{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 1.0, "eval_steps": 500, "global_step": 80000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "grad_norm": 1.121005654335022, "learning_rate": 2.25e-07, "loss": 1.5252, "step": 10 }, { "grad_norm": 1.098426103591919, "learning_rate": 4.75e-07, "loss": 1.5364, "step": 20 }, { "grad_norm": 1.2078971862792969, "learning_rate": 7.25e-07, "loss": 1.5046, "step": 30 }, { "grad_norm": 1.0568773746490479, "learning_rate": 9.75e-07, "loss": 1.5147, "step": 40 }, { "grad_norm": 1.256998062133789, "learning_rate": 1.2250000000000001e-06, "loss": 1.4773, "step": 50 }, { "grad_norm": 1.8394856452941895, "learning_rate": 1.475e-06, "loss": 1.385, "step": 60 }, { "grad_norm": 1.4321011304855347, "learning_rate": 1.7250000000000002e-06, "loss": 1.2835, "step": 70 }, { "grad_norm": 1.1531648635864258, "learning_rate": 1.975e-06, "loss": 1.2113, "step": 80 }, { "grad_norm": 0.834656834602356, "learning_rate": 2.225e-06, "loss": 1.1785, "step": 90 }, { "grad_norm": 1.5144314765930176, "learning_rate": 2.4750000000000004e-06, "loss": 1.1652, "step": 100 }, { "grad_norm": 1.0780751705169678, "learning_rate": 2.725e-06, "loss": 1.1364, "step": 110 }, { "grad_norm": 1.8253467082977295, "learning_rate": 2.975e-06, "loss": 1.1663, "step": 120 }, { "grad_norm": 1.637829065322876, "learning_rate": 3.225e-06, "loss": 1.1326, "step": 130 }, { "grad_norm": 1.2881423234939575, "learning_rate": 3.4750000000000006e-06, "loss": 1.1291, "step": 140 }, { "grad_norm": 1.515366792678833, "learning_rate": 3.725e-06, "loss": 1.1036, "step": 150 }, { "grad_norm": 1.1055456399917603, "learning_rate": 3.975e-06, "loss": 1.1311, "step": 160 }, { "grad_norm": 1.3808757066726685, "learning_rate": 4.225e-06, "loss": 1.1061, "step": 170 }, { "grad_norm": 1.5630948543548584, "learning_rate": 4.475e-06, "loss": 1.1127, "step": 180 }, { "grad_norm": 1.288834571838379, "learning_rate": 4.7250000000000005e-06, "loss": 1.0973, "step": 190 }, { "grad_norm": 1.3455849885940552, "learning_rate": 4.975000000000001e-06, "loss": 1.0989, "step": 200 }, { "grad_norm": 0.9715861082077026, "learning_rate": 5.225e-06, "loss": 1.0912, "step": 210 }, { "grad_norm": 1.2705224752426147, "learning_rate": 5.475e-06, "loss": 1.0976, "step": 220 }, { "grad_norm": 2.9808297157287598, "learning_rate": 5.725e-06, "loss": 1.126, "step": 230 }, { "grad_norm": 1.4939380884170532, "learning_rate": 5.975e-06, "loss": 1.0911, "step": 240 }, { "grad_norm": 1.4222742319107056, "learning_rate": 6.2250000000000005e-06, "loss": 1.0876, "step": 250 }, { "grad_norm": 1.5305079221725464, "learning_rate": 6.475000000000001e-06, "loss": 1.081, "step": 260 }, { "grad_norm": 1.1755304336547852, "learning_rate": 6.725000000000001e-06, "loss": 1.0824, "step": 270 }, { "grad_norm": 1.0639582872390747, "learning_rate": 6.975000000000001e-06, "loss": 1.0667, "step": 280 }, { "grad_norm": 1.266035795211792, "learning_rate": 7.2249999999999994e-06, "loss": 1.0536, "step": 290 }, { "grad_norm": 1.1176941394805908, "learning_rate": 7.4750000000000004e-06, "loss": 1.0568, "step": 300 }, { "grad_norm": 1.0939548015594482, "learning_rate": 7.725e-06, "loss": 1.0797, "step": 310 }, { "grad_norm": 1.3648834228515625, "learning_rate": 7.975e-06, "loss": 1.0407, "step": 320 }, { "grad_norm": 1.4159666299819946, "learning_rate": 8.225e-06, "loss": 1.0429, "step": 330 }, { "grad_norm": 1.3145655393600464, "learning_rate": 8.475000000000001e-06, "loss": 1.0128, "step": 340 }, { "grad_norm": 1.4442048072814941, "learning_rate": 8.725e-06, "loss": 0.9681, "step": 350 }, { "grad_norm": 1.3777692317962646, "learning_rate": 8.975e-06, "loss": 0.9454, "step": 360 }, { "grad_norm": 1.5711113214492798, "learning_rate": 9.225e-06, "loss": 0.9273, "step": 370 }, { "grad_norm": 1.754247784614563, "learning_rate": 9.475e-06, "loss": 0.8773, "step": 380 }, { "grad_norm": 2.223850727081299, "learning_rate": 9.725000000000001e-06, "loss": 0.8485, "step": 390 }, { "grad_norm": 1.9668300151824951, "learning_rate": 9.975e-06, "loss": 0.8109, "step": 400 }, { "grad_norm": 1.9332512617111206, "learning_rate": 1.0225e-05, "loss": 0.7665, "step": 410 }, { "grad_norm": 2.39975905418396, "learning_rate": 1.0475e-05, "loss": 0.7168, "step": 420 }, { "grad_norm": 2.356105089187622, "learning_rate": 1.0725e-05, "loss": 0.6988, "step": 430 }, { "grad_norm": 1.766808032989502, "learning_rate": 1.0975e-05, "loss": 0.6658, "step": 440 }, { "grad_norm": 1.8786314725875854, "learning_rate": 1.1225e-05, "loss": 0.6148, "step": 450 }, { "grad_norm": 2.140390396118164, "learning_rate": 1.1475000000000001e-05, "loss": 0.5867, "step": 460 }, { "grad_norm": 2.407991886138916, "learning_rate": 1.1725e-05, "loss": 0.5711, "step": 470 }, { "grad_norm": 2.0050857067108154, "learning_rate": 1.1975e-05, "loss": 0.5653, "step": 480 }, { "grad_norm": 1.7494064569473267, "learning_rate": 1.2225e-05, "loss": 0.531, "step": 490 }, { "grad_norm": 2.9363152980804443, "learning_rate": 1.2475e-05, "loss": 0.4941, "step": 500 }, { "grad_norm": 2.2491798400878906, "learning_rate": 1.2725000000000001e-05, "loss": 0.4959, "step": 510 }, { "grad_norm": 2.280500888824463, "learning_rate": 1.2975e-05, "loss": 0.4682, "step": 520 }, { "grad_norm": 2.267956018447876, "learning_rate": 1.3225000000000001e-05, "loss": 0.4613, "step": 530 }, { "grad_norm": 2.0224385261535645, "learning_rate": 1.3475000000000002e-05, "loss": 0.3907, "step": 540 }, { "grad_norm": 2.8898017406463623, "learning_rate": 1.3725000000000002e-05, "loss": 0.4178, "step": 550 }, { "grad_norm": 2.549264430999756, "learning_rate": 1.3975000000000003e-05, "loss": 0.392, "step": 560 }, { "grad_norm": 2.3589160442352295, "learning_rate": 1.4225e-05, "loss": 0.3473, "step": 570 }, { "grad_norm": 2.3888375759124756, "learning_rate": 1.4475e-05, "loss": 0.3222, "step": 580 }, { "grad_norm": 2.4678544998168945, "learning_rate": 1.4725e-05, "loss": 0.3277, "step": 590 }, { "grad_norm": 1.8492510318756104, "learning_rate": 1.4975e-05, "loss": 0.3198, "step": 600 }, { "grad_norm": 3.0117177963256836, "learning_rate": 1.5225e-05, "loss": 0.3091, "step": 610 }, { "grad_norm": 2.7009100914001465, "learning_rate": 1.5475e-05, "loss": 0.3119, "step": 620 }, { "grad_norm": 2.503520965576172, "learning_rate": 1.5725e-05, "loss": 0.2901, "step": 630 }, { "grad_norm": 2.8344008922576904, "learning_rate": 1.5975000000000002e-05, "loss": 0.2496, "step": 640 }, { "grad_norm": 3.0942108631134033, "learning_rate": 1.6225e-05, "loss": 0.2424, "step": 650 }, { "grad_norm": 2.917459011077881, "learning_rate": 1.6475e-05, "loss": 0.2542, "step": 660 }, { "grad_norm": 1.9132962226867676, "learning_rate": 1.6725000000000003e-05, "loss": 0.1959, "step": 670 }, { "grad_norm": 3.039884328842163, "learning_rate": 1.6975000000000003e-05, "loss": 0.2113, "step": 680 }, { "grad_norm": 2.1146152019500732, "learning_rate": 1.7225e-05, "loss": 0.1996, "step": 690 }, { "grad_norm": 2.2227330207824707, "learning_rate": 1.7475e-05, "loss": 0.2238, "step": 700 }, { "grad_norm": 2.2381155490875244, "learning_rate": 1.7725e-05, "loss": 0.1815, "step": 710 }, { "grad_norm": 2.783285617828369, "learning_rate": 1.7975e-05, "loss": 0.2012, "step": 720 }, { "grad_norm": 1.9433461427688599, "learning_rate": 1.8225e-05, "loss": 0.1866, "step": 730 }, { "grad_norm": 2.397273063659668, "learning_rate": 1.8475000000000002e-05, "loss": 0.1566, "step": 740 }, { "grad_norm": 1.808558702468872, "learning_rate": 1.8725e-05, "loss": 0.1846, "step": 750 }, { "grad_norm": 1.5180989503860474, "learning_rate": 1.8975e-05, "loss": 0.168, "step": 760 }, { "grad_norm": 1.878841757774353, "learning_rate": 1.9225e-05, "loss": 0.1741, "step": 770 }, { "grad_norm": 2.444445848464966, "learning_rate": 1.9475000000000002e-05, "loss": 0.2077, "step": 780 }, { "grad_norm": 2.151775360107422, "learning_rate": 1.9725000000000002e-05, "loss": 0.1586, "step": 790 }, { "grad_norm": 2.5324764251708984, "learning_rate": 1.9975e-05, "loss": 0.1616, "step": 800 }, { "grad_norm": 2.515255928039551, "learning_rate": 2.0225000000000004e-05, "loss": 0.168, "step": 810 }, { "grad_norm": 1.9926341772079468, "learning_rate": 2.0475e-05, "loss": 0.162, "step": 820 }, { "grad_norm": 2.0468251705169678, "learning_rate": 2.0725e-05, "loss": 0.1821, "step": 830 }, { "grad_norm": 1.687544345855713, "learning_rate": 2.0975e-05, "loss": 0.1753, "step": 840 }, { "grad_norm": 1.8437021970748901, "learning_rate": 2.1225e-05, "loss": 0.1682, "step": 850 }, { "grad_norm": 2.00057053565979, "learning_rate": 2.1475e-05, "loss": 0.1463, "step": 860 }, { "grad_norm": 2.1531760692596436, "learning_rate": 2.1725e-05, "loss": 0.1757, "step": 870 }, { "grad_norm": 2.084429979324341, "learning_rate": 2.1975000000000002e-05, "loss": 0.1488, "step": 880 }, { "grad_norm": 2.434541702270508, "learning_rate": 2.2225e-05, "loss": 0.1558, "step": 890 }, { "grad_norm": 1.6967642307281494, "learning_rate": 2.2475e-05, "loss": 0.1583, "step": 900 }, { "grad_norm": 1.675935983657837, "learning_rate": 2.2725000000000003e-05, "loss": 0.1572, "step": 910 }, { "grad_norm": 1.6067559719085693, "learning_rate": 2.2975000000000003e-05, "loss": 0.1244, "step": 920 }, { "grad_norm": 2.2007641792297363, "learning_rate": 2.3225000000000002e-05, "loss": 0.1617, "step": 930 }, { "grad_norm": 1.9188828468322754, "learning_rate": 2.3475e-05, "loss": 0.1622, "step": 940 }, { "grad_norm": 1.4582762718200684, "learning_rate": 2.3725e-05, "loss": 0.1356, "step": 950 }, { "grad_norm": 2.1646711826324463, "learning_rate": 2.3975e-05, "loss": 0.1427, "step": 960 }, { "grad_norm": 1.7772908210754395, "learning_rate": 2.4225e-05, "loss": 0.1734, "step": 970 }, { "grad_norm": 2.2620973587036133, "learning_rate": 2.4475000000000002e-05, "loss": 0.1293, "step": 980 }, { "grad_norm": 2.0178773403167725, "learning_rate": 2.4725e-05, "loss": 0.1336, "step": 990 }, { "grad_norm": 1.634515643119812, "learning_rate": 2.4975e-05, "loss": 0.1329, "step": 1000 }, { "grad_norm": 1.2787086963653564, "learning_rate": 2.5225e-05, "loss": 0.1341, "step": 1010 }, { "grad_norm": 1.5896923542022705, "learning_rate": 2.5475e-05, "loss": 0.1507, "step": 1020 }, { "grad_norm": 1.689558506011963, "learning_rate": 2.5725e-05, "loss": 0.1176, "step": 1030 }, { "grad_norm": 1.7368279695510864, "learning_rate": 2.5974999999999998e-05, "loss": 0.1343, "step": 1040 }, { "grad_norm": 2.012054204940796, "learning_rate": 2.6225e-05, "loss": 0.14, "step": 1050 }, { "grad_norm": 2.163816213607788, "learning_rate": 2.6475e-05, "loss": 0.1539, "step": 1060 }, { "grad_norm": 2.014840841293335, "learning_rate": 2.6725e-05, "loss": 0.1551, "step": 1070 }, { "grad_norm": 1.3771517276763916, "learning_rate": 2.6975000000000002e-05, "loss": 0.1542, "step": 1080 }, { "grad_norm": 1.7582107782363892, "learning_rate": 2.7225e-05, "loss": 0.1326, "step": 1090 }, { "grad_norm": 1.6140812635421753, "learning_rate": 2.7475e-05, "loss": 0.1185, "step": 1100 }, { "grad_norm": 2.1230320930480957, "learning_rate": 2.7725e-05, "loss": 0.1609, "step": 1110 }, { "grad_norm": 1.3654413223266602, "learning_rate": 2.7975000000000002e-05, "loss": 0.1194, "step": 1120 }, { "grad_norm": 1.4254071712493896, "learning_rate": 2.8225e-05, "loss": 0.1354, "step": 1130 }, { "grad_norm": 2.2245163917541504, "learning_rate": 2.8475e-05, "loss": 0.1691, "step": 1140 }, { "grad_norm": 1.9085265398025513, "learning_rate": 2.8725e-05, "loss": 0.1574, "step": 1150 }, { "grad_norm": 1.5015127658843994, "learning_rate": 2.8975000000000003e-05, "loss": 0.1352, "step": 1160 }, { "grad_norm": 1.62619948387146, "learning_rate": 2.9225000000000002e-05, "loss": 0.134, "step": 1170 }, { "grad_norm": 1.4004398584365845, "learning_rate": 2.9475e-05, "loss": 0.1178, "step": 1180 }, { "grad_norm": 1.4763094186782837, "learning_rate": 2.9725000000000004e-05, "loss": 0.1399, "step": 1190 }, { "grad_norm": 2.379767894744873, "learning_rate": 2.9975000000000004e-05, "loss": 0.1307, "step": 1200 }, { "grad_norm": 2.692941427230835, "learning_rate": 3.0225000000000003e-05, "loss": 0.1672, "step": 1210 }, { "grad_norm": 1.2952873706817627, "learning_rate": 3.0475000000000002e-05, "loss": 0.1384, "step": 1220 }, { "grad_norm": 1.5826505422592163, "learning_rate": 3.0725e-05, "loss": 0.1301, "step": 1230 }, { "grad_norm": 1.3424187898635864, "learning_rate": 3.0975e-05, "loss": 0.1413, "step": 1240 }, { "grad_norm": 1.9122514724731445, "learning_rate": 3.122500000000001e-05, "loss": 0.1229, "step": 1250 }, { "grad_norm": 1.63528573513031, "learning_rate": 3.1475e-05, "loss": 0.1281, "step": 1260 }, { "grad_norm": 1.78049635887146, "learning_rate": 3.1725e-05, "loss": 0.1068, "step": 1270 }, { "grad_norm": 1.3700839281082153, "learning_rate": 3.1975e-05, "loss": 0.1337, "step": 1280 }, { "grad_norm": 1.2153373956680298, "learning_rate": 3.2225e-05, "loss": 0.1317, "step": 1290 }, { "grad_norm": 1.609290599822998, "learning_rate": 3.2474999999999997e-05, "loss": 0.1222, "step": 1300 }, { "grad_norm": 1.2899553775787354, "learning_rate": 3.2725e-05, "loss": 0.1211, "step": 1310 }, { "grad_norm": 1.3369027376174927, "learning_rate": 3.2975e-05, "loss": 0.1208, "step": 1320 }, { "grad_norm": 1.3738212585449219, "learning_rate": 3.3225e-05, "loss": 0.115, "step": 1330 }, { "grad_norm": 1.2463552951812744, "learning_rate": 3.3475e-05, "loss": 0.1033, "step": 1340 }, { "grad_norm": 1.428085446357727, "learning_rate": 3.3725e-05, "loss": 0.1306, "step": 1350 }, { "grad_norm": 2.108917236328125, "learning_rate": 3.3975e-05, "loss": 0.1256, "step": 1360 }, { "grad_norm": 1.3310357332229614, "learning_rate": 3.4225e-05, "loss": 0.1021, "step": 1370 }, { "grad_norm": 1.4799468517303467, "learning_rate": 3.4475000000000005e-05, "loss": 0.1044, "step": 1380 }, { "grad_norm": 1.640622854232788, "learning_rate": 3.4725000000000004e-05, "loss": 0.1097, "step": 1390 }, { "grad_norm": 1.452348232269287, "learning_rate": 3.4975e-05, "loss": 0.1065, "step": 1400 }, { "grad_norm": 1.25355863571167, "learning_rate": 3.5225e-05, "loss": 0.1137, "step": 1410 }, { "grad_norm": 1.7678731679916382, "learning_rate": 3.5475e-05, "loss": 0.1205, "step": 1420 }, { "grad_norm": 1.2328757047653198, "learning_rate": 3.5725e-05, "loss": 0.0986, "step": 1430 }, { "grad_norm": 1.3111356496810913, "learning_rate": 3.5975e-05, "loss": 0.1258, "step": 1440 }, { "grad_norm": 1.2619454860687256, "learning_rate": 3.6225000000000006e-05, "loss": 0.1208, "step": 1450 }, { "grad_norm": 1.38468599319458, "learning_rate": 3.6475000000000006e-05, "loss": 0.1234, "step": 1460 }, { "grad_norm": 1.132032036781311, "learning_rate": 3.6725000000000005e-05, "loss": 0.1237, "step": 1470 }, { "grad_norm": 1.45026433467865, "learning_rate": 3.6975000000000004e-05, "loss": 0.108, "step": 1480 }, { "grad_norm": 1.2852901220321655, "learning_rate": 3.7225000000000004e-05, "loss": 0.1079, "step": 1490 }, { "grad_norm": 1.0655888319015503, "learning_rate": 3.7475e-05, "loss": 0.1176, "step": 1500 }, { "grad_norm": 0.9598363637924194, "learning_rate": 3.7725e-05, "loss": 0.0994, "step": 1510 }, { "grad_norm": 1.5474919080734253, "learning_rate": 3.7975e-05, "loss": 0.1175, "step": 1520 }, { "grad_norm": 1.2220356464385986, "learning_rate": 3.8225e-05, "loss": 0.1175, "step": 1530 }, { "grad_norm": 1.0901398658752441, "learning_rate": 3.8475e-05, "loss": 0.1065, "step": 1540 }, { "grad_norm": 1.1398667097091675, "learning_rate": 3.8725e-05, "loss": 0.0972, "step": 1550 }, { "grad_norm": 1.1007106304168701, "learning_rate": 3.8975e-05, "loss": 0.1161, "step": 1560 }, { "grad_norm": 1.2919716835021973, "learning_rate": 3.9225e-05, "loss": 0.0917, "step": 1570 }, { "grad_norm": 1.2142064571380615, "learning_rate": 3.9475000000000004e-05, "loss": 0.106, "step": 1580 }, { "grad_norm": 1.4552786350250244, "learning_rate": 3.9725e-05, "loss": 0.113, "step": 1590 }, { "grad_norm": 0.9106762409210205, "learning_rate": 3.9975e-05, "loss": 0.0985, "step": 1600 }, { "grad_norm": 1.1024417877197266, "learning_rate": 4.0225e-05, "loss": 0.0877, "step": 1610 }, { "grad_norm": 1.1327942609786987, "learning_rate": 4.0475e-05, "loss": 0.1136, "step": 1620 }, { "grad_norm": 1.0002038478851318, "learning_rate": 4.0725e-05, "loss": 0.136, "step": 1630 }, { "grad_norm": 1.1021066904067993, "learning_rate": 4.0975e-05, "loss": 0.1156, "step": 1640 }, { "grad_norm": 0.9444631338119507, "learning_rate": 4.1225e-05, "loss": 0.1096, "step": 1650 }, { "grad_norm": 1.1036098003387451, "learning_rate": 4.1475000000000005e-05, "loss": 0.1206, "step": 1660 }, { "grad_norm": 1.4367233514785767, "learning_rate": 4.1725000000000005e-05, "loss": 0.1059, "step": 1670 }, { "grad_norm": 1.304125428199768, "learning_rate": 4.1975000000000004e-05, "loss": 0.1036, "step": 1680 }, { "grad_norm": 0.9623640179634094, "learning_rate": 4.2225e-05, "loss": 0.0975, "step": 1690 }, { "grad_norm": 0.9920961260795593, "learning_rate": 4.2475e-05, "loss": 0.1077, "step": 1700 }, { "grad_norm": 1.520556926727295, "learning_rate": 4.2725e-05, "loss": 0.1214, "step": 1710 }, { "grad_norm": 0.9028767347335815, "learning_rate": 4.2975e-05, "loss": 0.1155, "step": 1720 }, { "grad_norm": 1.0145341157913208, "learning_rate": 4.322500000000001e-05, "loss": 0.1118, "step": 1730 }, { "grad_norm": 1.2038990259170532, "learning_rate": 4.3475000000000006e-05, "loss": 0.098, "step": 1740 }, { "grad_norm": 1.2264971733093262, "learning_rate": 4.3725000000000006e-05, "loss": 0.1134, "step": 1750 }, { "grad_norm": 1.16008722782135, "learning_rate": 4.3975e-05, "loss": 0.1164, "step": 1760 }, { "grad_norm": 1.1774541139602661, "learning_rate": 4.4225e-05, "loss": 0.1013, "step": 1770 }, { "grad_norm": 0.8039512634277344, "learning_rate": 4.4475e-05, "loss": 0.1205, "step": 1780 }, { "grad_norm": 1.2446894645690918, "learning_rate": 4.4725e-05, "loss": 0.0974, "step": 1790 }, { "grad_norm": 1.3224968910217285, "learning_rate": 4.4975e-05, "loss": 0.1105, "step": 1800 }, { "grad_norm": 0.7471984028816223, "learning_rate": 4.5225e-05, "loss": 0.1017, "step": 1810 }, { "grad_norm": 1.3821659088134766, "learning_rate": 4.5475e-05, "loss": 0.1089, "step": 1820 }, { "grad_norm": 1.5408413410186768, "learning_rate": 4.5725e-05, "loss": 0.1125, "step": 1830 }, { "grad_norm": 1.1606374979019165, "learning_rate": 4.5975e-05, "loss": 0.1151, "step": 1840 }, { "grad_norm": 1.0767736434936523, "learning_rate": 4.6225e-05, "loss": 0.1346, "step": 1850 }, { "grad_norm": 0.8326289057731628, "learning_rate": 4.6475000000000005e-05, "loss": 0.119, "step": 1860 }, { "grad_norm": 1.1896811723709106, "learning_rate": 4.6725000000000004e-05, "loss": 0.1144, "step": 1870 }, { "grad_norm": 1.0002437829971313, "learning_rate": 4.6975000000000003e-05, "loss": 0.0932, "step": 1880 }, { "grad_norm": 1.0180840492248535, "learning_rate": 4.7225e-05, "loss": 0.113, "step": 1890 }, { "grad_norm": 1.4022252559661865, "learning_rate": 4.7475e-05, "loss": 0.0901, "step": 1900 }, { "grad_norm": 1.5173518657684326, "learning_rate": 4.7725e-05, "loss": 0.1297, "step": 1910 }, { "grad_norm": 1.103956937789917, "learning_rate": 4.7975e-05, "loss": 0.1004, "step": 1920 }, { "grad_norm": 1.1607311964035034, "learning_rate": 4.822500000000001e-05, "loss": 0.0983, "step": 1930 }, { "grad_norm": 1.2889913320541382, "learning_rate": 4.8475000000000006e-05, "loss": 0.1077, "step": 1940 }, { "grad_norm": 0.6800610423088074, "learning_rate": 4.8725000000000005e-05, "loss": 0.0964, "step": 1950 }, { "grad_norm": 1.2305705547332764, "learning_rate": 4.8975000000000005e-05, "loss": 0.129, "step": 1960 }, { "grad_norm": 0.7882130742073059, "learning_rate": 4.9225000000000004e-05, "loss": 0.1012, "step": 1970 }, { "grad_norm": 1.5297895669937134, "learning_rate": 4.9475e-05, "loss": 0.1205, "step": 1980 }, { "grad_norm": 0.7949218153953552, "learning_rate": 4.9725e-05, "loss": 0.0987, "step": 1990 }, { "grad_norm": 1.024610996246338, "learning_rate": 4.9975e-05, "loss": 0.1059, "step": 2000 }, { "grad_norm": 0.9348684549331665, "learning_rate": 5.0225e-05, "loss": 0.1372, "step": 2010 }, { "grad_norm": 1.2875611782073975, "learning_rate": 5.047500000000001e-05, "loss": 0.1075, "step": 2020 }, { "grad_norm": 0.9901153445243835, "learning_rate": 5.0725e-05, "loss": 0.0846, "step": 2030 }, { "grad_norm": 1.085318684577942, "learning_rate": 5.0975000000000006e-05, "loss": 0.1157, "step": 2040 }, { "grad_norm": 1.0365221500396729, "learning_rate": 5.1225e-05, "loss": 0.1057, "step": 2050 }, { "grad_norm": 0.8770920634269714, "learning_rate": 5.1475000000000004e-05, "loss": 0.0931, "step": 2060 }, { "grad_norm": 0.9151886105537415, "learning_rate": 5.1725000000000004e-05, "loss": 0.0877, "step": 2070 }, { "grad_norm": 0.9144207239151001, "learning_rate": 5.197500000000001e-05, "loss": 0.1012, "step": 2080 }, { "grad_norm": 0.8643687963485718, "learning_rate": 5.2225e-05, "loss": 0.0944, "step": 2090 }, { "grad_norm": 0.9832541346549988, "learning_rate": 5.247500000000001e-05, "loss": 0.1097, "step": 2100 }, { "grad_norm": 0.7618874907493591, "learning_rate": 5.2725e-05, "loss": 0.1246, "step": 2110 }, { "grad_norm": 1.0956093072891235, "learning_rate": 5.297500000000001e-05, "loss": 0.094, "step": 2120 }, { "grad_norm": 0.7955130934715271, "learning_rate": 5.3225e-05, "loss": 0.097, "step": 2130 }, { "grad_norm": 0.9649744629859924, "learning_rate": 5.3475e-05, "loss": 0.0854, "step": 2140 }, { "grad_norm": 0.8775770664215088, "learning_rate": 5.3725000000000005e-05, "loss": 0.1096, "step": 2150 }, { "grad_norm": 0.8110630512237549, "learning_rate": 5.3975e-05, "loss": 0.0942, "step": 2160 }, { "grad_norm": 0.7850621342658997, "learning_rate": 5.4225000000000003e-05, "loss": 0.0937, "step": 2170 }, { "grad_norm": 0.9284189939498901, "learning_rate": 5.4474999999999996e-05, "loss": 0.117, "step": 2180 }, { "grad_norm": 1.1038529872894287, "learning_rate": 5.4725e-05, "loss": 0.1037, "step": 2190 }, { "grad_norm": 0.9372755289077759, "learning_rate": 5.4975e-05, "loss": 0.0965, "step": 2200 }, { "grad_norm": 0.9955427050590515, "learning_rate": 5.522500000000001e-05, "loss": 0.0957, "step": 2210 }, { "grad_norm": 1.1646369695663452, "learning_rate": 5.5475e-05, "loss": 0.0943, "step": 2220 }, { "grad_norm": 0.8345479965209961, "learning_rate": 5.5725000000000006e-05, "loss": 0.1034, "step": 2230 }, { "grad_norm": 1.082646369934082, "learning_rate": 5.5975e-05, "loss": 0.0869, "step": 2240 }, { "grad_norm": 0.956725537776947, "learning_rate": 5.6225000000000005e-05, "loss": 0.1028, "step": 2250 }, { "grad_norm": 0.8162146210670471, "learning_rate": 5.6475e-05, "loss": 0.0902, "step": 2260 }, { "grad_norm": 0.7683647871017456, "learning_rate": 5.6725e-05, "loss": 0.0853, "step": 2270 }, { "grad_norm": 0.9686477184295654, "learning_rate": 5.6975e-05, "loss": 0.0829, "step": 2280 }, { "grad_norm": 1.1406627893447876, "learning_rate": 5.722500000000001e-05, "loss": 0.109, "step": 2290 }, { "grad_norm": 0.8410414457321167, "learning_rate": 5.7475e-05, "loss": 0.0818, "step": 2300 }, { "grad_norm": 0.7203332781791687, "learning_rate": 5.772500000000001e-05, "loss": 0.0903, "step": 2310 }, { "grad_norm": 0.7163267731666565, "learning_rate": 5.7975e-05, "loss": 0.0832, "step": 2320 }, { "grad_norm": 0.9604745507240295, "learning_rate": 5.8225000000000006e-05, "loss": 0.1, "step": 2330 }, { "grad_norm": 0.9375787973403931, "learning_rate": 5.8475000000000005e-05, "loss": 0.1005, "step": 2340 }, { "grad_norm": 0.8483784198760986, "learning_rate": 5.8725000000000004e-05, "loss": 0.0912, "step": 2350 }, { "grad_norm": 0.8128497004508972, "learning_rate": 5.8975000000000004e-05, "loss": 0.0885, "step": 2360 }, { "grad_norm": 0.9040464162826538, "learning_rate": 5.922500000000001e-05, "loss": 0.0947, "step": 2370 }, { "grad_norm": 0.9976675510406494, "learning_rate": 5.9475e-05, "loss": 0.1069, "step": 2380 }, { "grad_norm": 0.7290956974029541, "learning_rate": 5.9724999999999995e-05, "loss": 0.1039, "step": 2390 }, { "grad_norm": 0.8591521978378296, "learning_rate": 5.9975e-05, "loss": 0.0949, "step": 2400 }, { "grad_norm": 0.7053890228271484, "learning_rate": 6.0225e-05, "loss": 0.0803, "step": 2410 }, { "grad_norm": 0.7643669247627258, "learning_rate": 6.0475000000000006e-05, "loss": 0.1254, "step": 2420 }, { "grad_norm": 0.917387843132019, "learning_rate": 6.0725e-05, "loss": 0.105, "step": 2430 }, { "grad_norm": 0.830765426158905, "learning_rate": 6.0975000000000005e-05, "loss": 0.1053, "step": 2440 }, { "grad_norm": 0.8387424349784851, "learning_rate": 6.1225e-05, "loss": 0.0963, "step": 2450 }, { "grad_norm": 0.6299511194229126, "learning_rate": 6.1475e-05, "loss": 0.087, "step": 2460 }, { "grad_norm": 0.7173672914505005, "learning_rate": 6.1725e-05, "loss": 0.0816, "step": 2470 }, { "grad_norm": 0.7432805895805359, "learning_rate": 6.1975e-05, "loss": 0.1067, "step": 2480 }, { "grad_norm": 0.5724557638168335, "learning_rate": 6.2225e-05, "loss": 0.0776, "step": 2490 }, { "grad_norm": 0.5489363670349121, "learning_rate": 6.2475e-05, "loss": 0.0798, "step": 2500 }, { "grad_norm": 0.7103375196456909, "learning_rate": 6.2725e-05, "loss": 0.0872, "step": 2510 }, { "grad_norm": 0.7310965657234192, "learning_rate": 6.297500000000001e-05, "loss": 0.0853, "step": 2520 }, { "grad_norm": 0.680713951587677, "learning_rate": 6.3225e-05, "loss": 0.0813, "step": 2530 }, { "grad_norm": 1.1431649923324585, "learning_rate": 6.347500000000001e-05, "loss": 0.0885, "step": 2540 }, { "grad_norm": 0.6525793671607971, "learning_rate": 6.3725e-05, "loss": 0.0923, "step": 2550 }, { "grad_norm": 0.8122071027755737, "learning_rate": 6.397500000000001e-05, "loss": 0.0939, "step": 2560 }, { "grad_norm": 0.7443684339523315, "learning_rate": 6.4225e-05, "loss": 0.0886, "step": 2570 }, { "grad_norm": 0.6866037845611572, "learning_rate": 6.447500000000001e-05, "loss": 0.0899, "step": 2580 }, { "grad_norm": 0.7951505780220032, "learning_rate": 6.4725e-05, "loss": 0.0904, "step": 2590 }, { "grad_norm": 0.7457119226455688, "learning_rate": 6.497500000000001e-05, "loss": 0.1036, "step": 2600 }, { "grad_norm": 0.843637228012085, "learning_rate": 6.5225e-05, "loss": 0.0967, "step": 2610 }, { "grad_norm": 0.640144944190979, "learning_rate": 6.5475e-05, "loss": 0.078, "step": 2620 }, { "grad_norm": 0.8904373049736023, "learning_rate": 6.5725e-05, "loss": 0.0896, "step": 2630 }, { "grad_norm": 0.7321547865867615, "learning_rate": 6.5975e-05, "loss": 0.1025, "step": 2640 }, { "grad_norm": 0.5895065665245056, "learning_rate": 6.6225e-05, "loss": 0.0978, "step": 2650 }, { "grad_norm": 0.8046767711639404, "learning_rate": 6.6475e-05, "loss": 0.1009, "step": 2660 }, { "grad_norm": 0.5937095284461975, "learning_rate": 6.672500000000001e-05, "loss": 0.1103, "step": 2670 }, { "grad_norm": 0.8541538715362549, "learning_rate": 6.6975e-05, "loss": 0.1268, "step": 2680 }, { "grad_norm": 0.8476661443710327, "learning_rate": 6.722500000000001e-05, "loss": 0.0969, "step": 2690 }, { "grad_norm": 0.7658106684684753, "learning_rate": 6.7475e-05, "loss": 0.0918, "step": 2700 }, { "grad_norm": 1.126037836074829, "learning_rate": 6.7725e-05, "loss": 0.095, "step": 2710 }, { "grad_norm": 0.8006020784378052, "learning_rate": 6.7975e-05, "loss": 0.1119, "step": 2720 }, { "grad_norm": 0.8583477139472961, "learning_rate": 6.8225e-05, "loss": 0.0991, "step": 2730 }, { "grad_norm": 0.719208836555481, "learning_rate": 6.8475e-05, "loss": 0.0764, "step": 2740 }, { "grad_norm": 0.8519750237464905, "learning_rate": 6.8725e-05, "loss": 0.0931, "step": 2750 }, { "grad_norm": 0.9024448990821838, "learning_rate": 6.8975e-05, "loss": 0.0942, "step": 2760 }, { "grad_norm": 0.7875032424926758, "learning_rate": 6.9225e-05, "loss": 0.0933, "step": 2770 }, { "grad_norm": 0.48339444398880005, "learning_rate": 6.9475e-05, "loss": 0.101, "step": 2780 }, { "grad_norm": 0.6177260279655457, "learning_rate": 6.9725e-05, "loss": 0.0998, "step": 2790 }, { "grad_norm": 1.0914751291275024, "learning_rate": 6.997500000000001e-05, "loss": 0.0959, "step": 2800 }, { "grad_norm": 0.861923098564148, "learning_rate": 7.022500000000001e-05, "loss": 0.1198, "step": 2810 }, { "grad_norm": 0.8874877691268921, "learning_rate": 7.0475e-05, "loss": 0.0993, "step": 2820 }, { "grad_norm": 0.675304114818573, "learning_rate": 7.072500000000001e-05, "loss": 0.0822, "step": 2830 }, { "grad_norm": 0.7052448391914368, "learning_rate": 7.0975e-05, "loss": 0.1053, "step": 2840 }, { "grad_norm": 0.5610954761505127, "learning_rate": 7.122500000000001e-05, "loss": 0.1042, "step": 2850 }, { "grad_norm": 0.672646164894104, "learning_rate": 7.1475e-05, "loss": 0.0883, "step": 2860 }, { "grad_norm": 0.7421826124191284, "learning_rate": 7.172500000000001e-05, "loss": 0.0848, "step": 2870 }, { "grad_norm": 0.60381019115448, "learning_rate": 7.1975e-05, "loss": 0.0815, "step": 2880 }, { "grad_norm": 0.4684419333934784, "learning_rate": 7.2225e-05, "loss": 0.0828, "step": 2890 }, { "grad_norm": 0.6876832246780396, "learning_rate": 7.2475e-05, "loss": 0.1014, "step": 2900 }, { "grad_norm": 0.7535187602043152, "learning_rate": 7.272499999999999e-05, "loss": 0.073, "step": 2910 }, { "grad_norm": 0.5865347981452942, "learning_rate": 7.2975e-05, "loss": 0.0788, "step": 2920 }, { "grad_norm": 0.7541764974594116, "learning_rate": 7.3225e-05, "loss": 0.0737, "step": 2930 }, { "grad_norm": 0.7093668580055237, "learning_rate": 7.347500000000001e-05, "loss": 0.0739, "step": 2940 }, { "grad_norm": 0.6786952018737793, "learning_rate": 7.3725e-05, "loss": 0.0812, "step": 2950 }, { "grad_norm": 0.5627505779266357, "learning_rate": 7.397500000000001e-05, "loss": 0.0908, "step": 2960 }, { "grad_norm": 0.5887742042541504, "learning_rate": 7.4225e-05, "loss": 0.0707, "step": 2970 }, { "grad_norm": 0.7134391665458679, "learning_rate": 7.447500000000001e-05, "loss": 0.0782, "step": 2980 }, { "grad_norm": 0.6507739424705505, "learning_rate": 7.4725e-05, "loss": 0.0867, "step": 2990 }, { "grad_norm": 0.9310378432273865, "learning_rate": 7.4975e-05, "loss": 0.0982, "step": 3000 }, { "grad_norm": 0.5831695199012756, "learning_rate": 7.5225e-05, "loss": 0.0738, "step": 3010 }, { "grad_norm": 0.7623891234397888, "learning_rate": 7.5475e-05, "loss": 0.0831, "step": 3020 }, { "grad_norm": 0.7273319363594055, "learning_rate": 7.5725e-05, "loss": 0.078, "step": 3030 }, { "grad_norm": 0.583897590637207, "learning_rate": 7.5975e-05, "loss": 0.0775, "step": 3040 }, { "grad_norm": 0.5178344249725342, "learning_rate": 7.6225e-05, "loss": 0.076, "step": 3050 }, { "grad_norm": 0.6884287595748901, "learning_rate": 7.6475e-05, "loss": 0.0898, "step": 3060 }, { "grad_norm": 0.6292582154273987, "learning_rate": 7.672500000000001e-05, "loss": 0.0716, "step": 3070 }, { "grad_norm": 0.6683777570724487, "learning_rate": 7.697500000000001e-05, "loss": 0.0875, "step": 3080 }, { "grad_norm": 0.9661931991577148, "learning_rate": 7.722500000000001e-05, "loss": 0.1076, "step": 3090 }, { "grad_norm": 0.7924360632896423, "learning_rate": 7.747500000000001e-05, "loss": 0.0807, "step": 3100 }, { "grad_norm": 0.617585301399231, "learning_rate": 7.7725e-05, "loss": 0.0767, "step": 3110 }, { "grad_norm": 0.59670090675354, "learning_rate": 7.797500000000001e-05, "loss": 0.0642, "step": 3120 }, { "grad_norm": 0.5468007922172546, "learning_rate": 7.8225e-05, "loss": 0.0698, "step": 3130 }, { "grad_norm": 0.6968387961387634, "learning_rate": 7.8475e-05, "loss": 0.0818, "step": 3140 }, { "grad_norm": 0.9179096817970276, "learning_rate": 7.8725e-05, "loss": 0.0744, "step": 3150 }, { "grad_norm": 0.580878734588623, "learning_rate": 7.8975e-05, "loss": 0.0761, "step": 3160 }, { "grad_norm": 0.8833786249160767, "learning_rate": 7.9225e-05, "loss": 0.0896, "step": 3170 }, { "grad_norm": 0.7404505014419556, "learning_rate": 7.9475e-05, "loss": 0.1007, "step": 3180 }, { "grad_norm": 0.7698076367378235, "learning_rate": 7.9725e-05, "loss": 0.0919, "step": 3190 }, { "grad_norm": 0.6636810898780823, "learning_rate": 7.9975e-05, "loss": 0.087, "step": 3200 }, { "grad_norm": 0.5139772295951843, "learning_rate": 8.022500000000001e-05, "loss": 0.0744, "step": 3210 }, { "grad_norm": 0.8803701996803284, "learning_rate": 8.0475e-05, "loss": 0.1052, "step": 3220 }, { "grad_norm": 0.7891584038734436, "learning_rate": 8.072500000000001e-05, "loss": 0.0939, "step": 3230 }, { "grad_norm": 0.7983344197273254, "learning_rate": 8.0975e-05, "loss": 0.0799, "step": 3240 }, { "grad_norm": 0.6959712505340576, "learning_rate": 8.122500000000001e-05, "loss": 0.0963, "step": 3250 }, { "grad_norm": 0.5858649611473083, "learning_rate": 8.1475e-05, "loss": 0.0823, "step": 3260 }, { "grad_norm": 0.7714118361473083, "learning_rate": 8.172500000000001e-05, "loss": 0.0777, "step": 3270 }, { "grad_norm": 0.6253095865249634, "learning_rate": 8.1975e-05, "loss": 0.0725, "step": 3280 }, { "grad_norm": 0.7248762249946594, "learning_rate": 8.2225e-05, "loss": 0.0746, "step": 3290 }, { "grad_norm": 0.7176856994628906, "learning_rate": 8.2475e-05, "loss": 0.0728, "step": 3300 }, { "grad_norm": 0.5930078625679016, "learning_rate": 8.2725e-05, "loss": 0.0881, "step": 3310 }, { "grad_norm": 0.6257420182228088, "learning_rate": 8.2975e-05, "loss": 0.0976, "step": 3320 }, { "grad_norm": 0.5716971755027771, "learning_rate": 8.3225e-05, "loss": 0.085, "step": 3330 }, { "grad_norm": 0.844603955745697, "learning_rate": 8.347500000000001e-05, "loss": 0.0842, "step": 3340 }, { "grad_norm": 0.5920970439910889, "learning_rate": 8.3725e-05, "loss": 0.0812, "step": 3350 }, { "grad_norm": 0.8423042297363281, "learning_rate": 8.397500000000001e-05, "loss": 0.0772, "step": 3360 }, { "grad_norm": 0.5249737501144409, "learning_rate": 8.422500000000001e-05, "loss": 0.0711, "step": 3370 }, { "grad_norm": 0.8096702098846436, "learning_rate": 8.447500000000001e-05, "loss": 0.1226, "step": 3380 }, { "grad_norm": 0.8382686376571655, "learning_rate": 8.4725e-05, "loss": 0.084, "step": 3390 }, { "grad_norm": 0.709370493888855, "learning_rate": 8.4975e-05, "loss": 0.0887, "step": 3400 }, { "grad_norm": 0.6225737929344177, "learning_rate": 8.5225e-05, "loss": 0.083, "step": 3410 }, { "grad_norm": 0.6167056560516357, "learning_rate": 8.5475e-05, "loss": 0.0812, "step": 3420 }, { "grad_norm": 0.777795672416687, "learning_rate": 8.5725e-05, "loss": 0.0931, "step": 3430 }, { "grad_norm": 0.6035148501396179, "learning_rate": 8.5975e-05, "loss": 0.1002, "step": 3440 }, { "grad_norm": 0.6990535855293274, "learning_rate": 8.6225e-05, "loss": 0.0907, "step": 3450 }, { "grad_norm": 0.5029811263084412, "learning_rate": 8.6475e-05, "loss": 0.0686, "step": 3460 }, { "grad_norm": 0.6488916873931885, "learning_rate": 8.672500000000001e-05, "loss": 0.1034, "step": 3470 }, { "grad_norm": 0.49904489517211914, "learning_rate": 8.6975e-05, "loss": 0.0905, "step": 3480 }, { "grad_norm": 0.6010817289352417, "learning_rate": 8.7225e-05, "loss": 0.0796, "step": 3490 }, { "grad_norm": 0.6295519471168518, "learning_rate": 8.747500000000001e-05, "loss": 0.0637, "step": 3500 }, { "grad_norm": 0.5233983993530273, "learning_rate": 8.7725e-05, "loss": 0.0756, "step": 3510 }, { "grad_norm": 0.7221945524215698, "learning_rate": 8.797500000000001e-05, "loss": 0.0875, "step": 3520 }, { "grad_norm": 0.8312153816223145, "learning_rate": 8.8225e-05, "loss": 0.0906, "step": 3530 }, { "grad_norm": 0.6109550595283508, "learning_rate": 8.847500000000001e-05, "loss": 0.0858, "step": 3540 }, { "grad_norm": 0.5949515104293823, "learning_rate": 8.8725e-05, "loss": 0.0882, "step": 3550 }, { "grad_norm": 0.6358239054679871, "learning_rate": 8.897500000000001e-05, "loss": 0.0739, "step": 3560 }, { "grad_norm": 0.6631335616111755, "learning_rate": 8.9225e-05, "loss": 0.0849, "step": 3570 }, { "grad_norm": 0.6995726823806763, "learning_rate": 8.9475e-05, "loss": 0.0918, "step": 3580 }, { "grad_norm": 0.7257386445999146, "learning_rate": 8.9725e-05, "loss": 0.0915, "step": 3590 }, { "grad_norm": 0.6811807155609131, "learning_rate": 8.9975e-05, "loss": 0.086, "step": 3600 }, { "grad_norm": 0.7444417476654053, "learning_rate": 9.0225e-05, "loss": 0.0813, "step": 3610 }, { "grad_norm": 0.5769206881523132, "learning_rate": 9.0475e-05, "loss": 0.0729, "step": 3620 }, { "grad_norm": 0.5957974195480347, "learning_rate": 9.072500000000001e-05, "loss": 0.0724, "step": 3630 }, { "grad_norm": 0.5645696520805359, "learning_rate": 9.0975e-05, "loss": 0.0806, "step": 3640 }, { "grad_norm": 0.6803743839263916, "learning_rate": 9.122500000000001e-05, "loss": 0.0851, "step": 3650 }, { "grad_norm": 0.6295881867408752, "learning_rate": 9.1475e-05, "loss": 0.0774, "step": 3660 }, { "grad_norm": 0.7616122961044312, "learning_rate": 9.172500000000001e-05, "loss": 0.0879, "step": 3670 }, { "grad_norm": 0.5569307208061218, "learning_rate": 9.1975e-05, "loss": 0.0853, "step": 3680 }, { "grad_norm": 0.6064307689666748, "learning_rate": 9.2225e-05, "loss": 0.0743, "step": 3690 }, { "grad_norm": 0.7083445191383362, "learning_rate": 9.2475e-05, "loss": 0.0764, "step": 3700 }, { "grad_norm": 0.7213942408561707, "learning_rate": 9.2725e-05, "loss": 0.0778, "step": 3710 }, { "grad_norm": 0.6001719236373901, "learning_rate": 9.2975e-05, "loss": 0.0835, "step": 3720 }, { "grad_norm": 0.6074764132499695, "learning_rate": 9.3225e-05, "loss": 0.072, "step": 3730 }, { "grad_norm": 0.6571146845817566, "learning_rate": 9.3475e-05, "loss": 0.0723, "step": 3740 }, { "grad_norm": 0.6015981435775757, "learning_rate": 9.3725e-05, "loss": 0.0866, "step": 3750 }, { "grad_norm": 0.5687327980995178, "learning_rate": 9.397500000000001e-05, "loss": 0.0706, "step": 3760 }, { "grad_norm": 0.7409697771072388, "learning_rate": 9.422500000000001e-05, "loss": 0.0902, "step": 3770 }, { "grad_norm": 0.6957749724388123, "learning_rate": 9.4475e-05, "loss": 0.0818, "step": 3780 }, { "grad_norm": 0.7960202097892761, "learning_rate": 9.472500000000001e-05, "loss": 0.0791, "step": 3790 }, { "grad_norm": 0.5338627099990845, "learning_rate": 9.4975e-05, "loss": 0.1027, "step": 3800 }, { "grad_norm": 0.6785304546356201, "learning_rate": 9.522500000000001e-05, "loss": 0.0698, "step": 3810 }, { "grad_norm": 0.6707538366317749, "learning_rate": 9.5475e-05, "loss": 0.0849, "step": 3820 }, { "grad_norm": 0.5107866525650024, "learning_rate": 9.572500000000001e-05, "loss": 0.0675, "step": 3830 }, { "grad_norm": 0.44539278745651245, "learning_rate": 9.5975e-05, "loss": 0.0764, "step": 3840 }, { "grad_norm": 0.5804737210273743, "learning_rate": 9.622500000000001e-05, "loss": 0.0868, "step": 3850 }, { "grad_norm": 0.631111741065979, "learning_rate": 9.6475e-05, "loss": 0.0744, "step": 3860 }, { "grad_norm": 0.5413681268692017, "learning_rate": 9.6725e-05, "loss": 0.0701, "step": 3870 }, { "grad_norm": 0.5145949721336365, "learning_rate": 9.6975e-05, "loss": 0.066, "step": 3880 }, { "grad_norm": 0.6486348509788513, "learning_rate": 9.7225e-05, "loss": 0.0801, "step": 3890 }, { "grad_norm": 0.5532105565071106, "learning_rate": 9.747500000000001e-05, "loss": 0.0758, "step": 3900 }, { "grad_norm": 0.5784058570861816, "learning_rate": 9.7725e-05, "loss": 0.0833, "step": 3910 }, { "grad_norm": 0.7483074069023132, "learning_rate": 9.797500000000001e-05, "loss": 0.0937, "step": 3920 }, { "grad_norm": 0.6623482704162598, "learning_rate": 9.8225e-05, "loss": 0.0814, "step": 3930 }, { "grad_norm": 0.6479174494743347, "learning_rate": 9.847500000000001e-05, "loss": 0.0916, "step": 3940 }, { "grad_norm": 0.6400222778320312, "learning_rate": 9.8725e-05, "loss": 0.0808, "step": 3950 }, { "grad_norm": 0.623907208442688, "learning_rate": 9.897500000000001e-05, "loss": 0.0886, "step": 3960 }, { "grad_norm": 0.7552863359451294, "learning_rate": 9.9225e-05, "loss": 0.0733, "step": 3970 }, { "grad_norm": 0.5994132161140442, "learning_rate": 9.9475e-05, "loss": 0.0804, "step": 3980 }, { "grad_norm": 0.738457977771759, "learning_rate": 9.9725e-05, "loss": 0.0867, "step": 3990 }, { "grad_norm": 0.7381294369697571, "learning_rate": 9.9975e-05, "loss": 0.0901, "step": 4000 }, { "grad_norm": 0.5476551651954651, "learning_rate": 9.999999653982884e-05, "loss": 0.0773, "step": 4010 }, { "grad_norm": 1.1369831562042236, "learning_rate": 9.999998457874392e-05, "loss": 0.0894, "step": 4020 }, { "grad_norm": 0.5373391509056091, "learning_rate": 9.999996407402913e-05, "loss": 0.091, "step": 4030 }, { "grad_norm": 0.4111018478870392, "learning_rate": 9.999993502568801e-05, "loss": 0.0665, "step": 4040 }, { "grad_norm": 0.6133153438568115, "learning_rate": 9.999989743372548e-05, "loss": 0.0777, "step": 4050 }, { "grad_norm": 0.6238013505935669, "learning_rate": 9.999985129814798e-05, "loss": 0.0733, "step": 4060 }, { "grad_norm": 0.5729981660842896, "learning_rate": 9.99997966189634e-05, "loss": 0.066, "step": 4070 }, { "grad_norm": 0.4398922324180603, "learning_rate": 9.999973339618107e-05, "loss": 0.0752, "step": 4080 }, { "grad_norm": 0.5192022323608398, "learning_rate": 9.999966162981179e-05, "loss": 0.07, "step": 4090 }, { "grad_norm": 0.6605653762817383, "learning_rate": 9.999958131986784e-05, "loss": 0.0797, "step": 4100 }, { "grad_norm": 0.6894550919532776, "learning_rate": 9.999949246636293e-05, "loss": 0.0847, "step": 4110 }, { "grad_norm": 0.6021091341972351, "learning_rate": 9.999939506931224e-05, "loss": 0.0912, "step": 4120 }, { "grad_norm": 0.6216923594474792, "learning_rate": 9.999928912873243e-05, "loss": 0.0749, "step": 4130 }, { "grad_norm": 0.4714604914188385, "learning_rate": 9.999917464464159e-05, "loss": 0.0724, "step": 4140 }, { "grad_norm": 0.5209283828735352, "learning_rate": 9.999905161705929e-05, "loss": 0.087, "step": 4150 }, { "grad_norm": 0.5970317125320435, "learning_rate": 9.999892004600653e-05, "loss": 0.0672, "step": 4160 }, { "grad_norm": 0.6313736438751221, "learning_rate": 9.999877993150581e-05, "loss": 0.074, "step": 4170 }, { "grad_norm": 0.460879385471344, "learning_rate": 9.999863127358108e-05, "loss": 0.0673, "step": 4180 }, { "grad_norm": 0.5210947394371033, "learning_rate": 9.999847407225773e-05, "loss": 0.0722, "step": 4190 }, { "grad_norm": 0.6232216358184814, "learning_rate": 9.999830832756262e-05, "loss": 0.0691, "step": 4200 }, { "grad_norm": 0.579931378364563, "learning_rate": 9.999813403952407e-05, "loss": 0.0702, "step": 4210 }, { "grad_norm": 0.4860551655292511, "learning_rate": 9.999795120817187e-05, "loss": 0.0778, "step": 4220 }, { "grad_norm": 0.6484834551811218, "learning_rate": 9.999775983353725e-05, "loss": 0.0931, "step": 4230 }, { "grad_norm": 0.45878586173057556, "learning_rate": 9.999755991565292e-05, "loss": 0.0924, "step": 4240 }, { "grad_norm": 0.4218076467514038, "learning_rate": 9.999735145455303e-05, "loss": 0.0821, "step": 4250 }, { "grad_norm": 0.5103121399879456, "learning_rate": 9.99971344502732e-05, "loss": 0.0857, "step": 4260 }, { "grad_norm": 0.5252031087875366, "learning_rate": 9.999690890285053e-05, "loss": 0.0747, "step": 4270 }, { "grad_norm": 1.0673086643218994, "learning_rate": 9.999667481232356e-05, "loss": 0.0657, "step": 4280 }, { "grad_norm": 0.5390567779541016, "learning_rate": 9.999643217873225e-05, "loss": 0.0644, "step": 4290 }, { "grad_norm": 0.5157567262649536, "learning_rate": 9.999618100211809e-05, "loss": 0.0664, "step": 4300 }, { "grad_norm": 0.6993318796157837, "learning_rate": 9.999592128252402e-05, "loss": 0.0703, "step": 4310 }, { "grad_norm": 0.5503459572792053, "learning_rate": 9.999565301999437e-05, "loss": 0.0974, "step": 4320 }, { "grad_norm": 0.4288609027862549, "learning_rate": 9.999537621457502e-05, "loss": 0.0688, "step": 4330 }, { "grad_norm": 0.5724561214447021, "learning_rate": 9.999509086631323e-05, "loss": 0.0662, "step": 4340 }, { "grad_norm": 0.41323956847190857, "learning_rate": 9.99947969752578e-05, "loss": 0.0613, "step": 4350 }, { "grad_norm": 0.44669052958488464, "learning_rate": 9.999449454145891e-05, "loss": 0.0611, "step": 4360 }, { "grad_norm": 0.539537787437439, "learning_rate": 9.999418356496827e-05, "loss": 0.0673, "step": 4370 }, { "grad_norm": 0.540908932685852, "learning_rate": 9.999386404583899e-05, "loss": 0.0761, "step": 4380 }, { "grad_norm": 0.5297500491142273, "learning_rate": 9.999353598412568e-05, "loss": 0.0689, "step": 4390 }, { "grad_norm": 0.5488400459289551, "learning_rate": 9.999319937988442e-05, "loss": 0.0871, "step": 4400 }, { "grad_norm": 0.5727903842926025, "learning_rate": 9.999285423317268e-05, "loss": 0.0652, "step": 4410 }, { "grad_norm": 0.5654491782188416, "learning_rate": 9.999250054404947e-05, "loss": 0.0896, "step": 4420 }, { "grad_norm": 0.5964249968528748, "learning_rate": 9.99921383125752e-05, "loss": 0.0831, "step": 4430 }, { "grad_norm": 0.49736884236335754, "learning_rate": 9.99917675388118e-05, "loss": 0.0788, "step": 4440 }, { "grad_norm": 0.5396902561187744, "learning_rate": 9.99913882228226e-05, "loss": 0.0868, "step": 4450 }, { "grad_norm": 0.5895163416862488, "learning_rate": 9.999100036467242e-05, "loss": 0.0719, "step": 4460 }, { "grad_norm": 0.4435919225215912, "learning_rate": 9.999060396442753e-05, "loss": 0.0711, "step": 4470 }, { "grad_norm": 0.5895199775695801, "learning_rate": 9.999019902215566e-05, "loss": 0.0788, "step": 4480 }, { "grad_norm": 0.6485477685928345, "learning_rate": 9.998978553792602e-05, "loss": 0.0753, "step": 4490 }, { "grad_norm": 0.5787511467933655, "learning_rate": 9.998936351180926e-05, "loss": 0.0752, "step": 4500 }, { "grad_norm": 0.5784372091293335, "learning_rate": 9.998893294387747e-05, "loss": 0.0911, "step": 4510 }, { "grad_norm": 0.47066962718963623, "learning_rate": 9.998849383420426e-05, "loss": 0.0707, "step": 4520 }, { "grad_norm": 0.5301247239112854, "learning_rate": 9.998804618286465e-05, "loss": 0.0698, "step": 4530 }, { "grad_norm": 0.5723125338554382, "learning_rate": 9.99875899899351e-05, "loss": 0.0823, "step": 4540 }, { "grad_norm": 0.6269090175628662, "learning_rate": 9.99871252554936e-05, "loss": 0.0855, "step": 4550 }, { "grad_norm": 0.5343359708786011, "learning_rate": 9.998665197961955e-05, "loss": 0.0717, "step": 4560 }, { "grad_norm": 0.5234417915344238, "learning_rate": 9.998617016239379e-05, "loss": 0.0778, "step": 4570 }, { "grad_norm": 0.6417948603630066, "learning_rate": 9.998567980389869e-05, "loss": 0.0723, "step": 4580 }, { "grad_norm": 0.464377760887146, "learning_rate": 9.998518090421802e-05, "loss": 0.0756, "step": 4590 }, { "grad_norm": 0.5526030659675598, "learning_rate": 9.998467346343703e-05, "loss": 0.0718, "step": 4600 }, { "grad_norm": 0.720876932144165, "learning_rate": 9.998415748164243e-05, "loss": 0.073, "step": 4610 }, { "grad_norm": 0.5605248212814331, "learning_rate": 9.998363295892238e-05, "loss": 0.0642, "step": 4620 }, { "grad_norm": 0.6542764902114868, "learning_rate": 9.998309989536652e-05, "loss": 0.0764, "step": 4630 }, { "grad_norm": 0.5314453840255737, "learning_rate": 9.998255829106593e-05, "loss": 0.0789, "step": 4640 }, { "grad_norm": 0.46513301134109497, "learning_rate": 9.998200814611316e-05, "loss": 0.0653, "step": 4650 }, { "grad_norm": 0.4885835349559784, "learning_rate": 9.998144946060219e-05, "loss": 0.0715, "step": 4660 }, { "grad_norm": 0.6612483859062195, "learning_rate": 9.998088223462852e-05, "loss": 0.0693, "step": 4670 }, { "grad_norm": 0.5539016723632812, "learning_rate": 9.998030646828905e-05, "loss": 0.0648, "step": 4680 }, { "grad_norm": 0.4756821393966675, "learning_rate": 9.997972216168217e-05, "loss": 0.0898, "step": 4690 }, { "grad_norm": 0.5500847697257996, "learning_rate": 9.997912931490771e-05, "loss": 0.0685, "step": 4700 }, { "grad_norm": 0.5684693455696106, "learning_rate": 9.9978527928067e-05, "loss": 0.0667, "step": 4710 }, { "grad_norm": 0.681466817855835, "learning_rate": 9.997791800126277e-05, "loss": 0.0632, "step": 4720 }, { "grad_norm": 0.8963981866836548, "learning_rate": 9.997729953459927e-05, "loss": 0.0829, "step": 4730 }, { "grad_norm": 0.7072616219520569, "learning_rate": 9.997667252818214e-05, "loss": 0.073, "step": 4740 }, { "grad_norm": 0.4385315477848053, "learning_rate": 9.997603698211855e-05, "loss": 0.0753, "step": 4750 }, { "grad_norm": 0.5545654892921448, "learning_rate": 9.99753928965171e-05, "loss": 0.0733, "step": 4760 }, { "grad_norm": 0.5754374861717224, "learning_rate": 9.997474027148781e-05, "loss": 0.0793, "step": 4770 }, { "grad_norm": 0.6834706664085388, "learning_rate": 9.997407910714223e-05, "loss": 0.0914, "step": 4780 }, { "grad_norm": 0.5543652176856995, "learning_rate": 9.997340940359332e-05, "loss": 0.0722, "step": 4790 }, { "grad_norm": 0.3383401334285736, "learning_rate": 9.997273116095552e-05, "loss": 0.0746, "step": 4800 }, { "grad_norm": 0.7580839991569519, "learning_rate": 9.997204437934473e-05, "loss": 0.0692, "step": 4810 }, { "grad_norm": 0.5374947786331177, "learning_rate": 9.997134905887829e-05, "loss": 0.0765, "step": 4820 }, { "grad_norm": 0.6061835885047913, "learning_rate": 9.997064519967501e-05, "loss": 0.0735, "step": 4830 }, { "grad_norm": 0.5358559489250183, "learning_rate": 9.996993280185517e-05, "loss": 0.0652, "step": 4840 }, { "grad_norm": 0.532904863357544, "learning_rate": 9.99692118655405e-05, "loss": 0.0652, "step": 4850 }, { "grad_norm": 0.5067242383956909, "learning_rate": 9.996848239085417e-05, "loss": 0.0632, "step": 4860 }, { "grad_norm": 0.45449939370155334, "learning_rate": 9.996774437792085e-05, "loss": 0.0663, "step": 4870 }, { "grad_norm": 0.47688454389572144, "learning_rate": 9.996699782686664e-05, "loss": 0.0567, "step": 4880 }, { "grad_norm": 0.5608693361282349, "learning_rate": 9.996624273781909e-05, "loss": 0.0808, "step": 4890 }, { "grad_norm": 0.41938316822052, "learning_rate": 9.996547911090725e-05, "loss": 0.0578, "step": 4900 }, { "grad_norm": 0.49188193678855896, "learning_rate": 9.996470694626157e-05, "loss": 0.0655, "step": 4910 }, { "grad_norm": 0.53224116563797, "learning_rate": 9.996392624401403e-05, "loss": 0.0689, "step": 4920 }, { "grad_norm": 0.42857202887535095, "learning_rate": 9.996313700429801e-05, "loss": 0.0635, "step": 4930 }, { "grad_norm": 0.39667418599128723, "learning_rate": 9.996233922724836e-05, "loss": 0.0536, "step": 4940 }, { "grad_norm": 0.6215639710426331, "learning_rate": 9.996153291300141e-05, "loss": 0.067, "step": 4950 }, { "grad_norm": 0.359880268573761, "learning_rate": 9.996071806169494e-05, "loss": 0.0632, "step": 4960 }, { "grad_norm": 0.7162643074989319, "learning_rate": 9.995989467346817e-05, "loss": 0.0812, "step": 4970 }, { "grad_norm": 0.409749299287796, "learning_rate": 9.995906274846183e-05, "loss": 0.0687, "step": 4980 }, { "grad_norm": 0.5297077894210815, "learning_rate": 9.995822228681803e-05, "loss": 0.0632, "step": 4990 }, { "grad_norm": 0.5239155292510986, "learning_rate": 9.99573732886804e-05, "loss": 0.0784, "step": 5000 }, { "grad_norm": 0.5372620224952698, "learning_rate": 9.995651575419402e-05, "loss": 0.0685, "step": 5010 }, { "grad_norm": 0.5811275839805603, "learning_rate": 9.995564968350541e-05, "loss": 0.0705, "step": 5020 }, { "grad_norm": 0.7433735132217407, "learning_rate": 9.995477507676256e-05, "loss": 0.065, "step": 5030 }, { "grad_norm": 0.6026665568351746, "learning_rate": 9.995389193411493e-05, "loss": 0.0676, "step": 5040 }, { "grad_norm": 0.4114062786102295, "learning_rate": 9.995300025571339e-05, "loss": 0.0752, "step": 5050 }, { "grad_norm": 0.5899551510810852, "learning_rate": 9.995210004171034e-05, "loss": 0.0565, "step": 5060 }, { "grad_norm": 0.6340254545211792, "learning_rate": 9.995119129225956e-05, "loss": 0.0815, "step": 5070 }, { "grad_norm": 0.5717336535453796, "learning_rate": 9.995027400751637e-05, "loss": 0.0753, "step": 5080 }, { "grad_norm": 0.49747952818870544, "learning_rate": 9.994934818763751e-05, "loss": 0.0591, "step": 5090 }, { "grad_norm": 0.6199728846549988, "learning_rate": 9.994841383278115e-05, "loss": 0.0689, "step": 5100 }, { "grad_norm": 0.5950746536254883, "learning_rate": 9.994747094310695e-05, "loss": 0.0762, "step": 5110 }, { "grad_norm": 0.4983644485473633, "learning_rate": 9.994651951877604e-05, "loss": 0.0873, "step": 5120 }, { "grad_norm": 0.6864198446273804, "learning_rate": 9.994555955995099e-05, "loss": 0.0767, "step": 5130 }, { "grad_norm": 0.5035277009010315, "learning_rate": 9.994459106679581e-05, "loss": 0.0709, "step": 5140 }, { "grad_norm": 0.7754384279251099, "learning_rate": 9.994361403947603e-05, "loss": 0.0586, "step": 5150 }, { "grad_norm": 0.486234188079834, "learning_rate": 9.994262847815854e-05, "loss": 0.0959, "step": 5160 }, { "grad_norm": 0.5189380049705505, "learning_rate": 9.99416343830118e-05, "loss": 0.0729, "step": 5170 }, { "grad_norm": 0.44119009375572205, "learning_rate": 9.994063175420565e-05, "loss": 0.0597, "step": 5180 }, { "grad_norm": 0.5307364463806152, "learning_rate": 9.99396205919114e-05, "loss": 0.0846, "step": 5190 }, { "grad_norm": 0.4302844703197479, "learning_rate": 9.993860089630185e-05, "loss": 0.0611, "step": 5200 }, { "grad_norm": 0.7849263548851013, "learning_rate": 9.993757266755123e-05, "loss": 0.0653, "step": 5210 }, { "grad_norm": 0.5272495150566101, "learning_rate": 9.993653590583522e-05, "loss": 0.0667, "step": 5220 }, { "grad_norm": 0.5266837477684021, "learning_rate": 9.993549061133102e-05, "loss": 0.0593, "step": 5230 }, { "grad_norm": 0.48270726203918457, "learning_rate": 9.993443678421719e-05, "loss": 0.0735, "step": 5240 }, { "grad_norm": 0.5952420234680176, "learning_rate": 9.993337442467384e-05, "loss": 0.0776, "step": 5250 }, { "grad_norm": 0.5688992738723755, "learning_rate": 9.993230353288248e-05, "loss": 0.0587, "step": 5260 }, { "grad_norm": 0.647667407989502, "learning_rate": 9.993122410902608e-05, "loss": 0.0674, "step": 5270 }, { "grad_norm": 0.4862939119338989, "learning_rate": 9.993013615328912e-05, "loss": 0.0672, "step": 5280 }, { "grad_norm": 0.41371843218803406, "learning_rate": 9.992903966585747e-05, "loss": 0.0647, "step": 5290 }, { "grad_norm": 0.42250362038612366, "learning_rate": 9.992793464691852e-05, "loss": 0.0582, "step": 5300 }, { "grad_norm": 0.6793314218521118, "learning_rate": 9.992682109666105e-05, "loss": 0.0539, "step": 5310 }, { "grad_norm": 0.37892210483551025, "learning_rate": 9.992569901527538e-05, "loss": 0.0616, "step": 5320 }, { "grad_norm": 0.5314620733261108, "learning_rate": 9.99245684029532e-05, "loss": 0.0748, "step": 5330 }, { "grad_norm": 0.5427423715591431, "learning_rate": 9.992342925988774e-05, "loss": 0.0639, "step": 5340 }, { "grad_norm": 0.4645208418369293, "learning_rate": 9.992228158627361e-05, "loss": 0.0553, "step": 5350 }, { "grad_norm": 0.49694910645484924, "learning_rate": 9.992112538230693e-05, "loss": 0.0674, "step": 5360 }, { "grad_norm": 0.4451981782913208, "learning_rate": 9.991996064818527e-05, "loss": 0.075, "step": 5370 }, { "grad_norm": 0.5257868766784668, "learning_rate": 9.991878738410768e-05, "loss": 0.072, "step": 5380 }, { "grad_norm": 0.5505748391151428, "learning_rate": 9.991760559027457e-05, "loss": 0.0613, "step": 5390 }, { "grad_norm": 0.5523684024810791, "learning_rate": 9.991641526688793e-05, "loss": 0.0615, "step": 5400 }, { "grad_norm": 0.4406563937664032, "learning_rate": 9.991521641415113e-05, "loss": 0.0615, "step": 5410 }, { "grad_norm": 0.6145594120025635, "learning_rate": 9.991400903226904e-05, "loss": 0.0673, "step": 5420 }, { "grad_norm": 0.6921329498291016, "learning_rate": 9.991279312144794e-05, "loss": 0.0686, "step": 5430 }, { "grad_norm": 0.37732645869255066, "learning_rate": 9.991156868189564e-05, "loss": 0.0739, "step": 5440 }, { "grad_norm": 0.4178144633769989, "learning_rate": 9.991033571382131e-05, "loss": 0.0516, "step": 5450 }, { "grad_norm": 0.720687747001648, "learning_rate": 9.990909421743569e-05, "loss": 0.0547, "step": 5460 }, { "grad_norm": 0.48502418398857117, "learning_rate": 9.990784419295085e-05, "loss": 0.0603, "step": 5470 }, { "grad_norm": 0.4466925263404846, "learning_rate": 9.990658564058044e-05, "loss": 0.0552, "step": 5480 }, { "grad_norm": 0.5351961851119995, "learning_rate": 9.990531856053948e-05, "loss": 0.0584, "step": 5490 }, { "grad_norm": 0.34217706322669983, "learning_rate": 9.99040429530445e-05, "loss": 0.0546, "step": 5500 }, { "grad_norm": 0.4230134189128876, "learning_rate": 9.990275881831346e-05, "loss": 0.059, "step": 5510 }, { "grad_norm": 0.36913546919822693, "learning_rate": 9.990146615656577e-05, "loss": 0.0529, "step": 5520 }, { "grad_norm": 0.4904607832431793, "learning_rate": 9.990016496802233e-05, "loss": 0.0614, "step": 5530 }, { "grad_norm": 0.3327493369579315, "learning_rate": 9.989885525290548e-05, "loss": 0.0527, "step": 5540 }, { "grad_norm": 0.6702595949172974, "learning_rate": 9.989753701143897e-05, "loss": 0.0611, "step": 5550 }, { "grad_norm": 0.4877476990222931, "learning_rate": 9.989621024384812e-05, "loss": 0.057, "step": 5560 }, { "grad_norm": 0.4400719106197357, "learning_rate": 9.989487495035959e-05, "loss": 0.0583, "step": 5570 }, { "grad_norm": 0.6307154893875122, "learning_rate": 9.989353113120156e-05, "loss": 0.0717, "step": 5580 }, { "grad_norm": 0.571216881275177, "learning_rate": 9.989217878660366e-05, "loss": 0.0573, "step": 5590 }, { "grad_norm": 0.4965840280056, "learning_rate": 9.989081791679695e-05, "loss": 0.0646, "step": 5600 }, { "grad_norm": 0.36924242973327637, "learning_rate": 9.988944852201397e-05, "loss": 0.0565, "step": 5610 }, { "grad_norm": 0.4087572693824768, "learning_rate": 9.988807060248873e-05, "loss": 0.0502, "step": 5620 }, { "grad_norm": 0.5647485256195068, "learning_rate": 9.988668415845665e-05, "loss": 0.0447, "step": 5630 }, { "grad_norm": 0.6256299018859863, "learning_rate": 9.988528919015466e-05, "loss": 0.0584, "step": 5640 }, { "grad_norm": 0.5265887379646301, "learning_rate": 9.988388569782112e-05, "loss": 0.0825, "step": 5650 }, { "grad_norm": 0.5373584032058716, "learning_rate": 9.988247368169583e-05, "loss": 0.0636, "step": 5660 }, { "grad_norm": 0.5365583896636963, "learning_rate": 9.988105314202007e-05, "loss": 0.0612, "step": 5670 }, { "grad_norm": 0.48864662647247314, "learning_rate": 9.987962407903659e-05, "loss": 0.0652, "step": 5680 }, { "grad_norm": 0.490559458732605, "learning_rate": 9.987818649298957e-05, "loss": 0.0577, "step": 5690 }, { "grad_norm": 0.4892744719982147, "learning_rate": 9.987674038412465e-05, "loss": 0.0596, "step": 5700 }, { "grad_norm": 0.46483537554740906, "learning_rate": 9.987528575268891e-05, "loss": 0.0693, "step": 5710 }, { "grad_norm": 0.5126171708106995, "learning_rate": 9.987382259893095e-05, "loss": 0.0606, "step": 5720 }, { "grad_norm": 0.4780590534210205, "learning_rate": 9.987235092310074e-05, "loss": 0.0666, "step": 5730 }, { "grad_norm": 0.45442256331443787, "learning_rate": 9.987087072544978e-05, "loss": 0.0638, "step": 5740 }, { "grad_norm": 0.39281365275382996, "learning_rate": 9.9869382006231e-05, "loss": 0.0587, "step": 5750 }, { "grad_norm": 0.4683944284915924, "learning_rate": 9.986788476569875e-05, "loss": 0.0563, "step": 5760 }, { "grad_norm": 0.3872678279876709, "learning_rate": 9.986637900410887e-05, "loss": 0.0559, "step": 5770 }, { "grad_norm": 0.4279055595397949, "learning_rate": 9.986486472171869e-05, "loss": 0.0632, "step": 5780 }, { "grad_norm": 0.5905216932296753, "learning_rate": 9.986334191878692e-05, "loss": 0.0509, "step": 5790 }, { "grad_norm": 0.5049157738685608, "learning_rate": 9.986181059557378e-05, "loss": 0.0661, "step": 5800 }, { "grad_norm": 0.4965856075286865, "learning_rate": 9.986027075234094e-05, "loss": 0.0718, "step": 5810 }, { "grad_norm": 0.3931066393852234, "learning_rate": 9.985872238935152e-05, "loss": 0.0563, "step": 5820 }, { "grad_norm": 0.48457834124565125, "learning_rate": 9.985716550687008e-05, "loss": 0.0626, "step": 5830 }, { "grad_norm": 0.4383699893951416, "learning_rate": 9.985560010516264e-05, "loss": 0.0578, "step": 5840 }, { "grad_norm": 0.39937883615493774, "learning_rate": 9.985402618449668e-05, "loss": 0.0529, "step": 5850 }, { "grad_norm": 0.6789182424545288, "learning_rate": 9.985244374514118e-05, "loss": 0.0714, "step": 5860 }, { "grad_norm": 0.4339824616909027, "learning_rate": 9.985085278736651e-05, "loss": 0.0665, "step": 5870 }, { "grad_norm": 0.5643037557601929, "learning_rate": 9.984925331144452e-05, "loss": 0.0703, "step": 5880 }, { "grad_norm": 0.4280531406402588, "learning_rate": 9.984764531764851e-05, "loss": 0.0642, "step": 5890 }, { "grad_norm": 0.47062763571739197, "learning_rate": 9.984602880625326e-05, "loss": 0.0725, "step": 5900 }, { "grad_norm": 0.40518510341644287, "learning_rate": 9.9844403777535e-05, "loss": 0.0673, "step": 5910 }, { "grad_norm": 0.41199472546577454, "learning_rate": 9.984277023177135e-05, "loss": 0.0507, "step": 5920 }, { "grad_norm": 0.47888314723968506, "learning_rate": 9.984112816924148e-05, "loss": 0.0564, "step": 5930 }, { "grad_norm": 0.4528254568576813, "learning_rate": 9.983947759022596e-05, "loss": 0.0607, "step": 5940 }, { "grad_norm": 0.5218254923820496, "learning_rate": 9.983781849500682e-05, "loss": 0.0566, "step": 5950 }, { "grad_norm": 0.4917014241218567, "learning_rate": 9.98361508838676e-05, "loss": 0.0508, "step": 5960 }, { "grad_norm": 0.5300503373146057, "learning_rate": 9.98344747570932e-05, "loss": 0.0618, "step": 5970 }, { "grad_norm": 0.4868715703487396, "learning_rate": 9.983279011497004e-05, "loss": 0.0642, "step": 5980 }, { "grad_norm": 0.4084424078464508, "learning_rate": 9.983109695778596e-05, "loss": 0.0656, "step": 5990 }, { "grad_norm": 0.43009403347969055, "learning_rate": 9.982939528583032e-05, "loss": 0.0796, "step": 6000 }, { "grad_norm": 0.534338653087616, "learning_rate": 9.982768509939385e-05, "loss": 0.0538, "step": 6010 }, { "grad_norm": 0.6411554217338562, "learning_rate": 9.982596639876879e-05, "loss": 0.0607, "step": 6020 }, { "grad_norm": 0.5079619884490967, "learning_rate": 9.982423918424881e-05, "loss": 0.0632, "step": 6030 }, { "grad_norm": 0.5552597045898438, "learning_rate": 9.982250345612908e-05, "loss": 0.0598, "step": 6040 }, { "grad_norm": 0.47022074460983276, "learning_rate": 9.982075921470611e-05, "loss": 0.0643, "step": 6050 }, { "grad_norm": 1.117583155632019, "learning_rate": 9.981900646027802e-05, "loss": 0.0629, "step": 6060 }, { "grad_norm": 0.5041999816894531, "learning_rate": 9.981724519314425e-05, "loss": 0.0669, "step": 6070 }, { "grad_norm": 0.6160788536071777, "learning_rate": 9.981547541360581e-05, "loss": 0.0629, "step": 6080 }, { "grad_norm": 0.4435524642467499, "learning_rate": 9.981369712196508e-05, "loss": 0.0673, "step": 6090 }, { "grad_norm": 0.49114254117012024, "learning_rate": 9.981191031852592e-05, "loss": 0.0664, "step": 6100 }, { "grad_norm": 0.5604746341705322, "learning_rate": 9.981011500359362e-05, "loss": 0.0526, "step": 6110 }, { "grad_norm": 0.46365615725517273, "learning_rate": 9.9808311177475e-05, "loss": 0.0649, "step": 6120 }, { "grad_norm": 0.5333303213119507, "learning_rate": 9.980649884047826e-05, "loss": 0.0436, "step": 6130 }, { "grad_norm": 0.32690221071243286, "learning_rate": 9.980467799291307e-05, "loss": 0.0557, "step": 6140 }, { "grad_norm": 0.43656450510025024, "learning_rate": 9.980284863509058e-05, "loss": 0.0714, "step": 6150 }, { "grad_norm": 0.5131890773773193, "learning_rate": 9.980101076732334e-05, "loss": 0.0707, "step": 6160 }, { "grad_norm": 0.41466617584228516, "learning_rate": 9.979916438992544e-05, "loss": 0.0716, "step": 6170 }, { "grad_norm": 0.4157552421092987, "learning_rate": 9.979730950321237e-05, "loss": 0.0543, "step": 6180 }, { "grad_norm": 0.41153571009635925, "learning_rate": 9.979544610750104e-05, "loss": 0.0717, "step": 6190 }, { "grad_norm": 0.66008061170578, "learning_rate": 9.97935742031099e-05, "loss": 0.0543, "step": 6200 }, { "grad_norm": 0.5267980694770813, "learning_rate": 9.979169379035878e-05, "loss": 0.0626, "step": 6210 }, { "grad_norm": 1.9233500957489014, "learning_rate": 9.978980486956899e-05, "loss": 0.0626, "step": 6220 }, { "grad_norm": 0.8111964464187622, "learning_rate": 9.978790744106332e-05, "loss": 0.0675, "step": 6230 }, { "grad_norm": 0.44212910532951355, "learning_rate": 9.978600150516594e-05, "loss": 0.0678, "step": 6240 }, { "grad_norm": 0.45173004269599915, "learning_rate": 9.978408706220259e-05, "loss": 0.061, "step": 6250 }, { "grad_norm": 0.4485941231250763, "learning_rate": 9.978216411250032e-05, "loss": 0.0454, "step": 6260 }, { "grad_norm": 0.42014047503471375, "learning_rate": 9.978023265638778e-05, "loss": 0.0576, "step": 6270 }, { "grad_norm": 0.4722181260585785, "learning_rate": 9.977829269419495e-05, "loss": 0.0688, "step": 6280 }, { "grad_norm": 0.5641476511955261, "learning_rate": 9.977634422625335e-05, "loss": 0.0578, "step": 6290 }, { "grad_norm": 0.5017063617706299, "learning_rate": 9.97743872528959e-05, "loss": 0.0658, "step": 6300 }, { "grad_norm": 0.5227478742599487, "learning_rate": 9.9772421774457e-05, "loss": 0.0608, "step": 6310 }, { "grad_norm": 0.5234737992286682, "learning_rate": 9.977044779127252e-05, "loss": 0.0615, "step": 6320 }, { "grad_norm": 0.42470496892929077, "learning_rate": 9.976846530367971e-05, "loss": 0.0515, "step": 6330 }, { "grad_norm": 0.5895519852638245, "learning_rate": 9.976647431201735e-05, "loss": 0.0704, "step": 6340 }, { "grad_norm": 0.5000032782554626, "learning_rate": 9.976447481662568e-05, "loss": 0.0599, "step": 6350 }, { "grad_norm": 0.4164268374443054, "learning_rate": 9.976246681784629e-05, "loss": 0.0644, "step": 6360 }, { "grad_norm": 0.5093873739242554, "learning_rate": 9.976045031602234e-05, "loss": 0.0528, "step": 6370 }, { "grad_norm": 0.4301096200942993, "learning_rate": 9.975842531149837e-05, "loss": 0.054, "step": 6380 }, { "grad_norm": 0.4822219908237457, "learning_rate": 9.975639180462043e-05, "loss": 0.0581, "step": 6390 }, { "grad_norm": 0.3846195340156555, "learning_rate": 9.975434979573596e-05, "loss": 0.0686, "step": 6400 }, { "grad_norm": 0.39010703563690186, "learning_rate": 9.97522992851939e-05, "loss": 0.0435, "step": 6410 }, { "grad_norm": 0.4002074897289276, "learning_rate": 9.975024027334461e-05, "loss": 0.0528, "step": 6420 }, { "grad_norm": 0.43588533997535706, "learning_rate": 9.974817276053993e-05, "loss": 0.0494, "step": 6430 }, { "grad_norm": 0.465758740901947, "learning_rate": 9.974609674713315e-05, "loss": 0.049, "step": 6440 }, { "grad_norm": 0.4168603718280792, "learning_rate": 9.9744012233479e-05, "loss": 0.0484, "step": 6450 }, { "grad_norm": 0.3158040940761566, "learning_rate": 9.974191921993366e-05, "loss": 0.0522, "step": 6460 }, { "grad_norm": 0.5963517427444458, "learning_rate": 9.973981770685474e-05, "loss": 0.051, "step": 6470 }, { "grad_norm": 0.42776358127593994, "learning_rate": 9.97377076946014e-05, "loss": 0.0454, "step": 6480 }, { "grad_norm": 0.5107532739639282, "learning_rate": 9.973558918353412e-05, "loss": 0.057, "step": 6490 }, { "grad_norm": 0.5184114575386047, "learning_rate": 9.973346217401494e-05, "loss": 0.0541, "step": 6500 }, { "grad_norm": 0.5172733068466187, "learning_rate": 9.973132666640726e-05, "loss": 0.0493, "step": 6510 }, { "grad_norm": 1.1689622402191162, "learning_rate": 9.972918266107602e-05, "loss": 0.0607, "step": 6520 }, { "grad_norm": 0.5960022211074829, "learning_rate": 9.972703015838756e-05, "loss": 0.0565, "step": 6530 }, { "grad_norm": 0.4154093563556671, "learning_rate": 9.97248691587097e-05, "loss": 0.0698, "step": 6540 }, { "grad_norm": 0.44820454716682434, "learning_rate": 9.972269966241166e-05, "loss": 0.0594, "step": 6550 }, { "grad_norm": 0.42310360074043274, "learning_rate": 9.972052166986417e-05, "loss": 0.0518, "step": 6560 }, { "grad_norm": 0.453490674495697, "learning_rate": 9.971833518143938e-05, "loss": 0.0713, "step": 6570 }, { "grad_norm": 0.5061421990394592, "learning_rate": 9.971614019751093e-05, "loss": 0.0581, "step": 6580 }, { "grad_norm": 0.6061375737190247, "learning_rate": 9.971393671845383e-05, "loss": 0.0688, "step": 6590 }, { "grad_norm": 0.2901016175746918, "learning_rate": 9.971172474464464e-05, "loss": 0.0634, "step": 6600 }, { "grad_norm": 0.5125653743743896, "learning_rate": 9.97095042764613e-05, "loss": 0.0521, "step": 6610 }, { "grad_norm": 0.37855395674705505, "learning_rate": 9.970727531428324e-05, "loss": 0.0474, "step": 6620 }, { "grad_norm": 0.4505496621131897, "learning_rate": 9.970503785849132e-05, "loss": 0.0501, "step": 6630 }, { "grad_norm": 0.4047142267227173, "learning_rate": 9.970279190946788e-05, "loss": 0.0476, "step": 6640 }, { "grad_norm": 0.46410080790519714, "learning_rate": 9.970053746759667e-05, "loss": 0.0654, "step": 6650 }, { "grad_norm": 0.5118327736854553, "learning_rate": 9.969827453326292e-05, "loss": 0.0449, "step": 6660 }, { "grad_norm": 0.3733236789703369, "learning_rate": 9.969600310685332e-05, "loss": 0.0491, "step": 6670 }, { "grad_norm": 0.6329568028450012, "learning_rate": 9.969372318875596e-05, "loss": 0.0606, "step": 6680 }, { "grad_norm": 0.4606161117553711, "learning_rate": 9.969143477936043e-05, "loss": 0.057, "step": 6690 }, { "grad_norm": 0.5360599756240845, "learning_rate": 9.968913787905775e-05, "loss": 0.0538, "step": 6700 }, { "grad_norm": 0.4834895133972168, "learning_rate": 9.968683248824045e-05, "loss": 0.0592, "step": 6710 }, { "grad_norm": 0.526343047618866, "learning_rate": 9.968451860730238e-05, "loss": 0.0613, "step": 6720 }, { "grad_norm": 0.4650905132293701, "learning_rate": 9.968219623663896e-05, "loss": 0.059, "step": 6730 }, { "grad_norm": 0.763775110244751, "learning_rate": 9.967986537664702e-05, "loss": 0.0791, "step": 6740 }, { "grad_norm": 0.480550080537796, "learning_rate": 9.967752602772483e-05, "loss": 0.0543, "step": 6750 }, { "grad_norm": 0.5963035821914673, "learning_rate": 9.967517819027212e-05, "loss": 0.0529, "step": 6760 }, { "grad_norm": 0.4227844178676605, "learning_rate": 9.967282186469009e-05, "loss": 0.0468, "step": 6770 }, { "grad_norm": 0.3667798936367035, "learning_rate": 9.967045705138135e-05, "loss": 0.054, "step": 6780 }, { "grad_norm": 0.49610865116119385, "learning_rate": 9.966808375074998e-05, "loss": 0.0524, "step": 6790 }, { "grad_norm": 0.4513457119464874, "learning_rate": 9.966570196320154e-05, "loss": 0.0633, "step": 6800 }, { "grad_norm": 0.45306289196014404, "learning_rate": 9.966331168914299e-05, "loss": 0.051, "step": 6810 }, { "grad_norm": 0.49693942070007324, "learning_rate": 9.966091292898277e-05, "loss": 0.0615, "step": 6820 }, { "grad_norm": 0.5551395416259766, "learning_rate": 9.965850568313076e-05, "loss": 0.0572, "step": 6830 }, { "grad_norm": 0.5164602994918823, "learning_rate": 9.965608995199827e-05, "loss": 0.0624, "step": 6840 }, { "grad_norm": 0.5769986510276794, "learning_rate": 9.965366573599812e-05, "loss": 0.0581, "step": 6850 }, { "grad_norm": 0.4350873529911041, "learning_rate": 9.965123303554453e-05, "loss": 0.0489, "step": 6860 }, { "grad_norm": 0.48755574226379395, "learning_rate": 9.964879185105317e-05, "loss": 0.0667, "step": 6870 }, { "grad_norm": 0.47472333908081055, "learning_rate": 9.964634218294119e-05, "loss": 0.0632, "step": 6880 }, { "grad_norm": 0.30464449524879456, "learning_rate": 9.964388403162714e-05, "loss": 0.0644, "step": 6890 }, { "grad_norm": 0.5188261270523071, "learning_rate": 9.96414173975311e-05, "loss": 0.063, "step": 6900 }, { "grad_norm": 0.43742379546165466, "learning_rate": 9.963894228107451e-05, "loss": 0.0518, "step": 6910 }, { "grad_norm": 0.3980954587459564, "learning_rate": 9.963645868268032e-05, "loss": 0.0664, "step": 6920 }, { "grad_norm": 0.5248470902442932, "learning_rate": 9.963396660277289e-05, "loss": 0.0615, "step": 6930 }, { "grad_norm": 0.4901803731918335, "learning_rate": 9.963146604177807e-05, "loss": 0.0546, "step": 6940 }, { "grad_norm": 0.4194282293319702, "learning_rate": 9.962895700012311e-05, "loss": 0.0518, "step": 6950 }, { "grad_norm": 0.6351339221000671, "learning_rate": 9.962643947823677e-05, "loss": 0.054, "step": 6960 }, { "grad_norm": 0.7785986661911011, "learning_rate": 9.962391347654921e-05, "loss": 0.057, "step": 6970 }, { "grad_norm": 1.0904313325881958, "learning_rate": 9.962137899549204e-05, "loss": 0.0682, "step": 6980 }, { "grad_norm": 0.41597819328308105, "learning_rate": 9.961883603549835e-05, "loss": 0.0492, "step": 6990 }, { "grad_norm": 0.4803171753883362, "learning_rate": 9.961628459700267e-05, "loss": 0.0619, "step": 7000 }, { "grad_norm": 0.44752416014671326, "learning_rate": 9.961372468044095e-05, "loss": 0.0801, "step": 7010 }, { "grad_norm": 0.46920692920684814, "learning_rate": 9.961115628625062e-05, "loss": 0.0605, "step": 7020 }, { "grad_norm": 0.47962141036987305, "learning_rate": 9.960857941487056e-05, "loss": 0.0762, "step": 7030 }, { "grad_norm": 0.4295434355735779, "learning_rate": 9.960599406674106e-05, "loss": 0.0601, "step": 7040 }, { "grad_norm": 0.4152180254459381, "learning_rate": 9.960340024230393e-05, "loss": 0.055, "step": 7050 }, { "grad_norm": 0.45870986580848694, "learning_rate": 9.960079794200232e-05, "loss": 0.0526, "step": 7060 }, { "grad_norm": 0.43755120038986206, "learning_rate": 9.959818716628096e-05, "loss": 0.0552, "step": 7070 }, { "grad_norm": 0.4135204255580902, "learning_rate": 9.95955679155859e-05, "loss": 0.0685, "step": 7080 }, { "grad_norm": 0.48615312576293945, "learning_rate": 9.959294019036472e-05, "loss": 0.0592, "step": 7090 }, { "grad_norm": 0.4029376208782196, "learning_rate": 9.959030399106646e-05, "loss": 0.0472, "step": 7100 }, { "grad_norm": 0.28906968235969543, "learning_rate": 9.958765931814153e-05, "loss": 0.0614, "step": 7110 }, { "grad_norm": 0.5595409274101257, "learning_rate": 9.958500617204184e-05, "loss": 0.0528, "step": 7120 }, { "grad_norm": 0.5683703422546387, "learning_rate": 9.958234455322075e-05, "loss": 0.0658, "step": 7130 }, { "grad_norm": 0.40671733021736145, "learning_rate": 9.957967446213308e-05, "loss": 0.0728, "step": 7140 }, { "grad_norm": 0.3788314759731293, "learning_rate": 9.957699589923501e-05, "loss": 0.0609, "step": 7150 }, { "grad_norm": 0.5236706733703613, "learning_rate": 9.957430886498431e-05, "loss": 0.0544, "step": 7160 }, { "grad_norm": 0.49853673577308655, "learning_rate": 9.957161335984008e-05, "loss": 0.0619, "step": 7170 }, { "grad_norm": 0.5220471024513245, "learning_rate": 9.956890938426291e-05, "loss": 0.0509, "step": 7180 }, { "grad_norm": 0.43738046288490295, "learning_rate": 9.956619693871482e-05, "loss": 0.0665, "step": 7190 }, { "grad_norm": 0.4298911690711975, "learning_rate": 9.956347602365934e-05, "loss": 0.0623, "step": 7200 }, { "grad_norm": 0.7180488705635071, "learning_rate": 9.956074663956135e-05, "loss": 0.0612, "step": 7210 }, { "grad_norm": 0.46178969740867615, "learning_rate": 9.955800878688726e-05, "loss": 0.059, "step": 7220 }, { "grad_norm": 0.5301060080528259, "learning_rate": 9.955526246610489e-05, "loss": 0.0616, "step": 7230 }, { "grad_norm": 0.5370058417320251, "learning_rate": 9.955250767768349e-05, "loss": 0.056, "step": 7240 }, { "grad_norm": 0.6483880877494812, "learning_rate": 9.95497444220938e-05, "loss": 0.0592, "step": 7250 }, { "grad_norm": 0.39193856716156006, "learning_rate": 9.954697269980797e-05, "loss": 0.0545, "step": 7260 }, { "grad_norm": 0.5097141265869141, "learning_rate": 9.954419251129962e-05, "loss": 0.0487, "step": 7270 }, { "grad_norm": 0.4403044581413269, "learning_rate": 9.95414038570438e-05, "loss": 0.0502, "step": 7280 }, { "grad_norm": 0.34792283177375793, "learning_rate": 9.953860673751703e-05, "loss": 0.0458, "step": 7290 }, { "grad_norm": 0.3990423381328583, "learning_rate": 9.953580115319725e-05, "loss": 0.0559, "step": 7300 }, { "grad_norm": 0.5418176651000977, "learning_rate": 9.953298710456387e-05, "loss": 0.0523, "step": 7310 }, { "grad_norm": 0.38780221343040466, "learning_rate": 9.953016459209771e-05, "loss": 0.0458, "step": 7320 }, { "grad_norm": 0.47377270460128784, "learning_rate": 9.952733361628108e-05, "loss": 0.0509, "step": 7330 }, { "grad_norm": 0.46651503443717957, "learning_rate": 9.952449417759772e-05, "loss": 0.056, "step": 7340 }, { "grad_norm": 0.3780425786972046, "learning_rate": 9.952164627653279e-05, "loss": 0.0466, "step": 7350 }, { "grad_norm": 0.6920761466026306, "learning_rate": 9.951878991357292e-05, "loss": 0.048, "step": 7360 }, { "grad_norm": 0.43319588899612427, "learning_rate": 9.951592508920622e-05, "loss": 0.0545, "step": 7370 }, { "grad_norm": 0.39163586497306824, "learning_rate": 9.951305180392219e-05, "loss": 0.0446, "step": 7380 }, { "grad_norm": 0.4361785054206848, "learning_rate": 9.951017005821178e-05, "loss": 0.0448, "step": 7390 }, { "grad_norm": 0.4890795648097992, "learning_rate": 9.95072798525674e-05, "loss": 0.0526, "step": 7400 }, { "grad_norm": 0.46251511573791504, "learning_rate": 9.950438118748293e-05, "loss": 0.0564, "step": 7410 }, { "grad_norm": 0.4302504360675812, "learning_rate": 9.950147406345366e-05, "loss": 0.0592, "step": 7420 }, { "grad_norm": 0.4901852607727051, "learning_rate": 9.949855848097635e-05, "loss": 0.0589, "step": 7430 }, { "grad_norm": 0.3970917761325836, "learning_rate": 9.949563444054916e-05, "loss": 0.0528, "step": 7440 }, { "grad_norm": 0.3932298421859741, "learning_rate": 9.949270194267178e-05, "loss": 0.0498, "step": 7450 }, { "grad_norm": 0.6126936674118042, "learning_rate": 9.948976098784526e-05, "loss": 0.0559, "step": 7460 }, { "grad_norm": 0.45734721422195435, "learning_rate": 9.948681157657213e-05, "loss": 0.0549, "step": 7470 }, { "grad_norm": 0.5389500260353088, "learning_rate": 9.948385370935638e-05, "loss": 0.0689, "step": 7480 }, { "grad_norm": 0.47119012475013733, "learning_rate": 9.94808873867034e-05, "loss": 0.057, "step": 7490 }, { "grad_norm": 0.41824886202812195, "learning_rate": 9.947791260912009e-05, "loss": 0.0493, "step": 7500 }, { "grad_norm": 0.3883272111415863, "learning_rate": 9.947492937711474e-05, "loss": 0.0469, "step": 7510 }, { "grad_norm": 0.36654627323150635, "learning_rate": 9.947193769119707e-05, "loss": 0.0464, "step": 7520 }, { "grad_norm": 0.4101833999156952, "learning_rate": 9.946893755187834e-05, "loss": 0.0446, "step": 7530 }, { "grad_norm": 0.38464251160621643, "learning_rate": 9.946592895967115e-05, "loss": 0.0575, "step": 7540 }, { "grad_norm": 0.3713603913784027, "learning_rate": 9.94629119150896e-05, "loss": 0.0496, "step": 7550 }, { "grad_norm": 0.4486082196235657, "learning_rate": 9.94598864186492e-05, "loss": 0.0532, "step": 7560 }, { "grad_norm": 0.39681583642959595, "learning_rate": 9.945685247086696e-05, "loss": 0.0568, "step": 7570 }, { "grad_norm": 0.5308398008346558, "learning_rate": 9.945381007226129e-05, "loss": 0.0537, "step": 7580 }, { "grad_norm": 0.32812726497650146, "learning_rate": 9.945075922335203e-05, "loss": 0.0514, "step": 7590 }, { "grad_norm": 0.9599292278289795, "learning_rate": 9.944769992466049e-05, "loss": 0.0678, "step": 7600 }, { "grad_norm": 0.40137532353401184, "learning_rate": 9.944463217670945e-05, "loss": 0.0561, "step": 7610 }, { "grad_norm": 0.3985302448272705, "learning_rate": 9.944155598002307e-05, "loss": 0.0557, "step": 7620 }, { "grad_norm": 0.5771428346633911, "learning_rate": 9.943847133512701e-05, "loss": 0.0626, "step": 7630 }, { "grad_norm": 0.551591157913208, "learning_rate": 9.943537824254834e-05, "loss": 0.0576, "step": 7640 }, { "grad_norm": 0.4876486659049988, "learning_rate": 9.943227670281559e-05, "loss": 0.058, "step": 7650 }, { "grad_norm": 0.778035044670105, "learning_rate": 9.942916671645873e-05, "loss": 0.0517, "step": 7660 }, { "grad_norm": 0.5517625212669373, "learning_rate": 9.942604828400916e-05, "loss": 0.0565, "step": 7670 }, { "grad_norm": 0.4577121138572693, "learning_rate": 9.942292140599975e-05, "loss": 0.0527, "step": 7680 }, { "grad_norm": 0.5684711933135986, "learning_rate": 9.94197860829648e-05, "loss": 0.0535, "step": 7690 }, { "grad_norm": 0.34396061301231384, "learning_rate": 9.941664231544004e-05, "loss": 0.0448, "step": 7700 }, { "grad_norm": 0.4350902736186981, "learning_rate": 9.941349010396264e-05, "loss": 0.066, "step": 7710 }, { "grad_norm": 0.3731021285057068, "learning_rate": 9.941032944907125e-05, "loss": 0.0535, "step": 7720 }, { "grad_norm": 0.35483720898628235, "learning_rate": 9.940716035130596e-05, "loss": 0.0556, "step": 7730 }, { "grad_norm": 0.469810426235199, "learning_rate": 9.940398281120821e-05, "loss": 0.0517, "step": 7740 }, { "grad_norm": 0.4952540099620819, "learning_rate": 9.940079682932102e-05, "loss": 0.0465, "step": 7750 }, { "grad_norm": 0.33593007922172546, "learning_rate": 9.939760240618877e-05, "loss": 0.0556, "step": 7760 }, { "grad_norm": 0.4810146689414978, "learning_rate": 9.939439954235729e-05, "loss": 0.0535, "step": 7770 }, { "grad_norm": 0.4122205376625061, "learning_rate": 9.939118823837387e-05, "loss": 0.0506, "step": 7780 }, { "grad_norm": 0.4752987325191498, "learning_rate": 9.938796849478725e-05, "loss": 0.0577, "step": 7790 }, { "grad_norm": 0.35641637444496155, "learning_rate": 9.938474031214755e-05, "loss": 0.0596, "step": 7800 }, { "grad_norm": 0.4955659806728363, "learning_rate": 9.938150369100643e-05, "loss": 0.0593, "step": 7810 }, { "grad_norm": 0.4338963031768799, "learning_rate": 9.93782586319169e-05, "loss": 0.0393, "step": 7820 }, { "grad_norm": 0.7847790718078613, "learning_rate": 9.937500513543348e-05, "loss": 0.0438, "step": 7830 }, { "grad_norm": 0.653581440448761, "learning_rate": 9.937174320211207e-05, "loss": 0.0484, "step": 7840 }, { "grad_norm": 0.4842633903026581, "learning_rate": 9.936847283251009e-05, "loss": 0.0568, "step": 7850 }, { "grad_norm": 0.4948282241821289, "learning_rate": 9.936519402718632e-05, "loss": 0.0578, "step": 7860 }, { "grad_norm": 0.33704689145088196, "learning_rate": 9.936190678670102e-05, "loss": 0.0543, "step": 7870 }, { "grad_norm": 0.5327231884002686, "learning_rate": 9.935861111161593e-05, "loss": 0.0487, "step": 7880 }, { "grad_norm": 0.31018370389938354, "learning_rate": 9.935530700249416e-05, "loss": 0.0452, "step": 7890 }, { "grad_norm": 0.33310267329216003, "learning_rate": 9.935199445990028e-05, "loss": 0.0591, "step": 7900 }, { "grad_norm": 0.46309277415275574, "learning_rate": 9.934867348440033e-05, "loss": 0.0555, "step": 7910 }, { "grad_norm": 0.4248230457305908, "learning_rate": 9.934534407656176e-05, "loss": 0.0731, "step": 7920 }, { "grad_norm": 0.369706928730011, "learning_rate": 9.93420062369535e-05, "loss": 0.0468, "step": 7930 }, { "grad_norm": 0.3708363175392151, "learning_rate": 9.933865996614589e-05, "loss": 0.0542, "step": 7940 }, { "grad_norm": 0.4882661700248718, "learning_rate": 9.933530526471068e-05, "loss": 0.0483, "step": 7950 }, { "grad_norm": 0.45900002121925354, "learning_rate": 9.933194213322114e-05, "loss": 0.0487, "step": 7960 }, { "grad_norm": 0.3007563352584839, "learning_rate": 9.932857057225192e-05, "loss": 0.0499, "step": 7970 }, { "grad_norm": 0.37241679430007935, "learning_rate": 9.932519058237912e-05, "loss": 0.0447, "step": 7980 }, { "grad_norm": 0.5192960500717163, "learning_rate": 9.932180216418032e-05, "loss": 0.0478, "step": 7990 }, { "grad_norm": 0.3230718672275543, "learning_rate": 9.931840531823446e-05, "loss": 0.0487, "step": 8000 }, { "grad_norm": 0.43019890785217285, "learning_rate": 9.9315000045122e-05, "loss": 0.0498, "step": 8010 }, { "grad_norm": 0.409746378660202, "learning_rate": 9.931158634542481e-05, "loss": 0.0502, "step": 8020 }, { "grad_norm": 0.64134681224823, "learning_rate": 9.930816421972617e-05, "loss": 0.048, "step": 8030 }, { "grad_norm": 0.3456425070762634, "learning_rate": 9.930473366861086e-05, "loss": 0.0486, "step": 8040 }, { "grad_norm": 0.4094483554363251, "learning_rate": 9.930129469266505e-05, "loss": 0.0503, "step": 8050 }, { "grad_norm": 0.3468760550022125, "learning_rate": 9.929784729247638e-05, "loss": 0.0463, "step": 8060 }, { "grad_norm": 0.3089451491832733, "learning_rate": 9.929439146863389e-05, "loss": 0.0446, "step": 8070 }, { "grad_norm": 0.43980079889297485, "learning_rate": 9.92909272217281e-05, "loss": 0.0566, "step": 8080 }, { "grad_norm": 0.475032776594162, "learning_rate": 9.928745455235097e-05, "loss": 0.0548, "step": 8090 }, { "grad_norm": 0.3859490752220154, "learning_rate": 9.928397346109588e-05, "loss": 0.0528, "step": 8100 }, { "grad_norm": 0.47194725275039673, "learning_rate": 9.928048394855762e-05, "loss": 0.0584, "step": 8110 }, { "grad_norm": 0.3603428602218628, "learning_rate": 9.92769860153325e-05, "loss": 0.0518, "step": 8120 }, { "grad_norm": 0.516917884349823, "learning_rate": 9.927347966201819e-05, "loss": 0.0656, "step": 8130 }, { "grad_norm": 0.4664272964000702, "learning_rate": 9.926996488921383e-05, "loss": 0.0581, "step": 8140 }, { "grad_norm": 0.3943425118923187, "learning_rate": 9.926644169752001e-05, "loss": 0.0549, "step": 8150 }, { "grad_norm": 0.5102285146713257, "learning_rate": 9.926291008753875e-05, "loss": 0.06, "step": 8160 }, { "grad_norm": 0.6716267466545105, "learning_rate": 9.92593700598735e-05, "loss": 0.0609, "step": 8170 }, { "grad_norm": 0.7549842596054077, "learning_rate": 9.925582161512915e-05, "loss": 0.0509, "step": 8180 }, { "grad_norm": 0.5180937647819519, "learning_rate": 9.925226475391205e-05, "loss": 0.0606, "step": 8190 }, { "grad_norm": 0.463859498500824, "learning_rate": 9.924869947682993e-05, "loss": 0.0576, "step": 8200 }, { "grad_norm": 0.5904533863067627, "learning_rate": 9.924512578449204e-05, "loss": 0.0592, "step": 8210 }, { "grad_norm": 0.32603612542152405, "learning_rate": 9.924154367750901e-05, "loss": 0.0564, "step": 8220 }, { "grad_norm": 0.5655185580253601, "learning_rate": 9.923795315649293e-05, "loss": 0.0572, "step": 8230 }, { "grad_norm": 1.142593502998352, "learning_rate": 9.92343542220573e-05, "loss": 0.0748, "step": 8240 }, { "grad_norm": 0.432991087436676, "learning_rate": 9.92307468748171e-05, "loss": 0.057, "step": 8250 }, { "grad_norm": 0.3619069755077362, "learning_rate": 9.922713111538873e-05, "loss": 0.0525, "step": 8260 }, { "grad_norm": 0.417672336101532, "learning_rate": 9.922350694439003e-05, "loss": 0.046, "step": 8270 }, { "grad_norm": 0.4748116433620453, "learning_rate": 9.921987436244024e-05, "loss": 0.0614, "step": 8280 }, { "grad_norm": 0.4268236458301544, "learning_rate": 9.921623337016008e-05, "loss": 0.0497, "step": 8290 }, { "grad_norm": 0.4961448907852173, "learning_rate": 9.921258396817172e-05, "loss": 0.052, "step": 8300 }, { "grad_norm": 0.43036940693855286, "learning_rate": 9.920892615709874e-05, "loss": 0.0435, "step": 8310 }, { "grad_norm": 0.4715198278427124, "learning_rate": 9.920525993756612e-05, "loss": 0.0526, "step": 8320 }, { "grad_norm": 0.6324524283409119, "learning_rate": 9.920158531020036e-05, "loss": 0.0529, "step": 8330 }, { "grad_norm": 0.4858635663986206, "learning_rate": 9.919790227562933e-05, "loss": 0.0473, "step": 8340 }, { "grad_norm": 0.500032901763916, "learning_rate": 9.919421083448237e-05, "loss": 0.054, "step": 8350 }, { "grad_norm": 0.5216325521469116, "learning_rate": 9.919051098739022e-05, "loss": 0.0519, "step": 8360 }, { "grad_norm": 0.2861941158771515, "learning_rate": 9.918680273498514e-05, "loss": 0.0535, "step": 8370 }, { "grad_norm": 0.3700065314769745, "learning_rate": 9.918308607790072e-05, "loss": 0.0537, "step": 8380 }, { "grad_norm": 0.3141019940376282, "learning_rate": 9.917936101677205e-05, "loss": 0.0461, "step": 8390 }, { "grad_norm": 0.4181690514087677, "learning_rate": 9.917562755223564e-05, "loss": 0.0392, "step": 8400 }, { "grad_norm": 0.29551389813423157, "learning_rate": 9.917188568492944e-05, "loss": 0.0525, "step": 8410 }, { "grad_norm": 0.6236687898635864, "learning_rate": 9.916813541549283e-05, "loss": 0.0487, "step": 8420 }, { "grad_norm": 0.454467236995697, "learning_rate": 9.916437674456663e-05, "loss": 0.0516, "step": 8430 }, { "grad_norm": 1.0988487005233765, "learning_rate": 9.916060967279308e-05, "loss": 0.0569, "step": 8440 }, { "grad_norm": 0.5044707655906677, "learning_rate": 9.91568342008159e-05, "loss": 0.0515, "step": 8450 }, { "grad_norm": 0.32781359553337097, "learning_rate": 9.915305032928019e-05, "loss": 0.0603, "step": 8460 }, { "grad_norm": 0.5122451782226562, "learning_rate": 9.914925805883253e-05, "loss": 0.0528, "step": 8470 }, { "grad_norm": 0.4339406490325928, "learning_rate": 9.914545739012088e-05, "loss": 0.0519, "step": 8480 }, { "grad_norm": 0.4981926679611206, "learning_rate": 9.91416483237947e-05, "loss": 0.055, "step": 8490 }, { "grad_norm": 0.4434828460216522, "learning_rate": 9.913783086050485e-05, "loss": 0.0554, "step": 8500 }, { "grad_norm": 2.267503261566162, "learning_rate": 9.913400500090364e-05, "loss": 0.0463, "step": 8510 }, { "grad_norm": 0.30473607778549194, "learning_rate": 9.913017074564479e-05, "loss": 0.0408, "step": 8520 }, { "grad_norm": 0.2973996102809906, "learning_rate": 9.912632809538348e-05, "loss": 0.0486, "step": 8530 }, { "grad_norm": 0.43758025765419006, "learning_rate": 9.912247705077629e-05, "loss": 0.0517, "step": 8540 }, { "grad_norm": 0.4086833894252777, "learning_rate": 9.911861761248127e-05, "loss": 0.0582, "step": 8550 }, { "grad_norm": 0.5561848282814026, "learning_rate": 9.91147497811579e-05, "loss": 0.0485, "step": 8560 }, { "grad_norm": 0.5525003671646118, "learning_rate": 9.911087355746709e-05, "loss": 0.0741, "step": 8570 }, { "grad_norm": 0.6147617697715759, "learning_rate": 9.910698894207117e-05, "loss": 0.0622, "step": 8580 }, { "grad_norm": 0.511444628238678, "learning_rate": 9.910309593563392e-05, "loss": 0.0579, "step": 8590 }, { "grad_norm": 0.393148273229599, "learning_rate": 9.909919453882057e-05, "loss": 0.0442, "step": 8600 }, { "grad_norm": 0.48511451482772827, "learning_rate": 9.90952847522977e-05, "loss": 0.0431, "step": 8610 }, { "grad_norm": 0.307079017162323, "learning_rate": 9.909136657673346e-05, "loss": 0.0494, "step": 8620 }, { "grad_norm": 0.30656060576438904, "learning_rate": 9.908744001279731e-05, "loss": 0.0476, "step": 8630 }, { "grad_norm": 0.45860520005226135, "learning_rate": 9.90835050611602e-05, "loss": 0.0519, "step": 8640 }, { "grad_norm": 0.4207751154899597, "learning_rate": 9.90795617224945e-05, "loss": 0.0573, "step": 8650 }, { "grad_norm": 0.4176954925060272, "learning_rate": 9.907560999747405e-05, "loss": 0.0473, "step": 8660 }, { "grad_norm": 0.8829623460769653, "learning_rate": 9.907164988677408e-05, "loss": 0.0498, "step": 8670 }, { "grad_norm": 0.3980009853839874, "learning_rate": 9.906768139107124e-05, "loss": 0.0493, "step": 8680 }, { "grad_norm": 0.5698482394218445, "learning_rate": 9.906370451104367e-05, "loss": 0.0539, "step": 8690 }, { "grad_norm": 0.6632389426231384, "learning_rate": 9.905971924737088e-05, "loss": 0.0531, "step": 8700 }, { "grad_norm": 0.6219661235809326, "learning_rate": 9.905572560073387e-05, "loss": 0.0565, "step": 8710 }, { "grad_norm": 0.5497992634773254, "learning_rate": 9.905172357181501e-05, "loss": 0.0659, "step": 8720 }, { "grad_norm": 1.3255695104599, "learning_rate": 9.904771316129817e-05, "loss": 0.0571, "step": 8730 }, { "grad_norm": 0.6716846823692322, "learning_rate": 9.904369436986862e-05, "loss": 0.0484, "step": 8740 }, { "grad_norm": 0.44652798771858215, "learning_rate": 9.903966719821303e-05, "loss": 0.0539, "step": 8750 }, { "grad_norm": 0.277895987033844, "learning_rate": 9.903563164701956e-05, "loss": 0.0499, "step": 8760 }, { "grad_norm": 0.906212329864502, "learning_rate": 9.903158771697778e-05, "loss": 0.0534, "step": 8770 }, { "grad_norm": 0.3330267667770386, "learning_rate": 9.902753540877867e-05, "loss": 0.057, "step": 8780 }, { "grad_norm": 0.3208003640174866, "learning_rate": 9.902347472311466e-05, "loss": 0.0436, "step": 8790 }, { "grad_norm": 0.29063570499420166, "learning_rate": 9.901940566067962e-05, "loss": 0.0418, "step": 8800 }, { "grad_norm": 0.3595385253429413, "learning_rate": 9.901532822216883e-05, "loss": 0.0447, "step": 8810 }, { "grad_norm": 0.5483381748199463, "learning_rate": 9.901124240827904e-05, "loss": 0.0409, "step": 8820 }, { "grad_norm": 0.6332729458808899, "learning_rate": 9.900714821970835e-05, "loss": 0.0424, "step": 8830 }, { "grad_norm": 0.463628351688385, "learning_rate": 9.900304565715641e-05, "loss": 0.0471, "step": 8840 }, { "grad_norm": 0.3398115336894989, "learning_rate": 9.899893472132419e-05, "loss": 0.0407, "step": 8850 }, { "grad_norm": 0.2943992614746094, "learning_rate": 9.899481541291415e-05, "loss": 0.0442, "step": 8860 }, { "grad_norm": 0.4661751985549927, "learning_rate": 9.899068773263016e-05, "loss": 0.0591, "step": 8870 }, { "grad_norm": 1.1132683753967285, "learning_rate": 9.898655168117754e-05, "loss": 0.0538, "step": 8880 }, { "grad_norm": 0.4058960974216461, "learning_rate": 9.898240725926302e-05, "loss": 0.0613, "step": 8890 }, { "grad_norm": 0.4126334488391876, "learning_rate": 9.897825446759478e-05, "loss": 0.0518, "step": 8900 }, { "grad_norm": 0.32559171319007874, "learning_rate": 9.897409330688241e-05, "loss": 0.0492, "step": 8910 }, { "grad_norm": 0.46423935890197754, "learning_rate": 9.896992377783692e-05, "loss": 0.0566, "step": 8920 }, { "grad_norm": 0.366713285446167, "learning_rate": 9.89657458811708e-05, "loss": 0.0459, "step": 8930 }, { "grad_norm": 0.33819395303726196, "learning_rate": 9.896155961759792e-05, "loss": 0.0474, "step": 8940 }, { "grad_norm": 0.4263189435005188, "learning_rate": 9.895736498783361e-05, "loss": 0.0471, "step": 8950 }, { "grad_norm": 0.2860603928565979, "learning_rate": 9.895316199259462e-05, "loss": 0.0416, "step": 8960 }, { "grad_norm": 0.3901812732219696, "learning_rate": 9.894895063259909e-05, "loss": 0.0421, "step": 8970 }, { "grad_norm": 0.5167029500007629, "learning_rate": 9.894473090856667e-05, "loss": 0.0437, "step": 8980 }, { "grad_norm": 0.48927199840545654, "learning_rate": 9.894050282121839e-05, "loss": 0.0703, "step": 8990 }, { "grad_norm": 0.6124585270881653, "learning_rate": 9.893626637127668e-05, "loss": 0.0492, "step": 9000 }, { "grad_norm": 0.4888843297958374, "learning_rate": 9.893202155946546e-05, "loss": 0.0513, "step": 9010 }, { "grad_norm": 0.4705168902873993, "learning_rate": 9.892776838651006e-05, "loss": 0.0571, "step": 9020 }, { "grad_norm": 0.4295760989189148, "learning_rate": 9.892350685313722e-05, "loss": 0.0621, "step": 9030 }, { "grad_norm": 0.4472285509109497, "learning_rate": 9.891923696007513e-05, "loss": 0.0526, "step": 9040 }, { "grad_norm": 0.38073208928108215, "learning_rate": 9.891495870805336e-05, "loss": 0.0422, "step": 9050 }, { "grad_norm": 0.5334897041320801, "learning_rate": 9.891067209780298e-05, "loss": 0.0564, "step": 9060 }, { "grad_norm": 0.4217931628227234, "learning_rate": 9.890637713005646e-05, "loss": 0.0511, "step": 9070 }, { "grad_norm": 0.5546966791152954, "learning_rate": 9.890207380554767e-05, "loss": 0.0481, "step": 9080 }, { "grad_norm": 0.41411978006362915, "learning_rate": 9.889776212501196e-05, "loss": 0.0636, "step": 9090 }, { "grad_norm": 0.3468710780143738, "learning_rate": 9.889344208918605e-05, "loss": 0.0537, "step": 9100 }, { "grad_norm": 0.5177452564239502, "learning_rate": 9.888911369880812e-05, "loss": 0.053, "step": 9110 }, { "grad_norm": 0.5012469291687012, "learning_rate": 9.888477695461777e-05, "loss": 0.0586, "step": 9120 }, { "grad_norm": 0.4553753733634949, "learning_rate": 9.888043185735607e-05, "loss": 0.0672, "step": 9130 }, { "grad_norm": 0.41134169697761536, "learning_rate": 9.887607840776542e-05, "loss": 0.0546, "step": 9140 }, { "grad_norm": 0.5725051164627075, "learning_rate": 9.887171660658975e-05, "loss": 0.0626, "step": 9150 }, { "grad_norm": 0.4136071801185608, "learning_rate": 9.886734645457435e-05, "loss": 0.0406, "step": 9160 }, { "grad_norm": 0.3994075059890747, "learning_rate": 9.886296795246597e-05, "loss": 0.0518, "step": 9170 }, { "grad_norm": 0.4506870210170746, "learning_rate": 9.885858110101276e-05, "loss": 0.0518, "step": 9180 }, { "grad_norm": 0.317841500043869, "learning_rate": 9.885418590096434e-05, "loss": 0.0387, "step": 9190 }, { "grad_norm": 0.394681453704834, "learning_rate": 9.88497823530717e-05, "loss": 0.0534, "step": 9200 }, { "grad_norm": 0.49785521626472473, "learning_rate": 9.884537045808732e-05, "loss": 0.0582, "step": 9210 }, { "grad_norm": 0.34930872917175293, "learning_rate": 9.884095021676502e-05, "loss": 0.0566, "step": 9220 }, { "grad_norm": 0.4494592845439911, "learning_rate": 9.883652162986017e-05, "loss": 0.0476, "step": 9230 }, { "grad_norm": 0.43744611740112305, "learning_rate": 9.883208469812943e-05, "loss": 0.0617, "step": 9240 }, { "grad_norm": 0.6158370971679688, "learning_rate": 9.882763942233098e-05, "loss": 0.0446, "step": 9250 }, { "grad_norm": 0.3778960704803467, "learning_rate": 9.882318580322441e-05, "loss": 0.0471, "step": 9260 }, { "grad_norm": 0.4284372925758362, "learning_rate": 9.881872384157067e-05, "loss": 0.07, "step": 9270 }, { "grad_norm": 0.5336896777153015, "learning_rate": 9.881425353813225e-05, "loss": 0.0611, "step": 9280 }, { "grad_norm": 0.5277944207191467, "learning_rate": 9.880977489367296e-05, "loss": 0.0683, "step": 9290 }, { "grad_norm": 0.40172433853149414, "learning_rate": 9.88052879089581e-05, "loss": 0.0598, "step": 9300 }, { "grad_norm": 0.3627793788909912, "learning_rate": 9.880079258475434e-05, "loss": 0.0503, "step": 9310 }, { "grad_norm": 0.8996289968490601, "learning_rate": 9.879628892182985e-05, "loss": 0.0643, "step": 9320 }, { "grad_norm": 0.4588569402694702, "learning_rate": 9.879177692095416e-05, "loss": 0.0529, "step": 9330 }, { "grad_norm": 0.33221927285194397, "learning_rate": 9.878725658289825e-05, "loss": 0.0461, "step": 9340 }, { "grad_norm": 0.39990103244781494, "learning_rate": 9.878272790843454e-05, "loss": 0.0432, "step": 9350 }, { "grad_norm": 0.3980273902416229, "learning_rate": 9.877819089833682e-05, "loss": 0.0418, "step": 9360 }, { "grad_norm": 0.779035210609436, "learning_rate": 9.877364555338038e-05, "loss": 0.0601, "step": 9370 }, { "grad_norm": 0.6651774644851685, "learning_rate": 9.876909187434186e-05, "loss": 0.0545, "step": 9380 }, { "grad_norm": 0.40578874945640564, "learning_rate": 9.876452986199939e-05, "loss": 0.0518, "step": 9390 }, { "grad_norm": 0.5025899410247803, "learning_rate": 9.875995951713248e-05, "loss": 0.056, "step": 9400 }, { "grad_norm": 0.4953238368034363, "learning_rate": 9.875538084052207e-05, "loss": 0.0581, "step": 9410 }, { "grad_norm": 0.5427589416503906, "learning_rate": 9.875079383295053e-05, "loss": 0.0566, "step": 9420 }, { "grad_norm": 0.2884726822376251, "learning_rate": 9.874619849520167e-05, "loss": 0.0372, "step": 9430 }, { "grad_norm": 0.43886733055114746, "learning_rate": 9.874159482806069e-05, "loss": 0.0483, "step": 9440 }, { "grad_norm": 0.39660510420799255, "learning_rate": 9.873698283231426e-05, "loss": 0.0403, "step": 9450 }, { "grad_norm": 0.44370073080062866, "learning_rate": 9.87323625087504e-05, "loss": 0.0363, "step": 9460 }, { "grad_norm": 1.0954163074493408, "learning_rate": 9.872773385815863e-05, "loss": 0.0629, "step": 9470 }, { "grad_norm": 0.3021570146083832, "learning_rate": 9.872309688132986e-05, "loss": 0.0407, "step": 9480 }, { "grad_norm": 0.2918310761451721, "learning_rate": 9.871845157905639e-05, "loss": 0.0462, "step": 9490 }, { "grad_norm": 0.3298071324825287, "learning_rate": 9.871379795213201e-05, "loss": 0.0388, "step": 9500 }, { "grad_norm": 0.5100157856941223, "learning_rate": 9.87091360013519e-05, "loss": 0.0426, "step": 9510 }, { "grad_norm": 0.4498734176158905, "learning_rate": 9.870446572751262e-05, "loss": 0.0461, "step": 9520 }, { "grad_norm": 0.42497339844703674, "learning_rate": 9.869978713141224e-05, "loss": 0.0431, "step": 9530 }, { "grad_norm": 0.43199849128723145, "learning_rate": 9.869510021385016e-05, "loss": 0.0483, "step": 9540 }, { "grad_norm": 0.4274902641773224, "learning_rate": 9.869040497562727e-05, "loss": 0.0472, "step": 9550 }, { "grad_norm": 0.49392226338386536, "learning_rate": 9.868570141754587e-05, "loss": 0.0543, "step": 9560 }, { "grad_norm": 1.1351197957992554, "learning_rate": 9.868098954040965e-05, "loss": 0.0728, "step": 9570 }, { "grad_norm": 0.4698277711868286, "learning_rate": 9.867626934502374e-05, "loss": 0.0538, "step": 9580 }, { "grad_norm": 0.3695233166217804, "learning_rate": 9.86715408321947e-05, "loss": 0.0545, "step": 9590 }, { "grad_norm": 0.4238811731338501, "learning_rate": 9.86668040027305e-05, "loss": 0.0445, "step": 9600 }, { "grad_norm": 0.4595392942428589, "learning_rate": 9.866205885744053e-05, "loss": 0.0465, "step": 9610 }, { "grad_norm": 0.4945903718471527, "learning_rate": 9.865730539713563e-05, "loss": 0.0502, "step": 9620 }, { "grad_norm": 0.4567844569683075, "learning_rate": 9.8652543622628e-05, "loss": 0.0457, "step": 9630 }, { "grad_norm": 0.440216600894928, "learning_rate": 9.864777353473132e-05, "loss": 0.0462, "step": 9640 }, { "grad_norm": 0.6258738040924072, "learning_rate": 9.864299513426068e-05, "loss": 0.0478, "step": 9650 }, { "grad_norm": 0.4571375548839569, "learning_rate": 9.863820842203254e-05, "loss": 0.0528, "step": 9660 }, { "grad_norm": 0.33492136001586914, "learning_rate": 9.863341339886483e-05, "loss": 0.0494, "step": 9670 }, { "grad_norm": 0.4404773414134979, "learning_rate": 9.86286100655769e-05, "loss": 0.0445, "step": 9680 }, { "grad_norm": 0.3160436451435089, "learning_rate": 9.862379842298953e-05, "loss": 0.0468, "step": 9690 }, { "grad_norm": 0.3735196888446808, "learning_rate": 9.861897847192485e-05, "loss": 0.0371, "step": 9700 }, { "grad_norm": 0.40666693449020386, "learning_rate": 9.86141502132065e-05, "loss": 0.0397, "step": 9710 }, { "grad_norm": 0.3942891061306, "learning_rate": 9.860931364765946e-05, "loss": 0.0511, "step": 9720 }, { "grad_norm": 0.5062488317489624, "learning_rate": 9.860446877611021e-05, "loss": 0.048, "step": 9730 }, { "grad_norm": 0.3572193682193756, "learning_rate": 9.859961559938655e-05, "loss": 0.0421, "step": 9740 }, { "grad_norm": 0.37302207946777344, "learning_rate": 9.85947541183178e-05, "loss": 0.0445, "step": 9750 }, { "grad_norm": 0.6562185287475586, "learning_rate": 9.858988433373463e-05, "loss": 0.0443, "step": 9760 }, { "grad_norm": 0.434842586517334, "learning_rate": 9.858500624646918e-05, "loss": 0.0578, "step": 9770 }, { "grad_norm": 0.35672712326049805, "learning_rate": 9.858011985735497e-05, "loss": 0.0484, "step": 9780 }, { "grad_norm": 0.3432375192642212, "learning_rate": 9.857522516722693e-05, "loss": 0.0423, "step": 9790 }, { "grad_norm": 0.39944425225257874, "learning_rate": 9.857032217692145e-05, "loss": 0.0437, "step": 9800 }, { "grad_norm": 0.4781692922115326, "learning_rate": 9.856541088727631e-05, "loss": 0.0474, "step": 9810 }, { "grad_norm": 0.5130788683891296, "learning_rate": 9.856049129913072e-05, "loss": 0.0493, "step": 9820 }, { "grad_norm": 0.28267887234687805, "learning_rate": 9.85555634133253e-05, "loss": 0.0417, "step": 9830 }, { "grad_norm": 0.5889216661453247, "learning_rate": 9.855062723070208e-05, "loss": 0.0568, "step": 9840 }, { "grad_norm": 0.3748871982097626, "learning_rate": 9.854568275210454e-05, "loss": 0.046, "step": 9850 }, { "grad_norm": 0.4047444760799408, "learning_rate": 9.854072997837754e-05, "loss": 0.0387, "step": 9860 }, { "grad_norm": 0.31107619404792786, "learning_rate": 9.853576891036737e-05, "loss": 0.0488, "step": 9870 }, { "grad_norm": 0.38590207695961, "learning_rate": 9.853079954892177e-05, "loss": 0.0426, "step": 9880 }, { "grad_norm": 0.3091963529586792, "learning_rate": 9.852582189488983e-05, "loss": 0.035, "step": 9890 }, { "grad_norm": 0.30563777685165405, "learning_rate": 9.852083594912212e-05, "loss": 0.0443, "step": 9900 }, { "grad_norm": 0.35854050517082214, "learning_rate": 9.851584171247058e-05, "loss": 0.0581, "step": 9910 }, { "grad_norm": 0.4479358196258545, "learning_rate": 9.851083918578863e-05, "loss": 0.0476, "step": 9920 }, { "grad_norm": 0.391989529132843, "learning_rate": 9.850582836993103e-05, "loss": 0.0411, "step": 9930 }, { "grad_norm": 1.2966432571411133, "learning_rate": 9.850080926575397e-05, "loss": 0.0442, "step": 9940 }, { "grad_norm": 0.29777318239212036, "learning_rate": 9.849578187411515e-05, "loss": 0.042, "step": 9950 }, { "grad_norm": 0.3744329810142517, "learning_rate": 9.849074619587354e-05, "loss": 0.0451, "step": 9960 }, { "grad_norm": 0.5300676226615906, "learning_rate": 9.848570223188964e-05, "loss": 0.0505, "step": 9970 }, { "grad_norm": 0.284749835729599, "learning_rate": 9.848064998302531e-05, "loss": 0.0466, "step": 9980 }, { "grad_norm": 0.39702123403549194, "learning_rate": 9.847558945014386e-05, "loss": 0.0508, "step": 9990 }, { "grad_norm": 0.5040485262870789, "learning_rate": 9.847052063410996e-05, "loss": 0.0551, "step": 10000 }, { "grad_norm": 0.4209670126438141, "learning_rate": 9.846544353578977e-05, "loss": 0.0537, "step": 10010 }, { "grad_norm": 0.46868595480918884, "learning_rate": 9.846035815605081e-05, "loss": 0.0473, "step": 10020 }, { "grad_norm": 0.3811141848564148, "learning_rate": 9.845526449576204e-05, "loss": 0.0423, "step": 10030 }, { "grad_norm": 0.39388981461524963, "learning_rate": 9.845016255579383e-05, "loss": 0.0476, "step": 10040 }, { "grad_norm": 0.39767926931381226, "learning_rate": 9.844505233701794e-05, "loss": 0.0483, "step": 10050 }, { "grad_norm": 0.44404712319374084, "learning_rate": 9.843993384030757e-05, "loss": 0.0548, "step": 10060 }, { "grad_norm": 0.3211003243923187, "learning_rate": 9.843480706653737e-05, "loss": 0.0403, "step": 10070 }, { "grad_norm": 0.24906675517559052, "learning_rate": 9.84296720165833e-05, "loss": 0.0474, "step": 10080 }, { "grad_norm": 0.409164696931839, "learning_rate": 9.842452869132286e-05, "loss": 0.0486, "step": 10090 }, { "grad_norm": 0.2584502398967743, "learning_rate": 9.841937709163489e-05, "loss": 0.042, "step": 10100 }, { "grad_norm": 0.509044885635376, "learning_rate": 9.841421721839962e-05, "loss": 0.0487, "step": 10110 }, { "grad_norm": 2.0360541343688965, "learning_rate": 9.840904907249879e-05, "loss": 0.045, "step": 10120 }, { "grad_norm": 0.5100451111793518, "learning_rate": 9.840387265481545e-05, "loss": 0.0512, "step": 10130 }, { "grad_norm": 0.30542704463005066, "learning_rate": 9.839868796623411e-05, "loss": 0.0437, "step": 10140 }, { "grad_norm": 0.5724145770072937, "learning_rate": 9.839349500764072e-05, "loss": 0.052, "step": 10150 }, { "grad_norm": 0.4697045683860779, "learning_rate": 9.83882937799226e-05, "loss": 0.0527, "step": 10160 }, { "grad_norm": 0.9861262440681458, "learning_rate": 9.838308428396849e-05, "loss": 0.0681, "step": 10170 }, { "grad_norm": 0.3715120851993561, "learning_rate": 9.837786652066854e-05, "loss": 0.0607, "step": 10180 }, { "grad_norm": 0.4417566657066345, "learning_rate": 9.837264049091437e-05, "loss": 0.0637, "step": 10190 }, { "grad_norm": 0.3630509376525879, "learning_rate": 9.836740619559893e-05, "loss": 0.0507, "step": 10200 }, { "grad_norm": 0.453513503074646, "learning_rate": 9.836216363561659e-05, "loss": 0.0456, "step": 10210 }, { "grad_norm": 0.37706688046455383, "learning_rate": 9.835691281186322e-05, "loss": 0.045, "step": 10220 }, { "grad_norm": 0.33923089504241943, "learning_rate": 9.8351653725236e-05, "loss": 0.049, "step": 10230 }, { "grad_norm": 0.6387913227081299, "learning_rate": 9.83463863766336e-05, "loss": 0.0471, "step": 10240 }, { "grad_norm": 0.2551979720592499, "learning_rate": 9.834111076695602e-05, "loss": 0.0471, "step": 10250 }, { "grad_norm": 0.380631685256958, "learning_rate": 9.833582689710477e-05, "loss": 0.0404, "step": 10260 }, { "grad_norm": 0.4575612545013428, "learning_rate": 9.833053476798268e-05, "loss": 0.0523, "step": 10270 }, { "grad_norm": 0.6407182216644287, "learning_rate": 9.832523438049404e-05, "loss": 0.0554, "step": 10280 }, { "grad_norm": 0.3874457776546478, "learning_rate": 9.831992573554454e-05, "loss": 0.0537, "step": 10290 }, { "grad_norm": 0.4605327844619751, "learning_rate": 9.831460883404128e-05, "loss": 0.0474, "step": 10300 }, { "grad_norm": 0.48765525221824646, "learning_rate": 9.830928367689278e-05, "loss": 0.039, "step": 10310 }, { "grad_norm": 0.3711389899253845, "learning_rate": 9.830395026500896e-05, "loss": 0.0454, "step": 10320 }, { "grad_norm": 0.4224874973297119, "learning_rate": 9.829860859930115e-05, "loss": 0.0436, "step": 10330 }, { "grad_norm": 0.3358304798603058, "learning_rate": 9.829325868068212e-05, "loss": 0.0446, "step": 10340 }, { "grad_norm": 0.3247297704219818, "learning_rate": 9.8287900510066e-05, "loss": 0.0499, "step": 10350 }, { "grad_norm": 0.3747510313987732, "learning_rate": 9.828253408836834e-05, "loss": 0.0404, "step": 10360 }, { "grad_norm": 0.5102689862251282, "learning_rate": 9.827715941650615e-05, "loss": 0.0436, "step": 10370 }, { "grad_norm": 