996, "step": 13000 }, { "epoch": 6.83, "learning_rate": 3.5414141414141416e-06, "loss": 0.0578, "step": 13010 }, { "epoch": 6.83, "learning_rate": 3.5363636363636367e-06, "loss": 0.0557, "step": 13020 }, { "epoch": 6.84, "learning_rate": 3.5323232323232325e-06, "loss": 0.0566, "step": 13030 }, { "epoch": 6.84, "learning_rate": 3.5272727272727276e-06, "loss": 0.0574, "step": 13040 }, { "epoch": 6.85, "learning_rate": 3.5222222222222223e-06, "loss": 0.0563, "step": 13050 }, { "epoch": 6.85, "learning_rate": 3.5171717171717175e-06, "loss": 0.0551, "step": 13060 }, { "epoch": 6.86, "learning_rate": 3.512121212121212e-06, "loss": 0.0551, "step": 13070 }, { "epoch": 6.86, "learning_rate": 3.5070707070707073e-06, "loss": 0.0576, "step": 13080 }, { "epoch": 6.87, "learning_rate": 3.502020202020202e-06, "loss": 0.0549, "step": 13090 }, { "epoch": 6.87, "learning_rate": 3.496969696969697e-06, "loss": 0.059, "step": 13100 }, { "epoch": 6.87, "eval_loss": 0.08966064453125, "eval_runtime": 173.8793, "eval_samples_per_second": 18.57, "eval_steps_per_second": 0.075, "eval_wer": 10.221191110173738, "step": 13100 }, { "epoch": 6.88, "learning_rate": 3.491919191919192e-06, "loss": 0.0566, "step": 13110 }, { "epoch": 6.88, "learning_rate": 3.4868686868686875e-06, "loss": 0.0562, "step": 13120 }, { "epoch": 6.89, "learning_rate": 3.481818181818182e-06, "loss": 0.0568, "step": 13130 }, { "epoch": 6.89, "learning_rate": 3.4767676767676774e-06, "loss": 0.0573, "step": 13140 }, { "epoch": 6.9, "learning_rate": 3.471717171717172e-06, "loss": 0.0561, "step": 13150 }, { "epoch": 6.9, "learning_rate": 3.4666666666666672e-06, "loss": 0.057, "step": 13160 }, { "epoch": 6.91, "learning_rate": 3.461616161616162e-06, "loss": 0.0563, "step": 13170 }, { "epoch": 6.92, "learning_rate": 3.456565656565657e-06, "loss": 0.0564, "step": 13180 }, { "epoch": 6.92, "learning_rate": 3.451515151515152e-06, "loss": 0.0554, "step": 13190 }, { "epoch": 6.93, "learning_rate": 3.446464646464647e-06, "loss": 0.0585, "step": 13200 }, { "epoch": 6.93, "eval_loss": 0.0902099609375, "eval_runtime": 157.852, "eval_samples_per_second": 20.456, "eval_steps_per_second": 0.082, "eval_wer": 10.189554717777016, "step": 13200 }, { "epoch": 6.93, "learning_rate": 3.4414141414141417e-06, "loss": 0.0545, "step": 13210 }, { "epoch": 6.94, "learning_rate": 3.4363636363636364e-06, "loss": 0.0575, "step": 13220 }, { "epoch": 6.94, "learning_rate": 3.4313131313131315e-06, "loss": 0.0572, "step": 13230 }, { "epoch": 6.95, "learning_rate": 3.4262626262626262e-06, "loss": 0.0566, "step": 13240 }, { "epoch": 6.95, "learning_rate": 3.4212121212121214e-06, "loss": 0.0567, "step": 13250 }, { "epoch": 6.96, "learning_rate": 3.416161616161616e-06, "loss": 0.0582, "step": 13260 }, { "epoch": 6.96, "learning_rate": 3.4111111111111113e-06, "loss": 0.0543, "step": 13270 }, { "epoch": 6.97, "learning_rate": 3.406060606060606e-06, "loss": 0.0576, "step": 13280 }, { "epoch": 6.97, "learning_rate": 3.4010101010101015e-06, "loss": 0.058, "step": 13290 }, { "epoch": 6.98, "learning_rate": 3.3959595959595963e-06, "loss": 0.0569, "step": 13300 }, { "epoch": 6.98, "eval_loss": 0.0904541015625, "eval_runtime": 161.1123, "eval_samples_per_second": 20.042, "eval_steps_per_second": 0.081, "eval_wer": 10.218554744140675, "step": 13300 }, { "epoch": 6.98, "learning_rate": 3.3909090909090914e-06, "loss": 0.0565, "step": 13310 }, { "epoch": 6.99, "learning_rate": 3.385858585858586e-06, "loss": 0.0557, "step": 13320 }, { "epoch": 6.99, "learning_rate": 3.3808080808080813e-06, "loss": 0.057, "step": 13330 }, { "epoch": 7.0, "learning_rate": 3.375757575757576e-06, "loss": 0.0579, "step": 13340 }, { "epoch": 7.0, "learning_rate": 3.370707070707071e-06, "loss": 0.0542, "step": 13350 }, { "epoch": 7.01, "learning_rate": 3.365656565656566e-06, "loss": 0.0547, "step": 13360 }, { "epoch": 7.01, "learning_rate": 3.360606060606061e-06, "loss": 0.0539, "step": 13370 }, { "epoch": 7.02, "learning_rate": 3.3555555555555557e-06, "loss": 0.0543, "step": 13380 }, { "epoch": 7.03, "learning_rate": 3.350505050505051e-06, "loss": 0.0543, "step": 13390 }, { "epoch": 7.03, "learning_rate": 3.3454545454545456e-06, "loss": 0.0535, "step": 13400 }, { "epoch": 7.03, "eval_loss": 0.09130859375, "eval_runtime": 211.7481, "eval_samples_per_second": 15.249, "eval_steps_per_second": 0.061, "eval_wer": 10.134191031082755, "step": 13400 }, { "epoch": 6.53, "learning_rate": 3.6632996632996636e-06, "loss": 0.3334, "step": 13410 }, { "epoch": 6.53, "learning_rate": 3.658489658489659e-06, "loss": 0.193, "step": 13420 }, { "epoch": 6.54, "learning_rate": 3.6536796536796538e-06, "loss": 0.1177, "step": 13430 }, { "epoch": 6.54, "learning_rate": 3.648869648869649e-06, "loss": 0.0629, "step": 13440 }, { "epoch": 6.55, "learning_rate": 3.6440596440596444e-06, "loss": 0.057, "step": 13450 }, { "epoch": 6.55, "learning_rate": 3.6392496392496397e-06, "loss": 0.0545, "step": 13460 }, { "epoch": 6.55, "learning_rate": 3.634439634439635e-06, "loss": 0.0523, "step": 13470 }, { "epoch": 6.56, "learning_rate": 3.6296296296296302e-06, "loss": 0.0575, "step": 13480 }, { "epoch": 6.56, "learning_rate": 3.6248196248196247e-06, "loss": 0.0531, "step": 13490 }, { "epoch": 6.57, "learning_rate": 3.62000962000962e-06, "loss": 0.0526, "step": 13500 }, { "epoch": 6.57, "learning_rate": 3.6151996151996153e-06, "loss": 0.0542, "step": 13510 }, { "epoch": 6.58, "learning_rate": 3.6103896103896106e-06, "loss": 0.0515, "step": 13520 }, { "epoch": 6.58, "learning_rate": 3.605579605579606e-06, "loss": 0.052, "step": 13530 }, { "epoch": 6.59, "learning_rate": 3.6007696007696007e-06, "loss": 0.0531, "step": 13540 }, { "epoch": 6.59, "learning_rate": 3.595959595959596e-06, "loss": 0.05, "step": 13550 }, { "epoch": 6.6, "learning_rate": 3.5911495911495913e-06, "loss": 0.0523, "step": 13560 }, { "epoch": 6.6, "learning_rate": 3.5863395863395866e-06, "loss": 0.0521, "step": 13570 }, { "epoch": 6.61, "learning_rate": 3.581529581529582e-06, "loss": 0.0524, "step": 13580 }, { "epoch": 6.61, "learning_rate": 3.5767195767195772e-06, "loss": 0.0509, "step": 13590 }, { "epoch": 6.62, "learning_rate": 3.571909571909572e-06, "loss": 0.0519, "step": 13600 }, { "epoch": 6.62, "eval_loss": 0.0928955078125, "eval_runtime": 212.5799, "eval_samples_per_second": 15.19, "eval_steps_per_second": 0.061, "eval_wer": 10.40310036645488, "step": 13600 }, { "epoch": 6.62, "learning_rate": 3.5670995670995674e-06, "loss": 0.0512, "step": 13610 }, { "epoch": 6.63, "learning_rate": 3.5622895622895627e-06, "loss": 0.0517, "step": 13620 }, { "epoch": 6.63, "learning_rate": 3.557479557479558e-06, "loss": 0.0497, "step": 13630 }, { "epoch": 6.64, "learning_rate": 3.5526695526695533e-06, "loss": 0.0515, "step": 13640 }, { "epoch": 6.64, "learning_rate": 3.5478595478595477e-06, "loss": 0.0531, "step": 13650 }, { "epoch": 6.65, "learning_rate": 3.543049543049543e-06, "loss": 0.0507, "step": 13660 }, { "epoch": 6.65, "learning_rate": 3.5382395382395383e-06, "loss": 0.0512, "step": 13670 }, { "epoch": 6.66, "learning_rate": 3.5334295334295336e-06, "loss": 0.0524, "step": 13680 }, { "epoch": 6.66, "learning_rate": 3.528619528619529e-06, "loss": 0.0519, "step": 13690 }, { "epoch": 6.67, "learning_rate": 3.523809523809524e-06, "loss": 0.0512, "step": 13700 }, { "epoch": 6.67, "learning_rate": 3.518999518999519e-06, "loss": 0.0516, "step": 13710 }, { "epoch": 6.68, "learning_rate": 3.5141895141895143e-06, "loss": 0.0497, "step": 13720 }, { "epoch": 6.68, "learning_rate": 3.5093795093795096e-06, "loss": 0.0516, "step": 13730 }, { "epoch": 6.69, "learning_rate": 3.504569504569505e-06, "loss": 0.0515, "step": 13740 }, { "epoch": 6.69, "learning_rate": 3.4997594997595002e-06, "loss": 0.0534, "step": 13750 }, { "epoch": 6.7, "learning_rate": 3.494949494949495e-06, "loss": 0.0516, "step": 13760 }, { "epoch": 6.7, "learning_rate": 3.4901394901394904e-06, "loss": 0.0515, "step": 13770 }, { "epoch": 6.71, "learning_rate": 3.4853294853294857e-06, "loss": 0.052, "step": 13780 }, { "epoch": 6.71, "learning_rate": 3.480519480519481e-06, "loss": 0.0502, "step": 13790 }, { "epoch": 6.72, "learning_rate": 3.4757094757094763e-06, "loss": 0.0534, "step": 13800 }, { "epoch": 6.72, "eval_loss": 0.09210205078125, "eval_runtime": 212.841, "eval_samples_per_second": 15.171, "eval_steps_per_second": 0.061, "eval_wer": 10.395191268355699, "step": 13800 }, { "epoch": 6.72, "learning_rate": 3.4708994708994716e-06, "loss": 0.0508, "step": 13810 }, { "epoch": 6.73, "learning_rate": 3.466089466089466e-06, "loss": 0.0511, "step": 13820 }, { "epoch": 6.73, "learning_rate": 3.4612794612794613e-06, "loss": 0.0508, "step": 13830 }, { "epoch": 6.73, "learning_rate": 3.4564694564694566e-06, "loss": 0.0509, "step": 13840 }, { "epoch": 6.74, "learning_rate": 3.451659451659452e-06, "loss": 0.0516, "step": 13850 }, { "epoch": 6.74, "learning_rate": 3.446849446849447e-06, "loss": 0.0515, "step": 13860 }, { "epoch": 6.75, "learning_rate": 3.442039442039442e-06, "loss": 0.0514, "step": 13870 }, { "epoch": 6.75, "learning_rate": 3.4372294372294373e-06, "loss": 0.0511, "step": 13880 }, { "epoch": 6.76, "learning_rate": 3.4324194324194326e-06, "loss": 0.0497, "step": 13890 }, { "epoch": 6.76, "learning_rate": 3.427609427609428e-06, "loss": 0.0498, "step": 13900 }, { "epoch": 6.77, "learning_rate": 3.4227994227994232e-06, "loss": 0.0529, "step": 13910 }, { "epoch": 6.77, "learning_rate": 3.4179894179894185e-06, "loss": 0.0507, "step": 13920 }, { "epoch": 6.78, "learning_rate": 3.4131794131794134e-06, "loss": 0.0505, "step": 13930 }, { "epoch": 6.78, "learning_rate": 3.4083694083694087e-06, "loss": 0.0517, "step": 13940 }, { "epoch": 6.79, "learning_rate": 3.403559403559404e-06, "loss": 0.0514, "step": 13950 }, { "epoch": 6.79, "learning_rate": 3.3987493987493993e-06, "loss": 0.0506, "step": 13960 }, { "epoch": 6.8, "learning_rate": 3.3939393939393946e-06, "loss": 0.0494, "step": 13970 }, { "epoch": 6.8, "learning_rate": 3.389129389129389e-06, "loss": 0.0528, "step": 13980 }, { "epoch": 6.81, "learning_rate": 