{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 0.6, "eval_steps": 500, "global_step": 60000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "grad_norm": 0.47260773181915283, "learning_rate": 1.8e-07, "loss": 1.1046, "step": 10 }, { "grad_norm": 0.42390117049217224, "learning_rate": 3.8e-07, "loss": 1.1073, "step": 20 }, { "grad_norm": 0.47471940517425537, "learning_rate": 5.8e-07, "loss": 1.0806, "step": 30 }, { "grad_norm": 0.4786790907382965, "learning_rate": 7.8e-07, "loss": 1.0771, "step": 40 }, { "grad_norm": 0.5104601383209229, "learning_rate": 9.8e-07, "loss": 1.0763, "step": 50 }, { "grad_norm": 0.4175357222557068, "learning_rate": 1.18e-06, "loss": 1.0802, "step": 60 }, { "grad_norm": 0.543953537940979, "learning_rate": 1.3800000000000001e-06, "loss": 1.0651, "step": 70 }, { "grad_norm": 0.6749354600906372, "learning_rate": 1.5800000000000003e-06, "loss": 1.0528, "step": 80 }, { "grad_norm": 0.5703180432319641, "learning_rate": 1.7800000000000001e-06, "loss": 1.0365, "step": 90 }, { "grad_norm": 0.5217125415802002, "learning_rate": 1.98e-06, "loss": 1.0242, "step": 100 }, { "grad_norm": 0.46530234813690186, "learning_rate": 2.1800000000000003e-06, "loss": 1.0424, "step": 110 }, { "grad_norm": 0.5628986954689026, "learning_rate": 2.38e-06, "loss": 1.0417, "step": 120 }, { "grad_norm": 0.46236786246299744, "learning_rate": 2.5800000000000003e-06, "loss": 1.0341, "step": 130 }, { "grad_norm": 0.485328733921051, "learning_rate": 2.78e-06, "loss": 1.0456, "step": 140 }, { "grad_norm": 0.41217294335365295, "learning_rate": 2.9800000000000003e-06, "loss": 1.0415, "step": 150 }, { "grad_norm": 0.4756689667701721, "learning_rate": 3.1800000000000005e-06, "loss": 1.0401, "step": 160 }, { "grad_norm": 0.5221025943756104, "learning_rate": 3.38e-06, "loss": 1.0334, "step": 170 }, { "grad_norm": 0.4560315012931824, "learning_rate": 3.58e-06, "loss": 1.054, "step": 180 }, { "grad_norm": 0.5379422307014465, "learning_rate": 3.7800000000000002e-06, "loss": 1.0469, "step": 190 }, { "grad_norm": 0.5234620571136475, "learning_rate": 3.98e-06, "loss": 1.0493, "step": 200 }, { "grad_norm": 0.5519675612449646, "learning_rate": 4.18e-06, "loss": 1.0337, "step": 210 }, { "grad_norm": 0.49558329582214355, "learning_rate": 4.38e-06, "loss": 1.0285, "step": 220 }, { "grad_norm": 0.49592623114585876, "learning_rate": 4.58e-06, "loss": 1.0281, "step": 230 }, { "grad_norm": 0.5485683679580688, "learning_rate": 4.780000000000001e-06, "loss": 1.0394, "step": 240 }, { "grad_norm": 0.4751640260219574, "learning_rate": 4.98e-06, "loss": 1.0213, "step": 250 }, { "grad_norm": 0.6961615085601807, "learning_rate": 5.18e-06, "loss": 1.0022, "step": 260 }, { "grad_norm": 0.896144449710846, "learning_rate": 5.38e-06, "loss": 1.0152, "step": 270 }, { "grad_norm": 0.9158169627189636, "learning_rate": 5.580000000000001e-06, "loss": 0.9863, "step": 280 }, { "grad_norm": 0.9867625832557678, "learning_rate": 5.78e-06, "loss": 0.9902, "step": 290 }, { "grad_norm": 1.052965760231018, "learning_rate": 5.98e-06, "loss": 0.9822, "step": 300 }, { "grad_norm": 1.1499738693237305, "learning_rate": 6.18e-06, "loss": 0.959, "step": 310 }, { "grad_norm": 1.2108310461044312, "learning_rate": 6.38e-06, "loss": 0.9245, "step": 320 }, { "grad_norm": 1.5613349676132202, "learning_rate": 6.58e-06, "loss": 0.9118, "step": 330 }, { "grad_norm": 1.7165347337722778, "learning_rate": 6.78e-06, "loss": 0.8826, "step": 340 }, { "grad_norm": 1.8094466924667358, "learning_rate": 6.98e-06, "loss": 0.8546, "step": 350 }, { "grad_norm": 1.695487141609192, "learning_rate": 7.180000000000001e-06, "loss": 0.8214, "step": 360 }, { "grad_norm": 1.7566548585891724, "learning_rate": 7.3800000000000005e-06, "loss": 0.8093, "step": 370 }, { "grad_norm": 2.205599784851074, "learning_rate": 7.580000000000001e-06, "loss": 0.7889, "step": 380 }, { "grad_norm": 1.8415162563323975, "learning_rate": 7.78e-06, "loss": 0.7516, "step": 390 }, { "grad_norm": 2.2959468364715576, "learning_rate": 7.98e-06, "loss": 0.7699, "step": 400 }, { "grad_norm": 1.8826522827148438, "learning_rate": 8.18e-06, "loss": 0.7232, "step": 410 }, { "grad_norm": 2.321441173553467, "learning_rate": 8.380000000000001e-06, "loss": 0.6952, "step": 420 }, { "grad_norm": 2.2520809173583984, "learning_rate": 8.580000000000001e-06, "loss": 0.6831, "step": 430 }, { "grad_norm": 2.074122428894043, "learning_rate": 8.78e-06, "loss": 0.6678, "step": 440 }, { "grad_norm": 1.6957366466522217, "learning_rate": 8.98e-06, "loss": 0.6229, "step": 450 }, { "grad_norm": 2.9986095428466797, "learning_rate": 9.180000000000002e-06, "loss": 0.6225, "step": 460 }, { "grad_norm": 2.0681660175323486, "learning_rate": 9.38e-06, "loss": 0.6048, "step": 470 }, { "grad_norm": 2.3952388763427734, "learning_rate": 9.58e-06, "loss": 0.616, "step": 480 }, { "grad_norm": 1.7885857820510864, "learning_rate": 9.78e-06, "loss": 0.5858, "step": 490 }, { "grad_norm": 2.843278646469116, "learning_rate": 9.980000000000001e-06, "loss": 0.5644, "step": 500 }, { "grad_norm": 2.2172935009002686, "learning_rate": 1.018e-05, "loss": 0.5593, "step": 510 }, { "grad_norm": 2.082463502883911, "learning_rate": 1.038e-05, "loss": 0.5376, "step": 520 }, { "grad_norm": 2.578920602798462, "learning_rate": 1.058e-05, "loss": 0.547, "step": 530 }, { "grad_norm": 1.6801835298538208, "learning_rate": 1.0780000000000002e-05, "loss": 0.5057, "step": 540 }, { "grad_norm": 2.6798758506774902, "learning_rate": 1.098e-05, "loss": 0.5241, "step": 550 }, { "grad_norm": 2.538792610168457, "learning_rate": 1.118e-05, "loss": 0.5069, "step": 560 }, { "grad_norm": 1.9280446767807007, "learning_rate": 1.1380000000000001e-05, "loss": 0.4547, "step": 570 }, { "grad_norm": 2.4564476013183594, "learning_rate": 1.1580000000000001e-05, "loss": 0.4401, "step": 580 }, { "grad_norm": 2.6823832988739014, "learning_rate": 1.178e-05, "loss": 0.4476, "step": 590 }, { "grad_norm": 2.351358652114868, "learning_rate": 1.198e-05, "loss": 0.4359, "step": 600 }, { "grad_norm": 2.232100009918213, "learning_rate": 1.2180000000000002e-05, "loss": 0.3977, "step": 610 }, { "grad_norm": 2.155564546585083, "learning_rate": 1.238e-05, "loss": 0.3948, "step": 620 }, { "grad_norm": 2.464561700820923, "learning_rate": 1.258e-05, "loss": 0.3953, "step": 630 }, { "grad_norm": 2.942869186401367, "learning_rate": 1.278e-05, "loss": 0.3833, "step": 640 }, { "grad_norm": 3.03479266166687, "learning_rate": 1.2980000000000001e-05, "loss": 0.3561, "step": 650 }, { "grad_norm": 2.9595820903778076, "learning_rate": 1.3180000000000001e-05, "loss": 0.3557, "step": 660 }, { "grad_norm": 1.838884711265564, "learning_rate": 1.338e-05, "loss": 0.3167, "step": 670 }, { "grad_norm": 2.79404616355896, "learning_rate": 1.358e-05, "loss": 0.3207, "step": 680 }, { "grad_norm": 2.477296829223633, "learning_rate": 1.3780000000000002e-05, "loss": 0.3024, "step": 690 }, { "grad_norm": 3.115823268890381, "learning_rate": 1.3980000000000002e-05, "loss": 0.3056, "step": 700 }, { "grad_norm": 2.735769271850586, "learning_rate": 1.4180000000000001e-05, "loss": 0.272, "step": 710 }, { "grad_norm": 3.117103099822998, "learning_rate": 1.4380000000000001e-05, "loss": 0.2747, "step": 720 }, { "grad_norm": 2.6757118701934814, "learning_rate": 1.4580000000000003e-05, "loss": 0.2549, "step": 730 }, { "grad_norm": 2.493833065032959, "learning_rate": 1.4779999999999999e-05, "loss": 0.2201, "step": 740 }, { "grad_norm": 2.2451515197753906, "learning_rate": 1.4979999999999999e-05, "loss": 0.2389, "step": 750 }, { "grad_norm": 1.8285349607467651, "learning_rate": 1.518e-05, "loss": 0.228, "step": 760 }, { "grad_norm": 2.2848563194274902, "learning_rate": 1.538e-05, "loss": 0.2203, "step": 770 }, { "grad_norm": 2.1363701820373535, "learning_rate": 1.558e-05, "loss": 0.2448, "step": 780 }, { "grad_norm": 2.1553306579589844, "learning_rate": 1.578e-05, "loss": 0.2, "step": 790 }, { "grad_norm": 2.4751365184783936, "learning_rate": 1.598e-05, "loss": 0.1876, "step": 800 }, { "grad_norm": 2.78115177154541, "learning_rate": 1.618e-05, "loss": 0.192, "step": 810 }, { "grad_norm": 1.955132007598877, "learning_rate": 1.6380000000000002e-05, "loss": 0.1901, "step": 820 }, { "grad_norm": 2.3743464946746826, "learning_rate": 1.658e-05, "loss": 0.1825, "step": 830 }, { "grad_norm": 2.0151000022888184, "learning_rate": 1.6780000000000002e-05, "loss": 0.1722, "step": 840 }, { "grad_norm": 1.8431248664855957, "learning_rate": 1.698e-05, "loss": 0.1789, "step": 850 }, { "grad_norm": 2.6401195526123047, "learning_rate": 1.718e-05, "loss": 0.1526, "step": 860 }, { "grad_norm": 2.4224345684051514, "learning_rate": 1.7380000000000003e-05, "loss": 0.1788, "step": 870 }, { "grad_norm": 2.9957642555236816, "learning_rate": 1.758e-05, "loss": 0.1517, "step": 880 }, { "grad_norm": 2.163836717605591, "learning_rate": 1.7780000000000003e-05, "loss": 0.1564, "step": 890 }, { "grad_norm": 1.9706273078918457, "learning_rate": 1.798e-05, "loss": 0.1495, "step": 900 }, { "grad_norm": 1.3324317932128906, "learning_rate": 1.818e-05, "loss": 0.143, "step": 910 }, { "grad_norm": 1.9211949110031128, "learning_rate": 1.838e-05, "loss": 0.1249, "step": 920 }, { "grad_norm": 1.936745047569275, "learning_rate": 1.858e-05, "loss": 0.1413, "step": 930 }, { "grad_norm": 2.6475751399993896, "learning_rate": 1.878e-05, "loss": 0.1451, "step": 940 }, { "grad_norm": 2.7994847297668457, "learning_rate": 1.898e-05, "loss": 0.1295, "step": 950 }, { "grad_norm": 2.1083269119262695, "learning_rate": 1.918e-05, "loss": 0.1353, "step": 960 }, { "grad_norm": 2.277811050415039, "learning_rate": 1.938e-05, "loss": 0.162, "step": 970 }, { "grad_norm": 1.83091139793396, "learning_rate": 1.9580000000000002e-05, "loss": 0.1166, "step": 980 }, { "grad_norm": 2.461174726486206, "learning_rate": 1.978e-05, "loss": 0.1142, "step": 990 }, { "grad_norm": 1.9571685791015625, "learning_rate": 1.9980000000000002e-05, "loss": 0.1199, "step": 1000 }, { "grad_norm": 1.3510910272598267, "learning_rate": 2.0180000000000003e-05, "loss": 0.1135, "step": 1010 }, { "grad_norm": 2.2749392986297607, "learning_rate": 2.038e-05, "loss": 0.1403, "step": 1020 }, { "grad_norm": 1.5413285493850708, "learning_rate": 2.0580000000000003e-05, "loss": 0.1073, "step": 1030 }, { "grad_norm": 2.104311227798462, "learning_rate": 2.078e-05, "loss": 0.1201, "step": 1040 }, { "grad_norm": 2.4890267848968506, "learning_rate": 2.098e-05, "loss": 0.1284, "step": 1050 }, { "grad_norm": 2.1119751930236816, "learning_rate": 2.118e-05, "loss": 0.136, "step": 1060 }, { "grad_norm": 2.0609400272369385, "learning_rate": 2.138e-05, "loss": 0.1422, "step": 1070 }, { "grad_norm": 2.0861833095550537, "learning_rate": 2.158e-05, "loss": 0.1467, "step": 1080 }, { "grad_norm": 1.7478619813919067, "learning_rate": 2.178e-05, "loss": 0.1226, "step": 1090 }, { "grad_norm": 2.1601767539978027, "learning_rate": 2.198e-05, "loss": 0.101, "step": 1100 }, { "grad_norm": 2.3549458980560303, "learning_rate": 2.218e-05, "loss": 0.151, "step": 1110 }, { "grad_norm": 1.9541536569595337, "learning_rate": 2.2380000000000003e-05, "loss": 0.1022, "step": 1120 }, { "grad_norm": 1.4987444877624512, "learning_rate": 2.258e-05, "loss": 0.1203, "step": 1130 }, { "grad_norm": 2.3660426139831543, "learning_rate": 2.2780000000000002e-05, "loss": 0.1582, "step": 1140 }, { "grad_norm": 2.187920570373535, "learning_rate": 2.298e-05, "loss": 0.1397, "step": 1150 }, { "grad_norm": 1.5966957807540894, "learning_rate": 2.318e-05, "loss": 0.1187, "step": 1160 }, { "grad_norm": 1.771752953529358, "learning_rate": 2.3380000000000003e-05, "loss": 0.1192, "step": 1170 }, { "grad_norm": 1.8625659942626953, "learning_rate": 2.358e-05, "loss": 0.1021, "step": 1180 }, { "grad_norm": 2.1129608154296875, "learning_rate": 2.3780000000000003e-05, "loss": 0.1215, "step": 1190 }, { "grad_norm": 4.023634433746338, "learning_rate": 2.398e-05, "loss": 0.1333, "step": 1200 }, { "grad_norm": 4.016610622406006, "learning_rate": 2.418e-05, "loss": 0.1476, "step": 1210 }, { "grad_norm": 1.625063419342041, "learning_rate": 2.438e-05, "loss": 0.1169, "step": 1220 }, { "grad_norm": 2.064836263656616, "learning_rate": 2.4580000000000002e-05, "loss": 0.1189, "step": 1230 }, { "grad_norm": 1.7055710554122925, "learning_rate": 2.478e-05, "loss": 0.1263, "step": 1240 }, { "grad_norm": 2.2800254821777344, "learning_rate": 2.498e-05, "loss": 0.103, "step": 1250 }, { "grad_norm": 2.2383344173431396, "learning_rate": 2.5180000000000003e-05, "loss": 0.1158, "step": 1260 }, { "grad_norm": 1.389495849609375, "learning_rate": 2.5380000000000004e-05, "loss": 0.0944, "step": 1270 }, { "grad_norm": 1.6747606992721558, "learning_rate": 2.5580000000000002e-05, "loss": 0.1176, "step": 1280 }, { "grad_norm": 1.481675624847412, "learning_rate": 2.5779999999999997e-05, "loss": 0.1151, "step": 1290 }, { "grad_norm": 1.34571373462677, "learning_rate": 2.598e-05, "loss": 0.1048, "step": 1300 }, { "grad_norm": 1.3815200328826904, "learning_rate": 2.618e-05, "loss": 0.1071, "step": 1310 }, { "grad_norm": 1.5357588529586792, "learning_rate": 2.6379999999999998e-05, "loss": 0.112, "step": 1320 }, { "grad_norm": 1.8370627164840698, "learning_rate": 2.658e-05, "loss": 0.0932, "step": 1330 }, { "grad_norm": 1.5234593152999878, "learning_rate": 2.678e-05, "loss": 0.0917, "step": 1340 }, { "grad_norm": 1.7394421100616455, "learning_rate": 2.698e-05, "loss": 0.1196, "step": 1350 }, { "grad_norm": 2.702491283416748, "learning_rate": 2.718e-05, "loss": 0.114, "step": 1360 }, { "grad_norm": 2.011516571044922, "learning_rate": 2.738e-05, "loss": 0.0894, "step": 1370 }, { "grad_norm": 1.5570334196090698, "learning_rate": 2.758e-05, "loss": 0.0856, "step": 1380 }, { "grad_norm": 1.8120388984680176, "learning_rate": 2.778e-05, "loss": 0.1016, "step": 1390 }, { "grad_norm": 1.8195563554763794, "learning_rate": 2.798e-05, "loss": 0.0869, "step": 1400 }, { "grad_norm": 1.4312589168548584, "learning_rate": 2.818e-05, "loss": 0.1016, "step": 1410 }, { "grad_norm": 1.8849586248397827, "learning_rate": 2.8380000000000003e-05, "loss": 0.1017, "step": 1420 }, { "grad_norm": 1.2504222393035889, "learning_rate": 2.858e-05, "loss": 0.0792, "step": 1430 }, { "grad_norm": 1.226036548614502, "learning_rate": 2.8780000000000002e-05, "loss": 0.104, "step": 1440 }, { "grad_norm": 1.414929747581482, "learning_rate": 2.898e-05, "loss": 0.1159, "step": 1450 }, { "grad_norm": 1.372789740562439, "learning_rate": 2.9180000000000002e-05, "loss": 0.1039, "step": 1460 }, { "grad_norm": 1.2401673793792725, "learning_rate": 2.9380000000000003e-05, "loss": 0.1143, "step": 1470 }, { "grad_norm": 1.5954829454421997, "learning_rate": 2.958e-05, "loss": 0.0943, "step": 1480 }, { "grad_norm": 1.263228178024292, "learning_rate": 2.9780000000000003e-05, "loss": 0.0946, "step": 1490 }, { "grad_norm": 1.3786497116088867, "learning_rate": 2.998e-05, "loss": 0.1104, "step": 1500 }, { "grad_norm": 1.401642084121704, "learning_rate": 3.0180000000000002e-05, "loss": 0.094, "step": 1510 }, { "grad_norm": 1.6227985620498657, "learning_rate": 3.0380000000000004e-05, "loss": 0.0996, "step": 1520 }, { "grad_norm": 1.100611925125122, "learning_rate": 3.058e-05, "loss": 0.1015, "step": 1530 }, { "grad_norm": 1.141063928604126, "learning_rate": 3.078e-05, "loss": 0.0978, "step": 1540 }, { "grad_norm": 1.3272008895874023, "learning_rate": 3.0980000000000005e-05, "loss": 0.0862, "step": 1550 }, { "grad_norm": 0.9812951683998108, "learning_rate": 3.118e-05, "loss": 0.1132, "step": 1560 }, { "grad_norm": 1.4331419467926025, "learning_rate": 3.138e-05, "loss": 0.0832, "step": 1570 }, { "grad_norm": 1.573462724685669, "learning_rate": 3.1580000000000006e-05, "loss": 0.0931, "step": 1580 }, { "grad_norm": 1.3922439813613892, "learning_rate": 3.1780000000000004e-05, "loss": 0.1091, "step": 1590 }, { "grad_norm": 1.2853431701660156, "learning_rate": 3.198e-05, "loss": 0.0934, "step": 1600 }, { "grad_norm": 1.0682400465011597, "learning_rate": 3.218e-05, "loss": 0.074, "step": 1610 }, { "grad_norm": 1.3752813339233398, "learning_rate": 3.238e-05, "loss": 0.0943, "step": 1620 }, { "grad_norm": 1.2353447675704956, "learning_rate": 3.2579999999999996e-05, "loss": 0.121, "step": 1630 }, { "grad_norm": 1.0755605697631836, "learning_rate": 3.278e-05, "loss": 0.1039, "step": 1640 }, { "grad_norm": 1.484345555305481, "learning_rate": 3.298e-05, "loss": 0.0969, "step": 1650 }, { "grad_norm": 1.1352086067199707, "learning_rate": 3.318e-05, "loss": 0.1019, "step": 1660 }, { "grad_norm": 1.2061045169830322, "learning_rate": 3.338e-05, "loss": 0.0734, "step": 1670 }, { "grad_norm": 1.2511520385742188, "learning_rate": 3.358e-05, "loss": 0.0847, "step": 1680 }, { "grad_norm": 1.172087550163269, "learning_rate": 3.378e-05, "loss": 0.0873, "step": 1690 }, { "grad_norm": 1.0568056106567383, "learning_rate": 3.398e-05, "loss": 0.0943, "step": 1700 }, { "grad_norm": 1.9926527738571167, "learning_rate": 3.418e-05, "loss": 0.1181, "step": 1710 }, { "grad_norm": 1.5254887342453003, "learning_rate": 3.438e-05, "loss": 0.0931, "step": 1720 }, { "grad_norm": 1.379210114479065, "learning_rate": 3.4580000000000004e-05, "loss": 0.0884, "step": 1730 }, { "grad_norm": 1.3378459215164185, "learning_rate": 3.478e-05, "loss": 0.0828, "step": 1740 }, { "grad_norm": 1.253860354423523, "learning_rate": 3.498e-05, "loss": 0.0941, "step": 1750 }, { "grad_norm": 2.0348479747772217, "learning_rate": 3.518e-05, "loss": 0.0965, "step": 1760 }, { "grad_norm": 1.2796845436096191, "learning_rate": 3.5380000000000003e-05, "loss": 0.0796, "step": 1770 }, { "grad_norm": 1.3572660684585571, "learning_rate": 3.558e-05, "loss": 0.1069, "step": 1780 }, { "grad_norm": 1.6282320022583008, "learning_rate": 3.578e-05, "loss": 0.0852, "step": 1790 }, { "grad_norm": 1.1375030279159546, "learning_rate": 3.5980000000000004e-05, "loss": 0.0885, "step": 1800 }, { "grad_norm": 1.0767956972122192, "learning_rate": 3.618e-05, "loss": 0.0781, "step": 1810 }, { "grad_norm": 1.8363488912582397, "learning_rate": 3.638e-05, "loss": 0.0989, "step": 1820 }, { "grad_norm": 1.751368761062622, "learning_rate": 3.6580000000000006e-05, "loss": 0.0923, "step": 1830 }, { "grad_norm": 1.1684935092926025, "learning_rate": 3.6780000000000004e-05, "loss": 0.0989, "step": 1840 }, { "grad_norm": 1.1917624473571777, "learning_rate": 3.698e-05, "loss": 0.1153, "step": 1850 }, { "grad_norm": 1.1194264888763428, "learning_rate": 3.7180000000000007e-05, "loss": 0.1086, "step": 1860 }, { "grad_norm": 1.0198911428451538, "learning_rate": 3.7380000000000005e-05, "loss": 0.0863, "step": 1870 }, { "grad_norm": 1.4159961938858032, "learning_rate": 3.758e-05, "loss": 0.0715, "step": 1880 }, { "grad_norm": 1.143539547920227, "learning_rate": 3.778000000000001e-05, "loss": 0.0843, "step": 1890 }, { "grad_norm": 0.7961069941520691, "learning_rate": 3.7980000000000006e-05, "loss": 0.0764, "step": 1900 }, { "grad_norm": 1.6038621664047241, "learning_rate": 3.818e-05, "loss": 0.1124, "step": 1910 }, { "grad_norm": 1.124564528465271, "learning_rate": 3.838e-05, "loss": 0.0885, "step": 1920 }, { "grad_norm": 1.0287461280822754, "learning_rate": 3.858e-05, "loss": 0.0778, "step": 1930 }, { "grad_norm": 1.2970155477523804, "learning_rate": 3.878e-05, "loss": 0.0892, "step": 1940 }, { "grad_norm": 1.0183141231536865, "learning_rate": 3.898e-05, "loss": 0.0713, "step": 1950 }, { "grad_norm": 1.7291549444198608, "learning_rate": 3.918e-05, "loss": 0.1031, "step": 1960 }, { "grad_norm": 1.0613744258880615, "learning_rate": 3.938e-05, "loss": 0.0704, "step": 1970 }, { "grad_norm": 1.7015715837478638, "learning_rate": 3.958e-05, "loss": 0.1, "step": 1980 }, { "grad_norm": 1.243877649307251, "learning_rate": 3.978e-05, "loss": 0.0754, "step": 1990 }, { "grad_norm": 0.9385055899620056, "learning_rate": 3.998e-05, "loss": 0.087, "step": 2000 }, { "grad_norm": 1.3054912090301514, "learning_rate": 4.018e-05, "loss": 0.1131, "step": 2010 }, { "grad_norm": 1.618214726448059, "learning_rate": 4.038e-05, "loss": 0.0906, "step": 2020 }, { "grad_norm": 1.4397448301315308, "learning_rate": 4.058e-05, "loss": 0.0707, "step": 2030 }, { "grad_norm": 0.9030084609985352, "learning_rate": 4.078e-05, "loss": 0.0906, "step": 2040 }, { "grad_norm": 1.160494089126587, "learning_rate": 4.0980000000000004e-05, "loss": 0.0885, "step": 2050 }, { "grad_norm": 1.7339483499526978, "learning_rate": 4.118e-05, "loss": 0.0782, "step": 2060 }, { "grad_norm": 0.810237467288971, "learning_rate": 4.138e-05, "loss": 0.0739, "step": 2070 }, { "grad_norm": 1.120898723602295, "learning_rate": 4.1580000000000005e-05, "loss": 0.0801, "step": 2080 }, { "grad_norm": 0.7503056526184082, "learning_rate": 4.178e-05, "loss": 0.08, "step": 2090 }, { "grad_norm": 1.1162068843841553, "learning_rate": 4.198e-05, "loss": 0.0809, "step": 2100 }, { "grad_norm": 1.0183351039886475, "learning_rate": 4.2180000000000006e-05, "loss": 0.1045, "step": 2110 }, { "grad_norm": 1.0034611225128174, "learning_rate": 4.2380000000000004e-05, "loss": 0.0793, "step": 2120 }, { "grad_norm": 1.0283640623092651, "learning_rate": 4.258e-05, "loss": 0.0789, "step": 2130 }, { "grad_norm": 0.7836875915527344, "learning_rate": 4.278e-05, "loss": 0.0595, "step": 2140 }, { "grad_norm": 0.8345230221748352, "learning_rate": 4.2980000000000005e-05, "loss": 0.0905, "step": 2150 }, { "grad_norm": 0.871344804763794, "learning_rate": 4.318e-05, "loss": 0.0736, "step": 2160 }, { "grad_norm": 0.7854212522506714, "learning_rate": 4.338e-05, "loss": 0.0701, "step": 2170 }, { "grad_norm": 0.9391396045684814, "learning_rate": 4.3580000000000006e-05, "loss": 0.0878, "step": 2180 }, { "grad_norm": 1.033511757850647, "learning_rate": 4.3780000000000004e-05, "loss": 0.0736, "step": 2190 }, { "grad_norm": 1.0480092763900757, "learning_rate": 4.398e-05, "loss": 0.0726, "step": 2200 }, { "grad_norm": 0.8528473377227783, "learning_rate": 4.418000000000001e-05, "loss": 0.0773, "step": 2210 }, { "grad_norm": 1.1313806772232056, "learning_rate": 4.438e-05, "loss": 0.0711, "step": 2220 }, { "grad_norm": 1.063657283782959, "learning_rate": 4.458e-05, "loss": 0.0782, "step": 2230 }, { "grad_norm": 0.9745526909828186, "learning_rate": 4.478e-05, "loss": 0.0702, "step": 2240 }, { "grad_norm": 0.7768816351890564, "learning_rate": 4.498e-05, "loss": 0.0845, "step": 2250 }, { "grad_norm": 0.8323822617530823, "learning_rate": 4.518e-05, "loss": 0.0672, "step": 2260 }, { "grad_norm": 0.7043272852897644, "learning_rate": 4.538e-05, "loss": 0.0645, "step": 2270 }, { "grad_norm": 0.8758164644241333, "learning_rate": 4.558e-05, "loss": 0.0668, "step": 2280 }, { "grad_norm": 1.6384285688400269, "learning_rate": 4.578e-05, "loss": 0.0992, "step": 2290 }, { "grad_norm": 0.8822786808013916, "learning_rate": 4.5980000000000004e-05, "loss": 0.0639, "step": 2300 }, { "grad_norm": 0.8866972327232361, "learning_rate": 4.618e-05, "loss": 0.0727, "step": 2310 }, { "grad_norm": 0.9160493612289429, "learning_rate": 4.638e-05, "loss": 0.0622, "step": 2320 }, { "grad_norm": 0.9186035394668579, "learning_rate": 4.6580000000000005e-05, "loss": 0.0837, "step": 2330 }, { "grad_norm": 0.8636658787727356, "learning_rate": 4.678e-05, "loss": 0.0722, "step": 2340 }, { "grad_norm": 0.921385645866394, "learning_rate": 4.698e-05, "loss": 0.0734, "step": 2350 }, { "grad_norm": 0.8534148335456848, "learning_rate": 4.718e-05, "loss": 0.077, "step": 2360 }, { "grad_norm": 0.8480973839759827, "learning_rate": 4.7380000000000004e-05, "loss": 0.0733, "step": 2370 }, { "grad_norm": 0.7912062406539917, "learning_rate": 4.758e-05, "loss": 0.0826, "step": 2380 }, { "grad_norm": 0.9433387517929077, "learning_rate": 4.778e-05, "loss": 0.0886, "step": 2390 }, { "grad_norm": 0.7897411584854126, "learning_rate": 4.7980000000000005e-05, "loss": 0.0788, "step": 2400 }, { "grad_norm": 0.7765304446220398, "learning_rate": 4.818e-05, "loss": 0.0732, "step": 2410 }, { "grad_norm": 0.7294731140136719, "learning_rate": 4.838e-05, "loss": 0.1094, "step": 2420 }, { "grad_norm": 0.8844425082206726, "learning_rate": 4.8580000000000006e-05, "loss": 0.0784, "step": 2430 }, { "grad_norm": 1.166480302810669, "learning_rate": 4.8780000000000004e-05, "loss": 0.0836, "step": 2440 }, { "grad_norm": 1.1558499336242676, "learning_rate": 4.898e-05, "loss": 0.0684, "step": 2450 }, { "grad_norm": 0.7138095498085022, "learning_rate": 4.918000000000001e-05, "loss": 0.0642, "step": 2460 }, { "grad_norm": 0.8647704720497131, "learning_rate": 4.9380000000000005e-05, "loss": 0.0654, "step": 2470 }, { "grad_norm": 0.8062322735786438, "learning_rate": 4.958e-05, "loss": 0.0873, "step": 2480 }, { "grad_norm": 0.6568583250045776, "learning_rate": 4.978e-05, "loss": 0.0662, "step": 2490 }, { "grad_norm": 0.81143718957901, "learning_rate": 4.9980000000000006e-05, "loss": 0.0592, "step": 2500 }, { "grad_norm": 0.7644357681274414, "learning_rate": 5.0180000000000004e-05, "loss": 0.06, "step": 2510 }, { "grad_norm": 0.7200099229812622, "learning_rate": 5.038e-05, "loss": 0.0682, "step": 2520 }, { "grad_norm": 0.651469886302948, "learning_rate": 5.058000000000001e-05, "loss": 0.0572, "step": 2530 }, { "grad_norm": 0.9762552976608276, "learning_rate": 5.0780000000000005e-05, "loss": 0.0673, "step": 2540 }, { "grad_norm": 0.7394102215766907, "learning_rate": 5.098e-05, "loss": 0.0751, "step": 2550 }, { "grad_norm": 0.7801147103309631, "learning_rate": 5.118000000000001e-05, "loss": 0.072, "step": 2560 }, { "grad_norm": 0.7554254531860352, "learning_rate": 5.1380000000000006e-05, "loss": 0.0696, "step": 2570 }, { "grad_norm": 0.7226750254631042, "learning_rate": 5.1580000000000004e-05, "loss": 0.0681, "step": 2580 }, { "grad_norm": 0.6029090881347656, "learning_rate": 5.178000000000001e-05, "loss": 0.0642, "step": 2590 }, { "grad_norm": 0.9211573600769043, "learning_rate": 5.198000000000001e-05, "loss": 0.0734, "step": 2600 }, { "grad_norm": 1.0821454524993896, "learning_rate": 5.2180000000000005e-05, "loss": 0.077, "step": 2610 }, { "grad_norm": 0.6425203084945679, "learning_rate": 5.238000000000001e-05, "loss": 0.0601, "step": 2620 }, { "grad_norm": 0.8494429588317871, "learning_rate": 5.258000000000001e-05, "loss": 0.0644, "step": 2630 }, { "grad_norm": 0.9872174263000488, "learning_rate": 5.2780000000000006e-05, "loss": 0.0806, "step": 2640 }, { "grad_norm": 0.7250400185585022, "learning_rate": 5.2980000000000004e-05, "loss": 0.0788, "step": 2650 }, { "grad_norm": 0.9690873622894287, "learning_rate": 5.318000000000001e-05, "loss": 0.0818, "step": 2660 }, { "grad_norm": 1.0315600633621216, "learning_rate": 5.338000000000001e-05, "loss": 0.0808, "step": 2670 }, { "grad_norm": 0.9829235672950745, "learning_rate": 5.3580000000000005e-05, "loss": 0.0965, "step": 2680 }, { "grad_norm": 0.8396091461181641, "learning_rate": 5.378e-05, "loss": 0.0691, "step": 2690 }, { "grad_norm": 0.6611124873161316, "learning_rate": 5.3979999999999995e-05, "loss": 0.0607, "step": 2700 }, { "grad_norm": 0.6435964703559875, "learning_rate": 5.418e-05, "loss": 0.0686, "step": 2710 }, { "grad_norm": 0.7727639079093933, "learning_rate": 5.438e-05, "loss": 0.0862, "step": 2720 }, { "grad_norm": 0.7547170519828796, "learning_rate": 5.4579999999999996e-05, "loss": 0.0695, "step": 2730 }, { "grad_norm": 0.7094339728355408, "learning_rate": 5.478e-05, "loss": 0.0635, "step": 2740 }, { "grad_norm": 0.8265712261199951, "learning_rate": 5.498e-05, "loss": 0.0739, "step": 2750 }, { "grad_norm": 0.8957444429397583, "learning_rate": 5.518e-05, "loss": 0.0663, "step": 2760 }, { "grad_norm": 0.7515813112258911, "learning_rate": 5.538e-05, "loss": 0.0668, "step": 2770 }, { "grad_norm": 0.5553003549575806, "learning_rate": 5.558e-05, "loss": 0.0719, "step": 2780 }, { "grad_norm": 0.7900295257568359, "learning_rate": 5.578e-05, "loss": 0.0708, "step": 2790 }, { "grad_norm": 0.712753415107727, "learning_rate": 5.5979999999999996e-05, "loss": 0.0617, "step": 2800 }, { "grad_norm": 1.4845716953277588, "learning_rate": 5.618e-05, "loss": 0.0954, "step": 2810 }, { "grad_norm": 1.0098481178283691, "learning_rate": 5.638e-05, "loss": 0.0801, "step": 2820 }, { "grad_norm": 0.7350061535835266, "learning_rate": 5.658e-05, "loss": 0.0645, "step": 2830 }, { "grad_norm": 0.8492476344108582, "learning_rate": 5.678e-05, "loss": 0.0863, "step": 2840 }, { "grad_norm": 0.6186723709106445, "learning_rate": 5.698e-05, "loss": 0.0788, "step": 2850 }, { "grad_norm": 0.6812577247619629, "learning_rate": 5.718e-05, "loss": 0.0666, "step": 2860 }, { "grad_norm": 0.7062024474143982, "learning_rate": 5.738e-05, "loss": 0.0603, "step": 2870 }, { "grad_norm": 0.5730915069580078, "learning_rate": 5.758e-05, "loss": 0.0554, "step": 2880 }, { "grad_norm": 0.8036779761314392, "learning_rate": 5.778e-05, "loss": 0.0631, "step": 2890 }, { "grad_norm": 0.529093861579895, "learning_rate": 5.7980000000000004e-05, "loss": 0.0754, "step": 2900 }, { "grad_norm": 0.940301239490509, "learning_rate": 5.818e-05, "loss": 0.0594, "step": 2910 }, { "grad_norm": 0.6929938793182373, "learning_rate": 5.838e-05, "loss": 0.0652, "step": 2920 }, { "grad_norm": 0.8173381686210632, "learning_rate": 5.858e-05, "loss": 0.0501, "step": 2930 }, { "grad_norm": 0.6694719195365906, "learning_rate": 5.878e-05, "loss": 0.054, "step": 2940 }, { "grad_norm": 0.6085557341575623, "learning_rate": 5.898e-05, "loss": 0.0513, "step": 2950 }, { "grad_norm": 0.6042953729629517, "learning_rate": 5.918e-05, "loss": 0.0676, "step": 2960 }, { "grad_norm": 0.6950425505638123, "learning_rate": 5.9380000000000004e-05, "loss": 0.055, "step": 2970 }, { "grad_norm": 0.7471460700035095, "learning_rate": 5.958e-05, "loss": 0.0656, "step": 2980 }, { "grad_norm": 0.8791922330856323, "learning_rate": 5.978e-05, "loss": 0.0733, "step": 2990 }, { "grad_norm": 0.5397282242774963, "learning_rate": 5.9980000000000005e-05, "loss": 0.0723, "step": 3000 }, { "grad_norm": 0.5662893056869507, "learning_rate": 6.018e-05, "loss": 0.0531, "step": 3010 }, { "grad_norm": 0.7680960893630981, "learning_rate": 6.038e-05, "loss": 0.0595, "step": 3020 }, { "grad_norm": 0.6801112294197083, "learning_rate": 6.0580000000000006e-05, "loss": 0.0625, "step": 3030 }, { "grad_norm": 0.6374152898788452, "learning_rate": 6.0780000000000004e-05, "loss": 0.0638, "step": 3040 }, { "grad_norm": 0.7891551852226257, "learning_rate": 6.098e-05, "loss": 0.0629, "step": 3050 }, { "grad_norm": 0.6096785664558411, "learning_rate": 6.118000000000001e-05, "loss": 0.068, "step": 3060 }, { "grad_norm": 0.5144187211990356, "learning_rate": 6.138e-05, "loss": 0.0502, "step": 3070 }, { "grad_norm": 0.6750665903091431, "learning_rate": 6.158e-05, "loss": 0.0616, "step": 3080 }, { "grad_norm": 1.11825430393219, "learning_rate": 6.178000000000001e-05, "loss": 0.0834, "step": 3090 }, { "grad_norm": 0.8286563754081726, "learning_rate": 6.198e-05, "loss": 0.0607, "step": 3100 }, { "grad_norm": 0.5850666165351868, "learning_rate": 6.218e-05, "loss": 0.052, "step": 3110 }, { "grad_norm": 0.703693687915802, "learning_rate": 6.238000000000001e-05, "loss": 0.0527, "step": 3120 }, { "grad_norm": 0.6720925569534302, "learning_rate": 6.258e-05, "loss": 0.0543, "step": 3130 }, { "grad_norm": 0.6664562821388245, "learning_rate": 6.278e-05, "loss": 0.0617, "step": 3140 }, { "grad_norm": 0.618125855922699, "learning_rate": 6.298000000000001e-05, "loss": 0.054, "step": 3150 }, { "grad_norm": 0.5815813541412354, "learning_rate": 6.318e-05, "loss": 0.0623, "step": 3160 }, { "grad_norm": 0.6762789487838745, "learning_rate": 6.338e-05, "loss": 0.0541, "step": 3170 }, { "grad_norm": 0.6341826319694519, "learning_rate": 6.358000000000001e-05, "loss": 0.0734, "step": 3180 }, { "grad_norm": 0.6021322011947632, "learning_rate": 6.378e-05, "loss": 0.0695, "step": 3190 }, { "grad_norm": 0.788998544216156, "learning_rate": 6.398000000000001e-05, "loss": 0.0701, "step": 3200 }, { "grad_norm": 0.5740426778793335, "learning_rate": 6.418000000000001e-05, "loss": 0.0555, "step": 3210 }, { "grad_norm": 1.0790995359420776, "learning_rate": 6.438e-05, "loss": 0.0837, "step": 3220 }, { "grad_norm": 0.7256030440330505, "learning_rate": 6.458000000000001e-05, "loss": 0.0742, "step": 3230 }, { "grad_norm": 0.6337389945983887, "learning_rate": 6.478000000000001e-05, "loss": 0.0601, "step": 3240 }, { "grad_norm": 0.49241265654563904, "learning_rate": 6.498e-05, "loss": 0.0675, "step": 3250 }, { "grad_norm": 0.8465797305107117, "learning_rate": 6.518000000000001e-05, "loss": 0.0632, "step": 3260 }, { "grad_norm": 0.9190468788146973, "learning_rate": 6.538000000000001e-05, "loss": 0.0621, "step": 3270 }, { "grad_norm": 0.6437724232673645, "learning_rate": 6.558e-05, "loss": 0.0527, "step": 3280 }, { "grad_norm": 0.6910668015480042, "learning_rate": 6.578000000000001e-05, "loss": 0.0545, "step": 3290 }, { "grad_norm": 0.5507031083106995, "learning_rate": 6.598e-05, "loss": 0.0545, "step": 3300 }, { "grad_norm": 0.6231828927993774, "learning_rate": 6.618e-05, "loss": 0.0676, "step": 3310 }, { "grad_norm": 0.6373624801635742, "learning_rate": 6.638e-05, "loss": 0.0783, "step": 3320 }, { "grad_norm": 0.6829156875610352, "learning_rate": 6.658e-05, "loss": 0.0603, "step": 3330 }, { "grad_norm": 0.7786504626274109, "learning_rate": 6.678e-05, "loss": 0.0603, "step": 3340 }, { "grad_norm": 0.5648596882820129, "learning_rate": 6.698e-05, "loss": 0.0621, "step": 3350 }, { "grad_norm": 0.9974026679992676, "learning_rate": 6.718e-05, "loss": 0.059, "step": 3360 }, { "grad_norm": 0.6867991089820862, "learning_rate": 6.738e-05, "loss": 0.0622, "step": 3370 }, { "grad_norm": 0.5894598960876465, "learning_rate": 6.758e-05, "loss": 0.0995, "step": 3380 }, { "grad_norm": 0.5104876160621643, "learning_rate": 6.778e-05, "loss": 0.0567, "step": 3390 }, { "grad_norm": 0.5778339505195618, "learning_rate": 6.798e-05, "loss": 0.0629, "step": 3400 }, { "grad_norm": 0.6494420170783997, "learning_rate": 6.818e-05, "loss": 0.0571, "step": 3410 }, { "grad_norm": 0.6783077716827393, "learning_rate": 6.838e-05, "loss": 0.0532, "step": 3420 }, { "grad_norm": 0.6452813148498535, "learning_rate": 6.858e-05, "loss": 0.0555, "step": 3430 }, { "grad_norm": 0.4845317602157593, "learning_rate": 6.878e-05, "loss": 0.0683, "step": 3440 }, { "grad_norm": 0.6348721981048584, "learning_rate": 6.898e-05, "loss": 0.0574, "step": 3450 }, { "grad_norm": 0.49397552013397217, "learning_rate": 6.918e-05, "loss": 0.0501, "step": 3460 }, { "grad_norm": 0.5389422178268433, "learning_rate": 6.938e-05, "loss": 0.065, "step": 3470 }, { "grad_norm": 0.558595597743988, "learning_rate": 6.958e-05, "loss": 0.0656, "step": 3480 }, { "grad_norm": 0.6908465623855591, "learning_rate": 6.978e-05, "loss": 0.0623, "step": 3490 }, { "grad_norm": 0.5981249213218689, "learning_rate": 6.998e-05, "loss": 0.0466, "step": 3500 }, { "grad_norm": 0.41275912523269653, "learning_rate": 7.018e-05, "loss": 0.0564, "step": 3510 }, { "grad_norm": 0.7131398916244507, "learning_rate": 7.038e-05, "loss": 0.0552, "step": 3520 }, { "grad_norm": 0.7148188352584839, "learning_rate": 7.058e-05, "loss": 0.0564, "step": 3530 }, { "grad_norm": 0.7656156420707703, "learning_rate": 7.078e-05, "loss": 0.0538, "step": 3540 }, { "grad_norm": 0.6428103446960449, "learning_rate": 7.098e-05, "loss": 0.0604, "step": 3550 }, { "grad_norm": 0.5123987197875977, "learning_rate": 7.118e-05, "loss": 0.0464, "step": 3560 }, { "grad_norm": 0.5153019428253174, "learning_rate": 7.138e-05, "loss": 0.0596, "step": 3570 }, { "grad_norm": 0.7362534999847412, "learning_rate": 7.158e-05, "loss": 0.0699, "step": 3580 }, { "grad_norm": 0.6907706260681152, "learning_rate": 7.178000000000001e-05, "loss": 0.0691, "step": 3590 }, { "grad_norm": 0.5810971260070801, "learning_rate": 7.198e-05, "loss": 0.0597, "step": 3600 }, { "grad_norm": 0.6145071387290955, "learning_rate": 7.218e-05, "loss": 0.0559, "step": 3610 }, { "grad_norm": 0.5577249526977539, "learning_rate": 7.238000000000001e-05, "loss": 0.0548, "step": 3620 }, { "grad_norm": 0.5186826586723328, "learning_rate": 7.258e-05, "loss": 0.0458, "step": 3630 }, { "grad_norm": 0.5082507729530334, "learning_rate": 7.278e-05, "loss": 0.0494, "step": 3640 }, { "grad_norm": 0.5848327875137329, "learning_rate": 7.298000000000001e-05, "loss": 0.0592, "step": 3650 }, { "grad_norm": 0.6643404960632324, "learning_rate": 7.318e-05, "loss": 0.0556, "step": 3660 }, { "grad_norm": 0.6264016032218933, "learning_rate": 7.338e-05, "loss": 0.0596, "step": 3670 }, { "grad_norm": 0.5300384163856506, "learning_rate": 7.358000000000001e-05, "loss": 0.0577, "step": 3680 }, { "grad_norm": 0.545089066028595, "learning_rate": 7.378e-05, "loss": 0.0482, "step": 3690 }, { "grad_norm": 0.576915979385376, "learning_rate": 7.398e-05, "loss": 0.0536, "step": 3700 }, { "grad_norm": 0.45437905192375183, "learning_rate": 7.418000000000001e-05, "loss": 0.0497, "step": 3710 }, { "grad_norm": 0.5010011792182922, "learning_rate": 7.438e-05, "loss": 0.0524, "step": 3720 }, { "grad_norm": 0.6177607178688049, "learning_rate": 7.458000000000001e-05, "loss": 0.047, "step": 3730 }, { "grad_norm": 0.697020947933197, "learning_rate": 7.478e-05, "loss": 0.051, "step": 3740 }, { "grad_norm": 0.6057714223861694, "learning_rate": 7.498e-05, "loss": 0.0607, "step": 3750 }, { "grad_norm": 0.4501596689224243, "learning_rate": 7.518000000000001e-05, "loss": 0.0445, "step": 3760 }, { "grad_norm": 0.6712278127670288, "learning_rate": 7.538e-05, "loss": 0.0592, "step": 3770 }, { "grad_norm": 0.5892327427864075, "learning_rate": 7.558e-05, "loss": 0.0593, "step": 3780 }, { "grad_norm": 0.4908471703529358, "learning_rate": 7.578000000000001e-05, "loss": 0.0514, "step": 3790 }, { "grad_norm": 0.5708381533622742, "learning_rate": 7.598e-05, "loss": 0.0875, "step": 3800 }, { "grad_norm": 0.49626606702804565, "learning_rate": 7.618e-05, "loss": 0.0502, "step": 3810 }, { "grad_norm": 0.6660820841789246, "learning_rate": 7.638000000000001e-05, "loss": 0.0608, "step": 3820 }, { "grad_norm": 0.5230540633201599, "learning_rate": 7.658e-05, "loss": 0.0495, "step": 3830 }, { "grad_norm": 0.39636409282684326, "learning_rate": 7.678000000000001e-05, "loss": 0.045, "step": 3840 }, { "grad_norm": 0.5371338725090027, "learning_rate": 7.698000000000001e-05, "loss": 0.0639, "step": 3850 }, { "grad_norm": 0.556281328201294, "learning_rate": 7.718e-05, "loss": 0.0469, "step": 3860 }, { "grad_norm": 0.506924569606781, "learning_rate": 7.738000000000001e-05, "loss": 0.0482, "step": 3870 }, { "grad_norm": 0.6010696291923523, "learning_rate": 7.758000000000001e-05, "loss": 0.0433, "step": 3880 }, { "grad_norm": 0.5692047476768494, "learning_rate": 7.778e-05, "loss": 0.0554, "step": 3890 }, { "grad_norm": 0.497881680727005, "learning_rate": 7.798000000000001e-05, "loss": 0.056, "step": 3900 }, { "grad_norm": 0.47154471278190613, "learning_rate": 7.818000000000001e-05, "loss": 0.0501, "step": 3910 }, { "grad_norm": 0.8493228554725647, "learning_rate": 7.838e-05, "loss": 0.0603, "step": 3920 }, { "grad_norm": 0.5262234807014465, "learning_rate": 7.858000000000001e-05, "loss": 0.0503, "step": 3930 }, { "grad_norm": 0.5761149525642395, "learning_rate": 7.878e-05, "loss": 0.0652, "step": 3940 }, { "grad_norm": 0.6267067193984985, "learning_rate": 7.897999999999999e-05, "loss": 0.0553, "step": 3950 }, { "grad_norm": 0.6152517199516296, "learning_rate": 7.918e-05, "loss": 0.0663, "step": 3960 }, { "grad_norm": 0.6686506271362305, "learning_rate": 7.938e-05, "loss": 0.0497, "step": 3970 }, { "grad_norm": 0.5765831470489502, "learning_rate": 7.958e-05, "loss": 0.0599, "step": 3980 }, { "grad_norm": 0.7720066905021667, "learning_rate": 7.978e-05, "loss": 0.0613, "step": 3990 }, { "grad_norm": 0.7942103147506714, "learning_rate": 7.998e-05, "loss": 0.0595, "step": 4000 }, { "grad_norm": 0.6579099297523499, "learning_rate": 8.018e-05, "loss": 0.0537, "step": 4010 }, { "grad_norm": 0.6541140675544739, "learning_rate": 8.038e-05, "loss": 0.0586, "step": 4020 }, { "grad_norm": 0.6016380190849304, "learning_rate": 8.058e-05, "loss": 0.0633, "step": 4030 }, { "grad_norm": 0.5494979023933411, "learning_rate": 8.078e-05, "loss": 0.0441, "step": 4040 }, { "grad_norm": 0.5002334117889404, "learning_rate": 8.098e-05, "loss": 0.0542, "step": 4050 }, { "grad_norm": 0.5327333211898804, "learning_rate": 8.118e-05, "loss": 0.0517, "step": 4060 }, { "grad_norm": 0.49719035625457764, "learning_rate": 8.138e-05, "loss": 0.0455, "step": 4070 }, { "grad_norm": 0.4989062547683716, "learning_rate": 8.158e-05, "loss": 0.0629, "step": 4080 }, { "grad_norm": 0.5061677694320679, "learning_rate": 8.178e-05, "loss": 0.0543, "step": 4090 }, { "grad_norm": 0.5559020638465881, "learning_rate": 8.198e-05, "loss": 0.0467, "step": 4100 }, { "grad_norm": 0.48339220881462097, "learning_rate": 8.218e-05, "loss": 0.0557, "step": 4110 }, { "grad_norm": 0.7481857538223267, "learning_rate": 8.238000000000001e-05, "loss": 0.0632, "step": 4120 }, { "grad_norm": 0.6493443250656128, "learning_rate": 8.258e-05, "loss": 0.056, "step": 4130 }, { "grad_norm": 0.5938178896903992, "learning_rate": 8.278e-05, "loss": 0.0554, "step": 4140 }, { "grad_norm": 0.5216941833496094, "learning_rate": 8.298000000000001e-05, "loss": 0.0667, "step": 4150 }, { "grad_norm": 0.533244252204895, "learning_rate": 8.318e-05, "loss": 0.0534, "step": 4160 }, { "grad_norm": 0.5469269752502441, "learning_rate": 8.338e-05, "loss": 0.0513, "step": 4170 }, { "grad_norm": 0.48941850662231445, "learning_rate": 8.358e-05, "loss": 0.0481, "step": 4180 }, { "grad_norm": 0.8649469017982483, "learning_rate": 8.378e-05, "loss": 0.0523, "step": 4190 }, { "grad_norm": 0.5576428771018982, "learning_rate": 8.398e-05, "loss": 0.0514, "step": 4200 }, { "grad_norm": 0.569622278213501, "learning_rate": 8.418e-05, "loss": 0.0495, "step": 4210 }, { "grad_norm": 0.4981507956981659, "learning_rate": 8.438e-05, "loss": 0.0639, "step": 4220 }, { "grad_norm": 0.9355514049530029, "learning_rate": 8.458e-05, "loss": 0.0753, "step": 4230 }, { "grad_norm": 0.5910239815711975, "learning_rate": 8.478e-05, "loss": 0.0647, "step": 4240 }, { "grad_norm": 0.4751816391944885, "learning_rate": 8.498e-05, "loss": 0.0642, "step": 4250 }, { "grad_norm": 0.4070935845375061, "learning_rate": 8.518000000000001e-05, "loss": 0.0597, "step": 4260 }, { "grad_norm": 0.5197845101356506, "learning_rate": 8.538e-05, "loss": 0.0553, "step": 4270 }, { "grad_norm": 0.49361908435821533, "learning_rate": 8.558e-05, "loss": 0.0481, "step": 4280 }, { "grad_norm": 0.5886411666870117, "learning_rate": 8.578000000000001e-05, "loss": 0.0527, "step": 4290 }, { "grad_norm": 0.4115513861179352, "learning_rate": 8.598e-05, "loss": 0.0447, "step": 4300 }, { "grad_norm": 0.4285655617713928, "learning_rate": 8.618e-05, "loss": 0.0534, "step": 4310 }, { "grad_norm": 0.6057738065719604, "learning_rate": 8.638000000000001e-05, "loss": 0.0852, "step": 4320 }, { "grad_norm": 0.5767436027526855, "learning_rate": 8.658e-05, "loss": 0.0525, "step": 4330 }, { "grad_norm": 0.42692211270332336, "learning_rate": 8.678e-05, "loss": 0.0455, "step": 4340 }, { "grad_norm": 0.4125378131866455, "learning_rate": 8.698000000000001e-05, "loss": 0.0427, "step": 4350 }, { "grad_norm": 0.446586549282074, "learning_rate": 8.718e-05, "loss": 0.0472, "step": 4360 }, { "grad_norm": 0.46220383048057556, "learning_rate": 8.738000000000001e-05, "loss": 0.0539, "step": 4370 }, { "grad_norm": 0.5928171873092651, "learning_rate": 8.758000000000001e-05, "loss": 0.0519, "step": 4380 }, { "grad_norm": 0.40769150853157043, "learning_rate": 8.778e-05, "loss": 0.0435, "step": 4390 }, { "grad_norm": 0.5142196416854858, "learning_rate": 8.798000000000001e-05, "loss": 0.0619, "step": 4400 }, { "grad_norm": 0.5184516310691833, "learning_rate": 8.818000000000001e-05, "loss": 0.0537, "step": 4410 }, { "grad_norm": 0.49128860235214233, "learning_rate": 8.838e-05, "loss": 0.0681, "step": 4420 }, { "grad_norm": 0.5231001973152161, "learning_rate": 8.858000000000001e-05, "loss": 0.0582, "step": 4430 }, { "grad_norm": 0.5427704453468323, "learning_rate": 8.878000000000001e-05, "loss": 0.0573, "step": 4440 }, { "grad_norm": 0.41063234210014343, "learning_rate": 8.898e-05, "loss": 0.0691, "step": 4450 }, { "grad_norm": 0.42444753646850586, "learning_rate": 8.918000000000001e-05, "loss": 0.0547, "step": 4460 }, { "grad_norm": 0.41333961486816406, "learning_rate": 8.938e-05, "loss": 0.0502, "step": 4470 }, { "grad_norm": 0.5266408324241638, "learning_rate": 8.958e-05, "loss": 0.0608, "step": 4480 }, { "grad_norm": 0.5012792944908142, "learning_rate": 8.978000000000001e-05, "loss": 0.0604, "step": 4490 }, { "grad_norm": 0.4658188819885254, "learning_rate": 8.998e-05, "loss": 0.0554, "step": 4500 }, { "grad_norm": 0.3979945182800293, "learning_rate": 9.018000000000001e-05, "loss": 0.0763, "step": 4510 }, { "grad_norm": 0.4621061384677887, "learning_rate": 9.038000000000001e-05, "loss": 0.0594, "step": 4520 }, { "grad_norm": 0.5162387490272522, "learning_rate": 9.058e-05, "loss": 0.0549, "step": 4530 }, { "grad_norm": 0.6951661109924316, "learning_rate": 9.078000000000001e-05, "loss": 0.0706, "step": 4540 }, { "grad_norm": 0.599094808101654, "learning_rate": 9.098000000000001e-05, "loss": 0.0678, "step": 4550 }, { "grad_norm": 0.5372647643089294, "learning_rate": 9.118e-05, "loss": 0.0513, "step": 4560 }, { "grad_norm": 0.5182415843009949, "learning_rate": 9.138e-05, "loss": 0.0624, "step": 4570 }, { "grad_norm": 0.7905921339988708, "learning_rate": 9.158e-05, "loss": 0.0564, "step": 4580 }, { "grad_norm": 0.5613818764686584, "learning_rate": 9.178e-05, "loss": 0.0543, "step": 4590 }, { "grad_norm": 0.5201306343078613, "learning_rate": 9.198e-05, "loss": 0.058, "step": 4600 }, { "grad_norm": 0.4433695673942566, "learning_rate": 9.218e-05, "loss": 0.0551, "step": 4610 }, { "grad_norm": 0.5856695175170898, "learning_rate": 9.238e-05, "loss": 0.0482, "step": 4620 }, { "grad_norm": 0.6758952140808105, "learning_rate": 9.258e-05, "loss": 0.0655, "step": 4630 }, { "grad_norm": 0.46996408700942993, "learning_rate": 9.278e-05, "loss": 0.0619, "step": 4640 }, { "grad_norm": 0.7021536231040955, "learning_rate": 9.298e-05, "loss": 0.0503, "step": 4650 }, { "grad_norm": 0.44561317563056946, "learning_rate": 9.318e-05, "loss": 0.0564, "step": 4660 }, { "grad_norm": 0.7978839874267578, "learning_rate": 9.338e-05, "loss": 0.0549, "step": 4670 }, { "grad_norm": 0.4006088078022003, "learning_rate": 9.358e-05, "loss": 0.0571, "step": 4680 }, { "grad_norm": 0.5049392580986023, "learning_rate": 9.378e-05, "loss": 0.0639, "step": 4690 }, { "grad_norm": 0.46481165289878845, "learning_rate": 9.398e-05, "loss": 0.0538, "step": 4700 }, { "grad_norm": 0.480698823928833, "learning_rate": 9.418e-05, "loss": 0.052, "step": 4710 }, { "grad_norm": 0.521399974822998, "learning_rate": 9.438e-05, "loss": 0.0485, "step": 4720 }, { "grad_norm": 0.3786125183105469, "learning_rate": 9.458e-05, "loss": 0.0606, "step": 4730 }, { "grad_norm": 0.3524838984012604, "learning_rate": 9.478e-05, "loss": 0.0548, "step": 4740 }, { "grad_norm": 0.5448586940765381, "learning_rate": 9.498e-05, "loss": 0.0452, "step": 4750 }, { "grad_norm": 0.522286057472229, "learning_rate": 9.518000000000001e-05, "loss": 0.0551, "step": 4760 }, { "grad_norm": 0.5830252766609192, "learning_rate": 9.538e-05, "loss": 0.06, "step": 4770 }, { "grad_norm": 0.6137187480926514, "learning_rate": 9.558e-05, "loss": 0.071, "step": 4780 }, { "grad_norm": 0.6045110821723938, "learning_rate": 9.578000000000001e-05, "loss": 0.0498, "step": 4790 }, { "grad_norm": 0.4724533259868622, "learning_rate": 9.598e-05, "loss": 0.0559, "step": 4800 }, { "grad_norm": 0.5263993740081787, "learning_rate": 9.618e-05, "loss": 0.0571, "step": 4810 }, { "grad_norm": 0.5049068927764893, "learning_rate": 9.638000000000001e-05, "loss": 0.0592, "step": 4820 }, { "grad_norm": 0.5686495900154114, "learning_rate": 9.658e-05, "loss": 0.0548, "step": 4830 }, { "grad_norm": 0.47514283657073975, "learning_rate": 9.678e-05, "loss": 0.0594, "step": 4840 }, { "grad_norm": 0.4446946978569031, "learning_rate": 9.698000000000001e-05, "loss": 0.0528, "step": 4850 }, { "grad_norm": 0.3966381847858429, "learning_rate": 9.718e-05, "loss": 0.0449, "step": 4860 }, { "grad_norm": 0.4581451714038849, "learning_rate": 9.738e-05, "loss": 0.0507, "step": 4870 }, { "grad_norm": 0.498658150434494, "learning_rate": 9.758000000000001e-05, "loss": 0.047, "step": 4880 }, { "grad_norm": 0.5827343463897705, "learning_rate": 9.778e-05, "loss": 0.0704, "step": 4890 }, { "grad_norm": 0.44770362973213196, "learning_rate": 9.798000000000001e-05, "loss": 0.0465, "step": 4900 }, { "grad_norm": 0.4378403425216675, "learning_rate": 9.818000000000001e-05, "loss": 0.0545, "step": 4910 }, { "grad_norm": 0.5712823271751404, "learning_rate": 9.838e-05, "loss": 0.0583, "step": 4920 }, { "grad_norm": 0.3837016820907593, "learning_rate": 9.858000000000001e-05, "loss": 0.0523, "step": 4930 }, { "grad_norm": 0.4336758852005005, "learning_rate": 9.878e-05, "loss": 0.0469, "step": 4940 }, { "grad_norm": 0.5145536065101624, "learning_rate": 9.898e-05, "loss": 0.0539, "step": 4950 }, { "grad_norm": 0.4326086938381195, "learning_rate": 9.918000000000001e-05, "loss": 0.0484, "step": 4960 }, { "grad_norm": 0.35272854566574097, "learning_rate": 9.938e-05, "loss": 0.0629, "step": 4970 }, { "grad_norm": 0.47148972749710083, "learning_rate": 9.958e-05, "loss": 0.0546, "step": 4980 }, { "grad_norm": 0.41919127106666565, "learning_rate": 9.978000000000001e-05, "loss": 0.0512, "step": 4990 }, { "grad_norm": 0.4382629990577698, "learning_rate": 9.998e-05, "loss": 0.0592, "step": 5000 }, { "grad_norm": 0.5707319974899292, "learning_rate": 9.999999778549045e-05, "loss": 0.052, "step": 5010 }, { "grad_norm": 0.4738978147506714, "learning_rate": 9.999999013039593e-05, "loss": 0.0505, "step": 5020 }, { "grad_norm": 0.3647136688232422, "learning_rate": 9.999997700737766e-05, "loss": 0.0431, "step": 5030 }, { "grad_norm": 0.3894270062446594, "learning_rate": 9.999995841643709e-05, "loss": 0.0464, "step": 5040 }, { "grad_norm": 0.5135908126831055, "learning_rate": 9.999993435757623e-05, "loss": 0.0556, "step": 5050 }, { "grad_norm": 0.428427517414093, "learning_rate": 9.999990483079773e-05, "loss": 0.0384, "step": 5060 }, { "grad_norm": 0.518785297870636, "learning_rate": 9.999986983610481e-05, "loss": 0.0651, "step": 5070 }, { "grad_norm": 0.37124010920524597, "learning_rate": 9.99998293735013e-05, "loss": 0.0592, "step": 5080 }, { "grad_norm": 0.4414944648742676, "learning_rate": 9.999978344299161e-05, "loss": 0.0447, "step": 5090 }, { "grad_norm": 0.41707149147987366, "learning_rate": 9.99997320445808e-05, "loss": 0.046, "step": 5100 }, { "grad_norm": 0.4625191390514374, "learning_rate": 9.999967517827444e-05, "loss": 0.0531, "step": 5110 }, { "grad_norm": 0.44894659519195557, "learning_rate": 9.999961284407879e-05, "loss": 0.0685, "step": 5120 }, { "grad_norm": 0.37109795212745667, "learning_rate": 9.999954504200067e-05, "loss": 0.0628, "step": 5130 }, { "grad_norm": 0.4711306095123291, "learning_rate": 9.999947177204744e-05, "loss": 0.0511, "step": 5140 }, { "grad_norm": 0.3386825919151306, "learning_rate": 9.999939303422718e-05, "loss": 0.0443, "step": 5150 }, { "grad_norm": 0.48417526483535767, "learning_rate": 9.999930882854847e-05, "loss": 0.0613, "step": 5160 }, { "grad_norm": 0.5174601674079895, "learning_rate": 9.999921915502051e-05, "loss": 0.0529, "step": 5170 }, { "grad_norm": 0.4613370895385742, "learning_rate": 9.99991240136531e-05, "loss": 0.0457, "step": 5180 }, { "grad_norm": 0.46207395195961, "learning_rate": 9.999902340445668e-05, "loss": 0.0722, "step": 5190 }, { "grad_norm": 0.3902304172515869, "learning_rate": 9.999891732744224e-05, "loss": 0.0415, "step": 5200 }, { "grad_norm": 0.5632699131965637, "learning_rate": 9.999880578262135e-05, "loss": 0.051, "step": 5210 }, { "grad_norm": 0.4487893581390381, "learning_rate": 9.999868877000624e-05, "loss": 0.0458, "step": 5220 }, { "grad_norm": 0.429827481508255, "learning_rate": 9.99985662896097e-05, "loss": 0.0446, "step": 5230 }, { "grad_norm": 0.4891599118709564, "learning_rate": 9.999843834144513e-05, "loss": 0.0467, "step": 5240 }, { "grad_norm": 0.4043658375740051, "learning_rate": 9.99983049255265e-05, "loss": 0.0605, "step": 5250 }, { "grad_norm": 0.5461553931236267, "learning_rate": 9.999816604186843e-05, "loss": 0.0467, "step": 5260 }, { "grad_norm": 0.4914673864841461, "learning_rate": 9.999802169048609e-05, "loss": 0.0495, "step": 5270 }, { "grad_norm": 0.4656551480293274, "learning_rate": 9.999787187139527e-05, "loss": 0.0563, "step": 5280 }, { "grad_norm": 0.5724367499351501, "learning_rate": 9.999771658461234e-05, "loss": 0.0572, "step": 5290 }, { "grad_norm": 0.40721842646598816, "learning_rate": 9.999755583015431e-05, "loss": 0.0459, "step": 5300 }, { "grad_norm": 0.35076120495796204, "learning_rate": 9.999738960803874e-05, "loss": 0.0446, "step": 5310 }, { "grad_norm": 0.5141162872314453, "learning_rate": 9.99972179182838e-05, "loss": 0.0512, "step": 5320 }, { "grad_norm": 0.5340847969055176, "learning_rate": 9.99970407609083e-05, "loss": 0.0634, "step": 5330 }, { "grad_norm": 0.4126587212085724, "learning_rate": 9.999685813593159e-05, "loss": 0.0502, "step": 5340 }, { "grad_norm": 0.3145295977592468, "learning_rate": 9.999667004337362e-05, "loss": 0.0419, "step": 5350 }, { "grad_norm": 0.40358588099479675, "learning_rate": 9.9996476483255e-05, "loss": 0.0466, "step": 5360 }, { "grad_norm": 0.3785283863544464, "learning_rate": 9.999627745559688e-05, "loss": 0.0401, "step": 5370 }, { "grad_norm": 0.47036078572273254, "learning_rate": 9.999607296042101e-05, "loss": 0.0428, "step": 5380 }, { "grad_norm": 0.31794309616088867, "learning_rate": 9.99958629977498e-05, "loss": 0.0417, "step": 5390 }, { "grad_norm": 0.364239364862442, "learning_rate": 9.999564756760615e-05, "loss": 0.0513, "step": 5400 }, { "grad_norm": 0.3183148205280304, "learning_rate": 9.999542667001366e-05, "loss": 0.0406, "step": 5410 }, { "grad_norm": 0.407442569732666, "learning_rate": 9.999520030499647e-05, "loss": 0.0441, "step": 5420 }, { "grad_norm": 0.45647570490837097, "learning_rate": 9.999496847257936e-05, "loss": 0.0411, "step": 5430 }, { "grad_norm": 0.4254451394081116, "learning_rate": 9.999473117278764e-05, "loss": 0.057, "step": 5440 }, { "grad_norm": 0.48760631680488586, "learning_rate": 9.999448840564731e-05, "loss": 0.0367, "step": 5450 }, { "grad_norm": 0.47136878967285156, "learning_rate": 9.999424017118488e-05, "loss": 0.041, "step": 5460 }, { "grad_norm": 0.49072396755218506, "learning_rate": 9.999398646942751e-05, "loss": 0.0403, "step": 5470 }, { "grad_norm": 0.4267460107803345, "learning_rate": 9.999372730040296e-05, "loss": 0.0451, "step": 5480 }, { "grad_norm": 0.3710397183895111, "learning_rate": 9.999346266413953e-05, "loss": 0.0422, "step": 5490 }, { "grad_norm": 0.49467381834983826, "learning_rate": 9.99931925606662e-05, "loss": 0.0454, "step": 5500 }, { "grad_norm": 0.40369993448257446, "learning_rate": 9.99929169900125e-05, "loss": 0.0472, "step": 5510 }, { "grad_norm": 0.33567243814468384, "learning_rate": 9.999263595220855e-05, "loss": 0.0396, "step": 5520 }, { "grad_norm": 0.43521517515182495, "learning_rate": 9.99923494472851e-05, "loss": 0.0455, "step": 5530 }, { "grad_norm": 0.4737412929534912, "learning_rate": 9.999205747527348e-05, "loss": 0.0378, "step": 5540 }, { "grad_norm": 0.2757497727870941, "learning_rate": 9.999176003620561e-05, "loss": 0.0418, "step": 5550 }, { "grad_norm": 0.45051488280296326, "learning_rate": 9.999145713011405e-05, "loss": 0.0424, "step": 5560 }, { "grad_norm": 0.42437633872032166, "learning_rate": 9.999114875703186e-05, "loss": 0.0424, "step": 5570 }, { "grad_norm": 0.4736972153186798, "learning_rate": 9.999083491699281e-05, "loss": 0.0565, "step": 5580 }, { "grad_norm": 0.5330377817153931, "learning_rate": 9.999051561003123e-05, "loss": 0.0428, "step": 5590 }, { "grad_norm": 0.42715874314308167, "learning_rate": 9.999019083618202e-05, "loss": 0.0497, "step": 5600 }, { "grad_norm": 0.4642340838909149, "learning_rate": 9.99898605954807e-05, "loss": 0.0421, "step": 5610 }, { "grad_norm": 0.4010845422744751, "learning_rate": 9.998952488796338e-05, "loss": 0.0446, "step": 5620 }, { "grad_norm": 0.35108205676078796, "learning_rate": 9.998918371366676e-05, "loss": 0.0411, "step": 5630 }, { "grad_norm": 0.4202430546283722, "learning_rate": 9.99888370726282e-05, "loss": 0.0452, "step": 5640 }, { "grad_norm": 0.4516030550003052, "learning_rate": 9.998848496488556e-05, "loss": 0.0579, "step": 5650 }, { "grad_norm": 0.4282025396823883, "learning_rate": 9.998812739047736e-05, "loss": 0.0451, "step": 5660 }, { "grad_norm": 0.5174033045768738, "learning_rate": 9.99877643494427e-05, "loss": 0.0456, "step": 5670 }, { "grad_norm": 0.40393194556236267, "learning_rate": 9.998739584182128e-05, "loss": 0.0559, "step": 5680 }, { "grad_norm": 0.47538459300994873, "learning_rate": 9.998702186765342e-05, "loss": 0.049, "step": 5690 }, { "grad_norm": 0.35080012679100037, "learning_rate": 9.998664242698e-05, "loss": 0.0459, "step": 5700 }, { "grad_norm": 0.4269775152206421, "learning_rate": 9.998625751984251e-05, "loss": 0.0546, "step": 5710 }, { "grad_norm": 0.49344927072525024, "learning_rate": 9.998586714628307e-05, "loss": 0.0434, "step": 5720 }, { "grad_norm": 0.45573025941848755, "learning_rate": 9.998547130634432e-05, "loss": 0.044, "step": 5730 }, { "grad_norm": 0.5111093521118164, "learning_rate": 9.99850700000696e-05, "loss": 0.0538, "step": 5740 }, { "grad_norm": 0.4222013056278229, "learning_rate": 9.998466322750278e-05, "loss": 0.0415, "step": 5750 }, { "grad_norm": 0.3406006693840027, "learning_rate": 9.998425098868834e-05, "loss": 0.0455, "step": 5760 }, { "grad_norm": 0.3829037547111511, "learning_rate": 9.998383328367136e-05, "loss": 0.0469, "step": 5770 }, { "grad_norm": 0.6126700639724731, "learning_rate": 9.99834101124975e-05, "loss": 0.0541, "step": 5780 }, { "grad_norm": 0.5062404870986938, "learning_rate": 9.998298147521309e-05, "loss": 0.045, "step": 5790 }, { "grad_norm": 0.4898018538951874, "learning_rate": 9.998254737186496e-05, "loss": 0.0581, "step": 5800 }, { "grad_norm": 0.2705562710762024, "learning_rate": 9.99821078025006e-05, "loss": 0.0499, "step": 5810 }, { "grad_norm": 0.3730393350124359, "learning_rate": 9.998166276716807e-05, "loss": 0.0467, "step": 5820 }, { "grad_norm": 0.39898380637168884, "learning_rate": 9.998121226591606e-05, "loss": 0.0479, "step": 5830 }, { "grad_norm": 0.3654755651950836, "learning_rate": 9.998075629879382e-05, "loss": 0.0443, "step": 5840 }, { "grad_norm": 0.3286758065223694, "learning_rate": 9.99802948658512e-05, "loss": 0.0384, "step": 5850 }, { "grad_norm": 0.4468033015727997, "learning_rate": 9.99798279671387e-05, "loss": 0.0523, "step": 5860 }, { "grad_norm": 0.4026978015899658, "learning_rate": 9.997935560270734e-05, "loss": 0.0482, "step": 5870 }, { "grad_norm": 0.37377166748046875, "learning_rate": 9.997887777260879e-05, "loss": 0.0574, "step": 5880 }, { "grad_norm": 0.36460286378860474, "learning_rate": 9.997839447689532e-05, "loss": 0.0509, "step": 5890 }, { "grad_norm": 0.45343393087387085, "learning_rate": 9.997790571561978e-05, "loss": 0.0686, "step": 5900 }, { "grad_norm": 0.3302176296710968, "learning_rate": 9.99774114888356e-05, "loss": 0.0577, "step": 5910 }, { "grad_norm": 0.4340757429599762, "learning_rate": 9.997691179659684e-05, "loss": 0.0459, "step": 5920 }, { "grad_norm": 0.412306547164917, "learning_rate": 9.997640663895815e-05, "loss": 0.0447, "step": 5930 }, { "grad_norm": 0.32650408148765564, "learning_rate": 9.997589601597477e-05, "loss": 0.0509, "step": 5940 }, { "grad_norm": 0.3649135231971741, "learning_rate": 9.997537992770252e-05, "loss": 0.0451, "step": 5950 }, { "grad_norm": 0.3830304443836212, "learning_rate": 9.997485837419788e-05, "loss": 0.0411, "step": 5960 }, { "grad_norm": 0.5429280996322632, "learning_rate": 9.997433135551786e-05, "loss": 0.0491, "step": 5970 }, { "grad_norm": 0.4142684042453766, "learning_rate": 9.997379887172009e-05, "loss": 0.0476, "step": 5980 }, { "grad_norm": 0.4203357696533203, "learning_rate": 9.997326092286281e-05, "loss": 0.0586, "step": 5990 }, { "grad_norm": 0.3931107223033905, "learning_rate": 9.997271750900486e-05, "loss": 0.0675, "step": 6000 }, { "grad_norm": 0.4046901762485504, "learning_rate": 9.997216863020565e-05, "loss": 0.0523, "step": 6010 }, { "grad_norm": 0.3847743570804596, "learning_rate": 9.99716142865252e-05, "loss": 0.0512, "step": 6020 }, { "grad_norm": 0.40956321358680725, "learning_rate": 9.997105447802415e-05, "loss": 0.0515, "step": 6030 }, { "grad_norm": 0.43596982955932617, "learning_rate": 9.997048920476373e-05, "loss": 0.0472, "step": 6040 }, { "grad_norm": 0.47383779287338257, "learning_rate": 9.996991846680572e-05, "loss": 0.0558, "step": 6050 }, { "grad_norm": 0.40006014704704285, "learning_rate": 9.996934226421257e-05, "loss": 0.0433, "step": 6060 }, { "grad_norm": 0.4636707007884979, "learning_rate": 9.996876059704726e-05, "loss": 0.0463, "step": 6070 }, { "grad_norm": 0.4405924379825592, "learning_rate": 9.996817346537343e-05, "loss": 0.0516, "step": 6080 }, { "grad_norm": 0.4029555022716522, "learning_rate": 9.996758086925526e-05, "loss": 0.0496, "step": 6090 }, { "grad_norm": 0.3879355490207672, "learning_rate": 9.996698280875759e-05, "loss": 0.0554, "step": 6100 }, { "grad_norm": 0.3140811026096344, "learning_rate": 9.99663792839458e-05, "loss": 0.0391, "step": 6110 }, { "grad_norm": 0.3553231656551361, "learning_rate": 9.99657702948859e-05, "loss": 0.0471, "step": 6120 }, { "grad_norm": 0.3365764319896698, "learning_rate": 9.996515584164448e-05, "loss": 0.0369, "step": 6130 }, { "grad_norm": 0.3679174780845642, "learning_rate": 9.996453592428873e-05, "loss": 0.0446, "step": 6140 }, { "grad_norm": 0.42453479766845703, "learning_rate": 9.996391054288646e-05, "loss": 0.0677, "step": 6150 }, { "grad_norm": 0.397403746843338, "learning_rate": 9.996327969750605e-05, "loss": 0.0632, "step": 6160 }, { "grad_norm": 0.3846220076084137, "learning_rate": 9.996264338821649e-05, "loss": 0.0572, "step": 6170 }, { "grad_norm": 0.4248979687690735, "learning_rate": 9.996200161508735e-05, "loss": 0.0444, "step": 6180 }, { "grad_norm": 0.42226332426071167, "learning_rate": 9.996135437818885e-05, "loss": 0.0616, "step": 6190 }, { "grad_norm": 0.3090721368789673, "learning_rate": 9.996070167759175e-05, "loss": 0.0495, "step": 6200 }, { "grad_norm": 0.43692177534103394, "learning_rate": 9.996004351336743e-05, "loss": 0.051, "step": 6210 }, { "grad_norm": 0.4358850419521332, "learning_rate": 9.995937988558785e-05, "loss": 0.0481, "step": 6220 }, { "grad_norm": 0.46531641483306885, "learning_rate": 9.995871079432561e-05, "loss": 0.0487, "step": 6230 }, { "grad_norm": 0.33212411403656006, "learning_rate": 9.995803623965389e-05, "loss": 0.0516, "step": 6240 }, { "grad_norm": 0.41344505548477173, "learning_rate": 9.995735622164641e-05, "loss": 0.0543, "step": 6250 }, { "grad_norm": 0.4109017252922058, "learning_rate": 9.995667074037758e-05, "loss": 0.0409, "step": 6260 }, { "grad_norm": 0.41950520873069763, "learning_rate": 9.995597979592232e-05, "loss": 0.0443, "step": 6270 }, { "grad_norm": 0.43028825521469116, "learning_rate": 9.995528338835625e-05, "loss": 0.0465, "step": 6280 }, { "grad_norm": 0.4430437982082367, "learning_rate": 9.995458151775547e-05, "loss": 0.0454, "step": 6290 }, { "grad_norm": 0.3235726058483124, "learning_rate": 9.995387418419677e-05, "loss": 0.0499, "step": 6300 }, { "grad_norm": 0.35633301734924316, "learning_rate": 9.99531613877575e-05, "loss": 0.0416, "step": 6310 }, { "grad_norm": 0.30018842220306396, "learning_rate": 9.995244312851559e-05, "loss": 0.0434, "step": 6320 }, { "grad_norm": 0.4742390811443329, "learning_rate": 9.995171940654961e-05, "loss": 0.0452, "step": 6330 }, { "grad_norm": 0.4612194895744324, "learning_rate": 9.995099022193871e-05, "loss": 0.0577, "step": 6340 }, { "grad_norm": 0.41178351640701294, "learning_rate": 9.995025557476261e-05, "loss": 0.0442, "step": 6350 }, { "grad_norm": 0.3527931571006775, "learning_rate": 9.994951546510165e-05, "loss": 0.0477, "step": 6360 }, { "grad_norm": 0.41524243354797363, "learning_rate": 9.994876989303679e-05, "loss": 0.044, "step": 6370 }, { "grad_norm": 0.3906223177909851, "learning_rate": 9.994801885864955e-05, "loss": 0.0425, "step": 6380 }, { "grad_norm": 0.4067162871360779, "learning_rate": 9.994726236202205e-05, "loss": 0.0508, "step": 6390 }, { "grad_norm": 0.4468441307544708, "learning_rate": 9.994650040323704e-05, "loss": 0.0583, "step": 6400 }, { "grad_norm": 0.4065152406692505, "learning_rate": 9.994573298237784e-05, "loss": 0.0409, "step": 6410 }, { "grad_norm": 0.3857818841934204, "learning_rate": 9.994496009952837e-05, "loss": 0.0503, "step": 6420 }, { "grad_norm": 0.47659850120544434, "learning_rate": 9.994418175477316e-05, "loss": 0.0434, "step": 6430 }, { "grad_norm": 0.3231425881385803, "learning_rate": 9.994339794819733e-05, "loss": 0.043, "step": 6440 }, { "grad_norm": 0.3614327311515808, "learning_rate": 9.994260867988658e-05, "loss": 0.0403, "step": 6450 }, { "grad_norm": 0.3596927523612976, "learning_rate": 9.994181394992723e-05, "loss": 0.0393, "step": 6460 }, { "grad_norm": 0.39424628019332886, "learning_rate": 9.994101375840618e-05, "loss": 0.0461, "step": 6470 }, { "grad_norm": 0.31146204471588135, "learning_rate": 9.994020810541098e-05, "loss": 0.0401, "step": 6480 }, { "grad_norm": 0.36957088112831116, "learning_rate": 9.99393969910297e-05, "loss": 0.0521, "step": 6490 }, { "grad_norm": 0.3451228737831116, "learning_rate": 9.993858041535104e-05, "loss": 0.0379, "step": 6500 }, { "grad_norm": 0.3140046298503876, "learning_rate": 9.99377583784643e-05, "loss": 0.038, "step": 6510 }, { "grad_norm": 0.3760920464992523, "learning_rate": 9.993693088045939e-05, "loss": 0.0413, "step": 6520 }, { "grad_norm": 0.3643108606338501, "learning_rate": 9.99360979214268e-05, "loss": 0.0418, "step": 6530 }, { "grad_norm": 0.36722424626350403, "learning_rate": 9.99352595014576e-05, "loss": 0.0508, "step": 6540 }, { "grad_norm": 0.4681377112865448, "learning_rate": 9.993441562064354e-05, "loss": 0.0397, "step": 6550 }, { "grad_norm": 0.4535677134990692, "learning_rate": 9.993356627907685e-05, "loss": 0.0483, "step": 6560 }, { "grad_norm": 0.34724894165992737, "learning_rate": 9.99327114768504e-05, "loss": 0.0549, "step": 6570 }, { "grad_norm": 0.3628024756908417, "learning_rate": 9.99318512140577e-05, "loss": 0.0377, "step": 6580 }, { "grad_norm": 0.4116303026676178, "learning_rate": 9.993098549079284e-05, "loss": 0.0529, "step": 6590 }, { "grad_norm": 0.3991818130016327, "learning_rate": 9.993011430715047e-05, "loss": 0.0535, "step": 6600 }, { "grad_norm": 0.40088164806365967, "learning_rate": 9.992923766322586e-05, "loss": 0.0408, "step": 6610 }, { "grad_norm": 0.36208608746528625, "learning_rate": 9.99283555591149e-05, "loss": 0.0385, "step": 6620 }, { "grad_norm": 0.4251891076564789, "learning_rate": 9.992746799491404e-05, "loss": 0.042, "step": 6630 }, { "grad_norm": 0.33983686566352844, "learning_rate": 9.992657497072033e-05, "loss": 0.0347, "step": 6640 }, { "grad_norm": 0.3440859019756317, "learning_rate": 9.992567648663147e-05, "loss": 0.0524, "step": 6650 }, { "grad_norm": 0.3224794268608093, "learning_rate": 9.992477254274568e-05, "loss": 0.0392, "step": 6660 }, { "grad_norm": 0.24391406774520874, "learning_rate": 9.992386313916183e-05, "loss": 0.0361, "step": 6670 }, { "grad_norm": 0.33934545516967773, "learning_rate": 9.992294827597934e-05, "loss": 0.0481, "step": 6680 }, { "grad_norm": 0.3591856360435486, "learning_rate": 9.992202795329831e-05, "loss": 0.0426, "step": 6690 }, { "grad_norm": 0.35857540369033813, "learning_rate": 9.992110217121936e-05, "loss": 0.0413, "step": 6700 }, { "grad_norm": 0.327396422624588, "learning_rate": 9.992017092984372e-05, "loss": 0.0441, "step": 6710 }, { "grad_norm": 0.29540014266967773, "learning_rate": 9.991923422927326e-05, "loss": 0.0476, "step": 6720 }, { "grad_norm": 0.4798617660999298, "learning_rate": 9.991829206961037e-05, "loss": 0.0448, "step": 6730 }, { "grad_norm": 0.3545166850090027, "learning_rate": 9.991734445095813e-05, "loss": 0.0651, "step": 6740 }, { "grad_norm": 0.32279735803604126, "learning_rate": 9.991639137342015e-05, "loss": 0.0435, "step": 6750 }, { "grad_norm": 0.3431774973869324, "learning_rate": 9.991543283710064e-05, "loss": 0.0422, "step": 6760 }, { "grad_norm": 0.3626813292503357, "learning_rate": 9.991446884210445e-05, "loss": 0.0357, "step": 6770 }, { "grad_norm": 0.39463168382644653, "learning_rate": 9.9913499388537e-05, "loss": 0.046, "step": 6780 }, { "grad_norm": 0.34548214077949524, "learning_rate": 9.99125244765043e-05, "loss": 0.0368, "step": 6790 }, { "grad_norm": 0.39606213569641113, "learning_rate": 9.991154410611296e-05, "loss": 0.047, "step": 6800 }, { "grad_norm": 0.29636433720588684, "learning_rate": 9.99105582774702e-05, "loss": 0.0334, "step": 6810 }, { "grad_norm": 0.3742251396179199, "learning_rate": 9.990956699068384e-05, "loss": 0.05, "step": 6820 }, { "grad_norm": 0.4872821569442749, "learning_rate": 9.990857024586224e-05, "loss": 0.0428, "step": 6830 }, { "grad_norm": 0.42600512504577637, "learning_rate": 9.990756804311446e-05, "loss": 0.0457, "step": 6840 }, { "grad_norm": 0.35475587844848633, "learning_rate": 9.990656038255006e-05, "loss": 0.0408, "step": 6850 }, { "grad_norm": 0.37880849838256836, "learning_rate": 9.990554726427926e-05, "loss": 0.0382, "step": 6860 }, { "grad_norm": 0.47162240743637085, "learning_rate": 9.990452868841284e-05, "loss": 0.0567, "step": 6870 }, { "grad_norm": 0.3342052400112152, "learning_rate": 9.99035046550622e-05, "loss": 0.0487, "step": 6880 }, { "grad_norm": 0.34389492869377136, "learning_rate": 9.99024751643393e-05, "loss": 0.0448, "step": 6890 }, { "grad_norm": 0.43661507964134216, "learning_rate": 9.990144021635677e-05, "loss": 0.0508, "step": 6900 }, { "grad_norm": 0.27062180638313293, "learning_rate": 9.990039981122775e-05, "loss": 0.0395, "step": 6910 }, { "grad_norm": 0.3993728756904602, "learning_rate": 9.989935394906602e-05, "loss": 0.048, "step": 6920 }, { "grad_norm": 0.37614598870277405, "learning_rate": 9.989830262998598e-05, "loss": 0.0461, "step": 6930 }, { "grad_norm": 0.36954253911972046, "learning_rate": 9.989724585410259e-05, "loss": 0.036, "step": 6940 }, { "grad_norm": 0.31904786825180054, "learning_rate": 9.989618362153139e-05, "loss": 0.0364, "step": 6950 }, { "grad_norm": 0.42138323187828064, "learning_rate": 9.989511593238859e-05, "loss": 0.0402, "step": 6960 }, { "grad_norm": 0.33299365639686584, "learning_rate": 9.98940427867909e-05, "loss": 0.0425, "step": 6970 }, { "grad_norm": 0.34589338302612305, "learning_rate": 9.989296418485573e-05, "loss": 0.051, "step": 6980 }, { "grad_norm": 0.37745821475982666, "learning_rate": 9.989188012670101e-05, "loss": 0.0384, "step": 6990 }, { "grad_norm": 0.3882427215576172, "learning_rate": 9.989079061244528e-05, "loss": 0.045, "step": 7000 }, { "grad_norm": 0.40526413917541504, "learning_rate": 9.988969564220769e-05, "loss": 0.053, "step": 7010 }, { "grad_norm": 0.3827071487903595, "learning_rate": 9.988859521610801e-05, "loss": 0.0456, "step": 7020 }, { "grad_norm": 0.43373745679855347, "learning_rate": 9.988748933426656e-05, "loss": 0.0628, "step": 7030 }, { "grad_norm": 0.35784727334976196, "learning_rate": 9.988637799680428e-05, "loss": 0.0423, "step": 7040 }, { "grad_norm": 0.3249938488006592, "learning_rate": 9.98852612038427e-05, "loss": 0.0387, "step": 7050 }, { "grad_norm": 0.3439911901950836, "learning_rate": 9.988413895550397e-05, "loss": 0.038, "step": 7060 }, { "grad_norm": 0.4060392379760742, "learning_rate": 9.98830112519108e-05, "loss": 0.0419, "step": 7070 }, { "grad_norm": 0.3137739300727844, "learning_rate": 9.98818780931865e-05, "loss": 0.0596, "step": 7080 }, { "grad_norm": 0.4604194462299347, "learning_rate": 9.988073947945502e-05, "loss": 0.0458, "step": 7090 }, { "grad_norm": 0.3840552866458893, "learning_rate": 9.987959541084087e-05, "loss": 0.0376, "step": 7100 }, { "grad_norm": 0.38270217180252075, "learning_rate": 9.987844588746915e-05, "loss": 0.0506, "step": 7110 }, { "grad_norm": 0.3380346894264221, "learning_rate": 9.987729090946558e-05, "loss": 0.037, "step": 7120 }, { "grad_norm": 0.28018835186958313, "learning_rate": 9.987613047695647e-05, "loss": 0.0373, "step": 7130 }, { "grad_norm": 0.34210023283958435, "learning_rate": 9.987496459006871e-05, "loss": 0.0611, "step": 7140 }, { "grad_norm": 0.2788534164428711, "learning_rate": 9.987379324892982e-05, "loss": 0.0553, "step": 7150 }, { "grad_norm": 0.7022762298583984, "learning_rate": 9.987261645366788e-05, "loss": 0.0521, "step": 7160 }, { "grad_norm": 0.4227756857872009, "learning_rate": 9.987143420441158e-05, "loss": 0.0487, "step": 7170 }, { "grad_norm": 0.392683744430542, "learning_rate": 9.987024650129022e-05, "loss": 0.0425, "step": 7180 }, { "grad_norm": 0.40142911672592163, "learning_rate": 9.986905334443368e-05, "loss": 0.0508, "step": 7190 }, { "grad_norm": 0.5014833807945251, "learning_rate": 9.986785473397245e-05, "loss": 0.0434, "step": 7200 }, { "grad_norm": 0.33746835589408875, "learning_rate": 9.98666506700376e-05, "loss": 0.0441, "step": 7210 }, { "grad_norm": 0.40838608145713806, "learning_rate": 9.986544115276081e-05, "loss": 0.0452, "step": 7220 }, { "grad_norm": 0.4393383264541626, "learning_rate": 9.986422618227433e-05, "loss": 0.0388, "step": 7230 }, { "grad_norm": 0.41910070180892944, "learning_rate": 9.986300575871106e-05, "loss": 0.0447, "step": 7240 }, { "grad_norm": 0.3243785798549652, "learning_rate": 9.986177988220444e-05, "loss": 0.0467, "step": 7250 }, { "grad_norm": 0.31424111127853394, "learning_rate": 9.986054855288856e-05, "loss": 0.039, "step": 7260 }, { "grad_norm": 0.3919713497161865, "learning_rate": 9.985931177089802e-05, "loss": 0.0376, "step": 7270 }, { "grad_norm": 0.3739698827266693, "learning_rate": 9.985806953636814e-05, "loss": 0.0391, "step": 7280 }, { "grad_norm": 0.29953524470329285, "learning_rate": 9.985682184943471e-05, "loss": 0.0442, "step": 7290 }, { "grad_norm": 0.31697380542755127, "learning_rate": 9.98555687102342e-05, "loss": 0.0398, "step": 7300 }, { "grad_norm": 0.4475407600402832, "learning_rate": 9.985431011890367e-05, "loss": 0.0425, "step": 7310 }, { "grad_norm": 0.33307838439941406, "learning_rate": 9.985304607558075e-05, "loss": 0.0349, "step": 7320 }, { "grad_norm": 0.3193512558937073, "learning_rate": 9.985177658040364e-05, "loss": 0.0384, "step": 7330 }, { "grad_norm": 0.2991916835308075, "learning_rate": 9.985050163351119e-05, "loss": 0.0442, "step": 7340 }, { "grad_norm": 0.31560567021369934, "learning_rate": 9.984922123504286e-05, "loss": 0.0358, "step": 7350 }, { "grad_norm": 0.39522138237953186, "learning_rate": 9.984793538513862e-05, "loss": 0.04, "step": 7360 }, { "grad_norm": 0.3382217586040497, "learning_rate": 9.984664408393912e-05, "loss": 0.0457, "step": 7370 }, { "grad_norm": 0.3163924515247345, "learning_rate": 9.984534733158556e-05, "loss": 0.0377, "step": 7380 }, { "grad_norm": 0.29219353199005127, "learning_rate": 9.984404512821977e-05, "loss": 0.0403, "step": 7390 }, { "grad_norm": 0.38121750950813293, "learning_rate": 9.984273747398411e-05, "loss": 0.0388, "step": 7400 }, { "grad_norm": 0.3408188223838806, "learning_rate": 9.984142436902165e-05, "loss": 0.0403, "step": 7410 }, { "grad_norm": 0.3558816611766815, "learning_rate": 9.984010581347596e-05, "loss": 0.0486, "step": 7420 }, { "grad_norm": 0.5145797729492188, "learning_rate": 9.983878180749121e-05, "loss": 0.0458, "step": 7430 }, { "grad_norm": 0.3821069598197937, "learning_rate": 9.983745235121222e-05, "loss": 0.0441, "step": 7440 }, { "grad_norm": 0.37060850858688354, "learning_rate": 9.983611744478438e-05, "loss": 0.0413, "step": 7450 }, { "grad_norm": 0.36323729157447815, "learning_rate": 9.983477708835365e-05, "loss": 0.0431, "step": 7460 }, { "grad_norm": 0.3744618892669678, "learning_rate": 9.983343128206664e-05, "loss": 0.0478, "step": 7470 }, { "grad_norm": 0.5419213771820068, "learning_rate": 9.983208002607049e-05, "loss": 0.0519, "step": 7480 }, { "grad_norm": 0.3565937876701355, "learning_rate": 9.9830723320513e-05, "loss": 0.0452, "step": 7490 }, { "grad_norm": 0.34490495920181274, "learning_rate": 9.982936116554254e-05, "loss": 0.0399, "step": 7500 }, { "grad_norm": 0.4626024663448334, "learning_rate": 9.982799356130803e-05, "loss": 0.0353, "step": 7510 }, { "grad_norm": 0.31364676356315613, "learning_rate": 9.982662050795908e-05, "loss": 0.0354, "step": 7520 }, { "grad_norm": 0.358704537153244, "learning_rate": 9.982524200564583e-05, "loss": 0.0336, "step": 7530 }, { "grad_norm": 0.3181699812412262, "learning_rate": 9.982385805451901e-05, "loss": 0.0418, "step": 7540 }, { "grad_norm": 0.30988606810569763, "learning_rate": 9.982246865472998e-05, "loss": 0.0372, "step": 7550 }, { "grad_norm": 0.21042683720588684, "learning_rate": 9.982107380643069e-05, "loss": 0.0379, "step": 7560 }, { "grad_norm": 0.28291553258895874, "learning_rate": 9.981967350977368e-05, "loss": 0.0471, "step": 7570 }, { "grad_norm": 0.5097411870956421, "learning_rate": 9.981826776491208e-05, "loss": 0.0501, "step": 7580 }, { "grad_norm": 0.4104539155960083, "learning_rate": 9.98168565719996e-05, "loss": 0.0494, "step": 7590 }, { "grad_norm": 0.30304208397865295, "learning_rate": 9.98154399311906e-05, "loss": 0.0557, "step": 7600 }, { "grad_norm": 0.3093979060649872, "learning_rate": 9.981401784263997e-05, "loss": 0.0488, "step": 7610 }, { "grad_norm": 0.3637687861919403, "learning_rate": 9.981259030650326e-05, "loss": 0.0441, "step": 7620 }, { "grad_norm": 0.3288261294364929, "learning_rate": 9.981115732293655e-05, "loss": 0.0462, "step": 7630 }, { "grad_norm": 0.49741479754447937, "learning_rate": 9.980971889209659e-05, "loss": 0.0418, "step": 7640 }, { "grad_norm": 0.38700082898139954, "learning_rate": 9.980827501414064e-05, "loss": 0.0398, "step": 7650 }, { "grad_norm": 0.2882310748100281, "learning_rate": 9.980682568922663e-05, "loss": 0.0389, "step": 7660 }, { "grad_norm": 0.33081334829330444, "learning_rate": 9.980537091751304e-05, "loss": 0.0438, "step": 7670 }, { "grad_norm": 0.3856220841407776, "learning_rate": 9.980391069915897e-05, "loss": 0.0443, "step": 7680 }, { "grad_norm": 0.3075283467769623, "learning_rate": 9.98024450343241e-05, "loss": 0.0417, "step": 7690 }, { "grad_norm": 0.24967753887176514, "learning_rate": 9.980097392316872e-05, "loss": 0.0318, "step": 7700 }, { "grad_norm": 0.4584141969680786, "learning_rate": 9.97994973658537e-05, "loss": 0.0531, "step": 7710 }, { "grad_norm": 0.3154069185256958, "learning_rate": 9.979801536254054e-05, "loss": 0.0479, "step": 7720 }, { "grad_norm": 0.36245307326316833, "learning_rate": 9.979652791339127e-05, "loss": 0.0467, "step": 7730 }, { "grad_norm": 0.3618753254413605, "learning_rate": 9.97950350185686e-05, "loss": 0.0442, "step": 7740 }, { "grad_norm": 0.3066999912261963, "learning_rate": 9.979353667823574e-05, "loss": 0.0373, "step": 7750 }, { "grad_norm": 0.28114521503448486, "learning_rate": 9.979203289255658e-05, "loss": 0.0417, "step": 7760 }, { "grad_norm": 0.5020354986190796, "learning_rate": 9.979052366169557e-05, "loss": 0.0434, "step": 7770 }, { "grad_norm": 0.435264527797699, "learning_rate": 9.978900898581775e-05, "loss": 0.046, "step": 7780 }, { "grad_norm": 0.4063189923763275, "learning_rate": 9.978748886508875e-05, "loss": 0.0434, "step": 7790 }, { "grad_norm": 0.30877408385276794, "learning_rate": 9.978596329967484e-05, "loss": 0.0394, "step": 7800 }, { "grad_norm": 0.31499147415161133, "learning_rate": 9.978443228974284e-05, "loss": 0.0388, "step": 7810 }, { "grad_norm": 0.29157888889312744, "learning_rate": 9.978289583546015e-05, "loss": 0.032, "step": 7820 }, { "grad_norm": 0.3037221133708954, "learning_rate": 9.978135393699484e-05, "loss": 0.0358, "step": 7830 }, { "grad_norm": 0.4697336256504059, "learning_rate": 9.977980659451548e-05, "loss": 0.0333, "step": 7840 }, { "grad_norm": 0.26235684752464294, "learning_rate": 9.977825380819135e-05, "loss": 0.038, "step": 7850 }, { "grad_norm": 0.3590366244316101, "learning_rate": 9.97766955781922e-05, "loss": 0.0438, "step": 7860 }, { "grad_norm": 0.394405335187912, "learning_rate": 9.977513190468848e-05, "loss": 0.0461, "step": 7870 }, { "grad_norm": 0.3886561691761017, "learning_rate": 9.977356278785116e-05, "loss": 0.0411, "step": 7880 }, { "grad_norm": 0.244771808385849, "learning_rate": 9.977198822785184e-05, "loss": 0.0353, "step": 7890 }, { "grad_norm": 0.38931792974472046, "learning_rate": 9.977040822486273e-05, "loss": 0.0521, "step": 7900 }, { "grad_norm": 0.26630884408950806, "learning_rate": 9.97688227790566e-05, "loss": 0.0481, "step": 7910 }, { "grad_norm": 0.2844429910182953, "learning_rate": 9.976723189060684e-05, "loss": 0.0641, "step": 7920 }, { "grad_norm": 