{ "best_metric": null, "best_model_checkpoint": null, "epoch": 1.8221674201945472, "eval_steps": 100, "global_step": 13300, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.0, "learning_rate": 5.399999999999999e-06, "loss": 11.7783, "step": 10 }, { "epoch": 0.01, "learning_rate": 1.0799999999999998e-05, "loss": 12.9293, "step": 20 }, { "epoch": 0.01, "learning_rate": 1.68e-05, "loss": 13.9825, "step": 30 }, { "epoch": 0.01, "learning_rate": 2.2199999999999998e-05, "loss": 16.1366, "step": 40 }, { "epoch": 0.01, "learning_rate": 2.7599999999999997e-05, "loss": 34.2401, "step": 50 }, { "epoch": 0.02, "learning_rate": 3.36e-05, "loss": 10.2244, "step": 60 }, { "epoch": 0.02, "learning_rate": 3.96e-05, "loss": 8.0319, "step": 70 }, { "epoch": 0.02, "learning_rate": 4.56e-05, "loss": 6.303, "step": 80 }, { "epoch": 0.02, "learning_rate": 5.1599999999999994e-05, "loss": 5.358, "step": 90 }, { "epoch": 0.03, "learning_rate": 5.76e-05, "loss": 7.887, "step": 100 }, { "epoch": 0.03, "eval_loss": 3.6606574058532715, "eval_runtime": 1231.7206, "eval_samples_per_second": 9.485, "eval_steps_per_second": 1.186, "eval_wer": 1.0, "step": 100 }, { "epoch": 0.03, "learning_rate": 6.359999999999999e-05, "loss": 3.6845, "step": 110 }, { "epoch": 0.03, "learning_rate": 6.96e-05, "loss": 3.4054, "step": 120 }, { "epoch": 0.04, "learning_rate": 7.56e-05, "loss": 3.4159, "step": 130 }, { "epoch": 0.04, "learning_rate": 8.16e-05, "loss": 3.3657, "step": 140 }, { "epoch": 0.04, "learning_rate": 8.759999999999999e-05, "loss": 4.0007, "step": 150 }, { "epoch": 0.04, "learning_rate": 9.36e-05, "loss": 3.3271, "step": 160 }, { "epoch": 0.05, "learning_rate": 9.96e-05, "loss": 3.185, "step": 170 }, { "epoch": 0.05, "learning_rate": 0.00010559999999999998, "loss": 3.1758, "step": 180 }, { "epoch": 0.05, "learning_rate": 0.00011159999999999999, "loss": 3.124, "step": 190 }, { "epoch": 0.05, "learning_rate": 0.0001176, "loss": 4.0646, "step": 200 }, { "epoch": 0.05, "eval_loss": 3.173530101776123, "eval_runtime": 1202.0614, "eval_samples_per_second": 9.719, "eval_steps_per_second": 1.215, "eval_wer": 1.0, "step": 200 }, { "epoch": 0.06, "learning_rate": 0.0001236, "loss": 4.6529, "step": 210 }, { "epoch": 0.06, "learning_rate": 0.000129, "loss": 3.8886, "step": 220 }, { "epoch": 0.06, "learning_rate": 0.000135, "loss": 3.6103, "step": 230 }, { "epoch": 0.07, "learning_rate": 0.00014099999999999998, "loss": 3.2837, "step": 240 }, { "epoch": 0.07, "learning_rate": 0.000147, "loss": 3.7727, "step": 250 }, { "epoch": 0.07, "learning_rate": 0.00015299999999999998, "loss": 3.7147, "step": 260 }, { "epoch": 0.07, "learning_rate": 0.000159, "loss": 3.3716, "step": 270 }, { "epoch": 0.08, "learning_rate": 0.000165, "loss": 3.135, "step": 280 }, { "epoch": 0.08, "learning_rate": 0.00017099999999999998, "loss": 3.1926, "step": 290 }, { "epoch": 0.08, "learning_rate": 0.00017699999999999997, "loss": 3.7509, "step": 300 }, { "epoch": 0.08, "eval_loss": 3.2335991859436035, "eval_runtime": 1163.562, "eval_samples_per_second": 10.041, "eval_steps_per_second": 1.256, "eval_wer": 1.0, "step": 300 }, { "epoch": 0.08, "learning_rate": 0.00018299999999999998, "loss": 3.3192, "step": 310 }, { "epoch": 0.09, "learning_rate": 0.00018899999999999999, "loss": 3.0958, "step": 320 }, { "epoch": 0.09, "learning_rate": 0.000195, "loss": 3.0755, "step": 330 }, { "epoch": 0.09, "learning_rate": 0.000201, "loss": 3.1681, "step": 340 }, { "epoch": 0.1, "learning_rate": 0.00020699999999999996, "loss": 3.7908, "step": 350 }, { "epoch": 0.1, "learning_rate": 0.00021299999999999997, "loss": 3.4041, "step": 360 }, { "epoch": 0.1, "learning_rate": 0.00021899999999999998, "loss": 3.1383, "step": 370 }, { "epoch": 0.1, "learning_rate": 0.000225, "loss": 3.1267, "step": 380 }, { "epoch": 0.11, "learning_rate": 0.00023099999999999998, "loss": 3.2255, "step": 390 }, { "epoch": 0.11, "learning_rate": 0.0002364, "loss": 3.8832, "step": 400 }, { "epoch": 0.11, "eval_loss": 3.147369623184204, "eval_runtime": 1163.4758, "eval_samples_per_second": 10.041, "eval_steps_per_second": 1.256, "eval_wer": 1.0, "step": 400 }, { "epoch": 0.11, "learning_rate": 5.999999999999999e-06, "loss": 4.262, "step": 410 }, { "epoch": 0.12, "learning_rate": 1.1999999999999999e-05, "loss": 4.2338, "step": 420 }, { "epoch": 0.12, "learning_rate": 1.7999999999999997e-05, "loss": 4.1712, "step": 430 }, { "epoch": 0.12, "learning_rate": 2.28e-05, "loss": 4.4312, "step": 440 }, { "epoch": 0.12, "learning_rate": 2.88e-05, "loss": 4.9701, "step": 450 }, { "epoch": 0.13, "learning_rate": 3.48e-05, "loss": 3.8904, "step": 460 }, { "epoch": 0.13, "learning_rate": 4.08e-05, "loss": 3.7617, "step": 470 }, { "epoch": 0.13, "learning_rate": 4.68e-05, "loss": 3.6052, "step": 480 }, { "epoch": 0.13, "learning_rate": 5.279999999999999e-05, "loss": 3.5166, "step": 490 }, { "epoch": 0.14, "learning_rate": 5.88e-05, "loss": 3.9646, "step": 500 }, { "epoch": 0.14, "eval_loss": 3.334455728530884, "eval_runtime": 1100.3955, "eval_samples_per_second": 10.617, "eval_steps_per_second": 1.328, "eval_wer": 1.0, "step": 500 }, { "epoch": 0.14, "learning_rate": 6.479999999999999e-05, "loss": 3.234, "step": 510 }, { "epoch": 0.14, "learning_rate": 7.079999999999999e-05, "loss": 3.1796, "step": 520 }, { "epoch": 0.15, "learning_rate": 7.68e-05, "loss": 3.1659, "step": 530 }, { "epoch": 0.15, "learning_rate": 8.28e-05, "loss": 3.1283, "step": 540 }, { "epoch": 0.15, "learning_rate": 8.879999999999999e-05, "loss": 3.2447, "step": 550 }, { "epoch": 0.15, "learning_rate": 9.479999999999999e-05, "loss": 3.0545, "step": 560 }, { "epoch": 0.16, "learning_rate": 0.0001008, "loss": 3.08, "step": 570 }, { "epoch": 0.16, "learning_rate": 0.00010679999999999998, "loss": 3.0577, "step": 580 }, { "epoch": 0.16, "learning_rate": 0.00011279999999999999, "loss": 3.0595, "step": 590 }, { "epoch": 0.16, "learning_rate": 0.0001188, "loss": 3.1657, "step": 600 }, { "epoch": 0.16, "eval_loss": 3.055232048034668, "eval_runtime": 1107.3435, "eval_samples_per_second": 10.55, "eval_steps_per_second": 1.319, "eval_wer": 1.0, "step": 600 }, { "epoch": 0.17, "learning_rate": 0.00012479999999999997, "loss": 3.0392, "step": 610 }, { "epoch": 0.17, "learning_rate": 0.00013079999999999998, "loss": 3.03, "step": 620 }, { "epoch": 0.17, "learning_rate": 0.0001368, "loss": 3.0355, "step": 630 }, { "epoch": 0.18, "learning_rate": 0.00014279999999999997, "loss": 3.104, "step": 640 }, { "epoch": 0.18, "learning_rate": 0.00014879999999999998, "loss": 3.3866, "step": 650 }, { "epoch": 0.18, "learning_rate": 0.0001548, "loss": 3.043, "step": 660 }, { "epoch": 0.18, "learning_rate": 0.0001608, "loss": 3.0396, "step": 670 }, { "epoch": 0.19, "learning_rate": 0.0001668, "loss": 3.0713, "step": 680 }, { "epoch": 0.19, "learning_rate": 0.00017279999999999997, "loss": 3.0453, "step": 690 }, { "epoch": 0.19, "learning_rate": 0.00017879999999999998, "loss": 3.0641, "step": 700 }, { "epoch": 0.19, "eval_loss": 3.091907024383545, "eval_runtime": 1074.3754, "eval_samples_per_second": 10.874, "eval_steps_per_second": 1.36, "eval_wer": 1.0, "step": 700 }, { "epoch": 0.19, "learning_rate": 0.0001848, "loss": 3.045, "step": 710 }, { "epoch": 0.2, "learning_rate": 0.00019079999999999998, "loss": 3.0312, "step": 720 }, { "epoch": 0.2, "learning_rate": 0.00019679999999999999, "loss": 3.0153, "step": 730 }, { "epoch": 0.2, "learning_rate": 0.0002028, "loss": 2.9866, "step": 740 }, { "epoch": 0.21, "learning_rate": 0.00020879999999999998, "loss": 2.9827, "step": 750 }, { "epoch": 0.21, "learning_rate": 0.00021479999999999996, "loss": 2.9405, "step": 760 }, { "epoch": 0.21, "learning_rate": 0.00022079999999999997, "loss": 2.7764, "step": 770 }, { "epoch": 0.21, "learning_rate": 0.00022679999999999998, "loss": 2.5073, "step": 780 }, { "epoch": 0.22, "learning_rate": 0.0002328, "loss": 2.1963, "step": 790 }, { "epoch": 0.22, "learning_rate": 0.0002388, "loss": 1.9443, "step": 800 }, { "epoch": 0.22, "eval_loss": 1.5123839378356934, "eval_runtime": 1077.0854, "eval_samples_per_second": 10.847, "eval_steps_per_second": 1.356, "eval_wer": 1.0, "step": 800 }, { "epoch": 0.22, "learning_rate": 0.0002448, "loss": 1.5669, "step": 810 }, { "epoch": 0.22, "learning_rate": 0.00025079999999999997, "loss": 1.2878, "step": 820 }, { "epoch": 0.23, "learning_rate": 0.00025679999999999995, "loss": 1.1475, "step": 830 }, { "epoch": 0.23, "learning_rate": 0.0002628, "loss": 1.0235, "step": 840 }, { "epoch": 0.23, "learning_rate": 0.0002688, "loss": 1.0418, "step": 850 }, { "epoch": 0.24, "learning_rate": 0.0002748, "loss": 0.916, "step": 860 }, { "epoch": 0.24, "learning_rate": 0.0002808, "loss": 0.8304, "step": 870 }, { "epoch": 0.24, "learning_rate": 0.0002868, "loss": 0.6735, "step": 880 }, { "epoch": 0.24, "learning_rate": 0.00029279999999999996, "loss": 0.6997, "step": 890 }, { "epoch": 0.25, "learning_rate": 0.0002988, "loss": 0.6772, "step": 900 }, { "epoch": 0.25, "eval_loss": 0.6391938328742981, "eval_runtime": 1088.2, "eval_samples_per_second": 10.736, "eval_steps_per_second": 1.343, "eval_wer": 0.8844946784317292, "step": 900 }, { "epoch": 0.25, "learning_rate": 0.0002999779755896118, "loss": 0.697, "step": 910 }, { "epoch": 0.25, "learning_rate": 0.00029995044507662654, "loss": 0.5928, "step": 920 }, { "epoch": 0.25, "learning_rate": 0.00029992291456364134, "loss": 0.5868, "step": 930 }, { "epoch": 0.26, "learning_rate": 0.00029989538405065613, "loss": 0.588, "step": 940 }, { "epoch": 0.26, "learning_rate": 0.0002998678535376709, "loss": 0.6109, "step": 950 }, { "epoch": 0.26, "learning_rate": 0.00029984032302468567, "loss": 0.5986, "step": 960 }, { "epoch": 0.27, "learning_rate": 0.00029981279251170047, "loss": 0.5192, "step": 970 }, { "epoch": 0.27, "learning_rate": 0.0002997852619987152, "loss": 0.5146, "step": 980 }, { "epoch": 0.27, "learning_rate": 0.00029975773148573, "loss": 0.5384, "step": 990 }, { "epoch": 0.27, "learning_rate": 0.00029973020097274474, "loss": 0.5144, "step": 1000 }, { "epoch": 0.27, "eval_loss": 0.4253242611885071, "eval_runtime": 1098.9682, "eval_samples_per_second": 10.631, "eval_steps_per_second": 1.329, "eval_wer": 0.7255526243973438, "step": 1000 }, { "epoch": 0.28, "learning_rate": 0.00029970267045975954, "loss": 0.5067, "step": 1010 }, { "epoch": 0.28, "learning_rate": 0.00029967513994677434, "loss": 0.4826, "step": 1020 }, { "epoch": 0.28, "learning_rate": 0.0002996476094337891, "loss": 0.5047, "step": 1030 }, { "epoch": 0.28, "learning_rate": 0.0002996200789208039, "loss": 0.4862, "step": 1040 }, { "epoch": 0.29, "learning_rate": 0.00029959254840781867, "loss": 0.4903, "step": 1050 }, { "epoch": 0.29, "learning_rate": 0.0002995650178948334, "loss": 0.524, "step": 1060 }, { "epoch": 0.29, "learning_rate": 0.0002995374873818482, "loss": 0.4635, "step": 1070 }, { "epoch": 0.3, "learning_rate": 0.00029950995686886295, "loss": 0.4518, "step": 1080 }, { "epoch": 0.3, "learning_rate": 0.00029948242635587774, "loss": 0.4541, "step": 1090 }, { "epoch": 0.3, "learning_rate": 0.00029945489584289254, "loss": 0.4361, "step": 1100 }, { "epoch": 0.3, "eval_loss": 0.3405020236968994, "eval_runtime": 1102.4518, "eval_samples_per_second": 10.597, "eval_steps_per_second": 1.325, "eval_wer": 0.6595788228872919, "step": 1100 }, { "epoch": 0.3, "learning_rate": 0.0002994273653299073, "loss": 0.4521, "step": 1110 }, { "epoch": 0.31, "learning_rate": 0.0002993998348169221, "loss": 0.4303, "step": 1120 }, { "epoch": 0.31, "learning_rate": 0.0002993723043039368, "loss": 0.4547, "step": 1130 }, { "epoch": 0.31, "learning_rate": 0.0002993447737909516, "loss": 0.4971, "step": 1140 }, { "epoch": 0.32, "learning_rate": 0.0002993172432779664, "loss": 0.4177, "step": 1150 }, { "epoch": 0.32, "learning_rate": 0.00029928971276498115, "loss": 0.4208, "step": 1160 }, { "epoch": 0.32, "learning_rate": 0.00029926218225199595, "loss": 0.4125, "step": 1170 }, { "epoch": 0.32, "learning_rate": 0.0002992346517390107, "loss": 0.3863, "step": 1180 }, { "epoch": 0.33, "learning_rate": 0.0002992071212260255, "loss": 0.4037, "step": 1190 }, { "epoch": 0.33, "learning_rate": 0.0002991795907130403, "loss": 0.4239, "step": 1200 }, { "epoch": 0.33, "eval_loss": 0.2857191860675812, "eval_runtime": 1081.4677, "eval_samples_per_second": 10.803, "eval_steps_per_second": 1.351, "eval_wer": 0.5968798326207586, "step": 1200 }, { "epoch": 0.33, "learning_rate": 0.000299152060200055, "loss": 0.3721, "step": 1210 }, { "epoch": 0.33, "learning_rate": 0.0002991245296870698, "loss": 0.3766, "step": 1220 }, { "epoch": 0.34, "learning_rate": 0.00029909699917408456, "loss": 0.4146, "step": 1230 }, { "epoch": 0.34, "learning_rate": 0.00029906946866109936, "loss": 0.4146, "step": 1240 }, { "epoch": 0.34, "learning_rate": 0.00029904193814811415, "loss": 0.3439, "step": 1250 }, { "epoch": 0.35, "learning_rate": 0.0002990144076351289, "loss": 0.3952, "step": 1260 }, { "epoch": 0.35, "learning_rate": 0.0002989868771221437, "loss": 0.3558, "step": 1270 }, { "epoch": 0.35, "learning_rate": 0.00029895934660915843, "loss": 0.3923, "step": 1280 }, { "epoch": 0.35, "learning_rate": 0.0002989318160961732, "loss": 0.3893, "step": 1290 }, { "epoch": 0.36, "learning_rate": 0.000298904285583188, "loss": 0.3226, "step": 1300 }, { "epoch": 0.36, "eval_loss": 0.279046893119812, "eval_runtime": 1089.4065, "eval_samples_per_second": 10.724, "eval_steps_per_second": 1.341, "eval_wer": 0.5880332939143091, "step": 1300 }, { "epoch": 0.36, "learning_rate": 0.00029887675507020276, "loss": 0.4045, "step": 1310 }, { "epoch": 0.36, "learning_rate": 0.00029884922455721756, "loss": 0.3552, "step": 1320 }, { "epoch": 0.36, "learning_rate": 0.00029882169404423235, "loss": 0.3442, "step": 1330 }, { "epoch": 0.37, "learning_rate": 0.0002987941635312471, "loss": 0.3695, "step": 1340 }, { "epoch": 0.37, "learning_rate": 0.0002987666330182619, "loss": 0.3378, "step": 1350 }, { "epoch": 0.37, "learning_rate": 0.00029873910250527663, "loss": 0.3464, "step": 1360 }, { "epoch": 0.38, "learning_rate": 0.00029871157199229143, "loss": 0.3532, "step": 1370 }, { "epoch": 0.38, "learning_rate": 0.0002986840414793062, "loss": 0.3477, "step": 1380 }, { "epoch": 0.38, "learning_rate": 0.00029865651096632097, "loss": 0.3431, "step": 1390 }, { "epoch": 0.38, "learning_rate": 0.00029862898045333576, "loss": 0.3508, "step": 1400 }, { "epoch": 0.38, "eval_loss": 0.24916307628154755, "eval_runtime": 1079.8194, "eval_samples_per_second": 10.819, "eval_steps_per_second": 1.353, "eval_wer": 0.5586168470845083, "step": 1400 }, { "epoch": 0.39, "learning_rate": 0.00029860144994035056, "loss": 0.372, "step": 1410 }, { "epoch": 0.39, "learning_rate": 0.0002985739194273653, "loss": 0.3597, "step": 1420 }, { "epoch": 0.39, "learning_rate": 0.0002985463889143801, "loss": 0.3585, "step": 1430 }, { "epoch": 0.39, "learning_rate": 0.00029851885840139484, "loss": 0.3502, "step": 1440 }, { "epoch": 0.4, "learning_rate": 0.00029849132788840963, "loss": 0.3097, "step": 1450 }, { "epoch": 0.4, "learning_rate": 0.00029846379737542443, "loss": 0.3237, "step": 1460 }, { "epoch": 0.4, "learning_rate": 0.00029843626686243917, "loss": 0.3071, "step": 1470 }, { "epoch": 0.41, "learning_rate": 0.00029840873634945397, "loss": 0.3479, "step": 1480 }, { "epoch": 0.41, "learning_rate": 0.0002983812058364687, "loss": 0.3608, "step": 1490 }, { "epoch": 0.41, "learning_rate": 0.0002983536753234835, "loss": 0.3091, "step": 1500 }, { "epoch": 0.41, "eval_loss": 0.25719356536865234, "eval_runtime": 1086.8697, "eval_samples_per_second": 10.749, "eval_steps_per_second": 1.344, "eval_wer": 0.5655075957427453, "step": 1500 }, { "epoch": 0.41, "learning_rate": 0.0002983261448104983, "loss": 0.3485, "step": 1510 }, { "epoch": 0.42, "learning_rate": 0.00029829861429751304, "loss": 0.3394, "step": 1520 }, { "epoch": 0.42, "learning_rate": 0.00029827108378452784, "loss": 0.3418, "step": 1530 }, { "epoch": 0.42, "learning_rate": 0.0002982435532715426, "loss": 0.3772, "step": 1540 }, { "epoch": 0.42, "learning_rate": 0.0002982160227585574, "loss": 0.3438, "step": 1550 }, { "epoch": 0.43, "learning_rate": 0.00029818849224557217, "loss": 0.3139, "step": 1560 }, { "epoch": 0.43, "learning_rate": 0.0002981609617325869, "loss": 0.3213, "step": 1570 }, { "epoch": 0.43, "learning_rate": 0.0002981334312196017, "loss": 0.3382, "step": 1580 }, { "epoch": 0.44, "learning_rate": 0.00029810590070661645, "loss": 0.3355, "step": 1590 }, { "epoch": 0.44, "learning_rate": 0.00029807837019363124, "loss": 0.3094, "step": 1600 }, { "epoch": 0.44, "eval_loss": 0.2237711250782013, "eval_runtime": 1040.3304, "eval_samples_per_second": 11.23, "eval_steps_per_second": 1.404, "eval_wer": 0.5121668334394615, "step": 1600 }, { "epoch": 0.44, "learning_rate": 0.00029805083968064604, "loss": 0.3198, "step": 1610 }, { "epoch": 0.44, "learning_rate": 0.0002980233091676608, "loss": 0.3048, "step": 1620 }, { "epoch": 0.45, "learning_rate": 0.0002979957786546756, "loss": 0.3149, "step": 1630 }, { "epoch": 0.45, "learning_rate": 0.00029796824814169037, "loss": 0.3298, "step": 1640 }, { "epoch": 0.45, "learning_rate": 0.0002979407176287051, "loss": 0.3102, "step": 1650 }, { "epoch": 0.45, "learning_rate": 0.0002979131871157199, "loss": 0.2989, "step": 1660 }, { "epoch": 0.46, "learning_rate": 0.00029788565660273465, "loss": 0.3399, "step": 1670 }, { "epoch": 0.46, "learning_rate": 0.00029785812608974945, "loss": 0.2812, "step": 1680 }, { "epoch": 0.46, "learning_rate": 0.00029783059557676424, "loss": 0.4134, "step": 1690 }, { "epoch": 0.47, "learning_rate": 0.000297803065063779, "loss": 0.305, "step": 1700 }, { "epoch": 0.47, "eval_loss": 0.23389825224876404, "eval_runtime": 1030.0489, "eval_samples_per_second": 11.342, "eval_steps_per_second": 1.418, "eval_wer": 0.5010688620030929, "step": 1700 }, { "epoch": 0.47, "learning_rate": 0.0002977755345507938, "loss": 0.3217, "step": 1710 }, { "epoch": 0.47, "learning_rate": 0.0002977480040378085, "loss": 0.2761, "step": 1720 }, { "epoch": 0.47, "learning_rate": 