0.42906278371810913, "learning_rate": 9.82717764953978e-05, "loss": 0.0438, "step": 10380 }, { "grad_norm": 0.4232748746871948, "learning_rate": 9.826638532596308e-05, "loss": 0.0409, "step": 10390 }, { "grad_norm": 0.40482914447784424, "learning_rate": 9.82609859091232e-05, "loss": 0.0448, "step": 10400 }, { "grad_norm": 0.3087441325187683, "learning_rate": 9.825557824580076e-05, "loss": 0.0371, "step": 10410 }, { "grad_norm": 0.40690895915031433, "learning_rate": 9.82501623369198e-05, "loss": 0.0418, "step": 10420 }, { "grad_norm": 0.4069916009902954, "learning_rate": 9.824473818340574e-05, "loss": 0.0531, "step": 10430 }, { "grad_norm": 0.42459696531295776, "learning_rate": 9.823930578618541e-05, "loss": 0.0434, "step": 10440 }, { "grad_norm": 0.525507390499115, "learning_rate": 9.823386514618709e-05, "loss": 0.0499, "step": 10450 }, { "grad_norm": 0.3363460302352905, "learning_rate": 9.82284162643404e-05, "loss": 0.0425, "step": 10460 }, { "grad_norm": 0.9264169931411743, "learning_rate": 9.822295914157642e-05, "loss": 0.0384, "step": 10470 }, { "grad_norm": 0.5347291827201843, "learning_rate": 9.821749377882763e-05, "loss": 0.0411, "step": 10480 }, { "grad_norm": 0.3713454008102417, "learning_rate": 9.821202017702791e-05, "loss": 0.0407, "step": 10490 }, { "grad_norm": 0.29294511675834656, "learning_rate": 9.820653833711253e-05, "loss": 0.0339, "step": 10500 }, { "grad_norm": 0.4179764986038208, "learning_rate": 9.820104826001822e-05, "loss": 0.049, "step": 10510 }, { "grad_norm": 0.35758042335510254, "learning_rate": 9.819554994668305e-05, "loss": 0.0373, "step": 10520 }, { "grad_norm": 0.40815019607543945, "learning_rate": 9.819004339804654e-05, "loss": 0.0406, "step": 10530 }, { "grad_norm": 0.40374329686164856, "learning_rate": 9.818452861504961e-05, "loss": 0.0416, "step": 10540 }, { "grad_norm": 0.42424115538597107, "learning_rate": 9.81790055986346e-05, "loss": 0.0364, "step": 10550 }, { "grad_norm": 0.29204151034355164, "learning_rate": 9.817347434974523e-05, "loss": 0.0381, "step": 10560 }, { "grad_norm": 0.496980220079422, "learning_rate": 9.816793486932664e-05, "loss": 0.0435, "step": 10570 }, { "grad_norm": 0.3223333954811096, "learning_rate": 9.816238715832538e-05, "loss": 0.0438, "step": 10580 }, { "grad_norm": 0.3776867985725403, "learning_rate": 9.815683121768939e-05, "loss": 0.0454, "step": 10590 }, { "grad_norm": 0.36483901739120483, "learning_rate": 9.815126704836804e-05, "loss": 0.0559, "step": 10600 }, { "grad_norm": 0.6464248895645142, "learning_rate": 9.81456946513121e-05, "loss": 0.0628, "step": 10610 }, { "grad_norm": 0.4823336601257324, "learning_rate": 9.814011402747373e-05, "loss": 0.0551, "step": 10620 }, { "grad_norm": 0.3177531957626343, "learning_rate": 9.813452517780651e-05, "loss": 0.0395, "step": 10630 }, { "grad_norm": 0.361348032951355, "learning_rate": 9.81289281032654e-05, "loss": 0.051, "step": 10640 }, { "grad_norm": 0.33712321519851685, "learning_rate": 9.812332280480683e-05, "loss": 0.0469, "step": 10650 }, { "grad_norm": 0.3233439326286316, "learning_rate": 9.811770928338854e-05, "loss": 0.0528, "step": 10660 }, { "grad_norm": 0.4209281802177429, "learning_rate": 9.811208753996979e-05, "loss": 0.0489, "step": 10670 }, { "grad_norm": 0.36625486612319946, "learning_rate": 9.810645757551113e-05, "loss": 0.0368, "step": 10680 }, { "grad_norm": 0.4496822655200958, "learning_rate": 9.810081939097459e-05, "loss": 0.0509, "step": 10690 }, { "grad_norm": 0.4693571627140045, "learning_rate": 9.809517298732356e-05, "loss": 0.0644, "step": 10700 }, { "grad_norm": 0.580208420753479, "learning_rate": 9.80895183655229e-05, "loss": 0.0502, "step": 10710 }, { "grad_norm": 0.4010489583015442, "learning_rate": 9.808385552653877e-05, "loss": 0.0414, "step": 10720 }, { "grad_norm": 0.5440316796302795, "learning_rate": 9.807818447133886e-05, "loss": 0.0379, "step": 10730 }, { "grad_norm": 0.3627488613128662, "learning_rate": 9.807250520089215e-05, "loss": 0.0457, "step": 10740 }, { "grad_norm": 0.4327663481235504, "learning_rate": 9.806681771616908e-05, "loss": 0.0536, "step": 10750 }, { "grad_norm": 0.44277700781822205, "learning_rate": 9.80611220181415e-05, "loss": 0.0571, "step": 10760 }, { "grad_norm": 0.4891164004802704, "learning_rate": 9.805541810778264e-05, "loss": 0.0548, "step": 10770 }, { "grad_norm": 0.3560030460357666, "learning_rate": 9.804970598606716e-05, "loss": 0.0478, "step": 10780 }, { "grad_norm": 0.5116910338401794, "learning_rate": 9.804398565397106e-05, "loss": 0.0432, "step": 10790 }, { "grad_norm": 0.7980298399925232, "learning_rate": 9.803825711247183e-05, "loss": 0.0536, "step": 10800 }, { "grad_norm": 0.4563007950782776, "learning_rate": 9.803252036254831e-05, "loss": 0.0403, "step": 10810 }, { "grad_norm": 0.37131550908088684, "learning_rate": 9.802677540518076e-05, "loss": 0.0511, "step": 10820 }, { "grad_norm": 0.24989289045333862, "learning_rate": 9.802102224135081e-05, "loss": 0.041, "step": 10830 }, { "grad_norm": 0.3450574576854706, "learning_rate": 9.801526087204155e-05, "loss": 0.0464, "step": 10840 }, { "grad_norm": 0.44704464077949524, "learning_rate": 9.800949129823743e-05, "loss": 0.0453, "step": 10850 }, { "grad_norm": 0.36731570959091187, "learning_rate": 9.80037135209243e-05, "loss": 0.0379, "step": 10860 }, { "grad_norm": 0.33476778864860535, "learning_rate": 9.799792754108946e-05, "loss": 0.0432, "step": 10870 }, { "grad_norm": 0.3301321566104889, "learning_rate": 9.799213335972152e-05, "loss": 0.0418, "step": 10880 }, { "grad_norm": 0.3813799023628235, "learning_rate": 9.798633097781058e-05, "loss": 0.0429, "step": 10890 }, { "grad_norm": 0.3480040431022644, "learning_rate": 9.79805203963481e-05, "loss": 0.0451, "step": 10900 }, { "grad_norm": 0.39128929376602173, "learning_rate": 9.797470161632697e-05, "loss": 0.0403, "step": 10910 }, { "grad_norm": 0.32867148518562317, "learning_rate": 9.796887463874145e-05, "loss": 0.0432, "step": 10920 }, { "grad_norm": 0.5208410024642944, "learning_rate": 9.796303946458718e-05, "loss": 0.0511, "step": 10930 }, { "grad_norm": 0.48192811012268066, "learning_rate": 9.795719609486127e-05, "loss": 0.0469, "step": 10940 }, { "grad_norm": 0.3366769254207611, "learning_rate": 9.795134453056219e-05, "loss": 0.0423, "step": 10950 }, { "grad_norm": 0.40628987550735474, "learning_rate": 9.794548477268979e-05, "loss": 0.0357, "step": 10960 }, { "grad_norm": 0.23404370248317719, "learning_rate": 9.793961682224537e-05, "loss": 0.0349, "step": 10970 }, { "grad_norm": 0.29443392157554626, "learning_rate": 9.793374068023156e-05, "loss": 0.0379, "step": 10980 }, { "grad_norm": 0.37123268842697144, "learning_rate": 9.792785634765247e-05, "loss": 0.0416, "step": 10990 }, { "grad_norm": 0.32498258352279663, "learning_rate": 9.792196382551357e-05, "loss": 0.0398, "step": 11000 }, { "grad_norm": 0.649732768535614, "learning_rate": 9.791606311482171e-05, "loss": 0.0418, "step": 11010 }, { "grad_norm": 0.7701237797737122, "learning_rate": 9.791015421658518e-05, "loss": 0.0397, "step": 11020 }, { "grad_norm": 0.5795014500617981, "learning_rate": 9.790423713181362e-05, "loss": 0.0463, "step": 11030 }, { "grad_norm": 0.38673698902130127, "learning_rate": 9.789831186151814e-05, "loss": 0.0463, "step": 11040 }, { "grad_norm": 0.4319928288459778, "learning_rate": 9.789237840671118e-05, "loss": 0.0442, "step": 11050 }, { "grad_norm": 0.47734901309013367, "learning_rate": 9.78864367684066e-05, "loss": 0.0505, "step": 11060 }, { "grad_norm": 0.5474944710731506, "learning_rate": 9.788048694761968e-05, "loss": 0.042, "step": 11070 }, { "grad_norm": 0.3965405821800232, "learning_rate": 9.787452894536709e-05, "loss": 0.0475, "step": 11080 }, { "grad_norm": 0.3876264989376068, "learning_rate": 9.786856276266685e-05, "loss": 0.0463, "step": 11090 }, { "grad_norm": 0.44125840067863464, "learning_rate": 9.786258840053845e-05, "loss": 0.0518, "step": 11100 }, { "grad_norm": 0.34104862809181213, "learning_rate": 9.785660586000273e-05, "loss": 0.0463, "step": 11110 }, { "grad_norm": 0.3618362247943878, "learning_rate": 9.785061514208196e-05, "loss": 0.0426, "step": 11120 }, { "grad_norm": 0.25956135988235474, "learning_rate": 9.784461624779977e-05, "loss": 0.048, "step": 11130 }, { "grad_norm": 0.4558226466178894, "learning_rate": 9.783860917818123e-05, "loss": 0.0443, "step": 11140 }, { "grad_norm": 0.375018835067749, "learning_rate": 9.783259393425277e-05, "loss": 0.0328, "step": 11150 }, { "grad_norm": 0.3355352580547333, "learning_rate": 9.782657051704221e-05, "loss": 0.041, "step": 11160 }, { "grad_norm": 0.38274675607681274, "learning_rate": 9.782053892757883e-05, "loss": 0.0398, "step": 11170 }, { "grad_norm": 0.34507516026496887, "learning_rate": 9.781449916689324e-05, "loss": 0.0415, "step": 11180 }, { "grad_norm": 0.3608453869819641, "learning_rate": 9.780845123601746e-05, "loss": 0.045, "step": 11190 }, { "grad_norm": 0.3968375325202942, "learning_rate": 9.780239513598492e-05, "loss": 0.0467, "step": 11200 }, { "grad_norm": 0.37690460681915283, "learning_rate": 9.779633086783047e-05, "loss": 0.048, "step": 11210 }, { "grad_norm": 0.4004824161529541, "learning_rate": 9.779025843259031e-05, "loss": 0.046, "step": 11220 }, { "grad_norm": 0.4174762964248657, "learning_rate": 9.778417783130204e-05, "loss": 0.0628, "step": 11230 }, { "grad_norm": 0.30992674827575684, "learning_rate": 9.777808906500468e-05, "loss": 0.0423, "step": 11240 }, { "grad_norm": 0.3745494484901428, "learning_rate": 9.777199213473862e-05, "loss": 0.0441, "step": 11250 }, { "grad_norm": 0.34088823199272156, "learning_rate": 9.77658870415457e-05, "loss": 0.0352, "step": 11260 }, { "grad_norm": 0.4625767767429352, "learning_rate": 9.775977378646906e-05, "loss": 0.0437, "step": 11270 }, { "grad_norm": 0.3811945617198944, "learning_rate": 9.775365237055331e-05, "loss": 0.0507, "step": 11280 }, { "grad_norm": 0.5522233247756958, "learning_rate": 9.774752279484445e-05, "loss": 0.0418, "step": 11290 }, { "grad_norm": 0.3511815667152405, "learning_rate": 9.774138506038984e-05, "loss": 0.0404, "step": 11300 }, { "grad_norm": 0.38376134634017944, "learning_rate": 9.773523916823826e-05, "loss": 0.0421, "step": 11310 }, { "grad_norm": 0.3767261505126953, "learning_rate": 9.772908511943986e-05, "loss": 0.0422, "step": 11320 }, { "grad_norm": 0.7356544733047485, "learning_rate": 9.77229229150462e-05, "loss": 0.0414, "step": 11330 }, { "grad_norm": 1.118933081626892, "learning_rate": 9.771675255611024e-05, "loss": 0.0362, "step": 11340 }, { "grad_norm": 0.34687867760658264, "learning_rate": 9.771057404368632e-05, "loss": 0.0589, "step": 11350 }, { "grad_norm": 0.4225882887840271, "learning_rate": 9.770438737883018e-05, "loss": 0.0441, "step": 11360 }, { "grad_norm": 0.4057016670703888, "learning_rate": 9.769819256259898e-05, "loss": 0.0374, "step": 11370 }, { "grad_norm": 0.5016815662384033, "learning_rate": 9.769198959605119e-05, "loss": 0.044, "step": 11380 }, { "grad_norm": 0.2966436743736267, "learning_rate": 9.768577848024678e-05, "loss": 0.05, "step": 11390 }, { "grad_norm": 0.2840026617050171, "learning_rate": 9.767955921624702e-05, "loss": 0.0503, "step": 11400 }, { "grad_norm": 0.44547516107559204, "learning_rate": 9.767333180511465e-05, "loss": 0.0435, "step": 11410 }, { "grad_norm": 0.3871667981147766, "learning_rate": 9.766709624791373e-05, "loss": 0.0484, "step": 11420 }, { "grad_norm": 0.3477819263935089, "learning_rate": 9.766085254570975e-05, "loss": 0.0439, "step": 11430 }, { "grad_norm": 0.32710254192352295, "learning_rate": 9.76546006995696e-05, "loss": 0.0407, "step": 11440 }, { "grad_norm": 0.6121784448623657, "learning_rate": 9.764834071056155e-05, "loss": 0.0409, "step": 11450 }, { "grad_norm": 0.4782467186450958, "learning_rate": 9.764207257975526e-05, "loss": 0.0442, "step": 11460 }, { "grad_norm": 0.32541462779045105, "learning_rate": 9.763579630822179e-05, "loss": 0.0364, "step": 11470 }, { "grad_norm": 0.25995200872421265, "learning_rate": 9.762951189703356e-05, "loss": 0.0378, "step": 11480 }, { "grad_norm": 0.3301006853580475, "learning_rate": 9.762321934726442e-05, "loss": 0.0421, "step": 11490 }, { "grad_norm": 0.3059641420841217, "learning_rate": 9.761691865998959e-05, "loss": 0.039, "step": 11500 }, { "grad_norm": 0.671093761920929, "learning_rate": 9.76106098362857e-05, "loss": 0.0401, "step": 11510 }, { "grad_norm": 0.3269205689430237, "learning_rate": 9.760429287723072e-05, "loss": 0.0389, "step": 11520 }, { "grad_norm": 0.4503408372402191, "learning_rate": 9.759796778390406e-05, "loss": 0.0494, "step": 11530 }, { "grad_norm": 0.6633976697921753, "learning_rate": 9.759163455738653e-05, "loss": 0.0451, "step": 11540 }, { "grad_norm": 0.37320584058761597, "learning_rate": 9.75852931987603e-05, "loss": 0.046, "step": 11550 }, { "grad_norm": 0.2778657078742981, "learning_rate": 9.757894370910891e-05, "loss": 0.0396, "step": 11560 }, { "grad_norm": 0.4349772036075592, "learning_rate": 9.757258608951733e-05, "loss": 0.0422, "step": 11570 }, { "grad_norm": 0.438893586397171, "learning_rate": 9.75662203410719e-05, "loss": 0.0437, "step": 11580 }, { "grad_norm": 0.43451210856437683, "learning_rate": 9.755984646486034e-05, "loss": 0.0463, "step": 11590 }, { "grad_norm": 0.3236117959022522, "learning_rate": 9.75534644619718e-05, "loss": 0.0431, "step": 11600 }, { "grad_norm": 0.4067523777484894, "learning_rate": 9.754707433349676e-05, "loss": 0.0553, "step": 11610 }, { "grad_norm": 0.39208340644836426, "learning_rate": 9.754067608052715e-05, "loss": 0.0533, "step": 11620 }, { "grad_norm": 0.5525875091552734, "learning_rate": 9.753426970415622e-05, "loss": 0.0643, "step": 11630 }, { "grad_norm": 1.5204483270645142, "learning_rate": 9.752785520547868e-05, "loss": 0.043, "step": 11640 }, { "grad_norm": 0.3257596790790558, "learning_rate": 9.752143258559056e-05, "loss": 0.0484, "step": 11650 }, { "grad_norm": 0.606766402721405, "learning_rate": 9.751500184558933e-05, "loss": 0.0426, "step": 11660 }, { "grad_norm": 0.5295853018760681, "learning_rate": 9.750856298657383e-05, "loss": 0.042, "step": 11670 }, { "grad_norm": 0.4119164049625397, "learning_rate": 9.750211600964428e-05, "loss": 0.0494, "step": 11680 }, { "grad_norm": 0.34326452016830444, "learning_rate": 9.749566091590226e-05, "loss": 0.0433, "step": 11690 }, { "grad_norm": 0.3584167957305908, "learning_rate": 9.748919770645083e-05, "loss": 0.0412, "step": 11700 }, { "grad_norm": 0.5710134506225586, "learning_rate": 9.748272638239432e-05, "loss": 0.045, "step": 11710 }, { "grad_norm": 1.1451373100280762, "learning_rate": 9.747624694483855e-05, "loss": 0.0498, "step": 11720 }, { "grad_norm": 0.4417473077774048, "learning_rate": 9.746975939489065e-05, "loss": 0.0485, "step": 11730 }, { "grad_norm": 0.3932982087135315, "learning_rate": 9.746326373365918e-05, "loss": 0.0405, "step": 11740 }, { "grad_norm": 0.2852630913257599, "learning_rate": 9.745675996225403e-05, "loss": 0.0555, "step": 11750 }, { "grad_norm": 0.43774470686912537, "learning_rate": 9.745024808178657e-05, "loss": 0.041, "step": 11760 }, { "grad_norm": 0.3344358801841736, "learning_rate": 9.744372809336947e-05, "loss": 0.045, "step": 11770 }, { "grad_norm": 0.45807233452796936, "learning_rate": 9.743719999811682e-05, "loss": 0.0442, "step": 11780 }, { "grad_norm": 0.5418578386306763, "learning_rate": 9.743066379714412e-05, "loss": 0.0495, "step": 11790 }, { "grad_norm": 0.32031920552253723, "learning_rate": 9.74241194915682e-05, "loss": 0.043, "step": 11800 }, { "grad_norm": 0.3330762982368469, "learning_rate": 9.741756708250731e-05, "loss": 0.048, "step": 11810 }, { "grad_norm": 0.38506242632865906, "learning_rate": 9.741100657108109e-05, "loss": 0.04, "step": 11820 }, { "grad_norm": 0.7344732284545898, "learning_rate": 9.740443795841054e-05, "loss": 0.0493, "step": 11830 }, { "grad_norm": 0.34319913387298584, "learning_rate": 9.739786124561805e-05, "loss": 0.0405, "step": 11840 }, { "grad_norm": 0.3227331042289734, "learning_rate": 9.73912764338274e-05, "loss": 0.0455, "step": 11850 }, { "grad_norm": 0.46654853224754333, "learning_rate": 9.738468352416377e-05, "loss": 0.0475, "step": 11860 }, { "grad_norm": 0.30626726150512695, "learning_rate": 9.737808251775369e-05, "loss": 0.0341, "step": 11870 }, { "grad_norm": 1.6833974123001099, "learning_rate": 9.737147341572512e-05, "loss": 0.0514, "step": 11880 }, { "grad_norm": 0.5961723327636719, "learning_rate": 9.736485621920735e-05, "loss": 0.0377, "step": 11890 }, { "grad_norm": 0.43959489464759827, "learning_rate": 9.735823092933108e-05, "loss": 0.0439, "step": 11900 }, { "grad_norm": 0.38908815383911133, "learning_rate": 9.735159754722838e-05, "loss": 0.0393, "step": 11910 }, { "grad_norm": 0.4623761475086212, "learning_rate": 9.734495607403275e-05, "loss": 0.0392, "step": 11920 }, { "grad_norm": 0.346216082572937, "learning_rate": 9.733830651087901e-05, "loss": 0.043, "step": 11930 }, { "grad_norm": 0.4112968146800995, "learning_rate": 9.733164885890338e-05, "loss": 0.0406, "step": 11940 }, { "grad_norm": 0.3436206579208374, "learning_rate": 9.732498311924349e-05, "loss": 0.0346, "step": 11950 }, { "grad_norm": 0.40212124586105347, "learning_rate": 9.731830929303833e-05, "loss": 0.0403, "step": 11960 }, { "grad_norm": 0.4200741946697235, "learning_rate": 9.731162738142827e-05, "loss": 0.0375, "step": 11970 }, { "grad_norm": 0.38448548316955566, "learning_rate": 9.730493738555506e-05, "loss": 0.0455, "step": 11980 }, { "grad_norm": 0.36234596371650696, "learning_rate": 9.729823930656186e-05, "loss": 0.0431, "step": 11990 }, { "grad_norm": 0.41602152585983276, "learning_rate": 9.729153314559316e-05, "loss": 0.0539, "step": 12000 }, { "grad_norm": 0.3648545742034912, "learning_rate": 9.728481890379486e-05, "loss": 0.0399, "step": 12010 }, { "grad_norm": 0.38861167430877686, "learning_rate": 9.727809658231428e-05, "loss": 0.038, "step": 12020 }, { "grad_norm": 0.41171756386756897, "learning_rate": 9.727136618230003e-05, "loss": 0.0399, "step": 12030 }, { "grad_norm": 0.48610517382621765, "learning_rate": 9.726462770490219e-05, "loss": 0.0522, "step": 12040 }, { "grad_norm": 0.30361080169677734, "learning_rate": 9.725788115127214e-05, "loss": 0.0553, "step": 12050 }, { "grad_norm": 0.818155825138092, "learning_rate": 9.725112652256274e-05, "loss": 0.0463, "step": 12060 }, { "grad_norm": 0.6620181798934937, "learning_rate": 9.724436381992812e-05, "loss": 0.0596, "step": 12070 }, { "grad_norm": 0.37999197840690613, "learning_rate": 9.723759304452387e-05, "loss": 0.0463, "step": 12080 }, { "grad_norm": 0.4077428877353668, "learning_rate": 9.72308141975069e-05, "loss": 0.0391, "step": 12090 }, { "grad_norm": 0.468179851770401, "learning_rate": 9.722402728003557e-05, "loss": 0.0511, "step": 12100 }, { "grad_norm": 0.48547884821891785, "learning_rate": 9.721723229326953e-05, "loss": 0.0462, "step": 12110 }, { "grad_norm": 0.4029490053653717, "learning_rate": 9.721042923836992e-05, "loss": 0.0355, "step": 12120 }, { "grad_norm": 0.32248610258102417, "learning_rate": 9.720361811649914e-05, "loss": 0.039, "step": 12130 }, { "grad_norm": 0.4084973633289337, "learning_rate": 9.719679892882106e-05, "loss": 0.0417, "step": 12140 }, { "grad_norm": 0.34754645824432373, "learning_rate": 9.718997167650085e-05, "loss": 0.046, "step": 12150 }, { "grad_norm": 0.4828566610813141, "learning_rate": 9.718313636070515e-05, "loss": 0.0394, "step": 12160 }, { "grad_norm": 0.3112391531467438, "learning_rate": 9.717629298260192e-05, "loss": 0.037, "step": 12170 }, { "grad_norm": 0.3310001492500305, "learning_rate": 9.716944154336047e-05, "loss": 0.0498, "step": 12180 }, { "grad_norm": 0.3819175362586975, "learning_rate": 9.716258204415157e-05, "loss": 0.0403, "step": 12190 }, { "grad_norm": 0.46003273129463196, "learning_rate": 9.715571448614728e-05, "loss": 0.0507, "step": 12200 }, { "grad_norm": 0.4942564368247986, "learning_rate": 9.71488388705211e-05, "loss": 0.061, "step": 12210 }, { "grad_norm": 0.3609003722667694, "learning_rate": 9.714195519844788e-05, "loss": 0.0435, "step": 12220 }, { "grad_norm": 0.6469032168388367, "learning_rate": 9.713506347110386e-05, "loss": 0.0429, "step": 12230 }, { "grad_norm": 0.34694358706474304, "learning_rate": 9.712816368966662e-05, "loss": 0.0395, "step": 12240 }, { "grad_norm": 0.35333746671676636, "learning_rate": 9.712125585531517e-05, "loss": 0.0486, "step": 12250 }, { "grad_norm": 0.3868130147457123, "learning_rate": 9.711433996922988e-05, "loss": 0.0467, "step": 12260 }, { "grad_norm": 0.9515573978424072, "learning_rate": 9.710741603259245e-05, "loss": 0.0448, "step": 12270 }, { "grad_norm": 0.34649282693862915, "learning_rate": 9.710048404658603e-05, "loss": 0.038, "step": 12280 }, { "grad_norm": 0.39094096422195435, "learning_rate": 9.709354401239508e-05, "loss": 0.0405, "step": 12290 }, { "grad_norm": 0.31523677706718445, "learning_rate": 9.708659593120546e-05, "loss": 0.0314, "step": 12300 }, { "grad_norm": 0.41629016399383545, "learning_rate": 9.707963980420443e-05, "loss": 0.0406, "step": 12310 }, { "grad_norm": 0.4176892638206482, "learning_rate": 9.707267563258058e-05, "loss": 0.0442, "step": 12320 }, { "grad_norm": 0.40577757358551025, "learning_rate": 9.70657034175239e-05, "loss": 0.0373, "step": 12330 }, { "grad_norm": 0.24805422127246857, "learning_rate": 9.705872316022577e-05, "loss": 0.0418, "step": 12340 }, { "grad_norm": 0.3678344488143921, "learning_rate": 9.705173486187891e-05, "loss": 0.0414, "step": 12350 }, { "grad_norm": 0.33329805731773376, "learning_rate": 9.704473852367741e-05, "loss": 0.0426, "step": 12360 }, { "grad_norm": 0.5761215090751648, "learning_rate": 9.70377341468168e-05, "loss": 0.0397, "step": 12370 }, { "grad_norm": 0.33387792110443115, "learning_rate": 9.703072173249389e-05, "loss": 0.0381, "step": 12380 }, { "grad_norm": 0.3102268576622009, "learning_rate": 9.702370128190693e-05, "loss": 0.0444, "step": 12390 }, { "grad_norm": 0.4649679958820343, "learning_rate": 9.701667279625552e-05, "loss": 0.0419, "step": 12400 }, { "grad_norm": 0.5040677785873413, "learning_rate": 9.700963627674065e-05, "loss": 0.0484, "step": 12410 }, { "grad_norm": 0.36119845509529114, "learning_rate": 9.700259172456466e-05, "loss": 0.0295, "step": 12420 }, { "grad_norm": 0.3895125091075897, "learning_rate": 9.699553914093124e-05, "loss": 0.0497, "step": 12430 }, { "grad_norm": 0.3734954595565796, "learning_rate": 9.698847852704553e-05, "loss": 0.0451, "step": 12440 }, { "grad_norm": 0.32811635732650757, "learning_rate": 9.6981409884114e-05, "loss": 0.0342, "step": 12450 }, { "grad_norm": 0.8866487145423889, "learning_rate": 9.697433321334443e-05, "loss": 0.0409, "step": 12460 }, { "grad_norm": 0.3807670474052429, "learning_rate": 9.696724851594607e-05, "loss": 0.0396, "step": 12470 }, { "grad_norm": 1.1068288087844849, "learning_rate": 9.696015579312952e-05, "loss": 0.0412, "step": 12480 }, { "grad_norm": 0.37381500005722046, "learning_rate": 9.695305504610668e-05, "loss": 0.0425, "step": 12490 }, { "grad_norm": 0.3661348521709442, "learning_rate": 9.694594627609092e-05, "loss": 0.062, "step": 12500 }, { "grad_norm": 0.39922311902046204, "learning_rate": 9.693882948429691e-05, "loss": 0.0587, "step": 12510 }, { "grad_norm": 0.590846061706543, "learning_rate": 9.693170467194071e-05, "loss": 0.0509, "step": 12520 }, { "grad_norm": 0.5958951711654663, "learning_rate": 9.692457184023977e-05, "loss": 0.0455, "step": 12530 }, { "grad_norm": 0.5030877590179443, "learning_rate": 9.691743099041291e-05, "loss": 0.0524, "step": 12540 }, { "grad_norm": 0.39698123931884766, "learning_rate": 9.691028212368027e-05, "loss": 0.0486, "step": 12550 }, { "grad_norm": 0.3540841341018677, "learning_rate": 9.690312524126342e-05, "loss": 0.0338, "step": 12560 }, { "grad_norm": 0.28128719329833984, "learning_rate": 9.689596034438527e-05, "loss": 0.0378, "step": 12570 }, { "grad_norm": 0.2303953617811203, "learning_rate": 9.688878743427012e-05, "loss": 0.0357, "step": 12580 }, { "grad_norm": 0.22692030668258667, "learning_rate": 9.688160651214359e-05, "loss": 0.0433, "step": 12590 }, { "grad_norm": 0.46077755093574524, "learning_rate": 9.687441757923273e-05, "loss": 0.0403, "step": 12600 }, { "grad_norm": 0.36415836215019226, "learning_rate": 9.68672206367659e-05, "loss": 0.0322, "step": 12610 }, { "grad_norm": 0.3560256063938141, "learning_rate": 9.686001568597291e-05, "loss": 0.0349, "step": 12620 }, { "grad_norm": 0.2832796275615692, "learning_rate": 9.685280272808486e-05, "loss": 0.0269, "step": 12630 }, { "grad_norm": 0.4097214937210083, "learning_rate": 9.684558176433424e-05, "loss": 0.0501, "step": 12640 }, { "grad_norm": 0.45912113785743713, "learning_rate": 9.683835279595495e-05, "loss": 0.0385, "step": 12650 }, { "grad_norm": 0.41483643651008606, "learning_rate": 9.683111582418216e-05, "loss": 0.0517, "step": 12660 }, { "grad_norm": 0.32688072323799133, "learning_rate": 9.682387085025254e-05, "loss": 0.0425, "step": 12670 }, { "grad_norm": 0.4765015244483948, "learning_rate": 9.681661787540401e-05, "loss": 0.0474, "step": 12680 }, { "grad_norm": 0.3834018409252167, "learning_rate": 9.680935690087593e-05, "loss": 0.0414, "step": 12690 }, { "grad_norm": 0.5711590647697449, "learning_rate": 9.680208792790901e-05, "loss": 0.0426, "step": 12700 }, { "grad_norm": 0.3220374286174774, "learning_rate": 9.679481095774529e-05, "loss": 0.0341, "step": 12710 }, { "grad_norm": 0.3552662432193756, "learning_rate": 9.678752599162822e-05, "loss": 0.04, "step": 12720 }, { "grad_norm": 3.2519829273223877, "learning_rate": 9.678023303080259e-05, "loss": 0.0359, "step": 12730 }, { "grad_norm": 0.40219011902809143, "learning_rate": 9.677293207651459e-05, "loss": 0.0384, "step": 12740 }, { "grad_norm": 0.3075340986251831, "learning_rate": 9.676562313001173e-05, "loss": 0.0355, "step": 12750 }, { "grad_norm": 0.3922475576400757, "learning_rate": 9.675830619254293e-05, "loss": 0.0514, "step": 12760 }, { "grad_norm": 0.36014324426651, "learning_rate": 9.675098126535843e-05, "loss": 0.0487, "step": 12770 }, { "grad_norm": 0.3379684388637543, "learning_rate": 9.674364834970988e-05, "loss": 0.0452, "step": 12780 }, { "grad_norm": 0.37071385979652405, "learning_rate": 9.673630744685028e-05, "loss": 0.0434, "step": 12790 }, { "grad_norm": 0.48798078298568726, "learning_rate": 9.672895855803397e-05, "loss": 0.047, "step": 12800 }, { "grad_norm": 0.3380456566810608, "learning_rate": 9.672160168451667e-05, "loss": 0.0396, "step": 12810 }, { "grad_norm": 0.3826441466808319, "learning_rate": 9.671423682755549e-05, "loss": 0.0387, "step": 12820 }, { "grad_norm": 0.5119278430938721, "learning_rate": 9.670686398840888e-05, "loss": 0.0405, "step": 12830 }, { "grad_norm": 0.4205002188682556, "learning_rate": 9.669948316833664e-05, "loss": 0.0371, "step": 12840 }, { "grad_norm": 0.4074099063873291, "learning_rate": 9.669209436859997e-05, "loss": 0.0404, "step": 12850 }, { "grad_norm": 0.32782167196273804, "learning_rate": 9.66846975904614e-05, "loss": 0.046, "step": 12860 }, { "grad_norm": 0.47221314907073975, "learning_rate": 9.667729283518483e-05, "loss": 0.0384, "step": 12870 }, { "grad_norm": 0.4883868098258972, "learning_rate": 9.666988010403557e-05, "loss": 0.0372, "step": 12880 }, { "grad_norm": 0.34806057810783386, "learning_rate": 9.66624593982802e-05, "loss": 0.0401, "step": 12890 }, { "grad_norm": 0.3229653835296631, "learning_rate": 9.665503071918675e-05, "loss": 0.047, "step": 12900 }, { "grad_norm": 0.3382420837879181, "learning_rate": 9.664759406802456e-05, "loss": 0.0368, "step": 12910 }, { "grad_norm": 0.3619127571582794, "learning_rate": 9.664014944606437e-05, "loss": 0.0329, "step": 12920 }, { "grad_norm": 0.49659809470176697, "learning_rate": 9.663269685457822e-05, "loss": 0.0493, "step": 12930 }, { "grad_norm": 0.35873788595199585, "learning_rate": 9.662523629483962e-05, "loss": 0.0412, "step": 12940 }, { "grad_norm": 0.28046345710754395, "learning_rate": 9.661776776812333e-05, "loss": 0.0324, "step": 12950 }, { "grad_norm": 0.4612137973308563, "learning_rate": 9.661029127570553e-05, "loss": 0.0315, "step": 12960 }, { "grad_norm": 0.33488354086875916, "learning_rate": 9.660280681886373e-05, "loss": 0.0349, "step": 12970 }, { "grad_norm": 0.36818188428878784, "learning_rate": 9.659531439887685e-05, "loss": 0.0305, "step": 12980 }, { "grad_norm": 0.35801932215690613, "learning_rate": 9.658781401702511e-05, "loss": 0.0327, "step": 12990 }, { "grad_norm": 0.39231258630752563, "learning_rate": 9.658030567459015e-05, "loss": 0.0369, "step": 13000 }, { "grad_norm": 1.4202344417572021, "learning_rate": 9.65727893728549e-05, "loss": 0.0577, "step": 13010 }, { "grad_norm": 0.390693724155426, "learning_rate": 9.656526511310375e-05, "loss": 0.0487, "step": 13020 }, { "grad_norm": 0.43774205446243286, "learning_rate": 9.655773289662233e-05, "loss": 0.0506, "step": 13030 }, { "grad_norm": 0.34946081042289734, "learning_rate": 9.655019272469772e-05, "loss": 0.0348, "step": 13040 }, { "grad_norm": 0.2594813108444214, "learning_rate": 9.654264459861832e-05, "loss": 0.0339, "step": 13050 }, { "grad_norm": 0.419364333152771, "learning_rate": 9.653508851967391e-05, "loss": 0.0342, "step": 13060 }, { "grad_norm": 0.3577711880207062, "learning_rate": 9.65275244891556e-05, "loss": 0.0473, "step": 13070 }, { "grad_norm": 0.33596062660217285, "learning_rate": 9.651995250835591e-05, "loss": 0.0449, "step": 13080 }, { "grad_norm": 0.41247642040252686, "learning_rate": 9.651237257856862e-05, "loss": 0.0421, "step": 13090 }, { "grad_norm": 0.2664695084095001, "learning_rate": 9.6504784701089e-05, "loss": 0.0366, "step": 13100 }, { "grad_norm": 0.4610074758529663, "learning_rate": 9.649718887721357e-05, "loss": 0.0353, "step": 13110 }, { "grad_norm": 0.3604312539100647, "learning_rate": 9.648958510824028e-05, "loss": 0.0281, "step": 13120 }, { "grad_norm": 0.4756093919277191, "learning_rate": 9.648197339546837e-05, "loss": 0.0394, "step": 13130 }, { "grad_norm": 0.3804180920124054, "learning_rate": 9.647435374019851e-05, "loss": 0.0441, "step": 13140 }, { "grad_norm": 0.352541446685791, "learning_rate": 9.646672614373266e-05, "loss": 0.0401, "step": 13150 }, { "grad_norm": 0.3217868506908417, "learning_rate": 9.645909060737418e-05, "loss": 0.0278, "step": 13160 }, { "grad_norm": 0.30538633465766907, "learning_rate": 9.645144713242778e-05, "loss": 0.0416, "step": 13170 }, { "grad_norm": 0.4209161400794983, "learning_rate": 9.64437957201995e-05, "loss": 0.0419, "step": 13180 }, { "grad_norm": 0.4187161922454834, "learning_rate": 9.643613637199678e-05, "loss": 0.0353, "step": 13190 }, { "grad_norm": 0.33333733677864075, "learning_rate": 9.642846908912839e-05, "loss": 0.0441, "step": 13200 }, { "grad_norm": 0.368192583322525, "learning_rate": 9.642079387290444e-05, "loss": 0.0325, "step": 13210 }, { "grad_norm": 0.328721821308136, "learning_rate": 9.641311072463644e-05, "loss": 0.033, "step": 13220 }, { "grad_norm": 0.9411596655845642, "learning_rate": 9.640541964563722e-05, "loss": 0.0462, "step": 13230 }, { "grad_norm": 0.3947920799255371, "learning_rate": 9.639772063722096e-05, "loss": 0.0516, "step": 13240 }, { "grad_norm": 0.3969920575618744, "learning_rate": 9.639001370070324e-05, "loss": 0.0402, "step": 13250 }, { "grad_norm": 0.4165513217449188, "learning_rate": 9.638229883740095e-05, "loss": 0.0499, "step": 13260 }, { "grad_norm": 0.3819199800491333, "learning_rate": 9.637457604863233e-05, "loss": 0.0408, "step": 13270 }, { "grad_norm": 0.40653225779533386, "learning_rate": 9.636684533571703e-05, "loss": 0.035, "step": 13280 }, { "grad_norm": 0.4030735492706299, "learning_rate": 9.635910669997599e-05, "loss": 0.0428, "step": 13290 }, { "grad_norm": 0.3894065022468567, "learning_rate": 9.635136014273154e-05, "loss": 0.0409, "step": 13300 }, { "grad_norm": 0.9202561378479004, "learning_rate": 9.634360566530735e-05, "loss": 0.0395, "step": 13310 }, { "grad_norm": 0.4241218566894531, "learning_rate": 9.633584326902845e-05, "loss": 0.0511, "step": 13320 }, { "grad_norm": 0.38334015011787415, "learning_rate": 9.632807295522124e-05, "loss": 0.0442, "step": 13330 }, { "grad_norm": 0.32002127170562744, "learning_rate": 9.632029472521342e-05, "loss": 0.0432, "step": 13340 }, { "grad_norm": 0.32931986451148987, "learning_rate": 9.631250858033409e-05, "loss": 0.0361, "step": 13350 }, { "grad_norm": 0.756092369556427, "learning_rate": 9.630471452191371e-05, "loss": 0.041, "step": 13360 }, { "grad_norm": 0.5080056190490723, "learning_rate": 9.629691255128405e-05, "loss": 0.0459, "step": 13370 }, { "grad_norm": 0.45423850417137146, "learning_rate": 9.628910266977825e-05, "loss": 0.0401, "step": 13380 }, { "grad_norm": 0.45244720578193665, "learning_rate": 9.628128487873083e-05, "loss": 0.0434, "step": 13390 }, { "grad_norm": 0.8717120885848999, "learning_rate": 9.627345917947761e-05, "loss": 0.0569, "step": 13400 }, { "grad_norm": 0.38650450110435486, "learning_rate": 9.626562557335579e-05, "loss": 0.0438, "step": 13410 }, { "grad_norm": 0.34985771775245667, "learning_rate": 9.625778406170393e-05, "loss": 0.0415, "step": 13420 }, { "grad_norm": 0.48259127140045166, "learning_rate": 9.624993464586193e-05, "loss": 0.0369, "step": 13430 }, { "grad_norm": 0.3768957257270813, "learning_rate": 9.624207732717105e-05, "loss": 0.0372, "step": 13440 }, { "grad_norm": 0.4156528115272522, "learning_rate": 9.623421210697386e-05, "loss": 0.0441, "step": 13450 }, { "grad_norm": 0.3136962056159973, "learning_rate": 9.622633898661434e-05, "loss": 0.043, "step": 13460 }, { "grad_norm": 0.4033057689666748, "learning_rate": 9.621845796743778e-05, "loss": 0.0506, "step": 13470 }, { "grad_norm": 0.3066248893737793, "learning_rate": 9.621056905079082e-05, "loss": 0.0454, "step": 13480 }, { "grad_norm": 0.6138759255409241, "learning_rate": 9.620267223802149e-05, "loss": 0.0485, "step": 13490 }, { "grad_norm": 0.49693500995635986, "learning_rate": 9.619476753047911e-05, "loss": 0.0365, "step": 13500 }, { "grad_norm": 0.3359806537628174, "learning_rate": 9.618685492951438e-05, "loss": 0.0436, "step": 13510 }, { "grad_norm": 0.3396463692188263, "learning_rate": 9.617893443647938e-05, "loss": 0.0485, "step": 13520 }, { "grad_norm": 0.30093663930892944, "learning_rate": 9.617100605272746e-05, "loss": 0.0492, "step": 13530 }, { "grad_norm": 2.2388312816619873, "learning_rate": 9.616306977961338e-05, "loss": 0.0465, "step": 13540 }, { "grad_norm": 0.26683953404426575, "learning_rate": 9.615512561849326e-05, "loss": 0.0357, "step": 13550 }, { "grad_norm": 0.4045572280883789, "learning_rate": 9.61471735707245e-05, "loss": 0.0453, "step": 13560 }, { "grad_norm": 0.42745232582092285, "learning_rate": 9.613921363766592e-05, "loss": 0.0302, "step": 13570 }, { "grad_norm": 0.4010466933250427, "learning_rate": 9.613124582067763e-05, "loss": 0.046, "step": 13580 }, { "grad_norm": 0.37215766310691833, "learning_rate": 9.612327012112112e-05, "loss": 0.0399, "step": 13590 }, { "grad_norm": 0.34109026193618774, "learning_rate": 9.611528654035921e-05, "loss": 0.0551, "step": 13600 }, { "grad_norm": 1.2063298225402832, "learning_rate": 9.610729507975611e-05, "loss": 0.0364, "step": 13610 }, { "grad_norm": 0.3227570652961731, "learning_rate": 9.609929574067731e-05, "loss": 0.0567, "step": 13620 }, { "grad_norm": 0.3122597336769104, "learning_rate": 9.609128852448967e-05, "loss": 0.041, "step": 13630 }, { "grad_norm": 0.43234366178512573, "learning_rate": 9.608327343256143e-05, "loss": 0.043, "step": 13640 }, { "grad_norm": 0.35500568151474, "learning_rate": 9.607525046626216e-05, "loss": 0.0411, "step": 13650 }, { "grad_norm": 0.41927385330200195, "learning_rate": 9.606721962696272e-05, "loss": 0.044, "step": 13660 }, { "grad_norm": 0.37987929582595825, "learning_rate": 9.60591809160354e-05, "loss": 0.0348, "step": 13670 }, { "grad_norm": 0.26177847385406494, "learning_rate": 9.605113433485378e-05, "loss": 0.0298, "step": 13680 }, { "grad_norm": 0.34613481163978577, "learning_rate": 9.604307988479279e-05, "loss": 0.035, "step": 13690 }, { "grad_norm": 0.3524908721446991, "learning_rate": 9.603501756722876e-05, "loss": 0.036, "step": 13700 }, { "grad_norm": 0.5528252720832825, "learning_rate": 9.602694738353927e-05, "loss": 0.0429, "step": 13710 }, { "grad_norm": 0.34562447667121887, "learning_rate": 9.601886933510331e-05, "loss": 0.0409, "step": 13720 }, { "grad_norm": 0.3856247663497925, "learning_rate": 9.60107834233012e-05, "loss": 0.0332, "step": 13730 }, { "grad_norm": 0.2898196280002594, "learning_rate": 9.60026896495146e-05, "loss": 0.0324, "step": 13740 }, { "grad_norm": 0.2868691384792328, "learning_rate": 9.599458801512652e-05, "loss": 0.0401, "step": 13750 }, { "grad_norm": 0.28549298644065857, "learning_rate": 9.598647852152129e-05, "loss": 0.0528, "step": 13760 }, { "grad_norm": 0.2651842534542084, "learning_rate": 9.597836117008462e-05, "loss": 0.0383, "step": 13770 }, { "grad_norm": 0.3247760236263275, "learning_rate": 9.597023596220356e-05, "loss": 0.0368, "step": 13780 }, { "grad_norm": 0.3534371554851532, "learning_rate": 9.596210289926643e-05, "loss": 0.0384, "step": 13790 }, { "grad_norm": 1.7245795726776123, "learning_rate": 9.5953961982663e-05, "loss": 0.0344, "step": 13800 }, { "grad_norm": 0.3456549644470215, "learning_rate": 9.594581321378431e-05, "loss": 0.043, "step": 13810 }, { "grad_norm": 0.24575696885585785, "learning_rate": 9.593765659402276e-05, "loss": 0.0427, "step": 13820 }, { "grad_norm": 0.4389885365962982, "learning_rate": 9.59294921247721e-05, "loss": 0.0397, "step": 13830 }, { "grad_norm": 0.460470587015152, "learning_rate": 9.59213198074274e-05, "loss": 0.0434, "step": 13840 }, { "grad_norm": 0.4435124099254608, "learning_rate": 9.59131396433851e-05, "loss": 0.0394, "step": 13850 }, { "grad_norm": 0.45424389839172363, "learning_rate": 9.590495163404297e-05, "loss": 0.0478, "step": 13860 }, { "grad_norm": 0.2540106773376465, "learning_rate": 9.589675578080009e-05, "loss": 0.0398, "step": 13870 }, { "grad_norm": 0.37296533584594727, "learning_rate": 9.588855208505694e-05, "loss": 0.0319, "step": 13880 }, { "grad_norm": 0.35198870301246643, "learning_rate": 9.588034054821529e-05, "loss": 0.0418, "step": 13890 }, { "grad_norm": 0.2683415114879608, "learning_rate": 9.587212117167826e-05, "loss": 0.0273, "step": 13900 }, { "grad_norm": 0.31596165895462036, "learning_rate": 9.586389395685033e-05, "loss": 0.0321, "step": 13910 }, { "grad_norm": 0.2718057334423065, "learning_rate": 9.585565890513733e-05, "loss": 0.0409, "step": 13920 }, { "grad_norm": 0.38819420337677, "learning_rate": 9.584741601794636e-05, "loss": 0.0376, "step": 13930 }, { "grad_norm": 0.3188600242137909, "learning_rate": 9.58391652966859e-05, "loss": 0.037, "step": 13940 }, { "grad_norm": 0.33562591671943665, "learning_rate": 9.583090674276583e-05, "loss": 0.0341, "step": 13950 }, { "grad_norm": 0.2734396457672119, "learning_rate": 9.582264035759726e-05, "loss": 0.0409, "step": 13960 }, { "grad_norm": 0.3759593069553375, "learning_rate": 9.58143661425927e-05, "loss": 0.0582, "step": 13970 }, { "grad_norm": 0.31697872281074524, "learning_rate": 9.580608409916601e-05, "loss": 0.0434, "step": 13980 }, { "grad_norm": 0.43109554052352905, "learning_rate": 9.579779422873233e-05, "loss": 0.0433, "step": 13990 }, { "grad_norm": 0.37831223011016846, "learning_rate": 9.578949653270819e-05, "loss": 0.0386, "step": 14000 }, { "grad_norm": 0.26929643750190735, "learning_rate": 9.578119101251144e-05, "loss": 0.0392, "step": 14010 }, { "grad_norm": 0.288017213344574, "learning_rate": 9.577287766956127e-05, "loss": 0.0412, "step": 14020 }, { "grad_norm": 0.3825142979621887, "learning_rate": 9.57645565052782e-05, "loss": 0.0412, "step": 14030 }, { "grad_norm": 0.558864414691925, "learning_rate": 9.575622752108407e-05, "loss": 0.0401, "step": 14040 }, { "grad_norm": 0.30373314023017883, "learning_rate": 9.57478907184021e-05, "loss": 0.0335, "step": 14050 }, { "grad_norm": 0.2786952555179596, "learning_rate": 9.573954609865681e-05, "loss": 0.0377, "step": 14060 }, { "grad_norm": 0.3090585172176361, "learning_rate": 9.573119366327408e-05, "loss": 0.0428, "step": 14070 }, { "grad_norm": 2.6423120498657227, "learning_rate": 9.57228334136811e-05, "loss": 0.0401, "step": 14080 }, { "grad_norm": 0.2567603290081024, "learning_rate": 9.571446535130641e-05, "loss": 0.0331, "step": 14090 }, { "grad_norm": 0.31431689858436584, "learning_rate": 9.570608947757988e-05, "loss": 0.0408, "step": 14100 }, { "grad_norm": 0.3529722988605499, "learning_rate": 9.569770579393274e-05, "loss": 0.0387, "step": 14110 }, { "grad_norm": 0.5211558938026428, "learning_rate": 9.56893143017975e-05, "loss": 0.033, "step": 14120 }, { "grad_norm": 0.3384382426738739, "learning_rate": 9.568091500260806e-05, "loss": 0.0442, "step": 14130 }, { "grad_norm": 0.3231615722179413, "learning_rate": 9.567250789779961e-05, "loss": 0.04, "step": 14140 }, { "grad_norm": 0.32349875569343567, "learning_rate": 9.566409298880872e-05, "loss": 0.0393, "step": 14150 }, { "grad_norm": 0.4122556447982788, "learning_rate": 9.565567027707326e-05, "loss": 0.0544, "step": 14160 }, { "grad_norm": 0.3237845301628113, "learning_rate": 9.56472397640324e-05, "loss": 0.0473, "step": 14170 }, { "grad_norm": 0.5598138570785522, "learning_rate": 9.563880145112675e-05, "loss": 0.0391, "step": 14180 }, { "grad_norm": 0.5060837864875793, "learning_rate": 9.563035533979814e-05, "loss": 0.0391, "step": 14190 }, { "grad_norm": 0.3525603413581848, "learning_rate": 9.562190143148981e-05, "loss": 0.0461, "step": 14200 }, { "grad_norm": 0.38563331961631775, "learning_rate": 9.561343972764627e-05, "loss": 0.0407, "step": 14210 }, { "grad_norm": 0.4115765690803528, "learning_rate": 9.560497022971343e-05, "loss": 0.0435, "step": 14220 }, { "grad_norm": 0.2788725793361664, "learning_rate": 9.559649293913847e-05, "loss": 0.038, "step": 14230 }, { "grad_norm": 0.49581006169319153, "learning_rate": 9.558800785736993e-05, "loss": 0.0383, "step": 14240 }, { "grad_norm": 0.38598039746284485, "learning_rate": 9.557951498585767e-05, "loss": 0.0486, "step": 14250 }, { "grad_norm": 0.6572080254554749, "learning_rate": 9.557101432605293e-05, "loss": 0.0469, "step": 14260 }, { "grad_norm": 0.6477527618408203, "learning_rate": 9.556250587940818e-05, "loss": 0.0462, "step": 14270 }, { "grad_norm": 0.3559754490852356, "learning_rate": 9.555398964737734e-05, "loss": 0.04, "step": 14280 }, { "grad_norm": 0.4471298158168793, "learning_rate": 9.554546563141555e-05, "loss": 0.0357, "step": 14290 }, { "grad_norm": 0.3479751944541931, "learning_rate": 9.553693383297937e-05, "loss": 0.0386, "step": 14300 }, { "grad_norm": 0.43018633127212524, "learning_rate": 9.552839425352663e-05, "loss": 0.0401, "step": 14310 }, { "grad_norm": 0.3148028254508972, "learning_rate": 9.551984689451652e-05, "loss": 0.0636, "step": 14320 }, { "grad_norm": 0.36341342329978943, "learning_rate": 9.551129175740953e-05, "loss": 0.0424, "step": 14330 }, { "grad_norm": 0.3624456822872162, "learning_rate": 9.550272884366754e-05, "loss": 0.0404, "step": 14340 }, { "grad_norm": 0.36834409832954407, "learning_rate": 9.549415815475369e-05, "loss": 0.0414, "step": 14350 }, { "grad_norm": 0.2422669380903244, "learning_rate": 9.548557969213247e-05, "loss": 0.0325, "step": 14360 }, { "grad_norm": 0.29011139273643494, "learning_rate": 9.547699345726972e-05, "loss": 0.0383, "step": 14370 }, { "grad_norm": 0.6114208102226257, "learning_rate": 9.546839945163257e-05, "loss": 0.0356, "step": 14380 }, { "grad_norm": 0.35662585496902466, "learning_rate": 9.545979767668953e-05, "loss": 0.0427, "step": 14390 }, { "grad_norm": 0.43905484676361084, "learning_rate": 9.54511881339104e-05, "loss": 0.0345, "step": 14400 }, { "grad_norm": 8.329068183898926, "learning_rate": 9.54425708247663e-05, "loss": 0.0387, "step": 14410 }, { "grad_norm": 0.27843883633613586, "learning_rate": 9.543394575072972e-05, "loss": 0.0307, "step": 14420 }, { "grad_norm": 0.674006998538971, "learning_rate": 9.542531291327441e-05, "loss": 0.047, "step": 14430 }, { "grad_norm": 0.5124675035476685, "learning_rate": 9.541667231387552e-05, "loss": 0.0373, "step": 14440 }, { "grad_norm": 0.32065442204475403, "learning_rate": 9.540802395400949e-05, "loss": 0.0416, "step": 14450 }, { "grad_norm": 0.3553861677646637, "learning_rate": 9.539936783515406e-05, "loss": 0.0388, "step": 14460 }, { "grad_norm": 0.2983005940914154, "learning_rate": 9.539070395878835e-05, "loss": 0.0396, "step": 14470 }, { "grad_norm": 0.36424434185028076, "learning_rate": 9.538203232639277e-05, "loss": 0.0313, "step": 14480 }, { "grad_norm": 0.3172913193702698, "learning_rate": 9.537335293944907e-05, "loss": 0.0386, "step": 14490 }, { "grad_norm": 0.3256101906299591, "learning_rate": 9.536466579944032e-05, "loss": 0.0464, "step": 14500 }, { "grad_norm": 0.278324156999588, "learning_rate": 9.535597090785091e-05, "loss": 0.0345, "step": 14510 }, { "grad_norm": 0.2882360816001892, "learning_rate": 9.534726826616656e-05, "loss": 0.0323, "step": 14520 }, { "grad_norm": 0.3566344678401947, "learning_rate": 9.53385578758743e-05, "loss": 0.0515, "step": 14530 }, { "grad_norm": 0.7464251518249512, "learning_rate": 9.532983973846252e-05, "loss": 0.0383, "step": 14540 }, { "grad_norm": 0.576220691204071, "learning_rate": 9.53211138554209e-05, "loss": 0.0396, "step": 14550 }, { "grad_norm": 0.36228805780410767, "learning_rate": 9.531238022824047e-05, "loss": 0.0434, "step": 14560 }, { "grad_norm": 0.38600999116897583, "learning_rate": 9.530363885841355e-05, "loss": 0.0395, "step": 14570 }, { "grad_norm": 0.5281086564064026, "learning_rate": 9.52948897474338e-05, "loss": 0.0509, "step": 14580 }, { "grad_norm": 1.066827416419983, "learning_rate": 9.528613289679622e-05, "loss": 0.053, "step": 14590 }, { "grad_norm": 0.3051888942718506, "learning_rate": 9.52773683079971e-05, "loss": 0.0422, "step": 14600 }, { "grad_norm": 0.4699893891811371, "learning_rate": 9.526859598253407e-05, "loss": 0.0359, "step": 14610 }, { "grad_norm": 0.3004939556121826, "learning_rate": 9.525981592190609e-05, "loss": 0.0401, "step": 14620 }, { "grad_norm": 0.6937693357467651, "learning_rate": 9.525102812761342e-05, "loss": 0.0428, "step": 14630 }, { "grad_norm": 0.34335073828697205, "learning_rate": 9.524223260115768e-05, "loss": 0.0374, "step": 14640 }, { "grad_norm": 0.3214642107486725, "learning_rate": 9.523342934404175e-05, "loss": 0.0374, "step": 14650 }, { "grad_norm": 0.900672435760498, "learning_rate": 9.522461835776989e-05, "loss": 0.0364, "step": 14660 }, { "grad_norm": 0.3604736328125, "learning_rate": 9.521579964384764e-05, "loss": 0.0391, "step": 14670 }, { "grad_norm": 0.3762270212173462, "learning_rate": 9.52069732037819e-05, "loss": 0.0338, "step": 14680 }, { "grad_norm": 0.41613566875457764, "learning_rate": 9.519813903908083e-05, "loss": 0.0409, "step": 14690 }, { "grad_norm": 0.3111479580402374, "learning_rate": 9.5189297151254e-05, "loss": 0.0389, "step": 14700 }, { "grad_norm": 0.3889046311378479, "learning_rate": 9.518044754181218e-05, "loss": 0.052, "step": 14710 }, { "grad_norm": 0.4096849858760834, "learning_rate": 9.51715902122676e-05, "loss": 0.0428, "step": 14720 }, { "grad_norm": 0.28269582986831665, "learning_rate": 9.516272516413368e-05, "loss": 0.0423, "step": 14730 }, { "grad_norm": 0.4571014642715454, "learning_rate": 9.515385239892525e-05, "loss": 0.0408, "step": 14740 }, { "grad_norm": 0.4076552391052246, "learning_rate": 9.514497191815839e-05, "loss": 0.0426, "step": 14750 }, { "grad_norm": 0.34403878450393677, "learning_rate": 9.513608372335055e-05, "loss": 0.0377, "step": 14760 }, { "grad_norm": 0.3811687231063843, "learning_rate": 9.512718781602045e-05, "loss": 0.0341, "step": 14770 }, { "grad_norm": 0.4837053120136261, "learning_rate": 9.511828419768823e-05, "loss": 0.0419, "step": 14780 }, { "grad_norm": 0.32494163513183594, "learning_rate": 9.510937286987521e-05, "loss": 0.033, "step": 14790 }, { "grad_norm": 0.3896945118904114, "learning_rate": 9.510045383410408e-05, "loss": 0.0417, "step": 14800 }, { "grad_norm": 0.31837549805641174, "learning_rate": 9.509152709189892e-05, "loss": 0.0324, "step": 14810 }, { "grad_norm": 0.3991513252258301, "learning_rate": 9.508259264478504e-05, "loss": 0.0433, "step": 14820 }, { "grad_norm": 0.2981654405593872, "learning_rate": 9.507365049428909e-05, "loss": 0.0428, "step": 14830 }, { "grad_norm": 0.29031720757484436, "learning_rate": 9.506470064193902e-05, "loss": 0.0517, "step": 14840 }, { "grad_norm": 0.3315223753452301, "learning_rate": 9.505574308926414e-05, "loss": 0.0498, "step": 14850 }, { "grad_norm": 0.42590126395225525, "learning_rate": 9.504677783779505e-05, "loss": 0.0427, "step": 14860 }, { "grad_norm": 0.3181162178516388, "learning_rate": 9.503780488906365e-05, "loss": 0.0399, "step": 14870 }, { "grad_norm": 0.4346207082271576, "learning_rate": 9.502882424460319e-05, "loss": 0.0393, "step": 14880 }, { "grad_norm": 0.4093276858329773, "learning_rate": 9.501983590594821e-05, "loss": 0.039, "step": 14890 }, { "grad_norm": 0.32245492935180664, "learning_rate": 9.501083987463455e-05, "loss": 0.032, "step": 14900 }, { "grad_norm": 0.5654948949813843, "learning_rate": 9.500183615219942e-05, "loss": 0.0398, "step": 14910 }, { "grad_norm": 0.3287525177001953, "learning_rate": 9.49928247401813e-05, "loss": 0.034, "step": 14920 }, { "grad_norm": 0.33790576457977295, "learning_rate": 9.498380564011997e-05, "loss": 0.0326, "step": 14930 }, { "grad_norm": 0.2899699807167053, "learning_rate": 9.497477885355656e-05, "loss": 0.0427, "step": 14940 }, { "grad_norm": 0.3409923017024994, "learning_rate": 9.496574438203353e-05, "loss": 0.0489, "step": 14950 }, { "grad_norm": 0.28351056575775146, "learning_rate": 9.495670222709459e-05, "loss": 0.0422, "step": 14960 }, { "grad_norm": 0.3973595201969147, "learning_rate": 9.494765239028483e-05, "loss": 0.0425, "step": 14970 }, { "grad_norm": 0.3580668568611145, "learning_rate": 9.493859487315057e-05, "loss": 0.0391, "step": 14980 }, { "grad_norm": 0.2800907790660858, "learning_rate": 9.492952967723953e-05, "loss": 0.0411, "step": 14990 }, { "grad_norm": 0.4419916272163391, "learning_rate": 9.492045680410068e-05, "loss": 0.0407, "step": 15000 }, { "grad_norm": 0.5761343240737915, "learning_rate": 9.491137625528436e-05, "loss": 0.0357, "step": 15010 }, { "grad_norm": 0.42477792501449585, "learning_rate": 9.490228803234215e-05, "loss": 0.0343, "step": 15020 }, { "grad_norm": 0.40767765045166016, "learning_rate": 9.489319213682701e-05, "loss": 0.0347, "step": 15030 }, { "grad_norm": 0.3701358139514923, "learning_rate": 9.488408857029316e-05, "loss": 0.0498, "step": 15040 }, { "grad_norm": 0.28548145294189453, "learning_rate": 9.487497733429616e-05, "loss": 0.0491, "step": 15050 }, { "grad_norm": 0.3768601417541504, "learning_rate": 9.486585843039286e-05, "loss": 0.0416, "step": 15060 }, { "grad_norm": 0.2682602107524872, "learning_rate": 9.485673186014143e-05, "loss": 0.034, "step": 15070 }, { "grad_norm": 0.41226717829704285, "learning_rate": 9.484759762510137e-05, "loss": 0.0387, "step": 15080 }, { "grad_norm": 0.3663342595100403, "learning_rate": 9.483845572683346e-05, "loss": 0.0432, "step": 15090 }, { "grad_norm": 0.45018452405929565, "learning_rate": 9.48293061668998e-05, "loss": 0.0398, "step": 15100 }, { "grad_norm": 0.3164404034614563, "learning_rate": 9.48201489468638e-05, "loss": 0.0361, "step": 15110 }, { "grad_norm": 0.38951027393341064, "learning_rate": 9.481098406829016e-05, "loss": 0.0297, "step": 15120 }, { "grad_norm": 0.31280916929244995, "learning_rate": 9.480181153274495e-05, "loss": 0.0315, "step": 15130 }, { "grad_norm": 0.3265438675880432, "learning_rate": 9.479263134179548e-05, "loss": 0.0278, "step": 15140 }, { "grad_norm": 0.3502233326435089, "learning_rate": 9.478344349701039e-05, "loss": 0.0357, "step": 15150 }, { "grad_norm": 0.24475540220737457, "learning_rate": 9.477424799995964e-05, "loss": 0.0348, "step": 15160 }, { "grad_norm": 0.26582571864128113, "learning_rate": 9.476504485221448e-05, "loss": 0.0413, "step": 15170 }, { "grad_norm": 0.30273863673210144, "learning_rate": 9.475583405534748e-05, "loss": 0.0337, "step": 15180 }, { "grad_norm": 0.4671550393104553, "learning_rate": 9.474661561093251e-05, "loss": 0.0333, "step": 15190 }, { "grad_norm": 0.34601765871047974, "learning_rate": 9.473738952054478e-05, "loss": 0.0458, "step": 15200 }, { "grad_norm": 0.3347501754760742, "learning_rate": 9.472815578576073e-05, "loss": 0.0366, "step": 15210 }, { "grad_norm": 0.3799239695072174, "learning_rate": 9.471891440815817e-05, "loss": 0.0338, "step": 15220 }, { "grad_norm": 0.34092196822166443, "learning_rate": 9.470966538931621e-05, "loss": 0.027, "step": 15230 }, { "grad_norm": 0.32307344675064087, "learning_rate": 9.470040873081525e-05, "loss": 0.0302, "step": 15240 }, { "grad_norm": 0.4559812843799591, "learning_rate": 9.469114443423698e-05, "loss": 0.0455, "step": 15250 }, { "grad_norm": 0.28711503744125366, "learning_rate": 9.468187250116445e-05, "loss": 0.0358, "step": 15260 }, { "grad_norm": 0.40016165375709534, "learning_rate": 9.467259293318197e-05, "loss": 0.0307, "step": 15270 }, { "grad_norm": 0.36422768235206604, "learning_rate": 9.466330573187514e-05, "loss": 0.0358, "step": 15280 }, { "grad_norm": 0.3362134099006653, "learning_rate": 9.46540108988309e-05, "loss": 0.0439, "step": 15290 }, { "grad_norm": 0.3644087314605713, "learning_rate": 9.46447084356375e-05, "loss": 0.038, "step": 15300 }, { "grad_norm": 0.2893296182155609, "learning_rate": 9.463539834388447e-05, "loss": 0.0437, "step": 15310 }, { "grad_norm": 0.2887659966945648, "learning_rate": 9.462608062516263e-05, "loss": 0.0295, "step": 15320 }, { "grad_norm": 0.2724611759185791, "learning_rate": 9.461675528106413e-05, "loss": 0.0292, "step": 15330 }, { "grad_norm": 0.3539057672023773, "learning_rate": 9.460742231318244e-05, "loss": 0.0327, "step": 15340 }, { "grad_norm": 0.3630081117153168, "learning_rate": 9.459808172311229e-05, "loss": 0.0407, "step": 15350 }, { "grad_norm": 0.43593013286590576, "learning_rate": 9.458873351244972e-05, "loss": 0.037, "step": 15360 }, { "grad_norm": 0.300568163394928, "learning_rate": 9.457937768279211e-05, "loss": 0.0389, "step": 15370 }, { "grad_norm": 0.3558381497859955, "learning_rate": 9.45700142357381e-05, "loss": 0.0296, "step": 15380 }, { "grad_norm": 0.4507303833961487, "learning_rate": 9.456064317288765e-05, "loss": 0.0341, "step": 15390 }, { "grad_norm": 0.36109888553619385, "learning_rate": 9.455126449584201e-05, "loss": 0.0414, "step": 15400 }, { "grad_norm": 0.27597227692604065, "learning_rate": 9.454187820620375e-05, "loss": 0.0316, "step": 15410 }, { "grad_norm": 0.469900906085968, "learning_rate": 9.453248430557673e-05, "loss": 0.039, "step": 15420 }, { "grad_norm": 0.3282761871814728, "learning_rate": 9.452308279556611e-05, "loss": 0.0508, "step": 15430 }, { "grad_norm": 0.3214319348335266, "learning_rate": 9.451367367777835e-05, "loss": 0.0324, "step": 15440 }, { "grad_norm": 0.27640295028686523, "learning_rate": 9.450425695382122e-05, "loss": 0.038, "step": 15450 }, { "grad_norm": 0.32608193159103394, "learning_rate": 9.449483262530375e-05, "loss": 0.03, "step": 15460 }, { "grad_norm": 0.305081844329834, "learning_rate": 9.448540069383633e-05, "loss": 0.03, "step": 15470 }, { "grad_norm": 0.3841964304447174, "learning_rate": 9.447596116103061e-05, "loss": 0.0329, "step": 15480 }, { "grad_norm": 0.30268189311027527, "learning_rate": 9.446651402849955e-05, "loss": 0.0389, "step": 15490 }, { "grad_norm": 0.3098703622817993, "learning_rate": 9.44570592978574e-05, "loss": 0.0335, "step": 15500 }, { "grad_norm": 0.37366247177124023, "learning_rate": 9.444759697071972e-05, "loss": 0.0381, "step": 15510 }, { "grad_norm": 0.3121759593486786, "learning_rate": 9.443812704870336e-05, "loss": 0.0316, "step": 15520 }, { "grad_norm": 0.3037852644920349, "learning_rate": 9.442864953342649e-05, "loss": 0.0453, "step": 15530 }, { "grad_norm": 0.7628637552261353, "learning_rate": 9.441916442650852e-05, "loss": 0.0391, "step": 15540 }, { "grad_norm": 0.26952019333839417, "learning_rate": 9.440967172957023e-05, "loss": 0.03, "step": 15550 }, { "grad_norm": 0.33807751536369324, "learning_rate": 9.440017144423364e-05, "loss": 0.0355, "step": 15560 }, { "grad_norm": 0.28471672534942627, "learning_rate": 9.439066357212209e-05, "loss": 0.038, "step": 15570 }, { "grad_norm": 0.2605871558189392, "learning_rate": 9.438114811486022e-05, "loss": 0.0412, "step": 15580 }, { "grad_norm": 0.2723153829574585, "learning_rate": 9.4371625074074e-05, "loss": 0.0342, "step": 15590 }, { "grad_norm": 0.36642125248908997, "learning_rate": 9.436209445139059e-05, "loss": 0.0311, "step": 15600 }, { "grad_norm": 1.5404753684997559, "learning_rate": 9.435255624843855e-05, "loss": 0.0352, "step": 15610 }, { "grad_norm": 0.3131641447544098, "learning_rate": 9.43430104668477e-05, "loss": 0.0297, "step": 15620 }, { "grad_norm": 0.41007715463638306, "learning_rate": 9.433345710824914e-05, "loss": 0.0349, "step": 15630 }, { "grad_norm": 0.32650047540664673, "learning_rate": 9.432389617427529e-05, "loss": 0.0357, "step": 15640 }, { "grad_norm": 0.24473561346530914, "learning_rate": 9.431432766655984e-05, "loss": 0.036, "step": 15650 }, { "grad_norm": 0.39699432253837585, "learning_rate": 9.430475158673778e-05, "loss": 0.0322, "step": 15660 }, { "grad_norm": 0.3112611174583435, "learning_rate": 9.429516793644542e-05, "loss": 0.0357, "step": 15670 }, { "grad_norm": 0.3199608027935028, "learning_rate": 9.428557671732034e-05, "loss": 0.0304, "step": 15680 }, { "grad_norm": 0.3079599142074585, "learning_rate": 9.42759779310014e-05, "loss": 0.0355, "step": 15690 }, { "grad_norm": 0.4427284896373749, "learning_rate": 9.426637157912879e-05, "loss": 0.0309, "step": 15700 }, { "grad_norm": 0.3130052983760834, "learning_rate": 9.425675766334397e-05, "loss": 0.0383, "step": 15710 }, { "grad_norm": 0.4056973159313202, "learning_rate": 9.424713618528968e-05, "loss": 0.0313, "step": 15720 }, { "grad_norm": 0.40866976976394653, "learning_rate": 9.423750714661e-05, "loss": 0.0386, "step": 15730 }, { "grad_norm": 0.3681586682796478, "learning_rate": 9.422787054895022e-05, "loss": 0.0303, "step": 15740 }, { "grad_norm": 0.3735371232032776, "learning_rate": 9.4218226393957e-05, "loss": 0.0421, "step": 15750 }, { "grad_norm": 0.2632262408733368, "learning_rate": 9.420857468327828e-05, "loss": 0.0403, "step": 15760 }, { "grad_norm": 0.3181803822517395, "learning_rate": 9.419891541856323e-05, "loss": 0.0403, "step": 15770 }, { "grad_norm": 0.31072086095809937, "learning_rate": 9.41892486014624e-05, "loss": 0.0378, "step": 15780 }, { "grad_norm": 0.30315059423446655, "learning_rate": 9.417957423362756e-05, "loss": 0.0432, "step": 15790 }, { "grad_norm": 0.36339491605758667, "learning_rate": 9.416989231671178e-05, "loss": 0.0385, "step": 15800 }, { "grad_norm": 0.32493162155151367, "learning_rate": 9.416020285236946e-05, "loss": 0.033, "step": 15810 }, { "grad_norm": 0.29466038942337036, "learning_rate": 9.415050584225626e-05, "loss": 0.0348, "step": 15820 }, { "grad_norm": 0.3442705571651459, "learning_rate": 9.414080128802914e-05, "loss": 0.0339, "step": 15830 }, { "grad_norm": 0.347953200340271, "learning_rate": 9.413108919134632e-05, "loss": 0.0303, "step": 15840 }, { "grad_norm": 0.33017498254776, "learning_rate": 9.412136955386734e-05, "loss": 0.0303, "step": 15850 }, { "grad_norm": 0.24049116671085358, "learning_rate": 9.411164237725303e-05, "loss": 0.0323, "step": 15860 }, { "grad_norm": 0.3936634957790375, "learning_rate": 9.41019076631655e-05, "loss": 0.0345, "step": 15870 }, { "grad_norm": 3.8821909427642822, "learning_rate": 9.409216541326815e-05, "loss": 0.0397, "step": 15880 }, { "grad_norm": 0.29726284742355347, "learning_rate": 9.408241562922564e-05, "loss": 0.0388, "step": 15890 }, { "grad_norm": 0.31673675775527954, "learning_rate": 9.407265831270395e-05, "loss": 0.031, "step": 15900 }, { "grad_norm": 0.3576529622077942, "learning_rate": 9.406289346537035e-05, "loss": 0.0311, "step": 15910 }, { "grad_norm": 0.30313587188720703, "learning_rate": 9.405312108889339e-05, "loss": 0.0433, "step": 15920 }, { "grad_norm": 0.37682369351387024, "learning_rate": 9.404334118494288e-05, "loss": 0.0444, "step": 15930 }, { "grad_norm": 0.48763394355773926, "learning_rate": 9.403355375518995e-05, "loss": 0.0408, "step": 15940 }, { "grad_norm": 0.3997094929218292, "learning_rate": 9.4023758801307e-05, "loss": 0.0286, "step": 15950 }, { "grad_norm": 0.333880752325058, "learning_rate": 9.401395632496774e-05, "loss": 0.037, "step": 15960 }, { "grad_norm": 0.33406931161880493, "learning_rate": 9.400414632784711e-05, "loss": 0.0425, "step": 15970 }, { "grad_norm": 0.36078718304634094, "learning_rate": 9.39943288116214e-05, "loss": 0.0371, "step": 15980 }, { "grad_norm": 0.38763874769210815, "learning_rate": 9.398450377796815e-05, "loss": 0.0431, "step": 15990 }, { "grad_norm": 0.41017165780067444, "learning_rate": 9.397467122856616e-05, "loss": 0.041, "step": 16000 }, { "grad_norm": 0.2867724895477295, "learning_rate": 9.396483116509558e-05, "loss": 0.0378, "step": 16010 }, { "grad_norm": 0.5043345093727112, "learning_rate": 9.39549835892378e-05, "loss": 0.0453, "step": 16020 }, { "grad_norm": 0.3842546343803406, "learning_rate": 9.39451285026755e-05, "loss": 0.0387, "step": 16030 }, { "grad_norm": 0.4961106777191162, "learning_rate": 9.393526590709262e-05, "loss": 0.038, "step": 16040 }, { "grad_norm": 0.4309074580669403, "learning_rate": 9.392539580417444e-05, "loss": 0.0359, "step": 16050 }, { "grad_norm": 0.34178122878074646, "learning_rate": 9.391551819560747e-05, "loss": 0.0401, "step": 16060 }, { "grad_norm": 0.3037175238132477, "learning_rate": 9.390563308307955e-05, "loss": 0.0345, "step": 16070 }, { "grad_norm": 0.4529064893722534, "learning_rate": 9.389574046827974e-05, "loss": 0.0368, "step": 16080 }, { "grad_norm": 0.35779011249542236, "learning_rate": 9.388584035289845e-05, "loss": 0.0372, "step": 16090 }, { "grad_norm": 0.35625162720680237, "learning_rate": 9.387593273862732e-05, "loss": 0.0384, "step": 16100 }, { "grad_norm": 0.3098684549331665, "learning_rate": 9.386601762715929e-05, "loss": 0.0403, "step": 16110 }, { "grad_norm": 0.3816545605659485, "learning_rate": 9.38560950201886e-05, "loss": 0.0436, "step": 16120 }, { "grad_norm": 0.34001022577285767, "learning_rate": 9.384616491941071e-05, "loss": 0.0383, "step": 16130 }, { "grad_norm": 0.33576762676239014, "learning_rate": 9.383622732652245e-05, "loss": 0.0357, "step": 16140 }, { "grad_norm": 0.3870428502559662, "learning_rate": 9.382628224322187e-05, "loss": 0.0384, "step": 16150 }, { "grad_norm": 0.30635640025138855, "learning_rate": 9.381632967120829e-05, "loss": 0.0361, "step": 16160 }, { "grad_norm": 0.3529347777366638, "learning_rate": 9.380636961218235e-05, "loss": 0.0341, "step": 16170 }, { "grad_norm": 0.3675538897514343, "learning_rate": 9.379640206784597e-05, "loss": 0.0374, "step": 16180 }, { "grad_norm": 0.3399518132209778, "learning_rate": 9.378642703990229e-05, "loss": 0.0339, "step": 16190 }, { "grad_norm": 0.29465165734291077, "learning_rate": 9.37764445300558e-05, "loss": 0.0359, "step": 16200 }, { "grad_norm": 0.26920199394226074, "learning_rate": 9.376645454001222e-05, "loss": 0.0407, "step": 16210 }, { "grad_norm": 0.30196669697761536, "learning_rate": 9.375645707147858e-05, "loss": 0.0332, "step": 16220 }, { "grad_norm": 0.42603036761283875, "learning_rate": 9.374645212616316e-05, "loss": 0.0493, "step": 16230 }, { "grad_norm": 0.6396195888519287, "learning_rate": 9.373643970577555e-05, "loss": 0.0485, "step": 16240 }, { "grad_norm": 0.3673569858074188, "learning_rate": 9.372641981202659e-05, "loss": 0.0412, "step": 16250 }, { "grad_norm": 0.445858895778656, "learning_rate": 9.37163924466284e-05, "loss": 0.0382, "step": 16260 }, { "grad_norm": 0.32017502188682556, "learning_rate": 9.370635761129438e-05, "loss": 0.0338, "step": 16270 }, { "grad_norm": 0.35425958037376404, "learning_rate": 9.36963153077392e-05, "loss": 0.0425, "step": 16280 }, { "grad_norm": 0.3525136709213257, "learning_rate": 9.368626553767888e-05, "loss": 0.0386, "step": 16290 }, { "grad_norm": 0.35423043370246887, "learning_rate": 9.367620830283057e-05, "loss": 0.0306, "step": 16300 }, { "grad_norm": 0.44054099917411804, "learning_rate": 9.366614360491281e-05, "loss": 0.042, "step": 16310 }, { "grad_norm": 0.34702950716018677, "learning_rate": 9.365607144564539e-05, "loss": 0.0364, "step": 16320 }, { "grad_norm": 0.3395412266254425, "learning_rate": 9.364599182674934e-05, "loss": 0.0441, "step": 16330 }, { "grad_norm": 0.5660090446472168, "learning_rate": 9.3635904749947e-05, "loss": 0.0438, "step": 16340 }, { "grad_norm": 0.26794686913490295, "learning_rate": 9.362581021696202e-05, "loss": 0.0384, "step": 16350 }, { "grad_norm": 0.2672222852706909, "learning_rate": 9.361570822951921e-05, "loss": 0.0434, "step": 16360 }, { "grad_norm": 0.3836769461631775, "learning_rate": 9.360559878934476e-05, "loss": 0.0484, "step": 16370 }, { "grad_norm": 0.36893418431282043, "learning_rate": 9.359548189816611e-05, "loss": 0.038, "step": 16380 }, { "grad_norm": 0.26762309670448303, "learning_rate": 9.358535755771193e-05, "loss": 0.0379, "step": 16390 }, { "grad_norm": 0.3787645399570465, "learning_rate": 9.357522576971221e-05, "loss": 0.0381, "step": 16400 }, { "grad_norm": 0.3965562880039215, "learning_rate": 9.356508653589819e-05, "loss": 0.0352, "step": 16410 }, { "grad_norm": 0.3970875144004822, "learning_rate": 9.355493985800237e-05, "loss": 0.0359, "step": 16420 }, { "grad_norm": 0.33413586020469666, "learning_rate": 9.354478573775857e-05, "loss": 0.0377, "step": 16430 }, { "grad_norm": 0.3900012671947479, "learning_rate": 9.353462417690186e-05, "loss": 0.035, "step": 16440 }, { "grad_norm": 0.3253329396247864, "learning_rate": 9.352445517716853e-05, "loss": 0.0303, "step": 16450 }, { "grad_norm": 0.35474923253059387, "learning_rate": 9.351427874029621e-05, "loss": 0.0378, "step": 16460 }, { "grad_norm": 0.391208291053772, "learning_rate": 9.350409486802379e-05, "loss": 0.0343, "step": 16470 }, { "grad_norm": 0.40508168935775757, "learning_rate": 9.349390356209138e-05, "loss": 0.0394, "step": 16480 }, { "grad_norm": 0.5439741015434265, "learning_rate": 9.348370482424042e-05, "loss": 0.0335, "step": 16490 }, { "grad_norm": 0.3304598331451416, "learning_rate": 9.347349865621357e-05, "loss": 0.0283, "step": 16500 }, { "grad_norm": 0.3181858956813812, "learning_rate": 9.346328505975481e-05, "loss": 0.0344, "step": 16510 }, { "grad_norm": 0.36800622940063477, "learning_rate": 9.345306403660936e-05, "loss": 0.0388, "step": 16520 }, { "grad_norm": 0.3710225820541382, "learning_rate": 9.344283558852371e-05, "loss": 0.0415, "step": 16530 }, { "grad_norm": 0.3223992586135864, "learning_rate": 9.343259971724563e-05, "loss": 0.0389, "step": 16540 }, { "grad_norm": 0.32188504934310913, "learning_rate": 9.342235642452413e-05, "loss": 0.0384, "step": 16550 }, { "grad_norm": 0.34014129638671875, "learning_rate": 9.341210571210954e-05, "loss": 0.0398, "step": 16560 }, { "grad_norm": 0.24908088147640228, "learning_rate": 9.340184758175338e-05, "loss": 0.0345, "step": 16570 }, { "grad_norm": 0.4046228229999542, "learning_rate": 9.339158203520854e-05, "loss": 0.0472, "step": 16580 }, { "grad_norm": 0.6750109195709229, "learning_rate": 9.338130907422908e-05, "loss": 0.0405, "step": 16590 }, { "grad_norm": 0.3107208013534546, "learning_rate": 9.337102870057037e-05, "loss": 0.0334, "step": 16600 }, { "grad_norm": 0.290387898683548, "learning_rate": 9.336074091598907e-05, "loss": 0.03, "step": 16610 }, { "grad_norm": 0.5485457181930542, "learning_rate": 9.335044572224306e-05, "loss": 0.0325, "step": 16620 }, { "grad_norm": 0.4032699763774872, "learning_rate": 9.334014312109151e-05, "loss": 0.036, "step": 16630 }, { "grad_norm": 0.42997264862060547, "learning_rate": 9.332983311429486e-05, "loss": 0.0395, "step": 16640 }, { "grad_norm": 0.3170957565307617, "learning_rate": 9.33195157036148e-05, "loss": 0.0376, "step": 16650 }, { "grad_norm": 0.2573501467704773, "learning_rate": 9.330919089081432e-05, "loss": 0.0338, "step": 16660 }, { "grad_norm": 0.2964933216571808, "learning_rate": 9.32988586776576e-05, "loss": 0.0485, "step": 16670 }, { "grad_norm": 0.44780102372169495, "learning_rate": 9.328851906591016e-05, "loss": 0.0321, "step": 16680 }, { "grad_norm": 0.41124847531318665, "learning_rate": 9.327817205733875e-05, "loss": 0.0329, "step": 16690 }, { "grad_norm": 0.40143832564353943, "learning_rate": 9.326781765371142e-05, "loss": 0.0305, "step": 16700 }, { "grad_norm": 0.270202100276947, "learning_rate": 9.325745585679741e-05, "loss": 0.0343, "step": 16710 }, { "grad_norm": 0.488158255815506, "learning_rate": 9.32470866683673e-05, "loss": 0.0505, "step": 16720 }, { "grad_norm": 0.3176603317260742, "learning_rate": 9.323671009019288e-05, "loss": 0.0451, "step": 16730 }, { "grad_norm": 0.44910573959350586, "learning_rate": 9.322632612404725e-05, "loss": 0.0339, "step": 16740 }, { "grad_norm": 0.3579404056072235, "learning_rate": 9.321593477170471e-05, "loss": 0.0444, "step": 16750 }, { "grad_norm": 0.3167540431022644, "learning_rate": 9.320553603494088e-05, "loss": 0.0339, "step": 16760 }, { "grad_norm": 0.5453197956085205, "learning_rate": 9.319512991553261e-05, "loss": 0.0346, "step": 16770 }, { "grad_norm": 0.34944167733192444, "learning_rate": 9.318471641525803e-05, "loss": 0.0367, "step": 16780 }, { "grad_norm": 0.21991825103759766, "learning_rate": 9.317429553589652e-05, "loss": 0.027, "step": 16790 }, { "grad_norm": 0.3427276611328125, "learning_rate": 9.316386727922873e-05, "loss": 0.036, "step": 16800 }, { "grad_norm": 0.34461015462875366, "learning_rate": 9.315343164703656e-05, "loss": 0.0301, "step": 16810 }, { "grad_norm": 0.33611905574798584, "learning_rate": 9.314298864110316e-05, "loss": 0.0297, "step": 16820 }, { "grad_norm": 0.24759025871753693, "learning_rate": 9.313253826321295e-05, "loss": 0.0297, "step": 16830 }, { "grad_norm": 0.48538631200790405, "learning_rate": 9.312208051515165e-05, "loss": 0.0299, "step": 16840 }, { "grad_norm": 0.4001196026802063, "learning_rate": 9.311161539870618e-05, "loss": 0.0285, "step": 16850 }, { "grad_norm": 0.24790066480636597, "learning_rate": 9.310114291566474e-05, "loss": 0.0396, "step": 16860 }, { "grad_norm": 0.3464120924472809, "learning_rate": 9.309066306781679e-05, "loss": 0.0271, "step": 16870 }, { "grad_norm": 0.3069632947444916, "learning_rate": 9.308017585695306e-05, "loss": 0.0282, "step": 16880 }, { "grad_norm": 0.4178350269794464, "learning_rate": 9.306968128486552e-05, "loss": 0.0399, "step": 16890 }, { "grad_norm": 0.32318761944770813, "learning_rate": 9.30591793533474e-05, "loss": 0.0285, "step": 16900 }, { "grad_norm": 0.29484426975250244, "learning_rate": 9.304867006419321e-05, "loss": 0.0338, "step": 16910 }, { "grad_norm": 0.28487637639045715, "learning_rate": 9.303815341919868e-05, "loss": 0.0273, "step": 16920 }, { "grad_norm": 0.43624818325042725, "learning_rate": 9.302762942016084e-05, "loss": 0.0339, "step": 16930 }, { "grad_norm": 0.32178494334220886, "learning_rate": 9.301709806887792e-05, "loss": 0.0326, "step": 16940 }, { "grad_norm": 0.3143419027328491, "learning_rate": 9.300655936714948e-05, "loss": 0.0402, "step": 16950 }, { "grad_norm": 0.391064316034317, "learning_rate": 9.299601331677627e-05, "loss": 0.034, "step": 16960 }, { "grad_norm": 0.3862215280532837, "learning_rate": 9.298545991956033e-05, "loss": 0.0353, "step": 16970 }, { "grad_norm": 0.31961575150489807, "learning_rate": 9.297489917730493e-05, "loss": 0.03, "step": 16980 }, { "grad_norm": 0.30517762899398804, "learning_rate": 9.296433109181464e-05, "loss": 0.0384, "step": 16990 }, { "grad_norm": 0.38263803720474243, "learning_rate": 9.295375566489523e-05, "loss": 0.0392, "step": 17000 }, { "grad_norm": 0.41997238993644714, "learning_rate": 9.294317289835379e-05, "loss": 0.0421, "step": 17010 }, { "grad_norm": 0.32456883788108826, "learning_rate": 9.293258279399859e-05, "loss": 0.044, "step": 17020 }, { "grad_norm": 0.7412046194076538, "learning_rate": 9.292198535363919e-05, "loss": 0.0321, "step": 17030 }, { "grad_norm": 0.33550190925598145, "learning_rate": 9.291138057908641e-05, "loss": 0.0297, "step": 17040 }, { "grad_norm": 0.21493151783943176, "learning_rate": 9.290076847215234e-05, "loss": 0.0292, "step": 17050 }, { "grad_norm": 0.6358821392059326, "learning_rate": 9.289014903465025e-05, "loss": 0.0499, "step": 17060 }, { "grad_norm": 0.31412431597709656, "learning_rate": 9.287952226839475e-05, "loss": 0.036, "step": 17070 }, { "grad_norm": 0.5088747143745422, "learning_rate": 9.286888817520164e-05, "loss": 0.0386, "step": 17080 }, { "grad_norm": 0.3594599664211273, "learning_rate": 9.285824675688803e-05, "loss": 0.0404, "step": 17090 }, { "grad_norm": 0.36289751529693604, "learning_rate": 9.28475980152722e-05, "loss": 0.0354, "step": 17100 }, { "grad_norm": 0.2863699197769165, "learning_rate": 9.283694195217379e-05, "loss": 0.0397, "step": 17110 }, { "grad_norm": 0.5031334757804871, "learning_rate": 9.282627856941356e-05, "loss": 0.0397, "step": 17120 }, { "grad_norm": 0.3220252990722656, "learning_rate": 9.281560786881363e-05, "loss": 0.0335, "step": 17130 }, { "grad_norm": 0.4717540144920349, "learning_rate": 9.280492985219733e-05, "loss": 0.0313, "step": 17140 }, { "grad_norm": 0.35942399501800537, "learning_rate": 9.279424452138924e-05, "loss": 0.0367, "step": 17150 }, { "grad_norm": 0.2713391184806824, "learning_rate": 9.278355187821517e-05, "loss": 0.0342, "step": 17160 }, { "grad_norm": 0.39280667901039124, "learning_rate": 9.277285192450224e-05, "loss": 0.0398, "step": 17170 }, { "grad_norm": 0.3686666786670685, "learning_rate": 9.276214466207875e-05, "loss": 0.0368, "step": 17180 }, { "grad_norm": 0.30834874510765076, "learning_rate": 9.275143009277427e-05, "loss": 0.0355, "step": 17190 }, { "grad_norm": 0.32892367243766785, "learning_rate": 9.274070821841964e-05, "loss": 0.0318, "step": 17200 }, { "grad_norm": 0.49303776025772095, "learning_rate": 9.272997904084696e-05, "loss": 0.0336, "step": 17210 }, { "grad_norm": 0.4516763985157013, "learning_rate": 9.271924256188951e-05, "loss": 0.0318, "step": 17220 }, { "grad_norm": 0.3052560091018677, "learning_rate": 9.270849878338189e-05, "loss": 0.0294, "step": 17230 }, { "grad_norm": 0.33670708537101746, "learning_rate": 9.269774770715991e-05, "loss": 0.0366, "step": 17240 }, { "grad_norm": 0.2889864146709442, "learning_rate": 9.268698933506061e-05, "loss": 0.0485, "step": 17250 }, { "grad_norm": 0.2877305746078491, "learning_rate": 9.267622366892235e-05, "loss": 0.0315, "step": 17260 }, { "grad_norm": 0.37044912576675415, "learning_rate": 9.266545071058465e-05, "loss": 0.0371, "step": 17270 }, { "grad_norm": 0.29754117131233215, "learning_rate": 9.265467046188833e-05, "loss": 0.0309, "step": 17280 }, { "grad_norm": 0.3257599174976349, "learning_rate": 9.264388292467543e-05, "loss": 0.0372, "step": 17290 }, { "grad_norm": 0.33477094769477844, "learning_rate": 9.263308810078926e-05, "loss": 0.0325, "step": 17300 }, { "grad_norm": 0.34570565819740295, "learning_rate": 9.262228599207434e-05, "loss": 0.0298, "step": 17310 }, { "grad_norm": 0.31736260652542114, "learning_rate": 9.261147660037647e-05, "loss": 0.0308, "step": 17320 }, { "grad_norm": 0.3574795126914978, "learning_rate": 9.26006599275427e-05, "loss": 0.0314, "step": 17330 }, { "grad_norm": 0.38376301527023315, "learning_rate": 9.258983597542124e-05, "loss": 0.0402, "step": 17340 }, { "grad_norm": 0.4114547073841095, "learning_rate": 9.257900474586167e-05, "loss": 0.0341, "step": 17350 }, { "grad_norm": 0.39493733644485474, "learning_rate": 9.256816624071471e-05, "loss": 0.0403, "step": 17360 }, { "grad_norm": 0.32671841979026794, "learning_rate": 9.25573204618324e-05, "loss": 0.038, "step": 17370 }, { "grad_norm": 0.3131188452243805, "learning_rate": 9.254646741106796e-05, "loss": 0.0399, "step": 17380 }, { "grad_norm": 0.4931085705757141, "learning_rate": 9.253560709027589e-05, "loss": 0.0309, "step": 17390 }, { "grad_norm": 0.2974867820739746, "learning_rate": 9.252473950131192e-05, "loss": 0.0357, "step": 17400 }, { "grad_norm": 0.331929475069046, "learning_rate": 9.251386464603302e-05, "loss": 0.0239, "step": 17410 }, { "grad_norm": 0.28070884943008423, "learning_rate": 9.250298252629741e-05, "loss": 0.0295, "step": 17420 }, { "grad_norm": 0.2654360234737396, "learning_rate": 9.249209314396454e-05, "loss": 0.0317, "step": 17430 }, { "grad_norm": 0.44542667269706726, "learning_rate": 9.248119650089513e-05, "loss": 0.0337, "step": 17440 }, { "grad_norm": 0.27970367670059204, "learning_rate": 9.247029259895108e-05, "loss": 0.0339, "step": 17450 }, { "grad_norm": 0.296068012714386, "learning_rate": 9.24593814399956e-05, "loss": 0.029, "step": 17460 }, { "grad_norm": 0.395966112613678, "learning_rate": 9.244846302589309e-05, "loss": 0.0305, "step": 17470 }, { "grad_norm": 0.5470436811447144, "learning_rate": 9.243753735850923e-05, "loss": 0.0388, "step": 17480 }, { "grad_norm": 0.3923816680908203, "learning_rate": 9.24266044397109e-05, "loss": 0.033, "step": 17490 }, { "grad_norm": 0.32852911949157715, "learning_rate": 9.241566427136624e-05, "loss": 0.0358, "step": 17500 }, { "grad_norm": 0.3315228521823883, "learning_rate": 9.240471685534463e-05, "loss": 0.0368, "step": 17510 }, { "grad_norm": 0.4222482740879059, "learning_rate": 9.239376219351667e-05, "loss": 0.0402, "step": 17520 }, { "grad_norm": 0.27180784940719604, "learning_rate": 9.238280028775425e-05, "loss": 0.0304, "step": 17530 }, { "grad_norm": 0.25508809089660645, "learning_rate": 9.237183113993041e-05, "loss": 0.0393, "step": 17540 }, { "grad_norm": 0.2326328605413437, "learning_rate": 9.236085475191952e-05, "loss": 0.0313, "step": 17550 }, { "grad_norm": 0.9490134716033936, "learning_rate": 9.234987112559709e-05, "loss": 0.0494, "step": 17560 }, { "grad_norm": 0.38245436549186707, "learning_rate": 9.233888026283999e-05, "loss": 0.0376, "step": 17570 }, { "grad_norm": 0.3065773844718933, "learning_rate": 9.232788216552619e-05, "loss": 0.0377, "step": 17580 }, { "grad_norm": 0.3296293616294861, "learning_rate": 9.231687683553502e-05, "loss": 0.0397, "step": 17590 }, { "grad_norm": 0.33600252866744995, "learning_rate": 9.230586427474698e-05, "loss": 0.0439, "step": 17600 }, { "grad_norm": 0.2852442264556885, "learning_rate": 9.229484448504379e-05, "loss": 0.0394, "step": 17610 }, { "grad_norm": 0.3781937062740326, "learning_rate": 9.228381746830843e-05, "loss": 0.0337, "step": 17620 }, { "grad_norm": 0.4480392038822174, "learning_rate": 9.227278322642514e-05, "loss": 0.0318, "step": 17630 }, { "grad_norm": 0.3098250925540924, "learning_rate": 9.226174176127937e-05, "loss": 0.0405, "step": 17640 }, { "grad_norm": 0.3771049678325653, "learning_rate": 9.22506930747578e-05, "loss": 0.0358, "step": 17650 }, { "grad_norm": 0.5128540396690369, "learning_rate": 9.223963716874831e-05, "loss": 0.0422, "step": 17660 }, { "grad_norm": 0.32274964451789856, "learning_rate": 9.222857404514012e-05, "loss": 0.0336, "step": 17670 }, { "grad_norm": 0.32227304577827454, "learning_rate": 9.221750370582355e-05, "loss": 0.0323, "step": 17680 }, { "grad_norm": 2.3549821376800537, "learning_rate": 9.220642615269028e-05, "loss": 0.0313, "step": 17690 }, { "grad_norm": 0.7287243008613586, "learning_rate": 9.219534138763311e-05, "loss": 0.0408, "step": 17700 }, { "grad_norm": 0.4676916301250458, "learning_rate": 9.218424941254613e-05, "loss": 0.0419, "step": 17710 }, { "grad_norm": 0.28502434492111206, "learning_rate": 9.217315022932468e-05, "loss": 0.0386, "step": 17720 }, { "grad_norm": 0.3168308436870575, "learning_rate": 9.216204383986528e-05, "loss": 0.0298, "step": 17730 }, { "grad_norm": 0.3108782172203064, "learning_rate": 9.215093024606574e-05, "loss": 0.0408, "step": 17740 }, { "grad_norm": 0.38804563879966736, "learning_rate": 9.213980944982506e-05, "loss": 0.0486, "step": 17750 }, { "grad_norm": 0.44675779342651367, "learning_rate": 9.212868145304346e-05, "loss": 0.0378, "step": 17760 }, { "grad_norm": 0.37403637170791626, "learning_rate": 9.211754625762241e-05, "loss": 0.0321, "step": 17770 }, { "grad_norm": 0.29975777864456177, "learning_rate": 9.210640386546463e-05, "loss": 0.0432, "step": 17780 }, { "grad_norm": 0.8909753561019897, "learning_rate": 9.209525427847405e-05, "loss": 0.0308, "step": 17790 }, { "grad_norm": 0.8802274465560913, "learning_rate": 9.208409749855583e-05, "loss": 0.0304, "step": 17800 }, { "grad_norm": 0.3471032977104187, "learning_rate": 9.207293352761633e-05, "loss": 0.0427, "step": 17810 }, { "grad_norm": 0.34345611929893494, "learning_rate": 9.206176236756319e-05, "loss": 0.0333, "step": 17820 }, { "grad_norm": 0.3006901144981384, "learning_rate": 9.205058402030525e-05, "loss": 0.0353, "step": 17830 }, { "grad_norm": 0.28435200452804565, "learning_rate": 9.203939848775259e-05, "loss": 0.0315, "step": 17840 }, { "grad_norm": 0.41724738478660583, "learning_rate": 9.202820577181652e-05, "loss": 0.0368, "step": 17850 }, { "grad_norm": 0.32767611742019653, "learning_rate": 9.201700587440953e-05, "loss": 0.0299, "step": 17860 }, { "grad_norm": 0.4090961515903473, "learning_rate": 9.200579879744544e-05, "loss": 0.0328, "step": 17870 }, { "grad_norm": 0.44457241892814636, "learning_rate": 9.199458454283918e-05, "loss": 0.032, "step": 17880 }, { "grad_norm": 0.4177357852458954, "learning_rate": 9.198336311250697e-05, "loss": 0.0349, "step": 17890 }, { "grad_norm": 0.3207743167877197, "learning_rate": 9.197213450836626e-05, "loss": 0.0313, "step": 17900 }, { "grad_norm": 0.28111663460731506, "learning_rate": 9.19608987323357e-05, "loss": 0.0411, "step": 17910 }, { "grad_norm": 0.22331824898719788, "learning_rate": 9.194965578633517e-05, "loss": 0.0322, "step": 17920 }, { "grad_norm": 0.37346309423446655, "learning_rate": 9.193840567228582e-05, "loss": 0.0346, "step": 17930 }, { "grad_norm": 0.4110884368419647, "learning_rate": 9.192714839210994e-05, "loss": 0.0363, "step": 17940 }, { "grad_norm": 0.3629226088523865, "learning_rate": 9.19158839477311e-05, "loss": 0.0402, "step": 17950 }, { "grad_norm": 0.4224226176738739, "learning_rate": 9.190461234107411e-05, "loss": 0.0409, "step": 17960 }, { "grad_norm": 0.3871932625770569, "learning_rate": 9.189333357406496e-05, "loss": 0.0294, "step": 17970 }, { "grad_norm": 0.5194762945175171, "learning_rate": 9.188204764863089e-05, "loss": 0.0406, "step": 17980 }, { "grad_norm": 0.3382445275783539, "learning_rate": 9.187075456670033e-05, "loss": 0.0283, "step": 17990 }, { "grad_norm": 0.4067932069301605, "learning_rate": 9.1859454330203e-05, "loss": 0.0348, "step": 18000 }, { "grad_norm": 0.3155721127986908, "learning_rate": 9.18481469410698e-05, "loss": 0.045, "step": 18010 }, { "grad_norm": 0.3278665840625763, "learning_rate": 9.183683240123281e-05, "loss": 0.0396, "step": 18020 }, { "grad_norm": 0.36757683753967285, "learning_rate": 9.182551071262541e-05, "loss": 0.0411, "step": 18030 }, { "grad_norm": 0.26106318831443787, "learning_rate": 9.181418187718218e-05, "loss": 0.0363, "step": 18040 }, { "grad_norm": 0.4469119906425476, "learning_rate": 9.180284589683888e-05, "loss": 0.0364, "step": 18050 }, { "grad_norm": 0.3757087290287018, "learning_rate": 9.17915027735325e-05, "loss": 0.0336, "step": 18060 }, { "grad_norm": 0.2848169505596161, "learning_rate": 9.178015250920133e-05, "loss": 0.0279, "step": 18070 }, { "grad_norm": 0.34491223096847534, "learning_rate": 9.176879510578477e-05, "loss": 0.0312, "step": 18080 }, { "grad_norm": 0.327808678150177, "learning_rate": 9.17574305652235e-05, "loss": 0.0358, "step": 18090 }, { "grad_norm": 0.45504826307296753, "learning_rate": 9.174605888945942e-05, "loss": 0.043, "step": 18100 }, { "grad_norm": 0.3268153667449951, "learning_rate": 9.173468008043564e-05, "loss": 0.0322, "step": 18110 }, { "grad_norm": 0.34641921520233154, "learning_rate": 9.172329414009648e-05, "loss": 0.0368, "step": 18120 }, { "grad_norm": 1.2424702644348145, "learning_rate": 9.171190107038747e-05, "loss": 0.0395, "step": 18130 }, { "grad_norm": 0.6615407466888428, "learning_rate": 9.170050087325541e-05, "loss": 0.0308, "step": 18140 }, { "grad_norm": 0.37731680274009705, "learning_rate": 9.168909355064824e-05, "loss": 0.0312, "step": 18150 }, { "grad_norm": 2.6988232135772705, "learning_rate": 9.167767910451519e-05, "loss": 0.0325, "step": 18160 }, { "grad_norm": 0.291227251291275, "learning_rate": 9.166625753680669e-05, "loss": 0.03, "step": 18170 }, { "grad_norm": 0.33250850439071655, "learning_rate": 9.165482884947431e-05, "loss": 0.0269, "step": 18180 }, { "grad_norm": 0.28453272581100464, "learning_rate": 9.164339304447098e-05, "loss": 0.0378, "step": 18190 }, { "grad_norm": 0.31300610303878784, "learning_rate": 9.163195012375072e-05, "loss": 0.0289, "step": 18200 }, { "grad_norm": 0.39496132731437683, "learning_rate": 9.16205000892688e-05, "loss": 0.0301, "step": 18210 }, { "grad_norm": 0.2401861697435379, "learning_rate": 9.160904294298175e-05, "loss": 0.0339, "step": 18220 }, { "grad_norm": 0.36980774998664856, "learning_rate": 9.159757868684727e-05, "loss": 0.0384, "step": 18230 }, { "grad_norm": 0.3085935711860657, "learning_rate": 9.15861073228243e-05, "loss": 0.041, "step": 18240 }, { "grad_norm": 0.3406094014644623, "learning_rate": 9.157462885287296e-05, "loss": 0.0362, "step": 18250 }, { "grad_norm": 0.27310654520988464, "learning_rate": 9.156314327895461e-05, "loss": 0.0354, "step": 18260 }, { "grad_norm": 0.3747425377368927, "learning_rate": 9.155165060303185e-05, "loss": 0.0344, "step": 18270 }, { "grad_norm": 0.3070754110813141, "learning_rate": 9.154015082706841e-05, "loss": 0.0285, "step": 18280 }, { "grad_norm": 0.34063562750816345, "learning_rate": 9.152864395302936e-05, "loss": 0.0324, "step": 18290 }, { "grad_norm": 0.26828011870384216, "learning_rate": 9.151712998288085e-05, "loss": 0.0279, "step": 18300 }, { "grad_norm": 0.5084806084632874, "learning_rate": 9.150560891859031e-05, "loss": 0.0374, "step": 18310 }, { "grad_norm": 0.3022250533103943, "learning_rate": 9.14940807621264e-05, "loss": 0.037, "step": 18320 }, { "grad_norm": 0.33432725071907043, "learning_rate": 9.148254551545894e-05, "loss": 0.029, "step": 18330 }, { "grad_norm": 0.513393759727478, "learning_rate": 9.147100318055901e-05, "loss": 0.037, "step": 18340 }, { "grad_norm": 0.3335705101490021, "learning_rate": 9.145945375939888e-05, "loss": 0.0347, "step": 18350 }, { "grad_norm": 0.32874301075935364, "learning_rate": 9.144789725395203e-05, "loss": 0.0346, "step": 18360 }, { "grad_norm": 0.350462943315506, "learning_rate": 9.14363336661931e-05, "loss": 0.0333, "step": 18370 }, { "grad_norm": 0.3716275095939636, "learning_rate": 9.142476299809806e-05, "loss": 0.0324, "step": 18380 }, { "grad_norm": 0.4717976450920105, "learning_rate": 9.1413185251644e-05, "loss": 0.0305, "step": 18390 }, { "grad_norm": 0.2759037911891937, "learning_rate": 9.140160042880923e-05, "loss": 0.0333, "step": 18400 }, { "grad_norm": 0.3507821261882782, "learning_rate": 9.139000853157327e-05, "loss": 0.0368, "step": 18410 }, { "grad_norm": 0.4000323414802551, "learning_rate": 9.137840956191688e-05, "loss": 0.0374, "step": 18420 }, { "grad_norm": 0.29430726170539856, "learning_rate": 9.136680352182199e-05, "loss": 0.0292, "step": 18430 }, { "grad_norm": 0.30561572313308716, "learning_rate": 9.135519041327177e-05, "loss": 0.0266, "step": 18440 }, { "grad_norm": 0.6744679808616638, "learning_rate": 9.134357023825058e-05, "loss": 0.0376, "step": 18450 }, { "grad_norm": 0.3720588684082031, "learning_rate": 9.133194299874398e-05, "loss": 0.029, "step": 18460 }, { "grad_norm": 0.3910042941570282, "learning_rate": 9.132030869673876e-05, "loss": 0.0344, "step": 18470 }, { "grad_norm": 0.30177369713783264, "learning_rate": 9.130866733422288e-05, "loss": 0.0308, "step": 18480 }, { "grad_norm": 0.3178838789463043, "learning_rate": 9.129701891318556e-05, "loss": 0.0281, "step": 18490 }, { "grad_norm": 0.3390294015407562, "learning_rate": 9.128536343561718e-05, "loss": 0.045, "step": 18500 }, { "grad_norm": 0.2655843198299408, "learning_rate": 9.127370090350934e-05, "loss": 0.0331, "step": 18510 }, { "grad_norm": 0.38829687237739563, "learning_rate": 9.126203131885487e-05, "loss": 0.0311, "step": 18520 }, { "grad_norm": 0.4870069921016693, "learning_rate": 9.125035468364775e-05, "loss": 0.0342, "step": 18530 }, { "grad_norm": 0.5663111805915833, "learning_rate": 9.123867099988322e-05, "loss": 0.032, "step": 18540 }, { "grad_norm": 0.5415154695510864, "learning_rate": 9.122698026955769e-05, "loss": 0.0383, "step": 18550 }, { "grad_norm": 0.31895509362220764, "learning_rate": 9.12152824946688e-05, "loss": 0.0521, "step": 18560 }, { "grad_norm": 0.27863723039627075, "learning_rate": 9.120357767721538e-05, "loss": 0.0353, "step": 18570 }, { "grad_norm": 0.4014824330806732, "learning_rate": 9.119186581919745e-05, "loss": 0.0353, "step": 18580 }, { "grad_norm": 0.3657088875770569, "learning_rate": 9.118014692261624e-05, "loss": 0.0346, "step": 18590 }, { "grad_norm": 0.4254082143306732, "learning_rate": 9.116842098947422e-05, "loss": 0.0434, "step": 18600 }, { "grad_norm": 0.25665420293807983, "learning_rate": 9.115668802177499e-05, "loss": 0.037, "step": 18610 }, { "grad_norm": 0.3831409215927124, "learning_rate": 9.114494802152342e-05, "loss": 0.0508, "step": 18620 }, { "grad_norm": 0.4543253481388092, "learning_rate": 9.113320099072555e-05, "loss": 0.0568, "step": 18630 }, { "grad_norm": 0.3265348970890045, "learning_rate": 9.112144693138864e-05, "loss": 0.0402, "step": 18640 }, { "grad_norm": 0.3300667107105255, "learning_rate": 9.110968584552111e-05, "loss": 0.0324, "step": 18650 }, { "grad_norm": 0.4632779061794281, "learning_rate": 9.109791773513264e-05, "loss": 0.033, "step": 18660 }, { "grad_norm": 0.44692549109458923, "learning_rate": 9.108614260223403e-05, "loss": 0.035, "step": 18670 }, { "grad_norm": 0.39159294962882996, "learning_rate": 9.107436044883738e-05, "loss": 0.033, "step": 18680 }, { "grad_norm": 1.4964145421981812, "learning_rate": 9.10625712769559e-05, "loss": 0.0288, "step": 18690 }, { "grad_norm": 0.36203745007514954, "learning_rate": 9.105077508860406e-05, "loss": 0.0346, "step": 18700 }, { "grad_norm": 0.5695075988769531, "learning_rate": 9.103897188579751e-05, "loss": 0.0292, "step": 18710 }, { "grad_norm": 0.2617528736591339, "learning_rate": 9.102716167055308e-05, "loss": 0.0297, "step": 18720 }, { "grad_norm": 0.4110233187675476, "learning_rate": 9.10153444448888e-05, "loss": 0.0561, "step": 18730 }, { "grad_norm": 0.32445642352104187, "learning_rate": 9.100352021082393e-05, "loss": 0.0279, "step": 18740 }, { "grad_norm": 0.3463228642940521, "learning_rate": 9.099168897037891e-05, "loss": 0.0306, "step": 18750 }, { "grad_norm": 0.41838130354881287, "learning_rate": 9.097985072557538e-05, "loss": 0.0342, "step": 18760 }, { "grad_norm": 0.5416314005851746, "learning_rate": 9.096800547843615e-05, "loss": 0.0524, "step": 18770 }, { "grad_norm": 0.33431321382522583, "learning_rate": 9.095615323098526e-05, "loss": 0.0439, "step": 18780 }, { "grad_norm": 0.8511385321617126, "learning_rate": 9.094429398524795e-05, "loss": 0.0329, "step": 18790 }, { "grad_norm": 0.29212382435798645, "learning_rate": 9.093242774325061e-05, "loss": 0.0295, "step": 18800 }, { "grad_norm": 0.33517101407051086, "learning_rate": 9.092055450702088e-05, "loss": 0.0347, "step": 18810 }, { "grad_norm": 0.377108633518219, "learning_rate": 9.090867427858756e-05, "loss": 0.0287, "step": 18820 }, { "grad_norm": 0.2569846212863922, "learning_rate": 9.089678705998066e-05, "loss": 0.0321, "step": 18830 }, { "grad_norm": 0.29994887113571167, "learning_rate": 9.088489285323139e-05, "loss": 0.036, "step": 18840 }, { "grad_norm": 0.5267969965934753, "learning_rate": 9.087299166037212e-05, "loss": 0.0388, "step": 18850 }, { "grad_norm": 0.349223792552948, "learning_rate": 9.086108348343647e-05, "loss": 0.0387, "step": 18860 }, { "grad_norm": 0.34769266843795776, "learning_rate": 9.08491683244592e-05, "loss": 0.0358, "step": 18870 }, { "grad_norm": 0.3696376383304596, "learning_rate": 9.08372461854763e-05, "loss": 0.0366, "step": 18880 }, { "grad_norm": 0.32415980100631714, "learning_rate": 9.082531706852492e-05, "loss": 0.0331, "step": 18890 }, { "grad_norm": 0.624326229095459, "learning_rate": 9.081338097564342e-05, "loss": 0.0422, "step": 18900 }, { "grad_norm": 0.29099783301353455, "learning_rate": 9.080143790887137e-05, "loss": 0.0293, "step": 18910 }, { "grad_norm": 0.3864312171936035, "learning_rate": 9.07894878702495e-05, "loss": 0.035, "step": 18920 }, { "grad_norm": 0.2399662882089615, "learning_rate": 9.077753086181974e-05, "loss": 0.0399, "step": 18930 }, { "grad_norm": 0.32123154401779175, "learning_rate": 9.076556688562524e-05, "loss": 0.0371, "step": 18940 }, { "grad_norm": 0.2614983320236206, "learning_rate": 9.075359594371029e-05, "loss": 0.0297, "step": 18950 }, { "grad_norm": 0.3648661971092224, "learning_rate": 9.07416180381204e-05, "loss": 0.0373, "step": 18960 }, { "grad_norm": 0.3933961093425751, "learning_rate": 9.072963317090228e-05, "loss": 0.0326, "step": 18970 }, { "grad_norm": 0.30743417143821716, "learning_rate": 9.071764134410382e-05, "loss": 0.0356, "step": 18980 }, { "grad_norm": 0.4528582692146301, "learning_rate": 9.070564255977407e-05, "loss": 0.0336, "step": 18990 }, { "grad_norm": 0.32262977957725525, "learning_rate": 9.06936368199633e-05, "loss": 0.0347, "step": 19000 }, { "grad_norm": 0.2800362706184387, "learning_rate": 9.0681624126723e-05, "loss": 0.0355, "step": 19010 }, { "grad_norm": 1.9005722999572754, "learning_rate": 9.066960448210576e-05, "loss": 0.0295, "step": 19020 }, { "grad_norm": 0.35600510239601135, "learning_rate": 9.065757788816543e-05, "loss": 0.0322, "step": 19030 }, { "grad_norm": 0.3562047481536865, "learning_rate": 9.064554434695705e-05, "loss": 0.0438, "step": 19040 }, { "grad_norm": 0.34994325041770935, "learning_rate": 9.063350386053677e-05, "loss": 0.0358, "step": 19050 }, { "grad_norm": 0.3974023163318634, "learning_rate": 9.062145643096202e-05, "loss": 0.0353, "step": 19060 }, { "grad_norm": 0.8473291397094727, "learning_rate": 9.060940206029136e-05, "loss": 0.0426, "step": 19070 }, { "grad_norm": 0.337381511926651, "learning_rate": 9.059734075058457e-05, "loss": 0.0369, "step": 19080 }, { "grad_norm": 0.31261545419692993, "learning_rate": 9.058527250390257e-05, "loss": 0.0436, "step": 19090 }, { "grad_norm": 0.3080952763557434, "learning_rate": 9.057319732230752e-05, "loss": 0.0393, "step": 19100 }, { "grad_norm": 0.5509666204452515, "learning_rate": 9.056111520786273e-05, "loss": 0.0352, "step": 19110 }, { "grad_norm": 0.39272716641426086, "learning_rate": 9.054902616263268e-05, "loss": 0.0371, "step": 19120 }, { "grad_norm": 0.33251482248306274, "learning_rate": 9.05369301886831e-05, "loss": 0.0413, "step": 19130 }, { "grad_norm": 0.672741174697876, "learning_rate": 9.052482728808083e-05, "loss": 0.0292, "step": 19140 }, { "grad_norm": 0.3715266287326813, "learning_rate": 9.051271746289391e-05, "loss": 0.0283, "step": 19150 }, { "grad_norm": 0.41550955176353455, "learning_rate": 9.050060071519162e-05, "loss": 0.0403, "step": 19160 }, { "grad_norm": 0.3063233196735382, "learning_rate": 9.048847704704437e-05, "loss": 0.0313, "step": 19170 }, { "grad_norm": 0.38180261850357056, "learning_rate": 9.047634646052376e-05, "loss": 0.0338, "step": 19180 }, { "grad_norm": 0.3494808077812195, "learning_rate": 9.046420895770256e-05, "loss": 0.0474, "step": 19190 }, { "grad_norm": 0.41006121039390564, "learning_rate": 9.045206454065473e-05, "loss": 0.0391, "step": 19200 }, { "grad_norm": 0.36753085255622864, "learning_rate": 9.043991321145546e-05, "loss": 0.0355, "step": 19210 }, { "grad_norm": 0.40542730689048767, "learning_rate": 9.042775497218105e-05, "loss": 0.0307, "step": 19220 }, { "grad_norm": 0.28286105394363403, "learning_rate": 9.041558982490901e-05, "loss": 0.0308, "step": 19230 }, { "grad_norm": 0.2924899458885193, "learning_rate": 9.040341777171805e-05, "loss": 0.0301, "step": 19240 }, { "grad_norm": 0.3015260100364685, "learning_rate": 9.039123881468802e-05, "loss": 0.0285, "step": 19250 }, { "grad_norm": 0.26809442043304443, "learning_rate": 9.037905295589998e-05, "loss": 0.029, "step": 19260 }, { "grad_norm": 0.2693219482898712, "learning_rate": 9.036686019743617e-05, "loss": 0.0245, "step": 19270 }, { "grad_norm": 0.2314576655626297, "learning_rate": 9.035466054137997e-05, "loss": 0.0269, "step": 19280 }, { "grad_norm": 0.3330080211162567, "learning_rate": 9.0342453989816e-05, "loss": 0.0319, "step": 19290 }, { "grad_norm": 0.287581205368042, "learning_rate": 9.033024054483e-05, "loss": 0.031, "step": 19300 }, { "grad_norm": 0.3740918040275574, "learning_rate": 9.031802020850894e-05, "loss": 0.029, "step": 19310 }, { "grad_norm": 0.22648847103118896, "learning_rate": 9.030579298294092e-05, "loss": 0.0321, "step": 19320 }, { "grad_norm": 0.330364465713501, "learning_rate": 9.029355887021524e-05, "loss": 0.0375, "step": 19330 }, { "grad_norm": 0.3542436957359314, "learning_rate": 9.028131787242238e-05, "loss": 0.0456, "step": 19340 }, { "grad_norm": 0.19135266542434692, "learning_rate": 9.026906999165399e-05, "loss": 0.0334, "step": 19350 }, { "grad_norm": 0.33862629532814026, "learning_rate": 9.025681523000291e-05, "loss": 0.031, "step": 19360 }, { "grad_norm": 0.2565687596797943, "learning_rate": 9.024455358956315e-05, "loss": 0.03, "step": 19370 }, { "grad_norm": 0.30596181750297546, "learning_rate": 9.023228507242984e-05, "loss": 0.0376, "step": 19380 }, { "grad_norm": 0.41329675912857056, "learning_rate": 9.022000968069937e-05, "loss": 0.0377, "step": 19390 }, { "grad_norm": 0.3813146948814392, "learning_rate": 9.020772741646928e-05, "loss": 0.0336, "step": 19400 }, { "grad_norm": 0.36259615421295166, "learning_rate": 9.019543828183826e-05, "loss": 0.0326, "step": 19410 }, { "grad_norm": 0.19260264933109283, "learning_rate": 9.018314227890616e-05, "loss": 0.0282, "step": 19420 }, { "grad_norm": 0.34651219844818115, "learning_rate": 9.017083940977408e-05, "loss": 0.0502, "step": 19430 }, { "grad_norm": 0.34921199083328247, "learning_rate": 9.015852967654422e-05, "loss": 0.0346, "step": 19440 }, { "grad_norm": 0.3474169075489044, "learning_rate": 9.014621308131996e-05, "loss": 0.0321, "step": 19450 }, { "grad_norm": 0.36555805802345276, "learning_rate": 9.01338896262059e-05, "loss": 0.0376, "step": 19460 }, { "grad_norm": 0.28207579255104065, "learning_rate": 9.012155931330777e-05, "loss": 0.042, "step": 19470 }, { "grad_norm": 0.2767799198627472, "learning_rate": 9.010922214473246e-05, "loss": 0.0259, "step": 19480 }, { "grad_norm": 0.3090585172176361, "learning_rate": 9.009687812258808e-05, "loss": 0.042, "step": 19490 }, { "grad_norm": 0.24491845071315765, "learning_rate": 9.00845272489839e-05, "loss": 0.029, "step": 19500 }, { "grad_norm": 0.26351144909858704, "learning_rate": 9.007216952603031e-05, "loss": 0.0453, "step": 19510 }, { "grad_norm": 0.2875118851661682, "learning_rate": 9.005980495583894e-05, "loss": 0.03, "step": 19520 }, { "grad_norm": 0.35444602370262146, "learning_rate": 9.004743354052252e-05, "loss": 0.0266, "step": 19530 }, { "grad_norm": 0.21677789092063904, "learning_rate": 9.003505528219503e-05, "loss": 0.0275, "step": 19540 }, { "grad_norm": 0.30111458897590637, "learning_rate": 9.002267018297154e-05, "loss": 0.0284, "step": 19550 }, { "grad_norm": 0.38702642917633057, "learning_rate": 9.001027824496834e-05, "loss": 0.0324, "step": 19560 }, { "grad_norm": 0.31170177459716797, "learning_rate": 8.999787947030287e-05, "loss": 0.0281, "step": 19570 }, { "grad_norm": 0.24424314498901367, "learning_rate": 8.998547386109376e-05, "loss": 0.0333, "step": 19580 }, { "grad_norm": 0.3201070725917816, "learning_rate": 8.997306141946073e-05, "loss": 0.028, "step": 19590 }, { "grad_norm": 0.32617896795272827, "learning_rate": 8.996064214752481e-05, "loss": 0.0326, "step": 19600 }, { "grad_norm": 0.24396611750125885, "learning_rate": 8.994821604740806e-05, "loss": 0.0296, "step": 19610 }, { "grad_norm": 0.3728139400482178, "learning_rate": 8.993578312123377e-05, "loss": 0.0311, "step": 19620 }, { "grad_norm": 0.3526402711868286, "learning_rate": 8.992334337112639e-05, "loss": 0.0351, "step": 19630 }, { "grad_norm": 0.32229405641555786, "learning_rate": 8.991089679921154e-05, "loss": 0.0303, "step": 19640 }, { "grad_norm": 0.29943618178367615, "learning_rate": 8.989844340761599e-05, "loss": 0.0354, "step": 19650 }, { "grad_norm": 1.3942638635635376, "learning_rate": 8.988598319846768e-05, "loss": 0.0314, "step": 19660 }, { "grad_norm": 0.295678049325943, "learning_rate": 8.987351617389574e-05, "loss": 0.034, "step": 19670 }, { "grad_norm": 0.38986632227897644, "learning_rate": 8.98610423360304e-05, "loss": 0.0308, "step": 19680 }, { "grad_norm": 0.3513506352901459, "learning_rate": 8.984856168700317e-05, "loss": 0.037, "step": 19690 }, { "grad_norm": 0.45471447706222534, "learning_rate": 8.983607422894658e-05, "loss": 0.0369, "step": 19700 }, { "grad_norm": 1.0049211978912354, "learning_rate": 8.982357996399442e-05, "loss": 0.0337, "step": 19710 }, { "grad_norm": 0.2856704294681549, "learning_rate": 8.981107889428164e-05, "loss": 0.0368, "step": 19720 }, { "grad_norm": 0.25255727767944336, "learning_rate": 8.979857102194428e-05, "loss": 0.035, "step": 19730 }, { "grad_norm": 0.30818748474121094, "learning_rate": 8.978605634911968e-05, "loss": 0.0372, "step": 19740 }, { "grad_norm": 0.40397292375564575, "learning_rate": 8.977353487794616e-05, "loss": 0.029, "step": 19750 }, { "grad_norm": 0.3518747389316559, "learning_rate": 8.976100661056334e-05, "loss": 0.036, "step": 19760 }, { "grad_norm": 0.2875145971775055, "learning_rate": 8.974847154911197e-05, "loss": 0.0284, "step": 19770 }, { "grad_norm": 0.3452318608760834, "learning_rate": 8.973592969573393e-05, "loss": 0.0325, "step": 19780 }, { "grad_norm": 0.9862524271011353, "learning_rate": 8.972338105257228e-05, "loss": 0.0374, "step": 19790 }, { "grad_norm": 0.23973001539707184, "learning_rate": 8.971082562177125e-05, "loss": 0.0315, "step": 19800 }, { "grad_norm": 0.3162707984447479, "learning_rate": 8.96982634054762e-05, "loss": 0.0295, "step": 19810 }, { "grad_norm": 0.35477909445762634, "learning_rate": 8.96856944058337e-05, "loss": 0.0318, "step": 19820 }, { "grad_norm": 0.37827402353286743, "learning_rate": 8.967311862499144e-05, "loss": 0.0263, "step": 19830 }, { "grad_norm": 0.36042454838752747, "learning_rate": 8.966053606509825e-05, "loss": 0.0343, "step": 19840 }, { "grad_norm": 0.6336508989334106, "learning_rate": 8.964794672830417e-05, "loss": 0.0418, "step": 19850 }, { "grad_norm": 0.29117220640182495, "learning_rate": 8.963535061676038e-05, "loss": 0.04, "step": 19860 }, { "grad_norm": 0.3284386396408081, "learning_rate": 8.962274773261918e-05, "loss": 0.0313, "step": 19870 }, { "grad_norm": 0.38947463035583496, "learning_rate": 8.961013807803409e-05, "loss": 0.035, "step": 19880 }, { "grad_norm": 0.2618682384490967, "learning_rate": 8.959752165515973e-05, "loss": 0.0301, "step": 19890 }, { "grad_norm": 2.2950315475463867, "learning_rate": 8.958489846615193e-05, "loss": 0.0559, "step": 19900 }, { "grad_norm": 0.3380938172340393, "learning_rate": 8.957226851316762e-05, "loss": 0.0336, "step": 19910 }, { "grad_norm": 0.3185369670391083, "learning_rate": 8.955963179836493e-05, "loss": 0.0327, "step": 19920 }, { "grad_norm": 0.3579740524291992, "learning_rate": 8.954698832390312e-05, "loss": 0.0326, "step": 19930 }, { "grad_norm": 0.3192031681537628, "learning_rate": 8.953433809194263e-05, "loss": 0.0306, "step": 19940 }, { "grad_norm": 0.43270641565322876, "learning_rate": 8.9521681104645e-05, "loss": 0.0339, "step": 19950 }, { "grad_norm": 0.30351749062538147, "learning_rate": 8.9509017364173e-05, "loss": 0.0285, "step": 19960 }, { "grad_norm": 0.30210280418395996, "learning_rate": 8.949634687269052e-05, "loss": 0.0309, "step": 19970 }, { "grad_norm": 0.31297537684440613, "learning_rate": 8.948366963236259e-05, "loss": 0.0356, "step": 19980 }, { "grad_norm": 0.3499440550804138, "learning_rate": 8.947098564535538e-05, "loss": 0.0302, "step": 19990 }, { "grad_norm": 0.382102370262146, "learning_rate": 8.945829491383627e-05, "loss": 0.034, "step": 20000 }, { "grad_norm": 0.3887753486633301, "learning_rate": 8.944559743997374e-05, "loss": 0.0295, "step": 20010 }, { "grad_norm": 0.8222836852073669, "learning_rate": 8.943289322593746e-05, "loss": 0.0334, "step": 20020 }, { "grad_norm": 0.2977130711078644, "learning_rate": 8.942018227389821e-05, "loss": 0.0297, "step": 20030 }, { "grad_norm": 0.36920106410980225, "learning_rate": 8.940746458602795e-05, "loss": 0.0496, "step": 20040 }, { "grad_norm": 0.3862466514110565, "learning_rate": 8.939474016449979e-05, "loss": 0.0374, "step": 20050 }, { "grad_norm": 0.7653757333755493, "learning_rate": 8.938200901148799e-05, "loss": 0.0292, "step": 20060 }, { "grad_norm": 0.4595189392566681, "learning_rate": 8.936927112916795e-05, "loss": 0.0371, "step": 20070 }, { "grad_norm": 0.34472009539604187, "learning_rate": 8.935652651971622e-05, "loss": 0.0349, "step": 20080 }, { "grad_norm": 0.29883697628974915, "learning_rate": 8.934377518531052e-05, "loss": 0.033, "step": 20090 }, { "grad_norm": 0.3153780996799469, "learning_rate": 8.933101712812967e-05, "loss": 0.0267, "step": 20100 }, { "grad_norm": 0.2397211343050003, "learning_rate": 8.931825235035374e-05, "loss": 0.0299, "step": 20110 }, { "grad_norm": 0.4204062819480896, "learning_rate": 8.930548085416382e-05, "loss": 0.0314, "step": 20120 }, { "grad_norm": 0.2390129715204239, "learning_rate": 8.92927026417422e-05, "loss": 0.0361, "step": 20130 }, { "grad_norm": 0.28742262721061707, "learning_rate": 8.92799177152724e-05, "loss": 0.0303, "step": 20140 }, { "grad_norm": 0.2986000180244446, "learning_rate": 8.926712607693895e-05, "loss": 0.0331, "step": 20150 }, { "grad_norm": 0.26622745394706726, "learning_rate": 8.925432772892762e-05, "loss": 0.0318, "step": 20160 }, { "grad_norm": 0.307851642370224, "learning_rate": 8.924152267342529e-05, "loss": 0.0347, "step": 20170 }, { "grad_norm": 0.38188132643699646, "learning_rate": 8.922871091261998e-05, "loss": 0.0316, "step": 20180 }, { "grad_norm": 0.3260954022407532, "learning_rate": 8.92158924487009e-05, "loss": 0.0319, "step": 20190 }, { "grad_norm": 0.3902558386325836, "learning_rate": 8.920306728385834e-05, "loss": 0.0294, "step": 20200 }, { "grad_norm": 0.28760963678359985, "learning_rate": 8.919023542028379e-05, "loss": 0.0331, "step": 20210 }, { "grad_norm": 0.7387185096740723, "learning_rate": 8.917739686016988e-05, "loss": 0.0367, "step": 20220 }, { "grad_norm": 0.36927542090415955, "learning_rate": 8.916455160571033e-05, "loss": 0.0364, "step": 20230 }, { "grad_norm": 0.3826165497303009, "learning_rate": 8.915169965910008e-05, "loss": 0.0333, "step": 20240 }, { "grad_norm": 0.30390268564224243, "learning_rate": 8.913884102253514e-05, "loss": 0.0275, "step": 20250 }, { "grad_norm": 0.328651487827301, "learning_rate": 8.912597569821273e-05, "loss": 0.0282, "step": 20260 }, { "grad_norm": 0.24347424507141113, "learning_rate": 8.911310368833118e-05, "loss": 0.0314, "step": 20270 }, { "grad_norm": 0.252301961183548, "learning_rate": 8.910022499508994e-05, "loss": 0.0299, "step": 20280 }, { "grad_norm": 0.34891486167907715, "learning_rate": 8.908733962068965e-05, "loss": 0.0407, "step": 20290 }, { "grad_norm": 0.38415291905403137, "learning_rate": 8.907444756733207e-05, "loss": 0.0335, "step": 20300 }, { "grad_norm": 0.33177649974823, "learning_rate": 8.906154883722006e-05, "loss": 0.0427, "step": 20310 }, { "grad_norm": 0.48109984397888184, "learning_rate": 8.904864343255773e-05, "loss": 0.0286, "step": 20320 }, { "grad_norm": 0.3373034596443176, "learning_rate": 8.90357313555502e-05, "loss": 0.0355, "step": 20330 }, { "grad_norm": 0.6679546236991882, "learning_rate": 8.90228126084038e-05, "loss": 0.0362, "step": 20340 }, { "grad_norm": 0.3817780613899231, "learning_rate": 8.900988719332601e-05, "loss": 0.036, "step": 20350 }, { "grad_norm": 0.32158729434013367, "learning_rate": 8.899695511252542e-05, "loss": 0.0361, "step": 20360 }, { "grad_norm": 0.5186676383018494, "learning_rate": 