0.30831411480903625, "learning_rate": 9.976563555968742e-05, "loss": 0.0412, "step": 7930 }, { "grad_norm": 0.3403984606266022, "learning_rate": 9.976403378647292e-05, "loss": 0.0436, "step": 7940 }, { "grad_norm": 0.37202802300453186, "learning_rate": 9.97624265711385e-05, "loss": 0.039, "step": 7950 }, { "grad_norm": 0.3810587227344513, "learning_rate": 9.976081391385993e-05, "loss": 0.0396, "step": 7960 }, { "grad_norm": 0.27541717886924744, "learning_rate": 9.975919581481356e-05, "loss": 0.0372, "step": 7970 }, { "grad_norm": 0.3339158594608307, "learning_rate": 9.975757227417634e-05, "loss": 0.0394, "step": 7980 }, { "grad_norm": 0.33269694447517395, "learning_rate": 9.975594329212586e-05, "loss": 0.0358, "step": 7990 }, { "grad_norm": 0.33172258734703064, "learning_rate": 9.97543088688402e-05, "loss": 0.0354, "step": 8000 }, { "grad_norm": 0.29533782601356506, "learning_rate": 9.975266900449814e-05, "loss": 0.0338, "step": 8010 }, { "grad_norm": 0.334717333316803, "learning_rate": 9.975102369927898e-05, "loss": 0.0366, "step": 8020 }, { "grad_norm": 0.27103033661842346, "learning_rate": 9.974937295336269e-05, "loss": 0.0342, "step": 8030 }, { "grad_norm": 0.32222798466682434, "learning_rate": 9.974771676692975e-05, "loss": 0.0351, "step": 8040 }, { "grad_norm": 0.35282987356185913, "learning_rate": 9.974605514016131e-05, "loss": 0.0399, "step": 8050 }, { "grad_norm": 0.2913660705089569, "learning_rate": 9.974438807323907e-05, "loss": 0.0328, "step": 8060 }, { "grad_norm": 0.31634610891342163, "learning_rate": 9.974271556634535e-05, "loss": 0.036, "step": 8070 }, { "grad_norm": 0.3847522437572479, "learning_rate": 9.974103761966302e-05, "loss": 0.0507, "step": 8080 }, { "grad_norm": 0.3583144545555115, "learning_rate": 9.973935423337563e-05, "loss": 0.0407, "step": 8090 }, { "grad_norm": 0.4359554946422577, "learning_rate": 9.973766540766722e-05, "loss": 0.0474, "step": 8100 }, { "grad_norm": 0.41042235493659973, "learning_rate": 9.97359711427225e-05, "loss": 0.0488, "step": 8110 }, { "grad_norm": 0.3547172248363495, "learning_rate": 9.973427143872677e-05, "loss": 0.0393, "step": 8120 }, { "grad_norm": 0.3480238616466522, "learning_rate": 9.973256629586589e-05, "loss": 0.0533, "step": 8130 }, { "grad_norm": 0.49000585079193115, "learning_rate": 9.973085571432632e-05, "loss": 0.0466, "step": 8140 }, { "grad_norm": 0.41198715567588806, "learning_rate": 9.972913969429513e-05, "loss": 0.0397, "step": 8150 }, { "grad_norm": 0.398416131734848, "learning_rate": 9.972741823596e-05, "loss": 0.0485, "step": 8160 }, { "grad_norm": 0.31005141139030457, "learning_rate": 9.972569133950917e-05, "loss": 0.044, "step": 8170 }, { "grad_norm": 0.28182998299598694, "learning_rate": 9.972395900513151e-05, "loss": 0.0365, "step": 8180 }, { "grad_norm": 0.3340926766395569, "learning_rate": 9.972222123301645e-05, "loss": 0.0379, "step": 8190 }, { "grad_norm": 0.32722029089927673, "learning_rate": 9.972047802335403e-05, "loss": 0.0376, "step": 8200 }, { "grad_norm": 0.2337450385093689, "learning_rate": 9.971872937633488e-05, "loss": 0.0324, "step": 8210 }, { "grad_norm": 0.2965572774410248, "learning_rate": 9.971697529215024e-05, "loss": 0.0374, "step": 8220 }, { "grad_norm": 0.2997048497200012, "learning_rate": 9.971521577099192e-05, "loss": 0.0386, "step": 8230 }, { "grad_norm": 0.34191620349884033, "learning_rate": 9.971345081305236e-05, "loss": 0.0378, "step": 8240 }, { "grad_norm": 0.29535654187202454, "learning_rate": 9.971168041852456e-05, "loss": 0.0365, "step": 8250 }, { "grad_norm": 0.38552847504615784, "learning_rate": 9.970990458760215e-05, "loss": 0.0309, "step": 8260 }, { "grad_norm": 0.3351072371006012, "learning_rate": 9.970812332047929e-05, "loss": 0.0334, "step": 8270 }, { "grad_norm": 0.3670240640640259, "learning_rate": 9.97063366173508e-05, "loss": 0.0468, "step": 8280 }, { "grad_norm": 0.3362395167350769, "learning_rate": 9.970454447841207e-05, "loss": 0.0333, "step": 8290 }, { "grad_norm": 0.3832468092441559, "learning_rate": 9.970274690385909e-05, "loss": 0.0413, "step": 8300 }, { "grad_norm": 0.3239048719406128, "learning_rate": 9.970094389388844e-05, "loss": 0.0352, "step": 8310 }, { "grad_norm": 0.3124125897884369, "learning_rate": 9.969913544869728e-05, "loss": 0.0324, "step": 8320 }, { "grad_norm": 0.30941829085350037, "learning_rate": 9.96973215684834e-05, "loss": 0.0387, "step": 8330 }, { "grad_norm": 0.2623038589954376, "learning_rate": 9.969550225344513e-05, "loss": 0.0324, "step": 8340 }, { "grad_norm": 0.3272646963596344, "learning_rate": 9.969367750378147e-05, "loss": 0.0406, "step": 8350 }, { "grad_norm": 0.3708786368370056, "learning_rate": 9.969184731969194e-05, "loss": 0.0331, "step": 8360 }, { "grad_norm": 0.3390074372291565, "learning_rate": 9.96900117013767e-05, "loss": 0.032, "step": 8370 }, { "grad_norm": 0.3591773211956024, "learning_rate": 9.96881706490365e-05, "loss": 0.0392, "step": 8380 }, { "grad_norm": 0.27105578780174255, "learning_rate": 9.968632416287265e-05, "loss": 0.0332, "step": 8390 }, { "grad_norm": 0.30863597989082336, "learning_rate": 9.96844722430871e-05, "loss": 0.0319, "step": 8400 }, { "grad_norm": 0.3635157346725464, "learning_rate": 9.968261488988235e-05, "loss": 0.0368, "step": 8410 }, { "grad_norm": 0.30256345868110657, "learning_rate": 9.968075210346155e-05, "loss": 0.0355, "step": 8420 }, { "grad_norm": 0.37057700753211975, "learning_rate": 9.967888388402839e-05, "loss": 0.0355, "step": 8430 }, { "grad_norm": 0.2905457019805908, "learning_rate": 9.967701023178717e-05, "loss": 0.0374, "step": 8440 }, { "grad_norm": 0.32139351963996887, "learning_rate": 9.967513114694282e-05, "loss": 0.0358, "step": 8450 }, { "grad_norm": 0.2704802453517914, "learning_rate": 9.967324662970079e-05, "loss": 0.0338, "step": 8460 }, { "grad_norm": 0.3388743996620178, "learning_rate": 9.96713566802672e-05, "loss": 0.0318, "step": 8470 }, { "grad_norm": 0.3950795531272888, "learning_rate": 9.966946129884873e-05, "loss": 0.0348, "step": 8480 }, { "grad_norm": 0.46915093064308167, "learning_rate": 9.966756048565265e-05, "loss": 0.0398, "step": 8490 }, { "grad_norm": 0.3695997893810272, "learning_rate": 9.966565424088681e-05, "loss": 0.0456, "step": 8500 }, { "grad_norm": 0.3334856927394867, "learning_rate": 9.96637425647597e-05, "loss": 0.0356, "step": 8510 }, { "grad_norm": 0.33253881335258484, "learning_rate": 9.966182545748038e-05, "loss": 0.0292, "step": 8520 }, { "grad_norm": 0.30552002787590027, "learning_rate": 9.96599029192585e-05, "loss": 0.0369, "step": 8530 }, { "grad_norm": 0.25112080574035645, "learning_rate": 9.965797495030428e-05, "loss": 0.037, "step": 8540 }, { "grad_norm": 0.3267311751842499, "learning_rate": 9.96560415508286e-05, "loss": 0.0397, "step": 8550 }, { "grad_norm": 0.33460885286331177, "learning_rate": 9.965410272104286e-05, "loss": 0.0362, "step": 8560 }, { "grad_norm": 0.4331735372543335, "learning_rate": 9.96521584611591e-05, "loss": 0.0641, "step": 8570 }, { "grad_norm": 0.4295107424259186, "learning_rate": 9.965020877138994e-05, "loss": 0.0455, "step": 8580 }, { "grad_norm": 0.364651620388031, "learning_rate": 9.964825365194861e-05, "loss": 0.04, "step": 8590 }, { "grad_norm": 0.30338847637176514, "learning_rate": 9.96462931030489e-05, "loss": 0.0309, "step": 8600 }, { "grad_norm": 0.4606633484363556, "learning_rate": 9.96443271249052e-05, "loss": 0.0331, "step": 8610 }, { "grad_norm": 0.3834570050239563, "learning_rate": 9.964235571773255e-05, "loss": 0.0421, "step": 8620 }, { "grad_norm": 0.2859823405742645, "learning_rate": 9.96403788817465e-05, "loss": 0.0334, "step": 8630 }, { "grad_norm": 0.3335959315299988, "learning_rate": 9.963839661716325e-05, "loss": 0.0394, "step": 8640 }, { "grad_norm": 0.28629931807518005, "learning_rate": 9.963640892419958e-05, "loss": 0.0408, "step": 8650 }, { "grad_norm": 0.3337242305278778, "learning_rate": 9.963441580307286e-05, "loss": 0.0362, "step": 8660 }, { "grad_norm": 0.3612450659275055, "learning_rate": 9.963241725400104e-05, "loss": 0.0335, "step": 8670 }, { "grad_norm": 0.2929411828517914, "learning_rate": 9.963041327720271e-05, "loss": 0.0288, "step": 8680 }, { "grad_norm": 0.26308977603912354, "learning_rate": 9.962840387289697e-05, "loss": 0.0322, "step": 8690 }, { "grad_norm": 0.26539215445518494, "learning_rate": 9.962638904130363e-05, "loss": 0.0342, "step": 8700 }, { "grad_norm": 0.37204745411872864, "learning_rate": 9.962436878264298e-05, "loss": 0.0409, "step": 8710 }, { "grad_norm": 0.4519767761230469, "learning_rate": 9.962234309713598e-05, "loss": 0.0522, "step": 8720 }, { "grad_norm": 0.2672363221645355, "learning_rate": 9.962031198500414e-05, "loss": 0.0353, "step": 8730 }, { "grad_norm": 0.20370937883853912, "learning_rate": 9.961827544646958e-05, "loss": 0.0333, "step": 8740 }, { "grad_norm": 0.3071867525577545, "learning_rate": 9.961623348175501e-05, "loss": 0.0406, "step": 8750 }, { "grad_norm": 0.33614328503608704, "learning_rate": 9.961418609108377e-05, "loss": 0.0333, "step": 8760 }, { "grad_norm": 0.33174073696136475, "learning_rate": 9.961213327467971e-05, "loss": 0.0405, "step": 8770 }, { "grad_norm": 0.3284136950969696, "learning_rate": 9.961007503276736e-05, "loss": 0.0465, "step": 8780 }, { "grad_norm": 0.44630923867225647, "learning_rate": 9.960801136557179e-05, "loss": 0.0382, "step": 8790 }, { "grad_norm": 0.3088971972465515, "learning_rate": 9.960594227331866e-05, "loss": 0.0314, "step": 8800 }, { "grad_norm": 0.29954108595848083, "learning_rate": 9.960386775623429e-05, "loss": 0.0344, "step": 8810 }, { "grad_norm": 0.3303561508655548, "learning_rate": 9.96017878145455e-05, "loss": 0.0342, "step": 8820 }, { "grad_norm": 0.35773032903671265, "learning_rate": 9.959970244847977e-05, "loss": 0.0363, "step": 8830 }, { "grad_norm": 0.3727858364582062, "learning_rate": 9.959761165826518e-05, "loss": 0.0351, "step": 8840 }, { "grad_norm": 0.26926320791244507, "learning_rate": 9.959551544413033e-05, "loss": 0.0309, "step": 8850 }, { "grad_norm": 0.32789376378059387, "learning_rate": 9.959341380630448e-05, "loss": 0.0357, "step": 8860 }, { "grad_norm": 0.34388038516044617, "learning_rate": 9.959130674501746e-05, "loss": 0.0378, "step": 8870 }, { "grad_norm": 0.24700604379177094, "learning_rate": 9.958919426049968e-05, "loss": 0.0349, "step": 8880 }, { "grad_norm": 0.2731369435787201, "learning_rate": 9.958707635298219e-05, "loss": 0.0396, "step": 8890 }, { "grad_norm": 0.29203397035598755, "learning_rate": 9.958495302269657e-05, "loss": 0.0367, "step": 8900 }, { "grad_norm": 0.3501965403556824, "learning_rate": 9.958282426987503e-05, "loss": 0.0333, "step": 8910 }, { "grad_norm": 0.4486680328845978, "learning_rate": 9.95806900947504e-05, "loss": 0.0463, "step": 8920 }, { "grad_norm": 0.27970534563064575, "learning_rate": 9.957855049755604e-05, "loss": 0.0333, "step": 8930 }, { "grad_norm": 0.3286975622177124, "learning_rate": 9.957640547852593e-05, "loss": 0.0398, "step": 8940 }, { "grad_norm": 0.3959602415561676, "learning_rate": 9.957425503789466e-05, "loss": 0.0343, "step": 8950 }, { "grad_norm": 0.26844412088394165, "learning_rate": 9.957209917589738e-05, "loss": 0.0318, "step": 8960 }, { "grad_norm": 0.40796127915382385, "learning_rate": 9.956993789276987e-05, "loss": 0.0323, "step": 8970 }, { "grad_norm": 0.2597719728946686, "learning_rate": 9.956777118874847e-05, "loss": 0.0383, "step": 8980 }, { "grad_norm": 0.49494802951812744, "learning_rate": 9.956559906407016e-05, "loss": 0.0596, "step": 8990 }, { "grad_norm": 0.3469149172306061, "learning_rate": 9.956342151897245e-05, "loss": 0.0413, "step": 9000 }, { "grad_norm": 0.36657553911209106, "learning_rate": 9.956123855369346e-05, "loss": 0.0322, "step": 9010 }, { "grad_norm": 0.33418238162994385, "learning_rate": 9.955905016847196e-05, "loss": 0.037, "step": 9020 }, { "grad_norm": 0.3453254699707031, "learning_rate": 9.955685636354723e-05, "loss": 0.0482, "step": 9030 }, { "grad_norm": 0.4433116912841797, "learning_rate": 9.95546571391592e-05, "loss": 0.0421, "step": 9040 }, { "grad_norm": 0.3410087525844574, "learning_rate": 9.955245249554837e-05, "loss": 0.0365, "step": 9050 }, { "grad_norm": 0.28882908821105957, "learning_rate": 9.955024243295582e-05, "loss": 0.0458, "step": 9060 }, { "grad_norm": 0.28956693410873413, "learning_rate": 9.954802695162328e-05, "loss": 0.0292, "step": 9070 }, { "grad_norm": 0.3114205002784729, "learning_rate": 9.954580605179302e-05, "loss": 0.0316, "step": 9080 }, { "grad_norm": 0.3409483730792999, "learning_rate": 9.954357973370788e-05, "loss": 0.0414, "step": 9090 }, { "grad_norm": 0.3909824788570404, "learning_rate": 9.954134799761135e-05, "loss": 0.0434, "step": 9100 }, { "grad_norm": 0.38976046442985535, "learning_rate": 9.953911084374748e-05, "loss": 0.0375, "step": 9110 }, { "grad_norm": 0.301056444644928, "learning_rate": 9.953686827236093e-05, "loss": 0.0362, "step": 9120 }, { "grad_norm": 0.24016357958316803, "learning_rate": 9.953462028369695e-05, "loss": 0.0427, "step": 9130 }, { "grad_norm": 0.34886083006858826, "learning_rate": 9.953236687800136e-05, "loss": 0.0299, "step": 9140 }, { "grad_norm": 0.3299098610877991, "learning_rate": 9.95301080555206e-05, "loss": 0.0427, "step": 9150 }, { "grad_norm": 0.3004816770553589, "learning_rate": 9.952784381650171e-05, "loss": 0.0291, "step": 9160 }, { "grad_norm": 0.2905387580394745, "learning_rate": 9.952557416119226e-05, "loss": 0.0313, "step": 9170 }, { "grad_norm": 0.2725183665752411, "learning_rate": 9.95232990898405e-05, "loss": 0.0312, "step": 9180 }, { "grad_norm": 0.27299216389656067, "learning_rate": 9.95210186026952e-05, "loss": 0.0262, "step": 9190 }, { "grad_norm": 0.2529684603214264, "learning_rate": 9.951873270000576e-05, "loss": 0.0338, "step": 9200 }, { "grad_norm": 0.29495760798454285, "learning_rate": 9.951644138202216e-05, "loss": 0.0443, "step": 9210 }, { "grad_norm": 0.2565673589706421, "learning_rate": 9.951414464899498e-05, "loss": 0.0403, "step": 9220 }, { "grad_norm": 0.3326755464076996, "learning_rate": 9.951184250117538e-05, "loss": 0.0353, "step": 9230 }, { "grad_norm": 0.23443564772605896, "learning_rate": 9.950953493881513e-05, "loss": 0.0423, "step": 9240 }, { "grad_norm": 0.31082484126091003, "learning_rate": 9.950722196216658e-05, "loss": 0.028, "step": 9250 }, { "grad_norm": 0.3123813271522522, "learning_rate": 9.950490357148265e-05, "loss": 0.0296, "step": 9260 }, { "grad_norm": 0.23592843115329742, "learning_rate": 9.950257976701692e-05, "loss": 0.0348, "step": 9270 }, { "grad_norm": 0.2945682406425476, "learning_rate": 9.950025054902348e-05, "loss": 0.0335, "step": 9280 }, { "grad_norm": 0.45515328645706177, "learning_rate": 9.949791591775706e-05, "loss": 0.046, "step": 9290 }, { "grad_norm": 0.29989951848983765, "learning_rate": 9.949557587347298e-05, "loss": 0.0485, "step": 9300 }, { "grad_norm": 0.36513447761535645, "learning_rate": 9.949323041642713e-05, "loss": 0.0369, "step": 9310 }, { "grad_norm": 0.3658713698387146, "learning_rate": 9.949087954687602e-05, "loss": 0.0527, "step": 9320 }, { "grad_norm": 0.3560214936733246, "learning_rate": 9.948852326507672e-05, "loss": 0.0444, "step": 9330 }, { "grad_norm": 0.34303465485572815, "learning_rate": 9.948616157128694e-05, "loss": 0.0437, "step": 9340 }, { "grad_norm": 0.3551710844039917, "learning_rate": 9.948379446576493e-05, "loss": 0.0375, "step": 9350 }, { "grad_norm": 0.2724393606185913, "learning_rate": 9.948142194876952e-05, "loss": 0.0307, "step": 9360 }, { "grad_norm": 0.3648245930671692, "learning_rate": 9.947904402056024e-05, "loss": 0.0487, "step": 9370 }, { "grad_norm": 0.2884514331817627, "learning_rate": 9.947666068139708e-05, "loss": 0.037, "step": 9380 }, { "grad_norm": 0.34478041529655457, "learning_rate": 9.947427193154071e-05, "loss": 0.0336, "step": 9390 }, { "grad_norm": 0.3712116777896881, "learning_rate": 9.947187777125233e-05, "loss": 0.0441, "step": 9400 }, { "grad_norm": 0.39745190739631653, "learning_rate": 9.946947820079377e-05, "loss": 0.0403, "step": 9410 }, { "grad_norm": 0.4495091140270233, "learning_rate": 9.946707322042747e-05, "loss": 0.0371, "step": 9420 }, { "grad_norm": 0.24971836805343628, "learning_rate": 9.94646628304164e-05, "loss": 0.0316, "step": 9430 }, { "grad_norm": 0.36190247535705566, "learning_rate": 9.946224703102418e-05, "loss": 0.0335, "step": 9440 }, { "grad_norm": 0.28841298818588257, "learning_rate": 9.945982582251498e-05, "loss": 0.0305, "step": 9450 }, { "grad_norm": 0.26972225308418274, "learning_rate": 9.94573992051536e-05, "loss": 0.0288, "step": 9460 }, { "grad_norm": 0.32452213764190674, "learning_rate": 9.94549671792054e-05, "loss": 0.0477, "step": 9470 }, { "grad_norm": 0.265654057264328, "learning_rate": 9.945252974493635e-05, "loss": 0.0334, "step": 9480 }, { "grad_norm": 0.33540230989456177, "learning_rate": 9.9450086902613e-05, "loss": 0.0319, "step": 9490 }, { "grad_norm": 0.2833904027938843, "learning_rate": 9.944763865250248e-05, "loss": 0.0297, "step": 9500 }, { "grad_norm": 0.21889053285121918, "learning_rate": 9.944518499487254e-05, "loss": 0.0325, "step": 9510 }, { "grad_norm": 0.3309599459171295, "learning_rate": 9.944272592999151e-05, "loss": 0.0363, "step": 9520 }, { "grad_norm": 0.24016353487968445, "learning_rate": 9.94402614581283e-05, "loss": 0.0321, "step": 9530 }, { "grad_norm": 0.5105071067810059, "learning_rate": 9.943779157955244e-05, "loss": 0.0427, "step": 9540 }, { "grad_norm": 0.28752365708351135, "learning_rate": 9.943531629453403e-05, "loss": 0.0356, "step": 9550 }, { "grad_norm": 0.38507986068725586, "learning_rate": 9.943283560334375e-05, "loss": 0.0426, "step": 9560 }, { "grad_norm": 0.270968496799469, "learning_rate": 9.943034950625288e-05, "loss": 0.0515, "step": 9570 }, { "grad_norm": 0.354124516248703, "learning_rate": 9.942785800353332e-05, "loss": 0.0412, "step": 9580 }, { "grad_norm": 0.34162798523902893, "learning_rate": 9.942536109545751e-05, "loss": 0.0355, "step": 9590 }, { "grad_norm": 0.30225980281829834, "learning_rate": 9.942285878229853e-05, "loss": 0.0329, "step": 9600 }, { "grad_norm": 0.31122031807899475, "learning_rate": 9.942035106433001e-05, "loss": 0.0318, "step": 9610 }, { "grad_norm": 0.30079641938209534, "learning_rate": 9.94178379418262e-05, "loss": 0.0364, "step": 9620 }, { "grad_norm": 0.23015478253364563, "learning_rate": 9.941531941506194e-05, "loss": 0.0305, "step": 9630 }, { "grad_norm": 0.28525885939598083, "learning_rate": 9.941279548431263e-05, "loss": 0.0276, "step": 9640 }, { "grad_norm": 0.19526536762714386, "learning_rate": 9.941026614985431e-05, "loss": 0.0314, "step": 9650 }, { "grad_norm": 0.3631061911582947, "learning_rate": 9.940773141196357e-05, "loss": 0.0328, "step": 9660 }, { "grad_norm": 0.3072575628757477, "learning_rate": 9.94051912709176e-05, "loss": 0.0382, "step": 9670 }, { "grad_norm": 0.2972184717655182, "learning_rate": 9.940264572699421e-05, "loss": 0.0275, "step": 9680 }, { "grad_norm": 0.29194778203964233, "learning_rate": 9.940009478047174e-05, "loss": 0.033, "step": 9690 }, { "grad_norm": 0.2711225748062134, "learning_rate": 9.939753843162918e-05, "loss": 0.0275, "step": 9700 }, { "grad_norm": 0.26688915491104126, "learning_rate": 9.939497668074609e-05, "loss": 0.0267, "step": 9710 }, { "grad_norm": 0.3410705029964447, "learning_rate": 9.93924095281026e-05, "loss": 0.0328, "step": 9720 }, { "grad_norm": 0.3340420126914978, "learning_rate": 9.938983697397948e-05, "loss": 0.038, "step": 9730 }, { "grad_norm": 0.4060855507850647, "learning_rate": 9.938725901865805e-05, "loss": 0.0314, "step": 9740 }, { "grad_norm": 0.25217095017433167, "learning_rate": 9.93846756624202e-05, "loss": 0.0311, "step": 9750 }, { "grad_norm": 0.25454795360565186, "learning_rate": 9.938208690554849e-05, "loss": 0.0332, "step": 9760 }, { "grad_norm": 0.34205904603004456, "learning_rate": 9.9379492748326e-05, "loss": 0.0329, "step": 9770 }, { "grad_norm": 0.3111865222454071, "learning_rate": 9.937689319103641e-05, "loss": 0.0327, "step": 9780 }, { "grad_norm": 0.33230724930763245, "learning_rate": 9.937428823396404e-05, "loss": 0.0292, "step": 9790 }, { "grad_norm": 0.36545032262802124, "learning_rate": 9.937167787739372e-05, "loss": 0.0277, "step": 9800 }, { "grad_norm": 0.3395206332206726, "learning_rate": 9.936906212161095e-05, "loss": 0.0321, "step": 9810 }, { "grad_norm": 0.31090906262397766, "learning_rate": 9.936644096690176e-05, "loss": 0.036, "step": 9820 }, { "grad_norm": 0.3097941279411316, "learning_rate": 9.936381441355282e-05, "loss": 0.033, "step": 9830 }, { "grad_norm": 0.2739008665084839, "learning_rate": 9.936118246185136e-05, "loss": 0.0422, "step": 9840 }, { "grad_norm": 0.35293304920196533, "learning_rate": 9.935854511208518e-05, "loss": 0.0395, "step": 9850 }, { "grad_norm": 0.2784968912601471, "learning_rate": 9.935590236454272e-05, "loss": 0.0272, "step": 9860 }, { "grad_norm": 0.17581260204315186, "learning_rate": 9.935325421951298e-05, "loss": 0.027, "step": 9870 }, { "grad_norm": 0.33297213912010193, "learning_rate": 9.935060067728557e-05, "loss": 0.0355, "step": 9880 }, { "grad_norm": 0.23992134630680084, "learning_rate": 9.934794173815067e-05, "loss": 0.0263, "step": 9890 }, { "grad_norm": 0.3258779048919678, "learning_rate": 9.934527740239906e-05, "loss": 0.0312, "step": 9900 }, { "grad_norm": 0.24677518010139465, "learning_rate": 9.934260767032209e-05, "loss": 0.0335, "step": 9910 }, { "grad_norm": 0.2853933274745941, "learning_rate": 9.933993254221172e-05, "loss": 0.0347, "step": 9920 }, { "grad_norm": 0.24607616662979126, "learning_rate": 9.933725201836053e-05, "loss": 0.0266, "step": 9930 }, { "grad_norm": 0.343521386384964, "learning_rate": 9.933456609906162e-05, "loss": 0.0306, "step": 9940 }, { "grad_norm": 0.26507964730262756, "learning_rate": 9.933187478460875e-05, "loss": 0.0258, "step": 9950 }, { "grad_norm": 0.26392966508865356, "learning_rate": 9.93291780752962e-05, "loss": 0.0263, "step": 9960 }, { "grad_norm": 0.28132596611976624, "learning_rate": 9.932647597141893e-05, "loss": 0.034, "step": 9970 }, { "grad_norm": 0.34164923429489136, "learning_rate": 9.932376847327239e-05, "loss": 0.0306, "step": 9980 }, { "grad_norm": 0.30036354064941406, "learning_rate": 9.932105558115268e-05, "loss": 0.035, "step": 9990 }, { "grad_norm": 0.39716631174087524, "learning_rate": 9.931833729535651e-05, "loss": 0.0374, "step": 10000 }, { "grad_norm": 0.330174058675766, "learning_rate": 9.931561361618111e-05, "loss": 0.0344, "step": 10010 }, { "grad_norm": 0.3047487735748291, "learning_rate": 9.931288454392435e-05, "loss": 0.0352, "step": 10020 }, { "grad_norm": 0.25069332122802734, "learning_rate": 9.931015007888467e-05, "loss": 0.0261, "step": 10030 }, { "grad_norm": 0.3021503686904907, "learning_rate": 9.930741022136112e-05, "loss": 0.0353, "step": 10040 }, { "grad_norm": 0.2836885452270508, "learning_rate": 9.930466497165333e-05, "loss": 0.0344, "step": 10050 }, { "grad_norm": 0.3988097608089447, "learning_rate": 9.93019143300615e-05, "loss": 0.0364, "step": 10060 }, { "grad_norm": 0.2566373348236084, "learning_rate": 9.929915829688644e-05, "loss": 0.0242, "step": 10070 }, { "grad_norm": 0.24227192997932434, "learning_rate": 9.929639687242955e-05, "loss": 0.0307, "step": 10080 }, { "grad_norm": 0.2809053063392639, "learning_rate": 9.929363005699281e-05, "loss": 0.0306, "step": 10090 }, { "grad_norm": 0.32953518629074097, "learning_rate": 9.92908578508788e-05, "loss": 0.0314, "step": 10100 }, { "grad_norm": 0.29236817359924316, "learning_rate": 9.928808025439069e-05, "loss": 0.0242, "step": 10110 }, { "grad_norm": 0.2837081551551819, "learning_rate": 9.928529726783223e-05, "loss": 0.028, "step": 10120 }, { "grad_norm": 0.33269596099853516, "learning_rate": 9.928250889150774e-05, "loss": 0.0332, "step": 10130 }, { "grad_norm": 0.3513176143169403, "learning_rate": 9.92797151257222e-05, "loss": 0.0308, "step": 10140 }, { "grad_norm": 0.3304891884326935, "learning_rate": 9.927691597078108e-05, "loss": 0.0363, "step": 10150 }, { "grad_norm": 0.29217857122421265, "learning_rate": 9.927411142699053e-05, "loss": 0.0308, "step": 10160 }, { "grad_norm": 0.21156974136829376, "learning_rate": 9.927130149465725e-05, "loss": 0.0294, "step": 10170 }, { "grad_norm": 0.27920594811439514, "learning_rate": 9.92684861740885e-05, "loss": 0.0352, "step": 10180 }, { "grad_norm": 0.38380929827690125, "learning_rate": 9.926566546559217e-05, "loss": 0.0409, "step": 10190 }, { "grad_norm": 0.28422802686691284, "learning_rate": 9.926283936947673e-05, "loss": 0.0341, "step": 10200 }, { "grad_norm": 0.19886669516563416, "learning_rate": 9.926000788605126e-05, "loss": 0.033, "step": 10210 }, { "grad_norm": 0.21634899079799652, "learning_rate": 9.92571710156254e-05, "loss": 0.0292, "step": 10220 }, { "grad_norm": 0.2808785140514374, "learning_rate": 9.925432875850936e-05, "loss": 0.0365, "step": 10230 }, { "grad_norm": 0.3412841260433197, "learning_rate": 9.925148111501396e-05, "loss": 0.031, "step": 10240 }, { "grad_norm": 0.25459370017051697, "learning_rate": 9.924862808545066e-05, "loss": 0.0311, "step": 10250 }, { "grad_norm": 0.2964625954627991, "learning_rate": 9.924576967013141e-05, "loss": 0.0271, "step": 10260 }, { "grad_norm": 0.2696995735168457, "learning_rate": 9.924290586936887e-05, "loss": 0.0309, "step": 10270 }, { "grad_norm": 0.2744312286376953, "learning_rate": 9.924003668347614e-05, "loss": 0.0335, "step": 10280 }, { "grad_norm": 0.27946504950523376, "learning_rate": 9.923716211276704e-05, "loss": 0.0397, "step": 10290 }, { "grad_norm": 0.3271830976009369, "learning_rate": 9.923428215755594e-05, "loss": 0.0302, "step": 10300 }, { "grad_norm": 0.3207140862941742, "learning_rate": 9.923139681815775e-05, "loss": 0.0288, "step": 10310 }, { "grad_norm": 0.3491053879261017, "learning_rate": 9.922850609488801e-05, "loss": 0.0336, "step": 10320 }, { "grad_norm": 0.4009222388267517, "learning_rate": 9.922560998806287e-05, "loss": 0.0287, "step": 10330 }, { "grad_norm": 0.2692245841026306, "learning_rate": 9.922270849799905e-05, "loss": 0.0323, "step": 10340 }, { "grad_norm": 0.33246734738349915, "learning_rate": 9.92198016250138e-05, "loss": 0.0374, "step": 10350 }, { "grad_norm": 0.24809333682060242, "learning_rate": 9.921688936942506e-05, "loss": 0.0281, "step": 10360 }, { "grad_norm": 0.33565112948417664, "learning_rate": 9.921397173155129e-05, "loss": 0.0279, "step": 10370 }, { "grad_norm": 0.2169666290283203, "learning_rate": 9.921104871171157e-05, "loss": 0.0289, "step": 10380 }, { "grad_norm": 0.21502365171909332, "learning_rate": 9.920812031022554e-05, "loss": 0.0275, "step": 10390 }, { "grad_norm": 0.24017955362796783, "learning_rate": 9.920518652741348e-05, "loss": 0.0315, "step": 10400 }, { "grad_norm": 0.24304181337356567, "learning_rate": 9.920224736359618e-05, "loss": 0.0276, "step": 10410 }, { "grad_norm": 0.2851497232913971, "learning_rate": 9.91993028190951e-05, "loss": 0.0329, "step": 10420 }, { "grad_norm": 0.3222920298576355, "learning_rate": 9.919635289423222e-05, "loss": 0.0241, "step": 10430 }, { "grad_norm": 0.29981088638305664, "learning_rate": 9.919339758933015e-05, "loss": 0.0259, "step": 10440 }, { "grad_norm": 0.28807932138442993, "learning_rate": 9.919043690471209e-05, "loss": 0.0329, "step": 10450 }, { "grad_norm": 0.2960050106048584, "learning_rate": 9.91874708407018e-05, "loss": 0.0285, "step": 10460 }, { "grad_norm": 0.28497815132141113, "learning_rate": 9.918449939762367e-05, "loss": 0.0291, "step": 10470 }, { "grad_norm": 0.27539312839508057, "learning_rate": 9.91815225758026e-05, "loss": 0.0274, "step": 10480 }, { "grad_norm": 0.2931763827800751, "learning_rate": 9.917854037556419e-05, "loss": 0.026, "step": 10490 }, { "grad_norm": 0.23315143585205078, "learning_rate": 9.917555279723454e-05, "loss": 0.0236, "step": 10500 }, { "grad_norm": 0.33411547541618347, "learning_rate": 9.917255984114036e-05, "loss": 0.036, "step": 10510 }, { "grad_norm": 0.2726949453353882, "learning_rate": 9.916956150760896e-05, "loss": 0.0236, "step": 10520 }, { "grad_norm": 0.24397622048854828, "learning_rate": 9.916655779696826e-05, "loss": 0.0289, "step": 10530 }, { "grad_norm": 0.23924513161182404, "learning_rate": 9.916354870954671e-05, "loss": 0.0275, "step": 10540 }, { "grad_norm": 0.2597835659980774, "learning_rate": 9.91605342456734e-05, "loss": 0.0231, "step": 10550 }, { "grad_norm": 0.21067467331886292, "learning_rate": 9.915751440567795e-05, "loss": 0.023, "step": 10560 }, { "grad_norm": 0.2799851894378662, "learning_rate": 9.915448918989066e-05, "loss": 0.0242, "step": 10570 }, { "grad_norm": 0.3198325037956238, "learning_rate": 9.915145859864232e-05, "loss": 0.0299, "step": 10580 }, { "grad_norm": 0.3449134826660156, "learning_rate": 9.914842263226437e-05, "loss": 0.0351, "step": 10590 }, { "grad_norm": 0.34474247694015503, "learning_rate": 9.914538129108882e-05, "loss": 0.0441, "step": 10600 }, { "grad_norm": 0.2361019253730774, "learning_rate": 9.914233457544825e-05, "loss": 0.0505, "step": 10610 }, { "grad_norm": 0.4100811779499054, "learning_rate": 9.913928248567586e-05, "loss": 0.0471, "step": 10620 }, { "grad_norm": 0.2370147556066513, "learning_rate": 9.913622502210542e-05, "loss": 0.0274, "step": 10630 }, { "grad_norm": 0.2691366970539093, "learning_rate": 9.913316218507128e-05, "loss": 0.0383, "step": 10640 }, { "grad_norm": 0.2531528174877167, "learning_rate": 9.91300939749084e-05, "loss": 0.0362, "step": 10650 }, { "grad_norm": 0.2953074276447296, "learning_rate": 9.91270203919523e-05, "loss": 0.0361, "step": 10660 }, { "grad_norm": 0.3236953914165497, "learning_rate": 9.912394143653912e-05, "loss": 0.035, "step": 10670 }, { "grad_norm": 0.21129684150218964, "learning_rate": 9.912085710900555e-05, "loss": 0.023, "step": 10680 }, { "grad_norm": 0.29229408502578735, "learning_rate": 9.911776740968892e-05, "loss": 0.0249, "step": 10690 }, { "grad_norm": 0.33939021825790405, "learning_rate": 9.911467233892709e-05, "loss": 0.0458, "step": 10700 }, { "grad_norm": 0.2851484715938568, "learning_rate": 9.911157189705853e-05, "loss": 0.0358, "step": 10710 }, { "grad_norm": 0.3798407316207886, "learning_rate": 9.910846608442229e-05, "loss": 0.0306, "step": 10720 }, { "grad_norm": 0.22885636985301971, "learning_rate": 9.910535490135805e-05, "loss": 0.0275, "step": 10730 }, { "grad_norm": 0.21877031028270721, "learning_rate": 9.910223834820603e-05, "loss": 0.0251, "step": 10740 }, { "grad_norm": 0.2199474275112152, "learning_rate": 9.909911642530703e-05, "loss": 0.0309, "step": 10750 }, { "grad_norm": 0.27377521991729736, "learning_rate": 9.909598913300249e-05, "loss": 0.0354, "step": 10760 }, { "grad_norm": 0.3761735260486603, "learning_rate": 9.909285647163438e-05, "loss": 0.0377, "step": 10770 }, { "grad_norm": 0.3350848853588104, "learning_rate": 9.908971844154531e-05, "loss": 0.034, "step": 10780 }, { "grad_norm": 0.33925604820251465, "learning_rate": 9.908657504307843e-05, "loss": 0.0284, "step": 10790 }, { "grad_norm": 0.3224477767944336, "learning_rate": 9.908342627657751e-05, "loss": 0.035, "step": 10800 }, { "grad_norm": 0.28962457180023193, "learning_rate": 9.908027214238689e-05, "loss": 0.0272, "step": 10810 }, { "grad_norm": 0.26894411444664, "learning_rate": 9.90771126408515e-05, "loss": 0.0305, "step": 10820 }, { "grad_norm": 0.22719092667102814, "learning_rate": 9.907394777231685e-05, "loss": 0.0281, "step": 10830 }, { "grad_norm": 0.23493607342243195, "learning_rate": 9.907077753712905e-05, "loss": 0.0334, "step": 10840 }, { "grad_norm": 0.29115235805511475, "learning_rate": 9.906760193563482e-05, "loss": 0.0296, "step": 10850 }, { "grad_norm": 0.27706727385520935, "learning_rate": 9.906442096818139e-05, "loss": 0.0242, "step": 10860 }, { "grad_norm": 0.3740626871585846, "learning_rate": 9.906123463511665e-05, "loss": 0.0275, "step": 10870 }, { "grad_norm": 0.2402338683605194, "learning_rate": 9.905804293678907e-05, "loss": 0.0308, "step": 10880 }, { "grad_norm": 0.24028706550598145, "learning_rate": 9.905484587354766e-05, "loss": 0.029, "step": 10890 }, { "grad_norm": 0.24769797921180725, "learning_rate": 9.905164344574205e-05, "loss": 0.0323, "step": 10900 }, { "grad_norm": 0.2260453999042511, "learning_rate": 9.904843565372248e-05, "loss": 0.0265, "step": 10910 }, { "grad_norm": 0.35861483216285706, "learning_rate": 9.904522249783972e-05, "loss": 0.0275, "step": 10920 }, { "grad_norm": 0.3076381981372833, "learning_rate": 9.904200397844517e-05, "loss": 0.0362, "step": 10930 }, { "grad_norm": 0.3288288414478302, "learning_rate": 9.903878009589078e-05, "loss": 0.0341, "step": 10940 }, { "grad_norm": 0.32169604301452637, "learning_rate": 9.903555085052915e-05, "loss": 0.0281, "step": 10950 }, { "grad_norm": 0.2450297325849533, "learning_rate": 9.903231624271338e-05, "loss": 0.023, "step": 10960 }, { "grad_norm": 0.24047352373600006, "learning_rate": 9.902907627279724e-05, "loss": 0.0237, "step": 10970 }, { "grad_norm": 0.2699143588542938, "learning_rate": 9.902583094113504e-05, "loss": 0.0269, "step": 10980 }, { "grad_norm": 0.21769626438617706, "learning_rate": 9.902258024808168e-05, "loss": 0.0257, "step": 10990 }, { "grad_norm": 0.2964402437210083, "learning_rate": 9.901932419399264e-05, "loss": 0.0273, "step": 11000 }, { "grad_norm": 0.24767838418483734, "learning_rate": 9.9016062779224e-05, "loss": 0.0259, "step": 11010 }, { "grad_norm": 0.23193813860416412, "learning_rate": 9.901279600413242e-05, "loss": 0.028, "step": 11020 }, { "grad_norm": 0.39232364296913147, "learning_rate": 9.900952386907518e-05, "loss": 0.0277, "step": 11030 }, { "grad_norm": 0.3339059352874756, "learning_rate": 9.90062463744101e-05, "loss": 0.0254, "step": 11040 }, { "grad_norm": 0.2749948501586914, "learning_rate": 9.900296352049558e-05, "loss": 0.0239, "step": 11050 }, { "grad_norm": 0.2982577681541443, "learning_rate": 9.899967530769065e-05, "loss": 0.0334, "step": 11060 }, { "grad_norm": 0.29079845547676086, "learning_rate": 9.899638173635489e-05, "loss": 0.0243, "step": 11070 }, { "grad_norm": 0.27889564633369446, "learning_rate": 9.899308280684849e-05, "loss": 0.0271, "step": 11080 }, { "grad_norm": 0.23941174149513245, "learning_rate": 9.898977851953222e-05, "loss": 0.0261, "step": 11090 }, { "grad_norm": 0.2442619651556015, "learning_rate": 9.898646887476741e-05, "loss": 0.0306, "step": 11100 }, { "grad_norm": 0.27276718616485596, "learning_rate": 9.898315387291603e-05, "loss": 0.032, "step": 11110 }, { "grad_norm": 0.2821882963180542, "learning_rate": 9.89798335143406e-05, "loss": 0.0267, "step": 11120 }, { "grad_norm": 0.1820041686296463, "learning_rate": 9.897650779940419e-05, "loss": 0.0331, "step": 11130 }, { "grad_norm": 0.19169996678829193, "learning_rate": 9.897317672847054e-05, "loss": 0.0254, "step": 11140 }, { "grad_norm": 0.21758083999156952, "learning_rate": 9.89698403019039e-05, "loss": 0.0178, "step": 11150 }, { "grad_norm": 0.27527281641960144, "learning_rate": 9.896649852006917e-05, "loss": 0.0248, "step": 11160 }, { "grad_norm": 0.2742769420146942, "learning_rate": 9.896315138333177e-05, "loss": 0.0225, "step": 11170 }, { "grad_norm": 0.2837849259376526, "learning_rate": 9.895979889205774e-05, "loss": 0.0264, "step": 11180 }, { "grad_norm": 0.2893587648868561, "learning_rate": 9.895644104661372e-05, "loss": 0.0301, "step": 11190 }, { "grad_norm": 0.27846235036849976, "learning_rate": 9.895307784736691e-05, "loss": 0.0299, "step": 11200 }, { "grad_norm": 0.25666549801826477, "learning_rate": 9.894970929468512e-05, "loss": 0.0288, "step": 11210 }, { "grad_norm": 0.22315971553325653, "learning_rate": 9.89463353889367e-05, "loss": 0.0253, "step": 11220 }, { "grad_norm": 0.3642069101333618, "learning_rate": 9.894295613049065e-05, "loss": 0.0351, "step": 11230 }, { "grad_norm": 0.29911914467811584, "learning_rate": 9.893957151971649e-05, "loss": 0.0267, "step": 11240 }, { "grad_norm": 0.32160696387290955, "learning_rate": 9.893618155698436e-05, "loss": 0.0222, "step": 11250 }, { "grad_norm": 0.2889440655708313, "learning_rate": 9.8932786242665e-05, "loss": 0.0207, "step": 11260 }, { "grad_norm": 0.24779224395751953, "learning_rate": 9.89293855771297e-05, "loss": 0.0248, "step": 11270 }, { "grad_norm": 0.33895987272262573, "learning_rate": 9.892597956075036e-05, "loss": 0.0346, "step": 11280 }, { "grad_norm": 0.30105769634246826, "learning_rate": 9.892256819389947e-05, "loss": 0.0286, "step": 11290 }, { "grad_norm": 0.30419066548347473, "learning_rate": 9.891915147695006e-05, "loss": 0.0257, "step": 11300 }, { "grad_norm": 0.30968770384788513, "learning_rate": 9.891572941027577e-05, "loss": 0.0275, "step": 11310 }, { "grad_norm": 0.2744768559932709, "learning_rate": 9.89123019942509e-05, "loss": 0.0273, "step": 11320 }, { "grad_norm": 0.25389498472213745, "learning_rate": 9.89088692292502e-05, "loss": 0.0248, "step": 11330 }, { "grad_norm": 0.24447143077850342, "learning_rate": 9.89054311156491e-05, "loss": 0.0217, "step": 11340 }, { "grad_norm": 0.26288101077079773, "learning_rate": 9.890198765382357e-05, "loss": 0.0442, "step": 11350 }, { "grad_norm": 0.2827719449996948, "learning_rate": 9.889853884415021e-05, "loss": 0.0283, "step": 11360 }, { "grad_norm": 0.22690457105636597, "learning_rate": 9.889508468700614e-05, "loss": 0.0218, "step": 11370 }, { "grad_norm": 0.2631964385509491, "learning_rate": 9.889162518276915e-05, "loss": 0.0258, "step": 11380 }, { "grad_norm": 0.28780853748321533, "learning_rate": 9.888816033181752e-05, "loss": 0.0292, "step": 11390 }, { "grad_norm": 0.24549120664596558, "learning_rate": 9.888469013453018e-05, "loss": 0.031, "step": 11400 }, { "grad_norm": 0.35106703639030457, "learning_rate": 9.888121459128663e-05, "loss": 0.0262, "step": 11410 }, { "grad_norm": 0.2660970091819763, "learning_rate": 9.887773370246693e-05, "loss": 0.0418, "step": 11420 }, { "grad_norm": 0.266541451215744, "learning_rate": 9.887424746845177e-05, "loss": 0.0289, "step": 11430 }, { "grad_norm": 0.15918764472007751, "learning_rate": 9.887075588962239e-05, "loss": 0.0277, "step": 11440 }, { "grad_norm": 0.2878464162349701, "learning_rate": 9.88672589663606e-05, "loss": 0.0286, "step": 11450 }, { "grad_norm": 0.37777185440063477, "learning_rate": 9.886375669904886e-05, "loss": 0.0371, "step": 11460 }, { "grad_norm": 0.25814419984817505, "learning_rate": 9.886024908807014e-05, "loss": 0.0246, "step": 11470 }, { "grad_norm": 0.299892783164978, "learning_rate": 9.885673613380806e-05, "loss": 0.0257, "step": 11480 }, { "grad_norm": 0.34381628036499023, "learning_rate": 9.885321783664676e-05, "loss": 0.0322, "step": 11490 }, { "grad_norm": 0.23633526265621185, "learning_rate": 9.884969419697101e-05, "loss": 0.0271, "step": 11500 }, { "grad_norm": 0.2678866684436798, "learning_rate": 9.884616521516614e-05, "loss": 0.0246, "step": 11510 }, { "grad_norm": 0.21925301849842072, "learning_rate": 9.88426308916181e-05, "loss": 0.0268, "step": 11520 }, { "grad_norm": 0.26946067810058594, "learning_rate": 9.883909122671335e-05, "loss": 0.0226, "step": 11530 }, { "grad_norm": 0.17290501296520233, "learning_rate": 9.883554622083904e-05, "loss": 0.0252, "step": 11540 }, { "grad_norm": 0.29957151412963867, "learning_rate": 9.88319958743828e-05, "loss": 0.0289, "step": 11550 }, { "grad_norm": 0.28545746207237244, "learning_rate": 9.882844018773291e-05, "loss": 0.0255, "step": 11560 }, { "grad_norm": 0.22005802392959595, "learning_rate": 9.882487916127823e-05, "loss": 0.0235, "step": 11570 }, { "grad_norm": 0.33401864767074585, "learning_rate": 9.882131279540815e-05, "loss": 0.0312, "step": 11580 }, { "grad_norm": 0.27221083641052246, "learning_rate": 9.881774109051271e-05, "loss": 0.0303, "step": 11590 }, { "grad_norm": 0.22274422645568848, "learning_rate": 9.881416404698252e-05, "loss": 0.0246, "step": 11600 }, { "grad_norm": 0.2991376519203186, "learning_rate": 9.881058166520873e-05, "loss": 0.0372, "step": 11610 }, { "grad_norm": 0.4067988991737366, "learning_rate": 9.880699394558311e-05, "loss": 0.0366, "step": 11620 }, { "grad_norm": 0.40485629439353943, "learning_rate": 9.880340088849801e-05, "loss": 0.0461, "step": 11630 }, { "grad_norm": 0.2956446409225464, "learning_rate": 9.879980249434637e-05, "loss": 0.0294, "step": 11640 }, { "grad_norm": 0.38686054944992065, "learning_rate": 9.879619876352168e-05, "loss": 0.032, "step": 11650 }, { "grad_norm": 0.2811795771121979, "learning_rate": 9.879258969641809e-05, "loss": 0.03, "step": 11660 }, { "grad_norm": 0.2507357895374298, "learning_rate": 9.878897529343023e-05, "loss": 0.023, "step": 11670 }, { "grad_norm": 0.22255492210388184, "learning_rate": 9.878535555495338e-05, "loss": 0.0287, "step": 11680 }, { "grad_norm": 0.3800530433654785, "learning_rate": 9.87817304813834e-05, "loss": 0.0265, "step": 11690 }, { "grad_norm": 0.2292756885290146, "learning_rate": 9.877810007311671e-05, "loss": 0.0269, "step": 11700 }, { "grad_norm": 0.20625123381614685, "learning_rate": 9.877446433055035e-05, "loss": 0.0266, "step": 11710 }, { "grad_norm": 0.2960132956504822, "learning_rate": 9.877082325408191e-05, "loss": 0.0267, "step": 11720 }, { "grad_norm": 0.21879413723945618, "learning_rate": 9.876717684410954e-05, "loss": 0.0237, "step": 11730 }, { "grad_norm": 0.2695523500442505, "learning_rate": 9.876352510103204e-05, "loss": 0.0242, "step": 11740 }, { "grad_norm": 0.24657005071640015, "learning_rate": 9.875986802524875e-05, "loss": 0.0333, "step": 11750 }, { "grad_norm": 0.274624764919281, "learning_rate": 9.87562056171596e-05, "loss": 0.0249, "step": 11760 }, { "grad_norm": 0.19260740280151367, "learning_rate": 9.875253787716511e-05, "loss": 0.0299, "step": 11770 }, { "grad_norm": 0.3125512897968292, "learning_rate": 9.874886480566637e-05, "loss": 0.0258, "step": 11780 }, { "grad_norm": 0.2353973686695099, "learning_rate": 9.874518640306507e-05, "loss": 0.0248, "step": 11790 }, { "grad_norm": 0.20015884935855865, "learning_rate": 9.874150266976347e-05, "loss": 0.0284, "step": 11800 }, { "grad_norm": 0.24734096229076385, "learning_rate": 9.873781360616443e-05, "loss": 0.0339, "step": 11810 }, { "grad_norm": 0.2784419059753418, "learning_rate": 9.873411921267137e-05, "loss": 0.0277, "step": 11820 }, { "grad_norm": 0.385623037815094, "learning_rate": 9.873041948968829e-05, "loss": 0.0383, "step": 11830 }, { "grad_norm": 0.3749149739742279, "learning_rate": 9.872671443761981e-05, "loss": 0.0255, "step": 11840 }, { "grad_norm": 0.24687236547470093, "learning_rate": 9.872300405687109e-05, "loss": 0.0278, "step": 11850 }, { "grad_norm": 0.3549979627132416, "learning_rate": 9.871928834784792e-05, "loss": 0.0306, "step": 11860 }, { "grad_norm": 0.22952622175216675, "learning_rate": 9.871556731095661e-05, "loss": 0.0258, "step": 11870 }, { "grad_norm": 0.2073226422071457, "learning_rate": 9.871184094660411e-05, "loss": 0.0393, "step": 11880 }, { "grad_norm": 0.2066120058298111, "learning_rate": 9.870810925519791e-05, "loss": 0.0232, "step": 11890 }, { "grad_norm": 0.25926247239112854, "learning_rate": 9.870437223714612e-05, "loss": 0.0261, "step": 11900 }, { "grad_norm": 0.3511803150177002, "learning_rate": 9.87006298928574e-05, "loss": 0.0278, "step": 11910 }, { "grad_norm": 0.2897498607635498, "learning_rate": 9.869688222274103e-05, "loss": 0.0232, "step": 11920 }, { "grad_norm": 0.24574415385723114, "learning_rate": 9.869312922720681e-05, "loss": 0.0287, "step": 11930 }, { "grad_norm": 0.1912657618522644, "learning_rate": 9.868937090666521e-05, "loss": 0.0255, "step": 11940 }, { "grad_norm": 0.21197649836540222, "learning_rate": 9.86856072615272e-05, "loss": 0.0194, "step": 11950 }, { "grad_norm": 0.30479663610458374, "learning_rate": 9.868183829220438e-05, "loss": 0.0225, "step": 11960 }, { "grad_norm": 0.29264917969703674, "learning_rate": 9.867806399910893e-05, "loss": 0.0248, "step": 11970 }, { "grad_norm": 0.32415151596069336, "learning_rate": 9.867428438265356e-05, "loss": 0.0264, "step": 11980 }, { "grad_norm": 0.33056968450546265, "learning_rate": 9.867049944325165e-05, "loss": 0.0281, "step": 11990 }, { "grad_norm": 0.34083229303359985, "learning_rate": 9.86667091813171e-05, "loss": 0.0355, "step": 12000 }, { "grad_norm": 0.2865141034126282, "learning_rate": 9.866291359726438e-05, "loss": 0.0279, "step": 12010 }, { "grad_norm": 0.339272677898407, "learning_rate": 9.865911269150861e-05, "loss": 0.0255, "step": 12020 }, { "grad_norm": 0.21350696682929993, "learning_rate": 9.865530646446544e-05, "loss": 0.0236, "step": 12030 }, { "grad_norm": 0.22039756178855896, "learning_rate": 9.86514949165511e-05, "loss": 0.0278, "step": 12040 }, { "grad_norm": 0.2141934037208557, "learning_rate": 9.864767804818243e-05, "loss": 0.0302, "step": 12050 }, { "grad_norm": 0.20844842493534088, "learning_rate": 9.86438558597768e-05, "loss": 0.0281, "step": 12060 }, { "grad_norm": 0.2950209081172943, "learning_rate": 9.864002835175225e-05, "loss": 0.0419, "step": 12070 }, { "grad_norm": 0.3114580512046814, "learning_rate": 9.863619552452734e-05, "loss": 0.0281, "step": 12080 }, { "grad_norm": 0.2108130306005478, "learning_rate": 9.863235737852119e-05, "loss": 0.0227, "step": 12090 }, { "grad_norm": 0.269625186920166, "learning_rate": 9.862851391415356e-05, "loss": 0.0312, "step": 12100 }, { "grad_norm": 0.2549295127391815, "learning_rate": 9.862466513184477e-05, "loss": 0.0247, "step": 12110 }, { "grad_norm": 0.2893507778644562, "learning_rate": 9.86208110320157e-05, "loss": 0.0212, "step": 12120 }, { "grad_norm": 0.26036033034324646, "learning_rate": 9.861695161508784e-05, "loss": 0.0254, "step": 12130 }, { "grad_norm": 0.19087868928909302, "learning_rate": 9.861308688148324e-05, "loss": 0.0221, "step": 12140 }, { "grad_norm": 0.19404517114162445, "learning_rate": 9.860921683162455e-05, "loss": 0.0237, "step": 12150 }, { "grad_norm": 0.2384820431470871, "learning_rate": 9.860534146593499e-05, "loss": 0.0204, "step": 12160 }, { "grad_norm": 0.21464090049266815, "learning_rate": 9.860146078483836e-05, "loss": 0.022, "step": 12170 }, { "grad_norm": 0.3612446188926697, "learning_rate": 9.859757478875905e-05, "loss": 0.0321, "step": 12180 }, { "grad_norm": 0.26707518100738525, "learning_rate": 9.859368347812204e-05, "loss": 0.0233, "step": 12190 }, { "grad_norm": 0.27291247248649597, "learning_rate": 9.858978685335285e-05, "loss": 0.0344, "step": 12200 }, { "grad_norm": 0.23887403309345245, "learning_rate": 9.858588491487763e-05, "loss": 0.0447, "step": 12210 }, { "grad_norm": 0.2571420669555664, "learning_rate": 9.858197766312308e-05, "loss": 0.0288, "step": 12220 }, { "grad_norm": 0.25641965866088867, "learning_rate": 9.857806509851649e-05, "loss": 0.03, "step": 12230 }, { "grad_norm": 0.3843052089214325, "learning_rate": 9.857414722148574e-05, "loss": 0.0257, "step": 12240 }, { "grad_norm": 0.26081156730651855, "learning_rate": 9.857022403245928e-05, "loss": 0.0346, "step": 12250 }, { "grad_norm": 0.3265989124774933, "learning_rate": 9.856629553186615e-05, "loss": 0.0366, "step": 12260 }, { "grad_norm": 0.22292488813400269, "learning_rate": 9.856236172013595e-05, "loss": 0.0286, "step": 12270 }, { "grad_norm": 0.2600381076335907, "learning_rate": 9.85584225976989e-05, "loss": 0.0258, "step": 12280 }, { "grad_norm": 0.24602670967578888, "learning_rate": 9.855447816498575e-05, "loss": 0.0289, "step": 12290 }, { "grad_norm": 0.1873142123222351, "learning_rate": 9.855052842242787e-05, "loss": 0.0204, "step": 12300 }, { "grad_norm": 0.22917529940605164, "learning_rate": 9.85465733704572e-05, "loss": 0.0243, "step": 12310 }, { "grad_norm": 0.36511433124542236, "learning_rate": 9.854261300950624e-05, "loss": 0.0311, "step": 12320 }, { "grad_norm": 0.19687813520431519, "learning_rate": 9.853864734000813e-05, "loss": 0.023, "step": 12330 }, { "grad_norm": 0.2097187042236328, "learning_rate": 9.85346763623965e-05, "loss": 0.0282, "step": 12340 }, { "grad_norm": 0.23862911760807037, "learning_rate": 9.853070007710564e-05, "loss": 0.0264, "step": 12350 }, { "grad_norm": 0.27330872416496277, "learning_rate": 9.85267184845704e-05, "loss": 0.0286, "step": 12360 }, { "grad_norm": 0.22480913996696472, "learning_rate": 9.852273158522616e-05, "loss": 0.0234, "step": 12370 }, { "grad_norm": 0.1904180347919464, "learning_rate": 9.851873937950896e-05, "loss": 0.025, "step": 12380 }, { "grad_norm": 0.22927452623844147, "learning_rate": 9.851474186785537e-05, "loss": 0.0264, "step": 12390 }, { "grad_norm": 0.25986993312835693, "learning_rate": 9.851073905070254e-05, "loss": 0.0294, "step": 12400 }, { "grad_norm": 0.3687087595462799, "learning_rate": 9.850673092848824e-05, "loss": 0.0316, "step": 12410 }, { "grad_norm": 0.2417517602443695, "learning_rate": 9.850271750165077e-05, "loss": 0.0219, "step": 12420 }, { "grad_norm": 0.33432257175445557, "learning_rate": 9.849869877062902e-05, "loss": 0.0368, "step": 12430 }, { "grad_norm": 0.30698394775390625, "learning_rate": 9.849467473586252e-05, "loss": 0.0336, "step": 12440 }, { "grad_norm": 0.29726481437683105, "learning_rate": 9.849064539779127e-05, "loss": 0.0261, "step": 12450 }, { "grad_norm": 0.2071291208267212, "learning_rate": 9.848661075685594e-05, "loss": 0.0302, "step": 12460 }, { "grad_norm": 0.19749793410301208, "learning_rate": 9.848257081349778e-05, "loss": 0.0248, "step": 12470 }, { "grad_norm": 0.22446005046367645, "learning_rate": 9.847852556815856e-05, "loss": 0.0216, "step": 12480 }, { "grad_norm": 0.23430003225803375, "learning_rate": 9.847447502128067e-05, "loss": 0.0231, "step": 12490 }, { "grad_norm": 0.22906410694122314, "learning_rate": 9.847041917330708e-05, "loss": 0.0276, "step": 12500 }, { "grad_norm": 0.3050019145011902, "learning_rate": 9.846635802468132e-05, "loss": 0.0363, "step": 12510 }, { "grad_norm": 0.2148897647857666, "learning_rate": 9.84622915758475e-05, "loss": 0.0287, "step": 12520 }, { "grad_norm": 0.28988438844680786, "learning_rate": 9.845821982725034e-05, "loss": 0.0289, "step": 12530 }, { "grad_norm": 0.28970327973365784, "learning_rate": 9.845414277933514e-05, "loss": 0.0339, "step": 12540 }, { "grad_norm": 0.3595861792564392, "learning_rate": 9.845006043254771e-05, "loss": 0.0286, "step": 12550 }, { "grad_norm": 0.2788763642311096, "learning_rate": 9.844597278733451e-05, "loss": 0.0251, "step": 12560 }, { "grad_norm": 0.20965830981731415, "learning_rate": 9.844187984414259e-05, "loss": 0.0246, "step": 12570 }, { "grad_norm": 0.2124074548482895, "learning_rate": 9.84377816034195e-05, "loss": 0.0241, "step": 12580 }, { "grad_norm": 0.21031206846237183, "learning_rate": 9.843367806561345e-05, "loss": 0.0307, "step": 12590 }, { "grad_norm": 0.2963016629219055, "learning_rate": 9.842956923117317e-05, "loss": 0.0284, "step": 12600 }, { "grad_norm": 0.26586923003196716, "learning_rate": 9.842545510054802e-05, "loss": 0.022, "step": 12610 }, { "grad_norm": 0.2456645667552948, "learning_rate": 9.842133567418792e-05, "loss": 0.0243, "step": 12620 }, { "grad_norm": 0.2597130239009857, "learning_rate": 9.841721095254333e-05, "loss": 0.0206, "step": 12630 }, { "grad_norm": 0.3086458146572113, "learning_rate": 9.841308093606537e-05, "loss": 0.0329, "step": 12640 }, { "grad_norm": 0.2504803538322449, "learning_rate": 9.840894562520565e-05, "loss": 0.0244, "step": 12650 }, { "grad_norm": 0.31069666147232056, "learning_rate": 9.840480502041642e-05, "loss": 0.0288, "step": 12660 }, { "grad_norm": 0.2334115207195282, "learning_rate": 9.840065912215049e-05, "loss": 0.0241, "step": 12670 }, { "grad_norm": 0.18536770343780518, "learning_rate": 9.839650793086124e-05, "loss": 0.0307, "step": 12680 }, { "grad_norm": 0.2622363567352295, "learning_rate": 9.839235144700265e-05, "loss": 0.0281, "step": 12690 }, { "grad_norm": 0.32741430401802063, "learning_rate": 9.838818967102926e-05, "loss": 0.025, "step": 12700 }, { "grad_norm": 0.242533877491951, "learning_rate": 9.83840226033962e-05, "loss": 0.0237, "step": 12710 }, { "grad_norm": 0.252637654542923, "learning_rate": 9.837985024455918e-05, "loss": 0.0244, "step": 12720 }, { "grad_norm": 0.2900477647781372, "learning_rate": 9.837567259497447e-05, "loss": 0.0229, "step": 12730 }, { "grad_norm": 0.25663888454437256, "learning_rate": 9.837148965509894e-05, "loss": 0.0282, "step": 12740 }, { "grad_norm": 0.2361534833908081, "learning_rate": 9.836730142539001e-05, "loss": 0.0191, "step": 12750 }, { "grad_norm": 0.29983681440353394, "learning_rate": 9.836310790630574e-05, "loss": 0.0309, "step": 12760 }, { "grad_norm": 0.2550269663333893, "learning_rate": 9.83589090983047e-05, "loss": 0.0313, "step": 12770 }, { "grad_norm": 0.3556704819202423, "learning_rate": 9.835470500184605e-05, "loss": 0.0294, "step": 12780 }, { "grad_norm": 0.2493724822998047, "learning_rate": 9.835049561738957e-05, "loss": 0.03, "step": 12790 }, { "grad_norm": 0.27516505122184753, "learning_rate": 9.834628094539558e-05, "loss": 0.0258, "step": 12800 }, { "grad_norm": 0.2957218885421753, "learning_rate": 9.834206098632499e-05, "loss": 0.0272, "step": 12810 }, { "grad_norm": 0.19588826596736908, "learning_rate": 9.833783574063931e-05, "loss": 0.0267, "step": 12820 }, { "grad_norm": 0.24853573739528656, "learning_rate": 9.833360520880058e-05, "loss": 0.0266, "step": 12830 }, { "grad_norm": 0.16596435010433197, "learning_rate": 9.832936939127144e-05, "loss": 0.0231, "step": 12840 }, { "grad_norm": 0.2897602617740631, "learning_rate": 9.832512828851515e-05, "loss": 0.0228, "step": 12850 }, { "grad_norm": 0.25227171182632446, "learning_rate": 9.832088190099546e-05, "loss": 0.0276, "step": 12860 }, { "grad_norm": 0.3438115417957306, "learning_rate": 9.831663022917679e-05, "loss": 0.0227, "step": 12870 }, { "grad_norm": 0.23413513600826263, "learning_rate": 9.831237327352407e-05, "loss": 0.0192, "step": 12880 }, { "grad_norm": 0.28450557589530945, "learning_rate": 9.830811103450286e-05, "loss": 0.0247, "step": 12890 }, { "grad_norm": 0.2086557000875473, "learning_rate": 9.830384351257924e-05, "loss": 0.0319, "step": 12900 }, { "grad_norm": 0.2476930022239685, "learning_rate": 9.829957070821993e-05, "loss": 0.0243, "step": 12910 }, { "grad_norm": 0.23696641623973846, "learning_rate": 9.829529262189218e-05, "loss": 0.0176, "step": 12920 }, { "grad_norm": 0.15594618022441864, "learning_rate": 9.829100925406385e-05, "loss": 0.0217, "step": 12930 }, { "grad_norm": 0.2375161200761795, "learning_rate": 9.828672060520333e-05, "loss": 0.0221, "step": 12940 }, { "grad_norm": 0.21297447383403778, "learning_rate": 9.828242667577966e-05, "loss": 0.018, "step": 12950 }, { "grad_norm": 0.20585991442203522, "learning_rate": 9.82781274662624e-05, "loss": 0.0195, "step": 12960 }, { "grad_norm": 0.25898638367652893, "learning_rate": 9.82738229771217e-05, "loss": 0.0206, "step": 12970 }, { "grad_norm": 0.20656999945640564, "learning_rate": 9.826951320882829e-05, "loss": 0.0176, "step": 12980 }, { "grad_norm": 0.24396593868732452, "learning_rate": 9.826519816185351e-05, "loss": 0.0213, "step": 12990 }, { "grad_norm": 0.25072425603866577, "learning_rate": 9.826087783666921e-05, "loss": 0.0238, "step": 13000 }, { "grad_norm": 0.26025325059890747, "learning_rate": 9.825655223374787e-05, "loss": 0.041, "step": 13010 }, { "grad_norm": 0.3751235902309418, "learning_rate": 9.825222135356253e-05, "loss": 0.0334, "step": 13020 }, { "grad_norm": 0.26793426275253296, "learning_rate": 9.82478851965868e-05, "loss": 0.0326, "step": 13030 }, { "grad_norm": 0.22703155875205994, "learning_rate": 9.82435437632949e-05, "loss": 0.0226, "step": 13040 }, { "grad_norm": 0.2131863534450531, "learning_rate": 9.823919705416158e-05, "loss": 0.0197, "step": 13050 }, { "grad_norm": 0.2669881582260132, "learning_rate": 9.82348450696622e-05, "loss": 0.0239, "step": 13060 }, { "grad_norm": 0.30083417892456055, "learning_rate": 9.823048781027268e-05, "loss": 0.0343, "step": 13070 }, { "grad_norm": 0.2917356491088867, "learning_rate": 9.822612527646953e-05, "loss": 0.024, "step": 13080 }, { "grad_norm": 0.2535250782966614, "learning_rate": 9.822175746872984e-05, "loss": 0.0255, "step": 13090 }, { "grad_norm": 0.17477960884571075, "learning_rate": 9.821738438753123e-05, "loss": 0.0257, "step": 13100 }, { "grad_norm": 0.18886259198188782, "learning_rate": 9.821300603335196e-05, "loss": 0.0188, "step": 13110 }, { "grad_norm": 0.2859015166759491, "learning_rate": 9.820862240667085e-05, "loss": 0.0197, "step": 13120 }, { "grad_norm": 0.2358391284942627, "learning_rate": 9.820423350796726e-05, "loss": 0.0291, "step": 13130 }, { "grad_norm": 0.31972062587738037, "learning_rate": 9.819983933772118e-05, "loss": 0.0265, "step": 13140 }, { "grad_norm": 0.38902920484542847, "learning_rate": 9.819543989641314e-05, "loss": 0.0267, "step": 13150 }, { "grad_norm": 0.1704118549823761, "learning_rate": 9.819103518452423e-05, "loss": 0.0182, "step": 13160 }, { "grad_norm": 0.2720170319080353, "learning_rate": 9.818662520253618e-05, "loss": 0.0227, "step": 13170 }, { "grad_norm": 0.17454299330711365, "learning_rate": 9.818220995093126e-05, "loss": 0.0242, "step": 13180 }, { "grad_norm": 0.2053280919790268, "learning_rate": 9.817778943019228e-05, "loss": 0.0185, "step": 13190 }, { "grad_norm": 0.26404112577438354, "learning_rate": 9.81733636408027e-05, "loss": 0.0252, "step": 13200 }, { "grad_norm": 0.19617900252342224, "learning_rate": 9.816893258324649e-05, "loss": 0.0171, "step": 13210 }, { "grad_norm": 0.2052229344844818, "learning_rate": 9.816449625800823e-05, "loss": 0.0187, "step": 13220 }, { "grad_norm": 0.24538040161132812, "learning_rate": 9.816005466557308e-05, "loss": 0.0232, "step": 13230 }, { "grad_norm": 0.31794649362564087, "learning_rate": 9.815560780642674e-05, "loss": 0.0307, "step": 13240 }, { "grad_norm": 0.29564833641052246, "learning_rate": 9.815115568105555e-05, "loss": 0.0223, "step": 13250 }, { "grad_norm": 0.35970234870910645, "learning_rate": 9.814669828994638e-05, "loss": 0.0247, "step": 13260 }, { "grad_norm": 0.19708514213562012, "learning_rate": 9.814223563358665e-05, "loss": 0.0233, "step": 13270 }, { "grad_norm": 0.23976783454418182, "learning_rate": 9.813776771246443e-05, "loss": 0.0224, "step": 13280 }, { "grad_norm": 0.19448062777519226, "learning_rate": 9.813329452706829e-05, "loss": 0.0226, "step": 13290 }, { "grad_norm": 0.2510097622871399, "learning_rate": 9.812881607788744e-05, "loss": 0.022, "step": 13300 }, { "grad_norm": 0.21282926201820374, "learning_rate": 9.812433236541163e-05, "loss": 0.0259, "step": 13310 }, { "grad_norm": 0.3583712875843048, "learning_rate": 9.811984339013116e-05, "loss": 0.0296, "step": 13320 }, { "grad_norm": 0.2862602770328522, "learning_rate": 9.811534915253698e-05, "loss": 0.0255, "step": 13330 }, { "grad_norm": 0.2781870663166046, "learning_rate": 9.811084965312056e-05, "loss": 0.0246, "step": 13340 }, { "grad_norm": 0.254109263420105, "learning_rate": 9.810634489237396e-05, "loss": 0.02, "step": 13350 }, { "grad_norm": 0.23133057355880737, "learning_rate": 9.81018348707898e-05, "loss": 0.0246, "step": 13360 }, { "grad_norm": 0.24516020715236664, "learning_rate": 9.809731958886131e-05, "loss": 0.0302, "step": 13370 }, { "grad_norm": 0.32079067826271057, "learning_rate": 9.809279904708224e-05, "loss": 0.0274, "step": 13380 }, { "grad_norm": 0.2785782516002655, "learning_rate": 9.808827324594699e-05, "loss": 0.0286, "step": 13390 }, { "grad_norm": 0.3018161654472351, "learning_rate": 9.808374218595046e-05, "loss": 0.0325, "step": 13400 }, { "grad_norm": 0.27855145931243896, "learning_rate": 9.80792058675882e-05, "loss": 0.0257, "step": 13410 }, { "grad_norm": 0.1936137080192566, "learning_rate": 9.807466429135627e-05, "loss": 0.0282, "step": 13420 }, { "grad_norm": 0.2179400622844696, "learning_rate": 9.807011745775132e-05, "loss": 0.0235, "step": 13430 }, { "grad_norm": 0.2516898512840271, "learning_rate": 9.806556536727061e-05, "loss": 0.022, "step": 13440 }, { "grad_norm": 0.29254236817359924, "learning_rate": 9.806100802041193e-05, "loss": 0.0297, "step": 13450 }, { "grad_norm": 0.2515927255153656, "learning_rate": 9.805644541767368e-05, "loss": 0.0352, "step": 13460 }, { "grad_norm": 0.18077228963375092, "learning_rate": 9.805187755955478e-05, "loss": 0.0393, "step": 13470 }, { "grad_norm": 0.22365079820156097, "learning_rate": 9.804730444655483e-05, "loss": 0.0316, "step": 13480 }, { "grad_norm": 0.20927375555038452, "learning_rate": 9.804272607917388e-05, "loss": 0.0299, "step": 13490 }, { "grad_norm": 0.22803808748722076, "learning_rate": 9.803814245791265e-05, "loss": 0.0183, "step": 13500 }, { "grad_norm": 0.3166435956954956, "learning_rate": 9.803355358327239e-05, "loss": 0.0307, "step": 13510 }, { "grad_norm": 0.21237944066524506, "learning_rate": 9.802895945575492e-05, "loss": 0.0316, "step": 13520 }, { "grad_norm": 0.18277901411056519, "learning_rate": 9.802436007586266e-05, "loss": 0.0327, "step": 13530 }, { "grad_norm": 0.23254330456256866, "learning_rate": 9.801975544409858e-05, "loss": 0.0301, "step": 13540 }, { "grad_norm": 0.21587525308132172, "learning_rate": 9.801514556096625e-05, "loss": 0.0225, "step": 13550 }, { "grad_norm": 0.2524784207344055, "learning_rate": 9.801053042696977e-05, "loss": 0.029, "step": 13560 }, { "grad_norm": 0.19743184745311737, "learning_rate": 9.800591004261388e-05, "loss": 0.0221, "step": 13570 }, { "grad_norm": 0.2690284252166748, "learning_rate": 9.800128440840385e-05, "loss": 0.0252, "step": 13580 }, { "grad_norm": 0.31175169348716736, "learning_rate": 9.799665352484552e-05, "loss": 0.0326, "step": 13590 }, { "grad_norm": 0.2643221914768219, "learning_rate": 9.799201739244532e-05, "loss": 0.0387, "step": 13600 }, { "grad_norm": 0.16185273230075836, "learning_rate": 9.798737601171025e-05, "loss": 0.0204, "step": 13610 }, { "grad_norm": 0.3384391963481903, "learning_rate": 9.79827293831479e-05, "loss": 0.0342, "step": 13620 }, { "grad_norm": 0.30835267901420593, "learning_rate": 9.797807750726638e-05, "loss": 0.0247, "step": 13630 }, { "grad_norm": 0.3381163775920868, "learning_rate": 9.797342038457446e-05, "loss": 0.0249, "step": 13640 }, { "grad_norm": 0.26115116477012634, "learning_rate": 9.796875801558141e-05, "loss": 0.0296, "step": 13650 }, { "grad_norm": 0.3198961019515991, "learning_rate": 9.79640904007971e-05, "loss": 0.0292, "step": 13660 }, { "grad_norm": 0.2786710560321808, "learning_rate": 9.795941754073199e-05, "loss": 0.0244, "step": 13670 }, { "grad_norm": 0.22767536342144012, "learning_rate": 9.795473943589705e-05, "loss": 0.0245, "step": 13680 }, { "grad_norm": 0.3202740252017975, "learning_rate": 9.795005608680394e-05, "loss": 0.0252, "step": 13690 }, { "grad_norm": 0.26886022090911865, "learning_rate": 9.794536749396477e-05, "loss": 0.0253, "step": 13700 }, { "grad_norm": 0.2580215036869049, "learning_rate": 9.79406736578923e-05, "loss": 0.0225, "step": 13710 }, { "grad_norm": 0.21289482712745667, "learning_rate": 9.793597457909984e-05, "loss": 0.0204, "step": 13720 }, { "grad_norm": 0.2761736512184143, "learning_rate": 9.793127025810127e-05, "loss": 0.021, "step": 13730 }, { "grad_norm": 0.15942947566509247, "learning_rate": 9.792656069541104e-05, "loss": 0.0194, "step": 13740 }, { "grad_norm": 0.14203186333179474, "learning_rate": 9.79218458915442e-05, "loss": 0.0227, "step": 13750 }, { "grad_norm": 0.3021998107433319, "learning_rate": 9.791712584701634e-05, "loss": 0.0395, "step": 13760 }, { "grad_norm": 0.21311956644058228, "learning_rate": 9.791240056234364e-05, "loss": 0.0251, "step": 13770 }, { "grad_norm": 0.23334667086601257, "learning_rate": 9.790767003804283e-05, "loss": 0.0235, "step": 13780 }, { "grad_norm": 0.29902034997940063, "learning_rate": 9.790293427463126e-05, "loss": 0.0248, "step": 13790 }, { "grad_norm": 0.23658190667629242, "learning_rate": 9.789819327262684e-05, "loss": 0.0233, "step": 13800 }, { "grad_norm": 0.2698666453361511, "learning_rate": 9.7893447032548e-05, "loss": 0.0309, "step": 13810 }, { "grad_norm": 0.22998982667922974, "learning_rate": 9.78886955549138e-05, "loss": 0.0263, "step": 13820 }, { "grad_norm": 0.2534676790237427, "learning_rate": 9.788393884024387e-05, "loss": 0.0283, "step": 13830 }, { "grad_norm": 0.30033156275749207, "learning_rate": 9.787917688905836e-05, "loss": 0.0275, "step": 13840 }, { "grad_norm": 0.3194705545902252, "learning_rate": 9.787440970187807e-05, "loss": 0.0211, "step": 13850 }, { "grad_norm": 0.15051810443401337, "learning_rate": 9.786963727922429e-05, "loss": 0.0252, "step": 13860 }, { "grad_norm": 0.22888287901878357, "learning_rate": 9.786485962161897e-05, "loss": 0.0261, "step": 13870 }, { "grad_norm": 0.21395210921764374, "learning_rate": 9.786007672958455e-05, "loss": 0.0185, "step": 13880 }, { "grad_norm": 0.2762719392776489, "learning_rate": 9.78552886036441e-05, "loss": 0.0293, "step": 13890 }, { "grad_norm": 0.23881955444812775, "learning_rate": 9.785049524432124e-05, "loss": 0.0192, "step": 13900 }, { "grad_norm": 0.26843294501304626, "learning_rate": 9.784569665214016e-05, "loss": 0.0228, "step": 13910 }, { "grad_norm": 0.2624998986721039, "learning_rate": 9.784089282762563e-05, "loss": 0.0278, "step": 13920 }, { "grad_norm": 0.2802102565765381, "learning_rate": 9.7836083771303e-05, "loss": 0.0241, "step": 13930 }, { "grad_norm": 0.21469047665596008, "learning_rate": 9.783126948369817e-05, "loss": 0.0243, "step": 13940 }, { "grad_norm": 0.322576642036438, "learning_rate": 9.78264499653376e-05, "loss": 0.0231, "step": 13950 }, { "grad_norm": 0.22318632900714874, "learning_rate": 9.782162521674838e-05, "loss": 0.0241, "step": 13960 }, { "grad_norm": 0.21502836048603058, "learning_rate": 9.781679523845812e-05, "loss": 0.0421, "step": 13970 }, { "grad_norm": 0.34585991501808167, "learning_rate": 9.781196003099502e-05, "loss": 0.0309, "step": 13980 }, { "grad_norm": 0.2040969729423523, "learning_rate": 9.780711959488786e-05, "loss": 0.0261, "step": 13990 }, { "grad_norm": 0.24814753234386444, "learning_rate": 9.780227393066599e-05, "loss": 0.0246, "step": 14000 }, { "grad_norm": 0.20829272270202637, "learning_rate": 9.77974230388593e-05, "loss": 0.022, "step": 14010 }, { "grad_norm": 0.28151029348373413, "learning_rate": 9.779256691999829e-05, "loss": 0.0308, "step": 14020 }, { "grad_norm": 0.2634847164154053, "learning_rate": 9.778770557461403e-05, "loss": 0.0272, "step": 14030 }, { "grad_norm": 0.28913217782974243, "learning_rate": 9.778283900323812e-05, "loss": 0.0195, "step": 14040 }, { "grad_norm": 0.20813043415546417, "learning_rate": 9.777796720640277e-05, "loss": 0.0203, "step": 14050 }, { "grad_norm": 0.1928267627954483, "learning_rate": 9.777309018464078e-05, "loss": 0.0265, "step": 14060 }, { "grad_norm": 0.24398626387119293, "learning_rate": 9.776820793848547e-05, "loss": 0.0254, "step": 14070 }, { "grad_norm": 0.21972431242465973, "learning_rate": 9.776332046847075e-05, "loss": 0.022, "step": 14080 }, { "grad_norm": 0.1933583915233612, "learning_rate": 9.775842777513111e-05, "loss": 0.0221, "step": 14090 }, { "grad_norm": 0.2278405725955963, "learning_rate": 9.775352985900163e-05, "loss": 0.0307, "step": 14100 }, { "grad_norm": 0.25023016333580017, "learning_rate": 9.774862672061791e-05, "loss": 0.0252, "step": 14110 }, { "grad_norm": 0.2096412032842636, "learning_rate": 9.774371836051616e-05, "loss": 0.0214, "step": 14120 }, { "grad_norm": 0.284763902425766, "learning_rate": 9.773880477923315e-05, "loss": 0.0317, "step": 14130 }, { "grad_norm": 0.24286113679409027, "learning_rate": 9.773388597730623e-05, "loss": 0.0286, "step": 14140 }, { "grad_norm": 0.18496005237102509, "learning_rate": 9.77289619552733e-05, "loss": 0.022, "step": 14150 }, { "grad_norm": 0.3246272802352905, "learning_rate": 9.772403271367285e-05, "loss": 0.0367, "step": 14160 }, { "grad_norm": 0.394121915102005, "learning_rate": 9.771909825304396e-05, "loss": 0.0262, "step": 14170 }, { "grad_norm": 0.2275812029838562, "learning_rate": 9.771415857392619e-05, "loss": 0.0243, "step": 14180 }, { "grad_norm": 0.32320263981819153, "learning_rate": 9.770921367685978e-05, "loss": 0.029, "step": 14190 }, { "grad_norm": 0.2715618312358856, "learning_rate": 9.770426356238551e-05, "loss": 0.0249, "step": 14200 }, { "grad_norm": 0.23606599867343903, "learning_rate": 9.769930823104469e-05, "loss": 0.0237, "step": 14210 }, { "grad_norm": 0.24154327809810638, "learning_rate": 9.769434768337926e-05, "loss": 0.0302, "step": 14220 }, { "grad_norm": 0.2895973026752472, "learning_rate": 9.768938191993164e-05, "loss": 0.0248, "step": 14230 }, { "grad_norm": 0.23531480133533478, "learning_rate": 9.768441094124494e-05, "loss": 0.0232, "step": 14240 }, { "grad_norm": 0.18447966873645782, "learning_rate": 9.767943474786275e-05, "loss": 0.0277, "step": 14250 }, { "grad_norm": 0.1978904753923416, "learning_rate": 9.767445334032923e-05, "loss": 0.0219, "step": 14260 }, { "grad_norm": 0.32616421580314636, "learning_rate": 9.766946671918919e-05, "loss": 0.029, "step": 14270 }, { "grad_norm": 0.29779085516929626, "learning_rate": 9.766447488498796e-05, "loss": 0.0232, "step": 14280 }, { "grad_norm": 0.2674468159675598, "learning_rate": 9.765947783827139e-05, "loss": 0.0207, "step": 14290 }, { "grad_norm": 0.23698203265666962, "learning_rate": 9.765447557958599e-05, "loss": 0.0262, "step": 14300 }, { "grad_norm": 0.22822529077529907, "learning_rate": 9.764946810947879e-05, "loss": 0.0225, "step": 14310 }, { "grad_norm": 0.23792432248592377, "learning_rate": 9.764445542849738e-05, "loss": 0.0368, "step": 14320 }, { "grad_norm": 0.2502062916755676, "learning_rate": 9.763943753718998e-05, "loss": 0.0234, "step": 14330 }, { "grad_norm": 0.2052524983882904, "learning_rate": 9.76344144361053e-05, "loss": 0.0258, "step": 14340 }, { "grad_norm": 0.28477028012275696, "learning_rate": 9.762938612579269e-05, "loss": 0.0259, "step": 14350 }, { "grad_norm": 0.18730103969573975, "learning_rate": 9.762435260680202e-05, "loss": 0.0216, "step": 14360 }, { "grad_norm": 0.22301578521728516, "learning_rate": 9.761931387968373e-05, "loss": 0.026, "step": 14370 }, { "grad_norm": 0.2859238088130951, "learning_rate": 9.76142699449889e-05, "loss": 0.023, "step": 14380 }, { "grad_norm": 0.262776255607605, "learning_rate": 9.760922080326908e-05, "loss": 0.029, "step": 14390 }, { "grad_norm": 0.23070736229419708, "learning_rate": 9.760416645507644e-05, "loss": 0.025, "step": 14400 }, { "grad_norm": 0.2424560785293579, "learning_rate": 9.759910690096375e-05, "loss": 0.0218, "step": 14410 }, { "grad_norm": 0.22718724608421326, "learning_rate": 9.759404214148429e-05, "loss": 0.0212, "step": 14420 }, { "grad_norm": 0.2285201996564865, "learning_rate": 9.758897217719191e-05, "loss": 0.0229, "step": 14430 }, { "grad_norm": 0.2426779866218567, "learning_rate": 9.758389700864113e-05, "loss": 0.0208, "step": 14440 }, { "grad_norm": 0.25084611773490906, "learning_rate": 9.757881663638688e-05, "loss": 0.0242, "step": 14450 }, { "grad_norm": 0.218692809343338, "learning_rate": 9.757373106098478e-05, "loss": 0.0214, "step": 14460 }, { "grad_norm": 0.2392587512731552, "learning_rate": 9.756864028299097e-05, "loss": 0.0238, "step": 14470 }, { "grad_norm": 0.1889304667711258, "learning_rate": 9.75635443029622e-05, "loss": 0.0193, "step": 14480 }, { "grad_norm": 0.2933620810508728, "learning_rate": 9.755844312145572e-05, "loss": 0.022, "step": 14490 }, { "grad_norm": 0.1660984456539154, "learning_rate": 9.755333673902941e-05, "loss": 0.0307, "step": 14500 }, { "grad_norm": 0.2013074904680252, "learning_rate": 9.75482251562417e-05, "loss": 0.0258, "step": 14510 }, { "grad_norm": 0.2304338961839676, "learning_rate": 9.754310837365155e-05, "loss": 0.0208, "step": 14520 }, { "grad_norm": 0.24693520367145538, "learning_rate": 9.753798639181856e-05, "loss": 0.0372, "step": 14530 }, { "grad_norm": 0.2033688724040985, "learning_rate": 9.753285921130286e-05, "loss": 0.0217, "step": 14540 }, { "grad_norm": 0.24370455741882324, "learning_rate": 9.752772683266512e-05, "loss": 0.0212, "step": 14550 }, { "grad_norm": 0.20126283168792725, "learning_rate": 9.752258925646665e-05, "loss": 0.0225, "step": 14560 }, { "grad_norm": 0.20771393179893494, "learning_rate": 9.751744648326926e-05, "loss": 0.0198, "step": 14570 }, { "grad_norm": 0.20671360194683075, "learning_rate": 9.751229851363536e-05, "loss": 0.0256, "step": 14580 }, { "grad_norm": 0.2310469150543213, "learning_rate": 9.750714534812793e-05, "loss": 0.0329, "step": 14590 }, { "grad_norm": 0.16901017725467682, "learning_rate": 9.750198698731053e-05, "loss": 0.0292, "step": 14600 }, { "grad_norm": 0.1489357054233551, "learning_rate": 9.749682343174722e-05, "loss": 0.0193, "step": 14610 }, { "grad_norm": 0.21046893298625946, "learning_rate": 9.749165468200272e-05, "loss": 0.0196, "step": 14620 }, { "grad_norm": 0.20572230219841003, "learning_rate": 9.748648073864229e-05, "loss": 0.0225, "step": 14630 }, { "grad_norm": 0.30018502473831177, "learning_rate": 9.748130160223168e-05, "loss": 0.0219, "step": 14640 }, { "grad_norm": 0.2483537346124649, "learning_rate": 9.747611727333734e-05, "loss": 0.0228, "step": 14650 }, { "grad_norm": 0.21430501341819763, "learning_rate": 9.74709277525262e-05, "loss": 0.0219, "step": 14660 }, { "grad_norm": 0.32866883277893066, "learning_rate": 9.746573304036576e-05, "loss": 0.0295, "step": 14670 }, { "grad_norm": 0.29449957609176636, "learning_rate": 9.746053313742412e-05, "loss": 0.0214, "step": 14680 }, { "grad_norm": 0.2365882396697998, "learning_rate": 9.745532804426994e-05, "loss": 0.0255, "step": 14690 }, { "grad_norm": 0.20043890178203583, "learning_rate": 9.745011776147242e-05, "loss": 0.0265, "step": 14700 }, { "grad_norm": 0.23013000190258026, "learning_rate": 9.744490228960138e-05, "loss": 0.0282, "step": 14710 }, { "grad_norm": 0.15829536318778992, "learning_rate": 9.743968162922713e-05, "loss": 0.0251, "step": 14720 }, { "grad_norm": 0.22303073108196259, "learning_rate": 9.743445578092064e-05, "loss": 0.0235, "step": 14730 }, { "grad_norm": 0.19718147814273834, "learning_rate": 9.742922474525338e-05, "loss": 0.0297, "step": 14740 }, { "grad_norm": 0.19284456968307495, "learning_rate": 9.742398852279741e-05, "loss": 0.026, "step": 14750 }, { "grad_norm": 0.2267046421766281, "learning_rate": 9.741874711412535e-05, "loss": 0.0273, "step": 14760 }, { "grad_norm": 0.2957620322704315, "learning_rate": 9.741350051981042e-05, "loss": 0.024, "step": 14770 }, { "grad_norm": 0.24603286385536194, "learning_rate": 9.740824874042633e-05, "loss": 0.0232, "step": 14780 }, { "grad_norm": 0.20581181347370148, "learning_rate": 9.740299177654746e-05, "loss": 0.0189, "step": 14790 }, { "grad_norm": 0.32639703154563904, "learning_rate": 9.739772962874867e-05, "loss": 0.0233, "step": 14800 }, { "grad_norm": 0.16095663607120514, "learning_rate": 9.739246229760541e-05, "loss": 0.0197, "step": 14810 }, { "grad_norm": 0.18651285767555237, "learning_rate": 9.738718978369376e-05, "loss": 0.0238, "step": 14820 }, { "grad_norm": 0.16945907473564148, "learning_rate": 9.738191208759025e-05, "loss": 0.0307, "step": 14830 }, { "grad_norm": 0.2165047526359558, "learning_rate": 9.73766292098721e-05, "loss": 0.0351, "step": 14840 }, { "grad_norm": 0.27451950311660767, "learning_rate": 9.737134115111699e-05, "loss": 0.0409, "step": 14850 }, { "grad_norm": 0.2888193726539612, "learning_rate": 9.736604791190323e-05, "loss": 0.0229, "step": 14860 }, { "grad_norm": 0.17105482518672943, "learning_rate": 9.73607494928097e-05, "loss": 0.0202, "step": 14870 }, { "grad_norm": 0.26147183775901794, "learning_rate": 9.735544589441581e-05, "loss": 0.0254, "step": 14880 }, { "grad_norm": 0.25057804584503174, "learning_rate": 9.735013711730154e-05, "loss": 0.0219, "step": 14890 }, { "grad_norm": 0.2623242437839508, "learning_rate": 9.734482316204747e-05, "loss": 0.0201, "step": 14900 }, { "grad_norm": 0.27881839871406555, "learning_rate": 9.733950402923473e-05, "loss": 0.0237, "step": 14910 }, { "grad_norm": 0.20905229449272156, "learning_rate": 9.7334179719445e-05, "loss": 0.0211, "step": 14920 }, { "grad_norm": 0.33797556161880493, "learning_rate": 9.732885023326053e-05, "loss": 0.0204, "step": 14930 }, { "grad_norm": 0.22466064989566803, "learning_rate": 9.732351557126418e-05, "loss": 0.0309, "step": 14940 }, { "grad_norm": 0.2469981163740158, "learning_rate": 9.731817573403929e-05, "loss": 0.0373, "step": 14950 }, { "grad_norm": 0.17968247830867767, "learning_rate": 9.731283072216985e-05, "loss": 0.0296, "step": 14960 }, { "grad_norm": 0.27748265862464905, "learning_rate": 9.730748053624039e-05, "loss": 0.0252, "step": 14970 }, { "grad_norm": 0.24269601702690125, "learning_rate": 9.730212517683598e-05, "loss": 0.0265, "step": 14980 }, { "grad_norm": 0.28463828563690186, "learning_rate": 9.729676464454228e-05, "loss": 0.0246, "step": 14990 }, { "grad_norm": 0.3419332206249237, "learning_rate": 9.72913989399455e-05, "loss": 0.0276, "step": 15000 }, { "grad_norm": 0.19957666099071503, "learning_rate": 9.728602806363242e-05, "loss": 0.0193, "step": 15010 }, { "grad_norm": 0.2896881401538849, "learning_rate": 9.728065201619043e-05, "loss": 0.0209, "step": 15020 }, { "grad_norm": 0.2486845999956131, "learning_rate": 9.727527079820742e-05, "loss": 0.0207, "step": 15030 }, { "grad_norm": 0.190235897898674, "learning_rate": 9.726988441027186e-05, "loss": 0.0317, "step": 15040 }, { "grad_norm": 0.25282880663871765, "learning_rate": 9.726449285297281e-05, "loss": 0.0356, "step": 15050 }, { "grad_norm": 0.3496107757091522, "learning_rate": 9.72590961268999e-05, "loss": 0.0361, "step": 15060 }, { "grad_norm": 0.20682382583618164, "learning_rate": 9.725369423264328e-05, "loss": 0.0254, "step": 15070 }, { "grad_norm": 0.1874798685312271, "learning_rate": 9.72482871707937e-05, "loss": 0.0267, "step": 15080 }, { "grad_norm": 0.28622063994407654, "learning_rate": 9.724287494194247e-05, "loss": 0.0283, "step": 15090 }, { "grad_norm": 0.21745572984218597, "learning_rate": 9.723745754668147e-05, "loss": 0.0243, "step": 15100 }, { "grad_norm": 0.24912241101264954, "learning_rate": 9.723203498560313e-05, "loss": 0.0245, "step": 15110 }, { "grad_norm": 0.252337247133255, "learning_rate": 9.722660725930046e-05, "loss": 0.0196, "step": 15120 }, { "grad_norm": 0.2759527862071991, "learning_rate": 9.722117436836702e-05, "loss": 0.0215, "step": 15130 }, { "grad_norm": 0.2416398823261261, "learning_rate": 9.721573631339696e-05, "loss": 0.0171, "step": 15140 }, { "grad_norm": 0.2521699368953705, "learning_rate": 9.721029309498494e-05, "loss": 0.0241, "step": 15150 }, { "grad_norm": 0.1720665544271469, "learning_rate": 9.720484471372627e-05, "loss": 0.0222, "step": 15160 }, { "grad_norm": 0.25933048129081726, "learning_rate": 9.719939117021673e-05, "loss": 0.0265, "step": 15170 }, { "grad_norm": 0.2771762013435364, "learning_rate": 9.719393246505275e-05, "loss": 0.0228, "step": 15180 }, { "grad_norm": 0.19077520072460175, "learning_rate": 9.718846859883128e-05, "loss": 0.0214, "step": 15190 }, { "grad_norm": 0.291400283575058, "learning_rate": 9.718299957214982e-05, "loss": 0.0289, "step": 15200 }, { "grad_norm": 0.1779903620481491, "learning_rate": 9.717752538560646e-05, "loss": 0.0238, "step": 15210 }, { "grad_norm": 0.37550076842308044, "learning_rate": 9.717204603979986e-05, "loss": 0.0235, "step": 15220 }, { "grad_norm": 0.1764754205942154, "learning_rate": 9.716656153532922e-05, "loss": 0.0178, "step": 15230 }, { "grad_norm": 0.2595665156841278, "learning_rate": 9.716107187279434e-05, "loss": 0.0205, "step": 15240 }, { "grad_norm": 0.35259392857551575, "learning_rate": 9.715557705279555e-05, "loss": 0.0364, "step": 15250 }, { "grad_norm": 0.25226590037345886, "learning_rate": 9.715007707593372e-05, "loss": 0.027, "step": 15260 }, { "grad_norm": 0.2109677493572235, "learning_rate": 9.714457194281036e-05, "loss": 0.0215, "step": 15270 }, { "grad_norm": 0.25441399216651917, "learning_rate": 9.713906165402751e-05, "loss": 0.0223, "step": 15280 }, { "grad_norm": 0.2899956703186035, "learning_rate": 9.713354621018774e-05, "loss": 0.0295, "step": 15290 }, { "grad_norm": 0.24542950093746185, "learning_rate": 9.712802561189422e-05, "loss": 0.0253, "step": 15300 }, { "grad_norm": 0.21808695793151855, "learning_rate": 9.712249985975069e-05, "loss": 0.0314, "step": 15310 }, { "grad_norm": 0.19013932347297668, "learning_rate": 9.71169689543614e-05, "loss": 0.0196, "step": 15320 }, { "grad_norm": 0.31495845317840576, "learning_rate": 9.711143289633123e-05, "loss": 0.0208, "step": 15330 }, { "grad_norm": 0.2495282143354416, "learning_rate": 9.710589168626561e-05, "loss": 0.0239, "step": 15340 }, { "grad_norm": 0.29330822825431824, "learning_rate": 9.710034532477048e-05, "loss": 0.0281, "step": 15350 }, { "grad_norm": 0.21229594945907593, "learning_rate": 9.709479381245239e-05, "loss": 0.0218, "step": 15360 }, { "grad_norm": 0.21089982986450195, "learning_rate": 9.708923714991847e-05, "loss": 0.0243, "step": 15370 }, { "grad_norm": 0.23127295076847076, "learning_rate": 9.708367533777638e-05, "loss": 0.0236, "step": 15380 }, { "grad_norm": 0.279473215341568, "learning_rate": 9.707810837663431e-05, "loss": 0.0214, "step": 15390 }, { "grad_norm": 0.22076457738876343, "learning_rate": 9.707253626710113e-05, "loss": 0.0227, "step": 15400 }, { "grad_norm": 