0.0002977204735248233, "loss": 0.3057, "step": 1730 }, { "epoch": 0.48, "learning_rate": 0.0002976929430118381, "loss": 0.3093, "step": 1740 }, { "epoch": 0.48, "learning_rate": 0.00029766541249885286, "loss": 0.2508, "step": 1750 }, { "epoch": 0.48, "learning_rate": 0.00029763788198586765, "loss": 0.3088, "step": 1760 }, { "epoch": 0.48, "learning_rate": 0.0002976103514728824, "loss": 0.2762, "step": 1770 }, { "epoch": 0.49, "learning_rate": 0.0002975828209598972, "loss": 0.2767, "step": 1780 }, { "epoch": 0.49, "learning_rate": 0.000297555290446912, "loss": 0.3105, "step": 1790 }, { "epoch": 0.49, "learning_rate": 0.0002975277599339267, "loss": 0.258, "step": 1800 }, { "epoch": 0.49, "eval_loss": 0.2067476361989975, "eval_runtime": 1030.7384, "eval_samples_per_second": 11.335, "eval_steps_per_second": 1.417, "eval_wer": 0.5007845901937596, "step": 1800 }, { "epoch": 0.5, "learning_rate": 0.0002975002294209415, "loss": 0.3283, "step": 1810 }, { "epoch": 0.5, "learning_rate": 0.0002974726989079563, "loss": 0.2985, "step": 1820 }, { "epoch": 0.5, "learning_rate": 0.00029744516839497106, "loss": 0.2884, "step": 1830 }, { "epoch": 0.5, "learning_rate": 0.00029741763788198585, "loss": 0.2912, "step": 1840 }, { "epoch": 0.51, "learning_rate": 0.0002973901073690006, "loss": 0.2704, "step": 1850 }, { "epoch": 0.51, "learning_rate": 0.0002973625768560154, "loss": 0.2642, "step": 1860 }, { "epoch": 0.51, "learning_rate": 0.0002973350463430302, "loss": 0.2884, "step": 1870 }, { "epoch": 0.52, "learning_rate": 0.00029730751583004493, "loss": 0.2991, "step": 1880 }, { "epoch": 0.52, "learning_rate": 0.0002972799853170597, "loss": 0.2994, "step": 1890 }, { "epoch": 0.52, "learning_rate": 0.00029725245480407447, "loss": 0.2673, "step": 1900 }, { "epoch": 0.52, "eval_loss": 0.2200978547334671, "eval_runtime": 1028.7302, "eval_samples_per_second": 11.357, "eval_steps_per_second": 1.42, "eval_wer": 0.5077321932138633, "step": 1900 }, { "epoch": 0.52, "learning_rate": 0.00029722492429108926, "loss": 0.3358, "step": 1910 }, { "epoch": 0.53, "learning_rate": 0.00029719739377810406, "loss": 0.2968, "step": 1920 }, { "epoch": 0.53, "learning_rate": 0.0002971698632651188, "loss": 0.2756, "step": 1930 }, { "epoch": 0.53, "learning_rate": 0.0002971423327521336, "loss": 0.3264, "step": 1940 }, { "epoch": 0.53, "learning_rate": 0.0002971148022391484, "loss": 0.2562, "step": 1950 }, { "epoch": 0.54, "learning_rate": 0.00029708727172616313, "loss": 0.2717, "step": 1960 }, { "epoch": 0.54, "learning_rate": 0.00029705974121317793, "loss": 0.283, "step": 1970 }, { "epoch": 0.54, "learning_rate": 0.00029703221070019267, "loss": 0.2728, "step": 1980 }, { "epoch": 0.55, "learning_rate": 0.00029700468018720747, "loss": 0.2867, "step": 1990 }, { "epoch": 0.55, "learning_rate": 0.00029697714967422226, "loss": 0.2256, "step": 2000 }, { "epoch": 0.55, "eval_loss": 0.18709534406661987, "eval_runtime": 1029.6709, "eval_samples_per_second": 11.346, "eval_steps_per_second": 1.419, "eval_wer": 0.46329482397889565, "step": 2000 }, { "epoch": 0.55, "learning_rate": 0.000296949619161237, "loss": 0.2914, "step": 2010 }, { "epoch": 0.55, "learning_rate": 0.0002969220886482518, "loss": 0.3636, "step": 2020 }, { "epoch": 0.56, "learning_rate": 0.00029689455813526654, "loss": 0.2804, "step": 2030 }, { "epoch": 0.56, "learning_rate": 0.00029686702762228134, "loss": 0.313, "step": 2040 }, { "epoch": 0.56, "learning_rate": 0.00029683949710929613, "loss": 0.2664, "step": 2050 }, { "epoch": 0.56, "learning_rate": 0.0002968119665963109, "loss": 0.2472, "step": 2060 }, { "epoch": 0.57, "learning_rate": 0.00029678443608332567, "loss": 0.2605, "step": 2070 }, { "epoch": 0.57, "learning_rate": 0.0002967569055703404, "loss": 0.2489, "step": 2080 }, { "epoch": 0.57, "learning_rate": 0.0002967293750573552, "loss": 0.2753, "step": 2090 }, { "epoch": 0.58, "learning_rate": 0.00029670184454437, "loss": 0.2697, "step": 2100 }, { "epoch": 0.58, "eval_loss": 0.19330532848834991, "eval_runtime": 1030.6142, "eval_samples_per_second": 11.336, "eval_steps_per_second": 1.418, "eval_wer": 0.4696283998908396, "step": 2100 }, { "epoch": 0.58, "learning_rate": 0.00029667431403138474, "loss": 0.2843, "step": 2110 }, { "epoch": 0.58, "learning_rate": 0.00029664678351839954, "loss": 0.2692, "step": 2120 }, { "epoch": 0.58, "learning_rate": 0.0002966192530054143, "loss": 0.2718, "step": 2130 }, { "epoch": 0.59, "learning_rate": 0.0002965917224924291, "loss": 0.2774, "step": 2140 }, { "epoch": 0.59, "learning_rate": 0.00029656419197944387, "loss": 0.2596, "step": 2150 }, { "epoch": 0.59, "learning_rate": 0.0002965366614664586, "loss": 0.272, "step": 2160 }, { "epoch": 0.59, "learning_rate": 0.0002965091309534734, "loss": 0.3172, "step": 2170 }, { "epoch": 0.6, "learning_rate": 0.0002964816004404882, "loss": 0.3228, "step": 2180 }, { "epoch": 0.6, "learning_rate": 0.00029645406992750295, "loss": 0.2719, "step": 2190 }, { "epoch": 0.6, "learning_rate": 0.00029642653941451774, "loss": 0.233, "step": 2200 }, { "epoch": 0.6, "eval_loss": 0.18349601328372955, "eval_runtime": 1030.0621, "eval_samples_per_second": 11.342, "eval_steps_per_second": 1.418, "eval_wer": 0.45199217683980714, "step": 2200 }, { "epoch": 0.61, "learning_rate": 0.0002963990089015325, "loss": 0.2861, "step": 2210 }, { "epoch": 0.61, "learning_rate": 0.0002963714783885473, "loss": 0.2984, "step": 2220 }, { "epoch": 0.61, "learning_rate": 0.0002963439478755621, "loss": 0.2395, "step": 2230 }, { "epoch": 0.61, "learning_rate": 0.0002963164173625768, "loss": 0.2607, "step": 2240 }, { "epoch": 0.62, "learning_rate": 0.0002962888868495916, "loss": 0.2739, "step": 2250 }, { "epoch": 0.62, "learning_rate": 0.00029626135633660636, "loss": 0.2549, "step": 2260 }, { "epoch": 0.62, "learning_rate": 0.00029623382582362115, "loss": 0.266, "step": 2270 }, { "epoch": 0.62, "learning_rate": 0.00029620629531063595, "loss": 0.2786, "step": 2280 }, { "epoch": 0.63, "learning_rate": 0.0002961787647976507, "loss": 0.2524, "step": 2290 }, { "epoch": 0.63, "learning_rate": 0.0002961512342846655, "loss": 0.2078, "step": 2300 }, { "epoch": 0.63, "eval_loss": 0.1761041283607483, "eval_runtime": 1030.565, "eval_samples_per_second": 11.337, "eval_steps_per_second": 1.418, "eval_wer": 0.4481601928499955, "step": 2300 }, { "epoch": 0.63, "learning_rate": 0.0002961237037716803, "loss": 0.2331, "step": 2310 }, { "epoch": 0.64, "learning_rate": 0.000296096173258695, "loss": 0.2197, "step": 2320 }, { "epoch": 0.64, "learning_rate": 0.0002960686427457098, "loss": 0.2595, "step": 2330 }, { "epoch": 0.64, "learning_rate": 0.00029604111223272456, "loss": 0.2571, "step": 2340 }, { "epoch": 0.64, "learning_rate": 0.00029601358171973935, "loss": 0.2074, "step": 2350 }, { "epoch": 0.65, "learning_rate": 0.00029598605120675415, "loss": 0.2587, "step": 2360 }, { "epoch": 0.65, "learning_rate": 0.0002959585206937689, "loss": 0.246, "step": 2370 }, { "epoch": 0.65, "learning_rate": 0.0002959309901807837, "loss": 0.2964, "step": 2380 }, { "epoch": 0.65, "learning_rate": 0.00029590345966779843, "loss": 0.2453, "step": 2390 }, { "epoch": 0.66, "learning_rate": 0.0002958759291548132, "loss": 0.2721, "step": 2400 }, { "epoch": 0.66, "eval_loss": 0.17422164976596832, "eval_runtime": 1036.0048, "eval_samples_per_second": 11.277, "eval_steps_per_second": 1.41, "eval_wer": 0.44059856272173203, "step": 2400 }, { "epoch": 0.66, "learning_rate": 0.000295848398641828, "loss": 0.2198, "step": 2410 }, { "epoch": 0.66, "learning_rate": 0.00029582086812884276, "loss": 0.2405, "step": 2420 }, { "epoch": 0.67, "learning_rate": 0.00029579333761585756, "loss": 0.2334, "step": 2430 }, { "epoch": 0.67, "learning_rate": 0.0002957658071028723, "loss": 0.2894, "step": 2440 }, { "epoch": 0.67, "learning_rate": 0.0002957382765898871, "loss": 0.2534, "step": 2450 }, { "epoch": 0.67, "learning_rate": 0.0002957107460769019, "loss": 0.2769, "step": 2460 }, { "epoch": 0.68, "learning_rate": 0.00029568321556391663, "loss": 0.284, "step": 2470 }, { "epoch": 0.68, "learning_rate": 0.00029565568505093143, "loss": 0.2619, "step": 2480 }, { "epoch": 0.68, "learning_rate": 0.00029562815453794617, "loss": 0.4885, "step": 2490 }, { "epoch": 0.69, "learning_rate": 0.00029560062402496097, "loss": 0.2807, "step": 2500 }, { "epoch": 0.69, "eval_loss": 0.1716814488172531, "eval_runtime": 1034.3139, "eval_samples_per_second": 11.295, "eval_steps_per_second": 1.413, "eval_wer": 0.4375966524151733, "step": 2500 }, { "epoch": 0.69, "learning_rate": 0.00029557309351197576, "loss": 0.264, "step": 2510 }, { "epoch": 0.69, "learning_rate": 0.0002955455629989905, "loss": 0.2517, "step": 2520 }, { "epoch": 0.69, "learning_rate": 0.0002955180324860053, "loss": 0.2351, "step": 2530 }, { "epoch": 0.7, "learning_rate": 0.00029549050197302004, "loss": 0.2626, "step": 2540 }, { "epoch": 0.7, "learning_rate": 0.00029546297146003484, "loss": 0.2721, "step": 2550 }, { "epoch": 0.7, "learning_rate": 0.00029543544094704963, "loss": 0.2779, "step": 2560 }, { "epoch": 0.7, "learning_rate": 0.0002954079104340644, "loss": 0.2684, "step": 2570 }, { "epoch": 0.71, "learning_rate": 0.00029538037992107917, "loss": 0.2444, "step": 2580 }, { "epoch": 0.71, "learning_rate": 0.00029535284940809397, "loss": 0.2643, "step": 2590 }, { "epoch": 0.71, "learning_rate": 0.0002953253188951087, "loss": 0.2397, "step": 2600 }, { "epoch": 0.71, "eval_loss": 0.178777813911438, "eval_runtime": 1033.0369, "eval_samples_per_second": 11.309, "eval_steps_per_second": 1.414, "eval_wer": 0.4444191758391704, "step": 2600 }, { "epoch": 0.72, "learning_rate": 0.0002952977883821235, "loss": 0.2436, "step": 2610 }, { "epoch": 0.72, "learning_rate": 0.0002952702578691383, "loss": 0.2767, "step": 2620 }, { "epoch": 0.72, "learning_rate": 0.00029524272735615304, "loss": 0.2188, "step": 2630 }, { "epoch": 0.72, "learning_rate": 0.00029521519684316784, "loss": 0.2281, "step": 2640 }, { "epoch": 0.73, "learning_rate": 0.0002951876663301826, "loss": 0.265, "step": 2650 }, { "epoch": 0.73, "learning_rate": 0.00029516013581719737, "loss": 0.27, "step": 2660 }, { "epoch": 0.73, "learning_rate": 0.00029513260530421217, "loss": 0.2213, "step": 2670 }, { "epoch": 0.73, "learning_rate": 0.0002951050747912269, "loss": 0.2646, "step": 2680 }, { "epoch": 0.74, "learning_rate": 0.0002950775442782417, "loss": 0.2378, "step": 2690 }, { "epoch": 0.74, "learning_rate": 0.00029505001376525645, "loss": 0.2672, "step": 2700 }, { "epoch": 0.74, "eval_loss": 0.15344244241714478, "eval_runtime": 1033.7172, "eval_samples_per_second": 11.302, "eval_steps_per_second": 1.413, "eval_wer": 0.4126262166833439, "step": 2700 }, { "epoch": 0.74, "learning_rate": 0.00029502248325227124, "loss": 0.2552, "step": 2710 }, { "epoch": 0.75, "learning_rate": 0.00029499495273928604, "loss": 0.2254, "step": 2720 }, { "epoch": 0.75, "learning_rate": 0.0002949674222263008, "loss": 0.252, "step": 2730 }, { "epoch": 0.75, "learning_rate": 0.0002949398917133156, "loss": 0.2361, "step": 2740 }, { "epoch": 0.75, "learning_rate": 0.0002949123612003303, "loss": 0.2479, "step": 2750 }, { "epoch": 0.76, "learning_rate": 0.0002948848306873451, "loss": 0.2512, "step": 2760 }, { "epoch": 0.76, "learning_rate": 0.0002948573001743599, "loss": 0.2675, "step": 2770 }, { "epoch": 0.76, "learning_rate": 0.00029482976966137465, "loss": 0.2157, "step": 2780 }, { "epoch": 0.76, "learning_rate": 0.00029480223914838945, "loss": 0.2625, "step": 2790 }, { "epoch": 0.77, "learning_rate": 0.0002947747086354042, "loss": 0.1954, "step": 2800 }, { "epoch": 0.77, "eval_loss": 0.1593170017004013, "eval_runtime": 1038.0053, "eval_samples_per_second": 11.255, "eval_steps_per_second": 1.408, "eval_wer": 0.40848721913945235, "step": 2800 }, { "epoch": 0.77, "learning_rate": 0.000294747178122419, "loss": 0.23, "step": 2810 }, { "epoch": 0.77, "learning_rate": 0.0002947196476094338, "loss": 0.2252, "step": 2820 }, { "epoch": 0.78, "learning_rate": 0.0002946921170964485, "loss": 0.2564, "step": 2830 }, { "epoch": 0.78, "learning_rate": 0.0002946645865834633, "loss": 0.2578, "step": 2840 }, { "epoch": 0.78, "learning_rate": 0.00029463705607047806, "loss": 0.2118, "step": 2850 }, { "epoch": 0.78, "learning_rate": 0.00029460952555749285, "loss": 0.2366, "step": 2860 }, { "epoch": 0.79, "learning_rate": 0.00029458199504450765, "loss": 0.2228, "step": 2870 }, { "epoch": 0.79, "learning_rate": 0.0002945544645315224, "loss": 0.252, "step": 2880 }, { "epoch": 0.79, "learning_rate": 0.0002945269340185372, "loss": 0.2554, "step": 2890 }, { "epoch": 0.79, "learning_rate": 0.00029449940350555193, "loss": 0.2753, "step": 2900 }, { "epoch": 0.79, "eval_loss": 0.1575159728527069, "eval_runtime": 1033.311, "eval_samples_per_second": 11.306, "eval_steps_per_second": 1.414, "eval_wer": 0.4128081506413172, "step": 2900 }, { "epoch": 0.8, "learning_rate": 0.0002944718729925667, "loss": 0.2448, "step": 2910 }, { "epoch": 0.8, "learning_rate": 0.0002944443424795815, "loss": 0.2356, "step": 2920 }, { "epoch": 0.8, "learning_rate": 0.00029441681196659626, "loss": 0.2523, "step": 2930 }, { "epoch": 0.81, "learning_rate": 0.00029438928145361106, "loss": 0.2386, "step": 2940 }, { "epoch": 0.81, "learning_rate": 0.0002943617509406258, "loss": 0.2327, "step": 2950 }, { "epoch": 0.81, "learning_rate": 0.0002943342204276406, "loss": 0.2543, "step": 2960 }, { "epoch": 0.81, "learning_rate": 0.0002943066899146554, "loss": 0.2409, "step": 2970 }, { "epoch": 0.82, "learning_rate": 0.0002942791594016702, "loss": 0.263, "step": 2980 }, { "epoch": 0.82, "learning_rate": 0.00029425162888868493, "loss": 0.231, "step": 2990 }, { "epoch": 0.82, "learning_rate": 0.0002942240983756997, "loss": 0.1975, "step": 3000 }, { "epoch": 0.82, "eval_loss": 0.14351560175418854, "eval_runtime": 1046.2091, "eval_samples_per_second": 11.167, "eval_steps_per_second": 1.396, "eval_wer": 0.39260211043391247, "step": 3000 }, { "epoch": 0.82, "learning_rate": 0.00029419656786271447, "loss": 0.2252, "step": 3010 }, { "epoch": 0.83, "learning_rate": 0.00029416903734972926, "loss": 0.2248, "step": 3020 }, { "epoch": 0.83, "learning_rate": 0.00029414150683674406, "loss": 0.2285, "step": 3030 }, { "epoch": 0.83, "learning_rate": 0.0002941139763237588, "loss": 0.2458, "step": 3040 }, { "epoch": 0.84, "learning_rate": 0.0002940864458107736, "loss": 0.2482, "step": 3050 }, { "epoch": 0.84, "learning_rate": 0.00029405891529778834, "loss": 0.2798, "step": 3060 }, { "epoch": 0.84, "learning_rate": 0.00029403138478480313, "loss": 0.2338, "step": 3070 }, { "epoch": 0.84, "learning_rate": 0.00029400385427181793, "loss": 0.2369, "step": 3080 }, { "epoch": 0.85, "learning_rate": 0.00029397632375883267, "loss": 0.2208, "step": 3090 }, { "epoch": 0.85, "learning_rate": 0.00029394879324584747, "loss": 0.2282, "step": 3100 }, { "epoch": 0.85, "eval_loss": 0.17016960680484772, "eval_runtime": 1038.5865, "eval_samples_per_second": 11.249, "eval_steps_per_second": 1.407, "eval_wer": 0.4270672245974711, "step": 3100 }, { "epoch": 0.85, "learning_rate": 0.0002939212627328622, "loss": 0.2531, "step": 3110 }, { "epoch": 0.85, "learning_rate": 0.000293893732219877, "loss": 0.3018, "step": 3120 }, { "epoch": 0.86, "learning_rate": 0.0002938662017068918, "loss": 0.2405, "step": 3130 }, { "epoch": 0.86, "learning_rate": 0.00029383867119390654, "loss": 0.2549, "step": 3140 }, { "epoch": 0.86, "learning_rate": 0.00029381114068092134, "loss": 0.2245, "step": 3150 }, { "epoch": 0.87, "learning_rate": 0.0002937836101679361, "loss": 0.2354, "step": 3160 }, { "epoch": 0.87, "learning_rate": 0.00029375607965495087, "loss": 0.2367, "step": 3170 }, { "epoch": 0.87, "learning_rate": 0.00029372854914196567, "loss": 0.2452, "step": 3180 }, { "epoch": 0.87, "learning_rate": 0.0002937010186289804, "loss": 0.2382, "step": 3190 }, { "epoch": 0.88, "learning_rate": 0.0002936734881159952, "loss": 0.2372, "step": 3200 }, { "epoch": 0.88, "eval_loss": 0.1416802853345871, "eval_runtime": 1032.3105, "eval_samples_per_second": 11.317, "eval_steps_per_second": 1.415, "eval_wer": 0.3964682070408442, "step": 3200 }, { "epoch": 0.88, "learning_rate": 0.00029364595760300995, "loss": 0.2478, "step": 3210 }, { "epoch": 0.88, "learning_rate": 0.00029361842709002474, "loss": 0.2127, "step": 3220 }, { "epoch": 0.89, "learning_rate": 0.00029359089657703954, "loss": 0.2021, "step": 3230 }, { "epoch": 0.89, "learning_rate": 0.0002935633660640543, "loss": 0.2422, "step": 3240 }, { "epoch": 0.89, "learning_rate": 0.0002935358355510691, "loss": 0.2399, "step": 3250 }, { "epoch": 0.89, "learning_rate": 0.0002935083050380838, "loss": 0.2405, "step": 3260 }, { "epoch": 0.9, "learning_rate": 0.0002934807745250986, "loss": 0.1949, "step": 3270 }, { "epoch": 0.9, "learning_rate": 0.0002934532440121134, "loss": 0.2234, "step": 3280 }, { "epoch": 0.9, "learning_rate": 0.0002934257134991282, "loss": 0.2403, "step": 3290 }, { "epoch": 0.9, "learning_rate": 0.00029339818298614295, "loss": 0.229, "step": 3300 }, { "epoch": 0.9, "eval_loss": 0.13848544657230377, "eval_runtime": 1033.0662, "eval_samples_per_second": 11.309, "eval_steps_per_second": 1.414, "eval_wer": 0.38596152096788866, "step": 3300 }, { "epoch": 0.91, "learning_rate": 0.0002933706524731577, "loss": 0.2132, "step": 3310 }, { "epoch": 0.91, "learning_rate": 0.0002933431219601725, "loss": 0.5408, "step": 3320 }, { "epoch": 0.91, "learning_rate": 0.0002933155914471873, "loss": 0.2344, "step": 3330 }, { "epoch": 0.92, "learning_rate": 0.0002932880609342021, "loss": 0.2454, "step": 3340 }, { "epoch": 0.92, "learning_rate": 0.0002932605304212168, "loss": 0.2326, "step": 3350 }, { "epoch": 0.92, "learning_rate": 0.0002932329999082316, "loss": 0.2076, "step": 3360 }, { "epoch": 0.92, "learning_rate": 0.00029320546939524635, "loss": 0.3867, "step": 3370 }, { "epoch": 0.93, "learning_rate": 0.00029317793888226115, "loss": 0.2038, "step": 3380 }, { "epoch": 0.93, "learning_rate": 0.00029315040836927595, "loss": 0.2371, "step": 3390 }, { "epoch": 0.93, "learning_rate": 