3.3843193843193843e-06, "loss": 0.0505, "step": 13990 }, { "epoch": 6.81, "learning_rate": 3.3795093795093796e-06, "loss": 0.0508, "step": 14000 }, { "epoch": 6.81, "eval_loss": 0.0933837890625, "eval_runtime": 212.0566, "eval_samples_per_second": 15.227, "eval_steps_per_second": 0.061, "eval_wer": 10.360918509925918, "step": 14000 }, { "epoch": 6.82, "learning_rate": 3.374699374699375e-06, "loss": 0.0491, "step": 14010 }, { "epoch": 6.82, "learning_rate": 3.36988936988937e-06, "loss": 0.0516, "step": 14020 }, { "epoch": 6.83, "learning_rate": 3.3650793650793655e-06, "loss": 0.0505, "step": 14030 }, { "epoch": 6.83, "learning_rate": 3.3602693602693604e-06, "loss": 0.0496, "step": 14040 }, { "epoch": 6.84, "learning_rate": 3.3554593554593556e-06, "loss": 0.052, "step": 14050 }, { "epoch": 6.84, "learning_rate": 3.350649350649351e-06, "loss": 0.0508, "step": 14060 }, { "epoch": 6.85, "learning_rate": 3.3458393458393462e-06, "loss": 0.052, "step": 14070 }, { "epoch": 6.85, "learning_rate": 3.3410293410293415e-06, "loss": 0.0513, "step": 14080 }, { "epoch": 6.86, "learning_rate": 3.3362193362193364e-06, "loss": 0.0497, "step": 14090 }, { "epoch": 6.86, "learning_rate": 3.3314093314093317e-06, "loss": 0.0514, "step": 14100 }, { "epoch": 6.87, "learning_rate": 3.326599326599327e-06, "loss": 0.0499, "step": 14110 }, { "epoch": 6.87, "learning_rate": 3.3217893217893223e-06, "loss": 0.0509, "step": 14120 }, { "epoch": 6.88, "learning_rate": 3.3169793169793176e-06, "loss": 0.0503, "step": 14130 }, { "epoch": 6.88, "learning_rate": 3.312169312169313e-06, "loss": 0.0509, "step": 14140 }, { "epoch": 6.89, "learning_rate": 3.3073593073593073e-06, "loss": 0.0509, "step": 14150 }, { "epoch": 6.89, "learning_rate": 3.3025493025493026e-06, "loss": 0.049, "step": 14160 }, { "epoch": 6.9, "learning_rate": 3.297739297739298e-06, "loss": 0.0497, "step": 14170 }, { "epoch": 6.9, "learning_rate": 3.292929292929293e-06, "loss": 0.05, "step": 14180 }, { "epoch": 6.91, "learning_rate": 3.2881192881192885e-06, "loss": 0.0521, "step": 14190 }, { "epoch": 6.91, "learning_rate": 3.2833092833092834e-06, "loss": 0.0514, "step": 14200 }, { "epoch": 6.91, "eval_loss": 0.09368896484375, "eval_runtime": 212.3474, "eval_samples_per_second": 15.206, "eval_steps_per_second": 0.061, "eval_wer": 10.440009490917719, "step": 14200 }, { "epoch": 6.91, "learning_rate": 3.2784992784992787e-06, "loss": 0.0503, "step": 14210 }, { "epoch": 6.92, "learning_rate": 3.273689273689274e-06, "loss": 0.0501, "step": 14220 }, { "epoch": 6.92, "learning_rate": 3.2688792688792692e-06, "loss": 0.0507, "step": 14230 }, { "epoch": 6.93, "learning_rate": 3.2640692640692645e-06, "loss": 0.0518, "step": 14240 }, { "epoch": 6.93, "learning_rate": 3.25925925925926e-06, "loss": 0.0506, "step": 14250 }, { "epoch": 6.94, "learning_rate": 3.2544492544492547e-06, "loss": 0.0521, "step": 14260 }, { "epoch": 6.94, "learning_rate": 3.24963924963925e-06, "loss": 0.0505, "step": 14270 }, { "epoch": 6.95, "learning_rate": 3.2448292448292453e-06, "loss": 0.0486, "step": 14280 }, { "epoch": 6.95, "learning_rate": 3.2400192400192406e-06, "loss": 0.0503, "step": 14290 }, { "epoch": 6.96, "learning_rate": 3.235209235209236e-06, "loss": 0.0512, "step": 14300 }, { "epoch": 6.96, "learning_rate": 3.2303992303992303e-06, "loss": 0.0523, "step": 14310 }, { "epoch": 6.97, "learning_rate": 3.2255892255892256e-06, "loss": 0.0502, "step": 14320 }, { "epoch": 6.97, "learning_rate": 3.220779220779221e-06, "loss": 0.05, "step": 14330 }, { "epoch": 6.98, "learning_rate": 3.215969215969216e-06, "loss": 0.0486, "step": 14340 }, { "epoch": 6.98, "learning_rate": 3.2111592111592115e-06, "loss": 0.0492, "step": 14350 }, { "epoch": 6.99, "learning_rate": 3.206349206349207e-06, "loss": 0.0481, "step": 14360 }, { "epoch": 6.99, "learning_rate": 3.2015392015392017e-06, "loss": 0.05, "step": 14370 }, { "epoch": 7.0, "learning_rate": 3.196729196729197e-06, "loss": 0.0504, "step": 14380 }, { "epoch": 7.0, "learning_rate": 3.1919191919191923e-06, "loss": 0.0497, "step": 14390 }, { "epoch": 7.01, "learning_rate": 3.1871091871091875e-06, "loss": 0.0488, "step": 14400 }, { "epoch": 7.01, "eval_loss": 0.0950927734375, "eval_runtime": 209.1646, "eval_samples_per_second": 15.438, "eval_steps_per_second": 0.062, "eval_wer": 10.4874640795128, "step": 14400 }, { "epoch": 7.01, "learning_rate": 3.182299182299183e-06, "loss": 0.0471, "step": 14410 }, { "epoch": 7.02, "learning_rate": 3.1774891774891777e-06, "loss": 0.049, "step": 14420 }, { "epoch": 7.02, "learning_rate": 3.172679172679173e-06, "loss": 0.0476, "step": 14430 }, { "epoch": 7.03, "learning_rate": 3.1678691678691683e-06, "loss": 0.0463, "step": 14440 }, { "epoch": 7.03, "learning_rate": 3.1630591630591636e-06, "loss": 0.0495, "step": 14450 }, { "epoch": 7.04, "learning_rate": 3.1582491582491585e-06, "loss": 0.0488, "step": 14460 }, { "epoch": 7.04, "learning_rate": 3.1534391534391538e-06, "loss": 0.0485, "step": 14470 }, { "epoch": 7.05, "learning_rate": 3.1486291486291486e-06, "loss": 0.0491, "step": 14480 }, { "epoch": 7.05, "learning_rate": 3.143819143819144e-06, "loss": 0.0476, "step": 14490 }, { "epoch": 7.06, "learning_rate": 3.1390091390091392e-06, "loss": 0.0493, "step": 14500 }, { "epoch": 7.06, "learning_rate": 3.1341991341991345e-06, "loss": 0.0471, "step": 14510 }, { "epoch": 7.07, "learning_rate": 3.12938912938913e-06, "loss": 0.0475, "step": 14520 }, { "epoch": 7.07, "learning_rate": 3.1245791245791247e-06, "loss": 0.0475, "step": 14530 }, { "epoch": 7.08, "learning_rate": 3.11976911976912e-06, "loss": 0.0472, "step": 14540 }, { "epoch": 7.08, "learning_rate": 3.1149591149591153e-06, "loss": 0.0475, "step": 14550 }, { "epoch": 7.09, "learning_rate": 3.1101491101491106e-06, "loss": 0.0471, "step": 14560 }, { "epoch": 7.09, "learning_rate": 3.105339105339106e-06, "loss": 0.0478, "step": 14570 }, { "epoch": 7.09, "learning_rate": 3.100529100529101e-06, "loss": 0.0493, "step": 14580 }, { "epoch": 7.1, "learning_rate": 3.0957190957190956e-06, "loss": 0.0461, "step": 14590 }, { "epoch": 7.1, "learning_rate": 3.090909090909091e-06, "loss": 0.047, "step": 14600 }, { "epoch": 7.1, "eval_loss": 0.0955810546875, "eval_runtime": 208.5869, "eval_samples_per_second": 15.48, "eval_steps_per_second": 0.062, "eval_wer": 10.5006459096781, "step": 14600 }, { "epoch": 7.11, "learning_rate": 3.086099086099086e-06, "loss": 0.0484, "step": 14610 }, { "epoch": 7.11, "learning_rate": 3.0812890812890815e-06, "loss": 0.0466, "step": 14620 }, { "epoch": 7.12, "learning_rate": 3.0764790764790768e-06, "loss": 0.047, "step": 14630 }, { "epoch": 7.12, "learning_rate": 3.0716690716690716e-06, "loss": 0.0463, "step": 14640 }, { "epoch": 7.13, "learning_rate": 3.066859066859067e-06, "loss": 0.0486, "step": 14650 }, { "epoch": 7.13, "learning_rate": 3.0620490620490622e-06, "loss": 0.0494, "step": 14660 }, { "epoch": 7.14, "learning_rate": 3.0572390572390575e-06, "loss": 0.0472, "step": 14670 }, { "epoch": 7.14, "learning_rate": 3.052429052429053e-06, "loss": 0.0471, "step": 14680 }, { "epoch": 7.15, "learning_rate": 3.047619047619048e-06, "loss": 0.0479, "step": 14690 }, { "epoch": 7.15, "learning_rate": 3.042809042809043e-06, "loss": 0.0482, "step": 14700 }, { "epoch": 7.16, "learning_rate": 3.0379990379990383e-06, "loss": 0.0478, "step": 14710 }, { "epoch": 7.16, "learning_rate": 3.0331890331890336e-06, "loss": 0.0489, "step": 14720 }, { "epoch": 7.17, "learning_rate": 3.028379028379029e-06, "loss": 0.0486, "step": 14730 }, { "epoch": 7.17, "learning_rate": 3.023569023569024e-06, "loss": 0.0472, "step": 14740 }, { "epoch": 7.18, "learning_rate": 3.0187590187590186e-06, "loss": 0.0462, "step": 14750 }, { "epoch": 7.18, "learning_rate": 3.013949013949014e-06, "loss": 0.0483, "step": 14760 }, { "epoch": 7.19, "learning_rate": 3.009139009139009e-06, "loss": 0.0464, "step": 14770 }, { "epoch": 7.19, "learning_rate": 3.0043290043290045e-06, "loss": 0.0484, "step": 14780 }, { "epoch": 7.2, "learning_rate": 2.9995189995189998e-06, "loss": 0.0479, "step": 14790 }, { "epoch": 7.2, "learning_rate": 2.9947089947089946e-06, "loss": 0.046, "step": 14800 }, { "epoch": 7.2, "eval_loss": 0.0948486328125, "eval_runtime": 203.4404, "eval_samples_per_second": 15.872, "eval_steps_per_second": 0.064, "eval_wer": 10.52173683794258, "step": 14800 }, { "epoch": 7.21, "learning_rate": 2.98989898989899e-06, "loss": 0.0464, "step": 14810 }, { "epoch": 7.21, "learning_rate": 2.9850889850889852e-06, "loss": 0.0496, "step": 14820 }, { "epoch": 7.22, "learning_rate": 2.9802789802789805e-06, "loss": 0.0467, "step": 14830 }, { "epoch": 7.22, "learning_rate": 2.975468975468976e-06, "loss": 0.0495, "step": 14840 }, { "epoch": 7.23, "learning_rate": 2.970658970658971e-06, "loss": 0.0475, "step": 14850 }, { "epoch": 7.23, "learning_rate": 2.965848965848966e-06, "loss": 0.0469, "step": 14860 }, { "epoch": 7.24, "learning_rate": 2.9610389610389613e-06, "loss": 0.0473, "step": 14870 }, { "epoch": 7.24, "learning_rate": 2.9562289562289566e-06, "loss": 0.0482, "step": 14880 }, { "epoch": 7.25, "learning_rate": 2.951418951418952e-06, "loss": 0.0463, "step": 14890 }, { "epoch": 7.25, "learning_rate": 2.946608946608947e-06, "loss": 0.0472, "step": 14900 }, { "epoch": 7.26, "learning_rate": 2.9417989417989416e-06, "loss": 0.0465, "step": 14910 }, { "epoch": 7.26, "learning_rate": 2.936988936988937e-06, "loss": 0.0479, "step": 14920 }, { "epoch": 7.27, "learning_rate": 2.932178932178932e-06, "loss": 0.0471, "step": 14930 }, { "epoch": 7.27, "learning_rate": 2.9273689273689275e-06, "loss": 0.0482, "step": 14940 }, { "epoch": 7.27, "learning_rate": 2.9225589225589228e-06, "loss": 0.049, "step": 14950 }, { "epoch": 7.28, "learning_rate": 2.917748917748918e-06, "loss": 0.046, "step": 14960 }, { "epoch": 7.28, "learning_rate": 2.912938912938913e-06, "loss": 0.0482, "step": 14970 }, { "epoch": 7.29, "learning_rate": 2.9081289081289082e-06, "loss": 0.0482, "step": 14980 }, { "epoch": 7.29, "learning_rate": 2.9033189033189035e-06, "loss": 0.0473, "step": 14990 }, { "epoch": 