0.2600025236606598, "learning_rate": 9.706695900978613e-05, "loss": 0.0199, "step": 15410 }, { "grad_norm": 0.20560555160045624, "learning_rate": 9.706137660529926e-05, "loss": 0.0251, "step": 15420 }, { "grad_norm": 0.31278160214424133, "learning_rate": 9.705578905425101e-05, "loss": 0.0425, "step": 15430 }, { "grad_norm": 0.2272595316171646, "learning_rate": 9.705019635725241e-05, "loss": 0.0269, "step": 15440 }, { "grad_norm": 0.20497113466262817, "learning_rate": 9.704459851491508e-05, "loss": 0.0331, "step": 15450 }, { "grad_norm": 0.24172233045101166, "learning_rate": 9.703899552785118e-05, "loss": 0.0254, "step": 15460 }, { "grad_norm": 0.2781061828136444, "learning_rate": 9.703338739667346e-05, "loss": 0.0214, "step": 15470 }, { "grad_norm": 0.3735010623931885, "learning_rate": 9.70277741219952e-05, "loss": 0.0278, "step": 15480 }, { "grad_norm": 0.2779931128025055, "learning_rate": 9.702215570443027e-05, "loss": 0.026, "step": 15490 }, { "grad_norm": 0.1607367992401123, "learning_rate": 9.701653214459309e-05, "loss": 0.0208, "step": 15500 }, { "grad_norm": 0.2050749510526657, "learning_rate": 9.701090344309865e-05, "loss": 0.0267, "step": 15510 }, { "grad_norm": 0.21985048055648804, "learning_rate": 9.700526960056247e-05, "loss": 0.0184, "step": 15520 }, { "grad_norm": 0.24512867629528046, "learning_rate": 9.699963061760068e-05, "loss": 0.0183, "step": 15530 }, { "grad_norm": 0.2292667031288147, "learning_rate": 9.699398649482997e-05, "loss": 0.0228, "step": 15540 }, { "grad_norm": 0.21019242703914642, "learning_rate": 9.698833723286753e-05, "loss": 0.0204, "step": 15550 }, { "grad_norm": 0.25162258744239807, "learning_rate": 9.698268283233118e-05, "loss": 0.0265, "step": 15560 }, { "grad_norm": 0.3474053144454956, "learning_rate": 9.697702329383929e-05, "loss": 0.0293, "step": 15570 }, { "grad_norm": 0.22462700307369232, "learning_rate": 9.697135861801074e-05, "loss": 0.0316, "step": 15580 }, { "grad_norm": 0.2310541719198227, "learning_rate": 9.696568880546505e-05, "loss": 0.0229, "step": 15590 }, { "grad_norm": 0.19083434343338013, "learning_rate": 9.696001385682223e-05, "loss": 0.0216, "step": 15600 }, { "grad_norm": 0.23968110978603363, "learning_rate": 9.695433377270291e-05, "loss": 0.0268, "step": 15610 }, { "grad_norm": 0.25786927342414856, "learning_rate": 9.694864855372824e-05, "loss": 0.0211, "step": 15620 }, { "grad_norm": 0.2107478380203247, "learning_rate": 9.694295820051995e-05, "loss": 0.0224, "step": 15630 }, { "grad_norm": 0.21976928412914276, "learning_rate": 9.693726271370032e-05, "loss": 0.0219, "step": 15640 }, { "grad_norm": 0.20567601919174194, "learning_rate": 9.693156209389221e-05, "loss": 0.0229, "step": 15650 }, { "grad_norm": 0.31277433037757874, "learning_rate": 9.692585634171905e-05, "loss": 0.0255, "step": 15660 }, { "grad_norm": 0.22213703393936157, "learning_rate": 9.692014545780476e-05, "loss": 0.022, "step": 15670 }, { "grad_norm": 0.16178648173809052, "learning_rate": 9.691442944277393e-05, "loss": 0.0182, "step": 15680 }, { "grad_norm": 0.20713871717453003, "learning_rate": 9.690870829725162e-05, "loss": 0.0215, "step": 15690 }, { "grad_norm": 0.2959996163845062, "learning_rate": 9.69029820218635e-05, "loss": 0.0232, "step": 15700 }, { "grad_norm": 0.2267557978630066, "learning_rate": 9.689725061723579e-05, "loss": 0.0263, "step": 15710 }, { "grad_norm": 0.2522988021373749, "learning_rate": 9.689151408399527e-05, "loss": 0.0191, "step": 15720 }, { "grad_norm": 0.3069360554218292, "learning_rate": 9.688577242276924e-05, "loss": 0.0311, "step": 15730 }, { "grad_norm": 0.19688650965690613, "learning_rate": 9.688002563418566e-05, "loss": 0.0228, "step": 15740 }, { "grad_norm": 0.22525599598884583, "learning_rate": 9.687427371887293e-05, "loss": 0.0277, "step": 15750 }, { "grad_norm": 0.23955842852592468, "learning_rate": 9.686851667746012e-05, "loss": 0.0238, "step": 15760 }, { "grad_norm": 0.2884877920150757, "learning_rate": 9.686275451057677e-05, "loss": 0.0246, "step": 15770 }, { "grad_norm": 0.22962050139904022, "learning_rate": 9.685698721885308e-05, "loss": 0.023, "step": 15780 }, { "grad_norm": 0.22283923625946045, "learning_rate": 9.68512148029197e-05, "loss": 0.0286, "step": 15790 }, { "grad_norm": 0.22234287858009338, "learning_rate": 9.684543726340791e-05, "loss": 0.0264, "step": 15800 }, { "grad_norm": 0.18761509656906128, "learning_rate": 9.683965460094952e-05, "loss": 0.023, "step": 15810 }, { "grad_norm": 0.2626989483833313, "learning_rate": 9.683386681617694e-05, "loss": 0.0213, "step": 15820 }, { "grad_norm": 0.2288074493408203, "learning_rate": 9.68280739097231e-05, "loss": 0.0211, "step": 15830 }, { "grad_norm": 0.18786223232746124, "learning_rate": 9.682227588222148e-05, "loss": 0.0197, "step": 15840 }, { "grad_norm": 0.23232035338878632, "learning_rate": 9.681647273430618e-05, "loss": 0.0195, "step": 15850 }, { "grad_norm": 0.24477063119411469, "learning_rate": 9.681066446661182e-05, "loss": 0.0217, "step": 15860 }, { "grad_norm": 0.25646746158599854, "learning_rate": 9.680485107977357e-05, "loss": 0.0271, "step": 15870 }, { "grad_norm": 0.23802921175956726, "learning_rate": 9.679903257442716e-05, "loss": 0.0246, "step": 15880 }, { "grad_norm": 0.24212859570980072, "learning_rate": 9.679320895120891e-05, "loss": 0.0249, "step": 15890 }, { "grad_norm": 0.18747539818286896, "learning_rate": 9.67873802107557e-05, "loss": 0.0178, "step": 15900 }, { "grad_norm": 0.22056367993354797, "learning_rate": 9.67815463537049e-05, "loss": 0.0204, "step": 15910 }, { "grad_norm": 0.23940449953079224, "learning_rate": 9.677570738069457e-05, "loss": 0.0237, "step": 15920 }, { "grad_norm": 0.23983295261859894, "learning_rate": 9.676986329236318e-05, "loss": 0.0259, "step": 15930 }, { "grad_norm": 0.21284569799900055, "learning_rate": 9.676401408934987e-05, "loss": 0.0229, "step": 15940 }, { "grad_norm": 0.1875373125076294, "learning_rate": 9.675815977229428e-05, "loss": 0.0193, "step": 15950 }, { "grad_norm": 0.17598313093185425, "learning_rate": 9.675230034183664e-05, "loss": 0.0182, "step": 15960 }, { "grad_norm": 0.21531447768211365, "learning_rate": 9.674643579861773e-05, "loss": 0.0244, "step": 15970 }, { "grad_norm": 0.26266878843307495, "learning_rate": 9.674056614327886e-05, "loss": 0.0184, "step": 15980 }, { "grad_norm": 0.20900431275367737, "learning_rate": 9.673469137646198e-05, "loss": 0.0248, "step": 15990 }, { "grad_norm": 0.21966899931430817, "learning_rate": 9.67288114988095e-05, "loss": 0.0209, "step": 16000 }, { "grad_norm": 0.21411024034023285, "learning_rate": 9.672292651096447e-05, "loss": 0.0208, "step": 16010 }, { "grad_norm": 0.18874625861644745, "learning_rate": 9.671703641357042e-05, "loss": 0.0207, "step": 16020 }, { "grad_norm": 0.19252750277519226, "learning_rate": 9.67111412072715e-05, "loss": 0.022, "step": 16030 }, { "grad_norm": 0.22302201390266418, "learning_rate": 9.670524089271242e-05, "loss": 0.0216, "step": 16040 }, { "grad_norm": 0.2163030505180359, "learning_rate": 9.669933547053842e-05, "loss": 0.0196, "step": 16050 }, { "grad_norm": 0.22854752838611603, "learning_rate": 9.669342494139531e-05, "loss": 0.021, "step": 16060 }, { "grad_norm": 0.23484258353710175, "learning_rate": 9.668750930592943e-05, "loss": 0.0198, "step": 16070 }, { "grad_norm": 0.3292390704154968, "learning_rate": 9.668158856478775e-05, "loss": 0.021, "step": 16080 }, { "grad_norm": 0.3117564916610718, "learning_rate": 9.66756627186177e-05, "loss": 0.0228, "step": 16090 }, { "grad_norm": 0.1699584573507309, "learning_rate": 9.666973176806737e-05, "loss": 0.0239, "step": 16100 }, { "grad_norm": 0.17493626475334167, "learning_rate": 9.666379571378534e-05, "loss": 0.0227, "step": 16110 }, { "grad_norm": 0.27698999643325806, "learning_rate": 9.665785455642076e-05, "loss": 0.0288, "step": 16120 }, { "grad_norm": 0.3279920220375061, "learning_rate": 9.665190829662337e-05, "loss": 0.0232, "step": 16130 }, { "grad_norm": 0.2742263674736023, "learning_rate": 9.664595693504342e-05, "loss": 0.0199, "step": 16140 }, { "grad_norm": 0.16349871456623077, "learning_rate": 9.664000047233175e-05, "loss": 0.0266, "step": 16150 }, { "grad_norm": 0.17749814689159393, "learning_rate": 9.663403890913976e-05, "loss": 0.0241, "step": 16160 }, { "grad_norm": 0.19023063778877258, "learning_rate": 9.662807224611938e-05, "loss": 0.0216, "step": 16170 }, { "grad_norm": 0.1769041121006012, "learning_rate": 9.662210048392311e-05, "loss": 0.0183, "step": 16180 }, { "grad_norm": 0.24602967500686646, "learning_rate": 9.661612362320405e-05, "loss": 0.0224, "step": 16190 }, { "grad_norm": 0.21377643942832947, "learning_rate": 9.661014166461579e-05, "loss": 0.0228, "step": 16200 }, { "grad_norm": 0.13903872668743134, "learning_rate": 9.66041546088125e-05, "loss": 0.0297, "step": 16210 }, { "grad_norm": 0.15127508342266083, "learning_rate": 9.659816245644895e-05, "loss": 0.0196, "step": 16220 }, { "grad_norm": 0.18163186311721802, "learning_rate": 9.65921652081804e-05, "loss": 0.0264, "step": 16230 }, { "grad_norm": 0.280267596244812, "learning_rate": 9.658616286466271e-05, "loss": 0.026, "step": 16240 }, { "grad_norm": 0.3457470238208771, "learning_rate": 9.65801554265523e-05, "loss": 0.0225, "step": 16250 }, { "grad_norm": 0.25707492232322693, "learning_rate": 9.657414289450612e-05, "loss": 0.0227, "step": 16260 }, { "grad_norm": 0.2094838172197342, "learning_rate": 9.656812526918171e-05, "loss": 0.0224, "step": 16270 }, { "grad_norm": 0.20395302772521973, "learning_rate": 9.656210255123712e-05, "loss": 0.0263, "step": 16280 }, { "grad_norm": 0.16504985094070435, "learning_rate": 9.6556074741331e-05, "loss": 0.0229, "step": 16290 }, { "grad_norm": 0.16999123990535736, "learning_rate": 9.655004184012256e-05, "loss": 0.0176, "step": 16300 }, { "grad_norm": 0.2629798948764801, "learning_rate": 9.654400384827152e-05, "loss": 0.0256, "step": 16310 }, { "grad_norm": 0.28476622700691223, "learning_rate": 9.653796076643818e-05, "loss": 0.0204, "step": 16320 }, { "grad_norm": 0.20561227202415466, "learning_rate": 9.653191259528344e-05, "loss": 0.0311, "step": 16330 }, { "grad_norm": 0.2302899956703186, "learning_rate": 9.65258593354687e-05, "loss": 0.0204, "step": 16340 }, { "grad_norm": 0.20187848806381226, "learning_rate": 9.651980098765591e-05, "loss": 0.0173, "step": 16350 }, { "grad_norm": 0.22739574313163757, "learning_rate": 9.651373755250765e-05, "loss": 0.0239, "step": 16360 }, { "grad_norm": 0.3176218867301941, "learning_rate": 9.650766903068697e-05, "loss": 0.029, "step": 16370 }, { "grad_norm": 0.2017831802368164, "learning_rate": 9.650159542285753e-05, "loss": 0.0227, "step": 16380 }, { "grad_norm": 0.21310217678546906, "learning_rate": 9.649551672968353e-05, "loss": 0.029, "step": 16390 }, { "grad_norm": 0.19352436065673828, "learning_rate": 9.648943295182973e-05, "loss": 0.0234, "step": 16400 }, { "grad_norm": 0.20844149589538574, "learning_rate": 9.648334408996144e-05, "loss": 0.023, "step": 16410 }, { "grad_norm": 0.2655220329761505, "learning_rate": 9.647725014474452e-05, "loss": 0.0223, "step": 16420 }, { "grad_norm": 0.2586666941642761, "learning_rate": 9.64711511168454e-05, "loss": 0.0261, "step": 16430 }, { "grad_norm": 0.20502418279647827, "learning_rate": 9.646504700693108e-05, "loss": 0.021, "step": 16440 }, { "grad_norm": 0.260814368724823, "learning_rate": 9.645893781566907e-05, "loss": 0.021, "step": 16450 }, { "grad_norm": 0.23046842217445374, "learning_rate": 9.645282354372744e-05, "loss": 0.023, "step": 16460 }, { "grad_norm": 0.19517722725868225, "learning_rate": 9.644670419177491e-05, "loss": 0.0199, "step": 16470 }, { "grad_norm": 0.2570740580558777, "learning_rate": 9.644057976048062e-05, "loss": 0.0301, "step": 16480 }, { "grad_norm": 0.20949184894561768, "learning_rate": 9.643445025051435e-05, "loss": 0.026, "step": 16490 }, { "grad_norm": 0.28931403160095215, "learning_rate": 9.642831566254641e-05, "loss": 0.0217, "step": 16500 }, { "grad_norm": 0.25614166259765625, "learning_rate": 9.642217599724769e-05, "loss": 0.0215, "step": 16510 }, { "grad_norm": 0.1920831948518753, "learning_rate": 9.64160312552896e-05, "loss": 0.0237, "step": 16520 }, { "grad_norm": 0.27031004428863525, "learning_rate": 9.64098814373441e-05, "loss": 0.0231, "step": 16530 }, { "grad_norm": 0.33867326378822327, "learning_rate": 9.640372654408374e-05, "loss": 0.0216, "step": 16540 }, { "grad_norm": 0.20583371818065643, "learning_rate": 9.639756657618162e-05, "loss": 0.023, "step": 16550 }, { "grad_norm": 0.3442153036594391, "learning_rate": 9.639140153431138e-05, "loss": 0.0264, "step": 16560 }, { "grad_norm": 0.2381899505853653, "learning_rate": 9.638523141914721e-05, "loss": 0.0236, "step": 16570 }, { "grad_norm": 0.21320347487926483, "learning_rate": 9.637905623136388e-05, "loss": 0.0257, "step": 16580 }, { "grad_norm": 0.19654503464698792, "learning_rate": 9.637287597163669e-05, "loss": 0.018, "step": 16590 }, { "grad_norm": 0.1390400230884552, "learning_rate": 9.63666906406415e-05, "loss": 0.0182, "step": 16600 }, { "grad_norm": 0.2361716330051422, "learning_rate": 9.636050023905473e-05, "loss": 0.0156, "step": 16610 }, { "grad_norm": 0.2000289261341095, "learning_rate": 9.635430476755336e-05, "loss": 0.0161, "step": 16620 }, { "grad_norm": 0.23434315621852875, "learning_rate": 9.63481042268149e-05, "loss": 0.0167, "step": 16630 }, { "grad_norm": 0.3153533637523651, "learning_rate": 9.634189861751745e-05, "loss": 0.0248, "step": 16640 }, { "grad_norm": 0.2719980776309967, "learning_rate": 9.633568794033967e-05, "loss": 0.0235, "step": 16650 }, { "grad_norm": 0.2225322425365448, "learning_rate": 9.63294721959607e-05, "loss": 0.0214, "step": 16660 }, { "grad_norm": 0.22863894701004028, "learning_rate": 9.63232513850603e-05, "loss": 0.0334, "step": 16670 }, { "grad_norm": 0.2389281839132309, "learning_rate": 9.631702550831878e-05, "loss": 0.0231, "step": 16680 }, { "grad_norm": 0.3893519341945648, "learning_rate": 9.631079456641698e-05, "loss": 0.0249, "step": 16690 }, { "grad_norm": 0.2640669047832489, "learning_rate": 9.630455856003632e-05, "loss": 0.0215, "step": 16700 }, { "grad_norm": 0.2296333909034729, "learning_rate": 9.629831748985876e-05, "loss": 0.0244, "step": 16710 }, { "grad_norm": 0.22797846794128418, "learning_rate": 9.629207135656679e-05, "loss": 0.0357, "step": 16720 }, { "grad_norm": 0.22505129873752594, "learning_rate": 9.628582016084353e-05, "loss": 0.0313, "step": 16730 }, { "grad_norm": 0.2623140215873718, "learning_rate": 9.627956390337254e-05, "loss": 0.0213, "step": 16740 }, { "grad_norm": 0.27261680364608765, "learning_rate": 9.627330258483802e-05, "loss": 0.0303, "step": 16750 }, { "grad_norm": 0.3099120557308197, "learning_rate": 9.62670362059247e-05, "loss": 0.0216, "step": 16760 }, { "grad_norm": 0.2017795443534851, "learning_rate": 9.626076476731786e-05, "loss": 0.0213, "step": 16770 }, { "grad_norm": 0.28539344668388367, "learning_rate": 9.625448826970336e-05, "loss": 0.0226, "step": 16780 }, { "grad_norm": 0.1640094518661499, "learning_rate": 9.624820671376755e-05, "loss": 0.0194, "step": 16790 }, { "grad_norm": 0.3264823853969574, "learning_rate": 9.62419201001974e-05, "loss": 0.0236, "step": 16800 }, { "grad_norm": 0.3045097291469574, "learning_rate": 9.623562842968037e-05, "loss": 0.02, "step": 16810 }, { "grad_norm": 0.25485700368881226, "learning_rate": 9.622933170290454e-05, "loss": 0.019, "step": 16820 }, { "grad_norm": 0.23218987882137299, "learning_rate": 9.622302992055849e-05, "loss": 0.0198, "step": 16830 }, { "grad_norm": 0.21516571938991547, "learning_rate": 9.62167230833314e-05, "loss": 0.0171, "step": 16840 }, { "grad_norm": 0.20393338799476624, "learning_rate": 9.621041119191295e-05, "loss": 0.0199, "step": 16850 }, { "grad_norm": 0.20222893357276917, "learning_rate": 9.620409424699342e-05, "loss": 0.0258, "step": 16860 }, { "grad_norm": 0.29511675238609314, "learning_rate": 9.619777224926359e-05, "loss": 0.0195, "step": 16870 }, { "grad_norm": 0.21260157227516174, "learning_rate": 9.619144519941485e-05, "loss": 0.0206, "step": 16880 }, { "grad_norm": 0.20829927921295166, "learning_rate": 9.618511309813912e-05, "loss": 0.0223, "step": 16890 }, { "grad_norm": 0.19520996510982513, "learning_rate": 9.617877594612886e-05, "loss": 0.0181, "step": 16900 }, { "grad_norm": 0.19796887040138245, "learning_rate": 9.617243374407707e-05, "loss": 0.0171, "step": 16910 }, { "grad_norm": 0.20133674144744873, "learning_rate": 9.616608649267736e-05, "loss": 0.0211, "step": 16920 }, { "grad_norm": 0.2608165144920349, "learning_rate": 9.615973419262385e-05, "loss": 0.0214, "step": 16930 }, { "grad_norm": 0.14789970219135284, "learning_rate": 9.615337684461119e-05, "loss": 0.0184, "step": 16940 }, { "grad_norm": 0.263294517993927, "learning_rate": 9.614701444933465e-05, "loss": 0.0282, "step": 16950 }, { "grad_norm": 0.22893117368221283, "learning_rate": 9.614064700748997e-05, "loss": 0.0225, "step": 16960 }, { "grad_norm": 0.23177549242973328, "learning_rate": 9.613427451977352e-05, "loss": 0.0232, "step": 16970 }, { "grad_norm": 0.1926964670419693, "learning_rate": 9.612789698688216e-05, "loss": 0.021, "step": 16980 }, { "grad_norm": 0.2906971573829651, "learning_rate": 9.612151440951334e-05, "loss": 0.0262, "step": 16990 }, { "grad_norm": 0.4140971601009369, "learning_rate": 9.611512678836506e-05, "loss": 0.024, "step": 17000 }, { "grad_norm": 0.25733935832977295, "learning_rate": 9.610873412413584e-05, "loss": 0.0337, "step": 17010 }, { "grad_norm": 0.3416774570941925, "learning_rate": 9.610233641752476e-05, "loss": 0.0315, "step": 17020 }, { "grad_norm": 0.2333684265613556, "learning_rate": 9.609593366923151e-05, "loss": 0.0206, "step": 17030 }, { "grad_norm": 0.1659603863954544, "learning_rate": 9.608952587995625e-05, "loss": 0.0184, "step": 17040 }, { "grad_norm": 0.2997359037399292, "learning_rate": 9.608311305039972e-05, "loss": 0.0216, "step": 17050 }, { "grad_norm": 0.3500133156776428, "learning_rate": 9.607669518126326e-05, "loss": 0.0291, "step": 17060 }, { "grad_norm": 0.15774907171726227, "learning_rate": 9.607027227324866e-05, "loss": 0.0192, "step": 17070 }, { "grad_norm": 0.24987205862998962, "learning_rate": 9.606384432705837e-05, "loss": 0.0196, "step": 17080 }, { "grad_norm": 0.1837649792432785, "learning_rate": 9.60574113433953e-05, "loss": 0.0235, "step": 17090 }, { "grad_norm": 0.1544526070356369, "learning_rate": 9.6050973322963e-05, "loss": 0.0196, "step": 17100 }, { "grad_norm": 0.2083483785390854, "learning_rate": 9.604453026646547e-05, "loss": 0.0226, "step": 17110 }, { "grad_norm": 0.24396087229251862, "learning_rate": 9.603808217460735e-05, "loss": 0.0195, "step": 17120 }, { "grad_norm": 0.22539688646793365, "learning_rate": 9.603162904809377e-05, "loss": 0.0196, "step": 17130 }, { "grad_norm": 0.15901462733745575, "learning_rate": 9.602517088763045e-05, "loss": 0.0171, "step": 17140 }, { "grad_norm": 0.17515578866004944, "learning_rate": 9.601870769392365e-05, "loss": 0.0235, "step": 17150 }, { "grad_norm": 0.1883973926305771, "learning_rate": 9.601223946768017e-05, "loss": 0.022, "step": 17160 }, { "grad_norm": 0.22877620160579681, "learning_rate": 9.600576620960734e-05, "loss": 0.0243, "step": 17170 }, { "grad_norm": 0.19257745146751404, "learning_rate": 9.599928792041308e-05, "loss": 0.0205, "step": 17180 }, { "grad_norm": 0.2029264271259308, "learning_rate": 9.599280460080587e-05, "loss": 0.0191, "step": 17190 }, { "grad_norm": 0.20502249896526337, "learning_rate": 9.59863162514947e-05, "loss": 0.0228, "step": 17200 }, { "grad_norm": 0.23082223534584045, "learning_rate": 9.597982287318911e-05, "loss": 0.0211, "step": 17210 }, { "grad_norm": 0.2611636519432068, "learning_rate": 9.597332446659923e-05, "loss": 0.0212, "step": 17220 }, { "grad_norm": 0.22786951065063477, "learning_rate": 9.59668210324357e-05, "loss": 0.0208, "step": 17230 }, { "grad_norm": 0.22298607230186462, "learning_rate": 9.596031257140974e-05, "loss": 0.0195, "step": 17240 }, { "grad_norm": 0.17596101760864258, "learning_rate": 9.59537990842331e-05, "loss": 0.0312, "step": 17250 }, { "grad_norm": 0.32225391268730164, "learning_rate": 9.594728057161806e-05, "loss": 0.0217, "step": 17260 }, { "grad_norm": 0.16980716586112976, "learning_rate": 9.594075703427752e-05, "loss": 0.0202, "step": 17270 }, { "grad_norm": 0.1576445996761322, "learning_rate": 9.593422847292486e-05, "loss": 0.0173, "step": 17280 }, { "grad_norm": 0.173395574092865, "learning_rate": 9.592769488827402e-05, "loss": 0.0217, "step": 17290 }, { "grad_norm": 0.2114131599664688, "learning_rate": 9.592115628103952e-05, "loss": 0.0207, "step": 17300 }, { "grad_norm": 0.17729230225086212, "learning_rate": 9.591461265193643e-05, "loss": 0.0193, "step": 17310 }, { "grad_norm": 0.23160719871520996, "learning_rate": 9.590806400168032e-05, "loss": 0.015, "step": 17320 }, { "grad_norm": 0.21311993896961212, "learning_rate": 9.590151033098735e-05, "loss": 0.0172, "step": 17330 }, { "grad_norm": 0.28468069434165955, "learning_rate": 9.589495164057423e-05, "loss": 0.022, "step": 17340 }, { "grad_norm": 0.20538027584552765, "learning_rate": 9.58883879311582e-05, "loss": 0.0179, "step": 17350 }, { "grad_norm": 0.2524874806404114, "learning_rate": 9.588181920345705e-05, "loss": 0.0242, "step": 17360 }, { "grad_norm": 0.17630890011787415, "learning_rate": 9.587524545818913e-05, "loss": 0.0196, "step": 17370 }, { "grad_norm": 0.18990768492221832, "learning_rate": 9.586866669607335e-05, "loss": 0.0201, "step": 17380 }, { "grad_norm": 0.15131911635398865, "learning_rate": 9.586208291782915e-05, "loss": 0.0181, "step": 17390 }, { "grad_norm": 0.22185005247592926, "learning_rate": 9.58554941241765e-05, "loss": 0.0235, "step": 17400 }, { "grad_norm": 0.3347283899784088, "learning_rate": 9.584890031583596e-05, "loss": 0.0172, "step": 17410 }, { "grad_norm": 0.2512843608856201, "learning_rate": 9.584230149352861e-05, "loss": 0.0218, "step": 17420 }, { "grad_norm": 0.2038719654083252, "learning_rate": 9.58356976579761e-05, "loss": 0.0181, "step": 17430 }, { "grad_norm": 0.30656448006629944, "learning_rate": 9.58290888099006e-05, "loss": 0.0249, "step": 17440 }, { "grad_norm": 0.22502164542675018, "learning_rate": 9.582247495002486e-05, "loss": 0.0227, "step": 17450 }, { "grad_norm": 0.23182132840156555, "learning_rate": 9.581585607907214e-05, "loss": 0.0206, "step": 17460 }, { "grad_norm": 0.13793103396892548, "learning_rate": 9.580923219776628e-05, "loss": 0.0183, "step": 17470 }, { "grad_norm": 0.19000419974327087, "learning_rate": 9.580260330683167e-05, "loss": 0.0213, "step": 17480 }, { "grad_norm": 0.19827218353748322, "learning_rate": 9.579596940699322e-05, "loss": 0.017, "step": 17490 }, { "grad_norm": 0.12744519114494324, "learning_rate": 9.578933049897643e-05, "loss": 0.0162, "step": 17500 }, { "grad_norm": 0.24880114197731018, "learning_rate": 9.578268658350728e-05, "loss": 0.021, "step": 17510 }, { "grad_norm": 0.18417878448963165, "learning_rate": 9.577603766131235e-05, "loss": 0.0206, "step": 17520 }, { "grad_norm": 0.21132266521453857, "learning_rate": 9.576938373311878e-05, "loss": 0.0199, "step": 17530 }, { "grad_norm": 0.2750726044178009, "learning_rate": 9.576272479965421e-05, "loss": 0.0251, "step": 17540 }, { "grad_norm": 0.28772613406181335, "learning_rate": 9.575606086164687e-05, "loss": 0.0193, "step": 17550 }, { "grad_norm": 0.23640571534633636, "learning_rate": 9.57493919198255e-05, "loss": 0.0371, "step": 17560 }, { "grad_norm": 0.23670803010463715, "learning_rate": 9.57427179749194e-05, "loss": 0.0291, "step": 17570 }, { "grad_norm": 0.22795608639717102, "learning_rate": 9.573603902765846e-05, "loss": 0.0233, "step": 17580 }, { "grad_norm": 0.19135384261608124, "learning_rate": 9.572935507877304e-05, "loss": 0.0267, "step": 17590 }, { "grad_norm": 0.2402127981185913, "learning_rate": 9.57226661289941e-05, "loss": 0.0338, "step": 17600 }, { "grad_norm": 0.2683235704898834, "learning_rate": 9.571597217905315e-05, "loss": 0.0286, "step": 17610 }, { "grad_norm": 0.2696508467197418, "learning_rate": 9.57092732296822e-05, "loss": 0.0225, "step": 17620 }, { "grad_norm": 0.1950652301311493, "learning_rate": 9.570256928161385e-05, "loss": 0.0176, "step": 17630 }, { "grad_norm": 0.2877263128757477, "learning_rate": 9.569586033558126e-05, "loss": 0.024, "step": 17640 }, { "grad_norm": 0.14501658082008362, "learning_rate": 9.568914639231807e-05, "loss": 0.02, "step": 17650 }, { "grad_norm": 0.2731291651725769, "learning_rate": 9.568242745255852e-05, "loss": 0.0311, "step": 17660 }, { "grad_norm": 0.17453086376190186, "learning_rate": 9.567570351703739e-05, "loss": 0.0226, "step": 17670 }, { "grad_norm": 0.2446589171886444, "learning_rate": 9.566897458649001e-05, "loss": 0.0235, "step": 17680 }, { "grad_norm": 0.17220504581928253, "learning_rate": 9.566224066165221e-05, "loss": 0.0218, "step": 17690 }, { "grad_norm": 0.16382788121700287, "learning_rate": 9.565550174326043e-05, "loss": 0.0267, "step": 17700 }, { "grad_norm": 0.211470365524292, "learning_rate": 9.564875783205162e-05, "loss": 0.0247, "step": 17710 }, { "grad_norm": 0.17239831387996674, "learning_rate": 9.564200892876328e-05, "loss": 0.023, "step": 17720 }, { "grad_norm": 0.16803207993507385, "learning_rate": 9.563525503413348e-05, "loss": 0.0204, "step": 17730 }, { "grad_norm": 0.22838640213012695, "learning_rate": 9.562849614890079e-05, "loss": 0.0299, "step": 17740 }, { "grad_norm": 0.2683706283569336, "learning_rate": 9.562173227380436e-05, "loss": 0.0316, "step": 17750 }, { "grad_norm": 0.24376346170902252, "learning_rate": 9.561496340958389e-05, "loss": 0.0265, "step": 17760 }, { "grad_norm": 0.2282353639602661, "learning_rate": 9.560818955697959e-05, "loss": 0.0194, "step": 17770 }, { "grad_norm": 0.2513279616832733, "learning_rate": 9.560141071673228e-05, "loss": 0.0287, "step": 17780 }, { "grad_norm": 0.15650516748428345, "learning_rate": 9.559462688958323e-05, "loss": 0.0152, "step": 17790 }, { "grad_norm": 0.2773297131061554, "learning_rate": 9.558783807627434e-05, "loss": 0.0196, "step": 17800 }, { "grad_norm": 0.28079211711883545, "learning_rate": 9.558104427754801e-05, "loss": 0.0265, "step": 17810 }, { "grad_norm": 0.28383079171180725, "learning_rate": 9.557424549414722e-05, "loss": 0.0226, "step": 17820 }, { "grad_norm": 0.21081598103046417, "learning_rate": 9.556744172681546e-05, "loss": 0.0229, "step": 17830 }, { "grad_norm": 0.23152944445610046, "learning_rate": 9.556063297629677e-05, "loss": 0.021, "step": 17840 }, { "grad_norm": 0.173445463180542, "learning_rate": 9.555381924333578e-05, "loss": 0.0204, "step": 17850 }, { "grad_norm": 0.23167748749256134, "learning_rate": 9.554700052867758e-05, "loss": 0.0217, "step": 17860 }, { "grad_norm": 0.2523997128009796, "learning_rate": 9.554017683306789e-05, "loss": 0.0206, "step": 17870 }, { "grad_norm": 0.21970833837985992, "learning_rate": 9.553334815725294e-05, "loss": 0.0201, "step": 17880 }, { "grad_norm": 0.2282545417547226, "learning_rate": 9.552651450197949e-05, "loss": 0.0213, "step": 17890 }, { "grad_norm": 0.24348290264606476, "learning_rate": 9.551967586799486e-05, "loss": 0.0186, "step": 17900 }, { "grad_norm": 0.3623904585838318, "learning_rate": 9.551283225604692e-05, "loss": 0.0221, "step": 17910 }, { "grad_norm": 0.18155770003795624, "learning_rate": 9.550598366688406e-05, "loss": 0.0193, "step": 17920 }, { "grad_norm": 0.24059394001960754, "learning_rate": 9.549913010125526e-05, "loss": 0.0201, "step": 17930 }, { "grad_norm": 0.256091833114624, "learning_rate": 9.549227155990999e-05, "loss": 0.0195, "step": 17940 }, { "grad_norm": 0.18766145408153534, "learning_rate": 9.548540804359828e-05, "loss": 0.027, "step": 17950 }, { "grad_norm": 0.2673878073692322, "learning_rate": 9.547853955307077e-05, "loss": 0.0253, "step": 17960 }, { "grad_norm": 0.1836382895708084, "learning_rate": 9.547166608907853e-05, "loss": 0.0166, "step": 17970 }, { "grad_norm": 0.2368585765361786, "learning_rate": 9.546478765237326e-05, "loss": 0.0305, "step": 17980 }, { "grad_norm": 0.1978825032711029, "learning_rate": 9.545790424370715e-05, "loss": 0.0187, "step": 17990 }, { "grad_norm": 0.15770618617534637, "learning_rate": 9.5451015863833e-05, "loss": 0.0253, "step": 18000 }, { "grad_norm": 0.18404537439346313, "learning_rate": 9.544412251350408e-05, "loss": 0.0296, "step": 18010 }, { "grad_norm": 0.18554316461086273, "learning_rate": 9.543722419347422e-05, "loss": 0.0244, "step": 18020 }, { "grad_norm": 0.260405957698822, "learning_rate": 9.543032090449788e-05, "loss": 0.0262, "step": 18030 }, { "grad_norm": 0.24719688296318054, "learning_rate": 9.542341264732992e-05, "loss": 0.0259, "step": 18040 }, { "grad_norm": 0.22334982454776764, "learning_rate": 9.541649942272585e-05, "loss": 0.0276, "step": 18050 }, { "grad_norm": 0.28912943601608276, "learning_rate": 9.54095812314417e-05, "loss": 0.0214, "step": 18060 }, { "grad_norm": 0.18731479346752167, "learning_rate": 9.540265807423401e-05, "loss": 0.0217, "step": 18070 }, { "grad_norm": 0.22352194786071777, "learning_rate": 9.53957299518599e-05, "loss": 0.0183, "step": 18080 }, { "grad_norm": 0.19138874113559723, "learning_rate": 9.5388796865077e-05, "loss": 0.0223, "step": 18090 }, { "grad_norm": 0.2054857611656189, "learning_rate": 9.538185881464353e-05, "loss": 0.0251, "step": 18100 }, { "grad_norm": 0.25252488255500793, "learning_rate": 9.537491580131821e-05, "loss": 0.0214, "step": 18110 }, { "grad_norm": 0.19445832073688507, "learning_rate": 9.53679678258603e-05, "loss": 0.0197, "step": 18120 }, { "grad_norm": 0.2574584484100342, "learning_rate": 9.536101488902966e-05, "loss": 0.02, "step": 18130 }, { "grad_norm": 0.14773067831993103, "learning_rate": 9.535405699158663e-05, "loss": 0.0172, "step": 18140 }, { "grad_norm": 0.20646683871746063, "learning_rate": 9.53470941342921e-05, "loss": 0.0187, "step": 18150 }, { "grad_norm": 0.19319307804107666, "learning_rate": 9.534012631790756e-05, "loss": 0.0206, "step": 18160 }, { "grad_norm": 0.21985849738121033, "learning_rate": 9.533315354319494e-05, "loss": 0.0196, "step": 18170 }, { "grad_norm": 0.2058899700641632, "learning_rate": 9.532617581091682e-05, "loss": 0.0188, "step": 18180 }, { "grad_norm": 0.13003945350646973, "learning_rate": 9.531919312183629e-05, "loss": 0.0209, "step": 18190 }, { "grad_norm": 0.1378670632839203, "learning_rate": 9.531220547671688e-05, "loss": 0.0161, "step": 18200 }, { "grad_norm": 0.18234750628471375, "learning_rate": 9.530521287632285e-05, "loss": 0.0197, "step": 18210 }, { "grad_norm": 0.16359075903892517, "learning_rate": 9.529821532141884e-05, "loss": 0.0164, "step": 18220 }, { "grad_norm": 0.14549627900123596, "learning_rate": 9.52912128127701e-05, "loss": 0.017, "step": 18230 }, { "grad_norm": 0.17217917740345, "learning_rate": 9.528420535114244e-05, "loss": 0.019, "step": 18240 }, { "grad_norm": 0.26051798462867737, "learning_rate": 9.527719293730215e-05, "loss": 0.0248, "step": 18250 }, { "grad_norm": 0.18480494618415833, "learning_rate": 9.527017557201611e-05, "loss": 0.0197, "step": 18260 }, { "grad_norm": 0.21862785518169403, "learning_rate": 9.526315325605176e-05, "loss": 0.0243, "step": 18270 }, { "grad_norm": 0.25802603363990784, "learning_rate": 9.525612599017699e-05, "loss": 0.0169, "step": 18280 }, { "grad_norm": 0.24266138672828674, "learning_rate": 9.524909377516033e-05, "loss": 0.0206, "step": 18290 }, { "grad_norm": 0.258450448513031, "learning_rate": 9.524205661177081e-05, "loss": 0.0189, "step": 18300 }, { "grad_norm": 0.16772693395614624, "learning_rate": 9.523501450077801e-05, "loss": 0.0216, "step": 18310 }, { "grad_norm": 0.25808578729629517, "learning_rate": 9.522796744295202e-05, "loss": 0.0225, "step": 18320 }, { "grad_norm": 0.17860563099384308, "learning_rate": 9.522091543906352e-05, "loss": 0.0199, "step": 18330 }, { "grad_norm": 0.28148385882377625, "learning_rate": 9.521385848988369e-05, "loss": 0.0199, "step": 18340 }, { "grad_norm": 0.1556948721408844, "learning_rate": 9.520679659618428e-05, "loss": 0.0177, "step": 18350 }, { "grad_norm": 0.24055969715118408, "learning_rate": 9.519972975873754e-05, "loss": 0.0197, "step": 18360 }, { "grad_norm": 0.21934881806373596, "learning_rate": 9.519265797831633e-05, "loss": 0.0178, "step": 18370 }, { "grad_norm": 0.22248351573944092, "learning_rate": 9.518558125569399e-05, "loss": 0.0197, "step": 18380 }, { "grad_norm": 0.15944376587867737, "learning_rate": 9.517849959164442e-05, "loss": 0.0173, "step": 18390 }, { "grad_norm": 0.23073242604732513, "learning_rate": 9.517141298694205e-05, "loss": 0.025, "step": 18400 }, { "grad_norm": 0.2366045117378235, "learning_rate": 9.516432144236188e-05, "loss": 0.0243, "step": 18410 }, { "grad_norm": 0.19486001133918762, "learning_rate": 9.515722495867941e-05, "loss": 0.0253, "step": 18420 }, { "grad_norm": 0.17806367576122284, "learning_rate": 9.515012353667072e-05, "loss": 0.0193, "step": 18430 }, { "grad_norm": 0.19369521737098694, "learning_rate": 9.51430171771124e-05, "loss": 0.0174, "step": 18440 }, { "grad_norm": 0.25673139095306396, "learning_rate": 9.513590588078159e-05, "loss": 0.0263, "step": 18450 }, { "grad_norm": 0.2950800061225891, "learning_rate": 9.512878964845597e-05, "loss": 0.0229, "step": 18460 }, { "grad_norm": 0.3413749933242798, "learning_rate": 9.512166848091377e-05, "loss": 0.0286, "step": 18470 }, { "grad_norm": 0.2001888006925583, "learning_rate": 9.511454237893376e-05, "loss": 0.0222, "step": 18480 }, { "grad_norm": 0.19157154858112335, "learning_rate": 9.51074113432952e-05, "loss": 0.0194, "step": 18490 }, { "grad_norm": 0.16805395483970642, "learning_rate": 9.510027537477797e-05, "loss": 0.029, "step": 18500 }, { "grad_norm": 0.17252285778522491, "learning_rate": 9.509313447416242e-05, "loss": 0.0224, "step": 18510 }, { "grad_norm": 0.22618085145950317, "learning_rate": 9.508598864222949e-05, "loss": 0.0197, "step": 18520 }, { "grad_norm": 0.2342989146709442, "learning_rate": 9.507883787976062e-05, "loss": 0.0226, "step": 18530 }, { "grad_norm": 0.20941317081451416, "learning_rate": 9.507168218753781e-05, "loss": 0.017, "step": 18540 }, { "grad_norm": 0.17616622149944305, "learning_rate": 9.506452156634362e-05, "loss": 0.0195, "step": 18550 }, { "grad_norm": 0.2163083404302597, "learning_rate": 9.505735601696109e-05, "loss": 0.0296, "step": 18560 }, { "grad_norm": 0.24890772998332977, "learning_rate": 9.505018554017385e-05, "loss": 0.0222, "step": 18570 }, { "grad_norm": 0.3414008319377899, "learning_rate": 9.504301013676604e-05, "loss": 0.0198, "step": 18580 }, { "grad_norm": 0.265550434589386, "learning_rate": 9.503582980752238e-05, "loss": 0.02, "step": 18590 }, { "grad_norm": 0.34012913703918457, "learning_rate": 9.502864455322809e-05, "loss": 0.0345, "step": 18600 }, { "grad_norm": 0.16547754406929016, "learning_rate": 9.502145437466891e-05, "loss": 0.0195, "step": 18610 }, { "grad_norm": 0.2319410890340805, "learning_rate": 9.501425927263116e-05, "loss": 0.0333, "step": 18620 }, { "grad_norm": 0.4346678555011749, "learning_rate": 9.500705924790172e-05, "loss": 0.0508, "step": 18630 }, { "grad_norm": 0.2646629810333252, "learning_rate": 9.499985430126794e-05, "loss": 0.0291, "step": 18640 }, { "grad_norm": 0.26710087060928345, "learning_rate": 9.499264443351775e-05, "loss": 0.0263, "step": 18650 }, { "grad_norm": 0.182930126786232, "learning_rate": 9.498542964543961e-05, "loss": 0.0249, "step": 18660 }, { "grad_norm": 0.2168819010257721, "learning_rate": 9.497820993782252e-05, "loss": 0.0245, "step": 18670 }, { "grad_norm": 0.28560495376586914, "learning_rate": 9.497098531145601e-05, "loss": 0.0245, "step": 18680 }, { "grad_norm": 0.21701680123806, "learning_rate": 9.496375576713017e-05, "loss": 0.0216, "step": 18690 }, { "grad_norm": 0.22851860523223877, "learning_rate": 9.49565213056356e-05, "loss": 0.0178, "step": 18700 }, { "grad_norm": 0.18136265873908997, "learning_rate": 9.494928192776342e-05, "loss": 0.0181, "step": 18710 }, { "grad_norm": 0.15216580033302307, "learning_rate": 9.494203763430538e-05, "loss": 0.0162, "step": 18720 }, { "grad_norm": 0.30140045285224915, "learning_rate": 9.493478842605366e-05, "loss": 0.0422, "step": 18730 }, { "grad_norm": 0.16871806979179382, "learning_rate": 9.492753430380105e-05, "loss": 0.0168, "step": 18740 }, { "grad_norm": 0.20138181746006012, "learning_rate": 9.492027526834083e-05, "loss": 0.0174, "step": 18750 }, { "grad_norm": 0.22961701452732086, "learning_rate": 9.491301132046684e-05, "loss": 0.02, "step": 18760 }, { "grad_norm": 0.16550928354263306, "learning_rate": 9.490574246097345e-05, "loss": 0.0365, "step": 18770 }, { "grad_norm": 0.3014562427997589, "learning_rate": 9.48984686906556e-05, "loss": 0.0327, "step": 18780 }, { "grad_norm": 0.29291802644729614, "learning_rate": 9.489119001030871e-05, "loss": 0.0228, "step": 18790 }, { "grad_norm": 0.2518567442893982, "learning_rate": 9.488390642072878e-05, "loss": 0.0201, "step": 18800 }, { "grad_norm": 0.21866931021213531, "learning_rate": 9.48766179227123e-05, "loss": 0.0218, "step": 18810 }, { "grad_norm": 0.22721697390079498, "learning_rate": 9.486932451705636e-05, "loss": 0.0223, "step": 18820 }, { "grad_norm": 0.19209323823451996, "learning_rate": 9.486202620455857e-05, "loss": 0.0201, "step": 18830 }, { "grad_norm": 0.22952277958393097, "learning_rate": 9.485472298601704e-05, "loss": 0.0194, "step": 18840 }, { "grad_norm": 0.21018904447555542, "learning_rate": 9.484741486223043e-05, "loss": 0.0217, "step": 18850 }, { "grad_norm": 0.20222115516662598, "learning_rate": 9.484010183399797e-05, "loss": 0.0253, "step": 18860 }, { "grad_norm": 0.2468768209218979, "learning_rate": 9.483278390211938e-05, "loss": 0.0231, "step": 18870 }, { "grad_norm": 0.1874411255121231, "learning_rate": 9.482546106739496e-05, "loss": 0.02, "step": 18880 }, { "grad_norm": 0.19859202206134796, "learning_rate": 9.48181333306255e-05, "loss": 0.0182, "step": 18890 }, { "grad_norm": 0.23497538268566132, "learning_rate": 9.481080069261237e-05, "loss": 0.0303, "step": 18900 }, { "grad_norm": 0.2026180922985077, "learning_rate": 9.480346315415745e-05, "loss": 0.0178, "step": 18910 }, { "grad_norm": 0.2654508650302887, "learning_rate": 9.479612071606314e-05, "loss": 0.0191, "step": 18920 }, { "grad_norm": 0.19468429684638977, "learning_rate": 9.478877337913244e-05, "loss": 0.024, "step": 18930 }, { "grad_norm": 0.20581850409507751, "learning_rate": 9.478142114416881e-05, "loss": 0.0186, "step": 18940 }, { "grad_norm": 0.15367262065410614, "learning_rate": 9.47740640119763e-05, "loss": 0.0163, "step": 18950 }, { "grad_norm": 0.2071981579065323, "learning_rate": 9.476670198335947e-05, "loss": 0.0208, "step": 18960 }, { "grad_norm": 0.2144143134355545, "learning_rate": 9.47593350591234e-05, "loss": 0.022, "step": 18970 }, { "grad_norm": 0.2270154356956482, "learning_rate": 9.475196324007376e-05, "loss": 0.0222, "step": 18980 }, { "grad_norm": 0.2869788110256195, "learning_rate": 9.474458652701669e-05, "loss": 0.0182, "step": 18990 }, { "grad_norm": 0.24617350101470947, "learning_rate": 9.473720492075892e-05, "loss": 0.024, "step": 19000 }, { "grad_norm": 0.291413813829422, "learning_rate": 9.472981842210768e-05, "loss": 0.0234, "step": 19010 }, { "grad_norm": 0.25664734840393066, "learning_rate": 9.472242703187074e-05, "loss": 0.0215, "step": 19020 }, { "grad_norm": 0.2095668464899063, "learning_rate": 9.471503075085643e-05, "loss": 0.0165, "step": 19030 }, { "grad_norm": 0.20559075474739075, "learning_rate": 9.470762957987359e-05, "loss": 0.0232, "step": 19040 }, { "grad_norm": 0.20575709640979767, "learning_rate": 9.470022351973158e-05, "loss": 0.0237, "step": 19050 }, { "grad_norm": 0.2626223564147949, "learning_rate": 9.469281257124034e-05, "loss": 0.0198, "step": 19060 }, { "grad_norm": 0.2057984173297882, "learning_rate": 9.46853967352103e-05, "loss": 0.0249, "step": 19070 }, { "grad_norm": 0.23726493120193481, "learning_rate": 9.467797601245246e-05, "loss": 0.0177, "step": 19080 }, { "grad_norm": 0.22519363462924957, "learning_rate": 9.467055040377834e-05, "loss": 0.0259, "step": 19090 }, { "grad_norm": 0.2012433409690857, "learning_rate": 9.466311990999999e-05, "loss": 0.0273, "step": 19100 }, { "grad_norm": 0.21543440222740173, "learning_rate": 9.465568453193e-05, "loss": 0.0207, "step": 19110 }, { "grad_norm": 0.1535288244485855, "learning_rate": 9.464824427038148e-05, "loss": 0.0176, "step": 19120 }, { "grad_norm": 0.15346480906009674, "learning_rate": 9.46407991261681e-05, "loss": 0.0277, "step": 19130 }, { "grad_norm": 0.16320325434207916, "learning_rate": 9.463334910010404e-05, "loss": 0.0166, "step": 19140 }, { "grad_norm": 0.21448054909706116, "learning_rate": 9.462589419300403e-05, "loss": 0.0131, "step": 19150 }, { "grad_norm": 0.22594919800758362, "learning_rate": 9.461843440568333e-05, "loss": 0.0219, "step": 19160 }, { "grad_norm": 0.15533927083015442, "learning_rate": 9.461096973895773e-05, "loss": 0.0175, "step": 19170 }, { "grad_norm": 0.16989697515964508, "learning_rate": 9.460350019364355e-05, "loss": 0.0214, "step": 19180 }, { "grad_norm": 0.2892799973487854, "learning_rate": 9.459602577055764e-05, "loss": 0.0367, "step": 19190 }, { "grad_norm": 0.2670445442199707, "learning_rate": 9.45885464705174e-05, "loss": 0.0216, "step": 19200 }, { "grad_norm": 0.2659730315208435, "learning_rate": 9.458106229434076e-05, "loss": 0.0189, "step": 19210 }, { "grad_norm": 0.2348109632730484, "learning_rate": 9.457357324284617e-05, "loss": 0.0179, "step": 19220 }, { "grad_norm": 0.2490309178829193, "learning_rate": 9.456607931685262e-05, "loss": 0.0231, "step": 19230 }, { "grad_norm": 0.18420226871967316, "learning_rate": 9.455858051717965e-05, "loss": 0.0206, "step": 19240 }, { "grad_norm": 0.23679938912391663, "learning_rate": 9.45510768446473e-05, "loss": 0.0195, "step": 19250 }, { "grad_norm": 0.1762048900127411, "learning_rate": 9.454356830007618e-05, "loss": 0.0191, "step": 19260 }, { "grad_norm": 0.14079336822032928, "learning_rate": 9.45360548842874e-05, "loss": 0.0149, "step": 19270 }, { "grad_norm": 0.1748078614473343, "learning_rate": 9.452853659810261e-05, "loss": 0.0195, "step": 19280 }, { "grad_norm": 0.2874862551689148, "learning_rate": 9.452101344234401e-05, "loss": 0.0255, "step": 19290 }, { "grad_norm": 0.19178834557533264, "learning_rate": 9.451348541783431e-05, "loss": 0.0205, "step": 19300 }, { "grad_norm": 0.24625755846500397, "learning_rate": 9.450595252539678e-05, "loss": 0.0183, "step": 19310 }, { "grad_norm": 0.17598050832748413, "learning_rate": 9.449841476585518e-05, "loss": 0.0266, "step": 19320 }, { "grad_norm": 0.2696547508239746, "learning_rate": 9.449087214003384e-05, "loss": 0.0362, "step": 19330 }, { "grad_norm": 0.3208564817905426, "learning_rate": 9.448332464875765e-05, "loss": 0.0374, "step": 19340 }, { "grad_norm": 0.21670958399772644, "learning_rate": 9.447577229285192e-05, "loss": 0.0258, "step": 19350 }, { "grad_norm": 0.17043329775333405, "learning_rate": 9.446821507314261e-05, "loss": 0.0167, "step": 19360 }, { "grad_norm": 0.22059570252895355, "learning_rate": 9.446065299045617e-05, "loss": 0.0186, "step": 19370 }, { "grad_norm": 0.2705771327018738, "learning_rate": 9.445308604561955e-05, "loss": 0.0306, "step": 19380 }, { "grad_norm": 0.1967778354883194, "learning_rate": 9.444551423946028e-05, "loss": 0.0233, "step": 19390 }, { "grad_norm": 0.1331602931022644, "learning_rate": 9.443793757280638e-05, "loss": 0.0196, "step": 19400 }, { "grad_norm": 0.20805591344833374, "learning_rate": 9.443035604648646e-05, "loss": 0.0177, "step": 19410 }, { "grad_norm": 0.17795272171497345, "learning_rate": 9.44227696613296e-05, "loss": 0.0177, "step": 19420 }, { "grad_norm": 0.25592532753944397, "learning_rate": 9.441517841816542e-05, "loss": 0.0316, "step": 19430 }, { "grad_norm": 0.213293194770813, "learning_rate": 9.440758231782413e-05, "loss": 0.0188, "step": 19440 }, { "grad_norm": 0.20883995294570923, "learning_rate": 9.439998136113639e-05, "loss": 0.0201, "step": 19450 }, { "grad_norm": 0.2135203629732132, "learning_rate": 9.439237554893344e-05, "loss": 0.0193, "step": 19460 }, { "grad_norm": 0.1479957401752472, "learning_rate": 9.438476488204705e-05, "loss": 0.0253, "step": 19470 }, { "grad_norm": 0.24195392429828644, "learning_rate": 9.43771493613095e-05, "loss": 0.0166, "step": 19480 }, { "grad_norm": 0.21614696085453033, "learning_rate": 9.436952898755362e-05, "loss": 0.0314, "step": 19490 }, { "grad_norm": 0.1854080855846405, "learning_rate": 9.436190376161276e-05, "loss": 0.0187, "step": 19500 }, { "grad_norm": 0.23211205005645752, "learning_rate": 9.43542736843208e-05, "loss": 0.0288, "step": 19510 }, { "grad_norm": 0.2361888736486435, "learning_rate": 9.434663875651216e-05, "loss": 0.0173, "step": 19520 }, { "grad_norm": 0.21142590045928955, "learning_rate": 9.433899897902177e-05, "loss": 0.0174, "step": 19530 }, { "grad_norm": 0.2643749713897705, "learning_rate": 9.433135435268511e-05, "loss": 0.0178, "step": 19540 }, { "grad_norm": 0.2501124441623688, "learning_rate": 9.432370487833819e-05, "loss": 0.0212, "step": 19550 }, { "grad_norm": 0.2937639355659485, "learning_rate": 9.431605055681756e-05, "loss": 0.0295, "step": 19560 }, { "grad_norm": 0.21697171032428741, "learning_rate": 9.430839138896026e-05, "loss": 0.0224, "step": 19570 }, { "grad_norm": 0.20717543363571167, "learning_rate": 9.43007273756039e-05, "loss": 0.0242, "step": 19580 }, { "grad_norm": 0.16019970178604126, "learning_rate": 9.429305851758658e-05, "loss": 0.0204, "step": 19590 }, { "grad_norm": 0.2301650196313858, "learning_rate": 9.428538481574699e-05, "loss": 0.0234, "step": 19600 }, { "grad_norm": 0.2386363297700882, "learning_rate": 9.42777062709243e-05, "loss": 0.0207, "step": 19610 }, { "grad_norm": 0.1892164796590805, "learning_rate": 9.427002288395821e-05, "loss": 0.0218, "step": 19620 }, { "grad_norm": 0.19338077306747437, "learning_rate": 9.426233465568898e-05, "loss": 0.0235, "step": 19630 }, { "grad_norm": 0.21491307020187378, "learning_rate": 9.42546415869574e-05, "loss": 0.0176, "step": 19640 }, { "grad_norm": 0.2510479986667633, "learning_rate": 9.424694367860473e-05, "loss": 0.0221, "step": 19650 }, { "grad_norm": 0.18599219620227814, "learning_rate": 9.423924093147284e-05, "loss": 0.0173, "step": 19660 }, { "grad_norm": 0.15810248255729675, "learning_rate": 9.423153334640407e-05, "loss": 0.0163, "step": 19670 }, { "grad_norm": 0.15638203918933868, "learning_rate": 9.42238209242413e-05, "loss": 0.0172, "step": 19680 }, { "grad_norm": 0.2116442322731018, "learning_rate": 9.421610366582798e-05, "loss": 0.0223, "step": 19690 }, { "grad_norm": 0.18089263141155243, "learning_rate": 9.420838157200803e-05, "loss": 0.0174, "step": 19700 }, { "grad_norm": 0.17905841767787933, "learning_rate": 9.420065464362594e-05, "loss": 0.0221, "step": 19710 }, { "grad_norm": 0.23214414715766907, "learning_rate": 9.419292288152673e-05, "loss": 0.0232, "step": 19720 }, { "grad_norm": 0.22945937514305115, "learning_rate": 9.418518628655588e-05, "loss": 0.0189, "step": 19730 }, { "grad_norm": 0.14317046105861664, "learning_rate": 9.417744485955951e-05, "loss": 0.018, "step": 19740 }, { "grad_norm": 0.17155402898788452, "learning_rate": 9.41696986013842e-05, "loss": 0.0157, "step": 19750 }, { "grad_norm": 0.20144151151180267, "learning_rate": 9.416194751287705e-05, "loss": 0.0243, "step": 19760 }, { "grad_norm": 0.19274020195007324, "learning_rate": 9.415419159488572e-05, "loss": 0.0177, "step": 19770 }, { "grad_norm": 0.22246462106704712, "learning_rate": 9.414643084825837e-05, "loss": 0.023, "step": 19780 }, { "grad_norm": 0.25790515542030334, "learning_rate": 9.413866527384372e-05, "loss": 0.0249, "step": 19790 }, { "grad_norm": 0.22966152429580688, "learning_rate": 9.4130894872491e-05, "loss": 0.0175, "step": 19800 }, { "grad_norm": 0.18453803658485413, "learning_rate": 9.412311964504998e-05, "loss": 0.0199, "step": 19810 }, { "grad_norm": 0.16644613444805145, "learning_rate": 9.411533959237091e-05, "loss": 0.0197, "step": 19820 }, { "grad_norm": 0.16314196586608887, "learning_rate": 9.410755471530464e-05, "loss": 0.014, "step": 19830 }, { "grad_norm": 0.1908222883939743, "learning_rate": 9.40997650147025e-05, "loss": 0.0177, "step": 19840 }, { "grad_norm": 0.34818896651268005, "learning_rate": 9.409197049141637e-05, "loss": 0.0272, "step": 19850 }, { "grad_norm": 0.24903902411460876, "learning_rate": 9.408417114629863e-05, "loss": 0.0325, "step": 19860 }, { "grad_norm": 0.21538856625556946, "learning_rate": 9.40763669802022e-05, "loss": 0.0199, "step": 19870 }, { "grad_norm": 0.1841929852962494, "learning_rate": 9.406855799398056e-05, "loss": 0.03, "step": 19880 }, { "grad_norm": 0.19905972480773926, "learning_rate": 9.406074418848767e-05, "loss": 0.0245, "step": 19890 }, { "grad_norm": 0.3352203667163849, "learning_rate": 9.405292556457805e-05, "loss": 0.0377, "step": 19900 }, { "grad_norm": 0.1867518126964569, "learning_rate": 9.404510212310671e-05, "loss": 0.0242, "step": 19910 }, { "grad_norm": 0.1938522905111313, "learning_rate": 9.403727386492924e-05, "loss": 0.0231, "step": 19920 }, { "grad_norm": 0.20682357251644135, "learning_rate": 9.40294407909017e-05, "loss": 0.0219, "step": 19930 }, { "grad_norm": 0.37606191635131836, "learning_rate": 9.40216029018807e-05, "loss": 0.0217, "step": 19940 }, { "grad_norm": 0.22459867596626282, "learning_rate": 9.401376019872338e-05, "loss": 0.0192, "step": 19950 }, { "grad_norm": 0.26934534311294556, "learning_rate": 9.400591268228746e-05, "loss": 0.0235, "step": 19960 }, { "grad_norm": 0.2357480227947235, "learning_rate": 9.399806035343106e-05, "loss": 0.0223, "step": 19970 }, { "grad_norm": 0.24796903133392334, "learning_rate": 9.399020321301294e-05, "loss": 0.0227, "step": 19980 }, { "grad_norm": 0.17918619513511658, "learning_rate": 9.398234126189234e-05, "loss": 0.0206, "step": 19990 }, { "grad_norm": 0.27356788516044617, "learning_rate": 9.397447450092902e-05, "loss": 0.0222, "step": 20000 }, { "grad_norm": 0.16773129999637604, "learning_rate": 9.39666029309833e-05, "loss": 0.0178, "step": 20010 }, { "grad_norm": 0.20665320754051208, "learning_rate": 9.395872655291596e-05, "loss": 0.0184, "step": 20020 }, { "grad_norm": 0.1789788156747818, "learning_rate": 9.395084536758838e-05, "loss": 0.0185, "step": 20030 }, { "grad_norm": 0.1930507868528366, "learning_rate": 9.394295937586243e-05, "loss": 0.034, "step": 20040 }, { "grad_norm": 0.18272492289543152, "learning_rate": 9.393506857860052e-05, "loss": 0.0237, "step": 20050 }, { "grad_norm": 0.2593914866447449, "learning_rate": 9.392717297666555e-05, "loss": 0.0166, "step": 20060 }, { "grad_norm": 0.281372994184494, "learning_rate": 9.391927257092101e-05, "loss": 0.0258, "step": 20070 }, { "grad_norm": 0.23871104419231415, "learning_rate": 9.391136736223085e-05, "loss": 0.019, "step": 20080 }, { "grad_norm": 0.2351095974445343, "learning_rate": 9.390345735145956e-05, "loss": 0.0259, "step": 20090 }, { "grad_norm": 0.17148229479789734, "learning_rate": 9.389554253947219e-05, "loss": 0.0196, "step": 20100 }, { "grad_norm": 0.13271325826644897, "learning_rate": 9.388762292713428e-05, "loss": 0.0181, "step": 20110 }, { "grad_norm": 0.19829830527305603, "learning_rate": 9.38796985153119e-05, "loss": 0.0196, "step": 20120 }, { "grad_norm": 0.1962597817182541, "learning_rate": 9.387176930487169e-05, "loss": 0.0201, "step": 20130 }, { "grad_norm": 0.15197737514972687, "learning_rate": 9.386383529668072e-05, "loss": 0.0164, "step": 20140 }, { "grad_norm": 0.16212593019008636, "learning_rate": 9.385589649160669e-05, "loss": 0.0198, "step": 20150 }, { "grad_norm": 0.17508605122566223, "learning_rate": 9.384795289051775e-05, "loss": 0.0214, "step": 20160 }, { "grad_norm": 0.19436796009540558, "learning_rate": 9.384000449428261e-05, "loss": 0.0263, "step": 20170 }, { "grad_norm": 0.16120901703834534, "learning_rate": 9.383205130377048e-05, "loss": 0.0251, "step": 20180 }, { "grad_norm": 0.16783468425273895, "learning_rate": 9.382409331985114e-05, "loss": 0.0183, "step": 20190 }, { "grad_norm": 0.15793700516223907, "learning_rate": 9.381613054339482e-05, "loss": 0.0183, "step": 20200 }, { "grad_norm": 0.1657642424106598, "learning_rate": 9.380816297527235e-05, "loss": 0.0195, "step": 20210 }, { "grad_norm": 0.19121761620044708, "learning_rate": 9.380019061635506e-05, "loss": 0.0192, "step": 20220 }, { "grad_norm": 0.21163304150104523, "learning_rate": 9.379221346751474e-05, "loss": 0.0244, "step": 20230 }, { "grad_norm": 0.25589174032211304, "learning_rate": 9.378423152962382e-05, "loss": 0.0196, "step": 20240 }, { "grad_norm": 0.24294735491275787, "learning_rate": 9.377624480355517e-05, "loss": 0.0194, "step": 20250 }, { "grad_norm": 0.14369501173496246, "learning_rate": 9.376825329018219e-05, "loss": 0.0167, "step": 20260 }, { "grad_norm": 0.16655583679676056, "learning_rate": 9.376025699037884e-05, "loss": 0.0161, "step": 20270 }, { "grad_norm": 0.1654357761144638, "learning_rate": 9.37522559050196e-05, "loss": 0.0161, "step": 20280 }, { "grad_norm": 0.19327622652053833, "learning_rate": 9.37442500349794e-05, "loss": 0.0248, "step": 20290 }, { "grad_norm": 0.1974593698978424, "learning_rate": 9.373623938113381e-05, "loss": 0.0174, "step": 20300 }, { "grad_norm": 0.18151913583278656, "learning_rate": 9.372822394435883e-05, "loss": 0.0232, "step": 20310 }, { "grad_norm": 0.21548116207122803, "learning_rate": 9.372020372553102e-05, "loss": 0.0193, "step": 20320 }, { "grad_norm": 0.22210273146629333, "learning_rate": 9.371217872552746e-05, "loss": 0.0176, "step": 20330 }, { "grad_norm": 0.1823291778564453, "learning_rate": 9.370414894522576e-05, "loss": 0.0213, "step": 20340 }, { "grad_norm": 0.2880922853946686, "learning_rate": 9.369611438550406e-05, "loss": 0.0227, "step": 20350 }, { "grad_norm": 0.18619664013385773, "learning_rate": 9.368807504724095e-05, "loss": 0.0191, "step": 20360 }, { "grad_norm": 0.19161809980869293, "learning_rate": 9.368003093131565e-05, "loss": 0.0235, "step": 20370 }, { "grad_norm": 0.22790586948394775, "learning_rate": 9.367198203860785e-05, "loss": 0.0318, "step": 20380 }, { "grad_norm": 0.16284625232219696, "learning_rate": 9.366392836999774e-05, "loss": 0.0194, "step": 20390 }, { "grad_norm": 0.22455663979053497, "learning_rate": 9.365586992636607e-05, "loss": 0.0185, "step": 20400 }, { "grad_norm": 0.23400649428367615, "learning_rate": 9.364780670859412e-05, "loss": 0.0202, "step": 20410 }, { "grad_norm": 0.22954674065113068, "learning_rate": 9.363973871756364e-05, "loss": 0.0172, "step": 20420 }, { "grad_norm": 0.19351281225681305, "learning_rate": 9.363166595415696e-05, "loss": 0.0159, "step": 20430 }, { "grad_norm": 0.2954654097557068, "learning_rate": 9.362358841925686e-05, "loss": 0.0226, "step": 20440 }, { "grad_norm": 0.1912432163953781, "learning_rate": 9.361550611374674e-05, "loss": 0.0211, "step": 20450 }, { "grad_norm": 0.17646053433418274, "learning_rate": 9.360741903851043e-05, "loss": 0.0348, "step": 20460 }, { "grad_norm": 0.14645333588123322, "learning_rate": 9.359932719443236e-05, "loss": 0.03, "step": 20470 }, { "grad_norm": 0.1980893313884735, "learning_rate": 9.35912305823974e-05, "loss": 0.0216, "step": 20480 }, { "grad_norm": 0.24533431231975555, "learning_rate": 9.358312920329101e-05, "loss": 0.0167, "step": 20490 }, { "grad_norm": 0.17367152869701385, "learning_rate": 9.357502305799914e-05, "loss": 0.0194, "step": 20500 }, { "grad_norm": 0.16247901320457458, "learning_rate": 9.356691214740824e-05, "loss": 0.0182, "step": 20510 }, { "grad_norm": 0.1664699912071228, "learning_rate": 9.355879647240535e-05, "loss": 0.0183, "step": 20520 }, { "grad_norm": 0.16490013897418976, "learning_rate": 9.355067603387798e-05, "loss": 0.0235, "step": 20530 }, { "grad_norm": 0.26734310388565063, "learning_rate": 9.354255083271412e-05, "loss": 0.0184, "step": 20540 }, { "grad_norm": 0.18668639659881592, "learning_rate": 9.353442086980239e-05, "loss": 0.0222, "step": 20550 }, { "grad_norm": 0.2652212083339691, "learning_rate": 9.352628614603185e-05, "loss": 0.019, "step": 20560 }, { "grad_norm": 0.23877757787704468, "learning_rate": 9.351814666229209e-05, "loss": 0.0175, "step": 20570 }, { "grad_norm": 0.19684246182441711, "learning_rate": 9.351000241947324e-05, "loss": 0.0171, "step": 20580 }, { "grad_norm": 0.196128711104393, "learning_rate": 9.350185341846594e-05, "loss": 0.0205, "step": 20590 }, { "grad_norm": 0.1871023178100586, "learning_rate": 9.349369966016134e-05, "loss": 0.0233, "step": 20600 }, { "grad_norm": 0.22805273532867432, "learning_rate": 9.348554114545117e-05, "loss": 0.0245, "step": 20610 }, { "grad_norm": 0.20164775848388672, "learning_rate": 9.347737787522758e-05, "loss": 0.0183, "step": 20620 }, { "grad_norm": 0.23910211026668549, "learning_rate": 9.346920985038332e-05, "loss": 0.023, "step": 20630 }, { "grad_norm": 0.2700845003128052, "learning_rate": 9.346103707181162e-05, "loss": 0.0228, "step": 20640 }, { "grad_norm": 0.18236739933490753, "learning_rate": 9.345285954040626e-05, "loss": 0.0164, "step": 20650 }, { "grad_norm": 0.17116478085517883, "learning_rate": 9.34446772570615e-05, "loss": 0.0172, "step": 20660 }, { "grad_norm": 0.37792325019836426, "learning_rate": 9.343649022267214e-05, "loss": 0.0164, "step": 20670 }, { "grad_norm": 0.17081661522388458, "learning_rate": 9.342829843813353e-05, "loss": 0.021, "step": 20680 }, { "grad_norm": 0.23584116995334625, "learning_rate": 9.342010190434149e-05, "loss": 0.0196, "step": 20690 }, { "grad_norm": 0.16527606546878815, "learning_rate": 9.34119006221924e-05, "loss": 0.0217, "step": 20700 }, { "grad_norm": 0.2069733440876007, "learning_rate": 9.340369459258313e-05, "loss": 0.0187, "step": 20710 }, { "grad_norm": 0.22019167244434357, "learning_rate": 9.339548381641106e-05, "loss": 0.0276, "step": 20720 }, { "grad_norm": 0.13735400140285492, "learning_rate": 9.338726829457413e-05, "loss": 0.0294, "step": 20730 }, { "grad_norm": 0.1514410376548767, "learning_rate": 9.337904802797078e-05, "loss": 0.0166, "step": 20740 }, { "grad_norm": 0.29582470655441284, "learning_rate": 9.337082301749993e-05, "loss": 0.0267, "step": 20750 }, { "grad_norm": 0.2501949071884155, "learning_rate": 9.336259326406109e-05, "loss": 0.0236, "step": 20760 }, { "grad_norm": 0.18795007467269897, "learning_rate": 9.335435876855427e-05, "loss": 0.028, "step": 20770 }, { "grad_norm": 0.2540923058986664, "learning_rate": 9.334611953187994e-05, "loss": 0.0264, "step": 20780 }, { "grad_norm": 0.16818712651729584, "learning_rate": 9.333787555493914e-05, "loss": 0.018, "step": 20790 }, { "grad_norm": 0.17640803754329681, "learning_rate": 9.332962683863345e-05, "loss": 0.0287, "step": 20800 }, { "grad_norm": 0.18873097002506256, "learning_rate": 9.332137338386489e-05, "loss": 0.0184, "step": 20810 }, { "grad_norm": 0.2510127127170563, "learning_rate": 9.33131151915361e-05, "loss": 0.0273, "step": 20820 }, { "grad_norm": 0.1791241616010666, "learning_rate": 9.330485226255012e-05, "loss": 0.0195, "step": 20830 }, { "grad_norm": 0.22522425651550293, "learning_rate": 9.329658459781061e-05, "loss": 0.0228, "step": 20840 }, { "grad_norm": 0.140732079744339, "learning_rate": 9.328831219822172e-05, "loss": 0.0221, "step": 20850 }, { "grad_norm": 0.18015499413013458, "learning_rate": 9.328003506468808e-05, "loss": 0.0222, "step": 20860 }, { "grad_norm": 0.22941535711288452, "learning_rate": 9.327175319811488e-05, "loss": 0.0193, "step": 20870 }, { "grad_norm": 0.19409909844398499, "learning_rate": 9.326346659940781e-05, "loss": 0.0218, "step": 20880 }, { "grad_norm": 0.17597684264183044, "learning_rate": 9.325517526947308e-05, "loss": 0.0237, "step": 20890 }, { "grad_norm": 0.23588043451309204, "learning_rate": 9.32468792092174e-05, "loss": 0.0196, "step": 20900 }, { "grad_norm": 0.20567426085472107, "learning_rate": 9.323857841954803e-05, "loss": 0.0214, "step": 20910 }, { "grad_norm": 0.1584627777338028, "learning_rate": 9.323027290137276e-05, "loss": 0.0188, "step": 20920 }, { "grad_norm": 0.18377771973609924, "learning_rate": 9.322196265559981e-05, "loss": 0.0166, "step": 20930 }, { "grad_norm": 0.12716257572174072, "learning_rate": 9.321364768313803e-05, "loss": 0.0192, "step": 20940 }, { "grad_norm": 0.23110705614089966, "learning_rate": 9.32053279848967e-05, "loss": 0.0211, "step": 20950 }, { "grad_norm": 0.15788058936595917, "learning_rate": 9.319700356178567e-05, "loss": 0.0203, "step": 20960 }, { "grad_norm": 0.24555842578411102, "learning_rate": 9.318867441471527e-05, "loss": 0.0177, "step": 20970 }, { "grad_norm": 0.22471392154693604, "learning_rate": 9.318034054459637e-05, "loss": 0.017, "step": 20980 }, { "grad_norm": 0.17804822325706482, "learning_rate": 9.317200195234034e-05, "loss": 0.0169, "step": 20990 }, { "grad_norm": 0.2457440346479416, "learning_rate": 9.316365863885909e-05, "loss": 0.0214, "step": 21000 }, { "grad_norm": 0.18318839371204376, "learning_rate": 9.315531060506502e-05, "loss": 0.0252, "step": 21010 }, { "grad_norm": 0.18305648863315582, "learning_rate": 9.314695785187108e-05, "loss": 0.0174, "step": 21020 }, { "grad_norm": 0.1802230030298233, "learning_rate": 9.313860038019069e-05, "loss": 0.0215, "step": 21030 }, { "grad_norm": 0.23827607929706573, "learning_rate": 9.313023819093782e-05, "loss": 0.0201, "step": 21040 }, { "grad_norm": 0.23948360979557037, "learning_rate": 9.312187128502695e-05, "loss": 0.0211, "step": 21050 }, { "grad_norm": 0.20901338756084442, "learning_rate": 9.311349966337307e-05, "loss": 0.0269, "step": 21060 }, { "grad_norm": 0.19048941135406494, "learning_rate": 9.310512332689169e-05, "loss": 0.0189, "step": 21070 }, { "grad_norm": 0.17194898426532745, "learning_rate": 9.309674227649883e-05, "loss": 0.0179, "step": 21080 }, { "grad_norm": 0.1541508287191391, "learning_rate": 9.308835651311103e-05, "loss": 0.022, "step": 21090 }, { "grad_norm": 0.25723034143447876, "learning_rate": 9.307996603764533e-05, "loss": 0.024, "step": 21100 }, { "grad_norm": 0.2623042166233063, "learning_rate": 9.307157085101932e-05, "loss": 0.02, "step": 21110 }, { "grad_norm": 0.2130751609802246, "learning_rate": 9.306317095415109e-05, "loss": 0.0189, "step": 21120 }, { "grad_norm": 0.2442706823348999, "learning_rate": 9.305476634795922e-05, "loss": 0.0248, "step": 21130 }, { "grad_norm": 0.2499615103006363, "learning_rate": 9.304635703336284e-05, "loss": 0.0195, "step": 21140 }, { "grad_norm": 0.16582567989826202, "learning_rate": 9.303794301128157e-05, "loss": 0.0199, "step": 21150 }, { "grad_norm": 0.19124722480773926, "learning_rate": 9.302952428263555e-05, "loss": 0.0187, "step": 21160 }, { "grad_norm": 0.1906554102897644, "learning_rate": 9.302110084834545e-05, "loss": 0.0204, "step": 21170 }, { "grad_norm": 0.1928076148033142, "learning_rate": 9.301267270933245e-05, "loss": 0.0166, "step": 21180 }, { "grad_norm": 0.18178482353687286, "learning_rate": 9.300423986651823e-05, "loss": 0.0192, "step": 21190 }, { "grad_norm": 0.21049202978610992, "learning_rate": 9.299580232082501e-05, "loss": 0.0198, "step": 21200 }, { "grad_norm": 0.165618434548378, "learning_rate": 9.298736007317547e-05, "loss": 0.0171, "step": 21210 }, { "grad_norm": 0.18796277046203613, "learning_rate": 9.297891312449288e-05, "loss": 0.0203, "step": 21220 }, { "grad_norm": 0.18785196542739868, "learning_rate": 9.297046147570094e-05, "loss": 0.0191, "step": 21230 }, { "grad_norm": 0.20919442176818848, "learning_rate": 9.296200512772396e-05, "loss": 0.0173, "step": 21240 }, { "grad_norm": 0.14448121190071106, "learning_rate": 9.295354408148668e-05, "loss": 0.0172, "step": 21250 }, { "grad_norm": 0.2271040380001068, "learning_rate": 9.294507833791441e-05, "loss": 0.0206, "step": 21260 }, { "grad_norm": 0.1672740876674652, "learning_rate": 9.293660789793295e-05, "loss": 0.0213, "step": 21270 }, { "grad_norm": 0.209634467959404, "learning_rate": 9.292813276246858e-05, "loss": 0.0204, "step": 21280 }, { "grad_norm": 0.19341576099395752, "learning_rate": 9.291965293244816e-05, "loss": 0.0194, "step": 21290 }, { "grad_norm": 0.1389055699110031, "learning_rate": 9.291116840879904e-05, "loss": 0.0169, "step": 21300 }, { "grad_norm": 0.21248936653137207, "learning_rate": 9.290267919244904e-05, "loss": 0.0179, "step": 21310 }, { "grad_norm": 0.16966648399829865, "learning_rate": 9.289418528432655e-05, "loss": 0.017, "step": 21320 }, { "grad_norm": 0.2336961328983307, "learning_rate": 9.288568668536045e-05, "loss": 0.0228, "step": 21330 }, { "grad_norm": 0.17590267956256866, "learning_rate": 9.287718339648013e-05, "loss": 0.0166, "step": 21340 }, { "grad_norm": 0.19452328979969025, "learning_rate": 9.28686754186155e-05, "loss": 0.0209, "step": 21350 }, { "grad_norm": 0.19847644865512848, "learning_rate": 9.286016275269698e-05, "loss": 0.0224, "step": 21360 }, { "grad_norm": 0.14867675304412842, "learning_rate": 9.285164539965551e-05, "loss": 0.0189, "step": 21370 }, { "grad_norm": 0.19329734146595, "learning_rate": 9.284312336042251e-05, "loss": 0.0164, "step": 21380 }, { "grad_norm": 0.19825606048107147, "learning_rate": 9.283459663592996e-05, "loss": 0.0172, "step": 21390 }, { "grad_norm": 0.26759952306747437, "learning_rate": 9.282606522711033e-05, "loss": 0.0228, "step": 21400 }, { "grad_norm": 0.14848944544792175, "learning_rate": 9.281752913489657e-05, "loss": 0.0267, "step": 21410 }, { "grad_norm": 0.15773902833461761, "learning_rate": 9.280898836022222e-05, "loss": 0.0203, "step": 21420 }, { "grad_norm": 0.18906599283218384, "learning_rate": 9.280044290402126e-05, "loss": 0.0183, "step": 21430 }, { "grad_norm": 0.14833855628967285, "learning_rate": 9.279189276722821e-05, "loss": 0.0195, "step": 21440 }, { "grad_norm": 0.1936030387878418, "learning_rate": 9.278333795077812e-05, "loss": 0.027, "step": 21450 }, { "grad_norm": 0.1916305273771286, "learning_rate": 9.27747784556065e-05, "loss": 0.0171, "step": 21460 }, { "grad_norm": 0.2989467978477478, "learning_rate": 9.276621428264942e-05, "loss": 0.028, "step": 21470 }, { "grad_norm": 0.23002012073993683, "learning_rate": 9.275764543284345e-05, "loss": 0.0185, "step": 21480 }, { "grad_norm": 0.20802202820777893, "learning_rate": 9.274907190712566e-05, "loss": 0.0199, "step": 21490 }, { "grad_norm": 0.2475631833076477, "learning_rate": 9.274049370643363e-05, "loss": 0.0198, "step": 21500 }, { "grad_norm": 0.16903415322303772, "learning_rate": 9.273191083170547e-05, "loss": 0.023, "step": 21510 }, { "grad_norm": 0.14116618037223816, "learning_rate": 9.27233232838798e-05, "loss": 0.0179, "step": 21520 }, { "grad_norm": 0.14634332060813904, "learning_rate": 9.27147310638957e-05, "loss": 0.0161, "step": 21530 }, { "grad_norm": 0.18158915638923645, "learning_rate": 9.270613417269286e-05, "loss": 0.0189, "step": 21540 }, { "grad_norm": 0.2492484748363495, "learning_rate": 9.269753261121138e-05, "loss": 0.0191, "step": 21550 }, { "grad_norm": 0.18451805412769318, "learning_rate": 9.268892638039194e-05, "loss": 0.0184, "step": 21560 }, { "grad_norm": 0.14461731910705566, "learning_rate": 9.268031548117569e-05, "loss": 0.0167, "step": 21570 }, { "grad_norm": 0.1852138787508011, "learning_rate": 9.26716999145043e-05, "loss": 0.0168, "step": 21580 }, { "grad_norm": 0.24372442066669464, "learning_rate": 9.266307968131998e-05, "loss": 0.0227, "step": 21590 }, { "grad_norm": 0.20454934239387512, "learning_rate": 9.26544547825654e-05, "loss": 0.0195, "step": 21600 }, { "grad_norm": 0.23756451904773712, "learning_rate": 9.264582521918376e-05, "loss": 0.0249, "step": 21610 }, { "grad_norm": 0.15158694982528687, "learning_rate": 9.263719099211881e-05, "loss": 0.0191, "step": 21620 }, { "grad_norm": 0.15016613900661469, "learning_rate": 9.262855210231476e-05, "loss": 0.0214, "step": 21630 }, { "grad_norm": 0.2341250330209732, "learning_rate": 9.261990855071633e-05, "loss": 0.0173, "step": 21640 }, { "grad_norm": 0.12801240384578705, "learning_rate": 9.261126033826878e-05, "loss": 0.0164, "step": 21650 }, { "grad_norm": 0.21185122430324554, "learning_rate": 9.260260746591786e-05, "loss": 0.0211, "step": 21660 }, { "grad_norm": 0.19476765394210815, "learning_rate": 9.259394993460985e-05, "loss": 0.0162, "step": 21670 }, { "grad_norm": 0.1745961457490921, "learning_rate": 9.258528774529151e-05, "loss": 0.0163, "step": 21680 }, { "grad_norm": 0.21025288105010986, "learning_rate": 9.257662089891013e-05, "loss": 0.0185, "step": 21690 }, { "grad_norm": 0.19654956459999084, "learning_rate": 9.25679493964135e-05, "loss": 0.0192, "step": 21700 }, { "grad_norm": 0.16840031743049622, "learning_rate": 9.255927323874994e-05, "loss": 0.0169, "step": 21710 }, { "grad_norm": 0.28832781314849854, "learning_rate": 9.255059242686822e-05, "loss": 0.0222, "step": 21720 }, { "grad_norm": 0.28465235233306885, "learning_rate": 9.254190696171769e-05, "loss": 0.0155, "step": 21730 }, { "grad_norm": 0.1485668420791626, "learning_rate": 9.25332168442482e-05, "loss": 0.0193, "step": 21740 }, { "grad_norm": 0.14029891788959503, "learning_rate": 9.252452207541004e-05, "loss": 0.0156, "step": 21750 }, { "grad_norm": 0.24865630269050598, "learning_rate": 9.251582265615409e-05, "loss": 0.0289, "step": 21760 }, { "grad_norm": 0.23123407363891602, "learning_rate": 9.250711858743169e-05, "loss": 0.0214, "step": 21770 }, { "grad_norm": 0.23432190716266632, "learning_rate": 9.24984098701947e-05, "loss": 0.0318, "step": 21780 }, { "grad_norm": 0.2510402798652649, "learning_rate": 9.248969650539552e-05, "loss": 0.0187, "step": 21790 }, { "grad_norm": 0.238267183303833, "learning_rate": 9.2480978493987e-05, "loss": 0.0249, "step": 21800 }, { "grad_norm": 0.15230421721935272, "learning_rate": 9.247225583692256e-05, "loss": 0.0151, "step": 21810 }, { "grad_norm": 0.14405380189418793, "learning_rate": 9.246352853515607e-05, "loss": 0.0156, "step": 21820 }, { "grad_norm": 0.29475903511047363, "learning_rate": 9.245479658964194e-05, "loss": 0.0222, "step": 21830 }, { "grad_norm": 0.18106190860271454, "learning_rate": 9.244606000133507e-05, "loss": 0.0182, "step": 21840 }, { "grad_norm": 0.19876623153686523, "learning_rate": 9.24373187711909e-05, "loss": 0.0178, "step": 21850 }, { "grad_norm": 0.2513927221298218, "learning_rate": 9.242857290016537e-05, "loss": 0.0225, "step": 21860 }, { "grad_norm": 0.178123340010643, "learning_rate": 9.241982238921488e-05, "loss": 0.0201, "step": 21870 }, { "grad_norm": 0.2386052906513214, "learning_rate": 9.24110672392964e-05, "loss": 0.0151, "step": 21880 }, { "grad_norm": 0.15954844653606415, "learning_rate": 9.240230745136737e-05, "loss": 0.0159, "step": 21890 }, { "grad_norm": 0.20471912622451782, "learning_rate": 9.239354302638575e-05, "loss": 0.016, "step": 21900 }, { "grad_norm": 0.2741299867630005, "learning_rate": 9.238477396531e-05, "loss": 0.0229, "step": 21910 }, { "grad_norm": 0.2273901104927063, "learning_rate": 9.23760002690991e-05, "loss": 0.0199, "step": 21920 }, { "grad_norm": 0.22330348193645477, "learning_rate": 9.236722193871252e-05, "loss": 0.0166, "step": 21930 }, { "grad_norm": 0.18766598403453827, "learning_rate": 9.235843897511023e-05, "loss": 0.0198, "step": 21940 }, { "grad_norm": 0.1225069910287857, "learning_rate": 9.234965137925276e-05, "loss": 0.015, "step": 21950 }, { "grad_norm": 0.17375211417675018, "learning_rate": 9.234085915210108e-05, "loss": 0.0169, "step": 21960 }, { "grad_norm": 0.152937114238739, "learning_rate": 9.23320622946167e-05, "loss": 0.0169, "step": 21970 }, { "grad_norm": 0.2090960443019867, "learning_rate": 9.232326080776163e-05, "loss": 0.0165, "step": 21980 }, { "grad_norm": 0.1887703537940979, "learning_rate": 9.23144546924984e-05, "loss": 0.0174, "step": 21990 }, { "grad_norm": 0.13931767642498016, "learning_rate": 9.230564394979e-05, "loss": 0.0174, "step": 22000 }, { "grad_norm": 0.17711196839809418, "learning_rate": 9.22968285806e-05, "loss": 0.0163, "step": 22010 }, { "grad_norm": 0.18638087809085846, "learning_rate": 9.228800858589242e-05, "loss": 0.0214, "step": 22020 }, { "grad_norm": 0.15894246101379395, "learning_rate": 9.227918396663179e-05, "loss": 0.0202, "step": 22030 }, { "grad_norm": 0.14324849843978882, "learning_rate": 9.227035472378319e-05, "loss": 0.0191, "step": 22040 }, { "grad_norm": 0.1554497629404068, "learning_rate": 9.226152085831213e-05, "loss": 0.0168, "step": 22050 }, { "grad_norm": 0.25168275833129883, "learning_rate": 9.22526823711847e-05, "loss": 0.0165, "step": 22060 }, { "grad_norm": 0.19068343937397003, "learning_rate": 9.224383926336745e-05, "loss": 0.017, "step": 22070 }, { "grad_norm": 0.18420244753360748, "learning_rate": 9.223499153582744e-05, "loss": 0.0152, "step": 22080 }, { "grad_norm": 0.1813991516828537, "learning_rate": 9.222613918953226e-05, "loss": 0.0243, "step": 22090 }, { "grad_norm": 0.20359590649604797, "learning_rate": 9.221728222544999e-05, "loss": 0.0183, "step": 22100 }, { "grad_norm": 0.2705857753753662, "learning_rate": 9.22084206445492e-05, "loss": 0.0266, "step": 22110 }, { "grad_norm": 0.19486083090305328, "learning_rate": 9.2199554447799e-05, "loss": 0.0302, "step": 22120 }, { "grad_norm": 0.17836415767669678, "learning_rate": 9.219068363616897e-05, "loss": 0.0222, "step": 22130 }, { "grad_norm": 0.18544583022594452, "learning_rate": 9.218180821062919e-05, "loss": 0.0194, "step": 22140 }, { "grad_norm": 0.20761561393737793, "learning_rate": 9.21729281721503e-05, "loss": 0.0197, "step": 22150 }, { "grad_norm": 0.1467994898557663, "learning_rate": 9.216404352170339e-05, "loss": 0.0168, "step": 22160 }, { "grad_norm": 0.23118633031845093, "learning_rate": 9.215515426026007e-05, "loss": 0.0143, "step": 22170 }, { "grad_norm": 0.2331753373146057, "learning_rate": 9.214626038879246e-05, "loss": 0.0161, "step": 22180 }, { "grad_norm": 0.2031596153974533, "learning_rate": 9.21373619082732e-05, "loss": 0.0163, "step": 22190 }, { "grad_norm": 0.1476721316576004, "learning_rate": 9.212845881967535e-05, "loss": 0.0169, "step": 22200 }, { "grad_norm": 0.152994304895401, "learning_rate": 9.211955112397262e-05, "loss": 0.0196, "step": 22210 }, { "grad_norm": 0.14489765465259552, "learning_rate": 9.211063882213909e-05, "loss": 0.0178, "step": 22220 }, { "grad_norm": 0.11610966920852661, "learning_rate": 9.210172191514942e-05, "loss": 0.0137, "step": 22230 }, { "grad_norm": 0.18106770515441895, "learning_rate": 9.209280040397874e-05, "loss": 0.0213, "step": 22240 }, { "grad_norm": 0.20216695964336395, "learning_rate": 9.208387428960268e-05, "loss": 0.0136, "step": 22250 }, { "grad_norm": 0.26958686113357544, "learning_rate": 9.20749435729974e-05, "loss": 0.0203, "step": 22260 }, { "grad_norm": 0.1871519684791565, "learning_rate": 9.206600825513957e-05, "loss": 0.0151, "step": 22270 }, { "grad_norm": 0.22110305726528168, "learning_rate": 9.20570683370063e-05, "loss": 0.0195, "step": 22280 }, { "grad_norm": 0.1737510710954666, "learning_rate": 9.204812381957528e-05, "loss": 0.0183, "step": 22290 }, { "grad_norm": 0.17654713988304138, "learning_rate": 9.203917470382465e-05, "loss": 0.0161, "step": 22300 }, { "grad_norm": 0.20323534309864044, "learning_rate": 9.203022099073309e-05, "loss": 0.0184, "step": 22310 }, { "grad_norm": 0.26166802644729614, "learning_rate": 9.202126268127976e-05, "loss": 0.0223, "step": 22320 }, { "grad_norm": 0.1868041306734085, "learning_rate": 9.20122997764443e-05, "loss": 0.0161, "step": 22330 }, { "grad_norm": 0.18471793830394745, "learning_rate": 9.200333227720692e-05, "loss": 0.0233, "step": 22340 }, { "grad_norm": 0.1377231329679489, "learning_rate": 9.199436018454826e-05, "loss": 0.0209, "step": 22350 }, { "grad_norm": 0.13916097581386566, "learning_rate": 9.198538349944952e-05, "loss": 0.0158, "step": 22360 }, { "grad_norm": 0.22149424254894257, "learning_rate": 9.197640222289234e-05, "loss": 0.0266, "step": 22370 }, { "grad_norm": 0.1765255481004715, "learning_rate": 9.196741635585895e-05, "loss": 0.0191, "step": 22380 }, { "grad_norm": 0.1499384045600891, "learning_rate": 9.195842589933199e-05, "loss": 0.0136, "step": 22390 }, { "grad_norm": 0.3121516704559326, "learning_rate": 9.194943085429466e-05, "loss": 0.0231, "step": 22400 }, { "grad_norm": 0.27690109610557556, "learning_rate": 9.194043122173065e-05, "loss": 0.032, "step": 22410 }, { "grad_norm": 0.2684941589832306, "learning_rate": 9.193142700262413e-05, "loss": 0.0241, "step": 22420 }, { "grad_norm": 0.16347646713256836, "learning_rate": 9.192241819795979e-05, "loss": 0.0173, "step": 22430 }, { "grad_norm": 0.14754223823547363, "learning_rate": 9.191340480872284e-05, "loss": 0.0172, "step": 22440 }, { "grad_norm": 0.23404763638973236, "learning_rate": 9.190438683589895e-05, "loss": 0.022, "step": 22450 }, { "grad_norm": 0.2210860401391983, "learning_rate": 9.189536428047432e-05, "loss": 0.0196, "step": 22460 }, { "grad_norm": 0.20582592487335205, "learning_rate": 9.188633714343564e-05, "loss": 0.0194, "step": 22470 }, { "grad_norm": 0.3006993234157562, "learning_rate": 9.18773054257701e-05, "loss": 0.0192, "step": 22480 }, { "grad_norm": 0.2145913541316986, "learning_rate": 9.18682691284654e-05, "loss": 0.0183, "step": 22490 }, { "grad_norm": 0.16720381379127502, "learning_rate": 9.185922825250974e-05, "loss": 0.0161, "step": 22500 }, { "grad_norm": 0.19680173695087433, "learning_rate": 9.185018279889181e-05, "loss": 0.0142, "step": 22510 }, { "grad_norm": 0.15216225385665894, "learning_rate": 9.184113276860082e-05, "loss": 0.0208, "step": 22520 }, { "grad_norm": 0.2592465877532959, "learning_rate": 9.183207816262645e-05, "loss": 0.0247, "step": 22530 }, { "grad_norm": 0.1468537598848343, "learning_rate": 9.182301898195891e-05, "loss": 0.0151, "step": 22540 }, { "grad_norm": 0.1296308934688568, "learning_rate": 9.181395522758889e-05, "loss": 0.0231, "step": 22550 }, { "grad_norm": 0.3095874488353729, "learning_rate": 9.180488690050759e-05, "loss": 0.0303, "step": 22560 }, { "grad_norm": 0.18264806270599365, "learning_rate": 9.179581400170671e-05, "loss": 0.0194, "step": 22570 }, { "grad_norm": 0.20848488807678223, "learning_rate": 9.178673653217845e-05, "loss": 0.0167, "step": 22580 }, { "grad_norm": 0.18933211266994476, "learning_rate": 9.177765449291551e-05, "loss": 0.0154, "step": 22590 }, { "grad_norm": 0.18803425133228302, "learning_rate": 9.176856788491109e-05, "loss": 0.0172, "step": 22600 }, { "grad_norm": 0.20412184298038483, "learning_rate": 9.175947670915887e-05, "loss": 0.0166, "step": 22610 }, { "grad_norm": 0.1807786524295807, "learning_rate": 9.175038096665309e-05, "loss": 0.0191, "step": 22620 }, { "grad_norm": 0.18988709151744843, "learning_rate": 9.17412806583884e-05, "loss": 0.0171, "step": 22630 }, { "grad_norm": 0.20780351758003235, "learning_rate": 9.173217578536002e-05, "loss": 0.0264, "step": 22640 }, { "grad_norm": 0.2668604850769043, "learning_rate": 9.172306634856362e-05, "loss": 0.0215, "step": 22650 }, { "grad_norm": 0.2212231308221817, "learning_rate": 9.171395234899545e-05, "loss": 0.0206, "step": 22660 }, { "grad_norm": 0.20054201781749725, "learning_rate": 9.170483378765214e-05, "loss": 0.0176, "step": 22670 }, { "grad_norm": 0.21266242861747742, "learning_rate": 9.169571066553091e-05, "loss": 0.0182, "step": 22680 }, { "grad_norm": 0.20836341381072998, "learning_rate": 9.168658298362946e-05, "loss": 0.0169, "step": 22690 }, { "grad_norm": 0.23093266785144806, "learning_rate": 9.167745074294598e-05, "loss": 0.0172, "step": 22700 }, { "grad_norm": 0.28520262241363525, "learning_rate": 9.166831394447913e-05, "loss": 0.0278, "step": 22710 }, { "grad_norm": 0.20910769701004028, "learning_rate": 9.165917258922812e-05, "loss": 0.0157, "step": 22720 }, { "grad_norm": 0.16426561772823334, "learning_rate": 9.165002667819262e-05, "loss": 0.0178, "step": 22730 }, { "grad_norm": 0.1876697838306427, "learning_rate": 9.164087621237282e-05, "loss": 0.0172, "step": 22740 }, { "grad_norm": 0.23442009091377258, "learning_rate": 9.163172119276942e-05, "loss": 0.0255, "step": 22750 }, { "grad_norm": 0.24424952268600464, "learning_rate": 9.162256162038358e-05, "loss": 0.0259, "step": 22760 }, { "grad_norm": 0.16080202162265778, "learning_rate": 9.161339749621698e-05, "loss": 0.0189, "step": 22770 }, { "grad_norm": 0.13187721371650696, "learning_rate": 9.160422882127177e-05, "loss": 0.0167, "step": 22780 }, { "grad_norm": 0.1800008863210678, "learning_rate": 9.159505559655069e-05, "loss": 0.0153, "step": 22790 }, { "grad_norm": 0.14862298965454102, "learning_rate": 9.158587782305684e-05, "loss": 0.0144, "step": 22800 }, { "grad_norm": 0.1932631880044937, "learning_rate": 9.157669550179391e-05, "loss": 0.0194, "step": 22810 }, { "grad_norm": 0.1980815976858139, "learning_rate": 9.156750863376609e-05, "loss": 0.0235, "step": 22820 }, { "grad_norm": 0.19837094843387604, "learning_rate": 9.155831721997801e-05, "loss": 0.0181, "step": 22830 }, { "grad_norm": 0.17484217882156372, "learning_rate": 9.154912126143484e-05, "loss": 0.0219, "step": 22840 }, { "grad_norm": 0.2687036693096161, "learning_rate": 9.153992075914224e-05, "loss": 0.0325, "step": 22850 }, { "grad_norm": 0.14511172473430634, "learning_rate": 9.153071571410635e-05, "loss": 