0.0002931228778562907, "loss": 0.2332, "step": 3400 }, { "epoch": 0.93, "eval_loss": 0.13254162669181824, "eval_runtime": 1032.4462, "eval_samples_per_second": 11.316, "eval_steps_per_second": 1.415, "eval_wer": 0.380503502228691, "step": 3400 }, { "epoch": 0.93, "learning_rate": 0.0002930953473433055, "loss": 0.3243, "step": 3410 }, { "epoch": 0.94, "learning_rate": 0.0002930678168303202, "loss": 0.2216, "step": 3420 }, { "epoch": 0.94, "learning_rate": 0.000293040286317335, "loss": 0.2173, "step": 3430 }, { "epoch": 0.94, "learning_rate": 0.0002930127558043498, "loss": 0.2095, "step": 3440 }, { "epoch": 0.95, "learning_rate": 0.00029298522529136456, "loss": 0.2069, "step": 3450 }, { "epoch": 0.95, "learning_rate": 0.00029295769477837935, "loss": 0.2326, "step": 3460 }, { "epoch": 0.95, "learning_rate": 0.0002929301642653941, "loss": 0.2067, "step": 3470 }, { "epoch": 0.95, "learning_rate": 0.0002929026337524089, "loss": 0.2236, "step": 3480 }, { "epoch": 0.96, "learning_rate": 0.0002928751032394237, "loss": 0.229, "step": 3490 }, { "epoch": 0.96, "learning_rate": 0.00029284757272643843, "loss": 0.2074, "step": 3500 }, { "epoch": 0.96, "eval_loss": 0.13085076212882996, "eval_runtime": 1031.5246, "eval_samples_per_second": 11.326, "eval_steps_per_second": 1.416, "eval_wer": 0.375, "step": 3500 }, { "epoch": 0.96, "learning_rate": 0.0002928200422134532, "loss": 0.2062, "step": 3510 }, { "epoch": 0.96, "learning_rate": 0.00029279251170046797, "loss": 0.3621, "step": 3520 }, { "epoch": 0.97, "learning_rate": 0.00029276498118748276, "loss": 0.2127, "step": 3530 }, { "epoch": 0.97, "learning_rate": 0.00029273745067449756, "loss": 0.2343, "step": 3540 }, { "epoch": 0.97, "learning_rate": 0.0002927099201615123, "loss": 0.1566, "step": 3550 }, { "epoch": 0.98, "learning_rate": 0.0002926823896485271, "loss": 0.2281, "step": 3560 }, { "epoch": 0.98, "learning_rate": 0.00029265485913554184, "loss": 0.2268, "step": 3570 }, { "epoch": 0.98, "learning_rate": 0.00029262732862255663, "loss": 0.1878, "step": 3580 }, { "epoch": 0.98, "learning_rate": 0.00029259979810957143, "loss": 0.1967, "step": 3590 }, { "epoch": 0.99, "learning_rate": 0.00029257226759658617, "loss": 0.1937, "step": 3600 }, { "epoch": 0.99, "eval_loss": 0.13056674599647522, "eval_runtime": 1033.4241, "eval_samples_per_second": 11.305, "eval_steps_per_second": 1.414, "eval_wer": 0.37841126171199857, "step": 3600 }, { "epoch": 0.99, "learning_rate": 0.00029254473708360097, "loss": 0.204, "step": 3610 }, { "epoch": 0.99, "learning_rate": 0.0002925172065706157, "loss": 0.2036, "step": 3620 }, { "epoch": 0.99, "learning_rate": 0.0002924896760576305, "loss": 0.2268, "step": 3630 }, { "epoch": 1.0, "learning_rate": 0.0002924621455446453, "loss": 0.2303, "step": 3640 }, { "epoch": 1.0, "learning_rate": 0.0002924346150316601, "loss": 0.2061, "step": 3650 }, { "epoch": 1.0, "learning_rate": 0.00029240708451867484, "loss": 0.222, "step": 3660 }, { "epoch": 1.01, "learning_rate": 0.0002923795540056896, "loss": 0.1853, "step": 3670 }, { "epoch": 1.01, "learning_rate": 0.00029235202349270437, "loss": 0.2252, "step": 3680 }, { "epoch": 1.01, "learning_rate": 0.00029232449297971917, "loss": 0.2318, "step": 3690 }, { "epoch": 1.01, "learning_rate": 0.00029229696246673396, "loss": 0.2211, "step": 3700 }, { "epoch": 1.01, "eval_loss": 0.14021088182926178, "eval_runtime": 1031.5194, "eval_samples_per_second": 11.326, "eval_steps_per_second": 1.416, "eval_wer": 0.383619121258983, "step": 3700 }, { "epoch": 1.02, "learning_rate": 0.0002922694319537487, "loss": 0.1938, "step": 3710 }, { "epoch": 1.02, "learning_rate": 0.00029224190144076345, "loss": 0.2073, "step": 3720 }, { "epoch": 1.02, "learning_rate": 0.00029221437092777824, "loss": 0.2023, "step": 3730 }, { "epoch": 1.02, "learning_rate": 0.00029218684041479304, "loss": 0.2171, "step": 3740 }, { "epoch": 1.03, "learning_rate": 0.00029215930990180783, "loss": 0.1932, "step": 3750 }, { "epoch": 1.03, "learning_rate": 0.0002921317793888226, "loss": 0.1921, "step": 3760 }, { "epoch": 1.03, "learning_rate": 0.00029210424887583737, "loss": 0.229, "step": 3770 }, { "epoch": 1.04, "learning_rate": 0.0002920767183628521, "loss": 0.1971, "step": 3780 }, { "epoch": 1.04, "learning_rate": 0.0002920491878498669, "loss": 0.2161, "step": 3790 }, { "epoch": 1.04, "learning_rate": 0.0002920216573368817, "loss": 0.1928, "step": 3800 }, { "epoch": 1.04, "eval_loss": 0.13391314446926117, "eval_runtime": 1039.3763, "eval_samples_per_second": 11.24, "eval_steps_per_second": 1.406, "eval_wer": 0.3714409169471482, "step": 3800 }, { "epoch": 1.04, "learning_rate": 0.00029199412682389645, "loss": 0.227, "step": 3810 }, { "epoch": 1.05, "learning_rate": 0.00029196659631091124, "loss": 0.1845, "step": 3820 }, { "epoch": 1.05, "learning_rate": 0.000291939065797926, "loss": 0.2082, "step": 3830 }, { "epoch": 1.05, "learning_rate": 0.0002919115352849408, "loss": 0.2131, "step": 3840 }, { "epoch": 1.05, "learning_rate": 0.0002918840047719556, "loss": 0.1836, "step": 3850 }, { "epoch": 1.06, "learning_rate": 0.0002918564742589703, "loss": 0.2096, "step": 3860 }, { "epoch": 1.06, "learning_rate": 0.0002918289437459851, "loss": 0.198, "step": 3870 }, { "epoch": 1.06, "learning_rate": 0.00029180141323299985, "loss": 0.194, "step": 3880 }, { "epoch": 1.07, "learning_rate": 0.00029177388272001465, "loss": 0.2517, "step": 3890 }, { "epoch": 1.07, "learning_rate": 0.00029174635220702945, "loss": 0.176, "step": 3900 }, { "epoch": 1.07, "eval_loss": 0.13185542821884155, "eval_runtime": 1034.2276, "eval_samples_per_second": 11.296, "eval_steps_per_second": 1.413, "eval_wer": 0.37426089329573364, "step": 3900 }, { "epoch": 1.07, "learning_rate": 0.0002917188216940442, "loss": 0.2084, "step": 3910 }, { "epoch": 1.07, "learning_rate": 0.000291691291181059, "loss": 0.1942, "step": 3920 }, { "epoch": 1.08, "learning_rate": 0.0002916637606680737, "loss": 0.2091, "step": 3930 }, { "epoch": 1.08, "learning_rate": 0.0002916362301550885, "loss": 0.2117, "step": 3940 }, { "epoch": 1.08, "learning_rate": 0.0002916086996421033, "loss": 0.1762, "step": 3950 }, { "epoch": 1.09, "learning_rate": 0.0002915811691291181, "loss": 0.2032, "step": 3960 }, { "epoch": 1.09, "learning_rate": 0.00029155363861613285, "loss": 0.2233, "step": 3970 }, { "epoch": 1.09, "learning_rate": 0.0002915261081031476, "loss": 0.1855, "step": 3980 }, { "epoch": 1.09, "learning_rate": 0.0002914985775901624, "loss": 0.1992, "step": 3990 }, { "epoch": 1.1, "learning_rate": 0.0002914710470771772, "loss": 0.1818, "step": 4000 }, { "epoch": 1.1, "eval_loss": 0.1326366364955902, "eval_runtime": 1036.1093, "eval_samples_per_second": 11.276, "eval_steps_per_second": 1.41, "eval_wer": 0.3759892658964796, "step": 4000 }, { "epoch": 1.1, "learning_rate": 0.000291443516564192, "loss": 0.1836, "step": 4010 }, { "epoch": 1.1, "learning_rate": 0.0002914159860512067, "loss": 0.2213, "step": 4020 }, { "epoch": 1.1, "learning_rate": 0.00029138845553822147, "loss": 0.1852, "step": 4030 }, { "epoch": 1.11, "learning_rate": 0.00029136092502523626, "loss": 0.2357, "step": 4040 }, { "epoch": 1.11, "learning_rate": 0.00029133339451225106, "loss": 0.1824, "step": 4050 }, { "epoch": 1.11, "learning_rate": 0.00029130586399926585, "loss": 0.1947, "step": 4060 }, { "epoch": 1.12, "learning_rate": 0.0002912783334862806, "loss": 0.2006, "step": 4070 }, { "epoch": 1.12, "learning_rate": 0.00029125080297329534, "loss": 0.1958, "step": 4080 }, { "epoch": 1.12, "learning_rate": 0.00029122327246031013, "loss": 0.2205, "step": 4090 }, { "epoch": 1.12, "learning_rate": 0.00029119574194732493, "loss": 0.2191, "step": 4100 }, { "epoch": 1.12, "eval_loss": 0.13117532432079315, "eval_runtime": 1035.211, "eval_samples_per_second": 11.286, "eval_steps_per_second": 1.411, "eval_wer": 0.37676248521786593, "step": 4100 }, { "epoch": 1.13, "learning_rate": 0.0002911682114343397, "loss": 0.1991, "step": 4110 }, { "epoch": 1.13, "learning_rate": 0.00029114068092135447, "loss": 0.2014, "step": 4120 }, { "epoch": 1.13, "learning_rate": 0.00029111315040836926, "loss": 0.2254, "step": 4130 }, { "epoch": 1.13, "learning_rate": 0.000291085619895384, "loss": 0.2056, "step": 4140 }, { "epoch": 1.14, "learning_rate": 0.0002910580893823988, "loss": 0.1819, "step": 4150 }, { "epoch": 1.14, "learning_rate": 0.0002910305588694136, "loss": 0.2095, "step": 4160 }, { "epoch": 1.14, "learning_rate": 0.00029100302835642834, "loss": 0.2117, "step": 4170 }, { "epoch": 1.15, "learning_rate": 0.00029097549784344313, "loss": 0.1921, "step": 4180 }, { "epoch": 1.15, "learning_rate": 0.0002909479673304579, "loss": 0.1961, "step": 4190 }, { "epoch": 1.15, "learning_rate": 0.00029092043681747267, "loss": 0.1848, "step": 4200 }, { "epoch": 1.15, "eval_loss": 0.12278643250465393, "eval_runtime": 1034.1658, "eval_samples_per_second": 11.297, "eval_steps_per_second": 1.413, "eval_wer": 0.35989948148821976, "step": 4200 }, { "epoch": 1.15, "learning_rate": 0.00029089290630448746, "loss": 0.1991, "step": 4210 }, { "epoch": 1.16, "learning_rate": 0.0002908653757915022, "loss": 0.1814, "step": 4220 }, { "epoch": 1.16, "learning_rate": 0.000290837845278517, "loss": 0.1836, "step": 4230 }, { "epoch": 1.16, "learning_rate": 0.00029081031476553174, "loss": 0.1921, "step": 4240 }, { "epoch": 1.16, "learning_rate": 0.00029078278425254654, "loss": 0.1855, "step": 4250 }, { "epoch": 1.17, "learning_rate": 0.00029075525373956133, "loss": 0.2016, "step": 4260 }, { "epoch": 1.17, "learning_rate": 0.00029072772322657613, "loss": 0.205, "step": 4270 }, { "epoch": 1.17, "learning_rate": 0.00029070019271359087, "loss": 0.1881, "step": 4280 }, { "epoch": 1.18, "learning_rate": 0.0002906726622006056, "loss": 0.2192, "step": 4290 }, { "epoch": 1.18, "learning_rate": 0.0002906451316876204, "loss": 0.1671, "step": 4300 }, { "epoch": 1.18, "eval_loss": 0.12893985211849213, "eval_runtime": 1038.4465, "eval_samples_per_second": 11.25, "eval_steps_per_second": 1.407, "eval_wer": 0.37314654780314743, "step": 4300 }, { "epoch": 1.18, "learning_rate": 0.0002906176011746352, "loss": 0.2075, "step": 4310 }, { "epoch": 1.18, "learning_rate": 0.00029059007066165, "loss": 0.225, "step": 4320 }, { "epoch": 1.19, "learning_rate": 0.00029056254014866474, "loss": 0.2183, "step": 4330 }, { "epoch": 1.19, "learning_rate": 0.0002905350096356795, "loss": 0.2136, "step": 4340 }, { "epoch": 1.19, "learning_rate": 0.0002905074791226943, "loss": 0.198, "step": 4350 }, { "epoch": 1.19, "learning_rate": 0.0002904799486097091, "loss": 0.2346, "step": 4360 }, { "epoch": 1.2, "learning_rate": 0.00029045241809672387, "loss": 0.2207, "step": 4370 }, { "epoch": 1.2, "learning_rate": 0.0002904248875837386, "loss": 0.1996, "step": 4380 }, { "epoch": 1.2, "learning_rate": 0.00029039735707075335, "loss": 0.1872, "step": 4390 }, { "epoch": 1.21, "learning_rate": 0.00029036982655776815, "loss": 0.1759, "step": 4400 }, { "epoch": 1.21, "eval_loss": 0.1245460957288742, "eval_runtime": 1036.0049, "eval_samples_per_second": 11.277, "eval_steps_per_second": 1.41, "eval_wer": 0.36324251796597834, "step": 4400 }, { "epoch": 1.21, "learning_rate": 0.00029034229604478295, "loss": 0.1972, "step": 4410 }, { "epoch": 1.21, "learning_rate": 0.00029031476553179774, "loss": 0.2011, "step": 4420 }, { "epoch": 1.21, "learning_rate": 0.0002902872350188125, "loss": 0.1978, "step": 4430 }, { "epoch": 1.22, "learning_rate": 0.0002902597045058272, "loss": 0.1881, "step": 4440 }, { "epoch": 1.22, "learning_rate": 0.000290232173992842, "loss": 0.1781, "step": 4450 }, { "epoch": 1.22, "learning_rate": 0.0002902046434798568, "loss": 0.2026, "step": 4460 }, { "epoch": 1.22, "learning_rate": 0.0002901771129668716, "loss": 0.2026, "step": 4470 }, { "epoch": 1.23, "learning_rate": 0.00029014958245388635, "loss": 0.2007, "step": 4480 }, { "epoch": 1.23, "learning_rate": 0.00029012205194090115, "loss": 0.4222, "step": 4490 }, { "epoch": 1.23, "learning_rate": 0.0002900945214279159, "loss": 0.1783, "step": 4500 }, { "epoch": 1.23, "eval_loss": 0.12668462097644806, "eval_runtime": 1032.5946, "eval_samples_per_second": 11.314, "eval_steps_per_second": 1.415, "eval_wer": 0.35453242972800875, "step": 4500 }, { "epoch": 1.24, "learning_rate": 0.0002900669909149307, "loss": 0.1804, "step": 4510 }, { "epoch": 1.24, "learning_rate": 0.0002900394604019455, "loss": 0.1893, "step": 4520 }, { "epoch": 1.24, "learning_rate": 0.0002900119298889602, "loss": 0.1818, "step": 4530 }, { "epoch": 1.24, "learning_rate": 0.000289984399375975, "loss": 0.1815, "step": 4540 }, { "epoch": 1.25, "learning_rate": 0.00028995686886298976, "loss": 0.2069, "step": 4550 }, { "epoch": 1.25, "learning_rate": 0.00028992933835000456, "loss": 0.2155, "step": 4560 }, { "epoch": 1.25, "learning_rate": 0.00028990180783701935, "loss": 0.1816, "step": 4570 }, { "epoch": 1.25, "learning_rate": 0.0002898742773240341, "loss": 0.1832, "step": 4580 }, { "epoch": 1.26, "learning_rate": 0.0002898467468110489, "loss": 0.1775, "step": 4590 }, { "epoch": 1.26, "learning_rate": 0.00028981921629806363, "loss": 0.1576, "step": 4600 }, { "epoch": 1.26, "eval_loss": 0.1280120611190796, "eval_runtime": 1036.0204, "eval_samples_per_second": 11.277, "eval_steps_per_second": 1.41, "eval_wer": 0.3581028836532339, "step": 4600 }, { "epoch": 1.26, "learning_rate": 0.00028979168578507843, "loss": 0.1803, "step": 4610 }, { "epoch": 1.27, "learning_rate": 0.0002897641552720932, "loss": 0.1562, "step": 4620 }, { "epoch": 1.27, "learning_rate": 0.000289736624759108, "loss": 0.1841, "step": 4630 }, { "epoch": 1.27, "learning_rate": 0.00028970909424612276, "loss": 0.1993, "step": 4640 }, { "epoch": 1.27, "learning_rate": 0.0002896815637331375, "loss": 0.1856, "step": 4650 }, { "epoch": 1.28, "learning_rate": 0.0002896540332201523, "loss": 0.1863, "step": 4660 }, { "epoch": 1.28, "learning_rate": 0.0002896265027071671, "loss": 0.197, "step": 4670 }, { "epoch": 1.28, "learning_rate": 0.0002895989721941819, "loss": 0.1745, "step": 4680 }, { "epoch": 1.29, "learning_rate": 0.00028957144168119663, "loss": 0.1993, "step": 4690 }, { "epoch": 1.29, "learning_rate": 0.0002895439111682114, "loss": 0.1726, "step": 4700 }, { "epoch": 1.29, "eval_loss": 0.12151551246643066, "eval_runtime": 1035.4623, "eval_samples_per_second": 11.283, "eval_steps_per_second": 1.411, "eval_wer": 0.35406622396070225, "step": 4700 }, { "epoch": 1.29, "learning_rate": 0.00028951638065522617, "loss": 0.1771, "step": 4710 }, { "epoch": 1.29, "learning_rate": 0.00028948885014224096, "loss": 0.177, "step": 4720 }, { "epoch": 1.3, "learning_rate": 0.00028946131962925576, "loss": 0.1863, "step": 4730 }, { "epoch": 1.3, "learning_rate": 0.0002894337891162705, "loss": 0.1879, "step": 4740 }, { "epoch": 1.3, "learning_rate": 0.00028940625860328524, "loss": 0.1581, "step": 4750 }, { "epoch": 1.3, "learning_rate": 0.00028937872809030004, "loss": 0.1907, "step": 4760 }, { "epoch": 1.31, "learning_rate": 0.00028935119757731483, "loss": 0.1756, "step": 4770 }, { "epoch": 1.31, "learning_rate": 0.00028932366706432963, "loss": 0.1793, "step": 4780 }, { "epoch": 1.31, "learning_rate": 0.00028929613655134437, "loss": 0.183, "step": 4790 }, { "epoch": 1.32, "learning_rate": 0.0002892686060383591, "loss": 0.1702, "step": 4800 }, { "epoch": 1.32, "eval_loss": 0.13416741788387299, "eval_runtime": 1029.9454, "eval_samples_per_second": 11.343, "eval_steps_per_second": 1.419, "eval_wer": 0.3690075502592559, "step": 4800 }, { "epoch": 1.32, "learning_rate": 0.0002892410755253739, "loss": 0.1963, "step": 4810 }, { "epoch": 1.32, "learning_rate": 0.0002892135450123887, "loss": 0.1787, "step": 4820 }, { "epoch": 1.32, "learning_rate": 0.0002891860144994035, "loss": 0.1877, "step": 4830 }, { "epoch": 1.33, "learning_rate": 0.00028915848398641824, "loss": 0.1848, "step": 4840 }, { "epoch": 1.33, "learning_rate": 0.00028913095347343304, "loss": 0.1763, "step": 4850 }, { "epoch": 1.33, "learning_rate": 0.0002891034229604478, "loss": 0.2131, "step": 4860 }, { "epoch": 1.33, "learning_rate": 0.0002890758924474626, "loss": 0.2014, "step": 4870 }, { "epoch": 1.34, "learning_rate": 0.00028904836193447737, "loss": 0.1825, "step": 4880 }, { "epoch": 1.34, "learning_rate": 0.0002890208314214921, "loss": 0.1761, "step": 4890 }, { "epoch": 1.34, "learning_rate": 0.0002889933009085069, "loss": 0.2667, "step": 4900 }, { "epoch": 1.34, "eval_loss": 0.12561985850334167, "eval_runtime": 1032.5452, "eval_samples_per_second": 11.315, "eval_steps_per_second": 1.415, "eval_wer": 0.3635381606476849, "step": 4900 }, { "epoch": 1.35, "learning_rate": 0.00028896577039552165, "loss": 0.1961, "step": 4910 }, { "epoch": 1.35, "learning_rate": 0.00028893823988253645, "loss": 0.4355, "step": 4920 }, { "epoch": 1.35, "learning_rate": 0.00028891070936955124, "loss": 0.1876, "step": 4930 }, { "epoch": 1.35, "learning_rate": 0.00028888317885656604, "loss": 0.182, "step": 4940 }, { "epoch": 1.36, "learning_rate": 0.0002888556483435808, "loss": 0.1972, "step": 4950 }, { "epoch": 1.36, "learning_rate": 0.0002888281178305955, "loss": 0.1987, "step": 4960 }, { "epoch": 1.36, "learning_rate": 0.0002888005873176103, "loss": 0.1898, "step": 4970 }, { "epoch": 1.36, "learning_rate": 0.0002887730568046251, "loss": 0.1877, "step": 4980 }, { "epoch": 1.37, "learning_rate": 0.0002887455262916399, "loss": 0.1746, "step": 4990 }, { "epoch": 1.37, "learning_rate": 0.00028871799577865465, "loss": 0.232, "step": 5000 }, { "epoch": 1.37, "eval_loss": 0.12278987467288971, "eval_runtime": 1033.2314, "eval_samples_per_second": 11.307, "eval_steps_per_second": 1.414, "eval_wer": 0.3580005458018739, "step": 5000 }, { "epoch": 1.37, "learning_rate": 0.0002886904652656694, "loss": 0.1778, "step": 5010 }, { "epoch": 1.38, "learning_rate": 0.0002886629347526842, "loss": 0.1862, "step": 5020 }, { "epoch": 1.38, "learning_rate": 0.000288635404239699, "loss": 0.1933, "step": 5030 }, { "epoch": 1.38, "learning_rate": 0.0002886078737267138, "loss": 0.1667, "step": 5040 }, { "epoch": 