{ "best_metric": 0.43008479475975037, "best_model_checkpoint": "ai-light-dance_singing_ft_wav2vec2-large-xlsr-53-5gram-v3/checkpoint-552", "epoch": 9.998756078253987, "global_step": 5520, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.02, "learning_rate": 2.0000000000000002e-07, "loss": 0.144, "step": 10 }, { "epoch": 0.04, "learning_rate": 4.0000000000000003e-07, "loss": 0.1475, "step": 20 }, { "epoch": 0.05, "learning_rate": 6.000000000000001e-07, "loss": 0.1668, "step": 30 }, { "epoch": 0.07, "learning_rate": 8.000000000000001e-07, "loss": 0.1406, "step": 40 }, { "epoch": 0.09, "learning_rate": 9.800000000000001e-07, "loss": 0.165, "step": 50 }, { "epoch": 0.11, "learning_rate": 1.1800000000000001e-06, "loss": 0.1566, "step": 60 }, { "epoch": 0.13, "learning_rate": 1.3800000000000001e-06, "loss": 0.1449, "step": 70 }, { "epoch": 0.14, "learning_rate": 1.5800000000000001e-06, "loss": 0.1674, "step": 80 }, { "epoch": 0.16, "learning_rate": 1.7800000000000001e-06, "loss": 0.1563, "step": 90 }, { "epoch": 0.18, "learning_rate": 1.98e-06, "loss": 0.1726, "step": 100 }, { "epoch": 0.2, "learning_rate": 2.1800000000000003e-06, "loss": 0.1373, "step": 110 }, { "epoch": 0.22, "learning_rate": 2.38e-06, "loss": 0.1516, "step": 120 }, { "epoch": 0.24, "learning_rate": 2.5800000000000003e-06, "loss": 0.157, "step": 130 }, { "epoch": 0.25, "learning_rate": 2.7800000000000005e-06, "loss": 0.1434, "step": 140 }, { "epoch": 0.27, "learning_rate": 2.9800000000000003e-06, "loss": 0.157, "step": 150 }, { "epoch": 0.29, "learning_rate": 3.1800000000000005e-06, "loss": 0.153, "step": 160 }, { "epoch": 0.31, "learning_rate": 3.3800000000000007e-06, "loss": 0.1445, "step": 170 }, { "epoch": 0.33, "learning_rate": 3.58e-06, "loss": 0.1579, "step": 180 }, { "epoch": 0.34, "learning_rate": 3.7800000000000002e-06, "loss": 0.1357, "step": 190 }, { "epoch": 0.36, "learning_rate": 3.980000000000001e-06, "loss": 0.145, "step": 200 }, { "epoch": 0.38, "learning_rate": 4.18e-06, "loss": 0.1465, "step": 210 }, { "epoch": 0.4, "learning_rate": 4.38e-06, "loss": 0.1298, "step": 220 }, { "epoch": 0.42, "learning_rate": 4.58e-06, "loss": 0.1416, "step": 230 }, { "epoch": 0.43, "learning_rate": 4.78e-06, "loss": 0.1402, "step": 240 }, { "epoch": 0.45, "learning_rate": 4.980000000000001e-06, "loss": 0.1361, "step": 250 }, { "epoch": 0.47, "learning_rate": 5.18e-06, "loss": 0.1479, "step": 260 }, { "epoch": 0.49, "learning_rate": 5.380000000000001e-06, "loss": 0.1333, "step": 270 }, { "epoch": 0.51, "learning_rate": 5.580000000000001e-06, "loss": 0.1527, "step": 280 }, { "epoch": 0.52, "learning_rate": 5.78e-06, "loss": 0.1352, "step": 290 }, { "epoch": 0.54, "learning_rate": 5.98e-06, "loss": 0.1596, "step": 300 }, { "epoch": 0.56, "learning_rate": 6.18e-06, "loss": 0.1446, "step": 310 }, { "epoch": 0.58, "learning_rate": 6.380000000000001e-06, "loss": 0.146, "step": 320 }, { "epoch": 0.6, "learning_rate": 6.5800000000000005e-06, "loss": 0.1631, "step": 330 }, { "epoch": 0.62, "learning_rate": 6.780000000000001e-06, "loss": 0.1377, "step": 340 }, { "epoch": 0.63, "learning_rate": 6.98e-06, "loss": 0.1541, "step": 350 }, { "epoch": 0.65, "learning_rate": 7.180000000000001e-06, "loss": 0.12, "step": 360 }, { "epoch": 0.67, "learning_rate": 7.3800000000000005e-06, "loss": 0.1318, "step": 370 }, { "epoch": 0.69, "learning_rate": 7.58e-06, "loss": 0.1396, "step": 380 }, { "epoch": 0.71, "learning_rate": 7.78e-06, "loss": 0.138, "step": 390 }, { "epoch": 0.72, "learning_rate": 7.980000000000002e-06, "loss": 0.1683, "step": 400 }, { "epoch": 0.74, "learning_rate": 8.18e-06, "loss": 0.1427, "step": 410 }, { "epoch": 0.76, "learning_rate": 8.380000000000001e-06, "loss": 0.1552, "step": 420 }, { "epoch": 0.78, "learning_rate": 8.580000000000001e-06, "loss": 0.1406, "step": 430 }, { "epoch": 0.8, "learning_rate": 8.78e-06, "loss": 0.1287, "step": 440 }, { "epoch": 0.81, "learning_rate": 8.96e-06, "loss": 0.1341, "step": 450 }, { "epoch": 0.83, "learning_rate": 9.16e-06, "loss": 0.1316, "step": 460 }, { "epoch": 0.85, "learning_rate": 9.360000000000002e-06, "loss": 0.1339, "step": 470 }, { "epoch": 0.87, "learning_rate": 9.56e-06, "loss": 0.1129, "step": 480 }, { "epoch": 0.89, "learning_rate": 9.760000000000001e-06, "loss": 0.1334, "step": 490 }, { "epoch": 0.9, "learning_rate": 9.960000000000001e-06, "loss": 0.1608, "step": 500 }, { "epoch": 0.92, "learning_rate": 9.999937336970615e-06, "loss": 0.1567, "step": 510 }, { "epoch": 0.94, "learning_rate": 9.999682771105642e-06, "loss": 0.1204, "step": 520 }, { "epoch": 0.96, "learning_rate": 9.99923239592798e-06, "loss": 0.128, "step": 530 }, { "epoch": 0.98, "learning_rate": 9.998586229076279e-06, "loss": 0.1416, "step": 540 }, { "epoch": 1.0, "learning_rate": 9.997744295857243e-06, "loss": 0.1517, "step": 550 }, { "epoch": 1.0, "eval_loss": 0.43008479475975037, "eval_runtime": 248.9762, "eval_samples_per_second": 7.892, "eval_steps_per_second": 3.948, "eval_wer": 0.1633294313350995, "step": 552 }, { "epoch": 1.01, "learning_rate": 9.996706629244636e-06, "loss": 0.1469, "step": 560 }, { "epoch": 1.03, "learning_rate": 9.995473269877994e-06, "loss": 0.1207, "step": 570 }, { "epoch": 1.05, "learning_rate": 9.994044266061033e-06, "loss": 0.1239, "step": 580 }, { "epoch": 1.07, "learning_rate": 9.992419673759752e-06, "loss": 0.1294, "step": 590 }, { "epoch": 1.09, "learning_rate": 9.990599556600252e-06, "loss": 0.1446, "step": 600 }, { "epoch": 1.1, "learning_rate": 9.98858398586623e-06, "loss": 0.1397, "step": 610 }, { "epoch": 1.12, "learning_rate": 9.9863730404962e-06, "loss": 0.149, "step": 620 }, { "epoch": 1.14, "learning_rate": 9.983966807080394e-06, "loss": 0.1448, "step": 630 }, { "epoch": 1.16, "learning_rate": 9.981365379857373e-06, "loss": 0.1295, "step": 640 }, { "epoch": 1.18, "learning_rate": 9.97856886071034e-06, "loss": 0.1278, "step": 650 }, { "epoch": 1.2, "learning_rate": 9.975577359163137e-06, "loss": 0.1421, "step": 660 }, { "epoch": 1.21, "learning_rate": 9.972390992375972e-06, "loss": 0.1156, "step": 670 }, { "epoch": 1.23, "learning_rate": 9.969009885140821e-06, "loss": 0.1355, "step": 680 }, { "epoch": 1.25, "learning_rate": 9.96543416987654e-06, "loss": 0.1135, "step": 690 }, { "epoch": 1.27, "learning_rate": 9.961663986623688e-06, "loss": 0.1506, "step": 700 }, { "epoch": 1.29, "learning_rate": 9.957699483039027e-06, "loss": 0.1439, "step": 710 }, { "epoch": 1.3, "learning_rate": 9.953540814389754e-06, "loss": 0.1195, "step": 720 }, { "epoch": 1.32, "learning_rate": 9.949188143547414e-06, "loss": 0.1406, "step": 730 }, { "epoch": 1.34, "learning_rate": 9.944641640981519e-06, "loss": 0.1145, "step": 740 }, { "epoch": 1.36, "learning_rate": 9.939901484752874e-06, "loss": 0.1502, "step": 750 }, { "epoch": 1.38, "learning_rate": 9.934967860506604e-06, "loss": 0.1428, "step": 760 }, { "epoch": 1.39, "learning_rate": 9.929840961464883e-06, "loss": 0.1237, "step": 770 }, { "epoch": 1.41, "learning_rate": 9.924520988419366e-06, "loss": 0.1519, "step": 780 }, { "epoch": 1.43, "learning_rate": 9.919008149723326e-06, "loss": 0.1247, "step": 790 }, { "epoch": 1.45, "learning_rate": 9.91330266128349e-06, "loss": 0.1406, "step": 800 }, { "epoch": 1.47, "learning_rate": 9.90740474655159e-06, "loss": 0.1437, "step": 810 }, { "epoch": 1.48, "learning_rate": 9.901314636515612e-06, "loss": 0.1324, "step": 820 }, { "epoch": 1.5, "learning_rate": 9.895032569690732e-06, "loss": 0.1244, "step": 830 }, { "epoch": 1.52, "learning_rate": 9.888558792110005e-06, "loss": 0.1244, "step": 840 }, { "epoch": 1.54, "learning_rate": 9.881893557314701e-06, "loss": 0.1293, "step": 850 }, { "epoch": 1.56, "learning_rate": 9.875037126344394e-06, "loss": 0.143, "step": 860 }, { "epoch": 1.58, "learning_rate": 9.867989767726722e-06, "loss": 0.1339, "step": 870 }, { "epoch": 1.59, "learning_rate": 9.86075175746689e-06, "loss": 0.14, "step": 880 }, { "epoch": 1.61, "learning_rate": 9.853323379036848e-06, "loss": 0.1424, "step": 890 }, { "epoch": 1.63, "learning_rate": 9.845704923364186e-06, "loss": 0.1209, "step": 900 }, { "epoch": 1.65, "learning_rate": 9.837896688820747e-06, "loss": 0.1518, "step": 910 }, { "epoch": 1.67, "learning_rate": 9.829898981210944e-06, "loss": 0.1168, "step": 920 }, { "epoch": 1.68, "learning_rate": 9.821712113759773e-06, "loss": 0.149, "step": 930 }, { "epoch": 1.7, "learning_rate": 9.813336407100558e-06, "loss": 0.1339, "step": 940 }, { "epoch": 1.72, "learning_rate": 9.80477218926238e-06, "loss": 0.1342, "step": 950 }, { "epoch": 1.74, "learning_rate": 9.796019795657244e-06, "loss": 0.1425, "step": 960 }, { "epoch": 1.76, "learning_rate": 9.787079569066933e-06, "loss": 0.1223, "step": 970 }, { "epoch": 1.77, "learning_rate": 9.777951859629584e-06, "loss": 0.1262, "step": 980 }, { "epoch": 1.79, "learning_rate": 9.768637024825985e-06, "loss": 0.1206, "step": 990 }, { "epoch": 1.81, "learning_rate": 9.759135429465556e-06, "loss": 0.1106, "step": 1000 }, { "epoch": 1.83, "learning_rate": 9.749447445672081e-06, "loss": 0.1214, "step": 1010 }, { "epoch": 1.85, "learning_rate": 9.739573452869123e-06, "loss": 0.119, "step": 1020 }, { "epoch": 1.86, "learning_rate": 9.729513837765166e-06, "loss": 0.1193, "step": 1030 }, { "epoch": 1.88, "learning_rate": 9.719268994338466e-06, "loss": 0.1407, "step": 1040 }, { "epoch": 1.9, "learning_rate": 9.708839323821635e-06, "loss": 0.1281, "step": 1050 }, { "epoch": 1.92, "learning_rate": 9.698225234685909e-06, "loss": 0.1402, "step": 1060 }, { "epoch": 1.94, "learning_rate": 9.687427142625163e-06, "loss": 0.1214, "step": 1070 }, { "epoch": 1.96, "learning_rate": 9.676445470539624e-06, "loss": 0.1193, "step": 1080 }, { "epoch": 1.97, "learning_rate": 9.665280648519311e-06, "loss": 0.154, "step": 1090 }, { "epoch": 1.99, "learning_rate": 9.653933113827196e-06, "loss": 0.1309, "step": 1100 }, { "epoch": 2.0, "eval_loss": 0.43477532267570496, "eval_runtime": 248.2664, "eval_samples_per_second": 7.915, "eval_steps_per_second": 3.959, "eval_wer": 0.1628673525968825, "step": 1104 }, { "epoch": 2.01, "learning_rate": 9.642403310882065e-06, "loss": 0.146, "step": 1110 }, { "epoch": 2.03, "learning_rate": 9.630691691241124e-06, "loss": 0.1305, "step": 1120 }, { "epoch": 2.05, "learning_rate": 9.61879871358232e-06, "loss": 0.1247, "step": 1130 }, { "epoch": 2.07, "learning_rate": 9.606724843686355e-06, "loss": 0.1237, "step": 1140 }, { "epoch": 2.08, "learning_rate": 9.594470554418464e-06, "loss": 0.1182, "step": 1150 }, { "epoch": 2.1, "learning_rate": 9.582036325709892e-06, "loss": 0.1338, "step": 1160 }, { "epoch": 2.12, "learning_rate": 9.569422644539086e-06, "loss": 0.1144, "step": 1170 }, { "epoch": 2.14, "learning_rate": 9.55663000491264e-06, "loss": 0.11, "step": 1180 }, { "epoch": 2.16, "learning_rate": 9.543658907845937e-06, "loss": 0.1153, "step": 1190 }, { "epoch": 2.17, "learning_rate": 9.530509861343524e-06, "loss": 0.1178, "step": 1200 }, { "epoch": 2.19, "learning_rate": 9.51718338037923e-06, "loss": 0.13, "step": 1210 }, { "epoch": 2.21, "learning_rate": 9.50367998687598e-06, "loss": 0.1302, "step": 1220 }, { "epoch": 2.23, "learning_rate": 9.49000020968537e-06, "loss": 0.1282, "step": 1230 }, { "epoch": 2.25, "learning_rate": 9.476144584566947e-06, "loss": 0.1275, "step": 1240 }, { "epoch": 2.26, "learning_rate": 9.462113654167222e-06, "loss": 0.1282, "step": 1250 }, { "epoch": 2.28, "learning_rate": 9.447907967998431e-06, "loss": 0.12, "step": 1260 }, { "epoch": 2.3, "learning_rate": 9.433528082416999e-06, "loss": 0.1156, "step": 1270 }, { "epoch": 2.32, "learning_rate": 9.418974560601761e-06, "loss": 0.1073, "step": 1280 }, { "epoch": 2.34, "learning_rate": 9.404247972531897e-06, "loss": 0.1139, "step": 1290 }, { "epoch": 2.35, "learning_rate": 9.389348894964624e-06, "loss": 0.1262, "step": 1300 }, { "epoch": 2.37, "learning_rate": 9.374277911412586e-06, "loss": 0.1226, "step": 1310 }, { "epoch": 2.39, "learning_rate": 9.359035612121018e-06, "loss": 0.1027, "step": 1320 }, { "epoch": 2.41, "learning_rate": 9.343622594044626e-06, "loss": 0.121, "step": 1330 }, { "epoch": 2.43, "learning_rate": 9.328039460824204e-06, "loss": 0.1153, "step": 1340 }, { "epoch": 2.45, "learning_rate": 9.31228682276299e-06, "loss": 0.126, "step": 1350 }, { "epoch": 2.46, "learning_rate": 9.296365296802778e-06, "loss": 0.1183, "step": 1360 }, { "epoch": 2.48, "learning_rate": 9.280275506499736e-06, "loss": 0.1305, "step": 1370 }, { "epoch": 2.5, "learning_rate": 9.264018082000007e-06, "loss": 0.1311, "step": 1380 }, { "epoch": 2.52, "learning_rate": 9.247593660015001e-06, "loss": 0.1342, "step": 1390 }, { "epoch": 2.54, "learning_rate": 9.231002883796495e-06, "loss": 0.1285, "step": 1400 }, { "epoch": 2.55, "learning_rate": 9.21424640311141e-06, "loss": 0.136, "step": 1410 }, { "epoch": 2.57, "learning_rate": 