0.0165, "step": 22860 }, { "grad_norm": 0.24973535537719727, "learning_rate": 9.152150612733384e-05, "loss": 0.0182, "step": 22870 }, { "grad_norm": 0.19910705089569092, "learning_rate": 9.151229199983184e-05, "loss": 0.0232, "step": 22880 }, { "grad_norm": 0.1700160950422287, "learning_rate": 9.150307333260802e-05, "loss": 0.0177, "step": 22890 }, { "grad_norm": 0.21578605473041534, "learning_rate": 9.149385012667048e-05, "loss": 0.0273, "step": 22900 }, { "grad_norm": 0.2327551394701004, "learning_rate": 9.148462238302788e-05, "loss": 0.0212, "step": 22910 }, { "grad_norm": 0.25745686888694763, "learning_rate": 9.147539010268936e-05, "loss": 0.0228, "step": 22920 }, { "grad_norm": 0.15802320837974548, "learning_rate": 9.14661532866645e-05, "loss": 0.0174, "step": 22930 }, { "grad_norm": 0.15069973468780518, "learning_rate": 9.145691193596348e-05, "loss": 0.0152, "step": 22940 }, { "grad_norm": 0.24291469156742096, "learning_rate": 9.144766605159691e-05, "loss": 0.0285, "step": 22950 }, { "grad_norm": 0.2131718099117279, "learning_rate": 9.14384156345759e-05, "loss": 0.0169, "step": 22960 }, { "grad_norm": 0.24095749855041504, "learning_rate": 9.142916068591204e-05, "loss": 0.0169, "step": 22970 }, { "grad_norm": 0.1825617104768753, "learning_rate": 9.141990120661746e-05, "loss": 0.0178, "step": 22980 }, { "grad_norm": 0.27057692408561707, "learning_rate": 9.141063719770475e-05, "loss": 0.0213, "step": 22990 }, { "grad_norm": 0.19974032044410706, "learning_rate": 9.140136866018704e-05, "loss": 0.0179, "step": 23000 }, { "grad_norm": 0.16890691220760345, "learning_rate": 9.139209559507788e-05, "loss": 0.0173, "step": 23010 }, { "grad_norm": 0.19410942494869232, "learning_rate": 9.13828180033914e-05, "loss": 0.0174, "step": 23020 }, { "grad_norm": 0.17868269979953766, "learning_rate": 9.137353588614212e-05, "loss": 0.0209, "step": 23030 }, { "grad_norm": 0.16867958009243011, "learning_rate": 9.136424924434519e-05, "loss": 0.0161, "step": 23040 }, { "grad_norm": 0.2646947205066681, "learning_rate": 9.135495807901615e-05, "loss": 0.0264, "step": 23050 }, { "grad_norm": 0.16622988879680634, "learning_rate": 9.134566239117108e-05, "loss": 0.0186, "step": 23060 }, { "grad_norm": 0.20909784734249115, "learning_rate": 9.13363621818265e-05, "loss": 0.0159, "step": 23070 }, { "grad_norm": 0.27152952551841736, "learning_rate": 9.132705745199953e-05, "loss": 0.0164, "step": 23080 }, { "grad_norm": 0.2324083149433136, "learning_rate": 9.131774820270768e-05, "loss": 0.0163, "step": 23090 }, { "grad_norm": 0.20653462409973145, "learning_rate": 9.130843443496901e-05, "loss": 0.0208, "step": 23100 }, { "grad_norm": 0.20811861753463745, "learning_rate": 9.129911614980206e-05, "loss": 0.0223, "step": 23110 }, { "grad_norm": 0.17294523119926453, "learning_rate": 9.128979334822584e-05, "loss": 0.0204, "step": 23120 }, { "grad_norm": 0.16808198392391205, "learning_rate": 9.128046603125992e-05, "loss": 0.0177, "step": 23130 }, { "grad_norm": 0.16452623903751373, "learning_rate": 9.12711341999243e-05, "loss": 0.0146, "step": 23140 }, { "grad_norm": 0.15865711867809296, "learning_rate": 9.12617978552395e-05, "loss": 0.0329, "step": 23150 }, { "grad_norm": 0.15851321816444397, "learning_rate": 9.12524569982265e-05, "loss": 0.0174, "step": 23160 }, { "grad_norm": 0.19674049317836761, "learning_rate": 9.124311162990684e-05, "loss": 0.0211, "step": 23170 }, { "grad_norm": 0.1609267145395279, "learning_rate": 9.12337617513025e-05, "loss": 0.0154, "step": 23180 }, { "grad_norm": 0.16716186702251434, "learning_rate": 9.122440736343596e-05, "loss": 0.0207, "step": 23190 }, { "grad_norm": 0.19012007117271423, "learning_rate": 9.12150484673302e-05, "loss": 0.0187, "step": 23200 }, { "grad_norm": 0.22036780416965485, "learning_rate": 9.120568506400873e-05, "loss": 0.0164, "step": 23210 }, { "grad_norm": 0.1597166657447815, "learning_rate": 9.119631715449548e-05, "loss": 0.0171, "step": 23220 }, { "grad_norm": 0.2483896166086197, "learning_rate": 9.118694473981493e-05, "loss": 0.019, "step": 23230 }, { "grad_norm": 0.2552315294742584, "learning_rate": 9.117756782099203e-05, "loss": 0.0172, "step": 23240 }, { "grad_norm": 0.1393563598394394, "learning_rate": 9.11681863990522e-05, "loss": 0.0192, "step": 23250 }, { "grad_norm": 0.14476029574871063, "learning_rate": 9.115880047502142e-05, "loss": 0.0162, "step": 23260 }, { "grad_norm": 0.24585705995559692, "learning_rate": 9.114941004992609e-05, "loss": 0.0189, "step": 23270 }, { "grad_norm": 0.21069397032260895, "learning_rate": 9.114001512479317e-05, "loss": 0.0193, "step": 23280 }, { "grad_norm": 0.15886159241199493, "learning_rate": 9.113061570065003e-05, "loss": 0.027, "step": 23290 }, { "grad_norm": 0.2104758769273758, "learning_rate": 9.112121177852459e-05, "loss": 0.0211, "step": 23300 }, { "grad_norm": 0.26653987169265747, "learning_rate": 9.111180335944527e-05, "loss": 0.0244, "step": 23310 }, { "grad_norm": 0.36207470297813416, "learning_rate": 9.110239044444093e-05, "loss": 0.02, "step": 23320 }, { "grad_norm": 0.19349397718906403, "learning_rate": 9.109297303454099e-05, "loss": 0.0187, "step": 23330 }, { "grad_norm": 0.238480806350708, "learning_rate": 9.108355113077526e-05, "loss": 0.0212, "step": 23340 }, { "grad_norm": 0.19843225181102753, "learning_rate": 9.107412473417419e-05, "loss": 0.0197, "step": 23350 }, { "grad_norm": 0.22838947176933289, "learning_rate": 9.106469384576858e-05, "loss": 0.017, "step": 23360 }, { "grad_norm": 0.15735605359077454, "learning_rate": 9.105525846658978e-05, "loss": 0.0225, "step": 23370 }, { "grad_norm": 0.20236994326114655, "learning_rate": 9.104581859766965e-05, "loss": 0.0251, "step": 23380 }, { "grad_norm": 0.1632431447505951, "learning_rate": 9.10363742400405e-05, "loss": 0.028, "step": 23390 }, { "grad_norm": 0.18756943941116333, "learning_rate": 9.102692539473518e-05, "loss": 0.0288, "step": 23400 }, { "grad_norm": 0.19873179495334625, "learning_rate": 9.101747206278697e-05, "loss": 0.0281, "step": 23410 }, { "grad_norm": 0.16395531594753265, "learning_rate": 9.100801424522968e-05, "loss": 0.0155, "step": 23420 }, { "grad_norm": 0.13927863538265228, "learning_rate": 9.099855194309762e-05, "loss": 0.0199, "step": 23430 }, { "grad_norm": 0.27597835659980774, "learning_rate": 9.098908515742554e-05, "loss": 0.0178, "step": 23440 }, { "grad_norm": 0.22373682260513306, "learning_rate": 9.097961388924873e-05, "loss": 0.0193, "step": 23450 }, { "grad_norm": 0.17502053081989288, "learning_rate": 9.097013813960298e-05, "loss": 0.0161, "step": 23460 }, { "grad_norm": 0.1598334163427353, "learning_rate": 9.09606579095245e-05, "loss": 0.0174, "step": 23470 }, { "grad_norm": 0.1984207034111023, "learning_rate": 9.095117320005008e-05, "loss": 0.0161, "step": 23480 }, { "grad_norm": 0.22936038672924042, "learning_rate": 9.094168401221691e-05, "loss": 0.0181, "step": 23490 }, { "grad_norm": 0.2261216938495636, "learning_rate": 9.093219034706273e-05, "loss": 0.0291, "step": 23500 }, { "grad_norm": 0.1744571477174759, "learning_rate": 9.092269220562577e-05, "loss": 0.0227, "step": 23510 }, { "grad_norm": 0.2645356357097626, "learning_rate": 9.09131895889447e-05, "loss": 0.0386, "step": 23520 }, { "grad_norm": 0.174827441573143, "learning_rate": 9.090368249805873e-05, "loss": 0.0178, "step": 23530 }, { "grad_norm": 0.19259478151798248, "learning_rate": 9.089417093400754e-05, "loss": 0.0216, "step": 23540 }, { "grad_norm": 0.16138985753059387, "learning_rate": 9.088465489783131e-05, "loss": 0.0177, "step": 23550 }, { "grad_norm": 0.20739826560020447, "learning_rate": 9.087513439057068e-05, "loss": 0.0169, "step": 23560 }, { "grad_norm": 0.161650612950325, "learning_rate": 9.08656094132668e-05, "loss": 0.0161, "step": 23570 }, { "grad_norm": 0.19606131315231323, "learning_rate": 9.085607996696134e-05, "loss": 0.0179, "step": 23580 }, { "grad_norm": 0.2341902107000351, "learning_rate": 9.084654605269639e-05, "loss": 0.0249, "step": 23590 }, { "grad_norm": 0.19989560544490814, "learning_rate": 9.083700767151457e-05, "loss": 0.0206, "step": 23600 }, { "grad_norm": 0.13260649144649506, "learning_rate": 9.082746482445898e-05, "loss": 0.0215, "step": 23610 }, { "grad_norm": 0.17786578834056854, "learning_rate": 9.081791751257325e-05, "loss": 0.0177, "step": 23620 }, { "grad_norm": 0.12228477001190186, "learning_rate": 9.080836573690142e-05, "loss": 0.015, "step": 23630 }, { "grad_norm": 0.19081413745880127, "learning_rate": 9.079880949848805e-05, "loss": 0.0231, "step": 23640 }, { "grad_norm": 0.1815364807844162, "learning_rate": 9.078924879837822e-05, "loss": 0.0185, "step": 23650 }, { "grad_norm": 0.13912904262542725, "learning_rate": 9.077968363761747e-05, "loss": 0.016, "step": 23660 }, { "grad_norm": 0.12205670773983002, "learning_rate": 9.077011401725182e-05, "loss": 0.0169, "step": 23670 }, { "grad_norm": 0.2164659947156906, "learning_rate": 9.07605399383278e-05, "loss": 0.0234, "step": 23680 }, { "grad_norm": 0.15701033174991608, "learning_rate": 9.075096140189243e-05, "loss": 0.0175, "step": 23690 }, { "grad_norm": 0.21436841785907745, "learning_rate": 9.074137840899318e-05, "loss": 0.0197, "step": 23700 }, { "grad_norm": 0.1852630078792572, "learning_rate": 9.073179096067804e-05, "loss": 0.0168, "step": 23710 }, { "grad_norm": 0.14577029645442963, "learning_rate": 9.072219905799549e-05, "loss": 0.016, "step": 23720 }, { "grad_norm": 0.2571631073951721, "learning_rate": 9.071260270199447e-05, "loss": 0.022, "step": 23730 }, { "grad_norm": 0.14433658123016357, "learning_rate": 9.070300189372441e-05, "loss": 0.0142, "step": 23740 }, { "grad_norm": 0.12259168177843094, "learning_rate": 9.069339663423528e-05, "loss": 0.0163, "step": 23750 }, { "grad_norm": 0.16163718700408936, "learning_rate": 9.068378692457747e-05, "loss": 0.0179, "step": 23760 }, { "grad_norm": 0.1728564351797104, "learning_rate": 9.067417276580189e-05, "loss": 0.0169, "step": 23770 }, { "grad_norm": 0.18361175060272217, "learning_rate": 9.066455415895993e-05, "loss": 0.0179, "step": 23780 }, { "grad_norm": 0.1970680207014084, "learning_rate": 9.065493110510346e-05, "loss": 0.017, "step": 23790 }, { "grad_norm": 0.12873108685016632, "learning_rate": 9.064530360528484e-05, "loss": 0.0185, "step": 23800 }, { "grad_norm": 0.21262627840042114, "learning_rate": 9.063567166055695e-05, "loss": 0.0172, "step": 23810 }, { "grad_norm": 0.23616577684879303, "learning_rate": 9.062603527197308e-05, "loss": 0.0166, "step": 23820 }, { "grad_norm": 0.16796435415744781, "learning_rate": 9.06163944405871e-05, "loss": 0.0151, "step": 23830 }, { "grad_norm": 0.1990448534488678, "learning_rate": 9.060674916745327e-05, "loss": 0.0181, "step": 23840 }, { "grad_norm": 0.13098594546318054, "learning_rate": 9.05970994536264e-05, "loss": 0.0147, "step": 23850 }, { "grad_norm": 0.18464666604995728, "learning_rate": 9.05874453001618e-05, "loss": 0.0157, "step": 23860 }, { "grad_norm": 0.24149183928966522, "learning_rate": 9.057778670811517e-05, "loss": 0.0209, "step": 23870 }, { "grad_norm": 0.14514707028865814, "learning_rate": 9.056812367854281e-05, "loss": 0.0152, "step": 23880 }, { "grad_norm": 0.18690091371536255, "learning_rate": 9.055845621250143e-05, "loss": 0.02, "step": 23890 }, { "grad_norm": 0.14485836029052734, "learning_rate": 9.054878431104825e-05, "loss": 0.0152, "step": 23900 }, { "grad_norm": 0.18769098818302155, "learning_rate": 9.0539107975241e-05, "loss": 0.0173, "step": 23910 }, { "grad_norm": 0.26756376028060913, "learning_rate": 9.052942720613784e-05, "loss": 0.023, "step": 23920 }, { "grad_norm": 0.14779187738895416, "learning_rate": 9.051974200479745e-05, "loss": 0.0168, "step": 23930 }, { "grad_norm": 0.22796356678009033, "learning_rate": 9.051005237227901e-05, "loss": 0.0146, "step": 23940 }, { "grad_norm": 0.13877025246620178, "learning_rate": 9.050035830964215e-05, "loss": 0.0133, "step": 23950 }, { "grad_norm": 0.24649794399738312, "learning_rate": 9.049065981794698e-05, "loss": 0.028, "step": 23960 }, { "grad_norm": 0.17536067962646484, "learning_rate": 9.048095689825414e-05, "loss": 0.0216, "step": 23970 }, { "grad_norm": 0.16051287949085236, "learning_rate": 9.047124955162472e-05, "loss": 0.018, "step": 23980 }, { "grad_norm": 0.29613274335861206, "learning_rate": 9.046153777912028e-05, "loss": 0.023, "step": 23990 }, { "grad_norm": 0.23229335248470306, "learning_rate": 9.045182158180292e-05, "loss": 0.0175, "step": 24000 }, { "grad_norm": 0.15965892374515533, "learning_rate": 9.044210096073516e-05, "loss": 0.0144, "step": 24010 }, { "grad_norm": 0.1714218407869339, "learning_rate": 9.043237591698004e-05, "loss": 0.0167, "step": 24020 }, { "grad_norm": 0.1918952316045761, "learning_rate": 9.04226464516011e-05, "loss": 0.0177, "step": 24030 }, { "grad_norm": 0.162299245595932, "learning_rate": 9.041291256566229e-05, "loss": 0.0171, "step": 24040 }, { "grad_norm": 0.15629799664020538, "learning_rate": 9.040317426022814e-05, "loss": 0.0179, "step": 24050 }, { "grad_norm": 0.1266280561685562, "learning_rate": 9.03934315363636e-05, "loss": 0.0204, "step": 24060 }, { "grad_norm": 0.20374645292758942, "learning_rate": 9.038368439513409e-05, "loss": 0.0144, "step": 24070 }, { "grad_norm": 0.16944991052150726, "learning_rate": 9.03739328376056e-05, "loss": 0.0211, "step": 24080 }, { "grad_norm": 0.14050449430942535, "learning_rate": 9.036417686484451e-05, "loss": 0.0146, "step": 24090 }, { "grad_norm": 0.2021453082561493, "learning_rate": 9.035441647791773e-05, "loss": 0.0207, "step": 24100 }, { "grad_norm": 0.15372030436992645, "learning_rate": 9.034465167789263e-05, "loss": 0.0172, "step": 24110 }, { "grad_norm": 0.1780446618795395, "learning_rate": 9.033488246583706e-05, "loss": 0.0154, "step": 24120 }, { "grad_norm": 0.1723700612783432, "learning_rate": 9.032510884281941e-05, "loss": 0.0217, "step": 24130 }, { "grad_norm": 0.2148437350988388, "learning_rate": 9.031533080990848e-05, "loss": 0.0214, "step": 24140 }, { "grad_norm": 0.2570120096206665, "learning_rate": 9.030554836817358e-05, "loss": 0.0171, "step": 24150 }, { "grad_norm": 0.14762242138385773, "learning_rate": 9.029576151868451e-05, "loss": 0.0175, "step": 24160 }, { "grad_norm": 0.2128148376941681, "learning_rate": 9.028597026251155e-05, "loss": 0.0156, "step": 24170 }, { "grad_norm": 0.16285988688468933, "learning_rate": 9.027617460072547e-05, "loss": 0.0172, "step": 24180 }, { "grad_norm": 0.15372449159622192, "learning_rate": 9.026637453439745e-05, "loss": 0.0147, "step": 24190 }, { "grad_norm": 0.17068952322006226, "learning_rate": 9.025657006459927e-05, "loss": 0.0149, "step": 24200 }, { "grad_norm": 0.18208716809749603, "learning_rate": 9.024676119240311e-05, "loss": 0.0146, "step": 24210 }, { "grad_norm": 0.14436310529708862, "learning_rate": 9.023694791888166e-05, "loss": 0.0214, "step": 24220 }, { "grad_norm": 0.12392332404851913, "learning_rate": 9.022713024510808e-05, "loss": 0.0209, "step": 24230 }, { "grad_norm": 0.13402004539966583, "learning_rate": 9.021730817215601e-05, "loss": 0.0164, "step": 24240 }, { "grad_norm": 0.3217780590057373, "learning_rate": 9.02074817010996e-05, "loss": 0.0195, "step": 24250 }, { "grad_norm": 0.2505490183830261, "learning_rate": 9.019765083301342e-05, "loss": 0.0166, "step": 24260 }, { "grad_norm": 0.19195754826068878, "learning_rate": 9.01878155689726e-05, "loss": 0.0165, "step": 24270 }, { "grad_norm": 0.17708593606948853, "learning_rate": 9.017797591005268e-05, "loss": 0.0174, "step": 24280 }, { "grad_norm": 0.23773004114627838, "learning_rate": 9.016813185732972e-05, "loss": 0.0222, "step": 24290 }, { "grad_norm": 0.16064880788326263, "learning_rate": 9.015828341188027e-05, "loss": 0.0225, "step": 24300 }, { "grad_norm": 0.2611379623413086, "learning_rate": 9.01484305747813e-05, "loss": 0.0218, "step": 24310 }, { "grad_norm": 0.21715478599071503, "learning_rate": 9.013857334711033e-05, "loss": 0.0184, "step": 24320 }, { "grad_norm": 0.13578101992607117, "learning_rate": 9.012871172994534e-05, "loss": 0.0136, "step": 24330 }, { "grad_norm": 0.18397103250026703, "learning_rate": 9.011884572436476e-05, "loss": 0.0187, "step": 24340 }, { "grad_norm": 0.20191577076911926, "learning_rate": 9.010897533144754e-05, "loss": 0.0236, "step": 24350 }, { "grad_norm": 0.17645400762557983, "learning_rate": 9.009910055227306e-05, "loss": 0.0162, "step": 24360 }, { "grad_norm": 0.14829197525978088, "learning_rate": 9.008922138792124e-05, "loss": 0.0168, "step": 24370 }, { "grad_norm": 0.14968927204608917, "learning_rate": 9.007933783947244e-05, "loss": 0.0138, "step": 24380 }, { "grad_norm": 0.10152511298656464, "learning_rate": 9.006944990800752e-05, "loss": 0.0139, "step": 24390 }, { "grad_norm": 0.22755788266658783, "learning_rate": 9.005955759460779e-05, "loss": 0.0184, "step": 24400 }, { "grad_norm": 0.16177473962306976, "learning_rate": 9.004966090035508e-05, "loss": 0.0147, "step": 24410 }, { "grad_norm": 0.17411571741104126, "learning_rate": 9.003975982633166e-05, "loss": 0.0168, "step": 24420 }, { "grad_norm": 0.22396357357501984, "learning_rate": 9.00298543736203e-05, "loss": 0.0189, "step": 24430 }, { "grad_norm": 0.11658065021038055, "learning_rate": 9.001994454330427e-05, "loss": 0.0124, "step": 24440 }, { "grad_norm": 0.19674213230609894, "learning_rate": 9.001003033646727e-05, "loss": 0.0196, "step": 24450 }, { "grad_norm": 0.17812176048755646, "learning_rate": 9.00001117541935e-05, "loss": 0.0179, "step": 24460 }, { "grad_norm": 0.16127145290374756, "learning_rate": 8.999018879756764e-05, "loss": 0.019, "step": 24470 }, { "grad_norm": 0.19321531057357788, "learning_rate": 8.998026146767487e-05, "loss": 0.0183, "step": 24480 }, { "grad_norm": 0.2208574116230011, "learning_rate": 8.99703297656008e-05, "loss": 0.0161, "step": 24490 }, { "grad_norm": 0.209761843085289, "learning_rate": 8.996039369243156e-05, "loss": 0.0199, "step": 24500 }, { "grad_norm": 0.19455984234809875, "learning_rate": 8.995045324925378e-05, "loss": 0.0152, "step": 24510 }, { "grad_norm": 0.23489077389240265, "learning_rate": 8.994050843715448e-05, "loss": 0.024, "step": 24520 }, { "grad_norm": 0.2187676876783371, "learning_rate": 8.993055925722121e-05, "loss": 0.016, "step": 24530 }, { "grad_norm": 0.21822284162044525, "learning_rate": 8.992060571054202e-05, "loss": 0.0181, "step": 24540 }, { "grad_norm": 0.17635169625282288, "learning_rate": 8.991064779820542e-05, "loss": 0.0189, "step": 24550 }, { "grad_norm": 0.25991982221603394, "learning_rate": 8.990068552130036e-05, "loss": 0.0144, "step": 24560 }, { "grad_norm": 0.15178583562374115, "learning_rate": 8.989071888091634e-05, "loss": 0.0152, "step": 24570 }, { "grad_norm": 0.21590174734592438, "learning_rate": 8.988074787814329e-05, "loss": 0.0162, "step": 24580 }, { "grad_norm": 0.17482855916023254, "learning_rate": 8.987077251407158e-05, "loss": 0.0155, "step": 24590 }, { "grad_norm": 0.12584562599658966, "learning_rate": 8.986079278979216e-05, "loss": 0.0157, "step": 24600 }, { "grad_norm": 0.19709157943725586, "learning_rate": 8.985080870639635e-05, "loss": 0.0231, "step": 24610 }, { "grad_norm": 0.16950874030590057, "learning_rate": 8.984082026497603e-05, "loss": 0.0154, "step": 24620 }, { "grad_norm": 0.21281279623508453, "learning_rate": 8.98308274666235e-05, "loss": 0.0268, "step": 24630 }, { "grad_norm": 0.18329782783985138, "learning_rate": 8.982083031243155e-05, "loss": 0.0208, "step": 24640 }, { "grad_norm": 0.25644490122795105, "learning_rate": 8.98108288034935e-05, "loss": 0.0179, "step": 24650 }, { "grad_norm": 0.17095035314559937, "learning_rate": 8.980082294090305e-05, "loss": 0.0211, "step": 24660 }, { "grad_norm": 0.14468777179718018, "learning_rate": 8.979081272575443e-05, "loss": 0.0201, "step": 24670 }, { "grad_norm": 0.21249239146709442, "learning_rate": 8.978079815914236e-05, "loss": 0.0185, "step": 24680 }, { "grad_norm": 0.2005542665719986, "learning_rate": 8.977077924216202e-05, "loss": 0.0179, "step": 24690 }, { "grad_norm": 0.2664791941642761, "learning_rate": 8.976075597590905e-05, "loss": 0.0218, "step": 24700 }, { "grad_norm": 0.24527104198932648, "learning_rate": 8.975072836147958e-05, "loss": 0.0191, "step": 24710 }, { "grad_norm": 0.17671065032482147, "learning_rate": 8.974069639997025e-05, "loss": 0.0175, "step": 24720 }, { "grad_norm": 0.1389608532190323, "learning_rate": 8.973066009247808e-05, "loss": 0.0175, "step": 24730 }, { "grad_norm": 0.14262276887893677, "learning_rate": 8.972061944010066e-05, "loss": 0.0197, "step": 24740 }, { "grad_norm": 0.17349213361740112, "learning_rate": 8.971057444393603e-05, "loss": 0.0198, "step": 24750 }, { "grad_norm": 0.1583123505115509, "learning_rate": 8.970052510508268e-05, "loss": 0.0228, "step": 24760 }, { "grad_norm": 0.18600797653198242, "learning_rate": 8.969047142463959e-05, "loss": 0.0197, "step": 24770 }, { "grad_norm": 0.1991412490606308, "learning_rate": 8.968041340370621e-05, "loss": 0.0193, "step": 24780 }, { "grad_norm": 0.18784697353839874, "learning_rate": 8.96703510433825e-05, "loss": 0.0177, "step": 24790 }, { "grad_norm": 0.11684083193540573, "learning_rate": 8.966028434476883e-05, "loss": 0.0137, "step": 24800 }, { "grad_norm": 0.195394828915596, "learning_rate": 8.96502133089661e-05, "loss": 0.0171, "step": 24810 }, { "grad_norm": 0.174282506108284, "learning_rate": 8.964013793707564e-05, "loss": 0.0167, "step": 24820 }, { "grad_norm": 0.18514452874660492, "learning_rate": 8.963005823019932e-05, "loss": 0.0192, "step": 24830 }, { "grad_norm": 0.2543599009513855, "learning_rate": 8.961997418943939e-05, "loss": 0.0176, "step": 24840 }, { "grad_norm": 0.16863664984703064, "learning_rate": 8.960988581589865e-05, "loss": 0.0159, "step": 24850 }, { "grad_norm": 0.1307218223810196, "learning_rate": 8.959979311068037e-05, "loss": 0.0139, "step": 24860 }, { "grad_norm": 0.21024653315544128, "learning_rate": 8.958969607488823e-05, "loss": 0.0241, "step": 24870 }, { "grad_norm": 0.14385563135147095, "learning_rate": 8.957959470962647e-05, "loss": 0.0147, "step": 24880 }, { "grad_norm": 0.14673635363578796, "learning_rate": 8.956948901599971e-05, "loss": 0.0166, "step": 24890 }, { "grad_norm": 0.2385166734457016, "learning_rate": 8.955937899511315e-05, "loss": 0.0217, "step": 24900 }, { "grad_norm": 0.20930688083171844, "learning_rate": 8.954926464807238e-05, "loss": 0.0169, "step": 24910 }, { "grad_norm": 0.15653590857982635, "learning_rate": 8.953914597598347e-05, "loss": 0.0146, "step": 24920 }, { "grad_norm": 0.11979705095291138, "learning_rate": 8.952902297995303e-05, "loss": 0.0151, "step": 24930 }, { "grad_norm": 0.17347308993339539, "learning_rate": 8.951889566108804e-05, "loss": 0.0167, "step": 24940 }, { "grad_norm": 0.20781934261322021, "learning_rate": 8.950876402049606e-05, "loss": 0.0197, "step": 24950 }, { "grad_norm": 0.18887479603290558, "learning_rate": 8.949862805928504e-05, "loss": 0.0168, "step": 24960 }, { "grad_norm": 0.1453150063753128, "learning_rate": 8.948848777856343e-05, "loss": 0.0181, "step": 24970 }, { "grad_norm": 0.2330017387866974, "learning_rate": 8.947834317944017e-05, "loss": 0.0162, "step": 24980 }, { "grad_norm": 0.18935111165046692, "learning_rate": 8.946819426302466e-05, "loss": 0.0318, "step": 24990 }, { "grad_norm": 0.1860373318195343, "learning_rate": 8.945804103042676e-05, "loss": 0.0179, "step": 25000 }, { "grad_norm": 0.14593058824539185, "learning_rate": 8.944788348275681e-05, "loss": 0.0198, "step": 25010 }, { "grad_norm": 0.14841598272323608, "learning_rate": 8.943772162112565e-05, "loss": 0.0179, "step": 25020 }, { "grad_norm": 0.19075189530849457, "learning_rate": 8.942755544664454e-05, "loss": 0.0275, "step": 25030 }, { "grad_norm": 0.20726639032363892, "learning_rate": 8.941738496042525e-05, "loss": 0.0164, "step": 25040 }, { "grad_norm": 0.18612809479236603, "learning_rate": 8.940721016357999e-05, "loss": 0.0164, "step": 25050 }, { "grad_norm": 0.22919858992099762, "learning_rate": 8.939703105722148e-05, "loss": 0.0203, "step": 25060 }, { "grad_norm": 0.14697431027889252, "learning_rate": 8.93868476424629e-05, "loss": 0.0171, "step": 25070 }, { "grad_norm": 0.15558326244354248, "learning_rate": 8.937665992041786e-05, "loss": 0.0179, "step": 25080 }, { "grad_norm": 0.27562403678894043, "learning_rate": 8.93664678922005e-05, "loss": 0.0205, "step": 25090 }, { "grad_norm": 0.1473923772573471, "learning_rate": 8.93562715589254e-05, "loss": 0.0173, "step": 25100 }, { "grad_norm": 0.1695377230644226, "learning_rate": 8.934607092170762e-05, "loss": 0.0183, "step": 25110 }, { "grad_norm": 0.18143992125988007, "learning_rate": 8.933586598166266e-05, "loss": 0.0189, "step": 25120 }, { "grad_norm": 0.2476176768541336, "learning_rate": 8.932565673990655e-05, "loss": 0.0336, "step": 25130 }, { "grad_norm": 0.1654650717973709, "learning_rate": 8.931544319755574e-05, "loss": 0.0197, "step": 25140 }, { "grad_norm": 0.16911858320236206, "learning_rate": 8.930522535572718e-05, "loss": 0.018, "step": 25150 }, { "grad_norm": 0.14800065755844116, "learning_rate": 8.929500321553826e-05, "loss": 0.0174, "step": 25160 }, { "grad_norm": 0.2499002069234848, "learning_rate": 8.928477677810686e-05, "loss": 0.0139, "step": 25170 }, { "grad_norm": 0.18443942070007324, "learning_rate": 8.927454604455137e-05, "loss": 0.0168, "step": 25180 }, { "grad_norm": 0.21258299052715302, "learning_rate": 8.926431101599053e-05, "loss": 0.0203, "step": 25190 }, { "grad_norm": 0.12800520658493042, "learning_rate": 8.925407169354369e-05, "loss": 0.0136, "step": 25200 }, { "grad_norm": 0.18343490362167358, "learning_rate": 8.92438280783306e-05, "loss": 0.0202, "step": 25210 }, { "grad_norm": 0.13716013729572296, "learning_rate": 8.923358017147146e-05, "loss": 0.0152, "step": 25220 }, { "grad_norm": 0.23163361847400665, "learning_rate": 8.922332797408697e-05, "loss": 0.0203, "step": 25230 }, { "grad_norm": 0.16558195650577545, "learning_rate": 8.921307148729831e-05, "loss": 0.0158, "step": 25240 }, { "grad_norm": 0.13365332782268524, "learning_rate": 8.920281071222712e-05, "loss": 0.0134, "step": 25250 }, { "grad_norm": 0.13370876014232635, "learning_rate": 8.919254564999548e-05, "loss": 0.0157, "step": 25260 }, { "grad_norm": 0.15216602385044098, "learning_rate": 8.918227630172598e-05, "loss": 0.0163, "step": 25270 }, { "grad_norm": 0.1314457356929779, "learning_rate": 8.917200266854165e-05, "loss": 0.0171, "step": 25280 }, { "grad_norm": 0.13430571556091309, "learning_rate": 8.9161724751566e-05, "loss": 0.0172, "step": 25290 }, { "grad_norm": 0.23165956139564514, "learning_rate": 8.915144255192302e-05, "loss": 0.0177, "step": 25300 }, { "grad_norm": 0.22297388315200806, "learning_rate": 8.914115607073714e-05, "loss": 0.0176, "step": 25310 }, { "grad_norm": 0.13964055478572845, "learning_rate": 8.913086530913327e-05, "loss": 0.0136, "step": 25320 }, { "grad_norm": 0.13756617903709412, "learning_rate": 8.912057026823681e-05, "loss": 0.0215, "step": 25330 }, { "grad_norm": 0.1496930718421936, "learning_rate": 8.91102709491736e-05, "loss": 0.0133, "step": 25340 }, { "grad_norm": 0.15638408064842224, "learning_rate": 8.909996735306996e-05, "loss": 0.0172, "step": 25350 }, { "grad_norm": 0.15519647300243378, "learning_rate": 8.908965948105268e-05, "loss": 0.014, "step": 25360 }, { "grad_norm": 0.21956871449947357, "learning_rate": 8.907934733424901e-05, "loss": 0.0183, "step": 25370 }, { "grad_norm": 0.11499602347612381, "learning_rate": 8.906903091378666e-05, "loss": 0.0144, "step": 25380 }, { "grad_norm": 0.14825540781021118, "learning_rate": 8.905871022079384e-05, "loss": 0.0146, "step": 25390 }, { "grad_norm": 0.13802145421504974, "learning_rate": 8.90483852563992e-05, "loss": 0.0139, "step": 25400 }, { "grad_norm": 0.20752154290676117, "learning_rate": 8.903805602173185e-05, "loss": 0.0173, "step": 25410 }, { "grad_norm": 0.1903468519449234, "learning_rate": 8.902772251792137e-05, "loss": 0.0143, "step": 25420 }, { "grad_norm": 0.2712548077106476, "learning_rate": 8.901738474609786e-05, "loss": 0.021, "step": 25430 }, { "grad_norm": 0.20049893856048584, "learning_rate": 8.900704270739179e-05, "loss": 0.0178, "step": 25440 }, { "grad_norm": 0.15945494174957275, "learning_rate": 8.89966964029342e-05, "loss": 0.0242, "step": 25450 }, { "grad_norm": 0.17735660076141357, "learning_rate": 8.898634583385652e-05, "loss": 0.0165, "step": 25460 }, { "grad_norm": 0.2550681531429291, "learning_rate": 8.897599100129065e-05, "loss": 0.0236, "step": 25470 }, { "grad_norm": 0.18982575833797455, "learning_rate": 8.896563190636903e-05, "loss": 0.0162, "step": 25480 }, { "grad_norm": 0.12189222872257233, "learning_rate": 8.895526855022448e-05, "loss": 0.0168, "step": 25490 }, { "grad_norm": 0.2406451404094696, "learning_rate": 8.894490093399033e-05, "loss": 0.0191, "step": 25500 }, { "grad_norm": 0.1767163723707199, "learning_rate": 8.893452905880035e-05, "loss": 0.0175, "step": 25510 }, { "grad_norm": 0.20445062220096588, "learning_rate": 8.892415292578883e-05, "loss": 0.0167, "step": 25520 }, { "grad_norm": 0.27104395627975464, "learning_rate": 8.891377253609046e-05, "loss": 0.021, "step": 25530 }, { "grad_norm": 0.31009650230407715, "learning_rate": 8.890338789084043e-05, "loss": 0.0197, "step": 25540 }, { "grad_norm": 0.20822733640670776, "learning_rate": 8.88929989911744e-05, "loss": 0.0176, "step": 25550 }, { "grad_norm": 0.2281998246908188, "learning_rate": 8.888260583822847e-05, "loss": 0.0206, "step": 25560 }, { "grad_norm": 0.25072795152664185, "learning_rate": 8.887220843313921e-05, "loss": 0.0179, "step": 25570 }, { "grad_norm": 0.16847269237041473, "learning_rate": 8.88618067770437e-05, "loss": 0.0167, "step": 25580 }, { "grad_norm": 0.1750144064426422, "learning_rate": 8.885140087107942e-05, "loss": 0.0178, "step": 25590 }, { "grad_norm": 0.2021254301071167, "learning_rate": 8.884099071638436e-05, "loss": 0.015, "step": 25600 }, { "grad_norm": 0.16563007235527039, "learning_rate": 8.883057631409695e-05, "loss": 0.0178, "step": 25610 }, { "grad_norm": 0.1804523915052414, "learning_rate": 8.882015766535608e-05, "loss": 0.0193, "step": 25620 }, { "grad_norm": 0.23675641417503357, "learning_rate": 8.880973477130115e-05, "loss": 0.0193, "step": 25630 }, { "grad_norm": 0.1805536448955536, "learning_rate": 8.879930763307197e-05, "loss": 0.016, "step": 25640 }, { "grad_norm": 0.19610749185085297, "learning_rate": 8.878887625180884e-05, "loss": 0.0201, "step": 25650 }, { "grad_norm": 0.11199513077735901, "learning_rate": 8.877844062865253e-05, "loss": 0.0127, "step": 25660 }, { "grad_norm": 0.1557471603155136, "learning_rate": 8.876800076474424e-05, "loss": 0.0242, "step": 25670 }, { "grad_norm": 0.1860453337430954, "learning_rate": 8.875755666122568e-05, "loss": 0.015, "step": 25680 }, { "grad_norm": 0.15183889865875244, "learning_rate": 8.8747108319239e-05, "loss": 0.0159, "step": 25690 }, { "grad_norm": 0.17102673649787903, "learning_rate": 8.87366557399268e-05, "loss": 0.0144, "step": 25700 }, { "grad_norm": 0.1246008649468422, "learning_rate": 8.872619892443217e-05, "loss": 0.0166, "step": 25710 }, { "grad_norm": 0.12848247587680817, "learning_rate": 8.871573787389865e-05, "loss": 0.0132, "step": 25720 }, { "grad_norm": 0.2193766087293625, "learning_rate": 8.870527258947024e-05, "loss": 0.0198, "step": 25730 }, { "grad_norm": 0.15972720086574554, "learning_rate": 8.869480307229143e-05, "loss": 0.016, "step": 25740 }, { "grad_norm": 0.15396730601787567, "learning_rate": 8.868432932350712e-05, "loss": 0.015, "step": 25750 }, { "grad_norm": 0.17872364819049835, "learning_rate": 8.867385134426272e-05, "loss": 0.0172, "step": 25760 }, { "grad_norm": 0.14807520806789398, "learning_rate": 8.866336913570407e-05, "loss": 0.015, "step": 25770 }, { "grad_norm": 0.14843964576721191, "learning_rate": 8.865288269897751e-05, "loss": 0.0126, "step": 25780 }, { "grad_norm": 0.21521469950675964, "learning_rate": 8.864239203522981e-05, "loss": 0.0153, "step": 25790 }, { "grad_norm": 0.20176202058792114, "learning_rate": 8.863189714560822e-05, "loss": 0.0159, "step": 25800 }, { "grad_norm": 0.16733093559741974, "learning_rate": 8.862139803126043e-05, "loss": 0.0166, "step": 25810 }, { "grad_norm": 0.10239288955926895, "learning_rate": 8.861089469333463e-05, "loss": 0.0138, "step": 25820 }, { "grad_norm": 0.1631157547235489, "learning_rate": 8.860038713297944e-05, "loss": 0.0144, "step": 25830 }, { "grad_norm": 0.16232772171497345, "learning_rate": 8.858987535134394e-05, "loss": 0.0205, "step": 25840 }, { "grad_norm": 0.15302789211273193, "learning_rate": 8.857935934957769e-05, "loss": 0.0173, "step": 25850 }, { "grad_norm": 0.20602168142795563, "learning_rate": 8.856883912883071e-05, "loss": 0.0162, "step": 25860 }, { "grad_norm": 0.16720172762870789, "learning_rate": 8.855831469025346e-05, "loss": 0.0173, "step": 25870 }, { "grad_norm": 0.23637349903583527, "learning_rate": 8.854778603499689e-05, "loss": 0.0256, "step": 25880 }, { "grad_norm": 0.159295916557312, "learning_rate": 8.85372531642124e-05, "loss": 0.0156, "step": 25890 }, { "grad_norm": 0.2662118375301361, "learning_rate": 8.852671607905185e-05, "loss": 0.0131, "step": 25900 }, { "grad_norm": 0.12560349702835083, "learning_rate": 8.851617478066754e-05, "loss": 0.0188, "step": 25910 }, { "grad_norm": 0.21995797753334045, "learning_rate": 8.850562927021227e-05, "loss": 0.0169, "step": 25920 }, { "grad_norm": 0.14283107221126556, "learning_rate": 8.849507954883928e-05, "loss": 0.0161, "step": 25930 }, { "grad_norm": 0.14500883221626282, "learning_rate": 8.848452561770226e-05, "loss": 0.0121, "step": 25940 }, { "grad_norm": 0.13923892378807068, "learning_rate": 8.847396747795538e-05, "loss": 0.0184, "step": 25950 }, { "grad_norm": 0.1544426828622818, "learning_rate": 8.846340513075327e-05, "loss": 0.0188, "step": 25960 }, { "grad_norm": 0.20142999291419983, "learning_rate": 8.845283857725099e-05, "loss": 0.0167, "step": 25970 }, { "grad_norm": 0.2698022723197937, "learning_rate": 8.844226781860409e-05, "loss": 0.0148, "step": 25980 }, { "grad_norm": 0.1302185207605362, "learning_rate": 8.84316928559686e-05, "loss": 0.0162, "step": 25990 }, { "grad_norm": 0.2250986546278, "learning_rate": 8.842111369050094e-05, "loss": 0.0155, "step": 26000 }, { "grad_norm": 0.2642551064491272, "learning_rate": 8.841053032335808e-05, "loss": 0.0207, "step": 26010 }, { "grad_norm": 0.13746900856494904, "learning_rate": 8.839994275569735e-05, "loss": 0.0156, "step": 26020 }, { "grad_norm": 0.21201904118061066, "learning_rate": 8.838935098867662e-05, "loss": 0.0161, "step": 26030 }, { "grad_norm": 0.16374170780181885, "learning_rate": 8.837875502345418e-05, "loss": 0.0196, "step": 26040 }, { "grad_norm": 0.17317365109920502, "learning_rate": 8.83681548611888e-05, "loss": 0.0194, "step": 26050 }, { "grad_norm": 0.17162954807281494, "learning_rate": 8.835755050303969e-05, "loss": 0.0142, "step": 26060 }, { "grad_norm": 0.18145856261253357, "learning_rate": 8.834694195016653e-05, "loss": 0.0211, "step": 26070 }, { "grad_norm": 0.30045679211616516, "learning_rate": 8.833632920372942e-05, "loss": 0.0255, "step": 26080 }, { "grad_norm": 0.1682860255241394, "learning_rate": 8.832571226488903e-05, "loss": 0.0236, "step": 26090 }, { "grad_norm": 0.21332545578479767, "learning_rate": 8.831509113480634e-05, "loss": 0.0203, "step": 26100 }, { "grad_norm": 0.17162422835826874, "learning_rate": 8.83044658146429e-05, "loss": 0.0157, "step": 26110 }, { "grad_norm": 0.16761496663093567, "learning_rate": 8.829383630556067e-05, "loss": 0.0167, "step": 26120 }, { "grad_norm": 0.21270298957824707, "learning_rate": 8.828320260872207e-05, "loss": 0.0213, "step": 26130 }, { "grad_norm": 0.18677465617656708, "learning_rate": 8.827256472529e-05, "loss": 0.0184, "step": 26140 }, { "grad_norm": 0.13967163860797882, "learning_rate": 8.826192265642778e-05, "loss": 0.0269, "step": 26150 }, { "grad_norm": 0.1279754489660263, "learning_rate": 8.825127640329923e-05, "loss": 0.0154, "step": 26160 }, { "grad_norm": 0.1502099186182022, "learning_rate": 8.824062596706861e-05, "loss": 0.0174, "step": 26170 }, { "grad_norm": 0.1956000179052353, "learning_rate": 8.822997134890062e-05, "loss": 0.0168, "step": 26180 }, { "grad_norm": 0.17377358675003052, "learning_rate": 8.821931254996044e-05, "loss": 0.0165, "step": 26190 }, { "grad_norm": 0.19084493815898895, "learning_rate": 8.82086495714137e-05, "loss": 0.0181, "step": 26200 }, { "grad_norm": 0.2050647884607315, "learning_rate": 8.81979824144265e-05, "loss": 0.0203, "step": 26210 }, { "grad_norm": 0.2391202598810196, "learning_rate": 8.818731108016536e-05, "loss": 0.0189, "step": 26220 }, { "grad_norm": 0.27525272965431213, "learning_rate": 8.81766355697973e-05, "loss": 0.0172, "step": 26230 }, { "grad_norm": 0.24618473649024963, "learning_rate": 8.816595588448977e-05, "loss": 0.0151, "step": 26240 }, { "grad_norm": 0.18201950192451477, "learning_rate": 8.81552720254107e-05, "loss": 0.019, "step": 26250 }, { "grad_norm": 0.30717360973358154, "learning_rate": 8.814458399372842e-05, "loss": 0.0269, "step": 26260 }, { "grad_norm": 0.17434392869472504, "learning_rate": 8.813389179061181e-05, "loss": 0.0167, "step": 26270 }, { "grad_norm": 0.16825367510318756, "learning_rate": 8.812319541723012e-05, "loss": 0.0303, "step": 26280 }, { "grad_norm": 0.16077837347984314, "learning_rate": 8.811249487475309e-05, "loss": 0.0268, "step": 26290 }, { "grad_norm": 0.1528255045413971, "learning_rate": 8.810179016435092e-05, "loss": 0.025, "step": 26300 }, { "grad_norm": 0.24284841120243073, "learning_rate": 8.809108128719428e-05, "loss": 0.0322, "step": 26310 }, { "grad_norm": 0.16552487015724182, "learning_rate": 8.808036824445424e-05, "loss": 0.018, "step": 26320 }, { "grad_norm": 0.1459854245185852, "learning_rate": 8.806965103730238e-05, "loss": 0.0191, "step": 26330 }, { "grad_norm": 0.19966968894004822, "learning_rate": 8.805892966691074e-05, "loss": 0.0178, "step": 26340 }, { "grad_norm": 0.13159117102622986, "learning_rate": 8.804820413445175e-05, "loss": 0.0152, "step": 26350 }, { "grad_norm": 0.1744564026594162, "learning_rate": 8.803747444109837e-05, "loss": 0.0191, "step": 26360 }, { "grad_norm": 0.14258167147636414, "learning_rate": 8.802674058802399e-05, "loss": 0.0142, "step": 26370 }, { "grad_norm": 0.1351248174905777, "learning_rate": 8.801600257640241e-05, "loss": 0.0163, "step": 26380 }, { "grad_norm": 0.15068550407886505, "learning_rate": 8.800526040740795e-05, "loss": 0.0191, "step": 26390 }, { "grad_norm": 0.21998350322246552, "learning_rate": 8.799451408221535e-05, "loss": 0.0174, "step": 26400 }, { "grad_norm": 0.16556838154792786, "learning_rate": 8.798376360199982e-05, "loss": 0.0163, "step": 26410 }, { "grad_norm": 0.15701645612716675, "learning_rate": 8.797300896793701e-05, "loss": 0.0154, "step": 26420 }, { "grad_norm": 0.19840221107006073, "learning_rate": 8.796225018120302e-05, "loss": 0.0207, "step": 26430 }, { "grad_norm": 0.17919741570949554, "learning_rate": 8.795148724297444e-05, "loss": 0.0182, "step": 26440 }, { "grad_norm": 0.16408474743366241, "learning_rate": 8.794072015442825e-05, "loss": 0.0188, "step": 26450 }, { "grad_norm": 0.1401500403881073, "learning_rate": 8.792994891674198e-05, "loss": 0.0203, "step": 26460 }, { "grad_norm": 0.1908600926399231, "learning_rate": 8.79191735310935e-05, "loss": 0.0151, "step": 26470 }, { "grad_norm": 0.17226773500442505, "learning_rate": 8.790839399866122e-05, "loss": 0.014, "step": 26480 }, { "grad_norm": 0.1360037624835968, "learning_rate": 8.789761032062397e-05, "loss": 0.0177, "step": 26490 }, { "grad_norm": 0.2563905715942383, "learning_rate": 8.788682249816103e-05, "loss": 0.0177, "step": 26500 }, { "grad_norm": 0.1563989222049713, "learning_rate": 8.787603053245215e-05, "loss": 0.0265, "step": 26510 }, { "grad_norm": 0.15163247287273407, "learning_rate": 8.78652344246775e-05, "loss": 0.0176, "step": 26520 }, { "grad_norm": 0.150889590382576, "learning_rate": 8.785443417601776e-05, "loss": 0.017, "step": 26530 }, { "grad_norm": 0.18381769955158234, "learning_rate": 8.784362978765401e-05, "loss": 0.0177, "step": 26540 }, { "grad_norm": 0.2257203310728073, "learning_rate": 8.783282126076779e-05, "loss": 0.0163, "step": 26550 }, { "grad_norm": 0.12828876078128815, "learning_rate": 8.782200859654112e-05, "loss": 0.0184, "step": 26560 }, { "grad_norm": 0.10072185844182968, "learning_rate": 8.781119179615646e-05, "loss": 0.0139, "step": 26570 }, { "grad_norm": 0.17810013890266418, "learning_rate": 8.780037086079674e-05, "loss": 0.0152, "step": 26580 }, { "grad_norm": 0.21821285784244537, "learning_rate": 8.778954579164527e-05, "loss": 0.0171, "step": 26590 }, { "grad_norm": 0.1494636833667755, "learning_rate": 8.777871658988588e-05, "loss": 0.0168, "step": 26600 }, { "grad_norm": 0.1248231753706932, "learning_rate": 8.776788325670285e-05, "loss": 0.0156, "step": 26610 }, { "grad_norm": 0.1429128497838974, "learning_rate": 8.775704579328089e-05, "loss": 0.02, "step": 26620 }, { "grad_norm": 0.16110952198505402, "learning_rate": 8.774620420080517e-05, "loss": 0.0178, "step": 26630 }, { "grad_norm": 0.16690246760845184, "learning_rate": 8.773535848046131e-05, "loss": 0.0169, "step": 26640 }, { "grad_norm": 0.12246102094650269, "learning_rate": 8.772450863343538e-05, "loss": 0.0124, "step": 26650 }, { "grad_norm": 0.27640894055366516, "learning_rate": 8.77136546609139e-05, "loss": 0.0216, "step": 26660 }, { "grad_norm": 0.1362057775259018, "learning_rate": 8.770279656408385e-05, "loss": 0.0163, "step": 26670 }, { "grad_norm": 0.1425909548997879, "learning_rate": 8.769193434413265e-05, "loss": 0.0173, "step": 26680 }, { "grad_norm": 0.17114557325839996, "learning_rate": 8.76810680022482e-05, "loss": 0.0159, "step": 26690 }, { "grad_norm": 0.2909027636051178, "learning_rate": 8.767019753961878e-05, "loss": 0.0249, "step": 26700 }, { "grad_norm": 0.1918639987707138, "learning_rate": 8.765932295743321e-05, "loss": 0.0192, "step": 26710 }, { "grad_norm": 0.17361004650592804, "learning_rate": 8.764844425688068e-05, "loss": 0.014, "step": 26720 }, { "grad_norm": 0.14839160442352295, "learning_rate": 8.763756143915092e-05, "loss": 0.0139, "step": 26730 }, { "grad_norm": 0.14565616846084595, "learning_rate": 8.7626674505434e-05, "loss": 0.0144, "step": 26740 }, { "grad_norm": 0.16870839893817902, "learning_rate": 8.761578345692053e-05, "loss": 0.0208, "step": 26750 }, { "grad_norm": 0.17821046710014343, "learning_rate": 8.760488829480156e-05, "loss": 0.0164, "step": 26760 }, { "grad_norm": 0.14230500161647797, "learning_rate": 8.759398902026854e-05, "loss": 0.0138, "step": 26770 }, { "grad_norm": 0.14792481064796448, "learning_rate": 8.758308563451339e-05, "loss": 0.0166, "step": 26780 }, { "grad_norm": 0.1979212462902069, "learning_rate": 8.75721781387285e-05, "loss": 0.0174, "step": 26790 }, { "grad_norm": 0.18429197371006012, "learning_rate": 8.75612665341067e-05, "loss": 0.0167, "step": 26800 }, { "grad_norm": 0.14603443443775177, "learning_rate": 8.755035082184126e-05, "loss": 0.0147, "step": 26810 }, { "grad_norm": 0.15642066299915314, "learning_rate": 8.753943100312592e-05, "loss": 0.014, "step": 26820 }, { "grad_norm": 0.13296030461788177, "learning_rate": 8.752850707915484e-05, "loss": 0.0143, "step": 26830 }, { "grad_norm": 0.26921766996383667, "learning_rate": 8.751757905112264e-05, "loss": 0.0194, "step": 26840 }, { "grad_norm": 0.19211626052856445, "learning_rate": 8.75066469202244e-05, "loss": 0.0164, "step": 26850 }, { "grad_norm": 0.14980195462703705, "learning_rate": 8.749571068765567e-05, "loss": 0.0157, "step": 26860 }, { "grad_norm": 0.22665175795555115, "learning_rate": 8.748477035461238e-05, "loss": 0.0146, "step": 26870 }, { "grad_norm": 0.1772104799747467, "learning_rate": 8.747382592229095e-05, "loss": 0.0178, "step": 26880 }, { "grad_norm": 0.17391501367092133, "learning_rate": 8.746287739188828e-05, "loss": 0.0159, "step": 26890 }, { "grad_norm": 0.14809133112430573, "learning_rate": 8.745192476460165e-05, "loss": 0.0156, "step": 26900 }, { "grad_norm": 0.1347099095582962, "learning_rate": 8.744096804162882e-05, "loss": 0.0189, "step": 26910 }, { "grad_norm": 0.31631699204444885, "learning_rate": 8.743000722416804e-05, "loss": 0.0198, "step": 26920 }, { "grad_norm": 0.23358893394470215, "learning_rate": 8.741904231341793e-05, "loss": 0.0158, "step": 26930 }, { "grad_norm": 0.30770811438560486, "learning_rate": 8.740807331057762e-05, "loss": 0.0274, "step": 26940 }, { "grad_norm": 0.15849687159061432, "learning_rate": 8.739710021684667e-05, "loss": 0.0189, "step": 26950 }, { "grad_norm": 0.15569916367530823, "learning_rate": 8.738612303342503e-05, "loss": 0.0154, "step": 26960 }, { "grad_norm": 0.13093189895153046, "learning_rate": 8.73751417615132e-05, "loss": 0.0189, "step": 26970 }, { "grad_norm": 0.09587867558002472, "learning_rate": 8.736415640231208e-05, "loss": 0.0119, "step": 26980 }, { "grad_norm": 0.1737622767686844, "learning_rate": 8.735316695702297e-05, "loss": 0.0171, "step": 26990 }, { "grad_norm": 0.13904103636741638, "learning_rate": 8.734217342684769e-05, "loss": 0.0132, "step": 27000 }, { "grad_norm": 0.20967482030391693, "learning_rate": 8.733117581298847e-05, "loss": 0.0119, "step": 27010 }, { "grad_norm": 0.14270882308483124, "learning_rate": 8.732017411664796e-05, "loss": 0.0151, "step": 27020 }, { "grad_norm": 0.15859679877758026, "learning_rate": 8.730916833902936e-05, "loss": 0.0132, "step": 27030 }, { "grad_norm": 0.1517093926668167, "learning_rate": 8.729815848133618e-05, "loss": 0.0218, "step": 27040 }, { "grad_norm": 0.16206324100494385, "learning_rate": 8.728714454477247e-05, "loss": 0.0148, "step": 27050 }, { "grad_norm": 0.15177921950817108, "learning_rate": 8.727612653054269e-05, "loss": 0.013, "step": 27060 }, { "grad_norm": 0.10890565067529678, "learning_rate": 8.726510443985176e-05, "loss": 0.0146, "step": 27070 }, { "grad_norm": 0.1682748943567276, "learning_rate": 8.725407827390503e-05, "loss": 0.0176, "step": 27080 }, { "grad_norm": 0.14348000288009644, "learning_rate": 8.724304803390833e-05, "loss": 0.0144, "step": 27090 }, { "grad_norm": 0.1651497483253479, "learning_rate": 8.723201372106788e-05, "loss": 0.0134, "step": 27100 }, { "grad_norm": 0.2716020941734314, "learning_rate": 8.722097533659038e-05, "loss": 0.0183, "step": 27110 }, { "grad_norm": 0.1426631063222885, "learning_rate": 8.720993288168299e-05, "loss": 0.017, "step": 27120 }, { "grad_norm": 0.15451762080192566, "learning_rate": 8.719888635755327e-05, "loss": 0.0133, "step": 27130 }, { "grad_norm": 0.23070602118968964, "learning_rate": 8.718783576540928e-05, "loss": 0.0215, "step": 27140 }, { "grad_norm": 0.18956607580184937, "learning_rate": 8.717678110645948e-05, "loss": 0.0151, "step": 27150 }, { "grad_norm": 0.14346325397491455, "learning_rate": 8.716572238191279e-05, "loss": 0.017, "step": 27160 }, { "grad_norm": 0.197982519865036, "learning_rate": 8.715465959297857e-05, "loss": 0.0194, "step": 27170 }, { "grad_norm": 0.2763586938381195, "learning_rate": 8.714359274086665e-05, "loss": 0.0204, "step": 27180 }, { "grad_norm": 0.15425780415534973, "learning_rate": 8.713252182678726e-05, "loss": 0.0183, "step": 27190 }, { "grad_norm": 0.23459620773792267, "learning_rate": 8.712144685195112e-05, "loss": 0.025, "step": 27200 }, { "grad_norm": 0.16189344227313995, "learning_rate": 8.711036781756936e-05, "loss": 0.0158, "step": 27210 }, { "grad_norm": 0.14097373187541962, "learning_rate": 8.709928472485357e-05, "loss": 0.0147, "step": 27220 }, { "grad_norm": 0.14787161350250244, "learning_rate": 8.708819757501579e-05, "loss": 0.0151, "step": 27230 }, { "grad_norm": 0.11567407101392746, "learning_rate": 8.707710636926846e-05, "loss": 0.0137, "step": 27240 }, { "grad_norm": 0.28210484981536865, "learning_rate": 8.706601110882455e-05, "loss": 0.025, "step": 27250 }, { "grad_norm": 0.14681075513362885, "learning_rate": 8.705491179489738e-05, "loss": 0.0138, "step": 27260 }, { "grad_norm": 0.3972674608230591, "learning_rate": 8.704380842870077e-05, "loss": 0.0117, "step": 27270 }, { "grad_norm": 0.21094311773777008, "learning_rate": 8.703270101144895e-05, "loss": 0.0155, "step": 27280 }, { "grad_norm": 0.10313285142183304, "learning_rate": 8.702158954435664e-05, "loss": 0.0132, "step": 27290 }, { "grad_norm": 0.09916100651025772, "learning_rate": 8.701047402863896e-05, "loss": 0.0126, "step": 27300 }, { "grad_norm": 0.18539820611476898, "learning_rate": 8.699935446551148e-05, "loss": 0.0155, "step": 27310 }, { "grad_norm": 0.0999218299984932, "learning_rate": 8.698823085619022e-05, "loss": 0.0151, "step": 27320 }, { "grad_norm": 0.13067324459552765, "learning_rate": 8.697710320189166e-05, "loss": 0.0109, "step": 27330 }, { "grad_norm": 0.14737647771835327, "learning_rate": 8.696597150383268e-05, "loss": 0.0142, "step": 27340 }, { "grad_norm": 0.16453813016414642, "learning_rate": 8.695483576323063e-05, "loss": 0.0193, "step": 27350 }, { "grad_norm": 0.1893981248140335, "learning_rate": 8.69436959813033e-05, "loss": 0.019, "step": 27360 }, { "grad_norm": 0.13286030292510986, "learning_rate": 8.693255215926892e-05, "loss": 0.0136, "step": 27370 }, { "grad_norm": 0.181280255317688, "learning_rate": 8.692140429834617e-05, "loss": 0.0157, "step": 27380 }, { "grad_norm": 0.17324498295783997, "learning_rate": 8.691025239975415e-05, "loss": 0.0153, "step": 27390 }, { "grad_norm": 0.1593199372291565, "learning_rate": 8.689909646471243e-05, "loss": 0.013, "step": 27400 }, { "grad_norm": 0.12824520468711853, "learning_rate": 8.688793649444099e-05, "loss": 0.0189, "step": 27410 }, { "grad_norm": 0.14777886867523193, "learning_rate": 8.687677249016029e-05, "loss": 0.0149, "step": 27420 }, { "grad_norm": 0.22277718782424927, "learning_rate": 8.686560445309118e-05, "loss": 0.0181, "step": 27430 }, { "grad_norm": 0.23598362505435944, "learning_rate": 8.685443238445499e-05, "loss": 0.0211, "step": 27440 }, { "grad_norm": 0.20966649055480957, "learning_rate": 8.68432562854735e-05, "loss": 0.0184, "step": 27450 }, { "grad_norm": 0.1560625433921814, "learning_rate": 8.683207615736887e-05, "loss": 0.016, "step": 27460 }, { "grad_norm": 0.2252187877893448, "learning_rate": 8.682089200136379e-05, "loss": 0.0246, "step": 27470 }, { "grad_norm": 0.260514497756958, "learning_rate": 8.680970381868132e-05, "loss": 0.02, "step": 27480 }, { "grad_norm": 0.16251416504383087, "learning_rate": 8.679851161054498e-05, "loss": 0.018, "step": 27490 }, { "grad_norm": 0.20480486750602722, "learning_rate": 8.678731537817873e-05, "loss": 0.024, "step": 27500 }, { "grad_norm": 0.3028400242328644, "learning_rate": 8.677611512280697e-05, "loss": 0.0231, "step": 27510 }, { "grad_norm": 0.1858925223350525, "learning_rate": 8.676491084565457e-05, "loss": 0.0235, "step": 27520 }, { "grad_norm": 0.21161609888076782, "learning_rate": 8.675370254794678e-05, "loss": 0.022, "step": 27530 }, { "grad_norm": 0.16427132487297058, "learning_rate": 8.674249023090935e-05, "loss": 0.0172, "step": 27540 }, { "grad_norm": 0.1978030800819397, "learning_rate": 8.673127389576843e-05, "loss": 0.0223, "step": 27550 }, { "grad_norm": 0.15998421609401703, "learning_rate": 8.67200535437506e-05, "loss": 0.0138, "step": 27560 }, { "grad_norm": 0.2089099884033203, "learning_rate": 8.670882917608296e-05, "loss": 0.0165, "step": 27570 }, { "grad_norm": 0.15548311173915863, "learning_rate": 8.669760079399292e-05, "loss": 0.0148, "step": 27580 }, { "grad_norm": 0.1548306941986084, "learning_rate": 8.668636839870845e-05, "loss": 0.0148, "step": 27590 }, { "grad_norm": 0.2014123946428299, "learning_rate": 8.667513199145789e-05, "loss": 0.0178, "step": 27600 }, { "grad_norm": 0.22407746315002441, "learning_rate": 8.666389157347002e-05, "loss": 0.0214, "step": 27610 }, { "grad_norm": 0.17515796422958374, "learning_rate": 8.66526471459741e-05, "loss": 0.017, "step": 27620 }, { "grad_norm": 0.1643729954957962, "learning_rate": 8.66413987101998e-05, "loss": 0.0144, "step": 27630 }, { "grad_norm": 0.17129777371883392, "learning_rate": 8.663014626737723e-05, "loss": 0.0155, "step": 27640 }, { "grad_norm": 0.14888520538806915, "learning_rate": 8.661888981873691e-05, "loss": 0.0233, "step": 27650 }, { "grad_norm": 0.14168725907802582, "learning_rate": 8.660762936550988e-05, "loss": 0.0257, "step": 27660 }, { "grad_norm": 0.17454960942268372, "learning_rate": 8.659636490892753e-05, "loss": 0.024, "step": 27670 }, { "grad_norm": 0.19564098119735718, "learning_rate": 8.658509645022174e-05, "loss": 0.0198, "step": 27680 }, { "grad_norm": 0.14463067054748535, "learning_rate": 8.657382399062481e-05, "loss": 0.0136, "step": 27690 }, { "grad_norm": 0.16332118213176727, "learning_rate": 8.656254753136946e-05, "loss": 0.0199, "step": 27700 }, { "grad_norm": 0.23165681958198547, "learning_rate": 8.655126707368891e-05, "loss": 0.0147, "step": 27710 }, { "grad_norm": 0.12902535498142242, "learning_rate": 8.653998261881672e-05, "loss": 0.0149, "step": 27720 }, { "grad_norm": 0.1621161252260208, "learning_rate": 8.652869416798699e-05, "loss": 0.0133, "step": 27730 }, { "grad_norm": 0.13570110499858856, "learning_rate": 8.651740172243417e-05, "loss": 0.0169, "step": 27740 }, { "grad_norm": 0.18202006816864014, "learning_rate": 8.65061052833932e-05, "loss": 0.0153, "step": 27750 }, { "grad_norm": 0.14410518109798431, "learning_rate": 8.649480485209945e-05, "loss": 0.0142, "step": 27760 }, { "grad_norm": 0.160082146525383, "learning_rate": 8.64835004297887e-05, "loss": 0.0231, "step": 27770 }, { "grad_norm": 0.1354362815618515, "learning_rate": 8.64721920176972e-05, "loss": 0.0164, "step": 27780 }, { "grad_norm": 0.2283909022808075, "learning_rate": 8.646087961706164e-05, "loss": 0.0202, "step": 27790 }, { "grad_norm": 0.20352256298065186, "learning_rate": 8.644956322911908e-05, "loss": 0.0167, "step": 27800 }, { "grad_norm": 0.2142694890499115, "learning_rate": 8.643824285510709e-05, "loss": 0.0196, "step": 27810 }, { "grad_norm": 0.12448513507843018, "learning_rate": 8.642691849626364e-05, "loss": 0.0147, "step": 27820 }, { "grad_norm": 0.12124097347259521, "learning_rate": 8.641559015382717e-05, "loss": 0.0156, "step": 27830 }, { "grad_norm": 0.1687634289264679, "learning_rate": 8.640425782903649e-05, "loss": 0.0189, "step": 27840 }, { "grad_norm": 0.1393219232559204, "learning_rate": 8.639292152313091e-05, "loss": 0.0127, "step": 27850 }, { "grad_norm": 0.11971649527549744, "learning_rate": 8.638158123735015e-05, "loss": 0.0138, "step": 27860 }, { "grad_norm": 0.15223994851112366, "learning_rate": 8.637023697293436e-05, "loss": 0.0131, "step": 27870 }, { "grad_norm": 0.2447865754365921, "learning_rate": 8.635888873112414e-05, "loss": 0.0305, "step": 27880 }, { "grad_norm": 0.144855797290802, "learning_rate": 8.634753651316052e-05, "loss": 0.0166, "step": 27890 }, { "grad_norm": 0.14612111449241638, "learning_rate": 8.633618032028496e-05, "loss": 0.0158, "step": 27900 }, { "grad_norm": 0.23855434358119965, "learning_rate": 8.632482015373934e-05, "loss": 0.0156, "step": 27910 }, { "grad_norm": 0.16497862339019775, "learning_rate": 8.6313456014766e-05, "loss": 0.0164, "step": 27920 }, { "grad_norm": 0.19286882877349854, "learning_rate": 8.630208790460771e-05, "loss": 0.0163, "step": 27930 }, { "grad_norm": 0.15995465219020844, "learning_rate": 8.629071582450768e-05, "loss": 0.0162, "step": 27940 }, { "grad_norm": 0.18845787644386292, "learning_rate": 8.62793397757095e-05, "loss": 0.0163, "step": 27950 }, { "grad_norm": 0.1517781913280487, "learning_rate": 8.626795975945729e-05, "loss": 0.0225, "step": 27960 }, { "grad_norm": 0.13017262518405914, "learning_rate": 8.625657577699551e-05, "loss": 0.0158, "step": 27970 }, { "grad_norm": 0.1210646778345108, "learning_rate": 8.624518782956914e-05, "loss": 0.0138, "step": 27980 }, { "grad_norm": 0.22568023204803467, "learning_rate": 8.62337959184235e-05, "loss": 0.0162, "step": 27990 }, { "grad_norm": 0.1658143252134323, "learning_rate": 8.622240004480441e-05, "loss": 0.0209, "step": 28000 }, { "grad_norm": 0.16291698813438416, "learning_rate": 8.621100020995814e-05, "loss": 0.0134, "step": 28010 }, { "grad_norm": 0.21719081699848175, "learning_rate": 8.619959641513132e-05, "loss": 0.02, "step": 28020 }, { "grad_norm": 0.1958620399236679, "learning_rate": 8.618818866157105e-05, "loss": 0.017, "step": 28030 }, { "grad_norm": 0.18167231976985931, "learning_rate": 8.617677695052487e-05, "loss": 0.0158, "step": 28040 }, { "grad_norm": 0.10003530979156494, "learning_rate": 8.616536128324078e-05, "loss": 0.0146, "step": 28050 }, { "grad_norm": 0.14018233120441437, "learning_rate": 8.615394166096712e-05, "loss": 0.014, "step": 28060 }, { "grad_norm": 0.17886406183242798, "learning_rate": 8.614251808495279e-05, "loss": 0.0138, "step": 28070 }, { "grad_norm": 0.16691873967647552, "learning_rate": 8.6131090556447e-05, "loss": 0.0138, "step": 28080 }, { "grad_norm": 0.19642110168933868, "learning_rate": 8.611965907669947e-05, "loss": 0.017, "step": 28090 }, { "grad_norm": 0.15766844153404236, "learning_rate": 8.610822364696034e-05, "loss": 0.0171, "step": 28100 }, { "grad_norm": 0.13150203227996826, "learning_rate": 8.609678426848015e-05, "loss": 0.0156, "step": 28110 }, { "grad_norm": 0.2489291876554489, "learning_rate": 8.60853409425099e-05, "loss": 0.019, "step": 28120 }, { "grad_norm": 0.12277286499738693, "learning_rate": 8.607389367030104e-05, "loss": 0.0141, "step": 28130 }, { "grad_norm": 0.16866731643676758, "learning_rate": 8.606244245310538e-05, "loss": 0.0153, "step": 28140 }, { "grad_norm": 0.22074154019355774, "learning_rate": 8.605098729217525e-05, "loss": 0.0252, "step": 28150 }, { "grad_norm": 0.13138334453105927, "learning_rate": 8.603952818876335e-05, "loss": 0.0183, "step": 28160 }, { "grad_norm": 0.10701949149370193, "learning_rate": 8.602806514412281e-05, "loss": 0.014, "step": 28170 }, { "grad_norm": 0.20144276320934296, "learning_rate": 8.601659815950726e-05, "loss": 0.0203, "step": 28180 }, { "grad_norm": 0.2051352560520172, "learning_rate": 8.600512723617067e-05, "loss": 0.0361, "step": 28190 }, { "grad_norm": 0.23183223605155945, "learning_rate": 8.59936523753675e-05, "loss": 0.0205, "step": 28200 }, { "grad_norm": 0.138701930642128, "learning_rate": 8.598217357835264e-05, "loss": 0.018, "step": 28210 }, { "grad_norm": 0.17278482019901276, "learning_rate": 8.597069084638135e-05, "loss": 0.0155, "step": 28220 }, { "grad_norm": 0.20316572487354279, "learning_rate": 8.595920418070939e-05, "loss": 0.0178, "step": 28230 }, { "grad_norm": 0.15790848433971405, "learning_rate": 8.594771358259295e-05, "loss": 0.0149, "step": 28240 }, { "grad_norm": 0.22828976809978485, "learning_rate": 8.593621905328858e-05, "loss": 0.0227, "step": 28250 }, { "grad_norm": 0.23237696290016174, "learning_rate": 8.592472059405333e-05, "loss": 0.0197, "step": 28260 }, { "grad_norm": 0.13759084045886993, "learning_rate": 8.591321820614464e-05, "loss": 0.0122, "step": 28270 }, { "grad_norm": 0.16772182285785675, "learning_rate": 8.590171189082041e-05, "loss": 0.0182, "step": 28280 }, { "grad_norm": 0.1172453910112381, "learning_rate": 8.589020164933894e-05, "loss": 0.0136, "step": 28290 }, { "grad_norm": 0.17846402525901794, "learning_rate": 8.587868748295898e-05, "loss": 0.0137, "step": 28300 }, { "grad_norm": 0.12748907506465912, "learning_rate": 8.586716939293971e-05, "loss": 0.0142, "step": 28310 }, { "grad_norm": 0.12753324210643768, "learning_rate": 8.58556473805407e-05, "loss": 0.0139, "step": 28320 }, { "grad_norm": 0.13211052119731903, "learning_rate": 8.584412144702202e-05, "loss": 0.0174, "step": 28330 }, { "grad_norm": 0.17038753628730774, "learning_rate": 8.58325915936441e-05, "loss": 0.0159, "step": 28340 }, { "grad_norm": 0.1351444572210312, "learning_rate": 8.582105782166783e-05, "loss": 0.0135, "step": 28350 }, { "grad_norm": 0.13130412995815277, "learning_rate": 8.580952013235455e-05, "loss": 0.0134, "step": 28360 }, { "grad_norm": 0.17395634949207306, "learning_rate": 8.579797852696596e-05, "loss": 0.0171, "step": 28370 }, { "grad_norm": 0.13889382779598236, "learning_rate": 8.578643300676428e-05, "loss": 0.0151, "step": 28380 }, { "grad_norm": 0.143227219581604, "learning_rate": 8.577488357301209e-05, "loss": 0.0171, "step": 28390 }, { "grad_norm": 0.15692658722400665, "learning_rate": 8.576333022697242e-05, "loss": 0.013, "step": 28400 }, { "grad_norm": 0.2628374397754669, "learning_rate": 8.575177296990873e-05, "loss": 0.0164, "step": 28410 }, { "grad_norm": 0.1677473783493042, "learning_rate": 8.574021180308489e-05, "loss": 0.0205, "step": 28420 }, { "grad_norm": 0.2656395435333252, "learning_rate": 8.572864672776523e-05, "loss": 0.0171, "step": 28430 }, { "grad_norm": 0.1790633499622345, "learning_rate": 8.571707774521447e-05, "loss": 0.0154, "step": 28440 }, { "grad_norm": 0.18820521235466003, "learning_rate": 8.57055048566978e-05, "loss": 0.0167, "step": 28450 }, { "grad_norm": 0.1861162632703781, "learning_rate": 8.569392806348078e-05, "loss": 0.0167, "step": 28460 }, { "grad_norm": 0.17648006975650787, "learning_rate": 8.568234736682947e-05, "loss": 0.0153, "step": 28470 }, { "grad_norm": 0.18880824744701385, "learning_rate": 8.567076276801029e-05, "loss": 0.0209, "step": 28480 }, { "grad_norm": 0.14333723485469818, "learning_rate": 8.565917426829013e-05, "loss": 0.0128, "step": 28490 }, { "grad_norm": 0.14207127690315247, "learning_rate": 8.564758186893628e-05, "loss": 0.0142, "step": 28500 }, { "grad_norm": 0.1650894433259964, "learning_rate": 8.563598557121649e-05, "loss": 0.0149, "step": 28510 }, { "grad_norm": 0.14876340329647064, "learning_rate": 8.562438537639888e-05, "loss": 0.0187, "step": 28520 }, { "grad_norm": 0.22798319160938263, "learning_rate": 8.561278128575206e-05, "loss": 0.017, "step": 28530 }, { "grad_norm": 0.16107435524463654, "learning_rate": 8.5601173300545e-05, "loss": 0.0156, "step": 28540 }, { "grad_norm": 0.1820375919342041, "learning_rate": 8.558956142204717e-05, "loss": 0.019, "step": 28550 }, { "grad_norm": 0.1493062674999237, "learning_rate": 8.55779456515284e-05, "loss": 0.0179, "step": 28560 }, { "grad_norm": 0.16499917209148407, "learning_rate": 8.556632599025898e-05, "loss": 0.0242, "step": 28570 }, { "grad_norm": 0.1570412963628769, "learning_rate": 8.555470243950964e-05, "loss": 0.021, "step": 28580 }, { "grad_norm": 0.18099914491176605, "learning_rate": 8.554307500055148e-05, "loss": 0.0178, "step": 28590 }, { "grad_norm": 0.14699001610279083, "learning_rate": 8.553144367465609e-05, "loss": 0.0145, "step": 28600 }, { "grad_norm": 0.17011219263076782, "learning_rate": 8.551980846309544e-05, "loss": 0.026, "step": 28610 }, { "grad_norm": 0.16458721458911896, "learning_rate": 8.550816936714193e-05, "loss": 0.0195, "step": 28620 }, { "grad_norm": 0.10658632218837738, "learning_rate": 8.549652638806841e-05, "loss": 0.0149, "step": 28630 }, { "grad_norm": 0.12133340537548065, "learning_rate": 8.548487952714812e-05, "loss": 0.0177, "step": 28640 }, { "grad_norm": 0.1970311999320984, "learning_rate": 8.547322878565478e-05, "loss": 0.0182, "step": 28650 }, { "grad_norm": 0.26510733366012573, "learning_rate": 8.546157416486245e-05, "loss": 0.0174, "step": 28660 }, { "grad_norm": 0.13369102776050568, "learning_rate": 8.54499156660457e-05, "loss": 0.0134, "step": 28670 }, { "grad_norm": 0.12369266152381897, "learning_rate": 8.543825329047947e-05, "loss": 0.0163, "step": 28680 }, { "grad_norm": 0.12361416220664978, "learning_rate": 8.542658703943913e-05, "loss": 0.0159, "step": 28690 }, { "grad_norm": 0.11518321931362152, "learning_rate": 8.541491691420051e-05, "loss": 0.0173, "step": 28700 }, { "grad_norm": 0.1471816897392273, "learning_rate": 8.54032429160398e-05, "loss": 0.0148, "step": 28710 }, { "grad_norm": 0.12768524885177612, "learning_rate": 8.539156504623369e-05, "loss": 0.019, "step": 28720 }, { "grad_norm": 0.1951921135187149, "learning_rate": 8.537988330605923e-05, "loss": 0.0314, "step": 28730 }, { "grad_norm": 0.5106045007705688, "learning_rate": 8.536819769679393e-05, "loss": 0.0145, "step": 28740 }, { "grad_norm": 0.19996042549610138, "learning_rate": 8.53565082197157e-05, "loss": 0.0176, "step": 28750 }, { "grad_norm": 0.17959676682949066, "learning_rate": 8.534481487610289e-05, "loss": 0.026, "step": 28760 }, { "grad_norm": 0.14557939767837524, "learning_rate": 8.533311766723428e-05, "loss": 0.0145, "step": 28770 }, { "grad_norm": 0.13888393342494965, "learning_rate": 8.532141659438901e-05, "loss": 0.0176, "step": 28780 }, { "grad_norm": 0.13896334171295166, "learning_rate": 8.530971165884675e-05, "loss": 0.0184, "step": 28790 }, { "grad_norm": 0.1332722306251526, "learning_rate": 8.529800286188752e-05, "loss": 0.0161, "step": 28800 }, { "grad_norm": 0.12112534046173096, "learning_rate": 8.528629020479175e-05, "loss": 0.0117, "step": 28810 }, { "grad_norm": 0.17020884156227112, "learning_rate": 8.527457368884033e-05, "loss": 0.0165, "step": 28820 }, { "grad_norm": 0.13366448879241943, "learning_rate": 8.526285331531458e-05, "loss": 0.0145, "step": 28830 }, { "grad_norm": 0.14007040858268738, "learning_rate": 8.525112908549621e-05, "loss": 0.0136, "step": 28840 }, { "grad_norm": 0.20370975136756897, "learning_rate": 8.523940100066735e-05, "loss": 0.0193, "step": 28850 }, { "grad_norm": 0.17013587057590485, "learning_rate": 8.52276690621106e-05, "loss": 0.0174, "step": 28860 }, { "grad_norm": 0.16576024889945984, "learning_rate": 8.521593327110889e-05, "loss": 0.018, "step": 28870 }, { "grad_norm": 0.275911808013916, "learning_rate": 8.520419362894569e-05, "loss": 0.026, "step": 28880 }, { "grad_norm": 0.18683040142059326, "learning_rate": 8.51924501369048e-05, "loss": 0.0163, "step": 28890 }, { "grad_norm": 0.1590772122144699, "learning_rate": 8.518070279627047e-05, "loss": 0.017, "step": 28900 }, { "grad_norm": 0.18372265994548798, "learning_rate": 8.516895160832737e-05, "loss": 0.0166, "step": 28910 }, { "grad_norm": 0.21964894235134125, "learning_rate": 8.515719657436061e-05, "loss": 0.0142, "step": 28920 }, { "grad_norm": 0.11174304783344269, "learning_rate": 8.514543769565568e-05, "loss": 0.0143, "step": 28930 }, { "grad_norm": 0.2610425055027008, "learning_rate": 8.513367497349853e-05, "loss": 0.0157, "step": 28940 }, { "grad_norm": 0.13998104631900787, "learning_rate": 8.51219084091755e-05, "loss": 0.0178, "step": 28950 }, { "grad_norm": 0.13673052191734314, "learning_rate": 8.511013800397338e-05, "loss": 0.0161, "step": 28960 }, { "grad_norm": 0.1833869367837906, "learning_rate": 8.509836375917937e-05, "loss": 0.0188, "step": 28970 }, { "grad_norm": 0.11801403015851974, "learning_rate": 8.508658567608104e-05, "loss": 0.0143, "step": 28980 }, { "grad_norm": 0.19487610459327698, "learning_rate": 8.507480375596647e-05, "loss": 0.0145, "step": 28990 }, { "grad_norm": 0.1823681741952896, "learning_rate": 8.506301800012408e-05, "loss": 0.0153, "step": 29000 }, { "grad_norm": 0.1533142328262329, "learning_rate": 8.505122840984278e-05, "loss": 0.0155, "step": 29010 }, { "grad_norm": 0.13114695250988007, "learning_rate": 8.503943498641182e-05, "loss": 0.0167, "step": 29020 }, { "grad_norm": 0.1297517716884613, "learning_rate": 8.502763773112095e-05, "loss": 0.0151, "step": 29030 }, { "grad_norm": 0.21554812788963318, "learning_rate": 8.501583664526026e-05, "loss": 0.0191, "step": 29040 }, { "grad_norm": 0.13074736297130585, "learning_rate": 8.500403173012032e-05, "loss": 0.0162, "step": 29050 }, { "grad_norm": 0.22311513125896454, "learning_rate": 8.499222298699211e-05, "loss": 0.0178, "step": 29060 }, { "grad_norm": 0.22421863675117493, "learning_rate": 8.498041041716701e-05, "loss": 0.016, "step": 29070 }, { "grad_norm": 0.16532550752162933, "learning_rate": 8.496859402193681e-05, "loss": 0.0142, "step": 29080 }, { "grad_norm": 0.15985533595085144, "learning_rate": 8.495677380259374e-05, "loss": 0.0134, "step": 29090 }, { "grad_norm": 0.15652625262737274, "learning_rate": 8.494494976043045e-05, "loss": 0.0174, "step": 29100 }, { "grad_norm": 0.14701421558856964, "learning_rate": 8.493312189673998e-05, "loss": 0.0136, "step": 29110 }, { "grad_norm": 0.14506292343139648, "learning_rate": 8.492129021281584e-05, "loss": 0.0155, "step": 29120 }, { "grad_norm": 0.14416669309139252, "learning_rate": 8.490945470995188e-05, "loss": 0.0117, "step": 29130 }, { "grad_norm": 0.12269096076488495, "learning_rate": 8.489761538944247e-05, "loss": 0.0157, "step": 29140 }, { "grad_norm": 0.11825773119926453, "learning_rate": 8.48857722525823e-05, "loss": 0.0146, "step": 29150 }, { "grad_norm": 0.17677666246891022, "learning_rate": 8.487392530066652e-05, "loss": 0.0202, "step": 29160 }, { "grad_norm": 0.18834541738033295, "learning_rate": 8.486207453499069e-05, "loss": 0.0223, "step": 29170 }, { "grad_norm": 0.233986034989357, "learning_rate": 8.485021995685082e-05, "loss": 0.0168, "step": 29180 }, { "grad_norm": 0.21597786247730255, "learning_rate": 8.483836156754328e-05, "loss": 0.0147, "step": 29190 }, { "grad_norm": 0.13975347578525543, "learning_rate": 8.482649936836491e-05, "loss": 0.015, "step": 29200 }, { "grad_norm": 0.18120142817497253, "learning_rate": 8.481463336061293e-05, "loss": 0.0153, "step": 29210 }, { "grad_norm": 0.14062151312828064, "learning_rate": 8.480276354558496e-05, "loss": 0.0238, "step": 29220 }, { "grad_norm": 0.1384837031364441, "learning_rate": 8.479088992457913e-05, "loss": 0.0147, "step": 29230 }, { "grad_norm": 0.11806603521108627, "learning_rate": 8.477901249889387e-05, "loss": 0.0246, "step": 29240 }, { "grad_norm": 0.17236123979091644, "learning_rate": 8.47671312698281e-05, "loss": 0.0149, "step": 29250 }, { "grad_norm": 0.16696326434612274, "learning_rate": 8.475524623868112e-05, "loss": 0.0177, "step": 29260 }, { "grad_norm": 0.20981314778327942, "learning_rate": 8.474335740675266e-05, "loss": 0.0169, "step": 29270 }, { "grad_norm": 0.20132461190223694, "learning_rate": 8.473146477534289e-05, "loss": 0.0181, "step": 29280 }, { "grad_norm": 0.1458989977836609, "learning_rate": 8.471956834575232e-05, "loss": 0.0177, "step": 29290 }, { "grad_norm": 0.17173554003238678, "learning_rate": 8.470766811928197e-05, "loss": 0.0158, "step": 29300 }, { "grad_norm": 0.1661941111087799, "learning_rate": 8.469576409723323e-05, "loss": 0.0235, "step": 29310 }, { "grad_norm": 0.19277358055114746, "learning_rate": 8.468385628090788e-05, "loss": 0.0175, "step": 29320 }, { "grad_norm": 0.21329329907894135, "learning_rate": 8.467194467160815e-05, "loss": 0.0155, "step": 29330 }, { "grad_norm": 0.1316782385110855, "learning_rate": 8.466002927063667e-05, "loss": 0.0144, "step": 29340 }, { "grad_norm": 0.20799270272254944, "learning_rate": 8.464811007929651e-05, "loss": 0.0199, "step": 29350 }, { "grad_norm": 0.18610408902168274, "learning_rate": 8.463618709889114e-05, "loss": 0.0163, "step": 29360 }, { "grad_norm": 0.16150464117527008, "learning_rate": 8.462426033072442e-05, "loss": 0.0173, "step": 29370 }, { "grad_norm": 0.1831352859735489, "learning_rate": 8.461232977610061e-05, "loss": 0.0154, "step": 29380 }, { "grad_norm": 0.11816773563623428, "learning_rate": 8.46003954363245e-05, "loss": 0.0119, "step": 29390 }, { "grad_norm": 0.14236508309841156, "learning_rate": 8.458845731270115e-05, "loss": 0.0153, "step": 29400 }, { "grad_norm": 0.2347380667924881, "learning_rate": 8.45765154065361e-05, "loss": 0.0162, "step": 29410 }, { "grad_norm": 0.15089638531208038, "learning_rate": 8.456456971913532e-05, "loss": 0.0134, "step": 29420 }, { "grad_norm": 0.1577099859714508, "learning_rate": 8.455262025180517e-05, "loss": 0.0167, "step": 29430 }, { "grad_norm": 0.09711679071187973, "learning_rate": 8.454066700585242e-05, "loss": 0.0157, "step": 29440 }, { "grad_norm": 0.1267976015806198, "learning_rate": 8.452870998258423e-05, "loss": 0.0172, "step": 29450 }, { "grad_norm": 0.1312425136566162, "learning_rate": 8.451674918330825e-05, "loss": 0.0119, "step": 29460 }, { "grad_norm": 0.18847352266311646, "learning_rate": 8.450478460933246e-05, "loss": 0.0155, "step": 29470 }, { "grad_norm": 0.1924211084842682, "learning_rate": 8.449281626196532e-05, "loss": 0.019, "step": 29480 }, { "grad_norm": 0.24569660425186157, "learning_rate": 8.448084414251564e-05, "loss": 0.0185, "step": 29490 }, { "grad_norm": 0.17172880470752716, "learning_rate": 8.446886825229271e-05, "loss": 0.0197, "step": 29500 }, { "grad_norm": 0.1789063662290573, "learning_rate": 8.445688859260615e-05, "loss": 0.0157, "step": 29510 }, { "grad_norm": 0.20502275228500366, "learning_rate": 8.444490516476606e-05, "loss": 0.0139, "step": 29520 }, { "grad_norm": 0.12855304777622223, "learning_rate": 8.443291797008293e-05, "loss": 0.0168, "step": 29530 }, { "grad_norm": 0.12319879233837128, "learning_rate": 8.442092700986765e-05, "loss": 0.0278, "step": 29540 }, { "grad_norm": 0.12630751729011536, "learning_rate": 8.440893228543156e-05, "loss": 0.0177, "step": 29550 }, { "grad_norm": 0.13037346303462982, "learning_rate": 8.439693379808638e-05, "loss": 0.0138, "step": 29560 }, { "grad_norm": 0.149660125374794, "learning_rate": 8.43849315491442e-05, "loss": 0.0168, "step": 29570 }, { "grad_norm": 0.1973385512828827, "learning_rate": 8.437292553991763e-05, "loss": 0.0196, "step": 29580 }, { "grad_norm": 0.15872257947921753, "learning_rate": 8.436091577171959e-05, "loss": 0.016, "step": 29590 }, { "grad_norm": 0.14537569880485535, "learning_rate": 8.434890224586347e-05, "loss": 0.0131, "step": 29600 }, { "grad_norm": 0.11614295095205307, "learning_rate": 8.433688496366303e-05, "loss": 0.0109, "step": 29610 }, { "grad_norm": 0.1977703869342804, "learning_rate": 8.432486392643248e-05, "loss": 0.0121, "step": 29620 }, { "grad_norm": 0.13933221995830536, "learning_rate": 8.431283913548643e-05, "loss": 0.0168, "step": 29630 }, { "grad_norm": 0.10463914275169373, "learning_rate": 8.430081059213985e-05, "loss": 0.0143, "step": 29640 }, { "grad_norm": 0.13165464997291565, "learning_rate": 8.428877829770823e-05, "loss": 0.016, "step": 29650 }, { "grad_norm": 0.16536009311676025, "learning_rate": 8.427674225350735e-05, "loss": 0.0177, "step": 29660 }, { "grad_norm": 0.19832274317741394, "learning_rate": 8.426470246085347e-05, "loss": 0.0186, "step": 29670 }, { "grad_norm": 0.18085259199142456, "learning_rate": 8.425265892106324e-05, "loss": 0.0185, "step": 29680 }, { "grad_norm": 0.16330720484256744, "learning_rate": 8.424061163545374e-05, "loss": 0.0183, "step": 29690 }, { "grad_norm": 0.1923598200082779, "learning_rate": 8.422856060534243e-05, "loss": 0.0134, "step": 29700 }, { "grad_norm": 0.275900274515152, "learning_rate": 8.421650583204718e-05, "loss": 0.0191, "step": 29710 }, { "grad_norm": 0.2285616248846054, "learning_rate": 8.420444731688633e-05, "loss": 0.0169, "step": 29720 }, { "grad_norm": 0.17809806764125824, "learning_rate": 8.419238506117852e-05, "loss": 0.0283, "step": 29730 }, { "grad_norm": 0.1473456174135208, "learning_rate": 8.418031906624289e-05, "loss": 0.0143, "step": 29740 }, { "grad_norm": 0.17585694789886475, "learning_rate": 8.416824933339898e-05, "loss": 0.0237, "step": 29750 }, { "grad_norm": 0.16058509051799774, "learning_rate": 8.415617586396667e-05, "loss": 0.0155, "step": 29760 }, { "grad_norm": 0.14994126558303833, "learning_rate": 8.414409865926632e-05, "loss": 0.0207, "step": 29770 }, { "grad_norm": 0.4797939956188202, "learning_rate": 8.413201772061867e-05, "loss": 0.0173, "step": 29780 }, { "grad_norm": 0.1282249242067337, "learning_rate": 8.411993304934488e-05, "loss": 0.014, "step": 29790 }, { "grad_norm": 0.18348689377307892, "learning_rate": 8.410784464676654e-05, "loss": 0.015, "step": 29800 }, { "grad_norm": 0.1314314752817154, "learning_rate": 8.409575251420556e-05, "loss": 0.0137, "step": 29810 }, { "grad_norm": 0.10706157982349396, "learning_rate": 8.408365665298435e-05, "loss": 0.0124, "step": 29820 }, { "grad_norm": 0.36988362669944763, "learning_rate": 8.40715570644257e-05, "loss": 0.0229, "step": 29830 }, { "grad_norm": 0.16812369227409363, "learning_rate": 8.40594537498528e-05, "loss": 0.0246, "step": 29840 }, { "grad_norm": 0.2406930923461914, "learning_rate": 8.404734671058924e-05, "loss": 0.0221, "step": 29850 }, { "grad_norm": 0.19249291718006134, "learning_rate": 8.403523594795902e-05, "loss": 0.0184, "step": 29860 }, { "grad_norm": 0.1398271769285202, "learning_rate": 8.402312146328659e-05, "loss": 0.014, "step": 29870 }, { "grad_norm": 0.22712966799736023, "learning_rate": 8.401100325789675e-05, "loss": 0.0181, "step": 29880 }, { "grad_norm": 0.15546002984046936, "learning_rate": 8.399888133311472e-05, "loss": 0.0141, "step": 29890 }, { "grad_norm": 0.1768646091222763, "learning_rate": 8.398675569026613e-05, "loss": 0.0266, "step": 29900 }, { "grad_norm": 0.13185833394527435, "learning_rate": 8.397462633067705e-05, "loss": 0.0161, "step": 29910 }, { "grad_norm": 0.11459401249885559, "learning_rate": 8.396249325567392e-05, "loss": 0.0159, "step": 29920 }, { "grad_norm": 0.1692984700202942, "learning_rate": 8.395035646658357e-05, "loss": 0.0235, "step": 29930 }, { "grad_norm": 0.17691418528556824, "learning_rate": 8.39382159647333e-05, "loss": 0.0163, "step": 29940 }, { "grad_norm": 0.17822811007499695, "learning_rate": 8.392607175145075e-05, "loss": 0.0152, "step": 29950 }, { "grad_norm": 0.17715008556842804, "learning_rate": 8.3913923828064e-05, "loss": 0.0189, "step": 29960 }, { "grad_norm": 0.22823335230350494, "learning_rate": 8.390177219590152e-05, "loss": 0.0152, "step": 29970 }, { "grad_norm": 0.19675596058368683, "learning_rate": 8.388961685629222e-05, "loss": 0.0178, "step": 29980 }, { "grad_norm": 0.152419775724411, "learning_rate": 8.387745781056536e-05, "loss": 0.0157, "step": 29990 }, { "grad_norm": 0.13573969900608063, "learning_rate": 8.386529506005065e-05, "loss": 0.0145, "step": 30000 }, { "grad_norm": 0.18449604511260986, "learning_rate": 8.38531286060782e-05, "loss": 0.0164, "step": 30010 }, { "grad_norm": 0.14980512857437134, "learning_rate": 8.384095844997849e-05, "loss": 0.0134, "step": 30020 }, { "grad_norm": 0.148381769657135, "learning_rate": 8.382878459308245e-05, "loss": 0.0127, "step": 30030 }, { "grad_norm": 0.13623632490634918, "learning_rate": 8.381660703672138e-05, "loss": 0.0143, "step": 30040 }, { "grad_norm": 0.14157287776470184, "learning_rate": 8.380442578222702e-05, "loss": 0.0236, "step": 30050 }, { "grad_norm": 0.13781365752220154, "learning_rate": 8.379224083093146e-05, "loss": 0.0178, "step": 30060 }, { "grad_norm": 0.15772905945777893, "learning_rate": 8.378005218416727e-05, "loss": 0.0157, "step": 30070 }, { "grad_norm": 0.151526540517807, "learning_rate": 8.376785984326735e-05, "loss": 0.0175, "step": 30080 }, { "grad_norm": 0.20565296709537506, "learning_rate": 8.375566380956506e-05, "loss": 0.0139, "step": 30090 }, { "grad_norm": 0.2473592460155487, "learning_rate": 8.374346408439411e-05, "loss": 0.0162, "step": 30100 }, { "grad_norm": 0.22125746309757233, "learning_rate": 8.373126066908868e-05, "loss": 0.0213, "step": 30110 }, { "grad_norm": 0.12792973220348358, "learning_rate": 8.371905356498326e-05, "loss": 0.017, "step": 30120 }, { "grad_norm": 0.1807451993227005, "learning_rate": 8.370684277341288e-05, "loss": 0.0161, "step": 30130 }, { "grad_norm": 0.13888512551784515, "learning_rate": 8.369462829571282e-05, "loss": 0.0154, "step": 30140 }, { "grad_norm": 0.20339109003543854, "learning_rate": 8.36824101332189e-05, "loss": 0.0147, "step": 30150 }, { "grad_norm": 0.12077663093805313, "learning_rate": 8.367018828726721e-05, "loss": 0.0127, "step": 30160 }, { "grad_norm": 0.1878005564212799, "learning_rate": 8.365796275919438e-05, "loss": 0.0174, "step": 30170 }, { "grad_norm": 0.11419723182916641, "learning_rate": 8.364573355033734e-05, "loss": 0.0124, "step": 30180 }, { "grad_norm": 0.16085383296012878, "learning_rate": 8.363350066203346e-05, "loss": 0.0122, "step": 30190 }, { "grad_norm": 0.17286799848079681, "learning_rate": 8.362126409562053e-05, "loss": 0.0134, "step": 30200 }, { "grad_norm": 0.2049308866262436, "learning_rate": 8.360902385243667e-05, "loss": 0.0148, "step": 30210 }, { "grad_norm": 0.11928606778383255, "learning_rate": 8.359677993382052e-05, "loss": 0.0123, "step": 30220 }, { "grad_norm": 0.2548709809780121, "learning_rate": 8.358453234111103e-05, "loss": 0.0145, "step": 30230 }, { "grad_norm": 0.1786654144525528, "learning_rate": 8.357228107564756e-05, "loss": 0.0193, "step": 30240 }, { "grad_norm": 0.18617956340312958, "learning_rate": 8.356002613876993e-05, "loss": 0.0133, "step": 30250 }, { "grad_norm": 0.12439808994531631, "learning_rate": 8.35477675318183e-05, "loss": 0.0108, "step": 30260 }, { "grad_norm": 0.2103133350610733, "learning_rate": 8.353550525613323e-05, "loss": 0.0169, "step": 30270 }, { "grad_norm": 0.16678619384765625, "learning_rate": 8.352323931305572e-05, "loss": 0.0257, "step": 30280 }, { "grad_norm": 0.1010483130812645, "learning_rate": 8.351096970392717e-05, "loss": 0.0183, "step": 30290 }, { "grad_norm": 0.14512120187282562, "learning_rate": 8.349869643008937e-05, "loss": 0.0136, "step": 30300 }, { "grad_norm": 0.17674289643764496, "learning_rate": 