7.3, "learning_rate": 2.898508898508899e-06, "loss": 0.0484, "step": 15000 }, { "epoch": 7.3, "eval_loss": 0.0960693359375, "eval_runtime": 211.5972, "eval_samples_per_second": 15.26, "eval_steps_per_second": 0.061, "eval_wer": 10.4874640795128, "step": 15000 }, { "epoch": 7.3, "learning_rate": 2.893698893698894e-06, "loss": 0.0467, "step": 15010 }, { "epoch": 7.31, "learning_rate": 2.888888888888889e-06, "loss": 0.0471, "step": 15020 }, { "epoch": 7.31, "learning_rate": 2.8840788840788843e-06, "loss": 0.0482, "step": 15030 }, { "epoch": 7.32, "learning_rate": 2.8792688792688796e-06, "loss": 0.0468, "step": 15040 }, { "epoch": 7.32, "learning_rate": 2.874458874458875e-06, "loss": 0.0472, "step": 15050 }, { "epoch": 7.33, "learning_rate": 2.86964886964887e-06, "loss": 0.0472, "step": 15060 }, { "epoch": 7.33, "learning_rate": 2.8648388648388655e-06, "loss": 0.0488, "step": 15070 }, { "epoch": 7.34, "learning_rate": 2.86002886002886e-06, "loss": 0.0485, "step": 15080 }, { "epoch": 7.34, "learning_rate": 2.855218855218855e-06, "loss": 0.0467, "step": 15090 }, { "epoch": 7.35, "learning_rate": 2.8504088504088505e-06, "loss": 0.0479, "step": 15100 }, { "epoch": 7.35, "learning_rate": 2.845598845598846e-06, "loss": 0.0482, "step": 15110 }, { "epoch": 7.36, "learning_rate": 2.840788840788841e-06, "loss": 0.049, "step": 15120 }, { "epoch": 7.36, "learning_rate": 2.835978835978836e-06, "loss": 0.047, "step": 15130 }, { "epoch": 7.37, "learning_rate": 2.8311688311688312e-06, "loss": 0.0483, "step": 15140 }, { "epoch": 7.37, "learning_rate": 2.8263588263588265e-06, "loss": 0.0466, "step": 15150 }, { "epoch": 7.38, "learning_rate": 2.821548821548822e-06, "loss": 0.049, "step": 15160 }, { "epoch": 7.38, "learning_rate": 2.816738816738817e-06, "loss": 0.0456, "step": 15170 }, { "epoch": 7.39, "learning_rate": 2.8119288119288124e-06, "loss": 0.0477, "step": 15180 }, { "epoch": 7.39, "learning_rate": 2.8071188071188073e-06, "loss": 0.0474, "step": 15190 }, { "epoch": 7.4, "learning_rate": 2.8023088023088026e-06, "loss": 0.0473, "step": 15200 }, { "epoch": 7.4, "eval_loss": 0.09576416015625, "eval_runtime": 212.7288, "eval_samples_per_second": 15.179, "eval_steps_per_second": 0.061, "eval_wer": 10.48219134744668, "step": 15200 }, { "epoch": 7.4, "learning_rate": 2.797498797498798e-06, "loss": 0.0469, "step": 15210 }, { "epoch": 7.41, "learning_rate": 2.792688792688793e-06, "loss": 0.0492, "step": 15220 }, { "epoch": 7.41, "learning_rate": 2.7878787878787885e-06, "loss": 0.0459, "step": 15230 }, { "epoch": 7.42, "learning_rate": 2.783068783068783e-06, "loss": 0.0473, "step": 15240 }, { "epoch": 7.42, "learning_rate": 2.7782587782587782e-06, "loss": 0.0482, "step": 15250 }, { "epoch": 7.43, "learning_rate": 2.7734487734487735e-06, "loss": 0.0477, "step": 15260 }, { "epoch": 7.43, "learning_rate": 2.768638768638769e-06, "loss": 0.0482, "step": 15270 }, { "epoch": 7.44, "learning_rate": 2.763828763828764e-06, "loss": 0.0479, "step": 15280 }, { "epoch": 7.44, "learning_rate": 2.7590187590187594e-06, "loss": 0.0493, "step": 15290 }, { "epoch": 7.45, "learning_rate": 2.7542087542087543e-06, "loss": 0.0482, "step": 15300 }, { "epoch": 7.45, "learning_rate": 2.7493987493987496e-06, "loss": 0.0478, "step": 15310 }, { "epoch": 7.45, "learning_rate": 2.744588744588745e-06, "loss": 0.0448, "step": 15320 }, { "epoch": 7.46, "learning_rate": 2.73977873977874e-06, "loss": 0.0467, "step": 15330 }, { "epoch": 7.46, "learning_rate": 2.7349687349687354e-06, "loss": 0.0469, "step": 15340 }, { "epoch": 7.47, "learning_rate": 2.7301587301587303e-06, "loss": 0.047, "step": 15350 }, { "epoch": 7.47, "learning_rate": 2.7253487253487256e-06, "loss": 0.0472, "step": 15360 }, { "epoch": 7.48, "learning_rate": 2.720538720538721e-06, "loss": 0.0467, "step": 15370 }, { "epoch": 7.48, "learning_rate": 2.715728715728716e-06, "loss": 0.0494, "step": 15380 }, { "epoch": 7.49, "learning_rate": 2.7109187109187115e-06, "loss": 0.048, "step": 15390 }, { "epoch": 7.49, "learning_rate": 2.7061087061087068e-06, "loss": 0.05, "step": 15400 }, { "epoch": 7.49, "eval_loss": 0.09564208984375, "eval_runtime": 205.1854, "eval_samples_per_second": 15.737, "eval_steps_per_second": 0.063, "eval_wer": 10.45055495504996, "step": 15400 }, { "epoch": 7.5, "learning_rate": 2.7012987012987012e-06, "loss": 0.0473, "step": 15410 }, { "epoch": 7.5, "learning_rate": 2.6964886964886965e-06, "loss": 0.0481, "step": 15420 }, { "epoch": 7.51, "learning_rate": 2.691678691678692e-06, "loss": 0.0454, "step": 15430 }, { "epoch": 7.51, "learning_rate": 2.686868686868687e-06, "loss": 0.0474, "step": 15440 }, { "epoch": 7.52, "learning_rate": 2.6820586820586824e-06, "loss": 0.0473, "step": 15450 }, { "epoch": 7.52, "learning_rate": 2.6772486772486773e-06, "loss": 0.0469, "step": 15460 }, { "epoch": 7.53, "learning_rate": 2.6724386724386726e-06, "loss": 0.047, "step": 15470 }, { "epoch": 7.53, "learning_rate": 2.667628667628668e-06, "loss": 0.0458, "step": 15480 }, { "epoch": 7.54, "learning_rate": 2.662818662818663e-06, "loss": 0.0451, "step": 15490 }, { "epoch": 7.54, "learning_rate": 2.6580086580086584e-06, "loss": 0.048, "step": 15500 }, { "epoch": 7.55, "learning_rate": 2.6531986531986537e-06, "loss": 0.0489, "step": 15510 }, { "epoch": 7.55, "learning_rate": 2.6483886483886486e-06, "loss": 0.048, "step": 15520 }, { "epoch": 7.56, "learning_rate": 2.643578643578644e-06, "loss": 0.048, "step": 15530 }, { "epoch": 7.56, "learning_rate": 2.6387686387686388e-06, "loss": 0.0468, "step": 15540 }, { "epoch": 7.57, "learning_rate": 2.633958633958634e-06, "loss": 0.0474, "step": 15550 }, { "epoch": 7.57, "learning_rate": 2.6291486291486294e-06, "loss": 0.0467, "step": 15560 }, { "epoch": 7.58, "learning_rate": 2.6243386243386242e-06, "loss": 0.0462, "step": 15570 }, { "epoch": 7.58, "learning_rate": 2.6195286195286195e-06, "loss": 0.0471, "step": 15580 }, { "epoch": 7.59, "learning_rate": 2.614718614718615e-06, "loss": 0.0473, "step": 15590 }, { "epoch": 7.59, "learning_rate": 2.60990860990861e-06, "loss": 0.0452, "step": 15600 }, { "epoch": 7.59, "eval_loss": 0.09564208984375, "eval_runtime": 208.2007, "eval_samples_per_second": 15.509, "eval_steps_per_second": 0.062, "eval_wer": 10.395191268355699, "step": 15600 }, { "epoch": 7.6, "learning_rate": 2.6050986050986054e-06, "loss": 0.0475, "step": 15610 }, { "epoch": 7.6, "learning_rate": 2.6002886002886007e-06, "loss": 0.048, "step": 15620 }, { "epoch": 7.61, "learning_rate": 2.5954785954785956e-06, "loss": 0.0475, "step": 15630 }, { "epoch": 7.61, "learning_rate": 2.590668590668591e-06, "loss": 0.0472, "step": 15640 }, { "epoch": 7.62, "learning_rate": 2.585858585858586e-06, "loss": 0.046, "step": 15650 }, { "epoch": 7.62, "learning_rate": 2.5810485810485815e-06, "loss": 0.0477, "step": 15660 }, { "epoch": 7.63, "learning_rate": 2.5762385762385767e-06, "loss": 0.0471, "step": 15670 }, { "epoch": 7.63, "learning_rate": 2.571428571428571e-06, "loss": 0.0472, "step": 15680 }, { "epoch": 7.64, "learning_rate": 2.5666185666185665e-06, "loss": 0.0476, "step": 15690 }, { "epoch": 7.64, "learning_rate": 2.5618085618085618e-06, "loss": 0.0475, "step": 15700 }, { "epoch": 7.64, "learning_rate": 2.556998556998557e-06, "loss": 0.048, "step": 15710 }, { "epoch": 7.65, "learning_rate": 2.5521885521885524e-06, "loss": 0.0474, "step": 15720 }, { "epoch": 7.65, "learning_rate": 2.5473785473785477e-06, "loss": 0.0454, "step": 15730 }, { "epoch": 7.66, "learning_rate": 2.5425685425685425e-06, "loss": 0.0476, "step": 15740 }, { "epoch": 7.66, "learning_rate": 2.537758537758538e-06, "loss": 0.0503, "step": 15750 }, { "epoch": 7.67, "learning_rate": 2.532948532948533e-06, "loss": 0.0476, "step": 15760 }, { "epoch": 7.67, "learning_rate": 2.5281385281385284e-06, "loss": 0.0467, "step": 15770 }, { "epoch": 7.68, "learning_rate": 2.5233285233285237e-06, "loss": 0.0481, "step": 15780 }, { "epoch": 7.68, "learning_rate": 2.5185185185185186e-06, "loss": 0.0466, "step": 15790 }, { "epoch": 7.69, "learning_rate": 2.513708513708514e-06, "loss": 0.0463, "step": 15800 }, { "epoch": 7.69, "eval_loss": 0.09539794921875, "eval_runtime": 207.8235, "eval_samples_per_second": 15.537, "eval_steps_per_second": 0.063, "eval_wer": 10.329282117529198, "step": 15800 }, { "epoch": 7.69, "learning_rate": 2.508898508898509e-06, "loss": 0.048, "step": 15810 }, { "epoch": 7.7, "learning_rate": 2.5040885040885045e-06, "loss": 0.0474, "step": 15820 }, { "epoch": 7.7, "learning_rate": 2.4992784992784993e-06, "loss": 0.0482, "step": 15830 }, { "epoch": 7.71, "learning_rate": 2.4944684944684946e-06, "loss": 0.0491, "step": 15840 }, { "epoch": 7.71, "learning_rate": 2.48965848965849e-06, "loss": 0.0451, "step": 15850 }, { "epoch": 7.72, "learning_rate": 2.4848484848484848e-06, "loss": 0.0469, "step": 15860 }, { "epoch": 7.72, "learning_rate": 2.48003848003848e-06, "loss": 0.0466, "step": 15870 }, { "epoch": 7.73, "learning_rate": 2.4752284752284754e-06, "loss": 0.0472, "step": 15880 }, { "epoch": 7.73, "learning_rate": 2.4704184704184707e-06, "loss": 0.0467, "step": 15890 }, { "epoch": 7.74, "learning_rate": 2.465608465608466e-06, "loss": 0.0466, "step": 15900 }, { "epoch": 7.74, "learning_rate": 2.4607984607984613e-06, "loss": 0.0454, "step": 15910 }, { "epoch": 7.75, "learning_rate": 2.455988455988456e-06, "loss": 0.0464, "step": 15920 }, { "epoch": 7.75, "learning_rate": 2.4511784511784514e-06, "loss": 0.0476, "step": 15930 }, { "epoch": 7.76, "learning_rate": 2.4463684463684463e-06, "loss": 0.0474, "step": 15940 }, { "epoch": 7.76, "learning_rate": 2.4415584415584416e-06, "loss": 0.0477, "step": 15950 }, { "epoch": 7.77, "learning_rate": 2.436748436748437e-06, "loss": 0.0488, "step": 15960 }, { "epoch": 7.77, "learning_rate": 2.431938431938432e-06, "loss": 0.0486, "step": 15970 }, { "epoch": 7.78, "learning_rate": 2.4271284271284275e-06, "loss": 0.0471, "step": 15980 }, { "epoch": 7.78, "learning_rate": 2.4223184223184228e-06, "loss": 0.048, "step": 15990 }, { "epoch": 7.79, "learning_rate": 2.4175084175084176e-06, "loss": 0.048, "step": 