8.898401636821176e-05, "loss": 0.0351, "step": 20370 }, { "grad_norm": 0.3295230567455292, "learning_rate": 8.897107096259593e-05, "loss": 0.0453, "step": 20380 }, { "grad_norm": 0.47911423444747925, "learning_rate": 8.895811889788994e-05, "loss": 0.0299, "step": 20390 }, { "grad_norm": 0.2697046995162964, "learning_rate": 8.894516017630692e-05, "loss": 0.0261, "step": 20400 }, { "grad_norm": 0.5680161714553833, "learning_rate": 8.893219480006118e-05, "loss": 0.0308, "step": 20410 }, { "grad_norm": 0.3049478828907013, "learning_rate": 8.891922277136817e-05, "loss": 0.0289, "step": 20420 }, { "grad_norm": 0.27939072251319885, "learning_rate": 8.890624409244441e-05, "loss": 0.0329, "step": 20430 }, { "grad_norm": 0.33631911873817444, "learning_rate": 8.889325876550763e-05, "loss": 0.0389, "step": 20440 }, { "grad_norm": 0.30466803908348083, "learning_rate": 8.888026679277666e-05, "loss": 0.0335, "step": 20450 }, { "grad_norm": 0.26096847653388977, "learning_rate": 8.886726817647147e-05, "loss": 0.0453, "step": 20460 }, { "grad_norm": 0.2640622854232788, "learning_rate": 8.885426291881319e-05, "loss": 0.0397, "step": 20470 }, { "grad_norm": 0.2919905185699463, "learning_rate": 8.884125102202401e-05, "loss": 0.0305, "step": 20480 }, { "grad_norm": 0.3496383726596832, "learning_rate": 8.882823248832736e-05, "loss": 0.0298, "step": 20490 }, { "grad_norm": 0.3121950328350067, "learning_rate": 8.881520731994772e-05, "loss": 0.0372, "step": 20500 }, { "grad_norm": 0.25048214197158813, "learning_rate": 8.880217551911077e-05, "loss": 0.0296, "step": 20510 }, { "grad_norm": 0.4313671290874481, "learning_rate": 8.878913708804323e-05, "loss": 0.0282, "step": 20520 }, { "grad_norm": 0.29393744468688965, "learning_rate": 8.877609202897308e-05, "loss": 0.0315, "step": 20530 }, { "grad_norm": 0.28255757689476013, "learning_rate": 8.876304034412933e-05, "loss": 0.0253, "step": 20540 }, { "grad_norm": 0.28218159079551697, "learning_rate": 8.874998203574214e-05, "loss": 0.0279, "step": 20550 }, { "grad_norm": 0.46566274762153625, "learning_rate": 8.873691710604284e-05, "loss": 0.0294, "step": 20560 }, { "grad_norm": 0.2821458578109741, "learning_rate": 8.872384555726387e-05, "loss": 0.023, "step": 20570 }, { "grad_norm": 0.42427247762680054, "learning_rate": 8.871076739163878e-05, "loss": 0.0244, "step": 20580 }, { "grad_norm": 0.38823312520980835, "learning_rate": 8.86976826114023e-05, "loss": 0.0287, "step": 20590 }, { "grad_norm": 0.32834360003471375, "learning_rate": 8.868459121879023e-05, "loss": 0.0365, "step": 20600 }, { "grad_norm": 0.3310912549495697, "learning_rate": 8.867149321603956e-05, "loss": 0.0418, "step": 20610 }, { "grad_norm": 0.3074679672718048, "learning_rate": 8.865838860538835e-05, "loss": 0.0256, "step": 20620 }, { "grad_norm": 0.35840389132499695, "learning_rate": 8.864527738907585e-05, "loss": 0.0314, "step": 20630 }, { "grad_norm": 0.3139476478099823, "learning_rate": 8.863215956934239e-05, "loss": 0.0321, "step": 20640 }, { "grad_norm": 0.27300339937210083, "learning_rate": 8.861903514842947e-05, "loss": 0.024, "step": 20650 }, { "grad_norm": 0.38522496819496155, "learning_rate": 8.860590412857966e-05, "loss": 0.028, "step": 20660 }, { "grad_norm": 0.3226935565471649, "learning_rate": 8.85927665120367e-05, "loss": 0.0268, "step": 20670 }, { "grad_norm": 0.3555399775505066, "learning_rate": 8.857962230104546e-05, "loss": 0.0285, "step": 20680 }, { "grad_norm": 0.22526055574417114, "learning_rate": 8.856647149785193e-05, "loss": 0.0275, "step": 20690 }, { "grad_norm": 0.2768881022930145, "learning_rate": 8.855331410470322e-05, "loss": 0.0296, "step": 20700 }, { "grad_norm": 0.33163198828697205, "learning_rate": 8.854015012384756e-05, "loss": 0.0302, "step": 20710 }, { "grad_norm": 0.339609831571579, "learning_rate": 8.852697955753433e-05, "loss": 0.0375, "step": 20720 }, { "grad_norm": 0.2819393575191498, "learning_rate": 8.851380240801399e-05, "loss": 0.0374, "step": 20730 }, { "grad_norm": 0.2661682963371277, "learning_rate": 8.850061867753818e-05, "loss": 0.0222, "step": 20740 }, { "grad_norm": 0.3722224533557892, "learning_rate": 8.848742836835963e-05, "loss": 0.0369, "step": 20750 }, { "grad_norm": 0.272428959608078, "learning_rate": 8.847423148273221e-05, "loss": 0.0348, "step": 20760 }, { "grad_norm": 0.35397806763648987, "learning_rate": 8.846102802291092e-05, "loss": 0.0436, "step": 20770 }, { "grad_norm": 0.333298921585083, "learning_rate": 8.844781799115183e-05, "loss": 0.042, "step": 20780 }, { "grad_norm": 0.37503287196159363, "learning_rate": 8.84346013897122e-05, "loss": 0.027, "step": 20790 }, { "grad_norm": 0.23980163037776947, "learning_rate": 8.842137822085038e-05, "loss": 0.0314, "step": 20800 }, { "grad_norm": 0.2592593729496002, "learning_rate": 8.840814848682585e-05, "loss": 0.0255, "step": 20810 }, { "grad_norm": 0.30425161123275757, "learning_rate": 8.83949121898992e-05, "loss": 0.0371, "step": 20820 }, { "grad_norm": 0.2980601191520691, "learning_rate": 8.838166933233217e-05, "loss": 0.0316, "step": 20830 }, { "grad_norm": 0.31404903531074524, "learning_rate": 8.83684199163876e-05, "loss": 0.0387, "step": 20840 }, { "grad_norm": 0.3339456617832184, "learning_rate": 8.835516394432943e-05, "loss": 0.0338, "step": 20850 }, { "grad_norm": 0.4448781907558441, "learning_rate": 8.834190141842276e-05, "loss": 0.0387, "step": 20860 }, { "grad_norm": 0.41568559408187866, "learning_rate": 8.83286323409338e-05, "loss": 0.0306, "step": 20870 }, { "grad_norm": 0.22875648736953735, "learning_rate": 8.831535671412986e-05, "loss": 0.0312, "step": 20880 }, { "grad_norm": 0.31764721870422363, "learning_rate": 8.830207454027938e-05, "loss": 0.0373, "step": 20890 }, { "grad_norm": 0.30251574516296387, "learning_rate": 8.828878582165192e-05, "loss": 0.0337, "step": 20900 }, { "grad_norm": 0.3468495011329651, "learning_rate": 8.827549056051818e-05, "loss": 0.0315, "step": 20910 }, { "grad_norm": 0.32025784254074097, "learning_rate": 8.826218875914993e-05, "loss": 0.0294, "step": 20920 }, { "grad_norm": 0.33566996455192566, "learning_rate": 8.82488804198201e-05, "loss": 0.0272, "step": 20930 }, { "grad_norm": 0.28546473383903503, "learning_rate": 8.82355655448027e-05, "loss": 0.0289, "step": 20940 }, { "grad_norm": 0.4394429326057434, "learning_rate": 8.822224413637293e-05, "loss": 0.0371, "step": 20950 }, { "grad_norm": 1.9726333618164062, "learning_rate": 8.820891619680697e-05, "loss": 0.0336, "step": 20960 }, { "grad_norm": 0.2889609634876251, "learning_rate": 8.819558172838227e-05, "loss": 0.0282, "step": 20970 }, { "grad_norm": 0.39251068234443665, "learning_rate": 8.818224073337731e-05, "loss": 0.0242, "step": 20980 }, { "grad_norm": 0.32516002655029297, "learning_rate": 8.816889321407169e-05, "loss": 0.029, "step": 20990 }, { "grad_norm": 0.2589583396911621, "learning_rate": 8.815553917274615e-05, "loss": 0.0336, "step": 21000 }, { "grad_norm": 0.33040159940719604, "learning_rate": 8.81421786116825e-05, "loss": 0.0359, "step": 21010 }, { "grad_norm": 0.30736222863197327, "learning_rate": 8.812881153316373e-05, "loss": 0.0325, "step": 21020 }, { "grad_norm": 0.28824567794799805, "learning_rate": 8.81154379394739e-05, "loss": 0.0376, "step": 21030 }, { "grad_norm": 0.2854727804660797, "learning_rate": 8.810205783289818e-05, "loss": 0.0322, "step": 21040 }, { "grad_norm": 0.3816985487937927, "learning_rate": 8.808867121572286e-05, "loss": 0.0381, "step": 21050 }, { "grad_norm": 0.27015888690948486, "learning_rate": 8.807527809023537e-05, "loss": 0.0409, "step": 21060 }, { "grad_norm": 0.29990965127944946, "learning_rate": 8.80618784587242e-05, "loss": 0.0306, "step": 21070 }, { "grad_norm": 0.2634992003440857, "learning_rate": 8.804847232347902e-05, "loss": 0.0263, "step": 21080 }, { "grad_norm": 0.23186878859996796, "learning_rate": 8.803505968679054e-05, "loss": 0.0371, "step": 21090 }, { "grad_norm": 0.2741934657096863, "learning_rate": 8.802164055095061e-05, "loss": 0.0384, "step": 21100 }, { "grad_norm": 0.30983707308769226, "learning_rate": 8.80082149182522e-05, "loss": 0.0307, "step": 21110 }, { "grad_norm": 0.293239027261734, "learning_rate": 8.79947827909894e-05, "loss": 0.0304, "step": 21120 }, { "grad_norm": 0.4595594108104706, "learning_rate": 8.798134417145738e-05, "loss": 0.0414, "step": 21130 }, { "grad_norm": 0.3264326751232147, "learning_rate": 8.796789906195243e-05, "loss": 0.0398, "step": 21140 }, { "grad_norm": 0.29974645376205444, "learning_rate": 8.795444746477195e-05, "loss": 0.0287, "step": 21150 }, { "grad_norm": 0.22455459833145142, "learning_rate": 8.794098938221446e-05, "loss": 0.0293, "step": 21160 }, { "grad_norm": 0.22937728464603424, "learning_rate": 8.792752481657957e-05, "loss": 0.0315, "step": 21170 }, { "grad_norm": 0.29950278997421265, "learning_rate": 8.791405377016802e-05, "loss": 0.0407, "step": 21180 }, { "grad_norm": 0.46181777119636536, "learning_rate": 8.790057624528163e-05, "loss": 0.0389, "step": 21190 }, { "grad_norm": 0.5205345749855042, "learning_rate": 8.788709224422333e-05, "loss": 0.0359, "step": 21200 }, { "grad_norm": 0.49218979477882385, "learning_rate": 8.787360176929717e-05, "loss": 0.0362, "step": 21210 }, { "grad_norm": 0.3209773600101471, "learning_rate": 8.786010482280834e-05, "loss": 0.0333, "step": 21220 }, { "grad_norm": 0.38339388370513916, "learning_rate": 8.784660140706306e-05, "loss": 0.0367, "step": 21230 }, { "grad_norm": 0.44004011154174805, "learning_rate": 8.783309152436872e-05, "loss": 0.0333, "step": 21240 }, { "grad_norm": 0.36828091740608215, "learning_rate": 8.781957517703375e-05, "loss": 0.0275, "step": 21250 }, { "grad_norm": 0.5496537089347839, "learning_rate": 8.780605236736776e-05, "loss": 0.0381, "step": 21260 }, { "grad_norm": 0.29039841890335083, "learning_rate": 8.779252309768142e-05, "loss": 0.0347, "step": 21270 }, { "grad_norm": 0.3330870270729065, "learning_rate": 8.777898737028652e-05, "loss": 0.0342, "step": 21280 }, { "grad_norm": 0.330508291721344, "learning_rate": 8.776544518749591e-05, "loss": 0.0331, "step": 21290 }, { "grad_norm": 0.4505854547023773, "learning_rate": 8.775189655162364e-05, "loss": 0.0264, "step": 21300 }, { "grad_norm": 0.3193078935146332, "learning_rate": 8.773834146498474e-05, "loss": 0.0299, "step": 21310 }, { "grad_norm": 0.25131756067276, "learning_rate": 8.772477992989545e-05, "loss": 0.0259, "step": 21320 }, { "grad_norm": 0.3933301568031311, "learning_rate": 8.771121194867304e-05, "loss": 0.0388, "step": 21330 }, { "grad_norm": 0.3002784550189972, "learning_rate": 8.769763752363589e-05, "loss": 0.0284, "step": 21340 }, { "grad_norm": 0.3697550594806671, "learning_rate": 8.768405665710352e-05, "loss": 0.0319, "step": 21350 }, { "grad_norm": 0.2985265851020813, "learning_rate": 8.767046935139655e-05, "loss": 0.0298, "step": 21360 }, { "grad_norm": 0.35096287727355957, "learning_rate": 8.765687560883666e-05, "loss": 0.0274, "step": 21370 }, { "grad_norm": 0.2961009740829468, "learning_rate": 8.764327543174664e-05, "loss": 0.0262, "step": 21380 }, { "grad_norm": 0.2868914306163788, "learning_rate": 8.762966882245038e-05, "loss": 0.0287, "step": 21390 }, { "grad_norm": 0.2874295711517334, "learning_rate": 8.761605578327291e-05, "loss": 0.0316, "step": 21400 }, { "grad_norm": 0.2473398745059967, "learning_rate": 8.76024363165403e-05, "loss": 0.0349, "step": 21410 }, { "grad_norm": 0.3020894229412079, "learning_rate": 8.758881042457976e-05, "loss": 0.0333, "step": 21420 }, { "grad_norm": 0.2442140132188797, "learning_rate": 8.757517810971957e-05, "loss": 0.0267, "step": 21430 }, { "grad_norm": 0.4168330729007721, "learning_rate": 8.756153937428913e-05, "loss": 0.0307, "step": 21440 }, { "grad_norm": 0.3056495785713196, "learning_rate": 8.754789422061889e-05, "loss": 0.0385, "step": 21450 }, { "grad_norm": 0.30699610710144043, "learning_rate": 8.753424265104052e-05, "loss": 0.0291, "step": 21460 }, { "grad_norm": 0.40037113428115845, "learning_rate": 8.752058466788659e-05, "loss": 0.0407, "step": 21470 }, { "grad_norm": 0.32032114267349243, "learning_rate": 8.750692027349097e-05, "loss": 0.0276, "step": 21480 }, { "grad_norm": 0.3108524680137634, "learning_rate": 8.749324947018847e-05, "loss": 0.032, "step": 21490 }, { "grad_norm": 0.23460587859153748, "learning_rate": 8.747957226031507e-05, "loss": 0.0321, "step": 21500 }, { "grad_norm": 0.38565173745155334, "learning_rate": 8.746588864620787e-05, "loss": 0.0338, "step": 21510 }, { "grad_norm": 1.0716183185577393, "learning_rate": 8.745219863020498e-05, "loss": 0.0292, "step": 21520 }, { "grad_norm": 0.36279863119125366, "learning_rate": 8.743850221464564e-05, "loss": 0.0319, "step": 21530 }, { "grad_norm": 0.2479841113090515, "learning_rate": 8.742479940187026e-05, "loss": 0.0293, "step": 21540 }, { "grad_norm": 0.31160247325897217, "learning_rate": 8.74110901942202e-05, "loss": 0.0344, "step": 21550 }, { "grad_norm": 0.35310763120651245, "learning_rate": 8.739737459403803e-05, "loss": 0.0312, "step": 21560 }, { "grad_norm": 4.953132629394531, "learning_rate": 8.738365260366737e-05, "loss": 0.0348, "step": 21570 }, { "grad_norm": 0.31332331895828247, "learning_rate": 8.736992422545292e-05, "loss": 0.0253, "step": 21580 }, { "grad_norm": 0.3315850794315338, "learning_rate": 8.73561894617405e-05, "loss": 0.0306, "step": 21590 }, { "grad_norm": 0.5137568712234497, "learning_rate": 8.734244831487697e-05, "loss": 0.0316, "step": 21600 }, { "grad_norm": 0.2829173505306244, "learning_rate": 8.732870078721035e-05, "loss": 0.0337, "step": 21610 }, { "grad_norm": 0.2938004732131958, "learning_rate": 8.731494688108972e-05, "loss": 0.0265, "step": 21620 }, { "grad_norm": 0.9732280969619751, "learning_rate": 8.730118659886523e-05, "loss": 0.0349, "step": 21630 }, { "grad_norm": 0.3318890631198883, "learning_rate": 8.728741994288814e-05, "loss": 0.0293, "step": 21640 }, { "grad_norm": 0.31600040197372437, "learning_rate": 8.727364691551079e-05, "loss": 0.0313, "step": 21650 }, { "grad_norm": 0.36223119497299194, "learning_rate": 8.725986751908661e-05, "loss": 0.036, "step": 21660 }, { "grad_norm": 0.33432021737098694, "learning_rate": 8.724608175597016e-05, "loss": 0.0297, "step": 21670 }, { "grad_norm": 0.22999761998653412, "learning_rate": 8.723228962851699e-05, "loss": 0.0286, "step": 21680 }, { "grad_norm": 0.28337299823760986, "learning_rate": 8.721849113908385e-05, "loss": 0.0298, "step": 21690 }, { "grad_norm": 0.3248438239097595, "learning_rate": 8.720468629002848e-05, "loss": 0.0331, "step": 21700 }, { "grad_norm": 0.231761172413826, "learning_rate": 8.719087508370978e-05, "loss": 0.0223, "step": 21710 }, { "grad_norm": 0.3228580951690674, "learning_rate": 8.717705752248772e-05, "loss": 0.0355, "step": 21720 }, { "grad_norm": 0.3366279900074005, "learning_rate": 8.71632336087233e-05, "loss": 0.0284, "step": 21730 }, { "grad_norm": 0.7699724435806274, "learning_rate": 8.71494033447787e-05, "loss": 0.037, "step": 21740 }, { "grad_norm": 0.2365482747554779, "learning_rate": 8.713556673301708e-05, "loss": 0.026, "step": 21750 }, { "grad_norm": 0.31425243616104126, "learning_rate": 8.712172377580278e-05, "loss": 0.0454, "step": 21760 }, { "grad_norm": 0.31192103028297424, "learning_rate": 8.710787447550114e-05, "loss": 0.0312, "step": 21770 }, { "grad_norm": 1.8885411024093628, "learning_rate": 8.70940188344787e-05, "loss": 0.0395, "step": 21780 }, { "grad_norm": 0.2888944447040558, "learning_rate": 8.708015685510293e-05, "loss": 0.0298, "step": 21790 }, { "grad_norm": 0.29289013147354126, "learning_rate": 8.706628853974252e-05, "loss": 0.0305, "step": 21800 }, { "grad_norm": 1.1501415967941284, "learning_rate": 8.705241389076715e-05, "loss": 0.0338, "step": 21810 }, { "grad_norm": 0.33916833996772766, "learning_rate": 8.703853291054764e-05, "loss": 0.0312, "step": 21820 }, { "grad_norm": 0.4855216145515442, "learning_rate": 8.702464560145587e-05, "loss": 0.0361, "step": 21830 }, { "grad_norm": 0.3521707057952881, "learning_rate": 8.701075196586476e-05, "loss": 0.028, "step": 21840 }, { "grad_norm": 0.344743549823761, "learning_rate": 8.699685200614842e-05, "loss": 0.0368, "step": 21850 }, { "grad_norm": 0.2444888800382614, "learning_rate": 8.698294572468193e-05, "loss": 0.037, "step": 21860 }, { "grad_norm": 0.20406734943389893, "learning_rate": 8.696903312384148e-05, "loss": 0.0335, "step": 21870 }, { "grad_norm": 0.48149725794792175, "learning_rate": 8.695511420600439e-05, "loss": 0.0286, "step": 21880 }, { "grad_norm": 0.31588104367256165, "learning_rate": 8.694118897354901e-05, "loss": 0.0251, "step": 21890 }, { "grad_norm": 1.0279476642608643, "learning_rate": 8.692725742885478e-05, "loss": 0.0276, "step": 21900 }, { "grad_norm": 0.35041648149490356, "learning_rate": 8.691331957430221e-05, "loss": 0.0358, "step": 21910 }, { "grad_norm": 0.7811600565910339, "learning_rate": 8.68993754122729e-05, "loss": 0.0349, "step": 21920 }, { "grad_norm": 0.5086073279380798, "learning_rate": 8.688542494514955e-05, "loss": 0.0288, "step": 21930 }, { "grad_norm": 0.6238490343093872, "learning_rate": 8.68714681753159e-05, "loss": 0.0271, "step": 21940 }, { "grad_norm": 0.2404600828886032, "learning_rate": 8.685750510515676e-05, "loss": 0.025, "step": 21950 }, { "grad_norm": 0.4565236568450928, "learning_rate": 8.684353573705805e-05, "loss": 0.0282, "step": 21960 }, { "grad_norm": 0.2539159059524536, "learning_rate": 8.682956007340677e-05, "loss": 0.0302, "step": 21970 }, { "grad_norm": 0.36069488525390625, "learning_rate": 8.681557811659095e-05, "loss": 0.025, "step": 21980 }, { "grad_norm": 0.477167010307312, "learning_rate": 8.680158986899974e-05, "loss": 0.0309, "step": 21990 }, { "grad_norm": 0.31236129999160767, "learning_rate": 8.678759533302335e-05, "loss": 0.0281, "step": 22000 }, { "grad_norm": 0.24952061474323273, "learning_rate": 8.677359451105308e-05, "loss": 0.0257, "step": 22010 }, { "grad_norm": 0.23106932640075684, "learning_rate": 8.675958740548123e-05, "loss": 0.0285, "step": 22020 }, { "grad_norm": 0.35763028264045715, "learning_rate": 8.674557401870129e-05, "loss": 0.0296, "step": 22030 }, { "grad_norm": 0.3594840168952942, "learning_rate": 8.673155435310775e-05, "loss": 0.0294, "step": 22040 }, { "grad_norm": 0.5551986694335938, "learning_rate": 8.671752841109617e-05, "loss": 0.0238, "step": 22050 }, { "grad_norm": 0.29665282368659973, "learning_rate": 8.670349619506321e-05, "loss": 0.0281, "step": 22060 }, { "grad_norm": 0.2576073408126831, "learning_rate": 8.66894577074066e-05, "loss": 0.0324, "step": 22070 }, { "grad_norm": 0.2531616985797882, "learning_rate": 8.667541295052513e-05, "loss": 0.0269, "step": 22080 }, { "grad_norm": 0.3328308165073395, "learning_rate": 8.666136192681865e-05, "loss": 0.0358, "step": 22090 }, { "grad_norm": 0.33307191729545593, "learning_rate": 8.664730463868811e-05, "loss": 0.0366, "step": 22100 }, { "grad_norm": 0.3205569088459015, "learning_rate": 8.663324108853552e-05, "loss": 0.04, "step": 22110 }, { "grad_norm": 0.2556741535663605, "learning_rate": 8.661917127876395e-05, "loss": 0.0426, "step": 22120 }, { "grad_norm": 0.3138792812824249, "learning_rate": 8.660509521177754e-05, "loss": 0.0362, "step": 22130 }, { "grad_norm": 0.7153605818748474, "learning_rate": 8.65910128899815e-05, "loss": 0.0366, "step": 22140 }, { "grad_norm": 0.24870489537715912, "learning_rate": 8.657692431578214e-05, "loss": 0.0324, "step": 22150 }, { "grad_norm": 0.2578502595424652, "learning_rate": 8.656282949158679e-05, "loss": 0.0288, "step": 22160 }, { "grad_norm": 0.3476344048976898, "learning_rate": 8.654872841980388e-05, "loss": 0.0275, "step": 22170 }, { "grad_norm": 0.3819178640842438, "learning_rate": 8.653462110284289e-05, "loss": 0.0284, "step": 22180 }, { "grad_norm": 0.3493456244468689, "learning_rate": 8.652050754311437e-05, "loss": 0.0331, "step": 22190 }, { "grad_norm": 0.3265051543712616, "learning_rate": 8.650638774302995e-05, "loss": 0.0367, "step": 22200 }, { "grad_norm": 0.3884110152721405, "learning_rate": 8.649226170500233e-05, "loss": 0.0304, "step": 22210 }, { "grad_norm": 0.2559434175491333, "learning_rate": 8.647812943144524e-05, "loss": 0.0331, "step": 22220 }, { "grad_norm": 0.3033057749271393, "learning_rate": 8.646399092477351e-05, "loss": 0.0291, "step": 22230 }, { "grad_norm": 0.5814343094825745, "learning_rate": 8.644984618740301e-05, "loss": 0.0389, "step": 22240 }, { "grad_norm": 0.2792385518550873, "learning_rate": 8.643569522175073e-05, "loss": 0.0251, "step": 22250 }, { "grad_norm": 0.45044684410095215, "learning_rate": 8.642153803023463e-05, "loss": 0.0392, "step": 22260 }, { "grad_norm": 0.32976382970809937, "learning_rate": 8.640737461527383e-05, "loss": 0.0287, "step": 22270 }, { "grad_norm": 0.30656033754348755, "learning_rate": 8.639320497928845e-05, "loss": 0.0371, "step": 22280 }, { "grad_norm": 0.23914161324501038, "learning_rate": 8.637902912469969e-05, "loss": 0.024, "step": 22290 }, { "grad_norm": 0.2866462171077728, "learning_rate": 8.636484705392982e-05, "loss": 0.0308, "step": 22300 }, { "grad_norm": 0.22141778469085693, "learning_rate": 8.635065876940216e-05, "loss": 0.0265, "step": 22310 }, { "grad_norm": 0.3453361690044403, "learning_rate": 8.633646427354112e-05, "loss": 0.0281, "step": 22320 }, { "grad_norm": 0.2496062070131302, "learning_rate": 8.632226356877213e-05, "loss": 0.024, "step": 22330 }, { "grad_norm": 0.35470983386039734, "learning_rate": 8.630805665752173e-05, "loss": 0.0345, "step": 22340 }, { "grad_norm": 0.2860497534275055, "learning_rate": 8.629384354221748e-05, "loss": 0.0353, "step": 22350 }, { "grad_norm": 0.2857305407524109, "learning_rate": 8.627962422528797e-05, "loss": 0.0283, "step": 22360 }, { "grad_norm": 0.4164443910121918, "learning_rate": 8.626539870916296e-05, "loss": 0.0475, "step": 22370 }, { "grad_norm": 0.26042407751083374, "learning_rate": 8.625116699627317e-05, "loss": 0.0393, "step": 22380 }, { "grad_norm": 0.4083437919616699, "learning_rate": 8.623692908905041e-05, "loss": 0.0284, "step": 22390 }, { "grad_norm": 0.37717023491859436, "learning_rate": 8.622268498992755e-05, "loss": 0.0352, "step": 22400 }, { "grad_norm": 3.3786609172821045, "learning_rate": 8.620843470133851e-05, "loss": 0.0454, "step": 22410 }, { "grad_norm": 0.3918411135673523, "learning_rate": 8.619417822571829e-05, "loss": 0.0316, "step": 22420 }, { "grad_norm": 0.38719260692596436, "learning_rate": 8.617991556550292e-05, "loss": 0.028, "step": 22430 }, { "grad_norm": 0.2453937977552414, "learning_rate": 8.616564672312952e-05, "loss": 0.0247, "step": 22440 }, { "grad_norm": 0.27969610691070557, "learning_rate": 8.61513717010362e-05, "loss": 0.0314, "step": 22450 }, { "grad_norm": 0.29741358757019043, "learning_rate": 8.613709050166221e-05, "loss": 0.0288, "step": 22460 }, { "grad_norm": 0.342084676027298, "learning_rate": 8.61228031274478e-05, "loss": 0.0351, "step": 22470 }, { "grad_norm": 0.2643446624279022, "learning_rate": 8.610850958083431e-05, "loss": 0.0276, "step": 22480 }, { "grad_norm": 0.2999390959739685, "learning_rate": 8.609420986426409e-05, "loss": 0.0299, "step": 22490 }, { "grad_norm": 0.2428479939699173, "learning_rate": 8.60799039801806e-05, "loss": 0.0259, "step": 22500 }, { "grad_norm": 0.37460997700691223, "learning_rate": 8.606559193102828e-05, "loss": 0.0212, "step": 22510 }, { "grad_norm": 0.5869011878967285, "learning_rate": 8.605127371925273e-05, "loss": 0.0344, "step": 22520 }, { "grad_norm": 0.3860783576965332, "learning_rate": 8.603694934730047e-05, "loss": 0.0373, "step": 22530 }, { "grad_norm": 0.2997685372829437, "learning_rate": 8.602261881761919e-05, "loss": 0.025, "step": 22540 }, { "grad_norm": 0.2583668828010559, "learning_rate": 8.600828213265759e-05, "loss": 0.0329, "step": 22550 }, { "grad_norm": 0.38886597752571106, "learning_rate": 8.599393929486539e-05, "loss": 0.0458, "step": 22560 }, { "grad_norm": 0.31855279207229614, "learning_rate": 8.59795903066934e-05, "loss": 0.0272, "step": 22570 }, { "grad_norm": 0.30133914947509766, "learning_rate": 8.596523517059347e-05, "loss": 0.0245, "step": 22580 }, { "grad_norm": 0.4294811189174652, "learning_rate": 8.59508738890185e-05, "loss": 0.0281, "step": 22590 }, { "grad_norm": 0.2287948578596115, "learning_rate": 8.593650646442246e-05, "loss": 0.026, "step": 22600 }, { "grad_norm": 0.25422120094299316, "learning_rate": 8.59221328992603e-05, "loss": 0.0263, "step": 22610 }, { "grad_norm": 0.2312415987253189, "learning_rate": 8.590775319598813e-05, "loss": 0.029, "step": 22620 }, { "grad_norm": 0.3275466859340668, "learning_rate": 8.589336735706301e-05, "loss": 0.0274, "step": 22630 }, { "grad_norm": 0.329448938369751, "learning_rate": 8.587897538494307e-05, "loss": 0.0374, "step": 22640 }, { "grad_norm": 0.25026580691337585, "learning_rate": 8.586457728208756e-05, "loss": 0.0299, "step": 22650 }, { "grad_norm": 0.34639161825180054, "learning_rate": 8.585017305095667e-05, "loss": 0.0276, "step": 22660 }, { "grad_norm": 0.21689964830875397, "learning_rate": 8.583576269401173e-05, "loss": 0.0241, "step": 22670 }, { "grad_norm": 0.357419490814209, "learning_rate": 8.582134621371504e-05, "loss": 0.0327, "step": 22680 }, { "grad_norm": 0.38312479853630066, "learning_rate": 8.580692361253e-05, "loss": 0.0282, "step": 22690 }, { "grad_norm": 0.4353541433811188, "learning_rate": 8.579249489292104e-05, "loss": 0.032, "step": 22700 }, { "grad_norm": 0.3224048912525177, "learning_rate": 8.577806005735363e-05, "loss": 0.04, "step": 22710 }, { "grad_norm": 0.35131439566612244, "learning_rate": 8.576361910829429e-05, "loss": 0.0257, "step": 22720 }, { "grad_norm": 0.24635067582130432, "learning_rate": 8.574917204821057e-05, "loss": 0.0274, "step": 22730 }, { "grad_norm": 0.2741357386112213, "learning_rate": 8.57347188795711e-05, "loss": 0.0283, "step": 22740 }, { "grad_norm": 0.35197386145591736, "learning_rate": 8.572025960484551e-05, "loss": 0.0369, "step": 22750 }, { "grad_norm": 0.3120858073234558, "learning_rate": 8.57057942265045e-05, "loss": 0.0385, "step": 22760 }, { "grad_norm": 0.3386951982975006, "learning_rate": 8.569132274701984e-05, "loss": 0.0318, "step": 22770 }, { "grad_norm": 0.3273901343345642, "learning_rate": 8.567684516886427e-05, "loss": 0.0263, "step": 22780 }, { "grad_norm": 0.29980143904685974, "learning_rate": 8.56623614945116e-05, "loss": 0.024, "step": 22790 }, { "grad_norm": 0.3552224636077881, "learning_rate": 8.564787172643675e-05, "loss": 0.0218, "step": 22800 }, { "grad_norm": 0.31834617257118225, "learning_rate": 8.563337586711559e-05, "loss": 0.0303, "step": 22810 }, { "grad_norm": 0.4122709035873413, "learning_rate": 8.561887391902506e-05, "loss": 0.0294, "step": 22820 }, { "grad_norm": 0.3066251873970032, "learning_rate": 8.560436588464316e-05, "loss": 0.0236, "step": 22830 }, { "grad_norm": 0.41872933506965637, "learning_rate": 8.55898517664489e-05, "loss": 0.0318, "step": 22840 }, { "grad_norm": 0.34405213594436646, "learning_rate": 8.557533156692236e-05, "loss": 0.042, "step": 22850 }, { "grad_norm": 0.278074711561203, "learning_rate": 8.556080528854467e-05, "loss": 0.0271, "step": 22860 }, { "grad_norm": 1.4472160339355469, "learning_rate": 8.554627293379791e-05, "loss": 0.0329, "step": 22870 }, { "grad_norm": 0.2741892635822296, "learning_rate": 8.553173450516531e-05, "loss": 0.0419, "step": 22880 }, { "grad_norm": 0.28579020500183105, "learning_rate": 8.551719000513108e-05, "loss": 0.0284, "step": 22890 }, { "grad_norm": 0.24115435779094696, "learning_rate": 8.550263943618049e-05, "loss": 0.0402, "step": 22900 }, { "grad_norm": 0.3653021454811096, "learning_rate": 8.54880828007998e-05, "loss": 0.0311, "step": 22910 }, { "grad_norm": 0.4404592514038086, "learning_rate": 8.547352010147637e-05, "loss": 0.0349, "step": 22920 }, { "grad_norm": 0.29194003343582153, "learning_rate": 8.545895134069855e-05, "loss": 0.0296, "step": 22930 }, { "grad_norm": 0.2347765862941742, "learning_rate": 8.544437652095576e-05, "loss": 0.0285, "step": 22940 }, { "grad_norm": 1.2666006088256836, "learning_rate": 8.542979564473843e-05, "loss": 0.0414, "step": 22950 }, { "grad_norm": 0.32713285088539124, "learning_rate": 8.541520871453802e-05, "loss": 0.0262, "step": 22960 }, { "grad_norm": 0.3516915440559387, "learning_rate": 8.540061573284705e-05, "loss": 0.028, "step": 22970 }, { "grad_norm": 0.2650880217552185, "learning_rate": 8.538601670215906e-05, "loss": 0.0313, "step": 22980 }, { "grad_norm": 0.2826676368713379, "learning_rate": 8.537141162496864e-05, "loss": 0.0303, "step": 22990 }, { "grad_norm": 2.0303456783294678, "learning_rate": 8.535680050377137e-05, "loss": 0.0297, "step": 23000 }, { "grad_norm": 0.6570334434509277, "learning_rate": 8.534218334106391e-05, "loss": 0.0274, "step": 23010 }, { "grad_norm": 0.9163411855697632, "learning_rate": 8.532756013934393e-05, "loss": 0.0306, "step": 23020 }, { "grad_norm": 0.2779492437839508, "learning_rate": 8.531293090111012e-05, "loss": 0.0388, "step": 23030 }, { "grad_norm": 0.3208373486995697, "learning_rate": 8.529829562886225e-05, "loss": 0.0292, "step": 23040 }, { "grad_norm": 0.32620537281036377, "learning_rate": 8.528365432510105e-05, "loss": 0.0376, "step": 23050 }, { "grad_norm": 0.31939804553985596, "learning_rate": 8.526900699232833e-05, "loss": 0.0301, "step": 23060 }, { "grad_norm": 0.27772223949432373, "learning_rate": 8.525435363304695e-05, "loss": 0.0273, "step": 23070 }, { "grad_norm": 0.5738397240638733, "learning_rate": 8.523969424976072e-05, "loss": 0.027, "step": 23080 }, { "grad_norm": 0.30586960911750793, "learning_rate": 8.522502884497457e-05, "loss": 0.0236, "step": 23090 }, { "grad_norm": 0.3611381947994232, "learning_rate": 8.521035742119437e-05, "loss": 0.0305, "step": 23100 }, { "grad_norm": 0.37678319215774536, "learning_rate": 8.519567998092712e-05, "loss": 0.0358, "step": 23110 }, { "grad_norm": 0.36136868596076965, "learning_rate": 8.518099652668075e-05, "loss": 0.0286, "step": 23120 }, { "grad_norm": 0.2445937693119049, "learning_rate": 8.516630706096429e-05, "loss": 0.0249, "step": 23130 }, { "grad_norm": 0.31996339559555054, "learning_rate": 8.515161158628773e-05, "loss": 0.0223, "step": 23140 }, { "grad_norm": 0.5412515997886658, "learning_rate": 8.513691010516216e-05, "loss": 0.0447, "step": 23150 }, { "grad_norm": 0.2347996085882187, "learning_rate": 8.512220262009966e-05, "loss": 0.0292, "step": 23160 }, { "grad_norm": 0.2374412566423416, "learning_rate": 8.510748913361332e-05, "loss": 0.0252, "step": 23170 }, { "grad_norm": 0.4192468822002411, "learning_rate": 8.509276964821726e-05, "loss": 0.0204, "step": 23180 }, { "grad_norm": 0.4006207585334778, "learning_rate": 8.507804416642669e-05, "loss": 0.0276, "step": 23190 }, { "grad_norm": 0.35659337043762207, "learning_rate": 8.506331269075774e-05, "loss": 0.0254, "step": 23200 }, { "grad_norm": 0.2612816393375397, "learning_rate": 8.504857522372765e-05, "loss": 0.0264, "step": 23210 }, { "grad_norm": 0.29446661472320557, "learning_rate": 8.503383176785461e-05, "loss": 0.0334, "step": 23220 }, { "grad_norm": 0.3071539103984833, "learning_rate": 8.501908232565792e-05, "loss": 0.0335, "step": 23230 }, { "grad_norm": 0.4022326171398163, "learning_rate": 8.50043268996578e-05, "loss": 0.0244, "step": 23240 }, { "grad_norm": 0.4133915901184082, "learning_rate": 8.498956549237562e-05, "loss": 0.0312, "step": 23250 }, { "grad_norm": 0.2860296666622162, "learning_rate": 8.497479810633366e-05, "loss": 0.029, "step": 23260 }, { "grad_norm": 0.3550276458263397, "learning_rate": 8.496002474405525e-05, "loss": 0.0295, "step": 23270 }, { "grad_norm": 0.28161853551864624, "learning_rate": 8.494524540806478e-05, "loss": 0.0305, "step": 23280 }, { "grad_norm": 0.25672611594200134, "learning_rate": 8.493046010088761e-05, "loss": 0.0346, "step": 23290 }, { "grad_norm": 0.5181224942207336, "learning_rate": 8.491566882505018e-05, "loss": 0.029, "step": 23300 }, { "grad_norm": 0.25320616364479065, "learning_rate": 8.490087158307988e-05, "loss": 0.0325, "step": 23310 }, { "grad_norm": 0.6173479557037354, "learning_rate": 8.488606837750518e-05, "loss": 0.0288, "step": 23320 }, { "grad_norm": 0.25434425473213196, "learning_rate": 8.487125921085552e-05, "loss": 0.0283, "step": 23330 }, { "grad_norm": 0.28857770562171936, "learning_rate": 8.485644408566141e-05, "loss": 0.0321, "step": 23340 }, { "grad_norm": 0.4247199594974518, "learning_rate": 8.484162300445431e-05, "loss": 0.0293, "step": 23350 }, { "grad_norm": 0.4298948347568512, "learning_rate": 8.482679596976676e-05, "loss": 0.026, "step": 23360 }, { "grad_norm": 0.31821271777153015, "learning_rate": 8.48119629841323e-05, "loss": 0.034, "step": 23370 }, { "grad_norm": 0.28272557258605957, "learning_rate": 8.479712405008547e-05, "loss": 0.0343, "step": 23380 }, { "grad_norm": 0.28232306241989136, "learning_rate": 8.478227917016184e-05, "loss": 0.0361, "step": 23390 }, { "grad_norm": 0.4655596613883972, "learning_rate": 8.476742834689801e-05, "loss": 0.0376, "step": 23400 }, { "grad_norm": 0.3412933349609375, "learning_rate": 8.475257158283157e-05, "loss": 0.0343, "step": 23410 }, { "grad_norm": 0.31690672039985657, "learning_rate": 8.473770888050112e-05, "loss": 0.028, "step": 23420 }, { "grad_norm": 0.2542780041694641, "learning_rate": 8.47228402424463e-05, "loss": 0.0327, "step": 23430 }, { "grad_norm": 0.5324639081954956, "learning_rate": 8.470796567120775e-05, "loss": 0.0319, "step": 23440 }, { "grad_norm": 0.4418598711490631, "learning_rate": 8.469308516932714e-05, "loss": 0.0308, "step": 23450 }, { "grad_norm": 0.3287464678287506, "learning_rate": 8.467819873934714e-05, "loss": 0.0251, "step": 23460 }, { "grad_norm": 0.2512708604335785, "learning_rate": 8.466330638381143e-05, "loss": 0.0308, "step": 23470 }, { "grad_norm": 0.20019584894180298, "learning_rate": 8.464840810526469e-05, "loss": 0.0245, "step": 23480 }, { "grad_norm": 0.24617430567741394, "learning_rate": 8.463350390625264e-05, "loss": 0.0283, "step": 23490 }, { "grad_norm": 0.26260823011398315, "learning_rate": 8.4618593789322e-05, "loss": 0.0394, "step": 23500 }, { "grad_norm": 0.5648883581161499, "learning_rate": 8.46036777570205e-05, "loss": 0.0371, "step": 23510 }, { "grad_norm": 0.2812351882457733, "learning_rate": 8.458875581189688e-05, "loss": 0.0481, "step": 23520 }, { "grad_norm": 0.2639346420764923, "learning_rate": 8.457382795650092e-05, "loss": 0.0291, "step": 23530 }, { "grad_norm": 0.32201501727104187, "learning_rate": 8.455889419338335e-05, "loss": 0.0337, "step": 23540 }, { "grad_norm": 0.2296740710735321, "learning_rate": 8.454395452509593e-05, "loss": 0.0303, "step": 23550 }, { "grad_norm": 0.3174663484096527, "learning_rate": 8.452900895419146e-05, "loss": 0.0302, "step": 23560 }, { "grad_norm": 0.43670541048049927, "learning_rate": 8.451405748322376e-05, "loss": 0.0336, "step": 23570 }, { "grad_norm": 0.35042500495910645, "learning_rate": 8.449910011474759e-05, "loss": 0.0303, "step": 23580 }, { "grad_norm": 0.31146296858787537, "learning_rate": 8.448413685131876e-05, "loss": 0.0375, "step": 23590 }, { "grad_norm": 0.32308652997016907, "learning_rate": 8.446916769549407e-05, "loss": 0.038, "step": 23600 }, { "grad_norm": 0.2788418233394623, "learning_rate": 8.445419264983136e-05, "loss": 0.028, "step": 23610 }, { "grad_norm": 0.2346746325492859, "learning_rate": 8.443921171688947e-05, "loss": 0.0256, "step": 23620 }, { "grad_norm": 0.2798503637313843, "learning_rate": 8.442422489922819e-05, "loss": 0.0278, "step": 23630 }, { "grad_norm": 0.6774947643280029, "learning_rate": 8.440923219940838e-05, "loss": 0.0306, "step": 23640 }, { "grad_norm": 0.317017138004303, "learning_rate": 8.439423361999189e-05, "loss": 0.0325, "step": 23650 }, { "grad_norm": 0.2777263820171356, "learning_rate": 8.437922916354155e-05, "loss": 0.029, "step": 23660 }, { "grad_norm": 0.34381672739982605, "learning_rate": 8.436421883262123e-05, "loss": 0.0283, "step": 23670 }, { "grad_norm": 0.3645610213279724, "learning_rate": 8.434920262979577e-05, "loss": 0.0398, "step": 23680 }, { "grad_norm": 0.2700718939304352, "learning_rate": 8.433418055763104e-05, "loss": 0.0252, "step": 23690 }, { "grad_norm": 0.593806266784668, "learning_rate": 8.431915261869389e-05, "loss": 0.0314, "step": 23700 }, { "grad_norm": 0.3930681049823761, "learning_rate": 8.43041188155522e-05, "loss": 0.0299, "step": 23710 }, { "grad_norm": 0.24833552539348602, "learning_rate": 8.428907915077481e-05, "loss": 0.0322, "step": 23720 }, { "grad_norm": 0.27634888887405396, "learning_rate": 8.42740336269316e-05, "loss": 0.0434, "step": 23730 }, { "grad_norm": 0.3237321972846985, "learning_rate": 8.425898224659345e-05, "loss": 0.0298, "step": 23740 }, { "grad_norm": 0.6152241826057434, "learning_rate": 8.42439250123322e-05, "loss": 0.0299, "step": 23750 }, { "grad_norm": 1.2218834161758423, "learning_rate": 8.422886192672076e-05, "loss": 0.0299, "step": 23760 }, { "grad_norm": 0.3474455773830414, "learning_rate": 8.421379299233297e-05, "loss": 0.0274, "step": 23770 }, { "grad_norm": 0.34744709730148315, "learning_rate": 8.419871821174371e-05, "loss": 0.0285, "step": 23780 }, { "grad_norm": 0.31949496269226074, "learning_rate": 8.418363758752884e-05, "loss": 0.0286, "step": 23790 }, { "grad_norm": 0.2060636579990387, "learning_rate": 8.416855112226523e-05, "loss": 0.0261, "step": 23800 }, { "grad_norm": 0.2536853551864624, "learning_rate": 8.415345881853075e-05, "loss": 0.0234, "step": 23810 }, { "grad_norm": 0.249368354678154, "learning_rate": 8.413836067890426e-05, "loss": 0.0233, "step": 23820 }, { "grad_norm": 0.31768134236335754, "learning_rate": 8.41232567059656e-05, "loss": 0.0233, "step": 23830 }, { "grad_norm": 1.734426498413086, "learning_rate": 8.410814690229565e-05, "loss": 0.0252, "step": 23840 }, { "grad_norm": 0.8275406956672668, "learning_rate": 8.409303127047626e-05, "loss": 0.023, "step": 23850 }, { "grad_norm": 0.33843517303466797, "learning_rate": 8.407790981309028e-05, "loss": 0.0257, "step": 23860 }, { "grad_norm": 0.4071366786956787, "learning_rate": 8.406278253272153e-05, "loss": 0.034, "step": 23870 }, { "grad_norm": 0.2681575417518616, "learning_rate": 8.404764943195487e-05, "loss": 0.0261, "step": 23880 }, { "grad_norm": 0.5319808125495911, "learning_rate": 8.403251051337613e-05, "loss": 0.0316, "step": 23890 }, { "grad_norm": 0.3170325458049774, "learning_rate": 8.401736577957214e-05, "loss": 0.0314, "step": 23900 }, { "grad_norm": 0.3695789873600006, "learning_rate": 8.40022152331307e-05, "loss": 0.0283, "step": 23910 }, { "grad_norm": 0.3397224247455597, "learning_rate": 8.398705887664064e-05, "loss": 0.0397, "step": 23920 }, { "grad_norm": 0.3795437812805176, "learning_rate": 8.397189671269177e-05, "loss": 0.0283, "step": 23930 }, { "grad_norm": 0.3457115888595581, "learning_rate": 8.395672874387488e-05, "loss": 0.0253, "step": 23940 }, { "grad_norm": 0.2745959460735321, "learning_rate": 8.394155497278177e-05, "loss": 0.0243, "step": 23950 }, { "grad_norm": 0.353927344083786, "learning_rate": 8.392637540200523e-05, "loss": 0.0465, "step": 23960 }, { "grad_norm": 0.2800356447696686, "learning_rate": 8.391119003413902e-05, "loss": 0.0316, "step": 23970 }, { "grad_norm": 0.35594287514686584, "learning_rate": 8.38959988717779e-05, "loss": 0.0285, "step": 23980 }, { "grad_norm": 0.3333165943622589, "learning_rate": 8.388080191751764e-05, "loss": 0.0318, "step": 23990 }, { "grad_norm": 0.3467680811882019, "learning_rate": 8.386559917395496e-05, "loss": 0.0287, "step": 24000 }, { "grad_norm": 0.27648550271987915, "learning_rate": 8.385039064368761e-05, "loss": 0.0231, "step": 24010 }, { "grad_norm": 0.23717062175273895, "learning_rate": 8.383517632931431e-05, "loss": 0.0274, "step": 24020 }, { "grad_norm": 0.4399145841598511, "learning_rate": 8.381995623343477e-05, "loss": 0.0244, "step": 24030 }, { "grad_norm": 0.2213096171617508, "learning_rate": 8.380473035864968e-05, "loss": 0.0292, "step": 24040 }, { "grad_norm": 0.20510610938072205, "learning_rate": 8.378949870756076e-05, "loss": 0.0275, "step": 24050 }, { "grad_norm": 0.3473796546459198, "learning_rate": 8.377426128277063e-05, "loss": 0.0295, "step": 24060 }, { "grad_norm": 0.2918533980846405, "learning_rate": 8.375901808688298e-05, "loss": 0.0241, "step": 24070 }, { "grad_norm": 0.3731593191623688, "learning_rate": 8.374376912250246e-05, "loss": 0.0279, "step": 24080 }, { "grad_norm": 0.35521799325942993, "learning_rate": 8.372851439223468e-05, "loss": 0.0274, "step": 24090 }, { "grad_norm": 0.3235905170440674, "learning_rate": 8.371325389868627e-05, "loss": 0.0305, "step": 24100 }, { "grad_norm": 0.24010980129241943, "learning_rate": 8.369798764446482e-05, "loss": 0.0279, "step": 24110 }, { "grad_norm": 0.2962825298309326, "learning_rate": 8.368271563217893e-05, "loss": 0.0229, "step": 24120 }, { "grad_norm": 0.2776317894458771, "learning_rate": 8.366743786443817e-05, "loss": 0.0366, "step": 24130 }, { "grad_norm": 0.3565210998058319, "learning_rate": 8.365215434385309e-05, "loss": 0.0313, "step": 24140 }, { "grad_norm": 0.28664785623550415, "learning_rate": 8.36368650730352e-05, "loss": 0.0295, "step": 24150 }, { "grad_norm": 0.27629899978637695, "learning_rate": 8.362157005459705e-05, "loss": 0.0282, "step": 24160 }, { "grad_norm": 0.3221476078033447, "learning_rate": 8.360626929115213e-05, "loss": 0.0284, "step": 24170 }, { "grad_norm": 0.24162203073501587, "learning_rate": 8.359096278531492e-05, "loss": 0.0263, "step": 24180 }, { "grad_norm": 0.35588088631629944, "learning_rate": 8.357565053970088e-05, "loss": 0.0259, "step": 24190 }, { "grad_norm": 0.36481571197509766, "learning_rate": 8.356033255692647e-05, "loss": 0.0234, "step": 24200 }, { "grad_norm": 1.8557193279266357, "learning_rate": 8.354500883960911e-05, "loss": 0.0261, "step": 24210 }, { "grad_norm": 0.2756928503513336, "learning_rate": 8.352967939036717e-05, "loss": 0.0328, "step": 24220 }, { "grad_norm": 0.33350351452827454, "learning_rate": 8.35143442118201e-05, "loss": 0.033, "step": 24230 }, { "grad_norm": 0.2934171259403229, "learning_rate": 8.349900330658819e-05, "loss": 0.026, "step": 24240 }, { "grad_norm": 0.3095951974391937, "learning_rate": 8.348365667729284e-05, "loss": 0.038, "step": 24250 }, { "grad_norm": 0.392260879278183, "learning_rate": 8.346830432655633e-05, "loss": 0.031, "step": 24260 }, { "grad_norm": 0.2942417860031128, "learning_rate": 8.345294625700195e-05, "loss": 0.0318, "step": 24270 }, { "grad_norm": 0.3256002366542816, "learning_rate": 8.343758247125402e-05, "loss": 0.0297, "step": 24280 }, { "grad_norm": 0.45162391662597656, "learning_rate": 8.342221297193776e-05, "loss": 0.0324, "step": 24290 }, { "grad_norm": 0.2296094000339508, "learning_rate": 8.34068377616794e-05, "loss": 0.032, "step": 24300 }, { "grad_norm": 0.4152502119541168, "learning_rate": 8.339145684310615e-05, "loss": 0.0288, "step": 24310 }, { "grad_norm": 0.2718745172023773, "learning_rate": 8.337607021884618e-05, "loss": 0.0278, "step": 24320 }, { "grad_norm": 0.28531405329704285, "learning_rate": 8.336067789152867e-05, "loss": 0.0221, "step": 24330 }, { "grad_norm": 0.34211382269859314, "learning_rate": 8.334527986378369e-05, "loss": 0.0295, "step": 24340 }, { "grad_norm": 0.37170642614364624, "learning_rate": 8.332987613824239e-05, "loss": 0.0281, "step": 24350 }, { "grad_norm": 0.26453858613967896, "learning_rate": 8.331446671753685e-05, "loss": 0.0238, "step": 24360 }, { "grad_norm": 0.3009037673473358, "learning_rate": 8.329905160430007e-05, "loss": 0.0234, "step": 24370 }, { "grad_norm": 0.3225686848163605, "learning_rate": 8.328363080116611e-05, "loss": 0.0244, "step": 24380 }, { "grad_norm": 0.6798415780067444, "learning_rate": 8.326820431076997e-05, "loss": 0.0305, "step": 24390 }, { "grad_norm": 0.2902754545211792, "learning_rate": 8.325277213574759e-05, "loss": 0.03, "step": 24400 }, { "grad_norm": 0.25590479373931885, "learning_rate": 8.32373342787359e-05, "loss": 0.0284, "step": 24410 }, { "grad_norm": 0.3683156967163086, "learning_rate": 8.322189074237285e-05, "loss": 0.0288, "step": 24420 }, { "grad_norm": 0.41287973523139954, "learning_rate": 8.32064415292973e-05, "loss": 0.0293, "step": 24430 }, { "grad_norm": 0.22209957242012024, "learning_rate": 8.319098664214907e-05, "loss": 0.0228, "step": 24440 }, { "grad_norm": 0.35926496982574463, "learning_rate": 8.3175526083569e-05, "loss": 0.0283, "step": 24450 }, { "grad_norm": 0.4346022605895996, "learning_rate": 8.316005985619889e-05, "loss": 0.0314, "step": 24460 }, { "grad_norm": 0.24459174275398254, "learning_rate": 8.314458796268147e-05, "loss": 0.0324, "step": 24470 }, { "grad_norm": 0.2472396045923233, "learning_rate": 8.312911040566047e-05, "loss": 0.0273, "step": 24480 }, { "grad_norm": 0.2872833013534546, "learning_rate": 8.31136271877806e-05, "loss": 0.0291, "step": 24490 }, { "grad_norm": 0.32901012897491455, "learning_rate": 8.309813831168748e-05, "loss": 0.032, "step": 24500 }, { "grad_norm": 0.3317587077617645, "learning_rate": 8.308264378002777e-05, "loss": 0.0253, "step": 24510 }, { "grad_norm": 0.4188922345638275, "learning_rate": 8.306714359544906e-05, "loss": 0.0316, "step": 24520 }, { "grad_norm": 0.3588489294052124, "learning_rate": 8.30516377605999e-05, "loss": 0.0297, "step": 24530 }, { "grad_norm": 0.43958792090415955, "learning_rate": 8.30361262781298e-05, "loss": 0.0245, "step": 24540 }, { "grad_norm": 0.398062527179718, "learning_rate": 8.302060915068924e-05, "loss": 0.0303, "step": 24550 }, { "grad_norm": 1.020459532737732, "learning_rate": 8.300508638092972e-05, "loss": 0.0339, "step": 24560 }, { "grad_norm": 0.40663936734199524, "learning_rate": 8.298955797150361e-05, "loss": 0.0379, "step": 24570 }, { "grad_norm": 0.39256516098976135, "learning_rate": 8.297402392506433e-05, "loss": 0.0318, "step": 24580 }, { "grad_norm": 0.3880354166030884, "learning_rate": 8.295848424426617e-05, "loss": 0.0295, "step": 24590 }, { "grad_norm": 0.2376030683517456, "learning_rate": 8.29429389317645e-05, "loss": 0.0256, "step": 24600 }, { "grad_norm": 0.2947590947151184, "learning_rate": 8.292738799021556e-05, "loss": 0.033, "step": 24610 }, { "grad_norm": 0.3953455090522766, "learning_rate": 8.291183142227656e-05, "loss": 0.0267, "step": 24620 }, { "grad_norm": 0.4293447434902191, "learning_rate": 8.289626923060572e-05, "loss": 0.0323, "step": 24630 }, { "grad_norm": 0.2511711120605469, "learning_rate": 8.288070141786218e-05, "loss": 0.0284, "step": 24640 }, { "grad_norm": 0.3428215980529785, "learning_rate": 8.286512798670605e-05, "loss": 0.0339, "step": 24650 }, { "grad_norm": 0.2589489817619324, "learning_rate": 8.284954893979842e-05, "loss": 0.0313, "step": 24660 }, { "grad_norm": 0.4129243493080139, "learning_rate": 8.283396427980131e-05, "loss": 0.0394, "step": 24670 }, { "grad_norm": 0.2767796218395233, "learning_rate": 8.281837400937771e-05, "loss": 0.0313, "step": 24680 }, { "grad_norm": 0.37372153997421265, "learning_rate": 8.28027781311916e-05, "loss": 0.0302, "step": 24690 }, { "grad_norm": 0.39318612217903137, "learning_rate": 8.278717664790785e-05, "loss": 0.0318, "step": 24700 }, { "grad_norm": 0.359208345413208, "learning_rate": 8.277156956219234e-05, "loss": 0.0311, "step": 24710 }, { "grad_norm": 0.2986433804035187, "learning_rate": 8.275595687671189e-05, "loss": 0.0317, "step": 24720 }, { "grad_norm": 0.2994198501110077, "learning_rate": 8.27403385941343e-05, "loss": 0.0316, "step": 24730 }, { "grad_norm": 0.26568034291267395, "learning_rate": 8.272471471712828e-05, "loss": 0.0322, "step": 24740 }, { "grad_norm": 0.26345446705818176, "learning_rate": 8.270908524836355e-05, "loss": 0.0328, "step": 24750 }, { "grad_norm": 0.25925377011299133, "learning_rate": 8.269345019051074e-05, "loss": 0.0308, "step": 24760 }, { "grad_norm": 1.9215036630630493, "learning_rate": 8.267780954624147e-05, "loss": 0.028, "step": 24770 }, { "grad_norm": 0.23128986358642578, "learning_rate": 8.266216331822827e-05, "loss": 0.031, "step": 24780 }, { "grad_norm": 0.24503587186336517, "learning_rate": 8.264651150914469e-05, "loss": 0.0298, "step": 24790 }, { "grad_norm": 0.3470582962036133, "learning_rate": 8.263085412166517e-05, "loss": 0.0256, "step": 24800 }, { "grad_norm": 0.21547235548496246, "learning_rate": 8.261519115846514e-05, "loss": 0.0312, "step": 24810 }, { "grad_norm": 0.2976394295692444, "learning_rate": 8.259952262222096e-05, "loss": 0.0286, "step": 24820 }, { "grad_norm": 0.2530001103878021, "learning_rate": 8.258384851560997e-05, "loss": 0.0263, "step": 24830 }, { "grad_norm": 0.2712734341621399, "learning_rate": 8.256816884131044e-05, "loss": 0.0296, "step": 24840 }, { "grad_norm": 0.28353703022003174, "learning_rate": 8.255248360200159e-05, "loss": 0.0224, "step": 24850 }, { "grad_norm": 0.3079797923564911, "learning_rate": 8.253679280036359e-05, "loss": 0.028, "step": 24860 }, { "grad_norm": 0.2875131368637085, "learning_rate": 8.252109643907762e-05, "loss": 0.0357, "step": 24870 }, { "grad_norm": 0.2738398313522339, "learning_rate": 8.250539452082569e-05, "loss": 0.0239, "step": 24880 }, { "grad_norm": 0.26217901706695557, "learning_rate": 8.248968704829087e-05, "loss": 0.0249, "step": 24890 }, { "grad_norm": 0.3527447283267975, "learning_rate": 8.247397402415714e-05, "loss": 0.0302, "step": 24900 }, { "grad_norm": 0.268315851688385, "learning_rate": 8.24582554511094e-05, "loss": 0.023, "step": 24910 }, { "grad_norm": 0.31016939878463745, "learning_rate": 8.244253133183355e-05, "loss": 0.0208, "step": 24920 }, { "grad_norm": 0.259391725063324, "learning_rate": 8.24268016690164e-05, "loss": 0.0252, "step": 24930 }, { "grad_norm": 0.29620546102523804, "learning_rate": 8.241106646534571e-05, "loss": 0.0266, "step": 24940 }, { "grad_norm": 0.3421007990837097, "learning_rate": 8.23953257235102e-05, "loss": 0.0282, "step": 24950 }, { "grad_norm": 0.3063160479068756, "learning_rate": 8.237957944619956e-05, "loss": 0.0246, "step": 24960 }, { "grad_norm": 0.2765139639377594, "learning_rate": 8.236382763610437e-05, "loss": 0.0247, "step": 24970 }, { "grad_norm": 0.3326495289802551, "learning_rate": 8.234807029591619e-05, "loss": 0.0228, "step": 24980 }, { "grad_norm": 0.3169604539871216, "learning_rate": 8.233230742832752e-05, "loss": 0.0372, "step": 24990 }, { "grad_norm": 0.2825509309768677, "learning_rate": 8.231653903603178e-05, "loss": 0.0283, "step": 25000 }, { "grad_norm": 0.3435817360877991, "learning_rate": 8.23007651217234e-05, "loss": 0.029, "step": 25010 }, { "grad_norm": 0.3922729790210724, "learning_rate": 8.228498568809769e-05, "loss": 0.0318, "step": 25020 }, { "grad_norm": 0.3764961361885071, "learning_rate": 8.22692007378509e-05, "loss": 0.0379, "step": 25030 }, { "grad_norm": 1.9496010541915894, "learning_rate": 8.225341027368028e-05, "loss": 0.0244, "step": 25040 }, { "grad_norm": 0.32344546914100647, "learning_rate": 8.223761429828399e-05, "loss": 0.0299, "step": 25050 }, { "grad_norm": 0.2329665571451187, "learning_rate": 8.22218128143611e-05, "loss": 0.0276, "step": 25060 }, { "grad_norm": 0.2599218785762787, "learning_rate": 8.220600582461166e-05, "loss": 0.0247, "step": 25070 }, { "grad_norm": 0.2934930920600891, "learning_rate": 8.219019333173668e-05, "loss": 0.0279, "step": 25080 }, { "grad_norm": 0.3362848460674286, "learning_rate": 8.217437533843805e-05, "loss": 0.0333, "step": 25090 }, { "grad_norm": 0.3043251931667328, "learning_rate": 8.215855184741867e-05, "loss": 0.028, "step": 25100 }, { "grad_norm": 0.4616616368293762, "learning_rate": 8.21427228613823e-05, "loss": 0.0275, "step": 25110 }, { "grad_norm": 0.37917107343673706, "learning_rate": 8.21268883830337e-05, "loss": 0.0275, "step": 25120 }, { "grad_norm": 0.3132621943950653, "learning_rate": 8.211104841507855e-05, "loss": 0.0403, "step": 25130 }, { "grad_norm": 0.23133544623851776, "learning_rate": 8.209520296022346e-05, "loss": 0.0273, "step": 25140 }, { "grad_norm": 0.6968553066253662, "learning_rate": 8.207935202117599e-05, "loss": 0.0305, "step": 25150 }, { "grad_norm": 0.2407916933298111, "learning_rate": 8.206349560064463e-05, "loss": 0.0276, "step": 25160 }, { "grad_norm": 0.392875611782074, "learning_rate": 8.204763370133881e-05, "loss": 0.0262, "step": 25170 }, { "grad_norm": 0.27904921770095825, "learning_rate": 8.203176632596892e-05, "loss": 0.028, "step": 25180 }, { "grad_norm": 0.2592708468437195, "learning_rate": 8.20158934772462e-05, "loss": 0.0339, "step": 25190 }, { "grad_norm": 0.31054216623306274, "learning_rate": 8.200001515788294e-05, "loss": 0.0246, "step": 25200 }, { "grad_norm": 0.2108478546142578, "learning_rate": 8.198413137059228e-05, "loss": 0.0243, "step": 25210 }, { "grad_norm": 0.42300209403038025, "learning_rate": 8.196824211808835e-05, "loss": 0.0292, "step": 25220 }, { "grad_norm": 0.2830105423927307, "learning_rate": 8.195234740308617e-05, "loss": 0.0281, "step": 25230 }, { "grad_norm": 0.2729164659976959, "learning_rate": 8.193644722830171e-05, "loss": 0.0216, "step": 25240 }, { "grad_norm": 0.21694284677505493, "learning_rate": 8.19205415964519e-05, "loss": 0.0247, "step": 25250 }, { "grad_norm": 0.20647576451301575, "learning_rate": 8.190463051025456e-05, "loss": 0.0215, "step": 25260 }, { "grad_norm": 0.45976370573043823, "learning_rate": 8.188871397242843e-05, "loss": 0.0261, "step": 25270 }, { "grad_norm": 0.38568514585494995, "learning_rate": 8.187279198569326e-05, "loss": 0.0283, "step": 25280 }, { "grad_norm": 0.8539336323738098, "learning_rate": 8.185686455276966e-05, "loss": 0.0234, "step": 25290 }, { "grad_norm": 0.26280808448791504, "learning_rate": 8.184093167637921e-05, "loss": 0.0323, "step": 25300 }, { "grad_norm": 0.42044302821159363, "learning_rate": 8.182499335924437e-05, "loss": 0.0275, "step": 25310 }, { "grad_norm": 0.3912908136844635, "learning_rate": 8.18090496040886e-05, "loss": 0.0267, "step": 25320 }, { "grad_norm": 0.32047978043556213, "learning_rate": 8.179310041363621e-05, "loss": 0.03, "step": 25330 }, { "grad_norm": 0.3118187189102173, "learning_rate": 8.17771457906125e-05, "loss": 0.0258, "step": 25340 }, { "grad_norm": 0.31964802742004395, "learning_rate": 8.176118573774371e-05, "loss": 0.0245, "step": 25350 }, { "grad_norm": 0.2794150114059448, "learning_rate": 8.174522025775692e-05, "loss": 0.0251, "step": 25360 }, { "grad_norm": 0.5411961078643799, "learning_rate": 8.172924935338022e-05, "loss": 0.0318, "step": 25370 }, { "grad_norm": 0.28937581181526184, "learning_rate": 8.171327302734262e-05, "loss": 0.0255, "step": 25380 }, { "grad_norm": 0.4082184433937073, "learning_rate": 8.169729128237401e-05, "loss": 0.0257, "step": 25390 }, { "grad_norm": 0.6626333594322205, "learning_rate": 8.168130412120525e-05, "loss": 0.0266, "step": 25400 }, { "grad_norm": 0.38228532671928406, "learning_rate": 8.16653115465681e-05, "loss": 0.0279, "step": 25410 }, { "grad_norm": 0.3020946681499481, "learning_rate": 8.164931356119526e-05, "loss": 0.0293, "step": 25420 }, { "grad_norm": 5.302785396575928, "learning_rate": 8.163331016782032e-05, "loss": 0.0301, "step": 25430 }, { "grad_norm": 0.27916428446769714, "learning_rate": 8.161730136917785e-05, "loss": 0.0259, "step": 25440 }, { "grad_norm": 0.5686575770378113, "learning_rate": 8.160128716800333e-05, "loss": 0.0359, "step": 25450 }, { "grad_norm": 0.31908875703811646, "learning_rate": 8.158526756703313e-05, "loss": 0.027, "step": 25460 }, { "grad_norm": 0.31180649995803833, "learning_rate": 8.156924256900455e-05, "loss": 0.0338, "step": 25470 }, { "grad_norm": 0.26859140396118164, "learning_rate": 8.155321217665584e-05, "loss": 0.0257, "step": 25480 }, { "grad_norm": 0.2305227518081665, "learning_rate": 8.153717639272614e-05, "loss": 0.0223, "step": 25490 }, { "grad_norm": 0.3158667981624603, "learning_rate": 8.152113521995555e-05, "loss": 0.0282, "step": 25500 }, { "grad_norm": 0.21616265177726746, "learning_rate": 8.150508866108505e-05, "loss": 0.0308, "step": 25510 }, { "grad_norm": 0.2548888623714447, "learning_rate": 8.148903671885657e-05, "loss": 0.0258, "step": 25520 }, { "grad_norm": 0.2890022099018097, "learning_rate": 8.147297939601292e-05, "loss": 0.0334, "step": 25530 }, { "grad_norm": 0.2646428048610687, "learning_rate": 8.145691669529792e-05, "loss": 0.0299, "step": 25540 }, { "grad_norm": 0.3459462821483612, "learning_rate": 8.144084861945618e-05, "loss": 0.032, "step": 25550 }, { "grad_norm": 1.7980154752731323, "learning_rate": 8.142477517123333e-05, "loss": 0.0354, "step": 25560 }, { "grad_norm": 0.29955625534057617, "learning_rate": 8.140869635337586e-05, "loss": 0.0258, "step": 25570 }, { "grad_norm": 0.2838893532752991, "learning_rate": 8.139261216863123e-05, "loss": 0.0257, "step": 25580 }, { "grad_norm": 0.2410224974155426, "learning_rate": 8.137652261974776e-05, "loss": 0.0225, "step": 25590 }, { "grad_norm": 0.31623879075050354, "learning_rate": 8.136042770947472e-05, "loss": 0.0258, "step": 25600 }, { "grad_norm": 0.24994878470897675, "learning_rate": 8.134432744056228e-05, "loss": 0.0314, "step": 25610 }, { "grad_norm": 0.3251395523548126, "learning_rate": 8.132822181576158e-05, "loss": 0.0314, "step": 25620 }, { "grad_norm": 0.34858059883117676, "learning_rate": 8.131211083782459e-05, "loss": 0.0297, "step": 25630 }, { "grad_norm": 0.40199747681617737, "learning_rate": 8.129599450950424e-05, "loss": 0.03, "step": 25640 }, { "grad_norm": 0.4241131544113159, "learning_rate": 8.127987283355438e-05, "loss": 0.0322, "step": 25650 }, { "grad_norm": 0.4312482476234436, "learning_rate": 8.126374581272976e-05, "loss": 0.0289, "step": 25660 }, { "grad_norm": 0.7624138593673706, "learning_rate": 8.124761344978605e-05, "loss": 0.04, "step": 25670 }, { "grad_norm": 0.334247887134552, "learning_rate": 8.12314757474798e-05, "loss": 0.0266, "step": 25680 }, { "grad_norm": 0.30726945400238037, "learning_rate": 8.121533270856856e-05, "loss": 0.0254, "step": 25690 }, { "grad_norm": 0.27132704854011536, "learning_rate": 8.119918433581069e-05, "loss": 0.0267, "step": 25700 }, { "grad_norm": 0.3220958113670349, "learning_rate": 8.118303063196551e-05, "loss": 0.0263, "step": 25710 }, { "grad_norm": 0.3241426646709442, "learning_rate": 8.116687159979326e-05, "loss": 0.0245, "step": 25720 }, { "grad_norm": 0.3257962465286255, "learning_rate": 8.115070724205508e-05, "loss": 0.0286, "step": 25730 }, { "grad_norm": 0.28535494208335876, "learning_rate": 8.113453756151296e-05, "loss": 0.0252, "step": 25740 }, { "grad_norm": 0.3448971211910248, "learning_rate": 8.111836256092995e-05, "loss": 0.0221, "step": 25750 }, { "grad_norm": 0.7423998117446899, "learning_rate": 8.110218224306985e-05, "loss": 0.0285, "step": 25760 }, { "grad_norm": 0.24165146052837372, "learning_rate": 8.108599661069745e-05, "loss": 0.0248, "step": 25770 }, { "grad_norm": 0.41174033284187317, "learning_rate": 8.106980566657845e-05, "loss": 0.0214, "step": 25780 }, { "grad_norm": 0.22850404679775238, "learning_rate": 8.10536094134794e-05, "loss": 0.0282, "step": 25790 }, { "grad_norm": 0.43440866470336914, "learning_rate": 8.103740785416783e-05, "loss": 0.0274, "step": 25800 }, { "grad_norm": 0.2497391551733017, "learning_rate": 8.102120099141212e-05, "loss": 0.0266, "step": 25810 }, { "grad_norm": 0.22366654872894287, "learning_rate": 8.100498882798163e-05, "loss": 0.0209, "step": 25820 }, { "grad_norm": 0.30295366048812866, "learning_rate": 8.09887713666465e-05, "loss": 0.0217, "step": 25830 }, { "grad_norm": 0.4951333999633789, "learning_rate": 8.09725486101779e-05, "loss": 0.0284, "step": 25840 }, { "grad_norm": 0.4792473614215851, "learning_rate": 8.095632056134784e-05, "loss": 0.0282, "step": 25850 }, { "grad_norm": 0.4317879378795624, "learning_rate": 8.094008722292925e-05, "loss": 0.0263, "step": 25860 }, { "grad_norm": 0.18734145164489746, "learning_rate": 8.092384859769598e-05, "loss": 0.0277, "step": 25870 }, { "grad_norm": 0.27150335907936096, "learning_rate": 8.090760468842275e-05, "loss": 0.0348, "step": 25880 }, { "grad_norm": 0.43488261103630066, "learning_rate": 8.089135549788521e-05, "loss": 0.0324, "step": 25890 }, { "grad_norm": 0.3466068208217621, "learning_rate": 8.087510102885987e-05, "loss": 0.0268, "step": 25900 }, { "grad_norm": 0.23545561730861664, "learning_rate": 8.085884128412422e-05, "loss": 0.0296, "step": 25910 }, { "grad_norm": 0.37607479095458984, "learning_rate": 8.084257626645659e-05, "loss": 0.0281, "step": 25920 }, { "grad_norm": 0.24013318121433258, "learning_rate": 8.08263059786362e-05, "loss": 0.0244, "step": 25930 }, { "grad_norm": 0.43207982182502747, "learning_rate": 8.081003042344325e-05, "loss": 0.0228, "step": 25940 }, { "grad_norm": 0.24455873668193817, "learning_rate": 8.079374960365872e-05, "loss": 0.0244, "step": 25950 }, { "grad_norm": 0.4325748085975647, "learning_rate": 8.077746352206463e-05, "loss": 0.0348, "step": 25960 }, { "grad_norm": 0.37112531065940857, "learning_rate": 8.076117218144377e-05, "loss": 0.0241, "step": 25970 }, { "grad_norm": 0.3518610894680023, "learning_rate": 8.074487558457991e-05, "loss": 0.0263, "step": 25980 }, { "grad_norm": 0.45614948868751526, "learning_rate": 8.072857373425768e-05, "loss": 0.0262, "step": 25990 }, { "grad_norm": 0.34660598635673523, "learning_rate": 8.071226663326264e-05, "loss": 0.0306, "step": 26000 }, { "grad_norm": 0.3042421042919159, "learning_rate": 8.069595428438121e-05, "loss": 0.0394, "step": 26010 }, { "grad_norm": 0.7260030508041382, "learning_rate": 8.067963669040072e-05, "loss": 0.031, "step": 26020 }, { "grad_norm": 0.4391445219516754, "learning_rate": 8.066331385410942e-05, "loss": 0.0321, "step": 26030 }, { "grad_norm": 0.25314000248908997, "learning_rate": 8.064698577829641e-05, "loss": 0.0297, "step": 26040 }, { "grad_norm": 0.22108101844787598, "learning_rate": 8.063065246575175e-05, "loss": 0.0301, "step": 26050 }, { "grad_norm": 0.35896536707878113, "learning_rate": 8.061431391926631e-05, "loss": 0.0302, "step": 26060 }, { "grad_norm": 0.2391963005065918, "learning_rate": 8.059797014163195e-05, "loss": 0.0284, "step": 26070 }, { "grad_norm": 0.2975029945373535, "learning_rate": 8.058162113564133e-05, "loss": 0.0416, "step": 26080 }, { "grad_norm": 0.35970550775527954, "learning_rate": 8.056526690408806e-05, "loss": 0.0325, "step": 26090 }, { "grad_norm": 0.31268510222435, "learning_rate": 8.054890744976666e-05, "loss": 0.0277, "step": 26100 }, { "grad_norm": 0.3295165002346039, "learning_rate": 8.053254277547248e-05, "loss": 0.0249, "step": 26110 }, { "grad_norm": 0.32978567481040955, "learning_rate": 8.051617288400182e-05, "loss": 0.0279, "step": 26120 }, { "grad_norm": 0.5210772752761841, "learning_rate": 8.049979777815182e-05, "loss": 0.0299, "step": 26130 }, { "grad_norm": 0.29212358593940735, "learning_rate": 8.048341746072054e-05, "loss": 0.0255, "step": 26140 }, { "grad_norm": 0.41633152961730957, "learning_rate": 8.046703193450696e-05, "loss": 0.0353, "step": 26150 }, { "grad_norm": 0.29655957221984863, "learning_rate": 8.04506412023109e-05, "loss": 0.0288, "step": 26160 }, { "grad_norm": 0.7298606038093567, "learning_rate": 8.043424526693306e-05, "loss": 0.0367, "step": 26170 }, { "grad_norm": 0.6404273509979248, "learning_rate": 8.04178441311751e-05, "loss": 0.0356, "step": 26180 }, { "grad_norm": 0.4488683342933655, "learning_rate": 8.04014377978395e-05, "loss": 0.0365, "step": 26190 }, { "grad_norm": 0.2693946957588196, "learning_rate": 8.038502626972967e-05, "loss": 0.0338, "step": 26200 }, { "grad_norm": 0.2869097888469696, "learning_rate": 8.036860954964989e-05, "loss": 0.0292, "step": 26210 }, { "grad_norm": 0.32631996273994446, "learning_rate": 8.035218764040531e-05, "loss": 0.0295, "step": 26220 }, { "grad_norm": 0.30758723616600037, "learning_rate": 8.033576054480199e-05, "loss": 0.0267, "step": 26230 }, { "grad_norm": 0.30875900387763977, "learning_rate": 8.031932826564688e-05, "loss": 0.0222, "step": 26240 }, { "grad_norm": 0.2366255521774292, "learning_rate": 8.030289080574782e-05, "loss": 0.0306, "step": 26250 }, { "grad_norm": 0.28043845295906067, "learning_rate": 8.028644816791349e-05, "loss": 0.038, "step": 26260 }, { "grad_norm": 0.8056156635284424, "learning_rate": 8.027000035495351e-05, "loss": 0.027, "step": 26270 }, { "grad_norm": 0.30714520812034607, "learning_rate": 8.025354736967836e-05, "loss": 0.041, "step": 26280 }, { "grad_norm": 2.199528217315674, "learning_rate": 8.023708921489941e-05, "loss": 0.033, "step": 26290 }, { "grad_norm": 0.28273940086364746, "learning_rate": 8.022062589342887e-05, "loss": 0.0319, "step": 26300 }, { "grad_norm": 0.27214229106903076, "learning_rate": 8.020415740807993e-05, "loss": 0.0351, "step": 26310 }, { "grad_norm": 0.2746025621891022, "learning_rate": 8.018768376166656e-05, "loss": 0.0252, "step": 26320 }, { "grad_norm": 0.28363677859306335, "learning_rate": 8.017120495700368e-05, "loss": 0.0236, "step": 26330 }, { "grad_norm": 0.8217446208000183, "learning_rate": 8.015472099690704e-05, "loss": 0.0223, "step": 26340 }, { "grad_norm": 0.247731551527977, "learning_rate": 8.013823188419332e-05, "loss": 0.0242, "step": 26350 }, { "grad_norm": 0.3198651373386383, "learning_rate": 8.012173762168006e-05, "loss": 0.0271, "step": 26360 }, { "grad_norm": 0.3017612397670746, "learning_rate": 8.010523821218567e-05, "loss": 0.0242, "step": 26370 }, { "grad_norm": 0.37631481885910034, "learning_rate": 8.008873365852945e-05, "loss": 0.031, "step": 26380 }, { "grad_norm": 0.2989370822906494, "learning_rate": 8.007222396353157e-05, "loss": 0.0283, "step": 26390 }, { "grad_norm": 0.28495872020721436, "learning_rate": 8.00557091300131e-05, "loss": 0.0269, "step": 26400 }, { "grad_norm": 0.2932027578353882, "learning_rate": 8.003918916079597e-05, "loss": 0.025, "step": 26410 }, { "grad_norm": 0.38741472363471985, "learning_rate": 8.002266405870298e-05, "loss": 0.0272, "step": 26420 }, { "grad_norm": 0.3434954881668091, "learning_rate": 8.000613382655782e-05, "loss": 0.031, "step": 26430 }, { "grad_norm": 0.3083735406398773, "learning_rate": 7.998959846718505e-05, "loss": 0.0313, "step": 26440 }, { "grad_norm": 0.2188553661108017, "learning_rate": 7.997305798341012e-05, "loss": 0.0271, "step": 26450 }, { "grad_norm": 0.4514496922492981, "learning_rate": 7.995651237805937e-05, "loss": 0.0306, "step": 26460 }, { "grad_norm": 0.357628732919693, "learning_rate": 7.993996165395996e-05, "loss": 0.0254, "step": 26470 }, { "grad_norm": 0.2698664367198944, "learning_rate": 7.992340581393996e-05, "loss": 0.0237, "step": 26480 }, { "grad_norm": 0.24753913283348083, "learning_rate": 7.990684486082831e-05, "loss": 0.0252, "step": 26490 }, { "grad_norm": 0.40723302960395813, "learning_rate": 7.989027879745482e-05, "loss": 0.0282, "step": 26500 }, { "grad_norm": 0.24836967885494232, "learning_rate": 7.98737076266502e-05, "loss": 0.0351, "step": 26510 }, { "grad_norm": 1.367685079574585, "learning_rate": 7.985713135124598e-05, "loss": 0.0264, "step": 26520 }, { "grad_norm": 0.37063273787498474, "learning_rate": 7.98405499740746e-05, "loss": 0.0295, "step": 26530 }, { "grad_norm": 0.8238252997398376, "learning_rate": 7.98239634979694e-05, "loss": 0.0303, "step": 26540 }, { "grad_norm": 0.35740622878074646, "learning_rate": 7.98073719257645e-05, "loss": 0.0264, "step": 26550 }, { "grad_norm": 0.34246134757995605, "learning_rate": 7.979077526029499e-05, "loss": 0.0309, "step": 26560 }, { "grad_norm": 0.2711082696914673, "learning_rate": 7.977417350439675e-05, "loss": 0.0254, "step": 26570 }, { "grad_norm": 0.32373881340026855, "learning_rate": 7.97575666609066e-05, "loss": 0.027, "step": 26580 }, { "grad_norm": 0.2395429164171219, "learning_rate": 7.974095473266216e-05, "loss": 0.0245, "step": 26590 }, { "grad_norm": 0.3751452565193176, "learning_rate": 7.972433772250198e-05, "loss": 0.0337, "step": 26600 }, { "grad_norm": 0.1797022521495819, "learning_rate": 7.970771563326544e-05, "loss": 0.0236, "step": 26610 }, { "grad_norm": 0.2962683439254761, "learning_rate": 7.96910884677928e-05, "loss": 0.0299, "step": 26620 }, { "grad_norm": 0.4292139410972595, "learning_rate": 7.967445622892523e-05, "loss": 0.0309, "step": 26630 }, { "grad_norm": 0.5442044138908386, "learning_rate": 7.965781891950465e-05, "loss": 0.0274, "step": 26640 }, { "grad_norm": 0.2321830540895462, "learning_rate": 7.964117654237397e-05, "loss": 0.0222, "step": 26650 }, { "grad_norm": 0.3194202482700348, "learning_rate": 7.962452910037692e-05, "loss": 0.0289, "step": 26660 }, { "grad_norm": 0.27043092250823975, "learning_rate": 7.96078765963581e-05, "loss": 0.0248, "step": 26670 }, { "grad_norm": 0.4234366714954376, "learning_rate": 7.95912190331629e-05, "loss": 0.0254, "step": 26680 }, { "grad_norm": 0.3801901340484619, "learning_rate": 7.957455641363772e-05, "loss": 0.0302, "step": 26690 }, { "grad_norm": 0.26723116636276245, "learning_rate": 7.955788874062968e-05, "loss": 0.0309, "step": 26700 }, { "grad_norm": 0.3200388252735138, "learning_rate": 7.95412160169869e-05, "loss": 0.0258, "step": 26710 }, { "grad_norm": 0.626234769821167, "learning_rate": 7.952453824555824e-05, "loss": 0.0258, "step": 26720 }, { "grad_norm": 0.42818716168403625, "learning_rate": 7.95078554291935e-05, "loss": 0.0277, "step": 26730 }, { "grad_norm": 0.29754382371902466, "learning_rate": 7.94911675707433e-05, "loss": 0.0288, "step": 26740 }, { "grad_norm": 0.3480367958545685, "learning_rate": 7.947447467305915e-05, "loss": 0.032, "step": 26750 }, { "grad_norm": 0.23015104234218597, "learning_rate": 7.94577767389934e-05, "loss": 0.0264, "step": 26760 }, { "grad_norm": 0.2783924341201782, "learning_rate": 7.944107377139928e-05, "loss": 0.0237, "step": 26770 }, { "grad_norm": 0.7308823466300964, "learning_rate": 7.942436577313088e-05, "loss": 0.031, "step": 26780 }, { "grad_norm": 0.2754218876361847, "learning_rate": 7.940765274704312e-05, "loss": 0.0289, "step": 26790 }, { "grad_norm": 0.27364879846572876, "learning_rate": 7.939093469599181e-05, "loss": 0.0289, "step": 26800 }, { "grad_norm": 0.2912551760673523, "learning_rate": 7.93742116228336e-05, "loss": 0.0237, "step": 26810 }, { "grad_norm": 0.6434603929519653, "learning_rate": 7.935748353042602e-05, "loss": 0.0291, "step": 26820 }, { "grad_norm": 0.5196155309677124, "learning_rate": 7.934075042162744e-05, "loss": 0.0321, "step": 26830 }, { "grad_norm": 0.29147306084632874, "learning_rate": 7.932401229929705e-05, "loss": 0.033, "step": 26840 }, { "grad_norm": 0.3272143006324768, "learning_rate": 7.9307269166295e-05, "loss": 0.0266, "step": 26850 }, { "grad_norm": 0.23778145015239716, "learning_rate": 7.92905210254822e-05, "loss": 0.0263, "step": 26860 }, { "grad_norm": 0.3283270299434662, "learning_rate": 7.927376787972045e-05, "loss": 0.0256, "step": 26870 }, { "grad_norm": 0.2893885374069214, "learning_rate": 7.92570097318724e-05, "loss": 0.0248, "step": 26880 }, { "grad_norm": 0.24871332943439484, "learning_rate": 7.924024658480158e-05, "loss": 0.0227, "step": 26890 }, { "grad_norm": 0.25122690200805664, "learning_rate": 7.922347844137233e-05, "loss": 0.0257, "step": 26900 }, { "grad_norm": 0.49674418568611145, "learning_rate": 7.92067053044499e-05, "loss": 0.0243, "step": 26910 }, { "grad_norm": 0.5326026082038879, "learning_rate": 7.918992717690031e-05, "loss": 0.0314, "step": 26920 }, { "grad_norm": 0.3346273899078369, "learning_rate": 7.917314406159053e-05, "loss": 0.0246, "step": 26930 }, { "grad_norm": 0.4044027030467987, "learning_rate": 7.915635596138832e-05, "loss": 0.033, "step": 26940 }, { "grad_norm": 0.2644941508769989, "learning_rate": 7.913956287916228e-05, "loss": 0.0316, "step": 26950 }, { "grad_norm": 0.2572774589061737, "learning_rate": 7.912276481778193e-05, "loss": 0.0241, "step": 26960 }, { "grad_norm": 0.3245455324649811, "learning_rate": 7.910596178011759e-05, "loss": 0.0322, "step": 26970 }, { "grad_norm": 0.3312346637248993, "learning_rate": 7.908915376904043e-05, "loss": 0.0202, "step": 26980 }, { "grad_norm": 0.2584068179130554, "learning_rate": 7.907234078742247e-05, "loss": 0.0269, "step": 26990 }, { "grad_norm": 0.3191561698913574, "learning_rate": 7.90555228381366e-05, "loss": 0.0209, "step": 27000 }, { "grad_norm": 0.2548160254955292, "learning_rate": 7.903869992405656e-05, "loss": 0.0195, "step": 27010 }, { "grad_norm": 0.2855788767337799, "learning_rate": 7.902187204805691e-05, "loss": 0.0231, "step": 27020 }, { "grad_norm": 0.2877477705478668, "learning_rate": 7.900503921301308e-05, "loss": 0.0208, "step": 27030 }, { "grad_norm": 0.2444644570350647, "learning_rate": 7.898820142180133e-05, "loss": 0.031, "step": 27040 }, { "grad_norm": 0.306071013212204, "learning_rate": 7.897135867729879e-05, "loss": 0.019, "step": 27050 }, { "grad_norm": 0.2959018349647522, "learning_rate": 7.89545109823834e-05, "loss": 0.0247, "step": 27060 }, { "grad_norm": 0.4063841998577118, "learning_rate": 7.8937658339934e-05, "loss": 0.0285, "step": 27070 }, { "grad_norm": 0.31638380885124207, "learning_rate": 7.892080075283026e-05, "loss": 0.0253, "step": 27080 }, { "grad_norm": 0.2298644781112671, "learning_rate": 7.890393822395263e-05, "loss": 0.0226, "step": 27090 }, { "grad_norm": 0.3473605811595917, "learning_rate": 7.88870707561825e-05, "loss": 0.0254, "step": 27100 }, { "grad_norm": 0.46357300877571106, "learning_rate": 7.887019835240203e-05, "loss": 0.0313, "step": 27110 }, { "grad_norm": 0.29372215270996094, "learning_rate": 7.885332101549427e-05, "loss": 0.0284, "step": 27120 }, { "grad_norm": 0.27448976039886475, "learning_rate": 7.883643874834308e-05, "loss": 0.0209, "step": 27130 }, { "grad_norm": 0.40210482478141785, "learning_rate": 7.881955155383321e-05, "loss": 0.0348, "step": 27140 }, { "grad_norm": 0.21364209055900574, "learning_rate": 7.880265943485017e-05, "loss": 0.0255, "step": 27150 }, { "grad_norm": 0.3611292839050293, "learning_rate": 7.878576239428038e-05, "loss": 0.0254, "step": 27160 }, { "grad_norm": 0.2599988579750061, "learning_rate": 7.87688604350111e-05, "loss": 0.0311, "step": 27170 }, { "grad_norm": 0.329531729221344, "learning_rate": 7.875195355993042e-05, "loss": 0.0323, "step": 27180 }, { "grad_norm": 0.4273085296154022, "learning_rate": 7.873504177192724e-05, "loss": 0.0261, "step": 27190 }, { "grad_norm": 0.28124290704727173, "learning_rate": 7.87181250738913e-05, "loss": 0.0317, "step": 27200 }, { "grad_norm": 0.26008808612823486, "learning_rate": 7.870120346871324e-05, "loss": 0.023, "step": 27210 }, { "grad_norm": 0.3132264316082001, "learning_rate": 7.86842769592845e-05, "loss": 0.0269, "step": 27220 }, { "grad_norm": 0.3408699631690979, "learning_rate": 7.866734554849732e-05, "loss": 0.0258, "step": 27230 }, { "grad_norm": 0.3713447153568268, "learning_rate": 7.865040923924486e-05, "loss": 0.0236, "step": 27240 }, { "grad_norm": 0.40031298995018005, "learning_rate": 7.863346803442104e-05, "loss": 0.0291, "step": 27250 }, { "grad_norm": 0.6311206817626953, "learning_rate": 7.861652193692067e-05, "loss": 0.023, "step": 27260 }, { "grad_norm": 0.2389308512210846, "learning_rate": 7.859957094963937e-05, "loss": 0.0186, "step": 27270 }, { "grad_norm": 0.25192350149154663, "learning_rate": 7.858261507547357e-05, "loss": 0.0216, "step": 27280 }, { "grad_norm": 0.2777305543422699, "learning_rate": 7.856565431732061e-05, "loss": 0.0204, "step": 27290 }, { "grad_norm": 0.21593455970287323, "learning_rate": 7.854868867807859e-05, "loss": 0.0171, "step": 27300 }, { "grad_norm": 0.2796812355518341, "learning_rate": 7.85317181606465e-05, "loss": 0.0235, "step": 27310 }, { "grad_norm": 0.24290256202220917, "learning_rate": 7.85147427679241e-05, "loss": 0.0226, "step": 27320 }, { "grad_norm": 0.4518110454082489, "learning_rate": 7.849776250281205e-05, "loss": 0.0209, "step": 27330 }, { "grad_norm": 0.2509656846523285, "learning_rate": 7.84807773682118e-05, "loss": 0.0215, "step": 27340 }, { "grad_norm": 0.5590590834617615, "learning_rate": 7.846378736702565e-05, "loss": 0.0269, "step": 27350 }, { "grad_norm": 0.26856449246406555, "learning_rate": 7.844679250215671e-05, "loss": 0.0245, "step": 27360 }, { "grad_norm": 0.2540501058101654, "learning_rate": 7.842979277650898e-05, "loss": 0.0208, "step": 27370 }, { "grad_norm": 0.276500940322876, "learning_rate": 7.84127881929872e-05, "loss": 0.0231, "step": 27380 }, { "grad_norm": 0.32736846804618835, "learning_rate": 7.839577875449704e-05, "loss": 0.0248, "step": 27390 }, { "grad_norm": 0.3067411780357361, "learning_rate": 7.837876446394489e-05, "loss": 0.0203, "step": 27400 }, { "grad_norm": 0.2755142152309418, "learning_rate": 7.836174532423805e-05, "loss": 0.0229, "step": 27410 }, { "grad_norm": 0.24656815826892853, "learning_rate": 7.834472133828466e-05, "loss": 0.0213, "step": 27420 }, { "grad_norm": 0.3239099085330963, "learning_rate": 7.832769250899359e-05, "loss": 0.0283, "step": 27430 }, { "grad_norm": 0.32760658860206604, "learning_rate": 7.831065883927464e-05, "loss": 0.0293, "step": 27440 }, { "grad_norm": 0.3642767369747162, "learning_rate": 7.829362033203841e-05, "loss": 0.0266, "step": 27450 }, { "grad_norm": 0.3692730963230133, "learning_rate": 7.827657699019628e-05, "loss": 0.0295, "step": 27460 }, { "grad_norm": 0.36414000391960144, "learning_rate": 7.825952881666052e-05, "loss": 0.0415, "step": 27470 }, { "grad_norm": 0.3073456883430481, "learning_rate": 7.824247581434418e-05, "loss": 0.0272, "step": 27480 }, { "grad_norm": 0.46151548624038696, "learning_rate": 7.822541798616116e-05, "loss": 0.0259, "step": 27490 }, { "grad_norm": 0.3796616792678833, "learning_rate": 7.820835533502617e-05, "loss": 0.0322, "step": 27500 }, { "grad_norm": 0.32876476645469666, "learning_rate": 7.819128786385475e-05, "loss": 0.0344, "step": 27510 }, { "grad_norm": 0.3449096381664276, "learning_rate": 7.817421557556329e-05, "loss": 0.0348, "step": 27520 }, { "grad_norm": 0.28835001587867737, "learning_rate": 7.815713847306893e-05, "loss": 0.0296, "step": 27530 }, { "grad_norm": 0.2821832299232483, "learning_rate": 7.81400565592897e-05, "loss": 0.0249, "step": 27540 }, { "grad_norm": 0.3262866735458374, "learning_rate": 7.812296983714444e-05, "loss": 0.0332, "step": 27550 }, { "grad_norm": 0.2662400007247925, "learning_rate": 7.810587830955281e-05, "loss": 0.024, "step": 27560 }, { "grad_norm": 0.31650277972221375, "learning_rate": 7.808878197943528e-05, "loss": 0.0314, "step": 27570 }, { "grad_norm": 0.24242694675922394, "learning_rate": 7.807168084971312e-05, "loss": 0.0248, "step": 27580 }, { "grad_norm": 0.2884470224380493, "learning_rate": 7.805457492330849e-05, "loss": 0.0232, "step": 27590 }, { "grad_norm": 0.2527514398097992, "learning_rate": 7.803746420314428e-05, "loss": 0.0236, "step": 27600 }, { "grad_norm": 0.47591257095336914, "learning_rate": 7.802034869214428e-05, "loss": 0.0333, "step": 27610 }, { "grad_norm": 0.27044206857681274, "learning_rate": 7.800322839323303e-05, "loss": 0.0228, "step": 27620 }, { "grad_norm": 0.295116662979126, "learning_rate": 7.798610330933593e-05, "loss": 0.0246, "step": 27630 }, { "grad_norm": 0.22545234858989716, "learning_rate": 7.796897344337922e-05, "loss": 0.0248, "step": 27640 }, { "grad_norm": 0.3324466943740845, "learning_rate": 7.795183879828989e-05, "loss": 0.0299, "step": 27650 }, { "grad_norm": 0.2917918264865875, "learning_rate": 7.793469937699579e-05, "loss": 0.0331, "step": 27660 }, { "grad_norm": 0.29844069480895996, "learning_rate": 7.791755518242558e-05, "loss": 0.0274, "step": 27670 }, { "grad_norm": 0.2642054259777069, "learning_rate": 7.790040621750876e-05, "loss": 0.0288, "step": 27680 }, { "grad_norm": 0.35240617394447327, "learning_rate": 7.788325248517558e-05, "loss": 0.0253, "step": 27690 }, { "grad_norm": 0.2680457532405853, "learning_rate": 7.786609398835715e-05, "loss": 0.0256, "step": 27700 }, { "grad_norm": 0.3305671811103821, "learning_rate": 7.784893072998541e-05, "loss": 0.0232, "step": 27710 }, { "grad_norm": 0.2644398510456085, "learning_rate": 7.783176271299306e-05, "loss": 0.0237, "step": 27720 }, { "grad_norm": 0.49170413613319397, "learning_rate": 7.781458994031368e-05, "loss": 0.0241, "step": 27730 }, { "grad_norm": 0.3636353313922882, "learning_rate": 7.779741241488161e-05, "loss": 0.0253, "step": 27740 }, { "grad_norm": 0.2894432842731476, "learning_rate": 7.7780230139632e-05, "loss": 0.025, "step": 27750 }, { "grad_norm": 0.38436755537986755, "learning_rate": 7.776304311750087e-05, "loss": 0.0251, "step": 27760 }, { "grad_norm": 0.3905974328517914, "learning_rate": 7.7745851351425e-05, "loss": 0.0323, "step": 27770 }, { "grad_norm": 0.27064862847328186, "learning_rate": 7.772865484434197e-05, "loss": 0.0248, "step": 27780 }, { "grad_norm": 0.3408285081386566, "learning_rate": 7.77114535991902e-05, "loss": 0.0282, "step": 27790 }, { "grad_norm": 0.31367310881614685, "learning_rate": 7.769424761890893e-05, "loss": 0.0279, "step": 27800 }, { "grad_norm": 0.43808168172836304, "learning_rate": 7.767703690643817e-05, "loss": 0.0298, "step": 27810 }, { "grad_norm": 1.66057550907135, "learning_rate": 7.76598214647188e-05, "loss": 0.0286, "step": 27820 }, { "grad_norm": 0.3680925965309143, "learning_rate": 7.764260129669241e-05, "loss": 0.0226, "step": 27830 }, { "grad_norm": 0.2896800935268402, "learning_rate": 7.76253764053015e-05, "loss": 0.0305, "step": 27840 }, { "grad_norm": 0.2310210019350052, "learning_rate": 7.760814679348932e-05, "loss": 0.02, "step": 27850 }, { "grad_norm": 0.22771795094013214, "learning_rate": 7.759091246419992e-05, "loss": 0.0228, "step": 27860 }, { "grad_norm": 0.34763044118881226, "learning_rate": 7.75736734203782e-05, "loss": 0.0226, "step": 27870 }, { "grad_norm": 0.3368540108203888, "learning_rate": 7.755642966496985e-05, "loss": 0.0422, "step": 27880 }, { "grad_norm": 0.34742194414138794, "learning_rate": 7.753918120092132e-05, "loss": 0.0278, "step": 27890 }, { "grad_norm": 0.28700995445251465, "learning_rate": 7.752192803117993e-05, "loss": 0.0231, "step": 27900 }, { "grad_norm": 0.3091486990451813, "learning_rate": 7.750467015869377e-05, "loss": 0.0294, "step": 27910 }, { "grad_norm": 0.35058972239494324, "learning_rate": 7.748740758641174e-05, "loss": 0.0295, "step": 27920 }, { "grad_norm": 0.25733110308647156, "learning_rate": 7.74701403172835e-05, "loss": 0.0281, "step": 27930 }, { "grad_norm": 0.2032008171081543, "learning_rate": 7.745286835425962e-05, "loss": 0.0271, "step": 27940 }, { "grad_norm": 0.3180219233036041, "learning_rate": 7.743559170029138e-05, "loss": 0.0341, "step": 27950 }, { "grad_norm": 0.2428389936685562, "learning_rate": 7.741831035833087e-05, "loss": 0.0368, "step": 27960 }, { "grad_norm": 0.38068363070487976, "learning_rate": 7.740102433133102e-05, "loss": 0.0262, "step": 27970 }, { "grad_norm": 0.36794719099998474, "learning_rate": 7.738373362224553e-05, "loss": 0.0263, "step": 27980 }, { "grad_norm": 1.2368296384811401, "learning_rate": 7.73664382340289e-05, "loss": 0.0262, "step": 27990 }, { "grad_norm": 0.2800149917602539, "learning_rate": 7.734913816963647e-05, "loss": 0.0323, "step": 28000 }, { "grad_norm": 0.4190833568572998, "learning_rate": 7.73318334320243e-05, "loss": 0.0289, "step": 28010 }, { "grad_norm": 0.4085351228713989, "learning_rate": 7.731452402414934e-05, "loss": 0.0334, "step": 28020 }, { "grad_norm": 0.3820250928401947, "learning_rate": 7.729720994896928e-05, "loss": 0.0274, "step": 28030 }, { "grad_norm": 0.4983803331851959, "learning_rate": 7.727989120944262e-05, "loss": 0.0301, "step": 28040 }, { "grad_norm": 0.20337407290935516, "learning_rate": 7.726256780852865e-05, "loss": 0.0227, "step": 28050 }, { "grad_norm": 0.33308130502700806, "learning_rate": 7.724523974918749e-05, "loss": 0.0233, "step": 28060 }, { "grad_norm": 0.2781769633293152, "learning_rate": 7.722790703438002e-05, "loss": 0.0217, "step": 28070 }, { "grad_norm": 0.2895805239677429, "learning_rate": 7.72105696670679e-05, "loss": 0.0241, "step": 28080 }, { "grad_norm": 0.3351166248321533, "learning_rate": 7.719322765021364e-05, "loss": 0.0263, "step": 28090 }, { "grad_norm": 0.32334670424461365, "learning_rate": 7.717588098678051e-05, "loss": 0.0243, "step": 28100 }, { "grad_norm": 0.3037010133266449, "learning_rate": 7.715852967973258e-05, "loss": 0.0289, "step": 28110 }, { "grad_norm": 3.7134792804718018, "learning_rate": 7.714117373203474e-05, "loss": 0.0259, "step": 28120 }, { "grad_norm": 0.41807258129119873, "learning_rate": 7.712381314665259e-05, "loss": 0.0256, "step": 28130 }, { "grad_norm": 0.27229204773902893, "learning_rate": 7.710644792655261e-05, "loss": 0.0267, "step": 28140 }, { "grad_norm": 0.4019816815853119, "learning_rate": 7.708907807470207e-05, "loss": 0.0367, "step": 28150 }, { "grad_norm": 0.3324771225452423, "learning_rate": 7.707170359406896e-05, "loss": 0.0282, "step": 28160 }, { "grad_norm": 0.34752553701400757, "learning_rate": 7.705432448762213e-05, "loss": 0.0229, "step": 28170 }, { "grad_norm": 0.2621327340602875, "learning_rate": 7.703694075833117e-05, "loss": 0.0308, "step": 28180 }, { "grad_norm": 0.3585147261619568, "learning_rate": 7.70195524091665e-05, "loss": 0.0416, "step": 28190 }, { "grad_norm": 0.40909749269485474, "learning_rate": 7.70021594430993e-05, "loss": 0.0298, "step": 28200 }, { "grad_norm": 2.987316846847534, "learning_rate": 7.698476186310157e-05, "loss": 0.0338, "step": 28210 }, { "grad_norm": 0.27550002932548523, "learning_rate": 7.696735967214608e-05, "loss": 0.0292, "step": 28220 }, { "grad_norm": 0.35231900215148926, "learning_rate": 7.694995287320636e-05, "loss": 0.0259, "step": 28230 }, { "grad_norm": 0.294406920671463, "learning_rate": 7.693254146925679e-05, "loss": 0.0211, "step": 28240 }, { "grad_norm": 0.34690266847610474, "learning_rate": 7.691512546327251e-05, "loss": 0.0253, "step": 28250 }, { "grad_norm": 0.29621443152427673, "learning_rate": 7.689770485822939e-05, "loss": 0.0229, "step": 28260 }, { "grad_norm": 0.20040467381477356, "learning_rate": 7.688027965710416e-05, "loss": 0.0197, "step": 28270 }, { "grad_norm": 0.41746193170547485, "learning_rate": 7.686284986287433e-05, "loss": 0.0274, "step": 28280 }, { "grad_norm": 0.2197502702474594, "learning_rate": 7.684541547851817e-05, "loss": 0.0225, "step": 28290 }, { "grad_norm": 0.24005304276943207, "learning_rate": 7.68279765070147e-05, "loss": 0.0231, "step": 28300 }, { "grad_norm": 0.29325422644615173, "learning_rate": 7.68105329513438e-05, "loss": 0.021, "step": 28310 }, { "grad_norm": 0.24774397909641266, "learning_rate": 7.67930848144861e-05, "loss": 0.0285, "step": 28320 }, { "grad_norm": 0.35097968578338623, "learning_rate": 7.6775632099423e-05, "loss": 0.0263, "step": 28330 }, { "grad_norm": 0.26519131660461426, "learning_rate": 7.675817480913667e-05, "loss": 0.0262, "step": 28340 }, { "grad_norm": 0.3203217089176178, "learning_rate": 7.674071294661011e-05, "loss": 0.0219, "step": 28350 }, { "grad_norm": 0.32283589243888855, "learning_rate": 7.672324651482707e-05, "loss": 0.0249, "step": 28360 }, { "grad_norm": 0.33513882756233215, "learning_rate": 7.670577551677209e-05, "loss": 0.0244, "step": 28370 }, { "grad_norm": 0.2858888506889343, "learning_rate": 7.668829995543047e-05, "loss": 0.0272, "step": 28380 }, { "grad_norm": 0.32573285698890686, "learning_rate": 7.667081983378832e-05, "loss": 0.0249, "step": 28390 }, { "grad_norm": 0.28782904148101807, "learning_rate": 7.66533351548325e-05, "loss": 0.0208, "step": 28400 }, { "grad_norm": 0.3908216059207916, "learning_rate": 7.663584592155069e-05, "loss": 0.0255, "step": 28410 }, { "grad_norm": 0.31970253586769104, "learning_rate": 7.661835213693129e-05, "loss": 0.0307, "step": 28420 }, { "grad_norm": 0.34071868658065796, "learning_rate": 7.660085380396353e-05, "loss": 0.03, "step": 28430 }, { "grad_norm": 0.3455551862716675, "learning_rate": 7.658335092563738e-05, "loss": 0.024, "step": 28440 }, { "grad_norm": 1.5076864957809448, "learning_rate": 7.656584350494362e-05, "loss": 0.0276, "step": 28450 }, { "grad_norm": 0.30114978551864624, "learning_rate": 7.654833154487378e-05, "loss": 0.0265, "step": 28460 }, { "grad_norm": 0.2668339014053345, "learning_rate": 7.653081504842017e-05, "loss": 0.0303, "step": 28470 }, { "grad_norm": 0.2953836917877197, "learning_rate": 7.65132940185759e-05, "loss": 0.0337, "step": 28480 }, { "grad_norm": 0.23585617542266846, "learning_rate": 7.649576845833481e-05, "loss": 0.0243, "step": 28490 }, { "grad_norm": 0.1971285492181778, "learning_rate": 7.647823837069156e-05, "loss": 0.0216, "step": 28500 }, { "grad_norm": 0.2947187125682831, "learning_rate": 7.646070375864156e-05, "loss": 0.0221, "step": 28510 }, { "grad_norm": 0.641483724117279, "learning_rate": 7.644316462518099e-05, "loss": 0.0275, "step": 28520 }, { "grad_norm": 0.3106439411640167, "learning_rate": 7.642562097330679e-05, "loss": 0.0289, "step": 28530 }, { "grad_norm": 0.23005716502666473, "learning_rate": 7.640807280601671e-05, "loss": 0.027, "step": 28540 }, { "grad_norm": 0.2332245111465454, "learning_rate": 7.639052012630927e-05, "loss": 0.0322, "step": 28550 }, { "grad_norm": 0.3740326464176178, "learning_rate": 7.63729629371837e-05, "loss": 0.0246, "step": 28560 }, { "grad_norm": 0.22232681512832642, "learning_rate": 7.635540124164009e-05, "loss": 0.0328, "step": 28570 }, { "grad_norm": 0.24453231692314148, "learning_rate": 7.633783504267922e-05, "loss": 0.0273, "step": 28580 }, { "grad_norm": 0.29753604531288147, "learning_rate": 7.632026434330269e-05, "loss": 0.022, "step": 28590 }, { "grad_norm": 0.23962928354740143, "learning_rate": 7.630268914651282e-05, "loss": 0.0243, "step": 28600 }, { "grad_norm": 0.2226066291332245, "learning_rate": 7.628510945531278e-05, "loss": 0.0395, "step": 28610 }, { "grad_norm": 0.28970175981521606, "learning_rate": 7.626752527270641e-05, "loss": 0.034, "step": 28620 }, { "grad_norm": 5.329678535461426, "learning_rate": 7.62499366016984e-05, "loss": 0.0277, "step": 28630 }, { "grad_norm": 0.5003214478492737, "learning_rate": 7.623234344529416e-05, "loss": 0.0281, "step": 28640 }, { "grad_norm": 0.28692004084587097, "learning_rate": 7.62147458064999e-05, "loss": 0.0266, "step": 28650 }, { "grad_norm": 0.26964035630226135, "learning_rate": 7.619714368832254e-05, "loss": 0.0305, "step": 28660 }, { "grad_norm": 0.20702341198921204, "learning_rate": 7.61795370937698e-05, "loss": 0.0207, "step": 28670 }, { "grad_norm": 0.609432578086853, "learning_rate": 7.61619260258502e-05, "loss": 0.0255, "step": 28680 }, { "grad_norm": 1.0868483781814575, "learning_rate": 7.614431048757298e-05, "loss": 0.0284, "step": 28690 }, { "grad_norm": 0.3014363944530487, "learning_rate": 7.612669048194814e-05, "loss": 0.0264, "step": 28700 }, { "grad_norm": 0.28301602602005005, "learning_rate": 7.610906601198646e-05, "loss": 0.0218, "step": 28710 }, { "grad_norm": 0.31366729736328125, "learning_rate": 7.60914370806995e-05, "loss": 0.0239, "step": 28720 }, { "grad_norm": 0.2773756980895996, "learning_rate": 7.607380369109953e-05, "loss": 0.0365, "step": 28730 }, { "grad_norm": 0.25022852420806885, "learning_rate": 7.605616584619961e-05, "loss": 0.0213, "step": 28740 }, { "grad_norm": 0.30669161677360535, "learning_rate": 7.603852354901362e-05, "loss": 0.0246, "step": 28750 }, { "grad_norm": 0.2618636190891266, "learning_rate": 7.602087680255609e-05, "loss": 0.034, "step": 28760 }, { "grad_norm": 0.30323532223701477, "learning_rate": 7.600322560984238e-05, "loss": 0.0212, "step": 28770 }, { "grad_norm": 0.3434722125530243, "learning_rate": 7.598556997388863e-05, "loss": 0.0214, "step": 28780 }, { "grad_norm": 0.2535940408706665, "learning_rate": 7.596790989771166e-05, "loss": 0.0235, "step": 28790 }, { "grad_norm": 0.27444717288017273, "learning_rate": 7.595024538432914e-05, "loss": 0.0215, "step": 28800 }, { "grad_norm": 0.27322959899902344, "learning_rate": 7.59325764367594e-05, "loss": 0.0197, "step": 28810 }, { "grad_norm": 0.37574559450149536, "learning_rate": 7.59149030580216e-05, "loss": 0.0267, "step": 28820 }, { "grad_norm": 0.29238811135292053, "learning_rate": 7.589722525113562e-05, "loss": 0.0247, "step": 28830 }, { "grad_norm": 0.3392059803009033, "learning_rate": 7.587954301912216e-05, "loss": 0.0245, "step": 28840 }, { "grad_norm": 0.40350043773651123, "learning_rate": 7.586185636500263e-05, "loss": 0.0353, "step": 28850 }, { "grad_norm": 0.4704602062702179, "learning_rate": 7.584416529179914e-05, "loss": 0.0358, "step": 28860 }, { "grad_norm": 0.35593077540397644, "learning_rate": 7.582646980253465e-05, "loss": 0.0312, "step": 28870 }, { "grad_norm": 0.35433223843574524, "learning_rate": 7.580876990023282e-05, "loss": 0.0407, "step": 28880 }, { "grad_norm": 0.4655781388282776, "learning_rate": 7.579106558791809e-05, "loss": 0.0298, "step": 28890 }, { "grad_norm": 0.5426051020622253, "learning_rate": 7.577335686861565e-05, "loss": 0.0278, "step": 28900 }, { "grad_norm": 0.38758552074432373, "learning_rate": 7.575564374535141e-05, "loss": 0.0268, "step": 28910 }, { "grad_norm": 0.9688271284103394, "learning_rate": 7.573792622115207e-05, "loss": 0.0249, "step": 28920 }, { "grad_norm": 0.5891004800796509, "learning_rate": 7.572020429904507e-05, "loss": 0.0281, "step": 28930 }, { "grad_norm": 0.39669549465179443, "learning_rate": 7.570247798205861e-05, "loss": 0.0287, "step": 28940 }, { "grad_norm": 0.2673846185207367, "learning_rate": 7.568474727322164e-05, "loss": 0.0363, "step": 28950 }, { "grad_norm": 0.25252142548561096, "learning_rate": 7.566701217556384e-05, "loss": 0.0298, "step": 28960 }, { "grad_norm": 0.43540066480636597, "learning_rate": 7.564927269211564e-05, "loss": 0.0291, "step": 28970 }, { "grad_norm": 0.25618776679039, "learning_rate": 7.563152882590824e-05, "loss": 0.0256, "step": 28980 }, { "grad_norm": 0.4157531261444092, "learning_rate": 7.56137805799736e-05, "loss": 0.0265, "step": 28990 }, { "grad_norm": 0.28894519805908203, "learning_rate": 7.559602795734439e-05, "loss": 0.0263, "step": 29000 }, { "grad_norm": 0.3185175657272339, "learning_rate": 7.557827096105408e-05, "loss": 0.025, "step": 29010 }, { "grad_norm": 0.3683930039405823, "learning_rate": 7.55605095941368e-05, "loss": 0.0253, "step": 29020 }, { "grad_norm": 0.23912641406059265, "learning_rate": 7.554274385962752e-05, "loss": 0.0225, "step": 29030 }, { "grad_norm": 0.3310268819332123, "learning_rate": 7.552497376056191e-05, "loss": 0.0265, "step": 29040 }, { "grad_norm": 0.30700087547302246, "learning_rate": 7.550719929997639e-05, "loss": 0.0242, "step": 29050 }, { "grad_norm": 0.27291277050971985, "learning_rate": 7.548942048090813e-05, "loss": 0.0251, "step": 29060 }, { "grad_norm": 0.3330865800380707, "learning_rate": 7.547163730639506e-05, "loss": 0.0231, "step": 29070 }, { "grad_norm": 0.34291934967041016, "learning_rate": 7.545384977947583e-05, "loss": 0.0256, "step": 29080 }, { "grad_norm": 0.24107429385185242, "learning_rate": 7.543605790318981e-05, "loss": 0.0257, "step": 29090 }, { "grad_norm": 0.27007564902305603, "learning_rate": 7.54182616805772e-05, "loss": 0.0327, "step": 29100 }, { "grad_norm": 0.2920527756214142, "learning_rate": 7.540046111467885e-05, "loss": 0.023, "step": 29110 }, { "grad_norm": 0.23954983055591583, "learning_rate": 7.53826562085364e-05, "loss": 0.0244, "step": 29120 }, { "grad_norm": 0.33418187499046326, "learning_rate": 7.536484696519221e-05, "loss": 0.0209, "step": 29130 }, { "grad_norm": 0.230316162109375, "learning_rate": 7.534703338768942e-05, "loss": 0.0267, "step": 29140 }, { "grad_norm": 0.2237173169851303, "learning_rate": 7.532921547907185e-05, "loss": 0.0273, "step": 29150 }, { "grad_norm": 0.2641003727912903, "learning_rate": 7.531139324238412e-05, "loss": 0.0312, "step": 29160 }, { "grad_norm": 0.3145398795604706, "learning_rate": 7.529356668067157e-05, "loss": 0.0342, "step": 29170 }, { "grad_norm": 0.3721446692943573, "learning_rate": 7.527573579698023e-05, "loss": 0.0289, "step": 29180 }, { "grad_norm": 0.44014692306518555, "learning_rate": 7.525790059435693e-05, "loss": 0.0217, "step": 29190 }, { "grad_norm": 0.27641865611076355, "learning_rate": 7.524006107584926e-05, "loss": 0.0227, "step": 29200 }, { "grad_norm": 0.2968360483646393, "learning_rate": 7.522221724450544e-05, "loss": 0.0242, "step": 29210 }, { "grad_norm": 0.5694006681442261, "learning_rate": 7.520436910337451e-05, "loss": 0.0303, "step": 29220 }, { "grad_norm": 0.21502625942230225, "learning_rate": 7.518651665550627e-05, "loss": 0.0187, "step": 29230 }, { "grad_norm": 0.3125355541706085, "learning_rate": 7.516865990395117e-05, "loss": 0.0363, "step": 29240 }, { "grad_norm": 0.33766311407089233, "learning_rate": 7.515079885176047e-05, "loss": 0.0226, "step": 29250 }, { "grad_norm": 0.25866377353668213, "learning_rate": 7.513293350198612e-05, "loss": 0.0239, "step": 29260 }, { "grad_norm": 0.3320462703704834, "learning_rate": 7.511506385768081e-05, "loss": 0.0192, "step": 29270 }, { "grad_norm": 0.33073046803474426, "learning_rate": 7.509718992189801e-05, "loss": 0.0238, "step": 29280 }, { "grad_norm": 0.26977893710136414, "learning_rate": 7.507931169769182e-05, "loss": 0.0242, "step": 29290 }, { "grad_norm": 0.2724124491214752, "learning_rate": 7.506142918811722e-05, "loss": 0.0278, "step": 29300 }, { "grad_norm": 0.32559818029403687, "learning_rate": 7.504354239622978e-05, "loss": 0.0303, "step": 29310 }, { "grad_norm": 0.25003308057785034, "learning_rate": 7.50256513250859e-05, "loss": 0.0261, "step": 29320 }, { "grad_norm": 0.30593326687812805, "learning_rate": 7.500775597774265e-05, "loss": 0.0242, "step": 29330 }, { "grad_norm": 0.22199858725070953, "learning_rate": 7.498985635725788e-05, "loss": 0.0233, "step": 29340 }, { "grad_norm": 0.2676449120044708, "learning_rate": 7.497195246669012e-05, "loss": 0.0265, "step": 29350 }, { "grad_norm": 0.3332822918891907, "learning_rate": 7.495404430909868e-05, "loss": 0.0259, "step": 29360 }, { "grad_norm": 0.29708364605903625, "learning_rate": 7.493613188754356e-05, "loss": 0.0263, "step": 29370 }, { "grad_norm": 0.3559577763080597, "learning_rate": 7.49182152050855e-05, "loss": 0.0239, "step": 29380 }, { "grad_norm": 0.2729189693927765, "learning_rate": 7.490029426478598e-05, "loss": 0.021, "step": 29390 }, { "grad_norm": 0.7006871700286865, "learning_rate": 7.488236906970719e-05, "loss": 0.0229, "step": 29400 }, { "grad_norm": 0.32186436653137207, "learning_rate": 7.486443962291207e-05, "loss": 0.0265, "step": 29410 }, { "grad_norm": 0.23723134398460388, "learning_rate": 7.484650592746424e-05, "loss": 0.0224, "step": 29420 }, { "grad_norm": 0.32838892936706543, "learning_rate": 7.482856798642811e-05, "loss": 0.0239, "step": 29430 }, { "grad_norm": 0.2707138955593109, "learning_rate": 7.481062580286878e-05, "loss": 0.0275, "step": 29440 }, { "grad_norm": 0.3070697784423828, "learning_rate": 7.479267937985208e-05, "loss": 0.0255, "step": 29450 }, { "grad_norm": 0.2821824252605438, "learning_rate": 7.477472872044456e-05, "loss": 0.0231, "step": 29460 }, { "grad_norm": 0.731458842754364, "learning_rate": 7.475677382771347e-05, "loss": 0.0228, "step": 29470 }, { "grad_norm": 0.34833911061286926, "learning_rate": 7.473881470472683e-05, "loss": 0.0256, "step": 29480 }, { "grad_norm": 0.3108293116092682, "learning_rate": 7.472085135455337e-05, "loss": 0.0256, "step": 29490 }, { "grad_norm": 0.27267488837242126, "learning_rate": 7.470288378026256e-05, "loss": 0.0294, "step": 29500 }, { "grad_norm": 0.31704843044281006, "learning_rate": 7.468491198492451e-05, "loss": 0.0245, "step": 29510 }, { "grad_norm": 0.23015518486499786, "learning_rate": 7.466693597161013e-05, "loss": 0.0234, "step": 29520 }, { "grad_norm": 0.314828485250473, "learning_rate": 7.464895574339104e-05, "loss": 0.0333, "step": 29530 }, { "grad_norm": 0.18683363497257233, "learning_rate": 7.463097130333958e-05, "loss": 0.0337, "step": 29540 }, { "grad_norm": 0.3592301905155182, "learning_rate": 7.461298265452876e-05, "loss": 0.024, "step": 29550 }, { "grad_norm": 0.3497019410133362, "learning_rate": 7.459498980003239e-05, "loss": 0.0238, "step": 29560 }, { "grad_norm": 0.2856887876987457, "learning_rate": 7.457699274292493e-05, "loss": 0.0319, "step": 29570 }, { "grad_norm": 0.24512958526611328, "learning_rate": 7.455899148628159e-05, "loss": 0.0295, "step": 29580 }, { "grad_norm": 0.3664008378982544, "learning_rate": 7.45409860331783e-05, "loss": 0.0253, "step": 29590 }, { "grad_norm": 0.262647807598114, "learning_rate": 7.452297638669169e-05, "loss": 0.0213, "step": 29600 }, { "grad_norm": 0.23805484175682068, "learning_rate": 7.450496254989911e-05, "loss": 0.0223, "step": 29610 }, { "grad_norm": 0.30348750948905945, "learning_rate": 7.448694452587866e-05, "loss": 0.0223, "step": 29620 }, { "grad_norm": 0.25707921385765076, "learning_rate": 7.44689223177091e-05, "loss": 0.0279, "step": 29630 }, { "grad_norm": 0.9167550802230835, "learning_rate": 7.445089592846994e-05, "loss": 0.0252, "step": 29640 }, { "grad_norm": 0.26267707347869873, "learning_rate": 7.443286536124141e-05, "loss": 0.0241, "step": 29650 }, { "grad_norm": 0.40588873624801636, "learning_rate": 7.441483061910443e-05, "loss": 0.0273, "step": 29660 }, { "grad_norm": 0.320885568857193, "learning_rate": 7.439679170514064e-05, "loss": 0.031, "step": 29670 }, { "grad_norm": 0.3466460704803467, "learning_rate": 7.43787486224324e-05, "loss": 0.0259, "step": 29680 }, { "grad_norm": 0.3374539911746979, "learning_rate": 7.436070137406276e-05, "loss": 0.0282, "step": 29690 }, { "grad_norm": 0.26226872205734253, "learning_rate": 7.434264996311556e-05, "loss": 0.019, "step": 29700 }, { "grad_norm": 0.2726500332355499, "learning_rate": 7.432459439267525e-05, "loss": 0.0255, "step": 29710 }, { "grad_norm": 0.2708204984664917, "learning_rate": 7.430653466582701e-05, "loss": 0.0244, "step": 29720 }, { "grad_norm": 0.3476639688014984, "learning_rate": 7.42884707856568e-05, "loss": 0.0359, "step": 29730 }, { "grad_norm": 1.5793424844741821, "learning_rate": 7.427040275525122e-05, "loss": 0.0273, "step": 29740 }, { "grad_norm": 0.235631063580513, "learning_rate": 7.42523305776976e-05, "loss": 0.0297, "step": 29750 }, { "grad_norm": 0.32114776968955994, "learning_rate": 7.4234254256084e-05, "loss": 0.0273, "step": 29760 }, { "grad_norm": 0.2871139347553253, "learning_rate": 7.421617379349915e-05, "loss": 0.0278, "step": 29770 }, { "grad_norm": 0.280791312456131, "learning_rate": 7.41980891930325e-05, "loss": 0.0248, "step": 29780 }, { "grad_norm": 0.16809724271297455, "learning_rate": 7.418000045777425e-05, "loss": 0.019, "step": 29790 }, { "grad_norm": 0.42767447233200073, "learning_rate": 7.416190759081523e-05, "loss": 0.0212, "step": 29800 }, { "grad_norm": 0.35398581624031067, "learning_rate": 7.414381059524704e-05, "loss": 0.023, "step": 29810 }, { "grad_norm": 0.2075231820344925, "learning_rate": 7.412570947416195e-05, "loss": 0.0189, "step": 29820 }, { "grad_norm": 0.38881605863571167, "learning_rate": 7.410760423065295e-05, "loss": 0.0268, "step": 29830 }, { "grad_norm": 0.23921695351600647, "learning_rate": 7.408949486781372e-05, "loss": 0.0262, "step": 29840 }, { "grad_norm": 0.4012291133403778, "learning_rate": 7.407138138873868e-05, "loss": 0.0256, "step": 29850 }, { "grad_norm": 0.43889376521110535, "learning_rate": 7.405326379652292e-05, "loss": 0.0254, "step": 29860 }, { "grad_norm": 0.29537710547447205, "learning_rate": 7.403514209426222e-05, "loss": 0.0236, "step": 29870 }, { "grad_norm": 0.25538843870162964, "learning_rate": 7.40170162850531e-05, "loss": 0.0271, "step": 29880 }, { "grad_norm": 0.24074313044548035, "learning_rate": 7.399888637199278e-05, "loss": 0.0184, "step": 29890 }, { "grad_norm": 0.25285086035728455, "learning_rate": 7.398075235817914e-05, "loss": 0.0306, "step": 29900 }, { "grad_norm": 0.2610253095626831, "learning_rate": 7.39626142467108e-05, "loss": 0.0238, "step": 29910 }, { "grad_norm": 0.365245521068573, "learning_rate": 7.394447204068706e-05, "loss": 0.022, "step": 29920 }, { "grad_norm": 0.3009619414806366, "learning_rate": 7.392632574320793e-05, "loss": 0.0287, "step": 29930 }, { "grad_norm": 0.3168794512748718, "learning_rate": 7.390817535737411e-05, "loss": 0.0209, "step": 29940 }, { "grad_norm": 0.25730955600738525, "learning_rate": 7.389002088628703e-05, "loss": 0.0224, "step": 29950 }, { "grad_norm": 0.24819141626358032, "learning_rate": 7.387186233304877e-05, "loss": 0.027, "step": 29960 }, { "grad_norm": 0.2586667239665985, "learning_rate": 7.385369970076212e-05, "loss": 0.0248, "step": 29970 }, { "grad_norm": 0.3626065254211426, "learning_rate": 7.38355329925306e-05, "loss": 0.0317, "step": 29980 }, { "grad_norm": 0.2766560912132263, "learning_rate": 7.381736221145838e-05, "loss": 0.0238, "step": 29990 }, { "grad_norm": 0.9459972977638245, "learning_rate": 7.37991873606504e-05, "loss": 0.025, "step": 30000 }, { "grad_norm": 0.27389073371887207, "learning_rate": 7.378100844321218e-05, "loss": 0.0251, "step": 30010 }, { "grad_norm": 0.6815117597579956, "learning_rate": 7.376282546225004e-05, "loss": 0.0272, "step": 30020 }, { "grad_norm": 0.3695964217185974, "learning_rate": 7.374463842087094e-05, "loss": 0.0221, "step": 30030 }, { "grad_norm": 0.4421423673629761, "learning_rate": 7.372644732218254e-05, "loss": 0.0306, "step": 30040 }, { "grad_norm": 0.24505412578582764, "learning_rate": 7.370825216929322e-05, "loss": 0.0327, "step": 30050 }, { "grad_norm": 0.42830246686935425, "learning_rate": 7.369005296531205e-05, "loss": 0.029, "step": 30060 }, { "grad_norm": 0.21768254041671753, "learning_rate": 7.367184971334873e-05, "loss": 0.024, "step": 30070 }, { "grad_norm": 0.24238570034503937, "learning_rate": 7.365364241651371e-05, "loss": 0.0279, "step": 30080 }, { "grad_norm": 0.33980193734169006, "learning_rate": 7.363543107791815e-05, "loss": 0.0238, "step": 30090 }, { "grad_norm": 0.3820502758026123, "learning_rate": 7.361721570067384e-05, "loss": 0.024, "step": 30100 }, { "grad_norm": 0.528258740901947, "learning_rate": 7.359899628789331e-05, "loss": 0.0341, "step": 30110 }, { "grad_norm": 0.32166948914527893, "learning_rate": 7.358077284268974e-05, "loss": 0.027, "step": 30120 }, { "grad_norm": 0.3998275101184845, "learning_rate": 7.356254536817702e-05, "loss": 0.0308, "step": 30130 }, { "grad_norm": 0.2593115568161011, "learning_rate": 7.354431386746973e-05, "loss": 0.0258, "step": 30140 }, { "grad_norm": 0.2276931256055832, "learning_rate": 7.352607834368316e-05, "loss": 0.026, "step": 30150 }, { "grad_norm": 0.2352541983127594, "learning_rate": 7.350783879993324e-05, "loss": 0.0166, "step": 30160 }, { "grad_norm": 0.2409018725156784, "learning_rate": 7.348959523933658e-05, "loss": 0.0242, "step": 30170 }, { "grad_norm": 0.2996119260787964, "learning_rate": 7.347134766501057e-05, "loss": 0.029, "step": 30180 }, { "grad_norm": 0.24991567432880402, "learning_rate": 7.345309608007315e-05, "loss": 0.0195, "step": 30190 }, { "grad_norm": 0.25022053718566895, "learning_rate": 7.343484048764308e-05, "loss": 0.0266, "step": 30200 }, { "grad_norm": 0.2583811283111572, "learning_rate": 7.341658089083972e-05, "loss": 0.0243, "step": 30210 }, { "grad_norm": 0.2152002900838852, "learning_rate": 7.339831729278313e-05, "loss": 0.0243, "step": 30220 }, { "grad_norm": 0.2859939932823181, "learning_rate": 7.338004969659404e-05, "loss": 0.0213, "step": 30230 }, { "grad_norm": 0.2803565263748169, "learning_rate": 7.336177810539391e-05, "loss": 0.027, "step": 30240 }, { "grad_norm": 0.2867649495601654, "learning_rate": 7.334350252230485e-05, "loss": 0.0254, "step": 30250 }, { "grad_norm": 0.21186266839504242, "learning_rate": 7.332522295044965e-05, "loss": 0.0209, "step": 30260 }, { "grad_norm": 0.41015902161598206, "learning_rate": 7.33069393929518e-05, "loss": 0.0305, "step": 30270 }, { "grad_norm": 0.3867523670196533, "learning_rate": 7.328865185293545e-05, "loss": 0.0425, "step": 30280 }, { "grad_norm": 0.24034950137138367, "learning_rate": 7.327036033352546e-05, "loss": 0.0337, "step": 30290 }, { "grad_norm": 0.38257548213005066, "learning_rate": 7.325206483784733e-05, "loss": 0.0252, "step": 30300 }, { "grad_norm": 0.21409033238887787, "learning_rate": 7.323376536902724e-05, "loss": 0.0328, "step": 30310 }, { "grad_norm": 0.26928797364234924, "learning_rate": 7.321546193019213e-05, "loss": 0.0293, "step": 30320 }, { "grad_norm": 0.3429236114025116, "learning_rate": 7.31971545244695e-05, "loss": 0.0293, "step": 30330 }, { "grad_norm": 0.30124104022979736, "learning_rate": 7.31788431549876e-05, "loss": 0.0275, "step": 30340 }, { "grad_norm": 0.25479793548583984, "learning_rate": 7.316052782487534e-05, "loss": 0.0246, "step": 30350 }, { "grad_norm": 0.2509849965572357, "learning_rate": 