8.348641949288449e-05, "loss": 0.0219, "step": 30310 }, { "grad_norm": 0.20758689939975739, "learning_rate": 8.34741388936551e-05, "loss": 0.0192, "step": 30320 }, { "grad_norm": 0.2254127860069275, "learning_rate": 8.346185463374423e-05, "loss": 0.0185, "step": 30330 }, { "grad_norm": 0.13264107704162598, "learning_rate": 8.344956671449524e-05, "loss": 0.0163, "step": 30340 }, { "grad_norm": 0.1138787567615509, "learning_rate": 8.343727513725192e-05, "loss": 0.0144, "step": 30350 }, { "grad_norm": 0.1839062124490738, "learning_rate": 8.342497990335847e-05, "loss": 0.0186, "step": 30360 }, { "grad_norm": 0.18065254390239716, "learning_rate": 8.341268101415946e-05, "loss": 0.0178, "step": 30370 }, { "grad_norm": 0.12221037596464157, "learning_rate": 8.34003784709999e-05, "loss": 0.0155, "step": 30380 }, { "grad_norm": 0.13913139700889587, "learning_rate": 8.338807227522516e-05, "loss": 0.0186, "step": 30390 }, { "grad_norm": 0.1460905820131302, "learning_rate": 8.337576242818103e-05, "loss": 0.015, "step": 30400 }, { "grad_norm": 0.12106359004974365, "learning_rate": 8.336344893121372e-05, "loss": 0.0123, "step": 30410 }, { "grad_norm": 0.16953885555267334, "learning_rate": 8.335113178566977e-05, "loss": 0.0153, "step": 30420 }, { "grad_norm": 0.18367275595664978, "learning_rate": 8.33388109928962e-05, "loss": 0.0162, "step": 30430 }, { "grad_norm": 0.24429242312908173, "learning_rate": 8.33264865542404e-05, "loss": 0.0147, "step": 30440 }, { "grad_norm": 0.12008996307849884, "learning_rate": 8.331415847105013e-05, "loss": 0.0151, "step": 30450 }, { "grad_norm": 0.12550315260887146, "learning_rate": 8.330182674467357e-05, "loss": 0.0143, "step": 30460 }, { "grad_norm": 0.1437915712594986, "learning_rate": 8.32894913764593e-05, "loss": 0.0145, "step": 30470 }, { "grad_norm": 0.08639545738697052, "learning_rate": 8.327715236775633e-05, "loss": 0.0131, "step": 30480 }, { "grad_norm": 0.17704372107982635, "learning_rate": 8.326480971991398e-05, "loss": 0.015, "step": 30490 }, { "grad_norm": 0.1684717983007431, "learning_rate": 8.325246343428206e-05, "loss": 0.0171, "step": 30500 }, { "grad_norm": 0.1520690619945526, "learning_rate": 8.324011351221072e-05, "loss": 0.0147, "step": 30510 }, { "grad_norm": 0.20819689333438873, "learning_rate": 8.322775995505057e-05, "loss": 0.0145, "step": 30520 }, { "grad_norm": 0.12181401997804642, "learning_rate": 8.321540276415254e-05, "loss": 0.0173, "step": 30530 }, { "grad_norm": 0.23670509457588196, "learning_rate": 8.320304194086798e-05, "loss": 0.0167, "step": 30540 }, { "grad_norm": 0.1275826394557953, "learning_rate": 8.31906774865487e-05, "loss": 0.0131, "step": 30550 }, { "grad_norm": 0.1585678607225418, "learning_rate": 8.317830940254682e-05, "loss": 0.0134, "step": 30560 }, { "grad_norm": 0.17321528494358063, "learning_rate": 8.316593769021491e-05, "loss": 0.0195, "step": 30570 }, { "grad_norm": 0.21201643347740173, "learning_rate": 8.315356235090592e-05, "loss": 0.0204, "step": 30580 }, { "grad_norm": 0.1365056335926056, "learning_rate": 8.314118338597319e-05, "loss": 0.0151, "step": 30590 }, { "grad_norm": 0.1482141613960266, "learning_rate": 8.312880079677048e-05, "loss": 0.0157, "step": 30600 }, { "grad_norm": 0.17823544144630432, "learning_rate": 8.311641458465191e-05, "loss": 0.0164, "step": 30610 }, { "grad_norm": 0.17861442267894745, "learning_rate": 8.310402475097205e-05, "loss": 0.0172, "step": 30620 }, { "grad_norm": 0.15740260481834412, "learning_rate": 8.309163129708581e-05, "loss": 0.0147, "step": 30630 }, { "grad_norm": 0.16223005950450897, "learning_rate": 8.307923422434852e-05, "loss": 0.0131, "step": 30640 }, { "grad_norm": 0.250876247882843, "learning_rate": 8.30668335341159e-05, "loss": 0.0207, "step": 30650 }, { "grad_norm": 0.13604854047298431, "learning_rate": 8.305442922774408e-05, "loss": 0.0145, "step": 30660 }, { "grad_norm": 0.15638291835784912, "learning_rate": 8.304202130658959e-05, "loss": 0.0156, "step": 30670 }, { "grad_norm": 0.14289969205856323, "learning_rate": 8.302960977200931e-05, "loss": 0.0133, "step": 30680 }, { "grad_norm": 0.16341377794742584, "learning_rate": 8.301719462536058e-05, "loss": 0.0133, "step": 30690 }, { "grad_norm": 0.19675840437412262, "learning_rate": 8.300477586800108e-05, "loss": 0.0174, "step": 30700 }, { "grad_norm": 0.11160849779844284, "learning_rate": 8.299235350128892e-05, "loss": 0.0141, "step": 30710 }, { "grad_norm": 0.20227192342281342, "learning_rate": 8.297992752658258e-05, "loss": 0.0142, "step": 30720 }, { "grad_norm": 0.11837785691022873, "learning_rate": 8.296749794524094e-05, "loss": 0.0205, "step": 30730 }, { "grad_norm": 0.10273600369691849, "learning_rate": 8.295506475862332e-05, "loss": 0.0127, "step": 30740 }, { "grad_norm": 0.12779824435710907, "learning_rate": 8.294262796808933e-05, "loss": 0.0122, "step": 30750 }, { "grad_norm": 0.13977113366127014, "learning_rate": 8.293018757499909e-05, "loss": 0.0131, "step": 30760 }, { "grad_norm": 0.12392912805080414, "learning_rate": 8.291774358071305e-05, "loss": 0.0128, "step": 30770 }, { "grad_norm": 0.18558856844902039, "learning_rate": 8.290529598659205e-05, "loss": 0.0152, "step": 30780 }, { "grad_norm": 0.13687914609909058, "learning_rate": 8.289284479399738e-05, "loss": 0.0149, "step": 30790 }, { "grad_norm": 0.1785397231578827, "learning_rate": 8.288039000429064e-05, "loss": 0.0138, "step": 30800 }, { "grad_norm": 0.11805180460214615, "learning_rate": 8.286793161883388e-05, "loss": 0.0154, "step": 30810 }, { "grad_norm": 0.14356622099876404, "learning_rate": 8.285546963898954e-05, "loss": 0.0148, "step": 30820 }, { "grad_norm": 0.19063401222229004, "learning_rate": 8.284300406612044e-05, "loss": 0.0168, "step": 30830 }, { "grad_norm": 0.15560080111026764, "learning_rate": 8.283053490158978e-05, "loss": 0.0173, "step": 30840 }, { "grad_norm": 0.12084954977035522, "learning_rate": 8.28180621467612e-05, "loss": 0.0102, "step": 30850 }, { "grad_norm": 0.15621639788150787, "learning_rate": 8.280558580299867e-05, "loss": 0.0163, "step": 30860 }, { "grad_norm": 0.13816848397254944, "learning_rate": 8.279310587166661e-05, "loss": 0.0134, "step": 30870 }, { "grad_norm": 0.12993381917476654, "learning_rate": 8.278062235412978e-05, "loss": 0.0152, "step": 30880 }, { "grad_norm": 0.1445852369070053, "learning_rate": 8.276813525175339e-05, "loss": 0.0157, "step": 30890 }, { "grad_norm": 0.14783664047718048, "learning_rate": 8.2755644565903e-05, "loss": 0.018, "step": 30900 }, { "grad_norm": 0.17715556919574738, "learning_rate": 8.274315029794454e-05, "loss": 0.0175, "step": 30910 }, { "grad_norm": 0.138163223862648, "learning_rate": 8.273065244924443e-05, "loss": 0.0173, "step": 30920 }, { "grad_norm": 0.14773309230804443, "learning_rate": 8.271815102116936e-05, "loss": 0.0132, "step": 30930 }, { "grad_norm": 0.13050726056098938, "learning_rate": 8.270564601508648e-05, "loss": 0.0146, "step": 30940 }, { "grad_norm": 0.12301746755838394, "learning_rate": 8.269313743236333e-05, "loss": 0.0184, "step": 30950 }, { "grad_norm": 0.12385182827711105, "learning_rate": 8.268062527436783e-05, "loss": 0.0142, "step": 30960 }, { "grad_norm": 0.2157222181558609, "learning_rate": 8.266810954246828e-05, "loss": 0.0165, "step": 30970 }, { "grad_norm": 0.15300413966178894, "learning_rate": 8.265559023803338e-05, "loss": 0.0137, "step": 30980 }, { "grad_norm": 0.2020850032567978, "learning_rate": 8.264306736243223e-05, "loss": 0.0148, "step": 30990 }, { "grad_norm": 0.1912722885608673, "learning_rate": 8.263054091703432e-05, "loss": 0.0168, "step": 31000 }, { "grad_norm": 0.130269393324852, "learning_rate": 8.26180109032095e-05, "loss": 0.0148, "step": 31010 }, { "grad_norm": 0.11475087702274323, "learning_rate": 8.260547732232807e-05, "loss": 0.0111, "step": 31020 }, { "grad_norm": 0.1490316241979599, "learning_rate": 8.259294017576063e-05, "loss": 0.0154, "step": 31030 }, { "grad_norm": 0.20377130806446075, "learning_rate": 8.258039946487828e-05, "loss": 0.0171, "step": 31040 }, { "grad_norm": 0.10543281584978104, "learning_rate": 8.256785519105241e-05, "loss": 0.0134, "step": 31050 }, { "grad_norm": 0.13591758906841278, "learning_rate": 8.255530735565488e-05, "loss": 0.0107, "step": 31060 }, { "grad_norm": 0.1661704182624817, "learning_rate": 8.254275596005787e-05, "loss": 0.0142, "step": 31070 }, { "grad_norm": 0.18815860152244568, "learning_rate": 8.2530201005634e-05, "loss": 0.0126, "step": 31080 }, { "grad_norm": 0.12632235884666443, "learning_rate": 8.251764249375626e-05, "loss": 0.0179, "step": 31090 }, { "grad_norm": 0.18896958231925964, "learning_rate": 8.250508042579803e-05, "loss": 0.0167, "step": 31100 }, { "grad_norm": 0.25462377071380615, "learning_rate": 8.249251480313307e-05, "loss": 0.0177, "step": 31110 }, { "grad_norm": 0.24988770484924316, "learning_rate": 8.247994562713555e-05, "loss": 0.0154, "step": 31120 }, { "grad_norm": 0.18553583323955536, "learning_rate": 8.246737289917998e-05, "loss": 0.0148, "step": 31130 }, { "grad_norm": 0.14271780848503113, "learning_rate": 8.245479662064135e-05, "loss": 0.0158, "step": 31140 }, { "grad_norm": 0.13061584532260895, "learning_rate": 8.244221679289496e-05, "loss": 0.0127, "step": 31150 }, { "grad_norm": 0.14954368770122528, "learning_rate": 8.242963341731652e-05, "loss": 0.0137, "step": 31160 }, { "grad_norm": 0.18185393512248993, "learning_rate": 8.24170464952821e-05, "loss": 0.0156, "step": 31170 }, { "grad_norm": 0.18166691064834595, "learning_rate": 8.240445602816824e-05, "loss": 0.0206, "step": 31180 }, { "grad_norm": 0.09820786118507385, "learning_rate": 8.239186201735179e-05, "loss": 0.0133, "step": 31190 }, { "grad_norm": 0.14792008697986603, "learning_rate": 8.237926446420998e-05, "loss": 0.0112, "step": 31200 }, { "grad_norm": 0.1363881528377533, "learning_rate": 8.236666337012052e-05, "loss": 0.0133, "step": 31210 }, { "grad_norm": 0.10949233174324036, "learning_rate": 8.23540587364614e-05, "loss": 0.0126, "step": 31220 }, { "grad_norm": 0.12311102449893951, "learning_rate": 8.234145056461107e-05, "loss": 0.0149, "step": 31230 }, { "grad_norm": 0.16510125994682312, "learning_rate": 8.232883885594832e-05, "loss": 0.0204, "step": 31240 }, { "grad_norm": 0.19816899299621582, "learning_rate": 8.231622361185236e-05, "loss": 0.0158, "step": 31250 }, { "grad_norm": 0.3181968331336975, "learning_rate": 8.230360483370278e-05, "loss": 0.0294, "step": 31260 }, { "grad_norm": 0.12098640948534012, "learning_rate": 8.229098252287953e-05, "loss": 0.0115, "step": 31270 }, { "grad_norm": 0.10433916002511978, "learning_rate": 8.2278356680763e-05, "loss": 0.0218, "step": 31280 }, { "grad_norm": 0.2293347418308258, "learning_rate": 8.22657273087339e-05, "loss": 0.0173, "step": 31290 }, { "grad_norm": 0.3142009377479553, "learning_rate": 8.225309440817336e-05, "loss": 0.0236, "step": 31300 }, { "grad_norm": 0.16976268589496613, "learning_rate": 8.224045798046293e-05, "loss": 0.0315, "step": 31310 }, { "grad_norm": 0.2196931391954422, "learning_rate": 8.22278180269845e-05, "loss": 0.0147, "step": 31320 }, { "grad_norm": 0.16628876328468323, "learning_rate": 8.221517454912031e-05, "loss": 0.0136, "step": 31330 }, { "grad_norm": 0.14589914679527283, "learning_rate": 8.220252754825308e-05, "loss": 0.0151, "step": 31340 }, { "grad_norm": 0.12026769667863846, "learning_rate": 8.218987702576586e-05, "loss": 0.0151, "step": 31350 }, { "grad_norm": 0.17089566588401794, "learning_rate": 8.217722298304207e-05, "loss": 0.0209, "step": 31360 }, { "grad_norm": 0.1741703450679779, "learning_rate": 8.216456542146557e-05, "loss": 0.0182, "step": 31370 }, { "grad_norm": 0.15633898973464966, "learning_rate": 8.215190434242055e-05, "loss": 0.0141, "step": 31380 }, { "grad_norm": 0.18184731900691986, "learning_rate": 8.213923974729161e-05, "loss": 0.0154, "step": 31390 }, { "grad_norm": 0.113535076379776, "learning_rate": 8.212657163746373e-05, "loss": 0.0126, "step": 31400 }, { "grad_norm": 0.15817177295684814, "learning_rate": 8.211390001432227e-05, "loss": 0.0138, "step": 31410 }, { "grad_norm": 0.17843103408813477, "learning_rate": 8.210122487925297e-05, "loss": 0.0123, "step": 31420 }, { "grad_norm": 0.23402860760688782, "learning_rate": 8.208854623364202e-05, "loss": 0.0159, "step": 31430 }, { "grad_norm": 0.15226945281028748, "learning_rate": 8.207586407887589e-05, "loss": 0.0136, "step": 31440 }, { "grad_norm": 0.18373161554336548, "learning_rate": 8.206317841634148e-05, "loss": 0.0154, "step": 31450 }, { "grad_norm": 0.17162111401557922, "learning_rate": 8.205048924742609e-05, "loss": 0.0158, "step": 31460 }, { "grad_norm": 0.1260475218296051, "learning_rate": 8.203779657351738e-05, "loss": 0.0112, "step": 31470 }, { "grad_norm": 0.15085278451442719, "learning_rate": 8.20251003960034e-05, "loss": 0.0179, "step": 31480 }, { "grad_norm": 0.13494142889976501, "learning_rate": 8.201240071627258e-05, "loss": 0.017, "step": 31490 }, { "grad_norm": 0.18428562581539154, "learning_rate": 8.199969753571377e-05, "loss": 0.0186, "step": 31500 }, { "grad_norm": 0.21965456008911133, "learning_rate": 8.198699085571615e-05, "loss": 0.0149, "step": 31510 }, { "grad_norm": 0.21447010338306427, "learning_rate": 8.197428067766928e-05, "loss": 0.0146, "step": 31520 }, { "grad_norm": 0.1798044890165329, "learning_rate": 8.196156700296316e-05, "loss": 0.0196, "step": 31530 }, { "grad_norm": 0.18135182559490204, "learning_rate": 8.194884983298814e-05, "loss": 0.0137, "step": 31540 }, { "grad_norm": 0.16148513555526733, "learning_rate": 8.193612916913491e-05, "loss": 0.0143, "step": 31550 }, { "grad_norm": 0.1736438125371933, "learning_rate": 8.192340501279463e-05, "loss": 0.0153, "step": 31560 }, { "grad_norm": 0.13550296425819397, "learning_rate": 8.191067736535876e-05, "loss": 0.0162, "step": 31570 }, { "grad_norm": 0.1617557555437088, "learning_rate": 8.18979462282192e-05, "loss": 0.0125, "step": 31580 }, { "grad_norm": 0.15130682289600372, "learning_rate": 8.188521160276819e-05, "loss": 0.0139, "step": 31590 }, { "grad_norm": 0.1068098172545433, "learning_rate": 8.187247349039837e-05, "loss": 0.0115, "step": 31600 }, { "grad_norm": 0.23565487563610077, "learning_rate": 8.185973189250278e-05, "loss": 0.0186, "step": 31610 }, { "grad_norm": 0.19959557056427002, "learning_rate": 8.184698681047482e-05, "loss": 0.018, "step": 31620 }, { "grad_norm": 0.12909047305583954, "learning_rate": 8.183423824570827e-05, "loss": 0.0148, "step": 31630 }, { "grad_norm": 0.16234177350997925, "learning_rate": 8.182148619959725e-05, "loss": 0.0167, "step": 31640 }, { "grad_norm": 0.12469124794006348, "learning_rate": 8.180873067353636e-05, "loss": 0.0161, "step": 31650 }, { "grad_norm": 0.16591891646385193, "learning_rate": 8.179597166892052e-05, "loss": 0.015, "step": 31660 }, { "grad_norm": 0.1807849109172821, "learning_rate": 8.178320918714501e-05, "loss": 0.0144, "step": 31670 }, { "grad_norm": 0.13133931159973145, "learning_rate": 8.177044322960554e-05, "loss": 0.014, "step": 31680 }, { "grad_norm": 0.13954201340675354, "learning_rate": 8.175767379769816e-05, "loss": 0.0118, "step": 31690 }, { "grad_norm": 0.14169637858867645, "learning_rate": 8.174490089281932e-05, "loss": 0.0142, "step": 31700 }, { "grad_norm": 0.11382963508367538, "learning_rate": 8.173212451636584e-05, "loss": 0.0153, "step": 31710 }, { "grad_norm": 0.17956571280956268, "learning_rate": 8.171934466973493e-05, "loss": 0.0189, "step": 31720 }, { "grad_norm": 0.17194052040576935, "learning_rate": 8.170656135432418e-05, "loss": 0.0134, "step": 31730 }, { "grad_norm": 0.1509227603673935, "learning_rate": 8.169377457153155e-05, "loss": 0.0179, "step": 31740 }, { "grad_norm": 0.1722368597984314, "learning_rate": 8.168098432275539e-05, "loss": 0.0147, "step": 31750 }, { "grad_norm": 0.1772075593471527, "learning_rate": 8.166819060939442e-05, "loss": 0.0179, "step": 31760 }, { "grad_norm": 0.20290528237819672, "learning_rate": 8.165539343284772e-05, "loss": 0.0182, "step": 31770 }, { "grad_norm": 0.20561754703521729, "learning_rate": 8.16425927945148e-05, "loss": 0.0196, "step": 31780 }, { "grad_norm": 0.1840229630470276, "learning_rate": 8.162978869579551e-05, "loss": 0.0284, "step": 31790 }, { "grad_norm": 0.15248467028141022, "learning_rate": 8.161698113809007e-05, "loss": 0.0158, "step": 31800 }, { "grad_norm": 0.17546291649341583, "learning_rate": 8.160417012279911e-05, "loss": 0.0261, "step": 31810 }, { "grad_norm": 0.1356450319290161, "learning_rate": 8.159135565132363e-05, "loss": 0.0144, "step": 31820 }, { "grad_norm": 0.1405523270368576, "learning_rate": 8.157853772506498e-05, "loss": 0.0227, "step": 31830 }, { "grad_norm": 0.14576540887355804, "learning_rate": 8.156571634542494e-05, "loss": 0.0239, "step": 31840 }, { "grad_norm": 0.23554202914237976, "learning_rate": 8.15528915138056e-05, "loss": 0.0199, "step": 31850 }, { "grad_norm": 0.152358278632164, "learning_rate": 8.154006323160949e-05, "loss": 0.0151, "step": 31860 }, { "grad_norm": 0.13935130834579468, "learning_rate": 8.152723150023949e-05, "loss": 0.0179, "step": 31870 }, { "grad_norm": 0.2645920217037201, "learning_rate": 8.151439632109886e-05, "loss": 0.0171, "step": 31880 }, { "grad_norm": 0.16194036602973938, "learning_rate": 8.150155769559122e-05, "loss": 0.0301, "step": 31890 }, { "grad_norm": 0.10884969681501389, "learning_rate": 8.148871562512058e-05, "loss": 0.015, "step": 31900 }, { "grad_norm": 0.14641964435577393, "learning_rate": 8.147587011109136e-05, "loss": 0.0235, "step": 31910 }, { "grad_norm": 0.18850119411945343, "learning_rate": 8.14630211549083e-05, "loss": 0.0312, "step": 31920 }, { "grad_norm": 0.11385685205459595, "learning_rate": 8.145016875797655e-05, "loss": 0.0142, "step": 31930 }, { "grad_norm": 0.10187400877475739, "learning_rate": 8.143731292170164e-05, "loss": 0.0139, "step": 31940 }, { "grad_norm": 0.17855457961559296, "learning_rate": 8.142445364748944e-05, "loss": 0.0184, "step": 31950 }, { "grad_norm": 0.20401360094547272, "learning_rate": 8.141159093674624e-05, "loss": 0.0197, "step": 31960 }, { "grad_norm": 0.1485712230205536, "learning_rate": 8.139872479087869e-05, "loss": 0.0147, "step": 31970 }, { "grad_norm": 0.12945401668548584, "learning_rate": 8.13858552112938e-05, "loss": 0.012, "step": 31980 }, { "grad_norm": 0.18906298279762268, "learning_rate": 8.137298219939895e-05, "loss": 0.0212, "step": 31990 }, { "grad_norm": 0.1515437215566635, "learning_rate": 8.136010575660196e-05, "loss": 0.0129, "step": 32000 }, { "grad_norm": 0.18659307062625885, "learning_rate": 8.134722588431095e-05, "loss": 0.0194, "step": 32010 }, { "grad_norm": 0.15512140095233917, "learning_rate": 8.133434258393444e-05, "loss": 0.0178, "step": 32020 }, { "grad_norm": 0.16848373413085938, "learning_rate": 8.132145585688134e-05, "loss": 0.0165, "step": 32030 }, { "grad_norm": 0.16092807054519653, "learning_rate": 8.130856570456093e-05, "loss": 0.0171, "step": 32040 }, { "grad_norm": 0.11548453569412231, "learning_rate": 8.129567212838283e-05, "loss": 0.0158, "step": 32050 }, { "grad_norm": 0.19681474566459656, "learning_rate": 8.128277512975708e-05, "loss": 0.0192, "step": 32060 }, { "grad_norm": 0.2885902225971222, "learning_rate": 8.126987471009408e-05, "loss": 0.0171, "step": 32070 }, { "grad_norm": 0.14149731397628784, "learning_rate": 8.125697087080459e-05, "loss": 0.0109, "step": 32080 }, { "grad_norm": 0.1903044730424881, "learning_rate": 8.124406361329976e-05, "loss": 0.0198, "step": 32090 }, { "grad_norm": 0.17306911945343018, "learning_rate": 8.123115293899111e-05, "loss": 0.0157, "step": 32100 }, { "grad_norm": 0.1235608458518982, "learning_rate": 8.121823884929055e-05, "loss": 0.014, "step": 32110 }, { "grad_norm": 0.10265597701072693, "learning_rate": 8.12053213456103e-05, "loss": 0.0129, "step": 32120 }, { "grad_norm": 0.1933118999004364, "learning_rate": 8.119240042936303e-05, "loss": 0.0162, "step": 32130 }, { "grad_norm": 0.21186469495296478, "learning_rate": 8.117947610196175e-05, "loss": 0.0144, "step": 32140 }, { "grad_norm": 0.1764698475599289, "learning_rate": 8.116654836481982e-05, "loss": 0.0124, "step": 32150 }, { "grad_norm": 0.16171178221702576, "learning_rate": 8.115361721935106e-05, "loss": 0.0125, "step": 32160 }, { "grad_norm": 0.13194194436073303, "learning_rate": 8.114068266696953e-05, "loss": 0.016, "step": 32170 }, { "grad_norm": 0.11153806000947952, "learning_rate": 8.112774470908978e-05, "loss": 0.0117, "step": 32180 }, { "grad_norm": 0.11475368589162827, "learning_rate": 8.111480334712665e-05, "loss": 0.0111, "step": 32190 }, { "grad_norm": 0.14734667539596558, "learning_rate": 8.110185858249542e-05, "loss": 0.0139, "step": 32200 }, { "grad_norm": 0.24344736337661743, "learning_rate": 8.108891041661168e-05, "loss": 0.0176, "step": 32210 }, { "grad_norm": 0.14460216462612152, "learning_rate": 8.107595885089146e-05, "loss": 0.0142, "step": 32220 }, { "grad_norm": 0.1478167176246643, "learning_rate": 8.106300388675108e-05, "loss": 0.0128, "step": 32230 }, { "grad_norm": 0.19024783372879028, "learning_rate": 8.105004552560731e-05, "loss": 0.0178, "step": 32240 }, { "grad_norm": 0.21320463716983795, "learning_rate": 8.103708376887724e-05, "loss": 0.0193, "step": 32250 }, { "grad_norm": 0.16703885793685913, "learning_rate": 8.102411861797836e-05, "loss": 0.0152, "step": 32260 }, { "grad_norm": 0.16709262132644653, "learning_rate": 8.101115007432851e-05, "loss": 0.0137, "step": 32270 }, { "grad_norm": 0.9113238453865051, "learning_rate": 8.09981781393459e-05, "loss": 0.0108, "step": 32280 }, { "grad_norm": 0.15839727222919464, "learning_rate": 8.098520281444915e-05, "loss": 0.0142, "step": 32290 }, { "grad_norm": 0.1830999106168747, "learning_rate": 8.09722241010572e-05, "loss": 0.0141, "step": 32300 }, { "grad_norm": 0.14024698734283447, "learning_rate": 8.095924200058939e-05, "loss": 0.0186, "step": 32310 }, { "grad_norm": 0.13402022421360016, "learning_rate": 8.094625651446541e-05, "loss": 0.0107, "step": 32320 }, { "grad_norm": 0.1124209612607956, "learning_rate": 8.093326764410536e-05, "loss": 0.0129, "step": 32330 }, { "grad_norm": 0.10891943424940109, "learning_rate": 8.092027539092966e-05, "loss": 0.0147, "step": 32340 }, { "grad_norm": 0.1546153575181961, "learning_rate": 8.090727975635913e-05, "loss": 0.0127, "step": 32350 }, { "grad_norm": 0.1616239696741104, "learning_rate": 8.089428074181497e-05, "loss": 0.0131, "step": 32360 }, { "grad_norm": 0.14622797071933746, "learning_rate": 8.088127834871871e-05, "loss": 0.013, "step": 32370 }, { "grad_norm": 0.11192550510168076, "learning_rate": 8.086827257849226e-05, "loss": 0.0147, "step": 32380 }, { "grad_norm": 0.0918840616941452, "learning_rate": 8.085526343255795e-05, "loss": 0.0093, "step": 32390 }, { "grad_norm": 0.19642551243305206, "learning_rate": 8.084225091233842e-05, "loss": 0.0142, "step": 32400 }, { "grad_norm": 0.19060373306274414, "learning_rate": 8.082923501925668e-05, "loss": 0.0139, "step": 32410 }, { "grad_norm": 0.14834727346897125, "learning_rate": 8.081621575473617e-05, "loss": 0.015, "step": 32420 }, { "grad_norm": 0.162302166223526, "learning_rate": 8.080319312020064e-05, "loss": 0.0201, "step": 32430 }, { "grad_norm": 0.15315553545951843, "learning_rate": 8.079016711707421e-05, "loss": 0.0127, "step": 32440 }, { "grad_norm": 0.14873041212558746, "learning_rate": 8.077713774678139e-05, "loss": 0.0146, "step": 32450 }, { "grad_norm": 0.15412671864032745, "learning_rate": 8.076410501074707e-05, "loss": 0.0129, "step": 32460 }, { "grad_norm": 0.19625645875930786, "learning_rate": 8.075106891039647e-05, "loss": 0.0198, "step": 32470 }, { "grad_norm": 0.21761170029640198, "learning_rate": 8.073802944715523e-05, "loss": 0.0161, "step": 32480 }, { "grad_norm": 0.24644440412521362, "learning_rate": 8.072498662244929e-05, "loss": 0.0134, "step": 32490 }, { "grad_norm": 0.1760413497686386, "learning_rate": 8.0711940437705e-05, "loss": 0.0151, "step": 32500 }, { "grad_norm": 0.13550330698490143, "learning_rate": 8.06988908943491e-05, "loss": 0.0119, "step": 32510 }, { "grad_norm": 0.11196909844875336, "learning_rate": 8.068583799380863e-05, "loss": 0.0116, "step": 32520 }, { "grad_norm": 0.12690602242946625, "learning_rate": 8.067278173751104e-05, "loss": 0.0121, "step": 32530 }, { "grad_norm": 0.25147783756256104, "learning_rate": 8.065972212688417e-05, "loss": 0.0156, "step": 32540 }, { "grad_norm": 0.16766630113124847, "learning_rate": 8.064665916335618e-05, "loss": 0.0198, "step": 32550 }, { "grad_norm": 0.13899868726730347, "learning_rate": 8.063359284835564e-05, "loss": 0.0164, "step": 32560 }, { "grad_norm": 0.15217484533786774, "learning_rate": 8.062052318331142e-05, "loss": 0.0168, "step": 32570 }, { "grad_norm": 0.1838143914937973, "learning_rate": 8.060745016965283e-05, "loss": 0.0149, "step": 32580 }, { "grad_norm": 0.1732724905014038, "learning_rate": 8.059437380880952e-05, "loss": 0.0142, "step": 32590 }, { "grad_norm": 0.16025635600090027, "learning_rate": 8.058129410221146e-05, "loss": 0.0166, "step": 32600 }, { "grad_norm": 0.14937523007392883, "learning_rate": 8.056821105128908e-05, "loss": 0.0156, "step": 32610 }, { "grad_norm": 0.190080925822258, "learning_rate": 8.05551246574731e-05, "loss": 0.0166, "step": 32620 }, { "grad_norm": 0.12561103701591492, "learning_rate": 8.05420349221946e-05, "loss": 0.0146, "step": 32630 }, { "grad_norm": 0.1270051896572113, "learning_rate": 8.05289418468851e-05, "loss": 0.014, "step": 32640 }, { "grad_norm": 0.14429400861263275, "learning_rate": 8.051584543297642e-05, "loss": 0.0152, "step": 32650 }, { "grad_norm": 0.2137008011341095, "learning_rate": 8.050274568190074e-05, "loss": 0.013, "step": 32660 }, { "grad_norm": 0.16504834592342377, "learning_rate": 8.048964259509067e-05, "loss": 0.0181, "step": 32670 }, { "grad_norm": 0.19447214901447296, "learning_rate": 8.047653617397914e-05, "loss": 0.0141, "step": 32680 }, { "grad_norm": 0.262847900390625, "learning_rate": 8.046342641999941e-05, "loss": 0.0171, "step": 32690 }, { "grad_norm": 0.19874665141105652, "learning_rate": 8.045031333458517e-05, "loss": 0.0157, "step": 32700 }, { "grad_norm": 0.2847769260406494, "learning_rate": 8.043719691917047e-05, "loss": 0.0204, "step": 32710 }, { "grad_norm": 0.13497519493103027, "learning_rate": 8.042407717518966e-05, "loss": 0.0133, "step": 32720 }, { "grad_norm": 0.13901059329509735, "learning_rate": 8.041095410407751e-05, "loss": 0.0162, "step": 32730 }, { "grad_norm": 0.1521959751844406, "learning_rate": 8.039782770726913e-05, "loss": 0.0165, "step": 32740 }, { "grad_norm": 0.11719818413257599, "learning_rate": 8.038469798620004e-05, "loss": 0.0149, "step": 32750 }, { "grad_norm": 0.20886211097240448, "learning_rate": 8.037156494230604e-05, "loss": 0.0142, "step": 32760 }, { "grad_norm": 0.15914329886436462, "learning_rate": 8.035842857702338e-05, "loss": 0.0145, "step": 32770 }, { "grad_norm": 0.1761186420917511, "learning_rate": 8.03452888917886e-05, "loss": 0.0175, "step": 32780 }, { "grad_norm": 0.11725594848394394, "learning_rate": 8.033214588803866e-05, "loss": 0.0116, "step": 32790 }, { "grad_norm": 0.1616867631673813, "learning_rate": 8.031899956721083e-05, "loss": 0.0143, "step": 32800 }, { "grad_norm": 0.18697018921375275, "learning_rate": 8.030584993074282e-05, "loss": 0.0128, "step": 32810 }, { "grad_norm": 0.1089986190199852, "learning_rate": 8.02926969800726e-05, "loss": 0.013, "step": 32820 }, { "grad_norm": 0.08184492588043213, "learning_rate": 8.027954071663859e-05, "loss": 0.019, "step": 32830 }, { "grad_norm": 0.15176475048065186, "learning_rate": 8.026638114187954e-05, "loss": 0.0129, "step": 32840 }, { "grad_norm": 0.17511162161827087, "learning_rate": 8.025321825723456e-05, "loss": 0.019, "step": 32850 }, { "grad_norm": 0.12988655269145966, "learning_rate": 8.02400520641431e-05, "loss": 0.0183, "step": 32860 }, { "grad_norm": 0.1399833709001541, "learning_rate": 8.022688256404501e-05, "loss": 0.0123, "step": 32870 }, { "grad_norm": 0.11862239241600037, "learning_rate": 8.02137097583805e-05, "loss": 0.0139, "step": 32880 }, { "grad_norm": 0.1452031433582306, "learning_rate": 8.02005336485901e-05, "loss": 0.0105, "step": 32890 }, { "grad_norm": 0.16126877069473267, "learning_rate": 8.018735423611476e-05, "loss": 0.0168, "step": 32900 }, { "grad_norm": 0.11408356577157974, "learning_rate": 8.017417152239574e-05, "loss": 0.0151, "step": 32910 }, { "grad_norm": 0.1243429034948349, "learning_rate": 8.01609855088747e-05, "loss": 0.0154, "step": 32920 }, { "grad_norm": 0.1435280740261078, "learning_rate": 8.014779619699362e-05, "loss": 0.0123, "step": 32930 }, { "grad_norm": 0.13208629190921783, "learning_rate": 8.013460358819489e-05, "loss": 0.0138, "step": 32940 }, { "grad_norm": 0.13685470819473267, "learning_rate": 8.01214076839212e-05, "loss": 0.015, "step": 32950 }, { "grad_norm": 0.10536746680736542, "learning_rate": 8.010820848561565e-05, "loss": 0.0134, "step": 32960 }, { "grad_norm": 0.12247555702924728, "learning_rate": 8.009500599472171e-05, "loss": 0.0132, "step": 32970 }, { "grad_norm": 0.21592576801776886, "learning_rate": 8.008180021268314e-05, "loss": 0.0121, "step": 32980 }, { "grad_norm": 0.12795771658420563, "learning_rate": 8.006859114094414e-05, "loss": 0.0134, "step": 32990 }, { "grad_norm": 0.1677226424217224, "learning_rate": 8.005537878094921e-05, "loss": 0.0143, "step": 33000 }, { "grad_norm": 0.14107894897460938, "learning_rate": 8.004216313414323e-05, "loss": 0.0164, "step": 33010 }, { "grad_norm": 0.10726600140333176, "learning_rate": 8.002894420197149e-05, "loss": 0.0107, "step": 33020 }, { "grad_norm": 0.14911052584648132, "learning_rate": 8.001572198587954e-05, "loss": 0.0147, "step": 33030 }, { "grad_norm": 0.1271044909954071, "learning_rate": 8.000249648731338e-05, "loss": 0.0118, "step": 33040 }, { "grad_norm": 0.29633039236068726, "learning_rate": 7.998926770771928e-05, "loss": 0.0208, "step": 33050 }, { "grad_norm": 0.1823190152645111, "learning_rate": 7.997603564854397e-05, "loss": 0.0182, "step": 33060 }, { "grad_norm": 0.19431470334529877, "learning_rate": 7.996280031123448e-05, "loss": 0.0173, "step": 33070 }, { "grad_norm": 0.21462175250053406, "learning_rate": 7.994956169723818e-05, "loss": 0.0257, "step": 33080 }, { "grad_norm": 0.15341973304748535, "learning_rate": 7.993631980800285e-05, "loss": 0.0135, "step": 33090 }, { "grad_norm": 0.152894526720047, "learning_rate": 7.992307464497659e-05, "loss": 0.02, "step": 33100 }, { "grad_norm": 0.13849857449531555, "learning_rate": 7.990982620960787e-05, "loss": 0.014, "step": 33110 }, { "grad_norm": 0.11612430959939957, "learning_rate": 7.989657450334554e-05, "loss": 0.0143, "step": 33120 }, { "grad_norm": 0.16003359854221344, "learning_rate": 7.988331952763877e-05, "loss": 0.0152, "step": 33130 }, { "grad_norm": 0.18860261142253876, "learning_rate": 7.987006128393709e-05, "loss": 0.0227, "step": 33140 }, { "grad_norm": 0.11668971180915833, "learning_rate": 7.985679977369043e-05, "loss": 0.0116, "step": 33150 }, { "grad_norm": 0.16676343977451324, "learning_rate": 7.984353499834902e-05, "loss": 0.0153, "step": 33160 }, { "grad_norm": 0.14033178985118866, "learning_rate": 7.983026695936351e-05, "loss": 0.0124, "step": 33170 }, { "grad_norm": 0.2154844105243683, "learning_rate": 7.981699565818486e-05, "loss": 0.0135, "step": 33180 }, { "grad_norm": 0.15608274936676025, "learning_rate": 7.980372109626437e-05, "loss": 0.0102, "step": 33190 }, { "grad_norm": 0.13301432132720947, "learning_rate": 7.979044327505375e-05, "loss": 0.0126, "step": 33200 }, { "grad_norm": 0.2349739819765091, "learning_rate": 7.977716219600506e-05, "loss": 0.0161, "step": 33210 }, { "grad_norm": 0.13931939005851746, "learning_rate": 7.97638778605707e-05, "loss": 0.0158, "step": 33220 }, { "grad_norm": 0.19417789578437805, "learning_rate": 7.975059027020338e-05, "loss": 0.0264, "step": 33230 }, { "grad_norm": 0.16319182515144348, "learning_rate": 7.973729942635623e-05, "loss": 0.0138, "step": 33240 }, { "grad_norm": 0.11241982132196426, "learning_rate": 7.972400533048273e-05, "loss": 0.0168, "step": 33250 }, { "grad_norm": 0.1439264714717865, "learning_rate": 7.97107079840367e-05, "loss": 0.0174, "step": 33260 }, { "grad_norm": 0.19337862730026245, "learning_rate": 7.969740738847231e-05, "loss": 0.0165, "step": 33270 }, { "grad_norm": 0.22109273076057434, "learning_rate": 7.968410354524411e-05, "loss": 0.0167, "step": 33280 }, { "grad_norm": 0.11708156764507294, "learning_rate": 7.967079645580697e-05, "loss": 0.0137, "step": 33290 }, { "grad_norm": 0.17506004869937897, "learning_rate": 7.965748612161612e-05, "loss": 0.0155, "step": 33300 }, { "grad_norm": 0.17507526278495789, "learning_rate": 7.96441725441272e-05, "loss": 0.0186, "step": 33310 }, { "grad_norm": 0.1812824010848999, "learning_rate": 7.963085572479614e-05, "loss": 0.0136, "step": 33320 }, { "grad_norm": 0.1448460966348648, "learning_rate": 7.961753566507924e-05, "loss": 0.0132, "step": 33330 }, { "grad_norm": 0.12789112329483032, "learning_rate": 7.960421236643316e-05, "loss": 0.0171, "step": 33340 }, { "grad_norm": 0.23540131747722626, "learning_rate": 7.959088583031496e-05, "loss": 0.0142, "step": 33350 }, { "grad_norm": 0.12215933948755264, "learning_rate": 7.957755605818194e-05, "loss": 0.0126, "step": 33360 }, { "grad_norm": 0.11950993537902832, "learning_rate": 7.956422305149185e-05, "loss": 0.0157, "step": 33370 }, { "grad_norm": 0.17512483894824982, "learning_rate": 7.95508868117028e-05, "loss": 0.0173, "step": 33380 }, { "grad_norm": 0.22466380894184113, "learning_rate": 7.953754734027318e-05, "loss": 0.0165, "step": 33390 }, { "grad_norm": 0.12851156294345856, "learning_rate": 7.952420463866182e-05, "loss": 0.0118, "step": 33400 }, { "grad_norm": 0.11967978626489639, "learning_rate": 7.951085870832782e-05, "loss": 0.015, "step": 33410 }, { "grad_norm": 0.13838866353034973, "learning_rate": 7.949750955073067e-05, "loss": 0.0128, "step": 33420 }, { "grad_norm": 0.18538019061088562, "learning_rate": 7.948415716733022e-05, "loss": 0.0149, "step": 33430 }, { "grad_norm": 0.13369233906269073, "learning_rate": 7.947080155958669e-05, "loss": 0.0144, "step": 33440 }, { "grad_norm": 0.14633123576641083, "learning_rate": 7.94574427289606e-05, "loss": 0.011, "step": 33450 }, { "grad_norm": 0.13192565739154816, "learning_rate": 7.944408067691284e-05, "loss": 0.013, "step": 33460 }, { "grad_norm": 0.12698668241500854, "learning_rate": 7.943071540490473e-05, "loss": 0.0115, "step": 33470 }, { "grad_norm": 0.14076174795627594, "learning_rate": 7.94173469143978e-05, "loss": 0.0131, "step": 33480 }, { "grad_norm": 0.10067231953144073, "learning_rate": 7.940397520685406e-05, "loss": 0.0119, "step": 33490 }, { "grad_norm": 0.18096619844436646, "learning_rate": 7.939060028373577e-05, "loss": 0.0161, "step": 33500 }, { "grad_norm": 0.1332755833864212, "learning_rate": 7.937722214650565e-05, "loss": 0.016, "step": 33510 }, { "grad_norm": 0.1336979866027832, "learning_rate": 7.936384079662666e-05, "loss": 0.0124, "step": 33520 }, { "grad_norm": 0.1420097053050995, "learning_rate": 7.93504562355622e-05, "loss": 0.0153, "step": 33530 }, { "grad_norm": 0.12370989471673965, "learning_rate": 7.933706846477599e-05, "loss": 0.0137, "step": 33540 }, { "grad_norm": 0.17912912368774414, "learning_rate": 7.932367748573206e-05, "loss": 0.016, "step": 33550 }, { "grad_norm": 0.15298885107040405, "learning_rate": 7.931028329989485e-05, "loss": 0.0151, "step": 33560 }, { "grad_norm": 0.16450342535972595, "learning_rate": 7.929688590872913e-05, "loss": 0.017, "step": 33570 }, { "grad_norm": 0.11823868006467819, "learning_rate": 7.928348531370003e-05, "loss": 0.0109, "step": 33580 }, { "grad_norm": 0.08456481248140335, "learning_rate": 7.927008151627297e-05, "loss": 0.0092, "step": 33590 }, { "grad_norm": 0.09376697987318039, "learning_rate": 7.925667451791383e-05, "loss": 0.0112, "step": 33600 }, { "grad_norm": 0.1728041023015976, "learning_rate": 7.924326432008874e-05, "loss": 0.0115, "step": 33610 }, { "grad_norm": 0.16240081191062927, "learning_rate": 7.922985092426422e-05, "loss": 0.0144, "step": 33620 }, { "grad_norm": 0.1628105640411377, "learning_rate": 7.921643433190717e-05, "loss": 0.0141, "step": 33630 }, { "grad_norm": 0.12551337480545044, "learning_rate": 7.920301454448478e-05, "loss": 0.0148, "step": 33640 }, { "grad_norm": 0.17008866369724274, "learning_rate": 7.918959156346461e-05, "loss": 0.0134, "step": 33650 }, { "grad_norm": 0.15027743577957153, "learning_rate": 7.91761653903146e-05, "loss": 0.0145, "step": 33660 }, { "grad_norm": 0.20891068875789642, "learning_rate": 7.916273602650302e-05, "loss": 0.0177, "step": 33670 }, { "grad_norm": 0.1683206707239151, "learning_rate": 7.914930347349847e-05, "loss": 0.0121, "step": 33680 }, { "grad_norm": 0.1405433863401413, "learning_rate": 7.913586773276992e-05, "loss": 0.014, "step": 33690 }, { "grad_norm": 0.15677045285701752, "learning_rate": 7.912242880578667e-05, "loss": 0.0112, "step": 33700 }, { "grad_norm": 0.12789367139339447, "learning_rate": 7.910898669401839e-05, "loss": 0.0133, "step": 33710 }, { "grad_norm": 0.16764597594738007, "learning_rate": 7.909554139893511e-05, "loss": 0.0139, "step": 33720 }, { "grad_norm": 0.14478057622909546, "learning_rate": 7.908209292200715e-05, "loss": 0.0135, "step": 33730 }, { "grad_norm": 0.14522898197174072, "learning_rate": 7.906864126470523e-05, "loss": 0.0099, "step": 33740 }, { "grad_norm": 0.16912995278835297, "learning_rate": 7.905518642850041e-05, "loss": 0.013, "step": 33750 }, { "grad_norm": 0.15117940306663513, "learning_rate": 7.904172841486409e-05, "loss": 0.0158, "step": 33760 }, { "grad_norm": 0.23264512419700623, "learning_rate": 7.902826722526801e-05, "loss": 0.015, "step": 33770 }, { "grad_norm": 0.10520673543214798, "learning_rate": 7.901480286118427e-05, "loss": 0.0129, "step": 33780 }, { "grad_norm": 0.14356227219104767, "learning_rate": 7.900133532408531e-05, "loss": 0.0149, "step": 33790 }, { "grad_norm": 0.16817203164100647, "learning_rate": 7.898786461544395e-05, "loss": 0.0138, "step": 33800 }, { "grad_norm": 0.09324843436479568, "learning_rate": 7.897439073673325e-05, "loss": 0.0117, "step": 33810 }, { "grad_norm": 0.1287604421377182, "learning_rate": 7.896091368942677e-05, "loss": 0.0127, "step": 33820 }, { "grad_norm": 0.08109401166439056, "learning_rate": 7.894743347499832e-05, "loss": 0.0124, "step": 33830 }, { "grad_norm": 0.1328718066215515, "learning_rate": 7.893395009492203e-05, "loss": 0.0131, "step": 33840 }, { "grad_norm": 0.13099847733974457, "learning_rate": 7.892046355067248e-05, "loss": 0.014, "step": 33850 }, { "grad_norm": 0.15971538424491882, "learning_rate": 7.890697384372451e-05, "loss": 0.0151, "step": 33860 }, { "grad_norm": 0.20670035481452942, "learning_rate": 7.889348097555336e-05, "loss": 0.0127, "step": 33870 }, { "grad_norm": 0.2842061519622803, "learning_rate": 7.887998494763455e-05, "loss": 0.0145, "step": 33880 }, { "grad_norm": 0.14072328805923462, "learning_rate": 7.886648576144404e-05, "loss": 0.0132, "step": 33890 }, { "grad_norm": 0.11123070865869522, "learning_rate": 7.885298341845802e-05, "loss": 0.0133, "step": 33900 }, { "grad_norm": 0.20202797651290894, "learning_rate": 7.883947792015311e-05, "loss": 0.0167, "step": 33910 }, { "grad_norm": 0.25498083233833313, "learning_rate": 7.882596926800628e-05, "loss": 0.0347, "step": 33920 }, { "grad_norm": 0.12949642539024353, "learning_rate": 7.881245746349477e-05, "loss": 0.0133, "step": 33930 }, { "grad_norm": 0.1281362920999527, "learning_rate": 7.879894250809623e-05, "loss": 0.0112, "step": 33940 }, { "grad_norm": 0.1723213493824005, "learning_rate": 7.878542440328865e-05, "loss": 0.0286, "step": 33950 }, { "grad_norm": 0.12375468760728836, "learning_rate": 7.877190315055031e-05, "loss": 0.0136, "step": 33960 }, { "grad_norm": 0.17017774283885956, "learning_rate": 7.875837875135991e-05, "loss": 0.0125, "step": 33970 }, { "grad_norm": 0.1139908879995346, "learning_rate": 7.874485120719646e-05, "loss": 0.013, "step": 33980 }, { "grad_norm": 0.12668286263942719, "learning_rate": 7.873132051953928e-05, "loss": 0.0106, "step": 33990 }, { "grad_norm": 0.18014158308506012, "learning_rate": 7.87177866898681e-05, "loss": 0.0177, "step": 34000 }, { "grad_norm": 0.11476244032382965, "learning_rate": 7.870424971966294e-05, "loss": 0.0147, "step": 34010 }, { "grad_norm": 0.1576281040906906, "learning_rate": 7.869070961040419e-05, "loss": 0.0154, "step": 34020 }, { "grad_norm": 0.20612552762031555, "learning_rate": 7.867716636357257e-05, "loss": 0.0163, "step": 34030 }, { "grad_norm": 0.1534276008605957, "learning_rate": 7.866361998064915e-05, "loss": 0.015, "step": 34040 }, { "grad_norm": 0.1523941159248352, "learning_rate": 7.865007046311534e-05, "loss": 0.0156, "step": 34050 }, { "grad_norm": 0.17643436789512634, "learning_rate": 7.86365178124529e-05, "loss": 0.0131, "step": 34060 }, { "grad_norm": 0.12647707760334015, "learning_rate": 7.862296203014394e-05, "loss": 0.015, "step": 34070 }, { "grad_norm": 0.16333842277526855, "learning_rate": 7.860940311767088e-05, "loss": 0.0137, "step": 34080 }, { "grad_norm": 0.10977315157651901, "learning_rate": 7.85958410765165e-05, "loss": 0.012, "step": 34090 }, { "grad_norm": 0.13275450468063354, "learning_rate": 7.858227590816394e-05, "loss": 0.0149, "step": 34100 }, { "grad_norm": 0.1553916186094284, "learning_rate": 7.856870761409664e-05, "loss": 0.0121, "step": 34110 }, { "grad_norm": 0.19745154678821564, "learning_rate": 7.855513619579846e-05, "loss": 0.0129, "step": 34120 }, { "grad_norm": 0.1462908536195755, "learning_rate": 7.85415616547535e-05, "loss": 0.0134, "step": 34130 }, { "grad_norm": 0.12033642828464508, "learning_rate": 7.852798399244627e-05, "loss": 0.0098, "step": 34140 }, { "grad_norm": 0.16079235076904297, "learning_rate": 7.851440321036161e-05, "loss": 0.0134, "step": 34150 }, { "grad_norm": 0.15797454118728638, "learning_rate": 7.850081930998468e-05, "loss": 0.0129, "step": 34160 }, { "grad_norm": 0.23334351181983948, "learning_rate": 7.8487232292801e-05, "loss": 0.0183, "step": 34170 }, { "grad_norm": 0.15337613224983215, "learning_rate": 7.847364216029642e-05, "loss": 0.0153, "step": 34180 }, { "grad_norm": 0.140402689576149, "learning_rate": 7.846004891395716e-05, "loss": 0.0136, "step": 34190 }, { "grad_norm": 0.167881578207016, "learning_rate": 7.844645255526972e-05, "loss": 0.0146, "step": 34200 }, { "grad_norm": 0.2204260379076004, "learning_rate": 7.843285308572101e-05, "loss": 0.0153, "step": 34210 }, { "grad_norm": 0.13324464857578278, "learning_rate": 7.841925050679823e-05, "loss": 0.0142, "step": 34220 }, { "grad_norm": 0.10134400427341461, "learning_rate": 7.840564481998895e-05, "loss": 0.0136, "step": 34230 }, { "grad_norm": 0.11975867301225662, "learning_rate": 7.839203602678105e-05, "loss": 0.0116, "step": 34240 }, { "grad_norm": 0.1614375114440918, "learning_rate": 7.837842412866279e-05, "loss": 0.0132, "step": 34250 }, { "grad_norm": 0.13569194078445435, "learning_rate": 7.836480912712272e-05, "loss": 0.0159, "step": 34260 }, { "grad_norm": 0.2744900584220886, "learning_rate": 7.835119102364976e-05, "loss": 0.0184, "step": 34270 }, { "grad_norm": 0.1313621699810028, "learning_rate": 7.83375698197332e-05, "loss": 0.0166, "step": 34280 }, { "grad_norm": 0.18418192863464355, "learning_rate": 7.83239455168626e-05, "loss": 0.0145, "step": 34290 }, { "grad_norm": 0.08503159880638123, "learning_rate": 7.83103181165279e-05, "loss": 0.0113, "step": 34300 }, { "grad_norm": 0.1695684939622879, "learning_rate": 7.829668762021937e-05, "loss": 0.0169, "step": 34310 }, { "grad_norm": 0.1475752890110016, "learning_rate": 7.828305402942764e-05, "loss": 0.0118, "step": 34320 }, { "grad_norm": 0.1738114356994629, "learning_rate": 7.826941734564363e-05, "loss": 0.0118, "step": 34330 }, { "grad_norm": 0.13041993975639343, "learning_rate": 7.825577757035865e-05, "loss": 0.0105, "step": 34340 }, { "grad_norm": 0.19693437218666077, "learning_rate": 7.824213470506431e-05, "loss": 0.016, "step": 34350 }, { "grad_norm": 0.2093646228313446, "learning_rate": 7.822848875125257e-05, "loss": 0.0142, "step": 34360 }, { "grad_norm": 0.19988389313220978, "learning_rate": 7.821483971041576e-05, "loss": 0.0189, "step": 34370 }, { "grad_norm": 0.1374722272157669, "learning_rate": 7.820118758404649e-05, "loss": 0.0173, "step": 34380 }, { "grad_norm": 0.17900823056697845, "learning_rate": 7.818753237363776e-05, "loss": 0.0156, "step": 34390 }, { "grad_norm": 0.16399414837360382, "learning_rate": 7.817387408068286e-05, "loss": 0.0144, "step": 34400 }, { "grad_norm": 0.12427852302789688, "learning_rate": 7.816021270667544e-05, "loss": 0.0134, "step": 34410 }, { "grad_norm": 0.12200812250375748, "learning_rate": 7.81465482531095e-05, "loss": 0.014, "step": 34420 }, { "grad_norm": 0.12028475850820541, "learning_rate": 7.813288072147938e-05, "loss": 0.015, "step": 34430 }, { "grad_norm": 0.13649532198905945, "learning_rate": 7.811921011327972e-05, "loss": 0.0187, "step": 34440 }, { "grad_norm": 0.11105813831090927, "learning_rate": 7.810553643000549e-05, "loss": 0.0177, "step": 34450 }, { "grad_norm": 0.0883539617061615, "learning_rate": 7.809185967315206e-05, "loss": 0.0143, "step": 34460 }, { "grad_norm": 0.19217677414417267, "learning_rate": 7.80781798442151e-05, "loss": 0.0201, "step": 34470 }, { "grad_norm": 0.17207330465316772, "learning_rate": 7.806449694469061e-05, "loss": 0.0139, "step": 34480 }, { "grad_norm": 0.11190572381019592, "learning_rate": 7.805081097607492e-05, "loss": 0.0147, "step": 34490 }, { "grad_norm": 0.11510040611028671, "learning_rate": 7.803712193986474e-05, "loss": 0.0143, "step": 34500 }, { "grad_norm": 0.1332448273897171, "learning_rate": 7.802342983755702e-05, "loss": 0.0126, "step": 34510 }, { "grad_norm": 0.22996321320533752, "learning_rate": 7.800973467064918e-05, "loss": 0.0159, "step": 34520 }, { "grad_norm": 0.18707066774368286, "learning_rate": 7.799603644063884e-05, "loss": 0.0178, "step": 34530 }, { "grad_norm": 0.16460224986076355, "learning_rate": 7.798233514902405e-05, "loss": 0.0132, "step": 34540 }, { "grad_norm": 0.16314910352230072, "learning_rate": 7.796863079730318e-05, "loss": 0.0145, "step": 34550 }, { "grad_norm": 0.1398579627275467, "learning_rate": 7.795492338697488e-05, "loss": 0.0111, "step": 34560 }, { "grad_norm": 0.14305591583251953, "learning_rate": 7.794121291953819e-05, "loss": 0.0147, "step": 34570 }, { "grad_norm": 0.14361634850502014, "learning_rate": 7.792749939649246e-05, "loss": 0.0153, "step": 34580 }, { "grad_norm": 0.10914843529462814, "learning_rate": 7.79137828193374e-05, "loss": 0.013, "step": 34590 }, { "grad_norm": 0.21694190800189972, "learning_rate": 7.790006318957301e-05, "loss": 0.0223, "step": 34600 }, { "grad_norm": 0.2542678415775299, "learning_rate": 7.788634050869965e-05, "loss": 0.0163, "step": 34610 }, { "grad_norm": 0.13792213797569275, "learning_rate": 7.787261477821803e-05, "loss": 0.0117, "step": 34620 }, { "grad_norm": 0.15544354915618896, "learning_rate": 7.785888599962916e-05, "loss": 0.0135, "step": 34630 }, { "grad_norm": 0.19968415796756744, "learning_rate": 7.784515417443439e-05, "loss": 0.0147, "step": 34640 }, { "grad_norm": 0.16233471035957336, "learning_rate": 7.783141930413545e-05, "loss": 0.0169, "step": 34650 }, { "grad_norm": 0.12755616009235382, "learning_rate": 7.78176813902343e-05, "loss": 0.0103, "step": 34660 }, { "grad_norm": 0.16248555481433868, "learning_rate": 7.780394043423336e-05, "loss": 0.0135, "step": 34670 }, { "grad_norm": 0.19330964982509613, "learning_rate": 7.77901964376353e-05, "loss": 0.0177, "step": 34680 }, { "grad_norm": 0.17362268269062042, "learning_rate": 7.777644940194316e-05, "loss": 0.0148, "step": 34690 }, { "grad_norm": 0.22284889221191406, "learning_rate": 7.776269932866023e-05, "loss": 0.0165, "step": 34700 }, { "grad_norm": 0.16200928390026093, "learning_rate": 7.774894621929026e-05, "loss": 0.015, "step": 34710 }, { "grad_norm": 0.09973954409360886, "learning_rate": 7.773519007533725e-05, "loss": 0.0247, "step": 34720 }, { "grad_norm": 0.1236182227730751, "learning_rate": 7.772143089830556e-05, "loss": 0.0133, "step": 34730 }, { "grad_norm": 0.13968342542648315, "learning_rate": 7.770766868969985e-05, "loss": 0.0145, "step": 34740 }, { "grad_norm": 0.13190335035324097, "learning_rate": 7.769390345102518e-05, "loss": 0.0135, "step": 34750 }, { "grad_norm": 0.16220742464065552, "learning_rate": 7.768013518378683e-05, "loss": 0.0161, "step": 34760 }, { "grad_norm": 0.11508989334106445, "learning_rate": 7.766636388949053e-05, "loss": 0.0112, "step": 34770 }, { "grad_norm": 0.11751638352870941, "learning_rate": 7.765258956964229e-05, "loss": 0.0215, "step": 34780 }, { "grad_norm": 0.1652015894651413, "learning_rate": 7.76388122257484e-05, "loss": 0.0143, "step": 34790 }, { "grad_norm": 0.103579081594944, "learning_rate": 7.762503185931558e-05, "loss": 0.0115, "step": 34800 }, { "grad_norm": 0.13203537464141846, "learning_rate": 7.76112484718508e-05, "loss": 0.0112, "step": 34810 }, { "grad_norm": 0.1369178593158722, "learning_rate": 7.75974620648614e-05, "loss": 0.0145, "step": 34820 }, { "grad_norm": 0.11808847635984421, "learning_rate": 7.758367263985503e-05, "loss": 0.0111, "step": 34830 }, { "grad_norm": 0.10630127787590027, "learning_rate": 7.75698801983397e-05, "loss": 0.012, "step": 34840 }, { "grad_norm": 0.12729616463184357, "learning_rate": 7.755608474182372e-05, "loss": 0.0132, "step": 34850 }, { "grad_norm": 0.2070319652557373, "learning_rate": 7.754228627181574e-05, "loss": 0.0164, "step": 34860 }, { "grad_norm": 0.13365310430526733, "learning_rate": 7.752848478982476e-05, "loss": 0.0153, "step": 34870 }, { "grad_norm": 0.1683093160390854, "learning_rate": 7.751468029736006e-05, "loss": 0.0133, "step": 34880 }, { "grad_norm": 0.14395952224731445, "learning_rate": 7.750087279593129e-05, "loss": 0.0181, "step": 34890 }, { "grad_norm": 0.22967587411403656, "learning_rate": 7.748706228704843e-05, "loss": 0.0126, "step": 34900 }, { "grad_norm": 0.11919987946748734, "learning_rate": 7.747324877222176e-05, "loss": 0.0136, "step": 34910 }, { "grad_norm": 0.13623511791229248, "learning_rate": 7.745943225296188e-05, "loss": 0.0153, "step": 34920 }, { "grad_norm": 0.12802539765834808, "learning_rate": 7.744561273077981e-05, "loss": 0.0113, "step": 34930 }, { "grad_norm": 0.1539163589477539, "learning_rate": 7.743179020718678e-05, "loss": 0.0128, "step": 34940 }, { "grad_norm": 0.1646731197834015, "learning_rate": 7.741796468369443e-05, "loss": 0.0165, "step": 34950 }, { "grad_norm": 0.16807326674461365, "learning_rate": 7.740413616181466e-05, "loss": 0.0172, "step": 34960 }, { "grad_norm": 0.16750214993953705, "learning_rate": 7.739030464305978e-05, "loss": 0.0151, "step": 34970 }, { "grad_norm": 0.1683291345834732, "learning_rate": 7.737647012894235e-05, "loss": 0.018, "step": 34980 }, { "grad_norm": 0.12250223755836487, "learning_rate": 7.736263262097532e-05, "loss": 0.0138, "step": 34990 }, { "grad_norm": 0.16780422627925873, "learning_rate": 7.734879212067192e-05, "loss": 0.0138, "step": 35000 }, { "grad_norm": 0.2309248447418213, "learning_rate": 7.733494862954573e-05, "loss": 0.0149, "step": 35010 }, { "grad_norm": 0.13133014738559723, "learning_rate": 7.732110214911066e-05, "loss": 0.0121, "step": 35020 }, { "grad_norm": 0.14228032529354095, "learning_rate": 7.730725268088094e-05, "loss": 0.0184, "step": 35030 }, { "grad_norm": 0.1496986597776413, "learning_rate": 7.729340022637112e-05, "loss": 0.0153, "step": 35040 }, { "grad_norm": 0.1482969969511032, "learning_rate": 7.727954478709607e-05, "loss": 0.0136, "step": 35050 }, { "grad_norm": 0.12231019139289856, "learning_rate": 7.726568636457103e-05, "loss": 0.0127, "step": 35060 }, { "grad_norm": 0.13492615520954132, "learning_rate": 7.725182496031153e-05, "loss": 0.0135, "step": 35070 }, { "grad_norm": 0.16821621358394623, "learning_rate": 7.723796057583341e-05, "loss": 0.0154, "step": 35080 }, { "grad_norm": 0.15727338194847107, "learning_rate": 7.722409321265287e-05, "loss": 0.0143, "step": 35090 }, { "grad_norm": 0.14312933385372162, "learning_rate": 7.721022287228645e-05, "loss": 0.0217, "step": 35100 }, { "grad_norm": 0.16833284497261047, "learning_rate": 7.719634955625094e-05, "loss": 0.0151, "step": 35110 }, { "grad_norm": 0.18234004080295563, "learning_rate": 7.718247326606355e-05, "loss": 0.0144, "step": 35120 }, { "grad_norm": 0.1281295269727707, "learning_rate": 7.716859400324175e-05, "loss": 0.015, "step": 35130 }, { "grad_norm": 0.22344274818897247, "learning_rate": 7.715471176930335e-05, "loss": 0.0257, "step": 35140 }, { "grad_norm": 0.1310625523328781, "learning_rate": 7.714082656576651e-05, "loss": 0.0112, "step": 35150 }, { "grad_norm": 0.19176316261291504, "learning_rate": 7.712693839414968e-05, "loss": 0.0134, "step": 35160 }, { "grad_norm": 0.14100788533687592, "learning_rate": 7.711304725597164e-05, "loss": 0.0137, "step": 35170 }, { "grad_norm": 0.12549187242984772, "learning_rate": 7.709915315275151e-05, "loss": 0.014, "step": 35180 }, { "grad_norm": 0.13156670331954956, "learning_rate": 7.708525608600876e-05, "loss": 0.0148, "step": 35190 }, { "grad_norm": 0.10493259131908417, "learning_rate": 7.707135605726311e-05, "loss": 0.0147, "step": 35200 }, { "grad_norm": 0.15171155333518982, "learning_rate": 7.705745306803466e-05, "loss": 0.0312, "step": 35210 }, { "grad_norm": 0.1358078420162201, "learning_rate": 7.704354711984383e-05, "loss": 0.0132, "step": 35220 }, { "grad_norm": 0.14436843991279602, "learning_rate": 7.702963821421133e-05, "loss": 0.0143, "step": 35230 }, { "grad_norm": 0.1625218689441681, "learning_rate": 7.701572635265826e-05, "loss": 0.0135, "step": 35240 }, { "grad_norm": 0.1162044033408165, "learning_rate": 7.700181153670596e-05, "loss": 0.0125, "step": 35250 }, { "grad_norm": 0.09090308845043182, "learning_rate": 7.698789376787616e-05, "loss": 0.0116, "step": 35260 }, { "grad_norm": 0.1482856124639511, "learning_rate": 7.697397304769085e-05, "loss": 0.0156, "step": 35270 }, { "grad_norm": 0.19472002983093262, "learning_rate": 7.696004937767241e-05, "loss": 0.0243, "step": 35280 }, { "grad_norm": 0.24864616990089417, "learning_rate": 7.694612275934352e-05, "loss": 0.0146, "step": 35290 }, { "grad_norm": 0.23091669380664825, "learning_rate": 7.693219319422714e-05, "loss": 0.0161, "step": 35300 }, { "grad_norm": 0.16637736558914185, "learning_rate": 7.69182606838466e-05, "loss": 0.0152, "step": 35310 }, { "grad_norm": 0.17939601838588715, "learning_rate": 7.690432522972558e-05, "loss": 0.0149, "step": 35320 }, { "grad_norm": 0.11970212310552597, "learning_rate": 7.689038683338796e-05, "loss": 0.0133, "step": 35330 }, { "grad_norm": 0.11249443888664246, "learning_rate": 7.687644549635808e-05, "loss": 0.0098, "step": 35340 }, { "grad_norm": 0.17029201984405518, "learning_rate": 7.686250122016053e-05, "loss": 0.0123, "step": 35350 }, { "grad_norm": 0.16180644929409027, "learning_rate": 7.684855400632022e-05, "loss": 0.0143, "step": 35360 }, { "grad_norm": 0.1553993672132492, "learning_rate": 7.683460385636243e-05, "loss": 0.012, "step": 35370 }, { "grad_norm": 0.16732066869735718, "learning_rate": 7.68206507718127e-05, "loss": 0.0145, "step": 35380 }, { "grad_norm": 0.1660892367362976, "learning_rate": 7.680669475419693e-05, "loss": 0.0162, "step": 35390 }, { "grad_norm": 0.15987694263458252, "learning_rate": 7.679273580504132e-05, "loss": 0.0191, "step": 35400 }, { "grad_norm": 0.21278902888298035, "learning_rate": 7.67787739258724e-05, "loss": 0.0135, "step": 35410 }, { "grad_norm": 0.13174687325954437, "learning_rate": 7.676480911821705e-05, "loss": 0.0176, "step": 35420 }, { "grad_norm": 0.14789927005767822, "learning_rate": 7.675084138360238e-05, "loss": 0.0165, "step": 35430 }, { "grad_norm": 0.1703503429889679, "learning_rate": 7.673687072355592e-05, "loss": 0.0139, "step": 35440 }, { "grad_norm": 0.14671380817890167, "learning_rate": 7.67228971396055e-05, "loss": 0.0118, "step": 35450 }, { "grad_norm": 0.14502595365047455, "learning_rate": 7.670892063327922e-05, "loss": 0.0153, "step": 35460 }, { "grad_norm": 0.1086328849196434, "learning_rate": 7.669494120610552e-05, "loss": 0.0111, "step": 35470 }, { "grad_norm": 0.10176727175712585, "learning_rate": 7.66809588596132e-05, "loss": 0.0111, "step": 35480 }, { "grad_norm": 0.11285190284252167, "learning_rate": 7.666697359533132e-05, "loss": 0.0125, "step": 35490 }, { "grad_norm": 0.19092074036598206, "learning_rate": 7.665298541478932e-05, "loss": 0.0134, "step": 35500 }, { "grad_norm": 0.13023142516613007, "learning_rate": 7.66389943195169e-05, "loss": 0.0135, "step": 35510 }, { "grad_norm": 0.1315358579158783, "learning_rate": 7.66250003110441e-05, "loss": 0.0163, "step": 35520 }, { "grad_norm": 0.16316384077072144, "learning_rate": 7.66110033909013e-05, "loss": 0.0121, "step": 35530 }, { "grad_norm": 0.13104362785816193, "learning_rate": 7.659700356061918e-05, "loss": 0.0147, "step": 35540 }, { "grad_norm": 0.11855103075504303, "learning_rate": 7.658300082172875e-05, "loss": 0.012, "step": 35550 }, { "grad_norm": 0.174359530210495, "learning_rate": 7.65689951757613e-05, "loss": 0.0145, "step": 35560 }, { "grad_norm": 0.1916588991880417, "learning_rate": 7.655498662424849e-05, "loss": 0.0119, "step": 35570 }, { "grad_norm": 0.20587831735610962, "learning_rate": 7.654097516872227e-05, "loss": 0.0128, "step": 35580 }, { "grad_norm": 0.1941736340522766, "learning_rate": 7.65269608107149e-05, "loss": 0.0169, "step": 35590 }, { "grad_norm": 0.13095957040786743, "learning_rate": 7.6512943551759e-05, "loss": 0.0245, "step": 35600 }, { "grad_norm": 0.14120842516422272, "learning_rate": 7.649892339338743e-05, "loss": 0.0137, "step": 35610 }, { "grad_norm": 0.10140277445316315, "learning_rate": 7.648490033713344e-05, "loss": 0.0104, "step": 35620 }, { "grad_norm": 0.11269274353981018, "learning_rate": 7.647087438453058e-05, "loss": 0.013, "step": 35630 }, { "grad_norm": 0.14837664365768433, "learning_rate": 7.64568455371127e-05, "loss": 0.0146, "step": 35640 }, { "grad_norm": 0.16306351125240326, "learning_rate": 7.644281379641396e-05, "loss": 0.0134, "step": 35650 }, { "grad_norm": 0.1238226667046547, "learning_rate": 7.642877916396887e-05, "loss": 0.0131, "step": 35660 }, { "grad_norm": 0.14443860948085785, "learning_rate": 7.641474164131221e-05, "loss": 0.0149, "step": 35670 }, { "grad_norm": 0.12877322733402252, "learning_rate": 7.640070122997913e-05, "loss": 0.0115, "step": 35680 }, { "grad_norm": 0.10532840341329575, "learning_rate": 7.638665793150505e-05, "loss": 0.0117, "step": 35690 }, { "grad_norm": 0.1576584279537201, "learning_rate": 7.637261174742574e-05, "loss": 0.0144, "step": 35700 }, { "grad_norm": 0.12166628986597061, "learning_rate": 7.635856267927725e-05, "loss": 0.0128, "step": 35710 }, { "grad_norm": 0.12010981887578964, "learning_rate": 7.634451072859597e-05, "loss": 0.0116, "step": 35720 }, { "grad_norm": 0.12738589942455292, "learning_rate": 7.633045589691863e-05, "loss": 0.0143, "step": 35730 }, { "grad_norm": 0.10368647426366806, "learning_rate": 7.63163981857822e-05, "loss": 0.012, "step": 35740 }, { "grad_norm": 0.15515364706516266, "learning_rate": 7.630233759672403e-05, "loss": 0.0151, "step": 35750 }, { "grad_norm": 0.10861969739198685, "learning_rate": 7.628827413128177e-05, "loss": 0.0127, "step": 35760 }, { "grad_norm": 0.12577615678310394, "learning_rate": 7.627420779099338e-05, "loss": 0.0139, "step": 35770 }, { "grad_norm": 0.13894866406917572, "learning_rate": 7.626013857739711e-05, "loss": 0.0171, "step": 35780 }, { "grad_norm": 0.1257079690694809, "learning_rate": 7.624606649203158e-05, "loss": 0.012, "step": 35790 }, { "grad_norm": 0.10169102996587753, "learning_rate": 7.623199153643569e-05, "loss": 0.0165, "step": 35800 }, { "grad_norm": 0.1233230009675026, "learning_rate": 7.621791371214863e-05, "loss": 0.0198, "step": 35810 }, { "grad_norm": 0.08999013900756836, "learning_rate": 7.620383302070995e-05, "loss": 0.0138, "step": 35820 }, { "grad_norm": 0.13354504108428955, "learning_rate": 7.61897494636595e-05, "loss": 0.0145, "step": 35830 }, { "grad_norm": 0.15836626291275024, "learning_rate": 7.617566304253739e-05, "loss": 0.0188, "step": 35840 }, { "grad_norm": 0.14323420822620392, "learning_rate": 7.616157375888416e-05, "loss": 0.014, "step": 35850 }, { "grad_norm": 0.13951514661312103, "learning_rate": 7.614748161424053e-05, "loss": 0.0122, "step": 35860 }, { "grad_norm": 0.17170123755931854, "learning_rate": 7.613338661014763e-05, "loss": 0.0128, "step": 35870 }, { "grad_norm": 0.2179766744375229, "learning_rate": 7.611928874814686e-05, "loss": 0.0151, "step": 35880 }, { "grad_norm": 0.1562669724225998, "learning_rate": 7.610518802977996e-05, "loss": 0.013, "step": 35890 }, { "grad_norm": 0.15130272507667542, "learning_rate": 7.609108445658893e-05, "loss": 0.0128, "step": 35900 }, { "grad_norm": 0.1370217651128769, "learning_rate": 7.607697803011612e-05, "loss": 0.0262, "step": 35910 }, { "grad_norm": 0.10977678000926971, "learning_rate": 7.606286875190421e-05, "loss": 0.0176, "step": 35920 }, { "grad_norm": 0.15862815082073212, "learning_rate": 7.604875662349617e-05, "loss": 0.0128, "step": 35930 }, { "grad_norm": 0.21740147471427917, "learning_rate": 7.603464164643525e-05, "loss": 0.0198, "step": 35940 }, { "grad_norm": 0.1142352819442749, "learning_rate": 7.602052382226507e-05, "loss": 0.0147, "step": 35950 }, { "grad_norm": 0.11507031321525574, "learning_rate": 7.600640315252954e-05, "loss": 0.0114, "step": 35960 }, { "grad_norm": 0.1836388111114502, "learning_rate": 7.599227963877284e-05, "loss": 0.0167, "step": 35970 }, { "grad_norm": 0.17982134222984314, "learning_rate": 7.597815328253954e-05, "loss": 0.0159, "step": 35980 }, { "grad_norm": 0.15364626049995422, "learning_rate": 7.596402408537443e-05, "loss": 0.0146, "step": 35990 }, { "grad_norm": 0.11279837042093277, "learning_rate": 7.59498920488227e-05, "loss": 0.0205, "step": 36000 }, { "grad_norm": 0.1343621462583542, "learning_rate": 7.593575717442979e-05, "loss": 0.0124, "step": 36010 }, { "grad_norm": 0.14606845378875732, "learning_rate": 7.592161946374147e-05, "loss": 0.0133, "step": 36020 }, { "grad_norm": 0.1311798393726349, "learning_rate": 7.590747891830381e-05, "loss": 0.0127, "step": 36030 }, { "grad_norm": 0.15611714124679565, "learning_rate": 7.58933355396632e-05, "loss": 0.015, "step": 36040 }, { "grad_norm": 0.11110946536064148, "learning_rate": 7.587918932936636e-05, "loss": 0.0106, "step": 36050 }, { "grad_norm": 0.14753735065460205, "learning_rate": 7.586504028896028e-05, "loss": 0.0129, "step": 36060 }, { "grad_norm": 0.11761118471622467, "learning_rate": 7.585088841999228e-05, "loss": 0.0116, "step": 36070 }, { "grad_norm": 0.10085001587867737, "learning_rate": 7.583673372400999e-05, "loss": 0.0142, "step": 36080 }, { "grad_norm": 0.14231616258621216, "learning_rate": 7.582257620256134e-05, "loss": 0.0131, "step": 36090 }, { "grad_norm": 0.18180608749389648, "learning_rate": 7.580841585719458e-05, "loss": 0.0177, "step": 36100 }, { "grad_norm": 0.13177862763404846, "learning_rate": 7.579425268945825e-05, "loss": 0.0102, "step": 36110 }, { "grad_norm": 0.1413739174604416, "learning_rate": 7.578008670090127e-05, "loss": 0.0121, "step": 36120 }, { "grad_norm": 0.11972688138484955, "learning_rate": 7.576591789307272e-05, "loss": 0.0137, "step": 36130 }, { "grad_norm": 0.10926686972379684, "learning_rate": 7.575174626752216e-05, "loss": 0.0121, "step": 36140 }, { "grad_norm": 0.0949702337384224, "learning_rate": 7.573757182579934e-05, "loss": 0.0113, "step": 36150 }, { "grad_norm": 0.11534049361944199, "learning_rate": 7.572339456945435e-05, "loss": 0.0138, "step": 36160 }, { "grad_norm": 0.1695273518562317, "learning_rate": 7.570921450003762e-05, "loss": 0.0137, "step": 36170 }, { "grad_norm": 0.11432068794965744, "learning_rate": 7.569503161909983e-05, "loss": 0.011, "step": 36180 }, { "grad_norm": 0.18795038759708405, "learning_rate": 7.568084592819202e-05, "loss": 0.0152, "step": 36190 }, { "grad_norm": 0.15040160715579987, "learning_rate": 7.566665742886551e-05, "loss": 0.0127, "step": 36200 }, { "grad_norm": 0.2685355246067047, "learning_rate": 7.565246612267194e-05, "loss": 0.0169, "step": 36210 }, { "grad_norm": 0.24080972373485565, "learning_rate": 7.563827201116325e-05, "loss": 0.0158, "step": 36220 }, { "grad_norm": 0.19044120609760284, "learning_rate": 7.562407509589164e-05, "loss": 0.0148, "step": 36230 }, { "grad_norm": 0.09338664263486862, "learning_rate": 7.560987537840973e-05, "loss": 0.0101, "step": 36240 }, { "grad_norm": 0.11271078884601593, "learning_rate": 7.559567286027036e-05, "loss": 0.0118, "step": 36250 }, { "grad_norm": 0.09000454097986221, "learning_rate": 7.558146754302668e-05, "loss": 0.0141, "step": 36260 }, { "grad_norm": 0.15312062203884125, "learning_rate": 7.556725942823217e-05, "loss": 0.0135, "step": 36270 }, { "grad_norm": 0.1910359263420105, "learning_rate": 7.555304851744061e-05, "loss": 0.0147, "step": 36280 }, { "grad_norm": 0.10766252130270004, "learning_rate": 7.553883481220608e-05, "loss": 0.0218, "step": 36290 }, { "grad_norm": 0.09023067355155945, "learning_rate": 7.552461831408298e-05, "loss": 0.01, "step": 36300 }, { "grad_norm": 0.18782036006450653, "learning_rate": 7.551039902462599e-05, "loss": 0.0121, "step": 36310 }, { "grad_norm": 0.13223420083522797, "learning_rate": 7.549617694539011e-05, "loss": 0.0128, "step": 36320 }, { "grad_norm": 0.12997126579284668, "learning_rate": 7.548195207793066e-05, "loss": 0.014, "step": 36330 }, { "grad_norm": 0.13137172162532806, "learning_rate": 7.546772442380323e-05, "loss": 0.0109, "step": 36340 }, { "grad_norm": 0.16865719854831696, "learning_rate": 7.545349398456376e-05, "loss": 0.0161, "step": 36350 }, { "grad_norm": 0.13340622186660767, "learning_rate": 7.543926076176845e-05, "loss": 0.0139, "step": 36360 }, { "grad_norm": 0.19101248681545258, "learning_rate": 7.542502475697385e-05, "loss": 0.0116, "step": 36370 }, { "grad_norm": 0.14880815148353577, "learning_rate": 7.541078597173675e-05, "loss": 0.0149, "step": 36380 }, { "grad_norm": 0.12445618957281113, "learning_rate": 7.539654440761431e-05, "loss": 0.0159, "step": 36390 }, { "grad_norm": 0.12181926518678665, "learning_rate": 7.538230006616395e-05, "loss": 0.0144, "step": 36400 }, { "grad_norm": 0.11938884109258652, "learning_rate": 7.536805294894342e-05, "loss": 0.0146, "step": 36410 }, { "grad_norm": 0.36882174015045166, "learning_rate": 7.535380305751076e-05, "loss": 0.0128, "step": 36420 }, { "grad_norm": 0.17905373871326447, "learning_rate": 7.533955039342431e-05, "loss": 0.0137, "step": 36430 }, { "grad_norm": 0.11840043216943741, "learning_rate": 7.532529495824274e-05, "loss": 0.0176, "step": 36440 }, { "grad_norm": 0.13998854160308838, "learning_rate": 7.531103675352497e-05, "loss": 0.0153, "step": 36450 }, { "grad_norm": 0.13994812965393066, "learning_rate": 7.529677578083029e-05, "loss": 0.0154, "step": 36460 }, { "grad_norm": 0.10167160630226135, "learning_rate": 7.528251204171823e-05, "loss": 0.0101, "step": 36470 }, { "grad_norm": 0.14264214038848877, "learning_rate": 7.526824553774867e-05, "loss": 0.0153, "step": 36480 }, { "grad_norm": 0.14592434465885162, "learning_rate": 7.525397627048178e-05, "loss": 0.0119, "step": 36490 }, { "grad_norm": 0.09917698800563812, "learning_rate": 7.523970424147802e-05, "loss": 0.0101, "step": 36500 }, { "grad_norm": 0.1257966309785843, "learning_rate": 7.522542945229813e-05, "loss": 0.0119, "step": 36510 }, { "grad_norm": 0.09757556766271591, "learning_rate": 7.521115190450321e-05, "loss": 0.0107, "step": 36520 }, { "grad_norm": 0.1623593121767044, "learning_rate": 7.519687159965462e-05, "loss": 0.013, "step": 36530 }, { "grad_norm": 0.17116722464561462, "learning_rate": 7.518258853931403e-05, "loss": 0.0131, "step": 36540 }, { "grad_norm": 0.1788906455039978, "learning_rate": 7.516830272504342e-05, "loss": 0.0115, "step": 36550 }, { "grad_norm": 0.17426742613315582, "learning_rate": 7.515401415840509e-05, "loss": 0.0174, "step": 36560 }, { "grad_norm": 0.17031709849834442, "learning_rate": 7.513972284096155e-05, "loss": 0.0111, "step": 36570 }, { "grad_norm": 0.1757601797580719, "learning_rate": 7.512542877427574e-05, "loss": 0.02, "step": 36580 }, { "grad_norm": 0.18848271667957306, "learning_rate": 7.51111319599108e-05, "loss": 0.0173, "step": 36590 }, { "grad_norm": 0.13819685578346252, "learning_rate": 7.509683239943024e-05, "loss": 0.0154, "step": 36600 }, { "grad_norm": 0.14433567225933075, "learning_rate": 7.50825300943978e-05, "loss": 0.0149, "step": 36610 }, { "grad_norm": 0.145375058054924, "learning_rate": 7.50682250463776e-05, "loss": 0.0135, "step": 36620 }, { "grad_norm": 0.14510072767734528, "learning_rate": 7.5053917256934e-05, "loss": 0.0156, "step": 36630 }, { "grad_norm": 0.13615265488624573, "learning_rate": 7.503960672763166e-05, "loss": 0.0134, "step": 36640 }, { "grad_norm": 0.09617490321397781, "learning_rate": 7.502529346003559e-05, "loss": 0.0145, "step": 36650 }, { "grad_norm": 0.13814029097557068, "learning_rate": 7.501097745571106e-05, "loss": 0.0129, "step": 36660 }, { "grad_norm": 0.19645337760448456, "learning_rate": 7.499665871622361e-05, "loss": 0.0215, "step": 36670 }, { "grad_norm": 0.1762530356645584, "learning_rate": 7.498233724313917e-05, "loss": 0.0189, "step": 36680 }, { "grad_norm": 0.22160322964191437, "learning_rate": 7.496801303802387e-05, "loss": 0.0195, "step": 36690 }, { "grad_norm": 0.13878600299358368, "learning_rate": 7.495368610244423e-05, "loss": 0.0119, "step": 36700 }, { "grad_norm": 0.175279900431633, "learning_rate": 7.493935643796697e-05, "loss": 0.0156, "step": 36710 }, { "grad_norm": 0.25113117694854736, "learning_rate": 7.49250240461592e-05, "loss": 0.0147, "step": 36720 }, { "grad_norm": 0.1939544677734375, "learning_rate": 7.491068892858826e-05, "loss": 0.0137, "step": 36730 }, { "grad_norm": 0.12061594426631927, "learning_rate": 7.489635108682184e-05, "loss": 0.0109, "step": 36740 }, { "grad_norm": 0.1252427101135254, "learning_rate": 7.48820105224279e-05, "loss": 0.0117, "step": 36750 }, { "grad_norm": 0.1025615856051445, "learning_rate": 7.48676672369747e-05, "loss": 0.0138, "step": 36760 }, { "grad_norm": 0.2517145574092865, "learning_rate": 7.485332123203079e-05, "loss": 0.0146, "step": 36770 }, { "grad_norm": 0.12179891765117645, "learning_rate": 7.483897250916504e-05, "loss": 0.0211, "step": 36780 }, { "grad_norm": 0.15208107233047485, "learning_rate": 7.48246210699466e-05, "loss": 0.0142, "step": 36790 }, { "grad_norm": 0.13906708359718323, "learning_rate": 7.481026691594492e-05, "loss": 0.0122, "step": 36800 }, { "grad_norm": 0.10028204321861267, "learning_rate": 7.479591004872975e-05, "loss": 0.012, "step": 36810 }, { "grad_norm": 0.187600240111351, "learning_rate": 7.478155046987114e-05, "loss": 0.0163, "step": 36820 }, { "grad_norm": 0.14491067826747894, "learning_rate": 7.476718818093944e-05, "loss": 0.0126, "step": 36830 }, { "grad_norm": 0.11604642122983932, "learning_rate": 7.475282318350527e-05, "loss": 0.0127, "step": 36840 }, { "grad_norm": 0.12826670706272125, "learning_rate": 7.473845547913957e-05, "loss": 0.0146, "step": 36850 }, { "grad_norm": 0.197089284658432, "learning_rate": 7.472408506941357e-05, "loss": 0.0183, "step": 36860 }, { "grad_norm": 0.15843576192855835, "learning_rate": 7.470971195589881e-05, "loss": 0.0136, "step": 36870 }, { "grad_norm": 0.12153293937444687, "learning_rate": 7.46953361401671e-05, "loss": 0.0212, "step": 36880 }, { "grad_norm": 0.12858465313911438, "learning_rate": 7.468095762379055e-05, "loss": 0.014, "step": 36890 }, { "grad_norm": 0.12988251447677612, "learning_rate": 7.466657640834158e-05, "loss": 0.0171, "step": 36900 }, { "grad_norm": 0.1425446718931198, "learning_rate": 7.46521924953929e-05, "loss": 0.0149, "step": 36910 }, { "grad_norm": 0.16371573507785797, "learning_rate": 7.463780588651752e-05, "loss": 0.0158, "step": 36920 }, { "grad_norm": 0.1292228251695633, "learning_rate": 7.462341658328873e-05, "loss": 0.0147, "step": 36930 }, { "grad_norm": 0.20088721811771393, "learning_rate": 7.460902458728012e-05, "loss": 0.0151, "step": 36940 }, { "grad_norm": 0.1102236732840538, "learning_rate": 7.459462990006558e-05, "loss": 0.0093, "step": 36950 }, { "grad_norm": 0.1542031466960907, "learning_rate": 7.458023252321929e-05, "loss": 0.0134, "step": 36960 }, { "grad_norm": 0.13850197196006775, "learning_rate": 7.456583245831573e-05, "loss": 0.0125, "step": 36970 }, { "grad_norm": 0.15751318633556366, "learning_rate": 7.455142970692968e-05, "loss": 0.0225, "step": 36980 }, { "grad_norm": 0.1466890573501587, "learning_rate": 7.453702427063615e-05, "loss": 0.0141, "step": 36990 }, { "grad_norm": 0.1621260792016983, "learning_rate": 7.452261615101057e-05, "loss": 0.0134, "step": 37000 }, { "grad_norm": 0.16819344460964203, "learning_rate": 7.450820534962856e-05, "loss": 0.0154, "step": 37010 }, { "grad_norm": 0.13957224786281586, "learning_rate": 7.449379186806604e-05, "loss": 0.0136, "step": 37020 }, { "grad_norm": 0.10311897099018097, "learning_rate": 7.44793757078993e-05, "loss": 0.0123, "step": 37030 }, { "grad_norm": 0.1891123652458191, "learning_rate": 7.44649568707048e-05, "loss": 0.0127, "step": 37040 }, { "grad_norm": 0.1550455391407013, "learning_rate": 7.445053535805942e-05, "loss": 0.0163, "step": 37050 }, { "grad_norm": 0.10475075244903564, "learning_rate": 7.443611117154024e-05, "loss": 0.0122, "step": 37060 }, { "grad_norm": 0.1398056447505951, "learning_rate": 7.442168431272469e-05, "loss": 0.0128, "step": 37070 }, { "grad_norm": 0.10216837376356125, "learning_rate": 7.440725478319044e-05, "loss": 0.0129, "step": 37080 }, { "grad_norm": 0.15213175117969513, "learning_rate": 7.439282258451553e-05, "loss": 0.0096, "step": 37090 }, { "grad_norm": 0.1550092250108719, "learning_rate": 7.43783877182782e-05, "loss": 0.0119, "step": 37100 }, { "grad_norm": 0.1598026156425476, "learning_rate": 7.436395018605707e-05, "loss": 0.0124, "step": 37110 }, { "grad_norm": 0.12745800614356995, "learning_rate": 7.434950998943094e-05, "loss": 0.0126, "step": 37120 }, { "grad_norm": 0.10021385550498962, "learning_rate": 7.433506712997902e-05, "loss": 0.0156, "step": 37130 }, { "grad_norm": 0.1415201872587204, "learning_rate": 7.432062160928077e-05, "loss": 0.012, "step": 37140 }, { "grad_norm": 0.104552261531353, "learning_rate": 7.430617342891588e-05, "loss": 0.0098, "step": 37150 }, { "grad_norm": 0.1086924597620964, "learning_rate": 7.429172259046442e-05, "loss": 0.0145, "step": 37160 }, { "grad_norm": 0.13844381272792816, "learning_rate": 7.42772690955067e-05, "loss": 0.0103, "step": 37170 }, { "grad_norm": 0.17307159304618835, "learning_rate": 7.426281294562332e-05, "loss": 0.0232, "step": 37180 }, { "grad_norm": 0.12540674209594727, "learning_rate": 7.424835414239519e-05, "loss": 0.0148, "step": 37190 }, { "grad_norm": 0.15942297875881195, "learning_rate": 7.423389268740352e-05, "loss": 0.0138, "step": 37200 }, { "grad_norm": 0.11913767457008362, "learning_rate": 7.42194285822298e-05, "loss": 0.012, "step": 37210 }, { "grad_norm": 0.12804731726646423, "learning_rate": 7.420496182845576e-05, "loss": 0.0146, "step": 37220 }, { "grad_norm": 0.11282625794410706, "learning_rate": 7.419049242766348e-05, "loss": 0.0176, "step": 37230 }, { "grad_norm": 0.18749304115772247, "learning_rate": 7.417602038143533e-05, "loss": 0.0127, "step": 37240 }, { "grad_norm": 0.121027372777462, "learning_rate": 7.416154569135393e-05, "loss": 0.0137, "step": 37250 }, { "grad_norm": 0.09945876151323318, "learning_rate": 7.414706835900222e-05, "loss": 0.0121, "step": 37260 }, { "grad_norm": 0.12091279029846191, "learning_rate": 7.413258838596345e-05, "loss": 0.0143, "step": 37270 }, { "grad_norm": 0.14442992210388184, "learning_rate": 7.411810577382106e-05, "loss": 0.0163, "step": 37280 }, { "grad_norm": 0.09909835457801819, "learning_rate": 7.410362052415889e-05, "loss": 0.013, "step": 37290 }, { "grad_norm": 0.1518879532814026, "learning_rate": 7.408913263856102e-05, "loss": 0.0122, "step": 37300 }, { "grad_norm": 0.15207256376743317, "learning_rate": 7.407464211861183e-05, "loss": 0.0172, "step": 37310 }, { "grad_norm": 0.1398969292640686, "learning_rate": 7.406014896589597e-05, "loss": 0.0188, "step": 37320 }, { "grad_norm": 0.1564943641424179, "learning_rate": 7.404565318199838e-05, "loss": 0.0135, "step": 37330 }, { "grad_norm": 0.11268486082553864, "learning_rate": 7.403115476850432e-05, "loss": 0.0166, "step": 37340 }, { "grad_norm": 0.12665176391601562, "learning_rate": 7.401665372699932e-05, "loss": 0.0121, "step": 37350 }, { "grad_norm": 0.10953173041343689, "learning_rate": 7.400215005906916e-05, "loss": 0.0136, "step": 37360 }, { "grad_norm": 0.1374029964208603, "learning_rate": 7.398764376629996e-05, "loss": 0.0129, "step": 37370 }, { "grad_norm": 0.10364102572202682, "learning_rate": 7.39731348502781e-05, "loss": 0.0126, "step": 37380 }, { "grad_norm": 0.07332503795623779, "learning_rate": 7.395862331259028e-05, "loss": 0.0108, "step": 37390 }, { "grad_norm": 0.11531106382608414, "learning_rate": 7.394410915482341e-05, "loss": 0.0114, "step": 37400 }, { "grad_norm": 0.10945329815149307, "learning_rate": 7.392959237856476e-05, "loss": 0.0148, "step": 37410 }, { "grad_norm": 0.1590920239686966, "learning_rate": 7.391507298540187e-05, "loss": 0.0124, "step": 37420 }, { "grad_norm": 0.12084680050611496, "learning_rate": 7.390055097692257e-05, "loss": 0.01, "step": 37430 }, { "grad_norm": 0.18157696723937988, "learning_rate": 7.388602635471494e-05, "loss": 0.0265, "step": 37440 }, { "grad_norm": 0.1544891893863678, "learning_rate": 7.387149912036738e-05, "loss": 0.0212, "step": 37450 }, { "grad_norm": 0.14699116349220276, "learning_rate": 7.385696927546854e-05, "loss": 0.0122, "step": 37460 }, { "grad_norm": 0.13518209755420685, "learning_rate": 7.384243682160743e-05, "loss": 0.0274, "step": 37470 }, { "grad_norm": 0.09602486342191696, "learning_rate": 7.382790176037329e-05, "loss": 0.012, "step": 37480 }, { "grad_norm": 0.19510313868522644, "learning_rate": 7.381336409335562e-05, "loss": 0.014, "step": 37490 }, { "grad_norm": 0.17103485763072968, "learning_rate": 7.379882382214426e-05, "loss": 0.0148, "step": 37500 }, { "grad_norm": 0.11689331382513046, "learning_rate": 7.378428094832931e-05, "loss": 0.0155, "step": 37510 }, { "grad_norm": 0.20693475008010864, "learning_rate": 7.376973547350114e-05, "loss": 0.0187, "step": 37520 }, { "grad_norm": 0.22058017551898956, "learning_rate": 7.375518739925046e-05, "loss": 0.014, "step": 37530 }, { "grad_norm": 0.1397780179977417, "learning_rate": 7.374063672716818e-05, "loss": 0.0149, "step": 37540 }, { "grad_norm": 0.16327348351478577, "learning_rate": 7.372608345884558e-05, "loss": 0.0142, "step": 37550 }, { "grad_norm": 0.21126100420951843, "learning_rate": 7.371152759587415e-05, "loss": 0.0165, "step": 37560 }, { "grad_norm": 0.136809304356575, "learning_rate": 7.369696913984572e-05, "loss": 0.0108, "step": 37570 }, { "grad_norm": 0.11314646899700165, "learning_rate": 7.368240809235237e-05, "loss": 0.0127, "step": 37580 }, { "grad_norm": 0.11308804154396057, "learning_rate": 7.366784445498649e-05, "loss": 0.0122, "step": 37590 }, { "grad_norm": 0.13760033249855042, "learning_rate": 7.36532782293407e-05, "loss": 0.0122, "step": 37600 }, { "grad_norm": 0.12878228724002838, "learning_rate": 7.363870941700797e-05, "loss": 0.0119, "step": 37610 }, { "grad_norm": 0.11371923238039017, "learning_rate": 7.362413801958152e-05, "loss": 0.0121, "step": 37620 }, { "grad_norm": 0.19611488282680511, "learning_rate": 7.360956403865486e-05, "loss": 0.013, "step": 37630 }, { "grad_norm": 0.19133217632770538, "learning_rate": 7.359498747582177e-05, "loss": 0.0111, "step": 37640 }, { "grad_norm": 0.08989164233207703, "learning_rate": 7.358040833267634e-05, "loss": 0.0117, "step": 37650 }, { "grad_norm": 0.17412103712558746, "learning_rate": 7.356582661081288e-05, "loss": 0.0176, "step": 37660 }, { "grad_norm": 0.11361213028430939, "learning_rate": 7.355124231182607e-05, "loss": 0.0117, "step": 37670 }, { "grad_norm": 0.15046265721321106, "learning_rate": 7.353665543731083e-05, "loss": 0.014, "step": 37680 }, { "grad_norm": 0.10483214259147644, "learning_rate": 7.352206598886229e-05, "loss": 0.0129, "step": 37690 }, { "grad_norm": 0.18087199330329895, "learning_rate": 7.350747396807601e-05, "loss": 0.0139, "step": 37700 }, { "grad_norm": 0.20149901509284973, "learning_rate": 7.349287937654772e-05, "loss": 0.0094, "step": 37710 }, { "grad_norm": 0.1463003158569336, "learning_rate": 7.347828221587345e-05, "loss": 0.0149, "step": 37720 }, { "grad_norm": 0.12610022723674774, "learning_rate": 7.346368248764955e-05, "loss": 0.0098, "step": 37730 }, { "grad_norm": 0.14958436787128448, "learning_rate": 7.34490801934726e-05, "loss": 0.0115, "step": 37740 }, { "grad_norm": 0.16937512159347534, "learning_rate": 7.343447533493947e-05, "loss": 0.0174, "step": 37750 }, { "grad_norm": 