16000 }, { "epoch": 7.79, "eval_loss": 0.09503173828125, "eval_runtime": 209.4297, "eval_samples_per_second": 15.418, "eval_steps_per_second": 0.062, "eval_wer": 10.395191268355699, "step": 16000 }, { "epoch": 7.79, "learning_rate": 2.412698412698413e-06, "loss": 0.0485, "step": 16010 }, { "epoch": 7.8, "learning_rate": 2.4078884078884082e-06, "loss": 0.0463, "step": 16020 }, { "epoch": 7.8, "learning_rate": 2.403078403078403e-06, "loss": 0.0472, "step": 16030 }, { "epoch": 7.81, "learning_rate": 2.3982683982683984e-06, "loss": 0.0458, "step": 16040 }, { "epoch": 7.81, "learning_rate": 2.3934583934583937e-06, "loss": 0.0472, "step": 16050 }, { "epoch": 7.82, "learning_rate": 2.388648388648389e-06, "loss": 0.0477, "step": 16060 }, { "epoch": 7.82, "learning_rate": 2.3838383838383843e-06, "loss": 0.0461, "step": 16070 }, { "epoch": 7.82, "learning_rate": 2.379028379028379e-06, "loss": 0.0463, "step": 16080 }, { "epoch": 7.83, "learning_rate": 2.3742183742183744e-06, "loss": 0.048, "step": 16090 }, { "epoch": 7.83, "learning_rate": 2.3694083694083697e-06, "loss": 0.0461, "step": 16100 }, { "epoch": 7.84, "learning_rate": 2.3645983645983646e-06, "loss": 0.0469, "step": 16110 }, { "epoch": 7.84, "learning_rate": 2.35978835978836e-06, "loss": 0.046, "step": 16120 }, { "epoch": 7.85, "learning_rate": 2.354978354978355e-06, "loss": 0.0471, "step": 16130 }, { "epoch": 7.85, "learning_rate": 2.3501683501683505e-06, "loss": 0.0469, "step": 16140 }, { "epoch": 7.86, "learning_rate": 2.3453583453583458e-06, "loss": 0.049, "step": 16150 }, { "epoch": 7.86, "learning_rate": 2.3405483405483406e-06, "loss": 0.0466, "step": 16160 }, { "epoch": 7.87, "learning_rate": 2.335738335738336e-06, "loss": 0.0469, "step": 16170 }, { "epoch": 7.87, "learning_rate": 2.3309283309283312e-06, "loss": 0.0475, "step": 16180 }, { "epoch": 7.88, "learning_rate": 2.326118326118326e-06, "loss": 0.0471, "step": 16190 }, { "epoch": 7.88, "learning_rate": 2.3213083213083214e-06, "loss": 0.0465, "step": 16200 }, { "epoch": 7.88, "eval_loss": 0.095458984375, "eval_runtime": 205.3318, "eval_samples_per_second": 15.726, "eval_steps_per_second": 0.063, "eval_wer": 10.424191294719359, "step": 16200 }, { "epoch": 7.89, "learning_rate": 2.3164983164983167e-06, "loss": 0.0466, "step": 16210 }, { "epoch": 7.89, "learning_rate": 2.311688311688312e-06, "loss": 0.0486, "step": 16220 }, { "epoch": 7.9, "learning_rate": 2.3068783068783073e-06, "loss": 0.0471, "step": 16230 }, { "epoch": 7.9, "learning_rate": 2.302068302068302e-06, "loss": 0.0471, "step": 16240 }, { "epoch": 7.91, "learning_rate": 2.2972582972582974e-06, "loss": 0.0464, "step": 16250 }, { "epoch": 7.91, "learning_rate": 2.2924482924482927e-06, "loss": 0.0469, "step": 16260 }, { "epoch": 7.92, "learning_rate": 2.2876382876382876e-06, "loss": 0.0471, "step": 16270 }, { "epoch": 7.92, "learning_rate": 2.282828282828283e-06, "loss": 0.0464, "step": 16280 }, { "epoch": 7.93, "learning_rate": 2.278018278018278e-06, "loss": 0.0452, "step": 16290 }, { "epoch": 7.93, "learning_rate": 2.2732082732082735e-06, "loss": 0.0471, "step": 16300 }, { "epoch": 7.94, "learning_rate": 2.2683982683982688e-06, "loss": 0.0448, "step": 16310 }, { "epoch": 7.94, "learning_rate": 2.263588263588264e-06, "loss": 0.0497, "step": 16320 }, { "epoch": 7.95, "learning_rate": 2.258778258778259e-06, "loss": 0.0472, "step": 16330 }, { "epoch": 7.95, "learning_rate": 2.2539682539682542e-06, "loss": 0.044, "step": 16340 }, { "epoch": 7.96, "learning_rate": 2.249158249158249e-06, "loss": 0.0497, "step": 16350 }, { "epoch": 7.96, "learning_rate": 2.2443482443482444e-06, "loss": 0.0463, "step": 16360 }, { "epoch": 7.97, "learning_rate": 2.2395382395382397e-06, "loss": 0.0469, "step": 16370 }, { "epoch": 7.97, "learning_rate": 2.234728234728235e-06, "loss": 0.0458, "step": 16380 }, { "epoch": 7.98, "learning_rate": 2.2299182299182303e-06, "loss": 0.046, "step": 16390 }, { "epoch": 7.98, "learning_rate": 2.2251082251082256e-06, "loss": 0.0478, "step": 16400 }, { "epoch": 7.98, "eval_loss": 0.09503173828125, "eval_runtime": 203.438, "eval_samples_per_second": 15.872, "eval_steps_per_second": 0.064, "eval_wer": 10.42682766075242, "step": 16400 }, { "epoch": 7.99, "learning_rate": 2.2212602212602215e-06, "loss": 0.0466, "step": 16410 }, { "epoch": 7.99, "learning_rate": 2.2164502164502168e-06, "loss": 0.0468, "step": 16420 }, { "epoch": 8.0, "learning_rate": 2.211640211640212e-06, "loss": 0.0475, "step": 16430 }, { "epoch": 8.0, "learning_rate": 2.206830206830207e-06, "loss": 0.0447, "step": 16440 }, { "epoch": 8.0, "learning_rate": 2.2020202020202022e-06, "loss": 0.0427, "step": 16450 }, { "epoch": 8.01, "learning_rate": 2.197691197691198e-06, "loss": 0.0431, "step": 16460 }, { "epoch": 8.01, "learning_rate": 2.192881192881193e-06, "loss": 0.0421, "step": 16470 }, { "epoch": 8.02, "learning_rate": 2.1880711880711884e-06, "loss": 0.0419, "step": 16480 }, { "epoch": 8.02, "learning_rate": 2.1832611832611833e-06, "loss": 0.0435, "step": 16490 }, { "epoch": 8.03, "learning_rate": 2.1784511784511786e-06, "loss": 0.0419, "step": 16500 }, { "epoch": 8.03, "learning_rate": 2.173641173641174e-06, "loss": 0.0418, "step": 16510 }, { "epoch": 8.04, "learning_rate": 2.1688311688311687e-06, "loss": 0.0436, "step": 16520 }, { "epoch": 8.04, "learning_rate": 2.164021164021164e-06, "loss": 0.0424, "step": 16530 }, { "epoch": 8.05, "learning_rate": 2.1592111592111593e-06, "loss": 0.0439, "step": 16540 }, { "epoch": 8.05, "learning_rate": 2.1544011544011546e-06, "loss": 0.0436, "step": 16550 }, { "epoch": 8.06, "learning_rate": 2.14959114959115e-06, "loss": 0.0427, "step": 16560 }, { "epoch": 8.06, "learning_rate": 2.144781144781145e-06, "loss": 0.043, "step": 16570 }, { "epoch": 8.07, "learning_rate": 2.13997113997114e-06, "loss": 0.0416, "step": 16580 }, { "epoch": 8.07, "learning_rate": 2.1351611351611354e-06, "loss": 0.0418, "step": 16590 }, { "epoch": 8.08, "learning_rate": 2.1303511303511302e-06, "loss": 0.0419, "step": 16600 }, { "epoch": 8.08, "eval_loss": 0.098876953125, "eval_runtime": 210.6944, "eval_samples_per_second": 15.326, "eval_steps_per_second": 0.062, "eval_wer": 10.57446415860378, "step": 16600 }, { "epoch": 8.08, "learning_rate": 2.1255411255411255e-06, "loss": 0.0418, "step": 16610 }, { "epoch": 8.09, "learning_rate": 2.120731120731121e-06, "loss": 0.0439, "step": 16620 }, { "epoch": 8.09, "learning_rate": 2.115921115921116e-06, "loss": 0.0418, "step": 16630 }, { "epoch": 8.1, "learning_rate": 2.1111111111111114e-06, "loss": 0.0427, "step": 16640 }, { "epoch": 8.1, "learning_rate": 2.1063011063011067e-06, "loss": 0.0425, "step": 16650 }, { "epoch": 8.11, "learning_rate": 2.1014911014911016e-06, "loss": 0.0429, "step": 16660 }, { "epoch": 8.11, "learning_rate": 2.096681096681097e-06, "loss": 0.0462, "step": 16670 }, { "epoch": 8.12, "learning_rate": 2.091871091871092e-06, "loss": 0.0432, "step": 16680 }, { "epoch": 8.12, "learning_rate": 2.087061087061087e-06, "loss": 0.0422, "step": 16690 }, { "epoch": 8.13, "learning_rate": 2.0822510822510823e-06, "loss": 0.043, "step": 16700 }, { "epoch": 8.13, "learning_rate": 2.0774410774410776e-06, "loss": 0.0421, "step": 16710 }, { "epoch": 8.14, "learning_rate": 2.072631072631073e-06, "loss": 0.044, "step": 16720 }, { "epoch": 8.14, "learning_rate": 2.0678210678210682e-06, "loss": 0.044, "step": 16730 }, { "epoch": 8.15, "learning_rate": 2.063011063011063e-06, "loss": 0.0438, "step": 16740 }, { "epoch": 8.15, "learning_rate": 2.0582010582010584e-06, "loss": 0.0441, "step": 16750 }, { "epoch": 8.16, "learning_rate": 2.0533910533910537e-06, "loss": 0.0421, "step": 16760 }, { "epoch": 8.16, "learning_rate": 2.0485810485810485e-06, "loss": 0.0434, "step": 16770 }, { "epoch": 8.17, "learning_rate": 2.043771043771044e-06, "loss": 0.0435, "step": 16780 }, { "epoch": 8.17, "learning_rate": 2.038961038961039e-06, "loss": 0.0418, "step": 16790 }, { "epoch": 8.18, "learning_rate": 2.0341510341510344e-06, "loss": 0.0426, "step": 16800 }, { "epoch": 8.18, "eval_loss": 0.09857177734375, "eval_runtime": 205.5003, "eval_samples_per_second": 15.713, "eval_steps_per_second": 0.063, "eval_wer": 10.516464105876459, "step": 16800 }, { "epoch": 8.18, "learning_rate": 2.0293410293410297e-06, "loss": 0.0451, "step": 16810 }, { "epoch": 8.18, "learning_rate": 2.0245310245310246e-06, "loss": 0.0453, "step": 16820 }, { "epoch": 8.19, "learning_rate": 2.01972101972102e-06, "loss": 0.0445, "step": 16830 }, { "epoch": 8.19, "learning_rate": 2.014911014911015e-06, "loss": 0.0442, "step": 16840 }, { "epoch": 8.2, "learning_rate": 2.01010101010101e-06, "loss": 0.0415, "step": 16850 }, { "epoch": 8.2, "learning_rate": 2.0052910052910053e-06, "loss": 0.0443, "step": 16860 }, { "epoch": 8.21, "learning_rate": 2.0004810004810006e-06, "loss": 0.0424, "step": 16870 }, { "epoch": 8.21, "learning_rate": 1.995670995670996e-06, "loss": 0.0439, "step": 16880 }, { "epoch": 8.22, "learning_rate": 1.9908609908609912e-06, "loss": 0.0435, "step": 16890 }, { "epoch": 8.22, "learning_rate": 1.9860509860509865e-06, "loss": 0.0427, "step": 16900 }, { "epoch": 8.23, "learning_rate": 1.9812409812409814e-06, "loss": 0.0419, "step": 16910 }, { "epoch": 8.23, "learning_rate": 1.9764309764309767e-06, "loss": 0.0429, "step": 16920 }, { "epoch": 8.24, "learning_rate": 1.9716209716209716e-06, "loss": 0.0438, "step": 16930 }, { "epoch": 8.24, "learning_rate": 1.966810966810967e-06, "loss": 0.0443, "step": 16940 }, { "epoch": 8.25, "learning_rate": 1.962000962000962e-06, "loss": 0.0443, "step": 16950 }, { "epoch": 8.25, "learning_rate": 1.9571909571909574e-06, "loss": 0.0439, "step": 16960 }, { "epoch": 8.26, "learning_rate": 1.9523809523809527e-06, "loss": 0.0415, "step": 16970 }, { "epoch": 8.26, "learning_rate": 1.947570947570948e-06, "loss": 0.0426, "step": 16980 }, { "epoch": 8.27, "learning_rate": 1.942760942760943e-06, "loss": 0.0422, "step": 16990 }, { "epoch": 8.27, "learning_rate": 1.937950937950938e-06, "loss": 0.0433, "step": 17000 }, { "epoch": 8.27, "eval_loss": 0.09942626953125, "eval_runtime": 