9.197324874216372e-06, "loss": 0.1104, "step": 1420 }, { "epoch": 2.59, "learning_rate": 9.18023895983202e-06, "loss": 0.1212, "step": 1430 }, { "epoch": 2.61, "learning_rate": 9.162989329117037e-06, "loss": 0.1107, "step": 1440 }, { "epoch": 2.63, "learning_rate": 9.145576657641954e-06, "loss": 0.0921, "step": 1450 }, { "epoch": 2.64, "learning_rate": 9.128001627362683e-06, "loss": 0.1335, "step": 1460 }, { "epoch": 2.66, "learning_rate": 9.110264926593817e-06, "loss": 0.1136, "step": 1470 }, { "epoch": 2.68, "learning_rate": 9.092367249981663e-06, "loss": 0.1241, "step": 1480 }, { "epoch": 2.7, "learning_rate": 9.074309298477048e-06, "loss": 0.1268, "step": 1490 }, { "epoch": 2.72, "learning_rate": 9.056091779307854e-06, "loss": 0.1104, "step": 1500 }, { "epoch": 2.73, "learning_rate": 9.037715405951332e-06, "loss": 0.1357, "step": 1510 }, { "epoch": 2.75, "learning_rate": 9.019180898106148e-06, "loss": 0.1361, "step": 1520 }, { "epoch": 2.77, "learning_rate": 9.00048898166421e-06, "loss": 0.1138, "step": 1530 }, { "epoch": 2.79, "learning_rate": 8.981640388682222e-06, "loss": 0.122, "step": 1540 }, { "epoch": 2.81, "learning_rate": 8.962635857353028e-06, "loss": 0.1273, "step": 1550 }, { "epoch": 2.83, "learning_rate": 8.943476131976694e-06, "loss": 0.1248, "step": 1560 }, { "epoch": 2.84, "learning_rate": 8.92416196293136e-06, "loss": 0.1027, "step": 1570 }, { "epoch": 2.86, "learning_rate": 8.904694106643852e-06, "loss": 0.1076, "step": 1580 }, { "epoch": 2.88, "learning_rate": 8.885073325560054e-06, "loss": 0.1234, "step": 1590 }, { "epoch": 2.9, "learning_rate": 8.865300388115053e-06, "loss": 0.1206, "step": 1600 }, { "epoch": 2.92, "learning_rate": 8.84537606870304e-06, "loss": 0.1253, "step": 1610 }, { "epoch": 2.93, "learning_rate": 8.825301147646978e-06, "loss": 0.1199, "step": 1620 }, { "epoch": 2.95, "learning_rate": 8.805076411168054e-06, "loss": 0.1119, "step": 1630 }, { "epoch": 2.97, "learning_rate": 8.784702651354866e-06, "loss": 0.1058, "step": 1640 }, { "epoch": 2.99, "learning_rate": 8.764180666132424e-06, "loss": 0.1237, "step": 1650 }, { "epoch": 3.0, "eval_loss": 0.4611044228076935, "eval_runtime": 252.1565, "eval_samples_per_second": 7.793, "eval_steps_per_second": 3.898, "eval_wer": 0.1604337379089397, "step": 1656 }, { "epoch": 3.01, "learning_rate": 8.743511259230885e-06, "loss": 0.1404, "step": 1660 }, { "epoch": 3.03, "learning_rate": 8.72269524015408e-06, "loss": 0.1113, "step": 1670 }, { "epoch": 3.04, "learning_rate": 8.701733424147808e-06, "loss": 0.1023, "step": 1680 }, { "epoch": 3.06, "learning_rate": 8.68062663216792e-06, "loss": 0.1219, "step": 1690 }, { "epoch": 3.08, "learning_rate": 8.65937569084814e-06, "loss": 0.1072, "step": 1700 }, { "epoch": 3.1, "learning_rate": 8.637981432467719e-06, "loss": 0.1304, "step": 1710 }, { "epoch": 3.12, "learning_rate": 8.616444694918821e-06, "loss": 0.0999, "step": 1720 }, { "epoch": 3.13, "learning_rate": 8.594766321673723e-06, "loss": 0.113, "step": 1730 }, { "epoch": 3.15, "learning_rate": 8.572947161751757e-06, "loss": 0.1113, "step": 1740 }, { "epoch": 3.17, "learning_rate": 8.550988069686085e-06, "loss": 0.1135, "step": 1750 }, { "epoch": 3.19, "learning_rate": 8.528889905490217e-06, "loss": 0.1201, "step": 1760 }, { "epoch": 3.21, "learning_rate": 8.506653534624327e-06, "loss": 0.1119, "step": 1770 }, { "epoch": 3.22, "learning_rate": 8.484279827961372e-06, "loss": 0.1103, "step": 1780 }, { "epoch": 3.24, "learning_rate": 8.461769661752968e-06, "loss": 0.102, "step": 1790 }, { "epoch": 3.26, "learning_rate": 8.439123917595086e-06, "loss": 0.1319, "step": 1800 }, { "epoch": 3.28, "learning_rate": 8.416343482393514e-06, "loss": 0.1242, "step": 1810 }, { "epoch": 3.3, "learning_rate": 8.393429248329131e-06, "loss": 0.1322, "step": 1820 }, { "epoch": 3.31, "learning_rate": 8.370382112822957e-06, "loss": 0.1248, "step": 1830 }, { "epoch": 3.33, "learning_rate": 8.347202978501013e-06, "loss": 0.1116, "step": 1840 }, { "epoch": 3.35, "learning_rate": 8.323892753158969e-06, "loss": 0.123, "step": 1850 }, { "epoch": 3.37, "learning_rate": 8.300452349726585e-06, "loss": 0.1335, "step": 1860 }, { "epoch": 3.39, "learning_rate": 8.276882686231961e-06, "loss": 0.1048, "step": 1870 }, { "epoch": 3.41, "learning_rate": 8.253184685765591e-06, "loss": 0.1112, "step": 1880 }, { "epoch": 3.42, "learning_rate": 8.22935927644419e-06, "loss": 0.1093, "step": 1890 }, { "epoch": 3.44, "learning_rate": 8.205407391374372e-06, "loss": 0.1108, "step": 1900 }, { "epoch": 3.46, "learning_rate": 8.181329968616077e-06, "loss": 0.1261, "step": 1910 }, { "epoch": 3.48, "learning_rate": 8.157127951145855e-06, "loss": 0.1012, "step": 1920 }, { "epoch": 3.5, "learning_rate": 8.132802286819925e-06, "loss": 0.1117, "step": 1930 }, { "epoch": 3.51, "learning_rate": 8.108353928337056e-06, "loss": 0.0956, "step": 1940 }, { "epoch": 3.53, "learning_rate": 8.083783833201249e-06, "loss": 0.1005, "step": 1950 }, { "epoch": 3.55, "learning_rate": 8.059092963684246e-06, "loss": 0.1093, "step": 1960 }, { "epoch": 3.57, "learning_rate": 8.034282286787837e-06, "loss": 0.1173, "step": 1970 }, { "epoch": 3.59, "learning_rate": 8.00935277420599e-06, "loss": 0.1014, "step": 1980 }, { "epoch": 3.6, "learning_rate": 7.984305402286793e-06, "loss": 0.1186, "step": 1990 }, { "epoch": 3.62, "learning_rate": 7.959141151994223e-06, "loss": 0.1099, "step": 2000 }, { "epoch": 3.64, "learning_rate": 7.933861008869718e-06, "loss": 0.0937, "step": 2010 }, { "epoch": 3.66, "learning_rate": 7.908465962993584e-06, "loss": 0.1138, "step": 2020 }, { "epoch": 3.68, "learning_rate": 7.88295700894622e-06, "loss": 0.0977, "step": 2030 }, { "epoch": 3.69, "learning_rate": 7.85733514576916e-06, "loss": 0.094, "step": 2040 }, { "epoch": 3.71, "learning_rate": 7.831601376925957e-06, "loss": 0.1073, "step": 2050 }, { "epoch": 3.73, "learning_rate": 7.80575671026287e-06, "loss": 0.1082, "step": 2060 }, { "epoch": 3.75, "learning_rate": 7.779802157969402e-06, "loss": 0.0955, "step": 2070 }, { "epoch": 3.77, "learning_rate": 7.753738736538654e-06, "loss": 0.1182, "step": 2080 }, { "epoch": 3.79, "learning_rate": 7.727567466727517e-06, "loss": 0.1198, "step": 2090 }, { "epoch": 3.8, "learning_rate": 7.701289373516692e-06, "loss": 0.1086, "step": 2100 }, { "epoch": 3.82, "learning_rate": 7.67490548607055e-06, "loss": 0.1268, "step": 2110 }, { "epoch": 3.84, "learning_rate": 