0.1787475347518921, "learning_rate": 7.341986791364739e-05, "loss": 0.0223, "step": 37760 }, { "grad_norm": 0.15276624262332916, "learning_rate": 7.340525793119374e-05, "loss": 0.0155, "step": 37770 }, { "grad_norm": 0.21617773175239563, "learning_rate": 7.33906453891763e-05, "loss": 0.0131, "step": 37780 }, { "grad_norm": 0.11063027381896973, "learning_rate": 7.337603028919303e-05, "loss": 0.0113, "step": 37790 }, { "grad_norm": 0.106223464012146, "learning_rate": 7.336141263284225e-05, "loss": 0.0117, "step": 37800 }, { "grad_norm": 0.1059565320611, "learning_rate": 7.334679242172249e-05, "loss": 0.0158, "step": 37810 }, { "grad_norm": 0.15094849467277527, "learning_rate": 7.33321696574326e-05, "loss": 0.0148, "step": 37820 }, { "grad_norm": 0.1000073179602623, "learning_rate": 7.331754434157174e-05, "loss": 0.0133, "step": 37830 }, { "grad_norm": 0.17194047570228577, "learning_rate": 7.330291647573928e-05, "loss": 0.0154, "step": 37840 }, { "grad_norm": 0.11989780515432358, "learning_rate": 7.32882860615349e-05, "loss": 0.0131, "step": 37850 }, { "grad_norm": 0.1611221879720688, "learning_rate": 7.327365310055858e-05, "loss": 0.0153, "step": 37860 }, { "grad_norm": 0.11750607192516327, "learning_rate": 7.325901759441055e-05, "loss": 0.0111, "step": 37870 }, { "grad_norm": 0.14333407580852509, "learning_rate": 7.324437954469132e-05, "loss": 0.0194, "step": 37880 }, { "grad_norm": 0.17293965816497803, "learning_rate": 7.32297389530017e-05, "loss": 0.013, "step": 37890 }, { "grad_norm": 0.11489316821098328, "learning_rate": 7.321509582094274e-05, "loss": 0.0126, "step": 37900 }, { "grad_norm": 0.12784676253795624, "learning_rate": 7.320045015011582e-05, "loss": 0.0159, "step": 37910 }, { "grad_norm": 0.17857234179973602, "learning_rate": 7.318580194212253e-05, "loss": 0.0149, "step": 37920 }, { "grad_norm": 0.14310146868228912, "learning_rate": 7.317115119856483e-05, "loss": 0.0155, "step": 37930 }, { "grad_norm": 0.13278958201408386, "learning_rate": 7.315649792104485e-05, "loss": 0.0145, "step": 37940 }, { "grad_norm": 0.14005273580551147, "learning_rate": 7.314184211116507e-05, "loss": 0.0191, "step": 37950 }, { "grad_norm": 0.18227048218250275, "learning_rate": 7.312718377052823e-05, "loss": 0.0143, "step": 37960 }, { "grad_norm": 0.11295033246278763, "learning_rate": 7.311252290073737e-05, "loss": 0.0144, "step": 37970 }, { "grad_norm": 0.0974084734916687, "learning_rate": 7.309785950339572e-05, "loss": 0.0112, "step": 37980 }, { "grad_norm": 0.1375725418329239, "learning_rate": 7.308319358010692e-05, "loss": 0.0114, "step": 37990 }, { "grad_norm": 0.13483776152133942, "learning_rate": 7.306852513247474e-05, "loss": 0.0315, "step": 38000 }, { "grad_norm": 0.13407854735851288, "learning_rate": 7.305385416210336e-05, "loss": 0.0221, "step": 38010 }, { "grad_norm": 0.1266838014125824, "learning_rate": 7.303918067059711e-05, "loss": 0.0113, "step": 38020 }, { "grad_norm": 0.1315474659204483, "learning_rate": 7.302450465956075e-05, "loss": 0.0143, "step": 38030 }, { "grad_norm": 0.07407040148973465, "learning_rate": 7.300982613059914e-05, "loss": 0.0113, "step": 38040 }, { "grad_norm": 0.13850794732570648, "learning_rate": 7.299514508531757e-05, "loss": 0.0219, "step": 38050 }, { "grad_norm": 0.1266389936208725, "learning_rate": 7.298046152532148e-05, "loss": 0.016, "step": 38060 }, { "grad_norm": 0.09627415984869003, "learning_rate": 7.296577545221668e-05, "loss": 0.0123, "step": 38070 }, { "grad_norm": 0.26361721754074097, "learning_rate": 7.29510868676092e-05, "loss": 0.019, "step": 38080 }, { "grad_norm": 0.14364176988601685, "learning_rate": 7.293639577310538e-05, "loss": 0.0122, "step": 38090 }, { "grad_norm": 0.1504969447851181, "learning_rate": 7.292170217031179e-05, "loss": 0.0149, "step": 38100 }, { "grad_norm": 0.21566040813922882, "learning_rate": 7.290700606083532e-05, "loss": 0.015, "step": 38110 }, { "grad_norm": 0.17464755475521088, "learning_rate": 7.28923074462831e-05, "loss": 0.0151, "step": 38120 }, { "grad_norm": 0.1283191740512848, "learning_rate": 7.287760632826255e-05, "loss": 0.0181, "step": 38130 }, { "grad_norm": 0.2713989317417145, "learning_rate": 7.286290270838139e-05, "loss": 0.0214, "step": 38140 }, { "grad_norm": 0.1701163500547409, "learning_rate": 7.284819658824756e-05, "loss": 0.0125, "step": 38150 }, { "grad_norm": 0.11153954267501831, "learning_rate": 7.283348796946929e-05, "loss": 0.0155, "step": 38160 }, { "grad_norm": 0.1524384319782257, "learning_rate": 7.28187768536551e-05, "loss": 0.0153, "step": 38170 }, { "grad_norm": 0.1366666555404663, "learning_rate": 7.280406324241379e-05, "loss": 0.0165, "step": 38180 }, { "grad_norm": 0.13390421867370605, "learning_rate": 7.278934713735441e-05, "loss": 0.0177, "step": 38190 }, { "grad_norm": 0.11070642620325089, "learning_rate": 7.277462854008629e-05, "loss": 0.015, "step": 38200 }, { "grad_norm": 0.20335660874843597, "learning_rate": 7.275990745221904e-05, "loss": 0.0137, "step": 38210 }, { "grad_norm": 0.12868773937225342, "learning_rate": 7.274518387536252e-05, "loss": 0.0132, "step": 38220 }, { "grad_norm": 0.21186527609825134, "learning_rate": 7.273045781112689e-05, "loss": 0.0184, "step": 38230 }, { "grad_norm": 0.1605752557516098, "learning_rate": 7.271572926112257e-05, "loss": 0.0145, "step": 38240 }, { "grad_norm": 0.1105654165148735, "learning_rate": 7.270099822696024e-05, "loss": 0.012, "step": 38250 }, { "grad_norm": 0.09596675634384155, "learning_rate": 7.26862647102509e-05, "loss": 0.011, "step": 38260 }, { "grad_norm": 0.1627744883298874, "learning_rate": 7.267152871260573e-05, "loss": 0.0175, "step": 38270 }, { "grad_norm": 0.19390439987182617, "learning_rate": 7.265679023563629e-05, "loss": 0.0144, "step": 38280 }, { "grad_norm": 0.13621750473976135, "learning_rate": 7.264204928095433e-05, "loss": 0.0112, "step": 38290 }, { "grad_norm": 0.23674948513507843, "learning_rate": 7.262730585017188e-05, "loss": 0.0164, "step": 38300 }, { "grad_norm": 0.14664611220359802, "learning_rate": 7.261255994490131e-05, "loss": 0.0146, "step": 38310 }, { "grad_norm": 0.12895910441875458, "learning_rate": 7.259781156675517e-05, "loss": 0.0176, "step": 38320 }, { "grad_norm": 0.1201181635260582, "learning_rate": 7.258306071734633e-05, "loss": 0.0133, "step": 38330 }, { "grad_norm": 0.12306084483861923, "learning_rate": 7.256830739828792e-05, "loss": 0.0143, "step": 38340 }, { "grad_norm": 0.08712329715490341, "learning_rate": 7.255355161119336e-05, "loss": 0.0091, "step": 38350 }, { "grad_norm": 0.0715397447347641, "learning_rate": 7.253879335767628e-05, "loss": 0.0081, "step": 38360 }, { "grad_norm": 0.09079896658658981, "learning_rate": 7.252403263935067e-05, "loss": 0.011, "step": 38370 }, { "grad_norm": 0.18698787689208984, "learning_rate": 7.25092694578307e-05, "loss": 0.0133, "step": 38380 }, { "grad_norm": 0.14482688903808594, "learning_rate": 7.249450381473087e-05, "loss": 0.012, "step": 38390 }, { "grad_norm": 0.12488937377929688, "learning_rate": 7.247973571166593e-05, "loss": 0.0159, "step": 38400 }, { "grad_norm": 0.15154369175434113, "learning_rate": 7.246496515025089e-05, "loss": 0.0132, "step": 38410 }, { "grad_norm": 0.215743750333786, "learning_rate": 7.245019213210101e-05, "loss": 0.0194, "step": 38420 }, { "grad_norm": 0.13386881351470947, "learning_rate": 7.24354166588319e-05, "loss": 0.0158, "step": 38430 }, { "grad_norm": 0.21223706007003784, "learning_rate": 7.242063873205935e-05, "loss": 0.0177, "step": 38440 }, { "grad_norm": 0.17751376330852509, "learning_rate": 7.240585835339946e-05, "loss": 0.0159, "step": 38450 }, { "grad_norm": 0.10816005617380142, "learning_rate": 7.239107552446857e-05, "loss": 0.0139, "step": 38460 }, { "grad_norm": 0.14033207297325134, "learning_rate": 7.237629024688334e-05, "loss": 0.014, "step": 38470 }, { "grad_norm": 0.13230498135089874, "learning_rate": 7.236150252226064e-05, "loss": 0.0164, "step": 38480 }, { "grad_norm": 0.13643041253089905, "learning_rate": 7.234671235221765e-05, "loss": 0.0209, "step": 38490 }, { "grad_norm": 0.21830140054225922, "learning_rate": 7.233191973837179e-05, "loss": 0.0154, "step": 38500 }, { "grad_norm": 0.15163959562778473, "learning_rate": 7.231712468234078e-05, "loss": 0.012, "step": 38510 }, { "grad_norm": 0.17082516849040985, "learning_rate": 7.230232718574254e-05, "loss": 0.015, "step": 38520 }, { "grad_norm": 0.11204662173986435, "learning_rate": 7.228752725019535e-05, "loss": 0.0131, "step": 38530 }, { "grad_norm": 0.13055524230003357, "learning_rate": 7.227272487731769e-05, "loss": 0.0133, "step": 38540 }, { "grad_norm": 0.11800076812505722, "learning_rate": 7.225792006872831e-05, "loss": 0.0104, "step": 38550 }, { "grad_norm": 0.10779824107885361, "learning_rate": 7.224311282604628e-05, "loss": 0.0117, "step": 38560 }, { "grad_norm": 0.14157143235206604, "learning_rate": 7.222830315089085e-05, "loss": 0.0153, "step": 38570 }, { "grad_norm": 0.13389867544174194, "learning_rate": 7.22134910448816e-05, "loss": 0.0136, "step": 38580 }, { "grad_norm": 0.17971667647361755, "learning_rate": 7.219867650963839e-05, "loss": 0.0137, "step": 38590 }, { "grad_norm": 0.12839476764202118, "learning_rate": 7.21838595467813e-05, "loss": 0.0203, "step": 38600 }, { "grad_norm": 0.0934281051158905, "learning_rate": 7.216904015793067e-05, "loss": 0.0132, "step": 38610 }, { "grad_norm": 0.10515926033258438, "learning_rate": 7.215421834470713e-05, "loss": 0.0102, "step": 38620 }, { "grad_norm": 0.14084526896476746, "learning_rate": 7.213939410873158e-05, "loss": 0.018, "step": 38630 }, { "grad_norm": 0.1249292641878128, "learning_rate": 7.212456745162519e-05, "loss": 0.0126, "step": 38640 }, { "grad_norm": 0.07949822396039963, "learning_rate": 7.210973837500937e-05, "loss": 0.0154, "step": 38650 }, { "grad_norm": 0.08391036838293076, "learning_rate": 7.209490688050578e-05, "loss": 0.0097, "step": 38660 }, { "grad_norm": 0.10840349644422531, "learning_rate": 7.208007296973641e-05, "loss": 0.0122, "step": 38670 }, { "grad_norm": 0.11976775527000427, "learning_rate": 7.206523664432345e-05, "loss": 0.0155, "step": 38680 }, { "grad_norm": 0.1012842208147049, "learning_rate": 7.205039790588939e-05, "loss": 0.0111, "step": 38690 }, { "grad_norm": 0.2353155016899109, "learning_rate": 7.203555675605697e-05, "loss": 0.0147, "step": 38700 }, { "grad_norm": 0.13947439193725586, "learning_rate": 7.202071319644917e-05, "loss": 0.0156, "step": 38710 }, { "grad_norm": 0.19663667678833008, "learning_rate": 7.200586722868932e-05, "loss": 0.0142, "step": 38720 }, { "grad_norm": 0.14613790810108185, "learning_rate": 7.199101885440088e-05, "loss": 0.0118, "step": 38730 }, { "grad_norm": 0.12855340540409088, "learning_rate": 7.197616807520771e-05, "loss": 0.0107, "step": 38740 }, { "grad_norm": 0.13031023740768433, "learning_rate": 7.196131489273381e-05, "loss": 0.0119, "step": 38750 }, { "grad_norm": 0.12470577657222748, "learning_rate": 7.194645930860355e-05, "loss": 0.0129, "step": 38760 }, { "grad_norm": 0.12846839427947998, "learning_rate": 7.19316013244415e-05, "loss": 0.014, "step": 38770 }, { "grad_norm": 0.16068890690803528, "learning_rate": 7.191674094187248e-05, "loss": 0.0116, "step": 38780 }, { "grad_norm": 0.25180861353874207, "learning_rate": 7.190187816252165e-05, "loss": 0.0118, "step": 38790 }, { "grad_norm": 0.11248548328876495, "learning_rate": 7.188701298801435e-05, "loss": 0.0108, "step": 38800 }, { "grad_norm": 0.1531849354505539, "learning_rate": 7.18721454199762e-05, "loss": 0.0147, "step": 38810 }, { "grad_norm": 0.11575910449028015, "learning_rate": 7.185727546003312e-05, "loss": 0.0138, "step": 38820 }, { "grad_norm": 0.24650290608406067, "learning_rate": 7.184240310981126e-05, "loss": 0.0153, "step": 38830 }, { "grad_norm": 0.1750471442937851, "learning_rate": 7.182752837093704e-05, "loss": 0.0149, "step": 38840 }, { "grad_norm": 0.17003493010997772, "learning_rate": 7.181265124503711e-05, "loss": 0.022, "step": 38850 }, { "grad_norm": 0.11178122460842133, "learning_rate": 7.179777173373847e-05, "loss": 0.012, "step": 38860 }, { "grad_norm": 0.12151578813791275, "learning_rate": 7.178288983866826e-05, "loss": 0.0102, "step": 38870 }, { "grad_norm": 0.17769493162631989, "learning_rate": 7.176800556145397e-05, "loss": 0.0166, "step": 38880 }, { "grad_norm": 0.12495467066764832, "learning_rate": 7.175311890372334e-05, "loss": 0.0149, "step": 38890 }, { "grad_norm": 0.1318967193365097, "learning_rate": 7.17382298671043e-05, "loss": 0.0128, "step": 38900 }, { "grad_norm": 0.21271729469299316, "learning_rate": 7.172333845322515e-05, "loss": 0.0147, "step": 38910 }, { "grad_norm": 0.14607152342796326, "learning_rate": 7.170844466371436e-05, "loss": 0.0167, "step": 38920 }, { "grad_norm": 0.17691786587238312, "learning_rate": 7.16935485002007e-05, "loss": 0.014, "step": 38930 }, { "grad_norm": 0.12424389272928238, "learning_rate": 7.167864996431319e-05, "loss": 0.0155, "step": 38940 }, { "grad_norm": 0.14305034279823303, "learning_rate": 7.166374905768111e-05, "loss": 0.0118, "step": 38950 }, { "grad_norm": 0.1074095219373703, "learning_rate": 7.164884578193401e-05, "loss": 0.012, "step": 38960 }, { "grad_norm": 0.1938868910074234, "learning_rate": 7.163394013870168e-05, "loss": 0.0121, "step": 38970 }, { "grad_norm": 0.1747424304485321, "learning_rate": 7.161903212961419e-05, "loss": 0.0167, "step": 38980 }, { "grad_norm": 0.15798138082027435, "learning_rate": 7.160412175630183e-05, "loss": 0.015, "step": 38990 }, { "grad_norm": 0.09030789136886597, "learning_rate": 7.158920902039521e-05, "loss": 0.0103, "step": 39000 }, { "grad_norm": 0.1155998557806015, "learning_rate": 7.157429392352514e-05, "loss": 0.0125, "step": 39010 }, { "grad_norm": 0.14670531451702118, "learning_rate": 7.155937646732275e-05, "loss": 0.0124, "step": 39020 }, { "grad_norm": 0.1628723293542862, "learning_rate": 7.154445665341933e-05, "loss": 0.0152, "step": 39030 }, { "grad_norm": 0.17097517848014832, "learning_rate": 7.152953448344654e-05, "loss": 0.0128, "step": 39040 }, { "grad_norm": 0.18786360323429108, "learning_rate": 7.151460995903624e-05, "loss": 0.0146, "step": 39050 }, { "grad_norm": 0.1484297662973404, "learning_rate": 7.149968308182052e-05, "loss": 0.0118, "step": 39060 }, { "grad_norm": 0.17936798930168152, "learning_rate": 7.14847538534318e-05, "loss": 0.0143, "step": 39070 }, { "grad_norm": 0.13373394310474396, "learning_rate": 7.14698222755027e-05, "loss": 0.0123, "step": 39080 }, { "grad_norm": 0.21652790904045105, "learning_rate": 7.14548883496661e-05, "loss": 0.0128, "step": 39090 }, { "grad_norm": 0.10200341045856476, "learning_rate": 7.143995207755517e-05, "loss": 0.0144, "step": 39100 }, { "grad_norm": 0.1334475725889206, "learning_rate": 7.142501346080333e-05, "loss": 0.0137, "step": 39110 }, { "grad_norm": 0.16285410523414612, "learning_rate": 7.141007250104421e-05, "loss": 0.0094, "step": 39120 }, { "grad_norm": 0.12255088984966278, "learning_rate": 7.139512919991176e-05, "loss": 0.0118, "step": 39130 }, { "grad_norm": 0.12907099723815918, "learning_rate": 7.138018355904014e-05, "loss": 0.0103, "step": 39140 }, { "grad_norm": 0.16493172943592072, "learning_rate": 7.13652355800638e-05, "loss": 0.0229, "step": 39150 }, { "grad_norm": 0.17645488679409027, "learning_rate": 7.135028526461743e-05, "loss": 0.0124, "step": 39160 }, { "grad_norm": 0.1374884992837906, "learning_rate": 7.133533261433594e-05, "loss": 0.0101, "step": 39170 }, { "grad_norm": 0.10610408335924149, "learning_rate": 7.132037763085457e-05, "loss": 0.0115, "step": 39180 }, { "grad_norm": 0.17212477326393127, "learning_rate": 7.130542031580875e-05, "loss": 0.0125, "step": 39190 }, { "grad_norm": 0.13879051804542542, "learning_rate": 7.12904606708342e-05, "loss": 0.0109, "step": 39200 }, { "grad_norm": 0.16803741455078125, "learning_rate": 7.127549869756687e-05, "loss": 0.0171, "step": 39210 }, { "grad_norm": 0.12154465913772583, "learning_rate": 7.126053439764299e-05, "loss": 0.0118, "step": 39220 }, { "grad_norm": 0.14169566333293915, "learning_rate": 7.124556777269904e-05, "loss": 0.0134, "step": 39230 }, { "grad_norm": 0.15783189237117767, "learning_rate": 7.123059882437174e-05, "loss": 0.0162, "step": 39240 }, { "grad_norm": 0.1174846813082695, "learning_rate": 7.121562755429807e-05, "loss": 0.0108, "step": 39250 }, { "grad_norm": 0.10166124254465103, "learning_rate": 7.120065396411528e-05, "loss": 0.0113, "step": 39260 }, { "grad_norm": 0.11970546096563339, "learning_rate": 7.118567805546084e-05, "loss": 0.0115, "step": 39270 }, { "grad_norm": 0.18070656061172485, "learning_rate": 7.117069982997248e-05, "loss": 0.0155, "step": 39280 }, { "grad_norm": 0.2842753529548645, "learning_rate": 7.115571928928825e-05, "loss": 0.0264, "step": 39290 }, { "grad_norm": 0.19988210499286652, "learning_rate": 7.114073643504635e-05, "loss": 0.0143, "step": 39300 }, { "grad_norm": 0.11497554928064346, "learning_rate": 7.11257512688853e-05, "loss": 0.0125, "step": 39310 }, { "grad_norm": 0.10817674547433853, "learning_rate": 7.111076379244384e-05, "loss": 0.0135, "step": 39320 }, { "grad_norm": 0.10005325824022293, "learning_rate": 7.109577400736101e-05, "loss": 0.0148, "step": 39330 }, { "grad_norm": 0.18416132032871246, "learning_rate": 7.108078191527605e-05, "loss": 0.0145, "step": 39340 }, { "grad_norm": 0.1556960493326187, "learning_rate": 7.106578751782847e-05, "loss": 0.018, "step": 39350 }, { "grad_norm": 0.13292062282562256, "learning_rate": 7.105079081665803e-05, "loss": 0.0131, "step": 39360 }, { "grad_norm": 0.19082798063755035, "learning_rate": 7.103579181340476e-05, "loss": 0.0134, "step": 39370 }, { "grad_norm": 0.12524616718292236, "learning_rate": 7.102079050970893e-05, "loss": 0.0122, "step": 39380 }, { "grad_norm": 0.10904527455568314, "learning_rate": 7.100578690721105e-05, "loss": 0.0151, "step": 39390 }, { "grad_norm": 0.15827980637550354, "learning_rate": 7.09907810075519e-05, "loss": 0.0212, "step": 39400 }, { "grad_norm": 0.11553139239549637, "learning_rate": 7.097577281237249e-05, "loss": 0.0244, "step": 39410 }, { "grad_norm": 0.11589419841766357, "learning_rate": 7.09607623233141e-05, "loss": 0.0109, "step": 39420 }, { "grad_norm": 0.20858754217624664, "learning_rate": 7.094574954201827e-05, "loss": 0.0118, "step": 39430 }, { "grad_norm": 0.17995549738407135, "learning_rate": 7.093073447012675e-05, "loss": 0.0144, "step": 39440 }, { "grad_norm": 0.13980630040168762, "learning_rate": 7.09157171092816e-05, "loss": 0.011, "step": 39450 }, { "grad_norm": 0.1492944359779358, "learning_rate": 7.090069746112504e-05, "loss": 0.0109, "step": 39460 }, { "grad_norm": 0.20591847598552704, "learning_rate": 7.088567552729965e-05, "loss": 0.0103, "step": 39470 }, { "grad_norm": 0.15074960887432098, "learning_rate": 7.087065130944818e-05, "loss": 0.015, "step": 39480 }, { "grad_norm": 0.16699238121509552, "learning_rate": 7.085562480921366e-05, "loss": 0.0221, "step": 39490 }, { "grad_norm": 0.14776355028152466, "learning_rate": 7.084059602823937e-05, "loss": 0.0102, "step": 39500 }, { "grad_norm": 0.15043002367019653, "learning_rate": 7.082556496816882e-05, "loss": 0.0112, "step": 39510 }, { "grad_norm": 0.17133209109306335, "learning_rate": 7.081053163064582e-05, "loss": 0.0152, "step": 39520 }, { "grad_norm": 0.19592930376529694, "learning_rate": 7.079549601731434e-05, "loss": 0.0138, "step": 39530 }, { "grad_norm": 0.16545647382736206, "learning_rate": 7.07804581298187e-05, "loss": 0.0134, "step": 39540 }, { "grad_norm": 0.2136538177728653, "learning_rate": 7.07654179698034e-05, "loss": 0.0188, "step": 39550 }, { "grad_norm": 0.2650187313556671, "learning_rate": 7.075037553891321e-05, "loss": 0.0184, "step": 39560 }, { "grad_norm": 0.12149874120950699, "learning_rate": 7.073533083879315e-05, "loss": 0.0101, "step": 39570 }, { "grad_norm": 0.1576312631368637, "learning_rate": 7.072028387108849e-05, "loss": 0.0149, "step": 39580 }, { "grad_norm": 0.16778530180454254, "learning_rate": 7.070523463744472e-05, "loss": 0.0177, "step": 39590 }, { "grad_norm": 0.09740392863750458, "learning_rate": 7.069018313950763e-05, "loss": 0.0127, "step": 39600 }, { "grad_norm": 0.1332392543554306, "learning_rate": 7.067512937892322e-05, "loss": 0.0116, "step": 39610 }, { "grad_norm": 0.11791837960481644, "learning_rate": 7.066007335733775e-05, "loss": 0.0111, "step": 39620 }, { "grad_norm": 0.08616424351930618, "learning_rate": 7.064501507639772e-05, "loss": 0.0156, "step": 39630 }, { "grad_norm": 0.11363820731639862, "learning_rate": 7.062995453774987e-05, "loss": 0.0216, "step": 39640 }, { "grad_norm": 0.15769396722316742, "learning_rate": 7.061489174304121e-05, "loss": 0.0265, "step": 39650 }, { "grad_norm": 0.13106676936149597, "learning_rate": 7.0599826693919e-05, "loss": 0.0129, "step": 39660 }, { "grad_norm": 0.12519890069961548, "learning_rate": 7.05847593920307e-05, "loss": 0.0123, "step": 39670 }, { "grad_norm": 0.15493907034397125, "learning_rate": 7.056968983902406e-05, "loss": 0.0127, "step": 39680 }, { "grad_norm": 0.10755806416273117, "learning_rate": 7.055461803654706e-05, "loss": 0.0119, "step": 39690 }, { "grad_norm": 0.11645198613405228, "learning_rate": 7.053954398624794e-05, "loss": 0.0092, "step": 39700 }, { "grad_norm": 0.10832381248474121, "learning_rate": 7.052446768977518e-05, "loss": 0.0124, "step": 39710 }, { "grad_norm": 0.11066807061433792, "learning_rate": 7.050938914877748e-05, "loss": 0.0101, "step": 39720 }, { "grad_norm": 0.14520876109600067, "learning_rate": 7.049430836490382e-05, "loss": 0.0195, "step": 39730 }, { "grad_norm": 0.09070251882076263, "learning_rate": 7.047922533980342e-05, "loss": 0.0126, "step": 39740 }, { "grad_norm": 0.15934215486049652, "learning_rate": 7.046414007512571e-05, "loss": 0.0147, "step": 39750 }, { "grad_norm": 0.10308230668306351, "learning_rate": 7.044905257252042e-05, "loss": 0.0138, "step": 39760 }, { "grad_norm": 0.11879683285951614, "learning_rate": 7.04339628336375e-05, "loss": 0.0158, "step": 39770 }, { "grad_norm": 0.12341247498989105, "learning_rate": 7.041887086012711e-05, "loss": 0.0147, "step": 39780 }, { "grad_norm": 0.10106221586465836, "learning_rate": 7.040377665363969e-05, "loss": 0.0133, "step": 39790 }, { "grad_norm": 0.14723029732704163, "learning_rate": 7.038868021582594e-05, "loss": 0.0169, "step": 39800 }, { "grad_norm": 0.10896766185760498, "learning_rate": 7.037358154833679e-05, "loss": 0.0142, "step": 39810 }, { "grad_norm": 0.138983353972435, "learning_rate": 7.035848065282339e-05, "loss": 0.0178, "step": 39820 }, { "grad_norm": 0.15677592158317566, "learning_rate": 7.034337753093714e-05, "loss": 0.0112, "step": 39830 }, { "grad_norm": 0.11263074725866318, "learning_rate": 7.032827218432972e-05, "loss": 0.0129, "step": 39840 }, { "grad_norm": 0.10902433842420578, "learning_rate": 7.031316461465302e-05, "loss": 0.0122, "step": 39850 }, { "grad_norm": 0.19567826390266418, "learning_rate": 7.029805482355915e-05, "loss": 0.0122, "step": 39860 }, { "grad_norm": 0.1389392763376236, "learning_rate": 7.028294281270055e-05, "loss": 0.0122, "step": 39870 }, { "grad_norm": 0.12462152540683746, "learning_rate": 7.026782858372982e-05, "loss": 0.0164, "step": 39880 }, { "grad_norm": 0.11892783641815186, "learning_rate": 7.025271213829982e-05, "loss": 0.0126, "step": 39890 }, { "grad_norm": 0.10144128650426865, "learning_rate": 7.023759347806366e-05, "loss": 0.0097, "step": 39900 }, { "grad_norm": 0.088493712246418, "learning_rate": 7.022247260467472e-05, "loss": 0.0146, "step": 39910 }, { "grad_norm": 0.1575644165277481, "learning_rate": 7.020734951978658e-05, "loss": 0.0149, "step": 39920 }, { "grad_norm": 0.10997825115919113, "learning_rate": 7.019222422505307e-05, "loss": 0.0157, "step": 39930 }, { "grad_norm": 0.1422276347875595, "learning_rate": 7.017709672212827e-05, "loss": 0.014, "step": 39940 }, { "grad_norm": 0.1413479447364807, "learning_rate": 7.016196701266652e-05, "loss": 0.0128, "step": 39950 }, { "grad_norm": 0.11039081960916519, "learning_rate": 7.014683509832238e-05, "loss": 0.0124, "step": 39960 }, { "grad_norm": 0.11131061613559723, "learning_rate": 7.013170098075063e-05, "loss": 0.016, "step": 39970 }, { "grad_norm": 0.09484359622001648, "learning_rate": 7.011656466160632e-05, "loss": 0.0155, "step": 39980 }, { "grad_norm": 0.1213250532746315, "learning_rate": 7.010142614254475e-05, "loss": 0.0133, "step": 39990 }, { "grad_norm": 0.15504902601242065, "learning_rate": 7.008628542522147e-05, "loss": 0.0171, "step": 40000 }, { "grad_norm": 0.13263897597789764, "learning_rate": 7.007114251129218e-05, "loss": 0.0142, "step": 40010 }, { "grad_norm": 0.1123414933681488, "learning_rate": 7.005599740241293e-05, "loss": 0.0118, "step": 40020 }, { "grad_norm": 0.156277596950531, "learning_rate": 7.004085010023996e-05, "loss": 0.0182, "step": 40030 }, { "grad_norm": 0.13091687858104706, "learning_rate": 7.002570060642976e-05, "loss": 0.0127, "step": 40040 }, { "grad_norm": 0.14826175570487976, "learning_rate": 7.001054892263903e-05, "loss": 0.0141, "step": 40050 }, { "grad_norm": 0.13555774092674255, "learning_rate": 6.999539505052477e-05, "loss": 0.0127, "step": 40060 }, { "grad_norm": 0.1473482996225357, "learning_rate": 6.998023899174418e-05, "loss": 0.0172, "step": 40070 }, { "grad_norm": 0.11470769345760345, "learning_rate": 6.996508074795467e-05, "loss": 0.0122, "step": 40080 }, { "grad_norm": 0.18524444103240967, "learning_rate": 6.994992032081396e-05, "loss": 0.0165, "step": 40090 }, { "grad_norm": 0.13979576528072357, "learning_rate": 6.993475771197995e-05, "loss": 0.0129, "step": 40100 }, { "grad_norm": 0.11218640953302383, "learning_rate": 6.991959292311082e-05, "loss": 0.0123, "step": 40110 }, { "grad_norm": 0.11328458786010742, "learning_rate": 6.990442595586495e-05, "loss": 0.0134, "step": 40120 }, { "grad_norm": 0.12696434557437897, "learning_rate": 6.988925681190098e-05, "loss": 0.0135, "step": 40130 }, { "grad_norm": 0.32529863715171814, "learning_rate": 6.987408549287778e-05, "loss": 0.0161, "step": 40140 }, { "grad_norm": 0.24715380370616913, "learning_rate": 6.985891200045449e-05, "loss": 0.017, "step": 40150 }, { "grad_norm": 0.1517224758863449, "learning_rate": 6.984373633629045e-05, "loss": 0.0159, "step": 40160 }, { "grad_norm": 0.12428157776594162, "learning_rate": 6.98285585020452e-05, "loss": 0.0117, "step": 40170 }, { "grad_norm": 0.12336643785238266, "learning_rate": 6.981337849937864e-05, "loss": 0.0126, "step": 40180 }, { "grad_norm": 0.1226772889494896, "learning_rate": 6.979819632995078e-05, "loss": 0.0147, "step": 40190 }, { "grad_norm": 0.20873840153217316, "learning_rate": 6.978301199542193e-05, "loss": 0.0104, "step": 40200 }, { "grad_norm": 0.1852899044752121, "learning_rate": 6.976782549745263e-05, "loss": 0.0148, "step": 40210 }, { "grad_norm": 0.1713477224111557, "learning_rate": 6.975263683770367e-05, "loss": 0.0162, "step": 40220 }, { "grad_norm": 0.08126157522201538, "learning_rate": 6.973744601783604e-05, "loss": 0.011, "step": 40230 }, { "grad_norm": 0.12164821475744247, "learning_rate": 6.972225303951097e-05, "loss": 0.0145, "step": 40240 }, { "grad_norm": 0.12252837419509888, "learning_rate": 6.970705790438998e-05, "loss": 0.0106, "step": 40250 }, { "grad_norm": 0.22078517079353333, "learning_rate": 6.969186061413477e-05, "loss": 0.0144, "step": 40260 }, { "grad_norm": 0.1552283614873886, "learning_rate": 6.967666117040727e-05, "loss": 0.0198, "step": 40270 }, { "grad_norm": 0.11138316988945007, "learning_rate": 6.966145957486972e-05, "loss": 0.0113, "step": 40280 }, { "grad_norm": 0.1204969733953476, "learning_rate": 6.964625582918449e-05, "loss": 0.0163, "step": 40290 }, { "grad_norm": 0.10544255375862122, "learning_rate": 6.963104993501425e-05, "loss": 0.0123, "step": 40300 }, { "grad_norm": 0.1789569854736328, "learning_rate": 6.961584189402192e-05, "loss": 0.0149, "step": 40310 }, { "grad_norm": 0.0972258448600769, "learning_rate": 6.96006317078706e-05, "loss": 0.0112, "step": 40320 }, { "grad_norm": 0.12666098773479462, "learning_rate": 6.958541937822367e-05, "loss": 0.0127, "step": 40330 }, { "grad_norm": 0.1395150125026703, "learning_rate": 6.957020490674473e-05, "loss": 0.0115, "step": 40340 }, { "grad_norm": 0.1017850711941719, "learning_rate": 6.95549882950976e-05, "loss": 0.0112, "step": 40350 }, { "grad_norm": 0.13619327545166016, "learning_rate": 6.953976954494635e-05, "loss": 0.0118, "step": 40360 }, { "grad_norm": 0.1285516768693924, "learning_rate": 6.952454865795528e-05, "loss": 0.0126, "step": 40370 }, { "grad_norm": 0.1363363265991211, "learning_rate": 6.950932563578892e-05, "loss": 0.0134, "step": 40380 }, { "grad_norm": 0.13545197248458862, "learning_rate": 6.949410048011206e-05, "loss": 0.0105, "step": 40390 }, { "grad_norm": 0.11510040611028671, "learning_rate": 6.947887319258966e-05, "loss": 0.0133, "step": 40400 }, { "grad_norm": 0.07956603914499283, "learning_rate": 6.946364377488696e-05, "loss": 0.0114, "step": 40410 }, { "grad_norm": 0.10014050453901291, "learning_rate": 6.944841222866947e-05, "loss": 0.0129, "step": 40420 }, { "grad_norm": 0.14843514561653137, "learning_rate": 6.943317855560284e-05, "loss": 0.0165, "step": 40430 }, { "grad_norm": 0.16591349244117737, "learning_rate": 6.941794275735302e-05, "loss": 0.013, "step": 40440 }, { "grad_norm": 0.12351690232753754, "learning_rate": 6.94027048355862e-05, "loss": 0.0125, "step": 40450 }, { "grad_norm": 0.12886206805706024, "learning_rate": 6.938746479196871e-05, "loss": 0.0132, "step": 40460 }, { "grad_norm": 0.10139632970094681, "learning_rate": 6.937222262816724e-05, "loss": 0.0105, "step": 40470 }, { "grad_norm": 0.175113245844841, "learning_rate": 6.935697834584865e-05, "loss": 0.015, "step": 40480 }, { "grad_norm": 0.09824590384960175, "learning_rate": 6.934173194667998e-05, "loss": 0.0101, "step": 40490 }, { "grad_norm": 0.1251617670059204, "learning_rate": 6.93264834323286e-05, "loss": 0.0112, "step": 40500 }, { "grad_norm": 0.09592065960168839, "learning_rate": 6.931123280446206e-05, "loss": 0.0144, "step": 40510 }, { "grad_norm": 0.10197791457176208, "learning_rate": 6.929598006474811e-05, "loss": 0.0125, "step": 40520 }, { "grad_norm": 0.10740215331315994, "learning_rate": 6.92807252148548e-05, "loss": 0.0118, "step": 40530 }, { "grad_norm": 0.11292636394500732, "learning_rate": 6.926546825645037e-05, "loss": 0.0098, "step": 40540 }, { "grad_norm": 0.12403862178325653, "learning_rate": 6.92502091912033e-05, "loss": 0.013, "step": 40550 }, { "grad_norm": 0.15079624950885773, "learning_rate": 6.923494802078229e-05, "loss": 0.0126, "step": 40560 }, { "grad_norm": 0.13348805904388428, "learning_rate": 6.921968474685628e-05, "loss": 0.0103, "step": 40570 }, { "grad_norm": 0.11188987642526627, "learning_rate": 6.920441937109445e-05, "loss": 0.0123, "step": 40580 }, { "grad_norm": 0.07382240891456604, "learning_rate": 6.91891518951662e-05, "loss": 0.0118, "step": 40590 }, { "grad_norm": 0.14526593685150146, "learning_rate": 6.917388232074114e-05, "loss": 0.0129, "step": 40600 }, { "grad_norm": 0.16019782423973083, "learning_rate": 6.915861064948914e-05, "loss": 0.0132, "step": 40610 }, { "grad_norm": 0.13610686361789703, "learning_rate": 6.914333688308025e-05, "loss": 0.0124, "step": 40620 }, { "grad_norm": 0.21460549533367157, "learning_rate": 6.912806102318487e-05, "loss": 0.0118, "step": 40630 }, { "grad_norm": 0.11186230182647705, "learning_rate": 6.911278307147347e-05, "loss": 0.0125, "step": 40640 }, { "grad_norm": 0.09247685968875885, "learning_rate": 6.909750302961684e-05, "loss": 0.0112, "step": 40650 }, { "grad_norm": 0.1571117639541626, "learning_rate": 6.9082220899286e-05, "loss": 0.0166, "step": 40660 }, { "grad_norm": 0.15352660417556763, "learning_rate": 6.906693668215219e-05, "loss": 0.0162, "step": 40670 }, { "grad_norm": 0.09839093685150146, "learning_rate": 6.905165037988683e-05, "loss": 0.0086, "step": 40680 }, { "grad_norm": 0.1044887900352478, "learning_rate": 6.903636199416164e-05, "loss": 0.0129, "step": 40690 }, { "grad_norm": 0.1856784224510193, "learning_rate": 6.902107152664851e-05, "loss": 0.0109, "step": 40700 }, { "grad_norm": 0.09473677724599838, "learning_rate": 6.90057789790196e-05, "loss": 0.0107, "step": 40710 }, { "grad_norm": 0.13711021840572357, "learning_rate": 6.899048435294728e-05, "loss": 0.0138, "step": 40720 }, { "grad_norm": 0.1303950399160385, "learning_rate": 6.897518765010415e-05, "loss": 0.0132, "step": 40730 }, { "grad_norm": 0.12519964575767517, "learning_rate": 6.895988887216302e-05, "loss": 0.0119, "step": 40740 }, { "grad_norm": 0.14051179587841034, "learning_rate": 6.894458802079694e-05, "loss": 0.0152, "step": 40750 }, { "grad_norm": 0.11185856908559799, "learning_rate": 6.892928509767922e-05, "loss": 0.0106, "step": 40760 }, { "grad_norm": 0.11213730275630951, "learning_rate": 6.891398010448333e-05, "loss": 0.0119, "step": 40770 }, { "grad_norm": 0.23484867811203003, "learning_rate": 6.8898673042883e-05, "loss": 0.0152, "step": 40780 }, { "grad_norm": 0.12567108869552612, "learning_rate": 6.888336391455222e-05, "loss": 0.0125, "step": 40790 }, { "grad_norm": 0.08305318653583527, "learning_rate": 6.886805272116513e-05, "loss": 0.011, "step": 40800 }, { "grad_norm": 0.16578012704849243, "learning_rate": 6.885273946439617e-05, "loss": 0.0171, "step": 40810 }, { "grad_norm": 0.16370271146297455, "learning_rate": 6.883742414591998e-05, "loss": 0.0128, "step": 40820 }, { "grad_norm": 0.16018274426460266, "learning_rate": 6.88221067674114e-05, "loss": 0.0103, "step": 40830 }, { "grad_norm": 0.10817090421915054, "learning_rate": 6.88067873305455e-05, "loss": 0.0145, "step": 40840 }, { "grad_norm": 0.09122972935438156, "learning_rate": 6.879146583699765e-05, "loss": 0.0096, "step": 40850 }, { "grad_norm": 0.12123257666826248, "learning_rate": 6.877614228844334e-05, "loss": 0.0116, "step": 40860 }, { "grad_norm": 0.1580544412136078, "learning_rate": 6.876081668655832e-05, "loss": 0.0137, "step": 40870 }, { "grad_norm": 0.1602715402841568, "learning_rate": 6.874548903301861e-05, "loss": 0.0164, "step": 40880 }, { "grad_norm": 0.11886641383171082, "learning_rate": 6.873015932950039e-05, "loss": 0.0149, "step": 40890 }, { "grad_norm": 0.12344121932983398, "learning_rate": 6.871482757768012e-05, "loss": 0.014, "step": 40900 }, { "grad_norm": 0.18305253982543945, "learning_rate": 6.869949377923442e-05, "loss": 0.0149, "step": 40910 }, { "grad_norm": 0.14525367319583893, "learning_rate": 6.868415793584022e-05, "loss": 0.0119, "step": 40920 }, { "grad_norm": 0.08388325572013855, "learning_rate": 6.86688200491746e-05, "loss": 0.0117, "step": 40930 }, { "grad_norm": 0.08742943406105042, "learning_rate": 6.865348012091488e-05, "loss": 0.0113, "step": 40940 }, { "grad_norm": 0.10226310789585114, "learning_rate": 6.86381381527386e-05, "loss": 0.0144, "step": 40950 }, { "grad_norm": 0.09324365109205246, "learning_rate": 6.862279414632359e-05, "loss": 0.0109, "step": 40960 }, { "grad_norm": 0.18151938915252686, "learning_rate": 6.860744810334778e-05, "loss": 0.0139, "step": 40970 }, { "grad_norm": 0.09979728609323502, "learning_rate": 6.859210002548943e-05, "loss": 0.0103, "step": 40980 }, { "grad_norm": 0.1365964114665985, "learning_rate": 6.8576749914427e-05, "loss": 0.0102, "step": 40990 }, { "grad_norm": 0.16278818249702454, "learning_rate": 6.85613977718391e-05, "loss": 0.014, "step": 41000 }, { "grad_norm": 0.12920096516609192, "learning_rate": 6.854604359940464e-05, "loss": 0.01, "step": 41010 }, { "grad_norm": 0.12382589280605316, "learning_rate": 6.853068739880276e-05, "loss": 0.0101, "step": 41020 }, { "grad_norm": 0.21639636158943176, "learning_rate": 6.851532917171276e-05, "loss": 0.018, "step": 41030 }, { "grad_norm": 0.1510554403066635, "learning_rate": 6.849996891981418e-05, "loss": 0.0156, "step": 41040 }, { "grad_norm": 0.14777536690235138, "learning_rate": 6.848460664478684e-05, "loss": 0.0171, "step": 41050 }, { "grad_norm": 0.1424853503704071, "learning_rate": 6.846924234831069e-05, "loss": 0.0108, "step": 41060 }, { "grad_norm": 0.12171002477407455, "learning_rate": 6.845387603206597e-05, "loss": 0.014, "step": 41070 }, { "grad_norm": 0.06050686910748482, "learning_rate": 6.843850769773312e-05, "loss": 0.0095, "step": 41080 }, { "grad_norm": 0.20407329499721527, "learning_rate": 6.842313734699278e-05, "loss": 0.0211, "step": 41090 }, { "grad_norm": 0.11224640905857086, "learning_rate": 6.840776498152584e-05, "loss": 0.0136, "step": 41100 }, { "grad_norm": 0.13802696764469147, "learning_rate": 6.83923906030134e-05, "loss": 0.013, "step": 41110 }, { "grad_norm": 0.18260003626346588, "learning_rate": 6.837701421313677e-05, "loss": 0.0191, "step": 41120 }, { "grad_norm": 0.1723165512084961, "learning_rate": 6.83616358135775e-05, "loss": 0.0141, "step": 41130 }, { "grad_norm": 0.19324417412281036, "learning_rate": 6.834625540601734e-05, "loss": 0.0168, "step": 41140 }, { "grad_norm": 0.1027955561876297, "learning_rate": 6.833087299213829e-05, "loss": 0.0109, "step": 41150 }, { "grad_norm": 0.10943940281867981, "learning_rate": 6.83154885736225e-05, "loss": 0.0178, "step": 41160 }, { "grad_norm": 0.2560480833053589, "learning_rate": 6.830010215215244e-05, "loss": 0.0145, "step": 41170 }, { "grad_norm": 0.1359318047761917, "learning_rate": 6.828471372941072e-05, "loss": 0.0126, "step": 41180 }, { "grad_norm": 0.1496620923280716, "learning_rate": 6.826932330708018e-05, "loss": 0.0127, "step": 41190 }, { "grad_norm": 0.12306775152683258, "learning_rate": 6.825393088684393e-05, "loss": 0.0107, "step": 41200 }, { "grad_norm": 0.11250198632478714, "learning_rate": 6.823853647038524e-05, "loss": 0.0106, "step": 41210 }, { "grad_norm": 0.09248264133930206, "learning_rate": 6.822314005938762e-05, "loss": 0.0097, "step": 41220 }, { "grad_norm": 0.14567086100578308, "learning_rate": 6.820774165553479e-05, "loss": 0.0172, "step": 41230 }, { "grad_norm": 0.10270217061042786, "learning_rate": 6.819234126051072e-05, "loss": 0.0099, "step": 41240 }, { "grad_norm": 0.15056900680065155, "learning_rate": 6.817693887599956e-05, "loss": 0.0127, "step": 41250 }, { "grad_norm": 0.16212786734104156, "learning_rate": 6.816153450368568e-05, "loss": 0.019, "step": 41260 }, { "grad_norm": 0.13115715980529785, "learning_rate": 6.814612814525372e-05, "loss": 0.0184, "step": 41270 }, { "grad_norm": 0.12051386386156082, "learning_rate": 6.813071980238846e-05, "loss": 0.0118, "step": 41280 }, { "grad_norm": 0.1190991997718811, "learning_rate": 6.811530947677492e-05, "loss": 0.0098, "step": 41290 }, { "grad_norm": 0.1090923622250557, "learning_rate": 6.809989717009839e-05, "loss": 0.0085, "step": 41300 }, { "grad_norm": 0.1611652970314026, "learning_rate": 6.808448288404431e-05, "loss": 0.0159, "step": 41310 }, { "grad_norm": 0.1121593713760376, "learning_rate": 6.806906662029838e-05, "loss": 0.0119, "step": 41320 }, { "grad_norm": 0.1461515724658966, "learning_rate": 6.805364838054647e-05, "loss": 0.0111, "step": 41330 }, { "grad_norm": 0.14775945246219635, "learning_rate": 6.803822816647474e-05, "loss": 0.01, "step": 41340 }, { "grad_norm": 0.1263197511434555, "learning_rate": 6.802280597976949e-05, "loss": 0.013, "step": 41350 }, { "grad_norm": 0.13802319765090942, "learning_rate": 6.800738182211727e-05, "loss": 0.0111, "step": 41360 }, { "grad_norm": 0.14630800485610962, "learning_rate": 6.799195569520485e-05, "loss": 0.0141, "step": 41370 }, { "grad_norm": 0.13064195215702057, "learning_rate": 6.79765276007192e-05, "loss": 0.0113, "step": 41380 }, { "grad_norm": 0.08174263685941696, "learning_rate": 6.796109754034753e-05, "loss": 0.0106, "step": 41390 }, { "grad_norm": 0.12775841355323792, "learning_rate": 6.794566551577724e-05, "loss": 0.0112, "step": 41400 }, { "grad_norm": 0.10228373855352402, "learning_rate": 6.793023152869594e-05, "loss": 0.0108, "step": 41410 }, { "grad_norm": 0.08088314533233643, "learning_rate": 6.791479558079147e-05, "loss": 0.0107, "step": 41420 }, { "grad_norm": 0.06324728578329086, "learning_rate": 6.789935767375191e-05, "loss": 0.0116, "step": 41430 }, { "grad_norm": 0.08018845319747925, "learning_rate": 6.78839178092655e-05, "loss": 0.0092, "step": 41440 }, { "grad_norm": 0.1260446012020111, "learning_rate": 6.786847598902072e-05, "loss": 0.0133, "step": 41450 }, { "grad_norm": 0.17017202079296112, "learning_rate": 6.785303221470627e-05, "loss": 0.015, "step": 41460 }, { "grad_norm": 0.13039253652095795, "learning_rate": 6.783758648801108e-05, "loss": 0.0118, "step": 41470 }, { "grad_norm": 0.25308331847190857, "learning_rate": 6.782213881062422e-05, "loss": 0.0179, "step": 41480 }, { "grad_norm": 0.1890164613723755, "learning_rate": 6.780668918423508e-05, "loss": 0.0123, "step": 41490 }, { "grad_norm": 0.16286540031433105, "learning_rate": 6.779123761053317e-05, "loss": 0.0111, "step": 41500 }, { "grad_norm": 0.15892094373703003, "learning_rate": 6.777578409120826e-05, "loss": 0.0137, "step": 41510 }, { "grad_norm": 0.23225435614585876, "learning_rate": 6.776032862795033e-05, "loss": 0.0166, "step": 41520 }, { "grad_norm": 0.12972640991210938, "learning_rate": 6.774487122244956e-05, "loss": 0.0154, "step": 41530 }, { "grad_norm": 0.15031152963638306, "learning_rate": 6.772941187639637e-05, "loss": 0.0122, "step": 41540 }, { "grad_norm": 0.10954799503087997, "learning_rate": 6.771395059148134e-05, "loss": 0.0111, "step": 41550 }, { "grad_norm": 0.1141083613038063, "learning_rate": 6.76984873693953e-05, "loss": 0.0107, "step": 41560 }, { "grad_norm": 0.21546998620033264, "learning_rate": 6.768302221182931e-05, "loss": 0.0164, "step": 41570 }, { "grad_norm": 0.17228257656097412, "learning_rate": 6.766755512047457e-05, "loss": 0.0202, "step": 41580 }, { "grad_norm": 0.1760707050561905, "learning_rate": 6.765208609702259e-05, "loss": 0.0124, "step": 41590 }, { "grad_norm": 0.10505620390176773, "learning_rate": 6.763661514316499e-05, "loss": 0.0095, "step": 41600 }, { "grad_norm": 0.14744918048381805, "learning_rate": 6.76211422605937e-05, "loss": 0.0147, "step": 41610 }, { "grad_norm": 0.22374974191188812, "learning_rate": 6.760566745100076e-05, "loss": 0.0132, "step": 41620 }, { "grad_norm": 0.1339452862739563, "learning_rate": 6.75901907160785e-05, "loss": 0.014, "step": 41630 }, { "grad_norm": 0.11815745383501053, "learning_rate": 6.757471205751943e-05, "loss": 0.0101, "step": 41640 }, { "grad_norm": 0.12678910791873932, "learning_rate": 6.75592314770163e-05, "loss": 0.0128, "step": 41650 }, { "grad_norm": 0.09192796051502228, "learning_rate": 6.7543748976262e-05, "loss": 0.0167, "step": 41660 }, { "grad_norm": 0.12168188393115997, "learning_rate": 6.752826455694968e-05, "loss": 0.0119, "step": 41670 }, { "grad_norm": 0.10259765386581421, "learning_rate": 6.751277822077271e-05, "loss": 0.0127, "step": 41680 }, { "grad_norm": 0.0801655650138855, "learning_rate": 6.749728996942465e-05, "loss": 0.0108, "step": 41690 }, { "grad_norm": 0.10447437316179276, "learning_rate": 6.748179980459924e-05, "loss": 0.0111, "step": 41700 }, { "grad_norm": 0.1439155489206314, "learning_rate": 6.746630772799052e-05, "loss": 0.0106, "step": 41710 }, { "grad_norm": 0.18105827271938324, "learning_rate": 6.745081374129262e-05, "loss": 0.0147, "step": 41720 }, { "grad_norm": 0.2540067732334137, "learning_rate": 6.743531784619998e-05, "loss": 0.0183, "step": 41730 }, { "grad_norm": 0.16201810538768768, "learning_rate": 6.741982004440719e-05, "loss": 0.011, "step": 41740 }, { "grad_norm": 0.1631482094526291, "learning_rate": 6.740432033760907e-05, "loss": 0.0145, "step": 41750 }, { "grad_norm": 0.11909011751413345, "learning_rate": 6.738881872750066e-05, "loss": 0.0115, "step": 41760 }, { "grad_norm": 0.12193771451711655, "learning_rate": 6.737331521577715e-05, "loss": 0.014, "step": 41770 }, { "grad_norm": 0.22537274658679962, "learning_rate": 6.735780980413403e-05, "loss": 0.0126, "step": 41780 }, { "grad_norm": 0.1438610851764679, "learning_rate": 6.734230249426692e-05, "loss": 0.0152, "step": 41790 }, { "grad_norm": 0.12156292796134949, "learning_rate": 6.732679328787168e-05, "loss": 0.0122, "step": 41800 }, { "grad_norm": 0.09166331589221954, "learning_rate": 6.731128218664438e-05, "loss": 0.0107, "step": 41810 }, { "grad_norm": 0.11378705501556396, "learning_rate": 6.729576919228129e-05, "loss": 0.0088, "step": 41820 }, { "grad_norm": 0.26866912841796875, "learning_rate": 6.728025430647888e-05, "loss": 0.0127, "step": 41830 }, { "grad_norm": 0.13864289224147797, "learning_rate": 6.726473753093383e-05, "loss": 0.0142, "step": 41840 }, { "grad_norm": 0.19547927379608154, "learning_rate": 6.724921886734305e-05, "loss": 0.0116, "step": 41850 }, { "grad_norm": 0.13174360990524292, "learning_rate": 6.723369831740363e-05, "loss": 0.0221, "step": 41860 }, { "grad_norm": 0.10187169164419174, "learning_rate": 6.721817588281288e-05, "loss": 0.0116, "step": 41870 }, { "grad_norm": 0.10357700288295746, "learning_rate": 6.720265156526828e-05, "loss": 0.0111, "step": 41880 }, { "grad_norm": 0.07910235971212387, "learning_rate": 6.718712536646758e-05, "loss": 0.0125, "step": 41890 }, { "grad_norm": 0.08639627695083618, "learning_rate": 6.71715972881087e-05, "loss": 0.0155, "step": 41900 }, { "grad_norm": 0.21192039549350739, "learning_rate": 6.715606733188974e-05, "loss": 0.0134, "step": 41910 }, { "grad_norm": 0.10480035096406937, "learning_rate": 6.714053549950909e-05, "loss": 0.0097, "step": 41920 }, { "grad_norm": 0.12984298169612885, "learning_rate": 6.71250017926652e-05, "loss": 0.0138, "step": 41930 }, { "grad_norm": 0.1469232439994812, "learning_rate": 6.71094662130569e-05, "loss": 0.0097, "step": 41940 }, { "grad_norm": 0.12057623267173767, "learning_rate": 6.709392876238307e-05, "loss": 0.0141, "step": 41950 }, { "grad_norm": 0.11990490555763245, "learning_rate": 6.707838944234289e-05, "loss": 0.0141, "step": 41960 }, { "grad_norm": 0.16873480379581451, "learning_rate": 6.706284825463574e-05, "loss": 0.0126, "step": 41970 }, { "grad_norm": 0.14141727983951569, "learning_rate": 6.704730520096114e-05, "loss": 0.011, "step": 41980 }, { "grad_norm": 0.13799379765987396, "learning_rate": 6.703176028301888e-05, "loss": 0.015, "step": 41990 }, { "grad_norm": 0.12094707041978836, "learning_rate": 6.701621350250892e-05, "loss": 0.0101, "step": 42000 }, { "grad_norm": 0.08468421548604965, "learning_rate": 6.700066486113144e-05, "loss": 0.0095, "step": 42010 }, { "grad_norm": 0.16118481755256653, "learning_rate": 6.69851143605868e-05, "loss": 0.0128, "step": 42020 }, { "grad_norm": 0.1367938220500946, "learning_rate": 6.696956200257561e-05, "loss": 0.0159, "step": 42030 }, { "grad_norm": 0.17070740461349487, "learning_rate": 6.695400778879863e-05, "loss": 0.0227, "step": 42040 }, { "grad_norm": 0.24439755082130432, "learning_rate": 6.693845172095683e-05, "loss": 0.0184, "step": 42050 }, { "grad_norm": 0.11304056644439697, "learning_rate": 6.692289380075142e-05, "loss": 0.0129, "step": 42060 }, { "grad_norm": 0.0911850854754448, "learning_rate": 6.690733402988379e-05, "loss": 0.0112, "step": 42070 }, { "grad_norm": 0.3488553762435913, "learning_rate": 6.689177241005553e-05, "loss": 0.0147, "step": 42080 }, { "grad_norm": 0.1747228056192398, "learning_rate": 6.687620894296844e-05, "loss": 0.0155, "step": 42090 }, { "grad_norm": 0.13539764285087585, "learning_rate": 6.686064363032451e-05, "loss": 0.0141, "step": 42100 }, { "grad_norm": 0.15016447007656097, "learning_rate": 6.684507647382596e-05, "loss": 0.0126, "step": 42110 }, { "grad_norm": 0.12988103926181793, "learning_rate": 6.682950747517513e-05, "loss": 0.0158, "step": 42120 }, { "grad_norm": 0.0769573226571083, "learning_rate": 6.68139366360747e-05, "loss": 0.0121, "step": 42130 }, { "grad_norm": 0.22088724374771118, "learning_rate": 6.679836395822744e-05, "loss": 0.0109, "step": 42140 }, { "grad_norm": 0.2023039162158966, "learning_rate": 6.678278944333633e-05, "loss": 0.0139, "step": 42150 }, { "grad_norm": 0.14247865974903107, "learning_rate": 6.676721309310462e-05, "loss": 0.012, "step": 42160 }, { "grad_norm": 0.0912599265575409, "learning_rate": 6.67516349092357e-05, "loss": 0.0108, "step": 42170 }, { "grad_norm": 0.14057138562202454, "learning_rate": 6.673605489343315e-05, "loss": 0.0099, "step": 42180 }, { "grad_norm": 0.12797416746616364, "learning_rate": 6.672047304740081e-05, "loss": 0.0116, "step": 42190 }, { "grad_norm": 0.12006757408380508, "learning_rate": 6.670488937284268e-05, "loss": 0.0104, "step": 42200 }, { "grad_norm": 0.15492357313632965, "learning_rate": 6.668930387146294e-05, "loss": 0.0119, "step": 42210 }, { "grad_norm": 0.18485109508037567, "learning_rate": 6.667371654496605e-05, "loss": 0.011, "step": 42220 }, { "grad_norm": 0.11264736950397491, "learning_rate": 6.665812739505659e-05, "loss": 0.0123, "step": 42230 }, { "grad_norm": 0.14479243755340576, "learning_rate": 6.664253642343935e-05, "loss": 0.0135, "step": 42240 }, { "grad_norm": 0.09195440262556076, "learning_rate": 6.662694363181935e-05, "loss": 0.0101, "step": 42250 }, { "grad_norm": 0.38602951169013977, "learning_rate": 6.66113490219018e-05, "loss": 0.0146, "step": 42260 }, { "grad_norm": 0.12938156723976135, "learning_rate": 6.659575259539207e-05, "loss": 0.0126, "step": 42270 }, { "grad_norm": 0.14286480844020844, "learning_rate": 6.65801543539958e-05, "loss": 0.0122, "step": 42280 }, { "grad_norm": 0.1398358941078186, "learning_rate": 6.65645542994188e-05, "loss": 0.0126, "step": 42290 }, { "grad_norm": 0.08452747017145157, "learning_rate": 6.654895243336702e-05, "loss": 0.0093, "step": 42300 }, { "grad_norm": 0.19300805032253265, "learning_rate": 6.65333487575467e-05, "loss": 0.0132, "step": 42310 }, { "grad_norm": 0.08493560552597046, "learning_rate": 6.651774327366422e-05, "loss": 0.0108, "step": 42320 }, { "grad_norm": 0.11954116076231003, "learning_rate": 6.650213598342613e-05, "loss": 0.0103, "step": 42330 }, { "grad_norm": 0.13160856068134308, "learning_rate": 6.648652688853931e-05, "loss": 0.0169, "step": 42340 }, { "grad_norm": 0.17807114124298096, "learning_rate": 6.647091599071066e-05, "loss": 0.0125, "step": 42350 }, { "grad_norm": 0.14809665083885193, "learning_rate": 6.645530329164743e-05, "loss": 0.0122, "step": 42360 }, { "grad_norm": 0.0865320935845375, "learning_rate": 6.643968879305694e-05, "loss": 0.013, "step": 42370 }, { "grad_norm": 0.11195749044418335, "learning_rate": 6.642407249664683e-05, "loss": 0.0104, "step": 42380 }, { "grad_norm": 0.13366638123989105, "learning_rate": 6.640845440412483e-05, "loss": 0.0141, "step": 42390 }, { "grad_norm": 0.17816533148288727, "learning_rate": 6.639283451719893e-05, "loss": 0.0122, "step": 42400 }, { "grad_norm": 0.12421491742134094, "learning_rate": 6.637721283757729e-05, "loss": 0.0119, "step": 42410 }, { "grad_norm": 0.06861208379268646, "learning_rate": 6.636158936696828e-05, "loss": 0.0107, "step": 42420 }, { "grad_norm": 0.2807604968547821, "learning_rate": 6.634596410708047e-05, "loss": 0.0135, "step": 42430 }, { "grad_norm": 0.11176956444978714, "learning_rate": 6.633033705962257e-05, "loss": 0.013, "step": 42440 }, { "grad_norm": 0.1455785036087036, "learning_rate": 6.631470822630359e-05, "loss": 0.0125, "step": 42450 }, { "grad_norm": 0.0961223766207695, "learning_rate": 6.629907760883263e-05, "loss": 0.0126, "step": 42460 }, { "grad_norm": 0.11817087233066559, "learning_rate": 6.628344520891905e-05, "loss": 0.0121, "step": 42470 }, { "grad_norm": 0.07807740569114685, "learning_rate": 6.62678110282724e-05, "loss": 0.0085, "step": 42480 }, { "grad_norm": 0.14197635650634766, "learning_rate": 6.625217506860237e-05, "loss": 0.0138, "step": 42490 }, { "grad_norm": 0.09486257284879684, "learning_rate": 6.623653733161892e-05, "loss": 0.0112, "step": 42500 }, { "grad_norm": 0.12217280268669128, "learning_rate": 6.622089781903217e-05, "loss": 0.0128, "step": 42510 }, { "grad_norm": 0.1101468876004219, "learning_rate": 6.620525653255241e-05, "loss": 0.0132, "step": 42520 }, { "grad_norm": 0.16667941212654114, "learning_rate": 6.618961347389015e-05, "loss": 0.0123, "step": 42530 }, { "grad_norm": 0.092415951192379, "learning_rate": 6.617396864475613e-05, "loss": 0.0103, "step": 42540 }, { "grad_norm": 0.11603278666734695, "learning_rate": 6.61583220468612e-05, "loss": 0.0114, "step": 42550 }, { "grad_norm": 0.12866099178791046, "learning_rate": 6.614267368191645e-05, "loss": 0.0108, "step": 42560 }, { "grad_norm": 0.13313449919223785, "learning_rate": 6.61270235516332e-05, "loss": 0.0098, "step": 42570 }, { "grad_norm": 0.1086239367723465, "learning_rate": 6.611137165772287e-05, "loss": 0.0091, "step": 42580 }, { "grad_norm": 0.15917854011058807, "learning_rate": 6.60957180018972e-05, "loss": 0.0136, "step": 42590 }, { "grad_norm": 0.14605212211608887, "learning_rate": 6.608006258586797e-05, "loss": 0.0101, "step": 42600 }, { "grad_norm": 0.10797256976366043, "learning_rate": 6.606440541134731e-05, "loss": 0.0098, "step": 42610 }, { "grad_norm": 0.12205308675765991, "learning_rate": 6.60487464800474e-05, "loss": 0.0204, "step": 42620 }, { "grad_norm": 0.2341097593307495, "learning_rate": 6.603308579368071e-05, "loss": 0.0137, "step": 42630 }, { "grad_norm": 0.2819857895374298, "learning_rate": 6.601742335395987e-05, "loss": 0.0118, "step": 42640 }, { "grad_norm": 0.16888733208179474, "learning_rate": 6.600175916259769e-05, "loss": 0.014, "step": 42650 }, { "grad_norm": 0.11808377504348755, "learning_rate": 6.598609322130718e-05, "loss": 0.0121, "step": 42660 }, { "grad_norm": 0.08668850362300873, "learning_rate": 6.597042553180155e-05, "loss": 0.0093, "step": 42670 }, { "grad_norm": 0.12399778515100479, "learning_rate": 6.59547560957942e-05, "loss": 0.0122, "step": 42680 }, { "grad_norm": 0.14700239896774292, "learning_rate": 6.59390849149987e-05, "loss": 0.0172, "step": 42690 }, { "grad_norm": 0.10941455513238907, "learning_rate": 6.592341199112886e-05, "loss": 0.0102, "step": 42700 }, { "grad_norm": 0.1278400868177414, "learning_rate": 6.590773732589861e-05, "loss": 0.0131, "step": 42710 }, { "grad_norm": 0.15844425559043884, "learning_rate": 6.589206092102213e-05, "loss": 0.0205, "step": 42720 }, { "grad_norm": 0.09758875519037247, "learning_rate": 6.587638277821375e-05, "loss": 0.0104, "step": 42730 }, { "grad_norm": 0.09602918475866318, "learning_rate": 6.586070289918803e-05, "loss": 0.0112, "step": 42740 }, { "grad_norm": 0.12262661755084991, "learning_rate": 6.584502128565968e-05, "loss": 0.0131, "step": 42750 }, { "grad_norm": 0.14794574677944183, "learning_rate": 6.582933793934363e-05, "loss": 0.0114, "step": 42760 }, { "grad_norm": 0.10959822684526443, "learning_rate": 6.581365286195498e-05, "loss": 0.0102, "step": 42770 }, { "grad_norm": 0.12023996561765671, "learning_rate": 6.579796605520903e-05, "loss": 0.0136, "step": 42780 }, { "grad_norm": 0.15561355650424957, "learning_rate": 6.578227752082127e-05, "loss": 0.0133, "step": 42790 }, { "grad_norm": 0.09289511293172836, "learning_rate": 6.576658726050735e-05, "loss": 0.0114, "step": 42800 }, { "grad_norm": 0.10796117782592773, "learning_rate": 6.575089527598316e-05, "loss": 0.0102, "step": 42810 }, { "grad_norm": 0.14300689101219177, "learning_rate": 6.573520156896476e-05, "loss": 0.012, "step": 42820 }, { "grad_norm": 0.19051703810691833, "learning_rate": 6.571950614116835e-05, "loss": 0.0141, "step": 42830 }, { "grad_norm": 0.12659716606140137, "learning_rate": 6.570380899431039e-05, "loss": 0.0235, "step": 42840 }, { "grad_norm": 0.11738480627536774, "learning_rate": 6.568811013010749e-05, "loss": 0.0099, "step": 42850 }, { "grad_norm": 0.13432557880878448, "learning_rate": 6.567240955027645e-05, "loss": 0.013, "step": 42860 }, { "grad_norm": 0.10584986954927444, "learning_rate": 6.565670725653427e-05, "loss": 0.0111, "step": 42870 }, { "grad_norm": 0.09848454594612122, "learning_rate": 6.564100325059812e-05, "loss": 0.0107, "step": 42880 }, { "grad_norm": 0.14063650369644165, "learning_rate": 6.562529753418538e-05, "loss": 0.0127, "step": 42890 }, { "grad_norm": 0.12645521759986877, "learning_rate": 6.56095901090136e-05, "loss": 0.0101, "step": 42900 }, { "grad_norm": 0.08276254683732986, "learning_rate": 6.55938809768005e-05, "loss": 0.0086, "step": 42910 }, { "grad_norm": 0.11827055364847183, "learning_rate": 6.557817013926402e-05, "loss": 0.0107, "step": 42920 }, { "grad_norm": 0.08681438863277435, "learning_rate": 6.556245759812227e-05, "loss": 0.0167, "step": 42930 }, { "grad_norm": 0.1513226479291916, "learning_rate": 6.554674335509357e-05, "loss": 0.0131, "step": 42940 }, { "grad_norm": 0.06725745648145676, "learning_rate": 6.553102741189638e-05, "loss": 0.0092, "step": 42950 }, { "grad_norm": 0.11867929250001907, "learning_rate": 6.55153097702494e-05, "loss": 0.0141, "step": 42960 }, { "grad_norm": 0.11084529757499695, "learning_rate": 6.549959043187146e-05, "loss": 0.0126, "step": 42970 }, { "grad_norm": 0.1494150608778, "learning_rate": 6.54838693984816e-05, "loss": 0.0103, "step": 42980 }, { "grad_norm": 0.1419880986213684, "learning_rate": 6.546814667179909e-05, "loss": 0.0127, "step": 42990 }, { "grad_norm": 0.14657634496688843, "learning_rate": 6.545242225354328e-05, "loss": 0.0122, "step": 43000 }, { "grad_norm": 0.10544062405824661, "learning_rate": 6.543669614543383e-05, "loss": 0.0117, "step": 43010 }, { "grad_norm": 0.13590820133686066, "learning_rate": 6.542096834919049e-05, "loss": 0.0112, "step": 43020 }, { "grad_norm": 0.10413060337305069, "learning_rate": 6.540523886653324e-05, "loss": 0.0151, "step": 43030 }, { "grad_norm": 0.13002237677574158, "learning_rate": 6.538950769918222e-05, "loss": 0.0127, "step": 43040 }, { "grad_norm": 0.10898787528276443, "learning_rate": 6.537377484885779e-05, "loss": 0.0115, "step": 43050 }, { "grad_norm": 0.1276533305644989, "learning_rate": 6.535804031728045e-05, "loss": 0.0131, "step": 43060 }, { "grad_norm": 0.11441117525100708, "learning_rate": 6.534230410617092e-05, "loss": 0.0143, "step": 43070 }, { "grad_norm": 0.11837901175022125, "learning_rate": 6.532656621725007e-05, "loss": 0.012, "step": 43080 }, { "grad_norm": 0.11864915490150452, "learning_rate": 6.531082665223898e-05, "loss": 0.0129, "step": 43090 }, { "grad_norm": 0.12421081960201263, "learning_rate": 6.529508541285889e-05, "loss": 0.0113, "step": 43100 }, { "grad_norm": 0.19929732382297516, "learning_rate": 6.527934250083128e-05, "loss": 0.0133, "step": 43110 }, { "grad_norm": 0.13217808306217194, "learning_rate": 6.526359791787775e-05, "loss": 0.0095, "step": 43120 }, { "grad_norm": 0.11598531156778336, "learning_rate": 6.524785166572009e-05, "loss": 0.0111, "step": 43130 }, { "grad_norm": 0.10005805641412735, "learning_rate": 6.523210374608029e-05, "loss": 0.0122, "step": 43140 }, { "grad_norm": 0.14400769770145416, "learning_rate": 6.521635416068054e-05, "loss": 0.0117, "step": 43150 }, { "grad_norm": 0.12663036584854126, "learning_rate": 6.520060291124317e-05, "loss": 0.0245, "step": 43160 }, { "grad_norm": 0.11582676321268082, "learning_rate": 6.518484999949072e-05, "loss": 0.012, "step": 43170 }, { "grad_norm": 0.10775405913591385, "learning_rate": 6.516909542714591e-05, "loss": 0.0145, "step": 43180 }, { "grad_norm": 0.12541237473487854, "learning_rate": 6.515333919593161e-05, "loss": 0.011, "step": 43190 }, { "grad_norm": 0.10310761630535126, "learning_rate": 6.513758130757094e-05, "loss": 0.012, "step": 43200 }, { "grad_norm": 0.11145896464586258, "learning_rate": 6.512182176378713e-05, "loss": 0.0114, "step": 43210 }, { "grad_norm": 0.1269989162683487, "learning_rate": 6.510606056630362e-05, "loss": 0.0084, "step": 43220 }, { "grad_norm": 0.11198743432760239, "learning_rate": 6.509029771684405e-05, "loss": 0.0134, "step": 43230 }, { "grad_norm": 0.10490661859512329, "learning_rate": 6.50745332171322e-05, "loss": 0.0112, "step": 43240 }, { "grad_norm": 0.11901833862066269, "learning_rate": 6.505876706889207e-05, "loss": 0.0149, "step": 43250 }, { "grad_norm": 0.12373895943164825, "learning_rate": 6.504299927384781e-05, "loss": 0.015, "step": 43260 }, { "grad_norm": 0.15125776827335358, "learning_rate": 6.502722983372378e-05, "loss": 0.0112, "step": 43270 }, { "grad_norm": 0.11466529965400696, "learning_rate": 6.501145875024446e-05, "loss": 0.0145, "step": 43280 }, { "grad_norm": 0.15039785206317902, "learning_rate": 6.499568602513459e-05, "loss": 0.0142, "step": 43290 }, { "grad_norm": 0.1392875760793686, "learning_rate": 6.497991166011903e-05, "loss": 0.0118, "step": 43300 }, { "grad_norm": 0.170091912150383, "learning_rate": 6.49641356569229e-05, "loss": 0.0129, "step": 43310 }, { "grad_norm": 0.09471197426319122, "learning_rate": 6.494835801727135e-05, "loss": 0.0135, "step": 43320 }, { "grad_norm": 0.1051516905426979, "learning_rate": 6.493257874288986e-05, "loss": 0.0127, "step": 43330 }, { "grad_norm": 0.12378174066543579, "learning_rate": 6.491679783550399e-05, "loss": 0.0143, "step": 43340 }, { "grad_norm": 0.14466224610805511, "learning_rate": 6.490101529683957e-05, "loss": 0.011, "step": 43350 }, { "grad_norm": 0.09391418844461441, "learning_rate": 6.488523112862248e-05, "loss": 0.0102, "step": 43360 }, { "grad_norm": 0.1074337288737297, "learning_rate": 6.486944533257891e-05, "loss": 0.0139, "step": 43370 }, { "grad_norm": 0.12867826223373413, "learning_rate": 6.485365791043517e-05, "loss": 0.0141, "step": 43380 }, { "grad_norm": 0.10796276479959488, "learning_rate": 6.483786886391772e-05, "loss": 0.01, "step": 43390 }, { "grad_norm": 0.12331471592187881, "learning_rate": 6.482207819475323e-05, "loss": 0.0092, "step": 43400 }, { "grad_norm": 0.09864349663257599, "learning_rate": 6.480628590466857e-05, "loss": 0.0089, "step": 43410 }, { "grad_norm": 0.13523079454898834, "learning_rate": 6.479049199539074e-05, "loss": 0.0112, "step": 43420 }, { "grad_norm": 0.16636046767234802, "learning_rate": 6.477469646864693e-05, "loss": 0.0122, "step": 43430 }, { "grad_norm": 0.1277722269296646, "learning_rate": 6.475889932616454e-05, "loss": 0.0157, "step": 43440 }, { "grad_norm": 0.12734384834766388, "learning_rate": 6.474310056967111e-05, "loss": 0.0138, "step": 43450 }, { "grad_norm": 0.15600550174713135, "learning_rate": 6.472730020089437e-05, "loss": 0.0129, "step": 43460 }, { "grad_norm": 0.11816271394491196, "learning_rate": 6.471149822156222e-05, "loss": 0.0118, "step": 43470 }, { "grad_norm": 0.07364518195390701, "learning_rate": 6.469569463340275e-05, "loss": 0.0084, "step": 43480 }, { "grad_norm": 0.0887410044670105, "learning_rate": 6.467988943814418e-05, "loss": 0.0118, "step": 43490 }, { "grad_norm": 0.18142977356910706, "learning_rate": 6.4664082637515e-05, "loss": 0.0127, "step": 43500 }, { "grad_norm": 0.16824784874916077, "learning_rate": 6.464827423324377e-05, "loss": 0.0125, "step": 43510 }, { "grad_norm": 0.09088525176048279, "learning_rate": 6.463246422705931e-05, "loss": 0.0115, "step": 43520 }, { "grad_norm": 0.1606583446264267, "learning_rate": 6.461665262069055e-05, "loss": 0.0125, "step": 43530 }, { "grad_norm": 0.10849899798631668, "learning_rate": 6.460083941586665e-05, "loss": 0.0112, "step": 43540 }, { "grad_norm": 0.2570941746234894, "learning_rate": 6.45850246143169e-05, "loss": 0.0094, "step": 43550 }, { "grad_norm": 0.13101966679096222, "learning_rate": 6.456920821777077e-05, "loss": 0.0133, "step": 43560 }, { "grad_norm": 0.1635444313287735, "learning_rate": 6.455339022795796e-05, "loss": 0.0144, "step": 43570 }, { "grad_norm": 0.10030627995729446, "learning_rate": 6.453757064660825e-05, "loss": 0.0121, "step": 43580 }, { "grad_norm": 0.09207304567098618, "learning_rate": 6.452174947545169e-05, "loss": 0.0079, "step": 43590 }, { "grad_norm": 0.10893889516592026, "learning_rate": 6.450592671621842e-05, "loss": 0.0112, "step": 43600 }, { "grad_norm": 0.1320948600769043, "learning_rate": 6.449010237063882e-05, "loss": 0.0127, "step": 43610 }, { "grad_norm": 0.1207612082362175, "learning_rate": 6.447427644044343e-05, "loss": 0.0118, "step": 43620 }, { "grad_norm": 0.1220526248216629, "learning_rate": 6.44584489273629e-05, "loss": 0.011, "step": 43630 }, { "grad_norm": 0.12965457141399384, "learning_rate": 6.444261983312817e-05, "loss": 0.0104, "step": 43640 }, { "grad_norm": 0.10173758119344711, "learning_rate": 6.442678915947023e-05, "loss": 0.0107, "step": 43650 }, { "grad_norm": 0.09277190268039703, "learning_rate": 6.441095690812032e-05, "loss": 0.0089, "step": 43660 }, { "grad_norm": 0.1047152578830719, "learning_rate": 6.439512308080982e-05, "loss": 0.0126, "step": 43670 }, { "grad_norm": 0.09066279232501984, "learning_rate": 6.43792876792703e-05, "loss": 0.0091, "step": 43680 }, { "grad_norm": 0.10071007162332535, "learning_rate": 6.43634507052335e-05, "loss": 0.0143, "step": 43690 }, { "grad_norm": 0.15283255279064178, "learning_rate": 6.434761216043135e-05, "loss": 0.0143, "step": 43700 }, { "grad_norm": 0.17713001370429993, "learning_rate": 6.433177204659589e-05, "loss": 0.0131, "step": 43710 }, { "grad_norm": 0.09102434664964676, "learning_rate": 6.431593036545938e-05, "loss": 0.0101, "step": 43720 }, { "grad_norm": 0.09071101248264313, "learning_rate": 6.430008711875426e-05, "loss": 0.0092, "step": 43730 }, { "grad_norm": 0.10320965945720673, "learning_rate": 6.42842423082131e-05, "loss": 0.0121, "step": 43740 }, { "grad_norm": 0.11281223595142365, "learning_rate": 6.426839593556868e-05, "loss": 0.0123, "step": 43750 }, { "grad_norm": 0.08729640394449234, "learning_rate": 6.425254800255392e-05, "loss": 0.0107, "step": 43760 }, { "grad_norm": 0.24603316187858582, "learning_rate": 6.423669851090196e-05, "loss": 0.0155, "step": 43770 }, { "grad_norm": 0.1421559751033783, "learning_rate": 6.422084746234604e-05, "loss": 0.0121, "step": 43780 }, { "grad_norm": 0.14361119270324707, "learning_rate": 6.420499485861965e-05, "loss": 0.0102, "step": 43790 }, { "grad_norm": 0.10732542723417282, "learning_rate": 6.418914070145636e-05, "loss": 0.0123, "step": 43800 }, { "grad_norm": 0.12446345388889313, "learning_rate": 6.417328499258996e-05, "loss": 0.0109, "step": 43810 }, { "grad_norm": 0.12439221143722534, "learning_rate": 6.415742773375446e-05, "loss": 0.0095, "step": 43820 }, { "grad_norm": 0.2881907522678375, "learning_rate": 6.414156892668393e-05, "loss": 0.0119, "step": 43830 }, { "grad_norm": 0.12757614254951477, "learning_rate": 6.412570857311267e-05, "loss": 0.0123, "step": 43840 }, { "grad_norm": 0.1257529854774475, "learning_rate": 6.410984667477518e-05, "loss": 0.0128, "step": 43850 }, { "grad_norm": 0.18482838571071625, "learning_rate": 6.409398323340607e-05, "loss": 0.0099, "step": 43860 }, { "grad_norm": 0.13610756397247314, "learning_rate": 6.407811825074012e-05, "loss": 0.0128, "step": 43870 }, { "grad_norm": 0.11524952948093414, "learning_rate": 6.406225172851234e-05, "loss": 0.0123, "step": 43880 }, { "grad_norm": 0.09752479195594788, "learning_rate": 6.404638366845786e-05, "loss": 0.0117, "step": 43890 }, { "grad_norm": 0.16968917846679688, "learning_rate": 6.403051407231196e-05, "loss": 0.0148, "step": 43900 }, { "grad_norm": 0.08278892189264297, "learning_rate": 6.401464294181016e-05, "loss": 0.0126, "step": 43910 }, { "grad_norm": 0.13787131011486053, "learning_rate": 6.399877027868808e-05, "loss": 0.0106, "step": 43920 }, { "grad_norm": 0.20818406343460083, "learning_rate": 6.39828960846815e-05, "loss": 0.0128, "step": 43930 }, { "grad_norm": 0.15963929891586304, "learning_rate": 6.396702036152644e-05, "loss": 0.0116, "step": 43940 }, { "grad_norm": 0.13277150690555573, "learning_rate": 6.395114311095905e-05, "loss": 0.0148, "step": 43950 }, { "grad_norm": 0.12166336178779602, "learning_rate": 6.393526433471559e-05, "loss": 0.0135, "step": 43960 }, { "grad_norm": 0.1568482220172882, "learning_rate": 6.391938403453262e-05, "loss": 0.0165, "step": 43970 }, { "grad_norm": 0.47462883591651917, "learning_rate": 6.390350221214671e-05, "loss": 0.0137, "step": 43980 }, { "grad_norm": 0.1436188519001007, "learning_rate": 6.38876188692947e-05, "loss": 0.0174, "step": 43990 }, { "grad_norm": 0.13573388755321503, "learning_rate": 6.387173400771358e-05, "loss": 0.0137, "step": 44000 }, { "grad_norm": 0.10559132695198059, "learning_rate": 6.385584762914047e-05, "loss": 0.0105, "step": 44010 }, { "grad_norm": 0.10537101328372955, "learning_rate": 6.38399597353127e-05, "loss": 0.0142, "step": 44020 }, { "grad_norm": 0.13379377126693726, "learning_rate": 6.382407032796775e-05, "loss": 0.0109, "step": 44030 }, { "grad_norm": 0.12641023099422455, "learning_rate": 6.380817940884325e-05, "loss": 0.0097, "step": 44040 }, { "grad_norm": 0.19524937868118286, "learning_rate": 6.379228697967702e-05, "loss": 0.013, "step": 44050 }, { "grad_norm": 0.1342102736234665, "learning_rate": 6.3776393042207e-05, "loss": 0.0107, "step": 44060 }, { "grad_norm": 0.129347026348114, "learning_rate": 6.376049759817137e-05, "loss": 0.011, "step": 44070 }, { "grad_norm": 0.11870156973600388, "learning_rate": 6.37446006493084e-05, "loss": 0.0122, "step": 44080 }, { "grad_norm": 0.12332697957754135, "learning_rate": 6.372870219735656e-05, "loss": 0.0151, "step": 44090 }, { "grad_norm": 0.1339881867170334, "learning_rate": 6.371280224405452e-05, "loss": 0.0161, "step": 44100 }, { "grad_norm": 0.18203210830688477, "learning_rate": 6.369690079114101e-05, "loss": 0.0124, "step": 44110 }, { "grad_norm": 0.1471843421459198, "learning_rate": 6.368099784035504e-05, "loss": 0.0115, "step": 44120 }, { "grad_norm": 0.06969362497329712, "learning_rate": 6.366509339343572e-05, "loss": 0.0114, "step": 44130 }, { "grad_norm": 0.10520727187395096, "learning_rate": 6.364918745212232e-05, "loss": 0.0093, "step": 44140 }, { "grad_norm": 0.14294949173927307, "learning_rate": 6.363328001815431e-05, "loss": 0.0126, "step": 44150 }, { "grad_norm": 0.10165908932685852, "learning_rate": 6.361737109327128e-05, "loss": 0.01, "step": 44160 }, { "grad_norm": 0.09197964519262314, "learning_rate": 6.360146067921301e-05, "loss": 0.0102, "step": 44170 }, { "grad_norm": 0.09207798540592194, "learning_rate": 6.35855487777195e-05, "loss": 0.0104, "step": 44180 }, { "grad_norm": 0.15747641026973724, "learning_rate": 6.356963539053076e-05, "loss": 0.0173, "step": 44190 }, { "grad_norm": 0.09848111867904663, "learning_rate": 6.355372051938711e-05, "loss": 0.0123, "step": 44200 }, { "grad_norm": 0.10446561127901077, "learning_rate": 6.353780416602894e-05, "loss": 0.0111, "step": 44210 }, { "grad_norm": 0.09496016800403595, "learning_rate": 6.352188633219689e-05, "loss": 0.0091, "step": 44220 }, { "grad_norm": 0.11704938858747482, "learning_rate": 6.350596701963166e-05, "loss": 0.0106, "step": 44230 }, { "grad_norm": 0.10744224488735199, "learning_rate": 6.349004623007419e-05, "loss": 0.0105, "step": 44240 }, { "grad_norm": 0.17707660794258118, "learning_rate": 6.347412396526555e-05, "loss": 0.0132, "step": 44250 }, { "grad_norm": 0.14466847479343414, "learning_rate": 6.345820022694696e-05, "loss": 0.0149, "step": 44260 }, { "grad_norm": 0.10461173951625824, "learning_rate": 6.344227501685984e-05, "loss": 0.011, "step": 44270 }, { "grad_norm": 0.1455589085817337, "learning_rate": 6.342634833674572e-05, "loss": 0.0139, "step": 44280 }, { "grad_norm": 0.16615153849124908, "learning_rate": 6.341042018834635e-05, "loss": 0.0133, "step": 44290 }, { "grad_norm": 0.11036805063486099, "learning_rate": 6.339449057340359e-05, "loss": 0.0122, "step": 44300 }, { "grad_norm": 0.08342356234788895, "learning_rate": 6.337855949365945e-05, "loss": 0.0101, "step": 44310 }, { "grad_norm": 0.13943332433700562, "learning_rate": 6.336262695085619e-05, "loss": 0.0132, "step": 44320 }, { "grad_norm": 0.16728976368904114, "learning_rate": 6.334669294673612e-05, "loss": 0.0108, "step": 44330 }, { "grad_norm": 0.11784276366233826, "learning_rate": 6.333075748304179e-05, "loss": 0.0147, "step": 44340 }, { "grad_norm": 0.15441004931926727, "learning_rate": 6.331482056151585e-05, "loss": 0.0124, "step": 44350 }, { "grad_norm": 0.08640658855438232, "learning_rate": 6.329888218390117e-05, "loss": 0.0119, "step": 44360 }, { "grad_norm": 0.09472007304430008, "learning_rate": 6.328294235194072e-05, "loss": 0.0093, "step": 44370 }, { "grad_norm": 0.14526759088039398, "learning_rate": 6.326700106737765e-05, "loss": 0.0145, "step": 44380 }, { "grad_norm": 0.12992388010025024, "learning_rate": 6.325105833195531e-05, "loss": 0.0117, "step": 44390 }, { "grad_norm": 0.11524694412946701, "learning_rate": 6.323511414741715e-05, "loss": 0.0141, "step": 44400 }, { "grad_norm": 0.2484978586435318, "learning_rate": 6.321916851550678e-05, "loss": 0.0197, "step": 44410 }, { "grad_norm": 0.17297030985355377, "learning_rate": 6.320322143796806e-05, "loss": 0.014, "step": 44420 }, { "grad_norm": 0.16261853277683258, "learning_rate": 6.318727291654488e-05, "loss": 0.0126, "step": 44430 }, { "grad_norm": 0.15808650851249695, "learning_rate": 6.317132295298134e-05, "loss": 0.0114, "step": 44440 }, { "grad_norm": 0.11736441403627396, "learning_rate": 6.315537154902173e-05, "loss": 0.0111, "step": 44450 }, { "grad_norm": 0.12831196188926697, "learning_rate": 6.313941870641048e-05, "loss": 0.0132, "step": 44460 }, { "grad_norm": 0.12277070432901382, "learning_rate": 6.312346442689214e-05, "loss": 0.0122, "step": 44470 }, { "grad_norm": 0.10665213316679001, "learning_rate": 6.310750871221147e-05, "loss": 0.0115, "step": 44480 }, { "grad_norm": 0.08963076025247574, "learning_rate": 6.309155156411335e-05, "loss": 0.01, "step": 44490 }, { "grad_norm": 0.1801973134279251, "learning_rate": 6.307559298434284e-05, "loss": 0.0117, "step": 44500 }, { "grad_norm": 0.06865935772657394, "learning_rate": 6.305963297464512e-05, "loss": 0.0121, "step": 44510 }, { "grad_norm": 0.08635375648736954, "learning_rate": 6.304367153676561e-05, "loss": 0.0113, "step": 44520 }, { "grad_norm": 0.11524934321641922, "learning_rate": 6.302770867244978e-05, "loss": 0.0113, "step": 44530 }, { "grad_norm": 0.12128973752260208, "learning_rate": 6.301174438344328e-05, "loss": 0.0148, "step": 44540 }, { "grad_norm": 0.1536005288362503, "learning_rate": 6.299577867149202e-05, "loss": 0.0188, "step": 44550 }, { "grad_norm": 0.19032713770866394, "learning_rate": 6.297981153834192e-05, "loss": 0.0122, "step": 44560 }, { "grad_norm": 0.10435830801725388, "learning_rate": 6.296384298573916e-05, "loss": 0.0178, "step": 44570 }, { "grad_norm": 0.14316654205322266, "learning_rate": 6.294787301543001e-05, "loss": 0.0248, "step": 44580 }, { "grad_norm": 0.16296802461147308, "learning_rate": 6.293190162916095e-05, "loss": 0.0115, "step": 44590 }, { "grad_norm": 0.1154433935880661, "learning_rate": 6.291592882867855e-05, "loss": 0.0147, "step": 44600 }, { "grad_norm": 0.24304969608783722, "learning_rate": 6.28999546157296e-05, "loss": 0.0155, "step": 44610 }, { "grad_norm": 0.14802725613117218, "learning_rate": 6.288397899206102e-05, "loss": 0.0117, "step": 44620 }, { "grad_norm": 0.2940356433391571, "learning_rate": 6.286800195941984e-05, "loss": 0.0159, "step": 44630 }, { "grad_norm": 0.1456303894519806, "learning_rate": 6.285202351955334e-05, "loss": 0.0181, "step": 44640 }, { "grad_norm": 0.10804209113121033, "learning_rate": 6.283604367420887e-05, "loss": 0.0135, "step": 44650 }, { "grad_norm": 0.15844516456127167, "learning_rate": 6.282006242513394e-05, "loss": 0.0122, "step": 44660 }, { "grad_norm": 0.17723160982131958, "learning_rate": 6.280407977407628e-05, "loss": 0.0145, "step": 44670 }, { "grad_norm": 0.1862906813621521, "learning_rate": 6.27880957227837e-05, "loss": 0.0158, "step": 44680 }, { "grad_norm": 0.14087077975273132, "learning_rate": 6.27721102730042e-05, "loss": 0.0127, "step": 44690 }, { "grad_norm": 0.19185206294059753, "learning_rate": 6.275612342648592e-05, "loss": 0.0172, "step": 44700 }, { "grad_norm": 0.13485503196716309, "learning_rate": 6.274013518497716e-05, "loss": 0.0123, "step": 44710 }, { "grad_norm": 0.08578348904848099, "learning_rate": 6.272414555022636e-05, "loss": 0.0085, "step": 44720 }, { "grad_norm": 0.20809520781040192, "learning_rate": 6.270815452398215e-05, "loss": 0.011, "step": 44730 }, { "grad_norm": 0.09320586174726486, "learning_rate": 6.269216210799326e-05, "loss": 0.0154, "step": 44740 }, { "grad_norm": 0.16919027268886566, "learning_rate": 6.26761683040086e-05, "loss": 0.0224, "step": 44750 }, { "grad_norm": 0.13220857083797455, "learning_rate": 6.266017311377723e-05, "loss": 0.0149, "step": 44760 }, { "grad_norm": 0.10741320252418518, "learning_rate": 6.264417653904839e-05, "loss": 0.0171, "step": 44770 }, { "grad_norm": 0.12219732254743576, "learning_rate": 6.262817858157139e-05, "loss": 0.0144, "step": 44780 }, { "grad_norm": 0.11445598304271698, "learning_rate": 6.261217924309576e-05, "loss": 0.0196, "step": 44790 }, { "grad_norm": 0.09544189274311066, "learning_rate": 6.259617852537118e-05, "loss": 0.0107, "step": 44800 }, { "grad_norm": 0.22330695390701294, "learning_rate": 6.258017643014747e-05, "loss": 0.0165, "step": 44810 }, { "grad_norm": 0.10502622276544571, "learning_rate": 6.256417295917456e-05, "loss": 0.0181, "step": 44820 }, { "grad_norm": 0.17138168215751648, "learning_rate": 6.254816811420258e-05, "loss": 0.017, "step": 44830 }, { "grad_norm": 0.12653440237045288, "learning_rate": 6.253216189698183e-05, "loss": 0.0124, "step": 44840 }, { "grad_norm": 0.1029638946056366, "learning_rate": 6.251615430926267e-05, "loss": 0.0114, "step": 44850 }, { "grad_norm": 0.12224224209785461, "learning_rate": 6.25001453527957e-05, "loss": 0.0102, "step": 44860 }, { "grad_norm": 0.15805456042289734, "learning_rate": 6.248413502933164e-05, "loss": 0.0147, "step": 44870 }, { "grad_norm": 0.14273601770401, "learning_rate": 6.246812334062133e-05, "loss": 0.0115, "step": 44880 }, { "grad_norm": 0.1946043223142624, "learning_rate": 6.245211028841579e-05, "loss": 0.0132, "step": 44890 }, { "grad_norm": 0.19215653836727142, "learning_rate": 6.24360958744662e-05, "loss": 0.0113, "step": 44900 }, { "grad_norm": 0.09842699766159058, "learning_rate": 6.242008010052387e-05, "loss": 0.0113, "step": 44910 }, { "grad_norm": 0.14533862471580505, "learning_rate": 6.240406296834024e-05, "loss": 0.0176, "step": 44920 }, { "grad_norm": 0.09683352708816528, "learning_rate": 6.238804447966694e-05, "loss": 0.011, "step": 44930 }, { "grad_norm": 0.08401256054639816, "learning_rate": 6.237202463625573e-05, "loss": 0.0116, "step": 44940 }, { "grad_norm": 0.1441829651594162, "learning_rate": 6.235600343985848e-05, "loss": 0.0124, "step": 44950 }, { "grad_norm": 0.08071936666965485, "learning_rate": 6.233998089222729e-05, "loss": 0.0087, "step": 44960 }, { "grad_norm": 0.16866254806518555, "learning_rate": 6.232395699511433e-05, "loss": 0.018, "step": 44970 }, { "grad_norm": 0.14178091287612915, "learning_rate": 6.230793175027199e-05, "loss": 0.0167, "step": 44980 }, { "grad_norm": 0.13468235731124878, "learning_rate": 6.22919051594527e-05, "loss": 0.0113, "step": 44990 }, { "grad_norm": 0.15801680088043213, "learning_rate": 6.227587722440917e-05, "loss": 0.0177, "step": 45000 }, { "grad_norm": 0.1034521833062172, "learning_rate": 6.225984794689414e-05, "loss": 0.0097, "step": 45010 }, { "grad_norm": 0.19654618203639984, "learning_rate": 6.22438173286606e-05, "loss": 0.0158, "step": 45020 }, { "grad_norm": 0.10289980471134186, "learning_rate": 6.222778537146159e-05, "loss": 0.0103, "step": 45030 }, { "grad_norm": 0.11890736222267151, "learning_rate": 6.221175207705037e-05, "loss": 0.0112, "step": 45040 }, { "grad_norm": 0.12661850452423096, "learning_rate": 6.219571744718027e-05, "loss": 0.0097, "step": 45050 }, { "grad_norm": 0.11459051072597504, "learning_rate": 6.217968148360487e-05, "loss": 0.0129, "step": 45060 }, { "grad_norm": 0.13824915885925293, "learning_rate": 6.216364418807782e-05, "loss": 0.0102, "step": 45070 }, { "grad_norm": 0.14939768612384796, "learning_rate": 6.214760556235292e-05, "loss": 0.0105, "step": 45080 }, { "grad_norm": 0.1061270609498024, "learning_rate": 6.213156560818414e-05, "loss": 0.0109, "step": 45090 }, { "grad_norm": 0.08403600007295609, "learning_rate": 6.211552432732559e-05, "loss": 0.012, "step": 45100 }, { "grad_norm": 0.0689878836274147, "learning_rate": 6.20994817215315e-05, "loss": 0.0111, "step": 45110 }, { "grad_norm": 0.21845076978206635, "learning_rate": 6.208343779255628e-05, "loss": 0.0122, "step": 45120 }, { "grad_norm": 0.18154014647006989, "learning_rate": 6.206739254215449e-05, "loss": 0.0133, "step": 45130 }, { "grad_norm": 0.07819167524576187, "learning_rate": 6.205134597208077e-05, "loss": 0.0112, "step": 45140 }, { "grad_norm": 0.13326038420200348, "learning_rate": 6.203529808409e-05, "loss": 0.0104, "step": 45150 }, { "grad_norm": 0.11642120033502579, "learning_rate": 6.201924887993712e-05, "loss": 0.0161, "step": 45160 }, { "grad_norm": 0.13563284277915955, "learning_rate": 6.200319836137725e-05, "loss": 0.0128, "step": 45170 }, { "grad_norm": 0.11979570984840393, "learning_rate": 6.198714653016565e-05, "loss": 0.0129, "step": 45180 }, { "grad_norm": 0.10667748749256134, "learning_rate": 6.197109338805774e-05, "loss": 0.0124, "step": 45190 }, { "grad_norm": 0.09042780101299286, "learning_rate": 