210.7375, "eval_samples_per_second": 15.322, "eval_steps_per_second": 0.062, "eval_wer": 10.45846405314914, "step": 17000 }, { "epoch": 8.28, "learning_rate": 1.9331409331409335e-06, "loss": 0.0431, "step": 17010 }, { "epoch": 8.28, "learning_rate": 1.9283309283309283e-06, "loss": 0.0437, "step": 17020 }, { "epoch": 8.29, "learning_rate": 1.9235209235209236e-06, "loss": 0.045, "step": 17030 }, { "epoch": 8.29, "learning_rate": 1.918710918710919e-06, "loss": 0.0445, "step": 17040 }, { "epoch": 8.3, "learning_rate": 1.9139009139009142e-06, "loss": 0.0435, "step": 17050 }, { "epoch": 8.3, "learning_rate": 1.9090909090909095e-06, "loss": 0.0426, "step": 17060 }, { "epoch": 8.31, "learning_rate": 1.9042809042809044e-06, "loss": 0.0425, "step": 17070 }, { "epoch": 8.31, "learning_rate": 1.8994708994708997e-06, "loss": 0.0446, "step": 17080 }, { "epoch": 8.32, "learning_rate": 1.894660894660895e-06, "loss": 0.0439, "step": 17090 }, { "epoch": 8.32, "learning_rate": 1.8898508898508899e-06, "loss": 0.0417, "step": 17100 }, { "epoch": 8.33, "learning_rate": 1.8850408850408851e-06, "loss": 0.0441, "step": 17110 }, { "epoch": 8.33, "learning_rate": 1.8802308802308804e-06, "loss": 0.0419, "step": 17120 }, { "epoch": 8.34, "learning_rate": 1.8754208754208755e-06, "loss": 0.0443, "step": 17130 }, { "epoch": 8.34, "learning_rate": 1.8706108706108708e-06, "loss": 0.0425, "step": 17140 }, { "epoch": 8.35, "learning_rate": 1.865800865800866e-06, "loss": 0.0435, "step": 17150 }, { "epoch": 8.35, "learning_rate": 1.8609908609908612e-06, "loss": 0.043, "step": 17160 }, { "epoch": 8.36, "learning_rate": 1.8561808561808565e-06, "loss": 0.0428, "step": 17170 }, { "epoch": 8.36, "learning_rate": 1.8513708513708514e-06, "loss": 0.0437, "step": 17180 }, { "epoch": 8.36, "learning_rate": 1.8465608465608467e-06, "loss": 0.0425, "step": 17190 }, { "epoch": 8.37, "learning_rate": 1.841750841750842e-06, "loss": 0.0439, "step": 17200 }, { "epoch": 8.37, "eval_loss": 0.09930419921875, "eval_runtime": 201.2589, "eval_samples_per_second": 16.044, "eval_steps_per_second": 0.065, "eval_wer": 10.53491866810788, "step": 17200 }, { "epoch": 8.37, "learning_rate": 1.836940836940837e-06, "loss": 0.0435, "step": 17210 }, { "epoch": 8.38, "learning_rate": 1.8321308321308323e-06, "loss": 0.0422, "step": 17220 }, { "epoch": 8.38, "learning_rate": 1.8273208273208276e-06, "loss": 0.0451, "step": 17230 }, { "epoch": 8.39, "learning_rate": 1.8225108225108227e-06, "loss": 0.0423, "step": 17240 }, { "epoch": 8.39, "learning_rate": 1.817700817700818e-06, "loss": 0.0434, "step": 17250 }, { "epoch": 8.4, "learning_rate": 1.8128908128908129e-06, "loss": 0.0442, "step": 17260 }, { "epoch": 8.4, "learning_rate": 1.8080808080808082e-06, "loss": 0.0463, "step": 17270 }, { "epoch": 8.41, "learning_rate": 1.8032708032708034e-06, "loss": 0.0436, "step": 17280 }, { "epoch": 8.41, "learning_rate": 1.7984607984607985e-06, "loss": 0.0429, "step": 17290 }, { "epoch": 8.42, "learning_rate": 1.7936507936507938e-06, "loss": 0.0438, "step": 17300 }, { "epoch": 8.42, "learning_rate": 1.7888407888407891e-06, "loss": 0.0435, "step": 17310 }, { "epoch": 8.43, "learning_rate": 1.7840307840307842e-06, "loss": 0.0427, "step": 17320 }, { "epoch": 8.43, "learning_rate": 1.7792207792207795e-06, "loss": 0.043, "step": 17330 }, { "epoch": 8.44, "learning_rate": 1.7744107744107748e-06, "loss": 0.0449, "step": 17340 }, { "epoch": 8.44, "learning_rate": 1.7696007696007697e-06, "loss": 0.0434, "step": 17350 }, { "epoch": 8.45, "learning_rate": 1.764790764790765e-06, "loss": 0.0428, "step": 17360 }, { "epoch": 8.45, "learning_rate": 1.75998075998076e-06, "loss": 0.0419, "step": 17370 }, { "epoch": 8.46, "learning_rate": 1.7551707551707553e-06, "loss": 0.0426, "step": 17380 }, { "epoch": 8.46, "learning_rate": 1.7503607503607506e-06, "loss": 0.0429, "step": 17390 }, { "epoch": 8.47, "learning_rate": 1.7455507455507457e-06, "loss": 0.043, "step": 17400 }, { "epoch": 8.47, "eval_loss": 0.09954833984375, "eval_runtime": 207.7286, "eval_samples_per_second": 15.544, "eval_steps_per_second": 0.063, "eval_wer": 10.445282222983838, "step": 17400 }, { "epoch": 8.47, "learning_rate": 1.740740740740741e-06, "loss": 0.0451, "step": 17410 }, { "epoch": 8.48, "learning_rate": 1.7359307359307363e-06, "loss": 0.0429, "step": 17420 }, { "epoch": 8.48, "learning_rate": 1.7311207311207312e-06, "loss": 0.0448, "step": 17430 }, { "epoch": 8.49, "learning_rate": 1.7263107263107265e-06, "loss": 0.0441, "step": 17440 }, { "epoch": 8.49, "learning_rate": 1.7215007215007215e-06, "loss": 0.0429, "step": 17450 }, { "epoch": 8.5, "learning_rate": 1.7166907166907168e-06, "loss": 0.0425, "step": 17460 }, { "epoch": 8.5, "learning_rate": 1.7118807118807121e-06, "loss": 0.0443, "step": 17470 }, { "epoch": 8.51, "learning_rate": 1.707070707070707e-06, "loss": 0.0435, "step": 17480 }, { "epoch": 8.51, "learning_rate": 1.7022607022607023e-06, "loss": 0.0422, "step": 17490 }, { "epoch": 8.52, "learning_rate": 1.6974506974506976e-06, "loss": 0.0434, "step": 17500 }, { "epoch": 8.52, "learning_rate": 1.6926406926406927e-06, "loss": 0.0458, "step": 17510 }, { "epoch": 8.53, "learning_rate": 1.687830687830688e-06, "loss": 0.0423, "step": 17520 }, { "epoch": 8.53, "learning_rate": 1.6830206830206833e-06, "loss": 0.0418, "step": 17530 }, { "epoch": 8.54, "learning_rate": 1.6782106782106783e-06, "loss": 0.0444, "step": 17540 }, { "epoch": 8.54, "learning_rate": 1.6734006734006736e-06, "loss": 0.0435, "step": 17550 }, { "epoch": 8.55, "learning_rate": 1.6685906685906685e-06, "loss": 0.0421, "step": 17560 }, { "epoch": 8.55, "learning_rate": 1.6637806637806638e-06, "loss": 0.0427, "step": 17570 }, { "epoch": 8.55, "learning_rate": 1.658970658970659e-06, "loss": 0.044, "step": 17580 }, { "epoch": 8.56, "learning_rate": 1.6541606541606542e-06, "loss": 0.043, "step": 17590 }, { "epoch": 8.56, "learning_rate": 1.6493506493506495e-06, "loss": 0.0428, "step": 17600 }, { "epoch": 8.56, "eval_loss": 0.10028076171875, "eval_runtime": 202.281, "eval_samples_per_second": 15.963, "eval_steps_per_second": 0.064, "eval_wer": 10.46637315124832, "step": 17600 }, { "epoch": 8.57, "learning_rate": 1.6445406445406448e-06, "loss": 0.0433, "step": 17610 }, { "epoch": 8.57, "learning_rate": 1.6397306397306398e-06, "loss": 0.0426, "step": 17620 }, { "epoch": 8.58, "learning_rate": 1.6349206349206351e-06, "loss": 0.0425, "step": 17630 }, { "epoch": 8.58, "learning_rate": 1.6301106301106304e-06, "loss": 0.0438, "step": 17640 }, { "epoch": 8.59, "learning_rate": 1.6253006253006253e-06, "loss": 0.0437, "step": 17650 }, { "epoch": 8.59, "learning_rate": 1.6204906204906206e-06, "loss": 0.0445, "step": 17660 }, { "epoch": 8.6, "learning_rate": 1.6156806156806157e-06, "loss": 0.0431, "step": 17670 }, { "epoch": 8.6, "learning_rate": 1.610870610870611e-06, "loss": 0.0453, "step": 17680 }, { "epoch": 8.61, "learning_rate": 1.6060606060606063e-06, "loss": 0.0437, "step": 17690 }, { "epoch": 8.61, "learning_rate": 1.6012506012506013e-06, "loss": 0.0423, "step": 17700 }, { "epoch": 8.62, "learning_rate": 1.5964405964405966e-06, "loss": 0.043, "step": 17710 }, { "epoch": 8.62, "learning_rate": 1.591630591630592e-06, "loss": 0.044, "step": 17720 }, { "epoch": 8.63, "learning_rate": 1.5868205868205868e-06, "loss": 0.0442, "step": 17730 }, { "epoch": 8.63, "learning_rate": 1.582010582010582e-06, "loss": 0.0423, "step": 17740 }, { "epoch": 8.64, "learning_rate": 1.5772005772005774e-06, "loss": 0.0424, "step": 17750 }, { "epoch": 8.64, "learning_rate": 1.5723905723905725e-06, "loss": 0.0439, "step": 17760 }, { "epoch": 8.65, "learning_rate": 1.5675805675805678e-06, "loss": 0.0442, "step": 17770 }, { "epoch": 8.65, "learning_rate": 1.5627705627705629e-06, "loss": 0.0429, "step": 17780 }, { "epoch": 8.66, "learning_rate": 1.5579605579605581e-06, "loss": 0.0437, "step": 17790 }, { "epoch": 8.66, "learning_rate": 1.5531505531505534e-06, "loss": 0.0434, "step": 17800 }, { "epoch": 8.66, "eval_loss": 0.0987548828125, "eval_runtime": 202.5537, "eval_samples_per_second": 15.941, "eval_steps_per_second": 0.064, "eval_wer": 10.50591864174422, "step": 17800 }, { "epoch": 8.67, "learning_rate": 1.5483405483405483e-06, "loss": 0.0444, "step": 17810 }, { "epoch": 8.67, "learning_rate": 1.5435305435305436e-06, "loss": 0.0432, "step": 17820 }, { "epoch": 8.68, "learning_rate": 1.538720538720539e-06, "loss": 0.0437, "step": 17830 }, { "epoch": 8.68, "learning_rate": 1.533910533910534e-06, "loss": 0.0432, "step": 17840 }, { "epoch": 8.69, "learning_rate": 1.5291005291005293e-06, "loss": 0.0443, "step": 17850 }, { "epoch": 8.69, "learning_rate": 1.5242905242905246e-06, "loss": 0.0426, "step": 17860 }, { "epoch": 8.7, "learning_rate": 1.5194805194805196e-06, "loss": 0.0442, "step": 17870 }, { "epoch": 8.7, "learning_rate": 1.514670514670515e-06, "loss": 0.0435, "step": 17880 }, { "epoch": 8.71, "learning_rate": 1.5098605098605098e-06, "loss": 0.0451, "step": 17890 }, { "epoch": 8.71, "learning_rate": 1.5050505050505051e-06, "loss": 0.0434, "step": 17900 }, { "epoch": 8.72, "learning_rate": 1.5002405002405004e-06, "loss": 0.0431, "step": 17910 }, { "epoch": 8.72, "learning_rate": 1.4954304954304955e-06, "loss": 0.0427, "step": 17920 }, { "epoch": 8.73, "learning_rate": 1.4906204906204908e-06, "loss": 0.0422, "step": 17930 }, { "epoch": 8.73, "learning_rate": 1.485810485810486e-06, "loss": 0.0434, "step": 17940 }, { "epoch": 8.73, "learning_rate": 1.4810004810004812e-06, "loss": 0.0445, "step": 17950 }, { "epoch": 8.74, "learning_rate": 1.4761904761904762e-06, "loss": 0.0442, "step": 17960 }, { "epoch": 8.74, "learning_rate": 1.4713804713804715e-06, "loss": 0.043, "step": 17970 }, { "epoch": 8.75, "learning_rate": 1.4665704665704666e-06, "loss": 0.0421, "step": 17980 }, { "epoch": 8.75, "learning_rate": 1.461760461760462e-06, "loss": 0.0443, "step": 17990 }, { "epoch": 8.76, "learning_rate": 1.456950456950457e-06, "loss": 0.0446, "step": 18000 }, { "epoch": 8.76, "eval_loss": 0.09881591796875, "eval_runtime": 198.7229, "eval_samples_per_second": 