1.38, "learning_rate": 0.0002885803432137285, "loss": 0.1822, "step": 5050 }, { "epoch": 1.39, "learning_rate": 0.00028855281270074326, "loss": 0.1534, "step": 5060 }, { "epoch": 1.39, "learning_rate": 0.00028852528218775806, "loss": 0.3339, "step": 5070 }, { "epoch": 1.39, "learning_rate": 0.00028849775167477285, "loss": 0.179, "step": 5080 }, { "epoch": 1.39, "learning_rate": 0.00028847022116178765, "loss": 0.2182, "step": 5090 }, { "epoch": 1.4, "learning_rate": 0.0002884426906488024, "loss": 0.2012, "step": 5100 }, { "epoch": 1.4, "eval_loss": 0.11825747042894363, "eval_runtime": 1037.9251, "eval_samples_per_second": 11.256, "eval_steps_per_second": 1.408, "eval_wer": 0.34755071409078503, "step": 5100 }, { "epoch": 1.4, "learning_rate": 0.00028841516013581713, "loss": 0.1791, "step": 5110 }, { "epoch": 1.4, "learning_rate": 0.00028838762962283193, "loss": 0.1971, "step": 5120 }, { "epoch": 1.41, "learning_rate": 0.0002883600991098467, "loss": 0.1989, "step": 5130 }, { "epoch": 1.41, "learning_rate": 0.0002883325685968615, "loss": 0.2292, "step": 5140 }, { "epoch": 1.41, "learning_rate": 0.00028830503808387626, "loss": 0.166, "step": 5150 }, { "epoch": 1.41, "learning_rate": 0.00028827750757089106, "loss": 0.1829, "step": 5160 }, { "epoch": 1.42, "learning_rate": 0.0002882499770579058, "loss": 0.1925, "step": 5170 }, { "epoch": 1.42, "learning_rate": 0.0002882224465449206, "loss": 0.1866, "step": 5180 }, { "epoch": 1.42, "learning_rate": 0.0002881949160319354, "loss": 0.2301, "step": 5190 }, { "epoch": 1.42, "learning_rate": 0.00028816738551895013, "loss": 0.1862, "step": 5200 }, { "epoch": 1.42, "eval_loss": 0.12098130583763123, "eval_runtime": 1035.7561, "eval_samples_per_second": 11.28, "eval_steps_per_second": 1.411, "eval_wer": 0.3535317929591558, "step": 5200 }, { "epoch": 1.43, "learning_rate": 0.00028813985500596493, "loss": 0.1951, "step": 5210 }, { "epoch": 1.43, "learning_rate": 0.00028811232449297967, "loss": 0.1684, "step": 5220 }, { "epoch": 1.43, "learning_rate": 0.00028808479397999446, "loss": 0.1903, "step": 5230 }, { "epoch": 1.44, "learning_rate": 0.00028805726346700926, "loss": 0.1911, "step": 5240 }, { "epoch": 1.44, "learning_rate": 0.000288029732954024, "loss": 0.1821, "step": 5250 }, { "epoch": 1.44, "learning_rate": 0.0002880022024410388, "loss": 0.2087, "step": 5260 }, { "epoch": 1.44, "learning_rate": 0.00028797467192805354, "loss": 0.1847, "step": 5270 }, { "epoch": 1.45, "learning_rate": 0.00028794714141506833, "loss": 0.2127, "step": 5280 }, { "epoch": 1.45, "learning_rate": 0.00028791961090208313, "loss": 0.1968, "step": 5290 }, { "epoch": 1.45, "learning_rate": 0.0002878920803890979, "loss": 0.1947, "step": 5300 }, { "epoch": 1.45, "eval_loss": 0.12408991158008575, "eval_runtime": 1033.7998, "eval_samples_per_second": 11.301, "eval_steps_per_second": 1.413, "eval_wer": 0.3513940689529701, "step": 5300 }, { "epoch": 1.45, "learning_rate": 0.00028786454987611267, "loss": 0.1778, "step": 5310 }, { "epoch": 1.46, "learning_rate": 0.0002878370193631274, "loss": 0.2122, "step": 5320 }, { "epoch": 1.46, "learning_rate": 0.0002878094888501422, "loss": 0.1956, "step": 5330 }, { "epoch": 1.46, "learning_rate": 0.000287781958337157, "loss": 0.1801, "step": 5340 }, { "epoch": 1.47, "learning_rate": 0.0002877544278241718, "loss": 0.1707, "step": 5350 }, { "epoch": 1.47, "learning_rate": 0.00028772689731118654, "loss": 0.1918, "step": 5360 }, { "epoch": 1.47, "learning_rate": 0.0002876993667982013, "loss": 0.1845, "step": 5370 }, { "epoch": 1.47, "learning_rate": 0.0002876718362852161, "loss": 0.1996, "step": 5380 }, { "epoch": 1.48, "learning_rate": 0.00028764430577223087, "loss": 0.1877, "step": 5390 }, { "epoch": 1.48, "learning_rate": 0.00028761677525924567, "loss": 0.1806, "step": 5400 }, { "epoch": 1.48, "eval_loss": 0.11507146060466766, "eval_runtime": 1039.021, "eval_samples_per_second": 11.244, "eval_steps_per_second": 1.406, "eval_wer": 0.33993222960065494, "step": 5400 }, { "epoch": 1.48, "learning_rate": 0.0002875892447462604, "loss": 0.1864, "step": 5410 }, { "epoch": 1.49, "learning_rate": 0.00028756171423327515, "loss": 0.1771, "step": 5420 }, { "epoch": 1.49, "learning_rate": 0.00028753418372028995, "loss": 0.1856, "step": 5430 }, { "epoch": 1.49, "learning_rate": 0.00028750665320730474, "loss": 0.2536, "step": 5440 }, { "epoch": 1.49, "learning_rate": 0.00028747912269431954, "loss": 0.1733, "step": 5450 }, { "epoch": 1.5, "learning_rate": 0.0002874515921813343, "loss": 0.1679, "step": 5460 }, { "epoch": 1.5, "learning_rate": 0.0002874240616683491, "loss": 0.1943, "step": 5470 }, { "epoch": 1.5, "learning_rate": 0.0002873965311553638, "loss": 0.1878, "step": 5480 }, { "epoch": 1.5, "learning_rate": 0.0002873690006423786, "loss": 0.1777, "step": 5490 }, { "epoch": 1.51, "learning_rate": 0.0002873414701293934, "loss": 0.1779, "step": 5500 }, { "epoch": 1.51, "eval_loss": 0.11272893100976944, "eval_runtime": 1051.3836, "eval_samples_per_second": 11.112, "eval_steps_per_second": 1.39, "eval_wer": 0.3426157554807605, "step": 5500 }, { "epoch": 1.51, "learning_rate": 0.00028731393961640815, "loss": 0.1695, "step": 5510 }, { "epoch": 1.51, "learning_rate": 0.00028728640910342295, "loss": 0.1919, "step": 5520 }, { "epoch": 1.52, "learning_rate": 0.0002872588785904377, "loss": 0.1786, "step": 5530 }, { "epoch": 1.52, "learning_rate": 0.0002872313480774525, "loss": 0.186, "step": 5540 }, { "epoch": 1.52, "learning_rate": 0.0002872038175644673, "loss": 0.1565, "step": 5550 }, { "epoch": 1.52, "learning_rate": 0.000287176287051482, "loss": 0.2021, "step": 5560 }, { "epoch": 1.53, "learning_rate": 0.0002871487565384968, "loss": 0.2016, "step": 5570 }, { "epoch": 1.53, "learning_rate": 0.00028712122602551156, "loss": 0.1918, "step": 5580 }, { "epoch": 1.53, "learning_rate": 0.00028709369551252635, "loss": 0.2019, "step": 5590 }, { "epoch": 1.53, "learning_rate": 0.00028706616499954115, "loss": 0.176, "step": 5600 }, { "epoch": 1.53, "eval_loss": 0.11052325367927551, "eval_runtime": 1047.3933, "eval_samples_per_second": 11.154, "eval_steps_per_second": 1.395, "eval_wer": 0.34463977076321295, "step": 5600 }, { "epoch": 1.54, "learning_rate": 0.00028703863448655594, "loss": 0.1895, "step": 5610 }, { "epoch": 1.54, "learning_rate": 0.0002870111039735707, "loss": 0.1773, "step": 5620 }, { "epoch": 1.54, "learning_rate": 0.00028698357346058543, "loss": 0.1707, "step": 5630 }, { "epoch": 1.55, "learning_rate": 0.0002869560429476002, "loss": 0.1953, "step": 5640 }, { "epoch": 1.55, "learning_rate": 0.000286928512434615, "loss": 0.1857, "step": 5650 }, { "epoch": 1.55, "learning_rate": 0.0002869009819216298, "loss": 0.1813, "step": 5660 }, { "epoch": 1.55, "learning_rate": 0.00028687345140864456, "loss": 0.3278, "step": 5670 }, { "epoch": 1.56, "learning_rate": 0.0002868459208956593, "loss": 0.1934, "step": 5680 }, { "epoch": 1.56, "learning_rate": 0.0002868183903826741, "loss": 0.1889, "step": 5690 }, { "epoch": 1.56, "learning_rate": 0.0002867908598696889, "loss": 0.1387, "step": 5700 }, { "epoch": 1.56, "eval_loss": 0.10185012966394424, "eval_runtime": 1037.8682, "eval_samples_per_second": 11.257, "eval_steps_per_second": 1.408, "eval_wer": 0.33014190848721914, "step": 5700 }, { "epoch": 1.56, "learning_rate": 0.0002867633293567037, "loss": 0.1744, "step": 5710 }, { "epoch": 1.57, "learning_rate": 0.00028673579884371843, "loss": 0.1746, "step": 5720 }, { "epoch": 1.57, "learning_rate": 0.00028670826833073317, "loss": 0.1832, "step": 5730 }, { "epoch": 1.57, "learning_rate": 0.00028668073781774796, "loss": 0.1929, "step": 5740 }, { "epoch": 1.58, "learning_rate": 0.00028665320730476276, "loss": 0.1672, "step": 5750 }, { "epoch": 1.58, "learning_rate": 0.00028662567679177756, "loss": 0.1748, "step": 5760 }, { "epoch": 1.58, "learning_rate": 0.0002865981462787923, "loss": 0.1814, "step": 5770 }, { "epoch": 1.58, "learning_rate": 0.00028657061576580704, "loss": 0.1771, "step": 5780 }, { "epoch": 1.59, "learning_rate": 0.00028654308525282183, "loss": 0.1724, "step": 5790 }, { "epoch": 1.59, "learning_rate": 0.00028651555473983663, "loss": 0.1621, "step": 5800 }, { "epoch": 1.59, "eval_loss": 0.10375826060771942, "eval_runtime": 1035.0749, "eval_samples_per_second": 11.287, "eval_steps_per_second": 1.411, "eval_wer": 0.33085827344673885, "step": 5800 }, { "epoch": 1.59, "learning_rate": 0.0002864880242268514, "loss": 0.1839, "step": 5810 }, { "epoch": 1.59, "learning_rate": 0.00028646049371386617, "loss": 0.1961, "step": 5820 }, { "epoch": 1.6, "learning_rate": 0.00028643296320088096, "loss": 0.1632, "step": 5830 }, { "epoch": 1.6, "learning_rate": 0.0002864054326878957, "loss": 0.1617, "step": 5840 }, { "epoch": 1.6, "learning_rate": 0.0002863779021749105, "loss": 0.1951, "step": 5850 }, { "epoch": 1.61, "learning_rate": 0.0002863503716619253, "loss": 0.166, "step": 5860 }, { "epoch": 1.61, "learning_rate": 0.00028632284114894004, "loss": 0.1689, "step": 5870 }, { "epoch": 1.61, "learning_rate": 0.00028629531063595483, "loss": 0.1943, "step": 5880 }, { "epoch": 1.61, "learning_rate": 0.0002862677801229696, "loss": 0.1775, "step": 5890 }, { "epoch": 1.62, "learning_rate": 0.00028624024960998437, "loss": 0.1527, "step": 5900 }, { "epoch": 1.62, "eval_loss": 0.11036413908004761, "eval_runtime": 1034.9926, "eval_samples_per_second": 11.288, "eval_steps_per_second": 1.412, "eval_wer": 0.33977303738742837, "step": 5900 }, { "epoch": 1.62, "learning_rate": 0.00028621271909699917, "loss": 0.2039, "step": 5910 }, { "epoch": 1.62, "learning_rate": 0.0002861851885840139, "loss": 0.1963, "step": 5920 }, { "epoch": 1.62, "learning_rate": 0.0002861576580710287, "loss": 0.1781, "step": 5930 }, { "epoch": 1.63, "learning_rate": 0.00028613012755804345, "loss": 0.1761, "step": 5940 }, { "epoch": 1.63, "learning_rate": 0.00028610259704505824, "loss": 0.1508, "step": 5950 }, { "epoch": 1.63, "learning_rate": 0.00028607506653207304, "loss": 0.2437, "step": 5960 }, { "epoch": 1.64, "learning_rate": 0.00028604753601908783, "loss": 0.1787, "step": 5970 }, { "epoch": 1.64, "learning_rate": 0.0002860200055061026, "loss": 0.1921, "step": 5980 }, { "epoch": 1.64, "learning_rate": 0.0002859924749931173, "loss": 0.1984, "step": 5990 }, { "epoch": 1.64, "learning_rate": 0.0002859649444801321, "loss": 0.1748, "step": 6000 }, { "epoch": 1.64, "eval_loss": 0.10706117004156113, "eval_runtime": 1037.1354, "eval_samples_per_second": 11.265, "eval_steps_per_second": 1.409, "eval_wer": 0.3348267079050305, "step": 6000 }, { "epoch": 1.65, "learning_rate": 0.0002859374139671469, "loss": 0.2553, "step": 6010 }, { "epoch": 1.65, "learning_rate": 0.0002859098834541617, "loss": 0.1911, "step": 6020 }, { "epoch": 1.65, "learning_rate": 0.00028588235294117645, "loss": 0.1553, "step": 6030 }, { "epoch": 1.66, "learning_rate": 0.0002858548224281912, "loss": 0.1948, "step": 6040 }, { "epoch": 1.66, "learning_rate": 0.000285827291915206, "loss": 0.2145, "step": 6050 }, { "epoch": 1.66, "learning_rate": 0.0002857997614022208, "loss": 0.1577, "step": 6060 }, { "epoch": 1.66, "learning_rate": 0.0002857722308892356, "loss": 0.1695, "step": 6070 }, { "epoch": 1.67, "learning_rate": 0.0002857447003762503, "loss": 0.2078, "step": 6080 }, { "epoch": 1.67, "learning_rate": 0.00028571716986326506, "loss": 0.2078, "step": 6090 }, { "epoch": 1.67, "learning_rate": 0.00028568963935027985, "loss": 0.165, "step": 6100 }, { "epoch": 1.67, "eval_loss": 0.10916991531848907, "eval_runtime": 1042.7945, "eval_samples_per_second": 11.204, "eval_steps_per_second": 1.401, "eval_wer": 0.33797643955244244, "step": 6100 }, { "epoch": 1.67, "learning_rate": 0.00028566210883729465, "loss": 0.1749, "step": 6110 }, { "epoch": 1.68, "learning_rate": 0.00028563457832430944, "loss": 0.165, "step": 6120 }, { "epoch": 1.68, "learning_rate": 0.0002856070478113242, "loss": 0.1728, "step": 6130 }, { "epoch": 1.68, "learning_rate": 0.000285579517298339, "loss": 0.1732, "step": 6140 }, { "epoch": 1.69, "learning_rate": 0.0002855519867853537, "loss": 0.1662, "step": 6150 }, { "epoch": 1.69, "learning_rate": 0.0002855244562723685, "loss": 0.1689, "step": 6160 }, { "epoch": 1.69, "learning_rate": 0.0002854969257593833, "loss": 0.1774, "step": 6170 }, { "epoch": 1.69, "learning_rate": 0.00028546939524639806, "loss": 0.1691, "step": 6180 }, { "epoch": 1.7, "learning_rate": 0.00028544186473341285, "loss": 0.176, "step": 6190 }, { "epoch": 1.7, "learning_rate": 0.0002854143342204276, "loss": 0.1384, "step": 6200 }, { "epoch": 1.7, "eval_loss": 0.10628872364759445, "eval_runtime": 1039.2196, "eval_samples_per_second": 11.242, "eval_steps_per_second": 1.406, "eval_wer": 0.3349404166287638, "step": 6200 }, { "epoch": 1.7, "learning_rate": 0.0002853868037074424, "loss": 0.2081, "step": 6210 }, { "epoch": 1.7, "learning_rate": 0.0002853592731944572, "loss": 0.1934, "step": 6220 }, { "epoch": 1.71, "learning_rate": 0.00028533174268147193, "loss": 0.1613, "step": 6230 }, { "epoch": 1.71, "learning_rate": 0.0002853042121684867, "loss": 0.1717, "step": 6240 }, { "epoch": 1.71, "learning_rate": 0.00028527668165550146, "loss": 0.1603, "step": 6250 }, { "epoch": 1.72, "learning_rate": 0.00028524915114251626, "loss": 0.1749, "step": 6260 }, { "epoch": 1.72, "learning_rate": 0.00028522162062953106, "loss": 0.1565, "step": 6270 }, { "epoch": 1.72, "learning_rate": 0.00028519409011654585, "loss": 0.1862, "step": 6280 }, { "epoch": 1.72, "learning_rate": 0.0002851665596035606, "loss": 0.164, "step": 6290 }, { "epoch": 1.73, "learning_rate": 0.00028513902909057534, "loss": 0.1859, "step": 6300 }, { "epoch": 1.73, "eval_loss": 0.10580063611268997, "eval_runtime": 1053.9327, "eval_samples_per_second": 11.085, "eval_steps_per_second": 1.386, "eval_wer": 0.32495679068498134, "step": 6300 }, { "epoch": 1.73, "learning_rate": 0.00028511149857759013, "loss": 0.1775, "step": 6310 }, { "epoch": 1.73, "learning_rate": 0.0002850839680646049, "loss": 0.1876, "step": 6320 }, { "epoch": 1.73, "learning_rate": 0.0002850564375516197, "loss": 0.1628, "step": 6330 }, { "epoch": 1.74, "learning_rate": 0.00028502890703863446, "loss": 0.1733, "step": 6340 }, { "epoch": 1.74, "learning_rate": 0.0002850013765256492, "loss": 0.1906, "step": 6350 }, { "epoch": 1.74, "learning_rate": 0.000284973846012664, "loss": 0.1974, "step": 6360 }, { "epoch": 1.75, "learning_rate": 0.0002849463154996788, "loss": 0.1655, "step": 6370 }, { "epoch": 1.75, "learning_rate": 0.0002849187849866936, "loss": 0.1646, "step": 6380 }, { "epoch": 1.75, "learning_rate": 0.00028489125447370833, "loss": 0.1941, "step": 6390 }, { "epoch": 1.75, "learning_rate": 0.0002848637239607231, "loss": 0.1919, "step": 6400 }, { "epoch": 1.75, "eval_loss": 0.1080482006072998, "eval_runtime": 1063.6328, "eval_samples_per_second": 10.984, "eval_steps_per_second": 1.374, "eval_wer": 0.34133084690257437, "step": 6400 }, { "epoch": 1.76, "learning_rate": 0.00028483619344773787, "loss": 0.1714, "step": 6410 }, { "epoch": 1.76, "learning_rate": 0.00028480866293475267, "loss": 0.1509, "step": 6420 }, { "epoch": 1.76, "learning_rate": 0.00028478113242176746, "loss": 0.1878, "step": 6430 }, { "epoch": 1.76, "learning_rate": 0.0002847536019087822, "loss": 0.1855, "step": 6440 }, { "epoch": 1.77, "learning_rate": 0.00028472607139579695, "loss": 0.1743, "step": 6450 }, { "epoch": 1.77, "learning_rate": 0.00028469854088281174, "loss": 0.1612, "step": 6460 }, { "epoch": 1.77, "learning_rate": 0.00028467101036982654, "loss": 0.1659, "step": 6470 }, { "epoch": 1.78, "learning_rate": 0.00028464347985684133, "loss": 0.1824, "step": 6480 }, { "epoch": 1.78, "learning_rate": 0.0002846159493438561, "loss": 0.1656, "step": 6490 }, { "epoch": 1.78, "learning_rate": 0.00028458841883087087, "loss": 0.1425, "step": 6500 }, { "epoch": 1.78, "eval_loss": 0.10503587871789932, "eval_runtime": 1052.5362, "eval_samples_per_second": 11.1, "eval_steps_per_second": 1.388, "eval_wer": 0.3284135358864732, "step": 6500 }, { "epoch": 1.78, "learning_rate": 0.0002845608883178856, "loss": 0.1625, "step": 6510 }, { "epoch": 1.79, "learning_rate": 0.0002845333578049004, "loss": 0.1626, "step": 6520 }, { "epoch": 1.79, "learning_rate": 0.0002845058272919152, "loss": 0.1669, "step": 6530 }, { "epoch": 1.79, "learning_rate": 0.00028447829677892995, "loss": 0.1951, "step": 6540 }, { "epoch": 1.79, "learning_rate": 0.00028445076626594474, "loss": 0.1674, "step": 6550 }, { "epoch": 1.8, "learning_rate": 0.0002844232357529595, "loss": 0.1615, "step": 6560 }, { "epoch": 1.8, "learning_rate": 0.0002843957052399743, "loss": 0.1559, "step": 6570 }, { "epoch": 1.8, "learning_rate": 0.0002843681747269891, "loss": 0.1697, "step": 6580 }, { "epoch": 1.81, "learning_rate": 0.00028434064421400387, "loss": 0.186, "step": 6590 }, { "epoch": 1.81, "learning_rate": 0.0002843131137010186, "loss": 0.1571, "step": 6600 }, { "epoch": 1.81, "eval_loss": 0.10015376657247543, "eval_runtime": 1050.3185, "eval_samples_per_second": 11.123, "eval_steps_per_second": 1.391, "eval_wer": 0.32436550532156827, "step": 6600 }, { "epoch": 1.81, "learning_rate": 0.00028428558318803335, "loss": 0.1643, "step": 6610 }, { "epoch": 1.81, "learning_rate": 0.00028425805267504815, "loss": 0.1636, "step": 6620 }, { "epoch": 1.82, "learning_rate": 0.00028423052216206294, "loss": 0.1646, "step": 6630 }, { "epoch": 1.82, "learning_rate": 0.00028420299164907774, "loss": 0.1628, "step": 6640 }, { "epoch": 1.82, "learning_rate": 0.0002841754611360925, "loss": 0.1288, "step": 6650 }, { "epoch": 1.82, "learning_rate": 0.0002841479306231072, "loss": 0.1523, "step": 6660 }, { "epoch": 1.83, "learning_rate": 0.000284120400110122, "loss": 0.1646, "step": 6670 }, { "epoch": 1.83, "learning_rate": 0.0002840928695971368, "loss": 0.1748, "step": 6680 }, { "epoch": 1.83, "learning_rate": 0.0002840653390841516, "loss": 0.164, "step": 6690 }, { "epoch": 1.84, "learning_rate": 0.00028403780857116635, "loss": 0.1591, "step": 6700 }, { "epoch": 1.84, "eval_loss": 0.09989971667528152, "eval_runtime": 