6.195503893680903e-05, "loss": 0.0123, "step": 45200 }, { "grad_norm": 0.10052649676799774, "learning_rate": 6.193898317817524e-05, "loss": 0.0089, "step": 45210 }, { "grad_norm": 0.11126342415809631, "learning_rate": 6.192292611391221e-05, "loss": 0.0084, "step": 45220 }, { "grad_norm": 0.11998993903398514, "learning_rate": 6.190686774577591e-05, "loss": 0.0122, "step": 45230 }, { "grad_norm": 0.08621089160442352, "learning_rate": 6.189080807552245e-05, "loss": 0.0148, "step": 45240 }, { "grad_norm": 0.10495129227638245, "learning_rate": 6.187474710490809e-05, "loss": 0.0121, "step": 45250 }, { "grad_norm": 0.10365491360425949, "learning_rate": 6.185868483568926e-05, "loss": 0.0101, "step": 45260 }, { "grad_norm": 0.11768859624862671, "learning_rate": 6.184262126962245e-05, "loss": 0.0104, "step": 45270 }, { "grad_norm": 0.14235015213489532, "learning_rate": 6.182655640846442e-05, "loss": 0.0104, "step": 45280 }, { "grad_norm": 0.10422223061323166, "learning_rate": 6.181049025397196e-05, "loss": 0.0108, "step": 45290 }, { "grad_norm": 0.14345373213291168, "learning_rate": 6.179442280790202e-05, "loss": 0.0192, "step": 45300 }, { "grad_norm": 0.10799402743577957, "learning_rate": 6.177835407201174e-05, "loss": 0.0146, "step": 45310 }, { "grad_norm": 0.10993222147226334, "learning_rate": 6.176228404805839e-05, "loss": 0.0113, "step": 45320 }, { "grad_norm": 0.1622518002986908, "learning_rate": 6.174621273779932e-05, "loss": 0.0115, "step": 45330 }, { "grad_norm": 0.12905775010585785, "learning_rate": 6.17301401429921e-05, "loss": 0.0129, "step": 45340 }, { "grad_norm": 0.13429316878318787, "learning_rate": 6.171406626539436e-05, "loss": 0.0114, "step": 45350 }, { "grad_norm": 0.23405632376670837, "learning_rate": 6.169799110676398e-05, "loss": 0.0175, "step": 45360 }, { "grad_norm": 0.14333310723304749, "learning_rate": 6.168191466885885e-05, "loss": 0.0106, "step": 45370 }, { "grad_norm": 0.14448651671409607, "learning_rate": 6.166583695343711e-05, "loss": 0.0096, "step": 45380 }, { "grad_norm": 0.11299438774585724, "learning_rate": 6.164975796225698e-05, "loss": 0.0123, "step": 45390 }, { "grad_norm": 0.07167326658964157, "learning_rate": 6.163367769707683e-05, "loss": 0.01, "step": 45400 }, { "grad_norm": 0.09968110173940659, "learning_rate": 6.161759615965519e-05, "loss": 0.0116, "step": 45410 }, { "grad_norm": 0.12293767184019089, "learning_rate": 6.16015133517507e-05, "loss": 0.0085, "step": 45420 }, { "grad_norm": 0.09722016751766205, "learning_rate": 6.158542927512214e-05, "loss": 0.0109, "step": 45430 }, { "grad_norm": 0.11268937587738037, "learning_rate": 6.156934393152846e-05, "loss": 0.0109, "step": 45440 }, { "grad_norm": 0.1215190514922142, "learning_rate": 6.15532573227287e-05, "loss": 0.0154, "step": 45450 }, { "grad_norm": 0.18162080645561218, "learning_rate": 6.153716945048212e-05, "loss": 0.0172, "step": 45460 }, { "grad_norm": 0.16719532012939453, "learning_rate": 6.152108031654802e-05, "loss": 0.0142, "step": 45470 }, { "grad_norm": 0.09039901942014694, "learning_rate": 6.15049899226859e-05, "loss": 0.0126, "step": 45480 }, { "grad_norm": 0.101052425801754, "learning_rate": 6.148889827065537e-05, "loss": 0.0104, "step": 45490 }, { "grad_norm": 0.19990494847297668, "learning_rate": 6.147280536221622e-05, "loss": 0.0112, "step": 45500 }, { "grad_norm": 0.14928719401359558, "learning_rate": 6.145671119912832e-05, "loss": 0.0106, "step": 45510 }, { "grad_norm": 0.1318427473306656, "learning_rate": 6.144061578315169e-05, "loss": 0.0104, "step": 45520 }, { "grad_norm": 0.0955086499452591, "learning_rate": 6.142451911604654e-05, "loss": 0.0084, "step": 45530 }, { "grad_norm": 0.09549441933631897, "learning_rate": 6.140842119957315e-05, "loss": 0.0094, "step": 45540 }, { "grad_norm": 0.0901409313082695, "learning_rate": 6.139232203549197e-05, "loss": 0.0126, "step": 45550 }, { "grad_norm": 0.0996464192867279, "learning_rate": 6.137622162556357e-05, "loss": 0.0118, "step": 45560 }, { "grad_norm": 0.0766233429312706, "learning_rate": 6.13601199715487e-05, "loss": 0.0096, "step": 45570 }, { "grad_norm": 0.24264563620090485, "learning_rate": 6.134401707520816e-05, "loss": 0.0107, "step": 45580 }, { "grad_norm": 0.22010372579097748, "learning_rate": 6.132791293830298e-05, "loss": 0.0079, "step": 45590 }, { "grad_norm": 0.08758355677127838, "learning_rate": 6.131180756259428e-05, "loss": 0.0101, "step": 45600 }, { "grad_norm": 0.09098420292139053, "learning_rate": 6.129570094984331e-05, "loss": 0.0103, "step": 45610 }, { "grad_norm": 0.15559807419776917, "learning_rate": 6.127959310181145e-05, "loss": 0.0148, "step": 45620 }, { "grad_norm": 0.16618578135967255, "learning_rate": 6.126348402026026e-05, "loss": 0.0115, "step": 45630 }, { "grad_norm": 0.13144472241401672, "learning_rate": 6.12473737069514e-05, "loss": 0.0118, "step": 45640 }, { "grad_norm": 0.09743750095367432, "learning_rate": 6.123126216364665e-05, "loss": 0.0112, "step": 45650 }, { "grad_norm": 0.11809732019901276, "learning_rate": 6.121514939210797e-05, "loss": 0.0108, "step": 45660 }, { "grad_norm": 0.19863326847553253, "learning_rate": 6.119903539409741e-05, "loss": 0.0125, "step": 45670 }, { "grad_norm": 0.11397455632686615, "learning_rate": 6.118292017137716e-05, "loss": 0.0154, "step": 45680 }, { "grad_norm": 0.17537975311279297, "learning_rate": 6.116680372570959e-05, "loss": 0.0112, "step": 45690 }, { "grad_norm": 0.08142314851284027, "learning_rate": 6.115068605885713e-05, "loss": 0.0101, "step": 45700 }, { "grad_norm": 0.09998941421508789, "learning_rate": 6.113456717258243e-05, "loss": 0.0121, "step": 45710 }, { "grad_norm": 0.11492525041103363, "learning_rate": 6.11184470686482e-05, "loss": 0.013, "step": 45720 }, { "grad_norm": 0.10754130780696869, "learning_rate": 6.110232574881731e-05, "loss": 0.0097, "step": 45730 }, { "grad_norm": 0.0905071571469307, "learning_rate": 6.108620321485277e-05, "loss": 0.0105, "step": 45740 }, { "grad_norm": 0.13608673214912415, "learning_rate": 6.107007946851773e-05, "loss": 0.013, "step": 45750 }, { "grad_norm": 0.10666771978139877, "learning_rate": 6.105395451157542e-05, "loss": 0.013, "step": 45760 }, { "grad_norm": 0.11677923798561096, "learning_rate": 6.103782834578928e-05, "loss": 0.0098, "step": 45770 }, { "grad_norm": 0.09412262588739395, "learning_rate": 6.102170097292281e-05, "loss": 0.0129, "step": 45780 }, { "grad_norm": 0.1880759298801422, "learning_rate": 6.1005572394739716e-05, "loss": 0.0147, "step": 45790 }, { "grad_norm": 0.21839337050914764, "learning_rate": 6.0989442613003765e-05, "loss": 0.0109, "step": 45800 }, { "grad_norm": 0.21319524943828583, "learning_rate": 6.097331162947888e-05, "loss": 0.0142, "step": 45810 }, { "grad_norm": 0.13207384943962097, "learning_rate": 6.095717944592914e-05, "loss": 0.011, "step": 45820 }, { "grad_norm": 0.10505381971597672, "learning_rate": 6.094104606411873e-05, "loss": 0.0182, "step": 45830 }, { "grad_norm": 0.1463295966386795, "learning_rate": 6.0924911485811966e-05, "loss": 0.0197, "step": 45840 }, { "grad_norm": 0.09575953334569931, "learning_rate": 6.090877571277331e-05, "loss": 0.017, "step": 45850 }, { "grad_norm": 0.10360788553953171, "learning_rate": 6.089263874676736e-05, "loss": 0.0122, "step": 45860 }, { "grad_norm": 0.16749906539916992, "learning_rate": 6.0876500589558796e-05, "loss": 0.0131, "step": 45870 }, { "grad_norm": 0.10618530213832855, "learning_rate": 6.086036124291248e-05, "loss": 0.0113, "step": 45880 }, { "grad_norm": 0.17377254366874695, "learning_rate": 6.084422070859339e-05, "loss": 0.0212, "step": 45890 }, { "grad_norm": 0.13434350490570068, "learning_rate": 6.082807898836663e-05, "loss": 0.012, "step": 45900 }, { "grad_norm": 0.09214203804731369, "learning_rate": 6.081193608399742e-05, "loss": 0.01, "step": 45910 }, { "grad_norm": 0.0840618684887886, "learning_rate": 6.0795791997251164e-05, "loss": 0.0104, "step": 45920 }, { "grad_norm": 0.1372019499540329, "learning_rate": 6.0779646729893294e-05, "loss": 0.0093, "step": 45930 }, { "grad_norm": 0.07999851554632187, "learning_rate": 6.0763500283689476e-05, "loss": 0.0082, "step": 45940 }, { "grad_norm": 0.13140340149402618, "learning_rate": 6.0747352660405455e-05, "loss": 0.0122, "step": 45950 }, { "grad_norm": 0.0805596187710762, "learning_rate": 6.073120386180709e-05, "loss": 0.0088, "step": 45960 }, { "grad_norm": 0.07328525185585022, "learning_rate": 6.0715053889660425e-05, "loss": 0.0099, "step": 45970 }, { "grad_norm": 0.1037088930606842, "learning_rate": 6.069890274573157e-05, "loss": 0.0109, "step": 45980 }, { "grad_norm": 0.14544595777988434, "learning_rate": 6.068275043178679e-05, "loss": 0.02, "step": 45990 }, { "grad_norm": 0.08839821815490723, "learning_rate": 6.066659694959248e-05, "loss": 0.0132, "step": 46000 }, { "grad_norm": 0.1113436371088028, "learning_rate": 6.0650442300915176e-05, "loss": 0.0116, "step": 46010 }, { "grad_norm": 0.11632556468248367, "learning_rate": 6.063428648752152e-05, "loss": 0.0101, "step": 46020 }, { "grad_norm": 0.1689666211605072, "learning_rate": 6.0618129511178266e-05, "loss": 0.0134, "step": 46030 }, { "grad_norm": 0.16396306455135345, "learning_rate": 6.060197137365234e-05, "loss": 0.0118, "step": 46040 }, { "grad_norm": 0.18985264003276825, "learning_rate": 6.058581207671077e-05, "loss": 0.0096, "step": 46050 }, { "grad_norm": 0.1351679563522339, "learning_rate": 6.056965162212072e-05, "loss": 0.0117, "step": 46060 }, { "grad_norm": 0.08857684582471848, "learning_rate": 6.055349001164943e-05, "loss": 0.0179, "step": 46070 }, { "grad_norm": 0.11023498326539993, "learning_rate": 6.0537327247064347e-05, "loss": 0.0133, "step": 46080 }, { "grad_norm": 0.3701610267162323, "learning_rate": 6.0521163330133014e-05, "loss": 0.0177, "step": 46090 }, { "grad_norm": 0.1830539107322693, "learning_rate": 6.050499826262306e-05, "loss": 0.0129, "step": 46100 }, { "grad_norm": 0.15451428294181824, "learning_rate": 6.0488832046302294e-05, "loss": 0.0169, "step": 46110 }, { "grad_norm": 0.13935165107250214, "learning_rate": 6.0472664682938626e-05, "loss": 0.0113, "step": 46120 }, { "grad_norm": 0.13484686613082886, "learning_rate": 6.045649617430009e-05, "loss": 0.0118, "step": 46130 }, { "grad_norm": 0.10805448889732361, "learning_rate": 6.0440326522154866e-05, "loss": 0.0109, "step": 46140 }, { "grad_norm": 0.13655824959278107, "learning_rate": 6.0424155728271224e-05, "loss": 0.0104, "step": 46150 }, { "grad_norm": 0.15744484961032867, "learning_rate": 6.040798379441758e-05, "loss": 0.0106, "step": 46160 }, { "grad_norm": 0.1589779108762741, "learning_rate": 6.0391810722362485e-05, "loss": 0.0127, "step": 46170 }, { "grad_norm": 0.17168602347373962, "learning_rate": 6.037563651387458e-05, "loss": 0.0129, "step": 46180 }, { "grad_norm": 0.1592707484960556, "learning_rate": 6.0359461170722666e-05, "loss": 0.0107, "step": 46190 }, { "grad_norm": 0.1399434357881546, "learning_rate": 6.034328469467566e-05, "loss": 0.0142, "step": 46200 }, { "grad_norm": 0.097935751080513, "learning_rate": 6.0327107087502596e-05, "loss": 0.0138, "step": 46210 }, { "grad_norm": 0.10695017874240875, "learning_rate": 6.031092835097262e-05, "loss": 0.0103, "step": 46220 }, { "grad_norm": 0.1399047076702118, "learning_rate": 6.0294748486855024e-05, "loss": 0.0117, "step": 46230 }, { "grad_norm": 0.09299742430448532, "learning_rate": 6.0278567496919216e-05, "loss": 0.0133, "step": 46240 }, { "grad_norm": 0.12445055693387985, "learning_rate": 6.026238538293472e-05, "loss": 0.0112, "step": 46250 }, { "grad_norm": 0.12016655504703522, "learning_rate": 6.024620214667118e-05, "loss": 0.0149, "step": 46260 }, { "grad_norm": 0.11246414482593536, "learning_rate": 6.0230017789898384e-05, "loss": 0.0092, "step": 46270 }, { "grad_norm": 0.1765104979276657, "learning_rate": 6.021383231438622e-05, "loss": 0.0105, "step": 46280 }, { "grad_norm": 0.17303520441055298, "learning_rate": 6.0197645721904704e-05, "loss": 0.0094, "step": 46290 }, { "grad_norm": 0.08014937490224838, "learning_rate": 6.0181458014224e-05, "loss": 0.0092, "step": 46300 }, { "grad_norm": 0.12281699478626251, "learning_rate": 6.016526919311435e-05, "loss": 0.0113, "step": 46310 }, { "grad_norm": 0.1273038536310196, "learning_rate": 6.014907926034613e-05, "loss": 0.0102, "step": 46320 }, { "grad_norm": 0.08924610167741776, "learning_rate": 6.0132888217689875e-05, "loss": 0.0105, "step": 46330 }, { "grad_norm": 0.19030213356018066, "learning_rate": 6.01166960669162e-05, "loss": 0.0164, "step": 46340 }, { "grad_norm": 0.0981844887137413, "learning_rate": 6.0100502809795845e-05, "loss": 0.0141, "step": 46350 }, { "grad_norm": 0.10283425450325012, "learning_rate": 6.00843084480997e-05, "loss": 0.0124, "step": 46360 }, { "grad_norm": 0.11184947937726974, "learning_rate": 6.0068112983598736e-05, "loss": 0.0114, "step": 46370 }, { "grad_norm": 0.11374538391828537, "learning_rate": 6.005191641806407e-05, "loss": 0.0126, "step": 46380 }, { "grad_norm": 0.17150582373142242, "learning_rate": 6.003571875326694e-05, "loss": 0.0132, "step": 46390 }, { "grad_norm": 0.11089678853750229, "learning_rate": 6.00195199909787e-05, "loss": 0.0103, "step": 46400 }, { "grad_norm": 0.14165282249450684, "learning_rate": 6.000332013297082e-05, "loss": 0.0114, "step": 46410 }, { "grad_norm": 0.12465088814496994, "learning_rate": 5.998711918101487e-05, "loss": 0.0117, "step": 46420 }, { "grad_norm": 0.07178284972906113, "learning_rate": 5.997091713688261e-05, "loss": 0.0094, "step": 46430 }, { "grad_norm": 0.09925208985805511, "learning_rate": 5.9954714002345836e-05, "loss": 0.0111, "step": 46440 }, { "grad_norm": 0.13604635000228882, "learning_rate": 5.993850977917649e-05, "loss": 0.0111, "step": 46450 }, { "grad_norm": 0.1694980263710022, "learning_rate": 5.992230446914667e-05, "loss": 0.0109, "step": 46460 }, { "grad_norm": 0.13252994418144226, "learning_rate": 5.990609807402855e-05, "loss": 0.0138, "step": 46470 }, { "grad_norm": 0.09372962266206741, "learning_rate": 5.988989059559443e-05, "loss": 0.0143, "step": 46480 }, { "grad_norm": 0.09128382802009583, "learning_rate": 5.987368203561675e-05, "loss": 0.0105, "step": 46490 }, { "grad_norm": 0.11532699316740036, "learning_rate": 5.9857472395868055e-05, "loss": 0.01, "step": 46500 }, { "grad_norm": 0.15164130926132202, "learning_rate": 5.9841261678120983e-05, "loss": 0.0149, "step": 46510 }, { "grad_norm": 0.08295407891273499, "learning_rate": 5.982504988414834e-05, "loss": 0.0094, "step": 46520 }, { "grad_norm": 0.1476958990097046, "learning_rate": 5.9808837015723015e-05, "loss": 0.0232, "step": 46530 }, { "grad_norm": 0.12649360299110413, "learning_rate": 5.9792623074618016e-05, "loss": 0.0121, "step": 46540 }, { "grad_norm": 0.13035517930984497, "learning_rate": 5.977640806260648e-05, "loss": 0.0099, "step": 46550 }, { "grad_norm": 0.09718529134988785, "learning_rate": 5.976019198146166e-05, "loss": 0.0077, "step": 46560 }, { "grad_norm": 0.09173562377691269, "learning_rate": 5.974397483295692e-05, "loss": 0.0107, "step": 46570 }, { "grad_norm": 0.09834695607423782, "learning_rate": 5.972775661886575e-05, "loss": 0.0134, "step": 46580 }, { "grad_norm": 0.07875876873731613, "learning_rate": 5.971153734096173e-05, "loss": 0.0089, "step": 46590 }, { "grad_norm": 0.0812503844499588, "learning_rate": 5.969531700101859e-05, "loss": 0.0104, "step": 46600 }, { "grad_norm": 0.11886310577392578, "learning_rate": 5.9679095600810155e-05, "loss": 0.0136, "step": 46610 }, { "grad_norm": 0.15436571836471558, "learning_rate": 5.9662873142110384e-05, "loss": 0.0103, "step": 46620 }, { "grad_norm": 0.1058235615491867, "learning_rate": 5.964664962669333e-05, "loss": 0.0092, "step": 46630 }, { "grad_norm": 0.1230865865945816, "learning_rate": 5.9630425056333186e-05, "loss": 0.0092, "step": 46640 }, { "grad_norm": 0.11671585589647293, "learning_rate": 5.961419943280422e-05, "loss": 0.0122, "step": 46650 }, { "grad_norm": 0.13867931067943573, "learning_rate": 5.959797275788087e-05, "loss": 0.0112, "step": 46660 }, { "grad_norm": 0.11842361837625504, "learning_rate": 5.958174503333765e-05, "loss": 0.0115, "step": 46670 }, { "grad_norm": 0.11207402497529984, "learning_rate": 5.9565516260949195e-05, "loss": 0.0144, "step": 46680 }, { "grad_norm": 0.1142866462469101, "learning_rate": 5.9549286442490273e-05, "loss": 0.0101, "step": 46690 }, { "grad_norm": 0.10525680333375931, "learning_rate": 5.953305557973572e-05, "loss": 0.0112, "step": 46700 }, { "grad_norm": 0.1800650954246521, "learning_rate": 5.9516823674460564e-05, "loss": 0.0123, "step": 46710 }, { "grad_norm": 0.10451540350914001, "learning_rate": 5.9500590728439875e-05, "loss": 0.0083, "step": 46720 }, { "grad_norm": 0.09773506224155426, "learning_rate": 5.948435674344886e-05, "loss": 0.0121, "step": 46730 }, { "grad_norm": 0.10749746114015579, "learning_rate": 5.946812172126285e-05, "loss": 0.0083, "step": 46740 }, { "grad_norm": 0.10861977189779282, "learning_rate": 5.945188566365729e-05, "loss": 0.0121, "step": 46750 }, { "grad_norm": 0.10584865510463715, "learning_rate": 5.943564857240773e-05, "loss": 0.0138, "step": 46760 }, { "grad_norm": 0.08013332635164261, "learning_rate": 5.941941044928981e-05, "loss": 0.0133, "step": 46770 }, { "grad_norm": 0.5320270657539368, "learning_rate": 5.940317129607935e-05, "loss": 0.0123, "step": 46780 }, { "grad_norm": 0.09876038134098053, "learning_rate": 5.9386931114552204e-05, "loss": 0.0092, "step": 46790 }, { "grad_norm": 0.10077053308486938, "learning_rate": 5.937068990648438e-05, "loss": 0.0122, "step": 46800 }, { "grad_norm": 0.16691182553768158, "learning_rate": 5.935444767365199e-05, "loss": 0.0221, "step": 46810 }, { "grad_norm": 0.12923561036586761, "learning_rate": 5.933820441783129e-05, "loss": 0.014, "step": 46820 }, { "grad_norm": 0.12315630912780762, "learning_rate": 5.932196014079857e-05, "loss": 0.0116, "step": 46830 }, { "grad_norm": 0.10512740164995193, "learning_rate": 5.930571484433032e-05, "loss": 0.0087, "step": 46840 }, { "grad_norm": 0.07888862490653992, "learning_rate": 5.928946853020309e-05, "loss": 0.0095, "step": 46850 }, { "grad_norm": 0.0811837911605835, "learning_rate": 5.927322120019352e-05, "loss": 0.0093, "step": 46860 }, { "grad_norm": 0.0941988080739975, "learning_rate": 5.925697285607844e-05, "loss": 0.0114, "step": 46870 }, { "grad_norm": 0.07216975092887878, "learning_rate": 5.924072349963472e-05, "loss": 0.0093, "step": 46880 }, { "grad_norm": 0.0761168971657753, "learning_rate": 5.922447313263938e-05, "loss": 0.0099, "step": 46890 }, { "grad_norm": 0.07568488270044327, "learning_rate": 5.920822175686952e-05, "loss": 0.0089, "step": 46900 }, { "grad_norm": 0.07736991345882416, "learning_rate": 5.919196937410239e-05, "loss": 0.0125, "step": 46910 }, { "grad_norm": 0.17780587077140808, "learning_rate": 5.917571598611529e-05, "loss": 0.0115, "step": 46920 }, { "grad_norm": 0.11031182110309601, "learning_rate": 5.9159461594685686e-05, "loss": 0.0097, "step": 46930 }, { "grad_norm": 0.1468464732170105, "learning_rate": 5.914320620159114e-05, "loss": 0.0134, "step": 46940 }, { "grad_norm": 0.12700040638446808, "learning_rate": 5.912694980860931e-05, "loss": 0.01, "step": 46950 }, { "grad_norm": 0.17152182757854462, "learning_rate": 5.9110692417517964e-05, "loss": 0.0088, "step": 46960 }, { "grad_norm": 0.178956538438797, "learning_rate": 5.9094434030095e-05, "loss": 0.0127, "step": 46970 }, { "grad_norm": 0.08981821686029434, "learning_rate": 5.907817464811841e-05, "loss": 0.0096, "step": 46980 }, { "grad_norm": 0.1265813559293747, "learning_rate": 5.906191427336627e-05, "loss": 0.0094, "step": 46990 }, { "grad_norm": 0.12483218312263489, "learning_rate": 5.9045652907616835e-05, "loss": 0.0117, "step": 47000 }, { "grad_norm": 0.15692749619483948, "learning_rate": 5.902939055264838e-05, "loss": 0.0128, "step": 47010 }, { "grad_norm": 0.12080808728933334, "learning_rate": 5.901312721023935e-05, "loss": 0.0116, "step": 47020 }, { "grad_norm": 0.11111669987440109, "learning_rate": 5.8996862882168294e-05, "loss": 0.0098, "step": 47030 }, { "grad_norm": 0.11881823092699051, "learning_rate": 5.8980597570213826e-05, "loss": 0.0134, "step": 47040 }, { "grad_norm": 0.12288931757211685, "learning_rate": 5.896433127615471e-05, "loss": 0.0085, "step": 47050 }, { "grad_norm": 0.07229039818048477, "learning_rate": 5.894806400176981e-05, "loss": 0.008, "step": 47060 }, { "grad_norm": 0.07200267165899277, "learning_rate": 5.893179574883808e-05, "loss": 0.0104, "step": 47070 }, { "grad_norm": 0.15615135431289673, "learning_rate": 5.8915526519138585e-05, "loss": 0.0104, "step": 47080 }, { "grad_norm": 0.14916452765464783, "learning_rate": 5.889925631445053e-05, "loss": 0.0099, "step": 47090 }, { "grad_norm": 0.11445271223783493, "learning_rate": 5.888298513655318e-05, "loss": 0.012, "step": 47100 }, { "grad_norm": 0.09510726481676102, "learning_rate": 5.886671298722592e-05, "loss": 0.0085, "step": 47110 }, { "grad_norm": 0.10183085501194, "learning_rate": 5.885043986824828e-05, "loss": 0.0088, "step": 47120 }, { "grad_norm": 0.09991710633039474, "learning_rate": 5.883416578139982e-05, "loss": 0.0104, "step": 47130 }, { "grad_norm": 0.09949153661727905, "learning_rate": 5.881789072846029e-05, "loss": 0.014, "step": 47140 }, { "grad_norm": 0.18872348964214325, "learning_rate": 5.8801614711209474e-05, "loss": 0.0099, "step": 47150 }, { "grad_norm": 0.14450085163116455, "learning_rate": 5.878533773142732e-05, "loss": 0.0134, "step": 47160 }, { "grad_norm": 0.17787982523441315, "learning_rate": 5.8769059790893834e-05, "loss": 0.0129, "step": 47170 }, { "grad_norm": 0.1215575560927391, "learning_rate": 5.8752780891389145e-05, "loss": 0.0127, "step": 47180 }, { "grad_norm": 0.11011651903390884, "learning_rate": 5.873650103469351e-05, "loss": 0.0098, "step": 47190 }, { "grad_norm": 0.11689971387386322, "learning_rate": 5.872022022258725e-05, "loss": 0.0086, "step": 47200 }, { "grad_norm": 0.07156598567962646, "learning_rate": 5.8703938456850814e-05, "loss": 0.0074, "step": 47210 }, { "grad_norm": 0.07223350554704666, "learning_rate": 5.868765573926477e-05, "loss": 0.0099, "step": 47220 }, { "grad_norm": 0.22900882363319397, "learning_rate": 5.867137207160974e-05, "loss": 0.0121, "step": 47230 }, { "grad_norm": 0.12671534717082977, "learning_rate": 5.86550874556665e-05, "loss": 0.0105, "step": 47240 }, { "grad_norm": 0.12627530097961426, "learning_rate": 5.863880189321592e-05, "loss": 0.0112, "step": 47250 }, { "grad_norm": 0.23240187764167786, "learning_rate": 5.862251538603894e-05, "loss": 0.0171, "step": 47260 }, { "grad_norm": 0.11487603187561035, "learning_rate": 5.8606227935916656e-05, "loss": 0.0109, "step": 47270 }, { "grad_norm": 0.08688101917505264, "learning_rate": 5.858993954463021e-05, "loss": 0.0107, "step": 47280 }, { "grad_norm": 0.17939147353172302, "learning_rate": 5.8573650213960896e-05, "loss": 0.0139, "step": 47290 }, { "grad_norm": 0.1408926546573639, "learning_rate": 5.8557359945690084e-05, "loss": 0.0108, "step": 47300 }, { "grad_norm": 0.11893472820520401, "learning_rate": 5.8541068741599246e-05, "loss": 0.0121, "step": 47310 }, { "grad_norm": 0.1454412043094635, "learning_rate": 5.8524776603469985e-05, "loss": 0.0105, "step": 47320 }, { "grad_norm": 0.10623026639223099, "learning_rate": 5.850848353308397e-05, "loss": 0.0085, "step": 47330 }, { "grad_norm": 0.09492337703704834, "learning_rate": 5.849218953222297e-05, "loss": 0.0089, "step": 47340 }, { "grad_norm": 0.11155088245868683, "learning_rate": 5.847589460266891e-05, "loss": 0.0104, "step": 47350 }, { "grad_norm": 0.13208875060081482, "learning_rate": 5.845959874620375e-05, "loss": 0.0116, "step": 47360 }, { "grad_norm": 0.11071866750717163, "learning_rate": 5.84433019646096e-05, "loss": 0.0195, "step": 47370 }, { "grad_norm": 0.15013568103313446, "learning_rate": 5.842700425966863e-05, "loss": 0.011, "step": 47380 }, { "grad_norm": 0.12864983081817627, "learning_rate": 5.841070563316315e-05, "loss": 0.0098, "step": 47390 }, { "grad_norm": 0.08901844918727875, "learning_rate": 5.839440608687554e-05, "loss": 0.0126, "step": 47400 }, { "grad_norm": 0.11324942857027054, "learning_rate": 5.837810562258831e-05, "loss": 0.011, "step": 47410 }, { "grad_norm": 0.1504608392715454, "learning_rate": 5.836180424208405e-05, "loss": 0.0152, "step": 47420 }, { "grad_norm": 0.1095733642578125, "learning_rate": 5.834550194714543e-05, "loss": 0.0118, "step": 47430 }, { "grad_norm": 0.1296623796224594, "learning_rate": 5.832919873955528e-05, "loss": 0.0165, "step": 47440 }, { "grad_norm": 0.10108216851949692, "learning_rate": 5.831289462109648e-05, "loss": 0.0107, "step": 47450 }, { "grad_norm": 0.0966997817158699, "learning_rate": 5.829658959355202e-05, "loss": 0.0191, "step": 47460 }, { "grad_norm": 0.23616637289524078, "learning_rate": 5.8280283658704995e-05, "loss": 0.0154, "step": 47470 }, { "grad_norm": 0.12753356993198395, "learning_rate": 5.826397681833859e-05, "loss": 0.0123, "step": 47480 }, { "grad_norm": 0.11688285320997238, "learning_rate": 5.824766907423612e-05, "loss": 0.0083, "step": 47490 }, { "grad_norm": 0.1254238337278366, "learning_rate": 5.8231360428180945e-05, "loss": 0.0123, "step": 47500 }, { "grad_norm": 0.10139849036931992, "learning_rate": 5.821505088195658e-05, "loss": 0.0097, "step": 47510 }, { "grad_norm": 0.18090589344501495, "learning_rate": 5.819874043734661e-05, "loss": 0.0121, "step": 47520 }, { "grad_norm": 0.1232772022485733, "learning_rate": 5.8182429096134695e-05, "loss": 0.0094, "step": 47530 }, { "grad_norm": 0.12020006030797958, "learning_rate": 5.816611686010465e-05, "loss": 0.0097, "step": 47540 }, { "grad_norm": 0.1080104187130928, "learning_rate": 5.814980373104033e-05, "loss": 0.0137, "step": 47550 }, { "grad_norm": 0.16992419958114624, "learning_rate": 5.813348971072572e-05, "loss": 0.0185, "step": 47560 }, { "grad_norm": 0.10379303246736526, "learning_rate": 5.811717480094492e-05, "loss": 0.0106, "step": 47570 }, { "grad_norm": 0.1293824315071106, "learning_rate": 5.810085900348209e-05, "loss": 0.0139, "step": 47580 }, { "grad_norm": 0.1858876645565033, "learning_rate": 5.8084542320121483e-05, "loss": 0.01, "step": 47590 }, { "grad_norm": 0.14827820658683777, "learning_rate": 5.8068224752647497e-05, "loss": 0.0253, "step": 47600 }, { "grad_norm": 0.10770472884178162, "learning_rate": 5.805190630284457e-05, "loss": 0.0098, "step": 47610 }, { "grad_norm": 0.11336692422628403, "learning_rate": 5.803558697249729e-05, "loss": 0.0118, "step": 47620 }, { "grad_norm": 0.1179216206073761, "learning_rate": 5.8019266763390276e-05, "loss": 0.0167, "step": 47630 }, { "grad_norm": 0.08382441848516464, "learning_rate": 5.800294567730833e-05, "loss": 0.0094, "step": 47640 }, { "grad_norm": 0.08840499073266983, "learning_rate": 5.798662371603628e-05, "loss": 0.0106, "step": 47650 }, { "grad_norm": 0.07966172695159912, "learning_rate": 5.797030088135904e-05, "loss": 0.0089, "step": 47660 }, { "grad_norm": 0.07479289919137955, "learning_rate": 5.7953977175061705e-05, "loss": 0.0084, "step": 47670 }, { "grad_norm": 0.12215878814458847, "learning_rate": 5.793765259892938e-05, "loss": 0.0138, "step": 47680 }, { "grad_norm": 0.10619819909334183, "learning_rate": 5.792132715474729e-05, "loss": 0.0093, "step": 47690 }, { "grad_norm": 0.09635020792484283, "learning_rate": 5.790500084430078e-05, "loss": 0.0093, "step": 47700 }, { "grad_norm": 0.12835609912872314, "learning_rate": 5.7888673669375274e-05, "loss": 0.0107, "step": 47710 }, { "grad_norm": 0.08447007089853287, "learning_rate": 5.787234563175625e-05, "loss": 0.0109, "step": 47720 }, { "grad_norm": 0.15397629141807556, "learning_rate": 5.7856016733229355e-05, "loss": 0.0116, "step": 47730 }, { "grad_norm": 0.10759416222572327, "learning_rate": 5.7839686975580297e-05, "loss": 0.0147, "step": 47740 }, { "grad_norm": 0.07919338345527649, "learning_rate": 5.782335636059484e-05, "loss": 0.0108, "step": 47750 }, { "grad_norm": 0.1621650904417038, "learning_rate": 5.780702489005889e-05, "loss": 0.0113, "step": 47760 }, { "grad_norm": 0.13331536948680878, "learning_rate": 5.779069256575845e-05, "loss": 0.0105, "step": 47770 }, { "grad_norm": 0.1520354300737381, "learning_rate": 5.7774359389479574e-05, "loss": 0.0163, "step": 47780 }, { "grad_norm": 0.14857399463653564, "learning_rate": 5.775802536300845e-05, "loss": 0.0103, "step": 47790 }, { "grad_norm": 0.1048167422413826, "learning_rate": 5.774169048813134e-05, "loss": 0.0097, "step": 47800 }, { "grad_norm": 0.10600671917200089, "learning_rate": 5.77253547666346e-05, "loss": 0.0095, "step": 47810 }, { "grad_norm": 0.19475407898426056, "learning_rate": 5.770901820030465e-05, "loss": 0.0155, "step": 47820 }, { "grad_norm": 0.14968034625053406, "learning_rate": 5.769268079092809e-05, "loss": 0.0127, "step": 47830 }, { "grad_norm": 0.08255764096975327, "learning_rate": 5.767634254029151e-05, "loss": 0.0093, "step": 47840 }, { "grad_norm": 0.23737384378910065, "learning_rate": 5.7660003450181655e-05, "loss": 0.0248, "step": 47850 }, { "grad_norm": 0.09775938838720322, "learning_rate": 5.764366352238534e-05, "loss": 0.0101, "step": 47860 }, { "grad_norm": 0.10132541507482529, "learning_rate": 5.7627322758689474e-05, "loss": 0.0089, "step": 47870 }, { "grad_norm": 0.13612431287765503, "learning_rate": 5.761098116088105e-05, "loss": 0.0105, "step": 47880 }, { "grad_norm": 0.1092410609126091, "learning_rate": 5.759463873074717e-05, "loss": 0.0114, "step": 47890 }, { "grad_norm": 0.12451212853193283, "learning_rate": 5.757829547007504e-05, "loss": 0.0113, "step": 47900 }, { "grad_norm": 0.09953074902296066, "learning_rate": 5.756195138065189e-05, "loss": 0.0095, "step": 47910 }, { "grad_norm": 0.18636390566825867, "learning_rate": 5.754560646426511e-05, "loss": 0.0135, "step": 47920 }, { "grad_norm": 0.1386461853981018, "learning_rate": 5.752926072270216e-05, "loss": 0.0107, "step": 47930 }, { "grad_norm": 0.14125463366508484, "learning_rate": 5.7512914157750563e-05, "loss": 0.0099, "step": 47940 }, { "grad_norm": 0.10820189118385315, "learning_rate": 5.749656677119798e-05, "loss": 0.0095, "step": 47950 }, { "grad_norm": 0.10925164073705673, "learning_rate": 5.7480218564832125e-05, "loss": 0.0148, "step": 47960 }, { "grad_norm": 0.0833311453461647, "learning_rate": 5.746386954044082e-05, "loss": 0.0101, "step": 47970 }, { "grad_norm": 0.1460621953010559, "learning_rate": 5.744751969981195e-05, "loss": 0.0077, "step": 47980 }, { "grad_norm": 0.11966503411531448, "learning_rate": 5.7431169044733526e-05, "loss": 0.0132, "step": 47990 }, { "grad_norm": 0.08264995366334915, "learning_rate": 5.741481757699364e-05, "loss": 0.0093, "step": 48000 }, { "grad_norm": 0.23271092772483826, "learning_rate": 5.7398465298380434e-05, "loss": 0.0098, "step": 48010 }, { "grad_norm": 0.11111143976449966, "learning_rate": 5.7382112210682193e-05, "loss": 0.0153, "step": 48020 }, { "grad_norm": 0.1830282062292099, "learning_rate": 5.7365758315687266e-05, "loss": 0.0107, "step": 48030 }, { "grad_norm": 0.11885082721710205, "learning_rate": 5.734940361518407e-05, "loss": 0.0117, "step": 48040 }, { "grad_norm": 0.1158476173877716, "learning_rate": 5.733304811096116e-05, "loss": 0.0101, "step": 48050 }, { "grad_norm": 0.1583596020936966, "learning_rate": 5.731669180480713e-05, "loss": 0.013, "step": 48060 }, { "grad_norm": 0.15643639862537384, "learning_rate": 5.730033469851067e-05, "loss": 0.018, "step": 48070 }, { "grad_norm": 0.16079887747764587, "learning_rate": 5.7283976793860607e-05, "loss": 0.0131, "step": 48080 }, { "grad_norm": 0.0915820300579071, "learning_rate": 5.726761809264577e-05, "loss": 0.0129, "step": 48090 }, { "grad_norm": 0.14064332842826843, "learning_rate": 5.7251258596655155e-05, "loss": 0.0109, "step": 48100 }, { "grad_norm": 0.11767873913049698, "learning_rate": 5.72348983076778e-05, "loss": 0.0128, "step": 48110 }, { "grad_norm": 0.09608717262744904, "learning_rate": 5.7218537227502854e-05, "loss": 0.0107, "step": 48120 }, { "grad_norm": 0.08945269137620926, "learning_rate": 5.720217535791951e-05, "loss": 0.009, "step": 48130 }, { "grad_norm": 0.09685178846120834, "learning_rate": 5.718581270071711e-05, "loss": 0.0091, "step": 48140 }, { "grad_norm": 0.08339918404817581, "learning_rate": 5.716944925768505e-05, "loss": 0.0108, "step": 48150 }, { "grad_norm": 0.1009560152888298, "learning_rate": 5.7153085030612786e-05, "loss": 0.011, "step": 48160 }, { "grad_norm": 0.11557895690202713, "learning_rate": 5.713672002128989e-05, "loss": 0.014, "step": 48170 }, { "grad_norm": 0.17345979809761047, "learning_rate": 5.712035423150602e-05, "loss": 0.0115, "step": 48180 }, { "grad_norm": 0.08005564659833908, "learning_rate": 5.710398766305094e-05, "loss": 0.0108, "step": 48190 }, { "grad_norm": 0.233443021774292, "learning_rate": 5.708762031771442e-05, "loss": 0.0124, "step": 48200 }, { "grad_norm": 0.11879555135965347, "learning_rate": 5.707125219728642e-05, "loss": 0.0102, "step": 48210 }, { "grad_norm": 0.09066988527774811, "learning_rate": 5.7054883303556905e-05, "loss": 0.0078, "step": 48220 }, { "grad_norm": 0.1644354909658432, "learning_rate": 5.703851363831595e-05, "loss": 0.015, "step": 48230 }, { "grad_norm": 0.17255963385105133, "learning_rate": 5.702214320335374e-05, "loss": 0.0186, "step": 48240 }, { "grad_norm": 0.14904053509235382, "learning_rate": 5.7005772000460514e-05, "loss": 0.014, "step": 48250 }, { "grad_norm": 0.08060509711503983, "learning_rate": 5.698940003142656e-05, "loss": 0.0113, "step": 48260 }, { "grad_norm": 0.09278449416160583, "learning_rate": 5.697302729804236e-05, "loss": 0.0098, "step": 48270 }, { "grad_norm": 0.22717595100402832, "learning_rate": 5.695665380209837e-05, "loss": 0.0096, "step": 48280 }, { "grad_norm": 0.09532999247312546, "learning_rate": 5.6940279545385165e-05, "loss": 0.0107, "step": 48290 }, { "grad_norm": 0.09239061176776886, "learning_rate": 5.692390452969344e-05, "loss": 0.0095, "step": 48300 }, { "grad_norm": 0.09486018866300583, "learning_rate": 5.690752875681392e-05, "loss": 0.0109, "step": 48310 }, { "grad_norm": 0.11735258996486664, "learning_rate": 5.6891152228537435e-05, "loss": 0.0122, "step": 48320 }, { "grad_norm": 0.09827776253223419, "learning_rate": 5.687477494665492e-05, "loss": 0.0082, "step": 48330 }, { "grad_norm": 0.08567526191473007, "learning_rate": 5.685839691295733e-05, "loss": 0.0107, "step": 48340 }, { "grad_norm": 0.14412619173526764, "learning_rate": 5.6842018129235786e-05, "loss": 0.0131, "step": 48350 }, { "grad_norm": 0.12229298055171967, "learning_rate": 5.6825638597281404e-05, "loss": 0.0084, "step": 48360 }, { "grad_norm": 0.06436452269554138, "learning_rate": 5.680925831888546e-05, "loss": 0.0137, "step": 48370 }, { "grad_norm": 0.11529866605997086, "learning_rate": 5.6792877295839274e-05, "loss": 0.0123, "step": 48380 }, { "grad_norm": 0.5566753149032593, "learning_rate": 5.6776495529934224e-05, "loss": 0.0134, "step": 48390 }, { "grad_norm": 0.07451900094747543, "learning_rate": 5.6760113022961824e-05, "loss": 0.0101, "step": 48400 }, { "grad_norm": 0.07067735493183136, "learning_rate": 5.6743729776713617e-05, "loss": 0.0089, "step": 48410 }, { "grad_norm": 0.07764959335327148, "learning_rate": 5.672734579298126e-05, "loss": 0.0098, "step": 48420 }, { "grad_norm": 0.1006825789809227, "learning_rate": 5.671096107355649e-05, "loss": 0.0124, "step": 48430 }, { "grad_norm": 0.18528975546360016, "learning_rate": 5.66945756202311e-05, "loss": 0.0091, "step": 48440 }, { "grad_norm": 0.11562716960906982, "learning_rate": 5.667818943479699e-05, "loss": 0.0101, "step": 48450 }, { "grad_norm": 0.1544829159975052, "learning_rate": 5.666180251904612e-05, "loss": 0.0141, "step": 48460 }, { "grad_norm": 0.21079914271831512, "learning_rate": 5.6645414874770555e-05, "loss": 0.0136, "step": 48470 }, { "grad_norm": 0.14001551270484924, "learning_rate": 5.66290265037624e-05, "loss": 0.0111, "step": 48480 }, { "grad_norm": 0.11814115941524506, "learning_rate": 5.661263740781386e-05, "loss": 0.01, "step": 48490 }, { "grad_norm": 0.09345117956399918, "learning_rate": 5.6596247588717254e-05, "loss": 0.0111, "step": 48500 }, { "grad_norm": 0.08391116559505463, "learning_rate": 5.6579857048264926e-05, "loss": 0.0104, "step": 48510 }, { "grad_norm": 0.13821767270565033, "learning_rate": 5.6563465788249314e-05, "loss": 0.0099, "step": 48520 }, { "grad_norm": 0.11656571924686432, "learning_rate": 5.6547073810462956e-05, "loss": 0.0109, "step": 48530 }, { "grad_norm": 0.11577320098876953, "learning_rate": 5.653068111669846e-05, "loss": 0.0093, "step": 48540 }, { "grad_norm": 0.13532228767871857, "learning_rate": 5.651428770874848e-05, "loss": 0.013, "step": 48550 }, { "grad_norm": 0.07017216086387634, "learning_rate": 5.64978935884058e-05, "loss": 0.0107, "step": 48560 }, { "grad_norm": 0.2088249772787094, "learning_rate": 5.6481498757463244e-05, "loss": 0.0135, "step": 48570 }, { "grad_norm": 0.13245290517807007, "learning_rate": 5.646510321771373e-05, "loss": 0.0136, "step": 48580 }, { "grad_norm": 0.09231284260749817, "learning_rate": 5.644870697095024e-05, "loss": 0.0109, "step": 48590 }, { "grad_norm": 0.08380449563264847, "learning_rate": 5.643231001896586e-05, "loss": 0.0091, "step": 48600 }, { "grad_norm": 0.13342474400997162, "learning_rate": 5.6415912363553726e-05, "loss": 0.0373, "step": 48610 }, { "grad_norm": 0.21183277666568756, "learning_rate": 5.639951400650706e-05, "loss": 0.0118, "step": 48620 }, { "grad_norm": 0.15980611741542816, "learning_rate": 5.6383114949619165e-05, "loss": 0.0115, "step": 48630 }, { "grad_norm": 0.09828075021505356, "learning_rate": 5.636671519468342e-05, "loss": 0.0103, "step": 48640 }, { "grad_norm": 0.09886132925748825, "learning_rate": 5.635031474349327e-05, "loss": 0.0122, "step": 48650 }, { "grad_norm": 0.09750045090913773, "learning_rate": 5.6333913597842246e-05, "loss": 0.0093, "step": 48660 }, { "grad_norm": 0.17595410346984863, "learning_rate": 5.6317511759523955e-05, "loss": 0.013, "step": 48670 }, { "grad_norm": 0.1483246386051178, "learning_rate": 5.630110923033207e-05, "loss": 0.0098, "step": 48680 }, { "grad_norm": 0.10290094465017319, "learning_rate": 5.628470601206036e-05, "loss": 0.0138, "step": 48690 }, { "grad_norm": 0.08803140372037888, "learning_rate": 5.626830210650266e-05, "loss": 0.0117, "step": 48700 }, { "grad_norm": 0.07367272675037384, "learning_rate": 5.625189751545285e-05, "loss": 0.018, "step": 48710 }, { "grad_norm": 0.12769846618175507, "learning_rate": 5.6235492240704936e-05, "loss": 0.0126, "step": 48720 }, { "grad_norm": 0.1589965671300888, "learning_rate": 5.621908628405296e-05, "loss": 0.0172, "step": 48730 }, { "grad_norm": 0.16246668994426727, "learning_rate": 5.620267964729106e-05, "loss": 0.0144, "step": 48740 }, { "grad_norm": 0.11669357120990753, "learning_rate": 5.618627233221344e-05, "loss": 0.0095, "step": 48750 }, { "grad_norm": 0.19210976362228394, "learning_rate": 5.616986434061438e-05, "loss": 0.0137, "step": 48760 }, { "grad_norm": 0.07729484140872955, "learning_rate": 5.615345567428822e-05, "loss": 0.0091, "step": 48770 }, { "grad_norm": 0.10038081556558609, "learning_rate": 5.613704633502941e-05, "loss": 0.0082, "step": 48780 }, { "grad_norm": 0.1163610890507698, "learning_rate": 5.612063632463245e-05, "loss": 0.0106, "step": 48790 }, { "grad_norm": 0.18621008098125458, "learning_rate": 5.610422564489188e-05, "loss": 0.0106, "step": 48800 }, { "grad_norm": 0.16290362179279327, "learning_rate": 5.608781429760239e-05, "loss": 0.0124, "step": 48810 }, { "grad_norm": 0.0792756900191307, "learning_rate": 5.607140228455866e-05, "loss": 0.011, "step": 48820 }, { "grad_norm": 0.11005035787820816, "learning_rate": 5.605498960755553e-05, "loss": 0.0124, "step": 48830 }, { "grad_norm": 0.16946746408939362, "learning_rate": 5.603857626838782e-05, "loss": 0.0141, "step": 48840 }, { "grad_norm": 0.13029468059539795, "learning_rate": 5.60221622688505e-05, "loss": 0.0116, "step": 48850 }, { "grad_norm": 0.0806681290268898, "learning_rate": 5.6005747610738565e-05, "loss": 0.0114, "step": 48860 }, { "grad_norm": 0.07518566399812698, "learning_rate": 5.59893322958471e-05, "loss": 0.0097, "step": 48870 }, { "grad_norm": 0.11800812929868698, "learning_rate": 5.5972916325971256e-05, "loss": 0.0137, "step": 48880 }, { "grad_norm": 0.10591144859790802, "learning_rate": 5.595649970290628e-05, "loss": 0.0108, "step": 48890 }, { "grad_norm": 0.09716417640447617, "learning_rate": 5.5940082428447426e-05, "loss": 0.0137, "step": 48900 }, { "grad_norm": 0.09807916730642319, "learning_rate": 5.592366450439012e-05, "loss": 0.0185, "step": 48910 }, { "grad_norm": 0.20056377351284027, "learning_rate": 5.590724593252975e-05, "loss": 0.0133, "step": 48920 }, { "grad_norm": 0.1329261213541031, "learning_rate": 5.589082671466184e-05, "loss": 0.01, "step": 48930 }, { "grad_norm": 0.09732671082019806, "learning_rate": 5.587440685258199e-05, "loss": 0.0105, "step": 48940 }, { "grad_norm": 0.16816991567611694, "learning_rate": 5.585798634808583e-05, "loss": 0.0117, "step": 48950 }, { "grad_norm": 0.11581403017044067, "learning_rate": 5.584156520296909e-05, "loss": 0.0106, "step": 48960 }, { "grad_norm": 0.11621420830488205, "learning_rate": 5.582514341902757e-05, "loss": 0.0104, "step": 48970 }, { "grad_norm": 0.12135504931211472, "learning_rate": 5.580872099805713e-05, "loss": 0.0114, "step": 48980 }, { "grad_norm": 0.1405058056116104, "learning_rate": 5.5792297941853674e-05, "loss": 0.0092, "step": 48990 }, { "grad_norm": 0.10091209411621094, "learning_rate": 5.5775874252213247e-05, "loss": 0.0117, "step": 49000 }, { "grad_norm": 0.12678125500679016, "learning_rate": 5.575944993093189e-05, "loss": 0.0144, "step": 49010 }, { "grad_norm": 0.14581330120563507, "learning_rate": 5.574302497980574e-05, "loss": 0.0106, "step": 49020 }, { "grad_norm": 0.17774519324302673, "learning_rate": 5.5726599400631005e-05, "loss": 0.011, "step": 49030 }, { "grad_norm": 0.17670725286006927, "learning_rate": 5.571017319520399e-05, "loss": 0.0124, "step": 49040 }, { "grad_norm": 0.12782946228981018, "learning_rate": 5.5693746365320984e-05, "loss": 0.0117, "step": 49050 }, { "grad_norm": 0.152556374669075, "learning_rate": 5.567731891277848e-05, "loss": 0.0103, "step": 49060 }, { "grad_norm": 0.1502469778060913, "learning_rate": 5.566089083937288e-05, "loss": 0.0125, "step": 49070 }, { "grad_norm": 0.12041334807872772, "learning_rate": 5.564446214690079e-05, "loss": 0.0135, "step": 49080 }, { "grad_norm": 0.15153954923152924, "learning_rate": 5.56280328371588e-05, "loss": 0.0112, "step": 49090 }, { "grad_norm": 0.14338339865207672, "learning_rate": 5.5611602911943596e-05, "loss": 0.0114, "step": 49100 }, { "grad_norm": 0.08240264654159546, "learning_rate": 5.559517237305194e-05, "loss": 0.0115, "step": 49110 }, { "grad_norm": 0.46658027172088623, "learning_rate": 5.557874122228064e-05, "loss": 0.0133, "step": 49120 }, { "grad_norm": 0.12220345437526703, "learning_rate": 5.55623094614266e-05, "loss": 0.0103, "step": 49130 }, { "grad_norm": 0.08813539892435074, "learning_rate": 5.5545877092286743e-05, "loss": 0.0111, "step": 49140 }, { "grad_norm": 0.13234926760196686, "learning_rate": 5.552944411665811e-05, "loss": 0.0089, "step": 49150 }, { "grad_norm": 0.20188170671463013, "learning_rate": 5.5513010536337795e-05, "loss": 0.011, "step": 49160 }, { "grad_norm": 0.1878749132156372, "learning_rate": 5.5496576353122934e-05, "loss": 0.0108, "step": 49170 }, { "grad_norm": 0.12109418958425522, "learning_rate": 5.548014156881074e-05, "loss": 0.0083, "step": 49180 }, { "grad_norm": 0.20500893890857697, "learning_rate": 5.5463706185198494e-05, "loss": 0.0106, "step": 49190 }, { "grad_norm": 0.06869497150182724, "learning_rate": 5.544727020408358e-05, "loss": 0.0085, "step": 49200 }, { "grad_norm": 0.11459677666425705, "learning_rate": 5.5430833627263366e-05, "loss": 0.0102, "step": 49210 }, { "grad_norm": 0.10358815640211105, "learning_rate": 5.5414396456535354e-05, "loss": 0.0097, "step": 49220 }, { "grad_norm": 0.10933726280927658, "learning_rate": 5.539795869369709e-05, "loss": 0.0093, "step": 49230 }, { "grad_norm": 0.15368390083312988, "learning_rate": 5.538152034054618e-05, "loss": 0.0113, "step": 49240 }, { "grad_norm": 0.09238513559103012, "learning_rate": 5.536508139888028e-05, "loss": 0.0104, "step": 49250 }, { "grad_norm": 0.14523372054100037, "learning_rate": 5.534864187049716e-05, "loss": 0.0101, "step": 49260 }, { "grad_norm": 0.17531488835811615, "learning_rate": 5.5332201757194615e-05, "loss": 0.0099, "step": 49270 }, { "grad_norm": 0.10539427399635315, "learning_rate": 5.531576106077049e-05, "loss": 0.0105, "step": 49280 }, { "grad_norm": 0.11702795326709747, "learning_rate": 5.529931978302272e-05, "loss": 0.0098, "step": 49290 }, { "grad_norm": 0.08977992087602615, "learning_rate": 5.528287792574932e-05, "loss": 0.0091, "step": 49300 }, { "grad_norm": 0.08434640616178513, "learning_rate": 5.5266435490748324e-05, "loss": 0.0096, "step": 49310 }, { "grad_norm": 0.1240873858332634, "learning_rate": 5.524999247981787e-05, "loss": 0.012, "step": 49320 }, { "grad_norm": 0.07796113938093185, "learning_rate": 5.523354889475613e-05, "loss": 0.0088, "step": 49330 }, { "grad_norm": 0.15517954528331757, "learning_rate": 5.521710473736134e-05, "loss": 0.0132, "step": 49340 }, { "grad_norm": 0.08633444458246231, "learning_rate": 5.5200660009431826e-05, "loss": 0.01, "step": 49350 }, { "grad_norm": 0.09218156337738037, "learning_rate": 5.518421471276596e-05, "loss": 0.0092, "step": 49360 }, { "grad_norm": 0.11940672248601913, "learning_rate": 5.5167768849162174e-05, "loss": 0.008, "step": 49370 }, { "grad_norm": 0.11159716546535492, "learning_rate": 5.515132242041893e-05, "loss": 0.013, "step": 49380 }, { "grad_norm": 0.10105002671480179, "learning_rate": 5.513487542833483e-05, "loss": 0.0117, "step": 49390 }, { "grad_norm": 0.12058492749929428, "learning_rate": 5.5118427874708466e-05, "loss": 0.0099, "step": 49400 }, { "grad_norm": 0.08021748811006546, "learning_rate": 5.510197976133853e-05, "loss": 0.0094, "step": 49410 }, { "grad_norm": 0.0956086665391922, "learning_rate": 5.508553109002376e-05, "loss": 0.009, "step": 49420 }, { "grad_norm": 0.1120973601937294, "learning_rate": 5.5069081862562957e-05, "loss": 0.009, "step": 49430 }, { "grad_norm": 0.14359235763549805, "learning_rate": 5.5052632080754965e-05, "loss": 0.0097, "step": 49440 }, { "grad_norm": 0.11264940351247787, "learning_rate": 5.503618174639874e-05, "loss": 0.0103, "step": 49450 }, { "grad_norm": 0.13949619233608246, "learning_rate": 5.501973086129325e-05, "loss": 0.0076, "step": 49460 }, { "grad_norm": 0.10529503226280212, "learning_rate": 5.500327942723753e-05, "loss": 0.008, "step": 49470 }, { "grad_norm": 0.06416859477758408, "learning_rate": 5.498682744603071e-05, "loss": 0.0099, "step": 49480 }, { "grad_norm": 0.06607826799154282, "learning_rate": 5.497037491947194e-05, "loss": 0.0094, "step": 49490 }, { "grad_norm": 0.13532589375972748, "learning_rate": 5.4953921849360424e-05, "loss": 0.0167, "step": 49500 }, { "grad_norm": 0.1267155408859253, "learning_rate": 5.493746823749547e-05, "loss": 0.0134, "step": 49510 }, { "grad_norm": 0.15978792309761047, "learning_rate": 5.49210140856764e-05, "loss": 0.013, "step": 49520 }, { "grad_norm": 0.10360200703144073, "learning_rate": 5.4904559395702647e-05, "loss": 0.0102, "step": 49530 }, { "grad_norm": 0.09717220813035965, "learning_rate": 5.488810416937364e-05, "loss": 0.0095, "step": 49540 }, { "grad_norm": 0.1694607436656952, "learning_rate": 5.4871648408488926e-05, "loss": 0.0128, "step": 49550 }, { "grad_norm": 0.13679538667201996, "learning_rate": 5.485519211484807e-05, "loss": 0.0092, "step": 49560 }, { "grad_norm": 0.17883576452732086, "learning_rate": 5.4838735290250676e-05, "loss": 0.0122, "step": 49570 }, { "grad_norm": 0.2872430086135864, "learning_rate": 5.482227793649649e-05, "loss": 0.0119, "step": 49580 }, { "grad_norm": 0.1314241737127304, "learning_rate": 5.480582005538524e-05, "loss": 0.0129, "step": 49590 }, { "grad_norm": 0.14115825295448303, "learning_rate": 5.478936164871671e-05, "loss": 0.0135, "step": 49600 }, { "grad_norm": 0.21916989982128143, "learning_rate": 5.477290271829081e-05, "loss": 0.0137, "step": 49610 }, { "grad_norm": 0.09184059500694275, "learning_rate": 5.475644326590744e-05, "loss": 0.0101, "step": 49620 }, { "grad_norm": 0.10121342539787292, "learning_rate": 5.473998329336658e-05, "loss": 0.0124, "step": 49630 }, { "grad_norm": 0.15966391563415527, "learning_rate": 5.4723522802468286e-05, "loss": 0.0102, "step": 49640 }, { "grad_norm": 0.10166360437870026, "learning_rate": 5.470706179501264e-05, "loss": 0.0111, "step": 49650 }, { "grad_norm": 0.11291874200105667, "learning_rate": 5.4690600272799776e-05, "loss": 0.0121, "step": 49660 }, { "grad_norm": 0.2004029005765915, "learning_rate": 5.467413823762993e-05, "loss": 0.0127, "step": 49670 }, { "grad_norm": 0.1219596192240715, "learning_rate": 5.465767569130335e-05, "loss": 0.011, "step": 49680 }, { "grad_norm": 0.10291938483715057, "learning_rate": 5.464121263562036e-05, "loss": 0.0118, "step": 49690 }, { "grad_norm": 0.12288539856672287, "learning_rate": 5.4624749072381343e-05, "loss": 0.0119, "step": 49700 }, { "grad_norm": 0.08010602742433548, "learning_rate": 5.460828500338672e-05, "loss": 0.0094, "step": 49710 }, { "grad_norm": 0.11414150148630142, "learning_rate": 5.459182043043698e-05, "loss": 0.0136, "step": 49720 }, { "grad_norm": 0.10164673626422882, "learning_rate": 5.457535535533265e-05, "loss": 0.0102, "step": 49730 }, { "grad_norm": 0.07914713770151138, "learning_rate": 5.4558889779874334e-05, "loss": 0.0102, "step": 49740 }, { "grad_norm": 0.10304322838783264, "learning_rate": 5.454242370586269e-05, "loss": 0.0128, "step": 49750 }, { "grad_norm": 0.11570337414741516, "learning_rate": 5.452595713509843e-05, "loss": 0.0099, "step": 49760 }, { "grad_norm": 0.15822255611419678, "learning_rate": 5.4509490069382274e-05, "loss": 0.0102, "step": 49770 }, { "grad_norm": 0.07788527756929398, "learning_rate": 5.449302251051509e-05, "loss": 0.0078, "step": 49780 }, { "grad_norm": 0.1055484414100647, "learning_rate": 5.44765544602977e-05, "loss": 0.0086, "step": 49790 }, { "grad_norm": 0.09758533537387848, "learning_rate": 5.446008592053107e-05, "loss": 0.01, "step": 49800 }, { "grad_norm": 0.13245809078216553, "learning_rate": 5.4443616893016135e-05, "loss": 0.0114, "step": 49810 }, { "grad_norm": 0.09027286618947983, "learning_rate": 5.4427147379553934e-05, "loss": 0.0136, "step": 49820 }, { "grad_norm": 0.0730450376868248, "learning_rate": 5.441067738194556e-05, "loss": 0.0099, "step": 49830 }, { "grad_norm": 0.12172739952802658, "learning_rate": 5.439420690199214e-05, "loss": 0.0127, "step": 49840 }, { "grad_norm": 0.12845365703105927, "learning_rate": 5.437773594149484e-05, "loss": 0.0119, "step": 49850 }, { "grad_norm": 0.1109226793050766, "learning_rate": 5.436126450225495e-05, "loss": 0.0092, "step": 49860 }, { "grad_norm": 0.11068897694349289, "learning_rate": 5.434479258607371e-05, "loss": 0.0089, "step": 49870 }, { "grad_norm": 0.10239489376544952, "learning_rate": 5.43283201947525e-05, "loss": 0.01, "step": 49880 }, { "grad_norm": 0.1157042384147644, "learning_rate": 5.431184733009268e-05, "loss": 0.0181, "step": 49890 }, { "grad_norm": 0.10741601139307022, "learning_rate": 5.4295373993895736e-05, "loss": 0.0116, "step": 49900 }, { "grad_norm": 0.12327767163515091, "learning_rate": 5.4278900187963157e-05, "loss": 0.0092, "step": 49910 }, { "grad_norm": 0.09704026579856873, "learning_rate": 5.426242591409646e-05, "loss": 0.0106, "step": 49920 }, { "grad_norm": 0.13311757147312164, "learning_rate": 5.42459511740973e-05, "loss": 0.0088, "step": 49930 }, { "grad_norm": 0.08400406688451767, "learning_rate": 5.422947596976729e-05, "loss": 0.01, "step": 49940 }, { "grad_norm": 0.20223300158977509, "learning_rate": 5.4213000302908134e-05, "loss": 0.0134, "step": 49950 }, { "grad_norm": 0.1196112111210823, "learning_rate": 5.419652417532162e-05, "loss": 0.0097, "step": 49960 }, { "grad_norm": 0.09759445488452911, "learning_rate": 5.4180047588809534e-05, "loss": 0.0087, "step": 49970 }, { "grad_norm": 0.1490548998117447, "learning_rate": 5.4163570545173704e-05, "loss": 0.0097, "step": 49980 }, { "grad_norm": 0.1617552936077118, "learning_rate": 5.414709304621608e-05, "loss": 0.0092, "step": 49990 }, { "grad_norm": 0.10282540321350098, "learning_rate": 5.413061509373858e-05, "loss": 0.0127, "step": 50000 }, { "grad_norm": 0.069659523665905, "learning_rate": 5.411413668954324e-05, "loss": 0.0101, "step": 50010 }, { "grad_norm": 0.1527710258960724, "learning_rate": 5.4097657835432083e-05, "loss": 0.0081, "step": 50020 }, { "grad_norm": 0.11351160705089569, "learning_rate": 5.408117853320723e-05, "loss": 0.0096, "step": 50030 }, { "grad_norm": 0.11273396015167236, "learning_rate": 5.406469878467084e-05, "loss": 0.0117, "step": 50040 }, { "grad_norm": 0.11549652367830276, "learning_rate": 5.404821859162509e-05, "loss": 0.0114, "step": 50050 }, { "grad_norm": 0.17329242825508118, "learning_rate": 5.403173795587225e-05, "loss": 0.0105, "step": 50060 }, { "grad_norm": 0.5218756198883057, "learning_rate": 5.4015256879214606e-05, "loss": 0.0084, "step": 50070 }, { "grad_norm": 0.129831925034523, "learning_rate": 5.39987753634545e-05, "loss": 0.011, "step": 50080 }, { "grad_norm": 0.1229616031050682, "learning_rate": 5.3982293410394335e-05, "loss": 0.0097, "step": 50090 }, { "grad_norm": 0.14955081045627594, "learning_rate": 5.396581102183655e-05, "loss": 0.0107, "step": 50100 }, { "grad_norm": 0.09513723850250244, "learning_rate": 5.3949328199583634e-05, "loss": 0.0088, "step": 50110 }, { "grad_norm": 0.07930514961481094, "learning_rate": 5.393284494543813e-05, "loss": 0.0106, "step": 50120 }, { "grad_norm": 0.10595913976430893, "learning_rate": 5.391636126120262e-05, "loss": 0.0106, "step": 50130 }, { "grad_norm": 0.09837274253368378, "learning_rate": 5.389987714867971e-05, "loss": 0.0083, "step": 50140 }, { "grad_norm": 0.15416692197322845, "learning_rate": 5.3883392609672114e-05, "loss": 0.0099, "step": 50150 }, { "grad_norm": 0.14723438024520874, "learning_rate": 5.3866907645982545e-05, "loss": 0.01, "step": 50160 }, { "grad_norm": 0.1101071834564209, "learning_rate": 5.385042225941375e-05, "loss": 0.0091, "step": 50170 }, { "grad_norm": 0.170040562748909, "learning_rate": 5.3833936451768585e-05, "loss": 0.0111, "step": 50180 }, { "grad_norm": 0.11585794389247894, "learning_rate": 5.3817450224849884e-05, "loss": 0.0089, "step": 50190 }, { "grad_norm": 0.27596548199653625, "learning_rate": 5.380096358046056e-05, "loss": 0.0186, "step": 50200 }, { "grad_norm": 0.11376781016588211, "learning_rate": 5.378447652040359e-05, "loss": 0.0104, "step": 50210 }, { "grad_norm": 0.07583281397819519, "learning_rate": 5.376798904648195e-05, "loss": 0.0089, "step": 50220 }, { "grad_norm": 0.06693553924560547, "learning_rate": 5.375150116049869e-05, "loss": 0.0078, "step": 50230 }, { "grad_norm": 0.1687353104352951, "learning_rate": 5.373501286425691e-05, "loss": 0.0091, "step": 50240 }, { "grad_norm": 0.07371701300144196, "learning_rate": 5.3718524159559726e-05, "loss": 0.0078, "step": 50250 }, { "grad_norm": 0.13566187024116516, "learning_rate": 5.370203504821034e-05, "loss": 0.0115, "step": 50260 }, { "grad_norm": 0.14219941198825836, "learning_rate": 5.368554553201196e-05, "loss": 0.0142, "step": 50270 }, { "grad_norm": 0.08846938610076904, "learning_rate": 5.3669055612767874e-05, "loss": 0.0161, "step": 50280 }, { "grad_norm": 0.1629030406475067, "learning_rate": 5.3652565292281376e-05, "loss": 0.0113, "step": 50290 }, { "grad_norm": 0.10410210490226746, "learning_rate": 5.3636074572355824e-05, "loss": 0.0099, "step": 50300 }, { "grad_norm": 0.13162748515605927, "learning_rate": 5.361958345479463e-05, "loss": 0.0105, "step": 50310 }, { "grad_norm": 0.12082788348197937, "learning_rate": 5.360309194140123e-05, "loss": 0.0124, "step": 50320 }, { "grad_norm": 0.10942437499761581, "learning_rate": 5.3586600033979106e-05, "loss": 0.0115, "step": 50330 }, { "grad_norm": 0.17630858719348907, "learning_rate": 5.357010773433181e-05, "loss": 0.0098, "step": 50340 }, { "grad_norm": 0.12369644641876221, "learning_rate": 5.3553615044262895e-05, "loss": 0.0078, "step": 50350 }, { "grad_norm": 0.11741109937429428, "learning_rate": 5.3537121965575964e-05, "loss": 0.0106, "step": 50360 }, { "grad_norm": 0.10555721819400787, "learning_rate": 5.352062850007471e-05, "loss": 0.011, "step": 50370 }, { "grad_norm": 0.08846256881952286, "learning_rate": 5.350413464956282e-05, "loss": 0.0099, "step": 50380 }, { "grad_norm": 0.19762547314167023, "learning_rate": 5.348764041584403e-05, "loss": 0.0135, "step": 50390 }, { "grad_norm": 0.18789543211460114, "learning_rate": 5.347114580072212e-05, "loss": 0.0102, "step": 50400 }, { "grad_norm": 0.0883064866065979, "learning_rate": 5.345465080600093e-05, "loss": 0.0125, "step": 50410 }, { "grad_norm": 0.1181473359465599, "learning_rate": 5.3438155433484326e-05, "loss": 0.0076, "step": 50420 }, { "grad_norm": 0.11136897653341293, "learning_rate": 5.3421659684976197e-05, "loss": 0.0101, "step": 50430 }, { "grad_norm": 0.22761228680610657, "learning_rate": 5.340516356228052e-05, "loss": 0.0091, "step": 50440 }, { "grad_norm": 0.08633360266685486, "learning_rate": 5.338866706720128e-05, "loss": 0.0111, "step": 50450 }, { "grad_norm": 0.07837136834859848, "learning_rate": 5.337217020154249e-05, "loss": 0.0112, "step": 50460 }, { "grad_norm": 0.17298856377601624, "learning_rate": 5.335567296710825e-05, "loss": 0.0102, "step": 50470 }, { "grad_norm": 0.07817268371582031, "learning_rate": 5.333917536570265e-05, "loss": 0.008, "step": 50480 }, { "grad_norm": 0.09088225662708282, "learning_rate": 5.332267739912986e-05, "loss": 0.0089, "step": 50490 }, { "grad_norm": 0.11287792026996613, "learning_rate": 5.330617906919405e-05, "loss": 0.011, "step": 50500 }, { "grad_norm": 0.07446907460689545, "learning_rate": 5.328968037769949e-05, "loss": 0.0111, "step": 50510 }, { "grad_norm": 0.08172079175710678, "learning_rate": 5.327318132645042e-05, "loss": 0.0135, "step": 50520 }, { "grad_norm": 0.08525394648313522, "learning_rate": 5.325668191725116e-05, "loss": 0.0096, "step": 50530 }, { "grad_norm": 0.1632707267999649, "learning_rate": 5.324018215190606e-05, "loss": 0.011, "step": 50540 }, { "grad_norm": 0.1462525874376297, "learning_rate": 5.3223682032219515e-05, "loss": 0.0085, "step": 50550 }, { "grad_norm": 0.11064892262220383, "learning_rate": 5.320718155999595e-05, "loss": 0.0106, "step": 50560 }, { "grad_norm": 0.1281232088804245, "learning_rate": 5.3190680737039835e-05, "loss": 0.0123, "step": 50570 }, { "grad_norm": 0.12611918151378632, "learning_rate": 5.317417956515567e-05, "loss": 0.0089, "step": 50580 }, { "grad_norm": 0.14092621207237244, "learning_rate": 5.315767804614798e-05, "loss": 0.0113, "step": 50590 }, { "grad_norm": 0.06879538297653198, "learning_rate": 5.3141176181821395e-05, "loss": 0.0084, "step": 50600 }, { "grad_norm": 0.09340942651033401, "learning_rate": 5.31246739739805e-05, "loss": 0.0071, "step": 50610 }, { "grad_norm": 0.18267588317394257, "learning_rate": 5.310817142442994e-05, "loss": 0.0128, "step": 50620 }, { "grad_norm": 0.2434052675962448, "learning_rate": 5.309166853497445e-05, "loss": 0.0136, "step": 50630 }, { "grad_norm": 0.10864737629890442, "learning_rate": 5.307516530741873e-05, "loss": 0.0124, "step": 50640 }, { "grad_norm": 0.07908504456281662, "learning_rate": 5.305866174356754e-05, "loss": 0.0103, "step": 50650 }, { "grad_norm": 0.12432467192411423, "learning_rate": 5.304215784522571e-05, "loss": 0.0109, "step": 50660 }, { "grad_norm": 0.10874205827713013, "learning_rate": 5.302565361419808e-05, "loss": 0.0124, "step": 50670 }, { "grad_norm": 0.1540619134902954, "learning_rate": 5.3009149052289507e-05, "loss": 0.0161, "step": 50680 }, { "grad_norm": 0.0824929028749466, "learning_rate": 5.299264416130493e-05, "loss": 0.0096, "step": 50690 }, { "grad_norm": 0.11804217845201492, "learning_rate": 5.297613894304928e-05, "loss": 0.0089, "step": 50700 }, { "grad_norm": 0.09906168282032013, "learning_rate": 5.2959633399327534e-05, "loss": 0.0093, "step": 50710 }, { "grad_norm": 0.19167645275592804, "learning_rate": 5.294312753194476e-05, "loss": 0.0091, "step": 50720 }, { "grad_norm": 0.16588854789733887, "learning_rate": 5.292662134270596e-05, "loss": 0.0109, "step": 50730 }, { "grad_norm": 0.09648104757070541, "learning_rate": 5.291011483341626e-05, "loss": 0.014, "step": 50740 }, { "grad_norm": 0.218181312084198, "learning_rate": 5.2893608005880767e-05, "loss": 0.015, "step": 50750 }, { "grad_norm": 0.12086549401283264, "learning_rate": 5.287710086190467e-05, "loss": 0.0126, "step": 50760 }, { "grad_norm": 0.07785576581954956, "learning_rate": 5.286059340329314e-05, "loss": 0.0126, "step": 50770 }, { "grad_norm": 0.14158444106578827, "learning_rate": 5.284408563185141e-05, "loss": 0.0118, "step": 50780 }, { "grad_norm": 0.07216200232505798, "learning_rate": 5.282757754938476e-05, "loss": 0.0117, "step": 50790 }, { "grad_norm": 0.14134399592876434, "learning_rate": 5.281106915769849e-05, "loss": 0.0176, "step": 50800 }, { "grad_norm": 0.096343494951725, "learning_rate": 5.2794560458597897e-05, "loss": 0.0098, "step": 50810 }, { "grad_norm": 0.08453629165887833, "learning_rate": 5.27780514538884e-05, "loss": 0.0098, "step": 50820 }, { "grad_norm": 0.13174517452716827, "learning_rate": 5.2761542145375365e-05, "loss": 0.0095, "step": 50830 }, { "grad_norm": 0.07222193479537964, "learning_rate": 5.274503253486421e-05, "loss": 0.0084, "step": 50840 }, { "grad_norm": 0.08804693818092346, "learning_rate": 5.272852262416046e-05, "loss": 0.0111, "step": 50850 }, { "grad_norm": 0.08760951459407806, "learning_rate": 5.2712012415069555e-05, "loss": 0.0123, "step": 50860 }, { "grad_norm": 0.23756426572799683, "learning_rate": 5.269550190939705e-05, "loss": 0.0133, "step": 50870 }, { "grad_norm": 0.22677099704742432, "learning_rate": 5.267899110894852e-05, "loss": 0.01, "step": 50880 }, { "grad_norm": 0.11301382631063461, "learning_rate": 5.266248001552955e-05, "loss": 0.0095, "step": 50890 }, { "grad_norm": 0.08043865114450455, "learning_rate": 5.264596863094575e-05, "loss": 0.0106, "step": 50900 }, { "grad_norm": 0.1197287067770958, "learning_rate": 5.262945695700282e-05, "loss": 0.0103, "step": 50910 }, { "grad_norm": 0.1294659823179245, "learning_rate": 5.261294499550643e-05, "loss": 0.0127, "step": 50920 }, { "grad_norm": 0.12958939373493195, "learning_rate": 5.25964327482623e-05, "loss": 0.0097, "step": 50930 }, { "grad_norm": 0.10815922915935516, "learning_rate": 5.257992021707617e-05, "loss": 0.0089, "step": 50940 }, { "grad_norm": 0.1253366470336914, "learning_rate": 5.256340740375387e-05, "loss": 0.0145, "step": 50950 }, { "grad_norm": 0.1337498426437378, "learning_rate": 5.254689431010117e-05, "loss": 0.0118, "step": 50960 }, { "grad_norm": 0.06883320957422256, "learning_rate": 5.253038093792395e-05, "loss": 0.0077, "step": 50970 }, { "grad_norm": 0.1042814627289772, "learning_rate": 5.251386728902806e-05, "loss": 0.0122, "step": 50980 }, { "grad_norm": 0.15268197655677795, "learning_rate": 5.2497353365219446e-05, "loss": 0.0095, "step": 50990 }, { "grad_norm": 0.11969038844108582, "learning_rate": 5.2480839168304e-05, "loss": 0.0112, "step": 51000 }, { "grad_norm": 0.1323579102754593, "learning_rate": 5.246432470008773e-05, "loss": 0.0113, "step": 51010 }, { "grad_norm": 0.11273649334907532, "learning_rate": 5.2447809962376616e-05, "loss": 0.0104, "step": 51020 }, { "grad_norm": 0.09700343012809753, "learning_rate": 5.243129495697666e-05, "loss": 0.0087, "step": 51030 }, { "grad_norm": 0.07285244017839432, "learning_rate": 5.241477968569397e-05, "loss": 0.012, "step": 51040 }, { "grad_norm": 0.18766118586063385, "learning_rate": 5.23982641503346e-05, "loss": 0.0097, "step": 51050 }, { "grad_norm": 0.17930907011032104, "learning_rate": 5.238174835270465e-05, "loss": 0.0129, "step": 51060 }, { "grad_norm": 0.07234922051429749, "learning_rate": 5.23652322946103e-05, "loss": 0.0067, "step": 51070 }, { "grad_norm": 0.07136836647987366, "learning_rate": 5.234871597785771e-05, "loss": 0.008, "step": 51080 }, { "grad_norm": 0.17037798464298248, "learning_rate": 5.233219940425306e-05, "loss": 0.0125, "step": 51090 }, { "grad_norm": 0.2593158185482025, "learning_rate": 5.2315682575602576e-05, "loss": 0.022, "step": 51100 }, { "grad_norm": 0.11819583177566528, "learning_rate": 5.229916549371254e-05, "loss": 0.0106, "step": 51110 }, { "grad_norm": 0.1174670159816742, "learning_rate": 5.228264816038924e-05, "loss": 0.0094, "step": 51120 }, { "grad_norm": 0.13715261220932007, "learning_rate": 5.2266130577438935e-05, "loss": 0.01, "step": 51130 }, { "grad_norm": 0.11649003624916077, "learning_rate": 5.224961274666802e-05, "loss": 0.0112, "step": 51140 }, { "grad_norm": 0.1080213263630867, "learning_rate": 5.223309466988283e-05, "loss": 0.0176, "step": 51150 }, { "grad_norm": 0.09024038910865784, "learning_rate": 5.221657634888976e-05, "loss": 0.0153, "step": 51160 }, { "grad_norm": 0.119744211435318, "learning_rate": 5.2200057785495235e-05, "loss": 0.016, "step": 51170 }, { "grad_norm": 0.09474345296621323, "learning_rate": 5.2183538981505706e-05, "loss": 0.0108, "step": 51180 }, { "grad_norm": 0.1028803139925003, "learning_rate": 5.216701993872762e-05, "loss": 0.0098, "step": 51190 }, { "grad_norm": 0.10321641713380814, "learning_rate": 5.215050065896752e-05, "loss": 0.015, "step": 51200 }, { "grad_norm": 0.14025090634822845, "learning_rate": 5.2133981144031875e-05, "loss": 0.0201, "step": 51210 }, { "grad_norm": 0.1269555389881134, "learning_rate": 5.211746139572727e-05, "loss": 0.0117, "step": 51220 }, { "grad_norm": 0.08694514632225037, "learning_rate": 5.210094141586026e-05, "loss": 0.0126, "step": 51230 }, { "grad_norm": 0.11416008323431015, "learning_rate": 5.2084421206237476e-05, "loss": 0.0111, "step": 51240 }, { "grad_norm": 0.12476760894060135, "learning_rate": 5.20679007686655e-05, "loss": 0.0129, "step": 51250 }, { "grad_norm": 0.0917336642742157, "learning_rate": 5.205138010495102e-05, "loss": 0.01, "step": 51260 }, { "grad_norm": 0.1385759562253952, "learning_rate": 5.20348592169007e-05, "loss": 0.0127, "step": 51270 }, { "grad_norm": 0.1484721601009369, "learning_rate": 5.201833810632123e-05, "loss": 0.0112, "step": 51280 }, { "grad_norm": 0.10301337391138077, "learning_rate": 5.200181677501933e-05, "loss": 0.0103, "step": 51290 }, { "grad_norm": 0.10082630068063736, "learning_rate": 5.1985295224801777e-05, "loss": 0.0111, "step": 51300 }, { "grad_norm": 0.1419612020254135, "learning_rate": 5.196877345747531e-05, "loss": 0.0112, "step": 51310 }, { "grad_norm": 0.12581853568553925, "learning_rate": 5.195225147484674e-05, "loss": 0.0134, "step": 51320 }, { "grad_norm": 0.0886138528585434, "learning_rate": 5.193572927872288e-05, "loss": 0.011, "step": 51330 }, { "grad_norm": 0.18141376972198486, "learning_rate": 5.191920687091059e-05, "loss": 0.0132, "step": 51340 }, { "grad_norm": 0.09736675769090652, "learning_rate": 5.1902684253216695e-05, "loss": 0.0118, "step": 51350 }, { "grad_norm": 0.11685513705015182, "learning_rate": 5.1886161427448134e-05, "loss": 0.0111, "step": 51360 }, { "grad_norm": 0.12149473279714584, "learning_rate": 5.186963839541179e-05, "loss": 0.0103, "step": 51370 }, { "grad_norm": 0.1730288565158844, "learning_rate": 5.185311515891459e-05, "loss": 0.0075, "step": 51380 }, { "grad_norm": 0.09701148420572281, "learning_rate": 5.1836591719763516e-05, "loss": 0.0104, "step": 51390 }, { "grad_norm": 0.07080359756946564, "learning_rate": 5.182006807976554e-05, "loss": 0.0101, "step": 51400 }, { "grad_norm": 0.07714831829071045, "learning_rate": 5.180354424072763e-05, "loss": 0.0085, "step": 51410 }, { "grad_norm": 0.17504505813121796, "learning_rate": 5.178702020445686e-05, "loss": 0.0107, "step": 51420 }, { "grad_norm": 0.14156202971935272, "learning_rate": 5.177049597276024e-05, "loss": 0.0105, "step": 51430 }, { "grad_norm": 0.1475515216588974, "learning_rate": 5.175397154744482e-05, "loss": 0.0117, "step": 51440 }, { "grad_norm": 0.1728581190109253, "learning_rate": 5.1737446930317736e-05, "loss": 0.0082, "step": 51450 }, { "grad_norm": 0.10482676327228546, "learning_rate": 5.172092212318604e-05, "loss": 0.0104, "step": 51460 }, { "grad_norm": 0.08118309825658798, "learning_rate": 5.170439712785692e-05, "loss": 0.0124, "step": 51470 }, { "grad_norm": 0.18821237981319427, "learning_rate": 5.1687871946137467e-05, "loss": 0.0106, "step": 51480 }, { "grad_norm": 0.11219239234924316, "learning_rate": 5.16713465798349e-05, "loss": 0.0095, "step": 51490 }, { "grad_norm": 0.17736473679542542, "learning_rate": 5.165482103075637e-05, "loss": 0.0097, "step": 51500 }, { "grad_norm": 0.12177419662475586, "learning_rate": 5.163829530070909e-05, "loss": 0.0109, "step": 51510 }, { "grad_norm": 0.19221912324428558, "learning_rate": 5.162176939150032e-05, "loss": 0.0128, "step": 51520 }, { "grad_norm": 0.17041872441768646, "learning_rate": 5.160524330493728e-05, "loss": 0.0116, "step": 51530 }, { "grad_norm": 0.3587492108345032, "learning_rate": 5.158871704282724e-05, "loss": 0.0079, "step": 51540 }, { "grad_norm": 0.11306940764188766, "learning_rate": 5.15721906069775e-05, "loss": 0.0135, "step": 51550 }, { "grad_norm": 0.1041005551815033, "learning_rate": 5.155566399919536e-05, "loss": 0.0094, "step": 51560 }, { "grad_norm": 0.1265542358160019, "learning_rate": 5.153913722128813e-05, "loss": 0.0126, "step": 51570 }, { "grad_norm": 0.0894809290766716, "learning_rate": 5.1522610275063176e-05, "loss": 0.0093, "step": 51580 }, { "grad_norm": 0.08993442356586456, "learning_rate": 5.150608316232786e-05, "loss": 0.007, "step": 51590 }, { "grad_norm": 0.16087020933628082, "learning_rate": 5.148955588488955e-05, "loss": 0.0086, "step": 51600 }, { "grad_norm": 0.12236166000366211, "learning_rate": 5.147302844455564e-05, "loss": 0.0089, "step": 51610 }, { "grad_norm": 0.10247170925140381, "learning_rate": 5.145650084313356e-05, "loss": 0.0107, "step": 51620 }, { "grad_norm": 0.11952120810747147, "learning_rate": 5.1439973082430734e-05, "loss": 0.0085, "step": 51630 }, { "grad_norm": 0.107596255838871, "learning_rate": 5.142344516425461e-05, "loss": 0.0099, "step": 51640 }, { "grad_norm": 0.08577266335487366, "learning_rate": 5.1406917090412673e-05, "loss": 0.0111, "step": 51650 }, { "grad_norm": 0.1776844710111618, "learning_rate": 5.1390388862712404e-05, "loss": 0.0119, "step": 51660 }, { "grad_norm": 0.10700107365846634, "learning_rate": 5.137386048296129e-05, "loss": 0.0089, "step": 51670 }, { "grad_norm": 0.20515954494476318, "learning_rate": 5.135733195296686e-05, "loss": 0.0163, "step": 51680 }, { "grad_norm": 0.09439630806446075, "learning_rate": 5.134080327453664e-05, "loss": 0.0112, "step": 51690 }, { "grad_norm": 0.11724022030830383, "learning_rate": 5.1324274449478204e-05, "loss": 0.0095, "step": 51700 }, { "grad_norm": 0.11456847190856934, "learning_rate": 5.130774547959909e-05, "loss": 0.0094, "step": 51710 }, { "grad_norm": 0.09545985609292984, "learning_rate": 5.129121636670692e-05, "loss": 0.0101, "step": 51720 }, { "grad_norm": 0.08459150046110153, "learning_rate": 5.1274687112609246e-05, "loss": 0.0078, "step": 51730 }, { "grad_norm": 0.09590082615613937, "learning_rate": 5.125815771911372e-05, "loss": 0.0084, "step": 51740 }, { "grad_norm": 0.3186180591583252, "learning_rate": 5.1241628188027956e-05, "loss": 0.0081, "step": 51750 }, { "grad_norm": 0.08881696313619614, "learning_rate": 5.1225098521159585e-05, "loss": 0.0104, "step": 51760 }, { "grad_norm": 0.10327328741550446, "learning_rate": 5.12085687203163e-05, "loss": 0.0099, "step": 51770 }, { "grad_norm": 0.07632823288440704, "learning_rate": 5.119203878730575e-05, "loss": 0.0096, "step": 51780 }, { "grad_norm": 0.08226723968982697, "learning_rate": 5.117550872393564e-05, "loss": 0.0115, "step": 51790 }, { "grad_norm": 0.06930995732545853, "learning_rate": 5.1158978532013645e-05, "loss": 0.0087, "step": 51800 }, { "grad_norm": 0.11943355947732925, "learning_rate": 5.114244821334751e-05, "loss": 0.0121, "step": 51810 }, { "grad_norm": 0.10166120529174805, "learning_rate": 5.112591776974497e-05, "loss": 0.009, "step": 51820 }, { "grad_norm": 0.09072492271661758, "learning_rate": 5.110938720301374e-05, "loss": 0.0089, "step": 51830 }, { "grad_norm": 0.12102916836738586, "learning_rate": 5.1092856514961595e-05, "loss": 0.0121, "step": 51840 }, { "grad_norm": 0.1435013711452484, "learning_rate": 5.107632570739631e-05, "loss": 0.0093, "step": 51850 }, { "grad_norm": 0.0859345942735672, "learning_rate": 5.1059794782125645e-05, "loss": 0.0108, "step": 51860 }, { "grad_norm": 0.2623209059238434, "learning_rate": 5.104326374095744e-05, "loss": 0.0156, "step": 51870 }, { "grad_norm": 0.1204453706741333, "learning_rate": 5.102673258569947e-05, "loss": 0.0109, "step": 51880 }, { "grad_norm": 0.1327551305294037, "learning_rate": 5.101020131815956e-05, "loss": 0.0133, "step": 51890 }, { "grad_norm": 0.09637093544006348, "learning_rate": 5.099366994014556e-05, "loss": 0.0083, "step": 51900 }, { "grad_norm": 0.1869071125984192, "learning_rate": 5.097713845346532e-05, "loss": 0.0167, "step": 51910 }, { "grad_norm": 0.11279024928808212, "learning_rate": 5.096060685992668e-05, "loss": 0.0097, "step": 51920 }, { "grad_norm": 0.15767377614974976, "learning_rate": 5.0944075161337514e-05, "loss": 0.0116, "step": 51930 }, { "grad_norm": 0.08623076230287552, "learning_rate": 5.09275433595057e-05, "loss": 0.0103, "step": 51940 }, { "grad_norm": 0.13787804543972015, "learning_rate": 5.0911011456239157e-05, "loss": 0.0119, "step": 51950 }, { "grad_norm": 0.18582944571971893, "learning_rate": 5.089447945334574e-05, "loss": 0.0114, "step": 51960 }, { "grad_norm": 0.0866110622882843, "learning_rate": 5.087794735263343e-05, "loss": 0.0084, "step": 51970 }, { "grad_norm": 0.11193197965621948, "learning_rate": 5.086141515591011e-05, "loss": 0.009, "step": 51980 }, { "grad_norm": 0.12437689304351807, "learning_rate": 5.084488286498371e-05, "loss": 0.012, "step": 51990 }, { "grad_norm": 0.16568323969841003, "learning_rate": 5.08283504816622e-05, "loss": 0.0147, "step": 52000 }, { "grad_norm": 0.1891525238752365, "learning_rate": 5.081181800775353e-05, "loss": 0.0193, "step": 52010 }, { "grad_norm": 0.3798968195915222, "learning_rate": 5.079528544506565e-05, "loss": 0.0098, "step": 52020 }, { "grad_norm": 0.11043898016214371, "learning_rate": 5.077875279540657e-05, "loss": 0.0095, "step": 52030 }, { "grad_norm": 0.09713903814554214, "learning_rate": 5.076222006058424e-05, "loss": 0.0208, "step": 52040 }, { "grad_norm": 0.10058341920375824, "learning_rate": 5.074568724240667e-05, "loss": 0.0101, "step": 52050 }, { "grad_norm": 0.078293077647686, "learning_rate": 5.072915434268187e-05, "loss": 0.0126, "step": 52060 }, { "grad_norm": 0.09955361485481262, "learning_rate": 5.071262136321785e-05, "loss": 0.012, "step": 52070 }, { "grad_norm": 0.11228996515274048, "learning_rate": 5.069608830582262e-05, "loss": 0.0095, "step": 52080 }, { "grad_norm": 0.11040035635232925, "learning_rate": 5.067955517230423e-05, "loss": 0.0096, "step": 52090 }, { "grad_norm": 0.08709273487329483, "learning_rate": 5.06630219644707e-05, "loss": 0.0133, "step": 52100 }, { "grad_norm": 0.08057442307472229, "learning_rate": 5.064648868413008e-05, "loss": 0.0097, "step": 52110 }, { "grad_norm": 0.14010415971279144, "learning_rate": 5.062995533309044e-05, "loss": 0.0096, "step": 52120 }, { "grad_norm": 0.06638393551111221, "learning_rate": 5.061342191315983e-05, "loss": 0.0129, "step": 52130 }, { "grad_norm": 0.13017961382865906, "learning_rate": 5.0596888426146325e-05, "loss": 0.0096, "step": 52140 }, { "grad_norm": 0.05976099520921707, "learning_rate": 5.058035487385798e-05, "loss": 0.0081, "step": 52150 }, { "grad_norm": 0.0992966964840889, "learning_rate": 5.0563821258102915e-05, "loss": 0.0082, "step": 52160 }, { "grad_norm": 0.13383464515209198, "learning_rate": 5.05472875806892e-05, "loss": 0.0075, "step": 52170 }, { "grad_norm": 0.12789064645767212, "learning_rate": 5.053075384342494e-05, "loss": 0.0122, "step": 52180 }, { "grad_norm": 0.09682276844978333, "learning_rate": 5.0514220048118234e-05, "loss": 0.0094, "step": 52190 }, { "grad_norm": 0.08843443542718887, "learning_rate": 5.049768619657721e-05, "loss": 0.0109, "step": 52200 }, { "grad_norm": 0.07735030353069305, "learning_rate": 5.048115229060996e-05, "loss": 0.0098, "step": 52210 }, { "grad_norm": 0.11205586045980453, "learning_rate": 5.046461833202462e-05, "loss": 0.0117, "step": 52220 }, { "grad_norm": 0.11616187542676926, "learning_rate": 5.044808432262932e-05, "loss": 0.0145, "step": 52230 }, { "grad_norm": 0.13399815559387207, "learning_rate": 5.043155026423219e-05, "loss": 0.0101, "step": 52240 }, { "grad_norm": 0.09836404025554657, "learning_rate": 5.041501615864137e-05, "loss": 0.0067, "step": 52250 }, { "grad_norm": 0.10480085760354996, "learning_rate": 5.039848200766502e-05, "loss": 0.0088, "step": 52260 }, { "grad_norm": 0.0741194412112236, "learning_rate": 5.038194781311125e-05, "loss": 0.0081, "step": 52270 }, { "grad_norm": 0.08490044623613358, "learning_rate": 5.0365413576788264e-05, "loss": 0.0085, "step": 52280 }, { "grad_norm": 0.10578721016645432, "learning_rate": 5.0348879300504194e-05, "loss": 0.01, "step": 52290 }, { "grad_norm": 0.05573024973273277, "learning_rate": 5.033234498606721e-05, "loss": 0.0078, "step": 52300 }, { "grad_norm": 0.1729273498058319, "learning_rate": 5.0315810635285476e-05, "loss": 0.0098, "step": 52310 }, { "grad_norm": 0.1518976390361786, "learning_rate": 5.0299276249967164e-05, "loss": 0.0102, "step": 52320 }, { "grad_norm": 0.0973798930644989, "learning_rate": 5.0282741831920454e-05, "loss": 0.0137, "step": 52330 }, { "grad_norm": 0.19302520155906677, "learning_rate": 5.026620738295351e-05, "loss": 0.0124, "step": 52340 }, { "grad_norm": 0.08043353259563446, "learning_rate": 5.024967290487455e-05, "loss": 0.0077, "step": 52350 }, { "grad_norm": 0.0872223824262619, "learning_rate": 5.023313839949172e-05, "loss": 0.0084, "step": 52360 }, { "grad_norm": 0.08188586682081223, "learning_rate": 5.0216603868613224e-05, "loss": 0.0097, "step": 52370 }, { "grad_norm": 0.11169745773077011, "learning_rate": 5.020006931404726e-05, "loss": 0.0126, "step": 52380 }, { "grad_norm": 0.08549628406763077, "learning_rate": 5.0183534737602014e-05, "loss": 0.0087, "step": 52390 }, { "grad_norm": 0.13663889467716217, "learning_rate": 5.0167000141085686e-05, "loss": 0.0118, "step": 52400 }, { "grad_norm": 0.09848953783512115, "learning_rate": 5.015046552630647e-05, "loss": 0.0093, "step": 52410 }, { "grad_norm": 0.12040964514017105, "learning_rate": 5.013393089507258e-05, "loss": 0.0137, "step": 52420 }, { "grad_norm": 0.13304324448108673, "learning_rate": 5.011739624919219e-05, "loss": 0.0098, "step": 52430 }, { "grad_norm": 0.10731640458106995, "learning_rate": 5.0100861590473545e-05, "loss": 0.0086, "step": 52440 }, { "grad_norm": 0.12147682905197144, "learning_rate": 5.008432692072482e-05, "loss": 0.0089, "step": 52450 }, { "grad_norm": 0.14444305002689362, "learning_rate": 5.006779224175423e-05, "loss": 0.0164, "step": 52460 }, { "grad_norm": 0.13096807897090912, "learning_rate": 5.005125755536998e-05, "loss": 0.0117, "step": 52470 }, { "grad_norm": 0.08256353437900543, "learning_rate": 5.003472286338029e-05, "loss": 0.012, "step": 52480 }, { "grad_norm": 0.08740247040987015, "learning_rate": 5.001818816759335e-05, "loss": 0.0228, "step": 52490 }, { "grad_norm": 0.12796801328659058, "learning_rate": 5.000165346981738e-05, "loss": 0.0095, "step": 52500 }, { "grad_norm": 0.1043025478720665, "learning_rate": 4.9985118771860595e-05, "loss": 0.0102, "step": 52510 }, { "grad_norm": 0.081805519759655, "learning_rate": 4.996858407553119e-05, "loss": 0.0095, "step": 52520 }, { "grad_norm": 0.08728757500648499, "learning_rate": 4.99520493826374e-05, "loss": 0.0085, "step": 52530 }, { "grad_norm": 0.1331944763660431, "learning_rate": 4.9935514694987384e-05, "loss": 0.0101, "step": 52540 }, { "grad_norm": 0.16273833811283112, "learning_rate": 4.991898001438941e-05, "loss": 0.0084, "step": 52550 }, { "grad_norm": 0.09534332901239395, "learning_rate": 4.990244534265166e-05, "loss": 0.0103, "step": 52560 }, { "grad_norm": 0.1309708058834076, "learning_rate": 4.988591068158234e-05, "loss": 0.0098, "step": 52570 }, { "grad_norm": 0.1227567121386528, "learning_rate": 4.986937603298965e-05, "loss": 0.0112, "step": 52580 }, { "grad_norm": 0.08330486714839935, "learning_rate": 4.985284139868181e-05, "loss": 0.0081, "step": 52590 }, { "grad_norm": 0.14994855225086212, "learning_rate": 4.9836306780467e-05, "loss": 0.0098, "step": 52600 }, { "grad_norm": 0.0901746079325676, "learning_rate": 4.981977218015343e-05, "loss": 0.0079, "step": 52610 }, { "grad_norm": 0.07873330265283585, "learning_rate": 