16.249, "eval_steps_per_second": 0.065, "eval_wer": 10.405736732487938, "step": 18000 }, { "epoch": 8.76, "learning_rate": 1.4521404521404523e-06, "loss": 0.0425, "step": 18010 }, { "epoch": 8.77, "learning_rate": 1.4473304473304476e-06, "loss": 0.043, "step": 18020 }, { "epoch": 8.77, "learning_rate": 1.4425204425204424e-06, "loss": 0.0417, "step": 18030 }, { "epoch": 8.78, "learning_rate": 1.4377104377104377e-06, "loss": 0.042, "step": 18040 }, { "epoch": 8.78, "learning_rate": 1.432900432900433e-06, "loss": 0.0436, "step": 18050 }, { "epoch": 8.79, "learning_rate": 1.4280904280904281e-06, "loss": 0.0431, "step": 18060 }, { "epoch": 8.79, "learning_rate": 1.4232804232804234e-06, "loss": 0.0435, "step": 18070 }, { "epoch": 8.8, "learning_rate": 1.4184704184704187e-06, "loss": 0.0444, "step": 18080 }, { "epoch": 8.8, "learning_rate": 1.4136604136604138e-06, "loss": 0.0434, "step": 18090 }, { "epoch": 8.81, "learning_rate": 1.408850408850409e-06, "loss": 0.0434, "step": 18100 }, { "epoch": 8.81, "learning_rate": 1.404040404040404e-06, "loss": 0.0431, "step": 18110 }, { "epoch": 8.82, "learning_rate": 1.3992303992303992e-06, "loss": 0.0427, "step": 18120 }, { "epoch": 8.82, "learning_rate": 1.3944203944203945e-06, "loss": 0.0453, "step": 18130 }, { "epoch": 8.83, "learning_rate": 1.3896103896103896e-06, "loss": 0.0434, "step": 18140 }, { "epoch": 8.83, "learning_rate": 1.384800384800385e-06, "loss": 0.0449, "step": 18150 }, { "epoch": 8.84, "learning_rate": 1.3799903799903802e-06, "loss": 0.0443, "step": 18160 }, { "epoch": 8.84, "learning_rate": 1.3751803751803753e-06, "loss": 0.0437, "step": 18170 }, { "epoch": 8.85, "learning_rate": 1.3703703703703706e-06, "loss": 0.0427, "step": 18180 }, { "epoch": 8.85, "learning_rate": 1.3655603655603659e-06, "loss": 0.043, "step": 18190 }, { "epoch": 8.86, "learning_rate": 1.3607503607503607e-06, "loss": 0.0432, "step": 18200 }, { "epoch": 8.86, "eval_loss": 0.09991455078125, "eval_runtime": 198.9099, "eval_samples_per_second": 16.233, "eval_steps_per_second": 0.065, "eval_wer": 10.455827687116079, "step": 18200 }, { "epoch": 8.86, "learning_rate": 1.355940355940356e-06, "loss": 0.042, "step": 18210 }, { "epoch": 8.87, "learning_rate": 1.3511303511303511e-06, "loss": 0.0413, "step": 18220 }, { "epoch": 8.87, "learning_rate": 1.3463203463203464e-06, "loss": 0.0405, "step": 18230 }, { "epoch": 8.88, "learning_rate": 1.3415103415103417e-06, "loss": 0.0397, "step": 18240 }, { "epoch": 8.88, "learning_rate": 1.3367003367003368e-06, "loss": 0.0412, "step": 18250 }, { "epoch": 8.89, "learning_rate": 1.331890331890332e-06, "loss": 0.0425, "step": 18260 }, { "epoch": 8.89, "learning_rate": 1.3270803270803274e-06, "loss": 0.0401, "step": 18270 }, { "epoch": 8.9, "learning_rate": 1.3222703222703223e-06, "loss": 0.0391, "step": 18280 }, { "epoch": 8.9, "learning_rate": 1.3174603174603175e-06, "loss": 0.0417, "step": 18290 }, { "epoch": 8.91, "learning_rate": 1.3126503126503126e-06, "loss": 0.0407, "step": 18300 }, { "epoch": 8.91, "learning_rate": 1.307840307840308e-06, "loss": 0.0424, "step": 18310 }, { "epoch": 8.91, "learning_rate": 1.3030303030303032e-06, "loss": 0.0405, "step": 18320 }, { "epoch": 8.92, "learning_rate": 1.2982202982202983e-06, "loss": 0.0403, "step": 18330 }, { "epoch": 8.92, "learning_rate": 1.2934102934102936e-06, "loss": 0.041, "step": 18340 }, { "epoch": 8.93, "learning_rate": 1.2886002886002889e-06, "loss": 0.041, "step": 18350 }, { "epoch": 8.93, "learning_rate": 1.2837902837902838e-06, "loss": 0.0404, "step": 18360 }, { "epoch": 8.94, "learning_rate": 1.278980278980279e-06, "loss": 0.0407, "step": 18370 }, { "epoch": 8.94, "learning_rate": 1.2741702741702743e-06, "loss": 0.0421, "step": 18380 }, { "epoch": 8.95, "learning_rate": 1.2693602693602694e-06, "loss": 0.0401, "step": 18390 }, { "epoch": 8.95, "learning_rate": 1.2645502645502647e-06, "loss": 0.0431, "step": 18400 }, { "epoch": 8.95, "eval_loss": 0.1015625, "eval_runtime": 197.9557, "eval_samples_per_second": 16.312, "eval_steps_per_second": 0.066, "eval_wer": 10.52173683794258, "step": 18400 }, { "epoch": 8.96, "learning_rate": 1.2597402597402598e-06, "loss": 0.0437, "step": 18410 }, { "epoch": 8.96, "learning_rate": 1.254930254930255e-06, "loss": 0.0424, "step": 18420 }, { "epoch": 8.97, "learning_rate": 1.2501202501202504e-06, "loss": 0.0416, "step": 18430 }, { "epoch": 8.97, "learning_rate": 1.2453102453102455e-06, "loss": 0.0419, "step": 18440 }, { "epoch": 8.98, "learning_rate": 1.2405002405002406e-06, "loss": 0.0407, "step": 18450 }, { "epoch": 8.98, "learning_rate": 1.2366522366522367e-06, "loss": 0.0399, "step": 18460 }, { "epoch": 8.99, "learning_rate": 1.231842231842232e-06, "loss": 0.0413, "step": 18470 }, { "epoch": 8.99, "learning_rate": 1.227032227032227e-06, "loss": 0.0414, "step": 18480 }, { "epoch": 9.0, "learning_rate": 1.2222222222222223e-06, "loss": 0.0412, "step": 18490 }, { "epoch": 9.0, "learning_rate": 1.2174122174122176e-06, "loss": 0.041, "step": 18500 }, { "epoch": 9.01, "learning_rate": 1.2126022126022127e-06, "loss": 0.0417, "step": 18510 }, { "epoch": 9.01, "learning_rate": 1.2077922077922078e-06, "loss": 0.0408, "step": 18520 }, { "epoch": 9.02, "learning_rate": 1.202982202982203e-06, "loss": 0.0425, "step": 18530 }, { "epoch": 9.02, "learning_rate": 1.1981721981721984e-06, "loss": 0.0398, "step": 18540 }, { "epoch": 9.03, "learning_rate": 1.1933621933621935e-06, "loss": 0.0405, "step": 18550 }, { "epoch": 9.03, "learning_rate": 1.1885521885521885e-06, "loss": 0.0398, "step": 18560 }, { "epoch": 9.04, "learning_rate": 1.1837421837421838e-06, "loss": 0.0398, "step": 18570 }, { "epoch": 9.04, "learning_rate": 1.1789321789321791e-06, "loss": 0.0412, "step": 18580 }, { "epoch": 9.05, "learning_rate": 1.1741221741221742e-06, "loss": 0.0398, "step": 18590 }, { "epoch": 9.05, "learning_rate": 1.1693121693121693e-06, "loss": 0.0398, "step": 18600 }, { "epoch": 9.05, "eval_loss": 0.101806640625, "eval_runtime": 196.0743, "eval_samples_per_second": 16.468, "eval_steps_per_second": 0.066, "eval_wer": 10.50591864174422, "step": 18600 }, { "epoch": 9.06, "learning_rate": 1.1645021645021646e-06, "loss": 0.0412, "step": 18610 }, { "epoch": 9.06, "learning_rate": 1.1596921596921599e-06, "loss": 0.0405, "step": 18620 }, { "epoch": 9.07, "learning_rate": 1.154882154882155e-06, "loss": 0.041, "step": 18630 }, { "epoch": 9.07, "learning_rate": 1.15007215007215e-06, "loss": 0.0409, "step": 18640 }, { "epoch": 9.08, "learning_rate": 1.1452621452621453e-06, "loss": 0.0411, "step": 18650 }, { "epoch": 9.08, "learning_rate": 1.1404521404521406e-06, "loss": 0.0401, "step": 18660 }, { "epoch": 9.09, "learning_rate": 1.1356421356421357e-06, "loss": 0.0387, "step": 18670 }, { "epoch": 9.09, "learning_rate": 1.1308321308321308e-06, "loss": 0.0414, "step": 18680 }, { "epoch": 9.09, "learning_rate": 1.126022126022126e-06, "loss": 0.0422, "step": 18690 }, { "epoch": 9.1, "learning_rate": 1.1212121212121214e-06, "loss": 0.0393, "step": 18700 }, { "epoch": 9.1, "learning_rate": 1.1164021164021165e-06, "loss": 0.0416, "step": 18710 }, { "epoch": 9.11, "learning_rate": 1.1115921115921118e-06, "loss": 0.042, "step": 18720 }, { "epoch": 9.11, "learning_rate": 1.1067821067821068e-06, "loss": 0.0422, "step": 18730 }, { "epoch": 9.12, "learning_rate": 1.1019721019721021e-06, "loss": 0.0407, "step": 18740 }, { "epoch": 9.12, "learning_rate": 1.0971620971620972e-06, "loss": 0.0415, "step": 18750 }, { "epoch": 9.13, "learning_rate": 1.0923520923520925e-06, "loss": 0.0391, "step": 18760 }, { "epoch": 9.13, "learning_rate": 1.0875420875420876e-06, "loss": 0.0411, "step": 18770 }, { "epoch": 9.14, "learning_rate": 1.0827320827320829e-06, "loss": 0.0413, "step": 18780 }, { "epoch": 9.14, "learning_rate": 1.077922077922078e-06, "loss": 0.0384, "step": 18790 }, { "epoch": 9.15, "learning_rate": 1.0731120731120733e-06, "loss": 0.0417, "step": 18800 }, { "epoch": 9.15, "eval_loss": 0.10235595703125, "eval_runtime": 200.0527, "eval_samples_per_second": 16.141, "eval_steps_per_second": 0.065, "eval_wer": 10.54810049827318, "step": 18800 }, { "epoch": 9.15, "learning_rate": 1.0683020683020683e-06, "loss": 0.04, "step": 18810 }, { "epoch": 9.16, "learning_rate": 1.0634920634920636e-06, "loss": 0.0399, "step": 18820 }, { "epoch": 9.16, "learning_rate": 1.0586820586820587e-06, "loss": 0.0402, "step": 18830 }, { "epoch": 9.17, "learning_rate": 1.053872053872054e-06, "loss": 0.0405, "step": 18840 }, { "epoch": 9.17, "learning_rate": 1.049062049062049e-06, "loss": 0.0402, "step": 18850 }, { "epoch": 9.18, "learning_rate": 1.0442520442520444e-06, "loss": 0.04, "step": 18860 }, { "epoch": 9.18, "learning_rate": 1.0394420394420397e-06, "loss": 0.0416, "step": 18870 }, { "epoch": 9.19, "learning_rate": 1.0346320346320348e-06, "loss": 0.0403, "step": 18880 }, { "epoch": 9.19, "learning_rate": 1.0298220298220299e-06, "loss": 0.0407, "step": 18890 }, { "epoch": 9.2, "learning_rate": 1.0250120250120251e-06, "loss": 0.041, "step": 18900 }, { "epoch": 9.2, "learning_rate": 1.0202020202020202e-06, "loss": 0.0415, "step": 18910 }, { "epoch": 9.21, "learning_rate": 1.0153920153920155e-06, "loss": 0.042, "step": 18920 }, { "epoch": 9.21, "learning_rate": 1.0105820105820106e-06, "loss": 0.0393, "step": 18930 }, { "epoch": 9.22, "learning_rate": 1.0057720057720057e-06, "loss": 0.0389, "step": 18940 }, { "epoch": 9.22, "learning_rate": 1.000962000962001e-06, "loss": 0.0416, "step": 18950 }, { "epoch": 9.23, "learning_rate": 9.961519961519963e-07, "loss": 0.0419, "step": 18960 }, { "epoch": 9.23, "learning_rate": 9.913419913419914e-07, "loss": 0.0409, "step": 18970 }, { "epoch": 9.24, "learning_rate": 9.865319865319867e-07, "loss": 0.0406, "step": 18980 }, { "epoch": 9.24, "learning_rate": 9.817219817219817e-07, "loss": 0.0415, "step": 18990 }, { "epoch": 9.25, "learning_rate": 9.76911976911977e-07, "loss": 0.04, "step": 19000 }, { "epoch": 9.25, "eval_loss": 0.1025390625, "eval_runtime": 196.9266, "eval_samples_per_second": 16.397, "eval_steps_per_second": 