1047.7519, "eval_samples_per_second": 11.151, "eval_steps_per_second": 1.394, "eval_wer": 0.3237742199581552, "step": 6700 }, { "epoch": 1.84, "learning_rate": 0.0002840102780581811, "loss": 0.1732, "step": 6710 }, { "epoch": 1.84, "learning_rate": 0.0002839827475451959, "loss": 0.1573, "step": 6720 }, { "epoch": 1.84, "learning_rate": 0.0002839552170322107, "loss": 0.1639, "step": 6730 }, { "epoch": 1.85, "learning_rate": 0.0002839276865192255, "loss": 0.1808, "step": 6740 }, { "epoch": 1.85, "learning_rate": 0.0002839001560062402, "loss": 0.1593, "step": 6750 }, { "epoch": 1.85, "learning_rate": 0.00028387262549325496, "loss": 0.168, "step": 6760 }, { "epoch": 1.86, "learning_rate": 0.00028384509498026976, "loss": 0.1756, "step": 6770 }, { "epoch": 1.86, "learning_rate": 0.00028381756446728456, "loss": 0.163, "step": 6780 }, { "epoch": 1.86, "learning_rate": 0.00028379003395429935, "loss": 0.1771, "step": 6790 }, { "epoch": 1.86, "learning_rate": 0.00028376525649261264, "loss": 0.1615, "step": 6800 }, { "epoch": 1.86, "eval_loss": 0.09614775329828262, "eval_runtime": 1045.8287, "eval_samples_per_second": 11.171, "eval_steps_per_second": 1.397, "eval_wer": 0.3156781588283453, "step": 6800 }, { "epoch": 1.87, "learning_rate": 0.0002837377259796274, "loss": 0.1792, "step": 6810 }, { "epoch": 1.87, "learning_rate": 0.0002837101954666422, "loss": 0.1612, "step": 6820 }, { "epoch": 1.87, "learning_rate": 0.00028368266495365697, "loss": 0.1664, "step": 6830 }, { "epoch": 1.87, "learning_rate": 0.0002836551344406717, "loss": 0.1894, "step": 6840 }, { "epoch": 1.88, "learning_rate": 0.0002836276039276865, "loss": 0.1345, "step": 6850 }, { "epoch": 1.88, "learning_rate": 0.00028360007341470125, "loss": 0.1623, "step": 6860 }, { "epoch": 1.88, "learning_rate": 0.00028357254290171604, "loss": 0.1799, "step": 6870 }, { "epoch": 1.89, "learning_rate": 0.00028354501238873084, "loss": 0.1759, "step": 6880 }, { "epoch": 1.89, "learning_rate": 0.0002835174818757456, "loss": 0.1627, "step": 6890 }, { "epoch": 1.89, "learning_rate": 0.0002834899513627604, "loss": 0.148, "step": 6900 }, { "epoch": 1.89, "eval_loss": 0.09930101782083511, "eval_runtime": 1069.3227, "eval_samples_per_second": 10.926, "eval_steps_per_second": 1.366, "eval_wer": 0.3197489311379969, "step": 6900 }, { "epoch": 1.89, "learning_rate": 0.0002834624208497751, "loss": 0.209, "step": 6910 }, { "epoch": 1.9, "learning_rate": 0.0002834348903367899, "loss": 0.1594, "step": 6920 }, { "epoch": 1.9, "learning_rate": 0.0002834073598238047, "loss": 0.1598, "step": 6930 }, { "epoch": 1.9, "learning_rate": 0.00028337982931081945, "loss": 0.1674, "step": 6940 }, { "epoch": 1.9, "learning_rate": 0.00028335229879783425, "loss": 0.1648, "step": 6950 }, { "epoch": 1.91, "learning_rate": 0.000283324768284849, "loss": 0.1673, "step": 6960 }, { "epoch": 1.91, "learning_rate": 0.0002832972377718638, "loss": 0.1777, "step": 6970 }, { "epoch": 1.91, "learning_rate": 0.0002832697072588786, "loss": 0.1568, "step": 6980 }, { "epoch": 1.92, "learning_rate": 0.0002832421767458934, "loss": 0.1883, "step": 6990 }, { "epoch": 1.92, "learning_rate": 0.0002832146462329081, "loss": 0.1893, "step": 7000 }, { "epoch": 1.92, "eval_loss": 0.09632071852684021, "eval_runtime": 1046.0027, "eval_samples_per_second": 11.169, "eval_steps_per_second": 1.397, "eval_wer": 0.32194350950604933, "step": 7000 }, { "epoch": 1.92, "learning_rate": 0.00028318711571992286, "loss": 0.1766, "step": 7010 }, { "epoch": 1.92, "learning_rate": 0.00028315958520693765, "loss": 0.1697, "step": 7020 }, { "epoch": 1.93, "learning_rate": 0.00028313205469395245, "loss": 0.1723, "step": 7030 }, { "epoch": 1.93, "learning_rate": 0.00028310452418096725, "loss": 0.1574, "step": 7040 }, { "epoch": 1.93, "learning_rate": 0.000283076993667982, "loss": 0.1456, "step": 7050 }, { "epoch": 1.93, "learning_rate": 0.0002830494631549968, "loss": 0.1686, "step": 7060 }, { "epoch": 1.94, "learning_rate": 0.0002830219326420115, "loss": 0.1706, "step": 7070 }, { "epoch": 1.94, "learning_rate": 0.0002829944021290263, "loss": 0.1645, "step": 7080 }, { "epoch": 1.94, "learning_rate": 0.0002829668716160411, "loss": 0.1866, "step": 7090 }, { "epoch": 1.95, "learning_rate": 0.00028293934110305586, "loss": 0.1586, "step": 7100 }, { "epoch": 1.95, "eval_loss": 0.0969810038805008, "eval_runtime": 1048.7511, "eval_samples_per_second": 11.14, "eval_steps_per_second": 1.393, "eval_wer": 0.3227394705721823, "step": 7100 }, { "epoch": 1.95, "learning_rate": 0.00028291181059007065, "loss": 0.1577, "step": 7110 }, { "epoch": 1.95, "learning_rate": 0.0002828842800770854, "loss": 0.1716, "step": 7120 }, { "epoch": 1.95, "learning_rate": 0.0002828567495641002, "loss": 0.1957, "step": 7130 }, { "epoch": 1.96, "learning_rate": 0.000282829219051115, "loss": 0.1704, "step": 7140 }, { "epoch": 1.96, "learning_rate": 0.00028280168853812973, "loss": 0.1522, "step": 7150 }, { "epoch": 1.96, "learning_rate": 0.0002827741580251445, "loss": 0.1692, "step": 7160 }, { "epoch": 1.96, "learning_rate": 0.00028274662751215927, "loss": 0.2043, "step": 7170 }, { "epoch": 1.97, "learning_rate": 0.00028271909699917406, "loss": 0.1745, "step": 7180 }, { "epoch": 1.97, "learning_rate": 0.00028269156648618886, "loss": 0.1994, "step": 7190 }, { "epoch": 1.97, "learning_rate": 0.0002826640359732036, "loss": 0.1932, "step": 7200 }, { "epoch": 1.97, "eval_loss": 0.09549184143543243, "eval_runtime": 1063.1244, "eval_samples_per_second": 10.989, "eval_steps_per_second": 1.374, "eval_wer": 0.3200104612025835, "step": 7200 }, { "epoch": 1.98, "learning_rate": 0.0002826365054602184, "loss": 0.1741, "step": 7210 }, { "epoch": 1.98, "learning_rate": 0.00028260897494723314, "loss": 0.1612, "step": 7220 }, { "epoch": 1.98, "learning_rate": 0.00028258144443424793, "loss": 0.1632, "step": 7230 }, { "epoch": 1.98, "learning_rate": 0.00028255391392126273, "loss": 0.1843, "step": 7240 }, { "epoch": 1.99, "learning_rate": 0.00028252638340827747, "loss": 0.1714, "step": 7250 }, { "epoch": 1.99, "learning_rate": 0.00028249885289529227, "loss": 0.1662, "step": 7260 }, { "epoch": 1.99, "learning_rate": 0.000282471322382307, "loss": 0.1622, "step": 7270 }, { "epoch": 1.99, "learning_rate": 0.0002824437918693218, "loss": 0.1642, "step": 7280 }, { "epoch": 2.0, "learning_rate": 0.0002824162613563366, "loss": 0.2103, "step": 7290 }, { "epoch": 2.0, "learning_rate": 0.00028238873084335134, "loss": 0.172, "step": 7300 }, { "epoch": 2.0, "eval_loss": 0.09298280626535416, "eval_runtime": 1067.22, "eval_samples_per_second": 10.947, "eval_steps_per_second": 1.369, "eval_wer": 0.3159169471481852, "step": 7300 }, { "epoch": 2.0, "learning_rate": 0.00028236120033036614, "loss": 0.1545, "step": 7310 }, { "epoch": 2.01, "learning_rate": 0.0002823336698173809, "loss": 0.1518, "step": 7320 }, { "epoch": 2.01, "learning_rate": 0.0002823061393043957, "loss": 0.1638, "step": 7330 }, { "epoch": 2.01, "learning_rate": 0.00028227860879141047, "loss": 0.3504, "step": 7340 }, { "epoch": 2.01, "learning_rate": 0.0002822510782784252, "loss": 0.1497, "step": 7350 }, { "epoch": 2.02, "learning_rate": 0.00028222354776544, "loss": 0.204, "step": 7360 }, { "epoch": 2.02, "learning_rate": 0.0002821960172524548, "loss": 0.1577, "step": 7370 }, { "epoch": 2.02, "learning_rate": 0.00028216848673946954, "loss": 0.1677, "step": 7380 }, { "epoch": 2.02, "learning_rate": 0.00028214095622648434, "loss": 0.1653, "step": 7390 }, { "epoch": 2.03, "learning_rate": 0.00028211342571349913, "loss": 0.1617, "step": 7400 }, { "epoch": 2.03, "eval_loss": 0.09724044799804688, "eval_runtime": 1067.6814, "eval_samples_per_second": 10.942, "eval_steps_per_second": 1.368, "eval_wer": 0.3169630674065314, "step": 7400 }, { "epoch": 2.03, "learning_rate": 0.0002820858952005139, "loss": 0.1405, "step": 7410 }, { "epoch": 2.03, "learning_rate": 0.00028205836468752867, "loss": 0.1483, "step": 7420 }, { "epoch": 2.04, "learning_rate": 0.0002820308341745434, "loss": 0.1764, "step": 7430 }, { "epoch": 2.04, "learning_rate": 0.0002820033036615582, "loss": 0.1624, "step": 7440 }, { "epoch": 2.04, "learning_rate": 0.000281975773148573, "loss": 0.1766, "step": 7450 }, { "epoch": 2.04, "learning_rate": 0.00028194824263558775, "loss": 0.1594, "step": 7460 }, { "epoch": 2.05, "learning_rate": 0.00028192071212260254, "loss": 0.1473, "step": 7470 }, { "epoch": 2.05, "learning_rate": 0.0002818931816096173, "loss": 0.1403, "step": 7480 }, { "epoch": 2.05, "learning_rate": 0.0002818656510966321, "loss": 0.1618, "step": 7490 }, { "epoch": 2.06, "learning_rate": 0.0002818381205836469, "loss": 0.1644, "step": 7500 }, { "epoch": 2.06, "eval_loss": 0.0922645702958107, "eval_runtime": 1046.6265, "eval_samples_per_second": 11.163, "eval_steps_per_second": 1.396, "eval_wer": 0.3116983534976803, "step": 7500 }, { "epoch": 2.06, "learning_rate": 0.0002818105900706616, "loss": 0.1947, "step": 7510 }, { "epoch": 2.06, "learning_rate": 0.0002817830595576764, "loss": 0.4217, "step": 7520 }, { "epoch": 2.06, "learning_rate": 0.00028175552904469115, "loss": 0.1694, "step": 7530 }, { "epoch": 2.07, "learning_rate": 0.00028172799853170595, "loss": 0.1781, "step": 7540 }, { "epoch": 2.07, "learning_rate": 0.00028170046801872075, "loss": 0.1502, "step": 7550 }, { "epoch": 2.07, "learning_rate": 0.0002816729375057355, "loss": 0.1733, "step": 7560 }, { "epoch": 2.07, "learning_rate": 0.0002816454069927503, "loss": 0.1506, "step": 7570 }, { "epoch": 2.08, "learning_rate": 0.000281617876479765, "loss": 0.1675, "step": 7580 }, { "epoch": 2.08, "learning_rate": 0.0002815903459667798, "loss": 0.1532, "step": 7590 }, { "epoch": 2.08, "learning_rate": 0.0002815628154537946, "loss": 0.1098, "step": 7600 }, { "epoch": 2.08, "eval_loss": 0.10008783638477325, "eval_runtime": 1043.3579, "eval_samples_per_second": 11.197, "eval_steps_per_second": 1.4, "eval_wer": 0.31861184390066405, "step": 7600 }, { "epoch": 2.09, "learning_rate": 0.00028153528494080936, "loss": 0.1547, "step": 7610 }, { "epoch": 2.09, "learning_rate": 0.00028150775442782415, "loss": 0.1447, "step": 7620 }, { "epoch": 2.09, "learning_rate": 0.0002814802239148389, "loss": 0.1553, "step": 7630 }, { "epoch": 2.09, "learning_rate": 0.0002814526934018537, "loss": 0.156, "step": 7640 }, { "epoch": 2.1, "learning_rate": 0.0002814251628888685, "loss": 0.1499, "step": 7650 }, { "epoch": 2.1, "learning_rate": 0.00028139763237588323, "loss": 0.2971, "step": 7660 }, { "epoch": 2.1, "learning_rate": 0.000281370101862898, "loss": 0.1428, "step": 7670 }, { "epoch": 2.1, "learning_rate": 0.0002813425713499128, "loss": 0.1695, "step": 7680 }, { "epoch": 2.11, "learning_rate": 0.00028131504083692756, "loss": 0.1704, "step": 7690 }, { "epoch": 2.11, "learning_rate": 0.00028128751032394236, "loss": 0.1456, "step": 7700 }, { "epoch": 2.11, "eval_loss": 0.10283832997083664, "eval_runtime": 1053.1465, "eval_samples_per_second": 11.093, "eval_steps_per_second": 1.387, "eval_wer": 0.32621895751842084, "step": 7700 }, { "epoch": 2.11, "learning_rate": 0.0002812599798109571, "loss": 0.1464, "step": 7710 }, { "epoch": 2.12, "learning_rate": 0.0002812324492979719, "loss": 0.1493, "step": 7720 }, { "epoch": 2.12, "learning_rate": 0.0002812049187849867, "loss": 0.161, "step": 7730 }, { "epoch": 2.12, "learning_rate": 0.00028117738827200143, "loss": 0.1637, "step": 7740 }, { "epoch": 2.12, "learning_rate": 0.00028114985775901623, "loss": 0.146, "step": 7750 }, { "epoch": 2.13, "learning_rate": 0.000281122327246031, "loss": 0.1448, "step": 7760 }, { "epoch": 2.13, "learning_rate": 0.00028109479673304577, "loss": 0.1378, "step": 7770 }, { "epoch": 2.13, "learning_rate": 0.00028106726622006056, "loss": 0.164, "step": 7780 }, { "epoch": 2.13, "learning_rate": 0.0002810397357070753, "loss": 0.1681, "step": 7790 }, { "epoch": 2.14, "learning_rate": 0.0002810122051940901, "loss": 0.1518, "step": 7800 }, { "epoch": 2.14, "eval_loss": 0.10179836302995682, "eval_runtime": 1062.7252, "eval_samples_per_second": 10.993, "eval_steps_per_second": 1.375, "eval_wer": 0.3176680614936778, "step": 7800 }, { "epoch": 2.14, "learning_rate": 0.0002809846746811049, "loss": 0.1422, "step": 7810 }, { "epoch": 2.14, "learning_rate": 0.00028095714416811964, "loss": 0.1459, "step": 7820 }, { "epoch": 2.15, "learning_rate": 0.00028092961365513443, "loss": 0.1555, "step": 7830 }, { "epoch": 2.15, "learning_rate": 0.0002809020831421492, "loss": 0.1484, "step": 7840 }, { "epoch": 2.15, "learning_rate": 0.00028087730568046246, "loss": 0.1217, "step": 7850 }, { "epoch": 2.15, "learning_rate": 0.00028084977516747725, "loss": 0.1407, "step": 7860 }, { "epoch": 2.16, "learning_rate": 0.00028082224465449205, "loss": 0.1494, "step": 7870 }, { "epoch": 2.16, "learning_rate": 0.0002807947141415068, "loss": 0.1643, "step": 7880 }, { "epoch": 2.16, "learning_rate": 0.0002807671836285216, "loss": 0.1696, "step": 7890 }, { "epoch": 2.16, "learning_rate": 0.0002807396531155364, "loss": 0.1485, "step": 7900 }, { "epoch": 2.16, "eval_loss": 0.09863564372062683, "eval_runtime": 1050.6891, "eval_samples_per_second": 11.119, "eval_steps_per_second": 1.391, "eval_wer": 0.3153256617847721, "step": 7900 }, { "epoch": 2.17, "learning_rate": 0.0002807121226025511, "loss": 0.1462, "step": 7910 }, { "epoch": 2.17, "learning_rate": 0.0002806845920895659, "loss": 0.1486, "step": 7920 }, { "epoch": 2.17, "learning_rate": 0.0002806570615765807, "loss": 0.172, "step": 7930 }, { "epoch": 2.18, "learning_rate": 0.00028062953106359546, "loss": 0.1719, "step": 7940 }, { "epoch": 2.18, "learning_rate": 0.00028060200055061025, "loss": 0.1477, "step": 7950 }, { "epoch": 2.18, "learning_rate": 0.000280574470037625, "loss": 0.1478, "step": 7960 }, { "epoch": 2.18, "learning_rate": 0.0002805469395246398, "loss": 0.1693, "step": 7970 }, { "epoch": 2.19, "learning_rate": 0.0002805194090116546, "loss": 0.1479, "step": 7980 }, { "epoch": 2.19, "learning_rate": 0.0002804918784986693, "loss": 0.1835, "step": 7990 }, { "epoch": 2.19, "learning_rate": 0.0002804643479856841, "loss": 0.1411, "step": 8000 }, { "epoch": 2.19, "eval_loss": 0.09377647936344147, "eval_runtime": 1071.9906, "eval_samples_per_second": 10.898, "eval_steps_per_second": 1.363, "eval_wer": 0.31165287000818703, "step": 8000 }, { "epoch": 2.19, "learning_rate": 0.00028043681747269886, "loss": 0.1486, "step": 8010 }, { "epoch": 2.2, "learning_rate": 0.00028040928695971366, "loss": 0.1452, "step": 8020 }, { "epoch": 2.2, "learning_rate": 0.00028038175644672846, "loss": 0.1612, "step": 8030 }, { "epoch": 2.2, "learning_rate": 0.0002803542259337432, "loss": 0.1725, "step": 8040 }, { "epoch": 2.21, "learning_rate": 0.000280326695420758, "loss": 0.1436, "step": 8050 }, { "epoch": 2.21, "learning_rate": 0.00028029916490777273, "loss": 0.1561, "step": 8060 }, { "epoch": 2.21, "learning_rate": 0.00028027163439478753, "loss": 0.1562, "step": 8070 }, { "epoch": 2.21, "learning_rate": 0.0002802441038818023, "loss": 0.1799, "step": 8080 }, { "epoch": 2.22, "learning_rate": 0.00028021657336881707, "loss": 0.1425, "step": 8090 }, { "epoch": 2.22, "learning_rate": 0.00028018904285583186, "loss": 0.1439, "step": 8100 }, { "epoch": 2.22, "eval_loss": 0.09374161064624786, "eval_runtime": 1057.5136, "eval_samples_per_second": 11.048, "eval_steps_per_second": 1.382, "eval_wer": 0.31238060584008004, "step": 8100 }, { "epoch": 2.22, "learning_rate": 0.0002801615123428466, "loss": 0.161, "step": 8110 }, { "epoch": 2.22, "learning_rate": 0.0002801339818298614, "loss": 0.1245, "step": 8120 }, { "epoch": 2.23, "learning_rate": 0.0002801064513168762, "loss": 0.1461, "step": 8130 }, { "epoch": 2.23, "learning_rate": 0.00028007892080389094, "loss": 0.1509, "step": 8140 }, { "epoch": 2.23, "learning_rate": 0.00028005139029090573, "loss": 0.1579, "step": 8150 }, { "epoch": 2.24, "learning_rate": 0.0002800238597779205, "loss": 0.1496, "step": 8160 }, { "epoch": 2.24, "learning_rate": 0.00027999632926493527, "loss": 0.1488, "step": 8170 }, { "epoch": 2.24, "learning_rate": 0.00027996879875195007, "loss": 0.1413, "step": 8180 }, { "epoch": 2.24, "learning_rate": 0.0002799412682389648, "loss": 0.1775, "step": 8190 }, { "epoch": 2.25, "learning_rate": 0.0002799137377259796, "loss": 0.1469, "step": 8200 }, { "epoch": 2.25, "eval_loss": 0.0919284000992775, "eval_runtime": 1040.1265, "eval_samples_per_second": 11.232, "eval_steps_per_second": 1.405, "eval_wer": 0.3096402255981079, "step": 8200 }, { "epoch": 2.25, "learning_rate": 0.00027988620721299435, "loss": 0.1613, "step": 8210 }, { "epoch": 2.25, "learning_rate": 0.00027985867670000914, "loss": 0.1405, "step": 8220 }, { "epoch": 2.26, "learning_rate": 0.00027983114618702394, "loss": 0.1403, "step": 8230 }, { "epoch": 2.26, "learning_rate": 0.00027980361567403873, "loss": 0.1688, "step": 8240 }, { "epoch": 2.26, "learning_rate": 0.0002797760851610535, "loss": 0.1495, "step": 8250 }, { "epoch": 2.26, "learning_rate": 0.0002797485546480682, "loss": 0.146, "step": 8260 }, { "epoch": 2.27, "learning_rate": 0.000279721024135083, "loss": 0.169, "step": 8270 }, { "epoch": 2.27, "learning_rate": 0.0002796934936220978, "loss": 0.1583, "step": 8280 }, { "epoch": 2.27, "learning_rate": 0.0002796659631091126, "loss": 0.1608, "step": 8290 }, { "epoch": 2.27, "learning_rate": 0.00027963843259612734, "loss": 0.1476, "step": 8300 }, { "epoch": 2.27, "eval_loss": 0.09394428133964539, "eval_runtime": 1047.8186, "eval_samples_per_second": 11.15, "eval_steps_per_second": 1.394, "eval_wer": 0.30986764304557446, "step": 8300 }, { "epoch": 2.28, "learning_rate": 0.00027961090208314214, "loss": 0.154, "step": 8310 }, { "epoch": 2.28, "learning_rate": 0.0002795833715701569, "loss": 0.1406, "step": 8320 }, { "epoch": 2.28, "learning_rate": 0.0002795558410571717, "loss": 0.1667, "step": 8330 }, { "epoch": 2.29, "learning_rate": 0.0002795283105441865, "loss": 0.1622, "step": 8340 }, { "epoch": 2.29, "learning_rate": 0.0002795007800312012, "loss": 0.161, "step": 8350 }, { "epoch": 2.29, "learning_rate": 0.000279473249518216, "loss": 