7.648416837696821e-06, "loss": 0.1117, "step": 2120 }, { "epoch": 3.86, "learning_rate": 7.621824465806134e-06, "loss": 0.1051, "step": 2130 }, { "epoch": 3.88, "learning_rate": 7.595129411871376e-06, "loss": 0.1251, "step": 2140 }, { "epoch": 3.89, "learning_rate": 7.568332721386913e-06, "loss": 0.1011, "step": 2150 }, { "epoch": 3.91, "learning_rate": 7.541435443827638e-06, "loss": 0.1136, "step": 2160 }, { "epoch": 3.93, "learning_rate": 7.5144386326078715e-06, "loss": 0.105, "step": 2170 }, { "epoch": 3.95, "learning_rate": 7.487343345040109e-06, "loss": 0.0933, "step": 2180 }, { "epoch": 3.97, "learning_rate": 7.460150642293601e-06, "loss": 0.1097, "step": 2190 }, { "epoch": 3.98, "learning_rate": 7.432861589352804e-06, "loss": 0.1056, "step": 2200 }, { "epoch": 4.0, "eval_loss": 0.454052597284317, "eval_runtime": 253.3273, "eval_samples_per_second": 7.757, "eval_steps_per_second": 3.88, "eval_wer": 0.15738401823670753, "step": 2208 }, { "epoch": 4.0, "learning_rate": 7.40547725497567e-06, "loss": 0.1269, "step": 2210 }, { "epoch": 4.02, "learning_rate": 7.3779987116517745e-06, "loss": 0.1024, "step": 2220 }, { "epoch": 4.04, "learning_rate": 7.35042703556034e-06, "loss": 0.094, "step": 2230 }, { "epoch": 4.06, "learning_rate": 7.322763306528063e-06, "loss": 0.0975, "step": 2240 }, { "epoch": 4.08, "learning_rate": 7.2950086079868375e-06, "loss": 0.1024, "step": 2250 }, { "epoch": 4.09, "learning_rate": 7.267164026931317e-06, "loss": 0.1076, "step": 2260 }, { "epoch": 4.11, "learning_rate": 7.239230653876345e-06, "loss": 0.1134, "step": 2270 }, { "epoch": 4.13, "learning_rate": 7.2112095828142484e-06, "loss": 0.0762, "step": 2280 }, { "epoch": 4.15, "learning_rate": 7.1831019111719855e-06, "loss": 0.0982, "step": 2290 }, { "epoch": 4.17, "learning_rate": 7.154908739768173e-06, "loss": 0.1, "step": 2300 }, { "epoch": 4.18, "learning_rate": 7.126631172769968e-06, "loss": 0.1076, "step": 2310 }, { "epoch": 4.2, "learning_rate": 7.098270317649827e-06, "loss": 0.0846, "step": 2320 }, { "epoch": 4.22, "learning_rate": 7.06982728514213e-06, "loss": 0.0915, "step": 2330 }, { "epoch": 4.24, "learning_rate": 7.041303189199684e-06, "loss": 0.0962, "step": 2340 }, { "epoch": 4.26, "learning_rate": 7.012699146950088e-06, "loss": 0.0853, "step": 2350 }, { "epoch": 4.28, "learning_rate": 6.98401627865199e-06, "loss": 0.0997, "step": 2360 }, { "epoch": 4.29, "learning_rate": 6.955255707651207e-06, "loss": 0.0927, "step": 2370 }, { "epoch": 4.31, "learning_rate": 6.926418560336731e-06, "loss": 0.0992, "step": 2380 }, { "epoch": 4.33, "learning_rate": 6.897505966096616e-06, "loss": 0.1307, "step": 2390 }, { "epoch": 4.35, "learning_rate": 6.868519057273744e-06, "loss": 0.1065, "step": 2400 }, { "epoch": 4.37, "learning_rate": 6.839458969121485e-06, "loss": 0.1049, "step": 2410 }, { "epoch": 4.38, "learning_rate": 6.810326839759221e-06, "loss": 0.1004, "step": 2420 }, { "epoch": 4.4, "learning_rate": 6.7811238101277876e-06, "loss": 0.0894, "step": 2430 }, { "epoch": 4.42, "learning_rate": 6.751851023944781e-06, "loss": 0.1119, "step": 2440 }, { "epoch": 4.44, "learning_rate": 6.722509627659768e-06, "loss": 0.0798, "step": 2450 }, { "epoch": 4.46, "learning_rate": 6.693100770409381e-06, "loss": 0.1138, "step": 2460 }, { "epoch": 4.47, "learning_rate": 6.663625603972323e-06, "loss": 0.0846, "step": 2470 }, { "epoch": 4.49, "learning_rate": 6.634085282724252e-06, "loss": 0.1088, "step": 2480 }, { "epoch": 4.51, "learning_rate": 6.604480963592567e-06, "loss": 0.094, "step": 2490 }, { "epoch": 4.53, "learning_rate": 6.574813806011107e-06, "loss": 0.1107, "step": 2500 }, { "epoch": 4.55, "learning_rate": 6.545084971874738e-06, "loss": 0.1171, "step": 2510 }, { "epoch": 4.56, "learning_rate": 6.515295625493844e-06, "loss": 0.0954, "step": 2520 }, { "epoch": 4.58, "learning_rate": 6.485446933548735e-06, "loss": 0.0925, "step": 2530 }, { "epoch": 4.6, "learning_rate": 6.45554006504395e-06, "loss": 0.0928, "step": 2540 }, { "epoch": 4.62, "learning_rate": 6.425576191262476e-06, "loss": 0.1076, "step": 2550 }, { "epoch": 4.64, "learning_rate": 6.395556485719873e-06, "loss": 0.0999, "step": 2560 }, { "epoch": 4.65, "learning_rate": 6.3654821241183155e-06, "loss": 0.1252, "step": 2570 }, { "epoch": 4.67, "learning_rate": 6.33535428430055e-06, "loss": 0.1073, "step": 2580 }, { "epoch": 4.69, "learning_rate": 6.305174146203757e-06, "loss": 0.0933, "step": 2590 }, { "epoch": 4.71, "learning_rate": 6.27494289181335e-06, "loss": 0.0947, "step": 2600 }, { "epoch": 4.73, "learning_rate": 6.244661705116678e-06, "loss": 0.1053, "step": 2610 }, { "epoch": 4.75, "learning_rate": 6.214331772056652e-06, "loss": 0.0978, "step": 2620 }, { "epoch": 4.76, "learning_rate": 6.183954280485307e-06, "loss": 0.0874, "step": 2630 }, { "epoch": 4.78, "learning_rate": 6.153530420117275e-06, "loss": 0.1032, "step": 2640 }, { "epoch": 4.8, "learning_rate": 6.123061382483189e-06, "loss": 0.1009, "step": 2650 }, { "epoch": 4.82, "learning_rate": 6.092548360883024e-06, "loss": 0.1125, "step": 2660 }, { "epoch": 4.84, "learning_rate": 6.061992550339357e-06, "loss": 0.0917, "step": 2670 }, { "epoch": 4.85, "learning_rate": 6.031395147550566e-06, "loss": 0.0977, "step": 2680 }, { "epoch": 4.87, "learning_rate": 6.00075735084396e-06, "loss": 0.1141, "step": 2690 }, { "epoch": 4.89, "learning_rate": 5.970080360128851e-06, "loss": 0.0985, "step": 2700 }, { "epoch": 4.91, "learning_rate": 5.939365376849564e-06, "loss": 0.1162, "step": 2710 }, { "epoch": 4.93, "learning_rate": 5.908613603938367e-06, "loss": 0.0979, "step": 2720 }, { "epoch": 4.94, "learning_rate": 5.8778262457683776e-06, "loss": 0.0816, "step": 2730 }, { "epoch": 4.96, "learning_rate": 5.847004508106386e-06, "loss": 0.109, "step": 2740 }, { "epoch": 4.98, "learning_rate": 5.816149598065632e-06, "loss": 0.105, "step": 2750 }, { "epoch": 5.0, "learning_rate": 5.78526272405853e-06, "loss": 0.1236, "step": 2760 }, { "epoch": 5.0, "eval_loss": 0.46689143776893616, "eval_runtime": 244.9078, "eval_samples_per_second": 8.023, "eval_steps_per_second": 4.014, "eval_wer": 0.1603413221612963, "step": 2760 }, { "epoch": 5.02, "learning_rate": 5.754345095749342e-06, "loss": 0.1158, "step": 2770 }, { "epoch": 5.04, "learning_rate": 5.723397924006802e-06, "loss": 0.1168, "step": 2780 }, { "epoch": 5.05, "learning_rate": 5.692422420856694e-06, "loss": 0.1023, "step": 2790 }, { "epoch": 