7.314220853726234e-05, "loss": 0.0253, "step": 30360 }, { "grad_norm": 0.2614823281764984, "learning_rate": 7.312388529527884e-05, "loss": 0.0288, "step": 30370 }, { "grad_norm": 0.33647507429122925, "learning_rate": 7.310555810205577e-05, "loss": 0.0232, "step": 30380 }, { "grad_norm": 0.24752633273601532, "learning_rate": 7.308722696072476e-05, "loss": 0.0292, "step": 30390 }, { "grad_norm": 0.19837135076522827, "learning_rate": 7.306889187441811e-05, "loss": 0.0242, "step": 30400 }, { "grad_norm": 0.2809811532497406, "learning_rate": 7.305055284626876e-05, "loss": 0.0217, "step": 30410 }, { "grad_norm": 0.22322362661361694, "learning_rate": 7.303220987941037e-05, "loss": 0.024, "step": 30420 }, { "grad_norm": 0.3394823968410492, "learning_rate": 7.301386297697726e-05, "loss": 0.0269, "step": 30430 }, { "grad_norm": 0.2888876795768738, "learning_rate": 7.299551214210438e-05, "loss": 0.0262, "step": 30440 }, { "grad_norm": 0.36535364389419556, "learning_rate": 7.297715737792738e-05, "loss": 0.0288, "step": 30450 }, { "grad_norm": 0.6026020050048828, "learning_rate": 7.295879868758265e-05, "loss": 0.0249, "step": 30460 }, { "grad_norm": 0.5096079707145691, "learning_rate": 7.294043607420713e-05, "loss": 0.0343, "step": 30470 }, { "grad_norm": 0.5078492164611816, "learning_rate": 7.292206954093852e-05, "loss": 0.0289, "step": 30480 }, { "grad_norm": 0.3223264217376709, "learning_rate": 7.290369909091515e-05, "loss": 0.0253, "step": 30490 }, { "grad_norm": 0.3123013973236084, "learning_rate": 7.2885324727276e-05, "loss": 0.0274, "step": 30500 }, { "grad_norm": 0.41317233443260193, "learning_rate": 7.286694645316076e-05, "loss": 0.0251, "step": 30510 }, { "grad_norm": 0.37306246161460876, "learning_rate": 7.284856427170982e-05, "loss": 0.0252, "step": 30520 }, { "grad_norm": 0.4937693178653717, "learning_rate": 7.283017818606414e-05, "loss": 0.0277, "step": 30530 }, { "grad_norm": 0.2909570038318634, "learning_rate": 7.28117881993654e-05, "loss": 0.0258, "step": 30540 }, { "grad_norm": 0.3851203918457031, "learning_rate": 7.279339431475598e-05, "loss": 0.0244, "step": 30550 }, { "grad_norm": 0.36653465032577515, "learning_rate": 7.277499653537887e-05, "loss": 0.0254, "step": 30560 }, { "grad_norm": 0.3622586727142334, "learning_rate": 7.275659486437776e-05, "loss": 0.0314, "step": 30570 }, { "grad_norm": 0.4027950167655945, "learning_rate": 7.273818930489695e-05, "loss": 0.0309, "step": 30580 }, { "grad_norm": 0.5328851342201233, "learning_rate": 7.271977986008151e-05, "loss": 0.0279, "step": 30590 }, { "grad_norm": 0.3970298767089844, "learning_rate": 7.270136653307705e-05, "loss": 0.0243, "step": 30600 }, { "grad_norm": 1.671484112739563, "learning_rate": 7.268294932702994e-05, "loss": 0.0278, "step": 30610 }, { "grad_norm": 0.27532631158828735, "learning_rate": 7.266452824508719e-05, "loss": 0.0271, "step": 30620 }, { "grad_norm": 0.19960442185401917, "learning_rate": 7.264610329039643e-05, "loss": 0.025, "step": 30630 }, { "grad_norm": 0.1841442734003067, "learning_rate": 7.262767446610599e-05, "loss": 0.02, "step": 30640 }, { "grad_norm": 0.2551132142543793, "learning_rate": 7.260924177536485e-05, "loss": 0.0257, "step": 30650 }, { "grad_norm": 0.30051130056381226, "learning_rate": 7.259080522132265e-05, "loss": 0.0246, "step": 30660 }, { "grad_norm": 0.292758971452713, "learning_rate": 7.257236480712972e-05, "loss": 0.0236, "step": 30670 }, { "grad_norm": 0.307275652885437, "learning_rate": 7.255392053593697e-05, "loss": 0.0247, "step": 30680 }, { "grad_norm": 1.0201433897018433, "learning_rate": 7.253547241089607e-05, "loss": 0.0224, "step": 30690 }, { "grad_norm": 0.7107295989990234, "learning_rate": 7.251702043515927e-05, "loss": 0.0292, "step": 30700 }, { "grad_norm": 0.19971364736557007, "learning_rate": 7.249856461187952e-05, "loss": 0.0218, "step": 30710 }, { "grad_norm": 0.2843027412891388, "learning_rate": 7.248010494421042e-05, "loss": 0.0222, "step": 30720 }, { "grad_norm": 0.27901768684387207, "learning_rate": 7.246164143530622e-05, "loss": 0.032, "step": 30730 }, { "grad_norm": 0.30153074860572815, "learning_rate": 7.244317408832181e-05, "loss": 0.0212, "step": 30740 }, { "grad_norm": 0.47625547647476196, "learning_rate": 7.242470290641279e-05, "loss": 0.022, "step": 30750 }, { "grad_norm": 0.2680101990699768, "learning_rate": 7.240622789273536e-05, "loss": 0.0182, "step": 30760 }, { "grad_norm": 0.30243921279907227, "learning_rate": 7.238774905044638e-05, "loss": 0.0251, "step": 30770 }, { "grad_norm": 0.25515440106391907, "learning_rate": 7.236926638270341e-05, "loss": 0.0244, "step": 30780 }, { "grad_norm": 0.31784340739250183, "learning_rate": 7.23507798926646e-05, "loss": 0.0277, "step": 30790 }, { "grad_norm": 0.5384448170661926, "learning_rate": 7.23322895834888e-05, "loss": 0.0244, "step": 30800 }, { "grad_norm": 0.26006704568862915, "learning_rate": 7.231379545833552e-05, "loss": 0.0223, "step": 30810 }, { "grad_norm": 0.23146744072437286, "learning_rate": 7.229529752036487e-05, "loss": 0.0228, "step": 30820 }, { "grad_norm": 0.2745399475097656, "learning_rate": 7.227679577273765e-05, "loss": 0.0236, "step": 30830 }, { "grad_norm": 0.26836875081062317, "learning_rate": 7.225829021861529e-05, "loss": 0.0251, "step": 30840 }, { "grad_norm": 0.32470494508743286, "learning_rate": 7.223978086115992e-05, "loss": 0.0216, "step": 30850 }, { "grad_norm": 0.24190263450145721, "learning_rate": 7.222126770353425e-05, "loss": 0.0249, "step": 30860 }, { "grad_norm": 0.3050483763217926, "learning_rate": 7.22027507489017e-05, "loss": 0.0228, "step": 30870 }, { "grad_norm": 0.3133561313152313, "learning_rate": 7.218423000042627e-05, "loss": 0.0234, "step": 30880 }, { "grad_norm": 0.2095816433429718, "learning_rate": 7.216570546127268e-05, "loss": 0.0213, "step": 30890 }, { "grad_norm": 0.2396795153617859, "learning_rate": 7.214717713460626e-05, "loss": 0.0256, "step": 30900 }, { "grad_norm": 0.4598729908466339, "learning_rate": 7.2128645023593e-05, "loss": 0.0267, "step": 30910 }, { "grad_norm": 0.31448832154273987, "learning_rate": 7.211010913139951e-05, "loss": 0.024, "step": 30920 }, { "grad_norm": 0.2371329367160797, "learning_rate": 7.209156946119308e-05, "loss": 0.0227, "step": 30930 }, { "grad_norm": 0.23870180547237396, "learning_rate": 7.207302601614166e-05, "loss": 0.0257, "step": 30940 }, { "grad_norm": 0.5415692329406738, "learning_rate": 7.205447879941378e-05, "loss": 0.0272, "step": 30950 }, { "grad_norm": 0.24344737827777863, "learning_rate": 7.203592781417866e-05, "loss": 0.0226, "step": 30960 }, { "grad_norm": 0.3282104730606079, "learning_rate": 7.201737306360617e-05, "loss": 0.0273, "step": 30970 }, { "grad_norm": 0.2833203673362732, "learning_rate": 7.19988145508668e-05, "loss": 0.0248, "step": 30980 }, { "grad_norm": 0.2518477439880371, "learning_rate": 7.198025227913168e-05, "loss": 0.0255, "step": 30990 }, { "grad_norm": 0.30333784222602844, "learning_rate": 7.196168625157261e-05, "loss": 0.0269, "step": 31000 }, { "grad_norm": 0.205159991979599, "learning_rate": 7.194311647136201e-05, "loss": 0.0209, "step": 31010 }, { "grad_norm": 0.9167172312736511, "learning_rate": 7.192454294167297e-05, "loss": 0.025, "step": 31020 }, { "grad_norm": 0.282909095287323, "learning_rate": 7.190596566567917e-05, "loss": 0.0252, "step": 31030 }, { "grad_norm": 0.23364375531673431, "learning_rate": 7.188738464655496e-05, "loss": 0.0236, "step": 31040 }, { "grad_norm": 0.186082124710083, "learning_rate": 7.186879988747533e-05, "loss": 0.0246, "step": 31050 }, { "grad_norm": 0.2776058316230774, "learning_rate": 7.185021139161592e-05, "loss": 0.0206, "step": 31060 }, { "grad_norm": 0.47543177008628845, "learning_rate": 7.1831619162153e-05, "loss": 0.0209, "step": 31070 }, { "grad_norm": 0.33816251158714294, "learning_rate": 7.181302320226345e-05, "loss": 0.0191, "step": 31080 }, { "grad_norm": 0.8168628811836243, "learning_rate": 7.179442351512482e-05, "loss": 0.0269, "step": 31090 }, { "grad_norm": 0.3877570629119873, "learning_rate": 7.177582010391528e-05, "loss": 0.0278, "step": 31100 }, { "grad_norm": 0.33299997448921204, "learning_rate": 7.175721297181366e-05, "loss": 0.0351, "step": 31110 }, { "grad_norm": 1.4083143472671509, "learning_rate": 7.173860212199942e-05, "loss": 0.0273, "step": 31120 }, { "grad_norm": 0.2986415922641754, "learning_rate": 7.171998755765263e-05, "loss": 0.0238, "step": 31130 }, { "grad_norm": 0.27040740847587585, "learning_rate": 7.170136928195398e-05, "loss": 0.023, "step": 31140 }, { "grad_norm": 0.5751345753669739, "learning_rate": 7.168274729808489e-05, "loss": 0.0232, "step": 31150 }, { "grad_norm": 0.21142764389514923, "learning_rate": 7.166412160922728e-05, "loss": 0.0242, "step": 31160 }, { "grad_norm": 0.25509122014045715, "learning_rate": 7.164549221856382e-05, "loss": 0.0254, "step": 31170 }, { "grad_norm": 0.2476942539215088, "learning_rate": 7.162685912927775e-05, "loss": 0.0309, "step": 31180 }, { "grad_norm": 0.4567784368991852, "learning_rate": 7.160822234455294e-05, "loss": 0.0277, "step": 31190 }, { "grad_norm": 0.3192633390426636, "learning_rate": 7.158958186757391e-05, "loss": 0.0175, "step": 31200 }, { "grad_norm": 0.357421875, "learning_rate": 7.157093770152582e-05, "loss": 0.0235, "step": 31210 }, { "grad_norm": 0.46314090490341187, "learning_rate": 7.155228984959446e-05, "loss": 0.0245, "step": 31220 }, { "grad_norm": 0.289144903421402, "learning_rate": 7.153363831496621e-05, "loss": 0.0274, "step": 31230 }, { "grad_norm": 0.24790887534618378, "learning_rate": 7.151498310082811e-05, "loss": 0.025, "step": 31240 }, { "grad_norm": 0.32013797760009766, "learning_rate": 7.149632421036784e-05, "loss": 0.0249, "step": 31250 }, { "grad_norm": 0.364799439907074, "learning_rate": 7.147766164677369e-05, "loss": 0.0343, "step": 31260 }, { "grad_norm": 0.23982858657836914, "learning_rate": 7.145899541323459e-05, "loss": 0.0196, "step": 31270 }, { "grad_norm": 0.26482540369033813, "learning_rate": 7.144032551294007e-05, "loss": 0.029, "step": 31280 }, { "grad_norm": 0.2687353193759918, "learning_rate": 7.14216519490803e-05, "loss": 0.0246, "step": 31290 }, { "grad_norm": 0.392220139503479, "learning_rate": 7.140297472484609e-05, "loss": 0.0291, "step": 31300 }, { "grad_norm": 0.25556856393814087, "learning_rate": 7.138429384342891e-05, "loss": 0.0346, "step": 31310 }, { "grad_norm": 0.27247676253318787, "learning_rate": 7.136560930802074e-05, "loss": 0.0219, "step": 31320 }, { "grad_norm": 0.3392590582370758, "learning_rate": 7.134692112181431e-05, "loss": 0.0192, "step": 31330 }, { "grad_norm": 0.8571900129318237, "learning_rate": 7.13282292880029e-05, "loss": 0.0255, "step": 31340 }, { "grad_norm": 0.6365697979927063, "learning_rate": 7.130953380978043e-05, "loss": 0.024, "step": 31350 }, { "grad_norm": 0.2607056200504303, "learning_rate": 7.129083469034144e-05, "loss": 0.0293, "step": 31360 }, { "grad_norm": 0.33357492089271545, "learning_rate": 7.127213193288112e-05, "loss": 0.0313, "step": 31370 }, { "grad_norm": 0.26250162720680237, "learning_rate": 7.125342554059522e-05, "loss": 0.0233, "step": 31380 }, { "grad_norm": 0.30653128027915955, "learning_rate": 7.12347155166802e-05, "loss": 0.0236, "step": 31390 }, { "grad_norm": 0.23753392696380615, "learning_rate": 7.121600186433306e-05, "loss": 0.0208, "step": 31400 }, { "grad_norm": 0.3042334020137787, "learning_rate": 7.119728458675148e-05, "loss": 0.0223, "step": 31410 }, { "grad_norm": 0.35551127791404724, "learning_rate": 7.117856368713369e-05, "loss": 0.0225, "step": 31420 }, { "grad_norm": 0.2948765158653259, "learning_rate": 7.115983916867861e-05, "loss": 0.0206, "step": 31430 }, { "grad_norm": 0.2703913450241089, "learning_rate": 7.114111103458574e-05, "loss": 0.0213, "step": 31440 }, { "grad_norm": 0.204854354262352, "learning_rate": 7.11223792880552e-05, "loss": 0.0213, "step": 31450 }, { "grad_norm": 0.1884782612323761, "learning_rate": 7.110364393228773e-05, "loss": 0.0261, "step": 31460 }, { "grad_norm": 0.34011954069137573, "learning_rate": 7.108490497048471e-05, "loss": 0.0168, "step": 31470 }, { "grad_norm": 0.29552924633026123, "learning_rate": 7.10661624058481e-05, "loss": 0.0242, "step": 31480 }, { "grad_norm": 0.27633610367774963, "learning_rate": 7.10474162415805e-05, "loss": 0.0251, "step": 31490 }, { "grad_norm": 0.25596293807029724, "learning_rate": 7.102866648088511e-05, "loss": 0.0271, "step": 31500 }, { "grad_norm": 0.26042696833610535, "learning_rate": 7.100991312696576e-05, "loss": 0.0213, "step": 31510 }, { "grad_norm": 0.2005808800458908, "learning_rate": 7.099115618302686e-05, "loss": 0.0269, "step": 31520 }, { "grad_norm": 0.20151519775390625, "learning_rate": 7.097239565227349e-05, "loss": 0.033, "step": 31530 }, { "grad_norm": 0.2419082373380661, "learning_rate": 7.09536315379113e-05, "loss": 0.0191, "step": 31540 }, { "grad_norm": 0.30893203616142273, "learning_rate": 7.093486384314656e-05, "loss": 0.0243, "step": 31550 }, { "grad_norm": 0.2918691337108612, "learning_rate": 7.091609257118616e-05, "loss": 0.0226, "step": 31560 }, { "grad_norm": 0.22372880578041077, "learning_rate": 7.08973177252376e-05, "loss": 0.0242, "step": 31570 }, { "grad_norm": 0.46769291162490845, "learning_rate": 7.087853930850898e-05, "loss": 0.0251, "step": 31580 }, { "grad_norm": 0.30796951055526733, "learning_rate": 7.085975732420903e-05, "loss": 0.0205, "step": 31590 }, { "grad_norm": 0.3016993999481201, "learning_rate": 7.084097177554706e-05, "loss": 0.0246, "step": 31600 }, { "grad_norm": 0.31902459263801575, "learning_rate": 7.082218266573301e-05, "loss": 0.0292, "step": 31610 }, { "grad_norm": 0.3818025290966034, "learning_rate": 7.080338999797743e-05, "loss": 0.022, "step": 31620 }, { "grad_norm": 0.3304878771305084, "learning_rate": 7.07845937754915e-05, "loss": 0.0221, "step": 31630 }, { "grad_norm": 0.23393355309963226, "learning_rate": 7.076579400148693e-05, "loss": 0.0222, "step": 31640 }, { "grad_norm": 0.21316684782505035, "learning_rate": 7.074699067917611e-05, "loss": 0.0277, "step": 31650 }, { "grad_norm": 0.2785017192363739, "learning_rate": 7.072818381177201e-05, "loss": 0.0262, "step": 31660 }, { "grad_norm": 0.23562932014465332, "learning_rate": 7.070937340248823e-05, "loss": 0.0232, "step": 31670 }, { "grad_norm": 0.2674501836299896, "learning_rate": 7.069055945453893e-05, "loss": 0.0253, "step": 31680 }, { "grad_norm": 0.2953905761241913, "learning_rate": 7.067174197113892e-05, "loss": 0.0187, "step": 31690 }, { "grad_norm": 0.32713475823402405, "learning_rate": 7.065292095550355e-05, "loss": 0.0244, "step": 31700 }, { "grad_norm": 0.27277159690856934, "learning_rate": 7.063409641084887e-05, "loss": 0.0219, "step": 31710 }, { "grad_norm": 0.2986147701740265, "learning_rate": 7.061526834039145e-05, "loss": 0.028, "step": 31720 }, { "grad_norm": 0.3416382372379303, "learning_rate": 7.05964367473485e-05, "loss": 0.02, "step": 31730 }, { "grad_norm": 0.24644827842712402, "learning_rate": 7.057760163493783e-05, "loss": 0.0267, "step": 31740 }, { "grad_norm": 0.5530922412872314, "learning_rate": 7.055876300637783e-05, "loss": 0.0229, "step": 31750 }, { "grad_norm": 0.336514413356781, "learning_rate": 7.053992086488753e-05, "loss": 0.0267, "step": 31760 }, { "grad_norm": 0.44128111004829407, "learning_rate": 7.052107521368651e-05, "loss": 0.0312, "step": 31770 }, { "grad_norm": 0.2570568323135376, "learning_rate": 7.0502226055995e-05, "loss": 0.0301, "step": 31780 }, { "grad_norm": 0.3385563790798187, "learning_rate": 7.048337339503379e-05, "loss": 0.0387, "step": 31790 }, { "grad_norm": 0.2852681279182434, "learning_rate": 7.046451723402427e-05, "loss": 0.0218, "step": 31800 }, { "grad_norm": 0.3206480145454407, "learning_rate": 7.044565757618848e-05, "loss": 0.0344, "step": 31810 }, { "grad_norm": 0.257396936416626, "learning_rate": 7.042679442474899e-05, "loss": 0.0194, "step": 31820 }, { "grad_norm": 0.23583875596523285, "learning_rate": 7.040792778292902e-05, "loss": 0.0328, "step": 31830 }, { "grad_norm": 0.25370946526527405, "learning_rate": 7.038905765395234e-05, "loss": 0.0294, "step": 31840 }, { "grad_norm": 0.2700718641281128, "learning_rate": 7.037018404104334e-05, "loss": 0.027, "step": 31850 }, { "grad_norm": 0.29403263330459595, "learning_rate": 7.035130694742702e-05, "loss": 0.0231, "step": 31860 }, { "grad_norm": 0.2982371151447296, "learning_rate": 7.033242637632897e-05, "loss": 0.0234, "step": 31870 }, { "grad_norm": 0.31901854276657104, "learning_rate": 7.031354233097534e-05, "loss": 0.0219, "step": 31880 }, { "grad_norm": 0.3536297380924225, "learning_rate": 7.029465481459289e-05, "loss": 0.0365, "step": 31890 }, { "grad_norm": 0.4436925947666168, "learning_rate": 7.027576383040898e-05, "loss": 0.0234, "step": 31900 }, { "grad_norm": 0.2693318724632263, "learning_rate": 7.025686938165159e-05, "loss": 0.0271, "step": 31910 }, { "grad_norm": 0.2877466082572937, "learning_rate": 7.023797147154924e-05, "loss": 0.0338, "step": 31920 }, { "grad_norm": 0.26130178570747375, "learning_rate": 7.021907010333111e-05, "loss": 0.0223, "step": 31930 }, { "grad_norm": 0.254723459482193, "learning_rate": 7.020016528022685e-05, "loss": 0.0215, "step": 31940 }, { "grad_norm": 0.2740348279476166, "learning_rate": 7.018125700546683e-05, "loss": 0.0262, "step": 31950 }, { "grad_norm": 0.30410367250442505, "learning_rate": 7.016234528228196e-05, "loss": 0.0283, "step": 31960 }, { "grad_norm": 0.27202191948890686, "learning_rate": 7.014343011390372e-05, "loss": 0.022, "step": 31970 }, { "grad_norm": 0.33337217569351196, "learning_rate": 7.01245115035642e-05, "loss": 0.0205, "step": 31980 }, { "grad_norm": 0.3329049348831177, "learning_rate": 7.010558945449606e-05, "loss": 0.0328, "step": 31990 }, { "grad_norm": 0.22892776131629944, "learning_rate": 7.008666396993258e-05, "loss": 0.0214, "step": 32000 }, { "grad_norm": 0.32985302805900574, "learning_rate": 7.006773505310759e-05, "loss": 0.0252, "step": 32010 }, { "grad_norm": 0.29622411727905273, "learning_rate": 7.004880270725553e-05, "loss": 0.0256, "step": 32020 }, { "grad_norm": 0.34058165550231934, "learning_rate": 7.002986693561144e-05, "loss": 0.0301, "step": 32030 }, { "grad_norm": 0.28127580881118774, "learning_rate": 7.001092774141089e-05, "loss": 0.0274, "step": 32040 }, { "grad_norm": 0.29072776436805725, "learning_rate": 6.999198512789009e-05, "loss": 0.0255, "step": 32050 }, { "grad_norm": 0.4101502597332001, "learning_rate": 6.997303909828584e-05, "loss": 0.0312, "step": 32060 }, { "grad_norm": 0.8565406203269958, "learning_rate": 6.995408965583544e-05, "loss": 0.0281, "step": 32070 }, { "grad_norm": 0.2852315306663513, "learning_rate": 6.993513680377688e-05, "loss": 0.0209, "step": 32080 }, { "grad_norm": 0.44094085693359375, "learning_rate": 6.991618054534868e-05, "loss": 0.025, "step": 32090 }, { "grad_norm": 0.543205738067627, "learning_rate": 6.989722088378991e-05, "loss": 0.0241, "step": 32100 }, { "grad_norm": 0.275343656539917, "learning_rate": 6.987825782234027e-05, "loss": 0.0227, "step": 32110 }, { "grad_norm": 0.3745787441730499, "learning_rate": 6.985929136424006e-05, "loss": 0.0242, "step": 32120 }, { "grad_norm": 0.3613685071468353, "learning_rate": 6.984032151273012e-05, "loss": 0.0211, "step": 32130 }, { "grad_norm": 0.20946359634399414, "learning_rate": 6.982134827105186e-05, "loss": 0.0215, "step": 32140 }, { "grad_norm": 0.28680679202079773, "learning_rate": 6.980237164244729e-05, "loss": 0.0225, "step": 32150 }, { "grad_norm": 0.3364614248275757, "learning_rate": 6.9783391630159e-05, "loss": 0.0255, "step": 32160 }, { "grad_norm": 0.28317150473594666, "learning_rate": 6.976440823743015e-05, "loss": 0.0284, "step": 32170 }, { "grad_norm": 0.2997041940689087, "learning_rate": 6.974542146750451e-05, "loss": 0.0213, "step": 32180 }, { "grad_norm": 0.23450569808483124, "learning_rate": 6.972643132362637e-05, "loss": 0.02, "step": 32190 }, { "grad_norm": 0.30929839611053467, "learning_rate": 6.970743780904064e-05, "loss": 0.0212, "step": 32200 }, { "grad_norm": 0.28332123160362244, "learning_rate": 6.968844092699277e-05, "loss": 0.0253, "step": 32210 }, { "grad_norm": 0.694901704788208, "learning_rate": 6.966944068072883e-05, "loss": 0.0256, "step": 32220 }, { "grad_norm": 0.7389973998069763, "learning_rate": 6.965043707349545e-05, "loss": 0.0225, "step": 32230 }, { "grad_norm": 0.2826254069805145, "learning_rate": 6.963143010853982e-05, "loss": 0.0249, "step": 32240 }, { "grad_norm": 0.4832274913787842, "learning_rate": 6.961241978910971e-05, "loss": 0.0262, "step": 32250 }, { "grad_norm": 0.22951507568359375, "learning_rate": 6.959340611845344e-05, "loss": 0.0243, "step": 32260 }, { "grad_norm": 0.3355550467967987, "learning_rate": 6.957438909981995e-05, "loss": 0.0251, "step": 32270 }, { "grad_norm": 0.23193243145942688, "learning_rate": 6.955536873645872e-05, "loss": 0.0227, "step": 32280 }, { "grad_norm": 0.6759763956069946, "learning_rate": 6.953634503161982e-05, "loss": 0.0242, "step": 32290 }, { "grad_norm": 0.38024815917015076, "learning_rate": 6.951731798855387e-05, "loss": 0.0228, "step": 32300 }, { "grad_norm": 0.32390096783638, "learning_rate": 6.949828761051208e-05, "loss": 0.0283, "step": 32310 }, { "grad_norm": 0.29058700799942017, "learning_rate": 6.947925390074622e-05, "loss": 0.0184, "step": 32320 }, { "grad_norm": 0.33102473616600037, "learning_rate": 6.946021686250863e-05, "loss": 0.0247, "step": 32330 }, { "grad_norm": 0.6934410333633423, "learning_rate": 6.94411764990522e-05, "loss": 0.0268, "step": 32340 }, { "grad_norm": 0.26402199268341064, "learning_rate": 6.942213281363044e-05, "loss": 0.0199, "step": 32350 }, { "grad_norm": 0.22647921741008759, "learning_rate": 6.940308580949737e-05, "loss": 0.0207, "step": 32360 }, { "grad_norm": 0.2607855796813965, "learning_rate": 6.93840354899076e-05, "loss": 0.0207, "step": 32370 }, { "grad_norm": 0.2809682786464691, "learning_rate": 6.936498185811633e-05, "loss": 0.021, "step": 32380 }, { "grad_norm": 0.19603101909160614, "learning_rate": 6.93459249173793e-05, "loss": 0.0168, "step": 32390 }, { "grad_norm": 0.2785611152648926, "learning_rate": 6.932686467095279e-05, "loss": 0.0208, "step": 32400 }, { "grad_norm": 0.2849911153316498, "learning_rate": 6.930780112209373e-05, "loss": 0.0214, "step": 32410 }, { "grad_norm": 0.2937357723712921, "learning_rate": 6.92887342740595e-05, "loss": 0.0221, "step": 32420 }, { "grad_norm": 0.28679877519607544, "learning_rate": 6.926966413010816e-05, "loss": 0.0263, "step": 32430 }, { "grad_norm": 0.2582932710647583, "learning_rate": 6.925059069349824e-05, "loss": 0.0223, "step": 32440 }, { "grad_norm": 0.3387605547904968, "learning_rate": 6.923151396748886e-05, "loss": 0.024, "step": 32450 }, { "grad_norm": 0.21073228120803833, "learning_rate": 6.921243395533974e-05, "loss": 0.0185, "step": 32460 }, { "grad_norm": 0.30891212821006775, "learning_rate": 6.919335066031109e-05, "loss": 0.0257, "step": 32470 }, { "grad_norm": 0.2918630838394165, "learning_rate": 6.917426408566379e-05, "loss": 0.0238, "step": 32480 }, { "grad_norm": 0.36708804965019226, "learning_rate": 6.915517423465916e-05, "loss": 0.0209, "step": 32490 }, { "grad_norm": 0.22925178706645966, "learning_rate": 6.913608111055914e-05, "loss": 0.0183, "step": 32500 }, { "grad_norm": 0.25405773520469666, "learning_rate": 6.911698471662623e-05, "loss": 0.02, "step": 32510 }, { "grad_norm": 0.42967623472213745, "learning_rate": 6.90978850561235e-05, "loss": 0.0175, "step": 32520 }, { "grad_norm": 0.2405688613653183, "learning_rate": 6.907878213231454e-05, "loss": 0.0164, "step": 32530 }, { "grad_norm": 0.28148654103279114, "learning_rate": 6.90596759484635e-05, "loss": 0.0191, "step": 32540 }, { "grad_norm": 0.30074572563171387, "learning_rate": 6.904056650783514e-05, "loss": 0.027, "step": 32550 }, { "grad_norm": 0.5946173071861267, "learning_rate": 6.902145381369471e-05, "loss": 0.0285, "step": 32560 }, { "grad_norm": 0.3423583507537842, "learning_rate": 6.900233786930808e-05, "loss": 0.0233, "step": 32570 }, { "grad_norm": 0.3121476173400879, "learning_rate": 6.898321867794161e-05, "loss": 0.0231, "step": 32580 }, { "grad_norm": 0.29734745621681213, "learning_rate": 6.896409624286226e-05, "loss": 0.024, "step": 32590 }, { "grad_norm": 0.2199058085680008, "learning_rate": 6.894497056733754e-05, "loss": 0.0238, "step": 32600 }, { "grad_norm": 0.33480679988861084, "learning_rate": 6.89258416546355e-05, "loss": 0.0201, "step": 32610 }, { "grad_norm": 0.2927069365978241, "learning_rate": 6.890670950802474e-05, "loss": 0.0228, "step": 32620 }, { "grad_norm": 0.26908013224601746, "learning_rate": 6.88875741307744e-05, "loss": 0.0188, "step": 32630 }, { "grad_norm": 0.2974194884300232, "learning_rate": 6.886843552615425e-05, "loss": 0.0231, "step": 32640 }, { "grad_norm": 0.2896302342414856, "learning_rate": 6.884929369743451e-05, "loss": 0.0217, "step": 32650 }, { "grad_norm": 0.363489031791687, "learning_rate": 6.8830148647886e-05, "loss": 0.0203, "step": 32660 }, { "grad_norm": 0.3721354603767395, "learning_rate": 6.88110003807801e-05, "loss": 0.0203, "step": 32670 }, { "grad_norm": 0.2922886908054352, "learning_rate": 6.87918488993887e-05, "loss": 0.0242, "step": 32680 }, { "grad_norm": 0.31953445076942444, "learning_rate": 6.877269420698431e-05, "loss": 0.0292, "step": 32690 }, { "grad_norm": 0.31789204478263855, "learning_rate": 6.875353630683989e-05, "loss": 0.0324, "step": 32700 }, { "grad_norm": 0.35056596994400024, "learning_rate": 6.873437520222905e-05, "loss": 0.0293, "step": 32710 }, { "grad_norm": 0.3301660418510437, "learning_rate": 6.871521089642585e-05, "loss": 0.0218, "step": 32720 }, { "grad_norm": 0.29770588874816895, "learning_rate": 6.869604339270498e-05, "loss": 0.0207, "step": 32730 }, { "grad_norm": 0.20585696399211884, "learning_rate": 6.867687269434164e-05, "loss": 0.023, "step": 32740 }, { "grad_norm": 0.41607666015625, "learning_rate": 6.865769880461156e-05, "loss": 0.0253, "step": 32750 }, { "grad_norm": 0.26584523916244507, "learning_rate": 6.863852172679104e-05, "loss": 0.0195, "step": 32760 }, { "grad_norm": 0.4419514834880829, "learning_rate": 6.861934146415693e-05, "loss": 0.0233, "step": 32770 }, { "grad_norm": 0.3053850829601288, "learning_rate": 6.86001580199866e-05, "loss": 0.0284, "step": 32780 }, { "grad_norm": 0.36195674538612366, "learning_rate": 6.858097139755798e-05, "loss": 0.0182, "step": 32790 }, { "grad_norm": 0.2548736035823822, "learning_rate": 6.856178160014955e-05, "loss": 0.023, "step": 32800 }, { "grad_norm": 0.3851962983608246, "learning_rate": 6.85425886310403e-05, "loss": 0.0216, "step": 32810 }, { "grad_norm": 0.4361734986305237, "learning_rate": 6.852339249350979e-05, "loss": 0.0227, "step": 32820 }, { "grad_norm": 0.9718189239501953, "learning_rate": 6.850419319083812e-05, "loss": 0.0358, "step": 32830 }, { "grad_norm": 0.3763953745365143, "learning_rate": 6.848499072630592e-05, "loss": 0.0212, "step": 32840 }, { "grad_norm": 0.6978529095649719, "learning_rate": 6.846578510319439e-05, "loss": 0.0273, "step": 32850 }, { "grad_norm": 1.1593654155731201, "learning_rate": 6.844657632478519e-05, "loss": 0.0323, "step": 32860 }, { "grad_norm": 0.31184500455856323, "learning_rate": 6.842736439436063e-05, "loss": 0.02, "step": 32870 }, { "grad_norm": 0.2915257513523102, "learning_rate": 6.84081493152035e-05, "loss": 0.0232, "step": 32880 }, { "grad_norm": 0.3788963854312897, "learning_rate": 6.83889310905971e-05, "loss": 0.0225, "step": 32890 }, { "grad_norm": 0.3314165771007538, "learning_rate": 6.836970972382533e-05, "loss": 0.0251, "step": 32900 }, { "grad_norm": 0.2893258333206177, "learning_rate": 6.835048521817257e-05, "loss": 0.0227, "step": 32910 }, { "grad_norm": 0.2820485830307007, "learning_rate": 6.833125757692379e-05, "loss": 0.0249, "step": 32920 }, { "grad_norm": 0.3136429786682129, "learning_rate": 6.831202680336441e-05, "loss": 0.0192, "step": 32930 }, { "grad_norm": 0.19443419575691223, "learning_rate": 6.829279290078052e-05, "loss": 0.0189, "step": 32940 }, { "grad_norm": 0.25136521458625793, "learning_rate": 6.827355587245863e-05, "loss": 0.0225, "step": 32950 }, { "grad_norm": 0.5385200381278992, "learning_rate": 6.82543157216858e-05, "loss": 0.0219, "step": 32960 }, { "grad_norm": 0.30837857723236084, "learning_rate": 6.823507245174969e-05, "loss": 0.0217, "step": 32970 }, { "grad_norm": 0.24569618701934814, "learning_rate": 6.821582606593841e-05, "loss": 0.0197, "step": 32980 }, { "grad_norm": 0.29099440574645996, "learning_rate": 6.81965765675407e-05, "loss": 0.0223, "step": 32990 }, { "grad_norm": 0.23774972558021545, "learning_rate": 6.81773239598457e-05, "loss": 0.0212, "step": 33000 }, { "grad_norm": 0.31890103220939636, "learning_rate": 6.815806824614319e-05, "loss": 0.0218, "step": 33010 }, { "grad_norm": 0.2125491350889206, "learning_rate": 6.813880942972343e-05, "loss": 0.0177, "step": 33020 }, { "grad_norm": 0.1870664656162262, "learning_rate": 6.811954751387726e-05, "loss": 0.0202, "step": 33030 }, { "grad_norm": 0.31217631697654724, "learning_rate": 6.810028250189598e-05, "loss": 0.0192, "step": 33040 }, { "grad_norm": 0.3524424433708191, "learning_rate": 6.808101439707147e-05, "loss": 0.0301, "step": 33050 }, { "grad_norm": 0.21401003003120422, "learning_rate": 6.806174320269609e-05, "loss": 0.0241, "step": 33060 }, { "grad_norm": 0.3202325701713562, "learning_rate": 6.804246892206281e-05, "loss": 0.0241, "step": 33070 }, { "grad_norm": 0.27461379766464233, "learning_rate": 6.802319155846506e-05, "loss": 0.0398, "step": 33080 }, { "grad_norm": 0.2696029245853424, "learning_rate": 6.800391111519679e-05, "loss": 0.0211, "step": 33090 }, { "grad_norm": 0.25369659066200256, "learning_rate": 6.798462759555253e-05, "loss": 0.0314, "step": 33100 }, { "grad_norm": 0.23422732949256897, "learning_rate": 6.79653410028273e-05, "loss": 0.0265, "step": 33110 }, { "grad_norm": 0.2548418641090393, "learning_rate": 6.794605134031663e-05, "loss": 0.0216, "step": 33120 }, { "grad_norm": 0.2126566618680954, "learning_rate": 6.792675861131661e-05, "loss": 0.0225, "step": 33130 }, { "grad_norm": 0.3764708340167999, "learning_rate": 6.790746281912386e-05, "loss": 0.0298, "step": 33140 }, { "grad_norm": 0.3023393452167511, "learning_rate": 6.788816396703546e-05, "loss": 0.0185, "step": 33150 }, { "grad_norm": 0.4510969817638397, "learning_rate": 6.78688620583491e-05, "loss": 0.0233, "step": 33160 }, { "grad_norm": 0.22872790694236755, "learning_rate": 6.784955709636292e-05, "loss": 0.021, "step": 33170 }, { "grad_norm": 0.33008116483688354, "learning_rate": 6.783024908437564e-05, "loss": 0.024, "step": 33180 }, { "grad_norm": 0.21928206086158752, "learning_rate": 6.781093802568641e-05, "loss": 0.0163, "step": 33190 }, { "grad_norm": 0.24278032779693604, "learning_rate": 6.779162392359504e-05, "loss": 0.0221, "step": 33200 }, { "grad_norm": 0.3652735948562622, "learning_rate": 6.777230678140172e-05, "loss": 0.0224, "step": 33210 }, { "grad_norm": 0.25000879168510437, "learning_rate": 6.775298660240726e-05, "loss": 0.0205, "step": 33220 }, { "grad_norm": 0.6060812473297119, "learning_rate": 6.773366338991292e-05, "loss": 0.0323, "step": 33230 }, { "grad_norm": 0.22361308336257935, "learning_rate": 6.771433714722052e-05, "loss": 0.0192, "step": 33240 }, { "grad_norm": 0.30831125378608704, "learning_rate": 6.769500787763239e-05, "loss": 0.0288, "step": 33250 }, { "grad_norm": 0.22486820816993713, "learning_rate": 6.76756755844514e-05, "loss": 0.0236, "step": 33260 }, { "grad_norm": 0.2546765208244324, "learning_rate": 6.765634027098087e-05, "loss": 0.0224, "step": 33270 }, { "grad_norm": 0.29028522968292236, "learning_rate": 6.763700194052468e-05, "loss": 0.0233, "step": 33280 }, { "grad_norm": 0.1907869279384613, "learning_rate": 6.761766059638723e-05, "loss": 0.0237, "step": 33290 }, { "grad_norm": 0.28563636541366577, "learning_rate": 6.759831624187345e-05, "loss": 0.0255, "step": 33300 }, { "grad_norm": 0.3349626958370209, "learning_rate": 6.757896888028871e-05, "loss": 0.0264, "step": 33310 }, { "grad_norm": 0.6285145282745361, "learning_rate": 6.7559618514939e-05, "loss": 0.0232, "step": 33320 }, { "grad_norm": 0.39355698227882385, "learning_rate": 6.754026514913073e-05, "loss": 0.0216, "step": 33330 }, { "grad_norm": 0.26565253734588623, "learning_rate": 6.752090878617087e-05, "loss": 0.0258, "step": 33340 }, { "grad_norm": 0.3426927328109741, "learning_rate": 6.75015494293669e-05, "loss": 0.0203, "step": 33350 }, { "grad_norm": 0.35656481981277466, "learning_rate": 6.74821870820268e-05, "loss": 0.0236, "step": 33360 }, { "grad_norm": 0.34881895780563354, "learning_rate": 6.746282174745907e-05, "loss": 0.0263, "step": 33370 }, { "grad_norm": 0.36703288555145264, "learning_rate": 6.744345342897271e-05, "loss": 0.0251, "step": 33380 }, { "grad_norm": 0.677095890045166, "learning_rate": 6.742408212987724e-05, "loss": 0.0263, "step": 33390 }, { "grad_norm": 0.2338668555021286, "learning_rate": 6.740470785348269e-05, "loss": 0.0184, "step": 33400 }, { "grad_norm": 0.22287242114543915, "learning_rate": 6.738533060309958e-05, "loss": 0.0248, "step": 33410 }, { "grad_norm": 0.272747278213501, "learning_rate": 6.736595038203894e-05, "loss": 0.0234, "step": 33420 }, { "grad_norm": 0.3452949821949005, "learning_rate": 6.734656719361236e-05, "loss": 0.0211, "step": 33430 }, { "grad_norm": 0.40147167444229126, "learning_rate": 6.732718104113189e-05, "loss": 0.0233, "step": 33440 }, { "grad_norm": 0.26475781202316284, "learning_rate": 6.730779192791006e-05, "loss": 0.021, "step": 33450 }, { "grad_norm": 0.5643903017044067, "learning_rate": 6.728839985725997e-05, "loss": 0.0228, "step": 33460 }, { "grad_norm": 0.43107181787490845, "learning_rate": 6.726900483249517e-05, "loss": 0.0214, "step": 33470 }, { "grad_norm": 0.24889886379241943, "learning_rate": 6.724960685692976e-05, "loss": 0.0221, "step": 33480 }, { "grad_norm": 0.2825273275375366, "learning_rate": 6.723020593387833e-05, "loss": 0.0202, "step": 33490 }, { "grad_norm": 0.20899413526058197, "learning_rate": 6.721080206665593e-05, "loss": 0.023, "step": 33500 }, { "grad_norm": 0.23009638488292694, "learning_rate": 6.719139525857819e-05, "loss": 0.0257, "step": 33510 }, { "grad_norm": 0.2660304009914398, "learning_rate": 6.717198551296117e-05, "loss": 0.0238, "step": 33520 }, { "grad_norm": 0.36907958984375, "learning_rate": 6.715257283312148e-05, "loss": 0.0222, "step": 33530 }, { "grad_norm": 0.2890045642852783, "learning_rate": 6.713315722237623e-05, "loss": 0.0235, "step": 33540 }, { "grad_norm": 0.21258346736431122, "learning_rate": 6.7113738684043e-05, "loss": 0.0245, "step": 33550 }, { "grad_norm": 0.30903661251068115, "learning_rate": 6.709431722143989e-05, "loss": 0.0239, "step": 33560 }, { "grad_norm": 0.3486895263195038, "learning_rate": 6.70748928378855e-05, "loss": 0.0278, "step": 33570 }, { "grad_norm": 0.34958064556121826, "learning_rate": 6.705546553669891e-05, "loss": 0.0173, "step": 33580 }, { "grad_norm": 0.2658672034740448, "learning_rate": 6.703603532119974e-05, "loss": 0.0152, "step": 33590 }, { "grad_norm": 0.21153725683689117, "learning_rate": 6.701660219470808e-05, "loss": 0.0195, "step": 33600 }, { "grad_norm": 0.30384358763694763, "learning_rate": 6.69971661605445e-05, "loss": 0.0212, "step": 33610 }, { "grad_norm": 1.1332987546920776, "learning_rate": 6.697772722203008e-05, "loss": 0.0273, "step": 33620 }, { "grad_norm": 0.253218412399292, "learning_rate": 6.695828538248643e-05, "loss": 0.0242, "step": 33630 }, { "grad_norm": 0.24700447916984558, "learning_rate": 6.693884064523563e-05, "loss": 0.0274, "step": 33640 }, { "grad_norm": 0.2697051167488098, "learning_rate": 6.691939301360023e-05, "loss": 0.0217, "step": 33650 }, { "grad_norm": 1.2940987348556519, "learning_rate": 6.689994249090333e-05, "loss": 0.0294, "step": 33660 }, { "grad_norm": 0.32214587926864624, "learning_rate": 6.688048908046845e-05, "loss": 0.0262, "step": 33670 }, { "grad_norm": 0.620838463306427, "learning_rate": 6.686103278561969e-05, "loss": 0.0199, "step": 33680 }, { "grad_norm": 0.22538258135318756, "learning_rate": 6.684157360968156e-05, "loss": 0.0238, "step": 33690 }, { "grad_norm": 0.24241101741790771, "learning_rate": 6.682211155597911e-05, "loss": 0.018, "step": 33700 }, { "grad_norm": 0.18114928901195526, "learning_rate": 6.680264662783789e-05, "loss": 0.0247, "step": 33710 }, { "grad_norm": 0.3400563597679138, "learning_rate": 6.678317882858391e-05, "loss": 0.0292, "step": 33720 }, { "grad_norm": 0.3408403694629669, "learning_rate": 6.67637081615437e-05, "loss": 0.024, "step": 33730 }, { "grad_norm": 0.235420823097229, "learning_rate": 6.674423463004427e-05, "loss": 0.0184, "step": 33740 }, { "grad_norm": 0.37490808963775635, "learning_rate": 6.672475823741308e-05, "loss": 0.0231, "step": 33750 }, { "grad_norm": 0.28938788175582886, "learning_rate": 6.670527898697811e-05, "loss": 0.0249, "step": 33760 }, { "grad_norm": 0.28023234009742737, "learning_rate": 6.668579688206788e-05, "loss": 0.0252, "step": 33770 }, { "grad_norm": 0.2683209776878357, "learning_rate": 6.666631192601131e-05, "loss": 0.0186, "step": 33780 }, { "grad_norm": 0.2361072301864624, "learning_rate": 6.664682412213785e-05, "loss": 0.0189, "step": 33790 }, { "grad_norm": 0.25504353642463684, "learning_rate": 6.662733347377745e-05, "loss": 0.0218, "step": 33800 }, { "grad_norm": 0.3428136706352234, "learning_rate": 6.660783998426051e-05, "loss": 0.0195, "step": 33810 }, { "grad_norm": 0.4698343276977539, "learning_rate": 6.658834365691794e-05, "loss": 0.0216, "step": 33820 }, { "grad_norm": 0.2604614496231079, "learning_rate": 6.656884449508115e-05, "loss": 0.0231, "step": 33830 }, { "grad_norm": 0.3066045939922333, "learning_rate": 6.654934250208198e-05, "loss": 0.0257, "step": 33840 }, { "grad_norm": 0.22354789078235626, "learning_rate": 6.65298376812528e-05, "loss": 0.0201, "step": 33850 }, { "grad_norm": 0.4542563855648041, "learning_rate": 6.651033003592646e-05, "loss": 0.0254, "step": 33860 }, { "grad_norm": 0.22095048427581787, "learning_rate": 6.649081956943626e-05, "loss": 0.0206, "step": 33870 }, { "grad_norm": 0.21433000266551971, "learning_rate": 6.647130628511604e-05, "loss": 0.0202, "step": 33880 }, { "grad_norm": 0.31726932525634766, "learning_rate": 6.645179018630005e-05, "loss": 0.0201, "step": 33890 }, { "grad_norm": 0.2033764123916626, "learning_rate": 6.643227127632309e-05, "loss": 0.0195, "step": 33900 }, { "grad_norm": 0.3238867223262787, "learning_rate": 6.641274955852038e-05, "loss": 0.0224, "step": 33910 }, { "grad_norm": 0.3724110424518585, "learning_rate": 6.639322503622768e-05, "loss": 0.0387, "step": 33920 }, { "grad_norm": 0.26058104634284973, "learning_rate": 6.637369771278116e-05, "loss": 0.0211, "step": 33930 }, { "grad_norm": 0.21747203171253204, "learning_rate": 6.635416759151751e-05, "loss": 0.0193, "step": 33940 }, { "grad_norm": 0.25562214851379395, "learning_rate": 6.633463467577394e-05, "loss": 0.0372, "step": 33950 }, { "grad_norm": 0.657529354095459, "learning_rate": 6.631509896888803e-05, "loss": 0.0233, "step": 33960 }, { "grad_norm": 0.3137167692184448, "learning_rate": 6.629556047419794e-05, "loss": 0.0251, "step": 33970 }, { "grad_norm": 0.29868456721305847, "learning_rate": 6.627601919504223e-05, "loss": 0.0226, "step": 33980 }, { "grad_norm": 0.5277777910232544, "learning_rate": 6.625647513476001e-05, "loss": 0.0234, "step": 33990 }, { "grad_norm": 0.37495866417884827, "learning_rate": 6.62369282966908e-05, "loss": 0.0289, "step": 34000 }, { "grad_norm": 0.3742222487926483, "learning_rate": 6.621737868417464e-05, "loss": 0.0215, "step": 34010 }, { "grad_norm": 0.3674008250236511, "learning_rate": 6.619782630055198e-05, "loss": 0.02, "step": 34020 }, { "grad_norm": 0.2314121574163437, "learning_rate": 6.617827114916382e-05, "loss": 0.0189, "step": 34030 }, { "grad_norm": 0.23329581320285797, "learning_rate": 6.615871323335161e-05, "loss": 0.0256, "step": 34040 }, { "grad_norm": 0.2926411032676697, "learning_rate": 6.613915255645725e-05, "loss": 0.0261, "step": 34050 }, { "grad_norm": 0.3005077838897705, "learning_rate": 6.611958912182312e-05, "loss": 0.0208, "step": 34060 }, { "grad_norm": 0.27819281816482544, "learning_rate": 6.610002293279207e-05, "loss": 0.0208, "step": 34070 }, { "grad_norm": 0.41657552123069763, "learning_rate": 6.608045399270746e-05, "loss": 0.0204, "step": 34080 }, { "grad_norm": 0.2681131064891815, "learning_rate": 6.606088230491304e-05, "loss": 0.0206, "step": 34090 }, { "grad_norm": 0.3155236542224884, "learning_rate": 6.604130787275312e-05, "loss": 0.0251, "step": 34100 }, { "grad_norm": 0.34873247146606445, "learning_rate": 6.602173069957242e-05, "loss": 0.0223, "step": 34110 }, { "grad_norm": 0.2847839891910553, "learning_rate": 6.600215078871612e-05, "loss": 0.0229, "step": 34120 }, { "grad_norm": 0.30900606513023376, "learning_rate": 6.598256814352992e-05, "loss": 0.0212, "step": 34130 }, { "grad_norm": 0.3303220272064209, "learning_rate": 6.596298276735995e-05, "loss": 0.0214, "step": 34140 }, { "grad_norm": 0.2199457585811615, "learning_rate": 6.594339466355282e-05, "loss": 0.0243, "step": 34150 }, { "grad_norm": 0.27992409467697144, "learning_rate": 6.592380383545558e-05, "loss": 0.0213, "step": 34160 }, { "grad_norm": 0.42063799500465393, "learning_rate": 6.590421028641577e-05, "loss": 0.0295, "step": 34170 }, { "grad_norm": 0.23037926852703094, "learning_rate": 6.588461401978143e-05, "loss": 0.0219, "step": 34180 }, { "grad_norm": 0.2818238139152527, "learning_rate": 6.586501503890099e-05, "loss": 0.0197, "step": 34190 }, { "grad_norm": 0.26451826095581055, "learning_rate": 6.584541334712338e-05, "loss": 0.0229, "step": 34200 }, { "grad_norm": 0.24964895844459534, "learning_rate": 6.582580894779802e-05, "loss": 0.0209, "step": 34210 }, { "grad_norm": 0.24989980459213257, "learning_rate": 6.580620184427473e-05, "loss": 0.0252, "step": 34220 }, { "grad_norm": 0.35059359669685364, "learning_rate": 6.578659203990385e-05, "loss": 0.0251, "step": 34230 }, { "grad_norm": 0.22977367043495178, "learning_rate": 6.576697953803615e-05, "loss": 0.0232, "step": 34240 }, { "grad_norm": 0.2849329710006714, "learning_rate": 6.57473643420229e-05, "loss": 0.0262, "step": 34250 }, { "grad_norm": 0.27056822180747986, "learning_rate": 6.572774645521574e-05, "loss": 0.027, "step": 34260 }, { "grad_norm": 0.35995152592658997, "learning_rate": 6.570812588096688e-05, "loss": 0.0238, "step": 34270 }, { "grad_norm": 0.37563687562942505, "learning_rate": 6.568850262262893e-05, "loss": 0.0192, "step": 34280 }, { "grad_norm": 0.3807852268218994, "learning_rate": 6.566887668355497e-05, "loss": 0.0221, "step": 34290 }, { "grad_norm": 0.4203088581562042, "learning_rate": 6.564924806709851e-05, "loss": 0.0218, "step": 34300 }, { "grad_norm": 0.290152907371521, "learning_rate": 6.562961677661359e-05, "loss": 0.024, "step": 34310 }, { "grad_norm": 0.3012923002243042, "learning_rate": 6.56099828154546e-05, "loss": 0.0199, "step": 34320 }, { "grad_norm": 0.39023253321647644, "learning_rate": 6.55903461869765e-05, "loss": 0.0232, "step": 34330 }, { "grad_norm": 0.2944435179233551, "learning_rate": 6.557070689453465e-05, "loss": 0.0194, "step": 34340 }, { "grad_norm": 0.5216808319091797, "learning_rate": 6.555106494148482e-05, "loss": 0.0251, "step": 34350 }, { "grad_norm": 0.351375550031662, "learning_rate": 6.553142033118333e-05, "loss": 0.0308, "step": 34360 }, { "grad_norm": 0.3283761143684387, "learning_rate": 6.551177306698688e-05, "loss": 0.0245, "step": 34370 }, { "grad_norm": 0.4102853238582611, "learning_rate": 6.549212315225267e-05, "loss": 0.0275, "step": 34380 }, { "grad_norm": 0.26508453488349915, "learning_rate": 6.547247059033833e-05, "loss": 0.0268, "step": 34390 }, { "grad_norm": 0.3150786757469177, "learning_rate": 6.545281538460193e-05, "loss": 0.0233, "step": 34400 }, { "grad_norm": 0.23749220371246338, "learning_rate": 6.543315753840202e-05, "loss": 0.0241, "step": 34410 }, { "grad_norm": 0.31889837980270386, "learning_rate": 6.541349705509758e-05, "loss": 0.0247, "step": 34420 }, { "grad_norm": 0.25303885340690613, "learning_rate": 6.539383393804805e-05, "loss": 0.0227, "step": 34430 }, { "grad_norm": 0.24035893380641937, "learning_rate": 6.537416819061333e-05, "loss": 0.0266, "step": 34440 }, { "grad_norm": 0.2612905204296112, "learning_rate": 6.535449981615375e-05, "loss": 0.0224, "step": 34450 }, { "grad_norm": 0.22959311306476593, "learning_rate": 6.53348288180301e-05, "loss": 0.0241, "step": 34460 }, { "grad_norm": 0.2870071530342102, "learning_rate": 6.531515519960361e-05, "loss": 0.0218, "step": 34470 }, { "grad_norm": 0.5230473875999451, "learning_rate": 6.529547896423597e-05, "loss": 0.0256, "step": 34480 }, { "grad_norm": 0.244575634598732, "learning_rate": 6.52758001152893e-05, "loss": 0.021, "step": 34490 }, { "grad_norm": 0.41913366317749023, "learning_rate": 6.525611865612618e-05, "loss": 0.0211, "step": 34500 }, { "grad_norm": 0.33148905634880066, "learning_rate": 6.523643459010966e-05, "loss": 0.0223, "step": 34510 }, { "grad_norm": 0.29598110914230347, "learning_rate": 6.521674792060317e-05, "loss": 0.0257, "step": 34520 }, { "grad_norm": 0.3246869444847107, "learning_rate": 6.519705865097063e-05, "loss": 0.0225, "step": 34530 }, { "grad_norm": 0.3828313648700714, "learning_rate": 6.517736678457641e-05, "loss": 0.0193, "step": 34540 }, { "grad_norm": 0.2654533088207245, "learning_rate": 6.515767232478534e-05, "loss": 0.0219, "step": 34550 }, { "grad_norm": 0.31955868005752563, "learning_rate": 6.51379752749626e-05, "loss": 0.0214, "step": 34560 }, { "grad_norm": 0.3023216426372528, "learning_rate": 6.511827563847393e-05, "loss": 0.022, "step": 34570 }, { "grad_norm": 0.2492656111717224, "learning_rate": 6.509857341868542e-05, "loss": 0.0195, "step": 34580 }, { "grad_norm": 0.18667703866958618, "learning_rate": 6.507886861896367e-05, "loss": 0.0184, "step": 34590 }, { "grad_norm": 0.4908706545829773, "learning_rate": 6.505916124267567e-05, "loss": 0.0272, "step": 34600 }, { "grad_norm": 0.3818565309047699, "learning_rate": 6.503945129318891e-05, "loss": 0.0217, "step": 34610 }, { "grad_norm": 0.21605852246284485, "learning_rate": 6.501973877387122e-05, "loss": 0.0183, "step": 34620 }, { "grad_norm": 0.3962500989437103, "learning_rate": 6.500002368809098e-05, "loss": 0.0241, "step": 34630 }, { "grad_norm": 0.29294490814208984, "learning_rate": 6.498030603921694e-05, "loss": 0.0229, "step": 34640 }, { "grad_norm": 0.2831243872642517, "learning_rate": 6.496058583061832e-05, "loss": 0.024, "step": 34650 }, { "grad_norm": 0.30862724781036377, "learning_rate": 6.494086306566475e-05, "loss": 0.0177, "step": 34660 }, { "grad_norm": 0.5011367201805115, "learning_rate": 6.492113774772632e-05, "loss": 0.0197, "step": 34670 }, { "grad_norm": 0.28089961409568787, "learning_rate": 6.490140988017354e-05, "loss": 0.0245, "step": 34680 }, { "grad_norm": 0.32579922676086426, "learning_rate": 6.488167946637736e-05, "loss": 0.0255, "step": 34690 }, { "grad_norm": 0.25588029623031616, "learning_rate": 6.486194650970915e-05, "loss": 0.0247, "step": 34700 }, { "grad_norm": 0.6047727465629578, "learning_rate": 6.48422110135408e-05, "loss": 0.0253, "step": 34710 }, { "grad_norm": 0.23561310768127441, "learning_rate": 6.482247298124451e-05, "loss": 0.0279, "step": 34720 }, { "grad_norm": 0.2235858291387558, "learning_rate": 6.480273241619297e-05, "loss": 0.0209, "step": 34730 }, { "grad_norm": 0.31157606840133667, "learning_rate": 6.478298932175933e-05, "loss": 0.0235, "step": 34740 }, { "grad_norm": 0.41251155734062195, "learning_rate": 6.476324370131712e-05, "loss": 0.021, "step": 34750 }, { "grad_norm": 0.29714784026145935, "learning_rate": 6.474349555824036e-05, "loss": 0.0236, "step": 34760 }, { "grad_norm": 0.28150615096092224, "learning_rate": 6.472374489590342e-05, "loss": 0.0204, "step": 34770 }, { "grad_norm": 0.3266979157924652, "learning_rate": 6.470399171768118e-05, "loss": 0.0281, "step": 34780 }, { "grad_norm": 0.3329634964466095, "learning_rate": 6.468423602694891e-05, "loss": 0.0232, "step": 34790 }, { "grad_norm": 0.2220097780227661, "learning_rate": 6.466447782708232e-05, "loss": 0.0179, "step": 34800 }, { "grad_norm": 0.27233070135116577, "learning_rate": 6.464471712145754e-05, "loss": 0.0207, "step": 34810 }, { "grad_norm": 0.26040300726890564, "learning_rate": 6.462495391345114e-05, "loss": 0.0232, "step": 34820 }, { "grad_norm": 0.32359376549720764, "learning_rate": 6.46051882064401e-05, "loss": 0.0191, "step": 34830 }, { "grad_norm": 0.223566934466362, "learning_rate": 6.458542000380186e-05, "loss": 0.0209, "step": 34840 }, { "grad_norm": 0.316848486661911, "learning_rate": 6.456564930891424e-05, "loss": 0.0204, "step": 34850 }, { "grad_norm": 0.21322216093540192, "learning_rate": 6.454587612515555e-05, "loss": 0.0216, "step": 34860 }, { "grad_norm": 0.19204771518707275, "learning_rate": 6.452610045590444e-05, "loss": 0.024, "step": 34870 }, { "grad_norm": 0.2585228681564331, "learning_rate": 6.450632230454005e-05, "loss": 0.0212, "step": 34880 }, { "grad_norm": 0.21593502163887024, "learning_rate": 6.448654167444195e-05, "loss": 0.0271, "step": 34890 }, { "grad_norm": 0.38535070419311523, "learning_rate": 6.446675856899005e-05, "loss": 0.0227, "step": 34900 }, { "grad_norm": 0.31430479884147644, "learning_rate": 6.444697299156481e-05, "loss": 0.0214, "step": 34910 }, { "grad_norm": 0.25568607449531555, "learning_rate": 6.442718494554701e-05, "loss": 0.0276, "step": 34920 }, { "grad_norm": 0.3045883774757385, "learning_rate": 6.440739443431787e-05, "loss": 0.0172, "step": 34930 }, { "grad_norm": 0.2393694519996643, "learning_rate": 6.438760146125906e-05, "loss": 0.0205, "step": 34940 }, { "grad_norm": 0.32616952061653137, "learning_rate": 6.436780602975267e-05, "loss": 0.0264, "step": 34950 }, { "grad_norm": 0.7222316265106201, "learning_rate": 6.43480081431812e-05, "loss": 0.0233, "step": 34960 }, { "grad_norm": 0.22973234951496124, "learning_rate": 6.432820780492756e-05, "loss": 0.0251, "step": 34970 }, { "grad_norm": 0.40391573309898376, "learning_rate": 6.430840501837506e-05, "loss": 0.03, "step": 34980 }, { "grad_norm": 0.9657613039016724, "learning_rate": 6.428859978690748e-05, "loss": 0.0249, "step": 34990 }, { "grad_norm": 0.22964760661125183, "learning_rate": 6.426879211390901e-05, "loss": 0.0216, "step": 35000 }, { "grad_norm": 0.2989429533481598, "learning_rate": 6.424898200276422e-05, "loss": 0.0225, "step": 35010 }, { "grad_norm": 0.2920977473258972, "learning_rate": 6.42291694568581e-05, "loss": 0.0257, "step": 35020 }, { "grad_norm": 0.29754406213760376, "learning_rate": 6.42093544795761e-05, "loss": 0.0275, "step": 35030 }, { "grad_norm": 0.2812727689743042, "learning_rate": 6.418953707430403e-05, "loss": 0.024, "step": 35040 }, { "grad_norm": 4.205252647399902, "learning_rate": 6.416971724442819e-05, "loss": 0.0183, "step": 35050 }, { "grad_norm": 0.28091755509376526, "learning_rate": 6.414989499333519e-05, "loss": 0.02, "step": 35060 }, { "grad_norm": 0.3521222770214081, "learning_rate": 6.413007032441214e-05, "loss": 0.0246, "step": 35070 }, { "grad_norm": 0.34095579385757446, "learning_rate": 6.411024324104653e-05, "loss": 0.0204, "step": 35080 }, { "grad_norm": 0.28815147280693054, "learning_rate": 6.409041374662628e-05, "loss": 0.0202, "step": 35090 }, { "grad_norm": 0.3257054090499878, "learning_rate": 6.407058184453967e-05, "loss": 0.0344, "step": 35100 }, { "grad_norm": 0.3047725558280945, "learning_rate": 6.405074753817548e-05, "loss": 0.0221, "step": 35110 }, { "grad_norm": 0.34384527802467346, "learning_rate": 6.40309108309228e-05, "loss": 0.0229, "step": 35120 }, { "grad_norm": 0.3678741455078125, "learning_rate": 6.401107172617122e-05, "loss": 0.0222, "step": 35130 }, { "grad_norm": 0.2866789400577545, "learning_rate": 6.399123022731068e-05, "loss": 0.0293, "step": 35140 }, { "grad_norm": 0.3099401295185089, "learning_rate": 6.397138633773157e-05, "loss": 0.0211, "step": 35150 }, { "grad_norm": 0.27342814207077026, "learning_rate": 6.395154006082463e-05, "loss": 0.0231, "step": 35160 }, { "grad_norm": 0.22802573442459106, "learning_rate": 6.393169139998109e-05, "loss": 0.0221, "step": 35170 }, { "grad_norm": 0.2874455153942108, "learning_rate": 6.39118403585925e-05, "loss": 0.0227, "step": 35180 }, { "grad_norm": 0.2919323742389679, "learning_rate": 6.38919869400509e-05, "loss": 0.0256, "step": 35190 }, { "grad_norm": 0.22234977781772614, "learning_rate": 6.387213114774865e-05, "loss": 0.0244, "step": 35200 }, { "grad_norm": 0.22892282903194427, "learning_rate": 6.385227298507863e-05, "loss": 0.0346, "step": 35210 }, { "grad_norm": 0.28208649158477783, "learning_rate": 6.3832412455434e-05, "loss": 0.0215, "step": 35220 }, { "grad_norm": 0.2739589512348175, "learning_rate": 6.381254956220841e-05, "loss": 0.0208, "step": 35230 }, { "grad_norm": 0.3206920921802521, "learning_rate": 6.379268430879586e-05, "loss": 0.019, "step": 35240 }, { "grad_norm": 0.2955656945705414, "learning_rate": 6.37728166985908e-05, "loss": 0.0201, "step": 35250 }, { "grad_norm": 0.5273125767707825, "learning_rate": 6.375294673498804e-05, "loss": 0.0209, "step": 35260 }, { "grad_norm": 0.28154751658439636, "learning_rate": 6.373307442138284e-05, "loss": 0.0267, "step": 35270 }, { "grad_norm": 0.335123747587204, "learning_rate": 6.371319976117081e-05, "loss": 0.0389, "step": 35280 }, { "grad_norm": 0.5827100276947021, "learning_rate": 6.3693322757748e-05, "loss": 0.0208, "step": 35290 }, { "grad_norm": 0.2475641667842865, "learning_rate": 6.367344341451086e-05, "loss": 0.0215, "step": 35300 }, { "grad_norm": 0.36327606439590454, "learning_rate": 6.36535617348562e-05, "loss": 0.0251, "step": 35310 }, { "grad_norm": 0.2640496492385864, "learning_rate": 6.363367772218128e-05, "loss": 0.0237, "step": 35320 }, { "grad_norm": 0.39375635981559753, "learning_rate": 6.36137913798837e-05, "loss": 0.0212, "step": 35330 }, { "grad_norm": 0.2945350110530853, "learning_rate": 6.359390271136151e-05, "loss": 0.0203, "step": 35340 }, { "grad_norm": 0.6722784638404846, "learning_rate": 6.357401172001314e-05, "loss": 0.0232, "step": 35350 }, { "grad_norm": 0.5822033882141113, "learning_rate": 6.355411840923742e-05, "loss": 0.0223, "step": 35360 }, { "grad_norm": 0.2920222580432892, "learning_rate": 6.353422278243358e-05, "loss": 0.0174, "step": 35370 }, { "grad_norm": 0.337007999420166, "learning_rate": 6.351432484300121e-05, "loss": 0.0225, "step": 35380 }, { "grad_norm": 0.2831382155418396, "learning_rate": 6.349442459434036e-05, "loss": 0.0251, "step": 35390 }, { "grad_norm": 0.3506985604763031, "learning_rate": 6.34745220398514e-05, "loss": 0.0288, "step": 35400 }, { "grad_norm": 0.2891084551811218, "learning_rate": 6.345461718293518e-05, "loss": 0.0235, "step": 35410 }, { "grad_norm": 0.6374152302742004, "learning_rate": 6.343471002699286e-05, "loss": 0.0183, "step": 35420 }, { "grad_norm": 0.32900694012641907, "learning_rate": 6.341480057542602e-05, "loss": 0.0234, "step": 35430 }, { "grad_norm": 0.20656298100948334, "learning_rate": 6.339488883163667e-05, "loss": 0.0238, "step": 35440 }, { "grad_norm": 0.3053187429904938, "learning_rate": 6.337497479902716e-05, "loss": 0.0212, "step": 35450 }, { "grad_norm": 0.20867577195167542, "learning_rate": 6.335505848100027e-05, "loss": 0.019, "step": 35460 }, { "grad_norm": 0.5361135601997375, "learning_rate": 6.333513988095915e-05, "loss": 0.0185, "step": 35470 }, { "grad_norm": 0.4821956157684326, "learning_rate": 6.331521900230735e-05, "loss": 0.0184, "step": 35480 }, { "grad_norm": 0.2552618086338043, "learning_rate": 6.329529584844878e-05, "loss": 0.016, "step": 35490 }, { "grad_norm": 0.30712059140205383, "learning_rate": 6.327537042278777e-05, "loss": 0.0206, "step": 35500 }, { "grad_norm": 1.9326560497283936, "learning_rate": 6.325544272872905e-05, "loss": 0.0214, "step": 35510 }, { "grad_norm": 0.42260104417800903, "learning_rate": 6.323551276967771e-05, "loss": 0.0206, "step": 35520 }, { "grad_norm": 0.3052416741847992, "learning_rate": 6.321558054903922e-05, "loss": 0.0209, "step": 35530 }, { "grad_norm": 0.19372428953647614, "learning_rate": 6.319564607021947e-05, "loss": 0.022, "step": 35540 }, { "grad_norm": 0.24230720102787018, "learning_rate": 6.31757093366247e-05, "loss": 0.016, "step": 35550 }, { "grad_norm": 0.27982297539711, "learning_rate": 6.315577035166154e-05, "loss": 0.0223, "step": 35560 }, { "grad_norm": 0.37176334857940674, "learning_rate": 6.313582911873708e-05, "loss": 0.0204, "step": 35570 }, { "grad_norm": 0.22555147111415863, "learning_rate": 6.311588564125865e-05, "loss": 0.0184, "step": 35580 }, { "grad_norm": 0.307597279548645, "learning_rate": 6.30959399226341e-05, "loss": 0.0214, "step": 35590 }, { "grad_norm": 0.3304135799407959, "learning_rate": 6.30759919662716e-05, "loss": 0.0259, "step": 35600 }, { "grad_norm": 0.25768914818763733, "learning_rate": 6.30560417755797e-05, "loss": 0.0212, "step": 35610 }, { "grad_norm": 0.2953198254108429, "learning_rate": 6.303608935396735e-05, "loss": 0.018, "step": 35620 }, { "grad_norm": 0.2793847918510437, "learning_rate": 6.301613470484386e-05, "loss": 0.0188, "step": 35630 }, { "grad_norm": 0.2396153211593628, "learning_rate": 6.299617783161893e-05, "loss": 0.019, "step": 35640 }, { "grad_norm": 0.42898887395858765, "learning_rate": 6.297621873770266e-05, "loss": 0.021, "step": 35650 }, { "grad_norm": 0.2356702983379364, "learning_rate": 6.29562574265055e-05, "loss": 0.0198, "step": 35660 }, { "grad_norm": 0.29336118698120117, "learning_rate": 6.293629390143834e-05, "loss": 0.0226, "step": 35670 }, { "grad_norm": 0.24103721976280212, "learning_rate": 6.291632816591232e-05, "loss": 0.0196, "step": 35680 }, { "grad_norm": 0.30920735001564026, "learning_rate": 6.28963602233391e-05, "loss": 0.0186, "step": 35690 }, { "grad_norm": 0.3107178509235382, "learning_rate": 6.287639007713062e-05, "loss": 0.0218, "step": 35700 }, { "grad_norm": 0.28635191917419434, "learning_rate": 6.285641773069926e-05, "loss": 0.0195, "step": 35710 }, { "grad_norm": 0.8067449331283569, "learning_rate": 6.283644318745773e-05, "loss": 0.0215, "step": 35720 }, { "grad_norm": 0.31003502011299133, "learning_rate": 6.281646645081912e-05, "loss": 0.0238, "step": 35730 }, { "grad_norm": 0.24448271095752716, "learning_rate": 6.279648752419693e-05, "loss": 0.0226, "step": 35740 }, { "grad_norm": 0.36745530366897583, "learning_rate": 6.2776506411005e-05, "loss": 0.0242, "step": 35750 }, { "grad_norm": 0.33614835143089294, "learning_rate": 6.275652311465758e-05, "loss": 0.02, "step": 35760 }, { "grad_norm": 0.5037506222724915, "learning_rate": 6.273653763856926e-05, "loss": 0.0229, "step": 35770 }, { "grad_norm": 0.254203736782074, "learning_rate": 6.271654998615501e-05, "loss": 0.0229, "step": 35780 }, { "grad_norm": 0.2634749710559845, "learning_rate": 6.269656016083013e-05, "loss": 0.0176, "step": 35790 }, { "grad_norm": 0.29494163393974304, "learning_rate": 6.267656816601038e-05, "loss": 0.0252, "step": 35800 }, { "grad_norm": 0.2959955632686615, "learning_rate": 6.265657400511185e-05, "loss": 0.0255, "step": 35810 }, { "grad_norm": 0.28134390711784363, "learning_rate": 6.263657768155098e-05, "loss": 0.0191, "step": 35820 }, { "grad_norm": 0.26618993282318115, "learning_rate": 6.261657919874457e-05, "loss": 0.0181, "step": 35830 }, { "grad_norm": 0.29694366455078125, "learning_rate": 6.259657856010986e-05, "loss": 0.0228, "step": 35840 }, { "grad_norm": 0.3688850998878479, "learning_rate": 6.257657576906439e-05, "loss": 0.0192, "step": 35850 }, { "grad_norm": 0.26185542345046997, "learning_rate": 6.255657082902609e-05, "loss": 0.0148, "step": 35860 }, { "grad_norm": 0.265211284160614, "learning_rate": 6.253656374341325e-05, "loss": 0.018, "step": 35870 }, { "grad_norm": 0.29134657979011536, "learning_rate": 6.251655451564457e-05, "loss": 0.0197, "step": 35880 }, { "grad_norm": 1.7839640378952026, "learning_rate": 6.249654314913902e-05, "loss": 0.0225, "step": 35890 }, { "grad_norm": 0.21802356839179993, "learning_rate": 6.247652964731604e-05, "loss": 0.0163, "step": 35900 }, { "grad_norm": 0.19065934419631958, "learning_rate": 6.245651401359537e-05, "loss": 0.0351, "step": 35910 }, { "grad_norm": 0.5332854986190796, "learning_rate": 6.243649625139715e-05, "loss": 0.0251, "step": 35920 }, { "grad_norm": 0.2775398790836334, "learning_rate": 6.241647636414185e-05, "loss": 0.0227, "step": 35930 }, { "grad_norm": 0.22209185361862183, "learning_rate": 6.239645435525034e-05, "loss": 0.0239, "step": 35940 }, { "grad_norm": 0.19832612574100494, "learning_rate": 6.237643022814381e-05, "loss": 0.0206, "step": 35950 }, { "grad_norm": 0.26524248719215393, "learning_rate": 6.235640398624386e-05, "loss": 0.0198, "step": 35960 }, { "grad_norm": 0.2369401454925537, "learning_rate": 6.233637563297243e-05, "loss": 0.0217, "step": 35970 }, { "grad_norm": 0.2643791139125824, "learning_rate": 6.23163451717518e-05, "loss": 0.0217, "step": 35980 }, { "grad_norm": 0.27756914496421814, "learning_rate": 6.229631260600463e-05, "loss": 0.0223, "step": 35990 }, { "grad_norm": 0.2783142626285553, "learning_rate": 6.227627793915392e-05, "loss": 0.0263, "step": 36000 }, { "grad_norm": 0.2005961686372757, "learning_rate": 6.225624117462309e-05, "loss": 0.0185, "step": 36010 }, { "grad_norm": 0.3736790120601654, "learning_rate": 6.223620231583586e-05, "loss": 0.018, "step": 36020 }, { "grad_norm": 0.20050813257694244, "learning_rate": 6.221616136621629e-05, "loss": 0.0158, "step": 36030 }, { "grad_norm": 0.3760208785533905, "learning_rate": 6.219611832918887e-05, "loss": 0.0172, "step": 36040 }, { "grad_norm": 0.16374854743480682, "learning_rate": 6.217607320817838e-05, "loss": 0.0173, "step": 36050 }, { "grad_norm": 0.21387405693531036, "learning_rate": 6.215602600661001e-05, "loss": 0.0186, "step": 36060 }, { "grad_norm": 0.2322227656841278, "learning_rate": 6.213597672790925e-05, "loss": 0.015, "step": 36070 }, { "grad_norm": 0.1718876212835312, "learning_rate": 6.2115925375502e-05, "loss": 0.0212, "step": 36080 }, { "grad_norm": 0.26287877559661865, "learning_rate": 6.209587195281447e-05, "loss": 0.018, "step": 36090 }, { "grad_norm": 0.36032161116600037, "learning_rate": 6.207581646327324e-05, "loss": 0.0257, "step": 36100 }, { "grad_norm": 0.9412070512771606, "learning_rate": 6.205575891030526e-05, "loss": 0.0143, "step": 36110 }, { "grad_norm": 0.19029156863689423, "learning_rate": 6.203569929733781e-05, "loss": 0.0202, "step": 36120 }, { "grad_norm": 0.24548661708831787, "learning_rate": 6.201563762779852e-05, "loss": 0.0204, "step": 36130 }, { "grad_norm": 0.24489976465702057, "learning_rate": 6.199557390511538e-05, "loss": 0.0243, "step": 36140 }, { "grad_norm": 0.28360307216644287, "learning_rate": 6.197550813271675e-05, "loss": 0.0188, "step": 36150 }, { "grad_norm": 0.3795824944972992, "learning_rate": 6.195544031403131e-05, "loss": 0.0211, "step": 36160 }, { "grad_norm": 0.3013797104358673, "learning_rate": 6.19353704524881e-05, "loss": 0.0208, "step": 36170 }, { "grad_norm": 0.36579403281211853, "learning_rate": 6.191529855151652e-05, "loss": 0.0192, "step": 36180 }, { "grad_norm": 0.7590603232383728, "learning_rate": 6.189522461454629e-05, "loss": 0.0217, "step": 36190 }, { "grad_norm": 0.27089160680770874, "learning_rate": 6.187514864500752e-05, "loss": 0.022, "step": 36200 }, { "grad_norm": 0.36621180176734924, "learning_rate": 6.185507064633062e-05, "loss": 0.0218, "step": 36210 }, { "grad_norm": 0.4735392928123474, "learning_rate": 6.18349906219464e-05, "loss": 0.0239, "step": 36220 }, { "grad_norm": 0.19558143615722656, "learning_rate": 6.181490857528596e-05, "loss": 0.0196, "step": 36230 }, { "grad_norm": 0.2717134654521942, "learning_rate": 6.179482450978077e-05, "loss": 0.0175, "step": 36240 }, { "grad_norm": 0.22636224329471588, "learning_rate": 6.177473842886269e-05, "loss": 0.0163, "step": 36250 }, { "grad_norm": 0.21679151058197021, "learning_rate": 6.175465033596382e-05, "loss": 0.0242, "step": 36260 }, { "grad_norm": 0.33229780197143555, "learning_rate": 6.173456023451671e-05, "loss": 0.0209, "step": 36270 }, { "grad_norm": 0.7207083106040955, "learning_rate": 6.171446812795422e-05, "loss": 0.026, "step": 36280 }, { "grad_norm": 0.2677493989467621, "learning_rate": 6.169437401970949e-05, "loss": 0.0281, "step": 36290 }, { "grad_norm": 0.4309983551502228, "learning_rate": 6.16742779132161e-05, "loss": 0.0165, "step": 36300 }, { "grad_norm": 0.2685604989528656, "learning_rate": 6.165417981190789e-05, "loss": 0.016, "step": 36310 }, { "grad_norm": 0.23098231852054596, "learning_rate": 6.16340797192191e-05, "loss": 0.0193, "step": 36320 }, { "grad_norm": 0.33366355299949646, "learning_rate": 6.161397763858427e-05, "loss": 0.0184, "step": 36330 }, { "grad_norm": 0.3470863103866577, "learning_rate": 6.159387357343834e-05, "loss": 0.0198, "step": 36340 }, { "grad_norm": 0.1844319850206375, "learning_rate": 6.157376752721648e-05, "loss": 0.0241, "step": 36350 }, { "grad_norm": 0.25591734051704407, "learning_rate": 6.155365950335428e-05, "loss": 0.0191, "step": 36360 }, { "grad_norm": 0.36108890175819397, "learning_rate": 6.153354950528768e-05, "loss": 0.0196, "step": 36370 }, { "grad_norm": 0.27219530940055847, "learning_rate": 6.151343753645293e-05, "loss": 0.0235, "step": 36380 }, { "grad_norm": 0.2720422148704529, "learning_rate": 6.149332360028657e-05, "loss": 0.0221, "step": 36390 }, { "grad_norm": 0.2276013344526291, "learning_rate": 6.147320770022555e-05, "loss": 0.0231, "step": 36400 }, { "grad_norm": 0.24085120856761932, "learning_rate": 6.145308983970715e-05, "loss": 0.0239, "step": 36410 }, { "grad_norm": 0.23292845487594604, "learning_rate": 6.143297002216892e-05, "loss": 0.0211, "step": 36420 }, { "grad_norm": 0.2591760754585266, "learning_rate": 6.141284825104882e-05, "loss": 0.0228, "step": 36430 }, { "grad_norm": 0.24259909987449646, "learning_rate": 6.13927245297851e-05, "loss": 0.0247, "step": 36440 }, { "grad_norm": 0.9553861618041992, "learning_rate": 6.137259886181633e-05, "loss": 0.0257, "step": 36450 }, { "grad_norm": 0.23544834554195404, "learning_rate": 6.135247125058145e-05, "loss": 0.0201, "step": 36460 }, { "grad_norm": 1.33760666847229, "learning_rate": 6.133234169951974e-05, "loss": 0.0187, "step": 36470 }, { "grad_norm": 0.2365623414516449, "learning_rate": 6.131221021207078e-05, "loss": 0.0226, "step": 36480 }, { "grad_norm": 0.23281633853912354, "learning_rate": 6.129207679167448e-05, "loss": 0.0169, "step": 36490 }, { "grad_norm": 0.22267912328243256, "learning_rate": 6.127194144177109e-05, "loss": 0.0197, "step": 36500 }, { "grad_norm": 0.25705933570861816, "learning_rate": 6.125180416580118e-05, "loss": 0.0186, "step": 36510 }, { "grad_norm": 0.17818152904510498, "learning_rate": 6.123166496720571e-05, "loss": 0.0178, "step": 36520 }, { "grad_norm": 0.28669801354408264, "learning_rate": 6.121152384942588e-05, "loss": 0.0197, "step": 36530 }, { "grad_norm": 0.9497382640838623, "learning_rate": 6.119138081590324e-05, "loss": 0.0214, "step": 36540 }, { "grad_norm": 0.26572662591934204, "learning_rate": 6.117123587007971e-05, "loss": 0.0221, "step": 36550 }, { "grad_norm": 0.37763556838035583, "learning_rate": 6.11510890153975e-05, "loss": 0.0282, "step": 36560 }, { "grad_norm": 0.25420573353767395, "learning_rate": 6.113094025529916e-05, "loss": 0.0221, "step": 36570 }, { "grad_norm": 0.25048941373825073, "learning_rate": 6.111078959322757e-05, "loss": 0.0274, "step": 36580 }, { "grad_norm": 0.31440263986587524, "learning_rate": 6.109063703262592e-05, "loss": 0.0246, "step": 36590 }, { "grad_norm": 0.29867738485336304, "learning_rate": 6.107048257693772e-05, "loss": 0.0209, "step": 36600 }, { "grad_norm": 0.48436439037323, "learning_rate": 6.105032622960683e-05, "loss": 0.0237, "step": 36610 }, { "grad_norm": 0.2509678900241852, "learning_rate": 6.103016799407743e-05, "loss": 0.0231, "step": 36620 }, { "grad_norm": 0.23744580149650574, "learning_rate": 6.1010007873793984e-05, "loss": 0.0253, "step": 36630 }, { "grad_norm": 0.21044360101222992, "learning_rate": 6.098984587220131e-05, "loss": 0.0198, "step": 36640 }, { "grad_norm": 0.318717360496521, "learning_rate": 6.096968199274456e-05, "loss": 0.0205, "step": 36650 }, { "grad_norm": 0.2641647458076477, "learning_rate": 6.0949516238869166e-05, "loss": 0.0182, "step": 36660 }, { "grad_norm": 0.40777942538261414, "learning_rate": 6.092934861402092e-05, "loss": 0.0262, "step": 36670 }, { "grad_norm": 0.26754578948020935, "learning_rate": 6.0909179121645924e-05, "loss": 0.0303, "step": 36680 }, { "grad_norm": 0.40467163920402527, "learning_rate": 6.0889007765190576e-05, "loss": 0.0276, "step": 36690 }, { "grad_norm": 0.23846712708473206, "learning_rate": 6.086883454810162e-05, "loss": 0.0208, "step": 36700 }, { "grad_norm": 0.22847510874271393, "learning_rate": 6.0848659473826084e-05, "loss": 0.0232, "step": 36710 }, { "grad_norm": 0.27730345726013184, "learning_rate": 6.082848254581138e-05, "loss": 0.0185, "step": 36720 }, { "grad_norm": 0.3186626434326172, "learning_rate": 6.080830376750517e-05, "loss": 0.0203, "step": 36730 }, { "grad_norm": 0.19820640981197357, "learning_rate": 6.0788123142355445e-05, "loss": 0.0191, "step": 36740 }, { "grad_norm": 0.2827867865562439, "learning_rate": 6.076794067381052e-05, "loss": 0.0174, "step": 36750 }, { "grad_norm": 0.2474321573972702, "learning_rate": 6.074775636531905e-05, "loss": 0.0228, "step": 36760 }, { "grad_norm": 0.31468665599823, "learning_rate": 6.072757022032997e-05, "loss": 0.0224, "step": 36770 }, { "grad_norm": 0.2974105179309845, "learning_rate": 6.070738224229253e-05, "loss": 0.0229, "step": 36780 }, { "grad_norm": 0.3047991991043091, "learning_rate": 6.0687192434656314e-05, "loss": 0.0229, "step": 36790 }, { "grad_norm": 0.2608214318752289, "learning_rate": 6.066700080087121e-05, "loss": 0.0184, "step": 36800 }, { "grad_norm": 0.1594327837228775, "learning_rate": 6.0646807344387424e-05, "loss": 0.0178, "step": 36810 }, { "grad_norm": 0.23907549679279327, "learning_rate": 6.062661206865543e-05, "loss": 0.0218, "step": 36820 }, { "grad_norm": 0.3120385706424713, "learning_rate": 6.06064149771261e-05, "loss": 0.0204, "step": 36830 }, { "grad_norm": 0.2536231577396393, "learning_rate": 6.058621607325051e-05, "loss": 0.0164, "step": 36840 }, { "grad_norm": 0.2717978358268738, "learning_rate": 6.056601536048014e-05, "loss": 0.0217, "step": 36850 }, { "grad_norm": 0.33924928307533264, "learning_rate": 6.0545812842266725e-05, "loss": 0.0263, "step": 36860 }, { "grad_norm": 0.3229560852050781, "learning_rate": 6.052560852206232e-05, "loss": 0.0243, "step": 36870 }, { "grad_norm": 0.33063825964927673, "learning_rate": 6.05054024033193e-05, "loss": 0.0293, "step": 36880 }, { "grad_norm": 0.7512877583503723, "learning_rate": 6.048519448949032e-05, "loss": 0.0229, "step": 36890 }, { "grad_norm": 0.3264157176017761, "learning_rate": 6.046498478402839e-05, "loss": 0.0288, "step": 36900 }, { "grad_norm": 0.2830311357975006, "learning_rate": 6.044477329038677e-05, "loss": 0.0244, "step": 36910 }, { "grad_norm": 0.45992720127105713, "learning_rate": 6.042456001201906e-05, "loss": 0.0203, "step": 36920 }, { "grad_norm": 0.2188897728919983, "learning_rate": 6.040434495237917e-05, "loss": 0.022, "step": 36930 }, { "grad_norm": 0.25037795305252075, "learning_rate": 6.0384128114921256e-05, "loss": 0.0246, "step": 36940 }, { "grad_norm": 0.2983565926551819, "learning_rate": 6.036390950309987e-05, "loss": 0.0169, "step": 36950 }, { "grad_norm": 0.6223394870758057, "learning_rate": 6.0343689120369805e-05, "loss": 0.0214, "step": 36960 }, { "grad_norm": 0.3913498818874359, "learning_rate": 6.032346697018616e-05, "loss": 0.0198, "step": 36970 }, { "grad_norm": 0.34529608488082886, "learning_rate": 6.0303243056004375e-05, "loss": 0.0305, "step": 36980 }, { "grad_norm": 0.6412616968154907, "learning_rate": 6.0283017381280136e-05, "loss": 0.0228, "step": 36990 }, { "grad_norm": 0.6349539160728455, "learning_rate": 6.026278994946948e-05, "loss": 0.0194, "step": 37000 }, { "grad_norm": 0.2880728840827942, "learning_rate": 6.02425607640287e-05, "loss": 0.0248, "step": 37010 }, { "grad_norm": 0.20998017489910126, "learning_rate": 6.022232982841441e-05, "loss": 0.0203, "step": 37020 }, { "grad_norm": 0.36091113090515137, "learning_rate": 6.020209714608355e-05, "loss": 0.0185, "step": 37030 }, { "grad_norm": 0.34766659140586853, "learning_rate": 6.018186272049331e-05, "loss": 0.0231, "step": 37040 }, { "grad_norm": 0.25932183861732483, "learning_rate": 6.01616265551012e-05, "loss": 0.0238, "step": 37050 }, { "grad_norm": 0.25419288873672485, "learning_rate": 6.014138865336503e-05, "loss": 0.016, "step": 37060 }, { "grad_norm": 0.3320760428905487, "learning_rate": 6.0121149018742905e-05, "loss": 0.0203, "step": 37070 }, { "grad_norm": 0.1966816633939743, "learning_rate": 6.010090765469325e-05, "loss": 0.0193, "step": 37080 }, { "grad_norm": 1.65847647190094, "learning_rate": 6.008066456467473e-05, "loss": 0.0236, "step": 37090 }, { "grad_norm": 0.26263228058815, "learning_rate": 6.0060419752146335e-05, "loss": 0.0199, "step": 37100 }, { "grad_norm": 0.34139129519462585, "learning_rate": 6.0040173220567353e-05, "loss": 0.0202, "step": 37110 }, { "grad_norm": 0.3470175862312317, "learning_rate": 6.001992497339737e-05, "loss": 0.0218, "step": 37120 }, { "grad_norm": 0.2443697601556778, "learning_rate": 5.999967501409626e-05, "loss": 0.0241, "step": 37130 }, { "grad_norm": 0.1812954545021057, "learning_rate": 5.997942334612418e-05, "loss": 0.0191, "step": 37140 }, { "grad_norm": 0.9012796878814697, "learning_rate": 5.995916997294158e-05, "loss": 0.0167, "step": 37150 }, { "grad_norm": 0.21305376291275024, "learning_rate": 5.9938914898009214e-05, "loss": 0.023, "step": 37160 }, { "grad_norm": 0.320258766412735, "learning_rate": 5.991865812478813e-05, "loss": 0.0151, "step": 37170 }, { "grad_norm": 0.4998883605003357, "learning_rate": 5.989839965673964e-05, "loss": 0.0318, "step": 37180 }, { "grad_norm": 0.24273979663848877, "learning_rate": 5.987813949732539e-05, "loss": 0.0208, "step": 37190 }, { "grad_norm": 0.2684570550918579, "learning_rate": 5.9857877650007255e-05, "loss": 0.0198, "step": 37200 }, { "grad_norm": 0.6574846506118774, "learning_rate": 5.983761411824744e-05, "loss": 0.0192, "step": 37210 }, { "grad_norm": 0.24427124857902527, "learning_rate": 5.981734890550844e-05, "loss": 0.0197, "step": 37220 }, { "grad_norm": 0.46491602063179016, "learning_rate": 5.979708201525301e-05, "loss": 0.0268, "step": 37230 }, { "grad_norm": 0.35976460576057434, "learning_rate": 5.977681345094422e-05, "loss": 0.018, "step": 37240 }, { "grad_norm": 0.30722475051879883, "learning_rate": 5.97565432160454e-05, "loss": 0.0235, "step": 37250 }, { "grad_norm": 0.27222320437431335, "learning_rate": 5.9736271314020186e-05, "loss": 0.0227, "step": 37260 }, { "grad_norm": 0.2961176335811615, "learning_rate": 5.971599774833251e-05, "loss": 0.0239, "step": 37270 }, { "grad_norm": 0.21811164915561676, "learning_rate": 5.9695722522446525e-05, "loss": 0.0301, "step": 37280 }, { "grad_norm": 0.31333205103874207, "learning_rate": 5.9675445639826765e-05, "loss": 0.0197, "step": 37290 }, { "grad_norm": 0.24018526077270508, "learning_rate": 5.965516710393796e-05, "loss": 0.0188, "step": 37300 }, { "grad_norm": 0.20145031809806824, "learning_rate": 5.963488691824516e-05, "loss": 0.0235, "step": 37310 }, { "grad_norm": 0.6565828323364258, "learning_rate": 5.96146050862137e-05, "loss": 0.0237, "step": 37320 }, { "grad_norm": 0.26126012206077576, "learning_rate": 5.959432161130919e-05, "loss": 0.023, "step": 37330 }, { "grad_norm": 0.22046701610088348, "learning_rate": 5.9574036496997545e-05, "loss": 0.0292, "step": 37340 }, { "grad_norm": 0.28706493973731995, "learning_rate": 5.955374974674488e-05, "loss": 0.0206, "step": 37350 }, { "grad_norm": 0.43083831667900085, "learning_rate": 5.9533461364017696e-05, "loss": 0.0212, "step": 37360 }, { "grad_norm": 0.2711464762687683, "learning_rate": 5.9513171352282716e-05, "loss": 0.0179, "step": 37370 }, { "grad_norm": 0.2540942132472992, "learning_rate": 5.949287971500692e-05, "loss": 0.0181, "step": 37380 }, { "grad_norm": 0.16486544907093048, "learning_rate": 5.947258645565762e-05, "loss": 0.0141, "step": 37390 }, { "grad_norm": 0.22816891968250275, "learning_rate": 5.945229157770237e-05, "loss": 0.0216, "step": 37400 }, { "grad_norm": 0.24753698706626892, "learning_rate": 5.9431995084609006e-05, "loss": 0.0229, "step": 37410 }, { "grad_norm": 0.23403844237327576, "learning_rate": 5.941169697984564e-05, "loss": 0.0225, "step": 37420 }, { "grad_norm": 0.39766353368759155, "learning_rate": 5.9391397266880675e-05, "loss": 0.0187, "step": 37430 }, { "grad_norm": 0.2248028814792633, "learning_rate": 5.937109594918279e-05, "loss": 0.035, "step": 37440 }, { "grad_norm": 0.27675795555114746, "learning_rate": 5.9350793030220884e-05, "loss": 0.0321, "step": 37450 }, { "grad_norm": 0.28910428285598755, "learning_rate": 5.933048851346421e-05, "loss": 0.0195, "step": 37460 }, { "grad_norm": 0.26356369256973267, "learning_rate": 5.931018240238224e-05, "loss": 0.0337, "step": 37470 }, { "grad_norm": 0.17576569318771362, "learning_rate": 5.928987470044471e-05, "loss": 0.016, "step": 37480 }, { "grad_norm": 0.30089887976646423, "learning_rate": 5.9269565411121695e-05, "loss": 0.0189, "step": 37490 }, { "grad_norm": 0.18334752321243286, "learning_rate": 5.924925453788347e-05, "loss": 0.0207, "step": 37500 }, { "grad_norm": 0.34206026792526245, "learning_rate": 5.92289420842006e-05, "loss": 0.0248, "step": 37510 }, { "grad_norm": 2.776883125305176, "learning_rate": 5.9208628053543945e-05, "loss": 0.0284, "step": 37520 }, { "grad_norm": 0.3332988917827606, "learning_rate": 5.918831244938462e-05, "loss": 0.0197, "step": 37530 }, { "grad_norm": 0.49583545327186584, "learning_rate": 5.9167995275194e-05, "loss": 0.0244, "step": 37540 }, { "grad_norm": 0.4540884494781494, "learning_rate": 5.914767653444373e-05, "loss": 0.0226, "step": 37550 }, { "grad_norm": 0.3915795385837555, "learning_rate": 5.912735623060572e-05, "loss": 0.0236, "step": 37560 }, { "grad_norm": 0.2659054100513458, "learning_rate": 5.910703436715217e-05, "loss": 0.0171, "step": 37570 }, { "grad_norm": 0.2114289104938507, "learning_rate": 5.908671094755552e-05, "loss": 0.0187, "step": 37580 }, { "grad_norm": 0.36861661076545715, "learning_rate": 5.906638597528851e-05, "loss": 0.0186, "step": 37590 }, { "grad_norm": 0.25730812549591064, "learning_rate": 5.9046059453824076e-05, "loss": 0.0194, "step": 37600 }, { "grad_norm": 0.2704424560070038, "learning_rate": 5.9025731386635505e-05, "loss": 0.0175, "step": 37610 }, { "grad_norm": 0.3872895836830139, "learning_rate": 5.900540177719629e-05, "loss": 0.0215, "step": 37620 }, { "grad_norm": 0.2897292673587799, "learning_rate": 5.898507062898021e-05, "loss": 0.0218, "step": 37630 }, { "grad_norm": 0.3822730481624603, "learning_rate": 5.8964737945461316e-05, "loss": 0.0228, "step": 37640 }, { "grad_norm": 0.32332557439804077, "learning_rate": 5.8944403730113885e-05, "loss": 0.017, "step": 37650 }, { "grad_norm": 0.28813305497169495, "learning_rate": 5.892406798641248e-05, "loss": 0.0304, "step": 37660 }, { "grad_norm": 0.27275460958480835, "learning_rate": 5.890373071783193e-05, "loss": 0.0187, "step": 37670 }, { "grad_norm": 0.2837999761104584, "learning_rate": 5.888339192784732e-05, "loss": 0.0215, "step": 37680 }, { "grad_norm": 0.43828848004341125, "learning_rate": 5.8863051619934003e-05, "loss": 0.0199, "step": 37690 }, { "grad_norm": 0.4732799530029297, "learning_rate": 5.8842709797567554e-05, "loss": 0.0201, "step": 37700 }, { "grad_norm": 0.27151137590408325, "learning_rate": 5.8822366464223855e-05, "loss": 0.0161, "step": 37710 }, { "grad_norm": 0.22352899610996246, "learning_rate": 5.880202162337901e-05, "loss": 0.0192, "step": 37720 }, { "grad_norm": 0.22171838581562042, "learning_rate": 5.8781675278509405e-05, "loss": 0.0161, "step": 37730 }, { "grad_norm": 0.2658006250858307, "learning_rate": 5.8761327433091696e-05, "loss": 0.0162, "step": 37740 }, { "grad_norm": 0.25607389211654663, "learning_rate": 5.874097809060275e-05, "loss": 0.0247, "step": 37750 }, { "grad_norm": 0.544990599155426, "learning_rate": 5.87206272545197e-05, "loss": 0.0261, "step": 37760 }, { "grad_norm": 0.3256561756134033, "learning_rate": 5.8700274928319955e-05, "loss": 0.0212, "step": 37770 }, { "grad_norm": 0.23249073326587677, "learning_rate": 5.867992111548118e-05, "loss": 0.0219, "step": 37780 }, { "grad_norm": 0.2216109335422516, "learning_rate": 5.865956581948131e-05, "loss": 0.0187, "step": 37790 }, { "grad_norm": 0.21897855401039124, "learning_rate": 5.863920904379845e-05, "loss": 0.0198, "step": 37800 }, { "grad_norm": 0.26452040672302246, "learning_rate": 5.8618850791911064e-05, "loss": 0.0238, "step": 37810 }, { "grad_norm": 0.2310384064912796, "learning_rate": 5.859849106729779e-05, "loss": 0.0177, "step": 37820 }, { "grad_norm": 0.40590155124664307, "learning_rate": 5.857812987343758e-05, "loss": 0.0192, "step": 37830 }, { "grad_norm": 0.2465759962797165, "learning_rate": 5.855776721380957e-05, "loss": 0.0245, "step": 37840 }, { "grad_norm": 0.2598631978034973, "learning_rate": 5.8537403091893217e-05, "loss": 0.0221, "step": 37850 }, { "grad_norm": 0.3016932010650635, "learning_rate": 5.851703751116816e-05, "loss": 0.0209, "step": 37860 }, { "grad_norm": 0.1838378608226776, "learning_rate": 5.8496670475114336e-05, "loss": 0.02, "step": 37870 }, { "grad_norm": 0.23090524971485138, "learning_rate": 5.84763019872119e-05, "loss": 0.0234, "step": 37880 }, { "grad_norm": 0.3083747923374176, "learning_rate": 5.845593205094131e-05, "loss": 0.0185, "step": 37890 }, { "grad_norm": 0.2940863370895386, "learning_rate": 5.843556066978318e-05, "loss": 0.0164, "step": 37900 }, { "grad_norm": 0.3877260982990265, "learning_rate": 5.8415187847218455e-05, "loss": 0.0243, "step": 37910 }, { "grad_norm": 0.21787019073963165, "learning_rate": 5.839481358672827e-05, "loss": 0.0163, "step": 37920 }, { "grad_norm": 0.3610555827617645, "learning_rate": 5.837443789179407e-05, "loss": 0.0184, "step": 37930 }, { "grad_norm": 0.17746010422706604, "learning_rate": 5.8354060765897445e-05, "loss": 0.0177, "step": 37940 }, { "grad_norm": 0.2303411215543747, "learning_rate": 5.8333682212520334e-05, "loss": 0.0262, "step": 37950 }, { "grad_norm": 0.27422383427619934, "learning_rate": 5.831330223514486e-05, "loss": 0.0242, "step": 37960 }, { "grad_norm": 0.9357696771621704, "learning_rate": 5.8292920837253396e-05, "loss": 0.0206, "step": 37970 }, { "grad_norm": 0.23204867541790009, "learning_rate": 5.827253802232857e-05, "loss": 0.0167, "step": 37980 }, { "grad_norm": 0.29016372561454773, "learning_rate": 5.825215379385327e-05, "loss": 0.0126, "step": 37990 }, { "grad_norm": 0.24990826845169067, "learning_rate": 5.823176815531057e-05, "loss": 0.0321, "step": 38000 }, { "grad_norm": 0.217636376619339, "learning_rate": 5.8211381110183826e-05, "loss": 0.0242, "step": 38010 }, { "grad_norm": 0.3140930235385895, "learning_rate": 5.8190992661956645e-05, "loss": 0.0172, "step": 38020 }, { "grad_norm": 0.23808598518371582, "learning_rate": 5.817060281411284e-05, "loss": 0.0195, "step": 38030 }, { "grad_norm": 0.27557823061943054, "learning_rate": 5.815021157013647e-05, "loss": 0.0189, "step": 38040 }, { "grad_norm": 0.22909623384475708, "learning_rate": 5.8129818933511856e-05, "loss": 0.0285, "step": 38050 }, { "grad_norm": 0.22870947420597076, "learning_rate": 5.8109424907723544e-05, "loss": 0.0231, "step": 38060 }, { "grad_norm": 0.19184990227222443, "learning_rate": 5.80890294962563e-05, "loss": 0.0162, "step": 38070 }, { "grad_norm": 0.28378984332084656, "learning_rate": 5.806863270259515e-05, "loss": 0.0197, "step": 38080 }, { "grad_norm": 0.24922505021095276, "learning_rate": 5.804823453022536e-05, "loss": 0.0167, "step": 38090 }, { "grad_norm": 0.2864890396595001, "learning_rate": 5.80278349826324e-05, "loss": 0.0183, "step": 38100 }, { "grad_norm": 0.23270350694656372, "learning_rate": 5.8007434063302014e-05, "loss": 0.0188, "step": 38110 }, { "grad_norm": 0.19143685698509216, "learning_rate": 5.7987031775720136e-05, "loss": 0.0158, "step": 38120 }, { "grad_norm": 0.3685165345668793, "learning_rate": 5.7966628123372976e-05, "loss": 0.0239, "step": 38130 }, { "grad_norm": 0.26214101910591125, "learning_rate": 5.7946223109746956e-05, "loss": 0.0226, "step": 38140 }, { "grad_norm": 0.2260659784078598, "learning_rate": 5.7925816738328754e-05, "loss": 0.0168, "step": 38150 }, { "grad_norm": 0.220880925655365, "learning_rate": 5.790540901260521e-05, "loss": 0.0197, "step": 38160 }, { "grad_norm": 0.2382669299840927, "learning_rate": 5.788499993606351e-05, "loss": 0.0195, "step": 38170 }, { "grad_norm": 0.21926942467689514, "learning_rate": 5.786458951219096e-05, "loss": 0.0231, "step": 38180 }, { "grad_norm": 0.20484274625778198, "learning_rate": 5.784417774447517e-05, "loss": 0.0241, "step": 38190 }, { "grad_norm": 0.23764370381832123, "learning_rate": 5.782376463640393e-05, "loss": 0.0224, "step": 38200 }, { "grad_norm": 0.26451563835144043, "learning_rate": 5.780335019146531e-05, "loss": 0.0204, "step": 38210 }, { "grad_norm": 0.48637428879737854, "learning_rate": 5.778293441314755e-05, "loss": 0.022, "step": 38220 }, { "grad_norm": 0.41778525710105896, "learning_rate": 5.776251730493917e-05, "loss": 0.0287, "step": 38230 }, { "grad_norm": 0.27715349197387695, "learning_rate": 5.774209887032887e-05, "loss": 0.0204, "step": 38240 }, { "grad_norm": 0.2059638500213623, "learning_rate": 5.772167911280565e-05, "loss": 0.019, "step": 38250 }, { "grad_norm": 0.2328348308801651, "learning_rate": 5.770125803585864e-05, "loss": 0.0189, "step": 38260 }, { "grad_norm": 0.2539002597332001, "learning_rate": 5.768083564297726e-05, "loss": 0.0218, "step": 38270 }, { "grad_norm": 0.2705242931842804, "learning_rate": 5.766041193765114e-05, "loss": 0.0219, "step": 38280 }, { "grad_norm": 0.2024107128381729, "learning_rate": 5.763998692337015e-05, "loss": 0.0184, "step": 38290 }, { "grad_norm": 0.2785445749759674, "learning_rate": 5.761956060362433e-05, "loss": 0.0206, "step": 38300 }, { "grad_norm": 0.2799449563026428, "learning_rate": 5.7599132981904e-05, "loss": 0.0238, "step": 38310 }, { "grad_norm": 0.2177489846944809, "learning_rate": 5.75787040616997e-05, "loss": 0.0212, "step": 38320 }, { "grad_norm": 0.21274471282958984, "learning_rate": 5.755827384650212e-05, "loss": 0.0208, "step": 38330 }, { "grad_norm": 0.5697671175003052, "learning_rate": 5.753784233980228e-05, "loss": 0.0243, "step": 38340 }, { "grad_norm": 0.33923739194869995, "learning_rate": 5.751740954509135e-05, "loss": 0.0201, "step": 38350 }, { "grad_norm": 0.36980393528938293, "learning_rate": 5.7496975465860715e-05, "loss": 0.0172, "step": 38360 }, { "grad_norm": 0.17100243270397186, "learning_rate": 5.747654010560202e-05, "loss": 0.0205, "step": 38370 }, { "grad_norm": 0.24879658222198486, "learning_rate": 5.7456103467807097e-05, "loss": 0.0228, "step": 38380 }, { "grad_norm": 0.20508074760437012, "learning_rate": 5.7435665555968046e-05, "loss": 0.0181, "step": 38390 }, { "grad_norm": 0.2606692612171173, "learning_rate": 5.74152263735771e-05, "loss": 0.0241, "step": 38400 }, { "grad_norm": 0.25832927227020264, "learning_rate": 5.739478592412677e-05, "loss": 0.0168, "step": 38410 }, { "grad_norm": 0.3201472759246826, "learning_rate": 5.7374344211109766e-05, "loss": 0.0235, "step": 38420 }, { "grad_norm": 0.345672607421875, "learning_rate": 5.735390123801904e-05, "loss": 0.0261, "step": 38430 }, { "grad_norm": 0.29693207144737244, "learning_rate": 5.7333457008347704e-05, "loss": 0.0225, "step": 38440 }, { "grad_norm": 0.3095987141132355, "learning_rate": 5.7313011525589156e-05, "loss": 0.0202, "step": 38450 }, { "grad_norm": 0.40798479318618774, "learning_rate": 5.729256479323694e-05, "loss": 0.0243, "step": 38460 }, { "grad_norm": 0.24762363731861115, "learning_rate": 5.727211681478485e-05, "loss": 0.0246, "step": 38470 }, { "grad_norm": 0.20438697934150696, "learning_rate": 5.7251667593726886e-05, "loss": 0.022, "step": 38480 }, { "grad_norm": 0.35583916306495667, "learning_rate": 5.723121713355728e-05, "loss": 0.0272, "step": 38490 }, { "grad_norm": 0.20747989416122437, "learning_rate": 5.721076543777044e-05, "loss": 0.0168, "step": 38500 }, { "grad_norm": 0.36409449577331543, "learning_rate": 5.7190312509860986e-05, "loss": 0.0209, "step": 38510 }, { "grad_norm": 0.2310967892408371, "learning_rate": 5.716985835332379e-05, "loss": 0.0207, "step": 38520 }, { "grad_norm": 0.23455102741718292, "learning_rate": 5.714940297165389e-05, "loss": 0.0187, "step": 38530 }, { "grad_norm": 0.29569879174232483, "learning_rate": 5.712894636834656e-05, "loss": 0.0181, "step": 38540 }, { "grad_norm": 0.25163471698760986, "learning_rate": 5.7108488546897276e-05, "loss": 0.0191, "step": 38550 }, { "grad_norm": 0.2660047709941864, "learning_rate": 5.708802951080172e-05, "loss": 0.0201, "step": 38560 }, { "grad_norm": 0.15783816576004028, "learning_rate": 5.706756926355576e-05, "loss": 0.0211, "step": 38570 }, { "grad_norm": 0.26575663685798645, "learning_rate": 5.704710780865554e-05, "loss": 0.0207, "step": 38580 }, { "grad_norm": 0.26232805848121643, "learning_rate": 5.7026645149597325e-05, "loss": 0.017, "step": 38590 }, { "grad_norm": 0.2680598199367523, "learning_rate": 5.700618128987764e-05, "loss": 0.0222, "step": 38600 }, { "grad_norm": 0.47395724058151245, "learning_rate": 5.698571623299317e-05, "loss": 0.0259, "step": 38610 }, { "grad_norm": 0.2024628221988678, "learning_rate": 5.696524998244086e-05, "loss": 0.0179, "step": 38620 }, { "grad_norm": 0.21616503596305847, "learning_rate": 5.6944782541717836e-05, "loss": 0.0262, "step": 38630 }, { "grad_norm": 0.24775777757167816, "learning_rate": 5.69243139143214e-05, "loss": 0.0199, "step": 38640 }, { "grad_norm": 0.234715536236763, "learning_rate": 5.6903844103749125e-05, "loss": 0.0232, "step": 38650 }, { "grad_norm": 0.32174548506736755, "learning_rate": 5.688337311349869e-05, "loss": 0.0178, "step": 38660 }, { "grad_norm": 0.23346322774887085, "learning_rate": 5.6862900947068074e-05, "loss": 0.0183, "step": 38670 }, { "grad_norm": 0.24867463111877441, "learning_rate": 5.6842427607955374e-05, "loss": 0.0252, "step": 38680 }, { "grad_norm": 0.2648284137248993, "learning_rate": 5.682195309965893e-05, "loss": 0.0204, "step": 38690 }, { "grad_norm": 0.28356876969337463, "learning_rate": 5.6801477425677294e-05, "loss": 0.0185, "step": 38700 }, { "grad_norm": 0.3163548707962036, "learning_rate": 5.678100058950917e-05, "loss": 0.0248, "step": 38710 }, { "grad_norm": 0.2641778588294983, "learning_rate": 5.676052259465352e-05, "loss": 0.0197, "step": 38720 }, { "grad_norm": 0.21304287016391754, "learning_rate": 5.674004344460945e-05, "loss": 0.0174, "step": 38730 }, { "grad_norm": 0.2160796821117401, "learning_rate": 5.6719563142876295e-05, "loss": 0.019, "step": 38740 }, { "grad_norm": 0.6028353571891785, "learning_rate": 5.669908169295359e-05, "loss": 0.0194, "step": 38750 }, { "grad_norm": 0.3398630917072296, "learning_rate": 5.667859909834105e-05, "loss": 0.0219, "step": 38760 }, { "grad_norm": 0.25747811794281006, "learning_rate": 5.6658115362538565e-05, "loss": 0.0193, "step": 38770 }, { "grad_norm": 0.3069658577442169, "learning_rate": 5.663763048904628e-05, "loss": 0.0192, "step": 38780 }, { "grad_norm": 0.28718361258506775, "learning_rate": 5.661714448136447e-05, "loss": 0.0192, "step": 38790 }, { "grad_norm": 0.27461764216423035, "learning_rate": 5.659665734299366e-05, "loss": 0.0176, "step": 38800 }, { "grad_norm": 0.2677569091320038, "learning_rate": 5.6576169077434516e-05, "loss": 0.0209, "step": 38810 }, { "grad_norm": 0.22376789152622223, "learning_rate": 5.6555679688187944e-05, "loss": 0.0201, "step": 38820 }, { "grad_norm": 0.3074408769607544, "learning_rate": 5.6535189178755e-05, "loss": 0.0234, "step": 38830 }, { "grad_norm": 0.5195147395133972, "learning_rate": 5.651469755263698e-05, "loss": 0.0206, "step": 38840 }, { "grad_norm": 0.5544376969337463, "learning_rate": 5.6494204813335316e-05, "loss": 0.0301, "step": 38850 }, { "grad_norm": 0.17726118862628937, "learning_rate": 5.647371096435168e-05, "loss": 0.0179, "step": 38860 }, { "grad_norm": 0.272112637758255, "learning_rate": 5.645321600918788e-05, "loss": 0.0211, "step": 38870 }, { "grad_norm": 0.838530957698822, "learning_rate": 5.643271995134597e-05, "loss": 0.0277, "step": 38880 }, { "grad_norm": 0.17899225652217865, "learning_rate": 5.641222279432814e-05, "loss": 0.0195, "step": 38890 }, { "grad_norm": 0.24504001438617706, "learning_rate": 5.6391724541636834e-05, "loss": 0.0166, "step": 38900 }, { "grad_norm": 0.24526244401931763, "learning_rate": 5.6371225196774605e-05, "loss": 0.02, "step": 38910 }, { "grad_norm": 0.2328873574733734, "learning_rate": 5.635072476324423e-05, "loss": 0.021, "step": 38920 }, { "grad_norm": 0.239574134349823, "learning_rate": 5.63302232445487e-05, "loss": 0.0168, "step": 38930 }, { "grad_norm": 0.22723768651485443, "learning_rate": 5.6309720644191144e-05, "loss": 0.0188, "step": 38940 }, { "grad_norm": 0.31024283170700073, "learning_rate": 5.628921696567491e-05, "loss": 0.0166, "step": 38950 }, { "grad_norm": 0.3429921567440033, "learning_rate": 5.62687122125035e-05, "loss": 0.0229, "step": 38960 }, { "grad_norm": 0.23150202631950378, "learning_rate": 5.624820638818062e-05, "loss": 0.0184, "step": 38970 }, { "grad_norm": 0.29262450337409973, "learning_rate": 5.6227699496210164e-05, "loss": 0.0263, "step": 38980 }, { "grad_norm": 0.1866903305053711, "learning_rate": 5.6207191540096195e-05, "loss": 0.0204, "step": 38990 }, { "grad_norm": 1.2352803945541382, "learning_rate": 5.618668252334296e-05, "loss": 0.0197, "step": 39000 }, { "grad_norm": 0.6912661790847778, "learning_rate": 5.616617244945488e-05, "loss": 0.0208, "step": 39010 }, { "grad_norm": 0.7594832181930542, "learning_rate": 5.614566132193656e-05, "loss": 0.0245, "step": 39020 }, { "grad_norm": 0.25497710704803467, "learning_rate": 5.612514914429282e-05, "loss": 0.0236, "step": 39030 }, { "grad_norm": 0.286928653717041, "learning_rate": 5.610463592002863e-05, "loss": 0.0197, "step": 39040 }, { "grad_norm": 0.2709594964981079, "learning_rate": 5.608412165264909e-05, "loss": 0.0177, "step": 39050 }, { "grad_norm": 0.21483679115772247, "learning_rate": 5.606360634565959e-05, "loss": 0.0172, "step": 39060 }, { "grad_norm": 0.2898222506046295, "learning_rate": 5.604309000256559e-05, "loss": 0.0214, "step": 39070 }, { "grad_norm": 0.22037313878536224, "learning_rate": 5.6022572626872785e-05, "loss": 0.0153, "step": 39080 }, { "grad_norm": 0.30099302530288696, "learning_rate": 5.600205422208704e-05, "loss": 0.0208, "step": 39090 }, { "grad_norm": 0.35336360335350037, "learning_rate": 5.59815347917144e-05, "loss": 0.0199, "step": 39100 }, { "grad_norm": 0.24648289382457733, "learning_rate": 5.596101433926103e-05, "loss": 0.0163, "step": 39110 }, { "grad_norm": 0.251578152179718, "learning_rate": 5.5940492868233364e-05, "loss": 0.0136, "step": 39120 }, { "grad_norm": 0.5534349083900452, "learning_rate": 5.591997038213793e-05, "loss": 0.0186, "step": 39130 }, { "grad_norm": 0.21269634366035461, "learning_rate": 5.5899446884481475e-05, "loss": 0.0143, "step": 39140 }, { "grad_norm": 0.25323182344436646, "learning_rate": 5.5878922378770906e-05, "loss": 0.0288, "step": 39150 }, { "grad_norm": 0.3729557693004608, "learning_rate": 5.5858396868513285e-05, "loss": 0.0231, "step": 39160 }, { "grad_norm": 0.34674176573753357, "learning_rate": 5.583787035721586e-05, "loss": 0.0181, "step": 39170 }, { "grad_norm": 0.20411045849323273, "learning_rate": 5.581734284838606e-05, "loss": 0.0187, "step": 39180 }, { "grad_norm": 0.3350410461425781, "learning_rate": 5.579681434553147e-05, "loss": 0.0186, "step": 39190 }, { "grad_norm": 0.2163783460855484, "learning_rate": 5.5776284852159854e-05, "loss": 0.0178, "step": 39200 }, { "grad_norm": 0.32130223512649536, "learning_rate": 5.575575437177913e-05, "loss": 0.0217, "step": 39210 }, { "grad_norm": 0.3492756187915802, "learning_rate": 5.573522290789742e-05, "loss": 0.0217, "step": 39220 }, { "grad_norm": 0.28370869159698486, "learning_rate": 5.571469046402298e-05, "loss": 0.0184, "step": 39230 }, { "grad_norm": 0.28164345026016235, "learning_rate": 5.5694157043664205e-05, "loss": 0.0191, "step": 39240 }, { "grad_norm": 0.28860583901405334, "learning_rate": 5.567362265032975e-05, "loss": 0.0165, "step": 39250 }, { "grad_norm": 0.14571429789066315, "learning_rate": 5.565308728752836e-05, "loss": 0.0175, "step": 39260 }, { "grad_norm": 0.2607364356517792, "learning_rate": 5.5632550958768956e-05, "loss": 0.0135, "step": 39270 }, { "grad_norm": 0.3064562678337097, "learning_rate": 5.5612013667560636e-05, "loss": 0.0193, "step": 39280 }, { "grad_norm": 0.33159494400024414, "learning_rate": 5.5591475417412676e-05, "loss": 0.0348, "step": 39290 }, { "grad_norm": 0.6337120532989502, "learning_rate": 5.557093621183451e-05, "loss": 0.0244, "step": 39300 }, { "grad_norm": 0.3086041212081909, "learning_rate": 5.55503960543357e-05, "loss": 0.0181, "step": 39310 }, { "grad_norm": 0.2499016970396042, "learning_rate": 5.552985494842601e-05, "loss": 0.0184, "step": 39320 }, { "grad_norm": 1.2157868146896362, "learning_rate": 5.550931289761534e-05, "loss": 0.0224, "step": 39330 }, { "grad_norm": 0.7845652103424072, "learning_rate": 5.548876990541378e-05, "loss": 0.0215, "step": 39340 }, { "grad_norm": 0.3019134998321533, "learning_rate": 5.5468225975331564e-05, "loss": 0.0241, "step": 39350 }, { "grad_norm": 0.24142153561115265, "learning_rate": 5.544768111087909e-05, "loss": 0.0174, "step": 39360 }, { "grad_norm": 0.31345441937446594, "learning_rate": 5.54271353155669e-05, "loss": 0.0214, "step": 39370 }, { "grad_norm": 0.24842685461044312, "learning_rate": 5.5406588592905715e-05, "loss": 0.0202, "step": 39380 }, { "grad_norm": 0.2087690681219101, "learning_rate": 5.5386040946406416e-05, "loss": 0.0223, "step": 39390 }, { "grad_norm": 0.25341954827308655, "learning_rate": 5.536549237958004e-05, "loss": 0.0235, "step": 39400 }, { "grad_norm": 0.2671382427215576, "learning_rate": 5.5344942895937744e-05, "loss": 0.0292, "step": 39410 }, { "grad_norm": 0.2908293902873993, "learning_rate": 5.5324392498990904e-05, "loss": 0.0203, "step": 39420 }, { "grad_norm": 0.43628886342048645, "learning_rate": 5.5303841192251e-05, "loss": 0.0189, "step": 39430 }, { "grad_norm": 0.2661730945110321, "learning_rate": 5.52832889792297e-05, "loss": 0.0242, "step": 39440 }, { "grad_norm": 0.30951768159866333, "learning_rate": 5.526273586343881e-05, "loss": 0.0193, "step": 39450 }, { "grad_norm": 0.3117400109767914, "learning_rate": 5.5242181848390306e-05, "loss": 0.0206, "step": 39460 }, { "grad_norm": 0.3837119936943054, "learning_rate": 5.5221626937596285e-05, "loss": 0.0204, "step": 39470 }, { "grad_norm": 0.23368678987026215, "learning_rate": 5.520107113456903e-05, "loss": 0.0229, "step": 39480 }, { "grad_norm": 0.3184764087200165, "learning_rate": 5.5180514442820974e-05, "loss": 0.0275, "step": 39490 }, { "grad_norm": 0.19842159748077393, "learning_rate": 5.515995686586469e-05, "loss": 0.018, "step": 39500 }, { "grad_norm": 0.309295654296875, "learning_rate": 5.5139398407212916e-05, "loss": 0.0222, "step": 39510 }, { "grad_norm": 0.29442062973976135, "learning_rate": 5.511883907037849e-05, "loss": 0.0226, "step": 39520 }, { "grad_norm": 0.30586928129196167, "learning_rate": 5.509827885887449e-05, "loss": 0.0163, "step": 39530 }, { "grad_norm": 0.28498193621635437, "learning_rate": 5.507771777621406e-05, "loss": 0.0182, "step": 39540 }, { "grad_norm": 0.46416693925857544, "learning_rate": 5.505715582591052e-05, "loss": 0.0254, "step": 39550 }, { "grad_norm": 0.24728408455848694, "learning_rate": 5.50365930114774e-05, "loss": 0.0211, "step": 39560 }, { "grad_norm": 0.22681747376918793, "learning_rate": 5.5016029336428255e-05, "loss": 0.0165, "step": 39570 }, { "grad_norm": 0.2451038658618927, "learning_rate": 5.49954648042769e-05, "loss": 0.02, "step": 39580 }, { "grad_norm": 0.24708187580108643, "learning_rate": 5.4974899418537226e-05, "loss": 0.0243, "step": 39590 }, { "grad_norm": 0.21367964148521423, "learning_rate": 5.4954333182723316e-05, "loss": 0.0193, "step": 39600 }, { "grad_norm": 0.4193442463874817, "learning_rate": 5.493376610034937e-05, "loss": 0.0217, "step": 39610 }, { "grad_norm": 0.3088887333869934, "learning_rate": 5.4913198174929735e-05, "loss": 0.0247, "step": 39620 }, { "grad_norm": 0.18800196051597595, "learning_rate": 5.489262940997889e-05, "loss": 0.0195, "step": 39630 }, { "grad_norm": 0.2592799663543701, "learning_rate": 5.487205980901151e-05, "loss": 0.0266, "step": 39640 }, { "grad_norm": 0.28994220495224, "learning_rate": 5.485148937554234e-05, "loss": 0.0342, "step": 39650 }, { "grad_norm": 0.9973123669624329, "learning_rate": 5.483091811308635e-05, "loss": 0.0211, "step": 39660 }, { "grad_norm": 0.33402419090270996, "learning_rate": 5.4810346025158564e-05, "loss": 0.0182, "step": 39670 }, { "grad_norm": 0.3286122679710388, "learning_rate": 5.478977311527421e-05, "loss": 0.0202, "step": 39680 }, { "grad_norm": 0.350434273481369, "learning_rate": 5.476919938694863e-05, "loss": 0.0194, "step": 39690 }, { "grad_norm": 1.1796934604644775, "learning_rate": 5.474862484369733e-05, "loss": 0.0148, "step": 39700 }, { "grad_norm": 0.24309691786766052, "learning_rate": 5.472804948903589e-05, "loss": 0.0203, "step": 39710 }, { "grad_norm": 0.22733937203884125, "learning_rate": 5.470747332648013e-05, "loss": 0.0163, "step": 39720 }, { "grad_norm": 0.34541869163513184, "learning_rate": 5.468689635954591e-05, "loss": 0.0301, "step": 39730 }, { "grad_norm": 0.1823350340127945, "learning_rate": 5.46663185917493e-05, "loss": 0.0246, "step": 39740 }, { "grad_norm": 0.26322993636131287, "learning_rate": 5.464574002660645e-05, "loss": 0.0209, "step": 39750 }, { "grad_norm": 0.2507983446121216, "learning_rate": 5.4625160667633724e-05, "loss": 0.0232, "step": 39760 }, { "grad_norm": 0.1804058998823166, "learning_rate": 5.4604580518347505e-05, "loss": 0.0205, "step": 39770 }, { "grad_norm": 0.20975224673748016, "learning_rate": 5.458399958226442e-05, "loss": 0.0213, "step": 39780 }, { "grad_norm": 0.22405722737312317, "learning_rate": 5.456341786290119e-05, "loss": 0.0193, "step": 39790 }, { "grad_norm": 0.24855485558509827, "learning_rate": 5.454283536377465e-05, "loss": 0.0252, "step": 39800 }, { "grad_norm": 0.3399214744567871, "learning_rate": 5.452225208840179e-05, "loss": 0.0228, "step": 39810 }, { "grad_norm": 0.314710795879364, "learning_rate": 5.450166804029975e-05, "loss": 0.0241, "step": 39820 }, { "grad_norm": 0.3243914842605591, "learning_rate": 5.448108322298574e-05, "loss": 0.0164, "step": 39830 }, { "grad_norm": 0.34273627400398254, "learning_rate": 5.446049763997717e-05, "loss": 0.0188, "step": 39840 }, { "grad_norm": 0.2422255277633667, "learning_rate": 5.4439911294791546e-05, "loss": 0.0185, "step": 39850 }, { "grad_norm": 0.2804301381111145, "learning_rate": 5.441932419094652e-05, "loss": 0.0149, "step": 39860 }, { "grad_norm": 0.19218851625919342, "learning_rate": 5.439873633195985e-05, "loss": 0.0187, "step": 39870 }, { "grad_norm": 0.9274442791938782, "learning_rate": 5.437814772134947e-05, "loss": 0.028, "step": 39880 }, { "grad_norm": 0.3036034405231476, "learning_rate": 5.4357558362633366e-05, "loss": 0.021, "step": 39890 }, { "grad_norm": 0.24026449024677277, "learning_rate": 5.4336968259329726e-05, "loss": 0.0192, "step": 39900 }, { "grad_norm": 0.29582610726356506, "learning_rate": 5.431637741495681e-05, "loss": 0.0211, "step": 39910 }, { "grad_norm": 0.268972247838974, "learning_rate": 5.429578583303307e-05, "loss": 0.0231, "step": 39920 }, { "grad_norm": 0.2070341408252716, "learning_rate": 5.427519351707701e-05, "loss": 0.0219, "step": 39930 }, { "grad_norm": 0.2506856918334961, "learning_rate": 5.42546004706073e-05, "loss": 0.0214, "step": 39940 }, { "grad_norm": 0.2609613835811615, "learning_rate": 5.4234006697142735e-05, "loss": 0.0198, "step": 39950 }, { "grad_norm": 0.26372814178466797, "learning_rate": 5.421341220020224e-05, "loss": 0.0237, "step": 39960 }, { "grad_norm": 0.20138709247112274, "learning_rate": 5.419281698330482e-05, "loss": 0.0288, "step": 39970 }, { "grad_norm": 0.23536276817321777, "learning_rate": 5.4172221049969665e-05, "loss": 0.0234, "step": 39980 }, { "grad_norm": 0.3736990988254547, "learning_rate": 5.415162440371604e-05, "loss": 0.0241, "step": 39990 }, { "grad_norm": 0.2866295874118805, "learning_rate": 5.413102704806334e-05, "loss": 0.0267, "step": 40000 }, { "grad_norm": 0.17228420078754425, "learning_rate": 5.41104289865311e-05, "loss": 0.0184, "step": 40010 }, { "grad_norm": 0.18951477110385895, "learning_rate": 5.408983022263898e-05, "loss": 0.0179, "step": 40020 }, { "grad_norm": 0.34379589557647705, "learning_rate": 5.406923075990673e-05, "loss": 0.0252, "step": 40030 }, { "grad_norm": 0.24657151103019714, "learning_rate": 5.404863060185423e-05, "loss": 0.0197, "step": 40040 }, { "grad_norm": 0.24157775938510895, "learning_rate": 5.40280297520015e-05, "loss": 0.0204, "step": 40050 }, { "grad_norm": 0.2850571274757385, "learning_rate": 5.400742821386865e-05, "loss": 0.0174, "step": 40060 }, { "grad_norm": 0.3087683320045471, "learning_rate": 5.398682599097592e-05, "loss": 0.0247, "step": 40070 }, { "grad_norm": 0.5097824335098267, "learning_rate": 5.396622308684367e-05, "loss": 0.021, "step": 40080 }, { "grad_norm": 0.3219623267650604, "learning_rate": 5.394561950499236e-05, "loss": 0.0295, "step": 40090 }, { "grad_norm": 0.5000823736190796, "learning_rate": 5.39250152489426e-05, "loss": 0.0235, "step": 40100 }, { "grad_norm": 0.391128271818161, "learning_rate": 5.390441032221507e-05, "loss": 0.0223, "step": 40110 }, { "grad_norm": 0.29443496465682983, "learning_rate": 5.388380472833062e-05, "loss": 0.021, "step": 40120 }, { "grad_norm": 0.22290565073490143, "learning_rate": 5.386319847081016e-05, "loss": 0.0213, "step": 40130 }, { "grad_norm": 0.2845292091369629, "learning_rate": 5.384259155317473e-05, "loss": 0.0193, "step": 40140 }, { "grad_norm": 0.3699212372303009, "learning_rate": 5.38219839789455e-05, "loss": 0.0237, "step": 40150 }, { "grad_norm": 0.40540263056755066, "learning_rate": 5.380137575164376e-05, "loss": 0.0197, "step": 40160 }, { "grad_norm": 0.28736135363578796, "learning_rate": 5.378076687479085e-05, "loss": 0.0165, "step": 40170 }, { "grad_norm": 0.3127382695674896, "learning_rate": 5.3760157351908305e-05, "loss": 0.0194, "step": 40180 }, { "grad_norm": 0.2649577260017395, "learning_rate": 5.373954718651768e-05, "loss": 0.0203, "step": 40190 }, { "grad_norm": 0.15359817445278168, "learning_rate": 5.371893638214074e-05, "loss": 0.0186, "step": 40200 }, { "grad_norm": 0.23634912073612213, "learning_rate": 5.369832494229927e-05, "loss": 0.0203, "step": 40210 }, { "grad_norm": 0.2630312442779541, "learning_rate": 5.367771287051524e-05, "loss": 0.0264, "step": 40220 }, { "grad_norm": 0.18966460227966309, "learning_rate": 5.365710017031065e-05, "loss": 0.018, "step": 40230 }, { "grad_norm": 0.24533770978450775, "learning_rate": 5.3636486845207654e-05, "loss": 0.0227, "step": 40240 }, { "grad_norm": 0.24856647849082947, "learning_rate": 5.3615872898728526e-05, "loss": 0.0189, "step": 40250 }, { "grad_norm": 0.30083587765693665, "learning_rate": 5.3595258334395614e-05, "loss": 0.0222, "step": 40260 }, { "grad_norm": 0.2689748704433441, "learning_rate": 5.3574643155731396e-05, "loss": 0.0277, "step": 40270 }, { "grad_norm": 0.27016744017601013, "learning_rate": 5.3554027366258405e-05, "loss": 0.0174, "step": 40280 }, { "grad_norm": 0.3796261250972748, "learning_rate": 5.3533410969499355e-05, "loss": 0.021, "step": 40290 }, { "grad_norm": 0.22256813943386078, "learning_rate": 5.3512793968977e-05, "loss": 0.0156, "step": 40300 }, { "grad_norm": 0.3000052571296692, "learning_rate": 5.349217636821423e-05, "loss": 0.0183, "step": 40310 }, { "grad_norm": 0.17911000549793243, "learning_rate": 5.347155817073405e-05, "loss": 0.0132, "step": 40320 }, { "grad_norm": 0.27539321780204773, "learning_rate": 