0.1515, "step": 8360 }, { "epoch": 2.29, "learning_rate": 0.00027944571900523075, "loss": 0.1702, "step": 8370 }, { "epoch": 2.3, "learning_rate": 0.00027941818849224555, "loss": 0.1435, "step": 8380 }, { "epoch": 2.3, "learning_rate": 0.00027939065797926034, "loss": 0.138, "step": 8390 }, { "epoch": 2.3, "learning_rate": 0.0002793631274662751, "loss": 0.1678, "step": 8400 }, { "epoch": 2.3, "eval_loss": 0.0926092192530632, "eval_runtime": 1050.1998, "eval_samples_per_second": 11.125, "eval_steps_per_second": 1.391, "eval_wer": 0.30953788774674795, "step": 8400 }, { "epoch": 2.3, "learning_rate": 0.0002793355969532899, "loss": 0.1456, "step": 8410 }, { "epoch": 2.31, "learning_rate": 0.0002793080664403046, "loss": 0.1874, "step": 8420 }, { "epoch": 2.31, "learning_rate": 0.0002792805359273194, "loss": 0.1674, "step": 8430 }, { "epoch": 2.31, "learning_rate": 0.0002792530054143342, "loss": 0.1929, "step": 8440 }, { "epoch": 2.32, "learning_rate": 0.00027922547490134896, "loss": 0.1529, "step": 8450 }, { "epoch": 2.32, "learning_rate": 0.00027919794438836375, "loss": 0.1721, "step": 8460 }, { "epoch": 2.32, "learning_rate": 0.0002791704138753785, "loss": 0.1678, "step": 8470 }, { "epoch": 2.32, "learning_rate": 0.0002791428833623933, "loss": 0.1474, "step": 8480 }, { "epoch": 2.33, "learning_rate": 0.0002791153528494081, "loss": 0.1838, "step": 8490 }, { "epoch": 2.33, "learning_rate": 0.0002790878223364228, "loss": 0.1705, "step": 8500 }, { "epoch": 2.33, "eval_loss": 0.0972880944609642, "eval_runtime": 1054.6585, "eval_samples_per_second": 11.078, "eval_steps_per_second": 1.385, "eval_wer": 0.3167129082143182, "step": 8500 }, { "epoch": 2.33, "learning_rate": 0.0002790602918234376, "loss": 0.1696, "step": 8510 }, { "epoch": 2.33, "learning_rate": 0.00027903276131045236, "loss": 0.145, "step": 8520 }, { "epoch": 2.34, "learning_rate": 0.00027900523079746716, "loss": 0.1273, "step": 8530 }, { "epoch": 2.34, "learning_rate": 0.00027897770028448196, "loss": 0.1459, "step": 8540 }, { "epoch": 2.34, "learning_rate": 0.0002789501697714967, "loss": 0.1447, "step": 8550 }, { "epoch": 2.35, "learning_rate": 0.0002789226392585115, "loss": 0.1587, "step": 8560 }, { "epoch": 2.35, "learning_rate": 0.00027889510874552623, "loss": 0.1653, "step": 8570 }, { "epoch": 2.35, "learning_rate": 0.00027886757823254103, "loss": 0.1427, "step": 8580 }, { "epoch": 2.35, "learning_rate": 0.0002788400477195558, "loss": 0.1642, "step": 8590 }, { "epoch": 2.36, "learning_rate": 0.0002788125172065706, "loss": 0.1323, "step": 8600 }, { "epoch": 2.36, "eval_loss": 0.09097806364297867, "eval_runtime": 1042.0814, "eval_samples_per_second": 11.211, "eval_steps_per_second": 1.402, "eval_wer": 0.30551259892658966, "step": 8600 }, { "epoch": 2.36, "learning_rate": 0.00027878498669358536, "loss": 0.185, "step": 8610 }, { "epoch": 2.36, "learning_rate": 0.0002787574561806001, "loss": 0.1429, "step": 8620 }, { "epoch": 2.36, "learning_rate": 0.0002787299256676149, "loss": 0.1356, "step": 8630 }, { "epoch": 2.37, "learning_rate": 0.0002787023951546297, "loss": 0.1454, "step": 8640 }, { "epoch": 2.37, "learning_rate": 0.0002786748646416445, "loss": 0.1389, "step": 8650 }, { "epoch": 2.37, "learning_rate": 0.00027864733412865923, "loss": 0.1469, "step": 8660 }, { "epoch": 2.38, "learning_rate": 0.000278619803615674, "loss": 0.149, "step": 8670 }, { "epoch": 2.38, "learning_rate": 0.00027859227310268877, "loss": 0.1541, "step": 8680 }, { "epoch": 2.38, "learning_rate": 0.00027856474258970357, "loss": 0.1547, "step": 8690 }, { "epoch": 2.38, "learning_rate": 0.00027853721207671836, "loss": 0.1258, "step": 8700 }, { "epoch": 2.38, "eval_loss": 0.0872187614440918, "eval_runtime": 1047.8054, "eval_samples_per_second": 11.15, "eval_steps_per_second": 1.394, "eval_wer": 0.30171472755389794, "step": 8700 }, { "epoch": 2.39, "learning_rate": 0.0002785096815637331, "loss": 0.1555, "step": 8710 }, { "epoch": 2.39, "learning_rate": 0.0002784821510507479, "loss": 0.1583, "step": 8720 }, { "epoch": 2.39, "learning_rate": 0.00027845462053776264, "loss": 0.1569, "step": 8730 }, { "epoch": 2.39, "learning_rate": 0.00027842709002477744, "loss": 0.1695, "step": 8740 }, { "epoch": 2.4, "learning_rate": 0.00027839955951179223, "loss": 0.1258, "step": 8750 }, { "epoch": 2.4, "learning_rate": 0.000278372028998807, "loss": 0.1616, "step": 8760 }, { "epoch": 2.4, "learning_rate": 0.00027834449848582177, "loss": 0.1586, "step": 8770 }, { "epoch": 2.41, "learning_rate": 0.0002783169679728365, "loss": 0.1733, "step": 8780 }, { "epoch": 2.41, "learning_rate": 0.0002782894374598513, "loss": 0.1467, "step": 8790 }, { "epoch": 2.41, "learning_rate": 0.0002782619069468661, "loss": 0.1536, "step": 8800 }, { "epoch": 2.41, "eval_loss": 0.09626557677984238, "eval_runtime": 1048.2253, "eval_samples_per_second": 11.146, "eval_steps_per_second": 1.394, "eval_wer": 0.3102542527062676, "step": 8800 }, { "epoch": 2.41, "learning_rate": 0.00027823437643388084, "loss": 0.1618, "step": 8810 }, { "epoch": 2.42, "learning_rate": 0.00027820684592089564, "loss": 0.1546, "step": 8820 }, { "epoch": 2.42, "learning_rate": 0.0002781793154079104, "loss": 0.1415, "step": 8830 }, { "epoch": 2.42, "learning_rate": 0.0002781517848949252, "loss": 0.1769, "step": 8840 }, { "epoch": 2.42, "learning_rate": 0.00027812425438194, "loss": 0.1657, "step": 8850 }, { "epoch": 2.43, "learning_rate": 0.0002780967238689547, "loss": 0.1708, "step": 8860 }, { "epoch": 2.43, "learning_rate": 0.0002780691933559695, "loss": 0.1486, "step": 8870 }, { "epoch": 2.43, "learning_rate": 0.00027804166284298425, "loss": 0.1504, "step": 8880 }, { "epoch": 2.44, "learning_rate": 0.00027801413232999905, "loss": 0.1773, "step": 8890 }, { "epoch": 2.44, "learning_rate": 0.00027798660181701384, "loss": 0.1628, "step": 8900 }, { "epoch": 2.44, "eval_loss": 0.08714758604764938, "eval_runtime": 1049.6925, "eval_samples_per_second": 11.13, "eval_steps_per_second": 1.392, "eval_wer": 0.30454607477485673, "step": 8900 }, { "epoch": 2.44, "learning_rate": 0.00027795907130402864, "loss": 0.1854, "step": 8910 }, { "epoch": 2.44, "learning_rate": 0.0002779315407910434, "loss": 0.1482, "step": 8920 }, { "epoch": 2.45, "learning_rate": 0.0002779040102780581, "loss": 0.1622, "step": 8930 }, { "epoch": 2.45, "learning_rate": 0.0002778764797650729, "loss": 0.1618, "step": 8940 }, { "epoch": 2.45, "learning_rate": 0.0002778489492520877, "loss": 0.1535, "step": 8950 }, { "epoch": 2.46, "learning_rate": 0.0002778214187391025, "loss": 0.1579, "step": 8960 }, { "epoch": 2.46, "learning_rate": 0.00027779388822611725, "loss": 0.144, "step": 8970 }, { "epoch": 2.46, "learning_rate": 0.000277766357713132, "loss": 0.1676, "step": 8980 }, { "epoch": 2.46, "learning_rate": 0.0002777388272001468, "loss": 0.1622, "step": 8990 }, { "epoch": 2.47, "learning_rate": 0.0002777112966871616, "loss": 0.1504, "step": 9000 }, { "epoch": 2.47, "eval_loss": 0.08976973593235016, "eval_runtime": 1046.6105, "eval_samples_per_second": 11.163, "eval_steps_per_second": 1.396, "eval_wer": 0.3038069680705904, "step": 9000 }, { "epoch": 2.47, "learning_rate": 0.0002776837661741764, "loss": 0.1543, "step": 9010 }, { "epoch": 2.47, "learning_rate": 0.0002776562356611911, "loss": 0.1569, "step": 9020 }, { "epoch": 2.47, "learning_rate": 0.00027762870514820586, "loss": 0.1346, "step": 9030 }, { "epoch": 2.48, "learning_rate": 0.00027760117463522066, "loss": 0.1653, "step": 9040 }, { "epoch": 2.48, "learning_rate": 0.00027757364412223546, "loss": 0.1364, "step": 9050 }, { "epoch": 2.48, "learning_rate": 0.00027754611360925025, "loss": 0.1346, "step": 9060 }, { "epoch": 2.49, "learning_rate": 0.000277518583096265, "loss": 0.1355, "step": 9070 }, { "epoch": 2.49, "learning_rate": 0.0002774910525832798, "loss": 0.147, "step": 9080 }, { "epoch": 2.49, "learning_rate": 0.00027746352207029453, "loss": 0.1582, "step": 9090 }, { "epoch": 2.49, "learning_rate": 0.0002774359915573093, "loss": 0.1301, "step": 9100 }, { "epoch": 2.49, "eval_loss": 0.08558912575244904, "eval_runtime": 1045.4399, "eval_samples_per_second": 11.175, "eval_steps_per_second": 1.397, "eval_wer": 0.29659783498590014, "step": 9100 }, { "epoch": 2.5, "learning_rate": 0.0002774084610443241, "loss": 0.1486, "step": 9110 }, { "epoch": 2.5, "learning_rate": 0.00027738093053133886, "loss": 0.1662, "step": 9120 }, { "epoch": 2.5, "learning_rate": 0.00027735340001835366, "loss": 0.1518, "step": 9130 }, { "epoch": 2.5, "learning_rate": 0.0002773258695053684, "loss": 0.1664, "step": 9140 }, { "epoch": 2.51, "learning_rate": 0.0002772983389923832, "loss": 0.1394, "step": 9150 }, { "epoch": 2.51, "learning_rate": 0.000277270808479398, "loss": 0.1721, "step": 9160 }, { "epoch": 2.51, "learning_rate": 0.00027724327796641273, "loss": 0.1455, "step": 9170 }, { "epoch": 2.52, "learning_rate": 0.00027721574745342753, "loss": 0.1723, "step": 9180 }, { "epoch": 2.52, "learning_rate": 0.00027718821694044227, "loss": 0.1507, "step": 9190 }, { "epoch": 2.52, "learning_rate": 0.00027716068642745707, "loss": 0.1488, "step": 9200 }, { "epoch": 2.52, "eval_loss": 0.08458111435174942, "eval_runtime": 1050.2942, "eval_samples_per_second": 11.124, "eval_steps_per_second": 1.391, "eval_wer": 0.2977804057127263, "step": 9200 }, { "epoch": 2.52, "learning_rate": 0.00027713315591447186, "loss": 0.1467, "step": 9210 }, { "epoch": 2.53, "learning_rate": 0.0002771056254014866, "loss": 0.1567, "step": 9220 }, { "epoch": 2.53, "learning_rate": 0.0002770780948885014, "loss": 0.1866, "step": 9230 }, { "epoch": 2.53, "learning_rate": 0.00027705056437551614, "loss": 0.167, "step": 9240 }, { "epoch": 2.53, "learning_rate": 0.00027702303386253094, "loss": 0.1382, "step": 9250 }, { "epoch": 2.54, "learning_rate": 0.00027699550334954573, "loss": 0.176, "step": 9260 }, { "epoch": 2.54, "learning_rate": 0.00027696797283656053, "loss": 0.15, "step": 9270 }, { "epoch": 2.54, "learning_rate": 0.00027694044232357527, "loss": 0.1464, "step": 9280 }, { "epoch": 2.55, "learning_rate": 0.00027691291181059, "loss": 0.1787, "step": 9290 }, { "epoch": 2.55, "learning_rate": 0.0002768853812976048, "loss": 0.1621, "step": 9300 }, { "epoch": 2.55, "eval_loss": 0.08554688096046448, "eval_runtime": 1061.0591, "eval_samples_per_second": 11.011, "eval_steps_per_second": 1.377, "eval_wer": 0.30240835076867095, "step": 9300 }, { "epoch": 2.55, "learning_rate": 0.0002768578507846196, "loss": 0.1544, "step": 9310 }, { "epoch": 2.55, "learning_rate": 0.0002768303202716344, "loss": 0.3587, "step": 9320 }, { "epoch": 2.56, "learning_rate": 0.00027680278975864914, "loss": 0.169, "step": 9330 }, { "epoch": 2.56, "learning_rate": 0.0002767752592456639, "loss": 0.1487, "step": 9340 }, { "epoch": 2.56, "learning_rate": 0.0002767477287326787, "loss": 0.1448, "step": 9350 }, { "epoch": 2.56, "learning_rate": 0.0002767201982196935, "loss": 0.1492, "step": 9360 }, { "epoch": 2.57, "learning_rate": 0.00027669266770670827, "loss": 0.1484, "step": 9370 }, { "epoch": 2.57, "learning_rate": 0.000276665137193723, "loss": 0.142, "step": 9380 }, { "epoch": 2.57, "learning_rate": 0.00027663760668073775, "loss": 0.1614, "step": 9390 }, { "epoch": 2.58, "learning_rate": 0.00027661007616775255, "loss": 0.1453, "step": 9400 }, { "epoch": 2.58, "eval_loss": 0.08304944634437561, "eval_runtime": 1053.8093, "eval_samples_per_second": 11.086, "eval_steps_per_second": 1.386, "eval_wer": 0.2940393887019012, "step": 9400 }, { "epoch": 2.58, "learning_rate": 0.00027658254565476734, "loss": 0.1516, "step": 9410 }, { "epoch": 2.58, "learning_rate": 0.00027655501514178214, "loss": 0.1336, "step": 9420 }, { "epoch": 2.58, "learning_rate": 0.0002765274846287969, "loss": 0.1627, "step": 9430 }, { "epoch": 2.59, "learning_rate": 0.0002764999541158117, "loss": 0.1617, "step": 9440 }, { "epoch": 2.59, "learning_rate": 0.0002764724236028264, "loss": 0.1532, "step": 9450 }, { "epoch": 2.59, "learning_rate": 0.0002764448930898412, "loss": 0.138, "step": 9460 }, { "epoch": 2.59, "learning_rate": 0.000276417362576856, "loss": 0.1381, "step": 9470 }, { "epoch": 2.6, "learning_rate": 0.00027638983206387075, "loss": 0.1609, "step": 9480 }, { "epoch": 2.6, "learning_rate": 0.00027636230155088555, "loss": 0.1559, "step": 9490 }, { "epoch": 2.6, "learning_rate": 0.0002763347710379003, "loss": 0.13, "step": 9500 }, { "epoch": 2.6, "eval_loss": 0.08238458633422852, "eval_runtime": 1057.6848, "eval_samples_per_second": 11.046, "eval_steps_per_second": 1.381, "eval_wer": 0.2919585190575821, "step": 9500 }, { "epoch": 2.61, "learning_rate": 0.0002763072405249151, "loss": 0.1517, "step": 9510 }, { "epoch": 2.61, "learning_rate": 0.0002762797100119299, "loss": 0.1294, "step": 9520 }, { "epoch": 2.61, "learning_rate": 0.0002762521794989446, "loss": 0.1423, "step": 9530 }, { "epoch": 2.61, "learning_rate": 0.0002762246489859594, "loss": 0.1446, "step": 9540 }, { "epoch": 2.62, "learning_rate": 0.00027619711847297416, "loss": 0.1273, "step": 9550 }, { "epoch": 2.62, "learning_rate": 0.00027616958795998896, "loss": 0.1656, "step": 9560 }, { "epoch": 2.62, "learning_rate": 0.00027614205744700375, "loss": 0.1377, "step": 9570 }, { "epoch": 2.63, "learning_rate": 0.00027611452693401855, "loss": 0.146, "step": 9580 }, { "epoch": 2.63, "learning_rate": 0.0002760869964210333, "loss": 0.1363, "step": 9590 }, { "epoch": 2.63, "learning_rate": 0.00027605946590804803, "loss": 0.1225, "step": 9600 }, { "epoch": 2.63, "eval_loss": 0.08304128050804138, "eval_runtime": 1050.786, "eval_samples_per_second": 11.118, "eval_steps_per_second": 1.39, "eval_wer": 0.29202674429182207, "step": 9600 }, { "epoch": 2.63, "learning_rate": 0.0002760319353950628, "loss": 0.146, "step": 9610 }, { "epoch": 2.64, "learning_rate": 0.0002760044048820776, "loss": 0.1588, "step": 9620 }, { "epoch": 2.64, "learning_rate": 0.0002759768743690924, "loss": 0.1451, "step": 9630 }, { "epoch": 2.64, "learning_rate": 0.00027594934385610716, "loss": 0.2022, "step": 9640 }, { "epoch": 2.64, "learning_rate": 0.0002759218133431219, "loss": 0.1589, "step": 9650 }, { "epoch": 2.65, "learning_rate": 0.0002758942828301367, "loss": 0.1438, "step": 9660 }, { "epoch": 2.65, "learning_rate": 0.0002758667523171515, "loss": 0.1744, "step": 9670 }, { "epoch": 2.65, "learning_rate": 0.0002758392218041663, "loss": 0.1413, "step": 9680 }, { "epoch": 2.66, "learning_rate": 0.00027581169129118103, "loss": 0.143, "step": 9690 }, { "epoch": 2.66, "learning_rate": 0.00027578416077819577, "loss": 0.1337, "step": 9700 }, { "epoch": 2.66, "eval_loss": 0.08384636044502258, "eval_runtime": 1047.8124, "eval_samples_per_second": 11.15, "eval_steps_per_second": 1.394, "eval_wer": 0.2913558628217957, "step": 9700 }, { "epoch": 2.66, "learning_rate": 0.00027575663026521057, "loss": 0.1633, "step": 9710 }, { "epoch": 2.66, "learning_rate": 0.00027572909975222536, "loss": 0.1326, "step": 9720 }, { "epoch": 2.67, "learning_rate": 0.00027570156923924016, "loss": 0.1437, "step": 9730 }, { "epoch": 2.67, "learning_rate": 0.0002756740387262549, "loss": 0.1832, "step": 9740 }, { "epoch": 2.67, "learning_rate": 0.00027564650821326964, "loss": 0.1174, "step": 9750 }, { "epoch": 2.67, "learning_rate": 0.00027561897770028444, "loss": 0.1536, "step": 9760 }, { "epoch": 2.68, "learning_rate": 0.00027559144718729923, "loss": 0.1433, "step": 9770 }, { "epoch": 2.68, "learning_rate": 0.00027556391667431403, "loss": 0.1524, "step": 9780 }, { "epoch": 2.68, "learning_rate": 0.00027553638616132877, "loss": 0.1646, "step": 9790 }, { "epoch": 2.69, "learning_rate": 0.00027550885564834357, "loss": 0.1192, "step": 9800 }, { "epoch": 2.69, "eval_loss": 0.0845814049243927, "eval_runtime": 1048.3361, "eval_samples_per_second": 11.144, "eval_steps_per_second": 1.394, "eval_wer": 0.29916765214227237, "step": 9800 }, { "epoch": 2.69, "learning_rate": 0.0002754813251353583, "loss": 0.1384, "step": 9810 }, { "epoch": 2.69, "learning_rate": 0.0002754537946223731, "loss": 0.1412, "step": 9820 }, { "epoch": 2.69, "learning_rate": 0.0002754262641093879, "loss": 0.149, "step": 9830 }, { "epoch": 2.7, "learning_rate": 0.00027539873359640264, "loss": 0.1552, "step": 9840 }, { "epoch": 2.7, "learning_rate": 0.00027537120308341744, "loss": 0.1293, "step": 9850 }, { "epoch": 2.7, "learning_rate": 0.0002753436725704322, "loss": 0.1591, "step": 9860 }, { "epoch": 2.7, "learning_rate": 0.000275316142057447, "loss": 0.1333, "step": 9870 }, { "epoch": 2.71, "learning_rate": 0.00027528861154446177, "loss": 0.144, "step": 9880 }, { "epoch": 2.71, "learning_rate": 0.0002752610810314765, "loss": 0.1652, "step": 9890 }, { "epoch": 2.71, "learning_rate": 0.0002752335505184913, "loss": 0.1478, "step": 9900 }, { "epoch": 2.71, "eval_loss": 0.07936351001262665, "eval_runtime": 1059.5079, "eval_samples_per_second": 11.027, "eval_steps_per_second": 1.379, "eval_wer": 0.292129082143182, "step": 9900 }, { "epoch": 2.72, "learning_rate": 0.00027520602000550605, "loss": 0.1519, "step": 9910 }, { "epoch": 2.72, "learning_rate": 0.00027517848949252084, "loss": 0.1477, "step": 9920 }, { "epoch": 2.72, "learning_rate": 0.00027515095897953564, "loss": 0.1609, "step": 9930 }, { "epoch": 2.72, "learning_rate": 0.00027512342846655044, "loss": 0.1521, "step": 9940 }, { "epoch": 2.73, "learning_rate": 0.0002750958979535652, "loss": 0.1191, "step": 9950 }, { "epoch": 2.73, "learning_rate": 0.0002750683674405799, "loss": 0.131, "step": 9960 }, { "epoch": 2.73, "learning_rate": 0.0002750408369275947, "loss": 0.1323, "step": 9970 }, { "epoch": 2.73, "learning_rate": 0.0002750133064146095, "loss": 0.1182, "step": 9980 }, { "epoch": 2.74, "learning_rate": 0.0002749857759016243, "loss": 0.1493, "step": 9990 }, { "epoch": 2.74, "learning_rate": 0.00027495824538863905, "loss": 0.1188, "step": 10000 }, { "epoch": 2.74, "eval_loss": 0.0771343931555748, "eval_runtime": 1057.7326, "eval_samples_per_second": 11.045, "eval_steps_per_second": 1.381, "eval_wer": 0.28751250795961064, "step": 10000 }, { "epoch": 2.74, "learning_rate": 0.0002749307148756538, "loss": 