4.980323759954932e-05, "loss": 0.0073, "step": 52620 }, { "grad_norm": 0.10188865661621094, "learning_rate": 4.978670304046284e-05, "loss": 0.0078, "step": 52630 }, { "grad_norm": 0.2462729513645172, "learning_rate": 4.9770168504702195e-05, "loss": 0.0082, "step": 52640 }, { "grad_norm": 0.14552249014377594, "learning_rate": 4.975363399407559e-05, "loss": 0.0105, "step": 52650 }, { "grad_norm": 0.09981777518987656, "learning_rate": 4.9737099510391194e-05, "loss": 0.0121, "step": 52660 }, { "grad_norm": 0.10406874120235443, "learning_rate": 4.972056505545719e-05, "loss": 0.0116, "step": 52670 }, { "grad_norm": 0.09549064934253693, "learning_rate": 4.9704030631081775e-05, "loss": 0.0094, "step": 52680 }, { "grad_norm": 0.16043713688850403, "learning_rate": 4.968749623907313e-05, "loss": 0.0078, "step": 52690 }, { "grad_norm": 0.0906938835978508, "learning_rate": 4.967096188123943e-05, "loss": 0.0126, "step": 52700 }, { "grad_norm": 0.1021309345960617, "learning_rate": 4.965442755938884e-05, "loss": 0.0083, "step": 52710 }, { "grad_norm": 0.08008964359760284, "learning_rate": 4.963789327532954e-05, "loss": 0.0091, "step": 52720 }, { "grad_norm": 0.13381588459014893, "learning_rate": 4.962135903086968e-05, "loss": 0.0083, "step": 52730 }, { "grad_norm": 0.08992419391870499, "learning_rate": 4.960482482781741e-05, "loss": 0.0078, "step": 52740 }, { "grad_norm": 0.10535196959972382, "learning_rate": 4.958829066798092e-05, "loss": 0.0085, "step": 52750 }, { "grad_norm": 0.07860993593931198, "learning_rate": 4.957175655316835e-05, "loss": 0.0084, "step": 52760 }, { "grad_norm": 0.09412870556116104, "learning_rate": 4.9555222485187834e-05, "loss": 0.0133, "step": 52770 }, { "grad_norm": 0.1060197576880455, "learning_rate": 4.953868846584752e-05, "loss": 0.0094, "step": 52780 }, { "grad_norm": 0.09346403926610947, "learning_rate": 4.9522154496955544e-05, "loss": 0.0154, "step": 52790 }, { "grad_norm": 0.1138492301106453, "learning_rate": 4.950562058032001e-05, "loss": 0.0117, "step": 52800 }, { "grad_norm": 0.07925308495759964, "learning_rate": 4.948908671774909e-05, "loss": 0.0118, "step": 52810 }, { "grad_norm": 0.20039436221122742, "learning_rate": 4.947255291105087e-05, "loss": 0.01, "step": 52820 }, { "grad_norm": 0.10813317447900772, "learning_rate": 4.945601916203348e-05, "loss": 0.0106, "step": 52830 }, { "grad_norm": 0.08908386528491974, "learning_rate": 4.9439485472505016e-05, "loss": 0.0098, "step": 52840 }, { "grad_norm": 0.07689917087554932, "learning_rate": 4.942295184427356e-05, "loss": 0.0098, "step": 52850 }, { "grad_norm": 0.10969863831996918, "learning_rate": 4.9406418279147234e-05, "loss": 0.0106, "step": 52860 }, { "grad_norm": 0.1048860177397728, "learning_rate": 4.9389884778934116e-05, "loss": 0.0113, "step": 52870 }, { "grad_norm": 0.09244661778211594, "learning_rate": 4.937335134544228e-05, "loss": 0.0095, "step": 52880 }, { "grad_norm": 0.15773369371891022, "learning_rate": 4.935681798047979e-05, "loss": 0.0126, "step": 52890 }, { "grad_norm": 0.09510476142168045, "learning_rate": 4.934028468585473e-05, "loss": 0.0114, "step": 52900 }, { "grad_norm": 0.08842550218105316, "learning_rate": 4.932375146337514e-05, "loss": 0.0106, "step": 52910 }, { "grad_norm": 0.07871349900960922, "learning_rate": 4.930721831484908e-05, "loss": 0.011, "step": 52920 }, { "grad_norm": 0.1291983425617218, "learning_rate": 4.929068524208456e-05, "loss": 0.012, "step": 52930 }, { "grad_norm": 0.09953099489212036, "learning_rate": 4.927415224688965e-05, "loss": 0.0151, "step": 52940 }, { "grad_norm": 0.09189358353614807, "learning_rate": 4.9257619331072375e-05, "loss": 0.0193, "step": 52950 }, { "grad_norm": 0.11025794595479965, "learning_rate": 4.924108649644072e-05, "loss": 0.0074, "step": 52960 }, { "grad_norm": 0.19708378612995148, "learning_rate": 4.922455374480271e-05, "loss": 0.0206, "step": 52970 }, { "grad_norm": 0.16844388842582703, "learning_rate": 4.920802107796634e-05, "loss": 0.0134, "step": 52980 }, { "grad_norm": 0.08960609883069992, "learning_rate": 4.9191488497739564e-05, "loss": 0.0098, "step": 52990 }, { "grad_norm": 0.10170576721429825, "learning_rate": 4.917495600593042e-05, "loss": 0.0109, "step": 53000 }, { "grad_norm": 0.12367818504571915, "learning_rate": 4.915842360434684e-05, "loss": 0.0106, "step": 53010 }, { "grad_norm": 0.19652485847473145, "learning_rate": 4.914189129479679e-05, "loss": 0.0162, "step": 53020 }, { "grad_norm": 0.12449301034212112, "learning_rate": 4.912535907908822e-05, "loss": 0.0104, "step": 53030 }, { "grad_norm": 0.17390041053295135, "learning_rate": 4.910882695902907e-05, "loss": 0.0137, "step": 53040 }, { "grad_norm": 0.10872449725866318, "learning_rate": 4.9092294936427234e-05, "loss": 0.0106, "step": 53050 }, { "grad_norm": 0.11763332039117813, "learning_rate": 4.9075763013090685e-05, "loss": 0.0089, "step": 53060 }, { "grad_norm": 0.10707349330186844, "learning_rate": 4.9059231190827305e-05, "loss": 0.0125, "step": 53070 }, { "grad_norm": 0.09589072316884995, "learning_rate": 4.904269947144499e-05, "loss": 0.0181, "step": 53080 }, { "grad_norm": 0.17878560721874237, "learning_rate": 4.902616785675161e-05, "loss": 0.0132, "step": 53090 }, { "grad_norm": 0.23360225558280945, "learning_rate": 4.900963634855505e-05, "loss": 0.0121, "step": 53100 }, { "grad_norm": 0.10032851248979568, "learning_rate": 4.8993104948663173e-05, "loss": 0.0095, "step": 53110 }, { "grad_norm": 0.10351736843585968, "learning_rate": 4.8976573658883823e-05, "loss": 0.0084, "step": 53120 }, { "grad_norm": 0.09013958275318146, "learning_rate": 4.896004248102484e-05, "loss": 0.0113, "step": 53130 }, { "grad_norm": 0.112424835562706, "learning_rate": 4.8943511416894045e-05, "loss": 0.0091, "step": 53140 }, { "grad_norm": 0.0870077833533287, "learning_rate": 4.892698046829926e-05, "loss": 0.0108, "step": 53150 }, { "grad_norm": 0.11848634481430054, "learning_rate": 4.8910449637048275e-05, "loss": 0.0175, "step": 53160 }, { "grad_norm": 0.1943790167570114, "learning_rate": 4.8893918924948884e-05, "loss": 0.0144, "step": 53170 }, { "grad_norm": 0.11604642122983932, "learning_rate": 4.8877388333808835e-05, "loss": 0.0148, "step": 53180 }, { "grad_norm": 0.16770631074905396, "learning_rate": 4.8860857865435925e-05, "loss": 0.0128, "step": 53190 }, { "grad_norm": 0.1378399282693863, "learning_rate": 4.884432752163789e-05, "loss": 0.0117, "step": 53200 }, { "grad_norm": 0.11528120189905167, "learning_rate": 4.8827797304222466e-05, "loss": 0.0091, "step": 53210 }, { "grad_norm": 0.10284435003995895, "learning_rate": 4.8811267214997366e-05, "loss": 0.0087, "step": 53220 }, { "grad_norm": 0.13430878520011902, "learning_rate": 4.879473725577029e-05, "loss": 0.0102, "step": 53230 }, { "grad_norm": 0.0887361541390419, "learning_rate": 4.877820742834891e-05, "loss": 0.0068, "step": 53240 }, { "grad_norm": 0.07273530960083008, "learning_rate": 4.8761677734540956e-05, "loss": 0.0095, "step": 53250 }, { "grad_norm": 0.11265426874160767, "learning_rate": 4.874514817615407e-05, "loss": 0.0094, "step": 53260 }, { "grad_norm": 0.09196499735116959, "learning_rate": 4.872861875499588e-05, "loss": 0.0099, "step": 53270 }, { "grad_norm": 0.10209427773952484, "learning_rate": 4.871208947287404e-05, "loss": 0.0138, "step": 53280 }, { "grad_norm": 0.3124416172504425, "learning_rate": 4.869556033159615e-05, "loss": 0.0105, "step": 53290 }, { "grad_norm": 0.09322074800729752, "learning_rate": 4.8679031332969835e-05, "loss": 0.0097, "step": 53300 }, { "grad_norm": 0.11153150349855423, "learning_rate": 4.8662502478802635e-05, "loss": 0.0143, "step": 53310 }, { "grad_norm": 0.12325698882341385, "learning_rate": 4.8645973770902176e-05, "loss": 0.0131, "step": 53320 }, { "grad_norm": 0.10664699971675873, "learning_rate": 4.862944521107599e-05, "loss": 0.0095, "step": 53330 }, { "grad_norm": 0.20412801206111908, "learning_rate": 4.861291680113162e-05, "loss": 0.013, "step": 53340 }, { "grad_norm": 0.09151830524206161, "learning_rate": 4.8596388542876563e-05, "loss": 0.0081, "step": 53350 }, { "grad_norm": 0.08298402279615402, "learning_rate": 4.857986043811836e-05, "loss": 0.0108, "step": 53360 }, { "grad_norm": 0.09435948729515076, "learning_rate": 4.856333248866448e-05, "loss": 0.0086, "step": 53370 }, { "grad_norm": 0.08362285047769547, "learning_rate": 4.85468046963224e-05, "loss": 0.0091, "step": 53380 }, { "grad_norm": 0.13550125062465668, "learning_rate": 4.853027706289957e-05, "loss": 0.0088, "step": 53390 }, { "grad_norm": 0.0794893205165863, "learning_rate": 4.851374959020344e-05, "loss": 0.0081, "step": 53400 }, { "grad_norm": 0.08534900099039078, "learning_rate": 4.849722228004141e-05, "loss": 0.0155, "step": 53410 }, { "grad_norm": 0.1260179877281189, "learning_rate": 4.848069513422091e-05, "loss": 0.0091, "step": 53420 }, { "grad_norm": 0.11439765244722366, "learning_rate": 4.8464168154549294e-05, "loss": 0.0089, "step": 53430 }, { "grad_norm": 0.12046555429697037, "learning_rate": 4.844764134283392e-05, "loss": 0.0101, "step": 53440 }, { "grad_norm": 0.13123507797718048, "learning_rate": 4.843111470088218e-05, "loss": 0.0109, "step": 53450 }, { "grad_norm": 0.115132637321949, "learning_rate": 4.841458823050137e-05, "loss": 0.0112, "step": 53460 }, { "grad_norm": 0.11869911104440689, "learning_rate": 4.8398061933498816e-05, "loss": 0.0087, "step": 53470 }, { "grad_norm": 0.22401051223278046, "learning_rate": 4.83815358116818e-05, "loss": 0.0102, "step": 53480 }, { "grad_norm": 0.1555614173412323, "learning_rate": 4.8365009866857584e-05, "loss": 0.0124, "step": 53490 }, { "grad_norm": 0.08981558680534363, "learning_rate": 4.834848410083342e-05, "loss": 0.0079, "step": 53500 }, { "grad_norm": 0.08619168400764465, "learning_rate": 4.833195851541657e-05, "loss": 0.0115, "step": 53510 }, { "grad_norm": 0.09566477686166763, "learning_rate": 4.831543311241423e-05, "loss": 0.0087, "step": 53520 }, { "grad_norm": 0.26160508394241333, "learning_rate": 4.829890789363359e-05, "loss": 0.0172, "step": 53530 }, { "grad_norm": 0.15501202642917633, "learning_rate": 4.828238286088182e-05, "loss": 0.0097, "step": 53540 }, { "grad_norm": 0.10392521321773529, "learning_rate": 4.826585801596609e-05, "loss": 0.012, "step": 53550 }, { "grad_norm": 0.088890939950943, "learning_rate": 4.8249333360693486e-05, "loss": 0.0098, "step": 53560 }, { "grad_norm": 0.11315624415874481, "learning_rate": 4.8232808896871185e-05, "loss": 0.0078, "step": 53570 }, { "grad_norm": 0.12403617799282074, "learning_rate": 4.821628462630624e-05, "loss": 0.0081, "step": 53580 }, { "grad_norm": 0.10256431996822357, "learning_rate": 4.8199760550805735e-05, "loss": 0.0092, "step": 53590 }, { "grad_norm": 0.18369559943675995, "learning_rate": 4.818323667217669e-05, "loss": 0.0098, "step": 53600 }, { "grad_norm": 0.19121339917182922, "learning_rate": 4.8166712992226165e-05, "loss": 0.0097, "step": 53610 }, { "grad_norm": 0.14749744534492493, "learning_rate": 4.8150189512761154e-05, "loss": 0.0105, "step": 53620 }, { "grad_norm": 0.09376010298728943, "learning_rate": 4.813366623558862e-05, "loss": 0.0098, "step": 53630 }, { "grad_norm": 0.10787726193666458, "learning_rate": 4.811714316251554e-05, "loss": 0.0097, "step": 53640 }, { "grad_norm": 0.15806154906749725, "learning_rate": 4.810062029534886e-05, "loss": 0.0104, "step": 53650 }, { "grad_norm": 0.11021211743354797, "learning_rate": 4.8084097635895505e-05, "loss": 0.0107, "step": 53660 }, { "grad_norm": 0.10759557783603668, "learning_rate": 4.8067575185962335e-05, "loss": 0.0102, "step": 53670 }, { "grad_norm": 0.11382919549942017, "learning_rate": 4.805105294735625e-05, "loss": 0.0114, "step": 53680 }, { "grad_norm": 0.14165517687797546, "learning_rate": 4.803453092188406e-05, "loss": 0.0138, "step": 53690 }, { "grad_norm": 0.10995909571647644, "learning_rate": 4.801800911135263e-05, "loss": 0.0107, "step": 53700 }, { "grad_norm": 0.0940459594130516, "learning_rate": 4.8001487517568744e-05, "loss": 0.01, "step": 53710 }, { "grad_norm": 0.12704552710056305, "learning_rate": 4.798496614233918e-05, "loss": 0.0115, "step": 53720 }, { "grad_norm": 0.12325501441955566, "learning_rate": 4.796844498747069e-05, "loss": 0.0153, "step": 53730 }, { "grad_norm": 0.1731531172990799, "learning_rate": 4.795192405477e-05, "loss": 0.0092, "step": 53740 }, { "grad_norm": 0.15909500420093536, "learning_rate": 4.79354033460438e-05, "loss": 0.0117, "step": 53750 }, { "grad_norm": 0.093590646982193, "learning_rate": 4.79188828630988e-05, "loss": 0.0165, "step": 53760 }, { "grad_norm": 0.10929247736930847, "learning_rate": 4.790236260774165e-05, "loss": 0.0094, "step": 53770 }, { "grad_norm": 0.1303306370973587, "learning_rate": 4.788584258177896e-05, "loss": 0.0106, "step": 53780 }, { "grad_norm": 0.1012304350733757, "learning_rate": 4.7869322787017356e-05, "loss": 0.0123, "step": 53790 }, { "grad_norm": 0.10242491215467453, "learning_rate": 4.78528032252634e-05, "loss": 0.009, "step": 53800 }, { "grad_norm": 0.10448324680328369, "learning_rate": 4.783628389832366e-05, "loss": 0.0108, "step": 53810 }, { "grad_norm": 0.11450324952602386, "learning_rate": 4.7819764808004644e-05, "loss": 0.0111, "step": 53820 }, { "grad_norm": 0.08750380575656891, "learning_rate": 4.780324595611289e-05, "loss": 0.0085, "step": 53830 }, { "grad_norm": 0.0952267050743103, "learning_rate": 4.7786727344454854e-05, "loss": 0.0164, "step": 53840 }, { "grad_norm": 0.06426442414522171, "learning_rate": 4.777020897483697e-05, "loss": 0.0132, "step": 53850 }, { "grad_norm": 0.0874306708574295, "learning_rate": 4.775369084906569e-05, "loss": 0.0088, "step": 53860 }, { "grad_norm": 0.10577883571386337, "learning_rate": 4.773717296894742e-05, "loss": 0.0156, "step": 53870 }, { "grad_norm": 0.09206506609916687, "learning_rate": 4.772065533628848e-05, "loss": 0.0092, "step": 53880 }, { "grad_norm": 0.11486020684242249, "learning_rate": 4.7704137952895265e-05, "loss": 0.0077, "step": 53890 }, { "grad_norm": 0.09683402627706528, "learning_rate": 4.768762082057407e-05, "loss": 0.0075, "step": 53900 }, { "grad_norm": 0.21089021861553192, "learning_rate": 4.7671103941131186e-05, "loss": 0.0122, "step": 53910 }, { "grad_norm": 0.10308129340410233, "learning_rate": 4.765458731637287e-05, "loss": 0.0085, "step": 53920 }, { "grad_norm": 0.12340079993009567, "learning_rate": 4.763807094810537e-05, "loss": 0.015, "step": 53930 }, { "grad_norm": 0.09714832901954651, "learning_rate": 4.762155483813487e-05, "loss": 0.0124, "step": 53940 }, { "grad_norm": 0.23427806794643402, "learning_rate": 4.760503898826754e-05, "loss": 0.0168, "step": 53950 }, { "grad_norm": 0.09198381751775742, "learning_rate": 4.758852340030957e-05, "loss": 0.0133, "step": 53960 }, { "grad_norm": 0.09071994572877884, "learning_rate": 4.757200807606705e-05, "loss": 0.0141, "step": 53970 }, { "grad_norm": 0.11145913600921631, "learning_rate": 4.755549301734607e-05, "loss": 0.0097, "step": 53980 }, { "grad_norm": 0.10622759908437729, "learning_rate": 4.753897822595271e-05, "loss": 0.018, "step": 53990 }, { "grad_norm": 0.11079499870538712, "learning_rate": 4.7522463703692994e-05, "loss": 0.0084, "step": 54000 }, { "grad_norm": 0.11014354974031448, "learning_rate": 4.75059494523729e-05, "loss": 0.0108, "step": 54010 }, { "grad_norm": 0.07416024804115295, "learning_rate": 4.748943547379843e-05, "loss": 0.0093, "step": 54020 }, { "grad_norm": 0.08391828835010529, "learning_rate": 4.7472921769775535e-05, "loss": 0.0169, "step": 54030 }, { "grad_norm": 0.1488305628299713, "learning_rate": 4.745640834211011e-05, "loss": 0.0143, "step": 54040 }, { "grad_norm": 0.10318297147750854, "learning_rate": 4.7439895192608054e-05, "loss": 0.0097, "step": 54050 }, { "grad_norm": 0.15923325717449188, "learning_rate": 4.7423382323075186e-05, "loss": 0.0101, "step": 54060 }, { "grad_norm": 0.08062246441841125, "learning_rate": 4.740686973531737e-05, "loss": 0.0073, "step": 54070 }, { "grad_norm": 0.13780167698860168, "learning_rate": 4.7390357431140375e-05, "loss": 0.0092, "step": 54080 }, { "grad_norm": 0.12289819866418839, "learning_rate": 4.737384541234998e-05, "loss": 0.0097, "step": 54090 }, { "grad_norm": 0.11518910527229309, "learning_rate": 4.735733368075189e-05, "loss": 0.0085, "step": 54100 }, { "grad_norm": 0.11941134929656982, "learning_rate": 4.734082223815182e-05, "loss": 0.0111, "step": 54110 }, { "grad_norm": 0.08103341609239578, "learning_rate": 4.7324311086355436e-05, "loss": 0.0084, "step": 54120 }, { "grad_norm": 0.111515112221241, "learning_rate": 4.730780022716836e-05, "loss": 0.0119, "step": 54130 }, { "grad_norm": 0.1774366796016693, "learning_rate": 4.729128966239619e-05, "loss": 0.0099, "step": 54140 }, { "grad_norm": 0.09527041763067245, "learning_rate": 4.7274779393844536e-05, "loss": 0.0105, "step": 54150 }, { "grad_norm": 0.25891464948654175, "learning_rate": 4.725826942331891e-05, "loss": 0.019, "step": 54160 }, { "grad_norm": 0.11296485364437103, "learning_rate": 4.724175975262481e-05, "loss": 0.0113, "step": 54170 }, { "grad_norm": 0.12278088182210922, "learning_rate": 4.722525038356772e-05, "loss": 0.0086, "step": 54180 }, { "grad_norm": 0.07368626445531845, "learning_rate": 4.7208741317953075e-05, "loss": 0.0101, "step": 54190 }, { "grad_norm": 0.12190617620944977, "learning_rate": 4.7192232557586264e-05, "loss": 0.0145, "step": 54200 }, { "grad_norm": 0.08269442617893219, "learning_rate": 4.717572410427269e-05, "loss": 0.0102, "step": 54210 }, { "grad_norm": 0.09944824874401093, "learning_rate": 4.71592159598177e-05, "loss": 0.0106, "step": 54220 }, { "grad_norm": 0.09297779202461243, "learning_rate": 4.714270812602657e-05, "loss": 0.0091, "step": 54230 }, { "grad_norm": 0.10485915094614029, "learning_rate": 4.712620060470458e-05, "loss": 0.0166, "step": 54240 }, { "grad_norm": 0.09287860244512558, "learning_rate": 4.7109693397656986e-05, "loss": 0.0083, "step": 54250 }, { "grad_norm": 0.08921574056148529, "learning_rate": 4.709318650668894e-05, "loss": 0.0079, "step": 54260 }, { "grad_norm": 0.07752286642789841, "learning_rate": 4.7076679933605676e-05, "loss": 0.0085, "step": 54270 }, { "grad_norm": 0.07767053693532944, "learning_rate": 4.70601736802123e-05, "loss": 0.0102, "step": 54280 }, { "grad_norm": 0.12940998375415802, "learning_rate": 4.704366774831391e-05, "loss": 0.025, "step": 54290 }, { "grad_norm": 0.12130244076251984, "learning_rate": 4.702716213971557e-05, "loss": 0.0107, "step": 54300 }, { "grad_norm": 0.13533900678157806, "learning_rate": 4.70106568562223e-05, "loss": 0.0117, "step": 54310 }, { "grad_norm": 0.10168176889419556, "learning_rate": 4.6994151899639105e-05, "loss": 0.0092, "step": 54320 }, { "grad_norm": 0.08267682790756226, "learning_rate": 4.697764727177093e-05, "loss": 0.0143, "step": 54330 }, { "grad_norm": 0.15206365287303925, "learning_rate": 4.696114297442272e-05, "loss": 0.0099, "step": 54340 }, { "grad_norm": 0.08420594036579132, "learning_rate": 4.694463900939933e-05, "loss": 0.008, "step": 54350 }, { "grad_norm": 0.12456194311380386, "learning_rate": 4.692813537850564e-05, "loss": 0.0098, "step": 54360 }, { "grad_norm": 0.07140886783599854, "learning_rate": 4.6911632083546445e-05, "loss": 0.008, "step": 54370 }, { "grad_norm": 0.09971826523542404, "learning_rate": 4.6895129126326525e-05, "loss": 0.0151, "step": 54380 }, { "grad_norm": 0.08103162795305252, "learning_rate": 4.687862650865059e-05, "loss": 0.008, "step": 54390 }, { "grad_norm": 0.10501670092344284, "learning_rate": 4.686212423232339e-05, "loss": 0.0124, "step": 54400 }, { "grad_norm": 0.06870319694280624, "learning_rate": 4.6845622299149565e-05, "loss": 0.0097, "step": 54410 }, { "grad_norm": 0.14812839031219482, "learning_rate": 4.682912071093374e-05, "loss": 0.0095, "step": 54420 }, { "grad_norm": 0.08969702571630478, "learning_rate": 4.681261946948052e-05, "loss": 0.0096, "step": 54430 }, { "grad_norm": 0.10522010177373886, "learning_rate": 4.679611857659443e-05, "loss": 0.0167, "step": 54440 }, { "grad_norm": 0.06187368556857109, "learning_rate": 4.6779618034079975e-05, "loss": 0.0105, "step": 54450 }, { "grad_norm": 0.1202457919716835, "learning_rate": 4.6763117843741675e-05, "loss": 0.0135, "step": 54460 }, { "grad_norm": 0.08610261976718903, "learning_rate": 4.674661800738394e-05, "loss": 0.0091, "step": 54470 }, { "grad_norm": 0.15211354196071625, "learning_rate": 4.673011852681117e-05, "loss": 0.0099, "step": 54480 }, { "grad_norm": 0.10377884656190872, "learning_rate": 4.671361940382772e-05, "loss": 0.0126, "step": 54490 }, { "grad_norm": 0.1181868389248848, "learning_rate": 4.669712064023791e-05, "loss": 0.012, "step": 54500 }, { "grad_norm": 0.13373403251171112, "learning_rate": 4.668062223784602e-05, "loss": 0.0088, "step": 54510 }, { "grad_norm": 0.11540796607732773, "learning_rate": 4.6664124198456274e-05, "loss": 0.0114, "step": 54520 }, { "grad_norm": 0.06808649748563766, "learning_rate": 4.6647626523872905e-05, "loss": 0.0113, "step": 54530 }, { "grad_norm": 0.10948681086301804, "learning_rate": 4.6631129215900066e-05, "loss": 0.0096, "step": 54540 }, { "grad_norm": 0.1059860959649086, "learning_rate": 4.661463227634187e-05, "loss": 0.0196, "step": 54550 }, { "grad_norm": 0.06967701762914658, "learning_rate": 4.659813570700238e-05, "loss": 0.0076, "step": 54560 }, { "grad_norm": 0.12257622927427292, "learning_rate": 4.658163950968568e-05, "loss": 0.0092, "step": 54570 }, { "grad_norm": 0.08334317803382874, "learning_rate": 4.6565143686195726e-05, "loss": 0.011, "step": 54580 }, { "grad_norm": 0.09644260257482529, "learning_rate": 4.654864823833649e-05, "loss": 0.0082, "step": 54590 }, { "grad_norm": 0.10145344585180283, "learning_rate": 4.65321531679119e-05, "loss": 0.0094, "step": 54600 }, { "grad_norm": 0.11522118747234344, "learning_rate": 4.6515658476725834e-05, "loss": 0.0148, "step": 54610 }, { "grad_norm": 0.10397769510746002, "learning_rate": 4.649916416658212e-05, "loss": 0.0131, "step": 54620 }, { "grad_norm": 0.1924877166748047, "learning_rate": 4.648267023928455e-05, "loss": 0.0094, "step": 54630 }, { "grad_norm": 0.11795109510421753, "learning_rate": 4.646617669663687e-05, "loss": 0.0079, "step": 54640 }, { "grad_norm": 0.08160930126905441, "learning_rate": 4.644968354044277e-05, "loss": 0.0092, "step": 54650 }, { "grad_norm": 0.12443458288908005, "learning_rate": 4.643319077250596e-05, "loss": 0.0088, "step": 54660 }, { "grad_norm": 0.07378966361284256, "learning_rate": 4.6416698394630055e-05, "loss": 0.0066, "step": 54670 }, { "grad_norm": 0.07338970899581909, "learning_rate": 4.640020640861862e-05, "loss": 0.0087, "step": 54680 }, { "grad_norm": 0.1416679471731186, "learning_rate": 4.6383714816275195e-05, "loss": 0.0141, "step": 54690 }, { "grad_norm": 0.09612210839986801, "learning_rate": 4.636722361940329e-05, "loss": 0.0117, "step": 54700 }, { "grad_norm": 0.15470172464847565, "learning_rate": 4.635073281980632e-05, "loss": 0.0103, "step": 54710 }, { "grad_norm": 0.11688365042209625, "learning_rate": 4.633424241928775e-05, "loss": 0.0135, "step": 54720 }, { "grad_norm": 0.09762097895145416, "learning_rate": 4.6317752419650914e-05, "loss": 0.0094, "step": 54730 }, { "grad_norm": 0.17569178342819214, "learning_rate": 4.630126282269914e-05, "loss": 0.012, "step": 54740 }, { "grad_norm": 0.09787074476480484, "learning_rate": 4.628477363023569e-05, "loss": 0.0087, "step": 54750 }, { "grad_norm": 0.12204486131668091, "learning_rate": 4.626828484406383e-05, "loss": 0.0098, "step": 54760 }, { "grad_norm": 0.13680647313594818, "learning_rate": 4.625179646598669e-05, "loss": 0.0132, "step": 54770 }, { "grad_norm": 0.09420043975114822, "learning_rate": 4.6235308497807487e-05, "loss": 0.0082, "step": 54780 }, { "grad_norm": 0.20182159543037415, "learning_rate": 4.6218820941329276e-05, "loss": 0.0102, "step": 54790 }, { "grad_norm": 0.09644510596990585, "learning_rate": 4.620233379835513e-05, "loss": 0.0085, "step": 54800 }, { "grad_norm": 0.720079779624939, "learning_rate": 4.618584707068804e-05, "loss": 0.0103, "step": 54810 }, { "grad_norm": 0.1814005970954895, "learning_rate": 4.616936076013099e-05, "loss": 0.0094, "step": 54820 }, { "grad_norm": 0.08921696245670319, "learning_rate": 4.615287486848688e-05, "loss": 0.0134, "step": 54830 }, { "grad_norm": 0.1939084380865097, "learning_rate": 4.6136389397558584e-05, "loss": 0.0112, "step": 54840 }, { "grad_norm": 0.11270651966333389, "learning_rate": 4.6119904349148935e-05, "loss": 0.0084, "step": 54850 }, { "grad_norm": 0.0865454226732254, "learning_rate": 4.6103419725060726e-05, "loss": 0.0079, "step": 54860 }, { "grad_norm": 0.08528583496809006, "learning_rate": 4.608693552709667e-05, "loss": 0.0085, "step": 54870 }, { "grad_norm": 0.1942417323589325, "learning_rate": 4.607045175705947e-05, "loss": 0.0125, "step": 54880 }, { "grad_norm": 0.2096753716468811, "learning_rate": 4.605396841675176e-05, "loss": 0.0109, "step": 54890 }, { "grad_norm": 0.16000184416770935, "learning_rate": 4.6037485507976094e-05, "loss": 0.0088, "step": 54900 }, { "grad_norm": 0.1293536275625229, "learning_rate": 4.602100303253509e-05, "loss": 0.008, "step": 54910 }, { "grad_norm": 0.12280440330505371, "learning_rate": 4.600452099223122e-05, "loss": 0.0138, "step": 54920 }, { "grad_norm": 0.14472831785678864, "learning_rate": 4.598803938886691e-05, "loss": 0.0087, "step": 54930 }, { "grad_norm": 0.07563156634569168, "learning_rate": 4.597155822424459e-05, "loss": 0.0065, "step": 54940 }, { "grad_norm": 0.1362244188785553, "learning_rate": 4.59550775001666e-05, "loss": 0.0129, "step": 54950 }, { "grad_norm": 0.08221267908811569, "learning_rate": 4.5938597218435234e-05, "loss": 0.0112, "step": 54960 }, { "grad_norm": 0.15209954977035522, "learning_rate": 4.592211738085279e-05, "loss": 0.017, "step": 54970 }, { "grad_norm": 0.08260732889175415, "learning_rate": 4.590563798922146e-05, "loss": 0.0083, "step": 54980 }, { "grad_norm": 0.10484954714775085, "learning_rate": 4.5889159045343404e-05, "loss": 0.0132, "step": 54990 }, { "grad_norm": 0.15704508125782013, "learning_rate": 4.5872680551020734e-05, "loss": 0.0115, "step": 55000 }, { "grad_norm": 0.17633533477783203, "learning_rate": 4.585620250805552e-05, "loss": 0.009, "step": 55010 }, { "grad_norm": 0.11469637602567673, "learning_rate": 4.583972491824975e-05, "loss": 0.0099, "step": 55020 }, { "grad_norm": 0.11771222949028015, "learning_rate": 4.582324778340541e-05, "loss": 0.0092, "step": 55030 }, { "grad_norm": 0.11669926345348358, "learning_rate": 4.5806771105324424e-05, "loss": 0.0112, "step": 55040 }, { "grad_norm": 0.0786919817328453, "learning_rate": 4.579029488580865e-05, "loss": 0.0104, "step": 55050 }, { "grad_norm": 0.08938879519701004, "learning_rate": 4.5773819126659876e-05, "loss": 0.0088, "step": 55060 }, { "grad_norm": 0.107463039457798, "learning_rate": 4.575734382967991e-05, "loss": 0.0096, "step": 55070 }, { "grad_norm": 0.10613147914409637, "learning_rate": 4.574086899667044e-05, "loss": 0.0091, "step": 55080 }, { "grad_norm": 0.10557638853788376, "learning_rate": 4.5724394629433106e-05, "loss": 0.0088, "step": 55090 }, { "grad_norm": 0.2552609145641327, "learning_rate": 4.570792072976958e-05, "loss": 0.0103, "step": 55100 }, { "grad_norm": 0.07488758116960526, "learning_rate": 4.569144729948138e-05, "loss": 0.0109, "step": 55110 }, { "grad_norm": 0.11471119523048401, "learning_rate": 4.5674974340370033e-05, "loss": 0.0083, "step": 55120 }, { "grad_norm": 0.10173806548118591, "learning_rate": 4.565850185423699e-05, "loss": 0.0116, "step": 55130 }, { "grad_norm": 0.1413361132144928, "learning_rate": 4.564202984288365e-05, "loss": 0.0132, "step": 55140 }, { "grad_norm": 0.12460115551948547, "learning_rate": 4.5625558308111356e-05, "loss": 0.0096, "step": 55150 }, { "grad_norm": 0.10335417836904526, "learning_rate": 4.560908725172145e-05, "loss": 0.0095, "step": 55160 }, { "grad_norm": 0.111023910343647, "learning_rate": 4.559261667551516e-05, "loss": 0.0085, "step": 55170 }, { "grad_norm": 0.11101797968149185, "learning_rate": 4.5576146581293685e-05, "loss": 0.0132, "step": 55180 }, { "grad_norm": 0.23173603415489197, "learning_rate": 4.5559676970858164e-05, "loss": 0.0096, "step": 55190 }, { "grad_norm": 0.22090333700180054, "learning_rate": 4.5543207846009705e-05, "loss": 0.0145, "step": 55200 }, { "grad_norm": 0.07365138828754425, "learning_rate": 4.552673920854933e-05, "loss": 0.0117, "step": 55210 }, { "grad_norm": 0.1994144767522812, "learning_rate": 4.5510271060278006e-05, "loss": 0.0145, "step": 55220 }, { "grad_norm": 0.09604348987340927, "learning_rate": 4.5493803402996704e-05, "loss": 0.0164, "step": 55230 }, { "grad_norm": 0.12625692784786224, "learning_rate": 4.5477336238506295e-05, "loss": 0.0099, "step": 55240 }, { "grad_norm": 0.08689260482788086, "learning_rate": 4.546086956860759e-05, "loss": 0.0095, "step": 55250 }, { "grad_norm": 0.07319135218858719, "learning_rate": 4.5444403395101374e-05, "loss": 0.0077, "step": 55260 }, { "grad_norm": 0.18645332753658295, "learning_rate": 4.542793771978834e-05, "loss": 0.0125, "step": 55270 }, { "grad_norm": 0.10833707451820374, "learning_rate": 4.541147254446916e-05, "loss": 0.0118, "step": 55280 }, { "grad_norm": 0.08063340932130814, "learning_rate": 4.539500787094445e-05, "loss": 0.0168, "step": 55290 }, { "grad_norm": 0.1239449605345726, "learning_rate": 4.537854370101475e-05, "loss": 0.0094, "step": 55300 }, { "grad_norm": 0.09660361707210541, "learning_rate": 4.5362080036480566e-05, "loss": 0.0086, "step": 55310 }, { "grad_norm": 0.11706873774528503, "learning_rate": 4.534561687914235e-05, "loss": 0.0109, "step": 55320 }, { "grad_norm": 0.179949089884758, "learning_rate": 4.5329154230800466e-05, "loss": 0.0112, "step": 55330 }, { "grad_norm": 0.09558986872434616, "learning_rate": 4.531269209325525e-05, "loss": 0.0098, "step": 55340 }, { "grad_norm": 0.13743506371974945, "learning_rate": 4.529623046830696e-05, "loss": 0.0097, "step": 55350 }, { "grad_norm": 0.20277869701385498, "learning_rate": 4.5279769357755866e-05, "loss": 0.0097, "step": 55360 }, { "grad_norm": 0.1478448063135147, "learning_rate": 4.5263308763402084e-05, "loss": 0.009, "step": 55370 }, { "grad_norm": 0.09135451912879944, "learning_rate": 4.524684868704574e-05, "loss": 0.0102, "step": 55380 }, { "grad_norm": 0.08430926501750946, "learning_rate": 4.523038913048688e-05, "loss": 0.0099, "step": 55390 }, { "grad_norm": 0.07266629487276077, "learning_rate": 4.521393009552549e-05, "loss": 0.0076, "step": 55400 }, { "grad_norm": 0.11454498022794724, "learning_rate": 4.5197471583961484e-05, "loss": 0.0133, "step": 55410 }, { "grad_norm": 0.09272274374961853, "learning_rate": 4.51810135975948e-05, "loss": 0.0138, "step": 55420 }, { "grad_norm": 0.4941570460796356, "learning_rate": 4.516455613822521e-05, "loss": 0.0079, "step": 55430 }, { "grad_norm": 0.08944126963615417, "learning_rate": 4.514809920765249e-05, "loss": 0.0131, "step": 55440 }, { "grad_norm": 0.15250809490680695, "learning_rate": 4.5131642807676356e-05, "loss": 0.0085, "step": 55450 }, { "grad_norm": 0.07513301819562912, "learning_rate": 4.511518694009644e-05, "loss": 0.009, "step": 55460 }, { "grad_norm": 0.09366878122091293, "learning_rate": 4.509873160671231e-05, "loss": 0.0105, "step": 55470 }, { "grad_norm": 0.11934559792280197, "learning_rate": 4.508227680932356e-05, "loss": 0.0085, "step": 55480 }, { "grad_norm": 0.07829441875219345, "learning_rate": 4.5065822549729605e-05, "loss": 0.0118, "step": 55490 }, { "grad_norm": 0.08933686465024948, "learning_rate": 4.504936882972989e-05, "loss": 0.0091, "step": 55500 }, { "grad_norm": 0.15783053636550903, "learning_rate": 4.503291565112374e-05, "loss": 0.0084, "step": 55510 }, { "grad_norm": 0.07210245728492737, "learning_rate": 4.501646301571048e-05, "loss": 0.0095, "step": 55520 }, { "grad_norm": 0.07790371030569077, "learning_rate": 4.500001092528933e-05, "loss": 0.0114, "step": 55530 }, { "grad_norm": 0.3838373124599457, "learning_rate": 4.4983559381659455e-05, "loss": 0.0102, "step": 55540 }, { "grad_norm": 0.13889649510383606, "learning_rate": 4.496710838661999e-05, "loss": 0.0091, "step": 55550 }, { "grad_norm": 0.1026957631111145, "learning_rate": 4.495065794196999e-05, "loss": 0.0107, "step": 55560 }, { "grad_norm": 0.10199332237243652, "learning_rate": 4.4934208049508455e-05, "loss": 0.008, "step": 55570 }, { "grad_norm": 0.16991344094276428, "learning_rate": 4.4917758711034303e-05, "loss": 0.0123, "step": 55580 }, { "grad_norm": 0.10477853566408157, "learning_rate": 4.490130992834642e-05, "loss": 0.0104, "step": 55590 }, { "grad_norm": 0.16858811676502228, "learning_rate": 4.488486170324359e-05, "loss": 0.0129, "step": 55600 }, { "grad_norm": 0.1051110178232193, "learning_rate": 4.486841403752463e-05, "loss": 0.0082, "step": 55610 }, { "grad_norm": 0.07288963347673416, "learning_rate": 4.4851966932988174e-05, "loss": 0.0088, "step": 55620 }, { "grad_norm": 0.12745323777198792, "learning_rate": 4.4835520391432894e-05, "loss": 0.011, "step": 55630 }, { "grad_norm": 0.08367404341697693, "learning_rate": 4.481907441465733e-05, "loss": 0.009, "step": 55640 }, { "grad_norm": 0.2521665394306183, "learning_rate": 4.480262900446e-05, "loss": 0.0114, "step": 55650 }, { "grad_norm": 0.13020431995391846, "learning_rate": 4.478618416263933e-05, "loss": 0.0098, "step": 55660 }, { "grad_norm": 0.1399562954902649, "learning_rate": 4.476973989099373e-05, "loss": 0.0094, "step": 55670 }, { "grad_norm": 0.07572420686483383, "learning_rate": 4.475329619132153e-05, "loss": 0.0143, "step": 55680 }, { "grad_norm": 0.10885932296514511, "learning_rate": 4.4736853065420956e-05, "loss": 0.0115, "step": 55690 }, { "grad_norm": 0.07716762274503708, "learning_rate": 4.472041051509023e-05, "loss": 0.0096, "step": 55700 }, { "grad_norm": 0.14391639828681946, "learning_rate": 4.470396854212746e-05, "loss": 0.0091, "step": 55710 }, { "grad_norm": 0.05605705454945564, "learning_rate": 4.468752714833072e-05, "loss": 0.0076, "step": 55720 }, { "grad_norm": 0.07378008961677551, "learning_rate": 4.467108633549803e-05, "loss": 0.0083, "step": 55730 }, { "grad_norm": 0.0947333499789238, "learning_rate": 4.4654646105427335e-05, "loss": 0.0153, "step": 55740 }, { "grad_norm": 0.08913413435220718, "learning_rate": 4.463820645991651e-05, "loss": 0.0082, "step": 55750 }, { "grad_norm": 0.12692371010780334, "learning_rate": 4.462176740076334e-05, "loss": 0.0079, "step": 55760 }, { "grad_norm": 0.1349647492170334, "learning_rate": 4.4605328929765624e-05, "loss": 0.0135, "step": 55770 }, { "grad_norm": 0.10529651492834091, "learning_rate": 4.458889104872102e-05, "loss": 0.0073, "step": 55780 }, { "grad_norm": 0.1135210394859314, "learning_rate": 4.457245375942715e-05, "loss": 0.0136, "step": 55790 }, { "grad_norm": 0.13100168108940125, "learning_rate": 4.455601706368157e-05, "loss": 0.0112, "step": 55800 }, { "grad_norm": 0.09762947261333466, "learning_rate": 4.453958096328179e-05, "loss": 0.0082, "step": 55810 }, { "grad_norm": 0.14055469632148743, "learning_rate": 4.452314546002523e-05, "loss": 0.0086, "step": 55820 }, { "grad_norm": 0.06639055907726288, "learning_rate": 4.4506710555709246e-05, "loss": 0.0086, "step": 55830 }, { "grad_norm": 0.11646538227796555, "learning_rate": 4.449027625213114e-05, "loss": 0.011, "step": 55840 }, { "grad_norm": 0.06851614266633987, "learning_rate": 4.4473842551088135e-05, "loss": 0.0079, "step": 55850 }, { "grad_norm": 0.10454846918582916, "learning_rate": 4.445740945437738e-05, "loss": 0.0088, "step": 55860 }, { "grad_norm": 0.08263284713029861, "learning_rate": 4.444097696379602e-05, "loss": 0.0093, "step": 55870 }, { "grad_norm": 0.10251130163669586, "learning_rate": 4.442454508114106e-05, "loss": 0.0103, "step": 55880 }, { "grad_norm": 0.09068969637155533, "learning_rate": 4.4408113808209465e-05, "loss": 0.0103, "step": 55890 }, { "grad_norm": 0.0795305073261261, "learning_rate": 4.439168314679813e-05, "loss": 0.0092, "step": 55900 }, { "grad_norm": 0.07469450682401657, "learning_rate": 4.43752530987039e-05, "loss": 0.0095, "step": 55910 }, { "grad_norm": 0.12014075368642807, "learning_rate": 4.4358823665723515e-05, "loss": 0.0079, "step": 55920 }, { "grad_norm": 0.11704084277153015, "learning_rate": 4.434239484965371e-05, "loss": 0.01, "step": 55930 }, { "grad_norm": 0.0759599581360817, "learning_rate": 4.4325966652291103e-05, "loss": 0.0105, "step": 55940 }, { "grad_norm": 0.0674363523721695, "learning_rate": 4.430953907543225e-05, "loss": 0.0092, "step": 55950 }, { "grad_norm": 0.10494726896286011, "learning_rate": 4.4293112120873645e-05, "loss": 0.0089, "step": 55960 }, { "grad_norm": 0.06929574906826019, "learning_rate": 4.427668579041172e-05, "loss": 0.0065, "step": 55970 }, { "grad_norm": 0.1134333461523056, "learning_rate": 4.4260260085842816e-05, "loss": 0.0103, "step": 55980 }, { "grad_norm": 0.08093876391649246, "learning_rate": 4.4243835008963256e-05, "loss": 0.0097, "step": 55990 }, { "grad_norm": 0.07263403385877609, "learning_rate": 4.422741056156925e-05, "loss": 0.0067, "step": 56000 }, { "grad_norm": 0.06981468945741653, "learning_rate": 4.421098674545693e-05, "loss": 0.0078, "step": 56010 }, { "grad_norm": 0.12420257180929184, "learning_rate": 4.419456356242241e-05, "loss": 0.0119, "step": 56020 }, { "grad_norm": 0.07078056037425995, "learning_rate": 4.4178141014261685e-05, "loss": 0.0104, "step": 56030 }, { "grad_norm": 0.14933554828166962, "learning_rate": 4.416171910277071e-05, "loss": 0.0094, "step": 56040 }, { "grad_norm": 0.13819915056228638, "learning_rate": 4.414529782974533e-05, "loss": 0.0088, "step": 56050 }, { "grad_norm": 0.07107599079608917, "learning_rate": 4.41288771969814e-05, "loss": 0.0112, "step": 56060 }, { "grad_norm": 0.29884177446365356, "learning_rate": 4.411245720627464e-05, "loss": 0.0112, "step": 56070 }, { "grad_norm": 0.13869169354438782, "learning_rate": 4.40960378594207e-05, "loss": 0.0114, "step": 56080 }, { "grad_norm": 0.08170995116233826, "learning_rate": 4.407961915821519e-05, "loss": 0.0101, "step": 56090 }, { "grad_norm": 0.06732628494501114, "learning_rate": 4.406320110445362e-05, "loss": 0.0087, "step": 56100 }, { "grad_norm": 0.0845717340707779, "learning_rate": 4.404678369993144e-05, "loss": 0.0071, "step": 56110 }, { "grad_norm": 0.3889535665512085, "learning_rate": 4.403036694644406e-05, "loss": 0.011, "step": 56120 }, { "grad_norm": 0.15120941400527954, "learning_rate": 4.4013950845786764e-05, "loss": 0.0083, "step": 56130 }, { "grad_norm": 0.2302733212709427, "learning_rate": 4.399753539975482e-05, "loss": 0.0116, "step": 56140 }, { "grad_norm": 0.10214262455701828, "learning_rate": 4.398112061014337e-05, "loss": 0.0077, "step": 56150 }, { "grad_norm": 0.14125187695026398, "learning_rate": 4.396470647874753e-05, "loss": 0.0092, "step": 56160 }, { "grad_norm": 0.06573652476072311, "learning_rate": 4.394829300736229e-05, "loss": 0.0081, "step": 56170 }, { "grad_norm": 0.11476192623376846, "learning_rate": 4.393188019778265e-05, "loss": 0.008, "step": 56180 }, { "grad_norm": 0.10095246136188507, "learning_rate": 4.391546805180347e-05, "loss": 0.0089, "step": 56190 }, { "grad_norm": 0.10479313135147095, "learning_rate": 4.389905657121955e-05, "loss": 0.0091, "step": 56200 }, { "grad_norm": 0.15662981569766998, "learning_rate": 4.3882645757825635e-05, "loss": 0.0108, "step": 56210 }, { "grad_norm": 0.11725171655416489, "learning_rate": 4.386623561341637e-05, "loss": 0.0077, "step": 56220 }, { "grad_norm": 0.08094992488622665, "learning_rate": 4.384982613978637e-05, "loss": 0.0102, "step": 56230 }, { "grad_norm": 0.16911107301712036, "learning_rate": 4.383341733873012e-05, "loss": 0.0104, "step": 56240 }, { "grad_norm": 0.10666953027248383, "learning_rate": 4.381700921204209e-05, "loss": 0.0084, "step": 56250 }, { "grad_norm": 0.18062803149223328, "learning_rate": 4.380060176151663e-05, "loss": 0.0108, "step": 56260 }, { "grad_norm": 0.11891214549541473, "learning_rate": 4.378419498894805e-05, "loss": 0.0122, "step": 56270 }, { "grad_norm": 0.13403740525245667, "learning_rate": 4.376778889613056e-05, "loss": 0.0104, "step": 56280 }, { "grad_norm": 0.1019006073474884, "learning_rate": 4.375138348485831e-05, "loss": 0.0103, "step": 56290 }, { "grad_norm": 0.11422418802976608, "learning_rate": 4.3734978756925346e-05, "loss": 0.0073, "step": 56300 }, { "grad_norm": 0.06680507957935333, "learning_rate": 4.3718574714125706e-05, "loss": 0.0087, "step": 56310 }, { "grad_norm": 0.331333190202713, "learning_rate": 4.370217135825329e-05, "loss": 0.0087, "step": 56320 }, { "grad_norm": 0.08366131037473679, "learning_rate": 4.368576869110194e-05, "loss": 0.008, "step": 56330 }, { "grad_norm": 0.10362832993268967, "learning_rate": 4.366936671446544e-05, "loss": 0.0085, "step": 56340 }, { "grad_norm": 0.12409587949514389, "learning_rate": 4.3652965430137474e-05, "loss": 0.0116, "step": 56350 }, { "grad_norm": 0.09570877999067307, "learning_rate": 4.3636564839911646e-05, "loss": 0.0084, "step": 56360 }, { "grad_norm": 0.11874457448720932, "learning_rate": 4.3620164945581545e-05, "loss": 0.0064, "step": 56370 }, { "grad_norm": 0.09287052601575851, "learning_rate": 4.360376574894061e-05, "loss": 0.0098, "step": 56380 }, { "grad_norm": 0.053866930305957794, "learning_rate": 4.358736725178224e-05, "loss": 0.0094, "step": 56390 }, { "grad_norm": 0.15233008563518524, "learning_rate": 4.357096945589974e-05, "loss": 0.0129, "step": 56400 }, { "grad_norm": 0.10461459308862686, "learning_rate": 4.355457236308636e-05, "loss": 0.0092, "step": 56410 }, { "grad_norm": 0.11789380013942719, "learning_rate": 4.353817597513526e-05, "loss": 0.0074, "step": 56420 }, { "grad_norm": 0.11583118140697479, "learning_rate": 4.352178029383948e-05, "loss": 0.0069, "step": 56430 }, { "grad_norm": 0.17948998510837555, "learning_rate": 4.3505385320992105e-05, "loss": 0.01, "step": 56440 }, { "grad_norm": 0.10404776781797409, "learning_rate": 4.348899105838602e-05, "loss": 0.0085, "step": 56450 }, { "grad_norm": 0.09781016409397125, "learning_rate": 4.3472597507814087e-05, "loss": 0.0094, "step": 56460 }, { "grad_norm": 0.12864255905151367, "learning_rate": 4.3456204671069066e-05, "loss": 0.0109, "step": 56470 }, { "grad_norm": 0.09289883822202682, "learning_rate": 4.343981254994367e-05, "loss": 0.0084, "step": 56480 }, { "grad_norm": 0.10875638574361801, "learning_rate": 4.34234211462305e-05, "loss": 0.01, "step": 56490 }, { "grad_norm": 0.08500020205974579, "learning_rate": 4.340703046172213e-05, "loss": 0.0117, "step": 56500 }, { "grad_norm": 0.07405742257833481, "learning_rate": 4.339064049821097e-05, "loss": 0.0085, "step": 56510 }, { "grad_norm": 0.09449878334999084, "learning_rate": 4.3374251257489446e-05, "loss": 0.0124, "step": 56520 }, { "grad_norm": 0.08459426462650299, "learning_rate": 4.335786274134984e-05, "loss": 0.0089, "step": 56530 }, { "grad_norm": 0.2584135830402374, "learning_rate": 4.334147495158438e-05, "loss": 0.0206, "step": 56540 }, { "grad_norm": 0.19954557716846466, "learning_rate": 4.33250878899852e-05, "loss": 0.0113, "step": 56550 }, { "grad_norm": 0.13590818643569946, "learning_rate": 4.330870155834437e-05, "loss": 0.0095, "step": 56560 }, { "grad_norm": 0.11347675323486328, "learning_rate": 4.3292315958453884e-05, "loss": 0.0092, "step": 56570 }, { "grad_norm": 0.11872760206460953, "learning_rate": 4.3275931092105646e-05, "loss": 0.0097, "step": 56580 }, { "grad_norm": 0.1040973961353302, "learning_rate": 4.325954696109147e-05, "loss": 0.0072, "step": 56590 }, { "grad_norm": 0.1216101199388504, "learning_rate": 4.32431635672031e-05, "loss": 0.0103, "step": 56600 }, { "grad_norm": 0.11436421424150467, "learning_rate": 4.3226780912232214e-05, "loss": 0.0157, "step": 56610 }, { "grad_norm": 0.11149685829877853, "learning_rate": 4.321039899797036e-05, "loss": 0.0083, "step": 56620 }, { "grad_norm": 0.1349642127752304, "learning_rate": 4.319401782620908e-05, "loss": 0.007, "step": 56630 }, { "grad_norm": 0.06854542344808578, "learning_rate": 4.317763739873978e-05, "loss": 0.0109, "step": 56640 }, { "grad_norm": 0.22291487455368042, "learning_rate": 4.31612577173538e-05, "loss": 0.0092, "step": 56650 }, { "grad_norm": 0.11448418349027634, "learning_rate": 4.314487878384239e-05, "loss": 0.0178, "step": 56660 }, { "grad_norm": 0.2532821595668793, "learning_rate": 4.3128500599996736e-05, "loss": 0.0138, "step": 56670 }, { "grad_norm": 0.1139758825302124, "learning_rate": 4.31121231676079e-05, "loss": 0.0084, "step": 56680 }, { "grad_norm": 0.12524209916591644, "learning_rate": 4.309574648846694e-05, "loss": 0.0106, "step": 56690 }, { "grad_norm": 0.1129000186920166, "learning_rate": 4.3079370564364755e-05, "loss": 0.0133, "step": 56700 }, { "grad_norm": 0.14334708452224731, "learning_rate": 4.306299539709221e-05, "loss": 0.0073, "step": 56710 }, { "grad_norm": 0.1198858916759491, "learning_rate": 4.3046620988440046e-05, "loss": 0.0102, "step": 56720 }, { "grad_norm": 0.0993010550737381, "learning_rate": 4.303024734019897e-05, "loss": 0.0071, "step": 56730 }, { "grad_norm": 0.08480726927518845, "learning_rate": 4.3013874454159565e-05, "loss": 0.0095, "step": 56740 }, { "grad_norm": 0.07061424851417542, "learning_rate": 4.299750233211233e-05, "loss": 0.0127, "step": 56750 }, { "grad_norm": 0.053073447197675705, "learning_rate": 4.2981130975847715e-05, "loss": 0.0069, "step": 56760 }, { "grad_norm": 0.10651955753564835, "learning_rate": 4.2964760387156076e-05, "loss": 0.0096, "step": 56770 }, { "grad_norm": 0.1641833782196045, "learning_rate": 4.294839056782767e-05, "loss": 0.0075, "step": 56780 }, { "grad_norm": 0.08491162210702896, "learning_rate": 4.293202151965266e-05, "loss": 0.0121, "step": 56790 }, { "grad_norm": 0.12812113761901855, "learning_rate": 4.291565324442115e-05, "loss": 0.0086, "step": 56800 }, { "grad_norm": 0.12942202389240265, "learning_rate": 4.2899285743923135e-05, "loss": 0.012, "step": 56810 }, { "grad_norm": 0.07966378331184387, "learning_rate": 4.288291901994857e-05, "loss": 0.0083, "step": 56820 }, { "grad_norm": 0.05681253597140312, "learning_rate": 4.286655307428729e-05, "loss": 0.0077, "step": 56830 }, { "grad_norm": 0.09927232563495636, "learning_rate": 4.285018790872903e-05, "loss": 0.0117, "step": 56840 }, { "grad_norm": 0.13117091357707977, "learning_rate": 4.283382352506347e-05, "loss": 0.0095, "step": 56850 }, { "grad_norm": 0.1129906177520752, "learning_rate": 4.2817459925080205e-05, "loss": 0.0148, "step": 56860 }, { "grad_norm": 0.11601254343986511, "learning_rate": 4.28010971105687e-05, "loss": 0.0092, "step": 56870 }, { "grad_norm": 0.19673162698745728, "learning_rate": 4.278473508331841e-05, "loss": 0.0074, "step": 56880 }, { "grad_norm": 0.1740909367799759, "learning_rate": 4.276837384511864e-05, "loss": 0.0118, "step": 56890 }, { "grad_norm": 0.0992249920964241, "learning_rate": 4.275201339775864e-05, "loss": 0.0085, "step": 56900 }, { "grad_norm": 0.10832412540912628, "learning_rate": 4.273565374302756e-05, "loss": 0.0115, "step": 56910 }, { "grad_norm": 0.08790691196918488, "learning_rate": 4.2719294882714454e-05, "loss": 0.0086, "step": 56920 }, { "grad_norm": 0.0936090275645256, "learning_rate": 4.270293681860831e-05, "loss": 0.009, "step": 56930 }, { "grad_norm": 0.06071518734097481, "learning_rate": 4.268657955249802e-05, "loss": 0.0076, "step": 56940 }, { "grad_norm": 0.17287695407867432, "learning_rate": 4.267022308617241e-05, "loss": 0.0078, "step": 56950 }, { "grad_norm": 0.1290295571088791, "learning_rate": 4.265386742142018e-05, "loss": 0.0085, "step": 56960 }, { "grad_norm": 0.13388122618198395, "learning_rate": 4.263751256002996e-05, "loss": 0.0111, "step": 56970 }, { "grad_norm": 0.1197488009929657, "learning_rate": 4.26211585037903e-05, "loss": 0.0109, "step": 56980 }, { "grad_norm": 0.14208664000034332, "learning_rate": 4.2604805254489656e-05, "loss": 0.0098, "step": 56990 }, { "grad_norm": 0.09380112588405609, "learning_rate": 4.258845281391638e-05, "loss": 0.0089, "step": 57000 }, { "grad_norm": 0.12505480647087097, "learning_rate": 4.257210118385876e-05, "loss": 0.0173, "step": 57010 }, { "grad_norm": 0.11005298048257828, "learning_rate": 4.2555750366104994e-05, "loss": 0.0123, "step": 57020 }, { "grad_norm": 0.10269795358181, "learning_rate": 4.2539400362443184e-05, "loss": 0.007, "step": 57030 }, { "grad_norm": 0.12113458663225174, "learning_rate": 4.252305117466133e-05, "loss": 0.0093, "step": 57040 }, { "grad_norm": 0.3074793815612793, "learning_rate": 4.2506702804547356e-05, "loss": 0.0127, "step": 57050 }, { "grad_norm": 0.06812845170497894, "learning_rate": 4.249035525388907e-05, "loss": 0.008, "step": 57060 }, { "grad_norm": 0.07072378695011139, "learning_rate": 4.247400852447427e-05, "loss": 0.0064, "step": 57070 }, { "grad_norm": 0.09133630990982056, "learning_rate": 4.245766261809059e-05, "loss": 0.008, "step": 57080 }, { "grad_norm": 0.08633018285036087, "learning_rate": 4.244131753652558e-05, "loss": 0.0098, "step": 57090 }, { "grad_norm": 0.1942128837108612, "learning_rate": 4.242497328156672e-05, "loss": 0.0077, "step": 57100 }, { "grad_norm": 0.09443029016256332, "learning_rate": 4.240862985500139e-05, "loss": 0.0133, "step": 57110 }, { "grad_norm": 0.11382681876420975, "learning_rate": 4.2392287258616894e-05, "loss": 0.0066, "step": 57120 }, { "grad_norm": 0.13654129207134247, "learning_rate": 4.2375945494200396e-05, "loss": 0.0065, "step": 57130 }, { "grad_norm": 0.11180204153060913, "learning_rate": 4.2359604563539054e-05, "loss": 0.0087, "step": 57140 }, { "grad_norm": 0.13472934067249298, "learning_rate": 4.2343264468419885e-05, "loss": 0.0101, "step": 57150 }, { "grad_norm": 0.06856672465801239, "learning_rate": 4.232692521062979e-05, "loss": 0.0063, "step": 57160 }, { "grad_norm": 0.08883081376552582, "learning_rate": 4.231058679195562e-05, "loss": 0.0086, "step": 57170 }, { "grad_norm": 0.061857134103775024, "learning_rate": 4.229424921418411e-05, "loss": 0.009, "step": 57180 }, { "grad_norm": 0.12957380712032318, "learning_rate": 4.2277912479101916e-05, "loss": 0.0119, "step": 57190 }, { "grad_norm": 0.10771837085485458, "learning_rate": 4.226157658849562e-05, "loss": 0.0084, "step": 57200 }, { "grad_norm": 0.10382237285375595, "learning_rate": 4.2245241544151664e-05, "loss": 0.0088, "step": 57210 }, { "grad_norm": 0.07561373710632324, "learning_rate": 4.2228907347856425e-05, "loss": 0.007, "step": 57220 }, { "grad_norm": 0.09895452111959457, "learning_rate": 4.2212574001396214e-05, "loss": 0.0088, "step": 57230 }, { "grad_norm": 0.1380269080400467, "learning_rate": 4.21962415065572e-05, "loss": 0.0095, "step": 57240 }, { "grad_norm": 0.2597120404243469, "learning_rate": 4.217990986512548e-05, "loss": 0.0112, "step": 57250 }, { "grad_norm": 0.11306583881378174, "learning_rate": 4.216357907888704e-05, "loss": 0.0154, "step": 57260 }, { "grad_norm": 0.09854190051555634, "learning_rate": 4.2147249149627824e-05, "loss": 0.01, "step": 57270 }, { "grad_norm": 0.07890378683805466, "learning_rate": 4.213092007913364e-05, "loss": 0.0105, "step": 57280 }, { "grad_norm": 0.08769223839044571, "learning_rate": 4.2114591869190205e-05, "loss": 0.0119, "step": 57290 }, { "grad_norm": 0.10167870670557022, "learning_rate": 4.209826452158315e-05, "loss": 0.0133, "step": 57300 }, { "grad_norm": 0.18860003352165222, "learning_rate": 4.2081938038098e-05, "loss": 0.01, "step": 57310 }, { "grad_norm": 0.15144133567810059, "learning_rate": 4.2065612420520185e-05, "loss": 0.0107, "step": 57320 }, { "grad_norm": 0.0844540074467659, "learning_rate": 4.204928767063509e-05, "loss": 0.0083, "step": 57330 }, { "grad_norm": 0.08428951352834702, "learning_rate": 4.2032963790227944e-05, "loss": 0.009, "step": 57340 }, { "grad_norm": 0.09799986332654953, "learning_rate": 4.2016640781083887e-05, "loss": 0.0077, "step": 57350 }, { "grad_norm": 0.11435892432928085, "learning_rate": 4.2000318644988004e-05, "loss": 0.0167, "step": 57360 }, { "grad_norm": 0.09790652990341187, "learning_rate": 4.198399738372524e-05, "loss": 0.011, "step": 57370 }, { "grad_norm": 0.08044960349798203, "learning_rate": 4.196767699908045e-05, "loss": 0.0085, "step": 57380 }, { "grad_norm": 0.13702067732810974, "learning_rate": 4.195135749283845e-05, "loss": 0.0086, "step": 57390 }, { "grad_norm": 0.07533449679613113, "learning_rate": 4.1935038866783894e-05, "loss": 0.0096, "step": 57400 }, { "grad_norm": 0.08333614468574524, "learning_rate": 4.1918721122701355e-05, "loss": 0.0094, "step": 57410 }, { "grad_norm": 0.18369810283184052, "learning_rate": 4.190240426237533e-05, "loss": 0.0122, "step": 57420 }, { "grad_norm": 0.08854595571756363, "learning_rate": 4.188608828759019e-05, "loss": 0.0078, "step": 57430 }, { "grad_norm": 0.0689253881573677, "learning_rate": 4.1869773200130246e-05, "loss": 0.0167, "step": 57440 }, { "grad_norm": 0.14028425514698029, "learning_rate": 4.185345900177967e-05, "loss": 0.0138, "step": 57450 }, { "grad_norm": 0.08501484990119934, "learning_rate": 4.183714569432258e-05, "loss": 0.0122, "step": 57460 }, { "grad_norm": 0.053058553487062454, "learning_rate": 4.182083327954295e-05, "loss": 0.0087, "step": 57470 }, { "grad_norm": 0.07163447141647339, "learning_rate": 4.1804521759224714e-05, "loss": 0.0103, "step": 57480 }, { "grad_norm": 0.06889350712299347, "learning_rate": 4.178821113515165e-05, "loss": 0.0094, "step": 57490 }, { "grad_norm": 0.08869127184152603, "learning_rate": 4.1771901409107476e-05, "loss": 0.0106, "step": 57500 }, { "grad_norm": 0.08206762373447418, "learning_rate": 4.175559258287577e-05, "loss": 0.0122, "step": 57510 }, { "grad_norm": 0.09390144050121307, "learning_rate": 4.173928465824008e-05, "loss": 0.0107, "step": 57520 }, { "grad_norm": 0.0932479202747345, "learning_rate": 4.17229776369838e-05, "loss": 0.0097, "step": 57530 }, { "grad_norm": 0.08866164088249207, "learning_rate": 4.170667152089025e-05, "loss": 0.0157, "step": 57540 }, { "grad_norm": 0.1817389875650406, "learning_rate": 4.1690366311742626e-05, "loss": 0.0076, "step": 57550 }, { "grad_norm": 0.07122845202684402, "learning_rate": 4.1674062011324055e-05, "loss": 0.0062, "step": 57560 }, { "grad_norm": 0.07357130944728851, "learning_rate": 4.165775862141752e-05, "loss": 0.0081, "step": 57570 }, { "grad_norm": 0.12540219724178314, "learning_rate": 4.164145614380598e-05, "loss": 0.0095, "step": 57580 }, { "grad_norm": 0.11091852933168411, "learning_rate": 4.162515458027223e-05, "loss": 0.01, "step": 57590 }, { "grad_norm": 0.07783021777868271, "learning_rate": 4.160885393259899e-05, "loss": 0.0102, "step": 57600 }, { "grad_norm": 0.0945575013756752, "learning_rate": 4.159255420256886e-05, "loss": 0.0097, "step": 57610 }, { "grad_norm": 0.08681576699018478, "learning_rate": 4.157625539196436e-05, "loss": 0.0091, "step": 57620 }, { "grad_norm": 0.0751522108912468, "learning_rate": 4.155995750256791e-05, "loss": 0.0146, "step": 57630 }, { "grad_norm": 0.12447579950094223, "learning_rate": 4.15436605361618e-05, "loss": 0.0086, "step": 57640 }, { "grad_norm": 0.16811531782150269, "learning_rate": 4.152736449452827e-05, "loss": 0.0128, "step": 57650 }, { "grad_norm": 0.16306918859481812, "learning_rate": 4.151106937944943e-05, "loss": 0.009, "step": 57660 }, { "grad_norm": 0.06740094721317291, "learning_rate": 4.149477519270727e-05, "loss": 0.0084, "step": 57670 }, { "grad_norm": 0.10477152466773987, "learning_rate": 4.147848193608371e-05, "loss": 0.0083, "step": 57680 }, { "grad_norm": 0.09617683291435242, "learning_rate": 4.146218961136056e-05, "loss": 0.0087, "step": 57690 }, { "grad_norm": 0.08073197305202484, "learning_rate": 4.14458982203195e-05, "loss": 0.0089, "step": 57700 }, { "grad_norm": 0.31277015805244446, "learning_rate": 4.1429607764742165e-05, "loss": 0.0074, "step": 57710 }, { "grad_norm": 0.07341092079877853, "learning_rate": 4.141331824641003e-05, "loss": 0.0072, "step": 57720 }, { "grad_norm": 0.10363585501909256, "learning_rate": 4.139702966710452e-05, "loss": 0.0069, "step": 57730 }, { "grad_norm": 0.135690838098526, "learning_rate": 4.13807420286069e-05, "loss": 0.009, "step": 57740 }, { "grad_norm": 0.09847906976938248, "learning_rate": 4.136445533269837e-05, "loss": 0.0092, "step": 57750 }, { "grad_norm": 0.09226787090301514, "learning_rate": 4.134816958116003e-05, "loss": 0.0083, "step": 57760 }, { "grad_norm": 0.19382742047309875, "learning_rate": 4.133188477577283e-05, "loss": 0.011, "step": 57770 }, { "grad_norm": 0.17006732523441315, "learning_rate": 4.131560091831771e-05, "loss": 0.0082, "step": 57780 }, { "grad_norm": 0.09057478606700897, "learning_rate": 4.1299318010575405e-05, "loss": 0.008, "step": 57790 }, { "grad_norm": 0.08375269919633865, "learning_rate": 4.128303605432661e-05, "loss": 0.0245, "step": 57800 }, { "grad_norm": 0.18332339823246002, "learning_rate": 4.126675505135188e-05, "loss": 0.0102, "step": 57810 }, { "grad_norm": 0.11742369085550308, "learning_rate": 4.125047500343169e-05, "loss": 0.0091, "step": 57820 }, { "grad_norm": 0.1148061528801918, "learning_rate": 4.1234195912346376e-05, "loss": 0.0094, "step": 57830 }, { "grad_norm": 0.10266222804784775, "learning_rate": 4.1217917779876235e-05, "loss": 0.0095, "step": 57840 }, { "grad_norm": 0.07674866914749146, "learning_rate": 4.120164060780141e-05, "loss": 0.0158, "step": 57850 }, { "grad_norm": 0.1333809345960617, "learning_rate": 4.1185364397901935e-05, "loss": 0.0124, "step": 57860 }, { "grad_norm": 0.0736469179391861, "learning_rate": 4.116908915195776e-05, "loss": 0.0085, "step": 57870 }, { "grad_norm": 0.15514957904815674, "learning_rate": 4.115281487174872e-05, "loss": 0.0121, "step": 57880 }, { "grad_norm": 0.08140132576227188, "learning_rate": 4.113654155905452e-05, "loss": 0.0115, "step": 57890 }, { "grad_norm": 0.07349856942892075, "learning_rate": 4.112026921565483e-05, "loss": 0.0094, "step": 57900 }, { "grad_norm": 0.06395766139030457, "learning_rate": 4.1103997843329147e-05, "loss": 0.0069, "step": 57910 }, { "grad_norm": 0.10536109656095505, "learning_rate": 4.108772744385689e-05, "loss": 0.025, "step": 57920 }, { "grad_norm": 0.15422405302524567, "learning_rate": 4.1071458019017346e-05, "loss": 0.0116, "step": 57930 }, { "grad_norm": 0.08632656186819077, "learning_rate": 4.1055189570589734e-05, "loss": 0.0085, "step": 57940 }, { "grad_norm": 0.09250091016292572, "learning_rate": 4.103892210035315e-05, "loss": 0.0078, "step": 57950 }, { "grad_norm": 0.13374358415603638, "learning_rate": 4.102265561008656e-05, "loss": 0.011, "step": 57960 }, { "grad_norm": 0.0870385617017746, "learning_rate": 4.100639010156885e-05, "loss": 0.0074, "step": 57970 }, { "grad_norm": 0.09414411336183548, "learning_rate": 4.099012557657882e-05, "loss": 0.0085, "step": 57980 }, { "grad_norm": 0.08810147643089294, "learning_rate": 4.09738620368951e-05, "loss": 0.0064, "step": 57990 }, { "grad_norm": 0.07324306666851044, "learning_rate": 4.095759948429626e-05, "loss": 0.0082, "step": 58000 }, { "grad_norm": 0.06973929703235626, "learning_rate": 4.094133792056075e-05, "loss": 0.0076, "step": 58010 }, { "grad_norm": 0.07626531273126602, "learning_rate": 4.092507734746688e-05, "loss": 0.0105, "step": 58020 }, { "grad_norm": 0.12243571132421494, "learning_rate": 4.090881776679293e-05, "loss": 0.0085, "step": 58030 }, { "grad_norm": 0.08979059755802155, "learning_rate": 4.0892559180316994e-05, "loss": 0.01, "step": 58040 }, { "grad_norm": 0.07355217635631561, "learning_rate": 4.08763015898171e-05, "loss": 0.0077, "step": 58050 }, { "grad_norm": 0.07836800813674927, "learning_rate": 4.0860044997071145e-05, "loss": 0.0076, "step": 58060 }, { "grad_norm": 0.08920186758041382, "learning_rate": 4.084378940385693e-05, "loss": 0.0094, "step": 58070 }, { "grad_norm": 0.11844460666179657, "learning_rate": 4.082753481195211e-05, "loss": 0.0096, "step": 58080 }, { "grad_norm": 0.09884504228830338, "learning_rate": 4.0811281223134324e-05, "loss": 0.0083, "step": 58090 }, { "grad_norm": 0.0680857002735138, "learning_rate": 4.0795028639181e-05, "loss": 0.0136, "step": 58100 }, { "grad_norm": 0.09298719465732574, "learning_rate": 4.077877706186951e-05, "loss": 0.0068, "step": 58110 }, { "grad_norm": 0.17808766663074493, "learning_rate": 4.076252649297709e-05, "loss": 0.0124, "step": 58120 }, { "grad_norm": 0.13467688858509064, "learning_rate": 4.07462769342809e-05, "loss": 0.011, "step": 58130 }, { "grad_norm": 0.14430277049541473, "learning_rate": 4.073002838755793e-05, "loss": 0.0106, "step": 58140 }, { "grad_norm": 0.0738183781504631, "learning_rate": 4.0713780854585115e-05, "loss": 0.0089, "step": 58150 }, { "grad_norm": 0.13703195750713348, "learning_rate": 4.069753433713927e-05, "loss": 0.0099, "step": 58160 }, { "grad_norm": 0.06889753043651581, "learning_rate": 4.068128883699709e-05, "loss": 0.0066, "step": 58170 }, { "grad_norm": 0.06077215075492859, "learning_rate": 4.0665044355935155e-05, "loss": 0.0092, "step": 58180 }, { "grad_norm": 0.1057736799120903, "learning_rate": 4.064880089572993e-05, "loss": 0.0086, "step": 58190 }, { "grad_norm": 0.12271417677402496, "learning_rate": 4.063255845815779e-05, "loss": 0.0194, "step": 58200 }, { "grad_norm": 0.11740810424089432, "learning_rate": 4.0616317044994945e-05, "loss": 0.0086, "step": 58210 }, { "grad_norm": 0.09763132780790329, "learning_rate": 4.0600076658017585e-05, "loss": 0.0097, "step": 58220 }, { "grad_norm": 0.07379946112632751, "learning_rate": 4.0583837299001706e-05, "loss": 0.0082, "step": 58230 }, { "grad_norm": 0.0825074315071106, "learning_rate": 4.056759896972322e-05, "loss": 0.0109, "step": 58240 }, { "grad_norm": 0.08318860083818436, "learning_rate": 4.055136167195793e-05, "loss": 0.0102, "step": 58250 }, { "grad_norm": 0.1387851983308792, "learning_rate": 4.053512540748152e-05, "loss": 0.0123, "step": 58260 }, { "grad_norm": 0.09540268778800964, "learning_rate": 4.051889017806955e-05, "loss": 0.0103, "step": 58270 }, { "grad_norm": 0.12207227945327759, "learning_rate": 4.0502655985497516e-05, "loss": 0.0087, "step": 58280 }, { "grad_norm": 0.12137173116207123, "learning_rate": 4.048642283154074e-05, "loss": 0.0116, "step": 58290 }, { "grad_norm": 0.12387321144342422, "learning_rate": 4.0470190717974464e-05, "loss": 0.0102, "step": 58300 }, { "grad_norm": 0.09753414243459702, "learning_rate": 4.045395964657381e-05, "loss": 0.0104, "step": 58310 }, { "grad_norm": 0.07753532379865646, "learning_rate": 4.043772961911376e-05, "loss": 0.0083, "step": 58320 }, { "grad_norm": 0.11443044245243073, "learning_rate": 4.042150063736924e-05, "loss": 0.011, "step": 58330 }, { "grad_norm": 0.14011789858341217, "learning_rate": 4.0405272703114976e-05, "loss": 0.0084, "step": 58340 }, { "grad_norm": 0.06609835475683212, "learning_rate": 4.038904581812569e-05, "loss": 0.008, "step": 58350 }, { "grad_norm": 0.08852776885032654, "learning_rate": 4.03728199841759e-05, "loss": 0.0069, "step": 58360 }, { "grad_norm": 0.06808008998632431, "learning_rate": 4.0356595203040044e-05, "loss": 0.009, "step": 58370 }, { "grad_norm": 0.07263316959142685, "learning_rate": 4.0340371476492426e-05, "loss": 0.0085, "step": 58380 }, { "grad_norm": 0.0812200978398323, "learning_rate": 4.032414880630727e-05, "loss": 0.0113, "step": 58390 }, { "grad_norm": 0.0727292075753212, "learning_rate": 4.0307927194258646e-05, "loss": 0.007, "step": 58400 }, { "grad_norm": 0.08379273116588593, "learning_rate": 4.029170664212054e-05, "loss": 0.0077, "step": 58410 }, { "grad_norm": 0.10456264764070511, "learning_rate": 4.0275487151666794e-05, "loss": 0.01, "step": 58420 }, { "grad_norm": 0.12670865654945374, "learning_rate": 4.025926872467116e-05, "loss": 0.0139, "step": 58430 }, { "grad_norm": 0.11292989552021027, "learning_rate": 4.024305136290725e-05, "loss": 0.0086, "step": 58440 }, { "grad_norm": 0.08390805870294571, "learning_rate": 4.022683506814858e-05, "loss": 0.0101, "step": 58450 }, { "grad_norm": 0.07562592625617981, "learning_rate": 4.021061984216853e-05, "loss": 0.0065, "step": 58460 }, { "grad_norm": 0.09915870428085327, "learning_rate": 4.019440568674035e-05, "loss": 0.01, "step": 58470 }, { "grad_norm": 0.15004689991474152, "learning_rate": 4.017819260363724e-05, "loss": 0.0097, "step": 58480 }, { "grad_norm": 0.09363526850938797, "learning_rate": 4.016198059463223e-05, "loss": 0.0181, "step": 58490 }, { "grad_norm": 0.09318125247955322, "learning_rate": 4.0145769661498233e-05, "loss": 0.0117, "step": 58500 }, { "grad_norm": 0.06868073344230652, "learning_rate": 4.012955980600804e-05, "loss": 0.0083, "step": 58510 }, { "grad_norm": 0.12698592245578766, "learning_rate": 4.0113351029934355e-05, "loss": 0.0105, "step": 58520 }, { "grad_norm": 0.4519055187702179, "learning_rate": 4.0097143335049714e-05, "loss": 0.0107, "step": 58530 }, { "grad_norm": 0.07234445959329605, "learning_rate": 4.0080936723126606e-05, "loss": 0.0091, "step": 58540 }, { "grad_norm": 0.1328752636909485, "learning_rate": 4.006473119593735e-05, "loss": 0.0092, "step": 58550 }, { "grad_norm": 0.11834052950143814, "learning_rate": 4.004852675525415e-05, "loss": 0.0092, "step": 58560 }, { "grad_norm": 0.08548837155103683, "learning_rate": 4.00323234028491e-05, "loss": 0.0088, "step": 58570 }, { "grad_norm": 0.08463374525308609, "learning_rate": 4.0016121140494186e-05, "loss": 0.0103, "step": 58580 }, { "grad_norm": 0.07684190571308136, "learning_rate": 3.999991996996122e-05, "loss": 0.0081, "step": 58590 }, { "grad_norm": 0.09592710435390472, "learning_rate": 3.9983719893022e-05, "loss": 0.0077, "step": 58600 }, { "grad_norm": 0.11115814745426178, "learning_rate": 3.996752091144812e-05, "loss": 0.0083, "step": 58610 }, { "grad_norm": 0.07246418297290802, "learning_rate": 3.9951323027011055e-05, "loss": 0.0089, "step": 58620 }, { "grad_norm": 0.09693262726068497, "learning_rate": 3.993512624148219e-05, "loss": 0.0102, "step": 58630 }, { "grad_norm": 0.07097925245761871, "learning_rate": 3.99189305566328e-05, "loss": 0.0065, "step": 58640 }, { "grad_norm": 0.06557465344667435, "learning_rate": 3.990273597423401e-05, "loss": 0.0098, "step": 58650 }, { "grad_norm": 0.11797311902046204, "learning_rate": 3.988654249605681e-05, "loss": 0.0075, "step": 58660 }, { "grad_norm": 0.12425190955400467, "learning_rate": 3.987035012387213e-05, "loss": 0.0086, "step": 58670 }, { "grad_norm": 0.10228970646858215, "learning_rate": 3.9854158859450735e-05, "loss": 0.0069, "step": 58680 }, { "grad_norm": 0.10435130447149277, "learning_rate": 3.983796870456328e-05, "loss": 0.0104, "step": 58690 }, { "grad_norm": 0.07837016880512238, "learning_rate": 3.982177966098028e-05, "loss": 0.0078, "step": 58700 }, { "grad_norm": 0.1523856669664383, "learning_rate": 3.980559173047216e-05, "loss": 0.013, "step": 58710 }, { "grad_norm": 0.08207741379737854, "learning_rate": 3.9789404914809176e-05, "loss": 0.0069, "step": 58720 }, { "grad_norm": 0.0734202191233635, "learning_rate": 3.9773219215761544e-05, "loss": 0.0083, "step": 58730 }, { "grad_norm": 0.09306254237890244, "learning_rate": 3.975703463509928e-05, "loss": 0.0076, "step": 58740 }, { "grad_norm": 0.11693891882896423, "learning_rate": 3.974085117459231e-05, "loss": 0.0092, "step": 58750 }, { "grad_norm": 0.06775710731744766, "learning_rate": 3.972466883601043e-05, "loss": 0.0086, "step": 58760 }, { "grad_norm": 0.0851990357041359, "learning_rate": 3.9708487621123316e-05, "loss": 0.0109, "step": 58770 }, { "grad_norm": 0.0685621052980423, "learning_rate": 3.969230753170051e-05, "loss": 0.0101, "step": 58780 }, { "grad_norm": 0.07870997488498688, "learning_rate": 3.967612856951146e-05, "loss": 0.0081, "step": 58790 }, { "grad_norm": 0.0967182144522667, "learning_rate": 3.9659950736325476e-05, "loss": 0.0088, "step": 58800 }, { "grad_norm": 0.07669518142938614, "learning_rate": 3.964377403391174e-05, "loss": 0.0265, "step": 58810 }, { "grad_norm": 0.11902662366628647, "learning_rate": 3.9627598464039303e-05, "loss": 0.015, "step": 58820 }, { "grad_norm": 0.08614178001880646, "learning_rate": 3.9611424028477096e-05, "loss": 0.0118, "step": 58830 }, { "grad_norm": 0.09893467277288437, "learning_rate": 3.959525072899394e-05, "loss": 0.0148, "step": 58840 }, { "grad_norm": 0.09798454493284225, "learning_rate": 3.9579078567358506e-05, "loss": 0.0077, "step": 58850 }, { "grad_norm": 0.10563384741544724, "learning_rate": 3.956290754533939e-05, "loss": 0.0117, "step": 58860 }, { "grad_norm": 0.21977142989635468, "learning_rate": 3.954673766470501e-05, "loss": 0.0127, "step": 58870 }, { "grad_norm": 0.12454673647880554, "learning_rate": 3.9530568927223674e-05, "loss": 0.0107, "step": 58880 }, { "grad_norm": 0.09872699528932571, "learning_rate": 3.95144013346636e-05, "loss": 0.0085, "step": 58890 }, { "grad_norm": 0.09079205989837646, "learning_rate": 3.949823488879282e-05, "loss": 0.0078, "step": 58900 }, { "grad_norm": 0.09234539419412613, "learning_rate": 3.9482069591379276e-05, "loss": 0.0119, "step": 58910 }, { "grad_norm": 0.16484348475933075, "learning_rate": 3.9465905444190784e-05, "loss": 0.0117, "step": 58920 }, { "grad_norm": 0.10131021589040756, "learning_rate": 3.944974244899505e-05, "loss": 0.0119, "step": 58930 }, { "grad_norm": 0.1274711638689041, "learning_rate": 3.9433580607559616e-05, "loss": 0.0114, "step": 58940 }, { "grad_norm": 0.11574392020702362, "learning_rate": 3.941741992165193e-05, "loss": 0.0186, "step": 58950 }, { "grad_norm": 0.15067318081855774, "learning_rate": 3.940126039303928e-05, "loss": 0.0113, "step": 58960 }, { "grad_norm": 0.1475943922996521, "learning_rate": 3.938510202348886e-05, "loss": 0.0094, "step": 58970 }, { "grad_norm": 0.08258821070194244, "learning_rate": 3.93689448147677e-05, "loss": 0.0075, "step": 58980 }, { "grad_norm": 0.07288000732660294, "learning_rate": 3.935278876864278e-05, "loss": 0.008, "step": 58990 }, { "grad_norm": 0.1219426766037941, "learning_rate": 3.933663388688087e-05, "loss": 0.0146, "step": 59000 }, { "grad_norm": 0.24230633676052094, "learning_rate": 3.9320480171248645e-05, "loss": 0.0073, "step": 59010 }, { "grad_norm": 0.18671639263629913, "learning_rate": 3.930432762351265e-05, "loss": 0.014, "step": 59020 }, { "grad_norm": 0.0832897424697876, "learning_rate": 3.928817624543931e-05, "loss": 0.0096, "step": 59030 }, { "grad_norm": 0.07871921360492706, "learning_rate": 3.9272026038794885e-05, "loss": 0.009, "step": 59040 }, { "grad_norm": 0.20860210061073303, "learning_rate": 3.9255877005345586e-05, "loss": 0.0136, "step": 59050 }, { "grad_norm": 0.08691664785146713, "learning_rate": 3.923972914685743e-05, "loss": 0.0108, "step": 59060 }, { "grad_norm": 0.11917842924594879, "learning_rate": 3.922358246509631e-05, "loss": 0.0103, "step": 59070 }, { "grad_norm": 0.14884041249752045, "learning_rate": 3.920743696182801e-05, "loss": 0.0109, "step": 59080 }, { "grad_norm": 0.07482940703630447, "learning_rate": 3.9191292638818164e-05, "loss": 0.0103, "step": 59090 }, { "grad_norm": 0.1022743210196495, "learning_rate": 3.9175149497832316e-05, "loss": 0.01, "step": 59100 }, { "grad_norm": 0.10807197540998459, "learning_rate": 3.915900754063584e-05, "loss": 0.0083, "step": 59110 }, { "grad_norm": 0.08133311569690704, "learning_rate": 3.9142866768994014e-05, "loss": 0.008, "step": 59120 }, { "grad_norm": 0.09061618149280548, "learning_rate": 3.912672718467193e-05, "loss": 0.011, "step": 59130 }, { "grad_norm": 0.14043863117694855, "learning_rate": 3.911058878943463e-05, "loss": 0.0092, "step": 59140 }, { "grad_norm": 0.08847064524888992, "learning_rate": 3.9094451585046974e-05, "loss": 0.0086, "step": 59150 }, { "grad_norm": 0.08325529843568802, "learning_rate": 3.90783155732737e-05, "loss": 0.012, "step": 59160 }, { "grad_norm": 0.11633841693401337, "learning_rate": 3.9062180755879373e-05, "loss": 0.0091, "step": 59170 }, { "grad_norm": 0.160925030708313, "learning_rate": 3.904604713462855e-05, "loss": 0.008, "step": 59180 }, { "grad_norm": 0.060526929795742035, "learning_rate": 3.902991471128554e-05, "loss": 0.0061, "step": 59190 }, { "grad_norm": 0.14493529498577118, "learning_rate": 3.901378348761456e-05, "loss": 0.0066, "step": 59200 }, { "grad_norm": 0.08923755586147308, "learning_rate": 3.899765346537971e-05, "loss": 0.0089, "step": 59210 }, { "grad_norm": 0.09578149020671844, "learning_rate": 3.898152464634493e-05, "loss": 0.0076, "step": 59220 }, { "grad_norm": 0.07152219861745834, "learning_rate": 3.896539703227402e-05, "loss": 0.0089, "step": 59230 }, { "grad_norm": 0.12731367349624634, "learning_rate": 3.894927062493073e-05, "loss": 0.0083, "step": 59240 }, { "grad_norm": 0.05817127600312233, "learning_rate": 3.893314542607859e-05, "loss": 0.0086, "step": 59250 }, { "grad_norm": 0.09360410273075104, "learning_rate": 3.891702143748103e-05, "loss": 0.0092, "step": 59260 }, { "grad_norm": 0.1422417014837265, "learning_rate": 3.890089866090133e-05, "loss": 0.0072, "step": 59270 }, { "grad_norm": 0.15781469643115997, "learning_rate": 3.888477709810267e-05, "loss": 0.01, "step": 59280 }, { "grad_norm": 0.08696311712265015, "learning_rate": 3.886865675084806e-05, "loss": 0.0075, "step": 59290 }, { "grad_norm": 0.07556881755590439, "learning_rate": 3.885253762090043e-05, "loss": 0.0077, "step": 59300 }, { "grad_norm": 0.2281668484210968, "learning_rate": 3.8836419710022534e-05, "loss": 0.0108, "step": 59310 }, { "grad_norm": 0.1260947436094284, "learning_rate": 3.882030301997698e-05, "loss": 0.0099, "step": 59320 }, { "grad_norm": 0.07620521634817123, "learning_rate": 3.880418755252629e-05, "loss": 0.0083, "step": 59330 }, { "grad_norm": 0.09102289378643036, "learning_rate": 3.87880733094328e-05, "loss": 0.0084, "step": 59340 }, { "grad_norm": 0.0516803078353405, "learning_rate": 3.877196029245877e-05, "loss": 0.0083, "step": 59350 }, { "grad_norm": 0.07904329150915146, "learning_rate": 3.875584850336627e-05, "loss": 0.0096, "step": 59360 }, { "grad_norm": 0.13006505370140076, "learning_rate": 3.873973794391728e-05, "loss": 0.0087, "step": 59370 }, { "grad_norm": 0.06065671145915985, "learning_rate": 3.872362861587361e-05, "loss": 0.0084, "step": 59380 }, { "grad_norm": 0.11698812246322632, "learning_rate": 3.870752052099698e-05, "loss": 0.0084, "step": 59390 }, { "grad_norm": 0.06572416424751282, "learning_rate": 3.8691413661048916e-05, "loss": 0.0082, "step": 59400 }, { "grad_norm": 0.09062846004962921, "learning_rate": 3.867530803779085e-05, "loss": 0.0089, "step": 59410 }, { "grad_norm": 0.07035857439041138, "learning_rate": 3.8659203652984054e-05, "loss": 0.0084, "step": 59420 }, { "grad_norm": 0.13125376403331757, "learning_rate": 3.864310050838972e-05, "loss": 0.0111, "step": 59430 }, { "grad_norm": 0.05025453865528107, "learning_rate": 3.8626998605768835e-05, "loss": 0.0077, "step": 59440 }, { "grad_norm": 0.07083598524332047, "learning_rate": 3.8610897946882294e-05, "loss": 0.0093, "step": 59450 }, { "grad_norm": 0.09439064562320709, "learning_rate": 3.859479853349082e-05, "loss": 0.0098, "step": 59460 }, { "grad_norm": 0.08634182810783386, "learning_rate": 3.857870036735503e-05, "loss": 0.0081, "step": 59470 }, { "grad_norm": 0.11426685005426407, "learning_rate": 3.856260345023538e-05, "loss": 0.0099, "step": 59480 }, { "grad_norm": 0.07921238988637924, "learning_rate": 3.854650778389224e-05, "loss": 0.0094, "step": 59490 }, { "grad_norm": 0.10803425312042236, "learning_rate": 3.8530413370085786e-05, "loss": 0.0102, "step": 59500 }, { "grad_norm": 0.06870637089014053, "learning_rate": 3.851432021057608e-05, "loss": 0.0073, "step": 59510 }, { "grad_norm": 0.0880126804113388, "learning_rate": 3.849822830712306e-05, "loss": 0.0091, "step": 59520 }, { "grad_norm": 0.11428850889205933, "learning_rate": 3.848213766148649e-05, "loss": 0.0078, "step": 59530 }, { "grad_norm": 0.10200279206037521, "learning_rate": 3.846604827542602e-05, "loss": 0.0077, "step": 59540 }, { "grad_norm": 0.12751223146915436, "learning_rate": 3.844996015070117e-05, "loss": 0.0164, "step": 59550 }, { "grad_norm": 0.08951503783464432, "learning_rate": 3.843387328907132e-05, "loss": 0.0069, "step": 59560 }, { "grad_norm": 0.11390617489814758, "learning_rate": 3.8417787692295696e-05, "loss": 0.0136, "step": 59570 }, { "grad_norm": 0.1215326115489006, "learning_rate": 3.84017033621334e-05, "loss": 0.0067, "step": 59580 }, { "grad_norm": 0.06252360343933105, "learning_rate": 3.838562030034338e-05, "loss": 0.0069, "step": 59590 }, { "grad_norm": 0.055021218955516815, "learning_rate": 3.836953850868447e-05, "loss": 0.0071, "step": 59600 }, { "grad_norm": 0.056705038994550705, "learning_rate": 3.835345798891533e-05, "loss": 0.0067, "step": 59610 }, { "grad_norm": 0.10484713315963745, "learning_rate": 3.833737874279452e-05, "loss": 0.0105, "step": 59620 }, { "grad_norm": 0.07017987966537476, "learning_rate": 3.8321300772080433e-05, "loss": 0.007, "step": 59630 }, { "grad_norm": 0.10517983138561249, "learning_rate": 3.8305224078531345e-05, "loss": 0.008, "step": 59640 }, { "grad_norm": 0.16674061119556427, "learning_rate": 3.828914866390537e-05, "loss": 0.0114, "step": 59650 }, { "grad_norm": 0.13066786527633667, "learning_rate": 3.827307452996048e-05, "loss": 0.0125, "step": 59660 }, { "grad_norm": 0.09899228066205978, "learning_rate": 3.825700167845454e-05, "loss": 0.0179, "step": 59670 }, { "grad_norm": 0.14406070113182068, "learning_rate": 3.8240930111145214e-05, "loss": 0.0146, "step": 59680 }, { "grad_norm": 0.09869715571403503, "learning_rate": 3.8224859829790116e-05, "loss": 0.0074, "step": 59690 }, { "grad_norm": 0.1028614267706871, "learning_rate": 3.820879083614664e-05, "loss": 0.0125, "step": 59700 }, { "grad_norm": 0.24911117553710938, "learning_rate": 3.819272313197208e-05, "loss": 0.0122, "step": 59710 }, { "grad_norm": 0.07021868228912354, "learning_rate": 3.817665671902355e-05, "loss": 0.0062, "step": 59720 }, { "grad_norm": 0.057513006031513214, "learning_rate": 3.816059159905808e-05, "loss": 0.009, "step": 59730 }, { "grad_norm": 0.08776011317968369, "learning_rate": 3.8144527773832476e-05, "loss": 0.0112, "step": 59740 }, { "grad_norm": 0.0842214822769165, "learning_rate": 3.812846524510352e-05, "loss": 0.0078, "step": 59750 }, { "grad_norm": 0.09288164973258972, "learning_rate": 3.811240401462775e-05, "loss": 0.0104, "step": 59760 }, { "grad_norm": 0.08629260212182999, "learning_rate": 3.809634408416162e-05, "loss": 0.0123, "step": 59770 }, { "grad_norm": 0.16567125916481018, "learning_rate": 3.8080285455461394e-05, "loss": 0.009, "step": 59780 }, { "grad_norm": 0.10674768686294556, "learning_rate": 3.806422813028323e-05, "loss": 0.0074, "step": 59790 }, { "grad_norm": 0.1259160041809082, "learning_rate": 3.8048172110383114e-05, "loss": 0.0083, "step": 59800 }, { "grad_norm": 0.07432392984628677, "learning_rate": 3.8032117397516944e-05, "loss": 0.0076, "step": 59810 }, { "grad_norm": 0.12839746475219727, "learning_rate": 3.801606399344042e-05, "loss": 0.009, "step": 59820 }, { "grad_norm": 0.09531337022781372, "learning_rate": 3.800001189990913e-05, "loss": 0.008, "step": 59830 }, { "grad_norm": 0.10222813487052917, "learning_rate": 3.798396111867847e-05, "loss": 0.0081, "step": 59840 }, { "grad_norm": 0.05547190457582474, "learning_rate": 3.796791165150378e-05, "loss": 0.007, "step": 59850 }, { "grad_norm": 0.10286195576190948, "learning_rate": 3.795186350014018e-05, "loss": 0.008, "step": 59860 }, { "grad_norm": 0.10828021913766861, "learning_rate": 3.793581666634266e-05, "loss": 0.0084, "step": 59870 }, { "grad_norm": 0.06114666163921356, "learning_rate": 3.791977115186609e-05, "loss": 0.0073, "step": 59880 }, { "grad_norm": 0.08108583837747574, "learning_rate": 3.79037269584652e-05, "loss": 0.0087, "step": 59890 }, { "grad_norm": 0.11023040115833282, "learning_rate": 3.7887684087894545e-05, "loss": 0.0107, "step": 59900 }, { "grad_norm": 0.08829966932535172, "learning_rate": 3.787164254190854e-05, "loss": 0.0067, "step": 59910 }, { "grad_norm": 0.07962602376937866, "learning_rate": 3.7855602322261473e-05, "loss": 0.0083, "step": 59920 }, { "grad_norm": 0.19879698753356934, "learning_rate": 3.783956343070746e-05, "loss": 0.0123, "step": 59930 }, { "grad_norm": 0.09931936115026474, "learning_rate": 3.7823525869000526e-05, "loss": 0.0085, "step": 59940 }, { "grad_norm": 0.12976442277431488, "learning_rate": 3.78074896388945e-05, "loss": 0.0118, "step": 59950 }, { "grad_norm": 0.11439649760723114, "learning_rate": 3.7791454742143075e-05, "loss": 0.0101, "step": 59960 }, { "grad_norm": 0.09161200374364853, "learning_rate": 3.7775421180499795e-05, "loss": 0.0068, "step": 59970 }, { "grad_norm": 0.10370295494794846, "learning_rate": 3.7759388955718086e-05, "loss": 0.0089, "step": 59980 }, { "grad_norm": 0.08525866270065308, "learning_rate": 3.774335806955117e-05, "loss": 0.0104, "step": 59990 }, { "grad_norm": 0.16893485188484192, "learning_rate": 3.7727328523752215e-05, "loss": 0.0081, "step": 60000 } ], "logging_steps": 10, "max_steps": 100000, "num_input_tokens_seen": 0, "num_train_epochs": 9223372036854775807, "save_steps": 20000, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": false }, "attributes": {} } }, "total_flos": 0.0, "train_batch_size": 8, "trial_name": null, "trial_params": null }