5.34509393800595e-05, "loss": 0.017, "step": 40330 }, { "grad_norm": 0.2483714371919632, "learning_rate": 5.343031999971381e-05, "loss": 0.0166, "step": 40340 }, { "grad_norm": 0.2574527859687805, "learning_rate": 5.340970003322026e-05, "loss": 0.0164, "step": 40350 }, { "grad_norm": 0.20733681321144104, "learning_rate": 5.33890794841022e-05, "loss": 0.0178, "step": 40360 }, { "grad_norm": 0.196502223610878, "learning_rate": 5.336845835588318e-05, "loss": 0.0209, "step": 40370 }, { "grad_norm": 0.21587282419204712, "learning_rate": 5.334783665208674e-05, "loss": 0.0209, "step": 40380 }, { "grad_norm": 0.26582270860671997, "learning_rate": 5.332721437623657e-05, "loss": 0.018, "step": 40390 }, { "grad_norm": 0.251494437456131, "learning_rate": 5.3306591531856464e-05, "loss": 0.0185, "step": 40400 }, { "grad_norm": 0.22455249726772308, "learning_rate": 5.3285968122470295e-05, "loss": 0.0186, "step": 40410 }, { "grad_norm": 0.304557204246521, "learning_rate": 5.326534415160207e-05, "loss": 0.0202, "step": 40420 }, { "grad_norm": 0.24723400175571442, "learning_rate": 5.324471962277582e-05, "loss": 0.0211, "step": 40430 }, { "grad_norm": 0.27698636054992676, "learning_rate": 5.3224094539515746e-05, "loss": 0.0166, "step": 40440 }, { "grad_norm": 0.4459763467311859, "learning_rate": 5.3203468905346076e-05, "loss": 0.0196, "step": 40450 }, { "grad_norm": 0.24891746044158936, "learning_rate": 5.3182842723791195e-05, "loss": 0.0177, "step": 40460 }, { "grad_norm": 0.19348011910915375, "learning_rate": 5.316221599837554e-05, "loss": 0.0211, "step": 40470 }, { "grad_norm": 0.3305526375770569, "learning_rate": 5.314158873262366e-05, "loss": 0.0257, "step": 40480 }, { "grad_norm": 0.209937185049057, "learning_rate": 5.312096093006018e-05, "loss": 0.0155, "step": 40490 }, { "grad_norm": 0.24820251762866974, "learning_rate": 5.3100332594209846e-05, "loss": 0.0186, "step": 40500 }, { "grad_norm": 0.21443119645118713, "learning_rate": 5.3079703728597454e-05, "loss": 0.0168, "step": 40510 }, { "grad_norm": 0.21410036087036133, "learning_rate": 5.305907433674794e-05, "loss": 0.0204, "step": 40520 }, { "grad_norm": 0.32218462228775024, "learning_rate": 5.3038444422186284e-05, "loss": 0.0181, "step": 40530 }, { "grad_norm": 0.20063655078411102, "learning_rate": 5.301781398843757e-05, "loss": 0.0184, "step": 40540 }, { "grad_norm": 0.5657885670661926, "learning_rate": 5.299718303902699e-05, "loss": 0.0223, "step": 40550 }, { "grad_norm": 0.1975337117910385, "learning_rate": 5.297655157747979e-05, "loss": 0.0202, "step": 40560 }, { "grad_norm": 0.3673538267612457, "learning_rate": 5.295591960732136e-05, "loss": 0.0151, "step": 40570 }, { "grad_norm": 0.28434786200523376, "learning_rate": 5.293528713207708e-05, "loss": 0.0191, "step": 40580 }, { "grad_norm": 1.4936585426330566, "learning_rate": 5.291465415527253e-05, "loss": 0.0215, "step": 40590 }, { "grad_norm": 0.24795855581760406, "learning_rate": 5.2894020680433296e-05, "loss": 0.0177, "step": 40600 }, { "grad_norm": 0.2617652714252472, "learning_rate": 5.287338671108507e-05, "loss": 0.0227, "step": 40610 }, { "grad_norm": 0.23504184186458588, "learning_rate": 5.285275225075367e-05, "loss": 0.0196, "step": 40620 }, { "grad_norm": 0.2442466765642166, "learning_rate": 5.283211730296492e-05, "loss": 0.0174, "step": 40630 }, { "grad_norm": 0.39934831857681274, "learning_rate": 5.281148187124477e-05, "loss": 0.0217, "step": 40640 }, { "grad_norm": 0.25117018818855286, "learning_rate": 5.279084595911927e-05, "loss": 0.0193, "step": 40650 }, { "grad_norm": 0.28898003697395325, "learning_rate": 5.277020957011453e-05, "loss": 0.0252, "step": 40660 }, { "grad_norm": 0.2661616802215576, "learning_rate": 5.274957270775673e-05, "loss": 0.0211, "step": 40670 }, { "grad_norm": 0.21106314659118652, "learning_rate": 5.2728935375572164e-05, "loss": 0.0165, "step": 40680 }, { "grad_norm": 0.3504481017589569, "learning_rate": 5.2708297577087176e-05, "loss": 0.019, "step": 40690 }, { "grad_norm": 0.3678848147392273, "learning_rate": 5.2687659315828196e-05, "loss": 0.0193, "step": 40700 }, { "grad_norm": 0.18229657411575317, "learning_rate": 5.266702059532175e-05, "loss": 0.0169, "step": 40710 }, { "grad_norm": 1.163973093032837, "learning_rate": 5.264638141909444e-05, "loss": 0.0219, "step": 40720 }, { "grad_norm": 0.33958345651626587, "learning_rate": 5.2625741790672925e-05, "loss": 0.02, "step": 40730 }, { "grad_norm": 0.34966930747032166, "learning_rate": 5.2605101713583935e-05, "loss": 0.024, "step": 40740 }, { "grad_norm": 0.30362311005592346, "learning_rate": 5.2584461191354315e-05, "loss": 0.022, "step": 40750 }, { "grad_norm": 0.19811008870601654, "learning_rate": 5.2563820227510964e-05, "loss": 0.0146, "step": 40760 }, { "grad_norm": 0.24162541329860687, "learning_rate": 5.2543178825580864e-05, "loss": 0.0155, "step": 40770 }, { "grad_norm": 0.47220274806022644, "learning_rate": 5.2522536989091054e-05, "loss": 0.0241, "step": 40780 }, { "grad_norm": 0.21811188757419586, "learning_rate": 5.2501894721568655e-05, "loss": 0.0165, "step": 40790 }, { "grad_norm": 0.24882104992866516, "learning_rate": 5.248125202654089e-05, "loss": 0.0191, "step": 40800 }, { "grad_norm": 0.24015262722969055, "learning_rate": 5.246060890753501e-05, "loss": 0.0237, "step": 40810 }, { "grad_norm": 0.22127699851989746, "learning_rate": 5.243996536807837e-05, "loss": 0.0142, "step": 40820 }, { "grad_norm": 0.21792525053024292, "learning_rate": 5.2419321411698384e-05, "loss": 0.0176, "step": 40830 }, { "grad_norm": 0.23306985199451447, "learning_rate": 5.239867704192253e-05, "loss": 0.0251, "step": 40840 }, { "grad_norm": 0.19310705363750458, "learning_rate": 5.237803226227838e-05, "loss": 0.0147, "step": 40850 }, { "grad_norm": 0.2058684378862381, "learning_rate": 5.235738707629354e-05, "loss": 0.018, "step": 40860 }, { "grad_norm": 0.2261742651462555, "learning_rate": 5.233674148749575e-05, "loss": 0.0197, "step": 40870 }, { "grad_norm": 0.2681918144226074, "learning_rate": 5.231609549941272e-05, "loss": 0.0184, "step": 40880 }, { "grad_norm": 0.14282774925231934, "learning_rate": 5.229544911557233e-05, "loss": 0.0168, "step": 40890 }, { "grad_norm": 0.16787664592266083, "learning_rate": 5.227480233950246e-05, "loss": 0.0195, "step": 40900 }, { "grad_norm": 0.2504044473171234, "learning_rate": 5.22541551747311e-05, "loss": 0.0221, "step": 40910 }, { "grad_norm": 0.2902292311191559, "learning_rate": 5.223350762478626e-05, "loss": 0.0207, "step": 40920 }, { "grad_norm": 0.25810113549232483, "learning_rate": 5.221285969319606e-05, "loss": 0.0164, "step": 40930 }, { "grad_norm": 0.20099741220474243, "learning_rate": 5.219221138348865e-05, "loss": 0.0159, "step": 40940 }, { "grad_norm": 0.20971132814884186, "learning_rate": 5.217156269919228e-05, "loss": 0.02, "step": 40950 }, { "grad_norm": 0.1827419102191925, "learning_rate": 5.215091364383523e-05, "loss": 0.015, "step": 40960 }, { "grad_norm": 0.30066898465156555, "learning_rate": 5.213026422094588e-05, "loss": 0.0205, "step": 40970 }, { "grad_norm": 0.4652920067310333, "learning_rate": 5.210961443405262e-05, "loss": 0.0188, "step": 40980 }, { "grad_norm": 0.24409958720207214, "learning_rate": 5.208896428668396e-05, "loss": 0.0176, "step": 40990 }, { "grad_norm": 0.243011012673378, "learning_rate": 5.206831378236845e-05, "loss": 0.0257, "step": 41000 }, { "grad_norm": 0.26447516679763794, "learning_rate": 5.2047662924634666e-05, "loss": 0.0166, "step": 41010 }, { "grad_norm": 0.24395179748535156, "learning_rate": 5.20270117170113e-05, "loss": 0.0164, "step": 41020 }, { "grad_norm": 0.33787819743156433, "learning_rate": 5.200636016302707e-05, "loss": 0.0225, "step": 41030 }, { "grad_norm": 0.3500593900680542, "learning_rate": 5.1985708266210754e-05, "loss": 0.024, "step": 41040 }, { "grad_norm": 0.23202131688594818, "learning_rate": 5.1965056030091206e-05, "loss": 0.0241, "step": 41050 }, { "grad_norm": 0.24687407910823822, "learning_rate": 5.194440345819732e-05, "loss": 0.0177, "step": 41060 }, { "grad_norm": 0.24386970698833466, "learning_rate": 5.1923750554058084e-05, "loss": 0.0206, "step": 41070 }, { "grad_norm": 0.3670593500137329, "learning_rate": 5.1903097321202476e-05, "loss": 0.0169, "step": 41080 }, { "grad_norm": 0.2895781397819519, "learning_rate": 5.18824437631596e-05, "loss": 0.0283, "step": 41090 }, { "grad_norm": 0.26783546805381775, "learning_rate": 5.186178988345856e-05, "loss": 0.0222, "step": 41100 }, { "grad_norm": 0.297214537858963, "learning_rate": 5.184113568562855e-05, "loss": 0.029, "step": 41110 }, { "grad_norm": 0.26985517144203186, "learning_rate": 5.18204811731988e-05, "loss": 0.0231, "step": 41120 }, { "grad_norm": 0.28916823863983154, "learning_rate": 5.179982634969862e-05, "loss": 0.0229, "step": 41130 }, { "grad_norm": 0.27699655294418335, "learning_rate": 5.177917121865734e-05, "loss": 0.0241, "step": 41140 }, { "grad_norm": 0.23574306070804596, "learning_rate": 5.1758515783604346e-05, "loss": 0.0154, "step": 41150 }, { "grad_norm": 0.29334360361099243, "learning_rate": 5.173786004806911e-05, "loss": 0.0226, "step": 41160 }, { "grad_norm": 0.2685334384441376, "learning_rate": 5.1717204015581135e-05, "loss": 0.0206, "step": 41170 }, { "grad_norm": 0.3939460515975952, "learning_rate": 5.169654768966994e-05, "loss": 0.0155, "step": 41180 }, { "grad_norm": 0.24526485800743103, "learning_rate": 5.1675891073865156e-05, "loss": 0.0185, "step": 41190 }, { "grad_norm": 0.16174004971981049, "learning_rate": 5.1655234171696424e-05, "loss": 0.0163, "step": 41200 }, { "grad_norm": 0.3084802031517029, "learning_rate": 5.163457698669343e-05, "loss": 0.0156, "step": 41210 }, { "grad_norm": 0.18751658499240875, "learning_rate": 5.1613919522385936e-05, "loss": 0.0166, "step": 41220 }, { "grad_norm": 0.3883068859577179, "learning_rate": 5.1593261782303746e-05, "loss": 0.025, "step": 41230 }, { "grad_norm": 0.948348343372345, "learning_rate": 5.157260376997669e-05, "loss": 0.0187, "step": 41240 }, { "grad_norm": 0.2465340942144394, "learning_rate": 5.155194548893466e-05, "loss": 0.0177, "step": 41250 }, { "grad_norm": 0.378131240606308, "learning_rate": 5.153128694270759e-05, "loss": 0.0279, "step": 41260 }, { "grad_norm": 0.3382014036178589, "learning_rate": 5.151062813482548e-05, "loss": 0.0232, "step": 41270 }, { "grad_norm": 0.21768419444561005, "learning_rate": 5.148996906881832e-05, "loss": 0.0172, "step": 41280 }, { "grad_norm": 0.2572985589504242, "learning_rate": 5.1469309748216196e-05, "loss": 0.0202, "step": 41290 }, { "grad_norm": 0.28660860657691956, "learning_rate": 5.144865017654923e-05, "loss": 0.0149, "step": 41300 }, { "grad_norm": 0.2974141538143158, "learning_rate": 5.1427990357347556e-05, "loss": 0.0224, "step": 41310 }, { "grad_norm": 0.2980635166168213, "learning_rate": 5.140733029414139e-05, "loss": 0.0172, "step": 41320 }, { "grad_norm": 0.38700228929519653, "learning_rate": 5.138666999046098e-05, "loss": 0.0176, "step": 41330 }, { "grad_norm": 0.22050096094608307, "learning_rate": 5.136600944983658e-05, "loss": 0.015, "step": 41340 }, { "grad_norm": 0.41534915566444397, "learning_rate": 5.134534867579853e-05, "loss": 0.0194, "step": 41350 }, { "grad_norm": 0.31589004397392273, "learning_rate": 5.1324687671877184e-05, "loss": 0.0158, "step": 41360 }, { "grad_norm": 0.3329075872898102, "learning_rate": 5.130402644160296e-05, "loss": 0.0215, "step": 41370 }, { "grad_norm": 0.182650625705719, "learning_rate": 5.128336498850628e-05, "loss": 0.0157, "step": 41380 }, { "grad_norm": 0.26813048124313354, "learning_rate": 5.126270331611761e-05, "loss": 0.0168, "step": 41390 }, { "grad_norm": 0.33608728647232056, "learning_rate": 5.124204142796748e-05, "loss": 0.0141, "step": 41400 }, { "grad_norm": 0.25734254717826843, "learning_rate": 5.122137932758644e-05, "loss": 0.0202, "step": 41410 }, { "grad_norm": 0.2590499818325043, "learning_rate": 5.1200717018505074e-05, "loss": 0.0173, "step": 41420 }, { "grad_norm": 0.19052407145500183, "learning_rate": 5.118005450425403e-05, "loss": 0.018, "step": 41430 }, { "grad_norm": 3.082240104675293, "learning_rate": 5.115939178836392e-05, "loss": 0.0149, "step": 41440 }, { "grad_norm": 0.1545058935880661, "learning_rate": 5.113872887436547e-05, "loss": 0.0193, "step": 41450 }, { "grad_norm": 0.23530319333076477, "learning_rate": 5.11180657657894e-05, "loss": 0.0193, "step": 41460 }, { "grad_norm": 0.2805802524089813, "learning_rate": 5.109740246616649e-05, "loss": 0.0173, "step": 41470 }, { "grad_norm": 0.41597771644592285, "learning_rate": 5.10767389790275e-05, "loss": 0.0248, "step": 41480 }, { "grad_norm": 0.28283748030662537, "learning_rate": 5.1056075307903265e-05, "loss": 0.0184, "step": 41490 }, { "grad_norm": 0.20370811223983765, "learning_rate": 5.1035411456324634e-05, "loss": 0.02, "step": 41500 }, { "grad_norm": 0.30722716450691223, "learning_rate": 5.1014747427822504e-05, "loss": 0.0187, "step": 41510 }, { "grad_norm": 0.2707784175872803, "learning_rate": 5.0994083225927804e-05, "loss": 0.0216, "step": 41520 }, { "grad_norm": 0.2597369849681854, "learning_rate": 5.0973418854171475e-05, "loss": 0.0213, "step": 41530 }, { "grad_norm": 0.25090736150741577, "learning_rate": 5.095275431608447e-05, "loss": 0.017, "step": 41540 }, { "grad_norm": 0.19469720125198364, "learning_rate": 5.093208961519782e-05, "loss": 0.0219, "step": 41550 }, { "grad_norm": 0.20332732796669006, "learning_rate": 5.091142475504255e-05, "loss": 0.0204, "step": 41560 }, { "grad_norm": 0.2969757616519928, "learning_rate": 5.089075973914971e-05, "loss": 0.0228, "step": 41570 }, { "grad_norm": 0.42176708579063416, "learning_rate": 5.08700945710504e-05, "loss": 0.0259, "step": 41580 }, { "grad_norm": 0.2700521647930145, "learning_rate": 5.0849429254275714e-05, "loss": 0.0207, "step": 41590 }, { "grad_norm": 1.3104153871536255, "learning_rate": 5.0828763792356804e-05, "loss": 0.0173, "step": 41600 }, { "grad_norm": 0.29708555340766907, "learning_rate": 5.080809818882483e-05, "loss": 0.0215, "step": 41610 }, { "grad_norm": 0.28318291902542114, "learning_rate": 5.078743244721097e-05, "loss": 0.018, "step": 41620 }, { "grad_norm": 0.209715336561203, "learning_rate": 5.0766766571046455e-05, "loss": 0.0263, "step": 41630 }, { "grad_norm": 0.2596905529499054, "learning_rate": 5.07461005638625e-05, "loss": 0.0176, "step": 41640 }, { "grad_norm": 0.23545244336128235, "learning_rate": 5.072543442919037e-05, "loss": 0.0164, "step": 41650 }, { "grad_norm": 0.24430014193058014, "learning_rate": 5.070476817056132e-05, "loss": 0.025, "step": 41660 }, { "grad_norm": 0.3895542025566101, "learning_rate": 5.068410179150668e-05, "loss": 0.0206, "step": 41670 }, { "grad_norm": 0.26077380776405334, "learning_rate": 5.066343529555775e-05, "loss": 0.0169, "step": 41680 }, { "grad_norm": 0.2199060320854187, "learning_rate": 5.064276868624587e-05, "loss": 0.0158, "step": 41690 }, { "grad_norm": 0.19113801419734955, "learning_rate": 5.0622101967102396e-05, "loss": 0.0173, "step": 41700 }, { "grad_norm": 0.2226005494594574, "learning_rate": 5.0601435141658714e-05, "loss": 0.0164, "step": 41710 }, { "grad_norm": 0.2684522569179535, "learning_rate": 5.058076821344622e-05, "loss": 0.023, "step": 41720 }, { "grad_norm": 0.3081246018409729, "learning_rate": 5.056010118599632e-05, "loss": 0.0232, "step": 41730 }, { "grad_norm": 0.2945026457309723, "learning_rate": 5.053943406284044e-05, "loss": 0.0186, "step": 41740 }, { "grad_norm": 0.260486900806427, "learning_rate": 5.051876684751006e-05, "loss": 0.0155, "step": 41750 }, { "grad_norm": 0.2876989543437958, "learning_rate": 5.0498099543536584e-05, "loss": 0.0188, "step": 41760 }, { "grad_norm": 0.2471098154783249, "learning_rate": 5.047743215445152e-05, "loss": 0.0245, "step": 41770 }, { "grad_norm": 0.34025102853775024, "learning_rate": 5.045676468378637e-05, "loss": 0.0193, "step": 41780 }, { "grad_norm": 0.37697482109069824, "learning_rate": 5.0436097135072626e-05, "loss": 0.0226, "step": 41790 }, { "grad_norm": 0.3381823003292084, "learning_rate": 5.041542951184181e-05, "loss": 0.0182, "step": 41800 }, { "grad_norm": 0.1870785653591156, "learning_rate": 5.039476181762545e-05, "loss": 0.0137, "step": 41810 }, { "grad_norm": 0.665259599685669, "learning_rate": 5.037409405595508e-05, "loss": 0.0164, "step": 41820 }, { "grad_norm": 0.3149176239967346, "learning_rate": 5.035342623036229e-05, "loss": 0.0232, "step": 41830 }, { "grad_norm": 0.4103599786758423, "learning_rate": 5.033275834437862e-05, "loss": 0.0183, "step": 41840 }, { "grad_norm": 0.23962023854255676, "learning_rate": 5.031209040153564e-05, "loss": 0.0186, "step": 41850 }, { "grad_norm": 0.43544241786003113, "learning_rate": 5.0291422405364955e-05, "loss": 0.0235, "step": 41860 }, { "grad_norm": 1.1356685161590576, "learning_rate": 5.0270754359398133e-05, "loss": 0.0174, "step": 41870 }, { "grad_norm": 0.19363772869110107, "learning_rate": 5.025008626716682e-05, "loss": 0.0186, "step": 41880 }, { "grad_norm": 0.30212846398353577, "learning_rate": 5.0229418132202585e-05, "loss": 0.017, "step": 41890 }, { "grad_norm": 0.22318139672279358, "learning_rate": 5.020874995803707e-05, "loss": 0.0186, "step": 41900 }, { "grad_norm": 0.2075495570898056, "learning_rate": 5.01880817482019e-05, "loss": 0.0171, "step": 41910 }, { "grad_norm": 0.20707081258296967, "learning_rate": 5.01674135062287e-05, "loss": 0.0175, "step": 41920 }, { "grad_norm": 0.2592337429523468, "learning_rate": 5.0146745235649115e-05, "loss": 0.0186, "step": 41930 }, { "grad_norm": 0.21162578463554382, "learning_rate": 5.0126076939994795e-05, "loss": 0.0172, "step": 41940 }, { "grad_norm": 0.20193825662136078, "learning_rate": 5.010540862279736e-05, "loss": 0.02, "step": 41950 }, { "grad_norm": 0.2070237249135971, "learning_rate": 5.008474028758846e-05, "loss": 0.0166, "step": 41960 }, { "grad_norm": 0.20068052411079407, "learning_rate": 5.0064071937899784e-05, "loss": 0.0175, "step": 41970 }, { "grad_norm": 0.34835729002952576, "learning_rate": 5.004340357726296e-05, "loss": 0.0185, "step": 41980 }, { "grad_norm": 0.4322966933250427, "learning_rate": 5.002273520920965e-05, "loss": 0.0232, "step": 41990 }, { "grad_norm": 0.8670461177825928, "learning_rate": 5.000206683727151e-05, "loss": 0.0195, "step": 42000 }, { "grad_norm": 0.3133798837661743, "learning_rate": 4.998139846498021e-05, "loss": 0.0195, "step": 42010 }, { "grad_norm": 0.236979141831398, "learning_rate": 4.9960730095867405e-05, "loss": 0.0228, "step": 42020 }, { "grad_norm": 0.18982374668121338, "learning_rate": 4.9940061733464755e-05, "loss": 0.0205, "step": 42030 }, { "grad_norm": 0.1454552859067917, "learning_rate": 4.991939338130392e-05, "loss": 0.0273, "step": 42040 }, { "grad_norm": 0.3451796770095825, "learning_rate": 4.989872504291653e-05, "loss": 0.0236, "step": 42050 }, { "grad_norm": 0.18255828320980072, "learning_rate": 4.9878056721834273e-05, "loss": 0.0192, "step": 42060 }, { "grad_norm": 0.3250386416912079, "learning_rate": 4.98573884215888e-05, "loss": 0.0184, "step": 42070 }, { "grad_norm": 0.26864194869995117, "learning_rate": 4.9836720145711715e-05, "loss": 0.0197, "step": 42080 }, { "grad_norm": 0.28746703267097473, "learning_rate": 4.9816051897734725e-05, "loss": 0.02, "step": 42090 }, { "grad_norm": 0.24357813596725464, "learning_rate": 4.979538368118941e-05, "loss": 0.0175, "step": 42100 }, { "grad_norm": 0.20147061347961426, "learning_rate": 4.9774715499607446e-05, "loss": 0.0174, "step": 42110 }, { "grad_norm": 0.2603920102119446, "learning_rate": 4.9754047356520416e-05, "loss": 0.0233, "step": 42120 }, { "grad_norm": 0.2987242639064789, "learning_rate": 4.973337925545999e-05, "loss": 0.0174, "step": 42130 }, { "grad_norm": 0.24446719884872437, "learning_rate": 4.9712711199957746e-05, "loss": 0.0202, "step": 42140 }, { "grad_norm": 0.3472447395324707, "learning_rate": 4.969204319354531e-05, "loss": 0.0233, "step": 42150 }, { "grad_norm": 0.30646181106567383, "learning_rate": 4.9671375239754267e-05, "loss": 0.0178, "step": 42160 }, { "grad_norm": 0.27731817960739136, "learning_rate": 4.9650707342116195e-05, "loss": 0.0216, "step": 42170 }, { "grad_norm": 0.2011021077632904, "learning_rate": 4.96300395041627e-05, "loss": 0.0156, "step": 42180 }, { "grad_norm": 0.30294033885002136, "learning_rate": 4.960937172942532e-05, "loss": 0.0178, "step": 42190 }, { "grad_norm": 0.20924663543701172, "learning_rate": 4.958870402143565e-05, "loss": 0.0154, "step": 42200 }, { "grad_norm": 0.2230026125907898, "learning_rate": 4.9568036383725186e-05, "loss": 0.0199, "step": 42210 }, { "grad_norm": 0.25932028889656067, "learning_rate": 4.9547368819825516e-05, "loss": 0.0184, "step": 42220 }, { "grad_norm": 0.2583252191543579, "learning_rate": 4.952670133326812e-05, "loss": 0.0171, "step": 42230 }, { "grad_norm": 0.26757296919822693, "learning_rate": 4.950603392758453e-05, "loss": 0.0184, "step": 42240 }, { "grad_norm": 0.22796104848384857, "learning_rate": 4.948536660630621e-05, "loss": 0.0136, "step": 42250 }, { "grad_norm": 0.28224411606788635, "learning_rate": 4.9464699372964685e-05, "loss": 0.0216, "step": 42260 }, { "grad_norm": 0.3065575957298279, "learning_rate": 4.9444032231091395e-05, "loss": 0.0229, "step": 42270 }, { "grad_norm": 0.26954206824302673, "learning_rate": 4.9423365184217776e-05, "loss": 0.0214, "step": 42280 }, { "grad_norm": 0.32957640290260315, "learning_rate": 4.940269823587529e-05, "loss": 0.0179, "step": 42290 }, { "grad_norm": 0.2564055919647217, "learning_rate": 4.938203138959533e-05, "loss": 0.0174, "step": 42300 }, { "grad_norm": 0.2946532964706421, "learning_rate": 4.9361364648909325e-05, "loss": 0.019, "step": 42310 }, { "grad_norm": 0.4849565327167511, "learning_rate": 4.9340698017348605e-05, "loss": 0.0168, "step": 42320 }, { "grad_norm": 0.8706334233283997, "learning_rate": 4.932003149844458e-05, "loss": 0.0179, "step": 42330 }, { "grad_norm": 0.9023663997650146, "learning_rate": 4.929936509572857e-05, "loss": 0.0267, "step": 42340 }, { "grad_norm": 0.2968117892742157, "learning_rate": 4.927869881273191e-05, "loss": 0.0289, "step": 42350 }, { "grad_norm": 0.22953712940216064, "learning_rate": 4.9258032652985894e-05, "loss": 0.0202, "step": 42360 }, { "grad_norm": 0.25055748224258423, "learning_rate": 4.9237366620021786e-05, "loss": 0.018, "step": 42370 }, { "grad_norm": 0.1997644007205963, "learning_rate": 4.921670071737089e-05, "loss": 0.0157, "step": 42380 }, { "grad_norm": 0.22931142151355743, "learning_rate": 4.91960349485644e-05, "loss": 0.0196, "step": 42390 }, { "grad_norm": 0.2163461446762085, "learning_rate": 4.9175369317133566e-05, "loss": 0.0163, "step": 42400 }, { "grad_norm": 0.2956235408782959, "learning_rate": 4.915470382660953e-05, "loss": 0.0162, "step": 42410 }, { "grad_norm": 0.2320258468389511, "learning_rate": 4.9134038480523524e-05, "loss": 0.0159, "step": 42420 }, { "grad_norm": 0.2521032691001892, "learning_rate": 4.911337328240664e-05, "loss": 0.0194, "step": 42430 }, { "grad_norm": 0.22697171568870544, "learning_rate": 4.909270823579003e-05, "loss": 0.0195, "step": 42440 }, { "grad_norm": 0.2548355758190155, "learning_rate": 4.907204334420476e-05, "loss": 0.0189, "step": 42450 }, { "grad_norm": 0.3654727339744568, "learning_rate": 4.9051378611181884e-05, "loss": 0.0201, "step": 42460 }, { "grad_norm": 0.24361583590507507, "learning_rate": 4.9030714040252486e-05, "loss": 0.0182, "step": 42470 }, { "grad_norm": 0.21829462051391602, "learning_rate": 4.901004963494752e-05, "loss": 0.0147, "step": 42480 }, { "grad_norm": 0.38182222843170166, "learning_rate": 4.898938539879802e-05, "loss": 0.0203, "step": 42490 }, { "grad_norm": 0.2723025679588318, "learning_rate": 4.8968721335334904e-05, "loss": 0.0196, "step": 42500 }, { "grad_norm": 0.2194550782442093, "learning_rate": 4.894805744808912e-05, "loss": 0.0242, "step": 42510 }, { "grad_norm": 0.21940521895885468, "learning_rate": 4.8927393740591524e-05, "loss": 0.0189, "step": 42520 }, { "grad_norm": 0.26050639152526855, "learning_rate": 4.890673021637302e-05, "loss": 0.0172, "step": 42530 }, { "grad_norm": 0.2031823843717575, "learning_rate": 4.888606687896442e-05, "loss": 0.0156, "step": 42540 }, { "grad_norm": 0.1892385482788086, "learning_rate": 4.886540373189652e-05, "loss": 0.0201, "step": 42550 }, { "grad_norm": 0.2558199167251587, "learning_rate": 4.88447407787001e-05, "loss": 0.0179, "step": 42560 }, { "grad_norm": 0.3123477101325989, "learning_rate": 4.8824078022905864e-05, "loss": 0.0183, "step": 42570 }, { "grad_norm": 0.5210453271865845, "learning_rate": 4.880341546804455e-05, "loss": 0.0181, "step": 42580 }, { "grad_norm": 0.25726860761642456, "learning_rate": 4.878275311764679e-05, "loss": 0.0179, "step": 42590 }, { "grad_norm": 0.3026622533798218, "learning_rate": 4.8762090975243254e-05, "loss": 0.0162, "step": 42600 }, { "grad_norm": 0.21585169434547424, "learning_rate": 4.874142904436448e-05, "loss": 0.0156, "step": 42610 }, { "grad_norm": 0.2248256653547287, "learning_rate": 4.8720767328541094e-05, "loss": 0.0227, "step": 42620 }, { "grad_norm": 0.7840204238891602, "learning_rate": 4.870010583130358e-05, "loss": 0.0185, "step": 42630 }, { "grad_norm": 0.1825999915599823, "learning_rate": 4.867944455618241e-05, "loss": 0.0156, "step": 42640 }, { "grad_norm": 0.20504911243915558, "learning_rate": 4.865878350670807e-05, "loss": 0.0189, "step": 42650 }, { "grad_norm": 0.2986091077327728, "learning_rate": 4.8638122686410914e-05, "loss": 0.0165, "step": 42660 }, { "grad_norm": 0.21306201815605164, "learning_rate": 4.861746209882137e-05, "loss": 0.0197, "step": 42670 }, { "grad_norm": 0.27191033959388733, "learning_rate": 4.859680174746972e-05, "loss": 0.0174, "step": 42680 }, { "grad_norm": 0.2910085916519165, "learning_rate": 4.857614163588629e-05, "loss": 0.0243, "step": 42690 }, { "grad_norm": 0.24923363327980042, "learning_rate": 4.855548176760131e-05, "loss": 0.0214, "step": 42700 }, { "grad_norm": 0.3804454505443573, "learning_rate": 4.8534822146144986e-05, "loss": 0.0278, "step": 42710 }, { "grad_norm": 0.23971621692180634, "learning_rate": 4.851416277504749e-05, "loss": 0.0291, "step": 42720 }, { "grad_norm": 0.2031698077917099, "learning_rate": 4.8493503657838923e-05, "loss": 0.0148, "step": 42730 }, { "grad_norm": 0.3456363379955292, "learning_rate": 4.84728447980494e-05, "loss": 0.0184, "step": 42740 }, { "grad_norm": 0.38083234429359436, "learning_rate": 4.8452186199208914e-05, "loss": 0.0208, "step": 42750 }, { "grad_norm": 0.4128211736679077, "learning_rate": 4.843152786484749e-05, "loss": 0.0189, "step": 42760 }, { "grad_norm": 0.266427606344223, "learning_rate": 4.8410869798495035e-05, "loss": 0.0156, "step": 42770 }, { "grad_norm": 0.3139861524105072, "learning_rate": 4.8390212003681486e-05, "loss": 0.0211, "step": 42780 }, { "grad_norm": 0.3602313697338104, "learning_rate": 4.836955448393667e-05, "loss": 0.0205, "step": 42790 }, { "grad_norm": 0.2106761783361435, "learning_rate": 4.8348897242790406e-05, "loss": 0.0183, "step": 42800 }, { "grad_norm": 0.21880364418029785, "learning_rate": 4.832824028377243e-05, "loss": 0.0212, "step": 42810 }, { "grad_norm": 0.27642586827278137, "learning_rate": 4.830758361041249e-05, "loss": 0.0164, "step": 42820 }, { "grad_norm": 0.2482910454273224, "learning_rate": 4.828692722624022e-05, "loss": 0.0194, "step": 42830 }, { "grad_norm": 0.23578861355781555, "learning_rate": 4.826627113478522e-05, "loss": 0.0306, "step": 42840 }, { "grad_norm": 0.22448673844337463, "learning_rate": 4.824561533957708e-05, "loss": 0.0191, "step": 42850 }, { "grad_norm": 0.23791606724262238, "learning_rate": 4.822495984414527e-05, "loss": 0.0192, "step": 42860 }, { "grad_norm": 0.2891470789909363, "learning_rate": 4.8204304652019304e-05, "loss": 0.0195, "step": 42870 }, { "grad_norm": 0.4284594655036926, "learning_rate": 4.8183649766728537e-05, "loss": 0.0186, "step": 42880 }, { "grad_norm": 0.31260064244270325, "learning_rate": 4.816299519180237e-05, "loss": 0.0203, "step": 42890 }, { "grad_norm": 0.26363784074783325, "learning_rate": 4.8142340930770066e-05, "loss": 0.0185, "step": 42900 }, { "grad_norm": 0.1760357767343521, "learning_rate": 4.81216869871609e-05, "loss": 0.0181, "step": 42910 }, { "grad_norm": 0.2823197841644287, "learning_rate": 4.8101033364504064e-05, "loss": 0.0161, "step": 42920 }, { "grad_norm": 0.28463298082351685, "learning_rate": 4.808038006632866e-05, "loss": 0.0237, "step": 42930 }, { "grad_norm": 0.4767812490463257, "learning_rate": 4.805972709616383e-05, "loss": 0.0191, "step": 42940 }, { "grad_norm": 0.17958635091781616, "learning_rate": 4.8039074457538556e-05, "loss": 0.0157, "step": 42950 }, { "grad_norm": 0.6953539252281189, "learning_rate": 4.801842215398184e-05, "loss": 0.0212, "step": 42960 }, { "grad_norm": 0.3001043200492859, "learning_rate": 4.799777018902256e-05, "loss": 0.0182, "step": 42970 }, { "grad_norm": 0.26122602820396423, "learning_rate": 4.797711856618961e-05, "loss": 0.0168, "step": 42980 }, { "grad_norm": 0.7114275693893433, "learning_rate": 4.7956467289011765e-05, "loss": 0.0166, "step": 42990 }, { "grad_norm": 0.2046930491924286, "learning_rate": 4.793581636101778e-05, "loss": 0.0147, "step": 43000 }, { "grad_norm": 0.17926953732967377, "learning_rate": 4.7915165785736326e-05, "loss": 0.0158, "step": 43010 }, { "grad_norm": 0.30803897976875305, "learning_rate": 4.789451556669599e-05, "loss": 0.0184, "step": 43020 }, { "grad_norm": 0.25323542952537537, "learning_rate": 4.7873865707425384e-05, "loss": 0.0233, "step": 43030 }, { "grad_norm": 0.2224389761686325, "learning_rate": 4.7853216211452975e-05, "loss": 0.02, "step": 43040 }, { "grad_norm": 0.48254963755607605, "learning_rate": 4.78325670823072e-05, "loss": 0.0218, "step": 43050 }, { "grad_norm": 0.17854546010494232, "learning_rate": 4.781191832351641e-05, "loss": 0.0149, "step": 43060 }, { "grad_norm": 0.21554800868034363, "learning_rate": 4.7791269938608965e-05, "loss": 0.0211, "step": 43070 }, { "grad_norm": 0.27053481340408325, "learning_rate": 4.777062193111305e-05, "loss": 0.0204, "step": 43080 }, { "grad_norm": 0.22232167422771454, "learning_rate": 4.77499743045569e-05, "loss": 0.0187, "step": 43090 }, { "grad_norm": 0.19296668469905853, "learning_rate": 4.7729327062468596e-05, "loss": 0.0165, "step": 43100 }, { "grad_norm": 0.23945511877536774, "learning_rate": 4.770868020837619e-05, "loss": 0.018, "step": 43110 }, { "grad_norm": 0.19105789065361023, "learning_rate": 4.768803374580768e-05, "loss": 0.0156, "step": 43120 }, { "grad_norm": 0.1930237114429474, "learning_rate": 4.766738767829094e-05, "loss": 0.0172, "step": 43130 }, { "grad_norm": 0.246732696890831, "learning_rate": 4.764674200935388e-05, "loss": 0.0189, "step": 43140 }, { "grad_norm": 0.3425185978412628, "learning_rate": 4.762609674252424e-05, "loss": 0.0156, "step": 43150 }, { "grad_norm": 0.2335970252752304, "learning_rate": 4.760545188132974e-05, "loss": 0.0274, "step": 43160 }, { "grad_norm": 0.24399754405021667, "learning_rate": 4.758480742929801e-05, "loss": 0.0199, "step": 43170 }, { "grad_norm": 0.46716269850730896, "learning_rate": 4.756416338995664e-05, "loss": 0.0191, "step": 43180 }, { "grad_norm": 0.3060495853424072, "learning_rate": 4.7543519766833145e-05, "loss": 0.014, "step": 43190 }, { "grad_norm": 0.19434423744678497, "learning_rate": 4.752287656345492e-05, "loss": 0.0185, "step": 43200 }, { "grad_norm": 0.45197242498397827, "learning_rate": 4.7502233783349355e-05, "loss": 0.0204, "step": 43210 }, { "grad_norm": 0.1795637458562851, "learning_rate": 4.7481591430043694e-05, "loss": 0.0141, "step": 43220 }, { "grad_norm": 0.20890536904335022, "learning_rate": 4.746094950706521e-05, "loss": 0.0179, "step": 43230 }, { "grad_norm": 0.21210402250289917, "learning_rate": 4.744030801794099e-05, "loss": 0.0151, "step": 43240 }, { "grad_norm": 0.4811718463897705, "learning_rate": 4.741966696619813e-05, "loss": 0.0234, "step": 43250 }, { "grad_norm": 0.2189328670501709, "learning_rate": 4.739902635536361e-05, "loss": 0.0227, "step": 43260 }, { "grad_norm": 0.532379686832428, "learning_rate": 4.737838618896436e-05, "loss": 0.0181, "step": 43270 }, { "grad_norm": 0.1911710947751999, "learning_rate": 4.7357746470527203e-05, "loss": 0.0211, "step": 43280 }, { "grad_norm": 0.25493454933166504, "learning_rate": 4.73371072035789e-05, "loss": 0.0235, "step": 43290 }, { "grad_norm": 0.32829374074935913, "learning_rate": 4.731646839164616e-05, "loss": 0.0191, "step": 43300 }, { "grad_norm": 0.30379581451416016, "learning_rate": 4.729583003825556e-05, "loss": 0.0192, "step": 43310 }, { "grad_norm": 0.24400775134563446, "learning_rate": 4.727519214693367e-05, "loss": 0.0155, "step": 43320 }, { "grad_norm": 0.24726715683937073, "learning_rate": 4.725455472120689e-05, "loss": 0.0191, "step": 43330 }, { "grad_norm": 0.21030761301517487, "learning_rate": 4.723391776460164e-05, "loss": 0.019, "step": 43340 }, { "grad_norm": 0.2275860756635666, "learning_rate": 4.7213281280644186e-05, "loss": 0.0178, "step": 43350 }, { "grad_norm": 0.2874888479709625, "learning_rate": 4.719264527286075e-05, "loss": 0.0154, "step": 43360 }, { "grad_norm": 0.25375697016716003, "learning_rate": 4.717200974477744e-05, "loss": 0.0226, "step": 43370 }, { "grad_norm": 0.31581079959869385, "learning_rate": 4.715137469992034e-05, "loss": 0.0172, "step": 43380 }, { "grad_norm": 0.3927260637283325, "learning_rate": 4.713074014181539e-05, "loss": 0.0212, "step": 43390 }, { "grad_norm": 0.22756826877593994, "learning_rate": 4.711010607398846e-05, "loss": 0.0151, "step": 43400 }, { "grad_norm": 0.37139832973480225, "learning_rate": 4.708947249996538e-05, "loss": 0.0133, "step": 43410 }, { "grad_norm": 0.19384826719760895, "learning_rate": 4.706883942327183e-05, "loss": 0.0168, "step": 43420 }, { "grad_norm": 0.18634527921676636, "learning_rate": 4.704820684743347e-05, "loss": 0.0175, "step": 43430 }, { "grad_norm": 0.3068481981754303, "learning_rate": 4.702757477597581e-05, "loss": 0.0236, "step": 43440 }, { "grad_norm": 0.19160237908363342, "learning_rate": 4.700694321242435e-05, "loss": 0.0179, "step": 43450 }, { "grad_norm": 0.4127812683582306, "learning_rate": 4.698631216030441e-05, "loss": 0.023, "step": 43460 }, { "grad_norm": 0.16222167015075684, "learning_rate": 4.6965681623141314e-05, "loss": 0.0161, "step": 43470 }, { "grad_norm": 0.1665978878736496, "learning_rate": 4.694505160446024e-05, "loss": 0.012, "step": 43480 }, { "grad_norm": 0.22016744315624237, "learning_rate": 4.692442210778628e-05, "loss": 0.015, "step": 43490 }, { "grad_norm": 0.2985580861568451, "learning_rate": 4.690379313664448e-05, "loss": 0.0204, "step": 43500 }, { "grad_norm": 0.3352065980434418, "learning_rate": 4.688316469455973e-05, "loss": 0.0194, "step": 43510 }, { "grad_norm": 0.42403435707092285, "learning_rate": 4.6862536785056914e-05, "loss": 0.0153, "step": 43520 }, { "grad_norm": 0.22835706174373627, "learning_rate": 4.684190941166073e-05, "loss": 0.0191, "step": 43530 }, { "grad_norm": 1.6029244661331177, "learning_rate": 4.682128257789587e-05, "loss": 0.0172, "step": 43540 }, { "grad_norm": 0.3717387318611145, "learning_rate": 4.680065628728687e-05, "loss": 0.0137, "step": 43550 }, { "grad_norm": 0.20024371147155762, "learning_rate": 4.678003054335822e-05, "loss": 0.0183, "step": 43560 }, { "grad_norm": 0.24704988300800323, "learning_rate": 4.675940534963428e-05, "loss": 0.021, "step": 43570 }, { "grad_norm": 0.2257295697927475, "learning_rate": 4.673878070963931e-05, "loss": 0.0174, "step": 43580 }, { "grad_norm": 0.2482125461101532, "learning_rate": 4.671815662689756e-05, "loss": 0.0117, "step": 43590 }, { "grad_norm": 0.19130562245845795, "learning_rate": 4.669753310493306e-05, "loss": 0.0147, "step": 43600 }, { "grad_norm": 0.20509029924869537, "learning_rate": 4.6676910147269845e-05, "loss": 0.0198, "step": 43610 }, { "grad_norm": 0.23511870205402374, "learning_rate": 4.665628775743177e-05, "loss": 0.0182, "step": 43620 }, { "grad_norm": 0.22913387417793274, "learning_rate": 4.66356659389427e-05, "loss": 0.0178, "step": 43630 }, { "grad_norm": 0.24491927027702332, "learning_rate": 4.661504469532629e-05, "loss": 0.0182, "step": 43640 }, { "grad_norm": 0.1732669621706009, "learning_rate": 4.659442403010617e-05, "loss": 0.0157, "step": 43650 }, { "grad_norm": 0.24818821251392365, "learning_rate": 4.6573803946805845e-05, "loss": 0.0209, "step": 43660 }, { "grad_norm": 0.3736559748649597, "learning_rate": 4.6553184448948695e-05, "loss": 0.0176, "step": 43670 }, { "grad_norm": 0.25876525044441223, "learning_rate": 4.653256554005807e-05, "loss": 0.0133, "step": 43680 }, { "grad_norm": 0.1800062209367752, "learning_rate": 4.6511947223657145e-05, "loss": 0.0244, "step": 43690 }, { "grad_norm": 0.399524986743927, "learning_rate": 4.649132950326906e-05, "loss": 0.0218, "step": 43700 }, { "grad_norm": 0.17935894429683685, "learning_rate": 4.647071238241679e-05, "loss": 0.0164, "step": 43710 }, { "grad_norm": 0.18512371182441711, "learning_rate": 4.6450095864623264e-05, "loss": 0.0174, "step": 43720 }, { "grad_norm": 0.2042349874973297, "learning_rate": 4.642947995341123e-05, "loss": 0.0168, "step": 43730 }, { "grad_norm": 0.2982701361179352, "learning_rate": 4.640886465230345e-05, "loss": 0.0186, "step": 43740 }, { "grad_norm": 0.27872052788734436, "learning_rate": 4.6388249964822485e-05, "loss": 0.0175, "step": 43750 }, { "grad_norm": 0.1933678388595581, "learning_rate": 4.63676358944908e-05, "loss": 0.0153, "step": 43760 }, { "grad_norm": 0.3478354811668396, "learning_rate": 4.634702244483081e-05, "loss": 0.0212, "step": 43770 }, { "grad_norm": 0.24902239441871643, "learning_rate": 4.632640961936475e-05, "loss": 0.019, "step": 43780 }, { "grad_norm": 0.245930016040802, "learning_rate": 4.6305797421614835e-05, "loss": 0.0193, "step": 43790 }, { "grad_norm": 0.23051707446575165, "learning_rate": 4.6285185855103105e-05, "loss": 0.0169, "step": 43800 }, { "grad_norm": 0.4117720127105713, "learning_rate": 4.626457492335151e-05, "loss": 0.0196, "step": 43810 }, { "grad_norm": 0.8799735903739929, "learning_rate": 4.624396462988188e-05, "loss": 0.0187, "step": 43820 }, { "grad_norm": 0.8297770023345947, "learning_rate": 4.6223354978215985e-05, "loss": 0.0209, "step": 43830 }, { "grad_norm": 0.33850690722465515, "learning_rate": 4.620274597187544e-05, "loss": 0.0183, "step": 43840 }, { "grad_norm": 0.23648987710475922, "learning_rate": 4.6182137614381736e-05, "loss": 0.0193, "step": 43850 }, { "grad_norm": 0.3616687059402466, "learning_rate": 4.616152990925631e-05, "loss": 0.0179, "step": 43860 }, { "grad_norm": 0.20273226499557495, "learning_rate": 4.614092286002041e-05, "loss": 0.0193, "step": 43870 }, { "grad_norm": 0.25078094005584717, "learning_rate": 4.612031647019528e-05, "loss": 0.021, "step": 43880 }, { "grad_norm": 0.2903795540332794, "learning_rate": 4.609971074330193e-05, "loss": 0.0188, "step": 43890 }, { "grad_norm": 0.3651714324951172, "learning_rate": 4.607910568286136e-05, "loss": 0.023, "step": 43900 }, { "grad_norm": 0.2426363229751587, "learning_rate": 4.605850129239438e-05, "loss": 0.0185, "step": 43910 }, { "grad_norm": 0.2523854374885559, "learning_rate": 4.603789757542174e-05, "loss": 0.0204, "step": 43920 }, { "grad_norm": 0.2888559103012085, "learning_rate": 4.601729453546404e-05, "loss": 0.0182, "step": 43930 }, { "grad_norm": 0.6420690417289734, "learning_rate": 4.599669217604177e-05, "loss": 0.0178, "step": 43940 }, { "grad_norm": 0.23007404804229736, "learning_rate": 4.597609050067532e-05, "loss": 0.0217, "step": 43950 }, { "grad_norm": 0.24350184202194214, "learning_rate": 4.595548951288495e-05, "loss": 0.0165, "step": 43960 }, { "grad_norm": 0.22507429122924805, "learning_rate": 4.593488921619081e-05, "loss": 0.0213, "step": 43970 }, { "grad_norm": 0.30582359433174133, "learning_rate": 4.591428961411289e-05, "loss": 0.018, "step": 43980 }, { "grad_norm": 0.2409704476594925, "learning_rate": 4.589369071017117e-05, "loss": 0.0229, "step": 43990 }, { "grad_norm": 0.19673606753349304, "learning_rate": 4.587309250788538e-05, "loss": 0.0182, "step": 44000 }, { "grad_norm": 0.27861201763153076, "learning_rate": 4.585249501077522e-05, "loss": 0.0161, "step": 44010 }, { "grad_norm": 0.17908895015716553, "learning_rate": 4.583189822236021e-05, "loss": 0.0187, "step": 44020 }, { "grad_norm": 0.29284507036209106, "learning_rate": 4.5811302146159816e-05, "loss": 0.0172, "step": 44030 }, { "grad_norm": 0.24405324459075928, "learning_rate": 4.579070678569332e-05, "loss": 0.0171, "step": 44040 }, { "grad_norm": 0.25671812891960144, "learning_rate": 4.5770112144479904e-05, "loss": 0.0167, "step": 44050 }, { "grad_norm": 0.28710949420928955, "learning_rate": 4.5749518226038645e-05, "loss": 0.0156, "step": 44060 }, { "grad_norm": 0.29712745547294617, "learning_rate": 4.572892503388845e-05, "loss": 0.0178, "step": 44070 }, { "grad_norm": 0.33914390206336975, "learning_rate": 4.570833257154817e-05, "loss": 0.0171, "step": 44080 }, { "grad_norm": 0.22326873242855072, "learning_rate": 4.568774084253646e-05, "loss": 0.0217, "step": 44090 }, { "grad_norm": 0.28293031454086304, "learning_rate": 4.566714985037191e-05, "loss": 0.0253, "step": 44100 }, { "grad_norm": 0.348170667886734, "learning_rate": 4.564655959857295e-05, "loss": 0.0237, "step": 44110 }, { "grad_norm": 0.2754281163215637, "learning_rate": 4.5625970090657893e-05, "loss": 0.0197, "step": 44120 }, { "grad_norm": 0.2290036976337433, "learning_rate": 4.560538133014491e-05, "loss": 0.0168, "step": 44130 }, { "grad_norm": 0.2403162717819214, "learning_rate": 4.5584793320552055e-05, "loss": 0.0168, "step": 44140 }, { "grad_norm": 0.2721412479877472, "learning_rate": 4.556420606539728e-05, "loss": 0.0185, "step": 44150 }, { "grad_norm": 0.3356305956840515, "learning_rate": 4.554361956819836e-05, "loss": 0.014, "step": 44160 }, { "grad_norm": 0.1932346373796463, "learning_rate": 4.552303383247299e-05, "loss": 0.0145, "step": 44170 }, { "grad_norm": 0.15947625041007996, "learning_rate": 4.5502448861738664e-05, "loss": 0.0143, "step": 44180 }, { "grad_norm": 0.21908602118492126, "learning_rate": 4.548186465951284e-05, "loss": 0.0194, "step": 44190 }, { "grad_norm": 0.18168121576309204, "learning_rate": 4.546128122931277e-05, "loss": 0.0177, "step": 44200 }, { "grad_norm": 0.1935957968235016, "learning_rate": 4.544069857465561e-05, "loss": 0.016, "step": 44210 }, { "grad_norm": 0.2166462391614914, "learning_rate": 4.542011669905837e-05, "loss": 0.0144, "step": 44220 }, { "grad_norm": 0.3043156564235687, "learning_rate": 4.53995356060379e-05, "loss": 0.0169, "step": 44230 }, { "grad_norm": 0.2330729365348816, "learning_rate": 4.5378955299110994e-05, "loss": 0.0177, "step": 44240 }, { "grad_norm": 0.22147968411445618, "learning_rate": 4.5358375781794233e-05, "loss": 0.0203, "step": 44250 }, { "grad_norm": 0.3003840744495392, "learning_rate": 4.533779705760411e-05, "loss": 0.0191, "step": 44260 }, { "grad_norm": 0.12391608953475952, "learning_rate": 4.5317219130056934e-05, "loss": 0.0171, "step": 44270 }, { "grad_norm": 0.4462156295776367, "learning_rate": 4.5296642002668964e-05, "loss": 0.0204, "step": 44280 }, { "grad_norm": 0.22572080790996552, "learning_rate": 4.5276065678956216e-05, "loss": 0.0226, "step": 44290 }, { "grad_norm": 0.2034844011068344, "learning_rate": 4.525549016243466e-05, "loss": 0.0197, "step": 44300 }, { "grad_norm": 0.47199711203575134, "learning_rate": 4.523491545662008e-05, "loss": 0.018, "step": 44310 }, { "grad_norm": 0.33985471725463867, "learning_rate": 4.5214341565028104e-05, "loss": 0.0202, "step": 44320 }, { "grad_norm": 0.309824675321579, "learning_rate": 4.519376849117428e-05, "loss": 0.0163, "step": 44330 }, { "grad_norm": 0.3668563961982727, "learning_rate": 4.517319623857395e-05, "loss": 0.0188, "step": 44340 }, { "grad_norm": 0.25438952445983887, "learning_rate": 4.5152624810742394e-05, "loss": 0.0166, "step": 44350 }, { "grad_norm": 0.1979197859764099, "learning_rate": 4.5132054211194664e-05, "loss": 0.0211, "step": 44360 }, { "grad_norm": 0.19999630749225616, "learning_rate": 4.511148444344574e-05, "loss": 0.0157, "step": 44370 }, { "grad_norm": 0.35145753622055054, "learning_rate": 4.509091551101041e-05, "loss": 0.0184, "step": 44380 }, { "grad_norm": 0.190191388130188, "learning_rate": 4.507034741740338e-05, "loss": 0.0158, "step": 44390 }, { "grad_norm": 0.260936439037323, "learning_rate": 4.5049780166139145e-05, "loss": 0.019, "step": 44400 }, { "grad_norm": 0.30875012278556824, "learning_rate": 4.5029213760732075e-05, "loss": 0.0294, "step": 44410 }, { "grad_norm": 0.5108751654624939, "learning_rate": 4.5008648204696434e-05, "loss": 0.0204, "step": 44420 }, { "grad_norm": 0.5202075839042664, "learning_rate": 4.4988083501546284e-05, "loss": 0.0173, "step": 44430 }, { "grad_norm": 0.15473566949367523, "learning_rate": 4.4967519654795606e-05, "loss": 0.0174, "step": 44440 }, { "grad_norm": 0.3257696330547333, "learning_rate": 4.494695666795816e-05, "loss": 0.018, "step": 44450 }, { "grad_norm": 0.19724829494953156, "learning_rate": 4.4926394544547644e-05, "loss": 0.0191, "step": 44460 }, { "grad_norm": 0.2135189324617386, "learning_rate": 4.490583328807752e-05, "loss": 0.0197, "step": 44470 }, { "grad_norm": 0.4026873707771301, "learning_rate": 4.488527290206117e-05, "loss": 0.0212, "step": 44480 }, { "grad_norm": 0.2777264714241028, "learning_rate": 4.48647133900118e-05, "loss": 0.0159, "step": 44490 }, { "grad_norm": 0.296632319688797, "learning_rate": 4.484415475544244e-05, "loss": 0.0187, "step": 44500 }, { "grad_norm": 0.19512799382209778, "learning_rate": 4.4823597001866035e-05, "loss": 0.0181, "step": 44510 }, { "grad_norm": 0.19621512293815613, "learning_rate": 4.480304013279532e-05, "loss": 0.0135, "step": 44520 }, { "grad_norm": 0.3803163468837738, "learning_rate": 4.478248415174292e-05, "loss": 0.0152, "step": 44530 }, { "grad_norm": 0.17246878147125244, "learning_rate": 4.476192906222126e-05, "loss": 0.02, "step": 44540 }, { "grad_norm": 0.21797215938568115, "learning_rate": 4.474137486774268e-05, "loss": 0.026, "step": 44550 }, { "grad_norm": 0.20397165417671204, "learning_rate": 4.4720821571819296e-05, "loss": 0.0207, "step": 44560 }, { "grad_norm": 0.24425598978996277, "learning_rate": 4.470026917796314e-05, "loss": 0.0253, "step": 44570 }, { "grad_norm": 0.3190232515335083, "learning_rate": 4.4679717689686005e-05, "loss": 0.0319, "step": 44580 }, { "grad_norm": 0.22910933196544647, "learning_rate": 4.465916711049963e-05, "loss": 0.015, "step": 44590 }, { "grad_norm": 0.16028232872486115, "learning_rate": 4.4638617443915524e-05, "loss": 0.0209, "step": 44600 }, { "grad_norm": 0.3121139109134674, "learning_rate": 4.4618068693445055e-05, "loss": 0.0168, "step": 44610 }, { "grad_norm": 0.2670271694660187, "learning_rate": 4.459752086259946e-05, "loss": 0.0186, "step": 44620 }, { "grad_norm": 0.3006736636161804, "learning_rate": 4.457697395488977e-05, "loss": 0.02, "step": 44630 }, { "grad_norm": 0.21476499736309052, "learning_rate": 4.455642797382693e-05, "loss": 0.0178, "step": 44640 }, { "grad_norm": 0.2478044629096985, "learning_rate": 4.4535882922921643e-05, "loss": 0.0162, "step": 44650 }, { "grad_norm": 0.35191088914871216, "learning_rate": 4.451533880568455e-05, "loss": 0.0162, "step": 44660 }, { "grad_norm": 0.21860137581825256, "learning_rate": 4.449479562562603e-05, "loss": 0.0196, "step": 44670 }, { "grad_norm": 0.3264029920101166, "learning_rate": 4.447425338625639e-05, "loss": 0.02, "step": 44680 }, { "grad_norm": 0.21370425820350647, "learning_rate": 4.4453712091085705e-05, "loss": 0.0174, "step": 44690 }, { "grad_norm": 0.2719903588294983, "learning_rate": 4.443317174362392e-05, "loss": 0.0208, "step": 44700 }, { "grad_norm": 0.2888801395893097, "learning_rate": 4.441263234738085e-05, "loss": 0.0169, "step": 44710 }, { "grad_norm": 0.2120313048362732, "learning_rate": 4.4392093905866086e-05, "loss": 0.0139, "step": 44720 }, { "grad_norm": 0.2376914918422699, "learning_rate": 4.437155642258911e-05, "loss": 0.0158, "step": 44730 }, { "grad_norm": 0.16083452105522156, "learning_rate": 4.4351019901059177e-05, "loss": 0.0223, "step": 44740 }, { "grad_norm": 0.23060530424118042, "learning_rate": 4.433048434478547e-05, "loss": 0.0192, "step": 44750 }, { "grad_norm": 0.202823668718338, "learning_rate": 4.4309949757276906e-05, "loss": 0.0173, "step": 44760 }, { "grad_norm": 0.20320181548595428, "learning_rate": 4.428941614204233e-05, "loss": 0.0271, "step": 44770 }, { "grad_norm": 0.2121322900056839, "learning_rate": 4.426888350259034e-05, "loss": 0.0183, "step": 44780 }, { "grad_norm": 0.13884402811527252, "learning_rate": 4.424835184242939e-05, "loss": 0.0252, "step": 44790 }, { "grad_norm": 1.065930724143982, "learning_rate": 4.422782116506784e-05, "loss": 0.0198, "step": 44800 }, { "grad_norm": 0.4712582528591156, "learning_rate": 4.4207291474013754e-05, "loss": 0.0252, "step": 44810 }, { "grad_norm": 0.2348434180021286, "learning_rate": 4.418676277277514e-05, "loss": 0.0233, "step": 44820 }, { "grad_norm": 0.2523181438446045, "learning_rate": 4.416623506485976e-05, "loss": 0.0271, "step": 44830 }, { "grad_norm": 0.3317185640335083, "learning_rate": 4.414570835377527e-05, "loss": 0.0151, "step": 44840 }, { "grad_norm": 0.18802958726882935, "learning_rate": 4.412518264302909e-05, "loss": 0.0197, "step": 44850 }, { "grad_norm": 0.32640576362609863, "learning_rate": 4.4104657936128535e-05, "loss": 0.0153, "step": 44860 }, { "grad_norm": 0.28154462575912476, "learning_rate": 4.408413423658071e-05, "loss": 0.0198, "step": 44870 }, { "grad_norm": 0.29615968465805054, "learning_rate": 4.406361154789253e-05, "loss": 0.015, "step": 44880 }, { "grad_norm": 0.29936400055885315, "learning_rate": 4.4043089873570776e-05, "loss": 0.0183, "step": 44890 }, { "grad_norm": 0.27183181047439575, "learning_rate": 4.4022569217122025e-05, "loss": 0.0188, "step": 44900 }, { "grad_norm": 0.21058137714862823, "learning_rate": 4.400204958205274e-05, "loss": 0.0126, "step": 44910 }, { "grad_norm": 0.18246690928936005, "learning_rate": 4.3981530971869125e-05, "loss": 0.0242, "step": 44920 }, { "grad_norm": 0.20556388795375824, "learning_rate": 4.396101339007727e-05, "loss": 0.0156, "step": 44930 }, { "grad_norm": 0.18568108975887299, "learning_rate": 4.394049684018304e-05, "loss": 0.0159, "step": 44940 }, { "grad_norm": 0.20938150584697723, "learning_rate": 4.3919981325692186e-05, "loss": 0.015, "step": 44950 }, { "grad_norm": 0.26110410690307617, "learning_rate": 4.389946685011024e-05, "loss": 0.0144, "step": 44960 }, { "grad_norm": 0.20407666265964508, "learning_rate": 4.387895341694255e-05, "loss": 0.0218, "step": 44970 }, { "grad_norm": 0.18266399204730988, "learning_rate": 4.3858441029694324e-05, "loss": 0.0167, "step": 44980 }, { "grad_norm": 0.3293518126010895, "learning_rate": 4.3837929691870527e-05, "loss": 0.0154, "step": 44990 }, { "grad_norm": 0.3179512023925781, "learning_rate": 4.381741940697604e-05, "loss": 0.0184, "step": 45000 }, { "grad_norm": 0.22744683921337128, "learning_rate": 4.379691017851547e-05, "loss": 0.0136, "step": 45010 }, { "grad_norm": 0.2046550065279007, "learning_rate": 4.3776402009993304e-05, "loss": 0.0188, "step": 45020 }, { "grad_norm": 0.2059420347213745, "learning_rate": 4.3755894904913794e-05, "loss": 0.0231, "step": 45030 }, { "grad_norm": 0.23268266022205353, "learning_rate": 4.373538886678109e-05, "loss": 0.0156, "step": 45040 }, { "grad_norm": 0.2697172462940216, "learning_rate": 4.371488389909909e-05, "loss": 0.0174, "step": 45050 }, { "grad_norm": 0.6145715713500977, "learning_rate": 4.3694380005371515e-05, "loss": 0.0204, "step": 45060 }, { "grad_norm": 0.39618122577667236, "learning_rate": 4.367387718910196e-05, "loss": 0.0164, "step": 45070 }, { "grad_norm": 0.18423610925674438, "learning_rate": 4.3653375453793764e-05, "loss": 0.0149, "step": 45080 }, { "grad_norm": 0.20102651417255402, "learning_rate": 4.3632874802950136e-05, "loss": 0.0178, "step": 45090 }, { "grad_norm": 0.27951329946517944, "learning_rate": 4.3612375240074034e-05, "loss": 0.0203, "step": 45100 }, { "grad_norm": 0.2184690684080124, "learning_rate": 4.3591876768668325e-05, "loss": 0.0162, "step": 45110 }, { "grad_norm": 0.4984557330608368, "learning_rate": 4.3571379392235605e-05, "loss": 0.0206, "step": 45120 }, { "grad_norm": 0.248573899269104, "learning_rate": 4.3550883114278335e-05, "loss": 0.0141, "step": 45130 }, { "grad_norm": 0.371432363986969, "learning_rate": 4.353038793829876e-05, "loss": 0.016, "step": 45140 }, { "grad_norm": 0.257864385843277, "learning_rate": 4.350989386779891e-05, "loss": 0.0178, "step": 45150 }, { "grad_norm": 0.1863301694393158, "learning_rate": 4.3489400906280724e-05, "loss": 0.023, "step": 45160 }, { "grad_norm": 0.21279506385326385, "learning_rate": 4.3468909057245845e-05, "loss": 0.018, "step": 45170 }, { "grad_norm": 0.2393822968006134, "learning_rate": 4.3448418324195794e-05, "loss": 0.0198, "step": 45180 }, { "grad_norm": 0.21497245132923126, "learning_rate": 4.342792871063184e-05, "loss": 0.0184, "step": 45190 }, { "grad_norm": 0.18518289923667908, "learning_rate": 4.3407440220055145e-05, "loss": 0.015, "step": 45200 }, { "grad_norm": 0.4175342619419098, "learning_rate": 4.33869528559666e-05, "loss": 0.0193, "step": 45210 }, { "grad_norm": 0.17639441788196564, "learning_rate": 4.336646662186696e-05, "loss": 0.0163, "step": 45220 }, { "grad_norm": 0.8703545331954956, "learning_rate": 4.334598152125672e-05, "loss": 0.0161, "step": 45230 }, { "grad_norm": 0.2738049030303955, "learning_rate": 4.3325497557636276e-05, "loss": 0.0232, "step": 45240 }, { "grad_norm": 0.2341957688331604, "learning_rate": 4.330501473450574e-05, "loss": 0.0211, "step": 45250 }, { "grad_norm": 0.2807331085205078, "learning_rate": 4.328453305536507e-05, "loss": 0.0142, "step": 45260 }, { "grad_norm": 0.22303591668605804, "learning_rate": 4.326405252371404e-05, "loss": 0.017, "step": 45270 }, { "grad_norm": 0.2400551736354828, "learning_rate": 4.324357314305221e-05, "loss": 0.0143, "step": 45280 }, { "grad_norm": 0.2538006007671356, "learning_rate": 4.3223094916878945e-05, "loss": 0.0165, "step": 45290 }, { "grad_norm": 0.22547881305217743, "learning_rate": 4.320261784869338e-05, "loss": 0.0239, "step": 45300 }, { "grad_norm": 0.3483096957206726, "learning_rate": 4.318214194199455e-05, "loss": 0.0179, "step": 45310 }, { "grad_norm": 0.20243918895721436, "learning_rate": 4.316166720028118e-05, "loss": 0.0166, "step": 45320 }, { "grad_norm": 0.26624587178230286, "learning_rate": 4.3141193627051864e-05, "loss": 0.0165, "step": 45330 }, { "grad_norm": 0.2022877037525177, "learning_rate": 4.312072122580496e-05, "loss": 0.0167, "step": 45340 }, { "grad_norm": 0.24811725318431854, "learning_rate": 4.3100250000038646e-05, "loss": 0.0161, "step": 45350 }, { "grad_norm": 0.35328516364097595, "learning_rate": 4.307977995325091e-05, "loss": 0.0252, "step": 45360 }, { "grad_norm": 0.3170638978481293, "learning_rate": 4.30593110889395e-05, "loss": 0.0175, "step": 45370 }, { "grad_norm": 0.33146893978118896, "learning_rate": 4.3038843410602016e-05, "loss": 0.0175, "step": 45380 }, { "grad_norm": 0.6706540584564209, "learning_rate": 4.3018376921735774e-05, "loss": 0.019, "step": 45390 }, { "grad_norm": 0.34774690866470337, "learning_rate": 4.299791162583799e-05, "loss": 0.0168, "step": 45400 }, { "grad_norm": 0.29975318908691406, "learning_rate": 4.29774475264056e-05, "loss": 0.0175, "step": 45410 }, { "grad_norm": 0.2179901897907257, "learning_rate": 4.2956984626935365e-05, "loss": 0.0145, "step": 45420 }, { "grad_norm": 0.3653479516506195, "learning_rate": 4.293652293092383e-05, "loss": 0.0194, "step": 45430 }, { "grad_norm": 0.25890496373176575, "learning_rate": 4.2916062441867324e-05, "loss": 0.0164, "step": 45440 }, { "grad_norm": 0.7305709719657898, "learning_rate": 4.289560316326201e-05, "loss": 0.022, "step": 45450 }, { "grad_norm": 0.27928876876831055, "learning_rate": 4.28751450986038e-05, "loss": 0.0176, "step": 45460 }, { "grad_norm": 0.2849401533603668, "learning_rate": 4.2854688251388444e-05, "loss": 0.0179, "step": 45470 }, { "grad_norm": 0.2704008221626282, "learning_rate": 4.2834232625111425e-05, "loss": 0.0246, "step": 45480 }, { "grad_norm": 0.2572838366031647, "learning_rate": 4.2813778223268086e-05, "loss": 0.0165, "step": 45490 }, { "grad_norm": 0.22594530880451202, "learning_rate": 4.2793325049353477e-05, "loss": 0.0171, "step": 45500 }, { "grad_norm": 0.2024126499891281, "learning_rate": 4.2772873106862535e-05, "loss": 0.0152, "step": 45510 }, { "grad_norm": 0.2133072465658188, "learning_rate": 4.275242239928993e-05, "loss": 0.0147, "step": 45520 }, { "grad_norm": 0.2401246875524521, "learning_rate": 4.273197293013009e-05, "loss": 0.0183, "step": 45530 }, { "grad_norm": 0.24709011614322662, "learning_rate": 4.271152470287731e-05, "loss": 0.0136, "step": 45540 }, { "grad_norm": 0.21638675034046173, "learning_rate": 4.2691077721025594e-05, "loss": 0.0194, "step": 45550 }, { "grad_norm": 0.2225571572780609, "learning_rate": 4.267063198806883e-05, "loss": 0.0156, "step": 45560 }, { "grad_norm": 0.2634811997413635, "learning_rate": 4.2650187507500574e-05, "loss": 0.0157, "step": 45570 }, { "grad_norm": 0.23774093389511108, "learning_rate": 4.2629744282814275e-05, "loss": 0.018, "step": 45580 }, { "grad_norm": 0.3085877597332001, "learning_rate": 4.2609302317503074e-05, "loss": 0.0161, "step": 45590 }, { "grad_norm": 0.24040839076042175, "learning_rate": 4.258886161505999e-05, "loss": 0.0151, "step": 45600 }, { "grad_norm": 0.2833147644996643, "learning_rate": 4.2568422178977775e-05, "loss": 0.0145, "step": 45610 }, { "grad_norm": 0.25663262605667114, "learning_rate": 4.254798401274894e-05, "loss": 0.0198, "step": 45620 }, { "grad_norm": 0.2508018910884857, "learning_rate": 4.252754711986583e-05, "loss": 0.0186, "step": 45630 }, { "grad_norm": 0.2604459226131439, "learning_rate": 4.250711150382052e-05, "loss": 0.013, "step": 45640 }, { "grad_norm": 0.18523511290550232, "learning_rate": 4.248667716810495e-05, "loss": 0.0135, "step": 45650 }, { "grad_norm": 1.0959166288375854, "learning_rate": 4.246624411621074e-05, "loss": 0.017, "step": 45660 }, { "grad_norm": 0.2526661157608032, "learning_rate": 4.244581235162938e-05, "loss": 0.0162, "step": 45670 }, { "grad_norm": 0.21108733117580414, "learning_rate": 4.2425381877852075e-05, "loss": 0.0207, "step": 45680 }, { "grad_norm": 0.24754203855991364, "learning_rate": 4.2404952698369856e-05, "loss": 0.0162, "step": 45690 }, { "grad_norm": 0.16403956711292267, "learning_rate": 4.238452481667349e-05, "loss": 0.0168, "step": 45700 }, { "grad_norm": 0.37145301699638367, "learning_rate": 4.2364098236253526e-05, "loss": 0.0216, "step": 45710 }, { "grad_norm": 0.2492321878671646, "learning_rate": 4.2343672960600356e-05, "loss": 0.0161, "step": 45720 }, { "grad_norm": 0.26119858026504517, "learning_rate": 4.232324899320406e-05, "loss": 0.0166, "step": 45730 }, { "grad_norm": 0.19812117516994476, "learning_rate": 4.230282633755457e-05, "loss": 0.0154, "step": 45740 }, { "grad_norm": 0.23142486810684204, "learning_rate": 4.2282404997141515e-05, "loss": 0.0191, "step": 45750 }, { "grad_norm": 0.18885913491249084, "learning_rate": 4.2261984975454397e-05, "loss": 0.0156, "step": 45760 }, { "grad_norm": 0.2557961642742157, "learning_rate": 4.224156627598239e-05, "loss": 0.0139, "step": 45770 }, { "grad_norm": 0.30231428146362305, "learning_rate": 4.222114890221453e-05, "loss": 0.0217, "step": 45780 }, { "grad_norm": 0.2543032467365265, "learning_rate": 4.2200732857639546e-05, "loss": 0.0235, "step": 45790 }, { "grad_norm": 0.2779066264629364, "learning_rate": 4.2180318145746035e-05, "loss": 0.0167, "step": 45800 }, { "grad_norm": 0.3997828960418701, "learning_rate": 4.215990477002227e-05, "loss": 0.0218, "step": 45810 }, { "grad_norm": 0.20918937027454376, "learning_rate": 4.2139492733956356e-05, "loss": 0.0124, "step": 45820 }, { "grad_norm": 0.22138920426368713, "learning_rate": 4.211908204103615e-05, "loss": 0.0257, "step": 45830 }, { "grad_norm": 0.3374978005886078, "learning_rate": 4.2098672694749265e-05, "loss": 0.0234, "step": 45840 }, { "grad_norm": 0.21976719796657562, "learning_rate": 4.2078264698583133e-05, "loss": 0.0181, "step": 45850 }, { "grad_norm": 0.20219917595386505, "learning_rate": 4.205785805602488e-05, "loss": 0.0176, "step": 45860 }, { "grad_norm": 0.29005110263824463, "learning_rate": 4.203745277056149e-05, "loss": 0.0195, "step": 45870 }, { "grad_norm": 0.19757488369941711, "learning_rate": 4.201704884567964e-05, "loss": 0.016, "step": 45880 }, { "grad_norm": 0.2522405982017517, "learning_rate": 4.1996646284865816e-05, "loss": 0.0187, "step": 45890 }, { "grad_norm": 0.2224990278482437, "learning_rate": 4.197624509160625e-05, "loss": 0.0172, "step": 45900 }, { "grad_norm": 0.20414209365844727, "learning_rate": 4.195584526938692e-05, "loss": 0.0154, "step": 45910 }, { "grad_norm": 0.22664663195610046, "learning_rate": 4.193544682169365e-05, "loss": 0.0144, "step": 45920 }, { "grad_norm": 0.3683168292045593, "learning_rate": 4.1915049752011946e-05, "loss": 0.014, "step": 45930 }, { "grad_norm": 0.18164224922657013, "learning_rate": 4.189465406382712e-05, "loss": 0.0113, "step": 45940 }, { "grad_norm": 0.1872967779636383, "learning_rate": 4.187425976062422e-05, "loss": 0.0171, "step": 45950 }, { "grad_norm": 0.16183869540691376, "learning_rate": 4.18538668458881e-05, "loss": 0.0153, "step": 45960 }, { "grad_norm": 0.4407925307750702, "learning_rate": 4.183347532310333e-05, "loss": 0.0173, "step": 45970 }, { "grad_norm": 0.29776671528816223, "learning_rate": 4.181308519575429e-05, "loss": 0.0187, "step": 45980 }, { "grad_norm": 0.20358286798000336, "learning_rate": 4.179269646732507e-05, "loss": 0.0221, "step": 45990 }, { "grad_norm": 0.21902164816856384, "learning_rate": 4.177230914129954e-05, "loss": 0.0155, "step": 46000 }, { "grad_norm": 0.21337498724460602, "learning_rate": 4.175192322116136e-05, "loss": 0.0195, "step": 46010 }, { "grad_norm": 0.2829506993293762, "learning_rate": 4.173153871039391e-05, "loss": 0.0145, "step": 46020 }, { "grad_norm": 0.35569053888320923, "learning_rate": 4.171115561248036e-05, "loss": 0.0146, "step": 46030 }, { "grad_norm": 0.35164546966552734, "learning_rate": 4.16907739309036e-05, "loss": 0.0165, "step": 46040 }, { "grad_norm": 0.28447961807250977, "learning_rate": 4.167039366914633e-05, "loss": 0.0145, "step": 46050 }, { "grad_norm": 0.35091203451156616, "learning_rate": 4.165001483069096e-05, "loss": 0.0148, "step": 46060 }, { "grad_norm": 0.19375407695770264, "learning_rate": 4.16296374190197e-05, "loss": 0.022, "step": 46070 }, { "grad_norm": 0.3054529130458832, "learning_rate": 4.1609261437614464e-05, "loss": 0.017, "step": 46080 }, { "grad_norm": 0.43570855259895325, "learning_rate": 4.158888688995696e-05, "loss": 0.0254, "step": 46090 }, { "grad_norm": 0.25832703709602356, "learning_rate": 4.1568513779528645e-05, "loss": 0.0154, "step": 46100 }, { "grad_norm": 0.7928063869476318, "learning_rate": 4.1548142109810704e-05, "loss": 0.024, "step": 46110 }, { "grad_norm": 2.0912632942199707, "learning_rate": 4.152777188428414e-05, "loss": 0.0193, "step": 46120 }, { "grad_norm": 0.3902992308139801, "learning_rate": 4.1507403106429646e-05, "loss": 0.0158, "step": 46130 }, { "grad_norm": 0.7039600610733032, "learning_rate": 4.148703577972768e-05, "loss": 0.0127, "step": 46140 }, { "grad_norm": 0.3022601306438446, "learning_rate": 4.146666990765846e-05, "loss": 0.0141, "step": 46150 }, { "grad_norm": 0.21506381034851074, "learning_rate": 4.1446305493701986e-05, "loss": 0.0154, "step": 46160 }, { "grad_norm": 0.4735294580459595, "learning_rate": 4.142594254133796e-05, "loss": 0.0182, "step": 46170 }, { "grad_norm": 0.2277783453464508, "learning_rate": 4.1405581054045835e-05, "loss": 0.0173, "step": 46180 }, { "grad_norm": 0.47502776980400085, "learning_rate": 4.1385221035304864e-05, "loss": 0.0159, "step": 46190 }, { "grad_norm": 0.23808841407299042, "learning_rate": 4.136486248859398e-05, "loss": 0.0163, "step": 46200 }, { "grad_norm": 0.18980422616004944, "learning_rate": 4.1344505417391955e-05, "loss": 0.0177, "step": 46210 }, { "grad_norm": 0.27198925614356995, "learning_rate": 4.132414982517721e-05, "loss": 0.0174, "step": 46220 }, { "grad_norm": 0.29816684126853943, "learning_rate": 4.130379571542798e-05, "loss": 0.0157, "step": 46230 }, { "grad_norm": 0.6384710073471069, "learning_rate": 4.12834430916222e-05, "loss": 0.0252, "step": 46240 }, { "grad_norm": 0.2686612606048584, "learning_rate": 4.126309195723763e-05, "loss": 0.0174, "step": 46250 }, { "grad_norm": 0.2151067554950714, "learning_rate": 4.124274231575168e-05, "loss": 0.0233, "step": 46260 }, { "grad_norm": 0.23260676860809326, "learning_rate": 4.122239417064154e-05, "loss": 0.0153, "step": 46270 }, { "grad_norm": 0.20061153173446655, "learning_rate": 4.1202047525384184e-05, "loss": 0.013, "step": 46280 }, { "grad_norm": 0.1942981779575348, "learning_rate": 4.118170238345627e-05, "loss": 0.0151, "step": 46290 }, { "grad_norm": 0.17380216717720032, "learning_rate": 4.1161358748334256e-05, "loss": 0.0139, "step": 46300 }, { "grad_norm": 0.22157958149909973, "learning_rate": 4.1141016623494266e-05, "loss": 0.0193, "step": 46310 }, { "grad_norm": 0.3620413541793823, "learning_rate": 4.112067601241227e-05, "loss": 0.0146, "step": 46320 }, { "grad_norm": 0.343697726726532, "learning_rate": 4.110033691856387e-05, "loss": 0.0148, "step": 46330 }, { "grad_norm": 0.2713141441345215, "learning_rate": 4.107999934542451e-05, "loss": 0.0242, "step": 46340 }, { "grad_norm": 0.8397331237792969, "learning_rate": 4.105966329646928e-05, "loss": 0.0155, "step": 46350 }, { "grad_norm": 0.21016442775726318, "learning_rate": 4.103932877517308e-05, "loss": 0.0183, "step": 46360 }, { "grad_norm": 0.20720933377742767, "learning_rate": 4.101899578501052e-05, "loss": 0.0171, "step": 46370 }, { "grad_norm": 0.22757188975811005, "learning_rate": 4.099866432945595e-05, "loss": 0.0182, "step": 46380 }, { "grad_norm": 0.2744070887565613, "learning_rate": 4.097833441198346e-05, "loss": 0.0241, "step": 46390 }, { "grad_norm": 0.22088509798049927, "learning_rate": 4.095800603606685e-05, "loss": 0.0159, "step": 46400 }, { "grad_norm": 0.3554452657699585, "learning_rate": 4.093767920517975e-05, "loss": 0.0217, "step": 46410 }, { "grad_norm": 0.24937665462493896, "learning_rate": 4.091735392279539e-05, "loss": 0.0176, "step": 46420 }, { "grad_norm": 0.14728589355945587, "learning_rate": 4.089703019238685e-05, "loss": 0.013, "step": 46430 }, { "grad_norm": 0.15577173233032227, "learning_rate": 4.0876708017426866e-05, "loss": 0.0139, "step": 46440 }, { "grad_norm": 0.22934794425964355, "learning_rate": 4.085638740138798e-05, "loss": 0.0153, "step": 46450 }, { "grad_norm": 0.26267245411872864, "learning_rate": 4.08360683477424e-05, "loss": 0.0173, "step": 46460 }, { "grad_norm": 0.19360297918319702, "learning_rate": 4.0815750859962085e-05, "loss": 0.0162, "step": 46470 }, { "grad_norm": 0.2818237543106079, "learning_rate": 4.079543494151879e-05, "loss": 0.021, "step": 46480 }, { "grad_norm": 0.2226586490869522, "learning_rate": 4.07751205958839e-05, "loss": 0.0167, "step": 46490 }, { "grad_norm": 0.28938135504722595, "learning_rate": 4.0754807826528615e-05, "loss": 0.0132, "step": 46500 }, { "grad_norm": 0.25881922245025635, "learning_rate": 4.073449663692379e-05, "loss": 0.0215, "step": 46510 }, { "grad_norm": 0.20845714211463928, "learning_rate": 4.07141870305401e-05, "loss": 0.0171, "step": 46520 }, { "grad_norm": 0.257737934589386, "learning_rate": 4.0693879010847866e-05, "loss": 0.0257, "step": 46530 }, { "grad_norm": 0.3034854829311371, "learning_rate": 4.06735725813172e-05, "loss": 0.0195, "step": 46540 }, { "grad_norm": 0.23002557456493378, "learning_rate": 4.0653267745417903e-05, "loss": 0.0152, "step": 46550 }, { "grad_norm": 0.19521643221378326, "learning_rate": 4.063296450661949e-05, "loss": 0.0126, "step": 46560 }, { "grad_norm": 0.24979035556316376, "learning_rate": 4.061266286839128e-05, "loss": 0.0169, "step": 46570 }, { "grad_norm": 0.26011794805526733, "learning_rate": 4.0592362834202225e-05, "loss": 0.0185, "step": 46580 }, { "grad_norm": 0.205164834856987, "learning_rate": 4.057206440752107e-05, "loss": 0.0136, "step": 46590 }, { "grad_norm": 0.2995176911354065, "learning_rate": 4.0551767591816245e-05, "loss": 0.0172, "step": 46600 }, { "grad_norm": 0.22178170084953308, "learning_rate": 4.0531472390555935e-05, "loss": 0.0186, "step": 46610 }, { "grad_norm": 0.891204297542572, "learning_rate": 4.051117880720802e-05, "loss": 0.0166, "step": 46620 }, { "grad_norm": 0.45119890570640564, "learning_rate": 4.049088684524015e-05, "loss": 0.014, "step": 46630 }, { "grad_norm": 0.1921590268611908, "learning_rate": 4.0470596508119636e-05, "loss": 0.019, "step": 46640 }, { "grad_norm": 0.20776453614234924, "learning_rate": 4.0450307799313524e-05, "loss": 0.021, "step": 46650 }, { "grad_norm": 0.261667400598526, "learning_rate": 4.0430020722288656e-05, "loss": 0.0175, "step": 46660 }, { "grad_norm": 0.4442867040634155, "learning_rate": 4.040973528051148e-05, "loss": 0.016, "step": 46670 }, { "grad_norm": 0.21625688672065735, "learning_rate": 4.038945147744827e-05, "loss": 0.0208, "step": 46680 }, { "grad_norm": 0.21167433261871338, "learning_rate": 4.0369169316564945e-05, "loss": 0.0145, "step": 46690 }, { "grad_norm": 0.2141904979944229, "learning_rate": 4.034888880132718e-05, "loss": 0.0139, "step": 46700 }, { "grad_norm": 0.27278783917427063, "learning_rate": 4.032860993520035e-05, "loss": 0.0194, "step": 46710 }, { "grad_norm": 0.30696582794189453, "learning_rate": 4.030833272164959e-05, "loss": 0.0142, "step": 46720 }, { "grad_norm": 0.3333962559700012, "learning_rate": 4.0288057164139705e-05, "loss": 0.018, "step": 46730 }, { "grad_norm": 0.3158529996871948, "learning_rate": 4.0267783266135205e-05, "loss": 0.0176, "step": 46740 }, { "grad_norm": 0.34440451860427856, "learning_rate": 4.024751103110039e-05, "loss": 0.0178, "step": 46750 }, { "grad_norm": 0.28798338770866394, "learning_rate": 4.0227240462499176e-05, "loss": 0.0175, "step": 46760 }, { "grad_norm": 1.5454283952713013, "learning_rate": 4.020697156379531e-05, "loss": 0.0191, "step": 46770 }, { "grad_norm": 0.2609409987926483, "learning_rate": 4.018670433845215e-05, "loss": 0.0193, "step": 46780 }, { "grad_norm": 0.2759361267089844, "learning_rate": 4.016643878993284e-05, "loss": 0.0206, "step": 46790 }, { "grad_norm": 0.30571290850639343, "learning_rate": 4.014617492170017e-05, "loss": 0.0175, "step": 46800 }, { "grad_norm": 0.2820833623409271, "learning_rate": 4.0125912737216726e-05, "loss": 0.0313, "step": 46810 }, { "grad_norm": 0.5711109042167664, "learning_rate": 4.0105652239944735e-05, "loss": 0.0213, "step": 46820 }, { "grad_norm": 0.2394634634256363, "learning_rate": 4.0085393433346144e-05, "loss": 0.0156, "step": 46830 }, { "grad_norm": 0.21480895578861237, "learning_rate": 4.006513632088268e-05, "loss": 0.0155, "step": 46840 }, { "grad_norm": 0.20648761093616486, "learning_rate": 4.004488090601567e-05, "loss": 0.0141, "step": 46850 }, { "grad_norm": 0.24497754871845245, "learning_rate": 4.002462719220626e-05, "loss": 0.0196, "step": 46860 }, { "grad_norm": 0.17962190508842468, "learning_rate": 4.000437518291522e-05, "loss": 0.017, "step": 46870 }, { "grad_norm": 0.2605946362018585, "learning_rate": 3.9984124881603094e-05, "loss": 0.0195, "step": 46880 }, { "grad_norm": 0.2689305543899536, "learning_rate": 3.9963876291730086e-05, "loss": 0.0174, "step": 46890 }, { "grad_norm": 0.23482903838157654, "learning_rate": 3.994362941675614e-05, "loss": 0.0194, "step": 46900 }, { "grad_norm": 0.32027989625930786, "learning_rate": 3.992338426014088e-05, "loss": 0.0208, "step": 46910 }, { "grad_norm": 0.21574531495571136, "learning_rate": 3.9903140825343636e-05, "loss": 0.0176, "step": 46920 }, { "grad_norm": 0.2625197470188141, "learning_rate": 3.98828991158235e-05, "loss": 0.0144, "step": 46930 }, { "grad_norm": 0.6419386267662048, "learning_rate": 3.9862659135039185e-05, "loss": 0.0173, "step": 46940 }, { "grad_norm": 0.2633129954338074, "learning_rate": 3.984242088644918e-05, "loss": 0.0189, "step": 46950 }, { "grad_norm": 0.21042697131633759, "learning_rate": 3.9822184373511615e-05, "loss": 0.0152, "step": 46960 }, { "grad_norm": 0.21257419884204865, "learning_rate": 3.98019495996844e-05, "loss": 0.0167, "step": 46970 }, { "grad_norm": 0.1942121833562851, "learning_rate": 3.978171656842507e-05, "loss": 0.0158, "step": 46980 }, { "grad_norm": 0.3160427510738373, "learning_rate": 3.976148528319091e-05, "loss": 0.0147, "step": 46990 }, { "grad_norm": 0.26574206352233887, "learning_rate": 3.974125574743888e-05, "loss": 0.0188, "step": 47000 }, { "grad_norm": 0.27885833382606506, "learning_rate": 3.9721027964625686e-05, "loss": 0.0177, "step": 47010 }, { "grad_norm": 0.199485182762146, "learning_rate": 3.9700801938207676e-05, "loss": 0.0192, "step": 47020 }, { "grad_norm": 0.2434556931257248, "learning_rate": 3.9680577671640916e-05, "loss": 0.0171, "step": 47030 }, { "grad_norm": 0.21879315376281738, "learning_rate": 3.966035516838121e-05, "loss": 0.0221, "step": 47040 }, { "grad_norm": 0.27414023876190186, "learning_rate": 3.9640134431884014e-05, "loss": 0.0163, "step": 47050 }, { "grad_norm": 0.16399572789669037, "learning_rate": 3.961991546560451e-05, "loss": 0.0136, "step": 47060 }, { "grad_norm": 0.2956876754760742, "learning_rate": 3.959969827299753e-05, "loss": 0.0166, "step": 47070 }, { "grad_norm": 0.3234042525291443, "learning_rate": 3.9579482857517684e-05, "loss": 0.016, "step": 47080 }, { "grad_norm": 1.2500120401382446, "learning_rate": 3.955926922261921e-05, "loss": 0.0174, "step": 47090 }, { "grad_norm": 0.20528994500637054, "learning_rate": 3.9539057371756084e-05, "loss": 0.0149, "step": 47100 }, { "grad_norm": 0.24979256093502045, "learning_rate": 3.951884730838195e-05, "loss": 0.0139, "step": 47110 }, { "grad_norm": 0.18420416116714478, "learning_rate": 3.949863903595012e-05, "loss": 0.0146, "step": 47120 }, { "grad_norm": 0.22922572493553162, "learning_rate": 3.947843255791369e-05, "loss": 0.0158, "step": 47130 }, { "grad_norm": 0.4116259217262268, "learning_rate": 3.9458227877725364e-05, "loss": 0.0189, "step": 47140 }, { "grad_norm": 0.21726392209529877, "learning_rate": 3.943802499883758e-05, "loss": 0.0129, "step": 47150 }, { "grad_norm": 0.25628015398979187, "learning_rate": 3.9417823924702437e-05, "loss": 0.0203, "step": 47160 }, { "grad_norm": 0.2816210389137268, "learning_rate": 3.939762465877178e-05, "loss": 0.0176, "step": 47170 }, { "grad_norm": 0.2685372233390808, "learning_rate": 3.937742720449708e-05, "loss": 0.0159, "step": 47180 }, { "grad_norm": 0.2133750170469284, "learning_rate": 3.9357231565329563e-05, "loss": 0.0116, "step": 47190 }, { "grad_norm": 0.2505757510662079, "learning_rate": 3.933703774472008e-05, "loss": 0.0152, "step": 47200 }, { "grad_norm": 0.2633025050163269, "learning_rate": 3.93168457461192e-05, "loss": 0.0109, "step": 47210 }, { "grad_norm": 0.14091935753822327, "learning_rate": 3.9296655572977216e-05, "loss": 0.0174, "step": 47220 }, { "grad_norm": 0.23759986460208893, "learning_rate": 3.927646722874404e-05, "loss": 0.0207, "step": 47230 }, { "grad_norm": 0.1946820169687271, "learning_rate": 3.925628071686934e-05, "loss": 0.0162, "step": 47240 }, { "grad_norm": 0.3154272139072418, "learning_rate": 3.9236096040802415e-05, "loss": 0.0178, "step": 47250 }, { "grad_norm": 0.29943352937698364, "learning_rate": 3.9215913203992294e-05, "loss": 0.0242, "step": 47260 }, { "grad_norm": 0.22934162616729736, "learning_rate": 3.9195732209887645e-05, "loss": 0.0152, "step": 47270 }, { "grad_norm": 0.2117643505334854, "learning_rate": 3.9175553061936875e-05, "loss": 0.0163, "step": 47280 }, { "grad_norm": 0.17074443399906158, "learning_rate": 3.9155375763588045e-05, "loss": 0.0192, "step": 47290 }, { "grad_norm": 0.2456876039505005, "learning_rate": 3.913520031828889e-05, "loss": 0.0192, "step": 47300 }, { "grad_norm": 0.2502914071083069, "learning_rate": 3.911502672948685e-05, "loss": 0.0183, "step": 47310 }, { "grad_norm": 0.17981210350990295, "learning_rate": 3.9094855000629014e-05, "loss": 0.0151, "step": 47320 }, { "grad_norm": 0.16034932434558868, "learning_rate": 3.907468513516223e-05, "loss": 0.0169, "step": 47330 }, { "grad_norm": 0.22955337166786194, "learning_rate": 3.905451713653294e-05, "loss": 0.0134, "step": 47340 }, { "grad_norm": 0.1768738478422165, "learning_rate": 3.903435100818731e-05, "loss": 0.0147, "step": 47350 }, { "grad_norm": 0.18795524537563324, "learning_rate": 3.901418675357117e-05, "loss": 0.0143, "step": 47360 }, { "grad_norm": 0.19684536755084991, "learning_rate": 3.8994024376130075e-05, "loss": 0.024, "step": 47370 }, { "grad_norm": 0.2543429136276245, "learning_rate": 3.8973863879309194e-05, "loss": 0.0173, "step": 47380 }, { "grad_norm": 0.30803388357162476, "learning_rate": 3.8953705266553394e-05, "loss": 0.0142, "step": 47390 }, { "grad_norm": 0.7492367625236511, "learning_rate": 3.893354854130727e-05, "loss": 0.0218, "step": 47400 }, { "grad_norm": 0.20995226502418518, "learning_rate": 3.8913393707015006e-05, "loss": 0.0177, "step": 47410 }, { "grad_norm": 0.25178420543670654, "learning_rate": 3.889324076712056e-05, "loss": 0.0241, "step": 47420 }, { "grad_norm": 0.2258310168981552, "learning_rate": 3.8873089725067476e-05, "loss": 0.0182, "step": 47430 }, { "grad_norm": 0.21931080520153046, "learning_rate": 3.885294058429905e-05, "loss": 0.0205, "step": 47440 }, { "grad_norm": 0.3223656117916107, "learning_rate": 3.8832793348258206e-05, "loss": 0.0161, "step": 47450 }, { "grad_norm": 1.2008092403411865, "learning_rate": 3.881264802038756e-05, "loss": 0.0256, "step": 47460 }, { "grad_norm": 0.3421173393726349, "learning_rate": 3.879250460412939e-05, "loss": 0.0184, "step": 47470 }, { "grad_norm": 0.2555123567581177, "learning_rate": 3.8772363102925644e-05, "loss": 0.0153, "step": 47480 }, { "grad_norm": 0.2124171257019043, "learning_rate": 3.875222352021798e-05, "loss": 0.0145, "step": 47490 }, { "grad_norm": 0.16978819668293, "learning_rate": 3.8732085859447686e-05, "loss": 0.0194, "step": 47500 }, { "grad_norm": 0.2572355270385742, "learning_rate": 3.871195012405575e-05, "loss": 0.0143, "step": 47510 }, { "grad_norm": 0.23802576959133148, "learning_rate": 3.869181631748278e-05, "loss": 0.0177, "step": 47520 }, { "grad_norm": 0.24421440064907074, "learning_rate": 3.867168444316915e-05, "loss": 0.0174, "step": 47530 }, { "grad_norm": 0.21756629645824432, "learning_rate": 3.865155450455481e-05, "loss": 0.0135, "step": 47540 }, { "grad_norm": 0.24490520358085632, "learning_rate": 3.8631426505079426e-05, "loss": 0.02, "step": 47550 }, { "grad_norm": 0.2997647523880005, "learning_rate": 3.8611300448182304e-05, "loss": 0.0221, "step": 47560 }, { "grad_norm": 0.1990589052438736, "learning_rate": 3.859117633730248e-05, "loss": 0.0137, "step": 47570 }, { "grad_norm": 0.32744425535202026, "learning_rate": 3.857105417587858e-05, "loss": 0.0199, "step": 47580 }, { "grad_norm": 0.2805992662906647, "learning_rate": 3.855093396734894e-05, "loss": 0.0163, "step": 47590 }, { "grad_norm": 0.19883230328559875, "learning_rate": 3.8530815715151545e-05, "loss": 0.0239, "step": 47600 }, { "grad_norm": 0.20009927451610565, "learning_rate": 3.851069942272405e-05, "loss": 0.0137, "step": 47610 }, { "grad_norm": 0.30491745471954346, "learning_rate": 3.849058509350382e-05, "loss": 0.0184, "step": 47620 }, { "grad_norm": 0.2565860152244568, "learning_rate": 3.8470472730927783e-05, "loss": 0.0221, "step": 47630 }, { "grad_norm": 0.17878925800323486, "learning_rate": 3.845036233843264e-05, "loss": 0.0174, "step": 47640 }, { "grad_norm": 0.22390054166316986, "learning_rate": 3.843025391945469e-05, "loss": 0.0159, "step": 47650 }, { "grad_norm": 0.13464303314685822, "learning_rate": 3.841014747742992e-05, "loss": 0.013, "step": 47660 }, { "grad_norm": 0.5277400612831116, "learning_rate": 3.839004301579397e-05, "loss": 0.0139, "step": 47670 }, { "grad_norm": 0.2116738110780716, "learning_rate": 3.8369940537982097e-05, "loss": 0.0172, "step": 47680 }, { "grad_norm": 0.3648624122142792, "learning_rate": 3.834984004742933e-05, "loss": 0.0129, "step": 47690 }, { "grad_norm": 0.1883350908756256, "learning_rate": 3.832974154757026e-05, "loss": 0.0157, "step": 47700 }, { "grad_norm": 0.2625851631164551, "learning_rate": 3.830964504183919e-05, "loss": 0.0143, "step": 47710 }, { "grad_norm": 0.22879894077777863, "learning_rate": 3.828955053367003e-05, "loss": 0.0166, "step": 47720 }, { "grad_norm": 0.23821888864040375, "learning_rate": 3.826945802649642e-05, "loss": 0.0179, "step": 47730 }, { "grad_norm": 0.1917216032743454, "learning_rate": 3.824936752375159e-05, "loss": 0.0212, "step": 47740 }, { "grad_norm": 0.2129204422235489, "learning_rate": 3.822927902886848e-05, "loss": 0.016, "step": 47750 }, { "grad_norm": 1.0833275318145752, "learning_rate": 3.8209192545279653e-05, "loss": 0.0172, "step": 47760 }, { "grad_norm": 0.2013714462518692, "learning_rate": 3.8189108076417326e-05, "loss": 0.0139, "step": 47770 }, { "grad_norm": 0.38750535249710083, "learning_rate": 3.816902562571342e-05, "loss": 0.0244, "step": 47780 }, { "grad_norm": 0.670425295829773, "learning_rate": 3.814894519659944e-05, "loss": 0.0169, "step": 47790 }, { "grad_norm": 0.21405325829982758, "learning_rate": 3.812886679250661e-05, "loss": 0.0144, "step": 47800 }, { "grad_norm": 1.821425437927246, "learning_rate": 3.810879041686575e-05, "loss": 0.0164, "step": 47810 }, { "grad_norm": 0.3119468092918396, "learning_rate": 3.808871607310741e-05, "loss": 0.0224, "step": 47820 }, { "grad_norm": 0.3135085999965668, "learning_rate": 3.806864376466169e-05, "loss": 0.0234, "step": 47830 }, { "grad_norm": 0.33549773693084717, "learning_rate": 3.804857349495845e-05, "loss": 0.0143, "step": 47840 }, { "grad_norm": 0.328898161649704, "learning_rate": 3.802850526742713e-05, "loss": 0.0297, "step": 47850 }, { "grad_norm": 0.22577591240406036, "learning_rate": 3.800843908549683e-05, "loss": 0.0157, "step": 47860 }, { "grad_norm": 0.3027772605419159, "learning_rate": 3.7988374952596325e-05, "loss": 0.0154, "step": 47870 }, { "grad_norm": 0.2250015139579773, "learning_rate": 3.7968312872154e-05, "loss": 0.0164, "step": 47880 }, { "grad_norm": 0.3132321536540985, "learning_rate": 3.7948252847597965e-05, "loss": 0.0194, "step": 47890 }, { "grad_norm": 0.2035585641860962, "learning_rate": 3.7928194882355885e-05, "loss": 0.0182, "step": 47900 }, { "grad_norm": 0.17059487104415894, "learning_rate": 3.790813897985515e-05, "loss": 0.0171, "step": 47910 }, { "grad_norm": 0.22402431070804596, "learning_rate": 3.7888085143522726e-05, "loss": 0.0166, "step": 47920 }, { "grad_norm": 0.31177037954330444, "learning_rate": 3.7868033376785314e-05, "loss": 0.0197, "step": 47930 }, { "grad_norm": 0.22615757584571838, "learning_rate": 3.784798368306919e-05, "loss": 0.0165, "step": 47940 }, { "grad_norm": 0.2833106815814972, "learning_rate": 3.782793606580029e-05, "loss": 0.0123, "step": 47950 }, { "grad_norm": 0.34970539808273315, "learning_rate": 3.7807890528404205e-05, "loss": 0.0176, "step": 47960 }, { "grad_norm": 0.14428722858428955, "learning_rate": 3.778784707430616e-05, "loss": 0.0167, "step": 47970 }, { "grad_norm": 0.217751145362854, "learning_rate": 3.776780570693107e-05, "loss": 0.0151, "step": 47980 }, { "grad_norm": 0.4396800696849823, "learning_rate": 3.774776642970342e-05, "loss": 0.0185, "step": 47990 }, { "grad_norm": 0.2663453221321106, "learning_rate": 3.77277292460474e-05, "loss": 0.013, "step": 48000 }, { "grad_norm": 0.21960346400737762, "learning_rate": 3.770769415938678e-05, "loss": 0.0173, "step": 48010 }, { "grad_norm": 0.1737799495458603, "learning_rate": 3.768766117314506e-05, "loss": 0.0207, "step": 48020 }, { "grad_norm": 0.2538631558418274, "learning_rate": 3.76676302907453e-05, "loss": 0.0152, "step": 48030 }, { "grad_norm": 0.22057731449604034, "learning_rate": 3.764760151561021e-05, "loss": 0.0143, "step": 48040 }, { "grad_norm": 0.23030178248882294, "learning_rate": 3.76275748511622e-05, "loss": 0.0164, "step": 48050 }, { "grad_norm": 0.2841729521751404, "learning_rate": 3.7607550300823255e-05, "loss": 0.0156, "step": 48060 }, { "grad_norm": 0.26821935176849365, "learning_rate": 3.7587527868015044e-05, "loss": 0.0266, "step": 48070 }, { "grad_norm": 0.4152112603187561, "learning_rate": 3.756750755615881e-05, "loss": 0.016, "step": 48080 }, { "grad_norm": 0.13796071708202362, "learning_rate": 3.754748936867553e-05, "loss": 0.0166, "step": 48090 }, { "grad_norm": 0.24755936861038208, "learning_rate": 3.752747330898573e-05, "loss": 0.0146, "step": 48100 }, { "grad_norm": 0.19905130565166473, "learning_rate": 3.750745938050962e-05, "loss": 0.0149, "step": 48110 }, { "grad_norm": 0.14817164838314056, "learning_rate": 3.7487447586667025e-05, "loss": 0.0161, "step": 48120 }, { "grad_norm": 0.28030747175216675, "learning_rate": 3.7467437930877426e-05, "loss": 0.0129, "step": 48130 }, { "grad_norm": 0.1599465012550354, "learning_rate": 3.744743041655992e-05, "loss": 0.0139, "step": 48140 }, { "grad_norm": 0.2159184068441391, "learning_rate": 3.742742504713324e-05, "loss": 0.0164, "step": 48150 }, { "grad_norm": 0.268259733915329, "learning_rate": 3.740742182601576e-05, "loss": 0.0165, "step": 48160 }, { "grad_norm": 0.17433643341064453, "learning_rate": 3.7387420756625464e-05, "loss": 0.0179, "step": 48170 }, { "grad_norm": 0.2352031171321869, "learning_rate": 3.736742184238002e-05, "loss": 0.012, "step": 48180 }, { "grad_norm": 0.18425945937633514, "learning_rate": 3.7347425086696684e-05, "loss": 0.015, "step": 48190 }, { "grad_norm": 0.2875756025314331, "learning_rate": 3.732743049299235e-05, "loss": 0.0165, "step": 48200 }, { "grad_norm": 0.34343069791793823, "learning_rate": 3.730743806468354e-05, "loss": 0.0148, "step": 48210 }, { "grad_norm": 0.199512779712677, "learning_rate": 3.7287447805186436e-05, "loss": 0.0126, "step": 48220 }, { "grad_norm": 0.19143377244472504, "learning_rate": 3.726745971791682e-05, "loss": 0.017, "step": 48230 }, { "grad_norm": 0.26613450050354004, "learning_rate": 3.724747380629008e-05, "loss": 0.0234, "step": 48240 }, { "grad_norm": 0.22971302270889282, "learning_rate": 3.72274900737213e-05, "loss": 0.0174, "step": 48250 }, { "grad_norm": 0.35984665155410767, "learning_rate": 3.7207508523625123e-05, "loss": 0.0178, "step": 48260 }, { "grad_norm": 0.23786661028862, "learning_rate": 3.718752915941588e-05, "loss": 0.0144, "step": 48270 }, { "grad_norm": 0.3134400248527527, "learning_rate": 3.7167551984507464e-05, "loss": 0.0167, "step": 48280 }, { "grad_norm": 0.1848989874124527, "learning_rate": 3.714757700231346e-05, "loss": 0.0141, "step": 48290 }, { "grad_norm": 0.22726207971572876, "learning_rate": 3.712760421624703e-05, "loss": 0.0142, "step": 48300 }, { "grad_norm": 0.24440234899520874, "learning_rate": 3.710763362972099e-05, "loss": 0.0155, "step": 48310 }, { "grad_norm": 0.20401784777641296, "learning_rate": 3.708766524614774e-05, "loss": 0.0131, "step": 48320 }, { "grad_norm": 0.1425066888332367, "learning_rate": 3.7067699068939335e-05, "loss": 0.0123, "step": 48330 }, { "grad_norm": 0.23185400664806366, "learning_rate": 3.704773510150748e-05, "loss": 0.0131, "step": 48340 }, { "grad_norm": 0.29063987731933594, "learning_rate": 3.702777334726344e-05, "loss": 0.0201, "step": 48350 }, { "grad_norm": 0.24692225456237793, "learning_rate": 3.7007813809618164e-05, "loss": 0.0141, "step": 48360 }, { "grad_norm": 0.25641122460365295, "learning_rate": 3.6987856491982145e-05, "loss": 0.019, "step": 48370 }, { "grad_norm": 0.36419063806533813, "learning_rate": 3.696790139776558e-05, "loss": 0.018, "step": 48380 }, { "grad_norm": 0.3225443959236145, "learning_rate": 3.6947948530378235e-05, "loss": 0.0181, "step": 48390 }, { "grad_norm": 0.1634906679391861, "learning_rate": 3.6927997893229516e-05, "loss": 0.0141, "step": 48400 }, { "grad_norm": 0.34617123007774353, "learning_rate": 3.690804948972845e-05, "loss": 0.0164, "step": 48410 }, { "grad_norm": 0.19322602450847626, "learning_rate": 3.688810332328363e-05, "loss": 0.0147, "step": 48420 }, { "grad_norm": 0.6564612984657288, "learning_rate": 3.686815939730336e-05, "loss": 0.0189, "step": 48430 }, { "grad_norm": 0.26476722955703735, "learning_rate": 3.684821771519548e-05, "loss": 0.0169, "step": 48440 }, { "grad_norm": 0.20041556656360626, "learning_rate": 3.68282782803675e-05, "loss": 0.0141, "step": 48450 }, { "grad_norm": 0.1986832320690155, "learning_rate": 3.6808341096226504e-05, "loss": 0.0208, "step": 48460 }, { "grad_norm": 0.3384213149547577, "learning_rate": 3.678840616617924e-05, "loss": 0.0152, "step": 48470 }, { "grad_norm": 0.20853760838508606, "learning_rate": 3.6768473493632e-05, "loss": 0.015, "step": 48480 }, { "grad_norm": 0.2062930315732956, "learning_rate": 3.6748543081990783e-05, "loss": 0.0165, "step": 48490 }, { "grad_norm": 0.21663689613342285, "learning_rate": 3.672861493466112e-05, "loss": 0.0173, "step": 48500 }, { "grad_norm": 0.2902206480503082, "learning_rate": 3.670868905504818e-05, "loss": 0.0178, "step": 48510 }, { "grad_norm": 0.24239586293697357, "learning_rate": 3.6688765446556784e-05, "loss": 0.0153, "step": 48520 }, { "grad_norm": 0.273469477891922, "learning_rate": 3.6668844112591276e-05, "loss": 0.0174, "step": 48530 }, { "grad_norm": 0.17206978797912598, "learning_rate": 3.664892505655573e-05, "loss": 0.0104, "step": 48540 }, { "grad_norm": 0.19642411172389984, "learning_rate": 3.662900828185373e-05, "loss": 0.0143, "step": 48550 }, { "grad_norm": 0.16579091548919678, "learning_rate": 3.6609093791888516e-05, "loss": 0.0121, "step": 48560 }, { "grad_norm": 0.26122376322746277, "learning_rate": 3.658918159006292e-05, "loss": 0.018, "step": 48570 }, { "grad_norm": 0.23766948282718658, "learning_rate": 3.656927167977942e-05, "loss": 0.0164, "step": 48580 }, { "grad_norm": 0.15455262362957, "learning_rate": 3.654936406444006e-05, "loss": 0.0142, "step": 48590 }, { "grad_norm": 0.20543314516544342, "learning_rate": 3.65294587474465e-05, "loss": 0.0161, "step": 48600 }, { "grad_norm": 0.2348306030035019, "learning_rate": 3.650955573220002e-05, "loss": 0.0418, "step": 48610 }, { "grad_norm": 0.2183796912431717, "learning_rate": 3.648965502210149e-05, "loss": 0.0153, "step": 48620 }, { "grad_norm": 0.29407235980033875, "learning_rate": 3.646975662055142e-05, "loss": 0.0138, "step": 48630 }, { "grad_norm": 0.7688180804252625, "learning_rate": 3.644986053094987e-05, "loss": 0.0163, "step": 48640 }, { "grad_norm": 0.21364448964595795, "learning_rate": 3.642996675669659e-05, "loss": 0.0179, "step": 48650 }, { "grad_norm": 0.21227723360061646, "learning_rate": 3.641007530119083e-05, "loss": 0.0148, "step": 48660 }, { "grad_norm": 0.2524050176143646, "learning_rate": 3.639018616783153e-05, "loss": 0.0187, "step": 48670 }, { "grad_norm": 0.2258373647928238, "learning_rate": 3.637029936001719e-05, "loss": 0.0128, "step": 48680 }, { "grad_norm": 0.2362746000289917, "learning_rate": 3.6350414881145886e-05, "loss": 0.0182, "step": 48690 }, { "grad_norm": 0.1934850960969925, "learning_rate": 3.6330532734615386e-05, "loss": 0.0147, "step": 48700 }, { "grad_norm": 0.18110327422618866, "learning_rate": 3.6310652923822975e-05, "loss": 0.0186, "step": 48710 }, { "grad_norm": 0.41050150990486145, "learning_rate": 3.629077545216558e-05, "loss": 0.0136, "step": 48720 }, { "grad_norm": 0.26504048705101013, "learning_rate": 3.627090032303969e-05, "loss": 0.0185, "step": 48730 }, { "grad_norm": 0.2934916615486145, "learning_rate": 3.625102753984146e-05, "loss": 0.0207, "step": 48740 }, { "grad_norm": 0.49465617537498474, "learning_rate": 3.623115710596659e-05, "loss": 0.014, "step": 48750 }, { "grad_norm": 0.26782354712486267, "learning_rate": 3.6211289024810395e-05, "loss": 0.0156, "step": 48760 }, { "grad_norm": 0.5588270425796509, "learning_rate": 3.619142329976777e-05, "loss": 0.0131, "step": 48770 }, { "grad_norm": 0.21802179515361786, "learning_rate": 3.6171559934233247e-05, "loss": 0.0134, "step": 48780 }, { "grad_norm": 0.3477492034435272, "learning_rate": 3.615169893160093e-05, "loss": 0.0158, "step": 48790 }, { "grad_norm": 0.27549445629119873, "learning_rate": 3.61318402952645e-05, "loss": 0.0132, "step": 48800 }, { "grad_norm": 0.23024778068065643, "learning_rate": 3.6111984028617285e-05, "loss": 0.0145, "step": 48810 }, { "grad_norm": 0.19350874423980713, "learning_rate": 3.6092130135052134e-05, "loss": 0.0155, "step": 48820 }, { "grad_norm": 0.41354554891586304, "learning_rate": 3.6072278617961584e-05, "loss": 0.0177, "step": 48830 }, { "grad_norm": 0.2471519410610199, "learning_rate": 3.605242948073767e-05, "loss": 0.0178, "step": 48840 }, { "grad_norm": 0.22610269486904144, "learning_rate": 3.603258272677212e-05, "loss": 0.0163, "step": 48850 }, { "grad_norm": 0.2576081156730652, "learning_rate": 3.601273835945616e-05, "loss": 0.0181, "step": 48860 }, { "grad_norm": 0.44014763832092285, "learning_rate": 3.5992896382180664e-05, "loss": 0.0158, "step": 48870 }, { "grad_norm": 0.1432044506072998, "learning_rate": 3.597305679833609e-05, "loss": 0.0201, "step": 48880 }, { "grad_norm": 0.2374284565448761, "learning_rate": 3.595321961131245e-05, "loss": 0.0178, "step": 48890 }, { "grad_norm": 0.24663689732551575, "learning_rate": 3.593338482449942e-05, "loss": 0.0197, "step": 48900 }, { "grad_norm": 0.37615343928337097, "learning_rate": 3.591355244128618e-05, "loss": 0.0237, "step": 48910 }, { "grad_norm": 0.25726521015167236, "learning_rate": 3.589372246506158e-05, "loss": 0.0211, "step": 48920 }, { "grad_norm": 0.26278021931648254, "learning_rate": 3.5873894899213984e-05, "loss": 0.0124, "step": 48930 }, { "grad_norm": 0.4201195538043976, "learning_rate": 3.5854069747131416e-05, "loss": 0.014, "step": 48940 }, { "grad_norm": 0.22546496987342834, "learning_rate": 3.583424701220143e-05, "loss": 0.0136, "step": 48950 }, { "grad_norm": 0.1877472698688507, "learning_rate": 3.581442669781121e-05, "loss": 0.0157, "step": 48960 }, { "grad_norm": 0.20060515403747559, "learning_rate": 3.579460880734749e-05, "loss": 0.0166, "step": 48970 }, { "grad_norm": 0.2386404573917389, "learning_rate": 3.577479334419657e-05, "loss": 0.015, "step": 48980 }, { "grad_norm": 0.2540108859539032, "learning_rate": 3.575498031174444e-05, "loss": 0.0141, "step": 48990 }, { "grad_norm": 0.27599552273750305, "learning_rate": 3.573516971337657e-05, "loss": 0.0205, "step": 49000 }, { "grad_norm": 0.21662501990795135, "learning_rate": 3.5715361552478046e-05, "loss": 0.0179, "step": 49010 }, { "grad_norm": 0.21503160893917084, "learning_rate": 3.5695555832433536e-05, "loss": 0.014, "step": 49020 }, { "grad_norm": 0.3004564046859741, "learning_rate": 3.5675752556627325e-05, "loss": 0.017, "step": 49030 }, { "grad_norm": 0.2584361135959625, "learning_rate": 3.565595172844322e-05, "loss": 0.0149, "step": 49040 }, { "grad_norm": 0.23018783330917358, "learning_rate": 3.5636153351264666e-05, "loss": 0.0144, "step": 49050 }, { "grad_norm": 0.2549397349357605, "learning_rate": 3.5616357428474655e-05, "loss": 0.0137, "step": 49060 }, { "grad_norm": 0.20276431739330292, "learning_rate": 3.559656396345575e-05, "loss": 0.023, "step": 49070 }, { "grad_norm": 3.5049099922180176, "learning_rate": 3.5576772959590146e-05, "loss": 0.0189, "step": 49080 }, { "grad_norm": 0.6124345064163208, "learning_rate": 3.5556984420259545e-05, "loss": 0.0147, "step": 49090 }, { "grad_norm": 0.18169313669204712, "learning_rate": 3.5537198348845305e-05, "loss": 0.0134, "step": 49100 }, { "grad_norm": 0.18796339631080627, "learning_rate": 3.551741474872831e-05, "loss": 0.0157, "step": 49110 }, { "grad_norm": 0.22567039728164673, "learning_rate": 3.549763362328903e-05, "loss": 0.0174, "step": 49120 }, { "grad_norm": 0.179195836186409, "learning_rate": 3.5477854975907515e-05, "loss": 0.0143, "step": 49130 }, { "grad_norm": 0.2065667361021042, "learning_rate": 3.5458078809963416e-05, "loss": 0.0136, "step": 49140 }, { "grad_norm": 0.2150622457265854, "learning_rate": 3.543830512883594e-05, "loss": 0.0144, "step": 49150 }, { "grad_norm": 0.3214762806892395, "learning_rate": 3.5418533935903824e-05, "loss": 0.0129, "step": 49160 }, { "grad_norm": 0.24053725600242615, "learning_rate": 3.539876523454547e-05, "loss": 0.0196, "step": 49170 }, { "grad_norm": 0.20982132852077484, "learning_rate": 3.537899902813878e-05, "loss": 0.0128, "step": 49180 }, { "grad_norm": 0.24084585905075073, "learning_rate": 3.5359235320061293e-05, "loss": 0.0146, "step": 49190 }, { "grad_norm": 0.12647701799869537, "learning_rate": 3.533947411369003e-05, "loss": 0.0126, "step": 49200 }, { "grad_norm": 0.1892482489347458, "learning_rate": 3.53197154124017e-05, "loss": 0.0142, "step": 49210 }, { "grad_norm": 0.24049322307109833, "learning_rate": 3.52999592195725e-05, "loss": 0.0169, "step": 49220 }, { "grad_norm": 0.1934511512517929, "learning_rate": 3.5280205538578224e-05, "loss": 0.0173, "step": 49230 }, { "grad_norm": 0.26533451676368713, "learning_rate": 3.5260454372794236e-05, "loss": 0.0146, "step": 49240 }, { "grad_norm": 0.2875840663909912, "learning_rate": 3.524070572559545e-05, "loss": 0.0162, "step": 49250 }, { "grad_norm": 0.248356893658638, "learning_rate": 3.5220959600356395e-05, "loss": 0.0145, "step": 49260 }, { "grad_norm": 0.19122587144374847, "learning_rate": 3.5201216000451145e-05, "loss": 0.0135, "step": 49270 }, { "grad_norm": 0.24536648392677307, "learning_rate": 3.5181474929253335e-05, "loss": 0.0166, "step": 49280 }, { "grad_norm": 0.1858508437871933, "learning_rate": 3.516173639013615e-05, "loss": 0.0127, "step": 49290 }, { "grad_norm": 0.2405201494693756, "learning_rate": 3.5142000386472406e-05, "loss": 0.0109, "step": 49300 }, { "grad_norm": 0.512056291103363, "learning_rate": 3.5122266921634427e-05, "loss": 0.0167, "step": 49310 }, { "grad_norm": 0.2020161747932434, "learning_rate": 3.510253599899413e-05, "loss": 0.0121, "step": 49320 }, { "grad_norm": 0.23865152895450592, "learning_rate": 3.5082807621922965e-05, "loss": 0.0139, "step": 49330 }, { "grad_norm": 0.28657591342926025, "learning_rate": 3.506308179379201e-05, "loss": 0.0156, "step": 49340 }, { "grad_norm": 0.18190239369869232, "learning_rate": 3.5043358517971844e-05, "loss": 0.0149, "step": 49350 }, { "grad_norm": 0.17331407964229584, "learning_rate": 3.502363779783264e-05, "loss": 0.0124, "step": 49360 }, { "grad_norm": 0.39552998542785645, "learning_rate": 3.500391963674415e-05, "loss": 0.0139, "step": 49370 }, { "grad_norm": 0.23577792942523956, "learning_rate": 3.4984204038075615e-05, "loss": 0.0239, "step": 49380 }, { "grad_norm": 0.26087725162506104, "learning_rate": 3.496449100519595e-05, "loss": 0.0208, "step": 49390 }, { "grad_norm": 0.26862791180610657, "learning_rate": 3.494478054147354e-05, "loss": 0.0144, "step": 49400 }, { "grad_norm": 0.22186902165412903, "learning_rate": 3.4925072650276395e-05, "loss": 0.0137, "step": 49410 }, { "grad_norm": 0.4484730064868927, "learning_rate": 3.4905367334972016e-05, "loss": 0.014, "step": 49420 }, { "grad_norm": 0.28334495425224304, "learning_rate": 3.488566459892752e-05, "loss": 0.0132, "step": 49430 }, { "grad_norm": 0.2638440728187561, "learning_rate": 3.4865964445509585e-05, "loss": 0.0186, "step": 49440 }, { "grad_norm": 0.25097620487213135, "learning_rate": 3.484626687808438e-05, "loss": 0.0177, "step": 49450 }, { "grad_norm": 0.2217305600643158, "learning_rate": 3.4826571900017735e-05, "loss": 0.0155, "step": 49460 }, { "grad_norm": 0.19734404981136322, "learning_rate": 3.480687951467495e-05, "loss": 0.0131, "step": 49470 }, { "grad_norm": 0.31922462582588196, "learning_rate": 3.4787189725420925e-05, "loss": 0.013, "step": 49480 }, { "grad_norm": 1.7205557823181152, "learning_rate": 3.4767502535620086e-05, "loss": 0.0174, "step": 49490 }, { "grad_norm": 0.25926417112350464, "learning_rate": 3.474781794863648e-05, "loss": 0.0183, "step": 49500 }, { "grad_norm": 0.2411523163318634, "learning_rate": 3.472813596783363e-05, "loss": 0.0155, "step": 49510 }, { "grad_norm": 0.747918426990509, "learning_rate": 3.470845659657466e-05, "loss": 0.0187, "step": 49520 }, { "grad_norm": 0.2676992416381836, "learning_rate": 3.468877983822223e-05, "loss": 0.0154, "step": 49530 }, { "grad_norm": 0.14226430654525757, "learning_rate": 3.466910569613855e-05, "loss": 0.0122, "step": 49540 }, { "grad_norm": 0.3709304928779602, "learning_rate": 3.464943417368542e-05, "loss": 0.0167, "step": 49550 }, { "grad_norm": 0.2381971925497055, "learning_rate": 3.462976527422415e-05, "loss": 0.013, "step": 49560 }, { "grad_norm": 0.1662101447582245, "learning_rate": 3.461009900111562e-05, "loss": 0.0164, "step": 49570 }, { "grad_norm": 0.27685457468032837, "learning_rate": 3.459043535772023e-05, "loss": 0.0171, "step": 49580 }, { "grad_norm": 0.3827061355113983, "learning_rate": 3.4570774347398014e-05, "loss": 0.0193, "step": 49590 }, { "grad_norm": 0.1972106546163559, "learning_rate": 3.4551115973508454e-05, "loss": 0.0156, "step": 49600 }, { "grad_norm": 0.2923627495765686, "learning_rate": 3.453146023941066e-05, "loss": 0.0203, "step": 49610 }, { "grad_norm": 0.1941230446100235, "learning_rate": 3.451180714846325e-05, "loss": 0.0164, "step": 49620 }, { "grad_norm": 0.28299015760421753, "learning_rate": 3.449215670402438e-05, "loss": 0.0145, "step": 49630 }, { "grad_norm": 0.29804936051368713, "learning_rate": 3.447250890945181e-05, "loss": 0.0164, "step": 49640 }, { "grad_norm": 0.2510727345943451, "learning_rate": 3.4452863768102754e-05, "loss": 0.0164, "step": 49650 }, { "grad_norm": 0.23797322809696198, "learning_rate": 3.443322128333409e-05, "loss": 0.0165, "step": 49660 }, { "grad_norm": 0.29649975895881653, "learning_rate": 3.441358145850215e-05, "loss": 0.017, "step": 49670 }, { "grad_norm": 0.21467193961143494, "learning_rate": 3.439394429696286e-05, "loss": 0.0159, "step": 49680 }, { "grad_norm": 0.17067773640155792, "learning_rate": 3.4374309802071644e-05, "loss": 0.0173, "step": 49690 }, { "grad_norm": 0.2995770275592804, "learning_rate": 3.435467797718353e-05, "loss": 0.0156, "step": 49700 }, { "grad_norm": 0.696832537651062, "learning_rate": 3.433504882565306e-05, "loss": 0.0176, "step": 49710 }, { "grad_norm": 0.21132570505142212, "learning_rate": 3.43154223508343e-05, "loss": 0.019, "step": 49720 }, { "grad_norm": 1.4134190082550049, "learning_rate": 3.429579855608089e-05, "loss": 0.0207, "step": 49730 }, { "grad_norm": 0.7152271866798401, "learning_rate": 3.427617744474597e-05, "loss": 0.018, "step": 49740 }, { "grad_norm": 0.2888839542865753, "learning_rate": 3.425655902018231e-05, "loss": 0.019, "step": 49750 }, { "grad_norm": 0.23906487226486206, "learning_rate": 3.423694328574211e-05, "loss": 0.0136, "step": 49760 }, { "grad_norm": 0.24382758140563965, "learning_rate": 3.42173302447772e-05, "loss": 0.0122, "step": 49770 }, { "grad_norm": 0.32557445764541626, "learning_rate": 3.419771990063886e-05, "loss": 0.0138, "step": 49780 }, { "grad_norm": 0.21771681308746338, "learning_rate": 3.417811225667803e-05, "loss": 0.0117, "step": 49790 }, { "grad_norm": 0.22617341578006744, "learning_rate": 3.415850731624508e-05, "loss": 0.0161, "step": 49800 }, { "grad_norm": 0.19111306965351105, "learning_rate": 3.4138905082689945e-05, "loss": 0.0153, "step": 49810 }, { "grad_norm": 0.23961885273456573, "learning_rate": 3.4119305559362145e-05, "loss": 0.0165, "step": 49820 }, { "grad_norm": 0.18210962414741516, "learning_rate": 3.4099708749610684e-05, "loss": 0.0186, "step": 49830 }, { "grad_norm": 0.23804782330989838, "learning_rate": 3.408011465678413e-05, "loss": 0.0191, "step": 49840 }, { "grad_norm": 0.2668055295944214, "learning_rate": 3.406052328423055e-05, "loss": 0.0203, "step": 49850 }, { "grad_norm": 0.2670437693595886, "learning_rate": 3.4040934635297615e-05, "loss": 0.0144, "step": 49860 }, { "grad_norm": 0.19383659958839417, "learning_rate": 3.4021348713332466e-05, "loss": 0.0154, "step": 49870 }, { "grad_norm": 0.3232167065143585, "learning_rate": 3.4001765521681807e-05, "loss": 0.017, "step": 49880 }, { "grad_norm": 0.22125759720802307, "learning_rate": 3.398218506369188e-05, "loss": 0.0263, "step": 49890 }, { "grad_norm": 0.3512011468410492, "learning_rate": 3.3962607342708404e-05, "loss": 0.016, "step": 49900 }, { "grad_norm": 0.2343003898859024, "learning_rate": 3.394303236207673e-05, "loss": 0.0161, "step": 49910 }, { "grad_norm": 0.24385014176368713, "learning_rate": 3.392346012514166e-05, "loss": 0.0152, "step": 49920 }, { "grad_norm": 0.21084631979465485, "learning_rate": 3.390389063524757e-05, "loss": 0.0117, "step": 49930 }, { "grad_norm": 7.8250732421875, "learning_rate": 3.3884323895738324e-05, "loss": 0.0149, "step": 49940 }, { "grad_norm": 0.32726559042930603, "learning_rate": 3.386475990995738e-05, "loss": 0.0198, "step": 49950 }, { "grad_norm": 0.19186201691627502, "learning_rate": 3.384519868124765e-05, "loss": 0.0144, "step": 49960 }, { "grad_norm": 0.18143774569034576, "learning_rate": 3.3825640212951645e-05, "loss": 0.0155, "step": 49970 }, { "grad_norm": 0.22317351400852203, "learning_rate": 3.380608450841134e-05, "loss": 0.0118, "step": 49980 }, { "grad_norm": 0.23857516050338745, "learning_rate": 3.3786531570968305e-05, "loss": 0.0152, "step": 49990 }, { "grad_norm": 3.7512400150299072, "learning_rate": 3.3766981403963584e-05, "loss": 0.0219, "step": 50000 }, { "grad_norm": 0.2542170584201813, "learning_rate": 3.374743401073775e-05, "loss": 0.0162, "step": 50010 }, { "grad_norm": 0.23868799209594727, "learning_rate": 3.372788939463095e-05, "loss": 0.0153, "step": 50020 }, { "grad_norm": 0.2967623174190521, "learning_rate": 3.370834755898281e-05, "loss": 0.0155, "step": 50030 }, { "grad_norm": 0.28274500370025635, "learning_rate": 3.3688808507132493e-05, "loss": 0.0157, "step": 50040 }, { "grad_norm": 0.19298884272575378, "learning_rate": 3.3669272242418685e-05, "loss": 0.0177, "step": 50050 }, { "grad_norm": 0.17436692118644714, "learning_rate": 3.364973876817961e-05, "loss": 0.0142, "step": 50060 }, { "grad_norm": 0.17841418087482452, "learning_rate": 3.363020808775299e-05, "loss": 0.0139, "step": 50070 }, { "grad_norm": 0.21445991098880768, "learning_rate": 3.361068020447611e-05, "loss": 0.0164, "step": 50080 }, { "grad_norm": 0.8740042448043823, "learning_rate": 3.3591155121685724e-05, "loss": 0.0138, "step": 50090 }, { "grad_norm": 0.7120270729064941, "learning_rate": 3.357163284271814e-05, "loss": 0.0153, "step": 50100 }, { "grad_norm": 0.2996841073036194, "learning_rate": 3.355211337090919e-05, "loss": 0.0178, "step": 50110 }, { "grad_norm": 0.20833086967468262, "learning_rate": 3.353259670959421e-05, "loss": 0.0165, "step": 50120 }, { "grad_norm": 0.21912626922130585, "learning_rate": 3.351308286210808e-05, "loss": 0.0146, "step": 50130 }, { "grad_norm": 0.6302359104156494, "learning_rate": 3.3493571831785156e-05, "loss": 0.021, "step": 50140 }, { "grad_norm": 0.2157498300075531, "learning_rate": 3.347406362195936e-05, "loss": 0.0181, "step": 50150 }, { "grad_norm": 0.20967037975788116, "learning_rate": 3.345455823596411e-05, "loss": 0.0168, "step": 50160 }, { "grad_norm": 0.4641346335411072, "learning_rate": 3.3435055677132346e-05, "loss": 0.015, "step": 50170 }, { "grad_norm": 0.20037348568439484, "learning_rate": 3.3415555948796505e-05, "loss": 0.0165, "step": 50180 }, { "grad_norm": 0.20215335488319397, "learning_rate": 3.3396059054288556e-05, "loss": 0.0133, "step": 50190 }, { "grad_norm": 0.16275210678577423, "learning_rate": 3.3376564996940015e-05, "loss": 0.027, "step": 50200 }, { "grad_norm": 0.15268820524215698, "learning_rate": 3.3357073780081836e-05, "loss": 0.0112, "step": 50210 }, { "grad_norm": 0.21369609236717224, "learning_rate": 3.333758540704459e-05, "loss": 0.0163, "step": 50220 }, { "grad_norm": 0.22649332880973816, "learning_rate": 3.331809988115827e-05, "loss": 0.015, "step": 50230 }, { "grad_norm": 0.9972896575927734, "learning_rate": 3.329861720575244e-05, "loss": 0.02, "step": 50240 }, { "grad_norm": 0.179194837808609, "learning_rate": 3.3279137384156126e-05, "loss": 0.0156, "step": 50250 }, { "grad_norm": 0.23832209408283234, "learning_rate": 3.3259660419697934e-05, "loss": 0.0174, "step": 50260 }, { "grad_norm": 