0.066, "eval_wer": 10.55600959637236, "step": 19000 }, { "epoch": 9.25, "learning_rate": 9.721019721019721e-07, "loss": 0.04, "step": 19010 }, { "epoch": 9.26, "learning_rate": 9.672919672919674e-07, "loss": 0.0419, "step": 19020 }, { "epoch": 9.26, "learning_rate": 9.624819624819625e-07, "loss": 0.0388, "step": 19030 }, { "epoch": 9.27, "learning_rate": 9.576719576719578e-07, "loss": 0.0394, "step": 19040 }, { "epoch": 9.27, "learning_rate": 9.528619528619529e-07, "loss": 0.0408, "step": 19050 }, { "epoch": 9.27, "learning_rate": 9.480519480519482e-07, "loss": 0.0395, "step": 19060 }, { "epoch": 9.28, "learning_rate": 9.432419432419433e-07, "loss": 0.0402, "step": 19070 }, { "epoch": 9.28, "learning_rate": 9.384319384319384e-07, "loss": 0.0395, "step": 19080 }, { "epoch": 9.29, "learning_rate": 9.336219336219337e-07, "loss": 0.0395, "step": 19090 }, { "epoch": 9.29, "learning_rate": 9.288119288119289e-07, "loss": 0.0418, "step": 19100 }, { "epoch": 9.3, "learning_rate": 9.240019240019241e-07, "loss": 0.0413, "step": 19110 }, { "epoch": 9.3, "learning_rate": 9.191919191919192e-07, "loss": 0.0417, "step": 19120 }, { "epoch": 9.31, "learning_rate": 9.143819143819145e-07, "loss": 0.039, "step": 19130 }, { "epoch": 9.31, "learning_rate": 9.095719095719097e-07, "loss": 0.0418, "step": 19140 }, { "epoch": 9.32, "learning_rate": 9.047619047619048e-07, "loss": 0.0405, "step": 19150 }, { "epoch": 9.32, "learning_rate": 8.999518999518999e-07, "loss": 0.0425, "step": 19160 }, { "epoch": 9.33, "learning_rate": 8.951418951418952e-07, "loss": 0.0405, "step": 19170 }, { "epoch": 9.33, "learning_rate": 8.903318903318904e-07, "loss": 0.0402, "step": 19180 }, { "epoch": 9.34, "learning_rate": 8.855218855218856e-07, "loss": 0.0395, "step": 19190 }, { "epoch": 9.34, "learning_rate": 8.807118807118809e-07, "loss": 0.04, "step": 19200 }, { "epoch": 9.34, "eval_loss": 0.103271484375, "eval_runtime": 200.5219, "eval_samples_per_second": 16.103, "eval_steps_per_second": 0.065, "eval_wer": 10.5138277398434, "step": 19200 }, { "epoch": 9.35, "learning_rate": 8.75901875901876e-07, "loss": 0.0411, "step": 19210 }, { "epoch": 9.35, "learning_rate": 8.710918710918712e-07, "loss": 0.0401, "step": 19220 }, { "epoch": 9.36, "learning_rate": 8.662818662818664e-07, "loss": 0.0406, "step": 19230 }, { "epoch": 9.36, "learning_rate": 8.614718614718616e-07, "loss": 0.0417, "step": 19240 }, { "epoch": 9.37, "learning_rate": 8.566618566618567e-07, "loss": 0.0406, "step": 19250 }, { "epoch": 9.37, "learning_rate": 8.518518518518519e-07, "loss": 0.0407, "step": 19260 }, { "epoch": 9.38, "learning_rate": 8.470418470418471e-07, "loss": 0.0407, "step": 19270 }, { "epoch": 9.38, "learning_rate": 8.422318422318424e-07, "loss": 0.0404, "step": 19280 }, { "epoch": 9.39, "learning_rate": 8.374218374218375e-07, "loss": 0.0426, "step": 19290 }, { "epoch": 9.39, "learning_rate": 8.326118326118327e-07, "loss": 0.0389, "step": 19300 }, { "epoch": 9.4, "learning_rate": 8.278018278018277e-07, "loss": 0.0414, "step": 19310 }, { "epoch": 9.4, "learning_rate": 8.22991822991823e-07, "loss": 0.0401, "step": 19320 }, { "epoch": 9.41, "learning_rate": 8.181818181818182e-07, "loss": 0.0398, "step": 19330 }, { "epoch": 9.41, "learning_rate": 8.133718133718134e-07, "loss": 0.0401, "step": 19340 }, { "epoch": 9.42, "learning_rate": 8.085618085618087e-07, "loss": 0.042, "step": 19350 }, { "epoch": 9.42, "learning_rate": 8.037518037518038e-07, "loss": 0.0404, "step": 19360 }, { "epoch": 9.43, "learning_rate": 7.98941798941799e-07, "loss": 0.0408, "step": 19370 }, { "epoch": 9.43, "learning_rate": 7.941317941317942e-07, "loss": 0.0407, "step": 19380 }, { "epoch": 9.44, "learning_rate": 7.893217893217895e-07, "loss": 0.0418, "step": 19390 }, { "epoch": 9.44, "learning_rate": 7.845117845117845e-07, "loss": 0.0395, "step": 19400 }, { "epoch": 9.44, "eval_loss": 0.10284423828125, "eval_runtime": 201.0026, "eval_samples_per_second": 16.064, "eval_steps_per_second": 0.065, "eval_wer": 10.59819145290132, "step": 19400 }, { "epoch": 9.45, "learning_rate": 7.797017797017797e-07, "loss": 0.0411, "step": 19410 }, { "epoch": 9.45, "learning_rate": 7.748917748917749e-07, "loss": 0.0404, "step": 19420 }, { "epoch": 9.45, "learning_rate": 7.700817700817702e-07, "loss": 0.0412, "step": 19430 }, { "epoch": 9.46, "learning_rate": 7.652717652717653e-07, "loss": 0.0404, "step": 19440 }, { "epoch": 9.46, "learning_rate": 7.604617604617605e-07, "loss": 0.0414, "step": 19450 }, { "epoch": 9.47, "learning_rate": 7.556517556517558e-07, "loss": 0.0402, "step": 19460 }, { "epoch": 9.47, "learning_rate": 7.518037518037519e-07, "loss": 0.0421, "step": 19470 }, { "epoch": 9.48, "learning_rate": 7.46993746993747e-07, "loss": 0.0401, "step": 19480 }, { "epoch": 9.48, "learning_rate": 7.421837421837423e-07, "loss": 0.0391, "step": 19490 }, { "epoch": 9.49, "learning_rate": 7.373737373737375e-07, "loss": 0.04, "step": 19500 }, { "epoch": 9.49, "learning_rate": 7.325637325637326e-07, "loss": 0.0412, "step": 19510 }, { "epoch": 9.5, "learning_rate": 7.277537277537277e-07, "loss": 0.0409, "step": 19520 }, { "epoch": 9.5, "learning_rate": 7.22943722943723e-07, "loss": 0.0416, "step": 19530 }, { "epoch": 9.51, "learning_rate": 7.181337181337182e-07, "loss": 0.041, "step": 19540 }, { "epoch": 9.51, "learning_rate": 7.133237133237134e-07, "loss": 0.0407, "step": 19550 }, { "epoch": 9.52, "learning_rate": 7.085137085137085e-07, "loss": 0.0393, "step": 19560 }, { "epoch": 9.52, "learning_rate": 7.037037037037038e-07, "loss": 0.0419, "step": 19570 }, { "epoch": 9.53, "learning_rate": 6.98893698893699e-07, "loss": 0.0395, "step": 19580 }, { "epoch": 9.53, "learning_rate": 6.940836940836941e-07, "loss": 0.0405, "step": 19590 }, { "epoch": 9.54, "learning_rate": 6.892736892736892e-07, "loss": 0.0396, "step": 19600 }, { "epoch": 9.54, "eval_loss": 0.10308837890625, "eval_runtime": 201.6325, "eval_samples_per_second": 16.014, "eval_steps_per_second": 0.064, "eval_wer": 10.563918694471539, "step": 19600 }, { "epoch": 9.54, "learning_rate": 6.844636844636845e-07, "loss": 0.041, "step": 19610 }, { "epoch": 9.55, "learning_rate": 6.796536796536797e-07, "loss": 0.0391, "step": 19620 }, { "epoch": 9.55, "learning_rate": 6.748436748436749e-07, "loss": 0.042, "step": 19630 }, { "epoch": 9.56, "learning_rate": 6.700336700336702e-07, "loss": 0.0414, "step": 19640 }, { "epoch": 9.56, "learning_rate": 6.652236652236653e-07, "loss": 0.0412, "step": 19650 }, { "epoch": 9.57, "learning_rate": 6.604136604136605e-07, "loss": 0.0396, "step": 19660 }, { "epoch": 9.57, "learning_rate": 6.556036556036557e-07, "loss": 0.0407, "step": 19670 }, { "epoch": 9.58, "learning_rate": 6.507936507936509e-07, "loss": 0.0411, "step": 19680 }, { "epoch": 9.58, "learning_rate": 6.45983645983646e-07, "loss": 0.0422, "step": 19690 }, { "epoch": 9.59, "learning_rate": 6.411736411736412e-07, "loss": 0.0409, "step": 19700 }, { "epoch": 9.59, "learning_rate": 6.363636363636364e-07, "loss": 0.0412, "step": 19710 }, { "epoch": 9.6, "learning_rate": 6.315536315536316e-07, "loss": 0.0407, "step": 19720 }, { "epoch": 9.6, "learning_rate": 6.267436267436268e-07, "loss": 0.041, "step": 19730 }, { "epoch": 9.61, "learning_rate": 6.21933621933622e-07, "loss": 0.0391, "step": 19740 }, { "epoch": 9.61, "learning_rate": 6.171236171236172e-07, "loss": 0.04, "step": 19750 }, { "epoch": 9.62, "learning_rate": 6.123136123136123e-07, "loss": 0.0423, "step": 19760 }, { "epoch": 9.62, "learning_rate": 6.075036075036075e-07, "loss": 0.0392, "step": 19770 }, { "epoch": 9.63, "learning_rate": 6.026936026936027e-07, "loss": 0.0428, "step": 19780 }, { "epoch": 9.63, "learning_rate": 5.978835978835979e-07, "loss": 0.0411, "step": 19790 }, { "epoch": 9.64, "learning_rate": 5.930735930735931e-07, "loss": 0.0408, "step": 19800 }, { "epoch": 9.64, "eval_loss": 0.10260009765625, "eval_runtime": 199.8563, "eval_samples_per_second": 16.157, "eval_steps_per_second": 0.065, "eval_wer": 10.587645988769081, "step": 19800 }, { "epoch": 9.64, "learning_rate": 5.882635882635883e-07, "loss": 0.0401, "step": 19810 }, { "epoch": 9.64, "learning_rate": 5.834535834535835e-07, "loss": 0.0404, "step": 19820 }, { "epoch": 9.65, "learning_rate": 5.786435786435787e-07, "loss": 0.0397, "step": 19830 }, { "epoch": 9.65, "learning_rate": 5.738335738335738e-07, "loss": 0.0402, "step": 19840 }, { "epoch": 9.66, "learning_rate": 5.690235690235691e-07, "loss": 0.0426, "step": 19850 }, { "epoch": 9.66, "learning_rate": 5.642135642135642e-07, "loss": 0.0399, "step": 19860 }, { "epoch": 9.67, "learning_rate": 5.594035594035595e-07, "loss": 0.0403, "step": 19870 }, { "epoch": 9.67, "learning_rate": 5.545935545935546e-07, "loss": 0.0411, "step": 19880 }, { "epoch": 9.68, "learning_rate": 5.497835497835499e-07, "loss": 0.0413, "step": 19890 }, { "epoch": 9.68, "learning_rate": 5.44973544973545e-07, "loss": 0.0417, "step": 19900 }, { "epoch": 9.69, "learning_rate": 5.401635401635403e-07, "loss": 0.0409, "step": 19910 }, { "epoch": 9.69, "learning_rate": 5.353535353535354e-07, "loss": 0.041, "step": 19920 }, { "epoch": 9.7, "learning_rate": 5.305435305435306e-07, "loss": 0.042, "step": 19930 }, { "epoch": 9.7, "learning_rate": 5.257335257335257e-07, "loss": 0.0406, "step": 19940 }, { "epoch": 9.71, "learning_rate": 5.20923520923521e-07, "loss": 0.04, "step": 19950 }, { "epoch": 9.71, "learning_rate": 5.161135161135161e-07, "loss": 0.0405, "step": 19960 }, { "epoch": 9.72, "learning_rate": 5.113035113035113e-07, "loss": 0.0413, "step": 19970 }, { "epoch": 9.72, "learning_rate": 5.064935064935066e-07, "loss": 0.0412, "step": 19980 }, { "epoch": 9.73, "learning_rate": 5.016835016835017e-07, "loss": 0.039, "step": 19990 }, { "epoch": 9.73, "learning_rate": 4.96873496873497e-07, "loss": 0.0403, "step": 20000 }, { "epoch": 9.73, "eval_loss": 0.10296630859375, "eval_runtime": 192.3242, "eval_samples_per_second": 16.789, "eval_steps_per_second": 0.068, "eval_wer": 10.55600959637236, "step": 20000 }, { "epoch": 9.74, "learning_rate": 4.92063492063492e-07, "loss": 0.0415, "step": 20010 }, { "epoch": 9.74, "learning_rate": 4.872534872534873e-07, "loss": 0.0403, "step": 20020 }, { "epoch": 9.75, "learning_rate": 4.824434824434824e-07, "loss": 0.0414, "step": 20030 }, { "epoch": 9.75, "learning_rate": 4.776334776334777e-07, "loss": 0.041, "step": 20040 }, { "epoch": 9.76, "learning_rate": 4.7282347282347285e-07, "loss": 0.0409, "step": 20050 }, { "epoch": 9.76, "learning_rate": 4.6801346801346804e-07, "loss": 0.0413, "step": 20060 }, { "epoch": 9.77, "learning_rate": 4.632034632034632e-07, "loss": 0.0402, "step": 20070 }, { "epoch": 9.77, "learning_rate": 4.583934583934584e-07, "loss": 0.0417, "step": 20080 }, { "epoch": 9.78, "learning_rate": 4.5358345358345365e-07, "loss": 0.0393, "step": 20090 }, { "epoch": 9.78, "learning_rate": 4.487734487734488e-07, "loss": 0.0395, "step": 20100 }, { "epoch": 9.79, "learning_rate": 4.4396344396344403e-07, "loss": 0.0391, "step": 20110 }, { "epoch": 9.79, "learning_rate": 4.3915343915343916e-07, "loss": 0.0411, "step": 20120 }, { "epoch": 9.8, "learning_rate": 4.343434343434344e-07, "loss": 0.0406, "step": 20130 }, { "epoch": 9.8, "learning_rate": 4.2953342953342954e-07, "loss": 0.0419, "step": 20140 }, { "epoch": 9.81, "learning_rate": 4.247234247234248e-07, "loss": 0.0421, "step": 20150 }, { "epoch": 9.81, "learning_rate": 4.199134199134199e-07, "loss": 0.0398, "step": 20160 }, { "epoch": 9.82, "learning_rate": 4.1510341510341516e-07, "loss": 0.04, "step": 20170 }, { "epoch": 9.82, "learning_rate": 4.102934102934103e-07, "loss": 0.0407, "step": 20180 }, { "epoch": 9.82, "learning_rate": 4.0548340548340553e-07, "loss": 0.0413, "step": 20190 }, { "epoch": 9.83, "learning_rate": 4.0067340067340067e-07, "loss": 0.0406, "step": 20200 }, { "epoch": 9.83, "eval_loss": 0.102783203125, "eval_runtime": 197.9977, "eval_samples_per_second": 16.308, "eval_steps_per_second": 0.066, "eval_wer": 10.55073686430624, "step": 20200 }, { "epoch": 9.83, "learning_rate": 3.958633958633959e-07, "loss": 0.0412, "step": 20210 }, { "epoch": 9.84, "learning_rate": 3.910533910533911e-07, "loss": 0.04, "step": 20220 }, { "epoch": 9.84, "learning_rate": 3.862433862433863e-07, "loss": 0.0394, "step": 20230 }, { "epoch": 9.85, "learning_rate": 3.8143338143338147e-07, "loss": 0.0396, "step": 20240 }, { "epoch": 9.85, "learning_rate": 3.7662337662337666e-07, "loss": 0.0395, "step": 20250 }, { "epoch": 9.86, "learning_rate": 3.7181337181337185e-07, "loss": 0.0398, "step": 20260 }, { "epoch": 9.86, "learning_rate": 3.6700336700336704e-07, "loss": 0.0402, "step": 20270 }, { "epoch": 9.87, "learning_rate": 3.621933621933622e-07, "loss": 0.041, "step": 20280 }, { "epoch": 9.87, "learning_rate": 3.573833573833574e-07, "loss": 0.0407, "step": 20290 }, { "epoch": 9.88, "learning_rate": 3.525733525733526e-07, "loss": 0.0404, "step": 20300 }, { "epoch": 9.88, "learning_rate": 3.4776334776334774e-07, "loss": 0.0415, "step": 20310 }, { "epoch": 9.89, "learning_rate": 3.42953342953343e-07, "loss": 0.0405, "step": 20320 }, { "epoch": 9.89, "learning_rate": 3.381433381433382e-07, "loss": 0.0407, "step": 20330 }, { "epoch": 9.9, "learning_rate": 3.3333333333333335e-07, "loss": 0.0413, "step": 20340 }, { "epoch": 9.9, "learning_rate": 3.285233285233286e-07, "loss": 0.0395, "step": 20350 }, { "epoch": 9.91, "learning_rate": 3.2371332371332373e-07, "loss": 0.0415, "step": 20360 }, { "epoch": 9.91, "learning_rate": 3.1890331890331897e-07, "loss": 0.039, "step": 20370 }, { "epoch": 9.92, "learning_rate": 3.140933140933141e-07, "loss": 0.0401, "step": 20380 }, { "epoch": 9.92, "learning_rate": 3.0928330928330935e-07, "loss": 0.041, "step": 20390 }, { "epoch": 9.93, "learning_rate": 3.0447330447330453e-07, "loss": 0.04, "step": 20400 }, { "epoch": 9.93, "eval_loss": 0.1026611328125, "eval_runtime": 201.8654, "eval_samples_per_second": 15.996, "eval_steps_per_second": 0.064, "eval_wer": 10.52173683794258, "step": 20400 }, { "epoch": 9.93, "learning_rate": 2.996632996632997e-07, "loss": 0.0388, "step": 20410 }, { "epoch": 9.94, "learning_rate": 2.948532948532949e-07, "loss": 0.0405, "step": 20420 }, { "epoch": 9.94, "learning_rate": 2.9004329004329005e-07, "loss": 0.0392, "step": 20430 }, { "epoch": 9.95, "learning_rate": 2.8523328523328523e-07, "loss": 0.0415, "step": 20440 }, { "epoch": 9.95, "learning_rate": 2.804232804232804e-07, "loss": 0.041, "step": 20450 }, { "epoch": 9.96, "learning_rate": 2.756132756132756e-07, "loss": 0.0404, "step": 20460 }, { "epoch": 9.96, "learning_rate": 2.7176527176527177e-07, "loss": 0.041, "step": 20470 }, { "epoch": 9.97, "learning_rate": 2.6695526695526696e-07, "loss": 0.0404, "step": 20480 }, { "epoch": 9.97, "learning_rate": 2.6214526214526215e-07, "loss": 0.0429, "step": 20490 }, { "epoch": 9.98, "learning_rate": 2.5733525733525733e-07, "loss": 0.0409, "step": 20500 }, { "epoch": 9.98, "learning_rate": 2.525252525252525e-07, "loss": 0.04, "step": 20510 }, { "epoch": 9.99, "learning_rate": 2.477152477152477e-07, "loss": 0.0399, "step": 20520 }, { "epoch": 9.99, "learning_rate": 2.429052429052429e-07, "loss": 0.0409, "step": 20530 }, { "epoch": 10.0, "learning_rate": 2.3809523809523811e-07, "loss": 0.0424, "step": 20540 }, { "epoch": 10.0, "learning_rate": 2.3328523328523328e-07, "loss": 0.0419, "step": 20550 }, { "epoch": 10.0, "learning_rate": 2.2847522847522852e-07, "loss": 0.0394, "step": 20560 }, { "epoch": 10.01, "learning_rate": 2.236652236652237e-07, "loss": 0.0391, "step": 20570 }, { "epoch": 10.01, "learning_rate": 2.188552188552189e-07, "loss": 0.039, "step": 20580 }, { "epoch": 10.02, "learning_rate": 2.1404521404521408e-07, "loss": 0.0384, "step": 20590 }, { "epoch": 10.02, "learning_rate": 2.0923520923520927e-07, "loss": 0.0383, "step": 20600 }, { "epoch": 10.02, "eval_loss": 0.103515625, "eval_runtime": 201.2189, "eval_samples_per_second": 16.047, "eval_steps_per_second": 0.065, "eval_wer": 10.56128232843848, "step": 20600 }, { "epoch": 10.03, "learning_rate": 2.0442520442520446e-07, "loss": 0.0403, "step": 20610 }, { "epoch": 10.03, "learning_rate": 1.9961519961519962e-07, "loss": 0.0393, "step": 20620 }, { "epoch": 10.04, "learning_rate": 1.948051948051948e-07, "loss": 0.0392, "step": 20630 }, { "epoch": 10.04, "learning_rate": 1.8999518999519e-07, "loss": 0.0393, "step": 20640 }, { "epoch": 10.05, "learning_rate": 1.8518518518518518e-07, "loss": 0.0382, "step": 20650 }, { "epoch": 10.05, "learning_rate": 1.8037518037518037e-07, "loss": 0.0391, "step": 20660 }, { "epoch": 10.06, "learning_rate": 1.7556517556517556e-07, "loss": 0.0398, "step": 20670 }, { "epoch": 10.06, "learning_rate": 1.7075517075517077e-07, "loss": 0.0395, "step": 20680 }, { "epoch": 10.07, "learning_rate": 1.6594516594516596e-07, "loss": 0.0386, "step": 20690 }, { "epoch": 10.07, "learning_rate": 1.6113516113516115e-07, "loss": 0.0398, "step": 20700 }, { "epoch": 10.08, "learning_rate": 1.5632515632515634e-07, "loss": 0.0378, "step": 20710 }, { "epoch": 10.08, "learning_rate": 1.5151515151515152e-07, "loss": 0.0389, "step": 20720 }, { "epoch": 10.09, "learning_rate": 1.467051467051467e-07, "loss": 0.039, "step": 20730 }, { "epoch": 10.09, "learning_rate": 1.418951418951419e-07, "loss": 0.0397, "step": 20740 }, { "epoch": 10.1, "learning_rate": 1.370851370851371e-07, "loss": 0.0384, "step": 20750 }, { "epoch": 10.1, "learning_rate": 1.3227513227513228e-07, "loss": 0.0384, "step": 20760 }, { "epoch": 10.11, "learning_rate": 1.274651274651275e-07, "loss": 0.0402, "step": 20770 }, { "epoch": 10.11, "learning_rate": 1.2265512265512268e-07, "loss": 0.0395, "step": 20780 }, { "epoch": 10.12, "learning_rate": 1.1784511784511785e-07, "loss": 0.0392, "step": 20790 }, { "epoch": 10.12, "learning_rate": 1.1303511303511304e-07, "loss": 0.04, "step": 20800 }, { "epoch": 10.12, "eval_loss": 0.1041259765625, "eval_runtime": 201.9109, "eval_samples_per_second": 15.992, "eval_steps_per_second": 0.064, "eval_wer": 10.5270095700087, "step": 20800 }, { "epoch": 10.13, "learning_rate": 1.0822510822510823e-07, "loss": 0.0392, "step": 20810 }, { "epoch": 10.13, "learning_rate": 1.0341510341510342e-07, "loss": 0.0393, "step": 20820 }, { "epoch": 10.14, "learning_rate": 9.860509860509862e-08, "loss": 0.0382, "step": 20830 }, { "epoch": 10.14, "learning_rate": 9.379509379509381e-08, "loss": 0.0382, "step": 20840 }, { "epoch": 10.15, "learning_rate": 8.8985088985089e-08, "loss": 0.0369, "step": 20850 }, { "epoch": 10.15, "learning_rate": 8.417508417508418e-08, "loss": 0.0379, "step": 20860 }, { "epoch": 10.16, "learning_rate": 7.936507936507937e-08, "loss": 0.0395, "step": 20870 }, { "epoch": 10.16, "learning_rate": 7.455507455507456e-08, "loss": 0.0395, "step": 20880 }, { "epoch": 10.17, "learning_rate": 6.974506974506975e-08, "loss": 0.0394, "step": 20890 }, { "epoch": 10.17, "learning_rate": 6.493506493506495e-08, "loss": 0.0399, "step": 20900 }, { "epoch": 10.18, "learning_rate": 6.012506012506014e-08, "loss": 0.0391, "step": 20910 }, { "epoch": 10.18, "learning_rate": 5.531505531505532e-08, "loss": 0.0401, "step": 20920 }, { "epoch": 10.18, "learning_rate": 5.050505050505051e-08, "loss": 0.0392, "step": 20930 }, { "epoch": 10.19, "learning_rate": 4.56950456950457e-08, "loss": 0.0396, "step": 20940 }, { "epoch": 10.19, "learning_rate": 4.088504088504089e-08, "loss": 0.0375, "step": 20950 }, { "epoch": 10.2, "learning_rate": 3.6075036075036077e-08, "loss": 0.0409, "step": 20960 }, { "epoch": 10.2, "learning_rate": 3.126503126503127e-08, "loss": 0.0393, "step": 20970 }, { "epoch": 10.21, "learning_rate": 2.6455026455026456e-08, "loss": 0.0376, "step": 20980 }, { "epoch": 10.21, "learning_rate": 2.1645021645021647e-08, "loss": 0.0383, "step": 20990 }, { "epoch": 10.22, "learning_rate": 1.6835016835016835e-08, "loss": 0.0382, "step": 21000 }, { "epoch": 10.22, "eval_loss": 0.10406494140625, "eval_runtime": 197.193, "eval_samples_per_second": 16.375, "eval_steps_per_second": 0.066, "eval_wer": 10.563918694471539, "step": 21000 } ], "max_steps": 21000, "num_train_epochs": 11, "total_flos": 1.0969543307456093e+22, "trial_name": null, "trial_params": null }