0.14, "step": 10010 }, { "epoch": 2.75, "learning_rate": 0.0002749031843626686, "loss": 0.1393, "step": 10020 }, { "epoch": 2.75, "learning_rate": 0.0002748756538496834, "loss": 0.1269, "step": 10030 }, { "epoch": 2.75, "learning_rate": 0.0002748481233366982, "loss": 0.152, "step": 10040 }, { "epoch": 2.75, "learning_rate": 0.0002748205928237129, "loss": 0.1153, "step": 10050 }, { "epoch": 2.76, "learning_rate": 0.00027479306231072766, "loss": 0.1319, "step": 10060 }, { "epoch": 2.76, "learning_rate": 0.00027476553179774246, "loss": 0.1534, "step": 10070 }, { "epoch": 2.76, "learning_rate": 0.00027473800128475725, "loss": 0.1336, "step": 10080 }, { "epoch": 2.76, "learning_rate": 0.00027471047077177205, "loss": 0.1335, "step": 10090 }, { "epoch": 2.77, "learning_rate": 0.0002746829402587868, "loss": 0.2217, "step": 10100 }, { "epoch": 2.77, "eval_loss": 0.08399680256843567, "eval_runtime": 1055.6916, "eval_samples_per_second": 11.067, "eval_steps_per_second": 1.384, "eval_wer": 0.30015691803875194, "step": 10100 }, { "epoch": 2.77, "learning_rate": 0.0002746554097458016, "loss": 0.1605, "step": 10110 }, { "epoch": 2.77, "learning_rate": 0.0002746278792328163, "loss": 0.1481, "step": 10120 }, { "epoch": 2.78, "learning_rate": 0.0002746003487198311, "loss": 0.1511, "step": 10130 }, { "epoch": 2.78, "learning_rate": 0.0002745728182068459, "loss": 0.1261, "step": 10140 }, { "epoch": 2.78, "learning_rate": 0.00027454528769386066, "loss": 0.1192, "step": 10150 }, { "epoch": 2.78, "learning_rate": 0.00027451775718087545, "loss": 0.1568, "step": 10160 }, { "epoch": 2.79, "learning_rate": 0.0002744902266678902, "loss": 0.1355, "step": 10170 }, { "epoch": 2.79, "learning_rate": 0.000274462696154905, "loss": 0.1283, "step": 10180 }, { "epoch": 2.79, "learning_rate": 0.0002744351656419198, "loss": 0.1693, "step": 10190 }, { "epoch": 2.79, "learning_rate": 0.00027440763512893453, "loss": 0.1169, "step": 10200 }, { "epoch": 2.79, "eval_loss": 0.07694612443447113, "eval_runtime": 1046.644, "eval_samples_per_second": 11.162, "eval_steps_per_second": 1.396, "eval_wer": 0.2876375875557173, "step": 10200 }, { "epoch": 2.8, "learning_rate": 0.0002743801046159493, "loss": 0.165, "step": 10210 }, { "epoch": 2.8, "learning_rate": 0.00027435257410296407, "loss": 0.1532, "step": 10220 }, { "epoch": 2.8, "learning_rate": 0.00027432504358997886, "loss": 0.1442, "step": 10230 }, { "epoch": 2.81, "learning_rate": 0.00027429751307699366, "loss": 0.1493, "step": 10240 }, { "epoch": 2.81, "learning_rate": 0.00027426998256400845, "loss": 0.1159, "step": 10250 }, { "epoch": 2.81, "learning_rate": 0.0002742424520510232, "loss": 0.1357, "step": 10260 }, { "epoch": 2.81, "learning_rate": 0.00027421492153803794, "loss": 0.1288, "step": 10270 }, { "epoch": 2.82, "learning_rate": 0.00027418739102505273, "loss": 0.1484, "step": 10280 }, { "epoch": 2.82, "learning_rate": 0.00027415986051206753, "loss": 0.1471, "step": 10290 }, { "epoch": 2.82, "learning_rate": 0.0002741323299990823, "loss": 0.1334, "step": 10300 }, { "epoch": 2.82, "eval_loss": 0.07819453626871109, "eval_runtime": 1048.7903, "eval_samples_per_second": 11.14, "eval_steps_per_second": 1.393, "eval_wer": 0.287455653597744, "step": 10300 }, { "epoch": 2.83, "learning_rate": 0.00027410479948609707, "loss": 0.1331, "step": 10310 }, { "epoch": 2.83, "learning_rate": 0.0002740772689731118, "loss": 0.151, "step": 10320 }, { "epoch": 2.83, "learning_rate": 0.0002740497384601266, "loss": 0.1388, "step": 10330 }, { "epoch": 2.83, "learning_rate": 0.0002740222079471414, "loss": 0.1825, "step": 10340 }, { "epoch": 2.84, "learning_rate": 0.0002739946774341562, "loss": 0.1315, "step": 10350 }, { "epoch": 2.84, "learning_rate": 0.00027396714692117094, "loss": 0.1362, "step": 10360 }, { "epoch": 2.84, "learning_rate": 0.0002739396164081857, "loss": 0.1836, "step": 10370 }, { "epoch": 2.84, "learning_rate": 0.0002739120858952005, "loss": 0.1388, "step": 10380 }, { "epoch": 2.85, "learning_rate": 0.00027388455538221527, "loss": 0.1705, "step": 10390 }, { "epoch": 2.85, "learning_rate": 0.00027385702486923007, "loss": 0.1623, "step": 10400 }, { "epoch": 2.85, "eval_loss": 0.08330324292182922, "eval_runtime": 1040.7374, "eval_samples_per_second": 11.226, "eval_steps_per_second": 1.404, "eval_wer": 0.2947557536614209, "step": 10400 }, { "epoch": 2.85, "learning_rate": 0.0002738294943562448, "loss": 0.1416, "step": 10410 }, { "epoch": 2.86, "learning_rate": 0.00027380196384325955, "loss": 0.1439, "step": 10420 }, { "epoch": 2.86, "learning_rate": 0.00027377443333027434, "loss": 0.13, "step": 10430 }, { "epoch": 2.86, "learning_rate": 0.00027374690281728914, "loss": 0.1491, "step": 10440 }, { "epoch": 2.86, "learning_rate": 0.00027371937230430394, "loss": 0.1427, "step": 10450 }, { "epoch": 2.87, "learning_rate": 0.0002736918417913187, "loss": 0.1386, "step": 10460 }, { "epoch": 2.87, "learning_rate": 0.0002736643112783335, "loss": 0.1472, "step": 10470 }, { "epoch": 2.87, "learning_rate": 0.0002736367807653482, "loss": 0.1345, "step": 10480 }, { "epoch": 2.87, "learning_rate": 0.000273609250252363, "loss": 0.1459, "step": 10490 }, { "epoch": 2.88, "learning_rate": 0.0002735817197393778, "loss": 0.2, "step": 10500 }, { "epoch": 2.88, "eval_loss": 0.07720213383436203, "eval_runtime": 1051.6199, "eval_samples_per_second": 11.11, "eval_steps_per_second": 1.389, "eval_wer": 0.28559083052851814, "step": 10500 }, { "epoch": 2.88, "learning_rate": 0.00027355418922639255, "loss": 0.1559, "step": 10510 }, { "epoch": 2.88, "learning_rate": 0.00027352665871340734, "loss": 0.1678, "step": 10520 }, { "epoch": 2.89, "learning_rate": 0.0002734991282004221, "loss": 0.1393, "step": 10530 }, { "epoch": 2.89, "learning_rate": 0.0002734715976874369, "loss": 0.1568, "step": 10540 }, { "epoch": 2.89, "learning_rate": 0.0002734440671744517, "loss": 0.1266, "step": 10550 }, { "epoch": 2.89, "learning_rate": 0.00027341653666146647, "loss": 0.1457, "step": 10560 }, { "epoch": 2.9, "learning_rate": 0.0002733890061484812, "loss": 0.1428, "step": 10570 }, { "epoch": 2.9, "learning_rate": 0.00027336147563549596, "loss": 0.1418, "step": 10580 }, { "epoch": 2.9, "learning_rate": 0.00027333394512251075, "loss": 0.1239, "step": 10590 }, { "epoch": 2.9, "learning_rate": 0.00027330641460952555, "loss": 0.1288, "step": 10600 }, { "epoch": 2.9, "eval_loss": 0.07732350379228592, "eval_runtime": 1052.8488, "eval_samples_per_second": 11.097, "eval_steps_per_second": 1.388, "eval_wer": 0.28539752569817156, "step": 10600 }, { "epoch": 2.91, "learning_rate": 0.00027327888409654034, "loss": 0.1591, "step": 10610 }, { "epoch": 2.91, "learning_rate": 0.0002732513535835551, "loss": 0.179, "step": 10620 }, { "epoch": 2.91, "learning_rate": 0.0002732238230705698, "loss": 0.1627, "step": 10630 }, { "epoch": 2.92, "learning_rate": 0.0002731962925575846, "loss": 0.151, "step": 10640 }, { "epoch": 2.92, "learning_rate": 0.0002731687620445994, "loss": 0.1237, "step": 10650 }, { "epoch": 2.92, "learning_rate": 0.0002731412315316142, "loss": 0.1334, "step": 10660 }, { "epoch": 2.92, "learning_rate": 0.00027311370101862895, "loss": 0.1311, "step": 10670 }, { "epoch": 2.93, "learning_rate": 0.0002730861705056437, "loss": 0.1412, "step": 10680 }, { "epoch": 2.93, "learning_rate": 0.0002730586399926585, "loss": 0.1705, "step": 10690 }, { "epoch": 2.93, "learning_rate": 0.0002730311094796733, "loss": 0.1201, "step": 10700 }, { "epoch": 2.93, "eval_loss": 0.07816141843795776, "eval_runtime": 1068.2164, "eval_samples_per_second": 10.937, "eval_steps_per_second": 1.368, "eval_wer": 0.2885017738560902, "step": 10700 }, { "epoch": 2.93, "learning_rate": 0.0002730035789666881, "loss": 0.1286, "step": 10710 }, { "epoch": 2.94, "learning_rate": 0.0002729760484537028, "loss": 0.1341, "step": 10720 }, { "epoch": 2.94, "learning_rate": 0.00027294851794071757, "loss": 0.1418, "step": 10730 }, { "epoch": 2.94, "learning_rate": 0.00027292098742773236, "loss": 0.1541, "step": 10740 }, { "epoch": 2.95, "learning_rate": 0.00027289345691474716, "loss": 0.1817, "step": 10750 }, { "epoch": 2.95, "learning_rate": 0.00027286592640176195, "loss": 0.1397, "step": 10760 }, { "epoch": 2.95, "learning_rate": 0.0002728383958887767, "loss": 0.1222, "step": 10770 }, { "epoch": 2.95, "learning_rate": 0.0002728108653757915, "loss": 0.1289, "step": 10780 }, { "epoch": 2.96, "learning_rate": 0.00027278333486280623, "loss": 0.141, "step": 10790 }, { "epoch": 2.96, "learning_rate": 0.00027275580434982103, "loss": 0.1467, "step": 10800 }, { "epoch": 2.96, "eval_loss": 0.07779738306999207, "eval_runtime": 1050.2055, "eval_samples_per_second": 11.124, "eval_steps_per_second": 1.391, "eval_wer": 0.2863640498499045, "step": 10800 }, { "epoch": 2.96, "learning_rate": 0.0002727282738368358, "loss": 0.1565, "step": 10810 }, { "epoch": 2.96, "learning_rate": 0.00027270074332385057, "loss": 0.1536, "step": 10820 }, { "epoch": 2.97, "learning_rate": 0.00027267321281086536, "loss": 0.1532, "step": 10830 }, { "epoch": 2.97, "learning_rate": 0.0002726456822978801, "loss": 0.1658, "step": 10840 }, { "epoch": 2.97, "learning_rate": 0.0002726181517848949, "loss": 0.1091, "step": 10850 }, { "epoch": 2.98, "learning_rate": 0.0002725906212719097, "loss": 0.158, "step": 10860 }, { "epoch": 2.98, "learning_rate": 0.00027256309075892444, "loss": 0.1538, "step": 10870 }, { "epoch": 2.98, "learning_rate": 0.00027253556024593923, "loss": 0.1713, "step": 10880 }, { "epoch": 2.98, "learning_rate": 0.000272508029732954, "loss": 0.1597, "step": 10890 }, { "epoch": 2.99, "learning_rate": 0.00027248049921996877, "loss": 0.1452, "step": 10900 }, { "epoch": 2.99, "eval_loss": 0.0746755301952362, "eval_runtime": 1043.7631, "eval_samples_per_second": 11.193, "eval_steps_per_second": 1.4, "eval_wer": 0.28247521149822613, "step": 10900 }, { "epoch": 2.99, "learning_rate": 0.00027245296870698357, "loss": 0.1357, "step": 10910 }, { "epoch": 2.99, "learning_rate": 0.00027242543819399836, "loss": 0.1308, "step": 10920 }, { "epoch": 2.99, "learning_rate": 0.0002723979076810131, "loss": 0.1342, "step": 10930 }, { "epoch": 3.0, "learning_rate": 0.00027237037716802784, "loss": 0.1607, "step": 10940 }, { "epoch": 3.0, "learning_rate": 0.00027234284665504264, "loss": 0.1492, "step": 10950 }, { "epoch": 3.0, "learning_rate": 0.00027231531614205744, "loss": 0.1423, "step": 10960 }, { "epoch": 3.01, "learning_rate": 0.00027228778562907223, "loss": 0.1281, "step": 10970 }, { "epoch": 3.01, "learning_rate": 0.000272260255116087, "loss": 0.1284, "step": 10980 }, { "epoch": 3.01, "learning_rate": 0.0002722327246031017, "loss": 0.1524, "step": 10990 }, { "epoch": 3.01, "learning_rate": 0.0002722051940901165, "loss": 0.1182, "step": 11000 }, { "epoch": 3.01, "eval_loss": 0.07591033726930618, "eval_runtime": 1055.6051, "eval_samples_per_second": 11.068, "eval_steps_per_second": 1.384, "eval_wer": 0.28122441553716, "step": 11000 }, { "epoch": 3.02, "learning_rate": 0.0002721776635771313, "loss": 0.1261, "step": 11010 }, { "epoch": 3.02, "learning_rate": 0.0002721501330641461, "loss": 0.1208, "step": 11020 }, { "epoch": 3.02, "learning_rate": 0.00027212260255116084, "loss": 0.1195, "step": 11030 }, { "epoch": 3.03, "learning_rate": 0.0002720950720381756, "loss": 0.1374, "step": 11040 }, { "epoch": 3.03, "learning_rate": 0.0002720675415251904, "loss": 0.1129, "step": 11050 }, { "epoch": 3.03, "learning_rate": 0.0002720400110122052, "loss": 0.1203, "step": 11060 }, { "epoch": 3.03, "learning_rate": 0.00027201248049921997, "loss": 0.1218, "step": 11070 }, { "epoch": 3.04, "learning_rate": 0.0002719849499862347, "loss": 0.1215, "step": 11080 }, { "epoch": 3.04, "learning_rate": 0.0002719574194732495, "loss": 0.1353, "step": 11090 }, { "epoch": 3.04, "learning_rate": 0.00027192988896026425, "loss": 0.1078, "step": 11100 }, { "epoch": 3.04, "eval_loss": 0.07436859607696533, "eval_runtime": 1060.0152, "eval_samples_per_second": 11.022, "eval_steps_per_second": 1.378, "eval_wer": 0.27713090148276176, "step": 11100 }, { "epoch": 3.04, "learning_rate": 0.00027190235844727905, "loss": 0.122, "step": 11110 }, { "epoch": 3.05, "learning_rate": 0.00027187482793429384, "loss": 0.1207, "step": 11120 }, { "epoch": 3.05, "learning_rate": 0.0002718472974213086, "loss": 0.1236, "step": 11130 }, { "epoch": 3.05, "learning_rate": 0.0002718197669083234, "loss": 0.1423, "step": 11140 }, { "epoch": 3.06, "learning_rate": 0.0002717922363953381, "loss": 0.1057, "step": 11150 }, { "epoch": 3.06, "learning_rate": 0.0002717647058823529, "loss": 0.1309, "step": 11160 }, { "epoch": 3.06, "learning_rate": 0.0002717371753693677, "loss": 0.1566, "step": 11170 }, { "epoch": 3.06, "learning_rate": 0.00027170964485638245, "loss": 0.1259, "step": 11180 }, { "epoch": 3.07, "learning_rate": 0.00027168211434339725, "loss": 0.1369, "step": 11190 }, { "epoch": 3.07, "learning_rate": 0.000271654583830412, "loss": 0.1426, "step": 11200 }, { "epoch": 3.07, "eval_loss": 0.07973187416791916, "eval_runtime": 1054.5609, "eval_samples_per_second": 11.079, "eval_steps_per_second": 1.385, "eval_wer": 0.28829709815337035, "step": 11200 }, { "epoch": 3.07, "learning_rate": 0.0002716270533174268, "loss": 0.135, "step": 11210 }, { "epoch": 3.07, "learning_rate": 0.0002715995228044416, "loss": 0.1249, "step": 11220 }, { "epoch": 3.08, "learning_rate": 0.0002715719922914564, "loss": 0.1414, "step": 11230 }, { "epoch": 3.08, "learning_rate": 0.0002715444617784711, "loss": 0.1297, "step": 11240 }, { "epoch": 3.08, "learning_rate": 0.00027151693126548586, "loss": 0.1273, "step": 11250 }, { "epoch": 3.09, "learning_rate": 0.00027148940075250066, "loss": 0.1372, "step": 11260 }, { "epoch": 3.09, "learning_rate": 0.00027146187023951545, "loss": 0.1226, "step": 11270 }, { "epoch": 3.09, "learning_rate": 0.00027143433972653025, "loss": 0.1658, "step": 11280 }, { "epoch": 3.09, "learning_rate": 0.000271406809213545, "loss": 0.1384, "step": 11290 }, { "epoch": 3.1, "learning_rate": 0.00027137927870055973, "loss": 0.1322, "step": 11300 }, { "epoch": 3.1, "eval_loss": 0.07649017870426178, "eval_runtime": 1041.843, "eval_samples_per_second": 11.214, "eval_steps_per_second": 1.402, "eval_wer": 0.2796552351496407, "step": 11300 }, { "epoch": 3.1, "learning_rate": 0.00027135174818757453, "loss": 0.1446, "step": 11310 }, { "epoch": 3.1, "learning_rate": 0.0002713242176745893, "loss": 0.1355, "step": 11320 }, { "epoch": 3.1, "learning_rate": 0.0002712966871616041, "loss": 0.1253, "step": 11330 }, { "epoch": 3.11, "learning_rate": 0.00027126915664861886, "loss": 0.1483, "step": 11340 }, { "epoch": 3.11, "learning_rate": 0.0002712416261356336, "loss": 0.1045, "step": 11350 }, { "epoch": 3.11, "learning_rate": 0.0002712140956226484, "loss": 0.1288, "step": 11360 }, { "epoch": 3.12, "learning_rate": 0.0002711865651096632, "loss": 0.1186, "step": 11370 }, { "epoch": 3.12, "learning_rate": 0.000271159034596678, "loss": 0.1238, "step": 11380 }, { "epoch": 3.12, "learning_rate": 0.00027113150408369273, "loss": 0.1288, "step": 11390 }, { "epoch": 3.12, "learning_rate": 0.0002711039735707075, "loss": 0.1655, "step": 11400 }, { "epoch": 3.12, "eval_loss": 0.07433360815048218, "eval_runtime": 1060.5558, "eval_samples_per_second": 11.016, "eval_steps_per_second": 1.378, "eval_wer": 0.28851314472846357, "step": 11400 }, { "epoch": 3.13, "learning_rate": 0.00027107644305772227, "loss": 0.1117, "step": 11410 }, { "epoch": 3.13, "learning_rate": 0.00027104891254473707, "loss": 0.1776, "step": 11420 }, { "epoch": 3.13, "learning_rate": 0.00027102138203175186, "loss": 0.1274, "step": 11430 }, { "epoch": 3.13, "learning_rate": 0.0002709938515187666, "loss": 0.1339, "step": 11440 }, { "epoch": 3.14, "learning_rate": 0.0002709663210057814, "loss": 0.1317, "step": 11450 }, { "epoch": 3.14, "learning_rate": 0.00027093879049279614, "loss": 0.1253, "step": 11460 }, { "epoch": 3.14, "learning_rate": 0.00027091125997981094, "loss": 0.1291, "step": 11470 }, { "epoch": 3.15, "learning_rate": 0.00027088372946682573, "loss": 0.132, "step": 11480 }, { "epoch": 3.15, "learning_rate": 0.0002708561989538405, "loss": 0.1593, "step": 11490 }, { "epoch": 3.15, "learning_rate": 0.00027082866844085527, "loss": 0.1243, "step": 11500 }, { "epoch": 3.15, "eval_loss": 0.07442672550678253, "eval_runtime": 1063.6769, "eval_samples_per_second": 10.984, "eval_steps_per_second": 1.374, "eval_wer": 0.2792231419994542, "step": 11500 }, { "epoch": 3.15, "learning_rate": 0.00027080113792787, "loss": 0.1272, "step": 11510 }, { "epoch": 3.16, "learning_rate": 0.0002707736074148848, "loss": 0.1373, "step": 11520 }, { "epoch": 3.16, "learning_rate": 0.0002707460769018996, "loss": 0.1241, "step": 11530 }, { "epoch": 3.16, "learning_rate": 0.00027071854638891434, "loss": 0.1344, "step": 11540 }, { "epoch": 3.16, "learning_rate": 0.00027069101587592914, "loss": 0.1718, "step": 11550 }, { "epoch": 3.17, "learning_rate": 0.0002706634853629439, "loss": 0.1483, "step": 11560 }, { "epoch": 3.17, "learning_rate": 0.0002706359548499587, "loss": 0.1443, "step": 11570 }, { "epoch": 3.17, "learning_rate": 0.00027060842433697347, "loss": 0.143, "step": 11580 }, { "epoch": 3.18, "learning_rate": 0.00027058089382398827, "loss": 0.1322, "step": 11590 }, { "epoch": 3.18, "learning_rate": 0.000270553363311003, "loss": 0.1724, "step": 11600 }, { "epoch": 3.18, "eval_loss": 0.07493375241756439, "eval_runtime": 1056.542, "eval_samples_per_second": 11.058, "eval_steps_per_second": 1.383, "eval_wer": 0.2777676703356681, "step": 11600 }, { "epoch": 3.18, "learning_rate": 0.00027052583279801775, "loss": 0.1393, "step": 11610 }, { "epoch": 3.18, "learning_rate": 0.00027049830228503255, "loss": 0.1258, "step": 11620 }, { "epoch": 3.19, "learning_rate": 0.00027047077177204734, "loss": 0.1246, "step": 11630 }, { "epoch": 3.19, "learning_rate": 0.00027044324125906214, "loss": 0.1178, "step": 11640 }, { "epoch": 3.19, "learning_rate": 0.0002704157107460769, "loss": 0.1238, "step": 11650 }, { "epoch": 3.19, "learning_rate": 