5.07, "learning_rate": 5.66141979943438e-06, "loss": 0.1093, "step": 2800 }, { "epoch": 5.09, "learning_rate": 5.6303912739373e-06, "loss": 0.1154, "step": 2810 }, { "epoch": 5.11, "learning_rate": 5.5993380595773995e-06, "loss": 0.1025, "step": 2820 }, { "epoch": 5.13, "learning_rate": 5.5682613725335534e-06, "loss": 0.1118, "step": 2830 }, { "epoch": 5.14, "learning_rate": 5.537162429903928e-06, "loss": 0.106, "step": 2840 }, { "epoch": 5.16, "learning_rate": 5.5060424496583145e-06, "loss": 0.1161, "step": 2850 }, { "epoch": 5.18, "learning_rate": 5.474902650590428e-06, "loss": 0.1126, "step": 2860 }, { "epoch": 5.2, "learning_rate": 5.443744252270178e-06, "loss": 0.1083, "step": 2870 }, { "epoch": 5.22, "learning_rate": 5.412568474995896e-06, "loss": 0.1042, "step": 2880 }, { "epoch": 5.24, "learning_rate": 5.381376539746553e-06, "loss": 0.1085, "step": 2890 }, { "epoch": 5.25, "learning_rate": 5.350169668133937e-06, "loss": 0.0952, "step": 2900 }, { "epoch": 5.27, "learning_rate": 5.318949082354808e-06, "loss": 0.1085, "step": 2910 }, { "epoch": 5.29, "learning_rate": 5.287716005143028e-06, "loss": 0.106, "step": 2920 }, { "epoch": 5.31, "learning_rate": 5.256471659721685e-06, "loss": 0.1054, "step": 2930 }, { "epoch": 5.33, "learning_rate": 5.225217269755176e-06, "loss": 0.0925, "step": 2940 }, { "epoch": 5.34, "learning_rate": 5.193954059301285e-06, "loss": 0.1117, "step": 2950 }, { "epoch": 5.36, "learning_rate": 5.162683252763246e-06, "loss": 0.1271, "step": 2960 }, { "epoch": 5.38, "learning_rate": 5.131406074841787e-06, "loss": 0.1072, "step": 2970 }, { "epoch": 5.4, "learning_rate": 5.100123750487168e-06, "loss": 0.0969, "step": 2980 }, { "epoch": 5.42, "learning_rate": 5.068837504851208e-06, "loss": 0.1127, "step": 2990 }, { "epoch": 5.43, "learning_rate": 5.0375485632392975e-06, "loss": 0.1013, "step": 3000 }, { "epoch": 5.45, "learning_rate": 5.006258151062413e-06, "loss": 0.1036, "step": 3010 }, { "epoch": 5.47, "learning_rate": 4.974967493789126e-06, "loss": 0.0939, "step": 3020 }, { "epoch": 5.49, "learning_rate": 4.943677816897609e-06, "loss": 0.0949, "step": 3030 }, { "epoch": 5.51, "learning_rate": 4.912390345827634e-06, "loss": 0.1036, "step": 3040 }, { "epoch": 5.52, "learning_rate": 4.881106305932585e-06, "loss": 0.1076, "step": 3050 }, { "epoch": 5.54, "learning_rate": 4.849826922431469e-06, "loss": 0.1034, "step": 3060 }, { "epoch": 5.56, "learning_rate": 4.818553420360927e-06, "loss": 0.1004, "step": 3070 }, { "epoch": 5.58, "learning_rate": 4.787287024527254e-06, "loss": 0.1096, "step": 3080 }, { "epoch": 5.6, "learning_rate": 4.756028959458438e-06, "loss": 0.09, "step": 3090 }, { "epoch": 5.62, "learning_rate": 4.724780449356199e-06, "loss": 0.1059, "step": 3100 }, { "epoch": 5.63, "learning_rate": 4.693542718048037e-06, "loss": 0.1041, "step": 3110 }, { "epoch": 5.65, "learning_rate": 4.662316988939314e-06, "loss": 0.1053, "step": 3120 }, { "epoch": 5.67, "learning_rate": 4.631104484965331e-06, "loss": 0.1033, "step": 3130 }, { "epoch": 5.69, "learning_rate": 4.599906428543435e-06, "loss": 0.0967, "step": 3140 }, { "epoch": 5.71, "learning_rate": 4.568724041525143e-06, "loss": 0.1109, "step": 3150 }, { "epoch": 5.72, "learning_rate": 4.537558545148292e-06, "loss": 0.1208, "step": 3160 }, { "epoch": 5.74, "learning_rate": 4.50641115998921e-06, "loss": 0.1004, "step": 3170 }, { "epoch": 5.76, "learning_rate": 4.475283105914905e-06, "loss": 0.1081, "step": 3180 }, { "epoch": 5.78, "learning_rate": 4.444175602035301e-06, "loss": 0.0996, "step": 3190 }, { "epoch": 5.8, "learning_rate": 4.4130898666554845e-06, "loss": 0.0956, "step": 3200 }, { "epoch": 5.81, "learning_rate": 4.382027117227993e-06, "loss": 0.1069, "step": 3210 }, { "epoch": 5.83, "learning_rate": 4.350988570305136e-06, "loss": 0.0872, "step": 3220 }, { "epoch": 5.85, "learning_rate": 4.3199754414913455e-06, "loss": 0.1086, "step": 3230 }, { "epoch": 5.87, "learning_rate": 4.288988945395572e-06, "loss": 0.1122, "step": 3240 }, { "epoch": 5.89, "learning_rate": 4.258030295583709e-06, "loss": 0.1044, "step": 3250 }, { "epoch": 5.9, "learning_rate": 4.227100704531075e-06, "loss": 0.1083, "step": 3260 }, { "epoch": 5.92, "learning_rate": 4.1962013835749145e-06, "loss": 0.1024, "step": 3270 }, { "epoch": 5.94, "learning_rate": 4.1653335428669675e-06, "loss": 0.1036, "step": 3280 }, { "epoch": 5.96, "learning_rate": 4.134498391326069e-06, "loss": 0.1133, "step": 3290 }, { "epoch": 5.98, "learning_rate": 4.1036971365908055e-06, "loss": 0.095, "step": 3300 }, { "epoch": 6.0, "learning_rate": 4.072930984972215e-06, "loss": 0.1118, "step": 3310 }, { "epoch": 6.0, "eval_loss": 0.46403050422668457, "eval_runtime": 245.9183, "eval_samples_per_second": 7.99, "eval_steps_per_second": 3.997, "eval_wer": 0.15673710800320376, "step": 3312 }, { "epoch": 6.01, "learning_rate": 4.042201141406547e-06, "loss": 0.1158, "step": 3320 }, { "epoch": 6.03, "learning_rate": 4.011508809408068e-06, "loss": 0.1031, "step": 3330 }, { "epoch": 6.05, "learning_rate": 3.980855191021931e-06, "loss": 0.0957, "step": 3340 }, { "epoch": 6.07, "learning_rate": 3.9502414867770916e-06, "loss": 0.1012, "step": 3350 }, { "epoch": 6.09, "learning_rate": 3.919668895639298e-06, "loss": 0.1092, "step": 3360 }, { "epoch": 6.1, "learning_rate": 3.8891386149641304e-06, "loss": 0.1065, "step": 3370 }, { "epoch": 6.12, "learning_rate": 3.858651840450103e-06, "loss": 0.102, "step": 3380 }, { "epoch": 6.14, "learning_rate": 3.828209766091844e-06, "loss": 0.1022, "step": 3390 }, { "epoch": 6.16, "learning_rate": 3.797813584133327e-06, "loss": 0.0941, "step": 3400 }, { "epoch": 6.18, "learning_rate": 3.7674644850211815e-06, "loss": 0.1002, "step": 3410 }, { "epoch": 6.2, "learning_rate": 3.7401915340728467e-06, "loss": 0.1012, "step": 3420 }, { "epoch": 6.21, "learning_rate": 3.7099351654075777e-06, "loss": 0.0918, "step": 3430 }, { "epoch": 6.23, "learning_rate": 3.6797293212895947e-06, "loss": 0.0987, "step": 3440 }, { "epoch": 6.25, "learning_rate": 3.649575184711069e-06, "loss": 0.0862, "step": 3450 }, { "epoch": 6.27, "learning_rate": 3.6194739366390806e-06, "loss": 0.0944, "step": 3460 }, { "epoch": 6.29, "learning_rate": 3.5894267559693628e-06, "loss": 0.1108, "step": 3470 }, { "epoch": 6.3, "learning_rate": 3.559434819480137e-06, "loss": 0.0942, "step": 3480 }, { "epoch": 6.32, "learning_rate": 3.5294993017860226e-06, "loss": 0.0873, "step": 3490 }, { "epoch": 6.34, "learning_rate": 3.499621375292031e-06, "loss": 0.0988, "step": 3500 }, { "epoch": 6.36, "learning_rate": 3.4698022101476548e-06, "loss": 0.0908, "step": 3510 }, { "epoch": 6.38, "learning_rate": 3.440042974201034e-06, "loss": 0.0977, "step": 3520 }, { "epoch": 6.39, "learning_rate": 3.410344832953224e-06, "loss": 0.0942, "step": 3530 }, { "epoch": 6.41, "learning_rate": 3.380708949512543e-06, "loss": 0.1131, "step": 3540 }, { "epoch": 6.43, "learning_rate": 3.3511364845490237e-06, "loss": 0.0919, "step": 3550 }, { "epoch": 6.45, "learning_rate": 3.321628596248958e-06, "loss": 0.0998, "step": 3560 }, { "epoch": 6.47, "learning_rate": 3.29218644026953e-06, "loss": 0.1176, "step": 3570 }, { "epoch": 6.48, "learning_rate": 3.2628111696935655e-06, "loss": 0.1156, "step": 3580 }, { "epoch": 6.5, "learning_rate": 3.2335039349843646e-06, "loss": 0.0825, "step": 3590 }, { "epoch": 6.52, "learning_rate": 3.2042658839406453e-06, "loss": 0.0967, "step": 3600 }, { "epoch": 6.54, "learning_rate": 3.1750981616515965e-06, "loss": 0.1033, "step": 3610 }, { "epoch": 6.56, "learning_rate": 3.1460019104520246e-06, "loss": 0.1165, "step": 3620 }, { "epoch": 6.58, "learning_rate": 3.1169782698776195e-06, "loss": 0.0958, "step": 3630 }, { "epoch": 6.59, "learning_rate": 3.0880283766203203e-06, "loss": 0.1033, "step": 3640 }, { "epoch": 6.61, "learning_rate": 3.0591533644838024e-06, "loss": 0.0975, "step": 3650 }, { "epoch": 6.63, "learning_rate": 3.0303543643390722e-06, "loss": 0.0931, "step": 3660 }, { "epoch": 6.65, "learning_rate": 3.001632504080173e-06, "loss": 0.1036, "step": 3670 }, { "epoch": 6.67, "learning_rate": 2.9729889085800174e-06, "loss": 0.1015, "step": 3680 }, { "epoch": 6.68, "learning_rate": 2.944424699646329e-06, "loss": 0.0972, "step": 3690 }, { "epoch": 6.7, "learning_rate": 2.9159409959777086e-06, "loss": 0.088, "step": 3700 }, { "epoch": 6.72, "learning_rate": 2.8875389131198174e-06, "loss": 0.1096, "step": 3710 }, { "epoch": 6.74, "learning_rate": 2.859219563421699e-06, "loss": 0.0976, "step": 3720 }, { "epoch": 6.76, "learning_rate": 2.830984055992201e-06, "loss": 0.112, "step": 3730 }, { "epoch": 6.77, "learning_rate": 2.8028334966565395e-06, "loss": 0.094, "step": 3740 }, { "epoch": 6.79, "learning_rate": 2.7747689879130024e-06, "loss": 0.1054, "step": 3750 }, { "epoch": 6.81, "learning_rate": 2.7467916288897567e-06, "loss": 0.1094, "step": 3760 }, { "epoch": 6.83, "learning_rate": 2.7189025153018095e-06, "loss": 0.1053, "step": 3770 }, { "epoch": 6.85, "learning_rate": 2.6911027394080923e-06, "loss": 0.0901, "step": 3780 }, { "epoch": 6.86, "learning_rate": 2.6633933899686858e-06, "loss": 0.0987, "step": 3790 }, { "epoch": 6.88, "learning_rate": 2.6357755522021767e-06, "loss": 0.1008, "step": 3800 }, { "epoch": 6.9, "learning_rate": 2.608250307743153e-06, "loss": 0.079, "step": 3810 }, { "epoch": 6.92, "learning_rate": 2.5808187345998513e-06, "loss": 0.0898, "step": 3820 }, { "epoch": 6.94, "learning_rate": 2.5534819071119315e-06, "loss": 0.0951, "step": 3830 }, { "epoch": 6.96, "learning_rate": 2.5262408959084e-06, "loss": 0.0952, "step": 3840 }, { "epoch": 6.97, "learning_rate": 2.499096767865682e-06, "loss": 0.0866, "step": 3850 }, { "epoch": 6.99, "learning_rate": 2.472050586065836e-06, "loss": 0.0916, "step": 3860 }, { "epoch": 7.0, "eval_loss": 0.46777281165122986, "eval_runtime": 245.4593, "eval_samples_per_second": 8.005, "eval_steps_per_second": 4.005, "eval_wer": 0.15547409278541063, "step": 3864 }, { "epoch": 7.01, "learning_rate": 2.4451034097549237e-06, "loss": 0.1281, "step": 3870 }, { "epoch": 7.03, "learning_rate": 2.418256294301513e-06, "loss": 0.0963, "step": 3880 }, { "epoch": 7.05, "learning_rate": 2.3915102911553616e-06, "loss": 0.1001, "step": 3890 }, { "epoch": 7.07, "learning_rate": 2.364866447806227e-06, "loss": 0.0893, "step": 3900 }, { "epoch": 7.08, "learning_rate": 2.338325807742845e-06, "loss": 0.0929, "step": 3910 }, { "epoch": 7.1, "learning_rate": 2.3118894104120614e-06, "loss": 0.0893, "step": 3920 }, { "epoch": 7.12, "learning_rate": 2.285558291178126e-06, "loss": 0.0957, "step": 3930 }, { "epoch": 7.14, "learning_rate": 2.2593334812821356e-06, "loss": 0.095, "step": 3940 }, { "epoch": 7.16, "learning_rate": 2.2332160078016546e-06, "loss": 0.0996, "step": 3950 }, { "epoch": 7.17, "learning_rate": 2.207206893610487e-06, "loss": 0.1002, "step": 3960 }, { "epoch": 7.19, "learning_rate": 2.181307157338615e-06, "loss": 0.1148, "step": 3970 }, { "epoch": 7.21, "learning_rate": 2.155517813332307e-06, "loss": 0.0909, "step": 3980 }, { "epoch": 7.23, "learning_rate": 2.1298398716143897e-06, "loss": 0.0808, "step": 3990 }, { "epoch": 7.25, "learning_rate": 2.104274337844692e-06, "loss": 0.1056, "step": 4000 }, { "epoch": 7.26, "learning_rate": 2.0788222132806566e-06, "loss": 0.0892, "step": 4010 }, { "epoch": 7.28, "learning_rate": 2.0534844947381322e-06, "loss": 0.1216, "step": 4020 }, { "epoch": 7.3, "learning_rate": 2.0282621745523285e-06, "loss": 0.0946, "step": 4030 }, { "epoch": 7.32, "learning_rate": 2.003156240538954e-06, "loss": 0.089, "step": 4040 }, { "epoch": 7.34, "learning_rate": 1.9781676759555295e-06, "loss": 0.0973, "step": 4050 }, { "epoch": 7.35, "learning_rate": 1.9532974594628783e-06, "loss": 0.0887, "step": 4060 }, { "epoch": 7.37, "learning_rate": 1.928546565086799e-06, "loss": 0.114, "step": 4070 }, { "epoch": 7.39, "learning_rate": 1.903915962179913e-06, "loss": 0.0899, "step": 4080 }, { "epoch": 7.41, "learning_rate": 1.8794066153837092e-06, "loss": 0.1057, "step": 4090 }, { "epoch": 7.43, "learning_rate": 1.85501948459076e-06, "loss": 0.1055, "step": 4100 }, { "epoch": 7.45, "learning_rate": 1.8307555249071252e-06, "loss": 0.0903, "step": 4110 }, { "epoch": 7.46, "learning_rate": 1.806615686614951e-06, "loss": 0.1048, "step": 4120 }, { "epoch": 7.48, "learning_rate": 1.7826009151352486e-06, "loss": 0.0992, "step": 4130 }, { "epoch": 7.5, "learning_rate": 1.7587121509908723e-06, "loss": 0.1014, "step": 4140 }, { "epoch": 7.52, "learning_rate": 1.7349503297696745e-06, "loss": 0.091, "step": 4150 }, { "epoch": 7.54, "learning_rate": 1.7113163820878775e-06, "loss": 0.1005, "step": 4160 }, { "epoch": 7.55, "learning_rate": 1.6878112335536173e-06, "loss": 0.1074, "step": 4170 }, { "epoch": 7.57, "learning_rate": 1.6644358047306957e-06, "loss": 0.0981, "step": 4180 }, { "epoch": 7.59, "learning_rate": 