0.22322577238082886, "learning_rate": 3.3240186315705926e-05, "loss": 0.0166, "step": 50270 }, { "grad_norm": 0.23651129007339478, "learning_rate": 3.322071507550769e-05, "loss": 0.0179, "step": 50280 }, { "grad_norm": 0.17272883653640747, "learning_rate": 3.320124670243033e-05, "loss": 0.0148, "step": 50290 }, { "grad_norm": 0.18663126230239868, "learning_rate": 3.318178119980045e-05, "loss": 0.0133, "step": 50300 }, { "grad_norm": 0.21021263301372528, "learning_rate": 3.316231857094421e-05, "loss": 0.015, "step": 50310 }, { "grad_norm": 0.2429860383272171, "learning_rate": 3.31428588191872e-05, "loss": 0.0159, "step": 50320 }, { "grad_norm": 0.35340988636016846, "learning_rate": 3.312340194785458e-05, "loss": 0.0153, "step": 50330 }, { "grad_norm": 0.2992362678050995, "learning_rate": 3.3103947960270975e-05, "loss": 0.0166, "step": 50340 }, { "grad_norm": 0.22478802502155304, "learning_rate": 3.308449685976058e-05, "loss": 0.0118, "step": 50350 }, { "grad_norm": 0.22063995897769928, "learning_rate": 3.3065048649647024e-05, "loss": 0.015, "step": 50360 }, { "grad_norm": 0.3352084457874298, "learning_rate": 3.304560333325348e-05, "loss": 0.014, "step": 50370 }, { "grad_norm": 0.3564188778400421, "learning_rate": 3.3026160913902624e-05, "loss": 0.0161, "step": 50380 }, { "grad_norm": 0.23155692219734192, "learning_rate": 3.300672139491662e-05, "loss": 0.0169, "step": 50390 }, { "grad_norm": 0.4675794243812561, "learning_rate": 3.298728477961717e-05, "loss": 0.0174, "step": 50400 }, { "grad_norm": 1.0523658990859985, "learning_rate": 3.296785107132544e-05, "loss": 0.0167, "step": 50410 }, { "grad_norm": 0.23836830258369446, "learning_rate": 3.2948420273362156e-05, "loss": 0.015, "step": 50420 }, { "grad_norm": 0.22492803633213043, "learning_rate": 3.292899238904747e-05, "loss": 0.0137, "step": 50430 }, { "grad_norm": 0.15877144038677216, "learning_rate": 3.29095674217011e-05, "loss": 0.0149, "step": 50440 }, { "grad_norm": 0.18736734986305237, "learning_rate": 3.289014537464224e-05, "loss": 0.0163, "step": 50450 }, { "grad_norm": 0.22224976122379303, "learning_rate": 3.287072625118955e-05, "loss": 0.0161, "step": 50460 }, { "grad_norm": 0.2187967300415039, "learning_rate": 3.285131005466129e-05, "loss": 0.0147, "step": 50470 }, { "grad_norm": 0.21661034226417542, "learning_rate": 3.2831896788375105e-05, "loss": 0.0106, "step": 50480 }, { "grad_norm": 0.23699992895126343, "learning_rate": 3.281248645564822e-05, "loss": 0.0137, "step": 50490 }, { "grad_norm": 0.22749614715576172, "learning_rate": 3.2793079059797306e-05, "loss": 0.0181, "step": 50500 }, { "grad_norm": 0.19577425718307495, "learning_rate": 3.277367460413859e-05, "loss": 0.0161, "step": 50510 }, { "grad_norm": 0.15607745945453644, "learning_rate": 3.275427309198773e-05, "loss": 0.0187, "step": 50520 }, { "grad_norm": 0.3200289309024811, "learning_rate": 3.273487452665993e-05, "loss": 0.0182, "step": 50530 }, { "grad_norm": 0.19935934245586395, "learning_rate": 3.271547891146986e-05, "loss": 0.0175, "step": 50540 }, { "grad_norm": 0.21762624382972717, "learning_rate": 3.269608624973173e-05, "loss": 0.0128, "step": 50550 }, { "grad_norm": 0.26646941900253296, "learning_rate": 3.26766965447592e-05, "loss": 0.0184, "step": 50560 }, { "grad_norm": 0.2016395628452301, "learning_rate": 3.2657309799865424e-05, "loss": 0.0166, "step": 50570 }, { "grad_norm": 0.21445119380950928, "learning_rate": 3.2637926018363084e-05, "loss": 0.0128, "step": 50580 }, { "grad_norm": 0.2418619841337204, "learning_rate": 3.261854520356432e-05, "loss": 0.0135, "step": 50590 }, { "grad_norm": 0.1589275747537613, "learning_rate": 3.2599167358780816e-05, "loss": 0.0128, "step": 50600 }, { "grad_norm": 0.36948704719543457, "learning_rate": 3.2579792487323676e-05, "loss": 0.0133, "step": 50610 }, { "grad_norm": 0.5239495038986206, "learning_rate": 3.256042059250358e-05, "loss": 0.0204, "step": 50620 }, { "grad_norm": 0.36003825068473816, "learning_rate": 3.254105167763062e-05, "loss": 0.0237, "step": 50630 }, { "grad_norm": 0.19322247803211212, "learning_rate": 3.252168574601443e-05, "loss": 0.016, "step": 50640 }, { "grad_norm": 0.2848495543003082, "learning_rate": 3.2502322800964115e-05, "loss": 0.0161, "step": 50650 }, { "grad_norm": 0.22514449059963226, "learning_rate": 3.248296284578825e-05, "loss": 0.0151, "step": 50660 }, { "grad_norm": 0.2901355028152466, "learning_rate": 3.246360588379497e-05, "loss": 0.0206, "step": 50670 }, { "grad_norm": 0.31828784942626953, "learning_rate": 3.24442519182918e-05, "loss": 0.0219, "step": 50680 }, { "grad_norm": 0.1719590276479721, "learning_rate": 3.2424900952585845e-05, "loss": 0.0164, "step": 50690 }, { "grad_norm": 0.1811172217130661, "learning_rate": 3.240555298998362e-05, "loss": 0.016, "step": 50700 }, { "grad_norm": 0.21348369121551514, "learning_rate": 3.2386208033791194e-05, "loss": 0.0133, "step": 50710 }, { "grad_norm": 0.2152653932571411, "learning_rate": 3.2366866087314064e-05, "loss": 0.0157, "step": 50720 }, { "grad_norm": 0.20678290724754333, "learning_rate": 3.234752715385727e-05, "loss": 0.016, "step": 50730 }, { "grad_norm": 0.2066020965576172, "learning_rate": 3.232819123672529e-05, "loss": 0.0185, "step": 50740 }, { "grad_norm": 0.28031209111213684, "learning_rate": 3.2308858339222076e-05, "loss": 0.0166, "step": 50750 }, { "grad_norm": 0.22708146274089813, "learning_rate": 3.2289528464651144e-05, "loss": 0.0179, "step": 50760 }, { "grad_norm": 0.2752761244773865, "learning_rate": 3.22702016163154e-05, "loss": 0.0182, "step": 50770 }, { "grad_norm": 0.27868667244911194, "learning_rate": 3.2250877797517295e-05, "loss": 0.0181, "step": 50780 }, { "grad_norm": 0.17990189790725708, "learning_rate": 3.223155701155872e-05, "loss": 0.0193, "step": 50790 }, { "grad_norm": 0.2270745187997818, "learning_rate": 3.2212239261741105e-05, "loss": 0.0188, "step": 50800 }, { "grad_norm": 0.1545436680316925, "learning_rate": 3.219292455136528e-05, "loss": 0.0129, "step": 50810 }, { "grad_norm": 0.2019096314907074, "learning_rate": 3.217361288373165e-05, "loss": 0.0176, "step": 50820 }, { "grad_norm": 0.20533983409404755, "learning_rate": 3.215430426214002e-05, "loss": 0.0143, "step": 50830 }, { "grad_norm": 0.4778682589530945, "learning_rate": 3.2134998689889705e-05, "loss": 0.0153, "step": 50840 }, { "grad_norm": 0.16160275042057037, "learning_rate": 3.211569617027952e-05, "loss": 0.0156, "step": 50850 }, { "grad_norm": 0.2287445366382599, "learning_rate": 3.209639670660769e-05, "loss": 0.0156, "step": 50860 }, { "grad_norm": 0.40877553820610046, "learning_rate": 3.207710030217202e-05, "loss": 0.0157, "step": 50870 }, { "grad_norm": 0.3351745903491974, "learning_rate": 3.205780696026972e-05, "loss": 0.0181, "step": 50880 }, { "grad_norm": 0.21174149215221405, "learning_rate": 3.203851668419749e-05, "loss": 0.0131, "step": 50890 }, { "grad_norm": 0.22378578782081604, "learning_rate": 3.201922947725149e-05, "loss": 0.0147, "step": 50900 }, { "grad_norm": 0.2733936607837677, "learning_rate": 3.199994534272742e-05, "loss": 0.017, "step": 50910 }, { "grad_norm": 0.20198433101177216, "learning_rate": 3.1980664283920394e-05, "loss": 0.0212, "step": 50920 }, { "grad_norm": 0.21633510291576385, "learning_rate": 3.196138630412499e-05, "loss": 0.0125, "step": 50930 }, { "grad_norm": 0.24514320492744446, "learning_rate": 3.194211140663532e-05, "loss": 0.0125, "step": 50940 }, { "grad_norm": 0.19139017164707184, "learning_rate": 3.1922839594744914e-05, "loss": 0.0196, "step": 50950 }, { "grad_norm": 0.1956922709941864, "learning_rate": 3.190357087174683e-05, "loss": 0.0157, "step": 50960 }, { "grad_norm": 0.1923018991947174, "learning_rate": 3.188430524093353e-05, "loss": 0.0176, "step": 50970 }, { "grad_norm": 0.2467687427997589, "learning_rate": 3.1865042705597014e-05, "loss": 0.0157, "step": 50980 }, { "grad_norm": 0.2354460060596466, "learning_rate": 3.184578326902868e-05, "loss": 0.014, "step": 50990 }, { "grad_norm": 0.3323843777179718, "learning_rate": 3.182652693451949e-05, "loss": 0.0171, "step": 51000 }, { "grad_norm": 0.27264249324798584, "learning_rate": 3.180727370535981e-05, "loss": 0.0172, "step": 51010 }, { "grad_norm": 0.17891398072242737, "learning_rate": 3.178802358483945e-05, "loss": 0.013, "step": 51020 }, { "grad_norm": 0.12916123867034912, "learning_rate": 3.17687765762478e-05, "loss": 0.0161, "step": 51030 }, { "grad_norm": 0.8377067446708679, "learning_rate": 3.174953268287357e-05, "loss": 0.0184, "step": 51040 }, { "grad_norm": 0.27073240280151367, "learning_rate": 3.173029190800508e-05, "loss": 0.0168, "step": 51050 }, { "grad_norm": 0.22033077478408813, "learning_rate": 3.171105425493e-05, "loss": 0.0125, "step": 51060 }, { "grad_norm": 0.37472641468048096, "learning_rate": 3.169181972693557e-05, "loss": 0.0131, "step": 51070 }, { "grad_norm": 0.6388717889785767, "learning_rate": 3.1672588327308404e-05, "loss": 0.0117, "step": 51080 }, { "grad_norm": 0.21348780393600464, "learning_rate": 3.1653360059334646e-05, "loss": 0.0136, "step": 51090 }, { "grad_norm": 0.3742675185203552, "learning_rate": 3.163413492629985e-05, "loss": 0.0279, "step": 51100 }, { "grad_norm": 0.2254713475704193, "learning_rate": 3.161491293148912e-05, "loss": 0.0144, "step": 51110 }, { "grad_norm": 0.22388383746147156, "learning_rate": 3.1595694078186925e-05, "loss": 0.0138, "step": 51120 }, { "grad_norm": 0.1936907321214676, "learning_rate": 3.157647836967724e-05, "loss": 0.015, "step": 51130 }, { "grad_norm": 0.3318159878253937, "learning_rate": 3.155726580924354e-05, "loss": 0.0167, "step": 51140 }, { "grad_norm": 0.20729966461658478, "learning_rate": 3.1538056400168676e-05, "loss": 0.0213, "step": 51150 }, { "grad_norm": 0.2332572340965271, "learning_rate": 3.151885014573506e-05, "loss": 0.0193, "step": 51160 }, { "grad_norm": 0.230331152677536, "learning_rate": 3.149964704922448e-05, "loss": 0.0206, "step": 51170 }, { "grad_norm": 0.2630431652069092, "learning_rate": 3.148044711391825e-05, "loss": 0.0161, "step": 51180 }, { "grad_norm": 0.16485895216464996, "learning_rate": 3.146125034309707e-05, "loss": 0.0139, "step": 51190 }, { "grad_norm": 0.26109158992767334, "learning_rate": 3.1442056740041195e-05, "loss": 0.0187, "step": 51200 }, { "grad_norm": 0.20827531814575195, "learning_rate": 3.142286630803026e-05, "loss": 0.0258, "step": 51210 }, { "grad_norm": 0.34113746881484985, "learning_rate": 3.140367905034337e-05, "loss": 0.0175, "step": 51220 }, { "grad_norm": 0.2189200073480606, "learning_rate": 3.138449497025914e-05, "loss": 0.0152, "step": 51230 }, { "grad_norm": 0.2244187891483307, "learning_rate": 3.136531407105557e-05, "loss": 0.0144, "step": 51240 }, { "grad_norm": 0.33643072843551636, "learning_rate": 3.1346136356010184e-05, "loss": 0.0206, "step": 51250 }, { "grad_norm": 0.23578400909900665, "learning_rate": 3.132696182839988e-05, "loss": 0.0123, "step": 51260 }, { "grad_norm": 0.37341436743736267, "learning_rate": 3.130779049150111e-05, "loss": 0.0145, "step": 51270 }, { "grad_norm": 0.3063383102416992, "learning_rate": 3.128862234858971e-05, "loss": 0.0133, "step": 51280 }, { "grad_norm": 0.23964878916740417, "learning_rate": 3.1269457402941005e-05, "loss": 0.0147, "step": 51290 }, { "grad_norm": 0.43639302253723145, "learning_rate": 3.125029565782974e-05, "loss": 0.016, "step": 51300 }, { "grad_norm": 0.26391538977622986, "learning_rate": 3.123113711653013e-05, "loss": 0.0181, "step": 51310 }, { "grad_norm": 0.22264990210533142, "learning_rate": 3.121198178231587e-05, "loss": 0.0164, "step": 51320 }, { "grad_norm": 0.25289386510849, "learning_rate": 3.119282965846006e-05, "loss": 0.0133, "step": 51330 }, { "grad_norm": 0.2502487003803253, "learning_rate": 3.1173680748235303e-05, "loss": 0.0162, "step": 51340 }, { "grad_norm": 0.20209956169128418, "learning_rate": 3.115453505491358e-05, "loss": 0.0165, "step": 51350 }, { "grad_norm": 0.23436924815177917, "learning_rate": 3.1135392581766404e-05, "loss": 0.0149, "step": 51360 }, { "grad_norm": 0.16539034247398376, "learning_rate": 3.111625333206467e-05, "loss": 0.0153, "step": 51370 }, { "grad_norm": 0.20054365694522858, "learning_rate": 3.109711730907881e-05, "loss": 0.01, "step": 51380 }, { "grad_norm": 0.24245557188987732, "learning_rate": 3.107798451607859e-05, "loss": 0.0133, "step": 51390 }, { "grad_norm": 0.13417232036590576, "learning_rate": 3.105885495633328e-05, "loss": 0.0107, "step": 51400 }, { "grad_norm": 1.153437852859497, "learning_rate": 3.103972863311163e-05, "loss": 0.0143, "step": 51410 }, { "grad_norm": 0.24988803267478943, "learning_rate": 3.102060554968178e-05, "loss": 0.0123, "step": 51420 }, { "grad_norm": 0.3555695712566376, "learning_rate": 3.100148570931137e-05, "loss": 0.0147, "step": 51430 }, { "grad_norm": 0.23348289728164673, "learning_rate": 3.098236911526744e-05, "loss": 0.013, "step": 51440 }, { "grad_norm": 0.23527680337429047, "learning_rate": 3.0963255770816495e-05, "loss": 0.0124, "step": 51450 }, { "grad_norm": 0.17908242344856262, "learning_rate": 3.094414567922447e-05, "loss": 0.0192, "step": 51460 }, { "grad_norm": 0.2056492120027542, "learning_rate": 3.092503884375678e-05, "loss": 0.0165, "step": 51470 }, { "grad_norm": 0.284132182598114, "learning_rate": 3.0905935267678255e-05, "loss": 0.0154, "step": 51480 }, { "grad_norm": 0.17031227052211761, "learning_rate": 3.0886834954253166e-05, "loss": 0.016, "step": 51490 }, { "grad_norm": 0.2360488623380661, "learning_rate": 3.086773790674524e-05, "loss": 0.0126, "step": 51500 }, { "grad_norm": 0.15442539751529694, "learning_rate": 3.0848644128417617e-05, "loss": 0.0146, "step": 51510 }, { "grad_norm": 0.4541492164134979, "learning_rate": 3.082955362253294e-05, "loss": 0.0193, "step": 51520 }, { "grad_norm": 0.3265664279460907, "learning_rate": 3.081046639235323e-05, "loss": 0.0171, "step": 51530 }, { "grad_norm": 0.2157924771308899, "learning_rate": 3.0791382441139986e-05, "loss": 0.0098, "step": 51540 }, { "grad_norm": 0.21859554946422577, "learning_rate": 3.077230177215411e-05, "loss": 0.0163, "step": 51550 }, { "grad_norm": 0.247928649187088, "learning_rate": 3.0753224388656e-05, "loss": 0.0169, "step": 51560 }, { "grad_norm": 0.21478061378002167, "learning_rate": 3.073415029390544e-05, "loss": 0.0186, "step": 51570 }, { "grad_norm": 0.20785099267959595, "learning_rate": 3.071507949116166e-05, "loss": 0.0152, "step": 51580 }, { "grad_norm": 0.23736652731895447, "learning_rate": 3.069601198368337e-05, "loss": 0.0118, "step": 51590 }, { "grad_norm": 0.266233891248703, "learning_rate": 3.067694777472864e-05, "loss": 0.0149, "step": 51600 }, { "grad_norm": 0.2361045777797699, "learning_rate": 3.065788686755508e-05, "loss": 0.0181, "step": 51610 }, { "grad_norm": 0.23657254874706268, "learning_rate": 3.063882926541961e-05, "loss": 0.018, "step": 51620 }, { "grad_norm": 0.6089599132537842, "learning_rate": 3.061977497157872e-05, "loss": 0.0137, "step": 51630 }, { "grad_norm": 0.1788933277130127, "learning_rate": 3.0600723989288224e-05, "loss": 0.0157, "step": 51640 }, { "grad_norm": 0.15307027101516724, "learning_rate": 3.058167632180343e-05, "loss": 0.0121, "step": 51650 }, { "grad_norm": 0.21776527166366577, "learning_rate": 3.0562631972379074e-05, "loss": 0.0155, "step": 51660 }, { "grad_norm": 0.20773448050022125, "learning_rate": 3.0543590944269276e-05, "loss": 0.0135, "step": 51670 }, { "grad_norm": 0.25253671407699585, "learning_rate": 3.052455324072766e-05, "loss": 0.0197, "step": 51680 }, { "grad_norm": 0.20474572479724884, "learning_rate": 3.0505518865007244e-05, "loss": 0.0153, "step": 51690 }, { "grad_norm": 0.8693388104438782, "learning_rate": 3.048648782036048e-05, "loss": 0.0173, "step": 51700 }, { "grad_norm": 0.2367953211069107, "learning_rate": 3.0467460110039236e-05, "loss": 0.0196, "step": 51710 }, { "grad_norm": 0.28667381405830383, "learning_rate": 3.0448435737294866e-05, "loss": 0.0154, "step": 51720 }, { "grad_norm": 0.3117258846759796, "learning_rate": 3.042941470537808e-05, "loss": 0.0111, "step": 51730 }, { "grad_norm": 0.22591854631900787, "learning_rate": 3.0410397017539072e-05, "loss": 0.0132, "step": 51740 }, { "grad_norm": 0.2149108350276947, "learning_rate": 3.039138267702742e-05, "loss": 0.0127, "step": 51750 }, { "grad_norm": 0.21141989529132843, "learning_rate": 3.0372371687092193e-05, "loss": 0.0127, "step": 51760 }, { "grad_norm": 0.20586274564266205, "learning_rate": 3.0353364050981823e-05, "loss": 0.0161, "step": 51770 }, { "grad_norm": 0.16644993424415588, "learning_rate": 3.033435977194418e-05, "loss": 0.0164, "step": 51780 }, { "grad_norm": 0.19722165167331696, "learning_rate": 3.0315358853226618e-05, "loss": 0.0161, "step": 51790 }, { "grad_norm": 0.23106569051742554, "learning_rate": 3.0296361298075844e-05, "loss": 0.0125, "step": 51800 }, { "grad_norm": 0.23393361270427704, "learning_rate": 3.027736710973803e-05, "loss": 0.017, "step": 51810 }, { "grad_norm": 0.1711883246898651, "learning_rate": 3.025837629145875e-05, "loss": 0.0097, "step": 51820 }, { "grad_norm": 0.22527380287647247, "learning_rate": 3.0239388846483048e-05, "loss": 0.0139, "step": 51830 }, { "grad_norm": 0.25032490491867065, "learning_rate": 3.022040477805532e-05, "loss": 0.0202, "step": 51840 }, { "grad_norm": 0.288666695356369, "learning_rate": 3.020142408941946e-05, "loss": 0.0122, "step": 51850 }, { "grad_norm": 0.2052503377199173, "learning_rate": 3.018244678381873e-05, "loss": 0.0156, "step": 51860 }, { "grad_norm": 0.28828758001327515, "learning_rate": 3.0163472864495812e-05, "loss": 0.0177, "step": 51870 }, { "grad_norm": 0.2752626836299896, "learning_rate": 3.014450233469287e-05, "loss": 0.0145, "step": 51880 }, { "grad_norm": 0.2715108394622803, "learning_rate": 3.0125535197651422e-05, "loss": 0.0168, "step": 51890 }, { "grad_norm": 0.1973939836025238, "learning_rate": 3.0106571456612448e-05, "loss": 0.0126, "step": 51900 }, { "grad_norm": 0.38283059000968933, "learning_rate": 3.0087611114816305e-05, "loss": 0.0248, "step": 51910 }, { "grad_norm": 0.20185472071170807, "learning_rate": 3.006865417550284e-05, "loss": 0.0139, "step": 51920 }, { "grad_norm": 0.21063071489334106, "learning_rate": 3.0049700641911242e-05, "loss": 0.0167, "step": 51930 }, { "grad_norm": 0.13700732588768005, "learning_rate": 3.0030750517280175e-05, "loss": 0.0147, "step": 51940 }, { "grad_norm": 0.17437304556369781, "learning_rate": 3.0011803804847682e-05, "loss": 0.0177, "step": 51950 }, { "grad_norm": 0.2457733154296875, "learning_rate": 2.9992860507851224e-05, "loss": 0.0164, "step": 51960 }, { "grad_norm": 0.18124394118785858, "learning_rate": 2.9973920629527723e-05, "loss": 0.0109, "step": 51970 }, { "grad_norm": 0.28392860293388367, "learning_rate": 2.9954984173113453e-05, "loss": 0.0159, "step": 51980 }, { "grad_norm": 0.2462446242570877, "learning_rate": 2.993605114184418e-05, "loss": 0.0131, "step": 51990 }, { "grad_norm": 0.29546406865119934, "learning_rate": 2.9917121538955005e-05, "loss": 0.0174, "step": 52000 }, { "grad_norm": 0.3419764041900635, "learning_rate": 2.98981953676805e-05, "loss": 0.0225, "step": 52010 }, { "grad_norm": 0.2069687396287918, "learning_rate": 2.9879272631254594e-05, "loss": 0.0143, "step": 52020 }, { "grad_norm": 0.13328105211257935, "learning_rate": 2.986035333291073e-05, "loss": 0.0122, "step": 52030 }, { "grad_norm": 0.21593013405799866, "learning_rate": 2.9841437475881646e-05, "loss": 0.0238, "step": 52040 }, { "grad_norm": 0.21339823305606842, "learning_rate": 2.9822525063399555e-05, "loss": 0.0142, "step": 52050 }, { "grad_norm": 0.22626568377017975, "learning_rate": 2.9803616098696087e-05, "loss": 0.0161, "step": 52060 }, { "grad_norm": 0.33095481991767883, "learning_rate": 2.9784710585002228e-05, "loss": 0.02, "step": 52070 }, { "grad_norm": 0.23122958838939667, "learning_rate": 2.9765808525548467e-05, "loss": 0.011, "step": 52080 }, { "grad_norm": 0.24946418404579163, "learning_rate": 2.974690992356461e-05, "loss": 0.0158, "step": 52090 }, { "grad_norm": 0.2729038894176483, "learning_rate": 2.9728014782279934e-05, "loss": 0.0181, "step": 52100 }, { "grad_norm": 0.27302902936935425, "learning_rate": 2.970912310492307e-05, "loss": 0.0132, "step": 52110 }, { "grad_norm": 0.3736557364463806, "learning_rate": 2.9690234894722134e-05, "loss": 0.0159, "step": 52120 }, { "grad_norm": 0.217841237783432, "learning_rate": 2.9671350154904577e-05, "loss": 0.0179, "step": 52130 }, { "grad_norm": 0.2817344069480896, "learning_rate": 2.9652468888697282e-05, "loss": 0.016, "step": 52140 }, { "grad_norm": 0.25961029529571533, "learning_rate": 2.9633591099326562e-05, "loss": 0.0149, "step": 52150 }, { "grad_norm": 0.22782190144062042, "learning_rate": 2.9614716790018072e-05, "loss": 0.0112, "step": 52160 }, { "grad_norm": 0.2178090214729309, "learning_rate": 2.959584596399697e-05, "loss": 0.0114, "step": 52170 }, { "grad_norm": 0.26993027329444885, "learning_rate": 2.9576978624487717e-05, "loss": 0.0198, "step": 52180 }, { "grad_norm": 0.15439528226852417, "learning_rate": 2.9558114774714263e-05, "loss": 0.0124, "step": 52190 }, { "grad_norm": 0.22445201873779297, "learning_rate": 2.9539254417899897e-05, "loss": 0.0161, "step": 52200 }, { "grad_norm": 1.1320512294769287, "learning_rate": 2.9520397557267365e-05, "loss": 0.0161, "step": 52210 }, { "grad_norm": 0.2908609211444855, "learning_rate": 2.9501544196038765e-05, "loss": 0.0182, "step": 52220 }, { "grad_norm": 0.36271995306015015, "learning_rate": 2.9482694337435614e-05, "loss": 0.0164, "step": 52230 }, { "grad_norm": 0.22535447776317596, "learning_rate": 2.946384798467887e-05, "loss": 0.0136, "step": 52240 }, { "grad_norm": 0.17221833765506744, "learning_rate": 2.9445005140988824e-05, "loss": 0.0107, "step": 52250 }, { "grad_norm": 0.29712995886802673, "learning_rate": 2.942616580958524e-05, "loss": 0.0128, "step": 52260 }, { "grad_norm": 0.220149427652359, "learning_rate": 2.9407329993687193e-05, "loss": 0.0141, "step": 52270 }, { "grad_norm": 0.24461527168750763, "learning_rate": 2.938849769651326e-05, "loss": 0.0133, "step": 52280 }, { "grad_norm": 0.19697757065296173, "learning_rate": 2.9369668921281345e-05, "loss": 0.0168, "step": 52290 }, { "grad_norm": 0.4206762909889221, "learning_rate": 2.9350843671208773e-05, "loss": 0.0128, "step": 52300 }, { "grad_norm": 0.292093962430954, "learning_rate": 2.9332021949512245e-05, "loss": 0.0144, "step": 52310 }, { "grad_norm": 0.24042804539203644, "learning_rate": 2.9313203759407908e-05, "loss": 0.0154, "step": 52320 }, { "grad_norm": 0.1966237723827362, "learning_rate": 2.929438910411127e-05, "loss": 0.0173, "step": 52330 }, { "grad_norm": 0.3116878867149353, "learning_rate": 2.927557798683722e-05, "loss": 0.0182, "step": 52340 }, { "grad_norm": 0.20431676506996155, "learning_rate": 2.9256770410800095e-05, "loss": 0.013, "step": 52350 }, { "grad_norm": 0.3012312352657318, "learning_rate": 2.9237966379213554e-05, "loss": 0.0142, "step": 52360 }, { "grad_norm": 0.1702134758234024, "learning_rate": 2.9219165895290736e-05, "loss": 0.0164, "step": 52370 }, { "grad_norm": 0.20000679790973663, "learning_rate": 2.9200368962244083e-05, "loss": 0.0177, "step": 52380 }, { "grad_norm": 0.2287249118089676, "learning_rate": 2.9181575583285536e-05, "loss": 0.0125, "step": 52390 }, { "grad_norm": 0.19866593182086945, "learning_rate": 2.916278576162632e-05, "loss": 0.0154, "step": 52400 }, { "grad_norm": 0.23485532402992249, "learning_rate": 2.9143999500477116e-05, "loss": 0.012, "step": 52410 }, { "grad_norm": 0.1934860795736313, "learning_rate": 2.9125216803048004e-05, "loss": 0.0188, "step": 52420 }, { "grad_norm": 0.22340217232704163, "learning_rate": 2.9106437672548403e-05, "loss": 0.016, "step": 52430 }, { "grad_norm": 0.21059292554855347, "learning_rate": 2.9087662112187154e-05, "loss": 0.0108, "step": 52440 }, { "grad_norm": 0.21563616394996643, "learning_rate": 2.9068890125172498e-05, "loss": 0.0133, "step": 52450 }, { "grad_norm": 0.21964643895626068, "learning_rate": 2.9050121714712058e-05, "loss": 0.0177, "step": 52460 }, { "grad_norm": 0.23155780136585236, "learning_rate": 2.9031356884012835e-05, "loss": 0.0166, "step": 52470 }, { "grad_norm": 0.28851041197776794, "learning_rate": 2.9012595636281247e-05, "loss": 0.0143, "step": 52480 }, { "grad_norm": 0.16568897664546967, "learning_rate": 2.8993837974723016e-05, "loss": 0.0255, "step": 52490 }, { "grad_norm": 0.1608954221010208, "learning_rate": 2.8975083902543394e-05, "loss": 0.0192, "step": 52500 }, { "grad_norm": 0.23662962019443512, "learning_rate": 2.895633342294688e-05, "loss": 0.013, "step": 52510 }, { "grad_norm": 0.2563936710357666, "learning_rate": 2.893758653913744e-05, "loss": 0.0133, "step": 52520 }, { "grad_norm": 0.280697762966156, "learning_rate": 2.8918843254318423e-05, "loss": 0.0176, "step": 52530 }, { "grad_norm": 0.3194034993648529, "learning_rate": 2.8900103571692483e-05, "loss": 0.0158, "step": 52540 }, { "grad_norm": 0.3599799871444702, "learning_rate": 2.8881367494461796e-05, "loss": 0.0134, "step": 52550 }, { "grad_norm": 0.2512679696083069, "learning_rate": 2.886263502582779e-05, "loss": 0.0141, "step": 52560 }, { "grad_norm": 0.4274950921535492, "learning_rate": 2.8843906168991353e-05, "loss": 0.0146, "step": 52570 }, { "grad_norm": 0.19396547973155975, "learning_rate": 2.8825180927152728e-05, "loss": 0.0153, "step": 52580 }, { "grad_norm": 0.19273245334625244, "learning_rate": 2.8806459303511547e-05, "loss": 0.013, "step": 52590 }, { "grad_norm": 0.3093356192111969, "learning_rate": 2.8787741301266852e-05, "loss": 0.0147, "step": 52600 }, { "grad_norm": 0.21577881276607513, "learning_rate": 2.876902692361697e-05, "loss": 0.0168, "step": 52610 }, { "grad_norm": 0.45946377515792847, "learning_rate": 2.875031617375976e-05, "loss": 0.0132, "step": 52620 }, { "grad_norm": 0.20796635746955872, "learning_rate": 2.8731609054892318e-05, "loss": 0.0146, "step": 52630 }, { "grad_norm": 0.5198735594749451, "learning_rate": 2.871290557021119e-05, "loss": 0.0141, "step": 52640 }, { "grad_norm": 0.21690084040164948, "learning_rate": 2.8694205722912298e-05, "loss": 0.0121, "step": 52650 }, { "grad_norm": 0.411465048789978, "learning_rate": 2.8675509516190936e-05, "loss": 0.0156, "step": 52660 }, { "grad_norm": 0.1627606600522995, "learning_rate": 2.8656816953241772e-05, "loss": 0.0172, "step": 52670 }, { "grad_norm": 0.20817698538303375, "learning_rate": 2.863812803725887e-05, "loss": 0.011, "step": 52680 }, { "grad_norm": 0.2600213885307312, "learning_rate": 2.8619442771435623e-05, "loss": 0.0121, "step": 52690 }, { "grad_norm": 0.2805878818035126, "learning_rate": 2.860076115896484e-05, "loss": 0.018, "step": 52700 }, { "grad_norm": 0.22614726424217224, "learning_rate": 2.8582083203038712e-05, "loss": 0.0137, "step": 52710 }, { "grad_norm": 0.2596060335636139, "learning_rate": 2.8563408906848778e-05, "loss": 0.0162, "step": 52720 }, { "grad_norm": 0.2128164917230606, "learning_rate": 2.8544738273585993e-05, "loss": 0.0105, "step": 52730 }, { "grad_norm": 0.14101988077163696, "learning_rate": 2.8526071306440595e-05, "loss": 0.0122, "step": 52740 }, { "grad_norm": 0.30250728130340576, "learning_rate": 2.850740800860233e-05, "loss": 0.0135, "step": 52750 }, { "grad_norm": 0.580085277557373, "learning_rate": 2.8488748383260178e-05, "loss": 0.0157, "step": 52760 }, { "grad_norm": 0.17241473495960236, "learning_rate": 2.8470092433602635e-05, "loss": 0.0228, "step": 52770 }, { "grad_norm": 0.1803371012210846, "learning_rate": 2.845144016281745e-05, "loss": 0.0149, "step": 52780 }, { "grad_norm": 0.238543301820755, "learning_rate": 2.8432791574091744e-05, "loss": 0.0232, "step": 52790 }, { "grad_norm": 0.2158220261335373, "learning_rate": 2.8414146670612134e-05, "loss": 0.0178, "step": 52800 }, { "grad_norm": 0.1753816157579422, "learning_rate": 2.8395505455564446e-05, "loss": 0.016, "step": 52810 }, { "grad_norm": 0.19943858683109283, "learning_rate": 2.837686793213403e-05, "loss": 0.0156, "step": 52820 }, { "grad_norm": 0.3119356036186218, "learning_rate": 2.8358234103505477e-05, "loss": 0.0144, "step": 52830 }, { "grad_norm": 0.19867154955863953, "learning_rate": 2.8339603972862806e-05, "loss": 0.0157, "step": 52840 }, { "grad_norm": 0.197772815823555, "learning_rate": 2.83209775433894e-05, "loss": 0.0141, "step": 52850 }, { "grad_norm": 0.2456091195344925, "learning_rate": 2.8302354818268013e-05, "loss": 0.0146, "step": 52860 }, { "grad_norm": 0.25530123710632324, "learning_rate": 2.8283735800680754e-05, "loss": 0.0142, "step": 52870 }, { "grad_norm": 0.27236804366111755, "learning_rate": 2.826512049380913e-05, "loss": 0.013, "step": 52880 }, { "grad_norm": 0.19154459238052368, "learning_rate": 2.8246508900833935e-05, "loss": 0.0184, "step": 52890 }, { "grad_norm": 0.22734712064266205, "learning_rate": 2.822790102493541e-05, "loss": 0.0166, "step": 52900 }, { "grad_norm": 0.23165260255336761, "learning_rate": 2.820929686929314e-05, "loss": 0.0159, "step": 52910 }, { "grad_norm": 0.17326603829860687, "learning_rate": 2.819069643708605e-05, "loss": 0.0119, "step": 52920 }, { "grad_norm": 0.346439927816391, "learning_rate": 2.8172099731492484e-05, "loss": 0.0179, "step": 52930 }, { "grad_norm": 0.2407607138156891, "learning_rate": 2.815350675569003e-05, "loss": 0.0186, "step": 52940 }, { "grad_norm": 0.2129305899143219, "learning_rate": 2.8134917512855807e-05, "loss": 0.0175, "step": 52950 }, { "grad_norm": 0.2609189748764038, "learning_rate": 2.8116332006166134e-05, "loss": 0.0115, "step": 52960 }, { "grad_norm": 0.26970231533050537, "learning_rate": 2.809775023879685e-05, "loss": 0.0207, "step": 52970 }, { "grad_norm": 0.2868492901325226, "learning_rate": 2.807917221392299e-05, "loss": 0.0145, "step": 52980 }, { "grad_norm": 0.18182195723056793, "learning_rate": 2.8060597934719067e-05, "loss": 0.0121, "step": 52990 }, { "grad_norm": 0.23497873544692993, "learning_rate": 2.8042027404358927e-05, "loss": 0.0131, "step": 53000 }, { "grad_norm": 1.8565514087677002, "learning_rate": 2.802346062601572e-05, "loss": 0.015, "step": 53010 }, { "grad_norm": 0.29769042134284973, "learning_rate": 2.800489760286207e-05, "loss": 0.0215, "step": 53020 }, { "grad_norm": 0.21931926906108856, "learning_rate": 2.7986338338069834e-05, "loss": 0.0114, "step": 53030 }, { "grad_norm": 0.24948027729988098, "learning_rate": 2.7967782834810297e-05, "loss": 0.0182, "step": 53040 }, { "grad_norm": 0.8673292398452759, "learning_rate": 2.794923109625409e-05, "loss": 0.0123, "step": 53050 }, { "grad_norm": 0.23712356388568878, "learning_rate": 2.7930683125571205e-05, "loss": 0.0129, "step": 53060 }, { "grad_norm": 0.2720944881439209, "learning_rate": 2.7912138925930997e-05, "loss": 0.0157, "step": 53070 }, { "grad_norm": 0.27013465762138367, "learning_rate": 2.7893598500502117e-05, "loss": 0.0216, "step": 53080 }, { "grad_norm": 0.24822700023651123, "learning_rate": 2.7875061852452644e-05, "loss": 0.0174, "step": 53090 }, { "grad_norm": 0.30021780729293823, "learning_rate": 2.7856528984949982e-05, "loss": 0.0126, "step": 53100 }, { "grad_norm": 0.18379762768745422, "learning_rate": 2.7837999901160888e-05, "loss": 0.0155, "step": 53110 }, { "grad_norm": 0.3117358982563019, "learning_rate": 2.7819474604251484e-05, "loss": 0.012, "step": 53120 }, { "grad_norm": 0.2424369752407074, "learning_rate": 2.780095309738725e-05, "loss": 0.0141, "step": 53130 }, { "grad_norm": 0.22498536109924316, "learning_rate": 2.778243538373294e-05, "loss": 0.0121, "step": 53140 }, { "grad_norm": 0.1889858990907669, "learning_rate": 2.7763921466452826e-05, "loss": 0.0144, "step": 53150 }, { "grad_norm": 0.2199745774269104, "learning_rate": 2.7745411348710336e-05, "loss": 0.0208, "step": 53160 }, { "grad_norm": 0.22792251408100128, "learning_rate": 2.7726905033668395e-05, "loss": 0.0151, "step": 53170 }, { "grad_norm": 1.1905931234359741, "learning_rate": 2.7708402524489214e-05, "loss": 0.0181, "step": 53180 }, { "grad_norm": 0.2077733725309372, "learning_rate": 2.7689903824334364e-05, "loss": 0.0151, "step": 53190 }, { "grad_norm": 0.49272841215133667, "learning_rate": 2.7671408936364785e-05, "loss": 0.0154, "step": 53200 }, { "grad_norm": 0.1883896142244339, "learning_rate": 2.765291786374069e-05, "loss": 0.0143, "step": 53210 }, { "grad_norm": 0.2733362317085266, "learning_rate": 2.7634430609621775e-05, "loss": 0.0125, "step": 53220 }, { "grad_norm": 0.4423409402370453, "learning_rate": 2.7615947177166956e-05, "loss": 0.0127, "step": 53230 }, { "grad_norm": 0.22652125358581543, "learning_rate": 2.7597467569534553e-05, "loss": 0.0128, "step": 53240 }, { "grad_norm": 0.33568447828292847, "learning_rate": 2.757899178988226e-05, "loss": 0.0126, "step": 53250 }, { "grad_norm": 0.3939739465713501, "learning_rate": 2.7560519841367005e-05, "loss": 0.0129, "step": 53260 }, { "grad_norm": 0.2107423096895218, "learning_rate": 2.7542051727145237e-05, "loss": 0.0141, "step": 53270 }, { "grad_norm": 0.20872819423675537, "learning_rate": 2.7523587450372578e-05, "loss": 0.0154, "step": 53280 }, { "grad_norm": 0.5521530508995056, "learning_rate": 2.750512701420409e-05, "loss": 0.0199, "step": 53290 }, { "grad_norm": 0.18793846666812897, "learning_rate": 2.7486670421794158e-05, "loss": 0.0154, "step": 53300 }, { "grad_norm": 0.20115727186203003, "learning_rate": 2.7468217676296515e-05, "loss": 0.0162, "step": 53310 }, { "grad_norm": 0.27278968691825867, "learning_rate": 2.7449768780864226e-05, "loss": 0.0158, "step": 53320 }, { "grad_norm": 0.2216760814189911, "learning_rate": 2.7431323738649724e-05, "loss": 0.0111, "step": 53330 }, { "grad_norm": 0.3024008274078369, "learning_rate": 2.7412882552804713e-05, "loss": 0.0173, "step": 53340 }, { "grad_norm": 0.16682939231395721, "learning_rate": 2.739444522648032e-05, "loss": 0.0117, "step": 53350 }, { "grad_norm": 0.15480850636959076, "learning_rate": 2.737601176282697e-05, "loss": 0.0125, "step": 53360 }, { "grad_norm": 0.2502470314502716, "learning_rate": 2.735758216499445e-05, "loss": 0.0144, "step": 53370 }, { "grad_norm": 0.16410449147224426, "learning_rate": 2.7339156436131864e-05, "loss": 0.0136, "step": 53380 }, { "grad_norm": 0.2152758687734604, "learning_rate": 2.7320734579387663e-05, "loss": 0.0142, "step": 53390 }, { "grad_norm": 0.1600075662136078, "learning_rate": 2.730231659790966e-05, "loss": 0.0114, "step": 53400 }, { "grad_norm": 0.22465632855892181, "learning_rate": 2.728390249484494e-05, "loss": 0.0224, "step": 53410 }, { "grad_norm": 0.2596275508403778, "learning_rate": 2.7265492273340032e-05, "loss": 0.0127, "step": 53420 }, { "grad_norm": 0.19234418869018555, "learning_rate": 2.7247085936540678e-05, "loss": 0.0116, "step": 53430 }, { "grad_norm": 0.29147812724113464, "learning_rate": 2.722868348759205e-05, "loss": 0.0133, "step": 53440 }, { "grad_norm": 0.19148355722427368, "learning_rate": 2.7210284929638635e-05, "loss": 0.0146, "step": 53450 }, { "grad_norm": 0.2198815792798996, "learning_rate": 2.7191890265824183e-05, "loss": 0.0129, "step": 53460 }, { "grad_norm": 0.222874253988266, "learning_rate": 2.7173499499291926e-05, "loss": 0.0154, "step": 53470 }, { "grad_norm": 0.42232686281204224, "learning_rate": 2.7155112633184277e-05, "loss": 0.0177, "step": 53480 }, { "grad_norm": 0.15467114746570587, "learning_rate": 2.713672967064307e-05, "loss": 0.0155, "step": 53490 }, { "grad_norm": 0.17654949426651, "learning_rate": 2.711835061480945e-05, "loss": 0.0115, "step": 53500 }, { "grad_norm": 0.170064315199852, "learning_rate": 2.7099975468823896e-05, "loss": 0.0163, "step": 53510 }, { "grad_norm": 0.22251489758491516, "learning_rate": 2.708160423582622e-05, "loss": 0.0116, "step": 53520 }, { "grad_norm": 0.5013467669487, "learning_rate": 2.706323691895557e-05, "loss": 0.0169, "step": 53530 }, { "grad_norm": 0.2725560963153839, "learning_rate": 2.70448735213504e-05, "loss": 0.0148, "step": 53540 }, { "grad_norm": 0.22015134990215302, "learning_rate": 2.702651404614852e-05, "loss": 0.0142, "step": 53550 }, { "grad_norm": 0.19890646636486053, "learning_rate": 2.7008158496487056e-05, "loss": 0.0154, "step": 53560 }, { "grad_norm": 0.1944604068994522, "learning_rate": 2.6989806875502487e-05, "loss": 0.0122, "step": 53570 }, { "grad_norm": 0.4387090504169464, "learning_rate": 2.6971459186330584e-05, "loss": 0.0122, "step": 53580 }, { "grad_norm": 0.23847392201423645, "learning_rate": 2.695311543210648e-05, "loss": 0.0131, "step": 53590 }, { "grad_norm": 0.24061313271522522, "learning_rate": 2.693477561596463e-05, "loss": 0.0138, "step": 53600 }, { "grad_norm": 0.3808591961860657, "learning_rate": 2.6916439741038756e-05, "loss": 0.0122, "step": 53610 }, { "grad_norm": 0.1994265615940094, "learning_rate": 2.689810781046203e-05, "loss": 0.0169, "step": 53620 }, { "grad_norm": 0.2636284828186035, "learning_rate": 2.6879779827366823e-05, "loss": 0.0123, "step": 53630 }, { "grad_norm": 0.2222166210412979, "learning_rate": 2.6861455794884904e-05, "loss": 0.0135, "step": 53640 }, { "grad_norm": 0.17020440101623535, "learning_rate": 2.6843135716147373e-05, "loss": 0.0132, "step": 53650 }, { "grad_norm": 0.168840691447258, "learning_rate": 2.6824819594284556e-05, "loss": 0.0147, "step": 53660 }, { "grad_norm": 0.14418448507785797, "learning_rate": 2.6806507432426275e-05, "loss": 0.0132, "step": 53670 }, { "grad_norm": 0.17769500613212585, "learning_rate": 2.6788199233701512e-05, "loss": 0.0132, "step": 53680 }, { "grad_norm": 0.18224197626113892, "learning_rate": 2.6769895001238652e-05, "loss": 0.0155, "step": 53690 }, { "grad_norm": 0.18940424919128418, "learning_rate": 2.67515947381654e-05, "loss": 0.014, "step": 53700 }, { "grad_norm": 0.2580348551273346, "learning_rate": 2.6733298447608745e-05, "loss": 0.0159, "step": 53710 }, { "grad_norm": 0.45090100169181824, "learning_rate": 2.6715006132695074e-05, "loss": 0.0141, "step": 53720 }, { "grad_norm": 0.26613649725914, "learning_rate": 2.6696717796549985e-05, "loss": 0.019, "step": 53730 }, { "grad_norm": 0.23609983921051025, "learning_rate": 2.667843344229848e-05, "loss": 0.0127, "step": 53740 }, { "grad_norm": 0.3315834403038025, "learning_rate": 2.6660153073064846e-05, "loss": 0.0145, "step": 53750 }, { "grad_norm": 0.18992814421653748, "learning_rate": 2.664187669197271e-05, "loss": 0.0215, "step": 53760 }, { "grad_norm": 0.24420127272605896, "learning_rate": 2.6623604302145005e-05, "loss": 0.0136, "step": 53770 }, { "grad_norm": 0.19166232645511627, "learning_rate": 2.6605335906703975e-05, "loss": 0.0157, "step": 53780 }, { "grad_norm": 0.2141086608171463, "learning_rate": 2.65870715087712e-05, "loss": 0.0142, "step": 53790 }, { "grad_norm": 0.26244789361953735, "learning_rate": 2.6568811111467573e-05, "loss": 0.0128, "step": 53800 }, { "grad_norm": 0.25438132882118225, "learning_rate": 2.6550554717913258e-05, "loss": 0.0136, "step": 53810 }, { "grad_norm": 0.17789818346500397, "learning_rate": 2.6532302331227805e-05, "loss": 0.0159, "step": 53820 }, { "grad_norm": 0.30363747477531433, "learning_rate": 2.651405395453004e-05, "loss": 0.0113, "step": 53830 }, { "grad_norm": 0.20210932195186615, "learning_rate": 2.6495809590938115e-05, "loss": 0.0188, "step": 53840 }, { "grad_norm": 0.1918298453092575, "learning_rate": 2.64775692435695e-05, "loss": 0.0193, "step": 53850 }, { "grad_norm": 0.18815365433692932, "learning_rate": 2.6459332915540928e-05, "loss": 0.0135, "step": 53860 }, { "grad_norm": 0.19710180163383484, "learning_rate": 2.644110060996856e-05, "loss": 0.0155, "step": 53870 }, { "grad_norm": 0.14202113449573517, "learning_rate": 2.642287232996774e-05, "loss": 0.0139, "step": 53880 }, { "grad_norm": 0.19717735052108765, "learning_rate": 2.6404648078653205e-05, "loss": 0.0119, "step": 53890 }, { "grad_norm": 0.2600823938846588, "learning_rate": 2.6386427859139002e-05, "loss": 0.0092, "step": 53900 }, { "grad_norm": 0.2536807358264923, "learning_rate": 2.63682116745384e-05, "loss": 0.0184, "step": 53910 }, { "grad_norm": 0.31184419989585876, "learning_rate": 2.6349999527964138e-05, "loss": 0.0124, "step": 53920 }, { "grad_norm": 0.1736505776643753, "learning_rate": 2.63317914225281e-05, "loss": 0.0181, "step": 53930 }, { "grad_norm": 0.300526887178421, "learning_rate": 2.6313587361341585e-05, "loss": 0.0168, "step": 53940 }, { "grad_norm": 0.39435434341430664, "learning_rate": 2.6295387347515165e-05, "loss": 0.0174, "step": 53950 }, { "grad_norm": 0.17458926141262054, "learning_rate": 2.6277191384158727e-05, "loss": 0.0145, "step": 53960 }, { "grad_norm": 0.20190727710723877, "learning_rate": 2.6258999474381457e-05, "loss": 0.0194, "step": 53970 }, { "grad_norm": 0.17606277763843536, "learning_rate": 2.624081162129186e-05, "loss": 0.0126, "step": 53980 }, { "grad_norm": 0.13848048448562622, "learning_rate": 2.6222627827997765e-05, "loss": 0.0201, "step": 53990 }, { "grad_norm": 0.21423381567001343, "learning_rate": 2.6204448097606236e-05, "loss": 0.0118, "step": 54000 }, { "grad_norm": 0.14984306693077087, "learning_rate": 2.6186272433223726e-05, "loss": 0.0131, "step": 54010 }, { "grad_norm": 0.15608972311019897, "learning_rate": 2.6168100837955943e-05, "loss": 0.0118, "step": 54020 }, { "grad_norm": 0.17731331288814545, "learning_rate": 2.6149933314907926e-05, "loss": 0.0201, "step": 54030 }, { "grad_norm": 0.24745380878448486, "learning_rate": 2.6131769867184e-05, "loss": 0.0174, "step": 54040 }, { "grad_norm": 0.2700579762458801, "learning_rate": 2.611361049788783e-05, "loss": 0.0153, "step": 54050 }, { "grad_norm": 0.24639229476451874, "learning_rate": 2.6095455210122292e-05, "loss": 0.0146, "step": 54060 }, { "grad_norm": 1.5356152057647705, "learning_rate": 2.6077304006989712e-05, "loss": 0.0118, "step": 54070 }, { "grad_norm": 0.25492215156555176, "learning_rate": 2.6059156891591562e-05, "loss": 0.014, "step": 54080 }, { "grad_norm": 0.2530164420604706, "learning_rate": 2.6041013867028718e-05, "loss": 0.0134, "step": 54090 }, { "grad_norm": 0.1856953352689743, "learning_rate": 2.6022874936401347e-05, "loss": 0.0124, "step": 54100 }, { "grad_norm": 0.1680210679769516, "learning_rate": 2.6004740102808832e-05, "loss": 0.0148, "step": 54110 }, { "grad_norm": 0.36962294578552246, "learning_rate": 2.598660936935e-05, "loss": 0.0121, "step": 54120 }, { "grad_norm": 0.17949765920639038, "learning_rate": 2.5968482739122845e-05, "loss": 0.0171, "step": 54130 }, { "grad_norm": 0.22110533714294434, "learning_rate": 2.595036021522472e-05, "loss": 0.0136, "step": 54140 }, { "grad_norm": 0.21317020058631897, "learning_rate": 2.5932241800752278e-05, "loss": 0.0136, "step": 54150 }, { "grad_norm": 0.22404402494430542, "learning_rate": 2.5914127498801453e-05, "loss": 0.0334, "step": 54160 }, { "grad_norm": 0.16472077369689941, "learning_rate": 2.5896017312467497e-05, "loss": 0.0173, "step": 54170 }, { "grad_norm": 0.2477312833070755, "learning_rate": 2.587791124484493e-05, "loss": 0.0142, "step": 54180 }, { "grad_norm": 0.7265496850013733, "learning_rate": 2.5859809299027615e-05, "loss": 0.0153, "step": 54190 }, { "grad_norm": 0.20430956780910492, "learning_rate": 2.5841711478108632e-05, "loss": 0.0223, "step": 54200 }, { "grad_norm": 0.2286343276500702, "learning_rate": 2.582361778518043e-05, "loss": 0.0175, "step": 54210 }, { "grad_norm": 0.27398794889450073, "learning_rate": 2.580552822333472e-05, "loss": 0.0161, "step": 54220 }, { "grad_norm": 0.2158748060464859, "learning_rate": 2.578744279566252e-05, "loss": 0.0141, "step": 54230 }, { "grad_norm": 0.2955623269081116, "learning_rate": 2.576936150525413e-05, "loss": 0.0202, "step": 54240 }, { "grad_norm": 0.14721395075321198, "learning_rate": 2.5751284355199168e-05, "loss": 0.0102, "step": 54250 }, { "grad_norm": 0.31519755721092224, "learning_rate": 2.573321134858646e-05, "loss": 0.0128, "step": 54260 }, { "grad_norm": 0.16768516600131989, "learning_rate": 2.5715142488504286e-05, "loss": 0.0112, "step": 54270 }, { "grad_norm": 0.16404087841510773, "learning_rate": 2.5697077778040042e-05, "loss": 0.0138, "step": 54280 }, { "grad_norm": 0.20390841364860535, "learning_rate": 2.5679017220280522e-05, "loss": 0.0265, "step": 54290 }, { "grad_norm": 0.2084565907716751, "learning_rate": 2.5660960818311796e-05, "loss": 0.0138, "step": 54300 }, { "grad_norm": 0.3269296884536743, "learning_rate": 2.564290857521915e-05, "loss": 0.0198, "step": 54310 }, { "grad_norm": 0.20723527669906616, "learning_rate": 2.5624860494087298e-05, "loss": 0.0125, "step": 54320 }, { "grad_norm": 0.16807018220424652, "learning_rate": 2.5606816578000115e-05, "loss": 0.0175, "step": 54330 }, { "grad_norm": 0.2336467206478119, "learning_rate": 2.558877683004082e-05, "loss": 0.0123, "step": 54340 }, { "grad_norm": 0.17407646775245667, "learning_rate": 2.557074125329192e-05, "loss": 0.0089, "step": 54350 }, { "grad_norm": 0.7632303237915039, "learning_rate": 2.5552709850835195e-05, "loss": 0.0136, "step": 54360 }, { "grad_norm": 0.1908697783946991, "learning_rate": 2.5534682625751738e-05, "loss": 0.0124, "step": 54370 }, { "grad_norm": 0.17044566571712494, "learning_rate": 2.551665958112186e-05, "loss": 0.0174, "step": 54380 }, { "grad_norm": 0.16162419319152832, "learning_rate": 2.549864072002527e-05, "loss": 0.0113, "step": 54390 }, { "grad_norm": 0.1735333353281021, "learning_rate": 2.5480626045540858e-05, "loss": 0.0126, "step": 54400 }, { "grad_norm": 1.6683294773101807, "learning_rate": 2.546261556074684e-05, "loss": 0.0151, "step": 54410 }, { "grad_norm": 0.23200169205665588, "learning_rate": 2.5444609268720726e-05, "loss": 0.0145, "step": 54420 }, { "grad_norm": 0.12401111423969269, "learning_rate": 2.5426607172539297e-05, "loss": 0.0148, "step": 54430 }, { "grad_norm": 0.1512974202632904, "learning_rate": 2.5408609275278617e-05, "loss": 0.0201, "step": 54440 }, { "grad_norm": 0.2319972664117813, "learning_rate": 2.5390615580014055e-05, "loss": 0.0159, "step": 54450 }, { "grad_norm": 0.2573959529399872, "learning_rate": 2.5372626089820207e-05, "loss": 0.0163, "step": 54460 }, { "grad_norm": 0.2539890706539154, "learning_rate": 2.5354640807770997e-05, "loss": 0.0123, "step": 54470 }, { "grad_norm": 9.401215553283691, "learning_rate": 2.5336659736939622e-05, "loss": 0.0133, "step": 54480 }, { "grad_norm": 0.2048739790916443, "learning_rate": 2.5318682880398554e-05, "loss": 0.0141, "step": 54490 }, { "grad_norm": 0.26822465658187866, "learning_rate": 2.530071024121956e-05, "loss": 0.0155, "step": 54500 }, { "grad_norm": 0.28817346692085266, "learning_rate": 2.5282741822473627e-05, "loss": 0.0147, "step": 54510 }, { "grad_norm": 0.2581118047237396, "learning_rate": 2.526477762723114e-05, "loss": 0.0131, "step": 54520 }, { "grad_norm": 0.2320556789636612, "learning_rate": 2.5246817658561618e-05, "loss": 0.0182, "step": 54530 }, { "grad_norm": 0.2975630462169647, "learning_rate": 2.5228861919533965e-05, "loss": 0.0127, "step": 54540 }, { "grad_norm": 0.21998487412929535, "learning_rate": 2.5210910413216326e-05, "loss": 0.0239, "step": 54550 }, { "grad_norm": 0.1856634020805359, "learning_rate": 2.5192963142676086e-05, "loss": 0.0091, "step": 54560 }, { "grad_norm": 0.198720321059227, "learning_rate": 2.517502011098001e-05, "loss": 0.0111, "step": 54570 }, { "grad_norm": 0.21233581006526947, "learning_rate": 2.5157081321193987e-05, "loss": 0.0136, "step": 54580 }, { "grad_norm": 0.16142627596855164, "learning_rate": 2.5139146776383356e-05, "loss": 0.0128, "step": 54590 }, { "grad_norm": 0.43088334798812866, "learning_rate": 2.5121216479612575e-05, "loss": 0.014, "step": 54600 }, { "grad_norm": 0.21876901388168335, "learning_rate": 2.510329043394546e-05, "loss": 0.0173, "step": 54610 }, { "grad_norm": 0.8961684107780457, "learning_rate": 2.508536864244508e-05, "loss": 0.0161, "step": 54620 }, { "grad_norm": 0.17585550248622894, "learning_rate": 2.5067451108173778e-05, "loss": 0.0135, "step": 54630 }, { "grad_norm": 0.2690485715866089, "learning_rate": 2.5049537834193204e-05, "loss": 0.0145, "step": 54640 }, { "grad_norm": 0.17119881510734558, "learning_rate": 2.5031628823564194e-05, "loss": 0.014, "step": 54650 }, { "grad_norm": 0.2745952308177948, "learning_rate": 2.5013724079346933e-05, "loss": 0.0132, "step": 54660 }, { "grad_norm": 0.16042746603488922, "learning_rate": 2.4995823604600854e-05, "loss": 0.011, "step": 54670 }, { "grad_norm": 1.0737048387527466, "learning_rate": 2.497792740238465e-05, "loss": 0.0167, "step": 54680 }, { "grad_norm": 0.22919386625289917, "learning_rate": 2.49600354757563e-05, "loss": 0.0205, "step": 54690 }, { "grad_norm": 0.3030923902988434, "learning_rate": 2.494214782777306e-05, "loss": 0.0183, "step": 54700 }, { "grad_norm": 0.8969524502754211, "learning_rate": 2.4924264461491386e-05, "loss": 0.0174, "step": 54710 }, { "grad_norm": 0.2863752841949463, "learning_rate": 2.4906385379967133e-05, "loss": 0.0179, "step": 54720 }, { "grad_norm": 0.25281256437301636, "learning_rate": 2.4888510586255285e-05, "loss": 0.0167, "step": 54730 }, { "grad_norm": 0.25996720790863037, "learning_rate": 2.487064008341018e-05, "loss": 0.0181, "step": 54740 }, { "grad_norm": 0.23687157034873962, "learning_rate": 2.4852773874485407e-05, "loss": 0.0138, "step": 54750 }, { "grad_norm": 0.24645249545574188, "learning_rate": 2.483491196253377e-05, "loss": 0.0132, "step": 54760 }, { "grad_norm": 0.20958060026168823, "learning_rate": 2.4817054350607443e-05, "loss": 0.0185, "step": 54770 }, { "grad_norm": 0.15524077415466309, "learning_rate": 2.4799201041757742e-05, "loss": 0.0118, "step": 54780 }, { "grad_norm": 0.27273839712142944, "learning_rate": 2.4781352039035373e-05, "loss": 0.0164, "step": 54790 }, { "grad_norm": 0.23079335689544678, "learning_rate": 2.4763507345490194e-05, "loss": 0.0129, "step": 54800 }, { "grad_norm": 0.23052607476711273, "learning_rate": 2.4745666964171386e-05, "loss": 0.0181, "step": 54810 }, { "grad_norm": 0.2556719183921814, "learning_rate": 2.4727830898127387e-05, "loss": 0.0135, "step": 54820 }, { "grad_norm": 0.19546249508857727, "learning_rate": 2.4709999150405895e-05, "loss": 0.0189, "step": 54830 }, { "grad_norm": 0.25068747997283936, "learning_rate": 2.469217172405388e-05, "loss": 0.0136, "step": 54840 }, { "grad_norm": 0.18394342064857483, "learning_rate": 2.4674348622117527e-05, "loss": 0.0113, "step": 54850 }, { "grad_norm": 0.22166472673416138, "learning_rate": 2.465652984764234e-05, "loss": 0.0124, "step": 54860 }, { "grad_norm": 0.17075525224208832, "learning_rate": 2.4638715403673056e-05, "loss": 0.0132, "step": 54870 }, { "grad_norm": 0.22229018807411194, "learning_rate": 2.4620905293253676e-05, "loss": 0.0119, "step": 54880 }, { "grad_norm": 0.9035730957984924, "learning_rate": 2.4603099519427463e-05, "loss": 0.0163, "step": 54890 }, { "grad_norm": 0.21346652507781982, "learning_rate": 2.458529808523695e-05, "loss": 0.0114, "step": 54900 }, { "grad_norm": 0.17058980464935303, "learning_rate": 2.456750099372387e-05, "loss": 0.0145, "step": 54910 }, { "grad_norm": 0.18897727131843567, "learning_rate": 2.4549708247929327e-05, "loss": 0.0171, "step": 54920 }, { "grad_norm": 0.2564786374568939, "learning_rate": 2.4531919850893554e-05, "loss": 0.0221, "step": 54930 }, { "grad_norm": 0.133762925863266, "learning_rate": 2.4514135805656125e-05, "loss": 0.0106, "step": 54940 }, { "grad_norm": 0.2714488208293915, "learning_rate": 2.449635611525587e-05, "loss": 0.0146, "step": 54950 }, { "grad_norm": 0.25367894768714905, "learning_rate": 2.447858078273079e-05, "loss": 0.0149, "step": 54960 }, { "grad_norm": 0.2810807526111603, "learning_rate": 2.4460809811118275e-05, "loss": 0.0229, "step": 54970 }, { "grad_norm": 0.1364702731370926, "learning_rate": 2.444304320345483e-05, "loss": 0.0133, "step": 54980 }, { "grad_norm": 0.2368326634168625, "learning_rate": 2.4425280962776354e-05, "loss": 0.0134, "step": 54990 }, { "grad_norm": 0.21115869283676147, "learning_rate": 2.4407523092117875e-05, "loss": 0.0141, "step": 55000 }, { "grad_norm": 0.39394423365592957, "learning_rate": 2.438976959451374e-05, "loss": 0.0125, "step": 55010 }, { "grad_norm": 0.2005806863307953, "learning_rate": 2.4372020472997565e-05, "loss": 0.0146, "step": 55020 }, { "grad_norm": 0.22896632552146912, "learning_rate": 2.4354275730602122e-05, "loss": 0.0141, "step": 55030 }, { "grad_norm": 0.1656307578086853, "learning_rate": 2.433653537035958e-05, "loss": 0.0151, "step": 55040 }, { "grad_norm": 0.27636268734931946, "learning_rate": 2.431879939530123e-05, "loss": 0.0144, "step": 55050 }, { "grad_norm": 0.20343971252441406, "learning_rate": 2.4301067808457684e-05, "loss": 0.0121, "step": 55060 }, { "grad_norm": 0.2147199511528015, "learning_rate": 2.428334061285878e-05, "loss": 0.0169, "step": 55070 }, { "grad_norm": 0.3616948425769806, "learning_rate": 2.426561781153361e-05, "loss": 0.0128, "step": 55080 }, { "grad_norm": 0.2093481868505478, "learning_rate": 2.424789940751052e-05, "loss": 0.0122, "step": 55090 }, { "grad_norm": 0.15202920138835907, "learning_rate": 2.423018540381712e-05, "loss": 0.0178, "step": 55100 }, { "grad_norm": 0.2030351310968399, "learning_rate": 2.421247580348021e-05, "loss": 0.0133, "step": 55110 }, { "grad_norm": 0.1975477635860443, "learning_rate": 2.4194770609525897e-05, "loss": 0.0112, "step": 55120 }, { "grad_norm": 0.1980598270893097, "learning_rate": 2.417706982497951e-05, "loss": 0.015, "step": 55130 }, { "grad_norm": 0.21433787047863007, "learning_rate": 2.415937345286563e-05, "loss": 0.013, "step": 55140 }, { "grad_norm": 0.2592848241329193, "learning_rate": 2.4141681496208087e-05, "loss": 0.0165, "step": 55150 }, { "grad_norm": 0.28735852241516113, "learning_rate": 2.4123993958029946e-05, "loss": 0.0163, "step": 55160 }, { "grad_norm": 0.22746926546096802, "learning_rate": 2.4106310841353548e-05, "loss": 0.0128, "step": 55170 }, { "grad_norm": 0.23436224460601807, "learning_rate": 2.4088632149200398e-05, "loss": 0.0173, "step": 55180 }, { "grad_norm": 0.23807024955749512, "learning_rate": 2.4070957884591367e-05, "loss": 0.0141, "step": 55190 }, { "grad_norm": 0.3861495852470398, "learning_rate": 2.4053288050546464e-05, "loss": 0.0167, "step": 55200 }, { "grad_norm": 0.1905253380537033, "learning_rate": 2.403562265008498e-05, "loss": 0.0131, "step": 55210 }, { "grad_norm": 0.2570320963859558, "learning_rate": 2.4017961686225483e-05, "loss": 0.0136, "step": 55220 }, { "grad_norm": 0.2700287699699402, "learning_rate": 2.400030516198568e-05, "loss": 0.0247, "step": 55230 }, { "grad_norm": 0.5374438762664795, "learning_rate": 2.3982653080382673e-05, "loss": 0.0155, "step": 55240 }, { "grad_norm": 0.2197490781545639, "learning_rate": 2.396500544443266e-05, "loss": 0.0129, "step": 55250 }, { "grad_norm": 0.17828461527824402, "learning_rate": 2.3947362257151156e-05, "loss": 0.0119, "step": 55260 }, { "grad_norm": 0.2749539017677307, "learning_rate": 2.39297235215529e-05, "loss": 0.0159, "step": 55270 }, { "grad_norm": 0.20089931786060333, "learning_rate": 2.3912089240651873e-05, "loss": 0.0164, "step": 55280 }, { "grad_norm": 0.13205184042453766, "learning_rate": 2.389445941746129e-05, "loss": 0.0172, "step": 55290 }, { "grad_norm": 0.3427870571613312, "learning_rate": 2.3876834054993625e-05, "loss": 0.0148, "step": 55300 }, { "grad_norm": 0.213277205824852, "learning_rate": 2.3859213156260522e-05, "loss": 0.0128, "step": 55310 }, { "grad_norm": 0.17212733626365662, "learning_rate": 2.3841596724272948e-05, "loss": 0.0127, "step": 55320 }, { "grad_norm": 0.15758024156093597, "learning_rate": 2.382398476204106e-05, "loss": 0.0159, "step": 55330 }, { "grad_norm": 0.2716313600540161, "learning_rate": 2.3806377272574254e-05, "loss": 0.0124, "step": 55340 }, { "grad_norm": 0.3060625493526459, "learning_rate": 2.3788774258881174e-05, "loss": 0.016, "step": 55350 }, { "grad_norm": 0.7237461805343628, "learning_rate": 2.37711757239697e-05, "loss": 0.0164, "step": 55360 }, { "grad_norm": 0.2507343292236328, "learning_rate": 2.3753581670846954e-05, "loss": 0.0181, "step": 55370 }, { "grad_norm": 0.12546326220035553, "learning_rate": 2.3735992102519216e-05, "loss": 0.0141, "step": 55380 }, { "grad_norm": 0.20585015416145325, "learning_rate": 2.371840702199215e-05, "loss": 0.0143, "step": 55390 }, { "grad_norm": 0.20084647834300995, "learning_rate": 2.3700826432270494e-05, "loss": 0.0114, "step": 55400 }, { "grad_norm": 0.3728393316268921, "learning_rate": 2.3683250336358326e-05, "loss": 0.0269, "step": 55410 }, { "grad_norm": 0.17214810848236084, "learning_rate": 2.3665678737258923e-05, "loss": 0.0203, "step": 55420 }, { "grad_norm": 0.22510887682437897, "learning_rate": 2.3648111637974745e-05, "loss": 0.0112, "step": 55430 }, { "grad_norm": 0.22387360036373138, "learning_rate": 2.36305490415076e-05, "loss": 0.0197, "step": 55440 }, { "grad_norm": 0.1800646185874939, "learning_rate": 2.3612990950858392e-05, "loss": 0.0117, "step": 55450 }, { "grad_norm": 0.15908952057361603, "learning_rate": 2.359543736902735e-05, "loss": 0.0106, "step": 55460 }, { "grad_norm": 0.24695339798927307, "learning_rate": 2.3577888299013896e-05, "loss": 0.0166, "step": 55470 }, { "grad_norm": 0.2480705827474594, "learning_rate": 2.3560343743816683e-05, "loss": 0.012, "step": 55480 }, { "grad_norm": 0.16748858988285065, "learning_rate": 2.354280370643362e-05, "loss": 0.0145, "step": 55490 }, { "grad_norm": 0.26394957304000854, "learning_rate": 2.3525268189861777e-05, "loss": 0.0167, "step": 55500 }, { "grad_norm": 0.2560245394706726, "learning_rate": 2.3507737197097513e-05, "loss": 0.0106, "step": 55510 }, { "grad_norm": 0.35142162442207336, "learning_rate": 2.34902107311364e-05, "loss": 0.0122, "step": 55520 }, { "grad_norm": 0.3511863052845001, "learning_rate": 2.3472688794973225e-05, "loss": 0.0123, "step": 55530 }, { "grad_norm": 0.1911543309688568, "learning_rate": 2.3455171391602016e-05, "loss": 0.0143, "step": 55540 }, { "grad_norm": 0.2875365614891052, "learning_rate": 2.343765852401601e-05, "loss": 0.0111, "step": 55550 }, { "grad_norm": 0.23989863693714142, "learning_rate": 2.342015019520768e-05, "loss": 0.0183, "step": 55560 }, { "grad_norm": 0.2040388435125351, "learning_rate": 2.3402646408168742e-05, "loss": 0.0153, "step": 55570 }, { "grad_norm": 0.25228655338287354, "learning_rate": 2.3385147165890074e-05, "loss": 0.0153, "step": 55580 }, { "grad_norm": 0.22323483228683472, "learning_rate": 2.336765247136184e-05, "loss": 0.0118, "step": 55590 }, { "grad_norm": 0.287201851606369, "learning_rate": 2.33501623275734e-05, "loss": 0.0152, "step": 55600 }, { "grad_norm": 0.10251301527023315, "learning_rate": 2.333267673751334e-05, "loss": 0.0089, "step": 55610 }, { "grad_norm": 7.472472667694092, "learning_rate": 2.33151957041695e-05, "loss": 0.0124, "step": 55620 }, { "grad_norm": 0.22155523300170898, "learning_rate": 2.329771923052884e-05, "loss": 0.0123, "step": 55630 }, { "grad_norm": 0.2559247314929962, "learning_rate": 2.3280247319577697e-05, "loss": 0.0125, "step": 55640 }, { "grad_norm": 0.26226016879081726, "learning_rate": 2.3262779974301473e-05, "loss": 0.0154, "step": 55650 }, { "grad_norm": 0.17285625636577606, "learning_rate": 2.3245317197684895e-05, "loss": 0.0121, "step": 55660 }, { "grad_norm": 0.20164267718791962, "learning_rate": 2.322785899271188e-05, "loss": 0.0139, "step": 55670 }, { "grad_norm": 2.1096129417419434, "learning_rate": 2.3210405362365507e-05, "loss": 0.0171, "step": 55680 }, { "grad_norm": 0.4379349648952484, "learning_rate": 2.31929563096282e-05, "loss": 0.016, "step": 55690 }, { "grad_norm": 0.38713499903678894, "learning_rate": 2.317551183748146e-05, "loss": 0.0143, "step": 55700 }, { "grad_norm": 0.19590196013450623, "learning_rate": 2.3158071948906103e-05, "loss": 0.0119, "step": 55710 }, { "grad_norm": 0.19618600606918335, "learning_rate": 2.314063664688212e-05, "loss": 0.0128, "step": 55720 }, { "grad_norm": 0.25174298882484436, "learning_rate": 2.3123205934388725e-05, "loss": 0.0121, "step": 55730 }, { "grad_norm": 0.2988602817058563, "learning_rate": 2.310577981440436e-05, "loss": 0.0224, "step": 55740 }, { "grad_norm": 0.1982811987400055, "learning_rate": 2.3088358289906665e-05, "loss": 0.0135, "step": 55750 }, { "grad_norm": 0.3155648112297058, "learning_rate": 2.307094136387252e-05, "loss": 0.0119, "step": 55760 }, { "grad_norm": 0.21624164283275604, "learning_rate": 2.305352903927796e-05, "loss": 0.0154, "step": 55770 }, { "grad_norm": 0.17028282582759857, "learning_rate": 2.303612131909831e-05, "loss": 0.012, "step": 55780 }, { "grad_norm": 0.37517616152763367, "learning_rate": 2.3018718206308054e-05, "loss": 0.0198, "step": 55790 }, { "grad_norm": 0.1562924087047577, "learning_rate": 2.3001319703880925e-05, "loss": 0.0135, "step": 55800 }, { "grad_norm": 0.19951725006103516, "learning_rate": 2.2983925814789835e-05, "loss": 0.0124, "step": 55810 }, { "grad_norm": 0.17399625480175018, "learning_rate": 2.2966536542006957e-05, "loss": 0.012, "step": 55820 }, { "grad_norm": 0.23011213541030884, "learning_rate": 2.294915188850358e-05, "loss": 0.0146, "step": 55830 }, { "grad_norm": 0.1843106746673584, "learning_rate": 2.2931771857250333e-05, "loss": 0.0135, "step": 55840 }, { "grad_norm": 0.25112685561180115, "learning_rate": 2.2914396451216946e-05, "loss": 0.0129, "step": 55850 }, { "grad_norm": 0.27775031328201294, "learning_rate": 2.2897025673372412e-05, "loss": 0.0106, "step": 55860 }, { "grad_norm": 0.1484326273202896, "learning_rate": 2.287965952668494e-05, "loss": 0.0127, "step": 55870 }, { "grad_norm": 0.18860742449760437, "learning_rate": 2.2862298014121873e-05, "loss": 0.0138, "step": 55880 }, { "grad_norm": 0.1856251060962677, "learning_rate": 2.28449411386499e-05, "loss": 0.0137, "step": 55890 }, { "grad_norm": 0.2470370978116989, "learning_rate": 2.282758890323477e-05, "loss": 0.0155, "step": 55900 }, { "grad_norm": 0.11112842708826065, "learning_rate": 2.2810241310841528e-05, "loss": 0.0132, "step": 55910 }, { "grad_norm": 0.2663952112197876, "learning_rate": 2.2792898364434412e-05, "loss": 0.0143, "step": 55920 }, { "grad_norm": 0.23586444556713104, "learning_rate": 2.2775560066976843e-05, "loss": 0.0144, "step": 55930 }, { "grad_norm": 0.18767690658569336, "learning_rate": 2.275822642143147e-05, "loss": 0.0162, "step": 55940 }, { "grad_norm": 0.16994695365428925, "learning_rate": 2.274089743076014e-05, "loss": 0.0121, "step": 55950 }, { "grad_norm": 0.4087406098842621, "learning_rate": 2.2723573097923917e-05, "loss": 0.0141, "step": 55960 }, { "grad_norm": 0.3250470757484436, "learning_rate": 2.2706253425883024e-05, "loss": 0.0111, "step": 55970 }, { "grad_norm": 0.24279269576072693, "learning_rate": 2.2688938417596933e-05, "loss": 0.0129, "step": 55980 }, { "grad_norm": 0.19072219729423523, "learning_rate": 2.2671628076024305e-05, "loss": 0.0142, "step": 55990 }, { "grad_norm": 0.15401697158813477, "learning_rate": 2.2654322404123008e-05, "loss": 0.0125, "step": 56000 }, { "grad_norm": 0.25346848368644714, "learning_rate": 2.2637021404850105e-05, "loss": 0.0108, "step": 56010 }, { "grad_norm": 0.29502931237220764, "learning_rate": 2.2619725081161885e-05, "loss": 0.0093, "step": 56020 }, { "grad_norm": 0.12383749336004257, "learning_rate": 2.260243343601376e-05, "loss": 0.0148, "step": 56030 }, { "grad_norm": 0.4838081896305084, "learning_rate": 2.2585146472360473e-05, "loss": 0.0155, "step": 56040 }, { "grad_norm": 0.2565600872039795, "learning_rate": 2.2567864193155834e-05, "loss": 0.0145, "step": 56050 }, { "grad_norm": 0.20783700048923492, "learning_rate": 2.255058660135294e-05, "loss": 0.0121, "step": 56060 }, { "grad_norm": 0.4364957809448242, "learning_rate": 2.253331369990407e-05, "loss": 0.0161, "step": 56070 }, { "grad_norm": 0.21860823035240173, "learning_rate": 2.251604549176063e-05, "loss": 0.0165, "step": 56080 }, { "grad_norm": 0.1662176251411438, "learning_rate": 2.249878197987337e-05, "loss": 0.0146, "step": 56090 }, { "grad_norm": 0.16859221458435059, "learning_rate": 2.2481523167192087e-05, "loss": 0.0113, "step": 56100 }, { "grad_norm": 0.28325775265693665, "learning_rate": 2.2464269056665867e-05, "loss": 0.0118, "step": 56110 }, { "grad_norm": 0.15200787782669067, "learning_rate": 2.2447019651242958e-05, "loss": 0.0164, "step": 56120 }, { "grad_norm": 0.27389296889305115, "learning_rate": 2.2429774953870818e-05, "loss": 0.0104, "step": 56130 }, { "grad_norm": 0.31023597717285156, "learning_rate": 2.241253496749611e-05, "loss": 0.0178, "step": 56140 }, { "grad_norm": 0.26999175548553467, "learning_rate": 2.2395299695064614e-05, "loss": 0.0103, "step": 56150 }, { "grad_norm": 0.22517119348049164, "learning_rate": 2.237806913952145e-05, "loss": 0.0132, "step": 56160 }, { "grad_norm": 0.2054581493139267, "learning_rate": 2.2360843303810798e-05, "loss": 0.0101, "step": 56170 }, { "grad_norm": 0.4011913537979126, "learning_rate": 2.2343622190876084e-05, "loss": 0.0125, "step": 56180 }, { "grad_norm": 0.19850315153598785, "learning_rate": 2.2326405803659935e-05, "loss": 0.0131, "step": 56190 }, { "grad_norm": 0.1972479671239853, "learning_rate": 2.230919414510416e-05, "loss": 0.0128, "step": 56200 }, { "grad_norm": 0.9589872360229492, "learning_rate": 2.229198721814976e-05, "loss": 0.0129, "step": 56210 }, { "grad_norm": 0.2024661749601364, "learning_rate": 2.2274785025736948e-05, "loss": 0.0119, "step": 56220 }, { "grad_norm": 0.26137351989746094, "learning_rate": 2.225758757080507e-05, "loss": 0.0144, "step": 56230 }, { "grad_norm": 0.3131663203239441, "learning_rate": 2.2240394856292723e-05, "loss": 0.013, "step": 56240 }, { "grad_norm": 0.2114236205816269, "learning_rate": 2.2223206885137664e-05, "loss": 0.0123, "step": 56250 }, { "grad_norm": 0.34414857625961304, "learning_rate": 2.2206023660276853e-05, "loss": 0.0127, "step": 56260 }, { "grad_norm": 0.14514844119548798, "learning_rate": 2.218884518464645e-05, "loss": 0.0136, "step": 56270 }, { "grad_norm": 0.2916734218597412, "learning_rate": 2.2171671461181732e-05, "loss": 0.0146, "step": 56280 }, { "grad_norm": 0.2307446300983429, "learning_rate": 2.2154502492817292e-05, "loss": 0.0147, "step": 56290 }, { "grad_norm": 0.1676570177078247, "learning_rate": 2.2137338282486782e-05, "loss": 0.0118, "step": 56300 }, { "grad_norm": 0.18825000524520874, "learning_rate": 2.2120178833123113e-05, "loss": 0.0113, "step": 56310 }, { "grad_norm": 0.25081610679626465, "learning_rate": 2.210302414765838e-05, "loss": 0.0142, "step": 56320 }, { "grad_norm": 0.19318029284477234, "learning_rate": 2.2085874229023808e-05, "loss": 0.0141, "step": 56330 }, { "grad_norm": 0.2363356500864029, "learning_rate": 2.2068729080149907e-05, "loss": 0.0132, "step": 56340 }, { "grad_norm": 0.19365718960762024, "learning_rate": 2.205158870396625e-05, "loss": 0.0128, "step": 56350 }, { "grad_norm": 0.1582014113664627, "learning_rate": 2.2034453103401732e-05, "loss": 0.0119, "step": 56360 }, { "grad_norm": 0.15035924315452576, "learning_rate": 2.20173222813843e-05, "loss": 0.0102, "step": 56370 }, { "grad_norm": 0.2678317427635193, "learning_rate": 2.200019624084116e-05, "loss": 0.0143, "step": 56380 }, { "grad_norm": 0.18207548558712006, "learning_rate": 2.1983074984698687e-05, "loss": 0.0121, "step": 56390 }, { "grad_norm": 0.21912966668605804, "learning_rate": 2.1965958515882433e-05, "loss": 0.0168, "step": 56400 }, { "grad_norm": 0.13448110222816467, "learning_rate": 2.1948846837317162e-05, "loss": 0.0112, "step": 56410 }, { "grad_norm": 0.264360249042511, "learning_rate": 2.1931739951926738e-05, "loss": 0.0123, "step": 56420 }, { "grad_norm": 0.3175358772277832, "learning_rate": 2.191463786263429e-05, "loss": 0.0153, "step": 56430 }, { "grad_norm": 0.2597081959247589, "learning_rate": 2.1897540572362095e-05, "loss": 0.0126, "step": 56440 }, { "grad_norm": 0.17355629801750183, "learning_rate": 2.1880448084031614e-05, "loss": 0.0128, "step": 56450 }, { "grad_norm": 0.2876336872577667, "learning_rate": 2.1863360400563482e-05, "loss": 0.0132, "step": 56460 }, { "grad_norm": 0.21444973349571228, "learning_rate": 2.184627752487754e-05, "loss": 0.0125, "step": 56470 }, { "grad_norm": 0.256390780210495, "learning_rate": 2.1829199459892725e-05, "loss": 0.0117, "step": 56480 }, { "grad_norm": 0.21950872242450714, "learning_rate": 2.1812126208527282e-05, "loss": 0.0125, "step": 56490 }, { "grad_norm": 0.23393593728542328, "learning_rate": 2.179505777369852e-05, "loss": 0.0176, "step": 56500 }, { "grad_norm": 0.22166740894317627, "learning_rate": 2.1777994158322974e-05, "loss": 0.0159, "step": 56510 }, { "grad_norm": 0.22987231612205505, "learning_rate": 2.176093536531637e-05, "loss": 0.0151, "step": 56520 }, { "grad_norm": 0.7725927829742432, "learning_rate": 2.1743881397593537e-05, "loss": 0.0137, "step": 56530 }, { "grad_norm": 0.35823917388916016, "learning_rate": 2.1726832258068595e-05, "loss": 0.0201, "step": 56540 }, { "grad_norm": 0.3213077783584595, "learning_rate": 2.170978794965472e-05, "loss": 0.015, "step": 56550 }, { "grad_norm": 0.15349456667900085, "learning_rate": 2.169274847526438e-05, "loss": 0.0129, "step": 56560 }, { "grad_norm": 0.2594589293003082, "learning_rate": 2.1675713837809103e-05, "loss": 0.0123, "step": 56570 }, { "grad_norm": 0.31363552808761597, "learning_rate": 2.1658684040199655e-05, "loss": 0.012, "step": 56580 }, { "grad_norm": 0.1761782169342041, "learning_rate": 2.1641659085345974e-05, "loss": 0.0101, "step": 56590 }, { "grad_norm": 0.20231039822101593, "learning_rate": 2.1624638976157154e-05, "loss": 0.0116, "step": 56600 }, { "grad_norm": 0.23923413455486298, "learning_rate": 2.1607623715541476e-05, "loss": 0.0183, "step": 56610 }, { "grad_norm": 0.19028806686401367, "learning_rate": 2.159061330640636e-05, "loss": 0.0109, "step": 56620 }, { "grad_norm": 0.19527721405029297, "learning_rate": 2.1573607751658425e-05, "loss": 0.012, "step": 56630 }, { "grad_norm": 0.8456212878227234, "learning_rate": 2.155660705420347e-05, "loss": 0.0151, "step": 56640 }, { "grad_norm": 0.3338630795478821, "learning_rate": 2.153961121694644e-05, "loss": 0.0158, "step": 56650 }, { "grad_norm": 0.19319963455200195, "learning_rate": 2.152262024279145e-05, "loss": 0.0238, "step": 56660 }, { "grad_norm": 0.31289342045783997, "learning_rate": 2.150563413464183e-05, "loss": 0.0158, "step": 56670 }, { "grad_norm": 0.10986454784870148, "learning_rate": 2.1488652895399974e-05, "loss": 0.012, "step": 56680 }, { "grad_norm": 0.22235627472400665, "learning_rate": 2.1471676527967587e-05, "loss": 0.0146, "step": 56690 }, { "grad_norm": 0.21278195083141327, "learning_rate": 2.1454705035245414e-05, "loss": 0.0154, "step": 56700 }, { "grad_norm": 1.5388818979263306, "learning_rate": 2.143773842013343e-05, "loss": 0.0126, "step": 56710 }, { "grad_norm": 0.17395399510860443, "learning_rate": 2.1420776685530796e-05, "loss": 0.0117, "step": 56720 }, { "grad_norm": 0.20790880918502808, "learning_rate": 2.1403819834335742e-05, "loss": 0.0094, "step": 56730 }, { "grad_norm": 0.1345568746328354, "learning_rate": 2.1386867869445808e-05, "loss": 0.0143, "step": 56740 }, { "grad_norm": 0.4664058983325958, "learning_rate": 2.1369920793757548e-05, "loss": 0.0178, "step": 56750 }, { "grad_norm": 0.18600232899188995, "learning_rate": 2.1352978610166828e-05, "loss": 0.0113, "step": 56760 }, { "grad_norm": 0.20673054456710815, "learning_rate": 2.1336041321568546e-05, "loss": 0.0113, "step": 56770 }, { "grad_norm": 0.22861935198307037, "learning_rate": 2.131910893085684e-05, "loss": 0.0141, "step": 56780 }, { "grad_norm": 0.21926449239253998, "learning_rate": 2.1302181440925012e-05, "loss": 0.0161, "step": 56790 }, { "grad_norm": 0.15076421201229095, "learning_rate": 2.128525885466546e-05, "loss": 0.0136, "step": 56800 }, { "grad_norm": 0.17671912908554077, "learning_rate": 2.1268341174969847e-05, "loss": 0.0184, "step": 56810 }, { "grad_norm": 0.2550899386405945, "learning_rate": 2.1251428404728894e-05, "loss": 0.0134, "step": 56820 }, { "grad_norm": 0.2781410813331604, "learning_rate": 2.123452054683256e-05, "loss": 0.0109, "step": 56830 }, { "grad_norm": 0.2341664880514145, "learning_rate": 2.1217617604169927e-05, "loss": 0.0202, "step": 56840 }, { "grad_norm": 0.24442516267299652, "learning_rate": 2.120071957962924e-05, "loss": 0.0145, "step": 56850 }, { "grad_norm": 0.4476759433746338, "learning_rate": 2.1183826476097917e-05, "loss": 0.0238, "step": 56860 }, { "grad_norm": 0.15059378743171692, "learning_rate": 2.116693829646254e-05, "loss": 0.0132, "step": 56870 }, { "grad_norm": 0.2961454689502716, "learning_rate": 2.1150055043608806e-05, "loss": 0.0109, "step": 56880 }, { "grad_norm": 0.32798975706100464, "learning_rate": 2.113317672042162e-05, "loss": 0.0142, "step": 56890 }, { "grad_norm": 0.16850817203521729, "learning_rate": 2.1116303329785025e-05, "loss": 0.0121, "step": 56900 }, { "grad_norm": 0.2536030113697052, "learning_rate": 2.109943487458222e-05, "loss": 0.0155, "step": 56910 }, { "grad_norm": 0.16877026855945587, "learning_rate": 2.1082571357695574e-05, "loss": 0.01, "step": 56920 }, { "grad_norm": 0.5723968744277954, "learning_rate": 2.1065712782006557e-05, "loss": 0.0146, "step": 56930 }, { "grad_norm": 0.1788696050643921, "learning_rate": 2.104885915039591e-05, "loss": 0.01, "step": 56940 }, { "grad_norm": 0.28894466161727905, "learning_rate": 2.103201046574338e-05, "loss": 0.0099, "step": 56950 }, { "grad_norm": 0.23153001070022583, "learning_rate": 2.1015166730928022e-05, "loss": 0.0123, "step": 56960 }, { "grad_norm": 0.3273821175098419, "learning_rate": 2.0998327948827912e-05, "loss": 0.0129, "step": 56970 }, { "grad_norm": 0.15856397151947021, "learning_rate": 2.0981494122320363e-05, "loss": 0.0124, "step": 56980 }, { "grad_norm": 0.20994409918785095, "learning_rate": 2.0964665254281822e-05, "loss": 0.0136, "step": 56990 }, { "grad_norm": 1.466927170753479, "learning_rate": 2.094784134758784e-05, "loss": 0.0111, "step": 57000 }, { "grad_norm": 0.7628477215766907, "learning_rate": 2.0931022405113226e-05, "loss": 0.0192, "step": 57010 }, { "grad_norm": 0.3691774904727936, "learning_rate": 2.091420842973183e-05, "loss": 0.0136, "step": 57020 }, { "grad_norm": 0.281061053276062, "learning_rate": 2.0897399424316715e-05, "loss": 0.014, "step": 57030 }, { "grad_norm": 0.19525863230228424, "learning_rate": 2.0880595391740078e-05, "loss": 0.014, "step": 57040 }, { "grad_norm": 0.24739989638328552, "learning_rate": 2.0863796334873277e-05, "loss": 0.0151, "step": 57050 }, { "grad_norm": 0.1080731451511383, "learning_rate": 2.08470022565868e-05, "loss": 0.0102, "step": 57060 }, { "grad_norm": 0.27391454577445984, "learning_rate": 2.0830213159750317e-05, "loss": 0.0097, "step": 57070 }, { "grad_norm": 0.1799994260072708, "learning_rate": 2.0813429047232596e-05, "loss": 0.0126, "step": 57080 }, { "grad_norm": 0.20167405903339386, "learning_rate": 2.0796649921901594e-05, "loss": 0.0134, "step": 57090 }, { "grad_norm": 0.1865749955177307, "learning_rate": 2.077987578662441e-05, "loss": 0.0112, "step": 57100 }, { "grad_norm": 0.18409255146980286, "learning_rate": 2.0763106644267277e-05, "loss": 0.0172, "step": 57110 }, { "grad_norm": 0.14819134771823883, "learning_rate": 2.0746342497695607e-05, "loss": 0.0112, "step": 57120 }, { "grad_norm": 0.20571927726268768, "learning_rate": 2.0729583349773886e-05, "loss": 0.0107, "step": 57130 }, { "grad_norm": 0.1994878053665161, "learning_rate": 2.0712829203365853e-05, "loss": 0.0096, "step": 57140 }, { "grad_norm": 0.19120973348617554, "learning_rate": 2.0696080061334267e-05, "loss": 0.0124, "step": 57150 }, { "grad_norm": 0.317583292722702, "learning_rate": 2.067933592654117e-05, "loss": 0.0114, "step": 57160 }, { "grad_norm": 0.20905208587646484, "learning_rate": 2.066259680184763e-05, "loss": 0.0116, "step": 57170 }, { "grad_norm": 0.20207221806049347, "learning_rate": 2.0645862690113908e-05, "loss": 0.0108, "step": 57180 }, { "grad_norm": 0.17761969566345215, "learning_rate": 2.0629133594199436e-05, "loss": 0.0155, "step": 57190 }, { "grad_norm": 0.14473135769367218, "learning_rate": 2.0612409516962704e-05, "loss": 0.0112, "step": 57200 }, { "grad_norm": 0.28053250908851624, "learning_rate": 2.0595690461261467e-05, "loss": 0.0106, "step": 57210 }, { "grad_norm": 0.19038942456245422, "learning_rate": 2.0578976429952503e-05, "loss": 0.0125, "step": 57220 }, { "grad_norm": 0.17889364063739777, "learning_rate": 2.0562267425891802e-05, "loss": 0.0148, "step": 57230 }, { "grad_norm": 0.22992804646492004, "learning_rate": 2.0545563451934467e-05, "loss": 0.014, "step": 57240 }, { "grad_norm": 0.25638601183891296, "learning_rate": 2.0528864510934764e-05, "loss": 0.0144, "step": 57250 }, { "grad_norm": 0.7709656953811646, "learning_rate": 2.0512170605746096e-05, "loss": 0.0203, "step": 57260 }, { "grad_norm": 0.17924301326274872, "learning_rate": 2.0495481739220963e-05, "loss": 0.0122, "step": 57270 }, { "grad_norm": 0.4434738755226135, "learning_rate": 2.0478797914211045e-05, "loss": 0.0142, "step": 57280 }, { "grad_norm": 0.1534392535686493, "learning_rate": 2.046211913356716e-05, "loss": 0.012, "step": 57290 }, { "grad_norm": 0.2702638506889343, "learning_rate": 2.0445445400139247e-05, "loss": 0.0167, "step": 57300 }, { "grad_norm": 0.2373490333557129, "learning_rate": 2.0428776716776405e-05, "loss": 0.0112, "step": 57310 }, { "grad_norm": 0.1697423756122589, "learning_rate": 2.0412113086326856e-05, "loss": 0.0127, "step": 57320 }, { "grad_norm": 0.1844865083694458, "learning_rate": 2.0395454511637918e-05, "loss": 0.0114, "step": 57330 }, { "grad_norm": 0.2367851436138153, "learning_rate": 2.037880099555616e-05, "loss": 0.0131, "step": 57340 }, { "grad_norm": 0.34163761138916016, "learning_rate": 2.0362152540927144e-05, "loss": 0.0151, "step": 57350 }, { "grad_norm": 0.28584396839141846, "learning_rate": 2.0345509150595666e-05, "loss": 0.0259, "step": 57360 }, { "grad_norm": 0.48539918661117554, "learning_rate": 2.0328870827405617e-05, "loss": 0.0161, "step": 57370 }, { "grad_norm": 0.21101827919483185, "learning_rate": 2.0312237574200043e-05, "loss": 0.0124, "step": 57380 }, { "grad_norm": 0.24510125815868378, "learning_rate": 2.029560939382112e-05, "loss": 0.013, "step": 57390 }, { "grad_norm": 0.2428520917892456, "learning_rate": 2.0278986289110097e-05, "loss": 0.0125, "step": 57400 }, { "grad_norm": 0.1686897873878479, "learning_rate": 2.0262368262907484e-05, "loss": 0.0117, "step": 57410 }, { "grad_norm": 0.30469250679016113, "learning_rate": 2.024575531805279e-05, "loss": 0.0145, "step": 57420 }, { "grad_norm": 0.266498863697052, "learning_rate": 2.0229147457384735e-05, "loss": 0.0099, "step": 57430 }, { "grad_norm": 0.34891489148139954, "learning_rate": 2.0212544683741157e-05, "loss": 0.0218, "step": 57440 }, { "grad_norm": 0.22400978207588196, "learning_rate": 2.0195946999958976e-05, "loss": 0.0225, "step": 57450 }, { "grad_norm": 0.20307041704654694, "learning_rate": 2.017935440887434e-05, "loss": 0.014, "step": 57460 }, { "grad_norm": 0.1177489385008812, "learning_rate": 2.0162766913322423e-05, "loss": 0.0102, "step": 57470 }, { "grad_norm": 0.18025971949100494, "learning_rate": 2.0146184516137588e-05, "loss": 0.0159, "step": 57480 }, { "grad_norm": 0.20169304311275482, "learning_rate": 2.012960722015332e-05, "loss": 0.0131, "step": 57490 }, { "grad_norm": 0.18688945472240448, "learning_rate": 2.0113035028202214e-05, "loss": 0.0152, "step": 57500 }, { "grad_norm": 0.2691194415092468, "learning_rate": 2.009646794311602e-05, "loss": 0.0159, "step": 57510 }, { "grad_norm": 0.20034374296665192, "learning_rate": 2.007990596772559e-05, "loss": 0.0146, "step": 57520 }, { "grad_norm": 0.28680419921875, "learning_rate": 2.0063349104860923e-05, "loss": 0.0146, "step": 57530 }, { "grad_norm": 0.2136138528585434, "learning_rate": 2.0046797357351116e-05, "loss": 0.0191, "step": 57540 }, { "grad_norm": 0.1638396829366684, "learning_rate": 2.0030250728024412e-05, "loss": 0.009, "step": 57550 }, { "grad_norm": 0.3873176872730255, "learning_rate": 2.001370921970819e-05, "loss": 0.0111, "step": 57560 }, { "grad_norm": 0.25391435623168945, "learning_rate": 1.9997172835228932e-05, "loss": 0.0132, "step": 57570 }, { "grad_norm": 0.31635719537734985, "learning_rate": 1.9980641577412262e-05, "loss": 0.0153, "step": 57580 }, { "grad_norm": 0.22321705520153046, "learning_rate": 1.9964115449082925e-05, "loss": 0.0137, "step": 57590 }, { "grad_norm": 0.1855694055557251, "learning_rate": 1.9947594453064742e-05, "loss": 0.0139, "step": 57600 }, { "grad_norm": 0.17320924997329712, "learning_rate": 1.9931078592180774e-05, "loss": 0.0135, "step": 57610 }, { "grad_norm": 0.16562391817569733, "learning_rate": 1.9914567869253065e-05, "loss": 0.0133, "step": 57620 }, { "grad_norm": 0.3031803071498871, "learning_rate": 1.989806228710287e-05, "loss": 0.0181, "step": 57630 }, { "grad_norm": 0.1892576366662979, "learning_rate": 1.9881561848550555e-05, "loss": 0.0124, "step": 57640 }, { "grad_norm": 0.16372723877429962, "learning_rate": 1.9865066556415544e-05, "loss": 0.0153, "step": 57650 }, { "grad_norm": 0.2005675733089447, "learning_rate": 1.98485764135165e-05, "loss": 0.0103, "step": 57660 }, { "grad_norm": 0.1527862846851349, "learning_rate": 1.983209142267109e-05, "loss": 0.0125, "step": 57670 }, { "grad_norm": 0.2523372173309326, "learning_rate": 1.9815611586696165e-05, "loss": 0.0101, "step": 57680 }, { "grad_norm": 0.9151371717453003, "learning_rate": 1.9799136908407667e-05, "loss": 0.0136, "step": 57690 }, { "grad_norm": 0.1649319976568222, "learning_rate": 1.9782667390620678e-05, "loss": 0.0141, "step": 57700 }, { "grad_norm": 0.22994665801525116, "learning_rate": 1.976620303614939e-05, "loss": 0.0113, "step": 57710 }, { "grad_norm": 0.1636018455028534, "learning_rate": 1.9749743847807108e-05, "loss": 0.0156, "step": 57720 }, { "grad_norm": 0.23454320430755615, "learning_rate": 1.9733289828406272e-05, "loss": 0.0112, "step": 57730 }, { "grad_norm": 0.2851254940032959, "learning_rate": 1.9716840980758382e-05, "loss": 0.0135, "step": 57740 }, { "grad_norm": 0.1434909701347351, "learning_rate": 1.9700397307674134e-05, "loss": 0.0123, "step": 57750 }, { "grad_norm": 0.17200352251529694, "learning_rate": 1.968395881196328e-05, "loss": 0.012, "step": 57760 }, { "grad_norm": 0.30758678913116455, "learning_rate": 1.966752549643473e-05, "loss": 0.0134, "step": 57770 }, { "grad_norm": 0.25146251916885376, "learning_rate": 1.965109736389647e-05, "loss": 0.0127, "step": 57780 }, { "grad_norm": 0.2158893495798111, "learning_rate": 1.9634674417155645e-05, "loss": 0.0141, "step": 57790 }, { "grad_norm": 0.39840391278266907, "learning_rate": 1.9618256659018434e-05, "loss": 0.0232, "step": 57800 }, { "grad_norm": 0.25939223170280457, "learning_rate": 1.9601844092290257e-05, "loss": 0.0117, "step": 57810 }, { "grad_norm": 0.19057296216487885, "learning_rate": 1.9585436719775512e-05, "loss": 0.0092, "step": 57820 }, { "grad_norm": 0.2499939650297165, "learning_rate": 1.9569034544277793e-05, "loss": 0.0121, "step": 57830 }, { "grad_norm": 0.22795474529266357, "learning_rate": 1.9552637568599798e-05, "loss": 0.0103, "step": 57840 }, { "grad_norm": 0.16906657814979553, "learning_rate": 1.953624579554327e-05, "loss": 0.0111, "step": 57850 }, { "grad_norm": 0.23687563836574554, "learning_rate": 1.951985922790918e-05, "loss": 0.0139, "step": 57860 }, { "grad_norm": 0.20058825612068176, "learning_rate": 1.9503477868497505e-05, "loss": 0.0121, "step": 57870 }, { "grad_norm": 0.1682852804660797, "learning_rate": 1.9487101720107375e-05, "loss": 0.0122, "step": 57880 }, { "grad_norm": 1.6623231172561646, "learning_rate": 1.9470730785537032e-05, "loss": 0.0187, "step": 57890 }, { "grad_norm": 0.15201427042484283, "learning_rate": 1.9454365067583823e-05, "loss": 0.0109, "step": 57900 }, { "grad_norm": 0.204877108335495, "learning_rate": 1.9438004569044215e-05, "loss": 0.0088, "step": 57910 }, { "grad_norm": 0.1454123556613922, "learning_rate": 1.9421649292713724e-05, "loss": 0.0255, "step": 57920 }, { "grad_norm": 0.2839978337287903, "learning_rate": 1.9405299241387076e-05, "loss": 0.014, "step": 57930 }, { "grad_norm": 0.2903919816017151, "learning_rate": 1.9388954417858007e-05, "loss": 0.0123, "step": 57940 }, { "grad_norm": 0.21071653068065643, "learning_rate": 1.9372614824919417e-05, "loss": 0.0129, "step": 57950 }, { "grad_norm": 0.25966352224349976, "learning_rate": 1.9356280465363284e-05, "loss": 0.0143, "step": 57960 }, { "grad_norm": 0.24621786177158356, "learning_rate": 1.9339951341980723e-05, "loss": 0.0119, "step": 57970 }, { "grad_norm": 0.15861466526985168, "learning_rate": 1.9323627457561916e-05, "loss": 0.0122, "step": 57980 }, { "grad_norm": 0.19045451283454895, "learning_rate": 1.9307308814896198e-05, "loss": 0.0115, "step": 57990 }, { "grad_norm": 0.14996755123138428, "learning_rate": 1.9290995416771935e-05, "loss": 0.0101, "step": 58000 }, { "grad_norm": 0.21380378305912018, "learning_rate": 1.9274687265976665e-05, "loss": 0.0112, "step": 58010 }, { "grad_norm": 0.42954403162002563, "learning_rate": 1.9258384365297e-05, "loss": 0.0125, "step": 58020 }, { "grad_norm": 0.3242936134338379, "learning_rate": 1.924208671751866e-05, "loss": 0.0129, "step": 58030 }, { "grad_norm": 0.18197119235992432, "learning_rate": 1.9225794325426492e-05, "loss": 0.0136, "step": 58040 }, { "grad_norm": 0.21192248165607452, "learning_rate": 1.920950719180436e-05, "loss": 0.0141, "step": 58050 }, { "grad_norm": 0.15943275392055511, "learning_rate": 1.919322531943536e-05, "loss": 0.0112, "step": 58060 }, { "grad_norm": 0.1343510001897812, "learning_rate": 1.917694871110157e-05, "loss": 0.0118, "step": 58070 }, { "grad_norm": 0.50112384557724, "learning_rate": 1.9160677369584234e-05, "loss": 0.0153, "step": 58080 }, { "grad_norm": 0.20486459136009216, "learning_rate": 1.9144411297663694e-05, "loss": 0.0108, "step": 58090 }, { "grad_norm": 0.6213868856430054, "learning_rate": 1.9128150498119328e-05, "loss": 0.0196, "step": 58100 }, { "grad_norm": 0.18224868178367615, "learning_rate": 1.9111894973729726e-05, "loss": 0.0117, "step": 58110 }, { "grad_norm": 0.24043607711791992, "learning_rate": 1.9095644727272454e-05, "loss": 0.0142, "step": 58120 }, { "grad_norm": 0.2383323609828949, "learning_rate": 1.907939976152429e-05, "loss": 0.0122, "step": 58130 }, { "grad_norm": 0.4236668050289154, "learning_rate": 1.906316007926101e-05, "loss": 0.0155, "step": 58140 }, { "grad_norm": 0.15040265023708344, "learning_rate": 1.904692568325755e-05, "loss": 0.0101, "step": 58150 }, { "grad_norm": 0.2533336579799652, "learning_rate": 1.9030696576287925e-05, "loss": 0.0151, "step": 58160 }, { "grad_norm": 7.543585777282715, "learning_rate": 1.9014472761125242e-05, "loss": 0.0109, "step": 58170 }, { "grad_norm": 0.11603830009698868, "learning_rate": 1.899825424054172e-05, "loss": 0.012, "step": 58180 }, { "grad_norm": 0.30329498648643494, "learning_rate": 1.898204101730863e-05, "loss": 0.0125, "step": 58190 }, { "grad_norm": 0.2551495134830475, "learning_rate": 1.8965833094196394e-05, "loss": 0.0199, "step": 58200 }, { "grad_norm": 0.18570350110530853, "learning_rate": 1.8949630473974495e-05, "loss": 0.0173, "step": 58210 }, { "grad_norm": 0.23036976158618927, "learning_rate": 1.8933433159411517e-05, "loss": 0.0145, "step": 58220 }, { "grad_norm": 0.12884752452373505, "learning_rate": 1.891724115327514e-05, "loss": 0.0111, "step": 58230 }, { "grad_norm": 0.14834076166152954, "learning_rate": 1.8901054458332152e-05, "loss": 0.0157, "step": 58240 }, { "grad_norm": 0.1832190603017807, "learning_rate": 1.8884873077348364e-05, "loss": 0.0156, "step": 58250 }, { "grad_norm": 0.22597160935401917, "learning_rate": 1.88686970130888e-05, "loss": 0.0162, "step": 58260 }, { "grad_norm": 0.1858009248971939, "learning_rate": 1.8852526268317456e-05, "loss": 0.0133, "step": 58270 }, { "grad_norm": 0.18796294927597046, "learning_rate": 1.883636084579749e-05, "loss": 0.0127, "step": 58280 }, { "grad_norm": 0.2665784955024719, "learning_rate": 1.8820200748291145e-05, "loss": 0.0114, "step": 58290 }, { "grad_norm": 0.17325136065483093, "learning_rate": 1.8804045978559686e-05, "loss": 0.0152, "step": 58300 }, { "grad_norm": 0.44175395369529724, "learning_rate": 1.8787896539363594e-05, "loss": 0.015, "step": 58310 }, { "grad_norm": 0.17956328392028809, "learning_rate": 1.8771752433462298e-05, "loss": 0.0113, "step": 58320 }, { "grad_norm": 0.417766809463501, "learning_rate": 1.8755613663614447e-05, "loss": 0.0124, "step": 58330 }, { "grad_norm": 0.2598809003829956, "learning_rate": 1.873948023257767e-05, "loss": 0.0127, "step": 58340 }, { "grad_norm": 0.1871032565832138, "learning_rate": 1.8723352143108747e-05, "loss": 0.0117, "step": 58350 }, { "grad_norm": 0.16999903321266174, "learning_rate": 1.870722939796352e-05, "loss": 0.0096, "step": 58360 }, { "grad_norm": 0.3401132822036743, "learning_rate": 1.869111199989693e-05, "loss": 0.0134, "step": 58370 }, { "grad_norm": 0.23059497773647308, "learning_rate": 1.8674999951663018e-05, "loss": 0.0132, "step": 58380 }, { "grad_norm": 0.17366530001163483, "learning_rate": 1.8658893256014857e-05, "loss": 0.0183, "step": 58390 }, { "grad_norm": 0.10794718563556671, "learning_rate": 1.8642791915704655e-05, "loss": 0.0117, "step": 58400 }, { "grad_norm": 0.2526891231536865, "learning_rate": 1.8626695933483697e-05, "loss": 0.0133, "step": 58410 }, { "grad_norm": 0.17905865609645844, "learning_rate": 1.8610605312102335e-05, "loss": 0.0116, "step": 58420 }, { "grad_norm": 0.18978658318519592, "learning_rate": 1.8594520054310032e-05, "loss": 0.0162, "step": 58430 }, { "grad_norm": 0.2719230353832245, "learning_rate": 1.8578440162855326e-05, "loss": 0.0138, "step": 58440 }, { "grad_norm": 0.17500707507133484, "learning_rate": 1.8562365640485774e-05, "loss": 0.0158, "step": 58450 }, { "grad_norm": 0.21854671835899353, "learning_rate": 1.854629648994815e-05, "loss": 0.0122, "step": 58460 }, { "grad_norm": 0.17021353542804718, "learning_rate": 1.8530232713988183e-05, "loss": 0.0106, "step": 58470 }, { "grad_norm": 0.2594446837902069, "learning_rate": 1.851417431535074e-05, "loss": 0.0103, "step": 58480 }, { "grad_norm": 0.29167163372039795, "learning_rate": 1.8498121296779785e-05, "loss": 0.0195, "step": 58490 }, { "grad_norm": 0.26328346133232117, "learning_rate": 1.848207366101829e-05, "loss": 0.0149, "step": 58500 }, { "grad_norm": 0.16491301357746124, "learning_rate": 1.8466031410808422e-05, "loss": 0.012, "step": 58510 }, { "grad_norm": 0.22348766028881073, "learning_rate": 1.844999454889129e-05, "loss": 0.0157, "step": 58520 }, { "grad_norm": 0.2322668582201004, "learning_rate": 1.8433963078007237e-05, "loss": 0.0147, "step": 58530 }, { "grad_norm": 0.18672972917556763, "learning_rate": 1.841793700089554e-05, "loss": 0.0138, "step": 58540 }, { "grad_norm": 2.1382803916931152, "learning_rate": 1.8401916320294644e-05, "loss": 0.0135, "step": 58550 }, { "grad_norm": 0.2516520321369171, "learning_rate": 1.838590103894205e-05, "loss": 0.0112, "step": 58560 }, { "grad_norm": 0.24226030707359314, "learning_rate": 1.8369891159574292e-05, "loss": 0.0129, "step": 58570 }, { "grad_norm": 0.1543715000152588, "learning_rate": 1.835388668492708e-05, "loss": 0.0158, "step": 58580 }, { "grad_norm": 0.16868825256824493, "learning_rate": 1.8337887617735095e-05, "loss": 0.0109, "step": 58590 }, { "grad_norm": 0.6171187162399292, "learning_rate": 1.8321893960732157e-05, "loss": 0.0133, "step": 58600 }, { "grad_norm": 0.26845329999923706, "learning_rate": 1.8305905716651138e-05, "loss": 0.0172, "step": 58610 }, { "grad_norm": 0.33662092685699463, "learning_rate": 1.8289922888224e-05, "loss": 0.012, "step": 58620 }, { "grad_norm": 0.38730737566947937, "learning_rate": 1.8273945478181765e-05, "loss": 0.0111, "step": 58630 }, { "grad_norm": 0.34464210271835327, "learning_rate": 1.8257973489254558e-05, "loss": 0.0113, "step": 58640 }, { "grad_norm": 0.22939401865005493, "learning_rate": 1.824200692417152e-05, "loss": 0.0136, "step": 58650 }, { "grad_norm": 0.2958913743495941, "learning_rate": 1.8226045785660912e-05, "loss": 0.0121, "step": 58660 }, { "grad_norm": 0.24810341000556946, "learning_rate": 1.821009007645006e-05, "loss": 0.0137, "step": 58670 }, { "grad_norm": 0.1546589434146881, "learning_rate": 1.8194139799265357e-05, "loss": 0.0094, "step": 58680 }, { "grad_norm": 0.2473984956741333, "learning_rate": 1.8178194956832295e-05, "loss": 0.0164, "step": 58690 }, { "grad_norm": 0.8858705163002014, "learning_rate": 1.8162255551875346e-05, "loss": 0.0171, "step": 58700 }, { "grad_norm": 0.18419599533081055, "learning_rate": 1.8146321587118194e-05, "loss": 0.0171, "step": 58710 }, { "grad_norm": 0.21826264262199402, "learning_rate": 1.8130393065283448e-05, "loss": 0.0114, "step": 58720 }, { "grad_norm": 0.2902563214302063, "learning_rate": 1.8114469989092925e-05, "loss": 0.0163, "step": 58730 }, { "grad_norm": 0.2539586126804352, "learning_rate": 1.8098552361267397e-05, "loss": 0.0149, "step": 58740 }, { "grad_norm": 0.1861601024866104, "learning_rate": 1.8082640184526763e-05, "loss": 0.0101, "step": 58750 }, { "grad_norm": 0.728585958480835, "learning_rate": 1.806673346158999e-05, "loss": 0.0125, "step": 58760 }, { "grad_norm": 0.19269293546676636, "learning_rate": 1.8050832195175067e-05, "loss": 0.0154, "step": 58770 }, { "grad_norm": 0.4378443658351898, "learning_rate": 1.8034936387999136e-05, "loss": 0.0153, "step": 58780 }, { "grad_norm": 0.15582942962646484, "learning_rate": 1.8019046042778315e-05, "loss": 0.0116, "step": 58790 }, { "grad_norm": 0.16328252851963043, "learning_rate": 1.800316116222785e-05, "loss": 0.0121, "step": 58800 }, { "grad_norm": 0.20017875730991364, "learning_rate": 1.7987281749062018e-05, "loss": 0.0231, "step": 58810 }, { "grad_norm": 0.1452527791261673, "learning_rate": 1.7971407805994195e-05, "loss": 0.0201, "step": 58820 }, { "grad_norm": 0.15383918583393097, "learning_rate": 1.7955539335736787e-05, "loss": 0.0118, "step": 58830 }, { "grad_norm": 0.17069821059703827, "learning_rate": 1.7939676341001304e-05, "loss": 0.0157, "step": 58840 }, { "grad_norm": 0.20782771706581116, "learning_rate": 1.7923818824498275e-05, "loss": 0.0124, "step": 58850 }, { "grad_norm": 0.33766183257102966, "learning_rate": 1.7907966788937315e-05, "loss": 0.0139, "step": 58860 }, { "grad_norm": 0.20614568889141083, "learning_rate": 1.7892120237027116e-05, "loss": 0.0145, "step": 58870 }, { "grad_norm": 0.26602235436439514, "learning_rate": 1.7876279171475413e-05, "loss": 0.0132, "step": 58880 }, { "grad_norm": 0.5144856572151184, "learning_rate": 1.7860443594989028e-05, "loss": 0.0137, "step": 58890 }, { "grad_norm": 0.2070867270231247, "learning_rate": 1.784461351027379e-05, "loss": 0.0114, "step": 58900 }, { "grad_norm": 0.14953267574310303, "learning_rate": 1.7828788920034677e-05, "loss": 0.0169, "step": 58910 }, { "grad_norm": 0.2920539677143097, "learning_rate": 1.7812969826975623e-05, "loss": 0.0151, "step": 58920 }, { "grad_norm": 3.5989186763763428, "learning_rate": 1.7797156233799738e-05, "loss": 0.0192, "step": 58930 }, { "grad_norm": 0.20085187256336212, "learning_rate": 1.7781348143209094e-05, "loss": 0.0148, "step": 58940 }, { "grad_norm": 0.21856316924095154, "learning_rate": 1.7765545557904862e-05, "loss": 0.0157, "step": 58950 }, { "grad_norm": 0.3212288022041321, "learning_rate": 1.7749748480587302e-05, "loss": 0.0154, "step": 58960 }, { "grad_norm": 0.24619658291339874, "learning_rate": 1.773395691395564e-05, "loss": 0.0124, "step": 58970 }, { "grad_norm": 0.8051977753639221, "learning_rate": 1.7718170860708305e-05, "loss": 0.0125, "step": 58980 }, { "grad_norm": 0.19620081782341003, "learning_rate": 1.770239032354264e-05, "loss": 0.0122, "step": 58990 }, { "grad_norm": 0.24184615910053253, "learning_rate": 1.7686615305155124e-05, "loss": 0.0184, "step": 59000 }, { "grad_norm": 0.1787133812904358, "learning_rate": 1.767084580824128e-05, "loss": 0.0127, "step": 59010 }, { "grad_norm": 0.43310457468032837, "learning_rate": 1.765508183549569e-05, "loss": 0.0168, "step": 59020 }, { "grad_norm": 0.21195930242538452, "learning_rate": 1.763932338961199e-05, "loss": 0.013, "step": 59030 }, { "grad_norm": 0.4949565827846527, "learning_rate": 1.762357047328284e-05, "loss": 0.0124, "step": 59040 }, { "grad_norm": 0.35344383120536804, "learning_rate": 1.7607823089199997e-05, "loss": 0.0164, "step": 59050 }, { "grad_norm": 0.18799801170825958, "learning_rate": 1.7592081240054265e-05, "loss": 0.015, "step": 59060 }, { "grad_norm": 0.1615419089794159, "learning_rate": 1.7576344928535488e-05, "loss": 0.0126, "step": 59070 }, { "grad_norm": 0.1703656017780304, "learning_rate": 1.7560614157332573e-05, "loss": 0.0163, "step": 59080 }, { "grad_norm": 0.1813991665840149, "learning_rate": 1.7544888929133495e-05, "loss": 0.015, "step": 59090 }, { "grad_norm": 0.28499501943588257, "learning_rate": 1.752916924662522e-05, "loss": 0.0161, "step": 59100 }, { "grad_norm": 0.11852412670850754, "learning_rate": 1.751345511249387e-05, "loss": 0.0126, "step": 59110 }, { "grad_norm": 0.23738816380500793, "learning_rate": 1.749774652942452e-05, "loss": 0.0133, "step": 59120 }, { "grad_norm": 0.5284872651100159, "learning_rate": 1.748204350010135e-05, "loss": 0.0164, "step": 59130 }, { "grad_norm": 0.23793505132198334, "learning_rate": 1.7466346027207574e-05, "loss": 0.01, "step": 59140 }, { "grad_norm": 0.16870063543319702, "learning_rate": 1.745065411342547e-05, "loss": 0.0147, "step": 59150 }, { "grad_norm": 0.7102389335632324, "learning_rate": 1.7434967761436366e-05, "loss": 0.0159, "step": 59160 }, { "grad_norm": 0.253239244222641, "learning_rate": 1.741928697392058e-05, "loss": 0.0108, "step": 59170 }, { "grad_norm": 0.23853570222854614, "learning_rate": 1.7403611753557597e-05, "loss": 0.0113, "step": 59180 }, { "grad_norm": 2.4464690685272217, "learning_rate": 1.738794210302584e-05, "loss": 0.0092, "step": 59190 }, { "grad_norm": 0.18117620050907135, "learning_rate": 1.7372278025002837e-05, "loss": 0.0145, "step": 59200 }, { "grad_norm": 0.26066991686820984, "learning_rate": 1.7356619522165164e-05, "loss": 0.014, "step": 59210 }, { "grad_norm": 0.15487806499004364, "learning_rate": 1.7340966597188378e-05, "loss": 0.0114, "step": 59220 }, { "grad_norm": 0.13920535147190094, "learning_rate": 1.732531925274722e-05, "loss": 0.0138, "step": 59230 }, { "grad_norm": 0.23960644006729126, "learning_rate": 1.7309677491515318e-05, "loss": 0.0111, "step": 59240 }, { "grad_norm": 0.20495851337909698, "learning_rate": 1.7294041316165455e-05, "loss": 0.0108, "step": 59250 }, { "grad_norm": 0.43017029762268066, "learning_rate": 1.727841072936942e-05, "loss": 0.0145, "step": 59260 }, { "grad_norm": 0.19397146999835968, "learning_rate": 1.7262785733798058e-05, "loss": 0.0111, "step": 59270 }, { "grad_norm": 0.2245008647441864, "learning_rate": 1.7247166332121246e-05, "loss": 0.0127, "step": 59280 }, { "grad_norm": 0.290264368057251, "learning_rate": 1.7231552527007933e-05, "loss": 0.0132, "step": 59290 }, { "grad_norm": 0.20452578365802765, "learning_rate": 1.721594432112606e-05, "loss": 0.0148, "step": 59300 }, { "grad_norm": 0.2754870355129242, "learning_rate": 1.7200341717142655e-05, "loss": 0.0177, "step": 59310 }, { "grad_norm": 0.2678052484989166, "learning_rate": 1.718474471772377e-05, "loss": 0.0135, "step": 59320 }, { "grad_norm": 0.2546776533126831, "learning_rate": 1.7169153325534528e-05, "loss": 0.0134, "step": 59330 }, { "grad_norm": 0.19744062423706055, "learning_rate": 1.7153567543239045e-05, "loss": 0.01, "step": 59340 }, { "grad_norm": 0.23432889580726624, "learning_rate": 1.7137987373500525e-05, "loss": 0.0124, "step": 59350 }, { "grad_norm": 0.21440312266349792, "learning_rate": 1.7122412818981198e-05, "loss": 0.0119, "step": 59360 }, { "grad_norm": 0.18515430390834808, "learning_rate": 1.7106843882342283e-05, "loss": 0.0101, "step": 59370 }, { "grad_norm": 0.23175226151943207, "learning_rate": 1.709128056624415e-05, "loss": 0.0127, "step": 59380 }, { "grad_norm": 0.25940293073654175, "learning_rate": 1.70757228733461e-05, "loss": 0.0134, "step": 59390 }, { "grad_norm": 0.30669069290161133, "learning_rate": 1.706017080630653e-05, "loss": 0.0149, "step": 59400 }, { "grad_norm": 0.21036554872989655, "learning_rate": 1.7044624367782873e-05, "loss": 0.011, "step": 59410 }, { "grad_norm": 0.21695783734321594, "learning_rate": 1.7029083560431553e-05, "loss": 0.0109, "step": 59420 }, { "grad_norm": 0.321238249540329, "learning_rate": 1.701354838690813e-05, "loss": 0.0121, "step": 59430 }, { "grad_norm": 0.2377803772687912, "learning_rate": 1.6998018849867087e-05, "loss": 0.0109, "step": 59440 }, { "grad_norm": 0.1904979646205902, "learning_rate": 1.698249495196202e-05, "loss": 0.0103, "step": 59450 }, { "grad_norm": 0.16262158751487732, "learning_rate": 1.6966976695845528e-05, "loss": 0.0159, "step": 59460 }, { "grad_norm": 0.20721040666103363, "learning_rate": 1.6951464084169268e-05, "loss": 0.0089, "step": 59470 }, { "grad_norm": 0.14965349435806274, "learning_rate": 1.6935957119583916e-05, "loss": 0.0131, "step": 59480 }, { "grad_norm": 0.23923878371715546, "learning_rate": 1.6920455804739205e-05, "loss": 0.013, "step": 59490 }, { "grad_norm": 0.21565794944763184, "learning_rate": 1.6904960142283855e-05, "loss": 0.0132, "step": 59500 }, { "grad_norm": 0.21344316005706787, "learning_rate": 1.6889470134865666e-05, "loss": 0.0104, "step": 59510 }, { "grad_norm": 0.15173864364624023, "learning_rate": 1.687398578513145e-05, "loss": 0.0101, "step": 59520 }, { "grad_norm": 0.21187342703342438, "learning_rate": 1.6858507095727066e-05, "loss": 0.01, "step": 59530 }, { "grad_norm": 0.49068784713745117, "learning_rate": 1.6843034069297403e-05, "loss": 0.0103, "step": 59540 }, { "grad_norm": 0.7662982940673828, "learning_rate": 1.682756670848637e-05, "loss": 0.0185, "step": 59550 }, { "grad_norm": 0.2184251993894577, "learning_rate": 1.6812105015936937e-05, "loss": 0.0096, "step": 59560 }, { "grad_norm": 0.23570120334625244, "learning_rate": 1.6796648994291027e-05, "loss": 0.016, "step": 59570 }, { "grad_norm": 0.20134179294109344, "learning_rate": 1.678119864618973e-05, "loss": 0.0095, "step": 59580 }, { "grad_norm": 0.14511635899543762, "learning_rate": 1.6765753974273023e-05, "loss": 0.0093, "step": 59590 }, { "grad_norm": 0.16349847614765167, "learning_rate": 1.675031498118001e-05, "loss": 0.0098, "step": 59600 }, { "grad_norm": 0.21438737213611603, "learning_rate": 1.6734881669548804e-05, "loss": 0.0086, "step": 59610 }, { "grad_norm": 0.190182626247406, "learning_rate": 1.671945404201648e-05, "loss": 0.0175, "step": 59620 }, { "grad_norm": 0.1358054131269455, "learning_rate": 1.670403210121927e-05, "loss": 0.0107, "step": 59630 }, { "grad_norm": 0.3006137013435364, "learning_rate": 1.6688615849792312e-05, "loss": 0.0103, "step": 59640 }, { "grad_norm": 0.3227679431438446, "learning_rate": 1.6673205290369832e-05, "loss": 0.0131, "step": 59650 }, { "grad_norm": 0.7917004823684692, "learning_rate": 1.6657800425585086e-05, "loss": 0.0123, "step": 59660 }, { "grad_norm": 0.264403373003006, "learning_rate": 1.6642401258070327e-05, "loss": 0.0187, "step": 59670 }, { "grad_norm": 0.23330116271972656, "learning_rate": 1.6627007790456884e-05, "loss": 0.0147, "step": 59680 }, { "grad_norm": 1.1930172443389893, "learning_rate": 1.6611620025375037e-05, "loss": 0.01, "step": 59690 }, { "grad_norm": 0.15495412051677704, "learning_rate": 1.6596237965454154e-05, "loss": 0.0142, "step": 59700 }, { "grad_norm": 0.27650168538093567, "learning_rate": 1.6580861613322607e-05, "loss": 0.0145, "step": 59710 }, { "grad_norm": 0.15274955332279205, "learning_rate": 1.6565490971607796e-05, "loss": 0.0103, "step": 59720 }, { "grad_norm": 0.24189393222332, "learning_rate": 1.655012604293614e-05, "loss": 0.0123, "step": 59730 }, { "grad_norm": 0.1754966825246811, "learning_rate": 1.653476682993309e-05, "loss": 0.0134, "step": 59740 }, { "grad_norm": 0.24144642055034637, "learning_rate": 1.651941333522311e-05, "loss": 0.0117, "step": 59750 }, { "grad_norm": 0.27399706840515137, "learning_rate": 1.6504065561429715e-05, "loss": 0.0124, "step": 59760 }, { "grad_norm": 0.22124122083187103, "learning_rate": 1.6488723511175385e-05, "loss": 0.0172, "step": 59770 }, { "grad_norm": 0.23181529343128204, "learning_rate": 1.6473387187081668e-05, "loss": 0.0135, "step": 59780 }, { "grad_norm": 0.2216060757637024, "learning_rate": 1.6458056591769123e-05, "loss": 0.0108, "step": 59790 }, { "grad_norm": 0.37229034304618835, "learning_rate": 1.6442731727857335e-05, "loss": 0.0139, "step": 59800 }, { "grad_norm": 0.2270936816930771, "learning_rate": 1.642741259796492e-05, "loss": 0.0088, "step": 59810 }, { "grad_norm": 0.20810072124004364, "learning_rate": 1.641209920470944e-05, "loss": 0.0138, "step": 59820 }, { "grad_norm": 0.1832522749900818, "learning_rate": 1.639679155070762e-05, "loss": 0.0141, "step": 59830 }, { "grad_norm": 0.4528424143791199, "learning_rate": 1.6381489638575048e-05, "loss": 0.0132, "step": 59840 }, { "grad_norm": 1.1811813116073608, "learning_rate": 1.636619347092643e-05, "loss": 0.0124, "step": 59850 }, { "grad_norm": 0.2733987867832184, "learning_rate": 1.6350903050375476e-05, "loss": 0.0114, "step": 59860 }, { "grad_norm": 0.22505612671375275, "learning_rate": 1.6335618379534856e-05, "loss": 0.0114, "step": 59870 }, { "grad_norm": 0.4100826680660248, "learning_rate": 1.6320339461016364e-05, "loss": 0.0111, "step": 59880 }, { "grad_norm": 0.2640096843242645, "learning_rate": 1.6305066297430687e-05, "loss": 0.0108, "step": 59890 }, { "grad_norm": 0.13613660633563995, "learning_rate": 1.6289798891387654e-05, "loss": 0.0154, "step": 59900 }, { "grad_norm": 0.23332807421684265, "learning_rate": 1.6274537245495995e-05, "loss": 0.0087, "step": 59910 }, { "grad_norm": 0.41297462582588196, "learning_rate": 1.6259281362363527e-05, "loss": 0.0119, "step": 59920 }, { "grad_norm": 0.5332258343696594, "learning_rate": 1.6244031244597068e-05, "loss": 0.0139, "step": 59930 }, { "grad_norm": 0.1276119202375412, "learning_rate": 1.622878689480244e-05, "loss": 0.0127, "step": 59940 }, { "grad_norm": 0.3058120012283325, "learning_rate": 1.6213548315584498e-05, "loss": 0.0149, "step": 59950 }, { "grad_norm": 0.1940488964319229, "learning_rate": 1.6198315509547074e-05, "loss": 0.0119, "step": 59960 }, { "grad_norm": 0.437066912651062, "learning_rate": 1.6183088479293056e-05, "loss": 0.0098, "step": 59970 }, { "grad_norm": 0.22917817533016205, "learning_rate": 1.6167867227424316e-05, "loss": 0.0129, "step": 59980 }, { "grad_norm": 0.1308698058128357, "learning_rate": 1.615265175654176e-05, "loss": 0.015, "step": 59990 }, { "grad_norm": 0.2102929651737213, "learning_rate": 1.613744206924529e-05, "loss": 0.0128, "step": 60000 }, { "grad_norm": 0.19624662399291992, "learning_rate": 1.6122238168133845e-05, "loss": 0.0165, "step": 60010 }, { "grad_norm": 0.17063939571380615, "learning_rate": 1.6107040055805305e-05, "loss": 0.0137, "step": 60020 }, { "grad_norm": 0.21255646646022797, "learning_rate": 1.6091847734856673e-05, "loss": 0.0101, "step": 60030 }, { "grad_norm": 0.2841113805770874, "learning_rate": 1.6076661207883865e-05, "loss": 0.0129, "step": 60040 }, { "grad_norm": 0.6407032608985901, "learning_rate": 1.6061480477481848e-05, "loss": 0.0119, "step": 60050 }, { "grad_norm": 0.22747235000133514, "learning_rate": 1.604630554624461e-05, "loss": 0.0141, "step": 60060 }, { "grad_norm": 0.30920958518981934, "learning_rate": 1.603113641676509e-05, "loss": 0.0096, "step": 60070 }, { "grad_norm": 0.18563610315322876, "learning_rate": 1.6015973091635338e-05, "loss": 0.01, "step": 60080 }, { "grad_norm": 0.23141206800937653, "learning_rate": 1.6000815573446288e-05, "loss": 0.0103, "step": 60090 }, { "grad_norm": 0.4180852174758911, "learning_rate": 1.5985663864788002e-05, "loss": 0.0183, "step": 60100 }, { "grad_norm": 0.2268695831298828, "learning_rate": 1.597051796824946e-05, "loss": 0.0109, "step": 60110 }, { "grad_norm": 0.2359750121831894, "learning_rate": 1.5955377886418682e-05, "loss": 0.0136, "step": 60120 }, { "grad_norm": 0.20025447010993958, "learning_rate": 1.5940243621882704e-05, "loss": 0.0143, "step": 60130 }, { "grad_norm": 0.1784631311893463, "learning_rate": 1.5925115177227555e-05, "loss": 0.0157, "step": 60140 }, { "grad_norm": 0.2475012093782425, "learning_rate": 1.5909992555038288e-05, "loss": 0.0116, "step": 60150 }, { "grad_norm": 0.2176446169614792, "learning_rate": 1.5894875757898912e-05, "loss": 0.0175, "step": 60160 }, { "grad_norm": 0.2297031283378601, "learning_rate": 1.58797647883925e-05, "loss": 0.0113, "step": 