0.0002703881802330916, "loss": 0.1403, "step": 11660 }, { "epoch": 3.2, "learning_rate": 0.0002703606497201064, "loss": 0.147, "step": 11670 }, { "epoch": 3.2, "learning_rate": 0.0002703331192071212, "loss": 0.139, "step": 11680 }, { "epoch": 3.2, "learning_rate": 0.000270305588694136, "loss": 0.1418, "step": 11690 }, { "epoch": 3.21, "learning_rate": 0.00027027805818115075, "loss": 0.1136, "step": 11700 }, { "epoch": 3.21, "eval_loss": 0.07298364490270615, "eval_runtime": 1068.3256, "eval_samples_per_second": 10.936, "eval_steps_per_second": 1.368, "eval_wer": 0.27636905303374876, "step": 11700 }, { "epoch": 3.21, "learning_rate": 0.0002702505276681655, "loss": 0.12, "step": 11710 }, { "epoch": 3.21, "learning_rate": 0.0002702229971551803, "loss": 0.3796, "step": 11720 }, { "epoch": 3.21, "learning_rate": 0.0002701954666421951, "loss": 0.125, "step": 11730 }, { "epoch": 3.22, "learning_rate": 0.0002701679361292099, "loss": 0.1521, "step": 11740 }, { "epoch": 3.22, "learning_rate": 0.0002701404056162246, "loss": 0.1114, "step": 11750 }, { "epoch": 3.22, "learning_rate": 0.0002701128751032394, "loss": 0.1502, "step": 11760 }, { "epoch": 3.23, "learning_rate": 0.00027008534459025416, "loss": 0.1327, "step": 11770 }, { "epoch": 3.23, "learning_rate": 0.00027005781407726895, "loss": 0.1269, "step": 11780 }, { "epoch": 3.23, "learning_rate": 0.00027003028356428375, "loss": 0.1491, "step": 11790 }, { "epoch": 3.23, "learning_rate": 0.0002700027530512985, "loss": 0.1428, "step": 11800 }, { "epoch": 3.23, "eval_loss": 0.07759026437997818, "eval_runtime": 1063.5053, "eval_samples_per_second": 10.985, "eval_steps_per_second": 1.374, "eval_wer": 0.283578186118439, "step": 11800 }, { "epoch": 3.24, "learning_rate": 0.0002699752225383133, "loss": 0.1508, "step": 11810 }, { "epoch": 3.24, "learning_rate": 0.00026994769202532803, "loss": 0.1398, "step": 11820 }, { "epoch": 3.24, "learning_rate": 0.0002699201615123428, "loss": 0.1143, "step": 11830 }, { "epoch": 3.24, "learning_rate": 0.0002698926309993576, "loss": 0.132, "step": 11840 }, { "epoch": 3.25, "learning_rate": 0.00026986510048637236, "loss": 0.1306, "step": 11850 }, { "epoch": 3.25, "learning_rate": 0.00026983756997338716, "loss": 0.1439, "step": 11860 }, { "epoch": 3.25, "learning_rate": 0.0002698100394604019, "loss": 0.1557, "step": 11870 }, { "epoch": 3.26, "learning_rate": 0.0002697825089474167, "loss": 0.1394, "step": 11880 }, { "epoch": 3.26, "learning_rate": 0.0002697549784344315, "loss": 0.1352, "step": 11890 }, { "epoch": 3.26, "learning_rate": 0.0002697274479214463, "loss": 0.1189, "step": 11900 }, { "epoch": 3.26, "eval_loss": 0.07876188308000565, "eval_runtime": 1048.6869, "eval_samples_per_second": 11.141, "eval_steps_per_second": 1.393, "eval_wer": 0.2892636223051033, "step": 11900 }, { "epoch": 3.26, "learning_rate": 0.00026969991740846103, "loss": 0.1477, "step": 11910 }, { "epoch": 3.27, "learning_rate": 0.00026967238689547577, "loss": 0.1309, "step": 11920 }, { "epoch": 3.27, "learning_rate": 0.00026964485638249057, "loss": 0.1291, "step": 11930 }, { "epoch": 3.27, "learning_rate": 0.00026961732586950536, "loss": 0.1233, "step": 11940 }, { "epoch": 3.27, "learning_rate": 0.00026958979535652016, "loss": 0.1144, "step": 11950 }, { "epoch": 3.28, "learning_rate": 0.0002695622648435349, "loss": 0.1471, "step": 11960 }, { "epoch": 3.28, "learning_rate": 0.00026953473433054964, "loss": 0.1418, "step": 11970 }, { "epoch": 3.28, "learning_rate": 0.00026950720381756444, "loss": 0.1284, "step": 11980 }, { "epoch": 3.29, "learning_rate": 0.00026947967330457923, "loss": 0.1463, "step": 11990 }, { "epoch": 3.29, "learning_rate": 0.00026945214279159403, "loss": 0.1065, "step": 12000 }, { "epoch": 3.29, "eval_loss": 0.0732453316450119, "eval_runtime": 1048.1918, "eval_samples_per_second": 11.146, "eval_steps_per_second": 1.394, "eval_wer": 0.2746179386882562, "step": 12000 }, { "epoch": 3.29, "learning_rate": 0.00026942461227860877, "loss": 0.1389, "step": 12010 }, { "epoch": 3.29, "learning_rate": 0.0002693970817656235, "loss": 0.1317, "step": 12020 }, { "epoch": 3.3, "learning_rate": 0.0002693695512526383, "loss": 0.1402, "step": 12030 }, { "epoch": 3.3, "learning_rate": 0.0002693420207396531, "loss": 0.1282, "step": 12040 }, { "epoch": 3.3, "learning_rate": 0.0002693144902266679, "loss": 0.1132, "step": 12050 }, { "epoch": 3.3, "learning_rate": 0.00026928695971368264, "loss": 0.1618, "step": 12060 }, { "epoch": 3.31, "learning_rate": 0.0002692594292006974, "loss": 0.1398, "step": 12070 }, { "epoch": 3.31, "learning_rate": 0.0002692318986877122, "loss": 0.1341, "step": 12080 }, { "epoch": 3.31, "learning_rate": 0.00026920436817472697, "loss": 0.1472, "step": 12090 }, { "epoch": 3.32, "learning_rate": 0.00026917683766174177, "loss": 0.14, "step": 12100 }, { "epoch": 3.32, "eval_loss": 0.07203283160924911, "eval_runtime": 1117.1414, "eval_samples_per_second": 10.458, "eval_steps_per_second": 1.308, "eval_wer": 0.27564131720185575, "step": 12100 }, { "epoch": 3.32, "learning_rate": 0.00010876454293628807, "loss": 8.3306, "step": 12110 }, { "epoch": 3.32, "learning_rate": 0.0001085983379501385, "loss": 3.3464, "step": 12120 }, { "epoch": 3.32, "learning_rate": 0.00010843213296398891, "loss": 3.1357, "step": 12130 }, { "epoch": 3.33, "learning_rate": 0.00010826592797783933, "loss": 3.1124, "step": 12140 }, { "epoch": 3.33, "learning_rate": 0.0001081163434903047, "loss": 3.1601, "step": 12150 }, { "epoch": 3.33, "learning_rate": 0.00010795013850415512, "loss": 3.103, "step": 12160 }, { "epoch": 3.33, "learning_rate": 0.00010778393351800553, "loss": 3.0144, "step": 12170 }, { "epoch": 3.34, "learning_rate": 0.00010761772853185595, "loss": 3.0122, "step": 12180 }, { "epoch": 3.34, "learning_rate": 0.00010745152354570635, "loss": 2.9556, "step": 12190 }, { "epoch": 3.34, "learning_rate": 0.00010728531855955679, "loss": 2.7696, "step": 12200 }, { "epoch": 3.34, "eval_loss": 2.8195414543151855, "eval_runtime": 1007.6166, "eval_samples_per_second": 11.595, "eval_steps_per_second": 1.45, "eval_wer": 1.0, "step": 12200 }, { "epoch": 3.35, "learning_rate": 0.00010711911357340719, "loss": 4.4042, "step": 12210 }, { "epoch": 3.35, "learning_rate": 0.00010695290858725761, "loss": 3.3022, "step": 12220 }, { "epoch": 3.35, "learning_rate": 0.00010678670360110802, "loss": 3.0588, "step": 12230 }, { "epoch": 3.35, "learning_rate": 0.00010662049861495845, "loss": 3.0238, "step": 12240 }, { "epoch": 3.36, "learning_rate": 0.00010645429362880886, "loss": 2.9607, "step": 12250 }, { "epoch": 3.36, "learning_rate": 0.00010628808864265928, "loss": 2.9219, "step": 12260 }, { "epoch": 3.36, "learning_rate": 0.00010612188365650968, "loss": 2.7164, "step": 12270 }, { "epoch": 3.36, "learning_rate": 0.0001059556786703601, "loss": 2.273, "step": 12280 }, { "epoch": 3.37, "learning_rate": 0.00010578947368421051, "loss": 1.479, "step": 12290 }, { "epoch": 3.37, "learning_rate": 0.00010562326869806094, "loss": 0.7388, "step": 12300 }, { "epoch": 3.37, "eval_loss": 0.42925623059272766, "eval_runtime": 992.4477, "eval_samples_per_second": 11.772, "eval_steps_per_second": 1.472, "eval_wer": 0.7220845083234786, "step": 12300 }, { "epoch": 3.37, "learning_rate": 0.00010545706371191135, "loss": 0.5404, "step": 12310 }, { "epoch": 3.38, "learning_rate": 0.00010529085872576177, "loss": 0.3253, "step": 12320 }, { "epoch": 3.38, "learning_rate": 0.00010512465373961217, "loss": 0.2805, "step": 12330 }, { "epoch": 3.38, "learning_rate": 0.00010495844875346259, "loss": 0.2625, "step": 12340 }, { "epoch": 3.38, "learning_rate": 0.00010479224376731301, "loss": 0.2047, "step": 12350 }, { "epoch": 3.39, "learning_rate": 0.00010462603878116343, "loss": 0.1923, "step": 12360 }, { "epoch": 3.39, "learning_rate": 0.00010445983379501384, "loss": 0.1929, "step": 12370 }, { "epoch": 3.39, "learning_rate": 0.00010429362880886426, "loss": 0.1813, "step": 12380 }, { "epoch": 3.39, "learning_rate": 0.00010412742382271466, "loss": 0.1785, "step": 12390 }, { "epoch": 3.4, "learning_rate": 0.0001039612188365651, "loss": 0.1829, "step": 12400 }, { "epoch": 3.4, "eval_loss": 0.09247562289237976, "eval_runtime": 991.3415, "eval_samples_per_second": 11.785, "eval_steps_per_second": 1.474, "eval_wer": 0.29330028199763486, "step": 12400 }, { "epoch": 3.4, "learning_rate": 0.0001037950138504155, "loss": 0.2012, "step": 12410 }, { "epoch": 3.4, "learning_rate": 0.00010362880886426592, "loss": 0.151, "step": 12420 }, { "epoch": 3.41, "learning_rate": 0.00010346260387811633, "loss": 0.1492, "step": 12430 }, { "epoch": 3.41, "learning_rate": 0.00010329639889196675, "loss": 0.1593, "step": 12440 }, { "epoch": 3.41, "learning_rate": 0.00010313019390581715, "loss": 0.1294, "step": 12450 }, { "epoch": 3.41, "learning_rate": 0.00010296398891966759, "loss": 0.1458, "step": 12460 }, { "epoch": 3.42, "learning_rate": 0.00010279778393351799, "loss": 0.1639, "step": 12470 }, { "epoch": 3.42, "learning_rate": 0.00010263157894736841, "loss": 0.1313, "step": 12480 }, { "epoch": 3.42, "learning_rate": 0.00010246537396121882, "loss": 0.1751, "step": 12490 }, { "epoch": 3.42, "learning_rate": 0.00010229916897506925, "loss": 0.1534, "step": 12500 }, { "epoch": 3.42, "eval_loss": 0.07944495230913162, "eval_runtime": 989.7561, "eval_samples_per_second": 11.804, "eval_steps_per_second": 1.476, "eval_wer": 0.2762780860547621, "step": 12500 }, { "epoch": 3.43, "learning_rate": 0.00010213296398891966, "loss": 0.1306, "step": 12510 }, { "epoch": 3.43, "learning_rate": 0.00010196675900277008, "loss": 0.1483, "step": 12520 }, { "epoch": 3.43, "learning_rate": 0.00010180055401662048, "loss": 0.1643, "step": 12530 }, { "epoch": 3.44, "learning_rate": 0.0001016343490304709, "loss": 0.1494, "step": 12540 }, { "epoch": 3.44, "learning_rate": 0.00010146814404432131, "loss": 0.1435, "step": 12550 }, { "epoch": 3.44, "learning_rate": 0.00010130193905817174, "loss": 0.135, "step": 12560 }, { "epoch": 3.44, "learning_rate": 0.00010113573407202215, "loss": 0.1873, "step": 12570 }, { "epoch": 3.45, "learning_rate": 0.00010096952908587257, "loss": 0.1526, "step": 12580 }, { "epoch": 3.45, "learning_rate": 0.00010080332409972297, "loss": 0.1725, "step": 12590 }, { "epoch": 3.45, "learning_rate": 0.0001006371191135734, "loss": 0.134, "step": 12600 }, { "epoch": 3.45, "eval_loss": 0.07680243253707886, "eval_runtime": 989.8703, "eval_samples_per_second": 11.803, "eval_steps_per_second": 1.476, "eval_wer": 0.2718548167015373, "step": 12600 }, { "epoch": 3.45, "learning_rate": 0.00010047091412742381, "loss": 0.1714, "step": 12610 }, { "epoch": 3.46, "learning_rate": 0.00010030470914127423, "loss": 0.1599, "step": 12620 }, { "epoch": 3.46, "learning_rate": 0.00010013850415512464, "loss": 0.1292, "step": 12630 }, { "epoch": 3.46, "learning_rate": 9.997229916897506e-05, "loss": 0.1634, "step": 12640 }, { "epoch": 3.47, "learning_rate": 9.980609418282546e-05, "loss": 0.1341, "step": 12650 }, { "epoch": 3.47, "learning_rate": 9.96398891966759e-05, "loss": 0.1636, "step": 12660 }, { "epoch": 3.47, "learning_rate": 9.94736842105263e-05, "loss": 0.1418, "step": 12670 }, { "epoch": 3.47, "learning_rate": 9.930747922437672e-05, "loss": 0.1301, "step": 12680 }, { "epoch": 3.48, "learning_rate": 9.914127423822713e-05, "loss": 0.1679, "step": 12690 }, { "epoch": 3.48, "learning_rate": 9.897506925207755e-05, "loss": 0.1095, "step": 12700 }, { "epoch": 3.48, "eval_loss": 0.07164399325847626, "eval_runtime": 991.1101, "eval_samples_per_second": 11.788, "eval_steps_per_second": 1.474, "eval_wer": 0.2641339943600473, "step": 12700 }, { "epoch": 3.48, "learning_rate": 9.880886426592796e-05, "loss": 0.1323, "step": 12710 }, { "epoch": 3.48, "learning_rate": 9.864265927977839e-05, "loss": 0.1354, "step": 12720 }, { "epoch": 3.49, "learning_rate": 9.84764542936288e-05, "loss": 0.1336, "step": 12730 }, { "epoch": 3.49, "learning_rate": 9.831024930747921e-05, "loss": 0.1182, "step": 12740 }, { "epoch": 3.49, "learning_rate": 9.814404432132962e-05, "loss": 0.1188, "step": 12750 }, { "epoch": 3.5, "learning_rate": 9.797783933518005e-05, "loss": 0.1438, "step": 12760 }, { "epoch": 3.5, "learning_rate": 9.781163434903046e-05, "loss": 0.1498, "step": 12770 }, { "epoch": 3.5, "learning_rate": 9.764542936288088e-05, "loss": 0.1255, "step": 12780 }, { "epoch": 3.5, "learning_rate": 9.747922437673129e-05, "loss": 0.1412, "step": 12790 }, { "epoch": 3.51, "learning_rate": 9.73130193905817e-05, "loss": 0.1185, "step": 12800 }, { "epoch": 3.51, "eval_loss": 0.07018042355775833, "eval_runtime": 991.2624, "eval_samples_per_second": 11.786, "eval_steps_per_second": 1.474, "eval_wer": 0.2602906394978623, "step": 12800 }, { "epoch": 1.76, "learning_rate": 0.0001, "loss": 9.6158, "step": 12810 }, { "epoch": 1.76, "learning_rate": 0.0002, "loss": 4.6242, "step": 12820 }, { "epoch": 1.76, "learning_rate": 0.0003, "loss": 3.2009, "step": 12830 }, { "epoch": 1.76, "learning_rate": 0.0004, "loss": 3.1427, "step": 12840 }, { "epoch": 1.76, "learning_rate": 0.0005, "loss": 3.1868, "step": 12850 }, { "epoch": 1.76, "learning_rate": 0.0006, "loss": 3.03, "step": 12860 }, { "epoch": 1.76, "learning_rate": 0.0007, "loss": 2.9279, "step": 12870 }, { "epoch": 1.76, "learning_rate": 0.0008, "loss": 2.2881, "step": 12880 }, { "epoch": 1.77, "learning_rate": 0.0009000000000000001, "loss": 1.0598, "step": 12890 }, { "epoch": 1.77, "learning_rate": 0.001, "loss": 0.8413, "step": 12900 }, { "epoch": 1.77, "eval_loss": 0.6381306648254395, "eval_runtime": 1032.069, "eval_samples_per_second": 11.32, "eval_steps_per_second": 1.416, "eval_wer": 0.9998197007031673, "step": 12900 }, { "epoch": 1.77, "learning_rate": 0.0011, "loss": 0.6591, "step": 12910 }, { "epoch": 1.77, "learning_rate": 0.0012, "loss": 0.5799, "step": 12920 }, { "epoch": 1.77, "learning_rate": 0.0013000000000000002, "loss": 0.5874, "step": 12930 }, { "epoch": 1.77, "learning_rate": 0.0014, "loss": 0.6337, "step": 12940 }, { "epoch": 1.77, "learning_rate": 0.0015, "loss": 0.5462, "step": 12950 }, { "epoch": 1.78, "learning_rate": 0.0016, "loss": 0.6603, "step": 12960 }, { "epoch": 1.78, "learning_rate": 0.0017, "loss": 0.559, "step": 12970 }, { "epoch": 1.78, "learning_rate": 0.0018000000000000002, "loss": 0.5104, "step": 12980 }, { "epoch": 1.78, "learning_rate": 0.0019, "loss": 0.64, "step": 12990 }, { "epoch": 1.78, "learning_rate": 0.002, "loss": 0.7279, "step": 13000 }, { "epoch": 1.78, "eval_loss": 0.7579346895217896, "eval_runtime": 1025.6366, "eval_samples_per_second": 11.391, "eval_steps_per_second": 1.424, "eval_wer": 0.9911653344551956, "step": 13000 }, { "epoch": 1.78, "learning_rate": 0.001998610918183081, "loss": 0.76, "step": 13010 }, { "epoch": 1.78, "learning_rate": 0.001997221836366162, "loss": 0.6129, "step": 13020 }, { "epoch": 1.79, "learning_rate": 0.001995832754549243, "loss": 0.5919, "step": 13030 }, { "epoch": 1.79, "learning_rate": 0.001994443672732324, "loss": 0.597, "step": 13040 }, { "epoch": 1.79, "learning_rate": 0.001993054590915405, "loss": 0.7811, "step": 13050 }, { "epoch": 1.79, "learning_rate": 0.0019916655090984857, "loss": 0.7572, "step": 13060 }, { "epoch": 1.79, "learning_rate": 0.001990276427281567, "loss": 0.6534, "step": 13070 }, { "epoch": 1.79, "learning_rate": 0.0019888873454646477, "loss": 0.5853, "step": 13080 }, { "epoch": 1.79, "learning_rate": 0.001987498263647729, "loss": 0.6896, "step": 13090 }, { "epoch": 1.79, "learning_rate": 0.0019861091818308097, "loss": 0.7529, "step": 13100 }, { "epoch": 1.79, "eval_loss": 0.7920202016830444, "eval_runtime": 1038.712, "eval_samples_per_second": 11.248, "eval_steps_per_second": 1.407, "eval_wer": 0.9996394014063346, "step": 13100 }, { "epoch": 1.8, "learning_rate": 0.001984720100013891, "loss": 0.8078, "step": 13110 }, { "epoch": 1.8, "learning_rate": 0.0019833310181969718, "loss": 0.6549, "step": 13120 }, { "epoch": 1.8, "learning_rate": 0.001981941936380053, "loss": 0.5914, "step": 13130 }, { "epoch": 1.8, "learning_rate": 0.001980552854563134, "loss": 0.6606, "step": 13140 }, { "epoch": 1.8, "learning_rate": 0.001979163772746215, "loss": 0.8138, "step": 13150 }, { "epoch": 1.8, "learning_rate": 0.001977774690929296, "loss": 0.7442, "step": 13160 }, { "epoch": 1.8, "learning_rate": 0.001976385609112377, "loss": 0.6086, "step": 13170 }, { "epoch": 1.81, "learning_rate": 0.001974996527295458, "loss": 0.5857, "step": 13180 }, { "epoch": 1.81, "learning_rate": 0.0019736074454785387, "loss": 0.7015, "step": 13190 }, { "epoch": 1.81, "learning_rate": 0.00197221836366162, "loss": 0.6813, "step": 13200 }, { "epoch": 1.81, "eval_loss": 0.6364935040473938, "eval_runtime": 1043.1171, "eval_samples_per_second": 11.2, "eval_steps_per_second": 1.401, "eval_wer": 1.0451950237394074, "step": 13200 }, { "epoch": 1.81, "learning_rate": 0.0019708292818447007, "loss": 0.6895, "step": 13210 }, { "epoch": 1.81, "learning_rate": 0.0019694402000277815, "loss": 0.6146, "step": 13220 }, { "epoch": 1.81, "learning_rate": 0.0019680511182108627, "loss": 0.6458, "step": 13230 }, { "epoch": 1.81, "learning_rate": 0.0019666620363939435, "loss": 0.5839, "step": 13240 }, { "epoch": 1.82, "learning_rate": 0.0019652729545770247, "loss": 0.5767, "step": 13250 }, { "epoch": 1.82, "learning_rate": 0.0019638838727601055, "loss": 0.6664, "step": 13260 }, { "epoch": 1.82, "learning_rate": 0.0019624947909431863, "loss": 0.562, "step": 13270 }, { "epoch": 1.82, "learning_rate": 0.0019611057091262676, "loss": 0.5177, "step": 13280 }, { "epoch": 1.82, "learning_rate": 0.0019597166273093484, "loss": 0.5938, "step": 13290 }, { "epoch": 1.82, "learning_rate": 0.0019583275454924296, "loss": 0.5342, "step": 13300 }, { "epoch": 1.82, "eval_loss": 0.6438644528388977, "eval_runtime": 1041.9947, "eval_samples_per_second": 11.212, "eval_steps_per_second": 1.402, "eval_wer": 1.025121702025362, "step": 13300 } ], "logging_steps": 10, "max_steps": 14598, "num_train_epochs": 2, "save_steps": 100, "total_flos": 6.684795440358097e+19, "trial_name": null, "trial_params": null }