1.6411910111025265e-06, "loss": 0.0943, "step": 4190 }, { "epoch": 7.61, "learning_rate": 1.6180777630362832e-06, "loss": 0.1027, "step": 4200 }, { "epoch": 7.63, "learning_rate": 1.5950969657472375e-06, "loss": 0.0869, "step": 4210 }, { "epoch": 7.64, "learning_rate": 1.5722495192633198e-06, "loss": 0.117, "step": 4220 }, { "epoch": 7.66, "learning_rate": 1.5495363183898594e-06, "loss": 0.1003, "step": 4230 }, { "epoch": 7.68, "learning_rate": 1.5269582526745448e-06, "loss": 0.0911, "step": 4240 }, { "epoch": 7.7, "learning_rate": 1.5045162063725838e-06, "loss": 0.0874, "step": 4250 }, { "epoch": 7.72, "learning_rate": 1.4822110584120735e-06, "loss": 0.1026, "step": 4260 }, { "epoch": 7.73, "learning_rate": 1.4600436823595787e-06, "loss": 0.0936, "step": 4270 }, { "epoch": 7.75, "learning_rate": 1.43801494638591e-06, "loss": 0.0835, "step": 4280 }, { "epoch": 7.77, "learning_rate": 1.4161257132321377e-06, "loss": 0.0989, "step": 4290 }, { "epoch": 7.79, "learning_rate": 1.3943768401757917e-06, "loss": 0.0931, "step": 4300 }, { "epoch": 7.81, "learning_rate": 1.3727691789972913e-06, "loss": 0.0815, "step": 4310 }, { "epoch": 7.83, "learning_rate": 1.3513035759465841e-06, "loss": 0.1001, "step": 4320 }, { "epoch": 7.84, "learning_rate": 1.3299808717100037e-06, "loss": 0.0907, "step": 4330 }, { "epoch": 7.86, "learning_rate": 1.3088019013773456e-06, "loss": 0.0899, "step": 4340 }, { "epoch": 7.88, "learning_rate": 1.2877674944091605e-06, "loss": 0.0953, "step": 4350 }, { "epoch": 7.9, "learning_rate": 1.266878474604269e-06, "loss": 0.081, "step": 4360 }, { "epoch": 7.92, "learning_rate": 1.2461356600674984e-06, "loss": 0.1002, "step": 4370 }, { "epoch": 7.93, "learning_rate": 1.2255398631776427e-06, "loss": 0.088, "step": 4380 }, { "epoch": 7.95, "learning_rate": 1.2050918905556463e-06, "loss": 0.1001, "step": 4390 }, { "epoch": 7.97, "learning_rate": 1.1847925430330093e-06, "loss": 0.1004, "step": 4400 }, { "epoch": 7.99, "learning_rate": 1.16464261562043e-06, "loss": 0.1, "step": 4410 }, { "epoch": 8.0, "eval_loss": 0.47049203515052795, "eval_runtime": 243.653, "eval_samples_per_second": 8.065, "eval_steps_per_second": 4.034, "eval_wer": 0.15495040354876471, "step": 4416 }, { "epoch": 8.01, "learning_rate": 1.1446428974766643e-06, "loss": 0.1086, "step": 4420 }, { "epoch": 8.03, "learning_rate": 1.1247941718776212e-06, "loss": 0.0893, "step": 4430 }, { "epoch": 8.04, "learning_rate": 1.1050972161856843e-06, "loss": 0.1016, "step": 4440 }, { "epoch": 8.06, "learning_rate": 1.0855528018192679e-06, "loss": 0.1021, "step": 4450 }, { "epoch": 8.08, "learning_rate": 1.066161694222606e-06, "loss": 0.1029, "step": 4460 }, { "epoch": 8.1, "learning_rate": 1.0469246528357707e-06, "loss": 0.1003, "step": 4470 }, { "epoch": 8.12, "learning_rate": 1.0278424310649327e-06, "loss": 0.0996, "step": 4480 }, { "epoch": 8.13, "learning_rate": 1.008915776252854e-06, "loss": 0.1053, "step": 4490 }, { "epoch": 8.15, "learning_rate": 9.901454296496182e-07, "loss": 0.0977, "step": 4500 }, { "epoch": 8.17, "learning_rate": 9.715321263835996e-07, "loss": 0.1032, "step": 4510 }, { "epoch": 8.19, "learning_rate": 9.530765954326726e-07, "loss": 0.1114, "step": 4520 }, { "epoch": 8.21, "learning_rate": 9.347795595956632e-07, "loss": 0.0806, "step": 4530 }, { "epoch": 8.22, "learning_rate": 9.16641735464035e-07, "loss": 0.0995, "step": 4540 }, { "epoch": 8.24, "learning_rate": 8.986638333938352e-07, "loss": 0.1056, "step": 4550 }, { "epoch": 8.26, "learning_rate": 8.808465574778635e-07, "loss": 0.0918, "step": 4560 }, { "epoch": 8.28, "learning_rate": 8.631906055181044e-07, "loss": 0.0987, "step": 4570 }, { "epoch": 8.3, "learning_rate": 8.45696668998392e-07, "loss": 0.0999, "step": 4580 }, { "epoch": 8.31, "learning_rate": 8.283654330573326e-07, "loss": 0.0911, "step": 4590 }, { "epoch": 8.33, "learning_rate": 8.111975764614732e-07, "loss": 0.1001, "step": 4600 }, { "epoch": 8.35, "learning_rate": 7.94193771578709e-07, "loss": 0.1019, "step": 4610 }, { "epoch": 8.37, "learning_rate": 7.773546843519619e-07, "loss": 0.0912, "step": 4620 }, { "epoch": 8.39, "learning_rate": 7.606809742730942e-07, "loss": 0.095, "step": 4630 }, { "epoch": 8.41, "learning_rate": 7.441732943570784e-07, "loss": 0.0907, "step": 4640 }, { "epoch": 8.42, "learning_rate": 7.278322911164243e-07, "loss": 0.0883, "step": 4650 }, { "epoch": 8.44, "learning_rate": 7.116586045358609e-07, "loss": 0.1234, "step": 4660 }, { "epoch": 8.46, "learning_rate": 6.956528680472641e-07, "loss": 0.1065, "step": 4670 }, { "epoch": 8.48, "learning_rate": 6.79815708504859e-07, "loss": 0.098, "step": 4680 }, { "epoch": 8.5, "learning_rate": 6.641477461606632e-07, "loss": 0.1049, "step": 4690 }, { "epoch": 8.51, "learning_rate": 6.48649594640196e-07, "loss": 0.0924, "step": 4700 }, { "epoch": 8.53, "learning_rate": 6.333218609184466e-07, "loss": 0.1016, "step": 4710 }, { "epoch": 8.55, "learning_rate": 6.181651452961023e-07, "loss": 0.1077, "step": 4720 }, { "epoch": 8.57, "learning_rate": 6.031800413760408e-07, "loss": 0.0862, "step": 4730 }, { "epoch": 8.59, "learning_rate": 5.883671360400739e-07, "loss": 0.0892, "step": 4740 }, { "epoch": 8.6, "learning_rate": 5.737270094259723e-07, "loss": 0.081, "step": 4750 }, { "epoch": 8.62, "learning_rate": 5.592602349047405e-07, "loss": 0.0864, "step": 4760 }, { "epoch": 8.64, "learning_rate": 5.449673790581611e-07, "loss": 0.1028, "step": 4770 }, { "epoch": 8.66, "learning_rate": 5.308490016566042e-07, "loss": 0.0988, "step": 4780 }, { "epoch": 8.68, "learning_rate": 5.169056556371077e-07, "loss": 0.114, "step": 4790 }, { "epoch": 8.69, "learning_rate": 5.031378870817182e-07, "loss": 0.107, "step": 4800 }, { "epoch": 8.71, "learning_rate": 4.895462351961039e-07, "loss": 0.0935, "step": 4810 }, { "epoch": 8.73, "learning_rate": 4.761312322884415e-07, "loss": 0.0963, "step": 4820 }, { "epoch": 8.75, "learning_rate": 4.628934037485644e-07, "loss": 0.0836, "step": 4830 }, { "epoch": 8.77, "learning_rate": 4.498332680273881e-07, "loss": 0.1062, "step": 4840 }, { "epoch": 8.79, "learning_rate": 4.369513366166039e-07, "loss": 0.0952, "step": 4850 }, { "epoch": 8.8, "learning_rate": 4.242481140286486e-07, "loss": 0.0869, "step": 4860 }, { "epoch": 8.82, "learning_rate": 4.1172409777694433e-07, "loss": 0.0906, "step": 4870 }, { "epoch": 8.84, "learning_rate": 3.9937977835641294e-07, "loss": 0.087, "step": 4880 }, { "epoch": 