60170 }, { "grad_norm": 0.17295552790164948, "learning_rate": 1.586465964910109e-05, "loss": 0.0123, "step": 60180 }, { "grad_norm": 0.13972067832946777, "learning_rate": 1.5849560342605734e-05, "loss": 0.0121, "step": 60190 }, { "grad_norm": 0.289445161819458, "learning_rate": 1.583446687148651e-05, "loss": 0.0094, "step": 60200 }, { "grad_norm": 0.18081195652484894, "learning_rate": 1.581937923832248e-05, "loss": 0.0112, "step": 60210 }, { "grad_norm": 0.1818121075630188, "learning_rate": 1.5804297445691663e-05, "loss": 0.0109, "step": 60220 }, { "grad_norm": 0.21097645163536072, "learning_rate": 1.578922149617119e-05, "loss": 0.0113, "step": 60230 }, { "grad_norm": 0.20659761130809784, "learning_rate": 1.5774151392337084e-05, "loss": 0.0117, "step": 60240 }, { "grad_norm": 0.2811064124107361, "learning_rate": 1.575908713676442e-05, "loss": 0.0122, "step": 60250 }, { "grad_norm": 0.21505148708820343, "learning_rate": 1.5744028732027293e-05, "loss": 0.0131, "step": 60260 }, { "grad_norm": 0.2792700231075287, "learning_rate": 1.5728976180698723e-05, "loss": 0.0159, "step": 60270 }, { "grad_norm": 0.22338438034057617, "learning_rate": 1.5713929485350836e-05, "loss": 0.0138, "step": 60280 }, { "grad_norm": 0.27722182869911194, "learning_rate": 1.569888864855465e-05, "loss": 0.0135, "step": 60290 }, { "grad_norm": 0.18297335505485535, "learning_rate": 1.568385367288028e-05, "loss": 0.0119, "step": 60300 }, { "grad_norm": 0.14041224122047424, "learning_rate": 1.5668824560896755e-05, "loss": 0.0095, "step": 60310 }, { "grad_norm": 0.3087468445301056, "learning_rate": 1.5653801315172156e-05, "loss": 0.014, "step": 60320 }, { "grad_norm": 0.18594679236412048, "learning_rate": 1.5638783938273554e-05, "loss": 0.0104, "step": 60330 }, { "grad_norm": 0.2387544810771942, "learning_rate": 1.5623772432766966e-05, "loss": 0.014, "step": 60340 }, { "grad_norm": 0.21374565362930298, "learning_rate": 1.5608766801217507e-05, "loss": 0.0109, "step": 60350 }, { "grad_norm": 0.4282018840312958, "learning_rate": 1.5593767046189184e-05, "loss": 0.011, "step": 60360 }, { "grad_norm": 0.16834379732608795, "learning_rate": 1.5578773170245064e-05, "loss": 0.0099, "step": 60370 }, { "grad_norm": 0.4061209261417389, "learning_rate": 1.5563785175947182e-05, "loss": 0.0133, "step": 60380 }, { "grad_norm": 0.3183901607990265, "learning_rate": 1.5548803065856582e-05, "loss": 0.0112, "step": 60390 }, { "grad_norm": 0.13958722352981567, "learning_rate": 1.55338268425333e-05, "loss": 0.0097, "step": 60400 }, { "grad_norm": 0.369266539812088, "learning_rate": 1.5518856508536373e-05, "loss": 0.0103, "step": 60410 }, { "grad_norm": 0.3189254105091095, "learning_rate": 1.5503892066423786e-05, "loss": 0.0129, "step": 60420 }, { "grad_norm": 0.2707727551460266, "learning_rate": 1.548893351875258e-05, "loss": 0.0195, "step": 60430 }, { "grad_norm": 0.30231204628944397, "learning_rate": 1.547398086807875e-05, "loss": 0.0148, "step": 60440 }, { "grad_norm": 0.2674354612827301, "learning_rate": 1.5459034116957305e-05, "loss": 0.0182, "step": 60450 }, { "grad_norm": 0.29322561621665955, "learning_rate": 1.544409326794225e-05, "loss": 0.0197, "step": 60460 }, { "grad_norm": 0.2817988097667694, "learning_rate": 1.542915832358651e-05, "loss": 0.0142, "step": 60470 }, { "grad_norm": 0.29478776454925537, "learning_rate": 1.541422928644214e-05, "loss": 0.0112, "step": 60480 }, { "grad_norm": 0.239105686545372, "learning_rate": 1.5399306159060024e-05, "loss": 0.0174, "step": 60490 }, { "grad_norm": 0.18214468657970428, "learning_rate": 1.5384388943990185e-05, "loss": 0.01, "step": 60500 }, { "grad_norm": 0.6102849841117859, "learning_rate": 1.5369477643781526e-05, "loss": 0.012, "step": 60510 }, { "grad_norm": 0.34878531098365784, "learning_rate": 1.5354572260981985e-05, "loss": 0.0131, "step": 60520 }, { "grad_norm": 0.22279447317123413, "learning_rate": 1.533967279813851e-05, "loss": 0.0109, "step": 60530 }, { "grad_norm": 0.24561052024364471, "learning_rate": 1.5324779257796956e-05, "loss": 0.0113, "step": 60540 }, { "grad_norm": 0.2911011278629303, "learning_rate": 1.5309891642502293e-05, "loss": 0.0117, "step": 60550 }, { "grad_norm": 0.23917609453201294, "learning_rate": 1.5295009954798357e-05, "loss": 0.0098, "step": 60560 }, { "grad_norm": 0.21909473836421967, "learning_rate": 1.5280134197228036e-05, "loss": 0.0122, "step": 60570 }, { "grad_norm": 0.16392695903778076, "learning_rate": 1.526526437233319e-05, "loss": 0.0122, "step": 60580 }, { "grad_norm": 0.24710457026958466, "learning_rate": 1.525040048265466e-05, "loss": 0.0124, "step": 60590 }, { "grad_norm": 0.16481897234916687, "learning_rate": 1.5235542530732288e-05, "loss": 0.0096, "step": 60600 }, { "grad_norm": 1.250624179840088, "learning_rate": 1.5220690519104901e-05, "loss": 0.0117, "step": 60610 }, { "grad_norm": 0.2047813981771469, "learning_rate": 1.5205844450310275e-05, "loss": 0.0127, "step": 60620 }, { "grad_norm": 0.2153279185295105, "learning_rate": 1.519100432688521e-05, "loss": 0.0113, "step": 60630 }, { "grad_norm": 0.17583657801151276, "learning_rate": 1.5176170151365476e-05, "loss": 0.0117, "step": 60640 }, { "grad_norm": 0.6623921990394592, "learning_rate": 1.5161341926285833e-05, "loss": 0.0096, "step": 60650 }, { "grad_norm": 0.25086215138435364, "learning_rate": 1.5146519654180025e-05, "loss": 0.0091, "step": 60660 }, { "grad_norm": 0.22120894491672516, "learning_rate": 1.5131703337580739e-05, "loss": 0.016, "step": 60670 }, { "grad_norm": 0.20522603392601013, "learning_rate": 1.5116892979019731e-05, "loss": 0.0153, "step": 60680 }, { "grad_norm": 0.10966094583272934, "learning_rate": 1.5102088581027623e-05, "loss": 0.0095, "step": 60690 }, { "grad_norm": 0.21309952437877655, "learning_rate": 1.5087290146134154e-05, "loss": 0.0117, "step": 60700 }, { "grad_norm": 0.23042568564414978, "learning_rate": 1.5072497676867915e-05, "loss": 0.0103, "step": 60710 }, { "grad_norm": 0.13375720381736755, "learning_rate": 1.505771117575655e-05, "loss": 0.0113, "step": 60720 }, { "grad_norm": 0.17540673911571503, "learning_rate": 1.5042930645326691e-05, "loss": 0.0093, "step": 60730 }, { "grad_norm": 0.3872169554233551, "learning_rate": 1.5028156088103879e-05, "loss": 0.0125, "step": 60740 }, { "grad_norm": 0.3333757519721985, "learning_rate": 1.5013387506612735e-05, "loss": 0.0153, "step": 60750 }, { "grad_norm": 0.8326510787010193, "learning_rate": 1.4998624903376767e-05, "loss": 0.0141, "step": 60760 }, { "grad_norm": 0.25215625762939453, "learning_rate": 1.4983868280918517e-05, "loss": 0.0144, "step": 60770 }, { "grad_norm": 0.3181987404823303, "learning_rate": 1.4969117641759478e-05, "loss": 0.0158, "step": 60780 }, { "grad_norm": 0.1899190992116928, "learning_rate": 1.495437298842014e-05, "loss": 0.0117, "step": 60790 }, { "grad_norm": 0.23987480998039246, "learning_rate": 1.4939634323419976e-05, "loss": 0.0097, "step": 60800 }, { "grad_norm": 0.2933933138847351, "learning_rate": 1.4924901649277384e-05, "loss": 0.0127, "step": 60810 }, { "grad_norm": 0.1939440816640854, "learning_rate": 1.4910174968509793e-05, "loss": 0.008, "step": 60820 }, { "grad_norm": 0.243228018283844, "learning_rate": 1.4895454283633598e-05, "loss": 0.0114, "step": 60830 }, { "grad_norm": 0.2514728307723999, "learning_rate": 1.488073959716415e-05, "loss": 0.0101, "step": 60840 }, { "grad_norm": 0.1640457808971405, "learning_rate": 1.4866030911615791e-05, "loss": 0.0121, "step": 60850 }, { "grad_norm": 0.09637077897787094, "learning_rate": 1.4851328229501849e-05, "loss": 0.0119, "step": 60860 }, { "grad_norm": 0.18924124538898468, "learning_rate": 1.4836631553334562e-05, "loss": 0.0096, "step": 60870 }, { "grad_norm": 0.13123783469200134, "learning_rate": 1.4821940885625251e-05, "loss": 0.0109, "step": 60880 }, { "grad_norm": 0.2629687190055847, "learning_rate": 1.4807256228884109e-05, "loss": 0.011, "step": 60890 }, { "grad_norm": 0.14482460916042328, "learning_rate": 1.4792577585620353e-05, "loss": 0.0115, "step": 60900 }, { "grad_norm": 0.31246376037597656, "learning_rate": 1.4777904958342164e-05, "loss": 0.0141, "step": 60910 }, { "grad_norm": 0.33688241243362427, "learning_rate": 1.4763238349556691e-05, "loss": 0.0106, "step": 60920 }, { "grad_norm": 0.17299389839172363, "learning_rate": 1.4748577761770072e-05, "loss": 0.0149, "step": 60930 }, { "grad_norm": 0.18209059536457062, "learning_rate": 1.4733923197487354e-05, "loss": 0.0112, "step": 60940 }, { "grad_norm": 0.4629580080509186, "learning_rate": 1.4719274659212662e-05, "loss": 0.014, "step": 60950 }, { "grad_norm": 0.13571590185165405, "learning_rate": 1.4704632149448983e-05, "loss": 0.0094, "step": 60960 }, { "grad_norm": 0.2803242802619934, "learning_rate": 1.4689995670698343e-05, "loss": 0.0176, "step": 60970 }, { "grad_norm": 0.2171669751405716, "learning_rate": 1.4675365225461728e-05, "loss": 0.0169, "step": 60980 }, { "grad_norm": 0.36941206455230713, "learning_rate": 1.4660740816239032e-05, "loss": 0.0094, "step": 60990 }, { "grad_norm": 0.4966183304786682, "learning_rate": 1.464612244552923e-05, "loss": 0.013, "step": 61000 }, { "grad_norm": 0.25165385007858276, "learning_rate": 1.4631510115830161e-05, "loss": 0.0145, "step": 61010 }, { "grad_norm": 0.18429847061634064, "learning_rate": 1.4616903829638679e-05, "loss": 0.0122, "step": 61020 }, { "grad_norm": 0.1882709264755249, "learning_rate": 1.46023035894506e-05, "loss": 0.0103, "step": 61030 }, { "grad_norm": 0.25661906599998474, "learning_rate": 1.4587709397760713e-05, "loss": 0.0173, "step": 61040 }, { "grad_norm": 1.4416441917419434, "learning_rate": 1.4573121257062755e-05, "loss": 0.0132, "step": 61050 }, { "grad_norm": 0.6193217635154724, "learning_rate": 1.4558539169849472e-05, "loss": 0.0161, "step": 61060 }, { "grad_norm": 0.2249704748392105, "learning_rate": 1.4543963138612499e-05, "loss": 0.0133, "step": 61070 }, { "grad_norm": 0.2721116542816162, "learning_rate": 1.4529393165842498e-05, "loss": 0.0126, "step": 61080 }, { "grad_norm": 0.21250909566879272, "learning_rate": 1.4514829254029084e-05, "loss": 0.0131, "step": 61090 }, { "grad_norm": 0.4603988528251648, "learning_rate": 1.450027140566083e-05, "loss": 0.0122, "step": 61100 }, { "grad_norm": 0.17619925737380981, "learning_rate": 1.4485719623225268e-05, "loss": 0.0131, "step": 61110 }, { "grad_norm": 0.14368240535259247, "learning_rate": 1.4471173909208912e-05, "loss": 0.0116, "step": 61120 }, { "grad_norm": 0.3395143449306488, "learning_rate": 1.4456634266097236e-05, "loss": 0.0122, "step": 61130 }, { "grad_norm": 0.1679978221654892, "learning_rate": 1.444210069637461e-05, "loss": 0.0152, "step": 61140 }, { "grad_norm": 0.21764087677001953, "learning_rate": 1.4427573202524502e-05, "loss": 0.0105, "step": 61150 }, { "grad_norm": 0.18942174315452576, "learning_rate": 1.4413051787029208e-05, "loss": 0.0159, "step": 61160 }, { "grad_norm": 0.2063475251197815, "learning_rate": 1.439853645237006e-05, "loss": 0.0118, "step": 61170 }, { "grad_norm": 0.14058052003383636, "learning_rate": 1.4384027201027344e-05, "loss": 0.0114, "step": 61180 }, { "grad_norm": 0.1729942113161087, "learning_rate": 1.4369524035480253e-05, "loss": 0.0122, "step": 61190 }, { "grad_norm": 0.18334493041038513, "learning_rate": 1.4355026958207035e-05, "loss": 0.0099, "step": 61200 }, { "grad_norm": 0.2346818596124649, "learning_rate": 1.434053597168481e-05, "loss": 0.0133, "step": 61210 }, { "grad_norm": 0.1654522716999054, "learning_rate": 1.432605107838969e-05, "loss": 0.0139, "step": 61220 }, { "grad_norm": 0.2496868520975113, "learning_rate": 1.4311572280796759e-05, "loss": 0.0119, "step": 61230 }, { "grad_norm": 0.1569395661354065, "learning_rate": 1.4297099581380047e-05, "loss": 0.0131, "step": 61240 }, { "grad_norm": 0.2060549259185791, "learning_rate": 1.4282632982612538e-05, "loss": 0.0131, "step": 61250 }, { "grad_norm": 0.14144399762153625, "learning_rate": 1.4268172486966197e-05, "loss": 0.0158, "step": 61260 }, { "grad_norm": 0.1998787224292755, "learning_rate": 1.4253718096911894e-05, "loss": 0.0139, "step": 61270 }, { "grad_norm": 0.288132905960083, "learning_rate": 1.4239269814919504e-05, "loss": 0.0124, "step": 61280 }, { "grad_norm": 0.5051769018173218, "learning_rate": 1.4224827643457843e-05, "loss": 0.0186, "step": 61290 }, { "grad_norm": 0.22757908701896667, "learning_rate": 1.4210391584994686e-05, "loss": 0.0154, "step": 61300 }, { "grad_norm": 0.18097622692584991, "learning_rate": 1.4195961641996763e-05, "loss": 0.0124, "step": 61310 }, { "grad_norm": 0.2360049933195114, "learning_rate": 1.4181537816929751e-05, "loss": 0.014, "step": 61320 }, { "grad_norm": 0.1899547576904297, "learning_rate": 1.4167120112258303e-05, "loss": 0.0139, "step": 61330 }, { "grad_norm": 0.22238855063915253, "learning_rate": 1.4152708530445963e-05, "loss": 0.011, "step": 61340 }, { "grad_norm": 0.1775747537612915, "learning_rate": 1.4138303073955344e-05, "loss": 0.0116, "step": 61350 }, { "grad_norm": 0.23966678977012634, "learning_rate": 1.4123903745247897e-05, "loss": 0.012, "step": 61360 }, { "grad_norm": 0.15359383821487427, "learning_rate": 1.4109510546784082e-05, "loss": 0.0101, "step": 61370 }, { "grad_norm": 0.21007345616817474, "learning_rate": 1.4095123481023325e-05, "loss": 0.0161, "step": 61380 }, { "grad_norm": 1.3431322574615479, "learning_rate": 1.4080742550423936e-05, "loss": 0.0142, "step": 61390 }, { "grad_norm": 0.2831932008266449, "learning_rate": 1.4066367757443278e-05, "loss": 0.0118, "step": 61400 }, { "grad_norm": 0.23114226758480072, "learning_rate": 1.4051999104537572e-05, "loss": 0.0111, "step": 61410 }, { "grad_norm": 0.21895411610603333, "learning_rate": 1.4037636594162034e-05, "loss": 0.0123, "step": 61420 }, { "grad_norm": 0.21477890014648438, "learning_rate": 1.402328022877083e-05, "loss": 0.0101, "step": 61430 }, { "grad_norm": 0.21241545677185059, "learning_rate": 1.4008930010817073e-05, "loss": 0.0145, "step": 61440 }, { "grad_norm": 0.21609419584274292, "learning_rate": 1.3994585942752831e-05, "loss": 0.0124, "step": 61450 }, { "grad_norm": 0.2378813475370407, "learning_rate": 1.398024802702909e-05, "loss": 0.0101, "step": 61460 }, { "grad_norm": 0.290225088596344, "learning_rate": 1.3965916266095819e-05, "loss": 0.0148, "step": 61470 }, { "grad_norm": 0.21812258660793304, "learning_rate": 1.395159066240192e-05, "loss": 0.0122, "step": 61480 }, { "grad_norm": 0.18783031404018402, "learning_rate": 1.3937271218395248e-05, "loss": 0.0149, "step": 61490 }, { "grad_norm": 0.2567687928676605, "learning_rate": 1.3922957936522613e-05, "loss": 0.0117, "step": 61500 }, { "grad_norm": 0.4287925660610199, "learning_rate": 1.3908650819229758e-05, "loss": 0.0181, "step": 61510 }, { "grad_norm": 0.1837548464536667, "learning_rate": 1.389434986896137e-05, "loss": 0.0106, "step": 61520 }, { "grad_norm": 0.13774611055850983, "learning_rate": 1.388005508816112e-05, "loss": 0.0101, "step": 61530 }, { "grad_norm": 0.26915258169174194, "learning_rate": 1.3865766479271557e-05, "loss": 0.0114, "step": 61540 }, { "grad_norm": 0.2321157306432724, "learning_rate": 1.3851484044734225e-05, "loss": 0.0201, "step": 61550 }, { "grad_norm": 0.26771122217178345, "learning_rate": 1.3837207786989608e-05, "loss": 0.0198, "step": 61560 }, { "grad_norm": 0.21015310287475586, "learning_rate": 1.382293770847713e-05, "loss": 0.0109, "step": 61570 }, { "grad_norm": 0.28239521384239197, "learning_rate": 1.3808673811635159e-05, "loss": 0.0122, "step": 61580 }, { "grad_norm": 0.32670941948890686, "learning_rate": 1.3794416098900975e-05, "loss": 0.0125, "step": 61590 }, { "grad_norm": 0.18720367550849915, "learning_rate": 1.3780164572710886e-05, "loss": 0.0139, "step": 61600 }, { "grad_norm": 0.2653541564941406, "learning_rate": 1.3765919235500035e-05, "loss": 0.013, "step": 61610 }, { "grad_norm": 0.2571779191493988, "learning_rate": 1.375168008970259e-05, "loss": 0.0116, "step": 61620 }, { "grad_norm": 0.1469680815935135, "learning_rate": 1.3737447137751635e-05, "loss": 0.0117, "step": 61630 }, { "grad_norm": 0.27691012620925903, "learning_rate": 1.372322038207915e-05, "loss": 0.0135, "step": 61640 }, { "grad_norm": 0.32204681634902954, "learning_rate": 1.3708999825116159e-05, "loss": 0.0162, "step": 61650 }, { "grad_norm": 0.2494441568851471, "learning_rate": 1.3694785469292526e-05, "loss": 0.0175, "step": 61660 }, { "grad_norm": 0.16568903625011444, "learning_rate": 1.3680577317037101e-05, "loss": 0.0096, "step": 61670 }, { "grad_norm": 0.23198923468589783, "learning_rate": 1.3666375370777678e-05, "loss": 0.0093, "step": 61680 }, { "grad_norm": 0.13619305193424225, "learning_rate": 1.3652179632940981e-05, "loss": 0.0073, "step": 61690 }, { "grad_norm": 0.18221767246723175, "learning_rate": 1.3637990105952664e-05, "loss": 0.0125, "step": 61700 }, { "grad_norm": 0.1757345050573349, "learning_rate": 1.3623806792237337e-05, "loss": 0.0099, "step": 61710 }, { "grad_norm": 0.21663248538970947, "learning_rate": 1.3609629694218551e-05, "loss": 0.0147, "step": 61720 }, { "grad_norm": 0.18161001801490784, "learning_rate": 1.3595458814318762e-05, "loss": 0.0104, "step": 61730 }, { "grad_norm": 0.8451244831085205, "learning_rate": 1.3581294154959389e-05, "loss": 0.0154, "step": 61740 }, { "grad_norm": 0.36290645599365234, "learning_rate": 1.3567135718560792e-05, "loss": 0.0131, "step": 61750 }, { "grad_norm": 0.2064133882522583, "learning_rate": 1.3552983507542261e-05, "loss": 0.0126, "step": 61760 }, { "grad_norm": 0.4087008237838745, "learning_rate": 1.3538837524322023e-05, "loss": 0.0144, "step": 61770 }, { "grad_norm": 0.17778882384300232, "learning_rate": 1.3524697771317251e-05, "loss": 0.0184, "step": 61780 }, { "grad_norm": 0.29012617468833923, "learning_rate": 1.3510564250943997e-05, "loss": 0.0113, "step": 61790 }, { "grad_norm": 0.29635369777679443, "learning_rate": 1.3496436965617353e-05, "loss": 0.0116, "step": 61800 }, { "grad_norm": 0.20408909022808075, "learning_rate": 1.3482315917751243e-05, "loss": 0.0143, "step": 61810 }, { "grad_norm": 0.21316145360469818, "learning_rate": 1.3468201109758582e-05, "loss": 0.0118, "step": 61820 }, { "grad_norm": 0.20366597175598145, "learning_rate": 1.3454092544051222e-05, "loss": 0.0126, "step": 61830 }, { "grad_norm": 0.17975717782974243, "learning_rate": 1.3439990223039878e-05, "loss": 0.0133, "step": 61840 }, { "grad_norm": 0.15114958584308624, "learning_rate": 1.3425894149134321e-05, "loss": 0.0111, "step": 61850 }, { "grad_norm": 0.23505470156669617, "learning_rate": 1.3411804324743132e-05, "loss": 0.0124, "step": 61860 }, { "grad_norm": 0.22411386668682098, "learning_rate": 1.3397720752273896e-05, "loss": 0.0129, "step": 61870 }, { "grad_norm": 0.42857450246810913, "learning_rate": 1.3383643434133108e-05, "loss": 0.0135, "step": 61880 }, { "grad_norm": 0.16558803617954254, "learning_rate": 1.3369572372726196e-05, "loss": 0.0136, "step": 61890 }, { "grad_norm": 0.18900512158870697, "learning_rate": 1.3355507570457525e-05, "loss": 0.0106, "step": 61900 }, { "grad_norm": 0.2458789050579071, "learning_rate": 1.334144902973038e-05, "loss": 0.0093, "step": 61910 }, { "grad_norm": 0.19533903896808624, "learning_rate": 1.3327396752946997e-05, "loss": 0.0121, "step": 61920 }, { "grad_norm": 0.17882481217384338, "learning_rate": 1.3313350742508496e-05, "loss": 0.0137, "step": 61930 }, { "grad_norm": 0.24349462985992432, "learning_rate": 1.3299311000814973e-05, "loss": 0.0136, "step": 61940 }, { "grad_norm": 0.1539268046617508, "learning_rate": 1.3285277530265432e-05, "loss": 0.0137, "step": 61950 }, { "grad_norm": 0.2227034717798233, "learning_rate": 1.3271250333257813e-05, "loss": 0.0139, "step": 61960 }, { "grad_norm": 2.3205833435058594, "learning_rate": 1.3257229412188977e-05, "loss": 0.0106, "step": 61970 }, { "grad_norm": 0.1918942779302597, "learning_rate": 1.3243214769454726e-05, "loss": 0.0114, "step": 61980 }, { "grad_norm": 0.1661357581615448, "learning_rate": 1.3229206407449751e-05, "loss": 0.0105, "step": 61990 }, { "grad_norm": 0.2506335377693176, "learning_rate": 1.3215204328567742e-05, "loss": 0.0124, "step": 62000 }, { "grad_norm": 0.161610409617424, "learning_rate": 1.3201208535201232e-05, "loss": 0.0109, "step": 62010 }, { "grad_norm": 0.2016134113073349, "learning_rate": 1.3187219029741737e-05, "loss": 0.0136, "step": 62020 }, { "grad_norm": 0.22659005224704742, "learning_rate": 1.3173235814579693e-05, "loss": 0.0128, "step": 62030 }, { "grad_norm": 0.2239694893360138, "learning_rate": 1.3159258892104398e-05, "loss": 0.0128, "step": 62040 }, { "grad_norm": 0.18164895474910736, "learning_rate": 1.3145288264704198e-05, "loss": 0.0157, "step": 62050 }, { "grad_norm": 0.26486659049987793, "learning_rate": 1.3131323934766237e-05, "loss": 0.0134, "step": 62060 }, { "grad_norm": 0.38831666111946106, "learning_rate": 1.311736590467666e-05, "loss": 0.0134, "step": 62070 }, { "grad_norm": 0.16965043544769287, "learning_rate": 1.3103414176820505e-05, "loss": 0.0126, "step": 62080 }, { "grad_norm": 0.2749275267124176, "learning_rate": 1.308946875358174e-05, "loss": 0.0194, "step": 62090 }, { "grad_norm": 0.24523556232452393, "learning_rate": 1.3075529637343276e-05, "loss": 0.0221, "step": 62100 }, { "grad_norm": 0.15497030317783356, "learning_rate": 1.3061596830486883e-05, "loss": 0.0135, "step": 62110 }, { "grad_norm": 0.4371624290943146, "learning_rate": 1.3047670335393353e-05, "loss": 0.0087, "step": 62120 }, { "grad_norm": 0.17122307419776917, "learning_rate": 1.3033750154442298e-05, "loss": 0.0143, "step": 62130 }, { "grad_norm": 0.48751702904701233, "learning_rate": 1.3019836290012316e-05, "loss": 0.0102, "step": 62140 }, { "grad_norm": 0.20788006484508514, "learning_rate": 1.30059287444809e-05, "loss": 0.0128, "step": 62150 }, { "grad_norm": 0.18697141110897064, "learning_rate": 1.299202752022447e-05, "loss": 0.0222, "step": 62160 }, { "grad_norm": 0.3019084632396698, "learning_rate": 1.2978132619618371e-05, "loss": 0.0093, "step": 62170 }, { "grad_norm": 0.25568267703056335, "learning_rate": 1.2964244045036866e-05, "loss": 0.014, "step": 62180 }, { "grad_norm": 0.12024892121553421, "learning_rate": 1.295036179885311e-05, "loss": 0.0097, "step": 62190 }, { "grad_norm": 0.15293273329734802, "learning_rate": 1.293648588343922e-05, "loss": 0.0147, "step": 62200 }, { "grad_norm": 0.19839780032634735, "learning_rate": 1.2922616301166196e-05, "loss": 0.0133, "step": 62210 }, { "grad_norm": 0.412830650806427, "learning_rate": 1.2908753054403976e-05, "loss": 0.0094, "step": 62220 }, { "grad_norm": 0.2750057280063629, "learning_rate": 1.2894896145521429e-05, "loss": 0.01, "step": 62230 }, { "grad_norm": 0.15993963181972504, "learning_rate": 1.2881045576886275e-05, "loss": 0.0121, "step": 62240 }, { "grad_norm": 0.29278284311294556, "learning_rate": 1.2867201350865254e-05, "loss": 0.0097, "step": 62250 }, { "grad_norm": 0.2619226574897766, "learning_rate": 1.2853363469823914e-05, "loss": 0.0118, "step": 62260 }, { "grad_norm": 0.15481194853782654, "learning_rate": 1.283953193612682e-05, "loss": 0.0073, "step": 62270 }, { "grad_norm": 0.16154219210147858, "learning_rate": 1.2825706752137372e-05, "loss": 0.011, "step": 62280 }, { "grad_norm": 0.2259739190340042, "learning_rate": 1.2811887920217896e-05, "loss": 0.0125, "step": 62290 }, { "grad_norm": 0.1300486922264099, "learning_rate": 1.2798075442729707e-05, "loss": 0.0118, "step": 62300 }, { "grad_norm": 0.35799580812454224, "learning_rate": 1.2784269322032922e-05, "loss": 0.0132, "step": 62310 }, { "grad_norm": 0.23130951821804047, "learning_rate": 1.2770469560486686e-05, "loss": 0.0126, "step": 62320 }, { "grad_norm": 0.26184096932411194, "learning_rate": 1.2756676160448956e-05, "loss": 0.0138, "step": 62330 }, { "grad_norm": 0.16686378419399261, "learning_rate": 1.2742889124276663e-05, "loss": 0.0108, "step": 62340 }, { "grad_norm": 0.2677748203277588, "learning_rate": 1.2729108454325639e-05, "loss": 0.0136, "step": 62350 }, { "grad_norm": 0.38466858863830566, "learning_rate": 1.2715334152950614e-05, "loss": 0.0124, "step": 62360 }, { "grad_norm": 0.20244447886943817, "learning_rate": 1.2701566222505246e-05, "loss": 0.0113, "step": 62370 }, { "grad_norm": 0.3315894603729248, "learning_rate": 1.2687804665342107e-05, "loss": 0.0121, "step": 62380 }, { "grad_norm": 0.22653180360794067, "learning_rate": 1.267404948381265e-05, "loss": 0.0114, "step": 62390 }, { "grad_norm": 0.16044482588768005, "learning_rate": 1.2660300680267267e-05, "loss": 0.0096, "step": 62400 }, { "grad_norm": 0.20795871317386627, "learning_rate": 1.2646558257055257e-05, "loss": 0.0102, "step": 62410 }, { "grad_norm": 0.15191736817359924, "learning_rate": 1.263282221652482e-05, "loss": 0.008, "step": 62420 }, { "grad_norm": 2.0391921997070312, "learning_rate": 1.2619092561023088e-05, "loss": 0.0151, "step": 62430 }, { "grad_norm": 0.21210944652557373, "learning_rate": 1.2605369292896036e-05, "loss": 0.0126, "step": 62440 }, { "grad_norm": 0.15370655059814453, "learning_rate": 1.2591652414488658e-05, "loss": 0.0143, "step": 62450 }, { "grad_norm": 0.2198103964328766, "learning_rate": 1.2577941928144732e-05, "loss": 0.0098, "step": 62460 }, { "grad_norm": 0.45811036229133606, "learning_rate": 1.256423783620706e-05, "loss": 0.0152, "step": 62470 }, { "grad_norm": 0.19621184468269348, "learning_rate": 1.2550540141017264e-05, "loss": 0.0103, "step": 62480 }, { "grad_norm": 0.1820758879184723, "learning_rate": 1.253684884491591e-05, "loss": 0.0101, "step": 62490 }, { "grad_norm": 0.5043376684188843, "learning_rate": 1.2523163950242483e-05, "loss": 0.0096, "step": 62500 }, { "grad_norm": 0.17890632152557373, "learning_rate": 1.2509485459335313e-05, "loss": 0.0095, "step": 62510 }, { "grad_norm": 0.2375732958316803, "learning_rate": 1.2495813374531739e-05, "loss": 0.0108, "step": 62520 }, { "grad_norm": 0.4703159034252167, "learning_rate": 1.2482147698167907e-05, "loss": 0.0136, "step": 62530 }, { "grad_norm": 0.22404763102531433, "learning_rate": 1.2468488432578911e-05, "loss": 0.0126, "step": 62540 }, { "grad_norm": 0.18261510133743286, "learning_rate": 1.2454835580098761e-05, "loss": 0.0099, "step": 62550 }, { "grad_norm": 0.2226700782775879, "learning_rate": 1.2441189143060338e-05, "loss": 0.0108, "step": 62560 }, { "grad_norm": 0.23326461017131805, "learning_rate": 1.242754912379548e-05, "loss": 0.0146, "step": 62570 }, { "grad_norm": 0.14853334426879883, "learning_rate": 1.2413915524634844e-05, "loss": 0.0231, "step": 62580 }, { "grad_norm": 0.19809386134147644, "learning_rate": 1.2400288347908073e-05, "loss": 0.0127, "step": 62590 }, { "grad_norm": 0.39470186829566956, "learning_rate": 1.2386667595943663e-05, "loss": 0.0127, "step": 62600 }, { "grad_norm": 0.2284717559814453, "learning_rate": 1.2373053271069035e-05, "loss": 0.0129, "step": 62610 }, { "grad_norm": 0.27775999903678894, "learning_rate": 1.2359445375610501e-05, "loss": 0.0122, "step": 62620 }, { "grad_norm": 0.21888525784015656, "learning_rate": 1.2345843911893301e-05, "loss": 0.009, "step": 62630 }, { "grad_norm": 0.38242319226264954, "learning_rate": 1.2332248882241498e-05, "loss": 0.0226, "step": 62640 }, { "grad_norm": 0.22852566838264465, "learning_rate": 1.2318660288978178e-05, "loss": 0.0119, "step": 62650 }, { "grad_norm": 0.1573808491230011, "learning_rate": 1.2305078134425213e-05, "loss": 0.0138, "step": 62660 }, { "grad_norm": 0.24214543402194977, "learning_rate": 1.2291502420903434e-05, "loss": 0.0119, "step": 62670 }, { "grad_norm": 0.1507832109928131, "learning_rate": 1.2277933150732567e-05, "loss": 0.0112, "step": 62680 }, { "grad_norm": 0.18622013926506042, "learning_rate": 1.2264370326231216e-05, "loss": 0.0096, "step": 62690 }, { "grad_norm": 0.2653580904006958, "learning_rate": 1.2250813949716927e-05, "loss": 0.0132, "step": 62700 }, { "grad_norm": 0.16558830440044403, "learning_rate": 1.2237264023506063e-05, "loss": 0.0106, "step": 62710 }, { "grad_norm": 0.22197188436985016, "learning_rate": 1.2223720549913987e-05, "loss": 0.01, "step": 62720 }, { "grad_norm": 0.9676179885864258, "learning_rate": 1.221018353125487e-05, "loss": 0.013, "step": 62730 }, { "grad_norm": 0.15089555084705353, "learning_rate": 1.2196652969841837e-05, "loss": 0.0125, "step": 62740 }, { "grad_norm": 0.2783139944076538, "learning_rate": 1.2183128867986904e-05, "loss": 0.0171, "step": 62750 }, { "grad_norm": 0.24385090172290802, "learning_rate": 1.2169611228000927e-05, "loss": 0.0104, "step": 62760 }, { "grad_norm": 0.15662312507629395, "learning_rate": 1.2156100052193752e-05, "loss": 0.0103, "step": 62770 }, { "grad_norm": 0.1993575096130371, "learning_rate": 1.214259534287403e-05, "loss": 0.0131, "step": 62780 }, { "grad_norm": 0.26445314288139343, "learning_rate": 1.2129097102349363e-05, "loss": 0.0096, "step": 62790 }, { "grad_norm": 0.16440241038799286, "learning_rate": 1.2115605332926227e-05, "loss": 0.0097, "step": 62800 }, { "grad_norm": 0.14957326650619507, "learning_rate": 1.2102120036910003e-05, "loss": 0.0134, "step": 62810 }, { "grad_norm": 0.15025834739208221, "learning_rate": 1.2088641216604956e-05, "loss": 0.0107, "step": 62820 }, { "grad_norm": 0.30867379903793335, "learning_rate": 1.2075168874314264e-05, "loss": 0.0098, "step": 62830 }, { "grad_norm": 0.18661633133888245, "learning_rate": 1.2061703012339942e-05, "loss": 0.0096, "step": 62840 }, { "grad_norm": 0.30996930599212646, "learning_rate": 1.204824363298297e-05, "loss": 0.0117, "step": 62850 }, { "grad_norm": 0.17999528348445892, "learning_rate": 1.2034790738543173e-05, "loss": 0.0118, "step": 62860 }, { "grad_norm": 0.2752012312412262, "learning_rate": 1.2021344331319284e-05, "loss": 0.0105, "step": 62870 }, { "grad_norm": 0.13065339624881744, "learning_rate": 1.200790441360894e-05, "loss": 0.013, "step": 62880 }, { "grad_norm": 0.13823202252388, "learning_rate": 1.199447098770864e-05, "loss": 0.0084, "step": 62890 }, { "grad_norm": 0.3357084095478058, "learning_rate": 1.198104405591381e-05, "loss": 0.01, "step": 62900 }, { "grad_norm": 0.5154978036880493, "learning_rate": 1.1967623620518697e-05, "loss": 0.01, "step": 62910 }, { "grad_norm": 0.25199201703071594, "learning_rate": 1.1954209683816554e-05, "loss": 0.0102, "step": 62920 }, { "grad_norm": 0.15488344430923462, "learning_rate": 1.1940802248099402e-05, "loss": 0.0114, "step": 62930 }, { "grad_norm": 0.19712109863758087, "learning_rate": 1.1927401315658232e-05, "loss": 0.0131, "step": 62940 }, { "grad_norm": 0.8260706663131714, "learning_rate": 1.1914006888782898e-05, "loss": 0.0128, "step": 62950 }, { "grad_norm": 0.28394752740859985, "learning_rate": 1.1900618969762106e-05, "loss": 0.0148, "step": 62960 }, { "grad_norm": 0.21041163802146912, "learning_rate": 1.188723756088354e-05, "loss": 0.0088, "step": 62970 }, { "grad_norm": 0.19355426728725433, "learning_rate": 1.1873862664433672e-05, "loss": 0.0125, "step": 62980 }, { "grad_norm": 0.3090001344680786, "learning_rate": 1.1860494282697927e-05, "loss": 0.0109, "step": 62990 }, { "grad_norm": 0.2015582025051117, "learning_rate": 1.1847132417960588e-05, "loss": 0.0112, "step": 63000 }, { "grad_norm": 2.512120246887207, "learning_rate": 1.1833777072504832e-05, "loss": 0.0147, "step": 63010 }, { "grad_norm": 0.15817460417747498, "learning_rate": 1.1820428248612731e-05, "loss": 0.0095, "step": 63020 }, { "grad_norm": 0.35978275537490845, "learning_rate": 1.1807085948565245e-05, "loss": 0.0128, "step": 63030 }, { "grad_norm": 0.49764275550842285, "learning_rate": 1.1793750174642172e-05, "loss": 0.01, "step": 63040 }, { "grad_norm": 0.1643235981464386, "learning_rate": 1.1780420929122254e-05, "loss": 0.0147, "step": 63050 }, { "grad_norm": 0.18997018039226532, "learning_rate": 1.176709821428309e-05, "loss": 0.011, "step": 63060 }, { "grad_norm": 0.2073937952518463, "learning_rate": 1.1753782032401173e-05, "loss": 0.0135, "step": 63070 }, { "grad_norm": 0.23022860288619995, "learning_rate": 1.1740472385751866e-05, "loss": 0.0114, "step": 63080 }, { "grad_norm": 0.29831233620643616, "learning_rate": 1.172716927660943e-05, "loss": 0.0102, "step": 63090 }, { "grad_norm": 0.22257229685783386, "learning_rate": 1.1713872707247015e-05, "loss": 0.0139, "step": 63100 }, { "grad_norm": 0.09699058532714844, "learning_rate": 1.1700582679936595e-05, "loss": 0.0099, "step": 63110 }, { "grad_norm": 0.20450760424137115, "learning_rate": 1.1687299196949136e-05, "loss": 0.0151, "step": 63120 }, { "grad_norm": 0.23675262928009033, "learning_rate": 1.1674022260554374e-05, "loss": 0.0115, "step": 63130 }, { "grad_norm": 0.1758752465248108, "learning_rate": 1.1660751873020987e-05, "loss": 0.012, "step": 63140 }, { "grad_norm": 0.23773270845413208, "learning_rate": 1.1647488036616538e-05, "loss": 0.0127, "step": 63150 }, { "grad_norm": 0.508956253528595, "learning_rate": 1.1634230753607417e-05, "loss": 0.0147, "step": 63160 }, { "grad_norm": 0.1970503032207489, "learning_rate": 1.1620980026258982e-05, "loss": 0.0136, "step": 63170 }, { "grad_norm": 0.21031174063682556, "learning_rate": 1.1607735856835373e-05, "loss": 0.0096, "step": 63180 }, { "grad_norm": 0.20541901886463165, "learning_rate": 1.1594498247599677e-05, "loss": 0.0087, "step": 63190 }, { "grad_norm": 0.2004765123128891, "learning_rate": 1.158126720081384e-05, "loss": 0.0141, "step": 63200 }, { "grad_norm": 0.2037145495414734, "learning_rate": 1.156804271873868e-05, "loss": 0.0127, "step": 63210 }, { "grad_norm": 0.14189405739307404, "learning_rate": 1.1554824803633924e-05, "loss": 0.0132, "step": 63220 }, { "grad_norm": 0.42054030299186707, "learning_rate": 1.154161345775811e-05, "loss": 0.0098, "step": 63230 }, { "grad_norm": 0.15437211096286774, "learning_rate": 1.1528408683368724e-05, "loss": 0.0131, "step": 63240 }, { "grad_norm": 0.2608034908771515, "learning_rate": 1.1515210482722088e-05, "loss": 0.0117, "step": 63250 }, { "grad_norm": 0.4280223846435547, "learning_rate": 1.150201885807342e-05, "loss": 0.0162, "step": 63260 }, { "grad_norm": 0.1615166962146759, "learning_rate": 1.1488833811676807e-05, "loss": 0.0097, "step": 63270 }, { "grad_norm": 0.17882458865642548, "learning_rate": 1.1475655345785213e-05, "loss": 0.011, "step": 63280 }, { "grad_norm": 0.12699803709983826, "learning_rate": 1.1462483462650481e-05, "loss": 0.0102, "step": 63290 }, { "grad_norm": 2.1093909740448, "learning_rate": 1.1449318164523331e-05, "loss": 0.0156, "step": 63300 }, { "grad_norm": 0.14849241077899933, "learning_rate": 1.1436159453653334e-05, "loss": 0.0099, "step": 63310 }, { "grad_norm": 0.12598605453968048, "learning_rate": 1.1423007332288955e-05, "loss": 0.0143, "step": 63320 }, { "grad_norm": 0.27474069595336914, "learning_rate": 1.1409861802677546e-05, "loss": 0.0114, "step": 63330 }, { "grad_norm": 0.14798617362976074, "learning_rate": 1.1396722867065313e-05, "loss": 0.0081, "step": 63340 }, { "grad_norm": 0.20703577995300293, "learning_rate": 1.1383590527697352e-05, "loss": 0.0133, "step": 63350 }, { "grad_norm": 0.24302539229393005, "learning_rate": 1.137046478681758e-05, "loss": 0.01, "step": 63360 }, { "grad_norm": 0.36935362219810486, "learning_rate": 1.1357345646668888e-05, "loss": 0.0246, "step": 63370 }, { "grad_norm": 0.16369472444057465, "learning_rate": 1.1344233109492924e-05, "loss": 0.0138, "step": 63380 }, { "grad_norm": 0.13805072009563446, "learning_rate": 1.1331127177530292e-05, "loss": 0.0139, "step": 63390 }, { "grad_norm": 0.4031008780002594, "learning_rate": 1.1318027853020441e-05, "loss": 0.0132, "step": 63400 }, { "grad_norm": 0.30045971274375916, "learning_rate": 1.1304935138201645e-05, "loss": 0.0124, "step": 63410 }, { "grad_norm": 4.519707679748535, "learning_rate": 1.1291849035311153e-05, "loss": 0.0116, "step": 63420 }, { "grad_norm": 0.2289082258939743, "learning_rate": 1.1278769546584972e-05, "loss": 0.0112, "step": 63430 }, { "grad_norm": 0.5005783438682556, "learning_rate": 1.1265696674258052e-05, "loss": 0.0096, "step": 63440 }, { "grad_norm": 0.7452289462089539, "learning_rate": 1.1252630420564186e-05, "loss": 0.0104, "step": 63450 }, { "grad_norm": 0.22778862714767456, "learning_rate": 1.1239570787736036e-05, "loss": 0.01, "step": 63460 }, { "grad_norm": 0.18507897853851318, "learning_rate": 1.1226517778005135e-05, "loss": 0.0113, "step": 63470 }, { "grad_norm": 0.18956641852855682, "learning_rate": 1.1213471393601893e-05, "loss": 0.0094, "step": 63480 }, { "grad_norm": 0.20603100955486298, "learning_rate": 1.1200431636755587e-05, "loss": 0.0141, "step": 63490 }, { "grad_norm": 0.23240125179290771, "learning_rate": 1.1187398509694336e-05, "loss": 0.0143, "step": 63500 }, { "grad_norm": 0.17568987607955933, "learning_rate": 1.1174372014645146e-05, "loss": 0.0087, "step": 63510 }, { "grad_norm": 0.2815225422382355, "learning_rate": 1.1161352153833899e-05, "loss": 0.0169, "step": 63520 }, { "grad_norm": 3.116011619567871, "learning_rate": 1.1148338929485325e-05, "loss": 0.0107, "step": 63530 }, { "grad_norm": 0.6484834551811218, "learning_rate": 1.113533234382304e-05, "loss": 0.0082, "step": 63540 }, { "grad_norm": 0.2294875979423523, "learning_rate": 1.1122332399069513e-05, "loss": 0.0118, "step": 63550 }, { "grad_norm": 0.2107953131198883, "learning_rate": 1.1109339097446047e-05, "loss": 0.0121, "step": 63560 }, { "grad_norm": 0.26872581243515015, "learning_rate": 1.1096352441172897e-05, "loss": 0.015, "step": 63570 }, { "grad_norm": 0.24348299205303192, "learning_rate": 1.1083372432469086e-05, "loss": 0.0102, "step": 63580 }, { "grad_norm": 0.25130414962768555, "learning_rate": 1.107039907355255e-05, "loss": 0.0143, "step": 63590 }, { "grad_norm": 0.2818983197212219, "learning_rate": 1.1057432366640103e-05, "loss": 0.0145, "step": 63600 }, { "grad_norm": 0.30831581354141235, "learning_rate": 1.1044472313947352e-05, "loss": 0.0105, "step": 63610 }, { "grad_norm": 0.16655592620372772, "learning_rate": 1.1031518917688877e-05, "loss": 0.0103, "step": 63620 }, { "grad_norm": 0.12395914644002914, "learning_rate": 1.1018572180078007e-05, "loss": 0.0185, "step": 63630 }, { "grad_norm": 0.23991873860359192, "learning_rate": 1.1005632103327018e-05, "loss": 0.0102, "step": 63640 }, { "grad_norm": 0.2237478345632553, "learning_rate": 1.0992698689646996e-05, "loss": 0.0141, "step": 63650 }, { "grad_norm": 0.195643812417984, "learning_rate": 1.0979771941247919e-05, "loss": 0.0153, "step": 63660 }, { "grad_norm": 0.23445644974708557, "learning_rate": 1.0966851860338611e-05, "loss": 0.0135, "step": 63670 }, { "grad_norm": 0.286918044090271, "learning_rate": 1.0953938449126766e-05, "loss": 0.0143, "step": 63680 }, { "grad_norm": 0.3489987552165985, "learning_rate": 1.0941031709818933e-05, "loss": 0.022, "step": 63690 }, { "grad_norm": 0.5225656032562256, "learning_rate": 1.0928131644620509e-05, "loss": 0.0189, "step": 63700 }, { "grad_norm": 0.5741316080093384, "learning_rate": 1.0915238255735766e-05, "loss": 0.0133, "step": 63710 }, { "grad_norm": 0.4925846755504608, "learning_rate": 1.0902351545367833e-05, "loss": 0.0117, "step": 63720 }, { "grad_norm": 0.18028239905834198, "learning_rate": 1.0889471515718702e-05, "loss": 0.0115, "step": 63730 }, { "grad_norm": 0.15727315843105316, "learning_rate": 1.087659816898921e-05, "loss": 0.0107, "step": 63740 }, { "grad_norm": 0.2810782194137573, "learning_rate": 1.0863731507379082e-05, "loss": 0.0139, "step": 63750 }, { "grad_norm": 0.29971858859062195, "learning_rate": 1.0850871533086827e-05, "loss": 0.0136, "step": 63760 }, { "grad_norm": 0.3112790584564209, "learning_rate": 1.0838018248309927e-05, "loss": 0.0093, "step": 63770 }, { "grad_norm": 0.4989219307899475, "learning_rate": 1.0825171655244615e-05, "loss": 0.0168, "step": 63780 }, { "grad_norm": 0.2061978578567505, "learning_rate": 1.0812331756086025e-05, "loss": 0.0116, "step": 63790 }, { "grad_norm": 0.24854397773742676, "learning_rate": 1.079949855302817e-05, "loss": 0.0127, "step": 63800 }, { "grad_norm": 0.21664389967918396, "learning_rate": 1.0786672048263852e-05, "loss": 0.0141, "step": 63810 }, { "grad_norm": 0.18709196150302887, "learning_rate": 1.0773852243984817e-05, "loss": 0.0101, "step": 63820 }, { "grad_norm": 0.17721867561340332, "learning_rate": 1.0761039142381586e-05, "loss": 0.0159, "step": 63830 }, { "grad_norm": 0.38693493604660034, "learning_rate": 1.0748232745643577e-05, "loss": 0.0106, "step": 63840 }, { "grad_norm": 0.2842578887939453, "learning_rate": 1.0735433055959055e-05, "loss": 0.0146, "step": 63850 }, { "grad_norm": 0.1722826361656189, "learning_rate": 1.0722640075515133e-05, "loss": 0.0106, "step": 63860 }, { "grad_norm": 0.3984491229057312, "learning_rate": 1.0709853806497795e-05, "loss": 0.0113, "step": 63870 }, { "grad_norm": 0.22496631741523743, "learning_rate": 1.0697074251091831e-05, "loss": 0.0101, "step": 63880 }, { "grad_norm": 0.09802662581205368, "learning_rate": 1.0684301411480962e-05, "loss": 0.011, "step": 63890 }, { "grad_norm": 0.1608787328004837, "learning_rate": 1.067153528984769e-05, "loss": 0.0118, "step": 63900 }, { "grad_norm": 0.20759186148643494, "learning_rate": 1.0658775888373395e-05, "loss": 0.0107, "step": 63910 }, { "grad_norm": 0.23178289830684662, "learning_rate": 1.0646023209238314e-05, "loss": 0.0132, "step": 63920 }, { "grad_norm": 0.1379113793373108, "learning_rate": 1.0633277254621537e-05, "loss": 0.0125, "step": 63930 }, { "grad_norm": 0.17773227393627167, "learning_rate": 1.0620538026700994e-05, "loss": 0.0106, "step": 63940 }, { "grad_norm": 0.7919522523880005, "learning_rate": 1.0607805527653486e-05, "loss": 0.0124, "step": 63950 }, { "grad_norm": 1.077813982963562, "learning_rate": 1.059507975965462e-05, "loss": 0.0104, "step": 63960 }, { "grad_norm": 0.17745883762836456, "learning_rate": 1.0582360724878898e-05, "loss": 0.0084, "step": 63970 }, { "grad_norm": 0.19851474463939667, "learning_rate": 1.0569648425499651e-05, "loss": 0.013, "step": 63980 }, { "grad_norm": 0.16586436331272125, "learning_rate": 1.0556942863689063e-05, "loss": 0.0125, "step": 63990 }, { "grad_norm": 0.1975862830877304, "learning_rate": 1.054424404161819e-05, "loss": 0.009, "step": 64000 }, { "grad_norm": 0.27365973591804504, "learning_rate": 1.053155196145686e-05, "loss": 0.01, "step": 64010 }, { "grad_norm": 0.4884350895881653, "learning_rate": 1.0518866625373863e-05, "loss": 0.0113, "step": 64020 }, { "grad_norm": 0.23648881912231445, "learning_rate": 1.0506188035536735e-05, "loss": 0.0095, "step": 64030 }, { "grad_norm": 0.10660717636346817, "learning_rate": 1.0493516194111919e-05, "loss": 0.0114, "step": 64040 }, { "grad_norm": 0.145731583237648, "learning_rate": 1.0480851103264688e-05, "loss": 0.0109, "step": 64050 }, { "grad_norm": 0.21386472880840302, "learning_rate": 1.046819276515913e-05, "loss": 0.0098, "step": 64060 }, { "grad_norm": 0.19078680872917175, "learning_rate": 1.0455541181958256e-05, "loss": 0.0112, "step": 64070 }, { "grad_norm": 0.17566317319869995, "learning_rate": 1.0442896355823822e-05, "loss": 0.0104, "step": 64080 }, { "grad_norm": 0.13495531678199768, "learning_rate": 1.0430258288916539e-05, "loss": 0.0133, "step": 64090 }, { "grad_norm": 0.29563504457473755, "learning_rate": 1.0417626983395868e-05, "loss": 0.0109, "step": 64100 }, { "grad_norm": 0.1852491945028305, "learning_rate": 1.0405002441420165e-05, "loss": 0.0092, "step": 64110 }, { "grad_norm": 0.20223724842071533, "learning_rate": 1.039238466514662e-05, "loss": 0.0119, "step": 64120 }, { "grad_norm": 0.19562534987926483, "learning_rate": 1.0379773656731262e-05, "loss": 0.0147, "step": 64130 }, { "grad_norm": 0.29174327850341797, "learning_rate": 1.0367169418328986e-05, "loss": 0.0088, "step": 64140 }, { "grad_norm": 0.3156810402870178, "learning_rate": 1.0354571952093484e-05, "loss": 0.0136, "step": 64150 }, { "grad_norm": 0.17239384353160858, "learning_rate": 1.034198126017733e-05, "loss": 0.0094, "step": 64160 }, { "grad_norm": 0.19501258432865143, "learning_rate": 1.0329397344731928e-05, "loss": 0.011, "step": 64170 }, { "grad_norm": 0.18833735585212708, "learning_rate": 1.0316820207907524e-05, "loss": 0.012, "step": 64180 }, { "grad_norm": 0.12829796969890594, "learning_rate": 1.0304249851853215e-05, "loss": 0.0107, "step": 64190 }, { "grad_norm": 0.19666442275047302, "learning_rate": 1.0291686278716933e-05, "loss": 0.0141, "step": 64200 }, { "grad_norm": 0.34660500288009644, "learning_rate": 1.0279129490645418e-05, "loss": 0.0117, "step": 64210 }, { "grad_norm": 0.16028770804405212, "learning_rate": 1.0266579489784328e-05, "loss": 0.0097, "step": 64220 }, { "grad_norm": 0.26651301980018616, "learning_rate": 1.0254036278278084e-05, "loss": 0.0102, "step": 64230 }, { "grad_norm": 0.22876380383968353, "learning_rate": 1.0241499858269982e-05, "loss": 0.0107, "step": 64240 }, { "grad_norm": 0.2727087140083313, "learning_rate": 1.0228970231902169e-05, "loss": 0.0091, "step": 64250 }, { "grad_norm": 0.1591852605342865, "learning_rate": 1.0216447401315582e-05, "loss": 0.0097, "step": 64260 }, { "grad_norm": 0.19052885472774506, "learning_rate": 1.020393136865007e-05, "loss": 0.0077, "step": 64270 }, { "grad_norm": 0.31335729360580444, "learning_rate": 1.0191422136044242e-05, "loss": 0.0213, "step": 64280 }, { "grad_norm": 0.24996905028820038, "learning_rate": 1.017891970563563e-05, "loss": 0.0124, "step": 64290 }, { "grad_norm": 1.0251688957214355, "learning_rate": 1.0166424079560516e-05, "loss": 0.0156, "step": 64300 }, { "grad_norm": 0.2579709589481354, "learning_rate": 1.0153935259954078e-05, "loss": 0.0114, "step": 64310 }, { "grad_norm": 0.3193080723285675, "learning_rate": 1.0141453248950311e-05, "loss": 0.0145, "step": 64320 }, { "grad_norm": 0.2446037232875824, "learning_rate": 1.0128978048682054e-05, "loss": 0.0138, "step": 64330 }, { "grad_norm": 0.29171377420425415, "learning_rate": 1.0116509661280982e-05, "loss": 0.0179, "step": 64340 }, { "grad_norm": 0.322128564119339, "learning_rate": 1.0104048088877576e-05, "loss": 0.0099, "step": 64350 }, { "grad_norm": 0.3167470395565033, "learning_rate": 1.0091593333601201e-05, "loss": 0.0131, "step": 64360 }, { "grad_norm": 0.174276664853096, "learning_rate": 1.0079145397580031e-05, "loss": 0.0127, "step": 64370 }, { "grad_norm": 0.30594587326049805, "learning_rate": 1.006670428294107e-05, "loss": 0.0156, "step": 64380 }, { "grad_norm": 0.22867654263973236, "learning_rate": 1.0054269991810166e-05, "loss": 0.0108, "step": 64390 }, { "grad_norm": 0.17203424870967865, "learning_rate": 1.0041842526312024e-05, "loss": 0.0123, "step": 64400 }, { "grad_norm": 0.18304255604743958, "learning_rate": 1.0029421888570101e-05, "loss": 0.0082, "step": 64410 }, { "grad_norm": 0.18097560107707977, "learning_rate": 1.0017008080706813e-05, "loss": 0.02, "step": 64420 }, { "grad_norm": 0.18584232032299042, "learning_rate": 1.0004601104843287e-05, "loss": 0.0111, "step": 64430 }, { "grad_norm": 0.11936480551958084, "learning_rate": 9.992200963099562e-06, "loss": 0.0115, "step": 64440 }, { "grad_norm": 0.6070599555969238, "learning_rate": 9.979807657594486e-06, "loss": 0.0115, "step": 64450 }, { "grad_norm": 0.2173444926738739, "learning_rate": 9.967421190445703e-06, "loss": 0.0092, "step": 64460 }, { "grad_norm": 0.22556523978710175, "learning_rate": 9.955041563769769e-06, "loss": 0.0129, "step": 64470 }, { "grad_norm": 0.3810001015663147, "learning_rate": 9.942668779681974e-06, "loss": 0.0188, "step": 64480 }, { "grad_norm": 0.1904875785112381, "learning_rate": 9.930302840296541e-06, "loss": 0.0109, "step": 64490 }, { "grad_norm": 0.18199564516544342, "learning_rate": 9.917943747726426e-06, "loss": 0.0087, "step": 64500 }, { "grad_norm": 0.20691440999507904, "learning_rate": 9.905591504083484e-06, "loss": 0.0108, "step": 64510 }, { "grad_norm": 0.16131716966629028, "learning_rate": 9.893246111478382e-06, "loss": 0.0103, "step": 64520 }, { "grad_norm": 0.31907546520233154, "learning_rate": 9.88090757202057e-06, "loss": 0.0133, "step": 64530 }, { "grad_norm": 0.2767044007778168, "learning_rate": 9.868575887818421e-06, "loss": 0.0119, "step": 64540 }, { "grad_norm": 0.2206254005432129, "learning_rate": 9.856251060979044e-06, "loss": 0.0101, "step": 64550 }, { "grad_norm": 0.1555090993642807, "learning_rate": 9.843933093608426e-06, "loss": 0.0124, "step": 64560 }, { "grad_norm": 0.22872595489025116, "learning_rate": 9.831621987811368e-06, "loss": 0.014, "step": 64570 }, { "grad_norm": 0.25344401597976685, "learning_rate": 9.819317745691509e-06, "loss": 0.0105, "step": 64580 }, { "grad_norm": 0.24019457399845123, "learning_rate": 9.8070203693513e-06, "loss": 0.0111, "step": 64590 }, { "grad_norm": 0.25013092160224915, "learning_rate": 9.794729860892048e-06, "loss": 0.0094, "step": 64600 }, { "grad_norm": 0.33055734634399414, "learning_rate": 9.782446222413827e-06, "loss": 0.0224, "step": 64610 }, { "grad_norm": 0.10590548068284988, "learning_rate": 9.770169456015598e-06, "loss": 0.0144, "step": 64620 }, { "grad_norm": 0.1585545390844345, "learning_rate": 9.75789956379512e-06, "loss": 0.0121, "step": 64630 }, { "grad_norm": 0.1615108847618103, "learning_rate": 9.74563654784898e-06, "loss": 0.009, "step": 64640 }, { "grad_norm": 0.19768938422203064, "learning_rate": 9.733380410272596e-06, "loss": 0.0117, "step": 64650 }, { "grad_norm": 0.2345661222934723, "learning_rate": 9.721131153160207e-06, "loss": 0.0115, "step": 64660 }, { "grad_norm": 0.19320042431354523, "learning_rate": 9.708888778604886e-06, "loss": 0.0099, "step": 64670 }, { "grad_norm": 0.1820460855960846, "learning_rate": 9.696653288698488e-06, "loss": 0.0095, "step": 64680 }, { "grad_norm": 0.39416050910949707, "learning_rate": 9.684424685531763e-06, "loss": 0.0174, "step": 64690 }, { "grad_norm": 0.16929912567138672, "learning_rate": 9.672202971194216e-06, "loss": 0.0138, "step": 64700 }, { "grad_norm": 0.11888871341943741, "learning_rate": 9.659988147774213e-06, "loss": 0.0107, "step": 64710 }, { "grad_norm": 0.1492615044116974, "learning_rate": 9.647780217358942e-06, "loss": 0.0148, "step": 64720 }, { "grad_norm": 0.13090912997722626, "learning_rate": 9.635579182034376e-06, "loss": 0.0106, "step": 64730 }, { "grad_norm": 0.19719186425209045, "learning_rate": 9.623385043885386e-06, "loss": 0.0153, "step": 64740 }, { "grad_norm": 0.1793718934059143, "learning_rate": 9.61119780499557e-06, "loss": 0.0112, "step": 64750 }, { "grad_norm": 0.1786314696073532, "learning_rate": 9.599017467447418e-06, "loss": 0.0129, "step": 64760 }, { "grad_norm": 0.2327021211385727, "learning_rate": 9.586844033322206e-06, "loss": 0.015, "step": 64770 }, { "grad_norm": 0.23790961503982544, "learning_rate": 9.574677504700052e-06, "loss": 0.0107, "step": 64780 }, { "grad_norm": 5.096726417541504, "learning_rate": 9.562517883659877e-06, "loss": 0.0174, "step": 64790 }, { "grad_norm": 0.15325036644935608, "learning_rate": 9.55036517227944e-06, "loss": 0.0137, "step": 64800 }, { "grad_norm": 0.15867632627487183, "learning_rate": 9.538219372635282e-06, "loss": 0.0096, "step": 64810 }, { "grad_norm": 0.23679834604263306, "learning_rate": 9.526080486802802e-06, "loss": 0.012, "step": 64820 }, { "grad_norm": 0.15191864967346191, "learning_rate": 9.513948516856203e-06, "loss": 0.0109, "step": 64830 }, { "grad_norm": 0.41955646872520447, "learning_rate": 9.501823464868504e-06, "loss": 0.0095, "step": 64840 }, { "grad_norm": 0.26782792806625366, "learning_rate": 9.489705332911547e-06, "loss": 0.01, "step": 64850 }, { "grad_norm": 0.20925010740756989, "learning_rate": 9.477594123055994e-06, "loss": 0.0104, "step": 64860 }, { "grad_norm": 0.18802610039710999, "learning_rate": 9.465489837371321e-06, "loss": 0.0157, "step": 64870 }, { "grad_norm": 0.31247299909591675, "learning_rate": 9.453392477925794e-06, "loss": 0.0096, "step": 64880 }, { "grad_norm": 0.1759735643863678, "learning_rate": 9.441302046786566e-06, "loss": 0.0112, "step": 64890 }, { "grad_norm": 0.13628199696540833, "learning_rate": 9.429218546019519e-06, "loss": 0.0089, "step": 64900 }, { "grad_norm": 0.28010398149490356, "learning_rate": 9.41714197768941e-06, "loss": 0.0155, "step": 64910 }, { "grad_norm": 0.24495573341846466, "learning_rate": 9.405072343859805e-06, "loss": 0.0118, "step": 64920 }, { "grad_norm": 0.20284436643123627, "learning_rate": 9.393009646593043e-06, "loss": 0.0138, "step": 64930 }, { "grad_norm": 0.21834589540958405, "learning_rate": 9.38095388795035e-06, "loss": 0.014, "step": 64940 }, { "grad_norm": 1.7132916450500488, "learning_rate": 9.36890506999169e-06, "loss": 0.014, "step": 64950 }, { "grad_norm": 0.17345164716243744, "learning_rate": 9.356863194775894e-06, "loss": 0.0132, "step": 64960 }, { "grad_norm": 0.17440856993198395, "learning_rate": 9.344828264360583e-06, "loss": 0.0122, "step": 64970 }, { "grad_norm": 0.17039363086223602, "learning_rate": 9.332800280802201e-06, "loss": 0.01, "step": 64980 }, { "grad_norm": 0.17298254370689392, "learning_rate": 9.32077924615602e-06, "loss": 0.0075, "step": 64990 }, { "grad_norm": 0.2608230710029602, "learning_rate": 9.308765162476063e-06, "loss": 0.011, "step": 65000 }, { "grad_norm": 1.4214155673980713, "learning_rate": 9.296758031815239e-06, "loss": 0.0157, "step": 65010 }, { "grad_norm": 0.19179779291152954, "learning_rate": 9.284757856225229e-06, "loss": 0.0113, "step": 65020 }, { "grad_norm": 0.19041863083839417, "learning_rate": 9.272764637756538e-06, "loss": 0.016, "step": 65030 }, { "grad_norm": 0.1992625743150711, "learning_rate": 9.260778378458479e-06, "loss": 0.0118, "step": 65040 }, { "grad_norm": 0.12416427582502365, "learning_rate": 9.248799080379172e-06, "loss": 0.0112, "step": 65050 }, { "grad_norm": 2.4188601970672607, "learning_rate": 9.236826745565558e-06, "loss": 0.014, "step": 65060 }, { "grad_norm": 0.4874115586280823, "learning_rate": 9.224861376063388e-06, "loss": 0.0146, "step": 65070 }, { "grad_norm": 0.1910007894039154, "learning_rate": 9.212902973917192e-06, "loss": 0.0091, "step": 65080 }, { "grad_norm": 0.2110128551721573, "learning_rate": 9.20095154117035e-06, "loss": 0.0106, "step": 65090 }, { "grad_norm": 0.507356584072113, "learning_rate": 9.189007079865036e-06, "loss": 0.0116, "step": 65100 }, { "grad_norm": 0.27505847811698914, "learning_rate": 9.177069592042226e-06, "loss": 0.0098, "step": 65110 }, { "grad_norm": 0.22859357297420502, "learning_rate": 9.165139079741724e-06, "loss": 0.0155, "step": 65120 }, { "grad_norm": 0.3135814368724823, "learning_rate": 9.153215545002098e-06, "loss": 0.0145, "step": 65130 }, { "grad_norm": 0.2123027741909027, "learning_rate": 9.141298989860798e-06, "loss": 0.0145, "step": 65140 }, { "grad_norm": 2.6358065605163574, "learning_rate": 9.129389416353994e-06, "loss": 0.0107, "step": 65150 }, { "grad_norm": 0.727959394454956, "learning_rate": 9.11748682651673e-06, "loss": 0.0094, "step": 65160 }, { "grad_norm": 0.22341714799404144, "learning_rate": 9.105591222382837e-06, "loss": 0.013, "step": 65170 }, { "grad_norm": 0.1476607620716095, "learning_rate": 9.093702605984915e-06, "loss": 0.0099, "step": 65180 }, { "grad_norm": 0.24608151614665985, "learning_rate": 9.081820979354455e-06, "loss": 0.0111, "step": 65190 }, { "grad_norm": 0.3203660547733307, "learning_rate": 9.069946344521663e-06, "loss": 0.0115, "step": 65200 }, { "grad_norm": 0.3053346276283264, "learning_rate": 9.058078703515598e-06, "loss": 0.0117, "step": 65210 }, { "grad_norm": 0.23243595659732819, "learning_rate": 9.046218058364125e-06, "loss": 0.014, "step": 65220 }, { "grad_norm": 0.16652733087539673, "learning_rate": 9.034364411093893e-06, "loss": 0.0125, "step": 65230 }, { "grad_norm": 0.23470132052898407, "learning_rate": 9.022517763730371e-06, "loss": 0.0112, "step": 65240 }, { "grad_norm": 0.14350004494190216, "learning_rate": 9.010678118297827e-06, "loss": 0.0129, "step": 65250 }, { "grad_norm": 0.17636170983314514, "learning_rate": 8.998845476819345e-06, "loss": 0.0113, "step": 65260 }, { "grad_norm": 0.44192031025886536, "learning_rate": 8.987019841316773e-06, "loss": 0.013, "step": 65270 }, { "grad_norm": 0.3348971903324127, "learning_rate": 8.975201213810802e-06, "loss": 0.0101, "step": 65280 }, { "grad_norm": 0.184496209025383, "learning_rate": 8.963389596320915e-06, "loss": 0.0109, "step": 65290 }, { "grad_norm": 0.27306732535362244, "learning_rate": 8.951584990865391e-06, "loss": 0.0098, "step": 65300 }, { "grad_norm": 0.3215599060058594, "learning_rate": 8.939787399461319e-06, "loss": 0.014, "step": 65310 }, { "grad_norm": 0.30487218499183655, "learning_rate": 8.927996824124591e-06, "loss": 0.0173, "step": 65320 }, { "grad_norm": 0.7571775913238525, "learning_rate": 8.916213266869854e-06, "loss": 0.0143, "step": 65330 }, { "grad_norm": 0.3324112892150879, "learning_rate": 8.904436729710658e-06, "loss": 0.0103, "step": 65340 }, { "grad_norm": 0.2617427408695221, "learning_rate": 8.892667214659245e-06, "loss": 0.0098, "step": 65350 }, { "grad_norm": 0.20284584164619446, "learning_rate": 8.880904723726713e-06, "loss": 0.0133, "step": 65360 }, { "grad_norm": 0.23634600639343262, "learning_rate": 8.869149258922971e-06, "loss": 0.0085, "step": 65370 }, { "grad_norm": 0.781433641910553, "learning_rate": 8.857400822256662e-06, "loss": 0.0096, "step": 65380 }, { "grad_norm": 0.13693654537200928, "learning_rate": 8.845659415735324e-06, "loss": 0.0169, "step": 65390 }, { "grad_norm": 0.19829562306404114, "learning_rate": 8.83392504136521e-06, "loss": 0.0102, "step": 65400 }, { "grad_norm": 0.299676775932312, "learning_rate": 8.822197701151407e-06, "loss": 0.0108, "step": 65410 }, { "grad_norm": 0.3023393750190735, "learning_rate": 8.810477397097804e-06, "loss": 0.0139, "step": 65420 }, { "grad_norm": 0.16981911659240723, "learning_rate": 8.798764131207077e-06, "loss": 0.0108, "step": 65430 }, { "grad_norm": 0.234358549118042, "learning_rate": 8.787057905480706e-06, "loss": 0.0143, "step": 65440 }, { "grad_norm": 0.1778605878353119, "learning_rate": 8.775358721918958e-06, "loss": 0.0128, "step": 65450 }, { "grad_norm": 0.2335875779390335, "learning_rate": 8.763666582520923e-06, "loss": 0.0123, "step": 65460 }, { "grad_norm": 0.15318459272384644, "learning_rate": 8.751981489284445e-06, "loss": 0.01, "step": 65470 }, { "grad_norm": 0.27922192215919495, "learning_rate": 8.740303444206188e-06, "loss": 0.0117, "step": 65480 }, { "grad_norm": 0.21541275084018707, "learning_rate": 8.72863244928162e-06, "loss": 0.0102, "step": 65490 }, { "grad_norm": 0.2185908854007721, "learning_rate": 8.716968506504991e-06, "loss": 0.0154, "step": 65500 }, { "grad_norm": 0.17837879061698914, "learning_rate": 8.70531161786935e-06, "loss": 0.0096, "step": 65510 }, { "grad_norm": 0.13002119958400726, "learning_rate": 8.693661785366558e-06, "loss": 0.0149, "step": 65520 }, { "grad_norm": 0.17758779227733612, "learning_rate": 8.682019010987208e-06, "loss": 0.0131, "step": 65530 }, { "grad_norm": 0.19120272994041443, "learning_rate": 8.670383296720786e-06, "loss": 0.0115, "step": 65540 }, { "grad_norm": 1.2940865755081177, "learning_rate": 8.658754644555478e-06, "loss": 0.0124, "step": 65550 }, { "grad_norm": 0.21589794754981995, "learning_rate": 8.647133056478313e-06, "loss": 0.0111, "step": 65560 }, { "grad_norm": 0.5287905931472778, "learning_rate": 8.635518534475123e-06, "loss": 0.0091, "step": 65570 }, { "grad_norm": 0.14063912630081177, "learning_rate": 8.623911080530467e-06, "loss": 0.0112, "step": 65580 }, { "grad_norm": 0.27187708020210266, "learning_rate": 8.61231069662779e-06, "loss": 0.0116, "step": 65590 }, { "grad_norm": 0.178871750831604, "learning_rate": 8.600717384749252e-06, "loss": 0.0094, "step": 65600 }, { "grad_norm": 0.19094756245613098, "learning_rate": 8.58913114687584e-06, "loss": 0.0081, "step": 65610 }, { "grad_norm": 0.29062619805336, "learning_rate": 8.57755198498732e-06, "loss": 0.0123, "step": 65620 }, { "grad_norm": 0.212113618850708, "learning_rate": 8.565979901062265e-06, "loss": 0.0218, "step": 65630 }, { "grad_norm": 0.2735804319381714, "learning_rate": 8.554414897078033e-06, "loss": 0.0092, "step": 65640 }, { "grad_norm": 0.24049502611160278, "learning_rate": 8.542856975010727e-06, "loss": 0.0093, "step": 65650 }, { "grad_norm": 0.26488086581230164, "learning_rate": 8.531306136835337e-06, "loss": 0.0096, "step": 65660 }, { "grad_norm": 0.10240689665079117, "learning_rate": 8.519762384525548e-06, "loss": 0.0097, "step": 65670 }, { "grad_norm": 0.5813885927200317, "learning_rate": 8.508225720053875e-06, "loss": 0.0099, "step": 65680 }, { "grad_norm": 0.22417870163917542, "learning_rate": 8.496696145391625e-06, "loss": 0.0087, "step": 65690 }, { "grad_norm": 0.17949847877025604, "learning_rate": 8.485173662508889e-06, "loss": 0.0089, "step": 65700 }, { "grad_norm": 0.14352482557296753, "learning_rate": 8.473658273374536e-06, "loss": 0.0104, "step": 65710 }, { "grad_norm": 0.17248843610286713, "learning_rate": 8.462149979956253e-06, "loss": 0.0103, "step": 65720 }, { "grad_norm": 0.16595935821533203, "learning_rate": 8.450648784220461e-06, "loss": 0.0128, "step": 65730 }, { "grad_norm": 0.2133156955242157, "learning_rate": 8.439154688132417e-06, "loss": 0.0098, "step": 65740 }, { "grad_norm": 0.2795158922672272, "learning_rate": 8.427667693656143e-06, "loss": 0.0117, "step": 65750 }, { "grad_norm": 0.32526928186416626, "learning_rate": 8.416187802754454e-06, "loss": 0.0098, "step": 65760 }, { "grad_norm": 0.14532005786895752, "learning_rate": 8.404715017388965e-06, "loss": 0.0089, "step": 65770 }, { "grad_norm": 0.17028936743736267, "learning_rate": 8.393249339520015e-06, "loss": 0.0117, "step": 65780 }, { "grad_norm": 0.23103398084640503, "learning_rate": 8.381790771106834e-06, "loss": 0.0083, "step": 65790 }, { "grad_norm": 0.9871547818183899, "learning_rate": 8.370339314107339e-06, "loss": 0.01, "step": 65800 }, { "grad_norm": 0.42937415838241577, "learning_rate": 8.358894970478281e-06, "loss": 0.0153, "step": 65810 }, { "grad_norm": 0.13368137180805206, "learning_rate": 8.347457742175196e-06, "loss": 0.0122, "step": 65820 }, { "grad_norm": 0.2088039517402649, "learning_rate": 8.33602763115236e-06, "loss": 0.0113, "step": 65830 }, { "grad_norm": 0.26864132285118103, "learning_rate": 8.324604639362916e-06, "loss": 0.0127, "step": 65840 }, { "grad_norm": 0.1546775996685028, "learning_rate": 8.31318876875869e-06, "loss": 0.0103, "step": 65850 }, { "grad_norm": 0.1988040655851364, "learning_rate": 8.30178002129039e-06, "loss": 0.0101, "step": 65860 }, { "grad_norm": 0.22668421268463135, "learning_rate": 8.290378398907423e-06, "loss": 0.0105, "step": 65870 }, { "grad_norm": 0.29718518257141113, "learning_rate": 8.278983903558029e-06, "loss": 0.0117, "step": 65880 }, { "grad_norm": 0.16408085823059082, "learning_rate": 8.26759653718921e-06, "loss": 0.0123, "step": 65890 }, { "grad_norm": 0.3019825220108032, "learning_rate": 8.25621630174676e-06, "loss": 0.0107, "step": 65900 }, { "grad_norm": 0.29722630977630615, "learning_rate": 8.244843199175261e-06, "loss": 0.0165, "step": 65910 }, { "grad_norm": 0.20772282779216766, "learning_rate": 8.233477231418046e-06, "loss": 0.0092, "step": 65920 }, { "grad_norm": 0.15793754160404205, "learning_rate": 8.22211840041725e-06, "loss": 0.0099, "step": 65930 }, { "grad_norm": 0.2926624119281769, "learning_rate": 8.210766708113792e-06, "loss": 0.015, "step": 65940 }, { "grad_norm": 0.20577573776245117, "learning_rate": 8.199422156447367e-06, "loss": 0.0116, "step": 65950 }, { "grad_norm": 0.11528152972459793, "learning_rate": 8.188084747356451e-06, "loss": 0.0106, "step": 65960 }, { "grad_norm": 0.2764969766139984, "learning_rate": 8.176754482778299e-06, "loss": 0.0105, "step": 65970 }, { "grad_norm": 0.3706948459148407, "learning_rate": 8.165431364648918e-06, "loss": 0.0135, "step": 65980 }, { "grad_norm": 0.40898293256759644, "learning_rate": 8.154115394903162e-06, "loss": 0.0116, "step": 65990 }, { "grad_norm": 0.21135295927524567, "learning_rate": 8.142806575474582e-06, "loss": 0.0137, "step": 66000 }, { "grad_norm": 0.12980498373508453, "learning_rate": 8.131504908295562e-06, "loss": 0.0076, "step": 66010 }, { "grad_norm": 0.17290934920310974, "learning_rate": 8.120210395297262e-06, "loss": 0.0068, "step": 66020 }, { "grad_norm": 0.2777569890022278, "learning_rate": 8.108923038409565e-06, "loss": 0.0121, "step": 66030 }, { "grad_norm": 0.22835834324359894, "learning_rate": 8.097642839561226e-06, "loss": 0.0109, "step": 66040 }, { "grad_norm": 0.22997848689556122, "learning_rate": 8.08636980067966e-06, "loss": 0.0122, "step": 66050 }, { "grad_norm": 0.15458837151527405, "learning_rate": 8.075103923691186e-06, "loss": 0.0079, "step": 66060 }, { "grad_norm": 0.3055761754512787, "learning_rate": 8.063845210520793e-06, "loss": 0.0089, "step": 66070 }, { "grad_norm": 0.38742467761039734, "learning_rate": 8.052593663092295e-06, "loss": 0.0129, "step": 66080 }, { "grad_norm": 0.1371811479330063, "learning_rate": 8.04134928332827e-06, "loss": 0.0091, "step": 66090 }, { "grad_norm": 0.18439242243766785, "learning_rate": 8.030112073150086e-06, "loss": 0.0133, "step": 66100 }, { "grad_norm": 0.21252569556236267, "learning_rate": 8.018882034477881e-06, "loss": 0.0102, "step": 66110 }, { "grad_norm": 0.19026358425617218, "learning_rate": 8.007659169230541e-06, "loss": 0.0089, "step": 66120 }, { "grad_norm": 0.18835054337978363, "learning_rate": 7.996443479325755e-06, "loss": 0.0095, "step": 66130 }, { "grad_norm": 0.11403535306453705, "learning_rate": 7.985234966679977e-06, "loss": 0.0115, "step": 66140 }, { "grad_norm": 0.4427872598171234, "learning_rate": 7.974033633208438e-06, "loss": 0.0145, "step": 66150 }, { "grad_norm": 3.279557704925537, "learning_rate": 7.962839480825135e-06, "loss": 0.0087, "step": 66160 }, { "grad_norm": 0.25162187218666077, "learning_rate": 7.951652511442858e-06, "loss": 0.0128, "step": 66170 }, { "grad_norm": 0.27783095836639404, "learning_rate": 7.940472726973125e-06, "loss": 0.0129, "step": 66180 }, { "grad_norm": 0.26160863041877747, "learning_rate": 7.929300129326289e-06, "loss": 0.0105, "step": 66190 }, { "grad_norm": 0.2446305751800537, "learning_rate": 7.91813472041142e-06, "loss": 0.0112, "step": 66200 }, { "grad_norm": 0.10620930045843124, "learning_rate": 7.906976502136376e-06, "loss": 0.0095, "step": 66210 }, { "grad_norm": 0.17617449164390564, "learning_rate": 7.89582547640782e-06, "loss": 0.0152, "step": 66220 }, { "grad_norm": 0.21457338333129883, "learning_rate": 7.884681645131115e-06, "loss": 0.0074, "step": 66230 }, { "grad_norm": 0.34448006749153137, "learning_rate": 7.87354501021048e-06, "loss": 0.0101, "step": 66240 }, { "grad_norm": 0.2102023959159851, "learning_rate": 7.86241557354882e-06, "loss": 0.0119, "step": 66250 }, { "grad_norm": 3.3517234325408936, "learning_rate": 7.8512933370479e-06, "loss": 0.0115, "step": 66260 }, { "grad_norm": 0.5507779717445374, "learning_rate": 7.840178302608158e-06, "loss": 0.0128, "step": 66270 }, { "grad_norm": 0.16962872445583344, "learning_rate": 7.82907047212888e-06, "loss": 0.0106, "step": 66280 }, { "grad_norm": 0.3564243018627167, "learning_rate": 7.81796984750809e-06, "loss": 0.0097, "step": 66290 }, { "grad_norm": 0.2022610902786255, "learning_rate": 7.806876430642546e-06, "loss": 0.0125, "step": 66300 }, { "grad_norm": 0.18549716472625732, "learning_rate": 7.795790223427862e-06, "loss": 0.0119, "step": 66310 }, { "grad_norm": 0.2513122260570526, "learning_rate": 7.784711227758324e-06, "loss": 0.0142, "step": 66320 }, { "grad_norm": 0.1470772922039032, "learning_rate": 7.773639445527053e-06, "loss": 0.0131, "step": 66330 }, { "grad_norm": 0.2068246454000473, "learning_rate": 7.762574878625905e-06, "loss": 0.0119, "step": 66340 }, { "grad_norm": 0.19666841626167297, "learning_rate": 7.751517528945513e-06, "loss": 0.0091, "step": 66350 }, { "grad_norm": 0.17011702060699463, "learning_rate": 7.740467398375278e-06, "loss": 0.011, "step": 66360 }, { "grad_norm": 0.3199268877506256, "learning_rate": 7.729424488803378e-06, "loss": 0.0087, "step": 66370 }, { "grad_norm": 0.1527780294418335, "learning_rate": 7.71838880211671e-06, "loss": 0.0131, "step": 66380 }, { "grad_norm": 0.19944420456886292, "learning_rate": 7.707360340200997e-06, "loss": 0.0081, "step": 66390 }, { "grad_norm": 0.2870611548423767, "learning_rate": 7.696339104940697e-06, "loss": 0.0119, "step": 66400 }, { "grad_norm": 0.25082746148109436, "learning_rate": 7.685325098219038e-06, "loss": 0.0099, "step": 66410 }, { "grad_norm": 0.15146520733833313, "learning_rate": 7.674318321918017e-06, "loss": 0.013, "step": 66420 }, { "grad_norm": 0.3501870632171631, "learning_rate": 7.663318777918366e-06, "loss": 0.0163, "step": 66430 }, { "grad_norm": 0.24018648266792297, "learning_rate": 7.652326468099647e-06, "loss": 0.0145, "step": 66440 }, { "grad_norm": 0.3163057565689087, "learning_rate": 7.641341394340096e-06, "loss": 0.0129, "step": 66450 }, { "grad_norm": 0.327591210603714, "learning_rate": 7.630363558516818e-06, "loss": 0.013, "step": 66460 }, { "grad_norm": 0.1774492710828781, "learning_rate": 7.619392962505578e-06, "loss": 0.0096, "step": 66470 }, { "grad_norm": 0.2645660936832428, "learning_rate": 7.6084296081809725e-06, "loss": 0.0148, "step": 66480 }, { "grad_norm": 0.2645414173603058, "learning_rate": 7.597473497416347e-06, "loss": 0.0122, "step": 66490 }, { "grad_norm": 0.1940375715494156, "learning_rate": 7.586524632083764e-06, "loss": 0.0098, "step": 66500 }, { "grad_norm": 0.36925268173217773, "learning_rate": 7.5755830140541326e-06, "loss": 0.013, "step": 66510 }, { "grad_norm": 0.110977403819561, "learning_rate": 7.5646486451970475e-06, "loss": 0.0105, "step": 66520 }, { "grad_norm": 0.13872183859348297, "learning_rate": 7.553721527380897e-06, "loss": 0.0121, "step": 66530 }, { "grad_norm": 0.19484496116638184, "learning_rate": 7.542801662472826e-06, "loss": 0.0092, "step": 66540 }, { "grad_norm": 0.20535939931869507, "learning_rate": 7.5318890523387475e-06, "loss": 0.0086, "step": 66550 }, { "grad_norm": 0.23825624585151672, "learning_rate": 7.52098369884332e-06, "loss": 0.0131, "step": 66560 }, { "grad_norm": 0.3118745684623718, "learning_rate": 7.510085603849992e-06, "loss": 0.0108, "step": 66570 }, { "grad_norm": 0.45760801434516907, "learning_rate": 7.499194769220918e-06, "loss": 0.0136, "step": 66580 }, { "grad_norm": 0.19472834467887878, "learning_rate": 7.48831119681706e-06, "loss": 0.0135, "step": 66590 }, { "grad_norm": 0.14103524386882782, "learning_rate": 7.477434888498119e-06, "loss": 0.0123, "step": 66600 }, { "grad_norm": 0.15022149682044983, "learning_rate": 7.466565846122564e-06, "loss": 0.0143, "step": 66610 }, { "grad_norm": 0.22572848200798035, "learning_rate": 7.455704071547626e-06, "loss": 0.0137, "step": 66620 }, { "grad_norm": 0.14712412655353546, "learning_rate": 7.444849566629247e-06, "loss": 0.0131, "step": 66630 }, { "grad_norm": 0.23455536365509033, "learning_rate": 7.434002333222212e-06, "loss": 0.0111, "step": 66640 }, { "grad_norm": 0.19685178995132446, "learning_rate": 7.423162373179976e-06, "loss": 0.0123, "step": 66650 }, { "grad_norm": 0.22152766585350037, "learning_rate": 7.412329688354835e-06, "loss": 0.0105, "step": 66660 }, { "grad_norm": 0.18211890757083893, "learning_rate": 7.40150428059776e-06, "loss": 0.0138, "step": 66670 }, { "grad_norm": 0.20844079554080963, "learning_rate": 7.3906861517585354e-06, "loss": 0.0133, "step": 66680 }, { "grad_norm": 0.21432343125343323, "learning_rate": 7.37987530368569e-06, "loss": 0.0099, "step": 66690 }, { "grad_norm": 0.1829962134361267, "learning_rate": 7.369071738226474e-06, "loss": 0.0104, "step": 66700 }, { "grad_norm": 0.2818503975868225, "learning_rate": 7.358275457226954e-06, "loss": 0.0134, "step": 66710 }, { "grad_norm": 0.21101818978786469, "learning_rate": 7.347486462531899e-06, "loss": 0.013, "step": 66720 }, { "grad_norm": 0.15903986990451813, "learning_rate": 7.336704755984858e-06, "loss": 0.0105, "step": 66730 }, { "grad_norm": 0.13965722918510437, "learning_rate": 7.325930339428133e-06, "loss": 0.0156, "step": 66740 }, { "grad_norm": 0.22346194088459015, "learning_rate": 7.315163214702769e-06, "loss": 0.0102, "step": 66750 }, { "grad_norm": 0.24959805607795715, "learning_rate": 7.304403383648595e-06, "loss": 0.0139, "step": 66760 }, { "grad_norm": 0.12162460386753082, "learning_rate": 7.293650848104139e-06, "loss": 0.01, "step": 66770 }, { "grad_norm": 0.32889029383659363, "learning_rate": 7.2829056099067374e-06, "loss": 0.0119, "step": 66780 }, { "grad_norm": 0.27104565501213074, "learning_rate": 7.2721676708924494e-06, "loss": 0.0103, "step": 66790 }, { "grad_norm": 0.29257118701934814, "learning_rate": 7.261437032896096e-06, "loss": 0.0146, "step": 66800 }, { "grad_norm": 0.16448234021663666, "learning_rate": 7.250713697751255e-06, "loss": 0.0096, "step": 66810 }, { "grad_norm": 0.14721493422985077, "learning_rate": 7.239997667290255e-06, "loss": 0.008, "step": 66820 }, { "grad_norm": 0.202128604054451, "learning_rate": 7.2292889433441425e-06, "loss": 0.0091, "step": 66830 }, { "grad_norm": 0.3065444529056549, "learning_rate": 7.218587527742793e-06, "loss": 0.0107, "step": 66840 }, { "grad_norm": 0.9394786357879639, "learning_rate": 7.207893422314749e-06, "loss": 0.0085, "step": 66850 }, { "grad_norm": 0.17396293580532074, "learning_rate": 7.1972066288873545e-06, "loss": 0.0101, "step": 66860 }, { "grad_norm": 0.23202867805957794, "learning_rate": 7.186527149286687e-06, "loss": 0.0238, "step": 66870 }, { "grad_norm": 0.14788702130317688, "learning_rate": 7.175854985337576e-06, "loss": 0.0097, "step": 66880 }, { "grad_norm": 0.21674573421478271, "learning_rate": 7.16519013886362e-06, "loss": 0.0131, "step": 66890 }, { "grad_norm": 0.2134566456079483, "learning_rate": 7.154532611687109e-06, "loss": 0.0114, "step": 66900 }, { "grad_norm": 0.2786247730255127, "learning_rate": 7.143882405629177e-06, "loss": 0.0105, "step": 66910 }, { "grad_norm": 0.24829329550266266, "learning_rate": 7.133239522509605e-06, "loss": 0.0095, "step": 66920 }, { "grad_norm": 0.24258407950401306, "learning_rate": 7.1226039641469956e-06, "loss": 0.01, "step": 66930 }, { "grad_norm": 0.5862144231796265, "learning_rate": 7.111975732358678e-06, "loss": 0.0105, "step": 66940 }, { "grad_norm": 0.20467515289783478, "learning_rate": 7.101354828960693e-06, "loss": 0.0095, "step": 66950 }, { "grad_norm": 0.21087335050106049, "learning_rate": 7.090741255767918e-06, "loss": 0.0091, "step": 66960 }, { "grad_norm": 0.15322500467300415, "learning_rate": 7.080135014593875e-06, "loss": 0.0109, "step": 66970 }, { "grad_norm": 0.2374422550201416, "learning_rate": 7.069536107250896e-06, "loss": 0.0102, "step": 66980 }, { "grad_norm": 0.5433485507965088, "learning_rate": 7.058944535550049e-06, "loss": 0.0079, "step": 66990 }, { "grad_norm": 0.1922667920589447, "learning_rate": 7.048360301301138e-06, "loss": 0.012, "step": 67000 }, { "grad_norm": 0.15292906761169434, "learning_rate": 7.03778340631272e-06, "loss": 0.0095, "step": 67010 }, { "grad_norm": 0.7333593964576721, "learning_rate": 7.0272138523921e-06, "loss": 0.0124, "step": 67020 }, { "grad_norm": 0.2620852589607239, "learning_rate": 7.016651641345334e-06, "loss": 0.0146, "step": 67030 }, { "grad_norm": 0.1763560175895691, "learning_rate": 7.0060967749772e-06, "loss": 0.0108, "step": 67040 }, { "grad_norm": 0.2880754768848419, "learning_rate": 6.995549255091238e-06, "loss": 0.0169, "step": 67050 }, { "grad_norm": 0.15011140704154968, "learning_rate": 6.98500908348973e-06, "loss": 0.012, "step": 67060 }, { "grad_norm": 0.15556959807872772, "learning_rate": 6.974476261973711e-06, "loss": 0.0094, "step": 67070 }, { "grad_norm": 0.15948955714702606, "learning_rate": 6.963950792342949e-06, "loss": 0.0076, "step": 67080 }, { "grad_norm": 0.20725050568580627, "learning_rate": 6.9534326763959715e-06, "loss": 0.0094, "step": 67090 }, { "grad_norm": 0.3163101077079773, "learning_rate": 6.942921915929995e-06, "loss": 0.0112, "step": 67100 }, { "grad_norm": 0.15363167226314545, "learning_rate": 6.9324185127410735e-06, "loss": 0.01, "step": 67110 }, { "grad_norm": 0.22901751101016998, "learning_rate": 6.921922468623921e-06, "loss": 0.0109, "step": 67120 }, { "grad_norm": 0.17064513266086578, "learning_rate": 6.911433785372023e-06, "loss": 0.0103, "step": 67130 }, { "grad_norm": 0.15389637649059296, "learning_rate": 6.900952464777632e-06, "loss": 0.0098, "step": 67140 }, { "grad_norm": 0.27761849761009216, "learning_rate": 6.8904785086316815e-06, "loss": 0.0115, "step": 67150 }, { "grad_norm": 0.20477771759033203, "learning_rate": 6.880011918723927e-06, "loss": 0.0077, "step": 67160 }, { "grad_norm": 0.22786983847618103, "learning_rate": 6.86955269684279e-06, "loss": 0.0127, "step": 67170 }, { "grad_norm": 0.3038426339626312, "learning_rate": 6.859100844775473e-06, "loss": 0.0157, "step": 67180 }, { "grad_norm": 0.1423265039920807, "learning_rate": 6.84865636430792e-06, "loss": 0.0109, "step": 67190 }, { "grad_norm": 0.23146648705005646, "learning_rate": 6.838219257224804e-06, "loss": 0.0088, "step": 67200 }, { "grad_norm": 0.1457759588956833, "learning_rate": 6.827789525309536e-06, "loss": 0.0164, "step": 67210 }, { "grad_norm": 0.1856161206960678, "learning_rate": 6.81736717034428e-06, "loss": 0.0095, "step": 67220 }, { "grad_norm": 0.20421282947063446, "learning_rate": 6.806952194109933e-06, "loss": 0.0136, "step": 67230 }, { "grad_norm": 0.18926040828227997, "learning_rate": 6.796544598386112e-06, "loss": 0.01, "step": 67240 }, { "grad_norm": 0.22253206372261047, "learning_rate": 6.786144384951204e-06, "loss": 0.0121, "step": 67250 }, { "grad_norm": 0.18816505372524261, "learning_rate": 6.775751555582322e-06, "loss": 0.0089, "step": 67260 }, { "grad_norm": 0.2253911942243576, "learning_rate": 6.76536611205531e-06, "loss": 0.0118, "step": 67270 }, { "grad_norm": 0.2964872419834137, "learning_rate": 6.754988056144762e-06, "loss": 0.0125, "step": 67280 }, { "grad_norm": 0.16524425148963928, "learning_rate": 6.744617389624014e-06, "loss": 0.0082, "step": 67290 }, { "grad_norm": 0.2288735955953598, "learning_rate": 6.734254114265087e-06, "loss": 0.0125, "step": 67300 }, { "grad_norm": 0.17995773255825043, "learning_rate": 6.723898231838843e-06, "loss": 0.0082, "step": 67310 }, { "grad_norm": 0.38695552945137024, "learning_rate": 6.713549744114766e-06, "loss": 0.0104, "step": 67320 }, { "grad_norm": 0.2390809804201126, "learning_rate": 6.703208652861159e-06, "loss": 0.0107, "step": 67330 }, { "grad_norm": 0.14091551303863525, "learning_rate": 6.69287495984503e-06, "loss": 0.0097, "step": 67340 }, { "grad_norm": 0.2815758287906647, "learning_rate": 6.6825486668321e-06, "loss": 0.0091, "step": 67350 }, { "grad_norm": 0.17695699632167816, "learning_rate": 6.672229775586886e-06, "loss": 0.012, "step": 67360 }, { "grad_norm": 0.15877951681613922, "learning_rate": 6.661918287872576e-06, "loss": 0.0086, "step": 67370 }, { "grad_norm": 0.21184809505939484, "learning_rate": 6.6516142054511346e-06, "loss": 0.0115, "step": 67380 }, { "grad_norm": 0.18967455625534058, "learning_rate": 6.641317530083241e-06, "loss": 0.0092, "step": 67390 }, { "grad_norm": 0.1754133552312851, "learning_rate": 6.631028263528322e-06, "loss": 0.011, "step": 67400 }, { "grad_norm": 0.20030030608177185, "learning_rate": 6.620746407544537e-06, "loss": 0.0101, "step": 67410 }, { "grad_norm": 0.28944849967956543, "learning_rate": 6.610471963888742e-06, "loss": 0.0135, "step": 67420 }, { "grad_norm": 0.22769978642463684, "learning_rate": 6.600204934316606e-06, "loss": 0.0111, "step": 67430 }, { "grad_norm": 0.2469559758901596, "learning_rate": 6.589945320582452e-06, "loss": 0.0097, "step": 67440 }, { "grad_norm": 0.2948490381240845, "learning_rate": 6.579693124439374e-06, "loss": 0.0091, "step": 67450 }, { "grad_norm": 0.20241348445415497, "learning_rate": 6.56944834763919e-06, "loss": 0.0096, "step": 67460 }, { "grad_norm": 0.30431368947029114, "learning_rate": 6.5592109919324575e-06, "loss": 0.0105, "step": 67470 }, { "grad_norm": 0.21310682594776154, "learning_rate": 6.54898105906846e-06, "loss": 0.0097, "step": 67480 }, { "grad_norm": 0.44026246666908264, "learning_rate": 6.5387585507952155e-06, "loss": 0.0163, "step": 67490 }, { "grad_norm": 0.23854321241378784, "learning_rate": 6.528543468859461e-06, "loss": 0.0108, "step": 67500 }, { "grad_norm": 0.21405909955501556, "learning_rate": 6.518335815006682e-06, "loss": 0.0147, "step": 67510 }, { "grad_norm": 0.4428975284099579, "learning_rate": 6.5081355909810845e-06, "loss": 0.0171, "step": 67520 }, { "grad_norm": 0.1643742322921753, "learning_rate": 6.497942798525608e-06, "loss": 0.0125, "step": 67530 }, { "grad_norm": 0.3092622756958008, "learning_rate": 6.487757439381936e-06, "loss": 0.0122, "step": 67540 }, { "grad_norm": 0.21706974506378174, "learning_rate": 6.477579515290433e-06, "loss": 0.0136, "step": 67550 }, { "grad_norm": 0.16567355394363403, "learning_rate": 6.467409027990273e-06, "loss": 0.0133, "step": 67560 }, { "grad_norm": 0.17623241245746613, "learning_rate": 6.457245979219279e-06, "loss": 0.0146, "step": 67570 }, { "grad_norm": 0.26153457164764404, "learning_rate": 6.447090370714054e-06, "loss": 0.0207, "step": 67580 }, { "grad_norm": 0.1917465329170227, "learning_rate": 6.436942204209917e-06, "loss": 0.0107, "step": 67590 }, { "grad_norm": 0.2764910161495209, "learning_rate": 6.4268014814408804e-06, "loss": 0.0146, "step": 67600 }, { "grad_norm": 0.18941931426525116, "learning_rate": 6.41666820413977e-06, "loss": 0.0091, "step": 67610 }, { "grad_norm": 0.19166253507137299, "learning_rate": 6.406542374038033e-06, "loss": 0.0103, "step": 67620 }, { "grad_norm": 0.19232767820358276, "learning_rate": 6.396423992865935e-06, "loss": 0.009, "step": 67630 }, { "grad_norm": 0.13438692688941956, "learning_rate": 6.386313062352412e-06, "loss": 0.0113, "step": 67640 }, { "grad_norm": 0.17626754939556122, "learning_rate": 6.376209584225152e-06, "loss": 0.0077, "step": 67650 }, { "grad_norm": 0.14759589731693268, "learning_rate": 6.366113560210557e-06, "loss": 0.0102, "step": 67660 }, { "grad_norm": 0.36666250228881836, "learning_rate": 6.356024992033766e-06, "loss": 0.0171, "step": 67670 }, { "grad_norm": 0.34005123376846313, "learning_rate": 6.345943881418648e-06, "loss": 0.0102, "step": 67680 }, { "grad_norm": 0.27298733592033386, "learning_rate": 6.33587023008777e-06, "loss": 0.0094, "step": 67690 }, { "grad_norm": 0.2246103733778, "learning_rate": 6.32580403976245e-06, "loss": 0.0082, "step": 67700 }, { "grad_norm": 0.17633451521396637, "learning_rate": 6.31574531216273e-06, "loss": 0.0102, "step": 67710 }, { "grad_norm": 0.18563011288642883, "learning_rate": 6.305694049007371e-06, "loss": 0.0125, "step": 67720 }, { "grad_norm": 0.26694685220718384, "learning_rate": 6.2956502520138575e-06, "loss": 0.0121, "step": 67730 }, { "grad_norm": 0.23271293938159943, "learning_rate": 6.285613922898409e-06, "loss": 0.0094, "step": 67740 }, { "grad_norm": 0.17857177555561066, "learning_rate": 6.275585063375927e-06, "loss": 0.0152, "step": 67750 }, { "grad_norm": 0.16593985259532928, "learning_rate": 6.265563675160113e-06, "loss": 0.0114, "step": 67760 }, { "grad_norm": 0.17825905978679657, "learning_rate": 6.25554975996332e-06, "loss": 0.0098, "step": 67770 }, { "grad_norm": 0.19010381400585175, "learning_rate": 6.245543319496661e-06, "loss": 0.0083, "step": 67780 }, { "grad_norm": 0.27491575479507446, "learning_rate": 6.2355443554699685e-06, "loss": 0.0127, "step": 67790 }, { "grad_norm": 0.18253497779369354, "learning_rate": 6.225552869591766e-06, "loss": 0.0104, "step": 67800 }, { "grad_norm": 0.1645561158657074, "learning_rate": 6.215568863569365e-06, "loss": 0.0094, "step": 67810 }, { "grad_norm": 0.3810318112373352, "learning_rate": 6.205592339108712e-06, "loss": 0.0177, "step": 67820 }, { "grad_norm": 0.183050736784935, "learning_rate": 6.195623297914577e-06, "loss": 0.013, "step": 67830 }, { "grad_norm": 0.2271665334701538, "learning_rate": 6.185661741690357e-06, "loss": 0.0096, "step": 67840 }, { "grad_norm": 0.2413688451051712, "learning_rate": 6.1757076721382145e-06, "loss": 0.0139, "step": 67850 }, { "grad_norm": 0.161363884806633, "learning_rate": 6.165761090959038e-06, "loss": 0.0143, "step": 67860 }, { "grad_norm": 0.22107714414596558, "learning_rate": 6.155821999852424e-06, "loss": 0.0099, "step": 67870 }, { "grad_norm": 0.1755983680486679, "learning_rate": 6.145890400516696e-06, "loss": 0.0088, "step": 67880 }, { "grad_norm": 0.26469141244888306, "learning_rate": 6.1359662946488816e-06, "loss": 0.0098, "step": 67890 }, { "grad_norm": 0.21975168585777283, "learning_rate": 6.12604968394474e-06, "loss": 0.008, "step": 67900 }, { "grad_norm": 0.2253289371728897, "learning_rate": 6.11614057009875e-06, "loss": 0.0117, "step": 67910 }, { "grad_norm": 0.19365641474723816, "learning_rate": 6.106238954804111e-06, "loss": 0.0107, "step": 67920 }, { "grad_norm": 0.31036415696144104, "learning_rate": 6.096344839752738e-06, "loss": 0.0089, "step": 67930 }, { "grad_norm": 0.13267478346824646, "learning_rate": 6.086458226635278e-06, "loss": 0.0106, "step": 67940 }, { "grad_norm": 0.15410080552101135, "learning_rate": 6.076579117141046e-06, "loss": 0.0169, "step": 67950 }, { "grad_norm": 0.2463812679052353, "learning_rate": 6.066707512958153e-06, "loss": 0.0113, "step": 67960 }, { "grad_norm": 0.24043142795562744, "learning_rate": 6.056843415773361e-06, "loss": 0.0108, "step": 67970 }, { "grad_norm": 0.20672927796840668, "learning_rate": 6.0469868272721776e-06, "loss": 0.0104, "step": 67980 }, { "grad_norm": 0.14358672499656677, "learning_rate": 6.037137749138844e-06, "loss": 0.0135, "step": 67990 }, { "grad_norm": 0.22839093208312988, "learning_rate": 6.027296183056252e-06, "loss": 0.0159, "step": 68000 }, { "grad_norm": 0.22389455139636993, "learning_rate": 6.017462130706114e-06, "loss": 0.0136, "step": 68010 }, { "grad_norm": 0.19919012486934662, "learning_rate": 6.0076355937687516e-06, "loss": 0.0086, "step": 68020 }, { "grad_norm": 0.15110985934734344, "learning_rate": 5.9978165739232925e-06, "loss": 0.0084, "step": 68030 }, { "grad_norm": 0.13469037413597107, "learning_rate": 5.988005072847508e-06, "loss": 0.0102, "step": 68040 }, { "grad_norm": 0.19332201778888702, "learning_rate": 5.978201092217928e-06, "loss": 0.014, "step": 68050 }, { "grad_norm": 0.20076917111873627, "learning_rate": 5.9684046337097895e-06, "loss": 0.0097, "step": 68060 }, { "grad_norm": 0.29877668619155884, "learning_rate": 5.958615698997016e-06, "loss": 0.0096, "step": 68070 }, { "grad_norm": 0.22225041687488556, "learning_rate": 5.948834289752303e-06, "loss": 0.013, "step": 68080 }, { "grad_norm": 0.13257066905498505, "learning_rate": 5.939060407646996e-06, "loss": 0.0122, "step": 68090 }, { "grad_norm": 0.15860159695148468, "learning_rate": 5.929294054351198e-06, "loss": 0.0122, "step": 68100 }, { "grad_norm": 0.12800566852092743, "learning_rate": 5.919535231533707e-06, "loss": 0.008, "step": 68110 }, { "grad_norm": 0.3304944932460785, "learning_rate": 5.909783940862046e-06, "loss": 0.0136, "step": 68120 }, { "grad_norm": 0.13019384443759918, "learning_rate": 5.900040184002436e-06, "loss": 0.0097, "step": 68130 }, { "grad_norm": 2.085798978805542, "learning_rate": 5.890303962619831e-06, "loss": 0.0114, "step": 68140 }, { "grad_norm": 0.18695871531963348, "learning_rate": 5.88057527837787e-06, "loss": 0.0109, "step": 68150 }, { "grad_norm": 0.14189191162586212, "learning_rate": 5.870854132938924e-06, "loss": 0.0096, "step": 68160 }, { "grad_norm": 0.18247075378894806, "learning_rate": 5.861140527964071e-06, "loss": 0.0132, "step": 68170 }, { "grad_norm": 0.11784061044454575, "learning_rate": 5.851434465113098e-06, "loss": 0.0099, "step": 68180 }, { "grad_norm": 0.18328562378883362, "learning_rate": 5.841735946044524e-06, "loss": 0.0095, "step": 68190 }, { "grad_norm": 0.16239826381206512, "learning_rate": 5.832044972415523e-06, "loss": 0.0102, "step": 68200 }, { "grad_norm": 0.12108021229505539, "learning_rate": 5.8223615458820655e-06, "loss": 0.0085, "step": 68210 }, { "grad_norm": 0.28381943702697754, "learning_rate": 5.812685668098733e-06, "loss": 0.0129, "step": 68220 }, { "grad_norm": 0.21115848422050476, "learning_rate": 5.803017340718913e-06, "loss": 0.0135, "step": 68230 }, { "grad_norm": 0.8116796612739563, "learning_rate": 5.793356565394636e-06, "loss": 0.0131, "step": 68240 }, { "grad_norm": 0.23001357913017273, "learning_rate": 5.783703343776658e-06, "loss": 0.0109, "step": 68250 }, { "grad_norm": 0.18142876029014587, "learning_rate": 5.774057677514477e-06, "loss": 0.0087, "step": 68260 }, { "grad_norm": 0.2493838220834732, "learning_rate": 5.764419568256235e-06, "loss": 0.0099, "step": 68270 }, { "grad_norm": 0.2831990718841553, "learning_rate": 5.754789017648865e-06, "loss": 0.0213, "step": 68280 }, { "grad_norm": 0.6352275609970093, "learning_rate": 5.745166027337934e-06, "loss": 0.0143, "step": 68290 }, { "grad_norm": 0.18881459534168243, "learning_rate": 5.735550598967754e-06, "loss": 0.0135, "step": 68300 }, { "grad_norm": 0.2664009630680084, "learning_rate": 5.725942734181339e-06, "loss": 0.011, "step": 68310 }, { "grad_norm": 0.21764768660068512, "learning_rate": 5.7163424346204146e-06, "loss": 0.0102, "step": 68320 }, { "grad_norm": 0.2178156077861786, "learning_rate": 5.706749701925407e-06, "loss": 0.0093, "step": 68330 }, { "grad_norm": 0.16017407178878784, "learning_rate": 5.697164537735461e-06, "loss": 0.0087, "step": 68340 }, { "grad_norm": 0.17182591557502747, "learning_rate": 5.687586943688406e-06, "loss": 0.0094, "step": 68350 }, { "grad_norm": 0.24269546568393707, "learning_rate": 5.678016921420792e-06, "loss": 0.0094, "step": 68360 }, { "grad_norm": 0.1816343516111374, "learning_rate": 5.668454472567875e-06, "loss": 0.0157, "step": 68370 }, { "grad_norm": 0.32036012411117554, "learning_rate": 5.658899598763617e-06, "loss": 0.0127, "step": 68380 }, { "grad_norm": 0.2109067142009735, "learning_rate": 5.6493523016407035e-06, "loss": 0.0094, "step": 68390 }, { "grad_norm": 0.16059285402297974, "learning_rate": 5.6398125828304615e-06, "loss": 0.0121, "step": 68400 }, { "grad_norm": 0.2829135060310364, "learning_rate": 5.630280443963015e-06, "loss": 0.0103, "step": 68410 }, { "grad_norm": 0.13902027904987335, "learning_rate": 5.620755886667112e-06, "loss": 0.0099, "step": 68420 }, { "grad_norm": 0.18973398208618164, "learning_rate": 5.611238912570271e-06, "loss": 0.0103, "step": 68430 }, { "grad_norm": 0.2334783375263214, "learning_rate": 5.601729523298649e-06, "loss": 0.0118, "step": 68440 }, { "grad_norm": 0.18081599473953247, "learning_rate": 5.592227720477161e-06, "loss": 0.0136, "step": 68450 }, { "grad_norm": 0.2870798408985138, "learning_rate": 5.582733505729415e-06, "loss": 0.011, "step": 68460 }, { "grad_norm": 0.1633376032114029, "learning_rate": 5.573246880677668e-06, "loss": 0.0088, "step": 68470 }, { "grad_norm": 0.27236542105674744, "learning_rate": 5.56376784694298e-06, "loss": 0.0112, "step": 68480 }, { "grad_norm": 0.18097449839115143, "learning_rate": 5.554296406145027e-06, "loss": 0.0106, "step": 68490 }, { "grad_norm": 0.27685514092445374, "learning_rate": 5.544832559902219e-06, "loss": 0.0126, "step": 68500 }, { "grad_norm": 0.2504315972328186, "learning_rate": 5.53537630983168e-06, "loss": 0.0137, "step": 68510 }, { "grad_norm": 0.13884371519088745, "learning_rate": 5.5259276575492125e-06, "loss": 0.0119, "step": 68520 }, { "grad_norm": 0.2021384984254837, "learning_rate": 5.516486604669357e-06, "loss": 0.0141, "step": 68530 }, { "grad_norm": 0.24277363717556, "learning_rate": 5.507053152805303e-06, "loss": 0.0155, "step": 68540 }, { "grad_norm": 0.3636234700679779, "learning_rate": 5.497627303568975e-06, "loss": 0.0157, "step": 68550 }, { "grad_norm": 0.12476564943790436, "learning_rate": 5.488209058571003e-06, "loss": 0.0071, "step": 68560 }, { "grad_norm": 0.20466341078281403, "learning_rate": 5.4787984194207054e-06, "loss": 0.0098, "step": 68570 }, { "grad_norm": 0.23436297476291656, "learning_rate": 5.469395387726095e-06, "loss": 0.0107, "step": 68580 }, { "grad_norm": 0.18231117725372314, "learning_rate": 5.45999996509391e-06, "loss": 0.0098, "step": 68590 }, { "grad_norm": 1.3484857082366943, "learning_rate": 5.450612153129536e-06, "loss": 0.0149, "step": 68600 }, { "grad_norm": 0.22266839444637299, "learning_rate": 5.4412319534371426e-06, "loss": 0.0103, "step": 68610 }, { "grad_norm": 0.13216234743595123, "learning_rate": 5.431859367619513e-06, "loss": 0.0095, "step": 68620 }, { "grad_norm": 0.1984628289937973, "learning_rate": 5.422494397278172e-06, "loss": 0.0118, "step": 68630 }, { "grad_norm": 0.32019463181495667, "learning_rate": 5.413137044013344e-06, "loss": 0.01, "step": 68640 }, { "grad_norm": 0.3138768970966339, "learning_rate": 5.403787309423941e-06, "loss": 0.0119, "step": 68650 }, { "grad_norm": 0.12158980965614319, "learning_rate": 5.3944451951075835e-06, "loss": 0.0095, "step": 68660 }, { "grad_norm": 0.12179300934076309, "learning_rate": 5.385110702660562e-06, "loss": 0.0096, "step": 68670 }, { "grad_norm": 0.21254298090934753, "learning_rate": 5.375783833677922e-06, "loss": 0.0101, "step": 68680 }, { "grad_norm": 0.27204838395118713, "learning_rate": 5.36646458975334e-06, "loss": 0.0106, "step": 68690 }, { "grad_norm": 0.3371385931968689, "learning_rate": 5.3571529724792294e-06, "loss": 0.0102, "step": 68700 }, { "grad_norm": 0.3519003987312317, "learning_rate": 5.347848983446702e-06, "loss": 0.0113, "step": 68710 }, { "grad_norm": 0.4696281850337982, "learning_rate": 5.3385526242455185e-06, "loss": 0.0105, "step": 68720 }, { "grad_norm": 0.12249357253313065, "learning_rate": 5.329263896464226e-06, "loss": 0.0083, "step": 68730 }, { "grad_norm": 0.25953617691993713, "learning_rate": 5.3199828016899715e-06, "loss": 0.0119, "step": 68740 }, { "grad_norm": 0.22786551713943481, "learning_rate": 5.310709341508657e-06, "loss": 0.0117, "step": 68750 }, { "grad_norm": 0.5537611842155457, "learning_rate": 5.301443517504861e-06, "loss": 0.01, "step": 68760 }, { "grad_norm": 0.12693822383880615, "learning_rate": 5.292185331261862e-06, "loss": 0.0105, "step": 68770 }, { "grad_norm": 0.2069275826215744, "learning_rate": 5.28293478436162e-06, "loss": 0.0096, "step": 68780 }, { "grad_norm": 0.1840781420469284, "learning_rate": 5.273691878384829e-06, "loss": 0.0161, "step": 68790 }, { "grad_norm": 0.21351782977581024, "learning_rate": 5.264456614910812e-06, "loss": 0.0088, "step": 68800 }, { "grad_norm": 0.4086020290851593, "learning_rate": 5.255228995517647e-06, "loss": 0.0204, "step": 68810 }, { "grad_norm": 0.21657976508140564, "learning_rate": 5.246009021782067e-06, "loss": 0.0119, "step": 68820 }, { "grad_norm": 0.3353801667690277, "learning_rate": 5.2367966952795225e-06, "loss": 0.0129, "step": 68830 }, { "grad_norm": 0.16703304648399353, "learning_rate": 5.2275920175841485e-06, "loss": 0.0101, "step": 68840 }, { "grad_norm": 0.3372226655483246, "learning_rate": 5.218394990268771e-06, "loss": 0.0079, "step": 68850 }, { "grad_norm": 1.0336159467697144, "learning_rate": 5.209205614904916e-06, "loss": 0.0121, "step": 68860 }, { "grad_norm": 0.23113404214382172, "learning_rate": 5.200023893062777e-06, "loss": 0.011, "step": 68870 }, { "grad_norm": 0.18740962445735931, "learning_rate": 5.190849826311289e-06, "loss": 0.0104, "step": 68880 }, { "grad_norm": 0.2978973090648651, "learning_rate": 5.181683416218025e-06, "loss": 0.0108, "step": 68890 }, { "grad_norm": 0.3227802813053131, "learning_rate": 5.172524664349276e-06, "loss": 0.0085, "step": 68900 }, { "grad_norm": 1.5706043243408203, "learning_rate": 5.1633735722700416e-06, "loss": 0.0121, "step": 68910 }, { "grad_norm": 0.10342849045991898, "learning_rate": 5.154230141543958e-06, "loss": 0.009, "step": 68920 }, { "grad_norm": 0.16071832180023193, "learning_rate": 5.145094373733433e-06, "loss": 0.009, "step": 68930 }, { "grad_norm": 0.15670046210289001, "learning_rate": 5.135966270399478e-06, "loss": 0.0075, "step": 68940 }, { "grad_norm": 0.36507561802864075, "learning_rate": 5.126845833101857e-06, "loss": 0.0142, "step": 68950 }, { "grad_norm": 0.15132805705070496, "learning_rate": 5.1177330633989994e-06, "loss": 0.0105, "step": 68960 }, { "grad_norm": 0.2658848166465759, "learning_rate": 5.108627962848033e-06, "loss": 0.0214, "step": 68970 }, { "grad_norm": 0.09091266989707947, "learning_rate": 5.099530533004759e-06, "loss": 0.0096, "step": 68980 }, { "grad_norm": 0.13814477622509003, "learning_rate": 5.090440775423699e-06, "loss": 0.0098, "step": 68990 }, { "grad_norm": 0.1225348636507988, "learning_rate": 5.081358691658022e-06, "loss": 0.0101, "step": 69000 }, { "grad_norm": 0.19369901716709137, "learning_rate": 5.072284283259621e-06, "loss": 0.0119, "step": 69010 }, { "grad_norm": 0.10731286555528641, "learning_rate": 5.063217551779053e-06, "loss": 0.0113, "step": 69020 }, { "grad_norm": 0.19817307591438293, "learning_rate": 5.054158498765588e-06, "loss": 0.0102, "step": 69030 }, { "grad_norm": 0.26900514960289, "learning_rate": 5.0451071257671625e-06, "loss": 0.01, "step": 69040 }, { "grad_norm": 0.2702317535877228, "learning_rate": 5.036063434330407e-06, "loss": 0.0111, "step": 69050 }, { "grad_norm": 0.11825031042098999, "learning_rate": 5.027027426000652e-06, "loss": 0.0093, "step": 69060 }, { "grad_norm": 0.07794928550720215, "learning_rate": 5.017999102321886e-06, "loss": 0.0099, "step": 69070 }, { "grad_norm": 0.2814249098300934, "learning_rate": 5.0089784648368224e-06, "loss": 0.0128, "step": 69080 }, { "grad_norm": 0.1531401127576828, "learning_rate": 4.999965515086829e-06, "loss": 0.0101, "step": 69090 }, { "grad_norm": 0.16813263297080994, "learning_rate": 4.990960254611976e-06, "loss": 0.0105, "step": 69100 }, { "grad_norm": 0.2897240221500397, "learning_rate": 4.9819626849510194e-06, "loss": 0.0116, "step": 69110 }, { "grad_norm": 0.21206417679786682, "learning_rate": 4.97297280764138e-06, "loss": 0.0182, "step": 69120 }, { "grad_norm": 0.1456936150789261, "learning_rate": 4.963990624219211e-06, "loss": 0.0109, "step": 69130 }, { "grad_norm": 0.13657894730567932, "learning_rate": 4.955016136219298e-06, "loss": 0.0102, "step": 69140 }, { "grad_norm": 0.17103074491024017, "learning_rate": 4.946049345175146e-06, "loss": 0.0081, "step": 69150 }, { "grad_norm": 0.1860608160495758, "learning_rate": 4.937090252618937e-06, "loss": 0.0122, "step": 69160 }, { "grad_norm": 0.2603994607925415, "learning_rate": 4.928138860081521e-06, "loss": 0.0132, "step": 69170 }, { "grad_norm": 0.08384133875370026, "learning_rate": 4.919195169092472e-06, "loss": 0.0104, "step": 69180 }, { "grad_norm": 0.1358296126127243, "learning_rate": 4.910259181179994e-06, "loss": 0.0109, "step": 69190 }, { "grad_norm": 0.35501816868782043, "learning_rate": 4.901330897871015e-06, "loss": 0.011, "step": 69200 }, { "grad_norm": 0.17316703498363495, "learning_rate": 4.8924103206911375e-06, "loss": 0.0097, "step": 69210 }, { "grad_norm": 0.20714879035949707, "learning_rate": 4.883497451164637e-06, "loss": 0.0143, "step": 69220 }, { "grad_norm": 0.16714520752429962, "learning_rate": 4.874592290814484e-06, "loss": 0.0097, "step": 69230 }, { "grad_norm": 0.3566061854362488, "learning_rate": 4.865694841162327e-06, "loss": 0.0141, "step": 69240 }, { "grad_norm": 0.17998971045017242, "learning_rate": 4.856805103728496e-06, "loss": 0.0117, "step": 69250 }, { "grad_norm": 0.185471311211586, "learning_rate": 4.847923080032007e-06, "loss": 0.0095, "step": 69260 }, { "grad_norm": 0.15237244963645935, "learning_rate": 4.8390487715905486e-06, "loss": 0.0092, "step": 69270 }, { "grad_norm": 0.21302969753742218, "learning_rate": 4.8301821799205e-06, "loss": 0.0108, "step": 69280 }, { "grad_norm": 0.12538595497608185, "learning_rate": 4.821323306536918e-06, "loss": 0.0101, "step": 69290 }, { "grad_norm": 0.17151467502117157, "learning_rate": 4.8124721529535455e-06, "loss": 0.0113, "step": 69300 }, { "grad_norm": 0.18320327997207642, "learning_rate": 4.803628720682807e-06, "loss": 0.0136, "step": 69310 }, { "grad_norm": 0.2251061201095581, "learning_rate": 4.794793011235776e-06, "loss": 0.0098, "step": 69320 }, { "grad_norm": 0.19004370272159576, "learning_rate": 4.78596502612228e-06, "loss": 0.015, "step": 69330 }, { "grad_norm": 0.21121491491794586, "learning_rate": 4.777144766850738e-06, "loss": 0.0125, "step": 69340 }, { "grad_norm": 0.20271463692188263, "learning_rate": 4.768332234928313e-06, "loss": 0.0114, "step": 69350 }, { "grad_norm": 0.18906161189079285, "learning_rate": 4.759527431860828e-06, "loss": 0.0112, "step": 69360 }, { "grad_norm": 0.3235240578651428, "learning_rate": 4.750730359152755e-06, "loss": 0.0119, "step": 69370 }, { "grad_norm": 0.19421342015266418, "learning_rate": 4.741941018307311e-06, "loss": 0.0094, "step": 69380 }, { "grad_norm": 0.3500887453556061, "learning_rate": 4.733159410826321e-06, "loss": 0.0102, "step": 69390 }, { "grad_norm": 0.2038400024175644, "learning_rate": 4.724385538210357e-06, "loss": 0.0114, "step": 69400 }, { "grad_norm": 0.13614264130592346, "learning_rate": 4.715619401958599e-06, "loss": 0.0101, "step": 69410 }, { "grad_norm": 0.2327244132757187, "learning_rate": 4.706861003568958e-06, "loss": 0.013, "step": 69420 }, { "grad_norm": 0.2637282907962799, "learning_rate": 4.698110344538003e-06, "loss": 0.012, "step": 69430 }, { "grad_norm": 0.2563045024871826, "learning_rate": 4.689367426360975e-06, "loss": 0.0128, "step": 69440 }, { "grad_norm": 0.36574405431747437, "learning_rate": 4.680632250531819e-06, "loss": 0.0086, "step": 69450 }, { "grad_norm": 0.19991740584373474, "learning_rate": 4.671904818543115e-06, "loss": 0.0144, "step": 69460 }, { "grad_norm": 0.2073388546705246, "learning_rate": 4.66318513188615e-06, "loss": 0.0093, "step": 69470 }, { "grad_norm": 0.1675173044204712, "learning_rate": 4.654473192050884e-06, "loss": 0.0105, "step": 69480 }, { "grad_norm": 0.25355300307273865, "learning_rate": 4.64576900052594e-06, "loss": 0.0088, "step": 69490 }, { "grad_norm": 0.1676579862833023, "learning_rate": 4.637072558798638e-06, "loss": 0.0105, "step": 69500 }, { "grad_norm": 0.2579445242881775, "learning_rate": 4.628383868354969e-06, "loss": 0.0137, "step": 69510 }, { "grad_norm": 0.12186267971992493, "learning_rate": 4.6197029306795595e-06, "loss": 0.0084, "step": 69520 }, { "grad_norm": 0.13521848618984222, "learning_rate": 4.611029747255779e-06, "loss": 0.009, "step": 69530 }, { "grad_norm": 0.13240060210227966, "learning_rate": 4.6023643195656164e-06, "loss": 0.0103, "step": 69540 }, { "grad_norm": 0.26223403215408325, "learning_rate": 4.593706649089768e-06, "loss": 0.0079, "step": 69550 }, { "grad_norm": 0.31653860211372375, "learning_rate": 4.585056737307597e-06, "loss": 0.0121, "step": 69560 }, { "grad_norm": 0.23037923872470856, "learning_rate": 4.576414585697103e-06, "loss": 0.0095, "step": 69570 }, { "grad_norm": 0.2588525414466858, "learning_rate": 4.567780195735044e-06, "loss": 0.0108, "step": 69580 }, { "grad_norm": 0.2531736493110657, "learning_rate": 4.559153568896757e-06, "loss": 0.0092, "step": 69590 }, { "grad_norm": 0.2574526071548462, "learning_rate": 4.550534706656329e-06, "loss": 0.0121, "step": 69600 }, { "grad_norm": 0.18854251503944397, "learning_rate": 4.541923610486465e-06, "loss": 0.0097, "step": 69610 }, { "grad_norm": 0.16382884979248047, "learning_rate": 4.533320281858578e-06, "loss": 0.0118, "step": 69620 }, { "grad_norm": 0.3498329818248749, "learning_rate": 4.52472472224274e-06, "loss": 0.0114, "step": 69630 }, { "grad_norm": 0.16257958114147186, "learning_rate": 4.51613693310769e-06, "loss": 0.011, "step": 69640 }, { "grad_norm": 0.28117164969444275, "learning_rate": 4.507556915920868e-06, "loss": 0.0109, "step": 69650 }, { "grad_norm": 0.335489422082901, "learning_rate": 4.498984672148332e-06, "loss": 0.0097, "step": 69660 }, { "grad_norm": 0.21435025334358215, "learning_rate": 4.490420203254864e-06, "loss": 0.0115, "step": 69670 }, { "grad_norm": 0.32679641246795654, "learning_rate": 4.481863510703893e-06, "loss": 0.0102, "step": 69680 }, { "grad_norm": 1.440402865409851, "learning_rate": 4.473314595957523e-06, "loss": 0.0112, "step": 69690 }, { "grad_norm": 0.3531342148780823, "learning_rate": 4.464773460476535e-06, "loss": 0.0122, "step": 69700 }, { "grad_norm": 0.14726842939853668, "learning_rate": 4.456240105720372e-06, "loss": 0.0089, "step": 69710 }, { "grad_norm": 0.19963225722312927, "learning_rate": 4.4477145331471405e-06, "loss": 0.0174, "step": 69720 }, { "grad_norm": 0.14961613714694977, "learning_rate": 4.439196744213653e-06, "loss": 0.0087, "step": 69730 }, { "grad_norm": 0.1616528183221817, "learning_rate": 4.430686740375339e-06, "loss": 0.0076, "step": 69740 }, { "grad_norm": 0.19982241094112396, "learning_rate": 4.422184523086342e-06, "loss": 0.0091, "step": 69750 }, { "grad_norm": 0.1997576206922531, "learning_rate": 4.41369009379946e-06, "loss": 0.0083, "step": 69760 }, { "grad_norm": 0.20633544027805328, "learning_rate": 4.405203453966139e-06, "loss": 0.0073, "step": 69770 }, { "grad_norm": 0.20838427543640137, "learning_rate": 4.396724605036539e-06, "loss": 0.0092, "step": 69780 }, { "grad_norm": 0.20427992939949036, "learning_rate": 4.388253548459437e-06, "loss": 0.0103, "step": 69790 }, { "grad_norm": 0.1117975190281868, "learning_rate": 4.3797902856823395e-06, "loss": 0.0078, "step": 69800 }, { "grad_norm": 0.14650580286979675, "learning_rate": 4.371334818151357e-06, "loss": 0.0083, "step": 69810 }, { "grad_norm": 0.5945644974708557, "learning_rate": 4.362887147311306e-06, "loss": 0.0108, "step": 69820 }, { "grad_norm": 0.17320461571216583, "learning_rate": 4.354447274605672e-06, "loss": 0.0101, "step": 69830 }, { "grad_norm": 0.1929624080657959, "learning_rate": 4.346015201476572e-06, "loss": 0.0155, "step": 69840 }, { "grad_norm": 0.2099342942237854, "learning_rate": 4.337590929364855e-06, "loss": 0.0114, "step": 69850 }, { "grad_norm": 0.16040146350860596, "learning_rate": 4.329174459709973e-06, "loss": 0.0101, "step": 69860 }, { "grad_norm": 0.65245121717453, "learning_rate": 4.320765793950071e-06, "loss": 0.0128, "step": 69870 }, { "grad_norm": 1.2391352653503418, "learning_rate": 4.312364933521962e-06, "loss": 0.0146, "step": 69880 }, { "grad_norm": 0.14180119335651398, "learning_rate": 4.303971879861129e-06, "loss": 0.0135, "step": 69890 }, { "grad_norm": 0.5626726746559143, "learning_rate": 4.295586634401716e-06, "loss": 0.0137, "step": 69900 }, { "grad_norm": 0.6105245351791382, "learning_rate": 4.287209198576536e-06, "loss": 0.0113, "step": 69910 }, { "grad_norm": 0.19947363436222076, "learning_rate": 4.278839573817045e-06, "loss": 0.0117, "step": 69920 }, { "grad_norm": 0.31729426980018616, "learning_rate": 4.270477761553399e-06, "loss": 0.0134, "step": 69930 }, { "grad_norm": 0.7901586890220642, "learning_rate": 4.262123763214393e-06, "loss": 0.0099, "step": 69940 }, { "grad_norm": 0.13499365746974945, "learning_rate": 4.253777580227508e-06, "loss": 0.0101, "step": 69950 }, { "grad_norm": 0.27172720432281494, "learning_rate": 4.2454392140188755e-06, "loss": 0.0108, "step": 69960 }, { "grad_norm": 0.14601045846939087, "learning_rate": 4.2371086660132785e-06, "loss": 0.0098, "step": 69970 }, { "grad_norm": 0.1704515963792801, "learning_rate": 4.228785937634205e-06, "loss": 0.0096, "step": 69980 }, { "grad_norm": 0.22066472470760345, "learning_rate": 4.220471030303758e-06, "loss": 0.0119, "step": 69990 }, { "grad_norm": 0.3784123957157135, "learning_rate": 4.212163945442754e-06, "loss": 0.0095, "step": 70000 }, { "grad_norm": 0.29069632291793823, "learning_rate": 4.203864684470621e-06, "loss": 0.0105, "step": 70010 }, { "grad_norm": 0.36987170577049255, "learning_rate": 4.19557324880549e-06, "loss": 0.0106, "step": 70020 }, { "grad_norm": 0.2047756314277649, "learning_rate": 4.187289639864145e-06, "loss": 0.009, "step": 70030 }, { "grad_norm": 0.10440466552972794, "learning_rate": 4.1790138590619974e-06, "loss": 0.0075, "step": 70040 }, { "grad_norm": 0.3413418233394623, "learning_rate": 4.170745907813195e-06, "loss": 0.0098, "step": 70050 }, { "grad_norm": 0.27382373809814453, "learning_rate": 4.162485787530479e-06, "loss": 0.011, "step": 70060 }, { "grad_norm": 0.16841115057468414, "learning_rate": 4.154233499625282e-06, "loss": 0.0107, "step": 70070 }, { "grad_norm": 0.2151634842157364, "learning_rate": 4.145989045507692e-06, "loss": 0.0104, "step": 70080 }, { "grad_norm": 0.22765155136585236, "learning_rate": 4.1377524265864666e-06, "loss": 0.0123, "step": 70090 }, { "grad_norm": 0.2595203220844269, "learning_rate": 4.1295236442690175e-06, "loss": 0.011, "step": 70100 }, { "grad_norm": 0.4217941164970398, "learning_rate": 4.121302699961421e-06, "loss": 0.0116, "step": 70110 }, { "grad_norm": 0.23118311166763306, "learning_rate": 4.113089595068403e-06, "loss": 0.013, "step": 70120 }, { "grad_norm": 0.12375681847333908, "learning_rate": 4.104884330993364e-06, "loss": 0.0121, "step": 70130 }, { "grad_norm": 0.4263235926628113, "learning_rate": 4.0966869091383585e-06, "loss": 0.0136, "step": 70140 }, { "grad_norm": 0.29027876257896423, "learning_rate": 4.088497330904101e-06, "loss": 0.0091, "step": 70150 }, { "grad_norm": 0.23396898806095123, "learning_rate": 4.080315597689976e-06, "loss": 0.0097, "step": 70160 }, { "grad_norm": 0.34383437037467957, "learning_rate": 4.072141710893995e-06, "loss": 0.0198, "step": 70170 }, { "grad_norm": 0.2603360414505005, "learning_rate": 4.063975671912879e-06, "loss": 0.0135, "step": 70180 }, { "grad_norm": 0.19304987788200378, "learning_rate": 4.055817482141949e-06, "loss": 0.0068, "step": 70190 }, { "grad_norm": 0.25377777218818665, "learning_rate": 4.047667142975259e-06, "loss": 0.0111, "step": 70200 }, { "grad_norm": 0.19058464467525482, "learning_rate": 4.039524655805443e-06, "loss": 0.0146, "step": 70210 }, { "grad_norm": 0.20031742751598358, "learning_rate": 4.03139002202384e-06, "loss": 0.0067, "step": 70220 }, { "grad_norm": 0.17914283275604248, "learning_rate": 4.023263243020447e-06, "loss": 0.0116, "step": 70230 }, { "grad_norm": 0.19475795328617096, "learning_rate": 4.015144320183884e-06, "loss": 0.0146, "step": 70240 }, { "grad_norm": 0.14676259458065033, "learning_rate": 4.007033254901482e-06, "loss": 0.0081, "step": 70250 }, { "grad_norm": 0.20630766451358795, "learning_rate": 3.9989300485591795e-06, "loss": 0.0099, "step": 70260 }, { "grad_norm": 0.18875890970230103, "learning_rate": 3.990834702541601e-06, "loss": 0.0126, "step": 70270 }, { "grad_norm": 0.20516903698444366, "learning_rate": 3.982747218232019e-06, "loss": 0.0093, "step": 70280 }, { "grad_norm": 0.20393748581409454, "learning_rate": 3.97466759701236e-06, "loss": 0.0122, "step": 70290 }, { "grad_norm": 0.37695959210395813, "learning_rate": 3.966595840263226e-06, "loss": 0.0114, "step": 70300 }, { "grad_norm": 0.2866908013820648, "learning_rate": 3.958531949363836e-06, "loss": 0.0137, "step": 70310 }, { "grad_norm": 0.15831789374351501, "learning_rate": 3.950475925692098e-06, "loss": 0.0122, "step": 70320 }, { "grad_norm": 0.22871389985084534, "learning_rate": 3.9424277706245685e-06, "loss": 0.012, "step": 70330 }, { "grad_norm": 0.17304804921150208, "learning_rate": 3.934387485536451e-06, "loss": 0.0097, "step": 70340 }, { "grad_norm": 0.3521248996257782, "learning_rate": 3.926355071801619e-06, "loss": 0.0113, "step": 70350 }, { "grad_norm": 0.11964356154203415, "learning_rate": 3.9183305307925965e-06, "loss": 0.007, "step": 70360 }, { "grad_norm": 0.23662105202674866, "learning_rate": 3.910313863880533e-06, "loss": 0.0086, "step": 70370 }, { "grad_norm": 0.21849332749843597, "learning_rate": 3.902305072435292e-06, "loss": 0.0086, "step": 70380 }, { "grad_norm": 0.20812590420246124, "learning_rate": 3.894304157825329e-06, "loss": 0.0091, "step": 70390 }, { "grad_norm": 0.1074320375919342, "learning_rate": 3.886311121417791e-06, "loss": 0.0123, "step": 70400 }, { "grad_norm": 0.18411323428153992, "learning_rate": 3.878325964578472e-06, "loss": 0.0107, "step": 70410 }, { "grad_norm": 0.4097709655761719, "learning_rate": 3.870348688671815e-06, "loss": 0.0096, "step": 70420 }, { "grad_norm": 0.16261638700962067, "learning_rate": 3.862379295060931e-06, "loss": 0.01, "step": 70430 }, { "grad_norm": 0.11282848566770554, "learning_rate": 3.854417785107534e-06, "loss": 0.0167, "step": 70440 }, { "grad_norm": 0.18659473955631256, "learning_rate": 3.8464641601720755e-06, "loss": 0.0099, "step": 70450 }, { "grad_norm": 0.2846921682357788, "learning_rate": 3.838518421613579e-06, "loss": 0.0112, "step": 70460 }, { "grad_norm": 0.21031144261360168, "learning_rate": 3.830580570789766e-06, "loss": 0.0131, "step": 70470 }, { "grad_norm": 0.2272602617740631, "learning_rate": 3.822650609057005e-06, "loss": 0.009, "step": 70480 }, { "grad_norm": 0.21610252559185028, "learning_rate": 3.814728537770285e-06, "loss": 0.0084, "step": 70490 }, { "grad_norm": 0.2817498743534088, "learning_rate": 3.806814358283306e-06, "loss": 0.0102, "step": 70500 }, { "grad_norm": 0.16953103244304657, "learning_rate": 3.7989080719483596e-06, "loss": 0.0073, "step": 70510 }, { "grad_norm": 0.2147461175918579, "learning_rate": 3.7910096801164143e-06, "loss": 0.0115, "step": 70520 }, { "grad_norm": 0.16477060317993164, "learning_rate": 3.7831191841371016e-06, "loss": 0.0105, "step": 70530 }, { "grad_norm": 0.27124524116516113, "learning_rate": 3.775236585358688e-06, "loss": 0.0118, "step": 70540 }, { "grad_norm": 0.19074997305870056, "learning_rate": 3.7673618851280843e-06, "loss": 0.0102, "step": 70550 }, { "grad_norm": 0.21122032403945923, "learning_rate": 3.759495084790887e-06, "loss": 0.0114, "step": 70560 }, { "grad_norm": 0.24560312926769257, "learning_rate": 3.751636185691282e-06, "loss": 0.0112, "step": 70570 }, { "grad_norm": 0.13150469958782196, "learning_rate": 3.7437851891721607e-06, "loss": 0.0087, "step": 70580 }, { "grad_norm": 0.23043324053287506, "learning_rate": 3.7359420965750404e-06, "loss": 0.0097, "step": 70590 }, { "grad_norm": 0.44158345460891724, "learning_rate": 3.7281069092400922e-06, "loss": 0.0117, "step": 70600 }, { "grad_norm": 0.29785388708114624, "learning_rate": 3.7202796285061348e-06, "loss": 0.0111, "step": 70610 }, { "grad_norm": 6.129216194152832, "learning_rate": 3.712460255710637e-06, "loss": 0.0158, "step": 70620 }, { "grad_norm": 0.16910609602928162, "learning_rate": 3.704648792189719e-06, "loss": 0.0099, "step": 70630 }, { "grad_norm": 0.23119348287582397, "learning_rate": 3.696845239278124e-06, "loss": 0.0089, "step": 70640 }, { "grad_norm": 0.2607388198375702, "learning_rate": 3.689049598309302e-06, "loss": 0.0129, "step": 70650 }, { "grad_norm": 0.4060022532939911, "learning_rate": 3.681261870615288e-06, "loss": 0.0175, "step": 70660 }, { "grad_norm": 0.1446448564529419, "learning_rate": 3.6734820575268004e-06, "loss": 0.013, "step": 70670 }, { "grad_norm": 0.13388751447200775, "learning_rate": 3.665710160373209e-06, "loss": 0.0085, "step": 70680 }, { "grad_norm": 0.24308231472969055, "learning_rate": 3.65794618048248e-06, "loss": 0.0089, "step": 70690 }, { "grad_norm": 0.147058367729187, "learning_rate": 3.6501901191813125e-06, "loss": 0.0086, "step": 70700 }, { "grad_norm": 0.21820105612277985, "learning_rate": 3.642441977794975e-06, "loss": 0.0123, "step": 70710 }, { "grad_norm": 0.20437218248844147, "learning_rate": 3.634701757647424e-06, "loss": 0.0112, "step": 70720 }, { "grad_norm": 0.12714587152004242, "learning_rate": 3.6269694600612468e-06, "loss": 0.0075, "step": 70730 }, { "grad_norm": 0.2719990909099579, "learning_rate": 3.619245086357681e-06, "loss": 0.0099, "step": 70740 }, { "grad_norm": 0.1933818757534027, "learning_rate": 3.611528637856615e-06, "loss": 0.0102, "step": 70750 }, { "grad_norm": 0.17307953536510468, "learning_rate": 3.6038201158765884e-06, "loss": 0.0074, "step": 70760 }, { "grad_norm": 0.29980769753456116, "learning_rate": 3.5961195217347534e-06, "loss": 0.0135, "step": 70770 }, { "grad_norm": 0.20111730694770813, "learning_rate": 3.588426856746946e-06, "loss": 0.0089, "step": 70780 }, { "grad_norm": 0.21000489592552185, "learning_rate": 3.5807421222276316e-06, "loss": 0.0175, "step": 70790 }, { "grad_norm": 0.1744832992553711, "learning_rate": 3.5730653194899156e-06, "loss": 0.0096, "step": 70800 }, { "grad_norm": 0.2011800855398178, "learning_rate": 3.565396449845554e-06, "loss": 0.0098, "step": 70810 }, { "grad_norm": 0.30800971388816833, "learning_rate": 3.557735514604954e-06, "loss": 0.0107, "step": 70820 }, { "grad_norm": 0.11530131101608276, "learning_rate": 3.5500825150771633e-06, "loss": 0.0093, "step": 70830 }, { "grad_norm": 0.1200852170586586, "learning_rate": 3.5424374525698466e-06, "loss": 0.009, "step": 70840 }, { "grad_norm": 0.11486569792032242, "learning_rate": 3.5348003283893704e-06, "loss": 0.0077, "step": 70850 }, { "grad_norm": 0.16886094212532043, "learning_rate": 3.527171143840685e-06, "loss": 0.0116, "step": 70860 }, { "grad_norm": 0.1476830542087555, "learning_rate": 3.519549900227409e-06, "loss": 0.0124, "step": 70870 }, { "grad_norm": 0.21397189795970917, "learning_rate": 3.511936598851828e-06, "loss": 0.0126, "step": 70880 }, { "grad_norm": 0.2223493456840515, "learning_rate": 3.504331241014813e-06, "loss": 0.0102, "step": 70890 }, { "grad_norm": 0.18827809393405914, "learning_rate": 3.4967338280159414e-06, "loss": 0.0102, "step": 70900 }, { "grad_norm": 0.2005602866411209, "learning_rate": 3.4891443611533846e-06, "loss": 0.0097, "step": 70910 }, { "grad_norm": 0.1843814253807068, "learning_rate": 3.481562841723984e-06, "loss": 0.0129, "step": 70920 }, { "grad_norm": 0.21715886890888214, "learning_rate": 3.473989271023215e-06, "loss": 0.0106, "step": 70930 }, { "grad_norm": 0.23882195353507996, "learning_rate": 3.4664236503451854e-06, "loss": 0.0107, "step": 70940 }, { "grad_norm": 0.15687255561351776, "learning_rate": 3.4588659809826674e-06, "loss": 0.0087, "step": 70950 }, { "grad_norm": 0.17383883893489838, "learning_rate": 3.451316264227039e-06, "loss": 0.0163, "step": 70960 }, { "grad_norm": 0.20794208347797394, "learning_rate": 3.4437745013683574e-06, "loss": 0.0103, "step": 70970 }, { "grad_norm": 0.2667214572429657, "learning_rate": 3.4362406936952916e-06, "loss": 0.0112, "step": 70980 }, { "grad_norm": 0.2381163388490677, "learning_rate": 3.4287148424951786e-06, "loss": 0.01, "step": 70990 }, { "grad_norm": 0.16839508712291718, "learning_rate": 3.4211969490539617e-06, "loss": 0.0122, "step": 71000 }, { "grad_norm": 0.22874784469604492, "learning_rate": 3.4136870146562584e-06, "loss": 0.0112, "step": 71010 }, { "grad_norm": 0.2135382741689682, "learning_rate": 3.406185040585308e-06, "loss": 0.0112, "step": 71020 }, { "grad_norm": 0.21029901504516602, "learning_rate": 3.3986910281229966e-06, "loss": 0.0141, "step": 71030 }, { "grad_norm": 0.7393901348114014, "learning_rate": 3.391204978549828e-06, "loss": 0.0125, "step": 71040 }, { "grad_norm": 0.13985475897789001, "learning_rate": 3.3837268931449785e-06, "loss": 0.015, "step": 71050 }, { "grad_norm": 0.14575554430484772, "learning_rate": 3.376256773186248e-06, "loss": 0.0134, "step": 71060 }, { "grad_norm": 0.15215134620666504, "learning_rate": 3.3687946199500664e-06, "loss": 0.0113, "step": 71070 }, { "grad_norm": 0.16304032504558563, "learning_rate": 3.3613404347115295e-06, "loss": 0.0099, "step": 71080 }, { "grad_norm": 0.2411176711320877, "learning_rate": 3.35389421874433e-06, "loss": 0.0076, "step": 71090 }, { "grad_norm": 1.567274808883667, "learning_rate": 3.3464559733208446e-06, "loss": 0.0108, "step": 71100 }, { "grad_norm": 0.2783743739128113, "learning_rate": 3.3390256997120505e-06, "loss": 0.0115, "step": 71110 }, { "grad_norm": 0.20355366170406342, "learning_rate": 3.331603399187583e-06, "loss": 0.0108, "step": 71120 }, { "grad_norm": 0.20717941224575043, "learning_rate": 3.324189073015721e-06, "loss": 0.013, "step": 71130 }, { "grad_norm": 0.22964227199554443, "learning_rate": 3.3167827224633465e-06, "loss": 0.0106, "step": 71140 }, { "grad_norm": 0.10855075716972351, "learning_rate": 3.309384348796024e-06, "loss": 0.0105, "step": 71150 }, { "grad_norm": 0.17232108116149902, "learning_rate": 3.301993953277921e-06, "loss": 0.0102, "step": 71160 }, { "grad_norm": 0.3713332414627075, "learning_rate": 3.29461153717186e-06, "loss": 0.0102, "step": 71170 }, { "grad_norm": 0.19947640597820282, "learning_rate": 3.287237101739293e-06, "loss": 0.0136, "step": 71180 }, { "grad_norm": 0.13500085473060608, "learning_rate": 3.2798706482403075e-06, "loss": 0.0179, "step": 71190 }, { "grad_norm": 0.20501770079135895, "learning_rate": 3.2725121779336285e-06, "loss": 0.0122, "step": 71200 }, { "grad_norm": 0.21181946992874146, "learning_rate": 3.265161692076618e-06, "loss": 0.0134, "step": 71210 }, { "grad_norm": 0.22935357689857483, "learning_rate": 3.257819191925282e-06, "loss": 0.0096, "step": 71220 }, { "grad_norm": 0.21670636534690857, "learning_rate": 3.2504846787342392e-06, "loss": 0.0091, "step": 71230 }, { "grad_norm": 0.21197877824306488, "learning_rate": 3.243158153756759e-06, "loss": 0.0096, "step": 71240 }, { "grad_norm": 0.21671350300312042, "learning_rate": 3.2358396182447514e-06, "loss": 0.0093, "step": 71250 }, { "grad_norm": 1.7205415964126587, "learning_rate": 3.2285290734487496e-06, "loss": 0.0101, "step": 71260 }, { "grad_norm": 0.2638133466243744, "learning_rate": 3.221226520617926e-06, "loss": 0.0095, "step": 71270 }, { "grad_norm": 0.20054101943969727, "learning_rate": 3.213931961000094e-06, "loss": 0.0158, "step": 71280 }, { "grad_norm": 0.30329185724258423, "learning_rate": 3.2066453958416733e-06, "loss": 0.0085, "step": 71290 }, { "grad_norm": 0.20152145624160767, "learning_rate": 3.199366826387773e-06, "loss": 0.0114, "step": 71300 }, { "grad_norm": 0.13174663484096527, "learning_rate": 3.192096253882071e-06, "loss": 0.0094, "step": 71310 }, { "grad_norm": 0.11047571152448654, "learning_rate": 3.1848336795669177e-06, "loss": 0.0105, "step": 71320 }, { "grad_norm": 0.22948966920375824, "learning_rate": 3.1775791046833035e-06, "loss": 0.0106, "step": 71330 }, { "grad_norm": 0.12159830331802368, "learning_rate": 3.170332530470804e-06, "loss": 0.0087, "step": 71340 }, { "grad_norm": 0.314900666475296, "learning_rate": 3.1630939581677012e-06, "loss": 0.0125, "step": 71350 }, { "grad_norm": 0.28759732842445374, "learning_rate": 3.155863389010838e-06, "loss": 0.0103, "step": 71360 }, { "grad_norm": 0.37170761823654175, "learning_rate": 3.1486408242357323e-06, "loss": 0.0119, "step": 71370 }, { "grad_norm": 0.3978363871574402, "learning_rate": 3.1414262650765248e-06, "loss": 0.0215, "step": 71380 }, { "grad_norm": 0.19683021306991577, "learning_rate": 3.134219712765979e-06, "loss": 0.0134, "step": 71390 }, { "grad_norm": 0.18618406355381012, "learning_rate": 3.1270211685355044e-06, "loss": 0.0129, "step": 71400 }, { "grad_norm": 0.24127382040023804, "learning_rate": 3.1198306336151338e-06, "loss": 0.0098, "step": 71410 }, { "grad_norm": 0.21381831169128418, "learning_rate": 3.11264810923354e-06, "loss": 0.0098, "step": 71420 }, { "grad_norm": 1.386317253112793, "learning_rate": 3.105473596618008e-06, "loss": 0.009, "step": 71430 }, { "grad_norm": 1.6097983121871948, "learning_rate": 3.0983070969944683e-06, "loss": 0.0126, "step": 71440 }, { "grad_norm": 0.22568148374557495, "learning_rate": 3.09114861158748e-06, "loss": 0.0091, "step": 71450 }, { "grad_norm": 0.27265480160713196, "learning_rate": 3.0839981416202314e-06, "loss": 0.0106, "step": 71460 }, { "grad_norm": 0.2607642114162445, "learning_rate": 3.0768556883145507e-06, "loss": 0.0122, "step": 71470 }, { "grad_norm": 0.22947636246681213, "learning_rate": 3.0697212528908834e-06, "loss": 0.0096, "step": 71480 }, { "grad_norm": 0.26936453580856323, "learning_rate": 3.0625948365683e-06, "loss": 0.0101, "step": 71490 }, { "grad_norm": 1.1653138399124146, "learning_rate": 3.055476440564531e-06, "loss": 0.0133, "step": 71500 }, { "grad_norm": 0.15943177044391632, "learning_rate": 3.0483660660958924e-06, "loss": 0.0098, "step": 71510 }, { "grad_norm": 0.23214180767536163, "learning_rate": 3.0412637143773624e-06, "loss": 0.0117, "step": 71520 }, { "grad_norm": 0.21687963604927063, "learning_rate": 3.034169386622554e-06, "loss": 0.0117, "step": 71530 }, { "grad_norm": 0.1864936500787735, "learning_rate": 3.027083084043658e-06, "loss": 0.0108, "step": 71540 }, { "grad_norm": 1.0128878355026245, "learning_rate": 3.0200048078515676e-06, "loss": 0.0109, "step": 71550 }, { "grad_norm": 0.15246905386447906, "learning_rate": 3.012934559255742e-06, "loss": 0.0094, "step": 71560 }, { "grad_norm": 0.2162957787513733, "learning_rate": 3.0058723394642994e-06, "loss": 0.0131, "step": 71570 }, { "grad_norm": 0.13740099966526031, "learning_rate": 2.998818149683985e-06, "loss": 0.0143, "step": 71580 }, { "grad_norm": 0.18080595135688782, "learning_rate": 2.9917719911201627e-06, "loss": 0.0103, "step": 71590 }, { "grad_norm": 0.19714339077472687, "learning_rate": 2.9847338649768352e-06, "loss": 0.0112, "step": 71600 }, { "grad_norm": 0.34077295660972595, "learning_rate": 2.977703772456608e-06, "loss": 0.0115, "step": 71610 }, { "grad_norm": 0.1680152863264084, "learning_rate": 2.9706817147607535e-06, "loss": 0.0081, "step": 71620 }, { "grad_norm": 0.140125572681427, "learning_rate": 2.9636676930891393e-06, "loss": 0.011, "step": 71630 }, { "grad_norm": 0.2187948375940323, "learning_rate": 2.9566617086402625e-06, "loss": 0.0095, "step": 71640 }, { "grad_norm": 0.14086377620697021, "learning_rate": 2.9496637626112655e-06, "loss": 0.0101, "step": 71650 }, { "grad_norm": 0.19305111467838287, "learning_rate": 2.9426738561979027e-06, "loss": 0.0106, "step": 71660 }, { "grad_norm": 0.3072102963924408, "learning_rate": 2.9356919905945524e-06, "loss": 0.0248, "step": 71670 }, { "grad_norm": 0.22950324416160583, "learning_rate": 2.928718166994243e-06, "loss": 0.0097, "step": 71680 }, { "grad_norm": 0.23768895864486694, "learning_rate": 2.9217523865885833e-06, "loss": 0.0161, "step": 71690 }, { "grad_norm": 0.21643322706222534, "learning_rate": 2.9147946505678537e-06, "loss": 0.0095, "step": 71700 }, { "grad_norm": 0.08945147693157196, "learning_rate": 2.9078449601209313e-06, "loss": 0.0072, "step": 71710 }, { "grad_norm": 0.2532840371131897, "learning_rate": 2.900903316435333e-06, "loss": 0.0113, "step": 71720 }, { "grad_norm": 0.2419721484184265, "learning_rate": 2.8939697206971983e-06, "loss": 0.012, "step": 71730 }, { "grad_norm": 0.29269418120384216, "learning_rate": 2.8870441740912746e-06, "loss": 0.0091, "step": 71740 }, { "grad_norm": 0.2063802033662796, "learning_rate": 2.8801266778009762e-06, "loss": 0.0102, "step": 71750 }, { "grad_norm": 0.16423505544662476, "learning_rate": 2.8732172330082798e-06, "loss": 0.0149, "step": 71760 }, { "grad_norm": 0.15068086981773376, "learning_rate": 2.8663158408938517e-06, "loss": 0.0104, "step": 71770 }, { "grad_norm": 0.12895973026752472, "learning_rate": 2.859422502636938e-06, "loss": 0.0102, "step": 71780 }, { "grad_norm": 0.40753477811813354, "learning_rate": 2.8525372194154076e-06, "loss": 0.0106, "step": 71790 }, { "grad_norm": 0.14392054080963135, "learning_rate": 2.8456599924057913e-06, "loss": 0.0091, "step": 71800 }, { "grad_norm": 0.12472301721572876, "learning_rate": 2.8387908227831995e-06, "loss": 0.0115, "step": 71810 }, { "grad_norm": 0.18155761063098907, "learning_rate": 2.831929711721404e-06, "loss": 0.0146, "step": 71820 }, { "grad_norm": 0.22430633008480072, "learning_rate": 2.825076660392767e-06, "loss": 0.0092, "step": 71830 }, { "grad_norm": 0.2383682280778885, "learning_rate": 2.8182316699682908e-06, "loss": 0.0105, "step": 71840 }, { "grad_norm": 0.1922089159488678, "learning_rate": 2.811394741617601e-06, "loss": 0.0119, "step": 71850 }, { "grad_norm": 5.076345920562744, "learning_rate": 2.8045658765089356e-06, "loss": 0.0172, "step": 71860 }, { "grad_norm": 0.2410132884979248, "learning_rate": 2.7977450758091605e-06, "loss": 0.0177, "step": 71870 }, { "grad_norm": 0.37851768732070923, "learning_rate": 2.790932340683783e-06, "loss": 0.0167, "step": 71880 }, { "grad_norm": 0.1568557322025299, "learning_rate": 2.7841276722968823e-06, "loss": 0.0104, "step": 71890 }, { "grad_norm": 0.25848400592803955, "learning_rate": 2.7773310718112067e-06, "loss": 0.0099, "step": 71900 }, { "grad_norm": 0.1839333325624466, "learning_rate": 2.7705425403881102e-06, "loss": 0.01, "step": 71910 }, { "grad_norm": 0.18960411846637726, "learning_rate": 2.763762079187565e-06, "loss": 0.0118, "step": 71920 }, { "grad_norm": 0.15015976130962372, "learning_rate": 2.756989689368178e-06, "loss": 0.0099, "step": 71930 }, { "grad_norm": 0.13591545820236206, "learning_rate": 2.750225372087134e-06, "loss": 0.0134, "step": 71940 }, { "grad_norm": 0.23212572932243347, "learning_rate": 2.7434691285003033e-06, "loss": 0.0107, "step": 71950 }, { "grad_norm": 1.0282758474349976, "learning_rate": 2.7367209597621178e-06, "loss": 0.0126, "step": 71960 }, { "grad_norm": 0.2664547264575958, "learning_rate": 2.729980867025683e-06, "loss": 0.0138, "step": 71970 }, { "grad_norm": 0.29316675662994385, "learning_rate": 2.7232488514426724e-06, "loss": 0.011, "step": 71980 }, { "grad_norm": 0.29186737537384033, "learning_rate": 2.7165249141634097e-06, "loss": 0.0101, "step": 71990 }, { "grad_norm": 0.1917719691991806, "learning_rate": 2.709809056336837e-06, "loss": 0.0138, "step": 72000 }, { "grad_norm": 1.527220368385315, "learning_rate": 2.703101279110498e-06, "loss": 0.0091, "step": 72010 }, { "grad_norm": 0.26490721106529236, "learning_rate": 2.696401583630587e-06, "loss": 0.0149, "step": 72020 }, { "grad_norm": 0.17656724154949188, "learning_rate": 2.689709971041887e-06, "loss": 0.0095, "step": 72030 }, { "grad_norm": 0.22242309153079987, "learning_rate": 2.6830264424878126e-06, "loss": 0.0119, "step": 72040 }, { "grad_norm": 0.5639452934265137, "learning_rate": 2.676350999110394e-06, "loss": 0.0153, "step": 72050 }, { "grad_norm": 0.5327843427658081, "learning_rate": 2.6696836420502856e-06, "loss": 0.0144, "step": 72060 }, { "grad_norm": 0.1778971403837204, "learning_rate": 2.663024372446765e-06, "loss": 0.0088, "step": 72070 }, { "grad_norm": 0.6345939040184021, "learning_rate": 2.6563731914377055e-06, "loss": 0.0118, "step": 72080 }, { "grad_norm": 0.17676255106925964, "learning_rate": 2.6497301001596087e-06, "loss": 0.013, "step": 72090 }, { "grad_norm": 0.4976039230823517, "learning_rate": 2.643095099747611e-06, "loss": 0.0142, "step": 72100 }, { "grad_norm": 0.6439076066017151, "learning_rate": 2.636468191335445e-06, "loss": 0.0133, "step": 72110 }, { "grad_norm": 0.34757593274116516, "learning_rate": 2.629849376055471e-06, "loss": 0.0115, "step": 72120 }, { "grad_norm": 0.10184559226036072, "learning_rate": 2.6232386550386678e-06, "loss": 0.0103, "step": 72130 }, { "grad_norm": 0.1345234215259552, "learning_rate": 2.6166360294146097e-06, "loss": 0.0105, "step": 72140 }, { "grad_norm": 0.6825857758522034, "learning_rate": 2.6100415003115275e-06, "loss": 0.012, "step": 72150 }, { "grad_norm": 0.2112586498260498, "learning_rate": 2.603455068856225e-06, "loss": 0.0084, "step": 72160 }, { "grad_norm": 0.9008141160011292, "learning_rate": 2.5968767361741584e-06, "loss": 0.0137, "step": 72170 }, { "grad_norm": 0.16298019886016846, "learning_rate": 2.590306503389378e-06, "loss": 0.0096, "step": 72180 }, { "grad_norm": 0.15865762531757355, "learning_rate": 2.583744371624558e-06, "loss": 0.0124, "step": 72190 }, { "grad_norm": 0.5418599247932434, "learning_rate": 2.57719034200099e-06, "loss": 0.0112, "step": 72200 }, { "grad_norm": 0.20886938273906708, "learning_rate": 2.570644415638568e-06, "loss": 0.0104, "step": 72210 }, { "grad_norm": 0.23191310465335846, "learning_rate": 2.5641065936558295e-06, "loss": 0.0179, "step": 72220 }, { "grad_norm": 0.25333261489868164, "learning_rate": 2.557576877169898e-06, "loss": 0.0124, "step": 72230 }, { "grad_norm": 0.3128688633441925, "learning_rate": 2.5510552672965205e-06, "loss": 0.0106, "step": 72240 }, { "grad_norm": 0.20665377378463745, "learning_rate": 2.544541765150077e-06, "loss": 0.0074, "step": 72250 }, { "grad_norm": 0.1666916161775589, "learning_rate": 2.5380363718435165e-06, "loss": 0.0125, "step": 72260 }, { "grad_norm": 0.19009363651275635, "learning_rate": 2.5315390884884714e-06, "loss": 0.0101, "step": 72270 }, { "grad_norm": 0.5893856883049011, "learning_rate": 2.5250499161951214e-06, "loss": 0.0116, "step": 72280 }, { "grad_norm": 0.17456120252609253, "learning_rate": 2.518568856072301e-06, "loss": 0.009, "step": 72290 }, { "grad_norm": 0.2594192326068878, "learning_rate": 2.512095909227441e-06, "loss": 0.0106, "step": 72300 }, { "grad_norm": 0.42799434065818787, "learning_rate": 2.50563107676659e-06, "loss": 0.0122, "step": 72310 }, { "grad_norm": 0.14334598183631897, "learning_rate": 2.4991743597944152e-06, "loss": 0.0119, "step": 72320 }, { "grad_norm": 0.13493165373802185, "learning_rate": 2.492725759414205e-06, "loss": 0.0131, "step": 72330 }, { "grad_norm": 0.1473153978586197, "learning_rate": 2.4862852767278234e-06, "loss": 0.0108, "step": 72340 }, { "grad_norm": 0.14625154435634613, "learning_rate": 2.479852912835784e-06, "loss": 0.007, "step": 72350 }, { "grad_norm": 0.20232053101062775, "learning_rate": 2.4734286688372075e-06, "loss": 0.0143, "step": 72360 }, { "grad_norm": 0.31740254163742065, "learning_rate": 2.467012545829811e-06, "loss": 0.0156, "step": 72370 }, { "grad_norm": 0.16666428744792938, "learning_rate": 2.460604544909945e-06, "loss": 0.0082, "step": 72380 }, { "grad_norm": 0.12784907221794128, "learning_rate": 2.454204667172555e-06, "loss": 0.0079, "step": 72390 }, { "grad_norm": 0.36979740858078003, "learning_rate": 2.447812913711217e-06, "loss": 0.0107, "step": 72400 }, { "grad_norm": 0.21305203437805176, "learning_rate": 2.44142928561808e-06, "loss": 0.0122, "step": 72410 }, { "grad_norm": 0.2896682024002075, "learning_rate": 2.435053783983965e-06, "loss": 0.0099, "step": 72420 }, { "grad_norm": 0.10948489606380463, "learning_rate": 2.4286864098982453e-06, "loss": 0.0125, "step": 72430 }, { "grad_norm": 1.200554609298706, "learning_rate": 2.4223271644489397e-06, "loss": 0.0136, "step": 72440 }, { "grad_norm": 0.09630735218524933, "learning_rate": 2.415976048722679e-06, "loss": 0.0098, "step": 72450 }, { "grad_norm": 0.16395437717437744, "learning_rate": 2.4096330638046673e-06, "loss": 0.0088, "step": 72460 }, { "grad_norm": 0.32293400168418884, "learning_rate": 2.4032982107787816e-06, "loss": 0.0159, "step": 72470 }, { "grad_norm": 0.5438575744628906, "learning_rate": 2.396971490727451e-06, "loss": 0.0117, "step": 72480 }, { "grad_norm": 1.7403055429458618, "learning_rate": 2.3906529047317493e-06, "loss": 0.0105, "step": 72490 }, { "grad_norm": 0.25490185618400574, "learning_rate": 2.3843424538713465e-06, "loss": 0.0079, "step": 72500 }, { "grad_norm": 0.21081435680389404, "learning_rate": 2.3780401392245298e-06, "loss": 0.0099, "step": 72510 }, { "grad_norm": 0.10712355375289917, "learning_rate": 2.3717459618681883e-06, "loss": 0.0096, "step": 72520 }, { "grad_norm": 0.18014350533485413, "learning_rate": 2.3654599228778274e-06, "loss": 0.0105, "step": 72530 }, { "grad_norm": 0.2966359555721283, "learning_rate": 2.3591820233275607e-06, "loss": 0.0114, "step": 72540 }, { "grad_norm": 0.10323711484670639, "learning_rate": 2.3529122642901024e-06, "loss": 0.0095, "step": 72550 }, { "grad_norm": 0.24490231275558472, "learning_rate": 2.3466506468367845e-06, "loss": 0.0129, "step": 72560 }, { "grad_norm": 0.16914311051368713, "learning_rate": 2.3403971720375506e-06, "loss": 0.0123, "step": 72570 }, { "grad_norm": 0.0954844206571579, "learning_rate": 2.334151840960952e-06, "loss": 0.0095, "step": 72580 }, { "grad_norm": 0.1336578130722046, "learning_rate": 2.3279146546741347e-06, "loss": 0.0107, "step": 72590 }, { "grad_norm": 0.16637654602527618, "learning_rate": 2.32168561424288e-06, "loss": 0.0087, "step": 72600 }, { "grad_norm": 0.3461584746837616, "learning_rate": 2.3154647207315307e-06, "loss": 0.0096, "step": 72610 }, { "grad_norm": 0.1828300952911377, "learning_rate": 2.309251975203103e-06, "loss": 0.0087, "step": 72620 }, { "grad_norm": 0.13435034453868866, "learning_rate": 2.303047378719159e-06, "loss": 0.0112, "step": 72630 }, { "grad_norm": 0.817726194858551, "learning_rate": 2.2968509323399e-06, "loss": 0.0153, "step": 72640 }, { "grad_norm": 0.21217294037342072, "learning_rate": 2.290662637124147e-06, "loss": 0.009, "step": 72650 }, { "grad_norm": 0.14717374742031097, "learning_rate": 2.2844824941292807e-06, "loss": 0.012, "step": 72660 }, { "grad_norm": 0.2229578047990799, "learning_rate": 2.2783105044113406e-06, "loss": 0.0116, "step": 72670 }, { "grad_norm": 0.13299821317195892, "learning_rate": 2.272146669024944e-06, "loss": 0.0067, "step": 72680 }, { "grad_norm": 0.10893267393112183, "learning_rate": 2.26599098902332e-06, "loss": 0.009, "step": 72690 }, { "grad_norm": 0.1613793522119522, "learning_rate": 2.2598434654583113e-06, "loss": 0.0095, "step": 72700 }, { "grad_norm": 0.3284178078174591, "learning_rate": 2.253704099380355e-06, "loss": 0.0165, "step": 72710 }, { "grad_norm": 0.20473432540893555, "learning_rate": 2.247572891838512e-06, "loss": 0.0136, "step": 72720 }, { "grad_norm": 0.13651758432388306, "learning_rate": 2.241449843880422e-06, "loss": 0.0111, "step": 72730 }, { "grad_norm": 0.1336631029844284, "learning_rate": 2.235334956552354e-06, "loss": 0.0131, "step": 72740 }, { "grad_norm": 0.23755548894405365, "learning_rate": 2.2292282308991775e-06, "loss": 0.0122, "step": 72750 }, { "grad_norm": 0.1789056807756424, "learning_rate": 2.223129667964363e-06, "loss": 0.0116, "step": 72760 }, { "grad_norm": 0.26516294479370117, "learning_rate": 2.2170392687899834e-06, "loss": 0.0089, "step": 72770 }, { "grad_norm": 0.20548415184020996, "learning_rate": 2.210957034416733e-06, "loss": 0.0093, "step": 72780 }, { "grad_norm": 0.21743489801883698, "learning_rate": 2.2048829658838867e-06, "loss": 0.0102, "step": 72790 }, { "grad_norm": 0.5786439180374146, "learning_rate": 2.1988170642293525e-06, "loss": 0.01, "step": 72800 }, { "grad_norm": 0.08372501283884048, "learning_rate": 2.1927593304896075e-06, "loss": 0.0081, "step": 72810 }, { "grad_norm": 0.17679622769355774, "learning_rate": 2.1867097656997626e-06, "loss": 0.0081, "step": 72820 }, { "grad_norm": 0.3634064197540283, "learning_rate": 2.180668370893524e-06, "loss": 0.0096, "step": 72830 }, { "grad_norm": 0.32674694061279297, "learning_rate": 2.174635147103199e-06, "loss": 0.012, "step": 72840 }, { "grad_norm": 0.36654284596443176, "learning_rate": 2.1686100953597075e-06, "loss": 0.0105, "step": 72850 }, { "grad_norm": 0.3931468725204468, "learning_rate": 2.1625932166925433e-06, "loss": 0.0094, "step": 72860 }, { "grad_norm": 0.14847971498966217, "learning_rate": 2.1565845121298556e-06, "loss": 0.0138, "step": 72870 }, { "grad_norm": 0.6283239722251892, "learning_rate": 2.150583982698351e-06, "loss": 0.009, "step": 72880 }, { "grad_norm": 0.35959985852241516, "learning_rate": 2.144591629423359e-06, "loss": 0.009, "step": 72890 }, { "grad_norm": 0.2210197001695633, "learning_rate": 2.138607453328817e-06, "loss": 0.0123, "step": 72900 }, { "grad_norm": 0.21913254261016846, "learning_rate": 2.132631455437234e-06, "loss": 0.0109, "step": 72910 }, { "grad_norm": 0.39180588722229004, "learning_rate": 2.1266636367697714e-06, "loss": 0.0175, "step": 72920 }, { "grad_norm": 0.22866807878017426, "learning_rate": 2.120703998346152e-06, "loss": 0.0086, "step": 72930 }, { "grad_norm": 0.21544328331947327, "learning_rate": 2.1147525411847114e-06, "loss": 0.0122, "step": 72940 }, { "grad_norm": 0.25608089566230774, "learning_rate": 2.108809266302403e-06, "loss": 0.0125, "step": 72950 }, { "grad_norm": 0.20382200181484222, "learning_rate": 2.102874174714764e-06, "loss": 0.0117, "step": 72960 }, { "grad_norm": 0.5292545557022095, "learning_rate": 2.0969472674359393e-06, "loss": 0.0117, "step": 72970 }, { "grad_norm": 0.2825775444507599, "learning_rate": 2.091028545478668e-06, "loss": 0.0114, "step": 72980 }, { "grad_norm": 0.17131108045578003, "learning_rate": 2.08511800985432e-06, "loss": 0.0096, "step": 72990 }, { "grad_norm": 0.2772677540779114, "learning_rate": 2.07921566157282e-06, "loss": 0.0147, "step": 73000 }, { "grad_norm": 0.22939573228359222, "learning_rate": 2.073321501642728e-06, "loss": 0.0088, "step": 73010 }, { "grad_norm": 0.44478774070739746, "learning_rate": 2.0674355310711937e-06, "loss": 0.0122, "step": 73020 }, { "grad_norm": 0.14657838642597198, "learning_rate": 2.0615577508639682e-06, "loss": 0.0075, "step": 73030 }, { "grad_norm": 0.3080083727836609, "learning_rate": 2.055688162025399e-06, "loss": 0.0112, "step": 73040 }, { "grad_norm": 0.23464670777320862, "learning_rate": 2.0498267655584546e-06, "loss": 0.0116, "step": 73050 }, { "grad_norm": 0.14680352807044983, "learning_rate": 2.0439735624646626e-06, "loss": 0.0108, "step": 73060 }, { "grad_norm": 0.15826983749866486, "learning_rate": 2.0381285537442008e-06, "loss": 0.01, "step": 73070 }, { "grad_norm": 0.25407204031944275, "learning_rate": 2.032291740395803e-06, "loss": 0.0111, "step": 73080 }, { "grad_norm": 0.21194590628147125, "learning_rate": 2.0264631234168276e-06, "loss": 0.011, "step": 73090 }, { "grad_norm": 0.5535181164741516, "learning_rate": 2.0206427038032326e-06, "loss": 0.0129, "step": 73100 }, { "grad_norm": 0.2077605128288269, "learning_rate": 2.0148304825495457e-06, "loss": 0.0075, "step": 73110 }, { "grad_norm": 0.15655282139778137, "learning_rate": 2.0090264606489496e-06, "loss": 0.0103, "step": 73120 }, { "grad_norm": 0.17700880765914917, "learning_rate": 2.003230639093162e-06, "loss": 0.0137, "step": 73130 }, { "grad_norm": 0.19147124886512756, "learning_rate": 1.9974430188725525e-06, "loss": 0.0126, "step": 73140 }, { "grad_norm": 0.30452725291252136, "learning_rate": 1.9916636009760513e-06, "loss": 0.0098, "step": 73150 }, { "grad_norm": 0.10634607821702957, "learning_rate": 1.985892386391208e-06, "loss": 0.0087, "step": 73160 }, { "grad_norm": 0.1635483056306839, "learning_rate": 1.980129376104173e-06, "loss": 0.01, "step": 73170 }, { "grad_norm": 0.11905800551176071, "learning_rate": 1.9743745710996796e-06, "loss": 0.0094, "step": 73180 }, { "grad_norm": 0.22088681161403656, "learning_rate": 1.9686279723610757e-06, "loss": 0.0095, "step": 73190 }, { "grad_norm": 0.26241159439086914, "learning_rate": 1.962889580870281e-06, "loss": 0.0151, "step": 73200 }, { "grad_norm": 0.27416154742240906, "learning_rate": 1.9571593976078385e-06, "loss": 0.0098, "step": 73210 }, { "grad_norm": 0.33958396315574646, "learning_rate": 1.9514374235528876e-06, "loss": 0.0139, "step": 73220 }, { "grad_norm": 0.16338738799095154, "learning_rate": 1.9457236596831408e-06, "loss": 0.0095, "step": 73230 }, { "grad_norm": 0.4635172188282013, "learning_rate": 1.940018106974939e-06, "loss": 0.014, "step": 73240 }, { "grad_norm": 0.3081735670566559, "learning_rate": 1.9343207664032025e-06, "loss": 0.0087, "step": 73250 }, { "grad_norm": 0.13664965331554413, "learning_rate": 1.928631638941436e-06, "loss": 0.0082, "step": 73260 }, { "grad_norm": 0.2685193717479706, "learning_rate": 1.922950725561784e-06, "loss": 0.0102, "step": 73270 }, { "grad_norm": 0.2518281936645508, "learning_rate": 1.9172780272349312e-06, "loss": 0.0081, "step": 73280 }, { "grad_norm": 0.12702520191669464, "learning_rate": 1.911613544930202e-06, "loss": 0.0104, "step": 73290 }, { "grad_norm": 0.7191539406776428, "learning_rate": 1.9059572796155e-06, "loss": 0.0118, "step": 73300 }, { "grad_norm": 0.16214630007743835, "learning_rate": 1.9003092322573192e-06, "loss": 0.0108, "step": 73310 }, { "grad_norm": 0.4473765790462494, "learning_rate": 1.8946694038207647e-06, "loss": 0.0144, "step": 73320 }, { "grad_norm": 0.22378169000148773, "learning_rate": 1.8890377952695215e-06, "loss": 0.0094, "step": 73330 }, { "grad_norm": 0.2166965901851654, "learning_rate": 1.8834144075658865e-06, "loss": 0.0081, "step": 73340 }, { "grad_norm": 0.1346517652273178, "learning_rate": 1.8777992416707357e-06, "loss": 0.0144, "step": 73350 }, { "grad_norm": 0.4823218286037445, "learning_rate": 1.8721922985435458e-06, "loss": 0.0082, "step": 73360 }, { "grad_norm": 0.24111677706241608, "learning_rate": 1.8665935791424006e-06, "loss": 0.0103, "step": 73370 }, { "grad_norm": 0.7997989654541016, "learning_rate": 1.8610030844239512e-06, "loss": 0.0094, "step": 73380 }, { "grad_norm": 0.22157903015613556, "learning_rate": 1.8554208153434838e-06, "loss": 0.0088, "step": 73390 }, { "grad_norm": 0.17002959549427032, "learning_rate": 1.8498467728548296e-06, "loss": 0.0085, "step": 73400 }, { "grad_norm": 0.11867775022983551, "learning_rate": 1.8442809579104547e-06, "loss": 0.0099, "step": 73410 }, { "grad_norm": 4.102721214294434, "learning_rate": 1.8387233714614039e-06, "loss": 0.0103, "step": 73420 }, { "grad_norm": 0.6344323754310608, "learning_rate": 1.8331740144573174e-06, "loss": 0.0124, "step": 73430 }, { "grad_norm": 0.15502984821796417, "learning_rate": 1.8276328878464255e-06, "loss": 0.0088, "step": 73440 }, { "grad_norm": 0.26037177443504333, "learning_rate": 1.82209999257556e-06, "loss": 0.014, "step": 73450 }, { "grad_norm": 0.7099740505218506, "learning_rate": 1.8165753295901311e-06, "loss": 0.0096, "step": 73460 }, { "grad_norm": 0.1954866349697113, "learning_rate": 1.8110588998341616e-06, "loss": 0.0114, "step": 73470 }, { "grad_norm": 0.21499530971050262, "learning_rate": 1.805550704250253e-06, "loss": 0.0151, "step": 73480 }, { "grad_norm": 0.28189852833747864, "learning_rate": 1.8000507437796077e-06, "loss": 0.0102, "step": 73490 }, { "grad_norm": 0.797357439994812, "learning_rate": 1.7945590193620242e-06, "loss": 0.01, "step": 73500 }, { "grad_norm": 0.16841654479503632, "learning_rate": 1.7890755319358742e-06, "loss": 0.0118, "step": 73510 }, { "grad_norm": 0.18641075491905212, "learning_rate": 1.7836002824381525e-06, "loss": 0.0111, "step": 73520 }, { "grad_norm": 0.3129250407218933, "learning_rate": 1.7781332718044165e-06, "loss": 0.008, "step": 73530 }, { "grad_norm": 0.14972995221614838, "learning_rate": 1.772674500968835e-06, "loss": 0.0098, "step": 73540 }, { "grad_norm": 0.27616244554519653, "learning_rate": 1.7672239708641624e-06, "loss": 0.0099, "step": 73550 }, { "grad_norm": 0.7511342763900757, "learning_rate": 1.7617816824217371e-06, "loss": 0.0094, "step": 73560 }, { "grad_norm": 0.3355643153190613, "learning_rate": 1.7563476365715148e-06, "loss": 0.0138, "step": 73570 }, { "grad_norm": 0.1437343806028366, "learning_rate": 1.750921834242003e-06, "loss": 0.0142, "step": 73580 }, { "grad_norm": 0.22129610180854797, "learning_rate": 1.7455042763603436e-06, "loss": 0.0089, "step": 73590 }, { "grad_norm": 0.2546861171722412, "learning_rate": 1.7400949638522345e-06, "loss": 0.0098, "step": 73600 }, { "grad_norm": 0.27222004532814026, "learning_rate": 1.734693897641987e-06, "loss": 0.0203, "step": 73610 }, { "grad_norm": 0.37518638372421265, "learning_rate": 1.7293010786524955e-06, "loss": 0.0086, "step": 73620 }, { "grad_norm": 0.21392187476158142, "learning_rate": 1.72391650780524e-06, "loss": 0.0137, "step": 73630 }, { "grad_norm": 0.1828063279390335, "learning_rate": 1.718540186020301e-06, "loss": 0.011, "step": 73640 }, { "grad_norm": 0.153733029961586, "learning_rate": 1.7131721142163437e-06, "loss": 0.0093, "step": 73650 }, { "grad_norm": 0.2171110361814499, "learning_rate": 1.7078122933106233e-06, "loss": 0.0117, "step": 73660 }, { "grad_norm": 0.19327214360237122, "learning_rate": 1.7024607242189905e-06, "loss": 0.0133, "step": 73670 }, { "grad_norm": 0.2054198682308197, "learning_rate": 1.6971174078558749e-06, "loss": 0.007, "step": 73680 }, { "grad_norm": 0.24220405519008636, "learning_rate": 1.6917823451343073e-06, "loss": 0.0122, "step": 73690 }, { "grad_norm": 0.17841163277626038, "learning_rate": 1.6864555369659142e-06, "loss": 0.0111, "step": 73700 }, { "grad_norm": 0.16698716580867767, "learning_rate": 1.681136984260878e-06, "loss": 0.0109, "step": 73710 }, { "grad_norm": 0.15538634359836578, "learning_rate": 1.6758266879280172e-06, "loss": 0.0096, "step": 73720 }, { "grad_norm": 0.17333225905895233, "learning_rate": 1.6705246488746996e-06, "loss": 0.008, "step": 73730 }, { "grad_norm": 0.4738391637802124, "learning_rate": 1.6652308680069062e-06, "loss": 0.0121, "step": 73740 }, { "grad_norm": 0.17391449213027954, "learning_rate": 1.6599453462292081e-06, "loss": 0.0169, "step": 73750 }, { "grad_norm": 0.2818523049354553, "learning_rate": 1.6546680844447326e-06, "loss": 0.0094, "step": 73760 }, { "grad_norm": 0.1469571888446808, "learning_rate": 1.6493990835552475e-06, "loss": 0.0116, "step": 73770 }, { "grad_norm": 0.15272434055805206, "learning_rate": 1.6441383444610547e-06, "loss": 0.0089, "step": 73780 }, { "grad_norm": 0.20415009558200836, "learning_rate": 1.6388858680610908e-06, "loss": 0.0111, "step": 73790 }, { "grad_norm": 0.1442633420228958, "learning_rate": 1.6336416552528543e-06, "loss": 0.0097, "step": 73800 }, { "grad_norm": 0.145221546292305, "learning_rate": 1.6284057069324288e-06, "loss": 0.0126, "step": 73810 }, { "grad_norm": 0.2535047233104706, "learning_rate": 1.623178023994504e-06, "loss": 0.0113, "step": 73820 }, { "grad_norm": 0.12880435585975647, "learning_rate": 1.6179586073323483e-06, "loss": 0.0109, "step": 73830 }, { "grad_norm": 0.23289801180362701, "learning_rate": 1.612747457837821e-06, "loss": 0.0104, "step": 73840 }, { "grad_norm": 0.39079049229621887, "learning_rate": 1.607544576401354e-06, "loss": 0.0108, "step": 73850 }, { "grad_norm": 0.8502674102783203, "learning_rate": 1.6023499639119754e-06, "loss": 0.0098, "step": 73860 }, { "grad_norm": 0.19424404203891754, "learning_rate": 1.5971636212573138e-06, "loss": 0.0088, "step": 73870 }, { "grad_norm": 0.12023600190877914, "learning_rate": 1.5919855493235714e-06, "loss": 0.0109, "step": 73880 }, { "grad_norm": 0.1966428905725479, "learning_rate": 1.586815748995535e-06, "loss": 0.0101, "step": 73890 }, { "grad_norm": 0.2080332487821579, "learning_rate": 1.5816542211565866e-06, "loss": 0.0124, "step": 73900 }, { "grad_norm": 0.09964557737112045, "learning_rate": 1.5765009666886765e-06, "loss": 0.0074, "step": 73910 }, { "grad_norm": 0.184514120221138, "learning_rate": 1.5713559864723781e-06, "loss": 0.0112, "step": 73920 }, { "grad_norm": 0.1717393547296524, "learning_rate": 1.5662192813868104e-06, "loss": 0.0089, "step": 73930 }, { "grad_norm": 0.1840617060661316, "learning_rate": 1.561090852309699e-06, "loss": 0.0129, "step": 73940 }, { "grad_norm": 0.42932283878326416, "learning_rate": 1.5559707001173651e-06, "loss": 0.008, "step": 73950 }, { "grad_norm": 0.17400167882442474, "learning_rate": 1.5508588256846757e-06, "loss": 0.01, "step": 73960 }, { "grad_norm": 0.25098520517349243, "learning_rate": 1.5457552298851319e-06, "loss": 0.0097, "step": 73970 }, { "grad_norm": 0.23899511992931366, "learning_rate": 1.5406599135907918e-06, "loss": 0.0108, "step": 73980 }, { "grad_norm": 0.17232948541641235, "learning_rate": 1.5355728776723088e-06, "loss": 0.0113, "step": 73990 }, { "grad_norm": 0.13113655149936676, "learning_rate": 1.5304941229989155e-06, "loss": 0.0094, "step": 74000 }, { "grad_norm": 0.28166988492012024, "learning_rate": 1.5254236504384345e-06, "loss": 0.011, "step": 74010 }, { "grad_norm": 0.21562640368938446, "learning_rate": 1.5203614608572726e-06, "loss": 0.0116, "step": 74020 }, { "grad_norm": 0.292778342962265, "learning_rate": 1.5153075551204044e-06, "loss": 0.0108, "step": 74030 }, { "grad_norm": 0.14260436594486237, "learning_rate": 1.5102619340914225e-06, "loss": 0.0184, "step": 74040 }, { "grad_norm": 0.17916488647460938, "learning_rate": 1.505224598632482e-06, "loss": 0.0077, "step": 74050 }, { "grad_norm": 0.3442004323005676, "learning_rate": 1.5001955496043164e-06, "loss": 0.0145, "step": 74060 }, { "grad_norm": 0.22295933961868286, "learning_rate": 1.4951747878662604e-06, "loss": 0.0095, "step": 74070 }, { "grad_norm": 0.13909095525741577, "learning_rate": 1.4901623142762221e-06, "loss": 0.0092, "step": 74080 }, { "grad_norm": 3.315913438796997, "learning_rate": 1.4851581296907001e-06, "loss": 0.0118, "step": 74090 }, { "grad_norm": 0.25061464309692383, "learning_rate": 1.4801622349647714e-06, "loss": 0.0084, "step": 74100 }, { "grad_norm": 0.9695092439651489, "learning_rate": 1.4751746309520976e-06, "loss": 0.0101, "step": 74110 }, { "grad_norm": 0.1156042143702507, "learning_rate": 1.4701953185049132e-06, "loss": 0.0152, "step": 74120 }, { "grad_norm": 0.6460617184638977, "learning_rate": 1.4652242984740661e-06, "loss": 0.011, "step": 74130 }, { "grad_norm": 0.17426720261573792, "learning_rate": 1.4602615717089484e-06, "loss": 0.0151, "step": 74140 }, { "grad_norm": 0.19332432746887207, "learning_rate": 1.4553071390575657e-06, "loss": 0.009, "step": 74150 }, { "grad_norm": 0.1963060349225998, "learning_rate": 1.450361001366496e-06, "loss": 0.0132, "step": 74160 }, { "grad_norm": 0.18371638655662537, "learning_rate": 1.4454231594809021e-06, "loss": 0.0102, "step": 74170 }, { "grad_norm": 0.32144850492477417, "learning_rate": 1.4404936142445036e-06, "loss": 0.0095, "step": 74180 }, { "grad_norm": 3.780851125717163, "learning_rate": 1.4355723664996546e-06, "loss": 0.009, "step": 74190 }, { "grad_norm": 0.2528015673160553, "learning_rate": 1.4306594170872433e-06, "loss": 0.0122, "step": 74200 }, { "grad_norm": 0.13697995245456696, "learning_rate": 1.4257547668467598e-06, "loss": 0.017, "step": 74210 }, { "grad_norm": 0.1486227810382843, "learning_rate": 1.4208584166162886e-06, "loss": 0.0115, "step": 74220 }, { "grad_norm": 0.5022885799407959, "learning_rate": 1.4159703672324553e-06, "loss": 0.0125, "step": 74230 }, { "grad_norm": 0.13070449233055115, "learning_rate": 1.4110906195305252e-06, "loss": 0.01, "step": 74240 }, { "grad_norm": 0.31007853150367737, "learning_rate": 1.406219174344292e-06, "loss": 0.011, "step": 74250 }, { "grad_norm": 0.16660797595977783, "learning_rate": 1.401356032506157e-06, "loss": 0.009, "step": 74260 }, { "grad_norm": 0.23991833627223969, "learning_rate": 1.3965011948471051e-06, "loss": 0.0175, "step": 74270 }, { "grad_norm": 0.17675939202308655, "learning_rate": 1.3916546621966842e-06, "loss": 0.0107, "step": 74280 }, { "grad_norm": 0.4803307056427002, "learning_rate": 1.3868164353830482e-06, "loss": 0.0078, "step": 74290 }, { "grad_norm": 0.18950998783111572, "learning_rate": 1.3819865152329082e-06, "loss": 0.0091, "step": 74300 }, { "grad_norm": 0.4204982817173004, "learning_rate": 1.377164902571565e-06, "loss": 0.0152, "step": 74310 }, { "grad_norm": 1.0879154205322266, "learning_rate": 1.372351598222904e-06, "loss": 0.0103, "step": 74320 }, { "grad_norm": 0.20684580504894257, "learning_rate": 1.3675466030093843e-06, "loss": 0.0103, "step": 74330 }, { "grad_norm": 0.18138007819652557, "learning_rate": 1.3627499177520486e-06, "loss": 0.0127, "step": 74340 }, { "grad_norm": 0.2025754153728485, "learning_rate": 1.3579615432705196e-06, "loss": 0.0095, "step": 74350 }, { "grad_norm": 0.14747178554534912, "learning_rate": 1.3531814803829978e-06, "loss": 0.0112, "step": 74360 }, { "grad_norm": 0.18533235788345337, "learning_rate": 1.3484097299062803e-06, "loss": 0.0091, "step": 74370 }, { "grad_norm": 0.11318143457174301, "learning_rate": 1.343646292655698e-06, "loss": 0.0119, "step": 74380 }, { "grad_norm": 0.21506188809871674, "learning_rate": 1.3388911694452277e-06, "loss": 0.0124, "step": 74390 }, { "grad_norm": 0.26110896468162537, "learning_rate": 1.3341443610873638e-06, "loss": 0.0125, "step": 74400 }, { "grad_norm": 0.25353294610977173, "learning_rate": 1.329405868393213e-06, "loss": 0.01, "step": 74410 }, { "grad_norm": 0.18583625555038452, "learning_rate": 1.3246756921724613e-06, "loss": 0.0115, "step": 74420 }, { "grad_norm": 0.1316080540418625, "learning_rate": 1.3199538332333506e-06, "loss": 0.0108, "step": 74430 }, { "grad_norm": 0.40463393926620483, "learning_rate": 1.3152402923827411e-06, "loss": 0.0104, "step": 74440 }, { "grad_norm": 0.16836485266685486, "learning_rate": 1.3105350704260277e-06, "loss": 0.0108, "step": 74450 }, { "grad_norm": 0.4479907751083374, "learning_rate": 1.305838168167206e-06, "loss": 0.0084, "step": 74460 }, { "grad_norm": 0.49078506231307983, "learning_rate": 1.301149586408862e-06, "loss": 0.0075, "step": 74470 }, { "grad_norm": 0.1530444324016571, "learning_rate": 1.2964693259521322e-06, "loss": 0.013, "step": 74480 }, { "grad_norm": 0.27178239822387695, "learning_rate": 1.2917973875967548e-06, "loss": 0.009, "step": 74490 }, { "grad_norm": 0.3711223304271698, "learning_rate": 1.2871337721410249e-06, "loss": 0.0114, "step": 74500 }, { "grad_norm": 0.5773574709892273, "learning_rate": 1.2824784803818379e-06, "loss": 0.0114, "step": 74510 }, { "grad_norm": 0.1555749475955963, "learning_rate": 1.2778315131146467e-06, "loss": 0.0084, "step": 74520 }, { "grad_norm": 0.21842122077941895, "learning_rate": 1.2731928711334994e-06, "loss": 0.0128, "step": 74530 }, { "grad_norm": 0.24243475496768951, "learning_rate": 1.2685625552310122e-06, "loss": 0.0103, "step": 74540 }, { "grad_norm": 0.17750631272792816, "learning_rate": 1.263940566198374e-06, "loss": 0.0132, "step": 74550 }, { "grad_norm": 0.252316415309906, "learning_rate": 1.259326904825353e-06, "loss": 0.0082, "step": 74560 }, { "grad_norm": 5.581615924835205, "learning_rate": 1.2547215719003137e-06, "loss": 0.0121, "step": 74570 }, { "grad_norm": 0.13957644999027252, "learning_rate": 1.2501245682101703e-06, "loss": 0.0079, "step": 74580 }, { "grad_norm": 0.1295253187417984, "learning_rate": 1.2455358945404171e-06, "loss": 0.0107, "step": 74590 }, { "grad_norm": 0.27102944254875183, "learning_rate": 1.2409555516751493e-06, "loss": 0.0093, "step": 74600 }, { "grad_norm": 0.9533193707466125, "learning_rate": 1.2363835403970125e-06, "loss": 0.0094, "step": 74610 }, { "grad_norm": 0.2813533544540405, "learning_rate": 1.2318198614872489e-06, "loss": 0.0114, "step": 74620 }, { "grad_norm": 0.183307945728302, "learning_rate": 1.2272645157256457e-06, "loss": 0.0098, "step": 74630 }, { "grad_norm": 0.2754093110561371, "learning_rate": 1.222717503890608e-06, "loss": 0.0134, "step": 74640 }, { "grad_norm": 0.3229982554912567, "learning_rate": 1.218178826759081e-06, "loss": 0.0143, "step": 74650 }, { "grad_norm": 0.21477609872817993, "learning_rate": 1.213648485106611e-06, "loss": 0.0104, "step": 74660 }, { "grad_norm": 3.4238226413726807, "learning_rate": 1.2091264797073066e-06, "loss": 0.0185, "step": 74670 }, { "grad_norm": 0.20058700442314148, "learning_rate": 1.2046128113338494e-06, "loss": 0.0243, "step": 74680 }, { "grad_norm": 0.6232661604881287, "learning_rate": 1.200107480757512e-06, "loss": 0.012, "step": 74690 }, { "grad_norm": 0.3345671594142914, "learning_rate": 1.1956104887481168e-06, "loss": 0.017, "step": 74700 }, { "grad_norm": 0.6589882373809814, "learning_rate": 1.191121836074094e-06, "loss": 0.0097, "step": 74710 }, { "grad_norm": 0.21540436148643494, "learning_rate": 1.1866415235024186e-06, "loss": 0.0069, "step": 74720 }, { "grad_norm": 0.16234417259693146, "learning_rate": 1.182169551798662e-06, "loss": 0.013, "step": 74730 }, { "grad_norm": 0.3916054666042328, "learning_rate": 1.177705921726957e-06, "loss": 0.0134, "step": 74740 }, { "grad_norm": 0.19777703285217285, "learning_rate": 1.173250634050016e-06, "loss": 0.0131, "step": 74750 }, { "grad_norm": 0.16065451502799988, "learning_rate": 1.1688036895291354e-06, "loss": 0.0141, "step": 74760 }, { "grad_norm": 0.16307571530342102, "learning_rate": 1.1643650889241574e-06, "loss": 0.0085, "step": 74770 }, { "grad_norm": 0.2586275041103363, "learning_rate": 1.1599348329935311e-06, "loss": 0.011, "step": 74780 }, { "grad_norm": 0.17107966542243958, "learning_rate": 1.1555129224942673e-06, "loss": 0.0148, "step": 74790 }, { "grad_norm": 0.1824377030134201, "learning_rate": 1.1510993581819396e-06, "loss": 0.009, "step": 74800 }, { "grad_norm": 0.2641605734825134, "learning_rate": 1.1466941408107112e-06, "loss": 0.0145, "step": 74810 }, { "grad_norm": 0.18834801018238068, "learning_rate": 1.1422972711333247e-06, "loss": 0.0087, "step": 74820 }, { "grad_norm": 0.2818794250488281, "learning_rate": 1.1379087499010565e-06, "loss": 0.0115, "step": 74830 }, { "grad_norm": 0.254031240940094, "learning_rate": 1.1335285778638127e-06, "loss": 0.0119, "step": 74840 }, { "grad_norm": 0.23252561688423157, "learning_rate": 1.1291567557700333e-06, "loss": 0.012, "step": 74850 }, { "grad_norm": 0.23002780973911285, "learning_rate": 1.124793284366743e-06, "loss": 0.0099, "step": 74860 }, { "grad_norm": 0.09863279014825821, "learning_rate": 1.1204381643995455e-06, "loss": 0.0096, "step": 74870 }, { "grad_norm": 0.13890215754508972, "learning_rate": 1.116091396612595e-06, "loss": 0.0094, "step": 74880 }, { "grad_norm": 0.14439213275909424, "learning_rate": 1.1117529817486538e-06, "loss": 0.0081, "step": 74890 }, { "grad_norm": 0.19118601083755493, "learning_rate": 1.107422920549034e-06, "loss": 0.0085, "step": 74900 }, { "grad_norm": 0.1940375119447708, "learning_rate": 1.1031012137536157e-06, "loss": 0.0104, "step": 74910 }, { "grad_norm": 0.2097800225019455, "learning_rate": 1.0987878621008695e-06, "loss": 0.0108, "step": 74920 }, { "grad_norm": 0.19421735405921936, "learning_rate": 1.0944828663278284e-06, "loss": 0.0096, "step": 74930 }, { "grad_norm": 0.2202843576669693, "learning_rate": 1.0901862271700925e-06, "loss": 0.012, "step": 74940 }, { "grad_norm": 0.2143850326538086, "learning_rate": 1.0858979453618467e-06, "loss": 0.0099, "step": 74950 }, { "grad_norm": 0.24833229184150696, "learning_rate": 1.0816180216358441e-06, "loss": 0.0086, "step": 74960 }, { "grad_norm": 0.2669066786766052, "learning_rate": 1.0773464567233937e-06, "loss": 0.0105, "step": 74970 }, { "grad_norm": 0.17623524367809296, "learning_rate": 1.0730832513543953e-06, "loss": 0.0106, "step": 74980 }, { "grad_norm": 1.6047691106796265, "learning_rate": 1.0688284062573218e-06, "loss": 0.0104, "step": 74990 }, { "grad_norm": 0.16411106288433075, "learning_rate": 1.0645819221591969e-06, "loss": 0.0075, "step": 75000 }, { "grad_norm": 0.20805048942565918, "learning_rate": 1.0603437997856347e-06, "loss": 0.011, "step": 75010 }, { "grad_norm": 0.2410765290260315, "learning_rate": 1.0561140398608228e-06, "loss": 0.0117, "step": 75020 }, { "grad_norm": 0.11669980734586716, "learning_rate": 1.0518926431074937e-06, "loss": 0.0098, "step": 75030 }, { "grad_norm": 0.4601820409297943, "learning_rate": 1.0476796102469877e-06, "loss": 0.0105, "step": 75040 }, { "grad_norm": 0.14533120393753052, "learning_rate": 1.0434749419991786e-06, "loss": 0.0113, "step": 75050 }, { "grad_norm": 0.12296734750270844, "learning_rate": 1.0392786390825415e-06, "loss": 0.01, "step": 75060 }, { "grad_norm": 0.18259233236312866, "learning_rate": 1.0350907022141087e-06, "loss": 0.0106, "step": 75070 }, { "grad_norm": 0.22139687836170197, "learning_rate": 1.0309111321094744e-06, "loss": 0.009, "step": 75080 }, { "grad_norm": 0.8562966585159302, "learning_rate": 1.026739929482834e-06, "loss": 0.0093, "step": 75090 }, { "grad_norm": 0.1359313577413559, "learning_rate": 1.022577095046906e-06, "loss": 0.0103, "step": 75100 }, { "grad_norm": 0.2713600993156433, "learning_rate": 1.0184226295130217e-06, "loss": 0.0108, "step": 75110 }, { "grad_norm": 0.23990701138973236, "learning_rate": 1.0142765335910575e-06, "loss": 0.0105, "step": 75120 }, { "grad_norm": 0.2989460229873657, "learning_rate": 1.0101388079894746e-06, "loss": 0.0095, "step": 75130 }, { "grad_norm": 0.24826502799987793, "learning_rate": 1.0060094534152908e-06, "loss": 0.009, "step": 75140 }, { "grad_norm": 0.175815150141716, "learning_rate": 1.0018884705741028e-06, "loss": 0.0107, "step": 75150 }, { "grad_norm": 0.22180861234664917, "learning_rate": 9.977758601700805e-07, "loss": 0.0104, "step": 75160 }, { "grad_norm": 0.7383211255073547, "learning_rate": 9.93671622905945e-07, "loss": 0.0144, "step": 75170 }, { "grad_norm": 0.21049608290195465, "learning_rate": 9.895757594830024e-07, "loss": 0.0093, "step": 75180 }, { "grad_norm": 0.3197147846221924, "learning_rate": 9.854882706011204e-07, "loss": 0.0098, "step": 75190 }, { "grad_norm": 0.20985586941242218, "learning_rate": 9.814091569587513e-07, "loss": 0.0085, "step": 75200 }, { "grad_norm": 0.3474196195602417, "learning_rate": 9.77338419252888e-07, "loss": 0.0125, "step": 75210 }, { "grad_norm": 0.24395117163658142, "learning_rate": 9.732760581791289e-07, "loss": 0.0097, "step": 75220 }, { "grad_norm": 0.17758908867835999, "learning_rate": 9.692220744315972e-07, "loss": 0.0105, "step": 75230 }, { "grad_norm": 1.3216630220413208, "learning_rate": 9.651764687030162e-07, "loss": 0.009, "step": 75240 }, { "grad_norm": 0.1942911297082901, "learning_rate": 9.611392416846776e-07, "loss": 0.0074, "step": 75250 }, { "grad_norm": 0.13625973463058472, "learning_rate": 9.571103940664238e-07, "loss": 0.011, "step": 75260 }, { "grad_norm": 0.22974826395511627, "learning_rate": 9.530899265366822e-07, "loss": 0.0103, "step": 75270 }, { "grad_norm": 0.20032423734664917, "learning_rate": 9.490778397824307e-07, "loss": 0.0108, "step": 75280 }, { "grad_norm": 0.35320043563842773, "learning_rate": 9.450741344892322e-07, "loss": 0.0084, "step": 75290 }, { "grad_norm": 0.2574051022529602, "learning_rate": 9.410788113412116e-07, "loss": 0.0097, "step": 75300 }, { "grad_norm": 0.32364362478256226, "learning_rate": 9.370918710210563e-07, "loss": 0.0102, "step": 75310 }, { "grad_norm": 0.1790417730808258, "learning_rate": 9.331133142100268e-07, "loss": 0.0101, "step": 75320 }, { "grad_norm": 0.16396977007389069, "learning_rate": 9.291431415879459e-07, "loss": 0.0068, "step": 75330 }, { "grad_norm": 0.2808658182621002, "learning_rate": 9.251813538332155e-07, "loss": 0.0104, "step": 75340 }, { "grad_norm": 2.358292579650879, "learning_rate": 9.212279516227884e-07, "loss": 0.0113, "step": 75350 }, { "grad_norm": 0.3138565719127655, "learning_rate": 9.172829356322021e-07, "loss": 0.0106, "step": 75360 }, { "grad_norm": 0.1925482153892517, "learning_rate": 9.133463065355452e-07, "loss": 0.0093, "step": 75370 }, { "grad_norm": 0.15485192835330963, "learning_rate": 9.094180650054796e-07, "loss": 0.0098, "step": 75380 }, { "grad_norm": 0.7052704691886902, "learning_rate": 9.054982117132404e-07, "loss": 0.0094, "step": 75390 }, { "grad_norm": 0.21323269605636597, "learning_rate": 9.015867473286143e-07, "loss": 0.0109, "step": 75400 }, { "grad_norm": 0.24328888952732086, "learning_rate": 8.976836725199778e-07, "loss": 0.0085, "step": 75410 }, { "grad_norm": 0.2861866056919098, "learning_rate": 8.937889879542416e-07, "loss": 0.0094, "step": 75420 }, { "grad_norm": 0.7449032068252563, "learning_rate": 8.899026942969068e-07, "loss": 0.0085, "step": 75430 }, { "grad_norm": 1.016787052154541, "learning_rate": 8.860247922120424e-07, "loss": 0.0106, "step": 75440 }, { "grad_norm": 0.27553603053092957, "learning_rate": 8.821552823622737e-07, "loss": 0.0072, "step": 75450 }, { "grad_norm": 0.28824758529663086, "learning_rate": 8.782941654087884e-07, "loss": 0.0095, "step": 75460 }, { "grad_norm": 0.2733476459980011, "learning_rate": 8.744414420113478e-07, "loss": 0.0102, "step": 75470 }, { "grad_norm": 0.14944125711917877, "learning_rate": 8.705971128282753e-07, "loss": 0.0098, "step": 75480 }, { "grad_norm": 0.19256439805030823, "learning_rate": 8.66761178516473e-07, "loss": 0.0081, "step": 75490 }, { "grad_norm": 0.14925995469093323, "learning_rate": 8.629336397313781e-07, "loss": 0.0087, "step": 75500 }, { "grad_norm": 0.3255559802055359, "learning_rate": 8.591144971270282e-07, "loss": 0.0097, "step": 75510 }, { "grad_norm": 0.17034614086151123, "learning_rate": 8.553037513560069e-07, "loss": 0.0164, "step": 75520 }, { "grad_norm": 0.16662323474884033, "learning_rate": 8.515014030694546e-07, "loss": 0.0071, "step": 75530 }, { "grad_norm": 0.15129488706588745, "learning_rate": 8.47707452917107e-07, "loss": 0.0089, "step": 75540 }, { "grad_norm": 0.24812051653862, "learning_rate": 8.439219015472343e-07, "loss": 0.0107, "step": 75550 }, { "grad_norm": 0.7562601566314697, "learning_rate": 8.401447496066861e-07, "loss": 0.009, "step": 75560 }, { "grad_norm": 0.21160945296287537, "learning_rate": 8.363759977408792e-07, "loss": 0.0111, "step": 75570 }, { "grad_norm": 0.3234516680240631, "learning_rate": 8.326156465937817e-07, "loss": 0.0085, "step": 75580 }, { "grad_norm": 0.33436501026153564, "learning_rate": 8.28863696807941e-07, "loss": 0.0156, "step": 75590 }, { "grad_norm": 0.8744014501571655, "learning_rate": 8.251201490244609e-07, "loss": 0.0121, "step": 75600 }, { "grad_norm": 0.32153886556625366, "learning_rate": 8.213850038830129e-07, "loss": 0.0094, "step": 75610 }, { "grad_norm": 0.4903121888637543, "learning_rate": 8.176582620218309e-07, "loss": 0.0111, "step": 75620 }, { "grad_norm": 0.27844932675361633, "learning_rate": 8.139399240777057e-07, "loss": 0.0075, "step": 75630 }, { "grad_norm": 0.3291384279727936, "learning_rate": 8.102299906860122e-07, "loss": 0.0118, "step": 75640 }, { "grad_norm": 0.2511135935783386, "learning_rate": 8.065284624806657e-07, "loss": 0.0134, "step": 75650 }, { "grad_norm": 0.6349096298217773, "learning_rate": 8.028353400941601e-07, "loss": 0.0085, "step": 75660 }, { "grad_norm": 0.1778196543455124, "learning_rate": 7.991506241575575e-07, "loss": 0.008, "step": 75670 }, { "grad_norm": 0.23285146057605743, "learning_rate": 7.954743153004541e-07, "loss": 0.0093, "step": 75680 }, { "grad_norm": 0.1541273593902588, "learning_rate": 7.918064141510528e-07, "loss": 0.0107, "step": 75690 }, { "grad_norm": 0.15995623171329498, "learning_rate": 7.881469213360859e-07, "loss": 0.0126, "step": 75700 }, { "grad_norm": 0.14648760855197906, "learning_rate": 7.844958374808642e-07, "loss": 0.0081, "step": 75710 }, { "grad_norm": 0.2337225377559662, "learning_rate": 7.808531632092608e-07, "loss": 0.009, "step": 75720 }, { "grad_norm": 0.1488625854253769, "learning_rate": 7.772188991436946e-07, "loss": 0.0097, "step": 75730 }, { "grad_norm": 0.08488760143518448, "learning_rate": 7.735930459051799e-07, "loss": 0.0085, "step": 75740 }, { "grad_norm": 0.164400115609169, "learning_rate": 7.699756041132655e-07, "loss": 0.01, "step": 75750 }, { "grad_norm": 0.2698666453361511, "learning_rate": 7.663665743860793e-07, "loss": 0.0074, "step": 75760 }, { "grad_norm": 0.49149957299232483, "learning_rate": 7.627659573403001e-07, "loss": 0.011, "step": 75770 }, { "grad_norm": 0.13620170950889587, "learning_rate": 7.591737535911802e-07, "loss": 0.0082, "step": 75780 }, { "grad_norm": 0.17739668488502502, "learning_rate": 7.555899637525288e-07, "loss": 0.0141, "step": 75790 }, { "grad_norm": 0.20831283926963806, "learning_rate": 7.520145884367058e-07, "loss": 0.008, "step": 75800 }, { "grad_norm": 0.2383134514093399, "learning_rate": 7.484476282546615e-07, "loss": 0.0101, "step": 75810 }, { "grad_norm": 0.284511536359787, "learning_rate": 7.448890838158806e-07, "loss": 0.0131, "step": 75820 }, { "grad_norm": 0.1396934688091278, "learning_rate": 7.413389557284267e-07, "loss": 0.01, "step": 75830 }, { "grad_norm": 0.15732112526893616, "learning_rate": 7.377972445989201e-07, "loss": 0.01, "step": 75840 }, { "grad_norm": 0.19327794015407562, "learning_rate": 7.342639510325377e-07, "loss": 0.0096, "step": 75850 }, { "grad_norm": 0.251003623008728, "learning_rate": 7.307390756330246e-07, "loss": 0.0121, "step": 75860 }, { "grad_norm": 0.2244129180908203, "learning_rate": 7.272226190026876e-07, "loss": 0.0095, "step": 75870 }, { "grad_norm": 0.38178905844688416, "learning_rate": 7.237145817423907e-07, "loss": 0.0094, "step": 75880 }, { "grad_norm": 0.14958947896957397, "learning_rate": 7.202149644515654e-07, "loss": 0.0091, "step": 75890 }, { "grad_norm": 0.238073468208313, "learning_rate": 7.167237677281946e-07, "loss": 0.0092, "step": 75900 }, { "grad_norm": 0.14784109592437744, "learning_rate": 7.132409921688288e-07, "loss": 0.0093, "step": 75910 }, { "grad_norm": 0.18409954011440277, "learning_rate": 7.09766638368592e-07, "loss": 0.0066, "step": 75920 }, { "grad_norm": 0.1601049304008484, "learning_rate": 7.063007069211313e-07, "loss": 0.0088, "step": 75930 }, { "grad_norm": 0.1932772547006607, "learning_rate": 7.028431984187067e-07, "loss": 0.0092, "step": 75940 }, { "grad_norm": 0.1426909863948822, "learning_rate": 6.993941134520898e-07, "loss": 0.0063, "step": 75950 }, { "grad_norm": 0.19194428622722626, "learning_rate": 6.959534526106537e-07, "loss": 0.0087, "step": 75960 }, { "grad_norm": 0.3294405937194824, "learning_rate": 6.925212164822948e-07, "loss": 0.008, "step": 75970 }, { "grad_norm": 0.36542248725891113, "learning_rate": 6.890974056535049e-07, "loss": 0.0113, "step": 75980 }, { "grad_norm": 0.13097970187664032, "learning_rate": 6.856820207093106e-07, "loss": 0.0091, "step": 75990 }, { "grad_norm": 0.20677906274795532, "learning_rate": 6.822750622333063e-07, "loss": 0.008, "step": 76000 }, { "grad_norm": 0.18536144495010376, "learning_rate": 6.788765308076539e-07, "loss": 0.0101, "step": 76010 }, { "grad_norm": 0.20870593190193176, "learning_rate": 6.754864270130668e-07, "loss": 0.009, "step": 76020 }, { "grad_norm": 0.18259228765964508, "learning_rate": 6.721047514288203e-07, "loss": 0.0102, "step": 76030 }, { "grad_norm": 0.3132913410663605, "learning_rate": 6.687315046327469e-07, "loss": 0.0114, "step": 76040 }, { "grad_norm": 0.19560080766677856, "learning_rate": 6.65366687201252e-07, "loss": 0.0103, "step": 76050 }, { "grad_norm": 0.15851759910583496, "learning_rate": 6.620102997092814e-07, "loss": 0.0122, "step": 76060 }, { "grad_norm": 0.17389103770256042, "learning_rate": 6.586623427303596e-07, "loss": 0.0091, "step": 76070 }, { "grad_norm": 0.13639496266841888, "learning_rate": 6.553228168365455e-07, "loss": 0.0117, "step": 76080 }, { "grad_norm": 0.40884366631507874, "learning_rate": 6.519917225984884e-07, "loss": 0.0152, "step": 76090 }, { "grad_norm": 0.3249998092651367, "learning_rate": 6.486690605853718e-07, "loss": 0.0101, "step": 76100 }, { "grad_norm": 0.20137126743793488, "learning_rate": 6.453548313649527e-07, "loss": 0.0098, "step": 76110 }, { "grad_norm": 0.38845494389533997, "learning_rate": 6.420490355035446e-07, "loss": 0.0121, "step": 76120 }, { "grad_norm": 0.22523778676986694, "learning_rate": 6.387516735660071e-07, "loss": 0.01, "step": 76130 }, { "grad_norm": 0.14870700240135193, "learning_rate": 6.35462746115778e-07, "loss": 0.0083, "step": 76140 }, { "grad_norm": 0.31292784214019775, "learning_rate": 6.321822537148358e-07, "loss": 0.0124, "step": 76150 }, { "grad_norm": 0.20376275479793549, "learning_rate": 6.289101969237432e-07, "loss": 0.0106, "step": 76160 }, { "grad_norm": 0.2124198079109192, "learning_rate": 6.256465763015918e-07, "loss": 0.0105, "step": 76170 }, { "grad_norm": 0.09724719822406769, "learning_rate": 6.22391392406052e-07, "loss": 0.0065, "step": 76180 }, { "grad_norm": 0.13821089267730713, "learning_rate": 6.191446457933403e-07, "loss": 0.0072, "step": 76190 }, { "grad_norm": 0.22909504175186157, "learning_rate": 6.159063370182406e-07, "loss": 0.0089, "step": 76200 }, { "grad_norm": 2.183230400085449, "learning_rate": 6.126764666340879e-07, "loss": 0.0094, "step": 76210 }, { "grad_norm": 0.4662759304046631, "learning_rate": 6.094550351927852e-07, "loss": 0.0113, "step": 76220 }, { "grad_norm": 0.28054991364479065, "learning_rate": 6.062420432447757e-07, "loss": 0.0114, "step": 76230 }, { "grad_norm": 0.22394336760044098, "learning_rate": 6.030374913390813e-07, "loss": 0.0083, "step": 76240 }, { "grad_norm": 0.3015693426132202, "learning_rate": 5.998413800232694e-07, "loss": 0.0124, "step": 76250 }, { "grad_norm": 0.1523837298154831, "learning_rate": 5.966537098434755e-07, "loss": 0.0113, "step": 76260 }, { "grad_norm": 0.14384856820106506, "learning_rate": 5.934744813443694e-07, "loss": 0.0088, "step": 76270 }, { "grad_norm": 0.21538929641246796, "learning_rate": 5.903036950692054e-07, "loss": 0.0113, "step": 76280 }, { "grad_norm": 0.4036827087402344, "learning_rate": 5.871413515597835e-07, "loss": 0.0127, "step": 76290 }, { "grad_norm": 0.1979941874742508, "learning_rate": 5.839874513564547e-07, "loss": 0.0091, "step": 76300 }, { "grad_norm": 0.965743362903595, "learning_rate": 5.808419949981436e-07, "loss": 0.0108, "step": 76310 }, { "grad_norm": 0.16619767248630524, "learning_rate": 5.777049830223257e-07, "loss": 0.0095, "step": 76320 }, { "grad_norm": 0.2856212258338928, "learning_rate": 5.745764159650114e-07, "loss": 0.0084, "step": 76330 }, { "grad_norm": 0.26948267221450806, "learning_rate": 5.71456294360806e-07, "loss": 0.0094, "step": 76340 }, { "grad_norm": 0.17579126358032227, "learning_rate": 5.683446187428443e-07, "loss": 0.0106, "step": 76350 }, { "grad_norm": 0.37201741337776184, "learning_rate": 5.652413896428288e-07, "loss": 0.0096, "step": 76360 }, { "grad_norm": 0.33055534958839417, "learning_rate": 5.621466075910131e-07, "loss": 0.0105, "step": 76370 }, { "grad_norm": 0.07241054624319077, "learning_rate": 5.590602731162187e-07, "loss": 0.0077, "step": 76380 }, { "grad_norm": 0.1768777221441269, "learning_rate": 5.559823867458125e-07, "loss": 0.0085, "step": 76390 }, { "grad_norm": 0.22084306180477142, "learning_rate": 5.529129490057128e-07, "loss": 0.0143, "step": 76400 }, { "grad_norm": 0.18125680088996887, "learning_rate": 5.498519604204167e-07, "loss": 0.0097, "step": 76410 }, { "grad_norm": 0.16788344085216522, "learning_rate": 5.467994215129501e-07, "loss": 0.0086, "step": 76420 }, { "grad_norm": 0.323152095079422, "learning_rate": 5.437553328049183e-07, "loss": 0.0111, "step": 76430 }, { "grad_norm": 0.2133968621492386, "learning_rate": 5.407196948164661e-07, "loss": 0.0085, "step": 76440 }, { "grad_norm": 0.19676519930362701, "learning_rate": 5.376925080663009e-07, "loss": 0.01, "step": 76450 }, { "grad_norm": 0.22838391363620758, "learning_rate": 5.346737730716977e-07, "loss": 0.0151, "step": 76460 }, { "grad_norm": 0.34174713492393494, "learning_rate": 5.316634903484607e-07, "loss": 0.011, "step": 76470 }, { "grad_norm": 0.18708375096321106, "learning_rate": 5.286616604109728e-07, "loss": 0.0082, "step": 76480 }, { "grad_norm": 0.19204267859458923, "learning_rate": 5.256682837721627e-07, "loss": 0.0124, "step": 76490 }, { "grad_norm": 0.10576333850622177, "learning_rate": 5.226833609435156e-07, "loss": 0.0083, "step": 76500 }, { "grad_norm": 0.16770736873149872, "learning_rate": 5.19706892435079e-07, "loss": 0.0097, "step": 76510 }, { "grad_norm": 0.2454923689365387, "learning_rate": 5.167388787554406e-07, "loss": 0.0104, "step": 76520 }, { "grad_norm": 0.40710920095443726, "learning_rate": 5.137793204117614e-07, "loss": 0.0119, "step": 76530 }, { "grad_norm": 0.3372180461883545, "learning_rate": 5.108282179097479e-07, "loss": 0.0114, "step": 76540 }, { "grad_norm": 0.1770881712436676, "learning_rate": 5.078855717536523e-07, "loss": 0.0111, "step": 76550 }, { "grad_norm": 0.1928170919418335, "learning_rate": 5.049513824463059e-07, "loss": 0.008, "step": 76560 }, { "grad_norm": 0.2936924695968628, "learning_rate": 5.020256504890742e-07, "loss": 0.0079, "step": 76570 }, { "grad_norm": 0.4625551700592041, "learning_rate": 4.991083763818849e-07, "loss": 0.0114, "step": 76580 }, { "grad_norm": 0.23199157416820526, "learning_rate": 4.961995606232228e-07, "loss": 0.0138, "step": 76590 }, { "grad_norm": 0.24463775753974915, "learning_rate": 4.932992037101236e-07, "loss": 0.0116, "step": 76600 }, { "grad_norm": 0.1678517907857895, "learning_rate": 4.904073061381798e-07, "loss": 0.014, "step": 76610 }, { "grad_norm": 0.17990843951702118, "learning_rate": 4.875238684015349e-07, "loss": 0.0097, "step": 76620 }, { "grad_norm": 0.2608296573162079, "learning_rate": 4.846488909928948e-07, "loss": 0.0086, "step": 76630 }, { "grad_norm": 0.1699821650981903, "learning_rate": 4.81782374403511e-07, "loss": 0.0116, "step": 76640 }, { "grad_norm": 0.2828372120857239, "learning_rate": 4.789243191231918e-07, "loss": 0.0086, "step": 76650 }, { "grad_norm": 0.14154557883739471, "learning_rate": 4.76074725640302e-07, "loss": 0.0085, "step": 76660 }, { "grad_norm": 0.17837180197238922, "learning_rate": 4.732335944417632e-07, "loss": 0.0096, "step": 76670 }, { "grad_norm": 0.27304989099502563, "learning_rate": 4.704009260130371e-07, "loss": 0.009, "step": 76680 }, { "grad_norm": 0.07419931143522263, "learning_rate": 4.675767208381587e-07, "loss": 0.0082, "step": 76690 }, { "grad_norm": 0.15970659255981445, "learning_rate": 4.647609793997032e-07, "loss": 0.0097, "step": 76700 }, { "grad_norm": 0.13986316323280334, "learning_rate": 4.619537021788023e-07, "loss": 0.0095, "step": 76710 }, { "grad_norm": 0.21530263125896454, "learning_rate": 4.5915488965515005e-07, "loss": 0.0088, "step": 76720 }, { "grad_norm": 0.13152341544628143, "learning_rate": 4.563645423069807e-07, "loss": 0.01, "step": 76730 }, { "grad_norm": 0.24410831928253174, "learning_rate": 4.535826606110849e-07, "loss": 0.0083, "step": 76740 }, { "grad_norm": 0.2851121127605438, "learning_rate": 4.508092450428214e-07, "loss": 0.0119, "step": 76750 }, { "grad_norm": 0.1654936820268631, "learning_rate": 4.480442960760778e-07, "loss": 0.0102, "step": 76760 }, { "grad_norm": 0.1635598987340927, "learning_rate": 4.4528781418332053e-07, "loss": 0.0075, "step": 76770 }, { "grad_norm": 0.15534886717796326, "learning_rate": 4.4253979983555073e-07, "loss": 0.0139, "step": 76780 }, { "grad_norm": 0.42539381980895996, "learning_rate": 4.3980025350233154e-07, "loss": 0.0102, "step": 76790 }, { "grad_norm": 0.6663260459899902, "learning_rate": 4.370691756517664e-07, "loss": 0.0104, "step": 76800 }, { "grad_norm": 0.45489299297332764, "learning_rate": 4.3434656675053753e-07, "loss": 0.0086, "step": 76810 }, { "grad_norm": 0.25971922278404236, "learning_rate": 4.3163242726385613e-07, "loss": 0.0088, "step": 76820 }, { "grad_norm": 0.12150012701749802, "learning_rate": 4.289267576554956e-07, "loss": 0.0088, "step": 76830 }, { "grad_norm": 0.1871723234653473, "learning_rate": 4.262295583877807e-07, "loss": 0.0087, "step": 76840 }, { "grad_norm": 0.23869161307811737, "learning_rate": 4.235408299215815e-07, "loss": 0.0125, "step": 76850 }, { "grad_norm": 0.186075821518898, "learning_rate": 4.2086057271634173e-07, "loss": 0.0077, "step": 76860 }, { "grad_norm": 0.18672913312911987, "learning_rate": 4.181887872300394e-07, "loss": 0.0094, "step": 76870 }, { "grad_norm": 0.5337352752685547, "learning_rate": 4.155254739192038e-07, "loss": 0.0105, "step": 76880 }, { "grad_norm": 0.3386293053627014, "learning_rate": 4.128706332389265e-07, "loss": 0.0165, "step": 76890 }, { "grad_norm": 0.34587422013282776, "learning_rate": 4.1022426564284453e-07, "loss": 0.0088, "step": 76900 }, { "grad_norm": 0.10106837749481201, "learning_rate": 4.075863715831574e-07, "loss": 0.009, "step": 76910 }, { "grad_norm": 0.31480512022972107, "learning_rate": 4.0495695151059887e-07, "loss": 0.0091, "step": 76920 }, { "grad_norm": 0.25049155950546265, "learning_rate": 4.023360058744763e-07, "loss": 0.0078, "step": 76930 }, { "grad_norm": 0.12772662937641144, "learning_rate": 3.997235351226314e-07, "loss": 0.0072, "step": 76940 }, { "grad_norm": 0.11945295333862305, "learning_rate": 3.971195397014571e-07, "loss": 0.0096, "step": 76950 }, { "grad_norm": 0.20950168371200562, "learning_rate": 3.9452402005591417e-07, "loss": 0.0144, "step": 76960 }, { "grad_norm": 0.2207559049129486, "learning_rate": 3.9193697662950334e-07, "loss": 0.0086, "step": 76970 }, { "grad_norm": 0.4459674060344696, "learning_rate": 3.893584098642822e-07, "loss": 0.0124, "step": 76980 }, { "grad_norm": 0.16372641921043396, "learning_rate": 3.867883202008538e-07, "loss": 0.0118, "step": 76990 }, { "grad_norm": 0.2535099685192108, "learning_rate": 3.842267080783779e-07, "loss": 0.0112, "step": 77000 }, { "grad_norm": 0.2675168216228485, "learning_rate": 3.816735739345656e-07, "loss": 0.0102, "step": 77010 }, { "grad_norm": 0.10717524588108063, "learning_rate": 3.791289182056679e-07, "loss": 0.0077, "step": 77020 }, { "grad_norm": 0.13338547945022583, "learning_rate": 3.7659274132650913e-07, "loss": 0.0104, "step": 77030 }, { "grad_norm": 0.18697986006736755, "learning_rate": 3.7406504373044826e-07, "loss": 0.0096, "step": 77040 }, { "grad_norm": 0.8623630404472351, "learning_rate": 3.7154582584939533e-07, "loss": 0.0089, "step": 77050 }, { "grad_norm": 0.40001699328422546, "learning_rate": 3.690350881138227e-07, "loss": 0.009, "step": 77060 }, { "grad_norm": 0.3267580568790436, "learning_rate": 3.665328309527427e-07, "loss": 0.0117, "step": 77070 }, { "grad_norm": 0.1626030057668686, "learning_rate": 3.640390547937245e-07, "loss": 0.008, "step": 77080 }, { "grad_norm": 0.16145387291908264, "learning_rate": 3.615537600628882e-07, "loss": 0.0127, "step": 77090 }, { "grad_norm": 0.20030871033668518, "learning_rate": 3.5907694718489427e-07, "loss": 0.0114, "step": 77100 }, { "grad_norm": 0.12585419416427612, "learning_rate": 3.566086165829707e-07, "loss": 0.0136, "step": 77110 }, { "grad_norm": 0.3407885432243347, "learning_rate": 3.5414876867888024e-07, "loss": 0.0124, "step": 77120 }, { "grad_norm": 0.30127814412117004, "learning_rate": 3.5169740389295326e-07, "loss": 0.011, "step": 77130 }, { "grad_norm": 0.28768426179885864, "learning_rate": 3.4925452264405466e-07, "loss": 0.0078, "step": 77140 }, { "grad_norm": 0.2111709862947464, "learning_rate": 3.468201253496062e-07, "loss": 0.01, "step": 77150 }, { "grad_norm": 0.20258884131908417, "learning_rate": 3.443942124255861e-07, "loss": 0.0078, "step": 77160 }, { "grad_norm": 0.11393942683935165, "learning_rate": 3.4197678428650183e-07, "loss": 0.0087, "step": 77170 }, { "grad_norm": 0.5432651042938232, "learning_rate": 3.3956784134544504e-07, "loss": 0.0096, "step": 77180 }, { "grad_norm": 0.14167220890522003, "learning_rate": 3.3716738401402547e-07, "loss": 0.006, "step": 77190 }, { "grad_norm": 0.918855607509613, "learning_rate": 3.3477541270241495e-07, "loss": 0.014, "step": 77200 }, { "grad_norm": 0.21877238154411316, "learning_rate": 3.3239192781934215e-07, "loss": 0.0103, "step": 77210 }, { "grad_norm": 0.12725061178207397, "learning_rate": 3.3001692977207563e-07, "loss": 0.0076, "step": 77220 }, { "grad_norm": 1.3321290016174316, "learning_rate": 3.276504189664409e-07, "loss": 0.012, "step": 77230 }, { "grad_norm": 0.30361297726631165, "learning_rate": 3.252923958068088e-07, "loss": 0.0084, "step": 77240 }, { "grad_norm": 0.19070833921432495, "learning_rate": 3.2294286069609046e-07, "loss": 0.0112, "step": 77250 }, { "grad_norm": 0.2666250765323639, "learning_rate": 3.2060181403577583e-07, "loss": 0.0091, "step": 77260 }, { "grad_norm": 0.26005423069000244, "learning_rate": 3.1826925622587823e-07, "loss": 0.0087, "step": 77270 }, { "grad_norm": 0.17064787447452545, "learning_rate": 3.159451876649622e-07, "loss": 0.0085, "step": 77280 }, { "grad_norm": 0.23873096704483032, "learning_rate": 3.1362960875015444e-07, "loss": 0.0082, "step": 77290 }, { "grad_norm": 0.25472187995910645, "learning_rate": 3.1132251987711637e-07, "loss": 0.0093, "step": 77300 }, { "grad_norm": 0.2669500708580017, "learning_rate": 3.0902392144007695e-07, "loss": 0.0113, "step": 77310 }, { "grad_norm": 0.2555221617221832, "learning_rate": 3.0673381383179436e-07, "loss": 0.0147, "step": 77320 }, { "grad_norm": 0.17779341340065002, "learning_rate": 3.0445219744358875e-07, "loss": 0.0082, "step": 77330 }, { "grad_norm": 0.35056933760643005, "learning_rate": 3.021790726653262e-07, "loss": 0.0123, "step": 77340 }, { "grad_norm": 0.5068352222442627, "learning_rate": 2.9991443988542366e-07, "loss": 0.0116, "step": 77350 }, { "grad_norm": 0.1634691208600998, "learning_rate": 2.976582994908439e-07, "loss": 0.0092, "step": 77360 }, { "grad_norm": 0.19893445074558258, "learning_rate": 2.954106518671007e-07, "loss": 0.008, "step": 77370 }, { "grad_norm": 0.25580719113349915, "learning_rate": 2.9317149739825356e-07, "loss": 0.01, "step": 77380 }, { "grad_norm": 0.24605338275432587, "learning_rate": 2.909408364669075e-07, "loss": 0.0113, "step": 77390 }, { "grad_norm": 0.274615079164505, "learning_rate": 2.887186694542299e-07, "loss": 0.016, "step": 77400 }, { "grad_norm": 0.15084616839885712, "learning_rate": 2.865049967399225e-07, "loss": 0.012, "step": 77410 }, { "grad_norm": 0.36756736040115356, "learning_rate": 2.842998187022439e-07, "loss": 0.0113, "step": 77420 }, { "grad_norm": 0.17469422519207, "learning_rate": 2.8210313571800374e-07, "loss": 0.009, "step": 77430 }, { "grad_norm": 0.23533135652542114, "learning_rate": 2.7991494816255184e-07, "loss": 0.0096, "step": 77440 }, { "grad_norm": 0.2285144031047821, "learning_rate": 2.777352564097779e-07, "loss": 0.0098, "step": 77450 }, { "grad_norm": 0.24685941636562347, "learning_rate": 2.755640608321508e-07, "loss": 0.0096, "step": 77460 }, { "grad_norm": 0.26373183727264404, "learning_rate": 2.73401361800657e-07, "loss": 0.0116, "step": 77470 }, { "grad_norm": 0.24925512075424194, "learning_rate": 2.7124715968484537e-07, "loss": 0.0115, "step": 77480 }, { "grad_norm": 0.6195213794708252, "learning_rate": 2.691014548528104e-07, "loss": 0.0098, "step": 77490 }, { "grad_norm": 0.8367946743965149, "learning_rate": 2.669642476711864e-07, "loss": 0.0137, "step": 77500 }, { "grad_norm": 0.18876256048679352, "learning_rate": 2.648355385051815e-07, "loss": 0.0083, "step": 77510 }, { "grad_norm": 0.45961710810661316, "learning_rate": 2.627153277185157e-07, "loss": 0.0103, "step": 77520 }, { "grad_norm": 0.2748337388038635, "learning_rate": 2.606036156734881e-07, "loss": 0.0082, "step": 77530 }, { "grad_norm": 0.28015536069869995, "learning_rate": 2.5850040273092127e-07, "loss": 0.0098, "step": 77540 }, { "grad_norm": 0.2146589607000351, "learning_rate": 2.5640568925020536e-07, "loss": 0.0149, "step": 77550 }, { "grad_norm": 0.176138237118721, "learning_rate": 2.5431947558927083e-07, "loss": 0.0111, "step": 77560 }, { "grad_norm": 0.24207597970962524, "learning_rate": 2.522417621045825e-07, "loss": 0.0071, "step": 77570 }, { "grad_norm": 0.20177331566810608, "learning_rate": 2.501725491511786e-07, "loss": 0.0127, "step": 77580 }, { "grad_norm": 0.14319807291030884, "learning_rate": 2.481118370826263e-07, "loss": 0.0101, "step": 77590 }, { "grad_norm": 2.113438129425049, "learning_rate": 2.4605962625104393e-07, "loss": 0.0094, "step": 77600 }, { "grad_norm": 0.2595762610435486, "learning_rate": 2.440159170070955e-07, "loss": 0.0138, "step": 77610 }, { "grad_norm": 0.1437874436378479, "learning_rate": 2.41980709699996e-07, "loss": 0.0087, "step": 77620 }, { "grad_norm": 0.1804685741662979, "learning_rate": 2.3995400467751174e-07, "loss": 0.0113, "step": 77630 }, { "grad_norm": 0.28918927907943726, "learning_rate": 2.3793580228594902e-07, "loss": 0.0117, "step": 77640 }, { "grad_norm": 0.32489901781082153, "learning_rate": 2.3592610287015982e-07, "loss": 0.0111, "step": 77650 }, { "grad_norm": 0.13365855813026428, "learning_rate": 2.3392490677354718e-07, "loss": 0.0092, "step": 77660 }, { "grad_norm": 1.5356284379959106, "learning_rate": 2.3193221433806533e-07, "loss": 0.012, "step": 77670 }, { "grad_norm": 1.1365816593170166, "learning_rate": 2.299480259042086e-07, "loss": 0.0075, "step": 77680 }, { "grad_norm": 0.3515854775905609, "learning_rate": 2.2797234181102244e-07, "loss": 0.0109, "step": 77690 }, { "grad_norm": 0.19287541508674622, "learning_rate": 2.2600516239609238e-07, "loss": 0.0086, "step": 77700 }, { "grad_norm": 0.1329481303691864, "learning_rate": 2.240464879955606e-07, "loss": 0.0092, "step": 77710 }, { "grad_norm": 0.20354367792606354, "learning_rate": 2.2209631894410387e-07, "loss": 0.0097, "step": 77720 }, { "grad_norm": 0.2001475691795349, "learning_rate": 2.2015465557496117e-07, "loss": 0.0093, "step": 77730 }, { "grad_norm": 0.19369475543498993, "learning_rate": 2.1822149821990602e-07, "loss": 0.0133, "step": 77740 }, { "grad_norm": 0.34720614552497864, "learning_rate": 2.1629684720926303e-07, "loss": 0.014, "step": 77750 }, { "grad_norm": 0.4969378113746643, "learning_rate": 2.1438070287189693e-07, "loss": 0.0115, "step": 77760 }, { "grad_norm": 0.3494555354118347, "learning_rate": 2.1247306553523472e-07, "loss": 0.0128, "step": 77770 }, { "grad_norm": 0.3470092713832855, "learning_rate": 2.105739355252323e-07, "loss": 0.0104, "step": 77780 }, { "grad_norm": 0.10390496999025345, "learning_rate": 2.0868331316639678e-07, "loss": 0.0111, "step": 77790 }, { "grad_norm": 0.13363419473171234, "learning_rate": 2.0680119878179193e-07, "loss": 0.0084, "step": 77800 }, { "grad_norm": 0.3378552794456482, "learning_rate": 2.0492759269301054e-07, "loss": 0.0107, "step": 77810 }, { "grad_norm": 0.17651547491550446, "learning_rate": 2.0306249522021315e-07, "loss": 0.0084, "step": 77820 }, { "grad_norm": 0.11777033656835556, "learning_rate": 2.0120590668207816e-07, "loss": 0.0077, "step": 77830 }, { "grad_norm": 0.17934173345565796, "learning_rate": 1.993578273958574e-07, "loss": 0.0093, "step": 77840 }, { "grad_norm": 0.14745275676250458, "learning_rate": 1.975182576773371e-07, "loss": 0.011, "step": 77850 }, { "grad_norm": 0.29405325651168823, "learning_rate": 1.9568719784083812e-07, "loss": 0.0109, "step": 77860 }, { "grad_norm": 0.595727264881134, "learning_rate": 1.9386464819924898e-07, "loss": 0.0104, "step": 77870 }, { "grad_norm": 0.1927463710308075, "learning_rate": 1.9205060906399285e-07, "loss": 0.008, "step": 77880 }, { "grad_norm": 0.25225189328193665, "learning_rate": 1.9024508074503845e-07, "loss": 0.0107, "step": 77890 }, { "grad_norm": 0.19824647903442383, "learning_rate": 1.8844806355089452e-07, "loss": 0.0141, "step": 77900 }, { "grad_norm": 0.15451151132583618, "learning_rate": 1.8665955778863208e-07, "loss": 0.018, "step": 77910 }, { "grad_norm": 0.1473952680826187, "learning_rate": 1.848795637638512e-07, "loss": 0.0095, "step": 77920 }, { "grad_norm": 0.17752040922641754, "learning_rate": 1.831080817807085e-07, "loss": 0.013, "step": 77930 }, { "grad_norm": 0.22376742959022522, "learning_rate": 1.813451121418952e-07, "loss": 0.0171, "step": 77940 }, { "grad_norm": 1.0496619939804077, "learning_rate": 1.795906551486648e-07, "loss": 0.0067, "step": 77950 }, { "grad_norm": 0.28982219099998474, "learning_rate": 1.7784471110079414e-07, "loss": 0.0075, "step": 77960 }, { "grad_norm": 0.3033883571624756, "learning_rate": 1.7610728029662793e-07, "loss": 0.0145, "step": 77970 }, { "grad_norm": 0.34665513038635254, "learning_rate": 1.7437836303303979e-07, "loss": 0.0093, "step": 77980 }, { "grad_norm": 0.14415517449378967, "learning_rate": 1.726579596054545e-07, "loss": 0.0073, "step": 77990 }, { "grad_norm": 1.7214428186416626, "learning_rate": 1.7094607030784803e-07, "loss": 0.0117, "step": 78000 }, { "grad_norm": 0.2921573221683502, "learning_rate": 1.6924269543272532e-07, "loss": 0.0115, "step": 78010 }, { "grad_norm": 0.26979753375053406, "learning_rate": 1.6754783527115348e-07, "loss": 0.0091, "step": 78020 }, { "grad_norm": 0.11769024282693863, "learning_rate": 1.6586149011273422e-07, "loss": 0.0113, "step": 78030 }, { "grad_norm": 0.24423439800739288, "learning_rate": 1.6418366024562038e-07, "loss": 0.015, "step": 78040 }, { "grad_norm": 0.22015583515167236, "learning_rate": 1.6251434595651037e-07, "loss": 0.0095, "step": 78050 }, { "grad_norm": 0.1556798368692398, "learning_rate": 1.608535475306372e-07, "loss": 0.0114, "step": 78060 }, { "grad_norm": 0.2271939069032669, "learning_rate": 1.5920126525179048e-07, "loss": 0.0096, "step": 78070 }, { "grad_norm": 2.1213133335113525, "learning_rate": 1.5755749940229435e-07, "loss": 0.0116, "step": 78080 }, { "grad_norm": 0.30518007278442383, "learning_rate": 1.559222502630353e-07, "loss": 0.0088, "step": 78090 }, { "grad_norm": 0.1693653017282486, "learning_rate": 1.5429551811341757e-07, "loss": 0.0105, "step": 78100 }, { "grad_norm": 0.20933562517166138, "learning_rate": 1.5267730323141882e-07, "loss": 0.0137, "step": 78110 }, { "grad_norm": 0.6613638401031494, "learning_rate": 1.5106760589353454e-07, "loss": 0.0099, "step": 78120 }, { "grad_norm": 0.20109307765960693, "learning_rate": 1.4946642637483355e-07, "loss": 0.0103, "step": 78130 }, { "grad_norm": 0.28674018383026123, "learning_rate": 1.4787376494889703e-07, "loss": 0.0089, "step": 78140 }, { "grad_norm": 1.6715127229690552, "learning_rate": 1.4628962188787955e-07, "loss": 0.0137, "step": 78150 }, { "grad_norm": 0.33872121572494507, "learning_rate": 1.4471399746247006e-07, "loss": 0.0116, "step": 78160 }, { "grad_norm": 0.14103420078754425, "learning_rate": 1.4314689194188103e-07, "loss": 0.0108, "step": 78170 }, { "grad_norm": 2.213705062866211, "learning_rate": 1.4158830559390933e-07, "loss": 0.0123, "step": 78180 }, { "grad_norm": 0.12128343433141708, "learning_rate": 1.4003823868486422e-07, "loss": 0.0169, "step": 78190 }, { "grad_norm": 0.13772541284561157, "learning_rate": 1.3849669147960598e-07, "loss": 0.0093, "step": 78200 }, { "grad_norm": 0.2016805112361908, "learning_rate": 1.3696366424155726e-07, "loss": 0.01, "step": 78210 }, { "grad_norm": 0.6907439231872559, "learning_rate": 1.35439157232653e-07, "loss": 0.0167, "step": 78220 }, { "grad_norm": 0.43903928995132446, "learning_rate": 1.3392317071340144e-07, "loss": 0.0121, "step": 78230 }, { "grad_norm": 0.1768397092819214, "learning_rate": 1.3241570494283984e-07, "loss": 0.0106, "step": 78240 }, { "grad_norm": 0.40566471219062805, "learning_rate": 1.3091676017855103e-07, "loss": 0.0137, "step": 78250 }, { "grad_norm": 0.25323429703712463, "learning_rate": 1.2942633667666904e-07, "loss": 0.0107, "step": 78260 }, { "grad_norm": 0.15038786828517914, "learning_rate": 1.2794443469185679e-07, "loss": 0.0153, "step": 78270 }, { "grad_norm": 0.22282493114471436, "learning_rate": 1.2647105447734508e-07, "loss": 0.0083, "step": 78280 }, { "grad_norm": 0.20607653260231018, "learning_rate": 1.2500619628488254e-07, "loss": 0.0089, "step": 78290 }, { "grad_norm": 0.2222352921962738, "learning_rate": 1.2354986036477446e-07, "loss": 0.012, "step": 78300 }, { "grad_norm": 0.24336525797843933, "learning_rate": 1.221020469658718e-07, "loss": 0.0121, "step": 78310 }, { "grad_norm": 0.21359828114509583, "learning_rate": 1.2066275633556556e-07, "loss": 0.0106, "step": 78320 }, { "grad_norm": 0.295719712972641, "learning_rate": 1.192319887197979e-07, "loss": 0.0079, "step": 78330 }, { "grad_norm": 0.16751837730407715, "learning_rate": 1.1780974436303438e-07, "loss": 0.0083, "step": 78340 }, { "grad_norm": 0.15936891734600067, "learning_rate": 1.1639602350830836e-07, "loss": 0.0092, "step": 78350 }, { "grad_norm": 0.1367432177066803, "learning_rate": 1.1499082639718217e-07, "loss": 0.009, "step": 78360 }, { "grad_norm": 0.2486015409231186, "learning_rate": 1.1359415326976374e-07, "loss": 0.0088, "step": 78370 }, { "grad_norm": 0.24942021071910858, "learning_rate": 1.1220600436470663e-07, "loss": 0.0083, "step": 78380 }, { "grad_norm": 0.15848931670188904, "learning_rate": 1.1082637991920996e-07, "loss": 0.0088, "step": 78390 }, { "grad_norm": 0.6099053025245667, "learning_rate": 1.0945528016901851e-07, "loss": 0.0115, "step": 78400 }, { "grad_norm": 2.2513327598571777, "learning_rate": 1.0809270534840599e-07, "loss": 0.0106, "step": 78410 }, { "grad_norm": 0.23035381734371185, "learning_rate": 1.067386556902028e-07, "loss": 0.0121, "step": 78420 }, { "grad_norm": 0.13150446116924286, "learning_rate": 1.05393131425785e-07, "loss": 0.0073, "step": 78430 }, { "grad_norm": 0.21185559034347534, "learning_rate": 1.0405613278505199e-07, "loss": 0.0083, "step": 78440 }, { "grad_norm": 0.20983344316482544, "learning_rate": 1.027276599964766e-07, "loss": 0.0117, "step": 78450 }, { "grad_norm": 0.1587871015071869, "learning_rate": 1.0140771328704391e-07, "loss": 0.0075, "step": 78460 }, { "grad_norm": 0.1777205765247345, "learning_rate": 1.0009629288231237e-07, "loss": 0.0107, "step": 78470 }, { "grad_norm": 0.15035097301006317, "learning_rate": 9.879339900635276e-08, "loss": 0.0116, "step": 78480 }, { "grad_norm": 0.3199528455734253, "learning_rate": 9.74990318817981e-08, "loss": 0.0085, "step": 78490 }, { "grad_norm": 0.2264299839735031, "learning_rate": 9.621319172982701e-08, "loss": 0.0094, "step": 78500 }, { "grad_norm": 0.2530492842197418, "learning_rate": 9.493587877015264e-08, "loss": 0.0089, "step": 78510 }, { "grad_norm": 0.22558145225048065, "learning_rate": 9.366709322102263e-08, "loss": 0.0096, "step": 78520 }, { "grad_norm": 0.4629805386066437, "learning_rate": 9.240683529924688e-08, "loss": 0.0128, "step": 78530 }, { "grad_norm": 0.2262207716703415, "learning_rate": 9.115510522016979e-08, "loss": 0.008, "step": 78540 }, { "grad_norm": 0.14719226956367493, "learning_rate": 8.991190319767583e-08, "loss": 0.0066, "step": 78550 }, { "grad_norm": 0.2015247344970703, "learning_rate": 8.867722944419509e-08, "loss": 0.0078, "step": 78560 }, { "grad_norm": 0.1431422233581543, "learning_rate": 8.745108417069214e-08, "loss": 0.0096, "step": 78570 }, { "grad_norm": 0.10436831414699554, "learning_rate": 8.623346758669381e-08, "loss": 0.0084, "step": 78580 }, { "grad_norm": 0.1477469950914383, "learning_rate": 8.502437990025037e-08, "loss": 0.0081, "step": 78590 }, { "grad_norm": 0.22537094354629517, "learning_rate": 8.382382131795764e-08, "loss": 0.0087, "step": 78600 }, { "grad_norm": 0.1919427216053009, "learning_rate": 8.263179204496818e-08, "loss": 0.0123, "step": 78610 }, { "grad_norm": 0.2690526247024536, "learning_rate": 8.144829228496354e-08, "loss": 0.0102, "step": 78620 }, { "grad_norm": 0.2291366159915924, "learning_rate": 8.027332224016526e-08, "loss": 0.0095, "step": 78630 }, { "grad_norm": 0.22499877214431763, "learning_rate": 7.910688211135164e-08, "loss": 0.0115, "step": 78640 }, { "grad_norm": 0.2890464663505554, "learning_rate": 7.794897209783546e-08, "loss": 0.0113, "step": 78650 }, { "grad_norm": 0.1313127726316452, "learning_rate": 7.679959239746403e-08, "loss": 0.0108, "step": 78660 }, { "grad_norm": 0.1724458932876587, "learning_rate": 7.565874320664135e-08, "loss": 0.0119, "step": 78670 }, { "grad_norm": 0.21921156346797943, "learning_rate": 7.45264247203059e-08, "loss": 0.0108, "step": 78680 }, { "grad_norm": 0.14972050487995148, "learning_rate": 7.340263713194184e-08, "loss": 0.008, "step": 78690 }, { "grad_norm": 0.2555469274520874, "learning_rate": 7.228738063356777e-08, "loss": 0.0095, "step": 78700 }, { "grad_norm": 0.2873364984989166, "learning_rate": 7.118065541575903e-08, "loss": 0.0133, "step": 78710 }, { "grad_norm": 0.21088451147079468, "learning_rate": 7.008246166761435e-08, "loss": 0.0087, "step": 78720 }, { "grad_norm": 0.34445714950561523, "learning_rate": 6.899279957679472e-08, "loss": 0.0093, "step": 78730 }, { "grad_norm": 0.20212121307849884, "learning_rate": 6.791166932949011e-08, "loss": 0.0089, "step": 78740 }, { "grad_norm": 0.19325999915599823, "learning_rate": 6.683907111043608e-08, "loss": 0.0132, "step": 78750 }, { "grad_norm": 0.4155859053134918, "learning_rate": 6.577500510290824e-08, "loss": 0.0093, "step": 78760 }, { "grad_norm": 0.23647484183311462, "learning_rate": 6.47194714887278e-08, "loss": 0.0178, "step": 78770 }, { "grad_norm": 0.20978353917598724, "learning_rate": 6.367247044825608e-08, "loss": 0.0092, "step": 78780 }, { "grad_norm": 0.5615097880363464, "learning_rate": 6.263400216039994e-08, "loss": 0.0086, "step": 78790 }, { "grad_norm": 0.2077104151248932, "learning_rate": 6.160406680260078e-08, "loss": 0.0226, "step": 78800 }, { "grad_norm": 0.18714672327041626, "learning_rate": 6.058266455084561e-08, "loss": 0.0077, "step": 78810 }, { "grad_norm": 0.10106883943080902, "learning_rate": 5.956979557967257e-08, "loss": 0.0081, "step": 78820 }, { "grad_norm": 0.14069432020187378, "learning_rate": 5.856546006214325e-08, "loss": 0.0103, "step": 78830 }, { "grad_norm": 0.3257128596305847, "learning_rate": 5.7569658169881466e-08, "loss": 0.0105, "step": 78840 }, { "grad_norm": 0.2512139081954956, "learning_rate": 5.658239007303445e-08, "loss": 0.0112, "step": 78850 }, { "grad_norm": 0.23827946186065674, "learning_rate": 5.560365594030059e-08, "loss": 0.0088, "step": 78860 }, { "grad_norm": 0.24186857044696808, "learning_rate": 5.463345593891833e-08, "loss": 0.0099, "step": 78870 }, { "grad_norm": 0.20110468566417694, "learning_rate": 5.367179023467173e-08, "loss": 0.0094, "step": 78880 }, { "grad_norm": 0.14603190124034882, "learning_rate": 5.271865899187378e-08, "loss": 0.0072, "step": 78890 }, { "grad_norm": 0.2951541841030121, "learning_rate": 5.177406237340532e-08, "loss": 0.0163, "step": 78900 }, { "grad_norm": 0.2500631809234619, "learning_rate": 5.083800054065946e-08, "loss": 0.0097, "step": 78910 }, { "grad_norm": 0.262818843126297, "learning_rate": 4.991047365358603e-08, "loss": 0.0079, "step": 78920 }, { "grad_norm": 0.4191896915435791, "learning_rate": 4.899148187067493e-08, "loss": 0.0142, "step": 78930 }, { "grad_norm": 0.1918829083442688, "learning_rate": 4.808102534895609e-08, "loss": 0.0082, "step": 78940 }, { "grad_norm": 0.28734830021858215, "learning_rate": 4.717910424400507e-08, "loss": 0.0097, "step": 78950 }, { "grad_norm": 0.24286644160747528, "learning_rate": 4.628571870993193e-08, "loss": 0.0116, "step": 78960 }, { "grad_norm": 0.2595328092575073, "learning_rate": 4.5400868899392324e-08, "loss": 0.0124, "step": 78970 }, { "grad_norm": 0.35099586844444275, "learning_rate": 4.4524554963576436e-08, "loss": 0.0139, "step": 78980 }, { "grad_norm": 0.1419249027967453, "learning_rate": 4.36567770522367e-08, "loss": 0.0079, "step": 78990 }, { "grad_norm": 0.25086382031440735, "learning_rate": 4.279753531364339e-08, "loss": 0.0101, "step": 79000 }, { "grad_norm": 0.23032307624816895, "learning_rate": 4.1946829894617955e-08, "loss": 0.01, "step": 79010 }, { "grad_norm": 0.2971918284893036, "learning_rate": 4.11046609405219e-08, "loss": 0.0103, "step": 79020 }, { "grad_norm": 0.4775250256061554, "learning_rate": 4.027102859526233e-08, "loss": 0.0085, "step": 79030 }, { "grad_norm": 0.16561180353164673, "learning_rate": 3.9445933001280856e-08, "loss": 0.011, "step": 79040 }, { "grad_norm": 0.14478202164173126, "learning_rate": 3.8629374299564704e-08, "loss": 0.0119, "step": 79050 }, { "grad_norm": 0.15395307540893555, "learning_rate": 3.78213526296467e-08, "loss": 0.0083, "step": 79060 }, { "grad_norm": 0.18470622599124908, "learning_rate": 3.702186812958308e-08, "loss": 0.0094, "step": 79070 }, { "grad_norm": 0.1623789370059967, "learning_rate": 3.623092093599789e-08, "loss": 0.0106, "step": 79080 }, { "grad_norm": 0.3152011036872864, "learning_rate": 3.544851118403303e-08, "loss": 0.0095, "step": 79090 }, { "grad_norm": 0.6707383990287781, "learning_rate": 3.46746390073871e-08, "loss": 0.0092, "step": 79100 }, { "grad_norm": 0.20287585258483887, "learning_rate": 3.390930453829322e-08, "loss": 0.019, "step": 79110 }, { "grad_norm": 0.2539604902267456, "learning_rate": 3.3152507907519005e-08, "loss": 0.0098, "step": 79120 }, { "grad_norm": 0.1144578605890274, "learning_rate": 3.240424924438323e-08, "loss": 0.0078, "step": 79130 }, { "grad_norm": 0.13679079711437225, "learning_rate": 3.1664528676750295e-08, "loss": 0.013, "step": 79140 }, { "grad_norm": 0.1767454296350479, "learning_rate": 3.093334633100797e-08, "loss": 0.0144, "step": 79150 }, { "grad_norm": 0.3059631288051605, "learning_rate": 3.021070233210077e-08, "loss": 0.0109, "step": 79160 }, { "grad_norm": 0.22760820388793945, "learning_rate": 2.9496596803507692e-08, "loss": 0.0104, "step": 79170 }, { "grad_norm": 0.17275409400463104, "learning_rate": 2.879102986725335e-08, "loss": 0.0106, "step": 79180 }, { "grad_norm": 0.13833759725093842, "learning_rate": 2.809400164389131e-08, "loss": 0.0099, "step": 79190 }, { "grad_norm": 0.10594966262578964, "learning_rate": 2.740551225253185e-08, "loss": 0.008, "step": 79200 }, { "grad_norm": 0.25978192687034607, "learning_rate": 2.6725561810819754e-08, "loss": 0.0132, "step": 79210 }, { "grad_norm": 0.18310193717479706, "learning_rate": 2.605415043493431e-08, "loss": 0.0096, "step": 79220 }, { "grad_norm": 1.1620254516601562, "learning_rate": 2.5391278239605965e-08, "loss": 0.0105, "step": 79230 }, { "grad_norm": 0.18807744979858398, "learning_rate": 2.4736945338094118e-08, "loss": 0.012, "step": 79240 }, { "grad_norm": 0.19590012729167938, "learning_rate": 2.4091151842214887e-08, "loss": 0.01, "step": 79250 }, { "grad_norm": 0.1502210795879364, "learning_rate": 2.3453897862318884e-08, "loss": 0.0097, "step": 79260 }, { "grad_norm": 0.15816709399223328, "learning_rate": 2.2825183507285686e-08, "loss": 0.0098, "step": 79270 }, { "grad_norm": 0.1711490899324417, "learning_rate": 2.220500888455157e-08, "loss": 0.0105, "step": 79280 }, { "grad_norm": 0.14004205167293549, "learning_rate": 2.1593374100081777e-08, "loss": 0.0091, "step": 79290 }, { "grad_norm": 0.2523864209651947, "learning_rate": 2.09902792583927e-08, "loss": 0.0136, "step": 79300 }, { "grad_norm": 0.21765001118183136, "learning_rate": 2.0395724462540788e-08, "loss": 0.0091, "step": 79310 }, { "grad_norm": 0.32755962014198303, "learning_rate": 1.9809709814111453e-08, "loss": 0.0077, "step": 79320 }, { "grad_norm": 0.27149027585983276, "learning_rate": 1.923223541324126e-08, "loss": 0.0125, "step": 79330 }, { "grad_norm": 0.22126255929470062, "learning_rate": 1.8663301358606833e-08, "loss": 0.0111, "step": 79340 }, { "grad_norm": 0.26085710525512695, "learning_rate": 1.8102907747419295e-08, "loss": 0.0119, "step": 79350 }, { "grad_norm": 0.2864775061607361, "learning_rate": 1.7551054675435385e-08, "loss": 0.0092, "step": 79360 }, { "grad_norm": 0.15255266427993774, "learning_rate": 1.7007742236957447e-08, "loss": 0.0096, "step": 79370 }, { "grad_norm": 0.18849608302116394, "learning_rate": 1.647297052481678e-08, "loss": 0.014, "step": 79380 }, { "grad_norm": 0.32597070932388306, "learning_rate": 1.5946739630390285e-08, "loss": 0.0166, "step": 79390 }, { "grad_norm": 0.2303100973367691, "learning_rate": 1.5429049643606032e-08, "loss": 0.016, "step": 79400 }, { "grad_norm": 0.21946413815021515, "learning_rate": 1.4919900652909935e-08, "loss": 0.0133, "step": 79410 }, { "grad_norm": 0.2387104630470276, "learning_rate": 1.4419292745310175e-08, "loss": 0.009, "step": 79420 }, { "grad_norm": 0.19147557020187378, "learning_rate": 1.3927226006343885e-08, "loss": 0.0095, "step": 79430 }, { "grad_norm": 0.2018377184867859, "learning_rate": 1.3443700520093805e-08, "loss": 0.0124, "step": 79440 }, { "grad_norm": 0.17554731667041779, "learning_rate": 1.296871636917718e-08, "loss": 0.0089, "step": 79450 }, { "grad_norm": 0.19409747421741486, "learning_rate": 1.2502273634762419e-08, "loss": 0.009, "step": 79460 }, { "grad_norm": 0.21550525724887848, "learning_rate": 1.2044372396546876e-08, "loss": 0.0101, "step": 79470 }, { "grad_norm": 0.14935050904750824, "learning_rate": 1.1595012732773524e-08, "loss": 0.0123, "step": 79480 }, { "grad_norm": 0.219086691737175, "learning_rate": 1.1154194720225386e-08, "loss": 0.0128, "step": 79490 }, { "grad_norm": 0.2860143780708313, "learning_rate": 1.0721918434231093e-08, "loss": 0.01, "step": 79500 }, { "grad_norm": 0.19882747530937195, "learning_rate": 1.0298183948648232e-08, "loss": 0.0112, "step": 79510 }, { "grad_norm": 0.4049135446548462, "learning_rate": 9.88299133588555e-09, "loss": 0.0085, "step": 79520 }, { "grad_norm": 0.16291654109954834, "learning_rate": 9.476340666891847e-09, "loss": 0.015, "step": 79530 }, { "grad_norm": 0.20324000716209412, "learning_rate": 9.078232011139332e-09, "loss": 0.0085, "step": 79540 }, { "grad_norm": 0.4608801007270813, "learning_rate": 8.68866543666802e-09, "loss": 0.0105, "step": 79550 }, { "grad_norm": 0.15285487473011017, "learning_rate": 8.307641010035782e-09, "loss": 0.0115, "step": 79560 }, { "grad_norm": 0.33812469244003296, "learning_rate": 7.93515879635165e-09, "loss": 0.0114, "step": 79570 }, { "grad_norm": 0.19015923142433167, "learning_rate": 7.571218859264706e-09, "loss": 0.01, "step": 79580 }, { "grad_norm": 0.26551854610443115, "learning_rate": 7.215821260958544e-09, "loss": 0.0131, "step": 79590 }, { "grad_norm": 0.2068266123533249, "learning_rate": 6.868966062162363e-09, "loss": 0.0095, "step": 79600 }, { "grad_norm": 0.17006777226924896, "learning_rate": 6.530653322145419e-09, "loss": 0.009, "step": 79610 }, { "grad_norm": 0.2810971736907959, "learning_rate": 6.200883098717025e-09, "loss": 0.0109, "step": 79620 }, { "grad_norm": 0.20502440631389618, "learning_rate": 5.879655448226551e-09, "loss": 0.0147, "step": 79630 }, { "grad_norm": 0.17649675905704498, "learning_rate": 5.566970425557872e-09, "loss": 0.0072, "step": 79640 }, { "grad_norm": 0.20849068462848663, "learning_rate": 5.26282808414047e-09, "loss": 0.0084, "step": 79650 }, { "grad_norm": 0.193684384226799, "learning_rate": 4.967228475949437e-09, "loss": 0.0106, "step": 79660 }, { "grad_norm": 0.2141066938638687, "learning_rate": 4.680171651494369e-09, "loss": 0.0081, "step": 79670 }, { "grad_norm": 0.17467030882835388, "learning_rate": 4.40165765981937e-09, "loss": 0.0099, "step": 79680 }, { "grad_norm": 0.5406684875488281, "learning_rate": 4.131686548519698e-09, "loss": 0.0061, "step": 79690 }, { "grad_norm": 0.17363247275352478, "learning_rate": 3.8702583637251214e-09, "loss": 0.0099, "step": 79700 }, { "grad_norm": 2.64184308052063, "learning_rate": 3.617373150105463e-09, "loss": 0.0092, "step": 79710 }, { "grad_norm": 0.23186776041984558, "learning_rate": 3.3730309508706036e-09, "loss": 0.0096, "step": 79720 }, { "grad_norm": 0.21083341538906097, "learning_rate": 3.137231807781582e-09, "loss": 0.0099, "step": 79730 }, { "grad_norm": 0.21712611615657806, "learning_rate": 2.9099757611172894e-09, "loss": 0.0109, "step": 79740 }, { "grad_norm": 0.1391201764345169, "learning_rate": 2.6912628497133275e-09, "loss": 0.0113, "step": 79750 }, { "grad_norm": 0.2720983028411865, "learning_rate": 2.481093110945354e-09, "loss": 0.0111, "step": 79760 }, { "grad_norm": 0.21719671785831451, "learning_rate": 2.2794665807235327e-09, "loss": 0.0115, "step": 79770 }, { "grad_norm": 0.23339802026748657, "learning_rate": 2.0863832934980843e-09, "loss": 0.0089, "step": 79780 }, { "grad_norm": 0.1841348260641098, "learning_rate": 1.901843282264837e-09, "loss": 0.0124, "step": 79790 }, { "grad_norm": 0.2804833650588989, "learning_rate": 1.7258465785541245e-09, "loss": 0.0141, "step": 79800 }, { "grad_norm": 0.1855643391609192, "learning_rate": 1.5583932124418887e-09, "loss": 0.0126, "step": 79810 }, { "grad_norm": 0.2467198222875595, "learning_rate": 1.3994832125441282e-09, "loss": 0.0077, "step": 79820 }, { "grad_norm": 0.17669306695461273, "learning_rate": 1.2491166060057958e-09, "loss": 0.0095, "step": 79830 }, { "grad_norm": 0.16376341879367828, "learning_rate": 1.1072934185230032e-09, "loss": 0.0111, "step": 79840 }, { "grad_norm": 0.6848379969596863, "learning_rate": 9.740136743319195e-10, "loss": 0.0114, "step": 79850 }, { "grad_norm": 0.8731968998908997, "learning_rate": 8.492773962087696e-10, "loss": 0.0075, "step": 79860 }, { "grad_norm": 0.2160760909318924, "learning_rate": 7.330846054587337e-10, "loss": 0.0132, "step": 79870 }, { "grad_norm": 0.20126387476921082, "learning_rate": 6.254353219492526e-10, "loss": 0.0116, "step": 79880 }, { "grad_norm": 0.21134448051452637, "learning_rate": 5.263295640600685e-10, "loss": 0.0095, "step": 79890 }, { "grad_norm": 0.2318161129951477, "learning_rate": 4.357673487387359e-10, "loss": 0.0117, "step": 79900 }, { "grad_norm": 0.20760656893253326, "learning_rate": 3.537486914506616e-10, "loss": 0.0108, "step": 79910 }, { "grad_norm": 1.2568483352661133, "learning_rate": 2.8027360621241117e-10, "loss": 0.0091, "step": 79920 }, { "grad_norm": 0.21613621711730957, "learning_rate": 2.153421055806071e-10, "loss": 0.0077, "step": 79930 }, { "grad_norm": 0.40772661566734314, "learning_rate": 1.5895420064637733e-10, "loss": 0.0126, "step": 79940 }, { "grad_norm": 0.17433375120162964, "learning_rate": 1.1110990105200891e-10, "loss": 0.0096, "step": 79950 }, { "grad_norm": 0.17548315227031708, "learning_rate": 7.180921496874326e-11, "loss": 0.0123, "step": 79960 }, { "grad_norm": 0.4090609550476074, "learning_rate": 4.10521491134297e-11, "loss": 0.0092, "step": 79970 }, { "grad_norm": 0.13241204619407654, "learning_rate": 1.883870874297422e-11, "loss": 0.0083, "step": 79980 }, { "grad_norm": 0.20954859256744385, "learning_rate": 5.1688976432373584e-12, "loss": 0.0108, "step": 79990 }, { "grad_norm": 0.19629168510437012, "learning_rate": 4.271816234080461e-14, "loss": 0.0088, "step": 80000 } ], "logging_steps": 10, "max_steps": 80000, "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": true }, "attributes": {} } }, "total_flos": 0.0, "train_batch_size": 8, "trial_name": null, "trial_params": null }