8.86, "learning_rate": 3.872156392242682e-07, "loss": 0.0754, "step": 4890 }, { "epoch": 8.88, "learning_rate": 3.7523215678108015e-07, "loss": 0.0993, "step": 4900 }, { "epoch": 8.89, "learning_rate": 3.634298003521186e-07, "loss": 0.0977, "step": 4910 }, { "epoch": 8.91, "learning_rate": 3.518090321689693e-07, "loss": 0.1168, "step": 4920 }, { "epoch": 8.93, "learning_rate": 3.4037030735143474e-07, "loss": 0.0901, "step": 4930 }, { "epoch": 8.95, "learning_rate": 3.291140738897064e-07, "loss": 0.1056, "step": 4940 }, { "epoch": 8.97, "learning_rate": 3.1804077262682085e-07, "loss": 0.0966, "step": 4950 }, { "epoch": 8.98, "learning_rate": 3.0715083724139503e-07, "loss": 0.1301, "step": 4960 }, { "epoch": 9.0, "eval_loss": 0.47403913736343384, "eval_runtime": 244.7056, "eval_samples_per_second": 8.03, "eval_steps_per_second": 4.017, "eval_wer": 0.1551352350440515, "step": 4968 }, { "epoch": 9.0, "learning_rate": 2.964446942306426e-07, "loss": 0.108, "step": 4970 }, { "epoch": 9.02, "learning_rate": 2.8592276289366563e-07, "loss": 0.0953, "step": 4980 }, { "epoch": 9.04, "learning_rate": 2.755854553150394e-07, "loss": 0.101, "step": 4990 }, { "epoch": 9.06, "learning_rate": 2.654331763486673e-07, "loss": 0.0919, "step": 5000 }, { "epoch": 9.08, "learning_rate": 2.5546632360192893e-07, "loss": 0.094, "step": 5010 }, { "epoch": 9.09, "learning_rate": 2.456852874201082e-07, "loss": 0.1065, "step": 5020 }, { "epoch": 9.11, "learning_rate": 2.3609045087110317e-07, "loss": 0.0769, "step": 5030 }, { "epoch": 9.13, "learning_rate": 2.26682189730425e-07, "loss": 0.097, "step": 5040 }, { "epoch": 9.15, "learning_rate": 2.1746087246648172e-07, "loss": 0.0975, "step": 5050 }, { "epoch": 9.17, "learning_rate": 2.0842686022614412e-07, "loss": 0.0946, "step": 5060 }, { "epoch": 9.18, "learning_rate": 1.9958050682060692e-07, "loss": 0.0989, "step": 5070 }, { "epoch": 9.2, "learning_rate": 1.9092215871152786e-07, "loss": 0.0874, "step": 5080 }, { "epoch": 9.22, "learning_rate": 1.8245215499745838e-07, "loss": 0.1016, "step": 5090 }, { "epoch": 9.24, "learning_rate": 1.7417082740056713e-07, "loss": 0.1044, "step": 5100 }, { "epoch": 9.26, "learning_rate": 1.660785002536447e-07, "loss": 0.1017, "step": 5110 }, { "epoch": 9.28, "learning_rate": 1.5817549048740166e-07, "loss": 0.0955, "step": 5120 }, { "epoch": 9.29, "learning_rate": 1.504621076180579e-07, "loss": 0.0875, "step": 5130 }, { "epoch": 9.31, "learning_rate": 1.4293865373521953e-07, "loss": 0.1041, "step": 5140 }, { "epoch": 9.33, "learning_rate": 1.3560542349004723e-07, "loss": 0.0988, "step": 5150 }, { "epoch": 9.35, "learning_rate": 1.2846270408371832e-07, "loss": 0.0909, "step": 5160 }, { "epoch": 9.37, "learning_rate": 1.215107752561756e-07, "loss": 0.1083, "step": 5170 }, { "epoch": 9.38, "learning_rate": 1.1474990927517527e-07, "loss": 0.1009, "step": 5180 }, { "epoch": 9.4, "learning_rate": 1.0818037092562128e-07, "loss": 0.1099, "step": 5190 }, { "epoch": 9.42, "learning_rate": 1.018024174991944e-07, "loss": 0.0862, "step": 5200 }, { "epoch": 9.44, "learning_rate": 9.561629878427847e-08, "loss": 0.0782, "step": 5210 }, { "epoch": 9.46, "learning_rate": 8.962225705617556e-08, "loss": 0.1061, "step": 5220 }, { "epoch": 9.47, "learning_rate": 8.38205270676179e-08, "loss": 0.0864, "step": 5230 }, { "epoch": 9.49, "learning_rate": 7.82113360395742e-08, "loss": 0.1151, "step": 5240 }, { "epoch": 9.51, "learning_rate": 7.279490365235065e-08, "loss": 0.0906, "step": 5250 }, { "epoch": 9.53, "learning_rate": 6.757144203698607e-08, "loss": 0.0912, "step": 5260 }, { "epoch": 9.55, "learning_rate": 6.25411557669453e-08, "loss": 0.105, "step": 5270 }, { "epoch": 9.56, "learning_rate": 5.770424185010726e-08, "loss": 0.1117, "step": 5280 }, { "epoch": 9.58, "learning_rate": 5.306088972104939e-08, "loss": 0.0891, "step": 5290 }, { "epoch": 9.6, "learning_rate": 4.861128123362646e-08, "loss": 0.0901, "step": 5300 }, { "epoch": 9.62, "learning_rate": 4.43555906538512e-08, "loss": 0.0931, "step": 5310 }, { "epoch": 9.64, "learning_rate": 4.0293984653068105e-08, "loss": 0.1138, "step": 5320 }, { "epoch": 9.65, "learning_rate": 3.64266223014248e-08, "loss": 0.1282, "step": 5330 }, { "epoch": 9.67, "learning_rate": 3.275365506164363e-08, "loss": 0.0919, "step": 5340 }, { "epoch": 9.69, "learning_rate": 2.9275226783089806e-08, "loss": 0.0988, "step": 5350 }, { "epoch": 9.71, "learning_rate": 2.5991473696136416e-08, "loss": 0.11, "step": 5360 }, { "epoch": 9.73, "learning_rate": 2.2902524406830385e-08, "loss": 0.0971, "step": 5370 }, { "epoch": 9.75, "learning_rate": 2.0008499891853715e-08, "loss": 0.1008, "step": 5380 }, { "epoch": 9.76, "learning_rate": 1.7309513493788955e-08, "loss": 0.0955, "step": 5390 }, { "epoch": 9.78, "learning_rate": 1.4805670916676063e-08, "loss": 0.0855, "step": 5400 }, { "epoch": 9.8, "learning_rate": 1.249707022187574e-08, "loss": 0.1014, "step": 5410 }, { "epoch": 9.82, "learning_rate": 1.0586336291851795e-08, "loss": 0.1153, "step": 5420 }, { "epoch": 9.84, "learning_rate": 8.64893794362054e-09, "loss": 0.0956, "step": 5430 }, { "epoch": 9.85, "learning_rate": 6.907022602130054e-09, "loss": 0.0805, "step": 5440 }, { "epoch": 9.87, "learning_rate": 5.36065848835865e-09, "loss": 0.0883, "step": 5450 }, { "epoch": 9.89, "learning_rate": 4.009906164646471e-09, "loss": 0.0965, "step": 5460 }, { "epoch": 9.91, "learning_rate": 2.8548185323279407e-09, "loss": 0.1014, "step": 5470 }, { "epoch": 9.93, "learning_rate": 1.895440829656203e-09, "loss": 0.0836, "step": 5480 }, { "epoch": 9.94, "learning_rate": 1.13181063003176e-09, "loss": 0.0809, "step": 5490 }, { "epoch": 9.96, "learning_rate": 5.639578405342017e-10, "loss": 0.0959, "step": 5500 }, { "epoch": 9.98, "learning_rate": 1.9190470074703648e-10, "loss": 0.1045, "step": 5510 }, { "epoch": 10.0, "learning_rate": 1.56657818878303e-11, "loss": 0.0885, "step": 5520 }, { "epoch": 10.0, "eval_loss": 0.47021371126174927, "eval_runtime": 245.8434, "eval_samples_per_second": 7.993, "eval_steps_per_second": 3.998, "eval_wer": 0.15458074055819113, "step": 5520 }, { "epoch": 10.0, "step": 5520, "total_flos": 5.359318129634805e+19, "train_loss": 0.11096900397020837, "train_runtime": 38440.8489, "train_samples_per_second": 4.601, "train_steps_per_second": 0.144 } ], "max_steps": 5520, "num_train_epochs": 10, "total_flos": 5.359318129634805e+19, "trial_name": null, "trial_params": null }